<%BANNER%>

Moisture and Nutrient Constraints to Ecosystem Processes in a Forest Regrowth Stand in Eastern Amazonia, Brazil


PAGE 1

1 MOISTURE AND NUTRIENT CONSTRAINTS TO ECOSYSTEM PROCESSES IN A FOREST REGROWTH STAND IN EASTERN AMAZONIA, BRAZIL By STEEL SILVA VASCONCELOS 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 2006

PAGE 2

2 Copyright 2006 by Steel Silva Vasconcelos

PAGE 3

3 This work is dedicated to my wi fe Lvia and my daughter Crita.

PAGE 4

4 ACKNOWLEDGMENTS I am especially thankful to Lvia and Crita for tolera ting my physical and sometimes mental absences during critical periods throughout this challengi ng journey that began in July 1999, as well as for their help dur ing fieldwork. I thank my parent s for their efforts to provide me with an adequate education that served as a foundation to meet the requirements of the PhD program at the University of Florida. I am very grateful to my advisor, Dr. Daniel Zarin, for believing that I could succeed in the program and especially for the efficient and invalu able guidance and support that were critical to my academic development. I also thank the other members of my committeeDr. Eric Davidson, Dr. Steve Mulkey, Dr. Ted Schuur, an d Dr. Wendell Cropperfor their input during the program and insightful suggestions that greatly improved this dissertation. I thank the School of Forest Resources and Conservation for overall support, the Andrew Mellon Foundation for research funding, and Empr esa Brasileira de Pesquisa Agropecuria (EMBRAPA) for support to conclude the program. This ecosystem-level work would not have been possible without the help of many research assistants, technicians, and underg raduate students of the MANFLORA Project in Brazil (Wilson Oliveira, Beatriz Ro sa, Joanna Tucker, Lucas Fortini, Roberta Veluci-Marlow, Dbora Arago, Gizelle Benigno, Ronaldo Oliveira, Tmara Lima, Roberta Coelho, Alexandre Modesto, Elisngela Santos, Ana Jlia Amaral, a nd Evandro da Silva) and in the US (Leandra Arago and Patrcia Sampaio) who contributed to data collec tion and processing and insightful discussions. I offer extra special thanks to Dr. Maristela Arajo for data collection and processing, to Osrio Oliveira, Glebson Sousa, a nd Paulo Alencar for their assistance in the field, to Raimundo Nonato da Silva (UFRA) and hi s team (Manoel, Gilson, Geraldo Jr., and Seu Geraldo) for logistical suppor t, and to the Brazilian coor dinators of MANFLORAProf.

PAGE 5

5 Francisco de Assis Oliveira and Prof. Izildinha Miranda. Laborato ry support at Embrapa Eastern Amazon by Dr. Cludio Carvalho and his team (esp ecially Ivanildo Trinda de), as well by Dr. Marcus Vasconcelos, is greatly a ppreciated. Thanks go also to Dr. Marinela Capanu and Prof. Ramon Littell for helping with statistical analysis. Finally, I would like to thank a ll the friends in Gainesville that made the days for me and my family extremely pleasant: Camila and Guto Paula; Carolina and Vi ctor; Darlene, Eduardo, Luiza and Lgia Carlos; Gergia, Jos Carlos and Vtor Dubeux; Graziela, Lucinda and Roberta Noronha; Joanna Tucker; Juliana and Flvio Silv estre; Lucas Fortini; Mara, Lus, gor, Andr and Rassa Lima; Mrs. Lilian Raye; Patrcia and Emlio Bruna; Marcelo and Aline Carvalho; Roberta and Brian Marlow; and Rutecleia Zarin.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES................................................................................................................ .........9 ABSTRACT....................................................................................................................... ............11 CHAPTER 1 INTRODUCTION..................................................................................................................12 Literature Review.............................................................................................................. .....14 Moisture and Nutrient Limita tions to Tropical Forests...................................................14 Observational and Manipulative Experime nts to Study Moisture and Nutrient Limitations in Tropical Forests....................................................................................15 Moisture Effects on Ecosystem Processes in Tropical Forests.......................................18 Aboveground processes............................................................................................19 Belowground processes............................................................................................22 Nutrient Effects on Ecosystem Pr ocesses in Tropical Forests........................................23 Aboveground processes............................................................................................24 Belowground processes............................................................................................25 Conclusions.................................................................................................................... .........26 2 STUDY SITE AND EXPERIMENTAL DESIGN.................................................................28 Study Site..................................................................................................................... ...........28 Experimental Design............................................................................................................ ..29 3 SEASONAL AND EXPERIMENTAL E FFECTS ON LITTERFALL QUANTITY AND QUALITY IN EASTERN AMAZ ONIAN FOREST REGROWTH...........................34 Introduction................................................................................................................... ..........34 Material and methods........................................................................................................... ..35 Litterfall..................................................................................................................... ......35 Litter Stock................................................................................................................... ...37 Statistical Analysis..........................................................................................................37 Results........................................................................................................................ .............38 Non-woody Litterfall.......................................................................................................38 Irrigation experiment................................................................................................38 Litter removal experiment........................................................................................40 Litter stock................................................................................................................... ....41 Discussion..................................................................................................................... ..........42 Seasonal Patterns.............................................................................................................42 Limited Impact of Dry-season Irrigation.........................................................................43 Litter Removal Reduces Litte rfall N Concentration.......................................................45

PAGE 7

7 4 LEAF DECOMPOSITION IN A DRY SEASON IRRIGATION EXPERIMENT IN EASTERN AMAZONIAN FOREST REGROWTH.............................................................60 Introduction................................................................................................................... ..........60 Study Site and Experimental Design......................................................................................61 Material and Methods........................................................................................................... ..61 Leaf Litter Decomposition..............................................................................................61 Initial Leaf Litter Chemistry............................................................................................63 Specific Leaf Area...........................................................................................................63 Statistical Analysis..........................................................................................................64 Results........................................................................................................................ .............64 Discussion..................................................................................................................... ..........65 5 MOISTURE AND SUBSTRATE AVAILABILI TY CONSTRAIN SOIL TRACE GAS FLUXES IN AN EASTERN AMAZ ONIAN REGROWTH FOREST.................................74 Introduction................................................................................................................... ..........74 Material and Methods........................................................................................................... ..76 Field Measurements.........................................................................................................76 Statistical Analyses..........................................................................................................78 Results........................................................................................................................ .............79 Irrigation Experiment......................................................................................................79 Litter Removal Experiment.............................................................................................82 Discussion..................................................................................................................... ..........83 Soil CO2 Efflux and Belowground C Allocation............................................................83 Nitrogen Oxide Emissions...............................................................................................87 Methane Emissions..........................................................................................................88 Conclusions.................................................................................................................... .........88 6 MOISTURE CONSTRAINTS TO ABOVEGROUND NET PRIMARY PRODUCTIVITY IN EASTERN AM AZONIAN FOREST REGROWTH.........................97 Introduction................................................................................................................... ..........97 Study Site and Experimental Design......................................................................................98 Material and Methods........................................................................................................... ..98 Aboveground Net Primary Productivity..........................................................................98 Statistical Analysis..........................................................................................................99 Results........................................................................................................................ .............99 Discussion..................................................................................................................... ........100 7 CONCLUSIONS..................................................................................................................106 LIST OF REFERENCES.............................................................................................................109 BIOGRAPHICAL SKETCH.......................................................................................................128

PAGE 8

8 LIST OF TABLES Table page 2-1 Characteristics of rainfall distribution and intensity duri ng the experimental period in the site....................................................................................................................... .........32 2-2 Dry-season irrigation intervals and associ ated rainfall intensity and distribution.............32 3-1 F statistics and associated significance le vels for the effects of treatments, sampling date, and their inter action on non-woody litterfa ll mass and nutrients..............................47 3-2 Estimates of annual non-woody litterfall (mass, nitrogen, and phosphorus), nonwoody litter stock, a nd litterfall:forest floor mass ratio (kL) in tropical forests................48 4-1 F statistics and significance levels for the effects of tr eatments (control and irrigation), species, and thei r interactions on leaf l itter decomposition rates....................68 4-2 Decomposition rates for overs tory species in a tropical regrowth forest stand in eastern Amazonia, Brazil...................................................................................................68 4-3 Decomposition rates for overstory species under c ontrol and irrigated plots (Experiment 3) in a tropical regrowth fo rest stand in eastern Amazonia, Brazil..............69 4-4 Initial quality parameters of leaves in cubated in litterbags fo r decomposition studies in a tropical regrowth forest in Eastern Amazonia, Brazil.................................................69 4-5 Pearson correlation coefficien ts between decomposition rate ( k ) and initial quality parameters of leaves of overstory tr ee species incubated in litterbags..............................70 4-6 Decomposition rates estimated from litterbag studies for so me tropical forest sites........70 4-7 Chemical composition of leaf litte r for some tropical forest sites.....................................71 5-1 F statistics and significance levels for the effect of treatments, sampling date, and their interaction on soil trace gas fluxes and non-woody litterfall.....................................90 5-2 Annual soil carbon efflux and non-woody litterf all carbon for control, irrigated and litter remova l plots........................................................................................................... ..91 5-3 Estimates of annual soil carbon efflux in ol d-growth and regrowth tropical forests.........92 6-1 Allometric equations used to estimate tr ee biomass in a tropical regrowth forest stand in eastern Amazonia, Brazil.............................................................................................103 7-1 Summary of ecosystem process responses to intrannual and inte rannual variability effects and resource manipulation effects........................................................................108

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Simplified conceptual diagram of likel y effects of drought on leafand ecosystemlevel processes addressed in this dissertation....................................................................27 2-1 Daily rainfall during the experimental period....................................................................33 2-2 Experimental plot layout................................................................................................... .33 3-1 Effects of rainfall patterns and dry-s eason irrigation on monthly non-woody litterfall mass........................................................................................................................... .........50 3-2 Effects of dry-season irrigation and litter removal on annua l non-woody litterfall mass........................................................................................................................... .........51 3-3 Effects of rainfall patterns and dryseason irrigation on non-woody litterfall nutrient concentration.................................................................................................................. ....52 3-4 Effects of rainfall patterns and dry-s eason irrigation on monthly non-woody litterfall nutrient return................................................................................................................ .....53 3-5 Effects of dry-season irrigation on annual non-woody litterfall nutrient return................54 3-6 Effects of rainfall patterns and litter removal on non-woody litterfall mass.....................55 3-7 Effects of rainfall patterns and li tter removal on monthl y non-woody litterfall nutrient concentration........................................................................................................56 3-8 Effects of rainfall patterns and li tter removal on monthl y non-woody litterfall nutrient return................................................................................................................ .....57 3-9 Effects of litter removal on annual non-woody litterfall nutrient return...........................58 3-10 Non-woody litter stock for c ontrol and irrigation plots.....................................................59 4-1 Effects of dry-season irriga tion on leaf litter decomposition.............................................72 4-2 Relation between decomposition rate and in itial leaf litter characteristics for tree species........................................................................................................................ ........73 5-1 Effects of rainfall patterns and dry-season irrigation on soil moisture status and soil respiration.................................................................................................................... ......93 5-2 Effects of rainfall patterns and dry-s eason irrigation on soil nitrogen oxide and methane effluxes............................................................................................................... .94

PAGE 10

10 5-3 Effects of rainfall patterns and litter removal on soil moisture status and soil respiration.................................................................................................................... ......95 5-4 Effects of rainfall patterns and litter removal on soil nitrogen oxide and methane effluxes....................................................................................................................... ........96 6-1 Effects of dry-season irrigation on ab oveground increment and non-woody litterfall....104 6-2 Relationship between aboveground net primary productivity and rainfall......................105

PAGE 11

11 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 MOISTURE AND NUTRIENT CONSTRAINTS TO ECOSYSTEM PROCESSES IN A FOREST REGROWTH STAND IN EASTERN AMAZONIA, BRAZIL By Steel Silva Vasconcelos December 2006 Chair: Daniel Jacob Zarin Major Department: Forest Resources and Conservation Changes in land-use and climate are likely to alter resource (e.g., mo isture and nutrient) availability in tropical forest so ils, but quantitative assessment of th e role of resource constraints as regulators of ecosystem proce sses is rather limited. In this dissertation, moisture and nutrient availability were altered thr ough dry-season irrigation and bi-week ly aboveground litter removal, respectively, to study how these resources control aboveground and belowground ecosystem processes in a forest regrowth stand in the Br azilian Amazon. Moisture availability strongly constrains soil respiration as indicated by the re sponses of soil carbon dioxide emissions to soil wet-up events and dry-season irri gation. Higher moisture availabi lity in irrigated plots also increased leaf litter decomposition and slightly increased soil nitrous oxide and methane emissions, but did not alter monthly litterfall quan tity and quality, and soil nitric oxide emission. Litter removal decreased carbon dioxide emissions and litterfall nitrogen c oncentration, but had no effects on litterfall quantit y, and soil nitrogen oxides and methane emissions. Aboveground net primary productivity was constrained by moisture availability as indi cated by the response of wood increment to interannual variation in dr y season rainfall and to irrigation, suggesting decreased potential of carbon sequestration from forest regrowth under an ticipated scenarios of reduced rainfall in Amazonia.

PAGE 12

12 CHAPTER 1 INTRODUCTION In many tropical areas, esp ecially in the Brazilian Am azon, old-growth forests are increasingly being converted to forest re growthalso known as secondary or successional forestsfollowing abandonment of slash-and-burn agriculture and cattle pasture. Fearnside (1996) estimated that about 50% of the defore sted Brazilian Amazon landscape was in some stage of forest regrowth in 1990. Forest regrowth provides important ecosystem services such as carbon sequestration, rees tablishment of nutrient and wate r cycles, and maintenance of biodiversity (Brown & Lugo 1990, Markewitz et al. 2004, Nepstad et al. 2001, Sommer et al. 2002). In addition, these forests often represen t an important source of income (woody and nonwoody forest products) to local people (Brown & Lugo 1990). Adequate management of forest regrowth can represent an importa nt alternative to reduce pressure on old-growth forest sites in the Amazon region (Brown & Lugo 1990). Forest regrowth can play an important role in regional and global carbon (C) dynamics because of their high rates of biomass accu mulationa proxy for C sequestration (Zarin et al. 2001)although the frequent cleari ng of regrowth results in small net C uptake compared to total emissions from deforestation in the Amaz on (Steininger 2004). Effo rts to determine the capacity of forest regrowth to sequester C at di fferent spatial and tempor al levels have been pursued by recent modeling efforts (Neeff 2005, Zarin et al. 2001). Analyses of the rates of and controls on biomass accumulation may contribute to improved modeling of the potential of forest C sequestration (Johnson et al. 2000). Observational studies ha ve shown that several factors control the rate of biomass accumulation in tropical regrowth sites; these include land-use history (including disturbance type and intensity) (Gehring et al. 2005, Moran et al. 2000, Uhl et al. 1988, Zarin et al. 2005), surrounding vegetati on, soil fertility (Gehring et al. 1999, Moran et al.

PAGE 13

13 2000), and climate (Zarin et al. 2001). Such a variety of c ontrolling factors complicates modeling efforts of forest regrowth rates, bu t a recent synthesis of observational studies has shown that soil texture and dry season length are strongly correlated with C accumulation by forest regrowth, possibly through their effects on the availability of soil moisture and nutrients (Zarin et al. 2001). Further understanding of the processes by wh ich moisture and nutrient availability constrain biomass accumulation rates may be critical to understand the role of forest regrowth on regional and global C dynamics under future la nd-use and climate change scenarios. Manipulative experiments are required to better comprehend the role of resource availability on forest ecosystem processes. Unfortunately such experiments have rarely been employed to study ecosystem processes in tropical forests in genera l and especially in Amaz onian forest regrowth, which represents an important component of the landscape in the regi on (Fearnside 1996, Neeff et al. 2006, Zarin et al. 2001). This study is part of the M ANFLORA project (Manipulation of Moisture and Nutrient Availability in Young Regrowth Forests in Easter n Amazonia), which is a collaborative research program among the University of Florida, the Universidade Federal Rural da Amaznia (Federal Rural University of AmazoniaUFRA), and EMBRAPA Amaznia Oriental (EMBRAPA Eastern Amazon) initiated in 1999 at the UFRA expe rimental station in Cast anhal, Par, Brazil. Since 2001, moisture and nutrient ava ilability have been altered in two separate experiments: (1) dry-season irrigation and (2) c ontinuous litter remova l. To my knowledge, the MANFLORA project represents the only long-term, large-scal e (stand-level) experime ntal manipulation of moisture availability in tropical forest regrow th, and one of the few tr opical forest regrowth nutrient manipulation studies.

PAGE 14

14 In this dissertation, I examine the influence of resource availability on carbon and nutrient dynamics associated with litterf all, leaf litter decomposition, so il trace gases, and aboveground net primary productivity (ANPP). The dissertati on is divided into 7 chapters, including this general introduction and literat ure review. The second chapte r describes the study site and experimental design. The effects of moisture an d nutrient availability on litterfall (Chapter 3), leaf decomposition (Chapter 4), soil trace ga ses (Chapter 5), and ANPP (Chapter 6) are examined separately. Conclusions are summarized in Chapter 7. Literature Review This literature review addresses the role of mo isture and nutrient availa bility as constraints on ecosystem processes related to C dynamics in tropical sites, with sp ecial emphasis on forest regrowth sites in the Brazilian Amazon. The grow th of adult trees is the main focus of this review since they contribute the most to stan d-level C balance, although the effects of abiotic stresses on the understory [e.g., drought (Arago et al. 2005, Fortini et al. 2003)] may be more dramatic than on overstory plants Finally, the review is direct ed to moist, lowland evergreen tropical forests (Whitmore 1992), even though releva nt information from dry, deciduous tropical forests is also included, since there is a great deal of literature on rainfall effects on tree growth for deciduous sites. Moisture and Nutrient Limitations to Tropical Forests Tropical forest formations occupy areas with limited variation in te mperature but a wide range in rainfall intensity and distribution (Wh itmore 1992), giving rise to differing degrees in deciduousness. Since tropical lowland evergr een forests are characterized by high annual rainfall and evergreeness, and often occur on highly weathered, dystrophic soils (Sanchez 1976), previous research has mainly overlooked the effects of mois ture on forest processes and, therefore, has focused on nutrient limitations a nd mechanisms of nutrien t conservation (Herrera

PAGE 15

15 et al. 1978, Jordan 1983). However, more recent studies have demonstrated that ecosystem processes in tropical forests, including Amazonian forests, can be substantially affected by strong seasonality in rainfall (e.g., Keller et al. 2004). During the dry season (monthly precipitation < 100 mm), old-growth forests in Amazonia rely on deep rooting to retain leaves (Nepstad et al. 1994). In extreme cases, prolonged droughts, usually associated with El Nio events, can result in higher tree mortality in tropical old-growth (Condit et al. 1995, Williamson et al. 2000) as well as regrowth forests (Chazdon et al. 2005), increasing forest susceptibility to fire (Nepstad et al. 1999). Observational and Manipulative Experiments to Study Moisture and Nutrient Limitations in Tropical Forests Researchers usually rely on observational studie s to infer moisture and nutrient limitations to ecosystem level processes in tropical forest s. Measurement of ecosystem processes during wet and dry seasons (Berish & Ewel 1988, Cornu et al. 1997, Dantas & Phillipson 1989, Davidson et al. 2000, Scott et al. 1992) or along rainfall gr adients (Santiago 2003, Schuur & Matson 2001) has been used to study moisture c onstraints. An importa nt limitation of studies based on rainfall seasonality is the lack of c ontrol over factors (e.g., light availability, vapor pressure deficit, and phenology) th at covary with rainfall seasona lity and may significantly affect ecosystem processes. Nutrient constraints may be investigat ed by comparing ecosystem processes among different soil types varying in nutrient availabi lity (e.g., Moran et al. 2000). This approach has the disadvantages of hindering th e identification of the most limiting nutrient(s) at a specific site because of inherent differences in soil characte ristics (e.g., soil organic matter, pH, structure, texture), as well as incomplete control over land-use history am ong sites. Such disadvantages may be overcome through fertilizer addition in nu trient manipulation studies. Substrate-age

PAGE 16

16 sequences (Vitousek & Farrington 1 997) with contrasting nutrient availability have also been used to study nutrient effects on ecosystem pr ocesses, but these sequences are spatially restricted. Although observational studies are useful for identifying general tre nds as well as key questions for further researc h, they often do not permit a pr ocess-based unde rstanding of resource control over ecosystem dynamics. An improved understanding of moisture and nutrient limitations on ecosystem processes may be obtained through manipulative experiments. Experimental manipulations of moisture availabi lity are usually carried through water addition or exclusion, whereas nutrient manipulation usuall y involves fertilizer addi tion or litter removal (Eviner et al. 2000, Hanson 2000). Long-term, large-scale moisture manipulation st udies in tropical fo rests include the dryseason irrigation study in the Barro Co lorado Island station, Panama (Cavelier et al. 1999, Wieder & Wright 1995, Wright & Corn ejo 1990, Yavitt & Wright 2001, Yavitt et al. 2004), and two throughfall exclusion studies at old-growth sites in the Br azilian Amazon (Carvalho et al. 2005, Nepstad et al. 2002). Dry-season irrigation pl ots are easier to esta blish and operate than throughfall exclusion plots, but th e latter are necessary to ultima tely simulate the effects of drought on ecosystem processes. To my knowle dge, there are no published reports for largescale, long-term moisture manipulative experiment s in tropical forest regrowth sites, except for the studies conducted within th e MANFLORA Project (Arago et al. 2005, Fortini et al. 2003, Vasconcelos et al. 2004, Veluci et al. In preparation). Fertilization experiments are the most common tools used to manipulate nutrient availability in forests (e.g., Davidson et al. 2004a, Mirmanto et al. 1999, Tanner et al. 1998) because they are relatively easy and inexpensive (Eviner et al. 2000). In addition, those

PAGE 17

17 experiments are believed to provide the most conc lusive evidences of nutrient limitation in forest ecosystems (Raich et al. 1994). However the interpretation of fertilization experiments may be confounded by several interactions of nutrients with microorga nisms and the soil (nutrient immobilization by litter microbes, adsorption, trace gas losses, vol atilization, and leaching) that reduce the pool of added nutri ents for the plants (Eviner et al. 2000) or that cause secondary effects such as nutrient imbalance or soil pH alteration (Marschner 1995). Litter removal is another technique for manipulat ing nutrient availability in forests. This technique avoids the problems related to nutri ent addition experiments as discussed above, but litter removal has additional and unavoidable di sadvantages including in direct effects on soil moisture and temperature variation due to th e lack of insulation by aboveground litter (Sayer 2005). Also, soil compaction due to trampling (in the case of litter removal by raking) and raindrop impact occurs in litte r removal plots. Such effects may alter soil microorganism activity and, ultimately, affect soil nutrient av ailability, representing, therefore, a potential confounding factor. Microbial activity may be furt her influenced by reduced input of labile C (Cleveland et al. 2002) with litter removal. To my knowledge there is no large-scale, long -term experimental manipulation of nutrient availability through fertilization in lowland old-growth tropical forests, but there are some reports for forest regrowth, in cluding a short-term experiment in Costa Rica (Harcombe 1977) and three other experiments at Am azonian sites. Nutrient additi on in Amazonian forest regrowth has been conducted within both short-term, smallplot (Uhl 1987) and re latively long-term, large-plot conditions (Davidson et al. 2004a, Gehring et al. 1999); however, there are no reports of litter removal studies in this region, excep t for the study conducted within the MANFLORA Project (Vasconcelos et al. 2004). Thus far, there are only two large-scale, long-term litter

PAGE 18

18 manipulation studies in tropical fo rests besides the MANFLORA Proj ect. The first study is part of the Gigante Litter Manipul ation Project (GLiMP) in Pana ma (Sayer 2005) and included monthly litter raking from 45 m x 45 m plots. The other study ex cluded litter with tents on 3 m x 3 m plots in a secondary forest in the Luquillo Experimental Forest (Puerto Rico) as part of the Soil Organic Matter Dynamics Project (Li et al. 2005). Moisture Effects on Ecosystem Processes in Tropical Forests Improved knowledge of the mechanisms by which tropical forest s respond to drought stressboth at the plant and commu nity levelis crucial to cu rrent understanding and future projections of forest dynamics, C sequestration, and fire susceptib ility in the context of ongoing land-use and climate changes. Anticipated c limate change for the Amazon region may include more frequent and severe dry seasons in res ponse to global warming (IPCC 2001), deforestation (Costa & Foley 2000), and more frequent El Nio episodes (Trenberth & Hoar 1997). Large scientific research in itiatives in the Amazon region sin ce the 1980s including the ABRACOS Project (Gash et al. 1996) and more recently the LBA Pr ogram (Davidson & Artaxo 2004, Keller et al. 2004) have generated a great deal of relevant informati on. However, there are few observational or manipulative studies aimed at investigating moisture controls on Amazonian forest regrowth. The understanding of how low soil moisture availa bility controls tropic al forest ecosystem processes is not straightforwar d because drought has many direct and indirect effects on plant and soil organisms (Figure 1-1). Below I review so me effects of low soil moisture availability on aboveand belowground processes; forest fl oor decomposition is included in aboveground processes for the purposes of this review.

PAGE 19

19 Aboveground processes Low soil moisture availability can affect aboveground C fluxes in tropical forests in various direct and indirect inte rrelated ways, mainly through mo isture effects on carbon dioxide (CO2) assimilation in photosynthesis. The effects of low soil moisture availability on leaf-level photosynthesis have been the subject of many recent reviews (e.g., Chaves et al. 2003, Lawlor 2002). The interacting direct effects of drought st ress on C assimilation include (a) decrease in stomatal conductance in response to low soil mois ture supply or high va por pressure deficit, leading to reduced CO2 assimilation (Flexas & Medrano 2002, Lawlor 2002, Malhi et al. 1998, Mulkey & Wright 1996), and (b) impairme nt of photosynthetic machinery (Chaves et al. 2003, Malhi et al. 1998). Drought may also affect C assimila tion through indirect effects: (a) xylem cavitation during dry periods reduces hydraulic conductivity constraining stomatal conductance and CO2 assimilation (Brodribb et al. 2002, Hubbard et al. 2001), (b) CO2 assimilation decreases in response to phenological change s that reduce leaf area (Malhi et al. 1998), and (c) low soil moisture decreases nutrient availabilityeither directly through reduci ng nutrient solubility, and/or indirectly through crea ting less favorable conditions for the microbial activity that is responsible for the decomposition of organic matter and release of nutrients in the soil (Cornejo et al. 1994, Malhi et al. 1998)reducing leaf nutrients and limiting CO2 assimilation. Some studies in the Amazon region have reported a decrease in CO2 uptake at the leaf level during the dry season for unde rstory forest species (Arago et al. 2005, Fortini et al. 2003), but similar data for overstory species are scarce. Induced drought in thro ughfall exclusion plots reduced canopy leaf-level CO2 assimilation for some tree species in an old-growth forest in Amazonia (Nepstad et al. 2002), consistent with a moisture limitation on leaf gas exchange.

PAGE 20

20 Comparable studies in Amazonian forest regrowth are lacking, indicating the need for more studies of moisture limitations on canopy leaf gas exchange for these forests. Drought effects at the leaf level may reflect pr ocesses at the individual tree and ultimately ecosystem (or stand-) levels in tropical forests (F igure 1-1). In old-growth forest sites in the Brazilian Amazon, higher stem di ameter growth rates (a compone nt of aboveground net primary productivity, ANPP) are associated with wetter periods (Higuchi et al. 2003, Rice et al. 2004, Vieira et al. 2004). Comparable data for regrowth sites are scarce in part because most published studies of forest regrowth in the tropics rely on one single in ventory campaign in stands of different ages to represent a successi onal chronosequence (e.g., Saldarriaga et al. 1988), resulting in few available data for comparison between periods with different moisture availability for the same stand. In Costa Rican secondary rain forest s, the mortality of trees (diameter at breast height 10 cm) increased significantly with lower dryseason rainfall, but not with total annual rainfall (Chazdon et al. 2005), suggesting that tropical fo rest regrowth may be extremely sensitive to rainfall seasonality. Carbon assimilation at the stand level can al so be influenced by phenological changes associated with water stress or a weakly d eciduous strategy adopted by some tropical trees (Malhi et al. 1998). In fact, higher litte rfall rates during the dry peri od in tropical forests have been reported in many studies (e .g., Dantas & Phillipson 1989, Scott et al. 1992, Wieder & Wright 1995), but irrigation during the dry season in a tropical forest in Panama did not affect the quantity or timing of litterfal l (Wieder & Wright 1995). Higher litt erfall rates associated with the dry season may be triggered by an increase in vapor pressure deficits (Wright & Cornejo 1990), a decrease in cloud cover and soil nutrient availability (Eamus & Prior 2001), or may reflect a genetic trait (Goulden et al. 2004). An important issue to consider is th e production of

PAGE 21

21 different leaf phenotypes associated with ra infall seasonality as reported by Kitajima et al. (1997) in a Panamanian seasonal dry forest. Kitajima et al (1997) found that leaves produced in the late wet season (measured in the dry seas on) had higher photosynthe tic rates than those produced in the early wet and measured in the wet season. Eddy covariance measurements of net CO2 exchangethe balance between gross primary productivity and ecosystem respiration (Roy & Saugier 2001)have shown different responses of net CO2 assimilation during dry periods in the Brazilian Amazon. Malhi et al. (1998) reported reduced net CO2 uptake during the dry season in centr al Amazonia. However, Saleska et al. (2003) and Goulden et al. (2004) found higher rates of net C sequestration during the dry season in an old-growth forest in east-central Amazoni a, probably because drought reduced forest floor decomposition, but not canopy photosynthesis. Thus litter decomposition apparently plays an important role in defining the direction of ch ange in net ecosystem exchange due to drought stress. Litterbag and mass balance studies have shown that low moisture availability reduces litter decomposition in tr opical forests (Cornejo et al. 1994, Cornu et al. 1997, Luizo & Schubart 1987, Wieder & Wright 1995), probabl y as a result of lower leaching and/or decomposer activity during dry periods. Modeling studies have also predicted drought constraint s on C dynamics of tropical forests. Several ecosystem modeling studies (Asner et al. 2000, Foley et al. 2002, Phillips et al. 1998, Potter et al. 2001, Potter et al. 2004, Prentice & Lloyd 1998, Tian et al. 1998) have analyzed the effects of recent El Nio-Southern Oscillation (ENSO) events on the C balance of Amazonian forests. All of these studies have indicated that the basin is a source of CO2 (negative net ecosystem productivity, NEP) during El Nio events and a sink of CO2 (positive NEP) during La Nia events. Changes in C bala nce due to ENSO events in these models are

PAGE 22

22 largely driven through changes in net primary productivity (NPP), and not through alterations in heterotrophic respiration (Foley et al. 2002, Tian et al. 2000). During El Nio years, lower precipitation and higher temperatures result in increased simulated annual drought stress that limits NPP. In contrast, the dry period may represent an opportunity for C gain due to increased light availability associated with reduced cloudiness. During typical, non-ENSO rainfall years, Huete et al. (2006) found widespread greening in the dry season for central and eastern Amazonian oldgrowth forests, suggesting that sunlight may represent a stronger control over forest phenology and productivity than moisture availability. Cons istent with a light lim itation to forest phenology and productivity, Graham et al. (2003) reported increas ed photosynthesis, vegetative growth, and reproduction for branches of a tropical tree s upplied with extra il lumination during cloudy periods in Panama. Further research on the cont rols of water and light over tropical forest functioning is needed to better comprehend the response of these forests to climate change. Belowground processes Belowground processes are constr ained by drought through the effects of low soil moisture availability on root and micr obial dynamics. Root growth declines under low soil water potential, as shown for a te mperate oak forest (Joslin et al. 2001), but root elongation may be stimulated by dry conditions (Akmal & Hirasawa 2004) if plants allocate a larger fraction of photosynthate to belowground biomass in response to drought. In tropical forests, fine root production decreases and mortality may in crease during the dry season as shown by observational (Berish & Ewel 1988) and manipula tive studies in old-growth sites (Cattnio et al. 2002, Cavelier et al. 1999, Yavitt & Wright 2001). Microbial activity is also constrained by low soil moisture availability in tropical forest soils (Cleveland et al. 2002, Luizao et al. 1992). Decreased root and/or microbial activities in the mineral soil and/ or aboveground litter are likely

PAGE 23

23 causes of reduced soil CO2 efflux during the dry season (Davidson et al. 2000, Vasconcelos et al. 2004). Drought may also have an indi rect effect on belowground pro cesses if reducti on in leaf C assimilation under low soil moisture conditions re sults in decreased export of photosynthates to roots. Such a reduced export may decrease the availability of C for root and rhizosphere microorganism activity (Hgberg et al. 2001). However, the negative impact of low soil moisture on belowground processes can be m itigated if hydraulically lifted water makes a significant contribution to delayi ng soil dry-down in tropical fo rests. This phenomenon would allow microorganisms to remain active for l onger periods (Horton & Hart 1998), therefore leading to an increase in nutri ent mineralization. Da Rocha et al. (2004) suggested that the lack of drought stress in an eastern Am azonian old-growth forest was pr obably related to deep rooting and water redistribution by hydrauli c lift. In the context of th e Tapajs Throughfall Exclusion Experiment, Romero-Saltos et al. (2005) did not find evidence for hydraulically lifted water by understory/midcanopy tree species using deuteriumlabeled soil water prof iles, while Oliveira et al. (2005) showed strong eviden ce for the occurrence of hydraulic redistribution based on the dynamics of peaks of water recharge between sh allow and deep soil layers, and sap flow data measured in tap and lateral roots. Nutrient Effects on Ecosystem Processes in Tropical Forests In the tropics, many soils are highly weathe red and consequently dystrophic (Sanchez 1976), which has led most past research to focu s on nutrient cycling and assume that nutrient availability limits tropical lo wland evergreen forest productiv ity (Vitousek 1984). However, evidence for such a constraint is rather limited. Malhi et al. (2004) reported that spatial variability of coarse wood productivi ty of neotropical forests was a pparently associated with soil fertility. Manipulative experiments involving nut rient addition are necess ary to show limitation

PAGE 24

24 by a specific nutrient (Tanner et al. 1998), but such experiments ar e scarce in tropical lowland forests. Most studies rely on soil nutrient inventories, aboveground biomass accumulation (Zarin et al. 2001), leaf and litter fall nutrient concentrations (Vitousek 1984, Wood et al. 2005), root growth responses (Cuevas & Medina 1988) and structural properties (Herrera et al. 1978) to infer nutrient limitation to tropical lowland forest processes. Below I re view some effects of nutrient availability on aboveand belowground processes. Aboveground processes Many essential mineral elements are directly or indirectly involved in plant tissue growth (Marschner 1995), but a key aspect of the relation between nutrient availability and plant growth and function is the positive corr elation between maximum net phot osynthesis and leaf nitrogen (N) concentration (e.g., Lambers et al. 1998). This relationship is a consequence of the high investment of leaf nitrogen in the enzyme responsible for carboxylati on (ribulose bisphosphate carboxylase, Rubisco) and in othe r photosynthetic enzymes (Chapin et al. 2002, Taiz & Zeiger 1998). However, the significance and form of the relationship between maximum net photosynthesis and leaf nitrog en concentration may depend upon the importance of other limiting nutrients including phosphorus (P) as reported for Amazonian tree species (Reich et al. 1994). Phosphorus is often hypothesized to be th e most limiting nutrient in old-growth and regrowth lowland tropical forests. Analyzing within-stand nutrient use efficiency and nutrient return in litterfall, Vitousek (1984) suggested th at P, but not N availab ility, constrains fine litterfall (an important co mponent of ANPP) in lowland tropical forests, especially at Amazonian sites. Davidson et al. (2004a), however, report ed N co-limitation to tree growth at a forest regrowth site subjected to several cycles of slash-and-burn in the Brazilian Amazon, and associated the limitation by N with substantial lo sses of this element through burning. Also, in

PAGE 25

25 an old-growth lowland evergreen forest in Indonesia, Mirmanto et al. (1999) reported increased fine litterfall in plots fert ilized with N, P, and N+P. Aboveground biomass in old-growth forests in Central Amazonia has been negatively correlated with soil sand content (Laurance et al. 1999), probably due to the low capacity of nutrient retention by sandy soils. However, sin ce sandy soils also have low moisture retention capacity (Brady 1989), moisture limita tion may also have contribute d to the results obtained by Laurance et al. (1999). Belowground processes Nutrient availability affects belowg round processes by altering root and soil microorganism activities. Root responses to low nutrient availability may not be straightforward. The increased allocation of resources to belowgro und structures may be associated with low soil fertility (Giardina et al. 2004, Gower 1987). However, higher proliferation of fine roots in fertilized ingrowth cores in tropical forests suggests that root grow th is limited by low soil nutrient availability (Cue vas & Medina 1988, Mcgrath et al. 2001, Ostertag 1998, Raich et al. 1994), although fine root growth did not respond to likely reduced nutrient availability in litter removal plots in an old-growth forest in Panama (Sayer et al. 2006). For a Panamanian lowland tropical forest, Cavelier et al. (1999) suggested that control of fine root production may be more complex, involving not only nutrient pulses, bu t also water pulses and aboveground biomass growth. Soil microbial activity is constrained unde r low soil nutrient conditions. Cleveland et al. (2002, 2003) have shown increased microbial resp iration with phosphorus addition to tropical forest soil samples in a laboratory experi ment. Cleveland and Townsend (2006) reported increased in situ soil respiration with phosphor us and nitrogen fertiliza tion in an old-growth

PAGE 26

26 forest in Costa Rica; these authors suggested that phosphorus increased microbial respiration while nitrogen probably affected soil resp iration through effects on fine root dynamics. Production of roots on top of the mineral so il has been considered as a structural characteristic of forests grow ing on low nutrient soils (Herrera et al. 1978). However a recent study has shown significant root growth in the fore st floor of a relatively fertile old-growth site in Panama, suggesting that prolif eration of roots on top of the mineral soil is not necessarily caused by low mineral soil nutrient levels, but may result from the availability of aboveground litter (Sayer et al. 2006). Conclusions Research on moisture and nutrien t constraints to tropical forest regrowth is rather limited in quantity, and results are sometimes divergen t from one study to the next. Manipulative studies to investigate soil and pl ant processes in tropical forest regrowth are lacking, and are an important tool for exploring the complex intera ctions that influence ecosystem response to resource limitations. Furthermore, these st udies are needed to better understand present conditions and to project future impacts of c limate and land-use changes on C dynamics in tropical forest regrowth.

PAGE 27

27 Figure 1-1. Simplified conceptual diagram of likely effects of drought on leafand ecosystemlevel processes addresse d in this dissertation. and symbols stand for increase and decrease, respectively; ANPP aboveg round net primary productivity. Numbers refer to some study evidences of drought effects on processes: 1Eamus (2003); 2Sperry (2000); 3Cleveland et al. (2002); 4Lawlor and Cornic (2002); 5Brodribb et al. (2002), Hubbard et al. (2001); 6Firestone and Davidson (1989);7Chapin et al. (2002); 8Lawlor and Cornic (2002); 9Rascher (2004); 10Nepstad et al. (2002); 11,12Hgberg et al. (2001); 13Davidson et al. (2000), Vasconcelos et al. (2004). drought soil moisture root water u p take1 xylem cavitation2 soil microbial activity3 stomatal conductance5 metabolic constraint4 nitrogen mineralization6 leaf nitrogen7 photosynthesis8 leaf area9 stem growth10 ANPP12 soil res p iration13 transport of C to belowground structures11 root g rowth10

PAGE 28

28 CHAPTER 2 STUDY SITE AND EXPERIMENTAL DESIGN Study Site This study was conducted at a fi eld station belonging to the Fe deral Rural University of Amazonia ( Universidade Federal Rural da Amaznia UFRA), Brazil, n ear the city of Castanhal (1 19 S, 47 57 W) in the state of Par. Since July 2001, daily rainfall was measured 500 m away from the experimental area using a standard rain gauge. Prior to July 2001, rainfall data reported here are from th e National Agency of Electrical Energy ( Agncia Nacional de Energia Eltrica ANEEL) network meteorologi cal station at Castanhal (1o 17' 53" S, 47o 56' 56" W) located ~3 km from our site, but no longer in operation. From 70 to 90% of annual rainfall occurs between January and Jul y, resulting in a dry period from August to December (Figure 2-1). Annual rainfall during the experimental period (Table 2-1) was consistent with the mean standard error value regist ered from 1990 to 1999 by ANEEL (2461 271 mm). The number of dry months (rainfall < 100 mm month-1) during the experimental period varied from 2 to 5; several authors (e.g., Vieira et al. 2004) consider dry season months as those with less than 100 mm ra infall for tropical sites. The soils are classified as Dystrophi c Yellow Latosol Stony Phase I (Tenrio et al. 1999) in the Brazilian Classification, corresponding to Sombriustox in U.S. Soil Taxonomy. Soil granulometric composition in the first 20 cm is 20% clay, 74% sand, and 6% silt. Concretions represent 16% of the soil volume in the upper 10 cm of soil. In the surface soil (0 10 cm), pH is 5.0, organic C is 2.2%, or ganic C stock is 2.9 kg m-2, total N is 0.15%, C:N is 14.4, and Mehlich-1 extractable phosphorus is 1.58 mg kg-1 (Rangel-Vasconcelos 2002). This level of extractable soil phosphorus suggests low availability at our study site compared to other soil types and land uses in Amazonia (Mcgrath et al. 2001).

PAGE 29

29 Forest regrowth, annual crops, and active and de graded pastures characterize the landscape surrounding the field station. The stand under study was last abandoned in 1987 following multiple cycles of shifting cultivation, beginni ng in the 1940s when the old-growth forest was cleared. Each cycle of 1 to 2 years included cultivati on of corn, manioc, and beans, followed by fallow. Typical shifting cultivati on cycles lasted seven to ten years (G. Silva e Souza & O.L. Oliveira pers. comm.). Trees are mo stly evergreen, with few species (e.g., Annona paludosa and Rollinia exsucca ) showing deciduousness during the dr y season. 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 November 1999, mean se stem density was 213 19.7 individuals per 100 m2, basal area was 13 6 m2 ha-1, height was 4.9 0.4 m for the stand (Coelho et al. 2004), and aboveground biomass was 51.1 2.5 Mg ha-1 for trees with diameter at breast height > 1 cm. Experimental Design Plots were established in August 1999, when th e forest regrowth was 12 years old. Each treatment plot is 20 m x 20 m with a centrally nested 10 m x 10 m measurement subplot. The area between the measurement subplot and the pl othereafter called outer areawas used for some destructive samplings of soil, root, and abov eground litter. There were four replicate plots for the irrigation treatment, four plots for the litter removal treatment, and four plots left untreated as controls (Figure 22). Adjacent treatment plots were spaced 10 m from each other. One tensiometer (Jet Fill Tensiometers, Soilm oisture Equipment Corp., Santa Barbara, CA, USA) was installed at a depth of 10 cm in each plot and soil water potential was recorded on a weekly basis in the morning. The number of actua l replicates per treatmen t varied due to loss of

PAGE 30

30 water column tension during the dry season. Soil suction vari ation in response to rainfall seasonality and manipulation treatmen ts is presented in Chapter 5. Irrigation was applied at a rate of 5 mm day-1, for about 30 minutes, during the dry seasons of 2001 to 2005 (Table 2-2) in the late afte rnoon. The amount of daily irrigation applied corresponds to regional estimates of daily evapotranspiration (Jipp et al. 1998, Lean et al. 1996, Shuttleworth et al. 1984). Irrigation water was distributed through tape s with microholes every 15 cm. In 2001, irrigation tapes were spaced 4 m fr om each other. In the subsequent irrigation periods we reduced the distance between tapes to 2 m to facilitate more even distribution of water. We used rainfall and soil suction data to de fine approximate boundaries for the dry and wet seasons. The start of the dry season was defi ned by total rainfall less than 150 mm in the previous 30 days and soil suction more negativ e than -0.010 MPa; the en d of the dry season was defined by total rainfall greater than 150 mm in the previous 30 days 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 lo cation of seasonal boundaries is about 7 days. The lowest tension value registered was -0.092 MPa, which may reflect the limited functional range of tensiometers (Hanson 2000), although lower tensions may have occurred towards the end of dry season. The installation of tensiometers deeper in the soil and of time domain reflectrometer sensors for measurement of soil moisture content were hindered by the shallow depth to laterite in the soil profile in Ape. In the litter removal plots, leaf and branch fa ll 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 = 8); carbon and nitrogen stocks of the pretreatment litter layer were

PAGE 31

31 222.9 14.6 and 7.3 0.5 g m-2, respectively (n = 8). Raking maintained very low, but not entirely absent litter standing crop. Total new non-woody litterfall remove d during the treatment period (from August 2001 to December 2005) was 3568 136 g m-2 (n = 12). Carbon and nitrogen concentrations of pre-treatment litterfall were 47.9 0.2 and 1.2 0.02%, respectively, corresponding to a C:N ratio of 40 0.7 (n = 12). Measurements of gravimetric soil moisture cont ent in the first 10 cm of soil for one date during the 2001 dry season indicate d that irrigated plot s had about twice as much moisture as control plots (22 2% vs. 10 2%); in the litter removal plots soil moisture was 11 2%. For one date during the 2001 wet season, grav itational soil moisture content was 27 2% for control and irrigated plots, and 31 2% for litter removal plots (Rangel-Vasconcelos 2002). The difference in soil moisture status between contro l and irrigated plots was reflected in dry-season differences in pre-dawn leaf water potential for an unde rstory species ( Miconia ciliata ); in November 2001 pre-dawn leaf water potential for control plants was about -1.2 MPa while irrigated plants were about 1 MPa less negative (Fortini et al. 2003).

PAGE 32

32 Table 2-1. Characteristics of ra infall distribution and intensity during the experimental period in the site. Year 1999 2000 2001 2002 2003 2004 2005 Annual rainfall (mm) 2577 2399 3179 2301 2895 3038 2793 Minimum monthly rainfall (mm) NAb 66 34 56 42 8 13 Maximum monthly rainfall (mm) NA 291 489 385 499 611 476 Number of dry season months a NA 3 5 4 2c 3 3c Total dry season rainfall (mm) d NA 694 304 400 647 445 615 a Rainfall < 100 mm month-1. b NA Not available. c Not consecutive months. d Dry season period = August to December. Table 2-2. Dry-season irrigation intervals and associated rain fall intensity and distribution. Dry-season irrigation Interval Total rainfall (mm) Maximum daily rainfall (mm) Number of days without rainfall 1st 10 Aug 2001 to 16 Jan 2002 453 54 101 (63%)a 2nd 16 Aug 2002 to 20 Jan 2003 516 66 93 (59%) 3rd 7 Aug 2003 to 20 Dec 2003 559 74 90 (66%) 4th 23 Sep 2004 to 26 Jan 2005 547 130 81 (64%) 5th 29 Jul 2005 to 12 Dec 2005 422 66 97 (71%) a Percentage of days without rainfa ll during dry-season irrigation period.

PAGE 33

33 Date Aug Dec Apr Aug Dec Apr Aug Dec Apr Aug Dec Apr Aug Dec Apr Aug Dec Apr Aug Dec Rainfall (mm d-1) 0 20 40 60 80 100 120 140 160 2000 19992001200220042005 2003 Figure 2-1. Daily rainfall duri ng the experimental period (data prior to July 2001 are from a meteorological station about 3 km away from the study site). Figure 2-2. Experimental plot layout showing the arrangement of treatments.

PAGE 34

34 CHAPTER 3 SEASONAL AND EXPERIMENTAL EFFE CTS ON LITTERFALL QUANTITY AND QUALITY IN EASTERN AMAZONIAN FOREST REGROWTH Introduction Litterfall represents the major process of nutrient transfer from aboveground forest vegetation to soils (Vitousek & Sanford 1986), and fine litterfall co mprises a significant fraction of aboveground net primary produc tivity in forests (Clark et al. 2001b). Litter nitrogen and phosphorus cycling are of particular importance since these nutrients usually are the most limiting for tropical forest productivity (Vitousek 1984). Low phosphorus avai lability is likely a common constraint for tropical fo rest regrowth, and nitrogen li mitation appears significant for forests reestablishing after severa l episodes of slash-and-burn, whic h lead to substantial losses of nitrogen through volat ilization (Davidson et al. 2004a, Gehring et al. 1999). Litterfall quantity usually shows distinct pattern s associated with rainfall seasonality, i.e., litterfall peaks during dry season (e .g., Wieder & Wright 1995). Howe ver, a direct effect of soil moisture availability on litterf all quantity and timing has not been demonstrated (Cavelier et al. 1999, Wieder & Wright 1995). The concentration of nutrients in leaf litte rfall may also vary with rainfall seasonality in tropical forests (Wood et al. 2005), but a 5-year irrigation experiment in Panama did not affect litterf all nutrient concentration (Yavitt et al. 2004). Litterfall production has been shown to be limited by nu trient availability (Vitousek 1984), with fertilization resulting in higher litterfall rates in a dry tropical forest in Mexico (Campo & Vzquez-Yanes 2004) and a wet tropi cal forest in Puerto Rico (Li et al. 2006). Fertilization also results in increased leaf litter nutrient concentration in tropical forests (Li et al. 2006). A better understanding of fluxes and pools of carbon, nitrogen, and phosphorus involved in litterfall can help to improve m odels of forest biogeochemistry. More appropriate quantification of the role of soil moisture and nutrients in the regulation of litterfall can facilitate predictions of

PAGE 35

35 carbon and nutrient dynamics under different conditions of resource availability. In this context, long-term (> 1 year) data on litterfall quantity and quality are equally important to understand interannual variability effects on carbon and nutrient dynamics, but such information is scarce for tropical forests. The primary objective of this chapter was to inve stigate the effects of moisture and nutrient availability on litterfall within the context of the dry-seas on irrigation and litter removal experiments described in Chap ter 2. We hypothesized that (a ) dry-season irrigation would increase non-woody litterfall qua ntity and quality, and (b) litter removal would reduce nonwoody litterfall quantity and quality. Material and methods Litterfall From October 1999 to December 2005, litterfall wa s collected weekly in each of three 1 m x 1 m screen litter traps in th e 10 m x 10 m measurement subplots. The weekly frequency of litterfall collection was chosen to minimize mass and nutrient losses due to leaching of trapped litter (Luizao 1989). The plant mate rial collected in each trap was air-dried in the laboratory to remove excess moisture before storage. At 4-w eek intervals, material from the same collector was composited and then separated into woody and non-woody fractions. Leaves and their petioles, foliar rachises, and reproductive parts were included in non-woody litterfall. Our nonwoody fraction corresponds to the fine litter (or s mall litter) fraction defined in several studies (e.g., Smith et al. 1998), except for the non-inclusion of woody material. In fine litter, small-diameter woody materialusu ally <1-2 cm diameter (Clark et al. 2001a, Proctor 1983) is included assuming that this woody fraction (1) has turnover times comparable to other components of non-woody material (mostly fo liar and reproductive ma terial) and (2) may

PAGE 36

36 represent material produced from the current years growth. Thus, our estimate of non-woody material may represent a slight underestimation of fine litter. We weighed woody and non-woody litt erfall after drying at 60-70 C until constant weight. Litterfall data for April 2003 was lost du e to a malfunction of th e oven that resulted in burning of litterfall samples; for this period, we used for each trap a value of litterfall estimated from the mean relative contribu tion of April to annual litter fall per trap as follows: MC 100 AL MC Est where: Est = estimated litterfall for April 2003 (in g m-2); MC = mean relative contribution of April to annual litterfa ll in 2000, 2001, 2002, and 2004 (in %), i.e., 100 n ) litterfall annual litterfall (April MCn 1 i where i = year; and AL = total 2003 litterfall except April (in g m-2). Mean se MC was 6.3 0.2% for all traps over four years. Composite samples of non-woody litterfall we re ground with a coffee grinder (Krups, US) and stored in 60 mL scintillation vials for subseq uent analysis of carbon (C), nitrogen (N), and phosphorus (P). Carbon was determined with an elemental carbon analyz er (Carlo Erba model CNS2500) at the School of Forest Resources a nd Conservation (Univers ity of Florida) in samples collected from October 1999 to Marc h 2001. We estimated that non-woody litterfall was 48% C based on the monthly non-w oody litterfall C con centration (47.9 0.2%, n = 18).

PAGE 37

37 Nitrogen and phosphorus concentrations were determined in the Laboratory of Plant Ecophysiology and Propagation at Embrapa Amaznia Oriental (Brazil) in samples collected from January 2000 to December 2004. The Kjeldahl digestion was used to determine total nitrogen (Anderson & Ingram 1996). Phos phorus concentrations were determined colorimetrically after digestion of 0.1 g sample in sulfuric acid and peroxide (Murphy & Riley 1962). Following the criteria in Boone et al. (1999), all the samples were analyzed in duplicate for P, while 10% of the samples were randomly selected for duplicate analyses for N. Mean coefficient of variation in duplicate analyses was 2.1% for N (n = 542) and 4.1% for P (n = 2096). Percent error in relation to standard re ference material (peach leaves, NIST SRM 1547) was -14 1.6% for N (n = 22) and 2.0 1.0% for P (n = 24). To calculate N and P fluxes in non-w oody litterfall (nutrient return), nutrient concentrations were multiplied by mass for each trap per month. Litter Stock At the end of the 2004 wet season (25-Augus t) and dry season (29-December), we collected samples (n = 4) of forest floor litter from randomly chosen areas (25 cm x 50 cm) in each of the control and ir rigated plots and processed as for li tterfall. Non-woody litter stock was calculated by dividing the amount of dr y material per collection area (g m-2). Statistical Analysis We used SAS version 9.00 to run the statis tical analyses. We analyzed with PROC MIXED the effects of treatment, date, and trea tment-by-date interaction on the variables nonwoody litterfall mass (monthly and annual), and nitrogen and phosphorus concentration and return using a repeated measures analysis with compound symmetric covariance structure. This structure assumes constant variance at all date s and equal correlations between all pairs of measures on the same experimental unit, i.e., litterfall trap for the litterfall variables and plot for

PAGE 38

38 litter stock. We ran separate te sts to compare each of the treatmen ts with the control. Within this analysis, significant treatment effects would ha ve indicated temporally consistent differences between treatment and control measurements bot h preand post-treatment and across seasons, significant date effects were generally indicative of seasonal trends that affected both treatment and control measurements, and treatment-by-date effects indicated a significant difference between treatment and control measurements that occurred after the treatment was initiated. Thus, the treatment-by-date effect represents the be st test of treatment effect when there were no pre-existing differences among plot s prior to the treatment. We used a priori CONTRAST statements to explicitly test whether the m easured variables differed between seasons and between treatments within each season (wet and dry). When necessary, we performed log and square root transformations to meet the model assumptions of normality, based on the criteria of P > 0.05 in the Kolmogorov-Smirnov test, and equal variances, based on the absence of a pattern of heteroscedasticity in the plots of residual versus predicted values. Means and standard errors were calculated on the basis of untransformed data. All results are reported as significant when P 0.05; we report marginal significance when 0.05 < P < 0.10. Multiple comparisons of means were performed with Tukeys test (P < 0.05). Results Non-woody Litterfall Irrigation experiment Non-woody litterfall mass was si gnificantly affected by date and the intera ction between treatment and date (Table 3-1, Figure 3-1B). Th e significant effect of the interaction was not associated with consistent differences between treatments during the pre-treatment period (P = 0.76) or within dry-season irri gation periods (P = 0.18). N on-woody litterfall was significantly

PAGE 39

39 higher in the dry season than in the wet season (79.5 1.3 and 58.5 0.9 g m-2 month-1, respectively; P < 0.0001). Annual non-woody litterf all mass was significantly affected by the interaction between treatment and date (Tab le 3-1); in 2003, annual non-woody litterfall mass (Figure 3-2A) in irrigated pl ots was significantly higher th an in control plots (899.2 55.3 and 742.4 63.1 g m-2 year-1, respectively; P < 0.01). Annual li tterfall in the control plots was not correlated with annual rainfall (r = 0.129, P = 0.808, Pearson correlation). Non-woody litterfall N concentration was significa ntly affected by date only (Table 3-1, Figure 3-3B). The effect of date was not relate d to a significant seasonal influence on litterfall N concentration (dry = 1.24 0.01 vs. wet = 1.27 0.01% N, P = 0.86). The input of N in non-woody litterfall was sign ificantly affected by date and treatment x date interaction; there was no significant effect of treatment (Table 3-1, Figure 3-4B). The significant effect of the interaction was not rela ted to a consistent difference between treatments within dry-season irrigation periods (P = 0.19). Non-woody litterfall P concentration was signifi cantly affected by date and treatment x date interaction (Table 3-1, Figure 3-3C). L itterfall P concentration was significantly higher in control plots than in irrigated plots for some months during earl yto mid-dry season (November 2001 and September 2002) and late-dry to early-we t seasons (January and February in 2002 and 2003). Litterfall P concentration wa s significantly lower in the dry season than in the wet season (0.38 <0.01 and 0.40 <0.01 mg P g-1, respectively; P < 0.0001), although the difference was slight. Phosphorus return in non-woody litterfall was significantly affected by date and treatment x date interaction (Table 3-1, Figure 3-4C). Treatment differen ces within dry-season irrigation

PAGE 40

40 periods were marginally significant (P = 0.08), largely due to differen ces during the 2003 dryseason irrigation period. Annual return of N and P were significantly affected by date and treatment x date interaction (Table 3-1); irrigation plots showed significantly higher N and P return than control plots in 2003 (Figure 3-5). Litter removal experiment Non-woody litterfall mass was signi ficantly affected only by date (Table 3-1, Figure 3-6B) and was significantly higher in the dr y season than in th e wet season (76.2 1.2 and 58.4 0.9 g m-2 month-1, respectively; P < 0.0001). Annual nonwoody litterfall mass was significantly affected by date only (Table 3-1), with the 2001 mean litterfall rates significantly higher than subsequent years, but not diffe rent from 2000 (Figure 3-2B). Non-woody litterfall N concentration was signifi cantly affected by treatment, date, and treatment x date interaction (Table 3-1, Figur e 3-7B). During the treatment period, mean litterfall N concentration was about 12% higher fo r control plots than fo r litter removal plots (1.26 0.01 and 1.13 0.01% N, respectively; P = 0.01). This difference was not homogenous throughout the manipulation period; with the progression of litter removal, the difference between treatments in annual N concentration increased from ~ 11% in 2002 to ~ 16% in 2004, which correspond to values of ~ 5% (2002) and ~ 11% (2004) after accoun ting for pretreatment differences. There was also a significant effect of treatment during the pretreatment period (P = 0.03); however, pretreatment diff erences between plots did not a ffect the significance (P = 0.04) of post-treatment differences (contrast test). The return of N in non-woody litterfall was sign ificantly affected by date and treatment x date interaction (Table 3-1, Figure 3-8B). Howeve r, the contrast test showed that the significant

PAGE 41

41 effect of the interaction did not reflect consiste nt differences between treatments during the litter removal period (P = 0.36). Non-woody litterfall P concentration was signi ficantly affected by date only, with a marginally significant effect of treatment (Table 3-1, Figure 3-7C). Phosphorus concentration during the wet season was sligh tly but significantly higher than during the dry season (0.40 <0.01 and 0.36 <0.01 mg P g-1, respectively; P < 0.0001). The return of P in litterfall was significantly affected by date and treatment x date interaction (Table 3-1, Figure 3-8C). The si gnificant effect of the interaction term was associated with occasionally higher values for co ntrol plots. Phosphorus return in the dry season was slightly, but significantly hi gher than in the wet season (0.027 <0.001 and 0.022 <0.001 g P m-2, respectively; P < 0.0001). Annual return of N was signifi cantly affected by date and tr eatment x date interaction (Table 3-1). However, there was no detectable significant or marginally significant difference between treatment means in each year (P > 0.10, Tukey test), although 2001 values were generally higher than other years, and control plots tended to ha ve higher N return than litter removal plots in 2002 (Figure 3-9A). Annual retu rn of P was significantly affected by date only (Table 3-1), with substantially higher return rates in 2003 a nd 2004 than in the other years (Figure 3-9B). Litter stock The stock of non-woody litter (Figure 3-10) was significantly higher towards the end of the dry season (December 2004) than at the e nd of the wet season (August 2004) (680 54 and 435 36 g m-2, n = 8, respectively; P < 0.001). There we re no significant effects of treatment (P = 0.203) or treatment x date interaction (P = 0.271).

PAGE 42

42 Discussion Seasonal Patterns Non-woody litterfall rates measur ed in this study are within the range reported for both regrowth and old-growth Amazonian and other tropical forests elsewhere (Table 3-2). The higher rates of litterfall during the dry season compared to the wet season are also consistent with other studies in tropical forests (Dantas & Phillipson 1989, Sanchez & Alvarez-Sanchez 1995, Scott et al. 1992, Smith et al. 1998, Wieder & Wright 1995). The magnitude of interannual variability over 6 years varied fr om 9% for irrigated plots to 16% for litter removal plots, lower than that reported for a Panamanian old-grow th forest (38%) (Wieder & Wright 1995). Annual litterfall was not related to annual rainfall, sugges ting that litterfall product ion is not controlled by rainfall intensity for this regrowth forest stand. However, Lawrence (2005) found a positive relationship between annual litter fall and annual rainfall for tropical seasonal forests at a global scale. There were no detectable effects of rainfa ll seasonality on litterfall N concentration, although Yavitt et al. (2004) reported higher N concentration in leaf fall during the wet season for a Panamanian old-growth forest, and Wood et al. (2005) reported a wet s eason decline in leaf litterfall N concentration for a Costa Rican old-growth forest. Non-woody litterfall P concentration was lowe r during the dry season than in the wet season in the present study, with some lower va lues of litterfall P associated with peaks in litterfall, and some higher valu es of P occurring during lower litterfall rates in the wet season. These results for litterfall P are consistent with data reported for a secondary dry tropical forest in Mexico (Read & Lawrence 2003) and an old-growth forest in Costa Rica (Wood et al. 2005). Most annual litterfall P peaks o ccurred during the first 1-2 mont hs of the wet season, when rapid decomposition of litter accumu lated during the dry season could have supplied a pulse of

PAGE 43

43 nutrients to plants with the onset of rainfall (Lodge et al. 1994, Wood et al. 2005). Lower litterfall P concentration in irrigated plots during dry-wet season transitions (2001-02 and 200203), associated with the strongest dry-season irrigation periods, are consistent with the pulse hypothesis, i.e., irrigation could have prevente d litter accumulation and, therefore, nutrient mineralization pulse with the onset of rainfall. However, increased soil P availability for both control and irrigation plots duri ng wet-up events in the 2004 dry s eason is not consistent with irrigation effects on nu trient pulse (Veluci et al. In preparation). The lack of litter removal effects on litterfall P peak further suggests that the pulse hypothesis may not be applicable. Alternatively, the seasonal and treatment effect s on litterfall P may be caused by differences in P resorption between treatments, and/or differences in the contribution of P-rich, reproductive litterfall (flowers and fruits) during dry-wet transitions. Repr oductive litterfall has been shown to have higher P concentration than l eaf litterfall for tropical forests (Scott et al. 1992, Zagt 1997), and to peak (number of seeds m-2) during dry-wet season transi tions for our experimental site, although no irrigation effects have been obs erved in two consecutive evaluation years (Dias 2006). Litter stock measured in this study is within the range reporte d for tropical forests (Table 3-2). Increased litter stock in the dry season is consistent with higher litterfall and lower decomposition rates during this peri od at the study site (Chapter 4), as also reported for an oldgrowth forest in Panama (Wieder & Wright 1995). Limited Impact of Dry-season Irrigation Irrigation did not impact litterfall rates in the dry season, except for higher rates in irrigated plots for a few dates, mostly in the 2003 dry-se ason. These results are consistent with those found for a dry-season irrigation experiment in a semideciduous lowland forest in Panama (Cavelier et al. 1999, Wieder & Wright 1995), and further co nfirm that soil moisture availability

PAGE 44

44 may not trigger increased litterfall during the dr y season in tropical forests. Higher dry-season litterfall rates may be linked to increased vapor pressure deficits (Wright & Cornejo 1990), decreased cloud cover, decreased so il nutrient availability (Eamus & Prior 2001), or variation in temperature (Breitsprecher & Bethel 1990). Although the exact trigger(s) of increased dryseason litterfall have not been already ascertained, it is very likely that tropical trees respond to more than one cue (Wright & Cornejo 1990). Dry-season irrigation did not al ter N and had only small effects on P concentrations in nonwoody litterfall, consistent with the results from a water manipulation study in a Panamanian old-growth forest (Yavitt et al. 2004). The small impacts of dr y-season irrigation in this study contrasts with the potential for increased N and P availability in irri gated plots due to the combination of (1) higher N and P inputs in litte rfall during the dry season and (2) higher litter decomposition in irrigated plots (Chapter 4). Th us, these results suggest that low litter quality indicated by the high C:N and lignin:N ratios of leaf litter (Chapter 4) and non-woody litterfall may be a stronger control over N (as well as P) ava ilability than soil water status at this site, favoring microbial immobilizati on of nutrients; Aerts (1997) s uggest that litter chemistry (especially the lignin:N ratio) represents the most important determinant of decomposition rates in tropical regions. Furthermore, consistent with results from an irrigation study in Panama (Yavitt & Wright 1996), dry-seas on irrigation had no influence on so il net nitrific ation rates at our site (Vasconcelos et al. 2004). We expected that long-term irrigation would have resu lted in increased aboveground productivity and, consequently, higher non-w oody litterfall ratesan index of ANPP (Clark et al. 2001a, Jordan 1983). However, after 5 years of dry-season irrigation, this effect has not occurred consistently. Higher litterfall rates did occur for the irrigated plots in 2003, but that was

PAGE 45

45 in the year with the weakest dry season over th e whole experimental period, when we would have expected the least effect of dry-season irrigation on forest processes. However, increased annual litterfall in irrigated plot s in 2003 may have resulted from a lag effect of the extended drought in the preceding dry season (Table 2-1), consistent with results for a temperate mixed deciduous forest (Newman et al. 2006). Nonetheless, the eff ects of the extended 2002 dry season were not sufficiently intense to affect tree mortality at the community level which actually decreased from 2002 to 2003, and remained constant in 2004 (Arajo et al. 2005). Litter Removal Reduces Litterfall N Concentration Nitrogen and phosphorus concentrations and inputs in litterfall are comparable to values reported for forests of the Brazilia n Amazon and elsewhere in the tropics (Table 3-2). Increased differences in N concentrations between control and litter removal plots ar e consistent with the recognized role of nutrient cycling in litter as a significan t source of N for tropical forest plants (Markewitz et al. 2004, Vitousek & Sanford 1986). Mean litterfall P concentration for the control plots in this forest regrowth stand (0. 04%) coincides with the value proposed by Vitousek (1984) to distinguish between high and low P levels for tropical forests. For most months from 2000 to 2003, litterfall P concentrations were belo w this threshold, which may reflect the low availability of soil phosphorus, as suggested by the low soil extr actable P reported for the site (Rangel-Vasconcelos 2002). The lack of treatment effects on litterfall P concentration may be explained by sufficient supply of P from soil sources. While weathering processes are not likely a substantial source of P in highly weathered tropical soils deprived of primary P-containing minerals (Sanchez 1976), mineralization of P from soil organic matter may represent a significant source of this nutrient for plants, even after 40 months of bi-weekly li tter removal. Recent studies have determined substantial amounts of labile orga nic-P fractions (NaOHand NaHCO3-extractable) for

PAGE 46

46 Amazonian forest regrowth sites in Brazil (Frizano et al. 2003, Markewitz et al. 2004), and a simulation study concluded that N and P stored in (deeply buried) soil organic matter can sustain C accumulation rates under conditions of limited i nput of such nutrients in tropical forest regrowth (Herbert et al. 2003). In addition, some regrowth fo rest trees colonizing sites with low soil P availability probably present mechanisms to improve P acquisition such as mycorrhizal associations and high phosphata se exudation rates (Marschner 1995). Uhl (1987) hypothesized that high incidence of mycorrhizal infection and efficient uptake and nutrient use may be necessary for establishment of successional trees under the limiting nutrient conditions typical of abandoned lands after slash-and-burning in the tropics. Similarly, Gehring et al. (1999) suggested that the growth of two early successional tree species in an Amazonian forest site was not limited by soil P availability because of efficient mycorrhizal associations. Since litter is the main source of most nutrients in tropical forests (Markewitz et al. 2004, Vitousek & Sanford 1986), we expected that chro nic litter removal woul d have resulted in nutrient deficiency, and consequently reduced ANPP (Harrington et al. 2001). Thus, we hypothesized that non-woody lit terfall rates would diminish for litter removal plots. This study thus far indicates that the qua ntity of non-woody litterfall was in sensitive to the reduction in nutrient availability (indicated by reduced litter N concentration) imposed by the litter manipulation treatment, consistent with the resu lts obtained by Sayer (2005) for a 2-yr litter removal study in Panama. It is possible that extending the litter removal period will further reduce nutrient concentrations in litter, leading to a critical point where productivity will be significantly constrained. Nutr ient manipulation effects on ecos ystem processes are usually not immediate, and litter removal st udies may have slower effect s on litterfall responses than fertilization studies (Campo & Vzquez-Yanes 2004, Mirmanto et al. 1999).

PAGE 47

47Table 3-1. F statistics a nd associated significance levelsa for the effects of treatments (irrigation and litter removal), sampling date, and their interaction on non-woody litterfall mass and nutrients in a tropical regrowth forest stand in eastern Amazonia, Brazil. Irrigation Experiment Litter Removal Experiment Non-woody litterfall Treatment Date Treatment x Date Treatment Date Treatment x Date Mass 0.72 ns 36.76*** 1.66*** 0.22ns 25.21*** 1.34* N concentration 0.18 ns 27.87*** 1.00ns 8.42** 29.22*** 2.53*** N return 0.49 ns 31.94*** 1.54*** 0.65ns 26.71*** 1.40* P concentration 0.95ns 52.11*** 2.56*** 3.74ns 53.08*** 1.23ns P return 0.30ns 27.37*** 1.91*** 0.39ns 20.20*** 1.36* Annual mass 0.71ns 1.78ns 3.46** 0.24ns 4.59*** 0.74ns Annual N return 0.82ns 24.47*** 3.66** 0.48ns 24.89*** 2.67* Annual P return 0.63ns 79.92*** 3.88** 0.15ns 31.99*** 1.45ns aThe level of significance is indicated (*: P < 0.05 **: P < 0.01, ***: P < 0.001, ns: not significant).

PAGE 48

48Table 3-2. Estimates of annual non-woody litterfall (mass, n itrogen, and phosphorus), non-woody lit ter stock, and litterfall:f orest floor mass ratio (kL) in tropical forests. Forest descriptiona Location Annual rainfall (mm) Annual litterfall (g m-2 yr-1)b Non-woody litter (g m-2) kL Annual nitrogen litterfall (g m-2 yr-1) Annual phosphorus litterfall (g m-2 yr-1) Sourcec Regrowth 3 Brazil 2600 504 1 12 Puerto Rico 3810 820 500 1.64 2 10 Brazil 2433 690 880 3 19 Brazil 1800 890 10.4 0.28 4 3 18 Brazil 1940 1040 1300 5 30 Brazil 2830 630 6 15 Brazil 2760 783 613 1.25 9.8 0.30 7 Old-growth Brazil 2100 640 8.63 8 Brazil 2600 804 1 Venezuela 3565 1025 12.1 0.21 9 Brazil 2100 825 15.1 0.31 10 Brazil 2300 928 463 2.01 11.8 0.67 11 Panama 2600 1240 12 Brazil 1900 970 720 1.34 11.5 13 Indonesia 3600 710 5.84 0.16 14 Brazil 2000 564 840 15 Brazil 1800 900 16 Brazil 2000 570 920 17 Brazil 2200 890 600 1.5 18

PAGE 49

49Table 3-2. (continued) Forest descriptiona Location Annual rainfall (mm) Annual litterfall (g m-2 yr-1)b Non-woody litter (g m-2) kL Annual nitrogen litterfall (g m-2 yr-1) Annual phosphorus litterfall (g m-2 yr-1) Sourcec Brazil 1800 1030 14.3 0.33 4 Brazil 1940 1100 1740 5 Brazil 2000 1380 1460 19 a For the forest regrowth sites, age (years after abandonment) is presented. Regrowth includes sites classified as secondary fo rests, while oldgrowth refers to primary and mature forest sites. b Litterfall dry mass estimated as two times litterfall carbon for the studies without direct report of mass. c For each source number, details of coordinates, soil type, and authors of each study are presented below: 1 1 44' S, 47 9' W, Capito Poo, Brazil, unspecified soil, Dantas and Phillipson (1989) 2 18 19' N, 65 49' W, Puerto Rico, Ultisol, Cuevas et al. (1991) 3 2 25' S, 59 50' W, Manaus, Brazil, Oxisol, Mesquita et al. (1998) 4 2 59 S, 47 31 W, Paragomin as, Brazil, Haplustox, Markewitz et al. (2004) 5 Southwestern Amazonia, Brazil, Ultiso ls with patches of Oxisols, Salimon et al. (2004) 6 1o 18' 6'' S, 48o 26' 35'' W, Belm, Brazil, Yellow Latosol, Oliveira (2005) 7 1 19 S, 47 57 W, Ape, Brazil, Distrophic Yellow Latosol, this study 8 2o 34 S, 60o 7 W, Manaus, Brazil, Yellow Lato sol, Klinge and Rodrigues (1968) 9 1 54 N, 67 3 W, San Carlos de Rio Negro, Oxisol, Cuevas and Medina (1986) 10 2o 34 S, 60o 7 W, Manaus, Brazil, Yellow Latosol, Luizao (1989) 11 Marac Island, Brazil, Scott et al. (1992) 12 9 o 09 N, 79 o 51 W, Barro Colorado Island, Panama, Alfisol, Wieder and Wright(1995) 13 Curu-Una Forest Reserve, Brazil, Oxisol, Smith et al. (1998) 14 0 o 6 S, 113 o 56 E, Kalimantan, Indonesia, Yellow sandy soil, Mirmanto et al. (1999) 15 Tapajs National Forest, Brazil, Ultisols and oxisols, Silver et al. (2000) 16 2 o 59 S, 47 o 31 W, Paragominas, Brazil, Haplustox, Davidson et al. (2000) 17 2.897o S, 54.952o W, Tapajs National Forest, Haplustox, Nepstad et al. (2002) 18 2 o 3521.08 S, 60o 06 53.63 W, Manaus, Brazil, Oxisol, Luizao et al. (2004) 19 2o 64S, 54o 59W, Tapajs National Forest, Brazil, Oxisols and Ultisols, Silver et al. (2005)

PAGE 50

50 Date 09/99 11/99 01/00 03/00 05/00 07/00 09/00 11/00 01/01 03/01 05/01 07/01 09/01 11/01 01/02 03/02 05/02 07/02 09/02 11/02 01/03 03/03 05/03 07/03 09/03 11/03 01/04 03/04 05/04 07/04 09/04 11/04 01/05 03/05 05/05 07/05 09/05 11/05 01/06 Non-woody litterfall (g m-2) 0 20 40 60 80 100 120 140 160 control irrigation B Monthly rainfall (mm) 0 100 200 300 400 500 600 700 800 A Figure 3-1. Effects of rain fall patterns and dry-season irri gation on non-woody litterfall in an Amazonian forest regrowth stand, Brazil. A) Monthly rainfall. B) Monthly nonwoody litterfall for control and irrigation plots. In Figure 3-1B, each symbol represents the mean standard error, n = 12. Vertical gray bars indicate the irrigation periods. White and black horizontal bars ma rk the dry and wet seasons, respectively.

PAGE 51

51 Annual non-woody litterfall (g m-2 yr-1) 0 200 400 600 800 1000 control irrigation A Year 200020012002200320042005 0 200 400 600 800 control litter removal B* Figure 3-2. Effects of dry-s eason irrigation and litter removal on annual non-woody litterfall in an Amazonian forest regrowth stand, Brazil. A) Non-woody litte rfall for control and irrigation plots. B) Non-woody litterfall fo r control and litter re moval plots. Each symbol represents the mean standard e rror, n = 12. Treatments began in August 2001. Asterisk indicates significan t treatment difference (P < 0.05).

PAGE 52

52 control irrigation Date 12/99 02/00 04/00 06/00 08/00 10/00 12/00 02/01 04/01 06/01 08/01 10/01 12/01 02/02 04/02 06/02 08/02 10/02 12/02 02/03 04/03 06/03 08/03 10/03 12/03 02/04 04/04 06/04 08/04 10/04 12/04 Non-woody litterfall P (mg g-1) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Non-woody litterfall N (%) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Monthly rainfall (mm) 0 200 400 600 800 A B C Figure 3-3. Effects of rain fall patterns and dry-season irri gation on non-woody litterfall nutrient concentrations in an Amazonian forest regr owth stand, Brazil. A) Monthly rainfall. B) Nitrogen (N) concentrati on for control and irrigation plots. C) Phosphorus (P) concentration for control and irrigation pl ots. In Figures 3-3B-C, each symbol represents the mean standard error, n = 12. Vertical gray bars indicate the irrigation periods. White and black horizontal bars ma rk the dry and wet seasons, respectively.

PAGE 53

53 Date 12/99 02/00 04/00 06/00 08/00 10/00 12/00 02/01 04/01 06/01 08/01 10/01 12/01 02/02 04/02 06/02 08/02 10/02 12/02 02/03 04/03 06/03 08/03 10/03 12/03 02/04 04/04 06/04 08/04 10/04 12/04 Non-woody litterfall P return (g m-2) 0.00 0.01 0.02 0.03 0.04 0.05 0.06 Non-woody litterfall N return (g m-2) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Monthly rainfall (mm) 0 200 400 600 800 A B C control irrigation Figure 3-4. Effects of rain fall patterns and dry-season irri gation on non-woody litterfall nutrient return in an Amazonian forest regrowth stand, Brazil. A) Monthly rainfall. B) Nitrogen (N) return for contro l and irrigation plots. C) Phosphorus (P) return for control and irrigation plots. In Figures 3-4B-C, each symbol represents the mean standard error, n = 12. Ver tical gray bars indicate the irrigation periods. White and black horizontal bars mark the dr y and wet seasons, respectively.

PAGE 54

54 Annual non-woody litterfall N return (g m-2 yr-1) 0 2 4 6 8 10 12 14 Year 20002001200220032004Annual non-woody litterfall P return (g m-2 yr-1) 0.0 0.1 0.2 0.3 0.4 control irrigation A B Figure 3-5. Effects of dry-s eason irrigation on annual non-woody litterfall nutrient return in an Amazonian forest regrowth stand, Brazil. A) Nitrogen (N) return for control and irrigation plots. B) Phosphorus (P) return for control and irri gation plots. Each symbol represents the mean standard e rror, n = 12. Treatments began in August 2001.

PAGE 55

55 Date 09/99 11/99 01/00 03/00 05/00 07/00 09/00 11/00 01/01 03/01 05/01 07/01 09/01 11/01 01/02 03/02 05/02 07/02 09/02 11/02 01/03 03/03 05/03 07/03 09/03 11/03 01/04 03/04 05/04 07/04 09/04 11/04 01/05 03/05 05/05 07/05 09/05 11/05 01/06 Non-woody litterfall (g m-2) 0 20 40 60 80 100 120 140 160 control litter removal B Monthly rainfall (mm) 0 100 200 300 400 500 600 700 800 A Figure 3-6. Effects of ra infall patterns and li tter removal on non-woody litterfall in an Amazonian forest regrowth stand, Brazil. A) Monthly rainfall. B) Monthly nonwoody litterfall for control and litter remova l plots. In Figure 3-6B, each symbol represents the mean standard error, n = 12. The vertical line indicates the beginning of the litter removal treatment. White and black horizontal bars mark the dry and wet seasons, respectively.

PAGE 56

56 Non-woody litterfall N (%) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 B CDate 12/99 02/00 04/00 06/00 08/00 10/00 12/00 02/01 04/01 06/01 08/01 10/01 12/01 02/02 04/02 06/02 08/02 10/02 12/02 02/03 04/03 06/03 08/03 10/03 12/03 02/04 04/04 06/04 08/04 10/04 12/04 Non-woody litterfall P (mg g-1) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 control litter removal Monthly rainfall (mm) 0 200 400 600 800 A Figure 3-7. Effects of rain fall patterns and litter remova l on non-woody litterfall nutrient concentrations in an Amazonian forest regr owth stand, Brazil. A) Monthly rainfall. B) Nitrogen (N) concentration for control and litter removal plots. C) Phosphorus (P) concentration for control and litter removal plots. In Figures 3-7B-C, each symbol represents the mean standard error, n = 12. The vertical line indicates the beginning of the litter removal treatment. White and black horizontal bars mark the dry and wet seasons, respectively.

PAGE 57

57 Non-woody litterfall N return (g m-2) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 B CDate 12/99 02/00 04/00 06/00 08/00 10/00 12/00 02/01 04/01 06/01 08/01 10/01 12/01 02/02 04/02 06/02 08/02 10/02 12/02 02/03 04/03 06/03 08/03 10/03 12/03 02/04 04/04 06/04 08/04 10/04 12/04 Non-woody litterfall P return (g m-2) 0.00 0.01 0.02 0.03 0.04 0.05 0.06 control litter removal Monthly rainfall (mm) 0 200 400 600 800 A Figure 3-8. Effects of rainfa ll patterns and li tter removal on non-woody litter fall nutrient return in an Amazonian forest regrow th stand, Brazil. A) Monthl y rainfall. B) Nitrogen (N) return for control and litter removal plots. C) Phosphorus (P) return for control and litter removal plots. In Figures 3-8B-C, each symbol represents the mean standard error, n = 12. The vertical line indicates the beginning of the litter removal treatment. White and black horizontal bars mark th e dry and wet seasons, respectively.

PAGE 58

58 Annual non-woody litterfall N return (g m-2 yr-1) 0 2 4 6 8 10 12 14 Year 20002001200220032004Annual non-woody litterfall P return (g m-2 yr-1) 0.0 0.1 0.2 0.3 control litter removal A B Figure 3-9. Effects of litter removal on annua l non-woody litterfall nut rient return in an Amazonian forest regrowth stand, Brazil. A) Nitrogen (N) return for control and litter removal plots. (B) Phosphorus (P) return fo r control and litter removal plots. Each symbol represents the mean standard e rror, n = 12. Treatments began in August 2001.

PAGE 59

59 Date Aug-04Dec-04 Non-woody litter stock (g m-2) 0 200 400 600 800 1000 control irrigation Figure 3-10. Non-woody litter stock for control and irrigation plots in an Amazonian forest regrowth stand, Brazil. Each bar repres ents the mean standard error, n = 4.

PAGE 60

60 CHAPTER 4 LEAF DECOMPOSITION IN A DRY SEASON IRRIGATION EXPERIMENT IN EASTERN AMAZONIAN FOREST REGROWTH Introduction Litter decomposition is a major nutrient cycling process in terrestrial ecosystems and is particularly important for forest ecosystems on low fertility soils, including many tropical forests (Golley 1983, Swift et al. 1979). The rate of litter decompositi on is controlled by interactions of litter quality, environmental cond itions, and soil organisms (Swift et al. 1979). Litter quality is defined as the amount and types of organic ca rbon compounds, nutrient concentrations, and ratios between carbon compounds and nutrients in lit ter; low-quality litter (e.g., low nutrient, high carbon compounds:nutrient ratio) usually shows lower decomposition rates than highquality litter (e.g., high nutrient, lo w carbon compounds:nutrient ratio) (Loranger et al. 2002, Mesquita et al. 1998, Songwe et al. 1995). For moist tropical fore sts moisture and temperature are assumed to be non-limiting, and litter qua lity is thought to be the dominant control on decomposition rates (Aerts 1997), a lthough seasonal drought (Cornejo et al. 1994) and excessive moisture (Schuur 2001) may retard decompositi on at some tropical forest sites. Water manipulation studies may help to clarify seas onal drought effects on litter decomposition. A considerable number of studies have investig ated decomposition respon ses of leaf litter from several plant species to rainfall s easonality in tropical forests (e.g., Cornu et al. 1997, Cuevas & Medina 1988, Luizo & Schubart 1987), but there are few such data for tropical regrowth sites (Mesquita et al. 1998). The effects of water manipulation on decomposition rates have been examined for old-growth tropical forests in Brazil (Nepstad et al. 2002) and Panama (Cornejo et al. 1994, Wieder & Wright 1995), but related st udies are lacking for regrowth stands. Studies to quantify decomposition rates and their contro ls can help to improve understanding of carbon and nutrient cycling in tropi cal forest regrowth sites.

PAGE 61

61 The primary objective of this chapter was to inve stigate the effects of moisture availability on litter decomposition within the context of the dry-season irrigation expe riment described in Chapter 2. We hypothesized that (a) decompositi on rates would be faster under higher moisture availability in the wet season a nd during dry-season irrigation peri ods in the treatment plots, and (2) decomposition rates would be faster fo r species with higher quality leaves. Study Site and Experimental Design Study site and experimental desi gn are described in Chapter 2. Material and Methods Leaf Litter Decomposition The litterbag method (Harmon et al. 1999) was used to study l eaf litter decomposition. This method is the most frequently employed fo r examining litter decomposition in terrestrial ecosystems (Wieder & Lang 1982), although it has se veral limitations that can significantly influence decomposition rates including alteration of litter microclimate and exclusion of certain decomposer organisms (but see Prescott 2005). However, the litterbag method is adequate for studies comparing species, sites, and the effect s of experimental manipulations on decomposition (Heal et al. 1997, Wieder & Lang 1982). Traps were put outside the treatment plots to collect leaves for the decomposition study. Fresh fallen leaves of Lacistema pubescens Mart., Ocotea guianensis Aubl., Stryphnodendron pulcherrimum (Willd) Hochr., and Annona paludosa Aubl. were collected every week for 3-4 months prior to installing each of 3 separate decomposition experiments (described below). Collected leaves were dried under ambient conditions and stored. L. pubescens was chosen because it is the most common sp ecies in the study area (Arajo et al. 2005) and A. paludosa, O. guianensis, and S. pulcherrimum were selected because they re present a wide range in leaf texture; O. guianensis leaves are thick and appears to be recalcitrant, while A. paludosa and S.

PAGE 62

62 pulcherrimum possess leaves that appear to decompose more rapidly. After the collection period, 6-8 subsamples of 10 g each were oven dried at 65-70 oC until constant weight and the dry weight conversion factor (a ir-dry mass:oven-dry mass) was cal culated. These subsamples were processed following procedures described for litterfall (Chapter 3) to determine the initial leaf litter chemical compositi on (described below). For the second and third decomposition experiments, S. pulcherrimum was replaced by Vismia guianensis (Aubl.) Choisy because we noticed that some leaflets of th e former species were smaller than the openings in the litterbag screen, which could overestimate decomposition rates. V. guianensis is another common pioneer species in the study area (Arajo et al. 2005). Bags of polypropylene with openings of 1 mm x 0.8 mm and measuring 20 cm x 20 cm received about 10 g of air-dried material of only one species. In each of the control and irrigation plots, 18 litterbags of each species were randomly placed in the surface of the litter layer. After 30, 60, 120, 180, 240, and 360 days (Experiments 1 and 2) and 13, 31, 45, 61, and 90 days (Experiment 3), three bags of each species were retrieved in each plot. After retrieval, litterbags were air dried to facilitate the remova l of adhering soil particle s and roots gently using forceps and small, soft brushes (Tigre, medium size, Brazil) Then, the material was oven dried and weighed to calculate remaining leaf mass. Samples for the last collection in Experiment 2 were discarded because it was not possible to se parate out soil particle s from leaf material. To investigate the effects of dry-season irrigation on the remaining mass of leaf litter under different stages of decomposition, the experi ments had different in stallation and duration periods. Experiments 1 and 2 lasted 12 months and were installed in the beginning of the 2002 wet (February 7) and 2003 dry (July 27) seasons, resp ectively, in order to determine the effect of seasonality on initial and later st ages of decomposition. For an improved temporal resolution of

PAGE 63

63 dry-season irrigation effects on de composition, Experiment 3 was carried out exclusively during three months in the 2004 dry season (September 24 to December 23), with more frequent sampling during that period than for Experiments 1 and 2. To assess seasonal effects on leaf litter decomposition irrespective of treatment, we compared the remaining leaf mass of control plots at 60 days in Experiment 1 and at 61 days in Experiment 2, which corresponded to wet and dry s easons, respectively. Total rainfall in these wet and dry seasons was 1311 and 359 mm, respectivel y. For this analysis we used data for A. paludosa, L. pubescens, and O. guianensis. Remaining leaf mass (percen t) was calculated as 1000 X Xt, where tX is the dry litter mass at the time t and 0X is the initial dry litter mass. Initial Leaf Litter Chemistry Phosphorus concentrations were determined colorimetrically afte r digestion of 0.1 g sample in sulfuric acid and peroxide (M urphy & Riley 1962). Carbon and nitrogen were determined with an automated dry combusti on instrument (LECO Model CNS-2000). Lignin and cellulose were determined by a sequential di gestion of fibres (Anderson & Ingram 1996). Specific Leaf Area The specific leaf area (SLA) was measured in individuals of Annona paludosa (n = 3), Lacistema pubescences (n = 4), Ocotea guianensis (n = 4), and Vismia guianensis (n = 4) located in the control plots. In each tree, three young, fully expanded leav es were chosen from different branches, and three discs (1.11 cm2) were collected from each leaf. The discs were dried at 65 oC for 48 h and individually wei ghed to the nearest 0.0001 g.

PAGE 64

64 Statistical Analysis We used SAS System version 9.00 to run the statistical analyses. Decomposition rates (k) were calculated by fitting the obser ved data (i.e., remaining leaf mass) to the single exponential model proposed by Olson (1963) using the PROC NLIN procedure. In the single exponential model, Xt = X0ek t; where Xt and X0 are the litter mass at the times t and 0 (initial), respectively, and k is the decomposition rate (yr-1). Although this model makes unrealistic assumptions (e.g., treats litter as a uniform, ho mogeneous substance) regardi ng the decomposition of litter, k values calculated with this model are useful for interpreting short-term (first year decomposition), comparative experiments (Paustian et al. 1997) such those in this study. For these analyses, data (XtX0) from three litterbags were averaged per plot for each sampling date because we considered individual plots as the experimental units, resulting in n = 4 for each combination of species, treatment, and sampling date. The eff ects of species, treatments, and the species x treatment interaction on k values were analyzed with a tw o-way ANOVA using PROC ANOVA. PROC CORR was used to anal yze the correlation between k and initial litter quality parameters. The effects of species on initial litter ch emistry were analyzed with one-way ANOVA using PROC ANOVA. The TTES T procedure was used to compare seasonal effects on remaining leaf mass for c ontrol plots. The statistical analys es were carried out using the mean SLA calculated for each leaf per species. Means a nd standard errors were calculated on the basis of untransformed data. All results are reported as si gnificant when P 0.05. Multiple comparisons of means were pe rformed with Tukeys test. Results Dry-season irrigation had no signi ficant effects (Table 4-1) on leaf litter decomposition rates (k) obtained by fitting curves to all of the co llection data in the tw elve-month experiments (Table 4-2, Figure 4-1). Ho wever, in Experiment 2, k values obtained from curve fitting to the

PAGE 65

65 dry season data only were significantly higher in irrigated than in control plots (1.04 0.06 and 0.86 0.06 yr-1, respectively; P < 0.01) (Table 4-1). In the three-month dry-season experiment, A. paludosa showed significantly higher decompos ition rates than the other species under irrigation, which did not differ significantly among them (Table 4-3), and within species, decomposition rates were significantly higher in irrigated plots than in control plots (Table 4-3). All of the experiments showed si gnificant effects of species on k (Table 4-1). Overall, A. paludosa showed the highest decom position rates (Table 4-2, 43). There were significant differences (P < 0.0001) in specific leaf area, carbon, nitrogen, phosphorus lignin, and cellulose concentrations, lignin:N ratio, and C:N ratio among species (Table 4-4), but there were no significant correlations between k and leaf quality parameters (Table 4-5, Figure 4-2). The analysis of seasonal effects on decompos ition showed that remaining leaf mass was significantly (P < 0.001) higher in the dry season than in the wet season for L. pubescens (87.6 0.9 and 76.0 0.9%, respectively) and O. guianensis (88.5 1.1 and 77.5 0.9%, respectively); there was a marginally significant effect (P < 0.052) of season on A. paludosa remaining leaf mass (dry = 80.4 3.1% vs. wet = and 71.8 1.7%). Discussion Decomposition rates measured in this study ar e within the range reported for tropical forests (Table 4-6). Decompos ition was faster during the wet s eason than the dry season, as observed in other studies in trop ical forests in Amazonia (Cornu et al. 1997, Luizo & Schubart 1987) and elsewhere (Cornejo et al. 1994, Wieder & Wright 1995). Moisture constraints on decomposition were further confirmed by higher ma ss loss rates in dry-se ason irrigated plots, except when irrigation was applied to litter previously exposed to field conditions for 180 days; in this case, the greater proporti on of recalcitrant compounds in a dvanced stages of litter decay

PAGE 66

66 (Swift et al. 1979, Wieder & Lang 1982) probably confe rred less susceptibility to decomposition in response to increased moisture availability. Although leaf decomposition is significantly constrained during the dry season, the greatest difference between mass loss in control and irrigated plots was 10 to 13% only, and between dry and wet seasons was 7 to 12% only. Such small differences could be due to exclusion of macrofauna activity in leaf decomposition in the 1 mm x 0.8 mm opening bags. Using 1-mm mesh litterbags with additional ope nings of about 10 mm, Luizo and Schubart (1987) suggested that surface fine root penetr ation and macroarthropod activity determined the great difference in leaf mass loss between the dry a nd wet seasons for 3-yr-old forest regrowth in central Amazonia. It is not likely that fine root colonization has b een constrained in our litterbags as we did observe fi ne root adhered to leaves. Ho wever, the 1 mm x 0.8 mm opening bags likely restricted macroarthropo ds to access leaf material and this could have contributed to the small differences between dry and wet as we ll as control and irrigation percent leaf mass losses in this study. Leaf chemical and structural traits in this study are also consistent with other studies in tropical forests (Table 4-7). The range of li gnin concentration (42.9 to 51.7%) found for the species investigated in this regrowth forest is high in comparison to reported values for oldgrowth forest tree species in Panama (Table 4-6), but similar to results of Mesquita et al. (1998) and Vasconcelos and Laurance (2005), who fou nd lignin concentrations of about 53% for regrowth forest species in two central Amazonian sites. The lack of correlation between decomposition ra tes and leaf quality parameters may result from the reduced number of leaf lit ter species tested in this study, as also observed by Fonte and Schowalter (2004) who investigated 8 different litter species for a Puerto Rican forest. For

PAGE 67

67 tropical forests, the lignin:N ratio was found to be the strongest predictor of decomposition in a Panamanian old-growth forest (Santiago 2003), while decompositi on rates decreased with higher tannin concentration for a regrowth forest site in Amazonia (Mesquita et al. 1998). However, the strongest leaf quality predictor may change according to the stage of the decomposition process (Loranger et al. 2002). Despite the lack of correlati on between decomposition rates an d litter quality parameters, the highest decomposition rates observed in Annona paludosa are probably explained by their higher leaf quality, i.e., high c oncentrations of nitrogen and phos phorus, the lowest concentration of lignin, and thin leaves (high specific leaf area). The low decomposition rates of Ocotea guianensis and Vismia guianensis leaves are associated with low N and P concentrations, high lignin concentration, the highest C:N and lignin:N ratios (> 50), and thicker leaves (low specific leaf area). Interestingly, Lacistema pubescens was often the outlier interfering with a strong linear relationship between k and litter quality; decomposition rates of Lacistema are lower than would be predicted by regressing the data from the other species, suggesting that decomposition of Lacistema leaf litter may be strongly controlled by some litter quality parameter not determined in this study. One potential explan ation is the pubescent habit of its leaves. Overall, moisture effects on k were comparatively higher than those related to litter quality; while k was on average 2.4 times higher in irrigated than in control plots during the three-month dry-season experiment, the greatest diffe rence between species maximum/minimum k was 1.5 considering all of the experiments.

PAGE 68

68 Table 4-1. F statistics and a ssociated significance levels (in pa rentheses) for the effects of treatments (control and irrigation), species and their interacti ons on leaf litter decomposition rates in a tropical regrowth forest in eastern Amazonia, Brazil. Experiment Treatment Species Treatment x Species 1 (started in wet season) 3.66ns 15.96*** 0.64ns 2 (started in dry season; full period included in analysis) 0.91ns 10.26*** 1.70ns 2 (started in dry season; dry season only included in analysis) 10.47** 11.56*** 1.64ns 3 (started in dry season; frequent sampling) 194.92*** 10.15*** 3.20* Table 4-2. Decomposition rates (mean standard error) for overstory species in a tropical regrowth forest stand in easte rn Amazonia, Brazil (n = 8). Experiment 2 (started in dry season) Experiment 1 (started in wet season) full period included dry season only Species k (yr-1) A. paludosa 0.97 0.051 a 1.26 0.09 a 1.21 0.09 a L. pubescens 0.91 0.03 ab 1.02 0.03 bc 0.93 0.04 b O. guianensis 0.78 0.03 bc 0.85 0.04 c 0.73 0.07 b S. pulcherrimum 0.65 0.04 c V. guianensis 1.08 0.04 ab 0.91 0.07 b 1 Within a column, different le tters indicate that means differ at P < 0.05 (Tukeys test).

PAGE 69

69 Table 4-3. Decomposition rates (mean standard error) for overstory species under control and irrigated plots (Experiment 3) in a tropical regrowth forest stand in eastern Amazonia, Brazil (n = 4). This experiment started in the dry season and encompasses frequent sampling during this period. Control Irrigation Species k (yr-1) A. paludosa 0.59 0.041 Aa 1.52 0.08 Ba L. pubescens 0.49 0.06 Aa 1.14 0.05 Bb O. guianensis 0.48 0.03 Aa 1.00 0.06 Bb V. guianensis 0.39 0.03 Aa 1.02 0.14 Bb 1 Within columns and rows, different lowera nd upper-case letters, respectively, indicate that means differ at P < 0.05 (Tukeys test). Table 4-4. Initial quality parame ters (mean standard error) of l eaves incubated in litterbags for decomposition studies in a trop ical regrowth forest in Ea stern Amazonia, Brazil (n = 6-8). Annona paludosa Lacistema pubescens Ocotea guianensis Vismia guianensis Carbon (%) 50.77 0.121 a 53.32 0.17 b 52.55 0.08 c 52.65 0.07 c Nitrogen (%) 1.05 0.02 a 1.66 0.01 b 0.90 0.02 c 1.02 0.01 a Phosphorus (mg g-1) 0.45 0.01 a 0.50 0.01 b 0.32 0.01 c 0.4 0.01 d Lignin (%) 42.9 0.7 a 46.3 0.6 b 47.4 0.3 b 51.7 1.1 c Cellulose (%) 37.9 0.3 a 42.0 0.4 b 29.4 1.1 c 41.8 0.8 d Carbon : nitrogen 48.51 0.89 a 32.16 0.21 b 58.57 1.18 c 52.65 0.27 d Lignin : nitrogen 40.98 1.03 a 28.05 0.33 b 52.82 1.15 c 50.24 0.90 c Specific leaf area (cm2 g-1) 165 13 a 191 15 a 66 2 b 122 6 c 1 Within a row, different letters indicate th at means differ at P < 0.05 (Tukeys test).

PAGE 70

70 Table 4-5. Pearson correlation coeffi cients between decomposition rate (k) and initial quality parameters of leaves of overstory tree speci es incubated in litterbags (Experiment 2) in a tropical regrowth forest in Eastern Amazonia, Brazil (n = 4). Leaf litter quality parameter Pearson coefficient Nitrogen (%) 0.06206 (0.9379)1 Phosphorus (mg g-1) 0.56678 (0.4332) Lignin (%) -0.44339 (0.5566) Cellulose (%) 0.55625 (0.4437) Carbon : nitrogen -0.27987 (0.7201) Lignin : nitrogen -0.32252 (0.6775) Specific leaf area (cm2 g-1) 0.62462 (0.3754) 1 Numbers in parentheses are significance values Table 4-6. Decomposition rates estim ated from litterbag studies for some tropical forest sites. Site Decomposition rate (yr-1) Source Regrowth Manaus, Brazil 0.47 0.61 Mesquita et al. (1998) Guadeloupe, French West Indies 0.41 2.39 Loranger et al. (2002) Ape, Brazil 0.39 1.52 This study Old-growth San Carlos de Rio Negro, Venezuela 0.58 5.00 Cuevas and Medina (1988) Marac Island, Brazil 2.01 Scott et al. (1992) Southern Bakundu Reserve Forest, Cameroon 1.55 4.6 Songwe et al. (1995) Barro Colorado Island, Panama 3.2 Wieder and Wright (1995) Marac Island, Brazil 0.61 2.58 Luizo et al. (1998)

PAGE 71

71Table 4-7. Chemical composition of leaf litter for some tropical forest sites. Site Carbon (%) Nitrogen (%) Lignin (%) Celllulose (%) C:N Lignin:N SLA (cm2 g-1) Source Venezuelaa 49 57 1.12 1.71 14.2 26.3 17.3 39.4 62 77 Cuevas and Medina (1988) Panamaa 69.2 122.3 Cornejo et al. (1994) Venezuelaa 78 114c Reich et al. (1995) Brazilb 47.4 48.0 1.2 1.3 53 54 Mesquita et al. (1998) Guadeloupeb 1.1 2.5 22.8 29.5 19.2 20.9 11.7 20.7 Loranger et al. (2002) Panamaa 47.3 43.2 0.90 1.22 16.0 13.7 18.4 18.0 58.4 39.2 11.8 19.9 Santiago et al. (2003) Brazild 48.4 0.94 53.5 51.7 44.0 Vasconcelos and Laurance (2005) a old-growth forest b regrowth forest c mid successional species d mixed leaf litter of successional species

PAGE 72

72 Experiment 1 Experiment 2Experiment 3 0 20 40 60 80 100 Remaining leaf mass (%) 0 20 40 60 80 100 060120180240300 020406080100 D EF G H I KL 0 20 40 60 80 100 control irrigation AB CTime (d) 060120180240300360 0 20 40 60 80 100 J Figure 4-1. Effects of dry-season irrigation on leaf litter decomposition in a forest regrowth stand in eastern Amazonia, Brazil. Remaining leaf mass of (A, B, C) Annona paludosa, (D, E, F) Lacistema pubescens, (G, H, I) Ocotea guianensis, (J) Stryphnodendron pulcherrimum, and (K, L) Vismia guianensis. Each symbol represents the mean standard error (n = 4). White and black horizontal bars mark dry and wet seasons, respectively. Note diffe rent scales on the x-axes. Experiment 1 started in the wet season, while Experime nts 2 and 3 started in the dry season; Experiment 3 had more frequent samp ling than the othe r experiments.

PAGE 73

73 AB C DE F G Phosphorus (mg g-1) 0.30.40.50.6 0.8 0.9 1.0 1.1 1.2 1.3 1.4 Cellulose (%) 27303336394245 0.8 0.9 1.0 1.1 1.2 1.3 1.4 Lignin (%) 4042444648505254 k (yr-1) 0.8 0.9 1.0 1.1 1.2 1.3 1.4 C:N 30405060 0.8 0.9 1.0 1.1 1.2 1.3 1.4 Lignin:N 2030405060 0.8 0.9 1.0 1.1 1.2 1.3 1.4 SLA (cm2 g-1) 6090120150180210240 k (yr-1) 0.8 0.9 1.0 1.1 1.2 1.3 1.4 Annona paludosa Lacistema pubescens Ocotea guianensis Vismia guianensis Nitrogen (%) 0.60.81.01.21.41.61.8 k (yr-1) 0.8 0.9 1.0 1.1 1.2 1.3 1.4 Figure 4-2. Relation betw een decomposition rate (k) and initial leaf litter characteristics for tree species in a forest regrowth stand in easte rn Amazonia, Experiment 2. A) Nitrogen concentration. B) Phosphorus concentration. C) Carbon:nitrogen ratio. D) Lignin concentration. E) Cellulose concentration. F) Lignin:nitrogen ratio. G) Specific leaf area. Each symbol represents the mean standa rd error for the y-axis (vertical error bar) and the x-axis (horizont al error bar); n = 8 for k and n = 6-8 for litter quality.

PAGE 74

74 CHAPTER 5 MOISTURE AND SUBSTRATE AVAILABILI TY CONSTRAIN SOIL TRACE GAS FLUXES IN AN EASTERN AMAZONIAN REGROWTH FOREST Introduction Tropical forests represent an important sour ce of atmospheric greenhouse gases including carbon dioxide (CO2), nitrous oxide (N2O), and methane (CH4), along with nitric oxide (NO), a precursor to the photochemical production of tr opospheric ozone (Vitousek & Matson 1992). The production and consumption of these gases ar e strongly linked to the availability of both soil moisture and decomposable substrate. Howeve r, seasonal cycles of precipitation, litterfall, and decomposition are often confounded in ways th at limit our ability to quantify the relative importance of these interacting factors from seasonal observations of gaseous fluxes. Observational studies in tropical forests have sh own that higher soil moisture availability during the wet season usually increases soil CO2 and N2O effluxes, decreases NO efflux, and decreases CH4 consumption rates (Davidson et al. 2000, Fernandes et al. 2002, Garcia-Montiel et al. 2001, Kiese & Butterbach-Bahl 2002, Kiese et al. 2003, Verchot et al. 2000, Verchot et al. 1999). Fewer studies have evaluated the respon se of soil trace gas fluxes to experimental manipulation of soil moisture av ailability in tropical forests. In a throughfall exclusion experiment in the Tapajs Nationa l Forest, Brazil, emissions of N2O and CH4 were reduced by the exclusion of about 50% of annual throughfa ll, but no treatment effect was observed for NO or CO2 emissions (Davidson et al. 2004b). Addition of water to dry soil in short-term, smallscale field studies has resulted in increased emissions of CO2, NO, and N2O in wet (GarciaMontiel et al. 2003b, Nobre et al. 2001) and seasonally dry (Davidson et al. 1993) tropical forest soils. Regarding manipulative experiments of substrat e availability (i.e., addition or removal of aboveground litter), there are only two reports of long-term, large-sc ale field studies (Li et al.

PAGE 75

75 2004, Sayer 2005) that have assessed emissions of soil CO2 in tropical forests in addition to this study. In both studies, reduction of substrate availability through litter removal decreased soil CO2 efflux, which is consistent with several related studies in temp erate forests (Bowden et al. 1993, Jandl & Sollins 1997, Rey et al. 2002, Sulzman et al. 2005), but we encountered no published reports of litter removal effects on NO, N2O, and CH4. Measurements of soil CO2 efflux and non-woody litterfall ca n be used to estimate total belowground carbon allocation (TBCA) in forest s (Raich & Nadelhoffer 1989). For mature forests, TBCA is about two times aboveground litte rfall, while for regrowth forests, TBCA is about three times abovegr ound litterfall (Davidson et al. 2002, Raich & Nadelhoffer 1989), indicating that regrowth forest s allocate a relativel y larger proportion of C to belowground structures than mature forests (Davidson et al. 2002). Although TBCA represents the single largest flux of C in forest ecosystems aside from canopy C assimilation (Davidson et al. 2002), little is known about this flux of C in tropical forests. A better understanding of how trace gas emissions from tropical forest soils are affected by moisture and substrate availability can help to improve current biogeochemical models that predict impacts of changes in climate and land-us e practices on the atmospheric concentrations of these gases (Potter & Klooster 1998). Such data, together with more estimates of total belowground C allocation in tr opical forests are also need ed to better understand carbon dynamics in regrowth forests (Johnson et al. 2000). Few such data are available for Amazonian regrowth forests, a significant and dy namic component of forest landscapes in this region (Fearnside 1996, Zarin et al. 2001). Our primary objective in this study was to quan tify the effects of moisture and substrate availability on soil trace gas emissions in an Amazonian regrowth forest stand. In one

PAGE 76

76 experiment, dry-season moisture limitation was reduced by irrigation. In the other experiment, substrate limitation was provoked by litter rem oval. The dry-season irrigation and litter removal experiments are described in Chapter 2. Material and Methods Field Measurements Since July 2001, daily rainfall has been m easured 500 m away from the experimental area using a standard rain gauge. Prior to July 2001, rainfall data re ported here are from the National Agency of Electrical Energy (ANEEL) network meteorological station at Castanhal (01o 17' 53" S, 47o 56' 56" W) which is no longer in opera tion and that was about 3 km away from our site. One tensiometer (Jet Fill Tensiometers, So ilmoisture Equipment Corp., Santa Barbara, CA, USA) was installed at a dept h of 10 cm in each plot and so il water potential was recorded on a weekly basis in the morning. The number of actual replicates per treatment varied due to loss of water column tension during the dry season. Soil CO2 efflux was generally measured bi-weekly, beginning in March 2000, with an LI6400 portable photosynthesis system fitted with an LI-6400-09 soil CO2 flux chamber (LI-COR Inc., Lincoln, NE, USA). The chamber was fit in to circular polyvinyl ch loride (PVC) collars (11.5 cm internal diameter x 5.5 cm deep), wh ich were installed approx imately 2 cm into the soil. Each plot contained three soil collars, spaced at least 1 m apart, totaling 12 collars per treatment and sampling date. No live vegeta tion was contained within the collars. Measurements were taken between 0630 and 1100 hours. To better understand the results of CO2 flux analyses within the context of stand-level C dynamics, we also collected data on litterfall (Chapter 3). We estimated that non-woody litterfall was 48% C based on the monthly non-woody litterfall C concentration (47.9 0.2%)

PAGE 77

77 during the period of October 1999 to March 2001. Non-woody litterf all was 80 to 90% of total litterfall. Woody litterf all data are not reported here beca use of its much smaller impact on short-term trace gas emissions due to its slow turnover rate. Two additional PVC collars with 20 cm diameter and 10 cm height were installed within each plot (total of 8 collars per treatment and sampling date) and inserted approximately 2-3 cm into the soil for measurement of soil NO, N2O, and CH4 gas fluxes. During the measurements, a vented PVC cover made from the end cap of a 20-cm diameter PVC pipe was fit into the collars. On average, NO, N2O, and CH4 flux measurements were made every two months, beginning in August 1999. The flux measuremen t technique for NO used a chemiluminescence detector (Scintrex LMA-3, Scintrex Limited, Concord, ON, Canada) as described by Verchot et al. (1999). N2O and CH4 fluxes were measured by gas chromatography analyses of four syringe samples extracted from the same chambers at 10-minute intervals (Verchot et al. 2000, Verchot et al. 1999). The PVC collars used for soil tr ace gas measurements were left in place throughout the course of the experiments. To augment our understanding of the N gas fluxes (NO and N2O), we also include here results of potential soil nitrification determined with a variation of the aerobic incubation method (Hart et al. 1994). Nitrification is th e precursor to the denitrif ication process, and both processes produce NO and N2O (Firestone & Davidson 1989). For each plot, we analyzed one composite sample made of four samples collect ed at a depth of 10 cm in October 2001. We estimated net N nitrification from changes in ni trate concentrations dur ing 7-day incubation of soil. We corrected soil gravimetric moisture to 75% field capacity befo re sample incubation at about 28 oC in an incubator (Isuku FR24BS, Isuku Se isakusho Ltd., Tokyo, Japan). We did extractions of samples in 2 M potassium chloride (KCl) three days after collection in the field

PAGE 78

78 and in incubated samples. We kept samples under refrigeration (4 oC) prior to the initial extraction. We filtered extracts through What man No. 42 filter paper before analysis of nitrite/nitrate using a flow-inj ection system on a Lachat QuikCh em AE autoanalyzer (Lachat Instruments, Milwaukee, WI, USA). Prior to th e extractions, we dried subsamples of soil for 24 hours at 105 oC to determine actual moisture content. Statistical Analyses We used the SAS System version 9.00 to run the statistical analyses. We analyzed with PROC MIXED the effects of treatment, date, an d treatment-by-date interaction on the variables trace gas flux, soil water potential, and non-woody lit terfall using a repeated measures analysis with compound symmetric covariance structure. This structure a ssumes constant variance at all dates and equal correlations betw een all pairs of measures on the same experimental unit, i.e., collar, tensiometer, or trap for the soil trace gase s, soil water potential, and litterfall variables, respectively. We ran separate tests to compare each of the treatments with the control. Within this analysis, significant treatment effects w ould have indicated temporally consistent differences between treatment and control meas urements both preand post-treatment and across seasons (none were observed), significant date effects were ge nerally indicative of seasonal trends that affected both treatment and control measurements, and treatment-by-date effects indicated a significant difference between treatment and control measurements that occurred after the treatment was initiated. We used CONTRAST st atements to explicitly test whether the measured variables differed between seasons and between treatments within each season (wet and dry). We used the TTEST pro cedure to compare treatments and control means for soil nitrification. When necessary, we performed log and square root transformations to meet the model assumptions of normality, based on the criteria of P > 0.05 in the Kolmogorov-Smirnov test,

PAGE 79

79 and equal variances, based on the absence of a pa ttern of heteroscedasticity in the plots of residual versus predicted values. Means and standard errors were calculated on the basis of untransformed data. All results are reported as significant when P 0.05; we report marginal significance when 0.05 < P < 0.10. We estimated annual soil C efflux by linear interpolation between sampling dates using the EXPAND procedure. To estimate annual soil C efflux, we assumed that the variation in soil CO2 efflux with time of day was minimal as previously reported by Davidson et al. (2000) for an eastern Amazonian primary forest. We tested for interannual and between treatment differences in annual soil C effl ux and annual litterfall C values for control and irrigated plots in 2001 and 2002 using the PROC MIXED procedure. For the litter removal vs. control plot comparison of annual soil C efflux and annual litt erfall C we used the TTEST procedure for 2002 data only; we did not include the 2001 data in the litter removal vs. control comparison because the treatment regime wa s not initiated until August 2001. We estimated the relative contribution of aboveground litter to soil resp iration by subtracting litter removal soil CO2 efflux from control soil CO2 efflux. Results Irrigation Experiment Rainfall declined from mid-July to early-Ja nuary (dry season) during each year of the study (Figure 5-1A), resulting in lower soil wate r potential during this period (Figure 5-1B). The dry-season irrigation resulted in significantly (P < 0.0001) le ss negative soil water potential in control plots for most of the dates in 2001 and 2002 (Figure 5-1B). During the 2001 dry season, soil water potential was -0.052 0.003 and -0.024 0.002 MPa in control and irrigated

PAGE 80

80 plots, respectively; corresponding valu es for the 2002 dry season were -0.046 0.003 and 0.013 0.002 MPa. There was a significant effect of date and th e interaction between treatment and date on soil CO2 efflux (Table 5-1). Soil CO2 efflux for irrigated plots was significantly higher than for control plots during the dry-se ason irrigation (P < 0.0001, Figur e 5-1C). There was also a significant effect of date and the interaction between treatment and date on soil CO2 efflux for the pretreatment period (P < 0.0001); however, pr etreatment differences between plots did not affect the significance of the dry-season irriga tion effect. In the 2001 dry season irrigation period, soil CO2 efflux values were 3.91 0.13 and 5.54 0.19 mol CO2 m-2 s-1 for control and irrigated plots, respectively; corres ponding values for the 2002 dry season were 4.76 0.19 and 6.21 0.25 mol CO2 m-2 s-1. The lowest mean soil CO2 efflux rate (2.33 0.19 mol CO2 m-2 s-1), which occurred in the control treatment on 24 October 2001 (Figure 51C), coincided with a successive decrease in soil water status (to -0.084 MPa) caused by a long dry spell of 24 days without rain out of a total of 31 days, with total precipitation of only 9 mm during the 31-day period. A 93% increase in the control plot soil CO2 efflux in the subsequent m easurement coincided with an increase in soil water status (to -0.008 MPa) following two consecutive rainy days (19 and 26 mm) after the long dry spell, and immediately prior to the soil resp iration measurement; no increased soil CO2 efflux was observed for irrigate d plots. The pulse in soil CO2 efflux was then followed by a decrease in CO2 emissions associated with another dry period. Annual soil C efflux was significantly higher in 2002 than in 2001 (P < 0.0001) (Table 52). The effects of treatment and the interacti on between treatment and date were marginally significant (P < 0.07 and P < 0.10, respectivel y). Annual litterfall C was not affected by

PAGE 81

81 treatment or year (Table 5-2); although the interaction between treatment and year was marginally significant (P < 0.053) it is not readily attributable to a treatment effect. The significant effect of date on NO efflux (T able 5-1, Figure 5-2B) was largely due to a single value measured in the end of July 2002; wet vs. dry season contrasts indicated nonsignificant seasonal differences in NO efflux. For N2O (Figure 5-2C), the wet season efflux was significantly higher than the dry season efflux (5.62 0.50 and 2.41 0.47 g N m-2 h-1, respectively; P< 0.0001). During dry-season ir rigation, treatment vs. control contrasts indicated that N2O efflux in irrigated plots was significantly higher than in control plots (4.18 0.87 and 2.34 0.75 g N m-2 h-1, respectively; P < 0.05). Date was again the only factor to have a significant effect on CH4 efflux (Table 5-1, Figure 5-2D). Methane efflux in the dry season was significantly lower than in the wet season (-0.348 0.118 and 0.128 0.118 mg CH4 m-2 d-1, respectively; P < 0.0001). During dryseason irrigation, treatment vs. c ontrol contrasts in dicated that CH4 efflux in irrigated plots was also significantly higher than in control plots (0.226 0.361 and -0.526 0.185 mg CH4 m-2 d1, respectively; P < 0.01). The net CH4 emissions were generally close to zero, with most chambers generally showing net uptake of CH4 (77% in control plots and 80% in irrigated plots). The range of CH4 efflux for the whole experimental period was -5.00 to 22.03 mg CH4 m-2 d-1. Two chambers with very high effluxes (5.93 and 9.97 mg CH4 m-2 d-1) drove the large variability in the mean efflux for the control pl ot in March 2001, while the high variability for the irrigation means in September and Oct ober 2001 was driven by one chamber (9.08 and 10.30 mg CH4 m-2 d-1). The apparent high mean net production of CH4 for irrigated plots in September (0.884 1.353 mg CH4 m-2 d-1) and October (0.879 1.187 mg CH4 m-2 d-1) 2001

PAGE 82

82 becomes net consumption (-0.461 0.172 and -0.292 0.223 mg CH4 m-2 d-1) if we exclude the high efflux chambers from the calculation of means. There was no significant effect of irrigati on on net nitrification rates for control and irrigated plots (0.11 0.02 and 0.11 0.03 g N g-1 soil d-1, respectively). Litter Removal Experiment Soil water potential (Figure 5-3B) was signifi cantly less negative in the wet season than in the dry season (P < 0.0001). Soil CO2 efflux during the pretreat ment period (Figure 5-3C) for litter removal and control plot s did not differ significantly (4.18 0.12 and 4.24 0.08 mol CO2 m-2 s-1, respectively; P = 0.87). During the litter manipulation period, soil CO2 efflux in litter removal plots was signifi cantly lower than in control plots (3.54 0.17 and 4.90 0.18 mol CO2 m-2 s-1, respectively; P < 0.001). This difference was not homogeneous throughout the experimental period and followed a trajectory that can be divided in three phases. In the first phase, corresponding with the dry season and the ea rly rainy season, the difference between treatment and control measur ements was apparent for nearly all of the measurements made during the first six months of litter removal. The second phase, from 6-10 months after the beginning of litter removal, co rresponded with the mid to late rainy season. During this phase, there were fewer measuremen ts in which the difference between treatment means was significant. In the third phase, corresponding with the following dry season, the difference in soil CO2 efflux between treatments was uni formly significant, and persisted through the end of the measurement period. Aboveground litter respir ation represented 22 2% of total soil respiration for the whole litter removal period and was 22 2, 16 4, and 28 2% of total soil respiration during the first, second and third phases, respectively. Annual soil C efflux was significantly lower (P <

PAGE 83

83 0.05) in litter removal than in control plots in 2002 (Table 5-2). There was no significant difference in annual litterfall C between contro l and litter removal treatments in 2002 (Table 52). The significant interaction effect on N oxide emissions (Table 5-1) was not related to a consistent effect of litter removal on either NO or N2O effluxes (Figures 5-4B-C, respectively); the difference between treatments for both ga ses during the litter re moval period was nonsignificant. Emissions of CH4 (Figure 5-4D) in the dry season were significantly lower than in the wet season (-0.420 0.164 and 0.287 0.113 mg CH4 m-2 d-1, respectively; P < 0.01). Mean net nitrification rates in control plots were marginally higher (P = 0.06) than in litter removal plots (0.11 0.02 and 0.07 0.01 g N g-1 soil d-1, respectively). Discussion Soil CO2 Efflux and Belowground C Allocation The soil CO2 efflux rates measured in our study are within the range of data reported for tropical forests and are consistent with severa l other studies in Amazonian forests (Cattnio et al. 2002, Davidson et al. 2004b, Davidson et al. 2000, Fernandes et al. 2002, Nepstad et al. 2002, Salimon et al. 2004, Verchot et al. 2000) and in tropical fo rests elsewhere (Ishizuka et al. 2002, Kiese & Butterbach-Bahl 2002) that reported higher emissions of CO2 during the wet season than in the dry season. We have also shown strong pulses of CO2 efflux in response to rain events during dry periods (soil wet-up events ), as observed in old-growth forests in the Brazilian Amazon (Davidson et al. 2000, Sotta et al. 2004) and in Costa Rica (Schwendenmann et al. 2003). Our dry-season irrigation experiment further demonstrates the constraint that moisture availability exerts on soil CO2 efflux.

PAGE 84

84 Soil CO2 efflux as measured in the field mainly in tegrates root and mi crobial respiration, and we have not determined if the reduction in soil respiratio n in the dry season was caused by decreased activity of microbes, roots or both. Ho wever, a laboratory study with soil from the same site showed a significant increase in microbial basal respir ation during the 2001 wet season compared to the previous dry season (R angel-Vasconcelos 2002), as observed in other tropical forests (Cleveland et al. 2003, Luizao et al. 1992). Although microbial respiration rates determined under laboratory conditions cannot be compared to rates obtained in the field with chamber techniques, those results suggest that reduction in soil microbial activity during the dry season likely contributed to the observed lowe r rates of soil respiratio n during this period at our site. Likewise, reduced activity of micr obes in decomposing aboveground litter during the dry season could have contributed to lower soil CO2 efflux in non-irrigated plots. Borken et al. (2003) have recently shown that microbial respiration of the O horizon can contribute significantly to CO2 pulses after soil wet-up events in a temperate forest and Goulden et al. (2004) reported that increased soil respira tion after a rainfall during the dry season was associated with surface litter rehydration in an Amazonian old-growth forest. Wieder and Wright (1995) have also observed higher litter mass loss under irrigation compared with no irrigation in a tropical forest in Panama. Finally, lower soil CO2 efflux during the dry season could also have resulted from constrained root respiration due to d ecreased root growth (Cattnio et al. 2002) or decreased flux of photosynthates to roots, which limits root respiration itself (Hgberg et al. 2001) and/or rhizospheric microbi al respiration (Kuzyakov & Cheng 2001). Further research on differentiating r oot from microbial resp iration and aboveground litter from soil respiration are needed to be tter understand how moisture constrains CO2 efflux from tropical forest soils, especi ally because likely concomitant and opposite variations in root

PAGE 85

85 and microbial dynamics under dry conditions (Davidson et al. 2004b) make it difficult to understand the mechanisms by which moistu re controls total soil respiration. The variation in the size of the difference of soil CO2 efflux between control and litter removal throughout the manipulation period followed a trajectory th at can be linked to altered substrate availability and variation in soil water st atus due to the seasonality of rainfall. In the first phase of this trajectory, the early impact of litter removal on soil respiration suggests that CO2 efflux associated with microbial decompos ition of aboveground litter and superficial root respiration represents a substa ntial proportion (about 22 % in th e present study) of total soil respiration (Raich & Schlesinger 1992). Duri ng the second phase, an interaction between substrate availability and rainfall seasonality appears to influence the variation in soil CO2 efflux. The difference between control and litte r removal plots decreased during some dates in the second phase, suggesting that the contribu tion of belowground respiration was relatively higher during the wet season. The third phase ma y be characterized by the depletion of labile soil carbon and, therefore, an in crease in the difference in soil CO2 efflux between treatments. Although this phase is also coincident with the 2002 dry season, its length and consistency (i.e., lack of responsiveness to dry-season wet-up events ) lead us to suspect that, due to the removal of the litter layer, substrate availability has become a larger constraint on soil respiration than reduced moisture availability. In 2002, litte r removal resulted in a 28% reduction in soil CO2 efflux, which is very similar to the 27% found in Costa Rica after 2 years of litter removal (Sayer 2005), but lower than the 54% reduction af ter 7 years of litter removal in Puerto Rico (Li et al. 2004). The estimated annual soil C efflux measured in our control plots is comparable to another estimate for eastern Amazonian forests in Brazil (Davidson et al. 2000) and is higher than

PAGE 86

86 estimates for tropical old-grow th forests elsewhere (Ishizuka et al. 2002, Schwendenmann et al. 2003); annual fluxes measured in other tropical forest sites are given in Table 5-3. We observed lower total rainfall and higher annua l soil C efflux in 2002 than in 2001, suggesting that the interannual variability in soil C efflux was not caused by di fferences in annual rainfall. Pulses of CO2 associated with rainfall events observe d in this study are co nsistent with the hypothesis that rainfall distributi on, rather than total rainfa ll, may better explain annual variability in soil C efflux. Differences in a nnual soil C efflux between irrigation and control plots are also consistent with a substantia l moisture constraint on soil respiration. Based on our annual soil C efflux and non-woody litterfall C estimates (Chapter 3), we can calculate a C efflux : litterfall C ratio of 4.0 5.2 for our cont rol plots, consistent with the mean value of 4.16 reported by Davidson et al. (2002) for young forests. Total belowground carbon allocation (TBCA) estimated by the diffe rence between annual basis C fluxes in soil respiration and litterfall (Raich & Nadelhoffer 1989) is underestimat ed for regrowth forests if C storage in roots and soil is not accounted for (Davidson et al. 2002). However, simple calculation of TBCA based only on soil respiration and litterfall can provide a lower limit of TBCA for regrowth forests. For our site, th e ratio between annual soil C efflux and annual litterfall C indicates that TBCA re lative to litterfall is similar to values for other regrowth forest site in the eastern Amazon (Davidson et al. 2002) and higher than t hose of mature forests (Davidson et al. 2002, Raich & Nadelhoffer 1989), consistent with increased a llocation of C to belowground structures as a mechanism by which regrowth forests cope with the demands for water and nutrients (Davidson et al. 2002). Differences in annual soil C efflux between li tter removal and control plots are consistent with a substantial substrate c onstraint on soil respiration. In 2002, the amount of carbon in

PAGE 87

87 litterfall (368 14 g C m-2 yr-1) was well within one standard e rror of the mean difference in soil C efflux between control and litter removal (559 291 g C m-2 yr-1). This substantial difference in soil C efflux also suggests that ~2 0% of total soil C effl ux is due to litter respiration, with the remaining ~80% due to be lowground respiration; this is consistent with results obtained in litter removal studies in forest ecosystems in the tropics (Li et al. 2004, Sayer 2005) and other clim atic regions (Bowden et al. 1993, Jandl & Sollins 1997, Rey et al. 2002). Nitrogen Oxide Emissions Nitric and nitrous oxide effl uxes measured in this study bot h in wet or dry seasons are among the lowest reported for either regrowth or old-growth tropical fo rests in the Brazilian Amazon (Cattnio et al. 2002, Davidson et al. 2004b, Garcia-Montiel et al. 2001, Nepstad et al. 2002, Verchot et al. 1999) and tropical forests elsewhere (Erickson et al. 2001, Ishizuka et al. 2002, Palm et al. 2002). These low N oxide effluxes may re sult from low rates of N cycling, as indicated by the very low net nitrification rates we found in both seasons compared to other studies for Amazonian forests (Garcia-Montiel et al. 2003a, Neill et al. 1997, Palm et al. 2002). The thin concretionary soils of this site, along with the recent history of repeated slash-andburn cycles and the high litterfal l C:N ratios, are consistent with a very conservative nitrogen cycle and low rates of both nitrif ication and denitrification. Although fluxes were consistently low, slightly higher efflux of N2O in the wet season compared to the dry season reported here has also been observed in ot her tropical forests (Cattnio et al. 2002, Erickson et al. 2002, Garcia-Montiel et al. 2001, Kiese & Butterbach-Bahl 2002, Nepstad et al. 2002, Verchot et al. 1999). Consistent with the results obtained by Nobre et al. (2001), we also found a significant effect of irrigation on N2O efflux. Higher N2O efflux associated with wetter soil conditions

PAGE 88

88 during both the wet season and dr y-season irrigation periods like ly resulted from increased denitrification (Davidson 1991). The effects of litter removal on N oxide fluxes are not clear and difficult to interpret since the fluxes are inherently very lo w at our site. If N trace gas em issions were already limited by N availability in this infertile soil, the removal of litter might be expect ed to decrease emissions further. However, that decrease would be di fficult to detect relative to the naturally low emissions that were already frequently near detection limits prior to litter removal. Methane Emissions Methane fluxes measured at our site are in the lower range of both net consumption and net production fluxes found for tropical forests (Kiese et al. 2003, Palm et al. 2002, Verchot et al. 2000). Higher net consumption of CH4 in the dry season and lower net consumption (or even small net production) in the wet season observe d in our study is consistent with the pattern of CH4 emissions measured in other Brazi lian Amazonian forests (Cattnio et al. 2002, Nepstad et al. 2002, Verchot et al. 2000) and tropical fore sts elsewhere (Kiese et al. 2003). Increased net CH4 production during the wet season as well as during the irrigation period in our study suggests that higher soil wate r status decreased soil aerati on leading to an increase in methanogenesis (Davidson & Schimel 1995). Although decreased aeration during the wet season could have resulted in higher efflux of CH4 and N2O, higher soil CO2 efflux associated with wetter soil conditions c ould also have contributed to the increased efflux of CH4 and N2O because of the consumption of O2 in the respiration process (Palm et al. 2002, Verchot et al. 2000). Conclusions We conclude that soil CO2 efflux is strongly linked to soil moisture and substrate availability as indicated from the responses of CO2 emissions to soil wet-up events, dry-season

PAGE 89

89 irrigation, and litter removal for th is tropical regrowth stand. On an annual basis, this regrowth stand allocates a large amount of C to belowgrou nd structures relative to litterfall C. Relieving dry season water limitation increased soil resp iration by about 40 and 30% in the two dry seasons studied, corresponding to annual increases of 27 and 13% in 2001 and 2002, respectively. Removing abovegr ound litter reduced a nnual soil respiration by 28% in 2002. In general, N oxide emissions were very low, probably due to the inherently low rates of nitrogen cycling at this site. Emissions of N2O and CH4 were constrained by low moisture availability, while emissions of NO were not affected by irrigation. We were unable to detect more severe substrate limitation induced by the litter removal treatment on N oxide and CH4 emissions. The substantial impacts of soil mo isture and aboveground litter on soil CO2 efflux shown in this study suggest that alterations in the availability of these resources that may result from climate and land-use changes in tropical regions could have significant effects on regional CO2 fluxes.

PAGE 90

90Table 5-1. F statistics and a ssociated significance levels for the effect of tr eatments (irrigation and litter removal), sampl ing date, and their interaction on soil trace gas fluxes and non-woody litterfall in a tropical regrowth forest stand in eastern Amazoniaa (PROC MIXED, SAS System version 9.0). Significant treatment effects (not observed) would indicate temporally consistent differences between treatment and control both preand post-treatme nt and across seasons, significant date effects are generally indicative of seasona l trends that affect both treatment a nd control measurements, and treatment x date effects indicate a significant difference between the tr eament and control measurements that occurs after the treatment was initiated. Irrigation experiment Litter removal experiment Variable Treatment Date Treatment x Date Treatment Date Treatment x Date CO2 efflux 2.55ns 9.48*** 5.02*** 3.10ns 9.63*** 3.24*** NO efflux 0.04ns 5.46*** 1.50ns 3.29ns 7.65*** 2.21* N2O efflux 0.93ns 4.20*** 1.00ns 0.32ns 6.42*** 1.68* CH4 efflux 0.91ns 2.14** 1.22ns < 0.01ns 2.21** 0.77ns Litterfall 0.24ns 45.27*** 1.62** 0.27ns 32.91*** 1.18ns aThe level of significance is indicated (*: P < 0.05, **: P < 0.01, ***: P < 0.001, ns: not significant).

PAGE 91

91 Table 5-2. Annual soil carbon e fflux and non-woody litterfall carbon for control, irrigated and litter removal plots in a tropi cal regrowth forest stand in eastern Amazonia (mean se, n = 12 per treatment). Soil C efflux (g m-2 yr-1) Non-woody litterfall C (g m-2 yr-1) Treament 2001 2002 2001 2002 Control 1593 74 1988 126 410 28 383 27 Irrigation 2021 154 2237 158 398 24 415 24 Litter removal 1429 165 368 14

PAGE 92

92 Table 5-3. Estimates of annual soil carbon (C) efflux in old-growth and regrowth tropical forests. Forest Location Rainfall (mm yr-1) Soil type Soil C efflux (g C m-2 yr-1) Reference Old-growth1 Par, Brazil (2o 59 S, 47o 31 W) 1800 Haplustox 2000 Davidson et al. (2000) Sumatra, Indonesia (1o 05.164 S, 102o 05.702 E) 2060 Ultisol 560-820 Ishizuka et al. (2002) La Selva, Costa Rica (10o 20 N, 83o 50 W) 4200 Typic Haploperox old alluvium 1060 Schwendenmann et al. (2003) Par, Brazil (2.8968 oS, 54.9519 oW) 2000 Haplustox 1000 Davidson et al. (2004b) Acre, Brazil 1940 dystrophic and eutrophic Ultisols with patches of Oxisols 1700 Salimon et al. (2004) Barro Colorado, Panama 2600 Oxisol 1740 Sayer (2005) Par, Brazil (2o 64 S, 54o 59 W) 2000 clayey (Ultisols and Oxisols) 1084 Silver et al. (2005) sandy (Ultisols) 1363 Silver et al. (2005) Regrowth2 Par, Brazil (2o 59 S, 47o 31 W), 20-year-old 1800 Oxisol and alfisol 1800 Davidson et al. (2000) Gran Sabana, Venezuela (5o 0 S, 61o 0 W) 2200 Acrohumox 896 (tall) 1241 (medium) 1024 (low) Priess and Flster (2001) Acre, Brazil, 3-18-yrold 1940 dystrophic and eutrophic Ultisols 1600 Salimon et al. (2004) Par, Brazil Oxisol 1790 This study 1 Includes sites classified as mature and primary forests. 2 Includes sites classified as secondary forests.

PAGE 93

93 A B C Date 02/00 08/00 02/01 08/01 02/02 08/02 02/03 Soil CO2 efflux ( mol m-2 s-1) 0 2 4 6 8 control irrigation Rainfall (mm) 0 20 40 60 80 100 120 Soil water potential (MPa) -0.10 -0.08 -0.06 -0.04 -0.02 0.00 Figure 5-1. Effects of rainfall patterns and dry-season irrigation on soil moisture status and soil respiration in an Amazonian regrowth forest stand, Brazil. A) Daily rainfall at the study site. B) Soil water potentia l. C) Soil carbon dioxide (CO2) efflux. In Figures B-C, circles represent means ( se); n = 4 for soil water potential and n = 12 for soil CO2 efflux per sampling date. Gray-shaded areas indicate the dry season irrigation periods. White and black horizontal bars mark dry and wet seasons, respectively.

PAGE 94

94 A B C D Soil NO efflux ( g N m-2 h-1) 0 2 4 6 8 10 12 14 control irrigation Soil N2O efflux ( g N m-2 h-1) -5 0 5 10 Date 07/99 01/00 07/00 01/01 07/ 01 01/02 07/02 01/03 Soil CH4 efflux (mg CH4 m-2 d-1) -1 0 1 2 3 Rainfall (mm) 0 20 40 60 80 100 120 Figure 5-2. Effects of rainfall patterns and dry-seas on irrigation on soil nitrogen oxide and methane effluxes in an Amazonian regrowth fore st stand, Brazil. A) Daily rainfall at the study site. B) Soil nitric oxide ( NO) efflux. C) Soil nitrous oxide (N2O) efflux. D) Soil methane (CH4) efflux. In Figures B-D, clos ed and open circles represent means ( se) for control and irrigation treatm ents, respectively (n = 8 per sampling date). Gray-shaded areas indicate the dr y season irrigation periods. White and black horizontal bars mark dry and wet seasons, respectively.

PAGE 95

95 A B CDate 02/00 08/00 02/01 08/01 02/02 08/02 02/03 Soil CO2 efflux ( mol m-2 s-1) 0 2 4 6 8 control litter removal Soil water potential (MPa) -0.10 -0.08 -0.06 -0.04 -0.02 0.00 Rainfall (mm) 0 20 40 60 80 100 120 Figure 5-3. Effects of rainfa ll patterns and litter removal on soil moisture status and soil respiration in an Amazonian regrowth forest stand, Brazil. A) Daily rainfall at the study site. B) Soil water potentia l. C) Soil carbon dioxide (CO2) efflux. In Figures B-C, circles represent means ( se); n = 4 for soil water potential and n = 12 for soil CO2 efflux per sampling date. The vertical line indicates the begi nning of the litter removal treatment. White and black horizontal bars mark dry and wet seasons, respectively.

PAGE 96

96 A Soil NO efflux ( g N m-2 h-1) 0 2 4 6 8 10 12 14 Soil N2O efflux ( g N m-2 h-1) -5 0 5 10 Date 07/99 01/00 07/00 01/01 07/01 01/02 07/02 01/03 Soil CH4 efflux (mg CH4 m-2 d-1) -1 0 1 2 3 control litter removal Rainfall (mm) 0 20 40 60 80 100 120 B C D Figure 5-4. Effects of rainfa ll patterns and litter removal on so il nitrogen oxide and methane effluxes in an Amazonian regrowth forest sta nd, Brazil. A) Daily rainfall at the study site. B) Soil nitric oxide (NO) efflux. C) Soil nitrous oxide (N2O) efflux. D) Soil methane (CH4) efflux. In Figures B-D, closed and open circles represent means ( se) for control and litter removal treatments respectively (n = 8 per sampling date). The vertical line indicates the beginning of the litter removal treatment. White and black horizontal bars mark dry a nd wet seasons, respectively.

PAGE 97

97 CHAPTER 6 MOISTURE CONSTRAINTS TO ABOVEGRO UND NET PRIMARY PRODUCTIVITY IN EASTERN AMAZONIAN FOREST REGROWTH Introduction Net primary productivity (NPP) is considered to be the best integrator measure of resource effects on ecosystem processes (Chapin & Evin er 2005). Improved understanding of temporal shifts in NPP may aid predictions of ecosyst em response to ongoing climate and land-use changes (Tian et al. 1998). In tropical forests, reliabl e estimates of NPP mostly involve measurements of aboveground net primary productivity (ANPP) components; due to methodological difficulties belowground NPP is rarely measured (Clark et al. 2001b). For tropical forest regrowth (e.g. following agricu ltural conversion and aba ndonment), there is a paucity of data even on ANPP, in part because th ese sites are very rarely measured over multiple years. Aboveground biomass increment in live tr ees (i.e., wood increment) and non-woody litterfall (a proxy for leaf production) are comm only used to estimate ANPP; both aboveground biomass increment and non-woody litterfall can be relatively easily measured and represent two significant components of total ANPP (Clark et al. 2001a). Stem diameter and height measures are usually used to estimate aboveground bioma ss (AGB) through allometr ic equations (e.g., Ducey et al. Submitted). Despite several reports on AG B for tropical forest regrowth (Gehring et al. 2005, Saldarriaga et al. 1988, Zarin et al. 2001), repeated measures of AGB and litterfall are rare and calculations of ANPP for th ese forests are therefore lacking. Observational and manipulative experiments suggest that moisture availability may be an important control over ANPP in tropical forests. At old-growth forest sites in the Brazilian Amazon, higher diameter growth rates are associated with wetter periods (Higuchi et al. 2003, Rice et al. 2004, Vieira et al. 2004). Nepstad et al. (2002) have previously shown that soil

PAGE 98

98 moisture depletion during a part ial throughfall exclusion experi ment reduced ANPP in an oldgrowth Amazonian forest. Conversely, exce ssive soil moisture may also decrease ANPP (Schuur & Matson 2001). Analogous data from both observational and ma nipulative studies are lacking for tropical forest regr owth, even though recent estimates indicate that there are ~38 million ha of regrowth in Latin America al one, and the area is growing as unproductive deforested land is abandoned (ITTO 2002). The primary objective of this chapter was to investigate the response of ANPP to experimentally increased dry-season moisture av ailability and inter-annual variability in dryseason precipitation during a four-year irrigation experiment described in Chapter 2. We hypothesized that dry-season irri gation would increase ANPP, a nd that ANPP would also be positively correlated with dry-season precipitation. Study Site and Experimental Design Study site and experimental desi gn are described in Chapter 2. Material and Methods Aboveground Net Primary Productivity Aboveground net primary productivity (ANPP) was estimated as the sum of annual increases in aboveground biomass (AGB) of trees (diameter at breast height 1 cm) and nonwoody litterfall (Clark et al. 2001a, Grace et al. 2001) between July 2001 and July 2005. To estimate AGB, we used site-specific mixed-species and species-specific allometric equations based on diameter measurements (Table 6-1; Ducey et al. In preparation). Diameter increments have previously been pub lished, in part, by Arajo et al. (2005). Non-woody litterfall data are reported on Chapter 3. Aboveground biomass increment (AGBI, in Mg ha-1 yr-1) was calculated for each plot as follows (Clark et al. 2001a):

PAGE 99

99 ABGI = ( increments of surviving trees) + ( increment(s) of ingrowth), where the increment of surviving tree s was calculated as the AGB in yearx+1 minus the AGB in the previous year (i.e., yearx), and the increment of ingrowth was calculated as the AGB in the ingrowth year minus AGB relative to the minimum diameter (1 cm). This method of calculating AGBI may underest imate its actual value if trees exhibit significant growth between their last measurement and their deat h. In a separate study, the increment in diameter at breast height (DBH) of trees with DBH 5 cm was measured every 1-2 months from November 2003 to December 2005 us ing dendrometer bands fabricated with aluminium tapes (data not presented). We observed that several months prior to tree death, stem increment was consistently equal to zero, sugge sting that unaccounted diameter increment prior to tree death (Clark et al. 2001a) may have little impact on biomass increment estimates. Statistical Analysis We used the SAS System version 9.00 to run the statistical analyses. We analyzed with PROC MIXED the effects of treatment, date, an d treatment-by-date interaction on ANPP using a repeated measures analysis with compound symm etric covariance structure. This structure assumes constant variance at all dates and equal correlations between all pairs of measures on the same experimental unit, i.e., plot. We used PROC NLIN for linear regr ession analysis between ANPP and rainfall (currentand previous-year a nnual rainfall and dry-se ason rainfall); annual rainfall corresponds to total rainfall in the inte rval between yearly diameter measurements. All results are reported as significant when P 0.05; we report marginal significance when 0.05 < P < 0.10. Multiple comparisons of means were performed with Tukeys test. Results ANPP was significantly affected by date (P = 0.034) and treatment (P = 0.026), with marginally significant effects of treatment x date interaction (P = 0.059). In the annual periods

PAGE 100

100 from July 2002 to July 2003, and July 2003 to July 2004, ANPP was significantly higher in irrigated plots than in control plots (Figure 6-1). ANPP was also positively correlated with previous-year dry-season rainfall (Figure 6-2; R2 = 0.45; P < 0.01). Discussion Aboveground net primary productivity range calculated for this site (12.3 0.5 to 16.6 2.1 Mg ha-1 yr-1, n = 4, control plots) is equivalent to the highest valu es reported by Clark et al. (2001b) for old-growth tropical fore sts. Our estimate of ANPP for this site represents a lower bound, because it only includes wood in crement and non-woody litterfall. Although our calculated ANPP values are re latively high, the aboveground biomass accumulated about 12 years after land abandonment (51.5 2.6 Mg ha-1) is 13% lower than the value obtained (59.2 Mg ha-1) with a model developed to predict aboveground biomass accumulation by Amazonian regrowth forests (Zarin et al. 2001), and substantially lower (> 70%) than the value predicted by the model develope d for regrowth forests recovering from firstcycle slash-and-burn in central Amazonian regrowth forests (Gehring et al. 2005). Lower biomass compared to model predictions may resu lt from (a) the history of repeated burning events (Zarin et al. 2005), (b) the inherent low fertility of the concretionary soil, and (c) relatively distinct dry season peri ods at the study site (Ape). While the difference from Zarin et al.s (2001) is within the mode l error, the great discrepa ncy in relation to Gehring et al.s (2005) work may be due to less severe dry seasons and more fertile soils for central Amazonian regrowth forests (Gehring et al. 2005). While the results of our dry-season irrigation experiment demonstrat e the constraint of moisture availability on ANPP at this site, re duced ANPP associated with lower previous-year dry-season rainfall indicates a lag effect of the influence of drought on ANPP. A recent

PAGE 101

101 dendrochronological study also found a lag effect of rainfall on stem growth for a tropical tree species in Bolivia (Brienen & Zuidema 2005). Possible explanations for lag effects on tree growth include rainfall controls on bud prefor mation (Critchfield 1960), storage of reserves under favorable conditions (Dnisch et al. 2003), and long-term water table storage (Borchert 1994). Decreased moisture availability can reduce productivity th rough effects on leaf hydration, stomatal conductance, and, ultimately, leaf photosynthesis (Chaves et al. 2003, Malhi et al. 1998, Mulkey & Wright 1996). In our study, irrigation may have stimulated plant productivity directly through d ecreasing drought limitations on photos ynthesis and/or indirectly through enhancing nutrient availability due to mo isture effects on litter decomposition (Cornejo et al. 1994, Wieder & Wright 1995); ir rigation may have also allowe d plants to better utilize higher light availability during less cloudy da ys typical of the dry seas on. Leaf water potential and gas exchange for Vismia guianensisa common species at the study area (Arajo et al. 2005)showed less negative water potential (hi gher leaf hydration) and sustained higher photosynthetic capacity for some dates under irrigation (Vasconcelos et al. 2002). Photosynthetic capacity in the understory at the si te has also been shown to respond positively to both irrigation and dry-seas on rainfall events (Fortini et al. 2003) and to be sensitive to interannual differences in dr y-season rainfall (Arago et al. 2005). Wood increment was more sensitive than litt erfall to altered moisture availability, consistent with greater reductions in stemwood grow th than fine litterfall in throughfall exclusion plots previously reported for an ol d-growth Amazonian forest (Nepstad et al. 2002). Greater sensitivity of stemwood increm ent to altered moisture avai lability may have important implications for the carbon bala nce of tropical forest regrowth since stemwood represents a large pool of carbon with a low turnover rate.

PAGE 102

102 Anticipated climate change for the Amazon re gion may include more frequent and severe dry seasons in response to global warming (IP CC 2001), deforestation (Costa & Foley 2000) and more frequent El Nio episodes (Trenberth & Hoar 1997). Our results indicate that the potential of forest regrowth to sequester carbon w ill decrease under that projected scenario.

PAGE 103

103 Table 6-1. Allometric equations used to estimate tree biomass in a tropica l regrowth forest stand in eastern Amazonia, Brazil. Species Equation1 Abarema jupunba Biomass = 0.014978 DBH3.5763 Casearia javitensis Biomass = 0.32982 DBH1.7336 Lacistema pubescens Biomass = 0.044856 DBH3.1285 Myrcia sylvatica Biomass = 0.14988 DBH3.093 Ocotea guianensis Biomass = 0.098412 DBH2.6117 Poecilanthe effusa Biomass = 0.28772 DBH2.2747 Vismia guianensis Biomass = 0.2897 DBH2.0468 All other taxa Biomass = 0.18598 DBH2.3155 1 Biomass = dry weight in kg; DBH = diameter at breast height in cm

PAGE 104

104 Aboveground increment Non-woody litterfall 2001-20022002-20032003-20042004-2005 Biomass (Mg ha-1 yr-1) 0 2 4 6 8 10 12 14 16 18 20 irr ctl ctl ctl ctl irr irr irr Figure 6-1. Effects of dryseason irrigation on aboveground in crement and non-woody litterfall for a tropical forest regrowth stand in eas tern Amazonia, Brazil. Each stacked bar represent means (n = 4). Ctl and irr refers to control and irrigation plots, respectively.

PAGE 105

105 A Annual rainfall 2000 2200 2400 2600 2800 3000 3200 3400ANPP (Mg ha-1 yr-1) 10 12 14 16 18 10 12 14 16 18 20 Dry-season rainfall 200 300 400 500 600 700 800 R2 = 0.1827 p = 0.0987 R2 = 0.1067 p = 0.2169 R2 = 0.4108 p = 0.075 R2 = 0.4537 p = 0.0042B D C Figure 6-2. Relationship between aboveground net primary productiv ity (ANPP) and rainfall in an Amazonian forest regrowth stand, Braz il. A) Current-year annual rainfall. B) Current year dry-season rainfall C) Previous year annual ra infall. D) Previous year dry-season rainfall. Symbols are mean s se for control plots (n = 4).

PAGE 106

106 CHAPTER 7 CONCLUSIONS This long-term (five continuous years), st and-level resource manipulative experiment consisting of daily dry-season irrigation and bi-weekly removal of aboveground litter demonstrated moisture and nutrient limitations to carbon and nutrient dynam ics associated with above and belowground ecosystem processes for a forest regrowth site in eastern Amazonia, Brazil. Aboveground (litterfall quantity and qua lity, and litter decomposition) and belowground (soil trace gas fluxes) processes showed marked intrannual variation asso ciated with rainfall seasonality (Table 7-1). Soil carbon dioxide (CO2) efflux and litter deco mposition rates were strongly linked to moisture availability as indicat ed from their responses to rainfall seasonality and dry-season irrigation; differential decompositi on rates among tree species were linked to leaf chemical and physical properties. Soil emissions of nitrous oxide (N2O) and methane (CH4) were slightly increased by dry-se ason irrigation, but soil nitric oxide (NO) emissions were not sensitive to changes in soil moisture availabil ity in irrigated plots. Aboveground net primary productivity an index that integrates resource effects on ecosystem processes was constrained by moisture availability as indicated by the resp onse of wood increment to interannual variation in dry season rainfall and to irrigation. The early impacts of aboveground litter removal on soil CO2 efflux are consistent with a substantial contributi on of microbial decomposition of aboveground litter (especially non-woody material) and superficial root respiration to soil CO2 efflux. Altered nutrient availability due to litter removal was detected as increasingly re duction of nitrogen concentration in non-woody litterfall over time, consistent with the importance of litter cycling as source of nitrogen in forest ecosystems. However, net primary productiv ity (non-woody litterfall quantity) has not been constrained by reduced nitrogen av ailability so far, s uggesting some capacity of trees to sustain

PAGE 107

107 the same levels of biomass production with reduced leaf nitrogen Non-woody litterfall phosphorus concentration was less sensitive to ch ronic litter removal probably because plants compensated for removed phosphorus by accessing soil organic sources. Nitrogen oxide emissions and methane emissions we re not affected during the ini tial period (first 18 months) of litter removal. In general, this forest regrowth stand showed high resistance to altered nutrient availability, which may be linked to mechanisms that allow trees to mobilize nutrients (e.g., phosphorus) from soil organic sources, and to main tain productivity even u nder reduced litterfall nitrogen in litter removal plots. On the ot her hand, reduced ANPP associated with moisture availability suggests decreased potential of car bon sequestration from forest regrowth under anticipated scenarios of reduced rainfall in Amazonia. Our results may help to improve predictions of forest regrowth in the Brazili an Amazonia derived from process-based models such as CARLUC (Hirsch et al. 2004).

PAGE 108

108Table 7-1. Summary of ecosystem process responses to intrannual and in terannual variability effects (for control plots) and re source manipulation (dry-season irrigation and litter removal) effects. In trannual variability refers to variations associated with rainfall seasonality. The degree of resource ma nipulation effects is re lative to control. Process/Variable Intrannual variability Interannual variability Dry-Season Irrigation Litter Removal Soil water availability Yes NA ++ o Litterfall quantity Yes Yes + o Litterfall nitrogen concentration No NA o -Litterfall phosphorus concentration Yes NA o o Leaf litter decomposition Yes NA ++ NA Soil CO2 efflux Yes Yes ++ -Soil NO efflux Yes NA o o Soil N2O efflux Yes NA + o Soil CH4 efflux Yes NA + o ANPP NA Yes + NA Yes: presence of variability No: absence of variability +: slight, but significant increase ++: significant increase --: significant decrease o: no significant variation NA: not available

PAGE 109

109 LIST OF REFERENCES AERTS, R. 1997. Climate, leaf litter chemistry and leaf litte r decomposition in terrestrial ecosystems: a triangular relationship. Oikos 79: 439-449. AKMAL, M., & HIRASAWA, T. 2004. Growth respons es of seminal roots of wheat seedlings to a reduction in the water potential of vermiculite. Plant and Soil 267: 319-328. ANDERSON, J. M., & INGRAM, J. S. I. 1996. Tropical soil biology and fertility. A handbook of methods. CAB International, Wallingford. ARAGO, D. V., FORTINI, L. B., MULKEY, S. S., ZARIN, D. J., ARAUJO, M. M., & CARVALHO, C. J. R. DE. 2005. Correlation but no causation between leaf nitrogen and maximum assimilation: the role of drought and reproduction in gas exchange in an understory tropical plant Miconia ciliata (Melastomataceae). American Journal of Botany 92: 456-461. ARAJO, M. M., TUCKER, J. M., VASCONCELOS, S. S., ZARI N, D. J., OLIVEIRA, W., SAMPAIO, P. D., RANGEL-VASCONCELOS, L. G. T., OLIVEIRA, F. DE A., COELHO, R. DE F. R., ARAGO, D. V., & MIRANDA, I. 2005. Padro e processo sucessionais em florestas s ecundrias de diferentes id ades na Amaznia oriental. Cincia Florestal 15: 343-357. ASNER, G. P., TOWNSEND, A. R., & BRASWELL, B. H. 2000. Satellite observation of El Nio effects on Amazon forest phenology and productivity. Geophysical Research Letters 27: 981-984. BERISH, C. W., & EWEL, J. J. 1988. Root de velopment in simple and complex tropical successional ecosystems. Plant and Soil 106: 73-84. BOONE, R. D., GRIGAL, D. F., SOLLINS, P., AHRENS, R. J., & ARMSTRONG, D. E. 1999. Soil sampling, preparation, arch iving, and quality control. Pp. 3-28. in ROBERTSON, G. P., COLEMAN, D. C., BLEDSOE, C. S., & SOLLINS, P. (ed.). Standard soil methods for long-term ecological research. Oxford University Press, Oxford. BORCHERT, R. 1994. Induction of rehydration and bud break by irri gation or rain in deciduous trees of a tropical dry forest in Costa Rica. Trees 8: 198-204. BORKEN, W., DAVIDSON, E. A., SAVAGE, K., GAUDINSKI, J., & TRUMBORE, S. E. 2003. Drying and wetting effects on carbon di oxide release from organic horizons. Soil Science Society of America Journal 67: 1888-1896. BOWDEN, R. D., NADELHOFFER, K. J., BOONE, R. D., MELILLO, J. M., & GARRISON, J. B. 1993. Contributions of aboveground litter, be lowground litter, and ro ot respiration to total soil respiration in a temperate mixed hardwood forest. Canadian Journal of Forest Research 23: 1402-1407. BRADY, N. C. 1989. Natureza e propriedade dos solos. Livraria Freitas Bast os, Rio de Janeiro.

PAGE 110

110 BREITSPRECHER, A., & BETHEL, J. S. 1990. Stem -growth periodicity of trees in a tropical wet forest of Costa Rica. Ecology 71: 1156-1164. BRIENEN, R. J. W., & ZUIDEMA, P. A. 2005. Relati ng tree growth to rainfall in Bolivian rain forests: a test for six spec ies using tree ring analysis. Oecologia 146: 1-12. BRODRIBB, T. J., HOLBROOK, N. M., & GUTIRREZ, M. V. 2002. Hydraulic and photosynthetic co-ordinati on in seasonally dry tropical forest trees. Plant, Cell and Environment 25: 1435-1444. BROWN, S., & LUGO, A. E. 1990. Tropical secondary forests. Journal of Tropical Ecology 6: 1-32. CAMPO, J., & VZQUEZ-YANES, C. 2004. Ef fects of nutrient limitation on aboveground carbon dynamics during tropical dry forest regeneration in Yucatn, Mexico. Ecosystems 7: 311-319. CARVALHO, C. J. R. DE, NEPSTAD, D., CH AVES, M. M., & MEIR, P. (2005). Simulando o El Nio: duas experincias de excluso de chuvas em florestas na Amaznia. Pp. 42-49. in NOGUEIRA, R. J. M. C., ARAJO, E. D. L., WILLADINO, L. G., & CAVALCANTE, U. M. T. (ed.). Estresses ambientais: danos e benefcios em plantas, Recife, UFRPE, Imprensa Universitria. CATTNIO, J. H., DAVIDSON, E. A., NEPST AD, D. C., VERCHOT, L. V., & ACKERMAN, I. L. 2002. Unexpected results of a pilo t throughfall exclusion experiment on soil emissions of CO2, CH4, N2O, and NO in eastern Amazonia. Biology and Fertility of Soils 36: 102-108. CAVELIER, J., WRIGHT, S. J., & SANTAMARA, J. 1999. Effects of irrigation on litterfall, fine root biomass and pr oduction in a semideciduous lowland forest in Panama. Plant and Soil 211: 207-213. CHAPIN, F. S., III, & EVINER, V. T. 2005. Bi ogeochemistry of terrestrial net primary production. Pp. 215-247. in SCHLESINGER, W. H. (ed.). Biogeochemistry: Treatise on geochemistry, Volume 8. Elsevier, Boston. CHAPIN, F. S., III, MATSON, P. A., & MOONEY, H. A. 2002. Principles of terrestrial ecosystem ecology. Springer, New York. CHAVES, M. M., MAROCO, J. P., & PEREIRA, J. S. 2003. Understanding plant responses to droughtfrom genes to the whole plant. Functional Plant Biology 30: 239-264. CHAZDON, R. L., BRENES, A. R., & ALVARADO, B. V. 2005. E ffects of climate and stand age on annual tree dynamics in tropi cal second-growth rain forests. Ecology 86: 18081815.

PAGE 111

111 CLARK, D. A., BROWN, S., KICKLIGHTER, D. W., CHAMBERS, J. Q., THOMLINSON, J. R., & NI, J. 2001a. Measuring net primary pr oduction in forests: concepts and field methods. Ecological Applications 11: 356-370. CLARK, D. A., BROWN, S., KICKLIGHTER, D. W., CHAMBERS, J. Q., THOMLINSON, J. R., NI, J., & HOLLAND, E. A. 2001b. Net prim ary production in trop ical forests: an evaluation and synthesis of existing field data. Ecological Applications 11: 371-384. CLEVELAND, C. C., & TOWNSEND, A. R. 2006. Nutrient additions to a tropical rain forest drive substantial soil carbon dioxi de losses to the atmosphere. Proceedings of the National Academy of Sciences 103: 10316-10321. CLEVELAND, C. C., TOWNSEND A. R., & SCHMIDT, S. K. 2002. Phosphorus limitation of microbial processes in moist tropical fore sts: evidence from short-term laboratory incubations and field studies. Ecosystems 5: 680-691. CLEVELAND, C. C., TOWNSEND A. R., SCHMIDT, S. K., & CONSTANCE, B. C. 2003. Soil microbial dynamics and biogeochemist ry in tropical forests and pastures, southwestern Costa Rica. Ecological Applications 13: 314-326. COELHO, R. DE F. R., ZARIN, D. J., MIRANDA, I. S., & TUCKER, J. M. 2004. Anlise florstica e estrutural de uma floresta em diferentes estgi os sucessionais no municpio de Castanhal, Par. Acta Amazonica 33: 563-582. CONDIT, R., HUBBELL, S. P., & FOSTER, R. B. 1995. Mortality rates of 205 neotropical tree and shrub species and the im pact of a severe drought. Ecological Monographs 65: 419439. CORNEJO, F. H., VARELA, A., & WRIGHT, S. J. 1994. Tropical forest litter decomposition under seasonal drought: nutrient re lease, fungi and bacteria. Oikos 70: 183-190. CORNU, S., LUIZAO, F., ROUILLER, J., & LU CAS, Y. 1997. Comparative study of litter decomposition and mineral element release in two Amazonian forest ecosystems: litter bag experiments. Pedobiologia 41: 456-471. COSTA, M. H., & FOLEY, J. A. 2000. Comb ined effects of deforestation and doubled atmospheric CO2 concentrations on the climate of Amazonia. Journal of Climate 13: 1834. CRITCHFIELD, W. B. 1960. Leaf dimorphism in Populus trichocarpa. American Journal of Botany 47: 699-711. CUEVAS, E., BROWN, S., & LUGO, A. E. 1991. Aboveand belowground organic matter storage and production in a trop ical pine plantation and a paired broadleaf secondary forest. Plant and Soil 135: 257-268.

PAGE 112

112 CUEVAS, E., & MEDINA, E. 1986. Nutrient dynam ics within Amazonian forest ecosystems I. Nutrient flux in fine litter fall and efficiency of nutrient utilization. Oecologia 68: 466472. CUEVAS, E., & MEDINA, E. 1988. Nutrient dynamic s within Amazonian forests II. Fine root growth, nutrient availability a nd leaf litter decomposition. Oecologia 76: 222-235. DA ROCHA, H. R., GOULDEN, M. L., MILLER, S. D., MENTON, M. C., PINTO, L. D. V. O., FREITAS, H. C. DE, & FIGUEIRA, A. M. E. S. 2004. Seasonality of water and heat fluxes over a tropical forest in eastern Amazonia. Ecological Applications 14: S22-S32. DANTAS, M., & PHILLIPSON, J. 1989. Litterfall and litter nutrient cont ent in primary and secondary Amazonian "terra firme" rain forest. Journal of Tropical Ecology 5: 27-36. DAVIDSON, E. A. 1991. Fluxes of nitrous oxide a nd nitric oxide from te rrestrial ecosystems. Pp. 219-235. in ROGERS, J. E., & WHITMANN, W. B. (ed.). Microbial production and consumption of greenhouse gases: met hane, nitrogen oxides, and halomethanes. American Society for Mi crobiology, Washington. DAVIDSON, E. A., & ARTAXO, P. 2004. Globally si gnificant changes in biological processes of the Amazon Basin: results of the Larg e-scale Biosphere-Atmosphere Experiment. Global Change Biology 10: 519-529. DAVIDSON, E. A., CARVALHO, C. J. R. DE, VI EIRA, I. C. G., FIGUEIREDO, R. DE O., MOUTINHO, P., ISHIDA, F. Y., SANTOS, M. T. P. DOS, GUERRERO, J. B., KALIF, K., & SAB, R. T. 2004a. Nitrogen and phos phorus limitation of biomass growth in a tropical secondary forest. Ecological Applications 14: S150-S163. DAVIDSON, E. A., ISHIDA, F. Y., & NEPSTAD, D. C. 2004b. Effects of an experimental drought on soil emissions of carbon dioxide, methan e, nitrous oxide, and nitric oxide in a moist tropical forest. Global Change Biol 10: 718-730. DAVIDSON, E. A., MATSON, P. A., VITOUSEK P. M., RILEY, R., DUNKIN, K., GARCAMNDEZ, G., & MAASS, J. M. 1993. Proce sses regulating soil emissions of NO and N2O in a seasonally dry tropical forest. Ecology 74: 130-139. DAVIDSON, E. A., SAVAGE, K., BOLSTAD, P., CLARK, D. A., CURTIS, P. S., ELLSWORTH, D. S., HANSON, P. J., LAW, B. E., LUO, Y., PREGITZER, K. S., RANDOLPH, J. C., & ZAK, D. 2002. Be lowground carbon allocation in forests estimated from litterfall and IRGA-based soil respiration measurements. Agricultural and Forest Meteorology 113: 39-51. DAVIDSON, E. A., & SCHIMEL, D. S. 1995. Microbial processes of production and consumption of nitric oxide, nitrous oxide and methane. Pp. 327-357. in MATSON, P. A., & HARRISS, R. C. (ed.). Biogenic trace gases: measur ing emissions from soil and water. Blackwell Science, Oxford.

PAGE 113

113 DAVIDSON, E. A., VERCHOT, L. V., CATTNIO, J. H., ACKERMAN, I. L., & CARVALHO, J. E. M. 2000. Effects of soil wa ter content on soil respiration in forests and cattle pastures of eastern Amazonia. Biogeochemistry 48: 53-69. DIAS, H. K. O. 2006. Vegetao, chuva de sementes e banco de sementes do solo em floresta secundria sob manipulao de gua, Par, Amaznia Oriental, Brasil. Master thesis, Universidade Federal Rural da Amaznia, Belm. DUCEY, M. J., ZARIN, D. J., VASCONCELOS, S. S., & ARAJO, M. M. Submitted. Biomass equations for forest regrowth in the easte rn Amazon using randomized branch sampling. Biotropica. DNISCH, O., MONTIA, V. R ., & BAUCH, J. 2003. Dendroeco logical investigations on Swietenia macrophylla King and Cedrela odorata L. (Meliaceae) in the central Amazon. Trees 17: 244-250. EAMUS, D. 2003. How does ecosystem water bala nce affect net primar y productivity of woody ecosystems? Functional Plant Biology 30: 187-205. EAMUS, D., & PRIOR, L. 2001. Ecophysiology of trees of seasonally dry tropics: comparisons among phenologies. Advances in Ecological Research 32: 113-197. ERICKSON, H., DAVIDSON, E., & KELLER, M. 2002. Former land-use and tree species affect nitrogen oxide emissions from a tropical dry forest. Oecologia 130: 297-308. ERICKSON, H., KELLER, M., & DAVIDSON, E. A. 2001. Nitrogen oxide fluxes and nitrogen cycling during postagricultural succession and forest fertilization in the humid tropics. Ecosystems 4: 67-84. EVINER, V. T., CHAPIN, F. S ., III, & VAUGHN, C. E. 2000. Nutrient manipulations in terrestrial ecosystems. Pp. 291-307. in SALA O. E., JACKSON, R. B., MOONEY, H. A., & HOWARTH, R. W. (ed.). Methods in ecosystem science. Springer-Verlag, New York. FEARNSIDE, P. M. 1996. Amazonian deforestat ion and global warming: carbon stocks in vegetation replacing Brazil's Amazon forest. Forest Ecology and Management 80: 21-34. FERNANDES, S. A. P., BERNOUX, M., CERRI, C. C., FEIGL, B. J., & PICCOLO, M. C. 2002. Seasonal variation of soil chemical properties and CO2 and CH4 fluxes in unfertilized and P-fertilized pastures in an Ultisol of the Brazilian Amazon. Geoderma 107: 227-241. FIRESTONE, M. K., & DAVIDSON, E. A. 198 9. Microbiological basis of NO and N2O production and consumption in soil. Pp. 7-21. in ANDREAE, M. O., & SCHIMEL, D. S. (ed.). Exchange of trace gases between te rrestrial ecosystems and the atmosphere. John Wiley & Sons, Chichester.

PAGE 114

114 FLEXAS, J., & MEDRANO, H. 2002. Dr ought inhibition of photosynthesis in C3 plants: stomatal and non-stomatal limitations revisited. Annals of Botany 89: 183-189. FOLEY, J. A., BOTTA, A., & COE, M. T. 2002. El Nio-Southern Oscillation and the climate, ecosystems and rivers of Amazonia. Global Biogeochemical Cycles 16: 1132, doi: 10.1029/2002GB001872. FONTE, S. J., & SCHOWALTER, T. D. 2004. Decomposition of greenfall vs. senescent foliage in a tropical forest ecosystem in Puerto Rico. Biotropica 36: 474-482. FORTINI, L. B., MULKEY, S. S., ZARIN, D. J., VASCONCELOS, S. S., & CARVALHO, C. J. R. DE. 2003. Drought constrai nts on leaf gas exchange by Miconia ciliata (Melastomataceae) in the understory of an eas tern Amazonian regrowth forest stand. American Journal of Botany 90: 1064-1070. FRIZANO, J., VANN, D. R., JO HNSON, A. H., JOHNSON, C. M., VIEIRA, I. C. G., & ZARIN, D. J. 2003. Labile phosphorus in soils of forest fallows and primary forest in the Bragantina region, Brazil. Biotropica 35: 2-11. GARCIA-MONTIEL, D. C., MELI LLO, J. M., STEUDLER, P. A., CERRI, C. C., & PICCOLO, M. C. 2003a. Carbon limitations to nitrous oxide emissions in a humid tropical forest of the Brazilian Amazon. Biology and Fertility of Soils 38: 267-272. GARCIA-MONTIEL, D. C., STEUDLER, P. A., PICCOLO, M. C., MELILLO, J. M., NEILL, C., & CERRI, C. C. 2001. Controls on soil ni trogen oxide emissions from forest and pastures in the Brazilian Amazon. Global Biogeochemical Cycles 15: 1021-1030. GARCIA-MONTIEL, D. C., STEUDLER, P. A., PICCOLO, M. C., NEILL, C., MELILLO, J. M., & CERRI, C. C. 2003b. Nitrogen oxide em issions following wetting of dry soils in forest and pastures in Rondnia, Brazil. Biogeochemistry 64: 319-336. GASH, J. H. C., NOBRE, C. A., ROBERT S, J. M., & VICTORIA, R. L. 1996. Amazonian deforestation and climate. John Wiley & Sons, New York. GEHRING, C., DENICH, M., KANASHIRO, M ., & VLEK, P. L. G. 1999. Response of secondary vegetation in Eastern Amazonia to relaxed nutrient availa bility constraints. Biogeochemistry 45: 223-241. GEHRING, C., DENICH, M., & VLEK, P. G. 2005. Resilience of secondary forest regrowth after slash-and-burn agriculture in central Amazonia. Journal of Tropical Ecology 21: 519-527. GIARDINA, C. P., BINKLEY, D., RYAN, M. G ., FOWNES, J. H., & SENOCK, R. S. 2004. Belowground carbon cycling in a humid tropical forest decreases with fertilization. Oecologia 139: 545-550.

PAGE 115

115 GOLLEY, F. B. 1983. Nutrient cycling and nutri ent conservation. Pp. 137-156. in GOLLEY, F. B. (ed.). Ecosystems of the world 14 A. Tropical rainforests ecosystems: structure and function. Elsevier Scientific Publishing Co., New York. GOULDEN, M. L., MILLER, S. D., ROCHA, H. R. DA, MENTON, M. C., FREITAS, H. C. DE, FIGUEIRA, A. M. E. S., & SOUSA, C. A. D. DE. 2004. Diel and seasonal patterns of tropical forest CO2 exchange. Ecological Applications 14: S42-S54. GOWER, S. T. 1987. Relations between mineral nut rient availability and fine root biomass in two Costa Rican tropical wet forests: a hypothesis. Biotropica 19: 171-175. GRACE, J., MALHI, Y., HIGUCHI, N., & MEIR, P. 2001. Productivity of tropical rain forests. Pp. 401-426. in ROY, J., SAUGIER, B., & MOONEY, H. A. (ed.). Terrestrial global productivity. Academic Press, San Diego. GRAHAM, E. A., MULKEY, S. S ., KITAJIMA, K., PHILLIPS, N. G., & WRIGHT, S. J. 2003. Cloud cover limits net CO2 uptake and growth of a rainfo rest tree during tropical rainy seasons. Proceedings of the Nati onal Academy of Sciences 100: 572-576. HANSON, P. J. 2000. Large-scale water mani pulations. Pp. 341-180. in SALA, O. E., JACKSON, R. B., MOONEY, H. A. & HOWARTH, R. W. (ed.). Methods in ecosystem science. Springer-Verlag, New York. HARCOMBE, P. A. 1977. The influence of fer tilization on some aspects of succession in a humid tropical forest. Ecology 58: 1375-1383. HARMON, M. E., NADELHOFER, K. J., & BLAIR, J. M. 1999. Measuring decomposition, nutrient turnover, and stores in plant lit ter. Pp. 202-240. in ROBERTSON, G. P., COLEMAN, D. C., BLEDSOE, C. S., & SOLLINS, P. (ed.). Standard soil methods for long-term ecological research. Oxford University Press, New York. HARRINGTON, R. A., FOWNES, J. H., & VITOUSEK, P. M. 2001. Production and resource use efficiencies in Nand P-limited tropical forests: A comparison of responses to longterm fertilization. Ecosystems 4: 646-657. HART, S. C., STARK, J. M., DAVIDSON, E. A., & FIRESTONE, M. K. 1994. Nitrogen mineralization, immobilization, and nitrif ication. Pp. 985-1018. in WEAVER, R. W., ANGLE, J. S., & BOTTOMLEY, P. S. (ed.). Methods of soil analysis, Part 2. Microbiological and biochemical properties. Soil Science Society of America, Madison. HEAL, O. W., ANDERSON, J. M., & SWIFT, M. J. 1997. Plant litter quality and decomposition: an historical overview. Pp 3-30. in CADISCH, G., & GILLER, K. E. (ed.). Driven by nature: Plant l itter quality and decomposition. CABI Publishing, Wallingford. HERBERT, D. A., WILLIAMS, M., & RASTETTER, E. B. 2003. A model analysis of N and P limitation on carbon accumulation in Amazonian secondary forest after alternate land-use abandonment. Biogeochemistry 65: 121-150.

PAGE 116

116 HERRERA, R., JORDAN, C. F., KLINGE, H ., & MEDINA, E. 1978. Amazon ecosystems. Their structure and functioning with particular emphasis on nutrients. Interciencia 3: 223231. HIGUCHI, N., CHAMBERS, J. Q., SILVA, R. P. DA, MIRANDA, E. V. DE, SANTOS, J. DOS, IIDA, S., PINTO, A. C. M., ROCHA, R. DE M., & SOUZA, C. A. S. DE. 2003. Uso de bandas metlicas e dendrmetros auto mticos para a definio do padro de crescimento individual das princi pais espcies arbreas da floresta primria da regio de Manaus, Amazonas, Brasil. Pp. 55-68. in HI GUCHI, N., SANTOS, J., & SAMPAIO, P. T. B. (ed.). Projeto Jacaranda Fase II: Pesquisas Florestais na Amaznia Central. Instituto Nacional de Pesquisas da Amaznia (INPA), Manaus. HIRSCH, A. I., LITTLE, W. S., HOUGHTON, R. A., SCOTT, N. A., & WHITE, J. D. 2004. The net carbon flux due to deforestation and forest re-growth in the Brazilian Amazon: analysis using a process-based model. Global Change Biol 10: 908-924. HGBERG, P., NORDGREN, A., BUCHMANN, N., TAYLOR, A. F. S., EKBLAD, A., HGBERG, M. N., NYBERG, G., OTTOSSON -LFVENIUS, N., & READ, D. J. 2001. Large-scale forest girdling shows that curr ent photosynthesis drives soil respiration. Nature 411: 789-792. HORTON, J. L., & HART, S. C. 1998. Hydraulic lift: a potentially important ecosystem process. Trends in Ecology and Evolution 13: 232-235. HUBBARD, R. M., RYAN, M. G., STILLER, V., & SPERRY, J. S. 2001. Stomatal conductance and photosynthesis vary linearly with plan t hydraulic conductan ce in ponderosa pine. Plant, Cell and Environment 24: 113-121. HUETE, A. R., DIDAN, K., SHIMABUKURO, Y. E., RATANA, P., SALESKA, S. R., HUTYRA, L. R., YANG, W., NEMANI, R. R., & MYNENI, R. 2006. Amazon rainforests green-up with sunlight in dry season. Geophysical Research Letters 33: L06405, doi:10.1029/2005GL025583. IPCC (Intergovernmental Panel on Climate Change) 2001. Climate change 2001: the scientific basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. ISHIZUKA, S., TSURUTA, H., & MURDIYARSO, D. 2002. An intensive field study on CO2, CH4, and N2O emissions from soils at four landuse types in Sumatra, Indonesia. Global Biogeochemical Cycles 16: 1049, doi:10.1029/2001GB001614. ITTO (International Tropical Timber Organization) 2002. ITTO guidelines for the restoration, management and rehabilitation of degr aded and secondary tropical forests. ITTO, Yokohama. JANDL, R., & SOLLINS, P. 1997. Water-extractable soil carbon in relation to the belowground carbon cycle. Biology and Fertility of Soils 25: 196-201.

PAGE 117

117 JIPP, P. H., NEPSTAD, D. C., CASSEL, D. K ., & CARVALHO, C. R. DE. 1998. Deep soil moisture storage and transpiration in forest s and pastures of seasonally-dry Amazonia. Climatic Change 39: 395-412. JOHNSON, C. H., ZARIN, D. J., & JOHNS ON, A. H. 2000. Post-disturbance aboveground biomass accumulation in global secondary forests. Ecology 81: 1395-1401. JORDAN, C. F. 1983. Productivity of tropical rain forest ecosystems and the implications for their use as future wood and energy supplie s. Pp. 117-136. in GOLLEY, F. B. (ed.). Ecosystems of the world 14 A. Tropical rain forests ecosystems: structure and function. Elsevier Scientific Publishing Co., New York. JOSLIN, J. D., WOLFE, M. H., & HANSON, P. J. 2001. Factors controlling the timing of root elongation intensity in a ma ture upland oak stand. Plant and Soil 228: 201-212. KELLER, M., ALENCAR, A., ASNER, G. P., BRASWELL, B., BUSTAMANTE, M., DAVIDSON, E., FELDPAUSCH, T., FERNANDES, E., GOULDEN, M., KABAT, P., KRUIJT, B., LUIZO, F., MILLER, S., MARK EWITZ, D., NOBRE, A. D., NOBRE, C. A., FILHO, N. P., ROCHA, H. DA, DIAS, P. S., RANDOW C. V., & VOURLITIS, G. L. 2004. Ecological research in the LargeScale Biosphere-Atmosphere Experiment in Amazonia: early results. Ecological Applications 14: S3-S16. KIESE, R., & BUTTERBACH-BAHL, K. 2002. N2O and CO2 emissions from three different tropical forest sites in the wet tropics of Queensland, Australia. Soil Biology & Biochemistry 34: 975-987. KIESE, R., HEWETT, B., GRAHAM, A., & BUTTERBACH-BAHL, K. 2003. Seasonal variability of N2O emissions and CH4 uptake by tropical rainforest soils of Queensland, Australia. Global Biogeochemical Cycles 17: 1043, doi:10.1029/2002GB002014. KITAJIMA, K., MULKEY, S. S., & WRIGHT, S. J. 1997. Seasonal leaf phenotypes in the canopy of a tropical dry fores t: photosynthetic characterist ics and associated traits. Oecologia 109: 490-498. KLINGE, H., & RODRIGUES, W. A. 1968. Litter production in an area of Amazonian Terra Firme forest. Part I. Litter-fall, organi c carbon and total nitrogen contents of litter. Amazoniana 4: 287-302. KUZYAKOV, Y., & CHENG, W. 20 01. Photosynthesis controls of rhizosphere respiration and organic matter decomposition. Soil Biology & Biochemistry 33: 1915-1925. LAMBERS, H., CHAPIN, F. S., III, & PONS, T. L. 1998. Plant physiological ecology. Springer, New York. LAURANCE, W. F., FEARNSIDE, P. M ., LAURANCE, S. G., DELAMONICA, P., LOVEJOY, T. E., MERONA, J. M. R.-D., CHAMBERS, J. Q., & GASCON, C. 1999. Relationship between soils and Amazon fo rest biomass: a landscape-scale study. Forest Ecology and Management 118: 127-138.

PAGE 118

118 LAWLOR, D. W. 2002. Limitation to photosynthe sis in water-stressed leaves: stomata vs. metabolism and the role of ATP. Annals of Botany 89: 871-885. LAWLOR, D. W., & CORNIC, G. 2002. Photosynt hetic carbon assimilation and associated metabolism in relation to water deficits in higher plants. Plant, Cell and Environment 25: 275-294. LAWRENCE, D. 2005. Regional-scal e variation in litter producti on and seasonality in tropical dry forests of southern Mexico. Biotropica 37: 561-570. LEAN, J., BUNTON, C. B., NOBRE, C. A., & ROWNTREE, P. R. 1996. The simulated impact of Amazonian deforestation on climat e using measured ABRACOS vegetation characteristics. Pp. 549-576. in GASH, J. H. C., NOBRE, C. A., ROBERTS, J. M., & VICTORIA, R. L. (ed.). Amazonian deforestation and climate. John Wiley & Sons, New York. LI, Y., XU, M., SUN, O. J., & CUI, W. 2004. Effects of root and li tter exclusion on soil CO2 efflux and microbial biomass in wet tropical forests. Soil Biology and Biochemistry 36: 2111-2114. LI, Y., XU, M., & ZOU, X. 2006. Effects of nut rient additions on ecosystem carbon cycle in a Puerto Rican tropical wet forest. Global Change Biology 12: 284-293. 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: 151160. LODGE, D. J., MCDOWELL, W. H., & MCSWIN EY, C. P. 1994. The importance of nutrient pulses in tropical forests. Trends in Ecology and Evolution 9: 384-387. LORANGER, G., PONGE, J., IMBERT, D., & LA VELLE, P. 2002. Leaf decomposition in two semi-evergreen tropical forests: influence of litter quality. Biology and Fertility of Soils 35: 247-252. LUIZO, F., PROCTOR, J., THOMPSON, J., LUIZ O, R. C. C., MARRS, R. H., SCOTT, D. A., & VIANA, V. 1998. Rain forest on Mara c Island, Roraima, Brazil: soil and litter process response to artificial gaps. Forest Ecology and Management 102: 291-303. LUIZAO, F. J. 1989. Litter production and mineral el ement input to the forest floor in a central Amazonian forest. GeoJournal 19: 407-417. LUIZO, F. J., & SCHUBART, H. O. R. 1987. Litter production and decomposition in a terrafirme forest of Central Amazonia. Experientia 43: 259-265. LUIZAO, R. C. C., BONDE, T. A., & ROSSWA LL, T. 1992. Seasonal variation of soil microbial biomassthe effects of clearfelling a tropical rainforest and establishment of pasture in the Central Amazon. Soil Biology & Biochemistry 24: 802-813.

PAGE 119

119 LUIZAO, R. C. C., LUIZAO, F. J., PAIVA, R. Q., MONTEIRO, T. F., SOUSA, L. S., & KRUIJT, B. 2004. Variation of carbon a nd nitrogen cycling processes along a topographic gradient in a central Amazonian forest. Global Change Biology 10: 592-600. MALHI, Y., BAKER, T. R., PHILLIPS, O. L., ALMEIDA, S., ALVAREZ, E., ARROYO, L., CHAVE, J., CZIMCZIK, C. I., FIORE, A. D., HIGUCHI, N., KILLEEN, T. J., LAURANCE, S. G., LAURANCE, W. F., LE WIS, S. L., MONTOYA, L. M. M., MONTEAGUDO, A., NEILL, D. A., VARGAS, P. N., PATINO, S., PITMAN, N. C. A., QUESADA, C. A., SALOMAO, R., SILVA, J. N. M., LEZAMA, A. T., MARTINEZ, R. V., TERBORGH, J., VINCETI, B., & LLOYD, J. 2004. The above-ground coarse wood productivity of 104 Neotropical forest plots. Global Change Biology 10: 563-591. MALHI, Y., NOBRE, A. D., GRACE, J., KRUI JT, B., PEREIRA, M. G. P., CULF, A., & SCOTT, S. 1998. Carbon dioxide transfer over a Central Amazonian rain forest. Journal of Geophysical Research 103: 31,593-31,612. MARKEWITZ, D., DAVIDSON, E., MOUTINHO, P., & NEPSTAD, D. 2004. Nutrient loss and redistribution after forest clearing on a highly weathered soil in Amazonia. Ecological Applications 14: S177-S199. MARSCHNER, H. 1995. Mineral nutrition of higher plants. Academic Press, San Diego. MCGRATH, D. A., DURYEA, M. L., & CROPPER W. P. 2001. Soil phosphorus availability and fine root proliferation in Amazonian ag roforests 6 years following forest conversion. Agriculture, Ecosystems and Environment 83: 271-284. MESQUITA, R. DE C. G., WORKMAN, S. W., & NEELY, C. L. 1998. Slow litter decomposition in a Cecropia-dominated s econdary forest of central Amazonia. Soil Biology & Biochemistry 30: 167-175. MIRMANTO, E., PROCTOR, J., GREEN, J., N AGY, L., & SURIANTATA 1999. Effects of nitrogen and phosphorus fertilization in a lowland evergreen rainforest. Philosophycal Transactions of the Royal Society of London B 354: 1825-1829. MORAN, E. F., BRONDIZIO, E. S., TUCK ER, J. M., SILVA-FORSBERG, M. C. DA, MCCRACKEN, S., & FALESI, I. 2000. Effects of soil fertility and land-use on forest succession in Amaznia. Forest Ecology and Management 139: 93-108. MULKEY, S. S., & WRIGHT, S. J. 1996. Influen ce of seasonal drought on the carbon balance of tropical forest plants. Pp. 187-216. in MU LKEY, S. S., CHAZDON, R., & SMITH, A. P. (ed.). Tropical forest plant ecophysiology. Chapman and Hall, New York. MURPHY, J., & RILEY, J. P. 1962. A modified single solution method fo r determination of phosphate in natural waters. Analytica Chimica Acta 27: 31-36. NEEFF, T. 2005. Spatial modeling of primary and secondary forest growth in Amazonia. Forest Ecology and Management 219: 149-168.

PAGE 120

120 NEEFF, T., LUCAS, R. M., SANTOS, J. R. DO S, BRONDIZIO, E. S., & FREITAS, C. C. 2006. Area and age of secondary forests in Brazilian Amazonia 1978: an empirical estimate. Ecosystems 9: 609-623. NEILL, C., PICCOLO, M. C., CERRI, C. C., STEUDLER, P. A., MELILLO, J. M., & BRITO, M. 1997. Net nitrogen mineralization and net nitrification rates in soils following deforestation for pasture across the southw estern Brazilian Amazon Basin landscape. Oecologia 110: 243-252. NEPSTAD, D., MOUTINHO, P. R. S., & MARKEW ITZ, D. 2001. The recovery of biomass, nutrient stocks, and deep soil functions in secondary forests. Pp. 139-155. in MCCLAIN, M. E., VICTORIA, R. L., & RICHEY, J. E. (ed.). The biogeochemistry of the Amazon basin. Oxford University Press, New York. NEPSTAD, D. C., CARVALHO, C. J. R. DE, D AVIDSON, E., JIPP, P. H., LEFEBVRE, P. A., NEGREIROS, G. H., SILVA, E. D. D., ST ONE, T. A., TRUMBORE, S. E., & VIEIRA, S. 1994. The role of deep roots in the hydrological and carbon cycles of Amazonian forests and pastures. Nature 372: 666-669. NEPSTAD, D. C., MOUTINHO, P., DIAS-FILHO M. B., DAVIDSON, E., CARDINOT, G., MARKEWITZ, D., FIGUEIREDO, R., VI ANNA, N., CHAMBERS J., RAY, D., GUERREIROS, J. B., LEFEBVRE, P., S TERNBERG, L., MOREIRA, M., BARROS, L., ISHIDA, F. Y., TOHLVER, I., BELK, E., KALIF, K., & SCHWALBE, K. 2002. The effects of partial throughfall exclusio n on canopy processes, aboveground production, and biogeochemistry of an Amazon forest. Journal of Geophysical Research 107(D20): doi:10.1029/2001JD000360. NEPSTAD, D. C., VERSSIMO, A., ALENCAR, A., NOBRE, C., LIMA, E., LEFEBVRE, P., SCHLESINGER, P., POTTER, C., MOUTINHO P., MENDOZA, E., COCHRANE, M., & BROOKS, V. 1999. Large-scale impoverishme nt of Amazonian forests by logging and fire. Nature 398: 505-508. NEWMAN, G. S., ARTHUR, M. A., & MULLER, R. N. 2006. Aboveand belowground net primary production in a temper ate mixed deciduous forest. Ecosystems 9: 317-329. NOBRE, A. D., KELLER, M., CRILL, P. M., & HARRISS, R. C. 2001. Short-term nitrous oxide profile dynamics and emissions response to water, nitrogen and carbon additions in two tropical soils. Biology and Fertility of Soils 34: 363-373. OLIVEIRA, F. DE A. 2005. Biogeoqumica comparativa de floresta sucessional e Virola surinamensis na regio dos tabuleiros co steiros do esturio guajarino, Amaznia oriental, Brasil. Doctoral dissertation, Universi dade Federal do Par, Belm. OLIVEIRA, R. S., DAWSON, T. E., BURGESS, S. S. O., & NEPSTAD, D. C. 2005. Hydraulic redistribution in three Amazonian trees. Oecologia 145: 354-363. OLSON, J. S. 1963. Energy storage and the bala nce of producers and decomposers in ecological systems. Ecology 44: 322-331.

PAGE 121

121 OSTERTAG, R. 1998. Belowground effects of canopy gaps in a tropical wet forest. Ecology 79: 1294-1304. PALM, C. A., ALEGRE, J. C., AREVALO, L., MUTUO, P. K., MOSIER, A. R., & COE, R. 2002. Nitrous oxide and methane fluxes in six different land use systems in the Peruvian Amazon. Global Biogeochemical Cycles 16: 1073, doi:10.1029/2001GB001855. PAUSTIAN, K., GREN, G. I ., & BOSATTA, E. 1997. Modelli ng litter quality effects on decomposition and soil organic matter dynamics. Pp. 313-335. in CADISCH, G., & GILLER, K. E. (ed.). Driven by nature: Plant li tter quality and decomposition. CABI Publishing, Wallingford. PHILLIPS, O. L., MALHI, Y., HIGUCHI, N., LAURANCE, W. F., NEZ, P. V., VSQUEZ, R. M., LAURANCE, S. G., FERREIRA, L. V., STERN, M., BROWN S., & GRACE, J. 1998. Changes in the carbon balan ce of tropical forests: evid ence from long-term plots. Science 282: 439-442. POTTER, C., KLOOSTER, S., CARVALHO, C. R. DE, GENOVESE, V. B., TORREGOSA, A., DUNGAN, J., BOBO, M., & COUGHL AN, J. 2001. Modeling seasonal and interannual variability in ecosystem carbon cycling for the Brazilian Amazon region. Journal of Geophysical Research 106: 10,423-10,446. POTTER, C., KLOOSTER, S., ST EINBACH, M., TAN, P.-N., KUMAR, V., SHEKHAR, S., & CARVALHO, C. R. D. 2004. Understanding global teleconnections of climate to regional model estimates of Amazon ecosystem carbon fluxes. Global Change Biology 10: 693-703. POTTER, C. S., & KLOOSTER, S. A. 1998. In terannual variability in soil trace gas (CO2, N2O, NO) fluxes and analysis of controll ers on regional to global scales. Global Biogeochemical Cycles 12: 621-635. PRENTICE, I. C., & LLOYD, J. 1998. C-quest in the Amazon basin. Nature 396: 619-620. PRESCOTT, C. E. 2005. Do rates of litter deco mposition tell us anything we really need to know? Forest Ecology and Management 220: 66-74. PRIESS, J. A., & FLSTER, H. 2001. Microbial properties and soil respiration in submontane forests of Venezuelian Guyana: characteristi cs and response to fertilizer treatments. Soil Biology and Biochemistry 33: 503-509. PROCTOR, J. 1983. Tropical forest litterfall. I. Problems of data comparison. Pp. 267-273. in SUTTON, S. L., WHITMORE, T. C ., & CHADWICK, A. C. (ed.). Tropical rain forest: ecology and management. Blackwell Scientific Publications, Oxford. RAICH, J. W., & NADELHOFFER, K. J. 1989. Belowground carbon allocation in forest ecosystems: global trends. Ecology 70: 1346-1354.

PAGE 122

122 RAICH, J. W., RILEY, R. H., & VITOUSEK, P. M. 1994. Use of root-ingrowth cores to assess nutrient limitations in forest ecosystems. Canadian Journal of Forest Research 24: 21352138. RAICH, J. W., & SCHLESINGER, W. H. 1992. The global carbon dioxide flux in soil respiration and its relationshi p to vegetation and climate. Tellus 44B: 81-99. RANGEL-VASCONCELOS, L. G. T. 2002. Biomassa microbiana de solo sob vegetao secundria na Amaznia oriental. Master thesis, Faculdade de Cincias Agrrias do Par, Belm. RASCHER, U., BOBICH, E. G., LIN, G. H., WALTER, A., MORRIS, T., NAUMANN, M., NICHOL, C. J., PIERCE, D., BIL, K., KUDEYAROV, V., & BERRY, J. A. 2004. Functional diversity of photos ynthesis during drought in a model tropical rainforestthe contributions of leaf area, photosynthetic electron transpor t and stomatal conductance to reduction in net ecosystem carbon exchange. Plant, Cell and Environment 27: 12391256. READ, L., & LAWRENCE, D. 2003. Litter nutrie nt dynamics during succession in dry tropical forests of the Yucatan: regional and seasonal effects. Ecosystems 6: 747-761. REICH, P. B., ELLSWORTH, D. S., & UHL, C. 1995. Leaf carbon and nutrient assimilation and conservation in species of differing succession al status in an o ligotrophic Amazonian forest. Functional Ecology 9: 65-76. REICH, P. B., WALTERS, M. B., ELLSWORTH, D. S., & UHL, C. 1994. Photosynthesisnitrogen relations in Amazonian tree species I. Patterns among species and communities. Oecologia 97: 62-72. REY, A., PEGORARO, E., TEDESCH I, V., PARRI, I. D., JARV IS, P. G., & VALENTINI, R. 2002. Annual variation in soil respiration and its components in a coppice oak forest in Central Italy. Global Change Biology 8: 851-866. RICE, A. H., PYLE, E. H., SALESKA, S. R., HUTYRA, L ., PALACE, M., KELLER, M., CAMARGO, P. B. DE., PORTILHO, K., MA RQUES, D. F., & WOFSY, S. C. 2004. Carbon balance and vegetation dynamics in an old-growth Amazonian forest. Ecological Applications 14: S55-S71. ROMERO-SALTOS, H., STERNBERG, L. DA S. L., MOREIRA, M. Z., & NEPSTAD, D. C. 2005. Rainfall exclusion in an eastern Amazoni an forest alters soil water movement and depth of water uptake. American Journal of Botany 92: 443-455. ROY, J., & SAUGIER, B. 2001. Te rrestrial primary productivity: de finitions and milestones. Pp. 1-6. in ROY, J., SAUGIER, B., & MOONEY, H. A. (ed.). Terrestrial global productivity. Academic Press, San Diego.

PAGE 123

123 SALDARRIAGA, J. G., WEST, D. C., THAR P, M. L., & UHL, C. 1988. Long-term chronosequence of forest succession in the uppe r Rio Negro of Colombia and Venezuela. Journal of Ecology 76: 938-958. SALESKA, S. R., MILLER, S. D., MATROSS, D. M., GOULDEN, M. L., WOFSY, S. C., ROCHA, H. R. DA, CAMARGO, P. B. DE, CR ILL, P., DAUBE, B. C., FREITAS, H. C. DE, HUTYRA, L., KELLER, M., KIRCHHOFF V., MENTON, M., MUNGER, J. W., PYLE, E. H., RICE, A. H., & SILVA, H. 2003. Carbon in Amazon forests: Unexpected seasonal fluxes and disturbance-induced losses. Science 302: 1554-1557. SALIMON, C. I., DAVIDSON, E. A., VICTOR IA, R. L., & MELO, A. W. F. 2004. CO2 flux from soil in pastures and fore sts in southwestern Amazonia. Global Change Biology 10: 833-843. SANCHEZ, G., & ALVAREZ-SANCHEZ, J. 1995. L itterfall in primary and secondary tropical forests of Mexico. Tropical Ecology 36: 191-201. SANCHEZ, P. A. 1976. Properties and management of soils in the tropics. John Wiley and Sons, New York. SANTIAGO, L. S. 2003. Leaf traits of canopy trees on a precipitation gradient in Panama: integrating plant physiological ecology and ecosystem science. Doctoral dissertation, University of Florida, Gainesville. SAYER, E. J. 2005. Leaf litter manipulati on in a tropical forest. Doctoral dissertation, University of Cambridge, Cambridge. SAYER, E. J., TANNER, E. V. J., & CHEESMAN, A. W. 2006. Increased litterfall changes fine root distribution in a moist tropical forest. Plant and Soil 281: 5-13. SCHUUR, E. A. G. 2001. The effect of wa ter on decomposition dynamics in mesic to wet Hawaiian montane forests. Ecosystems 4: 259-273. SCHUUR, E. A. G., & MATSON, P. A. 2001. Ne t primary productivity and nutrient cycling across a mesic to wet precipitation gr adient in Hawaiian montane forest. Oecologia 128: 431-442. SCHWENDENMANN, L., VELDKAMP, E., BREN ES, 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. SCOTT, D. A., PROCTOR, J., & THOMPSON J. 1992. Ecological studies on a lowland evergreen rain forest on Marac Island, Roraim a, Brazil. II. Litter and nutrient cycling. Journal of Ecology 80: 705-717.

PAGE 124

124 SHUTTLEWORTH, W. J., GASH, J. H. C., LLO YD, 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. Quarterly Journal of the Royal Meteorological Society 110: 1143-1162. SILVER, W. L., NEFF, J., MCGRODDY, M., VELDKAMP, E., KELLER, M., & COSME, R. 2000. Effects of soil texture on belowground carbon and nutrient storage in a lowland Amazonian forest ecosystem. Ecosystems 3: 193-209. SILVER, W. L., THOMPSON, A. W., MCGRODDY, M. E., VARNER, R. K., DIAS, J. D., SILVA, H., CRILL, P. M., & KELLER, M. 20 05. Fine root dynamics and trace gas fluxes in two lowland tropical forest soils. Global Change Biology 11: 290-306. SMITH, K., GHOLZ, H. L., & OLIVEIRA, F. DE A. 1998. Litterfall and nitrogen-use efficiency of plantations and primary forest in the eastern Brazilian Amazon. Forest Ecology and Management 109: 209-220. SOMMER, R., S, T. D. DE A., VIELHAUER, K., ARAJO, A. C. DE, FLSTER, H., & VLEK, P. L. G. 2002. Transpiration and ca nopy conductance of secondary vegetation in the eastern Amazon. Agricultural and Forest Meteorology 112: 103-121. SONGWE, N. C., OKALI, D. U. U., & FASE HUN, F. E. 1995. Litter decomposition and nutrient release in a tropical rainforest, Southern Bakundu Forest Reserve, Cameroon. Journal of Tropical Ecology 11: 333-350. SOTTA, E. D., MEIR, P., MALHI, Y., NOBRE A. D., HODNETT, M., & GRACE, J. 2004. Soil CO2 efflux in a tropical forest in the central Amazon. Global Change Biology 10: 601-617. SPERRY, J. S. 2000. Hydraulic c onstraints on plant gas exchange. Agricultural and Forest Meteorology 104: 13-23. STEININGER, M. K. 2004. Net carbon fluxes from forest clearance and regrowth in the Amazon. Ecological Applications 14: S313-S322. SULZMAN, E. W., BRANT, J. B., BOWDEN, R. D., & LAJTHA, K. 2005. Contribution of aboveground litter, belowground litter, and rhizosphere respiration to total soil CO2 efflux in an old growth coniferous forest. Biogeochemistry 73: 231-256. SWIFT, M. J., HEAL, O. W., & ANDERSON, J. M. 1979. Decomposition in terrestrial ecosystems. University of California Press, Berkeley. TAIZ, L., & ZEIGER, E. 1998. Plant physiology. Sinauer, Sunderland. TANNER, E. V. J., VITOUSEK, P. M., & CUEVAS E. 1998. Experimental investigation of nutrient limitation of forest grow th on wet tropical mountains. Ecology 79: 10-22.

PAGE 125

125 TENRIO, A. R. D. M., GRAA, J. J. D. C., G ES, 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. TIAN, H., MELILLO, J. M., KI CKLIGHTER, D. W., MCGUIRE, A. D., HELFRICH III, J., MOORE III, B., & VRSMARTY, C. J. 2000. Climatic and biotic controls on annual carbon storage in Amazonian ecosystems. Global Ecology & Biogeography 9: 315-335. TIAN, H., MELLILO, J. M., KICK LIGHTER, D. W., MCGUIRE, A. D., HELFRICH III, J. V. K., MOORE III, B., & VRSMARTY, C. J. 1998. Effect of interannual climate variability on carbon storage in Amazonian ecosystems. Nature 396: 664-667. TRENBERTH, K. E., & HOAR, T. J. 1997. El Nio and climate change. Geophysical Research Letters 24: 3057-3060. UHL, C. 1987. Factors controlli ng succession following slash-and-bur n agriculture in Amazonia. Journal of Ecology 75: 377-407. UHL, C., BUSCHBACHER, R., & SERRO, E. A. S. 1988. Abandoned pastures on Eastern Amazonia. I. Patterns of plant succession. Journal of Ecology 76: 663-681. VASCONCELOS, H. L., & LAURANCE, W. F. 2005. In fluence of habitat, litter type, and soil invertebrates on leaf-litter decomposition in a fragmented Amazonian landscape. Oecologia 144: 456-462. VASCONCELOS, S. S., ZARIN, D. J., CA PANU, M., LITTELL, R., DAVIDSON, E. A., ISHIDA, F. Y., SANTOS, E. B., ARA JO, M. M., ARAGO, D. V., RANGELVASCONCELOS, L. G. T., OLIVEIRA, F. DE A., MCDOWELL, W. H., & CARVALHO, C. J. R. DE. 2004. Moisture and substrate availa bility constrain soil trace gas fluxes in an eastern Amazonian regrowth forest. Global Biogeochemical Cycles 18: GB2009, doi:10.1029/2003GB002210. VASCONCELOS, S. S., ZARIN, D. J., MULK EY, S. S., CARVALHO, C. J. R. DE, & FORTINI, L. B. (2002). Water use efficien cy increases in resp onse to drought for Vismia guianensis in the overstory of an Eastern Amazoni an regrowth forest. Pp. 209. in. (ed.). International LBA Meeting, Manaus, Brazil. VELUCI, R. M., ZARIN, D. J ., MCDOWELL, W. H., CARVALHO C. R., & GEZAN, S. In preparation. Effects of irrigati on and litter removal on nutrient availability in a seasonally dry tropical forest. Soil Biology and Biochemistry. VERCHOT, L. V., DAVIDSON, E. A., CATTN IO, J. H., & ACKERMAN, I. L. 2000. Landuse change and biogeochemical controls of me thane fluxes in soils of eastern Amazonia. Ecosystems 3: 41-56.

PAGE 126

126 VERCHOT, L. V., DAVIDSON, E. A., CATTNIO J. H., ACKERMAN, I. L., ERICKSON, H. E., & KELLER, M. 1999. Land use change a nd biogeochemical controls of nitrogen oxide emissions from soils in eastern Amazonia. Global Biogeochemical Cycles 13: 3146. VIEIRA, S., CAMARGO, P. B. DE, SELH ORST, D., SILVA, R. DA, HUTYRA, L., CHAMBERS, J. Q., BROWN, I. F., HIGUCHI N., SANTOS, J. DOS, WOFSY, S. C., TRUMBORE, S. E., & MARTINELLI, L. A. 2 004. Forest structure and carbon dynamics in Amazonian tropical rain forests. Oecologia 140: 468-479. VITOUSEK, P. M. 1984. Litterfall, nutrient cycli ng, and nutrient limitation in tropical forests. Ecology 65: 285-298. VITOUSEK, P. M., & FARRINGTON, H. 1997. Nutrient limitation and soil development: Experimental test of a biogeochemical theory. Biogeochemistry 37: 63-75. VITOUSEK, P. M., & MATSON, P. A. 1992. Tr opical forests and trace gases: potential interactions between tropical biol ogy and the atmospheric sciences. Biotropica 24: 233239. VITOUSEK, P. M., & SANFORD, J., R L 1986. Nu trient cycling in moist tropical forest. Annual Review of Ecology and Systematics 17: 137-167. WHITMORE, T. C. 1992. An introduction to tropical rain forests. Oxford University Press, Oxford. WIEDER, R. K., & LANG, G. E. 1982. A critique of the analytical methods used in examining decomposition data obtained from litter bags. Ecology 63: 1636-1642. WIEDER, R. K., & WRIGHT, S. J. 1995. Tropi cal forest litter dynamics and dry season irrigation on Barro Colorado Island, Panama. Ecology 76: 1971-1979. WILLIAMSON, G. B., LAURANCE, W. F ., OLIVEIRA, A. A ., DELAMNICA, P., GASCON, C., LOVEJOY, T. E., & POHL, L. 2000. Amazonian tree mortality during the 1997 El Nio drought. Conservation Biology 14: 1538-1542. WOOD, T. E., LAWRENCE, D., & CLARK, D. A. 2005. Variation in leaf litter nutrients of a Costa Rican rain forest is related to precipitation. Biogeochemistry 73: 417-437. WRIGHT, S. J., & CORNEJO, F. H. 1990. Seasona l drought and leaf fall in a tropical forest. Ecology 71: 1165-1175. YAVITT, J. B., & WRIGHT, S. J. 1996. Temporal patterns of soil nutrients in a Panamanian moist forest revealed by ion-exchange resin and experime ntal irrigation. Plant and Soil 183: 117-129. YAVITT, J. B., & WRIGHT, S. J. 2001. Drought and irrigation effects on fine root dynamics in a tropical moist forest, Panama. Biotropica 33: 421-434.

PAGE 127

127 YAVITT, J. B., WRIGHT, S. J., & WIEDER, R. K. 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. ZAGT, R. J. 1997. Pre-dispersal and early postdispersal demography, and reproductive litter production, in the tropical tree Dicymbe altsonii in Guyana. Journal of Tropical Ecology 13: 511-526. ZARIN, D. J., DAVIDSON, E. A ., BRONDIZIO, E., VIEIRA, I. C. G., S, T., FELDPAUSCH, T., SCHUUR, E. A. G., MESQUITA, R., MORAN, E., DELAMONICA, P., DUCEY, M. J., HURTT, G. C., SALIMON, C., & DENICH M. 2005. Legacy of fire slows carbon accumulation in Amazonian forest regrowth. Frontiers in Ecology and Environment 3: 365-369. ZARIN, D. J., DUCEY, M. J., TUCKER, J. M ., & SALAS, W. A. 2001. Potential biomass accumulation in Amazonian regrowth forests. Ecosystems 4: 658-668.

PAGE 128

128 BIOGRAPHICAL SKETCH Steel Silva Vasconcelos was born on August 4, 1972, in Rio de Janeir o, RJ, Brazil. He attended the University Federal Rural do Rio de Janeiro (UFRRJ) where he graduated in agronomic engineering in 1995. Continuing at th e same university, Steel received a masters degree in soil science in 1997 studyi ng the tolerance of rice plants to aluminium toxicity. After graduation, he moved with his wi fe Lvia and daughter Crita to Par, north of Brazil, where he worked for about two years on th e selection of maize genotypes suitable for slash-and-mulch agriculture. In June 1999, Steel started to work as a research assistant for a forest ecology project coordinated by Dr. Daniel Zarinthe MANFLORA project ba sed out of Castanhal, Par. Then, in 2002, he moved to Gainesville, FL, with his family to start his Ph.D. program in the School of Forest Resources and Conservation at the University of Florida (UF) under the supervision of Dr. Zarin. His doctoral research was developed at the MANFLORA experimental site. During his Ph.D. program, Steel was hired by the Brazilian Agricultural Research Center (EMBRAPA) as a researcher to study soil-plan t-water relationships at the Eastern Amazon Research Center in Belm, Par. Upon comple ting his Ph.D. program, he intends to continue working with tropical forest ecology as pa rt of his research duties at EMBRAPA.


Permanent Link: http://ufdc.ufl.edu/UFE0016620/00001

Material Information

Title: Moisture and Nutrient Constraints to Ecosystem Processes in a Forest Regrowth Stand in Eastern Amazonia, Brazil
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0016620:00001

Permanent Link: http://ufdc.ufl.edu/UFE0016620/00001

Material Information

Title: Moisture and Nutrient Constraints to Ecosystem Processes in a Forest Regrowth Stand in Eastern Amazonia, Brazil
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0016620:00001


This item has the following downloads:


Full Text





MOISTURE AND NUTRIENT CONSTRAINTS TO ECOSYSTEM PROCESSES IN A
FOREST REGROWTH STAND IN EASTERN AMAZONIA, BRAZIL




















By

STEEL SILVA VASCONCELOS


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

2006

































Copyright 2006

by

Steel Silva Vasconcelos
































This work is dedicated to my wife Livia and my daughter Carita.









ACKNOWLEDGMENTS

I am especially thankful to Livia and Carita for tolerating my physical and sometimes

mental absences during critical periods throughout this challenging journey that began in July

1999, as well as for their help during fieldwork. I thank my parents for their efforts to provide

me with an adequate education that served as a foundation to meet the requirements of the PhD

program at the University of Florida.

I am very grateful to my advisor, Dr. Daniel Zarin, for believing that I could succeed in the

program and especially for the efficient and invaluable guidance and support that were critical to

my academic development. I also thank the other members of my committee-Dr. Eric

Davidson, Dr. Steve Mulkey, Dr. Ted Schuur, and Dr. Wendell Cropper-for their input during

the program and insightful suggestions that greatly improved this dissertation.

I thank the School of Forest Resources and Conservation for overall support, the Andrew

Mellon Foundation for research funding, and Empresa Brasileira de Pesquisa Agropecuaria

(EMBRAPA) for support to conclude the program.

This ecosystem-level work would not have been possible without the help of many

research assistants, technicians, and undergraduate students of the MANFLORA Project in

Brazil (Wilson Oliveira, Beatriz Rosa, Joanna Tucker, Lucas Fortini, Roberta Veluci-Marlow,

Debora Aragao, Gizelle Benigno, Ronaldo Oliveira, Tdmara Lima, Roberta Coelho, Alexandre

Modesto, Elisdngela Santos, Ana Julia Amaral, and Evandro da Silva) and in the US (Leandra

Aragao and Patricia Sampaio) who contributed to data collection and processing and insightful

discussions. I offer extra special thanks to Dr. Maristela Araujo for data collection and

processing, to Os6rio Oliveira, Glebson Sousa, and "Paulo" Alencar for their assistance in the

field, to Raimundo Nonato da Silva (UFRA) and his team (Manoel, Gilson, Geraldo Jr., and Seu

Geraldo) for logistical support, and to the Brazilian coordinators of MANFLORA-Prof.









Francisco de Assis Oliveira and Prof. Izildinha Miranda. Laboratory support at Embrapa Eastern

Amazon by Dr. Claudio Carvalho and his team (especially Ivanildo Trindade), as well by Dr.

Marcus Vasconcelos, is greatly appreciated. Thanks go also to Dr. Marinela Capanu and Prof.

Ramon Littell for helping with statistical analysis.

Finally, I would like to thank all the friends in Gainesville that made the days for me and

my family extremely pleasant: Camila and Guto Paula; Carolina and Victor; Darlene, Eduardo,

Luiza and Ligia Carlos; Ge6rgia, Jose Carlos and Vitor Dubeux; Graziela, Lucinda and Roberta

Noronha; Joanna Tucker; Juliana and Flavio Silvestre; Lucas Fortini; Mara, Luis, Igor, Andre

and Raissa Lima; Mrs. Lilian Raye; Patricia and Emilio Bruna; Marcelo and Aline Carvalho;

Roberta and Brian Marlow; and Rutecleia Zarin.









TABLE OF CONTENTS


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

L IS T O F T A B L E S ................................................................................................. ..................... 8

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

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

CHAPTER

1 INTRODUCTION ................ ... ................. .. ........... ..................................... 12

L iteratu re R ev iew ................ .... ... .... .. ..... .... ............................ ...................................14
M oisture and Nutrient Limitations to Tropical Forests.............................. ............... 14
Observational and Manipulative Experiments to Study Moisture and Nutrient
L im stations in Tropical F orests............................................................. .... ................ 15
Moisture Effects on Ecosystem Processes in Tropical Forests .................................18
A boveground processes........................................... ......................... ............... 19
B elow ground processes................... ............... .. ....... ................ 22
Nutrient Effects on Ecosystem Processes in Tropical Forests ..................................23
A boveground processes........................................... ......................... ................ 24
B elow ground processes........................................... ......................... ................ 25
Conclusions ..................................................... .................. 26

2 STUDY SITE AND EXPERIMENTAL DESIGN............................................................28

S tu d y S ite ................................................................................................................................ 2 8
Experimental Design .............................................. ............................. 29

3 SEASONAL AND EXPERIMENTAL EFFECTS ON LITTERFALL QUANTITY
AND QUALITY IN EASTERN AMAZONIAN FOREST REGROWTH ........................34

Introduction ..................................................... .................. 34
M material and methods ............................. .. ........... .....................................35
L itte rf a ll ........................................................................................................................... 3 5
L itte r S to c k ......................................................................................................................3 7
S tatistic al A n aly sis ..........................................................................................................3 7
R results ......................................................................................................38
N on-w oody Litterfall .................................................................................................. 38
Irrigation experim ent .............................................................................................38
Litter rem oval experim ent ..................................................................................40
Litter stock ............................................. ............................... 41
Discussion ...................................................... .................. 42
S e a so n a l P atte rn s .............................................................................................................4 2
Limited Impact of Dry-season Irrigation......................................................................43
Litter Removal Reduces Litterfall N Concentration ................................................. 45

6









4 LEAF DECOMPOSITION IN A DRY SEASON IRRIGATION EXPERIMENT IN
EASTERN AMAZONIAN FOREST REGROWTH .................................. ..................... 60

In tro d u c tio n ............................................................................................................................. 6 0
Study Site and Experim mental D design ................. ......................................................... 61
M material an d M eth od s ............................................................................................................. 6 1
L eaf L itter D ecom position .. .................................................................... ................ 61
Initial L eaf L itter C hem istry ........................................... ......................... ................ 63
Sp ecific L eaf A rea ..................................................... .............................................. 63
S statistical A n aly sis .......................................................................................................... 6 4
R esu lts ........................................................................................................... 64
D iscu ssio n .............................................................................................................. ........ .. 6 5

5 MOISTURE AND SUBSTRATE AVAILABILITY CONSTRAIN SOIL TRACE GAS
FLUXES IN AN EASTERN AMAZONIAN REGROWTH FOREST...............................74

In tro du ctio n ............................................................................................................ ........ .. 7 4
M material an d M eth od s ............................................................................................................. 76
F field M easurem ents ............. .. ................... .................. .............. ......... ... ............ 76
S statistical A n aly ses .......................................................................................................... 7 8
R e su lts ................................................................................... ........... ... ..................... 7 9
Irrigation Experiment ........................ .. ........... .....................................79
L itter R em oval Experim ent .................................................................... ................ 82
D iscu ssion ............... .. ....... .. .......................................................................... . 83
Soil CO2 Efflux and Belowground C Allocation ............... ................................... 83
N itrogen O xide E m issions.. .................................................................. ................ 87
M ethane E m missions .............. .. .................. .................. ............ ........ .... ............... 88
C o n clu sio n s............................................................................................................ ........ .. 8 8

6 MOISTURE CONSTRAINTS TO ABOVEGROUND NET PRIMARY
PRODUCTIVITY IN EASTERN AMAZONIAN FOREST REGROWTH .........................97

In tro du ctio n ............... .... ...................................................................................... ........ .. 9 7
Study Site and Experim mental D design ................. ......................................................... 98
M material and M methods ............................................................................................................... 98
A boveground N et Prim ary Productivity..................................................... ................ 98
S statistical A n aly sis .......................................................................................................... 9 9
R esu lts ........................................................................................................... 99
D discussion .................................................................................................... 100

7 CONCLUSIONS .................................... ........... .................... ....... .. 106

L IST O F R E F E R E N C E S ....................................................... ................................................ 109

B IO G R A PH IC A L SK E T C H .................................................... ............................................. 128









LIST OF TABLES


Table page

2-1 Characteristics of rainfall distribution and intensity during the experimental period in
th e site ............................................................................................................. ....... .. 3 2

2-2 Dry-season irrigation intervals and associated rainfall intensity and distribution.............32

3-1 F statistics and associated significance levels for the effects of treatments, sampling
date, and their interaction on non-woody litterfall mass and nutrients...........................47

3-2 Estimates of annual non-woody litterfall (mass, nitrogen, and phosphorus), non-
woody litter stock, and litterfall:forest floor mass ratio (kL) in tropical forests .............48

4-1 F statistics and significance levels for the effects of treatments (control and
irrigation), species, and their interactions on leaf litter decomposition rates .................68

4-2 Decomposition rates for overstory species in a tropical regrowth forest stand in
eastern A m azonia, B razil ................................................. ............................................ 68

4-3 Decomposition rates for overstory species under control and irrigated plots
(Experiment 3) in a tropical regrowth forest stand in eastern Amazonia, Brazil ..............69

4-4 Initial quality parameters of leaves incubated in litterbags for decomposition studies
in a tropical regrowth forest in Eastern Amazonia, Brazil............................................69

4-5 Pearson correlation coefficients between decomposition rate (k) and initial quality
parameters of leaves of overstory tree species incubated in litterbags............................. 70

4-6 Decomposition rates estimated from litterbag studies for some tropical forest sites ........70

4-7 Chemical composition of leaf litter for some tropical forest sites...............................71

5-1 F statistics and significance levels for the effect of treatments, sampling date, and
their interaction on soil trace gas fluxes and non-woody litterfall................................90

5-2 Annual soil carbon efflux and non-woody litterfall carbon for control, irrigated and
litter rem ov al p lots ............................................................................................................. 9 1

5-3 Estimates of annual soil carbon efflux in old-growth and regrowth tropical forests.........92

6-1 Allometric equations used to estimate tree biomass in a tropical regrowth forest stand
in eastern A m azonia, B razil ........................................ .......................... ............... 103

7-1 Summary of ecosystem process responses to intrannual and interannual variability
effects and resource manipulation effects.......................................... 108









LIST OF FIGURES


Figure page

1-1 Simplified conceptual diagram of likely effects of drought on leaf- and ecosystem-
level processes addressed in this dissertation ............................................... ................ 27

2-1 Daily rainfall during the experimental period..................................................... 33

2-2 E xperim mental plot layout.................................................. ............................................ 33

3-1 Effects of rainfall patterns and dry-season irrigation on monthly non-woody litterfall
m ass ........................................................................................................... 50

3-2 Effects of dry-season irrigation and litter removal on annual non-woody litterfall
m ass ........................................................................................................... 5 1

3-3 Effects of rainfall patterns and dry-season irrigation on non-woody litterfall nutrient
con centration .................................................................................................... ........ .. 52

3-4 Effects of rainfall patterns and dry-season irrigation on monthly non-woody litterfall
n u trie n t retu rn .................................................................................................................. ... 5 3

3-5 Effects of dry-season irrigation on annual non-woody litterfall nutrient return............. 54

3-6 Effects of rainfall patterns and litter removal on non-woody litterfall mass .................. 55

3-7 Effects of rainfall patterns and litter removal on monthly non-woody litterfall
nutrient concentration ............... .. .................. ................... ....................... ............... 56

3-8 Effects of rainfall patterns and litter removal on monthly non-woody litterfall
n u trie n t retu rn .................................................................................................................. ... 5 7

3-9 Effects of litter removal on annual non-woody litterfall nutrient return ........................ 58

3-10 Non-woody litter stock for control and irrigation plots................................................59

4-1 Effects of dry-season irrigation on leaf litter decomposition.......................................72

4-2 Relation between decomposition rate and initial leaf litter characteristics for tree
sp e cie s ............................................................................................................. ....... .. 7 3

5-1 Effects of rainfall patterns and dry-season irrigation on soil moisture status and soil
resp iratio n ........................................................................................................ ........ .. 9 3

5-2 Effects of rainfall patterns and dry-season irrigation on soil nitrogen oxide and
methane effluxes .................................... ............ ............................ 94









5-3 Effects of rainfall patterns and litter removal on soil moisture status and soil
resp iratio n ........................................................................................................ ........ .. 9 5

5-4 Effects of rainfall patterns and litter removal on soil nitrogen oxide and methane
efflu x e s ............................................................................................................ ........ .. 9 6

6-1 Effects of dry-season irrigation on aboveground increment and non-woody litterfall.... 104

6-2 Relationship between aboveground net primary productivity and rainfall ....................105









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

MOISTURE AND NUTRIENT CONSTRAINTS TO ECOSYSTEM PROCESSES IN A
FOREST REGROWTH STAND IN EASTERN AMAZONIA, BRAZIL

By

Steel Silva Vasconcelos

December 2006

Chair: Daniel Jacob Zarin
Major Department: Forest Resources and Conservation

Changes in land-use and climate are likely to alter resource (e.g., moisture and nutrient)

availability in tropical forest soils, but quantitative assessment of the role of resource constraints

as regulators of ecosystem processes is rather limited. In this dissertation, moisture and nutrient

availability were altered through dry-season irrigation and bi-weekly aboveground litter removal,

respectively, to study how these resources control aboveground and belowground ecosystem

processes in a forest regrowth stand in the Brazilian Amazon. Moisture availability strongly

constrains soil respiration as indicated by the responses of soil carbon dioxide emissions to soil

wet-up events and dry-season irrigation. Higher moisture availability in irrigated plots also

increased leaf litter decomposition and slightly increased soil nitrous oxide and methane

emissions, but did not alter monthly litterfall quantity and quality, and soil nitric oxide emission.

Litter removal decreased carbon dioxide emissions and litterfall nitrogen concentration, but had

no effects on litterfall quantity, and soil nitrogen oxides and methane emissions. Aboveground

net primary productivity was constrained by moisture availability as indicated by the response of

wood increment to interannual variation in dry season rainfall and to irrigation, suggesting

decreased potential of carbon sequestration from forest regrowth under anticipated scenarios of

reduced rainfall in Amazonia.









CHAPTER 1
INTRODUCTION

In many tropical areas, especially in the Brazilian Amazon, old-growth forests are

increasingly being converted to forest regrowth-also known as secondary or successional

forests-following abandonment of slash-and-burn agriculture and cattle pasture. Fearnside

(1996) estimated that about 50% of the deforested Brazilian Amazon landscape was in some

stage of forest regrowth in 1990. Forest regrowth provides important ecosystem services such as

carbon sequestration, reestablishment of nutrient and water cycles, and maintenance of

biodiversity (Brown & Lugo 1990, Markewitz et al. 2004, Nepstad et al. 2001, Sommer et al.

2002). In addition, these forests often represent an important source of income (woody and non-

woody forest products) to local people (Brown & Lugo 1990). Adequate management of forest

regrowth can represent an important alternative to reduce pressure on old-growth forest sites in

the Amazon region (Brown & Lugo 1990).

Forest regrowth can play an important role in regional and global carbon (C) dynamics

because of their high rates of biomass accumulation-a proxy for C sequestration (Zarin et al.

2001)-although the frequent clearing of regrowth results in small net C uptake compared to

total emissions from deforestation in the Amazon (Steininger 2004). Efforts to determine the

capacity of forest regrowth to sequester C at different spatial and temporal levels have been

pursued by recent modeling efforts (Neeff 2005, Zarin et al. 2001). Analyses of the rates of and

controls on biomass accumulation may contribute to improved modeling of the potential of forest

C sequestration (Johnson et al. 2000). Observational studies have shown that several factors

control the rate of biomass accumulation in tropical regrowth sites; these include land-use history

(including disturbance type and intensity) (Gehring et al. 2005, Moran et al. 2000, Uhl et al.

1988, Zarin et al. 2005), surrounding vegetation, soil fertility (Gehring et al. 1999, Moran et al.









2000), and climate (Zarin et al. 2001). Such a variety of controlling factors complicates

modeling efforts of forest regrowth rates, but a recent synthesis of observational studies has

shown that soil texture and dry season length are strongly correlated with C accumulation by

forest regrowth, possibly through their effects on the availability of soil moisture and nutrients

(Zarin et al. 2001).

Further understanding of the processes by which moisture and nutrient availability

constrain biomass accumulation rates may be critical to understand the role of forest regrowth on

regional and global C dynamics under future land-use and climate change scenarios.

Manipulative experiments are required to better comprehend the role of resource availability on

forest ecosystem processes. Unfortunately such experiments have rarely been employed to study

ecosystem processes in tropical forests in general and especially in Amazonian forest regrowth,

which represents an important component of the landscape in the region (Fearnside 1996, Neeff

et al. 2006, Zarin et al. 2001).

This study is part of the MANFLORA project (Manipulation of Moisture and Nutrient

Availability in Young Regrowth Forests in Eastern Amazonia), which is a collaborative research

program among the University of Florida, the Universidade Federal Rural da Amaz6nia (Federal

Rural University of Amazonia-UFRA), and EMBRAPA Amaz6nia Oriental (EMBRAPA

Eastern Amazon) initiated in 1999 at the UFRA experimental station in Castanhal, Para, Brazil.

Since 2001, moisture and nutrient availability have been altered in two separate experiments: (1)

dry-season irrigation and (2) continuous litter removal. To my knowledge, the MANFLORA

project represents the only long-term, large-scale (stand-level) experimental manipulation of

moisture availability in tropical forest regrowth, and one of the few tropical forest regrowth

nutrient manipulation studies.









In this dissertation, I examine the influence of resource availability on carbon and nutrient

dynamics associated with litterfall, leaf litter decomposition, soil trace gases, and aboveground

net primary productivity (ANPP). The dissertation is divided into 7 chapters, including this

general introduction and literature review. The second chapter describes the study site and

experimental design. The effects of moisture and nutrient availability on litterfall (Chapter 3),

leaf decomposition (Chapter 4), soil trace gases (Chapter 5), and ANPP (Chapter 6) are

examined separately. Conclusions are summarized in Chapter 7.

Literature Review

This literature review addresses the role of moisture and nutrient availability as constraints

on ecosystem processes related to C dynamics in tropical sites, with special emphasis on forest

regrowth sites in the Brazilian Amazon. The growth of adult trees is the main focus of this

review since they contribute the most to stand-level C balance, although the effects of abiotic

stresses on the understory [e.g., drought (Aragdo et al. 2005, Fortini et al. 2003)] may be more

dramatic than on overstory plants. Finally, the review is directed to moist, lowland evergreen

tropical forests (Whitmore 1992), even though relevant information from dry, deciduous tropical

forests is also included, since there is a great deal of literature on rainfall effects on tree growth

for deciduous sites.

Moisture and Nutrient Limitations to Tropical Forests

Tropical forest formations occupy areas with limited variation in temperature but a wide

range in rainfall intensity and distribution (Whitmore 1992), giving rise to differing degrees in

deciduousness. Since tropical lowland evergreen forests are characterized by high annual

rainfall and evergreeness, and often occur on highly weathered, dystrophic soils (Sanchez 1976),

previous research has mainly overlooked the effects of moisture on forest processes and,

therefore, has focused on nutrient limitations and mechanisms of nutrient conservation (Herrera









et al. 1978, Jordan 1983). However, more recent studies have demonstrated that ecosystem

processes in tropical forests, including Amazonian forests, can be substantially affected by strong

seasonality in rainfall (e.g., Keller et al. 2004). During the dry season (monthly precipitation <

100 mm), old-growth forests in Amazonia rely on deep rooting to retain leaves (Nepstad et al.

1994). In extreme cases, prolonged droughts, usually associated with El Nifio events, can result

in higher tree mortality in tropical old-growth (Condit et al. 1995, Williamson et al. 2000) as

well as regrowth forests (Chazdon et al. 2005), increasing forest susceptibility to fire (Nepstad et

al. 1999).

Observational and Manipulative Experiments to Study Moisture and Nutrient Limitations
in Tropical Forests

Researchers usually rely on observational studies to infer moisture and nutrient limitations

to ecosystem level processes in tropical forests. Measurement of ecosystem processes during

wet and dry seasons (Berish & Ewel 1988, Cornu et al. 1997, Dantas & Phillipson 1989,

Davidson et al. 2000, Scott et al. 1992) or along rainfall gradients (Santiago 2003, Schuur &

Matson 2001) has been used to study moisture constraints. An important limitation of studies

based on rainfall seasonality is the lack of control over factors (e.g., light availability, vapor

pressure deficit, and phenology) that covary with rainfall seasonality and may significantly affect

ecosystem processes.

Nutrient constraints may be investigated by comparing ecosystem processes among

different soil types varying in nutrient availability (e.g., Moran et al. 2000). This approach has

the disadvantages of hindering the identification of the most limiting nutrients) at a specific site

because of inherent differences in soil characteristics (e.g., soil organic matter, pH, structure,

texture), as well as incomplete control over land-use history among sites. Such disadvantages

may be overcome through fertilizer addition in nutrient manipulation studies. Substrate-age









sequences (Vitousek & Farrington 1997) with contrasting nutrient availability have also been

used to study nutrient effects on ecosystem processes, but these sequences are spatially

restricted.

Although observational studies are useful for identifying general trends as well as key

questions for further research, they often do not permit a process-based understanding of

resource control over ecosystem dynamics. An improved understanding of moisture and nutrient

limitations on ecosystem processes may be obtained through manipulative experiments.

Experimental manipulations of moisture availability are usually carried through water addition or

exclusion, whereas nutrient manipulation usually involves fertilizer addition or litter removal

(Eviner et al. 2000, Hanson 2000).

Long-term, large-scale moisture manipulation studies in tropical forests include the dry-

season irrigation study in the Barro Colorado Island station, Panama (Cavelier et al. 1999,

Wieder & Wright 1995, Wright & Cornejo 1990, Yavitt & Wright 2001, Yavitt et al. 2004), and

two throughfall exclusion studies at old-growth sites in the Brazilian Amazon (Carvalho et al.

2005, Nepstad et al. 2002). Dry-season irrigation plots are easier to establish and operate than

throughfall exclusion plots, but the latter are necessary to ultimately simulate the effects of

drought on ecosystem processes. To my knowledge, there are no published reports for large-

scale, long-term moisture manipulative experiments in tropical forest regrowth sites, except for

the studies conducted within the MANFLORA Project (Arag.o et al. 2005, Fortini et al. 2003,

Vasconcelos et al. 2004, Veluci et al. In preparation).

Fertilization experiments are the most common tools used to manipulate nutrient

availability in forests (e.g., Davidson et al. 2004a, Mirmanto et al. 1999, Tanner et al. 1998)

because they are relatively easy and inexpensive (Eviner et al. 2000). In addition, those









experiments are believed to provide the most conclusive evidences of nutrient limitation in forest

ecosystems (Raich et al. 1994). However the interpretation of fertilization experiments may be

confounded by several interactions of nutrients with microorganisms and the soil (nutrient

immobilization by litter microbes, adsorption, trace gas losses, volatilization, and leaching) that

reduce the pool of added nutrients for the plants (Eviner et al. 2000) or that cause secondary

effects such as nutrient imbalance or soil pH alteration (Marschner 1995).

Litter removal is another technique for manipulating nutrient availability in forests. This

technique avoids the problems related to nutrient addition experiments as discussed above, but

litter removal has additional and unavoidable disadvantages including indirect effects on soil

moisture and temperature variation due to the lack of insulation by aboveground litter (Sayer

2005). Also, soil compaction due to trampling (in the case of litter removal by raking) and

raindrop impact occurs in litter removal plots. Such effects may alter soil microorganism

activity and, ultimately, affect soil nutrient availability, representing, therefore, a potential

confounding factor. Microbial activity may be further influenced by reduced input of labile C

(Cleveland et al. 2002) with litter removal.

To my knowledge there is no large-scale, long-term experimental manipulation of nutrient

availability through fertilization in lowland old-growth tropical forests, but there are some

reports for forest regrowth, including a short-term experiment in Costa Rica (Harcombe 1977)

and three other experiments at Amazonian sites. Nutrient addition in Amazonian forest regrowth

has been conducted within both short-term, small-plot (Uhl 1987) and relatively long-term,

large-plot conditions (Davidson et al. 2004a, Gehring et al. 1999); however, there are no reports

of litter removal studies in this region, except for the study conducted within the MANFLORA

Project (Vasconcelos et al. 2004). Thus far, there are only two large-scale, long-term litter









manipulation studies in tropical forests besides the MANFLORA Project. The first study is part

of the Gigante Litter Manipulation Project (GLiMP) in Panama (Sayer 2005) and included

monthly litter raking from 45 m x 45 m plots. The other study excluded litter with tents on 3 m x

3 m plots in a secondary forest in the Luquillo Experimental Forest (Puerto Rico) as part of the

Soil Organic Matter Dynamics Project (Li et al. 2005).

Moisture Effects on Ecosystem Processes in Tropical Forests

Improved knowledge of the mechanisms by which tropical forests respond to drought

stress-both at the plant and community level-is crucial to current understanding and future

projections of forest dynamics, C sequestration, and fire susceptibility in the context of ongoing

land-use and climate changes. Anticipated climate change for the Amazon region may include

more frequent and severe dry seasons in response to global warming (IPCC 2001), deforestation

(Costa & Foley 2000), and more frequent El Nifio episodes (Trenberth & Hoar 1997). Large

scientific research initiatives in the Amazon region since the 1980's including the ABRACOS

Project (Gash et al. 1996) and more recently the LBA Program (Davidson & Artaxo 2004, Keller

et al. 2004) have generated a great deal of relevant information. However, there are few

observational or manipulative studies aimed at investigating moisture controls on Amazonian

forest regrowth.

The understanding of how low soil moisture availability controls tropical forest ecosystem

processes is not straightforward because drought has many direct and indirect effects on plant

and soil organisms (Figure 1-1). Below I review some effects of low soil moisture availability

on above- and belowground processes; forest floor decomposition is included in aboveground

processes for the purposes of this review.









Aboveground processes

Low soil moisture availability can affect aboveground C fluxes in tropical forests in

various direct and indirect interrelated ways, mainly through moisture effects on carbon dioxide

(CO2) assimilation in photosynthesis. The effects of low soil moisture availability on leaf-level

photosynthesis have been the subject of many recent reviews (e.g., Chaves et al. 2003, Lawlor

2002).

The interacting direct effects of drought stress on C assimilation include (a) decrease in

stomatal conductance in response to low soil moisture supply or high vapor pressure deficit,

leading to reduced CO2 assimilation (Flexas & Medrano 2002, Lawlor 2002, Malhi et al. 1998,

Mulkey & Wright 1996), and (b) impairment of photosynthetic machinery (Chaves et al. 2003,

Malhi et al. 1998). Drought may also affect C assimilation through indirect effects: (a) xylem

cavitation during dry periods reduces hydraulic conductivity constraining stomatal conductance

and CO2 assimilation (Brodribb et al. 2002, Hubbard et al. 2001), (b) CO2 assimilation decreases

in response to phenological changes that reduce leaf area (Malhi et al. 1998), and (c) low soil

moisture decreases nutrient availability-either directly through reducing nutrient solubility,

and/or indirectly through creating less favorable conditions for the microbial activity that is

responsible for the decomposition of organic matter and release of nutrients in the soil (Cornejo

et al. 1994, Malhi et al. 1998)-reducing leaf nutrients and limiting CO2 assimilation.

Some studies in the Amazon region have reported a decrease in CO2 uptake at the leaf

level during the dry season for understory forest species (Aragdo et al. 2005, Fortini et al. 2003),

but similar data for overstory species are scarce. Induced drought in throughfall exclusion plots

reduced canopy leaf-level CO2 assimilation for some tree species in an old-growth forest in

Amazonia (Nepstad et al. 2002), consistent with a moisture limitation on leaf gas exchange.









Comparable studies in Amazonian forest regrowth are lacking, indicating the need for more

studies of moisture limitations on canopy leaf gas exchange for these forests.

Drought effects at the leaf level may reflect processes at the individual tree and ultimately

ecosystem (or stand-) levels in tropical forests (Figure 1-1). In old-growth forest sites in the

Brazilian Amazon, higher stem diameter growth rates (a component of aboveground net primary

productivity, ANPP) are associated with wetter periods (Higuchi et al. 2003, Rice et al. 2004,

Vieira et al. 2004). Comparable data for regrowth sites are scarce in part because most published

studies of forest regrowth in the tropics rely on one single inventory campaign in stands of

different ages to represent a successional chronosequence (e.g., Saldarriaga et al. 1988), resulting

in few available data for comparison between periods with different moisture availability for the

same stand. In Costa Rican secondary rain forests, the mortality of trees (diameter at breast

height > 10 cm) increased significantly with lower dry-season rainfall, but not with total annual

rainfall (Chazdon et al. 2005), suggesting that tropical forest regrowth may be extremely

sensitive to rainfall seasonality.

Carbon assimilation at the stand level can also be influenced by phenological changes

associated with water stress or a weakly deciduous strategy adopted by some tropical trees

(Malhi et al. 1998). In fact, higher litterfall rates during the dry period in tropical forests have

been reported in many studies (e.g., Dantas & Phillipson 1989, Scott et al. 1992, Wieder &

Wright 1995), but irrigation during the dry season in a tropical forest in Panama did not affect

the quantity or timing of litterfall (Wieder & Wright 1995). Higher litterfall rates associated with

the dry season may be triggered by an increase in vapor pressure deficits (Wright & Cornejo

1990), a decrease in cloud cover and soil nutrient availability (Eamus & Prior 2001), or may

reflect a genetic trait (Goulden et al. 2004). An important issue to consider is the production of









different leaf phenotypes associated with rainfall seasonality as reported by Kitajima et al.

(1997) in a Panamanian seasonal dry forest. Kitajima et al. (1997) found that leaves produced in

the late wet season (measured in the dry season) had higher photosynthetic rates than those

produced in the early wet and measured in the wet season.

Eddy covariance measurements of net CO2 exchange-the balance between gross primary

productivity and ecosystem respiration (Roy & Saugier 2001)-have shown different responses

of net CO2 assimilation during dry periods in the Brazilian Amazon. Malhi et al. (1998) reported

reduced net CO2 uptake during the dry season in central Amazonia. However, Saleska et al.

(2003) and Goulden et al. (2004) found higher rates of net C sequestration during the dry season

in an old-growth forest in east-central Amazonia, probably because drought reduced forest floor

decomposition, but not canopy photosynthesis. Thus, litter decomposition apparently plays an

important role in defining the direction of change in net ecosystem exchange due to drought

stress. Litterbag and mass balance studies have shown that low moisture availability reduces

litter decomposition in tropical forests (Comejo et al. 1994, Cornu et al. 1997, Luizdo &

Schubart 1987, Wieder & Wright 1995), probably as a result of lower leaching and/or

decomposer activity during dry periods.

Modeling studies have also predicted drought constraints on C dynamics of tropical

forests. Several ecosystem modeling studies (Asner et al. 2000, Foley et al. 2002, Phillips et al.

1998, Potter et al. 2001, Potter et al. 2004, Prentice & Lloyd 1998, Tian et al. 1998) have

analyzed the effects of recent El Nifio-Southern Oscillation (ENSO) events on the C balance of

Amazonian forests. All of these studies have indicated that the basin is a source of CO2

(negative net ecosystem productivity, NEP) during El Nifio events and a sink of CO2 (positive

NEP) during La Nifia events. Changes in C balance due to ENSO events in these models are









largely driven through changes in net primary productivity (NPP), and not through alterations in

heterotrophic respiration (Foley et al. 2002, Tian et al. 2000). During El Nifio years, lower

precipitation and higher temperatures result in increased simulated annual drought stress that

limits NPP.

In contrast, the dry period may represent an opportunity for C gain due to increased light

availability associated with reduced cloudiness. During typical, non-ENSO rainfall years, Huete

et al. (2006) found widespread greening in the dry season for central and eastern Amazonian old-

growth forests, suggesting that sunlight may represent a stronger control over forest phenology

and productivity than moisture availability. Consistent with a light limitation to forest phenology

and productivity, Graham et al. (2003) reported increased photosynthesis, vegetative growth, and

reproduction for branches of a tropical tree supplied with extra illumination during cloudy

periods in Panama. Further research on the controls of water and light over tropical forest

functioning is needed to better comprehend the response of these forests to climate change.

Belowground processes

Belowground processes are constrained by drought through the effects of low soil moisture

availability on root and microbial dynamics. Root growth declines under low soil water

potential, as shown for a temperate oak forest (Joslin et al. 2001), but root elongation may be

stimulated by dry conditions (Akmal & Hirasawa 2004) if plants allocate a larger fraction of

photosynthate to belowground biomass in response to drought. In tropical forests, fine root

production decreases and mortality may increase during the dry season as shown by

observational (Berish & Ewel 1988) and manipulative studies in old-growth sites (Cattinio et al.

2002, Cavelier et al. 1999, Yavitt & Wright 2001). Microbial activity is also constrained by low

soil moisture availability in tropical forest soils (Cleveland et al. 2002, Luizao et al. 1992).

Decreased root and/or microbial activities in the mineral soil and/or aboveground litter are likely









causes of reduced soil CO2 efflux during the dry season (Davidson et al. 2000, Vasconcelos et al.

2004).

Drought may also have an indirect effect on belowground processes if reduction in leaf C

assimilation under low soil moisture conditions results in decreased export of photosynthates to

roots. Such a reduced export may decrease the availability of C for root and rhizosphere

microorganism activity (Hogberg et al. 2001). However, the negative impact of low soil

moisture on belowground processes can be mitigated if hydraulically lifted water makes a

significant contribution to delaying soil dry-down in tropical forests. This phenomenon would

allow microorganisms to remain active for longer periods (Horton & Hart 1998), therefore

leading to an increase in nutrient mineralization. Da Rocha et al. (2004) suggested that the lack

of drought stress in an eastern Amazonian old-growth forest was probably related to deep rooting

and water redistribution by hydraulic lift. In the context of the Tapaj6s Throughfall Exclusion

Experiment, Romero-Saltos et al. (2005) did not find evidence for hydraulically lifted water by

understory/midcanopy tree species using deuterium-labeled soil water profiles, while Oliveira et

al. (2005) showed strong evidence for the occurrence of hydraulic redistribution based on the

dynamics of peaks of water recharge between shallow and deep soil layers, and sap flow data

measured in tap and lateral roots.

Nutrient Effects on Ecosystem Processes in Tropical Forests

In the tropics, many soils are highly weathered and consequently dystrophic (Sanchez

1976), which has led most past research to focus on nutrient cycling and assume that nutrient

availability limits tropical lowland evergreen forest productivity (Vitousek 1984). However,

evidence for such a constraint is rather limited. Malhi et al. (2004) reported that spatial

variability of coarse wood productivity of neotropical forests was apparently associated with soil

fertility. Manipulative experiments involving nutrient addition are necessary to show limitation









by a specific nutrient (Tanner et al. 1998), but such experiments are scarce in tropical lowland

forests. Most studies rely on soil nutrient inventories, aboveground biomass accumulation (Zarin

et al. 2001), leaf and litterfall nutrient concentrations (Vitousek 1984, Wood et al. 2005), root

growth responses (Cuevas & Medina 1988), and structural properties (Herrera et al. 1978) to

infer nutrient limitation to tropical lowland forest processes. Below I review some effects of

nutrient availability on above- and belowground processes.

Aboveground processes

Many essential mineral elements are directly or indirectly involved in plant tissue growth

(Marschner 1995), but a key aspect of the relation between nutrient availability and plant growth

and function is the positive correlation between maximum net photosynthesis and leaf nitrogen

(N) concentration (e.g., Lambers et al. 1998). This relationship is a consequence of the high

investment of leaf nitrogen in the enzyme responsible for carboxylation (ribulose bisphosphate

carboxylase, Rubisco) and in other photosynthetic enzymes (Chapin et al. 2002, Taiz & Zeiger

1998). However, the significance and form of the relationship between maximum net

photosynthesis and leaf nitrogen concentration may depend upon the importance of other

limiting nutrients including phosphorus (P) as reported for Amazonian tree species (Reich et al.

1994).

Phosphorus is often hypothesized to be the most limiting nutrient in old-growth and

regrowth lowland tropical forests. Analyzing within-stand nutrient use efficiency and nutrient

return in litterfall, Vitousek (1984) suggested that P, but not N availability, constrains fine

litterfall (an important component of ANPP) in lowland tropical forests, especially at Amazonian

sites. Davidson et al. (2004a), however, reported N co-limitation to tree growth at a forest

regrowth site subjected to several cycles of slash-and-burn in the Brazilian Amazon, and

associated the limitation by N with substantial losses of this element through burning. Also, in









an old-growth lowland evergreen forest in Indonesia, Mirmanto et al. (1999) reported increased

fine litterfall in plots fertilized with N, P, and N+P.

Aboveground biomass in old-growth forests in Central Amazonia has been negatively

correlated with soil sand content (Laurance et al. 1999), probably due to the low capacity of

nutrient retention by sandy soils. However, since sandy soils also have low moisture retention

capacity (Brady 1989), moisture limitation may also have contributed to the results obtained by

Laurance et al. (1999).

Belowground processes

Nutrient availability affects belowground processes by altering root and soil

microorganism activities. Root responses to low nutrient availability may not be straightforward.

The increased allocation of resources to belowground structures may be associated with low soil

fertility (Giardina et al. 2004, Gower 1987). However, higher proliferation of fine roots in

fertilized ingrowth cores in tropical forests suggests that root growth is limited by low soil

nutrient availability (Cuevas & Medina 1988, Mcgrath et al. 2001, Ostertag 1998, Raich et al.

1994), although fine root growth did not respond to likely reduced nutrient availability in litter

removal plots in an old-growth forest in Panama (Sayer et al. 2006). For a Panamanian lowland

tropical forest, Cavelier et al. (1999) suggested that control of fine root production may be more

complex, involving not only nutrient pulses, but also water pulses and aboveground biomass

growth.

Soil microbial activity is constrained under low soil nutrient conditions. Cleveland et al.

(2002, 2003) have shown increased microbial respiration with phosphorus addition to tropical

forest soil samples in a laboratory experiment. Cleveland and Townsend (2006) reported

increased in situ soil respiration with phosphorus and nitrogen fertilization in an old-growth









forest in Costa Rica; these authors suggested that phosphorus increased microbial respiration

while nitrogen probably affected soil respiration through effects on fine root dynamics.

Production of roots on top of the mineral soil has been considered as a structural

characteristic of forests growing on low nutrient soils (Herrera et al. 1978). However a recent

study has shown significant root growth in the forest floor of a relatively fertile old-growth site

in Panama, suggesting that proliferation of roots on top of the mineral soil is not necessarily

caused by low mineral soil nutrient levels, but may result from the availability of aboveground

litter (Sayer et al. 2006).

Conclusions

Research on moisture and nutrient constraints to tropical forest regrowth is rather limited

in quantity, and results are sometimes divergent from one study to the next. Manipulative

studies to investigate soil and plant processes in tropical forest regrowth are lacking, and are an

important tool for exploring the complex interactions that influence ecosystem response to

resource limitations. Furthermore, these studies are needed to better understand present

conditions and to project future impacts of climate and land-use changes on C dynamics in

tropical forest regrowth.












































Figure 1-1. Simplified conceptual diagram of likely effects of drought on leaf- and ecosystem-
level processes addressed in this dissertation. 1 and 4- symbols stand for increase and
decrease, respectively; ANPP aboveground net primary productivity. Numbers
refer to some study evidences of drought effects on processes: 1Eamus (2003);
2Sperry (2000); 3Cleveland et al. (2002); 4Lawlor and Comic (2002); 5Brodribb et al.
(2002), Hubbard et al. (2001); 6Firestone and Davidson (1989);7Chapin et al. (2002);
8Lawlor and Comic (2002); 9Rascher (2004); 10Nepstad et al. (2002); 11,12Hogberg et
al. (2001); 13Davidson et al. (2000), Vasconcelos et al. (2004).





27









CHAPTER 2
STUDY SITE AND EXPERIMENTAL DESIGN

Study Site

This study was conducted at a field station belonging to the Federal Rural University of

Amazonia (Universidade Federal Rural da Amaz6nia-UFRA), Brazil, near the city of

Castanhal (10 19' S, 470 57' W) in the state of Para. Since July 2001, daily rainfall was

measured 500 m away from the experimental area using a standard rain gauge. Prior to July

2001, rainfall data reported here are from the National Agency of Electrical Energy (Agencia

Nacional de Energia Eltrica-ANEEL) network meteorological station at Castanhal (1 17' 53"

S, 47 56' 56" W) located ~3 km from our site, but no longer in operation. From 70 to 90% of

annual rainfall occurs between January and July, resulting in a dry period from August to

December (Figure 2-1). Annual rainfall during the experimental period (Table 2-1) was

consistent with the mean standard error value registered from 1990 to 1999 by ANEEL (2461

+ 271 mm). The number of dry months (rainfall < 100 mm month-1) during the experimental

period varied from 2 to 5; several authors (e.g., Vieira et al. 2004) consider dry season months as

those with less than 100 mm rainfall for tropical sites.

The soils are classified as Dystrophic Yellow Latosol Stony Phase I (Ten6rio et al. 1999)

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

granulometric composition in the first 20 cm is 20% clay, 74% sand, and 6% silt. Concretions

represent 16% of the soil volume in the upper 10 cm of soil. In the surface soil (0 10 cm), pH

is 5.0, organic C is 2.2%, organic C stock is 2.9 kg m-2, total N is 0.15%, C:N is 14.4, and

Mehlich-1 extractable phosphorus is 1.58 mg kg-1 (Rangel-Vasconcelos 2002). This level of

extractable soil phosphorus suggests low availability at our study site compared to other soil

types and land uses in Amazonia (Mcgrath et al. 2001).









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 following

multiple cycles of shifting cultivation, beginning in the 1940's when the old-growth forest was

cleared. Each cycle of 1 to 2 years included cultivation of corn, manioc, and beans, followed by

fallow. Typical shifting cultivation cycles lasted seven to ten years (G. Silva e Souza & O.L.

Oliveira pers. comm.). Trees are mostly evergreen, with few species (e.g., Annonapaludosa and

Rollinia exsucca) showing deciduousness during the dry season. The four most abundant

overstory species are Lacistemapubescens Mart., Myrcia sylvatica (G. Mey.) DC, Vismia

guianensis (Aubl.) Choisy, and Cupania scrobiculata Rich., representing 71% of all stems in the

stand. In November 1999, mean + se stem density was 213 19.7 individuals per 100 m2, basal

area was 13 6 m2 ha-1, height was 4.9 0.4 m for the stand (Coelho et al. 2004), and

aboveground biomass was 51.1 + 2.5 Mg ha-1 for trees with diameter at breast height > 1 cm.

Experimental Design

Plots were established in August 1999, when the forest regrowth was 12 years old. Each

treatment plot is 20 m x 20 m with a centrally nested 10 m x 10 m measurement subplot. The

area between the measurement subplot and the plot-hereafter called "outer area"-was used for

some destructive samplings of soil, root, and aboveground litter. There were four replicate plots

for the irrigation treatment, four plots for the litter removal treatment, and four plots left

untreated as controls (Figure 2-2). Adjacent treatment plots were spaced 10 m from each other.

One tensiometer (Jet Fill Tensiometers, Soilmoisture Equipment Corp., Santa Barbara, CA,

USA) was installed at a depth of 10 cm in each plot and soil water potential was recorded on a

weekly basis in the morning. The number of actual replicates per treatment varied due to loss of









water column tension during the dry season. Soil suction variation in response to rainfall

seasonality and manipulation treatments is presented in Chapter 5.

Irrigation was applied at a rate of 5 mm day-, for about 30 minutes, during the dry seasons

of 2001 to 2005 (Table 2-2) in the late afternoon. The amount of daily irrigation applied

corresponds to regional estimates of daily evapotranspiration (Jipp et al. 1998, Lean et al. 1996,

Shuttleworth et al. 1984). Irrigation water was distributed through tapes with microholes every

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

periods we reduced the distance between tapes to 2 m to facilitate more even distribution of

water.

We used rainfall and soil suction 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 days 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 days 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 days. The lowest tension

value registered was -0.092 MPa, which may reflect the limited functional range of tensiometers

(Hanson 2000), although lower tensions may have occurred towards the end of dry season. The

installation of tensiometers deeper in the soil and of time domain reflectrometer sensors for

measurement of soil moisture content were hindered by the shallow depth to laterite in the soil

profile in Apeu.

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 = 8); carbon and nitrogen stocks of the pretreatment litter layer were
layer (538 35 g M n =8); carbon and nitrogen stocks of the pretreatment litter layer were









222.9 + 14.6 and 7.3 0.5 g m-2, respectively (n = 8). Raking maintained very low, but not

entirely absent litter standing crop. Total new non-woody litterfall removed during the treatment

period (from August 2001 to December 2005) was 3568 136 g m-2 (n = 12). Carbon and

nitrogen concentrations of pre-treatment litterfall were 47.9 0.2 and 1.2 0.02%, respectively,

corresponding to a C:N ratio of 40 0.7 (n = 12).

Measurements of gravimetric soil moisture content in the first 10 cm of soil for one date

during the 2001 dry season indicated that irrigated plots had about twice as much moisture as

control plots (22 2% vs. 10 2%); in the litter removal plots soil moisture was 11 2%. For

one date during the 2001 wet season, gravitational soil moisture content was 27 2% for control

and irrigated plots, and 31 2% for litter removal plots (Rangel-Vasconcelos 2002). The

difference in soil moisture status between control and irrigated plots was reflected in dry-season

differences in pre-dawn leaf water potential for an understory species (Miconia ciliata); in

November 2001 pre-dawn leaf water potential for control plants was about -1.2 MPa while

irrigated plants were about 1 MPa less negative (Fortini et al. 2003).









Table 2-1. Characteristics of rainfall distribution and intensity during the experimental period in
the site.
Year
1999 2000 2001 2002 2003 2004 2005
Annual rainfall (mm) 2577 2399 3179 2301 2895 3038 2793
Minimum monthly rainfall (mm) NAb 66 34 56 42 8 13
Maximum monthly rainfall (mm) NA 291 489 385 499 611 476
Number of dry season months a NA 3 5 4 2c 3 3c
Total dry season rainfall (mm) d NA 694 304 400 647 445 615
a Rainfall < 100 mm month-'.
b NA Not available.
c Not consecutive months.
d Dry season period = August to December.

Table 2-2. Dry-season irrigation intervals and associated rainfall intensity and distribution.


Dry-season


Total rainfall


Maximum daily


Number of
days without


irrigation Interval (mm) rainfall (mm) rainfall
1st 10 Aug 2001 to 453 54 101 (63%)a
16 Jan 2002
2nd 16 Aug 2002 to 516 66 93 (59%)
20 Jan 2003
3rd 7 Aug 2003 to 559 74 90(66%)
20 Dec 2003
4th 23 Sep 2004 to 547 130 81(64%)
26 Jan 2005
5th 29 Jul 2005 to 422 66 97(71%)
12 Dec 2005
a Percentage of days without rainfall during dry-season irrigation period.
















140 -

120

100

80 -

60

40







< < < < < < < < < < < < < Q

Date


Figure 2-1. Daily rainfall during the experimental period (data prior to July 2001 are from a
meteorological station about 3 km away from the study site).


Legend


IRR irrigation
CTL control
LIT litter removal






10m
CTL I



20 m 10 m
30


4


[RR IERR CTL

E 3 6


S T A T I 0 N


Figure 2-2. Experimental plot layout showing the arrangement of treatments.


R 0 A D


I N


PT---71 F----l









CHAPTER 3
SEASONAL AND EXPERIMENTAL EFFECTS ON LITTERFALL QUANTITY AND
QUALITY IN EASTERN AMAZONIAN FOREST REGROWTH

Introduction

Litterfall represents the major process of nutrient transfer from aboveground forest

vegetation to soils (Vitousek & Sanford 1986), and fine litterfall comprises a significant fraction

of aboveground net primary productivity in forests (Clark et al. 200 ib). Litter nitrogen and

phosphorus cycling are of particular importance since these nutrients usually are the most

limiting for tropical forest productivity (Vitousek 1984). Low phosphorus availability is likely a

common constraint for tropical forest regrowth, and nitrogen limitation appears significant for

forests reestablishing after several episodes of slash-and-burn, which lead to substantial losses of

nitrogen through volatilization (Davidson et al. 2004a, Gehring et al. 1999).

Litterfall quantity usually shows distinct patterns associated with rainfall seasonality, i.e.,

litterfall peaks during dry season (e.g., Wieder & Wright 1995). However, a direct effect of soil

moisture availability on litterfall quantity and timing has not been demonstrated (Cavelier et al.

1999, Wieder & Wright 1995). The concentration of nutrients in leaf litterfall may also vary

with rainfall seasonality in tropical forests (Wood et al. 2005), but a 5-year irrigation experiment

in Panama did not affect litterfall nutrient concentration (Yavitt et al. 2004). Litterfall

production has been shown to be limited by nutrient availability (Vitousek 1984), with

fertilization resulting in higher litterfall rates in a dry tropical forest in Mexico (Campo &

Vazquez-Yanes 2004) and a wet tropical forest in Puerto Rico (Li et al. 2006). Fertilization also

results in increased leaf litter nutrient concentration in tropical forests (Li et al. 2006).

A better understanding of fluxes and pools of carbon, nitrogen, and phosphorus involved in

litterfall can help to improve models of forest biogeochemistry. More appropriate quantification

of the role of soil moisture and nutrients in the regulation of litterfall can facilitate predictions of









carbon and nutrient dynamics under different conditions of resource availability. In this context,

long-term (> 1 year) data on litterfall quantity and quality are equally important to understand

interannual variability effects on carbon and nutrient dynamics, but such information is scarce

for tropical forests.

The primary objective of this chapter was to investigate the effects of moisture and nutrient

availability on litterfall within the context of the dry-season irrigation and litter removal

experiments described in Chapter 2. We hypothesized that (a) dry-season irrigation would

increase non-woody litterfall quantity and quality, and (b) litter removal would reduce non-

woody litterfall quantity and quality.

Material and methods

Litterfall

From October 1999 to December 2005, litterfall was collected weekly in each of three 1 m

x 1 m screen litter traps in the 10 m x 10 m measurement subplots. The weekly frequency of

litterfall collection was chosen to minimize mass and nutrient losses due to leaching of trapped

litter (Luizao 1989). The plant material collected in each trap was air-dried in the laboratory to

remove excess moisture before storage. At 4-week intervals, material from the same collector

was composite and then separated into woody and non-woody fractions. Leaves and their

petioles, foliar rachises, and reproductive parts were included in non-woody litterfall. Our non-

woody fraction corresponds to the "fine litter" (or "small litter") fraction defined in several

studies (e.g., Smith et al. 1998), except for the non-inclusion of woody material. In "fine litter",

small-diameter woody material-usually <1-2 cm diameter (Clark et al. 2001a, Proctor 1983)-

is included assuming that this woody fraction (1) has turnover times comparable to other

components of non-woody material (mostly foliar and reproductive material) and (2) may









represent material produced from the current year's growth. Thus, our estimate of non-woody

material may represent a slight underestimation of "fine litter".

We weighed woody and non-woody litterfall after drying at 60-70 C until constant

weight. Litterfall data for April 2003 was lost due to a malfunction of the oven that resulted in

burning of litterfall samples; for this period, we used for each trap a value of litterfall estimated

from the mean relative contribution of April to annual litterfall per trap as follows:

MCxAL
Est = ,
100 -MC'

where:

Est = estimated litterfall for April 2003 (in g m-2);

MC = mean relative contribution of April to annual litterfall in 2000, 2001, 2002, and 2004 (in

%), i.e.,


n (April litterfall annual litterfall)
MC = 1 x 100,
n

where i = year; and

AL = total 2003 litterfall except April (in g m-2).

Mean se MC was 6.3 0.2% for all traps over four years.

Composite samples of non-woody litterfall were ground with a coffee grinder (Krups, US)

and stored in 60 mL scintillation vials for subsequent analysis of carbon (C), nitrogen (N), and

phosphorus (P). Carbon was determined with an elemental carbon analyzer (Carlo Erba model

CNS2500) at the School of Forest Resources and Conservation (University of Florida) in

samples collected from October 1999 to March 2001. We estimated that non-woody litterfall

was 48% C based on the monthly non-woody litterfall C concentration (47.9 0.2%, n = 18).









Nitrogen and phosphorus concentrations were determined in the Laboratory of Plant

Ecophysiology and Propagation at Embrapa Amaz6nia Oriental (Brazil) in samples collected

from January 2000 to December 2004. The Kjeldahl digestion was used to determine total

nitrogen (Anderson & Ingram 1996). Phosphorus concentrations were determined

colorimetrically after digestion of 0.1 g sample in sulfuric acid and peroxide (Murphy & Riley

1962). Following the criteria in Boone et al. (1999), all the samples were analyzed in duplicate

for P, while 10% of the samples were randomly selected for duplicate analyses for N. Mean

coefficient of variation in duplicate analyses was 2.1% for N (n = 542) and 4.1% for P (n =

2096). Percent error in relation to standard reference material (peach leaves, NIST SRM 1547)

was -14 1.6% for N (n = 22) and 2.0 1.0% for P (n = 24).

To calculate N and P fluxes in non-woody litterfall (nutrient return), nutrient

concentrations were multiplied by mass for each trap per month.

Litter Stock

At the end of the 2004 wet season (25-August) and dry season (29-December), we

collected samples (n = 4) of forest floor litter from randomly chosen areas (25 cm x 50 cm) in

each of the control and irrigated plots and processed as for litterfall. Non-woody litter stock was

calculated by dividing the amount of dry material per collection area (g m-2).

Statistical Analysis

We used SAS version 9.00 to run the statistical analyses. We analyzed with PROC

MIXED the effects of treatment, date, and treatment-by-date interaction on the variables non-

woody litterfall mass (monthly and annual), and nitrogen and phosphorus concentration and

return using a repeated measures analysis with compound symmetric covariance structure. This

structure assumes constant variance at all dates and equal correlations between all pairs of

measures on the same experimental unit, i.e., litterfall trap for the litterfall variables and plot for









litter stock. We ran separate tests to compare each of the treatments with the control. Within

this analysis, significant treatment effects would have indicated temporally consistent differences

between treatment and control measurements both pre- and post-treatment and across seasons,

significant date effects were generally indicative of seasonal trends that affected both treatment

and control measurements, and treatment-by-date effects indicated a significant difference

between treatment and control measurements that occurred after the treatment was initiated.

Thus, the treatment-by-date effect represents the best test of treatment effect when there were no

pre-existing differences among plots prior to the treatment. We used a priori CONTRAST

statements to explicitly test whether the measured variables differed between seasons and

between treatments within each season (wet and dry).

When necessary, we performed log and square root transformations to meet the model

assumptions of normality, based on the criteria ofP > 0.05 in the Kolmogorov-Smirnov test, and

equal variances, based on the absence of a pattern of heteroscedasticity in the plots of residual

versus predicted values. Means and standard errors were calculated on the basis of

untransformed data. All results are reported as significant when P < 0.05; we report marginal

significance when 0.05 < P < 0.10. Multiple comparisons of means were performed with

Tukey's test (P < 0.05).

Results

Non-woody Litterfall

Irrigation experiment

Non-woody litterfall mass was significantly affected by date and the interaction between

treatment and date (Table 3-1, Figure 3-1B). The significant effect of the interaction was not

associated with consistent differences between treatments during the pre-treatment period (P =

0.76) or within dry-season irrigation periods (P = 0.18). Non-woody litterfall was significantly









higher in the dry season than in the wet season (79.5 + 1.3 and 58.5 0.9 g m-2 month-1,

respectively; P < 0.0001). Annual non-woody litterfall mass was significantly affected by the

interaction between treatment and date (Table 3-1); in 2003, annual non-woody litterfall mass

(Figure 3-2A) in irrigated plots was significantly higher than in control plots (899.2 55.3 and

742.4 63.1 g m-2 year-1, respectively; P < 0.01). Annual litterfall in the control plots was not

correlated with annual rainfall (r = 0.129, P = 0.808, Pearson correlation).

Non-woody litterfall N concentration was significantly affected by date only (Table 3-1,

Figure 3-3B). The effect of date was not related to a significant seasonal influence on litterfall N

concentration (dry = 1.24 0.01 vs. wet = 1.27 0.01% N, P = 0.86).

The input of N in non-woody litterfall was significantly affected by date and treatment x

date interaction; there was no significant effect of treatment (Table 3-1, Figure 3-4B). The

significant effect of the interaction was not related to a consistent difference between treatments

within dry-season irrigation periods (P = 0.19).

Non-woody litterfall P concentration was significantly affected by date and treatment x

date interaction (Table 3-1, Figure 3-3C). Litterfall P concentration was significantly higher in

control plots than in irrigated plots for some months during early- to mid-dry season (November

2001 and September 2002) and late-dry to early-wet seasons (January and February in 2002 and

2003). Litterfall P concentration was significantly lower in the dry season than in the wet season

(0.38 <0.01 and 0.40 <0.01 mg P g-1, respectively; P < 0.0001), although the difference was

slight.

Phosphorus return in non-woody litterfall was significantly affected by date and treatment

x date interaction (Table 3-1, Figure 3-4C). Treatment differences within dry-season irrigation









periods were marginally significant (P = 0.08), largely due to differences during the 2003 dry-

season irrigation period.

Annual return of N and P were significantly affected by date and treatment x date

interaction (Table 3-1); irrigation plots showed significantly higher N and P return than control

plots in 2003 (Figure 3-5).

Litter removal experiment

Non-woody litterfall mass was significantly affected only by date (Table 3-1, Figure 3-6B)

and was significantly higher in the dry season than in the wet season (76.2 + 1.2 and 58.4 0.9 g

m-2 month-1, respectively; P < 0.0001). Annual non-woody litterfall mass was significantly

affected by date only (Table 3-1), with the 2001 mean litterfall rates significantly higher than

subsequent years, but not different from 2000 (Figure 3-2B).

Non-woody litterfall N concentration was significantly affected by treatment, date, and

treatment x date interaction (Table 3-1, Figure 3-7B). During the treatment period, mean

litterfall N concentration was about 12% higher for control plots than for litter removal plots

(1.26 0.01 and 1.13 0.01% N, respectively; P = 0.01). This difference was not homogenous

throughout the manipulation period; with the progression of litter removal, the difference

between treatments in annual N concentration increased from ~ 11% in 2002 to ~ 16% in 2004,

which correspond to values of 5% (2002) and ~ 11% (2004) after accounting for pretreatment

differences. There was also a significant effect of treatment during the pretreatment period (P =

0.03); however, pretreatment differences between plots did not affect the significance (P = 0.04)

of post-treatment differences (contrast test).

The return of N in non-woody litterfall was significantly affected by date and treatment x

date interaction (Table 3-1, Figure 3-8B). However, the contrast test showed that the significant









effect of the interaction did not reflect consistent differences between treatments during the litter

removal period (P = 0.36).

Non-woody litterfall P concentration was significantly affected by date only, with a

marginally significant effect of treatment (Table 3-1, Figure 3-7C). Phosphorus concentration

during the wet season was slightly but significantly higher than during the dry season (0.40 +

<0.01 and 0.36 <0.01 mg P g-1, respectively; P < 0.0001).

The return of P in litterfall was significantly affected by date and treatment x date

interaction (Table 3-1, Figure 3-8C). The significant effect of the interaction term was

associated with occasionally higher values for control plots. Phosphorus return in the dry season

was slightly, but significantly higher than in the wet season (0.027 <0.001 and 0.022 <0.001

g P m-2, respectively; P < 0.0001).

Annual return of N was significantly affected by date and treatment x date interaction

(Table 3-1). However, there was no detectable significant or marginally significant difference

between treatment means in each year (P > 0.10, Tukey test), although 2001 values were

generally higher than other years, and control plots tended to have higher N return than litter

removal plots in 2002 (Figure 3-9A). Annual return of P was significantly affected by date only

(Table 3-1), with substantially higher return rates in 2003 and 2004 than in the other years

(Figure 3-9B).

Litter stock

The stock of non-woody litter (Figure 3-10) was significantly higher towards the end of the

dry season (December 2004) than at the end of the wet season (August 2004) (680 54 and 435

-2
+ 36 g m-2, n = 8, respectively; P < 0.001). There were no significant effects of treatment (P =

0.203) or treatment x date interaction (P = 0.271).









Discussion


Seasonal Patterns

Non-woody litterfall rates measured in this study are within the range reported for both

regrowth and old-growth Amazonian and other tropical forests elsewhere (Table 3-2). The

higher rates of litterfall during the dry season compared to the wet season are also consistent with

other studies in tropical forests (Dantas & Phillipson 1989, Sanchez & Alvarez-Sanchez 1995,

Scott et al. 1992, Smith et al. 1998, Wieder & Wright 1995). The magnitude of interannual

variability over 6 years varied from 9% for irrigated plots to 16% for litter removal plots, lower

than that reported for a Panamanian old-growth forest (38%) (Wieder & Wright 1995). Annual

litterfall was not related to annual rainfall, suggesting that litterfall production is not controlled

by rainfall intensity for this regrowth forest stand. However, Lawrence (2005) found a positive

relationship between annual litterfall and annual rainfall for tropical seasonal forests at a global

scale.

There were no detectable effects of rainfall seasonality on litterfall N concentration,

although Yavitt et al. (2004) reported higher N concentration in leaf fall during the wet season

for a Panamanian old-growth forest, and Wood et al. (2005) reported a wet season decline in leaf

litterfall N concentration for a Costa Rican old-growth forest.

Non-woody litterfall P concentration was lower during the dry season than in the wet

season in the present study, with some lower values of litterfall P associated with peaks in

litterfall, and some higher values of P occurring during lower litterfall rates in the wet season.

These results for litterfall P are consistent with data reported for a secondary dry tropical forest

in Mexico (Read & Lawrence 2003) and an old-growth forest in Costa Rica (Wood et al. 2005).

Most annual litterfall P peaks occurred during the first 1-2 months of the wet season, when

rapid decomposition of litter accumulated during the dry season could have supplied a pulse of









nutrients to plants with the onset of rainfall (Lodge et al. 1994, Wood et al. 2005). Lower

litterfall P concentration in irrigated plots during dry-wet season transitions (2001-02 and 2002-

03), associated with the strongest dry-season irrigation periods, are consistent with the pulse

hypothesis, i.e., irrigation could have prevented litter accumulation and, therefore, nutrient

mineralization pulse with the onset of rainfall. However, increased soil P availability for both

control and irrigation plots during wet-up events in the 2004 dry season is not consistent with

irrigation effects on nutrient pulse (Veluci et al. In preparation). The lack of litter removal

effects on litterfall P peak further suggests that the pulse hypothesis may not be applicable.

Alternatively, the seasonal and treatment effects on litterfall P may be caused by differences in P

resorption between treatments, and/or differences in the contribution of P-rich, reproductive

litterfall (flowers and fruits) during dry-wet transitions. Reproductive litterfall has been shown

to have higher P concentration than leaf litterfall for tropical forests (Scott et al. 1992, Zagt

1997), and to peak (number of seeds m-2) during dry-wet season transitions for our experimental

site, although no irrigation effects have been observed in two consecutive evaluation years (Dias

2006).

Litter stock measured in this study is within the range reported for tropical forests (Table

3-2). Increased litter stock in the dry season is consistent with higher litterfall and lower

decomposition rates during this period at the study site (Chapter 4), as also reported for an old-

growth forest in Panama (Wieder & Wright 1995).

Limited Impact of Dry-season Irrigation

Irrigation did not impact litterfall rates in the dry season, except for higher rates in irrigated

plots for a few dates, mostly in the 2003 dry-season. These results are consistent with those

found for a dry-season irrigation experiment in a semideciduous lowland forest in Panama

(Cavelier et al. 1999, Wieder & Wright 1995), and further confirm that soil moisture availability









may not trigger increased litterfall during the dry season in tropical forests. Higher dry-season

litterfall rates may be linked to increased vapor pressure deficits (Wright & Comejo 1990),

decreased cloud cover, decreased soil nutrient availability (Eamus & Prior 2001), or variation in

temperature (Breitsprecher & Bethel 1990). Although the exact trigger(s) of increased dry-

season litterfall have not been already ascertained, it is very likely that tropical trees respond to

more than one cue (Wright & Cornejo 1990).

Dry-season irrigation did not alter N and had only small effects on P concentrations in non-

woody litterfall, consistent with the results from a water manipulation study in a Panamanian

old-growth forest (Yavitt et al. 2004). The small impacts of dry-season irrigation in this study

contrasts with the potential for increased N and P availability in irrigated plots due to the

combination of (1) higher N and P inputs in litterfall during the dry season and (2) higher litter

decomposition in irrigated plots (Chapter 4). Thus, these results suggest that low litter quality-

indicated by the high C:N and lignin:N ratios of leaf litter (Chapter 4) and non-woody litterfall-

may be a stronger control over N (as well as P) availability than soil water status at this site,

favoring microbial immobilization of nutrients; Aerts (1997) suggest that litter chemistry

(especially the lignin:N ratio) represents the most important determinant of decomposition rates

in tropical regions. Furthermore, consistent with results from an irrigation study in Panama

(Yavitt & Wright 1996), dry-season irrigation had no influence on soil net nitrification rates at

our site (Vasconcelos et al. 2004).

We expected that long-term irrigation would have resulted in increased aboveground

productivity and, consequently, higher non-woody litterfall rates-an index of ANPP (Clark et

al. 2001a, Jordan 1983). However, after 5 years of dry-season irrigation, this effect has not

occurred consistently. Higher litterfall rates did occur for the irrigated plots in 2003, but that was









in the year with the weakest dry season over the whole experimental period, when we would

have expected the least effect of dry-season irrigation on forest processes. However, increased

annual litterfall in irrigated plots in 2003 may have resulted from a lag effect of the extended

drought in the preceding dry season (Table 2-1), consistent with results for a temperate mixed

deciduous forest (Newman et al. 2006). Nonetheless, the effects of the extended 2002 dry

season were not sufficiently intense to affect tree mortality at the community level which

actually decreased from 2002 to 2003, and remained constant in 2004 (Araujo et al. 2005).

Litter Removal Reduces Litterfall N Concentration

Nitrogen and phosphorus concentrations and inputs in litterfall are comparable to values

reported for forests of the Brazilian Amazon and elsewhere in the tropics (Table 3-2). Increased

differences in N concentrations between control and litter removal plots are consistent with the

recognized role of nutrient cycling in litter as a significant source of N for tropical forest plants

(Markewitz et al. 2004, Vitousek & Sanford 1986). Mean litterfall P concentration for the

control plots in this forest regrowth stand (0.04%) coincides with the value proposed by Vitousek

(1984) to distinguish between high and low P levels for tropical forests. For most months from

2000 to 2003, litterfall P concentrations were below this threshold, which may reflect the low

availability of soil phosphorus, as suggested by the low soil extractable P reported for the site

(Rangel-Vasconcelos 2002).

The lack of treatment effects on litterfall P concentration may be explained by sufficient

supply of P from soil sources. While weathering processes are not likely a substantial source of

P in highly weathered tropical soils deprived of primary P-containing minerals (Sanchez 1976),

mineralization of P from soil organic matter may represent a significant source of this nutrient

for plants, even after 40 months of bi-weekly litter removal. Recent studies have determined

substantial amounts of labile organic-P fractions (NaOH- and NaHCO3-extractable) for









Amazonian forest regrowth sites in Brazil (Frizano et al. 2003, Markewitz et al. 2004), and a

simulation study concluded that N and P stored in (deeply buried) soil organic matter can sustain

C accumulation rates under conditions of limited input of such nutrients in tropical forest

regrowth (Herbert et al. 2003). In addition, some regrowth forest trees colonizing sites with low

soil P availability probably present mechanisms to improve P acquisition such as mycorrhizal

associations and high phosphatase exudation rates (Marschner 1995). Uhl (1987) hypothesized

that high incidence of mycorrhizal infection and efficient uptake and nutrient use may be

necessary for establishment of successional trees under the limiting nutrient conditions typical of

abandoned lands after slash-and-burning in the tropics. Similarly, Gehring et al. (1999)

suggested that the growth of two early successional tree species in an Amazonian forest site was

not limited by soil P availability because of efficient mycorrhizal associations.

Since litter is the main source of most nutrients in tropical forests (Markewitz et al. 2004,

Vitousek & Sanford 1986), we expected that chronic litter removal would have resulted in

nutrient deficiency, and consequently reduced ANPP (Harrington et al. 2001). Thus, we

hypothesized that non-woody litterfall rates would diminish for litter removal plots. This study

thus far indicates that the quantity of non-woody litterfall was insensitive to the reduction in

nutrient availability (indicated by reduced litter N concentration) imposed by the litter

manipulation treatment, consistent with the results obtained by Sayer (2005) for a 2-yr litter

removal study in Panama. It is possible that extending the litter removal period will further

reduce nutrient concentrations in litter, leading to a critical point where productivity will be

significantly constrained. Nutrient manipulation effects on ecosystem processes are usually not

immediate, and litter removal studies may have slower effects on litterfall responses than

fertilization studies (Campo & Vazquez-Yanes 2004, Mirmanto et al. 1999).









Table 3-1. F statistics and associated significance levels for the effects of treatments (irrigation and litter removal), sampling date,
and their interaction on non-woody litterfall mass and nutrients in a tropical regrowth forest stand in eastern Amazonia,
Brazil.
Irrigation Experiment Litter Removal Experiment
Non-woody litterfall Treatment Date Treatment x Date Treatment Date Treatment x Date
Mass 0.72 ns 36.76*** 1.66*** 0.22ns 25.21*** 1.34*
N concentration 0.18 ns 27.87*** 1.OOns 8.42** 29.22*** 2.53***
N return 0.49 ns 31.94*** 1.54*** 0.65ns 26.71*** 1.40*
P concentration 0.95ns 52.11*** 2.56*** 3.74ns 53.08*** 1.23ns
P return 0.30ns 27.37*** 1.91*** 0.39ns 20.20*** 1.36*
Annual mass 0.7Ins 1.78ns 3.46** 0.24ns 4.59*** 0.74ns
Annual N return 0.82ns 24.47*** 3.66** 0.48ns 24.89*** 2.67*
Annual P return 0.63ns 79.92*** 3.88** 0.15ns 31.99*** 1.45ns
aThe level of significance is indicated (*: P < 0.05, **: P < 0.01, ***: P < 0.001, ns: not significant).









Table 3-2. Estimates of annual non-woody litterfall (mass, nitrogen, and phosphorus), non-woody litter stock, and litterfall:forest
floor mass ratio (kL) in tropical forests.
Annual Annual nitrogen Annual phosphorus
Forest rainfall Annual litterfall Non-woody litterfall litterfall
description Location (mm) (g m-2 yr-1)b litter (g m-2) kL (g m-2 yr-1) (g m-2 yr-1) Sourc


Regi h i/t
3
12

10
19
3 18
30
15
0 O1,-glo Ill


Brazil
Puerto
Rico
Brazil
Brazil
Brazil
Brazil
Brazil


Brazil
Brazil
Venezuela
Brazil
Brazil
Panama
Brazil
Indonesia
Brazil
Brazil
Brazil
Brazil


2600
3810

2433
1800
1940
2830
2760


2100
2600
3565
2100
2300
2600
1900
3600
2000
1800
2000
2200


504
820

690
890
1040 -1300
630
783


640
804
1025
825
928
1240
970
710


564 840


900
570 920
890


1.64


e


10.4


0.28


1.25


0.30


8.63


12.1
15.1
11.8


11.5
5.84


2.01


1.34


0.21
0.31
0.67


0.16










Table 3-2. (continued)
Annual Non-woody Annual nitrogen Annual phosphorus
Forest rainfall Annual litterfall litter litterfall litterfall
description Location (mm) (g m-2 yr-1)b (g m-2) kL (g m-2 yr-1) (g m-2 yr-1) Sourcec
Brazil 1800 1030 14.3 0.33 4
Brazil 1940 1100 -1740 5
Brazil 2000 1380- 1460 19
a For the forest regrowth sites, age (years after abandonment) is presented. Regrowth includes sites classified as secondary forests, while old-
growth refers to primary and mature forest sites.
b Litterfall dry mass estimated as two times litterfall carbon for the studies without direct report of mass.
c For each source number, details of coordinates, soil type, and authors of each study are presented below:
1 1 44' S, 47 9' W, Capitao Pogo, Brazil, unspecified soil, Dantas and Phillipson (1989)
2 180 19' N, 650 49' W, Puerto Rico, Ultisol, Cuevas etal. (1991)
3 2 25' S, 590 50' W, Manaus, Brazil, Oxisol, Mesquita et al. (1998)
4 2 59' S, 470 31' W, Paragominas, Brazil, Haplustox, Markewitz et al. (2004)
5 Southwestern Amazonia, Brazil, Ultisols with patches of Oxisols, Salimon et al. (2004)
6 1 18' 6" S, 48 26' 35" W, Belem, Brazil, Yellow Latosol, Oliveira (2005)
7 1 19' S, 470 57' W, Ape6i, Brazil, Distrophic Yellow Latosol, this study
8 2 34' S, 60 7' W, Manaus, Brazil, Yellow Latosol, Klinge and Rodrigues (1968)
9 10 54' N, 670 3' W, San Carlos de Rio Negro, Oxisol, Cuevas and Medina (1986)
10 2 34' S, 60 7' W, Manaus, Brazil, Yellow Latosol, Luizao (1989)
11 Maraca Island, Brazil, Scott et al. (1992)
12 9 o 09' N, 79 o 51' W, Barro Colorado Island, Panama, Alfisol, Wieder and Wright(1995)
13 Curua-Una Forest Reserve, Brazil, Oxisol, Smith et al. (1998)
14 0 o 6' S, 113 o 56' E, Kalimantan, Indonesia, "Yellow sandy soil", Mirmanto et al. (1999)
15 Tapaj6s National Forest, Brazil, Ultisols and oxisols, Silver et al. (2000)
16 2 o 59' S, 47 o 31' W, Paragominas, Brazil, Haplustox, Davidson et al. (2000)
17 2.897 S, 54.952 W, Tapaj6s National Forest, Haplustox, Nepstad et al. (2002)
18 2 35'21.08" S, 60 06' 53.63" W, Manaus, Brazil, Oxisol, Luizao etal. (2004)
19 2 64'S, 540 59'W, Tapaj6s National Forest, Brazil, Oxisols and Ultisols, Silver et al. (2005)











800 -
700 A
E 600 -
S500 -
400 -
> 300 -
S200 -
100

; 160 B control
E 140 O irrigation

120 -
S100 -

80 -
o 60 -

S40 -
o
z 20 -


0')0O)0000 004EEE E ENN NN N NMM" M mO 1- '1-1- 000 mom
0 0000 00000 00000 00000 00000 00000 0

Date

Figure 3-1. Effects of rainfall patterns and dry-season irrigation on non-woody litterfall in an
Amazonian forest regrowth stand, Brazil. A) Monthly rainfall. B) Monthly non-
woody litterfall for control and irrigation plots. In Figure 3-1B, each symbol
represents the mean + standard error, n = 12. Vertical gray bars indicate the irrigation
periods. White and black horizontal bars mark the dry and wet seasons, respectively.










1000


800 -



600


400 -



200

0 -

0



800 -



600


400 -



200



0


control
irrigation


I I I I I I


control
litter removal


2000 2001 2002 2003 2004 2005

Year


Figure 3-2. Effects of dry-season irrigation and litter removal on annual non-woody litterfall in
an Amazonian forest regrowth stand, Brazil. A) Non-woody litterfall for control and
irrigation plots. B) Non-woody litterfall for control and litter removal plots. Each
symbol represents the mean + standard error, n = 12. Treatments began in August
2001. Asterisk indicates significant treatment difference (P < 0.05).











800 -4-
A

E 600 -


400 -


200 -
0



1.6 -


S .2 -
1.0 -
>, 0.8 -
o
o 0.6 -
0.4 0 control
o 0 irrigation
Z 0.2 -


0.5 -

0.4 -




S 0.3 -

o 0.2 -
0

0.1 -
o
z
0.0


N N ~ CO O C) N N O 0 C) N N O 0 C) N N ~ CO OC) N (N CO C) (N
(D- O O O ( (D (D ( ( C) C) C) O O O C) C) C) C) OC) C) C) C) O ,-

Date


Figure 3-3. Effects of rainfall patterns and dry-season irrigation on non-woody litterfall nutrient
concentrations in an Amazonian forest regrowth stand, Brazil. A) Monthly rainfall.
B) Nitrogen (N) concentration for control and irrigation plots. C) Phosphorus (P)
concentration for control and irrigation plots. In Figures 3-3B-C, each symbol
represents the mean + standard error, n = 12. Vertical gray bars indicate the irrigation
periods. White and black horizontal bars mark the dry and wet seasons, respectively.












800
A
E 600 -


I 400 -


S 200 -
0

0
B
1.6 -
2 1.4 -
Z
1.2 -
C? 1.0 -

0.8 -
0 0.6 -
0
0.4 -
o 0.2 -

0.0

S 0.06 -
W S control
a. 0.05 O irrigation

Sc; 0.04 -

2? 0.03 -
0
o 0.02 -











Figure 3-4. o 12. Vernscal dry-bason irrigation periods. Whiean
C 0.01
0
Z

0M 04 0 0 O0 0 04 CN CN CN CN CN CN CO) CO" C0 COO CO C" O IT NT
0) IT 0 00 0- N 0 0 0 00 0- N 0 0 0 00 0- 000- 0 00 0- 00 0 00 00 N

Date


Figure 3-4. Effects of rainfall patterns and dry-season irrigation on non-woody litterfall nutrient
return in an Amazonian forest regrowth stand, Brazil. A) Monthly rainfall. B)
Nitrogen (N) return for control and irrigation plots. C) Phosphorus (P) return for
control and irrigation plots. In Figures 3-4B-C, each symbol represents the mean +
standard error, n = 12. Vertical gray bars indicate the irrigation periods. White and
black horizontal bars mark the dry and wet seasons, respectively.


















































* control
o irrigation


2000 2001 2002 2003 2004

Year


Figure 3-5. Effects of dry-season irrigation on annual non-woody litterfall nutrient return in an
Amazonian forest regrowth stand, Brazil. A) Nitrogen (N) return for control and
irrigation plots. B) Phosphorus (P) return for control and irrigation plots. Each
symbol represents the mean + standard error, n = 12. Treatments began in August
2001.


%CM
E


z




IL
0




0


C
C


0



w" 0.4-



I
e-- 0.3
'00 'E*
-


oE
0 0.2

0
O

0.1
C
C











800 -- --- -
700 A
E 600 -
4 500 -
400 -
> 300 -
200-
0
S100 -


S160 B control
E 140 O litter removal
120 -
S100 -

80 -
o 60 -

s 40 -
o
z 20 -


0M0O M )4 )4 04 04 04 0 CO CO CO CO CO C 'I"T" 'I" T LO LO CO C O C O O



Date


Figure 3-6. Effects of rainfall patterns and litter removal on non-woody litterfall in an
Amazonian forest regrowth stand, Brazil. A) Monthly rainfall. B) Monthly non-
woody litterfall for control and litter removal plots. In Figure 3-6B, each symbol
represents the mean + standard error, n = 12. The vertical line indicates the beginning
of the litter removal treatment. White and black horizontal bars mark the dry and wet
seasons, respectively.


















































0 O00000OO O '- N N N N N (N CM) C C C M M T T T
(N N T O 0 CN (N T O M C N (N O O M C N (N O M C0 N (N 0 0 0 CN

Date


Figure 3-7. Effects of rainfall patterns and litter removal on non-woody litterfall nutrient
concentrations in an Amazonian forest regrowth stand, Brazil. A) Monthly rainfall.
B) Nitrogen (N) concentration for control and litter removal plots. C) Phosphorus (P)
concentration for control and litter removal plots. In Figures 3-7B-C, each symbol
represents the mean + standard error, n = 12. The vertical line indicates the beginning
of the litter removal treatment. White and black horizontal bars mark the dry and wet
seasons, respectively.












800 -

A
S 600

C -
E 400

0
200


0-

1.6 B
1.4
Z
1.2
C 1.0

S 0.8
0 0.6
0
0.4
o 0.2
Z
0.0

C f control
00.06 litter removal

0. 0.05

o 0.04

0.03
0
o 0.02

0.01
0



-N IN 0 0O 0 -N N IT 0 0 CN N I 0C 0 C-N N -I 0 0 O 0 N N 0 0 C N


Date


Figure 3-8. Effects of rainfall patterns and litter removal on non-woody litterfall nutrient return
in an Amazonian forest regrowth stand, Brazil. A) Monthly rainfall. B) Nitrogen (N)
return for control and litter removal plots. C) Phosphorus (P) return for control and
litter removal plots. In Figures 3-8B-C, each symbol represents the mean + standard
error, n = 12. The vertical line indicates the beginning of the litter removal treatment.
White and black horizontal bars mark the dry and wet seasons, respectively.
















































* control
o litter removal


J .%J I I I I I
2000 2001 2002 2003 2004

Year


Figure 3-9. Effects of litter removal on annual non-woody litterfall nutrient return in an
Amazonian forest regrowth stand, Brazil. A) Nitrogen (N) return for control and litter
removal plots. (B) Phosphorus (P) return for control and litter removal plots. Each
symbol represents the mean + standard error, n = 12. Treatments began in August
2001.


E





Ci
z

t


0)
e-
'= >
>C
CE


0




0.3-


'00 'E*
0.2-

oE


o0
0.1
C
<










1000



800 -



600 -



400 -



200 -


- control
m1 irrigation


0 1-


Aug-04


Dec-04


Date

Figure 3-10. Non-woody litter stock for control and irrigation plots in an Amazonian forest
regrowth stand, Brazil. Each bar represents the mean + standard error, n = 4.









CHAPTER 4
LEAF DECOMPOSITION IN A DRY SEASON IRRIGATION EXPERIMENT IN EASTERN
AMAZONIAN FOREST REGROWTH

Introduction

Litter decomposition is a major nutrient cycling process in terrestrial ecosystems and is

particularly important for forest ecosystems on low fertility soils, including many tropical forests

(Golley 1983, Swift et al. 1979). The rate of litter decomposition is controlled by interactions of

litter quality, environmental conditions, and soil organisms (Swift et al. 1979). Litter quality is

defined as the amount and types of organic carbon compounds, nutrient concentrations, and

ratios between carbon compounds and nutrients in litter; low-quality litter (e.g., low nutrient,

high carbon compounds:nutrient ratio) usually shows lower decomposition rates than high-

quality litter (e.g., high nutrient, low carbon compounds:nutrient ratio) (Loranger et al. 2002,

Mesquita et al. 1998, Songwe et al. 1995). For moist tropical forests moisture and temperature

are assumed to be non-limiting, and litter quality is thought to be the dominant control on

decomposition rates (Aerts 1997), although seasonal drought (Cornejo et al. 1994) and excessive

moisture (Schuur 2001) may retard decomposition at some tropical forest sites. Water

manipulation studies may help to clarify seasonal drought effects on litter decomposition.

A considerable number of studies have investigated decomposition responses of leaf litter

from several plant species to rainfall seasonality in tropical forests (e.g., Cornu et al. 1997,

Cuevas & Medina 1988, Luizdo & Schubart 1987), but there are few such data for tropical

regrowth sites (Mesquita et al. 1998). The effects of water manipulation on decomposition rates

have been examined for old-growth tropical forests in Brazil (Nepstad et al. 2002) and Panama

(Cornejo et al. 1994, Wieder & Wright 1995), but related studies are lacking for regrowth stands.

Studies to quantify decomposition rates and their controls can help to improve understanding of

carbon and nutrient cycling in tropical forest regrowth sites.









The primary objective of this chapter was to investigate the effects of moisture availability

on litter decomposition within the context of the dry-season irrigation experiment described in

Chapter 2. We hypothesized that (a) decomposition rates would be faster under higher moisture

availability in the wet season and during dry-season irrigation periods in the treatment plots, and

(2) decomposition rates would be faster for species with higher quality leaves.

Study Site and Experimental Design

Study site and experimental design are described in Chapter 2.

Material and Methods

Leaf Litter Decomposition

The litterbag method (Harmon et al. 1999) was used to study leaf litter decomposition.

This method is the most frequently employed for examining litter decomposition in terrestrial

ecosystems (Wieder & Lang 1982), although it has several limitations that can significantly

influence decomposition rates including alteration of litter microclimate and exclusion of certain

decomposer organisms (but see Prescott 2005). However, the litterbag method is adequate for

studies comparing species, sites, and the effects of experimental manipulations on decomposition

(Heal et al. 1997, Wieder & Lang 1982).

Traps were put outside the treatment plots to collect leaves for the decomposition study.

Fresh fallen leaves of Lacistema pubescens Mart., Ocotea guianensis Aubl., Stryphnodendron

pulcherrimum (Willd) Hochr., and Annona paludosa Aubl. were collected every week for 3-4

months prior to installing each of 3 separate decomposition experiments (described below).

Collected leaves were dried under ambient conditions and stored. L. pubescens was chosen

because it is the most common species in the study area (Araujo et al. 2005) and A. paludosa, 0.

guianensis, and S. pulcherrimum were selected because they represent a wide range in leaf

texture; 0. guianensis leaves are thick and appears to be recalcitrant, while A. paludosa and S.









pulcherrimum possess leaves that appear to decompose more rapidly. After the collection

period, 6-8 subsamples of 10 g each were oven dried at 65-70 C until constant weight and the

dry weight conversion factor (air-dry mass:oven-dry mass) was calculated. These subsamples

were processed following procedures described for litterfall (Chapter 3) to determine the initial

leaf litter chemical composition (described below). For the second and third decomposition

experiments, S. pulcherrimum was replaced by Vismia guianensis (Aubl.) Choisy because we

noticed that some leaflets of the former species were smaller than the openings in the litterbag

screen, which could overestimate decomposition rates. V. guianensis is another common pioneer

species in the study area (Araujo et al. 2005).

Bags of polypropylene with openings of 1 mm x 0.8 mm and measuring 20 cm x 20 cm

received about 10 g of air-dried material of only one species. In each of the control and

irrigation plots, 18 litterbags of each species were randomly placed in the surface of the litter

layer. After 30, 60, 120, 180, 240, and 360 days (Experiments 1 and 2) and 13, 31, 45, 61, and

90 days (Experiment 3), three bags of each species were retrieved in each plot. After retrieval,

litterbags were air dried to facilitate the removal of adhering soil particles and roots gently using

forceps and small, soft brushes (Tigre, medium size, Brazil). Then, the material was oven dried

and weighed to calculate remaining leaf mass. Samples for the last collection in Experiment 2

were discarded because it was not possible to separate out soil particles from leaf material.

To investigate the effects of dry-season irrigation on the remaining mass of leaf litter under

different stages of decomposition, the experiments had different installation and duration

periods. Experiments 1 and 2 lasted 12 months and were installed in the beginning of the 2002

wet (February 7) and 2003 dry (July 27) seasons, respectively, in order to determine the effect of

seasonality on initial and later stages of decomposition. For an improved temporal resolution of









dry-season irrigation effects on decomposition, Experiment 3 was carried out exclusively during

three months in the 2004 dry season (September 24 to December 23), with more frequent

sampling during that period than for Experiments 1 and 2.

To assess seasonal effects on leaf litter decomposition irrespective of treatment, we

compared the remaining leaf mass of control plots at 60 days in Experiment 1 and at 61 days in

Experiment 2, which corresponded to wet and dry seasons, respectively. Total rainfall in these

wet and dry seasons was 1311 and 359 mm, respectively. For this analysis we used data for A.

paludosa, L. pubescens, and 0. guianensis.

Remaining leaf mass (percent) was calculated as


t x 100,
X0

where X, is the dry litter mass at the time t and X0 is the initial dry litter mass.

Initial Leaf Litter Chemistry

Phosphorus concentrations were determined colorimetrically after digestion of 0.1 g

sample in sulfuric acid and peroxide (Murphy & Riley 1962). Carbon and nitrogen were

determined with an automated dry combustion instrument (LECO Model CNS-2000). Lignin

and cellulose were determined by a sequential digestion of fibres (Anderson & Ingram 1996).

Specific Leaf Area

The specific leaf area (SLA) was measured in individuals ofAnnonapaludosa (n = 3),

Lacistema pubescences (n = 4), Ocotea guianensis (n = 4), and Vismia guianensis (n = 4) located

in the control plots. In each tree, three young, fully expanded leaves were chosen from different

branches, and three discs (1.11 cm2) were collected from each leaf. The discs were dried at 65

C for 48 h and individually weighed to the nearest 0.0001 g.









Statistical Analysis

We used SAS System version 9.00 to run the statistical analyses. Decomposition rates (k)

were calculated by fitting the observed data (i.e., remaining leaf mass) to the single exponential

model proposed by Olson (1963) using the PROC NLIN procedure. In the single exponential

model, Xt = Xoe-k; where X, and X0 are the litter mass at the times t and 0 (initial), respectively,

and k is the decomposition rate (yr-1). Although this model makes unrealistic assumptions (e.g.,

treats litter as a uniform, homogeneous substance) regarding the decomposition of litter, k values

calculated with this model are useful for interpreting short-term (first year decomposition),

comparative experiments (Paustian et al. 1997) such those in this study. For these analyses, data

(Xt-Xo) from three litterbags were averaged per plot for each sampling date because we

considered individual plots as the experimental units, resulting in n = 4 for each combination of

species, treatment, and sampling date. The effects of species, treatments, and the species x

treatment interaction on k values were analyzed with a two-way ANOVA using PROC ANOVA.

PROC CORR was used to analyze the correlation between k and initial litter quality parameters.

The effects of species on initial litter chemistry were analyzed with one-way ANOVA

using PROC ANOVA. The TTEST procedure was used to compare seasonal effects on

remaining leaf mass for control plots. The statistical analyses were carried out using the mean

SLA calculated for each leaf per species. Means and standard errors were calculated on the basis

of untransformed data. All results are reported as significant when P < 0.05. Multiple

comparisons of means were performed with Tukey's test.

Results

Dry-season irrigation had no significant effects (Table 4-1) on leaf litter decomposition

rates (k) obtained by fitting curves to all of the collection data in the twelve-month experiments

(Table 4-2, Figure 4-1). However, in Experiment 2, k values obtained from curve fitting to the









dry season data only were significantly higher in irrigated than in control plots (1.04 0.06 and

0.86 0.06 yr1, respectively; P < 0.01) (Table 4-1). In the three-month dry-season experiment,

A. paludosa showed significantly higher decomposition rates than the other species under

irrigation, which did not differ significantly among them (Table 4-3), and within species,

decomposition rates were significantly higher in irrigated plots than in control plots (Table 4-3).

All of the experiments showed significant effects of species on k (Table 4-1). Overall, A.

paludosa showed the highest decomposition rates (Table 4-2, 4-3). There were significant

differences (P < 0.0001) in specific leaf area, carbon, nitrogen, phosphorus, lignin, and cellulose

concentrations, lignin:N ratio, and C:N ratio among species (Table 4-4), but there were no

significant correlations between k and leaf quality parameters (Table 4-5, Figure 4-2).

The analysis of seasonal effects on decomposition showed that remaining leaf mass was

significantly (P < 0.001) higher in the dry season than in the wet season forL. pubescens (87.6 +

0.9 and 76.0 0.9%, respectively) and 0. guianensis (88.5 1.1 and 77.5 0.9%, respectively);

there was a marginally significant effect (P < 0.052) of season on A. paludosa remaining leaf

mass (dry = 80.4 3.1% vs. wet = and 71.8 1.7%).

Discussion

Decomposition rates measured in this study are within the range reported for tropical

forests (Table 4-6). Decomposition was faster during the wet season than the dry season, as

observed in other studies in tropical forests in Amazonia (Cornu et al. 1997, Luizdo & Schubart

1987) and elsewhere (Cornejo et al. 1994, Wieder & Wright 1995). Moisture constraints on

decomposition were further confirmed by higher mass loss rates in dry-season irrigated plots,

except when irrigation was applied to litter previously exposed to field conditions for 180 days;

in this case, the greater proportion of recalcitrant compounds in advanced stages of litter decay









(Swift et al. 1979, Wieder & Lang 1982) probably conferred less susceptibility to decomposition

in response to increased moisture availability.

Although leaf decomposition is significantly constrained during the dry season, the

greatest difference between mass loss in control and irrigated plots was 10 to 13% only, and

between dry and wet seasons was 7 to 12% only. Such small differences could be due to

exclusion of macrofauna activity in leaf decomposition in the 1 mm x 0.8 mm opening bags.

Using 1-mm mesh litterbags with additional openings of about 10 mm, Luizdo and Schubart

(1987) suggested that surface fine root penetration and macroarthropod activity determined the

great difference in leaf mass loss between the dry and wet seasons for 3-yr-old forest regrowth in

central Amazonia. It is not likely that fine root colonization has been constrained in our

litterbags as we did observe fine root adhered to leaves. However, the 1 mm x 0.8 mm opening

bags likely restricted macroarthropods to access leaf material and this could have contributed to

the small differences between dry and wet as well as control and irrigation percent leaf mass

losses in this study.

Leaf chemical and structural traits in this study are also consistent with other studies in

tropical forests (Table 4-7). The range of lignin concentration (42.9 to 51.7%) found for the

species investigated in this regrowth forest is high in comparison to reported values for old-

growth forest tree species in Panama (Table 4-6), but similar to results of Mesquita et al. (1998)

and Vasconcelos and Laurance (2005), who found lignin concentrations of about 53% for

regrowth forest species in two central Amazonian sites.

The lack of correlation between decomposition rates and leaf quality parameters may result

from the reduced number of leaf litter species tested in this study, as also observed by Fonte and

Schowalter (2004) who investigated 8 different litter species for a Puerto Rican forest. For









tropical forests, the lignin:N ratio was found to be the strongest predictor of decomposition in a

Panamanian old-growth forest (Santiago 2003), while decomposition rates decreased with higher

tannin concentration for a regrowth forest site in Amazonia (Mesquita et al. 1998). However, the

strongest leaf quality predictor may change according to the stage of the decomposition process

(Loranger et al. 2002).

Despite the lack of correlation between decomposition rates and litter quality parameters,

the highest decomposition rates observed in Annonapaludosa are probably explained by their

higher leaf quality, i.e., high concentrations of nitrogen and phosphorus, the lowest concentration

of lignin, and thin leaves (high specific leaf area). The low decomposition rates of Ocotea

guianensis and Vismia guianensis leaves are associated with low N and P concentrations, high

lignin concentration, the highest C:N and lignin:N ratios (> 50), and thicker leaves (low specific

leaf area). Interestingly, Lacistemapubescens was often the "outlier" interfering with a strong

linear relationship between k and litter quality; decomposition rates of Lacistema are lower than

would be predicted by regressing the data from the other species, suggesting that decomposition

of Lacistema leaf litter may be strongly controlled by some litter quality parameter not

determined in this study. One potential explanation is the pubescent habit of its leaves.

Overall, moisture effects on k were comparatively higher than those related to litter quality;

while k was on average 2.4 times higher in irrigated than in control plots during the three-month

dry-season experiment, the greatest difference between species maximum/minimum k was 1.5

considering all of the experiments.









Table 4-1. F statistics and associated significance levels (in parentheses) for the effects of
treatments (control and irrigation), species, and their interactions on leaf litter
decomposition rates in a tropical regrowth forest in eastern Amazonia, Brazil.
Experiment Treatment Species Treatment x Species
1 (started in wet season) 3.66ns 15.96*** 0.64ns
2 (started in dry season; full period 0.91ns 10.26*** 1.7Ons
included in analysis)
2 (started in dry season; dry season 10.47** 11.56*** 1.64ns
only included in analysis)
3 (started in dry season; frequent 194.92*** 10.15*** 3.20*
sampling)

Table 4-2. Decomposition rates (mean standard error) for overstory species in a tropical
regrowth forest stand in eastern Amazonia, Brazil (n = 8).
Experiment 2
Experiment 1 (started in dry season)
(started in wet season) full period included dry season only
Species k (yr-1)
A. paludosa 0.97 + 0.051 a 1.26 + 0.09 a 1.21 + 0.09 a
L. pubescens 0.91 0.03 ab 1.02 0.03 bc 0.93 + 0.04 b
0. guianensis 0.78 0.03 bc 0.85 0.04 c 0.73 0.07 b
S. pulcherrimum 0.65 + 0.04 c
V guianensis 1.08 + 0.04 ab 0.91 + 0.07 b
1 Within a column, different letters indicate that means differ at P < 0.05 (Tukey's test).









Table 4-3. Decomposition rates (mean standard error) for overstory species under control and
irrigated plots (Experiment 3) in a tropical regrowth forest stand in eastern Amazonia,
Brazil (n = 4). This experiment started in the dry season and encompasses frequent
sampling during this period.
Control Irrigation
Species k (yr-1)
A. paludosa 0.59 + 0.041 Aa 1.52 0.08 Ba
L. pubescens 0.49 0.06 Aa 1.14 0.05 Bb
O. guianensis 0.48 0.03 Aa 1.00 0.06 Bb
V guianensis 0.39 0.03 Aa 1.02 + 0.14 Bb
1 Within columns and rows, different lower- and upper-case letters, respectively, indicate that
means differ at P < 0.05 (Tukey's test).

Table 4-4. Initial quality parameters (mean standard error) of leaves incubated in litterbags for
decomposition studies in a tropical regrowth forest in Eastern Amazonia, Brazil (n =
6-8).
Annona Lacistema Ocotea Vismia
paludosa pubescens guianensis guianensis
Carbon (%) 50.77 + 0.121 a 53.32 + 0.17 b 52.55 + 0.08 c 52.65 + 0.07 c
Nitrogen (%) 1.05 0.02 a 1.66 0.01 b 0.90 0.02 c 1.02 + 0.01 a
Phosphorus (mg g-1) 0.45 + 0.01 a 0.50 0.01 b 0.32 0.01 c 0.4 + 0.01 d
Lignin (%) 42.9 0.7 a 46.3 0.6 b 47.4 0.3 b 51.7 + 1.1 c
Cellulose (%) 37.9 + 0.3 a 42.0 0.4 b 29.4 1.1 c 41.8 0.8 d
Carbon : nitrogen 48.51 + 0.89 a 32.16 0.21 b 58.57 + 1.18 c 52.65 0.27 d
Lignin : nitrogen 40.98 + 1.03 a 28.05 0.33 b 52.82 + 1.15 c 50.24 0.90 c
Specific leaf area 165 13 a 191 15 a 66 +2b 122 + 6 c
(cm2 g1)
Within a row, different letters indicate that means differ at P < 0.05 (Tukey's test).









Table 4-5. Pearson correlation coefficients between decomposition rate (k) and initial quality
parameters of leaves of overstory tree species incubated in litterbags (Experiment 2)
in a tropical regrowth forest in Eastern Amazonia, Brazil (n = 4).
Leaf litter quality parameter Pearson coefficient
Nitrogen (%) 0.06206 (0.9379)1
Phosphorus (mg g-1) 0.56678 (0.4332)
Lignin (%) -0.44339 (0.5566)
Cellulose (%) 0.55625 (0.4437)
Carbon : nitrogen -0.27987 (0.7201)
Lignin : nitrogen -0.32252 (0.6775)
Specific leaf area (cm2 g1) 0.62462 (0.3754)
1 Numbers in parentheses are significance values

Table 4-6. Decomposition rates estimated from litterbag studies for some tropical forest sites.
Decomposition
Site rate (yr1) Source
Re'gi ,i\ Iih
Manaus, Brazil 0.47 -0.61 Mesquita et al. (1998)
Guadeloupe, French West Indies 0.41 2.39 Loranger et al. (2002)
Apeu, Brazil 0.39 1.52 This study
OA/1-g i Iil i/h
San Carlos de Rio Negro, Venezuela 0.58 -5.00 Cuevas and Medina (1988)
Maraca Island, Brazil 2.01 Scott et al. (1992)
Southern Bakundu Reserve Forest, Cameroon 1.55 4.6 Songwe et al. (1995)
Barro Colorado Island, Panama 3.2 Wieder and Wright (1995)
Maraca Island, Brazil 0.61 2.58 Luizdo et al. (1998)














Table 4-7. Chemical composition of leaf litter for some tropical forest sites.
Site Carbon (%) Nitrogen (%) Lignin (%) Celllulose (%) C:N Lignin:N SLA (cm2 g-1) Source
Venezuelaa 49 -57 1.12 -1.71 14.2 -26.3 17.3 -39.4 62 -77 Cuevas and Medina (1988)
Panamaa 69.2 -122.3 Comejo et al. (1994)
Venezuelaa 78 114c Reich et al. (1995)
Brazilb 47.4 -48.0 1.2 -1.3 53 -54 Mesquita et al. (1998)
Guadeloupeb 1.1 -2.5 22.8 -29.5 19.2 -20.9 11.7 -20.7 Loranger et al. (2002)
Panamaa 47.3 -43.2 0.90 -1.22 16.0 -13.7 18.4 -18.0 58.4 -39.2 11.8 -19.9 Santiago et al. (2003)
Brazild 48.4 0.94 53.5 51.7 44.0 Vasconcelos and Laurance
(2005)
a old-growth forest
b regrowth forest
c mid successional species
d mixed leaf litter of successional species











Experiment 1


0 control
S 0 irrigation


0 i
100
0
80-

60

40 -

20
G
0
100


0 60 120 180 240 300 360


Experiment 2



8

0


E


Experiment 3

o
O0Q
0'


0o
0


0


0 60 120 180 240 300 0 20 40 60 80 100


Time (d)
Figure 4-1. Effects of dry-season irrigation on leaf litter decomposition in a forest regrowth
stand in eastern Amazonia, Brazil. Remaining leaf mass of (A, B, C) Annona
paludosa, (D, E, F) Lacistema pubescens, (G, H, I) Ocotea guianensis, (J)
Stryphnodendron pulcherrimum, and (K, L) Vismia guianensis. Each symbol
represents the mean + standard error (n = 4). White and black horizontal bars mark
dry and wet seasons, respectively. Note different scales on the x-axes. Experiment 1
started in the wet season, while Experiments 2 and 3 started in the dry season;
Experiment 3 had more frequent sampling than the other experiments.


0 I
100


in 3
32 fi













A


0.9 0.9

0.8 0.8
0.6 0.8 1.0 1.2 1.4 1.6 1.8
Nitrogen (%)
1.4 1.4 -

1.3 D 1.3 -
HH
1.2 1.2 -

1.1 + 1.1 -

1.0 1.0 -

0.9 0.9 -
0.8 -- 0.8
40 42 44 46 48 50 52 54
Lignin (%)
1.4

1.3 G

1.2 -


I^


B


0.3 0.4 0.5 0.6
Phosphorus (mg g-1)

E *


C









30 40 50 60
C:N

F


+i 0.9
. 0.8 ,
27 30 33 36 39 42 45 20 30 40 50 60
Cellulose (%) Lignin:N


* Annona paludosa
v Lacistema pubescens
* Ocotea guianensis
0 Vismia guianensis


60 90 120 150 180210240
SLA (cm2 g-1)


Figure 4-2. Relation between decomposition rate (k) and initial leaf litter characteristics for tree
species in a forest regrowth stand in eastern Amazonia, Experiment 2. A) Nitrogen
concentration. B) Phosphorus concentration. C) Carbon:nitrogen ratio. D) Lignin
concentration. E) Cellulose concentration. F) Lignin:nitrogen ratio. G) Specific leaf
area. Each symbol represents the mean standard error for the y-axis (vertical error
bar) and the x-axis (horizontal error bar); n = 8 for k and n = 6-8 for litter quality.









CHAPTER 5
MOISTURE AND SUBSTRATE AVAILABILITY CONSTRAIN SOIL TRACE GAS
FLUXES IN AN EASTERN AMAZONIAN REGROWTH FOREST

Introduction

Tropical forests represent an important source of atmospheric greenhouse gases including

carbon dioxide (C02), nitrous oxide (N20), and methane (CH4), along with nitric oxide (NO), a

precursor to the photochemical production of tropospheric ozone (Vitousek & Matson 1992).

The production and consumption of these gases are strongly linked to the availability of both

soil moisture and decomposable substrate. However, seasonal cycles of precipitation, litterfall,

and decomposition are often confounded in ways that limit our ability to quantify the relative

importance of these interacting factors from seasonal observations of gaseous fluxes.

Observational studies in tropical forests have shown that higher soil moisture availability

during the wet season usually increases soil CO2 and N20 effluxes, decreases NO efflux, and

decreases CH4 consumption rates (Davidson et al. 2000, Fernandes et al. 2002, Garcia-Montiel

et al. 2001, Kiese & Butterbach-Bahl 2002, Kiese et al. 2003, Verchot et al. 2000, Verchot et

al. 1999). Fewer studies have evaluated the response of soil trace gas fluxes to experimental

manipulation of soil moisture availability in tropical forests. In a throughfall exclusion

experiment in the Tapaj6s National Forest, Brazil, emissions of N20 and CH4 were reduced by

the exclusion of about 50% of annual throughfall, but no treatment effect was observed for NO

or CO2 emissions (Davidson et al. 2004b). Addition of water to dry soil in short-term, small-

scale field studies has resulted in increased emissions of C02, NO, and N20 in wet (Garcia-

Montiel et al. 2003b, Nobre et al. 2001) and seasonally dry (Davidson et al. 1993) tropical

forest soils.

Regarding manipulative experiments of substrate availability (i.e., addition or removal of

aboveground litter), there are only two reports of long-term, large-scale field studies (Li et al.









2004, Sayer 2005) that have assessed emissions of soil CO2 in tropical forests in addition to this

study. In both studies, reduction of substrate availability through litter removal decreased soil

CO2 efflux, which is consistent with several related studies in temperate forests (Bowden et al.

1993, Jandl & Sollins 1997, Rey et al. 2002, Sulzman et al. 2005), but we encountered no

published reports of litter removal effects on NO, N20, and CH4.

Measurements of soil CO2 efflux and non-woody litterfall can be used to estimate total

belowground carbon allocation (TBCA) in forests (Raich & Nadelhoffer 1989). For mature

forests, TBCA is about two times aboveground litterfall, while for regrowth forests, TBCA is

about three times aboveground litterfall (Davidson et al. 2002, Raich & Nadelhoffer 1989),

indicating that regrowth forests allocate a relatively larger proportion of C to belowground

structures than mature forests (Davidson et al. 2002). Although TBCA represents the single

largest flux of C in forest ecosystems aside from canopy C assimilation (Davidson et al. 2002),

little is known about this flux of C in tropical forests.

A better understanding of how trace gas emissions from tropical forest soils are affected

by moisture and substrate availability can help to improve current biogeochemical models that

predict impacts of changes in climate and land-use practices on the atmospheric concentrations

of these gases (Potter & Klooster 1998). Such data, together with more estimates of total

belowground C allocation in tropical forests are also needed to better understand carbon

dynamics in regrowth forests (Johnson et al. 2000). Few such data are available for

Amazonian regrowth forests, a significant and dynamic component of forest landscapes in this

region (Fearnside 1996, Zarin et al. 2001).

Our primary objective in this study was to quantify the effects of moisture and substrate

availability on soil trace gas emissions in an Amazonian regrowth forest stand. In one









experiment, dry-season moisture limitation was reduced by irrigation. In the other experiment,

substrate limitation was provoked by litter removal. The dry-season irrigation and litter

removal experiments are described in Chapter 2.

Material and Methods

Field Measurements

Since July 2001, daily rainfall has been measured 500 m away from the experimental

area using a standard rain gauge. Prior to July 2001, rainfall data reported here are from the

National Agency of Electrical Energy (ANEEL) network meteorological station at Castanhal

(01 17' 53" S, 47 56' 56" W) which is no longer in operation and that was about 3 km away

from our site.

One tensiometer (Jet Fill Tensiometers, Soilmoisture Equipment Corp., Santa Barbara,

CA, USA) was installed at a depth of 10 cm in each plot and soil water potential was recorded

on a weekly basis in the morning. The number of actual replicates per treatment varied due to

loss of water column tension during the dry season.

Soil CO2 efflux was generally measured bi-weekly, beginning in March 2000, with an LI-

6400 portable photosynthesis system fitted with an LI-6400-09 soil CO2 flux chamber (LI-COR

Inc., Lincoln, NE, USA). The chamber was fit into circular polyvinyl chloride (PVC) collars

(11.5 cm internal diameter x 5.5 cm deep), which were installed approximately 2 cm into the

soil. Each plot contained three soil collars, spaced at least 1 m apart, totaling 12 collars per

treatment and sampling date. No live vegetation was contained within the collars.

Measurements were taken between 0630 and 1100 hours.

To better understand the results of CO2 flux analyses within the context of stand-level C

dynamics, we also collected data on litterfall (Chapter 3). We estimated that non-woody

litterfall was 48% C based on the monthly non-woody litterfall C concentration (47.9 + 0.2%)









during the period of October 1999 to March 2001. Non-woody litterfall was 80 to 90% of total

litterfall. Woody litterfall data are not reported here because of its much smaller impact on

short-term trace gas emissions due to its slow turnover rate.

Two additional PVC collars with 20 cm diameter and 10 cm height were installed within

each plot (total of 8 collars per treatment and sampling date) and inserted approximately 2-3 cm

into the soil for measurement of soil NO, N20, and CH4 gas fluxes. During the measurements,

a vented PVC cover made from the end cap of a 20-cm diameter PVC pipe was fit into the

collars. On average, NO, N20, and CH4 flux measurements were made every two months,

beginning in August 1999. The flux measurement technique for NO used a chemiluminescence

detector (Scintrex LMA-3, Scintrex Limited, Concord, ON, Canada) as described by Verchot et

al. (1999). N20 and CH4 fluxes were measured by gas chromatography analyses of four

syringe samples extracted from the same chambers at 10-minute intervals (Verchot et al. 2000,

Verchot et al. 1999). The PVC collars used for soil trace gas measurements were left in place

throughout the course of the experiments.

To augment our understanding of the N gas fluxes (NO and N20), we also include here

results of potential soil nitrification determined with a variation of the aerobic incubation

method (Hart et al. 1994). Nitrification is the precursor to the denitrification process, and both

processes produce NO and N20 (Firestone & Davidson 1989). For each plot, we analyzed one

composite sample made of four samples collected at a depth of 10 cm in October 2001. We

estimated net N nitrification from changes in nitrate concentrations during 7-day incubation of

soil. We corrected soil gravimetric moisture to 75% field capacity before sample incubation at

about 28 C in an incubator (Isuku FR24BS, Isuku Seisakusho Ltd., Tokyo, Japan). We did

extractions of samples in 2 M potassium chloride (KC1) three days after collection in the field









and in incubated samples. We kept samples under refrigeration (4 C) prior to the initial

extraction. We filtered extracts through Whatman No. 42 filter paper before analysis of

nitrite/nitrate using a flow-injection system on a Lachat QuikChem AE autoanalyzer (Lachat

Instruments, Milwaukee, WI, USA). Prior to the extractions, we dried subsamples of soil for

24 hours at 105 C to determine actual moisture content.

Statistical Analyses

We used the SAS System version 9.00 to run the statistical analyses. We analyzed with

PROC MIXED the effects of treatment, date, and treatment-by-date interaction on the variables

trace gas flux, soil water potential, and non-woody litterfall using a repeated measures analysis

with compound symmetric covariance structure. This structure assumes constant variance at all

dates and equal correlations between all pairs of measures on the same experimental unit, i.e.,

collar, tensiometer, or trap for the soil trace gases, soil water potential, and litterfall variables,

respectively. We ran separate tests to compare each of the treatments with the control. Within

this analysis, significant treatment effects would have indicated temporally consistent

differences between treatment and control measurements both pre- and post-treatment and

across seasons (none were observed), significant date effects were generally indicative of

seasonal trends that affected both treatment and control measurements, and treatment-by-date

effects indicated a significant difference between treatment and control measurements that

occurred after the treatment was initiated. We used CONTRAST statements to explicitly test

whether the measured variables differed between seasons and between treatments within each

season (wet and dry). We used the TTEST procedure to compare treatments and control means

for soil nitrification.

When necessary, we performed log and square root transformations to meet the model

assumptions of normality, based on the criteria of P > 0.05 in the Kolmogorov-Smirnov test,









and equal variances, based on the absence of a pattern of heteroscedasticity in the plots of

residual versus predicted values. Means and standard errors were calculated on the basis of

untransformed data. All results are reported as significant when P < 0.05; we report marginal

significance when 0.05 < P < 0.10.

We estimated annual soil C efflux by linear interpolation between sampling dates using

the EXPAND procedure. To estimate annual soil C efflux, we assumed that the variation in

soil CO2 efflux with time of day was minimal as previously reported by Davidson et al. (2000)

for an eastern Amazonian primary forest. We tested for interannual and between treatment

differences in annual soil C efflux and annual litterfall C values for control and irrigated plots

in 2001 and 2002 using the PROC MIXED procedure. For the litter removal vs. control plot

comparison of annual soil C efflux and annual litterfall C we used the TTEST procedure for

2002 data only; we did not include the 2001 data in the litter removal vs. control comparison

because the treatment regime was not initiated until August 2001. We estimated the relative

contribution of aboveground litter to soil respiration by subtracting litter removal soil CO2

efflux from control soil CO2 efflux.

Results

Irrigation Experiment

Rainfall declined from mid-July to early-January (dry season) during each year of the

study (Figure 5-1A), resulting in lower soil water potential during this period (Figure 5-1B).

The dry-season irrigation resulted in significantly (P < 0.0001) less negative soil water potential

in control plots for most of the dates in 2001 and 2002 (Figure 5-1B). During the 2001 dry

season, soil water potential was -0.052 + 0.003 and -0.024 + 0.002 MPa in control and irrigated









plots, respectively; corresponding values for the 2002 dry season were -0.046 0.003 and -

0.013 0.002 MPa.

There was a significant effect of date and the interaction between treatment and date on

soil CO2 efflux (Table 5-1). Soil CO2 efflux for irrigated plots was significantly higher than for

control plots during the dry-season irrigation (P < 0.0001, Figure 5-1C). There was also a

significant effect of date and the interaction between treatment and date on soil CO2 efflux for

the pretreatment period (P < 0.0001); however, pretreatment differences between plots did not

affect the significance of the dry-season irrigation effect. In the 2001 dry season irrigation

period, soil CO2 efflux values were 3.91 + 0.13 and 5.54 0.19 itmol CO2 m-2 s-1 for control

and irrigated plots, respectively; corresponding values for the 2002 dry season were 4.76 0.19

and 6.21 0.25 itmol CO2 m-2 s1.

The lowest mean soil CO2 efflux rate (2.33 0.19 imol CO2 m-2 S-1), which occurred in

the control treatment on 24 October 2001 (Figure 5-1C), coincided with a successive decrease

in soil water status (to -0.084 MPa) caused by a long dry spell of 24 days without rain out of a

total of 31 days, with total precipitation of only 9 mm during the 31-day period. A 93%

increase in the control plot soil CO2 efflux in the subsequent measurement coincided with an

increase in soil water status (to -0.008 MPa) following two consecutive rainy days (19 and 26

mm) after the long dry spell, and immediately prior to the soil respiration measurement; no

increased soil CO2 efflux was observed for irrigated plots. The pulse in soil CO2 efflux was

then followed by a decrease in CO2 emissions associated with another dry period.

Annual soil C efflux was significantly higher in 2002 than in 2001 (P < 0.0001) (Table 5-

2). The effects of treatment and the interaction between treatment and date were marginally

significant (P < 0.07 and P < 0.10, respectively). Annual litterfall C was not affected by









treatment or year (Table 5-2); although the interaction between treatment and year was

marginally significant (P < 0.053) it is not readily attributable to a treatment effect.

The significant effect of date on NO efflux (Table 5-1, Figure 5-2B) was largely due to a

single value measured in the end of July 2002; wet vs. dry season contrasts indicated non-

significant seasonal differences in NO efflux. For N20 (Figure 5-2C), the wet season efflux

was significantly higher than the dry season efflux (5.62 0.50 and 2.41 0.47 [tg N m-2 h-1,

respectively; P< 0.0001). During dry-season irrigation, treatment vs. control contrasts

indicated that N20 efflux in irrigated plots was significantly higher than in control plots (4.18 +

0.87 and 2.34 0.75 tg N m-2 h-1, respectively; P < 0.05).

Date was again the only factor to have a significant effect on CH4 efflux (Table 5-1,

Figure 5-2D). Methane efflux in the dry season was significantly lower than in the wet season

(-0.348 0.118 and 0.128 0.118 mg CH4 m-2 d-1, respectively; P < 0.0001). During dry-

season irrigation, treatment vs. control contrasts indicated that CH4 efflux in irrigated plots was

also significantly higher than in control plots (0.226 0.361 and -0.526 0.185 mg CH4 m-2 d-

, respectively; P < 0.01). The net CH4 emissions were generally close to zero, with most

chambers generally showing net uptake of CH4 (77% in control plots and 80% in irrigated

plots). The range of CH4 efflux for the whole experimental period was -5.00 to 22.03 mg CH4

m-2 d-1. Two chambers with very high effluxes (5.93 and 9.97 mg CH4 m-2 d-1) drove the large

variability in the mean efflux for the control plot in March 2001, while the high variability for

the irrigation means in September and October 2001 was driven by one chamber (9.08 and

10.30 mg CH4 m-2 d-1). The apparent high mean net production of CH4 for irrigated plots in

September (0.884 1.353 mg CH4 m-2 d-1) and October (0.879 1.187 mg CH4 m-2 d1) 2001









becomes net consumption (-0.461 0.172 and -0.292 0.223 mg CH4 m-2 d-1) if we exclude

the high efflux chambers from the calculation of means.

There was no significant effect of irrigation on net nitrification rates for control and

irrigated plots (0.11 0.02 and 0.11 0.03 |tg N g-1 soil d-1, respectively).

Litter Removal Experiment

Soil water potential (Figure 5-3B) was significantly less negative in the wet season than

in the dry season (P < 0.0001). Soil CO2 efflux during the pretreatment period (Figure 5-3C)

for litter removal and control plots did not differ significantly (4.18 0.12 and 4.24 0.08

mol CO2 m2 s-1, respectively; P = 0.87). During the litter manipulation period, soil CO2

efflux in litter removal plots was significantly lower than in control plots (3.54 0.17 and 4.90

+ 0.18 mol CO2 -2 s-1, respectively; P < 0.001). This difference was not homogeneous

throughout the experimental period and followed a trajectory that can be divided in three

phases. In the first phase, corresponding with the dry season and the early rainy season, the

difference between treatment and control measurements was apparent for nearly all of the

measurements made during the first six months of litter removal. The second phase, from 6-10

months after the beginning of litter removal, corresponded with the mid to late rainy season.

During this phase, there were fewer measurements in which the difference between treatment

means was significant. In the third phase, corresponding with the following dry season, the

difference in soil CO2 efflux between treatments was uniformly significant, and persisted

through the end of the measurement period.

Aboveground litter respiration represented 22 2% of total soil respiration for the whole

litter removal period and was 22 2, 16 4, and 28 2% of total soil respiration during the

first, second and third phases, respectively. Annual soil C efflux was significantly lower (P <









0.05) in litter removal than in control plots in 2002 (Table 5-2). There was no significant

difference in annual litterfall C between control and litter removal treatments in 2002 (Table 5-

2).

The significant interaction effect on N oxide emissions (Table 5-1) was not related to a

consistent effect of litter removal on either NO or N20 effluxes (Figures 5-4B-C, respectively);

the difference between treatments for both gases during the litter removal period was non-

significant. Emissions of CH4 (Figure 5-4D) in the dry season were significantly lower than in

the wet season (-0.420 0.164 and 0.287 + 0.113 mg CH4 m-2 d1, respectively; P < 0.01).

Mean net nitrification rates in control plots were marginally higher (P = 0.06) than in

litter removal plots (0.11 0.02 and 0.07 0.01 |tg N g-1 soil d-1, respectively).

Discussion

Soil CO2 Efflux and Belowground C Allocation

The soil CO2 efflux rates measured in our study are within the range of data reported for

tropical forests and are consistent with several other studies in Amazonian forests (Cattdnio et

al. 2002, Davidson et al. 2004b, Davidson et al. 2000, Fernandes et al. 2002, Nepstad et al.

2002, Salimon et al. 2004, Verchot et al. 2000) and in tropical forests elsewhere (Ishizuka et al.

2002, Kiese & Butterbach-Bahl 2002) that reported higher emissions of CO2 during the wet

season than in the dry season. We have also shown strong pulses of CO2 efflux in response to

rain events during dry periods (soil wet-up events), as observed in old-growth forests in the

Brazilian Amazon (Davidson et al. 2000, Sotta et al. 2004) and in Costa Rica (Schwendenmann

et al. 2003). Our dry-season irrigation experiment further demonstrates the constraint that

moisture availability exerts on soil CO2 efflux.









Soil CO2 efflux as measured in the field mainly integrates root and microbial respiration,

and we have not determined if the reduction in soil respiration in the dry season was caused by

decreased activity of microbes, roots or both. However, a laboratory study with soil from the

same site showed a significant increase in microbial basal respiration during the 2001 wet

season compared to the previous dry season (Rangel-Vasconcelos 2002), as observed in other

tropical forests (Cleveland et al. 2003, Luizao et al. 1992). Although microbial respiration rates

determined under laboratory conditions cannot be compared to rates obtained in the field with

chamber techniques, those results suggest that reduction in soil microbial activity during the

dry season likely contributed to the observed lower rates of soil respiration during this period at

our site. Likewise, reduced activity of microbes in decomposing aboveground litter during the

dry season could have contributed to lower soil CO2 efflux in non-irrigated plots. Borken et al.

(2003) have recently shown that microbial respiration of the 0 horizon can contribute

significantly to CO2 pulses after soil wet-up events in a temperate forest and Goulden et al.

(2004) reported that increased soil respiration after a rainfall during the dry season was

associated with surface litter rehydration in an Amazonian old-growth forest. Wieder and

Wright (1995) have also observed higher litter mass loss under irrigation compared with no

irrigation in a tropical forest in Panama. Finally, lower soil CO2 efflux during the dry season

could also have resulted from constrained root respiration due to decreased root growth

(Cattdnio et al. 2002) or decreased flux of photosynthates to roots, which limits root respiration

itself (Hogberg et al. 2001) and/or rhizospheric microbial respiration (Kuzyakov & Cheng

2001). Further research on differentiating root from microbial respiration and aboveground

litter from soil respiration are needed to better understand how moisture constrains CO2 efflux

from tropical forest soils, especially because likely concomitant and opposite variations in root









and microbial dynamics under dry conditions (Davidson et al. 2004b) make it difficult to

understand the mechanisms by which moisture controls total soil respiration.

The variation in the size of the difference of soil CO2 efflux between control and litter

removal throughout the manipulation period followed a trajectory that can be linked to altered

substrate availability and variation in soil water status due to the seasonality of rainfall. In the

first phase of this trajectory, the early impact of litter removal on soil respiration suggests that

CO2 efflux associated with microbial decomposition of aboveground litter and superficial root

respiration represents a substantial proportion (about 22 % in the present study) of total soil

respiration (Raich & Schlesinger 1992). During the second phase, an interaction between

substrate availability and rainfall seasonality appears to influence the variation in soil CO2

efflux. The difference between control and litter removal plots decreased during some dates in

the second phase, suggesting that the contribution of belowground respiration was relatively

higher during the wet season. The third phase may be characterized by the depletion of labile

soil carbon and, therefore, an increase in the difference in soil CO2 efflux between treatments.

Although this phase is also coincident with the 2002 dry season, its length and consistency (i.e.,

lack of responsiveness to dry-season wet-up events) lead us to suspect that, due to the removal

of the litter layer, substrate availability has become a larger constraint on soil respiration than

reduced moisture availability. In 2002, litter removal resulted in a 28% reduction in soil CO2

efflux, which is very similar to the 27% found in Costa Rica after 2 years of litter removal

(Sayer 2005), but lower than the 54% reduction after 7 years of litter removal in Puerto Rico

(Li et al. 2004).

The estimated annual soil C efflux measured in our control plots is comparable to another

estimate for eastern Amazonian forests in Brazil (Davidson et al. 2000) and is higher than









estimates for tropical old-growth forests elsewhere (Ishizuka et al. 2002, Schwendenmann et al.

2003); annual fluxes measured in other tropical forest sites are given in Table 5-3. We

observed lower total rainfall and higher annual soil C efflux in 2002 than in 2001, suggesting

that the interannual variability in soil C efflux was not caused by differences in annual rainfall.

Pulses of CO2 associated with rainfall events observed in this study are consistent with the

hypothesis that rainfall distribution, rather than total rainfall, may better explain annual

variability in soil C efflux. Differences in annual soil C efflux between irrigation and control

plots are also consistent with a substantial moisture constraint on soil respiration.

Based on our annual soil C efflux and non-woody litterfall C estimates (Chapter 3), we

can calculate a C efflux : litterfall C ratio of 4.0 5.2 for our control plots, consistent with the

mean value of 4.16 reported by Davidson et al. (2002) for young forests. Total belowground

carbon allocation (TBCA) estimated by the difference between annual basis C fluxes in soil

respiration and litterfall (Raich & Nadelhoffer 1989) is underestimated for regrowth forests if C

storage in roots and soil is not accounted for (Davidson et al. 2002). However, simple

calculation of TBCA based only on soil respiration and litterfall can provide a lower limit of

TBCA for regrowth forests. For our site, the ratio between annual soil C efflux and annual

litterfall C indicates that TBCA relative to litterfall is similar to values for other regrowth forest

site in the eastern Amazon (Davidson et al. 2002) and higher than those of mature forests

(Davidson et al. 2002, Raich & Nadelhoffer 1989), consistent with increased allocation of C to

belowground structures as a mechanism by which regrowth forests cope with the demands for

water and nutrients (Davidson et al. 2002).

Differences in annual soil C efflux between litter removal and control plots are consistent

with a substantial substrate constraint on soil respiration. In 2002, the amount of carbon in









litterfall (368 14 g C m-2 yr-1) was well within one standard error of the mean difference in

soil C efflux between control and litter removal (559 291 g C m-2 yr-1). This substantial

difference in soil C efflux also suggests that -20% of total soil C efflux is due to litter

respiration, with the remaining -80% due to belowground respiration; this is consistent with

results obtained in litter removal studies in forest ecosystems in the tropics (Li et al. 2004,

Sayer 2005) and other climatic regions (Bowden et al. 1993, Jandl & Sollins 1997, Rey et al.

2002).

Nitrogen Oxide Emissions

Nitric and nitrous oxide effluxes measured in this study both in wet or dry seasons are

among the lowest reported for either regrowth or old-growth tropical forests in the Brazilian

Amazon (Cattanio et al. 2002, Davidson et al. 2004b, Garcia-Montiel et al. 2001, Nepstad et al.

2002, Verchot et al. 1999) and tropical forests elsewhere (Erickson et al. 2001, Ishizuka et al.

2002, Palm et al. 2002). These low N oxide effluxes may result from low rates of N cycling, as

indicated by the very low net nitrification rates we found in both seasons compared to other

studies for Amazonian forests (Garcia-Montiel et al. 2003a, Neill et al. 1997, Palm et al. 2002).

The thin concretionary soils of this site, along with the recent history of repeated slash-and-

burn cycles and the high litterfall C:N ratios, are consistent with a very conservative nitrogen

cycle and low rates of both nitrification and denitrification. Although fluxes were consistently

low, slightly higher efflux of N20 in the wet season compared to the dry season reported here

has also been observed in other tropical forests (Cattanio et al. 2002, Erickson et al. 2002,

Garcia-Montiel et al. 2001, Kiese & Butterbach-Bahl 2002, Nepstad et al. 2002, Verchot et al.

1999). Consistent with the results obtained by Nobre et al. (2001), we also found a significant

effect of irrigation on N20 efflux. Higher N20 efflux associated with wetter soil conditions









during both the wet season and dry-season irrigation periods likely resulted from increased

denitrification (Davidson 1991).

The effects of litter removal on N oxide fluxes are not clear and difficult to interpret since

the fluxes are inherently very low at our site. If N trace gas emissions were already limited by

N availability in this infertile soil, the removal of litter might be expected to decrease emissions

further. However, that decrease would be difficult to detect relative to the naturally low

emissions that were already frequently near detection limits prior to litter removal.

Methane Emissions

Methane fluxes measured at our site are in the lower range of both net consumption and

net production fluxes found for tropical forests (Kiese et al. 2003, Palm et al. 2002, Verchot et

al. 2000). Higher net consumption of CH4 in the dry season and lower net consumption (or

even small net production) in the wet season observed in our study is consistent with the pattern

of CH4 emissions measured in other Brazilian Amazonian forests (Cattanio et al. 2002, Nepstad

et al. 2002, Verchot et al. 2000) and tropical forests elsewhere (Kiese et al. 2003). Increased

net CH4 production during the wet season as well as during the irrigation period in our study

suggests that higher soil water status decreased soil aeration leading to an increase in

methanogenesis (Davidson & Schimel 1995). Although decreased aeration during the wet

season could have resulted in higher efflux of CH4 and N20, higher soil CO2 efflux associated

with wetter soil conditions could also have contributed to the increased efflux of CH4 and N20

because of the consumption of 02 in the respiration process (Palm et al. 2002, Verchot et al.

2000).

Conclusions

We conclude that soil C02 efflux is strongly linked to soil moisture and substrate

availability as indicated from the responses of C02 emissions to soil wet-up events, dry-season









irrigation, and litter removal for this tropical regrowth stand. On an annual basis, this regrowth

stand allocates a large amount of C to belowground structures relative to litterfall C. Relieving

dry season water limitation increased soil respiration by about 40 and 30% in the two dry

seasons studied, corresponding to annual increases of 27 and 13% in 2001 and 2002,

respectively. Removing aboveground litter reduced annual soil respiration by 28% in 2002.

In general, N oxide emissions were very low, probably due to the inherently low rates of

nitrogen cycling at this site. Emissions of N20 and CH4 were constrained by low moisture

availability, while emissions of NO were not affected by irrigation. We were unable to detect

more severe substrate limitation induced by the litter removal treatment on N oxide and CH4

emissions.

The substantial impacts of soil moisture and aboveground litter on soil CO2 efflux shown

in this study suggest that alterations in the availability of these resources that may result from

climate and land-use changes in tropical regions could have significant effects on regional CO2

fluxes.












Table 5-1. F statistics and associated significance levels for the effect of treatments (irrigation and litter removal), sampling date, and
their interaction on soil trace gas fluxes and non-woody litterfall in a tropical regrowth forest stand in eastern Amazoniaa
(PROC MIXED, SAS System version 9.0). Significant "treatment" effects (not observed) would indicate temporally
consistent differences between treatment and control both pre- and post-treatment and across seasons, significant "date"
effects are generally indicative of seasonal trends that affect both treatment and control measurements, and "treatment x
date" effects indicate a significant difference between the treatment and control measurements that occurs after the
treatment was initiated.
Irrigation experiment Litter removal experiment
Variable Treatment Date Treatment x Date Treatment Date Treatment x Date
CO2 efflux 2.55ns 9.48*** 5.02*** 3.10ns 9.63*** 3.24***
NO efflux 0.04ns 5.46*** 1.50ns 3.29ns 7.65*** 2.21*
N20 efflux 0.93ns 4.20*** 1.00ns 0.32ns 6.42*** 1.68*
CH4 efflux 0.91ns 2.14** 1.22ns < 0.01ns 2.21** 0.77ns
Litterfall 0.24ns 45.27*** 1.62** 0.27 ns 32.91*** 1.18ns
aThe level of significance is indicated (*: P < 0.05, **: P < 0.01, ***: P < 0.001, ns: not significant).









Table 5-2. Annual soil carbon efflux and non-woody litterfall carbon for control, irrigated and
litter removal plots in a tropical regrowth forest stand in eastern Amazonia (mean +
se, n = 12 per treatment).
Soil C efflux (g m-2 yr-1) Non-woody litterfall C (g m-2 yr-1)
Treament 2001 2002 2001 2002
Control 1593 74 1988 126 410 28 383 27
Irrigation 2021 + 154 2237 158 398 24 415 24
Litter removal 1429 165 368 14









Table 5-3. Estimates of annual soil carbon (C) efflux in old-growth and regrowth tropical
forests.

Rainfall Soil C efflux
Forest Location (mm yr-1) Soil type (g C m-2 yr-1) Reference


OI,1-g1i1 1 il1
Para, Brazil (2 59' S, 1800
47o 31' W)
Sumatra, Indonesia (1 2060
05.164' S, 102 05.702'
E)
La Selva, Costa Rica 4200
(10 20' N, 83 50' W)

Para, Brazil (2.8968 oS, 2000
54.9519 oW)
Acre, Brazil 1940




Barro Colorado, 2600
Panama
Para, Brazil (2 64' S, 2000
54 59' W)


Regi,/i n iF12
Para, Brazil (2 59' S,
47 31' W), 20-year-old

Gran Sabana,
Venezuela (5 0' S, 61
O' W)


1800


2200


Acre, Brazil, 3-18-yr- 1940


Para, Brazil


Haplustox

Ultisol


Typic
Haploperox
"old alluvium
Haplustox


dystrophic and
eutrophic
Ultisols with
patches of
Oxisols
Oxisol


2000

560-820


1060


1000

1700


1740


clayey (Ultisols 1084
and Oxisols)
sandy (Ultisols) 1363


Oxisol and
alfisol


Acrohumox



dystrophic and
eutrophic
Ultisols
Oxisol


1800


896 (tall)
1241
(medium)
1024 (low)
1600


1790


Davidson et al.
(2000)
Ishizuka et al. (2002)



Schwendenmann et
al. (2003)

Davidson et al.
(2004b)
Salimon et al. (2004)


Sayer (2005)


Silver et al. (2005)

Silver et al. (2005)


Davidson et al.
(2000)


Priess and Folster
(2001)


Salimon et al. (2004)


This study


1 Includes sites classified as mature and primary forests.
2 Includes sites classified as secondary forests.













100 A

E 80 -







80
S 60 -

2 40 o



iIII iI.1I






-0.04
0o

-0.06

-0.08 control
0 irrigation
-0.10




E 6


4


o 2
0

0 i i i i i i i i i
02/00 08/00 02/01 08/01 02/02 08/02 02/03

Date


Figure 5-1. Effects of rainfall patterns and dry-season irrigation on soil moisture status and soil
respiration in an Amazonian regrowth forest stand, Brazil. A) Daily rainfall at the
study site. B) Soil water potential. C) Soil carbon dioxide (C02) efflux. In Figures
B-C, circles represent means (+ se); n = 4 for soil water potential and n = 12 for soil
CO2 efflux per sampling date. Gray-shaded areas indicate the dry season irrigation
periods. White and black horizontal bars mark dry and wet seasons, respectively.












120 -
100 A
80 -
-E 60 -
0 40
20
14 -
B
12 0 control
0 irrigation
10 -

0 E 8T
ZZ 6 I i i i ii


2
0-


S 10-



















07/99 01/00 07/00 01/01 07/01 01/02 07/02 01/03
-5

D
X 2





cE -1




Date


Figure 5-2. Effects of rainfall patterns and dry-season irrigation on soil nitrogen oxide and
methane effluxes in an Amazonian regrowth forest stand, Brazil. A) Daily rainfall at
the study site. B) Soil nitric oxide (NO) efflux. C) Soil nitrous oxide (N20) efflux.
D) Soil methane (CH4) efflux. In Figures B-D, closed and open circles represent
means (+ se) for control and irrigation treatments, respectively (n = 8 per sampling
date). Gray-shaded areas indicate the dry season irrigation periods. White and black
horizontal bars mark dry and wet seasons, respectively.













100 A

E 80 -

T 60 -

2 40 -





0.00

-0.02 -B
ii-i

-0.04
0

-0.06 -

o -0.08 control
stu0 litter removal
-0.10

C"





w 4


B 2-
0


02/00 08/00 02/01 08/01 02/02 08/02 02/03

Date

Figure 5-3. Effects of rainfall patterns and litter removal on soil moisture status and soil
respiration in an Amazonian regrowth forest stand, Brazil. A) Daily rainfall at the
study site. B) Soil water potential. C) Soil carbon dioxide (CO2) efflux. In Figures
B-C, circles represent means (+ se); n = 4 for soil water potential and n = 12 for soil
CO2 efflux per sampling date. The vertical line indicates the beginning of the litter
removal treatment. White and black horizontal bars mark dry and wet seasons,
respectively.










120 -
100- A
S 80 -
E~ 60 -
0 40







ZZ 6- i
o 0











10 -
0 Y




o 4



o E 5




S 1 -T

__ # 1 0



07/99 01/00 07/00 01/01 07/01 01/02 07/02 01/03
Date

Figure 5-4. Effects of rainfall patterns and litter removal on soil nitrogen oxide and methane
effluxes in an Amazonian regrowth forest stand, Brazil. A) Daily rainfall at the study
site. B) Soil nitric oxide (NO) efflux. C) Soil nitrous oxide (N20) efflux. D) Soil
methane (CH4) efflux. In Figures B-D, closed and open circles represent means (+
se) for control and litter removal treatments, respectively (n = 8 per sampling date).
The vertical line indicates the beginning of the litter removal treatment. White and
black horizontal bars mark dry and wet seasons, respectively.









CHAPTER 6
MOISTURE CONSTRAINTS TO ABOVEGROUND NET PRIMARY PRODUCTIVITY IN
EASTERN AMAZONIAN FOREST REGROWTH

Introduction

Net primary productivity (NPP) is considered to be the best integrator measure of resource

effects on ecosystem processes (Chapin & Eviner 2005). Improved understanding of temporal

shifts in NPP may aid predictions of ecosystem response to ongoing climate and land-use

changes (Tian et al. 1998). In tropical forests, reliable estimates of NPP mostly involve

measurements of aboveground net primary productivity (ANPP) components; due to

methodological difficulties belowground NPP is rarely measured (Clark et al. 2001b). For

tropical forest regrowth (e.g. following agricultural conversion and abandonment), there is a

paucity of data even on ANPP, in part because these sites are very rarely measured over multiple

years.

Aboveground biomass increment in live trees (i.e., wood increment) and non-woody

litterfall (a proxy for leaf production) are commonly used to estimate ANPP; both aboveground

biomass increment and non-woody litterfall can be relatively easily measured and represent two

significant components of total ANPP (Clark et al. 2001 la). Stem diameter and height measures

are usually used to estimate aboveground biomass (AGB) through allometric equations (e.g.,

Ducey et al. Submitted). Despite several reports on AGB for tropical forest regrowth (Gehring

et al. 2005, Saldarriaga et al. 1988, Zarin et al. 2001), repeated measures of AGB and litterfall

are rare and calculations of ANPP for these forests are therefore lacking.

Observational and manipulative experiments suggest that moisture availability may be an

important control over ANPP in tropical forests. At old-growth forest sites in the Brazilian

Amazon, higher diameter growth rates are associated with wetter periods (Higuchi et al. 2003,

Rice et al. 2004, Vieira et al. 2004). Nepstad et al. (2002) have previously shown that soil









moisture depletion during a partial throughfall exclusion experiment reduced ANPP in an old-

growth Amazonian forest. Conversely, excessive soil moisture may also decrease ANPP

(Schuur & Matson 2001). Analogous data from both observational and manipulative studies are

lacking for tropical forest regrowth, even though recent estimates indicate that there are -38

million ha of regrowth in Latin America alone, and the area is growing as unproductive

deforested land is abandoned (ITTO 2002).

The primary objective of this chapter was to investigate the response of ANPP to

experimentally increased dry-season moisture availability and inter-annual variability in dry-

season precipitation during a four-year irrigation experiment described in Chapter 2. We

hypothesized that dry-season irrigation would increase ANPP, and that ANPP would also be

positively correlated with dry-season precipitation.

Study Site and Experimental Design

Study site and experimental design are described in Chapter 2.

Material and Methods

Aboveground Net Primary Productivity

Aboveground net primary productivity (ANPP) was estimated as the sum of annual

increases in aboveground biomass (AGB) of trees (diameter at breast height > 1 cm) and non-

woody litterfall (Clark et al. 200 l1a, Grace et al. 2001) between July 2001 and July 2005. To

estimate AGB, we used site-specific mixed-species and species-specific allometric equations

based on diameter measurements (Table 6-1; Ducey et al. In preparation). Diameter increments

have previously been published, in part, by Araujo et al. (2005). Non-woody litterfall data are

reported on Chapter 3.

Aboveground biomass increment (AGBI, in Mg ha-1 yr-1) was calculated for each plot as

follows (Clark et al. 2001a):









ABGI = (I increments of surviving trees) + (I increment(s) of ingrowth),

where the increment of surviving trees was calculated as the AGB in yearx+1 minus the AGB in

the previous year (i.e., yearx, and the increment of ingrowth was calculated as the AGB in the

ingrowth year minus AGB relative to the minimum diameter (1 cm).

This method of calculating AGBI may underestimate its actual value if trees exhibit

significant growth between their last measurement and their death. In a separate study, the

increment in diameter at breast height (DBH) of trees with DBH _> 5 cm was measured every 1-2

months from November 2003 to December 2005 using dendrometer bands fabricated with

aluminium tapes (data not presented). We observed that several months prior to tree death, stem

increment was consistently equal to zero, suggesting that unaccounted diameter increment prior

to tree death (Clark et al. 200 l1a) may have little impact on biomass increment estimates.

Statistical Analysis

We used the SAS System version 9.00 to run the statistical analyses. We analyzed with

PROC MIXED the effects of treatment, date, and treatment-by-date interaction on ANPP using a

repeated measures analysis with compound symmetric covariance structure. This structure

assumes constant variance at all dates and equal correlations between all pairs of measures on the

same experimental unit, i.e., plot. We used PROC NLIN for linear regression analysis between

ANPP and rainfall (current- and previous-year annual rainfall and dry-season rainfall); annual

rainfall corresponds to total rainfall in the interval between yearly diameter measurements. All

results are reported as significant when P < 0.05; we report marginal significance when 0.05 < P

< 0.10. Multiple comparisons of means were performed with Tukey's test.

Results

ANPP was significantly affected by date (P = 0.034) and treatment (P = 0.026), with

marginally significant effects of treatment x date interaction (P = 0.059). In the annual periods









from July 2002 to July 2003, and July 2003 to July 2004, ANPP was significantly higher in

irrigated plots than in control plots (Figure 6-1). ANPP was also positively correlated with

previous-year dry-season rainfall (Figure 6-2; R2 = 0.45; P < 0.01).

Discussion

Aboveground net primary productivity range calculated for this site (12.3 + 0.5 to 16.6 +

2.1 Mg ha-1 yr-1, n = 4, control plots) is equivalent to the highest values reported by Clark et al.

(2001b) for old-growth tropical forests. Our estimate of ANPP for this site represents a lower

bound, because it only includes wood increment and non-woody litterfall.

Although our calculated ANPP values are relatively high, the aboveground biomass

accumulated about 12 years after land abandonment (51.5 2.6 Mg ha-1) is 13% lower than the

value obtained (59.2 Mg ha-1) with a model developed to predict aboveground biomass

accumulation by Amazonian regrowth forests (Zarin et al. 2001), and substantially lower (>

70%) than the value predicted by the model developed for regrowth forests recovering from first-

cycle slash-and-burn in central Amazonian regrowth forests (Gehring et al. 2005). Lower

biomass compared to model predictions may result from (a) the history of repeated burning

events (Zarin et al. 2005), (b) the inherent low fertility of the concretionary soil, and (c)

relatively distinct dry season periods at the study site (Apeu). While the difference from Zarin et

al. 's (2001) is within the model error, the great discrepancy in relation to Gehring et al. 's (2005)

work may be due to less severe dry seasons and more fertile soils for central Amazonian

regrowth forests (Gehring et al. 2005).

While the results of our dry-season irrigation experiment demonstrate the constraint of

moisture availability on ANPP at this site, reduced ANPP associated with lower previous-year

dry-season rainfall indicates a lag effect of the influence of drought on ANPP. A recent