<%BANNER%>

Biogenic Phosphorus in Palustrine Wetlands

Permanent Link: http://ufdc.ufl.edu/UFE0041534/00001

Material Information

Title: Biogenic Phosphorus in Palustrine Wetlands Sources and Stabilization
Physical Description: 1 online resource (319 p.)
Language: english
Creator: Cheesman, Alexander
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: 31p, nmr, nuclear, organic, palustrine, phosphorus, wetlands
Soil and Water Science -- Dissertations, Academic -- UF
Genre: Soil and Water Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Nutrient pollution, from both diffuse agricultural applications and point source contamination is a pressing concern for ecosystem integrity across the globe. Of particular concern, given the historical precedence of industrialization within the US, is that of phosphorus (P) pollution. Wetlands are both a victim and potential solution of this. Oligotrophic systems are suffering fundamental shifts in ecosystem functioning at the same time that constructed wetlands are being touted for their remediation value. The role that biological processing and sequestration plays in the P cycle of wetlands has long been recognized, yet it is only recently that analytical techniques have emerged that allow us to probe the functional nature and stability of these P forms in environmental samples. The nature of P functional groups has immediate and profound implications on the interaction and fate of P in the environment, from determining susceptibility to enzymatic and abiotic hydrolysis, to dictating long-term stabilization. Therefore, this dissertation has sought to provide an advance in our understanding of biogenic P in wetland soils by first reviewing the current science and then applying 31P nuclear magnetic resonance (NMR) spectroscopy to investigate the composition and mechanisms that determine that composition in wetland systems. Initial studies focused on the surveying of P composition in a broad range of palustrine wetland soils. This work not only showed the range of P forms found within wetland soils (e.g. phosphonates, phosphomonoesters (including inositol phosphates), phosphodiesters and long chain inorganic polyphosphates) but highlighted basic wetland properties that appear to impact P composition. Landscape position, vegetation and climate were shown to have little direct influence on P composition while biogeochemical characteristic such as; pH, organic matter content, and nutrient availability (themselves a product of wetland setting) appeared to be linked directly to the P composition of surface soils. Subsequent chapters sought to explore the mechanistic role of these biogeochemical characteristics on determining soil P composition. The trend, observed between wetlands, that soils with a higher organic matter content had a higher proportion of P found as phosphodiesters was explored by comparison of soils across a landscape continuum. By comparing P composition of soils under similar vegetation and management histories across the wetland upland transition, the mechanistic role of organic matter content in isolation was investigated. In the wetlands studied, depressional systems within an agricultural landscape north of Lake Okeechobee Florida, P composition was shown to be independent of landscape position and organic matter content. This was unexpected, but believed to be the result of the unique role organic matter plays in the P dynamics of the sandy, low P binding capacity soils of the region. This lack of distinction in the P composition associated with organic matter across a landscape transition was also seen in a study established to determine the redox sensitivity of certain P forms. In this mesocosm study, there was no substantial difference in the turnover rates of DNA and the phosphomonoester myo-Inositol hexakisphosphate (myo-IP-6) when considering their presence in a highly organic freshwater system. The role of microbial processing of soil organic matter in response to environmental conditions, specifically P availability, was determined by tracking the transformations of P forms within detrital organic matter entering a wetland system and by monitoring P composition in surficial soils across a profound nutrient gradient. In both cases it was apparent that P composition was independent of the major allochthonous inputs and represented P forms derived as a result of in-situ microbial processing of organic matter in direct response to environmental conditions. In conclusion, i use nformation derived by the study of a range of palustrine systems to develop a working model of biogenic P sources and stabilization in wetlands. This provides not only a significant advance in our understanding of P composition and cycling in wetlands but also provides insite into the biological processes associated with the P cycle of both wetland and terrestrial ecosystems.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Alexander Cheesman.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Reddy, Konda R.
Local: Co-adviser: Turner, Benjamin.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041534:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041534/00001

Material Information

Title: Biogenic Phosphorus in Palustrine Wetlands Sources and Stabilization
Physical Description: 1 online resource (319 p.)
Language: english
Creator: Cheesman, Alexander
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: 31p, nmr, nuclear, organic, palustrine, phosphorus, wetlands
Soil and Water Science -- Dissertations, Academic -- UF
Genre: Soil and Water Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Nutrient pollution, from both diffuse agricultural applications and point source contamination is a pressing concern for ecosystem integrity across the globe. Of particular concern, given the historical precedence of industrialization within the US, is that of phosphorus (P) pollution. Wetlands are both a victim and potential solution of this. Oligotrophic systems are suffering fundamental shifts in ecosystem functioning at the same time that constructed wetlands are being touted for their remediation value. The role that biological processing and sequestration plays in the P cycle of wetlands has long been recognized, yet it is only recently that analytical techniques have emerged that allow us to probe the functional nature and stability of these P forms in environmental samples. The nature of P functional groups has immediate and profound implications on the interaction and fate of P in the environment, from determining susceptibility to enzymatic and abiotic hydrolysis, to dictating long-term stabilization. Therefore, this dissertation has sought to provide an advance in our understanding of biogenic P in wetland soils by first reviewing the current science and then applying 31P nuclear magnetic resonance (NMR) spectroscopy to investigate the composition and mechanisms that determine that composition in wetland systems. Initial studies focused on the surveying of P composition in a broad range of palustrine wetland soils. This work not only showed the range of P forms found within wetland soils (e.g. phosphonates, phosphomonoesters (including inositol phosphates), phosphodiesters and long chain inorganic polyphosphates) but highlighted basic wetland properties that appear to impact P composition. Landscape position, vegetation and climate were shown to have little direct influence on P composition while biogeochemical characteristic such as; pH, organic matter content, and nutrient availability (themselves a product of wetland setting) appeared to be linked directly to the P composition of surface soils. Subsequent chapters sought to explore the mechanistic role of these biogeochemical characteristics on determining soil P composition. The trend, observed between wetlands, that soils with a higher organic matter content had a higher proportion of P found as phosphodiesters was explored by comparison of soils across a landscape continuum. By comparing P composition of soils under similar vegetation and management histories across the wetland upland transition, the mechanistic role of organic matter content in isolation was investigated. In the wetlands studied, depressional systems within an agricultural landscape north of Lake Okeechobee Florida, P composition was shown to be independent of landscape position and organic matter content. This was unexpected, but believed to be the result of the unique role organic matter plays in the P dynamics of the sandy, low P binding capacity soils of the region. This lack of distinction in the P composition associated with organic matter across a landscape transition was also seen in a study established to determine the redox sensitivity of certain P forms. In this mesocosm study, there was no substantial difference in the turnover rates of DNA and the phosphomonoester myo-Inositol hexakisphosphate (myo-IP-6) when considering their presence in a highly organic freshwater system. The role of microbial processing of soil organic matter in response to environmental conditions, specifically P availability, was determined by tracking the transformations of P forms within detrital organic matter entering a wetland system and by monitoring P composition in surficial soils across a profound nutrient gradient. In both cases it was apparent that P composition was independent of the major allochthonous inputs and represented P forms derived as a result of in-situ microbial processing of organic matter in direct response to environmental conditions. In conclusion, i use nformation derived by the study of a range of palustrine systems to develop a working model of biogenic P sources and stabilization in wetlands. This provides not only a significant advance in our understanding of P composition and cycling in wetlands but also provides insite into the biological processes associated with the P cycle of both wetland and terrestrial ecosystems.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Alexander Cheesman.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Reddy, Konda R.
Local: Co-adviser: Turner, Benjamin.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041534:00001


This item has the following downloads:


Full Text





BIOGENIC PHOSPHORUS IN PALUSTRINE WETLANDS: SOURCES AND
STABILIZATION




















By

ALEXANDER WILLIAM CHEESMAN


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2010







































2010 Alexander William Cheesman





















To those who make it possible









ACKNOWLEDGMENTS

To my advisers Dr. K. Ramesh Reddy and Dr. Benjamin Turner I would like to

extend my deepest thanks for their guidance and patience throughout. Your example

and tuition has helped mold the researcher I aspire to be. I thank the other members of

my committee, Dr. Sue Newman, Dr. Nick Comerford and Dr. Mark Brenner for

providing a sounding board for ideas, as well as a sense of purpose to my work.

I thank the many people who contributed to the collection of wetland samples

from otherwise inaccessible locations. In no particular order, I would like to thank Drs.

Tom and Lynn Saunders, Dr. Jim Sickman, Dr. Kathy Crowley, Dr. Sofie Sjogersten, Dr.

Robert Kadlec, Mr. Jason Vogel and Dr. Rebecca Sharitz. I extend my particular thanks

to Dr. Diane De Steven who was invaluable in assisting in the collection of samples

from South Carolina, as well as providing extensive ancillary data on Carolina Bay

ecosystems.

My thanks go to Dr. Patrick Inglett and Dr. Ed Dunne who provided access to

archived samples for analysis as well as discussion and ideas when considering about

the role of macrophytes and the nature of P cycling in wetlands. I am indebted to Ms. Yu

Wang, Mr. Gavin Wilson, Mrs. Tania Romero and Dr. Alexander Blumenfeld for training

and invaluable assistance during analytical lab work. As well as to Dr. Jim Rocca whose

patience and ready wit have allowed me to gain at least a passable understanding of

31P NMR analysis.

Throughout my work I have received friendship, advice and scientific discourse

from people too many to mention. To those, I say thank you.

Lastly, I would like to thank my family and my parents especially. Your love

support and example have led me to become the man, and the scientist I am today.









TABLE OF CONTENTS


page

A C K N O W LED G M ENTS ........... ................................. ......................................... 4

LIST OF FIGURES.................................. ......... 13

LIST OF ABBREVIATIONS..................... .......... .............................. 17

A B S T R A C T ...................................................................................................... 1 9

CHAPTER

1 BIOGENIC PHOSPHORUS IN WETLANDS: AN INTRODUCTION ................. 22

P hosphorus in W wetlands .............. ............................ ......... .................. .............. 22
W etland Soil.......................................... ..... 24
W etla nd P hospho rus C ycle ............................................................ ... .. ............... 26
Biogenic Phosphorus in W wetlands .......... ... ...... ............................ ............... 27
Dynam ic Biogenic Phosphorus ............ ................... ................. .. .... ............... 28
D issertation O overview .. ................................. ........................................... 30
D issertation O objectives ...................... ....... ........ .. ............................ 31
D issertation Layout.............................. ............... 31

2 BIOGENIC PHOSPHORUS IN WETLANDS: TOTAL POOL DETERMINATION
AND THE APPLICATION OF 31P NUCLEAR MAGNETIC RESONANCE
SPECTROSCOPY ...................................................... 35

Introduction ............. ... .... .. .. ..................................................... 35
Total Organic Phosphorus in Soils............................................ .................... 35
Methodological Legacy............................. ...... ......... 35
Hierarchical Classification of Methods...................................... ......... 37
In-situ analysis of organic phosphorus ................................................. 37
Ex-situ analysis of organic phosphorus.......................... ............. 38
Existing Reviews of Organic Phosphorus in Wetlands ............................... 39
Published Estimates of Organic Phosphorus in Wetland Soils ........................ 40
Total Polyphosphates in Soils ........................... ...................... ................ 43
In-situ Analysis of Polyphosphates .......................... ......... ..... .......... 43
Ex-situ Analysis of Polyphosphates....................... .... ....................... 44
Total Biogenic Phosphorus in Other Ecosystem Components........................... 45
W a te r C o lu m n ............. ......... .. .. ...... ............................. ..... ............... 4 5
Biota ................................................... ............... 47
Biogenic Phosphorus Functional Groups.......................... .. ....................... 48
Phosphom onoesters........... ................................................ ................. 49
P ho sp ho d ie ste rs ......... ......................................................... 5 0
Polyphosphates .................. ............... ........ .................. ............... 50









Phosphonates .............. ... ....... ... .. .. .............. ...... ......... 52
Application of 31P Nuclear Magnetic Resonance Spectroscopy in Wetlands.......... 52
Basic Principles ............................................................................ .............. 53
N ucle i Inte ra ctio ns ............. ............... ...... .... .... ................................. 5 5
Range of Applications in W etlands....... ................. ......... .... ........... 57
Water column........................................ .......... 58
B io ta .................. .................................. ................. 5 9
Soils and detritus ............. .............................. ............ .. .......... 59
Experimental Considerations for Application to Soils ................................... 62
C conclusions .............................................................. ................... 64

3 INTERACTION OF BIOGENIC PHOSPHORUS WITH ANION EXCHANGE
MEMBRANES: IMPLICATION FOR SOIL PHOSPHORUS ANALYSIS ................. 82

Introduction ........................... ............... 82
Methods ................. .............................. ............... 84
Anion Exchange Membranes ...................................................... 84
Phosphorus Determ nation ............. ........ ... .... .... ........ .......... ............... 85
Experim mental D design ............................ ............ ............... ...... ... ....... 86
Anion exchange membrane exchange capacity..................................... 86
Phosphorus recovery by anion exchange membrane strips ...................... 86
Purity and stability of organic and condensed phosphates in deionized
water................................ ... ... ... ..... .. ......................... 86
Extraction of phosphorus compounds from wetland soils by anion
exchange membranes ............ ....... ..... ............ ................... 87
Results ................................. ... .......... .................... 88
Anion Exchange Membrane Capacity ........................................... ................. 88
Organic and Condensed Phosphorus Recovery by Anion Exchange
M e m b ra ne S trips ......................... .. ..... .. ...................... ......... .... .......... 8 9
Purity and Stability of Phosphorus Compounds in Deionized Water ................ 89
Application to Wetland Soils for Exchangeable and Fumigation-Released
P hosphorus ...... ... ...................................................... .... ........ 90
D is c u s s io n ............. ......... .. .. ......... .. .. ......... .................................... 9 0

4 A SURVEY OF BIOGENIC PHOSPHORUS IN WETLAND SOILS: A
SOLUTION 31P NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY
S T U D Y ........... .. ......... .. .. ......... .. .. .......... ........................................ 9 9

Introduction ............................. ............... 99
M methods ............... ......................................... ........... ............... 100
Sampling ................... ........ .... ......... ....................... 100
Biogeochem ical Characterization ........ ........... .. ......... .. ............... 101
Phosphorus Composition ...... .... ............ ...................... ........ ....... 102
Data Analysis ...... .. ...................... ...... .............................. 105
Results and Discussion....................................... 106
W wetlands Sampled ................... ......... .. .. ......... .. ................ 106
Biogeochemical Characteristics of Wetlands Sampled ................................ 107


6









Solution 31P Nuclear Magnetic Resonance Spectroscopy............................. 109
Extraction of total phosphorus............... ................................ ....... 109
Phosphorus composition..... ..................... .................. 110
D is c u s s io n .............. ..... ............ ................. ........................................... 1 1 4
Conclusions ................ ......... .. ........... ................. ........... 117

5 PHOSPHORUS FORMS IN HYDROLOGICALLY ISOLATED WETLAND AND
SURROUNDING PASTURE SOILS ............................. .......... .... .............. 139

Intro d uctio n ............. ......... .. ................................. .................................. 13 9
Materials and Methods......................................... 142
Site Description ..................... ............ ................... 142
Soil Sampling ................ ......... ............... ............ ..... 143
Hydroperiod Determination...................................... 143
Soil Biogeochem ical Properties .............................. .................. 143
Phosphorus Characterization ............... ...................... .. ............ 144
Solution 31P Nuclear Magnetic Resonance Spectroscopy ............................ 146
Data Analysis ................ ......... ............... ............ ..... 147
Results and Discussion ..................................... ................... 147
Soil Biogeochemical Characteristics ........... .... ......... ...... .............. 147
Soil Phosphorus Composition .................. ....... ......... ................... 148
Solution 31P Nuclear Magnetic Resonance Spectroscopy ............................ 149
Im pact of H ydroperiod ............................. ......... ............... ............... 151
Phosphorus Storage......................................... ................ ............... 154
C o n c lu s io n s ............. ......... .. .. ......... ................................ 1 5 4

6 STABILITY OF SELECT BIOGENIC PHOSPHORUS COMPOUNDS UNDER
AEROBIC AND ANAEROBIC CONDITIONS................................... 164

Intro d u ctio n ............. ......... .. ................................. ................. ............... 16 4
M methods ................................................................................................................ 166
Microcosm Experiment ............................... ............_. ..... ............ 166
Biogenic Phosphorus Spikes ............................. ......... ... ........... 167
Biogeochemical Characterization ................. .. ............. 167
Phosphorus Com position ........................ .... .......... ................... 168
Solution 31P Nuclear Magnetic Resonance Spectroscopy........................... 168
Data Analysis ................ ......... ............... ............ ..... 169
Results and Discussion....................................... 170
Phosphorus Recovery .............................. ......... ...................... 170
Initial Biogenic Phosphorus Composition ............. ..................................... 171
Stability of Phosphorus Functional Groups................................. ........... 171
Stability of polyphosphates and DNA........................................... 172
Stability of myo-Inositol hexakisphosphate ......... ............................. 173
C o n c lu s io n s ......... .. ......................................... ............... 1 7 6

7 PHOSPHORUS TRANSFORMATIONS DURING DECOMPOSITION OF
W ETLAND MACROPHYTES ............................................ ......................... 189









In tro d u c tio n ........................................................................................................... 1 8 9
Materials and Methods......................................... 191
Site Description ..................... ............ ................... 191
Study Design ...... .......... ........... .................... 192
Litterbag Study ............. ..................................... ...... ......... 192
Analysis of Biogeochemical Properties ........ .... .............................. ....... 193
Phosphorus Com position ............ .................................................. 194
Phosphorus extraction in NaOH-EDTA ........................... .. ............. 194
Solution 31P nuclear magnetic resonance spectroscopy ......................... 195
D ata A na lysis ............. ......... .. .. ........ .... ........................................ 19 5
R e s u lts ............. ......... .. .............. .. ............................................. 1 9 6
Initial Litter Material .............................................. 196
Mass Loss ............... ............................................... ............... 197
Phosphorus C ontent......................................... ................ ............... 197
P hosphorus C om position ........................ .. ......... ............... ............... 199
D is c u s s io n ............. ......... .. .............. .. ................................................ 2 0 1
C conclusions ...... .. .. .......................... ............................204

8 PHOSPHORUS FORMS AND DYNAMICS ALONG A STRONG NUTRIENT
GRADIENT IN A TROPICAL OMBROTROPHIC WETLAND .............................. 218

Intro d uctio n .......................................... .................. ............... 2 18
M e th o d s ................................................................................................... 2 2 0
Study Site ........................................................................ ............................ 220
Sampling ...... ........................ .................. 220
Soil Properties ................ ......... ........ ......... 221
Phosphorus Characterization ............................... .................... ............... 222
Anion exchange membranes ................ ............. ......... ............ 222
Solution 31P nuclear magnetic resonance spectroscopy ....................... 222
Hydrolytic Enzyme Assay .......... ............................... 224
Results ................ ......... ................... .............. 225
Soil Biogeochemical Properties ....................................... 225
Phosphorus Biogeochem istry.............................. ........... 226
Phosphorus recovery in NaOH-EDTA............................... 227
Solution 31P NMR spectroscopy...................... ...................... 228
D iscussio n ...... .. .. ...... .... .......................................... .........................22 9

9 SUMMARY AND CONCLUSIONS....................................................... 244

Biogenic Phosphorus Composition in Wetlands (Experimental Objective 1) ........ 245
Influence of Landscape Position (Experimental Objective 2) ...... ........................ 246
Influence of Redox Conditions (Experimental Objective 3) ............... ............ 247
Influence of Phosphorus Availability (Experimental Objective 4) ...................... 247
Synthesis and Further Studies........................ ........... ..... .......... 249
Organic Matter and Redox Conditions ......... ............. .......................... 249
N utrient Status... ................ ......... ............................. ............... 251
Ite rative P processing ......... .. ......... ..................................... ............ 2 52









APPENDIX

A ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 1 ........ 257

B ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 2 ........ 261

C ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 4 ........ 276

D SIMPLIFIED METHOD FOR DETERMINATION OF TOTAL PHOSPHORUS IN
W ETLAND SAMPLES ............... ............... .... ...................... 280

LIST OF REFERENCES ........... ......... ......... ... .............. .... ........... 283

BIOGRAPHICAL SKETCH ............... ............ .... .......................... 319









LIST OF TABLES
Table page

2-1 Hierarchical classification of methods used in the study of organic
phosphorus in soils and sedim ents................ .................................. ....... 66

2-2 Soil biogeochemical characteristics and estimates of total organic
phosphorus as determined in surface (0-10 cm) soils from a range of wetland
units in South Florida (Reddy et al. 1998) ............. ................ ..................... 67

2-3 Estimates of total organic phosphorus as determined in three surface
sediment samples from a lagoon on the Po river delta, Saca di Goro Italy
(Barbanti et al. 1994) ................................................................ ... ......... 67

2-4 Relative distribution of biogenic phosphorus within living biota. ......................... 68

2-5 A selection of functional groups based upon phosphorus, and select
examples of compounds containing these groups mentioned in this
dissertation. ............. ......... ........ .......................... ......... 69

2-6 Studies employing 31P nuclear magnetic resonance spectroscopy in wetland
and aquatic systems.. ..................................... .......... 71

2-7 Methodological details of studies employing 31P nuclear magnetic resonance
spectroscopy in wetland soils ................ .......... ... ..... ............... 74

3-1 Phosphorus compounds tested on anion exchange membranes.................. 94

4-1 Wetland study sites sampled for characterization of biogenic phosphorus
com position. ............... ... .............. ........................... .......... 119

4-2 Soil biogeochemical properties of studied wetland systems .......................... 121

4-3 Phosphorus composition of surface soils as determined by solution 31P NMR
spectroscopy ........ ......... ............................ ..... .. .... ........... 123

4-4 Inorganic polyphosphates as determined by solution 31P NMR spectroscopy
of wetland soils .............. ............ ... ............... ... ................ 125

4-5 Eigen values of principal components determined on PCA applied to
phosphorus composition within wetland soils. ............. .......... ............ ... 126

4-6 Multiple linear regression models used to predict the ratio of
phosphomonoesters to phosphodiesters in wetland surface soils.................... 126

4-7 Parameter estimates for optimal model for predicting ratio of
phosphomonoetsrs to phosphodiesters........... ................... ..... ........... 126









4-8 Inositol hexakisphosphates as determined within group B wetlands. ............... 126

4-9 Correlation between microbial biomass phosphorus and phosphorus forms
determined by solution 31P NMR spectroscopy ......................................... 127

5-1 Specific integral ranges used in the classification of solution 31P NMR spectra
for lyophilized material re-suspended in 0.9 mL (1 mol L-1 NaOH 100 mmol L-
1) + 0.1 m L D20 ................ .............. .......... ..... ............. ........... 156

5-2 Soil characteristics and nutrients determined in samples across landscape
position. .............. ..... ........ ................................................... 157

5-3 Phosphorus forms across landscape position determined by acid extraction
and AEM-NMR method ........... .. ............ ...................... ... 157

5-4 Phosphorus forms determined by solution 31P NMR spectroscopy of
amalgamated alkaline extracts from across the landscape transition............... 158

5-5 Storage of total (n = 12) and organic phosphorus (n = 4) within the top 10 cm
of soil, across landscape positions. Values from each landscape position
represent averages 1 SD................................... ............... 158

6-1 Characterization of surface soil collected for spike incubation microcosm
study ................. ................................... ........................... 177

6-2 Total phosphorus, after addition of biogenic P spikes, and recovery by AEM-
NMR method of all microcosms............................................... ............... 177

6-3 Phosphorus composition of microcosms as determined by AEM-NMR
method. .............. ......... ........... ................................. 178

6-4 Phosphorus composition of soil samples as determined by parallel analysis
of lyophilized soil extracts using two distinct nuclear magnetic resonance
m ach ine s. ................................................. ........ ... .......... ....... 17 9

6-5 Phosphorus composition of microcosms spiked with myo-Inositol
hexakisphosphate with time, as determined by AEM-NMR method .............. 180

7-1 Site characteristics for enriched and unenriched study sites within WCA-2A... 205

7-2 Characterization of litter material used within the decomposition study,
consisting of two species (Cladium and Typha) from both unenriched and
enriched portions of WCA-2A. ................ ............................. 205

7-3 Four way Univariate ANOVA for mass remaining......... ... .. .................. 206









7-4 Simple exponential decay rate constant (x =100e-kt ) and leaf litter half life
calculated from material recovered over the course of 15 months within
W CA-2A .............. ......... ......... ................ ... ..... ........ 206

7-5 Four way Univariate ANOVA of phosphorus concentration in leaf litter. Model
adjusted R2 = 0.947................................................................. .... ........ 207

7-6 Linear regression of changes in mass of phosphorus within litterbags held in
the field for between 33 and 454 days ................................. 207

7-7 Phosphorus forms as determined by solution 31P NMR spectroscopy of
NaOH-EDTA extracts during macrophyte leaf litter decomposition ................. 208

7-8 Coefficients (one standard deviation) of linear increases in concentrations of
major phosphorus forms identified within leaf litter during decomposition at
the enriched study site..... ............................................... 208

8-1 Soil biogeochemical characteristics from nine sampling stations across an
om brotrophic peat dom e .......................................................... ............... .... 235

8-2 Phosphorus forms identified by anion exchange membrane technique
applied to fresh soil samples. ........................ ..... .. ............. ............... 235

8-3 Phosphorus forms identified by solution 31P NMR spectroscopy...................... 236

B4-1 Compounds and chemical shifts tested for stability during lypophilization........ 272

D-1 Standard biogenic P compounds tested for recovery by anion exchange
membrane strip.......................................... ........... 281









LIST OF FIGURES


Figure page

1-1 W etland phosphorus cycle, ......... ................. ........................... .............. .... 33

1-2 Dynamic biogenic phosphorus within wetland soils ................................... 34

2-1 Frequency histogram of estimated organic P within 117 wetlands (32
lacustrine, 85 palustrine) ........................................ .............. ........... 76

2-2 Frequency histogram of estimated total organic P within 117 wetlands broken
dow n by general m ethod grouping. .............. ................. ................. .............. 77

2-3 Structural comparison of myo-lnositol hexakisphosphate and a-d-
glucopyranose 6-phosphate (pyranose ring form of a-d-glucose 6-
phosphate). .................................................................. .... .... ........ 78

2-4 Basic phospholipid compound structure found within both eukaryotic and
prokaryotic cell m em branes ............... .......................................... ..... ........ 79

2-5 Response of phosphorus nuclei to an applied magnetic field .......................... 80

2-6 Solution 31P nuclear magnetic resonance spectra showing common
functional groups. ............ .......... .. ........ ............. ... ...... ......... 81

3-1 Exchange capacity of anion exchange membrane (AEM) strips..................... 95

3-2 Recovery of phosphorus compounds by anion exchange membrane (AEM)
s trip s ............. ......... .. .. ......... .. .. ......... ................................. 9 6

3-3 Solution 31P nuclear magnetic resonance spectra of P compounds measured
after 24 h in deionized water ...... ....... ............... ............. ............... 97

3-4 Comparison of total and molybdate-reactive P as detected in anion exchange
m em brane eluants........... ............ ......... ........................ .......... ........ 98

4-1 Solution 31P NMR spectra of surface soils collected from a Michigan
treatm ent w etland ..................................... .................. ................. 128

4-2 Average nutrient concentrations in wetland surface soils.............................. 129

4-3 Categorization of wetland sites based upon Wards hierarchical classification
of pH and organic matter (estimated by loss on ignition).............................. 130

4-4 Solution 31P NMR spectra of biogenic P composition within group A wetlands
(high organic low pH ) ............. .......... ................ ................ .......... 13 1









4-5 Solution 31P NMR spectra of biogenic P composition within group B wetlands
(low organic low pH)................ ...................... ............... 132

4-6 Solution 31P NMR spectra of biogenic P composition within group C wetlands
(high organic high pH). ............ ..... ............. ......... ........... ............... 133

4-7 Solution 31P NMR spectra of biogenic P composition within group D wetlands
(low organic high pH ) .......................... ............... ................ .......... 134

4-8 Principal component analysis of P composition within wetland soils as
determined by solution 31P NMR spectroscopy. ................... ...... ............ 135

4-9 Principal component analysis of P composition within wetland soils as
determined by solution 31P NMR analysis. ....... ...... ..... .................. 136

4-10 Region 8 to 3 ppm within group B wetland spectra and peak assignments for;
A) unidentified inositol phosphate, B) orthophosphate, C, D, E, F) myo-
Inositol hexakisphosphate, G) scyllo-lnositol hexakisphosphate. .............. .. 137

4-11 Scatter plot of microbial P against A) DNA and B) Polyphosphates. ............... 138

5-1 Location of study sites showing A) Florida outline with area of interest north
of Lake Okeechobee, and B) detail of ranch sites containing two study
wetlands each, within priority basins north of Lake Okeechobee.. ................... 159

5-2 Phosphorus pools determined by AEM-NMR method across the three
landscape positions .......... .......... ....... ...................... 160

5-3 Example solution 31P NMR spectra from amalgamated samples from the
Beaty North wetland. .............. .... ...... ....... ........... ............... 161

5-4 Phosphorus characteristics plotted against hydroperiod: A) total phosphorus
for all samples, B) total organic phosphorus determined by 31P NMR
spectroscopy, C) phosphomonoesters, and D) phosphodiesters. .................... 162

5-5 Total soil carbon plotted against phosphorus concentrations across
landscape positions from four wetland sites............. ...... .................. 163

6-1 Experimental setup for investigation of biogenic phosphorus stability under
aerobic and anaerobic conditions ....... ...................... .............. 181

6-2 Biogenic phosphorus compounds used in spiking experiment .................... 182

6-3 Exchangeable phosphorus, determined by anion exchange membranes
during microcosm study .... .. ..................................................... 183

6-4 Example solution 31P NMR spectra of soil samples spiked with biogenic
phosphorus .............. ......... ... ................ ......... ................ ......... 184









6-5 Spectral deconvolution and peak assignments in 8 to 3 ppm region of
solution 31P NM R spectra. ......................................... .. .................. 185

6-6 Detail of 8 to 3 ppm region of NMR spectra gathered on soil spiked with
biogenic phosphorus. ............. ..... ........................ .......... ............... 186

6-7 Solution 31P NMR spectra, including the region 8 to 3 ppm in detail, from
alkaline soil extracts. ............. ..... ........ ................................................ 187

6-8 Concentrations of myo-lnositol hexakisphoshate as determined within
microcosm soils under aerobic and anaerobic conditions for up to 48 days..... 188

7-1 Location of chapter 7 study sites within Water Conservation Area (WCA)-2A
in the northern Everglades. ....................... .............. ... ............... 209

7-2 Mass remaining of four litter types placed at two distinct sites within WCA-2A
and recovered at time intervals up to 454 days. ................... ...... ............ 210

7-3 Changes in litter phosphorus concentration over time at, A) enriched site
and, B) unenriched site ............. ....... ..... ................. 211

7-4 Changes in mass of phosphorus in macrophyte leaf litter held within
litterbags over the course of 454 days ................. .................................. 212

7-5 Initial phosphorus composition of Typha and Cladium leaf litter sourced from
the enriched portion of WCA-2A ...... .................. ............... 213

7-6 Initial phosphorus composition of detritus and surface soils from enriched
and unenriched study sites sampled on (10/20/03). .................. ........... .. 214

7-7 Example solution 31P NMR spectra showing changes in phosphorus
composition of Typha leaf litter during decomposition at both an unenriched
and enriched site over the course of 454 days. .............. ................... 215

7-8 Changes in proportion of major P pools found within macrophyte leaf litter
over the course of 454 days of decomposition in WCA-2A............................ 216

7-9 Conceptual model of phosphorus turnover in wetland macrophyte detritus
under, A) enriched and, B) unenriched conditions ....................... ............... 217

8-1 Overview of study transect and sampling sites with the Changuinola peat
deposit, San San Pond Sak N.W Panama. ....................................... ...... 237

8-2 Nutrient gradient; A) mass of total P, B) mass of total N, C) Molar ratio C:P,
and D) N:P from nine study sites within the Changuinola peat deposit.. .......... 238

8-3 Enzyme activity from nine study sites within the Changuinola peat deposit..... 239









8-4 Comparison of P recovered by alkaline extraction (0.25 mol L-1 NaOH and
0.05 mol L-1 EDTA) of air died soils (4 h and 16 h) or in addition to AEM
extraction of non fumigated and fumigated fresh soil. .................................... 240

8-5 Solution 31P NMR spectra showing range of P forms present in surface soils
from select sites across the study transect.............. ............ ............ ... 241

8-6 Detail of solution 31P NMR spectra from site seven soils............................... 242

8-7 Solution 31P NMR spectra of site one and nine soils after application of anion
exchange membranes with (F) and without (NF) fumigation step using
hexano l .................................................... ................... ............. 243

9-1 Comparison of conceptual model of dynamic biogenic P cycling in soil
(Figure 1-4) modified for A) systems dominated by interactions with the
mineral phase (e.g. uplands) and B) systems dominated by interactions with
organic matter (e.g. wetlands). .............. ................. ........... ............... 254

9-2 Influence of anaerobic conditions (typical of wetlands) upon both organic
matter and mineral phase stabilization of biogenic P in soils.............. ........... 255

9-3 Simplified linear progression of development of biogenic P from inputs to
w etland soil ......... .. .......... ......... .......................................................... 256

9-4 Iterative processing of P within wetland soils. Mediated by microbial
processes, biogenic P undergoes interactions with both organic matter and
m mineral phase ............................................. ............... .............. 256

B4-1 Experimental schemtaic for test of biogenic P stability during lyopilization....... 273

B4-2 Solution 31P NMR spectra in various matrix environments............................... 274

B4-3 Detail of solution 31 P NMR spectra of standard biogenic phosphorus
compounds after lyophilization A) Glucose 6 phosphate showing pH
dependant peak splitting B) Alkaline hydrolysis products of RNA..................... 275

D-1 Comparison of total P determined by TP Ash vial method and Andersen
(1976) procedure............... ........ ........... .... ........... 282











AEM

AIC

DI

DDI

D20

EBPR

EDTA

EPA

FID

FL

FT

GF-B

HDPE

ICP-OES

IP6

IQR

MDP

MRP

NEXAFS

NMR

NRCS

NUTS

P

PAO


LIST OF ABBREVIATIONS

Anion exchange membrane

Akaike information criterion

Deionized water

Distilled deionized water

Deuterium oxide

Enhanced biological P removal

Ethylenediaminetetraacetic acid

Environmental protection agency

Free induction decay

Florida

Fourier transform

Glass fiber-B grade

High density polyethylene

Inductively coupled plasma atomic emission spectroscopy

Inositol hexakisphosphate

Interquartile range

Methylenediphosphonic acid

Molybdate reactive P

Near edge X-ray absorption fine structure

Nuclear magnetic resonance

National resource conservation service

NMR utility transform software

Phosphorus

Polyphosphate accumulating organism









PAEM Exchangable P recovered by AEM method

PM Microbial P recovered by AEM method

ppm Parts per million

rpm Revolutions per minute

RSD Relative standard deviation

SD Standard deviation

SEPI Seqential extraction focused on determination of inorganic P

SEPO Sequential extraction focused on the determination of organic P

S/TEM Scanning / transmission electron microscopy

TC Total carbon

TN Total nitrogen

US United States

USDA United States department of agriculture

WCA-2A Water Conservation Area 2A

XANES X-ray adsorption near edge structure









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

BIOGENIC PHOSPHORUS IN WETLANDS: SOURCES AND STABILIZATION


By

Alexander William Cheesman

August 2010

Chair: K. Ramesh Reddy
Cochair: Benjamin L. Turner
Major: Soil and Water Science

Nutrient pollution, from both diffuse agricultural applications and point source

contamination is a pressing concern for ecosystem integrity across the globe. Of

particular concern, given the historical precedence of industrialization within the US, is

that of phosphorus (P) pollution. Wetlands are both a victim and potential solution of

this. Oligotrophic systems are suffering fundamental shifts in ecosystem functioning at

the same time that constructed wetlands are being touted for their remediation value.

The role that biological processing and sequestration plays in the P cycle of

wetlands has long been recognized, yet it is only recently that analytical techniques

have emerged that allow us to probe the functional nature and stability of these P forms

in environmental samples. The nature of P functional groups has immediate and

profound implications on the interaction and fate of P in the environment, from

determining susceptibility to enzymatic and abiotic hydrolysis, to dictating long-term

stabilization. Therefore, this dissertation has sought to provide an advance in our

understanding of biogenic P in wetland soils by first reviewing the current science and









then applying 31P nuclear magnetic resonance (NMR) spectroscopy to investigate the

composition and mechanisms that determine that composition in wetland systems.

Initial studies focused on the surveying of P composition in a broad range of

palustrine wetland soils. This work not only showed the range of P forms found within

wetland soils (e.g. phosphonates, phosphomonoesters (including inositol phosphates),

phosphodiesters and long chain inorganic polyphosphates) but highlighted basic

wetland properties that appear to impact P composition. Landscape position, vegetation

and climate were shown to have little direct influence on P composition while

biogeochemical characteristic such as; pH, organic matter content, and nutrient

availability (themselves a product of wetland setting) appeared to be linked directly to

the P composition of surface soils. Subsequent chapters sought to explore the

mechanistic role of these biogeochemical characteristics on determining soil P

composition.

The trend, observed between wetlands, that soils with a higher organic matter

content had a higher proportion of P found as phosphodiesters was explored by

comparison of soils across a landscape continuum. By comparing P composition of soils

under similar vegetation and management histories across the wetland-upland

transition, the mechanistic role of organic matter content in isolation was investigated. In

the wetlands studied, depressional systems within an agricultural landscape north of

Lake Okeechobee Florida, P composition was shown to be independent of landscape

position and organic matter content. This was unexpected, but believed to be the result

of the unique role organic matter plays in the P dynamics of the sandy, low P binding

capacity soils of the region. This lack of distinction in the P composition associated with









organic matter across a landscape transition was also seen in a study established to

determine the redox sensitivity of certain P forms. In this mesocosm study, there was no

substantial difference in the turnover rates of DNA and the phosphomonoester myo-

Inositol hexakisphosphate (myo-IP6) when considering their presence in a highly

organic freshwater system.

The role of microbial processing of soil organic matter in response to

environmental conditions, specifically P availability, was determined by tracking the

transformations of P forms within detrital organic matter entering a wetland system and

by monitoring P composition in surficial soils across a profound nutrient gradient. In both

cases it was apparent that P composition was independent of the major allochthonous

inputs and represented P forms derived as a result of in-situ microbial processing of

organic matter in direct response to environmental conditions.

In conclusion, i use information derived by the study of a range of palustrine

systems to develop a working model of biogenic P sources and stabilization in wetlands.

This provides not only a significant advance in our understanding of P composition and

cycling in wetlands but also provides insite into the biological processes associated with

the P cycle of both wetland and terrestrial ecosystems.









CHAPTER 1
BIOGENIC PHOSPHORUS IN WETLANDS: AN INTRODUCTION

Phosphorus in Wetlands

Phosphorus (P) is a multivalent nonmetal, with five potential oxidation states.

Highly reactive in the environment, its potential for polyatomic interaction has led to its

evolution as a vital component of cellular biochemistry (Brown and Kornberg 2004).

Phosphorus constitutes a significant proportion of nucleic acids, lipid membranes,

proteins and phosphorylated metabolic intermediates (Raghothama and Karthikeyan

2005), it is therefore a vital nutrient for biomass production and often limits growth in

freshwater (Reddy et al. 2005; Verhoeven et al. 2006) and, increasingly, coastal

(Sundareshwar et al. 2003; Turner et al. 2003g) wetlands and aquatic ecosystems.

Although aerial deposition of P may be important in highly weathered, isolated soil

systems (Chadwick et al. 1999), the predominant route of P input to soils, and ultimately

to biological uptake, is through in situ dissolution of the mineral phase (Walker and

Syers 1976) or external delivery via the hydrologic cycle. Wetlands typically occupy

'downstream' positions in the landscape, and therefore often receive large inputs of P

from 'upstream' sources (Richardson 1999). Anthropogenic alteration of global P cycling

has been profound and is accelerating, as a consequence of increased fertilizer

application (Carpenter et al. 1998), point source P discharges (Garcia-Pintado et al.

2007), and by indirect means, through human-mediated alteration of related

biogeochemical cycles and landscape ecological processes (Caraco 1993). Human

populations and resource consumption continue to grow and pressure on the

environment will increase (Godfray et al. 2010). As a consequence, we are likely to see

increased disruption of the P cycle. For example, Tilman et al. (2001) predicted a 2.4-









fold increase in global P fertilizer use between years 2000 and 2050. Such nutrient

redistribution will cause fundamental changes in the environment with wetlands a focal

point of this change. Global models suggest the current level of P retention within

freshwater wetlands is 3.1 Tg y-1, compared with pre-industrial levels of only 1.2 Tg y1

(Bennett et al. 2001). This increase in net P input to wetlands has caused many

wetlands, especially within agricultural watersheds, to surpass their P loading capacity,

putting them at risk of biological degradation (Hagerthey et al. 2008; Verhoeven et al.

2006). It is therefore likely that increased anthropogenic disruption of the P cycle will

cause further degradation of wetland ecological capital (Khan and Ansari 2005).

Wetlands and other aquatic ecosystems fulfill a vital role within a functioning

landscape continuum (Sheaves 2009), providing direct and indirect ecosystem services

(Costanza et al. 1997; Silvius and Giesen 1996), including retention of water and

nutrients at the landscape scale (Mitsch and Day 2006; Moreno et al. 2007; Paludan et

al. 2002; Perkins et al. 2005). Recognition of wetland value has generated initiatives

and programs designed to preserve these vital ecological functions in the landscape.

For instance, the US Army Corps of Engineers and Environmental Protection Agency

(EPA) administer a program; the Compensatory Mitigation for Losses of Aquatic

Resources (www.epa.gov/wetlandmitigation/) designed to deliver "no net loss" of

wetland ecosystems. Furthermore, 'artificial' wetland areas are being created worldwide

to impound and treat polluted waters before they are discharged to natural systems

(Babatunde et al. 2008; Day et al. 2004).

Although there is an appreciation for the roles that biological turnover and

sequestration of P play within wetland soils (Newman and Robinson 1999; Reddy et al.









2005; Wetzel 1999), there is currently little information on the forms of P found in

wetland soils (CHAPTER 2). The functional nature of P affects both its stability and

biological turnover (see Dynamic Biogenic Phosphorus) therefore, detailed information

on the forms, as well as the mechanisms that determine them, is vital if we are to

understand the consequences of human alterations of natural P cycling (Corstanje et al.

2007; Kuhn et al. 2002) and successfully model and manage wetland systems for the

future (Moustafa et al. 1999).

Wetland Soil

Wetlands represent a transitional ecotone between terrestrial and aquatic

ecosystems. There is, however, no single, accepted definition for the term 'wetland'

(Cowardin et al. 1979). Federal institutions in the USA, at the recommendation of the

Wetlands Subcommittee of the Federal Geographic Data Committee, use the Cowardin

system to define wetland and deepwater habitats (Federal Register 61, 29 July, 1996,

39465-39466). Under this scheme, five major systems are identified: marine, estuarine,

riverine, lacustrine and palustrine ecosystems. The first four contain both wetland and

deepwater ecosystems, while palustrine systems refer solely to wetlands (APPENDIX

Al). Considering the multidimensional continuum across landscapes (Euliss et al.

2004), the upper boundary of wetlands is distinguished from terrestrial systems by the

presence of water for a portion of the growing season. In the USA, wetland delineation,

for the purposes of Section 404 of the Clean Water Act (Title 33 U.S. Code, Sec. 1344),

is codified using present hydrology, vegetation and hydric soil indicators (USDA-NRCS

2010; Environmental Laboratory 1987). The distinction between wetlands and

deepwater habitats is less clear, and is sometimes open to interpretation based upon

the classification scheme used. The Cowardin system defines the lower boundary of









marine and estuarine wetlands as the low-water mark during spring tides, and the

boundary within inland waters as 2 m below the recorded low-water mark (Cowardin et

al. 1979), with the important caveat that if rooted macrophytes exist below this level, the

wetland boundary is extended to include them. The most comprehensive and globally

adopted treaty on wetland recognition and wise use, the Ramsar Convention on

Wetlands (1971), uses a broader definition of wetlands. Under the text of the

Convention (Article 1.1) wetlands are defined as:

"... areas of marsh, fen, peatland or water, whether natural or artificial,
permanent or temporary, with water that is static or flowing, fresh, brackish
or salt, including areas of marine water the depth of which at low tide does
not exceed six meters."

Signatories of the Ramsar Convention, including the USA, recognize that the term

'wetland' can also be applied to landscape features, such as lakes, which are

considered deepwater habitats in the Cowardin system (APPENDIX A2). Similarly,

some geomorphic approaches for classifying global wetlands (Semeniuk and Semeniuk

1995; Semeniuk and Semeniuk 1997) include fresh, deepwater habitats as a sub-

category of wetlands (APPENDIX A3).

Within this dissertation, I have chosen to study the dynamics of biogenic P in

palustrine wetlands as defined by Cowardin et al. (1979). In doing so, I accept the

commonly held notion of wetlands as transitional state between extremes of truly

terrestrial and aquatic ecosystems. These transition systems remain understudied

(CHAPTER 2), eschewed by pedologists and sedimentologists, despite the fact that

wetlands are characterized by an exciting interplay of processes found in upland and

deepwater environments (Deevey 1970). Throughout this dissertation I use the term

'wetland' to describe environments affected significantly by water, but not dominated









solely by its presence, yet the term 'wetland' is not restricted to a priori criteria, except

when used in conjunction with descriptive modifiers such as those of the Cowardin

system.

In 1998, the US Department of Agriculture-Natural Resources Conservation

Service (USDA-NRCS) extended the term 'soil' to include surface materials that are

permanently submerged by water, yet capable of supporting rooted plants (Soil Survey

Staff 2010). Since material substrates within wetlands show evidence for both

diagenesis and pedogenesis (Demas and Rabenhorst 1999, 2001) in this dissertation, I

have used a mechanistic interpretation of substrate materials, drawing on relevant

literature pertaining to both 'soils' and 'sediments.' If not otherwise specified by the

literature source, I use the term 'soil' to identify all benthic substrates within a wetland.

Wetland Phosphorus Cycle

Traditional consideration of the P cycle in wetland ecosystems contends that P

cycling is relatively simple compared to other nutrients e.g. nitrogen, given that P has no

significant gaseous phase or change in oxidation state (Reddy et al. 2005; Wetzel

1999). In fact, there are multiple oxidation states found within natural systems

(CHAPTER 2) including reduced inorganic phosphties (Pech et al. 2009), and highly

reduced phosphine gas (Devai et al. 1988). Which may lead to the gaseous

translocation of significant quantities of P in certain wetlands and benthic sediments

(Devai and Delaune 1995; Geng et al. 2005), although the unstable nature of such

highly reduced inorganic components (Hanrahan et al. 2005; Morton and Edwards

2005) suggests a role only under rare conditions (Gassmann and Glindemann 1993).

The transport of the more common, particulate and dissolved forms of P into and within

wetlands has been the subject of much study (Kadlec 1997; Mitchell and Baldwin 2005;









Reddy et al. 2005; Turner 2005), allowing the development of theoretical models of P

mass cycling within wetland ecosystems (Figure 1-1). In such a model, a modification of

the three features identified in all wetlands by Reddy and Patrick (1993), wetlands

consist of four essential ecosystem components: soil, water column, live biomass, and

detritus. Transfers between ecosystem components are varied (Figure 1-1), yet in

wetlands typified by water-induced anaerobosis and high rates of biomass production,

detrital organic matter plays a central role (McGill and Cole 1981). Not only does

detritus (exogenous and endogenous organic matter) act as a conduit for nutrient

cycling, it also forms an important component of accreting material. Many wetlands

accumulate detritus with surface layers typically containing greater amounts of organic

matter relative to both underlying material and adjacent terrestrial ecosystems (Axt and

Walbridge, 1999; Gathumbi et al., 2005; Pant and Reddy, 2001).

The generalized P cycle (Figure 1-1) contains biogenic forms (CHAPTER 2-

Biogenic Functional Groups), found within organisms (e.g. above-ground biomass and

microbial components of the detritus/soil), and within the extracellular environment,

stabilized by complex interactions with mineral and organic matter (Celi and Barberis

2005b), or as distinct extracellular inclusions [granules] (Diaz et al. 2008).

Biogenic Phosphorus in Wetlands

The descriptor 'biogenic' has been applied to amorphous silica found within

structures such as diatom frustules as a means of distinguishing it, alongside opal

phytoliths, from mineral silicates found within soils and sediments (Street-Perrott and

Barker 2008; Struyf and Conley 2009). Similarly, the term has been applied by those

studying sediment P as a means of distinguishing P from biological sources from P of

mineral origin (Ahlgren et al. 2006a; Liu et al. 2009). Sedimentologists have long









recognized the relationship between of Si and P elemental cycles joined through

processes of biological fixation and sedimentation (Conley et al. 1993; Schelske et al.

1986). In this dissertation, I use the term 'biogenic' to refer to P compounds synthesized

by organisms. Predominantly organic, biogenic P also includes inorganic forms such as

polyphosphate (see CHAPTER 2-Biogenic Functional Groups), which is increasingly

recognized as an important component of biologically mediated P cycling within wetland

and aquatic systems (Ahlgren et al. 2006; Hupfer et al. 2007).

Dynamic Biogenic Phosphorus

Biogenic P in wetland soils includes both intracellular P held within viable algal,

macrophyte, microbial and faunal biomass, and extracellular P held within the soils

matrix (Figure 1-2). Biogenic P may represent a high proportion (> 60%) of total soil P

(CHAPTER 2), and represents a dynamic pool transformed by both biotic processes

and abiotic factors. In subsequent chapters, I will explore in detail how biological and

abiotic processes within wetlands may influence biogenic P pool. Here, I briefly outline

the rationale that underlies the concept of dynamic biogenic P.

Although some aquatic heterotrophic bacteria can directly assimilate simple sugar

phosphates (Heath 2005) this ability appears to be rare and there is no significant

mechanism of organic P uptake equivalent to, for instance, the assimilation of organic N

seen in boreal plant/mycorrhizal associations (Nasholm et al. 1998). Therefore, for biotic

uptake of P, microbes along with plants must hydrolyze biogenic P to orthophosphate in

the extracellular or periplasmic environment (Oberson and Joner 2005; Raghothama

1999; Raghothama and Karthikeyan 2005; Schachtman et al. 1998). Active acquisition

of P by biological communities includes the exudation of phosphatase enzymes (Kuhn

et al. 2002), organic acids, solubilizing agents (Raghothama and Karthikeyan 2005),









and modification of microbial associations (Leake and Miles 1996; Richardson et al.

2009). All such processes are mediated by interactions with external factors. For

example, the presence of extracellular phosphatase, a vital component of biologically

mediated hydrolysis of phosphoesters (Turner and Haygarth 2005), is coupled to the

nutrient status of biological communities (Olander & Vitousek 2000; Wright & Reddy

2001) and has an activity modified by external abiotic factors. Adsorption of extracellular

enzymes to colloidal particles and phenolic humic substances in soils may infer greater

stability, often extending periods of activity, but it also reduces their catalytic capacity

(Boavida and Wetzel 1998; Huang et al. 2005; Quiquampoix and Mousain 2005; Wetzel

1999). The consequence of this tradeoff is currently unknown, but is likely to impact the

rates of P turnover.

In contrast, to the hydrolysis of extracellular biogenic P is its stabilization via

interaction with colloidal particles, complexation with humic substances, precipitation

with polyvalent cations, and/or physical incorporation into organic matter (Celi and

Barberis 2005b). The relative importance of various mechanisms depends upon the P

composition and biogeochemical characteristics of a wetland. Two processes that are

particular important to biogenic P stabilization in wetlands are (i) physical incorporation

of P into organic matter (typically high in wetlands relative to surrounding terrestrial

systems) and (ii) redox-sensitive stabilization of biogenic P with mineral components. It

has been suggested that the latter mechanism result in the redox sensitive stabilization

of both polyphosphates (Hupfer and Lewandowski 2008; Sannigrahi and Ingall 2005)

and the major phosphomonoester myo-lnositol hexakisphosphate (myo-IP6) (Celi and









Barberis 2007; McKelvie 2007; Suzumura and Kamatani 1995b; Turner and Newman

2005).

The functional forms and concentrations of intracellular microbial P vary due to

changing environmental conditions and species composition and (Heath 2005; Makarov

et al. 2005) (CHAPTER 2), yet extracellular biogenic P also shows a great deal of

variability. The balance between stabilized and labile extracellular biogenic P represents

an equilibrium state, the result of continuous inputs of organic detritus and exudation,

moderated by abiotic stabilization, and both abiotic and biologically mediated hydrolysis

(Figure 1-2). This dissertation utilizes solution 31P NMR spectroscopy to investigate the

dynamic nature of biogenic P in wetlands.

Dissertation Overview

Using the conceptual model of P cycling in Figure 1-1, I address the role of

biogenic P in wetlands by investigating P forms and amounts in the four major

ecosystem compartments: water column, biota, detritus, and soil. In doing so, it

becomes apparent that our present understanding of biogenic P composition in

wetlands is incomplete (CHAPTER 2). In particular, I found that although most

researchers rightfully contend that organic P, a sub-group of biogenic P, is the dominant

form in wetland soils and detritus (Newman and Robinson 1999; Reddy et al. 2005),

there is little information on the functional forms of P present (CHAPTER 2). Studies

that have identified biogenic P forms within wetland soils suggest the forms differ

fundamentally from those in terrestrial systems (Sundareshwar et al. 2009; Turner and

Newman 2005). This has implications with respect to our ability to estimate the stability

of sequestered P, understand ecological interactions, or model the effects of

anthropogenic perturbations.









Dissertation Objectives


Objective 1: Review the current knowledge of biogenic P in wetlands and identify
gaps in our current understanding.

Objective 2: Determine the influence of wetland characteristics and soil
physicochemical properties on forms of biogenic P in wetland soils.

Hypothesis: The composition of biogenic P in wetland soils varies systematically
with respect to wetland characteristics, landscape position, and/or soil
biogeochemical properties.

Objective 3: Determine how position in the landscape, i.e. from terrestrial to
wetland environments, influences biogenic P composition of soils.

Hypothesis: Landscape position impacts soil properties, which in turn influence
biogenic P composition. Specifically, higher productivity, a receiving position in
the landscape and reduced decomposition leads to increased organic matter
content within wetlands. This leads to differences in mechanisms of abiotic
stabilization leading to hydroperiod correlating with a decrease in the ratio of
phosphomonoesters to phosphodiesters (predominantly DNA).

Objective 4: Determine how anaerobic conditions impact biogenic P composition.

Hypothesis: Anaerobic conditions destabilize the phosphomonoester myo-IP6
and polyphosphates, and lead to reduced decomposition of phosphodiester DNA.

Objective 5: Assess the role of nutrient availability in determining biogenic P
composition within wetland detritus and soils.

Hypothesis: Increased P availability, due to elevated ambient conditions, will
reduce turnover of biogenic P by microbes, thereby altering P composition within
wetland materials.

Dissertation Layout

CHAPTER 1. Introduction to the basic concepts and rationale behind the
dissertation.

CHAPTER 2. A detailed review of published work on biogenic P in wetlands,
focused on amounts and proportions of organic P, methods used, and the
consideration of polyphosphates. The chapter introduces the concept of biogenic
functional groups and the application of 31P NMR spectroscopy to wetlands.
[Objective 1]

CHAPTER 3. Exploration of the methodological implications of observed
interactions between biogenic P and a commonly used anion exchange
membrane strip procedure. [Method development]









* CHAPTER 4. An exploratory survey of biogenic P forms in 28 diverse wetlands.
In addition to basic descriptions of P composition, the study establishes potential
mechanistic drivers behind observed patterns. [Objective 2]

* CHAPTER 5. A study of isolated depressional wetlands north of Lake
Okeechobee, FL, which utilizes a natural landscape transition to investigate the
impact of altered hydroperiod and concomitant organic matter content on
biogenic P composition. [Objective 3,4]

* CHAPTER 6. A microcosm-based study investigating how the differential stability
of biogenic P under aerobic and anaerobic conditions may account for observed
P composition in wetlands. [Objective 4]

* CHAPTER 7. An investigation into how nutrient status impacts the accumulation
and speciation of P during microbial processing of herbaceous organic matter.
[Objective 5]

* CHAPTER 8. A detailed study into the composition of biogenic P in an
ombrotrophic peat dome, Panama, specifically investigating how nutrients and
vegetation cover alter biogenic P composition. [Objective 5]

* CHAPTER 9. Final synthesis and conclusions drawn from the dissertation.





































Figure 1-1. Wetland phosphorus cycle, with areas of focus found in this dissertation highlighted in green. 1- Biotic uptake
2- Trophic progression 3- Senescence 4- Hydrolysis (enzymatic or abiotic) 5- Re-suspension 6-Organic matter
accretion 7- Deposition 8- Erosion 9- Pedogenesis 10- Exudation/ active transport 11- Abiotic
stabilization/adsorption 12- Destabilization/ desorption 13- Flux/ diffusion 14- Precipitation 15- Dissolution 16-
Polyphosphate accumulation 17- Polyphosphate utilization 18- Microbial mediated dissolution of mineral P 19-
Diagenesis 20- Export of biomass (migration, harvesting etc) 21- Import of biomass (migration) 22- Water
import/export


33


























v i Labile i Bige"icP O
Biogenic P Microbial
Biomass 0 U
Stabilized vi i. .) _
Biogenic P 0
-. viii ix
- --- -- I i:

Mineral component

Figure 1-2. Dynamic biogenic phosphorus within wetland soils, black arrows represent P flux. i = biological senescence
and exudation, ii = biological uptake of orthophosphate, iii = flux of orthophosphate within water column, iv =
adsorption of orthophosphate to mineral components, v = hydrolysis of biogenic P, vi = stabilization of biogenic
P with abiotic phase of soil, viii = stabilization of biogenic P to organic matter within soil, viii = formation of P
containing minerals, ix = phosphate solubilization, x = intracellular P cycling.


_rc~CC~C_ ~CCC--~- _









CHAPTER 2:
BIOGENIC PHOSPHORUS IN WETLANDS: TOTAL POOL DETERMINATION AND
THE APPLICATION OF 31P NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY

Introduction

The role of biological inputs and turnover within the P cycle has long been

recognized (Potter and Benton 1916). Yet it is only recently that analytical techniques

have developed to the point that researchers can define sources and track the

transformations of biogenic P in the environment (McKelvie 2005). Techniques such as

solution 31P NMR spectroscopy provide a powerful tool which allows researchers to

investigate the functional nature of P pools, while also providing the 'satisfactory

absolute standard' required when evaluating more routine operationally defined

measures of P pools (Kaila and Virtanen 1955; Mehta et al. 1954). This chapter

summarizes existing information on total biogenic P (organic phosphorus and

polyphosphates) within wetlands, and introduces 31P NMR as a technique used in the

study of all biogenic P functional forms.

Total Organic Phosphorus in Soils

Methodological Legacy

It is not the aim of this section to reiterate the work cataloguing the development of

methods used in the study of organic P in wetland soils and sediments, for which the

reader is directed to a number of seminal papers (Barbanti et al. 1994; Kuo and Sparks

2001; Mehta et al. 1954; Sommers et al. 1970; Turner et al. 2005). Yet it is important to

understand the range of techniques applied to the study of organic P when attempting

to reconcile variation seen between studies. Most studies characterize P by operational

criteria and attempt to attribute functionality by post hoc interpretation. With each new

generation of researchers methods are adopted, and then adapted to provide what is









perceived as a 'better' measure under a given set of parameters, be they high mineral

content, the presence of completing humic substances, or high anthropogenic loading.

Such methodological alterations may be profound, with the proposal of novel

combinations of extracts to produce new sequential procedures (see APPENDIX B1) or

subtle, from the inclusion of additional pretreatment (Schlichting and Leinweber 2002;

Turner et al. 2007b), to the use of dilute salt 'washes' to remove adsorbed P during

sequential extractions (Barbanti et al. 1994; Sommers et al. 1970). Yet such

modifications only exemplify the range of methods and the lack of any universally

applied standard (McKelvie 2005). When researchers have compared these procedures

and modifications, it has been common to regard the method that returns the highest

organic P values as the most accurate (Barbanti et al. 1994; Kaila and Virtanen 1955;

Saunders and Williams 1955). With direct ex-situ analysis methods (see Hierarchical

Classification of Methods), this comparative approach was considered valid since it was

believed that two main sources of analytical error, the hydrolysis of organic P, and its

incomplete extraction, would both lead to underestimation of the true pool. With modern

evidence demonstrating that post hoc interpretation of pools is often in error, such

comparative approaches may have been inappropriate. For example, routinely used

colorimetric techniques may lead to the overestimation of organic P in alkaline extracts

by initially 'missing' orthophosphate completed within the organic matrix (Turner et al.

2006b). Similarly, recent work has shown the HCI extracted fraction of the well

established Hedley et a1.(1982) scheme may contain substantial amounts of organic P

(He et al. 2006). Application of operationally defined procedures without verification of

their interpretation under sample-specific conditions may well produce incomplete and









erroneous determination of total organic P pools. The potential errors associated with

indirect ex-situ determination of organic P (see Hierarchical Classification of Methods)

have long been recognized, with biases shown to include the solubilization of Al and Fe

bound P by hydrogen peroxide treatments (Pearson 1940), and the alteration of the

mineral phase by high temperature oxidation procedures (Condron et al. 1990b;

Dormaar and Webster 1964; Frink 1969; Pearson 1940). Yet, their simplicity and ease

of application has meant that ignition-based estimates of organic P are still commonly

used especially within highly organic systems (Reddy et al. 1998) where error

associated with a modified mineral phase is assumed to be minimal. With no universally

accepted, or even appropriate standard method, estimates of total organic P should be

treated with caution, especially when based on upon ex-situ analysis and

uncorroborated post hoc interpretation.

Hierarchical Classification of Methods

In the determination of the P within soils and sediments, we are able to distinguish

methods at a number of levels (Table 2-1). Initially, a distinction must be made between

those techniques that allow for the non-destructive analysis of the P in-situ, i.e. within

the intact soil matrix, and those that require its extraction and analysis ex-situ.

In-situ analysis of organic phosphorus

With no, or little, pretreatment certain techniques are able to distinguish structural

and functional forms of P present in the complete soil. While techniques such as

Scanning / Transmission Electron Microscopy (S/TEM), X-Ray Absorption Near Edge

Structure (XANES) or Near Edge X-ray Absorption Fine Structure (NEXAFS)

spectroscopy may not be applicable to the determination of organic P (Harris and White

2008), recent developments within synchrotron-based X-ray spectromicroscopy may









provide a method of determining broad functional forms (Brandes et al. 2007). In

addition, solid-state 31P NMR spectroscopy allows for the determination of P chemical

bonding environments within bulk soils (Conte et al. 2008). While the later has been

applied to agricultural and calcareous marsh soils (Delgado et al. 2000; McDowell et al.

2002) as well as demineralized marine sediments (Sannigrahi and Ingall 2005)

limitations exist, given typically low spectral resolution a result of anisotropy within solids

(see Nuclei Interactions), the presence of paramagnetics, and typically high

concentrations of humic substances (Shand et al. 1999). Nevertheless, as

methodological techniques and data interpretation improves, in-situ approaches will

offer us a 'gold standard' of P analysis in wetlands as non-destructive methods able to

discern chemical and functional pools without alteration of environmental samples.

Ex-situ analysis of organic phosphorus

In comparison to the in-situ methods mentioned above, most techniques used to

estimate total organic P require some form of extraction, with inferences drawn from its

subsequent analysis. Extraction methods applied may be direct or indirect. With direct

extraction procedures, the aim is to remove and determine the nature of various organic

P fractions while minimizing their alteration. Indirect methods seek to establish the

nature of the inorganic P and use the difference between it and an established measure

of total P to determine the organic pool. Although subtle, the distinction is needed due to

the inherent bias associated with potential methodological error. Both direct and indirect

methods can be composed of single or sequential extraction steps using a range of

solutes. Methods have employed neutral salt, acidic, alkaline, organic, and chelating

solutes under ambient and reducing conditions in various combinations, and under

various experimental conditions (see Table 2-1, APPENDIX Al and A2). Analysis of the









resulting extracts themselves may also be direct or indirect. Direct methods, such as

solution 31P NMR spectroscopy (Cade-Menun and Preston 1996), chromatographic

techniques (Gilbin et al. 2000) and enzymatic hydrolysis (Bunemann 2008) seek to

determine the nature of the P directly in the extract. The more commonly applied indirect

method seeks to determine orthophosphate, or commonly used operational analogues

(e.g. molybdate reactive P), before and after oxidation of organic matter and attribute

the difference to organic bound P. Recent work has challenged the validity of

assumptions implicit in this approach, highlighting the potential for analytical errors in

determining true orthophosphate concentrations (Gerke 1992; Kowalenko and Babuin

2007; Turner et al. 2006b). Indeed even without the potential for interference in extract

solutions the commonly determined molybdate reactive P (MRP) (John 1970; Murphy

and Riley 1962) may include reactive organic P fractions (Baldwin 1998) often

overlooked when calculating 'organic' P.

Existing Reviews of Organic Phosphorus in Wetlands

With the notable exception of A.F Harrsion's (1987) review of soil organic P, I am

unaware of any concerted effort to collate information on levels of total organic P

determined within wetlands. Reliable data sets from a limited number of highly studied

wetland types do exist, including the peatlands of North Finland (Kaila 1956), the

organic wetlands of South Florida (Reddy et al. 1998) and isolated wetlands associated

with certain highly studied watersheds (Bruland and Richardson 2006; Reddy et al.

1996). Inferences drawn from this limited number of wetland types are routinely

considered 'typical' of a wetland setting (Condron et al. 2005; Newman and Robinson

1999; Reddy et al. 2005). For example, Newman and Robinson (1999) state, based on

experiences in Florida:









"The TP [total P] of aquatic and wetland ecosystems is often dominated by
OP [organic P] which can compromise >50% of sediment TP (Reddy et al.,
1998)..."

Often presumed to constitute a sizable portion of total P, the organic fraction is rarely

definitively established or reported in many wetland systems. This is in large part due to

the known variance associated with the application of available methods. For example,

Reddy et al. (1998) report differences in the organic P fraction of up to 13% of total P

when comparing direct ex-situ sequential fractionation and indirect ex-situ determination

using high temperature ashing in highly organic palustrine systems (Table 2-2).

Similarly, Barbanti et al. (1994) found differences of up to 20% of total P (14% when

adopting recommended method modifications) when comparing direct and indirect ex-

situ methods in organic matter-poor lagoon sediments (Table 2-3).

Published Estimates of Organic Phosphorus in Wetland Soils

Although the uncertainty inherent in estimates of total organic P (see above) by

necessity provokes caution when comparing between studies and sample types, it is

informative to note the range of estimates made for total organic P in wetland soils.

Published values purporting to represent total organic P in a range of natural and

impacted wetland surface soils were collated from over 30 peer-reviewed journal

articles. The effort sought to demonstrate the range of organic P within the context of

landscape and biogeochemical conditions, and the 'wetland' was considered the unit of

replication. Field replicates from a single location, sharing similar characteristics, were

averaged. Samples from a single named wetland, but which represented a significant

range of conditions, i.e. over a nutrient gradient within the highly studied Water

Conservation Area 2A (WCA-2A) (Vaithiyanathan and Richardson 1997) or after a

degree of anthropogenic impact (Cooke et al. 1990) were reported individually. Given









that many references did not tabulate data, when required data was extracted from

graphs using the graph digitizer tool Q-plot ver:1.2.4276 (Fang Jin, 2010

www.qplot.com). Where available, ancillary data were collected, including wetland name,

location, dominant system and subsystem class (Cowardin et al. 1979), depth of

sampled substrate, pH and method used to estimate organic P. Given the lack of

standardization, methods were split into three broad groupings; ignition-based (i.e.

Aspila 1976), SEPI-sequential extraction focused on the determination of inorganic P,

with organic P determined as a byproduct (i.e. Ruttenberg 1992), and SEPO sequential

extraction schemes focused on the determination of organic P (i.e. Ivanoff 1998).

Estimates of total organic P were typified by a high degree of variability. Of the 117

palustrine, lacustrine and riverine wetlands listed (APPENDIX A3) estimates ranged

from 0 to 94% of total P, with an overall average of 58% and a median of 64%.

Estimates from individual wetland systems, similarly showed a great deal of variation.

For example, values from WCA-2A, an herbaceous, highly organic system and possibly

one of the most highly studied wetlands on earth, ranged from 54 to 94% of total P

(Koch and Reddy 1992; Vaithiyanathan and Richardson 1997). Such a comparison may

be disingenuous, given the limitations of the applied methodologies recognized by the

respective authors, as well as the potential influence of anthropogenic P loading. Yet it

does clearly demonstrate the issues faced when collating information on wetlands,

historically an understudied component of the landscape. Of the 117 wetlands listed, 32

were considered lacustrine or riverine systems as defined by Cowardin et al. (1979),

with the remaining 85 identified variously as moss, forested, herbaceous, or cultivated

persistent emergent palustrine systems.









Estimates of total organic P in lacustrine/riverine systems ranged from 0 to 73%,

and within palustrine systems from between 16 and 94% (Figure 2-1). There was a

significant distinction between the two groups of wetlands (Wilcoxon Mann Whitney test

Chi2 = 39.4169 d.f. 1, p < 0.001) with average organic P contents of 37 and 67%

respectively. This difference reflects a distinction in the prevalence of organic matter

between system types, itself an effect of higher productivity and accretion rates within

low-energy, vegetated palustrine systems (Sahrawat 2004). Given the lack of complete

ancillary data, and known issues with inter-conversion of total C, organic C, and

estimates of organic matter such as loss on ignition (LOI) (Szava-Kovats 2009; Wright

et al. 2008) further analysis of the suggested relationship between organic matter and

organic P content was deemed impractical. When comparing broad analytical method

groupings (Figure 2-2), a significant difference was seen between Ignition, SEPI and

SEPO (Kruskal Wallis test Chi2 = 15.6458 d.f. 2, p < 0.0004). Yet in the context of

known methodological error (Barbanti et al. 1994; Turner et al. 2005), average

estimates based on Ignition and SEPO procedures were similar, 66 and 60%

respectively. In contrast, SEPI procedures were distinctly lower, at 46% of total P. This

is expected given the a priori decision by researchers to use SEPI methods in systems

that are believed to contain lower concentrations of organic P.

Although generalizations as to the importance of organic P in wetlands can be

made in a manner similar to Newman and Robinson (1999), it is clear that a high

degree of variation exists within published estimates of total organic P in wetland soils.

It is likely that fundamental differences in the biogeochemistry among wetland types

results in the distinct variation seen. Both methodological bias and erroneous









interpretation by researchers of operationally defined pools are known to impact

estimates, highlighting the lack of a universally appropriate or applied standard method

(McKelvie 2005; Turner et al. 2005).

Total Polyphosphates in Soils

The study of polyphosphates, the second component of biogenic P, is motivated

by the observation that rather than representing a molecular fossil, polyphosphates act

as dynamic cellular component in a wide range of metabolic processes (Kornberg et al.

1999; Kulaev and Kulakovskaya 2000). This appreciation has led to the identification of

polyphosphate-accumulating organisms (PAO) within wastewater treatment systems

(Zilles and Noguera 2002), soil (Ghonsika and Miller 1973) and freshwater sediments

(Davelaar 1993), as well as freshwater autotrophic plankton (Eixler et al. 2005) and

settling seston (Hupfer et al. 2004). The use of synthesized polyphosphates, within

detergents, inorganic fertilizers, and industrial processing has resulted in a significant

anthropogenic load to the environment (Khan and Ansari 2005; Rashchi and Finch

2000; Sundareshwar et al. 2001). Yet evidence of biogenic polyphosphates in

unimpacted lacustrine (Ahlgren et al. 2006a; Davelaar 1993; Hupfer et al. 2007), and

palustrine wetlands (Sundareshwar et al. 2009) suggests an important role in a wide

range of natural wetland systems and as component of geological P cycling (Diaz et al.

2008).

In-situ Analysis of Polyphosphates

Although extracellular polyphosphate granules have been identified from

diatomaceous sources (Diaz et al. 2008), in-situ analysis of polyphosphate is focused

on its identification and quantification within viable cells. Intracellular identification takes

advantage of polyphosphate localization within either granules, prokaryotes, or









specialized organelles, eukaryotes. Transmission electron microscopy and direct light

microscopy, in conjunction with vital stains, are routinely used to identify and isolate

potential polyphosphate inclusions and PAO (Hupfer et al. 2008; Serafim et al. 2002).

Further definitive identification can be achieved through the application of

immunohistochemistry (Werner et al. 2007), synchrotron based X-ray fluorescence

spectromicroscopy (Brandes et al. 2007; Diaz et al. 2008), or application of the

commonly used immunofluorescence stain DAPI ( 4',6-diamidino-2-phenylindole). The

stain DAPI, used for direct microbial cell visualization in soils (Turner et al. 2003c), has

been applied to the qualitative detection of polyphosphate in autotrophic plankton (Eixler

et al. 2005) and isolated microbes (Gunther et al. 2009; Hupfer et al. 2008). Recent

comparison of the emission profiles of the DAPI-polyphosphate complex with other

polymeric ions (lipids, DNA etc) has also allowed for more accurate and sensitive in vivo

determination of polyphosphate concentrations (Aschar-Sobbi et al. 2008; Klauth et al.

2006).

Ex-situ Analysis of Polyphosphates

This approach includes procedures developed to extract and identify

polyphosphate independent of the soil or PAO contained therein. Extractions have

ranged from the use of perchloric acid with polyphosphate identification using gel

filtration (Ghonsika and Miller 1973) and IR spectroscopy (Pepper et al. 1976), to

alkaline extractions with identification of polyphosphates by solution 31P NMR

spectroscopy (Ahlgren et al. 2007; Sundareshwar et al. 2001; Uhlmann et al. 1990).

Although its application to environmental samples may be inappropriate given the level

of interference, it should be noted that Ramen spectroscopy has also been applied









successfully to the study of polyphosphates in the aqueous phase (De Jager and Heyns

1998).

As more studies apply techniques such as 31P NMR spectroscopy to identify

polyphosphates in wetland and aquatic systems it has become apparent that a broad

range of systems contain them in significant concentrations. In addition to artificial

sludges and conditions of excess P, they have been identified in oligotrophic lake

sediments (Ahlgren et al. 2006a) Carolina Bay wetlands (Sundareshwar et al. 2009) and

highly organic tundra soils (Turner et al. 2004). Phosphorus NMR is providing a

powerful tool in the definitive determination of what appears to be an important and

dynamic P pool.

Total Biogenic Phosphorus in Other Ecosystem Components

Water Column

As with wetland soils, the analysis of biogenic P in the water column of wetlands

relies heavily upon operational definitions (Newman and Robinson 1999; Worsfold et al.

2005). Particulate seston can be filtered and analyzed directly for total biogenic P, in a

manner similar to biota or soils (Selig et al. 2002), while dissolved biogenic P,

operationally defined by passing through a 0.45 pm filter (Worsfold et al. 2005), can be

determined using either direct or indirect methods. Indirect methods use the parallel

analysis for orthophosphate (or surrogates such as molybdate reactive P) and total P to

calculate total biogenic P pool size by difference (McKelvie 2005; Mitchell and Baldwin

2005), while direct methods aim to determine true species of biogenic P present. For an

in depth discussion of the development of techniques the reader is directed to a number

of accomplished reviews (McKelvie 2005; Mitchell and Baldwin 2005; Worsfold et al.









2008), but a brief overview is given here of both indirect and direct methods applied to

the study of biogenic P in the water column.

Indirect methods include the determination of total P via indiscriminate atomic

analysis i.e. Inductively Coupled Plasma Optical Emission Spectroscopy, ICP-OES, or

by hydrolysis of biogenic P and detection of orthophosphate. Methods of hydrolysis

have included thermal/chemical oxidation using peroxydisulphate (Menzel and Corwin

1965), and magnesium nitrate (Cembella et al. 1986), high temperature ashing followed

by acidic hydrolysis (Monaghan and Ruttenberg 1999), and ultraviolet photo-oxidation

(Golimowski and Golimowska 1996; McKelvie et al. 1989). Given different susceptibility

of compounds to hydrolysis under different methods, inference of the nature of P pool

determined should be with caution (McKelvie 2005; Solorzano and Strickland 1968).

Specifically, while polyphosphates are hydrolyzed during acidic oxidation (Monaghan

and Ruttenberg 1999) they appear recalcitrant during photo-oxidation, at least without

the use of additional acidic oxidants (i.e. peroxydisulphate) (McKelvie 2005).

Direct determination of total dissolved biogenic P in the water column, attempts to

identify biogenic P compounds by their functionality or physical characteristics.

Techniques include the selective hydrolysis of compounds, separation due to physical

size, or reactivity, and detection via spectrometric techniques. Selective chemical

hydrolysis of non inositol organic P compounds by hypobromination (Eisenreich and

Armstrong 1977), and the specific targeting by free (Herbes et al. 1975), and

immobilized (Shan et al. 1993, 1994) enzymes have all been used to identify sub

categories of dissolved biogenic P. The physical separation of P forms by gel filtration

(Minear 1972; Steward and Tate 1971) or on the basis of interaction with ion exchange









media (Minear et al. 1988; Nanny et al. 1995) has also been used to separate and

identify P forms. With the continued development of chromatographic separation and

the advent of spectrometric techniques such as 31P NMR (see below) and mass

spectrometry (Cooper et al. 2005; Llewelyn et al. 2002) researchers are now able to

elucidate total biogenic P, and the functional nature of P forms present. This not only

provides information on what has proven to be a significant P pool (Mitchell and Baldwin

2005), but allows research into cycling of biogenic P (Ammerman and Azam 1985;

Nanny and Minear 1994a; Reitzel et al. 2009) as part of benthic flux and the aquatic

microbial loop (Azam et al. 1983).

Biota

Although total P concentrations in biota exhibit large inter- and intra-specific

variation, it is usually assumed that the vast majority of P from senescing biota, under

natural conditions, is in an organic form (Harrison 1987; Stewart and Tiessen 1987).

While such assertions may be right with respect to certain organisms, it is now known

that alongside significant orthophosphate concentrations (liberated during cell lysis),

biogenic polyphosphates can represent up to 40% of identified P within certain microbial

organisms (Table 2-4), thereby representing a potentially significant P input into wetland

soils. The use of techniques such as solution 31P NMR has allowed researchers to

determine the forms of P present within cultured microbial organisms (Bunemann et al.

2008b) and to attempt to partition biological sources of P (Makarov et al. 2005) within

terrestrial soils. Yet the full elucidation of P within biota is still problematic, with the

recalcitrant residual often eluding analysis. What is certain is that biogenic P inputs to

soils represent a variety of both organic and inorganic forms dependent upon the nature

and structure of the biotic community (Condron et al. 2005; Oberson and Joner 2005)









as well as the interaction of that community with its abiotic environment (Bakken 1997;

Oberson and Joner 2005).

Biogenic Phosphorus Functional Groups

As discussed previously, biogenic P represents both organic P and

polyphosphates. Compounds can be further classified by their P chemical functionality

(Table 2-5). This classification has utility over operational criteria of P pools since it

provides information on sources, potential chemical interactions, the susceptibility to

enzymatic hydrolysis, as well as the expected products of such a hydrolysis. Yet

reliance upon it must be with caution, given the risk of trivializing the impact of the

organic moiety or the potential for multiple functional groups. For example, both myo-

Inositol hexakisphosphate (myo-IP6) and its metabolic precursor glucose 6 phosphate

(Raboy 2007) are phosphomonoesters. While glucose 6-phosphate represents a simple

sugar phosphate, in which the hydrolysis of a phosphoester linkage would release both

phosphate and an easily metabolized carbon source, myo-IP6 represents a stable

inositol carbon ring, having undergone substitution by six phosphate groups (Figure 2-

3). The resulting compound has a high pH-dependent charge density, making it likely to

interact with both mineral and humic substances within the soil matrix (Celi and Barberis

2005a, 2007). This leads to a degree of recalcitrance, which is often invoked to explain

the predominance of various IP6 isomers within the identified P compounds of upland

soils (Cade-Menun 2005a; Harrison 1987; Turner et al. 2002b).

The determination of coarse functional groupings, even without further elucidation

of specific compounds, provides information on potential stability and bioavailability

within the environment. Listed below are the major biogenic P forms encountered within









the environment, alongside discussion of specific examples of relevance to subsequent

chapters.

Phosphomonoesters

Phosphomonoesters represent organic compounds linked by an ester bond to a

phosphate group (Table 2-5). The use of phosphorylation as a mechanism of energy

transfer, and signal transduction within cellular chemistry leads to a vast array of

biological molecules containing the phosphomonoester functional group (Raghothama

and Karthikeyan 2005; Turner et al. 2003d). This can be exemplified in the known roles

that phosphorylated inositol derivatives play within cells (Michell 2008). Lower order

inositol phosphates appear to be a ubiquitous component of eukaryotic cells, and

presumed significant components of biogenic P entering wetland systems through plant

and animal detritus (Weimer and Armstrong 1979). At the same time, certain

stereoisomers of the higher order phosphorylated inositols are found only at low

concentrations, or within certain specialized biological elements such as seeds (Raboy

2007; Turner et al. 2002b). As such, they may only represent a small proportion of total

P in the standing biomass. Yet these higher-order forms (i.e. myo-IP6) have been found

to represent a significant portion of identified organic P within terrestrial soils (Cosgrove

1966; Murphy et al. 2009; Turner et al. 2003f). This discrepancy has been attributed to

their preferential stabilization (Celi and Barberis 2007), but the presence of significant

concentrations of otherwise rare stereoisomers (i.e. scyllo and neo-IP6) suggests

additional sources and processing of inositol derivatives within as yet unidentified

microbial components of the soil (Turner 2007; Turner and Richardson 2004).









Phosphodiesters

The phosphodiester bond (Table 2-5) is found within a range of biological

molecules, potentially representing a significant proportion of biotic total P due its role in

a number of critical polymeric molecules and cellular membranes. Polymeric

nucleotides such as RNA and DNA constitute a sizeable proportion of total cellular P

(Table 2-4) and during cell senescence become a component of the extracellular

environment, stabilized by complex interactions with both mineral and organic

components (Niemeyer and Gessler 2002; Ogram et al. 1988). Specific to certain

prokaryotes, polymeric teichoic acids have been inferred as present within terrestrial

soils (Makarov et al. 2002a, b) and marine sediments (Ahlgren et al. 2006b), although

caution is required with the potential misassignment of phospholipids as teichoic acids

(Condron et al. 1990a; Makarov et al. 2002b). Consisting of a backbone of polymerized

glycerol and ribitol phosphate, they, along with various functional side chains, provide

the major constituent for cell wall processes within Gram-positive bacteria (Swoboda et

al. 2010), as well as acting as a dynamic P reserve (Grant 1979). In addition, P-

containing lipids phospholipidss) represent a major constituent of all biotic cell

membranes. The biosynthetic precursor to all glycolipids, phosphatidate, is a

phosphomonoester, yet the majority of glycolipids as well as the other major

phospholipid group, sphingomyelins (Figure 2-4) show additional esterification and are

therefore classed as phosphodiesters.

Polyphosphates

Polyphosphates are molecules containing multiple phosphate residues bound by

high energy acid anhydride bonds (Harold 1966). Adenosine triphosphate (ATP), a vital

component of biological energy transfer and found within all biological active soils









(Eiland 1983; Wen et al. 2005), can be classed as a polyphosphate (as well as a

phosphomonoester). To avoid confusion, references within this dissertation to

'polyphosphates' will be solely to inorganic compounds ranging in chain length from the

simple two-residue pyrophosphate to linear macromolecules of many hundreds of

phosphate residues (Table 2-5). Polyphosphates are found ubiquitously in both

eukaryotic and prokaryotic cells (Kornberg et al. 1999). Potentially a prebiotic

macromolecule (Brown and Kornberg 2004), they are now implicated in a range of

biochemical functions from phosphate and energy storage to providing biochemical

adaptation to extreme environments (Kornberg 1995; Seufferheld et al. 2008). The

biological accumulation of significant concentrations of polyphosphates was first

identified by the isolation of metachromatic granules in yeast cells (Liebrmann 1890 In

Kornberg et al. 1999). Subsequently the identification and isolation of so-called

polyphosphate-accumulating organisms (PAO) has been studied as part of enhanced

biological P removal (EBPR) within wastewater treatment facilities (Zilles & Noguera

2002) as well as terrestrial and aquatic environments in which there was a surplus of

phosphate (Gachter and Meyer 1993). The imporatnce of PAO in both biotic and abiotic

mediated P flux in lacustrine sediments has been clearly demonstrated (Gachter and

Meyer 1993; Hupfer et al. 2007; Hupfer et al. 2004; Sannigrahi and Ingall 2005), but

although noted as present in palustrine systems (Bedrock et al. 1994; Sundareshwar et

al. 2009) their role has yet to be established. Given a known presence in fungal

biomass (Koukol et al. 2008) and a growing recognition of the role that fungal

decomposition plays in wetland systems (Joergensen and Wichern 2008) it is









conceivable that polyphosphates represent a dynamic and highly important P pool

within all wetland systems.

Phosphonates

Phosphonates contain a direct C-P covalent linkage and as such are inherently

stable to both biotic and abiotic hydrolysis. First isolated in rumen protozoa (Horiguchi

and Kandatsu 1959), phosphonates have now been described in a range of biota

(Ternan et al. 1998), yet their biological role is still poorly understood. One compound

found in relatively high concentrations and in a range of organisms is 2-

aminoethylphosphonate, commonly found as a phosphonolipid wherein it acts in place

of its analogue ethanolamine phosphate (Figure 1-4) (Ternan et al. 1998). Given their

resilience, phosphonates have been employed as metabolic disrupters by both

microorganisms, as toxic secondary metabolites, and humans in such products as the

widely adopted herbicide glyphosate (N-(phosphonomethyl) glycine). The potential for

such xenobiotics to persist and cause disruption in the environment has spurred

research into potential degradation pathways (Ermakova et al. 2008) highlighting the

potential of certain soil bacteria to utilize phosphonates as a sole P source. This work,

as well as recent research suggesting phosphonates as an important and highly active

component of dissolved organic P in the marine water column (Martinez et al. 2010),

suggests phosphonates may play an active role in P cycling of many natural systems.

Application of 31P Nuclear Magnetic Resonance Spectroscopy in Wetlands

Nuclear magnetic resonance spectroscopy provides researchers with a powerful

analytical tool to elucidate the biogenic P functional groups found within wetlands.

Although its application requires consideration of a number of factors, it can provide a

non-destructive tool with the potential to determine the bonding environment of every P









atom within an environmental sample (Cade-Menun 2005b). This dissertation does not

seek to give a comprehensive overview of the theoretical background or application to

environmental samples, for which the reader is directed to a number of comprehensive

text books (Berger et al. 1997; Canet 1996) and seminal papers (Cade-Menun 2005a,

b; Condron et al. 1997; Knicker and Nanny 1997). Instead, it provides a basic

background to the theory required when considering the development of techniques and

their application to wetland ecosystems.

Basic Principles

All nuclei spin on their axis and can be considered a magnetic dipole containing a

magnetic moment, i expressed as Equation 2-1 where h = [ h (Planck's constant) / 27 ],

y is the gyromagnetic ratio (a fundamental nuclear constant) and I is the vector

representation of the nuclear spin I.

i = yhl (2-1)

When nuclei with a I ; 0 are placed in a static magnetic field Bo the magnetic

moment i aligns, and the nucleus processes around the axis of the applied field with a

Lamor frequency ( coo) as given by Equation 2-2.

coo = -yBo (2-2)

Quantum mechanics states that an object with aforementioned spin has a discrete

number of spin states and energy levels described by the magnetic spin quantum

number mi. It follows that 21 + 1 different spin energy levels are possible, each with an

energy level as described in Equation 2-3.

E = -yhmiBo (2-3)

Therefore nuclei with I = 1/ (H, 13C, 15N, 170, 27Al, 31P, etc.) when placed in a

magnetic field Bo have two distinct energy states (Figure 2-5). The population









distribution of spins in the two states under standard conditions is given by the

Boltzmann distribution law, and the energy difference between the two states, is equal

to Equation 2-4 (Figure 2-5 A)

AE = -yhAmiBo (2-4)

Using Bohr's frequency law it follows that the nuclear magnetic resonance

condition is as in Equation 2-5. Given quantum mechanical rules state that the only

transitions allowed are those in which Am = 1. We have a situation that for a spin

transition between energy levels in a magnetic field Bo an energy quantum of oo is

required.

o, = -yBoAmi (2-5)

During NMR spectroscopy experiments, this transitional energy is supplied by a

secondary oscillating radio frequency pulse field B1 perpendicular to Bo. If the frequency

of the pulse satisfies the resonance condition of Equation 2-5 the nuclei absorb the

energy and the spins move to their higher energy, less stable state (Figure 2-5 B). The

most commonly applied Fourier Transformed (FT)-NMR uses an intense B1 pulse of

fixed frequency to excite all nuclei within a sample. After the radio frequency pulse, the

emitted oscillating current of excited nuclei returning to their equilibrium state is

recorded as a free induction decay (FID). Given the often low concentrations of 31P,

sequential FID's are collected and the sum subjected to a Fourier Transformation,

transforming the time domain information to a frequency domain spectrum.

After the transitional energy is removed, and the spin system is relaxing to its

thermal equilibrium distribution the dissipated energy is released through two

processes, spin-spin relaxation and spin-lattice relaxation. Spin-spin is a randomized









entropic process governed by the spin-spin relaxation time constant T2, whereas during

spin-lattice relaxation the energy is released to the surrounding matrix, governed by the

spin-lattice constant T1. The process of relaxation is important since the sample must

return to its ground state to avoid saturation of the sample and achieve quantitative

resonances during signal acquisition. The presence of relaxation agents either as

natural paramagnetic materials with unpaired electrons (i.e. iron and manganese), or

artificial transition metal complexes (e.g. chromium acetylacetonate) provide an efficient

relaxation pathway, greatly reducing T1 (Cade-Menun et al. 2002; McDowell et al. 2006;

Nanny and Minear 1994a, b) and thereby allowing for rapid repetition of pulse

experiments and cumulative FID acquisition. Ensuring sample nuclei have returned to

their ground state can also be achieved through the use of a reduced B1 radio frequency

pulse. Ideally, a calibrated 90 pulse is applied to maximize the tip angle (the angle

between magnetic moment of perturbed nucleus and Bo). The use of a reduced pulse

length reduces the tip angle proportionally. While reducing the levels of energy returned

on a single iteration, the nuclei achieve their ground state disproportionately faster.

When considering experiments of many thousands of scans this faster recovery time

leads to shorter total experiment times.

Nuclei Interactions

Nuclei in a macroscopic sample, held within an applied field Bo actually experience

a specific local magnetic field Bloc dependent upon a nucleus's interactions with its

environment. These interactions alter a nuclei's perceived magnetic field and therefore

modify its transition energy, coo. The change in the frequency domain spectrum can then

be interpreted to provide information on both the chemical bonding and physicochemical

nature found within the sample. Potential interactions include; chemical shift, spin-spin









(scalar), dipole-dipole and quadrupole. The most readily interpreted and most pertinent

to 31P NMR is that of chemical shift interactions. Chemically nonequivalent nuclei

experience different degrees of electron shielding, this leads to the nuclei processing

with different frequencies when in a fixed field. This results in an ability to separate

nuclei on the basis of the atoms to which they are bound. Subsequently, the comparison

of resulting spectrum with known compounds allows for the identification of both distinct

functional groups (Figure 2-6) and specific P containing compounds (Turner et al.

2003f). It should be noted that, given the impact of deprotonation on altering the

chemical shift interactions, many resonances peaks are pH-dependent (McDowell and

Stewart 2005b) and comparison with spectral libraries should be carried out at a

standard pH. In addition, given Lamor frequencies are dependent upon the size of the

magnetic field being applied and magnet size is continuously evolving, resonance

frequencies are routinely reported in relation to an arbitrary standard. In solution 31P

NMR this standard is usually 85% H3P04 set as 0, with chemical shifts (6) reported as

shift in frequency of the samples vs from this reference vrf (Equation 2-6), normally in the

units of parts per million (ppm) (Wilson 1987). Therefore, when comparing spectral

libraries to identify P compounds, both the pH of the matrix and referencing applied

should be noted.

= [(vs Vf)/ Vf] X 106 (2-6)

Spin-spin coupling is a phenomenon brought about by the interaction of the

target nuclei, in our case 31P, with bonded nuclei that also have a half integer nuclear

spin (i.e. the ubiquitous 1H). Although useful in garnering secondary and tertiary

structural information in advanced analytical chemistry i.e. identifying structural









stereoisomers of higher order inositol polyphosphates (Murthy 2007) the complex signal

patterns that result from 31P 1H coupling are often hard to interpret within

environmental samples, therefore, broadband heteronuclear (proton) decoupling is often

applied. During decoupling, a secondary radio frequency pulse is applied to saturate the

1H nuclei. The rapid interconvertion of 1H spin states results in a sample average

thereby negating the 1H influence on the primary nuclei, 31P.

In any molecule, electron distribution is anisotropic. In solutions, rapid molecular

movement averages these distribution differences and the effective magnetic field

experienced by a nucleus is as discussed. In solids, or very viscous liquids, molecular

movement is slowed, leading to nuclei experiencing different localized fields and

therefore a range of Lamor frequencies leading to a broadening of the signal, and poor

spectral resolution (Cade-Menun 2005b). To avoid line broadening most environmental

studies apply an initial extract to isolate desirable components for identification and use

31P NMR in the solution phase. Within wetlands, solid-state 31P NMR has been applied

in the analysis of calcareous marshes (Delgado et al. 2000), peat soils (Shand et al.

1999) and de-mineralized marine sediments (Sannigrahi and Ingall 2005). Yet

methodological issues that exist with achieving high spectral resolution (Conte et al.

2008) have so far limited its widespread application. It should also be noted that for

similar reasons the concentration of solutes during solution 31P NMR spectroscopy will

impact line broadening (see Experimental Considerations for Application to Soils).

Range of Applications in Wetlands

As of Aug 2010, over 50 papers have been published utilizing 31P NMR in the

study of biogenic P within wetland and aquatic systems (Table 2-6). This tally draws

from a broad range of research topics and wetland types. Predominantly focused upon









freshwater systems, the review does include studies of biogenic P within select marine

systems (for a complete review see Sannigrahi et al. (2006)) that are considered

informative to the study of P in lacustrine and estuarine wetlands (Benitez-Nelson et al.

2004; Paytan et al. 2003). Research has been categorized by its application to distinct

ecosystem components, water column, biota and soil, with soil further divided into

methodological papers, and those applying 'standard' procedures to various wetland

systems.

Water column

Application of 31P NMR spectroscopy to the analysis of dissolved and particulate P

within the water column is problematic, given typically low concentrations. Researchers

have applied a range of techniques to increase P abundance prior to analysis, from ultra

filtration (Sannigrahi and Ingall 2005), and reverse osmosis (Nanny and Minear 1997),

to lypholization (Cade-Menun et al. 2006) and more recently precipitation with poly-

aluminum chloride (Reitzel et al. 2009). Not only have forms within particulates, often

including phytoplankton (see Biota) been elucidated, but research has revealed

evidence for significant dissolved biogenic P, including organic phosphomonoesters,

phosphodiesters and inorganic pyrophosphate (Cade-Menun et al. 2006; Reitzel et al.

2009). In addition, dissolved biogenic P has been shown to constitute distinctly different

forms from those found within particulates (Sannigrahi et al. 2006). Taken in conjunction

with evidence for the importance of dissolved organic P in the microbial loop of aquatic

systems (Ammerman and Azam 1985; Cotner and Wetzel 1991) it is clear that 31P NMR

offers a powerful tool in the study of what may represent an ecologically important, and

highly active P pool.









Biota


A limited number of studies have applied 31P NMR spectroscopy directly to the

study of wetland organisms, the most common being that of free-living plankton

community and polyphosphate accumulating bacteria within settling seston and benthic

sediments (Hupfer et al. 2004; Khoshmanesh et al. 2002; Reitzel et al. 2007; Reitzel et

al. 2006a). Yet, it has also been applied to the profiling of phospholipids in sediment

bacteria (Bardygulanonn et al. 1995; Watts et al. 2002) and the freshwater sponge

Eunapius fragilis Laidy (Early et al. 1996). Unfortunately, there has been only limited

characterization of P composition in more substantial flora and fauna, a limitation also

seen in terrestrial systems. Of the few completed studies in wetlands, 31P NMR

spectroscopy has been applied to the determination of biogenic P within macrophyte

leachate released upon rewetting (Pant and Reddy 2001) and as a means of tracking in

vivo acidification during cell elongation of Potamogeton pectinatus (Summers et al.

2000).

Soils and detritus

First applied to terrestrial soils over 30 years ago (Newman and Tate 1980), the

study of biogenic P in wetland soils has been greatly enhanced by the application of 31P

NMR spectroscopy. That said, it has only been applied to a limited range of wetland

systems, with published work typified by a focus on lacustrine sediments in Europe

(Ahlgren et al. 2005; Hupfer and Gachter 1995; Reitzel et al. 2007), and China (Bai et

al. 2009; Liu et al. 2009; Zhang et al. 2009b), as well as the highly organic palustrine

systems of south Florida (Robinson et al. 1998; Turner and Newman 2005; Turner et al.

2006a). Additional wetlands investigated to some degree have included 'Carolina Bays'

(Sundareshwar et al. 2009), Australian 'Billabongs' (Baldwin 1996), New Zealand









streams (McDowell 2009), and Scottish Blanket Bogs (Bedrock et al. 1994), yet it is

clear only a limited range of wetland types have been investigated via 31P NMR

spectroscopy.

Extrapolating from those systems to which 31P NMR has been applied, biogenic P

in wetlands shows a number of distinctions from the more highly studied terrestrial

systems (see below). This dissertation aims to explore these differences, to investigate

whether trends seen to date are indicative of all wetlands, as well as to investigate

mechanistic process that may explain the observed biogenic P composition.

Quantity. As suggested by previous operational estimates of total organic P,

wetlands generally have a larger proportion of total P in organic forms, as compared to

terrestrial systems (Turner and Newman 2005). In addition, biogenic polyphosphates

have been found to represent significant levels of total P, up to 12%, in certain

lacustrine (Ahlgren et al. 2006a; Hupfer and Gachter 1995) and palustrine wetlands

(Sundareshwar et al. 2009).

Prevalence of phosphodiesters. Studies in a range of upland soils have shown

biogenic P to be dominated by phosphomonoesters (Chapuis-Lardy et al. 2001;

McDowell and Stewart 2006; Turner et al. 2003a; Turner et al. 2003e), with ratios of

phosphomonoesters to phosphodiesters averaging 21.8 under various land

management practices in New Zealand (McDowell and Stewart 2006) and 10.4 in

temperate pasture soils (Turner et al. 2003e). This is in contrast to studies in wetland

soils which have shown a greater prevalence of phosphodiesters. Including Turner and

Newman (2005) which found ratios averaging just 0.82 in the highly organic palustrine

wetlands of south Florida and Zhang et al. (2009b) which found an average of 3.7 when









comparing lacustrine sediments from 7 lakes across a range of trophic-states. Studies

within terrestrial systems have attributed a positive correlation between the proportion of

P found as phosphodiesters and annual precipitation to their increased recalcitrance

under 'wetter' conditions (Condron et al. 1990a; Sumann et al. 1998; Tate and Newman

1982). It could be that similar mechanisms, along with a generally higher microbial

biomass, and the reduced stability of redox-sensitive phosphomonoester complexes,

could lead to the observed pattern in wetlands (CHAPTERS 4 and 6).

Presence of Inositol hexakisphosphate. A major component of biogenic P in

terrestrial systems (Turner et al. 2002b), the potential recalcitrance of higher order

inositol phosphates appears to be dependent upon both the mineralogy and

physicochemical nature of studied wetlands. There is evidence to suggest that under

the anaerobic conditions, prevalent in wetlands, IP6 undergoes rapid degradation

(Suzumura and Kamatani 1995a, b). Such a turnover of significant terrigenous inputs of

IP6 supports studies that have either failed, or have found only limited evidence for IP6

in wetland systems (EI-Rifai et al. 2008; Turner 2006; Turner and Newman 2005; Turner

et al. 2006a). Yet at the same time, recent research has found significant levels of IP6

under certain, presumed anaerobic conditions (McDowell 2009; Turner and Weckstrom

2009; Zhang et al. 2009b). This leads to the conclusion that, if unstable under anaerobic

conditions, it is anaerobosis in concert with site mineralogy, possibly due to iron redox

processes (Heighton et al. 2008), that determine the stability of extracellular IP6. In

addition, the confounding influence of site salinity on IP6 stability has been suggested

(Turner and Weckstrom 2009), since increasing salinity appears to result in release of

organic P from freshwater stream materials (Gardolinski et al. 2004). However, it is









unclear if the observed trend in decreasing IP6 concentrations with increased salinity

(Suzumura and Kamatani 1995b) represents a difference in stability or distance from

terrigenous inputs.

Experimental Considerations for Application to Soils

For reasons established above (see Nuclei interactions) the vast majority of

studies into biogenic P using 31P NMR spectroscopy do so in solution (Table 2-7). This

requires the extraction of P within a suitable medium. Ideally, the extraction process

would lead to the recovery of all P forms while minimizing alteration. Certain studies

focused on the lipid profile of ecosystem components have applied organic solvents

(Bardygulanonn et al. 1995; Watts et al. 2002), yet such targeting of a specific sub

group of biogenic P is unusual. The ability of 31P NMR spectroscopy to distinguish

diverse chemical forms leads to its common application in more generalized extracts

(Cade-Menun 2005b; Cade-Menun et al. 2002). The use of 31P NMR in characterizing

sequential extractions of a sample may yield useful information on the functional nature

of P recovered by operational procedures (Baldwin 1996; Turner et al. 2006b), but it

also encounters problems with low sample concentrations and stepwise modification of

P forms. More commonly, a single alkaline extraction step is applied. First developed in

terrestrial systems (Bowman and Moir 1993; Newman and Tate 1980), it was designed

to recover both organic P, and P held within organ-metal complexes (Cade-Menun and

Preston 1996; Turner et al. 2005).

Alkaline extractions have often been used in concert with the metal chelators

ethylenediaminetetraacetic acid (EDTA) and Chelex to improve recovery of biogenic P

(Ahlgren et al. 2007; Cade-Menun 2005b; Turner et al. 2005). Yet it should be noted

they differentially impact subsequent NMR analysis. Ethylenediaminetetraacetic acid









retains any chelated paramagnetics in solution (except at very high pH values where co

precipitation may occur (Turner 2004)) which although reducing spin-lattice T1 constants

and leading to more rapid nuclei relaxation may also lead to undesired line broadening

(McDowell et al. 2006). In comparison, the use of the solid Chelex resin removes

paramagnetics from the solution, but may also remove certain biogenic P forms (Cade-

Menun and Preston 1996). A desire to reduce the impact of sample associated

paramagnetics has also led to researchers applying pre-extraction steps (Table 2-7).

These have ranged from the use of an initial mineral acid (Sannigrahi and Ingall 2005;

Turner and Weckstrom 2009), or metal chelator, e.g. EDTA (Ahlgren et al. 2007; Hupfer

and Gachter 1995; Khoshmanesh et al. 2002) to the adoption of the reducing agent,

dithionite (DeGroot and Golterman 1990) potentially in concert with metal chelation

(Carman et al. 2000; McDowell and Stewart 2005a) to both reduce and remove

paramagnetic Fe3+. In addition, the treatment of final extracts with ion exchange media

(Pant et al. 2002; Robinson et al. 1998), reducing agents (Ahlgren et al. 2005; Reitzel et

al. 2007; Zhang et al. 2009b) and organic precipitants (Ding et al. 2010) have all been

employed to reduce the impact of paramagnetics and completing humic substances on

spectrum acquisition. In addition to removing certain biogenic P forms (Ahlgren et al.

2007; Cade-Menun 2005b) and altering T1 relaxation constants (Cade-Menun et al.

2002; Ding et al. 2010; McDowell et al. 2006), the use of treatments to reduce

paramagnetics may lead to direct modification of certain functional forms. For example,

within the study of benthic sediments the use of sample pre-extraction to remove

paramagnetics is routine (Ahlgren et al. 2006a; Reitzel et al. 2007), yet polyphosphates

known to be stable under alkaline extraction conditions are catalytically hydrolyzed by









the presence of divalent cations (Harold 1966). The removal of such cations in a pre-

extraction stage may lead directly to the preservation of polyphosphates, which would

otherwise be lost from the acquired spectrum.

Although precise, Fourier transformed NMR spectroscopy is not a sensitive

technique. The observable nuclei 31P is 100 % abundant in nature yet its direct analysis

at environmental concentrations of soil extracts would often require unfeasibly long run

times. As a result, a process for concentrating samples maybe required. Two main

methods are commonly applied, rotary evaporation and lyophilization (Table 2-7). Given

known matrix hydrolysis of biogenic P (Turner et al. 2003d), snap freezing (-800C ) to

avoid excessive alkaline conditions during crystallization and lyophilization has often

been suggested as the best method (see APPENDIX A4). The influence of rotary

evaporation on the speciation of biogenic P during the sample concentration step is

currently not known.

Conclusions

The majority of studies purporting to quantify organic P, considered the largest

proportion of biogenic P, are based upon interpretation of operationally defined

procedures. While giving an indication as to the potential important role organic and

biogenic P may play in wetlands, such approaches are limited. The application of

advanced techniques, such as 31P NMR spectroscopy, allow for a more detailed and

complete analysis of functional biogenic P groups within various ecosystem

components of wetlands.

Those studies that have used 31P NMR to explore the functional nature of biogenic

P in wetland soils have highlighted a biogenic P signature fundamentally different from

the more highly studied terrestrial systems, including a prevalence of phosphodiesters,









and a complex site-specific interaction of specific P compounds including IP6. The

potential for a significant and dynamic polyphosphate pool has also been found, with

accurate quantification in highly studied lacustrine systems, and at least an indication of

its presence in palustrine systems. While not unique to wetlands (Turner et al. 2004) it

appears to represent a more substantial pool in wetland systems (Hupfer and Gachter

1995). This dissertation seeks to develop our understanding of biogenic P by expanding

the range of palustrine wetlands studied with solution 31P NMR spectroscopy. In

addition to gaining basic information on the functional forms present within a greater

range of wetlands, this dissertation will advance our understanding of the mechanistic

drivers that determine biogenic speciation and turnover within wetland soils.









Table 2-1. Hierarchical classification of methods used in the
phosphorus in soils and sediments
Method
In-situ analysis

Solid state 31P NMR
Scanning/Transmission Electron Microscopy
X-ray Adsorbtion Near Edge Structure (XANES)


study of organic

Example Reference


(Shand et al. 1999)

(Harris and White 2008)


Ex-situ analysis

Direct determination of organic P


Single step
Colorimetric determination pre and post
oxidation
31P NMR Spectroscopy
Chromatographic separation
Enzymatic hydrolysis
Sequential extraction
Acid pretreatment, organic solvent
Acid, Alkali extraction
Salt, Alkali extraction
Salt, Acid, Alkali extraction
Acid, Alkali extraction under reducing
conditions


(Potter and Benton 1916)
(Newman and Tate 1980)
(Gilbin et al. 2000)
(Pant and Warman 2000)

(Halstead et al. 1966)
(Mehta et al. 1954)
(Ghani 1942)
(Hedley et al. 1982)
(Sommers et al. 1970)


Indirect determination of organic P


Single step
Acidic extraction pre and post oxidation of
organic matter
Chemical oxidation
Low temperature ignition
High temperature ignition


Sequential extraction
Salt, Acid, Alkali extraction under reducing
conditions

Chelating extract of inorganic P under
reducing conditions


Peterson 1911 In (Pearson
1940)
(Legg and Black 1955)
(Saunders and Williams
1955) (Aspila et al. 1976)



(Ruttenberg 1992)

(De Groot and Golterman
1990)










Table 2-2. Soil biogeochemical characteristics and estimates of total organic phosphorus as determined in surface (0-10
cm) soils from a range of wetland units in South Florida (Reddy et al. 1998). Values represent averages 1
standard deviation


Hydrological unit within
Everglades
Water Conservation Area 1
Water Conservation Area 2
Water Conservation Area 3
Holey Land Wildlife
Management Area


TC TN
(mg gl) (mg gl)


5.8 0.1
7.2 0.03
6.7 0.03

7.5 0.1


TP
(pg g1)


Aspila (1976)t
Organic P
n (% total P)


440 5 30 0.6 544 41 90
420 5 28 0.5 685 42 96
4106 29 0.5 457 14 188

400 9 25 0.64 543 53 36


Everglades Agriculture Area nr nr nr 654 77
t = organic P estimated by parallel extraction with 1 M HCI before and after ignition at 550C
$ = sequential fractionation 1 M KCI. 0.1 M NaOH, 0.5 M HCI
nr = not reported


Sequential frac.t
n Organic P Difference
(% total P) (% total P)


69 9
67 17
64 11

45 44

nr 7


Table 2-3. Estimates of total organic phosphorus as determined in three surface sediment samples from
Po river delta, Saca di Goro Italy (Barbanti et al. 1994).
Organic P (% total P)


a lagoon on the


Lagoon sediments
Saca di Goro, Itlay

G1
G8

G10


p Organic C CaCO3 Total P
S(mg g-) (mg g ) (pg g-1)


19.5
24.7

21.5


835
1121

1193


AACt


GB-
MgCl2washW


GB-
MgCl2wash
corrected


30.3
25.0

21.5


SEDEX'


48.1
41.3

41.5


t = organic P estimated by parallel extraction with 1 M HCI before and after ignition at 550 uC (Aspila 1976)
: = sequential fractionation based on modified Golterman and Booman (1988) using MgC2 washes to minimize resorption
= corrected for presence of significant apatite concentrations
% = sequential fractionation based on Ruttenberg et al (1992)


41.9
38.4

35.4


30.5
24.5
29.6









Table 2-4. Relative distribution of biogenic phosphorus within living biota.
Source Nucleic acids Phospholipids
(Magid et al. 1996)t
Escherichia coli 65 15
Fungi 58 20
Spirodella 60 30
Nicotina 52 23


Phosphomonoesters


Polyphosphates


(Makarov et al. 2005) *
Plants
Fungi
Bacteria


(Koukol et al. 2008) *
Mycorrhizal fungi
Saprotrophic fungi


(Bunemann et al. 2008c) Phosphodiesters
Bacteria 18
Fungi 2
t = % of organic P determined
: = % of P determined by alkaline extraction and solution 31P NMR










Table 2-5. A selection of functional groups based upon phosphorus, and select
examples of compounds containing these groups mentioned in this
dissertation. If appropriate only single canonical structure depicted, R stands
for akyl or aryl group.


Functional group


Generic structure


oxidation state


Phosphine


Phosphine oxide


Phosphinite


Phosphinate


Phosphites


RI


II

Ra




I N

0



Ra










o

R'O







R2


Phosphonate


Phosphate


Phosphoramidate


(-III)




(-1)




(-1)




(+1)




(+111)




(+111)





(+V)


(+V)










Table 2-5. Continued


Generic compound classification used in this dissertation based
upon phosphorus functional group set out above


Example compounds


Phosphomonoesters





Phosphodiesters


HO

, II

iHO


Adenosine 5' monophosphate (AMP)
Ethanolamine phosphate
a-glucose 1-phosphate
myo-lnositol hexakisphosphate
Phosphoenolpyruvate

Deoxyribonucleic acid (DNA)
Ribonucleic acid (RNA)
L-a-Phosphatidyl choline
L-a-Phosphatidyl ethanolamine
Adenosine 3',5'- cyclic
monophosphate


Table 2-5. Continued


Phosphonates


Phosphamide



Polyphosphates


0



HO


a

II


2-Aminoethyl phosphonic acid
N-(Phosphonomethyl)glycine
(Glyphosate)


Phosphocreatine


Phosphoanhydride






Organic polyphosphates


0 I0 0
HO no HO
n

0 0

II II
000 0P\ ^'OH
HO HO


Pyrophosphate (n=0)
Polyphosphate (n>1)





Adenosine 5' diphosphate, ADP
Adenosine 5' triphosphate ATP










Table 2-6. Studies employing
Reference
Water Column
(Knicker and Nanny 1997;
Nanny and Minear 1994a)
(Nanny and Minear 1994b)

(Nanny and Minear 1997)

(Clark et al. 1999)

(Kolowith et al. 2001)

(Selig et al. 2002)
(Benitez-Nelson et al. 2004)

(Paytan et al. 2003)

(Cade-Menun et al. 2005)
(Cade-Menun et al. 2006)

(Sannigrahi et al. 2006)
(Reitzel et al. 2009)


Biota
(Bardygulanonn et al. 1995)
(Early et al. 1996)


(Summers et al. 2000)

(Pant and Reddy 2001)
(Khoshmanesh et al. 2002)

(Watts et al. 2002)


31P nuclear magnetic resonance spectroscopy in wetland and aquatic systems


Focus

Overview of considerations when using NMR to study dissolved P in the
water column
Use of Lanthanide shift reagents to improve resolution when monitoring
dissolved organic P in solution
Use of size fractionation, bromination and Lanthanide shift reagents to
elucidate further structure
Solid state analysis of high molecular weight dissolved P shows high levels
of phosphoesters and phosphonates
Solid state analysis of marine ultra filtered waters. Showing high
proportion of phosphoesters and phosphonates
Nature of dissolved and particulate P in lake water
Particulate inorganic and organic P in oxic and anoxic marine water
column
Evidence of heterogeneity in organic and inorganic biogenic P within
seston
Storage and pretreatment of marine water column particulates
Using solution 3P NMR to characterization of dissolved and particulate P
within river and flood plain waters.
Distinct difference between dissolved and particulate P
Precipitation of soluble P from lake water using poly aluminum chloride.
Subsequent identification of phosphomonoesters, polyphosphates and
DNA using alkaline resuspension.


Profiling of phospholipids in lake sediments
Characterization of phospholipids within membranes of the sponge
Eunapius fragilis Leidy

In vivo determination of P as evidence of cellular acidification during cell
elongation of Potamogeton pectinatus
Impact of hydrology on water soluble lechate from macrophytes
Luxury uptake of P resulting in polyphosphate accumulation in benthic
bacteria
Profiling of phospholipids from microbial biomass present within stream
sediments


Wetlandt


Location


Lacustrine IL, USA

Lacustrine IL, USA


Lacustrine,
Riverine
Marine

Marine


IL, USA

Pacific

Various


Lacustrine Germany
Marine Pacific


Marine

Marine
Riverine,
Lacustrine
Marine
Lacustrine


Global

CA, USA
SC, USA

HI,USA
Denmark


Lacustrine WI, USA
Lacustrine Lake
Michigan,
USA
NA

Palustrine FL, USA
Palustrine Australia


Riverine


Ontario
Canada










Table 2-6. Continued (Biota)
Reference

(Hupfer et al. 2004)
(Reitzel et al. 2006a)

(Reitzel et al. 2007)

Soil (methodological papers)
(Shand et al. 1999)
(Delgado et al. 2000)
(McDowell and Stewart 2005a)

(Turner et al. 2006b)


(Ahlgren et al. 2007)

(Turner et al. 2007b)

(EI-Rifai et al. 2008)
(Turner and Weckstrdm 2009)

Soil
(Bedrock et al. 1994)


(Hupfer and Gachter 1995)

(Baldwin 1996)

(Robinson et al. 1998)
(Carman et al. 2000)


(Sundareshwar et al. 2001)


Focus


Location


Polyphosphates detected in settling lake seston
Evidence of DNA, pyrophosphate and phospholipids in settling seston

Presence of various biogenic P forms within plankton and settling seston


Potential for use of solid state 31P NMR in peat soils
Solid state analysis of calcareous marshes
Use of a Ca-EDTA dithonite pre-extraction step to reduce line broadening
in samples with high paramagnetics
Evidence that orthophosphate may complex with organic molecules
preventing detection by standard molybdate colorimetry, potentially leading
to an overestimation of organic P.
Comparison of NaOH and NaOH + EDTA extraction using bicarbonate
buffered dithionite or EDTA as a pre-extraction step.
Comparison of sample handling procedures prior to extraction and
identification of P forms
Parallel analysis with mass spectroscopy
Use of phytate within brackish sediments as a paleo-indicator


Forms of P present in blanket peat under different management and
vegetation regimes. Presence of phosphonates and polyphosphates taken
as evidence of microbial activity
Detection of polyphosphate as an important transient sink within benthic
sediments.
NMR analysis coupled to modified SEDEX (Ruttenberg 1992) sequential
extraction scheme.
NMR analysis coupled to pre-extraction of labile P in high organic soils
Samples from various Lakes and the Benthic sea analyzed. Oxic/anoxic
conditions and presence of various cations suggested as source of
variability.
Pyrophosphate accumulation associated with anthropogenic impact of
coastal systems


Lacustrine Europe
Lacustrine Denmark

Lacustrine Sweden


Palustrine Scotland
Palustrine Spain
Riverine New
Zealand
Palustrine FL, USA


Lacustrine Sweden

Palustrine FL, USA


Palustrine
Marine,
Riverine


FL, USA
Denmark


Palustrine Scotland


Lacustrine Switzerland

Lacustrine Australia


Palustrine
Lacustrine,
Marine

Marine


FL, USA
Sweden


SC, USA










Table 2-6. Continued (Soil)
Reference Focus
(Pant et al. 2002) Analysis of P forms within surface sediments within a submerged aquatic
vegetation SAV treatment wetland
(Hupfer et al. 2004) Origin and diagenesis of polyphosphates in lake across various trophic
states
(Ahlgren et al. 2005) Attenuation of P forms with depth in sediments, half-life times estimated for
pyrophosphate and organic forms
(Turner and Newman 2005) The importance of phosphodiesters as a P pool in calcareous wetland
systems
(Sannigrahi and Ingall 2005) Influence of oxic/anoxic bottom waters in determining stability of P forms,
esp. polyphosphates. Using solid-state 31P NMR on demineralized
sediments
(Ahlgren et al. 2006a) Analysis of three oligotrophic lakes showing high variability in the presence
of polyphosphate.
(Ahlgren et al. 2006b) Degradation and half-life time estimates for various biogenic P compounds
in marine benthic sediments
(Reitzel et al. 2006a) Determination within lake sediments from across arrange of trophic states
(Reitzel et al. 2006b) Analysis of changes in identified groups with time in the sediment and with
addition of Al as a lake management strategy
(Turner 2006) Analysis of soils under rice cultivation, including flood irrigation
(Turner et al. 2006a) Identified biogenic forms within treatment wetlands shown to be dominated
by phosphodiesters.
(Reitzel et al. 2007) Rapid degradation of polyphosphates with depth and recalcitrance of other
biogenic P forms
(Bai et al. 2009) Presence of organic P in a eutrophic lake
(Sundareshwar et al. 2009) Diversity of P forms used as a measure of ecosystem function

(Zhang et al. 2009b) Surficial sediments from 7 shallow lakes demonstrating various
trophic status. No evidence of polyphosphates, IP6 identified
(McDowell 2009) Changes in stream sediment P forms as a result of surrounding land use
change
(Liu et al. 2009) Dominance of orthophosphate and phophomonoesters in heavily eutrophic
lake systems
t = Dominant system type as designated by Cowardin et al. (1979)


Palustrine

Lacustrine

Lacustrine


Location
FL, USA

Europe

Sweden


Palustrine FL, USA


Marine


Lacustrine

Marine

Lacustrine
Lacustrine

Palustrine
Palustrine

Lacustrine

Lacustrine
Palustrine

Lacustrine

Riverine

Lacustrine


Canada


Sweden

Sweden

Denmark
Denmark

Madagascar
FL, USA

Sweden

China
NC & SD
USA
China

New Zeland

China












Table 2-7. Methodological details of studies employing 31P nuclear magnetic resonance spectroscopy in wetland soils


Reference


Soil- (Methodological)
(Shand et al. 1999)

(Delgado et al. 2000)

(McDowell and Stewart 2005a)

(Turner et al. 2006b)

(Ahlgren et al. 2007)

(Turner et al. 2007b)

(EI-Rifai et al. 2008)

(Turner and Weckstr6m 2009)

Soil
(Bedrock et al. 1994)

(Hupfer and Gachter 1995)

(Baldwin 1996)

(Robinson et al. 1998)

(Carman et al. 2000)

(Sundareshwar et al. 2001)

(Pant et al. 2002)

(Hupfer et al. 2004)

(Ahlgren et al. 2005)

(Sannigrahi and Ingall 2005)

(Turner and Newman 2005)

(Ahlgren et al. 2006a)


Vegetationt




PE-moss

Herb

BS-S

PE-Herb,SA

SA, BS

PE-Herb,SA

PE-Herb.

BS


PE-Moss

BS

BS

PE-Herb.

BS

PE-Herb

SA

BS

BS

BS

PE-Herb.

BS


Extraction method*


Solid state

Solid State

0.25 M NaOH 50 mM EDTA

0.5 M NaOH

0.1 M NaOHO.125 M NaOH 0.25
M EDTA
0.5M NaOH

0.25 M NaOH 50mM EDTA

0.25 M NaOH 50mM EDTA



0.5 M NaOH

0.2 M NaOH + 67 mM EDTA

SEDEX sequential (Ruttenberg et
al 1992)
0.25 M 50 mM EDTA (85 C)

0.5 M NaOH

0.5 M NaOH 100 mM EDTA

0.4 M NaOH

0.2 M NaOH 67 mM EDTA

0.1 M NaOH

Solid state

0.25M NaOH, 50mM EDTA

0.125 M NaOH + 0.25 mM EDTA


Paramagnetic treatmentO

Pre extraction Post extraction


Acetylacetone

CBD

EDTA-Dithonite

0.5 M NaHCO31 M HCI
(Ivanoff et al 1998)
BD or EDTA

0.5M NaHCO31 M HCI
(Ivanoff et al 1998)
na

HCI


na

EDTA

na

na

CDB,MgCl2 wash

na

na

EDTA



HCI/HF

na

BD


na

na

na

Chelex X-100
Column
na

na

G-25 Sephadex

na

Dithonite

na

na

na


Concentration$




Lyophilization

na

Lypholization

Lypholization

Rotary evap

Lypholization

Lyphilization

Lypholization



Rotary evap.

Rotary evap



Lypholization

Rotary evap.

na

Rotary evap.

Rotary evap.

Rotary evap.

na

Lypholization

Rotary evap.


Acquisition Pulse
Delay
Pulse width (sl


CPMAS, HP
(90)
5 us

450

450

63

450

450

450


900

45

900



72

CP-MAS

450

63











Table 2-7. Continued


Extraction method*


Paramagnetic treatmentO

Pre extraction Post extraction


Concentration$


Acquisition Pulse
Delay
Pulse width (s)


(Ahlgren et al. 2006b) BS 0.25 M NaOH 50mM EDTA Dithonite rotary evap 63 1.25

(Reitzel et al. 2006a) BS 0.11M BD + 0.1 M NAOH na na Rotary evap. 63 1.2

(Reitzel et al. 2006b) BS 0.125 M NaOH + 25 mM EDTA na BD Rotary evap. 63 1.2

(Turner 2006) PE-CS 0.25 M NaOH 50mM EDTA na na Lypholization 450 2

(Turner et al. 2006a) PE-Herb 0.25M NaOH, 50mM EDTA na na Lypholization 450 2

(Reitzel et al. 2007) BS 0.1 M NaOH BD BD Rotary Evap 63 1.2

(Bai et al. 2009) BS 0.1 M NaOH EDTA, Dithonite na Rotary evap 90 2

(Liu et al. 2009) BS 0.25 M NaOH 50mM EDTA na na Lypholization 450 2

(McDowell 2009) BS 0.25 M NaOH 50 mM EDTA na na Lyophilization 450 4

(Sundareshwar et al. 2009) PE-Herb/ CS 0.25M NaOH 0.1M EDTA na na Lypholization

(Zhang et al. 2009b) BS 0.25M NaOH na BD Lypholization 90 4

t vegetation; PE = persistent emergent, moss = moss peat, Herb = Herbaceous, CS = flooded cultivated soil BS = Benthic sediments, S =
stream, SA= submerged aquatic,
$ extraction method; single step extraction or steps used within 31P NMR studies
Paramagnetic treatment; method applied to minimize effect of paramagnetics, pre or post extraction; (EDTA)= Ethylenediaminetetraacetic acid,
(CDB)= Citrate+ Dithonite+ Bicarbonate, (BD)= Bicarbonate + Dithonite,
i Concentration; method used to concentrate sample prior to 31P NMR analysis,
Acquisition parameters used in 31P NMR analysis
na= not applicable
- =not noted by reference


Reference


Vegetationt














l-
W10-

59 5-

o_ 0
0


E10-

5-

0-


Organic


P (% of total P)


Figure 2-1. Frequency histogram of estimated organic P within 117 wetlands (32
lacustrine, 85 palustrine)


Lacustrine/Riverine
Average = 37
Range = (0 73)
Interquartile range = 21


--1


15-


0 20 40 60 80 100


Palustrine
Average = 67
Range = (16 94)
Interquartile range = 25













- IGNITION
Average = 66
SIQR= 32



SEPI
Average = 46
. IQR= 22



SEPO
Average =60
IQR= 40
I- i


20


40


100


Organic P (% of total P)

Figure 2-2. Frequency histogram of estimated total organic P within 117 wetlands
broken down by general method grouping. IGNITION = parallel extraction of
inorganic P pre and post oxidation of organic matter, organic P estimated by
difference, SEPI = sequential extraction procedures focused on characterizing
inorganic P, SEPO = sequential extraction procedure focused on
characterizing organic P. (IQR = Inter Quartile Range)


60


80









Ortho-P


\ rt" myo-inositol hexakisphosphate
S(pH 5.0-9.0)
Ortho-P rt




Ortho-P


a-d-glucopyranose 6-phosphate








Figure 2-3. Structural comparison of myo-lnositol hexakisphosphate and a-d-glucopyranose 6-phosphate (pyranose ring
form of a-d-glucose 6-phosphate ). Ortho-P represents orthophosphate group with various levels of
deprotonation dependent upon environmental pH. Based upon (Barrientos and Murthy 1996)










Phospholipids


S/
N*-
o-
O0


SPhophatidate


II o 2
R 0
0

0 0
---- Phosphatidylcholine (lecithin)

Phosphatide N*-


--- Phosphatidylethanolamine (cephalin)


O
NH*


- O


COO

NH*


OH
OH


OH OH


Phosphatide


Figure 2-4. Basic phospholipid compound structure found within both eukaryotic and
prokaryotic cell membranes. Phosphodiester containing phosphoglycerides
constitute phosphatide after additional esterification, by alcohol functional
groups. R = fatty acid chain.


Sphinomyelin O


R HN


0
-> Phosphoglycendes


Phosphatide


--- Phosphatidylserine

Phosphatide


-- Phosphatidyl inositol


Phosphatide


- Diphosphatidylglycerol (cardiolipin)
O 0Pho
Phosphatide
























No Bo field


AE =hco


Energy
absorbed

B, rf pulse


--------------- -----------
,- ------------ <---4------~'

Energy
Lost

I--"-- Ir-I- --1-"--- ----
~~3--'--- > -~------------


B field strength


B0
applied magnetic
field


Figure 2-5. Response of phosphorus nuclei to an applied magnetic field. A) Zeeman splitting of a mi = /2 system, B)
graphical representation of precessional orbit of 31P nuclear magnetic dipole around the applied magnetic field
with transition due to applied B1 radio frequency pulse. Adapted from (Cade-Menun 2005b; Knicker and Nanny
1997).


A






CL
Z
LU











Orthophosphate


MDP (internal standard)






Phosphonates



qIh*iWWN*


Orthophosphate monoesters


Phosphodiesters

DNA


Polyp

kl h4-


Polyphosphate (mid chain)


hosphate (end chain)

Pyrophosphate
^^^^\rAyS*Y.<^^M


-10


-20 PPM


Chemical shift (ppm)

Figure 2-6. Solution 31P nuclear magnetic resonance spectra showing common functional groups. Sample represents
surface soil from a Carolina Bay, SC USA extracted using standard procedure (0.25 mol L-1 NaOH 50 mmol L-1
EDTA) and concentration via lypholization. Spectra were acquired using a Bruker Avance 500 Console with a
Magnex 11.75 T/54 mm magnet using a 10 mm BBO Probe at a stabilized 250 C with a calibrated (~300) pulse
length, a zgig pulse program, and a 2 s pulse delay.


k










CHAPTER 3
INTERACTION OF BIOGENIC PHOSPHORUS WITH ANION EXCHANGE
MEMBRANES: IMPLICATION FOR SOIL PHOSPHORUS ANALYSIS1

Introduction

Anion exchange membranes are commonly employed to study P dynamics in soils

and sediments (Myers et al. 2005; Saunders 1964; Skogley and Dobermann 1996). In

particular, they are used to measure 'readily exchangeable' phosphate in soils

(Sibbesen 1978) and, in conjunction with hexanol fumigation, to determine P contained

within the soil microbial biomass (Kouno et al. 1995; Myers et al. 1999). Anion

exchange membranes offer potential benefits over conventional soil P tests, given their

action as passive ion sinks analogous to biological uptake (Qian and Schoenau 2002).

Their use in the field can provide an integrated measure of P availability (Cooperband

and Logan 1994; Drohan et al. 2005; Meason and Idol 2008), and has led to the

development of commercial products designed to determine nutrient supply rates (e.g.,

PRSTM probes; Western Ag development Innovations, Saskatoon, Saskatchwan,

Canada).

Despite numerous assessments of the practical influence of experimental design

on both batch and field-deployed membranes (Mason et al. 2008; Qian and Schoenau

1997, 2002; Qian et al. 1992; Sato and Comerford 2006), surprisingly few studies have

examined their interaction with organic and condensed inorganic P compounds. Yet

there is a clear potential for sorption and recovery of organic P to both resin beads and

membranes (Cooperband et al. 1999; McDowell et al. 2008; Rubaek and Sibbesen

1993). Of particular note is the ~100% recovery of both phytic acid (myo-lnositol

1 Accepted for publication in a modified format, Soil Science Society of America Journal 2010









hexakisphosphate; IP6) and glucose 6-phosphate by membranes, type 204-U-386

(lonics Inc Watertown, MA) when exposed at 16 pg P cm-2 and loaded with a chloride

counterion (Cooperband et al. 1999).

Consideration of potential interactions between anion exchange membranes and

organic and condensed inorganic P compounds is important given the assumptions

made during routine application of the exchange media to study P dynamics. First, if

significant levels of organic and condensed inorganic P are recovered by anion-

exchange membranes and transferred to the eluant solution, their inclusion in

subsequent measurements will lead to an overestimation of 'bioavailable' or 'readily-

exchangeable' orthophosphate.

Second, when anion exchange membranes are employed as an ion sink during

the measurement of microbial P (Kouno et al. 1995; Myers et al. 1999) the analysis of

total P in the eluant is typically omitted, because the difference between orthophosphate

and total P is assumed to be negligible (Brookes et al. 1984), as in fumigation-

extraction procedures (Brookes et al. 1982; Hedley et al. 1982). However, if organic or

condensed inorganic P compounds released from lysed microbial cells are adsorbed

directly to anion exchange membranes without conversion to orthophosphate by

enzymatic or matrix induced hydrolysis, this may lead to an under-estimation of

fumigation-released (microbial) P.

Lastly, if organic or condensed inorganic P are recovered by anion exchange

membranes, both the nature of the eluant used for P recovery and the method of P

detection may influence levels of P determined as 'recovered' from a sample. The use

of non-selective elemental analysis such as ICP-OES (inductively-coupled plasma









optical emission spectrometry), or the hydrolysis of P containing compounds during

elution by mineral acids may erroneously attribute complex forms of P as 'bioavailable'.

The aims of this study were to establish (i) the potential interactions between a

commonly used anion exchange membrane and a range of model organic and

condensed inorganic P compounds, and (ii) the implications of any interaction during

the routine application of anion exchange membrane procedures to wetland soils

expected to contain high levels of soluble organic and condensed inorganic P.

Methods

After establishing the exchange capacity for orthophosphate, anion exchange

membrane strips were exposed to a range of standard P-containing compounds to

determine the levels of relative ion exchange and compound recovery. Eluant solutions

were analyzed for both molybdate-reactive P and total P, while solution 31P nuclear

magnetic resonance (NMR) spectroscopy was used to investigate the stability of P

compounds during their exposure to the anion exchange membrane strips. Finally, a

series of anion exchange membrane extractions from soils were analyzed for both total

and molybdate-reactive P to determine the potential for previously-established

interactions to impact P recovery in field samples.

Anion Exchange Membranes

The anion exchange membrane characterized in this study (BDH Prolabo

Product number: 551642S) is used routinely in field and laboratory procedures

(McLaughlin et al. 1994; Myers et al. 2005; Roboredo and Coutinho 2006; Turner and

Romero 2009b). Available through VWR International, UK (Lutterworth, Leicestershire,

UK) and CTL Scientific Supply Corp. (Deer Park, NY, USA), the membranes (previously

sold directly under the BDH brand name) use a polystyrene / divinylbenzene copolymer

84









base doped with quaternary ammonium as the ionogenic group and are supplied in

sheets 125 x 125 mm preloaded with chloride counterions.

Anion exchange membranes were cut into 1.5 x 6.25 cm strips, yielding a reactive

area of 18.75 cm2 per strip. The strips were charged with HCO3 counter ions by shaking

25 strips in a 250 mL HDPE bottle with three sequential changes of 200 mL 0.5 mol L-1

NaHCO3 over a 24 h period. Strips were then rinsed three times in deionized water (DI)

to remove adhering solution. Bicarbonate was chosen as the counterion since it mimics

biological uptake from the rhizosphere (Qian and Schoenau 2002; Sibbesen 1978) and

is considered preferable when studying P due to its lower affinity for exchange sites

compared to other commonly used counter ions (e.g., chloride) (Skogley and

Dobermann 1996).

After exposure to a test solution, membranes were rinsed in DI, shaken dry of

excess water, and immersed in a conical tube containing 50 mL of 0.25 mol L-1 H2SO4

and shaken for 3 h, after which a subsample of the eluant was decanted into a 20 mL

scintillation vial. Membranes were cleaned using a secondary acidic wash (0.25 mol L1

H2SO4, 1 h) and rinsed with multiple changes of DI before regeneration with bicarbonate

as described above. Phosphorus recovery was unaffected by repeated use of the

membrane strips (data not shown), despite some discoloration following use with soil

samples.

Phosphorus Determination

Molybdate-reactive P, an operationally defined parameter, was determined in the

extracts by automated colorimetry with detection at 880 nm using a flow injection

analyzer (Lachat Quickchem 8500, Hach Ltd, Loveland, CO). Total P was determined in

the extracts by ICP-OES (Optima 2100, Perkin-Elmer Inc., Shelton, CT).

85









Orthophosphate standards were prepared in the same matrix as the samples (0.25 mol

L-1 H2SO4) for both analyses.

Experimental Design

Anion exchange membrane exchange capacity

The response of the exchange membranes to increased orthophosphate

concentrations was tested to determine the exchange capacity of the membranes and

the range over which the dynamic anion exchange mimics that of an infinite sink.

Replicate (n=3) preloaded and rinsed membranes were placed in 250 mL HDPE bottles

with 75 mL of orthophosphate standards between 0 and 2.5 mg P and shaken for 24 h.

Adsorbed P was eluted and determined as described above.

Phosphorus recovery by anion exchange membrane strips

A series of organic and condensed inorganic P compounds were prepared at an

approximate concentration of 200 pg P mL-1 in DI, with precise concentrations

determined subsequently by ICP-OES (Table 3-1). Duplicate preloaded and rinsed

anion exchange membrane strips were loaded individually into 250 mL HDPE centrifuge

bottles with 70 mL DI and 5 mL of standard P-containing solutions. After standard

exposure and elution from membranes (see above), eluants were stored in 20 mL

HDPE scintillation vials at 4C until analysis by molybdate colorimetry within 72 h and

total P within one month.

Purity and stability of organic and condensed phosphates in deionized water

Standard solutions were analyzed by solution 31P NMR spectroscopy over a period

of 24 h to determine the stability of P compounds during exposure to anion exchange

membranes. Solutions (~ 200 pg P mL-) were mixed 1:1 with an internal standard

(methylenediphosphonic acid; MDP) (~ 50 pg P mL-) in the same matrix. Of the

86









resulting mixture, 0.9 mL was added to 0.1 mL deuterium oxide, the solution was

vortexed, and then loaded into a 5 mm diameter NMR tube. Solution 31P spectra were

acquired after 15 min and 24 h using a Bruker Avance 500 Console, Magnex 11.75 T/54

mm magnet fitted with a 5 mm BBO probe. Acquisitions were run at a stabilized 25C

with a 4.833 ps (~300) pulse length and a 2 s recycle delay. An average 2000 scans

(run length 1 h 21 min) were acquired and the resulting spectra referenced against

internal MDP with a chemical shift (6) = 17.46 ppm, determined with reference to an

externally held 85% H3PO4 standard (6 = 0 ppm).

Extraction of phosphorus compounds from wetland soils by anion exchange
membranes

A comparison of total and molybdate reactive P in the membrane eluants from

wetland soil extractions was carried out to determine the potential for organic and

condensed inorganic P recovery when anion exchange membranes are used to

measure 'bioavailable' and microbial P. A batch process using anion exchange

membranes prepared as described above was applied to 27 soils collected from a

nutrient gradient in the freshwater San San Pond Sak wetland, a domed peatland in

Bocas del Toro province, western Panama (Phillips et al. 1997). Soils were all high in

organic matter (total C 41-50%), but contained a range of total P concentrations (388-

1028 mg P kg-1).

Two fresh samples of each soil (3.5 g dry weight equivalent) were weighed into

250 mL HDPE centrifuge bottles and sample-specific volumes of DI were added to bring

the total water content to 75 mL. Both samples received DI and a single anion exchange

membrane strip (1.5 x 6.25 cm), with one subsample also receiving 1 mL of hexanol

(95%, Sigma Aldrich, St Louis, MO). After 24 h shaking, the anion exchange membrane









strips were removed, rinsed, eluted, and the resulting solution analyzed for total and

molybdate-reactive P as described above. A significant difference between total and

molybdate-reactive P would indicate the retention of organic or condensed inorganic P

compounds that were not hydrolyzed to molybdate-reactive P during elution in 0.25 mol

L-1 H2S04.

In normal application of the method, the difference in molybdate-reactive P

between the fumigated and non-fumigated samples would be attributed to 'fumigation-

released P'. This could then be adjusted for extraction efficiency and microbial biomass

not killed by fumigation (k factor) to provide an estimate of microbial P (Kouno et al.

1995). However, given problems with both correction factors, some authors prefer

simply to use uncorrected fumigation-released values (Bunemann et al. 2008a). For this

study uncorrected total and molybdate-reactive P in the eluants of both unfumigated

and fumigated samples were compared as pg P g-1 dry weight.

Results

Anion Exchange Membrane Capacity

Orthophosphate recovered from a non-competitive environment was within a 5%

error for the majority of the exposure levels tested in this study (Figure 3-1). Only at

levels in excess of 117 pg P cm-2 was recovery unacceptable, defined here as < 95%.

Taking a conservative working maximum of 100 pg P cm-2 of membrane and the utilized

membrane area to soil ratio of 5.6, this would equate to an orthophosphate sorption

capacity (560 pg P g-) well above the maximum likely to be encountered in the natural

environment.









Organic and Condensed Phosphorus Recovery by Anion Exchange Membrane
Strips

Total P concentrations in eluants demonstrated that most of the P compounds

interacted with the anion exchange membranes (Figure 3-2). Three compounds were

recovered completely (sodium pyrophosphate, glucose 6-phosphate, and adenosine 5'-

monophosphate), while three others showed recoveries between 20 and 60% (sodium

hexametaphosphate, 2-aminoethylphosphonic acid, and phytate). The macromolecules

RNA and DNA showed limited recovery (~ 10% and 0.2%, respectively), although it was

unclear if this represented true ion exchange at the ionogenic sites or contamination

due to physical adherence of the macromolecule to the polymeric membrane. Under the

concentrated conditions of the experiment a 'sheen' was observed on the membrane

strips exposed to DNA. I therefore considered that RNA and DNA recoveries were

inconclusive in the context of this experiment.

Of the compounds showing significant interaction with the membranes, some were

recovered as molybdate-unreactive P (adenosine 5'-monophosphate, 2-

aminoethlphosphonic acid, and phytate), while large proportions of others were

recovered as molybdate-reactive P (Figure 3-2). For example, of the ~100% of

pyrophosphate and glucose 6-phosphate recovered by the anion exchange membrane,

36 and 69%, respectively, were detected as molybdate-reactive P.

Purity and Stability of Phosphorus Compounds in Deionized Water

Comparison of the 31P NMR spectra from test compounds with a known

orthophosphate standard (Figure 3-3) showed neither contamination of the original

compounds with orthophosphate, nor orthophosphate released by hydrolysis in DI over

a period of 24 h The presence of molybdate-reactive P in the eluant solutions was









therefore considered to be due entirely to acidic hydrolysis during elution of P from the

membranes and subsequent storage prior to colorimetric analysis. Hydrolysis during

molybdate colorimetry (< 1 min contact time) was likely to be negligible given the

relatively long period of elution/storage in 0.25 mol L-1 H2S04.

Application to Wetland Soils for Exchangeable and Fumigation-Released
Phosphorus

Comparing the total and molybdate-reactive P recovered by anion exchange

membranes in a standard extraction from wetland soils, paired t-tests show significant

(p < 0.05) differences between molybdate reactive and total P determined in both non-

fumigated and fumigated samples, but average differences of only 3.4 and 1.3%,

respectively, are within the margin of error due to calibration. I concluded that there was

no evidence organic or condensed inorganic P, undetected by colorimetric analysis,

was recovered in appreciable amounts by the anion exchange membranes.

Discussion

The orthophosphate exchange capacity for the studied anion exchange

membranes agrees well with the published information on the performance of similar

membranes, including the linear response observed to 22.13 pg P cm-2 (Schoenau and

Huang 2001) and the total exchange capacity of commercially available anion exchange

membrane products i.e.the P exchange capacity of PRSTM Plant Root Simulators,

calculated as 301.4 pg P cm-2 based upon meq charge capacity

(http://www.westernag.ca/innov/technical_l .php). The calculated window of acceptable

recovery appears more robust than that demonstrated by Mason (2008) who, assuming

that a 6.26 x 2.5 cm strip has a reactive surface of 31.3 cm2, found a linear response to

only 4.6 pg P cm-2. While this may be due to their use of a counter ion (chloride) with a









higher affinity for the membrane exchange sites, it is also due to their conservative

interpretation of acceptable recovery (S. Mason, 2008, University of Adelaide, personal

communication).

Several organic and condensed inorganic P compounds interacted, and in some

cases were fully recovered, by a routinely used anion exchange membrane method. Of

these, glucose 6-phosphate and the condensed inorganic P compounds pyrophosphate

and hexametaphosphate were recovered partially as molybdate-reactive P. Given that

orthophosphate contamination in the original compounds was negligible, and the

compounds were stable during exposure to the membranes, we conclude that some

compounds were hydrolyzed to orthophosphate during elution from the membranes in

0.25 mol L-1 H2SO4 and storage prior to colorimetric analysis. The potential for acidic

hydrolysis of organic and condensed inorganic P is a recognized source of error in the

determination of orthophosphate by the operationally-defined molybdate colorimetry

(Dick and Tabatabai 1977; Worsfold et al. 2005). Hydrolysis during the acidic molybdate

reaction (< 1 min contact time), however, was considered to be negligible, given the

relatively long period of elution/storage in 0.25 mol L-1 H2SO4.

Similar studies examining the hydrolysis of soil organic P during extraction in

strong acids reported variable results. For example, Ivanoff et al. (1998) reported 38%

hydrolysis of para-nitrophenyl phosphate during a 3 h extraction in 1 mol L-1 HCI,

whereas Bowman (1989) reported 'negligible' hydrolysis of a series of organic P

compounds (para-nitrophenyl phosphate, glycerophosphate, phytic acid, and bis-para-

nitrophenyl phosphate) during a short extraction in 1.08 mol L-1 H2SO4. Further analysis

using non-destructive direct monitoring by 31P NMR spectroscopy would be required to









establish the impact of pH-dependent hydrolysis on model compounds over eluant

holding times.

For a series of wetland soils with contrasting P contents, no discernible molybdate-

unreactive P was recovered by the anion exchange membranes (Figure 3-4).

Fumigated samples were expected to contain high concentrations of organic and

condensed inorganic P from lysed microbial cells. This study demonstrates that, given

the use of acidic eluant, there is no significant distinction between microbial P recovered

by anion exchange membranes as determined from total elemental P analysis (ICP-

OES) or operationally defined molybdate colorimetric analysis.

This does not rule out the inclusion of acid-labile organic and condensed inorganic

P within molybdate-reactive P, both of which have been reported in soil solutions

(Espinosa et al. 1999). Yet the absence of any discernible molybdate-unreactive P in

these samples may suggest that acid-labile compounds are likely to be negligible. The

use of anion exchange membranes with more benign eluants (e.g., chloride or

bicarbonate) would be needed to confirm this assumption.

In cases that have reported the recovery of organic P by exchange media, an

assumption was made, due to their mode of action, that resin extractable organic P

estimated soil labile organic P (McDowell et al. 2008; Rubaek and Sibbesen 1993). Yet

a previous study of the nature of organic P in pasture soils using phosphatase

hydrolysis reported negligible concentrations of labile phosphomonoesters (including

polyphosphoric compounds), presumed to indicate a rapid hydrolysis of these

compounds following release into the soil solution (Turner et al. 2002a). Further









investigation is needed to determine if organic P may be recovered from soils by the

anion exchange membranes tested here.

In conclusion, although anion exchange membranes can interact with a diverse

range of P forms, this does not appear to limit their application to the determination of

readily-exchangeable phosphate or microbial P in soils, particularly for samples from

natural environments. However, care must be taken in assigning P recovered by anion

exchange membranes to 'labile orthophosphate', because acid-labile organic and

condensed inorganic phosphates may also be included. This will be especially

problematic in-situations where membranes are deployed after the application of

condensed inorganic P fertilizers (Bertrand et al. 2006). Although not fully resolved the

adaptation of the anion exchange membrane procedure to include a benign membrane

eluant (e.g. chloride, bicarbonate) to minimize hydrolysis and total P determination may

allow for the parallel assessment of labile orthophosphate, and organic / condensed

inorganic P.









Table 3-1. Phosphorus compounds tested on anion exchange membranes. Stock
solution concentrations determined by ICP-OES analysis.

Standard
Compound Chemical Formula Concentration
(pg P mL-1)


Potassium dihydrogen phosphate KH2PO4 100.0
Sodium hexametaphosphate (NaPO3)n.Na20 171.1
Sodium pyrophosphate, decahydrate Na4P207.10H20 207.6
D-Glucose 6-phosphate, disodium salt hydrate C6H1109PNa2.xH20 169.0
Adenosine 5'-monophosphate, monohydrate C10H14N507P.H20 157.5
2-Aminoethylphosphonic acid C2H8NP03 195.6
Phytic acidt, dodecasodium salt C6H6024P6Na12 183.6
Ribonucleic acid, Type VI from tortula yeast -- 177.5
Deoxyribonucleic acid, from salmon testes -- 250.7
t myo-lnositol hexakisphosphate, dodecasodium salt (sodium phytate)
-:- not appropriate











120-


100-
W

%. 80- +.,
E A a' B

.. ,, 60-
>
O M
0 40- ,'-"

,ee'
G /





0 20 40 60 80 100 120 140

P supplied to AEM (pg P cm2)




Figure 3-1. Exchange capacity of anion exchange membrane (AEM) strips. Values
plotted alongside limits of 'reasonable' recovery 95-105% (----) of calculated
exposure. A typical exposure level A of up to 25 pg cm-2 equates to 469 pg
P g-1 of 'exchangeable' phosphate when using one strip (6.25 x 1.5 cm) with 1
g of soil. Limit of reasonable recovery rate, B equates to 117 pg P cm-2 in a
non-competitive environment.



















8E- [ODetected as MRP
S8- DDetected as non MRP

UJ
<

>% 60-
.O
.0


S40-
U



20-



0 1 I I II II

r r t "r r z z

.c .c 0 Q
a a a a 0
i. 1. Q. "

0 a a a-0C




E EI
'a 5
4X N


Figure 3-2. Recovery of phosphorus compounds by anion exchange membrane (AEM)
strips. Average (n=2) plotted, all repeats within 2.3%. Recovered P plotted as
molybdate-reactive P (MRP) or as non-molybdate-reactive P (difference
between total and molybdate-reactive P).


- F--- --I -- -- r. -- -- -- -- -















Orthophosphate

*v w s
S. ,-Pyrophosphate

Glucose 6-phosphate

.Adenosine 5' monophosphate

2- Aminoethylphosphonic acid

Phytate

RNA

DNA
11 llll 11 1 11111 11111 1111 ll 1111 11111 I
20 10 0 -10 -20
Chemical shift (ppm)
Figure 3-3. Solution 31P nuclear magnetic resonance spectra of P compounds measured after 24 h in deionized water.
Spectra were acquired using a 4.833 ps (~300) pulse length and a 2 s recycle delay, approximately 2000 scans
required to achieve suitable signal to noise. Spectra presented using 8 Hz line broadening and referenced to
internal standard methylenediphosphonic acid (6 = 17.46 ppm).













40-









U 1 0 .
S.x0-


I 0 10 20 30 40 50
S 300-


0 .,--6
1-- 250- 6
Po,,
200- 9"
...-^


150- o#


100-

501..
50- --I I I
50 100 150 200 250 300

Molybdate reactive P
(IPg P g soil-1)
Figure 3-4. Comparison of total and molybdate-reactive P as detected in anion
exchange membrane eluants. Non fumigated (X) and hexanol fumigated (0)
samples show significant differences (paired t-test p < 0.05) between P
determined by ICP-OES and colorimetric analysis. Yet linear correlation (non
fumigated Y = 0.98x + 0.36, R2 = 0.999, fumigated Y = 0.96x 4.08, R2=
0.997,) and the small relative differences (3.4 and 1.3%) suggest limited
recovery of molybdate unreactive P.









CHAPTER 4
A SURVEY OF BIOGENIC PHOSPHORUS IN WETLAND SOILS: A SOLUTION 31P
NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY STUDY

Introduction

Biologically sourced and cycled P represents a significant component of the

wetland P cycle (Newman and Robinson 1999; Reddy et al. 1999; Reddy et al. 2005),

with operational defined 'organic P' representing on average over 60% and up to 94% of

total P within certain wetland systems (CHAPTER 2). If consideration is also given to

polyphosphates, an often-neglected component of the P cycle (Davelaar 1993), it is

clear that biogenic P dominates total P in many wetland soils. The functional nature of

this biogenic P has profound implications upon its interaction and stability in the

environment (Celi and Barberis 2005b) as well as determining potential biological

turnover (Oberson and Joner 2005; Richardson et al. 2005; Turner 2008a). Analytical

techniques, such as solution 31P NMR spectroscopy, now allow researchers the tools to

identify and track the functional forms of this biogenic P in the environment (Cade-

Menun 2005b; McKelvie 2005). Research employing 31P NMR spectroscopy has

provided valuable insight into P cycling of wetland systems, yet has to date been limited

(CHAPTER 2). This study represents a unique effort to expand the application of the

technique to investigate the range of biogenic P found within palustrine wetland soils.

First applied to wetlands during the study of Scottish blanket bogs (Bedrock et al.

1994), solution 31P NMR spectroscopy has been applied to a number of natural and

artificial wetland systems (Cade-Menun 2005b) (CHAPTER 2). Yet, its application

requires a certain level of specialist knowledge as well as access to NMR facilities. This

has limited the type of wetlands studied to date. Published work on freshwater systems

is typified by a focus on lacustrine sediments in Europe (Ahlgren et al. 2005; Hupfer and

99









Gachter 1995; Reitzel et al. 2007) China (Bai et al. 2009; Liu et al. 2009; Zhang et al.

2009b), and Australia (Baldwin 1996), riverine systems in New Zealand (McDowell

2009) and a limited number of palustrine systems including: the highly organic

subtropical marshes of south Florida (Robinson et al. 1998; Turner and Newman 2005;

Turner et al. 2007b; Turner et al. 2006a), blanket bogs in Scotland (Bedrock et al.

1994), Carolina Bays in the USA (Sundareshwar et al. 2009), as well as flooded rice

paddies in Madagascar (Turner 2006). To date, work in palustrine systems has

highlighted both the diverse range of biogenic P forms, including polyphosphates, that

may be found (Sundareshwar et al. 2009) and a potential P composition fundamentally

different to terrestrial systems (Turner and Newman 2005) (CHAPTER 2- Range of

Applications in Wetlands).

Although it is apparent that certain wetland soils may contain fundamentally

different P compositions to upland systems, a lack of research in this transitional

ecotone leaves us unable to determine if such observations are universal. In this

chapter, I use solution 31P NMR spectroscopy to evaluate the nature of biogenic P from

a broad range of palustrine wetlands. Specific objectives include; 1) establish the nature

and diversity of biogenic P forms found within wetlands soils, 2) analyze forms identified

in the context of ancillary biogeochemical and environmental information to derive

potential controlling mechanisms or characteristics that could explain the observed P

composition.

Methods

Sampling

Surface samples were collected over the course of three years from a diverse

range of wetland systems (Table 4-1). Study site locations were dictated by available

100









access or established working collaborations, but were selected for sampling to obtain a

diverse range of climatic, hydro-geomorphic and vegetation types. Given the exploratory

nature of the study, soil sampling consisted of four surface cores (diameter 7.5 cm x 10

cm) collected from independent sites considered representative of the study wetland.

Samples were kept on ice for immediate shipment to the Univ. of Florida, or in two

cases were air dried on site (sites 25 and 26). Samples were processed by hand,

removing coarse inorganic and organic fragments >2 mm. Homogenized samples were

split, with subsamples being kept at 4C (fresh) and set out to air dry under ambient lab

conditions, nominally 10 days, under conditions of elevated air flow. Fresh samples

were analyzed for water content, pH, exchangeable P and microbial P. Air dried

samples were ground (8000D mixer mill, SPEX SamplePrep, NJ) and sieved (mesh 60,

0.250 mm) prior to analysis for total elemental composition (C, N, P, Al, Ca, and Fe) and

P composition by solution 31P NMR spectroscopy.

Biogeochemical Characterization

Fresh soil samples were analyzed for soil water content by gravimetric loss

following drying at 700 C for 72 h. Sample pH was determined on a 1:2, soil to water

suspension using a glass electrode. Exchangeable and microbial P were determined by

anion exchange membrane batch method (Kouno et al. 1995; Myers et al. 1999; Thien

and Myers 1992). In brief, two fresh samples of each soil (3.5 g dry weight equivalent)

were weighed into 250 mL HDPE centrifuge bottles and a sample-specific volume of

DDI was added to bring the total water content to 75 mL. All samples received a single

anion exchange membrane strip (1.5 x 6.25 cm, BDH Prolabo Product number:

551642S, VWR International, UK) preloaded with HC03- counter ions (CHAPTER 3),

with a parallel subsample also receiving 1 mL of 1-hexanol (95%, Sigma Aldrich, St

101









Louis, MO). Samples were sealed and placed on a reciprocating shaker for 24 h after

which membrane strips were removed, rinsed under running DDI and placed in a

conical tube with 50 mL 0.25 mol L-1 H2S04. Membranes were eluted for 3 h and the

resulting solution was analyzed for molybdate-reactive P using a discrete auto-analyzer

(AQ2+, SEAL Analytical, UK) and standard molybdate colorimetry (USEPA 1993). The

difference between P recovered between parallel samples is attributed to fumigation

released P (Bunemann et al. 2008a) and is used in this study as a proxy for microbial P

without correction factor (Jenkinson et al. 2004).

Dried and ground soils were analyzed for loss on ignition (an estimate of total

organic matter) and elemental concentrations. Total P and metals were determined by

combustion of soil at 550C in a muffle furnace for 4 h and dissolution of the ash in 6

mol L-1 HCI (Andersen 1976). Acid solutions were analyzed for molybdate P (as above)

and for AI, Ca, and Fe, using ICP-OES (Thermo Jarrell Ash ICAP 61 E, Franklin, MA).

Total soil C and N were measured by combustion and gas chromatography using a

Flash EA1112 (Thermo Scientific, Waltham, MA).

Phosphorus Composition

Phosphorus forms were characterized via a standard alkaline extract and

solution 31P NMR spectroscopy of air dried soils (Cade-Menun and Preston 1996;

Turner et al. 2005). Although pretreatment is expected to impact P composition (Turner

et al. 2007b) the use of air drying was considered preferable given the ease of

application by collaborators and the belief that slow drying processes are a realistic

scenario within the environment. Phosphorus was extracted by shaking 1.00 g 0.01 g

of soil with 30 mL of solution containing 0.25 mol L-1 NaOH and 50 mmol L-1 EDTA in a

50-mL centrifuge tube for 4 h, after which samples were centrifuged at 7000 rpm

102









(maximum RCF -7000 g) (Sorvall RC6, SL600 Rotor; Thermo Fisher Scientific,

Waltham, MA) for 10 min. Subsamples of supernatant were drawn off for individual

sample determination of total P using a double acid (HNO3-H2S04) digest (Rowland

and Haygarth 1997) and molybdate colorimetry. Secondary subsamples were combined

on an equal volume basis within field replicates, spiked with an internal standard

methylenediphosphonic acid (MDP), frozen (-80C) and lyophilized to await solution 31P

NMR spectroscopy.

Lyophilized extracts (-300 mg) were redissolved in 3 mL of 1 mol L-1 NaOH and

0.1 mol L-1 EDTA within 15 mL centrifuge tubes before vortexing for 1 min. Samples

were subsequently filtered using a prewashed 0.2 pm syringe filter (GF-B) to remove

fine particles that may result in poor field homogeneity and thereby cause unacceptable

line broadening (see APPENDIX C1). Comparison of samples with and without filtration

suggests no significant error is associated with the filtration step (see APPENDIX C2).

Subsequently, 2.7 mL of redissolved filtered sample and 0.3 mL D20 (for signal lock)

were loaded into a 10 mm NMR tube for spectra acquisition. The use of an alkaline

matrix ensured a final pH >13. Although potentially resulting in the degradation of

certain phosphodiester functional groups (i.e. RNA and phosphatidyl choline) (Turner et

al. 2003d), this allows for consistent chemical shift (McDowell and Stewart 2005b) and

confidence in peak assignment when comparing to existing spectral libraries (Turner et

al. 2003d).

Spectra were acquired immediately using an Avance-500 (500.4 MHz 1H),

Magnex 11.8 Telsa/54 mm Bore magnet (AMRIS facility, McKinght Brain Institute

University of Florida) at a controlled 250 C. Acquisition parameters included use of a


103









simple zgig pulse profile, a calibrated 300 pulse, 2 s pulse delay and broad

heteronuclear decoupling (waltz 16). Between 30,000 and 50,000 scans were required

to achieve a reasonable signal to noise ratio, dependent upon sample P concentrations.

Spectra were acquired using sequential blocks of scans (10,000) with subsequent

combination of FIDs using Bruker proprietary software. Spectra were analyzed using

wxNUTS vr 1.0.1 for Microsoft Windows (Acorn NMR Inc. 2007). Initially spectra were

processed using 15 Hz line broadening, phased and corrected for baseline shift, and

referenced using internal standard MDP (6 = 17.46 ppm), established by comparison of

a redissolved soil extract with an external standard, 85% H3P04 (0 ppm) (see

APPENDIX C3). Spectra were integrated over set intervals, corresponding to

established bonding environments (Table 5-1). The region between 3 and 8 ppm was

further investigated using the deconvolution utility of wxNUTS software. A best fit

deconvolution of the spectra was acquired using 2 Hz line broadening (Figure 4-1).

Peak picking parameters were adjusted dependent upon signal to noise ratio of specific

samples, but ranged between 1 and 8% of maximum peak height and used 0.5 for the

root mean squared noise parameter. The region was split into orthophosphate and

phosphomonoesters (all other peaks determined by the algorithm in the region 3 to 8

ppm). Peak proportions from the deconvolution protocol were applied to the integral

determined in the 15 Hz spectra. A similar procedure was applied to the region -3 to -5

ppm (Figure 4-1), to differentiate pyrophosphate (-4.37 ppm) and higher order

polyphosphate groups (-3.91 and -4.03ppm) based upon comparison with standard

biogenic P compounds in the same matrix (see APPENDIX B4).


104









Data Analysis

Statistical analysis was carried out using JMP 8/0 (SAS Institute Inc. 2008) with SPSS

for windows version 17.0.0 statistical software (SPSS Inc., 2008) used for graphical

representation of correlations. Given non-normal data, co-linearity in fundamental

biogeochemical characteristics was tested by application of Spearman's rank

correlation. Exploration of emergent patterns was carried out by delineating wetland

sites into four fundamental groups, using hierarchical (Wards) classification of organic

matter content and pH. Organic matter and pH were selected given their lack of co-

linearity, and known influence of both parameters on biogeochemical P cycling.

Ordination of P composition diversity was performed using principal components

analysis (PCA) and compared with fundamental characteristics, including previous

defined groupings. This informed the development of multiple linear regression models

that sought to fit the ratio of phosphomonoesters to phosphodiesters found within

wetland soils. The ratio was normalized using natural log transformation and the

parameters used were standardized and centered using Equation 4-1, where ix equals

the average and 5x equals the standard deviation of the parameter x. The model was

evaluated using change in Akaike information Criterion (AIC), with values >2 considered

strong evidence of model improvement (Burnham and Anderson 2004).

(x-x,) / (2x,) (4-1)

Subsequently the influence of microbial biomass P, as a % of total P on P composition

was explored using Spearman's rank correlation against major biogenic P groups.


105









Results and Discussion


Wetlands Sampled

The 28 wetlands analyzed included two large wetland systems, the Changuinola

peat dome, Panama (Sites 20, 21, 22) and Houghton Lake treatment wetland, Michigan

(sites 4,5,6) in which three separate locations were treated as individual wetland sites.

This was considered appropriate given physical distances and distinct biogeochemical

conditions found between sites (Kadlec and Mitsch 2009; Sjogersten et al. 2010). Basic

biogeochemical characterization and alkaline extraction for P analysis was carried out

on nominally four surface samples (Table 4-2). In certain select cases, this protocol was

not achievable. Specifically, a lack of sample material from 8 mile lake Alaska (site 1)

necessitated the combination of samples on an equal mass basis and the application of

alkaline extraction to a single homogenized sample. Similarly, material received from

Dr. Sofie Sjogersten from site 26 situated near the town of Abisko, Sweden represented

a single previously homogenized air dried sample of surface (0-10 cm) soil considered

representative of the study site.

Study sites represent a vast range of climatic conditions, landscape positions,

dominant vegetation types as well as known differences in nutrient status (Table 4-1),

from tropical ombrotrophic peat domes (sites 20, 21, and 22), to high latitude acidic

based peatlands (sites 1, and 27) and fens (sites 28), to calcareous tropical/ subtropical

systems (sites 17,18,19,23,24 and 29), temperate fens (sites 3,15, and 16) and Carolina

bays (sites 7-14). Study sites included both 'pristine' systems and those severely

impacted by up to 30 y (Kadlec and Mitsch 2009) of nutrient enrichment (sites 4, 5, and

6). In addition, the study included a number of often understudied wetland types, such

as wet tundra (site 25) and high-altitude Paramo (site 2).

106









Biogeochemical Characteristics of Wetlands Sampled

Characterization showed a high degree of variability among wetland sites (Table

4-2). Loss on ignition, an often applied estimate of organic matter, ranged from 90 mg g

1 in a highly mineral Carolina Bay, within the Francis Marion National Forest (site 9) to

estimated fully organic matter, peat-based systems of high-latitude wet tundra (site 26)

and ombrotrophic bogs (site 27). As expected, total C showed a close colinearity with

estimated organic matter (Spearman's rho = 0.9035, p < 0.0001). Deviation from a

direct relationship between estimated organic matter content and total C determined by

elemental analysis was taken as evidence of carbonates (Wright et al. 2008) in

calcareous samples from Belize (sites 17,18,19) and south Florida (site 23,24).

The macronutrients N and P also showed a large range in concentration between

wetland systems, from 2 to 36 mg N g-1 and 51 to 3516 pg P g1, respectively. Plots of

average molar ratios (Table 4-2, Figure 4-2) highlight a distinction between P and N

relationships with C. As expected, given the biological derivation of N in the

environment (McGill and Cole 1981), there appears to be generalized coupling of total C

and N (Spearman's rho = 0.669, p < 0.0001). On closer inspection, this coupling

appears to break down in highly organic systems (total C > ~360 mg C g-1). When

considering only systems in which C < 360 mg C g-1, total C and N show a higher

degree of correlation (Spearman's rho 0.886, p < 0.0001) in a fashion similar to

terrestrial soils (Cleveland and Liptzin 2007). In contrast P appears to show no

correlation with total C (Spearman's rho 0.226, p = 0.247). Although such a decoupling

could reflect the importance of organic P cycling (Cleveland and Liptzin 2007) in

wetlands, in such a diverse range of sites it is more likely to reflect fundamental


107









differences in underlying site mineralogy and anthropogenic inputs of P independent of

C/N sources.

Of the total metals analyzed, Al ranged from 0.4 to 77.1 mg Al g-1 and showed

significant negative correlation with organic matter (Spearman's rho, -0.800, p <

0.0001). This is to be expected given the significant role aluminosilicates are likely to

play in wetlands with a large mineral fraction. Calcium contents ranged from

undetectable within Carolina bays of South Carolina to 334 mg Ca g-1 in the Belizean

calcareous fen (site 19). Such high concentrations in sites 19 and 17 (232 mg Ca g1)

probably reflect the presence of shell fragments and calcareous cyanobacterial mats

within surface samples collected from these coastal, low-salinity sites (Macek and

Rejmankova 2007). Even if these sites were considered outliers and excluded from

analysis, there was clear correlation between Ca concentration and site pH

(Spearman's rho = 0.6971, p < 0.0001). The redox-sensitive metal, Fe, showed no

apparent correlation with other basic biogeochemical characteristics, and ranged from

detection limit of 0.2 mg Fe g-1 in a large number of wetland sites to a maximum of 18.9

mg Fe g-1 within the heavily impacted portion of the Houghton Lake treatment wetland.

Given strong colinearity between several basic biogeochemical characteristics,

and the fact that Fe content was at or below detection limits for a number of sites, I

made the decision to define four broad groupings of wetland sites based upon the

simple hierarchical classification (Wards) of organic matter and pH (Figure 4-3). This

would then be applied to investigate emergent trends within P composition diversity.

The first group of 6 wetlands (group A) consists of highly organic (836 1000 mg g-

loss on ignition), acidic (pH 3.6-4.6) systems. Typified by Sphagnum sp.-dominated,


108









high-latitude bogs and mires (i.e. sites 1, 26 and 27), it also included tropical

ombrotrophic systems with a range of vegetation types (sites 20, 21, and 22). The

second grouping of eight wetlands (group B) represents those with an acidic (3.5 4.4)

pH and lower organic matter content (92 688 mg g-1 loss on ignition) than the truly

peat like, group A wetlands. This group included only Carolina Bay wetlands from the

Southeast Coastal Plain, USA. Although representing a range of vegetation types,

including both Cypress-dominated forested systems (site 8) and herbaceous open water

systems (site 13), their similar hydrogeomorphic setting within the landscape resulted in

broadly similar biogeochemical characteristics(De Steven and Toner 2004; Gaiser et al.

2001). The third group (C) represents 10 wetlands found to have an approximately

neutral pH (5.9 7.3) and high organic matter content (560 941 mg g-1 loss on

ignition). It included calcareous fens from England (site 3), New York (sites 15, 16),

Canada (site 28) and South Florida (sites 23, 24), plus wet Paramo of Ecuador (site 2)

and the Houghton lake treatment wetland (sites 5,6, and 7). The last group of wetlands

considered (D) represented those with an alkaline pH (pH 7.0 7.6) and relatively low

organic matter content (165 300 mg g-1 loss on ignition). This group was dominated by

calcareous fens (Macek and Rejmankova 2007) situated near the coast of northern

Belize (sites 17,18 and 19), but also included an arctic tundra system (site 25) that has

seen heavy grazing by migrating pink-footed geese (Wal et al. 2007).

Solution 31P Nuclear Magnetic Resonance Spectroscopy

Extraction of total phosphorus

Extraction of total P by the NaOH-EDTA method ranged from 25 to 84%. One

site (9) with a very low total P concentration (51 pg P g-) was calculated to have an

extraction efficiency of 125% and was therefore removed from all subsequent analysis

109









of P composition. Extraction efficiencies were found to vary significantly between

wetland groups (Kruskal Wallis test Chi2 = 8.23 d.f. 3, p < 0.05) reflecting the known

influence of calcareous soils on the standard NaOH-EDTA extraction (McDowell and

Stewart 2006; Turner et al. 2003a). Given the extraction explicitly targets organic P

(Turner et al. 2005), it is likely to reflect the biogenic P composition of soils, with the

residual considered alkaline-stable, inorganic and recalcitrant organic P (Cade-Menun

2005b; Turner et al. 2005). Therefore, the operationally defined 'residual- P' is

considered a distinct P type, and is included here when considering patterns in P

composition.

Phosphorus composition

Solution 31P NMR spectroscopy of alkaline extracts identified a diverse range of

biogenic P forms within wetland soils (see Table 4-3 and Figure 4-4, 5, 6, 7). Two

calcareous, low-P sites from Belize (sites 18 and 19) showed no evidence of biogenic P,

with only orthophosphate identified. The remaining sites contained phosphonates (0 to

44 pg P g-') phosphomonoesters (8 to 461 pg P g-') DNA (3 to 144 pg P g-') other

phosphodiesters (6 to 67 pg P g-') and polyphosphates (0 to 197 pg P g-1). Total

inorganic polyphosphates were further delineated into pyrophosphate and the mid and

terminal chain residues of longer polyphosphates (Table 4-4).

Given the range of total P between sites, analysis of composition diversity was

based upon composition, as a percentage of total P. Ordination using PCA, produced

two axes which together accounted for 64.1 % of the observed variance in P

composition. Examination of the PCA score plot with fundamental wetland groupings

superimposed (Figure 4-8) clearly demonstrated the separation of groups B and D, with

groups A and C (high organic matter) failing to show any clear distinction in P

110









composition. Examination of PCA loading plots and associated Eigen vales (Figure 4-8,

Table 4-5) showed separation of group D wetlands upon axis 1 to be the result of

distinct differences in the proportion of residual P as compared to the major biogenic P

groups (phosphomonoesters, DNA, phosphodiesters and pyrophosphate) identified,

while separation of group B wetlands, upon PCA axis 2 appeared to be a result of

increased prevalence of phosphonates and phosphomonoesters. Similar examination of

P composition with reference to Cowardin 'class' (Figure 4-9) and climatic zone data

(not shown) failed to show clear clustering, thereby suggesting that soil P composition is

dependent upon basic biogeochemical characteristics including, to some degree, both

the pH and organic matter content groupings used in this study.

As an extension of initial ordination analysis, the influence of basic

biogeochemical characteristics on select aspects of P composition were explored.

These included the ratio of phosphomonoester to phosphodiesters, the presence of

Inositol hexakisphosphate and the relationship between measures of microbial P and

biogenic P composition.

Ratio of phosphomonoesters to phosphodiesters in wetlands. The ratio of

phosphomonoesters to total alkaline stable phosphodiesters in the 28 palustrine

wetlands averaged 2.8, ranging from 0.9 in an ombrotropic Canadian bog (site 27) to

10.6 in Norweigan wet tundra (site 25). The influence of basic biogeochemical

characteristics on this ratio were explored by multiple linear regression. Given the

apparent influence of organic matter (or its reciprocal mineral content) in delineating

high phosphomonoester-containing group B wetlands, this was selected as a model

basis. Additionally the C:P ratio, a measure of potential P limitation, was tested, given


111









the known influence of P availability on microbial eco-physiological responses

associated with organic P cycling (Corstanje and Reddy 2006; Corstanje et al. 2007).

The C:P ratio also fulfilled the required assumption of being independent from organic

matter. Comparison of AIC values (Table 4-6) gave strong evidence (Burnham and

Anderson 2004) for a loss of information when considering the reduced model [f(LOI)]

over the optimal model [f(LOI + C:P). This would suggest both Loss on ignition and C:P

play a significant role in determining the ratio of phosphomonoesters to phosphodiesters

within wetland soils. Parameter estimates from the optimal model (Table 4-7) show that

the ratio of phosphomonoesters to phosphodiesters decreases with both increasing

organic matter and an increasing C:P ratio. Therefore, increased mineral content and

reduced P limitation both result in an elevated proportion of P to be found as

phosphomonoesters instead of phosphodiesters.

Presence of Inositol hexakisphosphate. Spectral deconvolution of the 8 to 3

ppm region revealed that in some samples a substantial portion of phosphomonoesters

corresponded with known peak assignment of higher order Inositol phosphates (Turner

et al. 2003f; Turner and Richardson 2004). The use of a standard preparation and

spectra acquisition protocol in conjunction with a stable internal standard (MDP)

provided confidence in the assignments of both myo- and scyllo-IP6 (Figure 4-10, Table

4-8). Inositol groups appeared particularly prevalent in group B wetlands, with myo-IP6

accounting for between 16 and 39% of total phosphomonoesters and scyllo-IP6

constituting between 8 and 20% of total phosphomonoesters. Two group B wetlands

(site 7 and 11) also showed evidence of phosphomonoester peaks (6.68 and 6.88 ppm)


112









suspected of being lower-order inositol derivatives or as yet unidentified isomers of IP6

(Turner 2007).

Determination of IP6 within wetlands other than group B systems proved

problematic given the degree of peak overlap within the phosphomonoester region.

However, peaks coincident with that attributed to scyllo-IP6 in group B wetlands (4.241 +

0.023) were found in sites 1, 2, 6, 15, 16, and 25. Given the suspected role of myo-IP6

as a biosynthetic precursor to scy//o-IP6 within the soil matrix (L'Annunziata 2007), it is

likely that both isomers may be present. Further analysis utilizing hypobromination to

hydrolyze non-inositol phosphomonoesters (Irving and Cosgrove 1981) would be

required to both confirm the presence of scyllo-IP6 and reveal the presence of other

isomers.

Microbial biomass. The potential relationship between microbial P, determined

by anion exchange membranes, and P composition was explored by application of

Spearman's rank correlation (Table 4-9). This highlighted both DNA and long-chain

polyphosphates (as percentages of total P), to show significant positive correlations with

microbial P (Figure 4-11). A significant positive correlation between microbial P and

DNA (Spearman's rho = 0.61 p = 0.002) would be expected given a standard microbial

composition. Yet the broad range of microbial biomass P (inter-quartile range equals,

2.4 to 19.0% of total P) as compared to that of DNA (inter-quartile range 3.8 to 9.6% of

total P), would suggest confounding factors that influence the proportion of DNA found

within soils, including altered microbial P composition between systems (Makarov et al.

2005) and the influence of extracellular stabilization of DNA (Celi and Barberis 2005b;

Niemeyer and Gessler 2002). The highly significant correlation between microbial P and


113









long chain polyphosphates (Spearman's rho = 0.78, p < 0.0001) reflects their biological

synthesis (Harold 1966), yet there also appears to be a distinction with a marked

increase in polyphosphate composition in systems with a microbial P >15% of total P,

possibly reflecting the role of polyphosphates as a metabolic stress response or

strategy under competitive environmental conditions (Kulaev and Kulakovskaya 2000;

Seufferheld et al. 2008). That said, the known interaction anion exchange membranes

with certain biogenic P forms (Cheesman et al. 2010b) necessitates caution when

interpreting causation. The strong positive correlation seen between total

polyphosphates (pyrophosphate + longer chain length polyphosphates) and microbial P,

(Spearman's rho = 0.8103, p < 0.0001) may reflect the fact that operationally defined

microbial P is in a large part due to recovery of polyphosphates.

Discussion

This unique data set demonstrates the diverse range of biogenic P found within

wetland soils, while also providing the basis to begin to explore mechanistic drivers

behind the P composition of wetland soils. In addition to confirming the nature of P

within calcareous palustrine systems (Turner and Newman 2005; Turner et al. 2006a)

this work also demands caution when extrapolating isolated observations to 'wetland

systems' in general, highlighting differences between broad wetland types. The

investigation of emergent patterns in P composition identified here based upon basic

biogeochemical characteristics (organic matter, pH, and nutrient availability) will form

the focus of subsequent chapters.

Studies within terrestrial systems have attributed a positive correlation between

the proportion of P found as phosphodiesters and annual precipitation to their increased

recalcitrance under 'wetter' conditions (Condron et al. 1990a; Sumann et al. 1998; Tate

114









and Newman 1982). It has been suggested that a similar mechanism of reduced

turnover under anaerobic conditions could account for the increased prevalence of

phosphodiesters in palustrine wetlands studied to date (Turner and Newman 2005). In

addition, the phosphomonoester IP6, often a major component of P in terrestrial soils

(Cosgrove 1966; Murphy et al. 2009; Turner et al. 2003f; Turner et al. 2002b), had

thought to be absent from wetlands (Turner and Newman 2005; Turner et al. 2006a),

with evidence suggesting rapid degradation under anaerobic conditions, typical of

wetland soils, (Suzumura and Kamatani 1995a, b). As evident from this study, and from

other recent research on estuarine (Turner and Weckstr6m 2009), lacustrine (Zhang et

al. 2009b), and riverine (McDowell 2009) systems (presumed to experience anaerobic

conditions) IP6 may actually constitute a substantial proportion of biogenic P in

wetlands. It is therefore likely that its presence, or absence, is likely to be critical to

observed trends in phophmonoester to phosphodiester ratios. In terrestrial systems,

calcite, Fe/AI oxides, clay and organic matter have all been shown to increase soils IP6

sorbtion capacity (Celi and Barberis 2007), yet in wetlands it is likely these factors are

further impacted by ambient physicochemical conditions (such as anaerobosis). As an

example, in the comparison of ferrous and ferric salts of IP6, the reduced Fe(ll)-IP6

complex was shown to have a lower rate of enzymatic dephosphorylation than Fe(lll)-

IP6 (Heighton et al. 2008). The complex interactions between site properties and IP6

stabilization are likely to be profound and are explored further in this dissertation

(CHAPTER 6). An alternative to the differential stability of IP6 under ambient wetland

conditions, may be difference in the inputs or in-situ IP6 synthesis between wetlands.

Inositol phosphates have been found in a range of biological materials (Michell 2008)


115









with higher order forms such as myo-IP6 found to represent a substantial proportion of

total P in reproductive structures such as pollen (Jackson et al. 1982) and certain seeds

and fruiting bodies (Lott et al. 2000). Wetlands in group B of this study represent

Carolina Bays within the Savannah River Site and Francis Marion National Forest both

situated within the South-east Coastal Plain. As such, uplands surrounding the study

sites contain significant coniferous forests, that may represent a significant direct

(though deposition) or indirect (through runoff) source of myo-IP6 (Jackson and

Linskens 1982). In addition, the presence of significant concentrations of otherwise rare

stereoisomers (i.e. scyllo) within group B wetlands suggests additional sources and

processing of inositol derivatives within currently unidentified soil microbial components

(Turner 2007; Turner and Richardson 2004). It is clear that further work is needed to

elucidate the role of differential IP6 sources and stabilization on the P composition of

wetland soils.

Polyphosphates represent more than just a microbial response to excessive P

loading, so-called luxury uptake (Khoshmanesh et al. 2002), and have been implicated

in a diverse range of metabolic processes (Kornberg 1995; Kornberg et al. 1999; Kulaev

and Kulakovskaya 2000). This study identified substantial polyphosphate pools within a

diverse range of wetlands (although predominantly acidic high organic-matter systems),

including samples from a low-P tropical ombrotrophic peat dome considered (sites

20,21 and 22). Similarly, recent evidence of polyphosphates within unimpacted Carolina

Bays (Sundareshwar et al. 2009) and oligotrophic Swedish lake sediments (Ahlgren et

al. 2006a) demands further investigation into the role of polyphosphates within

palustrine systems. It has been suggested, that the relationship between terminal to mid


116









chain phosphate residues can be used to estimate polyphosphate chain length (see

Equation 4-2 where Ot + Om equate to the integrals of terminal and mid chain

phosphates, respectively).

Chain length = 2(Ot + Om) / Ot Equ. 4-2.

Given the potential for soil-mediated hydrolysis (Hupfer and Gachter 1995) and the

hypochromic effect seen in the 31P NMR spectra of model linear polyphosphates

(Krupyanko et al. 1998), such a direct calculation may be erroneous. Yet it is worth

noting that there appears to be a greater difference between mid-chain and terminal

residues in acidic systems, suggesting greater chain length. Further work would be

needed to investigate if this truly represents a range of polyphosphate chain lengths

across environmental gradients, or an artifact due to extraction and 31P NMR

spectroscopy method used.

Conclusions

This study represents the most diverse range of wetland sites to which solution 31P

NMR spectroscopy has ever been applied. In studying the forms of biogenic P found

within such a broad range of wetlands I am able to identify potential patterns as well as

to explore the validity of extrapolating patterns seen within the few wetlands studied to

date. It is apparent that biogenic P within wetland systems shows a great deal of

variation, and extrapolation from previous work (Turner and Newman 2005) to all

wetlands should be with caution. When exploring diversity of biogenic P within wetland

soils, it is apparent that basic characteristics including pH and organic matter content

play a significant role in determining the nature of functional P forms found, but that

climatic conditions and wetland vegetation types appear to have limited impact. As an

exploratory study this work highlights a number of potential mechanistic drivers,

117









including the role of mineral content and P availability, in determining the forms of

biogenic P found in wetlands. These mechanistic drivers are the focus of subsequent

chapters.


118











Table 4-1. Wetland study sites sampled for characterization of biogenic phosphorus composition.


Wetland

1 8 mile

2 Laguna Papallacta,


3 Wicken Fen


4 Houghton lake
(CT350)
5 Houghton lake
(550C)
6 Houghton lake (P)

7 Francis Marion
National Forest
(FMNF) Bay 1
8 Francis Marion
National Forest
(FMNF) Bay 2
9 Francis Marion
National Forest
(FMNF) Bay 3
10 Francis Marion
National Forest
(FMNF) Bay 4
11 Savannah River Site
(SRS) Bay 1
12 Savannah River Site
(SRS) Bay 2

13 Savannah River Site
(SRS) Bay 3
14 Svannah River Site
(SRS) Bay 4


Location

Al, USA

Ecuador


UK


MI, USA

MI, USA

MI, USA

SC,USA


SC,USA


SC,USA


SC,USA


SC,USA

SC,USA


SC,USA

SC,USA


Wetland
Typet
Bog

Cushion
forming,
Paramo
Fen


Treatment
wetland
Treatment
wetland
Treatment
wetland
Carolina Bay


Carolina Bay


Carolina Bay


Carolina Bay


Carolina Bay

Carolina Bay


Carolina Bay

Carolina Bay


Vegetation Type*

Persistent emergent
(Moss)
Persistent emergent
(Herbaceous)

Persistent emergent
(Herbaceous)

Persistent emergent
(Herbaceous)
Persistent emergent
(Herbaceous)
Persistent emergent
(Herbaceous)
Persistent emergent
(Forested)

Persistent emergent
(Forested)

Persistent emergent
(Herbaceous)

Persistent emergent
(Forested)

Persistent emergent
(forested/herbaceous)
Persistent emergent
(Herbaceous)

Open water
/herbaceous
Persistent emergent
(Forested)


Dominant species
Sphagnum, Carex sp.

Self emergent succulent species Distichia
muscoides, Spagnum

Cladium mariscus


Typha sp.

Cyperacea sp. Typha

Typha.


Acer rubrum (var. trilobum), Nyssa biflora and Nyssa
aquatica, Lyonia lucida, flex myrtifolia

Taxodium ascendans, Nyssa biflora, Lyonia lucida,
Carex striata, Woodwardia virginica

Ilex glabra, Iris tridentata, Amphicarpum
muhlenbergianum, Eleocharis spp., Melanocarpa,
tricostata, and Lachnanthes caroliniana
Nyssa biflora, Taxodium ascendans, Acer rubrum,
Lyonia lucida, Cyrilla racemiflora, Pinus taeda

Panicum hemitomon, Nyssa biflora, Cephalanthus
occidentalis, Utricularia spp., Sphagnum spp.,
Pontederia cordata var. lancifolia
Panicum hemitomon, Sphagnum spp., Pontederia
cordata var. lancifolia, Juncus canadensis,
Cephalanthus occidentalis, Acer rubrum (var.
trilobum)
Nymphaea odorata, Panicum hemitomon, Utricularia
spp., Leersia hexandra, Eleocharis melanocarpa
Liquidambar styraciflua, Acer rubrum (var. trilobum),
Nyssa biflora, Taxodium ascendans, Smilax
rotundifolia


Potential
impacts1


Cattle Grazing


Sedge
Harvesting
since 1419
Intermediate P
loading
Low P loading

High P loading


Periodic
during


119










Table 4-1. Continued
Wetland
15 Larry Fen

16 Fish Fen

17 Hidden


18 Quiet


19 Doubloon


20 Changuinola Site 1

21 Changuinola Site 2

22 Changuinola Site 3

23 WCA 3A

24 Everglades National
Park
25 NyAlesund

26 Stordalen

27 Bog 8

28 Fen 1


Location
NY, USA

NY, USA

Belize


Belize


Belize


Panama

Panama

Panama

Fl, USA

Fl, USA

Spitsbergen,
Norway
Abisko,
Sweden
Canada

Canada


Wetland
Typet


Vegetation Type*


Rich Fen Persistent emergent
(Herbaceous)
Rich Fen Persistent emergent
(Herbaceous)
Oligotrophic Persistent emergent
Sumpland (Herbaceous/
Cyanobacteria)
Sumpland Persistent emergent
(Herbaceous/
Cyanobacteria)
Sumpland Persistent emergent
(Herbaceous/
Cyanobacteria)
Tropical peat Persistent emergent
dome (Forested)
Tropical peat Persistent emergent
dome (Forested)
Tropical peat Persistent emergent
dome (Forested)
Calcareous Open water


Fen
Calcareous
Fen
Wet tundra

Mire

Ombrotrophic
Bog
Fen


Persistent emergent
(Herbaceous)
Persistent emergent
(Moss)
Persistent emergent
(Moss)
Persistent emergent
(Moss)
Persistent emergent
(Herbaceous)


Dominant species

Carex sp. Campylium stellatum

Typha angustifolia. Carex sp. Campylium stellatum,
Sphagnum spp.Calliergonella cuspidata
Eleocharis cellulosa, Cyanobacteria spp.


Eleocharis cellulosa, Cyanobacteria spp.


Eleocharis cellulosa, Cyanobacteria spp.


Potential
impacts1




salt intrusion


salt intrusion


salt intrusion


Raphia tadeiga

Campnosperma panamensis, Cassipourea elliptica
(Sw.) Poir, Drypetes standleyi G.L. Webster
Campnosperma panamensis, Cyrilla racemiflora,
sawgrass
Nymphaea spp. Utricularia

Cladium jamaicense

Calliergon richardsoni, Poa arctica, Dupotia species Geese grazing

Spagnum fuscum, Betula nana, Rubus
chamaemorus,Vaccinium vitis ideae, Empetrum
nigrum
Sphagnum fuscum, graminoids, Lichens

Carex sp.


t = common wetland description
: = vegetation descriptor based upon Cowardin (1979) classification
= dominant vegetation species noted in the field
= potential external impacts noted that may effect P cycling


120












Table 4-2. Soil biogeochemical properties in studied wetland systems.
Organic matter Total P Total C Total N Total Ca
pH mg g-1 pg g-1 mg g1
Group
A


15 2.5 3.0 0.5 1083
0.2 0.4 4596
0.3 0.0 0.5 0.2 3219
6.2 0.7 3.0 0.1 1125
3.7 0.6 1.3 0.1 1498
1.5 0.4 1.2 0.2 1928


2242
4596
1083


1
26*
27
20
21
22


Average
Max
Min

Group
B
7
12
10
8
9
11
14
13


Average
Max
Min


25.8 2.0 728
15.7 4.0 745
34.7 7.6 982
8.2 4.0 1275
2.9 1.5 2547
72.9 8.2 283
77.1 3.3 285
26.2 6.7 815


958
2547
283


Total Fe


Total Al


986
238
356
1124
852
579


Molar Ratio


0.5


412 4
424
436 6
489 2
485 5
424 13


12 1
6
8 0
28 0
25 1
22 1


689
1124
238


3.3 1.6
3.7 0.5
4.7 1.1
2.1 0.5
0.7 0.3
5.6 0.8
6.2 1.9
2.9 0.7


919 10
1000
974 8
878 10
927 9
836 47


922
1000
836


497 168
546 102
478 177
688 102
92 39
239 107
226 49
246 184


377
688
92


925
1056
752
750
51
918
918
347


260 43
307 33
264 53
377 65
44 8
105 29
94 14
117 44


28 3
56
51 9
55 3
66 6
84 5


57
84
28


35 6
44 6
45 8
56 4
98 25
21 4
19 8
57 5


15 2
21 2
15 3
19 3
2 0
9 2
7 1
9 3


715
1056
51


121












Table 4-2 Continued
Wetland Organic matter

pH mg g-1
Group C

2 6.7 835 24
16 7 854 21
3 7.3 783 154
4 7.2 935 5
6 6.4 695 57
15 7 560 54
5 6.1 941 4
23 5.9 929 11
24 6.1 896 13
28 6.3 891 8


Average 6.6 832
Max 7.3 941
Min 5.9 560


Total P

pg g-1 mg g


Total C


Total N


Total Ca
C:P


875
1184
937
1439
3516
1184
982
277
310
679


Total Fe


Total Al


1.2
1.6
1.5
0.2
5.4
1.0
0.8
0.0
0.0
+ 0.0


1212
929
976
812
260
591
1201
4170
3800
1546


1550
4170
260


Molar Ratio
N:P


230
120
60
31
15
75
87
346
845
20


Group D
17 7.3 245 20 192 32 162 7
18 7.5 300 7 116 16 70 3
25 7.0 286 330 1513 387 132 29
19 7.6 165 29 126 15 153 0


Average 7.4 249 487 129
Max 7.6 300 1513 162
Min 7.0 165 116 70
t = organic matter estimated from loss on ignition (550C, 4 h)


N:P


S4
S5
1
1
1
S5
S3
29
61
+ 15


0.0
0.0
0.0
+ 0.0


51.1
0.8
6.6
+ 15.4


30
19
3
17


122










Table 4-3. Phosphorus composition of surface soils as determined by solution 31P NMR spectroscopy
NaOH- TP* Phosph-P" Ortho-P' Mono-P" DNA Phospholipids norpghacs Organic P Mono:
Polphosphates Organic P Dies"


Group A
1 758 (77)
26 190 (80)
27 138 (39)
20 516 (46) 20
21 320 (38) 8
22 261 (45)

Average 364 (54) 14
Max 758 (80) 20
Min 138 (38) 8

Group B
7 637 (69) 29
12 722 (68) 44
10 497 (66) 18
8 476 (63) 16
9t 63 (125)
11 598 (65) 14
14 534 (58) 12
13 219 (63) 4

Average 526 (65) 20
Max 722 (69) 44
Min 219 (58) 4


Group C
2 609 (70)


201 (20) 337 (34)
70 (29) 44 (18)
36 (10) 31 (9)
(2) 171 (15) 162 (14)
(1) 95 (11) 55 (6)
54 (9) 68 (12)

(2) 105 (16) 116 (16)
(2) 201 (29) 337 (34)
(1) 36 (9) 31 (6)


(3) 137 (15) 368 (40)
(4) 179 (17) 390 (37)
(2) 120 (16) 252 (33)
(2) 138 (18) 192 (26)
14 (27) 38 (76)
(2) 128 (14) 408 (44)
(1) 267 (29) 189 (21)
(1) 56 (16) 110 (32)

(2) 146 (18) 273 (33)
(4) 267 (29) 408 (44)
(1) 56 (14) 110 (21)


244 (28) 176 (20)


135 (14)
25 (11)
23 (7)
35 (3)
31 (4)
57 (10)

51 (8)
135 (14)
23 (3)



56 (6)
65 (6)
57 (8)
67 (9)
6 (12)
32 (3)
47 (5)
25 (7)

50 (6)
67 (9)
25 (3)


43 (4) 514 (52)
41 (17) 79 (33)
38 (11) 64 (18)
123 (11) 222 (20)
118 (14) 107 (13)
79 (14) 128 (22)

74 (12) 186 (26)
123 (17) 514 (52)
38 (4) 64 (13)



30 (3) 470 (51)
25 (2) 518 (49)
26 (3) 351 (47)
50 (7) 288 (38)
5 (10) 45 (88)
4 (0) 466 (51)
9 (1) 258 (28)
9 (3) 154 (44)


(3) 358 (44)
(7) 518 (51)
(1) 154 (28)


76 (9) 31 (4) 81 (9) 284 (32)


123










Table 4-3. Continued (Group C)


NaOH-TP*


16 595 (50)
3 789 (84)
4 909 (63)
6 2569 (73)
15 753 (64)
5 593 (60)
23 102 (37)
24 130 (42)
28 283 (42)

Average 733 (59)
Max 2569 (84)
Min 102 (37)

Group D
17 47 (25)
18 53 (46)
25 534 (35)
19 33 (26)


Phosph-P0


Ortho-P'


Mono-P"


131 (11) 270 (23) 106
167 (18) 344 (37) 144


Total inorganic Mono:
DNA Phospholipids Polyphosphates Organic P Dies
(9) 28 (2) 60 (5) 404 (34) 2
(15) 62 (7) 73 (8) 549 (59) 1.7


295 (20) 317 (22) 142 (10) 67 (5) 88 (6) 526 (37) 1.5
1759 (50) 407 (12) 141 (4) 64 (2) 197 (6) 612 (17) 2
118 (10) 461 (39) 88 (7) 37 (3) 50 (4) 586 (49) 3.7
225 (23) 170 (17) 96 (10) 30 (3) 72 (7) 296 (30) 1.3
38 (14) 28 (10) 18 (7) 6 (2) 11 (4) 53 (19) 1.2
46 (15) 38 (12) 27 (9) 9 (3) 10 (3) 75 (24) 1.1
54 (8) 77 (11) 57 (8) 22 (3) 74 (11) 156 (23) 1


308 (20) 229 (20) 90 (9) 36 (3) 72 (6) 354 (32 2
1759 (50) 461 (39) 144 (15) 67 (7) 197 (11) 612 (59 4
38 (8) 28 (10) 18 (4) 6 (2) 10 (3) 53 (17) 1


36 (19)
53 (46)


8 (4)


3 (2)


292 (19) 221 (15) 11 (1) 10 (1)
33 (26)


11 (6) 2.3

242 (16) 10.6


Average 167 (33) 104 (28) 57 (5) 4 (1) 3 (0)
Max 534 (46) 292 (46) 221 (15) 11 (2) 10 (1)
Min 33 (25) 33 (19) 0 (0) 0 (0) 0 (0)
t suspected error associated with determination of total P composition, site removed from subsequent analysis.
$ Total P recovered by alkaline extraction
Total phosphonates
Total orthophosphate
tt Total phosphomonoesters
$$ Ratio of total phosphomonoesters: total phosphodiesters


124


127 (11)
242 (16)
11 (6)










Table 4-4. Inorganic polyphosphates as determined by solution 31P NMR spectroscopy
of wetland soils. Values represent total inorganic polyphosphates delineated
into pyrophosphate, and the terminal (TR) and mid-chain (MR) of long chain
(n>3) polyphosphates.
Pyrophosphate Long chain polyphosphate
Ig g-1 % total P pg g-1 % of total P
TR MR
Group A
1 7.5 0.8 trace 35.5 3.6
26 4.9 2.1 7.9 28.0 15.1
27 4.3 1.2 6.2 27.3 9.4
20 12.6 1.1 trace 110.1 9.8
21 19.5 2.3 trace 98.9 11.6
22 7.5 1.3 trace 71.6 12.4

Group B
7 9.1 1.0 6.3 14.2 2.2
12 5.6 0.5 5.6 13.6 1.8
10 9.7 1.3 6.2 9.9 2.2
8 5.4 0.7 7.0 37.6 6.0
9 4.9 9.8
11 4.3 0.5
14 8.8 1.0
13 9.2 2.6

Group C
2 30.9 3.5 19.5 31.1 5.8
16 46.2 3.9 6.3 7.8 1.2
3 41.6 4.4 25.9 5.9 3.4
4 40.2 2.8 17.5 30.6 3.3
6 136.3 3.9 30.1 31.1 1.7
15 32.3 2.7 10.6 6.8 1.5
5 24.6 2.5 12.1 35.2 4.8
23 6.6 2.4 1.7 3.1 1.7
24 9.6 3.1
28 10.3 1.5 23.6 39.8 9.3

Group D
17 -
18 -
25 -
19 -


125









Table 4-5. Eigen values of principal components determined on PCA applied to
phosphorus composition within wetland soils. Principal components, 1 and 2
account for 41.6 and 22.5% of total P composition diversity respectively.
P form PC 1 PC 2
Phosphonates 0.033965 -0.66096
Orthophosphate -0.102 0.09497
Phosphomonoesters 0.385128 -0.49027
DNA 0.489812 0.164466
Other phosphodiesters 0.502311 0.150683
Pyrophosphate 0.364425 0.359201
Long chain polyphosphates 0.088475 0.305069
Residual -0.45525 0.204691

Table 4-6. Multiple linear regression models used to predict the ratio of
phosphomonoesters to phosphodiesters in wetland surface soils.
Model DFe Adj R2 A
1 f(cLOI +,C:P) 23 0.6650 0
2 f(cLOI) 24 0.6108 2.197
3 f(cC:P) 24 0.2885 17.88

Table 4-7. Parameter estimates for optimal model for predicting ratio of
phosphomonoetsrs to phosphodiesters. Parameters centered and
standardized and response variable normalized by natural logarithm before
model run.
Term Estimate Std error t ratio p value
Intercept 0.8775483 0.078683 11.15 < 0.0001
cLOI -0.962098 0.181879 -5.29 < 0.0001
cC:P -0.384798 0.174035 -2.21 0.0372

Table 4-8. Inositol hexakisphosphates as determined within group B wetlands.
Concentrations pg g-1 (% of phosphomonoesters)
Site myo-IP6 scyllo-IP6
7 96.1 (26.1) 56.1 (15.2)
8 63.8 (16.3) 48.1 (12.3)
9 trace 36.3 (14.4)
10 40.7 (21.2) 37.9 (19.7)
11 14.8 (38.9) 2.5 (6.7)
12 131.6 (32.2) 55.4 (13.6)
13 50.1 (26.5) 19.1 (10.1)
14 20.6 (18.7) 7.2 (6.6)


126









Table 4-9. Correlation between microbial biomass phosphorus (% of total phosphorus)
and phosphorus forms determined by solution 31P NMR spectroscopy (% of
total phosphorus).


Phosphonate
Orthophosphate
Phosphomonoesters
DNA
Other phosphodiesters
Pyrophosphate
Long chain Polyphosphate
significant at the 0.05 level
**significant at the 0.01 level
*** significant at the 0.001 level


Spearman rho
correlation
-0.2359
-4426
-0.1348
0.6123
0.3037
0.3734
0.7756


p

0.2670
0.0303
0.5300
0.0015
0.1490
0.0723
< 0.0001


127










Pyrophosphate


Polyphosphate group
% ndecon


----1 ---P ^ ".


-3.0 -3.5 -4.0 -4.5






II
p
s
I

J'a-


7


of spectral region -3 to -5 ppm with
hosphateand polyphosphate
Identified by spectral
evolution, overlaid in red.











of spectral region 8 to 3 ppm with
homonoesters identified by
al deconvolution, overlaid in red.


6 5 4 I I
6 5 4


I I I r I I I I I I I I I I I I I I '
20 10 0 -10 -20
Chemical shift (ppm)


Figure 4-1. Solution 31P NMR spectra of surface soils collected from a Michigan treatment wetland (Site # 6). Detail
showing spectral deconvolution used to identify phosphomonoesters and inorganic polyphosphates.


128


~Y~dl~LI*IC*LL-------~rur~*r~*~IC1


-- ----


LU


14










Average molar ratio observed in terrestrial soils


All samples n = 28
rho = 0.669; p<0.0001 /
0 to30 mmol C g1, n = 14
rho = 0.886; p<0.0001)
o 2- o o
E o
0


t
I I
A0



0 10 20 30 40 50


12- Group A
120- Group B /
Group C
Group D /
100,
S 0/ Site 6: highly impacted
SHoughton Lake
7o 80-
E

,. 60-

2 40 o o

0
20- 00


0 10 20 30 40 50
Total C (mmol g1)

Figure 4-2. Average nutrient concentrations in wetland surface soils. Individual
wetlands averages and partial least squared regressions plotted alongside
average relationship observed in terrestrial systems. Molar concentrations of
C and N show similar significant positive coupling (p < 0.0001) improved
when considering just 'low' C (< 30 mmol C g1) sites.


129












Category Wetland


GroupA
High Organic, low pH


Group B
Low Organic, low pH








Group C
High Organic, high pH


Group D
Low Organic, high pH


8 mile,AI
Stordalen, Sweden
F Bog 8, Canada
Changuinola Site 1, Panama
Changuinola Site 3,Panama
Changuinola Site 2, Panama
FMNF Bay 1, SC
SRS Bay 2, SC
FMNF Bay 4, SC
FMNF Bay 2, SC
FMNF Bay 3, SC
SRS Bay 1, SC
SRS Bay 4, SC
SRS Bay 3, SC
Laguna Papallacta, Equador
Fish Fen, NY
Wicken Fen, England
Houghton Lake (CT350), MI
Houghton Lake (P),MI
Larry Fen, NY
Houghton Lake (550C), MI
Everglades slough WCA-3, FI
Everglades Nat Park, FI
Fen 1, Canada
Hidden, Belize
Quite, Belize
NyAlesund, Norway
Doubloon Belize


Figure 4-3. Categorization of wetland sites based upon Wards hierarchical
classification of pH and organic matter (estimated by loss on ignition).


130


Category


Wetland









































20 10 0 -10 -20
Chemical shift (ppm)

Figure 4-4. Solution 31P NMR spectra of biogenic P composition within group A
wetlands (high organic low pH). Spectra acquired using an Avance-500
(500.4 MHz 1 H), Magnex 11.8 Tesla/54 mm Bore, at pH > 13 using a simple
zgig pulse program and calibrated 30 pulse angle. Between 30-50,000 scans
were combined and presented here using 15 Hz line broadening scaled and
referenced to internal standard methylenediphosphonic acid (6 = 17.46 ppm).


131


^M















































20 10 0 -10 -20
Chemical shift (ppm)

Figure 4-5. Solution 31P NMR spectra of biogenic P composition within group B
wetlands (low organic low pH). Spectra acquired using an Avance-500 (500.4
MHz 1H), Magnex 11.8 Tesla/54 mm Bore, at pH > 13 using a simple zgig
pulse program and calibrated 30 pulse angle. Between 30-50,000 scans
were combined and presented here using 15 Hz line broadening scaled and
referenced to internal standard methylenediphosphonic acid (6 = 17.46 ppm).


132







































I I I I I I I I 11 I
20 10 0 -10 -20
Chemical shift (ppm)


Figure 4-6. Solution 31P NMR spectra of biogenic P composition within group C
wetlands (high organic high pH). Spectra acquired using an Avance-500
(500.4 MHz 1 H), Magnex 11.8 Tesla/54 mm Bore, at pH > 13 using a simple
zgig pulse program and calibrated 30 pulse angle. Between 30-50,000 scans
were combined and presented here using 15 Hz line broadening scaled and
referenced to internal standard methylenediphosphonic acid (6 = 17.46 ppm).
Note orthophosphate spike curtailed in wetlands 4 and 5.


133








































20 10 0 -10 -20
Chemical shift (ppm)


Figure 4-7. Solution 31P NMR spectra of biogenic P composition within group D
wetlands (low organic high pH). Spectra acquired using an Avance-500
(500.4 MHz 1 H), Magnex 11.8 Tesla/54 mm Bore, at pH > 13 using a simple
zgig pulse program and calibrated 30 pulse angle. Between 30-50,000 scans
were combined and presented here using 15 Hz line broadening scaled and
referenced to internal standard methylenediphosphonic acid (6 = 17.46 ppm).


134


'1~GI(Ylk~Y~





~J~LL~LL~I-~~lll~rU L~LY-~L I-LLl~rh~Lu~l
















25
26V127 025
20 22 6 27 4
16 15
A24 019 14< 01
0130
:13 D8
09
0 10
0 7
0 11


U~ ~ C~) C~4 ~- 0 ~- C~4 C~)


S 0 PC -
PC 1


O Group A
ol Group B
' Group C
A Group D


,I n


i = Residual P (%)
ii = Orthophosphate (%)
iii = Polyphosphate (%)
iv = Pyrophosphate (%)
v = DNA (%)
vi = Other phosphodiesters (%)
vii = Phosphomonoesters (%)
viii = Phosphonates (%)


Figure 4-8. Principal component analysis of P composition within wetland soils as determined by solution 31P NMR
spectroscopy. A) Score plot of PCA 1 and PCA 2 accounting for 41.7 and 22.5% of variance respectively,
superimposed using grouping of wetlands based upon organic matter and pH. B) Loading plot visualizing role of
P composition is distinguishing variation between sites. Note site nine removed prior to analysis, due to
potential error in proportion calculation.


135




























S CPC C1 C
PC1


0 Persistent emergent, Forested
SPersistent emergent, Herbaceous
0 Persistent emergent, Moss
0 Open Water


le 1O


Figure 4-9. Principal component analysis of P composition within wetland soils as determined by solution 31P NMR
analysis. Wetland grouping as based Cowardin et al. (1979) major vegetation classes. Note site nine removed
prior to analysis


136


5

4

3

2

1


-1

-2

-3

-4

-5


- ---- ---.. .-- .O-- ---- - -


0

O*
0.


























D E

i JF


*


Site 14



Site 13


Site
Site 12 14
13
12
Site 11 1

10

Site 10
8
7


B C D E F G


6.189
6.208
6.195
6.198
6.206
6.206
6.206
6.197


5.068 4.705 4.570 4.215
5.087 4.726 4.594 4.235
5.085 4.725 4.592 4.24
5.087 4.725 4.593 4.24
5.109 4.742 4.615 4.254
4.205
5.114 4.758 4.627 4.264
5.121 4 764 4.633 4274


Site 9



Site 8



Site 7


6.201 6.011 5.096 4.735 4.603 4.241
0.007 0.019 0.019 0.021 0.022 0.023


7 6 5 4
Chemical sift (ppm)


Figure 4-10. Region 8 to 3 ppm within group B wetland spectra and peak assignments
for; A) unidentified inositol phosphate, B) orthophosphate, C, D, E, F) myo-
Inositol hexakisphosphate, G) scyllo-lnositol hexakisphosphate.


137


~Sr. -.















































0 5 10


15 20 25 30


Microbial P (% of total P)


Figure 4-11. Scatter plot of microbial P against A) DNA and B) Polyphosphates. Both
showing significant positive correlation as determined by Spearman's rank
correlation (rho (DNA) = 0.61, p = 0.002, rho(polyphosphates) = 0.78, p < 0.0001).


138


20-


3.
-15-
0
4-
0
o0 10-



z
o 5-


0-


A


o




Eo o





5 10 15 20 25 30
|o I> o
0 50 0 1 25 3


12-

4-
o


8-

(U
O 6-
0
- 4-

2-

0-


B o
0
O Group A
SGroup B o
Group C 0 0
Group D




So
P 00
0 0 o O0
Kvc0











CHAPTER 5
PHOSPHORUS FORMS IN HYDROLOGICALLY ISOLATED WETLAND AND
SURROUNDING PASTURE SOILS1

Introduction

Diffuse phosphorus (P) loads from agricultural ecosystems impact the ecological

functioning of many inland waterways and wetland systems (Khan and Ansari 2005;

Verhoeven et al. 2006). Yet wetlands within agricultural landscapes also offer a

potential solution, by acting as a water and nutrient storage system at the landscape-

scale (Mitsch and Day 2006; Moreno et al. 2007; Paludan et al. 2002; Perkins et al.

2005). The functional role that wetlands play within the landscape has been recognized

at the federal level by programs such as the US Army Corps of Engineers and EPA-

administered Compensatory Mitigation for Losses of Aquatic Resources

(www.epa.gov/wetlandmitigation/), and the NRCS-administered Wetland Reserve

Program (www.nrcs.usda.gov/programs/wrp/). At the state level, recognizing this has

translated to the adoption of state-specific programs such as Florida's Lake

Okeechobee Isolated Wetland Restoration Program (LOIWRP). A cost-share program

under the mandate of the Lake Okeechobee Protection Program (LOPP) (FL Statute

373.4595), this initiative seeks to enhance and restore wetlands to retain increased

amounts of water and P within the four priority basins (S-65D, S-65E, S-154 and S-191)

north of Lake Okeechobee (Zhang et al. 2009a)

Within the south-central Florida region, isolated wetlands represent shallow (~1 m)

depressions within a landscape of very low (0-2%) topographic relief (Capece et al.


1 Published in a modified format by Journal of Environmental Quality, online 25 May 2010.
doi:10.2134/jeq2009.0398.


139









2007; Reddy et al. 1996). Although a significant component of the landscape, ~13,000

ha (13.8% of the land area) within the four priority basins (McKee 2005), these wetlands

have little connectivity to surrounding surface waters. Anthropogenic alteration of

drainage patterns during the expansion of cattle ranching (Steinman and Rosen 2000)

means that a large proportion of these wetlands (including the study sites) are now

classed as 'head of ditch' wetlands, with some degree of channelized outflow (Flaig and

Reddy 1995). It is estimated that the surface soils (0-10 cm) of these systems store ~

290 kg P ha-1 (McKee, 2005), with recent studies showing significant increases in both

soil and total ecosystem P storage with increased hydroperiod (Dunne et al. 2007). It is

believed that restoring former hydrological conditions will lead to increased water and

nutrient storage in the landscape and, therefore, reduce the P loading to downstream

waters that ultimately contribute water to Lake Okeechobee. It is important to determine

the impact of restoration efforts on current P storage in both wetland and upland soils

that may experience an increased hydroperiod following restoration.

Wetland ecosystems have surface soils that contain greater amounts of organic

matter relative to both underlying soils and adjacent terrestrial ecosystems (Axt and

Walbridge 1999; Gathumbi et al. 2005; Pant and Reddy 2001). This is due to increased

organic matter accumulation within wetlands, a result of high plant productivity, their

receiving landscape position, and relatively slow decomposition rate mediated by

anaerobic conditions (Craft and Richardson 1993; DeBusk and Reddy 2003). Such

conditions can promote the accumulation of organic P relative to inorganic P, leading to

the generalization that P in wetland soils is dominated by organic fractions (Newman

and Robinson 1999; Reddy et al. 2005). In a meta-analysis of terrestrial grassland soils,


140









organic P was found to represent 26-57% of total P (Harrison 1987). Similarly, studies

of specific landscape types have shown high levels of variation; for example, between

31 and 75% of total P occurred as organic P in a range of 29 temperate pasture soils

(Turner et al. 2003e). Such variation can be attributed to differences in land-use history,

local climate, and pedogenic development (Harrison 1987; McDowell and Stewart 2006;

Sumann et al. 1998; Turner et al. 2007a) and highlights the need to determine changes

in P forms between wetlands and uplands in the context of a given climatic zone and

agricultural management history.

The biogeochemical turnover of organic P is dependent upon its stability in the

environment, a result of both abiotic and biotic processes affected by landscape position

(Celi and Barberis 2005b; Wetzel 1999). Studies in a range of upland soils have shown

organic P to be dominated by phosphomonoesters (Chapuis-Lardy et al. 2001;

McDowell and Stewart 2006; Murphy et al. 2009; Turner et al. 2003a), whereas recent

work has indicated a greater prevalence of phosphodiesters in organic matter

dominated wetlands (Turner and Newman 2005; Turner et al. 2006a). Studies of upland

soils have attributed a positive correlation between the proportion of organic P found as

phosphodiesters (e.g., nucleic acids, phospholipids) and annual precipitation to the

increased recalcitrance of phosphodiesters under 'wetter' conditions (Condron et al.

1990a; Tate and Newman 1982). It is presumed that similar mechanisms, alongside the

reduced stability of redox-sensitive iron-bound inositol hexakisphosphate (Heighton et

al. 2008; Suzumura and Kamatani 1995a), may account for the differences observed

between the organic P forms of uplands and wetlands observed to date (Turner and

Newman 2005). In addition to a greater prevalence of organic P within wetlands, it


141









seems likely that organic P forms will vary systematically across the upland-wetland

transition in response to differences in biogeochemical properties.

Here I compare the composition and availability of P between wetland and upland

soils of an agricultural cow-calf grazing system in the subtropics. The specific objectives

were to: (1) determine basic soil biogeochemical and P characteristics in wetland and

uplands soils of grazed subtropical pasture; (2) quantify organic P composition across

the upland-wetland interface using solution 31P NMR spectroscopy; and (3) determine

the effect of hydrology on P composition and storage, with the aim of predicting impacts

of increased hydroperiod.

Materials and Methods

Site Description

Four study sites were located on two cow-calf ranches, Larson-Dixie and Beaty,

north of Lake Okeechobee in south-central Florida (Figure 5-1). Surrounding pasture

uplands are unimproved cow-calf operations with typically low stocking densities of ~1

animal unit ha-1 (Bhadha and Jawitz; Gornak and Zhang 1999) and ( M. Flinchum,

personal communication, 2004). The unfenced wetlands, two on each ranch, were

similar in size, 1 to 2 ha, and supported similar vegetation communities, with deep

marsh zones (see soil sampling) dominated by open water species (e.g. Pontedaria

cordata var. lancifolia (Muhl.)), shallow marsh zones dominated by flood-tolerant

species (e.g. Juncus effusus L.), and surrounding pasture uplands being predominantly

Paspalum notatum Flugge. Soils at Larson Dixie are mapped as siliceous, hyperthermic

Spodic Psammaquents (Basinger series). Soils in the Beaty uplands are sandy siliceous

hyperthermic Aeric Alaquods (Myaka sands), while the study wetland are delineated as

a 'Basinger and Placid soils depressional association' with Placid soils being sandy,

142









siliceous, hyperthermic Typic Humaquepts. All soils are sandy textured with high

infiltration rates yet low internal drainage given the typically high water table (Lewis et

al. 2003)

Soil Sampling

In March 2007, during a period of draw down, each wetland area was stratified into

three zones based on soil, vegetation and hydrologic indicators: pasture upland, shallow

marsh, and deep marsh. Within each zone, three locations were selected randomly for

quadrat sampling. Within each quadrat, three intact soil cores (diameter 7.5 cm x 10

cm) were collected and amalgamated. Samples were transported to the laboratory on

ice and homogenized prior to sorting. Roots and recognizable organic fragments > 2

mm were removed by hand and the soil stored at 40C. Subsamples were oven dried

(70C, 72 h), sieved and ground for total elemental analysis (see below).

Hydroperiod Determination

Soil sampling elevations were compared to the elevation of a groundwater well

located in the center of each wetland. Using wetland water level elevations, I

determined how many days per year a specific soil sampling location was saturated and

termed this estimate 'hydroperiod'. Water levels in Larson Dixie wells were recorded

from July 2003 to April 2006 and in Beaty wells from December 2003 to March 2006.

Soil Biogeochemical Properties

Soil water content was determined as weight loss following drying at 70C for 72 h.

Soil pH was determined on a 1:2 soil to water suspension using a glass electrode and

soil bulk density was calculated using the known sample volume (3 cores, 7.5 cm

diameter, and 10 cm deep) and determined water content. Total P was determined by

combustion of soil at 550C in a muffle furnace for 4 h, dissolution of the ash in 6 mol L-1

143









HCI (Andersen 1976), and then detection of molybdate reactive P (MRP) using a

segmented flow analyzer (AAII Technicon, SEAL Analytical, UK) and standard

molybdate colorimetry (USEPA, 1993). A sub-sample of the acid solution was analyzed

for AI, Ca, Fe, and Mg on an inductively-coupled plasma optical-emission spectrometer

(ICP-OES) (Thermo Jarrell Ash ICAP 61 E, Franklin, MA). Total soil C and N were

measured by combustion and gas chromatography using a Flash EA1112 (Thermo

Scientific, Waltham, MA).

Phosphorus Characterization

Phosphorus forms were determined by two parallel methods: (1) a single step

process to determine acid extractable inorganic P by extraction in 1 mol L-1 HCI for 3 h

(Reddy et al. 1998) and (2) a combination of anion exchange membranes (AEM) and

solution 31P Nuclear Magnetic Resonance (NMR) spectroscopy. This method utilizes

HCO3 loaded AEM strips as an initial extraction procedure for exchangeable P (Myers

et al. 1999), followed by the addition of a solution containing NaOH and EDTA to

proceed with a standard alkaline extraction for 31P NMR analysis (Cade-Menun and

Preston 1996).

In the AEM-NMR method, duplicate fresh samples (3.5 g dry weight equivalent)

were measured into pre-weighed 250 mL HDPE centrifuge bottles and distilled

deionized water (DDI) was added to bring the water content to 74 mL. One set of

samples received a further 1 mL of DDI (non fumigated), while a parallel set received 1

mL of 1-hexanol (95%, Sigma Aldrich, St Louis, MO)1 mL hexanol (fumigated). This

allows for the determination of both exchangeable P (PAEM) and, by difference, a

measure of fumigation-released or 'microbial' P (PM) (Kouno et al. 1995; Myers et al.

1999). All samples received a single 6.25 x 1.5 cm AEM strip (BDH Prolabo Product

144









number: 551642S, VWR International, UK), which were prepared by shaking 25 strips

for 24 h in three sequential changes of 200 mL of 0.5 mol L-1 NaHCO3. The bottles were

sealed and samples shaken on a reciprocating table for 24 h. The AEM strips were then

removed, rinsed of adhering material with DDI, shaken dry and the phosphate eluted by

shaking for 3 h in 50 mL of 0.25 mol L-1 H2S04 (Cheesman et al. 2010b; Turner and

Romero 2009a). The concentration of phosphate was determined using a discrete

autoanalyzer (AQ2+, SEAL Analytical, UK) and standard molybdate colorimetry

(USEPA 1993). The AEM strips recovered -100% of an orthophosphate standard

solution (500 pg P /membrane) and 97% of an orthophosphate spike added to replicate

soil samples (100 pg P to 3.5 g soil/membrane). It should be noted that in addition to

inorganic orthophosphate, the membrane strips may have recovered small

concentrations of labile organic P (CHAPTER 3)

After removing the AEM strips, DDI was added to the unfumigated samples to

bring the water content to 100 mL. To this, 5 mL of a solution containing 5.25 mol L1

NaOH and 1.05 mol L-1 EDTA was added to give a final concentration of 0.25 mol L1

NaOH and 50 mmol L-1 EDTA in 1:30 soil to solution ratio. The samples were shaken for

4 h and then centrifuged at 7000 rpm (maximum RCF -7500 g) (Sorvall RC6, SL1500

Rotor; Thermo Fisher Scientific, Waltham, MA) for 10 min. The supernatant was

decanted and each sample analyzed for total P (NaOHTp) using a modified double-acid

digest (Rowland and Haygarth 1997). Three mL of extract was pipetted into a boiling

tube to which 1 mL of conc. H2SO4 and 1 mL of conc. HNO3 were added. After placing

on a heat block to reduce the volume to -0.5 mL, samples were refluxed at 5500C for at

least 3 h before being diluted by a factor of 100 in DDI and analyzed for orthophosphate


145









by molybdate colorimetry (see above). Alkaline extracts were not analyzed for

molybdate reactive P due to error associated with phosphate detection by this

procedure in high organic matter soils (Turner et al. 2006b).

Solution 31P Nuclear Magnetic Resonance Spectroscopy

Replicate alkaline extracts were combined on an equal volume basis within each

of the three zones in a given wetland, resulting in 12 amalgamated samples for solution

31P NMR analysis. Each extract (15 mL) was spiked with 1 mL methylenediphosphonic

acid (MDP) (50 pg P mL-1) as an internal standard, frozen at -80 C, and lyophilized.

Approximately 100 mg of lyophilized material was resuspended in 0.1 mL D20 and

0.9 mL of a solution containing 1 mol L-1 NaOH and 100 mmol L-1 EDTA, and loaded

into a 5 mm NMR tube to ensure consistent chemical shift and strong signal lock.

Solution 31P spectra were acquired using a Bruker Avance 500 MHz spectrometer, with

a broadband probe operating at 202.45 MHz. Spectra were accumulated using waltz

decoupling (zgig program) with a 4.0 ps pulse (~300), 0.4 s acquisition time and 1.0 s

delay. Between 18,000 and 60,000 scans were needed for good signal to noise ratio

dependent upon the P concentration in the sample.

Spectra were analyzed with NMR Utility Software (NUTS) initially using 15 Hz line

broadening. Spectra were phased, corrected for baseline shift, and calibrated to the

MDP internal standard (chemical shift (6) = 17.47 ppm determined when a spiked

sample was calibrated against externally held 85% H3P04 set at 0 ppm). Spectra were

integrated over specific intervals to determine functional P groups (Table 5-1) based

upon established literature (Turner et al. 2003d). The region between 3 and 8 ppm was

further analyzed using the deconvolution utility of NUTS software. A best fit

deconvolution of the spectra was acquired using 3 Hz line broadening and peak picking
146









parameters of 5% of maximum peak height and 0.5 for the root mean squared noise

parameter. The region was split into orthophosphate and phosphomonoesters (all other

peaks determined by the algorithm in the region 3 to 8 ppm). Peak proportions from

deconvolution were then applied to the interval integration determined in the 15 Hz

spectra. Repeated integration of the same spectrum provided confidence in the

detection of signals within the pyrophosphate region equivalent to 1 mg P kg-1 soil.

Data Analysis

All statistical tests were performed in SPSS for windows version 17.0.0 statistical

software (SPSS Inc., 2008). Both visual inspection and the Shapiro-Wilk test were used

to test normality. If required, a natural log transformation was applied before statistical

analysis. Basic biogeochemical characteristics and forms of P were analyzed with a

simple univariate ANOVA with landscape position as the main factor and wetland site

(Larson East/West, Beaty North/South) as a random factor. Post hoc analysis (Tukey

HSD test) was applied to explore any significant relationships determined between

landscape positions. Regression between calculated hydroperiod and biogeochemical

characteristics were explored by the fitting of linear, exponential and linear segmented

curves to the data.

Results and Discussion

Soil Biogeochemical Characteristics

As expected, hydroperiod varied between landscape positions, with the deep

marsh zone being flooded on average 63% of the year, the shallow marsh being flooded

25% of the year, and the uplands being rarely if ever flooded (Table 5-2). Deep marsh

zones had significantly lower pH, lower bulk density and higher organic matter

concentration (as indicated by loss on ignition) than other landscape positions (p <

147









0.05). Of the total metals analyzed, Mg was at or below the practical quantification limit

for the majority of samples (data not shown). Calcium showed no significant trends

across landscape position, whereas Al and Fe concentrations showed a significant

increase within wetlands (p < 0.05). Post hoc analysis suggested that deep marsh areas

were the most distinct, containing concentrations of up to twice as much Fe and 10

times as much Al compared to surrounding uplands. This is consistent with positive

correlations between organic matter and Fe/AI within other wetland systems (Darke and

Walbridge 2000) and I hypothesize that this represents stable organo-metal complexes

in the more organic deep marsh soils. The expected gradient in accumulating organic

matter, was supported by analysis of total C and total N (Table 5-2), both of which

increased significantly (p < 0.05) with landscape position towards the wetlands deep

marsh. There was also significant (p < 0.05) interaction observed between landscape

position and wetland site, attributed to longer hydroperiods within the Beaty wetlands

deep marsh (average 239 days) compared to Larsons (194 days) (data not shown).

For most soil properties, significant differences among landscape positions were

the result of distinct deep marsh characteristics. With the exception of total N, post hoc

analysis demonstrated only limited, non-significant differences between the pasture

upland and shallow marsh zones (see: Impact of hydroperiod).

Soil Phosphorus Composition

There was a significant difference in total P concentration across landscape

position, with an approximate threefold increase from 117 mg P kg-1 in the pasture

uplands to 371 mg P kg-1 in the deep marsh (p < 0.05) (Table 5-3, Figure 5-2). This

significant gradient was also observed in the HCI-extractable inorganic P, which ranged

from 23 mg P kg-1 (20% total P) in the uplands to 61 mg P kg-1 (16% total P) in the deep

148









marsh. The AEM-NMR method extracted on average 53% 11% of total P in all soils

sampled, with no significant bias in the recovery rate among landscape positions.

Phosphorus not extracted (residual P) was considered a distinct unidentified pool,

hypothesized to consist of recalcitrant organic and alkaline-stable mineral forms (Cade-

Menun 2005b; Turner et al. 2005).

Exchangable phosphate (PAEM) showed a significant increase from 4.4 mg P kg-1

(3.8% total P) in the uplands to 23.9 mg P kg-1 (6.4% total P) in the deep marsh (p <

0.05). Microbial P showed a significant increase in concentration, but a reduction in its

proportion of total P (p < 0.05), from 22 mg P kg-1 (19% total P) in uplands to 37 mg P

kg-1 (10% total P) in the deep marsh. Although the concentration of NaOHTp changed,

the proportion of it relative to total P did not differ significantly with landscape position,

ranging from 56 mg P kg-1 (48% total P) in the uplands to 184 mg P kg-1 (50% total P) in

the deep marsh. This pattern was repeated in the residual P fraction, with a significant

(p < 0.05) difference in concentration, yet similar proportions of total P, from 56 mg P kg-

1 (47% total P) in uplands to 163 mg P kg-1 (43% total P) in the deep marsh. Although

concentrations of all P pools determined by the AEM-NMR method (PAEM, PM, NaOHTP

and residual) showed significant differences across landscape positions, the difference

between pasture uplands and shallow marsh were not significant. Given the trends in

data across landscape position, I attribute this lack of significance to the high variability

within soils collected from the shallow marsh zone (see: Impact of Hydroperiod and

Figure 5-3).

Solution 31P Nuclear Magnetic Resonance Spectroscopy

Solution 31P NMR spectra showed the presence of orthophosphate,

phosphomonoesters, phosphodiesters, pyrophosphate and trace concentrations of
149









phosphonates and polyphosphates (Figure 5-3, Table 5-4). Approximately 73% of the

extracted P was organic, with phosphomonoesters constituting the major fraction,

ranging from 50 mg P kg-1 in the uplands to 127 mg P kg-1 in the deep marsh. The lack

of a characteristic 1:2:2:1 signature within the phosphomonoester region indicated the

absence of detectable concentrations of myo-lnositol hexakisphosphate (Turner et al.

2003f). The remaining organic P included phosphodiesters (11-13% total soil P),

dominated by DNA, with the remainder representing various alkali-stable phospholipids.

Due to the rapid hydrolysis of certain compounds in alkaline extracts (i.e. RNA and

some phospholipids) the proportion of organic P determined as phosphodiesters is likely

to be underestimated (Turner et al. 2003d). Total inorganic P constituted between 16

and 19% of total P. This was dominated by orthophosphate, while pyrophosphate was

detected in all spectra at concentrations ranging from 3.4 mg P kg-1 (2.0% total P) in the

shallow marsh to 6.7 mg P kg-1 (1.8% total P) in the deep marsh. Polyphosphate was

detected at low concentrations in certain deep marsh soils (average 1.5 mg P kg-1 soil).

The concentrations of phosphomonoesters, DNA, and orthophosphate increased

significantly from uplands to deep marsh (p < 0.05; Table 5-4). However, there was a

striking lack of distinction between the relative proportions of P forms identified across

landscape position. Instead of the expected changes in forms and magnitudes of P

between wetlands and uplands, all sites contained similar forms of P, leading me to

reject my initial hypothesis. The proportion of P identified as phosphodiesters did not

increase in the deep marsh as expected from previous studies of wetland soils (Turner

and Newman 2005; Turner et al. 2006a). In addition, the marked lack of the distinctive

peak signature associated with the various isomers of inositol hexakisphosphate, even


150









in upland pasture samples, was in contrast to other studies of upland grasslands

(McDowell and Stewart 2006; Murphy et al. 2009; Tate and Newman 1982; Turner et al.

2003e). The abiotic sorption and stabilization of inositol phosphates in such

environments is attributed to their interaction with clay minerals and amorphous Fe and

Al oxides (Celi and Barberis 2005b, 2007; Heighton et al. 2008). In addition to data

presented, which reports the low Fe and Al concentrations throughout the system

(Table 5-2), existing soil surveys (Lewis et al. 2003) and detailed particle size

distribution analysis (Bhadha and Jawitz, 2010) show only minimal clay content (<

2.5%) in the uplands. Therefore, it is likely that inositol phosphates are not physically

stabilized within these soils and instead are hydrolyzed rapidly.

Impact of Hydroperiod

The use of a priori categorization using vegetation, soils, and hydrology of the

wetland-upland landscape continuum is a qualitative approach to classifying the

different landscape positions (deep marsh, shallow marsh, and upland). The use of a

calculated hydroperiod is a quantitative approach, giving an absolute value for a

hydroperiod of a given location.

The SPSS curve estimating procedure was applied to test the significance of linear

regressions between the calculated hydroperiod and various soil P characteristics.

There was a positive linear relationship between soil total P and hydroperiod (R2=

0.734; p < 0.001; Figure 5-4 A). The potential for a more complex segmented

relationship with a critical threshold hydroperiod at approximately 100 days was

explored by the fitting of a segmented linear regression; however, there was not a

substantially better fit to the data (R2 = 0.758).


151









Phosphorus forms determined by solution 31P NMR spectroscopy were plotted

against the average hydroperiod of the amalgamated soils used to generate the

spectra. Total organic P determined by 31P NMR analysis showed a significant positive

linear relationship with hydroperiod (R2 = 0.837; p < 0.001; Figure 5-4 B). Similar

relationships were found for phosphomonoesters and phosphodiesters (Figure 5-4 C,

D). Although total organic P and its major forms showed similar positive trends in

concentration with calculated hydroperiod, I had initially hypothesized that increasing

hydroperiod would result in increased stabilization of organic P and as a result, lead to

an increase in its proportion of total P, as well as the preferential stabilization of forms

characteristic of wetter soils (i.e. phosphodiesters). However, I observed no evidence of

a significant relationship between hydroperiod and organic P (as a proportion of total P),

or the relative proportions of phosphodiesters to phosphomonoesters.

Across landscape position there was a general increase in total soil P towards the

wetland deep marsh, but the concomitant increase in all partitioned forms of P suggests

a mechanism of general accumulation, as opposed to a preferential stabilization of

specific forms. The highly significant (p < 0.001) linear relationship between total C and

total P (Figure 5-5) suggests that the accumulation of P is concurrent with the

accumulation of organic matter. Previous analysis of manure-impacted surface soils in

the region showed that the upper soil horizons consist of uncoated quartz sand grains,

and the low clay concentrations are dominated by non-crystalline silica (Harris et al.

1996; Harris et al. 1994). Although deeper horizons may adsorb P strongly (Graetz and

Nair 1995), mineral components within the surface soils have a very low P binding

capacity, resulting in P dynamics driven by an association with organic matter.


152









The accumulation of organic matter in response to increased hydroperiod may be

a result of anaerobosis or increased endogenous and exogenous inputs. The

transportation of particulate P in runoff has long been recognized as a significant

component of P transfer (Daniel et al. 1998) and, therefore, a focus for management

practices to reduce P loss from agricultural landscapes (McDowell et al. 2007; Sharpley

et al. 2001). The forms and nature of P transported in surface runoff is dependent upon

soil biophysical characteristics and site hydrological properties (Ballantine et al. 2009;

Heathwaite and Dils 2000). The direct determination of surface runoff at dispersed

wetland fringes with low overall topography is problematic, although indirect modeling

suggests that overland flow may play a considerable role in the water budget of these

study wetlands (J. H. Min, personal communication, 2010). With direct overland flow

accounting for up to 25% of water inputs, often associated with isolated high intensity

precipitation events. Such high energy rainfall events contribute disproportionately to

increased movement of particulate P in highly sloped tilled silt/loam soils (Jin et al.

2009). I am unaware of studies that emulate the site conditions, i.e. that include

contiguous ground cover, a low topographic gradient, and soils dominated by siliceous

sand grains and variously sized organic matter particles. In addition, all study wetlands

were unfenced within grazed pastures. Given known cattle use of wetlands throughout

the year (Pandey et al. 2009) and associated increases in potential sedimentation,

vegetation turnover (McDowell et al. 2007), and direct fecal additions, the influence of

cattle on organic matter accumulation may be significant.

The importance of organic matter stabilization of P, even in the low organic matter

upland soils, as well as the potential translocation of this organic matter, could account


153









for the absence of any distinct shifts in P forms between landscape position. Further

work is required to investigate the role of endogenous and exogenous inputs to the

carbon and P dynamics of the receiving wetlands.

Phosphorus Storage

For the restoration of hydrologic conditions in isolated wetlands to be an effective

P management practice, wetlands would need to show evidence for increased nutrient

storage on an aerial basis (kg ha-1) relative to other landscape positions. Although

moderated by a reduced bulk density in wetlands relative to uplands (Table 5-2),

storage of both total and organic P is significantly (p < 0.05) elevated within the surface

soils of the wetlands deep marsh (Table 5-5), with a pronounced twofold increase in

total P from 114 kg ha-1 in the unimproved pasture to 236 kg ha-1 in the deep marsh

areas. The marginal, but not significant, increase in P content between pasture uplands

areas and shallow marsh suggests that meaningful nutrient storage requires a

substantial increase in hydroperiod. In addition, comparing current landscape storage

does not take into account the time-scale required for its development. It can be

assumed that an increase in hydroperiod would necessitate a period of adjustment

during which P and C would be accreted. The rate and controlling mechanisms of this

accretion would have significant impact on the efficacy of restored wetlands to reduce

downstream P loading.

Conclusions

Isolated wetlands exhibited an accumulation of organic matter as compared to

surrounding pasture uplands, which was closely associated with an accumulation of

total P. However, this was not related to differences in soil P forms predicted from

differential stability across the landscape. Instead, total P accumulation was

154









concomitant with an increase in all P forms, including exchangeable P, microbial P, and

forms identified by alkaline extraction and 31P NMR spectroscopy. Successful

management of isolated wetlands for the onsite sequestering of P from agricultural

runoff will depend on the accumulation and stabilization of organic matter. This study

suggests that a substantial hydroperiod increase is required to sequester increased

amounts of P.


155









Table 5-1. Specific integral ranges used in the classification of solution 31P NMR spectra for lyophilized material re-
suspended in 0.9 mL (1 mol L-1 NaOH 100 mmol L-) + 0.1 mL D20. Internal standard methylenediphosphonic
(MDP) calibrated to 6 = 17.46 ppm. Region 8 to 3 ppm is determined as orthophosphate or phosphomonoesters
after deconvolution of spectra at 3 Hz line broadening.


Chemical
shift
21.0 to 20.0
18.0 to 16.8


Attributed phosphorus group
Phosphonates
MDP (methylenediphosphonic acid)
Orthophosphate + phosphomonoesters


8.0 to 3.0 Line broadening =3Hz


3.0 to -1.5


-3.0 to -5.0


Phosphodiesters 0.5 to 1.5 DNA
[3.0 to -1.5] DNA Other posphomonoesters
Pyrophosphate and end-chain polyphosphate functional groups


-18.0 to -21.0 Polyphosphate mid chain functional groups


156


Orthophosphate 6.24 0.01 ppm
Phosphomonoesters = [8 to 3] orthophosphate









Table 5-2. Soil characteristics and nutrients determined in samples across landscape
position. Values represent averages (n = 12) 1 standard deviation taken as
three replicates from four wetland sites.


Hydroperiod (days)
pH


Pasture Upland
2a 6
4.8a 0.6


Bulk density (g cm-3) 0.98a 0.10
Organic matter (g kg1)* 86a 23
Total elements (g kg-1)
Carbon 44a + 12
Nitrogen* 2.8a 0.5
Calcium 1.1 0.5
Iron* 0.7a 0.3
Aluminum ** 0.3a 0.2
t Organic matter estimated by loss on ignition
Significant at the 0.05 probability level
*** Significant at the 0.001 probability level


Shallow Marsh
92b 67
4.7a 0.4
0.92a + 0.21
123a 71


58a
3.6 b
1.2
0.6a
0.7a


Deep Marsh
2310 57
4.3b + 0.3
0.65b + 0.19
219b 66


112b
4.8c
1.9
1.1b
3.6b


Table 5-3. Phosphorus forms across landscape position determined by acid extraction
and AEM-NMR method. Values represent averages (n=12) 1 standard
deviation.
Pasture Upland Shallow Marsh Deep Marsh
Total P (mg kg-1) 117a 25 171a 107 371b 61
HCI-Pi t (% total P) 20 5 19 4 16 2
AEM-NMR Method (% total P)
PAEM* 4 2 6 3 6 + 2
PM* 19a 5 16a 10 10 b + 4
NaOHTP 49 10 42 8 51 13
ResidualO 47 + 11 52 8 43 13
t Inorganic P extracted in 1 mol L- HCI


t Exchangeable (PAEM) and microbial (F
Alkaline extracted total P
= (PAEM + NaOHTP).
* significant at the 0.05 probability level


'M) determined by AEM extraction


157


)









Table 5-4. Phosphorus forms determined by solution 31P NMR spectroscopy of
amalgamated alkaline extracts from across the landscape transition. Values
from each landscape position represent averages (n= 4) 1 standard
deviation


Phosphorus form


Phosphonate
Orthophosphate*
Phosphomonoester
DNA
Other phosphodiesters
Pyrophosphate


Pa

0.
18
50
6.4
8.
5.


Polyphosphate nd
Organic P (% of total NMR) 73.7
significant at the 0.05 probability level
n.d not detected by 3P NMR spectroscopy


sture Upland Shallow Marsh Deep Mars
mg kg-1
6 1.2 trace 0.9
a 6 24a 11 57b +
)3 5 58a 16 127b +
4a 1.9 14.2a 2.2 29.7b 1
8 7.2 10.1 0.7 12.5
3 2.6 3.4 2.4 6.7


nd
74.9


h


1.4
9
21
1.0
7.6
3.8


1.5 2.9
72.2


Table 5-5. Storage of total (n = 12) and organic phosphorus (n = 4) within the top 10 cm
of soil, across landscape transition. Values from each landscape position
represent averages 1 standard deviation
Phosphorus (kg ha-1) Pasture Upland Shallow Marsh Deep Marsh
Total P 114a + 27 141a 50 236b 62
Organic P 65a + 10 75a + 19 108b + 33
significant at the 0.05 probability level


158










































Figure 5-1. Location of study sites showing A) Florida outline with area of interest north
of Lake Okeechobee, and B) detail of ranch sites containing two study
wetlands each, within priority basins north of Lake Okeechobee. Beaty Ranch
is within sub-basin S-65D of the Kissimmee River watershed, with hydraulic
connection through Cypress slough to Canal C38 and then to the Lake.
Larson Dixie ranch is located in sub-basin S-154 within the Taylor Creek
watershed, connected through Mosquito Creek to Taylor Creek itself.










159













0- MAEM Extractable P
MAEM Microbial P
DNaOH-EDTA Total P
o0- Residual P
0--
i Total P
I i

0-

S0
+ E

*- r -
r I


Pasture Upland


Shallow Marsh


Deep Marsh


Figure 5-2. Phosphorus pools determined by AEM-NMR method across the three
landscape positions. Values are averages (n=12) 1 standard deviation. No
significant differences were found between Pasture Upland and Shallow
Marsh. (*) denotes significant differences between pools of the different
landscape positions (p < 0.05).


160


40


30



20


0.

E
0.

Co


10






























Pasture Upland


Shallow Marsh


Deep Marsh

20 10 0 -10 -20
Chemical Shift (ppm)


Figure 5-3. Example solution 31P NMR spectra from amalgamated samples from the
Beaty North wetland. Spectra acquired using a Bruker Avance 500 MHz with
a 4.00 ps pulse (~300), 0.4 s acquisition time and 1 s delay. Spectra
referenced and integrated against internal methylenediphosphonic acid
(MDP) standard (6 = 17.46 ppm) displayed here using 15 Hz line broadening
and scaling to MDP.


161



























u


'U C 0 o




A---'


A A
i0-


R2 = 0.794
p < 0.001
000 20 3
0 100 200 3


A
A 0

Aao
8 OO


0
0

A. a


A R 0.734
p < 0.001


Hydroperiod (days)


Figure 5-4. Phosphorus characteristics plotted against hydroperiod: A) total phosphorus for all samples, B) total organic
phosphorus determined by 31P NMR spectroscopy, C) phosphomonoesters, and D) phosphodiesters.


162


250 B

o
200 I-J Pasture Upland
A Shallow Marsh 0 -
O Deep Marsh ,--
150
---' o

100 --
-" A

50 R = 0.837
p < 0.001

0 100 200 300



D
o
60


o --
40-
--' 0


20- .---" A
R = 0.697
p = 0.001
000 20 3
0 100 200 300












500- ,


400- 0O /'o
O "/
--
0. A -0
300 0 -
b30 0 O

E
200- /A El Pasture Upland
A Shallow Marsh
7 i, 0 Deep Marsh
S100- Total P= 2.9 x (total C) + 13.8
/ R2 = 0.82

0 50 100 150 200

Total C (g C kg-1)


Figure 5-5. Total soil carbon plotted against phosphorus concentrations across
landscape positions (n = 36) from four wetland sites. Characteristics show a
significant (p < 0.001) positive relationship (R2 = 0.819).


163









CHAPTER 6
STABILITY OF SELECT BIOGENIC PHOSPHORUS COMPOUNDS UNDER AEROBIC
AND ANAEROBIC CONDITIONS

Introduction

The differential stability of biogenic P under wetland conditions, compared to

terrestrial soils has been invoked as a potential mechanism to explain observed

patterns seen in wetlands (CHAPTER 4). In brief, anaerobic conditions are believed to

lead to reduced decomposition of certain biogenic P forms (i.e. phospholipids and

polynucleotides) while reducing the stability of others (i.e IP6 and polyphosphates). This

dichotomous impact of anaerobic conditions on biogenic P forms is currently unproven,

yet has been suggested in some form by a number of authors (Hupfer et al. 2007;

McKelvie 2007; Mitchell and Baldwin 2005; Turner and Newman 2005). Here I use

solution 31P NMR spectroscopy to test this by determining the difference in turnover

over rates of major biogenic P forms (DNA, myo-IP6, and polyphosphate) under both

aerobic and anaerobic conditions.

Several studies looking at biogenic P forms within terrestrial systems have

postulated a reduced decomposition of phosphodiesters (including DNA) under 'wetter'

and by extension more anaerobic conditions (Condron et al. 1990a; Tate and Newman

1982), with the changes in phosphodiester decomposition acting as a driver in

determining patterns of biogenic P seen across climatic gradients (Sumann et al. 1998).

In wetlands, typified by anaerobic conditions, there is a greater prevalence of

phosphodiesters (CHAPTER 2; Turner and Newman 2005), yet it is unclear whether this

represents decreased decomposition due to biotic hydrolysis, greater or altered

microbial biomass (Makarov et al. 2002a), abiotic stabilization of DNA to generally


164









increased organic matter content (Celi and Barberis 2005b), or a relative increase due

to reduced stabilization of phosphomonoesters.

Inositol hexakisphosphate, is a major component of biogenic P in many terrestrial

systems (Murphy et al. 2009; Turner 2007; Turner et al. 2002b), but appears to be

absent within many wetland soils studied to date (CHAPTER 2). Yet the fundamental

process that underlies the presence or absence of IP6 within wetlands is currently

unknown (McKelvie 2007). One possibility lies in the observations of Suzumura and

Kamanti (1995a,b). They observed that under anaerobic conditions in marine

substrates, terrigenously derived IP6 underwent rapid degradation (Suzumura and

Kamatani 1995b), with complete breakdown in 40 days, as compared to just 50%

degradation in 60 days under aerobic conditions (Suzumura and Kamatani 1995a).

Although, such anaerobic mediated degradation appears contrary to recent observation

of IP6 as a conservative paleoindicator within estuarine sediments (Turner and

Weckstrom 2009) and as a component of freshwater, lacustrine (Golterman et al. 1998;

McKelvie 2007), riverine (McDowell 2009) and palustrine soils studied as part of this

dissertation (CHAPTER 4). It therefore important to establish if this apparent redox

sensitivity is a universal mechanism in wetlands or represents a interaction between

redox conditions and other biogeochemical characteristics.

Finally, biogenic polyphosphates, although initially intracellular, appear to form

discrete extracellular complexes (Diaz et al. 2008) that are stabilized in the environment

by their interaction with the abiotic matrix, including redox-sensitive components such

as amorphous iron oxides (Hupfer et al. 2007; Reitzel et al. 2007; Sannigrahi and Ingall

2005). It has been proposed that the interaction of these stabilized polyphosphates with


165









fluctuating redox conditions in sediments could account for a sizable component of

benthic P flux (Hupfer et al. 2004; Sannigrahi and Ingall 2005). Since polyphosphates

have been observed in palustrine wetlands (Bedrock et al. 1994; Sundareshwar et al.

2009) it is important to determine if differential redox conditions may account for their

presence or absence within particular wetland systems (CHAPTER 4), as well as to

help establish the ecological importance of this P form within palustrine soils.

Anaerobic conditions are postulated to directly impact a number of important

biogenic P pools, yet little direct confirmation exists that stability and turnover of these

forms is impacted by the presence or absence of 02. This study seeks to determine the

short-term impact of both aerobic and anaerobic conditions on biogenic P within

freshwater wetland soils, specifically focused on the fate of DNA, polyphosphates, and

myo-IP6.

Methods

Microcosm Experiment

Stability of biogenic P was determined with the use of microcosms held under

aerobic and anaerobic conditions for up to 48 days. In detail (Figure 6-1), 1 g of a

standard homogenized wetland soil (Table 6-1) was weighed into a 30 mL HDPE

centrifuge tube. Distilled deionized water (DDI) was added to bring total water content to

10 mL volumetricc water content 91%) before tubes were sealed with a septum-

containing cap. Aerobic or anaerobic conditions were developed by purging tubes for 5

min using either purified and filtered (0.2 pm GF-B in line filter) hydrocarbon free air

(Praxair Inc., Danbury CT) or ultra high purity N2 (Airgas, Radnor, PA). Anaerobic

treatments were further sealed in a polythene glove bag also purged with N2. Samples

were allowed to equilibrate for 2 weeks in the dark at 270C and with gentle agitation


166









(G25 incubator shaker, New Brunswick Scientific Co. Inc, NJ). Head spaces were

purged in a similar manner every 3 to 4 days found to maintain an 02 free state in

anaerobic treatments by monitoring a subset using gas chromatography (GC-8A TCD,

Shimadzu, Japan).

Biogenic Phosphorus Spikes

After equilibration, samples were spiked with either 2 mL of DDI (control), or 2 mL

of a biogenic standard (~200 pg P g-') in a DDI matrix. Standards (Sigma-Aldrich

Corporation, St Louis, MO) were either myo-IP6 (dodecasodium salt) or a mixture of

DNA (deoxyribonucleic acid, from salmon testes) and polyphosphate (sodium

hexametaphosphate 96%) (Figure 6-2). Subsequent analysis of representative

microcosms for total P (see Biogeochemical Characterization) showed clear

repeatability of standard compound spike (Table 6-2). The experiment represented a

factorial design with duplicate samples being subjected to each treatment combination.

After spiking, microcosms were maintained under conditions as described above and

destructively sampled and analyzed for P composition at days 1, 8, 16, 30 and 48.

Biogeochemical Characterization

Water content was calculated using gravimetric loss after drying sub-samples at

750C for 72 h. Total P was determined by combustion of soil at 550C in a muffle

furnace for 4 h, dissolution of the ash in 6 mol L-1 HCI (Andersen, 1976) and the

detection of MRP using a discrete auto analyzer (AQ2, SEAL Analytical, UK) and

standard molybdate colorimetry (USEPA, 1993). A sub-sample of the acid solution was

analyzed for AI, Ca, Fe, and Mg using ICP-OES (Thermo Jarrell Ash ICAP 61 E,

Franklin, MA). Total soil C and N were measured by combustion and gas

chromatography using a Flash EA1112 (Thermo Scientific, Waltham, MA).


167









Phosphorus Composition

Destructive sampling and analysis for P composition consisted of a modified AEM-

NMR method (CHAPTER 5). Briefly, individual tubes received a sample-specific volume

of DDI to bring total water content to 20 mL. To this was added one 6.25 x 1.5 cm anion

exchange membrane strip (BDH Prolabo Product number: 551642S, VWR

International, UK) preloaded with a HCO3 counter ion. Samples were sealed and

shaken on a reciprocating shaker at low speeds for 24 h. Membrane strips were

removed, rinsed of adhering material and eluted using 50 mL 0.25 mol L-1 H2S04.

Membrane eluants were analyzed for MRP (PAEM) as above. Residual samples received

further sample-specific DDI to bring total water volume to 28.57 mL. To this, 1.43 mL of

5.25 mol L-1 NaOH 1.05 mol L-1 EDTA was added to result in a final alkaline extraction

of 0.25 mol L-1 NaOH and 50 mmol L-1 EDTA in a 1:30 soil to solution ratio. Samples

were shaken for 4 h, before being centrifuged at 7000 rpm (maximum RCF ~7000 g)

(Sorvall RC6, SL600 Rotor; Thermo Fisher Scientific, Waltham, MA) for 12 min. A

subsample of supernatant was drawn off and stored at 4C for determination of total P

(PNaOH) using a double acid (H2SO4-HNO3) digest (Rowland and Haygarth 1997) and

molybdate colorimetry. A second subsample (5 mL) was combined within duplicates,

mixed with 1 mL of an MDP standard (47.6 pg P g-) and immediately frozen (-800C)

prior to lyophilization and NMR analysis.

Solution 31P Nuclear Magnetic Resonance Spectroscopy

Given the prohibitive costs associated with 31P NMR spectroscopy, initial spectra

acquisition was restricted to day 1 and day 48 samples. Given the unexpected recovery

of polyphosphates by the anion exchange membrane strips (see Phosphorus

Recovery), and the limited differences in degradation of DNA between experimental


168









treatments further analysis concentrated on samples that had received the myo-IP6

spike and had been extracted on days 8, 16 and 30. All lyophilized samples were

reconstituted using ~ 100 mg of lyophilized material per mL with the re-suspension

media being 0.9 mL (1 mol L-1 NaOH and 100 mmol L-1 EDTA) and 0.1 mL D20.

Samples were vortexed for 1 min prior to loading into appropriate NMR tubes. Initial day

1 and 48 samples were analyzed using a 5 mm BBO probe, Bruker Avance 600

Console, Oxford 14.1T/51 mm Magnet with subsequent analysis carried out using a 10

mm BBO probe, Bruker Avance 500 MHz 11.8 T/ 54 mm bore. All analysis used a

simple zgig program employing broad band decoupling (waltz 16) a calibrated 450 pulse

width and 2 s T1 delay. Sample spectra were referenced using internal standard MDP (6

= 17.46 ppm) and integrated over standard intervals (Turner et al. 2003d). The region 8

to 3 ppm was also analyzed using spectral line deconvolution enabling the quantification

of orthophosphate, myo-IP6 (Turner et al. 2003f) and scyllo-IP6 (Turner 2007; Turner

and Richardson 2004).

Data Analysis

All statistical tests were performed in SPSS for windows version 17.0.0 statistical

software (SPSS Inc., 2008). Exchangeable P, as determined by anion exchange

membrane strips, was analyzed for trends using univariate ANOVA within each spike

treatment, with time steps (1, 8, 16, 30, 48) and condition (aerobic, anaerobic) as fixed

factors. Post hoc analysis (Tukey HSD) was used to explore significant differences.

Concentrations of myo-IP6 determined across all time points were explored by simple

linear regression with time.


169









Results and Discussion


Phosphorus Recovery

The AEM-NMR method showed a good ability to describe P composition within

samples, with recovery of total P (PAEM + PNaOH) across all treatments and time periods

of 62.4 10.0% ( 1 standard deviation), the residual being considered recalcitrant

organic and alkaline stable inorganic forms. Recovery of biogenic P spikes was good,

with an increase in concentration of P recovered in spiked samples equivalent to 90% of

the myo-IP6 and 95% of the DNA + polyphosphate spike (Table 6-2).

Exchangeable P determined by anion exchange membrane showed no significant

distinction between aerobic and anaerobic conditions (ANOVA; p > 0.05) within any

spike treatment, yet there were significant differences between time steps in most

treatments (ANOVA; p < 0.001) (Figure 6-3). Although this may represent systematic

error during iterations of the analytical method, the recovery (100 10%) of standard

orthophosphate solutions by the membranes in parallel to soil extractions suggests that

there was a true difference in exchangeable P between time steps. Of particular note is

the recovery of large quantities of P from soils spiked with polyphosphate (spike 2).

Subsequent determination of P composition in alkaline extracts (Figure 6-4) showed

limited evidence of polyphosphates. Therefore, polyphosphates were either rapidly

hydrolyzed to orthophosphate or extracted directly by the anion exchange membrane

strips. Subsequent investigation has shown an ability of the membranes used to recover

certain biogenic P forms, including polyphosphates (Cheesman et al. 2010b). Taken in

conjunction with the declining trend in recovered P (Figure 6-3), it would appear that

polyphosphate, although initially recovered by the deployed anion exchange membrane,

is increasingly stabilized within the soil. Study of P composition in subsequent alkaline


170









extractions (Table 6-3) showed a concomitant increase in both orthophosphate and

residual P. Although not conclusive, this suggests both biotic uptake and potential

abiotic stabilization of polyphosphate.

Initial Biogenic Phosphorus Composition

There was a striking lack of distinction between biogenic P identified within aerobic

and anaerobic control microcosm soils extracted on day 1. Averaging between

conditions, it is clear that initial soil contained large proportions of biogenic P, including

phosphonates (1.8 0.2% total P), phosphomonoesters (30.2 0.2% total P),

phosphodiesters (11.5 0.9% of total P) and trace concentrations of polyphosphates

(Table 6-3, Figure 6-4). Deconvolution of the region 8 to 3 ppm allowed the clear

identification of two isomers of IP6, myo and scyllo, within the initial soil (Figure 6-5).

The identification of myo-IP6 by a priori peak assignment has previously been criticized,

with certain researchers calling for compound spiking to confirm suspected presence

(Smernik and Dougherty 2007). Comparison of control soil extracts with those that

received spike mixture 1 (Figure 6-6) confirms the assignment, clearly highlighting the

presence of the indicative 1:2:2:1 signature and chemical shift of the C2 bound

phosphate (Turner et al. 2003f). The assignment of the up-field peak (4.272 ppm) to

scyllo-IP6 is based on previous work by Turner and Richardson (2004) (Figure 6-6) as

well as the informed consideration of its known presence in other systems (Turner

2007).

Stability of Phosphorus Functional Groups

Comparison of initial control soils and after 48 days showed little distinction

between aerobic and anaerobic conditions. It was impossible to determine if slight

differences seen within the concentrations of minor groups such as DNA, were a


171









relevant shift in functional forms or an experimental artifact. Although, the lack of

substantial changes observed within the controls did allow for major changes in the

composition of spiked soils to be attributed to true distinctions in the fate of extracellular

P forms used. The lack of repeated measurement inherent to destructive sampling of

individual mesocosms limited my ability to discern the origin of the observed variance,

especially when considering changes in the recovery of the DNA spike. A redesigned

experiment would make use of batch reactors with continuous monitoring and

modification of redox levels as used by Christophoridis and Fytianos (2006). Thereby,

allowing stable 'aerobic' or 'anaerobic' conditions to be maintained while providing

thorough mixing and the ability for repeated analysis of a stable reactors soil, reducing

potential sample variability. The study presented here still provides us an initial incite

into the role of redox conditions on the fate of certain P forms, a vital consideration

when looking into the patterns observed in P composition of soils in the landscape.

Stability of polyphosphates and DNA

The recovery of extracellular polyphosphates by the anion exchange membranes

used here as an initial extraction step was unforeseen, yet did offer an indication of

timescale associated with the residency of extracellular polyphosphate in the soil.

Assuming the decrease in PAEM observed (Figure 6-3) was due to irreversible abiotic

stabilization or biotic uptake, extracellular polyphosphate was found to have a half life of

~35 days under both aerobic and anaerobic conditions. Although, further work using

solution 31P NMR spectroscopy without pre-extraction would be required to confirm this.

Final concentrations of DNA in soils spiked with mixture 2 showed a decrease over the

course of 48 days (Table 6-3). Taking changes in DNA concentration of control soils as

a background trend, changes in spiked soils equate to 30% of the additional DNA under


172









aerobic conditions and just 2% under anaerobic conditions. This is in line with the

original working hypothesis that phosphodiesters see a reduced decomposition rate

under anaerobic conditions, yet the experimental design and the potential for sample

specific variance (see above) limits my ability to use this as definitive evidence. Further

studies would be needed employing solely DNA as a P spike and following a complete

time course before estimates could be made of the half life of extracellular DNA under

differential redox conditions.

Stability of myo-lnositol hexakisphosphate

Comparison of spectra acquired using both the 600 and 500 MHz Brucker systems

(Table 6-4, Figure 6-7) showed that while functional forms with low overall

concentrations (i.e. phosphonates, polyphosphates) showed poor repeatability between

spectra acquisitions, major biogenic P forms, especially myo-IP6 within spiked soils

showed consistent (RSD < 5%) calculated concentrations. This was taken as strong

evidence for the validity of using spectra acquired in both systems for analysis of myo-

IP6 stability with time (Table 6-5, Figure 6-8). It is apparent that myo-IP6 showed

evidence of degradation under both aerobic or anaerobic conditions but that samples

showed a high degree of variability. Samples taken on day 30 were excluded from

statistical analysis given substantially lower recoveries of total P than at all other time

steps (Table 6-5). The loss of myo-IP6 was found to be marginally significant (R2 =

0.713, p < 0.01) under anaerobic conditions, yet in either aerobic or anaerobic

conditions represented a decrease of just 10% of the total myo-IP6 spike used; although

it should be noted that if significant, this total IP6 decomposition is equivalent to that in

previous studies (Suzumura and Kamatani 1995a) wherein total myo-IP6 used was

equivalent to just 23 pg P g-1 wet weight.


173









The lack of complete myo-IP6 degradation or any discernable distinction in its

turnover between aerobic and anaerobic conditions as otherwise seen in marine

sediments (Suzumura and Kamatani 1995a) probably reflects distinct biogeochemical

characteristics of the soil and sediment material used. Assuming similar basic sediment

properties to those seen in Suzumura and Kamatani (1995b). Total P concentrations

within the marine study were approximately the same as used here (~800 pg g-1 ) but

contained an order of magnitude less carbon ( ~29 mg g-1). It is also worth noting that

native IP6 in the Suzumura and Kamatani (1995ab) study was low with inferred myo,

scyllo and cis isomers combined representing just 1.9 pg P g-1 (~0.2% of total P). In

contrast using day 1 aerobic and anaerobic control samples from this study native myo

and scyllo-IP6 were found to represent ~14% of total P (Table 6-3). Differences in the

carbon content, material origins and biogeochemical characteristics of the palustrine

soils used here as compared to marine sediments used previously may impact both

abiotic stabilization and biological degradation of myo-IP6.

Abiotic stabilization of myo-lnositol hexakisphosphate. The abiotic sorption

and stabilization of IP6 is complex, and governed by various aspects of the

environmental matrix (e.g. clay content, amorphous metal oxides, Ca2+, calcite and

organic matter content) (Celi and Barberis 2007). The redox-sensitive degradation of IP6

has been linked in certain systems to the reduction of Fe(lll) oxides (Celi and Barberis

2007; McKelvie 2007), though it has also been suggested that a IP6 could form insoluble

complexes with reduced Fe species (DeGroot and Golterman 1993). In contrast, the

interaction and stabilization of organic P with organic matter is largely, with the

exception of polyvalent cation complexes, redox-insensitive (Brannon and Sommers


174









1985). Therefore, the mechanisms dictating the stabilization of IP6 in mineral sediments

may not be equivalent to those in highly organic soils.

Biotic degradation of myo- Inositol hexakisphosphate. Four distinct classes of

enzyme have been identified with the ability to hydrolyze IP6 (Mullaney and Ullah 2007).

Their presence in soils and sediments (Quiquampoix and Mousain 2005) is linked to

their exudation by both plants (Lung et al. 2008) and, more importantly, microorganisms

(Hill and Richardson 2007). Although the ability to utilize IP6 has been noted in a broad

range of bacteria, fungi and yeasts (Pandey et al. 2001), it represents a substantial

specialization and is likely to show organism specificity. Richardson and Hadobas

(1997) noted only 0.5% of soil bacteria isolates were able to utilize IP6 as a sole C and

P source, with the percentage of bacteria able to utilize IP6 increasing in the presence of

additional labile C. In addition, anaerobic bacteria isolated from rumen have been

shown to utilize IP6 (Yanke et al. 1998) and have been a significant research focus,

given the potential for dietary manipulation of livestock (Lei and Porres 2007) as a

means of combating environmental P loading (Leytem and Maguire 2007). Therefore, it

is clear that bacteria from a range of environments are able to hydrolyze IP6 although

few, if any, studies to date have attempted to identify organisms capable of degrading

IP6 within wetlands. Fungal microorganisms, an important component of many highly

organic palustrine wetlands (Ipsilantis and Sylvia 2007; Kominkova et al. 2000; Kuehn

et al. 2000) may prove to be a significant component of biological IP6 degradation, with

screening for activity in microorganisms routinely finding fungal organisms to be the

most competent in utilizing IP6 (Lissitskaya et al. 1999; Volfova et al. 1994). It is likely

that the dynamic aerobic/anaerobic interface of wetlands and the role fungi may play in


175









certain systems results in differences in the competencies of native microbial

populations to degrade IP6. Therefore, differences in microbial populations between

study sites could in itself impact the presence or turnover rate of myo-IP6.

Conclusions

It appears that biogenic P forms are relatively stable in organic freshwater wetland

soils. While the use of anion exchange membranes as a pre-extraction confounded the

ability to track the fate of polyphosphates both DNA and myo-IP6 underwent

degradation over the course of 48 days. Extracellular DNA conformed to the initial

hypothesis by showing only limited degradation under anaerobic conditions as

compared to a substantial (30%) loss under aerobic conditions. While the tentative

conclusions drawn as to the difference in stabilization of DNA under altered redox

conditions would require further analysis, this study provides a tantalizing incite into a

mechanism that could explain the predominance of phosphodiesters in wetland soils

(Turner and Newman 2005). In contrast I did not observe a rapid, or differential,

turnover of myo-IP6 under aerobic / anaerobic conditions as seen in Suzumura and

Kamatani (1995a), suggesting only limited redox sensitivity this supports recent

research that has found significant levels of IP6 under, presumed anaerobic conditions

(McDowell 2009; Turner and Weckstrom 2009) and CHAPTER 4. Allowing me to

conclude it is anaerobosis in concert with site mineralogy and organic matter content,

which appears to determine the stability of extracellular myo-IP6 in wetland soils.


176









Table 6-1. Characterization of surface (0-10 cm) soil collected for spike incubation
microcosm study.
Wetland name Blue lake, Ordway Swisher Biological Reserve
Wetland type Sumpland, Emergent herbaceous vegetation
Sampling Location 29042'38.83" N 81 59'34.80" W
Possible impacts Burning activity in surrounding uplands
Basic Characterization
Moisture content (%) 38.2
pH 4.5
Bulk density g cm-3 0.15
Organic matter (%) 84.0
Elemental concentrations
Ig g-1
Phosphorus 609
Calcium 907
Magnesium 94
mg g-1
Carbon 314.6
Nitrogen 23.1
Iron 2.7
Aluminum 15.5
t Estimated by loss on ignition 550C for 4 h

Table 6-2. Total phosphorus, after addition of biogenic P spikes, and recovery by AEM-
NMR method of all microcosms.
Total Pt AEM-NMR recovery*
9 9-1 g g-1 %
Soil (Control) 610 10.8 347 62 57 10
Soil + Spike 1 (myo-IP6) 810 6.8 530 59 65 7
Soil + Spike 2 (DNA + polyphosphate) 768 11.4 498 79 65 10
t = total P determined by TP-Ash on 4 microcosms of each type
: = Recovery by AEM-NMR method across time and aerobic/anaerobic treatment. (n=20)


177











Table 6-3. Phosphorus composition of microcosms as determined by AEM-NMR
8 to 3 ppm region$


method. Results (pg gl)


Day Condition Sp.t


PAEM*


NMR Phos-P0 Total Ortho-P myo-IP6 scyllo-
Total IP6
p9g g-1


Other Lipd-
Ptt


DNA Poly-P" Residual"


1 Anaerobic



Aerobic



48 Anaerobic



Aerobic


12
1 14
2 98

15
1 15
2 97

6
1 6
2 40

5
1 8
2 43


475 14 283


64 27 94
259 23 68
68 24 60
61 19 103
281 32 50
81 25 58
66 21 87
256 19 49
65 23 53
59 26 67
218 20 95
63 27 62


t Biogenic spike added to microcosm, 1 = myo-lnositol hexakisphospate, 2= DNA + polyphosphate
$ Anion exchange membrane recovered P
Phosphonates,
i Region 8 to 3 ppm de-convoluted to determine orthophosphate myo- and scyllo-IP6 and all other posphomonoesters
tt Phospholipids
$$ Polyphosphates
Residual P = total P (PAEM + PNaOH)


178










Table 6-4. Phosphorus composition of soil samples as determined by parallel analysis of lyophilized soil extracts using
two distinct nuclear magnetic resonance machines.
8 to 3 ppm region
Sample Machinet Phos-PT Total Ortho-P myo-IP6 scyllo-IP6 Other Lipid-P1 DNA Poly-Ptt
9 -1

Day 1, Aerobic control 600 7 268 88 67 26 87 13 56 9
soils 500 16 258 87 81 27 63 11 58 9
RSD(%) 52 3 1 14 3 22 13 3 2


Day 48, Anaerobic soils 600 13 411 86 247 23 53 14 50 6
spiked with myo-lnositol 500 12 403 80 237 23 63 21 52 5
hexakisphosphate RSD (%) 4 1 5 3 2 11 29 3 15


Day 48, Aerobic soils 600 18 395 75 230 25 65 17 53 11
spiked with myo-lnositol 500 11 404 81 251 24 48 15 57 6
hexakisphosphate RSD (%) 35 2 5 6 3 21 6 6 40
t NMR spectra acquired using; (600) 600 MHz magnet and 5 mm BBO probe; (500) 500 MHz magnet and 10 mm BBO probe.
$ Phosphonates
Region 8 to 3 ppm, split into; Total, (Ortho-P) orthophosphate, myo-IP6, scyllo-IP6 and (other) other phosphomonoesters.
Phospholipids
tt Phosphonates


179










Table 6-5. Phosphorus composition
by AEM-NMR method.


of microcosms spiked with myo-lnositol hexakisphosphate with time, as determined


Condition Dayt


PAEM


Anaerobic






Aerobic


NMR
Total


Phos-
p


8 to 3 ppm region
Total Ortho-P myo- scyllo-
IP6 IP6
pg g-1


Other Lipid-
Other ptt


DNA P-
DNA p*Y


(1.9)
(2.0)
(0.4)
(0.4)
(0.4)
(0.4)
(0.0)
(0.8)
(1.9)
(0.9)


t Day microcosm sampled
$ Anion exchange membrane recovered P average (n=2) (1 standard deviation)
Phosphonates
i Region 8 to 3 ppm de-convoluted to determine orthophosphate myo and scyllo IP6 and all other posphomonoesters
tt Phospholipids
$$ Polyphosphates
redidual P = total P (PAEM + PNaOH)


180


Residual







1 g dry weight
equivalent of fresh soil

Sample specific vol. of
DDI to bring total
water content to 10 ml

Filtered gas injected into
vials through septum,
samples purged 5 mins
every 3 days

Comp.air



C2f ree N2 -


30 ml HDPE centrifuge bottles
Rubber septum fitted into caps


2


2


2


2


2


2


Control Mixture Mixture
DDI 1 2


Equilibration
14 days


Spiked with P at 200
pg/g


Duplicates
destructively sampled
at day 1,6,18,30,48


Figure 6-1. Experimental setup for investigation of biogenic phosphorus stability under aerobic and anaerobic conditions.
Spike mixture 1 is myo-lnositol hexakisphosphate, spike mixture 2 contains DNA and polyphosphate.


X 10


X 10


Aerobic



Anaerobic


Timeline



















myo- Inositol hexakisphosphate
+ orthophosphate


DNA
+ orthophosphate


Sodium polyphospate
+ orthophosphate


20 10 0 -10 -20
Chemical shift (ppm)

Figure 6-2. Biogenic phosphorus compounds used in spiking experiment. Standards
were mixed with orthophosphate and MDP (6 = 17.46 ppm) to provide dual
integration standards. Matrix represents 0.9 mL DDI 0.1 mL D20.


182









Aerobic


15H O


t 0


Anaerobic


b
a+ +


a


0,





I..C
0

CU
0.
e-
0t



t0



LU
LU


Control











Mixture 1












Mixture 2


0 10 20 30 40 50 6 10 20 30 40 50
Time (days)
Figure 6-3. Exchangeable phosphorus, determined by anion exchange membranes
during microcosm study. Superscript letters indicate homogeneous subsets
(Tukey HSD, ANOVA p < 0.01)


183


a

b
<>


a


o b


10-

5

0-

120-

100-


a



b d
4,

























Spike mixture 2


ISpike mixture 1




Control


20 10 0 -10 -20
Chemical shift (ppm)
Figure 6-4. Example solution 31P NMR spectra of soil samples spiked with biogenic phosphorus. Samples represent soils
extracted on day 1 and previously equilibrated for 2 weeks under anaerobic conditions.



















myo-lnositol hexakisphosphate
5 axial 1 equitorial (pH>12)


Recombined spectra



Deconvoluted line fit



Original Spectra


scyllo-lnositol hexakisphosphate
6 axial (pH>11)


Peak assignment
(ppm)
Average SD
- Orthophosphate 6.200 0.007

-- myo-IP P2 6-031 0-014

P4,P6 5.111 0.042
P1,P3 4-761 0-012
P5 4-621 0-015
-- scyllo-IP, P123.4.5. 4.272 0.011
Other phosphomonoesters


I I I I I I I I I I I I I I i
7 6 5 4
Chemical shift (ppm)
Figure 6-5. Spectral deconvolution and peak assignments in 8 to 3 ppm region of solution 31P NMR spectra. Exemplar
spectra of reconstituted anaerobic control soil extracted at t = 1 day. Peak assignments represent average and
standard deviation for all t = 1 and t = 48 day samples analyzed using 5 mm BBO probe and Avance II Brucker
600 MHz system. Model conformations and spectra assignments based upon Turner and Richardson (2004), P
= phosphate functional group.


185


I


n ,I i
= =,. I i




















I ke mixture 2





Spike mixture 1





Control



7 6 5 4
Chemical sift (ppm)

Figure 6-6. Detail of 8 to 3 ppm region of NMR spectra gathered on soil spiked with
biogenic phosphorus. Spectra represent soils extracted on day 1 of study and
previously equilibrated for 2 weeks under anaerobic conditions.


186


















600 MHz,
5 mm BBO Probe





500 MHz,
10 mm BBO Probe


KWCAVO


20 10 0 -10 -20
Chemical shift (ppm)


Figure 6-7. Solution 31P NMR spectra, including the region 8 to 3 ppm in detail, from alkaline soil extracts. Soil extract
represents control anaerobic microcosms, day 1, part of incubation study. Spectra processed using 8 Hz line
broadening and referenced and scaled using MDP (6 = 17.46) shown here offset for clarity.


7 6 5 4












300-

,0 ""**-.. -. -. .. -- ,.
S..... .-,...............


0)
0- 200-
II



H ] Aerobic conditions
150- Anaerobic conditions
Aerobic (R2 = 0.334; p = 0.308)
SAnaerobic (R2 = 0.713; p = 0.070)

100
0 10 20 30 40 50
Time (days)


Figure 6-8. Concentrations of myo-lnositol hexakisphoshate as determined within
microcosm soils under aerobic and anaerobic conditions for up to 48 days.


188









CHAPTER 7
PHOSPHORUS TRANSFORMATIONS DURING DECOMPOSITION OF WETLAND
MACROPHYTES1

Introduction

Decomposition of detritus influences the biocycling, retention, and downstream

release of nutrients in wetland systems. The often cited model of wetland macrophyte

decomposition set out by Webster and Benfield (1986) identifies three distinct yet

overlapping phases of decomposition: an initial rapid leaching of water-soluble

components, microbial colonization and decomposition, followed by the mechanical and

invertebrate mediated fragmentation of material. Of these, the second stage microbial

colonization and decomposition represents the most dynamic alteration of nutrient

forms. Whether microbes mineralize or sequester inorganic nutrients during

decomposition of senesced biomass has implications for nutrient dynamics in wetlands

(Reddy et al. 2005) and for nutrient sequestration, an important function in treatment

wetlands (Alvarez and Becares 2006).

Phenological characteristics of the litter appear to influence initial decomposition

processes, after which decomposition rates are increasingly governed by gross nutrient

ratios (Enriquez et al. 1993) and the nutrient status of the environment (DeBusk and

Reddy 2005; Rejmankova and Sirova 2007). Anthropogenic perturbation of nutrient

availability in aquatic systems can cause shifts in trophic status (Khan and Ansari 2005),

changes in the composition of plant communities (Hagerthey et al. 2008; Vaithiyanathan

and Richardson 1999), and alteration of microbial eco-physiological processes

(Corstanje and Reddy 2006; Corstanje et al. 2007; Wright and Reddy 2001a, 2008).


1 Submitted in a modified format 2010


189









Observed alterations in catabolic processes appear to follow predicted changes in

resource reallocation (Allison and Vitousek 2005; Sinsabaugh and Moorhead 1994) with

an increase in bioavailable P leading to a reduction in investment in P acquisition, such

as a reduced release of extracellular phosphatase enzymes by microbes (Newman et

al. 2003; Penton and Newman 2008; Wright and Reddy 2001 b).

Although numerous studies have investigated factors that influence changes in

tissue total P during wetland macrophyte decomposition (Brinson 1977; Corstanje et al.

2006; Davis 1991; Fennessy et al. 2008; Grace et al. 2008; Quails and Richardson

2000; Rejmankova and Houdkova 2006), transformations of chemical forms during

decomposition are not well understood. For terrestrial systems, attempts have been

made to track temporal changes in P functional groups from select plant materials, with

efforts demonstrating both the accumulation of microbial phosphodiesters (Miltner et al.

1998) and the fungal synthesis of polyphosphates (Koukol et al. 2006). Other studies

have sought to partition biogenic P in soils between various microbial and plant sources

(Makarov et al. 2002a; Makarov et al. 2005), and transposed position within a soil profile

for time, attempting to track general transformations within forest soils during organic

matter decomposition (Gressel et al. 1996). In wetlands, previous attempts have been

made to characterize leachate from macrophyte leaves (Pant and Reddy 2001), but

changes in P forms in the autochthonously derived organic matter of wetland systems

remain poorly understood.

The objectives of this study were to determine how litter quality and site

characteristics impact changes in the P content of decomposing macrophyte leaves. In

addition to tracking changes in total P, solution 31P nuclear magnetic resonance (NMR)


190









spectroscopy was used to identify changes in the forms of P during decomposition. It

was hypothesized that in a P-enriched setting, the accumulation of microbial biogenic P

and alteration of macrophyte P would result in net P sequestration, whereas in an

oligotrophic setting there would be close coupling of biogenic P production and its

subsequent hydrolysis, limiting the accumulation of microbial derived P forms.

Materials and Methods

Site Description

Water Conservation Area 2A (WCA-2A) is a diked and hydraulically controlled 424

km2 portion of the northern Everglades, characterized as a freshwater peat system

underlain by limestone bedrock. Historically, productivity in the northern Everglades has

been limited by P availability, but as a result of anthropogenic loading from upstream

agricultural practices WCA-2A has developed a distinct nutritional and concomitant

vegetation gradient (Jensen et al. 1995; Koch and Reddy 1992; Vaithiyanathan and

Richardson 1997). There is a distinct transition from native Everglades marsh

dominated by sawgrass (Cladiumjamaicense Crantz), to dominance by cattail (Typha

domingenis Pers.) in areas impacted by nutrient-rich inflow water (Vaithiyanathan and

Richardson 1999). The nutrient-enriched areas have increased rates of heterotrophic

decomposition (Davis 1991; DeBusk and Reddy 2005; Wright and Reddy 2001a) and a

reduced extracellular phosphatase activity (Corstanje et al. 2007; Wright and Reddy

2001 b).

Two sites (Table 7-1, Figure 7-1) along a P gradient were selected for this study:

one enriched site 1.3 km from the northern inflow structure S10-C, and a second from

within an area considered unenriched by P loading. Site detritus and surface soil

(diameter 15 cm x 1.5 cm) were sampled (n=4) from the study locations prior to


191









experimental setup and analyzed for total C, N, P as well as P composition by solution

31P NMR spectroscopy.

Study Design

Phosphorus dynamics during leaf litter decomposition were assessed using a

standard litterbag approach. Samples represented archived samples from a study

implemented and carried out by Dr. Patrick Inglett (2005), in which a full factorial

experiment was established at both enriched and unenriched locations using leaf litter

of Typha and Cladium from both enriched and unenriched regions of WCA-2A. Samples

were retrieved over 15 months to track changes in litter quantity and P content. Select

litter samples were further analyzed by solution 31P NMR spectroscopy to determine

changes in the forms of P during decomposition.

Litterbag Study

Senescent leaf material of both Typha and Cladium were collected in bulk from

locations proximate to both the enriched and unenriched sites. Material included only

standing dead intact lamina of unknown age, but presumed < 2 y. Bulk material was

rinsed of adhering particulates, cut into 10 cm sections and dried at 60C to a constant

weight. Litter was analyzed for total C, N, P and P forms by solution 31P NMR

spectroscopy (see: Analysis of biogeochemical properties) as well as litter quality using

a modified proximate forage analysis (Goering and Soest 1970) utilizing the semi-

automated Ankom A200 (Ankon Technology, Fairport, NY).

Litterbags (15 cm x 15 cm) were constructed from grey polyethylene mesh (1 mm)

allowing entry of micro fauna, in addition to local microbial assemblages. Although

recognized to provide an altered biogeochemical environment for decomposition

(Bradford et al. 2002), the use of litterbags provides a useful method in the study of


192









time-dependent detritus processing. Approximately 15 g dry weight of homogenized

litter was sealed into individual litter bags. Replicates of each litter type were then

contained within a larger bag (2.5 cm mesh) to aid recovery. Replicate bags were

placed (January 2003) on the surface of Typha (enriched site) or Cladium (unenriched

site) stands and secured with a polyethylene stake. Three replicate bags were

recovered from each site after 16, 33, 75, 204, and 454 days. Litterbags were removed

from the larger mesh bag and returned on ice to the laboratory, where they were gently

washed with deionized water to remove surface debris, frozen, and lyophilized. Litter

was then removed from the bags for mass loss determination and ground to pass a 2

mm mesh using a Wiley mini mill (Thomas Scientific, Swedesboro, NJ). Samples of

coarsely-ground material were then analyzed for total P and P composition, and after

further grinding, total C and N. All samples were stored in sealed containers under dark

ambient lab conditions until analysis.

Analysis of Biogeochemical Properties

Initial site detritus, surface soils and recovered litter were analyzed for total N and

C simultaneously using a Costech Model 4010 Elemental Analyzer (Costech Analytical

Industries, Inc., Valencia, CA). Loss on ignition (LOI) and total P were determined using

a modified ashing method (See APPENDIX C). Samples were weighed (~200 mg) into

borosilicate scintillation vials and ignited at 550C for 4 h. The remaining ash was

dissolved in 1 M H2SO4 and shaken for 24 h. Solutions were diluted and analyzed for P

by standard molybdate colorimetry (USEPA 1993) using a discrete autoanalyzer (AQ2+,

SEAL Analytical, UK). This method gave recoveries on a range of certified standards

comparable the more labor-intensive method of Andersen (1976).


193









Phosphorus Composition

To identify P forms present, initial site detritus, surface soil and select leaf litter

recovered from the decomposition study were analyzed by solution 31P NMR

spectroscopy. Given practical and financial constraints of analyzing multiple samples

with 31P NMR spectroscopy, only Cladium and Typha sourced from the enriched site

were analyzed. Samples were analyzed from time steps that were considered able to

provide insight into mechanistic processes (i.e. initial and final material, and mateiral

after maximum total P leaching). Additional samples from the enriched study site were

then analyzed to provide information on accumulation rates.

Phosphorus extraction in NaOH-EDTA

A standard alkaline extraction (Cade-Menun and Preston 1996; Turner et al.

2007b) was applied to detritus, soil, and litter samples. In brief, field replicates were

combined on an equal weight basis to give one homogenized sample of 3 g, weighed

out into a 250 mL HDPE centrifuge bottle. To this, 90 mL of a solution containing 0.25 M

NaOH and 50 mM EDTA was added, to give a 1:30 solid to solution ratio. Samples were

shaken at ambient room temperature for 3 h and then centrifuged (Sorvall RC6

centrifuge, SLA 1500 Rotor; Thermo Fisher Scientific, Waltham, MA, USA) at 6500 rpm

for 10 min. A subsample (20 mL) of the supernatant was removed to a scintillation vial

and combined with 1 mL of 50 mg P L-1 methylenediphosphonic acid (MDP) as an

internal standard (Turner 2008b). Mixed samples were immediately frozen (-80 C) and

lyophilized prior to NMR spectroscopy. A second subsample was analyzed for total P

(NaOHTp) by a modified double-acid digest using H2SO4 and HNO3 (Rowland and

Haygarth 1997) and a modified discrete molybdate colorimetric method (see above).

Residual P not recovered by the alkaline extraction is by definition unidentified and its


194









chemical stability presumably indicates its recalcitrance in the environment. For mineral

soils it has been assigned as recalcitrant organic (Cade-Menun 2005b) or alkali-stable

inorganic P (Turner et al. 2007a), although there is little information for wetland soils.

Solution 31P nuclear magnetic resonance spectroscopy

Spectra were acquired using a Bruker Avance 500 Console with a Magnex 11.75

T/54 mm magnet using a 10 mm BBO Probe. Lyophilized samples (~300 mg) were re-

suspended in 0.3 mL D20 and 2.7 mL of a solution containing 1 M NaOH and 0.1 M

EDTA, vortexed, and then transferred to a 10 mm tube. Spectra acquisition was carried

out at a stabilized 250 C with a calibrated (~300) pulse length, a zgig pulse program, and

a 2 s T1 delay. Results presented here are of ~30,000 scans accumulated as three

sequential experiments, with FIDs summed post acquisition by Bruker proprietary

software.

Spectra interpretation was carried out using Acorn NUTS (NMR Utility Transform

Software) (Acorn NMR Inc, Livermore, CA). After applying 15 Hz line broadening,

spectra were referenced and integrated against the internal standard, MDPA, set as 6 =

17.46 ppm (compared to externally held 85% H3P04, 0 ppm). Integration over set

spectral windows were chosen to correspond with known phosphorus bonding

environments (Turner et al. 2003d). The region between 8 and 3 ppm was further

elucidated after 3 Hz line broadening by a deconvolution subroutine applied to identify

and quantify orthophosphate (6.21 0.02 ppm) among orthophosphate monoesters.

Data Analysis

All statistical tests were performed in SPSS for windows version 17.0.0 statistical

software (SPSS Inc. 2008). Mass remaining, P concentration and mass of P were

analyzed by a 4 way univariate ANOVA using site of decomposition (site), species of


195









litter (species), source of litter (source) and time as independent variables. Given the

homogeneous nature of initial material, time = 0 was excluded from the statistical

analysis. Resulting model residuals were analyzed via P-P plot and visual inspection for

normality. If required, original data were normalized using a natural log transformation.

Significant results are reported alongside their calculated partial eta2 values. Given

complications of interpreting multiple interaction factors, non-significant higher order

terms are not presented here. Simple decomposition rates were calculated using an

exponential decay model (Webster and Benfield 1986). Linear regression of changes in

the mass of leaf litter P over time excluded data from the initial equilibration and

leaching period. Due to practical constraints, solution 31P NMR data were acquired for

pooled samples and as such are without a measure of variance at individual time steps.

All integrated results using MDP as an internal standard were within 20% of

determined NaOHTp concentrations. Calculated rates of accumulation of specific P

forms are based upon simple linear regression after initial leaching and equilibration

period.

Results

Initial Litter Material

Typha litter material from both study sites contained greater concentrations of total

N and P, as well as a higher LOI (indicating lower biosilica and mineral content) than

Cladium (Table 7-2). Although Typha litter had a noticeably larger proportion of neutral

detergent-extractable C, the proportion of lignin and cellulose was similar, with LCI

(Lignin to Cellulose Index) between 0.170 and 0.197 for all litters sampled. Within

species, there was a distinct difference between material collected at the enriched and

unenriched site, with Cladium showing an approximate four-fold increase in total P and


196









Typha showing a two-fold increase, respectively. Using both species, collected from

enriched and unenriched settings allowed four litter types to be used within the

subsequent decomposition study.

Mass Loss

Between 36 and 70% of original mass of litter remained after 15 months in the

field (Figure 7-2). Rates of decomposition varied significantly between litter types, in

terms of both macrophyte species and litter source, as well as between study sites

(Four way Univariate ANOVA, Table 7-3). In addition, decay rate constants based upon

a simple exponential decay (Table 7-4) followed expected patterns, with increased

mass loss in response to increased litter quality and exogenous nutrients (Corstanje et

al. 2006; Rejmankova and Houdkova 2006). The higher 'quality' Typha litter material

decomposed more rapidly than the Cladium litter (ANOVA; p < 0.001), with enriched

material of both species showing significantly higher rates of mass loss (ANOVA; p <

0.001). All litter materials showed significantly (ANOVA; p < 0.001) higher mass loss at

the enriched site. However, decomposition rate constants were relatively constrained,

varying from 0.00275 day-' for enriched Typha material at the enriched site to 0.00136

day-' for unenriched Cladium at the unenriched site. Significant interaction terms,

between time and both litter species and source demonstrated the continued and

differential influence of initial material characteristics on the long-term stability of organic

matter.

Phosphorus Content

Total P concentrations (Figure 7-3) changed significantly with time. At the enriched

site all litter types except unenriched Cladium showed an initial drop in P, as from a

rapid leaching event, but then subsequently increased to between 5 and 10 times the


197









original P concentration. In contrast, at the unenriched site after an initial decrease in P,

concentrations did rise, but with the exception of unenriched Cladium never recovered

to their initial levels (Figure 7-3 B). Univariate four-way ANOVA (Table 7-5) showed

significant differences not only dependent upon the site of decomposition (p < 0.001),

but on the nature of the initial litter, as shown by highly significant effect of both litter

species (p < 0.001) and source (p < 0.001). Significant interaction terms (litter quality

with time) suggested different responses in P concentration of litter types over time.

This relationship was explored by applying a two-factor univariate ANOVA using litter

type (enriched Cladium, unenriched Cladium, enriched Typha, and unenriched Typha)

and site at each time step. Post hoc analysis (Tukey HSD) showed that at 16 days there

were significant differences in the total P concentrations between litters of the same

species sourced from different locations. In contrast, at day 454 there were no

significant differences in the total P concentrations in Cladium, yet still a significant

distinction for Typha (p < 0.05) and between Typha sourced from enriched and

unenriched locations (p < 0.001).

Changes in P concentration during decomposition may be a result of two distinct

processes: a change in litter mass that results in alteration of the endogenous P

concentration, or the loss or gain of P from the environment. The distinction within this

study was explored by plotting changes in the mass of P over time (Figure 7-4). Given

an inability to interpolate data and the potential for sites to experience variation in initial

flooding I draw a conservative breakpoint at 33 days, marking the switch from initial

abiotic processes and site equilibration to long-term microbial action. From 33 days to

the end of the study there was a distinction in changes in mass of P between sites.


198









Linear regression (Table 7-6) demonstrated that the mass of P in all litter at the

enriched site increased significantly over time (p < 0.001). At the unenriched site, the

mass of P in both enriched Cladium and unenriched Typha did not change significantly

with time. The P content of unenriched Cladium increased slightly, yet significantly (B=

0.037 pg P g-1 initial material day-'; p = 0.001) during decomposition, whereas in

enriched Typha there was actually a significant (p < 0.001) decrease in the mass of P.

Phosphorus Composition

Analysis of initial leaf litter demonstrated a range of P forms present within both

Cladium and Typha (Figure 7-5). In both cases, NaOH-EDTA extractable P was

dominated by orthophosphate (35 and 39% of total litter P in Cladium and Typha,

respectively). In addition, considerable amounts of phosphomonoesters (13 and 23%

total litter P), phosphodiesters (4 and 9% total litter P), and inorganic pyrophosphate (2

and 6% total litter P) were identified. It should be noted that there was probably an

inherent bias in the analysis, because some phosphodiesters (i.e. RNA and some

phospholipids) decompose to phosphomonoesters in alkaline solution during extraction

and analysis (Turner et al. 2003d).

Comparison of initial P composition at the two study sites (Figure 7-6)

demonstrated distinct differences in the forms and proportions of P identified. Standing

detritus and surface soil from both sites contained orthophosphate,

phosphomonoesters, phosphodiesters (dominated by DNA) and pyrophosphate.

However, only detritus from the enriched site contained long chain inorganic

polyphosphate (~3% of total P). Concentrations of all P groups identified were higher in

detritus than in soil at both sites, and higher in enriched than in unenriched sites. There

was a distinct shift in the ratio of phosphomonoesters to phosphodiesters in material


199









from the enriched and unenriched sites, from 1.35 to 0.58 in detritus and 1.69 to 0.90 in

surface soil, indicating a greater proportion of P identifiable as phosphodiesters within

detritus and at unenriched sites. Signals within the phosphomonoester regions

indicative of the phosphomonoester myo-lnositol hexakisphosphate were not detected

in any sample. This is consistent with other studies of calcareous freshwater wetlands

(Turner and Newman 2005; Turner et al. 2006a), although inositol phosphates could

have been present at concentrations below the limit of detection by solution 31P NMR

spectroscopy (EI-Rifai et al. 2008).

There was a convergence in the proportion of P forms present based upon the site

of litter decomposition (Figure 7-7; Table 7-7). Identifiable P lost during equilibration

(presumably by leaching) at both the enriched and unenriched site was dominated by

orthophosphate (up to 50% of lost P), yet concentrations of phosphodiesters and

pyrophosphate also declined after 16 or 33 days. Concentrations of

phosphomonoesters increased or remained relatively stable in all litters between initial

material and samples at days 16 and 33, resulting in an increase in their proportional

contribution to the total P (Table 7-7).

Recovery of total litter P by NaOH-EDTA extraction averaged 56% for all samples.

Residual P was a relatively stable proportion of the total P, although the concentration

increased in litter at the enriched site. Trace concentrations of phosphonates and mid-

chain polyphosphate were detected in one sample (Typha from the enriched site after

16 d), but were suspected to originate from adhering cyanobacterial or phytoplankton

biomass (Eixler et al. 2006). After the initial equilibration, litter at the enriched site

accumulated all forms of P identified by 31P NMR analysis, in direct contrast to the


200









unenriched site, where there was little change in P composition after the initial leaching

period (Figure 7-8). The increase in concentration of P forms at the enriched site

showed a significant linear response (Table 7-8), with most major forms showing a net

change of between 89 and 107 mg P kg-ly-1. Changes in pyrophosphate concentration

appeared to be distinct from other forms, showing a significant (p < 0.05) linear

response at approximately half the rate of change for phosphomonoesters,

phosphodiesters or orthophosphate concentrations.

Discussion

Plant litter decomposition rates determined in this study over the course of 15

months were generally lower than values reported for other herbaceous litters

(Fennessy et al. 2008; Rejmankova and Houdkova 2006), but correspond well with

previous data from WCA-2A. For example, Debusk and Reddy (2005) found simple

decay rate constants of approximately 0.003 d-1 for Typha decomposing under P-rich

conditions. Rates of mass loss demonstrated previously established patterns with both

higher litter quality, as determined by structural C and nutrient content and higher

environmental nutrient availability, resulting in an increased rate of mass loss.

Over the course of the study, total P concentrations within litter at the two sites

showed distinctly different trajectories. At the enriched site, total P concentration in all

litters increased, resulting in an average molar C:P ratio of ~488 after 15 months. In

contrast, concentrations at the unenriched site generally did not return to the levels of

the initial material entering the system, with an average molar ratio of ~4150 after 454

days. Although conventional understanding of litter decomposition would assume a

continued sequestration of P at both sites in response to gross stoichiometry, analysis

of changes in mass of P (Figure 7-4) suggests that while microbial activity at the


201









enriched site resulted in the net sequestration of P from the endogenous environment,

at the unenriched site there was little net gain (Figure 7-9). This interpretation assumes

that mass loss due to fragmentation is minimal and may represent an underestimation

of actual P accumulation in both sites. Yet previous studies have also suggested that

oligotrophic systems such as the unenriched Everglades may present an environmental

gradient in which P sequestration predicted from detritus C:P ratios is unlikely to occur

(Longhi et al. 2008; Quails and Richardson 2000). Indeed, Quails and Richardson

(2000) demonstrated that an environmental P concentration of 5 pg P L-1 in the water

column resulted in the net loss of P to the environment from both Typha and Cladium

leaf litter during decomposition.

Although recovery and identification of initial plant P by alkaline extraction was

less than previous published values for live plant tissue (Makarov et al. 2005),

recoveries were considered good at 54 and 76% of Cladium and Typha, respectively.

Initial changes in litter P composition as a result of loss of leaching of labile P

correspond to the identification of water extractable forms by He et al. (2009), who

showed a predominance of orthophosphate, but significant organic P loss from plant

material. The apparent stability of phosphomonoesters during the initial leaching may be

due to the recalcitrant nature of phosphomonoesters remaining in plant material after

nutrient resorption and senescence of leaves, or due to the rapid synthesis of new

phosphomonoesters in establishing microbial populations. In addition to leaching from

the plant biomass, consideration should be given to the fact that some compounds (e.g.

pyrophosphate) may be lost from endophytic and aerial saprotrophic fungal biomass

during initial preparation and site equilibration (Kominkova et al. 2000; Kuehn et al.


202









2000). Indeed, the use of standing dead biomass of unknown age leaves us unable to

determine alterations of P composition during these initial stages of the decay

continuum (Kuehn et al. 1999).

Initial abiotic leaching is often considered rapid (Davis et al. 2006), yet given an

inability to interpolate data and the potential for sites to experience variation in initial

equilibration I drew a conservative breakpoint at 33 days marking the switch from initial

abiotic processes and site equilibration to long-term microbial action. After 33 days this

microbial action at the enriched site resulted in the accumulation of all forms of P

identified by solution 31P NMR spectroscopy. The estimated rates of change in P

concentrations (Table 7-8) include contributions from both stabilized original and newly

accumulated compounds. For example, the increase in recalcitrant P concentration (up

to 8-fold) is not accounted for by the passive accumulation of initial endogenous

recalcitrant P and suggests a mechanism of transformation by which accumulating

biogenic P is ultimately modified to an operationally recalcitrant organic form.

Differences in both C quality and total P of initial material appeared to be reflected

in P composition of litter material collected at the enriched site, with Typha consisting of

more labile forms. However, changes in composition during decomposition appeared to

be independent of initial material, rather changes were dependent upon site

characteristics, with major P forms showing similar changes in concentration at the

enriched site. Fungal biomass is a major contributor to heterotrophic decomposition in

emergent herbaceous systems (Hackney et al. 2000; Kuehn et al. 2000) and although

litter quality can affect the composition of the microfungal community, their generalist

nature may result in similar responses to the exogenous nutrient availability (Thormann


203









et al. 2003).

Within the enriched study site, the P composition of decomposing litter appears

to be on a trajectory towards that determined in standing detritus and surface soil. In

contrast, litter material within the unenriched site after 454 days of decomposition did

not appear to reflect standing detritus. This could reflect differences in the role of

macrophyte litter across the northern Everglades. Within nutrient-impacted portions of

WCA 2A, surface substrate originates from Typha detritus and a high standing

population of heterotrophic microbes, while oligotrophic regions often include only

minimal Cladium fragments, but a significant contribution from the periphyton

community (Wright and Reddy 2008). Further work would be needed to determine the

role of these additional sources of biogenic P in determination of P forms seen within

wetland substrates.

Conclusions

Decomposition processes are affected by both initial material and site

characteristics. The decomposition of macrophyte leaf litter results in the sequestration

of P from the environment via the accumulation of biogenic forms resulting from

microbial turnover. Low ambient P concentrations result in limited or no P sequestration

due to heterotrophic decomposition of leaf litter. Substrates in relatively enriched

herbaceous systems are presumed to contain a large proportion of modified

macrophyte leaf litter and accumulated microbial biomass. Within oligotrophic systems,

macrophyte detritus and associated heterotrophic biomass appears to have less

influence on the nature of P in surface substrates.


204









Table 7-1. Site characteristics for enriched and unenriched study sites within WCA-2A.
Detritus and soil samples average (n=4) one standard error. Overlying
water characteristics based upon published values.
WCA-2A Location


Enriched Site


Unenriched Site


Latitude (N)


Longitude (W)
Distance from inflow structure (km)
Average water depth (cm; max, min)*


2621.230
80 20.967


1.93


12.8 (11.9, 14.0)


26 16.382
8021.502


10.05


11.9 (11.0, 13.6)


Overlying watert


Detritus


Total P (pg P L- )
Ortho P (pJg P L-1)
Total P (pg g-l)


52.4 (6.7)
27.8 (4.4)
1334 (452)


Soil (0- 1.5 cm) Total P (pg g-1) 1312 (40) 468 (40)
SStage data from South Florida Water management (23/01/03 through 04/21/04). Sampling
stations WCA2E1 and WCA2U1.t Site water characteristics, South Florida Water management
(6/21/94 through 9/27/94). Sampling stations WCA2E1 and WCA2E5. Available through
DBHYDRO (http://my.sfwmd.gov/dbhydroplsql/)

Table 7-2. Characterization of litter material used within the decomposition study,


consisting of two species (Cladium and Typha) from
enriched portions of WCA-2A.


LOI (%)
Total C (%)
Total N (%)
Total P (pg g-)
Molar C:P
Forage Analysis (Ankom A200)
Neut. Det Extractable (%)
Hemi Cellulose (%)
Cellulose (%)
Lignin (%)
Un = Unenriched
En = Enriched .


Cladium
Un En
91.8 93.7
41.4 40.9
0.41 0.38
46.1 171.0
23200 6179


both unenriched and

Typha
Un En
95.6 95.6
43.1 43.2
0.45 0.54
135.4 261.2
8223 4273


205


9.6 (0.9)
1.7 (1.3)
206 (53)










Table 7-3. Four way Univariate ANOVA for mass remaining. Full factorial model
adjusted R2 = 0.973). Source and species of litter (litter quality) as well as site
of decomposition and time in the field all shown to be highly significant (p <
0.001). With litter quality parameters and site showing significant interactions
with time.
Partial eta
Source of variation DF F p-value sar e
squared
Site (S) 1 57.9 < 0.001 .420
Time (T) 4 931.8 < 0.001 .979
Source (So) 1 74.3 < 0.001 .482
Species (Sp) 1 129.0 < 0.001 .617
Tx S 4 34.2 < 0.001 .631
Tx So 4 5.4 0.01 .213
T x Sp. 4 23.3 < 0.001 .538
So x Sp 1 7.1 0.009 .082


Table 7-4. Simple exponential decay rate constant (x =100e-kt ) and leaf litter half life
calculated from material recovered over the course of 15 months within WCA-
2A (n=15).
Site Litter type k R2 T1/2
(days)
Enriched Cladium Enriched 0.965 0.00172 403
Unenriched 0.924 0.00148 468
Typha Enriched 0.943 0.00275 253
Unenriched 0.918 0.00202 343
Unenriched Cladium Enriched 0.896 0.00118 587
Unenriched 0.904 0.00093 745
Typha Enriched 0.960 0.00207 335
Unenriched 0.954 0.00136 510


206









Table 7-5. Four way Univariate ANOVA of phosphorus concentration in leaf litter.
Model adjusted R2 = 0.947. Source and species of litter (litter quality) as well
as site of decomposition and time in the field all shown to be highly
significant.


Source of variation DF


Site (S)
Time (T)
Source (So)
Species (Sp)
SxT
Sx So
Sx Sp
TxSo
TxSp
Sox Sp


Partial eta
F p value
squared
2169.4 < 0.001 .966
302.4 < 0.001 .940
438.8 < 0.001 .851
759.2 < 0.001 .908
130.4 < 0.001 .871
77.0 < 0.001 .500
86.0 < 0.001 .527
11.0 < 0.001 .363
5.6 < 0.001 .226
22.3 < 0.001 .225


Table 7-6. Linear regression of changes in mass of phosphorus within litterbags held in
the field for between 33 and 454 days.


Enriched Cladium Enriched
Unenriched
Typha Enriched
Unenriched
Unenriched Cladium Enriched
Unenriched


Enriched


Unenriched


B (pg P g
day-l)
0.637
0.685
0.447
0.566
-0.036
0.037
-0.175


-0.014


p value


0.873
0.950
0.795
0.770
0.201
0.726
0.851
0.050


0.000
0.000
0.000
0.000
0.143
0.001
0.000
0.248


207


Site


Litter type


Typha









Table 7-7. Phosphorus forms as determined by solution 31P NMR spectroscopy of
NaOH-EDTA extracts during macrophyte leaf litter decomposition
% Total P


Site Species
13


E

0
16
Cladium 75
-0 204
_c 454
0
LJ 16
Typha 75
204
454
0
Cladium 33
_-
0 454
D 0
S Typha 33
454


0)
a-


a-
0
1-

171
104
187
441
795
261
237
365
742
1257
171
67
85
261
195
239


0 0

U 0
0
0 "0
o *^



a. 0
34.7
26.9
23.5
18.7
19.5
38.5
0.3 18.4
15.1
10.9
12.3
34.7
33.0
29.4
38.5
18.9
15.5


0
o(

0
E
0
c
0.
0
a.




13.9
16.4
22.6
32.7
21.5
12.6
14.9
12.8
15.2
29.2
22.6
29.9
17.5


z

3.3
1.9
4.1
3.8
10.0
4.7
1.6
5.2
4.3
8.3
3.3
1.3
6.9
4.7
1.4
4.4


0
oC
0
0
a-



1.1
1.9
1.5
3.0
4.8
4.1
11.8
5.1
2.6
4.0
1.1
9.1
4.9
4.1
4.6
1.2


C.
U,
0
..
0

a.

2.5
1.4
5.3
3.9
4.4
5.6
1.0
2.7
2.6
3.7
2.5
13.6
0.0
5.6
3.7
2.4


a


0

0- 0

45.7
41.9
48.9
56.8
44.9
24.5
2.4 34.5
50.3
67.1
56.7
45.7
27.8
29.4
24.5
41.4
58.9


Table 7-8. Coefficients (one standard deviation) of linear increases in concentrations
of major phosphorus forms identified within leaf litter during decomposition at
the enriched study site.


Cladium


Typha


Orthophosphate 0.292* (0.02) 0.256 (0.018)

Phosphomonoesters 0.244* (0.20) 0.260* (0.054)

Phosphodiesters 0.265* (0.39) 0.285* (0.057)

Pyrophosphate 0.072* (0.006) 0.099* (0.006)
* Significant at the 0.05 level
** significant at the 0.001 level


208




















































Figure 7-1. Location of chapter 7 study sites within Water Conservation Area (WCA)-2A
in the northern Everglades. Before recent hydrological diversions through
Storm Water Areas, flow was predominantly north-south carrying nutrient
impacted water from the Everglades Agricultural Area.




209












100-


80-

S60-

*@ 40-
E
100-

a 80-
5.


60-

40-


100


I I
200 300
Time (days)


400


500


Figure 7-2. Mass remaining of four litter types placed at two distinct sites within WCA-
2A and recovered at time intervals up to 454 days. Symbols represent
averages (n=3) with error bars showing one standard error. Significant
differences in decomposition rates between sites and between litter types
(see Table 7-3) were reflected in the range of decay rates calculated for litters
used (Table 7-4).


210


* Enriched Cladium
* Erriched Typha
11 Unenriched Cladium
o Unenriched Typha


an
B


Unenriched Site
a



U

o


Enriched Site












1,500-


I-
01,000-



%.0 0
500-
0


r0 1 I 10 I5
0 100 200 300 400 500


300




200




100


0 100 200 300 400 500


Time (days)

Figure 7-3. Changes in litter phosphorus concentration over time at, A) enriched site and, B) unenriched site. Symbols
represent averages (n=3) with error bars showing one standard error.


211


A

Site detrtus totl P --






Enriched Cladium
Enriched Typha
l Unenriched Cladium
O Unenriched Typha


B


4
-----~-------------------------*-

o


0
o
* Site detritus total P






rn













S400-
S300-
E 200-
100-
0-
S-100-
S-200-
a-
m 400-
300-
a- 200-
S100-
0-
CU
o -100-


-200


100


200


300


400


500


Time (days)


Figure 7-4. Changes in mass of phosphorus in macrophyte leaf litter held within
litterbags over the course of 454 days. Data standardized for initial mass of
material placed in litterbags. Symbols represent averages (n=3) with error
bars showing one standard error.


212


Enriched Site


I:

w i
*3=1
t* -


Unenriched Site

Enriched Cladium
Enriched Typha
l Uneriched Cladium
o Unenriched Typha


B 7-- EF


1g----------
0


I' -


i


I


r












Phosphorus (pg P g-1 ) Cladium Typha

Orthophosphate 59.3 100.5

Phophomonoesters 21.9 59.1

DNA 5.7 12.3

Other phosphodiesters 1.8 10.7

Pyrophosphate 4.2 14.6
Residual 78 64
TP-Ash 171 261

Recovery by solution 31P NMR (%) 54 76






V Typha


Cladium
I I I I
10 0 -10 -20
Chemical shift (PPM)

Figure 7-5. Initial phosphorus composition of Typha and Cladium leaf litter sourced from the enriched portion of WCA-2A.


213











Phosphorus (pg P g-' )


Enriched Site
Detritus Soil


Unenrichedsite
Detritus Soil


Orthophosphate

Phophomonoesters

DNA

Other phosphodiesters

Pyrophosphate
Polyphosphate
Residual

TP-Ash

Recovery by solution 31P NMR (%)


10 0 -10 -20
Enriched Site C


Detritus

Soil


lemical shift (ppm)


10 0 -10 -20
Unenriched Site


Figure 7-6. Initial phosphorus composition of detritus and surface soils from enriched and unenriched study sites sampled
on (10/20/03). Spectra plotted using 15 Hz line broadening and scaled using height of internal standard MDP.


214


142.9

109_0

59.2

21.5
45.3
31.2
520

899

42


238.0

136_8

65.2

15.8
16.2
n.d.
840

1311

36


58.8

19.7

28.0

6.2
10
n.d
123

221

56


76.8

29_0

26.5

5.9
7.7
n.d
279

421

35













Enriched Site


Unenriched Site

i


M pin,-" W---- -- --,-- Day 454

Day 204

-- ----- --- -- ------------------------- Day 75
-------------- Day 33

------ Day 16

SF I Day 0
~ I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I '
10 0 -10 -20 10 0 -10 -20
Chemical shift (ppm)
Figure 7-7. Example solution 31P NMR spectra showing changes in phosphorus composition of Typha leaf litter during
decomposition at both an unenriched and enriched site over the course of 454 days Spectra plotted using 15 Hz
line broadening and scaled to height of MDP.


215










200


A DN ID
1n X
15-G- U
15 ,-- -' A







I 20 0-

4-
0 = Orthophosphate
10 78 =i Phosphomonoesters w
o15 X =DNA O
L Pyrophosphate

100-

50- -
-- ------------------. M


0 0 0 0 0 00 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
Ir- C< CM t I.O -- (N CeO "I LI)
Time (days)
Figure 7-8. Changes in proportion of major P pools found within macrophyte leaf litter over the course of 454 days of
decomposition in WCA-2A.


216


Clardiuim


Tvnha













B) Unenriched Site


WATERCOLUMN

WA-


Net P Flux


I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
111 1


WATERCOLUMN
[--


Figure 7-9. Conceptual model of phosphorus turnover in wetland macrophyte detritus under, A) enriched and, B)
unenriched conditions. 1= phosphate uptake, 11= organic matter accreation, III = re-suspension and flux, IV
=senecence, V= enzymatic hydrolysis


217


Net P Flux


A) Enriched Site









CHAPTER 8
PHOSPHORUS FORMS AND DYNAMICS ALONG A STRONG NUTRIENT
GRADIENT IN A TROPICAL OMBROTROPHIC WETLAND

Introduction

The tropical wetlands of Central America are recognized as ecosystems of social,

cultural and economic importance (CCAD 2002; Ellison 2004). Yet their inaccessible

nature has meant that few detailed investigations have been carried out on their

formation, ecology and biogeochemistry (Phillips 1998). Such an understanding is

critical in determining the implications of increasing direct and indirect anthropogenic

pressures upon them (Limpens et al. 2008; Phillips 1998) and those systems linked

through hydrologic or nutrient cycles (D'Croz et al. 2005). Tropical peat domes are 'self

emergent' organic wetlands within the humid tropics (Semeniuk and Semeniuk 1997).

Although historically associated with the swamps of maritime south east Asia, significant

peat deposits are present though out the Caribbean coastal plain (Ellison 2004; Phillips

and Bustin 1996) and the tropical Americas (Lahteenoja et al. 2009b). They represent

systems which have an organic layer greater than 50 cm and a organic content >75%

(Andriesse 1988) their upper surface shows a pronounced convex morphology leading

to their hydrologic isolation from surrounding riverine, or marine, systems and a truly

ombrotrophic state (Anderson 1983). Although the water table may fluctuate within the

upper peat layer (acrotelm), in unimpacted peat domes it remains close to, or above,

the surface for the entire year resulting in surface flow and lateral seepage from the

central portion. This is presumed to result in significant nutrient redistribution towards

the periphery. Current conceptual models of peatdome development and maintenance

suggest a complex adaptive system, with multiple feedback processes active at different

spatial and temporal scales (Belyea and Baird 2006; Belyea and Clymo 2001) resulting


218









in a predictable progression of concentric vegetation types (or phasic communities)

across the dome surface. Although changes in total P content have been noted as part

of a general trend in declining nutrient status towards the center of tropical peat domes

and between phasic vegetation types (Phillips et al. 1997; Phillips 1998; Troxler 2007).

Little is known about the nature and cycling of P within these tropical systems

(Sjogersten et al. 2010). In addition to implications on in situ vegetation dynamics, the

nature and bioavailability of P will have direct implications upon nutrients exported to

aquatic systems downstream from these wetlands.

Within established ombrotrophic wetlands it can be assumed that the vast majority

of organic matter within surface substrates is autocanthously derived. Phosphorus

inputs within this organic matter, consists of a variety of forms dependent upon the

nature and structure of the biotic community (Harrison 1987; Koukol et al. 2008;

Makarov et al. 2005). Although, given rapid microbial-mediated processes (Oberson

and Joner 2005) it is unclear as to the balance between P forms derived from primary

eukaryotic inputs and those as a result of microbial turnover and its standing biomass.

The turnover of biogenic P is dependent upon its stability within the organic matrix, in

part dependant on the activity of extracellular hydrolytic enzymes. Although commonly

applied assays for hydrolytic enzyme activity in soils may not reflect actual flux rates

from organic P pools (Wallenstein and Weintraub 2008) close coupling to the nutrient

status of biota in wetlands (Caldwell 2005; Wright and Reddy 2001b) makes them a

useful indicator of P limitation and relative rates of pool turn over.

In this study, I aimed to identify changes in P forms and potential pool turnover in

the surface soils of a tropical ombrotrophic peat dome. It was hypothesized that altered


219









vegetation communities and nutrient availability would influence both the nature of P

inputs and microbial turnover across a gradient from the periphery to the center of the

Changuinola peat dome. Specifically, I believed that there would be a monotonic

change in the functional nature of P forms identified by solution 31P nuclear magnetic

resonance spectroscopy because of altered environmental conditions.

Methods

Study Site

The Changuinola peat deposit, part of the internationally recognized San san pond

sak wetland in Bocas del Toro province NW Panama (Ramsar 2009, Site #611)

represents a near pristine example of a raised ombrotrophic peat dome (Cohen et al.

1989, 1990). Palynological work (Phillips and Bustin 1996; Phillips et al. 1997) suggests

that Caribbean coastal ombrotrophic systems, such as the Changuinola deposit, may

develop by different mechanisms and contain distinct community types than those

identified in the well studied peat domes of maritime south east Asia (Anderson and

Muller 1975). Yet there are still strong similarities with a visible soil-vegetation catena

across the convex surface, mirroring the site development over time (Phillips et al.

1997). Current vegetation communities range from monodominant Raphia taedigera

palm swamp, at the periphery, through mixed and monodominant Campnosperma

panamensis forest swamp to a central 'bog plain' community dominated by herbaceous

species such as sawgrass (Phillips et al. 1997; Sjogersten et al. 2010)

Sampling

A study transect was established (Sjogersten et al. 2010; Troxler 2007) making

use of an original leveling survey transects (Cohen et al. 1989). Site access runs almost

perpendicular from a bordering canal towards the geodesic center of the peatdome


220









(Figure 8-1, Table 8-1). Nine sampling sites were established at 300 m intervals along

the linear path of the access route. Soil sampling, carried out in Sep 07, consisted of

collecting three repeat samples from within 20 m of each site. With each repeat being

an amalgamated sample of three surface cores (diameter 7.5 cm, 0-10 cm) collected

from within 2 m of each other using a sharpened metal cutting head upon a ridged

polycarbonate tube. Samples were immediately transferred to pre-labeled ziplock bags

and put on ice for transportation back to the lab. Samples were transferred to Panama

city, with initial sample processing occurring within 72 h of field sampling. Initial

processing involved homogenizing and the removal, by hand, of recognizable roots (>

approx diameter 1 mm) and lignified structures, seeds twigs etc. It should be noted that

due to the very high concentration of fine roots from herbaceous species at sites eight

and nine a significant amount of root biomass might have been retained. Samples were

subsequently split with half being air dried (~22C 10 days) to constant weight and the

remainder being held at 4C in sealed ziplock bags (subsequently referred to as fresh

sample). Air dried samples were ball ground using Tungsten carbide vessels, before

storing in airtight containers under ambient lab conditions until analysis.

Soil Properties

Soil moisture was calculated on fresh samples by gravimetric loss after drying at

105 C for 24 h, pH was measured on fresh samples using a strict 1:20 soil to water

ratio and standard glass pH electrode. Total elemental concentrations were determined

on dried and ground samples. Total soil C and N by combustion and gas

chromatography using a Flash EA1112 (Thermo Scientific, Waltham, MA), total P, Ca, K

and Mg after a standard H202 + H2SO4 digestion procedure (Parkinson and Allen 1975)

and ICP-OES (Optima 2100, Perkin-Elmer Inc., Shelton, CT).


221









Phosphorus Characterization

Anion exchange membranes

Available and microbial P was determined using anion exchange membrane

(AEM) extraction (Cheesman et al. 2010b; Kouno et al. 1995; Myers et al. 1999) using

one 6.25 x 1.5 cm AEM strip (BDH Prolabo Product number: 551642S, VWR

International, UK), preloaded with bicarbonate and eluted in 0.25 M H2S04. The

concentration of eluted phosphate was determined by automated molybdate colorimetry

with detection at 880 nm using a flow injection analyzer (Lachat Quickchem 8500, Hach

Ltd, Loveland, CO) and difference between non fumigated and hexanol fumigated

samples attributed to fumigation release of microbial P. It should be noted that in

addition to inorganic orthophosphate the membrane strips used may have recovered

labile organic and inorganic poly-phosphoric P (Cheesman et al. 2010b).

Solution 31P nuclear magnetic resonance spectroscopy

Phosphorus composition in soil samples was determined by standard alkaline

extraction (Cade-Menun and Preston 1996; Turner et al. 2007b) and 31P NMR

spectroscopy. The standard alkaline extraction (0.25 mol L-1 NaOH and 50 mmol L-1

EDTA) was applied to air dried soils with shaking for both 4 and 16 h. As well as on both

fumigated and non-fumigated fresh samples after application of the AEM strips, see

above (Cheesman et al. 2010a). All samples were shaken at a 1:30 soil to solution ratio

at ambient room temperature for 4 or 16 h before being centrifuged and the supernatant

removed. Due to the prohibitive costs of 31P NMR analysis field replicates were

combined on an equal volume basis (15 mL) and combined with 1 mL of an internal

standard methylenediphosphonic acid (MDP) (50 mg L-1) mixed, immediately frozen (-

800 C) and lyophilized to await re-suspension and spectra acquisition. A second


222









independent subsample was analyzed for total P (NaOHTp) by a modified double acid

digest using concentrated H2SO4 and HNO3 (Rowland and Haygarth 1997) and analysis

for molybdate reactive P using automated molybdate colorimetry.

Spectra were acquired using a Bruker Avance 500 Console with a Magnex 11.75

T/51mm Magnet, using a 10 mm BBO Probe. Lyophilized samples (~300 mg) were re-

suspended 2.7 mL (1 mol L-1 NaOH 0.1 mol L-1 EDTA) and 0.3 mL D20 before vortexing

and transfer to a 10 mm probe. Spectra acquisition was carried out at a stabilized 250 C

with a calibrated (~300) pulse length, a zgig pulse program and a 2 s T1 delay. Results

presented here are of ~40,000 scans accumulated as 4 sequential experiments with

FID's summed post acquisition, by Bruker proprietary software. In addition, exemplar

spectra from alkaline extractions after application of the AEM's are presented. These

were determined using a 5-mm NMR probe and a Bruker Avance DRX 500 MHz

spectrometer (Bruker, Germany) using a 6 ps pulse (450), a delay time of 1.0 s, an

acquisition time of 0.4 s, and a zgig pulse program.

Spectra interpretation was carried out using wxNUTS vr 1.0.1 for Microsoft

Windows (Acorn NMR Inc. 2007). Spectra were referenced and integrated against the

internal standard (MDP) set as 6= 17.46 ppm after its comparison to an externally held

85% H3P04 set as 6 = 0 ppm. Integration over set spectral windows were chosen to

correspond with known phosphorus bonding classes (Turner et al. 2003d). With spectral

deconvolution applied to the region 8 to 3 ppm to separate orthophosphate from

phosphomonoesters, and -3 to -5 to separate pyrophosphate and polyphosphosphate

end groups.


223









Hydrolytic Enzyme Assay

The activities of two hydrolytic enzymes critical to P cycling

(phosphomonoesterase and phosphodiesterase) were determined using fluorogenic

substrates based on a standard microplate assay (Marx et al. 2001). For each sample,

soil suspensions were prepared in a 1:100 soil/water ratio (containing 1 mmol L-1 NaN3

to prevent microbial activity). Soil suspension (50 pL) was then pipetted into wells on a

micro-well plate (8 wells per substrate) containing 100 pL of 200 pmol L-1 substrate (4-

methylumbelliferyl phosphate and bis(4-methylumbelliferyl) phosphate respectively )

and 50 pL of 200 mmol L-1 sodium acetate-acetic acid buffer adjusted to pH 4.0. Plates

were incubated for 30 min at 26C to approximate the daytime soil temperature in the

Changinola peat deposit. The reaction was terminated by adding 50 pL of 0.5 mol L-1

NaOH (final solution pH > 11) and the plates were read immediately on a FLUOstar

Optima multi-detection plate reader (BMG Labtech, Offenburg, Germany), with

excitation at 360 nm and emission at 460 nm. Control plates containing substrate,

buffer, and 1 mmol L-1 NaN3 (no soil suspension) were prepared and analyzed

immediately prior to and after the analysis of soil samples to account for initial

fluorescence as well as pH induced instability of substrates. Each soil had

corresponding blanks, methylumbelliferone (MU) standards and soil specific quench

standards. All enzyme activities are expressed here as pmol MU h-1 g-1 total C or when

comparing with soil total P as pmol MU h-1 g-l soil.

Data Analysis

All statistical tests were performed in SPSS for windows version 17.0.0 statistical

software (SPSS Inc. 2008). Data was checked for normality by application of a Shapiro-

Wilk test and visual inspection. If the assumption of normality was improved, natural


224









logs were used for statistical analysis. Differences between basic biogeochemical

characteristic were explored via a simple ANOVA, with a univariate GLM and site

number as the fixed factor. Post hoc analysis (Tukey HSD) was used to identify

significant homogeneous subsets. After confirmation of the suspected nutrient gradient,

the relationship between phosphatase enzyme activity per gram of total C and total soil

P was explored by use of the SPSS curve estimation of an inverse function.

Comparison of extraction efficiencies between methods used a simple paired t-test as

appropriate. Investigation of patterns seen within functional groups as determined by
31P NMR analysis used average parameters of combined samples (n=3). Given a lack

of sample material 31P NMR spectra for site 9 were determined on additional samples

(collected Nov 07). Sample total P was shown to be not significantly different (t-test: p >

0.05) but given AEM extraction was carried out at a different time, site nine was

excluded from later correlation analysis of total inorganic polyphosphates and

fumigation-released microbial P

Results

Soil Biogeochemical Properties

Surface peats from across the wetland transect were highly acidic (pH 3.7 0.4)

and of low bulk density, ranging from 0.03 g cm-3 within the central bog plain to 0.08 g

cm-3 within forested portions of the transect (Table 8-1). Surface samples showed a

range of total C concentrations from 41- 54% with a significant difference between sites

(ANOVA: p < 0.001). Given external mineral inputs to this ombrotrophic system are

minimal (Phillips and Bustin 1996) such differences reflect differences in biosilica

deposition from opal phytoliths and various diatoms (Kokfelt et al. 2009; Lopez-Buendia

et al. 2007; Street-Perrott and Barker 2008) or variance in C decomposition and peat


225









accretion rates across the peat dome (Craft and Richardson 1993). Total Ca and K

showed no significant difference between sites (ANOVA: p > 0.05) whilst total Mg

showed a significant difference (ANOVA: p < 0.001), although there was no clear trend

across the transect. Total N and P demonstrated expected trends between sites

(ANOVA: p < 0.001) from a relatively nutrient enriched site 1 (total N= 29 mg g-1, total P

= 1.0 mg g-) to oligotrophic site 9 (total N =22 mg g-1, total P = 0.4 mg g-1). In both

cases post hoc analysis (Tukey HSD) showed 4 homogeneous subsets across the

gradient with proximate sites showing no significant difference to each other. When

nutrient concentrations were expressed on a volumetric basis the very low bulk

densities of the surface peat within the central regions exacerbates the biologically

relevant gradient in nutrient content of the peat across the transect (Figure 8-2 A, B).

With a highly significant (ANOVA: p < 0.001) fivefold increase in total P ( 14.6 to 70.9

pg. cm3 ) and almost three fold increase in total N (0.734 to 2.0 mg cm-3). Molar ratios

of total soil N to P also suggest a significant increase in the degree of P deficiency

towards the central portions of the peat dome. Yet the relatively high total elemental

molar ratios, N:P (ranging from 28 to 50) and C:P (ranging from 485 to 1291) at all

sampling sites (Figure 8-2 C, D) is suggestive of P limitation even at the peripheral

sites (Cleveland and Liptzin 2007).

Phosphorus Biogeochemistry

Levels of 'bioavailable' P recovered by AEM strips were below 3 pg g-1 for sites 5

through 9 with other sites showing an increasing recovery to a maximum of 29 pg g-

(2.8 % of total P) at site 1 (Table 8-2). Phosphorus released by hexanol fumigation

showed significant differences among sites (ANOVA: p < 0.05) and represented

between a remarkable 18 and 38% of soil total P. Although fine root biomass may have


226









been included in this fumigation released pool, it is clear that a significant proportion of

total P is present within viable cells.

Both hydrolytic enzyme activities assayed showed significantly differences

(ANOVA: p <0.001) between study sites, with central sites showing rates up to seven

times that of the peripheral site 1 (Figure 8-3). Given the known influence of nutrient

status on hydrolytic enzymes the relationship between total P and enzyme activity was

explored via line fitting of an inverse relationship. This showed highly significant (p <

0.001) inverse relationships between total soil total P and phsophomonoesterase (R2 =

0.545) and phosphodiesterase (R2 = 0.740) activity per g of soil.

Phosphorus recovery in NaOH-EDTA

Alkaline extraction of air dried soils or in conjunction with AEM of fresh soils

recovered up to an average of 67% of soil total P (Figure 8-4). A simple paired t-test

showed significant difference between extraction efficiency of non fumigated and

fumigated fresh soil (paired t-test: p < 0.001), with the majority (95% of the difference)

being attributable to P recovered by the AEM during the fumigation step. Although

significantly different (paired t-test: p = 0.016), it is interesting to note that the extraction

of fresh soils after hexanol fumigation recovered similar levels to that after air drying

suggesting straight alkaline extraction of air dried soils recovers P pools associated with

microbial biomass (Turner et al. 2003c). There was a significant increase in P recovery

between extraction of air dried soils for 4 h and 16 h (paired t-test: p < 0.001) but with

an average relative difference of only 6.6% (4.11% of total P). Given known hydrolysis

of various compounds in alkaline solution (Doolette et al. 2009; Turner et al. 2003d) the

decision was made to use the standard 4 h extraction for detailed 31P NMR analysis.

Repeat extractions using the standard 4 h provided a consistent recovery ( 10% of


227









total P) based upon molybdate colorimetry of digested alkaline solutions (Data not

shown).

Solution 31P NMR spectroscopy

Spectra acquired demonstrated a diverse range of P forms to be present in the

surface soils (Figure 8-5, Table 8-3) with significant levels of organic P (phosphonates,

phosphomonoesters and phosphodiesters) as well as inorganic poly-phosphoric

compounds. The concentration of all components, other than the inorganic poly-

phosphates, identified by 31P NMR analysis showed significant (p < 0.001) positive

linear correlation with total soil P. The concentration of residual P, not extracted or

identified by the NMR analysis showed significant (p < 0.05) positive correlation with

total P, but as a proportion of total P showed a highly significant (p < 0.005) negative

correlation, ranging from 29% of total P at the relatively enriched site 1 to 55% at site 9.

Detected phosphonates were found to represent up to 31 pg g-1 and ranged from

1.7 to 3.3 % of total P within sites 1-7. Their lack of detection at the low P sites (site 8

and 9) is probably attributable to low signal to noise ratio and an inability to resolve their

presence as opposed to a true absence. Phosphodiesters other than DNA were found

at similar low levels across all sites (0.6 to 3 % of total P) but it should be noted that,

due to alkaline hydrolysis, this may represent a significant underestimation of important

phosphodiesters such as phosphatidyl choline (McDowell and Stewart 2005b; Turner et

al. 2003d). Concentrations of DNA ranged from 105 to 47 pg P g-1 and represented

from between 8.7 and 13.3 % of total P. While phosphomonoesters were found to be a

significant component of the organic P pool ranging from 174 to 45 pg P g1, 11.5 and

17.1 % of total P, yet did not contain peaks characteristic of commonly found isomers of

inositol hexakisphosphate (Figure 8-6). There was a significant positive correlation


228









(Pearsons rho = 0.790, p < 0.05) between soil total P and the ratio of P identified as

phosphomonoesters and total phosphodiesters, ranging from 1.40 at site two to a low of

0.88 at site eight.

Inorganic polyphosphates were found to constitute a major fraction, up to 24% of

total soil P, of all soil samples tested (Table 8-3). The presence of inorganic

polyphosphates in native soils was confirmed, using a limited number spectra from fresh

soil extracts (Figure 8-7) which when coupled to the rapid air drying used (< 10 days)

suggests their presence was not an artifact due to fungal growth (Koukol et al. 2008) but

the fact that polyphosphates represent an important insitu pool. Analysis of fresh

samples after application of AEM strips also provided insight into the location of this

significant P pool. The AEM strips used in this study have been shown to recover

significant levels of polyphosphate from solution (Cheesman et al. 2010b), when

extracts were analyzed after application of AEM but without the use of hexanol as a

biocide, polyphosphates were recovered (NF samples). When coupled to sample

fumigation polyphosphates were no longer detected or were found only in trace

concentrations (Figure 8-7). This would suggest the presence of polyphosphates within

cellular structures (Kornberg et al. 1999) or complexes disrupted by the action of

hexanol (Myers et al. 1999). In addition, total inorganic polyphosphates

(pyrophosphates and long chain polyphosphates) determined by solution 31P NMR

spectroscopy of dried soil extracts showed significant positive correlation (Pearsons rho

= 0.804, p < 0.05) with fumigation released P as determined by AEM on fresh soils.

Discussion

Basic characterization of the sampling transect confirmed previous studies in

identifying the presence of a distinct and remarkable nutrient gradient across distinct


229









vegetation communities within the Changuinola peat dome (Sjogersten et al. 2010;

Troxler 2007). Total P was shown to be at the high end of the range observed in other

tropical ombrotrophic systems, with sites in Kalimantan ranging from 272 to 373 pg g-1

(Page et al. 1999), and the Peruvian lowland Amazonia ranging from 130 to 590 pg g

(Lahteenoja et al. 2009a). This may reflect either the relatively young age and shallow

depth of peripheral sites as compared to other more established peat profiles (Phillips et

al. 1997), or proximity to the coast given the potential for both direct and occult

deposition of oceanic P. Bioavailable, estimated by AEM extractable, and total P

demonstrated a significant increase towards peripheral Raphia taedigera sites, yet the

high molar ratios of total C:P, and total N:P, in addition to significant levels of hydrolytic

phosphatase enzyme activity, suggest the potential for P limitation within the accreting

organic matter of all sites (Cleveland and Liptzin 2007). Previous evidence based upon

foliar nutrient ratios and 5 15N values (Troxler 2007) has been used to highlight a

potential shift from N to P limitation of trees across the peat catena, further work is

therefore required to explore the potential for differential nutrient limitations between

above and below ground biomass within and across this wetland system

(Sundareshwar et al. 2003).

Phosphomonoesterase activity at peripheral sites (Sitel-3) was found to be similar

to the only study I am aware of, that has reported enzyme activity under similar

environmental conditions. Jackson et al (2009) found potential activities of ~2.75 pmol

MU h-1 gOM-1 in samples from a Malaysian forested peat swamp, suggesting the

potential for high organic P turnover in tropical peatdomes (Quiquampoix and Mousain

2005). Comparison of potential enzyme activities reported between studies using non


230









standard assays (Marx et al. 2001; Wallenstein and Weintraub 2008) and in soils of

different physiochemical characteristics (Drouillon and Merckx 2005) may be erroneous,

yet it is clear that within our study there is a clear increase in the enzyme potential

towards the more P deficient central sites suggesting an increase in potential organic P

turnover.

Across all sites hexanol fumigation and AEM extraction showed that a substantial

proportion of total P was held within live biomass. Although it is likely that fine roots and

matrix bound P liberated by the action of hexanol contributed to this pool, fumigation

methods are known to underestimate microbial biomass (Brookes et al. 1982; Myers et

al. 1999), it is therefore likely that a substantial proportion of P identified in the NMR

spectra represents a high contribution from live microbial biomass. A similar conclusion

is drawn when comparing the extraction efficiency of the coupled AEM and alkaline

extraction of fumigated fresh soil to that of air dried soil (Figure 8-4). The large and

significant increase in extraction efficiency between AEM and alkaline extraction of non-

fumigated soil versus air dried soil is contrary to the known influence of pretreatment in

other peat based soils. In Turner et al (2007b) direct alkaline extraction of fresh and air

dried soils for 16 h tended to result in a similar recovery of P. I interpret this as evidence

that in the highly fibrous peats of this study system hexanol fumigation, or drying and

grinding, disrupts physical structures or lyses live biomass otherwise not extracted

during the short 4 h extraction period used.

The nature of phosphorus forms identified by solution 31P NMR spectroscopy was

remarkably similar between all sites and phasic vegetation zones, yet their proportions

appeared to show progressive alteration in relation to basic biogeochemical


231









characteristics. This suggest that microbial processing (in response to physiochemical

conditions) dictates the standing pools of P found within organic soils rather than the

nature of the above ground biomass. All sites showed similarities to previously studied

acidic peatlands, as well as characteristics seen in other high organic wetland systems.

Two peaks (20.64 and 19.14 ppm) assigned to phosphonates, while not present in the

highly studied calcareous peatlands of south Florida (Turner and Newman 2005; Turner

et al. 2006a) have been found in acidic northern hemisphere blanket bogs (Bedrock et

al. 1994; Turner et al. 2003b). Suggesting that phosphonates are either more prevalent

within biomass present (Ternan et al. 1998) or that they experience greater extracellular

stability under acidic conditions. This study also showed that, similar to other organic

wetlands, a significant proportions of organic P was found as phosphodiesters or as

potential phosphodiester hydrolysis products (Turner and Newman 2005). This is in

contrast to terrestrial soils where up to 90% of organic P may be identified as

phosphomonoesters (Condron et al. 2005), with isomers of inositol hexakisphosphate

forming a substantial proportion of total organic P(Turner et al. 2003f). This distinction

has been attributed to differential stabilization in the organic matter and redox conditions

prevalent to wetlands (Celi and Barberis 2005a; Turner et al. 2006a), though the recent

detection of the phosphomonoester myo-IP6 within anaerobic sediments (McDowell

2009; Turner and Weckstrom 2009) would suggest a complex interplay between

substrates and physiochemical conditions. The distinction between terrestrial soils and

wetlands may represent merely site difference in the proportion of P found within viable

microbial biomass as compared to the extracellular environment (Oberson and Joner

2005).


232









Given evidence for increased P deficiency across the study transect an

unexpectedly large proportion of total P was found as inorganic polyphosphates from all

sites. Polyphosphates are known to play an integral role in archeal, prokaroyotic and

eukaryotic cells (Kornberg et al. 1999) and are often associated with luxury microbial P

uptake (Hupfer et al. 2007; Khoshmanesh et al. 2002) and activated sludge processing

(Reichert and Wehrli 2007), yet polyphosphates have been detected in a range of

natural wetland/aquatic systems, including oligotrophic lake sediments (Ahlgren et al.

2006a; Hupfer et al. 2004), Carolina bays (Sundareshwar et al. 2009), and the humic

acid fraction of re-seeded peatlands (Bedrock et al. 1994). The large concentrations

detected within this study may represent intracellular stores associated with P

homeostasis or microbe sporulation (Brown and Kornberg 2004) under conditions of

fluctuating redox (Davelaar 1993), or may represent a generalized metabolic response

to environmental stress or nutrient deficiency within an ombrotrophic environment

(Seufferheld et al. 2008). Further work would be needed to elucidate the source and

dynamic nature of polyphosphates under field conditions. It is interesting to note that

Malaysian peat domes have a highly active surface layers dominated by Acidobacteria

and Crenarchaeota (Jackson et al. 2009), if typical of tropical peatdome systems such

unusual assemblages may explain the elevated role of long chain polyphosphates as

contrasted with other oligotrophic freshwater peatlands (Turner and Newman 2005). It

should also be noted that polyphosphates, although stable under the alkaline extraction

conditions, are catalytically degraded by the presence of divalent cations (Harold

1966). The addition of EDTA used here should of improved recovery of polyphosphate

(Cade-Menun and Preston 1996) yet some researchers interested in their role in


233









lacustrine and marine sediments have adopted a pre-extraction steps with EDTA

(Hupfer and Gachter 1995). Therefore it must be considered that the detection of

significant levels within this ombrotrophic, and presumably low iron, study site may be a

result of their reduced degradation during extraction and NMR spectroscopy as

compared to other wetland sites.

Interpretation of basic biogeochemical parameters suggests a reduction in P

availability and an increase in potential organic P cycling towards the central portion of

the study wetland. This corresponds with a pronounced increase in the proportion of

alkaline stable (residual) P as well as a decrease in the phosphomonoester to

phosphodiester ratio. Microbial modification of P forms within the detritus material of

organic based systems has been shown to be dependent upon nutrient availability

(CHAPTER 7). I contend that there are two potential mechanisms that could account for

the observed patterns, yet given a recognized limitation of 31P NMR as applied to bulk

soil is its inability to differentiate between viable biomass P and that stabilized in the

extracellular environment, I am currently unable to differentiate between them. On the

one hand observed patterns could reflect the increased biological demand and turnover

of organic P within central portions of the wetland resulting in the relative loss of

phosphomonoesters which would be otherwise abiotically stabilized in the extracellular

environment (Celi and Barberis 2005a) and the accumulation of highly humified alkaline

stable organic forms of P. Secondly, that changes in composition of bulk soil may reflect

changes in constitutive components of microbial biomass itself a major fraction of total

soil P (Makarov et al. 2005).


234









1 Table 8-1. Soil biogeochemical characteristics from nine sampling stations across an ombrotrophic peat dome. Values
2 are averages standard deviations of three replicate samples at each site
Total Elements
Site Vegt p Bulk density C N P Ca K Mg
(g cm3) ---- mg g-1 ---g-----
------------------------ Pg g-_ ------------------------

1 1 3.6 .12 0.069 .003 498 5.7 29 0.3 1028 25 145 16 539 41 308 6
2 3.8 .07 0.064 .005 508 4.2 29 0.6 1014 51 138 14 484 11 432 27
3 3.9 .03 0.060 .005 515 3.9 28 0.4 956 69 134 63 546 25 414 89
4 2 3.7 .02 0.078 .007 535 5.6 26 0.8 655 81 85 54 409 72 334 98
5 3.9 .05 0.064 .004 506 3.6 28 0.1 710 21 132 6 476 15 491 25
6 3 3.6 .17 0.056 .003 507 2.6 25 0.3 659 31 176 17 400 25 761 40
7 4 3.7 .11 0.040 .003 458 13.1 23 1.1 672 47 169 28 561 102 787 158
8 3.8 .19 0.050 .002 500 1.8 19 1.1 388 13 248 80 452 63 1150 111
9 5 3.6 .16 0.033 .003 417 3.0 22 1.0 442 33 68 3 479 70 324 18
3 t = Permanent vegetation sampling plot established by Sj6gersten et al. (2010)

4
5 Table 8-2. Phosphorus forms identified by anion exchange membrane technique applied to fresh soil samples. Values
6 are averages standard deviations of three replicate samples at each site.
AEM extractable Fumigation released
Site
S eg9 g9pg g-1 % total P
1 29.5 9.6 184 40.5 18 4.6
2 17.6 4.2 232 20.1 23 0.4
3 27.6 7.9 193 19.4 20 2.1
4 5.4 8.1 138 52.5 21 6.0
5 0.1 0.0 267 24.2 38 3.9
6 0.9 0.7 190 25.3 29 2.7
7 2.9 2.1 148 12.4 22 3.6
8 0.2 0.0 110 8.5 28 2.5
9 0.3 0.2 151 6.0 35 4.3
7
8
9


235











Table 8-3. Phosphorus forms identified by solution 31P NMR spectroscopy. Values equal concentrations as determined by
proportion of spectra area applied to total P determined by digest of alkaline extracts.


Phosphonates Orthophosphate Phosphomonoesters


Other
DNA Phosphodiesters Pyrophosphate
Phosphodiesters


Polyphosphatet


pg P g1 soil (% of total P)


27 (3)
31 (3)
18 (2)


258 (25)
233 (23)
170 (18)


167 (16)
174 (17)
143 (15)


105 (10)
93 (9)
91 (10)


15 (2) 129 (20) 79 (12) 57 (9)
16 (2) 82 (12) 86 (12) 72 (10)
16 (2) 83 (13) 75 (11) 61 (9)
11 (2) 76 (11) 89 (13) 90 (13)
nd 29 (7) 45 (11) 47 (12)
nd 54 (9) 68 (12) 57 (10)
n.d = not detected, trace.
t = Polyphosphate EG= End residue. MG = Mid chain residue
: = Spectra acquired on additional sample


29 (3)
30 (3)
26 (3)
10 (1)
14 (2)
9 (1)
10 (2)
3 (1)
3 (1)


32 (3)
20 (2)
11 (1)
26 (4)
4 (1)
3 (<0.5)
3 (<0.5)
2 (1)
7 (1)


34 (3)
14 (1)
25 (3)
19 (3)
18 (3)
9 (1)
9 (1)
4 (1)
n.d


76 (7) 301 (29)
107 (11) 310 (31)
138 (14) 333 (35)
61 (9) 260 (40)
147 (21) 271 (38)
135 (20) 268 (41)
124 (19) 260 (39)
65 (17) 193 (50)
72 (12) 317 (55)


236


Residual


-------------------------------------------------------------------


-------------------------------------------------------------------






































- Canal
Approximate vegetation zones
I = Raphia taedigera palm swamp
II = Mixed forest swamp
III = Compnosperma panamensis forest swamp
IV = Myrica-Cyrrilla bog plain


Figure 8-1. Overview of study transect and sampling sites with the Changuinola peat
deposit, San San Pond Sak N.W. Panama. Access route originates at a
Canal cut in 1908 and approximates previous leveling transects of the site
(Phillips and Bustin, 1996).


237













Raptia
A taem~e
palm swamp


Med forest Campnosperma
swamp i panamenss
forest swamp


I tyrica-
SCynba
bog plain


Rpahia
C awadip ra
palm swamp


MKed forest Capospmema
swamp peanamasis
forest swamp i


a I
-I-
r ab abc



ed
I HJ de

0 1 ef











5-
ab a


IaI
I I
I I r









ab ab



bc'
5 e





cd d
SI




,I I


u3] U,


1500-


*U* '0 0
co&3&3il


r

U1


I I


ed d




ab bcd abc











a a 1







ab ab
I a I











ab
a'aT
II
I I
I I
I I
I I
I I
I I
I -I -


UI I
- ,,.
U 02


e
U)
io


Figure 8-2. Nutrient gradient; A) mass of total P, B) mass of total N, C) Molar ratio C:P, and D) N:P from nine study sites
within the Changuinola peat deposit. All attributes show significant overall differences between sites (ANOVA: p
< 0.001) with superscript indicating homogenous subsets as derived from post hoc (Tukey HSD: p < 0.05).
General vegetation groupings based upon phasicc communities' identified by Philips et al (1998).



238


Myrica-
CyrfS
bog plain










A
>- c









..a ab
_b- -
<0 60-




43_





0

15-
C, Db
O 0






o b
0 ab 10-E
0 E ab










o a a a -
2 0 0









0 i b i i












significant difference between sites (ANOVA: p < 0.001) with superscript
letters indicating homogeneous subsets as derived from (Tukey HSD: p <
0.05)
ro 0



0 a
0 5-












letters indicating homogeneous subsets as derived from jukey HSD- p <


239
































Treatment:
Fumigation:
Extraction time:


I I- ___________


Fresh
No
4h


Fresh
Yes
4h


Air Dried

4h


Air Dried

16 h


Figure 8-4. Comparison of P recovered by alkaline extraction (0.25 mol L-1 NaOH and
0.05 mol L-1 EDTA) of air died soils (4 h and 16 h) or in addition to AEM
extraction of non fumigated and fumigated fresh soil. Bars represent averages
(n =27), error bars = 1 standard deviation


240














^J\JJ^ _L, lSite 9





Site 7





Site 4





Site 1

20 10 0 -10 -20
Chemical shift (ppm)

Figure 8-5. Solution 31P NMR spectra showing range of P forms present in surface soils
from select sites across the study transect. Spectra plotted using 15 Hz line
broadening referenced and scaled using internal standard MDP (5 = 17.46
ppm). Due to lack of sample material, spectra ($) acquired on additional
samples collected in Nov 07. Additional samples total P shown to be not
significantly different to original samples (t-test: p > 0.05).


241









6.21


5.10
4.92
4.84
4.63
4.46






6 5 4
Chemical shift (ppm)

Figure 8-6. Detail of solution 31P NMR spectra from site seven soils. Spectra plotted
using 2 Hz line broadening and referenced using internal standard MDP (5 =
17.46)


242
















Site 9 F


Site 1 NF


yrt / 1, "-"Site 1 F
20 10 0 -10 -20
Chemical shift (ppm)

Figure 8-7. Solution 31P NMR spectra of site one and nine soils after application of
anion exchange membranes with (F) and without (NF) fumigation step using
hexanol. Spectra plotted using 15 Hz line broadening referenced and scaled
using internal standard MDP (5 = 17.46 ppm).


243


kwrWf- *^ Site 9 NF









CHAPTER 9
SUMMARY AND CONCLUSIONS

Anthropogenic alteration of global P cycling has been profound, and given

expected global population and consumption growth we are likely to see the continued

and increased impacts of a disrupted P cycling on natural ecosystems. Due to their

position in the landscape wetlands are often a focus of this disruption, with many

wetlands showing a degradation or shift in ecosystem function due to P loading. Yet,

wetlands also act to sequester P in the landscape, providing a mechanism by which

downstream systems may be protected from adverse impacts. The dynamic interaction

of biological communities and P within wetlands has been the focus of much study, and

although known to constitute a sizable proportion of total P (CHAPTER 2), little

information exists as to the functional forms biogenic P may represent in wetland soils.

This dissertation has employed solution 31P NMR spectroscopy to investigate both the

forms of P found within wetlands and the potential mechanistic drivers that determine

biogenic P composition in soils.

Given a lack of existing information on biogenic P composition in wetlands, initial

studies focused on surveying composition within a diverse range of wetland systems.

The observation of potential mechanistic groupings, in conjunction with recent literature

sources, resulted in working hypotheses, tested in subsequent chapters. The complete

dissertation aimed to elucidate the dynamic nature of biogenic P in wetland soils.

Specific experimental objectives were:

Determine the influence of wetland characteristics and soil physicochemical
properties on forms of biogenic P in wetland soils.

Hypothesis: The composition of biogenic P in wetland soils varies systematically
with respect to wetland characteristics, landscape position, and/or soil
biogeochemical properties.


244









Determine how position in the landscape, i.e. from terrestrial to wetland
environments, influences biogenic P composition of soils.

Hypothesis: Landscape position impacts soil properties, which in turn influence
biogenic P composition. Specifically, higher productivity, a receiving position in
the landscape and reduced decomposition leads to increased organic matter
content within wetlands. This leads to differences in mechanisms of abiotic
stabilization leading to hydroperiod correlating with a decrease in the ratio of
phosphomonoesters to phosphodiesters (predominantly DNA).

Determine how anaerobic conditions impact biogenic P composition.

Hypothesis: Anaerobic conditions destabilize the phosphomonoester myo-IP6
and polyphosphates, and lead to reduced decomposition of phosphodiester DNA.

Assess the role of nutrient availability in determining biogenic P composition
within wetland detritus and soils.

Hypothesis: Increased P availability, due to elevated ambient conditions, will
reduce turnover of biogenic P by microbes, thereby altering P composition within
wetland materials.

Biogenic Phosphorus Composition in Wetlands (Experimental Objective 1)

Initial studies evaluated biogenic P composition in the surface soils (0-10 cm) of

28 freshwater wetlands, representing a diverse range of climatic conditions and

hydrogeomorpic types. Characterization and biogeochemical analysis alongside

determination of biogenic P forms using solution 31P NMR spectroscopy, allowed for

simple ordination and the identification of emergent patterns with respect to the biogenic

P found. Although P composition was independent of direct influence from wetland

vegetation or climatic setting, biogeochemical characteristics, themselves a product of

wetland setting, could be used to group sites in term of the biogenic P forms found. The

simple delineation of wetland sites with respect to pH and organic matter content

showed clear distinctions between the nature and proportion of P forms found in

wetland soils. Distinctions included the presence of phosphonates and isomers of the

phosphomonoester IP6 within acidic mineral-rich wetlands, and the presence of


245









substantial concentrations (up to 17% total P) of polyphosphate within high organic

systems. The ratio of phosphomonoesters to phosphodiesters, often used as a proxy

variable when demonstrating changes in P composition between systems (McDowell

and Stewart 2006; Turner and Newman 2005), showed a strong negative and significant

relationship with both organic matter content and C:P ratio, suggesting greater

prevalence of phosphodiesters in organic-rich systems and those in which P is limiting.

Influence of Landscape Position (Experimental Objective 2)

The influence of increased organic matter upon biogenic P composition was

further evaluated in the detailed study of biogenic P across a distinct upland wetland

landscape transition. Surface soils (0-10 cm) from four wetlands within cow-calf

pastures north of Lake Okeechobee, FL were studied, and found to contain significantly

(p < 0.05) greater concentrations of organic matter (219 g C kg-1), total P (371 mg P kg-

1) and metals (Al, Fe) relative to surrounding pasture. The concentration of P forms,

determined by extraction with AEM strips, 1 mol L-1 HCI, and an alkaline extract (0.25

mol L-1 NaOH 50 mmol L-1 EDTA) showed significant differences between uplands

and wetlands, but did not alter as a proportion of total P. Speciation of NaOH-EDTA

extracts, by solution 31P NMR spectroscopy, revealed that organic P was dominated by

phosphomonoesters in both wetland and pasture soils, but that myo-IP6 was not

detected in any sample. The tight coupling of total C and P in the sandy soils of the

region, as well as the typically low binding capacity of siliceous parent material suggests

that P dynamics across this agricultural landscape is driven by association with organic

matter. This led to me not finding the expected shift in composition, with altered organic

matter content, since there was no transition from systems dominated by interactions

with the mineral phase to those dominated by organic matter.

246









Influence of Redox Conditions (Experimental Objective 3)

Soil characteristics typical of wetlands, specifically high organic matter and low

redox conditions, are the result of water-induced anaerobosis. The direct implication of

anaerobic soils upon the stability of biogenic P, suggested as a mechanism driving

observed patterns in wetlands and terrestrial systems, was tested by tracking the fate of

standard compounds under aerobic and anaerobic conditions within a mesocosm

experiment (CHAPTER 6). The study showed that over the course of 48 days, both

DNA and the phosphomonoester myo-IP6 were degraded in wetland soils. Although

DNA appeared to conform to the working hypothesis, showing an increased

decomposition under aerobic conditions, myo-IP6 showed no distinction in turnover rate

between aerobic/anaerobic conditions. This appears to contrary to the much cited work

of Suzumura and Kamatani (1995a; 1995b) and the observation that anaerobosis

increases the destabilization of terrigenous IP6 in marine sediments, yet fits the recent

observation of isomers of IP6 in various anaerobic wetland soils (CHAPTER 4).

Influence of Phosphorus Availability (Experimental Objective 4)

The influence of P availability on the transformations of organic matter entering

wetlands and its influence upon forms found within wetland soils were investigated by

the study of biogenic P across 'natural' nutrient gradients within two distinct wetland

systems. Specifically, I studied the microbial-mediated transformations of detritus as it

enters WCA-2A, Florida and the completed a detailed analysis of biogenic P across a

nutrient and concomitant vegetation gradient within the Changuinola ombrotrophic peat

dome, Panama. In tracking changes in the P forms found in Cladiumjamaicense Crantz

and Typha domingensis Pers leaf litter during 15 months of decomposition at two sites

of markedly within a hard-water subtropical wetland (WCA 2A, Florida) I was able to

247









investigate the significant role of litter quality and P availability as it impacts microbial

decomposition and biogenic P composition. Macrophyte decomposition at the nutrient-

enriched site resulted in net sequestration of P from the environment in forms

characteristic of microbial cells (i.e. phosphodiesters and pyrophosphate). Low P

concentrations at the unenriched site resulted in little or no net sequestration of P, with

changes in P forms limited to the loss of compounds present in the initial leaf litter

material. I concluded that under nutrient-rich conditions, P sequestration and the entry

of biogenic P into the soils occurs through the accumulation of microbial-derived

compounds, in addition to the presumed, but uncorroborated, concentration of

endogenous macrophyte biogenic P. While in contrast, under nutrient-limited conditions

standing P pools within wetland soils appear to be independent of the heterotrophic

decomposition of macrophyte leaf litter. Suggesting an elevated role of a secondary

microbial loop, or periphyton, in the determining the composition of biogenic P found

within P limited peats of the northern Everglades.

The integrated influence of altered P availability on biogenic P in wetland soils was

investigated by study of an established gradient within an ombrotrophic tropical peat

dome, in Bocas del Toro province, Panama. Basic biogeochemical characteristics and

extracellular enzyme activities demonstrated a strong gradient in P availability and

potential organic P turnover, concomitant to substantial changes in phasic vegetation.

The types of biogenic P present, including surprisingly high concentrations of long-chain

polyphosphates, remained constant across a diverse range in vegetation types.

Although the proportions of P forms present appeared to show a dynamic response to

nutrient availability, including a decrease in the ratio of phosphomonoester to


248









phosphodiesters with reduced P availability. This would suggest that either altered P

turnover in response to nutrient requirements or the modification of standing microbial

biomass dictates the composition of biogenic P in the soil, with only a limited direct

influence from organic matter inputs.

Synthesis and Further Studies

Biogenic P in wetlands has been found to represent a diverse suite of functional

forms, the work here greatly expanding upon the range of wetland soils observed to

date. In doing so, it has become obvious that patterns exist, enabling broad

generalization to be made about the composition of biogenic P found in wetland soils.

Further work is needed to investigate patterns noted, yet not explored in this

dissertation, e.g. the absence of phosphonates in soils above a pH of 4.4 (CHAPTER

4). Studies have begun to explore the fundamental interactions between organic matter

content, anaerobic conditions, nutrient status, and vegetation types in determining

biogenic P composition in soils.

Organic Matter and Redox Conditions

Wetlands are typified by inundation or saturation for a period of time long enough

to result in soil and/or vegetation characteristics considered typical of a wetland.

Typically, the transition from uplands to wetlands, is to move to more anaerobic

conditions, resulting in the accumulation of detritus, and thereby conditions of higher

organic matter content. There is also variance between sites, dependent upon site

location (e.g. climate, vegetation) and hydrogeomorphic characteristics (e.g. landscape,

hydrologic). We can expand the conceptual model of dynamic biogenic P seen within

wetlands (Figure 1-2), to include two extremes upon the multidimensional continuum

that exists across all systems. The first occurs under conditions of low organic matter

249









(Figure 9-1 A), possibly due to an upland position or high external loading of mineral

material, in this case the composition of biogenic P is determined by its interaction with

the mineral phase. The second scenario occurs under conditions of high organic matter

(Figure 9-1 B), either through a sites receiving position in the landscape, high biomass

productivity, or water induced anaerobosis. Anaerobic conditions lead to the reduction

of redox sensitive components, directly impacting both mineral and organic matter

stabilization of extracellular biogenic P (Figure 9-2). Although a number of factors

confound this simple interaction model, including altered biogenic P inputs, and

intracellular biogenic P seen within wetlands (see below), it may well account for the

observed alteration in biogenic P composition between wetlands varying in organic

matter content (CHAPTER 4).

When I investigated this relationship, I had hypothesized that across a landscape

transition I would observe a shift in biogenic P composition due to a change from

mineral phase to organic matter stabilization of biogenic P. I did not find this shift but

this may be attributed to the characteristics of the sandy, quartz based parent material

in which the depressional wetlands were embedded (CHAPTER 5). Further, a major

component of the observed shift in biogenic P composition between systems of varying

organic matter content (CHAPTER 4) was the presence of various isomers of the

phosphomonoester IP6, not observed within the upland soils of the studied agricultural

landscape (CHAPTER 5). Evidence presented within this dissertation (CHAPTER 6)

and emerging from recent observation in anaerobic systems (McDowell 2009; Turner

and Weckstrom 2009) suggests a complex interaction between IP6 and the mineral


250









phase which may have as much to do with the binding and stabilization of the enzymes

associated with its degradation (Giaveno et al. 2010) as the phosphomonoester itself.

Nutrient Status

Wetland soil conditions, such as nutrient status, are known to have a direct

influence on microbial populations (Smith 2007) and their eco-physiological expression

(Corstanje et al. 2007). It is therefore likely that it will impact both intracellular

composition of microbes and the turnover of extracellular biogenic P. This dissertation

identified not only the direct accumulation of 'sequestered' biogenic P (CHAPTER 7),

but the resulting integration of adjusted microbial sequestration and biological turnover

across a wetland nutrient gradient (CHAPTER 8). Although studies within this

dissertation were unable to differentiate the role of altered intracellular composition (see

below), many studies noted the increased prevalence of phosphodiesters under nutrient

limited conditions. In conjunction with biogeochemical characterization of soils this has

been taken to be indicative of how under P limiting conditions a greater proportion of P

is found as microbial biomass and that there is a rapid cycling of biogenic P forms, such

as phosphomonoesters within the extracellular environment.

This dissertation has highlighted the role of microbial populations acting in

response to abiotic conditions to determine soil P composition. The composition of

biogenic P in soils appears governed by site conditions (i.e. nutrient availability, mineral

content, pH) and shows little direct coupling to the nature of detritus entering wetlands

(CHAPTER 4, 8). This has lead to the development of a combined concept, iterative

processing, which I believe explains how increasingly biogenic P in soils represents the

result of microbial cycling and a biogenic P composition independent of the nature of

detritus entering the wetland system.


251









Iterative Processing

The notion of 'dynamic biogenic P' introduced in CHAPTER1 (Figure 1-2) has

been explored throughout this dissertation. Yet, there is an implicit assumption, and

flaw, in such a model that echoes models by Wetzel (1999) and Reddy et al (2005). It

assumes that composition of biogenic P in soils is an absolute and does not take into

account the context in which the P form may be found. This model assumes P enters

the wetland in a particular functional form from biological sources which leads directly to

the biogenic P composition of the soil (Figure 9-3). By extension, this results in an

assumption that all P found within a certain functional form, for example DNA, acts in a

similar manner within wetland soils. Yet, as has been shown by studies of DNA

stabilization (Ogram et al. 1988) and microbial utilization (Leake and Miles 1996) this

assumption is over simplistic. The cycling of biogenic P in wetland soils occurs as a

result of iterative processing of P containing compounds by successive microbial

generations (Figure 9-4). At each stage, environmental conditions impact microbial

turnover of existing forms, as well as the synthesis of novel compounds. Therefore,

there is a need to go beyond this basic approach, to partition out biogenic forms that

represent active microbial biomass, from those that represent abiotically stabilized

forms. The development of an approach used in this dissertation (CHAPTER 7) may

offer a solution, by tracking observable changes within a known ecosystem component.

Yet it is likely the problem will require the application of additional techniques such as

the use of labeled biogenic P compounds to trace their fate and turnover in the soil.

Given the limited half-life of P isotopes (32P T1/2 =14.3 days, 33P T1/2=24.4 days) this

may include the stable isotope 180 already applied as a tracer for inorganic phosphate

(Young et al. 2009) and in the study of organic P turn over (Blake et al. 2005). As we

252









begin to be able to track the iterative processing of biogenic P within wetland soils we

will begin to fully realize the mechanisms that underlie observed patterns in the biogenic

P composition of wetlands.


253









Hydroperiod
Anaerobosis
I Productivity
Organic matter (%)
A B
\

Biomass Biomass



Organic 3 Organic
matter PC, matter PO
.... ............ .... .. ..... ....... .

Labile BiogenicP Labile Bogmnic P
Biogenic P M cromal f Biogenic P Microbial "
Stab lizedtablzed Biomass 0)
BiogenicP Biogenic P

-- -I --- ----
Mineral component Mineral component

Figure 9-1. Comparison of conceptual model of dynamic biogenic P cycling in soil (Figure 1-4) modified for A) systems
dominated by interactions with the mineral phase (e.g. uplands) and B) systems dominated by interactions with
organic matter (e.g. wetlands).


254











Redox sensitive ve
organometal
complexes

Organic matter



Anaerobic Accumulation of
Condition Organic matter


ve Mineral matter
Stabilization
Redox sensitive
mineral
components e


Figure 9-2. Influence of anaerobic conditions (typical of wetlands) upon both organic
matter and mineral phase stabilization of biogenic P in soils.


255































Figure 9-3. Simplified
wetland soil.


linear progression of development of biogenic P from inputs to


Organic Carbon

-- _--t -


SInteraction Environment
*4- P movement

Figure 9-4. Iterative processing of P within wetland soils. Mediated by microbial
processes, biogenic P undergoes interactions with both organic matter and
mineral phase. With each iteration showing stabilization, hydrolysis or turn
over dependant upon environmental conditions.


256









APPENDIX A
ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 1



Al METHODS FOR THE DETERMINATION OF TOTAL ORGANIC P IN
SOILS AND SEDIMENTS

A2 METHODS FOR SINGLE STEP INDIRECT DETERMINATION OF TOTAL
ORGANIC P

A3 ESTIMATES OF TOTAL ORGANIC PHOSPHORUS IN WETLAND
SUBSTRATES


257












APPENDIX Al. Cowardin classification system of wetlands and deepwater habitats


System
Marine


Subsystem
Subtidal


Intertidal




Estuarine Subtidal



Intertidal


Riverine Tidal





Lower
Perennial



Upper
Perennial


Intermittent


Lacustrine


Limnetic


Littoral


Palustrine


Class
Rock Bottom
Unconsolidated Bottom
Aquatic Bed
Reef

Aquatic Bed
Reef
Rocky Shore
Unconsolidated Shore

Rock Bottom
Unconsolidated Bottom
Aquatic Bed
Reef
Aquatic Bed
Reef
Streambed
Rocky Shore
Unconsolidated Shore
Emergent Wetland
Scrub-Shrub Wetland
Forested Wetland

Rock Bottom
Unconsolidated Bottom
Aquatic Bed
Streambed
Rocky Shore
Unconsolidated Shore
Emergent Wetland
Rock Bottom
Unconsolidated Bottom
Aquatic Bed
Rocky Shore
Unconsolidated Shore
Emergent Wetlands
Rock Bottom
Unconsolidated Bottom
Aquatic Bed
Rocky Shore
Unconsolidated Shore
Streambed

Rock Bottom
Unconsolidated Bottom
Aquatic Bed
Rock Bottom
Unconsolidated Bottom
Aquatic Bed
Rocky Shore
Unconsolidated Shore
Emergent Wetland

Rock Bottom
Unconsolidated Bottom
Aquatic Bed
Unconsolidated Shore
Moss Lichen Wetland
Emergent Wetland
Scrub-Shrub Wetland
Forested Wetland


258











APPENDIX A2. Ramsar Classification System for Wetland Type as approved by
Recommendation 4.7 and amended by Resolutions VI.5 and VII.11 of the Conference of
the Contracting Parties (Ramsar Convention Secretariat 1971).


Marine/Coastal Wetlands
Saline water


Saline or brackish water


Saline, brackish or freshwater
Freshwater


Inland Wetlands
Freshwater


Saline brackish or alkaline
water


Fresh, saline brackish or
alkaline water


Permanent


Shores

Intertidal


Lagoons
Estuarine waters
Subterranean
Lagoons


Flowing water


Lakes and pools



Marshes on
inorganic soils




Marshes on peat
soil

Marshes on
inorganic or peat
soils


lakes


Marshes and
pools

Geothermal

Subterranean


<6 m deep
Underwater vegetation
Coral reefs
Rocky
Sand, shingle or pebble
Flats (mud, sand or salt)
Marshes
Forested


Permanent


Seasonal/ intermittent
Permanent

Seasonal/intermittent

Permanent

Permanent/Seasonal/
intermittent

Seasonal/intermittent
Permanent


High altitude (alpine)


Tundra
Permanent

Seasonal/ intermittent
Permanent

Seasonal/ intermittent


A
B
C
D
E
G
H

J
F
Zk(a)
K


Rivers, Streams,
Creeks
Deltas
Springs, oases
Rivers streams creek
>8 ha
< 8 ha
>8 ha
< 8 ha
Herb dominated

Shrub dominated

Tree dominated
Herb dominated
Non forested

Forested


Zk(b)


259










APPENDIX A3. Hydrogeomorphic classification of inland wetlands, based on Semeniuk
and Semeniuk (1995; 1997).


Non- emergent wetland types


Landform


Hydrology
Permanently
inundated


Basins Channels


Lake


Seasonally inundated Sumpland


Intermittently
inundated
Seasonally
waterlogged


Playa


River

Creek

Wadi


Flats


Slopes Highlands or
Slopes hills


Floodplain

Barlkarra


Dampland Trough Palusplain Paluslope Palusmont


Self emergent wetland types


Ground water


Geothermal heating
No Geothermal
heating
Ombrotrophic


Sinter Mounds

Mound Springs


Raised bogs


- combinations are not possible as prevailing wetlands
Waterlogged= soils that are saturated with water, but where the water does not inundate the surface.
Inundated= soils that are covered with free standing water, the soil below the surface in these
situations is also saturated (waterlogged)


260


Mineral


Organic









APPENDIX B
ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 2



B1 METHODS FOR THE DETERMINATION OF TOTAL ORGANIC P IN
SOILS AND SEDIMENTS

B2 METHODS FOR SINGLE STEP INDIRECT DETERMINATION OF TOTAL
ORGANIC P

B3 ESTIMATES OF TOTAL ORGANIC PHOSPHORUS IN WETLAND
SUBSTRATES

B4 STABILITY OF STANDARD BIOGENIC P DURING ALKALINE
EXTRACTION AND LYPHOLIZATION


261












APPENDIX B1.
METHODS FOR THE DETERMINATION OF TOTAL ORGANIC P IN SOILS AND
SEDIMENTS


The table below lists procedures developed for the operational determination of
total organic P in soils and sediments. (Po = organic P, Pi = inorganic P, Ca~ = calcium
bound, Fe/AI~ = iron and aluminium bound, NTA= nitrolacetic acid, M= mol L-1)


Refernece


(Potter and Benton 1916;
Potter and Snyder 1916)
(Dean 1938)


Wrenshall and Dryer (1939)
In (Kaila and Virtanen
1955)
(Pearson 1940)

(Ghani 1942)



(Mehta et al. 1954)


(Saunders and Williams
1955)
(Legg and Black 1955)

(Chang and Jackson 1957)







(Walker and Adams 1958)

(Halstead et al. 1966)



Psenner (1984) In (Paludan
and Jensen 1995)


(Williams et al. 1971;
Williams et al. 1980)


(Hieltjes and Lijklema
1980)


Extraction/ Procedure


Dilute NH3OH


Na -Acetate pre treatment
0.25 M NaOH (90oC)

0.25 M H2SO4
Residual P
3 M HCI
4 M NH4OH


0.1 M HCI (or H2SO4)
0.5 M NH4OH (90oC, 2 h)
0.5 M Acetic acid
0.25 M NaOH (repeated)
1 M H2SO4

Conc. HCI (hot)
0.5 M NaOH (cold)
0.5 M NaOH (90 C)
0.1 M HCI (hot)
0.1 M NaOH (16 h)
Low temperature ignition (240oC)
Conc. HCI
(1M NH4CI) (30min)
0.5M NH4F (pH 8.2)
0.1M NaOH
0.5M HCI
0.3 M Citrate 0.15 M Dithonite(90 C)
(optional) 0.5N NH4F
(optional) 0.1M NaOH

Ignition + 0.1M NaOH

0.1 M HCI
0.2 M Acetylacetone (pH 8.0) +
Ultrasonic desperation
0.2 M Acetylacetone (repeated)
H20
0.11M NaHCO3, 0.11 M Na2S204
0.1 M NaOH
0.5 M HCI
0.22 M Citrate, 0.11 M NaHCO3, 0.11 M
Na2S204
1 M NaOH
0.5 M HCI
Residual P
1 M NH4CI
0.1 M NaOH
0.5 M HCI
Residual P


Rational or targeted phosphorus pool

Shown to remove ~100% of nucelic acid spike added
to soils
Remove exchangable Ca
Alkaline extractable P, and Po, failure of bromination to
oxidize NaOHTP suggested as evidence of IP6
Extraction of P,
Residual P, and Po
Believed to remove most of total Po
Extraction of residual Po

Pretreatment to remove bases
Single step extraction of P, and Po
Exchangeable bases and labile P,
Fe/AI ~P, and Po
Ca~P,

Ca~ P, and Po
Initial NaOH removes Powith minimal hydrolysis
Aggressive removal of residual Po
Removal of bases
Extraction of Po
Parallel extraction with and without oxidation of organic
matter.
To remove exchangable Ca
AI-P,
Fe~ P, and Po
Ca~P,
Reductant soluble (Fe occluded phosphate)
Al occluded phosphate
AI/Fe occluded phosphate

Parallel extraction of P before and after oxidation of
organic matter
High efficiency at recovering Po from a range of soil
types.


Readily labile P, and Po
Reducing dithonite (Fe~ P, / Po)
Po
Residual P,
Reductant soluble Fe~P

AI/Fe ~P
Ca~P F/AI/Ca-P
Residual AI/Fe-P, and Po
Labile P
Fe, Al P
Ca~P
Organic P


262












(Hedley et al. 1982)


(Van Eck 1982)


(De Groot and Golterman
1990)

(Psenner and Pucsko
1988)





(Ruttenberg 1992)


(De Groot and Golterman
1993)






(Bowman and Moir 1993)
(Ivanoff et al. 1998)


(Reddy et al. 1998)


Anion exchange resin
0.5 M NaHCO3
fumigation- 0.5M Na2HCO3
0.1 M NaOH
0.1 M NaOH +sonication
0.1 M HCL
Digestion H2SO4 and H202
1 M KCI
0.1 M NaOH
0.5 M HCI
Digest of residue
Ca-NTA/ Na2S204 (pH 8.0)
Na-EDTA (pH 8.0)
Residual P
H20
0.11 M NaHCO3 + 0.11 M Na2S204
0.1 M NaOH
0.1 M NaOH + 2 M H2SO4
0.5 M HCI
Residual digested
SEDEX Procedure
1 M MgCI2
CBD (0.3 M Na3-citrate, 1 M
NaHCO3, 0.144 M Na2S204)
1 M Na-acetate (pH 4.0)

1M HCI
Ignition 5500C +1M HCI

0.02 M Ca-NTA 0.045 M Na2S204
0.05M NaEDTA
0.25 M H2SO4
2.0 M NaOH (90oC)
Acidification
Phytase
Residual

0.25 M NaOH 50 mM EDTA
0.5 M NaHCO3
CHCI3 + 0.5 M NaHCO3
1 HCI
0.5 M NaOH
Residue
1 M KCI
0.1 M NaOH


0.5 M HCI
Residual


Labile P
Labile P
Microbial P
Fe/AI ~P
Interaggrgate
Ca ~P
Residual P
Exchangable P
Fe/AI-P + hydrolyzable Po
Ca CO3~P,
Po
Fe~P
Ca~P
Po
Pore water P,+ Po
Fe~P, + Po
AI-P,+ Po
Humic acid ~P
Ca~P,
Recalcitrant P

Loosely sorbed/exchangable P

Ferric-P
Authigenic carbonate,flourapatite,
biogenic apatite, Ca~P
Detrital P of igneous or metamorphis origin
P,

Fe(OOH)~P, amorphous Fe(OOH)
CaCO3~P,
Acid hydrolysable organic phosphorus
Organic phosphorus
Humic, Fulvic acid P
Phytate-P
Residual P

Single step alkaline extraction for P
Labile P, and Po
Microbial P
HCI extractable P, and Po
NaOH extractable P,, Humic/ Fulvic acid ~ P
Residual Po
Labile-P,
Fe/AI ~P,
NaOH extractable P,, Humic/ Fulvic acid ~ P
Ca/Mg ~P,
Residual Po and ocludded- P,


263









APPENDIX B2
METHODS FOR SINGLE STEP INDIRECT DETERMINATION OF TOTAL ORGANIC P


Method


Reference


Chemical Oxidation Peterson 1911 in (Pearson 1940)

(Dickman and De Turk 1938)

(Bray and Kurtz 1945)


Extraction
solution

0.2 M HNO3

0.1 M H2SO4

Pre-
extraction
with 0.5 M
NH4FI
0.1 M HCI


Low temperature
oxidation


Schmoeger (1897) in (Dickman and
De Turk 1938)


(Legg and Black 1955)


12% HCI


1 M HCI


140-160C
under
pressure

250C under
enriched 02


High temperature
oxidation


Madanov (1940) in (Kaila and
Virtanen 1955)

Odynsky (1936) in (Dickman and
De Turk 1938)

(Saunders and Williams 1955)


(Walker and Adams 1958)


0.3% citric
acid

2 M H2SO4


0.1 M H2SO4


0.5 M H2SO4


(Aspila et al. 1976)


1 M HCI


Ignition +
conc. H2SO4 +
K2S208


264


Oxidation
procedure


H202

H202

H202


300C


600C


550C


550C


M = mol L-'
Note: In all methods inorganic P is extracted in parallel samples pre and post oxidation
of organic matter the difference in being attributed to liberation of organic P.












APPENDIX B3
ESTIMATES OF TOTAL ORGANIC PHOSPHORUS


IN WETLAND SUBSTRATES


Reference


Location


Kaila (1955) In
Harrison 1983



Kaila and Missila
(1956) In Harrison
1983





McDonell and Walsch
(1956) In Harrison
1983
Saunders and Hinsch
(1968) In Harrison
1983
(Sommers et al. 1970;
Williams et al. 1970)







Agboola and Oko
(1976) In Harrison
1983
(Hosomi et al. 1982)

(Dean 1938)


(Reddy 1983)

(Johnston et al.
1984a; Johnston et al.
1984b)
(Richardson and
Marshall 1986)
Gore, AJP In


Finland





Finland






Ireland


New Zealand


WI,USA








Nigeria


Japan

Peterborough,
UK
Ely UK
FI, USA

WI, USA


MI, USA

England


Wetland name


Bog (pine)- Rame
Bog (treeless (Neva)
Bog Sphagnum
Spruce deciduous swamp (Karpi)
Fen Spagnum
Bog (pine)- Rame
Fen (treeless) -Letto
Bog (treeless)-Neva
Bog (treeless) (neva)
Bog (Rimpi)
Bog (Sphagnum)
Fen (birch)
Fen (sphagnum)
Bog (blanket)
Bog Raised acidic

Bog (Sphagnum)


Lake Wingra
Lake Monona
Lake Mendota
Devils Lake
Crystal Lake
Little John Lake
Trout Lake
Softwater lakes WI
Hardwater lakes, WI
Swamp


Lake Kasumigaura


Black fen


Heavy fen
Zellwood 4, Monteverde muck
Belle Glade, Pahokee muck
White clay lake, wetland
White Clay lake, alluvial
White clay lake background
Houghton Lake


Bog (blanket)


Vegetationt
PE-Forested
PE
PE-Moss
PE-Forested
PE-Moss
PE-Forested
PE
PE-moss
PE-moss
PE
PE-Moss
PE-Forested
PE- Moss
PE-Mosss
PE-Moss


Depth
(cm)
0-30
0-20
0-30
0-50
0-20
0-30
0-30
0-50
S
S
S
S
S
0-7.6
0-7.6


pH
4.2
3.6-5.5
3.7-5.1
4.6-4.9
4.4-5.8
4.0-5.1
3.9-8.0
3.6-5.5
4.4-4.8
4.6-5.2
4.2-5.3
4.9-5.3
4.8-5.5
4.5
5.6


Total P
(mg g-l)
379-772
286-1429
368-912
593-1244
1337-1482
186-1471
605-1244
357
937-1113
528-1389
373-680
1188-1493
107-1053
379-772
286-1429


PE- Moss 0-20.3 4.4


Lacustrine
Lacustrine
Lacustrine
Lacustrine
Lacustrine
Lacustrine
Lacustrine
Lacustrine
Lacustrine
PE


Lacustrine 0-10


PE-Agriculture

PE-Agriculture
PE
PE
PE
PE
PE
PE


PE-Moss


650
1260
1460
1630
2230
4120
6090
2250
950


134-385

1335


S 6.3


1025


7.5 1125
5.5 1450
5.7 1530
559t
746t
263t
5.5-6.1 700


Organic P
(%)


Method


Pearson
(1940), and
Ignition


Pearson
(1940), and
Ignition


Pearson
(1940)


Ignition I


Mehta (1954) SEPO








Ignition I


(Aspila et al.
1976)
Dean (1938)


Olsen and
Dean (1965)
Mehta (1954)


SEPO


SEPO


SEPO

Ashing-450 I


S 3-3.8 30-100 69


SEPO


265


Method
Grouping
I


SEPO












(Harrison 1987)

Table B3. Continued
Reference
(Cooke et al. 1990)


(De Groot and
Golterman 1990)



(Giblin et al. 1991)

(Koch and Reddy
1992)

(Cooke 1992; Cooke
et al. 1992)

(De Groot and
Golterman 1993)






(Gale et al. 1994)









(Barbanti et al. 1994)


(Quails and
Richardson 1995)

(Reddy et al. 1995)


Location
Nordland, New
Zealand


Relongue,
France
Camargue,
France
Alaska, Brooks
ridge
Everglades, FI,
USA

Nordland, New
Zealand


Wetland name
Waitangi forest wetland (ref)
Waitangi forest wetland
(impacted- Site 1)
Relongue I
Relongue II

Rice field I

Sagavanirktok river

WCA-2A (northern)
WCA-2A (central)
WCA-2A (southern)
Waitangi forest wetland
(impacted- flow path)


Camargue, Garcines Nord I
France Garcines Nord II
Garcines Sud
Relongues
Buisson Verte
Baisse Salee
Ditch ORF
Etang de Vaccares
Orlando, FL Constructed Waste water
USA treatment facility, Orlando FI,
Constructed WWT facility,
Orlando FI
Constructed WWT facility,
Orlando FI
Constructed WWT facility,
Orlando FI
Reference WWT facility, Orlando,
FI


Saca di Goro,
Italy

Everglades,
FI,USA

Everglades,FI,
USA


Po river delta lagoon G1
Po river delta lagoon G8
Po river delta lagoon G10
WCA-2A (D1,C1,D2,C2),
WCA-2A (D6,C6,A5,A6)


Vegetationt
PE
PE

PE
PE

PE-Agriculture

PE

PE
PE
PE
PE


PE
PE
PE
PE
PE
PE
Riverine
Lacustrine
PE

PE

PE

PE


Depth
(cm)
0-10
0-10

S
S

S

0-15

0-10
0-10
0-10
0-9


S
S
S
S
S
S
S
S
0-15

0-15

0-15

0-15

0-15


Lacustrine
Lacustrine
Lacustrine
PE
PE


pH Total P
H (mg g-1)
4.2 2742
4.7 8450

8.0 709
8.1 747


Organic P
(%)


6.2 700


1785
900
553
9299


660
632
488
986
419
689
3230
639
49


Method
(Olsen and
Sommers
1982)
(De Groot and
Golterman
1990)


81 (Williams et al.
1980)
54 (Hieltjes and
66 Lijklema 1980)
69
19 (Olsen and
Sommers
1982)
58 (De Groot and
33 Golterman
57 1993)


(Van Eck
1982)


6.2 536

5.1 670

4.6 837


867
1110
1254
124t
44t


WCA-1
WCA-2A
WCA-3
Holey Land Wildlife Management
Area


(De Groot and
Golterman
1990)

(Hedley et al.
1982)
(Reddy et al.
1995)


266


Method
Grouping
SEPO


SEPI




I

SEPI


SEPO


SEPI







SEPO


SEPI


SEPO


SEPO














Table B3. Continued
Reference
(Vaithiyanathan and
Richardson 1997)









(Mayer et al. 1999)


(Nguyen 2000)



(Qiu and McComb
2000)





(Schlichting et al.
2002)
(Craft and Chiang
2002)
(Bruland and
Richardson 2006)


Location
Everglades, FI,
USA







Everglades, FI,
USA
Canada



Nordland, New
Zealand

Swan Coastal
Plain, Western
Australia




Saxony-Anhalt
Germany
SW Georgia,
Florida
MN, USA


Wetland name
WCA-2A C1 = 1.4 km
WCA-2A C2= 3.5 km
WCA-2A C4= 6.9 km
WCA-2A C1 = 1.4 km
WCA-2A C2= 3.5 km
WCA-2A C3= 4.3 km
WCA-2A C4= 6.9 km
WCA-2A C5= 8.8 km
WCA-2A C6= 10.5 km
ENRP, flooded Histol


Lake Eire I
Lake Eire II


Vegetationt
PE
PE
PE
PE
PE
PE
PE
PE
PE
PE


Depth
(cm)
0-10
0-10
0-10
0-2.5
0-2.5
0-2.5
0-2.5
0-2.5
0-2.5
0-15


Lacustrine 0-10
Lacustrine 0-11


Waitangi forest wetland (pre
confluence)
Waitangi forest wetland (post
confluence)
Jandabup
Monger
Booragoon
Murdoch swamp
North
Banganup
Forrestdale
Droemling Nature park-AL

Joseph W. Jones Ecological
Reseach Center II
North Highway 12
Katrina Lake East
Katrina Lake West
Katrina Marsh North
Katrina Marsh South
Carlson South
Potato Farm North
North Highway 6
Painter Marsh North
Painter Marsh West
Painter Marsh Middle
Painter Marsh South
East Highway 110
South Highway 26
West Jennings


pH Total P
H (mg g-1)
nr 1664
nr 1044
nr 531
nr 1474
nr 1022
nr 529
nr 271
nr 278
nr 547
6.3 450

9.1 1186
9.2 1200

5.7 2647


PE

Lacustrine
Lacustrine
Lacustrine
Lacustrine
Lacustrine
Lacustrine
Lacustrine
PE-Forrested

PE-forested
PE-forested
PE
PE
PE
PE
PE
PE
PE
PE
PE
PE
PE
PE
PE
PE
PE


25261

20
1205
939
53
286
586
14
949

79
53
945
1428
1216
1054
1339
755
1544
1794
1197
1863
1481
1496
523
1544
1743


Organic P
(%)


Method
(Aspila et al.
1976)


54 (Ivanoff et al.
1998)
48
51 (Williams et al.
1971)


(Olsen and
Sommers
1982)
(Walker and
Adams 1958)





(Hedley and
Stewart 1982)
(Hedley et al.
1982)
(Hedley et al.
1982)


267


Method
Grouping
I








SEPO

I


SEPO









I
SEPO

SEPO

SEPO













Table B3. Continued
Reference
(Diaz et al. 2006)







(Dunne et al. 2007)


Location
FI, USA







FI, USA


Wetland name
Canal L 7
Canal L 39
Canal L 40
Canal L 6
Canal L 5
Canal L 38
Miami Canal north
Miami Canal south
Larson depressional
Beaty depressional


Vegetationt
Riverine
Riverine
Riverine
Riverine
Riverine
Riverine
Riverine
Riverine
PE
PE


Depth
(cm)
0-10
0-10
0-10
0-10
0-10
0-10
0-10
0-10
0-10
0-10


t Vegetation type Cowardin (1979) PE = Persistent emergent (Herbaceou


pH Total P Organic P M d Me
S (mg g-) (%) Method Gri
nr 129t 37 (Hieltjes and SE
nr 74t 45 Lijklema 1980)
nr 69t 45
nr 60t 35
nr 89t 41
nr 164t 32
nr 669t 36
nr 137t 36
5.4 479 90 (Ivanoff et al. SE
5.0 432 92 1998)
s) PE- moss = persistent emergent (moss


thod
ouping
PI


PO


dominated), PE- forested = persistent emergent (woddy vegetation dominated), PE- agriculture = persistent emergent
(crop dominated), sediment = Lacustrine or riverine systems.
$ pg P cm-3
Sample depth s = surface


268









APPENDIX B4
STABILITY OF STANDARD BIOGENIC P DURING ALKALINE EXTRACTION AND
LYPHOLIZATION


Introduction

The hydrolysis of certain biogenic P compounds is recognized as a potential

source of error during routine alkaline extraction and solution 31P NMR spectra

acquisition (Turner et al. 2003d). The effect of freezing and lyopholization as a method

of concentrating sample extracts has also been suggested as an additional source of

error (Cade-Menun et al. 2006) this experiment saught to determine the impact of

freezing standard biogenic P compounds at both -20 and -80C before lypholization

and resuspension under standard conditions

Materials and Methods

Biogenic P standards (Table B4-1) were made up at ~200 pg P mL-1 in both DDI

and a standard alkaline extract solution (0.25 mol L-1 NaOH and 50 mmol L-1 EDTA).

Four replicate sub-samples (0.5 mL) were pipetted into 1.5 mL micro centrifuge tubes

and combined with an MDP standard (0.5 mL, ~50 pg P mL-1) made up in the

corresponding matrix. Of the four replicates; one was analyzed fresh, one was frozen at

-20C and two were frozen at -80C. Frozen samples were maintained for at least 24 h

before being lyophilized. Lyophilized samples were then fully re-suspended in either (1

mL DDI and 0.1 mL D20) or (1 mol L-1 NaOH 100 mmol L-1 EDTA and 0.1 mL D20)

before being immediately analyzed by solution 31P NMR spectroscopy (Figure B4-1). All

samples were also analyzed after 24 h to assess stability in final resuspension matrix

over time. Given the residual salts in lyophilized material and dilution due to D20

(required for NMR signal lock) final matrices are considered; either DDI, 0.23 mol L-1


269









NaOH (~pH 13), 0.909 mol L-1 NaOH, or 1.14 mol L-1 NaOH (pH >13). During the

acquisition process it became apparent that most compounds were not effected by the

freeze drying process. For the sake of expediency, if samples made up in an alkaline

solution and frozen at -20C showed no evidence of hydrolysis, -80C samples and

those frozen in original matrix 1 (DDI) were not analyzed.

Spectra Acquisition

Spectra were acquired using a Bruker Avance 500 Console, Magnex 11.75 T/54

mm Magnet, using a 5 mm BBO Probe. Spectra acquisition was carried out at a

stabilized 25C with a 4.833 psec (-300) pulse length and a 2.00 s recycle delay.

Results presented here are of 2000 scans (total run length 1 h 21 min). Post acquisition

processing was carried out using Acorn NUTS (NMR Utility Transform Software) (Acorn

NMR Inc, Livermore, CA) with 8 Hz of line broadening applied before integration and

graphical representation. All spectra were referenced against MDP as an internal

standard, set at 6 = 17.46 ppm. It should be remembered that when comparing between

matrices we are dealing with a situation where the 'real' chemical shift of the D20 lock,

the MDP and the compound of interest may all vary, depending upon pKa values and

the degree of shielding. The use of an internal reference standard, as apposed to an

externally held reference (i.e. H3P04 within a coaxial insert) leads to an inability to use

resulting spectra as a matrix signal library, yet it still provides information on general

chemical structure of the compounds and their potential hydrolysis which, when

compared to orthophosphate standard in the matching matrix, can be used to indicate

the nature of hydrolysis products.

Results and Conclusions


270









There was an expected influence of pH on the chemical shift on all compounds

tested, although the large organic macromolecule RNA showed only limited chemical

shift (Figure B4-2, Table B4-1). There was also a compound specific interaction, d-

glucose 6-phosphate demonstrated peak splitting when analyzed after resuspension

with DDI (Figure B4-3A), which although minimized at the pH often applied for routine

analysis may lead to erroneous peak identification and therefore demands care when

comparing studies using different re-suspension medium. Most compounds tested were

stable in DDI, 0.23, 0.909 and 1.14 mol L-1 NaOH over the course of 24 h (data not

shown). Exceptions were polyphosphate, which showed slight (<5%) hydrolysis to

orthophosphate in 1.02 mol L-1 and RNA which saw complete hydrolysis to specific

phosphomonoesters detectable even as the initial spectra were being acquired in 1.02

mol L-1 (i.e. <1h 22 min). These specific phophomonoesters (Figure B4-3) correspond

well with previous signals identified after alkaline hydrolysis (Turner et al. 2003d) and

probably represent phosphomonoester nucleotides. There was no hydrolysis of model

compounds tested as a result of the freezing and lypholization step, even when carried

out at -200C.

As a result it is concluded that lyophilization represents a mild and non destructive

method for increasing alkaline sample concentrations. Although -80C is recommended

for snap freezing of samples prior to lyophilization, it appears that use of the more

commonly available -20C is unlikely to result in significant sample alteration. Model

compounds tested showed expected stability in DDI and strongly alkaline conditions,

with the phosphodiester RNA showing rapid hydrolysis at pH>13. The spectra collected

of RNA within a matrix of 0.23 mol L-1 NaOH, showed surprisingly little hydrolysis.


271









Further work is required to test the veracity of this and the potential that RNA may be

stable under alkaline conditions with a pH <13. If so, the resuspension matrix used for

resuspension may be modified to allow for routine peak assignments but limit hydrolysis

of functional forms such as RNA..


Table B4-1 Compounds and chemical shifts tested for stability during lypophilization
0.23 mol L1 1.14 mol L1
Compound DDI
NaOH NaOH


Orthophosphate

2-Aminoethylphosphonic acid

Polyphosphate (individual
peaks an main peak cluster)
Pyrophosphate

d-Glucose 6-phosphate


AMP

RNA


1.17

20.00

-9.48
(-20.4)-(-22.9)
-6.16

1.012


1.115

-1.14


1.92

19.62

-7.71
(-20.0)-(21.3)
-7.40

3.74
3.64
2.94

0.66, 0.31


6.21

20.69

-3.80,-4.4
(-18.9-22.6)
-4.64

5.64
5.08
4.91

Decomposed
see Figure E -3


272











Biogenic P Compounds
1) Orthophosphate
2) 2-Aminoethylphosphonic acid
3) Sodium polyphosphate
4) Sodium pyrophosphate
5) d-Glucose 6 phosphate
6)AMP
7) RNA

MDPASpike
Matrix 1: DDI
Matrix 2:0.25 mol L1 NaOH
S50 mmol L- EDTA)
I


Matrix 2:0.25 mol L-' NaOH
50 mmol L EDTA


Fresh


Frozen @ -20 "C
Lyophilized


Frozen @ -80 "C
Lyophilized


Re-suspension


Solution "P NMR Spectra acquired


Immediate


lmL(lmol L-' NaOH ,100
mmol L-1 EDTA) + 0.1 mL D1O


SImL DDI + 0.1 mL D20



lmL(1mol L NaOH ,100
mmol L- EDTA)+ 0.1 mLD2 0 I "
VI


Figure B4-1. Experimental schemtaic for test of biogenic P stability during lyopilization





273


24hr


i 1 IJhl




~ iI -I-Lhl


I
I
I
I
I
I
I
t5;
















































- --b-l----- -W-__ _--..- ~ ..~r. .-


Orthophosphate





2-Aminoethylphosphonic acid





Polyphosphate





Pyrophosphate





d-Glucose 6-phosphate





AMP


Ji s.-- RNA


20 15 10 5 0 -5 -10 -15 -20
Chemical Shift (ppm)

Figure B4-2. Solution 31P NMR spectra in various matrix environments


274


Matrix
DDI
0.227 mol L-1 NaOH -
1.136 mol L-1 NaOH ......


~~)*LLI._I l^.i. C_~Y -0- ,


..---- 1__~._- .~~L__.__ c __-_.__.~












DDI
0.227 mol L-1 NaOH
1.136 mol L-1 NaOH ......


Chemical Shift (ppm)


^NVAAMWIe
I I I I I
6
Chemical Shift


Matrix Peak
(ppm)
1.136 mol L-1 NaOH 5.03
4.94
4.71
4.57
4.50


I I I


(ppm)


Figure B4-3. Detail of solution 31P NMR spectra of standard biogenic phosphorus
compounds after lyophilization A) Glucose 6 phosphate showing pH
dependant peak splitting B) Alkaline hydrolysis products of RNA


275


Matrix


I









APPENDIX C
ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 4



C1 COMPARISON OF 31P NMR SPECTRA FROM UNFILTERED AND
FILTERED REDISSOLVED SOIL EXTRACTS. SHOWING EVIDENCE
FOR FIELD HETEROGENEITY

C2 SOLUTION 31P NMR SPECTRA AND INTEGRALS COMPARING
UNFILTERED AND FILTERED LYOPHILIZED AND REDISSOLVED
EXTRACTS

C3 CHEMICAL SHIFTS OF ORTHOPHOSPHATE AND INTERNAL
STANDARD MDP DETERMINED BY COMPARISON TO EXTERNALLY
HELD STANDARD 85% H3PO4


276








APPENDIX C1. Comparison of solution 31P NMR spectra from unfiltered and filtered redissolved soil extracts. Sample
represents alkaline extract of calcareous fen, Belize. Redissolved in 2.7 mL (1 mol L-1 NaOH 100 mmol L- EDTA ) and
0.3 mL D20. Sample analyzed using 500 MHz Bruker 10 mm probe, with and without filtering using 0.2 pm GF-B filter.
Spectra were acquired using a 10 mm BBO probe and 500 MHz 11.4 T magnet, a calibrated 30 pulse angle and
~40,000 scans. Spectra were referenced and integrated against internal standard MDP set at 6 = 17.46 ppm.


Filtered sample


Unfiltered sample


277


20 10 0 -10 -20
Chemical shift (ppm)











APPEDNDIX C2. Solution 31P NMR spectra comparing unfiltered and filtered
redissolved lyophilized extracts. Sample represents A) ombrotrophic wetland soil B)
Raphia tadeiga plant tissue. Samples Lyophilized and redissolved in 1 mol L-1 NaOH
100 mmol L-1 EDTA samples either loaded directly or pre filtered using 0.2 pm GF-B
filter. Spectra were acquired using a 10 mm BBO probe and 500 MHz 11.4 T magnet, a
calibrated 30 pulse angle and ~40,000 scans. Spectra were referenced and integrated
against internal standard MDP set at 6 = 17.46 ppm.


Sample


Not filtered
A Filtered
RSD (%)


18.4
17.5
3.6


C.
0

0
0
o


178.8
157.3
9.1


0

0
E
e0
Q.
0
a-,

121.8
137.1
8.4


a-,

0



19.8
6.1
74.9


Z





74.8
70.4
4.3


C>

0
C:
Q.
0



101.3
82.0
14.8


Not filtered 0.0 61.1 27.5 9.0 18.4 22.0
B Filtered 0.0 62.5 29.3 10.8 18.2 24.2
RSD (%) 1.6 4.5 13.1 0.7 7.0


Z
z

I-

514.7
470.3
6.4

137.9
145.1
3.6


H*A^ Filtered



.Unfiltered


278


20 10 0 -10 -20
Chemical shift (ppm)







APPENDIX C3. Chemical shifts of orthophosphate and internal standard MDP determined by comparison to externally
held standard 85% H3P04. Spectra acquired using 500 MHz 11.7 T field strength magnet using a calibrated 90 pulse,
zgig pulse program and 2000 scans.


10 mm NMR tube and 9 mm coaxial insert


-300 mg mL-1 lyophilized soil extract redissolved
in 2.7 ml 1 mol L-1 NaOH 100 mmol L-1 EDTA

4 85 % H3P04



MDP =17.46 P04= 6.206


1 1


85% H3P04 = 0


[unidentified]- presumed
contamination of polyphosphates in
85% H3P04




4


Chemical


0
shift (ppm)


279









APPENDIX D
SIMPLIFIED METHOD FOR DETERMINATION OF TOTAL PHOSPHORUS IN
WETLAND SAMPLES

This represents a simplified method for the determination of total P in high organic
matter wetland soil samples. Based upon Saunders and Williams (1955) it employs a
single vessel for ashing of samples before applying 1 mol L-1 H2SO4 to solubilize
liberated P. Experimental evidence is provided which shows insignificant difference
between proposed method and established Anderson digestion procedure.

Chemicals
1. Sulfuric acid, concentrated (H2SO4; FW 98.08)
2. Primary orthophosphate standard 1000 mg P L-1

Reagents
1. 1 M H2SO4: In a 2 L volumetric flask containing 1.5 L distilled deionized water
(DDI) slowly add 111.1 mL of concentrated H2SO4. After it has cooled make
up to volume with (DDI).

2. Orthophosphate 2nd standard: In a 100 mL volumetric flasks dilute 10 mL
1000mg P L-1 with DDI water to create a 2nd standards of 100 mg P L-1

3. Certified or internal standard soil and plant materials:

QA/QC
Include; 5% method duplicates, 5% method spikes (include 1 mL of 100 mg P L-1
orthophosphate 20 standard) and at least one external standard soil and plant
material per procedural batch.

Method
1. Label and pre weigh clean borosilicate glass scintillation vials. (Note ensure that
vials are not chipped to avoid spillage)
2. Weigh out into pre-weighed vials ~0.2 g ( 0.01 g) of finely ground 'air dried'
sample. Record sample weight to the nearest mg.
3. Optional: Place racked vials within a controlled oven and dry samples at 70C for
72 h. Allow to cool within sealed desiccator with fresh desiccant before
reweighing.
4. Optional: Place racked vials within a controlled oven and dry samples at 105C
for 24 h. Allow to cool within sealed desiccator with fresh desiccant before
reweighing.
5. Load vials into muffle furnace and follow temperature program set out below.
(Note exact timings of heating/cooling stages will depend on specifics of furnace
being used)


280











550C


6. Remove samples form the muffle furnace and allow to cool in a sealed desiccator
before reweighing to determine ash weight.
7. To each vial slowly 20 mL 1mol L-1 H2S04 from a recalibrated repipettor (Note
care must be taken with carbonate containing samples given evolution of CO2)
8. After the evolution of any gases, cap with a HDPE cone containing cap and place
on a reciprocating shaker at low speed for 24 h.
9. After shaking allow to settle for 24 h
10. Carefully decant overlying solution into prelabeled plastic scintillation vials and
store at room temperature to await analysis
11. Colorimetric analysis is carried out on samples after a x10 dilution to bring
phosphorus and acidity levels to within the working range.

Experimental evidence
Eight standards including; a range of homogenized wetland soil samples and 4
National Institute for Standards and Technology (NIST) Standard References
Materials compared using proposed vial method and established Anderson (1976)
ashing technique. Paired t-test showed only one standard (a domestic sludge) to return
significantly different values for total P.

Table D-1. Standard biogenic P compounds tested for recovery by anion exchange
membrane strip


Std Name
1 Std Sand
2 Std Peat
3 Std Histol
4 SRM 1547
5 SRM 2710
6 Std Periphyton
7 SRM 2709
8 SRM 2781


Description
Acid washed sand
Peat, ombrotrophic peat dome, Panama
Internal lab standard, Everglades
NIST- Peach leaves
NIST Montana soil
Periphyton, WCA-2A
NIST San Joaquin Soil
NIST Domestic sludge


Certified total
P (pg g1)



1370 70
1060 150

620 50
24200 900


281












1,500-



1X-
1,000-



500-



0




30,000-



20,000-



10,000-



0-


vCv
- -0


C -ir i (D 0o
-0 -0 4-0 4-40 4-0 4-0
wO CO wO C wO


B

D Vial Method
E Anderson digest











'% 0 "" L) (D I"- 00

.) 0U) (1 0) 03 01 0" w


Figure D-1 Comparison of total P determined by TP Ash vial method and Andersen
(1976) procedure. Average (n=3) standard deviation (*) = significant
difference (Paired t-test p < 0.05)


282


Kt IJ IlJl :I i I I- I L Ii


0
I-
c5









LIST OF REFERENCES


Ahlgren J, De Brabandere H, Reitzel K, Rydin E, Gogoll A & Waldeback M (2007)
Sediment phosphorus extractants for phosphorus-31 nuclear magnetic resonance
analysis: A quantitative evaluation. J. Environ. Qual. 36: 892-898.

Ahlgren J, Reitzel K, Danielsson R, Gogoll A & Rydin E (2006a) Biogenic phosphorus in
oligotrophic mountain lake sediments: Differences in composition measured with
NMR spectroscopy. Water Res. 40: 3705-3712.

Ahlgren J, Reitzel K, Tranvik L, Gogoll A & Rydin E (2006b) Degradation of organic
phosphorus compounds in anoxic Baltic Sea sediments: A 31P nuclear magnetic
resonance study. Limnol. Oceanogr. 51: 2341-2348.

Ahlgren J, Tranvik L, Gogoll A, Waldeback M, Markides K & Rydin E (2005) Sediment
depth attenuation of biogenic phosphorus compounds measured by 31P NMR.
Environ. Sci. Technol. 39: 867-872.

Allison SD & Vitousek PM (2005) Responses of extracellular enzymes to simple and
complex nutrient inputs. Soil Biol. Biochem. 37: 937-944.

Alvarez JA & Becares E (2006) Seasonal decomposition of Typha latifolia in a free-
water surface constructed wetland. Ecol. Eng. 28: 99-105.

Ammerman JW & Azam F (1985) Bacterial 5'-nucleotidase in aquatic ecosystems a
novel mechanism of phosphorus regeneration. Science 227: 1338-1340.

Andersen JM (1976) An ignition method for determination of total phosphorus in lake
sediments. Water Res. 10: 329-331.

Anderson JAR (1983) The tropical peat swamps of western Malesia.ln: Mires: swamp,
bog, fen and moor. Regional studies. Volume 4B, pp 181-199. Elsevier Scientific
Publishing Company, Amsterdam Netherlands

Anderson JAR & Muller J (1975) Palynological Study of A Holocene Peat and A
Miocene Coal Deposit from Nw Borneo. Rev. Palaeobot. Palynol. 19: 291-351.

Andriesse JP (1988) Nature and management of tropical peat soils. FAO, Rome

Aschar-Sobbi R, Abramov AY, Diao C, Kargacin ME, Kargacin GJ, French RJ & Pavlov
E (2008) High sensitivity, quantitative measurements of polyphosphate using a
new DAPI-Based approach. J. Fluoresc. 18: 859-866.

Aspila KI, Agemian H & Chau ASY (1976) Semiautomated method for determination of
inorganic, organic and total phosphate in sediments. Analyst 101: 187-197.

Axt JR & Walbridge MR (1999) Phosphate removal capacity of palustrine forested
wetlands and adjacent uplands in Virginia. Soil Sci. Soc. Am. J. 63: 1019-1031.


283









Azam F, Fenchel T, Field JG, Gray JS, Meyerreil LA & Thingstad F (1983) The
ecological role of water-column microbes in the sea. Mar. Ecol. Prog. Ser. 10: 257-
263.

Babatunde AO, Zhao YQ, O'Neill M & O'Sullivan B (2008) Constructed wetlands for
environmental pollution control: A review of developments, research and practice
in Ireland. Environment International 34: 116-126.

Bai XL, Ding SM, Fan CX, Liu T, Shi D & Zhang L (2009) Organic phosphorus species
in surface sediments of a large, shallow, eutrophic lake, Lake Taihu, China.
Environ. Pollut. 157: 2507-2513.

Bakken LR (1997) Culturable and nonculturable bacteria in soil.In: Modern soil
microbiology., pp 47-61. Marcel Dekker Inc., New York USA

Baldwin DS (1996) The phosphorus composition of a diverse series of Australian
sediments. Hydrobiologia 335: 63-73.

Baldwin DS (1998) Reactive "organic" phosphorus revisited. Water Res. 32: 2265-2270.

Ballantine D, Walling DE & Leeks GJL (2009) Mobilisation and transport of sediment-
associated phosphorus by surface runoff. Water Air Soil Pollut. 196: 311-320.

Barbanti A, Bergamini MC, Frascari F, Miserocchi S & Rosso G (1994) Critical Aspects
of Sedimentary Phosphorus Chemical Fractionation. J. Environ. Qual. 23: 1093-
1102.

Bardygulanonn LG, Kaster JL & Glonek T (1995) Phospholipid profiling of sediments
using 31P nuclear-magnetic- resonance. Lipids 30: 1047-1051.

Barrientos LG & Murthy PPN (1996) Conformational studies of myo-inositol phosphates.
Carbohydr. Res. 296: 39-54.

Bedrock CN, Cheshire MV, Chudek JA, Goodman BA & Shand CA (1994) Use of 31P
NMR to study the forms of phosphorus in peat soils. Sci. Total Environ. 152: 1-8.

Belyea LR & Baird AJ (2006) Beyond "The limits to peat bog growth": Cross-scale
feedback in peatland development. Ecol. Monogr. 76: 299-322.

Belyea LR & Clymo RS (2001) Feedback control of the rate of peat formation. Proc. R.
Soc. Lond. Ser. B-Biol. Sci. 268: 1315-1321.

Benitez-Nelson CR, O'Neill L, Kolowith LC, Pellechia P & Thunell R (2004)
Phosphonates and particulate organic phosphorus cycling in an anoxic marine
basin. Limnol. Oceanogr. 49: 1593-1604.

Bennett EM, Carpenter SR & Caraco NF (2001) Human impact on erodable phosphorus
and eutrophication: A global perspective. Bioscience 51: 227.


284









Berger S, Braun S & Kalinowski HO (1997) NMR Spectroscopy of the Non Metallic
Elements. John Wiley & Sons, Hoboken

Bertrand I, McLaughlin M, Holloway R, Armstrong R & McBeath T (2006) Changes in P
bioavailability induced by the application of liquid and powder sources of P, N and
Zn fertilizers in alkaline soils. Nutr. Cycl. Agroecosyst. 74: 27-40.

Bhadha JH & Jawitz JW Characterizing deep soils from an impacted subtropical
isolated wetland: implications for phosphorus storage. J. Soils Sed. 10: 514-525.

Blake RE, O'Neil JR & Surkov AV (2005) Biogeochemical cycling of phosphorus:
Insights from oxygen isotope effects of phosphoenzymes. Am. J. Sci. 305: 596-
620.

Boavida MJ & Wetzel RG (1998) Inhibition of phosphatase activity by dissolved humic
substances and hydrolytic reactivation by natural ultraviolet light. Freshw. Biol. 40:
285-293.

Bowman RA (1989) A sequential extraction procedure with concentrated sulfuric acid
and dilute base for soil organic phosphorus. Soil Sci. Soc. Am. J. 53: 362-366.

Bowman RA & Moir JO (1993) Basic EDTA as an extractant for soil organic
phosphorus. Soil Sci. Soc. Am. J. 57: 1516-1518.

Bradford MA, Tordoff GM, Eggers T, Jones TH & Newington JE (2002) Microbiota,
fauna, and mesh size interactions in litter decomposition. Oikos 99: 317-323.

Brandes JA, Ingall E & Paterson D (2007) Characterization of minerals and organic
phosphorus species in marine sediments using soft X-ray fluorescence
spectromicroscopy. Mar. Chem. 103: 250-265.

Brannon CA & Sommers LE (1985) Stability and mineralization of organic phosphorus
incorporated into model humic polymers. Soil Biol. Biochem. 17: 221-227.

Bray RH & Kurtz LT (1945) Determination of total, organic, and available forms of
phosphorus in soils. Soil Sci. 59: 39-45.

Brinson MM (1977) Decomposition and nutrient exchange of litter in an alluvial swamp
forest. Ecology 58: 601-609.

Brookes PC, Powlson DS & Jenkinson DS (1982) Measurement of microbial biomass
phosphorus in soil. Soil Biol. Biochem. 14: 319-329.

Brookes PC, Powlson DS & Jenkinson DS (1984) Phosphorus in the soil microbial
biomass. Soil Biol. Biochem. 16: 169-175.

Brown MRW & Kornberg A (2004) Inorganic polyphosphate in the origin and survival of
species. Proc. Natl. Acad. Sci. USA 101: 16085-16087.


285









Bruland GL & Richardson CJ (2006) An assessment of the phosphorus retention
capacity of wetlands in the Painter Creek watershed, Minnesota, USA. Water Air
Soil Pollut. 171: 169-184.

Bunemann EK (2008) Enzyme additions as a tool to assess the potential bioavailability
of organically bound nutrients. Soil Biol. Biochem. 40: 2116-2129.

Bunemann EK, Marschner P, Smernik RJ, Conyers M & McNeill AM (2008a) Soil
organic phosphorus and microbial community composition as affected by 26 years
of different management strategies. Biol. Fertil. Soils 44: 717-726.

Bunemann EK, Smernik RJ, Doolette AL, Marschner P, Stonor R, Wakelin SA & McNeill
AM (2008b) Forms of phosphorus in bacteria and fungi isolated from two
Australian soils. Soil Biol. Biochem. 40: 1908-1915.

Bunemann EK, Smernik RJ, Marschner P & McNeill AM (2008c) Microbial synthesis of
organic and condensed forms of phosphorus in acid and calcareous soils. Soil
Biol. Biochem. 40: 932-946.

Burnham KP & Anderson DR (2004) Multimodel inference understanding AIC and BIC
in model selection. Sociological Methods & Research 33: 261-304.

Cade-Menun BJ (2005a) Characterizing phosphorus in environmental and agricultural
samples by 31P nuclear magnetic resonance spectroscopy. Talanta 66: 359-371.

Cade-Menun BJ (2005b) Using phosphorus 31 nuclear magnetic resonance
spectroscopy to characterize organic phosphorus in environmental samples. In:
Turner BL, Frossard E & Baldwin DS (Eds) Organic phosphorus in the
environment, pp 21-44. CABI Publishing, Wallingford UK

Cade-Menun BJ, Benitez-Nelson CR, Pellechia P & Paytan A (2005) Refining 31P
nuclear magnetic resonance spectroscopy for marine particulate samples: Storage
conditions and extraction recovery. Mar. Chem. 97: 293-306.

Cade-Menun BJ, Liu CW, Nunlist R & McColl JG (2002) Soil and litter phosphorus 31
nuclear magnetic resonance spectroscopy: Extractants, metals, and phosphorus
relaxation times. J. Environ. Qual. 31: 457-465.

Cade-Menun BJ, Navaratnam JA & Walbridge MR (2006) Characterizing dissolved and
particulate phosphorus in water with 31P nuclear magnetic resonance
spectroscopy. Environ. Sci. Technol. 40: 7874-7880.

Cade-Menun BJ & Preston CM (1996) A comparison of soil extraction procedures for
31P NMR spectroscopy. Soil Sci. 161: 770-785.

Caldwell BA (2005) Enzyme activities as a component of soil biodiversity: a review.
Pedobiologia 49: 637-644.


286









Canet D (1996) Nuclear Magnetic Resonance: Concepts and Methods. Wiley, Hoboken,
NJ

Capece JC, Campbell KL, Bohlen PJ, Graetz DA & Portier KM (2007) Soil phosphorus,
cattle stocking rates, and water quality in subtropical pastures in Florida, USA.
Rangeland Ecol. Manage. 60: 19-30.

Caraco NF (1993) Disturbance of the phosphorus cycle a case of indirect effects of
human activity. Trends Ecol. Evol. 8: 51-54.

Carman R, Edlund G & Damberg C (2000) Distribution of organic and inorganic
phosphorus compounds in marine and lacustrine sediments: a P-31 NMR study.
Chem. Geol. 163: 101-114.

Carpenter SR, Caraco NF, Correll DL, Howarth RW, Sharpley AN & Smith VH (1998)
Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl. 8:
559-568.

CCAD (2002) Politica Centroamericana para la Conservacion y Uso Racional de los
Humedales. CCAD, San Jose, Costa Rica

Celi L & Barberis E (2005a) Abiotic Stabilization of Organic Phosphorus in the
Environment.In: Turner BL, Frossard E & Baldwin DS (Eds) Organic Phosphorus
in the Environment, pp 113-132. CABI Publishing,

Celi L & Barberis E (2005b) Abiotic stabilization of organic posphorus in the
environment. In: Turner BL, Frossard E & Baldwin DS (Eds) Organic phosphorus in
the environment, pp 113-132. CABI Publishing, Wallingford, UK

Celi L & Barberis E (2007) Abiotic reactions of inositol phosphates in soil.In: Turner BL,
Richardson AE & Mullaney EJ (Eds) Inositol phosphates: Linking agriculture and
the environment, pp 207-220. Cabi, Wallingford UK

Cembella AD, Antia NJ & Taylor FJR (1986) The determination of total phosphorus in
seawater by nitrate oxidation of the organic-component. Water Res. 20: 1197-
1199.

Chadwick OA, Derry LA, Vitousek PM, Huebert BJ & Hedin LO (1999) Changing
sources of nutrients during four million years of ecosystem development. Nature
397: 491.

Chang SC & Jackson ML (1957) Fractionation of soil phosphorus. Soil Sci 84: 133-144.

Chapuis-Lardy L, Brossard M & Quiquampoix H (2001) Assessing organic phosphorus
status of Cerrado oxisols (Brazil) using 31P NMR spectroscopy and
phosphomonoesterase activity measurement. Can. J. Soil Sci. 81: 591-601.


287









Cheesman AW, Dunne EJ, Turner BL & Reddy KR (2010a) Soil phosphorus forms in
hydrologically isolated wetlands and surrounding pasture uplands. J. Environ.
Qual. (in press):

Cheesman AW, Turner BL & Reddy KR (2010b) Interaction of organic and condensed
inorganic phosphorus compounds with anion-exchange membranes: Implications
for soil phosphorus analysis. Soil Sci Soc Am J (in press):

Christophoridis C & Fytianos K (2006) Conditions affecting the release of phosphorus
from surface lake sediments. J. Environ. Qual. 35: 1181-1192.

Clark LL, Ingall ED & Benner R (1999) Marine organic phosphorus cycling: Novel
insights from nuclear magnetic resonance. Am. J. Sci. 299: 724-737.

Cleveland CC & Liptzin D (2007) C : N : P stoichiometry in soil: is there a "Redfield
ratio" for the microbial biomass? Biogeochemistry 85: 235-252.

Cohen AD, Raymond R, Ramirez A, Morales Z & Ponce F (1989) The Changuinola peat
deposit of northwestern Panama A tropical, back-barrier, peat (coal)-forming
environment. Int. J. Coal Geol. 12: 157-192.

Cohen AD, Raymond R, Ramirez A, Morales Z & Ponce F (1990) The Changuinola peat
deposit of northwestern Panama A tropical, domed, back-barrier coalforming
environment. Int. J. Coal Geol. 16: 139-142.

Condron LM, Frossard E, Newman RH, Tekely P & Morel JL (1997) Use of 31P NMR in
the study of soils and the environment. In: Nanny MA, Minear RA & Leenheer JA
(Eds) Nuclear Magnetic Resonance Spectroscopy in Environmental Chemistry, pp
247-271. Oxford University Press, Oxford

Condron LM, Frossard E, Tiessen H, Newman RH & Stewart JWB (1990a) Chemical
nature of organic phosphorus in cultivated and uncultivated soils under different
environmental conditions. J. Soil Sci. 41: 41-50.

Condron LM, Moir JO, Tiessen H & Stewart JWB (1990b) Critical-evaluation of methods
for determining total organic phosphorus in tropical soils. Soil Sci. Soc. Am. J. 54:
1261-1266.

Condron LM, Turner BL & Cade-Menun BJ (2005) Chemistry and dynamics of soil
organic phosphorus. In: Sims JT & Sharpley AN (Eds) Phosphorus: agriculture and
the environment, pp 87-121. American Society of Agronomy, Madison USA

Conley DJ, Schelske CL & Stoermer EF (1993) Modification of the biogeochemical
cycle of silica with eutrophication. Mar. Ecol. Prog. Ser. 101: 179-192.

Conte P, Smejkalova D, Piccolo A & Spaccini R (2008) Evaluation of the factors
affecting direct polarization solid state 31P NMR spectroscopy of bulk soils. Eur. J.
Soil Sci. 59: 584-591.


288









Cooke JG (1992) Phosphorus Removal Processes in a Wetland after a Decade of
Receiving a Sewage Effluent. J. Environ. Qual. 21: 733-739.

Cooke JG, Cooper AB & Clunie NMU (1990) Changes in the Water, Soil, and
Vegetation of a Wetland after a Decade of Receiving a Sewage Effluent. N. Z. J.
Ecol. 14: 37-47.

Cooke JG, Stub L & Mora N (1992) Fractionation of Phosphorus in the Sediment of a
Wetland after a Decade of Receiving Sewage Effluent. J. Environ. Qual. 21: 726-
732.

Cooper WT, Llewelyn JM, Bennett GL, Stenson AC & Salters VJM (2005) Organic
phosporus speciation in natural waters by mass spectrometry.In: Turner BL,
Frossard E & Baldwin DS (Eds) Organic phosphorus in the environment, pp 45-74.
CABI Publishing, Wallingford UK

Cooperband LR, Gale PM & Comerford NB (1999) Refinement of the anion exchange
membrane method for soluble phosphorus measurement. Soil Sci. Soc. Am. J. 63:
58-64.

Cooperband LR & Logan TJ (1994) Measuring in-situ changes in labile soil-phosphorus
with anion-exchange membranes. Soil Sci. Soc. Am. J. 58: 105-114.

Corstanje R & Reddy KR (2006) Microbial indicators of nutrient enrichment: A
mesocosm study. Soil Sci. Soc. Am. J. 70: 1652-1661.

Corstanje R, Reddy KR & Portier KM (2006) Typha latifolia and Cladiumjamaicense
litter decay in response to exogenous nutrient enrichment. Aquat. Bot. 84: 70-78.

Corstanje R, Reddy KR, Prenger JP, Newman S & Ogram AV (2007) Soil microbial eco-
physiological response to nutrient enrichment in a sub-tropical wetland. Ecol. Indic.
7: 277-289.

Cosgrove DJ (1966) Detection of Isomers of phytic acid in Some Scottish and
Californian Soils. Soil Sci. 102: 42-43.

Costanza R, dArge R, deGroot R, Farber S, Grasso M, Hannon B, Limburg K, Naeem
S, Oneill RV, Paruelo J, Raskin RG, Sutton P & vandenBelt M (1997) The value of
the world's ecosystem services and natural capital. Nature 387: 253-260.

Cotner JB & Wetzel RG (1991) 5'-nucleotidase activity in a eutrophic lake and an
oligotrophic lake. Appl. Environ. Microbiol. 57: 1306-1312.

Cowardin LM, Carter V, Golet FC & LaRoe ET (1979) Classification of wetlands and
deepwater habitats of the United States. U.S. Fish and Wildlife Service,
Washington


289









Craft CB & Chiang C (2002) Forms and amounts of soil nitrogen and phosphorus
across a longleaf pine-depressional wetland landscape. Soil Sci. Soc. Am. J. 66:
1713.

Craft CB & Richardson CJ (1993) Peat accretion and N, P, and organic C accumulation
in nutrient-enriched and unenriched Everglades peatlands. Ecol. Appl. 3: 446-458.

D'Croz L, Del Rosario JB & Gondola P (2005) The effect of fresh water runoff on the
distribution of dissolved inorganic nutrients and plankton in the Bocas del Toro
archipelago, Caribbean Panama. Caribbean Journal of Science 41: 414-429.

Daniel TC, Sharpley AN & Lemunyon JL (1998) Agricultural phosphorus and
eutrophication: A symposium overview. J. Environ. Qual. 27: 251-257.

Darke AK & Walbridge MR (2000) Al and Fe biogeochemistry in a floodplain forest:
Implications for P retention. Biogeochemistry 51: 1-32.

Davelaar D (1993) Ecological significance of bacterial polyphosphate metabolism in
sediments. Hydrobiologia 253: 179-192.

Davis SE, Childers DL & Noe GB (2006) The contribution of leaching to the rapid
release of nutrients and carbon in the early decay of wetland vegetation.
Hydrobiologia 569: 87-97.

Davis SM (1991) Growth, decomposition, and nutrient retention of Cladium-jamaicense
Crantz and Typha domingensis Pers in the Florida Everglades. Aquat. Bot. 40:
203-224.

Day JW, Ko JY, Rybczyk J, Sabins D, Bean R, Berthelot G, Brantley C, Cardoch L,
Conner W, Day JN, Englande AJ, Feagley S, Hyfield E, Lane R, Lindsey J, Mistich
J, Reyes E & Twilley R (2004) The use of wetlands in the Mississippi Delta for
wastewater assimilation: A review. Ocean Coast. Manage. 47: 671-691.

De Groot CJ & Golterman HL (1990) Sequential fractionation of sediment phosphate.
Hydrobiologia 192: 143-148.

De Groot CJ & Golterman HL (1993) On the presence of organic phosphate in some
Camargue sediments Evidence for the importance of phytate. Hydrobiologia 252:
117-126.

De Jager HJ & Heyns AM (1998) Study of the hydrolysis of sodium polyphosphate in
water using Raman spectroscopy. Appl. Spectrosc. 52: 808-814.

De Steven D & Toner MM (2004) Vegetation of upper coastal plain depression
wetlands: Environmental templates and wetland dynamics within a landscape
framework. Wetlands 24: 23-42.


290









Dean LA (1938) An attempted fractionation of the soil phosphorus. J. Agric. Sci. 28:
234-246.

DeBusk WF & Reddy KR (2003) Nutrient and hydrology effects on soil respiration in a
northern Everglades marsh. J. Environ. Qual. 32: 702-710.

DeBusk WF & Reddy KR (2005) Litter decomposition and nutrient dynamics in a
phosphorus enriched everglades marsh. Biogeochemistry 75: 217-240.

Deevey ES (1970) In defense of mud. Bull. Ecol. Soc. Am. 51: 5-8.

DeGroot CJ & Golterman HL (1990) Sequential fractionation of sedimnet phosphate.
Hydrobiologia 192: 143-148.

DeGroot CJ & Golterman HL (1993) On the Presence of Organic Phosphate in Some
Camargue Sediments Evidence for the Importance of Phytate. Hydrobiologia
252: 117-126.

Delgado A, Ruiz JR, del Campillo MD, Kassem S & Andreu L (2000) Calcium- and iron-
related phosphorus in calcareous and calcareous marsh soils: Sequential
chemical fractionation and P-31 nuclear magnetic resonance study. Commun. Soil
Sci. Plant Anal. 31: 2483-2499.

Demas GP & Rabenhorst MC (1999) Subaqueous soils: Pedogenesis in a submersed
environment. Soil Sci. Soc. Am. J. 63: 1250-1257.

Demas GP & Rabenhorst MC (2001) Factors of subaqueous soil formation: a system of
quantitative pedology for submersed environments. Geoderma 102: 189-204.

Devai I & Delaune RD (1995) Evidence for phosphine production and emission from
Louisiana and Florida marsh soils. Org. Geochem. 23: 277-279.

Devai I, Felfoldy L, Wittner I & Plosz S (1988) Detection of phosphine- new aspects of
the phosphorus cycle in the hydrosphere. Nature 333: 343-345.

Diaz J, Ingall E, Benitez-Nelson C, Paterson D, de Jonge MD, McNulty I & Brandes JA
(2008) Marine polyphosphate: A key player in geologic phosphorus sequestration.
Science 320: 652-655.

Diaz OA, Daroub SH, Stuck JD, Clark MW, Lang TA & Reddy KR (2006) Sediment
inventory and phosphorus fractions for water conservation area canals in the
everglades. Soil Sci. Soc. Am. J. 70: 863-871.

Dick WA & Tabatabai MA (1977) Determination of orthophosphate in aqueous solutions
containing labile organic and inorganic phosphorus compounds. J. Environ. Qual.
6: 82-85.


291









Dickman SR & De Turk EE (1938) A method for the determination of the organic
phosphorus of soils. Soil Sci. 45: 29-39.

Ding SM, Xu D, Li B, Fan CX & Zhang CS (2010) Improvement of 31P NMR spectral
resolution by 8-hydroxyquinoline precipitation of paramagnetic Fe and Mn in
environmental samples. Environ. Sci. Technol. 44: 2555-2561.

Doolette AL, Smernik RJ & Dougherty WJ (2009) Spiking improved solution
phosphorus-31 nuclear magnetic resonance identification of soil phosphorus
compounds. Soil Sci. Soc. Am. J. 73: 919-927.

Dormaar JF & Webster GR (1964) Losses inherent in ignition procedures for
determining total organic phosphorus. Can. J. Soil Sci. 44: 1.

Drohan PJ, Merkler DJ & Buck BJ (2005) Suitability of the plant root simulator probe for
use in the Mojave Desert. Soil Sci. Soc. Am. J. 69: 1482-1491.

Drouillon M & Merckx R (2005) Performance of para-nitrophenyl phosphate and 4-
methylumbelliferyl phosphate as substrate analogues for phosphomonoesterase in
soils with different organic matter content. Soil Biol. Biochem. 37: 1527-1534.

Dunne EJ, Smith J, Perkins DB, Clark MW, Jawitz JW & Reddy KR (2007) Phosphorus
storage in historically isolated wetland ecosystems and surrounding pasture
uplands. Ecol. Eng. 31: 16-28.

Early TA, Kundrat JT, Schorp T & Glonek T (1996) Lake Michigan sponge phospholipid
variations with habitat: A P-31 nuclear magnetic resonance study. Comp.
Biochem. Physiol. B-Biochem. Mol. Biol. 114: 77-89.

Eiland F (1983) A simple method for quantitative-determination of ATP in soil. Soil Biol.
Biochem. 15: 665-670.

Eisenreich SJ & Armstrong DE (1977) Chromatographic investigation of inositol
phosphate esters in lake waters. Environ. Sci. Technol. 11: 497-501.

Eixler S, Karsten U & Selig U (2006) Phosphorus storage in Chlorella vulgaris
(Trebouxiophyceae, Chlorophyta) cells and its dependence on phosphate supply.
Phycologia 45: 53-60.

Eixler S, Selig U & Karsten U (2005) Extraction and detection methods for
polyphosphate storage in autotrophic planktonic organisms. Hydrobiologia 533:
135-143.

EI-Rifai H, Heerboth M, Gedris TE, Newman S, Orem W & Cooper WT (2008) NMR and
mass spectrometry of phosphorus in wetlands. Eur. J. Soil Sci. 59: 517-525.

Ellison AM (2004) Wetlands of Central America. Wetlands Ecol. Manage. 12: 3-55.


292









Enriquez S, Duarte CM & Sandjensen K (1993) Patterns in decomposition rates among
photosynthetic organisms the importance of detritus C-N-P content. Oecologia
94: 457-471.

Ermakova IT, Shushkova TV & Leont'evskii AA (2008) Microbial degradation of
organophosphonates by soil bacteria. Microbiology 77: 615-620.

Espinosa M, Turner BL & Haygarth PM (1999) Preconcentration and separation of trace
phosphorus compounds in soil leachate. J. Environ. Qual. 28: 1497-1504.

Euliss NH, Labaugh JW, Fredrickson LH, Mushet DM, Laubhan MRK, Swanson GA,
Winter TC, Rosenberry DO & Nelson RD (2004) The wetland continuum: A
conceptual framework for interpreting biological studies. Wetlands 24: 448-458.

Fennessy MS, Rokosch A & Mack JJ (2008) Patterns of plant decomposition and
nutrient cycling in natural and created wetlands. Wetlands 28: 300-310.

Flaig EG & Reddy KR (1995) Fate of phosphorus in the Lake Okeechobee watershed,
Florida, USA: Overview and recommendations. Ecol. Eng. 5: 127-142.

Frink CR (1969) Fractionatton of phosphorus in lake sediments: Analytical evaluation.
Proc. Soil Sci. Soc. Am. 33: 326-328.

Gachter R & Meyer JS (1993) The role of microorganisms in mobilization and fixation of
phosporus in sediments. Hydrobiologia 253: 103-121.

Gaiser EE, Taylor BE & Brooks MJ (2001) Establishment of wetlands on the
southeastern Atlantic Coastal Plain: Paleolimnological evidence of a mid-Holocene
hydrologic threshold from a South Carolina pond. J. Paleolimnol. 26: 373-391.

Gale PM, Reddy KR & Graetz DA (1994) Phosphorus Retention by Wetland Soils Used
for Treated Waste-Water Disposal. J. Environ. Qual. 23: 370-377.

Garcia-Pintado J, Martinez-Mena M, Barbera GG, Albaladejo J & Castillo VM (2007)
Anthropogenic nutrient sources and loads from a Mediterranean catchment into a
coastal lagoon: Mar Menor, Spain. Sci. Total Environ. 373: 220-239.

Gardolinski P, Worsfold PJ & McKelvie ID (2004) Seawater induced release and
transformation of organic and inorganic phosphorus from river sediments. Water
Res. 38: 688-692.

Gassmann G & Glindemann D (1993) Phosphane (PH3) in the biosphere. Angew.
Chem.-Int. Edit. Engl. 32: 761-763.

Gathumbi SM, Bohlen PJ & Graetz DA (2005) Nutrient enrichment of wetland
vegetation and sediments in subtropical pastures. Soil Sci. Soc. Am. J. 69: 539-
548.


293









Geng JJ, Niu XJ, Jin XC, Wang XR, Gu XH, Edwards M & Glindemann D (2005)
Simultaneous monitoring of phosphine and of phosphorus species in Taihu Lake
sediments and phosphine emission from lake sediments. Biogeochemistry 76:
283-298.

Gerke J (1992) Orthophosphate and organic phosphate in the soil solution of 4 sandy
soils in relation to pH-Evidence for humic-Fe-(AI-) phosphate complexes.
Commun. Soil Sci. Plant Anal. 23: 601.

Ghani MO (1942) Determination of organic phosphorus in alkali extracts of soils. Indian
J. Agric. Sci. 12: 336-340.

Ghonsika CP & Miller RH (1973) Soil inorganic polyphosphates of microbial origin. Plant
Soil 38: 651-655.

Giaveno C, Celi L, Richardson AE, Simpson RJ & Barberis E (2010) Interaction of
phytases with minerals and availability of substrate affect the hydrolysis of inositol
phosphates. Soil Biol. Biochem. 42: 491-498.

Giblin AE, Nadelhoffer KJ, Shaver GR, Laundre JA & McKerrow AJ (1991)
Biogeochemical Diversity Along a Riverside Toposequence in Arctic Alaska. Ecol.
Monogr. 61: 415-435.

Gilbin R, Gomez E & Picot B (2000) Phosphorus and organic matter in wetland
sediments: analysis through gel permeation chromatography (GPC). Agronomie
20: 567-576.

Godfray HCJ, Beddington JR, Crute IR, Haddad L, Lawrence D, Muir JF, Pretty J,
Robinson S, Thomas SM & Toulmin C (2010) Food security: The challenge of
feeding 9 billion people. Science 327: 812-818.

Goering HK & Soest PJ (1970) VAN Forage fiber analyses. (Apparatus, reagents,
procedures, and some applications).In: Agriculture Handbook, United States
Department of Agriculture, p 20.

Golimowski J & Golimowska K (1996) UV-photooxidation as pretreatment step in
inorganic analysis of environmental samples. Analytica Chimica Acta 325: 111-
133.

Golterman H, Paing J, Serrano L & Gomez E (1998) Presence of and phosphate
release from polyphosphates or phytate phosphate in lake sediments.
Hydrobiologia 364: 99.

Gornak SI & Zhang J (1999) A summary of landowner surveys and water quality data
from the northern Lake Okeechobee watershed. Appl. Eng. Agric. 15: 121-127.

Grace KA, Dierherg FE, DeBusk TA & White JR (2008) Phosphorus uptake by Typha
leaf litter as affected by oxygen availability. Wetlands 28: 817-826.


294









Graetz DA & Nair VD (1995) Fate of phosphorus in Florida Spodosols contaminated
with cattle manure. Ecol. Eng. 5: 163-181.

Grant WD (1979) Cell-wall Teichoic- acid as a reserve phosphate source in Bacillus-
subtilis J. Bacteriol. 137: 35-43.

Gressel N, McColl JG, Preston CM, Newman RH & Powers RF (1996) Linkages
between phosphorus transformations and carbon decomposition in a forest soil.
Biogeochemistry 33: 97-123.

Gunther S, Trutnau M, Kleinsteuber S, Hause G, Bley T, Roske I, Harms H & Muller S
(2009) Dynamics of polyphosphate-accumulating bacteria in wastewater treatment
plant microbial communities detected via DAPI (4 ',6 '-diamidino-2-phenylindole)
and tetracycline labeling. Appl. Environ. Microbiol. 75: 2111-2121.

Hackney CT, Padgett DE & Posey MH (2000) Fungal and bacterial contributions to the
decomposition of Cladium and Typha leaves in nutrient enriched and nutrient poor
areas of the Everglades, with a note on ergosterol concentrations in Everglades
soils. Mycol. Res. 104: 666-670.

Hagerthey SE, Newman S, Rutchey K, Smith EP & Godin J (2008) Multiple regime
shifts in a subtropical peatland: community specific threshold to europhication
Ecol. Monogr. 78: 547-565.

Halstead RL, Anderson G & Scott NM (1966) Extraction of organic matter from soils by
means of ultrasonic dispersion in aqueous acetylacetone. Nature 211: 1430-1431.

Hanrahan G, Salmassi TM, Khachikian CS & Foster KL (2005) Reduced inorganic
phosphorus in the natural environment: Significance, speciation and
determination. Talanta 66: 435-444.

Harold FM (1966) Inorganic polyphophates in biology- structure metabolism and
function. Bacteriol. Rev. 30: 772-794.

Harris WG, Rhue RD, Kidder G, Brown RB & Littell R (1996) Phosphorus retention as
related to morphology of sandy coastal plain soil materials. Soil Sci. Soc. Am. J.
60: 1513-1521.

Harris WG, Wang HD & Reddy KR (1994) Dairy manure influence on soil and sediment
composition implications for phosphorus rentention. J. Environ. Qual. 23: 1071-
1081.

Harris WG & White GN (2008) X-ray diffraction analysis of soils. In: Ulery A & Drees R
(Eds) Methods of soil analysis: part 5 mineralogical methods, pp 81-115. Soil
Science Society of America, Madison USA

Harrison AF (1987) Soil organic phosphorus. A review of world literature. CAB
International, Wallingford, Oxon. UK


295









He ZQ, Fortuna AM, Senwo ZN, Tazisong IA, Honeycutt CW & Griffin TS (2006)
Hydrochloric fractions in Hedley fractionation may contain inorganic and organic
phosphates. Soil Sci. Soc. Am. J. 70: 893-899.

He ZQ, Mao JD, Honeycutt CW, Ohno T, Hunt JF & Cade-Menun BJ (2009)
Characterization of plant-derived water extractable organic matter by multiple
spectroscopic techniques. Biol. Fertil. Soils 45: 609-616.

Heath RT (2005) Microbial turnover of organic phosphorus in aquatic systems.In: Turner
BL, Frossard E & Baldwin DS (Eds) Organic phosphorus in the environment, pp
185-203. CABI Publishing, Wallingford, UK

Heathwaite AL & Dils RM (2000) Characterising phosphorus loss in surface and
subsurface hydrological pathways. Sci. Total Environ. 251: 523-538.

Hedley MJ & Stewart JWB (1982) Method to measure microbial phosphate in soils. Soil
Biol. Biochem. 14: 377-385.

Hedley MJ, Stewart JWB & Chauhan BS (1982) Changes in inorganic and organic soil-
phosphorus fractions induced by cultivation practices and by laboratory
incubations. Soil Sci. Soc. Am. J. 46: 970-976.

Heighton L, Schmidt WF & Siefert RL (2008) Kinetic and equilibrium constants of phytic
acid and ferric and ferrous phytate derived from nuclear magnetic resonance
spectroscopy. J. Agric. Food Chem. 56: 9543-9547.

Herbes SE, Allen HE & Mancy KH (1975) Enzymatic characterization of soluble organic
phosphorus in lake water. Science 187: 432-434.

Hieltjes AHM & Lijklema L (1980) Fractionation of inorganic phosphates in calcareous
sediments. J. Environ. Qual. 9: 405-407.

Hill JE & Richardson AE (2007) Isolation and assessment of microorganisms that utilize
phytate. In: Inositol phosphates: linking agriculture and the environment, pp 61-77.
Cabi, Wallingford UK

Horiguchi M & Kandatsu M (1959) Isolation of 2-aminoethane posphonic acid from
rumen protozoa. Nature 184: 901-902.

Hosomi M, Okada M & Sudo R (1982) Release of Phosphorus from Lake Sediments.
Environment International 7: 93-98.

Huang QY, Liang W & Cai P (2005) Adsorption, desorption and activities of acid
phosphatase on various colloidal particles from an Ultisol. Colloids and Surfaces
B-Biointerfaces 45: 209-214.

Hupfer M & Gachter R (1995) Polyphosphate in lake-sediments 31P NMR spectroscopy
as a tool for its identification. Limnol. Oceanogr. 40: 610-617.


296









Hupfer M, Gloess S & Grossart HP (2007) Polyphosphate-accumulating
microorganisms in aquatic sediments. Aquat. Microb. Ecol. 47: 299-311.

Hupfer M, Gloss S, Schmieder P & Grossart HP (2008) Methods for detection and
quantification of polyphosphate and polyphosphate accumulating microorganisms
in aquatic sediments. Int. Rev. Hydrobiol. 93: 1-30.

Hupfer M & Lewandowski J (2008) Oxygen controls the phosphorus release from lake
sediments a long-lasting paradigm in limnology. Int. Rev. Hydrobiol. 93: 415-432.

Hupfer M, Rube B & Schmieder P (2004) Origin and diagenesis of polyphosphate in
lake sediments: A P-31-NMR study. Limnol. Oceanogr. 49: 1-10.

Ipsilantis I & Sylvia DM (2007) Abundance of fungi and bacteria in a nutrient-impacted
Florida wetland. Appl. Soil Ecol. 35: 272-280.

Irving GCJ & Cosgrove DJ (1981) The use of hypobromite oxidation to evaluate two
current methods for the estimation of inositol polyphosphates in alkaline extracts of
soils. Commun. Soil Sci. Plant Anal. 12: 495-509.

Ivanoff DB, Reddy KR & Robinson S (1998) Chemical fractionation of organic
phosphorus in selected Histosols. Soil Sci. 163: 36-45.

Jackson CR, Liew KC & Yule CM (2009) Structural and functional changes with depth in
microbial communities in a tropical Malaysian peat swamp forest. Microb. Ecol. 57:
402-412.

Jackson JF, Jones G & Linskens HF (1982) Phytic acid in pollen. Phytochemistry 21:
1255-1258.

Jackson JF & Linskens HF (1982) Conifer pollen contains phytate and could be a major
source of phytate phosphorus in forest soils. Australian Forest Research 12: 11-
18.

Jenkinson DS, Brookes PC & Powlson DS (2004) Measuring soil microbial biomass.
Soil Biol. Biochem. 36: 5-7.

Jensen JR, Rutchey K, Koch MS & Narumalani S (1995) Inland wetland change
detection in the Everglades Water Conservation Area 2A using a time-series of
normalized remotely-sensed data. Photogramm. Eng. Remote Sens. 61: 199-209.

Jin K, Cornelis WM, Gabriels D, Baert M, Wu HJ, Schiettecatte W, Cai DX, De Neve S,
Jin JY, Hartmann R & Hofman G (2009) Residue cover and rainfall intensity
effects on runoff soil organic carbon losses. Catena 78: 81-86.

Joergensen RG & Wichern F (2008) Quantitative assessment of the fungal contribution
to microbial tissue in soil. Soil Biol. Biochem. 40: 2977-2991.


297









John MK (1970) Colorimetric determination of phosphorus in soil and plant materials
with ascorbic acid. Soil Sci. 109: 214-220.

Johnston CA, Bubenzer GD, Lee GB, Madison FW & McHenry JR (1984a) Nutrient
trapping by sediment deposition in a seasonally flooded lakeside wetland. J.
Environ. Qual. 13: 283-290.

Johnston CA, Lee GB & Madison FW (1984b) The stratigraphy and composition of a
lakeside wetland. Soil Sci. Soc. Am. J. 48: 347-354.

Kadlec RH (1997) An autobiotic wetland phosphorus model. Ecol. Eng. 8: 145-172.

Kadlec RH & Mitsch WJ (2009) Special Issue: The Houghton Lake wetland treatment
project. Ecol. Eng. 35: 1285-1366.

Kaila A (1956) Phosphorus in virgin peat soils. Maataloust. Aikak. 28: 142-167.

Kaila A & Virtanen O (1955) Determination of organic phosphorus in samples of peat
soils. Maataloust. Aikak. 27: 104-115.

Khan FA & Ansari AA (2005) Eutrophication: An ecological vision. Bot. Rev. 71: 449-
482.

Khoshmanesh A, Hart BT, Duncan A & Beckett R (2002) Luxury uptake of phosphorus
by sediment bacteria. Water Res. 36: 774-778.

Klauth P, Pallerla SR, Vidaurre D, Ralfs C, Wendisch VF & Schoberth SM (2006)
Determination of soluble and granular inorganic polyphosphate in
Corynebacterium glutamicum. Appl. Microbiol. Biotechnol. 72: 1099-1106.

Knicker H & Nanny MA (1997) Nuclear magnetic resonance spectroscopy. In: Nanny
MA, Minear RA & Leenheer JA (Eds) Nuclear Magnetic Resonance Spectroscopy
in Environmental Chemistry, pp 3-15. Oxford University Press, Oxford

Koch MS & Reddy KR (1992) Distribution of soil and plant nutrients along a trophic
gradient in the Florida Everglades. Soil Sci. Soc. Am. J. 56: 1492-1499.

Kokfelt U, Struyf E & Randsalu L (2009) Diatoms in peat Dominant producers in a
changing environment? Soil Biol. Biochem. 41: 1764-1766.

Kolowith LC, Ingall ED & Benner R (2001) Composition and cycling of marine organic
phosphorus. Limnol. Oceanogr. 46: 309-320.

Kominkova D, Kuehn KA, Busing N, Steiner D & Gessner MO (2000) Microbial biomass,
growth, and respiration associated with submerged litter of Phragmites australis
decomposing in a littoral reed stand of a large lake. Aquat. Microb. Ecol. 22: 271-
282.


298









Kornberg A (1995) Inorganic polyphosphate toward making a forgotten polymer
unforgettable. J. Bacteriol. 177: 491-496.

Kornberg A, Rao NN & Ault-Riche D (1999) Inorganic polyphosphate: A molecule of
many functions. Annu. Rev. Biochem. 68: 89-125.

Koukol O, Novak F & Hrabal R (2008) Composition of the organic phosphorus fraction
in basidiocarps of saprotrophic and mycorrhizal fungi. Soil Biol. Biochem. 40:
2464-2467.

Koukol O, Novak F, Hrabal R & Vosatka M (2006) Saprotrophic fungi transform organic
phosphorus from spruce needle litter. Soil Biol. Biochem. 38: 3372-3379.

Kouno K, Tuchiya Y & Ando T (1995) Measurement of soil microbial biomass
phosphorus by an anion-exchange membrane method. Soil Biol. Biochem. 27:
1353-1357.

Kowalenko CG & Babuin D (2007) Interference problems with
phosphoantimonylmolybdenum colorimetric measurement of phosphorus in soil
and plant materials. Commun. Soil Sci. Plant Anal. 38: 1299-1316.

Krupyanko VI, Vagabov VM, Trilisenko LV, Krupyanko PV, Shchipanova IN, Sibel'dina
LA & Kulaev IS (1998) A hypochromic effect of signal attenuation in the P-31-NMR
spectra of linear polyphosphates. Appl. Biochem. Microbiol. 34: 392-395.

Kuehn KA, Gessner MO, Wetzel RG & Suberkropp K (1999) Decomposition and CO2
evolution from standing litter of the emergent macrophyte Erianthus giganteus.
Microb. Ecol. 38: 50-57.

Kuehn KA, Lemke MJ, Suberkropp K & Wetzel RG (2000) Microbial biomass and
production associated with decaying leaf litter of the emergent macrophyte Juncus
effusus. Limnol. Oceanogr. 45: 862-870.

Kuhn NL, Mendelssohn IA, McKee KL, Lorenzen B, Brix H & Miao SL (2002) Root
phosphatase activity in Cladiumjamaicense and Typha domingensis grown in
everglades soil at ambient and elevated phosphorus levels. Wetlands 22: 794-800.

Kulaev I & Kulakovskaya T (2000) Polyphosphate and phosphate pump. Annu. Rev.
Microbiol. 54: 709-734.

Kuo S & Sparks DL (2001) Phosphorus. In: Part 3, Chemical Methods, p 869. Soil
Science Society of America, Madison, Wisconsin

L'Annunziata MF (2007) Origins and biochemical transformations of inositol
stereoisomers and their phosphorylated derivatives in soil.In: Inositol phosphates:
linking agriculture and the environment, pp 41-60. Cabi, Wallingford UK


299









Lahteenoja O, Ruokolainen K, Schulman L & Alvarez J (2009a) Amazonian floodplains
harbour minerotrophic and ombrotrophic peatlands. Catena 79: 140-145.

Lahteenoja O, Ruokolainen K, Schulman L & Oinonen M (2009b) Amazonian peatlands:
an ignored C sink and potential source. Glob. Change Biol. 15: 2311-2320.

Leake JR & Miles W(1996) Phosphodiesters as mycorrhizal P sources .1.
Phosphodiesterase production and the utilization of DNA as a phosphorus source
by the ericoid mycorrhizal fungus Hymenoscyphus ericae. New Phytologist 132:
435-443.

Legg JO & Black CA (1955) Determination of organic phosphorus in soils: n. Ignition
method. Soil Sci Soc Amer Proc 19: 139-143.

Lei XG & Porres JM (2007) Phytase and inositol phosphates in animal nutrition: dietary
manipulation and phosphorus excretion by animals.In: Inositol phosphates: linking
agriculture and the environment, pp 133-149. Cabi, Wallingford UK

Lewis DL, Liudahl KJ, Noble CV & Carter LJ (2003) Soil survey of Okeechobee County,
Florida. USDA-NRCS,

Leytem AB & Maguire RO (2007) Environmental implications of inositol phosphates in
animal manures.In: Inositol phosphates: linking agriculture and the environment,
pp 150-168. Cabi, Wallingford UK

Limpens J, Berendse F, Blodau C, Canadell JG, Freeman C, Holden J, Roulet N, Rydin
H & Schaepman-Strub G (2008) Peatlands and the carbon cycle: from local
processes to global implications a synthesis Biogeosci. 5: 1739-1739.

Lissitskaya TB, Shmeleva VG, Vardoian GS & Yakovlev VI (1999) Screening of
microorganisms producing phytase. Mikol. Fitopatol. 33: 402-405.

Liu JY, Wang H, Yang HJ, Ma YJ & Cai OC (2009) Detection of phosphorus species in
sediments of artificial landscape lakes in China by fractionation and phosphorus-
31 nuclear magnetic resonance spectroscopy. Environ. Pollut. 157: 49-56.

Llewelyn JM, Landing WM, Marshall AG & Cooper WT (2002) Electrospray ionization
Fourier transform ion cyclotron resonance mass spectrometry of dissolved organic
phosphorus species in a treatment wetland after selective isolation and
concentration. Anal. Chem. 74: 600-606.

Longhi D, Bartoli M & Viaroli P (2008) Decomposition of four macrophytes in wetland
sediments: Organic matter and nutrient decay and associated benthic processes.
Aquat. Bot. 89: 303-310.

Lopez-Buendia AM, Whateley MKG, Bastida J & Urquiola MM (2007) Origins of mineral
matter in peat marsh and peat bog deposits, Spain. Int. J. Coal Geol. 71: 246-262.


300









Lott JNA, Ockenden I, Raboy V & Batten GD (2000) Phytic acid and phosphorus in crop
seeds and fruits: a global estimate. Seed Sci. Res. 10: 11-33.

Lung SC, Leung A, Kuang R, Wang Y, Leung P & Lim BL (2008) Phytase activity in
tobacco (Nicotiana tabacum) root exudates is exhibited by a purple acid
phosphatase. Phytochemistry 69: 365-373.

Macek P & Rejmankova E (2007) Response of emergent macrophytes to experimental
nutrient and salinity additions. Funct. Ecol. 21: 478-488.

Magid J, Tiessen H & Condron LM (1996) Dynamics of organic phosphorus in soils
under natural and agricultural ecosystems.In: Piccolo A (Ed) Humic substances in
terestrial ecosystems, pp 429-466. Elsevier Sience B.V., Amsetrdam

Makarov MI, Haumaier L & Zech W (2002a) The nature and origins of diester
phosphates in soils: A 31P NMR study. Biol. Fertil. Soils 35: 136-146.

Makarov MI, Haumaier L & Zech W (2002b) Nature of soil organic phosphorus: an
assessment of peak assignments in the diester region of P-31 NMR spectra. Soil
Biol. Biochem. 34: 1467-1477.

Makarov MI, Haumaier L, Zech W, Marfenina OE & Lysak LV (2005) Can 31P NMR
spectroscopy be used to indicate the origins of soil organic phosphates? Soil Biol.
Biochem. 37: 15-25.

Martinez A, Tyson GW & DeLong EF (2010) Widespread known and novel phosphonate
utilization pathways in marine bacteria revealed by functional screening and
metagenomic analyses. Environ. Microbiol. 12: 222-238.

Marx MC, Wood M & Jarvis SC (2001) A microplate fluorimetric assay for the study of
enzyme diversity in soils. Soil Biol. Biochem. 33: 1633-1640.

Mason S, Harnon R, Zhang H & Anderson J (2008) Investigating chemical constraints to
the measurement of phosphorus in soils using diffusive gradients in thin films
(DGT) and resin methods. Talanta 74: 779-787.

Mayer T, Ptacek C & Zanini L (1999) Sediments as a source of nutrients to
hypereutrophic marshes of Point Pelee, Ontario, Canada. Water Res. 33: 1460-
1470.

McDowell RW (2009) Effect of land use and moisture on phosphorus forms in upland
stream beds in South Otago, New Zealand. Mar. Freshw. Res. 60: 619-625.

McDowell RW, Condron LM, Mahieu N, Brookes PC, Poulton PR & Sharpley AN (2002)
Analysis of potentially mobile phosphorus in arable soils using solid state nuclear
magnetic resonance. J. Environ. Qual. 31: 450-456.


301









McDowell RW, Condron LM & Stewart I (2008) An examination of potential extraction
methods to assess plant-available organic phosphorus in soil. Biol. Fertil. Soils 44:
707-715.

McDowell RW, Nash DM & Robertson F (2007) Sources of phosphorus lost from a
grazed pasture receiving simulated rainfall. J. Environ. Qual. 36: 1281-1288.

McDowell RW & Stewart I (2005a) An improved technique for the determination of
organic phosphorus in sediments and soils by P-31 nuclear magnetic resonance
spectroscopy. Chem. Ecol. 21: 11-22.

McDowell RW & Stewart I (2005b) Peak assignments for phosphorus-31 nuclear
magnetic resonance spectroscopy in pH range 5-13 and their application in
environmental samples. Chem. Ecol. 21: 211-226.

McDowell RW & Stewart I (2006) The phosphorus composition of contrasting soils in
pastoral, native and forest management in Otago, New Zealand: sequential
extraction and 31P NMR. Geoderma 130: 176-189.

McDowell RW, Stewart I & Cade-Menun BJ (2006) An examination of spin-lattice
relaxation times for analysis of soil and manure extracts by liquid state
phosphorus-31 nuclear magnetic resonance spectroscopy. J. Environ. Qual. 35:
293-302.

McGill WB & Cole CV (1981) Comparative aspects of cycling of organic C,N,S and P
through soil organic matter. Geoderma 26: 267-286.

McKee KL (2005) Predicting soil phosphorus storage in historically isolated wetlands
within the Lake Okeechobee priority basins. Soil and Water Science, University of
Florida (p 241). Florida, Gainesville

McKelvie ID (2005) Separation, preconcentration and speciation of organic phosphorus
in environmental samples. In: Turner BL, Frossard E & Baldwin DS (Eds) Organic
phosphorus in the environment, pp 1-20. CABI Publishing, Wallingford UK

McKelvie ID (2007) Inositol phosphates in aquatic systems.In: Turner BL, Richardson
AE & Mullaney EJ (Eds) Inositol phosphates: linking agriculture and the
environment, pp 261-277. CABI, Wallingford UK

McKelvie ID, Hart BT, Cardwell TJ & Cattrall RW (1989) Spectrophotometric
determination of dissolved organic phosphorus in natural- waters using in-line
photo-oxidation and flow-injection. Analyst 114: 1459-1463.

McLaughlin MJ, Lancaster PA, Sale PG, Uren NC & Peverill KI (1994) Comparison of
cation/anion exchange resin methods for multi-element testing of acidic soils. Aust.
J. Soil Res. 32: 229-240.


302









Meason DF & Idol TW (2008) Nutrient sorption dynamics of resin membranes and resin
bags in a tropical forest. Soil Sci. Soc. Am. J. 72: 1806-1814.

Mehta NC, Legg JO, Goring CAI & Black CA (1954) Determination of organic
phosphorus in soils. Soil Sci Soc Amer Proc 18: 443-449.

Menzel DW & Corwin N (1965) The measurement of total phosphorus in seawater
based on the liberation of organically bound fractions by persulfate oxidation.
Limnol Oceanogr 10: 280-282.

Michell RH (2008) Inositol derivatives: Evolution and functions. Nat. Rev. Mol. Cell Biol.
9: 151-161.

Miltner A, Haumaier L & Zech W (1998) Transformations of phosphorus during
incubation of beech leaf litter in the presence of oxides. Eur. J. Soil Sci. 49: 471-
475.

Minear RA (1972) Characterization of naturally occurring dissolved organophosphorus
compounds. Environ. Sci. Technol. 6: 431.

Minear RA, Segars JE, Elwood JW & Mulholland PJ (1988) Seperation of inositol
phosphates by high-performance ion-exchange chromatography Analyst 113: 645-
649.

Mitchell AM & Baldwin DS (2005) Organic phosphorus in the aquatic environment:
speciation, transformations and interactions with nutrient cycles.In: Turner BL,
Frossard E & Baldwin DS (Eds) Organic phosphorus in the environment, pp 309-
323. CABI Publishing, Wallingford UK

Mitsch WJ & Day JW (2006) Restoration of wetlands in the Mississippi-Ohio-Missouri
(MOM) river basin: Experience and needed research. Ecol. Eng. 26: 55-69.

Monaghan EJ & Ruttenberg KC (1999) Dissolved organic phosphorus in the coastal
ocean: Reassessment of available methods and seasonal phosphorus profiles
from the Eel River Shelf. Limnol. Oceanogr. 44: 1702-1714.

Moreno D, Pedrocchi C, Comin FA, Garcia M & Cabezas A (2007) Creating wetlands
for the improvement of water quality and landscape restoration in semi-arid zones
degraded by intensive agricultural use. Ecol. Eng. 30: 103-111.

Morton SC & Edwards M (2005) Reduced phosphorus compounds in the environment.
Crit. Rev. Environ. Sci. Technol. 35: 333-364.

Moustafa MZ, Newman S, Fontaine TD, Chimney MJ & Kosier TC (1999) Phosphorus
retention by the Everglades Nutrient Removal: An Everglades stormwater
treatment area.In: Reddy KR, O'Connor GA & Schlichting A (Eds) Phosphorus
Biogeochemistry of Sub-Tropical Ecosystems, pp 489-509. CRC Press LLC,


303









Mullaney EJ & Ullah AHJ (2007) Phytases: attributes, catalytic mechanisms and
applications. In: Inositol phosphates: linking agriculture and the environment, pp
97-110. Cabi, Wallingford UK

Murphy J & Riley JP (1962) A modified single solution method for determination of
phosphate in natural waters. Analytica Chimica Acta 26: 31-36.

Murphy PNC, Bell A & Turner BL (2009) Phosphorus speciation in temperate basaltic
grassland soils by solution 31P NMR spectroscopy. Eur. J. Soil Sci. 60: 638-651.

Murthy PPN (2007) Identification of inositol phosphates by nuclear magnetic resonance
spectroscopy: unravelling structural diversity.In: Inositol phosphates: linking
agriculture and the environment, pp 7-22. CABI, Wallingford UK

Myers RG, Sharpley AN, Thien SJ & Pierzynski GM (2005) Ion-sink phosphorus
extraction methods applied on 24 soils from the continental USA. Soil Sci. Soc.
Am. J. 69: 511-521.

Myers RG, Thien SJ & Pierzynski GM (1999) Using an ion sink to extract microbial
phosphorus from soil. Soil Sci. Soc. Am. J. 63: 1229-1237.

Nanny MA, Kim S & Minear RA (1995) Aquatic soluble unreactive phosphorus HPLC
studies on concentrated water samples. Water Res. 29: 2138-2148.

Nanny MA & Minear RA (1994a) Organic phosphorus in the hydrosphere -
Characterization via 31 P fourier transform nuclear magnetic resonance
spectroscopy In: Baker LA (Ed) Environmental Chemistry of Lakes and
Reservoirs, pp 161-191. Amer Chemical Soc, Washington

Nanny MA & Minear RA (1994b) Use of lanthanide shift-reagents with P-31 FT-NMR
spectroscopy to analyze concentrated lake-water samples. Environ. Sci. Technol.
28: 1521-1527.

Nanny MA & Minear RA (1997) Characterization of soluble unreactive phosphorus using
P-31 nuclear magnetic resonance spectroscopy. Mar. Geol. 139: 77-94.

Nasholm T, Ekblad A, Nordin A, Giesler R, Hogberg M & Hogberg P (1998) Boreal
forest plants take up organic nitrogen. Nature 392: 914-916.

Newman RH & Tate KR (1980) Soil-phosphorus characterization by P-31 nuclear
magnetic-resonance. Commun. Soil Sci. Plant Anal. 11: 835-842.

Newman S, McCormick PV & Backus JG (2003) Phosphatase activity as an early
warning indicator of wetland eutrophication: problems and prospects. J. Appl.
Phycol. 15: 45-59.


304









Newman S & Robinson JS (1999) Forms of organic phosphorus in water, soils, and
sediments.In: Reddy KR, O'Connor GA & Schelske CL (Eds) Phosphorus
biogeochemistry of subtropical ecosystems, pp 207-223. CRC Press LLC, Boca
Raton, Florida

Nguyen LM (2000) Phosphate incorporation and transformation in surface sediments of
a sewage-impacted wetland as influenced by sediment sites, sediment pH and
added phosphate concentration. Ecol. Eng. 14: 139-155.

Niemeyer J & Gessler F (2002) Determination of free DNA in soils. J. Plant Nutr. Soil
Sci.-Z. Pflanzenernahr. Bodenkd. 165: 121-124.

Oberson A & Joner EJ (2005) Microbial turnover of phosphorus in soil. In: Turner BL,
Frossard E & Baldwin DS (Eds) Organic phosphorus in the environment, pp 133-
164. CABI Publishing, Wallingford UK

Ogram A, Sayler GS, Gustin D & Lewis RJ (1988) DNA Adsorption to Soils and
Sediments. Environ. Sci. Technol. 22: 982-984.

Olsen SR & Sommers LE (1982) Phosphorus.In: Page AL, Miller RH & Kenney DR
(Eds) Methods of soil analysis. Part 2: Cemical and microbial properties, pp 403-
427. American Society of Agronomy, Madison, WI

Page SE, Rieley JO, Shotyk OW & Weiss D (1999) Interdependence of peat and
vegetation in a tropical peat swamp forest. Philos. Trans. R. Soc. Lond. Ser. B-
Biol. Sci. 354: 1885-1897.

Paludan C, Alexeyev FE, Drews H, Fleischer S, Fuglsang A, Kindt T, Kowalski P, Moos
M, Radlowki A, Stromfors G, Westberg V & Wolter K (2002) Wetland management
to reduce Baltic sea eutrophication. Water Sci. Technol. 45: 87-94.

Paludan C & Jensen HS (1995) Sequential extraction of phosphorus in freshwater
wetland and lake sediment: Significance of humic acids. Wetlands 15: 365-373.

Pandey A, Szakacs G, Soccol CR, Rodriguez-Leon JA & Soccol VT (2001) Production,
purification and properties of microbial phytases. Bioresource Technology 77: 203-
214.

Pandey V, Kiker GA, Campbell KL, Williams MJ & Coleman SW (2009) GPS monitoring
of cattle location near water features in south Florida. Appl. Eng. Agric. 25: 551-
562.

Pant HK & Reddy KR (2001) Hydrologic influence on stability of organic phosphorus in
wetland detritus. J. Environ. Qual. 30: 668-674.

Pant HK, Reddy KR & Dierberg FE (2002) Bioavailability of organic phosphorus in a
submerged aquatic vegetation-dominated treatment wetland. J. Environ. Qual. 31:
1748-1756.


305









Pant HK & Warman PR (2000) Enzymatic hydrolysis of soil organic phosphorus by
immobilized phosphatases. Biol. Fertil. Soils 30: 306-311.

Parkinson JA & Allen SE (1975) Wet oxidation procedure suitable for determination of
nitrogen and mineral nutrients in biological-material. Commun. Soil Sci. Plant Anal.
6: 1-11.

Paytan A, Cade-Menun BJ, McLaughlin K & Faul KL (2003) Selective phosphorus
regeneration of sinking marine particles: evidence from 31P NMR. Mar. Chem. 82:
55-70.

Pearson RW (1940) Determination of organic phosphorus in soils. Indust. and Engineer.
Chem. Analyt. Ed. 12: 198-200.

Pech H, Henry A, Khachikian CS, Salmassi TM, Hanrahan G & Foster KL (2009)
Detection of Geothermal Phosphite Using High-Performance Liquid
Chromatography. Environ. Sci. Technol. 43: 7671-7675.

Penton CR & Newman S (2008) Enzyme-based resource allocated decomposition and
landscape heterogeneity in the Florida Everglades. J. Environ. Qual. 37: 972-976.

Pepper IL, Miller RH & Ghonsikar CP (1976) Microbial inorganic polyphosphates -
factors influencing their accumulation in soil. Soil Sci. Soc. Am. J. 40: 872-875.

Perkins DB, Olsen AE & Jawitz JW (2005) Spatially distributed isolated wetlands for
watershed-scale treatment of agricultural runoff.In: Dunne EJ, Reddy KR & Carton
OT (Eds) Nutrient management in agricultural watersheds: a wetlands solution, pp
80-91. Wageningen Academic Publishers, Wageningen, Netherlands

Phillips S & Bustin RM (1996) Sedimentology of the Changuinola peat deposit: Organic
and plastic sedimentary response to punctuated coastal subsidence. Geol. Soc.
Am. Bull. 108: 794-814.

Phillips S, Rouse GE & Bustin RM (1997) Vegetation zones and diagnostic pollen
profiles of a coastal peat swamp, Bocas del Toro, Panama. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 128: 301-338.

Phillips VD (1998) Peatswamp ecology and sustainable development in Borneo.
Biodivers. Conserv. 7: 651-671.

Potter RS & Benton TH (1916) The organic phosphorus of soil. Soil Sci. 2: 291-298.

Potter RS & Snyder RS (1916) Soluble nonprotein nitrogen of soil. J. Agric. Res. 6: 61-
64.

Psenner R & Pucsko R (1988) Phosphorus fractionation advantages and limits of the
method for the study of sediment P origins and interactions. Ergebnisse der
Limnologie 30: 43-60.


306









Qian P & Schoenau JJ (1997) Recent developments in use of ion exchange
membranes in agricultural and environmental research. Recent Research
Developments in Soil Science 1: 43-54.

Qian P & Schoenau JJ (2002) Practical applications of ion exchange resins in
agricultural and environmental soil research. Can. J. Soil Sci. 82: 9-21.

Qian P, Schoenau JJ & Huang WZ (1992) Use of ion-exchange membranes in routine
soil testing. Commun. Soil Sci. Plant Anal. 23: 1791-1804.

Qiu S & McComb A (2000) Properties of sediment phosphorus in seven wetlands of the
Swan Coastal Plain, South-Western Australia. Wetlands 20: 267-279.

Quails RG & Richardson CJ (1995) Forms of soil-phosphorus along a nutrient
enrichment gradient in the northern Everglades. Soil Sci. 160: 183-198.

Quails RG & Richardson CJ (2000) Phosphorus enrichment affects litter decomposition,
immobilization, and soil microbial phosphorus in wetland mesocosms. Soil Sci.
Soc. Am. J. 64: 799-808.

Quiquampoix H & Mousain D (2005) Enzymatic hydrolysis of organic phosphorus.In:
Turner BL, Frossard E & Baldwin DS (Eds) Organic phosphorus in the
environment, pp 89-112. CABI Publishing, Wallingford UK

Raboy V (2007) Seed phosphorus and the development of low-phytate crops. In: Turner
BL, Richardson AE & Mullaney EJ (Eds) Inositol phosphates: linking agriculture
and the environment, pp 111-132. CABI, Wallingford UK

Raghothama KG (1999) Phosphate acquisition. Annual Review of Plant Physiology and
Plant Molecular Biology 50: 665-693.

Raghothama KG & Karthikeyan AS (2005) Phosphate acquisition. Plant Soil 274: 37-49.

Ramsar Convention Secretariat (2007). Designating Ramsar sites: The Strategic
Framework and guidelines for the future development of the List of Wetlands of
International Importance. Ramsar handbooks for the wise use of wetlands, 3rd
edition, vol. 14. Ramsar Convention Secretariat, Gland, Switzerland.

Rashchi F & Finch JA (2000) Polyphosphates: A review their chemistry and application
with particular reference to mineral processing. Miner. Eng. 13: 1019-1035.

Reddy KR (1983) Soluble Phosphorus Release from Organic Soils. Agric. Ecosyst.
Environ. 9: 373-382.

Reddy KR, Diaz OA, Scinto LJ & Agami M (1995) Phosphorus dynamics in selected
wetlands and streams of the lake Okeechobee Basin. Ecol. Eng. 5: 183-207.


307









Reddy KR, Flaig EG & Graetz DA (1996) Phosphorus storage capacity of uplands,
wetlands and streams of the Lake Okeechobee watershed, Florida. Agric. Ecosyst.
Environ. 59: 203-216.

Reddy KR, Kadlec RH, Flaig E & Gale PM (1999) Phosphorus retention in streams and
wetlands: A review. Crit. Rev. Environ. Sci. Technol. 29: 83.

Reddy KR & Patrick WH (1993) Wetland soils opportunities and challenges. Soil Sci.
Soc. Am. J. 57: 1145-1147.

Reddy KR, Wang Y, Debusk WF, Fisher MM & Newman S (1998) Forms of soil
phosphorus in selected hydrologic units of the Florida Everglades. Soil Sci. Soc.
Am. J. 62: 1134-1147.

Reddy KR, Wetzel RG & Kadlec RH (2005) Biogeochemistry of phosphorus in
wetlands.In: Sims JT & Sharpley AN (Eds) Phosphorus: agriculture and the
environment, pp 263-316. American Society of Agronomy, Madison USA

Reichert P & Wehrli B (2007) Modelling organic phosphorus transformations in aquatic
systems.In: Turner BL, Richardson AE & Mullaney EJ (Eds) Inositol phosphates:
linking agriculture and the environment, pp 349-376. CABI Publishing, Wallingford
UK

Reitzel K, Ahlgren J, DeBrabandere H, Waldeback M, Gogoll A, Tranvik L & Rydin E
(2007) Degradation rates of organic phosphorus in lake sediment.
Biogeochemistry 82: 15-28.

Reitzel K, Ahlgren J, Gogoll A, Jensen HS & Rydin E (2006a) Characterization of
phosphorus in sequential extracts from lake sediments using P-31 nuclear
magnetic resonance spectroscopy. Can. J. Fish. Aquat. Sci. 63: 1686-1699.

Reitzel K, Ahlgren J, Gogoll A & Rydin E (2006b) Effects of aluminum treatment on
phosphorus, carbon, and nitrogen distribution in lake sediment: A 31P NMR study.
Water Res. 40: 647-654.

Reitzel K, Jensen HS, Flindt M & Andersen FO (2009) Identification of dissolved
nonreactive phosphorus in freshwater by precipitation with aluminum and
subsequent P-31 NMR Analysis. Environ. Sci. Technol. 43: 5391-5397.

Rejmankova E & Houdkova K (2006) Wetland plant decomposition under different
nutrient conditions: what is more important, litter quality or site quality?
Biogeochemistry 80: 245-262.

Rejmankova E & Sirova D (2007) Wetland macrophyte decomposition under different
nutrient conditions: relationships between decomposition rate, enzyme activities
and microbial biomass. Soil Biol. Biochem. 39: 526-538.


308









Richardson AE, Barea JM, McNeill AM & Prigent-Combaret C (2009) Acquisition of
phosphorus and nitrogen in the rhizosphere and plant growth promotion by
microorganisms. Plant Soil 321: 305-339.

Richardson AE, George TS, Hens M & Simpson RJ (2005) Utilization of soil organic
phosphorus by higher plants.In: Turner BL, Frossard E & Baldwin DS (Eds)
Organic phosphorus in the environment, p 165. CABI Publishing,

Richardson AE & Hadobas PA (1997) Soil isolates of Pseudomonas spp. that utilize
inositol phosphates. Can. J. Microbiol. 43: 509-516.

Richardson CJ (1999) The role of wetlands in storage, release and cycling of
phosphorus on the landscape: a 25-year retrospective.In: Reddy KR, O'Connor
GA & Schelske CL (Eds) Phosphorus biogeochemistry of subtropical ecosystems,
pp 47-68. CRC Press LLC, Boca Raton, Florida

Richardson CJ & Marshall PE (1986) Processes controlling movement, storage, and
export of phosphorus in a fen peatland. Ecol. Monogr. 56: 279-302.

Robinson JS, Johnston CT & Reddy KR (1998) Combined chemical and P-31-NMR
spectroscopic analysis of phosphorus in wetland organic soils. Soil Sci. 163: 705-
713.

Roboredo M & Coutinho J (2006) Chemical characterization of inorganic phosphorus
desorbed by ion exchange membranes in short- and long-term extraction periods.
Commun. Soil Sci. Plant Anal. 37: 1611-1626.

Rowland AP & Haygarth PM (1997) Determination of total dissolved phosphorus in soil
solutions. J. Environ. Qual. 26: 410-415.

Rubaek GH & Sibbesen E (1993) Resin extraction of labile, soil organic phosphorus. J.
Soil Sci. 44: 467-478.

Ruttenberg KC (1992) Development of a sequential extraction method for different
forms of phosphorus in marine-sediments. Limnol. Oceanogr. 37: 1460-1482.

Sahrawat KL (2004) Organic matter accumultion in submerged soils.In: Sparks D (Ed)
Advances in Agronomy, pp 169-202. Newark, NY

Sannigrahi P & Ingall E (2005) Polyphosphates as a source of enhanced P fluxes in
marine sediments overlain by anoxic waters: evidence from 31P NMR. Geochem.
Trans. 6: 52-59.

Sannigrahi P, Ingall ED & Benner R (2006) Nature and dynamics of phosphorus-
containing components of marine dissolved and particulate organic matter.
Geochim. Cosmochim. Acta 70: 5868-5882.


309









Sato S & Comerford NB (2006) Assessing methods for developing phosphorus
desorption isotherms from soils using anion exchange membranes. Plant Soil 279:
107-117.

Saunders WMH (1964) Extraction of soil phosphate by anion-exchange membrane. N.
Z. J. Agric. Res. 7: 427-431.

Saunders WMH & Williams EG (1955) Observations on the determination of total
organic phosphorus in soils. J. Soil Sci. 6: 254-267.

Schachtman DP, Reid RJ & Ayling SM (1998) Phosphorus uptake by plants: from soil to
cell. Plant Physiology 116: 447-453.

Schelske CL, Conley DJ, Stoermer EF, Newberry TL & Campbell CD (1986) Biogenic
silica and phosphorus accumulation in sediments as indexes of eutrophication in
the Laurentian Great-Lakes. Hydrobiologia 143: 79-86.

Schlichting A & Leinweber P (2002) Effects of pretreatment on sequentially-extracted
phosphorus fractions from peat soils. Commun. Soil Sci. Plant Anal. 33: 1617-
1627.

Schlichting A, Leinweber P, Meissner R & Altermann M (2002) Sequentially extracted
phosphorus fractions in peat-derived soils. J. Plant Nutr. Soil Sci.-Z.
Pflanzenernahr. Bodenkd. 165: 290.

Schoenau JJ & Huang WZ (2001) Assessment of ion availability in heterogeneous
media using ion-exchange membranes. (p 62). University of Saskatchewan, US

Selig U, Hubener T & Michalik M (2002) Dissolved and particulate phosphorus forms in
a eutrophic shallow lake. Aquat. Sci. 64: 97-105.

Semeniuk CA & Semeniuk V (1995) A geomorphic approach to global classification for
inland wetlands. Vegetatio 118: 103-124.

Semeniuk V & Semeniuk CA (1997) A geomorphic approach to global classification for
natural inland wetlands and rationalization of the system used by the Ramsar
Convention a discussion. Wetlands Ecol. Manage. 5: 145-158.

Serafim LS, Lemos PC, Levantesi C, Tandoi V, Santos H & Reis MAM (2002) Methods
for detection and visualization of intracellular polymers stored by polyphosphate-
accumulating microorganisms. J. Microbiol. Methods 51: 1-18.

Seufferheld MJ, Alvarez HM & Farias ME (2008) Role of polyphosphates in microbial
adaptation to extreme environments. Appl. Environ. Microbiol. 74: 5867-5874.

Shan Y, McKelvie ID & Hart BT (1993) Characterization of immobilized escherichia-coli
alkaline-phosphatase reactors in flow-injection analysis. Anal. Chem. 65: 3053-
3060.


310









Shan Y, McKelvie ID & Hart BT (1994) Determination of alkaline phosphatase-
hydrolyzable phosphorus in natural-water by enzymatic flow-injection. Limnol.
Oceanogr. 39: 1993-2000.

Shand CA, Cheshire MV, Bedrock CN, Chapman PJ, Fraser AR & Chudek JA (1999)
Solid-phase P-31 NMR spectra of peat and mineral soils, humic acids and soil
solution components: influence of iron and manganese. Plant Soil 214: 153-163.

Sharpley AN, Kleinman P & McDowell R (2001) Innovative management of agricultural
phosphorus to protect soil and water resources. Commun. Soil Sci. Plant Anal. 32:
1071-1100.

Sheaves M (2009) Consequences of ecological connectivity: the coastal ecosystem
mosaic. Mar. Ecol. Prog. Ser. 391: 107-115.

Sibbesen E (1978) Investigation of anion-exchange resin method for soil phosphate
extraction. Plant Soil 50: 305-321.

Silvius MJ & Giesen W (1996) Towards integrated management of swamp forests: a
case study from Sumatra. Tropical lowland peatlands of Southeast Asia.
Proceedings of a workshop on integrated planning and management of tropical
lowland peatlands, Cisarua, Indonesia, 3-8 July 1992 pp 247-267). International
Union for Conservation of Nature and Natural Resources,

Sinsabaugh RL & Moorhead DL (1994) Resource allocation to extracellular enzyme
prodcution- A model for nitrogen and phosphorus control of litter decomposition
Soil Biol. Biochem. 26: 1305-1311.

Sjogersten S, Cheesman AW, Turner BL & Lopez O (2010) Biogeochemical processes
along a nutrient gradient in a tropical ombrotrophic peatland. Biogeochemistry (in
press):

Skogley EO & Dobermann A (1996) Synthetic ion-exchange resins: soil and
environmental studies. J. Environ. Qual. 25: 13-24.

Smernik RJ & Dougherty WJ (2007) Identification of phytate in phosphorus-31 nuclear
magnetic resonance spectra: the need for spiking. Soil Sci Soc Am J 71: 1045-
1050.

Smith VH (2007) Microbial diversity-productivity relationships in aquatic ecosystems.
FEMS Microbiol. Ecol. 62: 181-186.

Soil Survey Staff (2010) Keys to Soil Taxonomy, 11th ed. USDA-Natural Resources
Conservation Service, Washington, DC.

Solorzano L & Strickland JDH (1968) Polyphosphate in seawater. Limnol. Oceanogr.
13: 515-&.


311









Sommers LE, Harris RF, Williams JD, Armstron.De & Syers JK (1970) Determination of
total organic phosphorus in lake sediments. Limnol. Oceanogr. 15: 301-304.

SPSS for windows, Rel. 17.0.0. 2008 Chicago: SPSS Inc



Steinman AD & Rosen BH (2000) Lotic-lentic linkages associated with Lake
Okeechobee, Florida. J. N. Am. Benthol. Soc. 19: 733-741.

Steward JH & Tate ME (1971) Gel chromatography of soil organic phosphorus. Journal
of Chromatography 60: 75-82.

Stewart JWB & Tiessen H (1987) Dynamics of soil organic phosphorus.
Biogeochemistry 4: 41-60.

Street-Perrott FA & Barker PA (2008) Biogenic silica: a neglected component of the
coupled global continental biogeochemical cycles of carbon and silicon. Earth
Surf. Process. Landf. 33: 1436-1457.

Struyf E & Conley DJ (2009) Silica: an essential nutrient in wetland biogeochemistry.
Front. Ecol. Environ. 7: 88-94.

Sumann M, Amelung W, Haumaier L & Zech W (1998) Climatic effects on soil organic
phosphorus in the North American Great Plains identified by phosphorus-31
nuclear magnetic resonance. Soil Sci. Soc. Am. J. 62: 1580-1586.

Summers JE, Ratcliffe RG & Jackson MB (2000) Anoxia tolerance in the aquatic
monocot Potamogeton pectinatus: absence of oxygen stimulates elongation in
association with an unusually large Pasteur effect. J. Exp. Bot. 51: 1413-1422.

Sundareshwar PV, Morris JT, Koepfler EK & Fornwalt B (2003) Phosphorus limitation of
coastal ecosystem processes. Science 299: 563-565.

Sundareshwar PV, Morris JT, Pellechia PJ, Cohen HJ, Porter DE & Jones BC (2001)
Occurrence and ecological implications of pyrophosphate in estuaries. Limnol.
Oceanogr. 46: 1570-1577.

Sundareshwar PV, Richardson CJ, Gleason RA, Pellechia PJ & Honomichl S (2009)
Nature versus nurture: functional assessment of restoration effects on wetland
services using nuclear magnetic resonance spectroscopy. Geophys. Res. Lett. (p
5).

Suzumura M & Kamatani A (1995a) Mineralization of inositol hexaphosphate in aerobic
and anaerobic marine-sediments implications for the phosphorus cycle.
Geochim. Cosmochim. Acta 59: 1021-1026.


312









Suzumura M & Kamatani A (1995b) Origin and distribution of inositol hexaphosphate in
estuarine and coastal sediments. Limnol. Oceanogr. 40: 1254-1261.

Swoboda JG, Campbell J, Meredith TC & Walker S (2010) Wall teichoic acid function,
biosynthesis, and inhibition. ChemBioChem 11: 35-45.

Szava-Kovats R (2009) Re-analysis of the relationship between organic carbon and
loss-on-ignition in soil. Commun. Soil Sci. Plant Anal. 40: 2712-2724.

Tate KR & Newman RH (1982) Phosphorus fractions of a climosequence of soils in
New-Zealand tussock grassland. Soil Biol. Biochem. 14: 191-196.

Ternan NG, Mc Grath JW, Mc Mullan G & Quinn JP (1998) Organophosphonates:
occurrence, synthesis and biodegradation by microorganisms. World J. Microbiol.
Biotechnol. 14: 635-647.

Thien SJ & Myers R (1992) Determination of bioavailable phosphorus in soil. Soil Sci.
Soc. Am. J. 56: 814-818.

Thormann MN, Currah RS & Bayley SE (2003) Succession of microfungal assemblages
in decomposing peatland plants. Plant Soil 250: 323-333.

Tilman D, Fargione J, Wolff B, D'Antonio C, Dobson A, Howarth R, Schindler D,
Schlesinger WH, Simberloff D & Swackhamer D (2001) Forecasting agriculturally
driven global environmental change. Science 292: 281-284.

Troxler TG (2007) Patterns of phosphorus, nitrogen and delta N-15 along a peat
development gradient in a coastal mire, Panama. J. Trop. Ecol. 23: 683-691.

Turner BL (2004) Optimizing phosphorus characterization in animal manures by solution
phosphorus-31 nuclear magnetic resonance spectroscopy. J. Environ. Qual. 33:
757-766.

Turner BL (2005) Organic phosphorus transfer from terrestrial to aquatic
environments.In: Turner BL, Frossard E & Baldwin DS (Eds) Organic phosphorus
in the environment, pp 269-294. CABI Publishing, Wallingford, UK

Turner BL (2006) Organic phosphorus in Madagascan rice soils. Geoderma 136: 279-
288.

Turner BL (2007) Inositol phosphates in soil: amounts, forms and significance of the
phosphorylated inositol stereoisomers.ln: Inositol phosphates: linking agriculture
and the environment, pp 186-206. CABI Publishing, Wallingford UK

Turner BL (2008a) Resource partitioning for soil phosphorus: a hypothesis. J. Ecol. 96:
698-702.


313









Turner BL (2008b) Soil organic phosphorus in tropical forests: an assessment of the
NaOH-EDTA extraction procedure for quantitative analysis by solution P-31 NMR
spectroscopy. Eur. J. Soil Sci. 59: 453-466.

Turner BL, Baxter R, Mahieu N, Sjogersten S & Whitton BA (2004) Phosphorus
compounds in subarctic Fennoscandian soils at the mountain birch, (Betula
pubescens) tundra ecotone. Soil Biol. Biochem. 36: 815-823.

Turner BL, Cade-Menun BJ, Condron LM & Newman S (2005) Extraction of soil organic
phosphorus. Talanta 66: 294-306.

Turner BL, Cade-Menun BJ & Westermann DT (2003a) Organic phosphorus
composition and potential bioavailability in semi-arid arable soils of the western
United States. Soil Sci. Soc. Am. J. 67: 1168-1179.

Turner BL, Chudek JA, Whitton BA & Baxter R (2003b) Phosphorus composition of
upland soils polluted by long-term atmospheric nitrogen deposition.
Biogeochemistry 65: 259-274.

Turner BL, Condron LM, Richardson SJ, Peltzer DA & Allison VJ (2007a) Soil organic
phosphorus transformations during pedogenesis. Ecosystems 10: 1166-1181.

Turner BL, Driessen JP, Haygarth PM & McKelvie ID (2003c) Potential contribution of
lysed bacterial cells to phosphorus solubilisation in two rewetted Australian
pasture soils. Soil Biol. Biochem. 35: 187-189.

Turner BL & Haygarth PM (2005) Phosphatase activity in temperate pasture soils:
potential regulation of labile organic phosphorus turnover by phosphodiesterase
activity. Sci. Total Environ. 344: 27-36.

Turner BL, Mahieu N & Condron LM (2003d) Phosphorus-31 nuclear magnetic
resonance spectral assignments of phosphorus compounds in soil NaOH-EDTA
extracts. Soil Sci. Soc. Am. J. 67: 497-510.

Turner BL, Mahieu N & Condron LM (2003e) The phosphorus composition of temperate
pasture soils determined by NaOH-EDTA extraction and solution 31P NMR
spectroscopy. Org. Geochem. 34: 1199-1210.

Turner BL, Mahieu N & Condron LM (2003f) Quantification of myo-inositol
hexakisphosphate in alkaline soil extracts by solution 31P NMR spectroscopy and
spectral deconvolution. Soil Sci. 168: 469-478.

Turner BL, McKelvie ID & Haygarth PM (2002a) Characterisation of water-extractable
soil organic phosphorus by phosphatase hydrolysis. Soil Biol. Biochem. 34: 27-35.

Turner BL & Newman S (2005) Phosphorus cycling in wetland soils: the importance of
phosphate diesters. J. Environ. Qual. 34: 1921-1929.


314









Turner BL, Newman S, Cheesman AW & Reddy KR (2007b) Sample pretreatment and
phosphorus speciation in wetland soils. Soil Sci. Soc. Am. J. 71: 1538-1546.

Turner BL, Newman S & Newman JM (2006a) Organic phosphorus sequestration in
subtropical treatment wetlands. Environ. Sci. Technol. 40: 727-733.

Turner BL, Newman S & Reddy KR (2006b) Overestimation of organic phosphorus in
wetland soils by alkaline extraction and molybdate colorimetry. Environ. Sci.
Technol. 40: 3349-3354.

Turner BL, Paphazy MJ, Haygarth PM & McKelvie ID (2002b) Inositol phosphates in the
environment. Philos. Trans. R. Soc. Lond. Ser. B-Biol. Sci. 357: 449-469.

Turner BL & Richardson AE (2004) Identification of scyllo-inositol phosphates in soil by
solution phosphorus-31 nuclear magnetic resonance spectroscopy. Soil Sci. Soc.
Am. J. 68: 802-808.

Turner BL & Romero TE (2009a) Short-term changes in extractable inorganic nutrients
during storage of tropical rain forest soils. Soil Sci. Soc. Am. J. 73: 1972-1979.

Turner BL & Romero TE (2009b) Storage-induced changes in extractable inorganic
nutrients during transport and storage of tropical forest soils. Soil Sci. Soc. Am. J.
69: 630-633.

Turner BL & Weckstrom K (2009) Phytate as a novel phosphorus-specific paleo-
indicator in aquatic sediments. J. Paleolimnol. 42: 391-400.

Turner RE, Nancy NN, Justic D & Dortch Q (2003g) Future aquatic nutrient limitations.
Mar. Pollut. Bull. 46: 1032-1034.

Uhlmann D, Roske I, Hupfer M & Ohms G (1990) A simple method to distinguish
between polyphosphate and other phosphate fractions of activated sludge. Water
Res. 24: 1355-1360.

United States Department of Agriculture, Natural Resources Conservation Service.
2010. Field Indicators of Hydric Soils in the United States, Version 7.0. L.M.
Vasilas, G.W. Hurt, and C.V. Noble (eds.). USDA, NRCS, in cooperation with the
National Technical Committee for Hydric Soils

United States Enivrionmental Protection Agency (1993). Methods for chemical analysis
of water and wastes, Environmental Monitoring Support Laboratory Cincinnati, OH

Vaithiyanathan P & Richardson CJ (1997) Nutrient profiles in the Everglades:
examination along the eutrophication gradient. Sci. Total Environ. 205: 81-95.

Vaithiyanathan P & Richardson CJ (1999) Macrophyte species changes in the
Everglades: examination along a eutrophication gradient. J. Environ. Qual. 28:
1347-1358.


315









Van Eck GTM (1982) Forms of phosphorus in particulate matter from the Hollands Diep
Haringvliet the Netherlands. Hydrobiologia 91-92: 665-682.

Verhoeven JTA, Arheimer B, Yin C & Hefting MM (2006) Regional and global concerns
over wetlands and water quality. Trends Ecol. Evol. 21: 96-103.

Volfova O, Dvorakova J, Hanzlikova A & Jandera A (1994) Phytase from Aspergillus-
niger. Folia Microbiol. 39: 481-484.

Wal Rvd, Sjogersten S, Woodin SJ, Cooper EJ, Jonsdottir IS, Kuijper D, Fox TAD &
Huiskes AD (2007) Spring feeding by pink-footed geese reduces carbon stocks
and sink strength in tundra ecosystems. Glob. Change Biol. 13: 539-545.

Walker TW & Adams AFR (1958) Studies on soil organic matter: I. Influence of
phosphorus content of parent materials on accumulations of carbon, nitrogen,
sulfur and organic phosphorus in grassland soils. Soil Sci 85: 307-318.

Walker TW & Syers JK (1976) Fate of phosphorus during pedogenesis. Geoderma 15:
1-19.

Wallenstein MD & Weintraub MN (2008) Emerging tools for measuring and modeling
the in situ activity of soil extracellular enzymes. Soil Biol. Biochem. 40: 2098-2106.

Watts EE, Dean PAW & Martin RR (2002) P-31 nuclear magnetic resonance study of
sediment microbial phospholipids. Can. J. Anal. Sci. Spectrosc. 47: 127-133.

Webster JR & Benfield EF (1986) Vascular plant breakdown in fresh-water ecosystems.
Annu. Rev. Ecol. Syst. 17: 567-594.

Weimer WC & Armstrong DE (1979) Naturally occurring organic phosphorus-
compounds in aquatic plants. Environ. Sci. Technol. 13: 826-829.

Wen G, Voroney RP, Curtin D, Schoenau JJ, Qian PY & Inanaga S (2005) Modification
and application of a soil ATP determination method. Soil Biol. Biochem. 37: 1999-
2006.

Werner TP, Amrhein N & Freimoser FM (2007) Specific localization of inorganic
polyphosphate (poly P) in fungal cell walls by selective extraction and
immunohistochemistry. Fungal Genet. Biol. 44: 845-852.

Wetzel RG (1999) Organic phosphorus mineralization in soils and sediments.In: Reddy
KR, O'Connor GA & Schlichting A (Eds) Phosphorus biogeochemistry of sub-
tropical ecosystems, pp 225-245. CRC Press LLC,

Williams JD, Syers JK, Harris RF & Armstron.De (1970) Adsorption and desorption of
inorganic phosphorus by lake sediments in a 0.1 M NaCI system. Environ. Sci.
Technol. 4: 517-&.


316









Williams JD, Syers JK, Shukla SS & Harris RF (1971) Levels of inorganic and total
phosphorus in lake sediments as related to other sediment parameters. Environ.
Sci. Technol. 5: 1113-1120.

Williams JDH, Mayer T & Nriagu JO (1980) Extractability of phosphorus from phosphate
minerals common in soils and sediments. Soil Sci. Soc. Am. J. 44: 462-465.

Wilson MA (1987) NMR techniques and applications in geochemistry and soil chemistry.
Pergamon Press, Oxford UK

Worsfold PJ, Gimbert LJ, Mankasingh U, Omaka ON, Hanrahan G, Gardolinski PCFC,
Haygarth PM, Turner BL, Keith-Roach MJ & McKelvie ID (2005) Sampling, sample
treatment and quality assurance issues for the determination of phosphorus
species in natural waters and soils. Talanta 66: 273-293.

Worsfold PJ, Monbet P, Tappin AD, Fitzsimons MF, Stiles DA & McKelvie ID (2008)
Characterisation and quantification of organic phosphorus and organic nitrogen
components in aquatic systems: A review. Analytica Chimica Acta 624: 37-58.

Wright AL & Reddy KR (2001a) Heterotrophic microbial activity in northern Everglades
wetland soils. Soil Sci. Soc. Am. J. 65: 1856-1864.

Wright AL & Reddy KR (2001b) Phosphorus loading effects on extracellular enzyme
activity in Everglades wetland soils. Soil Sci. Soc. Am. J. 65: 588-595.

Wright AL & Reddy KR (2008) Catabolic diversity of periphyton and detritus microbial
communities in a subtropical wetland. Biogeochemistry 89: 199-207.

Wright AL, Wang Y & Reddy KR (2008) Loss-on-ignition method to assess soil organic
carbon in calcareous Everglades wetlands. Commun. Soil Sci. Plant Anal. 39:
3074-3083.

wxNUTS, vr 1.0.1 for Microsoft Windows (2007) Livermore, CA: Acorn NMR Inc

Yanke LJ, Bae HD, Selinger LB & Cheng KJ (1998) Phytase activity of anaerobic
ruminal bacteria. Microbiology-(UK) 144: 1565-1573.

Young MB, McLaughlin K, Kendall C, Stringfellow W, Rollog M, Elsbury K, Donald E &
Paytan A (2009) Characterizing the Oxygen Isotopic Composition of Phosphate
Sources to Aquatic Ecosystems. Environ. Sci. Technol. 43: 5190-5196.

Zhang J, James T & McCormick P (2009a) South Florida environmental report: Chapter
10: Lake Okeechobee Protection Program- state of the lake and watershed.
South Florida Environmental Report. SFWMD, FDEP,

Zhang RY, Wu FC, He ZQ, Zheng JA, Song BA & Jin LH (2009b) Phosphorus
composition in sediments from seven different trophic lakes, China: a phosphorus-
31 NMR study. J. Environ. Qual. 38: 353-359.


317









Zilles JL & Noguera DR (2002) Isolation and characterization of polyphosphate-
accumulating organisms from full-scale wastewater treatment plants. Abstracts of
the General Meeting of the American Society for Microbiology 102: 434.


318









BIOGRAPHICAL SKETCH

Alexander Cheesman was born on into the family Lawrence and Sheliagh

Cheesman. He was raised, and his family continues to live, near the village of Ardleigh,

on the outskirts of Colchester, Essex. Attending local schools until the age of 18, he

then had the privilege to attend Pembroke College, Cambridge; graduating in 2004 with

a BA in Plant Science. As part of his undergraduate training he had the opportunity to

work with Dr. Edmund Tanner. A collaboration which resulted in a position as a volunteer

field assistant on a long term project, GLIMP (Gigante Leaf Litter Manipulation Project)

with the Smithsonian Tropical Research Institute (STRI), Panama. While based at the

STRI field station he had the opportunity to interact with a range of researchers gaining

an appreciation of innovative research from a diverse collection of disciplines. During

this time, he was also introduced to Dr. Benjamin Turner, a relationship which resulted

in an offer of a PhD position working with Dr.K. Ramesh Reddy at the University of

Florida. The product of these working relationships form the nexus of this dissertation.


319





PAGE 1

BIOGENIC PHOSPHORUS IN PALUSTRINE WETLANDS: SOURCES AND STABILIZATION By ALEXANDER WILLIAM CHEESMAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORID A IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010 1

PAGE 2

2010 Alexander William Cheesman 2

PAGE 3

To those who make it possible 3

PAGE 4

4 ACKNOWLEDGMENTS To my advisers Dr. K. Ramesh Reddy and Dr. Benjamin Turner I would like to extend my deepest thanks for their guidance and patience throughout. Your example and tuition has helped mold the researcher I as pire to be. I thank the other members of my committee, Dr. Sue Newman, Dr. Nick Comerford and Dr. Mark Brenner for providing a sounding board for ideas, as well as a sense of purpose to my work. I thank the many people w ho contributed to the collection of wetland samples from otherwise inaccessible locations. In no particular order, I would like to thank Drs. Tom and Lynn Saunders, Dr. Jim Sickman, Dr. Ka thy Crowley, Dr. So fie Sjogersten, Dr. Robert Kadlec, Mr. Jason Vogel and Dr. Rebecca Sharitz. I extend my particular thanks to Dr. Diane De Steven who was invaluable in assisting in the collection of samples from South Carolina, as well as providi ng extensive ancillary data on Carolina Bay ecosystems. My thanks go to Dr. Patrick Inglett and Dr. Ed Dunne who provided access to archived samples for analysis as well as discussion and i deas when considering about the role of macrophytes and the nature of P cycling in wetlands. I am indebted to Ms. Yu Wang, Mr. Gavin Wilson, Mrs. Tania Romero and Dr. Alexander Blumenfeld for training and invaluable assistance during analytical l ab work. As well as to Dr. Jim Rocca whose patience and ready wit have allowed me to gain at least a passable understanding of 31P NMR analysis. Throughout my work I have received friends hip, advice and scientific discourse from people too many to mention. To those, I say thank you. Lastly, I would like to thank my family and my parents especially. Your love support and example have led me to become the man, and the sci entist I am today.

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDG MENTS .................................................................................................. 4LIST OF FI GURES ........................................................................................................ 13LIST OF ABBR EVIATION S ........................................................................................... 17ABSTRACT ................................................................................................................... 19 CHAPTER 1 BIOGENIC PHOSPHORUS IN WE TLANDS: AN IN TRODUCTIO N ....................... 22Phosphorus in Wetlands ......................................................................................... 22Wetland Soil ............................................................................................................ 24Wetland Phosphorus Cycle ..................................................................................... 26Biogenic Phosphorus in Wetl ands .......................................................................... 27Dynamic Biogenic Phosphorus ............................................................................... 28Dissertation Overview ............................................................................................. 30Dissertation Ob jectives ..................................................................................... 31Dissertation Layout ........................................................................................... 312 BIOGENIC PHOSPHORUS IN WETLA NDS: TOTAL POOL DETERMINATION AND THE APPLICATION OF 31P NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY .................................................................................................. 35Introducti on ............................................................................................................. 35Total Organic Phospho rus in So ils .......................................................................... 35Methodological Legacy ..................................................................................... 35Hierarchical Classificat ion of Me thods .............................................................. 37In-situ analysis of organic phosphor us ....................................................... 37Ex-situ analysis of organic phosphor us ...................................................... 38Existing Reviews of Organic Phosphorus in Wetlands ..................................... 39Published Estimates of Organic Phosphorus in We tland Soils ......................... 40Total Polyphosphates in Soils ................................................................................. 43In-situ Analysis of Po lyphosphates ................................................................... 43Ex-situ Analysis of Po lyphosphates .................................................................. 44Total Biogenic Phosphorus in Other Ecosystem Components ................................ 45Water Column .................................................................................................. 45Biota ................................................................................................................. 47Biogenic Phosphorus F unctional Gr oups ................................................................ 48Phosphomonoes ters .................................................................................. 49Phosphodies ters ........................................................................................ 50Polyphosp hates ......................................................................................... 50 5

PAGE 6

Phosphona tes ............................................................................................ 52Application of 31P Nuclear Magnetic Resonance Spectroscopy in Wetlands .......... 52Basic Princi ples ................................................................................................ 53Nuclei Intera ctions ............................................................................................ 55Range of Applications in Wetl ands ................................................................... 57Water column ............................................................................................. 58Biota ........................................................................................................... 59Soils and de tritus ....................................................................................... 59Experimental Considerations fo r Application to Soils ....................................... 62Conclusi ons ............................................................................................................ 643 INTERACTION OF BIOGENIC PHO SPHORUS WITH ANION EXCHANGE MEMBRANES: IMPLICATION FOR S OIL PHOSPHORUS ANALYSIS ................. 82Introducti on ............................................................................................................. 82Methods .................................................................................................................. 84Anion Exchange Me mbranes ........................................................................... 84Phosphorus De terminatio n ............................................................................... 85Experimental Design ........................................................................................ 86Anion exchange membrane exchange capacity ......................................... 86Phosphorus recovery by ani on exchange membr ane strips ...................... 86Purity and stability of organic and condensed phosphates in deionized water ....................................................................................................... 86Extraction of phosphorus compounds from wetland soils by anion exchange memb ranes ............................................................................ 87Results .................................................................................................................... 88Anion Exchange Membr ane Capacity .............................................................. 88Organic and Condensed Phosphorus Recovery by Anion Exchange Membrane St rips ........................................................................................... 89Purity and Stability of Phosphorus Compounds in Dei onized Water ................ 89Application to Wetland Soils for Exchangeable and Fu migation-Released Phosphor us ................................................................................................... 90Discussio n .............................................................................................................. 904 A SURVEY OF BIOGENIC PHOSPH ORUS IN WETLAND SOILS: A SOLUTION 31P NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY STUDY .................................................................................................................... 99Introducti on ............................................................................................................. 99Methods ................................................................................................................ 100Sampling ........................................................................................................ 100Biogeochemical Charac terization ................................................................... 101Phosphorus Com position ............................................................................... 102Data Analys is ................................................................................................. 105Results and Discussion ......................................................................................... 106Wetlands Samp led ......................................................................................... 106Biogeochemical Characteristi cs of Wetlands Sampled .................................. 107 6

PAGE 7

Solution 31P Nuclear Magnetic Resonanc e Spectrosc opy .............................. 109Extraction of total phosphorus .................................................................. 109Phosphorus com position .......................................................................... 110Discussio n ............................................................................................................ 114Conclusion s .......................................................................................................... 1175 PHOSPHORUS FORMS IN HYDROL OGICALLY ISOLATED WETLAND AND SURROUNDING PAST URE SOIL S ..................................................................... 139Introducti on ........................................................................................................... 139Materials and Methods .......................................................................................... 142Site Descrip tion .............................................................................................. 142Soil Sampli ng ................................................................................................. 143Hydroperiod Determination ............................................................................. 143Soil Biogeochemical Properties ...................................................................... 143Phosphorus Charac terization ......................................................................... 144Solution 31P Nuclear Magnetic Resonanc e Spectrosc opy .............................. 146Data Analys is ................................................................................................. 147Results and Discussion ......................................................................................... 147Soil Biogeochemical C haracterist ics .............................................................. 147Soil Phosphorus Co mpositio n ........................................................................ 148Solution 31P Nuclear Magnetic Resonanc e Spectrosc opy .............................. 149Impact of Hydr operiod .................................................................................... 151Phosphorus Storage ....................................................................................... 154Conclusion s .......................................................................................................... 1546 STABILITY OF SELECT BIOGENIC PHOSPHORUS COMPOUNDS UNDER AEROBIC AND ANAEROBIC CONDITIONS ........................................................ 164Introducti on ........................................................................................................... 164Methods ................................................................................................................ 166Microcosm Exper iment ................................................................................... 166Biogenic Phosphorus Spikes .......................................................................... 167Biogeochemical Charac terization ................................................................... 167Phosphorus Com position ............................................................................... 168Solution 31P Nuclear Magnetic Resonanc e Spectrosc opy .............................. 168Data Analys is ................................................................................................. 169Results and Discussion ......................................................................................... 170Phosphorus Re covery .................................................................................... 170Initial Biogenic Phosphor us Composit ion ....................................................... 171Stability of Phosphor us Functional Groups ..................................................... 171Stability of polypho sphates and DNA ....................................................... 172Stability of myo-Inositol hexaki sphosphate .............................................. 173Conclusion s .......................................................................................................... 1767 PHOSPHORUS TRANSFORMATION S DURING DECOMPOSITION OF WETLAND MACRO PHYTES ................................................................................ 189 7

PAGE 8

Introducti on ........................................................................................................... 189Materials and Methods .......................................................................................... 191Site Descrip tion .............................................................................................. 191Study Design .................................................................................................. 192Litterbag St udy ............................................................................................... 192Analysis of Biogeochemical Properti es .......................................................... 193Phosphorus Com position ............................................................................... 194Phosphorus extraction in NaOHE DTA ................................................... 194Solution 31P nuclear magnetic resonance spectrosc opy .......................... 195Data Analys is ................................................................................................. 195Results .................................................................................................................. 196Initial Litter Ma terial ........................................................................................ 196Mass Loss ...................................................................................................... 197Phosphorus C ontent ....................................................................................... 197Phosphorus Com position ............................................................................... 199Discussio n ............................................................................................................ 201Conclusion s .......................................................................................................... 2048 PHOSPHORUS FORMS AND DYNAMI CS ALONG A STRONG NUTRIENT GRADIENT IN A TROPICAL OM BROTROPHIC WE TLAND ............................... 218Introducti on ........................................................................................................... 218Methods ................................................................................................................ 220Study Site ....................................................................................................... 220Sampling ........................................................................................................ 220Soil Properti es ................................................................................................ 221Phosphorus Charac terization ......................................................................... 222Anion exchange me mbranes ................................................................... 222Solution 31P nuclear magnetic resonance spectroscopy .......................... 222Hydrolytic Enzym e Assay ............................................................................... 224Results .................................................................................................................. 225Soil Biogeochemical Properties ...................................................................... 225Phosphorus Biogeoc hemistry ......................................................................... 226Phosphorus recovery in NaOHE DTA ..................................................... 227Solution 31P NMR spectro scopy ............................................................... 228Discussio n ............................................................................................................ 2299 SUMMARY AND CO NCLUSION S ........................................................................ 244Biogenic Phosphorus Composit ion in Wetlands (Experimental Objective 1) ........ 245Influence of Landscape Position (Ex perimental Object ive 2) ................................ 246Influence of Redox Conditions (Ex perimental Object ive 3) ................................... 247Influence of Phosphorus Availability (Experimental Objective 4) .......................... 247Synthesis and Further Studies .............................................................................. 249Organic Matter and Redox Conditions ........................................................... 249Nutrient Status ................................................................................................ 251Iterative Proc essing ........................................................................................ 252 8

PAGE 9

APPENDIX A ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 1 ......... 257B ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 2 ......... 261C ADDITIONAL DATA AND INFORMATION PERTAINING TO CHAPTER 4 ......... 276D SIMPLIFIED METHOD FOR DETERMIN ATION OF TOTAL PHOSPHORUS IN WETLAND SAMPL ES .......................................................................................... 280LIST OF REFE RENCES ............................................................................................. 283BIOGRAPHICAL SK ETCH .......................................................................................... 319 9

PAGE 10

LIST OF TABLES Table page 2-1 Hierarchical classification of methods used in the study of organic phosphorus in soils and sediments ..................................................................... 66 2-2 Soil biogeochemical characteri stics and estimates of total organic phosphorus as determined in surface (0-10 cm) soils from a range of wetland units in South Florida (Reddy et al 1998) .......................................................... 67 2-3 Estimates of total organic phosp horus as determined in three surface sediment samples from a lagoon on the Po river delta, Saca di Goro Italy (Barbanti et al. 1994) ......................................................................................... 67 2-4 Relative distribution of biogeni c phosphorus within living biota. ......................... 68 2-5 A selection of functional groups based upon phosphorus, and select examples of compounds containing these groups mentioned in this dissertatio n. ........................................................................................................ 69 2-6 Studies employing 31P nuclear magnetic resonance spectroscopy in wetland and aquatic syst ems ........................................................................................... 71 2-7 Methodological details of studies employing 31P nuclear magnetic resonance spectroscopy in wetland so ils ............................................................................. 74 3-1 Phosphorus compounds tested on anion exchange me mbranes. ...................... 94 4-1 Wetland study sites sampled for characterization of biogenic phosphorus compositio n. ..................................................................................................... 119 4-2 Soil biogeochemical properties of studied wetl and systems. ............................ 121 4-3 Phosphorus composition of surf ace soils as determined by solution 31P NMR spectroscopy .................................................................................................... 123 4-4 Inorganic polyphosphates as determined by solution 31P NMR spectroscopy of wetland so ils.. ............................................................................................... 125 4-5 Eigen values of principal components determined on PCA applied to phosphorus composition withi n wetland soil s. .................................................. 126 4-6 Multiple linear regression m odels used to predict the ratio of phosphomonoesters to phosphodiesters in wetland surface soils. ................... 126 4-7 Parameter estimates for optimal model for predicting ratio of phosphomonoetsrs to phosphod iesters. .......................................................... 126 10

PAGE 11

4-8 Inositol hexakisphosphates as determined within group B wetlands. ............... 126 4-9 Correlation between microbial biomass phosphorus and phosphorus forms determined by solution 31P NMR spectro scopy ................................................ 127 5-1 Specific integral ranges used in the classification of solution 31P NMR spectra for lyophilized material re-suspended in 0.9 mL (1 mol L-1 NaOH 100 mmol L1) + 0.1 mL D2O. ............................................................................................... 156 5-2 Soil characteristics and nutrients determined in samples across landscape position. ............................................................................................................ 157 5-3 Phosphorus forms across landscape position determined by acid extraction and AEMNMR me thod.. .................................................................................. 157 5-4 Phosphorus forms determined by solution 31P NMR spectroscopy of amalgamated alkaline extracts from across the landscape transition ............... 158 5-5 Storage of total (n = 12) and organic phosphorus (n = 4) within the top 10 cm of soil, across landscape positions. Values from each landscape position represent averages 1 SD ............................................................................... 158 6-1 Characterization of surface soil collected for spike incubation microcosm study. ................................................................................................................ 177 6-2 Total phosphorus, after addition of bi ogenic P spikes, and recovery by AEMNMR method of all microcosms ....................................................................... 177 6-3 Phosphorus composition of microcosms as determined by AEM-NMR method. ............................................................................................................ 178 6-4 Phosphorus composition of soil sa mples as determined by parallel analysis of lyophilized soil extracts using two distinct nuclear magnetic resonance machines ......................................................................................................... 179 6-5 Phosphorus compositi on of microcosms spiked with myo -Inositol hexakisphosphate with time, as dete rmined by AEM-NMR method. ................ 180 7-1 Site characteristics for enriched a nd unenriched study sites within WCA-2A. .. 205 7-2 Characterization of litter materi al used within the decomposition study, consisting of two species ( Cladium and Typha ) from both unenriched and enriched portions of WCA-2A .......................................................................... 205 7-3 Four way Univariate AN OVA for mass rema ining.. ........................................... 206 11

PAGE 12

12 7-4 Simple exponential deca y rate constant (x =100e-kt ) and leaf litter half life calculated from material recovered ov er the course of 15 months within WCA-2A ........................................................................................................... 206 7-5 Four way Univariate ANOVA of phosphorus concentration in leaf litter. Model adjusted R2 = 0.947. ......................................................................................... 207 7-6 Linear regression of changes in mass of phosphorus within litterbags held in the field for between 33 and 454 days. ............................................................. 207 7-7 Phosphorus forms as determined by solution 31P NMR spectroscopy of NaOHEDTA extracts during macrophyte leaf litter dec omposition ................. 208 7-8 Coefficients (one standard deviation) of linear increases in concentrations of major phosphorus forms identified within leaf litter during decomposition at the enriched study site. ..................................................................................... 208 8-1 Soil biogeochemical characteristics from nine sampling stations across an ombrotrophic peat dome. .................................................................................. 235 8-2 Phosphorus forms identified by anion exchange membrane technique applied to fresh so il samples. ........................................................................... 235 8-3 Phosphorus forms identified by solution 31P NMR spectro scopy. ..................... 236 B4-1 Compounds and chemical shifts test ed for stability during lypophilization ........ 272 D-1 Standard biogenic P compounds test ed for recovery by anion exchange membrane strip ................................................................................................. 281

PAGE 13

LIST OF FIGURES Figure page 1-1 Wetland phos phorus cycl e, ................................................................................ 33 1-2 Dynamic biogenic phosphorus within wetlan d soils ............................................ 34 2-1 Frequency histogram of estima ted organic P within 117 wetlands (32 lacustrine, 85 pal ustrine) .................................................................................... 76 2-2 Frequency histogram of estimated to tal organic P within 117 wetlands broken down by general me thod groupi ng. .................................................................... 77 2-3 Structural comparison of myo -Inositol hexakisphosphate and -dglucopyranose 6-phosphate (pyranose ring form of -d-glucose 6phosphate) ........................................................................................................ 78 2-4 Basic phospholipid compound struct ure found within both eukaryotic and prokaryotic cell membranes ............................................................................... 79 2-5 Response of phosphorus nuclei to an applied magnetic fiel d. ............................ 80 2-6 Solution 31P nuclear magnetic resonance spectra showing common functional groups. ............................................................................................... 81 3-1 Exchange capacity of anion exchange membrane (AEM) strips.. ....................... 95 3-2 Recovery of phosphorus compounds by anion exchange membrane (AEM) strips.. ................................................................................................................. 96 3-3 Solution 31P nuclear magnetic resonance spec tra of P compounds measured after 24 h in deion ized wate r............................................................................... 97 3-4 Comparison of total and molybdate-r eactive P as detected in anion exchange membrane el uants. ............................................................................................. 98 4-1 Solution 31P NMR spectra of surface soils collected from a Michigan treatment we tland ............................................................................................. 128 4-2 Average nutrient concentrations in wetland surf ace soils ................................ 129 4-3 Categorization of wetland sites based upon Wards hierarchical classification of pH and organic matter (estimat ed by loss on i gnition). ................................. 130 4-4 Solution 31P NMR spectra of biogenic P composition within group A wetlands (high organic lo w pH). ....................................................................................... 131 13

PAGE 14

4-5 Solution 31P NMR spectra of biogenic P composition within group B wetlands (low organic low pH). ........................................................................................ 132 4-6 Solution 31P NMR spectra of biogenic P com position within group C wetlands (high organic hi gh pH). ..................................................................................... 133 4-7 Solution 31P NMR spectra of biogenic P com position within group D wetlands (low organic hi gh pH). ....................................................................................... 134 4-8 Principal component analysis of P composition within wetland soils as determined by solution 31P NMR spectro scopy. ............................................... 135 4-9 Principal component analysis of P composition within wetland soils as determined by solution 31P NMR analysi s. ....................................................... 136 4-10 Region 8 to 3 ppm within group B we tland spectra and peak assignments for; A) unidentified inositol phosphate, B) orthophosphate, C, D, E, F) myoInositol hexakisphosphate, G) scyllo -Inositol hexakisp hosphate. ..................... 137 4-11 Scatter plot of microbial P agai nst A) DNA and B) Po lyphosphates.. ............... 138 5-1 Location of study sites showing A) Florida outline with area of interest north of Lake Okeechobee, and B) detail of ranch sites containing two study wetlands each, within priority basins north of Lake Okeechobee. ................... 159 5-2 Phosphorus pools determined by AEMNMR method across the three landscape posit ions. ......................................................................................... 160 5-3 Example solution 31P NMR spectr a from amalgamated samples from the Beaty North we tland. ........................................................................................ 161 5-4 Phosphorus characteristics plott ed against hydroperiod: A) total phosphorus for all samples, B) total organic phosphorus determined by 31P NMR spectroscopy, C) phosphomonoesters, and D) phosphodi esters. .................... 162 5-5 Total soil carbon plotted against phosphorus concentrations across landscape positions from f our wetland si tes. .................................................... 163 6-1 Experimental setup for investigation of biogenic phosphorus stability under aerobic and anaerobic c onditions. .................................................................... 181 6-2 Biogenic phosphorus compounds used in spik ing experim ent. ........................ 182 6-3 Exchangeable phosphorus, deter mined by anion exchange membranes during microcos m study. ................................................................................... 183 6-4 Example solution 31P NMR spectra of soil samples spiked with biogenic phosphorus. ...................................................................................................... 184 14

PAGE 15

6-5 Spectral deconvolution and peak a ssignments in 8 to 3 ppm region of solution 31P NMR spectr a. ................................................................................ 185 6-6 Detail of 8 to 3 ppm region of NMR spectra gathered on soil spiked with biogenic phos phorus. ....................................................................................... 186 6-7 Solution 31P NMR spectra, including the regi on 8 to 3 ppm in detail, from alkaline soil ex tracts. ........................................................................................ 187 6-8 Concentrations of myoInositol hexakisphoshate as determined within microcosm soils under aer obic and anaerobic conditions for up to 48 days. .... 188 7-1 Location of chapter 7 study site s within Water Conservation Area (WCA)-2A in the northern Ev erglades .............................................................................. 209 7-2 Mass remaining of four litter types pl aced at two distinct sites within WCA-2A and recovered at time interv als up to 454 days. ............................................... 210 7-3 Changes in litter phosphorus concentra tion over time at, A) enriched site and, B) unenriched site.. ................................................................................... 211 7-4 Changes in mass of phosphorus in macrophyte leaf litter held within litterbags over the cour se of 454 da ys. ............................................................. 212 7-5 Initial phosphorus composition of Typha and Cladium leaf litter sourced from the enriched portion of WCA-2A ...................................................................... 213 7-6 Initial phosphorus composition of det ritus and surface soils from enriched and unenriched study sites samp led on (10/20/ 03). ......................................... 214 7-7 Example solution 31P NMR spectra showing changes in phosphorus composition of Typha leaf litter during decomposition at both an unenriched and enriched site over the course of 454 days. ................................................ 215 7-8 Changes in proportion of major P pools found within macrophyte leaf litter over the course of 454 days of decomposition in WCA-2A. .............................. 216 7-9 Conceptual model of phosphorus tu rnover in wetland macrophyte detritus under, A) enriched and, B) unenriched condit ions. ........................................... 217 8-1 Overview of study transect and sampling sites with the Changuinola peat deposit, San San Pond Sa k N.W. Panama. ..................................................... 237 8-2 Nutrient gradient; A) ma ss of total P, B) mass of to tal N, C) Molar ratio C:P, and D) N:P from nine study sites within the Changuinola peat deposit.. .......... 238 8-3 Enzyme activity from nine study site s within the Changuinola peat deposit. .... 239 15

PAGE 16

8-4 Comparison of P recovered by al kaline extraction (0.25 mol L-1 NaOH and 0.05 mol L-1 EDTA) of air died soils (4 h and 16 h) or in addition to AEM extraction of non fumigated and fumigated fresh soil. ...................................... 240 8-5 Solution 31P NMR spectra showing range of P forms present in surface soils from select sites across the study trans ect.. ..................................................... 241 8-6 Detail of solution 31P NMR spectra from si te seven so ils. ................................. 242 8-7 Solution 31P NMR spectra of site one and nine soils after application of anion exchange membranes with (F) and wit hout (NF) fumigation step using hexanol. ............................................................................................................ 243 9-1 Comparison of conceptual model of dynamic biogenic P cycling in soil (Figure 1-4) modified for A) systems dominated by interactions with the mineral phase (e.g. uplands) and B) systems dominated by interactions with organic matter (e.g. wetlands). ......................................................................... 254 9-2 Influence of anaerobic conditions (t ypical of wetlands) upon both organic matter and mineral phase stabilizati on of biogenic P in soils. ........................... 255 9-3 Simplified linear progression of development of biogenic P from inputs to wetland soil ...................................................................................................... 256 9-4 Iterative processing of P within wetland soils. Mediated by microbial processes, biogenic P undergoes intera ctions with both organic matter and mineral phase. .................................................................................................. 256 B4-1 Experimental schemtaic for test of biogenic P stability during lyopilization ....... 273 B4-2 Solution 31P NMR spectra in various matrix environments ................................ 274 B4-3 Detail of solution 31P NMR spectra of st andard biogenic phosphorus compounds after lyophilization A) Glucose 6 phosphate showing pH dependant peak splitting B) Alkaline hy drolysis products of RNA..................... 275 D-1 Comparison of total P determi ned by TP Ash vial method and Andersen (1976) procedur e. ............................................................................................. 282 16

PAGE 17

LIST OF ABBREVIATIONS AEM Anion exchange membrane AIC Akaike information criterion DI Deionized water DDI Distilled deionized water D2O Deuterium oxide EBPR Enhanced biological P removal EDTA Ethylenediami netetraacetic acid EPA Environmental protection agency FID Free induction decay FL Florida FT Fourier transform GF-B Glass fiber-B grade HDPE High density polyethylene ICPOES Inductively coupled plasma atomic emission spectroscopy IP6 Inositol hexakisphosphate IQR Interquartile range MDP Methylenediphosphonic acid MRP Molybdate reactive P NEXAFS Near edge X-ray absorption fine structure NMR Nuclear magnetic resonance NRCS National resource conservation service NUTS NMR utility transform software P Phosphorus PAO Polyphosphate accumulating organism 17

PAGE 18

PAEM Exchangable P recovered by AEM method PM Microbial P recovered by AEM method ppm Parts per million rpm Revolutions per minute RSD Relative standard deviation SD Standard deviation SEPI Seqential extraction focused on determination of inorganic P SEPO Sequential extraction focused on the determination of organic P S/TEM Scanning / transmission electron microscopy TC Total carbon TN Total nitrogen US United States USDA United States depar tment of agriculture WCA-2A Water Conservation Area 2A XANES X-ray adsorption near edge structure 18

PAGE 19

Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy BIOGENIC PHOSPHORUS IN WETLA NDS: SOURCES AND ST ABILIZATION By Alexander William Cheesman August 2010 Chair: K. Ramesh Reddy Cochair: Benjamin L. Turner Major: Soil and Water Science Nutrient pollution, from both diffuse agric ultural applications and point source contamination is a pressing concern for ecosystem integrity across the globe. Of particular concern, given the historical precedence of industrialization within the US, is that of phosphorus (P) pollution. Wetlands are both a victim and potential solution of this. Oligotrophic systems are su ffering fundamental shifts in ecosystem functioning at the same time that constructed wetlands ar e being touted for their remediation value. The role that biological processing and sequestration plays in the P cycle of wetlands has long been recognized, yet it is only recently that analytical techniques have emerged that allow us to probe the functional nature and stability of these P forms in environmental samples. The nature of P functional groups has immediate and profound implications on the interaction and fate of P in the environment, from determining susceptibility to enzymatic and abiotic hydrolysis, to dictating long-term stabilization. Therefore, this dissertation has sought to provide an advance in our understanding of biogenic P in wetland soils by first reviewing the current science and 19

PAGE 20

then applying 31P nuclear magnetic resonance (NMR) sp ectroscopy to investigate the composition and mechanisms that determine t hat composition in wetland systems. Initial studies focused on the surveying of P composition in a broad range of palustrine wetland soils. This work not only showed the range of P forms found within wetland soils (e.g. phosphonates, phosphomonoeste rs (including inositol phosphates), phosphodiesters and long chain inorganic polyphosphates) but highlighted basic wetland properties that appear to impact P co mposition. Landscape position, vegetation and climate were shown to have little di rect influence on P composition while biogeochemical characteristic such as ; pH, organic matter content, and nutrient availability (themselves a product of wetl and setting) appeared to be linked directly to the P composition of surface soils. S ubsequent chapters sought to explore the mechanistic role of these biogeochemical characteristics on determining soil P composition. The trend, observed between wetlands, that soils with a higher organic matter content had a higher proportion of P found as phosphodiesters was explored by comparison of soils across a landscape conti nuum. By comparing P co mposition of soils under similar vegetation and management histories across the wetlandupland transition, the mechanistic role of organic matte r content in isolation was investigated. In the wetlands studied, depressi onal systems within an agricultural landscape north of Lake Okeechobee Florida, P composition was shown to be independent of landscape position and organic matter content. This was unexpected, but believed to be the result of the unique role organic matter plays in t he P dynamics of the sandy, low P binding capacity soils of the region. This lack of di stinction in the P com position associated with 20

PAGE 21

21 organic matter across a landscape transition was also seen in a study established to determine the redox sensitivity of certain P forms. In this mesocosm study, there was no substantial difference in the turnov er rates of DNA and the phosphomonoester myo Inositol hexakisphosphate ( myo -IP6) when considering their presence in a highly organic freshwater system. The role of microbial processing of soil organic matter in response to environmental conditions, specifically P avai lability, was determined by tracking the transformations of P forms within detrital organic matter entering a wetland system and by monitoring P composition in surficial so ils across a profound nutrient gradient. In both cases it was apparent that P composition was independent of t he major allochthonous inputs and represented P forms derived as a resu lt of in-situ microbial processing of organic matter in direct respons e to environmental conditions. In conclusion, i use nformation derived by the study of a range of palustrine systems to develop a working model of biog enic P sources and stabilization in wetlands. This provides not only a significant advanc e in our understanding of P composition and cycling in wetlands but also provides insite into the biological processes associated with the P cycle of both wetland and terrestrial ecosystems.

PAGE 22

CHAPTER 1 BIOGENIC PHOSPHORUS IN WE TLANDS: AN INTRODUCTION Phosphorus in Wetlands Phosphorus (P) is a multiv alent nonmetal, with five potential oxidation states. Highly reactive in the environment, its potentia l for polyatomic interaction has led to its evolution as a vital component of cellula r biochemistry (Brown and Kornberg 2004). Phosphorus constitutes a significant propor tion of nucleic acids, lipid membranes, proteins and phosphorylated metabolic intermediates (Raghothama and Karthikeyan 2005), it is therefore a vita l nutrient for biomass producti on and often limits growth in freshwater (Reddy et al. 2005; Verhoeven et al. 2006) and, increasingly, coastal (Sundareshwar et al. 2003; Turner et al. 2003g) wetlands and aquatic ecosystems. Although aerial deposition of P may be important in highl y weathered, isolated soil systems (Chadwick et al. 1999), t he predominant route of P input to soils, and ultimately to biological uptake, is through in situ dissolution of the mi neral phase (Walker and Syers 1976) or external delivery via the hy drologic cycle. Wetlands typically occupy downstream positions in the landscape, and therefore often receiv e large inputs of P from upstream sources (Richardson 1999). Anthropogenic alteration of global P cycling has been profound and is accelerating, as a consequence of increased fertilizer application (Carpenter et al. 1998), point source P discharges (Garcia-Pintado et al. 2007), and by indirect means, through human -mediated alteration of related biogeochemical cycles and landscape ecolog ical processes (Caraco 1993). Human populations and resource consumption c ontinue to grow and pressure on the environment will increase (Godfray et al. 2010) As a consequence, we are likely to see increased disruption of the P cycle. For example, Tilman et al. (2001) predicted a 2.422

PAGE 23

fold increase in global P fertilizer use bet ween years 2000 and 2050. Such nutrient redistribution will cause fundam ental changes in the environment with wetlands a focal point of this change. Global models suggest the current level of P retention within freshwater wetlands is 3.1 Tg y-1, compared with preindustrial levels of only 1.2 Tg y-1 (Bennett et al. 2001). This increase in net P input to wetlands has caused many wetlands, especially within agricultural wate rsheds, to surpass their P loading capacity, putting them at risk of biol ogical degradation (Hager they et al. 2008; Verhoeven et al. 2006). It is therefore likely that increased anthropogenic disr uption of the P cycle will cause further degradation of wetland ecological capital (Khan and Ansari 2005). Wetlands and other aquatic ecosystems fulfill a vital role within a functioning landscape continuum (Sheaves 2009), providing direct and indirect ecosystem services (Costanza et al. 1997; Silvius and Gies en 1996), including retention of water and nutrients at the landscape scale (Mitsch and Day 2006; Moreno et al 2007; Paludan et al. 2002; Perkins et al. 2005). Recognition of wetland value has generated initiatives and programs designed to preserve these vita l ecological functions in the landscape. For instance, the US Army Corps of E ngineers and Environmental Protection Agency (EPA) administer a program; the Compensatory Mitigation for Losses of Aquatic Resources (www.epa.gov/wetlandmitigation/ ) designed to deliver no net loss of wetland ecosystems. Furthermore, artificial wetland areas are being created worldwide to impound and treat polluted waters befor e they are discharged to natural systems (Babatunde et al. 2008; Day et al. 2004). Although there is an apprecia tion for the roles that biological turnover and sequestration of P play within wetland soils (Newman and Robinson 1999; Reddy et al. 23

PAGE 24

2005; Wetzel 1999), there is currently littl e information on the forms of P found in wetland soils (CHAPTER 2). The functional na ture of P affects both its stability and biological turnover (see Dynamic Biogenic Phosphorus) ther efore, detailed information on the forms, as well as the mechanisms that determine them, is vital if we are to understand the consequences of human alterations of natural P cycling (Corstanje et al. 2007; Kuhn et al. 2002) and successfully model and manage wetl and systems for the future (Moustafa et al. 1999). Wetland Soil Wetlands represent a transitional ecotone between terrestrial and aquatic ecosystems. There is, however, no single, accepted definition for the term wetland (Cowardin et al. 1979). Federal institutions in the USA, at the recommendation of the Wetlands Subcommittee of the Federal Geog raphic Data Committee, use the Cowardin system to define wetland and deepwater habitats (Federal Register 61, 29 July, 1996, 39465-39466). Under this scheme, five major syst ems are identified: marine, estuarine, riverine, lacustrine and palustrine ecosystems. The first four contain both wetland and deepwater ecosystems, while palustrine systems refer solely to wetlands (APPENDIX A1). Considering the multid imensional continuum across landscapes (Euliss et al. 2004), the upper boundary of wetl ands is distinguished from terrestrial systems by the presence of water for a portion of the growing season. In t he USA, wetland delineation, for the purposes of Section 404 of the Clean Wa ter Act (Title 33 U.S. Code, Sec. 1344), is codified using present hydrology, vegetat ion and hydric soil indicators (USDA-NRCS 2010; Environmental Laboratory 1987). The distinction between wetlands and deepwater habitats is less clear, and is sometimes open to interpretation based upon the classification scheme used. The Coward in system defines the lower boundary of 24

PAGE 25

marine and estuarine wetlands as the low-water mark during spring tides, and the boundary within inland waters as 2 m below t he recorded low-water mark (Cowardin et al. 1979), with the important caveat that if rooted macrophyte s exist below this level, the wetland boundary is extended to include t hem. The most comprehensive and globally adopted treaty on wetland recognition and wis e use, the Ramsar Convention on Wetlands (1971), uses a broader definition of wetlands. Under the text of the Convention (Article 1.1) wetlands are defined as: areas of marsh, fen, peatland or water, whether natural or artificial, permanent or temporary, with water that is static or flowing, fresh, brackish or salt, including areas of marine wate r the depth of which at low tide does not exceed six meters. Signatories of the Ramsar Convention, including the USA, recognize that the term wetland can also be applied to landscape features, such as lakes, which are considered deepwater habitats in the Cowa rdin system (APPENDIX A2). Similarly, some geomorphic approaches for classifying global wetlands (Semeniuk and Semeniuk 1995; Semeniuk and Semeniuk 1997) include fresh, deepwat er habitats as a subcategory of wetlands (APPENDIX A3). Within this dissertation, I have chosen to study the dynamics of biogenic P in palustrine wetlands as defined by Cowardin et al. (1979). In doing so, I accept the commonly held notion of wetlands as transit ional state between extremes of truly terrestrial and aquatic ecosystems. These transition systems remain understudied (CHAPTER 2), eschewed by pedologists and s edimentologists, despite the fact that wetlands are characterized by an exciting interplay of processes found in upland and deepwater environments (Deevey 1970). Throughout this dissertation I use the term wetland to describe environm ents affected significantly by water, but not dominated 25

PAGE 26

solely by its presence, yet the term wetland is not restricted to a priori criteria, except when used in conjunction with descriptive modi fiers such as those of the Cowardin system. In 1998, the US Department of Agri cultureNatural Resources Conservation Service (USDANRCS) extended the term soi l to include surface materials that are permanently submerged by water, yet capable of supporting rooted plants (Soil Survey Staff 2010). Since material substrates wit hin wetlands show evidence for both diagenesis and pedogenesis (Dem as and Rabenhorst 1999, 2001) in this dissertation, I have used a mechanistic interpretation of s ubstrate materials, drawing on relevant literature pertaining to both soils and sediments. If not otherwise specified by the literature source, I use the term soil to i dentify all benthic substrates within a wetland. Wetland Phosphorus Cycle Traditional consideration of the P cycle in wetland ecosystems contends that P cycling is relatively simple compared to other nutrients e.g. nitrogen, given that P has no significant gaseous phase or change in oxi dation state (Reddy et al. 2005; Wetzel 1999). In fact, there are multiple oxi dation states found within natural systems (CHAPTER 2) including reduced inorganic phosphties (Pech et al. 2009), and highly reduced phosphine gas (Devai et al. 1988). Which may lead to the gaseous translocation of significant quantities of P in certain wetlands and benthic sediments (Devai and Delaune 1995; Geng et al. 2005), although the uns table nature of such highly reduced inorganic components (Hanrahan et al. 2005; Morton and Edwards 2005) suggests a role only under rare conditions (Gassmann and Glindemann 1993). The transport of the more common, particula te and dissolved forms of P into and within wetlands has been the subject of much stud y (Kadlec 1997; Mitchell and Baldwin 2005; 26

PAGE 27

Reddy et al. 2005; Turner 2005) allowing the development of theoret ical models of P mass cycling within wetland ecosystems (Figure 11). In such a model, a modification of the three features identified in all wetl ands by Reddy and Patrick (1993), wetlands consist of four essential ecosystem com ponents: soil, water column, live biomass, and detritus. Transfers between ecosystem co mponents are varied (Figure 1-1), yet in wetlands typified by water-induced anaerobosis and high rates of biomass production, detrital organic matter plays a central role (McGill and Cole 1981). Not only does detritus (exogenous and endogenous organic matte r) act as a conduit for nutrient cycling, it also forms an important component of accreting material. Many wetlands accumulate detritus with surface layers typically containing greater amounts of organic matter relative to both underlying material and adjacent terrestrial ecosystems (Axt and Walbridge, 1999; Gathumbi et al ., 2005; Pant and Reddy, 2001). The generalized P cycle (Figure 1-1) c ontains biogenic forms (CHAPTER 2Biogenic Functional Groups), found within organisms (e.g. above-ground biomass and microbial components of the detritus/soil), and within the extracellular environment, stabilized by complex interactions with mi neral and organic matter (Celi and Barberis 2005b), or as distinct extr acellular inclusions [granules] (Diaz et al. 2008). Biogenic Phosphorus in Wetlands The descriptor biogenic has been appl ied to amorphous silica found within structures such as diatom frustules as a means of distinguishing it, alongside opal phytoliths, from mineral silicates found wit hin soils and sediments (Street-Perrott and Barker 2008; Struyf and Conley 2009). Simi larly, the term has been applied by those studying sediment P as a means of distinguishing P from bi ological sources from P of mineral origin (Ahlgren et al. 2006a; Li u et al. 2009). Sedimentologists have long 27

PAGE 28

recognized the relationship between of Si and P elemental cycles joined through processes of biological fixation and sedim entation (Conley et al. 1993; Schelske et al. 1986). In this dissertation, I use the term biogenic to re fer to P compounds synthesized by organisms. Predominantly organic, biogenic P also includes inorganic forms such as polyphosphate (see CHAPTER 2-Biogenic F unctional Groups), which is increasingly recognized as an important component of biol ogically mediated P cycling within wetland and aquatic systems (Ahlgren et al 2006; Hupfer et al. 2007). Dynamic Biogenic Phosphorus Biogenic P in wetland soils includes both intracellular P held within viable algal, macrophyte, microbial and faunal biomass, and extracellular P held within the soils matrix (Figure 1-2). Biogenic P may represent a high proportion (> 60%) of total soil P (CHAPTER 2), and represents a dynamic pool transformed by both biotic processes and abiotic factors. In subsequent chapters, I will explore in detail how biological and abiotic processes within wetlands may influen ce biogenic P pool. Here, I briefly outline the rationale that underlies the conc ept of dynamic biogenic P. Although some aquatic heterotrophic bacteria can directly assimilate simple sugar phosphates (Heath 2005) this ability appears to be rare and there is no significant mechanism of organic P uptake equivalent to, fo r instance, the assimilation of organic N seen in boreal plant/mycorrhizal associations (Nasholm et al. 1998). Therefore, for biotic uptake of P, microbes along with plants must hydrolyze biogenic P to orthophosphate in the extracellular or periplasmic environment (Oberson and J oner 2005; Raghothama 1999; Raghothama and Karthikey an 2005; Schachtman et al. 1998). Active acquisition of P by biological communities includes the exudation of phos phatase enzymes (Kuhn et al. 2002), organic acids, solubiliz ing agents (Raghothama and Karthikeyan 2005), 28

PAGE 29

and modification of microbial associations (Leake and Miles 1996; Richardson et al. 2009). All such processes are mediated by in teractions with external factors. For example, the presence of extracellular phos phatase, a vital component of biologically mediated hydrolysis of phosphoe sters (Turner and Haygarth 2005), is coupled to the nutrient status of biological communities (Olander & Vitousek 2000; Wright & Reddy 2001) and has an activity modified by external abi otic factors. Adsorption of extracellular enzymes to colloidal particles and phenolic humic substances in soils may infer greater stability, often extending periods of activity, but it also r educes their catalytic capacity (Boavida and Wetzel 1998; Huang et al. 2005; Quiquampoix and Mousain 2005; Wetzel 1999). The consequence of this tradeoff is current ly unknown, but is likely to impact the rates of P turnover. In contrast, to the hydrolysis of extrac ellular biogenic P is its stabilization via interaction with colloidal particles, comple xation with humic subst ances, precipitation with polyvalent cations, and/or physical inco rporation into organic matter (Celi and Barberis 2005b). The relative importanc e of various mechanisms depends upon the P composition and biogeochemical characteristics of a wetland. Two processes that are particular important to biogenic P stabilization in wetlands are (i) physical incorporation of P into organic matter (typically high in wetlands relative to surrounding terrestrial systems) and (ii) redox-sensit ive stabilization of biogenic P with mineral components. It has been suggested that the latter mechanism re sult in the redox sensitive stabilization of both polyphosphates (Hupfer and Lewandowski 2008; Sannigrahi and Ingall 2005) and the major phosphomonoester myo -Inositol hexakisphosphate ( myo -IP6) (Celi and 29

PAGE 30

Barberis 2007; McKelvie 2007; Suzumu ra and Kamatani 1995b; Turner and Newman 2005). The functional forms and concentrations of intracellular microbial P vary due to changing environmental conditions and species composition and (Heath 2005; Makarov et al. 2005) (CHAPTER 2), yet extracellula r biogenic P also show s a great deal of variability. The balance between stabilized and l abile extracellular biogenic P represents an equilibrium state, the result of continuous inputs of organic detritus and exudation, moderated by abiotic stabilizat ion, and both abiotic and biolog ically mediated hydrolysis (Figure 1-2). This dissertation utilizes solution 31P NMR spectroscopy to investigate the dynamic nature of biogenic P in wetlands. Dissertation Overview Using the conceptual model of P cycling in Figure 1-1, I address the role of biogenic P in wetlands by investigati ng P forms and amounts in the four major ecosystem compartments: water column, bi ota, detritus, and soil. In doing so, it becomes apparent that our present underst anding of biogenic P composition in wetlands is incomplete (CHAPTER 2). In particular, I found that although most researchers rightfully contend that organic P, a sub-group of biogenic P, is the dominant form in wetland soils and detritus (Newman and Robinson 1999; Reddy et al. 2005), there is little information on the functional forms of P present (CHAPTER 2). Studies that have identified biogenic P forms within wetland soils suggest the forms differ fundamentally from those in terrestrial syst ems (Sundareshwar et al. 2009; Turner and Newman 2005). This has implications with respect to our ability to estimate the stability of sequestered P, understand ecological in teractions, or model the effects of anthropogenic perturbations. 30

PAGE 31

Dissertation Objectives Objective 1: Review the current knowl edge of biogenic P in wetlands and identify gaps in our current understanding. Objective 2: Determine the influence of wetland characteristics and soil physicochemical properties on forms of biogenic P in wetland soils. Hypothesis: The composition of biogenic P in wetland soils varies systematically with respect to wetland characterist ics, landscape posit ion, and/or soil biogeochemical properties. Objective 3: Determine how position in the landscape, i.e. fr om terrestrial to wetland environments, influences bi ogenic P composition of soils. Hypothesis: Landscape position impacts soil properties, which in turn influence biogenic P composition. Spec ifically, higher productivity, a receiving position in the landscape and reduced decomposition leads to increased organic matter content within wetlands. This leads to differences in mechanisms of abiotic stabilization leading to hydroperiod correlating with a decrease in the ratio of phosphomonoesters to phosphodiesters (predominantly DNA). Objective 4: Determine how anaerobic co nditions impact biogenic P composition. Hypothesis: Anaerobic conditions des tabilize the phosphomonoester myo -IP6 and polyphosphates, and lead to reduced dec omposition of phosphodiester DNA. Objective 5: Assess the role of nutri ent availability in determining biogenic P composition within wetland detritus and soils. Hypothesis: Increased P availability, due to elevated ambient conditions, will reduce turnover of biogenic P by microbes, thereby altering P composition within wetland materials. Dissertation Layout CHAPTER 1. Introduction to the basic concepts and rationale behind the dissertation. CHAPTER 2. A detailed review of publ ished work on biogenic P in wetlands, focused on amounts and proportions of organic P, methods used, and the consideration of polyphosphates. The chapt er introduces the concept of biogenic functional groups and the application of 31P NMR spectroscopy to wetlands. [Objective 1] CHAPTER 3. Exploration of the met hodological implications of observed interactions between biogenic P and a commonly used anion exchange membrane strip procedur e. [Method development] 31

PAGE 32

32 CHAPTER 4. An exploratory survey of biogenic P forms in 28 diverse wetlands. In addition to basic descriptions of P com position, the study establishes potential mechanistic drivers behind observed patterns. [Objective 2] CHAPTER 5. A study of isolated depressional wetlands north of Lake Okeechobee, FL, which utilizes a natural landscape transition to investigate the impact of altered hydroperiod and c oncomitant organic matter content on biogenic P composition. [Objective 3,4] CHAPTER 6. A microcosm-based study inve stigating how the differential stability of biogenic P under aerobic and anaerobic c onditions may account for observed P composition in wetlands. [Objective 4] CHAPTER 7. An investigation into how nutrient status impacts the accumulation and speciation of P during microbial proc essing of herbaceous organic matter. [Objective 5] CHAPTER 8. A detailed study into t he composition of biogenic P in an ombrotrophic peat dome, Panama, specifically investigating how nutrients and vegetation cover alter biogenic P composition. [Objective 5] CHAPTER 9. Final synthesis and conclu sions drawn from the dissertation.

PAGE 33

Figure 1-1. Wetland phosphorus cycle, with areas of focus found in this dissertati on highlighted in green. 1Biotic uptake 2Trophic progression 3Senescence 4Hydrolysis (en zymatic or abiotic) 5Re-suspension 6-Organic matter accretion 7Deposition 8Erosion 9Pedogenesis 10Exudation/ active transport 11Abiotic stabilization/adsorption 12Destabiliza tion/ desorption 13Flux/ diffusion 14Precipitation 15Dissolution 16Polyphosphate accumulation 17Polyphosphate utilization 18Microbial mediat ed dissolution of mineral P 19Diagenesis 20Export of biomass (migration, harvesting etc) 21Import of biomass (migration) 22Water import/export 33

PAGE 34

34 Figure 1-2. Dynamic biogenic phosphorus within wetland soils, black arrows repres ent P flux. i = biological senescence and exudation, ii = biological uptake of orthophosphate, iii = flux of orthophosphate within water column, iv = adsorption of orthophosphate to mineral components, v = hydrolysis of biogenic P, vi = stabilization of biogenic P with abiotic phase of soil, viii = stabilization of biogenic P to organic ma tter within soil, viii = formation of P containing minerals, ix = phosphate solubi lization, x = intracellular P cycling.

PAGE 35

CHAPTER 2: BIOGENIC PHOSPHORUS IN WETLANDS : TOTAL POOL DETERMINATION AND THE APPLICATION OF 31P NUCLEAR MAGNETIC R ESONANCE SPECTROSCOPY Introduction The role of biological inputs and tu rnover within the P cycle has long been recognized (Potter and Benton 1916). Yet it is onl y recently that analytical techniques have developed to the point that resear chers can define sources and track the transformations of biogenic P in the environm ent (McKelvie 2005). Techniques such as solution 31P NMR spectroscopy provide a powerful tool which allows researchers to investigate the functional nat ure of P pools, while also providing the satisfactory absolute standard required when evaluati ng more routine oper ationally defined measures of P pools (Kaila and Virtanen 195 5; Mehta et al. 1954). This chapter summarizes existing info rmation on total biogenic P (organic phosphorus and polyphosphates) within wetlands, and introduces 31P NMR as a technique used in the study of all biogenic P functional forms. Total Organic Phosphorus in Soils Methodological Legacy It is not the aim of this section to reiter ate the work cataloguing the development of methods used in the study of organic P in wetland soils and sediments, for which the reader is directed to a number of seminal papers (Barbanti et al. 1994; Kuo and Sparks 2001; Mehta et al. 1954; Sommers et al. 1970; Tu rner et al. 2005). Yet it is important to understand the range of techniques applied to the study of organic P when attempting to reconcile variation seen between studies. Mo st studies characterize P by operational criteria and attempt to a ttribute functionality by post hoc interpretation. With each new generation of researchers me thods are adopted, and then adapt ed to provide what is 35

PAGE 36

perceived as a better measure under a giv en set of parameters, be they high mineral content, the presence of complexing humic substances, or high anthropogenic loading. Such methodological alterations may be profound, with the proposal of novel combinations of extracts to produce new sequential procedures (see APPENDIX B1) or subtle, from the inclusion of additional pr etreatment (Schlichti ng and Leinweber 2002; Turner et al. 2007b), to the use of dilute salt washes to remove adsorbed P during sequential extractions (Bar banti et al. 1994; Sommers et al. 1970). Yet such modifications only exemplify the range of methods and the lack of any universally applied standard (McKelvie 2005). When researchers have compared these procedures and modifications, it has been common to re gard the method that returns the highest organic P values as the most accurate (B arbanti et al. 1994; Kaila and Virtanen 1955; Saunders and Williams 1955). With direct ex-situ analysis methods (see Hierarchical Classification of Methods), this comparativ e approach was considered valid since it was believed that two main sources of analytical error, the hydrolysis of organic P, and its incomplete extraction, would both lead to under estimation of the true pool. With modern evidence demonstrating that post hoc interpretation of pools is often in error, such comparative approaches may have been inappropr iate. For example, routinely used colorimetric techniques may lead to the over estimation of organic P in alkaline extracts by initially missing orthophosphate complexed within the organic matrix (Turner et al. 2006b). Similarly, recent work has show n the HCl extracted fraction of the well established Hedley et al. (1982) scheme may contain subs tantial amounts of organic P (He et al. 2006). Application of operationally defined procedures without verification of their interpretation under sample-specific conditions may well produce incomplete and 36

PAGE 37

erroneous determination of tota l organic P pools. The potentia l errors associated with indirect ex-situ determination of organic P (see Hiera rchical Classification of Methods) have long been recognized, with biases shown to include the solubilization of Al and Fe bound P by hydrogen peroxide treatments (Pears on 1940), and the alteration of the mineral phase by high temperature ox idation procedures (C ondron et al. 1990b; Dormaar and Webster 1964; Frink 1969; Pears on 1940). Yet, their simplicity and ease of application has meant that ignition-based estimates of organic P are still commonly used especially within highly organic syst ems (Reddy et al. 1998) where error associated with a modified mineral phase is assumed to be minimal. With no universally accepted, or even appropriate standard method, estimates of total organic P should be treated with caution, es pecially when based on upon ex-situ analysis and uncorroborated post hoc interpretation. Hierarchical Classi fication of Methods In the determination of the P within soils and sediments, we are able to distinguish methods at a number of levels (Table 2-1). In itially, a distinction must be made between those techniques that allow for t he non-destructive analysis of the P in-situ, i.e. within the intact soil matrix, and those that require its extraction and analysis ex-situ In-situ analysis of organic phosphorus With no, or little, pretreatm ent certain techniques are abl e to distinguish structural and functional forms of P present in the co mplete soil. While techniques such as Scanning / Transmission Electron Microscopy (S/TEM), X-Ray Absorption Near Edge Structure (XANES) or Near Edge X-ra y Absorption Fine Structure (NEXAFS) spectroscopy may not be applicable to the determination of organic P (Harris and White 2008), recent developments within synchro tron-based X-ray spectromicroscopy may 37

PAGE 38

provide a method of determini ng broad functional forms (B randes et al. 2007). In addition, solid-state 31P NMR spectroscopy allows for the determination of P chemical bonding environments within bulk soils (Cont e et al. 2008). While the later has been applied to agricultural and calcareous marsh so ils (Delgado et al. 2000; McDowell et al. 2002) as well as demineralized marine sediments (Sannigrah i and Ingall 2005) limitations exist, given typically low spectral resolution a result of anisotropy within solids (see Nuclei Interactions), the presenc e of paramagnetics, and typically high concentrations of humic substances (Shand et al. 1999). Nevertheless, as methodological techniques and data interpretation improves, in-situ approaches will offer us a gold standard of P analysis in wetlands as non-destructive methods able to discern chemical and functional pools without alteration of environmental samples. Ex-situ analysis of organic phosphorus In comparison to the in-situ methods mentioned above, most techniques used to estimate total organic P require some form of extraction, with inferences drawn from its subsequent analysis. Extraction methods applied may be direct or indirect. With direct extraction procedures, the aim is to remove and determine the nature of various organic P fractions while minimizing their alteration. Indirect methods seek to establish the nature of the inorganic P and use the differ ence between it and an established measure of total P to determine the or ganic pool. Although subtle, the distinction is needed due to the inherent bias associated wit h potential methodological error. Both direct and indirect methods can be composed of single or sequential extraction steps using a range of solutes. Methods have employed neutral salt, acidic, alkaline, organic, and chelating solutes under ambient and r educing conditions in various combinations, and under various experimental conditions (see Table 2-1, APPENDIX A1 and A2). Analysis of the 38

PAGE 39

resulting extracts themselves may also be di rect or indirect. Direct methods, such as solution 31P NMR spectroscopy (Cade-Menun and Preston 1996), chromatographic techniques (Gilbin et al. 2000) and enzymatic hydrolysis (Bunemann 2008) seek to determine the nature of the P directly in the extract. The mo re commonly applied indirect method seeks to determine orthophosphate, or commonly used operational analogues (e.g. molybdate reactive P), before and after oxidation of organic matter and attribute the difference to organic bound P. Recent work has challenged the validity of assumptions implicit in this approach, highlig hting the potential for analytical errors in determining true orthophosphate concentrations (Gerke 1992; Kowalenko and Babuin 2007; Turner et al. 2006b). I ndeed even without the potential fo r interference in extract solutions the commonly determined molybdat e reactive P (MRP) (John 1970; Murphy and Riley 1962) may include reactive or ganic P fractions (Baldwin 1998) often overlooked when calculating organic P. Existing Reviews of Organi c Phosphorus in Wetlands With the notable exception of A.F Harrsion s (1987) review of soil organic P, I am unaware of any concerted effort to collate information on levels of total organic P determined within wetlands. Reliable data sets from a limited number of highly studied wetland types do exist, including the peatla nds of North Finland (Kaila 1956), the organic wetlands of South Florida (Reddy et al. 1998) and isolated wetlands associated with certain highly studied watersheds (B ruland and Richardson 2006; Reddy et al. 1996). Inferences drawn from this limit ed number of wetland types are routinely considered typical of a wetland setti ng (Condron et al. 2005; Newman and Robinson 1999; Reddy et al. 2005). For example, Newman and Robinson (1999) state, based on experiences in Florida: 39

PAGE 40

The TP [total P] of aquatic and we tland ecosystems is often dominated by OP [organic P] which can compromise >50% of sediment TP (Reddy et al., 1998) Often presumed to constitute a sizable portion of total P, the organic fraction is rarely definitively established or reported in many we tland systems. This is in large part due to the known variance associated with the applic ation of available methods. For example, Reddy et al. (1998) report differences in t he organic P fraction of up to 13% of total P when comparing direct ex-situ sequential fractionation and indirect ex-situ determination using high temperature ashing in highly organic palustrine systems (Table 2-2). Similarly, Barbanti et al. (1994) found differ ences of up to 20% of total P (14% when adopting recommended method modifications) when comparing direct and indirect exsitu methods in organic matter-poor lagoon sediments (Table 2-3). Published Estimates of Organi c Phosphorus in Wetland Soils Although the uncertainty i nherent in estimates of to tal organic P (see above) by necessity provokes caution when compari ng between studies and sample types, it is informative to note the range of estimates made for total organic P in wetland soils. Published values purporting to represent total organic P in a range of natural and impacted wetland surface soils were colla ted from over 30 peer-reviewed journal articles. The effort sought to demonstrate the range of organic P wit hin the context of landscape and biogeochemical conditions, and the wetland was considered the unit of replication. Field replicates from a single location, sharing similar characteristics, were averaged. Samples from a single named wetl and, but which represented a significant range of conditions, i.e. over a nutrient gradient within the highly studied Water Conservation Area 2A (WCA-2A) (Vaithiy anathan and Richardson 1997) or after a degree of anthropogenic impact (C ooke et al. 1990) were r eported individually. Given 40

PAGE 41

that many references did not tabulate data, when requir ed data was extracted from graphs using the graph digitizer tool Q-plot ver:1.2.4276 (Fang Jin, 2010 www.qplot.com ). Where available, ancillary data were collected, including wetland name, location, dominant system and subsystem cl ass (Cowardin et al. 1979), depth of sampled substrate, pH and method used to estimate organic P. Given the lack of standardization, methods were split into three broad groupings; ignition-based (i.e. Aspila 1976), SEPI-sequential extraction focus ed on the determination of inorganic P, with organic P determined as a byproduct (i .e. Ruttenberg 1992), and SEPO sequential extraction schemes focused on the determinat ion of organic P (i.e. Ivanoff 1998). Estimates of total organic P were typifi ed by a high degree of variability. Of the 117 palustrine, lacustrine and riverine wetl ands listed (APPENDIX A3) estimates ranged from 0 to 94% of total P, with an overall average of 58% and a median of 64%. Estimates from individual wetland systems, similarly showed a great deal of variation. For example, values from WCA-2A, an her baceous, highly organic system and possibly one of the most highly studied wetlands on earth, ranged from 54 to 94% of total P (Koch and Reddy 1992; Vaithi yanathan and Richardson 1997). Such a comparison may be disingenuous, given the limitations of t he applied methodologies recognized by the respective authors, as well as the potential influence of anthropogenic P loading. Yet it does clearly demonstrate the issues fac ed when collating information on wetlands, historically an understudied com ponent of the landscape. Of t he 117 wetlands listed, 32 were considered lacustrine or riverine system s as defined by Cowardin et al. (1979), with the remaining 85 identified variously as moss, forested, herbaceous, or cultivated persistent emergent pal ustrine systems. 41

PAGE 42

Estimates of total organic P in lacust rine/riverine systems ranged from 0 to 73%, and within palustrine system s from between 16 and 94% (F igure 2-1). There was a significant distinction between the two groups of wetlands (Wilcoxon Mann Whitney test Chi2 = 39.4169 d.f. 1, p < 0.001) with aver age organic P contents of 37 and 67% respectively. This difference reflects a di stinction in the prevalence of organic matter between system types, itself an effect of higher productivity and accretion rates within low-energy, vegetated palustrine systems (Sahraw at 2004). Given the lack of complete ancillary data, and known issues with inter-conversion of total C, organic C, and estimates of organic matter such as loss on ignition (LOI) (Szava-Kovats 2009; Wright et al. 2008) further analysis of the suggest ed relationship between organic matter and organic P content was deemed impractical. When comparing broad analytical method groupings (Figure 2-2), a significant differ ence was seen between Ignition, SEPI and SEPO (Kruskal Wallis test Chi2 = 15.6458 d.f. 2, p < 0.0004). Yet in the context of known methodological error (Barbanti et al. 1994; Turner et al. 2005), average estimates based on Ignition and SEPO proc edures were similar, 66 and 60% respectively. In contrast, SEPI procedures were distinctly lowe r, at 46% of total P. This is expected given the a priori decision by researchers to use SEPI methods in systems that are believed to contain lower concentrations of organic P. Although generalizations as to the importance of orga nic P in wetlands can be made in a manner similar to Newman and R obinson (1999), it is clear that a high degree of variation exists within published estimates of total organic P in wetland soils. It is likely that fundamental differences in the biogeochemistry among wetland types results in the distinct variation seen. Both methodological bias and erroneous 42

PAGE 43

interpretation by researchers of operati onally defined pools are known to impact estimates, highlighting the la ck of a universally appropriate or applied standard method (McKelvie 2005; Turner et al. 2005). Total Polyphosphates in Soils The study of polyphosphates, the second co mponent of biogenic P, is motivated by the observation that rather than representing a molecular fossil, polyphosphates act as dynamic cellular component in a wide range of metabolic processes (Kornberg et al. 1999; Kulaev and Kulakovskaya 2000). This apprec iation has led to the identification of polyphosphate-accumulating organisms (PAO) within wastewater treatment systems (Zilles and Noguera 2002), soil (Ghonsika and Miller 1973) and freshwater sediments (Davelaar 1993), as well as freshwater autotrophic plankton (Eixler et al. 2005) and settling seston (Hupfer et al. 2004). The us e of synthesized polyphosphates, within detergents, inorganic fertilizers, and industrial processing has resulted in a significant anthropogenic load to the env ironment (Khan and Ansari 2005; Rashchi and Finch 2000; Sundareshwar et al. 2001). Yet ev idence of biogenic polyphosphates in unimpacted lacustrine (Ahlgren et al. 2006a; Davelaar 1993; Hupfer et al. 2007), and palustrine wetlands (Sundareshwar et al. 2 009) suggests an import ant role in a wide range of natural wetland system s and as component of geologic al P cycling (Diaz et al. 2008). In-situ Analysis of Polyphosphates Although extracellular polyphosphate granules have been identified from diatomaceous sources (Diaz et al. 2008), in-situ analysis of polyphosphate is focused on its identification and quantification within viabl e cells. Intracellular identification takes advantage of polyphosphate localization withi n either granules, prokaryotes, or 43

PAGE 44

specialized organelles, eukaryotes. Transmi ssion electron microscopy and direct light microscopy, in conjunction with vital stains, are routinely used to identify and isolate potential polyphosphate inclusions and PAO (H upfer et al. 2008; Serafim et al. 2002). Further definitive identification can be achieved through the application of immunohistochemistry (Werner et al. 2007) synchrotron based X-ray fluorescence spectromicroscopy (Brandes et al. 2007; Dia z et al. 2008), or application of the commonly used immunofluorescence stain DAPI ( 4',6-diamidino-2-phenylindole). The stain DAPI, used for direct micr obial cell visualization in so ils (Turner et al. 2003c), has been applied to the qualitative de tection of polyphosphate in autotrophic plankton (Eixler et al. 2005) and isolated microbes (Gunther et al. 2009; Hupfer et al. 2008). Recent comparison of the emission profiles of the DAPI-polyphosphate complex with other polymeric ions (lipids, DNA etc) has also allowed for more a ccurate and sensitive in vivo determination of polyphosphate concentrations (Aschar-Sobbi et al. 2008; Klauth et al. 2006). Ex-situ Analysis of Polyphosphates This approach includes procedures developed to extract and identify polyphosphate independent of the soil or PAO contained therein. Extractions have ranged from the use of perchloric acid wit h polyphosphate identification using gel filtration (Ghonsika and Miller 1973) and IR spectroscopy (Pepper et al. 1976), to alkaline extractions with identificat ion of polyphosphates by solution 31P NMR spectroscopy (Ahlgren et al. 2007; Sundar eshwar et al. 2001; Uhlmann et al. 1990). Although its application to env ironmental samples may be i nappropriate given the level of interference, it should be noted that Ramen spectroscopy has also been applied 44

PAGE 45

successfully to the study of polyphosphat es in the aqueous phase (De Jager and Heyns 1998). As more studies apply techniques such as 31P NMR spectroscopy to identify polyphosphates in wetland and aquatic system s it has become apparent that a broad range of systems contain them in significant concentrations. In addi tion to artificial sludges and conditions of excess P, they ha ve been identified in oligotrophic lake sediments (Ahlgren et al. 2006a) Carolina Ba y wetlands (Sundareshwar et al. 2009) and highly organic tundra soils (Turner et al 2004). Phosphorus NMR is providing a powerful tool in the definitive determinat ion of what appears to be an important and dynamic P pool. Total Biogenic Phosphorus in Other Ecosystem Components Water Column As with wetland soils, the analysis of biogenic P in the water column of wetlands relies heavily upon operational definitions (New man and Robinson 1999; Worsfold et al. 2005). Particulate seston can be filtered and analyzed directly fo r total biogenic P, in a manner similar to biota or soils (Selig et al. 2002), while dissolved biogenic P, operationally defined by passing through a 0.45 m filter (Worsfold et al. 2005), can be determined using either direct or indirect methods. Indirect methods use the parallel analysis for orthophosphate (or surrogates such as molybdate reactive P) and total P to calculate total biogenic P pool size by difference (McKelvie 2005; Mitchell and Baldwin 2005), while direct methods aim to determine true species of biogenic P present. For an in depth discussion of the development of techniques the reader is directed to a number of accomplished reviews (McKelvie 2005; Mitc hell and Baldwin 2005; Worsfold et al. 45

PAGE 46

2008), but a brief overview is given here of both indirect and direct methods applied to the study of biogenic P in the water column. Indirect methods include the determinati on of total P via indiscriminate atomic analysis i.e. Inductively Coupled Plasma Optical Emission Spectroscopy, ICP-OES, or by hydrolysis of biogenic P and detection of orthophosphate. Me thods of hydrolysis have included thermal/chemical oxidation using peroxydisulphate (Menzel and Corwin 1965), and magnesium nitrate (Cem bella et al. 1986), high temperature ashing followed by acidic hydrolysis (Monaghan and Ruttenber g 1999), and ultraviolet photo-oxidation (Golimowski and Golimowska 1996; McKelvie et al. 1989). Given different susceptibility of compounds to hydrolysis under different methods, inference of the nature of P pool determined should be with caution (McKe lvie 2005; Solorzano and Strickland 1968). Specifically, while polyphosphates are hydrolyzed during acidic oxidation (Monaghan and Ruttenberg 1999) they appear recalcitrant dur ing photo-oxidation, at least without the use of additional acidic oxidants (i .e. peroxydisulphate) (McKelvie 2005). Direct determination of total dissolved biogen ic P in the water column, attempts to identify biogenic P compounds by their func tionality or physical characteristics. Techniques include the selective hydrolysi s of compounds, separation due to physical size, or reactivity, and detection via spec trometric techniques. Selective chemical hydrolysis of non inositol organic P compoun ds by hypobromination (Eisenreich and Armstrong 1977), and the specific targeting by free (Herbes et al. 1975), and immobilized (Shan et al. 1993, 1994) enzymes have all been used to identify sub categories of dissolved biogenic P. The physi cal separation of P forms by gel filtration (Minear 1972; Steward and Tate 1971) or on the basis of interaction with ion exchange 46

PAGE 47

media (Minear et al. 1988; Nanny et al 1995) has also been used to separate and identify P forms. With the continued devel opment of chromatographic separation and the advent of spectrometric techniques such as 31P NMR (see below) and mass spectrometry (Cooper et al. 2005; Llewelyn et al. 2002) researchers are now able to elucidate total biogenic P, and the functional nature of P fo rms present. This not only provides information on what has proven to be a significant P pool (Mitchell and Baldwin 2005), but allows research into cycli ng of biogenic P (A mmerman and Azam 1985; Nanny and Minear 1994a; Reitzel et al. 2009) as part of benthic flux and the aquatic microbial loop (Azam et al. 1983). Biota Although total P concentrations in biota exhibit large interand intra-specific variation, it is usually assumed that the va st majority of P from senescing biota, under natural conditions, is in an organic form (Harrison 1987; Stewart and Tiessen 1987). While such assertions may be right with res pect to certain organisms, it is now known that alongside significant orthophosphate conc entrations (liberated during cell lysis), biogenic polyphosphates can represent up to 40% of identified P within certain microbial organisms (Table 2-4), thereby r epresenting a potentially signifi cant P input into wetland soils. The use of techniques such as solution 31P NMR has allowed researchers to determine the forms of P present within cu ltured microbial organisms (Bunemann et al. 2008b) and to attempt to parti tion biological sources of P (Makarov et al. 2005) within terrestrial soils. Yet the full elucidation of P within biota is st ill problematic, with the recalcitrant residual often eluding analysis. What is certain is that biogenic P inputs to soils represent a variety of both organic and inorganic forms dependent upon the nature and structure of the biotic community (Condron et al. 2005; Oberson and Joner 2005) 47

PAGE 48

as well as the interaction of that community with its abiotic envir onment (Bakken 1997; Oberson and Joner 2005). Biogenic Phosphorus Functional Groups As discussed previously, biogenic P represents both organic P and polyphosphates. Compounds can be further cla ssified by their P chemical functionality (Table 2-5). This classification has utility over operational criteria of P pools since it provides information on sources, potential chemical interact ions, the susceptibility to enzymatic hydrolysis, as well as the expe cted products of such a hydrolysis. Yet reliance upon it must be with ca ution, given the risk of tr ivializing the impact of the organic moiety or the potential for multip le functional groups. For example, both myoInositol hexakisphosphate ( myo -IP6) and its metabolic precursor glucose 6 phosphate (Raboy 2007) are phosphomonoesters. While glucose 6-phosphate represents a simple sugar phosphate, in which the hydrolysis of a phosphoester linkage would release both phosphate and an easily metabolized carbon source, myo-IP6 represents a stable inositol carbon ring, having undergone substi tution by six phosphate groups (Figure 23). The resulting compound has a high pH-depende nt charge density, making it likely to interact with both mineral and humic substances within the soil matrix (Celi and Barberis 2005a, 2007). This leads to a degree of recalcit rance, which is often invoked to explain the predominance of various IP6 isomers within the identified P compounds of upland soils (Cade-Menun 2005a; Harrison 1987; Turner et al. 2002b). The determination of coarse functional gr oupings, even without further elucidation of specific compounds, provides informa tion on potential stability and bioavailability within the environment. Listed below are the major biogenic P forms encountered within 48

PAGE 49

the environment, alongside discussion of specif ic examples of relevance to subsequent chapters. Phosphomonoesters Phosphomonoesters represent organic co mpounds linked by an ester bond to a phosphate group (Table 2-5). Th e use of phosphorylation as a mechanism of energy transfer, and signal transduction within cellula r chemistry leads to a vast array of biological molecules contai ning the phosphomonoester func tional group (Raghothama and Karthikeyan 2005; Turner et al. 2003d). This can be exemplified in the known roles that phosphorylated inositol der ivatives play within cells (Michell 2008). Lower order inositol phosphates appear to be a ubiqu itous component of eukaryotic cells, and presumed significant components of biogenic P entering wetland systems through plant and animal detritus (Weimer and Armstrong 1979). At the same time, certain stereoisomers of the higher order phosphorylated inositol s are found only at low concentrations, or within certain specialized biological elements such as seeds (Raboy 2007; Turner et al. 2002b). As such, they may only represent a small proportion of total P in the standing biomass. Yet these higher-order forms (i.e. myoIP6) have been found to represent a significant portion of identifi ed organic P within terrestrial soils (Cosgrove 1966; Murphy et al. 2009; Tur ner et al. 2003f). This discr epancy has been attributed to their preferential stabilization (Celi and Barb eris 2007), but the presence of significant concentrations of otherwise rare stereoisomers (i.e. scyllo and neo -IP6) suggests additional sources and processing of inos itol derivatives within as yet unidentified microbial components of the soil (Tur ner 2007; Turner and Richardson 2004). 49

PAGE 50

Phosphodiesters The phosphodiester bond (Table 2-5) is found within a range of biological molecules, potentially representing a significant proportion of biotic total P due its role in a number of critical polymeric mole cules and cellular membranes. Polymeric nucleotides such as RNA and DNA constitute a sizeable proportion of total cellular P (Table 2-4) and during cell senescence bec ome a component of the extracellular environment, stabilized by complex intera ctions with both mineral and organic components (Niemeyer and Gessler 2002; Ogra m et al. 1988). Specific to certain prokaryotes, polymeric teichoic acids have been inferred as present within terrestrial soils (Makarov et al. 2002a, b) and marine sediments (Ahlgren et al. 2006b), although caution is required with the potential misa ssignment of phospholipids as teichoic acids (Condron et al. 1990a; Makarov et al. 2002b). C onsisting of a backbone of polymerized glycerol and ribitol phosphate, they, along with various functional side chains, provide the major constituent for cell wall processe s within Gram-positive bacteria (Swoboda et al. 2010), as well as acting as a dynamic P reserve (Grant 1979). In addition, Pcontaining lipids (phospholipid s) represent a major constituent of all biotic cell membranes. The biosynthetic precursor to all glycolipids, phosphatidate, is a phosphomonoester, yet the majority of gl ycolipids as well as the other major phospholipid group, sphingomyelins (Figure 2-4) show additional esterification and are therefore classed as phosphodiesters. Polyphosphates Polyphosphates are molecules containi ng multiple phosphate residues bound by high energy acid anhydride bonds (Harold 1966) Adenosine triphosphate (ATP), a vital component of biological energy transfer and f ound within all biological active soils 50

PAGE 51

(Eiland 1983; Wen et al. 2005), can be cl assed as a polyphosphate (as well as a phosphomonoester). To avoid confusion, references within this dissertation to polyphosphates will be solely to inorganic compounds rangi ng in chain length from the simple two-residue pyrophosphate to linear macromolecules of many hundreds of phosphate residues (Table 2-5). Polyphosphates are found ubiquitously in both eukaryotic and prokaryotic cells (Kornberg et al. 1999). Potentially a prebiotic macromolecule (Brown and Kornberg 2004), t hey are now implicated in a range of biochemical functions from phosphate and energy storage to providing biochemical adaptation to extreme environments (Kor nberg 1995; Seufferheld et al. 2008). The biological accumulation of significant c oncentrations of polyphosphates was first identified by the isolation of metachroma tic granules in yeast cells (Liebrmann 1890 In Kornberg et al. 1999). Subsequently the identification an d isolation of so-called polyphosphate-accumulating organisms (PAO ) has been studied as part of enhanced biological P removal (EBPR) within wastewat er treatment facilities (Zilles & Noguera 2002) as well as terrestrial and aquatic envir onments in which there was a surplus of phosphate (Gachter and Meyer 1993). The imporatnce of PAO in bot h biotic and abiotic mediated P flux in lacustrine sediment s has been clearly demons trated (Gachter and Meyer 1993; Hupfer et al. 2007; Hupfer et al. 2004; Sanni grahi and Ingall 2005), but although noted as present in palustrine systems (Bedrock et al. 1994; Sundareshwar et al. 2009) their role has yet to be establis hed. Given a known presence in fungal biomass (Koukol et al. 2008) and a growi ng recognition of the role that fungal decomposition plays in wetland systems (Joergensen and Wichern 2008) it is 51

PAGE 52

conceivable that polyphosphates represent a dynamic and highly important P pool within all wetland systems. Phosphonates Phosphonates contain a direct C-P coval ent linkage and as such are inherently stable to both biotic and abiotic hydrolysis. First isolated in rumen protozoa (Horiguchi and Kandatsu 1959), phosphonates have now been described in a range of biota (Ternan et al. 1998), yet their biological ro le is still poorly understood. One compound found in relatively high concentrati ons and in a range of organisms is 2aminoethylphosphonate, commonl y found as a phosphonolipid wherein it acts in place of its analogue ethanolamine phos phate (Figure 1-4) (Ternan et al. 1998). Given their resilience, phosphonates have been employed as metabolic disrupters by both microorganisms, as toxic seco ndary metabolites, and humans in such products as the widely adopted herbicide glyphosate (N-(phos phonomethyl) glycine). The potential for such xenobiotics to persist and cause disruption in the environment has spurred research into potential degradation pathways (Ermakova et al. 2008) highlighting the potential of certain soil bacteria to utilize pho sphonates as a sole P source. This work, as well as recent research suggesting phosp honates as an important and highly active component of dissolved organic P in the mari ne water column (Martinez et al. 2010), suggests phosphonates may play an active role in P cycling of many natural systems. Application of 31P Nuclear Magnetic Resonance Spectroscopy in Wetlands Nuclear magnetic resonance spectroscopy provides researchers with a powerful analytical tool to elucidate the biogenic P functional groups found within wetlands. Although its application requires consideration of a number of factors, it can provide a non-destructive tool with the potential to determine the bonding envi ronment of every P 52

PAGE 53

atom within an environmental sample (C ade-Menun 2005b). This dissertation does not seek to give a comprehensive overview of the theoretical backgro und or application to environmental samples, for which the reader is directed to a number of comprehensive text books (Berger et al. 1997; Canet 1996) and seminal papers (Cade-Menun 2005a, b; Condron et al. 1997; Knicker and Nanny 1997). Instead, it provides a basic background to the theory requi red when considering the devel opment of te chniques and their application to wetland ecosystems. Basic Principles All nuclei spin on their axis and can be c onsidered a magnetic dipole containing a magnetic moment, expressed as Equation 2-1 where = [ h (Plancks constant) / 2 ], is the gyromagnetic ratio (a fundamental nuclear constant) and I is the vector representation of the nuclear spin I = I (2-1) When nuclei with a I 0 are placed in a static magnetic field B0 the magnetic moment aligns, and the nucleus precesses around the axis of the applied field with a Lamor frequency ( 0 ) as given by Equation 2-2. 0 = B0 (2-2) Quantum mechanics states t hat an object with aforementi oned spin has a discrete number of spin states and energy levels described by the magnetic spin quantum number mI. It follows that 2 I + 1 different spin energy levels are possible, each with an energy level as described in Equation 2-3. E = mIB0 (2-3) Therefore nuclei with I = (1H, 13C, 15N, 17O, 27Al, 31P, etc.) when placed in a magnetic field B0 have two distinct energy states (Figure 2-5). The population 53

PAGE 54

distribution of spins in the two states under standard conditions is given by the Boltzmann distribution law, and the energy difference between the two states, is equal to Equation 2-4 (Figure 2-5 A) E = mIB0 (2-4) Using Bohrs frequency law it follows that the nuclear magnetic resonance condition is as in Equation 2-5. Given quant um mechanical rules state that the only transitions allowed are those in which m = 1. We have a situation that for a spin transition between energy le vels in a magnetic field B0 an energy quantum of o is required. o = B0 mI (2-5) During NMR spectroscopy experiments, th is transitional energy is supplied by a secondary oscillating radio frequency pulse field B1 perpendicular to B0. If the frequency of the pulse satisfies the resonance condi tion of Equation 2-5 the nuclei absorb the energy and the spins move to their higher e nergy, less stable state (Figure 2-5 B). The most commonly applied Fourier Transfo rmed (FT)-NMR uses an intense B1 pulse of fixed frequency to excite all nuclei within a sample. After the radio frequency pulse, the emitted oscillating current of excited nuclei returning to their equilibrium state is recorded as a free induction decay (FID). Given the often low concentrations of 31P, sequential FIDs are collected and the sum subjected to a Fourier Transformation, transforming the time domain information to a frequency domain spectrum. After the transitional energy is removed, and the spin system is relaxing to its thermal equilibrium distribution the di ssipated energy is released through two processes, spin-spin relaxation and spin-lattice relaxation. Spin-spin is a randomized 54

PAGE 55

entropic process governed by the spin-spin relaxation time constant T2, whereas during spin-lattice relaxation the energy is released to the surrounding matrix, governed by the spin-lattice constant T1. The process of relaxation is im portant since the sample must return to its ground state to avoid saturation of the sample and achieve quantitative resonances during signal acquisition. The presence of relaxation agents either as natural paramagnetic materials with unpaired electrons (i.e. iron and manganese), or artificial transition metal complexes (e.g. ch romium acetylacetonate) provide an efficient relaxation pathway, greatly reducing T1 (Cade-Menun et al. 2002; McDowell et al. 2006; Nanny and Minear 1994a, b) and thereby a llowing for rapid repetition of pulse experiments and cumulative FID acquisition. Ensuring sample nuclei have returned to their ground state can also be achi eved through the use of a reduced B1 radio frequency pulse. Ideally, a calibrated 90 pulse is appl ied to maximize the tip angle (the angle between magnetic moment of perturbed nucleus and B0). The use of a reduced pulse length reduces the tip angle prop ortionally. While reducing the levels of energy returned on a single iteration, the nuclei achieve their ground state disproportionately faster. When considering experiments of many thous ands of scans this faster recovery time leads to shorter tota l experiment times. Nuclei Interactions Nuclei in a macroscopic samp le, held within an applied field B0 actually experience a specific local magnetic field Bloc dependent upon a nucleuss interactions with its environment. These interactions alter a nucle is perceived magnetic field and therefore modify its transition energy, 0. The change in the frequency domain spectrum can then be interpreted to provide information on both the chemical bonding and physicochemical nature found within the sample. Potential interactions include; chemical shift, spin-spin 55

PAGE 56

(scalar), dipole-dipole and quadrupole. The most readily interpreted and most pertinent to 31P NMR is that of chemical shift intera ctions. Chemically nonequivalent nuclei experience different degrees of electron shielding, this leads to the nuclei precessing with different frequencies when in a fixed field. This results in an ability to separate nuclei on the basis of the atoms to which they are bound. Subsequently, the comparison of resulting spectrum with k nown compounds allows for the i dentification of both distinct functional groups (Figure 2-6) and specif ic P containing compounds (Turner et al. 2003f). It should be noted that, given t he impact of deprotonation on altering the chemical shift interactions, many re sonances peaks are pHdependent (McDowell and Stewart 2005b) and comparison with spectral libraries should be carried out at a standard pH. In addition, given Lamor fr equencies are dependent upon the size of the magnetic field being applied and magnet size is continu ously evolving, resonance frequencies are routinely report ed in relation to an arbitrary standard. In solution 31P NMR this standard is usually 85% H3PO4 set as 0, with chemical shifts ( ) reported as shift in frequency of the samples vs from this reference vrf (Equation 2-6), normally in the units of parts per million (ppm) (Wilson 19 87). Therefore, when comparing spectral libraries to identify P compounds, both the pH of the matrix and referencing applied should be noted. = [(vs vrf)/ vrf] x 106 (2-6) Spin-spin coupling is a phenomenon brought about by the interaction of the target nuclei, in our case 31P, with bonded nuclei that also have a half integer nuclear spin (i.e. the ubiquitous 1H). Although useful in garnering secondary and tertiary structural information in advanced analytical chemistry i.e. identifying structural 56

PAGE 57

stereoisomers of higher order inositol poly phosphates (Murthy 2007) the complex signal patterns that result from 31P 1H coupling are often hard to interpret within environmental samples, ther efore, broadband heteronuclear (p roton) decoupling is often applied. During decoupling, a secondary radio frequency pulse is applied to saturate the 1H nuclei. The rapid interconvertion of 1H spin states results in a sample average thereby negating the 1H influence on the primary nuclei, 31P. In any molecule, electron distribution is anisotropic. In solutions, rapid molecular movement averages these distribution diffe rences and the effective magnetic field experienced by a nucleus is as discussed. In solids, or very viscous liquids, molecular movement is slowed, leading to nuclei ex periencing different localized fields and therefore a range of Lamor frequencies leading to a br oadening of the signal, and poor spectral resolution (Cade-Menun 2005b). To avoid line broadening most environmental studies apply an initial extract to isolat e desirable components for identification and use 31P NMR in the solution phase. Within wetlands, solid-state 31P NMR has been applied in the analysis of calcareous marshes (Del gado et al. 2000), peat soils (Shand et al. 1999) and de-mineralized marine sediment s (Sannigrahi and Ingall 2005). Yet methodological issues that exist with achievi ng high spectral resolution (Conte et al. 2008) have so far limited its widespread application. It should also be noted that for similar reasons the concentrati on of solutes during solution 31P NMR spectroscopy will impact line broadening (see Exper imental Considerations for Application to Soils). Range of Applications in Wetlands As of Aug 2010, over 50 papers have been published utilizing 31P NMR in the study of biogenic P within wetland and aquatic systems (Table 2-6). This tally draws from a broad range of research topics and wetland types. Predominantly focused upon 57

PAGE 58

freshwater systems, the review does include studies of biogenic P within select marine systems (for a complete review see Sannigr ahi et al. (2006)) that are considered informative to the study of P in lacustri ne and estuarine wetlands (Benitez-Nelson et al. 2004; Paytan et al. 2003). Research has been ca tegorized by its application to distinct ecosystem components, water column, biota and soil, with soil further divided into methodological papers, and those applying s tandard procedures to various wetland systems. Water column Application of 31P NMR spectroscopy to the analysis of dissolved and particulate P within the water column is problematic, giv en typically low concentrations. Researchers have applied a range of techniques to increase P abundance prior to analysis, from ultra filtration (Sannigrahi and Ingall 2005), and reverse osmosis (Nanny and Minear 1997), to lypholization (Cade-Menun et al. 2006) and mo re recently precipitation with polyaluminum chloride (Reitzel et al. 2009). No t only have forms within particulates, often including phytoplankton (see Biota) been el ucidated, but research has revealed evidence for significant dissolved biogenic P, including organic phosphomonoesters, phosphodiesters and inorganic pyrophosphate (Cade-Menun et al. 2006; Reitzel et al. 2009). In addition, dissolved biogenic P has been shown to constitute distinctly different forms from those found within particulates (S annigrahi et al. 2006). Taken in conjunction with evidence for the importance of dissolved organic P in t he microbial loop of aquatic systems (Ammerman and Azam 1985; Cotner and Wetzel 1991) it is clear that 31P NMR offers a powerful tool in the study of w hat may represent an ecologically important, and highly active P pool. 58

PAGE 59

Biota A limited number of studies have applied 31P NMR spectroscopy directly to the study of wetland organisms, the most co mmon being that of free-living plankton community and polyphosphate accumulating bacte ria within settling seston and benthic sediments (Hupfer et al. 2004; Khoshmanesh et al. 2002; Reitzel et al. 2007; Reitzel et al. 2006a). Yet, it has also been applied to t he profiling of phospholipids in sediment bacteria (Bardygulanonn et al. 1995; Watts et al. 2002) and t he freshwater sponge Eunapius fragilis Laidy (Early et al. 1996). Unfort unately, there has been only limited characterization of P composition in more substantial flora and f auna, a limitation also seen in terrestrial systems. Of the few completed studies in wetlands, 31P NMR spectroscopy has been applied to the deter mination of biogenic P within macrophyte leachate released upon rewetting (Pant and R eddy 2001) and as a means of tracking in vivo acidification during cell elongation of Potamogeton pectinatus (Summers et al. 2000). Soils and detritus First applied to terrestrial soils over 30 years ago (Newman and Tate 1980), the study of biogenic P in wetland soils has been greatly enhanced by the application of 31P NMR spectroscopy. That said, it has onl y been applied to a lim ited range of wetland systems, with published work typified by a focus on lacustrine sediments in Europe (Ahlgren et al. 2005; Hupfer and Gachter 1995; Reitzel et al. 2007), and China (Bai et al. 2009; Liu et al. 2009; Zhang et al. 2009b), as well as the highly organic palustrine systems of south Florida (Robi nson et al. 1998; Turner and Newman 2005; Turner et al. 2006a). Additional wetlands investigated to some degree have included Carolina Bays (Sundareshwar et al. 2009), Australian B illabongs (Baldwin 1996), New Zealand 59

PAGE 60

streams (McDowell 2009) and Scottish Blanket Bogs (Bedrock et al. 1994), yet it is clear only a limited range of wetland types have been investigated via 31P NMR spectroscopy. Extrapolating from t hose systems to which 31P NMR has been applied, biogenic P in wetlands shows a number of distinctions from the more highl y studied terrestrial systems (see below). This dissert ation aims to explore these differences, to investigate whether trends seen to date are indicative of all wetlands, as well as to investigate mechanistic process that may explain the observed biogenic P composition. Quantity. As suggested by previous operati onal estimates of total organic P, wetlands generally have a lar ger proportion of total P in or ganic forms, as compared to terrestrial systems (Turner and Newman 2005) In addition, biogenic polyphosphates have been found to represent significant levels of total P, up to 12%, in certain lacustrine (Ahlgren et al. 2006a; Hupfer and Gachter 1995) and palustrine wetlands (Sundareshwar et al. 2009). Prevalence of phosphodiesters Studies in a range of upland soils have shown biogenic P to be dominated by phosphomonoesters (Chapuis-Lardy et al. 2001; McDowell and Stewart 2006; Turner et al. 2 003a; Turner et al. 2003e), with ratios of phosphomonoesters to phosphodiesters averaging 21.8 under various land management practices in New Zealand (Mc Dowell and Stewart 2006) and 10.4 in temperate pasture soils (Turner et al. 2003e). This is in c ontrast to studies in wetland soils which have shown a great er prevalence of phosphodiesters. Including Turner and Newman (2005) which found ratios averaging just 0.82 in the highly organic palustrine wetlands of south Florida and Zhang et al. (2009b) which found an average of 3.7 when 60

PAGE 61

comparing lacustrine sediments from 7 lakes across a range of trop hic-states. Studies within terrestrial systems have attributed a positive correlation between the proportion of P found as phosphodiesters and annual precipitation to thei r increased recalcitrance under wetter conditions (Condron et al. 1990a; Sumann et al. 1998; Tate and Newman 1982). It could be that similar mechanism s, along with a generally higher microbial biomass, and the reduced stability of redox -sensitive phosphomonoester complexes, could lead to the observed pattern in wetlands (CHAPTERS 4 and 6). Presence of Inositol hexakisphosphate A major component of biogenic P in terrestrial systems (Turner et al. 2002b), t he potential recalcitr ance of higher order inositol phosphates appears to be dependent upon both the mineralogy and physicochemical nature of studied wetlands. There is evidence to suggest that under the anaerobic conditions, prevalent in wetlands, IP6 undergoes rapid degradation (Suzumura and Kamatani 1995a, b). Such a tur nover of significant terrigenous inputs of IP6 supports studies that have either fail ed, or have found only limited evidence for IP6 in wetland systems (El-Rifai et al. 2008; Turner 2006; Turner and Newman 2005; Turner et al. 2006a). Yet at the same time, recent research has found significant levels of IP6 under certain, presumed anaerobic conditi ons (McDowell 2009; Turner and Weckstrm 2009; Zhang et al. 2009b). This leads to the conclusion that, if unstable under anaerobic conditions, it is anaerobosis in concert with site mineralogy, possibly due to iron redox processes (Heighton et al. 2008), that deter mine the stability of extracellular IP6. In addition, the confounding influence of site salinity on IP6 stability has been suggested (Turner and Weckstrm 2009), since increasing salinity appears to result in release of organic P from freshwater stream materials (Gardolinski et al. 2004). However, it is 61

PAGE 62

unclear if the observed trend in decreasing IP6 concentrations with increased salinity (Suzumura and Kamatani 1995b) represents a difference in st ability or distance from terrigenous inputs. Experimental Considerations for Application to Soils For reasons established above (see Nuclei interactions) the vast majority of studies into biogenic P using 31P NMR spectroscopy do so in solution (Table 2-7). This requires the extraction of P within a suit able medium. Ideally, th e extraction process would lead to the recovery of all P forms while minimizing altera tion. Certain studies focused on the lipid profile of ecosyst em components have applied organic solvents (Bardygulanonn et al. 1995; Watts et al. 2002) yet such targeting of a specific sub group of biogenic P is unusual. The ability of 31P NMR spectroscopy to distinguish diverse chemical forms leads to its common application in more generalized extracts (Cade-Menun 2005b; Cade-Menun et al. 2002). The use of 31P NMR in characterizing sequential extractions of a sample may yiel d useful information on the functional nature of P recovered by operational procedures (Baldwin 1996; Turner et al. 2006b), but it also encounters problems with low sample concentrations and stepwise modification of P forms. More commonly, a single alkaline ex traction step is applied. First developed in terrestrial systems (Bowman and Moir 1993; Newman and Tate 1980), it was designed to recover both organic P, and P held with in organ-metal complexes (Cade-Menun and Preston 1996; Turner et al. 2005). Alkaline extractions have often been used in concert with the metal chelators ethylenediaminetetraacetic acid (EDTA) and Chelex to improve recovery of biogenic P (Ahlgren et al. 2007; Cade-Menun 2005b; Turner et al. 2005). Yet it should be noted they differentially impact subsequent NMR analysis. Ethylenediaminetetraacetic acid 62

PAGE 63

retains any chelated paramagnetics in solution (e xcept at very high pH values where co precipitation may occur (Turner 2004)) which although reducing spin-lattice T1 constants and leading to more rapid nuclei relaxation may also lead to undesired line broadening (McDowell et al. 2006). In comparis on, the use of the solid Chelex resin removes paramagnetics from the solution, but may also remove certain biogenic P forms (CadeMenun and Preston 1996). A desire to reduce the impact of sample associated paramagnetics has also led to researchers applying pre-extraction steps (Table 2-7). These have ranged from the use of an initial mineral acid (Sannigrahi and Ingall 2005; Turner and Weckstrm 2009), or metal chelator, e.g. EDTA (Ahlgren et al. 2007; Hupfer and Gachter 1995; Khoshmanesh et al. 2002) to the adoption of the reducing agent, dithionite (DeGroot and Golterman 1990) pot entially in concert with metal chelation (Carman et al. 2000; McDowell and St ewart 2005a) to both reduce and remove paramagnetic Fe3+. In addition, the treatment of final extrac ts with ion exchange media (Pant et al. 2002; Robinson et al. 1998), reducing agents (Ahlgr en et al. 2005; Reitzel et al. 2007; Zhang et al. 2009b) and organic precip itants (Ding et al. 2010) have all been employed to reduce the impact of paramagnetics and complexing humic substances on spectrum acquisition. In addition to removi ng certain biogenic P forms (Ahlgren et al. 2007; Cade-Menun 2005b) and altering T1 relaxation constants (Cade-Menun et al. 2002; Ding et al. 2010; McDowell et al. 2006), the use of treatments to reduce paramagnetics may lead to direct modification of certain functional forms. For example, within the study of benthic s ediments the use of sample pre-extraction to remove paramagnetics is routine (Ahlgren et al. 2006a; Reitzel et al. 2007), yet polyphosphates known to be stable under alkaline extraction conditions are catalytically hydrolyzed by 63

PAGE 64

the presence of divalent cations (Harold 196 6). The removal of such cations in a preextraction stage may lead directly to the pr eservation of polyphosphates, which would otherwise be lost from the acquired spectrum. Although precise, Fourier transformed NM R spectroscopy is not a sensitive technique. The observable nuclei 31P is 100 % abundant in nature yet its direct analysis at environmental concentrations of soil extracts would often require unfeasibly long run times. As a result, a process for concentrating samples maybe required. Two main methods are commonly applied, rotary evaporation and lyophilizati on (Table 2-7). Given known matrix hydrolysis of biogenic P (Turner et al. 2003d), snap freezing (-80C ) to avoid excessive alkaline conditions during crystallization and lyophilization has often been suggested as the best method (see APPE NDIX A4). The influence of rotary evaporation on the speciation of biogenic P during the sample concentration step is currently not known. Conclusions The majority of studies purporting to quant ify organic P, consid ered the largest proportion of biogenic P, are based upon inte rpretation of operationally defined procedures. While giving an indication as to the potential import ant role organic and biogenic P may play in wetlands, such appr oaches are limited. The application of advanced techniques, such as 31P NMR spectroscopy, allo w for a more detailed and complete analysis of functional biogeni c P groups within various ecosystem components of wetlands. Those studies that have used 31P NMR to explore the functional nature of biogenic P in wetland soils have highlighted a biogen ic P signature fundamentally different from the more highly studied terrestrial systems, including a prevalence of phosphodiesters, 64

PAGE 65

65 and a complex site-specific interaction of specific P compounds including IP6. The potential for a significant and dynamic pol yphosphate pool has also been found, with accurate quantification in highly studied lacustrine systems, and at least an indication of its presence in palustrine system s. While not unique to wetl ands (Turner et al. 2004) it appears to represent a more substantial pool in wetland systems (Hupfer and Gachter 1995). This dissertation seeks to develop our understanding of bio genic P by expanding the range of palustrine we tlands studied with solution 31P NMR spectroscopy. In addition to gaining basic information on the functional forms present within a greater range of wetlands, this dissertation will advance our understanding of the mechanistic drivers that determine bi ogenic speciation and tur nover within wetland soils.

PAGE 66

Table 2-1. Hierarchical classification of methods used in the study of organic phosphorus in soils and sediments Method Example Reference In-situ analysis Solid state 31P NMR (Shand et al. 1999) Scanning/Transmission Electron Microscopy X-ray Adsorbtion Near Edge Structure (XANES) (Harris and White 2008) Ex-situ analysis Direct determination of organic P APPENDIX A1 Single step Colorimetric determination pre and post oxidation (Potter and Benton 1916) 31P NMR Spectroscopy (Newman and Tate 1980) Chromatographic separation (Gilbin et al. 2000) Enzymatic hydrolysis (Pant and Warman 2000) Sequential extraction Acid pretreatment, organic solvent (Halstead et al. 1966) Acid, Alkali extraction (Mehta et al. 1954) Salt, Alkali extraction (Ghani 1942) Salt, Acid, Alkali extraction (Hedley et al. 1982) Acid, Alkali extraction under reducing conditions (Sommers et al. 1970) Indirect determination of organic P APPENDIX A2 Single step Acidic extraction pre and post oxidation of organic matter Chemical oxidation Peterson 1911 In (Pearson 1940) Low temperature ignition (Legg and Black 1955) High temperature ignition (Saunders and Williams 1955) (Aspila et al. 1976) Sequential extraction Salt, Acid, Alkali extraction under reducing conditions (Ruttenberg 1992) Chelating extract of inorganic P under reducing conditions (De Groot and Golterman 1990) 66

PAGE 67

Table 2-2. Soil biogeochemical characteristics and estimates of total organic phosphorus as determined in surface (0-10 cm) soils from a range of wetland units in South Flori da (Reddy et al. 1998). Values represent averages 1 standard deviation Aspila (1976) Sequential frac. Hydrological unit within Everglades pH TC (mg g-1) TN (mg g-1) TP (g g-1) n Organic P (% total P) n Organic P (% total P) Difference (% total P) Water Conservation Area 1 5.8 0.1 440 5 30 0.6 544 41 90 69 9 82 13 Water Conservation Area 2 7.2 0.03 420 5 28 0.5 685 42 96 67 17 77 10 Water Conservation Area 3 6.7 0.03 410 6 29 0.5 457 14 188 64 11 74 9 Holey Land Wildlife Management Area 7.5 0.1 400 9 25 0.64 543 53 36 45 44 54 9 Everglades Agriculture Area nr nr nr 654 77 nr 7 71 = organic P estimated by parallel extraction with 1 M HCl before and after ignition at 550C = sequential fractionation 1 M KCl. 0.1 M NaOH, 0.5 M HCl nr = not reported Table 2-3. Estimates of total organic phosphorus as determi ned in three surface sediment samples from a lagoon on the Po river delta, Saca di Goro Italy (Barbanti et al. 1994). Organic P (% total P) Lagoon sediments Saca di Goro, Itlay pH Organic C (mg g-1) CaCO3 (mg g-1) Total P (g g-1) AAC GBMgCl2wash GBMgCl2wash corrected SEDEX G1 7.8 19.5 171 835 30.3 48.1 41.9 30.5 G8 7.6 24.7 170 1121 25.0 41.3 38.4 24.5 G10 7.5 21.5 170 1193 21.5 41.5 35.4 29.6 = organic P estimated by parallel extraction with 1 M HCl before and after ignition at 550 OC (Aspila 1976) = sequential fractionation based on modified Golterman and Booman (1988) using MgCl2 washes to minimize resorption = corrected for presence of si gnificant apatite concentrations = sequential fractionation based on Ruttenberg et al (1992) 67

PAGE 68

Table 2-4. Relative distribution of biogenic phosphorus withi n living biota. Source Nucleic acids Phospholipids Phosphomonoesters Polyphosphates (Magid et al. 1996) Escherichia coli 65 15 20 Fungi 58 20 22 Spirodella 60 30 10 Nicotina 52 23 21 (Makarov et al. 2005) Plants 45 5 47 0 Fungi 4.5 1.4 19 44 Bacteria 25 13 34 1 (Koukol et al. 2008) Mycorrhizal fungi 30 5 29 13 Saprotrophic fungi 33 0.7 22 12 (Bunemann et al. 2008c) Phosphodiesters Bacteria 18 65 2.8 Fungi 2 30 28 = % of organic P determined = % of P determined by alkaline extraction and solution 31P NMR 68

PAGE 69

69 Table 2 Functio n Phosph i Phosph i Phosph i Phosph i Phosph i Phosph o Phosph a Phosph o 5. A sele c example dissertat for akyl o n al group i ne i ne oxide i nite i nate i tes o nate a te o ramidate c tion of fun c s of comp o ion. If appr o o r aryl grouGe n c tional grou o unds cont a o priate onl y p. ps based u a ining thes e y single ca n u pon phos p e groups m n onical str u p horus, and entioned i n u cture depi c n eric structur e e oxi d (-III (-I) (-I) (+I) (+II (+II (+ V (+ V d ation state ) I) I) V ) V ) select n this c ted, R sta n n ds

PAGE 70

70 Table 2 Generic upon p h Phosph o Phosph o Table 2 Phosph o Phosph a Polyph o Ph o Or g 2 -5. Continu e compound c h osphorus fu n o monoesters o diesters 5. Continue d o nates a mide o sphates o sphoanhydr i g anic polyph o e d lassification u n ctional grou p d de o sphates u sed in this d i p set out abo v i ssertation b a v e a sed Ex a A d e Eth -g my o Ph o De o Rib LLA d e mo n 2A N-( P (Gl y Ph o a mple compo e nosine 5 m o anolamine p h lucose 1-pho o -Inositol he x o sphoenolpy r o xyribonuclei c onucleic acid Phosphatid y Phosphatid y e nosine 3,5n ophosphate A minoethyl ph P hosphonom y phosate) o sphocreatin e Pyrophos p Polyphos p A denosin e A denosin e unds o nophosphat e h osphate sphate x akisphospha t r uvate c acid (DNA) (RNA) y l choline y l ethanolami n cyclic osphonic aci d ethyl)glycine e p hate (n=0) p hate (n>1) e 5 diphosph a e 5 triphosph a e (AMP) t e n e d a te, ADP a te ATP

PAGE 71

Table 2-6. Studies employing 31P nuclear magnetic resonance spectrosc opy in wetland and aquatic systems Reference Focus WetlandLocation Water Column (Knicker and Nanny 1997; Nanny and Minear 1994a) Overview of considerations when usin g NMR to study dissolved P in the water column Lacustrine IL, USA (Nanny and Minear 1994b) Use of Lanthanide shift re agents to improve resolution when monitoring dissolved organic P in solution Lacustrine IL, USA (Nanny and Minear 1997) Use of size fractionati on, bromination and Lanthanide shift reagents to elucidate further structure Lacustrine, Riverine IL, USA (Clark et al. 1999) Solid state analysis of high molecular weight dissolved P shows high levels of phosphoesters and phosphonates Marine Pacific (Kolowith et al. 2001) Solid state analysis of marine ultra filtered waters. Showing high proportion of phosphoesters and phosphonates Marine Various (Selig et al. 2002) Nature of dissolved and pa rticulate P in lake water Lacustrine Germany (Benitez-Nelson et al. 2004) Particulate inorganic and organic P in oxic and anoxic marine water column Marine Pacific (Paytan et al. 2003) Evidence of heterogeneity in organic and inorganic biogenic P within seston Marine Global (Cade-Menun et al. 2005) Storage and pretreatment of marine water colu mn particulates Marine CA, USA (Cade-Menun et al. 2006) Using solution 31 P NMR to characterization of dissolved and particulate P within river and flood plain waters. Riverine, Lacustrine SC, USA (Sannigrahi et al. 2006) Distinct difference between dissolved and particulate P Marine HI,USA (Reitzel et al. 2009) Precipitation of soluble P from lake water using poly aluminum chloride. Subsequent identification of phosphomonoesters, polyphosphates and DNA using alkaline resuspension. Lacustrine Denmark Biota (Bardygulanonn et al. 1995) Profiling of phospho lipids in lake sediments Lacustrine WI, USA (Early et al. 1996) Characterization of pho spholipids within membranes of the sponge Eunapius fragilis Leidy Lacustrine Lake Michigan, USA (Summers et al. 2000) In vivo determination of P as evidence of cellular acidification during cell elongation of Potamogeton pectinatus NA (Pant and Reddy 2001) Impact of hydrology on water sol uble lechate from macrophytes Palustrine FL, USA (Khoshmanesh et al. 2002) Luxury uptake of P resu lting in polyphosphate accumulation in benthic bacteria Palustrine Australia (Watts et al. 2002) Profiling of phospholipids fr om microbial biomass present within stream sediments Riverine Ontario Canada 71

PAGE 72

Table 2-6. Continued (Biota) Reference Focus Location (Hupfer et al. 2004) Polyphosphates detected in settling lake seston Lacustrine Europe (Reitzel et al. 2006a) Evidence of DNA, pyrophosp hate and phospholipids in settling seston Lacustrine Denmark (Reitzel et al. 2007) Presence of various biogenic P forms within plankton and settling seston Lacustrine Sweden Soil (methodological papers) (Shand et al. 1999) Potential for use of solid state 31 P NMR in peat soils Palustrine Scotland (Delgado et al. 2000) Solid state analysis of calcareous marshes Palustrine Spain (McDowell and Stewart 2005a) Use of a Ca-EDTA dithonite pre-extraction step to reduce line broadening in samples with high paramagnetics Riverine New Zealand (Turner et al. 2006b) Evidence that orthophosphate may complex with organic molecules preventing detection by standard molybd ate colorimetry, potentially leading to an overestimation of organic P. Palustrine FL, USA (Ahlgren et al. 2007) Comparison of NaOH and NaOH + EDTA extraction using bicarbonate buffered dithionite or EDTA as a pre-extraction step. Lacustrine Sweden (Turner et al. 2007b) Comparison of sample handling procedures prior to extraction and identification of P forms Palustrine FL, USA (El-Rifai et al. 2008) Parallel analysis with mass spectroscopy Palustrine FL, USA (Turner and Weckstrm 2009) Use of phytate within brackish sediments as a paleo-indicator Marine, Riverine Denmark Soil (Bedrock et al. 1994) Forms of P present in blanket peat under different management and vegetation regimes. Presence of ph osphonates and polyphosphates taken as evidence of microbial activity Palustrine Scotland (Hupfer and Gachter 1995) Detection of polyphosphate as an important transient sink within benthic sediments. Lacustrine Switzerland (Baldwin 1996) NMR analysis coupled to modified SEDEX (Ruttenberg 1992) sequential extraction scheme. Lacustrine Australia (Robinson et al. 1998) NMR analysis coupled to pre-extracti on of labile P in high organic soils Palustrine FL, USA (Carman et al. 2000) Samples from various La kes and the Benthic sea analyzed. Oxic/anoxic conditions and presence of various cations suggested as source of variability. Lacustrine, Marine Sweden (Sundareshwar et al. 2001) Pyrophosphate accumulation associated with anthropogenic impact of coastal systems Marine SC, USA 72

PAGE 73

73 Table 2-6. Continued (Soil) Reference Focus Location (Pant et al. 2002) Analysis of P forms within surface sediments within a submerged aquatic vegetation SAV treatment wetland Palustrine FL, USA (Hupfer et al. 2004) Origin and diagenesis of polyphosphates in lake across various trophic states Lacustrine Europe (Ahlgren et al. 2005) Attenuation of P forms with depth in sediments, half-life times estimated for pyrophosphate and organic forms Lacustrine Sweden (Turner and Newman 2005) The importance of phosphodiesters as a P pool in calcareous wetland systems Palustrine FL, USA (Sannigrahi and Ingall 2005) Influence of oxic/anoxic bottom waters in determining stability of P forms, esp. polyphosphates. Using solid-state 31P NMR on demineralized sediments Marine Canada (Ahlgren et al. 2006a) Analysis of three oligotroph ic lakes showing high variability in the presence of polyphosphate. Lacustrine Sweden (Ahlgren et al. 2006b) Degradation and half-life time estimates for various biogenic P compounds in marine benthic sediments Marine Sweden (Reitzel et al. 2006a) Determination within lake sediment s from across arrange of trophi c states Lacustrine Denmark (Reitzel et al. 2006b) Analysis of changes in identified groups with time in the sediment and with addition of Al as a lake management strategy Lacustrine Denmark (Turner 2006) Analysis of soils under rice cultivat ion, including flood irrigation Palustrine Madagascar (Turner et al. 2006a) Identified biogenic forms wi thin treatment wetlands shown to be dominated by phosphodiesters. Palustrine FL, USA (Reitzel et al. 2007) Rapid degradation of polyp hosphates with depth and recalcitrance of other biogenic P forms Lacustrine Sweden (Bai et al. 2009) Presence of organic P in a eutrophic lake Lacustrine China (Sundareshwar et al. 2009) Diversity of P forms used as a measure of ecosystem function Palustrine NC & SD USA (Zhang et al. 2009b) Surficial sediments from 7 shallow lakes demonstrating various trophic status. No evidence of polyphosphates, IP6 identified Lacustrine China (McDowell 2009) Changes in stream sediment P forms as a result of surrounding land use change Riverine New Zeland (Liu et al. 2009) Dominance of orthophos phate and phophomonoesters in heavily eutrophic lake systems Lacustrine China = Dominant system type as desi gnated by Cowardin et al. (1979)

PAGE 74

Table 2-7. Methodological details of studies employing 31P nuclear magnetic resonance spec troscopy in wetland soils Reference Vegetation Extraction method Paramagnetic treatment Concentration Acquisition Pulse Pre extraction Post extraction Pulse width Delay (s) Soil(Methodological) (Shand et al. 1999) PE-moss Solid state Acetylacetone na Lyophilization CPMAS, HP (90) 4 (Delgado et al. 2000) Herb Solid State CBD na 5 us 10 (McDowell and Stewart 2005a) BS-S 0.25 M NaOH 50 mM EDTA EDTA-Dithonite na Lypholization 45 5 (Turner et al. 2006b) PE-Herb,SA 0.5 M NaOH 0.5 M NaHCO3 1 M HCl (Ivanoff et al 1998) na Lypholization 45 2 (Ahlgren et al. 2007) SA, BS 0.1 M NaOH0.125 M NaOH 0.25 M EDTA BD or EDTA na Rotary evap 63 1.2 (Turner et al. 2007b) PE-Herb,SA 0.5M NaOH 0.5M NaHCO3 1 M HCl (Ivanoff et al 1998) na Lypholization 45 2 (El-Rifai et al. 2008) PE-Herb. 0.25 M NaOH 50mM EDTA na na Lyphilization 45 (Turner and Weckstrm 2009) BS 0.25 M NaOH 50mM EDTA HCl na Lypholization 45 2 Soil (Bedrock et al. 1994) PE-Moss 0.5 M NaOH na na Rotary evap. 90 0.2 (Hupfer and Gachter 1995) BS 0.2 M NaOH + 67 mM EDTA EDTA na Rotary evap 10 (Baldwin 1996) BS SEDEX sequential (Ruttenberg et al 1992) na na (Robinson et al. 1998) PE-Herb. 0.25 M 50 mM EDTA (85 C) na Chelex X-100 Column Lypholization 1.5 (Carman et al. 2000) BS 0.5 M NaOH CDB,MgCl2 wash na Rotary evap. 90 1 (Sundareshwar et al. 2001) PE-Herb 0.5 M NaOH 100 mM EDTA na na na 45 2.1 (Pant et al. 2002) SA 0.4 M NaOH na G-25 Sephadex Rotary evap. 90 5 (Hupfer et al. 2004) BS 0.2 M NaOH 67 mM EDTA EDTA na Rotary evap. 2 (Ahlgren et al. 2005) BS 0.1 M NaOH Dithonite Rotary evap. 72 0.2 (Sannigrahi and Ingall 2005) BS Solid state HCl/HF na na CP-MAS 4 (Turner and Newman 2005) PE-Herb. 0.25M NaOH, 50mM EDTA na na Lypholization 45 1 (Ahlgren et al. 2006a) BS 0.125 M NaOH + 0.25 mM EDTA BD na Rotary evap. 63 1.2 74

PAGE 75

75 Table 2-7. Continued Reference Vegetation Extraction method Paramagnetic treatment Concentration Acquisition Pulse Pre extraction Post extraction Pulse width Delay (s) (Ahlgren et al. 2006b) BS 0.25 M NaOH 50mM EDTA Dithonite rotary evap 63 1.25 (Reitzel et al. 2006a) BS 0.11M BD + 0.1 M NAOH na na Rotary evap. 63 1.2 (Reitzel et al. 2006b) BS 0.125 M NaOH + 25 mM EDTA na BD Rotary evap. 63 1.2 (Turner 2006) PE-CS 0.25 M NaOH 50mM EDTA na na Lypholization 45 2 (Turner et al. 2006a) PE-Herb 0.25M NaOH, 50mM EDTA na na Lypholization 45 2 (Reitzel et al. 2007) BS 0.1M NaOH BD BD Rotary Evap 63 1.2 (Bai et al. 2009) BS 0.1 M NaOH EDTA, Dithonite na Rotary evap 90 2 (Liu et al. 2009) BS 0.25 M NaOH 50mM EDTA na na Lypholization 45 2 (McDowell 2009) BS 0.25 M NaOH 50 mM EDTA na na Lyophilization 45 4 (Sundareshwar et al. 2009) PE-Herb/ CS 0.25M NaOH 0.1M EDTA na na Lypholization (Zhang et al. 2009b) BS 0.25M NaOH na BD Lypholization 90 4 vegetation; PE = persistent emergent moss = moss peat, Herb = Herbaceous, CS = flooded cultivated soil BS = Benthic sedim ents, S = stream, SA= submerged aquatic, extraction method; single step extraction or steps used within 31P NMR studies Paramagnetic treatment; method applied to minimize effect of paramagnetics, pre or post extrac tion; (EDTA)= Ethylenediaminetetraacetic acid, (CDB)= Citrate+ Dithonite+ Bicarbonat e, (BD)= Bicarbonate + Dithonite, Concentration; method used to concentrate sample prior to 31P NMR analysis, Acquisition parameters used in 31P NMR analysis na= not applicable =not noted by reference

PAGE 76

Figure 2-1. Frequency hist ogram of estimated organi c P within 117 wetlands (32 lacustrine, 85 palustrine) 76

PAGE 77

Figure 2-2. Frequency histogram of estimated total or ganic P within 117 wetlands broken down by general me thod grouping. IGNITION = parallel extraction of inorganic P pre and post oxidation of or ganic matter, organic P estimated by difference, SEPI = sequential extraction procedures focused on characterizing inorganic P, SEPO = sequential extraction procedure focused on characterizing organic P. (I QR = Inter Quartile Range) 77

PAGE 78

Figure 2-3. Struct ural comparison of myo -Inositol hexakisphosphate and -d-glucopyranose 6-phosphate (pyranose ring form of -d-glucose 6-phosphate ). Ortho-P represents orthophosphate group with various levels of deprotonation dependent upon environmental pH. Based upon (Barrientos and Murthy 1996) 78

PAGE 79

Figure 2-4. Basic phospholipid compound structure found within both eukaryotic and prokaryotic cell membranes. Phosphodiester containing phosphoglycerides consitute phosphatide after additional este rification, by alcohol functional groups. R = fatty acid chain. 79

PAGE 80

Figure 2-5. Response of phosphoru s nuclei to an applied magnetic field. A) Zeeman splitting of a mI = system, B) graphical representation of precessional orbit of 31P nuclear magnetic dipole around the applied magnetic field with transition due to applied B1 radio frequency pulse. Adapted from (Cade-Menun 2005b; Knicker and Nanny 1997). 80

PAGE 81

Figure 2-6. Solution 31P nuclear magnetic resonance spectra showing co mmon functional groups. Sample represents surface soil from a Carolina Bay, SC USA extracted using standard procedure (0.25 mol L-1 NaOH 50 mmol L-1 EDTA) and concentration via lypholization. Spectra were acquired using a Bruker Avance 500 Console with a Magnex 11.75 T/54 mm magnet using a 10 mm BBO Probe at a stabilized 25 C with a calibrated (~30) pulse length, a zgig pulse program, and a 2 s pulse delay. 81

PAGE 82

CHAPTER 3 INTERACTION OF BIOGENIC PHO SPHORUS WITH ANION EXCHANGE MEMBRANES: IMPLICATION FOR SOIL PHOSPHORUS ANALYSIS1 Introduction Anion exchange membranes are commonly employed to study P dynamics in soils and sediments (Myers et al. 2005; Saunder s 1964; Skogley and Dobermann 1996). In particular, they are used to measure readily exchangeable phosphate in soils (Sibbesen 1978) and, in conjunction with hex anol fumigation, to determine P contained within the soil microbial biomass (Kouno et al. 1995; Myers et al. 1999). Anion exchange membranes offer potentia l benefits over conventional soil P tests, given their action as passive ion sinks analogous to biological uptake (Qi an and Schoenau 2002). Their use in the field can provide an integr ated measure of P availability (Cooperband and Logan 1994; Drohan et al. 2005; Meason and Idol 2008), and has led to the development of commercial products designed to determine nutrient supply rates (e.g., PRS probes; Western Ag development Innovations, Saskatoon, Saskatchwan, Canada). Despite numerous assessment s of the practical influe nce of experimental design on both batch and field-deployed membranes (Mason et al. 2008; Qian and Schoenau 1997, 2002; Qian et al. 1992; Sato and Comerf ord 2006), surprisingly few studies have examined their interaction with organic and condensed inorganic P compounds. Yet there is a clear potential for sorption and recovery of organic P to both resin beads and membranes (Cooperband et al. 1999; McDowell et al. 2008; Rubaek and Sibbesen 1993). Of particular note is the ~100% recovery of both phytic acid ( myo -Inositol 1 Accepted for publication in a modified format, Soil Science Society of America Journal 2010 82

PAGE 83

hexakisphosphate; IP6) and glucose 6-phosphate by membranes, type 204-U-386 (Ionics Inc Watertown, MA) when exposed at 16 g P cm-2 and loaded with a chloride counterion (Cooperba nd et al. 1999). Consideration of potentia l interactions between anion exchange membranes and organic and condensed inorganic P compounds is important given the assumptions made during routine application of the exchange media to st udy P dynamics. First, if significant levels of organic and condensed inorganic P are recovered by anionexchange membranes and transfe rred to the eluant soluti on, their inclusion in subsequent measurements will lead to an overes timation of bioavailable or readilyexchangeable orthophosphate. Second, when anion exchange membranes ar e employed as an ion sink during the measurement of microbial P (Kouno et al. 1995; Myers et al. 1999) the analysis of total P in the eluant is typically omitted, because the difference between orthophosphate and total P is assumed to be negligible (B rookes et al. 1984), as in fumigation extraction procedures (Brookes et al. 1982; Hedley et al. 1982). However, if organic or condensed inorganic P compounds released from lysed microbial cells are adsorbed directly to anion exchange membranes wit hout conversion to orthophosphate by enzymatic or matrix induced hydrolysis, this may lead to an under-estimation of fumigation-released (microbial) P. Lastly, if organic or condensed inorgani c P are recovered by anion exchange membranes, both the nature of the eluant used for P recovery and the method of P detection may influence levels of P determi ned as recovered from a sample. The use of non-selective elemental analysis such as ICPOES (inductively-coupled plasma 83

PAGE 84

optical emission spectrometry ), or the hydrolysis of P containing compounds during elution by mineral acids may erroneously attribute complex forms of P as bioavailable. The aims of this study were to estab lish (i) the potential interactions between a commonly used anion exchange membrane and a range of m odel organic and condensed inorganic P compounds, and (ii) the implications of any interaction during the routine application of anion exchange membrane proc edures to wetland soils expected to contain high levels of soluble organic and condensed inorganic P. Methods After establishing the exch ange capacity for orthophosphate, anion exchange membrane strips were exposed to a r ange of standard P-contai ning compounds to determine the levels of relative ion exc hange and compound recovery. Eluant solutions were analyzed for both molybdate-reactive P and total P, while solution 31P nuclear magnetic resonance (NMR) spectroscopy was used to investigate the stability of P compounds during their exposure to the ani on exchange membrane strips. Finally, a series of anion exchange membrane extractions from soils were analyzed for both total and molybdate-reactive P to determine t he potential for previously-established interactions to impact P re covery in field samples. Anion Exchange Membranes The anion exchange membrane characteri zed in this study (BDH Prolabo Product number: 551642S) is used routinely in field and laboratory procedures (McLaughlin et al. 1994; Myers et al. 2005; Roboredo and Coutinho 2006; Turner and Romero 2009b). Available through VWR Internat ional, UK (Lutterworth, Leicestershire, UK) and CTL Scientific Supply Corp. (Deer Pa rk, NY, USA), the membranes (previously sold directly under the BDH brand name) use a polystyrene / divinylbenzene copolymer 84

PAGE 85

base doped with quaternary ammonium as t he ionogenic group and are supplied in sheets 125 x 125 mm preloaded wit h chloride counterions. Anion exchange membranes were cut into 1.5 x 6.25 cm strips, yielding a reactive area of 18.75 cm2 per strip. The strips were charged with HCO3 counter ions by shaking 25 strips in a 250 mL HDPE bottle with thr ee sequential changes of 200 mL 0.5 mol L-1 NaHCO3 over a 24 h period. Strips were then ri nsed three times in deionized water (DI) to remove adhering solution. Bicarbonate was chosen as the counterion since it mimics biological uptake from the rhizosphere (Qian and Schoenau 2002; Si bbesen 1978) and is considered preferable when studying P due to its lower affinity for exchange sites compared to other commonly used counter ions (e.g., chloride) (Skogley and Dobermann 1996). After exposure to a test so lution, membranes were rinsed in DI, shaken dry of excess water, and immersed in a conical tube containing 50 mL of 0.25 mol L-1 H2SO4 and shaken for 3 h, after which a subsampl e of the eluant was decanted into a 20 mL scintillation vial. Membranes were cleaned us ing a secondary acidic wash (0.25 mol L-1 H2SO4, 1 h) and rinsed with multiple changes of DI before regenerat ion with bicarbonate as described above. Phosphorus recovery was unaffected by repeated use of the membrane strips (data not shown), despite some discoloration following use with soil samples. Phosphorus Determination Molybdate-reactive P, an oper ationally defined paramet er, was determined in the extracts by automated colorimetry with detec tion at 880 nm using a flow injection analyzer (Lachat Quickchem 8500, Hach Ltd, Loveland, CO). Total P was determined in the extracts by ICPOES (Optima 2100, Perkin-Elmer Inc., Shelton, CT). 85

PAGE 86

Orthophosphate standards were prepared in the sa me matrix as the samples (0.25 mol L-1 H2SO4) for both analyses. Experimental Design Anion exchange membrane exchange capacity The response of the exchange me mbranes to increased orthophosphate concentrations was tested to determine t he exchange capacity of the membranes and the range over which the dynamic anion exch ange mimics that of an infinite sink. Replicate (n=3) preloaded and rinsed membranes were placed in 250 mL HDPE bottles with 75 mL of orthophosphate st andards between 0 and 2.5 mg P and shaken for 24 h. Adsorbed P was eluted and determined as described above. Phosphorus recovery by anion exchange membrane strips A series of organic and condensed inorganic P compounds were prepared at an approximate concentration of 200 g P mL-1 in DI, with precise concentrations determined subsequently by ICPOES (T able 3-1). Duplicate preloaded and rinsed anion exchange membrane strips were loaded individually into 250 mL HDPE centrifuge bottles with 70 mL DI and 5 mL of standar d P-containing solutions. After standard exposure and elution from membranes (see above), eluants were stored in 20 mL HDPE scintillation vials at 4C until analysis by molybdate colorimetry within 72 h and total P within one month. Purity and stability of organic and condensed phosphates in deionized water Standard solutions were analyzed by solution 31P NMR spectroscopy over a period of 24 h to determine the stability of P co mpounds during exposure to anion exchange membranes. Solutions (~ 200 g P mL-1) were mixed 1:1 with an internal standard (methylenediphosphonic acid; MDP) (~ 50 g P mL-1) in the same matrix. Of the 86

PAGE 87

resulting mixture, 0.9 mL was added to 0. 1 mL deuterium oxide, the solution was vortexed, and then loaded into a 5 mm diameter NMR tube. Solution 31P spectra were acquired after 15 min and 24 h using a Bruk er Avance 500 Console, Magnex 11.75 T/54 mm magnet fitted with a 5 mm BBO probe. Acquisitions were run at a stabilized 25C with a 4.833 s (~30) pulse length and a 2 s recycle delay. An average 2000 scans (run length 1 h 21 min) were acquired and t he resulting spectra referenced against internal MDP with a chemical shift ( ) = 17.46 ppm, determined with reference to an externally held 85% H3PO4 standard ( = 0 ppm). Extraction of phosphorus compounds from wetland soils by anion exchange membranes A comparison of total and molybdate reactive P in the membrane eluants from wetland soil extractions was carried out to determine the potential for organic and condensed inorganic P recovery when anio n exchange membranes are used to measure bioavailable and microbial P. A batch process using anion exchange membranes prepared as described above was applied to 27 soils collected from a nutrient gradient in the freshwater San San Pond Sak wetland, a domed peatland in Bocas del Toro province, western Panama (Phi llips et al. 1997). Soils were all high in organic matter (total C 41%), but contained a range of to tal P concentrations (388 1028 mg P kg-1). Two fresh samples of each soil (3.5 g dr y weight equivalent) were weighed into 250 mL HDPE centrifuge bottles and sample-specific volumes of DI were added to bring the total water content to 75 mL. Both sa mples received DI and a single anion exchange membrane strip (1.5 x 6.25 c m), with one subsample also re ceiving 1 mL of hexanol (95%, Sigma Aldrich, St Louis, MO). After 24 h shaking, the anion exchange membrane 87

PAGE 88

strips were removed, rinsed, eluted, and the resulting solution analyzed for total and molybdate-reactive P as described above. A significant difference between total and molybdate-reactive P would indicate the retention of organic or condensed inorganic P compounds that were not hydrolyzed to molybdat e-reactive P during elution in 0.25 mol L-1 H2SO4. In normal application of the method, the difference in molybdate-reactive P between the fumigated and non-fumigated sample s would be attributed to fumigationreleased P. This could then be adjusted for extraction efficiency an d microbial biomass not killed by fumigation ( k factor) to provide an estimate of microbial P (Kouno et al. 1995). However, given problems with both corre ction factors, some authors prefer simply to use uncorrected fumigation-releas ed values (Bunemann et al. 2008a). For this study uncorrected total and molybdate-reacti ve P in the eluants of both unfumigated and fumigated samples were compared as g P g-1 dry weight. Results Anion Exchange Membrane Capacity Orthophosphate recovered from a non-com petitive environment was within a 5% error for the majority of the exposure levels tested in this study (Figure 3-1). Only at levels in excess of 117 g P cm-2 was recovery unacceptable, defined here as < 95%. Taking a conservative working maximum of 100 g P cm-2 of membrane and the utilized membrane area to soil ratio of 5.6, th is would equate to an orthophosphate sorption capacity (560 g P g-1) well above the maximum likely to be encountered in the natural environment. 88

PAGE 89

Organic and Condensed Phosphorus Reco very by Anion Exchange Membrane Strips Total P concentrations in eluants demons trated that most of the P compounds interacted with the anion exchange membranes (Figure 3-2). Three compounds were recovered completely (sodium pyrophosphate, glucose 6-phosphate, and adenosine 5monophosphate), while three others showed recoveries between 20 and 60% (sodium hexametaphosphate, 2-aminoethy lphosphonic acid, and phytate). The macromolecules RNA and DNA showed limited recovery (~ 10% and 0.2%, respectively), although it was unclear if this represented true ion exchange at the ionogenic sites or contamination due to physical adherence of the macromolecul e to the polymeric membrane. Under the concentrated conditions of the experimen t a sheen was observed on the membrane strips exposed to DNA. I t herefore considered that RNA and DNA recoveries were inconclusive in the context of this experiment. Of the compounds showing significant inte raction with the membranes, some were recovered as molybdate-unreacti ve P (adenosine 5-monophosphate, 2aminoethlphosphonic acid, and phytate), while large proportions of others were recovered as molybdate-reactive P (Figur e 3-2). For example, of the ~100% of pyrophosphate and glucose 6phosphate recovered by t he anion exchange membrane, 36 and 69%, respectively, were detected as molybdate-reactive P. Purity and Stability of Phospho rus Compounds in Deionized Water Comparison of the 31P NMR spectra from test compounds with a known orthophosphate standard (Figure 3-3) showed neither contam ination of the original compounds with orthophosphate, nor orthophosphate released by hydrolysis in DI over a period of 24 h The presence of molybdate -reactive P in the eluant solutions was 89

PAGE 90

therefore considered to be due entirely to acidic hydrolysis during elution of P from the membranes and subsequent storage prior to colo rimetric analysis. Hydrolysis during molybdate colorimetry (< 1 min contact ti me) was likely to be negligible given the relatively long period of el ution/storage in 0.25 mol L-1 H2SO4. Application to Wetland Soils for E xchangeable and Fumigation-Released Phosphorus Comparing the total and molybdate-reacti ve P recovered by anion exchange membranes in a standard extraction from wetl and soils, paired t-tests show significant ( p < 0.05) differences between molybdate reac tive and total P determined in both nonfumigated and fumigated samples, but av erage differences of only 3.4 and 1.3%, respectively, are within the margin of error due to calibration. I concluded that there was no evidence organic or condensed inorganic P, undetected by colorimetric analysis, was recovered in appreciable amount s by the anion exchange membranes. Discussion The orthophosphate exchange capacit y for the studied anion exchange membranes agrees well with the published in formation on the performance of similar membranes, including the linear response observed to 22.13 g P cm-2 (Schoenau and Huang 2001) and the total exchange capacity of commercially available anion exchange membrane products i.e.the P exchange capaci ty of PRS Plant Root Simulators, calculated as 301.4 g P cm-2 based upon meq charge capacity (http://www.westernag.ca/innov/technical_1.php) The calculated window of acceptable recovery appears more robust than that demons trated by Mason (2008) who, assuming that a 6.26 x 2.5 cm strip has a reactive surface of 31.3 cm2, found a linear response to only 4.6 g P cm-2. While this may be due to their use of a counter ion (chloride) with a 90

PAGE 91

higher affinity for the membr ane exchange sites, it is also due to their conservative interpretation of acceptable re covery (S. Mason, 2008, Univer sity of Adelaide, personal communication). Several organic and condensed inorganic P compounds interacted, and in some cases were fully recovered, by a routi nely used anion exchange membrane method. Of these, glucose 6-phosphate and the condens ed inorganic P compounds pyrophosphate and hexametaphosphate were reco vered partially as molybdate-reactive P. Given that orthophosphate contaminati on in the original compounds was negligible, and the compounds were stable during exposure to t he membranes, we conclude that some compounds were hydrolyzed to orthophosphate during elution from the membranes in 0.25 mol L-1 H2SO4 and storage prior to colorimetric analysis. The potential for acidic hydrolysis of organic and condensed inorganic P is a recognized source of error in the determination of orthophosphat e by the operationally-defi ned molybdate colorimetry (Dick and Tabatabai 1977; Worsfold et al. 2005). Hydrolysis during the acidic molybdate reaction (< 1 min contact time), however, was considered to be negligible, given the relatively long period of el ution/storage in 0.25 mol L-1 H2SO4. Similar studies examining the hydrolysis of soil organic P during extraction in strong acids reported variable results. For example, Ivano ff et al. (1998) reported 38% hydrolysis of para-nitropheny l phosphate during a 3 h extraction in 1 mol L-1 HCl, whereas Bowman (1989) reported negligible hydrolysis of a series of organic P compounds (para-nitrophenyl ph osphate, glycerophosphate, phytic acid, and bis-paranitrophenyl phosphate) during a short extraction in 1.08 mol L-1 H2SO4. Further analysis using non-destructive direct monitoring by 31P NMR spectroscopy would be required to 91

PAGE 92

establish the impact of pH-dependent hydrolysis on model compounds over eluant holding times. For a series of wetland soils with contrast ing P contents, no discernible molybdateunreactive P was recovered by the anion exchange membranes (Figure 3-4). Fumigated samples were expected to cont ain high concentrations of organic and condensed inorganic P from lysed microbial cells. This study demonstrates that, given the use of acidic eluant, ther e is no significant distincti on between microbial P recovered by anion exchange membranes as determined fr om total elemental P analysis (ICP OES) or operationally defined mo lybdate colorimetric analysis. This does not rule out the inclusion of acid-labile organic and condensed inorganic P within molybdate-reactive P, both of wh ich have been reported in soil solutions (Espinosa et al. 1999). Yet the absence of any discernible molybdate-unreactive P in these samples may suggest that acid-labile compounds are likely to be negligible. The use of anion exchange membranes with more benign eluants (e.g., chloride or bicarbonate) would be needed to confirm this assumption. In cases that have reported the recove ry of organic P by exchange media, an assumption was made, due to their mode of action, that resin ex tractable organic P estimated soil labile organic P (McDowell et al. 2008; Rubaek and Sibbesen 1993). Yet a previous study of the nat ure of organic P in pasture soils using phosphatase hydrolysis reported negligible concentrations of labile phosphomonoesters (including polyphosphoric compounds), presumed to i ndicate a rapid hydrolysis of these compounds following release into the soil so lution (Turner et al. 2002a). Further 92

PAGE 93

93 investigation is needed to det ermine if organic P may be recovered from soils by the anion exchange membranes tested here. In conclusion, although anion exchange me mbranes can interact with a diverse range of P forms, this does not appear to limit their applicat ion to the determination of readily-exchangeable phosphate or microbial P in soils, parti cularly for samples from natural environments. However, care must be taken in assigning P recovered by anion exchange membranes to labile orthophosphate, because acid-labile organic and condensed inorganic phosphates may also be included. This will be especially problematic in-situ ations where membranes are depl oyed after the application of condensed inorganic P fertilizers (Bertrand et al. 2006). Although not fully resolved the adaptation of the anion exchange membrane procedure to include a benign membrane eluant (e.g. chloride, bicarbonate) to mini mize hydrolysis and total P determination may allow for the parallel asse ssment of labile orthophosph ate, and organic / condensed inorganic P.

PAGE 94

Table 3-1. Phosphorus compounds te sted on anion exchange membranes. Stock solution concentrations determined by ICPOES analysis. Compound Chemical Formula Standard Concentration (g P mL-1) Potassium dihydrogen phosphate KH2PO4 100.0 Sodium hexametaphosphate (NaPO3)n.Na2O 171.1 Sodium pyrophosphate, decahydrate Na4P2O7.10H2O 207.6 D-Glucose 6-phosphate, disodium salt hydrate C6H11O9PNa2.xH2O 169.0 Adenosine 5-monophosphate, monohydrate C10H14N5O7P.H2O 157.5 2-Aminoethylphosphonic acid C2H8NPO3 195.6 Phytic acid, dodecasodium salt C6H6O24P6Na12 183.6 Ribonucleic acid, Type VI from tortula yeast -:177.5 Deoxyribonucleic acid, from salmon testes -:250.7 myo -Inositol hexakisphosphate, dodecas odium salt (sodium phytate) -:not appropriate 94

PAGE 95

P recovered by AEM (g P cm-2) P supplied to AEM (g P cm-2) Figure 3-1. Exchange capac ity of anion exchange membrane (AEM) strips. Values plotted alongside limits of reasonable recovery 95% (----) of calculated exposure. A typical exposure level A of up to 25 g cm-2 equates to 469 g P g-1 of exchangeable phosphate when using one strip (6.25 x 1.5 cm) with 1 g of soil. Limit of reasonable recovery rate, B equates to 117 g P cm-2 in a non-competitive environment. 95

PAGE 96

96 Figure 3-2. Recovery of phosphorus co mpounds by anion exchange membrane (AEM) strips. Average (n=2) plotted, all repeat s within 2.3%. Recovered P plotted as molybdate-reactive P (MRP) or as non-molybdate-reactive P (difference between total and molybdate-reactive P). .

PAGE 97

F F igure 3-3. S o Spe c req u inte r 20 o lution 31P nu c c tra were acq u u ired to achiev e r nal standard m C h 10 c lear magnetic u ired using a 4 e suitable sig n m ethylenediph h emical shift 0resonance s p 4 .833 s (~30 n al to noise. S p osphonic acid (ppm) 97 -10 O r r thophosphat e e H e e xametaphos p p hate P y y rophosphate Gl ucose 6-pho s s phate Ad d enosine 5 m m onophospha t t e 2A minoethylp hosphonic a c c id P h h ytate R N D N N A N A 20 p ectra of P co m ) pulse length p ectra presen t ( = 17.46 pp m pounds mea s and a 2 s rec y t ed using 8 Hz m). s ured after 24 y cle delay, ap p line broadeni n h in deionized p roximately 2 0 n g and refere n water. 0 00 scans n ced to

PAGE 98

Total P (g P g soil-1) Molybdate reactive P (g P g soil-1 ) Figure 3-4. Comparison of total and moly bdate-reactive P as detected in anion exchange membrane eluants. Non fumigated (X) and hexanol fumigated (O) samples show significant differences (paired t-test p < 0.05) between P determined by ICPOES and colorimetric analysis. Yet linear correlation (non fumigated Y = 0.98x + 0.36, R2 = 0.999, fumigated Y = 0.96x 4.08, R2= 0.997,) and the small relative differences (3.4 and 1.3%) suggest limited recovery of molybdate unreactive P. 98

PAGE 99

CHAPTER 4 A SURVEY OF BIOGENIC PHOSPHORUS IN WETLAND SOILS: A SOLUTION 31P NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY STUDY Introduction Biologically sourced and cycled P represent s a significant component of the wetland P cycle (Newman and Robinson 1999; Reddy et al. 1999; Reddy et al. 2005), with operational defined organi c P representing on average over 60% and up to 94% of total P within certain wetland systems (CHAPTER 2). If consideration is also given to polyphosphates, an often-neglected component of the P cycle (Davelaar 1993), it is clear that biogenic P dominates total P in m any wetland soils. The functional nature of this biogenic P has profound implications upon its interaction and stability in the environment (Celi and Barberis 2005b) as well as determining potential biological turnover (Oberson and Joner 2005; Richards on et al. 2005; Turner 2008a). Analytical techniques, such as solution 31P NMR spectroscopy, now allow researchers the tools to identify and track the functi onal forms of this biogenic P in the environment (CadeMenun 2005b; McKelvie 2005). Research employing 31P NMR spectroscopy has provided valuable insight into P cycling of wetland systems, yet has to date been limited (CHAPTER 2). This study represents a unique effort to expand th e application of the technique to investigate the range of biogeni c P found within palustrine wetland soils. First applied to wetlands during the study of Scottish blanket bogs (Bedrock et al. 1994), solution 31P NMR spectroscopy has been appli ed to a number of natural and artificial wetland systems (Cade-Menun 2005b) (CHAPTER 2). Yet, its application requires a certain level of specialist knowled ge as well as access to NMR facilities. This has limited the type of wetlands studied to da te. Published work on freshwater systems is typified by a focus on lacustrine sediment s in Europe (Ahlgren et al. 2005; Hupfer and 99

PAGE 100

Gachter 1995; Reitzel et al. 2007) China (Bai et al. 2009; Liu et al. 2009; Zhang et al. 2009b), and Australia (Baldwin 1996), rive rine systems in New Zealand (McDowell 2009) and a limited number of palustrine systems including: the highly organic subtropical marshes of south Florida (R obinson et al. 1998; Turner and Newman 2005; Turner et al. 2007b; Turner et al. 2006a), blanket bogs in Scotland (Bedrock et al. 1994), Carolina Bays in the USA (Sundareshwar et al. 2009), as well as flooded rice paddies in Madagascar (Turner 2006). To dat e, work in palustrine systems has highlighted both the diverse r ange of biogenic P forms, includ ing polyphosphates, that may be found (Sundareshwar et al. 2009) and a potential P composition fundamentally different to terrestrial systems (Turner and Newman 2005) (CHAPTER 2Range of Applications in Wetlands). Although it is apparent that certain wetland soils may contain fundamentally different P compositions to upland systems, a lack of res earch in this transitional ecotone leaves us unable to determine if such observations are universal. In this chapter, I use solution 31P NMR spectroscopy to evaluate the nature of biogenic P from a broad range of palustrine wetlands. Specific obj ectives include; 1) establish the nature and diversity of biogenic P forms found within wetlands soils, 2) analyze forms identified in the context of ancillary biogeochemical and env ironmental information to derive potential controlling mechanisms or characteri stics that could explain the observed P composition. Methods Sampling Surface samples were collected over the course of three years from a diverse range of wetland systems (Table 4-1). Study site locations were dictated by available 100

PAGE 101

access or established working collaborations, but were selected for sampling to obtain a diverse range of climatic, hydro-geomorphic and vegetation types. Given the exploratory nature of the study, soil sampling consisted of four surface cores (diameter 7.5 cm x 10 cm) collected from independent sites consider ed representative of the study wetland. Samples were kept on ice for immediate shipm ent to the Univ. of Florida, or in two cases were air dried on site (sites 25 and 26). Samples were processed by hand, removing coarse inorganic and organic fragm ents >2 mm. Homogenized samples were split, with subsamples being kept at 4C (fre sh) and set out to air dry under ambient lab conditions, nominally 10 days, under conditi ons of elevated air flow. Fresh samples were analyzed for water content, pH, ex changeable P and microbial P. Air dried samples were ground (8000D mixer mill, SPEX SamplePrep, NJ) and sieved (mesh 60, 0.250 mm) prior to analysis for total elemental composition (C, N, P, Al, Ca, and Fe) and P composition by solution 31P NMR spectroscopy. Biogeochemical Characterization Fresh soil samples were analyzed for so il water content by gravimetric loss following drying at 70 C for 72 h. Sample pH was determined on a 1:2, soil to water suspension using a glass electrode. Excha ngeable and microbial P were determined by anion exchange membrane batch method (Kouno et al. 1995; Myers et al. 1999; Thien and Myers 1992). In brief, two fresh samples of each soil (3.5 g dry weight equivalent) were weighed into 250 mL HDPE centrifuge bottles and a sample-specific volume of DDI was added to bring the total water content to 75 mL. All samples received a single anion exchange membrane strip (1 .5 x 6.25 cm, BDH Prolabo Product number: 551642S, VWR International, UK) preloaded with HCO3 counter ions (CHAPTER 3), with a parallel subsample also receiving 1 mL of 1-hexanol (95%, Sigma Aldrich, St 101

PAGE 102

Louis, MO). Samples were sealed and placed on a reciprocating shaker for 24 h after which membrane strips were removed, rinsed under running DDI and placed in a conical tube with 50 mL 0.25 mol L-1 H2SO4. Membranes were eluted for 3 h and the resulting solution was analyzed for molybdate -reactive P using a discrete auto-analyzer (AQ2+, SEAL Analytical, UK) and standar d molybdate colorimetry (USEPA 1993). The difference between P recovered between paralle l samples is attributed to fumigation released P (Bunemann et al. 2008a) and is used in this study as a proxy for microbial P without correction factor (Jenkinson et al. 2004). Dried and ground soils were analyzed for loss on ignition (an estimate of total organic matter) and elemental concentrations. Total P and metals were determined by combustion of soil at 550C in a muffle fu rnace for 4 h and dissolu tion of the ash in 6 mol L-1 HCl (Andersen 1976). Acid solutions were analyzed for molybdate P (as above) and for Al, Ca, and Fe, using ICPOES (Thermo Jarrell Ash ICAP 61E, Franklin, MA). Total soil C and N were measured by combustion and gas chromatography using a Flash EA1112 (Thermo Scientific, Waltham, MA). Phosphorus Composition Phosphorus forms were characteriz ed via a standard alkaline extract and solution 31P NMR spectroscopy of air dried soils (Cade-Menun and Preston 1996; Turner et al. 2005). Although pr etreatment is expected to impact P composition (Turner et al. 2007b) the use of air drying was considered preferable given the ease of application by collaborators and the belief th at slow drying processes are a realistic scenario within the environmen t. Phosphorus was extracted by shaking 1.00 g 0.01 g of soil with 30 mL of solution containing 0.25 mol L-1 NaOH and 50 mmol L-1 EDTA in a 50-mL centrifuge tube for 4 h, after whic h samples were centrifuged at 7000 rpm 102

PAGE 103

(maximum RCF ~7000 g) (Sorvall RC6, S L600 Rotor; Thermo Fisher Scientific, Waltham, MA) for 10 min. Subsamples of s upernatant were drawn off for individual sample determination of total P using a double acid (HNO3H2SO4) digest (Rowland and Haygarth 1997) and molybdate colorimetry. Secondary subsamples were combined on an equal volume basis within field rep licates, spiked with an internal standard methylenediphosphonic acid (MDP), frozen ( 80C) and lyophilized to await solution 31P NMR spectroscopy. Lyophilized extracts (~300 mg) were redissolved in 3 mL of 1 mol L-1 NaOH and 0.1 mol L-1 EDTA within 15 mL centrifuge tubes before vortexing for 1 min. Samples were subsequently filtered using a prewashed 0.2 m syringe filter (GF-B) to remove fine particles that may result in poor fi eld homogeneity and thereby cause unacceptable line broadening (see APPENDIX C1 ). Comparison of samples with and without filtration suggests no significant error is associated with the filt ration step (see APPENDIX C2). Subsequently, 2.7 mL of redissolved filtered sample and 0.3 mL D2O (for signal lock) were loaded into a 10 mm NMR tube for spectra acquisition. The use of an alkaline matrix ensured a final pH >13. Although pot entially resulting in the degradation of certain phosphodiester functional groups (i.e. RNA and phosphatidyl choline) (Turner et al. 2003d), this allows for consistent chem ical shift (McDowell and Stewart 2005b) and confidence in peak assignment wh en comparing to existing spectral libraries (Turner et al. 2003d). Spectra were acquired immediately using an Avance-500 (500.4 MHz 1H), Magnex 11.8 Telsa/54 mm Bore magnet (AMR IS facility, McKinght Brain Institute University of Florida) at a controlled 25 C. Acquisition paramet ers included use of a 103

PAGE 104

simple zgig pulse profile, a calibra ted 30 pulse, 2 s pulse delay and broad heteronuclear decoupling (waltz 16). Betw een 30,000 and 50,000 scans were required to achieve a reasonable signal to noise ratio, dependent upon sample P concentrations. Spectra were acquired using sequential bl ocks of scans (10,000) with subsequent combination of FIDs using Bruker proprietary software. Spectra were analyzed using wxNUTS vr 1.0.1 for Microsoft Windows (Acorn NMR Inc. 2007). Initially spectra were processed using 15 Hz line broadening, phased and corrected for baseline shift, and referenced using internal standard MDP ( = 17.46 ppm), established by comparison of a redissolved soil extract with an external standard, 85% H3PO4 (0 ppm) (see APPENDIX C3). Spectra were integrated over set intervals, corresponding to established bonding environments (Table 5-1). The region between 3 and 8 ppm was further investigated using the deconvoluti on utility of wxNUTS software. A best fit deconvolution of the spectra was acquired using 2 Hz line broadening (Figure 4-1). Peak picking parameters were adjusted dependent upon signal to noise ratio of specific samples, but ranged between 1 and 8% of maximum peak height and used 0.5 for the root mean squared nois e parameter. The region was sp lit into orthophosphate and phosphomonoesters (all other peaks determined by the algorithm in the region 3 to 8 ppm). Peak proportions from the deconvoluti on protocol were applied to the integral determined in the 15 Hz spectra. A sim ilar procedure was applied to the region 3 to 5 ppm (Figure 4-1), to diffe rentiate pyrophosphate ( 4.37 ppm) and higher order polyphosphate groups ( 3.91 and 4.03ppm) based upon com parison with standard biogenic P compounds in the same matrix (see APPENDIX B4). 104

PAGE 105

Data Analysis Statistical analysis was carried out using JMP 8/0 (SAS Institute Inc. 2008) with SPSS for windows version 17.0.0 statistical software (SPSS Inc., 2008) used for graphical representation of correlations. Given non-normal data, co-linearity in fundamental biogeochemical characteristics was test ed by application of Spearmans rank correlation. Exploration of emergent patterns was carried out by delineating wetland sites into four fundamental groups, using hierarchical (Wards) classification of organic matter content and pH. Organic matter and pH were selected given their lack of colinearity, and known influence of both parameters on biogeochemical P cycling. Ordination of P composition diversity was performed using principal components analysis (PCA) and compared with fundamental characteristics, including previous defined groupings. This informed the development of multiple linear regression models that sought to fit the ratio of phos phomonoesters to phosphod iesters found within wetland soils. The ratio was normalized us ing natural log transformation and the parameters used were standardized and ce ntered using Equation 4-1, where x equals the average and x equals the standard deviation of the parameter x. The model was evaluated using change in Akaike information Cr iterion (AIC), with values >2 considered strong evidence of model improvement (Burnham and Anderson 2004). (x-x) / (2 x) (4-1) Subsequently the influence of mi crobial biomass P, as a % of total P on P composition was explored using Spearmans rank correlation against major biogenic P groups. 105

PAGE 106

Results and Discussion Wetlands Sampled The 28 wetlands analyzed included two large wetland systems, the Changuinola peat dome, Panama (Sites 20, 21, 22) and Houghton Lake treat ment wetland, Michigan (sites 4,5,6) in which three separate locations were treated as individual wetland sites. This was considered appropriate given physica l distances and distinct biogeochemical conditions found between sites (Kadlec and Mi tsch 2009; Sjgersten et al. 2010). Basic biogeochemical characterization and alkaline extraction for P analysis was carried out on nominally four surface samples (Table 4-2). In certain select cases, this protocol was not achievable. Specifically, a lack of sample material from 8 mile lake Alaska (site 1) necessitated the combination of samples on an equal mass basis and the application of alkaline extraction to a single homogenized samp le. Similarly, materi al received from Dr. Sofie Sjgersten from site 26 situated near the town of Abisko, Sweden represented a single previously homogenized air dried sample of surface (0-10 cm) soil considered representative of the study site. Study sites represent a vast range of climatic conditi ons, landscape positions, dominant vegetation types as well as known di fferences in nutrient status (Table 4-1), from tropical ombrotrophic peat domes (sites 20, 21, and 22), to high latitude acidic based peatlands (sites 1, and 27) and fens (sites 28), to calc areous tropical/ subtropical systems (sites 17,18,19,23,24 and 29), temperat e fens (sites 3,15, and 16) and Carolina bays (sites 7-14). Study sites included both pristine systems and those severely impacted by up to 30 y (Kadlec and Mitsch 2009) of nutrient enrichment (sites 4, 5, and 6). In addition, the study incl uded a number of o ften understudied wetland types, such as wet tundra (site 25) and highaltitude Paramo (site 2). 106

PAGE 107

Biogeochemical Characteristics of Wetlands Sampled Characterization showed a high degree of variability among wetland sites (Table 4-2). Loss on ignition, an often applied estimate of organic matter, ranged from 90 mg g1 in a highly mineral Carolina Bay, within the Francis Marion National Forest (site 9) to estimated fully organic matter, peat-based syst ems of high-latitude wet tundra (site 26) and ombrotrophic bogs (site 27). As expected, total C show ed a close colinearity with estimated organic matter (Spearmans rho = 0.9035, p < 0.0001). Deviation from a direct relationship between estimated organi c matter content and total C determined by elemental analysis was taken as evidence of carbonates (Wright et al. 2008) in calcareous samples from Belize (sites 17,18,19) and south Flor ida (site 23,24). The macronutrients N and P also showed a large range in concentration between wetland systems, from 2 to 36 mg N g-1 and 51 to 3516 g P g-1 respectively. Plots of average molar ratios (Table 4-2, Figure 4-2) highlight a distin ction between P and N relationships with C. As expected, giv en the biological derivation of N in the environment (McGill and Cole 1981), there a ppears to be generalized coupling of total C and N (Spearmans rho = 0.669, p < 0.0001). On closer inspection, this coupling appears to break down in highly organi c systems (total C > ~360 mg C g-1). When considering only systems in which C < 360 mg C g-1, total C and N show a higher degree of correlation (Spearmans rho 0.886, p < 0.0001) in a fashion similar to terrestrial soils (Cleveland and Liptzin 2007). In contrast P appears to show no correlation with total C (Spearmans rho 0.226, p = 0.247). Although such a decoupling could reflect the importance of organic P cycling (Cleveland and Liptzin 2007) in wetlands, in such a diverse range of sites it is more likely to reflect fundamental 107

PAGE 108

differences in underlying site mineralogy and anthropogenic inputs of P independent of C/N sources. Of the total metals analyzed, Al ranged from 0.4 to 77.1 mg Al g-1 and showed significant negative correlation with organic matter (Spearmans rho, -0.800, p < 0.0001). This is to be expected given the significant role aluminosilicates are likely to play in wetlands with a large mineral fraction. Calcium contents ranged from undetectable within Carolina bays of South Carolina to 334 mg Ca g-1 in the Belizean calcareous fen (site 19). Such high concent rations in sites 19 and 17 (232 mg Ca g-1) probably reflect the presence of shell fragments and calcareous cyanobacterial mats within surface samples collected from these coastal, low-salinity sites (Macek and Rejmankova 2007). Even if these sites were considered outlie rs and excluded from analysis, there was clear correlation between Ca concentration and site pH (Spearmans rho = 0.6971, p < 0.0001). The redox-sensitive metal, Fe, showed no apparent correlation with other basic biogeoc hemical characteristics, and ranged from detection limit of 0.2 mg Fe g-1 in a large number of wetland sites to a maximum of 18.9 mg Fe g-1 within the heavily impa cted portion of the Houghton Lake treatment wetland. Given strong colinearity between several basic biogeochemical characteristics, and the fact that Fe content was at or below detection limits for a number of sites, I made the decision to define four broad groupings of wetland sites based upon the simple hierarchical classification (Wards) of organic matter and pH (Figure 4-3). This would then be applied to invest igate emergent trends within P composition diversity. The first group of 6 wetlands (group A) cons ists of highly organic (836 1000 mg g-1 loss on ignition), acidic (pH 3.6-4.6) systems. Typified by Sphagnum sp.-dominated, 108

PAGE 109

high-latitude bogs and mire s (i.e. sites 1, 26 and 27), it also included tropical ombrotrophic systems with a r ange of vegetation types (sites 20, 21, and 22). The second grouping of eight wetland s (group B) represents those with an acidic (3.5 4.4) pH and lower organic matter content (92 688 mg g-1 loss on ignition) than the truly peat like, group A wetlands. This group incl uded only Carolina Bay wetlands from the Southeast Coastal Plain, USA. Although r epresenting a range of vegetation types, including both Cypress-dominated forested systems (site 8) and herbaceous open water systems (site 13), their similar hydrogeomorphic setting within the landscape resulted in broadly similar biogeochemical characteristics(De Steven and Toner 2004; Gaiser et al. 2001). The third group (C) represents 10 wetlands found to have an approximately neutral pH (5.9 7.3) and high organic matter content (560 941 mg g-1 loss on ignition). It included calcareous fens from England (site 3), New York (sites 15, 16), Canada (site 28) and South Flori da (sites 23, 24), plus wet Paramo of Ecuador (site 2) and the Houghton lake treatment wetland (sites 5,6, and 7). The last group of wetlands considered (D) represented t hose with an alkaline pH (pH 7. 0 7.6) and relatively low organic matter content (165 300 mg g-1 loss on ignition). This group was dominated by calcareous fens (Macek and Rejmankova 2007) situated near the coast of northern Belize (sites 17,18 and 19), but also included an arctic tundra system (site 25) that has seen heavy grazing by migrating pink -footed geese (Wal et al. 2007). Solution 31P Nuclear Magnetic Resonance Spectroscopy Extraction of total phosphorus Extraction of total P by the NaOHE DTA method ranged from 25 to 84%. One site (9) with a very low total P concentration (51 g P g-1) was calculated to have an extraction efficiency of 125% and was therefor e removed from all subsequent analysis 109

PAGE 110

of P composition. Extraction efficiencies were found to vary significantly between wetland groups (Kruskal Wallis test Chi2 = 8.23 d.f. 3, p < 0.05) reflecting the known influence of calcareous soils on the standard NaOHEDTA extraction (McDowell and Stewart 2006; Turner et al. 2003a). Given the extraction ex plicitly targets organic P (Turner et al. 2005), it is likely to reflect the biogenic P composit ion of soils, with the residual considered alkalinestable, inorganic and recalcit rant organic P (Cade-Menun 2005b; Turner et al. 2005). Therefore, the operationally defined residualP is considered a distinct P ty pe, and is included here when considering patterns in P composition. Phosphorus composition Solution 31P NMR spectroscopy of alkaline extracts identified a diverse range of biogenic P forms within wetland soils (see T able 4-3 and Figure 44, 5, 6, 7). Two calcareous, low-P sites from Belize (sites 18 and 19) showed no evidence of biogenic P, with only orthophosphate identif ied. The remaining sites contained phosphonates (0 to 44 g P g-1) phosphomonoesters (8 to 461 g P g-1) DNA (3 to 144 g P g-1) other phosphodiesters (6 to 67 g P g-1) and polyphosphates (0 to 197 g P g-1). Total inorganic polyphosphates were further delineated into pyrophosphate and the mid and terminal chain residues of longer polyphosphates (Table 4-4). Given the range of total P between sites, analysis of composition diversity was based upon composition, as a percentage of total P. Ordination using PCA, produced two axes which together accounted for 64.1% of the observed variance in P composition. Examination of the PCA score plot with fundamental wetland groupings superimposed (Figure 4-8) clearly demonstrated the separation of groups B and D, with groups A and C (high organic matter) failing to show any clear distinction in P 110

PAGE 111

composition. Examination of PCA loading plots and associated Eigen vales (Figure 4-8, Table 4-5) showed separation of group D wetlands upon axis 1 to be the result of distinct differences in the proportion of re sidual P as compared to the major biogenic P groups (phosphomonoesters, DNA, phosphodies ters and pyrophosphate) identified, while separation of group B wetlands, upon PCA axis 2 appeared to be a result of increased prevalence of phosphonates and phos phomonoesters. Similar examination of P composition with reference to Cowardin c lass (Figure 4-9) and climatic zone data (not shown) failed to show clear clustering, thereby suggesting that soil P composition is dependent upon basic biogeochemical characteristics including, to some degree, both the pH and organic matter content groupings used in this study. As an extension of initial ordinati on analysis, the influence of basic biogeochemical characteristics on select as pects of P composition were explored. These included the ratio of phosphomonoeste r to phosphodiesters, the presence of Inositol hexakisphosphate and the relations hip between measures of microbial P and biogenic P com position. Ratio of phosphomonoesters to phosphodiesters in wetlands. The ratio of phosphomonoesters to total alkaline stable phosphodiesters in the 28 palustrine wetlands averaged 2.8, ranging from 0.9 in an ombrotropic Canadian bog (site 27) to 10.6 in Norweigan wet tundra (site 25). The influence of basic biogeochemical characteristics on this ratio were explored by multiple linear regression. Given the apparent influence of organic matter (or its re ciprocal mineral content) in delineating high phosphomonoester-containing group B wetlands this was selected as a model basis. Additionally the C:P rati o, a measure of potential P lim itation, was tested, given 111

PAGE 112

the known influence of P availability on microbial eco-physiological responses associated with organic P cyc ling (Corstanje and Reddy 2006; Corstanje et al. 2007). The C:P ratio also fulfilled the required assumption of being independent from organic matter. Comparison of AIC values (Tabl e 4-6) gave strong evidence (Burnham and Anderson 2004) for a loss of information wh en considering the reduced model [f(LOI)] over the optimal model [f(LOI + C:P). Th is would suggest both Loss on ignition and C:P play a significant role in determining the ra tio of phosphomonoesters to phosphodiesters within wetland soils. Parameter estimates from the optimal model (Table 4-7) show that the ratio of phosphomonoesters to phosphodi esters decreases with both increasing organic matter and an increasing C:P ratio. T herefore, increased mineral content and reduced P limitation both result in an el evated proportion of P to be found as phosphomonoesters instead of phosphodiesters. Presence of Inositol hexakisphosphate. Spectral deconvolution of the 8 to 3 ppm region revealed that in some samp les a substantial portion of phosphomonoesters corresponded with known peak assignment of higher order Inositol phosphates (Turner et al. 2003f; Turner and Richardson 2004). The use of a standard preparation and spectra acquisition protocol in conjuncti on with a stable internal standard (MDP) provided confidence in the assignments of both myoand scyllo -IP6 (Figure 4-10, Table 4-8). Inositol groups appear ed particularly prevalent in group B wetlands, with myo-IP6 accounting for between 16 and 39% of total phosphomonoesters and scyllo -IP6 constituting between 8 and 20% of total phosphomonoesters. Tw o group B wetlands (site 7 and 11) also showed evidence of phosphomonoester peaks (6.68 and 6.88 ppm) 112

PAGE 113

suspected of being lower-order inositol deriv atives or as yet uni dentified isomers of IP6 (Turner 2007). Determination of IP6 within wetlands other than group B systems proved problematic given the degree of peak overlap within the phosphomonoester region. However, peaks coincident with that attributed to scyllo -IP6 in group B wetlands (4.241 0.023) were found in sites 1, 2, 6, 15, 16, and 25. Given the suspected role of myo-IP6 as a biosynthetic precursor to scyllo -IP6 within the soil matrix (L'Annunziata 2007), it is likely that both isomers may be present. Further analysis utilizing hypobromination to hydrolyze non-inositol phosphomonoesters (Irving and Cosgrove 1981) would be required to both confirm the presence of scyllo -IP6 and reveal the presence of other isomers. Microbial biomass The potential relationship betwe en microbial P, determined by anion exchange membranes, and P composit ion was explored by application of Spearmans rank correlation (Table 4-9) This highlighted both DNA and long-chain polyphosphates (as percentages of total P), to show significant positive correlations with microbial P (Figure 4-11). A significant positive correlation between microbial P and DNA (Spearmans rho = 0.61 p = 0.002) would be expected gi ven a standard microbial composition. Yet the broad r ange of microbial biomass P (inter-quartile range equals, 2.4 to 19.0% of total P) as compared to that of DNA (inter-quartile range 3.8 to 9.6% of total P), would suggest confounding factors t hat influence the proportion of DNA found within soils, including altered microbial P composition between systems (Makarov et al. 2005) and the influence of extracellular stabi lization of DNA (Cel i and Barberis 2005b; Niemeyer and Gessler 2002). The highly signi ficant correlation between microbial P and 113

PAGE 114

long chain polyphosphates (Spearmans rho = 0.78, p < 0.0001) reflects their biological synthesis (Harold 1966), yet there also appears to be a distin ction with a marked increase in polyphosphate composition in sy stems with a microbial P >15% of total P, possibly reflecting the role of polyphosphates as a metabolic stress response or strategy under competitiv e environmental conditions (Kulaev and Kulakovskaya 2000; Seufferheld et al. 2008). That said, the known interaction anion exchange membranes with certain biogenic P forms (Cheesman et al. 2010b) necessitates caution when interpreting causation. The strong positive correlation seen between total polyphosphates (pyrophosphate + longer chai n length polyphosphates) and microbial P, (Spearmans rho = 0.8103, p < 0.0001) may reflect the fa ct that operationally defined microbial P is in a large part due to recovery of polyphosphates. Discussion This unique data set demonstrates the diverse range of biogenic P found within wetland soils, while also providing the basis to begin to explore mechanistic drivers behind the P composition of wetland soils. In addition to confirming the nature of P within calcareous palustrine systems (Turne r and Newman 2005; Turner et al. 2006a) this work also demands caution when extr apolating isolated observations to wetland systems in general, highlighting differ ences between broad wetland types. The investigation of emergent patterns in P composition identified here based upon basic biogeochemical characteristics (organic matt er, pH, and nutrient availability) will form the focus of subs equent chapters. Studies within terrestrial systems have a ttributed a positive correlation between the proportion of P found as phosphodiesters and annual precipitation to their increased recalcitrance under wetter conditions (Cond ron et al. 1990a; Sumann et al. 1998; Tate 114

PAGE 115

and Newman 1982). It has been suggested that a similar mechanism of reduced turnover under anaerobic conditions could acc ount for the increased prevalence of phosphodiesters in palustrine wetlands studied to date (Turner and Newman 2005). In addition, the phosphomonoester IP6, often a major component of P in terrestrial soils (Cosgrove 1966; Murphy et al. 2009; Turner et al. 2003f; Turner et al. 2002b), had thought to be absent from wetlands (Turner and Newman 2005; Turner et al. 2006a), with evidence suggesting rapid degradation u nder anaerobic conditions, typical of wetland soils, (Suzumura and Kamatani 1995a, b) As evident from this study, and from other recent research on estuarine (Turner and Weckst rm 2009), lacustrine (Zhang et al. 2009b), and riverine (McDowell 2009) system s (presumed to experience anaerobic conditions) IP6 may actually constitute a subs tantial proportion of biogenic P in wetlands. It is therefore likely that its presenc e, or absence, is likely to be critical to observed trends in phophmonoester to phosphodiester ratios. In terrestrial systems, calcite, Fe/Al oxides, clay and organic ma tter have all been shown to increase soils IP6 sorbtion capacity (Celi and Barberis 2007), yet in wetlands it is likely these factors are further impacted by ambient physicochemic al conditions (such as anaerobosis). As an example, in the comparison of ferrous and ferric salts of IP6, the reduced Fe(II)-IP6 complex was shown to have a lower rate of enzymatic dephosphorylation than Fe(III)IP6 (Heighton et al. 2008). The complex inte ractions between site properties and IP6 stabilization are likely to be profound and ar e explored further in this dissertation (CHAPTER 6). An alternative to the differential stability of IP6 under ambient wetland conditions, may be difference in the inputs or in-situ IP6 synthesis between wetlands. Inositol phosphates have been found in a r ange of biological materials (Michell 2008) 115

PAGE 116

with higher order forms such as myo-IP6 found to represent a s ubstantial proportion of total P in reproductive structures such as pollen (Jackson et al. 1982) and certain seeds and fruiting bodies (Lott et al. 2000). Wetl ands in group B of this study represent Carolina Bays within the Savannah River Site and Francis Marion National Forest both situated within the South-east Coastal Plain. As such, uplands surrounding the study sites contain significant coniferous forests, that may represent a significant direct (though deposition) or indirect (through runoff) source of myo -IP6 (Jackson and Linskens 1982). In addition, the presence of significant concentrations of otherwise rare stereoisomers (i.e. scyllo) within group B wetlands suggests additional sources and processing of inositol derivatives within curre ntly unidentified soil microbial components (Turner 2007; Turner and Richardson 2004). It is clear that further work is needed to elucidate the role of differential IP6 sources and stabilization on the P composition of wetland soils. Polyphosphates represent more than just a microbial response to excessive P loading, so-called luxury uptake (Khoshm anesh et al. 2002), and have been implicated in a diverse range of metabolic processes (Kornberg 1995; Kornberg et al. 1999; Kulaev and Kulakovskaya 2000). This study identifie d substantial polyphosphate pools within a diverse range of wetlands (although predominant ly acidic high organic-matter systems), including samples from a low-P tropical ombrotrophic peat dome considered (sites 20,21 and 22). Similarly, recent evidence of polyphosphates within unimpacted Carolina Bays (Sundareshwar et al. 2009) and oligotr ophic Swedish lake sediments (Ahlgren et al. 2006a) demands further investigation in to the role of polyphosphates within palustrine systems. It has been suggested, that the relationship between terminal to mid 116

PAGE 117

chain phosphate residues can be used to estimate polyphosphate chain length (see Equation 4-2 where t + m equate to the integrals of terminal and mid chain phosphates, respectively). Chain length = 2( t + m) / t Equ. 4-2. Given the potential for soil-mediated hydr olysis (Hupfer and Gachter 1995) and the hypochromic effect seen in the 31P NMR spectra of model linear polyphosphates (Krupyanko et al. 1998), such a direct calc ulation may be erroneous. Yet it is worth noting that there appears to be a greater difference betw een mid-chain and terminal residues in acidic systems, suggesting greate r chain length. Further work would be needed to investigate if this truly represent s a range of polyphos phate chain lengths across environmental gradients, or an artifact due to extraction and 31P NMR spectroscopy method used. Conclusions This study represents the most diverse range of wetland sites to which solution 31P NMR spectroscopy has ever been applied. In studying the forms of biogenic P found within such a broad range of wetlands I am able to identify potential patterns as well as to explore the validity of ex trapolating patterns seen within the few wetlands studied to date. It is apparent that biogenic P within wetland system s shows a great deal of variation, and extrapolation from previous work (Turner and Newman 2005) to all wetlands should be with caution. When exploring diversity of biogenic P within wetland soils, it is apparent that basic characterist ics including pH and organic matter content play a significant role in determining the nature of functional P forms found, but that climatic conditions and wetland vegetation ty pes appear to have limited impact. As an exploratory study this work highlights a number of potential mechanistic drivers, 117

PAGE 118

118 including the role of mineral content and P availability, in determining the forms of biogenic P found in wetlands. T hese mechanistic drivers are the focus of subsequent chapters.

PAGE 119

Table 4-1. Wetland study sites sampled for c haracterization of biogenic phosphorus composition. Wetland Location Wetland Type Vegetation Type Dominant species Potential impacts 1 8 mile Al, USA Bog Persistent emergent (Moss) Sphagnum, Carex sp. 2 Laguna Papallacta, Ecuador Cushion forming, Paramo Persistent emergent (Herbaceous) Self emergent succulent species Distichia muscoides Spagnum Cattle Grazing 3 Wicken Fen UK Fen Persistent emergent (Herbaceous) Cladium mariscus Sedge Harvesting since 1419 4 Houghton lake (CT350) MI, USA Treatment wetland Persistent emergent (Herbaceous) Typha sp. Intermediate P loading 5 Houghton lake (550C) MI, USA Treatment wetland Persistent emergent (Herbaceous) Cyperacea sp. Typha Low P loading 6 Houghton lake (P) MI, USA Treatment wetland Persistent emergent (Herbaceous) Typha. High P loading 7 Francis Marion National Forest (FMNF) Bay 1 SC,USA Carolina Bay Persistent emergent (Forested) Acer rubrum (var trilobum ), Nyssa biflora and Nyssa aquatica, Lyonia lucida Ilex myrtifolia 8 Francis Marion National Forest (FMNF) Bay 2 SC,USA Carolina Bay Persistent emergent (Forested) Taxodium ascendans, Nyssa biflora, Lyonia lucida, Carex striata, Woodwardia virginica 9 Francis Marion National Forest (FMNF) Bay 3 SC,USA Carolina Bay Persistent emergent (Herbaceous) Ilex glabra, Iris tridentata, Amphicarpum muhlenbergianum Eleocharis spp., M elanocarpa tricostata, and Lachnanthes caroliniana Periodic buring 10 Francis Marion National Forest (FMNF) Bay 4 SC,USA Carolina Bay Persistent emergent (Forested) Nyssa biflora, Taxodium ascendans, Acer rubrum, Lyonia lucida, Cyrilla racemiflora, Pinus taeda 11 Savannah River Site (SRS) Bay 1 SC,USA Carolina Bay Persistent emergent (forested/herbaceous) Panicum hemitomon, Nyssa biflora, Cephalanthus occidentalis,Utricularia spp., Sphagnum spp ., Pontederia cordata var. lancifolia 12 Savannah River Site (SRS) Bay 2 SC,USA Carolina Bay Persistent emergent (Herbaceous) Panicum hemitomon, Sphagnum spp ., Pontederia cordata var. lancifolia, Juncus canadensis, Cephalanthus occidentalis, Acer rubrum (var. trilobum ) 13 Savannah River Site (SRS) Bay 3 SC,USA Carolina Bay Open water /herbaceous Nymphaea odorata, Panicum hemitomon, Utricularia spp., Leersia hexandra, Eleocharis melanocarpa 14 Svannah River Site (SRS) Bay 4 SC,USA Carolina Bay Persistent emergent (Forested) Liquidambar styraciflua, Acer rubrum (var. trilobum ), Nyssa biflora, Taxodium ascendans, Smilax rotundifolia 119

PAGE 120

120 Table 4-1. Continued Wetland Location Wetland Type Vegetation Type Dominant species Potential impacts 15 Larry Fen NY, USA Rich Fen Persistent emergent (Herbaceous) Carex sp. Campylium stellatum 16 Fish Fen NY, USA Rich Fen Persistent emergent (Herbaceous) Typha angustifolia. Carex sp. Campylium stellatum, Sphagnum spp.Calliergonella cuspidata 17 Hidden Belize Oligotrophic Sumpland Persistent emergent (Herbaceous/ Cyanobacteria) Eleocharis cellulosa, Cyanobacteria spp. salt intrusion 18 Quiet Belize Sumpland Persistent emergent (Herbaceous/ Cyanobacteria) Eleocharis cellulosa, Cyanobacteria spp. salt intrusion 19 Doubloon Belize Sumpland Persistent emergent (Herbaceous/ Cyanobacteria) Eleocharis cellulosa, Cyanobacteria spp. salt intrusion 20 Changuinola Site 1 Panama Tropical peat dome Persistent emergent (Forested) Raphia tadeiga 21 Changuinola Site 2 Panama Tropical peat dome Persistent emergent (Forested) Campnosperma panamensis, Cassipourea elliptica (Sw.) Poir, Drypetes standleyi G.L. Webster 22 Changuinola Site 3 Panama Tropical peat dome Persistent emergent (Forested) Campnosperma panamensis Cyrilla racemiflora sawgrass 23 WCA 3A Fl, USA Calcareous Fen Open water Nymphaea spp. Utricularia 24 Everglades National Park Fl, USA Calcareous Fen Persistent emergent (Herbaceous) Cladium jamaicense 25 Ny Alesund Spitsbergen, Norway Wet tundra Persistent emergent (Moss) Calliergon richardsoni, Poa arctica, Dupotia species Geese grazing 26 Stordalen Abisko, Sweden Mire Persistent emergent (Moss) Spagnum fuscum, Betula nana, Rubus chamaemorus,Vaccinium vitis ideae, Empetrum nigrum 27 Bog 8 Canada Ombrotrophic Bog Persistent emergent (Moss) Sphagnum fuscum, graminoids, Lichens 28 Fen 1 Canada Fen Persistent emergent (Herbaceous) Carex sp. = common wetland description = vegetation descriptor based upon Cowardin (1979) classification = dominant vegetation species noted in the field = potential external impacts noted that may effect P cycling

PAGE 121

Table 4-2. Soil biogeochemical properties in studied wetland systems. Organic matter Total P Total C Total N Total Ca Total Fe Total Al Molar Ratio pH mg g-1 g g-1 mg g-1 C:P N:P Group A 1 4.6 919 10 986 30 412 4 12 1 4.7 0.5 15 2.5 3.0 0.5 1083 85 28 3 26 4.1 1000 238 424 6 1.8 0.2 0.4 4596 56 27 3.9 974 8 356 36 436 6 8 0 3.2 0.7 0.3 0.0 0.5 0.2 3219 492 51 9 20 3.8 878 10 1124 17 489 2 28 0 1.4 0.2 6.2 0.7 3.0 0.1 1125 40 55 3 21 3.6 927 9 852 58 485 5 25 1 3.3 0.5 3.7 0.6 1.3 0.1 1498 233 66 6 22 3.7 836 47 579 45 424 13 22 1 1.0 0.1 1.5 0.4 1.2 0.2 1928 288 84 5 Average 4.0 922 689 445 17 3 4 2 2242 57 Max 4.6 1000 1124 489 28 5 15 3 4596 84 Min 3.6 836 238 412 6 1 0.2 0.4 1083 28 Group B 7 3.6 497 168 925 84 260 43 15 2 0.8 0.4 3.3 1.6 25.8 2.0 728 192 35 6 12 3.9 546 102 1056 60 307 33 21 2 1.2 0.6 3.7 0.5 15.7 4.0 745 96 44 6 10 4.4 478 177 752 211 264 53 15 3 0.3 0.3 4.7 1.1 34.7 7.6 982 199 45 8 8 3.5 688 102 750 82 377 65 19 3 2.0 0.6 2.1 0.5 8.2 4.0 1275 188 56 4 9 4.2 92 39 51 18 44 8 2 0 0.0 0.7 0.3 2.9 1.5 2547 752 98 25 11 4.3 239 107 918 160 105 29 9 2 0.4 0.3 5.6 0.8 72.9 8.2 283 59 21 4 14 4.4 226 49 918 146 94 14 7 1 0.2 0.1 6.2 1.9 77.1 3.3 285 121 19 8 13 4 246 184 347 116 117 44 9 3 0.2 0.2 2.9 0.7 26.2 6.7 815 118 57 5 Average 4.0 377 715 196 12 0.6 3.6 32.9 958 47 Max 4.4 688 1056 377 21 2.0 6.2 77.1 2547 98 Min 3.5 92 51 44 2 0.0 0.7 2.9 283 19 121

PAGE 122

122 Table 4-2 Continued Wetland Organic matter Total P Total C Total N Total Ca Total Fe Total Al Molar Ratio pH mg g-1 g g-1 mg g-1 C:P N:P N:P Group C 2 6.7 835 24 875 87 399 7 15 1 20 2.3 3.7 1.2 3.4 0.8 1212 230 38 4 16 7 854 21 1184 69 421 9 29 1 32 1.5 4.6 1.6 3.6 1.4 929 120 54 5 3 7.3 783 154 937 43 353 7 26 1 96 24.5 12.2 1.5 3.7 0.6 976 60 62 1 4 7.2 935 5 1439 34 452 1 26 0 14 1.1 6.0 0.2 1.0 0.2 812 31 41 1 6 6.4 695 57 3516 255 353 15 35 2 23 5.0 18.8 5.4 15.2 3.7 260 15 22 1 15 7 560 54 1184 26 270 12 21 1 27 2.6 18.8 1.0 20.4 4.2 591 75 38 5 5 6.1 941 4 982 39 455 3 20 1 8.6 0.8 7.2 0.8 0.9 0.1 1201 87 45 3 23 5.9 929 11 277 8 445 7 36 1 26 1.2 0.2 0.0 2.3 0.5 4170 346 289 29 24 6.1 896 13 310 28 443 6 31 1 26 1.7 0.2 0.0 3.5 1.6 3800 845 226 61 28 6.3 891 8 679 3 406 1 26 3 18 4.4 0.4 0.0 1.0 0.3 1546 20 86 15 Average 6.6 832 1138 400 27 29 7.2 5.5 1550 90 Max 7.3 941 3516 455 36 96 18.8 20.4 4170 289 Min 5.9 560 277 270 15 8.6 0.2 0.9 260 22 Group D 17 7.3 245 20 192 32 162 7 11 0 232 51.1 0.2 0.0 15.3 7.1 232 51.1 136 30 18 7.5 300 7 116 16 70 3 6 0 14 0.8 0.2 0.0 45.8 3.2 14 0.8 111 19 25 7.0 286 330 1513 387 132 29 6 1 15 6.6 0.2 0.0 19.9 9.0 15 6.6 9 3 19 7.6 165 29 126 15 153 0 5 0 334 15.4 0.2 0.0 2.3 0.7 334 15.4 90 17 Average 7.4 249 487 129 7 149 0.2 20.8 149 87 Max 7.6 300 1513 162 11 334 0.2 45.8 334 136 Min 7.0 165 116 70 5 14 0.2 2.3 14 9 = organic matter estimated from loss on ignition (550C, 4 h)

PAGE 123

Table 4-3. Phosphorus composition of surface soils as determined by solution 31P NMR spectroscopy NaOHTP Phosph-P Ortho-P Mono-P DNA Phospholipids Total inorganic Polyphosphates Organic P Mono: Dies Group A 1 758 (77) 201 (20) 337 (34) 135 (14) 42 (4) 43 (4) 514 (52) 1.9 26 190 (80) 70 (29) 44 (18) 25 (11) 10 (4) 41 (17) 79 (33) 1.3 27 138 (39) 36 (10) 31 (9) 23 (7) 9 (3) 38 (11) 64 (18) 0.9 20 516 (46) 20 (2) 171 (15) 162 (14) 35 (3) 6 (1) 123 (11) 222 (20) 3.9 21 320 (38) 8 (1) 95 (11) 55 (6) 31 (4) 12 (1) 118 (14) 107 (13) 1.3 22 261 (45) 54 (9) 68 (12) 57 (10) 3 (1) 79 (14) 128 (22) 1.1Average364 (54) 14 (2) 105 (16) 116 (16) 51 (8) 14 (2) 74 (12) 186 (26) 2 Max 758 (80) 20 (2) 201 (29) 337 (34) 135 (14) 42 (4) 123 (17) 514 (52) 4 Min 138 (38) 8 (1) 36 (9) 31 (6) 23 (3) 3 (1) 38 (4) 64 (13) 1 Group B 7 637 (69) 29 (3) 137 (15) 368 (40) 56 (6) 17 (2) 30 (3) 470 (51) 5 12 722 (68) 44 (4) 179 (17) 390 (37) 65 (6) 19 (2) 25 (2) 518 (49) 4.6 10 497 (66) 18 (2) 120 (16) 252 (33) 57 (8) 24 (3) 26 (3) 351 (47) 3.1 8 476 (63) 16 (2) 138 (18) 192 (26) 67 (9) 14 (2) 50 (7) 288 (38) 2.4 9 63 (125) 14 (27) 38 (76) 6 (12) 5 (10) 45 (88) 6.2 11 598 (65) 14 (2) 128 (14) 408 (44) 32 (3) 12 (1) 4 (0) 466 (51) 9.3 14 534 (58) 12 (1) 267 (29) 189 (21) 47 (5) 10 (1) 9 (1) 258 (28) 3.3 13 219 (63) 4 (1) 56 (16) 110 (32) 25 (7) 15 (4) 9 (3) 154 (44) 2.8 Average 526 (65) 20 (2) 146 (18) 273 (33) 50 (6) 16 (2) 22 (3) 358 (44) 4 Max 722 (69) 44 (4) 267 (29) 408 (44) 67 (9) 24 (4) 50 (7) 518 (51) 9.3 Min 219 (58) 4 (1) 56 (14) 110 (21) 25 (3) 10 (1) 4 (1) 154 (28) 2.4 Group C 2 609 (70) 244 (28) 176 (20) 76 (9) 31 (4) 81 (9) 284 (32) 1.6 123

PAGE 124

124 Table 4-3. Contin ued (Group C) NaOHTP Phosph-P Ortho-P Mono-P DNA Phospholipids Total inorganic Polyphosphates Organic P Mono: Dies 16 595 (50) 131 (11) 270 (23) 106 (9) 28 (2) 60 (5) 404 (34) 2 3 789 (84) 167 (18) 344 (37) 144 (15) 62 (7) 73 (8) 549 (59) 1.7 4 909 (63) 295 (20) 317 (22) 142 (10) 67 (5) 88 (6) 526 (37) 1.5 6 2569 (73) 1759 (50) 407 (12) 141 (4) 64 (2) 197 (6) 612 (17) 2 15 753 (64) 118 (10) 461 (39) 88 (7) 37 (3) 50 (4) 586 (49) 3.7 5 593 (60) 225 (23) 170 (17) 96 (10) 30 (3) 72 (7) 296 (30) 1.3 23 102 (37) 38 (14) 28 (10) 18 (7) 6 (2) 11 (4) 53 (19) 1.2 24 130 (42) 46 (15) 38 (12) 27 (9) 9 (3) 10 (3) 75 (24) 1.1 28 283 (42) 54 (8) 77 (11) 57 (8) 22 (3) 74 (11) 156 (23) 1Average733 (59) 308 (20) 229 (20) 90 (9) 36 (3) 72 (6) 354 (32 2Max2569 (84) 1759 (50) 461 (39) 144 (15) 67 (7) 197 (11) 612 (59 4Min102 (37) 38 (8) 28 (10) 18 (4) 6 (2) 10 (3) 53 (17) 1 Group D 17 47 (25) 36 (19) 8 (4) 3 (2) 11 (6) 2.3 18 53 (46) 53 (46) 25 534 (35) 292 (19) 221 (15) 11 (1) 10 (1) 242 (16) 10.6 19 33 (26) 33 (26)Average167(33)104 (28) 57 (5) 4 (1)3 (0) 127 (11) 6 Max534(46)292 (46) 221 (15) 11 (2)10 (1) 242 (16) 11 Min33(25)33 (19) 0 (0) 0 (0)0 (0) 11 (6) 2 suspected error associated with determination of total P composition, site removed from subsequent analysis. Total P recovered by alkaline extraction Total phosphonates Total orthophosphate Total phosphomonoesters Ratio of total phosphomonoesters: total phosphodiesters

PAGE 125

Table 4-4. Inorganic polyphosphates as determined by solution 31P NMR spectroscopy of wetland soils. Values represent total inorganic polyphosphates delineated into pyrophosphate, and the terminal (TR) and mid-chain (MR) of long chain (n>3) polyphosphates. Pyrophosphate Long chain polyphosphate g g-1 % total P g g-1 % of total P TR MR Group A 1 7.5 0.8 trace 35.5 3.6 26 4.9 2.1 7.9 28.0 15.1 27 4.3 1.2 6.2 27.3 9.4 20 12.6 1.1 trace 110.1 9.8 21 19.5 2.3 trace 98.9 11.6 22 7.5 1.3 trace 71.6 12.4 Group B 7 9.1 1.0 6.3 14.2 2.2 12 5.6 0.5 5.6 13.6 1.8 10 9.7 1.3 6.2 9.9 2.2 8 5.4 0.7 7.0 37.6 6.0 9 4.9 9.8 11 4.3 0.5 14 8.8 1.0 13 9.2 2.6 Group C 2 30.9 3.5 19.5 31.1 5.8 16 46.2 3.9 6.3 7.8 1.2 3 41.6 4.4 25.9 5.9 3.4 4 40.2 2.8 17.5 30.6 3.3 6 136.3 3.9 30.1 31.1 1.7 15 32.3 2.7 10.6 6.8 1.5 5 24.6 2.5 12.1 35.2 4.8 23 6.6 2.4 1.7 3.1 1.7 24 9.6 3.1 28 10.3 1.5 23.6 39.8 9.3 Group D 17 18 25 19 125

PAGE 126

Table 4-5. Eigen values of principal components determined on PCA applied to phosphorus composition within wetland so ils. Principal components, 1 and 2 account for 41.6 and 22.5% of total P co mposition diversity respectively. P form PC 1 PC 2 Phosphonates 0.033965 -0.66096 Orthophosphate -0.102 0.09497 Phosphomonoesters 0.385128 -0.49027 DNA 0.489812 0.164466 Other phosphodiesters 0.502311 0.150683 Pyrophosphate 0.364425 0.359201 Long chain polyphosphates 0.088475 0.305069 Residual -0.45525 0.204691 Table 4-6. Multiple linear regressi on models used to predict the ratio of phosphomonoesters to phosphodiesters in wetland surface soils. Model DFe Adj R 2 i 1 f(cLOI +cC:P) 23 0.6650 0 2 f(cLOI) 24 0.6108 2.197 3 f(cC:P) 24 0.2885 17.88 Table 4-7. Parameter estimates for optimal model for predicting ratio of phosphomonoetsrs to phosphodiesters. Parameters centered and standardized and response variable norma lized by natural logarithm before model run. Term Estimate Std error t ratio p value Intercept 0.8775483 0.078683 11.15 < 0.0001 cLOI -0.962098 0.181879 -5.29 < 0.0001 cC:P -0.384798 0.174035 -2.21 0.0372 Table 4-8. Inositol hexakisphosphat es as determined within group B wetlands. Concentrations g g-1 (% of phosphomonoesters) Site myo IP6 scyllo IP6 7 96.1 (26.1) 56.1 (15.2) 8 63.8 (16.3) 48.1 (12.3) 9 trace 36.3 (14.4) 10 40.7 (21.2) 37.9 (19.7) 11 14.8 (38.9) 2.5 (6.7) 12 131.6 (32.2) 55.4 (13.6) 13 50.1 (26.5) 19.1 (10.1) 14 20.6 (18.7) 7.2 (6.6) 126

PAGE 127

Table 4-9. Correlation between microbial biomass phosphorus (% of total phosphorus) and phosphorus forms determined by solution 31P NMR spectroscopy (% of total phosphorus). Spearman rho correlation p Phosphonate -0.2359 0.2670 Orthophosphate -4426 0.0303 Phosphomonoesters -0.1348 0.5300 DNA 0.6123 0.0015 ** Other phosphodiesters 0.3037 0.1490 Pyrophosphate 0.3734 0.0723 Long chain Polyphosphate 0.7756 < 0.0001 *** significant at the 0.05 level **significant at the 0.01 level *** significant at the 0.001 level 127

PAGE 128

Figure 4-1. Solution 31P NMR spectra of surface soils collected from a Michigan treatment wetland (Site # 6). Detail showing spectral deconvolution used to identify phosphomonoesters and inorganic polyphosphates. 128

PAGE 129

Average molar ratio observed in terrestrial soils All samples n = 28 rho = 0.669; p <0.0001 Total N (mmol g-1) 0 to 30 mmol C g1 n = 14 rho = 0.886; p <0.0001) Group A Group B Group C Group D Site 6: highly impacted Houghton Lake Total P (mol g-1) Total C (mmol g-1) Figure 4-2. Average nutrient concentrations in wetland surface soils. Individual wetlands averages and partial least s quared regressions plotted alongside average relationship observed in terrestri al systems. Molar concentrations of C and N show similar significant positive coupling ( p < 0.0001) improved when considering just low C (< 30 mmol C g-1) sites. 129

PAGE 130

Figure 4-3. Categorization of wetl and sites based upon Wards hierarchical classification of pH and organic matte r (estimated by loss on ignition). 130

PAGE 131

Figure 4-4. Solution 31P NMR spectra of biogenic P composition within group A wetlands (high organic low pH). Spectra acquired using an Avance-500 (500.4 MHz 1H), Magnex 11.8 Tesla/54 mm Bore, at pH > 13 using a simple zgig pulse program and calibrated 30 pulse angle. Between 30-50,000 scans were combined and presented here usi ng 15 Hz line broadening scaled and referenced to internal standar d methylenediphosphonic acid ( = 17.46 ppm). 131

PAGE 132

Figure 4-5. Solution 31P NMR spectra of biogenic P composition within group B wetlands (low organic low pH). Spectr a acquired using an Avance-500 (500.4 MHz 1H), Magnex 11.8 Tesl a/54 mm Bore, at pH > 13 using a simple zgig pulse program and calibrated 30 pul se angle. Between 30-50,000 scans were combined and presented here usi ng 15 Hz line broadening scaled and referenced to internal standar d methylenediphosphonic acid ( = 17.46 ppm). 132

PAGE 133

Figure 4-6. Solution 31P NMR spectra of biogenic P composition within group C wetlands (high organic high pH). S pectra acquired using an Avance-500 (500.4 MHz 1H), Magnex 11.8 Tesla/54 mm Bore, at pH > 13 using a simple zgig pulse program and calibrated 30 pulse angle. Between 30-50,000 scans were combined and presented here usi ng 15 Hz line broadening scaled and referenced to internal standar d methylenediphosphonic acid ( = 17.46 ppm). Note orthophosphate spike curta iled in wetlands 4 and 5. 133

PAGE 134

Figure 4-7. Solution 31P NMR spectra of biogenic P composition within group D wetlands (low organic high pH). S pectra acquired using an Avance-500 (500.4 MHz 1H), Magnex 11.8 Tesla/54 mm Bore, at pH > 13 using a simple zgig pulse program and calibrated 30 pulse angle. Between 30-50,000 scans were combined and presented here usi ng 15 Hz line broadening scaled and referenced to internal standar d methylenediphosphonic acid ( = 17.46 ppm). 134

PAGE 135

5 4 3 2 1 0 -1 -2 -3 -4 -5 -5 -4 -3 -2 -1 0 1 2 3 4 5 i Group A Group B Group C Group D PC 1 PC 2 1 19 24 17 16 13 18 20 6 22 2 11 25 4 5 15 23 9 8 10 7 14 3 ii ii vi iv vii v viii i = Residual P (%) ii = Orthophosphate (%) iii = Polyphosphate (%) iv = Pyrophosphate (%) v = DNA (%) vi = Other phosphodiesters (%) vii = Phosphomonoesters (%) viii = Phosphonates (%) 21 27 26 Figure 4-8. Principal component analys is of P composition within wetland soils as determined by solution 31P NMR spectroscopy. A) Score plot of PCA 1 and PCA 2 acco unting for 41.7 and 22.5% of variance respectively, superimposed using grouping of wetlands based upon organic matter and pH. B) Loading plot visualizing role of P composition is distinguish ing variation between sites. Note site nine removed prior to analysis, due to potential error in proportion calculation. 135

PAGE 136

5 4 3 2 1 0 -1 -2 -3 -4 -5 -5 -4 -3 -2 -1 0 1 2 3 4 5 Persistent emergent, Forested Persistent emergent, Herbaceous Persistent emergent, Moss Open Water PC 1 PC 2 Figure 4-9. Principal component analys is of P composition within wetland soils as determined by solution 31P NMR analysis. Wetland grouping as based Cowardin et al. (1979) major vegetation classes. Note site nine removed prior to analysis 136

PAGE 137

Site B C D E F G 14 6.189 5.989 5.068 4.705 4.570 4.215 13 6.208 6.006 5.087 4.726 4.594 4.235 12 6.195 6.001 5.085 4.725 4.592 4.24 11 6.198 6.005 5.087 4.725 4.593 4.24 10 6.206 6.018 5.109 4.742 4.615 4.254 9 6.206 5.990 4.205 8 6.206 6.035 5.114 4.758 4.627 4.264 7 6.197 6.04 5.121 4.764 4.633 4.274 Av. 6.201 6.011 5.096 4.735 4.603 4.241 SD 0.007 0.019 0.019 0.021 0.022 0.023 Figure 4-10. Region 8 to 3 ppm within group B wetland spectra and peak assignments for; A) unidentified inositol phosphate, B) orthophosphate, C, D, E, F) myo Inositol hexakisphosphate, G) scyllo -Inositol hexakisphosphate. 137

PAGE 138

Figure 4-11. Scatter plot of microbial P against A) DNA and B) Polyphosphates. Both showing significant positive correla tion as determined by Spearmans rank correlation (rho (DNA) = 0.61, p = 0.002, rho(polyphosphates) = 0.78, p < 0.0001). Microbial P (% of total P) Pol yp hos p hate ( % of total P ) DNA ( % of total P ) Group A Group B Group C Group D A B 138

PAGE 139

CHAPTER 5 PHOSPHORUS FORMS IN HYDROLOGI CALLY ISOLATED WETLAND AND SURROUNDING PASTURE SOILS1 Introduction Diffuse phosphorus (P) loads from agric ultural ecosystems impact the ecological functioning of many inland waterways and wetland systems (Khan and Ansari 2005; Verhoeven et al. 2006). Yet wetlands within agricultural landscapes also offer a potential solution, by acting as a water and nutrient storage system at the landscapescale (Mitsch and Day 2006; Moreno et al. 2007; Paludan et al. 2002; Perkins et al. 2005). The functional role that wetlands pl ay within the landscape has been recognized at the federal level by pr ograms such as the US Army Corps of Engineers and EPAadministered Compensatory Mitigation for Losses of Aquatic Resources (www.epa.gov/wetlandmitigation/), and t he NRCS-administered Wetland Reserve Program (www.nrcs.usda.gov /programs/wrp/ ). At the state level, recognizing this has translated to the adoption of state-specific programs such as Floridas Lake Okeechobee Isolated Wetland Restoration Progr am (LOIWRP). A cost-share program under the mandate of the Lake Okeechobee Protection Progr am (LOPP) (FL Statute 373.4595), this initiative seeks to enhance and restore wetlands to retain increased amounts of water and P within the four priority basins (S-65D, S-65E, S-154 and S-191) north of Lake Okeechobee (Zhang et al. 2009a) Within the south-central Florida region, isolated wetland s represent shallow (~1 m) depressions within a landscape of very lo w (0-2%) topographic relief (Capece et al. 1 Published in a modified format by Journal of Environmental Quality, online 25 May 2010. doi:10.2134/jeq2009.0398. 139

PAGE 140

2007; Reddy et al. 1996). Although a signific ant component of the landscape, ~13,000 ha (13.8% of the land area) within the four priority bas ins (McKee 2005), these wetlands have little connectivity to surrounding surface waters. Anthropogenic alteration of drainage patterns during the expansion of cattle ranching (Steinman and Rosen 2000) means that a large proportion of these wetl ands (including the study sites) are now classed as head of ditch wetlands, with some degree of channelized outflow (Flaig and Reddy 1995). It is estimated that the surface soils (0-10 cm) of these systems store ~ 290 kg P ha-1 (McKee, 2005), with recent studies sh owing significant increases in both soil and total ecosystem P storage with increas ed hydroperiod (Dunne et al. 2007). It is believed that restoring former hydrologica l conditions will lead to increased water and nutrient storage in the land scape and, therefore, reduce the P loading to downstream waters that ultimately contribute water to Lake Okeechobee. It is important to determine the impact of restoration e fforts on current P storage in both wetland and upland soils that may experience an increased hy droperiod following restoration. Wetland ecosystems have surface soils that contain greater amounts of organic matter relative to both underlying soils and adjacent terrestrial ecosystems (Axt and Walbridge 1999; Gathumbi et al. 2005; Pant and Reddy 2001). This is due to increased organic matter accumulation within wetlands, a result of high plant productivity, their receiving landscape position, and relatively slow decomposition rate mediated by anaerobic conditions (Craft and Richards on 1993; DeBusk and Reddy 2003). Such conditions can promote the accumulation of or ganic P relative to inorganic P, leading to the generalization that P in wetland soils is dominated by organic fractions (Newman and Robinson 1999; Reddy et al. 2005). In a me ta-analysis of terrestrial grassland soils, 140

PAGE 141

organic P was found to represent 26% of tota l P (Harrison 1987). Similarly, studies of specific landscape types have shown high le vels of variation; for example, between 31 and 75% of total P occurred as organic P in a range of 29 temperate pasture soils (Turner et al. 2003e). Such variation can be attr ibuted to differences in land-use history, local climate, and pedogenic development (Ha rrison 1987; McDowell and Stewart 2006; Sumann et al. 1998; Turner et al. 2007a) and highlights the need to determine changes in P forms between wetlands and uplands in the context of a given climatic zone and agricultural management history. The biogeochemical turnover of organi c P is dependent upon its stability in the environment, a result of both abi otic and biotic processes affected by landscape position (Celi and Barberis 2005b; Wetzel 1999). Studies in a range of upland soils have shown organic P to be dominated by phosphomonoes ters (Chapuis-Lardy et al. 2001; McDowell and Stewart 2006; Murphy et al. 2 009; Turner et al. 2003a), whereas recent work has indicated a greater prevalenc e of phosphodiesters in organic matter dominated wetlands (Turner and Newman 2005; Turner et al. 2006a). Studies of upland soils have attributed a positive correlation between the proportion of organic P found as phosphodiesters (e.g., nucleic acids, phosph olipids) and annual pr ecipitation to the increased recalcitrance of phosphodiesters under wetter conditions (Condron et al. 1990a; Tate and Newman 1982). It is presumed that similar mechanisms, alongside the reduced stability of redox-sensitive iron-bou nd inositol hexakisphosphate (Heighton et al. 2008; Suzumura and Kamatani 1995a), may account for the differences observed between the organic P forms of uplands a nd wetlands observed to date (Turner and Newman 2005). In addition to a greater pr evalence of organic P within wetlands, it 141

PAGE 142

seems likely that organic P forms will vary systematically across the uplandwetland transition in response to differenc es in biogeochemical properties. Here I compare the composition and availability of P between wetland and upland soils of an agricultural cow-calf grazing system in the subtropics. The specific objectives were to: (1) determine basic soil biogeochemical and P characteristics in wetland and uplands soils of grazed subtropical pastu re; (2) quantify organic P composition across the uplandwetland interface using solution 31P NMR spectroscopy; and (3) determine the effect of hydrology on P composition and storage, with the aim of predicting impacts of increased hydroperiod. Materials and Methods Site Description Four study sites were located on two cow-calf ranches, Lars on-Dixie and Beaty, north of Lake Okeechobee in south-central Florida (Figure 5-1) Surrounding pasture uplands are unimproved cow-calf operations wit h typically low stocking densities of ~1 animal unit ha-1 (Bhadha and Jawitz; Gornak and Zhang 1999) and ( M. Flinchum, personal communication, 2004). The unfenced wetlands, two on each ranch, were similar in size, 1 to 2 ha, and supported similar vegetation communities, with deep marsh zones (see soil sampling) dominated by open water species (e.g. Pontedaria cordata var. lancifolia (Muhl.)), shallow marsh zones dominated by flood-tolerant species (e.g. Juncus effusus L.), and surrounding pasture uplands being predominantly Paspalum notatum Flugge. Soils at Lars on Dixie are mapped as si liceous, hyperthermic Spodic Psammaquents (Basinger series). Soils in the Beaty uplands are sandy siliceous hyperthermic Aeric Alaquods (Myaka sands), while the study wetland are delineated as a Basinger and Placid soils depressional association with Placid soils being sandy, 142

PAGE 143

siliceous, hyperthermic Typic Humaquepts. A ll soils are sandy textured with high infiltration rates yet low internal drainage given the typically high water table (Lewis et al. 2003) Soil Sampling In March 2007, during a period of draw down, each wetland area was stratified into three zones based on soil, vegetation and hydrol ogic indicators: pasture upland, shallow marsh, and deep marsh. Within each zone, th ree locations were selected randomly for quadrat sampling. Within each quadrat, three intact soil cores (diameter 7.5 cm 10 cm) were collected and amalgamated. Sample s were transported to the laboratory on ice and homogenized prior to sorting. R oots and recognizable or ganic fragments > 2 mm were removed by hand and the soil stored at 4C. Subsamples were oven dried (70C, 72 h), sieved and ground for tota l elemental analysis (see below). Hydroperiod Determination Soil sampling elevations were compared to the elevation of a groundwater well located in the center of each wetland. Using wetland water level elevations, I determined how many days per year a specif ic soil sampling location was saturated and termed this estimate hydroperiod. Water le vels in Larson Dixie wells were recorded from July 2003 to April 2006 and in Beaty we lls from December 2003 to March 2006. Soil Biogeochemical Properties Soil water content was determined as weight loss following drying at 70C for 72 h. Soil pH was determined on a 1:2 soil to water suspension using a glass electrode and soil bulk density was calcul ated using the known sample volume (3 cores, 7.5 cm diameter, and 10 cm deep) and determined wa ter content. Total P was determined by combustion of soil at 550C in a muffle furnac e for 4 h, dissolution of the ash in 6 mol L-1 143

PAGE 144

HCl (Andersen 1976), and then detection of molybdate reactive P (MRP) using a segmented flow analyzer (AAII Technico n, SEAL Analytical, UK) and standard molybdate colorimetry (USEPA, 1993). A sub-sa mple of the acid solution was analyzed for Al, Ca, Fe, and Mg on an inductively-coup led plasma optical-emission spectrometer (ICPOES) (Thermo Jarrell Ash ICAP 61E, Franklin, MA). Total soil C and N were measured by combustion and gas chromato graphy using a Flash EA1112 (Thermo Scientific, Waltham, MA). Phosphorus Characterization Phosphorus forms were determined by two parallel methods: (1) a single step process to determine acid extractable inorganic P by extraction in 1 mol L-1 HCl for 3 h (Reddy et al. 1998) and (2) a combinati on of anion exchange membranes (AEM) and solution 31P Nuclear Magnetic Resonance (NMR) s pectroscopy. This method utilizes HCO3 loaded AEM strips as an initial extr action procedure for exchangeable P (Myers et al. 1999), followed by the addition of a solution containing NaOH and EDTA to proceed with a standard alkaline extraction for 31P NMR analysis (Cade-Menun and Preston 1996). In the AEMNMR method, duplicate fresh samples (3.5 g dry weight equivalent) were measured into pre-weighed 250 mL HDPE centrifuge bottles and distilled deionized water (DDI) was added to bring th e water content to 74 mL. One set of samples received a further 1 mL of DDI (non fumigated), while a parallel set received 1 mL of 1-hexanol (95%, Sigma Aldrich, St Louis, MO)1 mL hexanol (fumigated). This allows for the determinati on of both exchangeable P (PAEM) and, by difference, a measure of fumigation-released or microbial P (PM) (Kouno et al. 1995; Myers et al. 1999). All samples received a single 6.25 x 1.5 cm AEM strip (BDH Prolabo Product 144

PAGE 145

number: 551642S, VWR International, UK), which were prepared by shaking 25 strips for 24 h in three sequential c hanges of 200 mL of 0.5 mol L-1 NaHCO3. The bottles were sealed and samples shaken on a reciprocating table for 24 h. The AEM strips were then removed, rinsed of adhering material with DDI, shaken dry and the phosphate eluted by shaking for 3 h in 50 mL of 0.25 mol L-1 H2SO4 (Cheesman et al. 2010b; Turner and Romero 2009a). The concentration of phosphate was determined using a discrete autoanalyzer (AQ2+, SEAL Analytical, UK) and standard molybdate colorimetry (USEPA 1993). The AEM strips recove red ~100% of an orthophosphate standard solution (500 g P /membrane) and 97% of an orthophosphate spike added to replicate soil samples (100 g P to 3.5 g soil/membrane). It should be noted that in addition to inorganic orthophosphate, the membrane strips may have recovered small concentrations of labile organic P (CHAPTER 3) After removing the AEM strips, DDI wa s added to the unfumigated samples to bring the water content to 100 mL. To this, 5 mL of a solution containing 5.25 mol L-1 NaOH and 1.05 mol L-1 EDTA was added to give a final concentration of 0.25 mol L-1 NaOH and 50 mmol L-1 EDTA in 1:30 soil to solution rati o. The samples were shaken for 4 h and then centrifuged at 7000 rpm (maximum RCF ~7500 g) (Sorvall RC6, SL1500 Rotor; Thermo Fisher Scientific, Walt ham, MA) for 10 min. The supernatant was decanted and each sample analyzed for total P (NaOHTP) using a modified double-acid digest (Rowland and Haygarth 1997). Three mL of extract was pipetted into a boiling tube to which 1 mL of conc. H2SO4 and 1 mL of conc. HNO3 were added. After placing on a heat block to reduce the volume to ~0.5 mL, samples were refluxed at 550C for at least 3 h before being diluted by a factor of 100 in DDI and analyzed for orthophosphate 145

PAGE 146

by molybdate colorimetry (see above). Al kaline extracts were not analyzed for molybdate reactive P due to error associated with phosphate detection by this procedure in high organic matter soils (Turner et al. 2006b). Solution 31P Nuclear Magnetic Resonance Spectroscopy Replicate alkaline extracts were combi ned on an equal volume basis within each of the three zones in a giv en wetland, resulting in 12 am algamated samples for solution 31P NMR analysis. Each extract (15 mL) wa s spiked with 1 mL methylenediphosphonic acid (MDP) (50 g P mL-1) as an internal standard, frozen at C, and lyophilized. Approximately 100 mg of lyophilized material was resuspended in 0.1 mL D2O and 0.9 mL of a solution containing 1 mol L-1 NaOH and 100 mmol L-1 EDTA, and loaded into a 5 mm NMR tube to ensure consistent chemical shift and strong signal lock. Solution 31P spectra were acquired using a Bruk er Avance 500 MHz spectrometer, with a broadband probe operating at 202.45 MHz. Spectra were accumulated using waltz decoupling (zgig program) with a 4.0 s pulse (~30), 0.4 s acquisition time and 1.0 s delay. Between 18,000 and 60,000 scans were needed for good signal to noise ratio dependent upon the P concentration in the sample. Spectra were analyzed with NMR Utility Software (NUTS) initially using 15 Hz line broadening. Spectra were phased, corrected fo r baseline shift, and calibrated to the MDP internal standard (chemical shift ( ) = 17.47 ppm determined when a spiked sample was calibrated against externally held 85% H3PO4 set at 0 ppm). Spectra were integrated over specific intervals to det ermine functional P groups (Table 5-1) based upon established literature (T urner et al. 2003d). The region between 3 and 8 ppm was further analyzed using the deconvolution utility of NUTS software. A best fit deconvolution of the spectra was acquired us ing 3 Hz line broadening and peak picking 146

PAGE 147

parameters of 5% of maximum peak height and 0.5 for the root mean squared noise parameter. The region was split into orthophosphate and phosphomonoesters (all other peaks determined by the algorithm in the regi on 3 to 8 ppm). Peak proportions from deconvolution were then applied to the inte rval integration determined in the 15 Hz spectra. Repeated integration of the same spectrum provided confidence in the detection of signals within the pyrophosph ate region equivalent to 1 mg P kg-1 soil. Data Analysis All statistical tests were performed in SPSS for windows version 17.0.0 statistical software (SPSS Inc., 2008). Both visual inspection and the Shapiro-Wilk test were used to test normality. If requir ed, a natural log transformation was applied before statistical analysis. Basic biogeochemical characteri stics and forms of P were analyzed with a simple univariate ANOVA with landscape positi on as the main factor and wetland site (Larson East/West, Beaty Nort h/South) as a random factor. Post hoc analysis (Tukey HSD test) was applied to explore any si gnificant relationships determined between landscape positions. Regression between ca lculated hydroperiod and biogeochemical characteristics were explored by the fitti ng of linear, exponential and linear segmented curves to the data. Results and Discussion Soil Biogeochemical Characteristics As expected, hydroperiod varied between landscape positions, with the deep marsh zone being flooded on average 63% of the year, the shallow marsh being flooded 25% of the year, and the uplands being rarely if ever flooded (Table 5-2). Deep marsh zones had significantly lower pH, lower bulk density and higher organic matter concentration (as indicated by loss on ignition) than other landscape positions ( p < 147

PAGE 148

0.05). Of the total metals anal yzed, Mg was at or below t he practical quantification limit for the majority of samples (data not show n). Calcium showed no significant trends across landscape position, whereas Al and Fe concentrations showed a significant increase within wetlands (p < 0.05). Post hoc analysis suggested that deep marsh areas were the most distinct, cont aining concentrations of up to twice as much Fe and 10 times as much Al compared to surrounding uplands. This is consistent with positive correlations between organic matter and Fe/Al within other wetlan d systems (Darke and Walbridge 2000) and I hypothesize that this represents stable organo-metal complexes in the more organic deep marsh soils. The expected gradient in accumulating organic matter, was supported by analysis of tota l C and total N (Table 5-2), both of which increased significantly ( p < 0.05) with landscape posit ion towards the wetlands deep marsh. There was also significant ( p < 0.05) interaction observed between landscape position and wetland site, attributed to longer hydroperiods within the Beaty wetlands deep marsh (average 239 days) compared to Larsons (194 days) (data not shown). For most soil properties, significant di fferences among landscape positions were the result of distinct deep marsh characteri stics. With the exception of total N, post hoc analysis demonstrated only limited, non-significant differences between the pasture upland and shallow marsh zones (s ee: Impact of hydroperiod). Soil Phosphorus Composition There was a significant difference in total P concentration across landscape position, with an approximate threef old increase from 117 mg P kg-1 in the pasture uplands to 371 mg P kg-1 in the deep marsh ( p < 0.05) (Table 5-3, Figure 5-2). This significant gradient was also observed in the HCl-extractable inorganic P, which ranged from 23 mg P kg-1 (20% total P) in the uplands to 61 mg P kg-1 (16% total P) in the deep 148

PAGE 149

marsh. The AEMNMR method extracted on av erage 53% 11% of total P in all soils sampled, with no significant bias in the recovery rate among landscape positions. Phosphorus not extracted (residual P) was considered a distinct unidentified pool, hypothesized to consist of recalcitrant organic and alkaline-stable mineral forms (CadeMenun 2005b; Turner et al. 2005). Exchangable phosphate (PAEM) showed a significant in crease from 4.4 mg P kg-1 (3.8% total P) in the uplands to 23.9 mg P kg-1 (6.4% total P) in the deep marsh ( p < 0.05). Microbial P showed a significant increas e in concentration, but a reduction in its proportion of total P ( p < 0.05), from 22 mg P kg-1 (19% total P) in uplands to 37 mg P kg-1 (10% total P) in the deep marsh. Although the concentration of NaOHTP changed, the proportion of it relative to total P did not differ significantly with landscape position, ranging from 56 mg P kg-1 (48% total P) in the uplands to 184 mg P kg-1 (50% total P) in the deep marsh. This pattern was repeated in the residual P fraction, with a significant ( p < 0.05) difference in concentration, yet sim ilar proportions of total P, from 56 mg P kg1 (47% total P) in uplands to 163 mg P kg-1 (43% total P) in the deep marsh. Although concentrations of all P pools determined by the AEMNMR method (PAEM, PM, NaOHTP and residual) showed significant differences across landscape positions, the difference between pasture uplands and shallow marsh were not significant. Given the trends in data across landscape position, I attribute this lack of significance to the high variability within soils collected from the shallow marsh zone (see: Impact of Hydroperiod and Figure 5-3). Solution 31P Nuclear Magnetic Resonance Spectroscopy Solution 31P NMR spectra showed the pr esence of orthophosphate, phosphomonoesters, phosphodiesters, pyrophosph ate and trace concentrations of 149

PAGE 150

phosphonates and polyphosphates (Figure 5-3, Table 5-4). Approxim ately 73% of the extracted P was organic, wit h phosphomonoesters constituting the major fraction, ranging from 50 mg P kg-1 in the uplands to 127 mg P kg-1 in the deep marsh. The lack of a characteristic 1:2:2: 1 signature within the phosphomonoes ter region indicated the absence of detectable concentrations of myo -Inositol hexakisphosphate (Turner et al. 2003f). The remaining organic P included phosphodiesters (11% total soil P), dominated by DNA, with the re mainder representing various alkali-stable phospholipids. Due to the rapid hydrolysis of certain co mpounds in alkaline extracts (i.e. RNA and some phospholipids) the proportion of organic P determined as phosphod iesters is likely to be underestimated (Turner et al. 2003d). To tal inorganic P constituted between 16 and 19% of total P. This was dominated by orthophosphate, while pyrophosphate was detected in all spectra at concent rations ranging from 3.4 mg P kg-1 (2.0% total P) in the shallow marsh to 6.7 mg P kg-1 (1.8% total P) in the deep marsh. Polyphosphate was detected at low concentrations in cert ain deep marsh soils (average 1.5 mg P kg-1 soil). The concentrations of phosphomonoeste rs, DNA, and orthopho sphate increased significantly from uplands to deep marsh ( p < 0.05; Table 5-4). However, there was a striking lack of distinction between the rela tive proportions of P forms identified across landscape position. Instead of the expecte d changes in forms and magnitudes of P between wetlands and uplands, all sites contai ned similar forms of P, leading me to reject my initial hypothesis. The proporti on of P identified as phosphodiesters did not increase in the deep marsh as expected from previous studies of wetland soils (Turner and Newman 2005; Turner et al. 2006a). In additi on, the marked lack of the distinctive peak signature associated with the various is omers of inositol hexakisphosphate, even 150

PAGE 151

in upland pasture samples, was in contrast to other studies of upland grasslands (McDowell and Stewart 2006; Murphy et al. 2 009; Tate and Newman 1982; Turner et al. 2003e). The abiotic sorption and stabilization of inositol phosphates in such environments is attributed to their interact ion with clay minerals and amorphous Fe and Al oxides (Celi and Barberis 2005b, 2007; Heighton et al. 2008). In addition to data presented, which reports t he low Fe and Al concentrations throughout the system (Table 5-2), existing soil surveys (Lewis et al. 2003) and detailed particle size distribution analysis (Bhadha and Jawitz, 2010) show only minimal clay content (< 2.5%) in the uplands. Therefore, it is likely that inositol phosphates are not physically stabilized within these soils and in stead are hydrolyzed rapidly. Impact of Hydroperiod The use of a priori categorization using vegetation, soils, and hydrology of the wetlandupland landscape contin uum is a qualitative appr oach to classifying the different landscape positions (deep marsh, shallow marsh, and upland). The use of a calculated hydroperiod is a quantitative approach, giving an absolute value for a hydroperiod of a given location. The SPSS curve estimating procedure was appl ied to test the significance of linear regressions between the calc ulated hydroperiod and various soil P characteristics. There was a positive linear relationship between soil total P and hydroperiod (R2= 0.734; p < 0.001; Figure 5-4 A). The potential for a more complex segmented relationship with a critical threshold hydroperiod at approxim ately 100 days was explored by the fitting of a segmented li near regression; however, there was not a substantially better fit to the data (R2 = 0.758). 151

PAGE 152

Phosphorus forms determined by solution 31P NMR spectroscopy were plotted against the average hydroperi od of the amalgamated so ils used to generate the spectra. Total organic P determined by 31P NMR analysis showed a significant positive linear relationship with hydroperiod (R2 = 0.837; p < 0.001; Figure 5-4 B). Similar relationships were found for phosphomonoest ers and phosphodiesters (Figure 5-4 C, D). Although total organic P and its major forms showed similar positive trends in concentration with calculated hydroperiod, I had initially hypothesized that increasing hydroperiod would result in increased stabiliz ation of organic P and as a result, lead to an increase in its proportion of to tal P, as well as the pref erential stabilization of forms characteristic of wetter soils (i.e. phosphodie sters). However, I observed no evidence of a significant relationship between hydroperiod and organic P (as a proportion of total P), or the relative proportions of phosphodiesters to phosphomonoesters. Across landscape position there was a genera l increase in total soil P towards the wetland deep marsh, but the conc omitant increase in all parti tioned forms of P suggests a mechanism of general accumulation, as opposed to a preferentia l stabilization of specific forms. The highly significant ( p < 0.001) linear relationship between total C and total P (Figure 5-5) suggests that the a ccumulation of P is concurrent with the accumulation of organic matter. Previous analysis of manure-impacted surface soils in the region showed that the upper soil horizons consist of uncoated quartz sand grains, and the low clay concentrations are dominated by non-crystalline silica (Harris et al. 1996; Harris et al. 1994). Although deeper hor izons may adsorb P strongly (Graetz and Nair 1995), mineral components within the su rface soils have a very low P binding capacity, resulting in P dynamics driven by an association with organic matter. 152

PAGE 153

The accumulation of organic matter in response to increased hydroperiod may be a result of anaerobosis or increased endogenous and exogenous inputs. The transportation of particulate P in runoff has long been recognized as a significant component of P transfer (Daniel et al. 1998) and, therefore, a focus for management practices to reduce P loss from agricultural landscapes (McDowell et al. 2007; Sharpley et al. 2001). The forms and natur e of P transported in surf ace runoff is dependent upon soil biophysical characteristics and site hydr ological properties (Ba llantine et al. 2009; Heathwaite and Dils 2000). The direct determination of surface runoff at dispersed wetland fringes with low over all topography is problematic although indirect modeling suggests that overland flow may play a cons iderable role in the water budget of these study wetlands (J. H. Min, personal communica tion, 2010). With direct overland flow accounting for up to 25% of water inputs, ofte n associated with isolated high intensity precipitation events. Such hi gh energy rainfall events cont ribute disproportionately to increased movement of particulate P in high ly sloped tilled silt/loam soils (Jin et al. 2009). I am unaware of studies that emulate the site conditions, i.e. that include contiguous ground cover, a low topographic gr adient, and soils dominated by siliceous sand grains and variously sized organic matter particles. In addition, all study wetlands were unfenced within grazed pas tures. Given known cattle us e of wetlands throughout the year (Pandey et al. 2009) and associated increases in potential sedimentation, vegetation turnover (McDowell et al. 2007), and direct fecal additions, the influence of cattle on organic matter accumulation may be significant. The importance of organic matter stabilization of P, even in the low organic matter upland soils, as well as the potential translocation of this organic matter, could account 153

PAGE 154

for the absence of any distinct shifts in P forms between landscape position. Further work is required to investigate the ro le of endogenous and exogenous inputs to the carbon and P dynamics of the receiving wetlands. Phosphorus Storage For the restoration of hydrologic condition s in isolated wetlands to be an effective P management practice, wetl ands would need to show evidence for increased nutrient storage on an aerial basis (kg ha-1) relative to other landscape positions. Although moderated by a reduced bulk density in wetl ands relative to uplands (Table 5-2), storage of both total and organi c P is significantly ( p < 0.05) elevated within the surface soils of the wetlands deep marsh (Table 5-5) with a pronounced twofold increase in total P from 114 kg ha-1 in the unimproved pasture to 236 kg ha-1 in the deep marsh areas. The marginal, but not significant, in crease in P content between pasture uplands areas and shallow marsh suggests that me aningful nutrient storage requires a substantial increase in hydr operiod. In addition, compari ng current landscape storage does not take into account the time-scale required for its development. It can be assumed that an increase in hydroperiod w ould necessitate a period of adjustment during which P and C would be accreted. The rate and controlling mechanisms of this accretion would have significant impact on t he efficacy of restored wetlands to reduce downstream P loading. Conclusions Isolated wetlands exhibited an accumula tion of organic matter as compared to surrounding pasture uplands, which was clos ely associated with an accumulation of total P. However, this was not related to differences in soil P forms predicted from differential stability across the landsc ape. Instead, total P accumulation was 154

PAGE 155

155 concomitant with an increase in all P forms, including exchangeable P, microbial P, and forms identified by alkaline extraction and 31P NMR spectroscopy. Successful management of isolated wetlands for the onsit e sequestering of P from agricultural runoff will depend on the accumulation and stabilizat ion of organic matter. This study suggests that a substantial hydroperiod increase is required to sequester increased amounts of P.

PAGE 156

Table 5-1. Specific integral ranges used in the classification of solution 31P NMR spectra for lyophilized material resuspended in 0.9 mL (1 mol L-1 NaOH 100 mmol L-1) + 0.1 mL D2O. Internal standard methylenediphosphonic (MDP) calibrated to = 17.46 ppm. Region 8 to 3 ppm is determined as orthophosphate or phosphomonoesters after deconvolution of spectr a at 3 Hz line broadening. Chemical shift Attributed phosphorus group 21.0 to 20.0 Phosphonates 18.0 to 16.8 MDP (methyl enediphosphonic acid) 8.0 to 3.0 Orthophosphate + phosphomonoesters Line broadening =3Hz Orthophosphate 6.24 0.01 ppm Phosphomonoesters = [8 to 3] orthophosphate 3.0 to -1.5 Phosphodiesters 0.5 to 1.5 DNA [3.0 to -1.5] DNA Other posphomonoesters -3.0 to -5.0 Pyrophosphate and end-c hain polyphosphate functional groups -18.0 to -21.0 Poly phosphate mid chain functional groups 156

PAGE 157

Table 5-2. Soil characteristics and nutrient s determined in samples across landscape position. Values represent averages (n = 12) 1 standard deviation taken as three replicates from four wetland sites. Pasture Upland Shallow Marsh Deep Marsh Hydroperiod (days)*** 2a 6 92b 67 231c 57 pH* 4.8a 0.6 4.7a 0.4 4.3b 0.3 Bulk density (g cm-3 ) 0.98a 0.10 0.92a 0.21 0.65b 0.19 Organic matter (g kg-1) 86a 23 123a 71 219b 66 Total elements (g kg-1) Carbon* 44a 12 58a 34 112b 37 Nitrogen* 2.8a 0.5 3.6b 1.2 4.8c 0.9 Calcium 1.1 0.5 1.2 1.0 1.9 0.6 Iron* 0.7a 0.3 0.6a 0.4 1.1b 0.3 Aluminum *** 0.3a 0.2 0.7a 0.8 3.6b 0.8 Organic matter estimated by loss on ignition Significant at the 0.05 probability level *** Significant at the 0.001 probability level Table 5-3. Phosphorus forms across lands cape position determined by acid extraction and AEMNMR method. Values repres ent averages (n=12) 1 standard deviation. Pasture Upland Shallow Marsh Deep Marsh Total P *(mg kg-1) 117a 25 171a 107 371b 61 HCl-Pi (% total P) 20 5 19 4 16 2 AEM-NMR Method (% total P) PAEM 4 2 6 3 6 2 PM 19a 5 16a 10 10 b 4 NaOHTP 49 10 42 8 51 13 Residual 47 11 52 8 43 13 Inorganic P extracted in 1 mol L-1 HCl Exchangeable (PAEM) and microbial (PM) determined by AEM extraction Alkaline extracted total P = (PAEM + NaOHTP). significant at the 0.05 probability level 157

PAGE 158

158 Table 5-4. Phosphorus forms determined by solution 31P NMR spectroscopy of amalgamated alkaline extracts from ac ross the landscape transition. Values from each landscape position repres ent averages (n= 4) 1 standard deviation Phosphorus form Pasture Upland Shallow Marsh Deep Marsh mg kg-1 Phosphonate 0.6 1.2 trace 0.9 1.4 Orthophosphate* 18a 6 24a 11 57b 9 Phosphomonoester* 50a 5 58a 16 127b 21 DNA* 6.4a 1.9 14.2a 2.2 29.7b 11.0 Other phosphodiesters 8.8 7.2 10.1 0.7 12.5 7.6 Pyrophosphate 5.3 2.6 3.4 2.4 6.7 3.8 Polyphosphate nd nd 1.5 2.9 Organic P (% of total NMR) 73.7 74.9 72.2 significant at the 0.05 probability level n.d not detected by 31P NMR spectroscopy Table 5-5. Storage of total (n = 12) and organic phosphorus (n = 4) within the top 10 cm of soil, across landscape transition. Values from each landscape position represent averages 1 standard deviation Phosphorus (kg ha-1) Pasture Upland Shallow Marsh Deep Marsh Total P* 114a 27 141a 50 236b 62 Organic P* 65a 10 75a 19 108b 33 significant at the 0.05 probability level

PAGE 159

B A Figure 5-1. Location of study sites showing A) Florida outline with area of interest north of Lake Okeechobee, and B) detail of ranch sites containing two study wetlands each, within priority basins north of Lake Okeechobee. Beaty Ranch is within sub-basin S-65D of the Ki ssimmee River watershed, with hydraulic connection through Cypress slough to Canal C38 and then to the Lake. Larson Dixie ranch is located in sub-basin S-154 within the Taylor Creek watershed, connected through Mosquito Creek to Taylor Creek itself 159

PAGE 160

* Soil P (mg P kg-1) *Figure 5-2. Phosphorus pools determined by AEMNMR method across the three landscape positions. Values are averages (n=12) 1 standard deviation. No significant differences were found between Pasture Upland and Shallow Marsh. (*) denotes significant differ ences between pools of the different landscape positions ( p < 0.05). 160

PAGE 161

161 Pasture Upland Shallow Marsh Deep Marsh 0 -10 -20 10 20 Chemical Shift (pp m ) Figure 5-3. Example solution 31P NMR s pectra from amalgamated samples from the Beaty North wetland. Spectra acquired using a Bruker Av ance 500 MHz with a 4.00 s pulse (~30), 0.4 s acquisition time and 1 s delay. Spectra referenced and integrated against inte rnal methylenediphosphonic acid (MDP) standard ( = 17.46 ppm) displayed here using 15 Hz line broadening and scaling to MDP.

PAGE 162

A B Phosphorus (mg P kg -1) C R2 = 0.734 R2 = 0.837 p < 0.001 p < 0.001 R2 = 0.794 p < 0.001 D Hydroperiod (days) R2 = 0.697 p = 0.001 Figure 5-4. Phosphorus characteristics pl otted against hydroperiod: A) total phosphor us for all samples, B) total organic phosphorus determined by 31P NMR spectroscopy, C) phosphomonoeste rs, and D) phosphodiesters. 162

PAGE 163

Total C (g C kg 1)Total P ( m g P k g 1 ) Total P = 2.9 x (t otal C) + 13.8 R2 = 0.82 Figure 5-5. Total soil carbon plotted against phosphorus concentrations across landscape positions (n = 36) from four wetland sites. Characteristics show a significant ( p < 0.001) positive relationship (R2 = 0.819). 163

PAGE 164

CHAPTER 6 STABILITY OF SELECT BIOGENIC PHOS PHORUS COMPOUNDS UNDER AEROBIC AND ANAEROBIC CONDITIONS Introduction The differential stability of biogenic P under wetland conditions, compared to terrestrial soils has been invoked as a potential mechanism to explain observed patterns seen in wetlands (CHAPTER 4). In brief, anaerobic conditi ons are believed to lead to reduced decomposition of certain biogenic P forms (i.e. phospholipids and polynucleotides) while reducing the stability of others (i.e IP6 and polyphosphates). This dichotomous impact of anaerobic conditions on biogenic P forms is currently unproven, yet has been suggested in some form by a number of authors (Hupf er et al. 2007; McKelvie 2007; Mitchell and Baldwin 20 05; Turner and Newman 2005). Here I use solution 31P NMR spectroscopy to test this by determining the difference in turnover over rates of major biogenic P forms (DNA, myo-IP6, and polyphosphate) under both aerobic and anaerobic conditions. Several studies looking at biogenic P forms within terrestrial systems have postulated a reduced decomposition of phosphodiesters (including DNA) under wetter and by extension more anaerobic conditions (Condron et al. 1990a; Tate and Newman 1982), with the changes in phosphodiester deco mposition acting as a driver in determining patterns of biogenic P seen across climatic gradients (Sumann et al. 1998). In wetlands, typified by anaerobic condition s, there is a greater prevalence of phosphodiesters (CHAPTER 2; Turner and Newm an 2005), yet it is unclear whether this represents decreased decomposit ion due to biotic hydrolysis, greater or altered microbial biomass (Makarov et al. 2002a), abiotic stabilization of DNA to generally 164

PAGE 165

increased organic matter content (Celi and Bar beris 2005b), or a relative increase due to reduced stabilization of phosphomonoesters. Inositol hexakisphosphate, is a major co mponent of biogenic P in many terrestrial systems (Murphy et al. 2009; Turner 2007; Turner et al. 2002b), but appears to be absent within many wetland soils studied to date (CHAPTER 2). Yet the fundamental process that underlies the presence or absence of IP6 within wetlands is currently unknown (McKelvie 2007). One possibility lies in the observations of Suzumura and Kamanti (1995a,b). They observed that under anaerobic conditions in marine substrates, terrigenously derived IP6 underwent rapid degr adation (Suzumura and Kamatani 1995b), with complete breakdown in 40 days, as compared to just 50% degradation in 60 days under aerobic condit ions (Suzumura and Kamatani 1995a). Although, such anaerobic mediated degradati on appears contrary to recent observation of IP6 as a conservative paleoindicator wit hin estuarine sediments (Turner and Weckstrm 2009) and as a component of freshwater, lacustrine (Golterman et al. 1998; McKelvie 2007), riverine (McDowell 2009) and palustrine soils studied as part of this dissertation (CHAPTER 4). It therefore important to estab lish if this apparent redox sensitivity is a universal mechanism in wetlands or represents a interaction between redox conditions and other biogeoc hemical characteristics. Finally, biogenic polyphosphates although initially intracellular, appear to form discrete extracellular complexes (Diaz et al 2008) that are stabiliz ed in the environment by their interaction with the abiotic matr ix, including redox-sensitive components such as amorphous iron oxides (Hupfer et al. 2007; Reitzel et al. 2007; Sannigrahi and Ingall 2005). It has been proposed that the interacti on of these stabilized polyphosphates with 165

PAGE 166

fluctuating redox conditions in sediments could account for a sizable component of benthic P flux (Hupfer et al. 2004; Sanni grahi and Ingall 2005). Since polyphosphates have been observed in palustrine wetlands (Bedrock et al. 1994; Sundareshwar et al. 2009) it is important to dete rmine if differential redox condi tions may account for their presence or absence within particular wetla nd systems (CHAPTER 4), as well as to help establish the ecological importance of this P form within palustrine soils. Anaerobic conditions are postulated to directly impact a num ber of important biogenic P pools, yet little direct confirmation exists that stability and turnover of these forms is impacted by the presence or absence of O2. This study seeks to determine the short-term impact of both aerobic and anaer obic conditions on biogenic P within freshwater wetland soils, spec ifically focused on the fate of DNA, polyphosphates, and myo-IP6. Methods Microcosm Experiment Stability of biogenic P was determined wit h the use of microcosms held under aerobic and anaerobic conditions for up to 48 days. In deta il (Figure 6-1), 1 g of a standard homogenized wetland soil (Table 6-1) was weighed into a 30 mL HDPE centrifuge tube. Distilled deionized water (DDI ) was added to bring total water content to 10 mL (volumetric water content 91%) bef ore tubes were sealed with a septumcontaining cap. Aerobic or anaerobic conditi ons were developed by purging tubes for 5 min using either purified and filtered (0.2 m GF-B in line filter) hydrocarbon free air (Praxair Inc., Danbury CT) or ultra high purity N2 (Airgas, Radnor, PA). Anaerobic treatments were further sealed in a polythene glove bag also purged with N2. Samples were allowed to equilibrate fo r 2 weeks in the dark at 27C and with gentle agitation 166

PAGE 167

(G25 incubator shaker, New Brunswick Scient ific Co. Inc, NJ). Head spaces were purged in a similar manner every 3 to 4 days found to maintain an O2 free state in anaerobic treatments by monitoring a subset using gas chromatography (GC-8A TCD, Shimadzu, Japan). Biogenic Phosphorus Spikes After equilibration, samples were spiked with either 2 mL of DDI (control), or 2 mL of a biogenic standard (~200 g P g-1) in a DDI matrix. Standards (Sigma-Aldrich Corporation, St Louis, MO) were either myo -IP6 (dodecasodium salt) or a mixture of DNA (deoxyribonucleic acid, from salm on testes) and polyphosphate (sodium hexametaphosphate 96%) (Figure 6-2). Subsequent analysis of representative microcosms for total P (see Biogeochem ical Characterization) showed clear repeatability of standard compound spike (Tabl e 6-2). The experiment represented a factorial design with duplicate samples being subjected to each treatment combination. After spiking, microcosms were maintained under conditions as described above and destructively sampled and analyzed for P co mposition at days 1, 8, 16, 30 and 48. Biogeochemical Characterization Water content was calculated using gravim etric loss after drying sub-samples at 75C for 72 h. Total P was determined by combustion of soil at 550C in a muffle furnace for 4 h, dissoluti on of the ash in 6 mol L-1 HCl (Andersen, 1976) and the detection of MRP using a discrete auto analyzer (AQ2, SEAL Analytical, UK) and standard molybdate colorimetry (U SEPA, 1993). A sub-sample of the acid solution was analyzed for Al, Ca, Fe, and Mg using IC POES (Thermo Jarrell Ash ICAP 61E, Franklin, MA). Total soil C and N were measured by combustion and gas chromatography using a Flash EA1112 (T hermo Scientific, Waltham, MA). 167

PAGE 168

Phosphorus Composition Destructive sampling and analysis for P co mposition consisted of a modified AEMNMR method (CHAPTER 5). Briefly, individual tubes received a sample-specific volume of DDI to bring total water content to 20 mL. To this was added one 6.25 x 1.5 cm anion exchange membrane strip (BDH Prolabo Product number: 551642S, VWR International, UK) preloaded with a HCO3 counter ion. Samples were sealed and shaken on a reciprocating shaker at low speeds for 24 h. Membrane strips were removed, rinsed of adhering material and eluted using 50 mL 0.25 mol L-1 H2SO4. Membrane eluants were analyzed for MRP (PAEM) as above. Residual samples received further sample-specific DDI to bring total wate r volume to 28.57 mL. To this, 1.43 mL of 5.25 mol L-1 NaOH 1.05 mol L-1 EDTA was added to result in a final alkaline extraction of 0.25 mol L-1 NaOH and 50 mmol L-1 EDTA in a 1:30 soil to solution ratio. Samples were shaken for 4 h, before being centrifuged at 7000 rpm (maximum RCF ~7000 g) (Sorvall RC6, SL600 Rotor; Thermo Fisher Scientific, Waltham, MA) for 12 min. A subsample of supernatant was drawn off and st ored at 4C for determination of total P (PNaOH) using a double acid (H2SO4 HNO3) digest (Rowland and Haygarth 1997) and molybdate colorimetry. A second subsample (5 mL) was combined within duplicates, mixed with 1 mL of an MDP standard (47.6 g P g-1) and immediately frozen (-80C) prior to lyophilization and NMR analysis. Solution 31P Nuclear Magnetic Resonance Spectroscopy Given the prohibitive costs associated with 31P NMR spectroscopy, initial spectra acquisition was restricted to day 1 and day 48 samples. Given the unexpected recovery of polyphosphates by the anion exchange membrane strips (see Phosphorus Recovery), and the limited differences in degradation of DNA between experimental 168

PAGE 169

treatments further analysis concentrat ed on samples that had received the myo -IP6 spike and had been extracted on days 8, 16 and 30. All lyophilized samples were reconstituted using ~ 100 mg of lyophilized material per mL with the re-suspension media being 0.9 mL (1 mol L-1 NaOH and 100 mmol L-1 EDTA) and 0.1 mL D2O. Samples were vortexed for 1 mi n prior to loading into appropriate NMR tubes. Initial day 1 and 48 samples were analyzed using a 5 mm BBO probe, Bruker Avance 600 Console, Oxford 14.1T/51 mm Magnet with subsequent analysis carried out using a 10 mm BBO probe, Bruker Avance 500 MHz 11.8 T/ 54 mm bore. All analysis used a simple zgig program employing broad band dec oupling (waltz 16) a calibrated 45 pulse width and 2 s T1 delay. Sample spectra were referenced using internal standard MDP ( = 17.46 ppm) and integrated over standard interv als (Turner et al. 2003d). The region 8 to 3 ppm was also analyzed using spectral line deconvolution enabli ng the quantification of orthophosphate, myo -IP6 (Turner et al. 2003f) and scyllo -IP6 (Turner 2007; Turner and Richardson 2004). Data Analysis All statistical tests were performed in SPSS for windows version 17.0.0 statistical software (SPSS Inc., 2008). Exchangeable P, as determined by anion exchange membrane strips, was analyzed for trends using univariate ANOVA within each spike treatment, with time steps (1, 8, 16, 30, 48) and condition (aerobi c, anaerobic) as fixed factors. Post hoc analysis (Tukey HSD) was used to explore significant differences. Concentrations of myo -IP6 determined across all time points were explored by simple linear regression with time. 169

PAGE 170

Results and Discussion Phosphorus Recovery The AEM-NMR method showed a good ability to describe P composition within samples, with recovery of total P (PAEM + PNaOH) across all treatments and time periods of 62.4 10.0% ( 1 standar d deviation), the residual being considered recalcitrant organic and alkaline stable inorganic forms. Recovery of biogenic P spikes was good, with an increase in concentration of P recovered in spiked samples equivalent to 90% of the myo -IP6 and 95% of the DNA + polyphosphate spike (Table 6-2). Exchangeable P determined by anion exch ange membrane showed no significant distinction between aerobic and anaerobic conditions (ANOVA; p > 0.05) within any spike treatment, yet there were significant differences between time steps in most treatments (ANOVA; p < 0.001) (Figure 6-3). Although this may represent systematic error during iterations of the analytical method, the re covery (100 10%) of standard orthophosphate solutions by the membranes in parallel to soil extractions suggests that there was a true difference in exchangeable P between time steps. Of particular note is the recovery of large quantities of P from soils spiked with polyphosphate (spike 2). Subsequent determination of P composition in alkaline extracts (Figure 6-4) showed limited evidence of polyphosphat es. Therefore, polyphosphates were either rapidly hydrolyzed to orthophosphate or extracted directly by the anion exchange membrane strips. Subsequent investigation has shown an ability of the membranes used to recover certain biogenic P forms, including polyphosphates (Cheesman et al. 2010b). Taken in conjunction with the declining trend in recove red P (Figure 6-3), it would appear that polyphosphate, although initially recovered by the deployed anion exchange membrane, is increasingly stabilized withi n the soil. Study of P composition in subsequent alkaline 170

PAGE 171

extractions (Table 6-3) showed a concom itant increase in both orthophosphate and residual P. Although not conclusive, this suggests both biotic uptake and potential abiotic stabilization of polyphosphate. Initial Biogenic Phosphorus Composition There was a striking lack of distinction bet ween biogenic P identified within aerobic and anaerobic control microcosm soils ex tracted on day 1. Averaging between conditions, it is clear that initial soil cont ained large proportions of biogenic P, including phosphonates (1.8 0.2% total P), phos phomonoesters (30.2 0.2% total P), phosphodiesters (11.5 0.9% of total P) and trace concentrations of polyphosphates (Table 6-3, Figure 6-4). Deconvolution of the region 8 to 3 ppm allowed the clear identification of two isomers of IP6, myo and scyllo, within the initial soil (Figure 6-5). The identification of myo-IP6 by a priori peak assignment has prev iously been criticized, with certain researchers calling for compound spiking to confirm suspected presence (Smernik and Dougherty 2007). Comparison of control soil extracts with those that received spike mixture 1 (Figur e 6-6) confirms the assignment, clearly highlighting the presence of the indicative 1:2:2:1 si gnature and chemical shift of the C2 bound phosphate (Turner et al. 2003f). The assignment of the upfield peak (4.272 ppm) to scylloIP6 is based on previous work by Turner and Richardson (2004) (Figure 6-6) as well as the informed consi deration of its known presenc e in other systems (Turner 2007). Stability of Phosphorus Functional Groups Comparison of initial control soils and after 48 days showed little distinction between aerobic and anaerobic conditions. It wa s impossible to determine if slight differences seen within the concentrations of minor groups such as DNA, were a 171

PAGE 172

relevant shift in functional forms or an experimental artifact. Although, the lack of substantial changes observed within the cont rols did allow for major changes in the composition of spiked soils to be attributed to tr ue distinctions in the fate of extracellular P forms used. The lack of repeated measurement inherent to destructive sampling of individual mesocosms limited my ability to disce rn the origin of the observed variance, especially when considering ch anges in the recovery of the DNA spike. A redesigned experiment would make use of batch reactors with continuous monitoring and modification of redox levels as used by Christophoridis and Fytian os (2006). Thereby, allowing stable aer obic or anaerobic conditions to be maintained while providing thorough mixing and the ability for repeated analysis of a stabl e reactors soil, reducing potential sample variability. The study pres ented here still provides us an initial incite into the role of redox conditions on the fate of certain P forms, a vital consideration when looking into the patterns observed in P composition of soils in the landscape. Stability of polyphosphates and DNA The recovery of extracellular polyphos phates by the anion exchange membranes used here as an initial extraction step was unf oreseen, yet did offer an indication of timescale associated with the residency of ex tracellular polyphosphate in the soil. Assuming the decrease in PAEM observed (Figure 6-3) was due to irreversible abiotic stabilization or biotic uptake, extracellular polyphosphate was found to have a half life of ~35 days under both aerobic and anaerobic conditi ons. Although, further work using solution 31P NMR spectroscopy without pre-extraction would be required to confirm this. Final concentrations of DNA in soils spik ed with mixture 2 showed a decrease over the course of 48 days (Table 6-3). Taking changes in DNA concentration of control soils as a background trend, changes in spiked soils equat e to 30% of the additional DNA under 172

PAGE 173

aerobic conditions and just 2% under anaerobic conditions. This is in line with the original working hypothesis that phosphodie sters see a reduced decomposition rate under anaerobic conditions, yet the experimen tal design and the potential for sample specific variance (see above) limits my ability to use this as definitive evidence. Further studies would be needed employing solely DNA as a P spike and following a complete time course before estimates could be made of the half life of ex tracellular DNA under differential redox conditions. Stability of myo-Inositol hexakisphosphate Comparison of spectra acquired usi ng both the 600 and 500 MHz Brucker systems (Table 6-4, Figure 6-7) showed that wh ile functional forms with low overall concentrations (i.e. phosphonates, polyphosphates) showed poor repeatability between spectra acquisitions, major biogenic P forms, especially myoIP6 within spiked soils showed consistent (RSD < 5%) calculated c oncentrations. This was taken as strong evidence for the validity of using spectra acquired in both systems for analysis of myo IP6 stability with time (Table 6-5, Figure 6-8). It is apparent that myo -IP6 showed evidence of degradation under both aerobic or anaerobic conditions but that samples showed a high degree of variability. Samples taken on day 30 were excluded from statistical analysis given substantially lower re coveries of total P than at all other time steps (Table 6-5). The loss of myo -IP6 was found to be marginally significant (R2 = 0.713, p < 0.01) under anaerobic conditions, ye t in either aerobic or anaerobic conditions represented a decrease of just 10% of the total myo -IP6 spike used; although it should be noted that if significant, this total IP6 decomposition is equivalent to that in previous studies (Suzumura and Kamatani 1995a) wherein total myo -IP6 used was equivalent to just 23 g P g-1 wet weight. 173

PAGE 174

The lack of complete myo-IP6 degradation or any discer nable distinction in its turnover between aerobic and anaerobic co nditions as otherwise seen in marine sediments (Suzumura and Kamatani 1995a) probably reflects distinct biogeochemical characteristics of the soil and sediment material used. Assu ming similar basic sediment properties to those seen in Suzumura and Ka matani (1995b). Total P concentrations within the marine study were approximatel y the same as used here (~800 g g-1 ) but contained an order of magnit ude less carbon ( ~29 mg g-1). It is also worth noting that native IP6 in the Suzumura and Kamatani (1995ab) study was low with inferred myo, scyllo and cis isomers combined repres enting just 1.9 g P g-1 (~0.2% of total P). In contrast using day 1 aerobic and anaerobic c ontrol samples from this study native myo and scyllo -IP6 were found to represent ~14% of to tal P (Table 6-3). Differences in the carbon content, material origins and biogeoc hemical characteristics of the palustrine soils used here as compared to marine sedi ments used previously may impact both abiotic stabilization and biological degradation of myo -IP6. Abiotic stabilization of myo-Inositol hexakisphosphate. The abiotic sorption and stabilization of IP6 is complex, and governed by various aspects of the environmental matrix (e.g. clay content, amorphous metal oxides, Ca2+ calcite and organic matter content) (Celi and Barberis 20 07). The redox-sensitive degradation of IP6 has been linked in certain systems to the reduc tion of Fe(III) oxides (Celi and Barberis 2007; McKelvie 2007), though it has also been suggested that a IP6 could form insoluble complexes with reduced Fe species (DeGro ot and Golterman 1993). In contrast, the interaction and stabilization of organic P with organic matter is largely, with the exception of polyvalent cation complexe s, redox-insensitive (Brannon and Sommers 174

PAGE 175

1985). Therefore, the mechanisms dictating the stabilization of IP6 in mineral sediments may not be equivalent to thos e in highly organic soils. Biotic degradation of myoInositol hexakisphosphate. Four distinct classes of enzyme have been identified with the ability to hydrolyze IP6 (Mullaney and Ullah 2007). Their presence in soils and sediments (Q uiquampoix and Mousain 2005) is linked to their exudation by both plants (Lung et al. 2008) and, more importantly, microorganisms (Hill and Richardson 2007). Al though the ability to utilize IP6 has been noted in a broad range of bacteria, fungi and yeasts (Pandey et al. 2001), it represents a substantial specialization and is likely to show organism specificity. Richardson and Hadobas (1997) noted only 0.5% of soil bacteria isolates were able to utilize IP6 as a sole C and P source, with the percentage of bacteria able to utilize IP6 increasing in the presence of additional labile C. In addition, anaerobi c bacteria isolated from rumen have been shown to utilize IP6 (Yanke et al. 1998) and have been a significant research focus, given the potential for dietar y manipulation of livestock (Lei and Porres 2007) as a means of combating environm ental P loading (Leytem and M aguire 2007). Therefore, it is clear that bacteria from a range of environments are able to hydrolyze IP6 although few, if any, studies to date have attempt ed to identify organisms capable of degrading IP6 within wetlands. Fungal microorganisms, an important component of many highly organic palustrine wetlands (Ipsilantis and Sylv ia 2007; Kominkova et al. 2000; Kuehn et al. 2000) may prove to be a sign ificant component of biological IP6 degradation, with screening for activity in microorganisms rout inely finding fungal organisms to be the most competent in utilizing IP6 (Lissitskaya et al. 1999; Volfov a et al. 1994). It is likely that the dynamic aerobic/anaerobic interface of wetlands and the role fungi may play in 175

PAGE 176

176 certain systems results in differences in the competencies of native microbial populations to degrade IP6. Therefore, differences in microbial populations between study sites could in itself impact the presence or turnover rate of myo-IP6. Conclusions It appears that biogenic P forms are relative ly stable in organic freshwater wetland soils. While the use of anion exchange memb ranes as a pre-extraction confounded the ability to track the fate of polyphosphates both DNA and myo -IP6 underwent degradation over the course of 48 days. Extracellular DNA conformed to the initial hypothesis by showing only limited degradation under anaerobic conditions as compared to a substantial (30%) loss under aerobic conditions. While the tentative conclusions drawn as to the difference in stabilization of DNA under altered redox conditions would require further analysis, this study provides a tantalizing incite into a mechanism that could explain the predominance of phosphodiesters in wetland soils (Turner and Newman 2005). In contrast I did not observe a rapid, or differential, turnover of myo-IP6 under aerobic / anaerobic conditions as seen in Suzumura and Kamatani (1995a), suggesting only limited r edox sensitivity this supports recent research that has found significant levels of IP6 under, presumed anaerobic conditions (McDowell 2009; Turner and Weckstrm 2009) and CHAPTER 4. Allowing me to conclude it is anaerobosis in concert with site mineralogy and organic matter content, which appears to determine the stability of extracellular myo-IP6 in wetland soils.

PAGE 177

Table 6-1. Characterization of surface (0-10 cm) soil collected for spike incubation microcosm study. Wetland name Blue lake, Ordway Swisher Biological Reserve Wetland type Sumpland, Emergent herbaceous vegetation Sampling Location 292.83 N 81594.80 W Possible impacts Burning activity in surrounding uplands Basic Characterization Moisture content (%) 38.2 pH 4.5 Bulk density g cm-30.15 Organic matter (%) 84.0 Elemental concentrations g g-1 Phosphorus 609 Calcium 907 Magnesium 94 mg g-1 Carbon 314.6 Nitrogen 23.1 Iron 2.7 Aluminum 15.5 Estimated by loss on ignition 550C for 4 h Table 6-2. Total phosphorus, after addition of biogenic P spikes, and recovery by AEMNMR method of all microcosms. Total P g g-1 AEM-NMR recovery g g-1 % Soil (Control) 610 10.8 347 62 57 10 Soil + Spike 1 ( myo -IP6) 810 6.8 530 59 65 7 Soil + Spike 2 (DNA + polyphosphate) 768 11.4 498 79 65 10 = total P determined by TP-Ash on 4 microcosms of each type = Recovery by AEM-NMR method across time and aerobic/anaerobic treatment. (n=20) 177

PAGE 178

Table 6-3. Phosphorus composit ion of microcosms as determined by AEM-NMR method. Results (g g-1) 8 to 3 ppm region Day Condition Sp. PAEM NMR Total Phos-P Total Ortho-P myoIP6 scylloIP6 Other LipidP DNA Poly-P Residual g g-1 1 Anaerobic 12 (0.1) 353 8 272 87 64 27 94 13 53 7 245 1 14 (1.9) 522 15 424 73 259 23 68 20 57 6 232 2 98 (0.1) 437 17 234 81 68 24 60 12 174 0 276 Aerobic 15 (0.0) 347 10 263 80 61 19 103 18 56 0 248 1 15 (0.4) 525 13 441 78 281 32 50 17 54 0 228 2 97 (5.4) 465 10 259 95 81 25 58 19 171 6 248 48 Anaerobic 6 (3.1) 344 9 258 84 66 21 87 16 52 9 260 1 6 (0.4) 494 19 406 83 256 19 49 12 48 8 268 2 40 (1.0) 424 7 266 125 65 23 53 7 137 7 346 Aerobic 5 (0.8) 308 19 224 73 59 26 67 12 43 10 297 1 8 (0.9) 494 17 402 69 218 20 95 17 51 7 267 2 43 (2.6) 475 14 283 131 63 27 62 11 156 11 292 Biogenic spike added to microcosm, 1 = myo -Inositol hexakisphospate, 2= DNA + polyphosphate Anion exchange membrane recovered P Phosphonates, Region 8 to 3 ppm de-convoluted to determine orthophosphate myoand scylloIP6 and all other posphomonoesters Phospholipids Polyphosphates Residual P = total P (PAEM + PNaOH) 178

PAGE 179

Table 6-4. Phosphorus composition of soil samples as dete rmined by parallel analysis of lyophilized soil extracts using two distinct nuclear magnetic resonance machines. 8 to 3 ppm region Sample Machine Phos-P Total Ortho-P myo -IP6scyllo-IP6Othe r Lipid-P DN A Poly-P g g-1 Day 1, Aerobic control soils 600 7 268 88 67 26 87 13 56 9 500 16 258 87 81 27 63 11 58 9RSD (%) 5231143221332 415321129315 Day 48, Anaerobic soils spiked with myo -Inositol hexakisphosphate 600 13 411 86 247 23 53 14 50 6500 12 403 80 237 23 63 21 52 5RSD (%) Day 48, Aerobic soils spiked with myo -Inositol hexakisphosphate 600 18 395 75 230 25 65 17 53 11500 11 404 81 251 24 48 15 57 6RSD (%) 35 2 5 6 3 21 6 6 40 NMR spectra acquired using; (600) 600 MHz magnet and 5 mm BBO probe; (500) 500 MHz magnet and 10 mm BBO probe. Phosphonates Region 8 to 3 ppm, split into; Total, (Ortho-P) orthophosphate, myo-IP6, scyllo -IP6 and (other) other phosphomonoesters. Phospholipids Phosphonates 179

PAGE 180

Table 6-5. Phosphorus composit ion of microcosms spiked with myo -Inositol hexakisphosphate with time, as determined by AEM-NMR method. 8 to 3 ppm region Condition Day PAEM NMR Total PhosP Total Ortho-P myoIP6scylloIP6 Other LipidP DNA PolyP Residual g g-1 Anaerobic 1 13.8 (1.9) 522 15 424 73 259 23 68 20 57 6 232 8 13.0 (2.0) 525 6 449 78 261 21 89 18 51 0 230 16 15.2 (0.4) 524 22 426 74 263 25 63 16 52 9 229 30 13.2 (0.4) 444 10 372 69 235 22 47 11 51 0 311 48 5.8 (0.4) 494 12 404 80 237 24 63 21 52 5 268 Aerobic 1 14.5 (0.4) 525 13 441 78 281 32 50 17 54 0 228 8 11.8 (0.0) 506 13 419 73 261 23 62 19 55 0 250 16 13.8 (0.8) 527 13 431 77 273 28 53 20 63 0 227 30 14.1 (1.9) 476 14 392 69 210 18 95 17 48 4 278 48 7.7 (0.9) 494 11 404 81 251 24 48 15 57 6 266 Day microcosm sampled Anion exchange membrane recovered P average (n=2) (1 standard deviation) Phosphonates Region 8 to 3 ppm de-convoluted to determine orthophosphate myo and scyllo IP6 and all other posphomonoesters Phospholipids Polyphosphates redidual P = total P (PAEM + PNaOH) 180

PAGE 181

Figure 6-1. Experimental setup for investigation of biogenic phosphorus stability under aerobic and anaerobic conditions. Spike mixture 1 is myo -Inositol hexakisphosphate, spike mixt ure 2 contains DNA and polyphosphate. 181

PAGE 182

Figure 6-2. Biogenic phos phorus compounds used in spik ing experiment. Standards were mixed with orthophosphate and MDP ( = 17.46 ppm) to provide dual integration standards. Matrix r epresents 0.9 mL DDI 0.1 mL D2O. 182

PAGE 183

Figure 6-3. Exchangeabl e phosphorus, determined by anion exchange membranes during microcosm study. Superscript letters indicate homogeneous subsets ( Tukey HSD ANOVA p < 0.01) 183

PAGE 184

Figure 6-4. Example solution 31P NMR spectra of soil samples spiked with biogenic phosphorus. Samples represent soils extracted on day 1 and previously equilibr ated for 2 weeks under anaerobic conditions. 184

PAGE 185

Figure 6-5. Spectral deconvolution and peak assi gnments in 8 to 3 ppm region of solution 31P NMR spectra. Exemplar spectra of reconstituted anaerobic cont rol soil extracted at t = 1 day. Peak assignments represent average and standard deviation for all t = 1 and t = 48 day samples analyzed using 5 mm BBO probe and Avance II Brucker 600 MHz system. Model conformations and spectra assi gnments based upon Turner and Richardson (2004), P = phosphate functional group. 185

PAGE 186

Figure 6-6. Detail of 8 to 3 ppm regi on of NMR spectra gathered on soil spiked with biogenic phosphorus. Spectra represent soils extracted on day 1 of study and previously equilibrated for 2 weeks under anaerobic conditions. 186

PAGE 187

Figure 6-7. Solution 31P NMR spectra, including the region 8 to 3 ppm in detail, from alkaline soil extracts. Soil extract represents control anaerobic microcosms, day 1, part of incubation st udy. Spectra processed using 8 Hz line broadening and referenced and scaled using MDP ( = 17.46) shown here of fset for clarity. 187

PAGE 188

myo-IP6 (g P g-1) Aerobic conditions Anaerobic conditions Aerobic (R2 = 0.334; p = 0.308) Anaerobic (R2 = 0.713; p = 0.070) Time (days) Figure 6-8. Concentrations of myoInositol hexakisphoshate as determined within microcosm soils under aerobic and anaerobic conditions for up to 48 days. 188

PAGE 189

CHAPTER 7 PHOSPHORUS TRANSFORMATIONS DUR ING DECOMPOSITION OF WETLAND MACROPHYTES1 Introduction Decomposition of detritus influences t he biocycling, retention, and downstream release of nutrients in wetland systems. The often cited model of wetland macrophyte decomposition set out by Webster and Benfie ld (1986) identifies three distinct yet overlapping phases of decom position: an initial rapid leaching of water-soluble components, microbial colonization and decom position, followed by the mechanical and invertebrate mediated fragment ation of material. Of these, the second stage microbial colonization and decomposition represents the most dynamic alteration of nutrient forms. Whether microbes mineralize or sequester inorganic nutrients during decomposition of senesced biomass has implic ations for nutrient dynamics in wetlands (Reddy et al. 2005) and for nutrient sequestrat ion, an important f unction in treatment wetlands (Alvarez and Becares 2006). Phenological characteristics of the litte r appear to influence initial decomposition processes, after which decomposition rates are increasingly governed by gross nutrient ratios (Enriquez et al. 1993) and the nutri ent status of the environment (DeBusk and Reddy 2005; Rejmankova and Sirova 2007). An thropogenic perturbation of nutrient availability in aquatic systems can cause shifts in trophic status (Khan and Ansari 2005), changes in the composition of plant communiti es (Hagerthey et al. 2008; Vaithiyanathan and Richardson 1999), and alteration of mi crobial eco-physiological processes (Corstanje and Reddy 2006; Co rstanje et al. 2007; Wr ight and Reddy 2001a, 2008). 1 Submitted in a modified format 2010 189

PAGE 190

Observed alterations in catabolic processes appear to follow predicted changes in resource reallocation (Allison and Vitousek 2005; Sinsabaugh and Moorhead 1994) with an increase in bioavailable P leading to a reduc tion in investment in P acquisition, such as a reduced release of extracellular phos phatase enzymes by microbes (Newman et al. 2003; Penton and Newman 2008; Wright and Reddy 2001b). Although numerous studies have investi gated factors that influence changes in tissue total P during wetland macrophyte decom position (Brinson 1977; Corstanje et al. 2006; Davis 1991; Fennessy et al. 2008; Grac e et al. 2008; Qualls and Richardson 2000; Rejmankova and Houdkova 2006), transfo rmations of chemical forms during decomposition are not well understood. For terrestrial systems attempts have been made to track temporal changes in P functional groups from select plant materials, with efforts demonstrating both the accumulation of microbial phosphodiesters (Miltner et al. 1998) and the fungal synthesis of polyphospha tes (Koukol et al. 2006). Other studies have sought to partition biogenic P in soils between various microbial and plant sources (Makarov et al. 2002a; Makarov et al. 2005), and transposed position wit hin a soil profile for time, attempting to track general transformations wit hin forest soils during organic matter decomposition (Gressel et al. 1996). In wetlands, previous attempts have been made to characterize leachate from ma crophyte leaves (Pant and Reddy 2001), but changes in P forms in the autochthonously derived organic matter of wetland systems remain poorly understood. The objectives of this study were to determine how litter quality and site characteristics impact changes in the P cont ent of decomposing macrophyte leaves. In addition to tracking changes in total P, solution 31P nuclear magnetic resonance (NMR) 190

PAGE 191

spectroscopy was used to identify changes in the forms of P during decomposition. It was hypothesized that in a P-enriched setting, the accumulation of microbial biogenic P and alteration of macrophyte P would result in net P sequestration, whereas in an oligotrophic setting there would be close coupling of biogenic P production and its subsequent hydrolysis, limiting the accumulation of microbial derived P forms. Materials and Methods Site Description Water Conservation Area 2A (WCA-2A) is a diked and hydraulically controlled 424 km2 portion of the northern Ev erglades, characterized as a freshwater peat system underlain by limestone bedrock. Historically, pr oductivity in the northern Everglades has been limited by P availability, but as a result of anthr opogenic loading from upstream agricultural practices WCA-2A has develop ed a distinct nutritional and concomitant vegetation gradient (Jensen et al. 1995; Koch and Reddy 1992; Vaithiyanathan and Richardson 1997). There is a distinct trans ition from native Everglades marsh dominated by sawgrass ( Cladium jamaicense Crantz), to dominance by cattail ( Typha domingenis Pers.) in areas impacted by nutrient-rich inflow water (Vaithiyanathan and Richardson 1999). The nutrient-enriched areas have increased rates of heterotrophic decomposition (Davis 1991; DeBusk and Reddy 2005; Wright and Reddy 2001a) and a reduced extracellular phosphatase activity (Corstanje et al. 2007; Wright and Reddy 2001b). Two sites (Table 7-1, Figure 7-1) along a P gradient were selected for this study: one enriched site 1.3 km from the northern inflow structur e S10-C, and a second from within an area considered unenriched by P lo ading. Site detritus and surface soil (diameter 15 cm x 1.5 cm) were sampled (n=4) from the study locations prior to 191

PAGE 192

experimental setup and analyzed for total C, N, P as well as P composition by solution 31P NMR spectroscopy. Study Design Phosphorus dynamics during leaf litter decomposition were assessed using a standard litterbag approach. Samples repres ented archived samples from a study implemented and carried out by Dr. Patrick I nglett (2005), in which a full factorial experiment was established at both enriched and unenriched locations using leaf litter of Typha and Cladium from both enriched and unenriched regions of WCA-2A. Samples were retrieved over 15 months to track ch anges in litter quantity and P content. Select litter samples were further analyzed by solution 31P NMR spectroscopy to determine changes in the forms of P during decomposition. Litterbag Study Senescent leaf material of both Typha and Cladium were collected in bulk from locations proximate to both the enriched and unenriched sites. Material included only standing dead intact lamina of unknown age, but presumed < 2 y. Bulk material was rinsed of adhering particulates, cut in to 10 cm sections and dried at 60 C to a constant weight. Litter was analyzed for tota l C, N, P and P forms by solution 31P NMR spectroscopy (see: Analysis of biogeochemical properties) as well as litter quality using a modified proximate forage analysis (Goer ing and Soest 1970) utilizing the semiautomated Ankom A200 (Ankon Tec hnology, Fairport, NY). Litterbags (15 cm x 15 cm) were construct ed from grey polyeth ylene mesh (1 mm) allowing entry of mi cro fauna, in addition to local microbial assemblages. Although recognized to provide an altered bioge ochemical environment for decomposition (Bradford et al. 2002), the use of litterbags provides a useful method in the study of 192

PAGE 193

time-dependent detritus processing. Approxim ately 15 g dry weight of homogenized litter was sealed into individual litter bags. Replicates of each litter type were then contained within a larger bag (2.5 cm mesh ) to aid recovery. Replicate bags were placed (January 2003) on the surface of Typha (enriched site) or Cladium (unenriched site) stands and secured with a polyethylene stake. Three replicate bags were recovered from each site after 16, 33, 75, 204, and 454 days. Litterbags were removed from the larger mesh bag and returned on ice to the laboratory, where they were gently washed with deionized water to remove surface debris, frozen, and lyophilized. Litter was then removed from the bags for mass loss determination and ground to pass a 2 mm mesh using a Wiley mini mill (Thomas Scientific, Swedesboro, NJ). Samples of coarsely-ground material were then analyzed for total P and P composition, and after further grinding, total C and N. All samples were stored in sealed containers under dark ambient lab conditions until analysis. Analysis of Biogeochemical Properties Initial site detritus, surface soils and re covered litter were analyzed for total N and C simultaneously using a Cost ech Model 4010 Elemental Anal yzer (Costech Analytical Industries, Inc., Valencia, CA). Loss on igni tion (LOI) and total P were determined using a modified ashing method (See APPENDIX C). Samples were weighed (~200 mg) into borosilicate scintillation vial s and ignited at 550C for 4 h. The remaining ash was dissolved in 1 M H2SO4 and shaken for 24 h. Solutions were diluted and analyzed for P by standard molybdate colorimetry (USEPA 1993) using a discrete autoanalyzer (AQ2+, SEAL Analytical, UK). This method gave re coveries on a range of certified standards comparable the more labor-inten sive method of Andersen (1976). 193

PAGE 194

Phosphorus Composition To identify P forms present, initial site det ritus, surface soil and select leaf litter recovered from the decompositio n study were analyzed by solution 31P NMR spectroscopy. Given practical and financial c onstraints of analyzing multiple samples with 31P NMR spectroscopy, only Cladium and Typha sourced from the enriched site were analyzed. Samples were analyzed from time steps that were considered able to provide insight into mechanistic processes (i.e. initial and final material, and mateiral after maximum total P leachi ng). Additional samples from the enriched study site were then analyzed to provide information on accumulation rates. Phosphorus extraction in NaOHEDTA A standard alkaline extraction (Cade-M enun and Preston 1996; Turner et al. 2007b) was applied to detritus, soil, and litter sa mples. In brief, field replicates were combined on an equal weight basis to give one homogenized sample of 3 g, weighed out into a 250 mL HDPE centrifuge bottle. To this, 90 mL of a solution containing 0.25 M NaOH and 50 mM EDTA was added, to give a 1: 30 solid to solution ratio. Samples were shaken at ambient room temperature fo r 3 h and then centri fuged (Sorvall RC6 centrifuge, SLA 1500 Rotor; Thermo Fisher Scientific, Waltham, MA, USA ) at 6500 rpm for 10 min. A subsample (20 mL) of the super natant was removed to a scintillation vial and combined with 1 mL of 50 mg P L-1 me thylenediphosphonic acid (MDP) as an internal standard (Turner 2008b). Mixed samples were immediately frozen ( C) and lyophilized prior to NMR spectroscopy. A second subsample was analyzed for total P (NaOHTP) by a modified double-acid digest using H2SO4 and HNO3 (Rowland and Haygarth 1997) and a modified discrete moly bdate colorimetric method (see above). Residual P not recovered by the alkaline extraction is by definition unidentified and its 194

PAGE 195

chemical stability presumably indicates its recalcitrance in the environment. For mineral soils it has been assigned as recalcitrant organic (Cade-Menun 2005b) or alkali-stable inorganic P (Turner et al. 2007a), although there is little in formation for wetland soils. Solution 31P nuclear magnetic res onance spectroscopy Spectra were acquired using a Bruker Avance 500 Console with a Magnex 11.75 T/54 mm magnet using a 10 mm BBO Probe. Lyophilized samples (~300 mg) were resuspended in 0.3 mL D2O and 2.7 mL of a solution containing 1 M NaOH and 0.1 M EDTA, vortexed, and then transferred to a 10 mm tube. Spectra acquisition was carried out at a stabilized 25 C with a calibrated (~30) pulse length, a zgig pulse program, and a 2 s T1 delay. Results presented here are of ~30,000 scans accumulated as three sequential experiments, with FIDs summed post acquisition by Bruker proprietary software. Spectra interpretation was carried out using Acorn NUTS (NMR Utility Transform Software) (Acorn NMR Inc, Livermore, CA). After applying 15 Hz line broadening, spectra were referenced and integrated agains t the internal standard, MDPA, set as = 17.46 ppm (compared to externally held 85% H3PO4 0 ppm). Integration over set spectral windows were chosen to correspond with known phosphorus bonding environments (Turner et al. 2003d). The r egion between 8 and 3 ppm was further elucidated after 3 Hz line broadening by a dec onvolution subroutine applied to identify and quantify orthophosphate (6.21 0.02 ppm) among orthophosphate monoesters. Data Analysis All statistical tests were performed in SPSS for windows version 17.0.0 statistical software (SPSS Inc. 2008). Mass remaini ng, P concentration and mass of P were analyzed by a 4 way univariate ANOVA using si te of decomposition (site), species of 195

PAGE 196

litter (species), source of litter (source) and time as independent variables. Given the homogeneous nature of initial material, time = 0 was excluded from the statistical analysis. Resulting model residuals were analyz ed via P-P plot and visual inspection for normality. If required, original data were normalized us ing a natural log transformation. Significant results are reported alongside their calculated partial eta2 values. Given complications of interpreting multiple intera ction factors, non-sign ificant higher order terms are not presented here. Simple decomposition rates were calculated using an exponential decay model (Webster and Benfie ld 1986). Linear regression of changes in the mass of leaf litter P over time excluded data from the initia l equilibration and leaching period. Due to practical constraints, solution 31P NMR data were acquired for pooled samples and as such are without a measure of variance at indi vidual time steps. All integrated results using MDP as an internal standard were within 20% of determined NaOHTP concentrations. Calculated rates of accumulation of specific P forms are based upon simple li near regression after initial leaching and equilibration period. Results Initial Litter Material Typha litter material from both study sites c ontained greater concentrations of total N and P, as well as a higher LOI (indicating lower biosilica and mineral content) than Cladium (Table 7-2). Although Typha litter had a noticeably larger proportion of neutral detergent-extractable C, the proportion of lignin and ce llulose was similar, with LCI (Lignin to Cellulose Index) between 0.170 and 0.197 for all litters sampled. Within species, there was a distinct difference bet ween material collected at the enriched and unenriched site, with Cladium showing an approximat e four-fold increase in total P and 196

PAGE 197

Typha showing a two-fold increase, respectivl y. Using both species, collected from enriched and unenriched settings allowed four litter types to be used within the subsequent decomposition study. Mass Loss Between 36 and 70% of original mass of litter remained after 15 months in the field (Figure 7-2). Rates of decomposition varied significantly between litter types, in terms of both macrophyte species and litter source, as well as between study sites (Four way Univariate ANOVA, Table 7-3). In addition, decay rate constants based upon a simple exponential decay (Table 7-4) followed expected patterns, with increased mass loss in response to increased litter quality and exogenous nutrients (Corstanje et al. 2006; Rejmankova and Houdkova 2006). The higher quality Typha litter material decomposed more rapidly than the Cladium litter (ANOVA ; p < 0.001), with enriched material of both species showing signifi cantly higher rates of mass loss (ANOVA; p < 0.001). All litter materials s howed significantly (ANOVA; p < 0.001) higher mass loss at the enriched site. However, decomposition rate constants were relatively constrained, varying from 0.00275 day-1 for enriched Typha material at the enriched site to 0.00136 day-1 for unenriched Cladium at the unenriched site. Signi ficant interaction terms, between time and both litter species and source demonstrated the continued and differential influence of initial material char acteristics on the long-term stability of organic matter. Phosphorus Content Total P concentrations (Figure 7-3) changed significantly with time. At the enriched site all litter types except unenriched Cladium showed an initial dr op in P, as from a rapid leaching event, but then subsequently increased to between 5 and 10 times the 197

PAGE 198

original P concentration. In contrast, at the unenriched site after an initial decrease in P, concentrations did rise, but with the exception of unenriched Cladium never recovered to their initial levels (Figure 7-3 B). Un ivariate four-way ANOVA (Table 7-5) showed significant differences not only dependent upon the site of decomposition ( p < 0.001), but on the nature of the initial litter, as shown by highly significant effect of both litter species ( p < 0.001) and source ( p < 0.001). Significant intera ction terms (litter quality with time) suggested different responses in P concentration of litter types over time. This relationship was explor ed by applying a two-factor univariate ANOVA using litter type (enriched Cladium unenriched Cladium enriched Typha, and unenriched Typha) and site at each time step. Post hoc analysis (Tukey HSD) showed that at 16 days there were significant differences in the total P concentrations between litters of the same species sourced from different locations. In contrast, at day 454 there were no significant differences in the total P concentrations in Cladium yet still a significant distinction for Typha ( p < 0.05) and between Typha sourced from enriched and unenriched locations ( p < 0.001). Changes in P concentration during decomposition may be a result of two distinct processes: a change in litter mass that results in alteration of the endogenous P concentration, or the loss or gain of P from the environment. The distinction within this study was explored by plotting changes in the mass of P over time (Figure 7-4). Given an inability to interpolate data and the potential for sites to experience variation in initial flooding I draw a conservative breakpoint at 33 days, marking the switch from initial abiotic processes and site equilibration to l ong-term microbial acti on. From 33 days to the end of the study there was a distincti on in changes in mass of P between sites. 198

PAGE 199

Linear regression (Table 7-6) demonstrated that the mass of P in all litter at the enriched site increased significantly over time ( p < 0.001). At the unenriched site, the mass of P in both enriched Cladium and unenriched Typha did not change significantly with time. The P content of unenriched Cladium increased slightly, yet significantly (B= 0.037 g P g-1 initial material day-1; p = 0.001) during decompos ition, whereas in enriched Typha there was actually a significant ( p < 0.001) decrease in the mass of P. Phosphorus Composition Analysis of initial leaf litter demonstra ted a range of P forms present within both Cladium and Typha (Figure 7-5). In both cases, NaOHEDTA extractable P was dominated by orthophosphate (35 and 39% of total litter P in Cladium and Typha, respectively). In addition, considerabl e amounts of phosphomonoe sters (13 and 23% total litter P), phosphodiesters (4 and 9% tota l litter P), and inorga nic pyrophosphate (2 and 6% total litter P) were identified. It should be noted that t here was probably an inherent bias in the analysis, because so me phosphodiesters (i.e. RNA and some phospholipids) decompose to phosphomonoesters in alkaline solution during extraction and analysis (Turner et al. 2003d). Comparison of initial P composition at the two study sites (Figure 7-6) demonstrated distinct differences in the fo rms and proportions of P identified. Standing detritus and surface soil from both sites contained orthophosphate, phosphomonoesters, phosphodiesters (dom inated by DNA) and pyrophosphate. However, only detritus from the enr iched site contained long chain inorganic polyphosphate (~3% of total P). Concentrations of all P groups identified were higher in detritus than in soil at both sites, and high er in enriched than in unenriched sites. There was a distinct shift in the ratio of phos phomonoesters to phosphodiesters in material 199

PAGE 200

from the enriched and unenriched site s, from 1.35 to 0.58 in detritus and 1.69 to 0.90 in surface soil, indicating a gr eater proportion of P identifiable as phosphodiesters within detritus and at unenriched sites. Signal s within the phosphomonoester regions indicative of the phosphomonoester myo -Inositol hexakisphosphat e were not detected in any sample. This is consistent with other studies of calcareous freshwater wetlands (Turner and Newman 2005; Turner et al. 2006a), although inositol phosphates could have been present at concentrations below the limit of detection by solution 31P NMR spectroscopy (El-Rifai et al. 2008). There was a convergence in the proportion of P forms present based upon the site of litter decomposition (Figure 7-7; Table 77). Identifiable P lost during equilibration (presumably by leaching) at both the enr iched and unenriched site was dominated by orthophosphate (up to 50% of lost P), yet concentrations of phosphodiesters and pyrophosphate also declined after 16 or 33 days. Concentrations of phosphomonoesters increased or remained relatively stable in all li tters between initial material and samples at days 16 and 33, result ing in an increase in their proportional contribution to the to tal P (Table 7-7). Recovery of total litter P by NaOHEDT A extraction averaged 56% for all samples. Residual P was a relatively stable proportion of the total P, alt hough the concentration increased in litter at the enriched site. Trace concentrations of phosphonates and midchain polyphosphate were detected in one sample ( Typha from the enriched site after 16 d), but were suspected to originate fr om adhering cyanobacterial or phytoplankton biomass (Eixler et al. 2006). After the initia l equilibration, litter at the enriched site accumulated all forms of P identified by 31P NMR analysis, in direct contrast to the 200

PAGE 201

unenriched site, where there was little change in P composition after the initial leaching period (Figure 7-8). The increase in concen tration of P forms at the enriched site showed a significant linear response (Table 78), with most major forms showing a net change of between 89 and 107 mg P kg-1y-1. Changes in pyrophosphate concentration appeared to be distinct from other forms, showing a significant ( p < 0.05) linear response at approximately half the rate of change for phosphomonoesters, phosphodiesters or ort hophosphate concentrations. Discussion Plant litter decomposition rates determined in this study over the course of 15 months were generally lower than values reported for other herbaceous litters (Fennessy et al. 2008; Rejmankova and Houd kova 2006), but correspond well with previous data from WCA-2A. For exampl e, Debusk and Reddy (2005) found simple decay rate constants of approximately 0.003 d-1 for Typha decomposing under P-rich conditions. Rates of mass loss demonstrated previously established patterns with both higher litter quality, as determined by stru ctural C and nutrient content and higher environmental nutrient availability, resulti ng in an increased rate of mass loss. Over the course of the study, total P c oncentrations within litter at the two sites showed distinctly different trajectories. At the enriched site, total P concentration in all litters increased, resulting in an average mo lar C:P ratio of ~488 after 15 months. In contrast, concentrations at t he unenriched site generally did not return to the levels of the initial material entering the system, wit h an average molar ratio of ~4150 after 454 days. Although conventional understanding of litter decomposition would assume a continued sequestration of P at both sites in response to gross stoichiometry, analysis of changes in mass of P (Figure 7-4) suggests that while microbial activity at the 201

PAGE 202

enriched site resulted in the net sequestr ation of P from the endogenous environment, at the unenriched site there was little net gain (Figure 7-9). This interpretation assumes that mass loss due to fragmentation is mi nimal and may represent an underestimation of actual P accumulation in both sites. Yet previous studies have also suggested that oligotrophic systems such as the unenriched Everglades ma y present an environmental gradient in which P sequestratio n predicted from detritus C:P ratios is unlikely to occur (Longhi et al. 2008; Qualls and Richar dson 2000). Indeed, Qualls and Richardson (2000) demonstrated that an environment al P concentration of 5 g P L-1 in the water column resulted in the net loss of P to the environment from both Typha and Cladium leaf litter during decomposition. Although recovery and identif ication of initial plant P by alkaline extraction was less than previous published values for live plant tissue (Makarov et al. 2005), recoveries were considered good at 54 and 76% of Cladium and Typha, respectively. Initial changes in litter P composition as a result of loss of leaching of labile P correspond to the identification of water ex tractable forms by He et al. (2009), who showed a predominance of orthophosphate, but significant organic P loss from plant material. The apparent stability of phosphomonoes ters during the initial leaching may be due to the recalcitrant nature of phosphomonoesters remaini ng in plant material after nutrient resorption and senescence of leaves, or due to the rapid synthesis of new phosphomonoesters in establishing microbial popu lations. In addition to leaching from the plant biomass, consideration should be gi ven to the fact that some compounds (e.g. pyrophosphate) may be lost from endophytic and aerial s aprotrophic fungal biomass during initial prepar ation and site equilibration (Kominkova et al. 2000; Kuehn et al. 202

PAGE 203

2000). Indeed, the use of standing dead biomass of unknown age leaves us unable to determine alterations of P composition dur ing these initial stages of the decay continuum (Kuehn et al. 1999). Initial abiotic leaching is often consi dered rapid (Davis et al. 2006), yet given an inability to interpolate data and the potential fo r sites to experience variation in initial equilibration I drew a conservative breakpoint at 33 days marking the switch from initial abiotic processes and site equilibration to longterm microbial action After 33 days this microbial action at the enriched site result ed in the accumulation of all forms of P identified by solution 31P NMR spectroscopy. The estimated rates of change in P concentrations (Table 7-8) include contributi ons from both stabilized original and newly accumulated compounds. For example, the increase in recalcitrant P concentration (up to 8-fold) is not accounted for by the passive accumulation of initial endogenous recalcitrant P and suggests a mechanism of transformation by which accumulating biogenic P is ultimately modified to an operationally recalcit rant organic form. Differences in both C quality and total P of initial material appeared to be reflected in P composition of litter material collected at the enriched site, with Typha consisting of more labile forms. However, changes in composition during decomposition appeared to be independent of initial material, ra ther changes were dependent upon site characteristics, with major P forms showi ng similar changes in concentration at the enriched site. Fungal biomass is a major cont ributor to heterotrophic decomposition in emergent herbaceous systems (Hackney et al. 2000; Kuehn et al. 2000) and although litter quality can affect the composition of the microfungal community, their generalist nature may result in similar responses to the exogenous nutrient av ailability (Thormann 203

PAGE 204

et al. 2003). Within the enriched study site, the P co mposition of decomposing litter appears to be on a trajectory towards that determined in standing det ritus and surface soil. In contrast, litter material wit hin the unenriched site after 454 days of decomposition did not appear to reflect standing detritus. This could reflect differences in the role of macrophyte litter across the northern Evergl ades. Within nutrient-impacted portions of WCA 2A, surface substrate originates from Typha detritus and a high standing population of heterotrophic mi crobes, while oligotrophic regions often include only minimal Cladium fragments, but a significant contribution from the periphyton community (Wright and Reddy 2008). Furt her work would be needed to determine the role of these additional sources of biogenic P in determination of P forms seen within wetland substrates. Conclusions Decomposition processes are affected by both initial material and site characteristics. The decomposition of macrophyte leaf litter results in the sequestration of P from the envir onment via the accumulation of biogenic forms resulting from microbial turnover. Low ambient P concentrations result in limited or no P sequestration due to heterotrophic decomposition of leaf litter. Substrates in relatively enriched herbaceous systems are presumed to contain a large proportion of modified macrophyte leaf litter and accumulated microbi al biomass. Within oligotrophic systems, macrophyte detritus and associated heter otrophic biomass appears to have less influence on the nature of P in surface substrates. 204

PAGE 205

Table 7-1. Site characteristics for enr iched and unenriched study sites within WCA-2A. Detritus and soil samples average (n=4) one standard error. Overlying water characteristics based upon published values. WCA-2A Location Enriched Site Unenriched Site Latitude (N) 26 21.230 26 16.382 Longitude (W) 80 20.967 80 21.502 Distance from inflow st ructure (km) 1.93 10.05 Average water depth (cm; max, min) 12.8 (11.9, 14.0) 11.9 (11.0, 13.6) Overlying water Total P (g P L 1 ) 52.4 (6.7) 9.6 (0.9) Ortho P (g P L 1 ) 27.8 (4.4) 1.7 (1.3) Detritus Total P (g g1 ) 1334 (452) 206 (53) Soil (01.5 cm) Total P (g g1 ) 1312 (40) 468 (40) Stage data from South Florida Water managem ent (23/01/03 through 04/21/04). Sampling stations WCA2E1 and WCA2U1. Site water characteristics, South Florida Water management (6/21/94 through 9/27/94). Sampling staions WCA2E1 and WCA2E5. Available through DBHYDRO (http://my.sfwmd.gov/dbhydroplsql/) Table 7-2. Characterization of litter mate rial used within the decomposition study, consisting of two species ( Cladium and Typha ) from both unenriched and enriched portions of WCA-2A. Cladium Typha Un En Un En LOI (%) 91.8 93.7 95.6 95.6 Total C (%) 41.4 40.9 43.1 43.2 Total N (%) 0.41 0.38 0.45 0.54 Total P (g g-1) 46.1 171.0 135.4 261.2 Molar C:P 23200 6179 8223 4273 Forage Analysis (Ankom A200) Neut. Det Extractable (%) 25 22 30 30 Hemi Cellulose (%) 29 33 23 25 Cellulose (%) 35 36 37 37 Lignin (%) 8.6 8.1 9.1 7.6 Un = Unenriched En = Enriched 205

PAGE 206

Table 7-3. Four way Univariate ANOVA for mass remaining. Full factorial model adjusted R2 = 0.973). Source and species of litte r (litter quality) as well as site of decomposition and time in the field all shown to be highly significant (p < 0.001). With litter quality parameters and si te showing significant interactions with time. Source of variation DF F p -value Partial eta squared Site (S) 1 57.9 < 0.001 .420 Time (T) 4 931.8 < 0.001 .979 Source (So) 1 74.3 < 0.001 .482 Species (Sp) 1 129.0 < 0.001 .617 T x S 4 34.2 < 0.001 .631 T x So 4 5.4 0.01 .213 T x Sp. 4 23.3 < 0.001 .538 So x Sp 1 7.1 0.009 .082 Table 7-4. Simple exponential decay rate constant (x =100e-kt ) and leaf litter half life calculated from material recovered over the course of 15 months within WCA2A (n=15). Site Litter type k R 2 T1/2 (days) Enriched Cladium Enriched 0.965 0.00172 403 Unenriched 0.924 0.00148 468 Typha Enriched 0.943 0.00275 253 Unenriched 0.918 0.00202 343 Unenriched Cladium Enriched 0.896 0.00118 587 Unenriched 0.904 0.00093 745 Typha Enriched 0.960 0.00207 335 Unenriched 0.954 0.00136 510 206

PAGE 207

207 Table 7-5. Four way Univariate ANOVA of phosphorus concentration in leaf litter. Model adjusted R2 = 0.947. Source and species of litter (litter quality) as well as site of decomposition and time in the field all shown to be highly significant. Source of variation DF F p value Partial eta squared Site (S) 1 2169.4 < 0.001 .966 Time (T) 4 302.4 < 0.001 .940 Source (So) 1 438.8 < 0.001 .851 Species (Sp) 1 759.2 < 0.001 .908 S x T 4 130.4 < 0.001 .871 S x So 1 77.0 < 0.001 .500 S x Sp 1 86.0 < 0.001 .527 T x So 4 11.0 < 0.001 .363 T x Sp 4 5.6 < 0.001 .226 So x Sp 1 22.3 < 0.001 .225 Table 7-6. Linear regression of changes in ma ss of phosphorus within litterbags held in the field for between 33 and 454 days. Site Litter type B (g P g1 day-1) R2 p value Enriched Cladium Enriched 0.637 0.873 0.000 Unenriched 0.685 0.950 0.000 Typha Enriched 0.447 0.795 0.000 Unenriched 0.566 0.770 0.000 Unenriched Cladium Enriched -0.036 0.201 0.143 Unenriched 0.037 0.726 0.001 Typha Enriched -0.175 0.851 0.000 Unenriched -0.014 0.050 0.248

PAGE 208

Table 7-7. Phosphorus forms as determined by solution 31P NMR spectroscopy of NaOHEDTA extracts during macr ophyte leaf litter decomposition Site Species Time (days) Total P ((g P g-1) % Total P Phosphonate Orthophosphate Phosphomonoesters DNA Other Phosphodiesters Pyrophosphate Polyphosphate Residual Enriched Cladium 0 171 34.7 12.8 3.3 1.1 2.5 45.7 16 104 26.9 26.0 1.9 1.9 1.4 41.9 75 187 23.5 16.6 4.1 1.5 5.3 48.9 204 441 18.7 13.9 3.8 3.0 3.9 56.8 454 795 19.5 16.4 10.0 4.8 4.4 44.9 Typha 0 261 38.5 22.6 4.7 4.1 5.6 24.5 16 237 0.3 18.4 32.7 1.6 11.8 1.0 2.4 34.5 75 365 15.1 21.5 5.2 5.1 2.7 50.3 204 742 10.9 12.6 4.3 2.6 2.6 67.1 454 1257 12.3 14.9 8.3 4.0 3.7 56.7Unenriched Cladium 0 171 34.7 12.8 3.3 1.1 2.5 45.7 33 67 33.0 15.2 1.3 9.1 13.6 27.8 454 85 29.4 29.2 6.9 4.9 0.0 29.4 Typha 0 261 38.5 22.6 4.7 4.1 5.6 24.5 33 195 18.9 29.9 1.4 4.6 3.7 41.4 454 239 15.5 17.5 4.4 1.2 2.4 58.9 Table 7-8. Coefficients (one standard deviati on) of linear increases in concentrations of major phosphorus forms identified within leaf litter during decomposition at the enriched study site. Cladium Typha Orthophosphate 0.292**(.02) 0.256*(.018) Phosphomonoesters 0.244*(.20) 0.260*(.054) Phosphodiesters 0.265*(.39) 0.285*(.057) Pyrophosphate 0.072*(.006) 0.099*(.006) Significant at the 0.05 level ** significant at the 0.001 level 208

PAGE 209

Figure 7-1. Location of chapter 7 study sites within Wate r Conservation Area (WCA)-2A in the northern Everglades. Before re cent hydrological diversions through Storm Water Areas, flow was predomi nantly northsouth carrying nutrient impacted water from the Ever glades Agricultural Area. 209

PAGE 210

210 Enriched Site Unenriched Site Mass remaining (%) Time (days) Figure 7-2. Mass remaining of four litter types placed at two distinct sites within WCA2A and recovered at time intervals up to 454 days. Symbols represent averages (n=3) with error bars sho wing one standard error. Significant differences in decomposition rates between sites and between litter types (see Table 7-3) were reflected in the range of decay rates calculated for litters used (Table 7-4).

PAGE 211

A B Site detritus total P Total P (g g-1) Site detritus total P Enriched Cladium Enriched Typha Unenriched Cladium Unenriched Typha Time (days) Figure 7-3. Changes in litter phosphorus concentration over time at, A) enriched site and, B) unenriched site. Symbols represent averages (n=3) with erro r bars showing one standard error. 211

PAGE 212

Enriched Site Unenriched Site Change in P mass (g P g-1 initial material) Time (days) Figure 7-4. Changes in mass of phosphorus in macrophyte leaf litter held within litterbags over the course of 454 days. Data standardized for initial mass of material placed in litterbags. Symbol s represent averages (n=3) with error bars showing one standard error. 212

PAGE 213

Figure 7-5. Initial phosphorus composition of Typha and Cladium leaf litter sourced from th e enriched portion of WCA-2A. 213

PAGE 214

Figure 7-6. Initial phosphorus compositi on of detritus and surface soils from en riched and unenriched study sites sampled on (10/20/03). Spectra plotted using 15 Hz line broadening and scal ed using height of internal standard MDP. 214

PAGE 215

Figure 7-7. Example solution 31P NMR spectra showing changes in phosphorus composition of Typha leaf litter during decomposition at both an unenriched and enri ched site over the course of 454 days Spectra plotted using 15 Hz line broadening and scaled to height of MDP. 215

PAGE 216

Cladium Typha Time ( da y s ) P form (g g-1) Unenriched Site = Orthophosphate = Phosphomonoesters = DNA = Pyrophosphate Enriched Site Figure 7-8. Changes in proportion of ma jor P pools found within macrophyte leaf litter over the course of 454 days of decomposition in WCA-2A. 216

PAGE 217

217 Figure 7-9. Conceptual model of phosphor us turnover in wetland macrophyte det ritus under, A) enriched and, B) unenriched conditions. I= phosphate uptake, II= organic ma tter accreation, III = re-s uspension and flux, IV =senecence, V= enzymatic hydrolysis

PAGE 218

CHAPTER 8 PHOSPHORUS FORMS AND DYNAMI CS ALONG A STRONG NUTRIENT GRADIENT IN A TROPICAL OMBROTROPHIC WETLAND Introduction The tropical wetlands of Central America are recognized as ecosystems of social, cultural and economic importance (CCAD 2 002; Ellison 2004). Yet their inaccessible nature has meant that few detailed investigations have been carried out on their formation, ecology and biogeochemistry (Phillips 1998). Such an understanding is critical in determining the implications of increasing direct and indirect anthropogenic pressures upon them (Limpens et al. 2008; Phillips 1998) and those systems linked through hydrologic or nutrient cycles (D'Croz et al. 2005) Tropical peat domes are self emergent organic wetlands within the humid tropics (Semeniuk and Semeniuk 1997). Although historically a ssociated with the swamps of mariti me south east Asia, significant peat deposits are present though out the Caribbean coastal plain (Ellison 2004; Phillips and Bustin 1996) and the tropical Americas (Lahteenoja et al. 2009b). They represent systems which have an organic layer greater than 50 cm and a organic content >75% (Andriesse 1988) their upper surface shows a pronounced convex morphology leading to their hydrologic isolation from surrounding riverine, or marine, systems and a truly ombrotrophic state (Anderson 1983). Although the water tabl e may fluctuate within the upper peat layer (acrotelm), in unimpacted peat domes it remains close to, or above, the surface for the entire y ear resulting in surface flow and lateral seepage from the central portion. This is presumed to result in significant nutrient redistribution towards the periphery. Current conc eptual models of peatdome development and maintenance suggest a complex adaptive system with multiple feedback processes active at different spatial and temporal scales (Belyea and Ba ird 2006; Belyea and Clymo 2001) resulting 218

PAGE 219

in a predictable progression of concentric vegetation types (or phasic communities) across the dome surface. Alt hough changes in total P content have been noted as part of a general trend in declining nutrient status towards the center of tropical peat domes and between phasic vegetation types (Phillips et al. 1997; Phillips 1998; Troxler 2007). Little is known about the nature and cycli ng of P within thes e tropical systems (Sjgersten et al. 2010). In addition to implicat ions on in situ vegetation dynamics, the nature and bioavailability of P will have direct implications upon nutrients exported to aquatic systems downstream fr om these wetlands. Within established ombrotrophic wetlands it can be assumed that the vast majority of organic matter within surface substrates is autocanthously derived. Phosphorus inputs within this organic matter, consists of a variety of forms dependent upon the nature and structure of the biotic communi ty (Harrison 1987; Koukol et al. 2008; Makarov et al. 2005). Although, given rapi d microbial-mediated processes (Oberson and Joner 2005) it is unclear as to the balance between P forms derived from primary eukaryotic inputs and those as a result of microbial turnover and its standing biomass. The turnover of biogenic P is dependent upon its stability within the organic matrix, in part dependant on the activity of extracellular hydrolytic enzymes. Although commonly applied assays for hydrolytic enzyme activity in soils may not reflect actual flux rates from organic P pools (Wallenstein and Weintr aub 2008) close coupling to the nutrient status of biota in wetlands (Caldwell 2005; Wright and Reddy 2001b) makes them a useful indicator of P limitation and re lative rates of pool turn over. In this study, I aimed to identify changes in P forms and potential pool turnover in the surface soils of a tropica l ombrotrophic peat dome. It wa s hypothesized that altered 219

PAGE 220

vegetation communities and nutrient availabilit y would influence both the nature of P inputs and microbial turnover across a gradient from the periphery to the center of the Changuinola peat dome. Specifically, I believed that there would be a monotonic change in the functional nature of P forms identified by solution 31P nuclear magnetic resonance spectroscopy because of altered environmental conditions. Methods Study Site The Changuinola peat deposit, part of the internationally recognized San san pond sak wetland in Bocas del Toro province NW Panama (Ramsar 2009, Site #611) represents a near pristine exam ple of a raised ombrotrophic peat dome (Cohen et al. 1989, 1990). Palynological work (Phillips and Bustin 1996; Phillips et al. 1997) suggests that Caribbean coastal ombrotrophic systems, such as the Chang uinola deposit, may develop by different mechanisms and contain distinct community types than those identified in the well studied peat domes of maritime south east Asia (Anderson and Muller 1975). Yet there are still strong similarities with a visible soil-vegetation catena across the convex surface, mirroring the site development over time (Phillips et al. 1997). Current vegetation communiti es range from monodominant Raphia taedigera palm swamp, at the periphery, through mixed and monodominant Campnosperma panamensis forest swamp to a central bog plain community dominated by herbaceous species such as sawgrass (Phillips et al. 1997; Sjgersten et al. 2010) Sampling A study transect was established (Sj gersten et al. 2010; Troxler 2007) making use of an original leveling survey transects (Cohen et al. 1989). Site access runs almost perpendicular from a bordering canal towa rds the geodesic center of the peatdome 220

PAGE 221

(Figure 8-1, Table 8-1). Nine sampling sites were established at 300 m intervals along the linear path of the access route. Soil sampling, carried out in Sep 07, consisted of collecting three repeat samples from withi n 20 m of each site. With each repeat being an amalgamated sample of three su rface cores (diameter 7.5 cm, 0 10 cm) collected from within 2 m of each other using a sharpened metal cutting head upon a ridged polycarbonate tube. Samples were immediatel y transferred to pre-labeled ziplock bags and put on ice for transportation back to the lab. Samples were transferred to Panama city, with initial sample processing occurri ng within 72 h of field sampling. Initial processing involved homogenizing and the removal, by hand, of recognizable roots (> approx diameter 1 mm) and ligni fied structures, seeds twigs etc. It should be noted that due to the very high concentration of fine ro ots from herbaceous species at sites eight and nine a significant amount of root biomass might have been retained. Samples were subsequently split with half being air dried (~ 22C 10 days) to constant weight and the remainder being held at 4C in sealed zipl ock bags (subsequently referred to as fresh sample). Air dried samples were ball gr ound using Tungsten carbide vessels, before storing in airtight containers under ambient lab conditions until analysis. Soil Properties Soil moisture was calculated on fresh samp les by gravimetric loss after drying at 105 C for 24 h, pH was measured on fresh samples using a strict 1:20 soil to water ratio and standard glass pH electrode. Total elemental concentrations were determined on dried and ground samples. Total soil C and N by combustion and gas chromatography using a Flash EA1112 (Thermo Sc ientific, Waltham, MA), total P, Ca, K and Mg after a standard H2O2 + H2SO4 digestion procedure (P arkinson and Allen 1975) and ICPOES (Optima 2100, Perkin-E lmer Inc., Shelton, CT). 221

PAGE 222

Phosphorus Characterization Anion exchange membranes Available and microbial P was determined using anion exchange membrane (AEM) extraction (Cheesman et al. 2010b; K ouno et al. 1995; Myers et al. 1999) using one 6.25 x 1.5 cm AEM strip (BDH Prolabo Product number: 551642S, VWR International, UK), preloaded with bicarbonate and eluted in 0.25 M H2SO4. The concentration of eluted phosphate was dete rmined by automated molybdate colorimetry with detection at 880 nm using a flow inje ction analyzer (Lachat Quickchem 8500, Hach Ltd, Loveland, CO) and difference betw een non fumigated and hexanol fumigated samples attributed to fumigation release of microbial P. It should be noted that in addition to inorganic orthophosphate the me mbrane strips used may have recovered labile organic and inorganic poly-phos phoric P (Cheesman et al. 2010b). Solution 31P nuclear magnetic resonance spectroscopy Phosphorus composition in soil sample s was determined by standard alkaline extraction (Cade-Menun and Prest on 1996; Turner et al. 2007b) and 31P NMR spectroscopy. The standard alkaline extraction (0.25 mol L-1 NaOH and 50 mmol L-1 EDTA) was applied to air dried soils with shak ing for both 4 and 16 h. As well as on both fumigated and non-fumigated fresh samples after application of the AEM strips, see above (Cheesman et al. 2010a). All samples were shaken at a 1:30 soil to solution ratio at ambient room temp erature for 4 or 16 h before being centrifuged and the supernatant removed. Due to the prohibitive costs of 31P NMR analysis field replicates were combined on an equal volume basis (15 mL) and combined with 1 mL of an internal standard methylenediphosphonic acid (MDP) (50 mg L-1) mixed, immediately frozen (80 C) and lyophilized to await re-suspension and spectra acquisition. A second 222

PAGE 223

independent subsample was an alyzed for total P (NaOHTP) by a modified double acid digest using concentrated H2SO4 and HNO3 (Rowland and Haygarth 1997) and analysis for molybdate reactive P using au tomated molybdate colorimetry. Spectra were acquired using a Bruker Avance 500 Console with a Magnex 11.75 T/51mm Magnet, using a 10 mm BBO Probe. Lyophilized samples (~300 mg) were resuspended 2.7 mL (1 mol L-1 NaOH 0.1 mol L-1 EDTA) and 0.3 mL D2O before vortexing and transfer to a 10 mm probe. Spectra acquisi tion was carried out at a stabilized 25 C with a calibrated (~30) pulse length a zgig pulse program and a 2 s T1 delay. Results presented here are of ~40,000 scans accumulated as 4 sequential experiments with FIDs summed post acquisition, by Bruker pr oprietary software. In addition, exemplar spectra from alkaline extractions after applic ation of the AEMs are presented. These were determined using a 5-mm NMR probe and a Bruker Avance DRX 500 MHz spectrometer (Bruker, Germany) using a 6 s pulse (45), a delay time of 1.0 s, an acquisition time of 0.4 s, and a zgig pulse program. Spectra interpretation was carried out using wxNUTS vr 1.0.1 for Microsoft Windows (Acorn NMR Inc. 2007). Spectra we re referenced and integrated against the internal standard (MDP) set as = 17.46 ppm after its comparis on to an externally held 85% H3PO4 set as = 0 ppm. Integration over set spectral windows were chosen to correspond with known phosphorus bonding classes (Turner et al. 2003d). With spectral deconvolution applied to the region 8 to 3 ppm to separate orthophosphate from phosphomonoesters, and -3 to -5 to separ ate pyrophosphate and polyphosphosphate end groups. 223

PAGE 224

Hydrolytic Enzyme Assay The activities of two hydrolytic enzymes critical to P cycling (phosphomonoesterase and phosphodiesterase) were determined using fluorogenic substrates based on a standard microplate assay (Marx et al. 2001). For each sample, soil suspensions were prepared in a 1: 100 soil/water ratio (containing 1 mmol L-1 NaN3 to prevent microbial activity). Soil suspensio n (50 L) was then pipetted into wells on a micro-well plate (8 wells per substr ate) containing 100 L of 200 mol L-1 substrate (4methylumbelliferyl phosphate and bis(4-methylumbelliferyl) phosphate respectively ) and 50 L of 200 mmol L-1 sodium acetateacetic acid bu ffer adjusted to pH 4.0. Plates were incubated for 30 min at 26C to approx imate the daytime soil temperature in the Changinola peat deposit. The reaction was terminated by adding 50 L of 0.5 mol L-1 NaOH (final solution pH > 11) and the pl ates were read immediately on a FLUOstar Optima multi-detection plate reader (BMG Labtech, Offenburg, Germany), with excitation at 360 nm and emission at 460 nm. Control plates containing substrate, buffer, and 1 mmol L-1 NaN3 (no soil suspension) were prepared and analyzed immediately prior to and a fter the analysis of soil samples to account for initial fluorescence as well as pH induced instability of substrates. Each soil had corresponding blanks, methylumbellifer one (MU) standards and soil specific quench standards. All enzyme activities ar e expressed here as mol MU h 1 g 1 total C or when comparing with soil total P as mol MU h 1 g 1 soil. Data Analysis All statistical tests were performed in SPSS for windows version 17.0.0 statistical software (SPSS Inc. 2008). Data was checked for normality by application of a ShapiroWilk test and visual inspection. If the a ssumption of normality was improved, natural 224

PAGE 225

logs were used for statistical analysis. Differences between basic biogeochemical characteristic were explored via a si mple ANOVA, with a univariate GLM and site number as the fixed factor. Post hoc analysis (Tukey HSD) was used to identify significant homogeneous subsets. After confirmation of the suspected nutrient gradient, the relationship between phosph atase enzyme activity per gram of total C and total soil P was explored by use of the SPSS curve estimation of an inverse function. Comparison of extraction efficiencies betwe en methods used a simple paired t-test as appropriate. Investigation of patterns seen within functional groups as determined by 31P NMR analysis used average parameters of combined samples (n=3). Given a lack of sample material 31P NMR spectra for site 9 were determined on additional samples (collected Nov 07). Sample total P was shown to be not significantly different (t-test: p > 0.05) but given AEM extraction was carried out at a different time, site nine was excluded from later correla tion analysis of total i norganic polyphosphates and fumigation-released microbial P Results Soil Biogeochemical Properties Surface peats from across the wetland transec t were highly acidic (pH 3.7 0.4) and of low bulk density, ranging from 0.03 g cm-3 within the central bog plain to 0.08 g cm-3 within forested portions of the transect (Table 8-1). Surface samples showed a range of total C concentrations from 4154% with a significant difference between sites (ANOVA: p < 0.001). Given external mineral i nputs to this ombrotrophic system are minimal (Phillips and Bustin 1996) such di fferences reflect differences in biosilica deposition from opal phytoliths and various di atoms (Kokfelt et al. 2009; Lopez-Buendia et al. 2007; Street-Perrott and Barker 2008) or variance in C decomposition and peat 225

PAGE 226

accretion rates across the peat dome (Craft and Richardson 1993). Total Ca and K showed no significant difference between sites (ANOVA: p > 0.05) whilst total Mg showed a significant difference (ANOVA: p < 0.001), although there was no clear trend across the transect. Total N and P demons trated expected trends between sites (ANOVA: p < 0.001) from a relatively nutrient enriched site 1 (total N= 29 mg g-1, total P = 1.0 mg g-1) to oligotrophic site 9 (total N =22 mg g-1, total P = 0.4 mg g-1). In both cases post hoc analysis (Tukey HSD) showed 4 homogeneous subsets across the gradient with proximate sites showing no si gnificant difference to each other. When nutrient concentrations were expressed on a volumetric basis the very low bulk densities of the surface peat within the central r egions exacerbates the biologically relevant gradient in nutrient content of t he peat across the transec t (Figure 8-2 A, B). With a highly significant (ANOVA: p < 0.001) fivefold increase in total P ( 14.6 to 70.9 g. cm-3 ) and almost three fold increase in total N (0.734 to 2.0 mg cm-3). Molar ratios of total soil N to P also suggest a signifi cant increase in the degree of P deficiency towards the central portions of the peat dome. Yet the relatively high total elemental molar ratios, N:P (ranging from 28 to 50) and C:P (ranging from 485 to 1291) at all sampling sites (Figure 8-2 C, D) is suggestive of P limitation even at the peripheral sites (Cleveland and Liptzin 2007). Phosphorus Biogeochemistry Levels of bioavailable P recovered by AEM strips were below 3 g g-1 for sites 5 through 9 with other sites showing an increa sing recovery to a maximum of 29 g g-1 (2.8 % of total P) at site 1 (Table 8-2) Phosphorus released by hexanol fumigation showed significant differences among sites (ANOVA: p < 0.05) and represented between a remarkable 18 and 38% of soil total P. Although fine root biomass may have 226

PAGE 227

been included in this fumigation released pool, it is clear that a significant proportion of total P is present wit hin viable cells. Both hydrolytic enzyme activities assayed showed significantly differences (ANOVA: p <0.001) between study sites, with central sites showing rates up to seven times that of the peripheral site 1 (Figur e 8-3). Given the known influence of nutrient status on hydrolytic enzymes the relations hip between total P and enzyme activity was explored via line fitting of an inverse rela tionship. This showed highly significant ( p < 0.001) inverse relationships between to tal soil total P and phsophomonoesterase (R2 = 0.545) and phosphodiesterase (R2 = 0.740) activity per g of soil. Phosphorus recovery in NaOHEDTA Alkaline extraction of air dried soils or in conjunction with AEM of fresh soils recovered up to an average of 67% of soil tota l P (Figure 8-4). A simple paired t-test showed significant difference between extr action efficiency of non fumigated and fumigated fresh soil (paired t-test: p < 0.001), with the majority (95% of the difference) being attributable to P recovered by the AEM during the fumigation step. Although significantly differ ent (paired t-test: p = 0.016), it is interesting to note that the extraction of fresh soils after hexanol fumigation recovered similar levels to that after air drying suggesting straight alkaline extraction of air dried soils recovers P pools associated with microbial biomass (Turner et al. 2003c). Ther e was a significant increase in P recovery between extraction of air dried soils for 4 h and 16 h (paired t-test: p < 0.001) but with an average relative difference of only 6.6% (4.11% of total P). Given known hydrolysis of various compounds in alkaline solution (Dool ette et al. 2009; Turner et al. 2003d) the decision was made to use the standard 4 h extraction for detailed 31P NMR analysis. Repeat extractions using the standard 4 h provided a consistent recovery ( 10% of 227

PAGE 228

total P) based upon molybdate colorimetry of digested alkaline solutions (Data not shown). Solution 31P NMR spectroscopy Spectra acquired demonstrated a diverse r ange of P forms to be present in the surface soils (Figure 8-5, Table 8-3) with si gnificant levels of organic P (phosphonates, phosphomonoesters and phosphodiesters) as well as inorganic poly-phosphoric compounds. The concentra tion of all components, ot her than the inorganic polyphosphates, identified by 31P NMR analysis showed significant ( p < 0.001) positive linear correlation with total soil P. The conc entration of residual P, not extracted or identified by the NMR analys is showed significant ( p < 0.05) positive correlation with total P, but as a proportion of to tal P showed a highly significant ( p < 0.005) negative correlation, ranging from 29% of total P at the re latively enriched site 1 to 55% at site 9. Detected phosphonates were found to represent up to 31 g g-1 and ranged from 1.7 to 3.3 % of total P within sites 1-7. Their lack of detecti on at the low P sites (site 8 and 9) is probably attributable to low signal to noise ratio and an inability to resolve their presence as opposed to a true absence. P hosphodiesters other than DNA were found at similar low levels across all sites (0.6 to 3 % of total P) but it should be noted that, due to alkaline hydrolysis, this may represent a significant underestimation of important phosphodiesters such as phosphatidyl cho line (McDowell and Stewart 2005b; Turner et al. 2003d). Concentrations of DNA ranged from 105 to 47 g P g-1 and represented from between 8.7 and 13.3 % of total P. While phosphomonoesters were found to be a significant component of the organic P pool ranging from 174 to 45 g P g-1 11.5 and 17.1 % of total P, yet did not contain peaks characteristic of commonly found isomers of inositol hexakisphosphate (Figure 8-6). T here was a significant positive correlation 228

PAGE 229

(Pearsons rho = 0.790, p < 0.05) between soil total P and t he ratio of P identified as phosphomonoesters and total phosphodiesters, ranging from 1.40 at site two to a low of 0.88 at site eight. Inorganic polyphosphates were found to consti tute a major fraction, up to 24% of total soil P, of all soil samples tested (Table 8-3). The presence of inorganic polyphosphates in native soils was confirmed, using a limited number spectra from fresh soil extracts (Figure 8-7) which when coupled to the rapid air drying used (< 10 days) suggests their presence was not an artifact d ue to fungal growth (Koukol et al. 2008) but the fact that polyphosphates represent an important insitu pool. Analysis of fresh samples after application of AEM strips also provided insight into the location of this significant P pool. The AEM strips used in this study have been shown to recover significant levels of polyphosphate fr om solution (Cheesman et al. 2010b), when extracts were analyzed after application of AEM but without the use of hexanol as a biocide, polyphosphates were recovered (NF samples). When coupled to sample fumigation polyphosphates were no longer detected or were found only in trace concentrations (Figure 8-7). This would suggest the presence of polyphosphates within cellular structures (Kornberg et al. 1999) or complexes disrupt ed by the action of hexanol (Myers et al. 1999) In addi tion, total inorganic polyphosphates (pyrophosphates and long chain polypho sphates) determined by solution 31P NMR spectroscopy of dried soil extracts showed si gnificant positive correlation (Pearsons rho = 0.804, p < 0.05) with fumigati on released P as determined by AEM on fresh soils. Discussion Basic characterization of the sampling transect confirmed previous studies in identifying the presence of a distinct and remarkable nutrient gradient across distinct 229

PAGE 230

vegetation communities within the Changu inola peat dome (Sjgersten et al. 2010; Troxler 2007). Total P was shown to be at t he high end of the range observed in other tropical ombrotrophic systems, with sites in Kalimantan ranging from 272 to 373 g g-1 (Page et al. 1999), and the Peruvian lowland Amazonia ranging from 130 to 590 g g1(Lahteenoja et al. 2009a). This may reflect eit her the relatively young age and shallow depth of peripheral sites as compared to other more established peat profiles (Phillips et al. 1997), or proximity to the coast giv en the potential for bot h direct and occult deposition of oceanic P. Bioavailable, esti mated by AEM extractable, and total P demonstrated a significant increase towards peripheral Raphia taedigera sites, yet the high molar ratios of total C:P, and total N:P, in addition to significant levels of hydrolytic phosphatase enzyme activity, suggest the potentia l for P limitation wit hin the accreting organic matter of all sites (Cleveland and Li ptzin 2007). Previous evidence based upon foliar nutrient ratios and 15N values (Troxler 2007) has been used to highlight a potential shift from N to P lim itation of trees across the peat catena, further work is therefore required to explore the potential for differential nutrient limitations between above and below ground biomass within and across this wetland system (Sundareshwar et al. 2003). Phosphomonoesterase activity at peripheral sites (Site1-3) was found to be similar to the only study I am aware of, that has reported enzyme activity under similar environmental conditions. Jackson et al (2009) found potential activities of ~2.75 mol MU h-1 gOM-1 in samples from a Malaysian forested peat swamp, suggesting the potential for high organic P turnover in tr opical peatdomes (Quiq uampoix and Mousain 2005). Comparison of potential enzyme activi ties reported between studies using non 230

PAGE 231

standard assays (Marx et al. 2001; Wallenstein and Weintraub 2008) and in soils of different physiochemical characteristics (D rouillon and Merckx 2005) may be erroneous, yet it is clear that within our study there is a clear increase in the enzyme potential towards the more P deficient central sites suggesting an increase in potential organic P turnover. Across all sites hexanol fumigation and AEM extraction showed that a substantial proportion of total P was held within live bioma ss. Although it is likely that fine roots and matrix bound P liberated by the action of hexanol contributed to this pool, fumigation methods are known to underestimate microbial biomass (Brookes et al. 1982; Myers et al. 1999), it is therefore likely that a substantial proportion of P identified in the NMR spectra represents a high contribution from live microbial biomass. A similar conclusion is drawn when comparing the extraction efficiency of the coupled AEM and alkaline extraction of fumigated fresh soil to that of air dried soil (Figure 8-4). The large and significant increase in extr action efficiency between AEM and alkaline extraction of nonfumigated soil versus air dried soil is contrary to the known influence of pretreatment in other peat based soils. In Turner et al (2007b) direct alkalin e extraction of fresh and air dried soils for 16 h tended to result in a similar recovery of P. I interpret this as evidence that in the highly fibrous peats of this study system hexanol fumi gation, or drying and grinding, disrupts physical structures or lyses live biomass ot herwise not extracted during the short 4 h extraction period used. The nature of phosphorus forms identified by solution 31P NMR spectroscopy was remarkably similar between all sites and phasi c vegetation zones, yet their proportions appeared to show progressive alteration in relation to basic biogeochemical 231

PAGE 232

characteristics. This suggest that microbial processing (in response to physiochemical conditions) dictates the standing pools of P found within organic so ils rather than the nature of the above ground biomass. All sites showed similarities to previously studied acidic peatlands, as well as characteristi cs seen in other high organic wetland systems. Two peaks (20.64 and 19.14 ppm) assigned to phos phonates, while not present in the highly studied calcareous peat lands of south Florida (Tur ner and Newman 2005; Turner et al. 2006a) have been found in acidic nort hern hemisphere blanket bogs (Bedrock et al. 1994; Turner et al. 2003b). Suggesting that phosphonates are either more prevalent within biomass present (Ternan et al. 1998) or that they expe rience greater extracellular stability under acidic conditions. This study also showed that, similar to other organic wetlands, a significant proportions of or ganic P was found as phosphodiesters or as potential phosphodiester hydr olysis products (Turner and Newman 2005). This is in contrast to terrestrial so ils where up to 90% of organic P may be identified as phosphomonoesters (Condron et al. 2005), with is omers of inositol hexakisphosphate forming a substantial proportion of total organic P(Turner et al. 2003f). This distinction has been attributed to different ial stabilization in the organi c matter and redox conditions prevalent to wetlands (Celi and Barberis 2005a; Turner et al. 2006a), though the recent detection of the phosphomonoester myo -IP6 within anaerobic sediments (McDowell 2009; Turner and Weckstrm 2009) would suggest a complex interplay between substrates and physiochemical conditions. The distinction between terrestrial soils and wetlands may represent merely site differenc e in the proportion of P found within viable microbial biomass as compared to the ex tracellular environment (Oberson and Joner 2005). 232

PAGE 233

Given evidence for increased P defic iency across the study transect an unexpectedly large proportion of total P wa s found as inorganic polyphosphates from all sites. Polyphosphates are known to play an in tegral role in archeal, prokaroyotic and eukaryotic cells (Kornberg et al. 1999) and are often associ ated with luxury microbial P uptake (Hupfer et al. 2007; Khoshmanesh et al. 2002) and activated sludge processing (Reichert and Wehrli 2007), yet polyphos phates have been detected in a range of natural wetland/aquatic systems, including oligotrophic lake sediments (Ahlgren et al. 2006a; Hupfer et al. 2004), Carolina bays (S undareshwar et al. 2009), and the humic acid fraction of re-seeded peatlands (Bedro ck et al. 1994). The large concentrations detected within this study may represent intracellular stores associated with P homeostasis or microbe sporulation (Bro wn and Kornberg 2004) under conditions of fluctuating redox (Davelaar 1993), or may represent a generalized metabolic response to environmental stress or nutrient defic iency within an ombrotrophic environment (Seufferheld et al. 2008). Further work w ould be needed to elucidate the source and dynamic nature of polyphosphates under field conditions. It is interesting to note that Malaysian peat domes have a highly active su rface layers dominated by Acidobacteria and Crenarchaeota (Jackson et al. 2009), if ty pical of tropical peatdome systems such unusual assemblages may explain the elevat ed role of long chain polyphosphates as contrasted with other oligot rophic freshwater peatlands (Turner and Newman 2005). It should also be noted that polyphosphates, alt hough stable under the alkaline extraction conditions, are catalytically degraded by t he presence of divalent cations (Harold 1966). The addition of EDTA used here should of improved recovery of polyphosphate (Cade-Menun and Preston 1996) yet some resear chers interested in their role in 233

PAGE 234

234 lacustrine and marine sediments have adopted a pre-extraction steps with EDTA (Hupfer and Gachter 1995). Ther efore it must be consider ed that the detection of significant levels within this ombrotrophic, and presumably low iron, study site may be a result of their reduced degradation during extraction and NMR spectroscopy as compared to other wetland sites. Interpretation of basic biogeochemical parameters suggests a reduction in P availability and an increase in potential organic P cycling towa rds the central portion of the study wetland. This corresponds with a pronounced increase in the proportion of alkaline stable (residual) P as well as a decrease in the phosphomonoester to phosphodiester ratio. Microbial modification of P forms withi n the detritus material of organic based systems has been shown to be dependent upon nutrient availability (CHAPTER 7). I contend that there are two potential mechani sms that could account for the observed patterns, yet giv en a recognized limitation of 31P NMR as applied to bulk soil is its inability to differentiate betw een viable biomass P and that stabilized in the extracellular environment, I am currently unabl e to differentiate between them. On the one hand observed patterns could reflect the increased biological demand and turnover of organic P within central porti ons of the wetland resulting in the relative loss of phosphomonoesters which would be otherwise abioti cally stabilized in the extracellular environment (Celi and Barberis 2005a) and the accumulation of highly humified alkaline stable organic forms of P. Secondly, that ch anges in composition of bulk soil may reflect changes in constitutive components of microbia l biomass itself a major fraction of total soil P (Makarov et al. 2005).

PAGE 235

1 2 Table 8-1. Soil biogeochemical characteristics from nine sampling stations across an om brotrophic peat dome. Values are averages standard deviations of th ree replicate samples at each site Site Veg pH Bulk density (g cm-3) Total Elements C N P Ca K Mg ---------mg g-1 -------------------------------g g--1 -----------------------1 1 3.6 .12 0.069 .003 498 .7 29 .3 1028 145 539 308 2 3.8 .07 0.064 .005 508 .2 29 .6 1014 138 484 432 3 3.9 .03 0.060 .005 515 .9 28 .4 956 134 546 414 4 2 3.7 .02 0.078 .007 535 .6 26 .8 655 85 54 409 334 5 3.9 .05 0.064 .004 506 .6 28 .1 710 132 476 491 6 3 3.6 .17 0.056 .003 507 .6 25 .3 659 176 400 761 7 4 3.7 .11 0.040 .003 458 .1 23 1.1 672 169 561 787 8 3.8 .19 0.050 .002 500 .8 19 .1 388 248 452 1150 9 5 3.6 .16 0.033 .003 417 .0 22 .0 442 68 3 479 324 3 4 5 6 = Permanent vegetation sampling plot established by Sjgersten et al. (2010) Table 8-2. Phosphorus forms identif ied by anion exchange membrane technique applied to fresh soil samples. Values are averages standard deviations of th ree replicate samples at each site. Site AEM extractable g g-1 Fumigation released g g-1 % total P 1 29.5 9.6 184 .5 18 .6 2 17.6 4.2 232 .1 23 .4 3 27.6 7.9 193 .4 20 .1 4 5.4 8.1 138 .5 21 .0 5 0.1 0.0 267 .2 38 .9 6 0.9 0.7 190 .3 29 .7 7 2.9 2.1 148 .4 22 .6 8 0.2 0.0 110 .5 28 .5 9 0.3 0.2 151 .0 35 .37 8 9 235

PAGE 236

Table 8-3. Phosphorus forms identified by solution 31P NMR spectroscopy. Values equal concentrations as determined by proportion of spectra area applied to total P determined by di gest of alkaline extracts. Site Phosphonates Orthophosphat e Phosphomonoesters DNA Other Phosphodiesters Pyrophosphate Polyphosphate Residual EG MG ------------------------------------------------------------------g P g-1 soil (% of total P) ------------------------------------------------------------------1 27 (3) 258 (25) 167 (16) 105 (10) 29 (3) 32 (3) 34 (3) 76 (7) 301 (29) 2 31 (3) 233 (23) 174 (17) 93 (9) 30 (3) 20 (2) 14 (1) 107 (11) 310 (31) 3 18 (2) 170 (18) 143 (15) 91 (10) 26 (3) 11 (1) 25 (3) 138 (14) 333 (35) 4 15 (2) 129 (20) 79 (12) 57 (9) 10 (1) 26 (4) 19 (3) 61 (9) 260 (40) 5 16 (2) 82 (12) 86 (12) 72 (10) 14 (2) 4 (1) 18 (3) 147 (21) 271 (38) 6 16 (2) 83 (13) 75 (11) 61 (9) 9 (1) 3 (<0.5) 9 (1) 135 (20) 268 (41) 7 11 (2) 76 (11) 89 (13) 90 (13) 10 (2) 3 (<0.5) 9 (1) 124 (19) 260 (39) 8 nd 29 (7) 45 (11) 47 (12) 3 (1) 2 (1) 4 (1) 65 (17) 193 (50) 9 nd 54 (9) 68 (12) 57 (10) 3 (1) 7 (1) n.d 72 (12) 317 (55) n.d = not detected, trace. = Polyphosphate EG= End residue. MG = Mid chain residue = Spectra acquired on additional sample 236

PAGE 237

Figure 8-1. Overview of study transect and sampling sites with the Changuinola peat deposit, San San Pond Sak N.W. Panam a. Access route originates at a Canal cut in 1908 and approximates previ ous leveling transects of the site (Phillips and Bustin, 1996). 237

PAGE 238

Figure 8-2. Nutrient gradient; A) mass of to tal P, B) mass of total N, C) Molar ratio C:P, and D) N:P from nine study sites within the Changuinola peat deposit. All attributes show significant overall differences between sites (ANOVA: p < 0.001) with superscript indicating homogenous subs ets as derived from post hoc (Tukey HSD: p < 0.05). General vegetation groupings based upon phasic communities identifi ed by Philips et al (1998). C A B D 238

PAGE 239

Figure 8-3. Enzyme activity from nine st udy sites within the Changuinola peat deposit. Both A) phosphomonoesterase and B) phosphordiesterase activity show significant difference between sites (ANOVA: p < 0.001) with superscript letters indicating homogeneous subsets as derived from (Tukey HSD: p < 0.05) 239

PAGE 240

4 Recovery (% total P) Fresh No 4h Fresh Yes 4h Air Dried 4h Air Dried 16 h Treatment: Fumigation: Extraction time: Figure 8-4. Comparison of P recovered by alkaline extraction (0.25 mol L-1 NaOH and 0.05 mol L-1 EDTA) of air died soils (4 h and 16 h) or in addition to AEM extraction of non fumigated and fumigated fr esh soil. Bars represent averages (n =27), error bars = 1 standard deviation 240

PAGE 241

Figure 8-5. Solution 31P NMR spectra showing range of P forms present in surface soils from select sites across the study tr ansect. Spectra plotted using 15 Hz line broadening referenced and scaled using internal standard MDP ( = 17.46 ppm). Due to lack of sample materi al, spectra () acquired on additional samples collected in Nov 07. Additional samples total P shown to be not significantly different to original samples (t-test: p > 0.05). 241

PAGE 242

Figure 8-6. Detail of solution 31P NMR spectra from site se ven soils. Spectra plotted using 2 Hz line broadening and referenced using internal standard MDP ( = 17.46) 242

PAGE 243

Figure 8-7. Solution 31P NMR spectra of site one and ni ne soils after application of anion exchange membranes with (F) and wit hout (NF) fumigation step using hexanol. Spectra plotted using 15 Hz line broadening referenced and scaled using internal standard MDP ( = 17.46 ppm). 243

PAGE 244

CHAPTER 9 SUMMARY AND CONCLUSIONS Anthropogenic alteration of global P cycling has been profound, and given expected global population and consumption growth we are lik ely to see the continued and increased impacts of a disrupted P cycl ing on natural ecosystems. Due to their position in the landscape wetlands are often a focus of this di sruption, with many wetlands showing a degradation or shift in ec osystem function due to P loading. Yet, wetlands also act to sequester P in the l andscape, providing a mechanism by which downstream systems may be prot ected from adverse impacts. The dynamic interaction of biological communities and P within wetlands has been the focus of much study, and although known to constitute a sizable pr oportion of total P (CHAPTER 2), little information exists as to the functional forms biogenic P may represent in wetland soils. This dissertation has employed solution 31P NMR spectroscopy to investigate both the forms of P found within wetlands and the potent ial mechanistic drivers that determine biogenic P composition in soils. Given a lack of existing information on biogenic P composition in wetlands, initial studies focused on surveying composition within a diverse range of wetland systems. The observation of potential mech anistic groupings, in conjunction with recent literature sources, resulted in working hypotheses, tested in subsequent chapters. The complete dissertation aimed to elucidate the dynamic nature of biogenic P in wetland soils. Specific experimental objectives were: Determine the influence of wetland char acteristics and soil physicochemical properties on forms of biogenic P in wetland soils. Hypothesis: The composition of biogenic P in wetland soils varies systematically with respect to wetland characterist ics, landscape posit ion, and/or soil biogeochemical properties. 244

PAGE 245

Determine how position in the landscape, i.e. from terrestrial to wetland environments, influences biogenic P composition of soils. Hypothesis: Landscape position impacts soil properties, which in turn influence biogenic P composition. Spec ifically, higher productivity, a receiving position in the landscape and reduced decomposition leads to increased organic matter content within wetlands. This leads to differences in mechanisms of abiotic stabilization leading to hydroperiod correlating with a decrease in the ratio of phosphomonoesters to phosphodiesters (predominantly DNA). Determine how anaerobic conditions impact biogenic P composition. Hypothesis: Anaerobic conditions des tabilize the phosphomonoester myo -IP6 and polyphosphates, and lead to reduced decomposition of phosphodiester DNA. Assess the role of nutrient availabilit y in determining biogenic P composition within wetland detritus and soils. Hypothesis: Increased P availability, due to elevated ambient conditions, will reduce turnover of biogenic P by microbes, thereby altering P composition within wetland materials. Biogenic Phosphorus Composition in We tlands (Experimental Objective 1) Initial studies evaluated biogenic P compos ition in the surface soils (0 cm) of 28 freshwater wetlands, representing a di verse range of climatic conditions and hydrogeomorpic types. Characterization and biogeochemical analysis alongside determination of biogenic P forms using solution 31P NMR spectroscopy, allowed for simple ordination and the identif ication of emergent patterns with respect to the biogenic P found. Although P compositi on was independent of dire ct influence from wetland vegetation or climatic setting, biogeochemical characteristics, themselves a product of wetland setting, could be used to group sites in term of the biogeni c P forms found. The simple delineation of wetland sites with respect to pH and or ganic matter content showed clear distinctions between the