<%BANNER%>

Seasonal and Experimental Effects on Microbial Composition and Dynamics in a Tropical Secondary Forest in the Eastern Am...

Permanent Link: http://ufdc.ufl.edu/UFE0016540/00001

Material Information

Title: Seasonal and Experimental Effects on Microbial Composition and Dynamics in a Tropical Secondary Forest in the Eastern Amazon, Brazil
Physical Description: 1 online resource (116 p.)
Language: english
Creator: Veluci-Marlow, Roberta M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: bacteria, carbon, exchange, fungi, ion, latossolo, microorganisms, mineralization, mycorrhizae, nitrification, nitrogen, nutrients, phosphatase, phosphorus, recovery, regrowth
Forest Resources and Conservation -- Dissertations, Academic -- UF
Genre: Forest Resources and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Tropical secondary forests are an increasingly important land cover in the Brazilian Amazon, with 30 to 50% of the deforested area of the Brazilian Amazon in some stage of abandonment. This study investigated water and nutrient constraints on microbial dynamics and nutrient availability in a tropical secondary forest in the Eastern Amazon with manipulative experiments, dry-season irrigation and bi-weekly litter removal, using frequent sampling to capture seasonal and intra-annual fluctuations in Castanhal, Para, Brazil. Irrigation did not consistently alter microbial dynamics, except for lowered NH4+ availability and fungal densities, and increased phosphatase activity. Litter removal decreased microbial biomass C and P, N-mineralization, phosphatase activity and NH4+ availability but increased NO3- availability. Intra-annual variability was mainly driven by wet-up events in the dry season that were not minimized by continuous irrigation (except for NH4+ availability), suggesting either that seasonal drought may not constrain the availability of nutrients or that irrigation was insufficient to cause a more significant effect. These results confirm the critical role of litterfall in tropical forest nutrient cycling and the importance of fluctuations in soil moisture status to nutrient availability. How these belowground results interact with aboveground processes including C uptake, is a fertile area for future research and modeling.
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 Roberta M Veluci-Marlow.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Zarin, Daniel J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0016540:00001

Permanent Link: http://ufdc.ufl.edu/UFE0016540/00001

Material Information

Title: Seasonal and Experimental Effects on Microbial Composition and Dynamics in a Tropical Secondary Forest in the Eastern Amazon, Brazil
Physical Description: 1 online resource (116 p.)
Language: english
Creator: Veluci-Marlow, Roberta M
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: bacteria, carbon, exchange, fungi, ion, latossolo, microorganisms, mineralization, mycorrhizae, nitrification, nitrogen, nutrients, phosphatase, phosphorus, recovery, regrowth
Forest Resources and Conservation -- Dissertations, Academic -- UF
Genre: Forest Resources and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Tropical secondary forests are an increasingly important land cover in the Brazilian Amazon, with 30 to 50% of the deforested area of the Brazilian Amazon in some stage of abandonment. This study investigated water and nutrient constraints on microbial dynamics and nutrient availability in a tropical secondary forest in the Eastern Amazon with manipulative experiments, dry-season irrigation and bi-weekly litter removal, using frequent sampling to capture seasonal and intra-annual fluctuations in Castanhal, Para, Brazil. Irrigation did not consistently alter microbial dynamics, except for lowered NH4+ availability and fungal densities, and increased phosphatase activity. Litter removal decreased microbial biomass C and P, N-mineralization, phosphatase activity and NH4+ availability but increased NO3- availability. Intra-annual variability was mainly driven by wet-up events in the dry season that were not minimized by continuous irrigation (except for NH4+ availability), suggesting either that seasonal drought may not constrain the availability of nutrients or that irrigation was insufficient to cause a more significant effect. These results confirm the critical role of litterfall in tropical forest nutrient cycling and the importance of fluctuations in soil moisture status to nutrient availability. How these belowground results interact with aboveground processes including C uptake, is a fertile area for future research and modeling.
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 Roberta M Veluci-Marlow.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Zarin, Daniel J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0016540:00001


This item has the following downloads:


Full Text





SEASONAL AND EXPERIMENTAL EFFECTS ON MICROBIAL COMPOSITION AND
DYNAMICS IN A TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON,
BRAZIL



















By

ROBERTA M. VELUCI-MARLOW


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

UNIVERSITY OF FLORIDA

2007

































2007 Roberta M. Veluci-Marlow
































To my mom who taught me by example the value of completing a project.









ACKNOWLEDGMENTS

I wish to express sincere appreciation to my husband Brian for his endless assistance

during all laboratory analysis and for postponing his own career and dreams to help me achieve

mine. In addition, special thanks are to due to my major advisor Daniel Zarin whose patience and

perseverance helped me throughout this process, and committee members Michelle Mack, Nick

Commerford, Tim Martin and William McDowell, whose critical eyes, and enlightened

mentoring were instrumental and inspiring.

This dissertation would not have been completed without the thrust and support of many

collaborators and friends, especially Steel Vasconcelos.

For guidance during laboratory work I thank Elizabeth Chu, Claudio Carvalho. For field

and laboratorial work, I thank Glebson and Branco, Robson Canuto, Fabio and Bruno, Jesus,

Tereza, Ronaldo, Ivanildo, Tenilson, Maristela, Ana Vania, Marcus, Malcher, and especially

Deborah Aragdo and her family for continuous field, laboratorial and emotional support. I also

thank Livia Rangel-Vasconcelos whose work inspired the proposal of this dissertation.

Thanks also for Patricia Sampaio, Leslie, Leandra, Bob Buschbacher, Lucas Fortini,

Marisa and Carol Boaventura for the endless help provided throughout this whole process. Thank

you Cherie Arias, for enduring by my side at all times.

The research was financially supported by an Andrew Mellon Foundation grant to Daniel

Zarin and was conducted under cooperative agreements between the University of Florida,

Universidade Federal Rural da Amaz6nia, and Embrapa Eastern Amazon.









TABLE OF CONTENTS



A C K N O W L E D G M E N T S ..............................................................................................................4

L IST O F T A B L E S ......................................................................................................... ........ .. 7

LIST OF FIGURES ...................................................... .9

A B S T R A C T .......................................................................................................... ..................... 10

CHAPTER

1 SOIL MICROBIAL AND NUTRIENT DYNAMICS IN SEASONAL TROPICAL
SECONDARY FOREST: RESPONSES TO CHANGES IN RESOURCE
AVAILABILITY IN THE EASTERN AMAZON ................................................................11

In tro d u ctio n ............................................................................................................................. 1 1
Stu dy Site ............................................................................................ ........ 14
Experimental Design .............................................. ............................. 15

2 SEASONAL AND EXPERIMENTAL EFFECTS OF MOISTURE AND SUBSTRATE
AVAILABILITY ON MICROBIAL STRUCTURE AND COMPOSITION IN
SEASONAL TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON..........20

In tro du ctio n ............................................................................................................. ........ .. 2 0
M materials and M methods .............. .............................................................................. 2 1
Soil Sam pling and P processing ......................................... ........................ ................ 2 1
M icrobial B iom ass C and N ................................................................... ................ 22
M icrob ial B iom ass P .................................................. ............................................. 22
B acterial and Fungal D ensities ..................................... ........................ ................ 24
A M F R oot C olonization ... ....................................................................... ................ 25
AM F Spores Q uantification ................... .............................................................. 26
S statistic al A n aly sis................................................................................................................ .. 2 7
R esu lts ................................................................................... ..... .. ..................... 2 8
S ea so n al E ffects.............................................................................................................. 2 8
Irrig atio n E effects .............................................................................................................. 2 8
Litter Removal Effects ............................ ........... ........................ 29
C correlation A naly ses .................................................. ............................................. 29
D iscu ssio n .............................................................................................................. ......... .. 3 1
S ea so n al E ffects.............................................................................................................. 3 1
Irrig atio n E effects .............................................................................................................. 3 3
Litter Removal Effects .................................... ........ ...... .............. 34









3 SEASONAL AND EXPERIMENTAL EFFECTS ON MICROBIAL PROCESSES IN
SEASONAL TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON.......... 52

In tro d u c tio n ............................................................................................................................. 5 2
M materials an d M eth o d s .................................................... .............................. .. ... ............54
N et N -M ineralization and N itrification ...................................................... ................ 54
A cid Phosphatase A activity .................... ............................................................... 55
B a sa l R e sp iratio n .............................................................................................................5 6
S statistic a l A n a ly sis..................................................................................................................5 6
R results .................................................. .................. ....................57
S e a so n a l E ffe c ts ...............................................................................................................5 7
Irrigation E effects ....................................................................................................... 57
Litter R em oval Effects ............................................................................................... 58
C correlation A analyses ................................................................................................. 58
Discussion ...................................................... .................. 59
S e a so n a l E ffe c ts ...............................................................................................................5 9
Irrigation E effects ....................................................................................................... 60
Litter R em oval Effects ............................................................................................... 61

4 IRRIGATION AND LITTER REMOVAL EFFECTS ON SOIL NUTRIENT
AVAILABILITY IN A SEASONAL TROPICAL SECONDARY FOREST IN THE
EA STERN AM AZON ............................................................................................. . 73

Introduction .......................................................... .... ..................... 73
Study site and experim ental design ....................... ............................................ ...............75
M materials and m methods ............................................................................................. . 75
Ion exchange resins .......................................................................................... . 75
A nion exchange m em brane ..........................................................................................77
S o il C :N ratio s .................................................................................................................7 8
Soil w ater potential (SW P) ...........................................................................................79
S statistic a l a n a ly sis ...................................................................................................................7 9
R results .................................................. .................. ....................80
S e a so n a l E ffe c ts ...............................................................................................................8 0
Irrigation effects .............. .......................................................................................81
Litter rem oval effects ................................................................................................ 82
C o rrelatio n an aly se s ........................................................................................................8 3
Discussion ...................................................... .................. 84

5 C O N C L U S IO N S .................................................................................................................. 10 1

APPENDIX MICROBIAL BIOMASS COMPARISSONS...........................................105

L IS T O F R E F E R E N C E S ............................................................................................................. 10 8

B IO G R A P H IC A L SK E T C H .......................................................................................................116





6









LIST OF TABLES


Table page

1-1 Pre-treatment microbial biomass, basal respiration, and metabolic quotient ................. 17

2-1 F-statistics for irrigation vs. control for soil microbial biomass C, N and P, bacteria
and fungi colony form ing units, and their ratio ............................................ ................ 37

2-2. Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the irrigation experim ent...................................... ................ 37

2-3 F-statistics for litter removal vs. control for soil microbial biomass microbial
biomass C, N and P, bacteria and fungi colony forming units, and their ratio...............38

2-4 Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the litter removal experiment...............................................38

2-5 Pearson correlation analysis across treatments among microbial biomass microbial
biomass C, N and P, bacteria and fungi colony forming units, and their ratio...............39

2-6 Pearson correlation analysis across treatments among variables reported in chapter 2
and variables reported in chapters 3 and 4.................................................... ................ 40

2-7 Comparative seasonal and/or annual mean for extractable microbial biomass C, N
and P across differing tropical systems, site, depth and soil type................................42

3-1 F-statistics for irrigation vs. control for microbial processes .......................................64

3-2 Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the irrigation experim ent...................................... ................ 64

3-3 F-statistics for litter removal vs. control for microbial processes.................................65

3-4 Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the litter removal experiment ...............................................65

3-5 Pearson correlation analysis across treatments among N-mineralization rates,
nitrification rates, phosphatase activity and basal respiration ......................................66

3-6 Pearson correlation analysis across treatments among variables reported in Chapter 3
and variables reported in chapters 2 and 4.................................................... ................ 67

3-7 Seasonality, litter removal, and/or irrigation effects on N-mineralization,
nitrification, phosphatase activity, basal respiration, substrate induced respiration,
and soil C 0 2 efflux across studies................................................................ ................ 68

4-1 F-statistics for irrigation vs. control for soil solution nutrients and soil C:N .................90









4-2 Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the irrigation experim ent...................................... ................ 90

4-3 F-statistics for litter removal vs. control for soil solution nutrients and soil C:N. ............91

4-4 Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the litter removal experiment ............................... ................ 91

4-5 Pearson correlation analysis across treatments among ammonium, nitrate,
phosphorus availability, and soil C :N ........................................................... ................ 92

4-6 Pearson correlation analysis across treatments among variables reported in Chapter 4
and variables reported in chapters 2 and 3.................................................... ................ 93

5-1 Summary of ecosystem processes responses to intrannual variability of rainfall
seasonality and to resource manipulations...... .... ........................ 104









LIST OF FIGURES


Figure page

1-1 Seasonality of soil CO2 efflux and daily rainfall at the study site................................. 17

1-2 Soil C0 2 efflux at the study site: irrigation ................................................... ................ 18

1-3 Soil C02 efflux at the study site: litter rem oval............................................. ................ 18

1.4 Experimental plot design showing the distribution of treatments. .............................. 19

2-1 Soil microbial biomass: effects of rainfall patterns and irrigation.................................44

2-2 Soil microflora: effects of rainfall patterns and irrigation ............................................46

2-3 AMF root colonization and spores: effects of rainfall patterns and irrigation ..................47

2-4 Effects of rainfall patterns and litter-removal on soil microbial biomass.......................48

2-5 Effects of rainfall patterns and 1 litter removal on soil microflora ................................50

2-6 Effects of rainfall patterns and 1 litter removal on AMF root colonization and spores .....51

3-1 Microbial processes: effects of rainfall patterns and irrigation ....................................69

3-2 Effects of rainfall patterns and litter removal on microbial processes...............................71

4-1 Nutrient availability: effects of rainfall patterns on control and irrigation ...................... 95

4-2 Effects of rainfall patterns on control and litter removal on nutrient availability ............97

4-3 Soil C:N : effects of rainfall patterns and irrigation ...................................... ................ 99

4-4 Effects of rainfall patterns and litter removal on soil C:N..........................................100









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

SEASONAL AND EXPERIMENTAL EFFECTS ON MICROBIAL COMPOSITION AND
DYNAMICS IN A TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON, BRAZIL

By

Roberta M. Veluci-Marlow

August 2007

Chair: Daniel J. Zarin
Major: Forest Resources and Conservation

Tropical secondary forests are an increasingly important land cover in the Brazilian

Amazon, with 30 to 50% of the deforested area of the Brazilian Amazon in some stage of

abandonment. This study investigated water and nutrient constraints on microbial dynamics and

nutrient availability in a tropical secondary forest in the Eastern Amazon with manipulative

experiments dry-season irrigation and bi-weekly litter removal using frequent sampling to

capture seasonal and intra-annual fluctuations in Castanhal, Para, Brazil. Irrigation did not

consistently alter microbial dynamics, except for lowered NH4+ availability and fungal densities,

and increased phosphatase activity. Litter removal decreased microbial biomass C and P, N-

mineralization, phosphatase activity and NH4+ availability but increased NO3- availability. Intra-

annual variability was mainly driven by wet-up events in the dry season that were not minimized

by continuous irrigation (except for NH4+ availability), suggesting either that seasonal drought

may not constrain the availability of nutrients or that irrigation was insufficient to cause a more

significant effect. These results confirm the critical role of litterfall in tropical forest nutrient

cycling and the importance of fluctuations in soil moisture status to nutrient availability. How

these belowground results interact with aboveground processes including C uptake, is a fertile

area for future research and modeling.









CHAPTER 1
SOIL MICROBIAL AND NUTRIENT DYNAMICS IN SEASONAL TROPICAL
SECONDARY FOREST: RESPONSES TO CHANGES IN RESOURCE AVAILABILITY IN
THE EASTERN AMAZON

Introduction

This chapter serves as an overall introduction to the study site and experimental design

used for the research described in Chapters 2, 3 and 4. The main objective of this study was to

examine how changes in moisture and substrate availability alter microbial community processes

and nutrient dynamics in a seasonal tropical secondary forest following agricultural

abandonment. The overall hypothesis is that moisture and substrate availability constrain

microbial and nutrient dynamics. This study addressed the effects of the following:

* Intra-annual variation associated with the seasonality of rainfall or litterfall
* Wet-up and dry-down events
* Increased dry-season moisture availability induced experimentally by daily irrigation when
monthly PPT < 150 mm
* Reduced substrate availability, induced experimentally by bi-weekly litterfall removal.

In general, seasonal variation occurs in soil microbial C, N, and P in tropical forests but the

direction of seasonal change is not consistent across studies. Some studies show a negative

correlation between biomass C, N and/or P and soil moisture (Ross 1987; Singh et al 1989;

Srivastava 1992), whereas others have found a positive correlation between soil moisture and

microbial biomass(Luizao et al 1992; Marschner et al 2002). At the study site used for the

research described in this dissertation, pre-treatment results showed higher microbial biomass

and higher C:N ratio in the dry season (Rangel-Vasconcelos 2002; Table 1.1), and showed

elevated soil CO2 efflux during the wet season (Vasconcelos et al. 2004; Figure 1.1). Reduced

microbial biomass in the wet season may be due to a combination of lyses at the onset of the

rains and increased microherbivory as the wet season progresses (Lodge et al 1994). In

seasonally dry tropical forests, the microbial biomass may conserve nutrients in "biologically









active" forms during dry periods that are characterized by high microbial biomass and low

turnover. Subsequent nutrient releases can occur rapidly during wetter periods that are

characterized by low microbial biomass and high turnover; such "pulse" events can serve to

stimulate plant growth in nutrient-poor tropical forest and savannah (Singh et al., 1989).

Prolonged drought followed by flushes of rainfall can impose osmotic stress on

microorganisms and result in pulses of nutrients from microbial mineralization, death and/or

rapid turnover (Lodge et al., 1994; Wardle 2002). Dry season pulses may cause short-term

changes in microbial composition and "transient" pulses of nutrients(Yavitt and Wright 1996).

Irrigation to field capacity in an old-growth tropical forest on Barro Colorado Island,

Panama, has shown that alleviating moisture stress reduced the amount of forest floor mass

throughout the year, increased net decomposition and decay during the dry-season (Wieder and

Wright 1995), and caused changes in microbial composition (Comejo et al., 1994). Bacterial

counts tended to be higher during the drier months of January, February and March and lower in

April and May for irrigated plots. Fungi were depressed by irrigation and fungal counts were

greatest when conditions were driest (February and March) and declined after light rains in April

and heavier rains in May (Comejo et al., 1994).

A dry-season irrigation experiment at the study site has maintained the soil moisture status

relatively constant, favoring continuous decomposition (Vasconcelos et al., 2004). Soil CO2

efflux during the dry-season irrigation plots was 40% and 30% higher than in control plots in

2001 and 2002, respectively (Figure 1.2). By the end of the dry season irrigation period forest

floor may be left with lower litter quality and quantity as previously shown in other irrigation

studies (Yavitt et al., 1993; Wieder and Wright 1995; Yavitt and Wright 1996; Yavitt et al.,

2004), but there was no evidence of that occurring at this site (Vasconcelos et al. 2004).









Microbial biomass and its composition may also change in response to the quality and

quantity of organic matter (Orchard and Cook 1983; Wardle 2002). Substrate composition,

including C/N ratio, lignin content and soluble compounds are important factors regulating the

decomposition and mineralization rates of SOM (Mamilov and Dilly 2002), and the composition

of the microbial biomass (Wardle 2002).

Litter removal experiments in a tropical wet forest in Puerto Rico showed that monthly

changes in soil microbial biomass were not synchronized with aboveground litter inputs, but

rather preceded litterfall by one month (Ruan et al., 2004). There were also no correlations

between soil microbial biomass and soil temperature, moisture or rainfall. They suggested that

lower plant nutrient uptake or retranslocation of carbon and nutrients to roots and stems prior to

senescence could have triggered increases on microbial biomass one month prior to litterfall;

first by lessening microbial competition for nutrients and later by increasing root exudates.

Another study examining the effects of litter removal during 8 years in the White

Mountain National Forest, New Hampshire (a young northern hardwood forest), concluded that

litter removal had no effects on the microbial biomass, other than lowered respiration rates (Fisk

and Fahey 2001). They concluded that belowground C supply exerted greater control of forest

floor microbial processes than fresh leaf litter inputs.

During litter manipulations at the study site, Vasconcelos et al. (2004) showed that soil

CO2 efflux in litter removal plots was significantly lower than in control plots (Figure 1-3). This

trend is apt to worsen with time as labile SOC is consumed and not replenished.

Ongoing controlled manipulations at the study site, including (1) irrigating in the dry

season and (2) bi-weekly litter removal provided an excellent opportunity to study how moisture

availability and substrate removal affect microbial and nutrient dynamics. The following









chapters describe several experiments designed to examine seasonal and experimental effects on

microbial composition and structure (Chapter 2), microbial processes (Chapter 3), and nutrient

availability (Chapter 4).

Study Site

This study was conducted within the MANFLORA project, a collaborative research

program that includes the University of Florida and two Brazilian institutions, the Universidade

Federal Rural da Amaz6nia Federal Rural University of Amaz6nia UFRA, and EMBRAPA

Amaz6nia Oriental. Initiated in 1999, the overall goal of the project was to determine how

changes in resource availability affect forest recovery following agricultural abandonment.

Experimental treatments were dry-season irrigation and litterfall removal in tropical forest

regrowth and were implemented at the UFRA field station in Castanhal, Para, Brazil 10 19' S,

470 57' W. Mean SE annual rainfall received in the last 10 years in this area was 2539 280

mm, most of which falls between January and June. Mean daily temperatures fluctuate between

24 and 270C.

The soils are classified as Distrophic Yellow Latosol Stony Phase I (Tenorio et al., 1999)

in the Brazilian Classification, corresponding to Sombriustox in U.S. Soil Taxonomy. Soil

particle size distribution in the first 20 cm is 20% clay, 74% sand, and 6% silt. In the surface soil

(0 10 cm), pH is 5.0, total C is 2.2%, total N is 0.15%, C:N is 14.4, and Mehlich-1 extractable

phosphorus is 1.58 mg kg-1 (Rangel-Vasconcelos et al., 2005).

Forest regrowth, annual crops, and active and degraded pastures characterize the landscape

surrounding the field station. The stand under study was last abandoned in 1987 after multiple

cycles of shifting cultivation, beginning about sixty years ago when the old-growth forest was

first cleared. Each cycle included cultivation of corn, manioc, and beans, for one to two years









followed by fallow. Typical shifting cultivation cycles lasted seven to ten years (G. Silva e Souza

& O.L. Oliveira pers. comm.). The four most abundant overstory species are Lacistema

pubescens Mart., Myrcia sylvatica (G Mey) DC, Vismia guianensis (Aubl.) Choisy, and Cupania

scrobiculata Rich., representing 71% of all stems in the stand. In July 2000, mean stem density

was 2130 individuals per 100 m2, mean basal area was 13 m2ha-1, and mean height was 4.9 m for

the stand (Coelho et al., 2003).

Experimental Design

Plots were established in 1999 in 12-year-old forest regrowth. There were four replicate

plots for the irrigation treatment, four plots for the litter removal treatment, and four control plots.

Each plot was 20 x 20 m with a centrally nested 10 x 10 m measurement subplot (Figure 1-4).

Irrigation was applied in the late afternoon at a rate of 5 mm day-, for about 30 min,

during the dry seasons of 2001, 2002, 2003, and 2004. The amount of daily irrigation applied

corresponds to regional estimates of daily evapotranspiration (Shuttleworth et al., 1984; Lean et

al., 1996; Jipp et al., 1998). Irrigation water was distributed through tapes with microholes every

15 cm. In 2001, irrigation tapes were spaced 4 m from each other. In 2002 the distance between

tapes was reduced to 2 m to facilitate more even distribution of water (Vasconcelos et al., 2004).

In 2004, irrigation was implemented between 21st September 2004 and 19 January 2005.

We used rainfall and soil water potential data to define approximate boundaries for the dry and

wet seasons. The start of the dry season was defined by total rainfall less than 150 mm in the

previous 30 d and soil suction more negative than -0.010 MPa; the end of the dry season was

defined by total rainfall greater than 150 mm in the previous 30 d and soil suction less negative

than -0.010 MPa. Since the soil suction data were obtained on a weekly basis, we estimate that

the error in the location of seasonal boundaries is about 7 d (Vasconcelos et al., 2004).









In the litter removal plots, leaf and branch fall were removed from the forest with plastic

rakes every two weeks, beginning in August 2001 with the removal of the pretreatment litter

layer (538 35 g m-2, N = 4); C and N concentrations of the pretreatment litter layer were 41.0

+ 0.9 and 1.3 < 0.01%, respectively (N = 8). Total new non-woody litterfall removed during

the treatment period (August 2001 to May 2005) was 3051 111 g m-2 (Vasconcelos et al.,

2004).









Table 1-1. Microbial biomass (BMS) carbon (C-BMS), nitrogen (N-BMS), carbon to nitrogen
microbial ratio (C:N MICROBIAL), basal respiration (CO2-BMS), and metabolic quotient
(qCO2), from the study site (mean standard error). Data from Rangel-Vasconcelos
(2002).
VARIABLES Nov-00 (Dry-season) Apr-01 (Wet-season)
C-BMS ([tg C g-1 soil)* 924.8 + 60 425.25 59.04
N-BMS (tg N g-1 soil) 65.8 + 5.3 50.25 + 6.13
C:N MICROBIAL* 14.5 + 1.85 9.094 + 1.97
C02-BMS ([tg C-CO2 g-1 soil h-1)* 1.7 + 0.08 2.6 0.20
qCO2 0.002 0 0.007 0
(*) indicate significant seasonal differences P < 0.05


100-
80-
2E 60-
40-



02/00 05/00 08/00 11/00 02/01 05/01 08/01 11/01 02/02 05/02 08/02 11/02 02/03
Date
Figure 1-1. Seasonality of soil CO2 efflux (mean + se) and daily rainfall at the study site (N= 12).
White and black horizontal bars indicate dry and wet seasons, respectively. Data from
Vasconcelos (2006).














c
E
0 0
00
0 E
(/) =


02/00 05/00 08/00 11/00 02/01 05/01 08/01 11/01 02/02 05/02 08/02 11/02 02/03
Date
Figure 1-2. Effects of dry-season irrigation on soil C02 efflux (roots and microorganisms) in
comparison to control plots at the study site (N= 12). White and black horizontal bars
indicate dry and wet seasons, respectively. Gray bars represent irrigation period in the
dry season. Standard errors are represented ( se). Data from Vasconcelos (2006).


x
' E
0 0

0 E
(U) =L


0 i 1
02/00 05/00 08/00 11/00 02/01 05/01 08/01 11/01 02/02 05/02 08/02 11/02 02/03
Date
Figure 1-3. Effects of litter removal on soil C02 efflux (roots and microorganisms) in
comparison to control plots at the study site (N= 12). Litter removal started on August
2001 (indicated by vertical line). Standard errors are represented ( se). Data from
Vasconcelos (2006).





































20m 10m


!*-path


20hrn


S T A T I 0 N


R 0 A D


Figure 1-4. Experimental plot design showing the distribution of treatments.


LEGEND:


I i









CHAPTER 2
SEASONAL AND EXPERIMENTAL EFFECTS OF MOISTURE AND SUBSTRATE
AVAILABILITY ON MICROBIAL STRUCTURE AND COMPOSITION IN SEASONAL
TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON

Introduction

Moisture and litterfall seasonality influence the seasonal variation of soil microbial C, N

and P in tropical forests but the direction of seasonal change is not consistent across studies.

Some studies have shown a negative correlation between biomass C, N and/or P and soil

moisture (Ross 1987; Singh et al 1989; Srivastava 1992), while others found a positive

correlation (Luizao et al 1992; Marschner et al., 2002). The relationship between soil microbial

biomass and litterfall seasonality is easily confounded with other variables, including root

exudates (Ruan et al., 2004), shifts in microbial community structure (Cornejo et al., 1994,

Marschner et al., 2002, Li et al., 2004 and 2005), N and/or P limitations (Fisk and Fahey 2001,

Davidson et al., 2004), soil type (Feighl 1995, Villar et al., 2003, Cleveland 2004), nutrient

pulses (Lodge et al., 1994), and litter quality (Hodge et al., 2000).

For this chapter, I measured the responses of microbial biomass C, N, and P and microbial

composition (bacterial and fungal densities, mycorrhizal root-colonization and spore availability)

to seasonal changes and wet-up events. Frequent drying and rewetting of soils favor is the

portion of the microbial community best adapted to coping with that stress (Gollan et al., 1992).

Fast growing microbes capable of rapid growth on the labile substrates released from litter or

microbial leaching into the soil during a rewetting event may be predominantly composed of

fungi and bacteria capable of withstanding sudden changes in soil water potential (Gollan et al.,

1992). I also examine substrate and water constraints on microbial biomass and composition

within two ongoing manipulative experiments designed to alter resource availability-- dry-season

irrigation and litterfall removal.









Irrigation to field capacity in an old-growth tropical forest in Barro Colorado Island,

Panama, has previously shown that increasing moisture caused changes in microbial composition

(Cornejo et al., 1994). Fungal densities decreased, while bacterial densities first increased and

then decreased in response to irrigation, likely as a result of reduced litter quality since more

recalcitrant material is degraded by specialized groups of fungi.

Litter removal has been previously employed to examine the consequences of reducing

substrate availability on microbial biomass and composition in a tropical wet forest in Puerto

Rico (Ruan et al., 2004). Their results showed that monthly changes in soil microbial biomass

were not synchronized with aboveground litter inputs, but rather preceded litterfall by one

month. They found no correlations between soil microbial biomass and soil temperature,

moisture or rainfall. Another study in Puerto Rico showed that microbial biomass C was

significantly depressed by litter removal (Li et al., 2004), including bacterial and fungal biomass

(Li et al., 2005). In another experimental litter manipulation in Panama, litter removal changed

fungal species composition and diversity, decreased soil fauna, decomposition rates, and nitrogen

and phosphorus from incoming litterfall, while litter addition produced no corresponding

increases for those variables (Sayer et al., 2006).

Materials and Methods

Study site and experimental design are described in Chapter 1.

Soil Sampling and Processing

During each of 13 monthly harvests, seven soil cores were taken per plot using a bulk-

density corer with 6 cm diameter and 5 cm depth. These samples were composite by plot, sifted

through a 2 mm mesh, placed in double-folded, tightly closed plastic bags, transported to the lab

in a cooler no longer than 4 h after harvest, and stored at 4C until analyzed. Gravimetric water









content was determined using parallel subsamples dried at 1050C for 24 h so that results could be

expressed on a dry weight basis.

Microbial Biomass C and N

The "fumigation-extraction" method was used to estimate microbial biomass carbon and

nitrogen (Vance et al., 1987). Chloroform fumigation (72 h) and direct extraction (agitation at

180 rpm in 0.5 M K2SO4 followed by filtration through Whatman No. 42 paper) was performed

on 20 g subsamples (Vance et al. 1987). Unfumigated subsamples were similarly extracted.

The C present in fumigated and unfumigated extracts was determined colorimetrically by

adaptation of a method described by Islam and Weil (1998). Aliquots of 2 mL were mixed with

0.75 mL of 0.17M K2Cr207 and 2 mL H2SO4 in a 75 mL Pyrex digestion tube. The covered

tubes were manually agitated and then heated to 1500C for 10 min with glass beads. The

samples were allowed to cool, 10 mL distilled water was added, and then the samples were

analyzed spectrophotometrically at 590 nm, comparing concentrations to a standard curve with

samples containing 0-2 mg C as sucrose dissolved in 0.5 M K2SO4. Microbial biomass-C was

estimated from the chloroform-labile C, using a Kec factor of 0.35 (Voroney et al., 1991).

Organic-N in the extract was digested using hydrogen peroxide and sulfuric acid at 260 C, and

then determined colorimetrically (Mulvaney 1996). Microbial biomass N was estimated from the

chloroform-labile N, using KN factor of 0.54 (Brookes et al., 1985).

Microbial Biomass P

Microbial biomass phosphorus was determined using a modification of the method

described by Lajtha et al. (1999) for Long Term Ecological Research (LTER). Subsamples were

analyzed in duplicate for fumigated and non-fumigated samples. Conversion of the total

dissolved, and organic dissolved phosphorus in fumigated and non-fumigated samples to

inorganic phosphorus was done by Acid Persulfate Digestion and then determined









colorimetrically as described by Murphy and Riley (1962). Microbial biomass P was estimated

using a Kp factor of 0.40 (Lajtha et al., 1999).

Extraction: One gram of soil was weighted in duplicates for each sample and transferred

to 50 mL centrifuge tubes. All samples were agitated with DI-water in centrifuge tubes for 16 h

with 2 pre-conditioned BDH membrane strips (Gallard-Schlesinger Ind., Product # 55164 2S,

Plainview, NY, 1-888-686-3454), to remove any inorganic P. Strips were then removed and

samples were centrifuged for 10 minutes at 5,000 rpm to discard supernatant. Non-fumigated

samples were stored at room temperature for 24 h while paired samples were fumigated using

1 mL of chloroform (CHCl3) in each tube and placed uncapped in vacuum desiccators. A beaker

was placed in the center of desiccators containing 25 mL of chloroform, closed air tight,

vacuumed for 10 minutes and then left under vacuum for 24 h. After this period, the vacuum was

released, chloroform volatiles eliminated, and 30 mL of 0.5 MNaHCO3 (sodium bicarbonate

solution) pH 8.5 was added to all samples and left shaking for 16 hours in a platform shaker.

Samples were then centrifuged for 10 minutes at 5,000 rpm and supernatant reserved for

persulfate digestion and analysis.

Persulfate digestion: A mixture of 0.8 g of potassium persulfate (K2S208), 10 mL of 0.9

MH2SO4 and 5 mL of sample extract was added to 50 mL volumetric flasks. Flasks were capped

with aluminum foil and placed in an autoclave at 1210 C, 17 psi for 50 minutes. Murphy & Riley

(1962) colorimetric analysis for inorganic P was developed directly in each flask after samples

were allowed to cool down. All samples were read under a UV spectrophotometer set at 885 nm

(Murphy & Riley 1962).









Bacterial and Fungal Densities

The serial dilution technique was used to quantify densities of fungi and bacterial colony

forming units (CFUs) at the study site (Comejo et al., 1994). The method measures only a small

portion of the total population. Nonetheless, it is quite useful for studying changes in the

population that grows on the chosen medium. Although more recent procedures offer whole

community analysis through microbial lipid analysis, substrate utilization, enzyme assays, and

nucleic acid analyses, among others (Sinsabaugh et al. 1999), none was logistically appropriate

due to the unavailability of material and equipment at the study site.

Each month, at the day of soil harvest, subsamples from composite soil samples from each

plot were separated for analysis. The soil subsamples were transported in open plastic bags and a

3 g sample from each plot was weighed, transferred to a 200 mL Erlenmeyer, and agitated in a

platform shaker at 300 rpm for 20 minutes in 97 mL of sterile water. All materials used during

the procedure were autoclaved for lh at 1210 C, 17 psi., and prepared inside of an ultraviolet-

laminar flow chamber to avoid potential contamination. After 3 months of preliminary tests the

best serial dilutions were -1 for fungi and -2 for bacteria. A 100 [.L aliquot of the appropriate

dilution was added to 20 mL of solid medium, either bacterial or fungal, and swirled. Each liter

of bacterial medium included 0.5 g dibasic-potassium phosphate (K2HPO4), 0.2 g of magnesium

sulfate (MgSO4.7H20), 0.01 g of Fe-Na-EDTA, 0.25 g of egg albumin, 1 g of glucose, 1 g of

yeast extract (levedura), and 20 g of TSA-agar dissolved up to volume in a water-bath, then

autoclaved. Prior to agar addition, the medium was brought up to pH 6.8. Each liter of fungal

medium included monobasic-potassium phosphate (KH2PO4), 1 g of magnesium sulfate

(MgSO4.7H20), 5 g of peptone, 10 g of dextrose, and 20 g of TSA-agar all dissolved up to

volume in a water-bath and then 3.3 mL of 0.1% Rose Bengal was added. After autoclaving, 0.3









g of streptomycin was added under the ultraviolet-laminar flow chamber and mixed slowly to

avoid bubble formation. Analysis was run in triplicates per plot (n = 12 per treatment), and

several blanks were incubated with either medium with sterile water under the same conditions

to check for possible contamination. No contamination was found during the study. Gravimetric

water content was determined by drying soil subsamples at 1050C for 24 h so that results could

be expressed on a dry weight basis. Only one type of agar medium was used for either bacteria or

fungi, incubated at 280C for 1 and 2 days, respectively.

AMF Root Colonization

During each soil sampling at the study site, fine roots (< Icm) were collected to quantify

arbuscular mycorrhizal fungi (AM) colonization (Giovannetti and Mosse 1980; Johnson et al

1999). The process was divided in five phases: i) selecting roots; ii) conserving roots until

analysis; iii) clearing roots; iv) staining roots and finally, v) preparing slides for quantifying

AMF colonization. Since AMF can easily degrade after their removal from soil, after soil was

sifted and roots were separated (1-2 g of roots < 1mm diameter), they were carefully washed in

tap water and conserved in a FAA solution (formalin, ethanol and CH3COOH glacial acetic

acid) to conserve the root until analysis. For clearing, roots were soaked and resoaked in KOH

10%, until very clear. Then, after transfer to a water-bath for 15 minutes at 900C, KOH was

discarded; roots were rinsed with water several times and then soaked in 2% HCl for 10 minutes

prior to staining. Tryplanblue (0.01 0, 05%) was used to stain and lactoglicerol to remove

excess stain (a combination of glycerin, lactic acid and water, 1:1:1). Using a sharp blade and

forceps, 20-25 root segments 1cm in length were separated to prepare microscope slides. Slides

were prepared in duplicates per plot. Finally, the number of inoculations was verified using a

compound or dissecting microscope (fitted with a hairline graticule in the eyepiece). Results









were expressed as the average number of AMF per linear cm of root in two slides per plot (n =

50).

AMF Spores Quantification

The method for quantification of mycorrhizal-fungi spores in soil was derived from a

composite of the methods suggested by Gerdemann and Nicolson (1963), Lopes et al. (1983) and

Johnson et al. (1999), with some modifications. A 30 g soil sample was transferred to a beaker

with water, mixed with a stirring rod to separate the supernatant, and then filtered through two

sieves a coarser sieve (0.71 mm) on top of a thinner sieve (0.053mm). This process was

repeated 4 times (add water to the soil, mix, wait for settling and then filter through the sieves).

In the end, the material retained on the thinner sieve was collected (residual with spores), and the

remaining soil material discarded (material remaining on the coarser sieve). During this process,

the residual with spores was carefully transferred to a centrifuge tube and centrifuged at 2.000

R.P.M. for 3 minutes. Supernatant was then discarded and approximately 40 mL of 45% sucrose

solution was added to the remaining residual containing the spores, and centrifuged again. First

spore collection consisted of decanting the sucrose solution supernatant through a smaller sieve

(0.021mm) where minuscule spores were retained. Thereafter, spores were rinsed with water to

remove excess sucrose and then transferred to a capped Nalgene bottle with water. Using the

same centrifuged sample, sucrose was added for the second time and the same process was

repeated. The second spore collection of each sample was then transferred to the respective

bottle of the first collection and samples were aspired through a vacuum pump through a porous

porcelain funnel where grid filter paper was inserted and spores retained for count. Drained filter

paper with spores was transferred to a labeled Petri dish slightly wet for best cohesion and read

under a compound or dissecting microscope. Results were expressed as the number of spores per

30 g of soil.









Statistical Analysis

The SAS System for Windows V8 (2) was used for statistical analyses. All response

variables of interest were log-transformed to meet the model assumptions of normality. PROC

MIXED was used using a repeated measures analysis with a heterogeneous-autoregressive error

structure. This structure allowed modeling within sample correlation over time and calculation of

individual error variances for each sampling date. Linear models were fitted on the variables

microbial biomass C, N, P, fungal and bacterial densities (and their ratio) with the following

effects: season, date, treatment, treatment by season, treatment by date, plot and plot by date.

CONTRAST statements were used to determine the significance of each fixed effect for each

pair of treatment comparisons (i.e., control vs. irrigation and control vs. litter removal) and least-

squares means were used to compare treatments and control means for the effects of season,

treatment and treatment by season interaction on microbial biomass nutrients and their ratios, and

on microbial densities and their ratios. For root colonization by AMF and spore counts, PROC

MEANS was used for paired comparisons (Control vs. irrigation or Control vs. Litter removal)

for each harvest date on log transformed variables. A new variable was created (e.g., DIFF)

containing the differences between the paired variables and then the T and PRT options of PROC

MEANS was used to test whether the mean difference significantly differed from zero.

Within each treatment (control, irrigation, litter removal), Pearson correlation analyses

were used to explore the bivariate relationships of results reported in subsequent chapters with

those reported here. Specifically, I tested for correlations between microbial biomass C, N, P

(and their ratios), fungal and bacterial densities (and their ratio), and : mineralization rates,

nitrification rates, phosphatase activity and basal respiration (Chapter 3), NH 4, NO-3 and PO-4

availability, total soil carbon and nitrogen (Chapter 4) and rainfall and soil water potential.









Results

There were no consistent effects of dry-season irrigation on the measured variables

(Figures 2-1, 2-2, 2-3) but litter removal significantly reduced microbial biomass carbon and, to

a lesser extent, microbial phosphorus (Figure 2-4), and bacterial and fungal densities (Figure 2-

5), and had mixed but significant effects on arbuscular mycorrhizal fungi root-colonization

(Figure 2-6).

Seasonal Effects

Microbial biomass C (MBc) was higher in the dry than wet season in control plots (Table

2-2), particularly after wet up events following an extended dry period (Figure 2-la). There was

no significant intra-annual variability for bacterial densities, but fungal densities were

significantly higher in the dry than wet seasons (Table 2-1 and 2-2).

Irrigation Effects

Microbial biomass C (MBc) was elevated in the dry season following an extended dry

period (Figure 2-la, Table 2-2). MBc, microbial biomass nitrogen (MBN) and phosphorus (MBp)

varied significantly in response to date and the treatment by date interaction; MBc was also

significantly affected by season and the treatment-by-season interaction (Table 2.1). The MBp

pool was lower in irrigation than control plots during the first harvest of the dry season on Oct. 14

2004, then higher in irrigation plots towards the end of the dry season on Dec. 15 2004, following

a substantial wet-up event after a long dry period (Figure 2. le).

Bacterial and fungal CFUs varied significantly in response to all main effects and their

interactions, except for the main effect of treatment which was significant for fungal CFUs only.

There was also a significant effect of date and the treatment by date interaction on B:F ratios

(Table 2.1; Figure 2.2). Bacterial CFUs were higher in the dry than wet season in irrigation plots

whereas fungal CFUs were higher in the dry season in control plots and lower in irrigated than









control plots in the dry season (Table 2-2). As a result, B:F ratio was higher in irrigated than

control plots in the dry season (Table 2-2). There was an abrupt increase in bacterial CFU during

a wet up event in December, even in irrigated plots. However, sampling after the first rains

following extended drought in March had no apparent effects on CFU counts. Increased B:F

ratios towards the end of the wet season resulted from low fungi-CFU and relatively constant

bacteria-CFU.

AMF root colonization in irrigated plots was significantly higher than in control plots

during one sampling in the dry-season, but there was no effect in three wet-season samples.

Continuous sampling showed a steady decrease in AMF spores in control plots towards the dry

season but there were no significant differences in AMF spore counts between irrigated and

control plots (Figure 2.3).

Litter Removal Effects

MBC, and to a lesser extent MBP, were significantly reduced by litter removal; MBN was

largely unaffected by this treatment (Figure 2-3; Tables 2-3, 2-4). Bacterial and fungal CFUs

were reduced by litter-removal, and fungal CFUs were higher in the dry than wet season. As a

result, bacterial to fungal ratios were higher in the wet season (Figure 2-5, Tables 2-3, 2-4).

AMF root colonization in litter removal plots was significantly higher than in control plots

in two samples and lower in one (Figure 2.6a); there were no treatment effects on AMF spores

(Figure 2.6b).

Correlation Analyses

Among the variables reported in this study, MBc in control and irrigation plots was

negatively correlated with microbial biomass P (Table 2-5A and B). Under litter removal plots,

MBc was positively correlated with fungi and correspondingly correlated with bacterial to fungal

ratios (Table 2-5C).









Including variables measured in Chapters 3 and 4, MBc was negatively correlated with

basal respiration and soil water potential across treatments (Table 2-6), with N-mineralization

rates and phosphatase activity in control plots (Table 2-6A), and with NH4+ under irrigation plots

(Table 2-6B). In the contrary, MBc was positively correlated with NH4+ and P04- in litter

removal plots (Table 2-6C).

MBN was positively correlated with fungal densities across treatments, and additionally

with bacterial densities in irrigated plots only (Table 2-5). The variation in MBN across

treatments was also attributed to phosphatase activity (Table 2-6). MBN was also positively

correlated with soil water potential under irrigation (Table 2-6 B), and negatively correlated with

basal respiration and soil C:N in litter removal plots (Table 2-6C).

The increase in microbial biomass P in control plots was associated with decreases in

fungal densities, and in litter removal plots with increases in bacterial densities (Table 2-5).

Microbial phosphorus was not correlated with microbial composition in irrigated plots (Table 2-5

B), but was positively correlated with nitrate availability, phosphatase activity, rainfall and soil

water potential (Table 2-6B).

Under litter removal, MBP was also positively correlated with phosphatase activity and

basal respiration. Microbial biomass C:N was negatively correlated with phosphatase activity

and soil water potential across treatments (Table 2-6), positively correlated with NO3-

availability across treatments, and negatively correlated with NH4+ availability in irrigation and

litter removal plots (Table 2-6). Microbial biomass N:P was negatively correlated with rainfall

and NO3-, and positively correlated with NH4+ in irrigated plots. Under litter removal, MBN:P was

negatively correlated with basal respiration, soil C:N and rainfall (Table 2-6). Biomass C:P ratios

was negatively correlated with basal respiration and mineralization in control plots, with basal









respiration in litter removal plots, and with phosphatase and soil water potential in irrigation

plots (Table 2-6). Under litter removal, MBc:p was positively correlated with phosphorus

availability (Table 2-6).

Changes in bacteria to fungal ratios were positively correlated with bacteria in control plots

only; negatively correlated with fungi across treatments (Table 2-5), and with N-nitrification

irrigated plots (Table 2-6). Fungal and bacterial densities were positively correlated with NH4

availability and N-mineralization rates in control plots (Table 2-6). Fungal densities were also

positively correlated with N-mineralization and N-nitrification rates in irrigated plots, and with

N-mineralization and phosphatase in litter removal plots. In addition to the positive correlation

with N-mineralization, bacterial densities were also positively correlated with NO3-, and rainfall

in litter removal plots.

Discussion

Seasonal Effects

Previous assessments of microbial biomass in tropical forests tend to report one sampling

period in the dry season and one in the wet season (Table 2.7). In the present study, microbial

biomass C, N and P was relatively low compared to previous reports.

Pre-treatment results at the study site also revealed greater microbial biomass carbon and

nitrogen in the dry than wet season (Rangel-Vasconcelos et al., 2005). Monthly sampling

throughout the year confirms the same seasonal trend for microbial biomass carbon across all

treatments, but the opposite for biomass phosphorus in litter removal plots.

The seasonal effect on microbial biomass C was associated with decreased basal

respiration rates as a result of lowering soil water potential (Chapter 3). Lower microbial activity

was accompanied by concomitant decreases in N-mineralization and phosphatase activity as

microbial biomass C accumulated. These results support a very common conserving strategy in









tropical forests with low nutrient availability, i.e. high immobilization of nutrients in the

microbial biomass in the dry season and subsequent release in the wet season when plant growth

is greatest (Singh et al., 1989). In addition, higher fungal densities in the dry than wet season but

no significant changes in bacterial densities may indicate a predominantly fungal community in

that period. This may help explain the corresponding effects that fungal densities had on

microbial biomass N and P, microbial nutrient ratios, N-mineralization rates and NH4+

availability.

Increased root activity in the dry-season may also explain the boost on microbial biomass

C shown in this study. Vasconcelos (2006) showed that fine root mass density across treatments

was approximately twice as high in the dry season of 2004 as in the wet season of 2005 at the

study site, which may lead to higher microbial activity and leakage of carbon-rich exudates that

becomes available to the microbial community.

Contrary to my results, greater microbial biomass estimates corresponded to greater

microbial activity in an oxisol forest in Costa Rica; but the active portion of the microbial

biomass remained relatively low regardless of soil C availability (Cleveland et al., 2004).

Alternatively, the persistent increase in microbial biomass C followed by decreases in microbial

respiration may indicate the microbial biomass increased efficiency by lowering their

metabolism in that period (less CO2 evolved per unit biomass). In a seasonally dry tropical forest

near Manaus (AM), a large proportion of the microbial biomass died off in the dry season, and

the activity of remaining microorganisms were reduced (Marschner et al., 2002). These effects

could be enhanced by decreased root exudation, changes in microbial community structure and

substrate utilization patterns.









In general, soil microbial C or N co-varied with microbial C:N, and C:N ratios higher in

the dry season across treatments (~ 30:1 in control plots and in irrigated plots, and 22:1 in litter

removal plots), than in the wet season across treatments (-10:1), but maintained an intra-annual

range similar to other tropical studies (Table 2-7). The seasonal difference in microbial C:N

ratios suggest that N availability was perhaps recycled more efficiently in the wet season by the

microbial biomass, than carbon.

Irrigation Effects

Although decomposition rates (k) were 2.4 times higher in irrigated than control plots

during the sampling period (Vasconcelos 2006), irrigation caused no consistent enhancement of

microbial biomass C, N, and P, or bacteria:fungal ratios, but significantly reduced fungal

densities.

Overall, fungal densities were lower in irrigated than control plots, but the decrease in

fungal densities in irrigated plots was significantly greater in the wet season. The long-term

effects of irrigation have probably affected fungal counts in agreement with studies that suggest

that fungi thrives best under lower soil water potentials (Cornejo et al., 1994). Nonetheless,

greater fungal and bacterial abundances in irrigated plots in the dry-season coincided with higher

microbial biomass C, higher litter decomposition and fine root mass growth in the same period

(Vasconcelos 2006); which may also explain the positive influence of fungal densities on N-

mineralization and nitrification rates in irrigated plots.

Although there were no effects of dry-season irrigation on fine root mass growth

(Vasconcelos 2006), the indirect effects of greater root C inputs may have enhanced microbial

abundance and therefore the microbial biomass. However, there were no concomitant increases

in microbial activity, as can be observed by the lowest rates of microbial basal respiration in the

same period (Chapter 3). Similar results were found by William and Rice (2007) under









continuously wet soils in an upland tallgrass prairie. The decrease in fungal and bacterial

densities in irrigated plots in the wet season could be a result of decreased litter quality due to

continuous decomposition in the dry season, but irrigation did not change litterfall quality

(Vasconcelos et al., 2006).

Microbial biomass phosphorus was not affected by irrigation but responded well to wet up

events, explaining the positive correlations with rainfall and soil water potential. The positive

correlation between MBp and phosphatase activity may indicate that biotic demand for P drives

phosphatase activity, and hence P mineralization (Vitousek et al., 2002), but this correlation was

observed only in control plots (Chapter 3 and 4), which suggests that irrigation minimized this

potential. The lack of correlation between MBp and phosphatase activity in irrigated plots may be

due to higher phosphatase activity in irrigated than control plots (Chapter 3), which could

alleviate the microbial demand for P through mineralization, by the higher potential of P

available through solubilization of inorganic/mineral P.

In general, the relationship among MBp, microbial N:P and C:P across treatments,

confirms the intertwined covariance among C, N and P-demand for microbial growth, and to

enzymatic activities (Paul and Clark 1997, Treseder and Vitousek 2001). Similar results was

found by Cleveland et al. (2004), with the highest microbial C:P ratio occurring when microbial

biomass C was at a maximum. The negative relationship between fungal densities and MBp in

control plots may indicate different P-demands as microbial composition shifts or as fungi

become more predominant.

Litter Removal Effects

Microbial biomass C and P were significantly lower in litter removal than control plots, but

only microbial biomass C varied seasonally. Similar to the other treatments, increased MBc in

the dry season seems related to wet up events, to the indirect effects that lower soil water









potential had on basal respiration, and to increased fine root abundance and root C inputs in that

period (as described in the previous section). The decrease in MIBc in litter removal plots stems

from the lack of substrate to sustain aboveground and belowground microbial decomposition and

microbial abundance.

Bacterial and fungal abundances were lower in litter removal than control plots, but in

agreement with other similar studies, fungal densities were higher within the dry than wet

seasons (Cornejo et al., 1994; Willian and Rice 2007).

The concomitant decrease in MBp, basal respiration and phosphatase activity between

October and December in the dry season suggests that lower microbial activity may result in

lower phosphatase activity, and perhaps, less soluble P for microbial uptake. Increased

mycorrhizal associations in response to litter removal may indicate that investments in

mycorrhizal associations were beneficial; alleviating the demand for P or N, since these

associations can be highly costly for the plant (Paul and Clark 1997; Rilling 2004).

This result suggests that arbuscular mycorrhizal fungi could be an important nutrient

conduit in this tropical secondary forest, consistent with their role in nutrient sequestration in

nutrient poor soils (Rilling et al., 2004; Hart and Trevors 2005, Sayer 2006). The deprivation of

the litter layer and consequent impoverishment of the mineral soil could trigger investments in

mycorrhizal associations; but the extent to which this impoverishment will impair microbial

activity remains unclear. A review by Sayer (2006) showed that litter manipulation affected

fungal growth and diversity by altering decomposers substrate, changing microclimate, leaching,

availability of nutrients, and spore abundances although the direction of those changes were not

consistent and often site-specific. In one of those studies, the removal of the organic layers

returned the soil to an earlier successional stage, resulting in greater abundance and diversity of









fungal species. Forest succession studies in eastern Para showed that AMF spores and

mycorrhizal infections decrease with increasing forest stand age and become similar to the

mature forest within 8 years of secondary forest succession (Carvalho et al., 2004, Chu and

Diekmann 2002).

Overall, biomass C:N was lower in litter removal than control plots, suggesting that the

microbial biomass was not as N-deprived as expected in the absence of aboveground litter (Table

2-7). In the same period, litter removal substantially decreased N-mineralization (Chapter 3), and

litter nitrogen concentration, although this reduction had not subsequent effects on litterfall mass

(Vasconcelos 2006). These findings suggest microbial access to N from an alternative nutrient

pathway. Increased NO-3 availability in litter removal plots (Chapter 4) may have compensated

for the increased biomass-N in relation to biomass-C in litter removal plots. Greater mycorrhizal

infections in litter removal plots may have contributed to increased microbial N-availability but

imposed higher costs for the symbiotic host (Paul and Clark 1997), as observed by the decrease

in litter N. In a nearby secondary forest, N amendments significantly improved tree growth,

while P or N+P amendments did not, suggesting that N was the most limiting nutrient (Davidson

et al., 2004).

Although soil C:N ratios were not significantly affected by treatments during the sampling

period (Chapter 4), correlation results suggested that increases in microbial biomass N and N:P

corresponded to decreases in soil C:N in litter removal plots. This may indicate that soil nitrogen

stocks may sustain microbial growth in the absence of litter substrate, and for that reason there

were no litter removal effects on MBN. These results may have further implications in litter

removal plots since without litter input the fraction of available carbon and nitrogen may

diminishes over time, and become critical for the microbial community.









Table 2-1. F-statistics for the effects of treatment (Irrigation vs. control), seasons (Wet vs. Dry),
date, and the interactions between treatment by season and treatment by date on soil
microbial biomass carbon, nitrogen and phosphorus (MBC, MBN and MBP), bacteria
and fungi colony forming units (B and F CFU's) and their ratio. = P < 0.05; ** = P
< 0.01; ***=p < 0.001.
Irrigation vs control
Variable Season Date Treat Treat x Season Treat x Date
MBc 50.2*** 9.1*** 1.2 17.2** 16.2***
MBN 4.4 40.6*** 2.1 2.0 39.0***
MBp 1.6 19.9*** 0.3 0.7 12.1***
B (CFU's) 7.4* 712.6*** <0.1 3.8* 3608.8***
F (CFU's) 10.0* 166.1*** 6.2* 5.6* 815.3***
B:F 2.0 49.1*** 3.7 2.4 267.0***

Table 2-2. Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the irrigation experiment. (Lower and upper bounds of the
95% confidence interval are provided in parenthesis). Lower case letters indicate
differences at P < 0.05 between treatments both annually and within each season.
Upper case letters indicate significant seasonal differences at P < 0.05 within each
treatment.


MBc
(mgC/K- soil)


Bacteria
(CFU/-1 dry soil)


Fungi
(CFU/-1 dry soil)


Bacteria:Fungi


Treatment LSM
299 162754 1034a 148
(269-328) (138862-190757) (841-1270) (117- 187)
275 161135 1499b 108
(246-305) (138291-187752) (1220-1841) (86-135)
Treatment by Season LSM
n


391A
Irrigation (340-441)

Control 63
(313-413)
Wet Season
241B
Irrigation 24)
(208-274)
221B8
Control _
(188-253)


211081A
(170979-260589)
168047
(137974-204674)

125492B
(99756-157868)
157314
(125051-197899)


1283a
(1000-1646)
1906bA
(1485-2445)

8322a
(604-1148)
11785aB
(855-1625)


Irrigation

Control


Dry Seaso


146a
(105-202)
88b
(64-119)

151
(110-208)
133
(97-183)









Table 2-3. F-statistics and associate significant levels (p-value) for the effect of treatment (Litter
removal vs. control), seasons (Wet vs. Dry), date, and the interactions between
treatment by season or treatment by date on soil microbial biomass carbon, nitrogen
and phosphorus (MBc, MBN and MBp), bacteria and fungi colony forming units (B
and F CFU's) and their ratio. = P < 0.05; ** = P < 0.01; ***=p < 0.001.
Litter removal vs. control
Variable Season Date Treat Treat x Season Treat x Date
MBc 40.06 ** 29.75*** 11.4 ** 17.16** 74.27***
MBN 1.04 44.45 *** 0.01 0.63 185.42 ***
MBp 2.28 16.91*** 4.89* 2.69 11.2 ***
B (CFU's) 0.12 7.94 *** 9.87 ** 3.52 3627.63 ***
F (CFU's) 16.81** 4.05 *** 6.04 7.69 ** 810.60 ***
B:F 12.67 ** 3.52 *** 0 4.49 278.44 ***

Table 2-4. Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the litter removal experiment. (Lower and upper bounds of
the 95% confidence interval are provided in parenthesis). Lower case letters indicate
differences at P < 0.05 between treatments both annually and within each season.
Upper case letters indicate significant seasonal differences at P < 0.05 within each


treatment.
MBc
(mgC/Kg- soil)


MBp
(mgP/Kg- soil)


Bacteria
(CFU/g-1 dry soil)


Fungi
(CFU/g-1 dry
soil)


Treatment LSM
204a 0.46a 114576 a 1038a 109
(174-233) (0.40-0.52) (98198-133686) (845-1276) (87-137
275b 0.56b 162592b 1499b 108
(246-305) (0.50-0.62) (139541-189449) (1220-1841) (86-135)
Treatment by Season LSM


Dry Season
Litter
removal
Control
Wet Season
Litter
removal
Control


276aA
(226-327)
363bA
(313-413)

158aB
(125-191)
2213bB
(188-253)


0.41
(0.33-0.49)
0.53
(0.45-0.61)

0.50
(0.41-0.57)
0.58
(0.49-0.66)


115128a'
(94525-140221)
168181b
(138084-204838)

114119
(90715-143561)
157329
(125064-197919)


1482A
(1149-1912)
1906A
(1485-2445)

727B
(527-1003)
11785 B
(854-1624)


Litter
removal
Control


B:F


76A
(55-105)
88
(64-119)

156
(113-214)
132B
(96-182)











Table 2-5. Pearson correlation analysis for each treatment Control, Irrigation and Litter removal, among variables reported in this chapter:
microbial biomass carbon, nitrogen and phosphorus (MBC, N, P) bacteria (B), fungi (F) and their ratio B:F. = P < 0.05; ** = P <
0.01; ***=p < 0.001.
Control
MBc MBN MBp B F B:F
MBc ns -0.31* ns ns ns
MBN ns ns 0.44** ns
MBp ns -0.47** ns
B ns 0.64***
F -0.67
B:F
Irrigation
MBc ns -0.34** ns ns ns
MBN ns 0.38* 0.49** ns
MBp ns ns ns
B 0.61*** ns
F -0.58***
B:F
Litter removal
MBc ns ns ns 0.53** -0.44**
MBN ns ns 0.35* ns
MBp 0.37 ns ns
B ns ns
F -0.77***
B:F











Table 2-6. Pearson correlation analysis for each treatment among variables reported in this chapter and variables reported in other
chapters: microbial biomass carbon, nitrogen and phosphorus (MBC, N and P), and their ratios (MBC:N, C:P, and N:P),
fungi (F), bacteria (B) and their ratio (B:F), mineralization (MIN), nitrification (NIT), basal respiration (B. resp.), soil C:N


ratio, rainfall and soil water potential (SWP).* = P < 0.05; **


P < 0.01; *** = p < 0.001


MBN
ns
ns
ns
ns
ns
0.44***
ns
ns
ns
ns


MBp
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns


MBC:N
ns
0.34**
ns
ns
ns
-0.45***
ns
ns
ns
-0.31*


MBN:P
ns
ns
ns
ns


MBc:p
ns
ns
ns
-0.41***


B
0.38*
ns
ns
0.42**


F
0.48**
ns
ns
0.54***


ns ns ns ns
ns ns ns ns
ns -0.36* ns ns
ns ns ns ns
ns ns ns ns
ns ns ns ns


B:F
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns


Irrigation
NH4+ -0.31 ns ns -0.55*** 0.49*** ns ns ns ns
N03- ns ns 0.29* 0.31* -0.39** ns ns ns ns
PO43- ns ns ns ns ns ns ns ns ns
MIN ns ns ns ns ns ns ns 0.47** ns
NIT ns ns ns ns ns ns ns 0.48** -0.51***
Phosp. ns 0.32** 0.27* -0.35** ns -0.29* ns ns ns
B. resp. -0.44** ns ns ns ns ns ns ns ns
Soil C:N ns ns ns ns ns ns ns ns ns
Rain ns ns 0.37** ns -0.32* ns ns ns ns
SWP -0.32* 0.34** 0.39** -0.41** ns -0.47*** ns ns ns


Control


NH4+
N03-
P043-
MIN
NIT
Phosp.
B. resp.
Soil C:N
Rain
SWP


MBc
ns
ns
ns
-0.34**
ns
-0.26*
-0.37**
ns
ns
-0.44***









Table 2-6. Continued
Litter Removal
MBc MBN MBP MBc:N MBN:P MBc:p B F B:F
NH4+ 0.46*** ns ns -0.41** ns ns ns ns ns
NO3- ns ns ns 0.33* ns ns 0.45** ns ns
PO43- 0.45** ns ns 0.33* ns 0.30* ns ns ns
MIN ns ns ns ns ns ns 0.44** 0.52** ns
NIT ns ns ns ns ns ns ns ns ns
Phosp. ns 0.58*** 0.35** -0.37** ns ns ns 0.34* ns
B. resp. -0.34* -0.30* 0.30* ns -0.32* -0.34 ns ns ns
Soil C:N ns -0.40** ns ns -0.50*** ns ns ns ns
Rain ns ns ns ns -0.33* ns 0.33* ns ns
SWP -0.54*** ns ns -0.40** ns ns ns ns ns









Table 2-7. Comparative seasonal and/or annual mean for extractable microbial biomass C, N and P (mgC/kg-soil) across differing
tropical systems, site, depth and soil type (modified w/ permission from Rangel-Vasconcelos 2002). Superscripts depict
notes on footnote (for more details refer to Appendix A).
Study Depth Soil Period/ MBc MBN MBp Treatment Authors
Site (cm) Description Season mgC/kg-1 soil mgC/kg-1 soil mgC/kg-1 soil
India 0-10 Ultisol Rainy 487 51 20 Singh et al


(Vindhyan
Hill Tract)
Manaus
(AM)
Paragominas
(PA)


Manaus
(AM)
Paragominas
tQ (PA)
Paragominas
(PA)

SW Costa
Rica

NE Puerto
Rico


0-5

0-10


Oxisol

Yellow
Latosol


0-10 Latosol


0-10

0-2.5
15-20


Latosol

Latosol


0-10 Oxisol


0-25 Mixed
Isothermic
Tropohum
ult


Winter
Summer
Annual
(avg)
Annual
(avg)

Annual
(avg)
Annual
(avg)
Wet


662
744
1287


476


659

695


700- 1500 70-80
280-460 50-60


2000
1000


Wet
Dry

Wet
Dry
Wet
Dry


250
325


920
300
275
120


Fertilization
(N, P, N+P)


Nutrient
Gradient

Control


(1989)

Luizdo et
al (1992)
Geraldes
et al
(1995)
Feigl et al
(1995)


Davidson
et al
(2004)
Cleveland
et al
(2004)
Li et al
(2004)









Table 2-7. Continued
Study Depth Soil Period MBc MBN MBp Treatment Authors
Site (cm) Description mgC/kg-1 soil mgC/kg-'soil mgC/kg-'soil
MANFLORA 0-10 Yellow Wet 348.40 54 Pre-Treat Rangel-


Castanhal-PA


NE Puerto
Rico


BCI (Panama)



NE Puerto
Rico


MANFLORA
Castanhal-PA


Latossol/
Oxisol


0-10 Oxisol-
Ultisol


0-15 Alfisol



0-10 Mixed
Isothermic
Tropohumult


Yellow
Latossol/
Oxissol


Dry
Dry
Annual
range
Annual
range
Wet
Dry
Wet
Dry


686.40
395.20
404.11
1080-
1710
1050-
1550


200
150
300
100


1000
< 10
1000
50


Control
lyIRR
Control


lyLR


Control

5yIRR

Control


7y LR


Wet
Dry
Wet
Dry
Wet
Dry


221.14
485.06
247.22
587.03
158.63
468.97


21.26
15.83
23.09
17.87
17.74
21.01


: estimated from a figure plot. y: 7 years of litter removal. ly: 7 years of litter removal. 5y


0.59 Control
0.53
0.61 4yIRR
0.55
0.50 4yLR
0.42
5 years of dry-season irrigation


Vasconcelos
et al (2004)

Ruan et al
(2004)


Yavitt et al
(2004)


Li et al
(2005)



This study

























Figure 2-1. Effects of rainfall patterns on control (*) and long-term dry-season irrigation (o) plots
in seasonally dry tropical forest. a) Daily rainfall at the study site, b) Soil water
potential (SWP), c) Microbial biomass C, d) Microbial biomass N, and e) Microbial
biomass P. In b-e, values are means (+ se) for n= 4 plots. White and black horizontal
bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet seasons (May to Sept
20th 2004, and Jan 20t to Aug 5th 2005), respectively. Vertical dashed lines indicate
the dry season irrigation period 23rd September 2004 to 26 January 2005). ANOVA
and treatment contrasts with fixed effects by each collection date (*P < 0.05; ** P =
0.0001; *** P < 0.0001).
















140 I
a I
120 -
E 100 -
E
= 80 -
S60-
40
20


0.00 b
-0.02
-0.04
- 0.06 -
o -0.08 -

-0.10 -
-0.12 -
600 -




E |
S1300 -

S200 -

E 0I
.o 100- d I

0 -

z
60 -
z 40 -

o 20 -
E
S1




0_
1.50-
1.0 e
1E .0 I



I 0.0- I I
May Jun Jul Aug Sep Ot Nov De Jan eb Mar Ap May Jun


May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun


Time (months)














Sa 120
6e+5 100
"- -80
0) E
4e+5 60
40 4000
S3000 20



M 2 0 0 0 ....... : :. 0
0 -0
b 120
4000 -
100
S3000- 80 E
60 U
4--
2000 a
I 40
20
1000 -:.. :::: .
U- 0...
0 *

500 C 120
100
5 400 -
80

.C 300 60
200 ..- 40

1 0 0 .. ../...

0
Dec Jan Feb Mar Apr May Jun Jul
Time (months)
Figure 2-2. Effects of rainfall patterns on control (.) and long-term dry-season irrigation (o) plots
in seasonally dry tropical forest. Left y-axis: a) Bacterial colony forming units, b)
Fungal colony forming units, and c) Bacteria to fungal colony forming units ratio.
Right y-axis: daily rainfall at the study site during the harvest period. (n = 4). In a-c,
values are means (+ se) for n= 4 plots. White and black horizontal bars represent dry
(Sept 21st 2004 to Jan 19th 2005), and wet season (Jan 20th to Aug 5th 2005),
respectively. Vertical dashed line represents dry-season irrigation period (Sept 23rd
2004 to Jan 26th 2005). ANOVA contrasts with fixed effects by each harvest date (*p
< 0.05).















0a T
o I
4- 30 I

'E
.N 20 -



I 10 I1
4--
SIn I
0

b
S50 -
if)
5 40 -
w 0
o -
0
C 20-

< 10


Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun
Time (months)
Figure 2-3. Effects of rainfall patterns on control (solid bars) and long-term irrigation (grey bars)
plots in seasonally dry tropical forest. a) AMF root colonization and b) AMF spores.
In a-b values are means (+ se) for n= 4 plots. White and black horizontal bars
represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (Jan 20th to Aug 5th
2005), respectively. Vertical dashed line represents dry-season irrigation period (Sept
23rd 2004 to Jan 26th 2005), *p < 0.05.






























Figure 2-4. Effects of rainfall patterns on control (*) and long-term litter-removal (o) plots in
seasonally dry tropical forest. a) Daily rainfall at the study site, b) Soil water potential
(SWP), c) Microbial biomass C, d) Microbial biomass N, and e) Microbial biomass P.
In b-e, values are means (+ se) for n= 4 plots. White and black horizontal bars
represent dry (Sept 21st 2004 to Jan 19th 2005), and wet seasons (May to Sept 20th
2004, and Jan 20th to Aug 5th 2005), respectively. ANOVA and treatment contrasts
with fixed effects by each collection date (*P < 0.05; ** P = 0.0001; *** P < 0.0001).
















140-
120 -
E 100 -

S80 -
S60 -
40 [


0
0.oo b

-0.02

" -0.04-
-0.06 -

-0.08 -
-0.10 -

-0.12 -

600-
C
500

S400
E 300

200 -
(/A
E 100-

0

100 d

0) 80-
z 60
S40-
z
20 -

S0-


S2.0 e
(W
1.5 -

0) 1.0 -

0.5

E 0.0



May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun


Time (months)














a 120
6e+5 100
S80 E
D 4e+5 60 =
"40
2e+5 20


0
b 120
o 4000 -
100

3000 80
60
D 60 -U
u- 2000 -
o 40
20
1000

0
c 120
500 -
100
S400 -
F 80 E
S300 60

S200 -'"40

100 -: : .. 20



axis: a) Bacterial colony forming units, b) Fungal colony forming units, and c)
Dec Jan Feb Mar Apr May Jun Jul
Time (months)
Figure 2-5. Effects of rainfall patterns on control (.) and long-term litter removal (o) plots in
seasonally dry tropical forest. In a-c, values are means (+ se) for n= 4 plots. Left y-
axis: a) Bacterial colony forming units, b) Fungal colony forming units, and c)
Bacteria to fungal colony forming units ratio. Right y-axis: daily rainfall at the study
site during the harvest period. (n = 4). White and black horizontal bars represent dry
(Sept 21st 2004 to Jan 19th 2005), and wet season (Jan 20th to Aug 5th 2005),
respectively. ANOVA contrasts with fixed effects by each harvest date (*p < 0.05).
















a
0
30



.2 200
4 --

10 -*




b
50 -
0
o
V)
40-
5 40 -
0

Y) 30

Y) 20 -

< 10



Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

Time (months)
Figure 2-6. Effects of rainfall patterns on control (solid bars) and long-term litter removal (grey
bars) plots in seasonally dry tropical forest. a) AMF root colonization and b) AMF
spores. In a-b values are means (+ se) for n= 4 plots. White and black horizontal bars
represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (Jan 20th to Aug 5th
2005), respectively (*p < 0.05).









CHAPTER 3
SEASONAL AND EXPERIMENTAL EFFECTS ON MICROBIAL PROCESSES IN
SEASONAL TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON

Introduction

Soil microorganisms exert control over nutrient availability through the processes of

decomposition, immobilization, and mineralization (Singh et al., 1989; Roy and Singh 1995;

Chander et al., 1998; Mcgroddy et al., 2004). The temporal variability of microbial processes in

seasonally dry tropical forests is highly influenced by water availability, litterfall quality and

quantity (Vitousek & Sanford 1986; Wardle 2002; Roy and Singh 1995). Lucas et al. (1993)

suggested that soil moisture controls a delicate balance between the processes of immobilization

and mineralization after observing that rewetting of seasonally dry Amazonian soils resulted in

induced net immobilization, whereas dry periods allowed mineral-N to accumulate (Lucas et al.,

1993). Alternatively, wet and dry cycles may stimulate microbial activity (Cornejo et al., 1994) ,

promote microbial turnover, desiccation, and osmotic stress, and dramatically affect detrital food

chains leading to pulses of nutrient mineralization (Lodge 1994, Wardle 2002). Others have

shown that CO2 evolution increases after rewetting compared to soil kept constantly moist

(Orchard and Cook 1983; Vasconcelos et al. 2004; Davidson et al. 2000; Kiese and Butterbach-

Bahl 2002; Schwendenmann et al., 2003).

Changes in rainfall and litterfall seasonality are known to influence the physiological and

structural functioning of the soil microbial community (Yavitt 2004, William and Rice 2007),

but their relative contribution to regulating shifts in microbial community composition, microbial

processes, and soil nutrient availability, is poorly understood.

This chapter examines water and substrate constraints on N-mineralization and

nitrification, basal respiration, and acid-phosphatase activity, and their temporal variability in a

17-year-old secondary forest in the Eastern Amazon during dry and wet seasons within two









ongoing manipulative experiments designed to alter resource availability -- dry-season irrigation

and litterfall removal, described in Chapter 1.

Together, net mineralization and net nitrification assays provide insight into soil fertility

and ecosystem function. Nitrogen mineralization is often used as an index of nitrogen available

to plants in terrestrial ecosystems (Robertson et al., 1999). Microbial mineralization of soil

organic matter (SOM) can provide a significant amount of the annual nutrient requirement of

plants (Smith et al 1994) but there are other potential N-available sources/pathways such as the

direct uptake of amino acid DON without microbial mineralization of organic N to ammonium

(NH 4) in tropical soils (Schimell and Bennett 2004). Direct uptake of organic N by plants and

mycorrhizae has been demonstrated across differing ecosystems and proven as an important

conduit to plant nutrition (Kaye and Hart 1997; Schimell and Bennett 2004).

Net nitrification refers to the conversion of ammonium to nitrate by nitrifiers, bacteria that

oxidize ammonium to nitrite and then, nitrate (Robertson et al., 1999). Nitrification tends to

dominate in systems with relatively high N availability, with lower plant and heterotroph

competition for NH+4, allowing nitrifiers to flourish, slowly shifting the N economy of the system

to NO-3 dominated (Schimell and Bennett 2004).

Basal respiration is defined as the respiration without the addition of organic substrate to

soil under laboratory conditions. Basal respiration is frequently used as an index of microbial

activities in soils and often correlates positively with soil organic matter and water availability

(Alef 1995; Forster 1995; Chapin III et al., 2002).

Phosphatase activity improves the availability of soil phosphorus in nutrient poor soils

such as tropical oxisols. Phosphatases catalyse the hydrolysis of organic phosphomonoester to

inorganic phosphorus which can be taken up by plants. Acid phosphatase is predominant in acid









soils (4-6.5 as pH optimum) and has been detected in animal, plant and microbial cells (Eivazi

and Tabatabai 1977).

This chapter also integrates results from Chapter 2 and 4 to show how microbial processes

may be influenced by microbial biomass structure and composition (Chapter 2), and ultimately

affect soil solution nutrients (Chapter 5). Most measurements were analyzed in response to 1)

seasonal changes in moisture regime (wet vs. dry season); 2) reduced moisture stress during the

dry season (IRR) and, 3) reduced substrate availability (LR).

Materials and Methods

Study site and experimental design are described in Chapter 1.

Net N-Mineralization and Nitrification

Net N mineralization was estimated from changes in NH4 and NO3- concentrations during

7-day aerobic incubations at 270C of 20 g subsamples (Hart et al., 1994). All extracts were stored

frozen until analyzed. Ammonium concentrations were determined colorimetrically by the

salicylate/nitroprusside method (Mulvaney 1996). Nitrate concentrations were determined using

a simple spectrophotometric procedure described by Yang et al. (1998). Results from this and all

other assays are expressed on a dry weight basis. Paired soil samples were first extracted with

100 mL 1MKCl for 2 h, then filtered, decanted and reserved for analysis. Paired subsamples

were incubated for 7 days, and then extracted, filtered and decanted. Aliquots were analyzed for

ammonium and nitrate contents. Net N-mineralized was expressed as a function of final nitrate

and ammonium concentrations, minus initial ammonium and nitrate concentrations, divided by

incubation time (7 days). The net N-mineralization equation is shown in Equation 3-1.









Nmineralized = [(Nitratef + Ammoniumf) (Nitrateo + Ammoniumo)]/Tdays (3-1)
Where Nmineralized = net N mineralization rate, expressed as tg N g-1 d-
Nitratef = final nitrate concentration, expressed as pg NO3- -N/g soil
Ammoniumf = final ammonium concentration, expressed as pg NH4+ -N/g soil
Nitrateo = initial nitrate concentration, expressed as tg NO3- -N/g soil
Ammoniumo = initial ammonium concentration, expressed as pg NH4+ -N/g soil
Tdays = incubation time/days

Net nitrification potentials were measured by the soil shaken slurry method (Hart et al.,

1994). Fresh soils were sieved (<2 mm) and 15 g of moist soils were weighed into 250-mL

flasks. The flasks received 100 mL of phosphate buffer (1 mM potassium phosphate, pH 7.2) and

were continuously shaken for 24 h at the high speed (200 RPM). Ten-milliliter aliquots were

sampled at 2 and 24 h (to represent initial and final nitrate concentrations), then filtered,

decanted, and analyzed colorimetrically as described above. Net nitrification was expressed as

final nitrate minus initial nitrate concentration divided by incubation time (Id = 24 hours). The

net nitrification equation is shown in Equation 3-2.

Nnitrified = (Nitratef Nitrateo)/Tdays (3-2)
Where Nnitrified = net nitrification rate, expressed as tg NO3 N g1 d'

Acid Phosphatase Activity


The acid-phosphatase assay was used due to the acidic nature of the study site as suggested

by Alef et al. (1995). The method was based on the determination ofp-nitrophenol (PNP)

released after the incubation of soil with the artificial substrate p-nitrophenyl phosphate for lh at

37C. Results were expressed as [tgPNP/g- dry soil/h-1. Paired soil samples (ig) were placed in a

10 mL flask with 4mL of Universal Modified Buffer solution at pH 6.5 and placed in desiccators

with vacuum for 5 minutes to increase solution absorption among soil aggregates. After samples

were incubated at 37C for 10 minutes, ImL p-nitrophenyl phosphate solution (PNP substrate)

was added, mixed, and re-incubated for lh followed by the addition of 4mL CaCl2 (0.5M) and









ImL NaOH (0.5M), and transferred to centrifuge tubes. After centrifuging, supernatant was

saved and diluted to 40X (e.g, 0.25 mL sample to 9.75mL water) prior to absorbance reading on

a spectrophotometer set at 400nm. Results were corrected using control samples run in parallel,

and p-nitrophenol concentration was calculated per milliliter of sample in reference to a

calibration curve. Results are expressed on a dry weight basis.

Basal Respiration

Basal respiration is defined as the respiration without the addition of organic substrate to

soil and was determined using the traditional procedure described by Isermeyer (1952) revised in

Alef and Nannipieri (1995). Soil samples (25g) were incubated for 6 d in closed jars with 25 mL

0.05MNaOH solution. CO2 trapped in NaOH was determined by HCI titration. Results are

expressed on a dry weight basis.

Statistical Analysis

The SAS System for Windows V8 (2) was used for statistical analyses. PROC MIXED

was used for a repeated measures analysis with a heterogeneous-autoregressive error structure

[arh (1)]. This structure allowed modeling within sample correlation over time and calculation of

individual error variances for each sampling date. Linear models were fitted on the variables

mineralization rates, nitrification rates, phosphatase activity and basal respiration with the

following effects: season, date, treatment, treatment by season, treatment by date, plot and plot

by date. CONTRAST statements were used to determine the significance of each fixed effect for

each pair of treatment comparisons (i.e., control vs. irrigation and control vs. litter removal) and

least-squares means were used to compare treatments and control means for the effects of

season, treatment and treatment by season interaction on mineralization rates, nitrification rates,

phosphatase activity and basal respiration. Within each treatment (control, irrigation, litter

removal), Pearson correlation analyses were used to explore the bivariate relationships of results









reported in previous and subsequent chapters to those reported here. Specifically, I tested for

correlations between mineralization rates, nitrification rates, phosphatase activity and basal

respiration with microbial biomass C, N, and P and their ratios, fungal and bacterial densities

(Chapter 2), NH+4, NO-3 and PO-4 and (Chapter 3), as well as rainfall and soil water potential.

Results

Seasonal Effects

Basal respiration rates were lower in the dry than wet seasons in control plots, and within

treatments (Tables 3-2 and 3-4). There was no seasonal effect on any of the other measured

processes.

Irrigation Effects

Net nitrogen mineralization was unaffected by the irrigation treatment, but varied

substantially at different sampling dates (Figure 3-1), and that is reflected in the significant effect

of date on mineralization (Table 3-1). During one occasion in the wet season of 2004, N-

mineralization was higher in irrigated than control plots (Figure 3-1C), and that is reflected by

the significance of the treatment-by-date interaction (Table 3-1). Nitrification rates were

significantly lower in irrigated plots on one sample date at the onset of the dry season (Figure 3-

1), and were also significantly affected by treatment and the treatment-by-date interaction (Table

3-1).

Phosphatase activity was elevated by irrigation on three sample dates (Figure 3-1). Date,

treatment, and treatment-by-date effects were significant for phosphatase activity (Table 3-1).

Across all sampling dates, phosphatase activity, on average, was slightly but significant higher in

the irrigated plots (Table 3-2).

Basal respiration was not significantly affected by the irrigation treatment, although small

enhancements of basal respiration in the irrigation plots were apparent on three sample dates









(Figure 3-1). Season, date, treatment-by-season, and treatment-by-date effects were significant

for basal respoiration (Table 3-1), with values on average about 50% higher in the west season

than in the dry season (Table 3-2).

Litter Removal Effects

Nitrogen mineralization was consistently and significantly reduced by litter removal

(Figure 3-2); treatment, treatment-by-season, and treatment-by-date effects were significant

(Table 3-3). On average, the reduction in N-mineralization due to litter removal was about 40%

(Table 3-4). In contrast, nitrification rates were largely unresponsive to litter removal (Figure 3-

2), and only the treatment-by-date interaction effect was significant (Table 3-3).

Phosphatase activity was also consistently and significantly reduced by litter removal

(Figure 3-2); date, treatment, treatment-by-season, and treatment-by-date effects were significant

(Table 3-3). On average, the reduction in phophatase activity due to litter removal was -30%

(Table 3-4). Basal respiration also tended to be significantly lower in litter removal than control

plots, but seasonal and date effects on treatment were dominant (Table 3-4) and the effect was

inconsistent over time (Figure 3-2). All main and interaction effects were significant for basal

respiration (Table 3-3).

Correlation Analyses

N-mineralization was positively correlated with phosphatase activity, basal respiration

(Table 3-5), fungal and bacterial densities (Table 3-6) across treatments; and negatively

correlated with N-nitrification across treatments (Table 3-5). N-mineralization was positively

correlated with NH4+ availability in control plots only, and negatively correlated with microbial

biomass C and C:P (Table 3-6). In irrigated plots, N-mineralization was also negatively

correlated with soil water potential. Under litter removal, N-mineralization was also positively

correlated with phosphorus availability (Table 3-6).









Nitrification rates were negatively correlated with bacterial to fungal ratio in irrigated

plots, and with NO3- in control plots. Under irrigation nitrification rates were positively

correlated with fungal densities; and with rainfall in control plot (Table 3-6).

Phosphatase activity was negatively correlated with microbial biomass C:N and positively

correlated with NH4+ availability and microbial biomass N and P across treatments (Table 3-6).

In control plots, phosphatase activity was also negatively correlated with microbial biomass C

and phosphorus availability (Table 3-6). Under irrigation, phosphatase activity was also negative

correlated with C:P ratios and positively correlated soil water potential; and in litter removal

plots also positively correlated with fungal densities (Table 3-6).

Basal respiration was negatively correlated with microbial biomass C across treatments

and with C:P in control and litter removal plots (Table 3-6). Basal respiration also varied

positively with soil water potential in control and litter removal plots, but not in irrigated plots

(Table 3-6). In litter removal plots, basal respiration was also negatively correlated with

microbial biomass N, N:P, bacterial to fungal ratios and P-availability, and positively correlated

with microbial biomass P and rainfall (Table 3-6).

Discussion

Seasonal Effects

Rainfall seasonality imposed a consistent decrease in microbial basal respiration in the dry

season across all treatments, even in irrigated plots. Similar results was reported for this site by

Rangel-Vasconcelos (2004), with higher microbial basal respiration during the 2001 wet season

compared to the previous dry season, as observed in other tropical forests (Luizao et al., 1992;

Cleveland et al., 2004).

Unlike measurements of soil CO2 efflux, which does not discriminate between autotrophic

(e.g., roots) and heterotrophic respiration (e.g., microbial), decreases in basal respiration









indicated lower microbial activity in that period. Soil CO2 efflux previously measured at this site

was also significantly lower in the dry than wet season in control and litter removal plots, but

significantly enhanced in irrigated plots (Vasconcelos et al., 2004; Table 3-7). Although CO2

efflux may not be compared to microbial basal respiration rates measured under laboratory

incubations, these results are suggestive that the increase in CO2 efflux in irrigated plots in the

dry season was attributed to increased root respiration and/or increased activity of microbes in

decomposing aboveground litter rather than soil microbes.

Irrigation Effects

Dry-season irrigation had a limited and inconsistent impact on microbial processes.

Although decomposition rates (k) were 2.4 times higher in irrigated than control plots during the

sampling period (Vasconcelos 2006), irrigation had no significant effects on N-mineralization,

nitrification and basal respiration, but slightly increased phosphatase activity.

The weak response to water availability can be observed by the lack of seasonal

differences in phosphastase activity, although it was positively correlated with soil water

potential. In contrast, phosphatase activity was higher in the dry than wet season in a wet tropical

forest in Costa Rica (Cleveland et al., 2004), and in a mature tropical forest on the BCI following

"closed" laboratory incubations that simulated the dry-season (Yavitt et al., 2004). However,

dry-season irrigation had no effects on phosphatase activity on the BCI study (Table 3-7). Lower

phosphatase activity in the wet season was attributed to "aging" of litter under prolonged wet

conditions; microorganisms decomposing fresher litter, immobilized more P and synthesized

more phosphatase enzymes.

P-availability may increase as a result of phosphatase activity (Malcolm et al., 1983), and

this potential was observed as the positive correlation between phosphatase activity and biomass

P, and consequent decreases in the microbial biomass C:P ratio, although phosphatase activity









had no direct effects on P-availability (Chapter IV). The slight increase in phosphatase activity in

irrigated plots could also be due to higher mycorrhizal infections during the irrigation period

(Chapter 2). Cornejo et al. (2007) showed that phosphatase activity was enhanced in mycorrhizal

soils where N was supplied as NO3- although P uptake by arburscular mycorrhizae increased by

25% irrespective of the N source supplied.

The consistent correlation between phosphatase activity and NH4+ availability and biomass

N may be related to the role of N in enzyme production. When P is in short supply, microbial N

investments for enzyme production to acquire more N may not be as rewarding as using N to

acquire more P (Vitousek et al., 2002).

Litter Removal Effects

The decrease in N-mineralization in response to litter removal occurred without

corresponding decreases in soil organic matter (Chapter 4), and may reflect a direct effect of

decreased litter substrate for microbial activity. Hence decreases in bacterial and fungal

densities, and microbial biomass C in litter removal plots (Chapter 2), may have contributed to

decreases in N-mineralization.

Changes in microbial immobilization and mineralization rates can also be related to

changes in microbial C:nutrient ratios in relation to their substrate's ratios (Hodge et al., 2000).

In this study, the consistent decrease in N-mineralization rates was not equivalent to decreases in

microbial biomass N (Chapter 2), or nitrogen availability (Chapter IV), but was positively related

to P-availability, phosphatase activity and basal respiration, and thus, increases in microbial

activity. For discussion of the relationship among N-mineralization, biomass-N, and litterfall N,

refer to Chapter 2 (pp. 38-38).

Litter removal had no effects on N-oxide emissions between 2000 and 2002, but

nitrification rates were marginally lower than in control plots in that period (Vasconcelos et al.,









2004). Prolonged litter-removal has had no further effects on nitrification rates based on my

laboratory incubation results. As expected, net N-nitrification covaried with N-mineralization

across treatments, but no other variable affected nitrification rates even though nitrate

availability was higher in litter removal than in control plots (Chapter 4).

Variability in microbial activity reflected by basal respiration may explain variations in

phosphatase activity, which then causes changes in phosphorus availability and microbial

biomass C, N, and P (and their ratios). However the causal relationships among these variables

remain unclear.

Although phosphatase was lower in litter removal plots, the pattern of variation was very

similar to control and irrigation plots, but not related to soil water potential or rainfall events.

Phosphatase activity was related positively with fungal densities, and perhaps influenced by the

substantial increase in mycorrhizal infections in litter removal than in control plots. The

significant decrease in fungal densities in litter removal plots (Chapter 2) may account to the

decrease in phosphatase activity, but also to the decrease in N-mineralization since these

processes appeared to be co-varying altogether.

The positive effect of rainfall and soil water potential on basal respiration in litter removal

plots contradicted the lack of response to continuous water availability in irrigated plots, but it is

in agreement with the seasonal effect that resulted in lower rates in the dry season. If water was a

major factor deriving this pattern, a corresponding increase in basal respiration should be seen in

irrigation plots, unless irrigation was insufficient. Nonetheless, this decrease was not

accompanied by decreases in biomass C, root biomass, and microbial densities as previously

discussed in Chapter 2 (pp. 35-36). Decreases in soil respiration following litter removal in a

mature tropical forest on the BCI were attributed to a substantial decline in the soil microbial









biomass, and to a slight decrease in fine root biomass. Higher soil respiration in the wet season

was attributed to a concomitant increase of the same variables (Sayer 2004, Table 3-7).

Likewise, there was a 2-3 fold decrease in soil CO2 efflux in a secondary forest in Puerto Rico

due to root and litter removal, followed by root or litter removal alone. This decrease was

followed by substantial decreases in microbial biomass following the same pattern of CO2 efflux

across treatments (Li et al., 2004, Table 3-7).

This chapter illustrates the role of litter as an important conduit for nutrient transfer to

microorganisms and microbial processes that can ultimately affect the recovery and productivity

of secondary forests in Eastern Amazon. My results also reveal inter-relationships among

microbial processes, microbial structure/composition and nutrient availability.









Table 3-1. F-statistics and associated significance levels (p-value) for the effects of treatment
(Irrigation vs. Control), season (wet vs. dry), date, and the interactions between
treatment by season and treatment by date on microbial processes mineralizationn,
nitrification, acid-phosphatase activity and basal respiration). = P < 0.05; ** = P <
0.01; *** = p < 0.001.
Irrigation vs. Control
Variable Season Date Treat Treat x Season Treat x Date
Mineralization 0.27 16.25*** 0 0.17 8.4***
Nitrification 0.16 28.87*** 1.35 1.07 15.59***
Phosphatase 0.35 10.32*** 6.03** 2.22 21.25***
Basal Resp 335.98*** 45.10*** 2.61 112.48*** 45.10***

Table 3-2. Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the irrigation experiment. (Lower and upper bounds of the
95% confidence interval are provided in parenthesis). Lower case letters indicate
differences at P < 0.05 between treatments both annually and within each season.
Upper case letters indicate significant seasonal differences at P < 0.05 within each
treatment.


Irrigation

Control

Dry Season
Irrigation

Control
Wet Season
Irrigation

Control


Phosphatase Basal Respiration
(igPNP/g-'lsoil/h- ) (gCO2/g-1soil)
Treatment LSM
2425a 156
(2318-2532) (151-160)
2236 b 161
(2129-2343) (156-166)
Treatment by Season LSM


2421
(2243-2599)
2180
(2002-2358)

2427
(2302-2551)
2261
(2137-2386)


128A
(123-134)
124A
(118-129)

194B
(185-203)
189B
(181-197)









Table 3-3. F-statistics and associate significant levels (p-value) for the effects of treatment
(Litter Removal vs. Control), season (wet vs. dry), date, and the interactions between
treatment by season or treatment by date on microbial processes mineralizationn and
nitrification rates, acid-phosphatase activity and basal respiration). = P < 0.05; ** =
P < 0.01; *** = p < 0.001
Litter Removal vs. Control
Variable Season Date Treat Treat x Season Treat x Date
Mineralization 0.81 16.04 68.04*** 23.84** 11.35***
Nitrification 0.05 21.57 0.22 0.54 11.23***
Phosphatase 0.30 15.02*** 63.7*** 22.11** 171.13***
Basal Resp 305.20*** 147.25*** 17.08** 107.58*** 124.10***

Table 3-4. Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the litter removal experiment. (Lower and upper bounds of
the 95% confidence interval are provided in parenthesis). Lower case letters indicate
differences at P < 0.05 between treatments both annually and within each season.
Upper case letters indicate significant seasonal differences at P < 0.05 within each
treatment.


Litter Removal

Control

Dry Season
Litter Removal

Control
Wet Season
Litter Removal

Control


Mineralization
([tgN/g-'soil/d-

4.92a
(4.37, 5.47)
8.21b
(7.65, 8.76)


5.09a
(4.38, 5.80)
8.42b
(7.68, 9.15)

4.79a
(4.01, 5.56)
8.058b
(7.28, 8.81)


n Phosphatase Ba
') (igPNP/g-lsoil/h-) (j
Treatment LSM
1614a
(1507, 1721)
2236b
(2129, 2343)
Treatment by Season LSM


1632a
(1454, 1810)
2180b
(2056, 2304)

1606
(1482, 1730)
2261b
(2137,2386)


isal Respiration
tgCO2/g-1soil)


142a
(138, 147)
156b
(151, 161)


116A
(111, 121)
124A
(118, 129)

169aB
(162, 176)
189bB
(181, 197)









Table 3-5. Pearson correlation analysis for each treatment, Control, Irrigation or Litter removal between variables reported in this
chapter: N-mineralization rates (Min), nitrification rates (Nit.), phosphatase activity (Phosp.) and basal respiration (Resp.).
= P < 0.05; ** =P < 0.01; *** = p < 0.001.
Control
Min. Nit. Phosp. B. resp
Mineralization -0.26* 0.56* ns
Nitrification ns ns
Phosphatase ns
B. respiration

Irrigation
Mineralization -0.26* 0.39** 0.49***
Nitrification ns ns
Phosphatase ns
B. respiration

Litter removal
Mineralization -0.30** 0.42** 0.35**
Nitrification ns ns
Phosphatase ns
B. respiration











Table 3-6. Pearson correlation analysis for each treatment (Control, Irrigation or Litter removal), between variables reported in this
chapter and variables reported in other chapters: soil C:N ratio, microbial biomass carbon, nitrogen and phosphorus (MBC,
N and P), and their ratios (MBC:N, C:P, and N:P), fungi (F), bacteria (B) and their ratio (B:F), ammonium (NH4+), nitrate
(N03-) and phosphorus (P043-), rainfall and soil water potential (SWP).* = P < 0.05; ** = P < 0.01; *** = p < 0.001.
Control Irrigation Litter Removal
Min Nit Phosp. Resp. Min Nit Phosp. Resp. Min Nit Phosp. Resp.


MBC -0.34** ns -0.26* -0.37** ns ns ns
MBN ns ns 0.44*** ns ns ns 0.32**
MBP ns ns 0.35** ns ns ns 0.27*
MBC:N ns ns -0.45*** ns ns ns -0.35**
MBN:P ns ns ns ns ns ns ns
MBC:P -0.41*** ns ns -0.36* ns ns -0.29*
B 0.42* ns ns ns 0.57*** ns ns
F 0.54*** ns ns ns 0.47** 0.48** ns
B:F ns ns ns ns ns -0.51*** ns
NHF4 0.47*** ns 0.36** ns ns ns 0.47***
NO3- ns -0.29* ns ns ns ns ns
PO43- ns ns -0.32* ns ns ns ns
Soil C:N ns ns ns ns ns ns ns
SWP ns ns ns 0.33* -0.31* ns 0.29*
Rainfall ns 0.25* ns ns ns ns ns


-0.44**
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns


ns
ns
ns
ns
ns
ns
0.44**
0.52**
ns
ns
ns
0.40**
ns
ns
ns


ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns


ns
0.58***
0.35**
-0.37**
ns
ns
ns
0.34*
ns
0.35**
ns
ns
ns
ns
ns


-0.34*
-0.30*
0.30*
ns
-0.32*
-0.34*
ns
ns
ns
-0.42**
ns
-0.34*
ns
0.39**
0.44**










Table 3-7. The effects of seasonality (Dry-D vs. Wet-W season), litter removal (LR), and irrigation (IRR) on N-mineralization (Net
Min.), nitrification (Net Nit.), phosphatase activity (Phosp.), basal respiration (Basal Resp.), substrate induced respiration
(SIR), and soil CO2 efflux across studies.


Seasonal Effect


Treatment Effect


Net Min.



Net Nit.



Phosp.



Basal Resp.


SIR Resp.
CO2 efflux


Study Location
BCI
BCI
Eastern Amazonia
BCI
BCI
Eastern Amazonia
BCI
Eastern Amazonia
Eastern Amazonia
Eastern Amazonia
Eastern Amazonia
Konza Prairie
Luquilo (Puerto Rico)
Luquilo (Puerto Rico)
BCI
Eastern Amazonia


Control Plots
D N.D.
ns
D>W
N.D.
ns
D>W
ns
ns
D D D D D D D D

Irrigation
D N.D.
ns
D>W
N.D.
ns
D>W
ns


D D D N.D.
N.D.
N.D.
D>W


Litter Removal
N.D.
N.D.
ns
N.D.
N.D.
ns
N.D.


ns
D N.D.
N.D.
D D D

D

ns
ns
LR ns
LR ns
ns
IRR>Control
LR LR IRR IRR LR LR LR IRR>Control
LR

Source
Yavitt et al. (2004)
Sayer (2004)
This study
Yavitt et al. (2004)
Sayer (2004)
This study
Yavitt et al. (2004)
This study
This study
This study
Rangel et al. (2004)
Williams and Rice (2004)
Li et al. (2004)
Li et al. (2005)
Sayer (2004)
Vasconcelos et al. (2004)
Vasconcelos et al. (2004)


ns = non-significant. N.D. = not determined































Figure 3-1. Effects of rainfall patterns and dry-season irrigation on microbial-mediate processes:
(A) daily rainfall at the study site, (B) soil water potential (SWP), (C) N-
mineralization, (D) N-nitrification and (E) Acid-phosphatase, and (F) Basal
respiration. In B-F, solid and open circles represent means (+ se) for control and
irrigation treatments, respectively (N = 4 for soil water potential, N=4 to microbial
processes). Vertical dashed lines indicate the dry season irrigation period (Sept 23rd
2004 to Jan 26th 2005). White and black horizontal bars represent dry (Sept 21st 2004
th th 'h th
to Jan 19 2005), and wet season (May to Sept 20 2004, or Jan 20t to Aug 5
2005), respectively. ANOVA and treatment contrasts with fixed effects by each
collection date (*P < 0.05; **P > 0.01; *** P<0.001).
















140-
120- A
T 100-
80
i 60-
S40
20


0.00 B
-0.02 -
-0.04
0 -0.06
-0.08
-0.10-
-0.12 -

14-

12 -

8-
6-
z 4-
2


0.4 D

U U 0.2

Z 0.0-
I IL

-0.2 -


3500- E

O 3000
U 2500

0 z 2000
1500-

1000-


200- F
0 150

*.. 1 0 0 -

aoo 50 I I
C-)
I 0I
2| o- F



May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

Time (months)





























Figure 3-2. Effects of rainfall patterns and litter removal on microbial-mediate processes: (A)
daily rainfall at the study site, (B) soil water potential (SWP), (C) N-mineralization,
(D) N-nitrification and (E) Acid-phosphatase, and (F) Basal respiration. In B-F, solid
and open circles represent means (+ se) for control and litter removal treatments,
respectively (N= 4 for soil water potential, N= 4 to microbial processes). White and
black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season
(May to Sept 20th 2004, or Jan 20th to Aug 5th 2005), respectively. ANOVA and
treatment contrasts with fixed effects by each collection date (*P < 0.05; **P > 0.01;
*** P<0.001).

























-1 11 iIlul


0 -0.04 -
o_ -0.06 -
S-0.08 -
-0.10 -
-0.12 -

14- C *
12-
10
10-...
S8-
E ; 6-
Z 4-
2-

D
0.4

U S 0.2

0.0-

-0.2 -


S3500- E
. 3000-



l.. 1500 \ j
S2000-

5 200 -
0-





100 -
F



S50 -

50-
-a-




May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

Time (months)









CHAPTER 4
IRRIGATION AND LITTER REMOVAL EFFECTS ON SOIL NUTRIENT AVAILABILITY
IN A SEASONAL TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON

Introduction

Temporal variability of soil nutrient concentrations in tropical forests is poorly understood,

although it has been directly linked to microbial activity and ecosystem productivity (Yavitt and

Wright 1996; Cleveland et al., 2002; Ruan et al., 2004). Pronounced dry seasons produce

reductions in organic matter decomposition, reduced plant uptake of soil nutrients, and increased

soil nutrient pools (Luizao and Schubart 1987; Singh et al., 1989; Yavitt et al., 1993; Yavitt and

Wright 1996). Nutrient and moisture availability are often associated with litterfall dynamics

(Vitousek 1984), but the direct effects of litter and water on soil nutrient availability are not

consistent across studies. Pulses of nutrient mineralization and immobilization fluctuations in

microbial populations have a direct impact on nutrients in solution. Lucas et al. (1993) suggested

that soil moisture controls a delicate balance between the processes of immobilization and

mineralization after observing that rewetting of seasonally dry Amazonian soils resulted in net

immobilization, whereas dry periods allowed mineral-N to accumulate. Maximum root growth

may actually occur in the transition from wet to dry and from dry to wet seasons as a response to

water and/or to nutrient pulses (Cavelier et al., 1999). In tropical dry forests, drying and

rewetting causes crashes in microbial populations and induces pulses of nutrient release from

epiphytes and dead microbial biomass (Lodge et al. 1994). The mechanism apparently relates to

the wetting of dry soil that disrupts the osmotic balance of soil microorganisms, causing nutrient

release to the soil, with nutrients accumulating in plant biomass during the wet season (Lodge et

al., 1994; Yavitt and Wright 1996).

A rich and informative literature on the effects of continuous irrigation on an old-growth

tropical-moist forest (a well-drained Alfisol) resulted from a five-year study on Barro Colorado









Island (BCI), Panama. Within that study, irrigation had little effect on concentrations of

inorganic N (Yavitt et al., 1993, Yavitt and Wright 1996). Irrigation altered microbial

composition by decreasing fungal densities, and bacterial densities after five months of exposure

(Cornejo et al., 1994), enhanced decomposition rates of the forest floor, and reduced forest floor

mass throughout the year (Wieder and Wright 1995), but did not affect nutrient concentrations in

leaf-fall or nutrient return from forest trees to the forest floor (Yavitt et al., 2004). Sayer (2005)

separately examined the effects of litter manipulations on BCI and found that two years of litter

removal had no effects on nutrient concentrations in the mineral soil, except to increase NO3-

availability in litter addition plots, Soil CO2 efflux decreased by 27% in litter removal plots,

accompanied by a significant decline in total microbial biomass, but increased approximately

25% in litter addition plots after accounting for changes in root biomass (Sayer 2005). Litter

removal reduced the abundance of meso-arthropods in Simaroub litter, slowed leaf-litter

decomposition, and significantly reduced the concentration of N and P in Cecropia litter (Sayer

et al., 2006). Correspondingly, litter addition accelerated the decay of wood and increased

nutrient concentrations of Cecropia litter (N, P and K), but had no effects on leaf litter

decomposition and meso-arthropod abundance (Sayer et al., 2006).

Whether the results from BCI are indicative of generalized responses of tropical forests to

altered moisture and substrate availability remains unclear. There have been two other litter-

removal studies in a tropical wet forest in the Luquillo Experimental Forest in north-eastern

Puerto Rico. The first study consisted of parallel measurements in a pine plantation and a

secondary forest between 1996 and 1997 under three treatments (root exclusion, litter exclusion,

and root-and-litter exclusion) that were initiated in 1990 (Li et al., 2004 and 2005). Their results

showed that litter removal significantly reduced soil respiration and microbial biomass in both









the pine plantation and the secondary forest, but litter removal had a greater effect on soil CO2

efflux than root exclusion in the secondary forest. The second study found that microbial

biomass was highly correlated with aboveground litter inputs from the preceding month,

suggesting that enhanced root exudates prior to senescence, could have influenced microbial

biomass abundance (Ruan et al., 2004).

The main objectives in the present study were to (1) determine the responses of NH4+,

NO3-, and PO43- availability, and soil C:N to seasonal changes and wet-up events in a tropical

secondary forest in the Eastern Amazon, and (2) to examine substrate and water constraints to

the availability of those nutrient species within two ongoing manipulative experiments designed

to alter resource availability in that forest -- dry-season irrigation and litterfall removal. The site

selected for this work contrasts with the BCI site in that it is (1) secondary v. old-growth forest;

(2) seasonally-dry v. continuously moist; and (3) underlain by a shallow, coarse, and relatively

infertile soil v. a the deep, fine-textured, relatively fertile soil present at the BCI site. These

contrasts allow us to draw inferences about both the similarities and differences in the responses

of soil nutrient availability to resource manipulations at the two sites.

Study Site and Experimental Design

Study site and experimental design are described in Chapter 1.

Materials and Methods

Ion Exchange Resins

The Dowex 50W-X8 cation exchange resins, 50-100 mesh, H form and 1.9 mmolc cm-3

(Sigma Aldrich Family, Catalog # 217492, Milwalkee, WI) was used to measure ammonium

availability (NH4+), in the mineral soil (5 cm depth) of all 12 plots.

The resin-bag technique consisting of a specified quantity of loose resin beads placed onto

a piece of porous fabric (mesh) and sealed into a convenient shape or size (Skogley and









Dobermann 1996) was deployed in the field to quantify N-H4 availability. The 25 cm2 resin bag

containing 2.2g + 0.02 of dry weight resins was hand-made of non-adsorbent fabric (07-40/25

Pecap, Sefar America Inc. Filtration Division, www.sefaramerica.com) re-used after extraction

procedures for 2 harvest cycles and then discarded.

Resin bags were recharged prior to deployment in the field by placing them in Nalgene

bottles with 100 ml 0.3M HC1, shaking for lhr on a platform shaker, and then rinsing 3 times

with deionized water. Researchers that have previously used Dowex resins (25-50 mesh)

recharged, eluted and sometimes re-used them by shaking each bag in a platform shaker for one

hour in 100 ml of 2 M NaCl in 0.1 M HCI (Giblin et al., 1994) or soaking for at least 18 h in 100

ml 1.2 M HCI (Yavitt et.al., 1996), then rinsing with de-ionized water before use. For Dowex

resins (50-100 mesh), resin extraction efficiency was greatest using 100 ml of 2 M NaCl in 0.3

M HCI (unpublished results).

During the first month of installation, 5 resin bags were installed randomly per

experimental plot, for a total of 20 units per treatment. A small slit was carefully lifted from the

top 5cm of the mineral soil using a flat, sharp and clean plastic spatula. Each bag location was

identified with a colored flag, so that bags would be replaced at the same location in the

following months. After each monthly harvest of resin bags installed in the field, resins were

transported to the laboratory facility in individual Nalgene bottles containing only de-ionized

water. Then, after excess organic matter and residuals were rinsed off the bags with de-ionized

water, 100 ml of 2 M NaCl in 0.3 M HCI was used for each bag to extract NH4+ ions. Bags were

shaken on a platform shaker for 1.5 h, removed, and an aliquot reserved for colorimetric

analysis. Ammonium concentrations were determined colorimetrically by the salicylate-









nitroprusside method (Mulvaney 1996) on duplicate samples. An average of the 2 samples was

used for statistical analysis; duplicates mean differed by < 8% ( 0.4).

Anion Exchange Membrane

We used BDH Anion exchange membranes (Gallard-Schlesinger Ind., Product # 55164 2S,

Plainview, NY, 1-888-686-3454) to measure nitrate (NO3-), and phosphorus availability (P04-) in

the mineral soil (5 cm depth) of all plots. This membrane is supplied in the chloride form, with

exchange capacity estimated at 0.2 mmol P g-1 (Turrion et al 1997).

Testing and deployment of anionic membranes for quantification of nitrate and phosphate

followed the method described by Turrion et al. (1997), modified to use 16cm2 membrane strips

during in situ incubations. They suggested that the resin bag technique have several

disadvantages over anion membranes. Resin bags may tear and wear out, loose resin beads, trap

fine roots, fungi, and soil particles that may interfere with analysis, and add diffusion problems

due to their three-dimensional spherical structure, while membranes' flatness impose no

diffusion problems and have greater surface area that improves contact with soil surface.

Before use, chloride-saturated anion-exchange membranes were converted to the

bicarbonate form set at 8.5 pH. Strips were shaken for 1 hour for each of three successive washes

in 50 ml 0.5M NaHCO3 /per strip (sodium bicarbonate), and then rinsed the strips three times in

deionized water. Strips were taken to the field in Nalgene bottles filled with deionized water.

During the first month of installation, 4 membranes were installed randomly per

experimental plot, for a total of 16 units per treatment. Each strip was labeled by placing a nylon

tread through a hole in one corner of the strip. A small slit was carefully lifted from the top 5cm

of the mineral soil using a flat, sharp and clean plastic spatula. Each membrane location was

labeled with a colored flag, so repeated collections could be carried out during the following

months. After each monthly harvest of membrane strips installed in the field, strips were









transported to the laboratory facility in individual Nalgene bottles containing only deionized

water. Then, after excess organic matter and residuals were rinsed off the bags with deionized

water, strings were cut and 32 ml of 0.3 M HCI (hydrochloric acid) was used to extract the

anions of interest (NO3- and P04-). Strips were shaken on a platform shaker for 2 h, removed, and

an aliquot reserved for colorimetric analysis. Nitrate was analyzed using a simple

spectrophotometric method to detect nitrate in water, resin or soil extracts as described by Yang

et al (1998), modified by the addition of 0.2ml 3 M KOH for each 2ml of aliquot (matrix 0.3M

HC1) to neutralize [Cl-] interference. Phosphorus determination was done using the standard

Murphy and Riley procedure (1962). Samples were analyzed in duplicates, and their average was

used for statistical analysis; on average, duplicates differed by <10% ( 0.5).

Soil C:N Ratios

Soil sampling and processing: In each plot, seven soil cores were taken using a bulk-

density corer with 6 cm diameter and 5 cm depth. These samples were composite per plot, sifted

through a 2 mm mesh, placed in double-folded, tightly closed plastic bags, transported to the lab

in a cooler no later than 4 hours after harvest, and stored at 4C until analyzed. Subsamples taken

to run total soil carbon were dried at 40C and total soil nitrogen was left air-drying prior to

analysis.

Determination of soil organic carbon (SOC): I used the Nelson-Somers method (1975)

revised in Forster (1995). This is a routine carbon analysis generally recommended for mineral

soils. Paired soil samples were digested using wet potassium dichromate, followed by titrimetric

measurement of unreacted dichromate.









Determination of Total N: I used the Keeney and Nelson's method (1982) revised in

Forster (1995). Paired soil samples were digested using sulfuric acid (the Kjedahl procedure)

followed by titrimetric analysis.

Soil Water Potential (SWP)

We used a four-channel datalogger with four gypsum resistance block sensors installed in

each plot at 10cm depth (Model 220, Spectrum Technologies, Inc) between 12 May and 17

December 2004. Measurements were taken hourly and then averaged per day and downloaded to

our database monthly. Due to technical problems with the dataloggers, between 18 of December

2004 and 30 of June 2005, SWP was recorded on a weekly basis in the morning using one

tensiometer installed per plot at 10cm depth (Model 2710 ARL, Forestry Supplier, Inc. Catalog

53, Jackson, Mississippi 39284-8397). Resistance block sensors have close correlations with

tensiometer readings of soil water potential (SWP) depending on soil type (Shock et al 2001).

For example, in an irrigation field, silt loam soil (Malheur Experiment Station, Oregan Sate

University), the resistance block closely tracked those obtained with a tensiometer (r2= 0.83). At

our study site, r2 values were 0.80, 0.81, and 0.60) for control, litter-removal, and irrigation plots,

respectively.

Statistical Analysis

The SAS System for Windows V8 (2) was used for statistical analyses. The response

variables of interest NH+4, NO-3 and PO-4 were log-transformed to meet the model

assumptions of normality. PROC MIXED was used using a repeated measures analysis with a

heterogeneous-autoregressive error structure [arh (1)]. This structure allowed modeling within

sample correlation over time and calculation of individual error variances for each sampling date.

Linear models were fitted on the variables NH+4, NO-3 and PO-4 and soil C:N with the following

effects: season, date, treatment, treatment by season, treatment by date, plot and plot by date. All









effects were considered fixed with the exception of plot and plot by date for NH 4, NO-3 and

PO-4, thereby allowing assessment of correlation between samples in the same plot on any date.

CONTRAST statements were used to determine the significance of each fixed effect for each

pair of treatment comparisons (i.e., control vs. irrigation and control vs. litter removal) and least-

squares means were used to compare treatments and control means for the effects of season,

treatment and treatment by season interaction on NH+4, NO-3 and PO-4 availability, and soil C:N

ratio. In order to correct for multiple testing the Bonferroni correction was used for each

response variable considering an experiment-wise significance level of 5%. Within each

treatment (control, irrigation, litter removal), Pearson correlation analyses were used to explore

the bivariate relationships of results reported in previous chapters to those reported here.

Specifically, I tested for correlations for NH+4, NO-3 and PO-4 and soil C:N with microbial

biomass C, N, and P and their ratios, fungal and bacterial densities (Chapter 2); mineralization

rates, nitrification rates, phosphatase activity, basal respiration (Chapter 3), as well as rainfall

and soil water potential.

Results

Dry-season irrigation reduced NH4 (Table 4-1, Figure 4-1c) and litter removal reduced

NH4 and P043- availability, while enhancing N03- availability (Table 4-5, Figure 4-2). There

were no seasonal or treatment effects on soil C:N ratios (Table 4-1). Correlations between NH4+,

N03-, and P043- availability and measurements reported in previous chapters were strongly

influenced by treatment.

Seasonal Effects

The only seasonal effect was increased phosphorus availability in the dry than wet season

in control plots (Tables 4-1, 4-2, 4-3 or 4-4). There was also intra-annual variation observed









within treatment for PO43- availability in irrigated plots (Table 4-2), and for NF4+ availability in

litter removal plots (Table 4-3).

Irrigation Effects

The dry-season extended from mid-July 2004 to early-January2005 (Figure 4-la), resulting

in lower water potential during this period (Figure 4-1b). November was the driest month of the

year, when there was only 8 mm of rainfall in the 27 d prior to our next harvest. Mean soil water

potential for the dry season was -0.046 0.001 and -0.018 0.001MPa (mean SE), for control

and irrigation plots with corresponding gravimetric soil water content of 0.67 0.07 and 0.83 +

0.10 ( SE), respectively. In the wet season of 2004, mean soil water potential was -0.016

0.001 and 0.011 0.001 MPa, for control and irrigation plots with corresponding gravimetric

soil water content of 0.81 0.05 and 0.65 0.10, respectively. In the wet season of 2005, mean

soil water potential was -0.010 + 0.005 and 0.006 0.003 MPa, for control and irrigation plots

with corresponding gravimetric soil water content of 1.38 0.04 and 1.64 0.05, respectively.

Ammonium availability declined due to irrigation (Figure 4-1c). Toward the end of the

dry season, there were two wet-up events that significantly increased the availability ofNH4+ in

control plots with little effect on irrigation plot values. The first pulse occurred after a 52 mm

rainfall event on 14 December 2004 (two days prior to our sampling). This event represented

double the amount of precipitation for the month prior to this date (25 mm from 14 November to

13 December 2004). On the 11 January 2005 there was a second pulse that followed smaller rain

events on 10 January and 11 January 2005 (13 mm and 17 mm), respectively (Figure la). These

events are responsible for the significant treatment by season interactions shown in Table 4-1.

Overall, the dampening effect of dry-season irrigation on NH4+ availability is apparent in both

the dry and the wet season (Table 4-2).









The two wet up events toward the end of the dry season also substantially increased NO3-

availability in both control and irrigation plots. NO3- availability was not significantly affected by

the irrigation treatment for any single sampling date (Figure Id). There were significant effects

of date and treatment-by-date interactions (Table 4-1), but overall there was no significant effect

of irrigation on NO3- availability during either season, although annual and seasonal mean values

trended lower for the irrigation plots (Table 4-2). Phosphate availability in both control and

irrigation plots was also increased by the two wet up events, and the increase was significantly

greater for the irrigation treatment (Figure le). There were significant main effects of date and

season, and both treatment-by date and treatment-by-season interactions on PO43- availability

(Table 4-1). Overall, there was no significant effect of irrigation on PO43- availability during

either season, but values were significantly higher in the dry season in both irrigation and control

plots (Table 4-2). There were significant effects of date and treatment by date interaction on soil

C:N ratios (Table 4-1).

Litter Removal Effects

Soil water potential varied seasonally and in response to precipitation during the dry

season, but was unaffected by litter removal (Figure 4-2a, and b). Ammonium availability was

significantly lower in litter removal than control plots in 9 out of 15 sampling dates, 8 in the wet

season (Figure 4-2c). These results are responsible for the significant main effects of date and

season, and the treatment-by-date and treatment-by-season interactions shown in Table 4-4. The

two wet-up events in the late dry season increased NH4 in the litter removal plots, but the effect

was smaller than in the control plots (Figure 4-2c). Overall NH4+ availability was significantly

reduced by litter removal, and that overall effect was driven by the significant treatment effect in

the wet season (Table 4-5).









Nitrate availability was significantly higher in litter removal than control plots in 6 out of

15 sampling dates, 5 in the wet season (Figure 4-2c). These results are responsible for the

significant main effects of treatment and date, and the treatment-by-date interactions shown in

Table 4-4. The one significantly different dry-season sampling occurred at the second wet-up

event toward the end of dry season, which increased NO3- availability in the litter removal plots

to a substantially greater extent than that exhibited in the control plots (Figure 4-2d). Overall,

NO3- availability was significantly increased by litter removal and that overall effect was driven

by the significant treatment effect in the wet season (Table 4-5).

Phosphate availability was significantly higher in litter removal than control plots in 4 out

of 15 sampling dates (Figure 4-2e). All of the main and interaction effects tested were

significant (Table 4-4). The first wet-up event toward the end of the dry season significantly

increased phosphate availability in the litter removal plots but the second wet up had the opposite

effect (Figure 4-2e). Overall P043- availability was slightly but significantly reduced by litter

removal, and the effect was consistent in both dry and wet seasons (Table 4-5). There were

significant effects of date and treatment by date interaction on soil C:N ratios (Table 4-3).

Correlation Analyses

The only significant correlations among variables measured in this study were between

NH4+ and PO43- availability in control plots, and between NO3- and PO43- availability in irrigated

plots (Table 4-5).

Control plot NH4+ availability was positively correlated with bacterial, fungal densities and

N-mineralization, and negatively correlated nitrification rates (Table 4-6). Under irrigation, NH4+

availability was positively correlated with phosphatase activity and microbial biomass N:P, and

negatively correlated with microbial biomass C and C:N (Table 4-6). Contrary to the results in

irrigation plots, NH4+ was positively correlated with microbial biomass C in litter removal plots,









but positively correlated with phosphatase activity. Additional results showed a negative

correlation between NH4+ and basal respiration, soil water potential and biomass C:N (Table 4-

6C).

NO3- availability was negatively correlated with nitrification rates in control plots only

(Table 4-6A), with microbial biomass N:P in irrigated plots (Table 4-6B), and with soil water

potential in litter removal plots (Table 4-6C). The positive correlation between NO3- and

microbial phosphors was consistent in control and irrigated plots; and between NO3- and

microbial C:N was consistent across treatments. Under irrigation, NO3- availability was also

positively correlated with rainfall.

PO43- availability was negatively correlated with phosphatase activity in control plots only

(Table 4-6A), and with soil water potential in irrigated plots (Table 4-6B). Under irrigation plots,

PO43- availability was also positively correlated with bacterial densities. Likewise in irrigated

plots, PO43- availability was negatively correlated with soil water potential in litter removal plots,

and with microbial basal respiration (Table 4-6C). Under litter removal, PO43- availability was

also positively correlated with microbial biomass C, C:N, C:P and with N-mineralization.

Soil C:N ratios appeared to be significantly affected by the effects of litter removal only.

Under litter removal, soil C:N was positively correlated with microbial biomass C:N, and

negatively correlated with soil water potential, microbial biomass N, and N:P (Table 4-6C).

Discussion

This study showed that dry season wet-ups were a principal source of temporal variability

in nutrient status resulting in higher nutrient availability in this period. The observed peaks in

NH4+, NO3- and P043- availability in the dry season as a result of wet up events is typical of dry

tropical forests with monsoonal climate, when pool sizes of these nutrients tend to increase

towards the dry season, as plants senesce, then decrease during the wet growing-season (Singh et









al., 1989, Lodge et al., 1994). In this study, the four peaks in ammonium throughout the year

accounted for half of the annual NH4 availability, thus wet up events appear to greatly

contribute to the total annual N-pool in this tropical seasonal forest, perhaps due to a positive net

mineralization response to wet-up (Table 4-6). The positive correlation between NH4

availability with fungal and bacterial densities, and N-mineralization in control plots could be a

result of their intrinsic relationship, observed by concomitant increases of these variables during

wet-up events.

The reduction in NH4 availability in irrigated plots could indicate that continuous water

availability resulted in higher nutrient immobilization (lower mineralization), but additional

results showed no effects of treatment on mineralization and nitrification rates (Chapter 3), even

though leaf decomposition was 2.4 times faster in irrigated than control plots (Vasconcelos et al.,

2006). Lodge et al. (1994) suggested that environments with dampened microclimate

fluctuations may enhance microbial N-immobilization, increase the competition for N, and even

affect primary productivity if microbial N-mineralization does not synchronize with plant N-

uptake. Nonetheless, under irrigated conditions, the association between microbial biomass

nutrient concentrations with NH4 availability suggested that continuous water availability have

increased the potential for microbial NH4 uptake, also indicated by the negative correlation

between NO3- and microbial N:P ratios (Table 4-6). The additional positive correlation between

NO3- availability with microbial biomass P and rainfall was probably driven by concomitant

responses of these variables to wet-up events. Changes in the ratio of nutrient concentrations of

the substrate in relation to that of the decomposing microbial biomass may also trigger changes

in the ratios of the latter (Hodge et al., 2000).









Reduced NH4 and P availability as a result of litter removal suggest that recycling of

nutrients in litter is an important pathway for nutrient availability as this secondary forest is

nutrient poor and a large proportion of potentially available nutrients are retained in the living

biomass and recycled in the litter (Singh et al., 1989). Some studies have suggested that prior to

leaf senescence, some tropical forests experience nutrient translocation from shoot to roots that

may result in increased root-exudates and nutrient availability (Ruan et al., 2004), which may

explain the high peak in NH4+ and NO3- availability prior to the onset of the dry-season when soil

water potential was very high (- 0.06 MPa, Figure 4-1b). Yavitt and Wright (1996) report that

fluctuations in soil nutrient availability imposed by the timing of leaf litterfall and nutrient

leaching from forest floor litter (also measured with ion-exchange resins), essentially

disappeared with depth in the mineral soil.

Soil NH4+ availability was likely reduced in litter removal plots because long-term effects

of four years of litter removal consistently reduced litterfall N (Vasconcelos et al., 2006),

confirming that litterfall is a significant source of N for tropical forest plants. The next plausible

explanation can be drawn by the positive feedback that increased water availability had on

microbial basal respiration (Chapter 3, Table 3-6C), which led to more NH4+ uptake, and

increased microbial biomass C in litter removal plots (Table 4-6C). Based on these correlations,

NH4+ availability increases as the soil dries because basal respiration also decreases, and there is

less NH4+ uptake. The same assumption can be applied to the availability of phosphorus, which

followed a similar pattern as NH4+, and covaried with it. The reduction of NH4 and P-

availability in litter removal plots could also be attributed to removal of the microflora biomass

(mostly fungi and bacteria) inhabiting the litter that contributes to immobilization and

mineralization of these nutrients in the interface between the mineral soil and the litter layer, as









suggested by Sayer (2005) in a similar study on the BCI. In fact, this study has shown that

bacterial and fungal densities significantly decreased in the mineral soil in litter removal

compared to control plots (Chapter 2).

There were however, no significant reductions on litter P concentrations although

Vasconcelos et al. (2006) characterized this site as having low litterfall P concentration in

accordance to values proposed by Vitousek (1986). They also suggested that this site may have

sufficient P-supply through soil organic matter mineralization based on recent studies that

showed substantial amounts of labile P-fractions for secondary forests in the Amazon (Frizano et

al., 2003, Markewitz et al., 2004). Alternatively, Vasconcelos et al. (2006) suggested that litter P

concentrations were not affected by litter removal likely due to enhanced P-acquisition through

mycorrhizal associations and high phosphatase exudation rates as previously observed in

secondary forests with low soil P (Marschnner 1995). Results from this study showed that litter

removal in fact enhanced arbuscular mycorrhizal fungi associations (Chapter 2), but phosphatase

activity was lower than in control plots and did not respond to seasonal changes (Chapter 3).

Arbuscular mycorrhizal fungi may enhance the reabsorption of nutrients lost through root

exudation, influence biochemical reactions in soil including mineralization of organic matter and

nitrification, and improve the capacity of their host plant to use organic sources of P and N

(Hamel 2004).

The litter layer may absorb N that would otherwise be available in the mineral soil, here

evident with higher NO3- availability in litter removal than control plots. Results from a recent

study in a seasonally dry tropical forest in Mexico showed that N-dynamics in the litter layer was

influenced by rainfall seasonality and labile C, with lowest C and N immobilization rates in the

rainy season (Anaya et al., 2007). Low substrate availability, could have resulted in less









microbial immobilization of NH4 allowing higher nitrification rates and therefore, higher NO3

availability, but nitrification rates were not significantly affected by litter removal (Chapter 3).

Soil microsites may even shift from NH4+ to NO3- dominated depending on demand and supply

of these nutrients (Schimel and Bennett 2004). The positive correlation between NO3- and

biomass C:N may further indicate that NO3- is the preferred N-source for microbial organisms in

litter removal plots, whereas the contrary relationship was observed between NH4+ and microbial

biomass C:N. The positive correlation between nitrate and bacterial densities in this study may

be related to nitrifiers, bacteria that oxidize ammonium to nitrite and then, nitrate (Robertson et

al., 1999).

Higher NO3- availability for litter removal plots could also indicate that litter is

intercepting NO3- in throughfall, as previously observed in a Northwestern Amazonian forest

(Tobon et al., 1994). The fact that NO3- availability was negatively correlated with soil water

potential even though it responded to wet up events contradicts nitrate's usual higher mobility in

water, and discard its potential relation to throughfall in this study site. Another study using the

same ionic membrane showed that NO3-diffusion to the membrane was impaired as the soil dried

(Turrion et al., 1997).

The linkage between soil C:N and microbial biomass nutrients in litter removal plots, may

suggest that the organic soil carbon and nitrogen pools has become an active reservoir of

nutrients for microbial uptake. However, further studies would be necessary to explore these

findings, and perhaps investigate how long soil carbon and nitrogen pools would sustain forest

and microbial activity under continuously litter removal.

In contrast with the results obtained in the BCI experiments (Yavitt et al., 1993; Yavitt and

Wright 1996, Yavitt et al., 2004), I found that dry-season irrigation reduced the availability of









NH4+ and that litter removal reduced N-4+ and P043- availability, while enhancing NO3-

availability. The results by Sayer (2006), showed that litter removal lowered the concentration of

N and P in Cecropia litter, but this results was not followed by reductions in nutrient availability

in the mineral soil (NH4+, NO3-, and P043- ), except to an increase in NO3- in litter-addition plots.

This study also showed that other nutrient pools may temporary counteract the nutrient

deficit imposed by litter removal, such as the contribution of nutrients stored on microbial

biomass, derived from soil C:N pools or compounds from root and mycorrhizal exudates, but

further studies would be necessary to identify the magnitude of these contributions.

Thus, nutrient availability could ultimately be affected by changes in microbial

composition, structure and activity as shown by their interactive effects in this seasonal tropical

forest.

On individual collection dates in this study, differences between treated and untreated plots

varied from non-significant to upwards of four-fold, suggesting that complex interactions shape

the responsiveness of nutrient dynamics to changes in resource availability. Frequent sampling is

needed to adequately capture intra-annual variability in soil nutrient availability.









Table 4-1. F-statistics and associated significance levels (p-value) for the effects of treatment
(Irrigation vs. control), season (wet vs. dry), date, and the interactions between
treatment by season and treatment by date on soil solution nutrients (NH+4, NO-3 and
PO-4), and soil C:N. = P < 0.05; ** = P < 0.01; *** = p < 0.001.
Irrigation vs. Control
Variable Season Date Treat Treat x Season Treat x Date
NH 4 0.640 15.33*** 6.97** 2.58* 12.60***
NO-3 0.006 33.06*** 0.84 0.47 16.33***
PO-4 25.60*** 10.55*** 0.34 8.71*** 11.76***
Soil C:N 0.30 5.46*** 0.32 0.87 6.12***

Table 4-2. Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the irrigation experiment. (Lower and upper bounds of the
95% confidence interval are provided in parenthesis). Lower case letters indicate
differences at P < 0.05 between treatments both annually and within each season.
Upper case letters indicate significant seasonal differences at P < 0.05 within each
treatment.


NH4 NO3-
(igN/ml/bag) (igN/ml/strip)
Treatment LSM
S3.36a 0.30a
irrigation (2.83-3.98) (0.16-0.55)

Control 4.65b 0.45 a
(3.92-5.52) (0.25-0.82)
Treatment by Season


Dry Season
Irrigation

Control
Wet Season


3.32 a
(2.59-4.25)
5.122b
(3.98-6.59)


3.38a
Irrigation (2.85-4.00)
Cnt (4.45b
Control .5
(3.76-5.27)


0.33
(0.17-0.62)
0.43
(0.22-0.81)

0.29
(0.16-0.54)
0.47
(0.25-0.87)


P04
(igP/ml/strip)

0.07a
(0.06-0.08)
0.07a
(0.06-0.08)
LSM


0.10A
(0.08-0.12)
0.10A
(0.08-0.12)

0.06B
(0.06-0.08)
0.06B
(0.05-0.07)









Table 4-3. F-statistics and associate significant levels (P-value) for the effects of treatment
(Litter removal vs. control), season (wet vs. dry), date, and the interactions between
treatment by season or treatment by date on soil solution nutrients (NH+4, NO-3 and
PO-4), and soil C:N. = P < 0.05; ** = P < 0.01; *** = p < 0.001.
Litter Removal vs. control
Variable Season Date Treat Treat x Season Treat x Date
NH+4 23.42*** 9.26*** 22.44*** 21.29*** 10.85***
NO-3 2.10 34.77*** 4.06* 2.21 17.15***
PO-4 33.64*** 27.67*** 20.34*** 18.07*** 166.56***
Soil C:N 0.01 17.05*** 0.42 1.37 147.93***

Table 4-4. Least square mean values for significant season, treatment, and treatment by season
contrasts associated with the litter removal experiment. (Lower and upper bounds of
the 95% confidence interval are provided in parenthesis). Lower case letters indicate
differences at P < 0.05 between treatments both annually and within each season.
Upper case letters indicate significant seasonal differences at P < 0.05 within each
treatment.
NH4+ N03- P04-
(igN/ bag) (igN/ strip) (pgP/ strip)
Treatment LSM
Litter removal 2.59a 1.09a 0.05a
(2.18-3.07) (0.59-1.99) (0.04-0.05)
Control 4.65b 0.45b 0.07b
(3.92-5.52) (0.25-0.82) (0.06-0.08)
Treatment by Season LSM
Dry Season
3.95A 0.91 0.07aA
Litter removal 0.91 0.07
(3.07-5.07) (0.48-1.72) (0.05-0.08)
5.12 0.43 0.10bA
(3.98-6.59) (0.22-0.81) (0.08-0.12)
Wet Season
2.14aB 1.20a 0.04aB
(1.81-2.53) (0.65-2.24) (0.03-0.05)
4.45b 0.47b 0.06bB
(3.76-5.27) (0.25-0.87) (0.05-0.07)









Table 4-5. Pearson correlation analysis for each treatment, Control, Irrigation and Litter removal
among variables reported in this chapter: ammonium (NH4 ), nitrate (N03-),
phosphorus availability (P043-), and soil C:N. = P < 0.05; ** = P < 0.01; *** = p <
0.001.
Control
NH4+ N03- P043- Soil C:N
NH4+ ns 0.34** ns
NO3- ns ns
PO43- ns
Soil C:N
Irrigation
NH4+ N03- PO43- Soil C:N
NI-4 ns ns ns
N03- 0.57*** ns
PO43- ns
Soil C:N
Litter removal
NH4+ N03- PO43- Soil C:N
NH4+ ns ns ns
N03- ns ns
PO43- ns
Soil C:N









Table 4-6. Pearson correlation analysis for each treatment among variables reported in this chapter (NH4 NO3-, P043-, and soil C:N),
and variables reported in other chapters: microbial biomass carbon, nitrogen and phosphorus (MBC, N and P), and their
ratios (MBC:N, C:P, and N:P), fungi (F), bacteria (B) and their ratio (B:F), mineralization (MIN), nitrification (NIT), basal
respiration (B. resp.), soil C:N ratio, rainfall and soil water potential (SWP).* = P < 0.05; ** = P < 0.01; *** = p < 0.001.

Control Irrigation Litter Removal


NH4


N03-


P043-


NH4


N03-


P04 3


NH4+ N03- P043-


MBC ns ns ns -0.31* ns ns 0.46*** ns 0.45**

MBN ns ns ns ns ns ns ns ns ns

MBP ns 0.29* ns ns 0.29* ns ns ns ns

MBc:N ns 0.24* ns -0.55*** 0.31* ns -0.41 0.33* 0.33*

MBN:P ns ns ns 0.49*** -0.39** ns ns ns ns

MBc:p ns ns ns ns ns ns ns ns 0.30*

B 0.38* ns ns ns ns 0.45** ns 0.45** ns

F 0.48** ns ns ns ns ns ns ns ns

B:F ns ns ns ns ns ns ns ns ns

MIN 0.47*** ns ns ns ns ns ns ns 0.40**

NIT -0.27* -0.29* ns ns ns ns ns ns ns

PNP ns ns -0.32* 0.47*** ns ns 0.35** ns ns

B. resp. ns ns ns ns ns ns -0.42** ns -0.34*

SWP ns ns ns ns ns -0.43** -0.41** -0.47*** -0.40**


ns ns ns ns 0.38** ns ns


Rain


ns ns


I


I





Litter Removal
Soil C:N


Table 4-6. Continued

Control Irrigation
Soil C:N Soil C:N
MBC ns ns

MBN ns ns

MBP ns ns

MBc:N ns ns

MBN:P ns ns

MBc:P ns ns

B ns ns

F ns ns

B:F ns ns

MIN ns ns

NIT ns ns

PNP ns ns

B. resp. ns ns

SWP ns ns

Rain ns ns


-0.40**


0.46***

-0.50***















0.27*

ns





























Figure 4-1. Effects of rainfall patterns on control (*) and long-term dry-season irrigation (o) plots
in seasonally dry tropical forest. a) Daily rainfall at the study site, b) Soil water
potential (SWP), c) NH+4, d) NO-3 and e) PO-4 availability. In b-e, values are means (
se) for n= 4 plots. White and black horizontal bars represent dry (21st September 2004
to 19 January 2005), and wet seasons (20 May to September 2004, and 20 Jan to 5
August 2005), respectively. Vertical dashed lines indicate the dry season irrigation
period 23rd September 2004 to 26 January 2005). ANOVA and treatment contrasts
with fixed effects by each collection date (*P < 0.05; ** P = 0.0001; *** P < 0.0001).














140
120
E
E 100
= 80
= 60
n 40
20

0.00
-0.02
-0.04
0. -0.06
) -0.08
-0.10
-0.12
14
12
S) 10
+ Co
T- 8
0 6

2
0
5
4
O~ 3
-z 2
<1 1
0


0 .-
- Q.


May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

Time (months)




























Figure 4-2. Effects of rainfall patterns on control (*) and long-term litter removal (o) plots in
seasonally dry tropical forest. a) Daily rainfall at the study site, b) Soil water potential
(SWP), c) NH+4, d) NO-3 and e) PO-4 availability. In b-e, values are means (+ se) for
n= 4 plots. White and black horizontal bars represent dry (21st September 2004 to 19
January 2005), and wet seasons (20 May to September 2004, and 20 Jan to 5 August
2005), respectively. ANOVA and treatment contrasts with fixed effects by each
collection date (*P < 0.05; ** P = 0.0001; *** P < 0.0001).















140 -
a
120 -
E
E 100 -
= 80 -
S60 -
CU
r 40 -

20 I ILI 1 i .I. L I. 1..11111 I .11
0.00 -
-0.02 b
S-0.04
-0.06 -
-0.08 0-- Control
-0.10 --o- LR

-0.12 -
14 -
12- c
10- *
+ 0)
SCo 8 -
z 6 .-
4 6-

2-
0
10 -
1 d
8 -
1 60 *

Sz 4-
0-

0.5 -
e
0.4 -
S0.3 -


, =L 0.1 -
0.0 -

May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

Time (months)












140 -
E 120 -A
E 100 -
80 -
4--
60 -
40 -

0
0.00 B
-0.02
-0.04
-0.06 -
-0.08



50 C
I Control
F4 f Irrigation


30

O 20


10



May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun
Time (months)
Figure 4-3. Effects of rainfall patterns and dry-season irrigation on soil organic carbon
(C) to total nitrogen (N) ratios: (A) daily rainfall at the study site, (B) soil
water potential (SWP), (C) C:N ratios. A-B solid and open circles represent
means ( se) for control and irrigation treatments, respectively (N = 4 for
soil water potential and N=4 soil C:N ratio). Vertical dashed lines indicate
the dry season irrigation period (Sept 23rd 2004 to Jan 26th 2005). White
and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005),
and wet season (May to Sept 20th 2004, or Jan 20th to Aug 5th 2005),
respectively.













140
E 120 -
E 100
80
4--
.% 60
40
20

0.00 B
-0.02 -
-0.04 -
0 -0.06 -
-0.08 -
-0.10 -
-0.12 -

50 C
i Control
40 7 Litter removal

Z
30

0
20 lB




0
May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun

Time (months)
Figure 4-4. Effects of rainfall patterns and litter removal on soil organic carbon (C) to total
nitrogen (N) ratios: (A) daily rainfall at the study site, (B) soil water potential (SWP),
(C) C:N ratios. A-B solid and open bars represent means ( se) for control and litter
removal treatments, respectively (N = 4 for soil water potential and N=4 soil C:N
ratio). White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th
2005), and wet season (May to Sept 20th 2004, or Jan 20th to Aug 5th 2005),
respectively.









CHAPTER 5
CONCLUSIONS

Table 5-1 summarizes the responsiveness to intra-annual variability, dry-season irrigation,

and litter removal, exhibited by the suite of variables related to soil microbial structure and

composition, and microbial and nutrient dynamics, that were measured for this dissertation.

Overall my results reveal complex inter-relationships among soil microbial processes, microbial

structure/composition and nutrient availability in a seasonal tropical secondary forest.

Microbial biomass carbon, bacterial and fungal densities, basal respiration, soil phosphorus

and water availability showed marked intrannual variation associated with rainfall seasonality

(Table 5-1). Although rainfall seasonality affected these variables, dry-season irrigation had no

corresponding responses on the same variables (except to increased soil water potential). Higher

microbial biomass C in the dry season was possibly fostered by greater root mass density in the

same period (Vasconcelos 2006-dissertation), and linked to lower basal respiration as the soil

dried. Or else, nutrients accumulated in the microbial biomass in the dry season, become

available at the onset of the wet season, when plant growth is usually at its peak, a well known

nutrient-conserving strategy (Singh et al., 1989). Lower microbial C:N ratios in the wet season

across treatments gives to the latter alternative a plausible explanation to the pattern seen here.

Nonetheless, irrigation increased mycorrhizal-root infections and phosphatase activity, and

decreased fungal densities and ammonium availability (Table 5-1).

The long-term effects of litter removal compromised microbial structure and dynamics,

microbial processes, and soil solution nutrients; combined, these effects could ultimately impair

or delay aboveground processes. These effects included decreased microbial biomass C and P,

bacterial and fungal densities; net N-mineralization, phosphatase activity, basal respiration,

ammonium and phosphorus availability; but increases in mycorrhizal infections and nitrate









availability (Table 5-1). Although arbuscular mycorrhizal fungi could be an important nutrient

conduit in this tropical secondary forest, the extent to which litter removal will cause

impoverishment of the mineral soil that impairs microbial activity remains unclear.

The decrease in NH4+ and P availability was probably related to the fact that available

nutrients were retained in the living biomass and recycled in the litter, and removing the litter

had a negative effect on soil nutrients. In fact, litter removal consistently reduced litterfall N but

had no effects on litterfall P (Vasconcelos 2006). NH4+ availability was also linked to increased

microbial activity, but with decreased microbial biomass C as soil water availability increased.

Litter P concentrations were not affected by litter removal possibly because P-acquisition was

enhanced by greater mycorrhizal associations in litter removal plots as demonstrated in this

study, and in other secondary forests with low soil P (Marschner 1995). Microbial biomass N

was not affected by litter removal although N-mineralization and NH4+ availability decreased,

but was significantly linked with soil C:N ratios, suggesting that organic nitrogen sources may

sustain microbial demand for N in the absence of aboveground litter; microbial C:N ratios were

substantially reduced litter removal. The decrease in MBc in litter removal plots may have

resulted from the lack of substrate to sustain aboveground and belowground microbial processes,

and was correlated with the decrease in bacterial and fungal densities, and basal respiration rates.

Collectively, these findings confirm that water and substrate availability, seasonal droughts

and wet-up events have an important influence on the physiological state of the soil microbial

community and on nutrient availability in this tropical secondary forest. The evident and

consistent response to litter removal confirms its role as a conduit of nutrient and as a habitat for

microorganisms that actively serve as a reservoir for nutrients, and as substrate for fine roots and

mycorrhizal associations. The potential increase in nutrient uptake by fine roots and mycorrhizae









at the soil interface may have compensated for the lack of nutrients percolating through and from

the litter to the mineral soil in the litter removal plots, and may help explain why aboveground

measurements showed high resistance to altered nutrient availability through litter removal

(Vaconcelos 2006), and belowground measurements were more sensitive. Conversely, the

overall lack of response that belowground processes had to increased moisture availability may

indicate the great plasticity of these native microorganisms to seasonal drought, whereas

moisture availability at the study site significantly constrained ANPP in the year prior to the

sampling period of this study, interannual changes in ANPP were not followed by significant

changes in litterfall quantity or quality, nor significantly linked to belowground processes

measured in this study (Vasconcelos 2006).

The steady and reduced availability of NH4+ under irrigation may reflect that of an

environment without microclimate oscillations, buffered by the lack of seasonal droughts, and

intermittent decomposition. Although it is unclear how this response may ultimately affect forest

productivity, it is a strong confirmation of the pulse hypothesis articulated by Lodge et al.

(1994), especially when combined with the overall responsiveness to wet up events reported

throughout this dissertation.

Future research efforts should focus on the impacts of moisture and nutrient constraints on

microbial dynamics affects on aboveground processes (including C uptake) in tropical secondary

forests, as their importance to tropical landscapes continues to grow.









Table 5-1. Summary of ecosystem processes responses to intrannual variability of rainfall
seasonality (control plots), and to resource manipulations (dry-season irrigation and
bi-weekly litterfall removal). The responses of manipulative experiments are relative
to control treatments.
Process/Variable Intrannual Variability Dry-season Irrigation Litter Removal
Biomass C Yes 0 --
Biomass N No 0 0
Biomass P No 0 --
Bacteria Yes 0 --
Fungi Yes -- --
B:F Yes 0 0
Mycorrhizae No + ++
Spores No 0 0
Net Mineralization No 0 --
Net Nitrification No 0 0
Phosphatase No ++ --
Basal Respiration Yes 0 --
NH4 Yes -- --
N03- No 0 ++
P04- Yes 0 --
Soil C:N No 0 0
Soil water availability Yes ++ 0
Yes: presence of variability
No: absence of variability
+: slight, but significant increase
++: significant increase
--: significant decrease
0: no significant variation









APPENDIX
MICROBIAL BIOMASS COMPARISONS











Table A-1. Comparative seasonal and/or annual mean for extractable microbial biomass C, N and P (mgC/kg-soil) across differing
tropical systems, site, depth and soil type (modified w/ permission from Rangel-Vasconcelos 2002).
MBc MBEN MBp Proportionality
Cover Type Period Extraction Type mgC/kg-1 mgC/kg1 mgC/kg-1 soil constant (Kc, KN, Treatment Authors
Description soil soil Kp)

Rainy .. 487 51 20 KN = 0.68


r lu
Dry Tropical Forest


FTAp
Primary Amazon
Tropical Forest
FN
Native Forest
FTAp
Primary Amazon
Tropical Forest
FTAp
Primary Amazon
S Tropical Forest
c> FTsl3
Tropical-Dry
Secondary Forest
(13y Abandoned
Pasture)


FTRp
Primary Tropical
Rain Forest

FTws20
Tropical-Wet
Secondary Forest


662
744


70
88


1287


Kc = 0.35



Kc = 0.41


r tiugation-
incubation

Fumigation-
incubation

Fumigation-
extraction
Fumigation-
extraction

Fumigation-
incubation



Fumigation-
extraction


280-460


2000

1000
920
300
275
120


0.35
0.68


0.35
0.68


50-60


Fertilization
(N, P, N+P)


Nutrient
Gradient


Control


OiSigi CLt al.
(1989)

Luizto et al.
(1992)

Geraldes et al.
(1995)


Feigl et al.
(1995)


Davidson et al.
2'",1I4)


Cleveland et al.
(21i" 14)



Li et al. -(21" 4)


Winter
Summer

Annual
(avg)

Annual
(avg)
Annual
(avg)

Annual
(avg)


700-1500 70-80


Fumigation-
extraction


Fumigation-
incubation


Wet Season


Wet

Dry
Wet
Dry
Wet
Dry












Table A-1. Continued
MBc MBT MBp Proportionality
Cover Type Period Extraction Type mgC/kg-1 mgC/kg1 mgC/kg-1 soil constant (Kc, KN, Treatment Authors
Description soil soil Kp)
Wf7ot A8R Aln --/


FTwsl4
Tropical-Wet
Secondary Forest


FTw
Wet Tropical Forest


FTm
Moist Tropical
Forest (old growth)

FTws20
" Tropical-Wet
^- Secondary Forest


FTwsl7
Tropical-Wet
Secondary Forest


686.40
395.20
404.11
1080-1710

1050-1550


28
16
14.22


1000
< 10
1000
50


Dry
Dry
Dry
Wet+Dry
season
Wet+Dry
season
Wet
Dry
Wet
Dry
Wet+Dry
season
Wet+Dry
season
Wet
Dry
Wet
Dry
Wet
Dry


Fumigation-
extraction


Not reported

Not reported


Fumigation-
extraction


Fumigation-
incubation



Fumigation-
extraction


Kc= 0.35
KN = 0.54


KN = 0.54
Kp= 0.37


=0.35
= 0.54
= 0.40


Pre-treatment

Control
lyIRR
Control


Control

5ylRR

Control

7yLR
Control


4yIRR


4yLR


Rangel-
Vasconcelos et
al. (21 14)


Ruan et al.
(21", 14)


Yavitt et al.
(2"1114)



Li et al. (2005)


Veluci 2006


: estimated from a figure plot. 7y: 7 years of litter removal. ly: 7 years of litter removal. 5y: 5 years of dry-season irrigation


202
221.14
485.06
247.22
587.03
158.63
468.97


21.26
15.83
23.09
17.87
17.74
21.01









LIST OF REFERENCES


Alef, K., Nannipieri, P., Cepeda-Trazar, C. (1995) Phosphatase activity. Alef, K. and Nannipieri,
P., ed. Methods in Applied Soil Microbiology and Biochemistry. San Diego, CA,
Academic Press, 335-344.

Alef, K. (1995) Soil Respiration. Alef K., Nannipieri P., ed. Methods in Applied Soil
Microbiology and Biochemistry. San Diego, CA, Academic Press, 214-219.

Anaya, C. A.; Garcia-Oliva, F., and Jaramillo V.J. (2007) Rainfall and labile carbon availability
control litter nitrogen dynamics in a tropical dry forest. Oecologia 150 (4), 602-610.

Brookes, P.C., A. Landman, G. Pruden, and D.S. Jenkinson (1985) Chloroform fumigation and
the release of soil nitrogen: A rapid direct extraction method to measure microbial biomass
nitrogen in soil. Soil Biology and Biochemistry 17, 837-843.

Carvalho, C. J. R., Davidson, E. A., Santos, T. P., Dutra, F. C. and Serrao, B. 0. (2004)
Changing enzymatic activities and mycorrhizal infections in a chronosequence of
secondary and mature forests of eastern Amazonia. III Conferencia Cientifica do LBA.

Cavelier, J., S.J. Wright, and J. Santamaria (1999) Effects of irrigation on litterfall, fine root
biomass and production in a semideciduous lowland forest in Panama, Plant and Soil 211,
207-213.

Chander, K., Goyal, S., Nandal, D. P., Kapoor, K. K. (1998) Soil organic matter, microbial
biomass and enzyme activities in a tropical agroforestry system. Biology and Fertility of
Soils 27, 168-172.

Chapin III, F. S., Matson, P. A., Mooney, H. A. (2002) Terrestrial Plant Nutrient Use. In
Principles of Terrestrial Ecosystem Ecology, ed. F. S. Chapin III, P. A. Matson and H. A.
Mooney, Springer New York 176-196.

Chu, E. And Diekmann, U. (2002) Efeitos de Usos Alternativos do Solo sobre a Populacao de
Fungos Micorrizicos Arbusculares na Amazonia. Embrapa Amazonia Oriental, 1-20.

Cleveland, C. C., Towsend, A. R., Constance, B. C. (2004) Soil Microbial Dynamics in Costa
Rica: Seasonal and Biogeochemical Constraints. Biotropica 36, 184-195.

Coelho, R. F. R., Zarin, D. J., Miranda, I. S., Tucker, J. M. (2003) Ingresso e mortalidade de uma
floresta em diferentes estagios sucessionais no municipio de Castanhal, Para (Recruitment
and mortality of different stages of successional forest in the municipality of Castanhal,
Para). Acta Amazonica 33, 619-630.

Cornejo, F. H., Varela, A., Wright, S. J. (1994) Tropical forest litter decomposition under
seasonal drought: nutrient release, fungi and bacteria. Oikos 70, 183-190.









Cornejo, P.; Borie, F.; Rubio, R., and Azcon, R. (2007) Influence of nitrogen source on the
viability, funcionality and persistence of Glomus etunicatum fungal propagules in an
Andisol. Applied Soil Ecology 35, 423-431.

Davidson, E. A., Carvalho, C. R., Vieira, I. C. G., Figueiredo, R., Moutinho, P., Ishida, F. Y.,
Santos, M. T. P., Guerrero, J. B., Kernel, K. (2004) Nitrogen and Phosphorus Limitation of
Biomass Growth in a Tropical Secondary Forest. Ecological Applications 14, S150-S163.

Davidson, E. A., Verchot, L. V., Cattdnio, J. H., Ackerman, I. L., Carvalho, J. E. M. (2000)
Effects of soil water content on soil respiration in forests and cattle pastures of eastern
Amazonia. Biogeochemistry 48, 53-69.

Feigl, B. J., Sparling, G. P., Ross, D. G. and Cerri, C. C. (1995) Soil Microbial Biomass in
Amazonian soils: evaluation of methods and estimates of pool sizes. Soil Biology and
Biochemistry 27, 1467-1472.

Fisk, M. C. and Fahey, T. J. (2001) Microbial biomass and nitrogen cycling responses to
fertilization and litter removal in young northern hardwood forests. Biogeochemistry 53,
201-223.

Forster, J. C. (1995) Organic Carbon. Alef, K. and Nannipieri, P., ed. Methods in Applied Soil
Microbiology and Biochemistry. San Diego, CA, Academic Press, 59-65.

Frizano, J.; Vann, D. R.; Johnson, A. H., and Johnson, C. M. (2003) Labile phosphorus in soils
of forest fallows and primary forest in the Bragantina Region, Brazil. Biotropica, 35(1), 2-
11.

Geraldes, P. A., Cerri, C. C., Feigl, B. J. (1995) Biomassa microbiana de solo sob pastagens na
Amaz6nia. Revista Brasileira de Ci6ncia do Solo, 1, 55-60.

Gerdemann, J. W. and Nicolson, T. H. (1963) Spores of mycorrhizal endogene species extracted
from soil by wet sieving and decanting. Transaction British Mycology Society, 26, 235-
246.

Giovannetti N. and Mosse B. (1980) An evaluation of techniques for measuring VA mycorrhizal
infection in roots. New Phytologist, 84, 489-500

Giblin, A. E., Laundre, J. A., Nadelhoffer, K. J., Shaver, G. R., (1994) Measuring nutrient
availability in Arctic soils using ion-exchange resins A field test. Soil Science Society of
America Journal 584, 1154-1162.

Gollan, Schurr, U. and Schulze, E. D. (1992) Stomatal response to drying soil in relation to
changes in xylem sap composition of Helianthus annuus. Plant Cell and Environment 15,
551-559.

Hamel C. (2004) Impact of arbuscular mycorrhizal fungi on N and P cycling in the root zone.
Canadian Journal of Soil Science 31; 84(4), 383-395.









Hart, S. C., Stark, J. M., Davidson, E. A., and Firestone, M. K. (1994) Nitrogen mineralization,
immobilization, and nitrification. Methods of Soil Analysis. Microbiological and
Biogeochemical Properties SSSA, Madison, WI.

Hart, M., Trevors, J. T. (2005) Microbe management: application of mycorrhizal fungi in
sustainable agriculture. Frontiers of Ecology and Environment 3, 533-539.

Hodge, A. Robinson D., and Fitter, (2000). Are microorganisms more effective than plants at
competing for nitrogen? Trends in Ecology and Evolution 5(7), 305-308.

Islam, K. R., Weil, R. R. (1998) Microwave irradiation of soil for routine measurement of
microbial biomass carbon. Biology and Fertility of Soils 27, 408-416.

Jipp, P. H., Nepstad, D. C., Cassel, D. K., Carvalho, C. R. D. (1998) Deep soil moisture storage
and transpiration in forests and pastures of seasonally-dry Amazonia. Climatic Change 39,
395-412.

Johnson, N. C., O'Dell, T. E. and Bledsoe, C. S. (1999) Methods for Ecological Studies of
Mycorrhizae. Robertson, G. P., Coleman, D. C., Bledsoe, C. S. and Sollins, P., ed.
Standard Soil Methods for Long-Term Ecological Research. Oxford University Press, Inc.,
Oxford. 378-412.

Kaye J. P. and Hart S.C (1997) Competition For Nitrogen Between Plants And Soil
Microorganisms. Trends In Ecology & Evolution 12(4), 139-143.

Kiese, R., Butterbach-Bahl, K. (2002) N20 and CO2 emissions from three different tropical
forest sites in the wet tropics of Queensland, Australia. Soil Biology and Biochemistry 34,
975-987.

Lajtha, K., Driscoll, C. T., Jarrell, W. M., and Elliot, E. T. (1999) Soil Phosphorus:
Characterization and Total Elemental Analysis. In Standard Soil Methods for Long-Term
Ecological Research, ed. G. P. Robertson, D. C. Coleman, C. S. Bledsoe and P. Sollins,
115-142. Oxford University Press, Inc., Oxford.

Lean, J., Bunton, C. B., Nobre, C. A. and Rowntree, P. R. (1996) The simulated impact of
Amazonian deforestation on climate using measured ABRACOS vegetation
characteristics, in Amazonian deforestation and climate., ed. J.H.C. Gash, John Wiley,
Hoboken, NJ., 549-576.

Li, Y., Xu, M., Sun, 0. J., Cui, W. (2004) Effects of root and litter removal on soil CO2 effux
and microbial biomass in wet tropical forests. Soil Biology & Biochemistry 36, 2111-2114.

Li, Y., Xu, M., Zou, X., Xia, Y. (2005) Soil CO2 efflux and fungal and bacterial biomass in a
plantation and a secondary forest in wet tropics in Puerto Rico. Plant and Soil 268, 151-
160.

Lodge, D.J., McDowell, W.H., McSwiney, C.P. (1994) The importance of nutrient pulses in
tropical forests. Trends in Ecology and Evolution 9, 384-387.









Lopes, E.S., Oliveira, E., Dias, R., Scheck, N.C. (1983) Occurrence and distribution of vesicular-
arbuscular mycorrhizal fungi in coffee (Coffee arabica L.) plantations in central Sao Paulo
state, Brazil. Turrialba, 33(4), 417-422.

Lucas, Y., Luizao, F. J., Chauvel, A., Rouiller, J., Nahon, D. (1993) The relations between
biological activity of the rain forest and mineral composition of soils. Science 260, 521-
523.

Luizao, C. C. R., Bonde, T. A., Rosswall, T. (1992) Seasonal variation of soil microbial biomass
the effects of clearfelling a tropical rainforest and establishment of pasture in the central
Amazon. Soil Biology and Biochemistry 24, 805-813.

Luizao, F. J., Schubart, H. 0. R., (1987). Litter production and decomposition in a terra-firme
forest of Central Amazonia. Experientia 43, 259-265.

Malcolm, R. E. (1983) Assessment of phosphatase-activity in soils. Soil Biology & Biochemistry
15(4), 403-408.

Mamilov, A. S., Dilly, 0. M. (2002) Soil Microbial ecophysiology as affected by short-term
variations in environmental conditions. Soil Biology and Biogeochemistry 34, 1283-1290.

Markewitz, D., Davidson, E., Moutinho, P., and Nepstad, D. (2004) Nutrient loss and
redistribution after forest clearing on a highly weathered soil in Amazonia. Ecological
Applications 14(4) S177-S199.

Marschner, P., Marino, W., Lieberei, R. (2002) Seasonal Effects On Microorganisms In The
Rhizosphere Of Two Tropical Plants In A Polyculture Agroforestry System In Central
Amazonia, Brazil. Biology and Fertility of Soils 35, 68-71.

Mcgroddy, M. E., Silver, W. L., De Oliveira, R. C. (2004) The Effect of Phosphorus Availability
on Decomposition Dynamics in a Seasonal Lowland Amazonian Forest. Ecosystems 7,
172-179.

Mulvaney, R. L. (1996) Colorimetric Methods. In Methods of Soil Analysis, ed. D. L. Sparks, A.
L. Page, P. A. Helmke, R. H. Loeppert, P. N. Soltanpour, Tabatabai, M. A., A. E. Summer,
Vol. Chemical and Microbial Properties Agronomy Monograph no. 9, 1123-1184.
Chemical Methods SSSA, Madison, Wisconsin, USA.

Murphy, J., Riley, J. P. (1962) A modified single solution method for the determination of
phosphate in natural waters. Analytical Chemistry Acta 27, 31-36.

Orchard, V. A., Cook, F. J. (1983) Relationship between soil respiration and soil moisture. Soil
Biology and Biochemistry 15, 447-453.

Paul, E. A. and Clark F. E. (1997) Phosphorus transformations. Soil Microbiology and
Biochemistry. 2 ed. New York: Academic Press, 289-298.









Rangel-Vasconcelos, L. G. T., Zarin, D., Carvalho, C. J. R., Santos, M. M. L., Vasconcelos, S. S.
and Oliveira, F. A. (2005) Carbono, nitrogenio e atividade da biomassa microbiana de um
solo sob vegetacgo secundaria de diferentes idades na Amaz6nia Oriental. Revista de
Ciencias Agrarias.

Rilling, M. C. (2004) Arbuscular mycorrhizae and terrestrial ecosystem processes. Ecology
Letters 7, 740-754.

Robertson, G. P., Wedin D., Groffman P. M., Blair J. M., Holland E. A., Nadelholffer K. J. and
Harris D. (1999) Soil carbon and nitrogen availability: nitrogen mineralization,
nitrification, and soil respiration potentials. Robertson, G. P.; Coleman, D. C.; Bledsoe, C.
S., and Sollins, P., Editors. Standard Soil Methods for Long-Term Ecological Research.
Oxford: Oxford University Press, Inc. 258-271.

Ross, D. J. (1987) Soil Microbial Biomass Estimated by the Fumigation Incubation Procedure -
Seasonal Fluctuations and Influence of Soil-Moisture Content Soil. Biology &
Biochemistry, 19(4), 397-404.

Roy, S., Singh, J. S. (1995) Seasonal and spatial dynamics of plant-available N and P pools and
N-mineralization in relation to fine roots in a dry tropical forest habitat. Soil Biology &
Biochemistry 27, 33-40.

Ruan, H. H., Zou, X. M., Scatena, F. N., Zimmerman, J. K. (2004) Asynchronous fluctuation of
soil microbial biomass and plant litterfall in a tropical wet forest. Plant and Soil 260, 147-
154.

Sayer, E. J. (2005a) Consequences of changes in litterfall for soil respiration in tropical forests.
Chapter IV, dissertation for the degree of Doctor of Philosophy: Leaf litter manipulation in
a tropical forest. University of Cambridge, UK, 93-109.

Sayer, E. J. (2005b) Changes in nutrient cycling following increased or decreased litterfall in a
moist tropical forest. Chapter V, dissertation for the degree of Doctor of Philosophy: Leaf
litter manipulation in a tropical forest. University of Cambridge, UK, 93-109.

Sayer, E. J., Tanner, E.V. J., Lacey, A. L. (2006) Effects of litter manipulation on early-stage
decomposition and meso-arthropod abundance in a tropical moist forest. Forest Ecology
and Management 229, 285-293.

Sayer, E. J. (2006) Using experimental manipulation to assess the roles of leaf litter in the
functioning of forest ecosystems. Biological Reviews 81, 1-31.

Schimel, J. P. and Bennett, J. (2004) Nitrogen mineralization: challenges of a changing
paradigm. Ecology 85(3), 591-602.

Schwendenmann, L.,Veldkamp, E., Brenes, T., O'brien, J. J., Mackensen, J. (2003) Spatial and
temporal variation in soil CO2 efflux in an old-growth neotropical rain forest, La Selva,
Costa Rica. Biogeochemistry 64, 11-128.









Shock, C. C., Akin, I. A., Unlenen, L. A., Feibert, E. B. G., Cedric, A., (2003) Precise irrigation
scheduling using soil moisture sensors. International Irrigation Show 2003 Proceedings,
The Irrigation Association. San Diego, CA, 18-20.

Shuttleworth, W. J., Gash, J. H. C., Lloyd, C. R., Moore, C. J., Roberts, J., Marques, A. D.,
Fisch, G., Silva, V. D., Ribeiro, M. D. G., Molion, L. C. B., Sa, L. D. D., Nobre, J. C. A.,
Cabral, 0. M. R., Patel, S. R., Demoraes J. C., (1984) Eddy correlation measurements of
energy partition for Amazonian forests. Quarterly Journal of the Royal Meteorological
Society 110, 1143-1162.

Singh, J. S., Raghbanshi A. S., Singh, R. S., Srivastava, S. C. (1989) Microbial biomass acts as a
sink of plant nutrients in dry tropical forests and savanna. Nature 338, 499-500.

Sinsabaugh, R. L., Klug, M. J., Collins, H. P., Yeager, P. E., Peterson, S. 0. (1999)
Characterizing the Soil Microbial Communities. In Standard Soil Methods for Long-Term
Ecological Research, ed. Robertson, G.P., Coleman, D.C., Bledsoe, C. S., Sollins, P., Vol.
2, 318-348. Oxford University Press, New York.

Skogley, E. 0., Dobermann, A., (1996) Synthetic ion-exchange resins: soil and environmental
studies. Journal of Environmental Quality 25, 13-24.

Smith J.L., Halvorson J. J., Bolton H. (1994) Spatial Relationships of Soil Microbial Biomass
and C and N Mineralization in a Semiarid Shrub-Steppe Ecosystem. Soil Biology &
Biochemistry 26, 1151-1159.

Srivastava, S. C. (1992) Microbial C, N and P in dry tropical soils: seasonal changes and
influence of soil moisture. Soil Biology and Biochemistry 24, 711-714.

Ten6rio, A. R. D. M., Graga, J. J. D. C., G6es, J. E. M., Mendez, J. G. R., Gama, J. R. M. F.,
Silva, P. R. 0. D., Chagas, P. S. M. D., Silva, R. N. P. D., Americo, R. R., Pereira, W. L.
M., (1999). Mapeamento dos solos da estacgo de piscicultura de Castanhal, PA. FCAP
Informe Tecnico 25, 5-26.

Tobon, C., Sevink, J., Verstraten, J. M., (2004) Litterflow chemistry and nutrient uptake from the
forest floor in Northwest Amazonian forest ecosystems. Biogeochemistry 69, 315-339.

Treseder, K. K. and Vitousek P. M. (2001) Effects of soil nutrient availability on investment in
acquisition of N and P in Hawaiian Rain Forests. Ecology 82(4), 946-954.

Turrion, M. B., Gallardo, J., Gonzalez, M., (1997) Nutrient availability in forest soils as
measured with anion-exchange membranes. Geomicrobiology Journal 14, 51-64.

Vance, E. D., Brookes, P. C., Jenkinson, D. S. (1987) An improved method for measuring soil
microbial biomass C. Soil Biology and Biochemistry 19,703-707.









Vasconcelos, S. S., Zarin, D. J., Capanu, M., Littell, R., Davidson, E. A., Ishilda, F. Y., Santos,
E. B., Araujo, M. M., Aragao, D. V., Rangel-Vasconcelos, L. G. T., Oliveira, F. A.,
McDowell, W. H., Carvalho, C. J. R. (2004) Moisture and substrate availability constrain
soil trace gas fluxes in an eastern Amazonian re-growth forest. Global Biogeochemical
Cycles 18, 1-10.

Vasconcelos, S.S., Zarin, D.J., Rosa, M.B.S., Oliveira, F.A., Carvalho C.J.R. (2006). Leaf
decomposition in a dry season irrigation experiment in Eastern Amazonian forest
regrowth. To be submitted to Biotropica.

Vasconcelos, S. S. (2006) Mositure and nutrient constraints to ecosystem processes in a forest
regrowth stand in Eastern Amazonia, Brasil. Dissertation for the degree of Doctor of
Philosophy. University of Florida.

Veluci, R. M., (2006) Effects of irrigation and litter removal on microbial processes in a
seasonally dry tropical forest. Chapter III, Dissertation for the degree of Doctor of
Philosophy: Irrigation and litter manipulation in a seasonally dry tropical forest. University
of Florida.

Vitousek, P. M., Sanford. R. L., (1986) Nutrient cycling in moist tropical forest. Annual Review
of Ecology and Systematics 17, 137-67.

Vitousek, P. M., Hattenschwiler, S., Olander, L., and Allison S. (2002) Nitrogen and Nature.
Royal Swedish Academy of Sciences 31(2), 97-101.

Voroney, R. P., Winter, J. P., Gregorich, E. G. (1991) Microbe and plant interactions. In Carbon
isotope techniques, edts. D.C. Coleman and B. Fry, 77-99. Academic Press, New York.

Wardle, D. A. (2002) Communities and Ecosystems: Linking the Aboveground and
Belowground Components. In Global Change Phenomena in an Aboveground-
Belowground Context, ed. D. A. Wardle, Princeton University Press, Princeton, New
Jersey, 239-292.

Wieder, R. K., Wright, S. J. (1995). Tropical forest litter dynamics and dry season irrigation on
Barro Colorado Island, Panama. Ecology 76: 1971-1979.

Williams, M. A. and Rice, C. W. (2007) Seven years of enhanced water availability influences
the physiological, structural, and functional attributes of a soil microbial community.
Applied Soil Ecology 35, 535-545.

Yang, J. E., Skogley, E. 0., Schaff, B. E., Kim, J. J. (1998) A simple spectrophotometric
determination of nitrate in water, resin and soil extracts. Soil Science Society of America
62, 1108-1115.

Yavitt, J. B., Wieder, R. K., Wright, S. J. (1993) Soil nutrient dynamics in response to irrigation
of a Panamanian tropical moist forest. Biogeochemistry 19, 1-25.









Yavitt, J., Wright, S. (1996) Temporal patterns of soil nutrients in a Panamanian moist forest
revealed by ion-exchange resin and experimental irrigation. Plant Soil 183, 117-129.

Yavitt, J., Wright, S., Wieder, R. (2004) Seasonal drought and dry-season irrigation influence
leaf-litter nutrients and soil enzymes in a moist, lowland forest in Panama. Austral Ecology
29, 177-188.









BIOGRAPHICAL SKETCH

Roberta Medeiros Veluci-Marlow was born in July 27th 1976, in Franca SP, Brazil. She

graduated with a bachelor's degree in biology from the University of Franca in 1997 and

engaged in an exchange program in the US to learn English between fall 1998 and 1999.

Thereafter, Roberta acquired a master's degree in soil ecology from the University of Toledo in

2002 where she worked on N-cycling and microbiotic crusts. Her future goals include a career

as a researcher and developer of research-oriented educational programs. Roberta's biggest

accomplishment in life was giving birth to Tiago in February 2006.





PAGE 1

1 SEASONAL AND EXPERIMENTAL EFFECT S ON MICROBIAL COMPOSITION AND DYNAMICS IN A TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON, BRAZIL By ROBERTA M. VELUCI-MARLOW 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 Roberta M. Veluci-Marlow

PAGE 3

3 To my mom who taught me by example the value of completing a project.

PAGE 4

4 ACKNOWLEDGMENTS I wish to express sincere appreciation to my husband Brian for his endless assistance during all laboratory analysis a nd for postponing his own career and dreams to help me achieve mine. In addition, special thanks are to due to my major advisor Daniel Zarin whose patience and perseverance helped me throughout this proce ss, and committee members Michelle Mack, Nick Commerford, Tim Martin and William McDowe ll, whose critical eyes, and enlightened mentoring were instrumental and inspiring. This dissertation would not have been comp leted without the thrust and support of many collaborators and friends, esp ecially Steel Vasconcelos. For guidance during laboratory work I tha nk Elizabeth Chu, Claudio Carvalho. For field and laboratorial work, I thank Glebson and Branco, Robson Canuto, Fabio and Bruno, Jesus, Tereza, Ronaldo, Ivanildo, Tenilson, Maristela, An a Vania, Marcus, Malcher, and especially Deborah Arago and her family for continuous fi eld, laboratorial and em otional support. I also thank Livia Rangel-Vasconcelos whose work insp ired the proposal of this dissertation. Thanks also for Patricia Sampaio, Leslie, Leandra, Bob Buschbacher, Lucas Fortini, Marisa and Carol Boaventura for the endless he lp provided throughout this whole process. Thank you Cherie Arias, for enduring by my side at all times. The research was financially supported by an Andrew Mellon Foundatio n grant to Daniel Zarin and was conducted under cooperative agreem ents between the University of Florida, Universidade Federal Rural da Amaznia, and Embrapa Eastern Amazon.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES................................................................................................................ .........9 ABSTRACT....................................................................................................................... ............10 CHAPTER 1 SOIL MICROBIAL AND NUTRIENT DYNA MICS IN SEASONAL TROPICAL SECONDARY FOREST: RESPONSES TO CHANGES IN RESOURCE AVAILABILITY IN THE EASTERN AMAZON................................................................11 Introduction................................................................................................................... ..........11 Study Site..................................................................................................................... ...........14 Experimental Design............................................................................................................ ..15 2 SEASONAL AND EXPERIMENTAL EFFECTS OF MOISTURE AND SUBSTRATE AVAILABILITY ON MICROBIAL STRU CTURE AND COMPOSITION IN SEASONAL TROPICAL SECONDARY FO REST IN THE EASTERN AMAZON..........20 Introduction................................................................................................................... ..........20 Materials and Methods.......................................................................................................... .21 Soil Sampling and Processing.........................................................................................21 Microbial Biomass C and N............................................................................................22 Microbial Biomass P.......................................................................................................22 Bacterial and Fungal Densities........................................................................................24 AMF Root Colonization..................................................................................................25 AMF Spores Quantification............................................................................................26 Statistical Analysis........................................................................................................... .......27 Results........................................................................................................................ .............28 Seasonal Effects...............................................................................................................28 Irrigation Effects............................................................................................................. .28 Litter Removal Effects....................................................................................................29 Correlation Analyses.......................................................................................................29 Discussion..................................................................................................................... ..........31 Seasonal Effects...............................................................................................................31 Irrigation Effects............................................................................................................. .33 Litter Removal Effects....................................................................................................34

PAGE 6

6 3 SEASONAL AND EXPERIMENTAL EFFECTS ON MICROBIAL PROCESSES IN SEASONAL TROPICAL SECONDARY FO REST IN THE EASTERN AMAZON..........52 Introduction................................................................................................................... ..........52 Materials and Methods.......................................................................................................... .54 Net N-Mineralization and Nitrification...........................................................................54 Acid Phosphatase Activity..............................................................................................55 Basal Respiration.............................................................................................................56 Statistical Analysis........................................................................................................... .......56 Results........................................................................................................................ .............57 Seasonal Effects...............................................................................................................57 Irrigation Effects............................................................................................................. .57 Litter Removal Effects....................................................................................................58 Correlation Analyses.......................................................................................................58 Discussion..................................................................................................................... ..........59 Seasonal Effects...............................................................................................................59 Irrigation Effects............................................................................................................. .60 Litter Removal Effects....................................................................................................61 4 IRRIGATION AND LITTER REMOVAL EFFECTS ON SOIL NUTRIENT AVAILABILITY IN A SEASONAL TROP ICAL SECONDARY FOREST IN THE EASTERN AMAZON............................................................................................................73 Introduction................................................................................................................... ..........73 Study site and experimental design........................................................................................75 Materials and methods.......................................................................................................... ..75 Ion exchange resins.........................................................................................................75 Anion exchange membrane.............................................................................................77 Soil C:N ratios................................................................................................................ .78 Soil water potential (SWP)..............................................................................................79 Statistical analysis........................................................................................................... ........79 Results........................................................................................................................ .............80 Seasonal Effects...............................................................................................................80 Irrigation effects............................................................................................................. .81 Litter removal effects......................................................................................................82 Correlation analyses........................................................................................................83 Discussion..................................................................................................................... ..........84 5 CONCLUSIONS..................................................................................................................101 APPENDIX MICROBIAL BIOMASS COMPARISSONS........................................................105 LIST OF REFERENCES.............................................................................................................108 BIOGRAPHICAL SKETCH.......................................................................................................116

PAGE 7

7 LIST OF TABLES Table page 1-1 Pre-treatment microbial biomass, ba sal respiration, and metabolic quotient....................17 2-1 F-statistics for irrigation vs. control for soil microbial biomass C, N and P, bacteria and fungi colony forming units, and their ratio.................................................................37 2-2. Least square mean values for signifi cant season, treatment, and treatment by season contrasts associated with the irrigation experiment...........................................................37 2-3 F-statistics for litter removal vs. cont rol for soil microbial biomass microbial biomass C, N and P, bacteria and fungi colony forming units, and their ratio..................38 2-4 Least square mean values for significa nt season, treatment, and treatment by season contrasts associated with th e litter removal experiment....................................................38 2-5 Pearson correlation analysis across tr eatments among microbial biomass microbial biomass C, N and P, bacteria and fungi colony forming units, and their ratio..................39 2-6 Pearson correlation analysis across treatme nts among variables repo rted in chapter 2 and variables reported in chapters 3 and 4.........................................................................40 2-7 Comparative seasonal and/ or annual mean for extractable microbial biomass C, N and P across differing tropical syst ems, site, depth and soil type.....................................42 3-1 F-statistics for irrigation vs. control for microbial processes............................................64 3-2 Least square mean values for significa nt season, treatment, and treatment by season contrasts associated with the irrigation experiment...........................................................64 3-3 F-statistics for litter removal vs. control for microbial processes......................................65 3-4 Least square mean values for significa nt season, treatment, and treatment by season contrasts associated with th e litter removal experiment....................................................65 3-5 Pearson correlation analysis across treatments among N-mineralization rates, nitrification rates, phosphatase activity a nd basal respiration...........................................66 3-6 Pearson correlation analysis across treatme nts among variables reported in Chapter 3 and variables reported in chapters 2 and 4.........................................................................67 3-7 Seasonality, litter removal, and/or irrigation effects on N-mineralization, nitrification, phosphatase activity, basal re spiration, substrate induced respiration, and soil CO2 efflux across studies.....................................................................................68 4-1 F-statistics for irrigation vs. contro l for soil solution nutri ents and soil C:N....................90

PAGE 8

8 4-2 Least square mean values for significa nt season, treatment, and treatment by season contrasts associated with the irrigation experiment...........................................................90 4-3 F-statistics for litter removal vs. contro l for soil solution nutrients and soil C:N.............91 4-4 Least square mean values for significa nt season, treatment, and treatment by season contrasts associated with th e litter removal experiment....................................................91 4-5 Pearson correlation analysis acro ss treatments among ammonium, nitrate, phosphorus availability, and soil C:N................................................................................92 4-6 Pearson correlation analysis across treatme nts among variables reported in Chapter 4 and variables reported in chapters 2 and 3.........................................................................93 5-1 Summary of ecosystem processes respons es to intrannual vari ability of rainfall seasonality and to resource manipulations.......................................................................104

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Seasonality of soil CO2 efflux and daily rainfall at the study site.......................................17 1-2 S oil CO2 efflux at the study site: irrigation........................................................................18 1-3 S oil CO2 efflux at the study site: litter removal..................................................................18 1.4 Experimental plot design showi ng the distribution of treatments.....................................19 2-1 Soil microbial biomass: effects of rainfall patterns and irrigation......................................44 2-2 Soil microflora: effects of ra infall patterns and irrigation.................................................46 2-3 AMF root colonization and spores: effect s of rainfall patterns and irrigation.....................47 2-4 Effects of rainfall patterns and litter-removal on soil microbial biomass..........................48 2-5 Effects of rainfall patterns and l litter removal on soil microflora.....................................50 2-6 Effects of rainfall patterns and l litter re moval on AMF root colonization and spores.....51 3-1 Microbial processes: effects of rainfall patterns and irrigation .........................................69 3-2 Effects of rainfall patterns and litter removal on microbial processes...............................71 4-1 Nutrient availability: effects of rain fall patterns on cont rol and ir rigation.........................95 4-2 Effects of rainfall patterns on control and litter removal on nutrient availability..............97 4-3 Soil C:N: effects of rainfall patterns and irrigation...........................................................99 4-4 Effects of rainfall patterns and litter removal on soil C:N...............................................100

PAGE 10

10 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 SEASONAL AND EXPERIMENTAL EFFECT S ON MICROBIAL COMPOSITION AND DYNAMICS IN A TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON, BRAZIL By Roberta M. Veluci-Marlow August 2007 Chair: Daniel J. Zarin Major: Forest Resources and Conservation Tropical secondary forests are an increasin gly important land cover in the Brazilian Amazon, with 30 to 50% of the deforested area of the Brazilian Amazon in some stage of abandonment. This study investigated water and nutrient constraints on microbial dynamics and nutrient availability in a tropical secondary fo rest in the Eastern Amazon with manipulative experiments dry-season irriga tion and bi-weekly litte r removal using frequent sampling to capture seasonal and intra-annual fluctuations in Castanhal, Pa r, Brazil. Irrigation did not consistently alter microbial dynamics, except for lowered NH4 + availability and fungal densities, and increased phosphatase activit y. Litter removal decreased microbial biomass C and P, Nmineralization, phosphatase activity and NH4 + availability but increased NO3 availability. Intraannual variability was mainly driven by wet-up even ts in the dry season th at were not minimized by continuous irrigation (except for NH4 + availability), suggesting either that seasonal drought may not constrain the availability of nutrients or that irrigation wa s insufficient to cause a more significant effect. These results confirm the critical role of litterfall in tropical forest nutrient cycling and the importance of fluc tuations in soil moisture stat us to nutrient availability. How these belowground results interact with abovegr ound processes including C uptake, is a fertile area for future research and modeling.

PAGE 11

11 CHAPTER 1 SOIL MICROBIAL AND NUTRIENT DYNA MICS IN SEASONAL TROPICAL SECONDARY FOREST: RESPONSES TO CH ANGES IN RESOURCE AVAILABILITY IN THE EASTERN AMAZON Introduction This chapter serves as an overall introducti on to the study site and experimental design used for the research described in Chapters 2, 3 and 4. The main objective of this study was to examine how changes in moisture and substrate av ailability alter microbial community processes and nutrient dynamics in a seasonal tropical secondary forest following agricultural abandonment. The overall hypothesis is that mois ture and substrate av ailability constrain microbial and nutrient dynamics. This study a ddressed the effects of the following: Intra-annual variation associated with th e seasonality of ra infall or litterfall Wet-up and dry-down events Increased dry-season moisture availability i nduced experimentally by daily irrigation when monthly PPT < 150 mm Reduced substrate availability, induced experimentally by bi-weekly litterfall removal. In general, seasonal variation occurs in soil micr obial C, N, and P in tropical forests but the direction of seasonal change is not consistent across studies. Some studies show a negative correlation between biomass C, N and/or P a nd soil moisture (Ross 1987; Singh et al 1989; Srivastava 1992), whereas others have found a positive correlation between soil moisture and microbial biomass(Luizao et al 1992; Marschner et al 2002). At the st udy site used for the research described in this di ssertation, pre-treatment results showed higher microbial biomass and higher C:N ratio in the dry season (Rangel-Vasconcelos 2002; Table 1.1), and showed elevated soil CO2 efflux during the wet season (Vasconc elos et al. 2004; Figure 1.1). Reduced microbial biomass in the wet season may be due to a combination of lyses at the onset of the rains and increased microherbivory as the we t season progresses (Lodge et al 1994). In seasonally dry tropical forests, the microbial biomass may conser ve nutrients in biologically

PAGE 12

12 active forms during dry periods that are characterized by high microbial biomass and low turnover. Subsequent nutrient releases can occur rapidly during wetter periods that are characterized by low microbial bi omass and high turnover; such pulse events can serve to stimulate plant growth in nutri ent-poor tropical forest and savannah (Singh et al., 1989). Prolonged drought followed by flushes of rainfall can impose osmotic stress on microorganisms and result in pulses of nutrients from microbial minera lization, death and/or rapid turnover (Lodge et al., 1994; Wardle 2002). Dry season pulses may cause short-term changes in microbial composition and transient pulses of nutrients(Yavitt and Wright 1996). Irrigation to field capacity in an old-grow th tropical forest on Barro Colorado Island, Panama, has shown that alleviating moisture stress reduced the amount of forest floor mass throughout the year, increased ne t decomposition and decay duri ng the dry-season (Wieder and Wright 1995), and caused changes in microbial composition (Cornejo et al., 1994). Bacterial counts tended to be higher during the drier months of January, February and March and lower in April and May for irrigated plots. Fungi were depressed by irrigation and fungal counts were greatest when conditions were drie st (February and March) and dec lined after light rains in April and heavier rains in Ma y (Cornejo et al., 1994). A dry-season irrigation experime nt at the study site has mainta ined the soil moisture status relatively constant, favoring c ontinuous decomposition (Vasc oncelos et al., 2004). Soil CO2 efflux during the dry-season irrigation plots was 40% and 30% higher than in control plots in 2001 and 2002, respectively (Figure 1.2). By the end of the dry season irrigation period forest floor may be left with lower litter quality and qu antity as previously sh own in other irrigation studies (Yavitt et al., 1993; Wieder and Wright 1995; Yavitt and Wri ght 1996; Yavitt et al., 2004), but there was no evidence of that occurring at this site (Vasconcelos et al. 2004).

PAGE 13

13 Microbial biomass and its composition may also change in response to the quality and quantity of organic matter (Orchard and Cook 1983; Wardle 2002). Substrate composition, including C/N ratio, lignin cont ent and soluble compounds are im portant factors regulating the decomposition and mineralization rates of SOM (Mamilov and Dilly 2002), and the composition of the microbial biomass (Wardle 2002). Litter removal experiments in a tropical wet forest in Puerto Rico showed that monthly changes in soil microbial biomass were not sy nchronized with aboveground litter inputs, but rather preceded litterfall by one month (Ruan et al., 2004). There were also no correlations between soil microbial biomass and soil temperature, moisture or rainfall. They suggested that lower plant nutrient uptake or retranslocation of carbon and nutr ients to roots and stems prior to senescence could have triggered increases on microbial biomass one month prior to litterfall; first by lessening microbial competition for nutrien ts and later by increasing root exudates. Another study examining the effects of litter removal during 8 years in the White Mountain National Forest, New Hampshire (a young northern hardwood forest), concluded that litter removal had no effects on the microbial biomass, other than lowered respiration rates (Fisk and Fahey 2001). They concluded that belowgr ound C supply exerted greater control of forest floor microbial processes than fresh leaf litter inputs. During litter manipulations at the study site, Vasconcelos et al. (2004) showed that soil CO2 efflux in litter removal plots was significantly lo wer than in control plot s (Figure 1-3). This trend is apt to worsen with time as lab ile SOC is consumed and not replenished. Ongoing controlled manipulations at the study site, including (1) irrigating in the dry season and (2) bi-weekly litter removal provided an excellent opportunity to study how moisture availability and substrate removal affect microbial and nutrient dynamics. The following

PAGE 14

14 chapters describe several experiments designed to examine seasonal and experimental effects on microbial composition and structur e (Chapter 2), microbial proce sses (Chapter 3), and nutrient availability (Chapter 4). Study Site This study was conducted within the MANFLO RA project, a collaborative research program that includes the University of Florida and two Brazilian institutions, the Universidade Federal Rural da Amaznia Federal Rural Un iversity of Amaznia UFRA, and EMBRAPA Amaznia Oriental. Initiated in 1999, the overa ll goal of the project was to determine how changes in resource availability affect forest recovery following agricultural abandonment. Experimental treatments were dry-season irriga tion and litterfall removal in tropical forest regrowth and were implemented at the UFRA field station in Castanhal, Par, Brazil 1 19 S, 47 57 W. Mean SE annual rainfall received in the last 10 years in this area was 2539 280 mm, most of which falls between January and Ju ne. Mean daily temperatures fluctuate between 24 and 27C. The soils are classified as Distrophic Yellow Latosol Stony Phase I (Tenorio et al., 1999) in the Brazilian Classification, corresponding to Sombriustox in U.S. Soil Taxonomy. Soil particle size distribution in the fi rst 20 cm is 20% clay, 74% sand, and 6% silt. In the surface soil (0 10 cm), pH is 5.0, total C is 2.2%, total N is 0.15%, C:N is 14.4, and Mehlich-1 extractable phosphorus is 1.58 mg kg-1 (Rangel-Vasconcelos et al., 2005). Forest regrowth, annual crops, and active and de graded pastures characterize the landscape surrounding the field station. The stand under st udy was last abando ned in 1987 after mu ltiple cycles of shifting cultivation, beginning about si xty years ago when the old-growth forest was first cleared. Each cycle included cultivation of corn, manioc, and beans, for one to two years

PAGE 15

15 followed by fallow. T ypical shifting cultivation cycles lasted seven to ten years (G. Silva e Souza & O.L. Oliveira pers. comm.). The four most abundant overstory species are Lacistema pubescens Mart., Myrcia sylvatica (G Mey) DC, Vismia guianensis (Aubl.) Choisy, and Cupania scrobiculata Rich., representing 71% of all stems in th e stand. In July 2000, mean stem density was 2130 individuals per 100 m2, mean basal area was 13 m2 ha-1, and mean height was 4.9 m for the stand (Coelho et al., 2003). Experimental Design Plots were established in 1999 in 12-year-old fo rest regrowth. There were four replicate plots for the irrigation treatment, four plots for the litter removal treatment, and four control plots. Each plot was 20 x 20 m with a centrally nested 10 x 10 m meas urement subplot (Figure 1-4). Irrigation was applied in the late afternoon at a rate of 5 mm day-1, for about 30 min, during the dry seasons of 2001, 2002, 2003, and 2004. The amount of daily irrigation applied corresponds to regional estimates of daily evapotra nspiration (Shuttleworth et al., 1984; Lean et al., 1996; Jipp et al., 1998). Irrigation water was distributed through tapes with microholes every 15 cm. In 2001, irrigation tapes were spaced 4 m from each other. In 2002 the distance between tapes was reduced to 2 m to facilitate more even distribution of water (Vasconcelos et al., 2004). In 2004, irrigation was implemented between 21st September 2004 and 19 January 2005. We used rainfall and soil water potential data to define approximate boundaries for the dry and wet seasons. The start of the dry season was de fined by total rainfall less than 150 mm in the previous 30 d and soil suction more negative than .010 MPa; the end of the dry season was defined by total rainfall greater than 150 mm in the previous 30 d and soil suction less negative than .010 MPa. Since the soil suction data were obtained on a weekly basis, we estimate that the error in the location of seasonal boundaries is about 7 d (Vasconcelos et al., 2004).

PAGE 16

16 In the litter removal plots, leaf and branch fall were removed from the forest with plastic rakes every two weeks, beginning in August 2001 w ith the removal of th e pretreatment litter layer (538 35 g m-2, N = 4); C and N concentrations of the pretreatment litter layer were 41.0 0.9 and 1.3 < 0.01%, respectively (N = 8). Total new non-woody litterfall removed during the treatment period (August 2001 to May 2005) was 3051 111 g m-2 (Vasconcelos et al., 2004).

PAGE 17

17 Table 1-1 Microbial biomass (BMS) carbon (C-BMS), nitrogen (N-BMS), carbon to nitrogen microbial ratio (C:N MICROBIAL), basal respiration (CO2-BMS), and metabolic quotient (qCO2), from the study site (mean standard error). Data fro m Rangel-Vasconcelos (2002). VARIABLES Nov-00 (Dry-season)Apr-01 (Wet-season) C-BMS ( g C g-1 soil)* 924.8 60 425.25 59.04 N-BMS ( g N g-1 soil) 65.8 5.3 50.25 6.13 C:N MICROBIAL* 14.5 1.85 9.094 1.97 CO2-BMS ( g C-CO2 g-1 soil h-1)*1.7 0.08 2.6 0.20 q CO2 0.002 0 0.007 0 (*) indicate significant s easonal differences P < 0.05 Date 05/00 11/00 05/01 11/01 05/02 11/02 02/00 08/00 02/01 08/01 02/02 08/02 02/03 Rainfall (mm) 0 20 40 60 80 100 120 Soil CO2 efflux ( mol CO2 m-2 s-1) 0 2 4 6 8 Figure 1-1 Seasonality of soil CO2 efflux (mean se) and daily rainfall at the study site (N = 12). White and black horizontal bars indicate dry and wet seasons, respectively. Data from Vasconcelos (2006).

PAGE 18

18 Date 02/00 05/00 08/00 11/00 02/01 05/01 08/01 11/01 02/02 05/02 08/02 11/02 02/03 Soil CO2 efflux ( mol CO2 m-2 s-1) 0 2 4 6 8 Figure 1-2 Effects of dry-season irrigation on soil CO2 efflux (roots and microorganisms) in comparison to control plots at the study site (N = 12). White and black horizontal bars indicate dry and wet seasons, respectively. Gr ay bars represent irrigation period in the dry season. Standard errors are represented ( se). Data from Vasconcelos (2006). Date 02/00 05/00 08/00 11/00 02/01 05/01 08/01 11/01 02/02 05/02 08/02 11/02 02/03 Soil CO2 efflux ( mol CO2 m-2 s-1) 0 2 4 6 8 Figure 1-3 Effects of litter removal on soil CO2 efflux (roots and microorganisms) in comparison to control plots at the study site (N = 12). Litter remova l started on August 2001 (indicated by vertical line). St andard errors are represented ( se). Data from Vasconcelos (2006).

PAGE 19

19 Figure 1-4. Experimental pl ot design showing the di stribution of treatments.

PAGE 20

20 CHAPTER 2 SEASONAL AND EXPERIMENTAL EFFECTS OF MOISTURE AND SUBSTRATE AVAILABILITY ON MICROBIAL STRUCTURE AND COMPOSITION IN SEASONAL TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON Introduction Moisture and litterfall seasonality influence th e seasonal variation of soil microbial C, N and P in tropical forests but the direction of s easonal change is not c onsistent across studies. Some studies have shown a negative correla tion between biomass C, N and/or P and soil moisture (Ross 1987; Singh et al 1989; Sriv astava 1992), while others found a positive correlation (Luizao et al 1992; Ma rschner et al., 2002). The relationship between soil microbial biomass and litterfall seasonality is easily c onfounded with other variab les, including root exudates (Ruan et al., 2004), shifts in microbi al community structure (Cornejo et al., 1994, Marschner et al., 2002, Li et al., 2004 and 2005), N and/or P limitations (Fisk and Fahey 2001, Davidson et al., 2004), soil type (Feighl 1995, Villar et al., 2003, Cleveland 2004), nutrient pulses (Lodge et al., 1994), and l itter quality (Hodge et al., 2000). For this chapter, I measured the responses of microbial biom ass C, N, and P and microbial composition (bacterial and fungal densities, mycorrhiz al root-colonization a nd spore availability) to seasonal changes and wet-up events. Frequent drying and rewetting of soils favor is the portion of the microbial community best adapted to coping with th at stress (Gollan et al., 1992). Fast growing microbes capable of rapid growth on the labile substrates released from litter or microbial leaching into the soil during a rewet ting event may be predominantly composed of fungi and bacteria capable of w ithstanding sudden changes in soil water potential (Gollan et al., 1992). I also examine substrate and water cons traints on microbial biomass and composition within two ongoing manipulative experiments designe d to alter resource availability-dry-season irrigation and litterf all removal.

PAGE 21

21 Irrigation to field capacity in an old-growth tropical forest in Barro Colorado Island, Panama, has previously shown th at increasing moisture caused ch anges in microbial composition (Cornejo et al., 1994). Fungal dens ities decreased, while bacterial densities first increased and then decreased in response to irrigation, likely as a result of reduced li tter quality since more recalcitrant material is degraded by specialized groups of fungi. Litter removal has been previously employed to examine the consequences of reducing substrate availablity on microbial biomass and co mposition in a tropical wet forest in Puerto Rico (Ruan et al., 2004). Their results showed that monthly ch anges in soil microbial biomass were not synchronized with aboveground litter in puts, but rather preceded litterfall by one month. They found no correlations between soil microbial biomass and soil temperature, moisture or rainfall. Another study in Puerto Rico showed that microbial biomass C was significantly depressed by litter removal (Li et al., 2004), includi ng bacterial and fungal biomass (Li et al., 2005). In another experimental litte r manipulation in Panama, litter removal changed fungal species composition and diversity, decrease d soil fauna, decomposition rates, and nitrogen and phosphorus from incoming litterfall, wh ile litter addition produced no corresponding increases for those variables (Sayer et al., 2006). Materials and Methods Study site and experimental desi gn are described in Chapter 1. Soil Sampling and Processing During each of 13 monthly harvests, seven soil cores were taken per plot using a bulkdensity corer with 6 cm diameter and 5 cm dept h. These samples were composited by plot, sifted through a 2 mm mesh, placed in doubl e-folded, tightly closed plasti c bags, transported to the lab in a cooler no longer than 4 h after harvest, and stored at 4oC until analyzed. Gravimetric water

PAGE 22

22 content was determined using parallel subsamples dried at 105C for 24 h so that results could be expressed on a dry weight basis. Microbial Biomass C and N The fumigation-extraction method was used to estimate microbial biomass carbon and nitrogen (Vance et al., 1987). Chloroform fumiga tion (72 h) and direct extraction (agitation at 180 rpm in 0.5 M K2SO4 followed by filtration through Whatma n No. 42 paper) was performed on 20 g subsamples (Vance et al. 1987). Unfumiga ted subsamples were similarly extracted. The C present in fumigated and unfumigated extracts was determined colorimetrically by adaptation of a method describe d by Islam and Weil (1998). Aliquot s of 2 mL were mixed with 0.75 mL of 0.17M K2Cr2O7 and 2 mL H2SO4 in a 75 mL Pyrex digestion tube. The covered tubes were manually agitated and then heated to 150C for 10 min with glass beads. The samples were allowed to cool, 10 mL distille d water was added, and then the samples were analyzed spectrophotometrically at 590 nm, comparing concentrati ons to a standard curve with samples containing 0-2 mg C as sucrose dissolved in 0.5 M K2SO4. Microbial biomass-C was estimated from the chloroform-labile C, using a Kec factor of 0.35 (Voroney et al., 1991). Organic-N in the extract was digested using hydroge n peroxide and sulfuric acid at 260 C, and then determined colorimetrically (Mulvaney 199 6). Microbial biomass N was estimated from the chloroform-labile N, using KN factor of 0.54 (Brookes et al., 1985). Microbial Biomass P Microbial biomass phosphorus was determin ed using a modification of the method described by Lajtha et al. (1999) for Long Term Ecological Resear ch (LTER). Subsamples were analyzed in duplicate for fumi gated and non-fumigated samples. Conversion of the total dissolved, and organic dissolved phosphorus in fumigated and non-fumigated samples to inorganic phosphorus was done by Acid Pers ulfate Digestion and then determined

PAGE 23

23 colorimetrically as described by Murphy and R iley (1962). Microbial biomass P was estimated using a KP factor of 0.40 (Lajtha et al., 1999). Extraction : One gram of soil was weighted in d uplicates for each sample and transferred to 50 mL centrifuge tubes. All samples were agit ated with DI-water in centrifuge tubes for 16 h with 2 pre-conditioned BDH membrane strips (Gallard-Schlesinger Ind., Product # 55164 2S, Plainview, NY, 1-888-686-3454), to remove any inor ganic P. Strips were then removed and samples were centrifuged for 10 minutes at 5,000 rpm to discard supernatant. Non-fumigated samples were stored at room temperature for 24 h while paired samples were fumigated using 1 mL of chloroform (CHCl3) in each tube and placed uncapped in vacuum desiccators. A beaker was placed in the center of desiccators containi ng 25 mL of chloroform, closed air tight, vacuumed for 10 minutes and then left under vacu um for 24 h. After this period, the vacuum was released, chloroform volatiles eliminated, and 30 mL of 0.5 M NaHCO3 (sodium bicarbonate solution) pH 8.5 was added to all samples and le ft shaking for 16 hours in a platform shaker. Samples were then centrifuged for 10 minutes at 5,000 rpm and supernatant reserved for persulfate digestion and analysis. Persulfate digestion: A mixture of 0.8 g of potassium persulfate (K2S2O8), 10 mL of 0.9 M H2SO4 and 5 mL of sample extract was added to 50 mL volumetric flasks. Flasks were capped with aluminum foil and placed in an autoclave at 121 C, 17 psi for 50 minutes. Murphy & Riley (1962) colorimetric analysis for inorganic P was developed directly in ea ch flask after samples were allowed to cool down. All samples were read under a UV spectrophotometer set at 885 nm (Murphy & Riley 1962).

PAGE 24

24 Bacterial and Fungal Densities The serial dilution technique was used to quant ify densities of fungi and bacterial colony forming units (CFUs) at the st udy site (Cornejo et al., 1994). The method measures only a small portion of the total population. N onetheless, it is quite useful for studying changes in the population that grows on the chosen medium. Although more recent procedures offer whole community analysis thro ugh microbial lipid analysis, substrat e utilization, enzyme assays, and nucleic acid analyses, among others (Sinsabaugh et al. 1999), none was logi stically appropriate due to the unavailability of material and equipment at the study site. Each month, at the day of soil harvest, subsamples from composite soil samples from each plot were separated for analysis The soil subsamples were trans ported in open plastic bags and a 3 g sample from each plot was weighed, transferre d to a 200 mL Erlenmeyer, and agitated in a platform shaker at 300 rpm for 20 minutes in 97 mL of sterile water. Al l materials used during the procedure were autoclaved for 1h at 121 C, 17 psi., and prepared inside of an ultravioletlaminar flow chamber to avoid potential contamin ation. After 3 months of preliminary tests the best serial dilutions were -1 for fungi and -2 for bacteria. A 100 L aliquot of the appropriate dilution was added to 20 mL of solid medium, eith er bacterial or fungal, and swirled. Each liter of bacterial medium included 0.5 g dibasic-potassium phosphate (K2HPO4), 0.2 g of magnesium sulfate (MgSO4.7H20), 0.01 g of Fe-Na-EDTA, 0.25 g of egg albumin, 1 g of glucose, 1 g of yeast extract (levedura), and 20 g of TSA-agar dissolved up to volume in a water-bath, then autoclaved. Prior to agar addi tion, the medium was brought up to pH 6.8. Each liter of fungal medium included monobasic-potassium phosphate (KH2PO4), 1 g of magnesium sulfate (MgSO4.7H20), 5 g of peptone, 10 g of dextrose, and 20 g of TSA-agar all dissolved up to volume in a water-bath and then 3.3 mL of 0.1% Rose Bengal was added. After autoclaving, 0.3

PAGE 25

25 g of streptomycin was added under the ultraviole t-laminar flow chamber and mixed slowly to avoid bubble formation. Analysis was run in trip licates per plot (n = 12 per treatment), and several blanks were incubated with either medi um with sterile water under the same conditions to check for possible contamination. No contam ination was found during the study. Gravimetric water content was determined by drying soil subsam ples at 105C for 24 h so that results could be expressed on a dry weight basis. Only one type of agar medium was used for either bacteria or fungi, incubated at 28C for 1 and 2 days, respectively. AMF Root Colonization During each soil sampling at the study site, fine roots (< 1cm) were collected to quantify arbuscular mycorrhizal fungi (AM) colonizatio n (Giovannetti and Mosse 1980; Johnson et al 1999). The process was divided in five phases: i) selecting roots; ii) conserving roots until analysis; iii) clearing roots; iv ) staining roots and finally, v) preparing slides for quantifying AMF colonization. Since AMF can easily degrade af ter their removal from soil, after soil was sifted and roots were separated (1-2 g of roots 1mm diameter), they were carefully washed in tap water and conserved in a FAA solution (formalin, ethanol and CH3COOH glacial acetic acid) to conserve the root un til analysis. For clearing, roots we re soaked and resoaked in KOH 10%, until very clear. Then, after transfer to a water-bath for 15 minutes at 90 C, KOH was discarded; roots were rinsed w ith water several times and then soaked in 2% HCl for 10 minutes prior to staining. Tryplanblue (0.01 0, 05%) was used to stain and lactoglicerol to remove excess stain (a combination of glycerin, lactic acid and water, 1:1:1). Using a sharp blade and forceps, 20-25 root segments 1cm in length were separated to prepare microscope slides. Slides were prepared in duplicates per plot. Finally, the numb er of inoculations was verified using a compound or dissecting microscope (fitted with a hairline gratic ule in the eyepiece). Results

PAGE 26

26 were expressed as the average number of AMF per linear cm of root in two slides per plot ( n = 50). AMF Spores Quantification The method for quantification of mycorrhizal -fungi spores in soil was derived from a composite of the methods suggested by Gerdema nn and Nicolson (1963), L opes et al. (1983) and Johnson et al. (1999), with some modifications. A 30 g soil sample was transferred to a beaker with water, mixed with a stirring rod to separa te the supernatant, and then filtered through two sieves a coarser siev e (0.71 mm) on top of a thinner si eve (0.053mm). This process was repeated 4 times (add water to the soil, mix, wait for settling and then filt er through the sieves). In the end, the material retained on the thinner si eve was collected (residual with spores), and the remaining soil material discarded (material remain ing on the coarser sieve) During this process, the residual with spores was car efully transferred to a centrif uge tube and centrifuged at 2.000 R.P.M. for 3 minutes. Supernatant was then disc arded and approximately 40 mL of 45% sucrose solution was added to the remaining residual cont aining the spores, and centrifuged again. First spore collection consisted of decanting the sucr ose solution supernatant through a smaller sieve (0.021mm) where minuscule spores we re retained. Thereafter, spores were rinsed with water to remove excess sucrose and then transferred to a capped Nalgene bottle with water. Using the same centrifuged sample, sucrose was added for the second time and the same process was repeated. The second spore collection of each sa mple was then transferred to the respective bottle of the first collection a nd samples were aspired through a vacuum pump through a porous porcelain funnel where grid filter paper was inserted and spores re tained for count. Drained filter paper with spores was transferred to a labeled Petri dish slightly wet for best cohesion and read under a compound or dissecting microscope. Results were expressed as the number of spores per 30 g of soil.

PAGE 27

27 Statistical Analysis The SAS System for Windows V8 (2) was used for statistical analyses. All response variables of interest were l og-transformed to meet the mode l assumptions of normality. PROC MIXED was used using a repeated measures analys is with a heterogeneou s-autoregressive error structure. This structure allowed modeling within sample correlation ove r time and calculation of individual error variances for each sampling date Linear models were fitted on the variables microbial biomass C, N, P, fungal and bacterial densities (and their ra tio) with the following effects: season, date, treatment, treatment by s eason, treatment by date, plot and plot by date. CONTRAST statements were used to determine the significance of each fixed effect for each pair of treatment comparisons (i .e., control vs. irrigation and cont rol vs. litter removal) and leastsquares means were used to compare treatments and control means for the effects of season, treatment and treatment by season interaction on mi crobial biomass nutrients and their ratios, and on microbial densities and their ratios. For root colonization by AMF and spore counts, PROC MEANS was used for paired comparisons (Control vs. irrigation or Contro l vs. Litter removal) for each harvest date on log transformed variab les. A new variable was created (e.g., DIFF) containing the differences between the paired vari ables and then the T an d PRT options of PROC MEANS was used to test whether the mean difference significantly differed from zero. Within each treatment (control, irrigation, litter removal), Pearson correlation analyses were used to explore the bivariat e relationships of results reported in subsequent chapters with those reported here. Specificall y, I tested for correlations betw een microbial biomass C, N, P (and their ratios), fungal and bact erial densities (and th eir ratio), and : mineralization rates, nitrification rates, phosphatase activit y and basal respiration (Chapter 3), NH+ 4, NO3 and PO4 availability, total soil carbon and nitrogen (Chapter 4) and rainfa ll and soil water potential.

PAGE 28

28 Results There were no consistent effects of dryseason irrigation on the measured variables (Figures 2-1, 2-2, 2-3) but litter removal signif icantly reduced microbial biomass carbon and, to a lesser extent, microbial phosphorus (Figure 2-4), and bacterial and fungal densities (Figure 25), and had mixed but significant effects on ar buscular mycorrhizal fungi root-colonization (Figure 2-6). Seasonal Effects Microbial biomass C (MBC) was higher in the dry than wet season in control plots (Table 2-2), particularly after wet up events following an extended dry period (Figure 2-1a). There was no significant intra-annual variability for b acterial densities, but fungal densities were significantly higher in th e dry than wet seasons (Table 2-1 and 2-2). Irrigation Effects Microbial biomass C (MBC) was elevated in the dry s eason following an extended dry period (Figure 2-1a, Table 2-2). MBC, microbial biomass nitrogen (MBN) and phosphorus (MBP) varied significantly in response to date and the treatment by da te interaction; MBC was also significantly affected by season and the treatm ent-by-season interaction (Table 2.1). The MBP pool was lower in irrigation than control plots during the first ha rvest of the dry season on Oct.14 2004, then higher in irrigation plots towards th e end of the dry season on Dec.15 2004, following a substantial wet-up event after a long dry period (Figure 2.1e). Bacterial and fungal CFUs varied significantly in response to all main effects and their interactions, except for the main effect of trea tment which was significant for fungal CFUs only. There was also a significant effect of date a nd the treatment by date in teraction on B:F ratios (Table 2.1; Figure 2.2). Bacterial CF Us were higher in the dry than wet season in irrigation plots whereas fungal CFUs were higher in the dry season in control plot s and lower in irrigated than

PAGE 29

29 control plots in the dry season (Table 2-2). As a result, B:F ratio was higher in irrigated than control plots in the dry season (Table 2-2). Ther e was an abrupt increase in bacterial CFU during a wet up event in December, even in irrigated plots. However, sampling after the first rains following extended drought in March had no appa rent effects on CFU counts. Increased B:F ratios towards the end of the wet season result ed from low fungi-CFU and relatively constant bacteria-CFU. AMF root colonization in irri gated plots was significantly higher than in control plots during one sampling in the dry-season, but ther e was no effect in three wet-season samples. Continuous sampling showed a stea dy decrease in AMF spores in control plots towards the dry season but there were no signi ficant differences in AMF spor e counts between irrigated and control plots (Figure 2.3). Litter Removal Effects MBC, and to a lesser extent MBP, were si gnificantly reduced by litter removal; MBN was largely unaffected by this treatment (Figure 2-3; Tables 2-3, 2-4). Bact erial and fungal CFUs were reduced by litter-removal, and fungal CFUs were higher in the dry than wet season. As a resulty, bacterial to fungal ratio s were higher in the wet season (Figure 2-5, Tables 2-3, 2-4). AMF root colonization in litter removal plots was significantly higher than in control plots in two samples and lower in one (Figure 2.6a); there were no treatment effects on AMF spores (Figure 2.6b). Correlation Analyses Among the variables reported in this study, MBC in control and irrigation plots was negatively correlated with microbial biomass P (Table 2-5A and B) Under litter removal plots, MBC was positively correlated with fungi and corr espondingly correlated with bacterial to fungal ratios (Table 2-5C).

PAGE 30

30 Including variables measured in Chapters 3 and 4, MBC was negatively correlated with basal respiration and soil water potential across treatments (Table 2-6), with N-mineralization rates and phosphatase activity in cont rol plots (Table 2-6A), and with NH4 + under irrigation plots (Table 2-6B). In the contrary, MBC was positively correlated with NH4 + and PO4 in litter removal plots (Table 2-6C). MBN was positively correlated with fungal dens ities across treatments, and additionally with bacterial densities in irrigated pl ots only (Table 2-5). The variation in MBN across treatments was also attributed to phosphatase activit y (Table 2-6). MBN was also positively correlated with soil water potential under irrigati on (Table 2-6 B), and negatively correlated with basal respiration and soil C:N in lit ter removal plots (Table 2-6C). The increase in microbial biomass P in control plots was associated with decreases in fungal densities, and in litter re moval plots with increases in b acterial densities (Table 2-5). Microbial phosphorus was not correlated with micr obial composition in irrigated plots (Table 2-5 B), but was positively correlated with nitrate ava ilability, phosphatase activity, rainfall and soil water potential (Table 2-6B). Under litter removal, MBP was also positivel y correlated with phosphatase activity and basal respiration. Microbial biomass C:N was negatively correlated with phosphatase activity and soil water potential across treatments (T able 2-6), positively correlated with NO3 availability across treatments, and negatively correlated with NH4 + availability in irrigation and litter removal plots (Table 2-6) Microbial biomass N:P was nega tively correlated with rainfall and NO3 -, and positively correlated with NH4 + in irrigated plots. Under litter removal, MBN:P was negatively correlated with basal respiration, soil C:N and rainfall (Table 26). Biomass C:P ratios was negatively correlated with ba sal respiration and mineralization in control plots, with basal

PAGE 31

31 respiration in litter removal pl ots, and with phosphatase and so il water potential in irrigation plots (Table 2-6). Under litter removal, MBC:P was positively correlated with phosphorus availability (Table 2-6). Changes in bacteria to fungal ra tios were positively correlated w ith bacteria in control plots only; negatively correlated with fungi across treatments (Table 2-5), and with N-nitrification irrigated plots (Table 2-6). Fungal and bacteria l densities were positively correlated with NH4 + availability and N-mineralization rates in control plots (Table 26). Fungal densities were also positively correlated with N-minerali zation and N-nitrification rates in irrigated plots, and with N-mineralization and phosphatase in litter remova l plots. In addition to the positive correlation with N-mineralization, bacter ial densities were also pos itively correlated with NO3 -, and rainfall in litter removal plots. Discussion Seasonal Effects Previous assessments of microbial biomass in tropical forests tend to report one sampling period in the dry season and one in the wet season (Table 2.7). In the present study, microbial biomass C, N and P was relatively low compared to previous reports. Pre-treatment results at the st udy site also revealed greate r microbial biomass carbon and nitrogen in the dry than wet season (Range l-Vasconcelos et al., 2005). Monthly sampling throughout the year confirms the same seasona l trend for microbial biomass carbon across all treatments, but the opposite for biomass phosphorus in litter removal plots. The seasonal effect on microbial biomass C was associated with decreased basal respiration rates as a result of lowering soil wate r potential (Chapter 3). Lower microbial activity was accompanied by concomitant decreases in N-mineralization and phosphatase activity as microbial biomass C accumulated. These results support a very common conserving strategy in

PAGE 32

32 tropical forests with low nutrien t availability, i.e. high immob ilization of nutri ents in the microbial biomass in the dry season and subsequent release in the wet season when plant growth is greatest (Singh et al., 1989). In addition, highe r fungal densities in the dry than wet season but no significant changes in bacterial densities ma y indicate a predominantly fungal community in that period. This may help explain the corr esponding effects that fungal densities had on microbial biomass N and P, microbial nutri ent ratios, N-mineralization rates and NH4 + availability. Increased root activity in th e dry-season may also explain the boost on microbial biomass C shown in this study. Vasconcelos (2006) showed that fine root mass density across treatments was approximately twice as high in the dry seas on of 2004 as in the wet season of 2005 at the study site, which may lead to higher microbial ac tivity and leakage of carbon-rich exudates that becomes available to the microbial community. Contrary to my results, greater microbial biomass estimates corresponded to greater microbial activity in an oxisol forest in Co sta Rica; but the active portion of the microbial biomass remained relatively low regardless of soil C availability (Cleveland et al., 2004). Alternatively, the persistent increase in microbi al biomass C followed by decreases in microbial respiration may indicate the microbial biom ass increased efficiency by lowering their metabolism in that period (less CO2 evolved per unit biomass). In a seasonally dry tropical forest near Manaus (AM), a large proportion of the micr obial biomass died off in the dry season, and the activity of remaining microor ganisms were reduced (Marschne r et al., 2002). These effects could be enhanced by decreased root exudation, changes in microbial community structure and substrate utilization patterns.

PAGE 33

33 In general, soil microbial C or N co-varied wi th microbial C:N, and C:N ratios higher in the dry season across treatments (~ 30:1 in control plots and in irri gated plots, and 22:1 in litter removal plots), than in the wet season across treatments (~10:1), but maintained an intra-annual range similar to other tropical studies (Table 2-7). The seasonal diffe rence in microbial C:N ratios suggest that N availability was perhaps re cycled more efficiently in the wet season by the microbial biomass, than carbon. Irrigation Effects Although decomposition rates (k) were 2.4 times higher in irrigated than control plots during the sampling period (Vasconcelos 2006), i rrigation caused no consis tent enhancement of microbial biomass C, N, and P, or bacteria :fungal ratios, but signifi cantly reduced fungal densities. Overall, fungal densities were lower in irriga ted than control plots, but the decrease in fungal densities in irrigated pl ots was significantly greater in the wet season. The long-term effects of irrigation have probably affected funga l counts in agreement with studies that suggest that fungi thrives best under lower soil water potentials (Corne jo et al., 1994). Nonetheless, greater fungal and bacterial abundances in irrigated plots in the dry-season coincided with higher microbial biomass C, higher litter decomposition and fine root mass growth in the same period (Vasconcelos 2006); which may also explain th e positive influence of fungal densities on Nmineralization and nitrification rates in irrigated plots. Although there were no effects of dry-seas on irrigation on fine root mass growth (Vasconcelos 2006), the indirect e ffects of greater root C inputs may have enhanced microbial abundance and therefore the microbial biomass. However, there were no concomitant increases in microbial activity, as can be observed by the lowe st rates of microbial basal respiration in the same period (Chapter 3). Similar results were found by William and Rice (2007) under

PAGE 34

34 continuously wet soils in an upland tallgrass prairie. The decrease in fungal and bacterial densities in irrigated plots in the wet season could be a result of decreased litter quality due to continuous decomposition in the dry season, but irrigation did not change litterfall quality (Vasconcelos et al., 2006). Microbial biomass phosphorus was not affected by irrigation but responded well to wet up events, explaining the positive correlations with rainfall and soil water potential. The positive correlation between MBP and phosphatase activity may indicate that biotic demand for P drives phosphatase activity, and hence P mi neralization (Vitousek et al., 2002), but this correlation was observed only in control plots (Chapter 3 and 4) which suggests that ir rigation minimized this potential. The lack of correlation between MBP and phosphatase activity in irrigated plots may be due to higher phosphatase activity in irrigated than control plots (Chapter 3), which could alleviate the microbial demand for P through mi neralization, by the hi gher potential of P available through solubilization of inorganic/mineral P. In general, the relationship among MBP, microbial N:P and C:P across treatments, confirms the intertwined covariance among C, N and P-demand for microbial growth, and to enzymatic activities (Paul and Clark 1997, Tresed er and Vitousek 2001). Similar results was found by Cleveland et al. (2004), with the highest microbial C:P ratio o ccurring when microbial biomass C was at a maximum. The negative relationship between f ungal densities and MBP in control plots may indicate differe nt P-demands as microbial co mposition shifts or as fungi become more predominant. Litter Removal Effects Microbial biomass C and P were significantly lo wer in litter removal th an control plots, but only microbial biomass C varied seasonally. Si milar to the other treatments, increased MBC in the dry season seems related to wet up events, to the indirect effects that lower soil water

PAGE 35

35 potential had on basal respiration, and to increased fine root abundance and root C inputs in that period (as described in the previous section). The decrease in MBC in litter removal plots stems from the lack of substrate to sustain above ground and belowground microbial decomposition and microbial abundance. Bacterial and fungal abundances were lower in litter removal than c ontrol plots, but in agreement with other similar studies, fungal de nsities were higher within the dry than wet seasons (Cornejo et al., 1994; Willian and Rice 2007). The concomitant decrease in MBP, basal respiration and phos phatase activity between October and December in the dry season suggests that lower microbial activity may result in lower phosphatase activity, and perhaps, less soluble P for microbia l uptake. Increased mycorrhizal associations in response to litte r removal may indicate that investments in mycorrhizal associations were beneficial; al leviating the demand for P or N, since these associations can be highly costly for th e plant (Paul and Clark 1997; Rilling 2004). This result suggests that arbus cular mycorrhizal fungi coul d be an important nutrient conduit in this tropical secondary forest, consistent with their ro le in nutrient sequestration in nutrient poor soils (Rilli ng et al., 2004; Hart and Trevors 200 5, Sayer 2006). The deprivation of the litter layer and consequent impoverishment of the mineral soil could trigger investments in mycorrhizal associations; but the extent to wh ich this impoverishment will impair microbial activity remains unclear. A review by Sayer (200 6) showed that litter manipulation affected fungal growth and diversity by a ltering decomposers substrate, changing microclimate, leaching, availability of nutrients, and spore abundances although the direct ion of those changes were not consistent and often site-specifi c. In one of those studies, the removal of the organic layers returned the soil to an earlier su ccessional stage, resulting in gr eater abundance and diversity of

PAGE 36

36 fungal species. Forest succession studies in eas tern Par showed that AMF spores and mycorrhizal infections decrease with increasi ng forest stand age and become similar to the mature forest within 8 years of secondary forest succession (Carva lho et al., 2004, Chu and Diekmann 2002). Overall, biomass C:N was lower in litter rem oval than control plots, suggesting that the microbial biomass was not as N-deprived as expe cted in the absence of aboveground litter (Table 2-7). In the same period, litter removal substantia lly decreased N-minerali zation (Chapter 3), and litter nitrogen concentration, although this reducti on had not subsequent effects on litterfall mass (Vasconcelos 2006). These findings suggest microbi al access to N from an alternative nutrient pathway. Increased NO3 availability in litter removal plot s (Chapter 4) may have compensated for the increased biomass-N in relation to biomassC in litter removal plots. Greater mycorrhizal infections in litter removal plot s may have contributed to increas ed microbial N-availability but imposed higher costs for the symbiotic host (Pau l and Clark 1997), as observed by the decrease in litter N. In a nearby secondary forest, N amendments significantly improved tree growth, while P or N+P amendments did not, suggesting that N was the most limiting nutrient (Davidson et al., 2004). Although soil C:N ratios were not significantly affected by treatments during the sampling period (Chapter 4), correlation re sults suggested that increases in microbial biomass N and N:P corresponded to decreases in soil C:N in litter removal pl ots. This may indicate that soil nitrogen stocks may sustain microbial growth in the absence of litter substrate, and for that reason there were no litter removal effects on MBN. These results may have furt her implications in litter removal plots since without li tter input the fraction of ava ilable carbon and nitrogen may diminishes over time, and become cr itical for the microbial community.

PAGE 37

37 Table 2-1. F-statistics for the effects of treatment (Irrigation vs. control) seasons (Wet vs. Dry), date, and the interactions between treatm ent by season and treatment by date on soil microbial biomass carbon, nitrogen and phosphorus (MBC, MBN and MBP), bacteria and fungi colony forming units (B and F CF Us) and their ratio. = P < 0.05; ** = P < 0.01; *** = p < 0.001. Irrigation vs control Variable Season Date Treat Treat x Season Treat x Date MBC 50.2*** 9.1*** 1.2 17.2** 16.2*** MBN 4.4 40.6*** 2.1 2.0 39.0*** MBP 1.6 19.9*** 0.3 0.7 12.1*** B (CFUs) 7.4* 712.6*** <0.1 3.8* 3608.8*** F (CFUs) 10.0* 166.1*** 6.2* 5.6* 815.3*** B:F 2.0 49.1*** 3.7 2.4 267.0*** Table 2-2. Least square mean values for signi ficant season, treatment, and treatment by season contrasts associated with the irrigation e xperiment. (Lower and upper bounds of the 95% confidence interval are provided in parenthesis). Lower case letters indicate differences at P < 0.05 between treatment s both annually and within each season. Upper case letters indicate significant seasonal differen ces at P < 0.05 within each treatment. MBC (mgC/Kg-1soil) Bacteria (CFU/g-1 dry soil) Fungi (CFU/g-1 dry soil) Bacteria:Fungi Treatment LSM Irrigation 299 (269-328) 162754 (138862-190757) 1034a (841-1270) 148 (117187) Control 275 (246-305) 161135 (138291-187752) 1499b (1220-1841) 108 (86-135) Treatment by Season LSM Dry Season Irrigation 391A (340-441) 211081A (170979-260589) 1283a (1000-1646) 146a (105-202) Control 363A (313-413) 168047 (137974-204674) 1906bA (1485-2445) 88b (64-119) Wet Season Irrigation 241B (208-274) 125492B (99756-157868) 832a (604-1148) 151 (110-208) Control 221B (188-253) 157314 (125051-197899) 1178aB (855-1625) 133 (97-183)

PAGE 38

38 Table 2-3 F -statistics and associate significant levels ( p -value) for the effect of treatment (Litter removal vs. control), seasons (Wet vs. Dr y), date, and the interactions between treatment by season or treatment by date on soil microbial biom ass carbon, nitrogen and phosphorus (MBC, MBN and MBP), bacteria and fungi colony forming units (B and F CFUs) and their ratio. = P < 0.05; ** = P < 0.01; *** = p < 0.001. Litter removal vs. control Variable Season Date Treat Treat x Season Treat x Date MBC 40.06 ** 29.75 *** 11.4 ** 17.16 ** 74.27 *** MBN 1.04 44.45 *** 0.01 0.63 185.42 *** MBP 2.28 16.91 *** 4.89* 2.69 11.2 *** B (CFUs) 0.12 7.94 *** 9.87 ** 3.52 3627.63 *** F (CFUs) 16.81** 4.05 *** 6.04 7.69 ** 810.60 *** B:F 12.67 ** 3.52 *** 0 4.49 278.44 *** Table 2-4. Least square mean values for signi ficant season, treatment, and treatment by season contrasts associated with the litter rem oval experiment. (Lower and upper bounds of the 95% confidence interval are provided in parenthesis). Lower case letters indicate differences at P < 0.05 between treatment s both annually and within each season. Upper case letters indicate significant seasonal differen ces at P < 0.05 within each treatment. MBC (mgC/Kg-1soil) MBP (mgP/Kg-1soil) Bacteria (CFU/g-1 dry soil) Fungi (CFU/g-1 dry soil) B:F Treatment LSM Litter removal 204a (174-233) 0.46a (0.40-0.52) 114576 a (98198-133686) 1038a (845-1276) 109 (87-137 Control 275b (246-305) 0.56b (0.50-0.62) 162592b (139541-189449) 1499b (1220-1841) 108 (86-135) Treatment by Season LSM Dry Season Litter removal 276aA (226-327) 0.41 (0.33-0.49) 115128a (94525-140221) 1482A (1149-1912) 76A (55-105) Control 363bA (313-413) 0.53 (0.45-0.61) 168181b (138084-204838) 1906A (1485-2445) 88 (64-119) Wet Season Litter removal 158aB (125-191) 0.50 (0.41-0.57) 114119 (90715-143561) 727B (527-1003) 156 (113-214) Control 221bB (188-253) 0.58 (0.49-0.66) 157329 (125064-197919) 1178B (854-1624) 132B (96-182)

PAGE 39

39Table 2-5. Pearson correlation analysis for each treatment Contro l, Irrigation and Litter remova l, among variables reported in this chapter: microbial biomass carbon, nitrogen and phosphor us (MBC, N, P) bacteria (B), fungi (F ) and their ratio B:F. = P < 0.05; ** = P < 0.01; *** = p < 0.001. Control MBCMBNMBP B F B:F MBC ns -0.31* ns ns ns MBN ns ns 0.44** ns MBP ns -0.47** ns B ns 0.64*** F -0.67 B:F Irrigation MBC ns -0.34** ns ns ns MBN ns 0.38*0.49** ns MBP ns ns ns B 0.61***ns F -0.58*** B:F Litter removal MBC ns ns ns 0.53** -0.44** MBN ns ns 0.35* ns MBP 0.37 ns ns B ns ns F -0.77*** B:F

PAGE 40

40Table 2-6. Pearson correlation analysis for each treatment among variables reported in this ch apter and variables reported in other chapters: microbial biomass carbon, nitr ogen and phosphorus (MBC, N and P), and their ratios (MBC:N, C:P, and N:P), fungi (F), bacteria (B) and their ratio (B:F), mi neralization (MIN), nitrif ication (NIT), basal resp iration (B. resp.), soil C: N ratio, rainfall and soil water potential (SWP).* = P < 0.05; ** = P < 0.01; *** = p < 0.001. Control MBC MBN MBP MBC:N MBN:P MBC:P B F B:F NH4 + ns ns ns ns ns ns 0.38* 0.48** ns NO3 ns ns ns 0.34** ns ns ns ns ns PO4 3ns ns ns ns ns ns ns ns ns MIN -0.34** ns ns ns ns -0.41***0.42** 0.54***ns NIT ns ns ns ns ns ns ns ns ns Phosp. -0.26* 0.44*** ns -0.45***ns ns ns ns ns B. resp. -0.37** ns ns ns ns -0.36* ns ns ns Soil C:N ns ns ns ns ns ns ns ns ns Rain ns ns ns ns ns ns ns ns ns SWP -0.44*** ns ns -0.31* ns ns ns ns ns Irrigation NH4 + -0.31 ns ns -0.55***0.49***ns ns ns ns NO3 ns ns 0.29* 0.31* -0.39** ns ns ns ns PO4 3ns ns ns ns ns ns ns ns ns MIN ns ns ns ns ns ns ns 0.47** ns NIT ns ns ns ns ns ns ns 0.48** -0.51*** Phosp. ns 0.32** 0.27* -0.35** ns -0.29* ns ns ns B. resp. -0.44** ns ns ns ns ns ns ns ns Soil C:N ns ns ns ns ns ns ns ns ns Rain ns ns 0.37**ns -0.32* ns ns ns ns SWP -0.32* 0.34** 0.39**-0.41** ns -0.47***ns ns ns

PAGE 41

41Table 2-6. Continued Litter Removal MBC MBN MBP MBC:N MBN:P MBC:PB F B:F NH4 + 0.46*** ns ns -0.41**ns ns ns ns ns NO3 ns ns ns 0.33* ns ns 0.45**ns ns PO4 30.45** ns ns 0.33* ns 0.30* ns ns ns MIN ns ns ns ns ns ns 0.44**0.52**ns NIT ns ns ns ns ns ns ns ns ns Phosp. ns 0.58*** 0.35**-0.37**ns ns ns 0.34* ns B. resp. -0.34* -0.30* 0.30* ns -0.32* -0.34 ns ns ns Soil C:N ns -0.40** ns ns -0.50***ns ns ns ns Rain ns ns ns ns -0.33* ns 0.33* ns ns SWP -0.54*** ns ns -0.40**ns ns ns ns ns

PAGE 42

42Table 2-7. Comparative seasonal and/ or annual mean for extractable microbial biomass C, N and P (mgC/kg-1soil) across differing tropical systems, site, depth and soil t ype (modified w/ permission from RangelVasconcelos 2002). Superscripts depict notes on footnote (for more de tails refer to Appendix A). Study Site Depth (cm) Soil Description Period/ Season MBC mgC/kg-1 soil MBN mgC/kg-1 soil MBP mgC/kg-1 soil Treatment Authors Rainy 487 51 20 Winter 662 70 29 India (Vindhyan Hill Tract) 0-10 Ultisol Summer 744 88 31 Singh et al (1989) Manaus (AM) 0-5 Oxisol Annual (avg) 1287 Luizo et al (1992) Paragominas (PA) 0-10 Yellow Latosol Annual (avg) 476 35 Geraldes et al (1995) Manaus (AM) 0-10 Latosol Annual (avg) 659 102 Paragominas (PA) 0-10 Latosol Annual (avg) 695 111 Feigl et al (1995) 0-2.5 7001500 70-80 Paragominas (PA) 15-20 Latosol Wet 280-460 50-60 Fertilization (N, P, N+P) Davidson et al (2004) Wet 2000 250 8.5 SW Costa Rica 0-10 Oxisol Dry 1000 325 5.5 Nutrient Gradient Cleveland et al (2004) Wet 920 Dry 300 Control Wet 275 NE Puerto Rico 0-25 Mixed Isothermic Tropohum ult Dry 120 7y LR Li et al (2004)

PAGE 43

43Table 2-7. Continued Study Site Depth (cm) Soil Description Period MBC mgC/kg-1 soil MBN mgC/kg-1soil MBP mgC/kg-1soil Treatment Authors Wet 348.40 54 Dry 686.40 63 Pre-Treat 395.20 36 Control MANFLORA Castanhal-PA 0-10 Yellow Latossol/ Oxisol Dry 404.11 32 1y IRR RangelVasconcelos et al (2004) Annual range 10801710 Control NE Puerto Rico 0-10 OxisolUltisol Annual range 10501550 1y LR Ruan et al (2004) Wet 200 1000 Dry 150 < 10 Control Wet 300 1000 BCI (Panama) 0-15 Alfisol Dry 100 50 5yIRR Yavitt et al (2004) 618 Control NE Puerto Rico 0-10 Mixed Isothermic Tropohumult 202 7y LR Li et al (2005) Wet 221.14 21.26 0.59 Dry 485.06 15.83 0.53 Control Wet 247.22 23.09 0.61 Dry 587.03 17.87 0.55 4yIRR Wet 158.63 17.74 0.50 MANFLORA Castanhal-PA 0-5 Yellow Latossol/ Oxissol Dry 468.97 21.01 0.42 4yLR This study : estimated from a figure plot. 7y : 7 years of litter removal. 1y : 7 years of litter removal. 5y : 5 years of dry-season irrigation

PAGE 44

44 Figure 2-1. Effects of rainfall patterns on control ( ) and long-term dry-season irrigation ( ) plots in seasonally dry tropical forest. a) Daily rainfall at the study site, b) Soil water potential (SWP), c) Microbial bi omass C, d) Microbial biomass N, and e) Microbial biomass P. In b-e, values are means ( se) for n = 4 plots. White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet seasons (May to Sept 20th 2004, and Jan 20th to Aug 5th 2005), respectively. Vertical dashed lines indicate the dry season irrigation period 23rd September 2004 to 26 January 2005). ANOVA and treatment contrasts with fixe d effects by each collection date (* P < 0.05; ** P = 0.0001; *** P < 0.0001).

PAGE 45

45 Rainfall (mm) 0 20 40 60 80 100 120 140 SWP (MPa) -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 Biomass C (mg C/kg-1 soil) 0 100 200 300 400 500 600 Biomass N (mg N/ kg-1 soil) 0 20 40 60 80 100 May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Biomass P (mg P/kg-1 soil) 0.0 0.5 1.0 1.5 2.0 Time (months) a c b d e *

PAGE 46

46 Bacteria ( CFU/g -1 dry soil) 0 2e+5 4e+5 6e+5 Rainfall (mm) 0 20 40 60 80 100 120 Fungi (CFU/g -1 dry soil) 0 1000 2000 3000 4000 Rainfall (mm) 0 20 40 60 80 100 120 Dec Jan Feb Mar Apr May Jun Jul Bacteria:Fungi 0 100 200 300 400 500 Rainfall (mm) 0 20 40 60 80 100 120 Time (months)a b c Figure 2-2 Effects of rainfall patterns on control ( ) and long-term dry-season irrigation ( ) plots in seasonally dry tropical forest. Left yaxis: a) Bacterial colo ny forming units, b) Fungal colony forming units, a nd c) Bacteria to fungal colony forming units ratio. Right y-axis: daily rainfall at the st udy site during th e harvest period. ( n = 4). In a-c, values are means ( se) for n = 4 plots. White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (Jan 20th to Aug 5th 2005), respectively. Vertical dashed line repres ents dry-season irrigation period (Sept 23rd 2004 to Jan 26th 2005). ANOVA contrasts w ith fixed effects by each harvest date (*p < 0.05).

PAGE 47

47 0 10 20 30 Mycorrhizae (n of infections/cm -1 of root) Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun AMF spores (g-30 of soil) 0 10 20 30 40 50 Time (months)*a b Figure 2-3 Effects of rainfall patterns on control (solid bars) and long-term irrigation (grey bars) plots in seasonally dry tropical forest. a) AMF root colonization and b) AMF spores. In a-b values are means ( se) for n = 4 plots. White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (Jan 20th to Aug 5th 2005), respectively. Vertical dashed line repr esents dry-season irrigation period (Sept 23rd 2004 to Jan 26th 2005), *p < 0.05.

PAGE 48

48 Figure 2-4 Effects of rainfall patterns on control ( ) and long-term litter-removal ( ) plots in seasonally dry tropical forest. a) Daily rain fall at the study site, b) Soil water potential (SWP), c) Microbial biomass C, d) Microbial biomass N, and e) Microbial biomass P. In b-e, values are means ( se) for n = 4 plots. White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet seasons (May to Sept 20th 2004, and Jan 20th to Aug 5th 2005), respectively. ANOVA and treatment contrasts with fixed effects by each collection date (* P < 0.05; ** P = 0.0001; *** P < 0.0001).

PAGE 49

49 Rainfall (mm) 0 20 40 60 80 100 120 140 SWP (MPa) -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 Biomass N (mg N/ kg-1 soil) 0 20 40 60 80 100 Time (months) May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Biomass P (mg P /kg-1soil) 0.0 0.5 1.0 1.5 2.0 0 100 200 300 400 500 600 * **Biomass C (mg C/Kg-1 soil) a b d e c

PAGE 50

50 Bacteria ( CFU/g -1 dry soil) 0 2e+5 4e+5 6e+5 Rainfall (mm) 0 20 40 60 80 100 120 Fungi (CFU/g -1 dry soil) 0 1000 2000 3000 4000 Rainfall (mm) 0 20 40 60 80 100 120 Time (months) Dec Jan Feb Mar Apr May Jun Jul Bacteria:Fungi 0 100 200 300 400 500 Rainfall (mm) 0 20 40 60 80 100 120 *a b c Figure 2-5. Effects of rainfall patterns on control ( ) and long-term litter removal ( ) plots in seasonally dry tropical forest. In a-c, values are means ( se) for n= 4 plots. Left yaxis: a) Bacterial colony forming units b) Fungal colony forming units, and c) Bacteria to fungal colony forming units rati o. Right y-axis: daily rainfall at the study site during the harvest period. ( n = 4). White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (Jan 20th to Aug 5th 2005), respectively. ANOVA contrast s with fixed effects by each harvest date (*p < 0.05).

PAGE 51

51 Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Mycorrhizae (n of infections/cm-1 root) 0 10 20 30 Time (months) Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun 0 10 20 30 40 50 AMF spores (g-30 of soil)* a b Figure 2-6 Effects of rainfall patterns on control (solid bars) and long-term litter removal (grey bars) plots in seasonally dry tropical fore st. a) AMF root colonization and b) AMF spores. In a-b values are means ( se) for n = 4 plots. White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (Jan 20th to Aug 5th 2005), respectively (*p < 0.05).

PAGE 52

52 CHAPTER 3 SEASONAL AND EXPERIMENTAL EFFECTS ON MICROBIAL PROCESSES IN SEASONAL TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON Introduction Soil microorganisms exert control over nutri ent availability thr ough the processes of decomposition, immobilization, and mineralizat ion (Singh et al., 1989; Roy and Singh 1995; Chander et al., 1998; Mcgroddy et al., 2004). The tem poral variability of mi crobial processes in seasonally dry tropical forests is highly influe nced by water availabil ity, litterfall quality and quantity (Vitousek & Sanford 1986; Wardle 200 2; Roy and Singh 1995). Lucas et al. (1993) suggested that soil moisture controls a delicat e balance between the proc esses of immobilization and mineralization after observing that rewetting of seasonally dry Amazonian soils resulted in induced net immobilization, whereas dry periods a llowed mineral-N to accumulate (Lucas et al., 1993). Alternatively, wet and dry cycles may stimula te microbial activity (Cornejo et al., 1994) promote microbial turnover, desiccation, and osmotic stress, and dramatica lly affect detrital food chains leading to pulses of nutrient minera lization (Lodge 1994, Wardle 2002). Others have shown that CO2 evolution increases after rewetting comp ared to soil kept constantly moist (Orchard and Cook 1983; Vasconcelos et al. 200 4; Davidson et al. 2000; Kiese and ButterbachBahl 2002; Schwendenmann et al., 2003). Changes in rainfall and litterfall seasonality are known to influen ce the physiological and structural functioning of the soil microbial community (Yavitt 2004, William and Rice 2007), but their relative contribution to regulating shifts in microbial community composition, microbial processes, and soil nutrient ava ilability, is poorly understood. This chapter examines water and substr ate constraints on N-mineralization and nitrification, basal respiration, and acid-phosphatase activity, and th eir temporal variability in a 17-year-old secondary forest in the Eastern Amazon during dry and we t seasons within two

PAGE 53

53 ongoing manipulative experiments desi gned to alter resource availabi lity -dry-season irrigation and litterfall removal, de scribed in Chapter 1. Together, net mineralization and net nitrification assays provide insight into soil fertility and ecosystem function. Nitrogen mi neralization is often used as an index of nitrogen available to plants in terrestrial ecos ystems (Robertson et al., 1999). Mi crobial mineralization of soil organic matter (SOM) can provide a significant amount of the annual nut rient requirement of plants (Smith et al 1994) but there are other po tential N-available sources/pathways such as the direct uptake of amino acid DON without microbial mineralization of organic N to ammonium (NH+ 4) in tropical soils (Schimell and Bennett 2004). Direct uptake of organic N by plants and mycorrhizae has been demonstrated across diffe ring ecosystems and proven as an important conduit to plant nutrition (Kaye and Hart 1997; Schimell and Bennett 2004). Net nitrification refers to the conversion of ammonium to nitrat e by nitrifiers, bacteria that oxidize ammonium to nitrite and then, nitrate (Robertson et al ., 1999). Nitrification tends to dominate in systems with relatively high N av ailability, with lowe r plant and heterotroph competition for NH+ 4, allowing nitrifiers to flourish, slowly shifting the N economy of the system to NO3 dominated (Schimell and Bennett 2004). Basal respiration is defined as the respiration without the ad dition of organic substrate to soil under laboratory conditions. Basal respiration is frequently used as an index of microbial activities in soils and often correlates positively with soil organic matter and water availability (Alef 1995; Forster 1995; Ch apin III et al., 2002). Phosphatase activity improves the availability of soil phos phorus in nutrient poor soils such as tropical oxisols. Phosphatases catalyse the hydrolysis of organi c phosphomonoester to inorganic phosphorus which can be taken up by plan ts. Acid phosphatase is predominant in acid

PAGE 54

54 soils (4-6.5 as pH optimum) and ha s been detected in animal, pl ant and microbial cells (Eivazi and Tabatabai 1977). This chapter also integrates results from Ch apter 2 and 4 to show how microbial processes may be influenced by microbial biomass structur e and composition (Chapter 2), and ultimately affect soil solution nutrients (Cha pter 5). Most measurements were analyzed in response to 1) seasonal changes in moisture regime (wet vs. dr y season); 2) reduced moisture stress during the dry season (IRR) and, 3) reduced substrate availability (LR). Materials and Methods Study site and experimental desi gn are described in Chapter 1. Net N-Mineralization and Nitrification Net N mineralization was estimated from changes in NH4 + and NO3 concentrations during 7-day aerobic incubations at 27C of 20 g subsampl es (Hart et al., 1994). A ll extracts were stored frozen until analyzed. Ammonium concentratio ns were determined colorimetrically by the salicylate/nitroprusside method (Mulvaney 1996). Nitrate concentr ations were determined using a simple spectrophotometric procedure described by Yang et al. (1998). Results from this and all other assays are expressed on a dr y weight basis. Paired soil samples were first extracted with 100 mL 1 M KCl for 2 h, then filtered, decanted and re served for analysis. Paired subsamples were incubated for 7 days, and then extracted, fi ltered and decanted. Aliquots were analyzed for ammonium and nitrate contents. Net N-mineralized was expressed as a function of final nitrate and ammonium concentrations, minus initial a mmonium and nitrate concentrations, divided by incubation time (7 days). The net N-minera lization equation is shown in Equation 3-1.

PAGE 55

55 Nmineralized = [(Nitratef + Ammoniumf) (Nitrate0 + Ammonium0)]/Tdays (3-1) Where Nmineralized = net N mineralization rate, expressed as g N g-1 d-1 Nitratef = final nitrate concentra tion, expressed as g NO3 -N/g soil Ammoniumf = final ammonium concentr ation, expressed as g NH4 + -N/g soil Nitrate0 = initial nitrate concentr ation, expressed as g NO3 -N/g soil Ammonium0 = initial ammonium concentration, expressed as g NH4 + -N/g soil Tdays = incubation time/days Net nitrification potentials were measured by th e soil shaken slurry method (Hart et al., 1994). Fresh soils were sieved (<2 mm) and 15 g of moist soils were weighed into 250-mL flasks. The flasks received 100 mL of phosphate buffer (1 mM potassium phosphate, pH 7.2) and were continuously shaken for 24 h at the high speed (200 RPM). Ten-milliliter aliquots were sampled at 2 and 24 h (to represent initial and final nitrate concentra tions), then filtered, decanted, and analyzed colorimetrically as desc ribed above. Net nitrification was expressed as final nitrate minus initial nitrate concentrati on divided by incubation time (1d = 24 hours). The net nitrification equation is shown in Equation 3-2. Nnitrified = (Nitratef Nitrate0)/Tdays (3-2) Where Nnitrified = net nitrification ra te, expressed as g NO3 N g-1 d-1 Acid Phosphatase Activity The acid-phosphatase assay was used due to the acidic nature of the st udy site as suggested by Alef et al. (1995). The method wa s based on the determination of p -nitrophenol (PNP) released after the incubation of soil with the artificial substrate p -nitrophenyl phosphate for 1h at 37C. Results were expressed as gPNP/g-1 dry soil/h-1. Paired soil samples (1g) were placed in a 10 mL flask with 4mL of Universal Modified Buffe r solution at pH 6.5 and placed in desiccators with vacuum for 5 minutes to increase solution absorption among soil aggregates. After samples were incubated at 37C for 10 minutes, 1mL pnitrophenyl phosphate solu tion (PNP substrate) was added, mixed, and re-incubated for 1h followed by the addition of 4mL CaCl2 (0.5 M ) and

PAGE 56

56 1mL NaOH (0.5 M ), and transferred to centrifuge tube s. After centrifuging supernatant was saved and diluted to 40X (e.g, 0.25 mL sample to 9.75mL water) prior to absorbance reading on a spectrophotometer set at 400nm. Results were co rrected using control samples run in parallel, and p-nitrophenol concentration was calculated per milliliter of sample in reference to a calibration curve. Results are expr essed on a dry weight basis. Basal Respiration Basal respiration is defined as the respiration without the ad dition of organic substrate to soil and was determined using the traditional proc edure described by Isermeyer (1952) revised in Alef and Nannipieri (1995). Soil sa mples (25g) were incubated for 6 d in closed jars with 25 mL 0.05 M NaOH solution. CO2 trapped in NaOH was determined by HCl titration. Results are expressed on a dry weight basis. Statistical Analysis The SAS System for Windows V8 (2) was used for statistical analyses. PROC MIXED was used for a repeated measures analysis with a heterogeneous-autoreg ressive error structure arh (1)]. This structure allowed modeling within sample correlation ove r time and calculation of individual error variances for each sampling date Linear models were fitted on the variables mineralization rates, nitrification rates, phospha tase activity and basal respiration with the following effects: season, date, treatment, treatme nt by season, treatment by date, plot and plot by date. CONTRAST statements were used to de termine the significance of each fixed effect for each pair of treatment comparisons (i.e., control vs. irrigation and control vs. litter removal) and least-squares means were used to compare trea tments and control means for the effects of season, treatment and treatment by season interacti on on mineralization rates, nitrification rates, phosphatase activity and basal respiration. Within each treatment (control, irrigation, litter removal), Pearson correlation analyses were used to explore the bivariate relationships of results

PAGE 57

57 reported in previous and subseque nt chapters to those reported he re. Specifically, I tested for correlations between mineraliza tion rates, nitrification rates, phosphatase activity and basal respiration with microbial biomass C, N, and P and their ratios, fungal and bacterial densities (Chapter 2), NH+ 4, NO3 and PO4 and (Chapter 3), as well as ra infall and soil water potential. Results Seasonal Effects Basal respiration rates were lower in the dry th an wet seasons in control plots, and within treatments (Tables 3-2 and 3-4). There was no seasonal effect on any of the other measured processes. Irrigation Effects Net nitrogen mineralization was unaffected by the irrigation treatment, but varied substantially at different sampling dates (Figure 3-1), and that is reflected in the significant effect of date on mineralization (Table 3-1). Duri ng one occasion in the wet season of 2004, Nmineralization was higher in irrigated than contro l plots (Figure 3-1C), a nd that is reflected by the significance of the treatment-by-date inter action (Table 3-1). Ni trification rates were significantly lower in irrigated pl ots on one sample date at the onset of the dry season (Figure 31), and were also significantly affected by treatment and the treat ment-by-date interaction (Table 3-1). Phosphatase activity was elevated by irrigation on three sample dates (Figure 3-1). Date, treatment, and treatment-by-date effects were significant for phosphatase activity (Table 3-1). Across all sampling dates, phosphatase activity, on average, was slightly but significant higher in the irrigated plots (Table 3-2). Basal respiration was not significantly affect ed by the irrigation treatment, although small enhancements of basal respiration in the irriga tion plots were apparent on three sample dates

PAGE 58

58 (Figure 3-1). Season, date, treatment-by-season, a nd treatment-by-date effects were significant for basal respoiration (Table 3-1), with values on average about 50% higher in the west season than in the dry season (Table 3-2). Litter Removal Effects Nitrogen mineralization was consistently a nd significantly reduced by litter removal (Figure 3-2); treatment, treatment-by-season, and treatment-by-date eff ects were significant (Table 3-3). On average, the reduction in N-mineralization due to litter removal was about 40% (Table 3-4). In contrast, nitrifi cation rates were largel y unresponsive to litter removal (Figure 32), and only the treatment-bydate interaction effect wa s significant (Table 3-3). Phosphatase activity was also consistently and significantly reduced by litter removal (Figure 3-2); date, treatment, tr eatment-by-season, and treatment-bydate effects were significant (Table 3-3). On average, the reduction in phophatase activity due to litter removal was ~30% (Table 3-4). Basal respiration also tended to be significantly lower in litt er removal than control plots, but seasonal and date effects on treatmen t were dominant (Table 3-4) and the effect was inconsistent over time (Figure 32). All main and interaction e ffects were significant for basal respiration (Table 3-3). Correlation Analyses N-mineralization was positively correlated with phosphatase activit y, basal respiration (Table 3-5), fungal and bacter ial densities (Table 3-6) across treatme nts; and negatively correlated with N-nitrification across treatments (Table 3-5). N-mineralization was positively correlated with NH4 + availability in contro l plots only, and negatively correlated with microbial biomass C and C:P (Table 3-6). In irrigated plots, N-mineralizati on was also negatively correlated with soil water potential. Under litte r removal, N-mineralization was also positively correlated with phosphorus av ailability (Table 3-6).

PAGE 59

59 Nitrification rates were negativ ely correlated with bacterial to fungal ratio in irrigated plots, and with NO3 in control plots. Under irrigation nitrification rates were positively correlated with fungal densities ; and with rainfall in cont rol plot (Table 3-6). Phosphatase activity was negatively correlated with microbial biomass C:N and positively correlated with NH4 + availability and microbi al biomass N and P across treatments (Table 3-6). In control plots, phosphatase ac tivity was also negatively correl ated with microbial biomass C and phosphorus availability (Table 3-6). Under i rrigation, phosphatase activ ity was also negative correlated with C:P ratios and positively correlat ed soil water potential; and in litter removal plots also positively correlated w ith fungal densities (Table 3-6). Basal respiration was negatively correlated with microbial biomass C across treatments and with C:P in control and li tter removal plots (Table 3-6). Basal respiration also varied positively with soil water potential in control and litter removal plots, but not in irrigated plots (Table 3-6). In litter removal plots, basal re spiration was also negatively correlated with microbial biomass N, N:P, bacterial to fungal ratios and P-availability, and positively correlated with microbial biomass P and rainfall (Table 3-6). Discussion Seasonal Effects Rainfall seasonality imposed a consistent decreas e in microbial basal respiration in the dry season across all treatments, even in irrigated plot s. Similar results was reported for this site by Rangel-Vasconcelos (2004), with higher microbi al basal respiration during the 2001 wet season compared to the previous dry season, as observe d in other tropical fore sts (Luizao et al., 1992; Cleveland et al., 2004). Unlike measurements of soil CO2 efflux, which does not discriminate between autotrophic (e.g., roots) and heterotrophic re spiration (e.g., microbial), decr eases in basal respiration

PAGE 60

60 indicated lower microbial ac tivity in that period. Soil CO2 efflux previously meas ured at this site was also significantly lower in the dry than we t season in control and li tter removal plots, but significantly enhanced in irri gated plots (Vasconcelos et al., 2004; Table 3-7). Although CO2 efflux may not be compared to microbial basa l respiration rates meas ured under laboratory incubations, these results are s uggestive that the increase in CO2 efflux in irrigated plots in the dry season was attributed to increased root resp iration and/or increased activity of microbes in decomposing aboveground litter ra ther than soil microbes. Irrigation Effects Dry-season irrigation had a limited and incons istent impact on microbial processes. Although decomposition rates (k) were 2.4 times highe r in irrigated than c ontrol plots during the sampling period (Vasconcelos 2006 ), irrigation had no significan t effects on N-mineralization, nitrification and basal re spiration, but slightly in creased phosphatase activity. The weak response to water availability can be observed by the lack of seasonal differences in phosphastase activity, although it was positively correlated with soil water potential. In contrast, phosphatase activity was higher in the dry th an wet season in a wet tropical forest in Costa Rica (Cleveland et al., 2004), and in a mature tropical forest on the BCI following closed laboratory incubations that simulated the dry-season (Yavitt et al., 2004). However, dry-season irrigation had no effects on phosphatase activity on the BCI study (Table 3-7). Lower phosphatase activity in the wet season was attrib uted to aging of litter under prolonged wet conditions; microorganisms decomposing fresher litter, immobilized more P and synthesized more phosphatase enzymes. P-availability may increase as a result of phosphatase activity (Malcolm et al., 1983), and this potential was observed as the positive co rrelation between phosphata se activity and biomass P, and consequent decreases in the microbial biomass C:P ratio, alt hough phosphatase activity

PAGE 61

61 had no direct effects on P-availabi lity (Chapter IV). The slight in crease in phosphatase activity in irrigated plots could also be due to higher mycorrhizal infec tions during the irrigation period (Chapter 2). Cornejo et al. (2007) showed that phosphatase ac tivity was enhanced in mycorrhizal soils where N was supplied as NO3 although P uptake by arburscular mycorrhizae increased by 25% irrespective of the N source supplied. The consistent correl ation between phospha tase activity and NH4 + availability and biomass N may be related to the role of N in enzyme production. When P is in short supply, microbial N investments for enzyme production to acquire more N may not be as rewarding as using N to acquire more P (Vitousek et al., 2002). Litter Removal Effects The decrease in N-mineralization in res ponse to litter removal occurred without corresponding decreases in soil organic matter (Cha pter 4), and may reflect a direct effect of decreased litter substrate for microbial activity. Hence decreases in bacterial and fungal densities, and microbial biomass C in litter removal plots (Chapter 2), may have contributed to decreases in N-mineralization. Changes in microbial immobilization and mine ralization rates can also be related to changes in microbial C:nutrient ratios in relation to their substrates ra tios (Hodge et al., 2000). In this study, the consistent decrease in N-minera lization rates was not equi valent to decreases in microbial biomass N (Chapter 2), or nitrogen ava ilability (Chapter IV), but was positively related to P-availability, phosphat ase activity and basal respiration, and thus, increases in microbial activity. For discussion of the re lationship among N-mineralization, biomass-N, and litterfall N, refer to Chapter 2 (pp. 38-38). Litter removal had no effects on N-oxide emissions between 2000 and 2002, but nitrification rates were marginally lower than in control plots in that peri od (Vasconcelos et al.,

PAGE 62

62 2004). Prolonged litter-removal has had no further effects on nitrificat ion rates based on my laboratory incubation results. As expected, net N-nitrif ication covaried with N-mineralization across treatments, but no other variable aff ected nitrification ra tes even though nitrate availability was higher in litter remova l than in control plots (Chapter 4). Variability in microbial activity reflected by basal respir ation may explain variations in phosphatase activity, which then causes changes in phosphorus availability and microbial biomass C, N, and P (and their ratios). Howeve r the causal relationships among these variables remain unclear. Although phosphatase was lower in litter removal plots, the pattern of variation was very similar to control and irrigation pl ots, but not related to soil wa ter potential or rainfall events. Phosphatase activity was related positively with fungal densities, and perhaps influenced by the substantial increase in mycorrhi zal infections in litter removal than in control plots. The significant decrease in fungal dens ities in litter removal plots (Chapter 2) may account to the decrease in phosphatase activity, but also to the decrease in N-mineralization since these processes appeared to be co-varying altogether. The positive effect of rainfall and soil water potential on basal respiration in litter removal plots contradicted the lack of res ponse to continuous water availabil ity in irrigated plots, but it is in agreement with the seasonal eff ect that resulted in lower rates in the dry season. If water was a major factor deriving this pattern, a corresponding in crease in basal respiration should be seen in irrigation plots, unless irrigation was insuffici ent. Nonetheless, this decrease was not accompanied by decreases in biomass C, root biom ass, and microbial densities as previously discussed in Chapter 2 (pp. 35-36) Decreases in soil respiratio n following litter removal in a mature tropical forest on the BCI were attributed to a substantial decline in the soil microbial

PAGE 63

63 biomass, and to a slight decrease in fine root biomass. Higher soil respiration in the wet season was attributed to a concomita nt increase of the same vari ables (Sayer 2004, Table 3-7). Likewise, there was a 2-3 fold decrease in soil CO2 efflux in a secondary forest in Puerto Rico due to root and litter removal, followed by root or litter removal alone. This decrease was followed by substantial decrease s in microbial biomass following the same pattern of CO2 efflux across treatments (Li et al., 2004, Table 3-7). This chapter illustrates the role of litter as an important conduit for nutrient transfer to microorganisms and microbial processes that can ultimately affect the recovery and productivity of secondary forests in Eastern Amazon. My re sults also reveal in ter-relationships among microbial processes, microbial structur e/composition and nutrient availability.

PAGE 64

64 Table 3-1. F -statistics and associated significance levels ( p -value) for the effects of treatment (Irrigation vs. Control), season (wet vs. dr y), date, and the in teractions between treatment by season and treatment by date on microbial processes (mineralization, nitrification, acid-phosphatas e activity and basal respirat ion). = P < 0.05; ** = P < 0.01; *** = p < 0.001. Irrigation vs. Control Variable Season Date Treat Treat x SeasonTreat x Date Mineralization0.27 16.25***0 0.17 8.4*** Nitrification 0.16 28.87***1.35 1.07 15.59*** Phosphatase 0.35 10.32***6.03**2.22 21.25*** Basal Resp 335.98***45.10***2.61 112.48*** 45.10*** Table 3-2. Least square mean values for signi ficant season, treatment, and treatment by season contrasts associated with the irrigation e xperiment. (Lower and upper bounds of the 95% confidence interval are provided in parenthesis). Lower case letters indicate differences at P < 0.05 between treatment s both annually and within each season. Upper case letters indicate significant seasonal differen ces at P < 0.05 within each treatment. Phosphatase (gPNP/g-1soil/h-1) Basal Respiration (gCO2/g-1soil) Treatment LSM Irrigation 2425a (2318-2532) 156 (151-160) Control 2236b (2129-2343) 161 (156-166) Treatment by Season LSM Dry Season Irrigation 2421 (2243-2599) 128A (123-134) Control 2180 (2002-2358) 124A (118-129) Wet Season Irrigation 2427 (2302-2551) 194B (185-203) Control 2261 (2137-2386) 189B (181-197)

PAGE 65

65 Table 3-3. F -statistics and associate significant levels ( p -value) for the effects of treatment (Litter Removal vs. Control), season (wet vs dry), date, and the interactions between treatment by season or treatment by date on microbial processes (mineralization and nitrification rates, acid-phos phatase activity and basal re spiration). = P < 0.05; ** = P < 0.01; *** = p < 0.001 Litter Removal vs. Control Variable Season Date Treat Treat x Season Treat x Date Mineralization 0.81 16.04 68.04***23.84** 11.35*** Nitrification 0.05 21.57 0.22 0.54 11.23*** Phosphatase 0.30 15.02*** 63.7*** 22.11** 171.13*** Basal Resp 305.20*** 147.25***17.08** 107.58*** 124.10*** Table 3-4. Least square mean values for signi ficant season, treatment, and treatment by season contrasts associated with the litter rem oval experiment. (Lower and upper bounds of the 95% confidence interval are provided in parenthesis). Lower case letters indicate differences at P < 0.05 between treatment s both annually and within each season. Upper case letters indicate significant seasonal differen ces at P < 0.05 within each treatment. Mineralization (gN/g-1soil/d-1) Phosphatase (gPNP/g-1soil/h-1) Basal Respiration (gCO2/g-1soil) Treatment LSM Litter Removal 4.92a (4.37, 5.47) 1614a (1507, 1721) 142a (138, 147) Control 8.21b (7.65, 8.76) 2236b (2129, 2343) 156b (151, 161) Treatment by Season LSM Dry Season Litter Removal 5.09a (4.38, 5.80) 1632a (1454, 1810) 116A (111, 121) Control 8.42b (7.68, 9.15) 2180b (2056, 2304) 124A (118, 129) Wet Season Litter Removal 4.79a (4.01, 5.56) 1606a (1482, 1730) 169aB (162, 176) Control 8.05b (7.28, 8.81) 2261b (2137, 2386) 189bB (181, 197)

PAGE 66

66Table 3-5. Pearson correlation analysis for each treatment, Contro l, Irrigation or Litter removal between variables reported i n this chapter: N-mineralization rates (Min), ni trification rates (Nit.), phosphatase activit y (Phosp.) and basal respiration (Resp.). = P < 0.05; ** = P < 0.01; *** = p < 0.001. Control Min. Nit. Phosp. B. resp Mineralization -0.26* 0.56* ns Nitrification ns ns Phosphatase ns B. respiration Irrigation Mineralization -0.26* 0.39** 0.49*** Nitrification ns ns Phosphatase ns B. respiration Litter removal Mineralization -0.30** 0.42** 0.35** Nitrification ns ns Phosphatase ns B. respiration

PAGE 67

67Table 3-6. Pearson correlation analysis for each treatment (Contro l, Irrigation or Litter removal), between variables reported in this chapter and variables reported in other chapters: soil C:N ra tio, microbial biomass carbon, nitrogen and phosphorus (MBC, N and P), and their ratios (MBC:N, C:P, and N:P), fungi (F ), bacteria (B) and their ratio (B:F), ammonium (NH4 +), nitrate (NO3 -) and phosphorus (PO4 3-), rainfall and soil water potential (S WP).* = P < 0.05; ** = P < 0.01; *** = p < 0.001. Control Irrigation Litter Removal Min Nit Phosp. Resp. Min Nit Phosp. Resp. Min NitPhosp. Resp. MBC -0.34** ns -0.26* -0.37**ns ns ns -0.44**ns ns ns -0.34* MBN ns ns 0.44*** ns ns ns 0.32** ns ns ns 0.58***-0.30* MBP ns ns 0.35** ns ns ns 0.27* ns ns ns 0.35** 0.30* MBC:N ns ns -0.45***ns ns ns -0.35** ns ns ns -0.37** ns MBN:P ns ns ns ns ns ns ns ns ns ns ns -0.32* MBC:P -0.41*** ns ns -0.36* ns ns -0.29* ns ns ns ns -0.34* B 0.42* ns ns ns 0.57***ns ns ns 0.44**ns ns ns F 0.54*** ns ns ns 0.47** 0.48** ns ns 0.52**ns 0.34* ns B:F ns ns ns ns ns -0.51***ns ns ns ns ns ns NH4 + 0.47*** ns 0.36** ns ns ns 0.47*** ns ns ns 0.35** -0.42** NO3 ns -0.29* ns ns ns ns ns ns ns ns ns ns PO4 3ns ns -0.32* ns ns ns ns ns 0.40**ns ns -0.34* Soil C:N ns ns ns ns ns ns ns ns ns ns ns ns SWP ns ns ns 0.33* -0.31* ns 0.29* ns ns ns ns 0.39** Rainfall ns 0.25* ns ns ns ns ns ns ns ns ns 0.44**

PAGE 68

68Table 3-7. The effects of seasonality (D ry-D vs. Wet-W season), litte r removal (LR), and irrigation (IRR) on N-mineralization (Net Min.), nitrification (Net Nit.), phosphatase activity (Phosp.), basal respirati on (Basal Resp.), substr ate induced respiration (SIR), and soil CO2 efflux across studies. Seasonal Effect Treatment Effect Study Location Control PlotsIrrig ationLitter Removal Source Net Min. BCI DW D>W N.D. ns Yavitt et al. (2004) BCI N.D. N.D. N.D. LRW D>W N.D. ns Yavitt et al. (2004) Eastern Amazonia ns ns IRR>ControlThis study Eastern Amazonia ns ns LRW IRR>C ontrolVasconcelos et al. (2004) D
PAGE 69

69 Figure 3-1. Effects of rainfall patterns and dr y-season irrigation on microbi al-mediate processes: (A) daily rainfall at the study site, (B ) soil water potential (SWP), (C) Nmineralization, (D) N-nitrification a nd (E) Acid-phosphatase, and (F) Basal respiration. In B-F, solid and open circles represent means ( se) for control and irrigation treatments, respectively (N = 4 for soil water potential, N=4 to microbial processes). Vertical dashed lines indicat e the dry season irrigation period (Sept 23rd 2004 to Jan 26th 2005). White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (May to Sept 20th 2004, or Jan 20th to Aug 5th 2005), respectively. ANOVA and treatment c ontrasts with fixed effects by each collection date (* P < 0.05; ** P > 0.01; *** P <0.001).

PAGE 70

70 Rainfall (mm) 0 20 40 60 80 100 120 140 SWP (MPa) -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 Net mineralization (N/g -1 dry soil d -1 ) 2 4 6 8 10 12 14 Net nitrification (N/g -1 dry soil d -1 ) -0.2 0.0 0.2 0.4 Phosphatase activity (g PNP.g dry soil -1 h -1 ) 1000 1500 2000 2500 3000 3500 Time (months) May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Basal respiration (g CO 2 /g -1 dry soil) 0 50 100 150 200 A B C D E F* *

PAGE 71

71 Figure 3-2. Effects of rainfall patterns and lit ter removal on microbial-mediate processes: (A) daily rainfall at the study site, (B) soil wa ter potential (SWP), (C ) N-mineralization, (D) N-nitrification and (E) Acid-phosphatase and (F) Basal respir ation. In B-F, solid and open circles represent means ( se) for control and litter removal treatments, respectively ( N = 4 for soil water potential, N = 4 to microbial processes). White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (May to Sept 20th 2004, or Jan 20th to Aug 5th 2005), respectively. ANOVA and treatment contrasts with fixed effects by each collection date (* P < 0.05; ** P > 0.01; *** P <0.001).

PAGE 72

72 Rainfall (mm) 0 20 40 60 80 100 120 140 SWP (MPa) -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 Net mineralization (N/g -1 dry soil d -1 ) 2 4 6 8 10 12 14 Net nitrification (N/g -1 dry soil d -1 ) -0.2 0.0 0.2 0.4 1000 1500 2000 2500 3000 3500 Time (months) May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun 0 50 100 150 200 * * * * * * ** Phosphatase activity (g PNP.g dry soil -1 h -1 ) Basal respiration (g CO 2 /g -1 dry soil)A B C D E F

PAGE 73

73 CHAPTER 4 IRRIGATION AND LITTER REMOVAL EFFECT S ON SOIL NUTRIENT AVAILABILITY IN A SEASONAL TROPICAL SECONDARY FOREST IN THE EASTERN AMAZON Introduction Temporal variability of soil nut rient concentrations in tropic al forests is poorly understood, although it has been directly linke d to microbial activity and ecosystem productivity (Yavitt and Wright 1996; Cleveland et al., 2002; Ruan et al., 2004). Pronounced dry seasons produce reductions in organic matter decomposition, reduced plant uptake of soil nutrients, and increased soil nutrient pools (Luizao and Sc hubart 1987; Singh et al., 1989; Ya vitt et al., 1993; Yavitt and Wright 1996). Nutrient and moisture availability are often associated w ith litterfall dynamics (Vitousek 1984), but the direct effects of litter and water on soil nutrient availability are not consistent across studies. Pulses of nutrient mineralization and immobilization fluctuations in microbial populations have a dir ect impact on nutrients in soluti on. Lucas et al. (1993) suggested that soil moisture controls a delicate balan ce between the processes of immobilization and mineralization after observing that rewetting of seasonally dry Am azonian soils resulted in net immobilization, whereas dry periods allowed mine ral-N to accumulate. Maximum root growth may actually occur in the transition from wet to dr y and from dry to wet seasons as a response to water and/or to nutrient pulses (Cavelier et al., 1999) In tropical dry forests, drying and rewetting causes crashes in microbial populations and induces pulses of nutrient release from epiphytes and dead microbial biomass (Lodge et al. 1994). The mechanism apparently relates to the wetting of dry soil that disrupts the osmotic balance of soil microorganisms, causing nutrient release to the soil, with nutrients accumulating in plant bi omass during the wet season (Lodge et al., 1994; Yavitt and Wright 1996). A rich and informative literature on the effect s of continuous irrigation on an old-growth tropical-moist forest (a well-dr ained Alfisol) resulted from a five-year study on Barro Colorado

PAGE 74

74 Island (BCI), Panama. Within that study, irrigation had little effect on concentrations of inorganic N (Yavitt et al., 1993, Yavitt and Wright 1996). Irrigati on altered microbial composition by decreasing fungal densities, and bacter ial densities af ter five months of exposure (Cornejo et al., 1994), enhanced de composition rates of the forest floor, and reduced forest floor mass throughout the year (Wieder and Wright 1995), but did not affect nutri ent concentrations in leaf-fall or nutrient return from forest trees to the forest floor (Yavitt et al., 2004). Sayer (2005) separately examined the effects of litter manipulations on BCI a nd found that two years of litter removal had no effects on nutrient concentrations in the mineral soil, except to increase NO3 availability in litter addition plots, Soil CO2 efflux decreased by 27% in litter removal plots, accompanied by a significant decline in total micr obial biomass, but increased approximately 25% in litter addition plots after accounting fo r changes in root biomass (Sayer 2005). Litter removal reduced the abundance of meso-arthropods in Simaroub litter, slowed leaf-litter decomposition, and significantly reduced the concentration of N and P in Cecropia litter (Sayer et al., 2006). Correspondingly, li tter addition accelerated the decay of wood and increased nutrient concentrations of Cecropia litter (N, P and K), but ha d no effects on leaf litter decomposition and meso-arthropod abundance (Sayer et al., 2006). Whether the results from BCI are indicative of generalized responses of tropical forests to altered moisture and substrate availability rema ins unclear. There have been two other litterremoval studies in a tropical wet forest in th e Luquillo Experimental Forest in north-eastern Puerto Rico. The first study consisted of para llel measurements in a pine plantation and a secondary forest between 1996 and 1997 under three tr eatments (root exclus ion, litter exclusion, and root-and-litter exclusion) that were initiated in 1990 (Li et al., 2004 and 2005). Their results showed that litter removal significantly reduced soil respiration and microbial biomass in both

PAGE 75

75 the pine plantation and the secondary forest, but litter removal had a gr eater effect on soil CO2 efflux than root exclusion in the secondary forest. The second study found that microbial biomass was highly correlated with abovegr ound litter inputs from the preceding month, suggesting that enhanced root exudates prior to senescence, could have influenced microbial biomass abundance (Ruan et al., 2004). The main objectives in the present study were to (1) determine the responses of NH4 +, NO3 -, and PO4 3availability, and soil C:N to seasonal ch anges and wet-up events in a tropical secondary forest in the Eastern Amazon, and (2) to examine substrate and water constraints to the availability of those nutri ent species within two ongoing manipulative experiments designed to alter resource availability in that forest -dry-season irrigation and litterfall removal. The site selected for this work contrasts with the BCI site in that it is (1) secondary v. old-growth forest; (2) seasonally-dry v. continuously moist; and (3) underlain by a shal low, coarse, and relatively infertile soil v. a the deep, fine-textured, relativel y fertile soil present at the BCI site. These contrasts allow us to draw inferences about both the similarities and differences in the responses of soil nutrient availability to resource manipulations at the two sites. Study Site and Experimental Design Study site and experimental desi gn are described in Chapter 1. Materials and Methods Ion Exchange Resins The Dowex 50W-X8 cation exch ange resins, 50-100 mesh, H+ form and 1.9 mmolc cm-3 (Sigma Aldrich Family, Catalog # 217492, Milwal kee, WI) was used to measure ammonium availability (NH4 +), in the mineral soil (5 cm depth) of all 12 plots. The resin-bag technique consisti ng of a specified quantity of loose resin beads placed onto a piece of porous fabric (mesh) and sealed into a convenient shape or size (Skogley and

PAGE 76

76 Dobermann 1996) was deployed in the field to quantify NH4 + availability. The 25 cm2 resin bag containing 2.2g 0.02 of dry weight resins wa s hand-made of non-adsorbent fabric (07-40/25 Pecap, Sefar America Inc. Filtration Division, www.sefaramerica.com ) re-used after extraction procedures for 2 harvest cy cles and then discarded. Resin bags were recharged pr ior to deployment in the fiel d by placing them in Nalgene bottles with 100 ml 0.3M HCl, shaking for 1hr on a platform shaker, and then rinsing 3 times with deionized water. Research ers that have previously us ed Dowex resins (25-50 mesh) recharged, eluted and sometimes re-used them by sh aking each bag in a platform shaker for one hour in 100 ml of 2 M NaCl in 0.1 M HCl (Giblin et al., 1994) or soaking for at least 18 h in 100 ml 1.2 M HCl (Yavitt et.al., 1996), then rinsing with de-ionized water before use. For Dowex resins (50-100 mesh), resin ex traction efficiency was greatest using 100 ml of 2 M NaCl in 0.3 M HCl (unpublished results). During the first month of installation, 5 resin bags were installed randomly per experimental plot, for a total of 20 units per treat ment. A small slit was carefully lifted from the top 5cm of the mineral soil using a flat, sharp an d clean plastic spatula. Each bag location was identified with a colored flag, so that bags would be replaced at the same location in the following months. After each monthly harvest of re sin bags installed in the field, resins were transported to the laboratory facility in i ndividual Nalgene bottles containing only de-ionized water. Then, after excess organic matter and resi duals were rinsed off the bags with de-ionized water, 100 ml of 2 M NaCl in 0.3 M HCl was used for each bag to extract NH4 + ions. Bags were shaken on a platform shaker for 1.5 h, removed, and an aliquot reserved for colorimetric analysis. Ammonium concentrations were determined colorimetrically by the salicylate-

PAGE 77

77 nitroprusside method (Mulvaney 1996) on duplicate samples. An average of the 2 samples was used for statistical anal ysis; duplicates mean differed by < 8% ( 0.4). Anion Exchange Membrane We used BDH Anion exchange membranes (G allard-Schlesinger Ind., Product # 55164 2S, Plainview, NY, 1-888-686-3454) to measure nitrate (NO3 -), and phosphorus availability (PO4 -) in the mineral soil (5 cm depth) of all plots. This membrane is supp lied in the chloride form, with exchange capacity estimated at 0.2 mmol P g-1 (Turrion et al 1997). Testing and deployment of anionic membranes for quantification of nitrate and phosphate followed the method described by Turrion et al. (1997), modified to use 16cm2 membrane strips during in situ incubations. They suggested that the resin bag technique have several disadvantages over anion membranes. Resin bags ma y tear and wear out, loose resin beads, trap fine roots, fungi, and soil particles that may in terfere with analysis, and add diffusion problems due to their three-dimensional spherical st ructure, while membranes flatness impose no diffusion problems and have greater surface area that improves contact with soil surface. Before use, chloride-saturated anion-ex change membranes were converted to the bicarbonate form set at 8.5 pH. St rips were shaken for 1 hour for each of three successive washes in 50 ml 0.5M NaHCO3 /per strip (sodium bicarbonate), and then rinsed the strips three times in deionized water. Strips were ta ken to the field in Nalgene bottles filled with deionized water. During the first month of installation, 4 membranes were installed randomly per experimental plot, for a total of 16 units per treatment. Each strip was labeled by placing a nylon tread through a hole in one corner of the strip. A small slit was carefully lifted from the top 5cm of the mineral soil using a flat sharp and clean plastic spatula. Each membrane location was labeled with a colored flag, so repeated coll ections could be carried out during the following months. After each monthly harv est of membrane strips insta lled in the field, strips were

PAGE 78

78 transported to the laboratory facility in i ndividual Nalgene bottles containing only deionized water. Then, after excess organic matter and resi duals were rinsed off the bags with deionized water, strings were cut and 32 ml of 0.3 M HCl (hydrochloric aci d) was used to extract the anions of interest (NO3 and PO4 -). Strips were shaken on a plat form shaker for 2 h, removed, and an aliquot reserved for colorimetric anal ysis. Nitrate was analyzed using a simple spectrophotometric method to detect nitrate in wa ter, resin or soil extrac ts as described by Yang et al (1998), modified by the addition of 0.2ml 3 M KOH for each 2ml of aliquot (matrix 0.3M HCl) to neutralize Clinterference. Phosphorus determin ation was done using the standard Murphy and Riley procedure (1962). Samples were analyzed in duplicates, and their average was used for statistical analysis; on average, duplicates differed by <10% ( 0.5). Soil C:N Ratios Soil sampling and processing: In each plot, seven soil co res were taken using a bulkdensity corer with 6 cm diameter and 5 cm dept h. These samples were composite per plot, sifted through a 2 mm mesh, placed in doubl e-folded, tightly closed plasti c bags, transported to the lab in a cooler no later than 4 hours after harvest, and stored at 4oC until analyzed. Subsamples taken to run total soil carbon were drie d at 40C and total soil nitrogen was left air-drying prior to analysis. Determination of soil organic carbon (SOC) : I used the Nelson-Somers method (1975) revised in Forster (1995). This is a routine carbon analysis generally recommended for mineral soils. Paired soil samples were digested using wet potassium dichromate, followed by titrimetric measurement of unreacted dichromate.

PAGE 79

79 Determination of Total N : I used the Keeney and Nelsons method (1982) revised in Forster (1995). Paired soil samples were digested using sulfuric acid (t he Kjedahl procedure) followed by titrimetric analysis. Soil Water Potential (SWP) We used a four-channel datal ogger with four gypsum resistan ce block sensors installed in each plot at 10cm depth (Model 220, Spectru m Technologies, Inc) between 12 May and 17 December 2004. Measurements were taken hourly a nd then averaged per day and downloaded to our database monthly. Due to technical problems with the dataloggers, between 18 of December 2004 and 30 of June 2005, SWP was recorded on a weekly basis in the morning using one tensiometer installed per plot at 10cm depth (M odel 2710 ARL, Forestry Supplier, Inc. Catalog 53, Jackson, Mississippi 39284-8397). Re sistance block sensors have close correlations with tensiometer readings of soil water potential (SWP) depending on soil type (Shock et al 2001). For example, in an irrigation field, silt loam soil (Malheur Experiment Station, Oregan Sate University), the resistance block closely tracked those obtained with a tensiometer (r2= 0.83). At our study site, r2 values were 0.80, 0.81, and 0.60) for control, litter-removal, and irrigation plots, respectively. Statistical Analysis The SAS System for Windows V8 (2) was used for statistical analyses. The response variables of interest NH+ 4, NO3 and PO4 were log-transformed to meet the model assumptions of normality. PROC MIXED was used using a repeated measures analysis with a heterogeneous-autoregressive error structure arh (1)]. This structure allowed modeling within sample correlation over time and calculation of i ndividual error variances for each sampling date. Linear models were fitted on the variables NH+ 4, NO3 and PO4 and soil C:N with the following effects: season, date, treatment, treatment by seas on, treatment by date, plot and plot by date. All

PAGE 80

80 effects were considered fixed with the exception of plot and plot by date for NH+ 4, NO3 and PO4, thereby allowing assessment of correlation between samples in the same plot on any date. CONTRAST statements were used to determine the significance of each fixed effect for each pair of treatment comparisons (i .e., control vs. irrigation and cont rol vs. litter removal) and leastsquares means were used to compare treatments and control means for the effects of season, treatment and treatment by season interaction on NH+ 4, NO3 and PO4 availability, and soil C:N ratio. In order to correct for multiple testing the Bonferroni correction was used for each response variable considering an experiment -wise significance level of 5%. Within each treatment (control, irrigation, litt er removal), Pearson correlation analyses we re used to explore the bivariate relationships of re sults reported in previous chap ters to those reported here. Specifically, I tested for correlations for NH+ 4, NO3 and PO4 and soil C:N with microbial biomass C, N, and P and their ratios, fungal and bacterial densities (Cha pter 2); mineralization rates, nitrification rates, phospha tase activity, basal resp iration (Chapter 3), as well as rainfall and soil water potential. Results Dry-season irrigation reduced NH4 + (Table 4-1, Figure 4-1c) and litter rem oval reduced NH4 + and PO4 3availability, while enhancing NO3 availability (Table 4-5, Figure 4-2). There were no seasonal or treatment effects on soil C: N ratios (Table 4-1). Correlations between NH4 +, NO3 -, and PO4 3availability and measurements reported in previous chapters were strongly influenced by treatment. Seasonal Effects The only seasonal effect was in creased phosphorus availability in the dry than wet season in control plots (Tables 4-1, 4-2, 4-3 or 4-4). There was also intra-a nnual variation observed

PAGE 81

81 within treatment for PO4 3availability in irrigated plots (Table 4-2), and for NH4 + availability in litter removal plots (Table 4-3). Irrigation Effects The dry-season extended from mid-July 2004 to early-January2005 (Figure 4-1a), resulting in lower water potential during th is period (Figure 4-1b). November was the driest month of the year, when there was only 8 mm of rainfall in the 27 d prior to our next harvest. Mean soil water potential for the dry season was -0.046 0.001 a nd -0.018 0.001MPa (mean SE), for control and irrigation plots with corre sponding gravimetric soil wate r content of 0.67 0.07 and 0.83 0.10 ( SE), respectively. In the wet season of 2004, mean soil water potential was -0.016 0.001 and 0.011 0.001 MPa, for control and irri gation plots with corresponding gravimetric soil water content of 0.81 0.05 and 0.65 0.10, re spectively. In the we t season of 2005, mean soil water potential was -0.010 0.005 and 0.006 0.003 MPa, for control and irrigation plots with corresponding gravimetri c soil water content of 1.38 0.04 and 1.64 0.05, respectively. Ammonium availability declined due to irrigation (Figure 4-1c ). Toward the end of the dry season, there were two wet-up events that significantly increased th e availability of NH4 + in control plots with little effect on irrigation plot values. The fi rst pulse occurred after a 52 mm rainfall event on 14 December 2004 (two days prio r to our sampling). This event represented double the amount of precipitation for the month prio r to this date (25 mm from 14 November to 13 December 2004). On the 11 January 2005 there was a second pulse that followed smaller rain events on 10 January and 11 January 2005 (13 mm and 17 mm), resp ectively (Figure 1a). These events are responsible for the significant treatme nt by season interactions shown in Table 4-1. Overall, the dampening effect of dry-season irrigation on NH4 + availability is apparent in both the dry and the wet season (Table 4-2).

PAGE 82

82 The two wet up events toward the end of th e dry season also substantially increased NO3 availability in both contro l and irrigation plots. NO3 availability was not significantly affected by the irrigation treatment for any single sampling date (Figure 1d). There were significant effects of date and treatment-by-date interactions (Table 4-1), but overall there was no significant effect of irrigation on NO3 availability during either season, a lthough annual and seasonal mean values trended lower for the irrigation plots (Table 4-2). Phosphate av ailability in both control and irrigation plots was also increase d by the two wet up events, and the increase was significantly greater for the irrigation treatment (Figure 1e). There were significant main effects of date and season, and both treatment-by date and treatmentby-season interactions on PO4 3availability (Table 4-1). Overall, there was no significant effect of irrigation on PO4 3availability during either season, but values were significantly higher in the dry season in both irrigation and control plots (Table 4-2). There were si gnificant effects of date and treat ment by date interaction on soil C:N ratios (Table 4-1). Litter Removal Effects Soil water potential varied s easonally and in response to precipitation during the dry season, but was unaffected by litter removal (Figur e 4-2a, and b). Ammonium availability was significantly lower in litter removal than control pl ots in 9 out of 15 sampli ng dates, 8 in the wet season (Figure 4-2c). These resu lts are responsible for the significant main effects of date and season, and the treatment-by-date and treatment-byseason interactions show n in Table 4-4. The two wet-up events in the late dry season increased NH4 + in the litter removal plots, but the effect was smaller than in the control plots (Figure 4-2c). Overall NH4 + availability was significantly reduced by litter removal, and that overall effect was driven by the significant treatment effect in the wet season (Table 4-5).

PAGE 83

83 Nitrate availability was significantly higher in li tter removal than cont rol plots in 6 out of 15 sampling dates, 5 in the wet season (Figure 4-2c). These results are responsible for the significant main effects of treatment and date, an d the treatment-by-date interactions shown in Table 4-4. The one significantl y different dry-season sampling o ccurred at the second wet-up event toward the end of dry season, which increased NO3 availability in the litter removal plots to a substantially greater extent than that exhi bited in the control plot s (Figure 4-2d). Overall, NO3 availability was significantly increased by litter removal and th at overall effect was driven by the significant treatment effect in the wet season (Table 4-5). Phosphate availability was significantly higher in litter removal than control plots in 4 out of 15 sampling dates (Figure 4-2e ). All of the main and interaction effects tested were significant (Table 4-4). The first wet-up event to ward the end of the dry season significantly increased phosphate availability in the litter removal plots but the second wet up had the opposite effect (Figure 4-2e). Overall PO4 3availability was slightly but significantly reduced by litter removal, and the effect was consistent in bot h dry and wet seasons (Table 4-5). There were significant effects of date and treatment by date interaction on soil C:N ratios (Table 4-3). Correlation Analyses The only significant correlations among variab les measured in this study were between NH4 + and PO4 3availability in control plots, and between NO3 and PO4 3availability in irrigated plots (Table 4-5). Control plot NH4 + availability was positively correlated with bacterial, fungal densities and N-mineralization, and negatively co rrelated nitrification rates (T able 4-6). Under irrigation, NH4 + availability was positively correlated with phosph atase activity and microbial biomass N:P, and negatively correlated with microbial biomass C and C:N (Table 4-6) Contrary to the results in irrigation plots, NH4 + was positively correlated with microbial biomass C in litte r removal plots,

PAGE 84

84 but positively correlated with phosphatase activity. Additional results showed a negative correlation between NH4 + and basal respiration, soil water potential and biomass C:N (Table 46C). NO3 availability was negatively correlated with nitrification rates in control plots only (Table 4-6A), with microbial biomass N:P in ir rigated plots (Table 4-6B ), and with soil water potential in litter remova l plots (Table 4-6C). The positive correlation between NO3 and microbial phosphors was consis tent in control and irriga ted plots; and between NO3 and microbial C:N was consistent acro ss treatments. Under irrigation, NO3 availability was also positively correlated with rainfall. PO4 3availability was negatively correlated with phosphatase activity in control plots only (Table 4-6A), and with soil water potential in irrigated plots (Table 4-6B). Under irrigation plots, PO4 3availability was also positively correlated w ith bacterial densities. Likewise in irrigated plots, PO4 3availability was negatively correlated with so il water potential in litter removal plots, and with microbial basal respiration (T able 4-6C). Under litter removal, PO4 3availability was also positively correlated with microbial bi omass C, C:N, C:P and with N-mineralization. Soil C:N ratios appeared to be significantly affected by the effects of litter removal only. Under litter removal, soil C:N was positively co rrelated with microbial biomass C:N, and negatively correlated with soil water potential, mi crobial biomass N, and N:P (Table 4-6C). Discussion This study showed that dry season wet-ups were a principal source of temporal variability in nutrient status resul ting in higher nutrient availability in this period. The observed peaks in NH4 +, NO3 and PO4 3availability in the dry season as a re sult of wet up events is typical of dry tropical forests with monsoonal climate, when pool sizes of these nutri ents tend to increase towards the dry season, as plants senesce, then decrease during the wet growing-season (Singh et

PAGE 85

85 al., 1989, Lodge et al., 1994). In this study, the four peaks in ammonium throughout the year accounted for half of the annual NH4 + availability, thus wet up events appear to greatly contribute to the total annual N-pool in this tropic al seasonal forest, perhaps due to a positive net mineralization response to wet-up (Table 4-6). The positive correlation between NH4 + availability with fungal and bact erial densities, and N-mineraliza tion in control plots could be a result of their intrinsic relationship, observed by concomitant increases of these variables during wet-up events. The reduction in NH4 + availability in irrigated plots c ould indicate that continuous water availability resulted in higher nutrient immob ilization (lower mineralization), but additional results showed no effects of treatment on minerali zation and nitrification ra tes (Chapter 3), even though leaf decomposition was 2.4 times faster in irri gated than control plots (Vasconcelos et al., 2006). Lodge et al. (1994) suggested that environments with da mpened microclimate fluctuations may enhance microbi al N-immobilization, increase th e competition for N, and even affect primary productivity if microbial N-mi neralization does not synchronize with plant Nuptake. Nonetheless, under irrigated conditions the association between microbial biomass nutrient concentrations with NH4 + availability suggested that cont inuous water availability have increased the potential for microbial NH4 + uptake, also indicated by the negative correlation between NO3 and microbial N:P ratios (Table 4-6). The additional positive correlation between NO3 availability with microbial biomass P and rainfall was probably driven by concomitant responses of these variables to wet-up events. Ch anges in the ratio of nut rient concentrations of the substrate in relation to that of the decom posing microbial biomass ma y also trigger changes in the ratios of the la tter (Hodge et al., 2000).

PAGE 86

86 Reduced NH4 + and P availability as a result of lit ter removal suggest that recycling of nutrients in litter is an important pathway for nutrient availability as this secondary forest is nutrient poor and a large proportion of potentially available nutrient s are retained in the living biomass and recycled in the litter (Singh et al., 1 989). Some studies have suggested that prior to leaf senescence, some tropical forests experience nutrient translocation fro m shoot to roots that may result in increased root-exudates and nutrien t availability (Ruan et al., 2004), which may explain the high peak in NH4 + and NO3 availability prior to the onset of the dry-season when soil water potential was very high (0.06 MPa, Figure 4-1b). Yavitt a nd Wright (1996) report that fluctuations in soil nutrient availability impos ed by the timing of leaf litterfall and nutrient leaching from forest floor litter (also measur ed with ion-exchange resins), essentially disappeared with depth in the mineral soil. Soil NH4 + availability was likely reduced in litter removal plots because long-term effects of four years of litter removal consistently reduced litterfall N (Vasconcelos et al., 2006), confirming that litterfall is a significant source of N for tropical forest pl ants. The next plausible explanation can be drawn by the positive feedback that increased water availability had on microbial basal respiration (Chapter 3, Table 3-6C), which led to more NH4 + uptake, and increased microbial biomass C in litter removal plots (Table 4-6C). Base d on these correlations, NH4 + availability increases as the soil dries because basal respiration also decreases, and there is less NH4 + uptake. The same assumption can be applie d to the availability of phosphorus, which followed a similar pattern as NH4 +, and covaried with it. The reduction of NH4 + and Pavailability in litter removal plot s could also be attributed to re moval of the microflora biomass (mostly fungi and bacteria) inhabiting the litt er that contributes to immobilization and mineralization of these nutrients in the interface between the minera l soil and the litter layer, as

PAGE 87

87 suggested by Sayer (2005) in a similar study on the BCI. In fact, this study has shown that bacterial and fungal densities si gnificantly decreased in the mineral soil in litter removal compared to control plots (Chapter 2). There were however, no significant reductio ns on litter P conc entrations although Vasconcelos et al. (2006) charac terized this site as having lo w litterfall P concentration in accordance to values proposed by Vitousek (1986). They also suggested that this site may have sufficient P-supply through soil organic matter mineralization based on recent studies that showed substantial amounts of labile P-fractions for secondary forests in the Amazon (Frizano et al., 2003, Markewitz et al., 2004). Alternatively, Vascon celos et al. (2006) sugge sted that litter P concentrations were not affected by litter removal likely due to enhanced P-acquisition through mycorrhizal associations and high phosphatase exudation rates as pr eviously observed in secondary forests with low soil P (Marschnner 1995) Results from this study showed that litter removal in fact enhanced arbuscular mycorrhizal fungi associations (Chapt er 2), but phosphatase activity was lower than in control plots and di d not respond to seasonal changes (Chapter 3). Arbuscular mycorrhizal fungi may enhance the reabsorption of nutrients lost through root exudation, influence biochemical re actions in soil including minera lization of organic matter and nitrification, and improve the cap acity of their host plant to use organic sources of P and N (Hamel 2004). The litter layer may absorb N that would otherw ise be available in the mineral soil, here evident with higher NO3 availability in litter removal than control plots. Results from a recent study in a seasonally dry tropical fo rest in Mexico showed that N-dynamics in the litter layer was influenced by rainfall seasonality and labile C, with lowest C and N immobilization rates in the rainy season (Anaya et al., 2007). Low substrat e availability, could ha ve resulted in less

PAGE 88

88 microbial immobilization of NH4 + allowing higher nitrification rates and therefore, higher NO3 availability, but nitrification ra tes were not significantly affected by litter removal (Chapter 3). Soil microsites may even shift from NH4 + to NO3 dominated depending on demand and supply of these nutrients (Sch imel and Bennett 2004). The positive correlation between NO3 and biomass C:N may further indicate that NO3 is the preferred N-source for microbial organisms in litter removal plots, whereas the contra ry relationship was observed between NH4 + and microbial biomass C:N. The positive correlation between n itrate and bacterial dens ities in this study may be related to nitrifiers, bacteria that oxidize a mmonium to nitrite and th en, nitrate (Robertson et al., 1999). Higher NO3 availability for litter removal plots could also indicate that litter is intercepting NO3 in throughfall, as previously observed in a Northwestern Amazonian forest (Tobon et al., 1994). The fact that NO3 availability was negatively correlated with soil water potential even though it responded to wet up events contradicts nitrates us ual higher mobility in water, and discard its potential relation to throughfall in this study site. Another study using the same ionic membrane showed that NO3 -diffusion to the membrane was impaired as the soil dried (Turrion et al., 1997). The linkage between soil C:N a nd microbial biomass nutrients in litter removal plots, may suggest that the organic soil carbon and nitrog en pools has become an active reservoir of nutrients for microbial uptake. However, further studies would be necessary to explore these findings, and perhaps investigat e how long soil carbon and nitrogen pools would sustain forest and microbial activity under c ontinuously litter removal. In contrast with the results obtained in the BCI experiments (Yavitt et al., 1993; Yavitt and Wright 1996, Yavitt et al., 2004), I found that dry-s eason irrigation reduced the availability of

PAGE 89

89 NH4 + and that litter removal reduced NH4 + and PO4 3availability, while enhancing NO3 availability. The results by Sayer (2006), showed that litter removal lowered the concentration of N and P in Cecropia litter, but this results was not followe d by reductions in nut rient availability in the mineral soil (NH4 + NO3 -, and PO4 3), except to an increase in NO3 in litter-addition plots. This study also showed that other nutrient pools may temporary count eract the nutrient deficit imposed by litter removal, such as the contribution of nutrients stored on microbial biomass, derived from soil C:N pools or compoun ds from root and mycorrhizal exudates, but further studies would be nece ssary to identify the magnit ude of these contributions. Thus, nutrient availability could ultimate ly be affected by changes in microbial composition, structure and activity as shown by their interactive effects in this seasonal tropical forest. On individual collection dates in this study, di fferences between treated and untreated plots varied from non-significant to upwards of four-f old, suggesting that comp lex interactions shape the responsiveness of nutrient dynamics to changes in resource availability. Frequent sampling is needed to adequately capture intra-annual variability in soil nutrient availability.

PAGE 90

90 Table 4-1. F-statistics and asso ciated significance levels (p-value ) for the effects of treatment (Irrigation vs. control), season (wet vs. dr y), date, and the in teractions between treatment by season and treatment by date on soil solution nutrients (NH+ 4, NO3 and PO4), and soil C:N. = P < 0.05; ** = P < 0.01; *** = p < 0.001. Table 4-2. Least square mean values for significant season, treatment, and treatment by season contrasts associated with the irrigation e xperiment. (Lower and upper bounds of the 95% confidence interval are provided in parenthesis). Lower case letters indicate differences at P < 0.05 between treatment s both annually and within each season. Upper case letters indicate significant seasonal differen ces at P < 0.05 within each treatment. NH4 + (gN/ml/bag) NO3 (gN/ml/strip) PO4 (gP/ml/strip) Treatment LSM Irrigation 3.36a (2.83-3.98) 0.30a (0.16-0.55) 0.07a (0.06-0.08) Control 4.65b (3.92-5.52) 0.45 a (0.25-0.82) 0.07a (0.06-0.08) Treatment by Season LSM Dry Season Irrigation 3.32 a (2.59-4.25) 0.33 (0.17-0.62) 0.10A (0.08-0.12) Control 5.12b (3.98-6.59) 0.43 (0.22-0.81) 0.10A (0.08-0.12) Wet Season Irrigation 3.38a (2.85-4.00) 0.29 (0.16-0.54) 0.06B (0.06-0.08) Control 4.45b (3.76-5.27) 0.47 (0.25-0.87) 0.06B (0.05-0.07) Irrigation vs. Control Variable Season Date Treat Treat x SeasonTreat x Date NH+ 4 0.640 15.33***6.97** 2.58* 12.60*** NO3 0.006 33.06***0.84 0.47 16.33*** PO4 25.60*** 10.55***0.34 8.71*** 11.76*** Soil C:N 0.30 5.46*** 0.32 0.87 6.12***

PAGE 91

91 Table 4-3. F-statistics and associ ate significant levels ( P -value) for the effects of treatment (Litter removal vs. control), season (wet vs dry), date, and the interactions between treatment by season or treatment by date on soil solution nutrients (NH+ 4, NO3 and PO4), and soil C:N. = P < 0.05; ** = P < 0.01; *** = p < 0.001. Litter Removal vs. control Variable Season Date Treat Treat x SeasonTreat x Date NH+ 4 23.42*** 9.26*** 22.44***21.29*** 10.85*** NO3 2.10 34.77***4.06 2.21 17.15*** PO4 33.64*** 27.67***20.34***18.07*** 166.56*** Soil C:N 0.01 17.05***0.42 1.37 147.93*** Table 4-4. Least square mean values for signi ficant season, treatment, and treatment by season contrasts associated with the litter rem oval experiment. (Lower and upper bounds of the 95% confidence interval are provided in parenthesis). Lower case letters indicate differences at P < 0.05 between treatment s both annually and within each season. Upper case letters indicate significant seasonal differen ces at P < 0.05 within each treatment. NH4 + (gN/ bag) NO3 (gN/ strip) PO4 (gP/ strip) Treatment LSM Litter removal 2.59a (2.18-3.07) 1.09a (0.59-1.99) 0.05a (0.04-0.05) Control 4.65b (3.92-5.52) 0.45b (0.25-0.82) 0.07b (0.06-0.08) Treatment by Season LSM Dry Season Litter removal 3.95A (3.07-5.07) 0.91 (0.48-1.72) 0.07aA (0.05-0.08) Control 5.12 (3.98-6.59) 0.43 (0.22-0.81) 0.10bA (0.08-0.12) Wet Season Litter removal 2.14aB (1.81-2.53) 1.20a (0.65-2.24) 0.04aB (0.03-0.05) Control 4.45b (3.76-5.27) 0.47b (0.25-0.87) 0.06bB (0.05-0.07)

PAGE 92

92 Table 4-5. Pearson correlation analysis for each tr eatment, Control, Irrigation and Litter removal among variables reported in this chapter: ammonium (NH4 +), nitrate (NO3 -), phosphorus availability (PO4 3-), and soil C:N. = P < 0.05; ** = P < 0.01; *** = p < 0.001. Control NH4 + NO3 PO4 3Soil C:N NH4 + ns 0.34** ns NO3 ns ns PO4 3ns Soil C:N Irrigation NH4 + NO3 PO4 3Soil C:N NH4 + ns ns ns NO3 0.57*** ns PO4 3ns Soil C:N Litter removal NH4 + NO3 PO4 3Soil C:N NH4 + ns ns ns NO3 ns ns PO4 3ns Soil C:N

PAGE 93

93Table 4-6. Pearson correlation analysis for each treatm ent among variables report ed in this chapter (NH4 +, NO3 -, PO4 3-, and soil C:N), and variables reported in ot her chapters: microbial biomass carbon, nitrog en and phosphorus (MBC, N and P), and their ratios (MBC:N, C:P, and N:P), fungi (F), bacteria (B) and thei r ratio (B:F), mineralization (MIN ), nitrification (NIT), basal respiration (B. resp.), soil C:N rati o, rainfall and soil water potential (SWP) .* = P < 0.05; ** = P < 0.01; *** = p < 0.001. Control Irrigation Litter Removal NH4 + NO3 PO4 3NH4 + NO3 PO4 3NH4 + NO3 PO4 3MBC ns ns ns -0.31* ns ns 0.46***ns 0.45** MBN ns ns ns ns ns ns ns ns ns MBP ns 0.29* ns ns 0.29* ns ns ns ns MBC:N ns 0.24* ns -0.55***0.31* ns -0.41 0.33* 0.33* MBN:P ns ns ns 0.49*** -0.39**ns ns ns ns MBC:P ns ns ns ns ns ns ns ns 0.30* B 0.38* ns ns ns ns 0.45** ns 0.45** ns F 0.48** ns ns ns ns ns ns ns ns B:F ns ns ns ns ns ns ns ns ns MIN 0.47***ns ns ns ns ns ns ns 0.40** NIT -0.27* -0.29*ns ns ns ns ns ns ns PNP ns ns -0.32*0.47*** ns ns 0.35** ns ns B. resp. ns ns ns ns ns ns -0.42** ns -0.34* SWP ns ns ns ns ns -0.43** -0.41** -0.47***-0.40** Rain ns ns ns ns 0.38** ns ns ns ns

PAGE 94

94 Table 4-6. Continued Control Irrigation Litter Removal Soil C:N Soil C:N Soil C:N MBC ns ns MBN ns ns -0.40** MBP ns ns MBC:N ns ns 0.46*** MBN:P ns ns -0.50*** MBC:P ns ns B ns ns F ns ns B:F ns ns MIN ns ns NIT ns ns PNP ns ns B. resp. ns ns SWP ns ns 0.27* Rain ns ns ns

PAGE 95

95 Figure 4-1. Effects of rainfall patterns on control ( ) and long-term dry-season irrigation ( ) plots in seasonally dry tropical forest. a) Daily rainfall at the study site, b) Soil water potential (SWP), c) NH+ 4, d) NO3 and e) PO4 availability. In b-e, values are means ( se) for n = 4 plots. White and black horizontal bars represent dry (21st September 2004 to 19 January 2005), and wet seasons (20 May to September 2004, and 20 Jan to 5 August 2005), respectively. Vertical dashed lines indicate the dry season irrigation period 23rd September 2004 to 26 January 2005) ANOVA and treatment contrasts with fixed effects by each collection date (* P < 0.05; ** P = 0.0001; *** P < 0.0001).

PAGE 96

96 Rainfall (mm) 20 40 60 80 100 120 140 SWP (MPa) -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 Control IRR IER-NH + 4 (gN/bag) 0 2 4 6 8 10 12 14 May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun AEM-NO 3 (gN/strip) 0 1 2 3 4 5 Time (months) May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun AEM-PO 4 (gP/strip) 0.0 0.1 0.2 0.3 0.4 0.5 * a c d b e

PAGE 97

97 Figure 4-2. Effects of rainfall patterns on control ( ) and long-term litter removal ( ) plots in seasonally dry tropical forest. a) Daily rain fall at the study site, b) Soil water potential (SWP), c) NH+ 4, d) NO3 and e) PO4 availability. In b-e, values are means ( se) for n= 4 plots. White and black hor izontal bars represent dry (21st September 2004 to 19 January 2005), and wet seasons (20 May to September 2004, and 20 Jan to 5 August 2005), respectively. ANOVA and treatment c ontrasts with fixed effects by each collection date (* P < 0.05; ** P = 0.0001; *** P < 0.0001).

PAGE 98

98 Rainfall (mm) 20 40 60 80 100 120 140 Date SWP (MPa) -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 0 2 4 6 8 10 12 14 Control LR 0 2 4 6 8 10 Time (months) May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun 0.0 0.1 0.2 0.3 0.4 0.5 AEM-PO 4 (gP/strip) AEM-NO 3 (gN/strip) IER-NH + 4 (gN/bag) * * * e d c b a *

PAGE 99

99 Rainfall (mm) 0 20 40 60 80 100 120 140 SWP (MPa) -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 Time (months) A B May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Soil C:N 0 10 20 30 40 50 Control Irrigation C Figure 4-3. Effects of rainfall patterns and dry-season irrigation on soil organic carbon (C) to total nitrogen (N) ratios: (A) da ily rainfall at the study site, (B) soil water potential (SWP), (C) C:N ratios. A-B solid and open circles represent means ( se) for control and irrigatio n treatments, respectively (N = 4 for soil water potential and N=4 soil C:N ratio ). Vertical dashed lines indicate the dry season irrigation period (S ept 23rd 2004 to Jan 26th 2005). White and black horizontal bars represen t dry (Sept 21st 2004 to Jan 19th 2005), and wet season (May to Sept 20th 2004, or Jan 20th to Aug 5th 2005), respectively.

PAGE 100

100 Rainfall (mm) 0 20 40 60 80 100 120 140 A SWP (MPa) -0.12 -0.10 -0.08 -0.06 -0.04 -0.02 0.00 B Time (months) May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Soil C:N 0 10 20 30 40 50 Control Litter removal C Figure 4-4. Effects of rainfa ll patterns and litter removal on soil organic carbon (C) to total nitrogen (N) ratios: (A) daily rainfall at the study site, (B) soil water potential (SWP), (C) C:N ratios. A-B solid and open bars represent means ( se) for control and litter removal treatments, respectively (N = 4 for soil water potential and N=4 soil C:N ratio). White and black horizontal bars represent dry (Sept 21st 2004 to Jan 19th 2005), and wet season (May to Sept 20th 2004, or Jan 20th to Aug 5th 2005), respectively.

PAGE 101

101 CHAPTER 5 CONCLUSIONS Table 5-1 summarizes the responsiveness to in tra-annual variability, dry-season irrigation, and litter removal, exhibited by the suite of variables related to soil microbial structure and composition, and microbial and nutrient dynamics, that were measured for this dissertation. Overall my results reveal comp lex inter-relationships among soil microbial processes, microbial structure/composition and nutrient availability in a seasonal tropical secondary forest. Microbial biomass carbon, bacter ial and fungal densities, basa l respiration, soil phosphorus and water availability sh owed marked intrannual variation asso ciated with rainfall seasonality (Table 5-1). Although rainfall seas onality affected these variable s, dry-season irrigation had no corresponding responses on the same variables (ex cept to increased soil wa ter potential). Higher microbial biomass C in the dry season was possibl y fostered by greater root mass density in the same period (Vasconcelos 2006-diss ertation), and linked to lower basal respiration as the soil dried. Or else, nutrients accumulated in the microbial biomass in the dry season, become available at the onset of the wet season, when pl ant growth is usually at its peak, a well known nutrient-conserving strategy (Singh et al., 1989). Lower microbial C:N ratios in the wet season across treatments gives to the latter alternative a plausible explanation to the pattern seen here. Nonetheless, irrigation increased mycorrhizal-ro ot infections and phosphatase activity, and decreased fungal densities and ammoni um availability (Table 5-1). The long-term effects of litter removal co mpromised microbial structure and dynamics, microbial processes, and soil solution nutrients; combined, these effects could ultimately impair or delay aboveground processes. These effects included decreased microbial biomass C and P, bacterial and fungal densities; net N-mineralization, phosphatase activity, basal respiration, ammonium and phosphorus availability; but increa ses in mycorrhizal infections and nitrate

PAGE 102

102 availability (Table 5-1). Although arbuscular myco rrhizal fungi could be an important nutrient conduit in this tropical secondary forest, th e extent to which litter removal will cause impoverishment of the mineral soil that impa irs microbial activity remains unclear. The decrease in NH4 + and P availability was probably related to the fact that available nutrients were retained in the living biomass and recycled in th e litter, and removing the litter had a negative effect on soil nutrients. In fact, li tter removal consistently reduced litterfall N but had no effects on litterfall P (Vasconcelos 2006). NH4 + availability was also linked to increased microbial activity, but with decr eased microbial biomass C as so il water availability increased. Litter P concentrations were not affected by litter removal possibly because P-acquisition was enhanced by greater mycorrhizal associations in litter removal plots as demonstrated in this study, and in other secondary forests with lo w soil P (Marschner 1995). Microbial biomass N was not affected by litter rem oval although N-mineralization and NH4 + availability decreased, but was significantly linked with soil C:N ratios suggesting that organi c nitrogen sources may sustain microbial demand for N in the absence of aboveground litter; microbial C:N ratios were substantially reduced litter re moval. The decrease in MBC in litter removal plots may have resulted from the lack of substrate to sust ain aboveground and belowgro und microbial processes, and was correlated with the decrease in bacterial and fungal densities, and basal respiration rates. Collectively, these findings confirm that wate r and substrate availabi lity, seasonal droughts and wet-up events have an important influence on the physiological stat e of the soil microbial community and on nutrient availability in this tropical secondary forest. The evident and consistent response to litter removal confirms its ro le as a conduit of nutri ent and as a habitat for microorganisms that actively serve as a reservoir fo r nutrients, and as substrate for fine roots and mycorrhizal associations. The potential increase in nutrient uptake by fine roots and mycorrhizae

PAGE 103

103 at the soil interface may have compensated for the lack of nutrients percolating through and from the litter to the mineral soil in the litter re moval plots, and may help explain why aboveground measurements showed high resistance to alte red nutrient availability through litter removal (Vaconcelos 2006), and belowgroun d measurements were more sensitive. Conversely, the overall lack of response that belowground processe s had to increased mois ture availability may indicate the great pl asticity of these native microorgan isms to seasonal drought, whereas moisture availability at the study site significan tly constrained ANPP in the year prior to the sampling period of this study, interannual change s in ANPP were not followed by significant changes in litterfall quantity or quality, nor significantly linked to belowground processes measured in this study (Vasconcelos 2006). The steady and reduced availability of NH4 + under irrigation may reflect that of an environment without microclimate oscillations, buffered by the lack of seasonal droughts, and intermittent decomposition. Although it is unclear how this response may ultimately affect forest productivity, it is a strong c onfirmation of the pulse hypothesi s articulated by Lodge et al. (1994), especially when combined with the ove rall responsiveness to wet up events reported throughout this dissertation. Future research efforts should focus on the im pacts of moisture and nutrient constraints on microbial dynamics affects on aboveground processe s (including C uptake) in tropical secondary forests, as their importance to tr opical landscapes continues to grow.

PAGE 104

104 Table 5-1. Summary of ecosystem processes resp onses to intrannual va riability of rainfall seasonality (control plots), and to resour ce manipulations (dry-season irrigation and bi-weekly litterfall removal) The responses of manipula tive experiments are relative to control treatments. Process/Variable Intrannual VariabilityDry-season IrrigationLitter Removal Biomass C Yes 0 -Biomass N No 0 0 Biomass P No 0 -Bacteria Yes 0 -Fungi Yes --B:F Yes 0 0 Mycorrhizae No + ++ Spores No 0 0 Net Mineralization No 0 -Net Nitrification No 0 0 Phosphatase No ++ -Basal Respiration Yes 0 -NH4 + Yes --NO3 No 0 ++ PO4 Yes 0 -Soil C:N No 0 0 Soil water availability Yes ++ 0 Yes: presence of variability No: absence of variability +: slight, but signi ficant increase ++: significant increase --: significant decrease 0: no significan t variation

PAGE 105

105 APPENDIX MICROBIAL BIOMASS COMPARISSONS

PAGE 106

106Table A-1. Comparative seasonal and/ or annual mean for extractable microbial biomass C, N and P (mgC/kg-1soil) across differing tropical systems, site, depth and soil type (modifi ed w/ permission from Ra ngel-Vasconcelos 2002). Cover Type Description Period Extraction Type MBC mgC/kg-1 soil MBN mgC/kg-1 soil MBP mgC/kg-1 soil Proportionality constant (KC, KN, KP) Treatment Authors Rainy 487 51 20 KN = 0.68 Winter 662 70 29 FTd Dry Tropical Forest Summer Fumigationincubation 744 88 31 Singh et al. (1989) FTAp Primary Amazon Tropical Forest Annual (avg) Fumigationincubation 1287 Luizo et al. (1992) FN Native Forest Annual (avg) Fumigationextraction 476 35 KC = 0.33 KN = 0.54 Geraldes et al. (1995) FTAp Primary Amazon Tropical Forest Annual (avg) Fumigationextraction 659 102 KC = 0.35 FTAp Primary Amazon Tropical Forest Annual (avg) Fumigationincubation 695 111 KC = 0.41 Feigl et al. (1995) 7001500 70-80 KC = 0.35 KN = 0.68 FTs13 Tropical-Dry Secondary Forest (13y Abandoned Pasture) Wet Season Fumigationextraction 280-460 50-60 KC = 0.35 KN = 0.68 Fertilization (N, P, N+P) Davidson et al. (2004) Wet 2000 250 8.5 KC = 0.45 KN = 0.54 FTRp Primary Tropical Rain Forest Dry Fumigationextraction 1000 325 5.5 -Nutrient Gradient Cleveland et al. (2004) Wet 920 ---Dry 300 ---Control Wet 275 ---FTws20 Tropical-Wet Secondary Forest Dry Fumigationincubation 120 ---7y LR Li et al. (2004)

PAGE 107

107 Table A-1. Continued Cover Type Description Period Extraction Type MBC mgC/kg-1 soil MBN mgC/kg-1 soil MBP mgC/kg-1 soil Proportionality constant (KC, KN, KP) Treatment Authors Wet 348.40 24 Dry 686.40 28 -Pre-treatment Dry 395.20 16 Control FTws14 Tropical-Wet Secondary Forest Dry Fumigationextraction 404.11 14.22 -KC = 0.35 KN = 0.54 1yIRR RangelVasconcelos et al. (2004) Wet+Dry season Not reported 1080-1710 ---Control FTw Wet Tropical Forest Wet+Dry season Not reported 1050-1550 ---1y LR Ruan et al. (2004) Wet 200 1000 Dry 150 < 10 Control Wet 300 1000 FTm Moist Tropical Forest (old growth) Dry Fumigationextraction -100 50 KN = 0.54 KP = 0.37 5yIRR Yavitt et al. (2004) Wet+Dry season 618 --Control FTws20 Tropical-Wet Secondary Forest Wet+Dry season Fumigationincubation 202 ---7y LR Li et al. (2005) Wet 221.14 21.26 0.59 Dry 485.06 15.83 0.53 Control Wet 247.22 23.09 0.61 Dry 587.03 17.87 0.55 4yIRR Wet 158.63 17.74 0.50 FTws17 Tropical-Wet Secondary Forest Dry Fumigationextraction 468.97 21.01 0.42 KC = 0.35 KN = 0.54 KP = 0.40 4yLR Veluci 2006 : estimated from a figure plot. 7y : 7 years of litter removal. 1y : 7 years of litter removal. 5y : 5 years of dry-season irrigation

PAGE 108

108 LIST OF REFERENCES Alef, K., Nannipieri, P., Cepeda-T razar, C. (1995) Phosphatase ac tivity. Alef, K. and Nannipieri, P., ed. Methods in Applied Soil Microbi ology and Biochemistry. San Diego, CA, Academic Press, 335-344. Alef, K. (1995) Soil Respiration. Alef K., Na nnipieri P., ed. Methods in Applied Soil Microbiology and Biochemistry. San Diego, CA, Academic Press, 214-219. Anaya C. A.; Garcia-Oliva, F., and Jaramillo V.J. (2007) Rainfall and labile carbon availability control litter nitrogen dynamics in a tropical dry forest. Oecologia 150 (4), 602-610. Brookes, P.C., A. Landman, G. Pruden, and D.S. Jenkinson (1985) Chloroform fumigation and the release of soil nitrogen: A rapid direct extraction method to meas ure microbial biomass nitrogen in soil. Soil Biology and Biochemistry 1 7 837-843. Carvalho, C. J. R., Davidson, E. A., Santos, T. P., Dutra, F. C. and Serrao, B. O. (2004) Changing enzymatic activities and mycorrhi zal infections in a chronosequence of secondary and mature forests of eastern Am azonia. III Conferencia Cientifica do LBA. Cavelier, J., S.J. Wright, and J. Santamara (1999) Effects of irrigation on litterfall, fine root biomass and production in a semideciduous lowl and forest in Panama, Plant and Soil 211, 207-213. Chander, K., Goyal, S., Nandal, D. P., Kapoor K. K. (1998) Soil organic matter, microbial biomass and enzyme activities in a tropical ag roforestry system. Bi ology and Fertility of Soils 27, 168-172. Chapin III, F. S., Matson, P. A., Mooney, H. A. (2002) Terrestrial Plant Nutrient Use. In Principles of Terrestrial Ecosystem Ecology, ed. F. S. Chapin III, P. A. Matson and H. A. Mooney, Springer New York 176-196. Chu, E. And Diekmann, U. (2002) Efeitos de Usos Alternativos do Solo sobre a Populacao de Fungos Micorrizicos Arbusculares na Amaz onia. Embrapa Amazonia Oriental, 1-20. Cleveland, C. C., Towsend, A. R., Constance, B. C. (2004) Soil Microbial Dynamics in Costa Rica: Seasonal and Biogeochemical Constraints. Biotropica 36, 184-195. Coelho, R. F. R., Zarin, D. J., Miranda, I. S., Tu cker, J. M. (2003) Ingresso e mortalidade de uma floresta em diferentes estagi os sucessionais no municpio de Castanhal, Par (Recruitment and mortality of different stages of successional forest in th e municipality of Castanhal, Par). Acta Amazonica 33, 619-630. Cornejo, F. H., Varela, A., Wright, S. J. ( 1994) Tropical forest li tter decomposition under seasonal drought: nutrient release, fungi and bacteria. Oikos 70, 183-190.

PAGE 109

109 Cornejo, P.; Borie, F.; Rubio, R., and Azcon, R. (2007) Influence of nitrogen source on the viability, funcionality and persistence of Glomus etunicatum fungal propagules in an Andisol. Applied Soil Ecology 35, 423-431. Davidson, E. A., Carvalho, C. R., Vieira, I. C. G ., Figueiredo, R., Moutinho, P., Ishida, F. Y., Santos, M. T. P., Guerrero, J. B., Kemel, K. (2004) Nitrogen and Phosphorus Limitation of Biomass Growth in a Tropical Secondary Fo rest. Ecological App lications 14, S150-S163. Davidson, E. A., Verchot, L. V., Cattnio, J. H., Ackerman, I. L., Carvalho, J. E. M. (2000) Effects of soil water content on soil respirati on in forests and cattle pastures of eastern Amazonia. Biogeochemistry 48, 53-69. Feigl, B. J., Sparling, G. P., Ross, D. G. a nd Cerri, C. C. (1995) Soil Microbial Biomass in Amazonian soils: evaluation of methods and es timates of pool sizes. Soil Biology and Biochemistry 27, 1467-1472. Fisk, M. C. and Fahey, T. J. (2001) Microbia l biomass and nitrogen cycling responses to fertilization and litte r removal in young northern hardw ood forests. Biogeochemistry 53, 201-223. Forster, J. C. (1995) Organic Ca rbon. Alef, K. and Nannipieri, P ., ed. Methods in Applied Soil Microbiology and Biochemistry. Sa n Diego, CA, Academic Press, 59-65. Frizano, J.; Vann, D. R.; Johnson, A. H., and J ohnson, C. M. (2003) Labile phosphorus in soils of forest fallows and primary forest in the Bragantina Region, Brazil. Biotropica, 35(1), 211. Geraldes, P. A., Cerri, C. C., Feigl, B. J. (1995) Biomassa microbiana de solo sob pastagens na Amaznia. Revista Brasileira de Cincia do Solo, 1, 55-60. Gerdemann, J. W. and Nicolson, T. H. (1963) S pores of mycorrhizal en dogene species extracted from soil by wet sieving and decanting. Tr ansaction British Mycology Society, 26, 235246. Giovannetti N. and Mosse B. ( 1980) An evaluation of technique s for measuring VA mycorrhizal infection in roots. New Phytologist, 84, 489-500 Giblin, A. E., Laundre, J. A., Nadelhoffer, K. J., Shaver, G. R., (1994) Measuring nutrient availability in Arctic soils using ion-exchange resins A field test. Soil Science Society of America Journal 584, 1154-1162. Gollan, Schurr, U. and Schulze, E. D. (1992) St omatal response to dryi ng soil in relation to changes in xylem sap composition of Helia nthus annuus. Plant Cell and Environment 15, 551-559. Hamel C. (2004) Impact of arbuscular mycorrhizal fungi on N and P cycling in the root zone. Canadian Journal of Soil Science 31; 84(4), 383-395.

PAGE 110

110 Hart, S. C., Stark, J. M., Davidson, E. A., and Firestone, M. K. (1994) Nitrogen mineralization, immobilization, and nitrification. Methods of Soil Analysis. Microbiological and Biogeochemical Propertie s SSSA, Madison, WI. Hart, M., Trevors, J. T. (2005) Microbe mana gement: application of mycorrhizal fungi in sustainable agriculture. Frontiers of Ecology and Environment 3, 533-539. Hodge, A. Robinson D., and Fitter, (2000). Are mi croorganisms more effec tive than plants at competing for nitrogen? Trends in Ecology and Evolution 5(7), 305-308. Islam, K. R., Weil, R. R. (1998) Microwave ir radiation of soil for routine measurement of microbial biomass carbon. Biol ogy and Fertility of Soils 27 408-416. Jipp, P. H., Nepstad, D. C., Cassel, D. K., Carval ho, C. R. D. (1998) Deep soil moisture storage and transpiration in forests and pastures of seasonally-dry Amazonia. Climatic Change 39, 395-412. Johnson, N. C., O'Dell, T. E. and Bledsoe, C. S. (1999) Methods for Ecological Studies of Mycorrhizae. Robertson, G. P., Coleman, D. C., Bledsoe, C. S. and Sollins, P., ed. Standard Soil Methods for Long-Term Ecological Research. Oxford University Press, Inc., Oxford. 378-412. Kaye J. P. and Hart S.C (1997) Competi tion For Nitrogen Betw een Plants And Soil Microorganisms Trends In Ecology & Evolution 12(4), 139-143. Kiese, R., Butterbach-Bahl, K. (2002) N2O and CO2 emissions from three different tropical forest sites in the wet tropics of Queensla nd, Australia. Soil Biology and Biochemistry 34, 975-987. Lajtha, K., Driscoll, C. T., Jarrell, W. M ., and Elliot, E. T. (1999) Soil Phosphorus: Characterization and Total Elemental Analysis In Standard Soil Methods for Long-Term Ecological Research, ed. G. P. Robertson, D. C. Coleman, C. S. Bledsoe and P. Sollins, 115-142. Oxford University Press, Inc., Oxford. Lean, J., Bunton, C. B., Nobre, C. A. and Rowntree, P. R. (1996) The simulated impact of Amazonian deforestation on climate using measured ABRACOS vegetation characteristics, in Amazonian deforestati on and climate., ed. J.H.C. Gash, John Wiley, Hoboken, NJ., 549-576. Li, Y., Xu, M., Sun, O. J., Cui, W. (2004) E ffects of root and litter removal on soil CO2 effux and microbial biomass in wet tropical fo rests. Soil Biology & Biochemistry 36, 2111-2114. Li, Y., Xu, M., Zou, X., Xia, Y. (2005) Soil CO2 efflux and fungal and bacterial biomass in a plantation and a secondary fore st in wet tropics in Puerto Rico. Plant and Soil 268, 151160. Lodge, D.J., McDowell, W.H., McSwiney, C.P. (1994) The importance of nutrient pulses in tropical forests. Trends in Ecology and Evolution 9, 384-387.

PAGE 111

111 Lopes, E.S., Oliveira, E., Dias, R., Scheck, N.C. (1983) Occurrence and distribution of vesiculararbuscular mycorrhizal fungi in coffee (Coffee arabica L.) plantations in central So Paulo state, Brazil. Turrialba, 33(4), 417-422. Lucas, Y., Luizao, F. J., Chauvel, A., Rouille r, J., Nahon, D. (1993) The relations between biological activity of the rain forest and mineral composition of soils. Science 260, 521523. Luizao, C. C. R., Bonde, T. A., Rosswall, T. (1 992) Seasonal variation of soil microbial biomass the effects of clearfelling a tropical rainfore st and establishment of pasture in the central Amazon. Soil Biology and Bi ochemistry 24, 805-813. Luizao, F. J., Schubart, H. O. R., (1987). Litt er production and decompos ition in a terra-firme forest of Central Amazonia. Experientia 43, 259-265. Malcolm, R. E. (1983) Assessment of phosphatase-a ctivity in soils. Soil Biology & Biochemistry 15(4), 403-408. Mamilov, A. S., Dilly, O. M. (2002) Soil Micr obial ecophysiology as affected by short-term variations in environmental conditions Soil Biology and Biogeochemistry 34, 1283-1290. Markewitz, D., Davidson, E., Moutinho, P., and Nepstad, D. (2004) Nutrient loss and redistribution after forest clearing on a highly weathered soil in Amazonia. Ecological Applications 14(4) S177-S199. Marschner, P., Marino, W., Lieberei, R. (2002) Seasonal Effects On Microorganisms In The Rhizosphere Of Two Tropical Plants In A Po lyculture Agroforestry System In Central Amazonia, Brazil. Biology and Fertility of Soils 35, 68-71. Mcgroddy, M. E., Silver, W. L., De Oliveira, R. C. (2004) The Effect of Phosphorus Availability on Decomposition Dynamics in a Seasonal Lowland Amazonian Forest. Ecosystems 7, 172-179. Mulvaney, R. L. (1996) Colorimetric Methods. In Methods of Soil Analysis, ed. D. L. Sparks, A. L. Page, P. A. Helmke, R. H. Loeppert, P. N. Soltanpour, Tabatabai, M. A., A. E. Summer, Vol. Chemical and Microbial Prop erties Agronomy Monograph no. 9, 1123-1184. Chemical Methods SSSA, Madison, Wisconsin, USA. Murphy, J., Riley, J. P. (1962) A modified sing le solution method for the determination of phosphate in natural waters. Anal ytical Chemistry Acta 27, 31-36. Orchard, V. A., Cook, F. J. (1983) Relationship between soil respir ation and soil moisture. Soil Biology and Biochemistry 15, 447-453. Paul, E. A. and Clark F. E. (1997) Phos phorus transformations. Soil Microbiology and Biochemistry. 2 ed. New York: Academic Press, 289-298.

PAGE 112

112 Rangel-Vasconcelos, L. G. T., Zari n, D., Carvalho, C. J. R., Santos, M. M. L., Vasconcelos, S. S. and Oliveira, F. A. (2005) Carbono, nitrognio e atividade da bioma ssa microbiana de um solo sob vegetao secundria de diferentes idades na Amaznia Oriental. Revista de Ciencias Agrrias. Rilling, M. C. (2004) Arbuscular mycorrhizae a nd terrestrial ecosystem processes. Ecology Letters 7, 740-754. Robertson, G. P., Wedin D., Groffman P. M., Blai r J. M., Holland E. A., Nadelholffer K. J. and Harris D. (1999) Soil carbon and nitrogen av ailability: nitrog en mineralization, nitrification, and soil respira tion potentials. Robertson, G. P.; Coleman, D. C.; Bledsoe, C. S., and Sollins, P., Editors. Standard Soil Methods for Long-Term Ecological Research. Oxford: Oxford University Press, Inc. 258-271. Ross, D. J. (1987) Soil Microbial Biomass Estim ated by the Fumigation Incubation Procedure Seasonal Fluctuations and Influence of Soil-Moisture Content Soil. Biology & Biochemistry, 19(4), 397-404. Roy, S., Singh, J. S. (1995) Seasonal and spatial dynamics of plant-available N and P pools and N-mineralization in relation to fine roots in a dry tropical forest ha bitat. Soil Biology & Biochemistry 27, 33-40. Ruan, H. H., Zou, X. M., Scatena, F. N., Zimm erman, J. K. (2004) Asynchronous fluctuation of soil microbial biomass and plan t litterfall in a tropical we t forest. Plant and Soil 260, 147154. Sayer, E. J. (2005a) Consequences of changes in li tterfall for soil respiratio n in tropical forests. Chapter IV, dissertation for the degree of Doct or of Philosophy: Leaf litter manipulation in a tropical forest. University of Cambridge, UK, 93-109. Sayer, E. J. (2005b) Changes in nutrient cycling following increased or de creased litterfall in a moist tropical forest. Chapter V, dissertation for the degree of Doctor of Philosophy: Leaf litter manipulation in a tropical forest University of Cambridge, UK, 93-109. Sayer, E. J., Tanner, E.V. J., Lacey, A. L. (2006) Effects of litter manipulation on early-stage decomposition and meso-arthropod abundance in a tropical moist forest. Forest Ecology and Management 229, 285-293. Sayer, E. J. (2006) Using experimental manipulat ion to assess the roles of leaf litter in the functioning of forest ecosystem s. Biological Reviews 81, 1-31. Schimel, J. P. and Bennett, J. (2004) Nitroge n mineralization: challenges of a changing paradigm. Ecology 85(3), 591-602. Schwendenmann, L.,Veldkamp, E., Brenes, T., O'br ien, J. J., Mackensen, J. (2003) Spatial and temporal variation in soil CO2 efflux in an old-growth neotro pical rain forest, La Selva, Costa Rica. Biogeochemistry 64, 11-128.

PAGE 113

113 Shock, C. C., Akin, I. A., Unlenen, L. A., Feibert, E. B. G., Cedric, A., (2003) Precise irrigation scheduling using soil moisture sensors. In ternational Irrigation Show 2003 Proceedings, The Irrigation Associat ion. San Diego, CA, 18-20. Shuttleworth, W. J., Gash, J. H. C., Lloyd, C. R., Moore, C. J., Roberts, J., Marques, A. D., Fisch, G., Silva, V. D., Ribeiro, M. D. G., Molion, L. C. B., Sa, L. D. D., Nobre, J. C. A., Cabral, O. M. R., Patel, S. R., Demoraes J. C., (1984) Eddy correlation measurements of energy partition for Amazonian forests. Quar terly Journal of the Royal Meteorological Society 110, 1143-1162. Singh, J. S., Raghbanshi A. S., Singh, R. S., Srivas tava, S. C. (1989) Microbial biomass acts as a sink of plant nutrients in dry tropical forests a nd savanna. Nature 338, 499-500. Sinsabaugh, R. L., Klug, M. J., Collins, H. P., Yeager, P. E., Peterson, S. O. (1999) Characterizing the Soil Microbial Communities. In Standard Soil Methods for Long-Term Ecological Research, ed. Robertson, G.P., Colema n, D.C., Bledsoe, C. S., Sollins, P., Vol. 2, 318-348. Oxford University Press, New York. Skogley, E. O., Dobermann, A., (1996) Synthetic ion-exchange resins: soil and environmental studies. Journal of Environmental Quality 25, 13-24. Smith J.L., Halvorson J. J., Bolton H. (1994) Sp atial Relationships of Soil Microbial Biomass and C and N Mineralization in a Semiarid Shrub-Steppe Ecosystem. Soil Biology & Biochemistry 26, 1151-1159. Srivastava, S. C. (1992) Microbial C, N and P in dry tropical soils: seasonal changes and influence of soil moisture. Soil Biology and Biochemistry 24, 711-714. Tenrio, A. R. D. M., Graa, J. J. D. C., Ges, J. E. M., Mendez, J. G. R., Gama, J. R. M. F., Silva, P. R. O. D., Chagas, P. S. M. D., Silva, R. N. P. D., Amrico, R. R., Pereira, W. L. M., (1999). Mapeamento dos solos da estao de piscicultura de Castanhal, PA. FCAP Informe Tcnico 25, 5-26. Tobon, C., Sevink, J., Verstraten, J. M., (2004) Litterflow chemistry and nutrient uptake from the forest floor in Northwest Amazonian fo rest ecosystems. Biogeochemistry 69, 315-339. Treseder, K. K. and Vitousek P. M. (2001) Eff ects of soil nutrient availability on investment in acquisition of N and P in Hawaiian Ra in Forests. Ecology 82(4), 946-954. Turrion, M. B., Gallardo, J., Gonzalez, M., (1997) Nutrient availability in forest soils as measured with anion-exchange membra nes. Geomicrobiology Journal 14, 51-64. Vance, E. D., Brookes, P. C., Jenkinson, D. S. (1987) An improved method for measuring soil microbial biomass C. Soil Bi ology and Biochemistry 19,703-707.

PAGE 114

114 Vasconcelos, S. S., Zarin, D. J., Capanu, M., Litte ll, R., Davidson, E. A., Ishilda, F. Y., Santos, E. B., Araujo, M. M., Aragao, D. V., Range l-Vasconcelos, L. G. T., Oliveira, F. A., McDowell, W. H., Carvalho, C. J. R. (2004) Mo isture and substrate av ailability constrain soil trace gas fluxes in an eastern Amazoni an re-growth forest. Global Biogeochemical Cycles 18, 1-10. Vasconcelos, S.S., Zarin, D.J., Rosa, M.B.S., Oliveira, F.A., Carvalho C.J.R. (2006). Leaf decomposition in a dry season irrigation e xperiment in Eastern Amazonian forest regrowth. To be submitted to Biotropica. Vasconcelos, S. S. (2006) Mositure and nutrient c onstraints to ecosystem processes in a forest regrowth stand in Eastern Amazonia, Brasil. Dissertation for the degree of Doctor of Philosophy. University of Florida. Veluci, R. M., (2006) Effects of irrigation and litter removal on microbial processes in a seasonally dry tropical forest. Chapter III, Dissertation for the degree of Doctor of Philosophy: Irrigation and litter ma nipulation in a seasonally dry tropical forest. University of Florida. Vitousek, P. M., Sanford. R. L., (1986) Nutrient cycling in moist tropical forest. Annual Review of Ecology and Systematics 17, 137-67. Vitousek, P. M., Hattenschwiler, S., Olander, L., and Allison S. (2002) Nitrogen and Nature. Royal Swedish Academy of Sciences 31(2), 97-101. Voroney, R. P., Winter, J. P., Gregorich, E. G. (1991) Microbe and plant interactions. In Carbon isotope techniques, edts. D.C. Coleman a nd B. Fry, 77-99. Academic Press, New York. Wardle, D. A. (2002) Communities and Ecosystems: Linking the Aboveground and Belowground Components. In Global Change Phenomena in an AbovegroundBelowground Context, ed. D. A. Wardle, Pr inceton University Press, Princeton, New Jersey, 239-292. Wieder, R. K., Wright, S. J. (1995). Tropical forest litter dyn amics and dry season irrigation on Barro Colorado Island, Panama. Ecology 76: 1971-1979. Williams, M. A. and Rice, C. W. (2007) Seven y ears of enhanced water availability influences the physiological, struct ural, and functional attributes of a soil microbial community. Applied Soil Ecology 35, 535-545. Yang, J. E., Skogley, E. O., Schaff, B. E., Ki m, J. J. (1998) A simple spectrophotometric determination of nitrate in water, resin and soil extracts. Soil Science Society of America 62, 1108-1115. Yavitt, J. B., Wieder, R. K., Wright, S. J. ( 1993) Soil nutrient dynamics in response to irrigation of a Panamanian tropical moist forest. Biogeochemistry 19, 1-25.

PAGE 115

115 Yavitt, J., Wright, S. (1996) Temporal patterns of soil nutrients in a Panamanian moist forest revealed by ion-exchange resin and expe rimental irrigation. Plant Soil 183, 117-129. Yavitt, J., Wright, S., Wieder, R. (2004) Seas onal drought and dry-season irrigation influence leaf-litter nutrients and soil en zymes in a moist, lowland forest in Panama. Austral Ecology 29, 177-188.

PAGE 116

116 BIOGRAPHICAL SKETCH Roberta Medeiros Veluci-Marlow was born in July 27th 1976, in Franca SP, Brazil. She graduated with a bachelors degree in biology from the Univ ersity of Franca in 1997 and engaged in an exchange program in the US to learn English between fall 1998 and 1999. Thereafter, Roberta acquired a masters degree in soil ecology from the University of Toledo in 2002 where she worked on N-cycling and microbiotic crusts. Her future goals include a career as a researcher and developer of research-ori ented educational program s. Robertas biggest accomplishment in life was giving birth to Tiago in February 2006.