<%BANNER%>

Investigating Re-Vegetation Patterns, Nutrient Limitation, and Storages of Nitrogen as Determinants of Restoration Success

Permanent Link: http://ufdc.ufl.edu/UFE0021540/00001

Material Information

Title: Investigating Re-Vegetation Patterns, Nutrient Limitation, and Storages of Nitrogen as Determinants of Restoration Success
Physical Description: 1 online resource (278 p.)
Language: english
Creator: Keppler, Angelique M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: decomposition, diversity, ecosystem, enzymes, invasion, isotopes, limitations, nitrogen, nutrient, phosphorus, plant, restoration, 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: The role of biodiversity in the function and stability of ecosystems has long been the object of scientific debate. Past research has found that variations in ecosystem function are related to differences in the functional characteristics, especially resource capture and utilization, of the dominant plants. The Hole-in-the-Donut (HID) region of the Everglades National Park (ENP) offers a unique opportunity to investigate the successional development of vegetation and the impact of macrophyte diversity on ecosystem functions, including productivity, nutrient-use efficiency and turnover. Historically, the HID was dominated by short hydroperiod prairies and pinelands. After 1916, farming practices were employed that altered approximately 4000 ha of natural vegetation. After 1975, farming ended and the HID was aggressively colonized by a non-native pest plant Schinus terebinthifolius (Brazilian pepper). The goal of my study was to determine; limitations on nitrogen and phosphorus, the nutrient-use efficiency of the vegetation, and long-term nitrogen storages in each restored site as well as the surrounding native communities. Additionally, I evaluated relationships between macrophyte diversity and soil characteristics and processes. Results of this study offer evidence that, with time, a diverse plant community similar to a native wetland community can develop after restoration by means of scraping away all soil. Additionally, this study indicated that soil processes have a greater influence on ecosystem N dynamics than does the plant community composition. We found that the soil in the native community stores significantly more N, is P-limited, and the dominate vegetation will regenerate more nutrients as compared to the restored wetland ecosystem soil and vegetation.
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 Angelique M Keppler.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Reddy, Konda R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2008-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021540:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021540/00001

Material Information

Title: Investigating Re-Vegetation Patterns, Nutrient Limitation, and Storages of Nitrogen as Determinants of Restoration Success
Physical Description: 1 online resource (278 p.)
Language: english
Creator: Keppler, Angelique M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: decomposition, diversity, ecosystem, enzymes, invasion, isotopes, limitations, nitrogen, nutrient, phosphorus, plant, restoration, 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: The role of biodiversity in the function and stability of ecosystems has long been the object of scientific debate. Past research has found that variations in ecosystem function are related to differences in the functional characteristics, especially resource capture and utilization, of the dominant plants. The Hole-in-the-Donut (HID) region of the Everglades National Park (ENP) offers a unique opportunity to investigate the successional development of vegetation and the impact of macrophyte diversity on ecosystem functions, including productivity, nutrient-use efficiency and turnover. Historically, the HID was dominated by short hydroperiod prairies and pinelands. After 1916, farming practices were employed that altered approximately 4000 ha of natural vegetation. After 1975, farming ended and the HID was aggressively colonized by a non-native pest plant Schinus terebinthifolius (Brazilian pepper). The goal of my study was to determine; limitations on nitrogen and phosphorus, the nutrient-use efficiency of the vegetation, and long-term nitrogen storages in each restored site as well as the surrounding native communities. Additionally, I evaluated relationships between macrophyte diversity and soil characteristics and processes. Results of this study offer evidence that, with time, a diverse plant community similar to a native wetland community can develop after restoration by means of scraping away all soil. Additionally, this study indicated that soil processes have a greater influence on ecosystem N dynamics than does the plant community composition. We found that the soil in the native community stores significantly more N, is P-limited, and the dominate vegetation will regenerate more nutrients as compared to the restored wetland ecosystem soil and vegetation.
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 Angelique M Keppler.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Reddy, Konda R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2008-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021540:00001


This item has the following downloads:


Full Text

PAGE 1

1 INVESTIGATING RE-VEGETATION PATTE RNS, NUTRIENT LIMITATIONS, AND STORAGES OF NITROGEN AS DETERMI NANTS OF RESTORATION SUCCESS By ANGELIQUE MARIE KEPPLER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Angelique M. Keppler

PAGE 3

3 To my daughter, Victoria, for all your patience and love!

PAGE 4

4 ACKNOWLEDGMENTS First and forem ost, I thank my daughter, Victor ia, for her understandi ng of the hard work and time commitment this took to complete. I coul d not have done this wit hout her. I thank my grandparent, Helen and Bernard Keppler, for giving me the hope that I could do more for myself than my parents provided. Without their life-long love and support, I would not have made it to this point in my life. Many thanks go to Rich ard Bochnak for his encouragement that I could do this regardless of the obstacles that stood in my way. Thank you for allowing me to share my work, enthusiasm, and frustrations with you no matter how bored you found the information. To all my family members who never lo st faith in me, you know who you are! To Dr Ramesh Reddy: I would never have been able to design and implement this project anywhere else. This has truly been a rare oppo rtunity. You have given me a chance to design, manage and complete the research study of my dreams. Thank you for supporting this work and encouraging me through all the good and bad times that came my way. I will cherish the time we worked together throughout my career. I thank my research committee, Drs. Nick Comerford, Mark Clark, Michelle Mack, and Jim Sickman. You have been a great committee to work with and aspire to impress. Thank you for challenging me to be my best. I have en joyed the classes, sampli ng trips or individual conversations with each of you! To Gavin Wilson, thank you for all your advice and help in the lab, I couldnt have done this without you. Your knowledge is like no other I have come across thus far. Thank you to Dawn Lucas for teaching me all you know about nitrogen analysis. You have been a great support even though you were not required to help. To Jason and Kathy Curtis, thank you for all your help with the mass spec analysis. Because of your assistance and training chapters 6 and 7 were able to be completed.

PAGE 5

5 To the many students of the WBL: I tha nk Melissa Martin for keeping me grounded. I will miss our intellectual conversations; they have helped me to understand my work better than anything else. I thank Adrienne Frisbee, Isabel a Torres, Caitlin Hicks, and Karen Hupp for the countless conversations and support you each pr ovided. I thank each and every person who endured the HID sampling trips. It was hot and the mosquitoes were thick, thanks for helping me out. Finally, I would like to thank the U.S. Depa rtment of Interior, National Park Service Everglades National Park for partial funding to support this res earch. I would like to give a special thanks to the memory of Michael Norlan d for his initial contribution to the development of this project.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ........10 LIST OF FIGURES.......................................................................................................................13 ABSTRACT...................................................................................................................................18 1 INTRODUCTION..................................................................................................................20 Rationale and Significance.....................................................................................................21 Problem Statement.............................................................................................................. ....25 History.............................................................................................................................25 Preliminary Work............................................................................................................ 28 Project Objectives............................................................................................................29 2 RELATIONSHIPS AMONG TIME, PLANT SP ECIES RIC HNESS, COMPOSITION AND ECOSYSTEM FUNCTION IN REST ORED SUBTROPICAL WETLANDS........... 37 Introduction................................................................................................................... ..........37 Methods..................................................................................................................................39 Site Description...............................................................................................................39 Species Diversity Indices................................................................................................40 Above-ground Biomass and Species Dominance............................................................ 41 Results.....................................................................................................................................42 Species Composition and Diversity.................................................................................42 Nutrient Concentrations and Pools..................................................................................44 Diversity and Biomass Relationships.............................................................................. 45 Discussion...............................................................................................................................45 Diversity and Species Composition Patterns................................................................... 46 Ecosystem Function......................................................................................................... 48 Applications to Wetland Restoration and Mitigation ...................................................... 50 Conclusions.............................................................................................................................51 3 MULTIVARIATE ANALYSIS OF PLANT COMMUNITY STRUCTURE AND RELATIONSHIPS TO ECOSYSTEM CHAR ACTERISTICS IN SUBTROPICAL RESTORED W ETLANDS..................................................................................................... 68 Introduction................................................................................................................... ..........68 Methods..................................................................................................................................70 Site Description...............................................................................................................70

PAGE 7

7 Soil Physical and Chemical Analysis.............................................................................. 71 Vegetation Nutrient Analysis.......................................................................................... 72 Statistical Analysis.......................................................................................................... 73 Results.....................................................................................................................................73 Species Composition.......................................................................................................73 Environmental Parameters............................................................................................... 74 Soil and Plant Relationships............................................................................................ 75 Discussion...............................................................................................................................78 Conclusions.............................................................................................................................81 4 NUTRIENT-USE EFFICIENCY AND POTEN TIAL NUTRIENT LIMITATIONS IN SUBTROPI CAL RESTORED WETLANDS........................................................................ 93 Introduction................................................................................................................... ..........93 Methods..................................................................................................................................96 Site Description...............................................................................................................96 Soil Analysis....................................................................................................................96 Vegetation Analysis......................................................................................................... 97 Nutrients...................................................................................................................97 Nutrient-use efficiency and nutrient-resorption efficiency ......................................98 Statistical Analysis.......................................................................................................... 99 Results.....................................................................................................................................99 N and P Concentrations...................................................................................................99 Nutrient Limitations...................................................................................................... 100 Discussion.............................................................................................................................104 Community Level Limitations...................................................................................... 105 Species Level Limitations............................................................................................. 107 Conclusions...........................................................................................................................109 5 CONTROLS ON REGENERATION OF NUTRIENTS FROM THE DOMINANT VEGET ATION IN NATIVE AND RESTORED WELTANDS......................................... 123 Introduction................................................................................................................... ........123 Methods................................................................................................................................126 Site Description.............................................................................................................126 Decomposition Experiment........................................................................................... 126 Litter Fractionation........................................................................................................ 127 Assessment of Litter Quality......................................................................................... 128 Microbial Biomass and Extracellular Enzym e Activities............................................ 129 Statistical Analysis........................................................................................................ 130 Results...................................................................................................................................130 Decomposition Rates..................................................................................................... 130 Litter Quality.................................................................................................................131 Microbial Activity......................................................................................................... 134 Discussion.............................................................................................................................135 Conclusions...........................................................................................................................140

PAGE 8

8 6 RETENTION OF 15N IN SOIL AND VEGETATION: SEASONAL VARIATION AND LONG TERM NITROGEN STORAGE IN RESTORED WETLANDS................... 161 Introduction................................................................................................................... ........161 Methods................................................................................................................................164 Site Description.............................................................................................................164 Experimental Design..................................................................................................... 164 Soil Sampling and Analysis........................................................................................... 165 Vegetation Sampling and Analysis............................................................................... 167 Calculations...................................................................................................................168 Statistical Analysis........................................................................................................ 169 Results...................................................................................................................................169 15N Retention after 24 Hours.........................................................................................169 Soil 15N Retention after 365 Days................................................................................. 170 Soil Processes................................................................................................................171 Species-level 15N Retention........................................................................................... 171 Community-level 15N Retention....................................................................................173 Seasonal Change in 15N Storage and N Budget............................................................. 174 Discussion.............................................................................................................................175 Soil 15N Retention..........................................................................................................175 Plant 15N Retention........................................................................................................177 Conclusions...........................................................................................................................180 7 VARIATIONS IN THE NAT URAL ABUNDANCE OF 15N IN RESTORED SUBTROPICALWETLAND VEGE TATION COMMUNITIES....................................... 202 Introduction................................................................................................................... ........202 Methods................................................................................................................................204 Site Description.............................................................................................................204 Soil Analysis..................................................................................................................204 Statistical Analysis........................................................................................................ 206 Results...................................................................................................................................207 Discussion.............................................................................................................................209 Conclusions...........................................................................................................................213 8 NUTRIENT MEMORY IN INVADED AB ANDONE D FARMLAND WITHIN THE EVERGLADES NATIONAL PARK: CONSIDERATIONS AND IMPLICATIONS FOR RESTORATION..........................................................................................................225 Introduction................................................................................................................... ........225 Methods................................................................................................................................227 Site Description.............................................................................................................227 Sample Collection and Analysis.................................................................................... 228 Decomposition Experiment........................................................................................... 229 Assessment of Litter Quality......................................................................................... 229 15N Tracer Experimental Design................................................................................... 230 Statistical Analysis........................................................................................................ 230

PAGE 9

9 Results...................................................................................................................................230 Soil Characteristics........................................................................................................230 Vegetation Characteristics............................................................................................. 231 Ecosystem Functions..................................................................................................... 232 Discussion.............................................................................................................................234 Conclusions...........................................................................................................................237 9 SYNTHESIS AND CONCLUSION.................................................................................... 251 Nitrogen and Phosphorus Limitations Drive Community Development............................. 251 Nutrient Regeneration.......................................................................................................... .254 Long Term Nitrogen Storage................................................................................................ 255 Applications to Wetland Restoration and Mitigation ........................................................... 256 Future Recommendations.....................................................................................................257 LIST OF REFERENCES.............................................................................................................259 BIOGRAPHICAL SKETCH.......................................................................................................277

PAGE 10

10 LIST OF TABLES Table page 2-1 Complete species list for all plants identified in the HID during 2005. An X indicates presence in site location. All sp ecies are listed in al phabetical order. (n=60) .................................................................................................................................53 2-2 Diversity indices calculated from comple te spec ies list (Table 2-1) for individual sites and the HID as a whole.............................................................................................. 56 2-3 Summary of results from ANOVA tests with dep endant variables of as either each site or year of biomass sampling........................................................................................ 57 2-4 Summary of species carb on, nitrogen, and phosphorus cont ent, pool sizes, and pool weighted ratios. ..................................................................................................................58 3-1 List of all species identi fied during vegetation biom ass co llection in the restored sites and the native communities. (n=60 plots).......................................................................... 82 3-2 Average values for soil chemical parameters for each site used in non-metric multidim ensional ordination with vegetation co mmunity data (n=10 for each site)......... 83 3-3 Average values for soil physical proper ties for each site used in non-m etric multidimensional ordination with vegetation community data. (n=10 for each site)........ 84 3-4 Average values for vegetation nutrient parameters for each site used in non-metric multidim ensional ordination with vegetation community data. (n=10 for each site)........ 85 4-1 Summary of the average nitrogen, phosphorus, and N:P ratios of the soil for each of the sites during the dry a nd wet season. (n=10) .............................................................. 111 4-2 Summary of results from a two-way ANOVA test with dependant variables of nitrogen (N) concentration, phosphorus (P) concentration, and N:P ratios for the soil, community level vegetation (com posite biomass), and species level.............................. 112 4-3 Summary of the average nitrogen, ph osphorus, and N:P ratios of the vegetation community (com posite biomass) for each of the sites during the dry and wet season. (n=10 for dry season, n=20 for wet season)..................................................................... 113 4-4 Comparison of nutrient characteristic s at the species level for nitrogen (N) concentration, phosphorus (P) concentration, and N:P ratio. ..........................................114 4-5 Summary of results from a two-way ANOVA test with dependant variables of nutrient-resorption efficiency of nitrog en (NRE-N) and phosphorus (NRE-P), and nutrient-use efficiencies of nitrogen (NUE-N) and phosphorus (NUE-P) for the community level vegetation (com posite biomass), and species level.............................. 115

PAGE 11

11 4-6 Pearsons correlation coefficients for the seven plant comm unity level (composite biomass) characteristics addressed in this study.............................................................. 116 4-7 Pearsons correlation coefficients for the seven sp ecies le vel characteristics addressed in this study..................................................................................................... 117 5-1 Summary of initial litter quality of Cladium jamaicense and T ypha domingensis for microbial decomposition. ................................................................................................ 142 5-2 Summary of results from a two-way ANOVA test with vari ab les as species, site, and time for the litter fractionation of Cladium jamaicense and Typha domingensis ............143 5-3 Summary of results from a two-way ANOVA test with dependant variables of C:N, C:P, and N:P ratios f or the litter material of Cladium jamaicense and Typha domingensis ......................................................................................................................144 5-4 Summary of microbial biomass carbo n (MBC), nitrogen (MBN), and phosphorus (MBP) as well as the ratios of MBC:N, MBC:P, and MBN:P. ....................................... 145 5-5 Summary of results from a two-way ANOVA test with dependant variables of m icrobial biomass (C, N, and P), and microbi al biomass ratios (C:N, N:P, and C:P) for the litter material of Cladium jamaicense and Typha domingensis ...........................146 5-6 Summary of results from a two-way ANOVA test with dependant variables of GA, L-LAA, and APA f or the litter material of Cladium jamaicense and Typha domingensis ......................................................................................................................147 6-1 Summary of soil physio-chemical parameters at the start of the tracer study. (n=10) ....182 6-2 Summary of vegetation community and litte r laye r chemical parameters at the start of the tracer study. (n=10)................................................................................................ 183 6-3 Percent of total 15N as NH4Cl recovered after 24 hours of application. (n=3)............... 184 6-4 Percent of total 15N as NH4Cl recovered after 365 days of application. (n=3)................ 185 6-5 Summary of results from a two-way ANOVA test with dependant variables of total nitrogen (N) and 15Nitrogen (15N) pools..........................................................................186 6-6 A summary of individual sp ecies biom ass, nitrogen and 15N data for each site. Species are ordered from most to least abundant. (n=3)................................................. 187 7-1 Chemical and isotopic values for se lected dom inant species from each site................... 215 8-1 Summary of physical, chemical, and biol ogical param eters for the invaded, native and 2003 restored communities. (n=10).........................................................................238

PAGE 12

12 8-2 Correlation coefficients from a principa l com ponents analysis on soil environmental characteristics for the invaded, native, and 2003 restored wetland communities............ 239 8-3 Summary of the decomposition constants, k (yr-1), and turnover rates, 1/k (yr), as determined by the mass loss from a field decomposition study...................................... 240

PAGE 13

13 LIST OF FIGURES Figure page 1-1 Conceptual model of the in te rlinking relationships betw een ecosystem processes, soil development, and vegetation community as well as primary driving forces that influence vegetative community structure......................................................................... 31 1-2 False image photo of southern Florida i ndicating land cover and vegetation types. The green area within th e black circle is th e location of the restoration project of the Hole-in-the-donut...............................................................................................................32 1-3 Delineation of the region of the Everglades Natio nal Park known as the Hole-in-theDonut to be restored after the e ffects of farming and invasion of Schinus terebinthifolius...................................................................................................................33 1-4 Location of the restored wetlands within the Hole -in-the-Donut in the Everglades National Park.....................................................................................................................34 1-5 Plot locations of Everglades Research Groups reoccurring vegetation surveys in both the res tored wetlands and the surrounding native communities........................................ 35 1-6 Location of each plot included in this study in the Hole-in-the-Donut region of the Everglades National Park. Im age courte sy of the Everglades National Park................... 36 2-1 Repeated above-ground biomass co llection reported as g dry wt m-2 for both the native communities and the restored wetlands.................................................................. 59 2-2 Repeated above-ground biomass co llection reported as g dry wt m-2 and separated by species contribution for both the native communities and the restored wetlands.............. 60 2-3 Relative frequency of vege tation sp ecies collected in the biomass sampling within the restored and native communities of the HID............................................................... 61 2-4 Mean tissue content for the community le vel vegetation for each site in the HID during wet season, Y2. .......................................................................................................62 2-5 Mean carbon, nitrogen, and phosphorus accum ulation in above-ground biomass and relative contribution of plan t species found in the HID..................................................... 63 2-6 Trends in community weighted carb on, nitrogen, and phos phorus ratios between native and restored wetland communities. ......................................................................... 64 2-7 Relationship between above-ground bioma ss production and species richness for the HID during the wet season, Y2. .........................................................................................65 2-8 Relationships with the age of each restored s ite for the restored wetland communities in the HID..........................................................................................................................66

PAGE 14

14 2-9 Relationships with above-ground bioma ss production for the sites in the H ID................ 67 3-1 Non-metric multidimensional scaling (NMS ) ordination of vegetation community for each restored site at six months after restoration...............................................................86 3-2 Non-metric multidimensional scaling (N MS) ordination of vegetation community data for the 1997 and 1998 restored site s at 8 years after restoration................................ 87 3-3 Non-metric multidimensional scaling (N MS) ordination of vegetation community data for restored and native sites........................................................................................88 3-4 Non-metric multidimensional scaling (N MS) ordination of vegetation community data for restored and native sites........................................................................................89 3-5 Non-metric multidimensional scaling or dina tion (NMS) of vegetation community data for restored and native sites........................................................................................90 3-6 Relationship between soil depth and s oil N:P and C:P ratios for the native and restored wetlands within the HID......................................................................................91 3-7 Relationship between soil depth and vege tation N:P and C:P ratios for the native and restored wetlands within the HID.. ....................................................................................92 4-1 Relationship between community level ve getation N and P concentration for each site in the H ID................................................................................................................ ..118 4-2 Relationships between nitrogen (N) and phosphorus (P) concentration and N:P rations.. .............................................................................................................................119 4-3 Nutrient-use efficiency and nutrient-resorption efficiency of nitrogen and phosphorus of the vegetation community (com posite biomass)......................................................... 120 4-4 Linear regression betwee n N:P ratio of the vegetation community (com posite biomass). ..................................................................................................................... ....121 4-5 Nutrient-use efficiency and nutrient-resorption efficiency of nitrogen and phosphorus for the individual species included in this study. ............................................................. 122 5-1 Climate data from 2006 obtained from Florida Autom ated Weather Network............... 148 5-2 Conceptual diagram of temporal patterns of nitrogen release from litter material of differing quality available for m icrobial decomposition................................................. 149 5-3 Percent mass remaining for Cladium jamaicense and Typha domingensis from time zero to 365 days for each site........................................................................................... 150 5-4 Initial nutrient ratios for Cladium jamaicense and Typha domingensis plant com partments................................................................................................................... 151

PAGE 15

15 5-5 Litter fractionation for initial Cladium jamaicense and Typha domingensis senescent litter and at 168 and 365 days. .........................................................................................152 5-6 Relationships between litter quality indices of Cla dium jamaicense and Typha domingensis ......................................................................................................................153 5-7 Relationships between litter quality indices of Cla dium jamaicense and Typha domingensis ......................................................................................................................154 5-8 Final ratios after 365 days of decomposition for Cladium jamaicense and Typha domingensis ......................................................................................................................155 5-9 Change in nutrients for Cladium jamaicense and Typha domingensis for each tim e period analyzed................................................................................................................156 5-10 Relationship between Lignin:N ratio and the change in nitrogen c ontent in the litter m aterial....................................................................................................................... .....157 5-11 Relationship between Lignin:P ratio and th e change in phophorus content in the litter m aterial....................................................................................................................... .....158 5-12 Enzyme activities associated with Cladium jamaicense and T ypha domingensis during decomposition.......................................................................................................159 5-13 Alkaline phosphatase (APA) enzy m e activities associated with Cladium jamaicense and Typha domingensis during decomposition................................................................ 160 6-1 Hole-in-the-Donut hydroperiod for 2006 reported as groundwat er level above NAVD 1988 (obtained from USGS Nationa l Water Information System).................................. 188 6-2 Plot layout for the ap plication and sampling of 15N tracer study.....................................189 6-3 Total amount of 15N applied remaining after 365 days....................................................190 6-4 Total amount of 15N applied remaining after 365 days....................................................191 6-5 Potentially mineralizable nitrogen (PMN) reported as mg NH4 kg-1 soil day-1 for each sample period and each site............................................................................................. 192 6-6 Estimates of volat ilizatio n rates of NH3-N in native and restored wetland communities; d.f.=3, F=5.9. (n=3)...................................................................................193 6-7 Amount of 15N recovered in Cladium jamaicense and Typha domingensis in each site at 168 days after application............................................................................................ 194 6-8 Amount of 15N recovered in dominant vegetation sp ecies in each site after 365 days of application...................................................................................................................195

PAGE 16

16 6-9 Amount of nitrogen stored in above(liv e plus senescent) and below-ground biom ass in each site at 365 days of the study................................................................................ 196 6-10 Relationship between the bulk soil 15N retention and the above-ground biomass 15N retention for each site....................................................................................................... 197 6-11 Amount of nitrogen stored in the vegetation community pools in each site after 365 days. .................................................................................................................................198 6-12 Changes in total nitrogen storages during the year long 15N tracer study for each pool..................................................................................................................................199 6-13 Changes in total 15N nitrogen retention during the year long 15N tracer study for each pool..................................................................................................................................200 6-14 Total nitrogen and 15N budget for each component of this study. The number in parenthesis is the amount of 15N retained in each pool from the initial 650 mg m-2 applied........................................................................................................................ ......201 7-1 Hole-in-the-Donut hydroperiod for 2005 reported as groundwat er level above NAVD 1988 (obtained from USGS Nationa l Water Information System).................................. 216 7-2 Relationship between 15N values and N:P ratios of th e soil in each restored wetland and the native community................................................................................................217 7-3 Relationship between community level plant 15N values measured during the dry season (April 2005) and those measured dur ing the wet season (late July 2005)........... 218 7-4 Relationship between community level plant 15N values and the N:P ratios which correspond to each value. Each site is code d to indicate site relationships. A) Dry season. B) Wet season. (N=60)...................................................................................... 219 7-5 Relationship between the wet season community level plant 15N values and the N:P ratios which correspond to each value............................................................................. 220 7-6 Relationship between the dry season community level plant 15N values and the N:P ratios which correspond to each value fo r the 2004 restored wetland community.......... 221 7-7 Relationship between species level 15N values and nitrog en, phosphorus and N:P ratios of dominant species................................................................................................ 222 7-8 Inorganic nitrogen in each site in the HID during the wet and dry seasons. ................... 223 7-9 Relationship between the community level vegetation 15N to the inorganic nitrogen and phosphorus in each restored site............................................................................... 224 8-1 Principal component analysis for soil environm ental characteri stics in the invaded, native, and 2003 restored wetland communities.............................................................. 241

PAGE 17

17 8-2 A comparison of soil nutrients for the invaded, native, and 2003 restored wetland communities. ....................................................................................................................242 8-3 A comparison of plant tissue chemistry for S. terebinthifolius T. domingensis and S. nigricans fo r the invaded, native, and 2003 restored wetland communities................... 243 8-4 Litter cellular fractionation after 365 days of decomposition for S. tereb inthifolius T. domingensis and S. nigricans ..........................................................................................244 8-5 Percent mass remaining for Schinus terebinthifolius Typha domingensis and Schoenus nigricans from tim e zero to 365 days.............................................................. 245 8-6 Change in nutrients for Schinus terebinthifolius Typha domingensis and Schoenus nigricans fo r each time period analyzed.......................................................................... 246 8-7 Enzyme activities associated with Schinus terebinthifolius T ypha domingensis and Schoenus nigricans during decomposition...................................................................... 247 8-8 Alkaline phosphatase enzyme activity (APA) associated with Schinus terebinthifolius, Typha domingensis and Schoenus nigricans during decomposition.... 248 8-9 The percent 15N recovered in each soil pool fr action from the initial 650 mg 15N m-2 that was applied................................................................................................................249 8-10 The amount of 15N recovered after 365 days in each soil pool fraction from the initial 650 mg 15N m-2 that was applied...................................................................................... 250

PAGE 18

18 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy INVESTIGATING RE-VEGETATION PATTE RNS, NUTRIENT LIMITATIONS, AND STORAGES OF NITROGEN AS DETERMI NANTS OF RESTORATION SUCCESS By Angelique Marie Keppler December 2007 Chair: K. Ramesh Reddy Major: Soil and Water Science The role of biodiversity on th e function and stability of ecosystems has long been the object of scientific debate. Past research has found that variations in ecosystem function are related to differences in the functional characteris tics, especially resource capture and utilization, of the dominant plants. The Hole-in-the-Donut (HID) region of the Everglades National Park (ENP) offers a unique opportunity to investigate the successional development of vegetation and the impact of macrophyte diversity on ecosystem functions, incl uding productivity, nutrien t-use efficiency and turnover. Historically, the HID was dominate d by short hydroperiod prairies and pinelands. After 1916, farming practices were employed that altered approximately 4000 ha of natural vegetation. After 1975, farming ended and the HI D was aggressively colonized by a non-native pest plant Schinus terebinthifolius (Brazilian pepper). The goal of my dissertation was to determin e; limitations on nitrogen and phosphorus, the nutrient-use efficiency of the vege tation, and long term nitrogen stor ages in each restored site as well as the surrounding native communities. Add itionally, I evaluated relationships between macrophyte diversity and soil charac teristics and processes.

PAGE 19

19 The results of this study offers evidence that with time, a diverse plant community similar to a native wetland community can develop after re storation by means of sc raping away all soil. Additionally, this study indicated that soil proc esses have a greater influence on ecosystem N dynamics than does the plant community compos ition. We found that the soil in the native community stores significantly more N, is P-lim ited, and the dominate ve getation will regenerate more nutrients as compared to the restor ed wetland ecosystem soil and vegetation.

PAGE 20

20 CHAPTER 1 INTRODUCTION The role of biodiversity on the function and stability of ecosy stem s has been the object of recent scientific debate. Past re search has found that variations in ecosystem function are related to differences in the functional characteristics, especially resource capture and utilization, of the dominant plants. The implication that ecosystem processes are dependen t on higher levels of biodiversity and vise versa has become mo re evident (Loreau 2000, Chabrerie et al. 2001, Engelhardt and Ritchie 2001, Hooper et al. 2005). Research has shown that importance s hould be placed on the impact of species interactions, rather than species number alone, on ecosystem function (Johnson et al. 1996). The importance of macrophyte diversity in wetland ec osystems and its link to ecosystem function (i.e., productivity, decom position, and nutrient-use efficiency) has received little attention. A positive relationship between plant diversity and primary productivity has been demonstrated in grasslands (Naeem et al. 1994, Tilman and Downing 1994, Walker 1995, Tilman et al. 1996), but little research has been done in wetlands. Due to the high level of disturbance and destruction of natural wetlands, the loss of eco system function and diversity is inevitable. Through restoration and mitigation efforts, the loss of wetlands is decr easing, but the success of these efforts in terms of function and diversity is still unclear. To understand the relationships between the vegetative community and ecosystem processes, it is important to understand what factors govern ecosystem structure. Several models have been conceptualized to explain ecosystem development and controls on soil processes and vegetative communities (Grime 1977, van der Valk 1981, Chapin et al. 1996, Lortie et al. 2004). The conceptual model proposed here combines several concepts from these existing models, but focuses more on the close linkage between the vegetation community, soil development and

PAGE 21

21 ecosystem processes as well as the primary drivi ng forces (i.e., life characteristics, competition, potential vegetation) influenci ng vegetative community structure (Figure 1-1). The vegetation community present in an ecosystem is closely link ed to the soil properties, and the interactions between the vegetation and the soil environment influence the dominant ecosystem processes through feedback cycles. Wetland plant communities are heavily in fluenced by their hydroperiod. Wetland ecosystems that are continually flooded verses seasonally flooded can have very different vegetative community structures (van der Valk 1981). The life characteristics (i.e., life span, propagule longevity, and propagule establishmen t requirements) of wetland vegetation will determine their success under di fferent hydrologic conditions. In the case of restored wetland systems, the seed bank (potential vegetation) can become an important driving force controlling the community structure that develops. However, with time these ea rly successional plant species can be driven out through competition for nutrient and other important resources by later successional plant species (Gri me 1977, van der Valk 1981). Rationale and Significance Scientists have long acknowledged the importance of vegetation on wetland ecosystem function. Several studies have investigated the impact of nutrient enrich ment (nitrogen (N) and phosphorus (P)) on vegetative communities in wetlands (Zedler 1993, W illis and Mitsch 1995, Bobbink et al. 1998, Boyer and Zedler 1998, Bedford et al. 1999, Mahaney et al. 2004, Rickey and Anderson 2004) or the nutrient removal pot ential of wetland plants (Reddy and DeBusk 1985, DeBusk et al. 1995, Tanner et al. 1995, Tanner 1996, Oomes et al. 1997, DeBusk et al. 2001), but few studies have been able to iden tify the role of vegeta tion diversity on wetland ecosystem processes (i.e., nutrient cycling). Ch anges in species composition could have long term impacts on N cycling in wetlands. From studi es in grasslands, fens, and forest ecosystems,

PAGE 22

22 it has been suggested that species composition ma y affect N cycling and storage by influencing the rate of N uptake, the use efficiency of N, the litter quality (Vitousek 1982, Johnson et al. 1996, Aerts and deCaluwe 1997, Grime 1997, Aerts et al. 1999), and therefore the rates of mineralization of soil organic N. The amount of N available for plant uptake is determined by the balance between external inputs and outputs and by the internal cycling of nutrients (Aerts et al. 1999). As N is cycled through an ecosystem it is continuously transforme d between different chemical species, mostly through biological processes. The conversion of N into various fo rms is important to maintain the biological requirement necessary for plant an d microbial growth. Mi neralization is a key process regulating the bioavailabil ity of N for plant assimilation. It is a microbial mediated process that converts orga nic N to inorganic N (NH4); therefore rates of mineralization can be significantly correlated to microbial biomass (White and Reddy 2001). Due to the anaerobic nature of wetland soils, NH4 + is the most stable inorganic form of N. The oxidation of NH4 + to NO3 (nitrification) is limited to the water-soil inte rface where oxygen diffuses from the water column into the soil or at the root-soil interface. Through diffusion and advection flow, wetland plants can transport oxyge n into the root zone creating an oxygenated rhizosphere where nitrification can occur (Reddy et al. 1989). This NO3 is then removed by plant uptake or by microbial communities via den itrification. Denitrification is an anaerobic microbially driven process by which NO3 is reduced to nitrous oxide (N2O) or N2 gas which is then released to the atmosphere. The N availability in the soil can affect the productivity of vegetati on. Plants growing in high N soils tend to have a higher tissue N concen tration and higher photos ynthetic rates than do plants growing in low N soils (Chapin et al. 2002). Nutrient (N or P) limitation is defined as the

PAGE 23

23 increase in growth in response to the addition of the limiting nutrient (Chapin et al. 1986). However, it should be noted that plant communities adapted to nutrient poor ecosystems generally have a lower growth response to nutri ent addition than to plant communities adapted to nutrient rich soils. This differe nce in response has been contri buted to the maximum potential growth rate of the plant species adapted to the soil nutrient conditions (Gri me 1977, Chapin et al. 1986). In wetland ecosystems, studies have sh own that productivity has increased with N additions but at the cost of sh ifting the plant community from a species-rich to a species-poor community (Fojt and Harding 1995, El-Kahloun et al. 2003). This suggests that while nutrient enrichment may increase productivity it may be at the cost of decreased plant diversity. The response of vegetation to different levels of nutrient availability is often evaluated by considering their nutrient-use efficiency (NUE). In short-lived plants (annuals), the NUE has been defined as the organic matter produced per unit of nutrient taken up or more simply the inverse of nutrient concentrati on in plant tissue (Chapin 1980). Fo r long-lived perennial plants, however, it has been argued that the NUE cannot be taken simply as the i nverse of plant nutrient concentration (Vitousek 1982). The NUE for perenni als is thus defined as the amount (in grams) of organic matter lost from plants or permanently stored within plants per unit (in grams) of nutrient lost or permanently stor ed (Vitousek 1982, Birk and Vitous ek 1986). In other words, the NUE is the ratio between above-ground bioma ss production and nutrient loss in litterfall. Berendse and Aerts (1987), however, have suggested that the aforementioned NUE definitions are inappropriate for assessing the ef ficiency of N for dry matter production at the species level. They suggest that the NUE-N (NUE of nitrogen) of individual plants should include the mean residence time of the N in the plant as well as the rate of carbon fixation per unit of N in the plant (N produc tivity). The mean residence time of N is defined by 1/Ln, where

PAGE 24

24 Ln is the N requirement per unit of N in the plant (g N g-1 N yr-1). The N requirement is the amount of N that is needed to maintain each un it of biomass during a given time period (g N g-1 dry weight yr-1). The N productivity (A) is defined as dry matter production per unit of N in the plant. They suggest that N productivity is impor tant in terms of NUE because the amount of N in the leaves of plants is one of the primary properties that determine the rate of photosynthesis. By combining the concepts of mean residen ce time and N productivity, the NUE at the species level is the product of the two, A/Ln (Berendse and Aerts 1987). Yet another term often used along with NUE is that of nutrient reso rption from senescing leaves. The nutrient-resorption efficiency (NRE) is defined as the ratio of the amount of nutrients resorbed from mature leaves to th e maximum nutrient pool in the mature leaves expressed as a percent (Aerts et al. 1999). The NRE of N (NRE -N) from senescing leaves is typically around 40-50%, but NRE-Ns as low as 0% and as high as 90% have been reported (Aerts 1996, Aerts et al. 1999, Chapin et al. 2002). It has been suggested that the large variation in NRE is in response to nutri ent availability status (Aerts 1996, Killingbeck 1996, Aerts et al. 1999, Richardson et al. 1999), in other words, in nutrient limited systems the NRE of senescing leaves will be greater than in nutrient rich envi ronments. However, reviews of the literature indicate that no such trends have been successfully supported (Chapin 1980, Aerts 1996). At the ecosystem level, NRE could have sign ificant implications in terms of nutrient cycling. To decrease dependen ce on soil nutrient availability and nutrient uptake, plants will resorb nutrients during senescence so that they are readily available for future plant growth. Past studies have suggested that efficient retranslocation or low losses of nutrients can increase the fitness of plant species in nutrien t limited ecosystems (Grime 1977, Berendse 1994, Richardson et al. 1999). In addition, hi gh NRE and NUE of vegetation can limit the

PAGE 25

25 remineralization (decomposition) of nutrients in the ecosystem due to low litter nutrient content (poor litter quality) (Bridgham et al. 1995). Decomposition of plant material is a key process for nutrient cycling within ecosystems. There are several factors that control the rates of decomposition. Mean annual temperature, precipitation, soil moisture, micr obial biomass, chemical compos ition of the soil (i.e., available N, P, and C) and litter material (i.e., lignin, tannin, amino aci ds, carbohydrates, C:N:P ratios) are some of the most important factors in consid ering decomposition rates (Aerts and deCaluwe 1997, Gartner and Cardon 2004). Nitrogen or P concentrations can limit the rates of decomposition depending on which one is a limiting resource within the ecosystem (Aerts and deCaluwe 1997, Feller et al. 2003). Feller et al. (2002) found in sites that were N-rich and Plimited that with P fertilization the NUE-P decreased and decomposition rates increased by almost 40%. This suggests that by increasing a lim iting nutrient in the soil, the litter quality can be altered and in turn increase decomposition rates and nutrient turnover. By investigating N cycling (mineralization, ni trification, and vegetativ e uptake), nutrientuse efficiency, decomposition and the diversity and abundance of vegetation, inferences on mediators of vegetation communities or the influence of plant diversity on ecosystem function can be made. Controls on nutrient availability and turnover can influence to nutrient dynamics of an ecosystem and limit the amount of nutrients available for plant uptake. In addition, the ability of some plant species to capture and e fficiently use these nutrients can increase their productivity and fitness allowing them to out-compete less effici ent plant species. Problem Statement History Invasive exo tic plants are a threat to Florid as natural areas. The problems associated with foreign aquatic invasions are in fringing on both disturbed and pristi ne ecosystems. The Exotic

PAGE 26

26 Pest Plant Council (EPPC) has identified Schinus terebinthifolius Raddi (Brazilian pepper or Florida holly) as one of Floridas most invasive species. Schinus terebinthifolius is native to Brazil, Argentina, and Paraguay and was first intr oduced into the United States in the 1840s as an ornamental. This evergreen dioecious tree be longs to the Anacardiaceae family and is related to poisonwood, poison oak, poison ivy, mango, and pistachio, etc. Schinus terebinthifolius produces dense clusters of small ( 1.5 mm) white flowers usually in spring. Their fruit is a cluster of small berries (6 mm diameter) which change fro m green to bright red as they ripen, hence the misnomer "Florida holly." These berries can have narcotic or toxic effects when eaten by birds and other wildlife (Clark 1997). The habit of S. terebinthifolius is a small tree (typically to 10 feet height but can reach 40 feet in height) and it is abundant in disturbed mo ist to mesic sites in th e southern half of the Florida peninsula. It forms dense thickets which exclude native vegetation by shading and chemical inhibition of their growth, and provide relatively poor wildlife habitat. Trees are moderately salt tolerant, withst and flooding, fire, drought, and quickly re-sprout after being cut. The root system is not considered invasive. Schinus terebinthifolius is a pioneer of disturbed site s in Florida ranging from highways, canals, fallow fields, and drained cypress stands (Ewel et al. 1982). Furthermore it successfully colonizes many native plant communities, including pine flatwoods, tropical hardwood hammocks, and mangrove forests (Ewel et al. 1982). In addition to its threat to Floridas natural ecosystems, it poses some potential health threat s. Being a relative of poison ivy, direct skin contact with its sap can cause severe skin irrita tion, airborne chemical emissions can result in sinus and nasal congestion and headaches. Cons umption by horses and cattle has resulted in hemorrhaging, intestinal compaction, and fatal colic (Clark 1997).

PAGE 27

27 In spite of S. terebinthifolius invasive and hazardous environmental qualities, it has several economic uses as well. As its common name (B razilian pepper) suggests, the dried fruits are used as a spice and sold in the United States as pink peppercorn. In areas of South America where it is native, the plant is used as a tonic and astringent. In Brazil it is considered medicinal and used in remedies to treat ulcers, respir atory problems, wounds, rheumatism, gout, tumors, diarrhea, skin ailments and arthritis. Other products include: toothpick s and a pollen source for honey bees (Clark 1997). The invasion if S. terebinthifolius is recorded in the Everglad es National Park (ENP) in the region called the Hole-in-the-Donut (HID) as ea rly as 1959 (Figure 1-2). Historically, the HID was dominated by short hydroperiod prairies and pinela nds. After 1916, farming practices were employed that altered approximately 4000 ha of natural vegetation. After 1975, farming ended and the HID was aggressively colonized by S. terebinthifolius (Dalrymple et al. 2003). The rapid spread of S. terebinthifolius resulted in population growth rates with increases of 20 times its density per year (Loope and Dunevitz 1981). Th is dense canopy in mature stands limits the ability of understory vegetation to exist or for desired indigenous vegeta tion to compete (Doren and Whiteaker 1990). Attempts to control S. terebinthifolius through use of prescribed fire, mechanical removal in conjunction with native plantings and chemical treatment all proved un successful. The dense canopy and lack of understory litter material made prescribed fire difficult and the high germination rate, high survival rate of seedli ngs, and rapid growth made chemical control difficult and costly (Doren and Whiteaker 1990, Dalr ymple et al. 2003). These failures led to the use of mitigation funds from Miami-Dade Count y to employ a scraping method to restore wetlands in the HID. This method invol ves the mechanical removal of existing S.

PAGE 28

28 terebinthifolius and underlying rock-plowed rubble and substrate leaving behind bedrock with pockets of captured substrate material. Thes e pockets provide enough substrate for hydrophytic, herbaceous vegetation to develop on the scraped sites. A pilot project of 18 ha in 1989 proved successful in the natural recolonization of indi genous wetland vegetation and the prevention of S. terebinthifolius invasion (Dalrymple et al. 2003). Th is success led to the long-term wetland restoration project to restore the entire HID by use of the scraping method (Figure 1-3). Yearly restoration sites began in 1997 a nd will continue to completion in 2010. The HID region of the ENP (Figure 1-4) offe rs a unique opportunity to investigate the successional development of vegetation and the relationship between macrophyte diversity and ecosystem functions, including productivity, nutrien t-use efficiency and turnover. The desired dominant species within the HID is Cladium jamaicense (sawgrass). Cladium jamaicense is the dominant species within natural, undisturbed portions of the ENP. It is present within restored sites of the HID but it is not a dom inant species. It is more abunda nt in the earliest restored site (1989) than it is in sites that are only a few years old. In the first few years after restoration begins, sites are dominated by weedy generalists w ho can tolerate harsh conditions. In the most recent survey of the restored si tes, the dominate vegetation are Typha domingensis Andropogon glomeratus and Sagittaria lancifolia and very little C. jamaicense is found (O'Hare and Dalrymple 2003). One important question in the successional development of the HID is what are colonization patterns for C. jamaicense and why is it not dominating the restored sites? Preliminary Work Currently, the Everglades Research Gr oup (ERG) conducts annual monitoring of the vegetation community, soil depth, and hydrology of each of the restored sites, as well as a reference and S. terebinthifolius site (Figure 1-5). For th e vegetation sampling, the BraunBlanquet method is employed to evaluate the overall vegetation community structure in

PAGE 29

29 permanently establish plots every fall at peak bi omass. On each restor ed site there are 20-10 m2 plots (10 randomly located near natural vege tation; 10 randomly located far from natural vegetation) and 40-1 m2 plots (20 nested in the northwest co rner of each of the large plots; 20 randomly located in the intermedia te strata of the site). The large plots were established to address broad characteristics of vegetation assemblages and the small plots are used to evaluate species composition with regard to soil dept h, elevation, and hydrology (O'Hare and Dalrymple 2003). These data were re-evaluated for stat istical relationships between the vegetation community structure, soil depth, elevation, and hydrology as well as with additional soil biogeochemical and vegeta tion properties collected in this study. Project Objectives The goal of my research was to determ ine th e role of vegetation diversity composition on wetland ecosystem function throughout successiona l development after restoration and the influence of nutrient availability on macrophyte nu trient-use efficiency (NUE) of N and P. Anthropogenic impacts, such as farming practices, can drastically change the nutrient dynamics of ecosystems. Wetlands created on abandoned agricultural land are t ypically P-rich and Nlimited. This could in turn impact the diversity and dominance of the macrophytes present. As the system develops, more N can be introduced via fixation and organic matter accretion while the P becomes tied up in the substrate shifting th e dynamics of the system to a P-limited system. In response to the change in nutrient dynamics the vegetation will also change. To assess changes in vegetation diversity and dominance the following studies were investigated in field plots of 0, 1, 2, 4, 8, and 16 years after restoration (Figure 1-6). Study 1. Spatial and Temporal Patterns

PAGE 30

30 Objective 1. Relate the diversity indices of macr ophytes as well as the abundance of the dominant macrophytes and Cladium jamaicense to environmental char acteristics by use of multivariate statistical analysis. Objective 2. Determine productivity for aboveand below-ground macrophytes and storages of carbon, nitrogen, and phosphorus of both the vegetation and the soil. Study 2. Nitrogen Availabil ity and Use Efficiency Objective 1. Determine nitrogen av ailability and long te rm retention using 15N stable isotope techniques. Objective 2. Determine the nutrient-use efficiency (NUE) of nitrogen and phosphorus for community level vegetation and the dominant plant species. Study 3. Organic Matter Turnover and Nitrogen Budget Objective 1. Determine decomposition rates, litter quality, and nitrogen and phosphorous regeneration potential of the dominant vegetation, Schinus terebinthifolius Cladium jamaicense Objective 2. Construct a nitrogen budget utilizing the information obtained from previous studies.

PAGE 31

31 Figure 1-1. Conceptual model of the interlinki ng relationships between ecosystem processes, soil development, and vegetation community as well as primary driving forces that influence vegetative community structure.

PAGE 32

32 Figure 1-2. False image photo of southern Florid a indicating land cover a nd vegetation types. The green area within the black circle is th e location of the restoration project of the Hole-in-the-donut. Courtesy of South Florida Water Management District

PAGE 33

33 Figure 1-3. Delineation of the regi on of the Everglades National Park known as the Hole-in-theDonut to be restored after the e ffects of farming and invasion of Schinus terebinthifolius. Image courtesy of the Everglades National Park.

PAGE 34

34 Figure 1-4. Location of the restored wetlands within the Hole-in-the-Donut in the Everglades National Park. Sites are labeled by the y ear in which they were scraped. Image courtesy of the Everglades National Park.

PAGE 35

35 Figure 1-5. Plot locations of Everglades Resear ch Groups reoccurring vegetation surveys in both the restored wetlands and the surrounding native communities. Image courtesy of the Everglades National Park.

PAGE 36

36 Figure 1-6. Location of each plot included in th is study in the Hole-in-the-Donut region of the Everglades National Park. Image courte sy of the Everglades National Park.

PAGE 37

37 CHAPTER 2 RELATIONSHIPS AMONG TIME, PLANT SPEC IES RIC HNESS, COMPOSITION AND ECOSYSTEM FUNCTION IN RESTOR ED SUBTROPICAL WETLANDS Introduction The loss of more the half of the original wetl ands in the United States has resulted in efforts to restore or recreate previously drained wetlands (W higham 1999). While efforts to restore and mitigate wetlands are well intended, few projects have successfully achieved natural ecosystem status in terms of vegetation st ructure and diversity of undisturbed wetland ecosystems; therefore, loss of wetland ecosyst em function and biodiversity is inevitable. Through restoration and mitigation efforts, the lo ss of wetlands is decreasing, but the success of these efforts in terms of function and diversity is still unclear. The importance of macrophyte diversity in wetland ecosystems and its link to ecosystem function (i.e., productivity, decomposition, and nutrient storages) has receiv ed increasing attention (Bornette et al. 1998, Gessner et al. 2004, Gusewell et al. 2005, Whitehouse and Bayley 2005, Boers et al. 2007). Positive relationships between plant diversity a nd primary productivity have been demonstrated in grasslands (Naeem et al. 1994, Tilman and Downing 1994, Walker 1995, Tilman et al. 1996), but relatively little research of this kind has been done in wetlands. In order to successfully restore ecosystem f unctionality, we need to understand the factors that govern the development of ecosystem function and the establishment of plant community structure. Several studies have investigated th e relationships between pl ant community structure and ecosystem functions such as: soil organic matter accumulation (Craft and Richardson 1993, Vymazal and Richardson 1995, Callaway et al. 2 003), soil nutrient pools (Ehrenfeld et al. 2001, Bengtsson et al. 2003, Zak et al. 2003, Fitter et al. 2005), biomass production (Catovsky et al. 2002, Pauli et al. 2002, Callaway et al. 2003), an d decomposition (Hector et al. 2000, Catovsky et al. 2002, Gartner and Cardon 2004).

PAGE 38

38 There are three common ways of measuring di versity over spatial sc ales: alpha, beta, and gamma diversity. Alpha diversity refers to the diversity within a particular area or ecosystem and is usually expressed as the average number of species (species richness) present in the sample units (i.e., plots) for a given ecosystem (McCune and Grace 2002). Beta diversity is a measure of the change in species diversity be tween two ecosystems or sample units, whereas gamma diversity is the total species pool with in an ecosystem or region (McCune and Grace 2002). Alpha diversity (or species richness) has been used extensively as a measure of species diversity (Tilman et al. 1996, Tilman et al. 1997, Bornette et al. 1998, Pollock et al. 1998, Seabloom and van der Valk 2003, Zak et al. 2003). Beta diversity (or dissimilarity in diversity) has been shown to be more important at largescales (i.e., biomes, geographical regions with latitudinal gradients). Studies have found that beta diversity and productivity are positively related on a temperate-tropical ecosystem gradient (Francis and Currie 2003, Hawkins et al. 2003), whereas on a local or even regional scale this relationshi p has been weak or unimodal (Grime 1973, Tilman 1982, Grace 1999, Waide et al. 1999). In addition to biodiversity, it is necessary to understand the importance of species composition as it relates to ecosystem func tion. The functional im portance of species composition can be examined by determining the re lative occurrences of i ndividual plant species which provides information on species abundanc e (Locky and Bayley 2006). Furthermore, differences in species abundance often allude to environmental conditions controlling species composition. In subtropical Florida we tlands, for example, the dominance of Cladium jamaicense is associated with nutrient-poor en vironments, whereas, the dominance of Typha domingensis is found under nutrient-rich conditio ns (Koch and Reddy 1992, Craft et al. 1995,

PAGE 39

39 Newman et al. 1996, Craft and Richardson 1997, Doren et al. 1997). Therefore, it can be useful to understand the desired habitat/environmental conditions of individual species to predict controls on species composition. In this study, we examined patterns of species richness and com position, productivity and nutrient accumulation in undisturbe d and restored subtropical wetland ecosystems that varied in time since restoration from 1 to 16 years. Our objectives were to1) determ ine patterns of plant species diversity and dominance in the native and restored wetland communities, 2) determine biomass production at both species and community levels, 3) quantify car bon (C), nitrogen (N), and phosphorus (P) accumulation in the biomass at the species and community level, and 4) relate species diversity indices to ecosystem function (i.e., biomass production, C, N, and P accumulation). Methods Site Description This study was conducted in wetland system s restored in the Hole-in-the-Donut (HID) region of the Everglades National Park (ENP) by Miami-Dade County mitigation funds. Past farming and management practices in the areas th at were restored left these systems open to invasion by Schinus terebinthifolius (Brazilian pepper). The nut rient enriched soil, higher elevation (resulting in short hydroperiods) and su btropical conditions of Florida made these disturbed areas an ideal location for invasion by S. terebinthifolius The natural surrounding marl prairie wetlands are inundated for approxima tely six months of the summer season. The goal of the restoration of the HID was to remove the enriched soil and lower the elevation to increase the hydroperiod to control S. terebinthifolius re-invasion (see Chapter 1 for a more detailed site description).

PAGE 40

40 Currently, the Everglades Research Gr oup (ERG) conducts annual monitoring of the vegetation community, soil depth, and hydrology of each of the restored sites, as well as a native community. For the vegetation sampling, they em ploy the Braun-Blanquet method to evaluate the overall vegetation community structure in pe rmanently establish plots every fall at peak biomass. On each restored site there are 20-10 m2 plots and 40-1 m2 plots (20 nested in the northwest corner of each of the large plots; 20 ra ndomly located in the intermediate strata of the site). The large plots were establish to address the broad charact eristics of vegetation assemblages and the small plots are used to evaluate species composition with regard to soil depth, elevation, and hydrology. This data will be re-evaluated for statistical relationships between the vegetation communities, soil depth, elevation, and hydrology to determine how well the current data explains the re-v egetation of the restored sites. From the intermediate strata, 10 of the pre-established small plots (1 m2) in the restored (1, 2, 4, 8, 16 years), native sites were utilized for characterization of vegetation and soil physical and chemical properties. Species Diversity Indices Micro-topography variability is high in these we tland system s. In an attempt to eliminate topography and hydrology differences as variables controlling ve getation patterns, we chose 10 plots in each site at approximately 0.5 m el evation for sites restored in 2004, 2003, 2001, 1997, and 1989 and the native community. The species composition data generated was used to calculate alpha, beta and gamma diversity. Alpha diversity wa s determined as the average number as species found in each plot at each si te. Gamma diversity was determined for the HID as a whole region (total species richness combined ) and for each individual site (total species richness in each site). This allowed us to calculate beta diversity in two ways, between and within sites. Beta diversity ( ) was calculated using Whittakers measure,

PAGE 41

41 = / (2-1) where is the gamma diversity of th e HID or individual sites and is alpha diversity for the average species richness observed in all the plots from each site. To determine how similar or dissimilar each restored site was from the native community, we calculated Sorensens similarity index (Cs), Cs = 2*J / ( 1 + 2), (2-2) where J is the number of shared species and 1 is the gamma diversity for the native community and 2 is the gamma diversity for the restored site being compared. This index value ranges from 0 where there are no shared species to 1 where the exact species are found in both communities. Above-ground Biomass and Species Dominance Above-ground biom ass was determined in Ap ril 2005 (dry season; Y1), July 2005 (wet season; Y2) and July 2006 (wet season; Y3). Plant shoots within 1 m2 plots adjacent to the preestablished plots will be clipped at the sediment surface. The plant matter was dried and weighed at 65C to a constant mass with results expressed as g dry weight m-2. For the July 2005 and 2006 biomass sampling, each m2 plot was divided by species contribution to biomass. The species collected in each biomass sampling were grouped based on abundance and importance to the restoration goal and ranked on frequency and percent cover. Because we were primarily interested in the co ntributions of dominant species to the overall biomass at each site, we grouped infrequent specie s (contributed less than 1% to the total above ground biomass) together in the category Othe r. Species that were most abundant were grouped by themselves. Several Poaceae species (grasses) were observed, however, unidentifiable. Members of the Poaceae fam ily, with the exception of Andropogon spp., found in Y3 (but not in Y2) were grouped into one cate gory called Poaceae and counted as one species.

PAGE 42

42 Relative frequency of individual species was de termined as a percent of the total biomass collected for each site This was done for the HID as a whole and for each site individually to determine overall dominance and site dominance. Both species and community level vegetation samples were analyzed for C, N, and P pool sizes. Total C and total N were determined by dry combustion with a Thermo Electron Corporation Flash EA NC Soil Analyzer for the bulk aboveground plant tissue. Values for species and community level vegetation were reported as a concentration (mg g-1) or pool size (mg m-2). Results Species Composition and Diversity The com plete spatial survey of the vege tation community conducted by the Everglades Research Group found 112 different plant species (gamma divers ity) present within the HID (Table 2-1and 2-2). They found 62 species in the native community and in the 1989, 1997, 2001, 2003, and 2004 restored wetlands they found 48, 51, 38, 40, and 38 species, respectively (Table 2-1 and 2-2). The alpha diversity (or species richness) was th e highest in the native community at an average of 18 species per plot followed by the 1989 and 1997 at 17, 2001 at 11, 2003 at 12 and the 2004 at 9 (Table 2-2). The beta diversity for the HID as a whole system (or region) with the gamma diversity of 112, was 6 for the native site with decreasing numbers as sites increased in age since restoration (Table 2-2). Beta diversity calculated with site-specific estimates of gamma showed that the 2004 site was the most variable with a beta diversity of 7 (Table 2-2). All other sites had similar beta dive rsities of 3 or 4. Sorensens similarity index ranged from 0.24 to 0.46, indicating that the oldest site (1989) was the most similar to the native community and the 2003 site was th e least similar (Table 2-2). Year 1 (Y1) biomass production resulted in the greatest bioma ss production in the native and 1989 sites as compared to the biomass production in year 2 (Y2) and y ear 3 (Y3) within the

PAGE 43

43 same sites as well as across all sites (p=0.0005; Figure 2-1 and Table 23). Biomass production in Y3 was highest in the 1997 and 2003 sites as co mpared to Y1 and Y2. There was a significant decrease in biomass production with time in the 1989 sites (p<0.0001) whereas in the 2003 site, the Y3 biomass resulted in a significant increase (p=0.0077; Fi gure 2-1 and Table 2-3). No significant change occurred in the 2001 and 2004 between each sampling (0.7796 and 0.7233, respectively, Table 2-3). In Y1, the biom ass production in the native and 1989 sites were significantly higher than the biomass produc tion in the 2001, 2003, and 2004 sites, but the 1989 site was not significantly different from the 1997 site (Table 2-3 and Figure 2-1). The same was true for Y2. During Y3, the biomass production in the native community was significantly greater than all other sites, the 1989 site was significantly greater than the 1997 and 2003 and the 1997 and 2003 were the same (p<0.0001; Table 2-3 and Figure 2-1). Figure 2-2 represents the biomass contributi on of each species group identified in each biomass sampling for each site (Table 2-1; sp ecies indicated by ). The amount of biomass contribution between species in th e native site did not change signi ficantly between Y2 and Y3. In the 1989 site, the amount of C. jamaicense did not change; the decrease in biomass production in Y3 was a result of less Other production. The significant increase in Y3 for both the 1997 and 2003 sites was primarily a result of an increase in T. domingensis In addition to total above-ground biomass, the relative frequency of each species was determined for the HID as a whole and for each site individually. The most frequent species occurring in the HID was T. domingensis followed closely by C. jamaicense. Schoenus nigricans and Poaceae were also frequent (Figure 2-3a). In the native community C. jamaicense and S. nigricans were co-dominant (Figure 2-3b). In the 1989 site, C. jamaicense and Poaceae are the two most frequently occurring species (Figure 2-3c ). The 1997 and 2003 sites were

PAGE 44

44 mostly dominated by T. domingensis ; however, in the 1997 site Andropogon spp. were also abundant and in the 2003 site S. sempervirens was abundant (Figures 2-3d and e). Nutrient Concentrations and Pools Community level carbon and nitrogen concen trations varied from 382-441 and 6.3.8-10.6.1 mg g-1 across sites, respectively (Figure 24a and b). The community level carbon content was the lowest in the 2001 site and the highest in the native site, whereas the nitrogen content was the lowest in the 2003 site and the highest in the 2001 site. The community level phosphorus content was the lowest in the native plant communities at 0.18.04 mg g-1 and gradually increase from oldest to yo ungest with the 2004 site at 0.6.09 mg g-1 (Figure 2-4c). Individual species tissue nutri ent content varied both betw een species and among sites. Cladium jamaicense had the lowest C content in the 1989 site, the lowest N content in the 1997 site and the lowest P content in the native community (Table 2-4), whereas it had the highest C, N and P content in the 2004, 1989, and 2004 sites, respectively. Schoenus nigricans was only found in the native community. As compared to C. jamaicense (its co-dominate species), S. nigricans had higher C content and lower N and P content (Table 2-4). Typha domingensis was not found in the native or 1989 sites, but among rest ored sites, it had the lowest C, N, and P in the 2004, 2004, and 1997 sites, respectivel y (Table 2-4). When present, S. lancifolia had the highest N and P content as compared to all othe r species except in the 2003 (Table 2-4) where J. megacephalus had the highest N and Andropogon spp. had the highest P content. Community level differences in C, N and P pools were a reflection of differences in species level contribu tions (Figure 2-5). Schoenus nigricans stored the most C and N of all species investigated (Figure 2-5d). Phosphorus storage was the highest in the group of Other species and in T. domingensis and Andropogon spp when present (Figure 2-5f). Sagittaria lancifolia had the lowest C, N and P storage prim arily as a result of having a low biomass

PAGE 45

45 contribution (Figure 2-5d, e, and f). Community weighted C:N ratios revealed that the vegetation community present in the 2003 site wa s significantly higher th an the ratios of the vegetation found in all other site s (Figure 2-6a). The community weighted N:P ratios of the vegetation in the 2003 and 2004 sites we re significantly lower than th e ratios in the vegetation in the native, 1989, 1997, and 2001 sites (Figure 2-6b). The community weighted C:P ratios of the vegetation gradually increased with age with the 2004 site having the lowest and the native vegetation with the highest; howev er, no significant differences we re observed (Figure 2-6c). Diversity and Biomass Relationships Richness was weakly po sitively related to aboveground biomass production across the sites (r2=0.56, d.f.=59, F=73.8, p<0.0001, n=60), with the 2004 site having the lowest above-ground biomass production and species richness and the native community having the highest (Figure 27). The age of each site contributed to the differences found in species richness (r2 = 0.40, d.f=49, F=28.5, p<0.001; Figure 2-8a), whereas the age did not contribute to the trend in biomass production found across restored wetland communities (r2 = 0.16, d.f.=49, F=8.9, p=0.0040; Figure 2-8b). Gamma and -diversities and biomass production were highly positively related (r2 = 0.87, d.f.=5, F=26.4, p=0.0070 and 0.93, d.f.=5, F=53.3, p=0.0020, respectively; Figure 2-9a and b). A negative relationship was observed between the (within) beta diversit y and biomass production (r2 = -0.79, d.f.=5, F=15.3, p=0.0200; Figure 2-9c). Discussion The Hole-in-the-Donut (HID) region of the Everglades National Park (E NP) offers a unique opportunity to study temporal patterns of relationships between species diversity and ecosystem structure and function by examining a chronosequence of restored wetland communities. Several diversity and ecosystem function experiments have been conducted under

PAGE 46

46 controlled environments where the species diversity levels were planted and maintained (Tilman and Downing 1994, Tilman et al. 1996, Tilman et al. 1997, Baldwin et al. 2001, Kellogg and Bridgham 2002, Zak et al. 2003). In order to improve restor ation efforts and management of large scale systems, it is important to understand how natural recruitment of plant species will develop and potentially impact ecosystem function. All communities experience some level of di sturbance which alters plant communities via removal or additions of indivi dual species (Sti ling 1999). In some cases, severe levels of disturbance (induced by either natu ral disasters or anthropogenic impacts) occur which result in large scale changes in ecosystem function and pl ant community structure which can include the invasion of non-native plan t species. The loss of more the half of the original wetlands in the United States has resulted in efforts to restore and recreate previously drained wetlands (Mitsch and Gosselink 2000). While efforts to restore and mitigate wetlands are well intended, few projects have been successful in achieving na tural ecosystem status in terms of vegetation composition of undisturbed wetland ecosystems. Due to the high level of disturbance and destruction of natural wetlands, the loss of wetland plant diversity is inevitable and in turn a loss of function could occur. The results of this study offer insights on the factors controlling and maintaining species composition in wetland ecosy stems beginning at primary succession. Diversity and Species Composition Patterns Community developm ent depends on many conf ounding factors ranging from soil nutrient availability, hydrology, and seed dispersal and recruitment just to name a few (van der Valk 1981). In this study, we found that th e lack of soil did not inhibit plant recolonization in restored wetland ecosystems. Within six months of comp lete soil removal, the 2 004 cleared site (age, 1 year) had a gamma diversity of 38 plant species and a -diversity of 9 species per plot. After

PAGE 47

47 this initial colonization of species, the introducti on of additional species a ppeared to be slow. The site restored in 2001 (age, 4 years) resulted in a -diversity increase by 2 species and the gamma diversity remained unchanged. However, by 16 years after rest oration (1989 site) the diversity per plot almost doubled to 17 species and the gamma dive rsity reached 48 plant species. While this -diversity is similar to the native co mmunities, the gamma diversity is still significantly lower by 14 species. While the gamma diversity may be lower in the 1989 site, the species composition is similar to the native plant community. This restored wetland had equal contributions of C. jamaicense to the above-ground biomass as did the native community. Additionally, T. domingensis was not present in any of the biomass samplings even though it has been found in the 1989 site in previous year s (O'Hare and Dalrymple 2003). The most notable difference between the native plant communities and the restor ed wetlands is the lack of colonization by S. nigricans in the restored plant communities. Schoenus nigricans is a dominant plant species in the surrounding native communities; however it has not been recorded as present in any of the restored wetland communities since this restor ation project began in 1989 (Dalrymple et al. 2003, O'Hare and Dalrymple 2003). Additional research is needed to understand the colonization patterns and environmental conditions required for this plant species to understand why it is not colonizing the restored wetland communities. The species diversity indices may be similar dur ing the first few years after restoration, but the species composition differs considerably. Mixed grasses contribute 50% or more to the biomass production of the 2004 and 2001 restored wetland communities, while the 2003 site is mostly dominated by T. domingensis Additionally, T. domingensis is the dominant species found in the 1997 site (age, 8). There are some concerns on the dominance of T. domingensis in

PAGE 48

48 these earlier restored wetland communities. While T. domingensis is native to the state of Florida, it is not native to this region of the ENP (Craft et al. 1 995, Miao et al. 2000). It has been suggested that T. domingensis has allelopathic properties whic h inhibit the seed survival and propagation of other plant species (Prindle a nd Martin 1996, Domenech et al. 1997), however this has not been adequately conf irmed with scientific research. More important is the invasion of T. domingensis into areas where the nutrient availability has been altered or enriched. Numerous studies have been conducted which have shown that Typha spp. will invade into nutrient enriched areas (Craft et al. 1995, Gophen 2000, Miao et al. 2000, Woo and Zedler 2002, Johnson and Rejmankova 2005). The restored wetla nd systems in the HID are P-rich compared to the native communities as a result of both pr evious farming practices and the destructive restoration technique. Typha domingensis has been shown to create a monoculture under such conditions (Koch and Reddy 1992, Craft et al. 1995, Doren et al. 1997, Daoust and Childers 1999, Bruland et al. 2006). While this has not occurred during the time this study was conducted, the 2004 restored wetland community has since become a T. domingensis monoculture (unpublished data; Everglades National Park). Ecosystem Function Our results indicate that the res tored wetland sy stems with the lowest species richness also had the lowest biomass production. Additionally we found a consistent increase in species richness with age development. Species divers ity predicted 34% of the variability in total biomass production. The age of the sites had a strong influence on specie s richness development explaining 40% of the variability; whereas the age of each site did not explain the variability in above-ground biomass production. These results indica te that an age-driven gradient in species richness may be linked to increases in biomass production. Fu rthermore, this relationship between species diversity and biomass production results in increases in ecosystem function in

PAGE 49

49 terms of C, N, and P accumulation. Other experi mental studies that have controlled species richness levels have found similar relations hips between species diversity and biomass production (Vermeer and Berendse 1983, Clemen t and Maltby 1996, Engelhardt and Ritchie 2001, Callaway et al. 2003, Olde Vent erink et al. 2003). Our study offers further support of this relationship through natural r ecruitment of plant species in wetland communities. In addition to species diversity effects on eco system function, the species composition of a system can have significant effects as we ll (Hooper and Vitousek 1997, Tilman et al. 1997, Engelhardt and Ritchie 2001, Callaway et al. 2003). Historically, the marl prairie wetlands of the ENP have been dominated by C. jamaicense and S. nigricans (Lodge 2005). Numerous studies have been conducted to conclude that C. jamaicense thrives under P-limited conditions and its growth and survival is inhibited when P availability in creases (Davis 1991, Craft et al. 1995, Newman et al. 1996, Craft and Richards on 1997, Richardson et al 1999, Chiang et al. 2000, Noe et al. 2001), unfortunately this same level of attention ha s not been given to S. nigricans In another study conducted in these wetland systems it was determined that C. jamaicense and S. nigricans had high nutrient-use efficiency of P which help to maintain the Plimited environment in which they thrive by retaining more P in standing biomass (see Chapter 4). Typha domingensis however, has a lower nu trient-use efficiency, potentially leading to higher litter quality and more ra pid release of nutrien ts into the soil solution, which could reinforce high P availability under enriched cond itions. All three species use of P could have significant effects on the ecosystem functi on of P availability and storage. More P is stored in the above-ground biomass of T. domingensis which over time would result in more P being stored in both standing biomass and litter making it unavailable for further uptake. This suggests that T. domingensis would facilitate the early development of the restored

PAGE 50

50 wetland communities. As a result, by 16 years after restoration, the wetlands are no longer dominated by T. domingensis but by the native C. jamaicense Earlier vegetation surveys in these wetlands have concluded that T. domingensis was abundant in the 1989 site, however to what extent is unclear since these studies focused primarily on the removal of invasive species Schinus terebinthifolius (Dalrymple et al. 2003, O'Hare and Dalrymple 2003). Further analysis is needed on mechanistic controls of T. domingensis on nutrient cycling to determine the longterm effects of its P use on ecosystem function. Applications to Wetland Restoration and Mitigation It has been suggested that plan ting desired plant species and di versity levels is needed to facilitate vegetation co mposition development in restored wetland communities (Zedler 1993, Kellogg and Bridgham 2002, Callaway et al. 2003). In large-scal e restoration applications, planting desired plant species is not always feasible. Therefore it is necessary to gain a better understanding of the natural recruitment of plant species from primary succession and the development of ecosystem function to increase the success of large-scale rest oration projects. In this study, we found that natural re cruitment would result in incr eases in species richness with time, and that the species composition would develop similarly to the native community provided enough time has past. We saw an immediate recruitment of a diverse plant community consisting of 38 individual plant species within six months of restoration. While additional increases in species diversity we re slow, the diversity did increas e significantly after 8 years. Additionally, it took between 8 to 16 years befo re the plant community developed into one representative of the native community. Howeve r, the ecosystem functi on was not restored in this time period, indicating that more time is needed for developmen t of native ecosystem function.

PAGE 51

51 The species composition changed significantly fr om site to site indicating that the plant communities are very dynamic and unstable. Understanding the seed dispersal, recruitment mechanisms, propagation requirements, and growth and survival rates of the native plant species could aid in the success of restor ation efforts. If a desired native plant species has a limited seed dispersal mechanism, it could take several years before that sp ecies colonizes or dominates a restored wetland. For example, the seed dispersal mechanism of T. domingensis is relatively fast compared to C. jamaicense (van der Valk and Rosburg 1997) which could contribute to the slower colonization patterns observed for C. jamaicense in this study. The results of this study offers evidence that with time, a diverse plant community similar to a native wetland community can develop without human intervention. However, more time is clearly needed to restore ecosystem function to th e level of the native system. The key here is time. Unfortunately, wetland mitigation laws require that wetlands created or restored that serve as mitigation projects are only required 5-10 years of monitoring (Clean Water Act, Section 404). This study along with many others provide s ample evidence that this monitoring time period is may not be long enough to restore and maintain plant community structure or ecosystem function (Whigham 1999, Brinson and Malvarez 2002, Kellogg and Bridgham 2002, Callaway et al. 2003, Dalrymple et al. 2003, Seabloom and van der Valk 2003, Polley et al. 2005). Conclusions The links between divers ity and function dur ing the successional development of the wetlands in this study has implications to the management of restored ecosystems. Landscape and watershed alterations can result in severe degradation of wetland systems which result in species compositional changes and loss of biodiversity. Wetland systems are driven

PAGE 52

52 predominantly by hydrology and many plant species will respond differently to fluctuations or changes in water level and flow. To maximize species diversity and compos ition development similar to a native (or reference) system, it is important to understand the factors governi ng the native system. For this restoration project, the native vegetation community is P-limited with extremely low levels of N and P. To enhance the potential for native plant species in colonize, the nutrient rich soil was complete removed to eliminate the effects from previous farming practices. This restoration method was destructive and labor intensive, the result is the de velopment of herbaceous wetland plant communities that, with time, has been colonized with some native species; however, overall the vegetation community st ructure is very different. Furt hermore, more time is needed for development of ecosystem functions simila r to those found in the native communites.

PAGE 53

53 Table 2-1. Complete species list for all plants identified in th e HID during 2005. An X indicates presence in site location. All species are listed in alphabetic al order. (n=60) Species Name Native19891997200120032004 Agalinis fasciculata X Agalinis linifolia X Alestris lutea X Alternanthera philoxeroides X Ammania coccinea X Ammania latifolia XXXX Ampelopsis arborea X Andropogon glomeratus XXXXX Andropogon virginicus XX Aristida purpurascens X Aster bracei XX X Aster subulatus XXXX Axonopus furcatus X Baccharis angustifolia X Baccharis glomeruliflora XXXXX Baccharis halimifolia X Bacopa monnieri XXXXX Cassytha filiformis X Centella asiatica XXXXX Chara light X Chara unidentified species XXXXX Cirsium horridulum X Cladium jamaicense XXXX Coelorachis rugosa X Conoclinium coelestinum XXXXX Cyperus elegans X Cyperus haspan XXXX Cyperus ochraceus X Cyperus odoratus X Cyperus polystachyos XXXX Cyperus surinamensis XX Dichanthelium dichotomum XXXX Diodia virginiana XXXXXX Dyschoriste angusta X Eclipta prostrata X Eleocharis geniculata XXXX Elytraria caroliniensis X Eragrostis elliottii XXXX

PAGE 54

54 Table 2-1. Continued. Species Name Native19891997200120032004 Erigeron quercifolius X Eupatorium capillifolium X Eupatorium leptophyllum XX XX Eustachys glauca XXXX Evolvulus sericeus X Fuirena breviseta XXXXX Heliotropium polyphyllum X Hydrocotyle umbellata XXXX Hymenocallis palmeri X Hypericum hypericoides XX Hyptis alata XXXX Hyptis spicigera X Ilex cassine X Ipomoea sagittata X Ipomoea triloba X Iva microcephala XX Juncus megacephalus XXXX Kosteletzkya virginica X Leptochloa fascicularis XXX Lobelia glandulosa X Ludwigia alata XX X Ludwigia microcarpa XXXXXX Ludwigia octovalvis XXXXX Ludwigia peruviana XX X Ludwigia repens XXX Lythrum alatum XXXX Mecardonia acuminata X Melanthera nivea X Mikania scandens XXXXXX Mitreola petiolata XXXXX Muhlenbergia capillaris XXX Oxypolis filiformis X Panicum dichotomiflorum XX Panicum hians X Panicum rigidulum XXX Panicum tenerum X Paspalum urvillei X Periphyton XXXXXX Persea palustris X Phyla nodiflora XXX

PAGE 55

55 Table 2-1. Continued. Species Name Native19891997200120032004 Phyllanthus caroliniensis XXXX Pluchea odorata X Pluchea rosea XXXXX Polygala balduinii X Polygala balduinii X Polygala grandiflora X Polygonum hydropiperoides X Polygonum punctatum X Proserpinaca palustris XXXX Rapanea punctata X Rhynchospora colorata XXXXX Rhynchospora divergens X Rhynchospora microcarpa XXXX Rhynchospora odorata XXX Rhynchospora tracyi X Saccharum giganteum XXX Sagittaria lancifolia XXXXX Salix caroliniana XXXX Sarcostemma clausum XX Schizachyrium scoparium X Schoenus nigricans X Scleria verticillata XX Sesbania herbacea XXX Setaria parviflora XXXXXX Solidago sempervirens XX X Spermacoce floridana XX Spermacoce prostrata X Spermacoce terminalis X Stemodia durantifolia X Typha domingensis XXXXX Verbena scabra X Vernonia blodgettii XX Vigna luteola XX Vitis shuttleworthii X Waltheria indica X Gamma Diversity624851384038 Total = 112 Represents species identified in biomass production sampling plots

PAGE 56

56 Table 2-2. Diversity indi ces calculated from complete species list (Table 2-1) for individual sites and the HID as a whole. Site gamma gamma CsCAveSDAveSD AveSD Native184.561.511240.862 1989174.172.0 30.8480.46 1997174.271.9 30.8510.36 2001113.0113.6 41.1380.44 2003121.991.6 30.5400.24 200495.21915.9 74.7380.40ABased on Whittaker's measur e for beta-diversity for the HID as a whole region.BBased on Whittaker's measure fo r beta-diversity with region defined as individual sites.CBased on Sorensen's similarity coefficient for be ta-diversity comparing the native site to each individual resored site. betaBSites alphabetaAHID

PAGE 57

57 Table 2-3. Summary of results from ANOVA tests with dependant variables of as either each site or year of biomass sampling. Site or y ear with shared lower case letters are not significantly different at p<0.05 base d on Tukeys multiple comparisons. Source of Variationd.f.prob. > F 1Year Native19891997200120032004 Y1 5<0.0001aababcbcc Y2 5<0.0001aababbcbcc Y3 3<0.0001abc c 2Sited.f.prob. > FY1Y2Y3 Native20.0005abb 19892<0.0001abb 199720.4224n.s.n.s.n.s. 200120.7796n.s.n.s.n.s. 200320.0077abba 200420.7233n.sn.sn.s n.s. = not significant

PAGE 58

58 Table 2-4. Summary of species carbon, nitrogen, and phosphorus content, pool sizes, and pool weighted ratios. SpeciesSiteAveSEAveSEAveSEAveSEAveSEAveSEAveSEAveSEAveSE C. jamaicenseNative447(50)6.2(0.7)0.1(0.01)23.3(2.3)0.35(0.04)7.2(0.9)73(1)4209(169)57(1.8)1989405(68)8.9(1.5)0.3(0.04)40.6(7.2)0.99(0.18)25.7(4.3)62(5)1623(99)41(7.1)1997428(58)5.1(2.4)0.2(0.05)5.5(2.3)0.07(0.33)2.8(0.8)84(3)1955(76)23(1.5)2001449(74)6.2(1.6)0.1(0.01)31.9(7.9)0.44(0.47)9.5(2.5)73(14)3366(169)46(7.1)2004453(113)6.8(1.7)0.3(0.07)6.8(2.5)0.08(0.03)3.3(1.1)70(4)1669(120)24(0.7)S. nigricansNative461(58)5.3(0.7)0.1(0.01)112.6(3.1)1.31(0.05)19.6(0.6)87(1)5796(80)67(1.5)T. domingensis1997441(74)5.6(0.9)0.3(0.05)54.4(8.5)0.68(0.11)31.7(6.2)81(3)1908(167)23(1.5)2001443(111)8.6(2.1)0.4(0.10)14.6(4.3)0.22(0.04)9.4(1.4)56(5)1298(182)22(1.1)2003456(76)5.6(0.9)0.4(0.07)43.1(2.4)0.71(0.10)59.4(10.4)75(3)1083(58)14(0.5)2004422(83)5.4(1.2)0.3(0.01)1.2(0.2)0.02(0.07)0.9(0.0)78(1)1297(77)17(0.5)S. lancifoliaNative450(74)19.3(4.3)1.1(0.23)0.9(0.1)0.04(0.02)2.1(0.6)23(2)427(23)18(1.5)1989458(92)12.1(2.4)0.6(0.11)5.0(1.0)0.09(0.01)4.0(0.6)47(5)1205(174)24(1.2)1997415(104)8.4(2.1)0.5(0.13)0.9(0.1)0.02(0.004)1.3(0.3)53(4)878(76)16(0.5)2001395(52)12.6(2.4)0.8(0.12)2.6(0.1)0.08(0.01)5.2(0.9)31(1)501(23)16(0.4)2003466(84)4.3(0.8)0.2(0.03)1.2(0.1)0.01(0.003)0.6(0.03)109(2)2097(23)19(0.4)2004405(27)11.9(1.6)0.7(0.08)1.9(0.1)0.06(0.005)3.2(1.1)34(1)601(55)18(0.4)Andropogon spp.1989470(235)5.8(2.9)0.3(0.14)8.1(1.8)0.10(0.03)5.4(2.4)83(8)2003(555)23(4.6)1997414(83)8.9(1.8)0.6(0.12)37.7(5.9)0.76(0.12)48.4(7.9)54(5)1120(194)18(1.5)2001425(142)6.8(2.3)0.2(0.07)23.6(6.7)0.35(0.10)11.0(3.2)74(14)2323(393)32(0.9)2003437(2)8.5(1.2)0.7(0.06)5.9(2.1)0.10(0.02)8.9(2.5)54(8)627(55)12(0.7)2004439(69)5.7(0.9)0.3(0.06)0.2(0.1)0.003(0.001)0.1(0.03)77(3)1318(77)17(0.5)J. megacephalus2001438(146)6.4(2.1)0.2(0.08)46.3(20.6)0.58(0.25)34.0(17.8)71(5)2241(449)32(5.9)2003430(103)9.8(0.5)0.4(0.05)6.1(1.3)0.14(0.03)5.6(1.2)44(5)1096(77)25(1.0)2004428(86)8.2(1.6)0.4(0.09)10.4(1.1)0.19(0.01)9.3(0.9)56(3)1301(206)22(2.1)OtherNative442(88)6.4(1.3)0.3(0.05)9.6(2.8)0.17(0.06)4.3(1.2)72(3)1827(129)27(2.4)1989445(45)8.9(0.9)0.5(0.05)81.2(10.4)1.31(0.12)55.6(5.3)58(2)3653(782)51(9.1)1997443(44)6.3(0.6)0.4(0.04)45.5(3.0)0.62(0.04)35.8(1.9)72(1)1259(30)18(0.4)2001434(43)7.5(0.7)0.3(0.03)26.7(2.3)0.45(0.04)19.9(1.8)60(1)1697(77)28(1.0)2003456(46)5.6(0.6)0.7(0.07)29.2(2.1)0.38(0.04)40.7(3.0)85(2)744(25)10(0.5)2004423(42)8.2(0.8)0.5(0.05)23.3(1.9)0.43(0.03)24.3(1.9)54(1)982(23)19(0.5)C Content (mg g-1) N Content (mg g-1) P Content (mg g-1) C Pool (g m-2) Pool Wtd N:P N Pool (g m-2) P Pool (mg m-2) Pool Wtd C:NPool Wtd C:P

PAGE 59

59 Site Native19891997200120032004 Above-ground Biomass (g m-2) 0 100 200 300 400 500 600 Y1 Y2 Y3 Figure 2-1. Repeated above-ground bioma ss collection reported as g dry wt m-2 for both the native communities and the restored wetlands. Y1 = dry season 2005, Y2 = wet season 2005, Y3 = wet season 2006, *not included in year 2 sampling analysis

PAGE 60

60 Figure 2-2. Repeated above-ground bioma ss collection reported as g dry wt m-2 and separated by species contribution for both the native comm unities and the restored wetlands. Y2 = wet season 2005, Y3 = wet season 2006, *not included in year 2 sampling analysis. Y1 = not separated into species contribution.

PAGE 61

61 Figure 2-3. Relative frequency of vegetation species collected in the biomass sampling within the restored and native communities of the HID. Species are ordered from left to right as most frequent to least frequent. A) All sites combined in HID. B) Native. C) 1989. D) 1997. E) 2003.

PAGE 62

62 Figure 2-4. Mean tissue content for the commun ity level vegetation for each site in the HID during wet season, Y2. A) Ca rbon; d.f.=5, F=9.7. B) Nitrogen; d.f.=5, F=17.2. C) Phosphorus, d.f.=5, F=2.7. n=60, 10 for each site average.

PAGE 63

63 Figure 2-5. Mean carbon, nitrogen, and phos phorus accumulation in above-ground biomass and relative contribution of plan t species found in the HID. A) Carbon pool; d.f.=5, F=6.8. B) Nitrogen pool; d.f.=5, F=7.7. C) Phosphorus pool; d.f.=5, F=3.2. D) Relative % carbon. E) Relative % nitrogen. F) Relative % phosphorus.

PAGE 64

64 Figure 2-6. Trends in community weighted carbon, nitrogen, and phosphorus ratios between native and restored wetland communities. A) Community weighted C:N ratio; d.f.=5, F=27.1, p<0.0001. B) Community weighted N:P ratios; d.f.=5, F=4.2, p=0.0030. C) Community weighted C:P ratios; d.f.=5, F=1.7, p=0.1500.

PAGE 65

65 Species Richness 0 5 10 15 20 25 Above-ground Biomass (g m-2) 0 100 200 300 400 Native 1989 1997 2001 2003 2004 r =0.56 Figure 2-7. Relationship between above-ground biomass production and species richness for the HID during the wet season, Y2. Additionally, each site has been coded to indicate site differences and relationships. (d.f.=59, F=73.8, p<0.001, N=60)

PAGE 66

66 Figure 2-8. Relationships with the age of each restored site for the restored wetland communities in the HID. A) Species richness, d.f.=49, 28.5, p<0.0001. B) Above-ground biomass production, d.f.=49, F=8.9, p=0.0040. n=50, 10 for each site.

PAGE 67

67 Figure 2-9. Relationships with above-ground biomass production for the sites in the HID. A) Gamma diversity; d.f.=5, F=26.4, p=0.007. B) Species richness (a-diversity), d.f=5, F=53.3, p=0.002. C) Beta diversity (within), d.f.=5, F=15.3, p=0.02. n=60, 10 for each site.

PAGE 68

68 CHAPTER 3 MULTIVARIATE ANALYSIS OF PLANT COMMUNITY STRUCTURE AND RELATIONSHIPS TO ECOSYSTEM CHARACTER ISTICS IN SUBTROPICAL RESTORED WETLANDS Introduction The loss of more the half of the original wetl ands in the United States has resulted in efforts to restore and recreate previously drai ned wetlands (Whigham 1999). While efforts to restore and mitigate wetlands are well intended, few projects have been successful in achieving natural ecosystem status in terms of vegetation composition of undisturbed reference wetland ecosystems. Hydrology, nutrient availability, a nd invasion of exotic plant species are the dominant factors impacting the spatial variati on of wetland plant communities. Several studies have investigated the impact of these factors on vegetative community st ructure in wetlands and the nutrient removal potential of wetland plants (Bornette et al. 1998, Kellogg and Bridgham 2002, Olde Venterink et al. 2002, Murphy et al. 2003, Seabloom and van der Valk 2003, Ehrenfeld 2004). Due to the hi gh level of disturbance and destru ction of natural wetlands, the loss of wetland plant diversity is inevitable and in turn a loss of function could occur. Therefore, it is important to understand the factors contro lling and maintaining species composition in wetland ecosystems. Hydrology is a dominant factor controlling the development of spatial variation in wetland plant community structure and can be the primary control over propagation and recruitment of wetland vegetation. Experiment al studies have shown that variability in hydrological regime can result in different patter ns of species development from the seed bank (van der Valk 1981, Pollock et al. 1998, Baldwin et al. 2001). In a series of field and greenhouse controlled seed bank experiments, Baldwin et al. (2001) found that continuously non-flooded soils had almost twice the species richness a nd 55% greater total stem length than did the

PAGE 69

69 continuously flooded soils. Whereas in studie s of riparian wetlands, the wetlands which experienced greater number of flood events or days of inundation pe r year resulted in a greater species richness (Bornette et al. 1998, Pollock et al 1998, Murphy et al. 2003). Hydrology can drive the nutrien t availability which can also alter the plant species composition of wetlands. River water is often nutrient rich and an increase in flood events can increase the amount of nutrients made availa ble to the vegetation co mmunity in riparian wetlands. In a study of riparian wetlands with high and low connectivity to the river, wetlands with the highest connectivity (inc reased hydroperiod) resulted in greater species richness and an increase in nutrient availability (Bornette et al. 1998). A wetland with a hydroperiod of 3 d y-1 resulted in species richness of 10 and nitrate concentratio ns of 1.15 mg L-1, whereas a wetland with a hydroperiod of 37 d y-1 resulted in species richness of 38 and nitrate concentrations of 10 mg L-1 (Bornette et al. 1998). In this case, the increase in flooding had an indirect positive effect on species richness due to the import of more nutrients This suggests that nutrient availability could also govern plant species com position and structure. Despite the increase in wetland protection, th ey continue to be threatened by nutrient enrichment due to anthropogenic impacts. This nutrient enrichment can be in the form of nitrogen (N), phosphorus (P) or bot h. Agricultural nutrient loadi ng can result in the decrease of plant diversity in wetland ecosystems. The mechan ism responsible for this decrease is generally accepted as increased competition or a heightened productivity in a nutrient loving plant (i.e., cattail) that shades and crowds out other smaller or native plants (Craft et al. 1995, Rivero et al. 2007). An increase in one nutrient can result in a li mitation of another. Some wetland ecosystems can naturally be P limited and increasing the N availability in this system can augment this

PAGE 70

70 problem and vise versa. This ch ange in nutrient availability ca n result in changes in species composition and decreases in species diversity. In a review of fens and bogs, the wetlands that were relatively undisturbed showed that high spec ies diversity was frequently associated with low nutrient status and that sp ecies rich wetlands typically ha ve moderate productivity and standing crop (Bedford et al. 1999 ). However, this relationship has not always been observed. Changes in species composition are not always associated with increa ses in productivity. Several studies have shown that with increase d nutrients, species di versity decreases but productivity increases (Willis and Mitsch 1995, Boyer and Zedler 1998, Mahaney et al. 2004, Rickey and Anderson 2004). Some wetland plants, like Typha spp. or Phragmites spp., will produce tall dense monospecific stands (hi gh above-ground biomass) under nutrient rich environments and in response they out-compete smaller plants for light which decreases species diversity. In this study, we combined vegetation composition analysis with environmental characteristics of the soil and plants in rest ored and native wetland communities to discern relationships among the two. By gaining a clear understanding of what factors govern vegetation community structure we can then evaluate the le vel of restoration success. The objectives were to 1) determine differences and/or similaritie s within the development of plant community structure in each restored site; and 2) relate the vegetation co mmunity structure of these sites to soil and plant properties. Methods Site Description This study was conducted in wetland systems restored in the Hole -in-the-Donut (HID) region of the Everglades National Park (ENP). Past farming and management practices in the areas that were restored left these systems open to invasion by Schinus terebinthifolius (Brazilian

PAGE 71

71 pepper). The nutrient enriched soil, higher elevation (resulti ng in short hydroperiods) and subtropical conditions of Florida made these disturbed areas an ideal location for invasion by S. terebinthifolius. The natural surrounding marl prairie wetlands are inundated for approximately six months of the summer season. The goal of the restoration of the HID was to remove the enriched soil and lower the elevati on to increase the hydroperiod to control S. terebinthifolius reinvasion (see Chapter 1 for a more detailed site description). Currently, the Everglades Research Gr oup (ERG) conducts annual monitoring of the vegetation community, soil depth, and hydrology of each of the restored sites, as well as a native community. For the vegetation sampling, the Br aun-Blanquet method was employed to evaluate the overall vegetation community structure in perm anently established plots every fall at peak biomass. On each restored site there are 20-10 m2 plots and 40-1 m2 plots (20 nested in the northwest corner of each of the large plots; 20 ra ndomly located in the inte rmediate strata of the site). The large plots were establish to address the broad characteristics of vegetation assemblages and the small plots were used to ev aluate species composition with regard to soil depth, elevation, and hydrology. This data will al so be used in conjunction with additional data collected in this study. From the intermediate strata, 10 of the pre-established small plots (1 m2) in the restored (1, 2, 4, 8, 16 years), native sites were utilized for characterization of ve getation and soil physical and chemical properties. Soil Physical and Chemical Analysis Soil samples were collected during July 2005 (w et season) from each plot via a 7.6 cm diameter PVC soil core collector to th e depth of bedrock and stored at 4C until laboratory analysis was performed. Before analyses were performed, all rocks, roots and litter material

PAGE 72

72 were removed from the soil. Bulk density was ca lculated by determining the moisture content of each core by drying a subsample at 60C until a constant weight was achieved. Soil depth was measured from the top of the soil to the bedroc k. Soil exposure was defined as the percent of bare soil without vegetation cove r. Organic matter content was determined as loss on ignition (LOI) by combusting 0.5 g of dry soil at 550C for 4 hours. LOI was calcu lated as the percent of organic matter lost after combustion. Ammonium and nitrate were determined by K2SO4 extraction (Bundy and Meisinger 1994) and analyzed by flow injection with a Bran Luebbe Auto Analyzer 3 Digital Colorimeter for NH4 (EPA Method 350.1) and a Alpkem Rapi d Flow Analyzer 300 Series for NO3 (EPA Method 353.2). A subsample of the K2SO4 extract was digested fo r total kjeldahl nitrogen (TKN) via kjeldahl block digest ion and analyzed by flow injec tion with a Bran Luebbe Auto Analyzer 2 Colorimeter (EPA Method 351.1). Total organic carbon (TOC) was analyzed from the extract with a Shimadzu TOC-5050A Total Organic Carbon An alyzer equipped with a ASI5000A auto sampler. Total phosphorus was dete rmined via HCl ash extraction and analyzed with a Seal AQ2+ Automated Discrete Anal yzer (EPA Method 119-A rev3) (Anderson 1976). Total carbon (TC) and total nitrogen (TN) we re determined by dry combustion with a Thermo Electron Corporation Flash EA NC Soil Analyzer. Carbon, N and P ratios were calculated on a mass basis as C:N, C:P, and N:P. Vegetation Nutrient Analysis Total C and TN were determined by dry com bustion with a Thermo Electron Corporation Flash EA NC Soil Analyzer for the bulk above -ground plant tissue. Total phosphorus was determined via HCl ash extraction and analyzed with a Seal AQ2+ Automated Discrete Analyzer

PAGE 73

73 (EPA Method 119-A rev3) (Anderson 1976). Car bon, N and P ratios were calculated on a mass basis as C:N, C:P, and N:P. Statistical Analysis A non-metric multidimensional scaling (NMS ) ordination was used with PC-ORD (autopilot, thorough mode) (McC une and Mefford 1999) to determine differences and/or similarities in the plant community struct ure within the 1997, 2001, 2003, and 2004 restored wetlands at six months after they were restored. Data for the 1989 site was not made available for six months after restoration; therefore it was not included in this analysis. To evaluate the 1989 site, a NMS ordination was app lied to investigate the 1989 and 1997 site at eight years after restoration to determine differen ces and/or similarities in long term developmental patterns in plant community composition. The NMS ordination technique was used to determine relationships between environmental characters and vegetation community structure for restored sites compared to native sites performed with PC -ORD (autopilot, thorough mode) (McCune and Mefford 1999). The environmental characters were divided into three groups, soil chemical, soil physical, and vegetation chemical parameters with three separa te NMS ordinations performed. The parameters from each analysis that indicated the strongest correlation with respect to differences in native communities verses restored sites were them subj ected to further analysis. Regressions were performed to determine if any relationships ex isted between these parameters among the three divided groups (SAS, JMP) (SAS 2005). Results Species Composition A total of 22 vegetation species were identified within the 60 1-m-2 plots to determine biomass production in the restored and native co mmunities included in this study (Table 3-1).

PAGE 74

74 These species were grouped based on abundance and importance to the restoration goal and ranked based by frequency and percent cover. Species that were that most abundant were grouped by themselves. Several species observed in the sites were rare an d contributed less than 1% to the total above ground biomass. In an effo rt to eliminate clutter and confusion these rare species were grouped together and called Other for ordinatio n analysis. Several Poaceae species (grasses) were observed, however, uni dentifiable to the spec ies level. All Poaceae species were grouped into one group called Poace ae and counted as one species and weighted based on its relative frequency in each site. Environmental Parameters The results of the soil chemical analysis are su mmarized as averages with standard errors and listed for each site in Table 32. Nitrate had non detectable li mits for all sites which was not surprising since this analysis was performed during the wet season. The 2004 site had considerably less extractable NH4, TKN, and TOC concentration as compared to all the other sites. For TC and TN there is a split in the site s with the Native, 1989, and 1997 sites having similar values higher than what was observed in the 2001, 2003, and 2004 sites. This resulted in the same trend with the C:N ratios. The TP valu es for the sites showed no trend with restoration age. However, the native communities had the lowest TP values of 0.33 g kg-1 and the 2003 site had the highest at 0.87 g kg-1. Interestingly, the native site had the highest TN values and the 2003 site had the lowest, 11.0 and 6.2 g kg-1, respectively. This resulted in N:P ratios following the same trend with 48.3 in the na tive and 7.6 in the 2003 sites. The soil physical properties did not result in th e same variability that was seen with the soil chemical parameters (Table 3-3). There wa s not a significant difference in the elevation across the sites. However, the 2004 site had a lowe r elevation than all the others du e to its lower geographical location relative to the other sites. The native sites had an average soil depth of

PAGE 75

75 14.6 cm, whereas the restored sites had a range of soil depths from 0.93 to 2.8 cm. The percent moisture was similar across sites except for the 2004 site which had the lowest moisture content of 46.4%. The native and 2004 sites had si milar bulk densities of 0.41 and 0.39 g cm-3 and the 1989, 1997, 2001, and 2003 sites had similar densities of 0.28, 0.27, 0.27, and 0.30 g cm-3. The low bulk densities are a result of high levels of organic matter present in the soil. This is supported by the high levels of %LOI which is a measure of organic matter content. The vegetation nutrient analysis results are found in Table 34. The lowest TN levels were found in the vegetation in the 2003 at 6.2 g kg-1, which is the same as what was found in the soil in the 2003 site. The vege tation in the native sites had the lowest amount of TP at 0.31 g kg-1 and the vegetation in the 2004 site had the highest at 0.57 g kg-1. The N:P ratios were similar for the native, 1989, 1997, and 2001 sites ranging from 34.8 to 37.7 with the native being the highest. The 2003 and 2004 sites had relatively low N:P ratios comparably at 16.8 and 18.1. Soil and Plant Relationships The NMS ordinations comparing the devel opment of the 1997, 2001, 2003, and 2004 sites at six months of development concluded that a 2-dimensional solution (two axis) was best for each ordination performed with a Monte Carlo test p-value = 0.092. The final stress for the 2dimensional solutions was 12.4 after 90 iteration s. The NMS ordination concluded that there were some difference in initial plant commun ity development within the 1997, 2001, 2003, and 2004 restored wetlands (Figure 3-1). The initia l plant colonization was similar in the 2001 and 2003 restored wetlands at six months after comple te soil removal. Thes e two sites were grouped with overlapping plots in ordination space. The 1997 site plant colonization at six months was clearly different from the 2001 and 2003 sites. The site restored in 2004 had the greatest variability in species composition across plots. This site had si milarities in species composition with the 1997, 2001 and 2003 restored wetland communities (Figure 3-1).

PAGE 76

76 The NMS ordinations comparing the 1989 and 1997 sites at eight years of development concluded that a 2-dimensional solution (two axis ) was best for each ordination performed with a Monte Carlo test p-value = 0.073. The final stress for the 2-dimensional solutions was 12.9 after 90 iterations. The NMS ordination investigating differences the relati onships between plant community development of the 1989 and 1997 sites at eight years after rest oration indicates that theses two sites are similar in species composition at this age of developm ent (Figure 3-2). The ordination suggests variability be tween plots in both sites. Ho wever due to the mixed single group of plots, it is concluded that these two sites are similar in species composition. The NMS ordinations of species composition comparison to environmental characteristics concluded that a 2-dimensional solution (two axis ) was best for each ordination performed with a Monte Carlo test p-value = 0.0196. The final st ress for the 2-dimensional solutions was 10.36 after 91 iterations. With vegetation community structure, the ordination resulted in the native communities clearly being different in vegetation composition as compared to the restored sites (Figures 3-3, 3-4, and 3-5). There is grouping with the restored sites, but they all overlap at some point in ordination space indicating areas of similarity. Schoenus nigricans was the dominant species present in the native communities followed by C. jamaicense. Schoenus nigricans was not found in any of the restored sites, but C. jamaicense was found in small numbers in all sites except the 2003 site. The absence of S. nigricans in any of the restored sites is largely responsible for the distinct grouping of the native communities as indicated by the placement of S. nigricans in ordination space (Figures 3-3, 3-4, and 3-5). Additional species found in the native sites include S. lancifolia, Poaceae, and Other. Poaceae occurred in similar frequencies in all sites. Sagittaria lancifolia were found in all sites, but in small numbers with the fewest in the native communities.

PAGE 77

77 Typha domingensis, Andropogon glomeratus, Andropogon virginicus (Andropogon spp.), and Juncus megacephalus were all absent from the native communities. Typha domingensis was present in all restored sites but the most a bundant in the 1997 and 2003 which is supported by the placement of T. domingensis between the groupings of the 1997 and 2003 sites in the ordination (Figures 3-3, 3-4, and 3-5). Andropogon spp. was also present in all restored sites but the most abundant in the 1997 site. Juncus megacephalus was limited to the 2001, 2003, and 2004 sites and was most abundant in th e 2001 restored site. The overlain joint plot for soil chemical para meters on the NMS ordination indicates that difference in vegetation community composition was driven by C:P, N:P, and TP variables (Figure 3-3). The length of each vector is pr oportional to correlations with the vegetation composition, the longer the length th e stronger the relationship. Th ese vectors also indicate that TP is inversely related to the C:P and N:P ratios (r2 = -0.773 and -0.738, respectively). In addition, due to the weak correla tion of TN with N:P ratios (r2 = 0.308), this suggests that the N:P ratios are driven by the amount of TP in the system. Of the soil physical properties that were investigated, only soil depth correlated with the vegetation composition (Figure 3-4). The direction and length of this vector indicates that the soil depth maintained in the native communities is an important parameter controlling overall vegetation composition. The overlain joint plot for the vegetation nutri ent parameters did not indicate any strong correlations with the vegetation composition (Figure 3-5). The le ngths of all the vectors are short relative to what was found w ith the other ordinations. The di rection of the vectors for the nutrient ratio, however, is simila r to the direction of the same ratios for the soil chemical analysis.

PAGE 78

78 To evaluate relationships between soil chemical, physical and vegetation nutrient parameters, regressions were performed between the parameters that were identified from each of the NMS ordinations. The C:P and N:P ratios of both the soil and vegetation and the soil depth was chosen for regression analysis. A st rong correlation was found to exist between the soil N:P and C:P ratios and the soil depth with an r2 = 0.76 (F=49.2, p<0.0001, d.f.=59) and 0.59 (F=36.0, p<0.0001, d.f.=59), respectively (Figure 36a, and b). Due to the large differences in soil depth between the native and restored sites, th ere was a cluster of point at the low soil depths with the linear regression extended by the deeper depths of the native communities. To investigate if this was a false relationship, the regressions were ran without the native sites to see if this relationship still existed within the cluste r of restored sites. The soil N:P ratios of the restored sites resulted in an r2 = 0.39 (F=5.7, p=0.02, d.f.=49; Figure 3-6c). While this correlation is not as strong as with the inclusion of the native site data, the result is the same positive relationship. The same was found to be true for the C:P ratio and soil depth with the exclusion of the native sites (r2 = 0.36, figure not shown). A relationship between vegetation N:P and C:P ratios with the soil depth was not seen. Both ratios resulted in a positive relationshi p, but the correlation was not significant (r2 = 0.18; F=6.2. p=0.05, d.f.=59 and 0.26; F=9.8, p=0.003, d.f.=59 for N:P and C:P, respectively; Figure 3-7a, and b). Discussion The desired species composition of the restored wetlands in the HID is the community found in the native si tes investigated, a Schoenus nigricans and Cladium jamaicense dominated system. In the first few years after restoration, sites are comprised of species associated with disturbance; i.e. Typha domingensis and Baccharis spp (O'Hare and Dalrymple 2003). However, as the sites develop with time, the native plant species C. jamaicense is becoming more

PAGE 79

79 abundant. Cladium jamaicense is present in small numbers within restored sites but S. nigricans is not present at all. The increased abundance of C. jamaicense in the 1989 site relative to the more recently restored sites, gives encouragement that C. jamaicense may become more abundant in all site s with time. Research has shown that planting desired vegetation communities in restored wetlands could result in faster community development similar to a reference wetland ecosystem (Kellogg and Bridgham 2002). Restored wetlands that re lied on natural re-vegetation had much lower species richness and an increase in exotics after a five to seve n year period as compared to reference wetlands (Seabloom and van der Valk 2003) This suggests that planting is necessary to insure the desired composition in wetland ve getative community structure of restored wetlands. An analysis of species diversity de velopment in restored wetlands of the HID indicated that planting was not necessary to restore high levels of species diversity and composition (See Chapter 2). However, the NM S ordination performed in this study did not indicate similarities between the 1989 and th e native plant communities. There is a clear differentiation between these two sites in ordinati on space. The most distinct contributing factor to this difference in ordination analysis is due to the presence of S. nigricans in the native community. The dominance of this plant species in the native community and the lack of its presence in any restored wetla nd system outweigh any similariti es found between the native and 1989 site when analyzed statistically. Research on S. nigricans within the ENP has been limited and no information on its recruitment and germination requirements in the ENP are known. Studies have been performed in the Netherlands and Mediterr anean regions on its behavior in salt marsh dunes (Ernst and Vanderham 1988, Ernst and Piccoli 1995, Er nst et al. 1995, Ba kker et al. 2007). Schoenus

PAGE 80

80 nigricans has been shown to be dependent on waterl ogged, silicon rich conditions in order to produce fruit (Ernst et al. 1995). Additionally, it has been show n to tolerate dry conditions, however optimal growth occurs under wet conditions (Ernst and Piccoli 1995). In opposition, a more recent study found that S. nigricans had optimal germination and recruitment at moist to dry conditions, suggesting that a retreating groun dwater table would provide optimal conditions for new recruitment (Bakker et al. 2007). While the fluctuating water table of the HID may be an ideal environment for S. nigricans, the silicon concentration could be limiting seed production. Silicon is typically high during earl y stages of primary succession in areas where new soil is derive from volcanic material a nd have been shown decrease with weathering processes (Hedin et al. 2003). In the HID, all soil has been removed and soil development is dependent upon biological growth and processes which could result in soil limited in silicon. By comparing the native and 2003 communities (the two extremes), we found that the native communities are P-limited and that the 2003 restored site may be N-limited immediately following restoration. The soil and vegetation in the native community indicates a P-limited system (N:P ratios of 48 and 38, respectively), whereas the soil and vegetation in the 2003 site indicates a possible N-limited system (N:P of 8 and 17). It has been suggested that a N:P < 14 results in an N-limitation and a N:P > 16 result s in a P-limitation (Koerselman and Meuleman 1996). However, more importantly is the signific ant difference between the N:P ratios from the native and 2003 sites. The farming practices that previously occurr ed in the HID resulted in elevated P and decreased N concentrations relativ e to the native regions of the EN P. The differences in ratios are a result of the native site having the highest N and the lowe st P values and the 2003 site having the lowest N and the highest P values. This imbalance in the N and P concentrations is a

PAGE 81

81 contributing factor that le d to the initial invasion by S. terebinthifolius. After restoration, this imbalance in nutrient concentrations still exists bu t it is not as extreme. In northern parts of the ENP, Typha spp. has been found to invade areas that have been disturbed via enrichments of P concentrations. A good example is Water Conser vation Area 2A (WCA-2A), levels of P have increased resulting in a shift from a C. jamaicense dominated system to one dominated by Typha spp. (either T. domingensis or T. latifolia) (Reddy et al. 1999, Ri vero et al. 2007). Conclusions One important question in the successiona l development of the HID is what are colonization patterns for C. jamaicense and S. nigricans and why are they not dominating the restored sites? Schoenus nigricans has not been found to coloni ze any of the restored wetland communities. This brings up some interesting life history question on the required conditions in order for S. nigricans to successfully recruit, propagate and survive. More research is needed on this species in the ENP to determine germination, recruitment and seed production requirements. Cladium jamaicense, however, is dominant in the 1989 site and is present in more recently restored sites, indicating that with time it will inhabit these wetland areas even if the environmental conditions are not restored to na tive conditions. This st udy provides evidence that a nutrient limitation could be responsible fo r plant community composition in subtropical restored wetland communities. Therefore, restor ation projects need to take into account the environmental nutrient condition needed to de velop and maintain native plant community structure.

PAGE 82

82 Table 3-1. List of all species identified during vegetation biomass collection in the restored sites and the native communities. (n=60 plots) Species NameCSpecies Group Species Name Macrophyte CommunityAAlgae MatsBAndropogon glomeratus AndropogonChara lightBAndropogon virginicus AndropogonChara unidentified species Baccharis angustifolia Other Periphyton unidentified species Baccharis glomeruliflora Other Baccharis halimifolia Other Centella asiatica Other Cladium jamaicense Cladium Eupatorium capillifolium Other Eupatorium leptophyllum Other Fuirena breviseta Other Hydrocotyle umbellata Other Juncus megacephalus Juncus Ludwigia peruviana Other Ludwigia repens Other Mikania scandens OtherBPoaceae Poaceae Sagittaria lancifolia Sagittaria Sarcostemma clausum Other Schoenus nigricans Schoenus Solidago sempervirens Other Typha domingensis TyphaANot included in NMS ordindati on, but present at all sitesBAll Poaceae species were grouped together with the exception of the two Andropogon species which were group together as one group named AndropogonCAll species in species group Other contribut ed less than 1% of the total above ground biomass at the time of samplin g and therefore were weighted with less importance over the most abundant plant species observed

PAGE 83

83 Table 3-2. Average values for soil chemical parameters for each site used in non-metric multidimensional ordination with vegetation community data (n=10 for each site). NDL = non detectable limits. AveSEAveSEAveSEAveSEAveSEAveSE NH4-N (ug g1 ) 33(1.3)53(0.9)72(3)70(4)58(3)16(0.4) NO3-N (ug g1 ) ND L NDLNDLNDLNDLNDL TKN (ug g1 ) 62(2.0)107(4)93(4)129(10)100(6)21(2) TOC (ug g1 ) 1073(23.2)1236(30)1467(32)1217(77)1132(28)806(57) TC (g kg1 ) 188(1.7)178(2)186(2)157(2)153(0.9)158(3) TN (g kg1 ) 11(0.2)10(0.1)10(0.1)7(0.1)6(0.1)7(0.2) TP (g kg1 ) 0.3(0.03)0.7(0.03)0.7(0.03)0.4(0.02)0.9(0.02)0.5(0.02) C:N 17(0.2)18(0.2)19(0.1)25(0.3)25(0.4)22(0.3) C:P 865(44.3)344(19)383(46)597(45)191(7)393(14) N:P 48(2.1)18(0.8)20(2)25(2)8(0.3)19(0.8) Native 1989 1997 2001 2003 2004

PAGE 84

84 Table 3-3. Average values for soil physical properties for each site used in non-metric multidimensional ordination with vegetation community data. (n=10 for each site) AveSEAveSEAveSEAveSEAveSEAveSE Elevation (m)0.5(0.004)0.5(0.005)0.5(0.04)0.5(0.05)0.5(0.004)0.4(0.01) Soil depth (cm)15(0.7)3(0.1)1(0.1)3(0.2)1(0.1)1(0.00) Moisture (%)61(5)64(4)69(7)56(1)59(8)46(5) Bulk Densit y (g cm-30.41(0.02)0.28(0.01)0.27(0.04)0.27(0.02)0.30(0.01)0.39(0.01) LOI (%)22(0.6)23(3.1)24(4.2)15(0.4)14(4.0)20(1.1) 2003 2004 Native 1989 1997 2001

PAGE 85

85 Table 3-4. Average values for vegetation nutrient parameters for each site used in non-metric multidimensional ordination with vegetation community data. (n=10 for each site) AveSEAveSEAveSEAveSEAveSEAveSE TC (g kg-1) 436(21)432(12)418(18)376(33)433(23)395(26) TN (g kg-1) 9(1)11(0.1)11(2)12(1)6(1)10(1) TP (g kg-1) 0.31(0.01)0.34(0.01)0.35(0.02)0.35(0.01)0.43(0.02)0.57(0.2) C:N47(7)41(5)40(1)33(6)71(9)42(1) C:P1723(104)1500(79)1383(54)1202(46)1200(50)769(32) N:P38(2)36(2)35(1)37(1)17(1)18(1) 2003 2004 Native 1989 1997 2001

PAGE 86

86 Figure 3-1. Non-metric multidimensional scal ing (NMS) ordination of vegetation community for each restored site at six months after restoration. The symbols represent each site with the year being the year the site was cleared. Da ta for the 1989 site was not available for six months after restoration.

PAGE 87

87 Figure 3-2. Non-metric multidimensional scal ing (NMS) ordination of vegetation community data for the 1997 and 1998 restored sites at 8 years after restoration. The symbols represent each site with the year be ing the year the site was cleared.

PAGE 88

88 Figure 3-3. Non-metric multidimensional scalin g (NMS) ordination of vegetation community data for restored and native sites. The sy mbols represent each site with the year being the year the site was cleared. The overlain line vectors represent significant correlations between vegetation composition and each soil chemical parameter.

PAGE 89

89 Figure 3-4. Non-metric multidimensional scalin g (NMS) ordination of vegetation community data for restored and native sites. The sy mbols represent each site with the year being the year the site was cleared. The overlain line vectors represent significant correlations between vegetation composition and each soil physical properties.

PAGE 90

90 Figure 3-5. Non-metric multidimensional scalin g ordination (NMS) of vegetation community data for restored and native sites. The sy mbols represent each site with the year being the year the site was cleared. The overlain line vectors represent significant correlations between vegetation composition an d each vegetation nutrient parameter.

PAGE 91

91 Figure 3-6. Relationship between soil depth and soil N:P and C:P ratios for the native and restored wetlands within the HID. A) So il N:P ratio; d.f.=59, F=49.2, p<0.0001. B) Soil C:P ratio; d.f=59, F=36.0, p<0.0001. C) Soil N:P ratio without native site; d.f.=49, F=5.7, p=0.0214.

PAGE 92

92 Figure 3-7. Relationship between soil depth and vegetation N:P a nd C:P ratios for the native and restored wetlands within the HID. A) Vegetation N:P ratio; d.f.=59, F=6.2, p=0.0156. B) Vegetation C:P ratio, d.f.=59, F=9.8, p=0.0027.

PAGE 93

93 CHAPTER 4 NUTRIENT-USE EFFICIENCY AND POTEN TIAL NUTRIENT LI MITATIONS IN SUBT ROPICAL RESTORED WETLANDS Introduction A common indicator used to assess nutrient limitations to vegetation communities is nitrogen to phosphorus (N:P) ratios. A high N:P ra tio at the vegetation level suggests a phosphorus (P)-limited ecosystem and a low N:P ratio suggests a nitrogen (N)-limited ecosystem. A review of fertilization studies in wetland vegetation communities found that N:P ratios < 14 resulted in a N-limited system whereas a N:P ratio > 16 would be P-limited (Koerselman and Meuleman 1996). This finding was further supported by a more extensive review of wetland communities, however, indicating that a critical ratio lo wer than 14 might be more suggestive of a N-limited system (Gsewe ll and Koerselman 2002). A more resent review of fertilization studies which included a wider range of terrestrial plant communities suggests that N:P ratio < 10 and > 20 are more suggestiv e of N and P limitations, respectively (Gsewell 2004). The ranges of N:P ratios between either 14 to 16 or 10 to 20 is suggested to indicate a possible co-limitation in N and P. When evaluating N:P ratios as a tool to assess nutrient limita tions there are several factors to keep in mind. These factors include spatial and temporal variations, elevational differences, and potential potassium limitations, all of which can result in variati ons in tissue N and P concentrations (Shaver and Chapin 1995, Gsewell and Koerselman 2002, Olde Venterink et al. 2002, Olde Venterink et al. 2003). Spatial, tem poral and differences in elevation can have localized affects on water levels which can in turn control nutrient availability. Studies have shown that under dry conditions, vegetation N con centrations will increase resulting in higher N:P ratios and under wet conditions N concentrations will decease re sulting in a lower N:P ratio (Gsewell et al. 2000, Olde Venterink et al. 2002).

PAGE 94

94 The response of vegetation to different levels of nutrient availability is often evaluated by considering their nutrient-use efficiency (NUE) and nutrient-resorption efficiency (NRE). In annuals, the NUE has been defined as the organi c matter produced per un it of nutrient taken up or more simply the inverse of nutrient concentration in plant tissue (Chapi n 1980). For perennial plants, however, it has been argued that the NUE cannot be taken simply as the inverse of plant nutrient concentration (Vitousek 19 82). The NUE for perennials is thus defined as the amount of organic matter lost from plants or permanently st ored within plants per unit of nutrient lost or permanently stored (Vitousek 1982, Birk and Vit ousek 1986). In other words, the NUE is the ratio between above-ground biomass produc tion and nutrient loss in litterfall. Berendse and Aerts (1987), however, have suggested that the aforementioned NUE definitions are inappropriate for assessing the ef ficiency of N for dry matter production at the species level. They suggest the NUE of individual plants should include the mean residence time of the N in the plant as well as the rate of carbon (C) fixation per unit of N in the plant (N productivity). The mean residence time of N is defined by 1/Ln, where Ln is the N requirement per unit of N in the plant (g N g-1 N yr-1). The N requirement is the amount of N that is needed to maintain each unit of biomass during a given time period (g N g-1 dry weight yr-1). The N productivity (A) is defined as dry matter production pe r unit of N in the plant. They suggest that N productivity is important in terms of NUE because the amount of N in the leaves of plants is one of the primary properties that determine th e rate of photosynthesis. By combining the concepts of mean residence time and N productivity the NUE at the species level is the product of the two, A/Ln (Berendse and Aerts 1987). The nutrient-resorption efficiency (NRE) is defi ned as the ratio of the amount of nutrients resorbed from mature leaves to the maximum nutr ient pool in the mature leaves expressed as a

PAGE 95

95 percent (Aerts et al. 1999). The NRE of N from senesci ng leaves is typically around 40-50%, but NREs as low as 0% and as high as 90% have been reported (Aerts 1996, Aerts et al. 1999, Chapin et al. 2002). It has been suggested that the large variation in NRE is in response to nutrient availability status (Aer ts 1996, Killingbeck 1996, Aerts et al. 1999, Richardson et al. 1999), in other words, in nutrient limited systems the NRE of senescing l eaves will be greater than in nutrient rich environments. While earlier reviews of the literature indicated that no such trends have been successfully supported (C hapin 1980, Aerts 1996), a recent study on wetland graminoids found that the NRE of both N and P were on average higher in P-limited systems (Gsewell 2005). At the ecosystem level, NRE could have significant implications in terms of nutrient cycling. To decrease dependence on soil nutrien t availability and nutrient uptake, plants will resorb nutrients during senescence so that they are readily availa ble for future plant growth. Past studies have suggested that efficient retranslocation or low losses of nutrients can increase the fitness of plant specie s in nutrient limited ecosystem s (Grime 1977, Berendse 1994, Richardson et al. 1999). In a ddition, high NRE and NUE of vege tation can limit the regeneration (decomposition) of nutrients in the ecosystem due to low litter nutrien t content (poor litter quality) (Bridgham et al. 1995). The goal of this study was to determine if a nut rient limitation gradient existed in the Holein-the-Donut (HID) region of the Everglades National Park (ENP). This region contains a chronosequence of restored wetland communities that involved complete soil removal to bedrock. The HID was heavily farmed in the earl y to mid 1900s resulting in a P-rich area in the middle of a naturally P-limited ENP. Anthropoge nic impacts, such as farming practices, can drastically change the nutrient dynamics of ecosystems. Wetlands created on abandoned

PAGE 96

96 agricultural land are typically P-rich and N-limite d. As the system develops, more N can be introduced via fixation and organic matter accre tion while the P can become tied up in the substrate shifting the dynamics of the system to a P-limited system. We hypothesized that immediately following soil removal that the restored sites would be N-limited and that with time they would shift to a P-limited system. The objec tives of the study were to 1) determine if a N:P ratio gradient existed in the ch ronosequence of restored wetlands ; 2) determine if a vegetation community level Nor P-limitation exists; and 3) determine if a species level Nor P-limitation exists. Methods Site Description This study was conducted in wetland systems re stored within the Hole-in-the-donut region of the Everglades National Park. Past farming and management practices in the areas that were restored left these systems open to invasion by Schinus terebinthifolius (Brazilian pepper). The nutrient enriched soil, higher elevation (result ing in short hydroperiods) and subtropical conditions of Florida made these disturbe d areas an ideal lo cation for invasion by S. terebinthifolius. The natural surrounding marl prairie wetlands are inundated for approximately six months of the summer season. The goal of the restoration of the HID was to remove the enriched soil and lower the elevati on to increase the hydroperiod to control S. terebinthifolius reinvasion (see Chapter 1 for a more detailed site description). Soil Analysis In April 2005 (dry season) and July 2005 (wet season), soil samples were collected with a 7.6 cm diameter PVC core from 10 plots random ly distributed throughout sites restored in 1989 (16), 1997 (8), 2001 (4), 2003 (2), and 2004 (1) as well as the surrounding native communities (see Chapter 1 for a more detailed site descript ion). The number in pa rentheses indicates the

PAGE 97

97 number of years since restorati on was completed at the time of th is study. Elevation was kept constant at 0.5 m to eliminate hydrology differences as a driving factor of nutrient availability. The soil cores were transported to the laboratory and stored at 4C until analysis. Before analyses were performed, all rocks, roots, and l itter material was removed from the soil. Within 24 to 48 hours of sample collection, each soil sample was extracted for ammonium (NH4) with K2SO4 (Bundy and Meisinger 1994) and set up for in cubation for potentially mineralizable nitrogen (PMN) (or biologically available nitrogen) (Keeney 1982, Bundy and Meisinger 1994, White and Reddy 2000). Soil extracts were analyzed via flow injection analysis with a Bran Luebbe Auto Analyzer 3 Digital Colorimeter (EPA Method 350.1). A subsample of each soil was dried at 60C for 3 days then ground with a ball grinder to a fine powder for total N and P analysis. Dry soil samples were analyzed for total N with a Thermo Electron Corp. Flash EA 1112 Series NC Soil Analyzer. Total P was dete rmined via HCl ash extraction and analyzed with a Seal AQ2+ Automated Discrete Anal yzer (EPA Method 119-A rev3) (Anderson 1976). Nitrogen and P ratios were calc ulated on a mass basis as N:P. Vegetation Analysis Nutrients Nitrogen and P were determined for both the composite biomass (co mmunity level) and selected individual plant species (species level) within 10, 1 m2 plots in each site. To determine nutrients in composite biomass, all the vegetation in a 1 m2 plot was cut at the soil surface, separated by live and senescent plant tissue, bu lked and dried at 70C until all moisture was removed. Once dry, all vegetation from each plot was passed through a Wiley Mill tissue grinder equipped with a 2-mm mesh screen to achieve homogeneity. A subsample was ball ground to a fine powder for N and P analysis. Individual plant species of Cladium jamaicense, Schoenus nigricans, Typha domingensis, and Sagittaria lancifolia were collected from each site

PAGE 98

98 near each plot when available. Each species was not always presen t at each site. Plant fractions of live, senescence, and litter were collected fo r each species. The samples were dried at 70C until moisture was removed and ground with a Wiley Mill tissue grinder equipped with a 2-mm mesh screen. A subsample was ba ll ground to a fine powder. A ll plant samples were analyzed for total N with a Thermo Electron Corp. Flash EA 1112 Series NC Soil Analyzer. Total P was determined via HCl ash extraction and analyzed with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 119-A rev3) (Anderson 1976). Nitroge n and P ratios were calculated on a mass basis as N:P. Nutrient-use efficiency and nutrient-resorption efficiency The NUE and NRE were determined followi ng methods outlined by Berendse and Aerts (1987), Aerts et al. (1999), and Feller et al. (2002). The liv e and senescent fraction of each biomass collection and individual species were used in the following calculations. The NRE was calculated as the percentage of N (or P) recove red from senescing leaves before stem fall: NRE = (Nlive stems Nsenescent stems)/(Nlive stems) 100 (%) (4-1) The NUE of the individual plants was calculated as: NUE = A/Ln, (g biomass mg-1 N) (4-2) where A is the N productivity, dry matter production per unit of N in the pl ant and is calculated as: A = biomass production (g dry wt m-2 yr-1)/biomass N (mg N m-2 y-1), (g dry wt mg-1 N) (4-3) and Ln is the N requirement per unit of nitroge n in the plant and is calculated as: Ln = Nlive plant (mg m-2) / Nsenescent plant (mg m-2), (unitless) (4-4) The N requirement is the fraction of N that is remaining in the biomass during a given time period. The NUE at the ecosystem level was taken as biomass production per unit of N in senescent leaves (g dry wt biomass mg-1 N).

PAGE 99

99 Statistical Analysis All data collected were analyzed statistica lly using Fit Model in JMP Version 5.1 (SAS 2005). Analysis of variance (ANOVA) was pe rformed to investigat e site and seasonal differences in soil, composite biomass, and i ndividual plant species for nutrients and each nutrient index. Regressions were performed to determine if any strong relationships existed between variables. Multiple comparisons were made using the Least Squa re Means test and to determine Pearsons correlation coefficients between all variables. Results N and P Concentrations Nitrogen and P concentrations in the so il ranged from 5.8-11 and 0.29-0.93 g kg-1, with the native site having the highest N and lowest P concentrations (Table 4-1). Significant differences were observed for soil N, P, and N:P ratios across sites, but not between seasons (Table 4-2). The native, 1989, and 1997 sites were not significantly different from each other in N concentration but were signifi cantly different from the 2001, 2003, and 2004 site (p<0.0001). For P, the native, 2001, and 2004 sites were not si gnificantly different from each other, but were significantly different from the 1989, 1997, a nd 2003 sites (p<0.0001). The soil N:P ratios in the native site were significantly higher than all the restored si tes with a p<0.0001. Within the restored sites the soil N:P was significantly di fferent between the 2001 and 2003 sites but neither were significantly different for the 1989, 1997, or 2004 sites (p<0.001). While soil N and P concentrations were unaffected by seasonality (wet verses dry season), the composite vegetation N concentrations were affected by dry and wet seasonal changes (Table 4-3). The biomass N concentration resulted in a significant difference between site and season (Table 4-2). With the exception of the 2003 restored site, the N concentration in the vegetation was significantly less in the we t season as compared to the dry season (p<0.0001). A similar

PAGE 100

100 trend was observed for the vege tation N:P ratios (p=0.0473). Phos phorus concentrations in the vegetation did not result in a significant difference between season (p=0.6563), but were significantly different between si tes (p<0.0001). The native vege tation P was significantly less than all restored sites except for 2001. In a ddition, during the wet seas on the native community N:P ratio was significantly higher than the restored communities (p<0.0001). At the species level, the two native species (C. jamaicense and S. nigricans) had significantly higher N:P ratios th an the two restored site species included in this study (p=0.0003; Table 4-2 and 4-4). Sagittaria lancifolia contained almost twice the concentration of N and two to four times as much P over the othe r three species with an average between all sites at 18.9 and 1.4 g kg-1, respectively (p<0.0001; Table 4-4). None of the species included were significantly different across sites for N, P or N:P (Table 4-2). Nutrient Limitations To determine if a nutrient limitation gradient ex isted between native and restored sites, we evaluated relationships between N and P concentra tions of the vegetation a nd soil, drivers of N:P ratios, nutrient-resorption efficiencies (NRE), and nutrient-use efficiencies (NUE) of the vegetation at both the community level and species level. Relationships between N and P concentrati ons for vegetation samples collected are represented in Figure 4-1. We included lines to represent the cut-off for N:P of 14 and 16 to assess where Nand P-limitations may occur. The critical N:P ratios we used are N:P<14 (Nlimited) and N:P>16 (P-limited). These critical ra tios have been proven va lid for several wetland vegetation communities (Koerselman and Meuleman 1996, Gsewell and Koerselman 2002). Any points which fall between the N:P 14 and 16 lin es represent possible co-limitations. During the dry season, few points fell between the lines or above the N:P = 14 line indicating little Nlimitation exists. The majority of the points fell below the N:P = 16 line, which strongly

PAGE 101

101 suggests a P-limitation in most sites (Figure 4-1a). For the wet season, the data suggests that the 2003 site and portions of the 1997 and 2004 site ha ve an N-limitation or co-limitation (Figure 41b). During both seasons, the native vegetation communities have N:P ratios greater than 16 and therefore are P-limited. To gain more insight on the type of limitations present within the HID, we plotted N and P concentrations against the N:P ra tios for both the vegetation biom ass and soil (Figure 4-2). The N:P ratios in both the soil and the vegetati on biomass are controlled by P alone. Weak relationships exists between N concentration and the N:P ratios, no definite trends were observed (Figures 4-2b and d). Interestingly, the same exponential change in P co ncentration to N:P ratio exists in both the vegetation biomass and the so il (Figures 4-2a and c) This relationship suggests that there is a critical P concentration that could result in a shift from an N-limitation towards a P-limitation. In an attempt to determine the point at which a critical P concentration was reach, we performed linear regressions on each half of the curve (Figures 4-2a and c). We statistical determined the mid-way point with st ep-wise regressions until maximum r2 values were achieved for each slope. The regression equations were so lved to determine at wh ich point the two lines would cross. This is the critical P concentra tion and N:P ratio. For the vegetation biomass, the critical P concentrations and N:P ratio is 0.44 g kg-1 and 18.3, respectively (Figure 4-2a), and the soil is 0.55 g kg-1 and 13.7, respectively (Figure 4-2c). The vegetation is then P-limited at N:P ratios great than 18.3. Due to the lack of a re lationship between N concentration and N:P ratios, it is hard to say with certainty that at N:P ratios less than 18.3 would result in an N-limitation. The community level nutrient-r esorption efficiency of N (NRE-N) was low for the native sites (appox. 20%) and as high as 60% for the 1989 restored site (Figure 4-3a). Significant

PAGE 102

102 difference were found between sites for NRE-N, but not between dry and wet season (p=0.0034 and 0.1874, respectively; Table 4-5). The NRE -N in the vegetation biomass in 1989 site was found to be significantly higher th an that of the native vegetati on (Figure 4-3a). The native vegetation was not significantl y different from any of the other restored sites. Significant differences were found between site s and wet and dry seas ons for the nutrientresorption efficiency of P (NRE-P) of the ve getation biomass (Table 4-5). There were no significant differences found duri ng the dry season across sites for NRE-P, but during the wet season the 2004 sites had significantly lower NREP as compared to the 1989 and 1997 sites, but were not significantly different from the native 2001 or 2003 sites (Figure 4-3b). In addition the NRE-P was considerably higher than the NRE-N (Figures 4-3a and b). The community level nutrient-use efficiency of N (NUE-N) was significantly lower in the 2004 site as compared to the vegetation found in the 1989, 1997, and 2003 (Figure 4-3c and Table 4-5). This suggests that the vegetation found in the 2004 site is more efficient with nitrogen because it is less available. The NUE-N of the native site vegetation was not significantly different from any of the vegetation in the restored sites. Significant seasonal differences were observed between the 1989, 2001, and 2003 sites (p=0.0172; Table 4-5 and Figure 4-3c). For all three of these sites, th e NUE-N was significantly less during the dry season as compared to the wet season. The nutrient-use efficiency of P (NUE-P) was significantly greater in the native vegetation over all the vegetation found in th e restored sites (p<0.0001; Tabl e 4-5 and Figure 4-3d). This increased efficiency with P in the native communiti es could be a result of a greater P-limitation. There was no significant differen ces found between dry and wet s eason for NUE-P for any of the vegetation communities (p=0.7591; Table 4-5).

PAGE 103

103 We regressed the N:P ratios of the vege tation biomass with the NUE-N and -P to determine if any relationships existed that could further indicate a Nor P-limitation. We found that a positive relationship exists between N:P ratio and NUE-P (r2=0.48; F=29.6, p<0.0001, d.f.=33; Figure 4-4b) and no relationship exists between N:P ratio and NUE-N (r2=0.016; F=0.53, p=0.47, d.f.=33; Figure 4-4a). With in creasing N:P ratios of the vegetation in the HID, P use becomes more efficient, suggesting that when P is limited the vegetation use it more wisely. To investigate species level response to pot ential nutrient limitations, we determined the NRE-N and -P and the NUE-N and -P for the dominant plant species found in each site. The dominant species found in the native communities was C. jamaicense and S. nigricans and the two dominant restored species found were T. domingensis, and S. lancifolia. C. jamaicense was also found in the 1989 restored site and S. nigricans was only present in the native communities. No significant differences were found between sites or species for NR E-N or -P (Table 45). The NRE-N and -P was at the species level (Figure 4-5a) was similar to what was found at the community level (Figure 4-3a). The NREN ranges from 17.6.2 to 44.1.6% and the NRE-P ranged from 66.3.2 to 82.6.3%. The NUE-N in S. lancifolia is significantly less than what was found for C. jamaicense, S. nigricans, or T. domingensis (p<0.0001; Figure 4-5c and Tabl e 4-5). No significant difference was found for NUE-N between sites at the species level (p=0.2005). Significant differences were observed for NUE-P at the species level both between sites and individual species. The native community species had significantly higher NUE-P over all the species in each restored site (p=0.0199). Sagittaria ha d significantly less NUE-P than Cladium or Schoenus but not Typha. The NUE-P of Typha was significan tly less than Cladium

PAGE 104

104 and Schoenus that was present in the native communities but not the Cladium found in the 1989 site (Figure 4-5d). Multivariate analyses were performed to obtain Pearsons correlation coefficients for each vegetation parameter against soil and vegetation properties. We conducted this multivariate test twice, at the community level (Table 4-6) and species level (Table 4-7). We considered any correlation coefficient that was .5 significan t. At the community level, the NRE-N was strongly correlated to both plan t N and P concentration. NRE-P was not correlated to any of the variables included. NUE-N was pos itively correlated to available NH4, but not to PMN. NUE-P was positively correlated to soil and plant N:P and was negatively correlated to plant P concentration. The plant N:P ratio was positively correlated to the soil N:P ratio and negatively correlated to soil and plant P concentration. In addition to being correla ted to NRE-N, plant N was also positively correlated to plant P. At the species level (Table 4-7), NRE-N was ne gatively correlated to plant N:P ratios. The NUE-N was positively correlated to NRE-P and NUE -P and negatively correlated to plant N and P concentration. The NUE-P was positively correlated to plant N:P ratios and negatively correlated to plant N and P concentrations. In addition, the NUE-P was positively correlated to soil N:P ratios but negatively correlated to soil P concentration. Plant N:P ratios negatively correlated to plant P, soil P, and soil NH4. However, plant N:P ratios positively correlated to soil N:P ratios. Plant N concentration positively correlated to plant P concentrations. Additionally, plant P positively correlated to soil P and negatively correlated to soil N:P ratios. Discussion We investigated the three obj ective of this study by looking at relationships between soil and plant N and P concentration and N:P ratios as well the NRE and NUE of the vegetation at

PAGE 105

105 the community and species level. We utilized thes e tools to evaluate the success of restoration in terms of nutrient limitations as a control over vegetation community structure. Community Level Limitations Large variations occur in N content of the vegetation, but the variation was small for P content. It has been suggested that these variatio ns are a result of differences in the availability of P and N in the soil (Koerselman and Meuleman 1996). This implies that plants will respond accordingly to increases in nutrient availability. In other words, if the availability of N or P increases, it will result in an in crease in plant N or P content. In this study, no relationship was found between soil N availability indices and plant N or N:P ratios. Howe ver, we did find that with increases in soil P there was an increase in plant P and a decrease in plant N:P ratios. While the soil N and P concentration are not direct meas ures of nutrient availability, they provide an indication that the HI D is driven by P. From N:P ratios, we find that a community leve l P-limitation persists in all sites except the restored 2003 site. Based solely on N:P ratios, the 2003 site appears to be N-limited (Figure 41). Past research has shown that N:P ratios of community level vegetation clearly differentiate between Nand P-limitations (Koerselman and Meuleman 1996, Gsewell and Koerselman 2002, Tessier and Raynal 2003, Gsewell 2004). The evidence to support the application of the N:P ratio as an indicator of nutrient limitation comes from a series of fert ilization studies. These studies have allowed scientists to develop critical N:P ratios that could indi cate shifts from Nto P-limitations. No relationship was found between the N:P rati os and the N concentrations, therefore we were unable to define a critical N concentration level (Figures 4-2b and d). The lack of this relationship limits us from concluding with certainty that a N-limitation exists. However, since P is clearly limiting and driving the system, we were able to determin e a critical N:P ratio. As long

PAGE 106

106 as either N or P is limiting, a critical N:P ratio can be determined (Aer ts and Chapin 2000). The community level critical N:P ratio is 18 for the HID (Figure 4-2a). This is slightly higher than what has been shown in the litera ture, most likely as a result of seasonal vari ation in hydrology. Gsewell (2004) found that critical N:P ratios as high as 20 were more appropriate when dealing with upland ecosystems. These wetlands are dry for approximately half of the year and as a result the critical N:P ratio could be higher. It could be argued that due to the nature of the restoration method (complete soil removal) that the strong relations hip found between the soil P and plant P (and soil N:P and plant N:P) is because some of the soil P is derived from the vegetation community. However, by taking a closer look at the ratios themselves, the soil and plant N:P ratios are ve ry different with the exception of the 2003 site (Table 4-1 and 4-3). In addition, yearly su rveys of the vegetation community structure indicate plant communities ha ve changed with time (O'Hare and Dalrymple 2003), therefore the vegetation contribution to th e soil nutrient pools coul d also change with time. While significant differences were found betw een the sites for both community level NREN and NUE-N, these differences did not relate to the nutrient li mitations suggested by the N:P ratios. In addition, we found that as N av ailability increased (based on extractable NH4; Gsewell and Koerselman 2002) so did the NUE-N, supplying further evidence that N is not limiting the community level vegetation. As a resu lt, N is not limiting in the native or restored plant communities and the differences observe d in NRE-N and NUE-N are likely driven by individual species traits (i.e., partitioning, pr oductivity, reproduction and translocation), not a nutrient limitation. These results are the opposit e of what has been observed in experimental

PAGE 107

107 fertilization studies in wetland systems where the NUE-N of the vegetation was found to increase when N is limiting (Meuleman et al. 2002, Feller et al. 2003). Examining the NRE-P and NUE-P at the commun ity level provides additional support that P controls the plant community nutrient availabil ity. Not only did the vegetation have high rates of NRE-P, the native communities have a much gr eater NUE-P over all the restored sites. We found that the native communities had the highest N:P ratios and therefore are the most Plimited. Since the ENP is naturally P-limited, it was not surprising that the vegetation community of the native sites had the highest NUE-P. Furthermore, there was a strong relationship between soil and plant N:P ratios an d the NUE-P, concluding that when P is limiting the vegetation community will use P more efficiently (Figure 4-4). This relationship also reveals that the 2003 plant community has the lowest N:P ratios and NUE-P while the native community has the highest N:P and NUE-P. We see no eviden ce within the other restored communities to suggest that a gradient in nutrient li mitation exists based on site age. Species Level Limitations It has been found that indivi dual species can differ from community level N:P ratios, therefore, making it difficult to determine species level nutrient limitations (Aerts and Chapin 2000). Koerselman and Meuleman (1996) attempte d to address the question of whether or not species level critical N:P ratios were identical to that of comm unity level. They argued that many of the studies they included in their re view were near monoc ultures and therefore interspecific differences in critical N:P ratios were likely to be in significant. In a later study, Gsewell and Koerselman (2002) f ound that the N:P ratios of indivi dual plants species varied considerably and that interspecific variation of species coul d make it difficult to determine species level nutrient limitations.

PAGE 108

108 We found that the N:P ratios of the species included in th is study were similar to the community level N:P ratios in the 1997 and 2003 sites, but not in th e native and 1989 sites (Tables 4-3 and 4-4). The community pres ent in the 1997 and 2003 s ite are dominated by T. domingensis (~50-60% of the above-ground biomass; see Chapter 2) and therefore the community level N:P ratios are driven by T. domingensis. The native site is dominated by S. nigricans (~60% of the above-ground biomass) and C. jamaicense (~30 of the above-ground biomass), whereas the 1989 site is dominated by C. jamaicense (~40-60%; see Chapter 2). At this dominance level, the community level N:P ratios were higher than the species level N:P ratios. Even though all three of these species have similar dominance levels in their perspective sites, their N:P ratios do not necessarily predic t community level N:P ratios. Therefore, we conclude that species level critical N:P ratio may be different than community level critical ratios. The NRE-N and NUE-N indicate differences between species but not across sites (Figure 4-5a and c). This suggests th at they are driven by species differences, not site nutrient limitations. At the species level, we found that with increases in both plant N and P concentrations the NUE-N would decrease (Table 4-7), however, this wa s not observed at the community level. This indicates, that at the sp ecies level, the NUE-N is driven by species N and P requirements not nutrient availability. In addition, this su ggests that there was not a Nlimitation at the species level within the HID ecosystem. The NUE-P offers a different conclusion a bout P-limitations at the species level. We observed that the NUE-P varies between species as well as across sites (Table 4-5 and Figure 45). Cladium jamaicense has a lower NUE-P in the 1989 site than in the native communities. Since the 1989 site is not as P-limited as the native site based on N:P ratios, C. jamaicense was

PAGE 109

109 not as efficient with P. A similar relationshi p was observed between species level plant P and community level plant P with the soil P and soil N:P ratios (Table 4-6 and 4-7). From this, we concluded at both the community level and specie s level, as soil P becomes more limited, the plants became more efficien t with their use of P. It should be noted, without conducting a N and P fertilization study directly, it is difficult to say with certainty that a Nor P-limitation exists (Vitousek 2004). Over the past decade, several review papers of fertil ization studies in terrestrial systems have been conducted to develop critical N:P ratios which allow us to ma ke inferences about nutrient limitations without conducting a fertilization study (Koerselman and Meuleman 1996, Gsewell and Koerselman 2002, Gsewell 2004). These studies focused primar ily on community level limitations that did not consider multi-species interactions. Therefor e, it is difficult to make conclusions about species level nutrient limitations solely by considering their N:P ratios. The use of additional tools, such as the nutrient-reso rption efficiency (NRE) and nutrien t-use efficiency (NUE), could be used in conjunction with the N:P ratios of in dividual species to potentially determine species level nutrient limitations (Vit ousek 1982, Vitousek 1984, Berendse and Aerts 1987, Aerts and Decaluwe 1994, Feller et al. 2003, Gsewell 2005). Conclusions The vegetation communities which have developed in the restored wetlands of the HID are very different than the surrounding desired native plant communities. Several factors can control re-vegetation patterns after distur bances. In this study we eval uated potential nutrient limitations as a control over vegetation community structure and restoration success. We found that soil N and P cont ent varied considerably within the restored sites but the N:P ratios were less variable. The N:P ratios of the native plant community were two to three times greater than the ratios found in the restor ed communities. Additionally, soil P and soil N:P

PAGE 110

110 ratios exhibited controls on plan t N:P at both community and species levels. No such conclusion can be made in terms of soil N. At the vegetation community level, the nativ e plant community has N:P ratios and NUE-P that are two to four times greater than that of restored plant communities. The species level N:P ratios and NUE-P of the native communities were al so two to four times greater than the species found in the restored communities. This suggest s at the community and species levels, native sites are more P-limited than the restored sites. Our study shows that a P-limitation is prevalent in the native communities as well as most of the restored communities. Little evidence was found to support a N-limitation in any of the sites. The N:P ratios of the site restored in 2003 imply that it may be N-limited not P-limited, however, no other data collected at the community or species level indicates a N-limitation is prevalent. At the community level, we f ound no differences in the NUE-N between sites, additionally; the differences observed in species level NUE-N could be due to plant traits not a N-limitation. To conclude, due to differences in the level of P-limitations in the restored sites verses the native communities, the increased levels of soil P could influence the re-vegetation patterns of the restored wetlands. At th ese higher levels of P, the desired vegetation composition (Cladium jamaicense and Schoenus nigricans co-dominance) is not achieve d. A greater understanding of the lasting impacts of the residu al P in the restored wetlands is necessary to determine if the desired native vegetation species will inhabit these wetlands with time.

PAGE 111

111 Table 4-1. Summary of the av erage nitrogen, phosphorus, and N:P ratios of the soil for each of the sites during the dry and wet season. (n=10) SiteSeasonAveSDAveSDAveSD NativeDry 11.01.60.290.154519.1 Wet11.02.20.330.314821.0 1989Dry 10.51.60.810.40155.0 Wet9.71.20.660.32188.1 1997Dry 9.41.30.930.35126.2 Wet10.11.30.740.302023.5 2001Dry 7.01.20.460.22189.5 Wet6.51.40.370.192520.4 2003Dry 5.81.00.900.1972.0 Wet6.21.40.870.2482.7 2004Dry 7.65.00.440.111810.1 Wet7.52.50.450.17198.3 N (g kg -1 ) P (g kg -1 ) N:P

PAGE 112

112 Table 4-2. Summary of resu lts from a two-way ANOVA test with dependant variables of nitrogen (N) concentration, phosphorus (P) co ncentration, and N:P ratios for the soil, community level vegetation (composite biomass), and species level Source of Variationd.f.F statprob. > Fd.f.F statprob. > Fd.f.F statprob. > F Soil Site521.7<0.0001516.1<0.0001518.9<0.0001 Season10.010.928411.90.170612.10.1467 Site*Season50.40.859550.60.716850.30.9414 Vegetation Community Site510.5<0.000157.5<0.0001522.2<0.0001 Season163.8<0.000110.20.656314.00.0473 Site*Season54.30.001053.00.013456.5<0.0001 Species Site31.10.374130.70.564230.40.7870 Species452.6<0.0001416.5<0.000149.60.0003 N (g kg -1 ) P (g kg -1 ) N:P

PAGE 113

113 Table 4-3. Summary of the average nitrogen, phosphorus, an d N:P ratios of the vegetation community (composite biomass) for each of the sites during the dry and wet season. (n=10 for dry season, n=20 for wet season) SiteSeasonAveSDAveSDAveSD NativeDry 9.41.10.310.123823.2 Wet5.91.00.140.085420.4 1989Dry 10.61.20.340.133615.7 Wet8.23.30.420.31238.6 1997Dry 11.02.20.350.163510.5 Wet6.92.40.410.23196.1 2001Dry 11.71.30.350.143711.8 Wet7.62.00.310.172910.6 2003Dry 6.20.90.430.21176.1 Wet6.32.00.580.25125.6 2004Dry 9.61.80.580.22188.0 Wet7.81.90.420.14206.1 N (g kg -1 ) P (g kg -1 ) N:P

PAGE 114

114 Table 4-4. Comparison of nut rient characteristics at the species level for nitrogen (N) concentration, phosphorus (P) concentration, and N:P ratio. Biomass numbers are an average for all plots at each site. (n=3) Biomass% Total SpeciesSiteTypeAveSDAveSDAveSD (g m2 ) BiomassC. jamaicenseNativeLive8.40.980.220.03390.378.224.0 Std dead5.80.790.050.0212925.7 Littern.a.n.a.n.a.n.a.n.a.n.a. 1989Live9.61.070.260.073912.469.428.9 Std dead8.01.990.080.019612.3 Litter8.51.780.130.016919.7S. nigricansNativeLive8.01.150.230.02355.1153.147.1 Std dead5.70.600.050.001131.5 Littern.a.n.a.n.a.n.a.n.a.n.a.S. lancifolia1989Live19.10.321.300.60167.30.90.4 Std dead16.78.580.841.014846.5 Litter13.7-0.34-401997Live18.51.851.330.13140.13.01.1 Std dead15.15.640.470.26347.3 Litter11.6-0.12-942003Live19.9-1.53-13-1.10.5 Std dead14.7-0.39-38Littern.a.n.a.n.a.n.a.n.a.n.a.T. domingensis1989Live10.71.380.650.06160.893.839.0 Std dead6.61.180.170.094316.2 Litter7.31.150.140.107136.6 1997Live9.72.120.680.23151.7199.872.5 Std dead6.00.400.080.01788.4 Litter7.10.580.100.037821.2 2003Live10.30.290.790.02130.5156.374.7 Std dead6.80.470.160.03446.3 Litter7.20.440.180.04428.3 N (g kg -1 ) N:P P (g kg -1 )

PAGE 115

115 Table 4-5. Summary of resu lts from a two-way ANOVA test with dependant variables of nutrient-resorption efficiency of nitrog en (NRE-N) and phosphorus (NRE-P), and nutrient-use efficiencies of nitroge n (NUE-N) and phosphorus (NUE-P) for the community level vegetation (composite biomass), and species level. Source of Variationd.f.F statpr ob. > Fd.f.F statprob. > Fd.f.F statprob. > Fd.f.F statprob. > F Vegetation Community Site 55.00.003453.30.022953.30.0228511.3<0.0001 Season 11.90.187419.40.005616.60.017210.10.7591 Site*Season50.60.7015510.439351.40.270850.20.9726 Species Site 31.20.354231.10.393831.70.200534.30.0199 Species 42.80.059841.30.3025413.0<0.000143.20.0379 NUE-N NUE-P NRE-P NRE-N

PAGE 116

116 Table 4-6. Pearsons correlati on coefficients for the seven pl ant community level (composite biomass) characteristics addr essed in this study. Tabl e abbreviations are nutrientresorption efficiency of n itrogen (NRE-N) and phosphorus (NRE-P), and nutrient-use efficiencies of nitrogen (NUE -N) and phosphorus (NUE-P), NH4 is an extractable ammonium, potential minerali zable nitrogen (PMN). NRE-NNRE-PNUE-NNUE-PPlant N:PPlant NPlant P Soil Variables NH40.16040.03050.60740.0066-0.0468-0.1210-0.0080 PMN0.47270.47790.0598-0.0212-0.07760.34370.2796 Soil N0.25080.38490.13710.35190.32010.1623-0.0560 Soil P0.48500.29280.1598-0.4159-0.58390.29650.6798 Soil N:P-0.3620-0.04510.01110.57580.7282-0.2998-0.6675 Plant Variables NRE-N 0.48100.2799-0.1825-0.26310.78320.6290 NRE-P 0.37160.4671-0.02670.18880.1947 NUE-N 0.2135-0.1273-0.3205-0.0518 NUE-P 0.6930-0.3019-0.6307 Plant N:P -0.1613-0.7637 Plant N 0.6146 Plant P

PAGE 117

117 Table 4-7. Pearsons correlation coefficients for the seven species level characteristics addressed in this study. Table abbreviations are nutrient-resorption efficiency of nitrogen (NRE-N) and phosphorus (NRE-P), and nutrien t-use efficiencies of nitrogen (NUEN) and phosphorus (NUE-P), NH4 is an extractable ammonium, potential mineralizable nitrogen (PMN). NRE-NNRE-PNUE-NNUE-PPlant N:PPlant NPlant P Soil Variables NH40.18260.0310-0.1915-0.4383-0.56370.27330.4540 PMN-0.2020-0.1249-0.3172-0.2919-0.03540.21160.0591 Soil N0.05230.01300.23720.29500.1390-0.2380-0.2833 Soil P0.43120.0245-0.0966-0.5488-0.73570.30020.5682 Soil N:P-0.23120.08740.33650.76800.7166-0.4282-0.6095 Plant Variables NRE-N 0.18500.1696-0.1481-0.58850.35550.4733 NRE-P 0.69750.4313-0.0403-0.4052-0.3749 NUE-N 0.70050.1761-0.7975-0.6513 NUE-P 0.6694-0.6782-0.7861 Plant N:P -0.4411-0.7668 Plant N 0.8667 Plant P

PAGE 118

118 Figure 4-1. Relationship between community leve l vegetation N and P concentration for each site in the HID. A) Dry season. B) Wet season. The solid line on each graph depicts N:P ratio of 14 and the dashed line de picts N:P ratio of 16 (mass basis).

PAGE 119

119 Figure 4-2. Relationships between nitrogen (N) and phosphorus (P) c oncentration and N:P rations. A) Vegetation P. B) Vegetation N. C) Soil P. D) Soil N. Data points are coded by sites. The lines on A and C depi ct linear regressions for the two slopes found by splitting the data at the midway poi nt. The point in which the two lines cross as labeled as the critical P concentration and N:P ratio.

PAGE 120

120 Figure 4-3. Nutrient-use e fficiency and nutrient-resorpti on efficiency of nitrogen and phosphorus of the vegetation community (c omposite biomass). A) Nitrogenresorption efficiency (NRE-N). B) phosphorus-resorption efficiency (NRE-P). C) Nitrogen-use efficiency (NUE-N ). D) Phosphorus-use efficiency (NUE-P). Bars within each site with the same lowercase le tters that are not significantly different at p<0.05. (n=6)

PAGE 121

121 Figure 4-4. Linear regression between N:P ratio of the vegetation community (composite biomass). A) nutrient-use efficiency of nitrogen (NUE-N); F=0.53, p=0.47, d.f.=33. B) Nutrient-use efficiency of phosphorus (NUE-P); F=29.6, p<0.0001, d.f.=33.

PAGE 122

122 Figure 4-5. Nutrient-use efficien cy and nutrient-resorption efficiency of nitrogen and phosphorus for the individual species included in this study. A) Nitrogen -resorption efficiency (NRE-N). B) Phosphorus-resorption efficien cy (NRE-P). C) Nitrogen-use efficiency (NUE-N). D) Phosphorus-use efficiency ( NUE-P). Bars for each species with the same lowercase letters that are not significantly different at p<0.05. (n=3)

PAGE 123

123 CHAPTER 5 CONTROLS ON REGENERATION OF NUTRIENTS FROM THE DOMINANT VEGET ATION IN NATIVE AND RESTORED WELTANDS Introduction Decomposition of plant litter material is a key process for nutrient cycling within ecosystems. Wetland ecosystems are often de scribed as detritus-based systems where decomposition of macrophyte litter is considered an important source of energy (Mann 1988). Emergent macrophytes often constitute a major portion of the primary production and in turn contribute largely to the organic matter accu mulation in wetlands (Wetzel 1990). Decay of senescent plant material begins in the standing dead plants befo re falling to the ground (Kuehn and Suberkropp 1998). Plant matter decomposition of fallen dead material can be divided into three distinct phases: an initial rapid loss due to leaching, microbial colonization and degradation, and physical and bi ological fragmentation (Valiela et al. 1985, Megonigal et al. 1996). There are several factors that control the rates of decomposition. Mean annual temperature, precipitation, soil moisture, mi crobial activity (i.e., enzymes), chemical composition of the soil and litter material (i.e., available nitrogen (N), phosphorus (P), carbon (C), and C:N:P ratios), and litter cellular quality (i.e., lignin, ta nnin, amino acids, carbohydrates,) are some of the most important factors that can control decompos ition rates (Aerts and deCaluwe 1997, Gartner and Cardon 2004). Extra-cellular enzymes have long been recognized as key contributors that control rates of decomposition (Schimel and Weintraub 2003). These enzymes are the primary means by which complex organic compounds are degraded by micr obial organisms. Enzyme production is both energy and N intensive (Koch 1985). When nutrien ts are limiting, microbes will produce extracellular enzymes to mobilize nutrients from de trital material. In contrast, high nutrient

PAGE 124

124 availability often negatively corr elates with enzyme ac tivities that release nutrients (Crst 1991, Sinsabaugh and Moorhead 1994). Plus, the quality of litter material can also be an important determinant in both microbial production of enzymes and rates of decay (Corstanje et al. 2006). The microbial communities responsible for decomposition will both mineralize and immobilize nutrients from the lit ter material. Furthermore, d ecomposition rates have been found to increase with the additions of N and P (Jordan et al. 1989). Rates of decomposition may be limited if either N or P concentrations are a lim iting resource within the ecosystem (Aerts and deCaluwe 1997, Feller et al. 2003). By increasing a limiting nutrient in the soil, the litter quality can be altered and in turn in crease decomposition rates and nu trient turnover. Nutrient limitations in plant communities have been assessed by calculating N:P ratios of the plant material (Koerselman and Meuleman 1996, Tessier and Raynal 2003, Gsewell 2004). Additionally, N:P ratios can be utilized to determine which nutrient (N or P) may limit decomposition (Gsewell and Freeman 2005, Gsewell and Verhoeven 2006). In addition to N and P ratios being good predictors of potential decay, the lignin concentrations, lignin:N a nd lignin:P ratios of plant litter materi al can also be useful indices to predict decay rates (Meentemeye r 1978, Melillo et al. 1982, Aert s and deCaluwe 1997). A large portion of the C stored in emerge nt macrophytes is found in recalcitrant material (i.e, lignin) that is not easily degraded (Mann 1988). In respons e, the mass loss of litter during the decomposition process is often correlated with secondary comp ounds like cellulose and lignin or lignin:N and :P ratios (Vitousek et al. 1994, Aerts and deCaluwe 1997, Rejmankova and Houdkova 2006). Isolated, closed wetlands systems (i.e., depressional wetlands) primarily receive water and nutrients from precipitation (Bedford et al. 1999, Battle and Golladay 2007). Frequently these wetland systems are nutrient-limited and tend to have minimal abiotic influences (i.e., waves,

PAGE 125

125 currents) that limited physical fragmentation of litte r material. It comes to reason that if these wetland systems are nutrient-limited, the vegetation community in these wetlands would also be nutrient limited. As a result, th e litter quality of the vegetation would be poor due to influences of the nutrient-limited environment. Additiona lly, these nutrient-lim ited environments depend on the decomposition of litter materi al for availability of nutrients (Aerts and deCaluwe 1997). In this study, we aimed to compare the decomposability of litter material of Cladium jamaicense and Typha domingensis across a chronosequence of re stored wetland communities in the Everglades National Park (ENP). Typha domingensis is the dominant plant species found in restored wetlands and C. jamaicense is the dominant native plant species. The primary objectives were to: 1) determine initial litter quality differences between these two species, 2) determine if site conditions had controls on rates of decomposition (restored verses native wetlands), 3) evaluate potential litter quality effects on decomposition ra tes between species and sites, 4) determine the regeneration of vital nutrients (N and P) from the litter material and 5) assess potential microbial limitations on decomposition between species and sites. We tested the these objectives with the investigation of the following hypothesis: H1) Due to the natural oligotrophic (P-l imited) conditions of the ENP, C. jamaicense would have poorer litter quality than T. domingensis and, therefore, would have sl ower rates of decomposition than T. domingensis, H2) The restored wetland communities woul d have greater rates of decay due to elevated nutrient concentrations relative to the native conditions, and H3) the microbial activity associated with the litter material will be greater with C. jamaicense and the native site due to greater nutrient limitations.

PAGE 126

126 Methods Site Description This study was conducted in wetland systems re stored within the Hole-in-the-Donut region of the Everglades National Park. Past farming and management practices in the areas that were restored left these systems open to invasion by Schinus terebinthifolius (Brazilian pepper). The nutrient enriched soil, higher elevation (result ing in short hydroperiods) and subtropical conditions of Florida made these disturbe d areas an ideal lo cation for invasion by S. terebinthifolius. The natural surrounding marl prairie wetlands are inundated for approximately six months of the summer season (Figure 5-1). The goal of the restoration of the HID was to remove the enriched soil and lower the el evation to increase the hydroperiod to control S. terebinthifolius re-invasion (see Chapter 1 for a more detailed site description). Decomposition Experiment We used an in situ litterbag decomposition experiment to estimate mass loss and nutrient regeneration potentials of C. jamaicense and T. domingensis. Sites included in the study were the surrounding native marl prairie grassland co mmunities of the ENP and the sites restored in 1989, 1997, and 2003. In December 2005, we collected live above-g round, live roots, senescent, and litter material for C. jamaicense and T. domingensis to determine partitioning of N and P in each plant component. For the litter bags, re cently senescent material that was still attach ed to the plant and standing was collected for C. jamaicense and T. domingensis. C. jamaicense material was only collected from the native communities were it dominated. T. domingensis material was collected throughout the 1997 and 2003 restored communities wher e it dominated. The litter material was dried at 30C for 48 hours and cut into 6 cm segments for placement in bags. Each bag was 10 x

PAGE 127

127 10 cm in size and made with 1-mm mesh fibergla ss screen. Each bag co ntained 5 g dry weight of either C. jamaicense or T. domingensis. In each of the fours sites, three 4 m2 plots were established for a total of 12 plots. The litter bags were placed on top of the soil substrate within each 4 m2 plots to be collected in triplicate at times 6, 12, 24, and 52 weeks e qualing 72 litter bags per sample period. Upon sampling the bags were stored at 4C until analysis In the laboratory, any soil and root material was separated from the remaining litter material. The wet weight plus mesh bag was recorded before any analysis was performed. Within 24 hours of collection, 1.0 g wet weight was separated for extra-cellular enzyme activities (EEA) and 1.0 g wet weight was separated for microbial biomass N and C analysis (MBN and M BC). The remaining litter was then reweighed, washed, frozen and freeze dried to remove all moisture. After drying, the litter material collected from each bag as well as the initial li tter collected was weighed to determine remaining litter mass dry weight. Once dry, the litter was passed through a Wiley Mill tissue grinde r equipped with a 1-mm mesh sc reen to achieve homogeneity. A subsample was ball ground to a fine powder for to tal C, N and P analysis. All litter samples were analyzed for total C and N with a Ther mo Electron Corp. Flash EA 1112 Series NC Soil Analyzer. Total P was determined via HCl ash extraction and analyzed with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 119-A rev3) (Anderson 197 6). Nitrogen and P ratios were calculated on a mass basis as N:P. Litter Fractionation Initial litter samples of each species as well as all li tterbags collected at time 168 and 365 days were analyzed for litter quality. A sequential extraction method was used to determine soluble cellular content (s ugars, carbohydrates, lipids, etc.), hemi-cellulose, -cellulose, and

PAGE 128

128 lignin (Rowland and Roberts 1994). The coarse ly ground litter material (Wiley Mill, 1-mm mesh screen) was weighed at 1.0 g each and sealed into filter bags that were extracted with neutral detergents to remove soluble cellular contents followe d by an acid detergent extraction for removal of hemi-cellulose using a Ankom A200 Fiber Analyzer. To remove -cellulose a 72% H2SO4 regent was used leaving residual litter containing lignin and ash. The residual litter was combusted at 550C for 4 hours to determine as h content. Each plan t fraction was reported as mg g-1 litter. Assessment of Litter Quality We classified C. jamaicense and T. domingensis as high or low quality based on percent lignin, lignin:N, lignin:P ratios, C:N and C:P ratios and decomposition rate constants. Rates of N mineralization or immobilization are highly influenced by lignin c ontent and C:N ratios (Figure 5-2) (Brady and Weil 1999). Litt er with high lignin content a nd high C:N ratios are considered poor in quality and would have slow rates of decomposition. Lignin contents of 20-25% and C:N ratios greater than 30 would be considered high (poor quality) (Brady and Weil 1999). Differing amounts of lignin in combination with varying C:N ratios can effect rates of N regeneration from litter material. Figure 4-2 illu strates the combined effects of lignin and C:N content and whether or not we can predict net immobilization or net mineralization of N. To determine a decomposition constant, k, for C. jamaicense and T. domingensis we assumed an exponential rate of mass loss for bot h species. The following equation was utilized to calculate k; Mf = Mi e-kt (5-1) where Mi is the initial mass of the litter, Mf is the final mass of the litter and t is the time at Mf. The decomposition constant, k, was calculated fo r both litter materials at each site. The mean

PAGE 129

129 residence time, or time required for the litter to decompose under steady state, was calculated as 1/k (Chapin et al. 2002). Microbial Biomass and Extracellular Enzyme Activities Microbial biomass C and N (MBC and MBN) were extracted from each litter sample during all sample times and microbial biomass P (MBP) was extracted at 365 days only with a chloroform fumigation-extraction incubation method (Brookes et al. 1985). Both the chloroform fumigated and non-fumigated sample s were extracted with 0.5M K2SO4 solution for MBC and N and a NaHCO3 solution for MBP. To determine MBC, each sample was analyzed for TOC with a Schimadzu TOC-5050A Total Or ganic Carbon Analyzer equippe d with an ASI-5000A autosampler. To determine MBN, each sample extract was digested for TKN via kjeldahl digestion (Brookes et al. 1985) and MBP ex tracts were digested via H2SO4 via block digestion. Both digests were analyzed for N and P content wi th a Seal AQ2+ Automated Discrete Analyzer (TKN: EPA Method 111-A rev1, and TP: EPA 11 9-Arev3). The MBC, MBN, and MBP were computed as the difference between the fumigate d and non-fumigated samples. No efficiency correction factors were used in th e calculation of this data. Enzyme activities were performed via a common fluorometric method (Hoppe 1993). Fluorescent artificial substrates methyl-umbelliferone (MUF)--D-glucoside, MUF-phosphate, and L-leucine-7-amino-4-methylcoumarin we re used to determine the EEA of -1,4-glucosidase (GA; EC 3.2.1.21), alkaline phospha tase (APA; EC 3.1.3.1), and L-leucine-aminopeptidase (LLAA; EC 3.4.11.1), respectively. Sample suspensions were made by homogenizing 1.0 g of litter material in 50 ml of ultra pure water with a tissue shredder. Each sample suspension was dispensed in triplicate in white, 96-well microplates by pipetting 100 l of sample and 100 l of the appropriate substrate in each

PAGE 130

130 well. Each plate was allowed to incubate in the dark at room te mperature for 2 hours. This time was determined by conducting an enzyme kinetic te st on each litter type with each substrate. After 2 hours, the plates were read with a Bio-Tek FL600 fluorometric plate reader with excitation wavelengths of 360/40 and emission wavelengths of 460/40. Enzymes activities were expressed as g MUF g-1 litter hr-1 for GA and APA and g AMC g-1 litter hr-1 for L-LAA. Each enzyme activity was normalized for MBC. Statistical Analysis All data were analyzed statistically usi ng Fit Model in JMP Version 5.1 (SAS 2005). Analysis of variance (ANOVA) was performed to i nvestigate site and species differences related to decomposition rates, litter qua lity and microbial activity. Re gressions were performed to determine relationships between variables. Results Decomposition Rates Cladium jamaicense and T. domingensis had similar rates of decomposability (Table 5-1). Differences in decomposition constants were s een between sites, but were minimal between species. Both species had the highest k in the 1997 site and the lowest in the 1989 site, 1.5.05 and 0.76.21 yr-1, respectively. The residence time of each litter was also similar between species but varied across sites. The 1997 site had the lowest and the 1989 site had the highest residence time at 0.65 and 1.31 years, respectively. The seasonal changes in hydrology had an impact on the rate of mass loss across all sites (Figure 5-3). The sites were dr y for the first two sampling periods of this study. We observed little change in the mass of either species duri ng this time period. Within two weeks following day 84 sample time, the wet season began and all sites were inundated with water.

PAGE 131

131 Litter Quality We assessed initial litter quality based on per cent lignin, C:N and C:P ratios (Table 5-1). We found no significant differences between C. jamaicense and T. domingensis for %Lignin and C:N, but there was signifi cantly higher C:P ratio for T. domingensis compare to C. jamaicense (p<0.0001). Both C. jamaicense and T. domingensis had low lignin content at 6.7 and 4.1% and high C:N ratios of 75.5 and 84.6, respectivel y. Based on these results, the litter for C. jamaicense and T. domingensis has a mid level quality rating and indicates that N would be immobilized initially to decrease the C:N ratio of the litter. Once the C:N ratio has decreased to a ratio more suitable for microbial decompos ition, net mineralization will occur. The initial C:N, C:P, and N:P ratios varied considerable for both C. jamaicense and T. domingensis for each plant component analyzed (Figure 5-4). No significant differences were found for C:N ratios between C. jamaicense or T. domingensis for any of the plant components analyzed. Cladium jamaicense had significantly higher C:P and N:P ratios in all plant parts compared to T. domingensis except for senescent ma terial (p<0.0001). There was no significant difference observed in the initial litter fr actionation of soluble cellular content, hemi-cellulose, -cellulose and lignin between C. jamaicense and T. domingensis (Figure 5-4a). After 168 and 365 days, a significant differe nce was found in the litter fractionation between species and sites (Figure 5-5b and 5-5c). Additionally, we found soluble cellular content for T. domingensis was on average higher th an the content found in C. jamaicense litter (p<0.0001 for both times) (Table 5-2). However, the lignin content of T. domingensis was found to be significantly less than C. jamaicense (p<0.0001 for both times) (Table 5-2). No signifi cant differences were found for hemi-cellulose or -cellulose content between species after 168 or 365 days of decomposition. For comparisons of each species

PAGE 132

132 between sites, a significant diffe rence was found between sites for the hemi-cellulose fraction of C. jamaicense at 168 days (p=0.0166), but no signi ficant differences were found for any fractions of C. jamaicense at 365 days. For T. domingensis, a significant difference was observed for all fractions of litter quality between sites at both 168 and 365 days of decomposition (Table 5-2). In addition to differences between species a nd sites, significant differences between the two time periods were investigated. For C. jamaicense, a significant change in all litter fractions between the two time periods was observed (Table 5-2). There wa s a significant increase in both soluble cellular content a nd lignin and significantly less hemi-cellulose and -cellulose at 365 days for C. jamaicense (Figure 5-5b and 5-5d). For T. domingensis, a significant change in all litter fractions except hemi-cellulose between th e two time periods was found (Table 5-2). There was a significant increase in both soluble ce llular content and ligni n and significantly less cellulose at 365 days for T. domingensis (Figure 5-5c and 5-5e). Relationships between lignin, N, and P as an indi cator of litter quality were evaluated. In addition to the lignin content, th e Lignin:N ratio can be a good pr edictor of the recalcitrance of the litter material. Both C. jamaicense and T. domingensis followed the same trends with respect to Lignin:N ratios. An inverse relationship be tween the N content and the Lignin:N ratio was observed (r2=0.37) (Figure 5-6a) and a positive rela tionship between lignin content and the Lignin:N ratio was found (r2=0.80) (Figure 5-6b). We also calculated Lignin:P ratios and investigated relationships between this ratio and lignin and P c ontent for each species. With respect to Lignin:P ratios, C. jamaicense and T. domingensis behaved differently. No relationship was observed between the Lignin:P ratio and the lignin content for each species (Figure 5-7b). For C. jamaicense there was an inverse relations hip between P content and the

PAGE 133

133 Lignin:P ratio (r2=0.62) (Figure 5-7a). An inve rse relationship was observed for T. domingensis as well, but the relationship was weaker (r2=0.39) (Figure 5-7a). The C:N, C:P and N:P ratios were significantly less for T. domingensis than C. jamaicense in most sites after 365 days (Figure 5-8 and Tabl e 5-3). The N:P ratio was the only ratio that resulted in a significant difference between sites as well as between species (Table 5-3). The C:N ratios of C. jamaicense after 365 days was similar to the initial C:N ratio determined, whereas the C:N ratio of T. domingensis was half the initial ratio (F igures 5-4a, 5-8a and 5-8b). The final C:P ratio of C. jamaicense was three times less than the initial and the C:P ratio of T. domingensis was half the initial ratio (Figures 5-4b 5-8c and 5-8d). The final N:P ratio for C. jamaicense was as much as five times less than the in itial ratio (Figures 5-4c and 5-8e). The final N:P ratio of T. domingensis was the same as the initial in th e native site and half the initial ratio in the 2003 site (F igures 5-4c and 5-8f). Cladium jamaicense mineralized both N and P for all times sampled during this decomposition study (Figure 5-9a and 5-9b). Typha domingensis immobilized N for the first 168 days of decomposition, but by 365 days N was mine ralized from this litte r material (Figure 59a). Phosphorus was immobilized in T. domingensis litter throughout the year long study (Figure 5-9b). To investigate potential contro ls on nutrient mineralization or immobilization, regressions between the change in N or P with their respec tive Lignin:N or :P ratio were performed. A positive relationship was observed between the change in N and the Lignin:N ratio (p=0.71; Figure 5-10). This relati onship was the same for both C. jamaicense and T. domingensis; however, T. domingensis had a lower range of data than C. jamaicense. The relationship between change in P and Lignin:P rati o was different for both species with T. domingensis

PAGE 134

134 having a lower range than C. jamaicense (Figure 5-11). Both species had a weak positive relationship, T. domingensis with an r2=0.35 and C. jamaicense with an r2=0.42. Microbial Activity The microbial biomass (MBC, N, and P) asso ciated with the litt er material varied considerably between species and sites during the decomposition study (Table 5-4). In addition, there was a considerable increase in MBC and N with time. For example, the MBC associated with C. jamaicense was 2815 mg kg-1 at 42 days and 18,739 mg kg-1 at 365 days in the native community (Table 5-4). This trend was found for C. jamaicense in all sites for both MBC and MBN. Typha domingensis litter material had a similar trend for MBC and MBN in the native site (Table 5-4); however, in the restored sites, the MBC and MBN associated with T. domingensis litter increased considerab ly by 168 days with an addi tional increase by 365 days. The MBP associated with C. jamaicense was similar for the native and 1989 sites (54 and 56 mg kg-1, respectively), whereas, it was almost double in the 1997 and 2003 sites (91 and 111 mg kg1, respectively) (Table 5-4). Significant differences were observed for MBC and MBN betw een species during all time periods sampled (Table 5-5). No significant differences we re observed between sites or site*species interactions for MBC and MBN except during the we t season (time period 168 days) (Table 5-5). However, no signi ficant differences we re observed at 168 days for MBC:N ratios between species, sites or site*species interactions. The MBP was significantly different between sites and species but had no site*species interaction effects (Table 5-5). For the enzyme GA no significant differences we re found between sites for C. jamaicense or T. domingensis for the first two sample periods of 48 and 84 days; however, there was a significant difference between sites for the two species after 168 and 365 days of decomposition

PAGE 135

135 (Table 5-6 and Figure 5-12). Significant differences in GA were observed between both species at each sample period (Table 5-6, Figure 5-12a and b). Site and species interactions were not observed for GA. We found significant difference between sites for L-LAA at both 84 and 168 days, but not at 48 and 365 days (Table 5-6). Significantly less L-LAA was found in association with T. domingensis litter material as compared to C. jamaicense litter at 42 and 84 days, but no significance was observed between species at 168 and 365 days (Table 5-6 and Figure 5-12c and d). Site and species interactions were only found to be significan t at day 84 (p=0.0005) (Table 56). Enzyme APA was only determined at 365 da ys of decomposition. We found no significant difference between sites or species (p=0.3021 a nd 0.9232, respectively) (Table 5-6 and Figure 513). Additionally no site and species interac tions were found (p=0.4548) (Table 5-6). Discussion The decomposition rates of C. jamaicense and T. domingensis did not differ from each other even though there were significant differences in litter quality indices of each species. This suggests that decomposition of these two species was not controlled by differences in species litter quality indices investigated. While di fferences were not found between species, we observed variation in decay coeffi cients between sites (Table 5-1) suggesting that there may be site characteristics (i.e., nutrien t availability, moisture, and microbial community) that alter rates of decay equally between litter types. Both species had rapi ds rates of decay in the 1997 and 2003 sites, however, slower rates were observed in the native communities and the 1989 site. In a study comparing C. jamaicense and T. latifolia, they found that thes e two litter types had significantly different decay coefficients even unde r the same soil characteristics (Corstanje et al. 2006). One notable difference between this study and ours was the initial N:P ratios of each

PAGE 136

136 litter type. Corstanje et al (2006) found initial litter N: P ratios to be 12 and 26 for T. latifolia and C. jamaicense, respectively. These ratios are consid erably lower than the ratios of 90 and 290 for C. jamaicense and T. domingensis, respectively, found in the litter material of our study suggesting that the P-limitation in the litter fr om our study had significa nt controls on nutrient regeneration but not decay coefficients. Based on similarities in initial litter quality of soluble cellular content, hemi-cellulose, cellulose, lignin (Figure 5-5a) and initial C:N ratio s (Figure 5-4a), we expected to see similar patterns in N regeneration between C. jamaicense and T. domingensis. However, the decomposition patterns of N were different in these two species. The microbial community associated with C. jamaicense mineralized N steadily during the decomposition period suggesting that N did not limit the decay of C. jamaicense (Figure 5-9a). However, T. domingensis litter immobilized N during the first 168 days of the decomposition study and mineralization of organic N did not occur until da y 365 (Figure 5-9a). Furthermore, the amount of N mineralized from T. domingensis at day 365 was approximately 3.5 times less than what was mineralized from C. jamaicense. The significant differences obs erved between the initial C:P and N:P ratios suggested that T. domingensis would immobilized P at much faster rates than C. jamaicense (Figure 5-4b and 54c). The initial C:P ratio for C. jamaicense was approximately 5 times less than the C:P ratio of T. domingensis and the initial N:P ratio was 3 times less (Figure 5-1). Surpri singly, P was found to be mineralized from C. jamaicense at every time period duri ng the decomposition study, whereas P was found to be immobilized by T. domingensis through out this st udy (Figure 5-9b). No differences were found for either species in te rms of N or P regeneration between sites.

PAGE 137

137 These findings on N and P indicate that secondary compound content of C. jamaicense and T. domingensis have some controls over N or P regene ration regardless of nutrient content. An increase in N or P content is usually attributed to immobilization by the microbes associated with the litter material because the material that is being decomposed is limited in either N or P content (Gsewell and Freeman 2005). Secondary compounds such as lignin have been shown to limit decomposition rates (Webster and Benf ield 1986) and in turn could limit nutrient regeneration. Lignin is more recalcitrant that other plant compounds and is often found to increase over time (Webster and Benfield 1986). The differences found in Lignin:N and :P ratios of C. jamaicense and T. domingensis did not affect the decomposition rate of each litter type. While di fferences were found in nutrient regeneration of both N and P for each litter type, an opposite relati onship between N and P regeneration and the Lignin:N a nd :P ratios were observed. Cladium jamaicense had the highest Lignin:N and :P ratios (more r ecalcitrant material) which suggests that N and P would more likely be immobilized compared to T. domingensis. Conversely, we found the opposite to be true. We expected that the Ligni n:N ratios would be ne gatively related to the change in N, but a positive relationship was observed (Figure 5-10). As a result, more N was mineralized (regenerated) from C. jamaicense than T. domingensis (Figure 5-9a). The lack of organic N mineralization from litter material that appeared to be N-limited due to litter quality indices has been observed in other studies (H arris et al. 1995, Scheffer and Ae rts 2000, Corstanje et al. 2006, Gsewell and Verhoeven 2006). This implies that l itter quality indices ar e a poor predictor of N mineralization. With increases in P content th e litter quality of each species increases (Figure 4-7a). Again, C. jamaicense and T. domingensis formed two distinct groups; however, each species

PAGE 138

138 relationship followed difference curves. In addi tion, no relationships were observed between the Lignin:P ratios and the lignin cont ent (Figure 5-7b). These results give further support that a Plimitation in the litter could have signif icant controls on nutrient regeneration. Typha domingensis had significantly lower Ligni n:P ratios as compared to C. jamaicense. As was the case with the change in N, we anticipated that the Lignin:P ratio would be negatively correlated with the change in P, a lthough the opposite was observed. We had hypothesized that T. domingensis would mineralize more P than C. jamaicense due to higher initi al quality, but the opposite was found (Figure 5-9b). Not only was the initial nut rient quality of C. jamaicense greater, P was consistently mineralized from this litter material throughout the study. It has been shown the P rarely limits deco mposition in field experiments even with extremely limiting N:P ratios (Aerts et al. 2001, Aerts et al. 2003, Gsewell and Verhoeven 2006). Additionally, it has been shown that P doe s not limit decay or re sult in immobilization until P concentrations of 0.3 mg g-1 have been reach (Xu and Hirata 2005, Gsewell and Verhoeven 2006). In this study, the P concentrations of both th ese litter types were well below this limit. The initial P content of C. jamaicense and T. domingensis was 0.10 and 0.02 mg g-1, respectively. While the apparent P-limitations of the litter material of C. jamaicense and T. domingensis did not limit rates of decay, they did a ffect the ability of both N and P to be mineralized from T. domingensis. Final nutrient ratios of C. jamaicense and T. domingensis do not offer any conclusions on the similarity in decay rates for these two li tter types. Considerable variability was found between each litter type and be tween sites. In general, C. jamaicense had higher final C:N and C:P ratios as compared to T. domingensis (Figure 5-8). Cladium jamaicense had much higher final N:P ratios in the native site but had similar ratios to T. domingensis in all other sites. This

PAGE 139

139 difference in final nutrient ratios implies that these tw o litter types would have different decay coefficients not only across site s but between species. We had hypothesized that due to the natural oligotrophic (P-limite d) conditions of the ENP, C. jamaicense would have poorer litter quality than T. domingensis and, therefore, would have sl ower rates of decomposition than T. domingensis. This hypothesis was not confirmed by th e initial nutrient or cellular content analysis. However, the final litter material of C. jamaicense is indeed of poorer nutrient quality. In a comparable study, litter decomposition was found to be limited by P only at N:P ratios greater than 22, but when litter had lower nutrien t ratios, decomposition c ould be either N or P limited (Gsewell and Freeman 2005). In addition, they found that establishing critical N:P ratios could be difficult due to threshold differences among speci es and that species-specific critical N:P ratios could be hi ghly influenced by growth conditions. This suggests that the site differences in which C. jamaicense and T. domingensis are grown in could have considerable controls on rates of decompositi on rather than nutrient ratios al one. Furthermore, since both initial litter types are extremely P-limited, the effects of this limitation could have equal weight on the rates of decomposition. Enzyme production is often the limiting step in litter decomposition in aquatic systems (Crst 1991). In this study, th e initial nutrient anal ysis of both litter types suggested that T. domingensis would result in N mineralization as compared to C. jamaicense, however, the opposite was found. By investigating the -1,4-glucosidase (GA) and L-leucineaminopeptidase (L-LLA) activities, we found signi ficantly greater activities of GA associated with the litter of C. jamaicense as compared to the litter of T. domingensis during the first six months of decomposition and a greater association of L-LLA through the first 84 days. After six months the L-LLA production with each litter type was the same. While differences in L-LLA

PAGE 140

140 production was not found between species at six months, the litter ma terial placed in the native community has significantly higher L-LLA production compared to the other sites. External environmental factors ha ve a greater impact on L-LLA production as compared to internal litter quality. Other studies have also demonstrated external environmental controlled on L-LLA production (Burns and Ryder 2001, Rejmankova and Sirova 2007). Unfortunately, the enzyme activity of alka line phosphatase (APA) was not included until the final sampling at 365 days; therefore, we can only speculate that this activity would also have had a greater association with C. jamaicense over T. domingensis. With greater enzyme activity more N and P would be released from the litter material. The microbial communities associated with C. jamaicense are activity acquiring more nutrients fr om this litter material and as a result more N and P is mineralized. Conclusions Even though C. jamaicense and T. domingensis had different initial nutrient contents and differences in litter quality throughout the decomposition study, their decay rates and coefficients were the same. Regardless of the path each litte r type followed, the end result in terms of organic matter input into each system was equivalent. With regards to nut rient regeneration, C. jamaicense indicated a much greater potential to release N and P for further utilization as compared to T. domingensis regardless of site location and litter quality indices. This indicates that in the restored communities, where T. domingensis is dominating, fewer nutrients would be regenera ted and available for futu re plant or microbial uptake. This has severe implications in te rms of native plant community establishment in restored wetlands. It is clear that T. domingensis has an impact on both N and P cycling which could prohibit native plant communities from co lonizing these areas. Competition studies for

PAGE 141

141 uptake of N and P is needed to determine how C. jamaicense and T. domingensis interact for nutrients to determine what effect T. domingensis litter decay may have. Under native conditions, both the microbial communities and C. jamaicense are thriving under nutrient limited conditions. In response, the microbes produce extra-cellular enzymes to acquire needed nutrients from litter associated with C. jamaicense. A greater understanding of this plant-microbe interaction is needed to ga in more insight to why the microbial communities across all sites are putting more energy into nutrient acquisition fr om a litter source of inferior quality.

PAGE 142

142 Table 5-1. Summary of initial litter quality of Cladium jamaicense and Typha domingensis for microbial decomposition. The decomposition constant, k, was determined from mass loss curves from field decomposition study. The residence time of the litter material at each site was determined as 1/k. Species Lignin % Site1/k (yr) C. jamaicense 6.775.5(11.3)3874(581)Native1.05(0.05)0.96 19890.91(0.07)1.10 19971.54(0.08)0.65 20031.38(0.17)0.72 T. domingensis 4.184.6(12.7)18729(2809)Native1.04(0.01)0.96 19890.76(0.21)1.31 19971.53(0.10)0.65 20031.22(0.03)0.82 k (yr-1) C:NC:P

PAGE 143

143 Table 5-2. Summary of results from a two-way ANOVA test with variables as species, site, and time for the litter fractionation of Cladium jamaicense and Typha domingensis. Source of Variationd.f.F-statp rob. > Fd.f.F-statprob. > F Soluble Cellular Content Species 129.2<0.0001119.1<0.0001 Site C. jamaicense 32.90.068030.80.5240 T. domingensis 323.1<0.000133.40.0483 Time* C. jamaicense 143.9<0.0001 T. domingensis 110.10.0032 Hemi-cellulose Species 11.60.214712.70.1091 Site C. jamaicense 34.80.016632.10.1433 T. domingensis 321.4<0.000134.80.0172 Time* C. jamaicense 124.2<0.0001 T. domingensis 14.00.0532-Cellulose Species10.10.708710.70.4094 Site C. jamaicense 32.70.085330.30.8287 T. domingensis 311.90.000434.10.0281 Time* C. jamaicense 164.4<0.0001 T. domingensis 129.5<0.0001 Lignin Species 151.8<0.0001123.9<0.0001 Site C. jamaicense 30.30.842830.90.4253 T. domingensis 36.60.005434.10.0272 Time* C. jamaicense 1108.9<0.0001 T. domingensis 1108.7<0.0001 *Time is a comparison between analysis of litter at sample periods 168 and 365 days 168 days 365 days

PAGE 144

144 Table 5-3. Summary of result s from a two-way ANOVA test with dependant variables of C:N, C:P, and N:P ratios for the litter material of Cladium jamaicense and Typha domingensis. Source of Variationd.f.F-statprob. > Fd.f.F-statprob. > Fd.f.F-statprob. > F Site32.30.090631.90.143234.20.0137 Species174.8<0.0001121.6<0.0001115.00.0006 Site*Species31.10.339030.90.465831.30.3029 C:N ratio C:P Ratio N:P Ratio

PAGE 145

145 Table 5-4. Summary of microbial biomass carbon (MBC), nitrogen (MBN), and phosphorus (MBP) as well as the ratios of MBC:N, MBC:P, and MBN:P. SpeciesDaysSite MBC:PC. jamaicense 42Native2815(91)370(4) 8(0.3)84Native3713(4)207(20) 19(1.9)168Native3631(84)217(11) 19(0.6)365Native18739(1347)708(41)54.1(2.2)26(0.9)13(0.4)3464219891786(218)238(21) 7(0.4)8419893834(162)215(18) 18(0.8)16819896559(926)433(79) 16(1.6)365198919092(715)850(27)55.5(1.6)22(0.3)15(0.2)3444219973107(278)308(32) 11(1.3)8419973635(208)281(20) 13(0.2)16819979730(194)982(27) 10(0.4)365199726677(1471)1018(114)90.8(5.6)28(2.0)11(1.3)2944220032235(78)240(15) 12(0.9)8420033709(106)267(15) 14(0.4)16820038636(119)681(32) 13(0.7)365200326744(2704)1123(131)111.3(4.5)24(0.6)10(0.8)240T. domingensis 42Native5391(132)438(16) 13(0.3)84Native7917(499)433(5) 18(1.2)168Native5590(952)295(65) 20(1.2)365Native26418(238)1041(56)24.5(2.1)26(1.5)43(2.1)10764219895155(196)327(19) 16(1.5)8419897455(129)347(8) 22(0.3)168198923486(1327)2050(212) 12(1.0)365198929607(4334)1325(155)74.4(4.7)22(1.3)18(1.4)3984219975597(167)359(36) 17(1.8)8419977509(444)366(5) 20(1.0)168199719894(530)1322(43) 15(0.2)365199733068(1642)1065(66)147.8(0.4)31(1.3)7(0.4)2244220035793(253)429(55) 15(2.5)8420036518(240)346(14) 19(1.4)168200319124(1166)1278(106) 16(1.7)365200332171(612)1484(40)150.6(6.9)22(0.4)11(0.4)214 MBP (mg kg-1)M B N : P MBC (mg kg-1) MBN (mg kg-1)M B C : N

PAGE 146

146 Table 5-5. Summary of resu lts from a two-way ANOVA test with dependant variables of microbial biomass (C, N, and P), and microbi al biomass ratios (C:N, N:P, and C:P) for the litter material of Cladium jamaicense and Typha domingensis. Source of VariationdfF-statprob> FdfF-statprob> FdfF-statprob> FdfF-statprob> F MBC Site31.20.334930.60.6437521.5<0.000132.20.1048 Species190.8<0.0001188.6<0.0001171.5<0.000119.70.0041 Site*Species30.80.507630.50.656157.20.001030.30.8564 MBN Site31.90.139931.60.2150319<0.000132.50.0840 Species150.033119.50.0054135.2<0.000116.90.0134 Site*Species30.40.722830.30.789838.40.000430.60.6250 MBP Site36.90.0013 Species14.50.0430 Site*Species30.50.6889 MBC:N Site33.10.044134.40.014451.10.362436.60.0017 Species12.30.1443112.70.001810.010.905410.030.8630 Site*Species30.90.478331.70.201150.40.786730.80.5137 MBN:P Site35.70.0036 Species13.60.0665 Site*Species33.70.0236 MBC:P Site34.80.0081 Species12.70.1108 Site*Species32.80.0580 42 days 84 days 168 days 365 days

PAGE 147

147 Table 5-6. Summary of results from a twoway ANOVA test with dependant variables of GA, L-LAA, and APA for the litter material of Cladium jamaicense and Typha domingensis. Source of Variatio n dfF-statprob> FdfF-statprob> FdfF-statprob> FdfF-statprob> F GA Site 31.90.145832.50.0834525.5<0.000132.80.0033 Species114.90.0006139.0<0.000112.30.029016.30.0184 Site*Species32.40.086831.90.162650.60.619131.20.3174L-LAA Site 31.20.320433.40.036139.40.000231.00.4174 Species128.5<0.0001140.3<0.000114.50.435011.70.2045 Site*Species31.20.341338.80.000532.40.088731.60.2028 APA Site 31.30.3021 Species 10.0090.9232 Site*Species 30.90.4548 84 days 168 days 365 days 42 days

PAGE 148

148 Figure 5-1. Climate data from 2006 obtained from Florida Auto mated Weather Network. A) Hydroperiod reported as groundwater leve l above NAVD 1988 (obt ained from USGS National Water Information System). B) Daily temperature. C) Daily precipitation.

PAGE 149

149 Figure 5-2. Conceptual diagram of temporal patter ns of nitrogen release from litter material of differing quality available for microbial deco mposition. Lignin content greater than 20% and C:N ratios greater than 30 would be considered high in content and be considered poor quality and therefore would be limited in microbial mineralization of nitrogen. High, mid, and low indicated litte r quality rating (modified from Brady and Weil, 1999)

PAGE 150

150 Figure 5-3. Percent mass remaining for Cladium jamaicense and Typha domingensis from time zero to 365 days for each site. A) Native. B) 1989. C) 1997. D) 2003. The grey areas on each graph illustrates when the sites were dry and the white areas illustrates when the sites were inundated with water.

PAGE 151

151 Figure 5-4. Initial nutrient ratios for Cladium jamaicense and Typha domingensis plant compartments. A) C:N. B) C:P. C) N:P.

PAGE 152

152 Figure 5-5. Litter fractionation for initial Cladium jamaicense and Typha domingensis senescent litter and at 168 and 365 days. A) Initial content of Cladium jamaicense and Typha domingensis. B) Cladium jamaicense for each site at 168 days. C) Typha domingensis for each site at 168 days. D) Cladium jamaicense for each site at 365 days. E) Typha domingensis for each site at 365 days.

PAGE 153

153 Figure 5-6. Relationships between litter quality indices of Cladium jamaicense and Typha domingensis. A) N content, d.f.=71, F=52.5, p<0.0001. B) Lignin content to the Lignin:N ratio; d.f.=71, F=232.9, p<0.0001.

PAGE 154

154 Figure 5-7. Relationships between litter quality indices of Cladium jamaicense and Typha domingensis. A) P content to lignin:P ratio; C. jamaicense d.f.=35, F=42.9, p<0.0001, T. domingensis d.f.=35, F=32.5, p<0.0001. B) Lignin content to lignin:P ratio; C. jamaicense d.f.=35, F=0.4, p=0.5273, T. domingensis d.f.=35, F=7.0, p=0.0122.

PAGE 155

155 Figure 5-8. Final ratios after 365 days of decomposition for Cladium jamaicense and Typha domingensis. A) C:N of Cladium jamaicense. B) C:N of Typha domingensis. C) C:P of Cladium jamaicense. D) C:P of Typha domingensis. E) N:P of Cladium jamaicense. F) N:P of Typha domingensis.

PAGE 156

156 Figure 5-9. Change in nutrients for Cladium jamaicense and Typha domingensis for each time period analyzed. A) Nitrogen. B) Phosphorus. Positive numbers above the line represent net mineralizati on of each nutrient and negative numbers below the line represent net immobilization of each nutrient.

PAGE 157

157 Lignin:N Ratio 0102030405060 Change in Nitrogen (mg N g -1 litter) -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05 C. jamaicense T. domingensis r =0.71 Figure 5-10. Relationship between Lignin:N ratio and the change in nitrogen content in the litter material. Points associated with Cladium jamaicense and Typha domingensis are designated; d.f.=83. F=198.6, p<0.0001.

PAGE 158

158 Lignin:P Ratio 01000200030004000500060007000 Change in Phop horus (mg P g-1 litter) -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 C. jamaicense T. domingensis r = 0.42 r = 0.35 Figure 5-11. Relationship between Lignin:P ratio and the change in phophorus content in the litter material. Points associated with Cladium jamaicense and Typha domingensis are designated; C. jamaicense d.f.=41, F=7.51, p=0.0098, T. domingensis d.f.=41, F=21.6, p<0.0001.

PAGE 159

159 Figure 5-12. Enzyme activities associated with Cladium jamaicense and Typha domingensis during decomposition.-glucosidase (GA) of Cladium jamaicense B) glucosidase (GA) of Typha domingensis C) L-leucine-aminopeptidase (L-LLA) of Cladium jamaicense D) L-leucine-aminopeptidase (L-LLA) of Typha domingensis

PAGE 160

160 Species C. jamaicenseT. domingensis APA ( g MUF kg-1 hr-1) 0 20 40 60 80 100 Native 1989 1997 2003 Figure 5-13. Alkaline phosphatase (APA) enzyme activities associated with Cladium jamaicense and Typha domingensis during decomposition.

PAGE 161

161 CHAPTER 6 RETENTION OF 15N IN SOIL AND VEGETATION: SEASONAL VARIATION AND LONG TERM NITROGEN STORAGE IN RESTORED WETLANDS Introduction Hydrologically isolated wetland systems are of ten nitrogen (N) poor and rely on N inputs via precipitation, deposition, a nd internal nutrient cy cling (i.e., minerali zation) (Fennessy and Cronk 2001). Nitrogen plays a major role in the functioning of wetland soils and plant communities (Verhoeven et al. 1996, Olde Venter ink et al. 2002, Picking and Veneman 2004). Therefore it is important to understand seasonal variability an d long term storages of N to adequately assess restorati on success of these types of wetland systems. Internal rates of N transformations are importa nt determinants of nutrient availability in wetland plant communities (Mitsch and Gosselink 2000). The continuous transformations and numerous chemical species of N can complicat e the study of N budgets (White and Howes 1994, Martin and Reddy 1997). Most of these transforma tions involve microbial activities, which may either facilitate access to or compete with plan ts for available N. The latter may be driven by incorporation of N into microbi al biomass (Kuehn et al. 2000, Catovsky et al. 2002, Inubushi and Acquaye 2004) or removal of N from the system via gaseous transformati ons (Firestone et al. 1980, Reddy et al. 1989, Bodelier et al. 1996, Chiu et al. 2004). Additionally, rates of these microbially driven processes can alter long term N storage in various ecosystem pools (i.e., NH4 + and NO3 -). Several studies have determined rates of N transformations and quantified N pools for wetland ecosystems (Hemond 1983, Martin and Re ddy 1997, Oomes et al. 1997, Senzia et al. 2002). The amalgamation of this information into whole system budgets to summarize N cycling and storage has been utilized extensivel y to create ecosystem mass balances which can assist in determining net N retention and lo ss (Martin and Reddy 1997, Oomes et al. 1997).

PAGE 162

162 However, this approach has many shortcomings in that the determination of key N processes that control N transformation carry a significant amount of error a nd variability (White and Howes 1994, Gribsholt et al. 2005). Principally, the de termination of N flux rates is predominantly restricted to laboratory assays, which only estimate rates of transformations. There are several different methods which have been utilized to assess denitrification, mineralization, and nitrification rates (Keeney 1982, Bundy and Meisinger 1994, Hart et al. 1994, Mosier and Klemedtsson 1994, Weaver and Danso 1994) an d different methodologies can result in contradictory rates. While these rates may provide representative estimates for N processes in these ecosystems, when utilized together these estimates can have confounding effects on errors in calculating total ecosystem net N retention and loss. In addition to microbially driven soil N pro cesses, plant communities and individual plant species can have considerable direct and indir ect effects on ecosystem N cycling (Aerts et al. 1999, Engelhardt and Ritchie 2001, Epstein et al. 2001, Ehre nfeld 2003). Community and species level N uptake and use can directly control rates of nutrient flux through competition with microbes for different N species and the turnove r time of N in plant biomass. Plant species may vary in their capacity to uptake NH4 +, NO3 or even amino acids as a source of N (Jackson et al. 1989, Schimel and Chapin 1996, Streeter et al. 2000, Henry and Jefferies 2003a, Schimel and Bennett 2004) and plant species have been shown to vary substant ially in the use and residence time of N in plant biomass. Plants can also in directly affect nutrient mineralization via the effects of their litter quality on microbial deco mposition (Mason and Bryant 1975, Hector et al. 2000, Gartner and Cardon 2004). The use of in situ stable is otope tracers to follow N thr oughout an ecosystem has provided a greater understanding of N processes and long term storages across a multitude of ecosystems

PAGE 163

163 (Hobbie and Chapin 1998, Mulholland et al. 2000, Fry and Smith 2002, Gerzabek et al. 2004, Fry 2006). Through additions of NH4 + or NO3 highly enriched in 15N relative to natural abundance levels can assist in determining rates of plant uptake (Merbach et al. 2000, Dinkelmeyer et al. 2003, Ruckauf et al. 2004, Templer and Dawson 2004), transformations (Mulholland et al. 2000, Tank et al. 2000, Temple r et al. 2003, Gribsholt et al. 2005), and storage (White and Howes 1994, Epstein et al. 2001). The primary objective of this study was to in vestigate the differences in ecosystem N retention and partitioning in restored wetland systems. Many times in the restoration or mitigation of wetland systems the focus is solely on the vegetation community; ecosystem function is often overlooked. The restored wetlands investigated in this study have very different plant community composition and diversity compar ed to the native or desired plant community structure. The native communities are co-dominated by Schoenus nigricans and Cladium jamaicense whereas the restored wetlands are dominated by Typha domingensis This study aimed to 1) determine which soil N pools had the greatest influence on ecos ystem N retention, 2) determine potential influences of both community and species level contributions to ecosystem N retention and 3) construct an N budget to assess the success of ecosystem function restoration. This study was conducted in a chro nosequence of restored wetlands in the Everglades National Park. We hypothesized that 1) the soil pools in th e native community would retain more N over time than the soil pools of the restored wetland communities and 2) that the vegetation community would acquire more N from the soil in the native communities compared to the vegetation in the restored wetlands and therefore would retain more N. The basis of these two hypotheses is due to the native soil having a greater soil depth a nd stability due to lack of

PAGE 164

164 disturbance. The restored wetla nds soil depth is very shallow ranging from 0-2 cm, whereas the native community soil depth ranges from 8-20 cm. Methods Site Description This study was conducted in wetland systems re stored within the Hole-in-the-donut (HID) region of the Everglades National Park (ENP). Past farming and management practices in the areas that were restored left these systems open to invasion by Schinus terebinthifolius (Brazilian pepper). The nutrient enriched soil, higher elevation (resulti ng in short hydroperiods) and subtropical conditions of Florida made these disturbed areas an ideal location for invasion by S. terebinthifolius The natural surrounding marl prairie wetlands are inundated for approximately six months of the summer season (Figure 6-1). The goal of the restoration of the HID was to remove the enriched soil and lower the el evation to increase the hydroperiod and control S. terebinthifolius re-invasion (see Chapter 1 for a mo re detailed site description). Experimental Design In 2006, we examined the recovery of N in field plots in the native community and the 1989, 1997, and 2003 restored wetlands. Two treatmen ts were investigated within three replicate field plots; enriched with 15N and a control, where no enriched 15N was added. Each plot was 4m2 in size and at 0.5 m elevation above sea leve l. We chose plots based on a stratified random design with the following criteria : 1) presence or dominance of Cladium jamaicense 2) presence or dominance of Typha domingensis and 3) presence of both C. jamaicense and T. domingensis Existing data collected by the Everglades Research Group was utilized to locate and establish these plots (O'Hare and Dalrympl e 2003). Plots in the native co mmunities were established in vegetation communities with co-dominance of Schoenus nigricans and C. jamaicense. Plots in the 1989 restored site were established in areas where both C. jamaicense and T. domingensis

PAGE 165

165 were present. Plots established in the 1997 and 2003 were established in areas dominated by T. domingensis The 15N tracer was applied to each field plot as a liquid form of 15NH4Cl (99 atom % 15N, Cambridge Isotopes) at a rate of 650 mg 15N m-2 during the dry season of January 2006. All samples were collected from the center 1-m2 nested in the 4-m2 plot (Figure 5-2). The outer 0.5m2 perimeter served as a buffer around the 1-m2 sample plot to reduce dilution of the applied 15N by the surrounding environment (Hauck et al. 1994). Each field plot received 10 grams of 15NH4Cl (equivalent to 2.6 g N m-2) dissolved in 7.5 liters of DI water. The solution was applied to the soil surface using a hand pumped pressurized sprayer. The solution was appl ied evenly over the entire 4-m2 plot keeping the spray nozzle close to the soil surface to avoid standing vegetation and litter material. Soil Sampling and Analysis Soil and litter samples were collected at 24 hour s after application and at days 42, 84, 168, and 365. Litter samples were collected in re plicates of three by ra ndomly clearing a 15 cm2 area within each 1 m2 plot. Soil samples were collected in rep licates of three from each plot via a 7.6 cm diameter PVC soil core collector to the depth of 5 cm for the native soil and to bedrock in restored wetlands and stored at 4C until laboratory analysis was performed. Before analyses were performed all litter and root material was removed from soil samples. The live roots were separated and kept for further analysis. Bulk de nsity was calculated by determining the moisture content of each core by drying a subsample at 60C until a constant weight was achieved. Soil depth was measured from the top of the soil to the bedrock. Organic matter content was determined as loss on ignition (LOI) by combusting 0.5 g of dry soil at 550C for 4 hours. LOI was calculated as the percent of organi c matter lost after combustion.

PAGE 166

166 Within 24 to 48 hours of sample collection, e ach soil sample was extracted for ammonium (NH4) with K2SO4 (Bundy and Meisinger 1994) and set up for incubation for potentially mineralizable nitrogen (PMN) (or biologically available N) (Keeney 1982, Bundy and Meisinger 1994, White and Reddy 2000). Ammonium and nitr ate were analyzed with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 350.1 for NH4 and EPA Method 353.2 for NO3). A subsample of the K2SO4 extract was digested for total kjelda hl nitrogen (TKN) via kjeldahl block digestion and analyzed by flow injection with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 351.1). Total organic carbon (TOC) was analyzed from the extract with a Shimadzu TOC-5050A Total Organic Carbon An alyzer equipped with a ASI-5000A auto sampler. Total P was determined via HCl ash extraction and anal yzed with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 119-A rev3). Total carb on (TC) and total N were determined by dry combustion with a Thermo Electron Corporatio n Flash EA NC Soil Analyzer. The isotope analysis of solid soil samples was performed via isotope ratio determination with a Thermo Finnigan MAT Delta Plus XL Mass Spectrophoto meter equipped with a Costech Instrument Elemental Analyzer for flash combustion of solid material for N and C analysis. To obtain a crude estimate of potential volatiliz ation rates within each site, in situ 0.1M HCl traps were used to capture NH3 gas. Two liter glass containe rs with an affixed sample cup containing a 50 ml solution of 0.1M HCl was placed in each site. Each sample cup was collected after a 24 hour period and stored as 4C until analysis was performed. The samples were analyzed for ammonia with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 350.1 for NH4). Results were expressed as mg N m-2 d-1. Microbial biomass carbon and nitrogen (MBC and MBN) were extracted from each soil sample via the chloroform fumigation-extraction incubation method (Brookes et al. 1985). Both

PAGE 167

167 the chloroform fumigated and non-fumigate d samples were extracted with 0.5M K2SO4 solution. To determine MBC, each subsample from each extr actant was acidified with concentrated ultrapure H2SO4 and analyzed for TOC with a Schi madzu TOC-5050A Total Organic Carbon Analyzer equipped with a ASI-5000A auto-sampl er. To determine MBN, each sample was digested for TKN via kjeldahl digestion (Brookes et al. 1985) a nd analyzed with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 111 -A rev1). The MBC and MBN were computed as the difference between the fumigated and non-fumigated samples. No correction factors were used in the calculation of this data. The 15N in the soil extractions of NH4, NO3, and TKN (both fumigated and non-fumigated) was determined af ter concentrating the nitrogen in each sample onto an acidified filter paper w ith a diffusion technique (Stark and Hart 1996). The isotope analysis of the filter papers was performed via isotope ratio determination with a Thermo Finnigan MAT Delta Plus XL Mass Spectrophoto meter equipped with a Costech Instrument Elemental Analyzer for flash combustion of solid material for N and C analysis. Vegetation Sampling and Analysis Subsamples of individual sp ecies were collected from each plot 168 days after the application of the 15N tracer. Species were chosen for collection based on dominance in individual plots. The dominant species were: S. nigricans and C. jamaicense in the native community, C. jamaicense and T. domingensis in the 1989 site, and T. domingensis in the 1997 and 2003 sites. Since biomass was destructively ha rvested, this was not sampled at this time so as not to disrupt the final sample time at 365 da ys. Therefore, biomass contribution of these species was estimated as an average of bioma ss contribution determined during the prior years sampling (see Chapter 2) and the sampling at the 365 day. At 365 days after the applica tion of the tracer, above-ground biomass (separated as live and senescent) was determined. Both live and senescent plant shoots within the center 1-m2

PAGE 168

168 plots was clipped at the soil surface. In addi tion to being separated by live and senescent, each biomass sampling was separated into species to determine species contri bution and reported as g dry weight m-2. Litter production was determined within each center 1-m-2 plot by collecting all the litter material laying on the soil surface and reported as g dry weight m-2. Root biomass was determined in each plot by collecting three rep licate 7.6 cm diameter cores (Fennessy and Cronk 2001). The live roots were separated from the so il, washed and dried at 50C and reported as g dry weight m-2. All plant samples collected were analyzed for total C and N via dry combustion with a Thermo Electron Corporation Flash EA NC Soil Analyzer for the bulk above-ground plant tissue. Total P was determined via HCl as h extraction and analyzed with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 119 -A rev3) (Anderson 1976). Carbon, N and P ratios were calculated as C:N, C:P, and N:P. The isotope an alysis of vegetation samples was performed via isotope ratio determination with a Thermo Finnigan MAT Delta Plus XL Mass Spectrophotometer equipped with a Costech In strument Elemental Analyzer for flash combustion of solid material for N and C analysis. Calculations The amount of the 15N tracer recovered in each of the pools analyzed was calculated as: 15Nmixed = (15Nsample 15Nbackground) / (15Ntracer 15Nbackground)*100, (6-1) where 15Nmixed is the atom % of the tr acer recovered in each pool, 15Nsample is the atom % of 15N in each sample, 15Nbackground is the natural abundance level of 15N (0.37 atom %), and 15Ntracer is the amount of 15N in the tracer applied (99 atom %) (F ry 2006). The indivi dual total nitrogen pools (TNpool) were calculated as mg N m-2, and the pool size of the 15N tracer applied recovered in the total N pool was calculated as: 15Npool = TNpool 15Nmixed, (6-2)

PAGE 169

169 where the 15Npool is the pool size of 15N recovered in total N pool as mg 15N m-2. The percent of 15Ntracer recovered was calculated as: %15N recovered = 15Npool (mg 15N m-2) / 650 (mg 15N m-2) 100, (6-3) The amount of 15N in the microbial biomass pool was calculated with the following equation: MB15N = (MBNfum*atom % 15Nfum / 100) (MBNnonfum*atom % 15Nnonfum / 100), (6-4) where MB15N is the amount of 15N stored in the microbial biomass N pool, atom % 15Nfum is the amount of 15N in the fumigated extractions, and the atom % 15Nnonfum is the amount of 15N in the non-fumigated control extractions (Templer et al. 2003). Statistical Analysis All data were analyzed statistically usi ng Fit Model in JMP Version 5.1 (SAS 2005). Analysis of variance (ANOVA) was performed to i nvestigate site and species differences related to the recovery of the 15N tracer applied. Regressions were performed to determine if any strong relationships existed between variables. Results 15N Retention after 24 Hours A summary of the site characte ristics before the initiation of the tracer study is found in Table 6-1 and a summary of the prior vegetation a nd litter characteristics are found in Table 6-2. Estimates of 15N retention in the bulk soil and litter layer pools after 24hr s were the highest in the native community at 79%, whereas the in the restored sites 15N retention was 46, 40, and 50% in the 1989, 1997, and 2003 sites, respectively (Table 6-3) The majority of the 15N retained at this time was found in the bulk soil. Of the bulk so il parameters that we analyzed, the MBN pool contained the majority of the 15N at 31, 24, 22, and 25% for the native, 1989, 1997, and 2003

PAGE 170

170 sites, respectively. The percentage of 15N retained in the NO3 pool was 24%, 18, 9 and 1% in the native, 1989, 1997, and 2003 sites, respectively (Table 6-3). In all sites except the 2003 site the percentage recovered in the NO3 pool was greater than what was recovered in the NH4 pool. Soil 15N Retention after 365 Days Patterns of N retention were gr eatly influenced by the restora tion process. After 365 days, the native community ecosystem retained 42% of the initial 650 mg 15N m-2 that was applied as compared to the restored wetland systems which retained 26, 13, and 29% in the 1989, 1997, and 2003 sites, respectively (Table 6-4). The majority of the 15N recovered at this time period was found in the bulk soil pool. Th e organic nitrogen (ON) pool retained the majority of the 15N found in the bulk soil at 22, 8, 4, and 16% in the native, 1989, 1997, and 2003 sites respectively. The MBN pool followed the ON pool in the next largest in percent recovered (Table 6-4). The 2003 restored sites litter layer retained more 15N at 4% over all other sites with 1% in the native site, and 2% in the 1989 and 1997 sites. The 1997 sites bulk soil pool reta ined significantly less of the 15N tracer than the native communities, but the 1989 and 2003 was not signif icantly different than either pool (p=0.098,level = 0.1; Figure 6-3a). The litte r layer in the 2003 site retained significantly more of the tracer as compared to all other s ites (p=0.03; Figure 6-3b). The bulk soil pool was separated into to four groups: NH4 +, NO3 -, ON, and MBN. The ON pool retained more of the tracer in all sites as compared to the other pools (Figure 6-4). The ON in the native community and the 2003 restored wetland retained significantly more 15N over the 1989 and 1997 restored wetlands (p=0.027; Figure 6-4c). The ON pool was followed by the MBN pool retaining a significant amount of the 15N tracer. The MBN in the native and 1989 sites retained significantly more 15N than the 1997 and 2003 sites (p=0.039; Figure 6-4d). The

PAGE 171

171 amount of 15N retained in the NH4 + and NO3 pools was small compared to all other pools. The amount of the tracer remaining in the NH4 + pool ranges from 0.05 to 0.5 mg 15N m-2 with no significant differences found between s ites (p=0.32; Figure 6-4a). The NO3 pool ranges from 0.2 to 2.4 mg 15N m-2 with the native community being signif icantly higher than all the restored communities (p<0.0001; Figure 6-4b). Soil Processes To assess microbial activity associated with mineralization of N, the potentially mineralizable N (PMN) was determined as mg NH4 + kg-1 soil day-1 (Figure 6-5). At all time periods except days 84 and 365, the native site has significantly lower PMN compared to the restored sites. At 84 days the native site PMN was not significantly diffe rent from the 2003 site and at 365 days the native site PMN was not signif icantly different form the 1997 or 2003 sites. The results of the volatilizati on study indicate that the 1997 s ite has significantly higher potential rates of NH3 + volatilization compared to all ot her sites (p=0.0154; Figure 6-6). Additionally, the native community has significantly lower rates of NH3 + volatilization than the restored wetland systems. Species-level 15N Retention By 168 days, selected plant species had in corporated between 4-7% of the 650 mg 15N m-2 that was applied. Both total N and 15N storage of the above-ground biomass was significantly more than the below-ground biomass for all spec ies across all sites (p<0 .0001; Figure 6-7, Table 6-5). The above-ground biomass N pool was significantly higher for T. domingensis in the 1997 and 2003 restored sites than all other species (p<0.0001; Figure 6-7a, Table 6-5). The aboveground N pool of S. nigricans was significantly higher than the N in C. jamaicense (p<0.0001; Figure 6-7a, Table 6-5). There was no significant difference in the amount of N in the belowground biomass between species (p=0.6699; Figure 6-7a, Table 6-5). The above-ground biomass

PAGE 172

172 of T. domingensis in the 1997 and S. nigricans in the native community retained significantly more of the tracer as compared to C. jamaicense and T. domingensis in all the other sites (p=0.0116; Figure 6-7b, Table 65). The below-ground biomass 15N retention of T. domingensis in the 2003 site was significantly less than the S. nigricans and the C. jamaicense found in the 1989 site but not to any of the ot her species in other sites (p=0. 0275; Figure 6-7b, Table 6-5). No significant differences were found be tween other species in other sites for 15N retention in the below-ground biomass pool. A summary of biomass (g m-2), total N (mg m-2), 15N (mg m-2) and %15N for all the individual plant species found at 365 days after application of the tra cer is listed in Table 6-6. The species within each site ar e ordered from highest to lowe st in biomass production. In general, the species with the highest biomass production also had the highest total N and 15N retention. Exceptions to this trend were th e group of unidentified species in the 1989 site, Sarcostemma clausum and Hypericum myrtifolium in the 1997 site, and Solidago sempervirens in the 2003 site (Table 6-6). Of all these species only S. sempervirens contributed significantly to the above-ground biomass pool following T. domingensis in biomass production in the 2003 site. Typha domingensis produced almost twice the biomass and slightly more total N than S sempervirens, but S. sempervirens retained approximately four times more 15N (Table 6-6). To take a closer look at sp ecies and site differences, th e dominant species as well as Poaceae (because it occurred in large numbers in all sites) were plotted for total N and 15N pools (Figure 6-8). We found that S. nigricans and T. domingensis (1997 site only) had significantly higher N pools than the other species (p=0.0001; Figure 6-8a, Table 6-5). Cladium jamaicense in the native and 1989 site and T. domingensis in the 2003 site had similar N pools and were

PAGE 173

173 significantly higher than Poaceae. There was no significant difference observed for Poaceae between sites. Schoenus nigricans had significantly higher 15N retention in its biomass as compared to all the other species (p=0.0007; Figure 6-8b, Table 6-5). Cladium jamaicense in the native and 1989 site and T. domingensis in the 1997 site had similar 15N retention but were significantly higher than T. domingensis in the 2003 site and Poaceae. There was no significant difference observed for T. domingensis in the 2003 site and Poaceae between sites. Community-level 15N Retention The above-ground biomass pool retained a greater perc entage of the 15N tracer than the below-ground biomass pool (T able 6-4). The amount of 15N recovered in the above-ground biomass was 5.5, 3.1, 3.8, and 4.7% for the na tive, 1989, 1997, and 2003 sites, respectively. The below-ground biomass only retained 0.8, 1.3, 1.1, and 1.5% of the 15N tracer in the native, 1989, 1997, and 2003 sites, respectively. We found no significant diffe rences for total N or 15N pools for the aboveor belowground community level biomass across sites (Figure 6-9, Table 6-5). When comparing the total above-ground pool (live plus senescent) to the to tal below-ground pool (live plus dead roots), we found the above-ground biomass N was significantly higher than the below-ground biomass N in the native and 2003 sites (p= 0.0250 and 0.0326, resp ectively; Table 6-5) but not in the 1989 and 1997 sites (p=0.5326 and 0.0976; Figu re 6-9a, Table 6-5). For 15N retention we found a significant difference between the abovea nd below-ground biomass in the native site (p=0.0298; Table 6-5) but not in any of the other sites (Figure 69b). To assess relationships in 15N retention, the bulk soil 15N retained in each site was regressed with the 15N retained in the biomass and found a positive relationship (r2 = 0.41; Figure 6-10). When the bulk soil 15N was the greatest, so was the 15N retention in the above-ground biomass.

PAGE 174

174 In addition to investigating differences between aboveand belowground vegetation, the differences between live, senescent, litter, and root plan t fractions were compared (Figure 6-11). We found no significant difference for N pools between plant fractions except for the 2003 restored site where the litter layer N pool was si gnificantly higher than the live pool but neither were significantly different from the senescent or litter pool (p = 0.0197; Figure 6-11a, Table 65). We found no signifi cant differences for 15N retention between the pl ant fractions (Figure 611b, Table 6-5). Seasonal Change in 15N Storage and N Budget Little change occurred in the bulk soil N pool over time except in the native community. There was a significant decrease in the amount of N stored in the native community soil during the wet season (168 days sample period) than fo r all other sample periods (Figure 6-12a). Additionally, the native commun ity bulk soil stored significan tly higher amounts of N as compared to all the restored sites followed by the 1989 restored wetland. The ON pool followed a similar trend as the bulk soil pool (Figure 6-12 e). The amount of N stored in the 1997 and 2003 were not significantly differently. In the native and 1989 sites, the NO3 pool resulted in an initial increase followed by a sharp decline during the wet season (Figure 6-12b). In all sites the NH4 + pool gradually decreased over time with the 1989 site indicating an increase during the wet season sampling period (F igure 6-12c). In all sites, the MBN pool increased during the dry season (fir st three sample period) with the native community resulting in the highest increase followed by a sharp declin e in the wet season (Figure 6-12d). The litter N pool resulted in the greatest va riability between sites (Figure 6-12f). The 2003 site had the largest N pool and the 1989 sites had the lowest The initial N was similar for the 1997 and native community; however, the 1997 site N was larger than the native community. All sites had an initial decline in N in the litter pool followed by and increa sed as time moved into the wet

PAGE 175

175 season. In addition, for all sites and all parameters except NH4, the final N pool is similar to the initial pool. In all pools except for the ON pool, the amount of 15N recovered decreased over time (Figure 6-13). Some variation, however, wa s observed in some sites where increases in 15N occurred in later sample periods. The bulk soil pool in the 2003 site had an increase in 15N after day 42 with day 1 and 84 having the same amount of 15N (Figure 6-13a). Additionally, the native and 1989 sites litter layer resulted in a slig ht increase in 15N at the 84 day sample period (Figure 6-13d). These variations could be a result in uneven distribution of the tracer during application; however, th e increase observed in the 2003 site bulk soil is significant suggesting translocation of 15N into bulk soil litter or vegeta tion exudates. The amount of 15N in the ON pool increased in the first 42 to 84 days and then decreased over time (Figure 6-13e). In native community, we observed and increase in the ON 15N pool in day 365 compared to what was observed at 168 days. A comple te budget of N pool sizes and 15N retention in each pool for each site at 365 days is found in Figure 6-14. Discussion Soil 15N Retention Our hypothesis that the native community soil pools would retain more N over time than the restored wetland communities was supported by this study. Soil 15N retention in the native community was 2-6% greater than the 15N retention in the restored communities with large differences observed in soil 15N retention across sites. The seasonal variation in 15N retention in the native community bulk soil indicates an increase in 15N after 365 days. This increase observed could be a result in un even distribution of the tracer during applica tion, organic N mineralization from new litter material or translocation of 15N into bulk soil litt er or vegetation exudates.

PAGE 176

176 Several potential losses could al so contribute to the unaccounted 15N applied in our study. Possible gaseous fluxes that c ould result in losses of the 15N tracer include volatilization of NH3, nitrification/denitr ification of NO, N2O and N2 and possibly leaching. Since the tracer was applied during the dry season, de nitrification was limited to anaerobic microsites and therefore was most likely limited until the wet season began. Furthermore, since these soils are calcareous marl wet soils with average pH values at 7.7, ther e is an increased potential for elevated rates of volatilization to occur. We found that the 1997 site had the pote ntial to lose si gnificantly more N via volatilization as compared to the other sites (Figure 6-6). Why this site has significantly higher rates of volatilization th an the other sites is still unknown. All sites had similar pH, moisture content, and bulk density values and were exposed to the same climatic environment (Table 6-1) all of which are fact ors that can result in elevated volatilization rates (Martin and Reddy 1997). As a result of high rates of poten tial loss in conjunction w ith a dilution effect, both the native and the restored wetl and soils retained significantly less 15N as compared to other studies. Recovery rates as high as 38% afte r 7 years were found in a salt water tidal marsh (White and Howes 1994) and rates between 62-75 % were found after 300 days for a forested system (Templer et al. 2005), wher eas in this study, after 365 da ys we only recovered 13-42% of the 15N applied. Studies have shown that the microbial community can immobi lize a considerable amount of 15N tracers that are applied (Schimel 1988, Temp ler et al. 2003). Our results demonstrates that the microbial biomass communities greatly influence soil N retention in these restored wetland ecosystems via immobilization. After 24 hours of application, the microbial biomass had immobilized on average 25-30% of the 15N that was applied in these systems. In addition to microbial immobilization, a c onsiderable amount of the 15NH4 + was nitrified to 15NO3 in the

PAGE 177

177 native and 1989 sites within the first 24 hours (approximately 17-24%) demonstrating that nitrification rates are high duri ng the dry season for these two site s (Table 6-3). This can have considerable consequences on long term retentio n of N in wetland systems. Nitrate is highly susceptible to leaching and denitrification; th erefore N could be permanently lost from the system. Conversely, either rate s of nitrification were not as high in the 1997 and 2003 sites or there was significantly more loss via leaching and deni trification in these two sites. In fact, more 15N remained in the 15NH4 + pool than what was nitrified and stored in the 15NO3 pool for the 2003 restored wetland. In the 1997 site, the elev ated rates of volatili zation could also be contributing to this difference observed in potential nitrification. From estimations of N mineralization, we f ound that in general the restored sites had greater levels of PMN activity (Figure 6-5). Th is indicates that the microbial communities are actively acquiring more N in the restored wetlands than in the native communities. This is an indication that N is more limiting in the restored wetlands and in response the microbes need to mineralize more N to meet their demands. This would result in more N being transformed to labile forms and increase potential lo sses of the N from the system. Plant 15N Retention Our hypothesis that plant communities in the native site will acquire more 15N from the soil than plant communities in the restored sy stems is not fully suppor ted. Community level biomass does not have a significant influence on total ecosystem N retention. While plant community level biomass production was different among all sites (Table 6-6), the amount of N and 15N retained in the community level biomass wa s not significantly diffe rent (Figure 6-8). However, individual plant species are evidently important, since species level differences were found between S. nigricans, C. jamaicense, and T. domingensis all three of which are dominant

PAGE 178

178 species in their respective sites. This indicates that shifts in species dominance could result in changes in overall N retention in the system. In addition to differences quantified betw een species, we found some evidence of differences within a single species at differe nt sites. At 168 days after application of 15N, T. domingensis present in the 1997 site acquire d approximately 4 times more 15N than T. domingensis present in the 1989 or 2003 site indicating site influences on 15N uptake at the species level. However, this di fference was not as apparent for T. domingensis in the 1997 and 2003 sites after 365 days. Additionally, the amount of N and 15N found in T. domingensis was significantly less during the dry se ason (365 days) as compared to the wet season (168 days). The wet season is when wetland plants are most activ e. It comes to reason that more N would be incorporated into biomass during th is time period due to increased availability and plant growth. In the 2003 site, T. domingensis is the dominant plant species; however, it has retained four times less 15N into biomass than S. sempervirens which contributes appr oximately 50% less in biomass production. Furthermore, T. domingensis in the 2003 sites produced 40% less aboveground biomass compared to T. domingensis in the 1997 site. Consequently, the amount of 15N retained in T. domingensis in the 2003 site was 60% less th an in the 1997 restored wetland (Table 6-6). Evidently, T. domingensis present in the 2003 site is not as good of a competitor for N as it is in the 1997 site. While the 1997 and 2003 sites contain several shared species, the relative abundance of ea ch is different. The two most no ticeable differences are the abundance of S. sempervirens and Baccharis spp. which contribute 28 and 8% to the above-ground biomass, respectively (Table 6-6). The amount of biomass produced by S. sempervirens in the 2003 site is 99% greater than that in the 1997 site; additionally Baccharis spp. is not present in the 1997 site. Therefore, T. domingensis dominating the 1997 site is not co mpeting for N against these two

PAGE 179

179 species. In the 2003 site, however, both S. sempervirens and Baccharis spp. have retained more 15N into biomass compared to T. domingensis indicating they are out competing T. domingensis for N. This behavior was not observed for C. jamaicense and S. nigricans Both species had similar amount of N during both the wet and dry season. Furthermore, C. jamaicense acquired comparable amounts of 15N in both the native and 1989 sites. This indicates th at the 1989 site plant community is shifting to a functionally similar plant comm unity found in the native site. Not only is C. jamaicense the dominant species in the 1989 site, no signs of T. domingensis was found at the end of this study (day 365) (Table 6-6). Typha domingensis was present during the wet season sampling but the overall cont ribution to above-ground biomass and 15N retention was small (Figure 6-6). Plant species influence on above-ground biomass 15N retention did not influence community level biomass 15N retention. Plant community structure did not impact community level influences on ecosystem N retention. But rather community level above-ground biomass was influenced by soil 15N retention (Figure 6-10). Severa l studies have reported that plant species richness results in greater ecosystem N retention (Naeem et al. 1994, Tilman et al. 1996). However, in this study we found that the 1997 site had the highest species richness but the lowest ecosystem 15N retention. The 1997 site had the hi ghest above-ground biomass production and biomass N storage above all other sites. The biomass in the 1997 site was approximately 1.5 times greater than the native comm unity, yet we found 1.5 times less 15N retained in the biomass in this site. This implies that species richness had no influence on ecosystem 15N retention. This is further supported by a similar study conducted in grassland ecosystems where they found that

PAGE 180

180 communities with greater functiona l group diversity yielded lower 15N retention than the communities with only one functional group present (Epstein et al. 1998). Conclusions Our results indicate that soil processes have a greater influence on ecosystem N retention than does the plant community composition. The soil in the native community retained significantly more 15N compared to the restored wetland ecosystems. This evidence supported our first hypothesis which stated that the native community soil pools would store more N over time than the restored wetland soil pools. This was not surprising since the restoration technique employed by the ENP to remove all the soil was a severely destructive means of restoration. As a result, the restored wetlands had considerably less N and decreased N availability. Consequently, the microbial activity in the restor ed wetlands was elevated relative to the native community which demonstrated that the microbial biomass communities greatly influence soil N retention in these restored wetland ecosystems. Vegetation community level 15N retention differences were insignificant across sites indicating that re gardless of soil 15N retention the community level vegetation was able to acquire similar amounts of N from the soil. The plant communities inhabiting the restored wetlands benefited from the elev ated microbial activities which resulted in higher levels of mineralized N. Accordingly, the vegetation communities in the restored sites were not more Nlimited than the community in the native site. However, species level effects on 15N retention were significant. The amount of biomass produced by S. sempervirens in the 2003 site is 99% grea ter than that in the 1997 site; additionally Baccharis spp. is not present in the 1997 site. Typha domingensis dominated the 2003 site, however, both S. sempervirens and Baccharis spp. (which cont ributed significantly

PAGE 181

181 less to above-ground biomass produc tion) have retained more 15N into biomass compared to T. domingensis indicating they are out competing T. domingensis for N. There are indications of restoration success in the 1989 restored wetland. Cladium jamaicense acquired comparable amounts of 15N in both the native and 1989 sites. This indicates that the 1989 site plant community is shifting to a functionally similar plant community found in the native site. Additionally, C. jamaicense the dominant species in the 1989 site, and no signs of T. domingensis was found in this site at the end of the study.

PAGE 182

182 Table 6-1. Summary of soil physio -chemical parameters at the star t of the tracer study. (n=10) Parameter AveSEAveSEAveSEAveSE Physical pH 7.6(0.03)7.6(0.02)7.7(0.02)7.6(0.03) Moisture Content (%)46.7(0.40)56.2(0.84)51.4(0.93)40.4(0.82) Bulk Density (g cm-3) 0.4(0.03)0.3(0.02)0.3(0.01)0.3(0.01) Soil Depth (cm) 5.0 2.7(0.13)1.0(0.08)1.3(0.08) LOI (%) 18.5(0.64)23.9(1.01)22.4(0.76)17.0(0.46) Chemical Total C (g kg-1) 155.6(0.68)169.9(2.83)158.2(2.72)151.5(2.13) Total N (g kg-1) 9.5(0.13)10.4(0.28)9.4(0.26)7.8(0.19) Total P (g kg-1) 0.2(0.01)0.6(0.02)0.7(0.05)0.8(0.01) TOC (mg kg-1) 528.6(21.01)858.2(28.09)940.5(63.25)971.8(69.84) Extractable TKN (mg kg-1) 115.1(5.93)184.4(10.33)180.1(14.35)191.3(23.05) NO3-N (mg kg-1) 19.7(0.66)32.8(2.69)68.3(5.36)7.7(0.53) NH4-N (mg kg-1) 25.7(2.81)38.7(7.21)44.8(6.14)79.3(15.96) Native 1989 1997 2003

PAGE 183

183 Table 6-2. Summary of vegetati on community and litter layer chemical parameters at the start of the tracer study. (n=10) Parameter AveSEAveSEAveSEAveSE Above-ground Biomass (g m-2) 253.3(6.8)186.4(8.5)221.9(10.8)136.9(6.1) Total C (g kg-1) 435.6(20.5)431.5(11.6)417.6(17.5)432.9(23.1) Total N (g kg-1) 9.4(1.1)10.6(0.1)11.0(2.2)6.2(0.9) Total P (g kg-1) 0.3(0.01)0.3(0.01)0.4(0.02)0.4(0.02) C:N46.9(6.8)41.3(5.4)39.9(1.0)70.6(9.1) C:P1723.4(103.6)1500.2(79.0)1382.9(54.0)1200.4(49.6) N:P37.7(2.3)35.8(1.6)34.8(1.0)16.8(0.6) Litter Layer Pool Size (g m-2) 96.9(26.5)54.1(4.1)100.4(5.1)157.7(6.6) Total C (g kg-1) 374.1(2.8)363.9(2.5)361.0(2.2)368.1(6.5) Total N (g kg-1) 11.2(0.2)13.3(0.2)12.9(0.1)13.1(0.2) Total P (g kg-1) 0.06(0.003)0.2(0.004)0.2(0.009)0.2(0.008) Native 1989 1997 2003

PAGE 184

184 Table 6-3. Percent of total 15N as NH4Cl recovered after 24 hours of application. (n=3) Parameters (%) Native198919972003 Bulk Soil 73.144.136.441.5 Soil Pools NO3-N 23.517.79.11.1 NH4-N 16.41.44.53.3 MBN31.424.222.225.0 Org-N1.80.80.612.2 Litter Pool5.52.33.58.5 Vegetation Pools*n.d.n.d.n.d.n.d. Percent Recovered78.746.439.950.0 Un-accounted21.353.660.150.0 *n.d. Not determined for 24 hours after application

PAGE 185

185 Table 6-4. Percent of total 15N as NH4Cl recovered after 365 days of application. (n=3) Parameters (%) Native198919972003 Bulk Soil 33.919.65.718.5 Soil Pools NO3-N 0.370.090.020.02 NH4-N 0.070.040.010.08 MBN11.311.51.22.3 Org-N22.28.04.416.1 Litter Pool1.41.61.94.3 Vegetation Pools Above-ground Biomass5.53.13.84.7 Below-ground Biomass0.81.31.11.5 Percent Recovered41.525.612.529.0 Un-accounted58.574.487.571.0

PAGE 186

186 Table 6-5. Summary of result s from a two-way ANOVA test with dependant variables of total nitrogen (N) and 15Nitrogen (15N) pools. Test Variable d.f.F-statP-value Sites Differences Species-level (168 days) Total N Above-ground336.8<0.0001 Below-ground32.90.079015N Above-ground33.80.0389 Below-ground36.90.0066 Community-level (365 days) Total N Above-ground32.40.1429 Below-ground31.40.322115N Above-ground30.70.5564 Below-ground31.20.3624 Abovevs Below-ground Total N Native 312.20.0250 1989 30.50.5326 1997 34.60.0976 2003 310.30.032615N Native 310.90.0298 1989 31.40.3062 1997 37.00.0567 2003 35.20.0855 Plant Fractions Total N Native 30.40.7401 1989 33.80.0582 1997 31.80.2211 2003 35.90.019715N Native 32.80.1055 1989 30.20.9140 1997 32.40.1424 2003 32.20.1718

PAGE 187

187 Table 6-6. A summary of individual species biomass, nitrogen and 15N data for each site. Species are ordered from most to least abundant. (n=3) BiomassTotal NTotal 15N% Applied 15N% Plant Total Species (g m2 )SE(mg m2 )SE (mg m2 )SE% 15 NSE % 15 N Native Schoenus nigricans 153.1(15.3)1033.3(67.4)22.7(2.0)3.50(0.31) 63.8 Cladium jamaicense 78.2(20.3)542.1(142.1)10.6(3.0)1.63(0.46) 29.7 *Poaceae (10) 15.0(6.4)105.2(40.2)2.1(0.8)0.32(0.13) 5.9 Cassytha filiformis 1.5(0.8)10.6(6.1)0.1(0.1)0.02(0.01) 0.3 Centella asiatica 0.6(0.3)6.9(4.0)0.1(0.1)0.02(0.01) 0.3 Total 248.31698.0 35.6 5.48 1989 Cladium jamaicense 69.4(20.4)472.2(123.4)13.37(5.4)2.06(0.83) 65.6 *Poaceae (3) 21.7(1.0)172.2(12.5)2.56(0.1)0.39(0.02) 12.6 Centella asiatica 5.7(0.9)93.6(14.2)1.29(0.1)0.20(0.02) 6.3 Eupatorium capillifolium 5.3(3.1)96.3(55.6)1.31(0.8)0.20(0.12) 6.4 Sagittaria lancifolia 2.7(1.5)31.4(18.1)0.80(0.5)0.12(0.07) 3.9 Sarcostemma clausum 2.2(0.3)36.7(8.7)0.42(0.1)0.06(0.02) 2.0 Ludwigia repens 2.0(0.7)21.1(6.8)0.26(0.1)0.04(0.01) 1.3 Axonopus furcatus 1.6(0.5)8.0(5.3)0.08(0.1)0.01(0.01) 0.4 Unidentified 1.4(0.6)22.3(8.1)0.30(0.2)0.05(0.03) 1.5 Total 111.9 953.8 20.38 3.14 1997 Typha domingensis 199.8(40.4)853.2(176.9)9.8(1.5)1.5(0.23) 39.9 Andropogon spp. 57.9(33.4)397.4(229.5)8.0(4.6)1.2(0.71) 32.9 *Poaceae (3) 23.5(6.4)182.9(49.9)2.3(0.6)0.3(0.09) 9.2 Unidentified 19.9(1.4)145.8(14.3)2.0(0.2)0.3(0.03) 8.0 Sarcostemma clausum 14.0(4.3)180.3(56.9)1.4(0.6)0.2(0.10) 5.6 Sagittaria lancifolia 3.0(1.7)39.8(23.0)0.4(0.2)0.06(0.04) 1.7 Centella asiatica 1.9(1.1)37.0(21.3)0.4(0.2)0.06(0.04) 1.6 Ludwigia peruviana 1.6(0.9)15.1(8.7)0.07(0.04)0.01(0.01) 0.3 Ludwigia repens 1.2(0.3)10.3(2.5)0.08(0.02)0.01(0.003) 0.3 Hypericum myrtifolium 0.8(0.5)5.0(2.9)0.03(0.02)0.01(0.003) 0.1 Axonopus furcatus 0.7(0.2)6.7(1.4)0.03(0.005)0.005(0.001) 0.1 Solidago sempervirens 0.4(0.2)3.8(2.2)0.03(0.02)0.004(0.003) 0.1 Hydrocotyle umbellata 0.2(0.1)2.1(1.2)0.01(0.01)0.002(0.001)0.05 Eupatorium capillifolium 0.2(0.1)2.2(1.2)0.02(0.01)0.003(0.001)0.07 Total 325.01881.5 24.5 3.8 2003 Typha domingensis 126.7(23.0)543.5(95.4)4.0(0.7)0.62(0.1) 13.4 Solidago sempervirens 78.6(14.9)484.4(86.0)15.6(4.2)2.40(0.6) 51.7 *Poaceae (2) 35.4(13.2)266.3(72.1)4.4(1.2)0.68(0.2) 14.6 Baccharis spp. 21.6(1.5)131.1(2.5)4.3(0.8)0.66(0.1) 14.2 Ludwigia peruviana 5.2(3.0)50.5(29.2)0.2(0.1)0.02(0.01) 0.5 Ludwigia repens 4.9(2.5)49.3(23.9)0.8(0.4)0.12(0.06) 2.5 Sagittaria lancifolia 3.4(1.9)29.5(17.0)0.4(0.2)0.06(0.03) 1.2 Fuirena breviseta 2.3(0.8)11.7(3.5)0.1(0.06)0.02(0.01) 0.5 Axonopus furcatus 1.7(1.0)12.4(7.2)0.2(0.09)0.02(0.01) 0.5 Mikania scandens 1.4(0.3)14.4(2.4)0.2(0.02)0.03(0.003) 0.6 Juncus megacephalus 0.9(0.5)6.0(3.5)0.06(0.03)0.01(0.01) 0.2 Eupatorium capillifolium 0.4(0.1)3.9(0.3)0.03(0.0004)0.01(0.0001)0.1 Total 282.51603.0 30.2 4.6 *Grasses could only be indentified to the family level. The number in ( ) is the number of difference species found.

PAGE 188

188 Figure 6-1. Hole-in-the-D onut hydroperiod for 2006 reported as groundwater level above NAVD 1988 (obtained from USGS National Water Information System).

PAGE 189

189 1-m2 nested sample plot 2-m 2-m Figure 6-2. Plot layout for th e application and sampling of 15N tracer study.

PAGE 190

190 Figure 6-3. Total amount of 15N applied remaining after 365 days A) Bulk soil pool; d.f.=3, F=2.96, p=0.0975. B) Litter layer pool; d.f.=3, F=3.8, p=0.0300. Amount of 15N remaining from initial 650 mg m-2 applied. (n=3)

PAGE 191

191 Figure 6-4. Total amount of 15N applied remaining after 365 days. A) Ammonium (NH4); d.f.=3, F=1.4. B) Nitrate (NO3); d.f.=3, F=32.2. C) Organic nitrogen (ON); d.f.=3, F=10.8. D) Microbial biomass nitrogen (MBN); d.f.=3, F=4.1. Amount of 15N remaining from initial 650 mg m-2 applied. (n=3)

PAGE 192

192 Figure 6-5. Potentially mineralizab le nitrogen (PMN) reported as mg NH4 kg-1 soil day-1 for each sample period and each site. The lo wer case letters indicate significant difference between sites within a given time periods. Appropriate p-values are given above the figure for each time; 1 d.f.=3, F=4.3; 42 d.f.=3, F=3.8; 84 d.f=3, F=11.8, 168 d.f.=3, F=10.6; 365 d.f.=3, F=6.9. (n=6)

PAGE 193

193 Site Native198919972003 NH3-N (mg m-2 d-1) 0 2 4 6 8 10 12 14 16 a b c b p = 0.0154 Figure 6-6. Estimates of volatilization rates of NH3-N in native and restored wetland communities; d.f.=3, F=5.9. (n=3)

PAGE 194

194 Figure 6-7. Amount of 15N recovered in Cladium jamaicense and Typha domingensis in each site at 168 days after application. A) Sp ecies level nitrogen pool B) Species level 15Nitrogen pool; see Table 65 for statistics. (n=3)

PAGE 195

195 Figure 6-8. Amount of 15N recovered in dominant vegetation sp ecies in each site after 365 days of application. A) Species bioma ss nitrogen pool. B) Species biomass 15Nitrogen pool; d.f.=8, F=16.0. (n=3)

PAGE 196

196 Figure 6-9. Amount of nitrogen stored in above(live plus senescent) and below-ground biomass in each site at 365 days of the study. A) Nitrogen storage. B) 15Nitrogen storage. (n=3)

PAGE 197

197 Bulk Soil 15N (mg 15N m-2) 0 50100150200250300 Above-ground Biomass 15N (mg 15N m-2) 10 15 20 25 30 35 40 45 r = 0.41Native 1989 2003 1997 Figure 6-10. Relationship between the bulk soil 15N retention and the above-ground biomass 15N retention for each site. Values are plotted as averages in order to compare relationship between sites as well as the two parameters; d.f.=3, F=1.4, p=0.3597. (n=3)

PAGE 198

198 Figure 6-11. Amount of nitrogen stored in the vegetation commun ity pools in each site after 365 days. A) Nitrogen pool. B) 15Nitrogen pool. (n=3)

PAGE 199

199 Figure 6-12. Changes in total nitrogen storages during the year long 15N tracer study for each pool. A) Bulk soil nitrog en pool. B) Nitrate (NO3) pool. C) Ammonium (NH4) pool. D) Microbial biomass nitrogen (MBN) pool. E) Soil organic nitrog en (ON) pool. F) Litter layer pool. (n=3)

PAGE 200

200 Figure 6-13. Changes in total 15N nitrogen retention during the year long 15N tracer study for each pool. A) Bulk soil nitrogen pool. B) Nitrate (NO3) pool. C) Ammonium (NH4) pool. D) Microbial biomass nitrogen (MBN) pool. E) So il organic nitrogen (ON) pool. (n=3)

PAGE 201

201 Figure 6-14. Total nitrogen and 15N budget for each component of this study. The number in parenthesis is the amount of 15N retained in each pool fr om the initial 650 mg m-2 applied. A) Native site. B) 1989 s ite. C) 1997 site. D) 2003 site.

PAGE 202

202 CHAPTER 7 VARIATIONS IN THE NAT URAL ABUNDANCE OF 15N IN RESTORED SUBTROPICALWETLAND VEGETATION COMMUNITIES Introduction In most terrestrial systems, nitrogen (N) is the element that most limits plant productivity (Chapin et al. 2002). Many studies have been conducted which show relationships between N content in plants a nd their corresponding 15N value. If we can gain insight on the mechanisms controlling whole-plant and foliar N isotopic co mposition, we will advance our knowledge of plant N acquisition and allocation (Evans 2001). Very little work has been done to date which has investigated th e relationship of 15N values of the vegetation under extremely P-limited conditions. The 15N values of plant communities and individual species has been used to determine plant N assimilation and translocation (Handl ey and Scrimegeour 1997, Evans 2001). Past research has attempted to use 15N signatures of the plants and soil N pools to determine the N source to the plant communities (Fry 2006, Temple r et al. 2007). Additionally, numerous studies have shown that plant 15N values and soil N pools or pro cess rates are strongly correlated (Garten 1993, Jones et al. 2004, Templer et al. 2007) It has been argued that these correlations are found due to the microbial discri mination against the heavier isotope 15N. The microbial community will discriminate against 15N during several soil N processes such as nitrification (Nadelhoffer and Fry 1994, Hgberg 1997), den itrification (Piccolo et al. 1996), and mineralization or decomposition (Nadelhoffer a nd Fry 1994). When discrimination occurs, the amount of 15N in the product is relatively lighter than the amount of 15N in the substrate form which it was formed. As a result, the substr ate becomes enriched relative to the product produced.

PAGE 203

203 Fractionation occurs during N proces ses in which the heavier isotope 15N is discriminated against. Commonly studied soil processes like mineralization of organic N and nitrification of NH4 + will result in fractionation rates of 05 and 15-35 respectively (Robinson 2001). Denitrification cans result in fractionation rates between 28-33 while NH4 + volatilization fractionation rates can be as high as 60 (Robinson 2001). Microbia l assimilation of NH4 + and NO3 can have fractionation rates of 14-20 and 13 respectively. The fractionation NH4 + and NO3 that occurs during plant assimilation range s from 9-18 and 0-19 respectively (Robinson 2001). As a result of the numerous possibilities for fractionation during the cycling of N, it becomes very difficult to us e the natural abundance of 15N as a tracer to determine N sources for plant uptake and assimilation. Therefore, it ha s been suggested that we cannot infer that the isotopic signatures of various ecosystem pools re late to each other (Handley and Scrimegeour 1997). In spite of these shortcomings, we believe we can utilize seasona l and temporal changes in plant 15N values along with changes in soil N pools (i.e. NH4 + and NO3 pools sizes) and what we know about fractionation rates to better understand how shifts in N cycling potentially affect plant N availability and uptake. This paper reports the natural abundance of 15N of the vegetation at the community and species level for restored wetland communities and how they compare to the native undisturbed communities. We also investigated seasonal and temporal variations in 15N values for plant communities and how they relate to changes in soil N pool sizes (i.e. NH4 + and NO3 -). Additionally, we investigated whether natural abundance 15N was driven by a potential Plimitation which persists in the native community and how the alte red nutrient dynamics in the resorted wetlands may contribute to differences in 15N values of the vegetation.

PAGE 204

204 Methods Site Description This study was conducted in wetland systems restored within the Hole-in-the-donut region of the Everglades National Park. Past farming and management practices in the areas that were restored left these systems open to invasion by Schinus terebinthifolius (Brazilian pepper). The nutrient enriched soil, higher elevation (result ing in short hydroperiods) and subtropical conditions of Florida made these disturbe d areas an ideal lo cation for invasion by S. terebinthifolius. The natural surrounding marl prairie we tlands are inundated for approximately six months of the summer season (Figure 7-1). The goal of the restoration of the HID was to remove the enriched soil and lower the el evation to increase the hydroperiod to control S. terebinthifolius re-invasion (see Chapter 1 for a more detailed site description). Soil Analysis In April (dry season) and July (wet seas on) of 2005, soil and vegetation samples were collected with a 7.6 cm diameter PVC core from 10 plots randomly dist ributed throughout sites restored in 1989 (16), 1997 (8), 2001 (4), 2003 (2), and 2004 (1) as well as the surrounding native communities. The number in parentheses i ndicates the number of years since restoration was completed at the time of this study. Elevation was kept constant at 0.5 m to eliminate hydrology differences as a driving fact or of nutrient availability. The soil cores were transported to the laborator y and stored at 4C until analysis. Before analyses were performed all rocks, roots, and li tter material was removed for the soil samples. Within 24 to 48 hours of sample collection, e ach soil sample was extracted for ammonium (NH4 +) and nitrate (NO3 -) with K2SO4 (Bundy and Meisinger 1994) and set up for incubation for potentially mineralizable nitrogen (PMN) (or bi ologically available N) (Keeney 1982, Bundy and Meisinger 1994, White and Reddy 2000). Soil extracts were analyzed via fl ow injection analysis

PAGE 205

205 with a Bran Luebbe Auto Analyzer 3 Digital Colorimeter (EPA Method 350.1). A subsample of the K2SO4 extract was digested for total kjeldahl nitr ogen (TKN) via kjeldahl block digestion and analyzed by flow injection with a Seal AQ2 + Automated Discrete Analyzer (EPA Method 351.1). A subsample of each soil was dried at 60C for 3 days then ground with a ball grinder to a fine powder for total N and P analysis. Dry soil samples were analyzed for total N with a Thermo Electron Corp. Flash EA 1112 Series NC Soil Analyzer. Total P was determined via HCl ash extraction and analyzed with a Seal AQ2 + Automated Discrete Analyzer (EPA Method 119-A rev3) (Anderson 1976). Nitrogen and P ratio s were calculated on a mass basis as N:P. Nitrogen and P were determined for both the composite biomass (co mmunity level) and selected individual plant species (species level) within 10, 1 m2 plots in each site. To determine nutrients in composite biomass, all the vegetation in a 1 m2 plot was cut at the soil surface, separated by live and senescent plant tissue, bu lked and dried at 70C until all moisture was removed. Once dry, all vegetation from each plot was passed through a Wiley Mill tissue grinder equipped with a 2-mm mesh screen to achieve homogeneity. A subsample was ball ground to a fine powder for N and P anal ysis. Individual plant species of Cladium jamaicense, Schoenus nigricans, Typha domingensis, and Sagittaria lancifolia were collected from each site near each plot when available. In some cases these species were not always present. Plant fractions of live, senescence, and litter were co llected for each species. The samples were dried at 70C until all moisture was removed and grou nd with a Wiley Mill ti ssue grinder equipped with a 2-mm mesh screen. A subsample was ba ll ground to a fine powder. All plant samples were analyzed for total N with a Thermo Electron Corp. Flash EA 1112 Series NC Soil Analyzer. Total P was determined via HCl ash extraction and analyzed with a Seal AQ2+

PAGE 206

206 Automated Discrete Analyzer (EPA Method 119-A rev3). Nitrogen and P ratios were calculated on a mass basis as N:P. The isotope analysis of soil and vegetati on samples was performed via isotope ratio determination with a Thermo Finnigan MAT Delta Plus XL Mass Spectrophotometer equipped with a Costech Instrument Elemental Analyzer fo r flash combustion of solid material for N and C analysis. The 15N results were reported in -notation as the deviation from the international standards (air): 15N () = [(Rsample / Rstandard) 1] 1000 (7-1) where Rsample and Rstandard are the 15N/14N ratio of the samples and standard, respectively. Replicate anlaysis of both random soil and vegetation samples resulted in a precision of <0.1 Reference materials from the Na tional Institute of Standards and Technology (NIST) were used for accuracy (ammonium sulfate standards N1 and USGS25) as well as NIST standard material peach. Statistical Analysis All data collected were analyzed statistica lly using Fit Model in JMP Version 5.1 (SAS 2005). Analysis of variance (ANOVA) and regressi ons were performed to investigate site and seasonal relationships in 15N values and plant N, P and N:P ratios and differences in soil NH4 + and NO3 -. Stepwise regressions were performed to determine which variables most significantly explained variability in 15N values of the community level ve getation. Variables included in stepwise regression were soil and plant N, P, N:P ratios, soil extracts of NH4 +, NO3 -, TKN, and PMN. Multiple regressions were performed with the variables determined significant from the stepwise regression.

PAGE 207

207 Results The 15N values of the soil during the wet season ranged from -1 to 3 for all the sites. A relationship was not found betw een the N:P ratios and the 15N values for the soil (r2 = 0.0009; Figure 7-2). However, the native community had a higher N:P ratio compared to the restored wetlands indicating that the native community would be more P-limited. The 15N values for the community level vegetati on during the dry season was similar to the 15N values during the wet season, suggesting little change in N use with seasonal variability (r2 = 0.80; Figure 7-3). Additionally, the native community le vel vegetation had significantly higher 15N values compared to the community level ve getation present in the restored wetlands. When combining all sites together, a relationship between the N:P and 15N values of the community level vegetation in the dry or wet season was not observed (r2 = -0.08 and -0.24, respectively; Figure 7-4). However, when the sites were separated and in vestigated individually, we did find relationships between N:P ratios and 15N values in some of the sites. The 15N of the vegetation community during the wet season in the native, 1989, 1997, and 2004 sites were negatively correlated with N:P ratios (r2 = -0.74, -0.38, -0.60, and -0.31, respectively; Figure 75). During the dry season a relationship between 15N and N:P ratios of the vegetation community of individual sites was only observed in the 2004 site (r2 = 0.77, Figure 7-6). Individual plant species collected during he we t season from each site varied considerably in 15N values (Figure 7-7). The relationship betw een N and P content was the same for most species. Schoenus nigricans (a native plant sp ecies) had the lowest 15N value at -2.6 along with the lowest N (5.3 g kg-1) and P (0.08 g kg-1) content (Table 7-1, Figures 7-7a). Sagittaria lancifolia had the highest 15N value at 1.6 and her highest N and P content at 10.9 and 0.59 g kg-1, respectively (Table 7-1, Figure 7-7a and b). Cladium jamaicense, T. domingensis, and J.

PAGE 208

208 megacephalus all had similar 15N signatures and N content (Tab le 7-1), The P content was lower for C. jamaicense at 0.20 g kg-1 where T. domingensis and J. megacephalus had similar P content at 0.40 and 0.37 g kg-1, respectively (Table 7-1). Poaceae and Andropogon spp. had similar N and P content and similar 15N values (Table 7-1). Ammonium concentrations were significantly lower during the dry season for all sites except the 2004 restored wetland (Figure 7-8a). During the we t season, the 1989, 1997, 2001, and 2003 restored wetlands had significantly higher NH4 + concentrations than the native site, whereas the 2004 site had significantly lower NH4 + concentrations (Figure 7-8a). Nitrate concentrations were non-detecta ble during the wet season. Du ring the dry season, the 1997 and 2001 site had significantly higher NO3 concentration as compared to all other sites (Figure 78b). The native and 1989 site had similar NO3 concentrations and the 2003 and 2004 site had similar NO3 concentrations that were significantly lower than all other sites (Figure 7-8b). TKN concentrations were not significantly differe nt between sites during the dry season (Figure 7-8c). In general, TKN concentrations duri ng the wet season were lower than dry season concentrations. During the wet season the 2004 site had significantly lower TKN concentrations than all other sites except for th e native site (Figure 7-8c). A stepwise regression resulted in plant P and soil NH4 + as the variables that most significantly predicted 15N values of the community level vegetation during the dry season (plant P p-value=0.005, r2=0.89; NH4 + p-value=0.10, r2=0.96). The 15N values of the wet season community level vegetation were most significantly predicted by plant P and soil TKN (plant P p-value=0.04, r2=0.92; TKN p-value=0.07; r2=0.89). Multiple regressions were performed with the variables that were selected from the stepwise regressions. The multiple regressions performed on the dry s eason data resulted in plant P and NH4 + explaining 96% of the

PAGE 209

209 variation in 15N values (Figure 7-9a), while the we t season variables (plant P and TKN) explained 92% of the variation (Figure 7-9b). Discussion The variability in isotopic signatures for both the community level and species level vegetation indicate that plant species within and across sites potentially access different sources of N and their use of N may result in different fractionation rates during N assimilation, transfer, and translocation processes. While differen ces were observed between sites, wet and dry seasonal variation in the community level vegetation 15N values was minimal. This indicates that community level vegetation isotopic signatur e of N was relatively unchanged with changes in hydrology. Since the presence of different chemical species of N (i.e., NH4 + and NO3 -) is controlled by hydrology in seasonally flooded wetlands (Martin and Reddy 1997, Hefting et al. 2004, Troxler Gann and Childers 2006), this indicates that the source of N to the vegetation may remain unchanged throughout the year even though NH4 + and NO3 concentrations may vary. The non-detectable limits of NO3 concentrations during the wet season can be expected due to favorable conditions for denitrification (Figure 7-8). Competition for NO3 in wetlands is high and any NO3 present will quickly be taken up by microbes or plants (Jackson et al. 1989, Olsson and Falkengren-Grerup 2000, Henry and Jefferies 2002). The stepwise regression performed with this data indicated that the 15N signatures of the vegetation communities in this study were not dependent on NO3 concentrations found during the dry season. It has been shown that if nitrification is high then the 15N of the plants will be dependent upon NO3 -, whereas if nitrification is low no relationship between plant 15N and NO3 will be observed (Koba et al. 2003). This indicates that the nitrification rates in the HID may be low and that the microbes may out-compete th e plants for available NO3 -.

PAGE 210

210 Nitrification has been shown to be li mited in wetland systems resulting in NH4 + being a more important source of N to plant communitie s (Hgberg 1997). In this study, we found that the community level vegetation 15N was strongly dependent upon NH4 + during the dry season (r2=0.96, p=0.10), whereas during the wet season, 15N was dependent on TKN (r2=0.92, p=0.04). Mineralization can be limiting in wetland systems due to anaerobic conditions limiting microbial activity (Oomes et al. 1997, Bridgha m and Richardson 2003), therefore during the wet season, NH4 + may be limiting due to decreased mineralization rates and as a result 15N would be less dependent on NH4 + alone. The strong relationship found between 15N and TKN indicates the vegetation may be utilizing organic forms of N during the wet season. Under conditions where inorganic forms of N are limiting, plants have been shown to preferentially take up amino acids as a source of N (Streete r et al. 2000, Henry and Jefferi es 2003b, a, Schimel and Bennett 2004, Weigelt et al. 2005). However, the majority of these studies have been performed in artic, boreal systems with little applicat ion in subtropical systems; therefore limited information is available on the importance of organic fo rms of N in subtropical wetlands. In this study, we found the 15N signature of the community le vel vegetation resulted in a 6 decrease in sites with higher N:P ratios (increased P-limitation). This change in 15N signature was most pronounced between the vege tation community in the 2004 site (highest P availability) and native si te (lowest P availability). Additiona lly, this same relationship between 15N and N:P ratios was found in all sites excep t the 2001 and 2003 sites during the wet season (Figure 7-5). During the dry season, however, this relations hip was only found within the 2004 site (Figure 7-6). Other studi es have also shown that the 15N signature of plant biomass is affected by the availability of P. It has b een observed that red mangroves can result in a 5 decrease in 15N under P-limited conditions as compared to N-limited conditions (McKee et al.

PAGE 211

211 2002), lake macrophytes nave been shown to vary in 15N values by 4 under differing N availability (Jones et al. 2004), and in a comparison of Cladium jamaicense in Typha domingensis across a P nutrient gradient a 8 and 4 increase, respectively, in 15N signature was found when P levels were enri ched (Inglett and Reddy 2006). Increased levels of P can result in in creases in organic N mineralization and NH4 + flux by stimulating the microbial activity associated w ith these processes (White and Reddy 2000). It has been proposed that the process of organic ma tter mineralization will result in the preferential mineralization of the lighter isotope t hus resulting in an increase in the 15N of the remaining organic matter (Fogel and Tuross 1999, Novak et al. 1999). Fr om a previous study, we found that organic N mineralization was higher in the restored sites (s ee Chapter 6) thus potentially resulting in organic matter with increased 15N values and in turn could result in a vegetation community with increased 15N values. However, while this explanation may seem likely, it may only partially contribute the increase in vegetation 15N in the restored sites. The bulk soil N (~95% organic N) only resulted in a ~3 change in 15N within the sites as compared to the vegetation 6 change. Additiona lly, there is no indication that this change in bulk soil 15N was driven by a P-limitation (Figure 7-2). Additional contributions to this shift in community level 15N values could be a result of mechanistic controls of individual species. A limitation in P availability could result in differing demands and use of N by plant species as well as controls on vegetation community structure. Previous research has shown that C. jamaicense is well adapted to and prefers P-limited environments (Newman et al. 1996, Craft a nd Richardson 1997, Richardson et al. 1999), whereas, T. domingensis prefers nutrient enriched environmen ts and will rapi dly take up both N and P under enriched conditions (Miao et al. 2000, Lorenzen et al. 2001). In a previous study,

PAGE 212

212 we found that the nutrient-use efficiency of N (NUE-N) for C. jamaicense and T. domingensis were not significantly different at 0.17.03 and 0.15.02 g biomass mg-1 N, respectively (see Chapter 4). As a result of similar N use, their corresponding 15N values are also similar, 0.31.08 and 0.21.07 respectively (Table 7-1). Schoenus nigricans also had a similar NUE-N at 0.18.02 but had a much lower 15N signature at -2.5.1 In contrast, the NUEN of S. lancifolia is 0.07.03 g biomass mg-1 N (see Chapter 4) and its corresponding 15N signature is 1.56.13 (Table 7-1). These results suggest that species that are less efficient with the use of N (those which require less N per unit of biomass produced) may discriminate less against the heavier isotope than the speci es which have significantly higher NUE-N. Nitrogen and P content also contributes to the 15N signatures of both the community and species level vegetation. At the species level, S. lancifolia had the highest N and P content at 10.9 and 0.59 g kg-1, respectively, and the highest 15N signature at 1.56.13 while S. nigricans had the lowest N and P content at 5.3 and 0.08 g kg-1, respectively, and the lowest 15N signature at -2.5.1 Cladium jamaicense and T. domingensis which had similar 15N signatures also had similar N content at 7.0.2 and 6.6.1 g kg-1, respectively; however the P content of C. jamaicense was 50% less than that of T. domingensis at 0.20.1 as compared to 0.40.2 g kg-1, respectively (Table 7-1). The diffe rence in P content did not affect the 15N signature of either species, indicating that th eir similarities in N c ontent and NUE-N outweigh the limitation of P in C. jamaicense. This was not the case with S. nigricans. Not only did it have lower N content, its P content was ~60% less than C. jamaicense and 80% less than T. domingensis. The potential for vegetation P content to affect 15N signature is further supported by strong relationship of the community level 15N signature with plant P (F igure 7-9). During both

PAGE 213

213 the dry and wet season, plant P content account for 80 and 92% of the va riation in community level 15N. The same stepwise regression indicated that plant N did not significantly contribute to the variability in community level 15N, suggesting that plant P is a more important determinant than plant N. These resu lts on the affects of N and P content on 15N signatures is further supported by a study on C. jamaicense and T. domingensis indicating similar correlations with 15N (Inglett and Reddy 2006). Conclusions The use of natural abundance 15N signatures can be a useful tool in assessing soil N cycling and availability and N uptake in plant communities. Additionally, it is important to investigate both species level 15N values as well as community level. Large differences were observed between the community level and species level 15N values in this study, indicating that individual species will re spond differently to changes in both N and P availability. In this study we found that NH4 + was an important source of N to the vegetation community as compared to NO3 -. Additionally, during the wet season the 15N signatures were dependent on TKN concentrations in the soil indicating that organic forms of N may be a source of N to plant communities under anaerobic conditions Little research has been performed on the importance of organic N as a source of N to plants in subtropical wetlands. Additional research is needed on the preferential uptake of NH4 +, NO3 and organic N to determine the importance of each form of N to plant communities in subtropical wetlands. The availability of P also had some controls over the 15N of the plant communities. At both the species and community levels, the 15N was dependent on the plant P content. Plants inhabiting restored wetlands that were rich in P compared to the native system resulted in higher plant P content and 15N signatures. Additiona lly, plant mechanisms like NUE-N may affect

PAGE 214

214 15N signatures of plants. Plant species like S. nigricans, C. jamaicense and T. domingensis were more efficient with their use of N as compared to S. lancifolia and as a result they had much lower 15N signatures. Little research has been done to compare the NUE-N and the 15N signature of plants to definitively link this mechan ism to N use. Additional research is needed to test this potential mechanistic control on 15N signatures of plant species.

PAGE 215

215 Table 7-1. Chemical and isotopic values fo r selected dominant species from each site. Species Site* AverageSEAverageSEAverageSEAverageSE Cladium jamaicense Native, 19897.0(0.2)0.20(0.1)35.8(2.9)0.31(0.08) Schoenus nigricans Native 5.3(0.05)0.08(0.01)66.2(1.3)-2.51(0.10) Sagittaria lancifolia 1989, 1997, 200310.9(0.4)0.59(0.3)18.4(2.8)1.56(0.13) Typha domingensis 1997, 2003, 20046.6(0.1)0.40(0.2)16.6(1.1)0.21(0.07) Juncus megacephalus 2003, 2004 7.8(0.2)0.37(0.2)20.9(1.0)0.32(0.25) Andropogon spp. 1997, 2001, 20037.6(0.2)0.45(0.3)16.8(3.0)-0.31(0.19) Poaceae All 8.0(0.1)0.44(0.2)18.3(9.5)-0.99(0.03) *all values reported are an average for each species in all sites. No significant differences were found for species present in different sites. N (g kg-1)P (g kg-1)N:P 15N

PAGE 216

216 Figure 7-1. Hole-in-the-D onut hydroperiod for 2005 reported as groundwater level above NAVD 1988 (obtained from USGS National Water Information System).

PAGE 217

217 Figure 7-2. Relationship between 15N values and N:P ratios of th e soil in each restored wetland and the native community. A) All data point s plotted. B) Averages of plots from each site. (n=10 for each site)

PAGE 218

218 Dry Season 15N () -3 -2 -1 0 1 2 3 Wet Season 15N () -2 -1 0 1 2 3 Native 1989 1997 2001 2003 2004 r = 0.80 Figure 7-3. Relationship be tween community level plant 15N values measured during the dry season (April 2005) and those measured during the wet season (late July 2005). (n=10 for each site)

PAGE 219

219 Figure 7-4. Relationship be tween community level plant 15N values and the N:P ratios which correspond to each value. Each site is code d to indicate site relationships. A) Dry season. B) Wet season. (N=60)

PAGE 220

220 Figure 7-5. Relationship between th e wet season community level plant 15N values and the N:P ratios which correspond to each value. A) Native. B) 1989. C) 1997. D) 2001. E) 2003. F) 2004. (N=10)

PAGE 221

221 N:P ratio 101520253035 15N () 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 r = 0.77 Figure 7-6. Relationship between the dry season community level plant 15N values and the N:P ratios which correspond to each value fo r the 2004 restored wetland community. Individual graphs for other sites not s hown due to lack of relationship: native (r2=0.02), 1989 (r2=0.06), 1997 (r2=0.16), 2001 (r2=0.03), and 2003 (r2=0.0018). (N=10)

PAGE 222

222 Figure 7-7. Relationshi p between species level 15N values and nitrog en, phosphorus and N:P ratios of dominant species. A) N Content. B) P content. C) N: P content. (N=10)

PAGE 223

223 Figure 7-8. Inorganic nitrogen in each site in the HID during the wet and dry seasons. A) NH4N; dry season F=2.92 and p=0.02, wet season F=6.29 and p=0.0001. B) NO3-N; dry season F=1.9 and p=0.010. C) TKN; dr y season F=1.07 and p=0.39, wet season F=4.5 and p=0.002. During the wet season NO3-N levels were non-detectable (B). (N=10)

PAGE 224

224 Figure 7-9. Relationship between the community level vegetation 15N to the inorganic nitrogen and phosphorus in each restored site. A) Actual and predicted 15N to NH4-N and plant P during dry season; F=36.9, p=0.008, d.f.=5. B) Actual and predicted 15N to TKN and plant P during wet season; F=17.5, p=0.02, d.f.=5. (N=6)

PAGE 225

225 CHAPTER 8 NUTRIENT MEMORY IN INVADED AB ANDONE D FARMLAND WITHIN THE EVERGLADES NATIONAL PARK: CONSIDERATIONS AND IMPLICATIONS FOR RESTORATION Introduction It has long been understood that anthropogenic alterations of nutrient dynamics can result in shifts in plant community structure. These ecosystem alterations can result in substitution one or more native plant species with an exotic speci es (Ehrenfeld 2003). While the visual effect of the replacement of native plant species by the invasion of exotics speci es is apparent, the consequences of invasion on soil processes is less obvious. Numerous studies have been conducted in natural communities showing the li nks between species composition on ecosystem function (Hector et al. 1999, Diaz and Cabido 2001, Loreau et al. 2001, Catovsky et al. 2002, Hooper et al. 2005), therefore, it can reasonably be deduced that the invasion of exotics will have consequent effects on ecosystem function incl uding but not limited to: nutrient processes, decomposition, storage pools, productiv ity and microbial activities. Ecosystem disturbance is often followed by the invasion of a non-native plant species. The sudden change in the resources available to plan t communities often times result in the loss of biodiversity which weakens the stability of plan t community structure allowing for the ease of invasion (Meekins and McCarthy 2001). Because invasion is not foreseen, most studies are conducted after the invasion has occurred making it difficult to determine the direct cause of initial entry into the sy stem (Wiser et al. 1998). Previous farming in the Hole-in-the-Donut (HID) region of the Everglades National Park (ENP) which altered approximately 4000 ha of continuous natural vegetation and soil has resulted in drastic nutrient dyn amic shifts and aggressive colonizati on by exotic species Schinus terebinthifolius (Dalrymple et al. 2003).

PAGE 226

226Schinus terebinthifolius, Raddi (Brazilian pepper or Flor ida holly) is a Category I most invasive plant species in the state of Florida. It has invaded many di sturbed ecosystems where the soil and vegetation community has been drastic ally altered due to an thropogenic activities. Schinus terebinthifolius is a small escaped ornamental tree (typically 10 feet in height but can reach 40 feet) native to Brazil and Paraguay ( Clark 1997). First introduced in 1840s, it is now abundant in disturbed moist to me sic sites in the southern half of the Florida peninsula. The canopy of S. terebinthifolius forms a closed, dense thicket which excludes native vegetation via shading and chemical inhibition of their growth, and provide relatively poor wildlife habitat (Clark 1997). Schinus terebinthifolius is related to poi sonwood, poison oak, poison ivy, mango, and pistachio, etc. and the berrie s it produces have been found to ha ve narcotic or toxic effects on birds and other wildlife. Schinus terebinthifolius is moderately salt tolerant, w ithstand flooding, fire, drought, and quickly re-sprout after being cut (Clark 1997). It is considered a se rious threat to natural Florida ecosystems and therefore its control and erad ication is critical. Attempt to control S. terebinthifolius in the ENP have included mechanical removal, fire management, herbicide treatments and biological control ag ents all of which have been unsuccessful in this region of the park (Dalrymple et al. 2003). Restoration eff ects led to a cooperative agreement between the ENP and Miami-Dade County via use if mitig ation funds to restore the HID region by implementing a scraping method to remove all vegetation and soil down to the calciumlimestone bedrock. The effect of previous management and inva sion has created an ecosy stem nutrient legacy which could impede ecosystem restoration. The goal of this study was to evaluate the environmental conditions of the invaded ab andoned farmland and how the invasion if Schinus

PAGE 227

227terebinthifolius has potential impacted ecosystem nutrien t dynamics. The objectives of this study were to: 1) determine differences in soil characteristics and re lationships between the invaded, native, and 2003 restored wetland communities, 2) determine the nutrient and cellular fractionation of S. terebinthifolius and the dominant plant species in the native community (Schoenus nigricans) and the 2003 restored wetland (Typha domingensis), and 3) evaluate functional differences in nutrient regeneration an d storage between the invaded, native, and 2003 restored communities. Methods Site Description This study was conducted in wetland systems rest ored within the HID region of the ENP. Past farming and management practices in the areas that were restored left these systems open to invasion by Schinus terebinthifolius (Brazilian pepper). The nut rient enriched soil, higher elevation (resulting in short hydroperiods) and su btropical conditions of Florida made these disturbed areas an ideal location for invasion by S. terebinthifolius. The natural surrounding marl prairie wetlands are inundated for approximately six months of the summer season. The goal of the restoration of the HID was to remove the enriched soil and lower the elevation to increase the hydroperiod to control S. terebinthifolius re-invasion (see Chapter 1 for a more detailed site description). To determine the legacy of invasion and pr evious farming practices within HID the methodology from chapters 4-5 were applied to areas that have yet to be re stored. A summary of the methodology is provided here, however, detail ed methodology is found in previous chapters. Site comparisons will be made between the inva ded and native communities as well as the 2003 restored wetland community.

PAGE 228

228Sample Collection and Analysis In April 2005 (dry season) and July 2005 (wet season), soil samples were collected with a 7.6 cm PVC core from 10 plots randomly dist ributed throughout the s ites The soil cores were transported to the laboratory and stored at 4 C until analysis. Within 24 to 48 hours of sample collection, each soil sample was extracted for ammonium (NH4 +) with K2SO4 (Bundy and Meisinger 1994) and set up for incubation for pote ntially mineralizable N (PMN) (or biologically available N) (Keeney 1982, Bundy and Meisinger 1994, White and Reddy 2000). Ammonium extracts were analyzed via flow injection analys is with a Bran Luebbe Auto Analyzer 3 Digital Colorimeter (EPA Method 350.1). A subsample of each soil was dried at 60C for 3 days then ground with a ball grinder to a fine powder for total N and P analysis. Dry soil samples were analyzed for total N with a Thermo Electron Corp. Flash EA 1112 Series NC Soil Analyzer. Total P was determined via HCl ash extracti on and analyzed with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 119-A rev3) (Anderson 1976). Nitrogen and P ratios were calculated on a mass basis as N:P. To determine nutrient content, random live a nd senescent samples were collected, bulked and dried at 70C until all moisture was removed. Once dry, all vegetation was passed through a Wiley Mill tissue grinder equipped with a 1-mm mesh screen to achieve homogeneity. A subsample was ball ground to a fine powder for N and P analysis. All plant samples were analyzed for total N with a Thermo Electron Corp. Flash EA 1112 Series NC Soil Analyzer. Total P was determined via HCl ash extracti on and analyzed with a Seal AQ2+ Automated Discrete Analyzer (EPA Method 119-A rev3) (Anderson 1976). Nitrogen and P ratios were calculated on a mass basis as N:P.

PAGE 229

229Decomposition Experiment We used the same in situ litterbag decomposition experiment to estimate mass loss and nutrient regeneration potentials Schinus terebinthifolius, Schoenus nigricans and Typha domingensis as described in Chapter 5. All sample s were analyzed for extra-cellular enzyme activities (EEA) and microbial biomass nitr ogen and carbon analysis (MBN and MBC). Additionally, samples were analyzed for total C, N and P analysis and soluble cellular content (sugars, carbohydrates, lipids, etc.), hemi-cellulose, -cellulose, and lignin. See Chapter 4 for detailed description of methodology. Assessment of Litter Quality We classified S. terebinthifolius, S. nigricans, and T. domingensis as high or low quality based on percent lignin, C:N and C:P ratios and a decomposition rate constant. Rates of N mineralization or immobilization are highly influenced by lignin c ontent and C:N ratios (Figure 5-2) (Brady and Weil 1999). Litt er with high lignin content a nd high C:N ratios are considered poor in quality and would have slow rates of decomposition. Lignin contents of 20-25% and C:N ratios greater than 30 would be considered high (poor quality). Differing amounts of lignin in combination with varying C:N ratios can effect rates of N regeneration from litter material. To determine a decomposition constant, k, for S. terebinthifolius, S. nigricans, and T. domingensis we assumed an exponential rate of ma ss loss for both species. The following equation was utilized to calculate k; Mf = Mi e-kt (8-1) where Mi is the initial mass of the litter, Mf is the final mass of the litter and t is the time at Mf. The decomposition constant, k, was determined both litter at each site. Th e mean residence time, or time required for the litter to decompose under stea dy state, was calculate d as 1/k (Chapin et al. 2002).

PAGE 230

23015N Tracer Experimental Design In 2006, we examined the retention of N via the use if 15N stable isotope techniques in three replicate field plots in the invaded and native communiti es and the 2003 restored wetland (see Chapter 6 for methodology and experimental design). Soil and vegetation samples were collected at time 24 hours after application an on days 42, 84, 168, and 365. Samples were analyzed for total N, NH4, NO3, MBN, ON, and 15N in previously mentions pools as well as PMN, and volatilization (see Chapte r 5 for detailed methodology). Statistical Analysis All data collected were analyzed statistica lly using Fit Model in JMP Version 5.1 (SAS 2005). Analysis of variance (ANOVA) was pe rformed to investigat e site and seasonal differences in soil. Regressions were performed to determine relationships between variables. Multiple comparisons were made using the Least Square Means test and to determine Pearsons correlation coefficients between all variables. Results Soil Characteristics Differences were observed between many of the soil physical, chemical and biological parameters analyzed (Table 8-1). A few notable differences include %LOI, NO3 -, carbon (C), N, P, C:P and N:P ratios, and PMN. The invaded sites had higher %LOI (34%) compared to the native community (22%) and the 2003 restored site (16%). Non-detectable limits of NO3 were found in the 2003 and native communities whereas 6.33 g g-1 was detected in the invaded sites. The invaded soil had the highe st levels of C (214 g kg-1) followed by the native community soil (188 g kg-1). Small differences were found between the N content in the invaded, 2003, and native sites, 8.8, 6.2, and 11 g kg-1, respectively. However, the ra nge of P content was large for all three sites; invaded (1.6 g kg-1), 2003 (0.87 g kg-1), and native (0.33 g kg-1). The C:P and N:P

PAGE 231

231 ratios were significantly greater in the native community soil as compare to the invaded and 2003 soil. Finally, the PMN in the soils were higher in the invaded site (20.2 mg kg-1 d-1) as compared to the native (10.3 mg kg-1 d-1) and the 2003 site (14 mg kg-1 d-1). The principal components analysis revealed that there are di fferences among soil environmental parameters in the invaded, na tive, and 2003 restored co mmunities. Principal component 1 described 37% of the variation an d principal component 2 described 35% of the variation (Figure 8-1). This an alysis has factored the native community out from the invaded and 2003 restored site based on soil environmenta l characters. The invaded and 2003 sites are grouped closely together with overlapping error bars indicating that th ere are similarities between the two sites. The C:P and N:P ratios were highly correlated with the native community and primarily responsible for the distinct separa tion (Table 8-2 and Figure 8-2c). Additionally, the N:P ratios were positively correlated with the MBC suggesting that with increases in N:P ratios (as found in the native community) the M BC increases. Additional correlations can be found in Table 8-2. Distinct differences were not visible in soil N content across sites, but the soil P content is clearly greater in the invaded co mmunity as compared to the na tive community (Figure 8-2a and b). The P content in the 2003 sites soil falls between the invaded and native sites concentrations. A similar trend is demonstrated by the PMN data with the invaded site having the greatest PMN activity and the native site the lowest (Figure 8-2d). Vegetation Characteristics The above-ground live, root and senescent tissue N content for S. terebinthifolius was higher than what was found for T. domingensis and S. nigricans whereas the litter layer for all species had similar values for N content (Figure 8-3a). The tissue P content was considerably higher in S. terebinthifolius as compared to the other tw o species (Figure 8-3b). In S.

PAGE 232

232terebinthifolius plant tissue, little differe nce is found in P content for the above-ground live, roots, and senescent whereas the litter layer had approximately 90% less P than the live fraction. For both T. domingensis and S. nigricans, the root contained more P than the above-ground live portion of the plant and the senescing tissue had significantly less P than the live. However, T. domingensis contained greater amounts of P in all plant parts as compared to S. nigricans (Figure 8-3b). The C:P and N:P ratios for S. nigricans was much greater in all pl ant parts than for either T. domingensis or S. terebinthifolius (Figure 8-3d and e). Additionally, both the C:P and N:P ratios of the senescent material for S. nigricans was considerably higher than other plant parts. The cellular fractionation of the senescent pl ant material of each species shows that S. terebinthifolius consists of approximate ly 75% lignin and that T. domingensis and S. nigricans are only 5 and 17% lignin, resp ectively (Figure 8-4a). Ecosystem Functions Typha domingensis and S. nigricans had similar decomposition rates of mass loss (Table 83 and Figure 8-5). The turnover time for each sp ecies was 0.86 and 0.89 years, respectively. The decay rate of S. terebinthifolius was much faster with a turnove r rate of 0.44 years (Table 83 and Figure 8-5). After 365 days of decompos ition the cellular fractionation of each species was very different than the initial. Schinus terebinthifolius litter consisted of approximately 50% soluble cellular content and onl y 30% lignin indicating a 60% lo ss in initial lignin content (Figure 8-4b). A larger portion of the remaining T. domingensis litter consisted of lignin than what was present in the initial anal ysis whereas the fraction of lignin in S. nigricans remained unchanged (Figure 8-4b). The amount of P and N regenera ted via mineralization from S. terebinthifolius was considerably greater than that of T. domingensis and S. nigricans (Figure 8-6a and b). After 365

PAGE 233

233 days of decomposition S. terebinthifolius regenerated approximately 50 mg N g-1 litter material whereas T. domingensis only regenerated about 10 mg N g-1 litter and S. nigricans regenerated 30 mg N g-1 litter (Figure 8-6a). For P regeneration, S. terebinthifolius released 12 mg P g-1 litter and S. nigricans only regenerated about 0.4 mg P g-1 litter (Figure 8-6b). Typha domingensis did not regenerate any P during the 365 da y study, but resulted in 0.4 mg P g-1 litter being immobilized. The GA associated with the decomposition of S. terebinthifolius was similar to that associated with T. domingensis up to the 365 days sampling period. After 365 days, the GA was significantly for S. terebinthifolius than for T. domingensis (Figure 8-7a). Initially, the GA association with S. nigricans was considerably higher than fo r the other two species. After 168 and 365 days the activities were low fo r all species (Figure 8-7a). The L-LAA associated with each litter type was very si milar to the association of GA (Figure 8-7b). Additionally, the activities of L-LAA were significantly lower than the activities of GA. The enzyme activity for APA was only determined at 365 days. The APA activity associated with S. nigricans is significantly greater than that of T. domingensis and S. terebinthifolius (Figure 8-8). The APA associated with S. terebinthifolius was near zero at 365 days of decomposition. The native site retained more N from the 15N tracer study after 24 hours at 72% than either the invaded (50%) or the 2003 rest ored wetland (41%) (Figure 8-9a). At 24 hours, more of the 15N was in the MBN pool compared to the other pools for all three sites. After 365 days the native soil retained 33% of the initial 15N tracer that was applied closely followed by the invaded site at 28% (Figure 8-9b ). The 2003 restored site only retain ed 18% in the soil of the initial 15N applied. After 365 days, most of the 15N was recovered in the ON pool (Figure 8-10a). The MB15N pool was the highest in the native commun ity and the lowest in the 2003 restored

PAGE 234

234 wetland (Figure 8-10b). The amount of 15N retained in the NH4 + pool was small and the same for all sites (Figure 8-10 c). The total amount of 15N remaining in the NO3 pool was small, however, the native community retained significan tly more as compared to the invaded and 2003 sites (Figure 8-10d). Discussion The effects of previous farm ing practices and invasion of Schinus terebinthifolius on ecosystem function has resulted in drastic a lterations of the soil nutrient dynamics. The disturbance from the farming and invasion resulted in highly enriched levels of P relative to the native community. The outcome of increased P con centrations has resulted in a shift away from the historical oligotrophic P-limited cond itions found in the native communities. The elevated levels of P found in the inva ded communities have resulted in high plant tissue P content in S. terebinthifolius as compared the native S. nigricans (Figure 8-3). Aboveground live tissue P content is a pproximately 95% greater in S. terebinthifolius compared to S. nigricans. This difference in tissue content between these species has demonstrated direct effects on the mineralization of P. The amount of P regenerated from S. terebinthifolius is about 97% greater than the amount regenerated from S. nigricans (Figure 8-6). The majority of the studies on invasion and soil processes have fo cused on or have found greatest significance on soil C and N processes, however some herbaceous sp ecies have been shown to result in increases in both soil and plant P (Ehren feld 2003, Vanderhoeven et al. 2006). More comparably, an exotic tree in Hawaii was found to release significant amounts of both N and P during decomposition as compared to the native tree species (Rothstein et al. 2004). This investigation found that the soil P c ontent in the 2003 restored wetland was about 50% less than what was found in the invaded soils and about 60% greater than what is indicative of the native communities (Table 8-1). This indi cates that the restoration process was able to

PAGE 235

235 remove half of the P legacy from farming and invasion. The consequences of the additional P remaining above the native levels could hinder the colonization of nativ e plant communities. Several studies on native plant species of the ENP have indicated that the native plants will not thrive when the natural oligotrophic conditions are altered (Newman et al. 1996, Richardson et al. 1999) leaving the system open to invasion by undesirable species like T. domingensis (Urban et al. 1993, Craft et al. 1995, Craft and Rich ardson 1997). In addition to the destructive restoration technique, the residu al P present in the 2003 site could be contributing to the domination by T. domingensis. The dominance T. domingensis also has long term effects on nutrient availa bility in the restored ecosystems. After 365 days of decomposition, T. domingensis immobilized approximately 5 mg P g-1 litter (Figure 8-6b). The rate of immobilization coupled with the slower decay rates of T. domingensis demonstrate that the residual P could become permanently buried in the soil profile making it unavailable for further uptake. If this pattern in nutrient decomposition continued over a long period of ti me, the nutrient dynamics could become similar to the native oligotrophic P-limited condition ideal for native plant community development. In addition to alterations in P dynamics, there are indications that the legacy of farming and invasion has altered the processes of N cycling. Schinus terebinthifolius litter material is a source of N to the microbial community. Both the above-ground live and senescent plant tissue of S. terebinthifolius is greater than the live and senescent tissue of T. domingensis and S. nigricans (Figure 8-3). As a result, this produced different N regeneration patterns for each species. Both S. terebinthifolius and S. nigricans resulted in N regeneration; however, the contribution from S. terebinthifolius was considerably higher (Figure 8-6). Due to this difference in litter decomposition, more N is made availabl e for further microbial or vegetation uptake.

PAGE 236

236 This effect of invasion on alterations in N rege neration has been demonstrated in other studies. It has been demonstrated that invasive grasses can increase N regeneration as compared to native woodland species (Mack et al. 2001) as well as in several herbaceous species (Ehrenfeld 2003). In contrast, N was immobilized into the litter of T. domingensis during the majority of the decomposition study. Mineralization of N did not occur until the final 6 months of the study (Figure 8-6a). While the total soil N content was not significantly difference in the 2003 site compared to the invaded and native soils, the amount of N stored in the 2003 restored site soil was lower (Table 8-1 and Figure 8-2). This sm all difference, however, could be significant enough to result in N-limitations to the vegetation community growing in the 2003 restored wetland. A previous study indi cated a potential N-limitation to the vegetation community found in the 2003 restored wetland (see chapter 4 for more information on nutrient limitation). In response to an N-limitation, the litter materi al produced by the vegetation found in the 2003 restored wetland would also be N-limited. Therefore, limiting the microbial communities ability to regenerate N. The total N storage capacity of the soil im pacted by farming and invasion was not significantly different from the so il in the native community (Fi gure 8-10e). In contrast, the 2003 stored significantly less N over the duration of the 15N tracer study. This concludes that long term total N storage capabilit ies are not altered by farming or invasion in this area however; the restoration technique has altered total N stor age capacity. While the total N storage capacity has not been affected by farming and invasion, the processes of N transformation are altered. Differences in rates of nitrification are apparent from the comparison of the %15N retained in the NO3 pool across sites (Figures 8-9 and 8-10d). The amount of 15N nitrified to NO3 in the invaded site within the first 24 hours was less than half of what it was in the native community

PAGE 237

237 (Figure 8-9). Therefore, nitrifi cation rates are lower has a result of farming and invasion. In a review of the effects of exotic species on soil processes, select species were shown to result in decreases in nitrification while the majority re sulted in increases in nitrification (Ehrenfeld 2003). Additionally, in grassland systems nitrif ication rates were found to be significantly higher in invaded grasses as compared to na tive woodland areas (M ack and D'Antonio 2003). Nitrification is a microbial medi ated process and changes in ra tes of nitrification indicate that the microbial communities found in the invade communities ha ve also been altered. In addition to altered nitrification rates, the MBC and the enzyme activities responsible for nutrient regeneration and OM decomposition have also been altered in the invaded communities. As a result of increased N and P concentrations in the S. terebinthifolius litter material the amount of enzyme activity associated with decompositi on is significantly lowe r than that for both T. domingensis and S. nigricans (Figure 8-7 and 8-8). With highe r levels of nutrient content the microbial community does not need to put energy into producing enzymes for hydrolysis (Chrst 1991). Conclusions The memory of enriched P lingering in the inva ded sites will be a challenge to overcome. The elevated levels of soil and plant P found in the invaded communities have resulted in high P availability as compare to the oligotrophic c onditions of the native communities. Previous farming and the consequent invasion by S. terebinthifolius has resulted in increase P availability, altered microbial activities and reduced N transformation rates. Additionally, the effect of the restoration method utilized will also carry with it a legacy which the native plant species will have to over come in order to colonize these sites.

PAGE 238

238 Table 8-1. Summary of physical, chemical, and biological pa rameters for the invaded, native and 2003 restored communities. (n=10) Parameter AveSE AveSE AveSE Physical Moisture (%) 38.24(0.65)59.52(0.91)61.19(0.45) Bulk Density (g cm-3) 0.31(0.004)0.30(0.01)0.41(0.01) LOI (%) 34.53(1.87)15.47(0.51)21.99(0.63) Chemical NO3-N (mg kg-1) 6.33(0.47)n.d.* n.d.* NH4-N (mg kg-1) 53.21(3.36)62.53(4.23)33.88(1.22) TKN (mg kg-1) 103.59(9.99)111.36(7.97)64.30(1.97) TOC (mg kg-1) 1818.76(140.09)1185.44(36.19)1069.90(23.37) DON (mg kg-1) 44.05(6.83)48.83(4.05)30.42(1.31) C (g kg-1) 214.52(7.43)152.83(0.89)187.52(1.75) N (g kg-1) 8.77(0.39)6.19(0.14)11.04(0.22) P (g kg-1) 1.59(0.03)0.87(0.02)0.33(0.03) C:N 25.21(0.23)25.48(0.41)17.32(0.19) C:P 153.54(11.80)190.88(6.67)865.09(44.26) N:P 6.33(0.53)7.60(0.27)48.27(2.10) Biological PMN (mg NH4 kg-1 d-1) 20.15(0.93)13.95(1.62)10.29(0.34) MBC (mg kg-1) 3775.79(164.43)6255.62(274.09)5589.40(99.55) MBN (mg kg-1) 502.64(26.78)495.36(19.55)492.03(16.79) MBC:N 8.25(0.28)13.18(0.50)11.92(0.21) *non-detectable limits Invaded 2003 Native

PAGE 239

239 Table 8-2. Correlation coefficients from a principal components analysis on soil envi ronmental characteristics for the invaded native, and 2003 restored wetland communities. The table abbreviati ons are: BD = bulk density, LOI = loss on ignition, TOC = total organic carbon, P = phosphorus, N = n itrogen, C = carbon, PMN = potentially mineralizable nitrogen, and MBC and N = microbial biomass carbon or nitrogen. Boldface indicates a correlation coefficient >0.70. ParameterSiteMoistureBDLOINH4TKNTOCPNCC:NC:PN:PPMNMBCMBN Moisture-0.27 BD-0.500.28 LOI-0.73-0.180.73 NH40.370.37-0.600.46 TKN0.280.32-0.750.610.93 TOC0.15-0.100.140.890.650.81 P0.57-0.64-0.220.280.190.730.16 N-0.590.330.370.750.330.460.60-0.24 C-0.17-0.800.130.980.420.620.900.110.81 C:N0.78-0.63-0.38-0.21-0.16-0.17-0.790.52-0.76-0.28 C:P-0.750.320.21-0.12-0.29-0.15-0.10-0.790.290.29-0.53 N:P-0.820.370.24-0.11-0.29-0.16-0.12-0.800.360.42-0.620.99 PMN0.220.76-0.140.620.750.730.650.300.400.56-0.14-0.25-0.25 MBC0.250.820.140.170.750.670.36-0.340.420.20-0.530.630.990.53 MBN0.120.410.240.650.680.700.70-0.240.630.65-0.35-0.58-0.320.560.67 MBC:N-0.470.41-0.14-0.390.920.82-0.20-0.38-0.14-0.33-0.170.170.17-0.830.33-0.39 239

PAGE 240

240 Table 8-3. Summary of the decomposition constants, k (yr-1), and turnover rates, 1/k (yr), as determined by the mass loss from a field decomposition study. Species AverageSE AverageSE Schinus terebinthifolius 2.31(0.19) 0.44(0.04) Schoenus nigricans 1.12(0.04) 0.89(0.03) Typha domingensis 1.22(0.29) 0.86(0.19) Decompostion Rate (k) Turnover (1/k)

PAGE 241

241 PC1 (37%) -3-2-10123 PC2 (35%) -4 -2 0 2 4 6 Invaded 2003 Native Figure 8-1. Principal component an alysis for soil environmental ch aracteristics in the invaded, native, and 2003 restored wetland communitie s. The first two principal components explained 72% of the total variance in the soil environment (n=20).

PAGE 242

242 Figure 8-2. A comparison of soil nutrients for the invaded, native, and 2003 restored wetland communities. A) Nitrogen content. B) Phosphorus content. C) N:P ratios. D) Potentially mineralizable nitrogen (PMN). (n=10)

PAGE 243

243 Figure 8-3. A comparison of plant tissue chemistry for S. terebinthifolius, T. domingensis, and S. nigricans for the invaded, native, and 2003 re stored wetland communities. A) Nitrogen content. B) Phosphorus content. C) C:N ratios. D) C:P ratios. D) N:P ratios. (n=10)

PAGE 244

244 Figure 8-4. Litter cellular fractionation after 365 days of decomposition for S. terebinthifolius, T. domingensis, and S. nigricans. A) Initial content. B) Final content.

PAGE 245

245 Time (days) 0 100 200 300 400 Mass Remaining (%) 0 20 40 60 80 100 120 Schinus/Invaded Typha/2003 Schoenus/Native Figure 8-5. Percent mass remaining for Schinus terebinthifolius, Typha domingensis, and Schoenus nigricans from time zero to 365 days.

PAGE 246

246 Figure 8-6. Change in nutrients for Schinus terebinthifolius, Typha domingensis, and Schoenus nigricans for each time period analyzed. A) Nitrogen. B) Phosphorus. Positive numbers indicate nutrient mineralization a nd negative numbers indicate nutrient immobilization.

PAGE 247

247 Figure 8-7. Enzyme activities associated with Schinus terebinthifolius, Typha domingensis, and Schoenus nigricans during decomposition. A) -glucosidase (GA). B) L-leucineaminopeptidase (L-LLA).

PAGE 248

248 Time (days) 365 APA ( g MUF kg-1 hr-1) 0.0 0.5 1.0 1.5 40.0 60.0 80.0 100.0 120.0 S. terebinthifolius T. domingensis S. nigricans Figure 8-8. Alkaline phosphatase enzy me activity (APA) associated with Schinus terebinthifolius, Typha domingensis and Schoenus nigricans during decomposition.

PAGE 249

249 Figure 8-9. The percent 15N recovered in each soil pool fr action from the initial 650 mg 15N m-2 that was applied. A) 24 hours. B) 365 days.

PAGE 250

250 Figure 8-10. The amount of 15N recovered after 365 days in e ach soil pool fraction from the initial 650 mg 15N m-2 that was applied. A) Organic 15N pool. B) MB15N pool. C) 15NH4 pool. D) 15NO3 pool. E) Bulk soil 15N pool.

PAGE 251

251 CHAPTER 9 SYNTHESIS AND CONCLUSION One im portant question in the successiona l development of the HID is what are colonization patterns for Cladium jamaicense and Schoenus nigricans and why are they not dominating the restored sites? Schoenus nigricans has not been found to repopulate any of the restored wetland communities. This brings up some interesting life history question on the required conditions in order for S. nigricans to successfully propagate and survive. C. jamaicense, however, is dominant in th e 1989 site and starting to become more abundant in more recently restored sites, indicati ng that with time it w ill inhabit these wetland areas even if the environmental conditions are not re stored to native conditions. Nitrogen and Phosphorus Limitations Drive Community Development The vegetation communities which have developed in the restored wetlands of the HID are very different than the surrounding desired native plant communities. Several factors can control re-vegetation patte rns after disturbances. In this study we evaluated potential nutrient limitations as a control over vegetation community structure and restoration success. The links between diversity and function dur ing the successional development of the wetlands in this study has implications to the management of restored ecosystems. Landscape and watershed alterations can result in severe degradation of wetland systems which result in species compositional changes and loss of biodiversity. Wetland systems are driven predominantly by hydrology and many plant species will respond differently to fluctuations or changes in water level and flow. To maximize species diversity and composition development similar to a native (or reference) system, it is important to understand the factors governing the native system. For this restoration project, the native site is a P-limited system with extremely low levels of N and P. To enhance the potential for native plant species in colonize, the nutrient

PAGE 252

252 rich soil was complete removed to eliminate the effects from previous farming practices. While this restoration method was destructive and labor intensive, the result is the development of herbaceous wetland plant communities that, with time (i.e., the 1989 site), have developed a species composition similar to the native plant comm unities. However, more time is needed for development of ecosystem function. The NMS ordination performed in this study indicated there is a clear differentiation between the restored and native sites in ordination space. The most distinct contributing factor to this difference in ordination analys is is due to the presence of S. nigricans in the native community. The dominance of this plant species in the native community and the lack of its presence in any restored wetla nd system outweigh any similariti es found between the native and 1989 site when analyzed statistically. By comparing the native and 2003 communities (the two extremes), we found that the native communities are P-limited and that the 2003 restored site may be N-limited within the first few years after restoration. The soil and vegetation in the native community indicates a Plimited system (N:P ratios of 48.3 and 37.7, respec tively), whereas the soil and vegetation in the 2003 site indicates a possibl e N-limited system (N:P of 7.6 and 16. 8). It has been suggested that a N:P < 14 results in an N-limitation and a N:P > 16 results in a P-limitation (Koerselman and Meuleman 1996). However, more importantly is the significant diffe rence between the N:P ratios from the native and 2003 sites. The much lower ratio in 2003 site suggests that either more phosphorus and/or less nitrogen are available in the soil. The differences in ratios are a result of the native site having the highest TN and the lowest TP values and the 2003 site having the lowest TN and the highest TP values.

PAGE 253

253 We found that soil N and P cont ent varied considerably within the restored sites but the N:P ratios were less variable. The N:P ratios of the native plant community were two to three times greater than the ratios found in the restor ed communities. Additionally, the soil P and soil N:P ratios exhibited controls on th e plant and plant N:P at both th e community and species level. No such conclusion can be made in terms of soil N. At the vegetation community level, the nativ e plant community has N:P ratios and NUE-P that are two to four times greater than that of the restored plant communities. The species level N:P ratios and NUE-P of the native communities were also two to four greater than the species found in the restored communities. This conclude s that at the community and species level, the native site is more P-limited th an the restored sites. We can say with certainty that a P-limitation is prevalent in the native communities as well as most of the restored communities. Little evidence was found to support a N-limitation in any of the sites. The N:P ratios of the site restored in 2003 imply that it is N-limited not P-limited, however, no other data collected at the community or species level indicates a N-limitation is prevalent. At the community level, we find no differences in the NUE-N between sites and the differences observed in species leve l NUE-N appear to be due to plant traits not a N-limitation. Due to differences in the level of P-limitations in the restored sites verses the native communities, the increased levels of soil P coul d influence the re-vegetation patterns of the restored wetlands. At these higher levels of P, the desired vegetation composition ( Cladium jamaicense and Schoenus nigricans co-dominance) is not achieve d. A greater understanding of the lasting impacts of the residu al P in the restored wetlands is necessary to determine if the desired native vegetation species will inhabit these wetlands with time.

PAGE 254

254 Furthermore, without conducting a N and P fertilizat ion study directly, it is difficult to say with certainty that a Nor P-limitation exists (Vitousek 2004). Over the past decade, several review papers of fertilization studies in terrestrial systems ha ve been conducted to develop critical N:P ratios which allo w us to make inferences about nutrient limitations without conducting a fertilization study (Koerselman and Meuleman 1996, Gsewell and Koerselman 2002, Gsewell 2004). These studies focused primar ily on community level limitations that did not consider multi-species interactions. Therefor e, it is difficult to make conclusions about species level nutrient limitations solely by considering their N:P ratios. The use of additional tools, such as the nutrient-reso rption efficiency (NRE) and nutrien t-use efficiency (NUE), could be used in conjunction with the N:P ratios of in dividual species to potentially determine species level nutrient limitations (Vit ousek 1982, Vitousek 1984, Berendse and Aerts 1987, Aerts and Decaluwe 1994, Feller et al. 2003, Gsewell 2005). Nutrient Regeneration The potential for N and P to be regenerated from the dominant plant species (Cladium jamaicense and Typha domingensis ) found in he HID was investigated via a in situ litter bag decomposition study. Cladium jamaicense and T. domingensis had different initial nutrient contents and differences in litter quality th roughout the decomposition study; however, their decay rates and coefficients were the same. Regard less of the path each litter type followed, the end result in terms of organic matter i nput into each system was the same. In regards to the nutrient regeneration, C. jamaicense indicated a much greater potential to release N and P for further utilization as compared to T. domingensis regardless of site location. This would indicate that in the restored communities where T. domingensis is dominating fewer nutrients would be recycled and available for futu re plant or microbial up take. This has severe implications in terms of native plant community es tablishment in restored wetlands. It is clear

PAGE 255

255 that T. domingensis has an impact on both N and P cycli ng which could prohibit native plant communities from colonizing these areas. Competition studies for uptake of N and P is needed to determine how C. jamaicense and T. domingensis interact for nutrients to determine what effect T. domingensis litter decay may have. Under native conditions, both the microbial communities and C. jamaicense are thriving under nutrient limited conditions. In response th e microbes produce extra-cellular enzymes to acquire needed nutrients from litter associated with C. jamaicense A greater understanding of this plant-microbe interaction is needed to ga in more insight to why the microbial communities across all sites are putting more energy into nutrient acquisition from a litter source of inferior quality. Long Term Nitrogen Storage The results of this study indica ted that soil processes have a greater influence on ecosystem N dynamics than does the plant community composition. With the use of and 15N tracer study, we found that the soil in the native community stores significantly more N compared to the restored wetland ecosystems. This was not su rprising since the restor ation technique employed by the ENP to remove all the soil was a severely destructive means of restoration. As a result, the restored wetlands had considerably less total N and decreased N availability. Consequently, the microbial activity in the restored wetlands was elevated relative to the native community which demonstrated that the microbial bioma ss communities greatly influence soil N dynamics in these restored wetland ecosystems. Vegetation community level N retention di fferences were insignificant across sites indicating that regardless of soil N retention the community le vel vegetation was able to acquire similar amounts of N from the soil. The pl ant communities inhabiting the restored wetlands benefited from the elevated microbial activities whic h resulted in higher levels of mineralized N.

PAGE 256

256 Accordingly, the vegetation communities in the re stored sites were not more N-limited than the community in the native site. Applications to Wetland Restoration and Mitigation It has been suggested that plan ting desired plant species and di versity levels is needed to facilitate vegetation composition development in restored wetland communities (Zedler 1993, Kellogg and Bridgham 2002, Callawa y et al. 2003). In large-s cale restoration applications, planting desired plant species is not always feasible. Therefore it is necessary to gain a better understanding of the natural recruitment of plant species from primary succession and the development of ecosystem function to increase the success of large-scale restoration projects. In this study, we found that natural recruitment would resu lt in increases in species richness with time, and that the species composition would develop similarly to the native community provided enough time has past. We saw an immedi ate recruitment of a diverse plant community consisting of 38 individual plant species within six months of restorat ion. While additional increases in species diversity were slow, the dive rsity did increase significantly after 8 years. Additionally, it took between 8 to 16 years befo re the plant community developed into one representative of the native community. However, the ecosystem functi on was not restored in this time period, indicating that more time is needed for development of native ecosystem function. The species composition changed significantly fr om site to site indicating that the plant communities are very dynamic and unstable. Understanding the seed dispersal, recruitment mechanisms, propagation requirements, and growth and survival rates of the native plant species could aid in the success of restor ation efforts. If a desired nati ve plant species has a limited seed dispersal mechanism, it could take several years before that sp ecies colonizes or dominates a restored wetland. For example, the seed dispersal mechanism of T. domingensis is relatively fast

PAGE 257

257 compared to C. jamaicense (van der Valk and Rosburg 1997) which could c ontribute to the slower colonization patterns observed for C. jamaicense in this study. The results of this study offers evidence that with time, a diverse plant community similar to a native wetland community can develop without human intervention. However, more time is clearly needed to restore ecosystem function to th e level of the native system. The key here is time. Unfortunately, wetland mitigation laws require that wetlands created or restored that serve as mitigation projects are only required 5-10 years of monitoring (Clean Water Act, Section 404). This study along with many others provides ample evidence that this monitoring time period is may not be long enough to restore and maintain plant community structure or ecosystem function (Whigham 1999, Brinson and Malvarez 2002, Kellogg and Bridgham 2002, Callaway et al. 2003, Dalrymple et al. 2003, Seabloom and van der Valk 2003, Polley et al. 2005). Future Recommendations It is still unclear of the long lasting effect s of this residual P on vegetation community structure in the restoration of the HID. The lega cy effect of enriched P lingering in the invaded sites will be a challenge to overcome. The el evated levels of soil and plant P found in the invaded communities have resulted in high P ava ilability as compare to the oligotrophic conditions of the native communities. Previ ous farming and the consequent invasion by S. terebinthifolius has resulted in increase P availability, altered microbial activities and reduced N transformation rates. Additionally, the effect of the restoration method utilized will also carry with it a legacy which the native plant species will have to over come in order to colonize these sites. The combination of field and laboratory re search has provided an indication of the influence of nutrient availability and limita tions on macrophyte diversity, dominance and

PAGE 258

258 composition. The results have helped eluc idate links between macrophyte diversity and ecosystem functions (i.e., productivity, decompos ition, nutrient availability, and nutrient-use efficiency). The knowledge gained from this resear ch can be used to inform scientists and policy makers about the importance of macrophyte divers ity in freshwater wetla nds as it relates to wetland mitigation and restoration.

PAGE 259

259 LIST OF REFERENCES Aerts, R. 1996. Nutrient resorp tion from senescing leaves of perennials: Are there general patterns? Journal of Ecology 84:597-608. Aerts, R., and F. S. Chapin. 2000. The mineral nutri tion of wild plants revi sited: A re-evaluation of processes and patterns. Pages 1-67 in Advances in Ecological Research, Vol 30. Aerts, R., H. de Caluwe, and B. Beltman. 2003. Plant community mediated vs. nutritional controls on litter decompositi on rates in grasslands. Ecology 84:3198-3208. Aerts, R., and H. Decaluwe. 1994. Nitrogen Use Efficiency of Carex Species in Relation to Nitrogen Supply. Ecology 75:2362-2372. Aerts, R., and H. deCaluwe. 1997. Nutritional and plant-mediated controls on leaf litter decomposition of Carex species. Ecology 78:244-260. Aerts, R., J. T. A. Verhoeven, and D. F. Wh igham. 1999. Plant-mediat ed controls on nutrient cycling in temperate fens and bogs. Ecology 80:2170-2181. Aerts, R., B. Wallen, N. Malmer, and H. de Caluwe. 2001. Nutritional constraints on Sphagnum growth and potential decay in north ern peatlands. Journal of Ecology 89:292-299. Anderson, J. M. 1976. An ignition method for determination of totla phosphorus in lake sediments. Water Research 10:329-331. Bakker, C., P. M. van Bodegom, H. J. M. Nelis sen, R. Aerts, and W. H. O. Ernst. 2007. Preference of wet dune species for waterlogged conditions can be explained by adaptations and specific recruitment requirements. Aquatic Botany 86:37-45. Baldwin, A. H., M. S. Egnotovich, and E. Clarke. 2001. Hydrologic cha nge and vegetation of tidal freshwater marshes: Field, greenhouse, and seed-b ank experiments. Wetlands 21:519531. Battle, J. M., and S. W. Golladay. 2007. How h ydrology, habitat type, and litter quality affect leaf breakdown in wetlands on the gulf coastal plain of Georgia. Wetlands 27:251-260. Bedford, B. L., M. R. Walbridge and A. Aldous. 1999. Patterns in nutrient availabi lity and plant diversity of temperate Nort h American wetlands. Ecology 80:2151-2169. Bengtsson, G., P. Bengtson, and K. F. Mansson. 2003. Gross nitrogen mineralization-, immobilization-, and nitrification rates as a function of soil C/N ratio and microbial activity. Soil Biology & Biochemistry 35:143-154. Berendse, F. 1994. Competition between Plant-P opulations at Low and Hi gh Nutrient Supplies. Oikos 71:253-260.

PAGE 260

260 Berendse, F., and R. Aerts. 1987. Nitrogen-use-e fficiency: a biologically meaningful definition. Functional Ecology 1:293-296. Birk, E. M., and P. M. Vitousek. 1986. Nitroge n Availability and Nitrogen Use Efficiency in Loblolly-Pine Stands. Ecology 67:69-79. Bobbink, R., M. Hornung, and J. G. M. Roel ofs. 1998. The effects of air-borne nitrogen pollutants on species diversity in natural and semi-natural European vegetation. Journal of Ecology 86:717-738. Bodelier, P. L. E., J. A. Li bochant, C. Blom, and H. J. Laanbroek. 1996. Dynamics of nitrification and denitrifica tion in rootoxygenated sediment s and adaptation of ammoniaoxidizing bacteria to lowoxygen or anoxic habitats. Applied and Environmental Microbiology 62:4100-4107. Boers, A. M., R. L. D. Veltman, and J. B. Zedler. 2007. Typha x glauca dominance and extended hydroperiod constrain restor ation of wetland diversity. Ecological Engineering 29:232244. Bornette, G., C. Amoros, and N. L. Lamouroux. 1998. Aquatic plant di versity in riverine wetlands: The role of conn ectivity. Freshwater Biology 39:267-283. Boyer, K. E., and J. B. Zedler. 1998. Effects of nitrogen additions on the ve rtical structure of a constructed cordgrass marsh. Ecological Applications 8:692-705. Brady, N. C., and R. R. Weil. 1999. The Nature and Properties of Soils 12th edition. Prentice Hall, Upper Saddle River, N.J. Bridgham, S. D., J. Pastor, C. A. McClaughert y, and C. J. Richardson. 1995. Nutrient-Use Efficiency a Litterfall Index, a Model, a nd a Test Along a Nutrient-Availability Gradient in North-Carolina Peatlands. American Naturalist 145:1-21. Bridgham, S. D., and C. J. Richardson. 2003. Endogenous versus exogenous nutrient control over decomposition and mineralization in No rth Carolina peatlands. Biogeochemistry 65:151-178. Brinson, M. M., and A. I. Malvarez. 2002. Temp erate freshwater wetland s: types, status, and threats. Environmental Conservation 29:115-133. Brookes, P. C., A. Landman, G. Pruden, and D. S. Jenkinson. 1985. Chlo roform Fumigation and the Release of Soil-Nitrogen a Rapid Dire ct Extraction Method to Measure Microbial Biomass Nitrogen in Soil. So il Biology & Biochemistry 17:837-842. Bruland, G. L., S. Grunwald, T. Z. Osborne, K. R. Reddy, and S. Newman. 2006. Spatial distribution of soil properties in Water C onservation Area 3 of the Everglades. Soil Science Society of America Journal 70:1662-1676.

PAGE 261

261 Bundy, L. G., and J. J. Meisinger. 1994. Nitrogen Availability Indices. Pages 951-984 in R. W. Weaver, S. Angle, P. Bottomley, D. Bezdicek, S. Smith, A. Tabatabai, A. Wollum, S. H. Mickelson, and J. M. Bigham, editors. Methods of Soil Analysis, Part 2: Microbiological and Biochemical Properties. Soil Science Society of America, Madison, WI. Burns, A., and D. S. Ryder. 2001. Response of bacterial extracellular en zymes to inundation of floodplain sediments. Freshwater Biology 46:1299-1307. Callaway, J. C., G. Sullivan, and J. B. Zedler. 2003. Species-rich plantings increase biomass and nitrogen accumulation in a wetland restoration experiment Ecological Applications 13:1626-1639. Catovsky, S., M. A. Bradford, and A. Hector 2002. Biodiversity a nd ecosystem productivity: implications for carbon storage. Oikos 97:443-448. Chabrerie, O., I. Poudevigne, F. Bureau, M. Vinceslas-Akpa, S. Nebbache, M. Aubert, A. Bourcier, and D. Alard. 2001. Biodiversity and ecosystem func tions in wetlands: A case study in the estuary of the Seine river, France. Estuaries 24:1088-1096. Chapin, F. S. 1980. The mineral nutrition of wild plants. Annual Review of Ecology and Systematics 11:233-260. Chapin, F. S., P. A. Matson, and H. A. Moone y. 2002. Principles of Terrestrial Ecosystem Ecology. Springer, New York. Chapin, F. S., M. S. Torn, and M. Tateno. 1996. Principles of ecosy stem sustainability. American Naturalist 148:1016-1037. Chapin, F. S., P. M. Vitousek, and K. Vancleve 1986. The Nature of Nutrient Limitation in Plant-Communities. American Naturalist 127:48-58. Chiang, C., C. B. Craft, D. W. Rogers, and C. J. Richardson. 2000. Effects of 4 years of nitrogen and phosphorus additions on Everglades plant communities. Aquatic Botany 68:61-78. Chiu, C. Y., S. C. Lee, T. H. Chen, and G. L. Tian. 2004. Denitrification associated N loss in mangrove soil. Nutrient Cycling in Agroecosystems 69:185-189. Chrst, R. J. 1991. Environmental control of th e synthesis and activity of aquatic microbial ectoenzymes. Pages 29-59 in R. J. Crst, editor. Microbial enzymes in aquatic systems. Springer-Verlag, New York. Clark, D. 1997. Brazilian pepper management pl an for Florida. Brazilian Pepper Task Force, Sanibel, Fl. Clement, B., and E. Maltby. 1996. Plant divers ity and ecological variables in moist and wet grasslands of riverine we tlands. Acta Botanica Gallica 143:309-316.

PAGE 262

262 Corstanje, R., K. R. Reddy, and K. M. Portie r. 2006. Typha latifolia and Cladium jamaicense litter decay in response to exogenous nutrient enrichment. Aquatic Botany 84:70-78. Craft, C. B., and C. J. Richardson. 1993. P eat Accretion and Phosphorus Accumulation Along a Eutrophication Gradient in the Nort hern Everglades. Biogeochemistry 22:133-156. Craft, C. B., and C. J. Richardson. 1997. Rela tionships between soil nutrients and plant species composition in everglades peatlands. Journal of Environmental Quality 26:224-232. Craft, C. B., J. Vymazal, and C. J. Rich ardson. 1995. Response of Everglades PlantCommunities to Nitrogen and Phosphorus Additions. Wetlands 15:258-271. Crst, R. J. 1991. Environmental control of th e synthesis and activity of aquatic microbial ectoenzymes. Pages 29-59 in R. J. Crst, editor. Microbial enzymes in aquatic systems. Springer-Verlag, New York. Dalrymple, G. H., R. F. Doren, N. K. O'Hare, M. R. Norland, and T. V. Armentano. 2003. Plant colonization after complete and partial removal of disturbed soils for wetland restoration of former agricultural fields in Everglades National Park. Wetlands 23:1015-1029. Daoust, R. J., and D. L. Childers. 1999. Controls on emergent macrophyte composition, abundance, and productivity in freshwater everglades wetland communities. Wetlands 19:262-275. Davis, S. M. 1991. Growth, Decomposition, and Nutrient Retention of Cladium-Jamaicense Crantz and Typha-Domingensis Pers in the Florida Everglades. Aquatic Botany 40:203224. DeBusk, T. A., F. E. Dierberg, and K. R. Re ddy. 2001. The use of macrophyte-based systems for phosphorus removal: an overview of 25 years of research and operational results in Florida. Water Science and Technology 44:39-46. DeBusk, T. A., J. E. Peterson, and K. R. Reddy. 1995. Use of aquatic and terrestrial plants for removing phosphorus from dairy wastewaters. Ecological Engineering 5:371-390. Diaz, S., and M. Cabido. 2001. Vive la differe nce: plant functional diversity matters to ecosystem processes. Trends in Ecology & Evolution 16:646-655. Dinkelmeyer, H., J. Lehmann, A. Renck, L. Trujillo J. P. da Silva, G. Gebauer, and K. Kaiser. 2003. Nitrogen uptake from N-15-enriched fertil izer by four tree crops in an Amazonian agroforest. Agroforestry Systems 57:213-224. Domenech, J. A., E. Perez, M. T. Gallardo, and D. F. Martin. 1997. Identi fication of allelopathic chemicals in a pond of ca ttail. Florida Scientist 60:202-203. Doren, R. F., T. V. Armentano, L. D. Whiteaker, and R. D. Jones. 1997. Marsh vegetation patterns and soil phosphorus grad ients in the Everglades ecosystem. Aquatic Botany 56:145-163.

PAGE 263

263 Doren, R. F., and L. D. Whiteaker. 1990. Comp arison of economic feasibility of chemical control strategies on differi ng age and density classes of Schinus terebinthifolius. Natural Areas Journal 10:28-34. Ehrenfeld, J. G. 2003. Effects of exotic plant invasions on soil nutrient cycling processes. Ecosystems 6:503-523. Ehrenfeld, J. G. 2004. The expression of multiple functions in urban forested wetlands. Wetlands 24:719-733. Ehrenfeld, J. G., P. Kourtev, and W. Z. Hu ang. 2001. Changes in soil functions following invasions of exotic understor y plants in deciduous forest s. Ecological Applications 11:1287-1300. El-Kahloun, M., D. Boeye, V. Van Haesebroeck, and B. Verhagen. 2003. Differential recovery of aboveand below-ground rich fen vegeta tion following fertilization. Journal of Vegetation Science 14:451-458. Engelhardt, K. A. M., and M. E. Ritchie. 2001. Effects of macrophyte species richness on wetland ecosystem functioning and services. Nature 411:687-689. Epstein, H. E., I. Burke, and A. R. Mosier. 1998. Plant effects on spatial and temporal patterns of nitrogen cycling in shortgrass steppe. Ecosystems 1:374-385. Epstein, H. E., I. Burke, and A. R. Mosier. 2001. Plant effects on nitrogen retention in shortgrass steppe 2 years after 15N addition. Oecologia 128:422-430. Ernst, W. H. O., and F. Piccoli. 1995. Fru it-Development and Performance of Schoenus Nigricans in Coastal Dune Sl acks of Europe an Extension of Baker,H.G. Seed MassAltitude to a Seed Mass-Latitude Relationship. Acta Botanica Neerlandica 44:41-53. Ernst, W. H. O., and N. F. Vanderham. 1988. Po pulation-Structure and Re juvenation Potential of Schoenus-Nigricans in Coastal Wet Dune Slacks. Acta Botanica Neerlandica 37:451-465. Ernst, W. H. O., R. D. Vis, and F. Piccoli. 1995. Silicon in Developing Nuts of the Sedge Schoenus Nigricans. Jour nal of Plant Physiology 146:481-488. Evans, R. D. 2001. Physiological mechanisms influencing plant nitrogen isotope compositon. Trends in Plant Science 6:121-126. Ewel, J. J., D. S. Ojima, D. A. Karl, a nd T. A. DeBusk. 1982. Schinus in successional ecosystems of the Everglades National Park South Florida Research Center Report T676:141. Feller, I. C., K. L. McKee, D. F. Whigham, and J. P. O'Neill. 2003. Nitrogen vs. phosphorus limitation across an ecotonal gradient in a mangrove forest. Biogeochemistry 62:145-175.

PAGE 264

264 Fennessy, M. S., and J. K. Cronk. 2001. Wetland Pl ants: Biology and Ecology, 1st edition. Lewis Publishers, New York. Firestone, M. K., R. B. Firestone, and J. M. Tiedje. 1980. Nitrous-Oxide from Soil Denitrification Factor s Controlling Its Biolog ical Production. Science 208:749-751. Fitter, A. H., C. A. Gilligan, K. Hollingworth A. Kleczkowski, R. M. Twyman, and J. W. Pitchford. 2005. Biodiversity and ecosystem function in soil. Functional Ecology 19:369377. Fogel, M. L., and N. Tuross. 1999. Transfor mation of plant biochemicals to geological macromolecules during early diagenesis. Oecologia 120:336-346. Fojt, W., and M. Harding. 1995. Thirty years of change in the vegetation communities of three valley mires in Suffulk, England. Journal of Applied Ecology 32:561-577. Francis, A. P., and D. J. Currie. 2003. A globa lly consistant richnessclimate relationship for angiosperms. American Naturalist 161:523-536. Fry, B. 2006. Stable Isotope Ecology. Springer, New York. Fry, B., and T. J. Smith. 2002. Stable isotope studies of red mangroves and filter feeders from the Shark River Estuary, Florida. Bulletin of Marine Science 70:871-890. Garten, C. T. 1993. Variation in Foliar N-15 A bundance and the Availability of Soil-Nitrogen on Walker Branch Watershed. Ecology 74:2098-2113. Gartner, T. B., and Z. G. Cardon. 2004. Decomp osition dynamics in mixed-species leaf litter. Oikos 104:230-246. Gerzabek, M. H., G. Haberhauer, M. Stemmer, S. Klepsch, and E. Haunold. 2004. Long-term behaviour of N-15 in an alpine gr assland ecosystem. Biogeochemistry 70:59-69. Gessner, M. O., P. Inchausti, L. Persson, D. G. Raffaelli, and P. S. Giller. 2004. Biodiversity effects on ecosystem functioning: insi ghts from aquatic systems. Oikos 104:419-422. Gophen, M. 2000. Nutrient and plant dynamics in Lake Agmon Wetlands (Hul a Valley, Israel): a review with emphasis on Typha domi ngensis (1994-1999). Hydrobiologia 441:25-36. Grace, J. B. 1999. The factors controlling specie s density in herbaceous plant communities: an assessment. Perspectives in Plant Ecology, Evolution and Systematics 2:1-28. Gribsholt, B., H. T. S. Boschker, E. Struyf, M. Andersson, A. Tramper, L. De Brabandere, S. van Damme, N. Brion, P. Meire, F. Dehairs, J. J. Middelbur g, and C. H. R. Heip. 2005. Nitrogen processing in a tidal freshwater marsh: A wholeecosystem N-15 labeling study. Limnology and Oceanography 50:1945-1959. Grime, J. P. 1973. Competitive exclus ion in herbaceous vegetation. Nature 242:344-347.

PAGE 265

265 Grime, J. P. 1977. Evidence for Existence of 3 Prim ary Strategies in Plants and Its Relevance to Ecological and Evolutionary Th eory. American Naturalist 111:1169-1194. Grime, J. P. 1997. Biodiversity and ecosyste m function: the debate deepens. Science 277:12601261. Gsewell, S. 2004. N:P ratios in terrestrial plants: variations and func tional significance. New Phytologist 164:243-266. Gsewell, S. 2005. Nutrient resorption of wetland gr aminoids is related to the type of nutrient limitation. Functional Ecology 19:344-354. Gsewell, S., and C. Freeman. 2005. Nutrient limitation and enzyme activities during litter decomposition of nine wetland species in rela tion to litter N : P ratios. Functional Ecology 19:582-593. Gsewell, S., and W. Koerselman. 2002. Variatio n in nitrogen and phosphorus concentrations of wetland plants. Perspectives in Plan t Ecology, Evolution and Systematics 5:37-61. Gsewell, S., and J. T. A. Verhoeven. 2006. Litter N : P ratios indicate whether N or P limits the decomposability of graminoid leaf litter. Plant and Soil 287:131-143. Gsewell, S., A. Zorzi, and A. Gigon. 2000. Mowing in early summer as a remedy to eutrophication in Swiss fen mea dows: are really more nutrients removed? Bulletin of the Geobotonical Institute ETH 66:11-24. Gusewell, S., K. M. Bailey, W. J. Roem, and B. L. Bedford. 2005. Nu trient limitation and botanical diversity in wetlands: Can fer tilisation raise species richness? Oikos 109:71-80. Handley, L. L., and C. M. Scrimege our. 1997. Terrestrial plant ecology and 15N natural abundance: the present limits to interpretation for uncultivated systems with original data from a Scottish old field. Advances in Ecological Research 27:133-212. Harris, T. T., K. A. Williges, and P. V. Zimb a. 1995. Primary productivity and decomposition of five emergent macrophytes communities in Lake Okeechobee marsh ecosystems. Archivum Hydrobiologii, Species Issues, Advances in Limnology 45:63-78. Hart, S. C., J. M. Stark, E. A. Davidson, a nd M. K. Firestone. 1994. N itrogen Mineralization, Immobilization, and Nitrification. Pages 985-1018 in R. W. Weaver, S. Angle, P. Bottomley, D. Bezdicek, S. Smith, A. Tabatabai, A. Wollum, S. H. Mickelson, and J. M. Bigham, editors. Methods of Soil Analysis Part 2: Microbiologi cal and Biochemical Properties. Soil Science Societ y of America, Madison, WI. Hauck, R. D., J. J. Meisinger, and R. L. Mulv aney. 1994. Practical consid erations in the use of nitrogen tracers in agricultura l and environmental research. in R. W. Weaver, S. Angle, P. Bottomley, D. Bezdicek, S. Smith, A. Tabatabai, A. Wollum, S. H. Mickelson, and J. M. Bigham, editors. Methods of Soil Analysis Part 2: Microbiologi cal and Biochemical Propertie. Soil Science Society of America, Madison, WI.

PAGE 266

266 Hawkins, B. A., R. Field, H. V. Cornell, D. J. Currie, J. F. Guegan, and D. M. Kaufman. 2003. Energy, water, and broad-scale patt erns of species richness. Ecology 84:3105-3117. Hector, A., A. J. Beale, A. Minns, S. J. Otwa y, and J. H. Lawton. 2000. Consequences of the reduction of plant diversity for litter decomposition: eff ects through litter quality and microenvironment. Oikos 90:357-371. Hector, A., B. Schmid, C. Beierkuhnlein, M. C. Caldeira, M. Diemer, P. G. Dimitrakopoulos, J. A. Finn, H. Freitas, P. S. Giller, J. Good, R. Harris, P. Hogberg, K. Huss-Danell, J. Joshi, A. Jumpponen, C. Korner, P. W. Leadley, M. Loreau, A. Minns, C. P. H. Mulder, G. O'Donovan, S. J. Otway, J. S. Pereira, A. Pr inz, D. J. Read, M. Scherer-Lorenzen, E. D. Schulze, A. S. D. Siamantziouras, E. M. Sp ehn, A. C. Terry, A. Y. Troumbis, F. I. Woodward, S. Yachi, and J. H. Lawton. 1999. Plant diversity and pr oductivity experiments in European grasslands. Science 286:1123-1127. Hedin, L. O., P. M. Vitousek, and P. A. Mats on. 2003. Nutrient losses over four million years of tropical forest deve lopment. Ecology 84:2231-2255. Hefting, M., J. C. Clement, D. Dowrick, A. C. Co sandey, S. Bernal, C. Cimpian, A. Tatur, T. P. Burt, and G. Pinay. 2004. Water table elev ation controls on soil nitrogen cycling in riparian wetlands along a European cl imatic gradient. Biogeochemistry 67:113-134. Hemond, H. F. 1983. The Nitrogen Budget of Thoreau Bog. Ecology 64:99-109. Henry, H. A. L., and R. L. Jefferies. 2002. Free amino acid, ammonium and nitrate concentrations in soil solutions of a grazed co astal marsh in relation to plant growth. Plant Cell and Environment 25:665-675. Henry, H. A. L., and R. L. Jefferies. 2003a. Inte ractions in the uptake of amino acids, ammonium and nitrate ions in the Arctic salt-marsh grass, Puccinellia phryga nodes. Plant Cell and Environment 26:419-428. Henry, H. A. L., and R. L. Jefferies. 2003b. Plant amino acid uptake, soluble N turnover and microbial N capture in soils of a grazed Arctic salt marsh. Journal of Ecology 91:627-636. Hobbie, S. E., and F. S. Chapin. 1998. The response of tundra plant biomass, aboveground production, nitrogen, and CO2 flux to experimental warming. Ecology 79. Hgberg, P. 1997. 15N natural abundance in soil-plan t systems. New Phytologist 137:179-203. Hooper, D. U., F. S. Chapin, J. J. Ewel, A. Hector P. Inchausti, S. Lavorel, J. H. Lawton, D. M. Lodge, M. Loreau, S. Naeem, B. Schmid, H. Se tala, A. J. Symstad, J. Vandermeer, and D. A. Wardle. 2005. Effects of biodiversity on ecosystem functioning: A consensus of current knowledge. Ecological Monographs 75:3-35. Hooper, D. U., and P. M. Vitousek. 1997. The effects of plant composition and diversity on ecosystem processes. Science 277:1302-1305.

PAGE 267

267 Hoppe, H. G. 1993. Use of fluorogenic model substr ates for ectracellular enzyme activity (EEA) measurement of bacteria. Pages 423-431 in P. F. Kemp, B. F. Sherr, E. B. Sherr, and J. J. Cole, editors. Handbook of methods in aquatic microbial ecology. Lewis Publishing, Boca Raton, FL. Inglett, P. W., and K. R. Reddy. 2006. Invest igating the use of macrophyte stable C and N isotopic ratios as indicators of wetland eutrophication: Patterns in the P-affected Everglades. Limnology and Oceanography 51:2380-2387. Inubushi, K., and S. Acquaye. 2004. Role of micr obial biomass in biogeochemical processes in paddy soil environments. Soil Sc ience and Plant Nutrition 50:793-805. Jackson, L. E., J. P. Schimel, and M. K. Fi restone. 1989. Short-Term Partitioning of Ammonium and Nitrate between Plants and Microbes in an Annual Grassland. Soil Biology & Biochemistry 21:409-415. Johnson, K. H., K. A. Vogt, H. J. Clark, O. J. Schmitz, and D. J. Vogt. 1996. Biodiversity and the productivity and stability of ecosyst ems. Trends in Ecology & Evolution 11:372-377. Johnson, S., and E. Rejmankova. 2005. Impact s of land use on nutri ent distribution and vegetation composition of freshwater wetlands in Northern Belize. Wetlands 25:89-100. Jones, R., L. King, M. M. Dent, S. C. Maberly, and C. E. Gibson. 2004. Nitrogen stable isotope ratios in surface sediments, epilithon and m acrophytes from upland lakes with differing nutrient status. Freshwater Biology 49:382-391. Jordan, T. E., D. F. Whigham, and C. D.L. 1989. The role of litter a nd nutrient cycling in brackish tidal marsh. Ecology 70:1906-1915. Keeney, D. R. 1982. Nitrogen-available indices. in A. L. Page and et al., editors. Methods of Soil Analysis, Part 2: Agron. M onogr. 9. ASA, Madison, WI. Kellogg, C. H., and S. D. Bridgham. 2002. Col onization during early succession of restored freshwater marshes. Canadian Journal of Botany-Revue Canadienne De Botanique 80:176185. Killingbeck, K. T. 1996. Nutrients in senesced leav es: Keys to the search for potential resorption and resorption proficiency. Ecology 77:1716-1727. Koba, K., M. Hirobe, L. Koyama, A. Kohzu, N. Tokuchi, K. J. Nadelhoffer, E. Wada, and H. Takeda. 2003. Natural 15N abundance of plants and soil N in a temperate coniferous forest. Ecosystems 6:457-469. Koch, A. L. 1985. The macroeconomic s of bacterial growth. Pages 1-42 in M. Fletcher and G. D. Floodgate, editors. Bacteria in their natura l environments. Acedemic Press, London.

PAGE 268

268 Koch, M. S., and K. R. Reddy. 1992. Distribu tion of Soil and Plant Nutrients Along a Trophic Gradient in the Florida Everglades. So il Science Society of America Journal 56:14921499. Koerselman, W., and A. F. M. Meuleman. 1996. Th e vegetation N:P ratio: a new tool to detect the nature of nutrien t limitation. Journal of Applied Ecology 33:1441-1450. Kuehn, K. A., M. J. Lemke, K. Suberkropp, and R. G. Wetzel. 2000. Microbial biomass and production associated with decay ing leaf litter of the emerge nt macrophyte Juncus effusus. Limnology and Oceanography 45:862-870. Kuehn, K. A., and K. Suberkropp. 1998. Decomp osition of standing litter of the freshwater emergent macrophyte Juncus e ffusus. Freshwater Biology 40:717-727. Locky, D. A., and S. E. Bayley. 2006. Plant di versity, composition, and rarity in the southern boreal peatlands of Manitoba, Canada. Canadi an Journal of Botany-Revue Canadienne De Botanique 84:940-955. Lodge, T. E. 2005. The Everglades Handbook: Understanding the Ecosystem, 2nd edition. CRC Press, Boca Raton. Loope, L. L., and V. L. Dunevitz. 1981. Investigations of early plant succession on abandoned farmland in the Everglades National Park. Sout h Florida Research Ce nter Report T-644:65. Loreau, M. 2000. Biodiversity a nd ecosystem functioning: recent theoretical a dvances. Oikos 91:3-17. Loreau, M., S. Naeem, P. Inchausti, J. Bengtsson, J. P. Grime, A. Hector, D. U. Hooper, M. A. Huston, D. Raffaelli, B. Schmid, D. Tilman, and D. A. Wardle. 2001. Ecology Biodiversity and ecosystem func tioning: Current knowledge and future challenges. Science 294:804-808. Lorenzen, B., H. Brix, I. A. Mendelssohn, K. L. McKee, and S. L. Miao. 2001. Growth, biomass allocation and nutrient use efficiency in Cladium jamaicense and Typha domingensis as affected by phosphorus and oxygen availability. Aquatic Botany 70:117-133. Lortie, C. J., R. W. Brooker, P. Choler, Z. Kikvidze, R. Michal et, F. I. Pugnaire, and R. M. Callaway. 2004. Rethinking plant community theory. Oikos 107:433-438. Mack, M. C., and C. M. D'Antonio. 2003. Exotic grasses alter controls over soil nitrogen dynamics in a Hawaiian woodland. Ecological Applications 13:154-166. Mack, M. C., C. M. D'Antonio, and R. E. Le y. 2001. Alteration of ecosystem nitrogen dynamics by exotic plants: A case study of C-4 gra sses in Hawaii. Ecological Applications 11:13231335. Mahaney, W. M., D. H. Wardrop, and R. P. Br ooks. 2004. Impacts of stressors on the emergence and growth of wetland plant species in Pennsylvania, USA. Wetlands 24:538-549.

PAGE 269

269 Mann, K. H. 1988. Production and use of detritus in various freshwater, estuarine and coastal marine ecosystems. Limnology and Oceanography 33:910-930. Martin, J. F., and K. R. Reddy. 1997. Interact ion and spatial distribut ion of wetland nitrogen processes. Ecological Modelling 105:1-21. Mason, C. F., and R. J. Bryant. 1975. Production, nutrient content and decomposition of Phragmites communis trin. and Typha angustifolia L. Journal of Ecology 63:71-95. McCune, B., and J. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, Oregon, USA. McCune, B., and M. J. Mefford. 1999. PC-ORD: Multivariate Analysis of Ecological Data. in MjM Software Design, Glen eden Beach, Oregon, USA. McKee, K. L., I. C. Feller, M. Pop, and W. Wanek. 2002. Mangrove isotopic (15N and 13C) fractionation across a nitrogen vs. phos phorus limitation gradient. Ecology 83:1065-1075. Meekins, J. F., and B. C. McCarthy. 2001. Eff ect of environmental variation on the invasive success of a nonindigenous forest herb. Ecological Applications 11:1336-1348. Meentemeyer, V. 1978. Macroclimate and Lignin Control of Litter Decomposition Rates. Ecology 59:465-472. Megonigal, J. P., S. P. Faulkne r, and W. H. Patrick. 1996. Th e microbial activity season in southeastern hydric so ils. Soil Science Society of America Journal 60:1263-1266. Melillo, J. M., J. D. Aber, and J. F. Muratore. 1982. Ntrogen and lignin c ontrol of hardwood leaf litter decomposition dynamics. Ecology 63:621-626. Merbach, W., J. Schulze, M. Richert, E. Rr occo, and K. Mengel. 2000. A comparison of different N-15 application techniques to st udy the N net rhizodeposition in the plant-soil system. Journal of Plant Nutrition and Soil Sc ience-Zeitschrift Fur Pflanzenernahrung Und Bodenkunde 163:375-379. Meuleman, A. F., J. P. Beekman, and J. T. A. Verhoeven. 2002. Nutrient retention and nutrient use efficiency in Phagmites australis stands after wastewater application. Wetlands 22:712-721. Miao, S. L., S. Newman, and F. H. Sklar. 2000. Effects of habitat nutrien ts and seed sources on growth and expansion of Typha domingensis. Aquatic Botany 68:297-311. Mitsch, W. J., and J. G. Gosselink. 2000. Wetlands, 3rd edition. John Wiley and Sons, New York.

PAGE 270

270 Mosier, A. R., and L. Klemedtsson. 1994. Meas uring denitrification in the field. Pages 10471065 in R. W. Weaver, S. Angle, P. Bottomley, D. Bezdicek, S. Smith, A. Tabatabai, A. Wollum, S. H. Mickelson, and J. M. Bigham, editors. Methods of So il Analysis, Part 2: Microbiological and Biochemical Properties. Soil Science So ciety of America, Madison, WI. Mulholland, P. J., J. L. Tank, D. M. Sanzone, W. M. Wollheim, B. J. Peterson, J. R. Webster, and J. L. Meyer. 2000. Nitrogen cycling in a forest stream determined by a N-15 tracer addition. Ecological Monographs 70:471-493. Murphy, K. J., G. Dickinson, S. M. Thomaz, L. M. Bini, K. Dick, K. Greaves, M. P. Kennedy, S. Livingstone, H. McFerran, and J. M. M ilne. 2003. Aquatic plant communities and predictors of diversity in a sub-tropical river floodplain: the upper Rio Parana, Brazil. Aquatic Botany 77:257-276. Nadelhoffer, K. J., and B. Fry. 1994. Nitrogen is otope studies in forest ecosystems. Pages 22-44 in K. Lajtha and R. H. Michener, editors. St able isotopes in ecol ogy and environemtal science. Blackwell Scientif ic Publications, London. Naeem, S., L. J. Thompson, S. P. Lawler, J. H. Lawton, and R. M. Woodfin. 1994. Declining Biodiversity Can Alter the Perfor mance of Ecosystems. Nature 368:734-737. Newman, S., J. B. Grace, and J. W. Koebel 1996. Effects of nutrients and hydroperiod on Typha, Cladium, and Eleocharis: Implications for Everglades restoration. Ecological Applications 6:774-783. Noe, G. B., D. L. Childers, and R. D. Jones. 2001. Phosphorus biogeochemistry and the impact of phosphorus enrichment: Why is the everglades so unique? Ecosystems 4:603-624. Novak, M., F. Buzek, and M. Adam ova. 1999. Vertical trends in 13C, 14N and 34S ratios in bulk Sphagnum peat. Soil Biology & Biochemistry 31:1343-1346. O'Hare, N. K., and G. H. Dalrymple. 2003. Final annual report: Biol ogical monitoring of restored wetlands in the Hole -in-the-Donut, Everglades Nation al Park, July 1, 2001 June 30, 2002. Submitted to Everglades National Park, Homestead, Florida. Olde Venterink, H., N. M. Pieter se, J. D. M. Belgers, M. J. Wa ssen, and O. D. de Ruiter. 2002. N, P and K budgets along nutrien t availability and productivi ty gradients in wetlands. Ecological Applications 12:1010-1026. Olde Venterink, H., M. J. Wasse n, A. W. M. Verkroost, and P. C. de Ruiter. 2003. Species richness-productivity patterns differ between N-, P-, and K-limited wetlands. Ecology 84:2191-2199. Olsson, M. O., and U. Falkengren-Grerup. 2000. Potential nitrification as an indicator of preferential uptake of ammoni um or nitrate by plants in an oak woodland understorey. Annals of Botany 85:299-305.

PAGE 271

271 Oomes, M. J. M., P. J. Kuikman, and F. H. H. Jacobs. 1997. Nitrogen availability and uptake by grassland in mesocosms at two water leve ls and two water qualities. Plant and Soil 192:249-259. Pauli, D., M. Peintinger, and B. Schmid. 2002. Nutr ient enrichment in calcareous fens: effects on plant species and community stru cture. Basic and Applied Ecology 3:255-266. Piccolo, M. C., C. Neill, J. M. Melillo, C. C. Cerri, and P. A. Steudler. 1996. 15N natural abundance in forest and pasture soils of the Brazilian Amazon basin. Plant and Soil 182:249-258. Picking, D. J., and P. L. M. Veneman. 2004. Ve getation patterns in a calcareous sloping fen of Southwestern Massachusetts, USA. Wetlands 24:514-528. Polley, H. W., J. D. Derner, and B. J. Wilsey. 2 005. Patterns of plant species diversity in remnant and restored tallgrass prai ries. Restoration Ecology 13:480-487. Pollock, M. M., R. J. Naiman, and T. A. Ha nley. 1998. Plant species richness in riparian wetlands A test of biodiversity theory. Ecology 79:94-105. Prindle, V., and D. F. Martin. 1996. Allelopathic properties of cattails, Typha domingensis, in Hillsborough County, Florida. Florida Scientist 59:141-154. Reddy, K. R., and W. F. DeBusk. 1985. Nutrie nt removal potential of selected aquatic macrophytes. Journal of Environmental Quality 14:459-462. Reddy, K. R., G. A. O'Conner, and C. L. Sche lske, editors. 1999. Phosphorus biogeochemistry in subtropical ecosystems. Lewis. Reddy, K. R., W. H. Patrick, and C. W. Lindau. 1989. Nitrification-Denitr ification at the Plant Root-Sediment Interface in Wetlands. Limnology and Oceanography 34:1004-1013. Rejmankova, E., and K. Houdkova. 2006. Wetland pl ant decomposition under different nutrient conditions: what is more important, litter quality or site qualit y? Biogeochemistry 80:245262. Rejmankova, E., and D. Sirova. 2007. Wetla nd macrophyte decomposition under different nutrient conditions: Relationships between decomposition rate, enzyme activities and microbial biomass. Soil Biology & Biochemistry 39:526-538. Richardson, C. J., G. M. Ferrell and P. Vaithiyanathan. 1999. Nutr ient effects on stand structure, resorption efficiency, and secondary co mpounds in Everglades sawgrass. Ecology 80:2182-2192. Rickey, M. A., and R. C. Anderson. 2004. Eff ects of nitrogen addition on the invasive grass Phragmites australis and a native competitor Spartina pectinata. Journal of Applied Ecology 41:888-896.

PAGE 272

272 Rivero, R. G., S. Grunwald, T. Z. Osborne, K. R. Reddy, and S. Newman. 2007. Characterization of the spatial distribu tion of soil properties in water conservation area 2A, Everglades, Florida. Soil Science 172:149-166. Robinson, D. 2001. 15N as an integrator of the nitrogen cycle. Tre nds in Ecology & Evolution 16:153-162. Rothstein, D. E., P. M. Vitousek, and B. L. Simmons. 2004. An exotic tree alters decomposition and nutrient cycling in a Hawaiia n montane forest. Ecosystems 7:805-814. Rowland, A. P., and J. D. Roberts. 1994. Li gnin and cellulose fractionation in decomposition studies using acid-detergent fiber methods. Communications in So il Science and Plant Analysis 25:269-277. Ruckauf, U., J. Augustin, R. Russow, and W. Merbach. 2004. Nitrate removal from drained and reflooded fen soils affected by soil N transf ormation processes and plant uptake. Soil Biology & Biochemistry 36:77-90. SAS. 2005. JMP. in SAS Institute, Inc., Cary, North Carolina, USA. Scheffer, R. A., and R. Aerts. 2000. Root d ecomposition and soil nutrient and carbon cycling in two temperate fen ecosystems. Oikos 91:541-549. Schimel, D. S. 1988. Calculations of microbial growth efficiency from 15N immobilization. Biogeochemistry 6:239-243. Schimel, J. P., and J. Bennett. 2004. Nitroge n mineralization: Challenges of a changing paradigm. Ecology 85:591-602. Schimel, J. P., and F. S. Chapin. 1996. Tundra plant uptake of amino acid and NH4+ nitrogen in situ: Plants compete well for amino acid N. Ecology 77:2142-2147. Schimel, J. P., and M. N. Weintraub. 2003. The implications of exoenz yme activity on microbial carbon and nitrogen limitation in soil: a theoretical model. Soil Biology & Biochemistry 35:549-563. Seabloom, E. W., and A. G. van der Valk. 2003. Plant diversity, composition, and invasion of restored and natural prairie pothole wetlands: Implications for restoration. Wetlands 23:112. Senzia, M. A., A. W. Mayo, T. S. A. Mbwette, J. H. Y. Katima, and S. E. Jorgensen. 2002. Modelling nitrogen transfor mation and removal in primar y facultative ponds. Ecological Modelling 154:207-215. Shaver, G. R., and F. S. Chapin. 1995. Long-term response to factorial NPK fertilizer treatment by Alaskan wet and moist t undra sedge species. Ecography 18:259-275.

PAGE 273

273 Sinsabaugh, R. L., and D. L. Moorhead. 1994. Resource allocation to extracellular enzyme production: a model for nitrogen and phosphor us control of litter decomposition. Soil Biology & Biochemistry 26:1305-1311. Stark, J. M., and S. C. Hart. 1996. Diffusion t echnique for preparing sa lt solutions, Kjeldahl digests, and persulfate digest s for nitrogen-15 analysis. Soil Science Society of America Journal 60:1846-1855. Stiling, P. 1999. Ecology: Theories and Applic ations, 3rd edition. Prentice Hall, Upper Saddle River, New Jersey. Streeter, T. C., R. Bol, and R. D. Bardgett. 2000. Amino acids as a nitr ogen source in temperate upland grasslands: the use of dua l labelled (C-13, N-15) glycine to test for direct uptake by dominant grasses. Rapid Communications in Mass Spectrometry 14:1351-1355. Tank, J. L., J. L. Meyer, D. M. Sanzone, P. J. Mu lholland, J. R. Webster, B. J. Peterson, W. M. Wollheim, and N. E. Leonard. 2000. Analysis of nitrogen cycling in a forest stream during autumn using a N-15-tracer ad dition. Limnology and Oceanography 45:1013-1029. Tanner, C. C. 1996. Plants for constructed we tland treatment systems a comparison of the growth and nutrient uptake of eight emergent species. Ecolog ical Engineering 7:59-83. Tanner, C. C., J. S. Clayton, and M. P. Upsd ell. 1995. Effect of loading rate and planting on treatment of dairy farm wastewaters in cons tructed wetlands II. Removal of nitrogen and phosphorus. Water Research 29:27-34. Templer, P., S. Findlay, and G. Lovett. 2003. Soil microbial biomass and nitrogen transformations among five tree species of th e Catskill Mountains, New York, USA. Soil Biology & Biochemistry 35:607-613. Templer, P. H., M. A. Arthur, G. M. Lovett, and K. C. Weathers. 2007. Plant and soil natural abundance delta N-15: indicators of relative ra tes of nitrogen cycling in temperate forest ecosystems. Oecologia 153:399-406. Templer, P. H., and T. E. Dawson. 2004. Nitr ogen uptake by four tree sp ecies of the Catskill Mountains, New York: Implications fo r forest N dynamics. Plant and Soil 262:251-261. Templer, P. H., G. M. Lovett, K. C. Weathers, S. E. Findlay, and T. E. Dawson. 2005. Influence of tree species on forest nitr ogen retention in the Catski ll Mountains, New York, USA. Ecosystems 8:1-16. Tessier, J. T., and D. J. Raynal. 2003. Use of n itrogen to phosphorus ratios in plant tissue as an indicator of nutrient limitation and nitroge n saturation. Journal of Applied Ecology 40:523534. Tilman, D. 1982. Resource Competition and Community Structure. Princeton University Press, Princeton.

PAGE 274

274 Tilman, D., and J. A. Downing. 1994. Biodive rsity and Stability in Grasslands. Nature 367:363365. Tilman, D., J. Knops, D. Wedin, P. Reich, M. R itchie, and E. Siemann. 1997. The influence of functional diversity and composition on ecosystem processes. Science 277:1300-1302. Tilman, D., D. Wedin, and J. Knops. 1996. Productivity and sustaina bility influenced by biodiversity in grassland ecosystems. Nature 379:718-720. Troxler Gann, T. G., and D. L. Childers. 2006. Relationships betw een hydrology and soils describe vegetation patterns in seasonally fl ooded tree islands of the southern Everglades, Florida. Plant and Soil 279:271-286. Urban, N. H., S. M. Davis, and N. G. Aumen. 1993. Fluctuations in Sawgrass and Cattail Densities in Everglades-Wat er-Conservation-Area-2a under Va rying Nutrient, Hydrologic and Fire Regimes. Aquatic Botany 46:203-223. Valiela, I., J. M. Teal, S. D. Allen, R. Van Etten, D. Goehringer, and S. Volkmann. 1985. Decomposition in salt marsh ecosystems: th e phases and major factors affecting the disappearance of above-ground organic matter. Journal of Marine Biological Ecology 89:29-54. van der Valk, A. G. 1981. Succession in wetlands: a Gleasonian approach. Ecology 62:688-696. van der Valk, A. G., and T. R. Rosburg. 1997. Seed bank composition along a phosphorus gradient in the northern Fl orida Everglades. Wetlands 17:228-236. Vanderhoeven, S., N. Dassonville, L. Chapuis-Lar dy, M. Hayez, and P. Meerts. 2006. Impact of the invasive alien plant Solidago giganteaon primary productivity, plant nutrient content and soil mineral nutrient concentrations. Plant and Soil 286:259-268. Verhoeven, J. T. A., W. Koerselman, and A. F. M. Meuleman. 1996. Nitrogenor phosphoruslimited growth in herbaceous, wet vegetation: Relations with atmospheric inputs and management regimes. Trends in Ecology & Evolution 11:494-497. Vermeer, J. G., and F. Berendse. 1983. The Rela tionship between Nutrient Availability, Shoot Biomass and Species Richness in Grassland and Wetland Communities. Vegetatio 53:121126. Vitousek, P. 1982. Nutrient Cycling and Nutr ient Use Efficiency. American Naturalist 119:553572. Vitousek, P. 1984. Litterfall, nutrient cycling, an d nutrient limitation in tropical forests. Ecology 65:285-298. Vitousek, P. 2004. Nutrient Cycling and Lim itation: Hawai'i as a Model System. Princeton University Press, Princeton, N.J.

PAGE 275

275 Vitousek, P., T. D.M., W. J. Parton, and R. L. Sanford. 1994. Litter decomposition on the Mauna-Loa environmental matrix, Hawaii pa tterens, mechanisms, and models. Ecology 75:418-429. Vymazal, J., and C. J. Richardson. 1995. Species Composition, Biomass, and Nutrient Content of Periphyton in the Florida Ever glades. Journal of Phycology 31:343-354. Waide, R., M. R. Willig, D. Steiner, C. F. Mittlebach, L. Gough, and S. I. Dodson. 1999. The relationship between productiv ity and species richness. A nnual Review of Ecology and Systematics 30:257-300. Walker, B. 1995. Conserving Biol ogical Diversity through Ecosystem Resilience. Conservation Biology 9:747-752. Weaver, R. W., and S. K. Danso. 1994. Dinitrogen Fixation. Pages 1019-1043 in R. W. Weaver, S. Angle, P. Bottomley, D. Bezdicek, S. Smith, A. Tabatabai, and A. Wollum, editors. Methods of Soil Analysis, Part 2: Microbi ological and Biochemical Processes. Soil Science Society of America, Madison, WI. Webster, J. R., and E. F. Benfield. 1986. Vasc ular plant breakdown in freshwater ecosystems. Annual Review of Ecology and Systematics 17:567-594. Weigelt, A., R. Bol, and R. D. Bardgett. 2005. Preferential uptake of soil nitrogen forms by grassland plant species. Oecologia 142:627-635. Wetzel, P. R. 1990. Land-water interfaces: meta bolic and limnological regulators. Verh. Int. Verein. Limnology 24:6-24. Whigham, D. F. 1999. Ecological i ssues related to wetland preserva tion, restorati on, creation and assessment. Science of the Total Environment 240:31-40. White, D. S., and B. L. Howes. 1994. Long-Te rm N-15-Nitrogen Retention in the Vegetated Sediments of a New-England Salt -Marsh. Limnology and Oceanography 39:1878-1892. White, J. R., and K. R. Reddy. 2000. Influen ce of phosphorus loading on organic nitrogen mineralization of everglades soils. So il Science Society of America Journal 64:1525-1534. White, J. R., and K. R. Reddy. 2001. Influence of selected inorganic electron acceptors on organic nitrogen mineralization in everglades soils. Soil Science Society of America Journal 65:941-948. Whitehouse, H. E., and S. E. Bayley. 2005. Ve getation patterns and bi odiversity of peatland plant communities surrounding mid-boreal we tland ponds in Alberta, Canada. Canadian Journal of Botany-Revue Canadienne De Botanique 83:621-637. Willis, C., and W. J. Mitsch. 1995. Effects of Hydrology and Nutrients on Seedling Emergence and Biomass of Aquatic Macrophytes from Natu ral and Artificial Seed Banks. Ecological Engineering 4:65-76.

PAGE 276

276 Wiser, S. K., R. B. Allen, P. W. Clinton, and K. H. Platt. 1998. Communi ty structure and forest invasion by an exotic herb over 23 years. Ecology 79:2071-2081. Woo, I., and J. B. Zedler. 2002. Can nutrients alone shift a sedge meadow towards dominance by the invasive Typha x glauca? Wetlands 22:509-521. Xu, X., and E. Hirata. 2005. Decomposition pa tterns of leaf litter of seven common canopy species in a subtropical forest: N and P dynamics. Plant and Soil 273:279-289. Zak, D. R., W. E. Holmes, D. C. White, A. D. Peacock, and D. Tilman. 2003. Plant diversity, soil microbial communities, and ecosystem function: Are there any links? Ecology 84:2042-2050. Zedler, J. B. 1993. Canopy Architecture of Natura l and Planted Cordgras s Marshes Selecting Habitat Evaluation Criteria. Ecological Applications 3:123-138.

PAGE 277

277 BIOGRAPHICAL SKETCH Angelique Marie Kepp ler is the second oldest of seven siblings, she grew up in the hills of southern Ohio in the small Villa ge of Beaver. Angelique spent her childhood days playing in the woods that surrounded a small privately own lake Her childhood surroundings and her love for nature help guide her future educational goals After graduating with honors from Eastern High School in 1993, she attended Ohio University studying chemical engineering for a brief period. Her time as a chemical engineering student only helped her to realize even more how much in need our environment was of protecting. After 4 years, Angelique chose to change her educational focus to natural resour ces and environmental science. In 2000, Angelique began her environmenta l science education at The Ohio State University where she graduated with Distinc tion in Environmental Science in 2002. Upon completion of her bachelors degree, she immedi ately began work on her masters at The Ohio State University. Her masters work was focused in ecological engineering. Her research involved the use of ecological tr eatment tanks to trea t highly concentrated manure wash water from the universities dairy milk house. Angeliq ue was awarded her masters degree in 2004. Angeliques next stop in life took her to the Un iversity of Florida in Gainesville where she continued her path in environmental scie nce focusing her resear ch on the ecology and biogeochemical processes of wetlands systems w ith a minor in botany. Her research was focused in a chronosequence of restored wetlands in the Everglades National Park. The overall goal of her project was to gain a greater understanding of the controls on plant composition and species diversity in relation to coupled biogeochemical processes of nitrogen. Her future plans include obtai ning a position as a professor c ontinuing to conduct research in natural systems. She hopes to develop a re search program that focuses on plant species diversity conservation with an emphasis on multi-tropic level relationships and coupled

PAGE 278

278 biogeochemical processes related to carbon, nitrogen, and phosphorus that could pote ntially alter plant species composition. Additionally, she look s forward to starting a teaching program that will benefit both higher education and K-12 level students.