Carbon Storage Abstracts - MASTER
Transcrição
Carbon Storage Abstracts - MASTER
Carbon Storage and Exchange 79 Carbon dioxide measurements over pasture and Amazonian forest during the wet season Atmospheric Measuring Campaign LBA Niek J. Bink 1 Jan Elbers 2 Friso Holwerda 3 Pavel Kabat 4 Bart Kruijt 5 Rosivaldo Lelis da Silva 6 Beatriz Machado Gomes 7 Antonio Ocimar Manzi 8 Paulo Jorge Oliveira 9 Celso von Randow 10 Maarten Waterloo 11 12 Fabrício B. Zanchi During the season of 1999 carbon dioxide measurements were carried out over two pasture sites and one forest site in Rondonia. The pasture sites were at the Fazenda Nossa Senhora and near the University of Rolim de Moura, the forest site was Reserva Jaru. Turbulence measurements of carbon dioxide were carried out. The measuring method, data processing procedure and the results of daily variations in concentration and fluxes will be presented. 1 VU, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands, Tel: +31204447362, Fax: +31206462457, 2 3 Email: [email protected]; ALTERRA, Email: [email protected]; VU, Email: 4 5 [email protected]; ALTERRA, Email: [email protected]; ALTERRA, Email: [email protected]; 6 UNIR, Email: [email protected]; 7 UNIR, Email: [email protected]; 8 9 CPTEC/INPE, Email: [email protected]; CPTEC/INPE, Email: [email protected]; 10 11 CPTEC/INPE, Email: [email protected]; ALTERRA, Email: [email protected]; 12 UNIR, Email: [email protected]. Carbon Storage and Exchange 80 The Spatial Distribution of Forest Biomass in Amazonia R. A. Houghton 1 The amount of carbon released to the atmosphere as a result of deforestation is determined, in part, by the amount of carbon held in the biomass of the forests converted to other uses. Uncertainty in forest biomass is responsible for much of the uncertainty in current estimates of the flux of carbon from land-use change. Here we compare several estimates of biomass for Brazilian Amazonia, based on spatial extrapolations of direct measurements, climatic variables, and remote sensing data, and ask three questions. First, do the methods yield similar estimates? Second, do they yield similar spatial distributions of biomass? And, third, what factors need most attention if we are to predict more accurately the distribution of biomass over large areas? The answer to the first two questions is that estimates of biomass (including dead and belowground biomass) vary widely, from a low of 39 PgC to a high of 93 PgC for Brazilian Amazonia. Furthermore, the estimates disagree as to the regions of high and low biomass. The lack of agreement among estimates, despite extensive calibrations and validations for each, confirms the need for reliable methods for determining aboveground biomass over large areas. Candidates include direct measurement of aboveground biomass from satellite (for example, using LIDAR) or dynamic modeling of observed stand-replacing disturbances. 1 WHRC, P.O. Box 296, Woods Hole, MA 02543 USA, Phone:508-540-9900, Fax: 508-540-9700, Email: [email protected]. Carbon Storage and Exchange 81 Impact of climate variability and human activity on terrestrial carbon dynamics in Amazonian ecosystems Hanqin Tian 1 Jerry M. Melillo 2 David W. Kicklighter 3 Berrien Moore III 4 Charles V. Vörösmarty 5 Previous analyses suggest that undisturbed Amazonian ecosystems, on average, acted as a sink of atmospheric CO2 as a result of climate variability and increasing CO2 (Tian et al. 1998; Prentice and Lloyd 1998). Other analyses show that Amazonian ecosystems release a large quantity of carbon to the atmosphere as a result of deforestation (Houghton et al. 2000). In this study, we have investigated how the combination of climate variability, increasing CO2 , deforestation and associated changes in cropland and pasture have affected terrestrial carbon dynamics in the Amazon Basin by using the Terrestrial Ecosystem Model (TEM). To address effects of land-use change on carbon storage, we have incorporated algorithms from MBL TCM (Houghton et al.1983; Melillo et al. 1988) into TEM. We ran TEM in transient mode using historical input data including: 1) historical mean atmospheric CO2 concentration; 2) gridded historical monthly data for air temperature and precipitation; and 3) gridded historical yearly data for cropland and pasture area based on remote sensing analyses and land use model. Our preliminary analyses have indicated that the carbon storage of Amazonian ecosystems during 1980-1995 remain unchanged as a result of the combined effects of CO2 , climate, deforestation and abandonment of agricultural lands. Our simulated results also suggest that the carbon sink induced by climate variability and increasing CO2 offsets the carbon source from land-use change in the Basin. The modeled net carbon storage in Amazonian ecosystems, however, varies more than 100% from year to year due to interannual variability in climate and deforested rates. 1 The Ecosystems Center, Marine Biological Laboratory, Woods Hole, MA02543, USA, Tel: (508) 2897498, Fax: (508) 457-1548, E-mail: [email protected] ,; 2 The Ecosystems Center, MBL Woods Hole,E-mail: [email protected] ; 3 The Ecosystems Center, MBL Woods Hole, E-mail: [email protected]; 4 Institute for the Study of Earth, Oceans and Space, UNH, E-mail: [email protected]; 5 Institute for the Study of Earth, Oceans and Space, UNH, E-mail: [email protected]. Carbon Storage and Exchange 82 Spatial and Historical Trends in the Amazon Carbon Sink Oliver Phillips 1 Yadvinder Malhi 2 Samuel da Almeida 3 Niro Higuchi 4 William Laurance 5 Percy Nuñez V. 6 Agustín Rudas 7 Natalino da Silva 8 Rodolfo Vásquez M. 9 By providing regularly recensused samples of all trees with significant biomass, permanent sample tree plots can yield accurate local measures of growth and death across time and space. In Amazonia, tree plots have recently been used to quantify changes in forest biomass, revealing that most old-growth Amazonian tree plots were accumulating biomass at the end of the twentieth century (Phillips et al.: Science 282:439-442, 1998). However the existing tree plot network is patchy both in space and time.Yet, if replicated and located across the environmental conditions of Amazonia, tree plots have the potential to reveal both the ecological processes involved in the carbon sink and the environmental factors controlling them. As part of the EU-funded CARBONSINK-LBA project, we are therefore collaborating to draw together local studies into a more systematic panAmazonian network of permanent tree plots. The project will concentrate on remeasuring existing plots, some of which may have been abandoned due to lack of funding or research interest. It will focus on the East-West LBA transect in Brazil and the natural extension of this transect to the aseasonal forests in western Amazonia (Colombia, Ecuador, Peru). Along the transect, we will work at sites where there is a critical mass of sample plots, providing local replicates across varying edaphic conditions. This will provide insight into the process and persistence of carbon sequestration, and a unique spatial and temporal data set with which to test and calibrate basin-wide carbon balance models. 1 University of Leeds, School of Geography, Tel: 44-1132-2336832; Fax: 44-1132-333308, E-mail: [email protected]; 2 University of Edinburgh, E-mail: [email protected]; 3 Museu Goeldi, 5 E-mail: [email protected]; 4 INPA, E-mail: [email protected]; INPA, E-mail: 6 7 [email protected] ; UNSAAC, E-mail: [email protected]; Instituto de Ciencias Naturales, E-mail: [email protected] ; 8 EMBRAPA, E-mail: [email protected]; 9 Proyecto Flora del Peru, Missouri Botanical Garden, E-mail: [email protected]. Carbon Storage and Exchange 83 Heterogeneity of soils and vegetation in the Tapajós National Forest: Implications for scaling up Mathew Williams 1 Y. E. Shimabokuro 2 D. A. Herbert 3 S Pardi Lacruz 4 C. D Renno 5 E. B. Rastetter 6 A critical goal of the LBA project is to understand the Amazon basin as a regional entity. This requires that detailed, site-specific investigations of ecosystem function are extended across the entire region. The extension process must take account of the regional variability in the key controls on ecosystem dynamics. For example, regional gross primary production will be constrained by spatial and temporal variation in both vegetation and soil characteristics. To determine the heterogeneity of these constraints, we undertook detailed surveys of several ecosystem parameters within the Tapajós National Forest on thirteen 0.25 ha plots, during November 1999. Our results showed that leaf area index was more variable within sites (0.0-8.9) than among sites (means varied from 5.2-7.0). In the two sites with the lowest gravimetric soil water content and very sandy soils (>80%), averaged –1 MPa. Apart from this link, there were no other clear correlations between soil parameters and vegetation characteristics. The variation in soil texture among sites was notable; at four sites clay content >80%, while at five sites the sand content > 75%. In summary, our surveys of vegetation structure suggest considerable heterogeneity at fine spatial scales (50 m or less). While soil characteristics were consistent within sites, the extreme variability in texture among sites has important implications for drought stress, local hydrology, soil biogeochemistry, and response to disturbance. 1 7 MBL, Tel: +1 508 289 7489; Fax: +1 508 457 1548, E-mail: [email protected]; 2 INPE, E-mail: [email protected] ; 3 MBL, E-mail: [email protected] ; 4 INPE, E-mail: [email protected]; 5 INPE, Email: [email protected]; 6 MBL, E-mail: [email protected] . Carbon Storage and Exchange 84 Estoque e Fluxos de Carbono e Nutrientes nas Raízes de Pastagens Abandonadas e Sistemas Agroflorestais na Amazônia Central Jorge Luis Enrique Gallardo-Ordinola 1 Flávio Luizão 2 Erick C. M. Fernandes 3 O trabalho será desenvolvido na Estação Experimental do CPAA/EMBRAPA, 60 km ao norte de Manaus. Pretende-se, comparar a produção de biomassa e a distribuição das raízes em sistemas agroflorestais com as da capoeira natural nas diferentes camadas do solo até 1,5 m de profundidade para avaliar seu papel no seqüestro de carbono, na estocagem e fluxo de nutrientes, e na conseqüente recuperação de pastagens degradadas. Os tratamentos incluem um sistema agroflorestal com cupuaçu e pupunha e um Sistema Agrossilvopastoril: com “altos insumos”, com ingá , mogno e paricá, e um controle de capoeira (CAP), correspondente à vegetação secundária que cresceu espontaneamente sobre o solo da pastagem após a queima e abandono. Para a amostragem de raízes, serão combinadas duas técnicas de amostragem: a técnica do trado para raízes finas e das trincheiras para raízes grossas. As raízes nos é 150 cm. As raízes maiores serão separadas em raízes de 2-5 mm e de 5 mm, por espécie de plantas. Nas raízes separadas do solo e secas a 65-70 ºC, serão analisados os teores de carbono e de macronutrientes, que serão também usados para calcular as relações C:N, C:P e N:P; serão também analisados os conteúdos de polifenóis e de lignina das raízes. Taxa de renovação de raizes será avaliado por minirhizotron e trado. 1 Instituto Nacional de Pesquisas da Amazônia, Rua Nelson Batista Sales 114, Cj Petro - Aleixo, Manaus, AM, Tel: 92 643 1911, Fax: 92 642 1838, E-mail: [email protected]; 2 INPA, E-mail: fluizã[email protected] ; 3 Cornell University, E-mail: [email protected]. Carbon Storage and Exchange 85 Biomass combustion experiments conducted in Fazenda Caiabi, town of Alta Floresta, MT, Brazil João Andrade de Carvalho Jr. 1 Fernando de Souza Costa 2 Ralf Gielow 3 David V. Sandberg 4 Ernesto Alvarado 5 Carlos Alberto Gurgel Veras 6 Aguinaldo Martins Serra Jr. 7 José Carlos dos Santos 8 Uncertainties in the estimates of emission rates from land use are very high, mainly due to a) measurement uncertainties in deforestation rates and clearing areas by fire, (b) ifficulties in determining biomass stock and (c) destination of deforested land. A large uncertainty to assess the impact of biomass burning in the atmosphere chemistry and in the global climate is the amount of biomass burning from each eco-system. These difficulties and uncertainties are more significant in Brazil, the country with the arger carbon stock in the vegetation in the world (five times larger than Zaire’s stock, with the second larger stock in the world). CO2 emissions from deforestation and biomass combustion are of great concern not only in Brazil, but also globally. This paper describes the work to estimate the completeness of biomass burning and the amount of carbon released to the atmosphere due the process of forest clearing in five test areas. The experiments were carried out at Fazenda Caiabi, in the town of Alta Floresta, state of Mato Grosso, Brazil, which is located in the Amazonian deforestation arc. The tests were performed in areas of 1 ha each, denominated A, B, C, D, and E. The following are the main characteristics of each area: A. located in the frontier with a pasture, with three sides directly on the forest, cut and burned in 1997; B. located approximately 400 m inside the forest, with four sides directly on the forest, cut and burned in 1997; C. located inside a deforested 9-hectare area, all cut and burned in 1998; D. located inside a deforested 4-hectare area, all cut in 1998 and burned in 1999; and E. located inside a deforested 4-hectare area, all cut and burned in 1999. 1 Av. Ariberto Pereira da Cunha, 333, Guaratinguetá, São Paulo, Tel: 12-5252800/r.149; Fax: 12-5252466 3 E-mail: [email protected] ; 2 INPE, E-mail: [email protected]; INPE, E-mail: 4 5 [email protected]; USDA, E-mail: [email protected]; University of Washington, E-mail: [email protected]; 6 UnB, E-mail: [email protected]; 7 INPE, E-mail: [email protected]; 8 INPE, E-mail: [email protected] . Carbon Storage and Exchange 86 One year of water and carbon fluxes from central Amazonia (ZF2), 1995-1996 Pegoraro Emiliano 1 Yadvinder Malhi 2 Antonio D. Nobre 3 A. Marques Filho 4 Maria G.P. Perreira 5 We report data on water and energy fluxes collected over one year at a dense terra firme forest in central Amazonia in 1995-1996. CO2 fluxes have been reported in a previous publication. In the dry season the mean net radiation peaks at 550 Wm-2, as opposed to 450Wm-2 in the wet season. Changes in long-wave radiation balance are more important than changes in forest albedo. The direct measurements of latent and sensible heat fluxes seemed too small to close the surface energy budget. This is not likely to be due to nonmeasurement of within-canopy heat storage. We corrected the data using the assumption of energy budget closure. The canopy conductance ranged between 10 mol m-2 s-1 in the wet season to 6 mol m-2 s-1 in the dry season. The dry season reduction was strongly correlated to reductions in soil water content. Although the Bowen ratio of the forest canopy increases significantly in the dry season, overall transpiration rates remained approximately constant because of the increased solar radiation. A comparison of the results with basin water balance studies suggested that the errors were likely to be purely in measurements of water fluxes, rather than in sensible heat fluxes. 1 Darwin Building, King's Buildings, Maifield Road, EH9 3JU, Edinburgh (Scotland, UK), Tel: +44-01316595744, Fax: +44-0131-6620478, E-mail: [email protected] ; 2 Edinburgh University, E-mail: [email protected]; 3 INPA, E-mail: [email protected]; 4 INPA; 5 INPA. Carbon Storage and Exchange 87 Estimating the Exchange of CO2 and its Stable Isotopes Between Amazonian Ecosystems and the Atmosphere at Multiple Spatial Scales Scott Denning 1 Lara Prihodko 2 Neil Suits 3 Marek Uliasz 4 Melville Nicholls 5 Elicia Inazawa 6 Niall Hanan 7 Pier-Luigi Vidale 8 We have developed a series of methods for estimating temporal and spatial variations in the exchange of CO2 and its stable isotopes. A numerical model of biophysical, hydrological, and biogeochemical processes (SiB2) has been adapted for Amazonian forests and grasslands. The model predicts exchanges of water, energy, carbon, and momentum on time scales of several minutes from environmental conditions, and also predicts the fractionation and changes in storage of 13 C and 18O in CO2 . The model was evaluated by comparison with flux data collected during ABRACOS, and has been coupled to a mesoscale model (CSU RAMS). Spatial and temporal variations in vegetation parameters are estimated using AVHRR imagery and soils are parameterized from a spatial database (IGBP-DIS). The model is reasonably successful in capturing the major features of the limited data available to date. The coupled model is used to estimate variations in CO2 and its stable isotopic composition in the atmosphere over several scales. These include diurnal cycles in the canopy air; regional patterns simulated in RAMS; and basin-wide variations simulated by coupling to a global general circulation model. Using a Bayesian synthesis inversion technique, we test a method for direct estimation of regional fluxes from hypothetical aircraft data. Results suggest that atmospheric properties can form a significant integral constraint to regional carbon fluxes. A campaign of aircraft sampling to implement this method is planned for 2001. 1 Department of Atmospheric Science, Tel: +1-970-491-6936; Fax: +1-970-491-8449, E-mail: [email protected] ; 2 Colorado State University, E-mail: [email protected] ; 3 Colorado 4 State University, E-mail: [email protected]; Colorado State University, E-mail: 5 [email protected] ; Colorado State University, E-mail: [email protected] ; 6 Colorado State University, E-mail: [email protected]; 7 Colorado State University, E-mail: [email protected]; 8 Colorado State University,E-mail: [email protected] . Carbon Storage and Exchange 88 Radon-222 Determination of Forest Canopy-Atmosphere Gas Exchange Rates: New Methods and Initial Results Christopher S. Martens 1 Howard Mendlovitz 2 Osvaldo L.L. Moraes 3 Patrick M. Crill 4 The primary goal of our LBA project is to quantify air exchange rates and gas fluxes between old growth and selectively logged forests and the overlying atmosphere. We use continuous, in situ radon-222 measurements and compare calculated vertical gas fluxes with values derived from independent eddy correlation and flux chamber measurements. Arrays of specially designed flow-through radon detectors are being deployed as of late April 2000 at both the primary forest (km 67) and selective logging (km 83) sites in the Flona Tapajos, following exhaustive laboratory and field calibrations and experiments in an experimental forest in North Carolina,. Our field tests in North Carolina have provided an initial data set that illustrates the capabilities of the new detectors to accurately resolve 0.01 pCi/l/m radon activity gradients within the forest canopy using15 minute counting intervals. A continuous six month data set from the North Carolina reveals both diel and seasonal changes in canopy inventories. Soil radon fluxes are measured using an adaption of flow-through detector designs to produce a fluxometer capable of repeated flux determinations taking approximately one hour each using existing flux chamber collars located around the tower sites. 1 CB-3300, Department of Marine Sciences, Chapel Hill, NC 27599-3300 USA, Tel: 919 962 0152; Fax: 919 962 1254, E-mail: [email protected] ; 2 UNC, E-mail: [email protected]; 3 UFMS, E-mail: [email protected]; 4 UNH, E-mail: [email protected] . Carbon Storage and Exchange 89 Variabilidade espacial do carbono do solo em área com pastagem na Amazônia Ocidental (RONDÔNIA) Carlos Eduardo Pellegrino Cerri 1 Carlos Clemente Cerri 2 Reynaldo L. Victoria 3 Martial Bernoux 4 Vincent Chaplot 5 Boris Volkoff 6 Maria Victoria Ballester 7 Marisa C. Piccolo 8 Brigitte J. Feigl 9 Jerry M. Melillo 10 Christopher Neill 11 Paul A. Steudler 12 Diana Garcia 13 O objetivo da pesquisa é avaliar a variabilidade espacial dos estoques de carbono do solo em uma área de 63 ha com pastagem na Amazônia Ocidental, Rondônia (Fazenda Nova Vida, 10o 10'05"S e 64o 49'27"W) e correlacionar os estoques de C do solo, com os fluxos de CO2 e carbono imobilizado na biomassa microbiana. Efetuou-se uma amostragem sistemática coletando-se amostras de solo de acordo com uma malha regular de 25 x 25 metros. Em cada um dos 985 pontos referenciados geograficamente coletou-se amostras nas camadas 0-10, 10-20 e 20-30 cm de profundidade, totalizando 2955 amostras. Foram realizadas tradagens espaçadas a cada 100m, nas profundidades 50-60, 70-80 e 90-100cm, totalizando 198 amostras. No decorrer da amostragem sistemática verificou-se que o terreno apresenta heterogeneidade entre as distâncias de 25 m, justificando uma amostragem mais adensada. Adotou-se então um sistema de coleta idêntico ao utilizado na amostragem sistemática, entretanto, utilizando grades regulares de 5 x 5 m e de 1 x 1 m. Foram amostradas as camadas 0-10, 10-20 e 20-30cm em quatro sub-áreas com 68 pontos cada, totalizando 816 amostras. Durante a amostragem verificou-se que existem cerca de cinco variações de solos na área de estudo. Com o intuito de conhecer e caracterizar tais solos, foram abertas 7 trincheiras com 2m de profundidade. Em cada uma das trincheiras foram descritos morfologicamente os horizontes diagnósticos e coletadas amostras para serem análisadas em laboratório.Esse levantamento servirá como base para a instalação de um experimento de intensificação de manejo de pastagens de baixa fertilidade. Atualmente as amostras estão sendo processadas. Assim que os resultados foram sendo disponibilizados serão incorporados ao banco de dados do LBA. 1 Avenida Centenário, 303 - Cx. Postal 96 Piracicaba, São Paulo, Tel: (19) 429-4708; Fax: (19) 429-4610, E-mail: [email protected] ; 2 CENA/USP, E-mail: [email protected]; 3 CENA/USP, E-mail: [email protected]; 4 IRD/CENA; 5 IRD/CENA; 6 IRD/CENA; 7 CENA/USP; 8 CENA/USP; 9 CENA/USP; 10 Marine Biological Laboratory (MBL) - The Ecossystem Center; 11 MBL - The Ecossystem Center; 12 MBL - The Ecossystem Center; 13 MBL - The Ecossystem Center. Carbon Storage and Exchange 90 Daily Course of Carbon Release from Stems and Branches J. Viviana Horna 1 Reiner Zimmermann 2 John Tenhunen 3 Inundated forests of Central Amazon are believed to have a high carbon release since net wood productivity is low in spite of favorable nutrient supply and climate. Diurnal courses of carbon release were measured in the main stem and branches of deciduous trees (Albizia multiflora, Tabebuia barbata, Pseudobombax munguba, Crataeva benthami, Vitex cymosa) and evergreen tree species (Nectandra amazonum, Laetia corymbulosa, Pouteria glomerata). Measurements were taken between March 1999 and March 2000. Stem and branch cuvettes were used in an open system with an integrated infrared gas analyzer running in differential mode. From March until September the area adjacent to the Amazon River is flooded by several meters. Rates for woody tissue respiration where higher during the beginning of the flooding season (March-April) and varied from 2 to 14 mmol CO2 m-2 s-1 among species. Carbon release was highest in the main stem of the evergreen Nectandra amazonum. During the dry season, from October to February, the values of respiration varied from 1 to 8 mmol CO2 m-2 s-1 among species. During this time carbon release was higher in the stem of the deciduous Albizia multiflora. Carbon release measurements for the complete annual cycle still need to be separated into maintenance and growth respiration that follows closely the phenology of the species. The final step of this study will be to develop a model to estimate carbon loss by respiration of the species studied. 1 University of Bayreuth, Plant Ecology. D-95440 Bayreuth, Germany, Tel: 0049-921-552573; Fax: 0049921-552564, E-mail: [email protected]; 2 Forest Ecology and Remote Sensing Group, University of Bayreuth, E-mail: [email protected]; 3 Plant Ecology, University of Bayreuth, E-mail: [email protected] . Carbon Storage and Exchange 91 The Ecosystem Demography Model (ED) Paul Moorcroft 1 George C. Hurtt 2 Steve Pacala 3 Elena Shevliakova 4 Berrien Moore 5 Natural disturbances and human land-use activities have a large impact on the structure, dynamics and fluxes of ecosystems at a variety of scales. As a result, ecosystems are more heterogeneous in important ways than is explained by climatic and edaphic factors alone. A critical impediment to incorporating relevant finescale detail in large-scale models is the problem of scaling-up. We developed a new individual-based terrestrial ecosystem model, the Ecosystem Demography Model (ED), that addresses these issues. ED predicts both ecosystem structure (e.g. above and below-ground biomass, vegetation height and basal area, and soil carbon stocks) and corresponding ecosystem fluxes (e.g. NPP, NEP and evapotranspiration) from climate, soil and land-use inputs. The model consists of integrated sub-models governing processes such as leaf-level physiology, plant allocation, allometry, phenology, dispersal, the effects of fire disturbances, and belowground sub-models for soil carbon dynamics and hydrology. Using a new method for scaling-up it is possible to predict ED's large-scale behavior without simulating the fate of every plant individually. ED is being used to examine how climate and edaphic factors, natural disturbances, and human land-use practices affect vegetation structure and fluxes within the Amazon Basin. Coupled versions are in development to consider potential feedbacks to the atmosphere. 1 Princeton University, Dept. Ecology and Evolutionary Biology, Princeton, NJ 08544-1003 USA Tel: 609-258-6886; Fax: 609-258-1334, E-mail: [email protected]; 2 University of New Hampshire, E-mail: [email protected]; 3 Princeton University, E-mail: [email protected]; 4 Princeton University, E-mail: [email protected] ; 5 University of New Hampshire, E-mail: [email protected] . Carbon Storage and Exchange 92 Non-CO2 volatile carbon emissions and net ecosystem exchange in Amazônia Jim Greenberg 1 Alex Guenther 2 Oscar Veja 3 Peter Harley 4 Luciana Vanni-Gatti 5 At present, the global CO2 budget requires the addition of a large, unidentified terrestrial sink in order to balance. After peatlands, forests contain the largest terrestrial reservoirs of carbon. Landuse changes in forested areas, therefore, represent a significant potential impact on the global carbon cycle. The net exchange of CO2 between ecosystems and the atmosphere is a measure of the net ecosystem exchange of carbon (NEE). The net flux of CO2 should be into young, rapidly growing forests; out of areas of recently cleared forest areas (as the organic matter decomposes); but near zero for mature or primary forests. The estimation of the net carbon exchange is complicated by the potential for missing a significant exchange of non-CO2 carbon. Living and dead organic matter emit a variety of volatile organic compounds (VOCs), including carbon monoxide, methane, isoprene, monoterpenes, methanol, acetone, acetaldehyde, formaldehyde, many VOCs related to damage of plant tissue, etc. We have measured several of these non-CO2 VOC emissions at the landscape level in various states in Amazônia, including Amazônas (near Balbina, primary and secondary forest), Rondônia (recently cleared forest and pasture as well as primary forest) and Pará (near Santarém, primary and secondary forest). The atmospheric concentration profiles of these VOC emissions were used to estimate hourly and daily emissions; a biogenic emission model was used to extrapolate these emissions to na annual flux. This annual emission estimate was then used to gauge the importance of non-CO2 VOC emissions to NEE in Amazônia. 1 P.O.Box 3000, Boulder, Colorado, USA 80307- 3000, Tel: 303-497-1454; Fax: 303-497-1477, E-mail: [email protected] ; 2 NCAR; 3 IPEN; 4 NCAR; 5 IPEN. Carbon Storage and Exchange 93 Decrease in CO2 emissions related to increase in the iron level of Brazilian soils Newton La Scala Júnior 1 José Marques Júnior 2 Caio Vinícius dos Santos 3 Gener Tadeu Pereira 4 Brazilian soils may be characterised by color. A great part of Brazilian soils have a dark red color, due to the iron level. The range if iron content is from 4% to as high as 33%, and the coloration is related directly to the iron content. In this work we have studied the CO2 emissions of a bare soil having a great contrast in the iron level. In our experimental field we have installed a transect, where CO2 emission was measured using an IRGA-based soil chamber, at points separated 10 meters from each other, ranging from a yellow to a red dark soil. We have also collected a 20 cm depth soil layer in all the measurement points and analyzed pH, organic carbon, Fed level among other soil properties. Our results are presented in terms of linear correlation and spatial variability analyzes of CO2 emissions and the soil properties studied. In particular, we have observed a significant negative linear correlation between soil respiration and iron level of the soil (r = -0.31, P < 0.05). Finally, the similarity of the spatial variability models of the CO2 emission, when compared to the iron level also supports the relationship between these two properties. 1 Rod. Paulo Donato Castellane s/n, 14870-000, Jaboticabal-SP, Tel: 16 + 323-2500 ext. 112; Fax: 16 + 3224275, E-mail: [email protected] , 2 FCAV/UNESP; 3 FCAV/UNESP; 4 FCAV/UNESP. Carbon Storage and Exchange 94 Photosynthetic response of canopy species in a forest area in eastern Brazilian Amazonia Moacyr B. Dias-Filho 1 José Benito Guerrero M. 2 Daniel C. Nepstad 3 The functional changes to experimental rainfall exclusion of a one-hectare area of primary forest in Santarém, State of Pará, Brazil, will be studied during a three-year period. The objective of the present study was to gather baseline data on the photosynthetic response of canopy (> 10m) species in the area, prior to implementation of the rainfall exclusion treatment. Measurements on individual species were made during the rainy and dry seasons of 1999, between 8:30h and 15:30h, local time, with an open portable photosynthesis system (Li-Cor LI-6400) under standard conditions. Wood towers provided access to the high forest canopy. Mean PPFD during gas exchange measurements were 212.1 ± 323.2 (mean ± standard deviation) and 298.1 ± 455.7 µmol m-2 s-1 for the rainy and dry seasons. Mean photosynthetic rates for the rainy and dry seasons were 5.37 ± 2.73 and 7.62 ± 3.96 µmol m-2 s-1, and maximum photosynthetic rates were 21.1 and 22.7 µmol m-2 s-1. 1 Lab. Ecofisiologia Vegtal, C. Postal 48, Belém PA, 66017-970, E-mail: [email protected]; 2 IPAM; 3 The Woods Hole Research Center. Carbon Storage and Exchange 95 Characterization of the Esecaflor/Caxiuanã Experimental Sites, PA Maria de Lourdes Pinheiro Ruivo 1 Antônio Fernado S. Marques 2 Everton da S. Cunha 3 Beatriz Quanz 4 Among several researches developed by the LBA project at Estação Científica Ferreira Penna (Caxiuanã, PA), the ESECAFLOR project is investigating the influence of the water exclusion from the soil on the forest cycle. This experiment is in its initial phase of collecting basic data related to soil, vegetation and climate. We will present the preliminary results on soil characteristics in tree plots (sites 1, 2 and 3). Samples were taken from 5m soil pits. The soils, classified as yellow Latosol, present A, B and C horizonts, are moderate to well drained, have pH values ranging from overly acid (3,5) to moderately acid (5,5), colors are dark yellow brown (10YR, 3/6) to yellowish red (7,5 YR, 6/6). Texture varies from sandy to clay, organic carbon (OC) varies from high (109 g/kg) to low (1.9 g/kg) and there is a high proportion of fulvic acid in relation to humic acids. Site 3 shows higher accumulation of OC and humic substances in relation to the others. This fact may be related to variation in drainage, topography or vegetation cover of the sites since the material of origin is the clay-sand sedimentary rocks of the Alter do Chão Formation. 1 Museu Paraense Emílio Goeldi , Av. Magalhães Barata, 376; Belém, Pará, Brasil, Tel: 091.249-0793; Fax: 091. 249-0466, E-mail: [email protected]; 2 Museu Paraense Emílio Goeldi; 3 Faculdade de Ciências Agrárias do Pará; 4 Museu Paraense Emílio Goeldi . Carbon Storage and Exchange 96 Carbon Budget in different components of a woody savanna in Brazilian Cerrado Roberto Engel Aduan Carlos Augusto Klink Eric A. Davidson 1 2 3 The Cerrado savanna covers 2,000,000 km2 in Central Brazil. The human occupation has been accelerating over the past few decades, and currently around 45% of its natural area have already been converted into agro-ecosystems, mainly planted pastures and cash crops. Despite the magnitude of human occupation, the effects of land use change on carbon balance of the converted areas, are still poorly understood. The aim of this work is to evaluate the carbon budget of a "Cerrado denso" (savanna dominated by trees and shrubs) area in the Reserva Ecologica do Roncador (RECOR/IBGE), near Brasília, by monitoring key processes related to carbon dynamics in the ecosystem: soil respiration (using the dynamic chamber IRGA technique), litterfall (using litter bags) and plant growth (dendrometry). The spatial and temporal variation of this processes will be correlated to the variation of key environmental factors (precipitation, humidity, soil water content and temperature), and which may be expected to vary with changes in land use and climate. Preliminary results of soil respiration show that CO2 fluxes from soil surface, in the rain season, are similar to other tropical savannas (0,11-0,22gC m-2 h-1) and to eastern Amazon primary forests in the dry season. 1 Campus Universitário - ICC - Ala Sul Departamento de Ecologia, Tel: 61 2721132, E-mail: [email protected] ; 2 Universidade de Brasília; 3 Woods Hole Research Center. Carbon Storage and Exchange 97 Carbon stocks and dynamics in a neotropical rain forest Deborah A. Clark 1 Steven F. Oberbauer 2 David B. Clark 3 At the La Selva Biological Station (Costa Rica), we are carrying out a long-term, landscape-scale assessment of carbon stocks and the climatic controls of carbon fluxes in a lowland neotropical rain forest (the "CARBONO" project). Since January 1998, we have continuously measured forest-level CO2 exchange with the atmosphere from a 42 m tower with eddy covariance techniques. For complementary ground-based studies, we stratified 18 0.5 ha plots over 600 ha of old-growth forest (six plots each on: steep ultisol slopes; ultisol plateau; and old alluvial terraces). Measurements began in the 18 plots in September 1997 for: SOC stocks (to 4 m) and coarse woody debris (CWD) stocks (at project initiation); fine roots (every 6-8 weeks); fine litterfall (biweekly); fine woody litterfall (monthly); CWD inputs (annual); tree growth and survival and aboveground biomass increment (annua ix 4-m deep soil pits. Soil respiration is measured at the surface biweekly in 6 plots, and bole and soil respiration are measured continuously with automated chambers at the tower site. Short-term (day-month) climatic responses of 10 canopy tree species are being quantified with xylem sapflow and dendrometer studies. Additional pilot studies and currently proposed work focus on forest- and species-level isoprene emissions, dry deposition of carbon, precipitation throughfall, litter chemistry, CWD decomposition, and microbial activity. The results from these studies to date reveal substantial temporal and spatial variation in the carbon dynamics of this forest. 1 Interlink-341, P.O. Box 02-5635, Miami FL 33102, USA, Tel: 506-766-6565x146; Fax: 506-766-6535, E-mail: [email protected] ; 2 Florida International University, E-mail: [email protected] ; 3 University of Missouri-St. Louis, E-mail: [email protected] . Carbon Storage and Exchange 98 Torres metálicas para instalação de sistemas de medição de fluxos sobre florestas George Sanches Suli 1 Nicolau Priante Filho 2 A tomada de dados por aparelhos de precisão instalados nas torres metálicas em regiões de estudo requer a adaptação da estrutura dessas torres de modo a proporcionar funcionalidade e segurança na instalação de aparelhos, algumas vezes tão caros quanto a própria estrutura. O fato do serviço de manutenção e coleta de dados ser executado por pessoas não qualificadas para o manejo de peças e deslocamento em grandes alturas, como bolsistas e professores, torna desejável que as torres apresentem certa ergonomia, como escadas inclinadas e patamares de descanso, e ofereçam sensação de segurança a quem deva executar qualquer operação em qualquer nível. Além disso, deve apresentar solução estrutural para possibilitar sua construção em regiões de difícil acesso, sem a derrubada de árvores e seja resistente às forças da natureza, como esforços devidos ao vento, corrosão e eventual que exigências mínimas quanto a peso e volume de fundações, baixo impacto ambiental, apresentando boa resistência à queda de grandes árvores, a idealização de um modelo otimizado precisa ser efetuada. Serão apresentados projetos de torres otimizados quanto à adaptabilidade de suportes à estrutura da torre e materiais empregados, como aço (A-36) galvanizado ou pintado e alumínio (6061-T6) para torres de até 120 m, visando a maior relação custo/benefício, com a adoção de um modelo geométrico único o que se verifica interessante no caso de previsão de verbas para uma estrutura de forma final definida e coleta de orçamentos sobre um projeto estabelecido. 1 Universidade Federal de Mato Grosso, Rua França n. 12, Q. 1, J. Europa, 78065-440 Cuiabá, Mato Grosso, Tel: (0xx65)6342050, E-mail: [email protected]; 2 UFMT, E-mail: [email protected]. Carbon Storage and Exchange 99 The net CO2 exchange of a transitional tropical forest (cerradão) during portions of the wet and dry season George L. Vourlitis 1 Nicolau Priante Filho 2 Mauro M. S. Hayashi 3 José de S.Nogueira 4 José Holanda Campelo Jr 5 Eddy covariance and measurements of the vertical CO2 gradient within the canopy were used to quantify the net CO2 exchange of a 30m tall transitional tropical forest (cerradão) over portions of the wet and dry season. The study was conducted near the city of Sinop in NW Mato Grosso, Brazil (11°24.75'S: 55°19.50'W), which is located near the ecotone of two major regional ecosystem types of South America (wet evergreen rain forest and cerrado). The eddy covariance system consisted of a 3-dimensional sonic anemometerthermometer for measuring the mean and fluctuating quantities of wind speed and temperature and an open-path infrared gas analyzer for measuring the mean and fluctuating quantities of CO2 and H2 O vapor. The eddy covariance sensors were mounted at a height of 42 m above ground level, which corresponds to a height of ca. 12m above the canopy. The vertical CO2 concentration was measured using a closed-path CO2 analyzer. Canopy CO2 storage was quantified by determining the rate of change of CO2 in the air column between the ground surface and the eddy covariance sensors. This term was important because of the tall and dense forest canopy and the relatively low nocturnal wind speed typical of the Sinop area. Preliminary data suggest that the cerradão stand was approximately in balance over the 97 d measurement period (average ±1SD daily net CO2 flux = 0.6 ± 1.3 gC m-2 d1; positive values denote net CO2 loss from the ecosystem). Net CO2 accumulation increased during the beginning of the wet-season measurement period (average ±1SD daily net CO2 flux = -1.7 ± 0.4 gC m-2 d-1), suggesting that seasonal variations in rainfall have important implications on the seasonal pattern of net CO2 exchange. 1 Biological Sciences Program, California State University, Tel: 760-750-4119, Fax: 760-750-4111, E-mail: 2 3 [email protected]; UFMT, E-mail: [email protected] ; UFMT, E-mail: 4 5 [email protected]; UFMT, E-mail:j [email protected] ; UFMT, E-mail: [email protected]. Carbon Storage and Exchange 100 Biomass of moist tropical forests decreases with increasing rainfall Jon Lloyd 1 A global analysis has been undertaken of the relationship between above-ground biomass and rainfall for tropical forests and woodlands. Once effects of disturbance and soil type are taken into account, a clear non-linear relationship emerges. Tropical forest biomass increases linearly up to about 2000 mm rainfall but declines significantly thereafter. Possible causes for this decline in biomass at high rainfall will be discussed. 1 Max-PLanck-Institut für Biogeochemie; POstfach 100164, 07743 Jena Germany, Tel: 49-3641-643711, Fax: 49-3641-643710, E-mail: [email protected]. Carbon Storage and Exchange 101 Soil organic carbon pool, soil CO2 efflux and CO2 profile concentration under old growth neotropical rainforest Luitgard Schwendenmann 1 Edzo Veldkamp 2 Tania Brenes 3 Jens Mackensen 4 Deborah A. Clark 5 David B. Clark 6 Steven F. Oberbauer 7 Estimates of carbon pools and fluxes in tropical forests are important in the current discussion of the world's carbon balance of forests. The main objectives of this study are (i) to present results on the soil carbon stocks, (ii) CO2 efflux and (iii) soil CO2 profile concentration under undisturbed tropical rainforest in the Atlantic zone of Costa Rica. In three Inceptisol plots (old alluvial terraces) and three Ultisol plots (highly weathered residual soils) 4 m deep shafts were installed. Soil samples were taken by depth for bulk density, soil organic cabon (SOC) and for 13 C and 14 C analysis. Since March 1998 we measure CO2 efflux bi-weekly from 8 permanent soil chambers per plot using an infrared CO2 analyzer. At the same time we measure soil CO2 concentrations at 5, 20, 40, 75, 150, 250 and 350 cm depth using a GC. In addition controlling environmental factors like soil temperature and soil moisture are determined. Soil organic carbon stock to 3 m depth for inceptisol ranged from 242 to 306 Mg C/ha. A significantly higher amount of carbon (315 to 394 Mg C/ha) is stored in the Ultisol sites. At the Inceptisol sites, soil CO2 efflux rates ranged from 10.4 to 11.7 Mg CO2 -C ha/yr. Significantly higher flux rates occurred in the less fertile Ultisols where efflux rates ranged between 12.4 to 15.8 Mg CO2 -C ha/yr. Higher soil CO2 concentrations in the upper soil profile under moist conditions (up to 2 % in Inceptisols) supports the idea that higher soil moisture contents partially inhibit CO2 diffusion to the soil surface. 1 Institute of Soil Science and Forest Nutrition, University of Goettingen, Germany La Selva Biological Station, Costa Rica, INTERLINK-341, P.O. Box 02-5635, Miami FL 33102, USA Tel: ++506-766-6565, Fax: ++506-766-6535, E-mail: [email protected] ; 2 University of Goettingen, E-mail: [email protected] ; 3 La Selva Biological Station, Costa Rica; 4 University of Goettingen; 5 University of Missouri-St.Louis; 6 University of Missouri-St.Louis; 7 Florida International University. Carbon Storage and Exchange 102 Biomass in the Tapajos National Forest, Brazil: Examination of Uncertainties arising from Sampling and Allometric Relationships Michael Keller 1 Michael Palace 2 George Hurtt 3 Changes in the biomass of Amazon region forests represent an important component of the global carbon cycle but the biomass of these forests remains poorly quantified. We examined forest survey data for trees with a diameter at breast height (DBH) greater than 35 cm from 392 ha in the Tapajos National Forest near Santarem, Para, Brazil. Based on tree diameter data, allometric relations, and published relations for biomass in other compartments besides trees, we estimated a total biomass density of 372 ± 59 Mg ha-1 (mean ± standard error). This estimate includes all live and dead plant material above and below ground with the exception of finely divided soil organic matter. We propagated errors in sampling and those associated with allometric relations and other ratios used to estimate biomass of roots, lianas and epiphytes, and necromass. The major sources of error in our estimate were found in the allometric relations for trees with DBH greater than 35 cm, in the estimates of numbers of trees with DBH less than 35 cm, and in root biomass. Simulated sampling based on our full survey, suggests that we could have estimated mean tree biomass (DBH > = 35 cm) to within 20% with 95% confidence by sampling 23 randomly selected 0.25 ha plots in our study area. 1 International Institute of Tropical Forestry, USDA; Forest Service, Rio Piedras, PR 00928-5000 USA and Complex Systems Research Center, Morse Hall, University of New Hampshire, Durham, NH 03824 USA E-mail: [email protected]; 2 UNH, Complex System Research Center, E-mail: [email protected]; 3 UNH, Complex Systems Research Center, E-mail: [email protected] . Carbon Storage and Exchange 103 How sensitive are calculations of CO2 and water? Yadvinder Malhi 1 Robert Clement 2 Antonio Nobre 3 Flux data are being collected at a number of sites within LBA. There are a number of possible methods for processing and analysing such data, including different procedures for detrending time series, rotating data, averaging periods and high and low frequency corrections. A number of sites report some problems with the resulting flux measurements, such as nonclosure of energy balance or low night-time CO2 fluxes. Here we examine raw data for heat, water and CO2 fluxes collected from ZF2 (near Manaus) in 1995, to examine the nature and spectral structure of turbulent flux transport, and to estimate the sensitivity of the resulting calculated fluxes to changes in analysis procedure. We conclude that a significant proportion of flux transport in tropical regions is at very low frequencies. 1 IERM, University of Edinburgh; Mayfield Road, Edinburgh EH9 3JU, Tel: +44 131 650 5744; Fax: +44 131 662 0478, E-mail: [email protected]; 2 IERM, E-mail: [email protected]; 3 INPA, E-mail: [email protected]. Carbon Storage and Exchange 104 CO2 , water and energy fluxes at Caxiuana National Forest, Para: the first year of results Fiona Carswell 1 Marcia Palheta 2 Antonio Lola da Costa 3 Yadvinder Malhi 4 Jose Maria da Costa 5 Maria das Gracas Pires Pereira 6 John Grace 7 Eddy covariance measurements have been carried out at a 56 m tower at Caxiuana National Forest, Para, since April 1999, as part of the LBA-EUSTACH and ECOBIOMA projects. We report here on the first year of data from this site. We examine the results and compare them with previous studies in Rondonia and Manaus The uptake of CO2 in the day was up to 20 or 30 µmol m-2 s-1, consistent with what has been observed in Rondonia and Manaus. This similarity suggests that bulk photosynthesis and respiration rates are very similar across the terra firme forests of Amazonia, irrespective of species composition. Overall, the data imply an annual uptake of about approximately 7 t C ha-1 year-1. 1 Gerald St., PO Box 69, Lincoln 8152, New Zealand, E-mail: [email protected] ; 2 UFPA, E-mail: [email protected]; 3 MPEG, E-mail: [email protected] ; 4 University of Edinburgh, E-mail: [email protected]; 5 6 Universidade de Vicosa, E-mail:[email protected]; Universidade de Vicosa, E-mail: 7 [email protected]; University of Edinburgh, E-mail:[email protected]. Carbon Storage and Exchange 105 Measurements of soil moisture, CO2 efflux and leaf area index in Rondonia during 1999 Humberto R. da Rocha 1 Maria A.F. Silva Dias 2 Helber C. Freitas 3 Amaury Caruzzo 4 Jorge M. Lopes 5 Maarten Waterloo 6 Osvaldo M.R. Cabral 7 Regina Alvala 8 Ralf Gielow 9 Antonio Manzi 10 Celso von Randow 11 Beatriz Gomes 12 During the AMC-LBA (Jan-May,1999) and 1999 in Rondônia, we collected specific cophysiological data and installed automatic instrumentation to provide further interpretation of aerological, surface fluxes and numerical modelling data. To this end, soil moisture, soil CO2 efflux and leaf area index measurements were made at a pasture (Fazenda N. S. Aparecida) and a forest (Reserva Biológica do Jarú) site near Ji-Paraná, Rondônia. The variation in the 95% continuous soil moisture data series reflects the seasonal and intra-seasonal variability of rainfall, in addition to single rainfall events within the dry season. The soil respiration data showed the correlation of respiration with soil temperature and moisture status is in principle evident. Other issues from the analysis are a low correlation of single measurements against temperature and, more important, a suggested difference in the mechanisms controlling CO2 efflux in the dry and wet seasons. In addition, evidence of periodic events (compatible with intra-seasonal temperature variations) have shown to result in a negative trend from the early dry season (June) to the middle wet season (January). Preliminary results of the leaf area index profile in the forest suggest a pattern consistent with the exctintion radiation profile obtained simultaneously. In January, the average LAI values for forest and pasture were 4.9 and 2.7 m2 m-2, respectively. 1 Universidade de Sao Paulo; Rua do Matao, 1226, Tel: (011)8184713, Fax: (011)8184714, E-mail: [email protected] ; 2 USP, 3 USP, 4 USP, 5 USP, 6 Alterra G.W. Research; 7 Embrapa 8 INPE; 9 INPE; 10 INPE; 11 INPE; 12 UNIR. Carbon Storage and Exchange 106 Preliminary assessment of VOC fluxes from a primary rain forest performed in the LBA site at Manaus Paolo Stefani 1 Riccardo Valentini 2 Paolo Ciccioli 3 Enzo Brancaleoni 4 Massimiliano Frattoni 5 Antonio D. Nobre 6 Alessandro C. De Araujo 7 To precisely assess the turnover of carbon from vegetation covered areas, it is essential to quantify the amount of assimilated carbon that is released in the atmosphere in the form of isoprenoid compounds. A suitable way to do it is to measure fluxes of biogenic volatile organic compounds (BVOC) by relaxed eddy accumulation. To gain information on the carbon balance of primary rain forests of the Amazon basin, a measuring site for VOC and CO2 was built in the test site managed by the INPA of Manaus. Eddy correlation (EC) measurements were combined with relaxed eddy accumulation (REA) determinations in a 40 m tower located inside a primary rain forest located 80 km from the city of Manaus (North direction). Preliminary data on BVOC fluxes collected from August 1999 to January 2000 indicate that the most abundant compound released by this forest ecosystem was isoprene with average normalized fluxes ranging from 1000 to 1500 ng m-2 s-1. Main monoterpene compounds released were pinenes and sabinene. Their fluxes ranged between 200 and 250 ng m-2 s-1. In terms of carbon, a net emission of reduced carbon ranging from 90 to 120 nmol m-2 s-1 was measured under standard conditions of PAR and temperature. Based on these preliminary data, fraction of emitted vs. assimilated varying from 0.46 and 0.6% were measured for the type of rain forest existing in Manaus. More extensive investigations are planned to confirm our preliminary observations and to follow the seasonality of BVOC emissions. They will be performed within the frame of the LBACarbosink project sponsored by the European Union. 1 Department of Forest Science (DISAFRI), Università della Tuscia, Viterbo, ITALYvia C. De Lollis, 01100, Tel: ++390761357394 Fax: ++390761357389, E-mail: [email protected]; 2 DISAFRI, E-mail: 3 [email protected] ; Istituto sull’Inquinamento Atmosferico del CNR, E-mail: [email protected]; 4 Istituto sull’Inquinamento Atmosferico del CNR; 5 Instituto sull’Inquinamento Atmosferico del CNR; 6 INPA, E-mail: [email protected] ; 7 INPA, E-mail: [email protected]. Carbon Storage and Exchange 107 Annual carbon dioxide fluxes across the Amazon basin: Preliminary comparison of 1999 results for some LBA-flux tower sites Pavel Kabat 1 Antonio Nobre 2 John Grace 3 Antonio O. Manzi 4 Bart Kruijt 5 Celso von Randow 6 Jan A. Elbers 7 Yadvinder Malhi 8 Alistair Culf 9 Carlos Nobre 10 Humberto R. da Rocha 11 Maarten J. Waterloo 12 Alessandro C. de Araujo 13 Paulo Jorge Oliveira 14 Helber C. Freitas 15 During 1999, continuous measurements of carbon dioxide fluxes were carried out at several LBA sites, set up jointly by the Europe – Brazil flux tower consortium: Rondonia (1 forest and 1 pasture site); Manaus (2 forest sites) and Caixuana (forest site). In addition to these continuous CO2 measurements, towers in both forest and pasture were set up to collect micrometeorological and soil data during the Wet AMC campaign in Rondonia. The coverage of the seasonal cycle by direct measurements was satisfactory; we have collected data on 70 - 90 % of the days, depending on respective site and starting date of the measurements. After preliminary analysis, data for all 5 forest sites indicate a high CO2 uptake during 1999, ranging between 3 to 7 ton C ha-1 y-1 . We will demonstrate several ways of analysis applied to the 1999 data, including the role of flux corrections and the sensitivity of the results to the treatment of night time CO2 fluxes and canopy storage. Through systematic inter-comparison between the sites and comparison with results obtained in previous studies, we attempt to further narrow down possible uncertainties in these potentially very high carbon uptake values. 1 Alterra Green World Research, P.O. Box 47, 6700 AA, Wageningen, The Netherlands, Tel: +31 317 474314; Fax: +31 317 424812, E-mail: [email protected]; 2 INPA, E-mail: [email protected] ; 3 University of Edinburgh, E-mail: [email protected]; 4 CPTEC-INPE, E-mail: [email protected]; (5) Alterra Green World Research, E-mail: [email protected] ; 6 CPTEC-INPE, E-mail: [email protected] ; 7 Alterra Green World Research, E-mail: [email protected]; 8 University of Edinburgh, E-mail: [email protected]; 9 NERC-Institute of Hydrology, E-mail: [email protected]; 10 CPTEC-INPE, E-mail: [email protected]; 11 Universidade de São Paulo, E-mail: [email protected]; 12 Alterra Green World Research, E-mail: [email protected] ; 13 INPA, E-mail: [email protected] ; 14 CPTEC-INPE, E-mail: [email protected]; 15 Universidade de São Paulo, E-mail: [email protected] Carbon Storage and Exchange 108 The sensitivity of NEP to radiation, CO2, temperature, air and soil humidity in three Amazon rain forests Bart Kruijt 1 Celso Von Randow 2 Antonio Donato Nobre 3 Fiona Carswell 4 Lourdes Ruivo 5 Yadvinder Malhi 6 Antonio Manzi 7 Pavel Kabat 8 John Grace 9 Alistair Culf 10 Jan Elbers 11 Maarten Waterloo 12 Allessandro Carioca 13 Paulo George 14 Paulo Stefani 15 Patrick Meir 16 Humberto Da Rocha 17 Since early 1999, net ecosystem exchange of CO2 as well as fluxes of water and heat and meteorological variables have been measured almost continuously in three Amazon forest sites: in the Cuieiras basin near Manaus, in Reserva Jaru, Rondonia and in the Reserva Caxiuana, Para. Of the first two of these sites, earlier data sets exist as well. In this study, we give an overview of these data sets and explore the relationships between measured variables. In particular, we will look at the ecosystem physiology, as represented by the response curves of NEP to incident radiation, air and soil humidity and nocturnal NEP (ecosystem respiration) to temperature. We will also calculate surface conductance using the evaporative and momentum flux measurements, and explore the sensitivity of this variable to the environment, comparing with earlier parameterisations for this variable at leaf empirical model of NEP, and use this to explain seasonal variations observed as well as inter-site differences. 1 Alterra, P.O. Box 47 , 6700 AA Wageningen, The Netherlands, Tel: +31 317 474331, Fax: +31 317 419000, E-mail: [email protected] ; 2 CPTEC-INPE, E-mail: [email protected] ; 3 INPA, Email: [email protected] ; 4 University of Edinburgh, E-mail: [email protected] ; 5 Alterra Green 6 World Research, MPEG, E-mail: [email protected]; University of Edinburgh, E-mail: 7 8 [email protected] ; CPTEC-INPE, E-mail: [email protected] ; Alterra Green World Research, E-mail: [email protected] ; 9 University of Edinburgh, E-mail: [email protected] ; 10 NERC-Institute of Hydrology, E-mail: [email protected] ; 11 Alterra Green World Research, E-mail: [email protected] ; 12 Alterra Green World Research, E-mail: [email protected] ; 13 INPA, E-mail: [email protected] ; 14 CPTEC-INPE, E-mail: [email protected] ; 15 University of 16 Tuscia, E-mail: [email protected] ; University of Edinburgh, E-mail: [email protected] 17 Universidade de São Paulo, E-mail: [email protected]. Carbon Storage and Exchange 109 Seqüestro de carbono uma experiência concreta: Resultados iniciais do projeto de seqüestro de carbono da Ilha do Bananal e seu entorno Divaldo Rezende 1 O principal objetivo do Projeto de Seqüestro de Carbono da Ilha do Bananal e seu Entorno (PSCIB) é desenvolver e implementar um sistema inovador, eqüitativo e sustentável para equilibrar as emissões dos gases causadores do efeito estufa através do Seqüestro de Carbono, compatível com as realidades sociais e ambientais da região da Ilha do Bananal. Um das características do PSCIB é que o projeto está inserido em uma zona ecotonal, ou seja onde ocorre uma interação entre dois ou mais ecossistemas, desta forma foram obtidos resultados específicos para 4 tipos de vegetação abrangendo 3 diferentes ecossistemas; Floresta de terra seca, floresta inundada, cerrado e várzeas. Os cálculos de estoque de Carbono da área do projeto foram calculados considerando duas variáveis. Primeiramente o geoprocessamento de imagens temporais nos anos de 1986, 1991,1995 e 1998 para definição da dinâmica de desmatamento e análise do uso da terra. E a Segunda variável, considerada neste estudo, são os teores de biomassa dos diversos ecossistemas da Ilha do Bananal e seu entorno. As perdas que ocorrem na área sem o projeto é de algo aproximado em torno de 993.888 toneladas de carbono, considerando somente a área tampão (as áreas entre os parques do Araguaia e do Cantão) estes valores são reduzidos a 615.531 toneladas de carbono que deixam de ser absorvidas a cada ano, mesmo assim são valores significativos que se extrapolados para os 25 anos de vida do projeto somarão algo em torno de 15.388.275 toneladas de carbono. Neste trabalho não foram consideradas as possíveis perdas em função das queimadas, mas esta variável deverá ser monitorada ao longo do projeto. 1 ECOLOGICA ASSESSORIA; ACSV-SE 22 LT 20 sls. 103/104 Palmas -TO Cep 77125-220 Tel: 063-215-1279, Fax: 063-225-3101, E-mail: [email protected]. Carbon Storage and Exchange 110 Avaliação de modelos biofísicos nas trocas de energia, água e carbono: PROJETO LBA-TRMM/1999 Julio Tota 1 Humberto R. Rocha 2 Carlos Nobre 3 Antonio Ocimar Manzi 4 Lianhong Gu 5 D. Fuentes 6 Gilberto Fisch 7 Fotossíntese e respiração da vegecao e solos têm um papel importante no ciclo do carbono em florestas úmidas da região da Amazônica. Um grande esforco tem sido feito para medir e modelar esses processos, os quais formam uma importante componente das interacoes entre a biosfera-atmosfera dentro do ecossistema amazonico. Como parte do projeto TRMM-LBA, foi realizada uma campanha de campo com medidas meteorológicas de superfície (temperatura, pressão, umidade e CO2 , etc.) em uma área de pastagem no Estado de Rondônia durante a estação chuvosa de 1999. Esta constou de medidas de superfície em uma torre micrometeorológica. As medidas de resposta rápida dos fluxos de CO2 foram feitas a 6 metros de altura por um sistema de "Eddy Covariance". Tambem foram feitas medidas das concentrações de CO2 em 4 níveis (0,5; 1,5; 4,0 e 9,5 metros) através de um perfilador de gás (Gas Profile). Neste trabalho sao realizadas simulacoes com modelos biofisicos (SiB2c e "Gu model") e a performance dos modelos para os principais fluxos de superficie sao avaliadas junto aos dados observados durante a campanha de campo do projeto LBA-TRMM/1999. O comportamento do ciclo diurno desses fluxos (em particular o CO2 ) associados as diferentes condicoes de estabilidade sao tambem analisados. Uma comparacao entre os fluxos de CO2 obtidos em dois pontos numa area de pastagem e em uma area de floresta sao analisados e comparados a simulacoes dos modelos biofisicos. 1 CPTEC, Rodovia Presidente Dutra, Km 40, Cachoeira Paulista, São Paulo, Tel: 5608400, E-mail: [email protected]; 2 USP, E-mail: [email protected]; 3 CPTEC, E-mail: [email protected]; 4 CPTEC, E-mail: [email protected] ; 5 University of Virginia, E-mail: [email protected] ; 6 University of Virginia, E-mail: [email protected]; 7 IAE-CTA-ACA, E-mail: [email protected] . Carbon Storage and Exchange 111 Influence of fire and soil water content on the aboveground phytomass in the “campo sujo” ecosystem Alexandre José Barbosa Santos 1 Gabriela Tunes da Silva 2 Carlos Alberto Quesada 3 Lacê Medeiros Breyer 4 Heloísa Sinátora Miranda 5 Antonio Carlos Miranda 6 The biomass production in the cerrados region presents strong seasonal changes which are related mainly with rain distribution along the year, in addition these ecosystems are often subject to fire that has influence in the productivity. This study was carried out in two adjacent areas of “campo sujo”, a type of grassland with sparse shrubs. The aboveground phytomass (AGP) productivity was compared for a recently burned area and for an area that had been kept unburned for one year. Along with the assessment of the phytomass production, the soil water content was measured at depths from 5 to 360 cm with a neutron probe. In the burned area, the estimated AGP production was 36 g m-2 two months after the fire, during the same period the unburned area presented a AGP increase of 56 g m-2. With the progression of the rainy season, the burned area presented more productivity as soil water content increased. From November to February the AGP accumulation was 202 g m2 in the burned area while in the unburned area it was only 40 g m-2. Both areas presented similar behavior relative to soil water storage, the water availability for the soil was calculated as 120 mm m-1 and the estimated evaporation rate 27 days after the fire was 0.4 mm day-1. 1 Universidade de Brasília, SQN 406 bloco F apto 208, E-mail: UnB; 5 UnB; 6 UnB; 7 UnB. [email protected] ; 2 UnB; 3 UnB; 4 Carbon Storage and Exchange 112 Measuring carbon balances in the Amazon Basin: II. A new instrument for obtaining long-term eddy-covariance fluxes and high-accuracy [CO2 ] in the tropics Bruce C. Daube 1 Scott R. Saleska 2 J. William Munger 3 Steven C. Wofsy 4 V.W.J.H. Kirchhoff 5 In part II of a two-part presentation (see also Saleska et.al.), we describe a new instrument for measuring eddy-covariance fluxes of CO2 , water vapor, momentum and sensible heat, and for making continuous >high-accuracy measurements of [CO2 ] in terrestrial and marine boundary >layer air. Fluxes will be measured in an old-growth Amazonian rainforest >(Tapajos National Forest, km 67, Santarem, PA), and high- accuracy [CO2 ] will be measured continuously both at the forest site and in Atlantic marine boundary layer air as it advects into the Amazon basin from the South American coastline (at Maxaranguape observatory, just north of Natal, RN - see Kirchoff et al.). The instruments consist of modular tower-mounted units (approximately 1m x 0.6m x 0.2m) which contain all key measurement and datalogging instrumentation, thus minimizing sample tube length (~2m in our case) between air intake and the closed-path pressure-controlled infra-red gas analyzer (Licor 6262). This eddy-covariance instrument maintains advantages of closedpath designs (e.g. precise instrument calibration) while also adding some advantages (e.g. minimal disturbance of the air sample before measurement) attributed to open-path designs. This system is particularly suitable for very tall vegetation ecosystems where sample tube-wall adsorption problems would be exacerbated by the long travel distance from the top of the tower to a ground-based measurement device. The high-accuracy [CO2] measurements (<0.5 ppm) will be compatible with the global CMDL sampling network, and can provide valuable near-surface boundary conditions for airborne campaigns designed to estimate regional or basin-wide CO2 fluxes. 1 Harvard University, 20 Oxford Street, Tel: 617-496-2766, Fax: 617-495-2768, E-mail: [email protected] ; 2 Harvard University, E-mail: [email protected] ; 3 Harvard University, E-mail: [email protected] ; 4 Harvard University, E-mail: [email protected] ; 5 INPE, E-mail: [email protected] . Carbon Storage and Exchange 113 Measuring carbon balances in the Amazon Basin: I. Woody vegetation dynamics in an old-growth tropical rainforest Scott R. Saleska 1 Lucy Hutyra 2 Elizabeth Hammond-Pyle 3 Edna G.T. Guimarães 4 Steven C. Wofsy 5 A major scientific question is the present status of Amazonia as a source or a sink for atmospheric carbon dioxide. We are using ground-based measurements of woody tree increment (dendromtery), combined with whole-system CO2 fluxes (eddy covariance), to elucidate ecological and climatic controls on the interannual carbon balance at na oldgrowth tropical forest (Tapajos National Forest, km 67, Santarem). Here, in part I of a two-part presentation (see also Daube, et al.), we characterize woody vegetation structure, and present results of the first few months of wood growth measurements. We inventoried a total of 945 large (DBH * 35 cm) trees on four 5-ha plots (number density: 47 large trees/ha), and sub-sampled 1643 small trees (10cm < DBH <35cm) on these same plots (number density: 413 small trees/ha). Most trees were provisionally identified to species: 217 species were distributed over 47 families. A stratified weighted random subsample of 1000 trees were selected for intensive study via dendrometry. Preliminary results indicate both high growth rates and high damage and mortality (~3% per year) among subclasses of trees. 1 Harvard University, 20 Oxford Street, Cambridge, MA 02138, USA, Tel: 617-495-9624, Fax: 617-4952768, E-mail: [email protected], ; 2 Harvard University, E-mail: [email protected] ; 3 Harvard University, E-mail: [email protected] ; 4 Fundação Floresta Tropical; 5 Harvard University, E-mail: [email protected] . Carbon Storage and Exchange 114 Comparison of the specific leaf area of 32 woody species in three Cerrado savanna plant communities in Brasilia Carlos A. Klink 1 Dulce A. da Silva 2 In order to predict the behavior of species, communities and whole ecosystems to changes in land use and climate, ecologists are seeking simple ways to describe key plant traits. Specific leaf area (SLA, leaf area per unit dry mass) is easy to measure and is an indicator of the capacity of plants to exploit resources in the environment. Our objective in this study was to compare the SLA of 32 woody species in three Cerrado savanna plant communities in the Reserva Ecologica do IBGE, Brasilia, Federal District, that differ from each other in woody plant density: a grassland with scattered shrub and trees; a Cerrado savanna; and a savanna woodland. Leaf area was measured with an electronic leaf area meter in the field and leaves were oven dried and weighed in the laboratory. SLA ranged from a maximum of 375.8 for the shrub Vellozia flavicans to a minimum of 1.8 for the tree Pali (P<0.001). No significant difference was found between the Cerrado savanna and the grassland but the average SLA of the savanna woodland was significantly higher (P<0.001). 1 Dept. of Ecology, University of Brasilia (UnB), C.P. 04631, UnB, Brasilia DF 70919-970, Tel: 61-3072182, Fax: 61-273-4571, E-mail: [email protected]; 2 Dept. of Ecology, UnB, E-mail: [email protected] . Carbon Storage and Exchange 115 CO2 flux between soil and atmosphere in forests and pastures of Western Amazonia Cleber Ibraim Salimon 1 Antonio Willian Flores de Mello 2 Eric A. Davidson 3 Irving Foster Brown 4 Reynaldo Luis Victoria 5 Conversion of forest to cattle pastures and subsequent abandonment of those pastures are occurring throughout the Amazon Basin, including regions with eutrophic (nutrient rich) and dystrophic (nutrient poor) soils. The objectives of this project are to determine the effects of land-use change and native soil fertility on soil respiration in pastures, secondary forests, and primary forests near Rio Branco, Acre. We are also investigating the relationship between CO2 flux and soil carbon stocks and whether CO2 flux can be used as an indicator of the degree of ecosystem degradation. CO2 flux measurements are made in the field with an infrared gas analyzer on a monthly basis in sixteen sites. Preliminary data analyses from June/99 to Aug/99 shows that the greatest CO2 fluxes are observed in pastures (193 mgC m-2h-1) and not in primary forests (127 mgC m-2h-1). Regarding to soil nutritional status, the preliminary results do not yet show a consistent pattern. For example pastures on eutrophic soils had higher flux in Jul/99 and lower flux in Aug/99 compared to pastures on dystrophic soils. Complementary analysis of microbial biomass, aboveground plant biomass, and root biomass and soil carbon stocks in the three types of land cover and two types of soils are planned. 1 Centro de Energia Nuclear na Agricultura/USP, Av. Centenário, 303 Cx.Postal 96, Piracicaba,SP, 13400-970 Brasil, Tel: (0--19)429-770, Fax: (0--19)4294610, E-mail: [email protected]; 2 Universidade Federal do Acre, E-amil: [email protected] ; 3 Woods Hole Research Center, E-mail: [email protected] ; 4 Woods hole Research Center, E-mail: [email protected] ; 5 CENA/USP, E-mail: [email protected] . Carbon Storage and Exchange 116 Enhancing The Simplified Biosphere Model (SSiB) And A Test With Amazonian Field Measurements Xiwu Zhan 1 Yongkang Xue 2 George J. Collatz 3 Terrestrial ecosystems is recognized as one of the most important parts of the global carbon cycle and the climate system, and accurate assessments of the CO2 fluxes between the terrestrial ecosystems and the atmospphere are pressingly needed. The current version of the Simplified biosphere model (SSIB) does not consider the photosynthetic activities of CO2 assimilation of plants and thus is unable to model the CO2 fluxes. The work to be presented here is to enhance the SSiB model with the capability of modeling CO2 fluxes. The equations and parameters of the photosynthesis-stomatal conductance coupled models by Collatz et al. for C3 or C4 plants are adopted for simulating the uptake or release of CO2 by land surface ecosystems. These models have sound physiological basis and fine description to the interactive effects of environmental factors of plant photosynthesis and stomatal movement. In the current literature, however, the equations of these models were solved numerically with iterations, which consumes a great amount of computational time. In some cases, the iteration process may not be convergent and an appropriate solution may be difficult to obtain. In this study, a new procedure is developed to quasi-analytically solve the equations of these models. When na appropriate solution of these models does not exist under some night time environmental conditions, the procedure applies simpler empirical equations from literature to simulate the processes because the land surface fluxes in these cases are not significant. Off-line testing of the enhanced SSIB model for modeling the land surface fluxes of CO2 , water vapor and sensible heat is conducted against the field measurements from an Amazonian Eddy Covariance Study in 1995 and 1996. The results of the test will be presented. 1 University of Maryland, 2181 LeFrak Hall, Department of Geography, College Park, MD 20742, Tel: 301405-7168, Fax: 603-806-8375, E-mail: [email protected] ; 2 University of California Los Angeles, E3 mail: [email protected] ; Laboratory for Terrestrial Physics, NASA/GSFC, E-mail: [email protected]. Carbon Storage and Exchange 117 Carbon Stocks in Agroforestry Systems Karen A McCaffery 1 Erick C M Fernandes 2 In the past three decades an estimated 25-40 Mha of forest have been cleared in the Amazon basin (Serrao et al., 1995) and approximately 70% of this land has been converted to pasture. Under current land-management practices, some 60% of new pastures degrade to woody secondary forest within fifteen years of establishment and the remaining 40% to unproductive grassland (Fearnside 1995). There are na estimated 20-35 million hectares of abandoned pastures in the Amazon, with low rates of regeneration and carbon sequestration and low biodiversity. While many studies have investigated the effects of deforestation and land conversion on carbon and nutrients (Fernandes et al. 1997; Cerri et al. 1991; Kang and Juo 1986), few studies have measured carbon sequestration in recovering agroecosystems (De Camargo et al. 1999). In 1992 a series of four prototype agroforestry systems we underway to quantify carbon and nutrient stocks in these agroforestry systems compared to unmanaged naturally regenerating secondary forest. Biomass is being estimated through a destructive harvest and the development of species-specific allometric equations. This study is generating information on the contribution of managed agroecosystems to carbon sequestration, nutrient levels and distribution. Agroforestry systems merit attention because they represent (1) an ecosystem management approach that can increase carbon and nutrient sequestration and; (2) an economically viable farming approach that can enhance the productive potential of cleared, degraded land and may reduce the need to clear additional forest. 1 Cornell University/Embrapa, CD Bougainville, Bloco Sandalo, Apto. 505, Rua Rei Arthur, PQ Dez, Manaus, AM, Brasil, Tel: 92-55-236-0042 (Manaus), E-mail: [email protected]; 2 Cornell University, Email: [email protected] . Carbon Storage and Exchange 118 Modelling Rainforest Carbon and Water Dynamics in the Colombian Amazon Lina María Mercado Montoya 1 Sharon A.Cowling 2 Conrado Tobón Marín 3 Germán Poveda 4 Tropical rainforests play a crucial role in regulating global climate processes, with tropical deforestation potentially having severe adverse effects on both regional and global climate conditions and on the global carbon budget . The Colombian Amazon is one of the tropical regions that remains poorly studied in terms of deforestation and its effects on the carbon and hydrological cycles. The objective of this study was to develop a simplified model that can simulate basic processes of carbon and water cycling in an undisturbed mature Colombian rainforest, to compare relative trends in carbon and water dynamics under different land-use change scenarios. Modeling data show conclusive evidence that deforestation followed by intensive agriculture (with subsequent pasture development) has long-term effects on net CO2 emissions to the atmosphere. Alternatively, deforestation followed by less-intensive agricultural practices such as shifting cultivation (with subsequent secondary vegetation regrowth) has much shorter and less severe consequences for net CO2 release to the atmosphere, because of carbon conservation induced by secondary forest regrowth. Modeling trends, such as presented in this study, can be of significant use for policy-makers and forest managers in Colombia, because not only does it help to illustrate some of the general impacts of tropical deforestation and land-use systems on global CO2 budgets, but it plays an even greater role in highlighting the need for more detailed and site-specific research addressing carbon and hydrological cycling in the Colombian Amazon rainforest. 1 Max-Planck Institute fuer Biogeochemie, Postfach 10 01 64, 07701 jena, Tel: 49 3641 643711, E-mail: [email protected]; 2 Lund University, Department of Plant Ecology, E-mail: [email protected] ; 3 University of Amsterdam, E-mail: [email protected]; 4 Universidad Nacional de Colombia, Posgrado de Recursos Hidráulicos, E-mail: [email protected] . Carbon Storage and Exchange 119 Forest dynamics and biomass estimation in bamboo-dominated forest in Acre State, southwestern Amazonia Marcos Silveira 1 Carlos Augusto Klink 2 Irving Foster Brown 3 José Marcelo Domingues Torezan 4 Arborescent bamboo species (Guadua spp.) dominate much of the canopy of forests in the state of Acre. These bamboo-dominated forests cover an area of 122.000 km2 of southwestern Amazonia, including parts of Peru and Bolivia. Our objective in this study is to understand the dynamics of bamboo-dominated forests and the contribution of bamboo species to the total forest biomass. We study forest dynamics at three ecological scales: landscape (edge geometry from Landsat TM images); vegetation structure and tree diversity (field surveys in Acre), and population dynamics (growth, natality and mortality of tagget seedlings and culms). The bamboo-dominated forest dynamics are characterized by flowering, die-back and colonization events. The patches show distinctive features that permit using time sequence Landsat imagery to follow their evolution, as B. Nelson of INPA has shown. Surprisingly, tree diversity is among the highest of forests studied in Amazonia. G. weberbaueri biomass was estimated through regression analyses on the dry above-ground biomass; this species is among the 7 tree species with the highest biomass. From a carbon-stock perspective, bamboo contributes only a small proportion of the total biomass, but in terms of resource partitioning, it is important due to the understory dominance. The culms of G. weberbaeuri grows rapidly in rainy season, reaching 15 m in height in few months. Their fast growth, coupled with arborescent habit appears to alter the development of certain tree species. 1 Universidade de Brasília, Universidade Federal do Acre, Campus Universitário, 70910-900, E-mail: 2 3 [email protected]; Universidade de Brasília, E-mail: [email protected] ; Woods Hole Research Center/Universidade Federal do Acre/Universidade Federal Fluminense, E-mail: [email protected]; 4 Universidade Estadual de Londrina, E-mail: [email protected]. Carbon Storage and Exchange 120 Spatial and temporal variation in carbon cycling dynamics in central Amazon forests Jeffrey Q. Chambers 1 Roseana Pereira da Silva 2 Luis Claudio de Oliveira 3 Joaquim dos Santos 4 Niro Higuchi 5 Central Amazon forests that lie off the floodplains of major rivers are often designated as terrafirme. Many of these forests, however, are characterized by small-scale (< 1 km) changes in elevation, soil type, water balance, and floristic composition that can alter carbon cycling dynamics. Soil texture, for example, shifts from ~80% clay on plateaus (Oxisols) to ~5% clay in valleys associated with ubiquitous small streams ("baxio"). The baxio represents about 1/3 of the forests spatial exent, and the slopes about another 1/3, however, most studies have focused on plateaus. We are measuring tree growth, soil respiration, and stem wood respiration in two 20 x 2,500 m permanent transect plots (toposequences) ~60 km north of Manaus Brazil that representatively sample the landscape. We find that the cycling of carbon changes significantly along this toposequence, and that, in some respects, forests associated with baxio behave more like floodplain than terra-firme forests. Monthly changes in tree diameter using highly accurate (±0.01 mm) tree dendrometers bands (~400), for example, show a significant seasonal change in carbon allocation to wood production associated with precipitation variability and topography. In August, the driest month, forest-wide wood production averaged 5.0 g kg-1 day-1, whereas, in December, one of the wettest months, average wood production increased to 60.5 g kg-1 day-1. Growth rates were log-normally distributed, however, a number of trees, at any given time, were not producing stem wood. In the historically driest month, August, only 54% of trees grew in diameter, whereas by December (one of the wettest months), 72% of trees were producing wood. This seasonal shift in carbon allocation varied with topography. Forests in the baxio produced most wood in the late dry season, when the water table had fallen, and on plateaus the peak in wood production occurred in the early wet season. When these signals are combined, December (considered both late-dry and early-wet), is the most productive month because the entire forest experienced a peak in wood production. This peak dropped by about 50% in January, from 60.5 g kg-1 day-1 to 34.5 g kg-1 day-1, as baxio wood production diminished. These results agree with previous eddy covariance measurements at a site near our plots that demonstrated a significant decline in forest productivity as the wet season progressed. With respect to other carbon cycling measures, soil respiration also demonstrated strong spatial variability in the mid wet season, with the baxio showing extremely low respiration rates associated with soil moisture saturation. Surprisingly, there was only a weak correlation between wood production and wood respiration rates, and variability in this carbon cycling component is still being explored. Overall, we are finding that spatial and temporal heterogeneity associated with topography influences large-scale carbon cycling patterns in central Amazon forests, and that these forests are better characterized as islands of terra-firme surrounded by a dendritic network of seasonally inundated forests. 1 Univeristy of California and Instituto Nacional de Pesquisas da Amazônia INPA/CPST CP 478, Manaus, Amazonas, Brasil, 69070-110, Tel: 92-642-3990, Fax: 92-642-4078, E-mail: [email protected]; 2 INPA, Email: [email protected]; 3 INPA, E-mail: [email protected]; 4 INPA, E-mail: [email protected]; 5 INPA, E-mail: [email protected] . Carbon Storage and Exchange 121 Prediction of delta 13 C for Yellow Latosolos of Amazon under primary forest Everaldo de Carvalho Conceição Telles 1 Plinio Barbosa de Camargo 2 Susan E. Trumbore 3 Luiz Antonio Martinelli 4 Reynaldo Luis Victória 5 Marcelo M. Zacharias 6 Joaquim dos Santos 7 This work analyzes the del-13C and the carbon and nitrogen content of soils using multiple linear regression and non-linear regression. The samples were collected in triplicate in pits at depths of 0-5, 5-10, 10-20, 20-30, 30-40 and 40-50 cm in two yellow Latosol plateaus under primary forest near to Manaus (ZF-2). In the first plateau the spacing between pits was 30 meters, and in the second, three pits were chosen according to morphologic variation. The elemental and isotopic analyses were accomplished using a massspectrometer (Delta plus-Finnigan Mat), coupled to an elemental analyzer (Carlo Erba EA1110). The generated models explain the variation of the del-13C accurately to within about 0,2‰, similar in magnitude to the analytical error. The best model to predict del-13C based on the two plateaus of Manaus, with the variables C and depth was tested in others Yellow Latosols with similar vegetation in the Amazon region at Santarem, Altamira, and Paragominas in Para; Manaus and Humaita in Amazonas; and Pimenta Bueno in Rondonia. The model predicted soil properties with an error below 1‰. The correlation of the obtained results and the available data in the literature through modeling are important because it increases the possibilities of extrapolation of the data obtained in Manaus and Santarém in relation to the del-13C in the soil associated with mechanisms of physical and chemical protection of the carbon and the time of cycling in the soil. 1 Av. Centenário 303, 13416-000, SP, Brazil, Tel: 55-19-4294600, Fax: 55-19-4294610, E-mail: [email protected]; 2 CENA-USP, E-mail: [email protected] ; 3 UC- Irvine, E-mail: [email protected]; 4 CENA-USP, E-mail: [email protected]; 5 CENA-USP, E-mail: [email protected]; 6 CENA-USP, E-mail: [email protected]; 7 INPA. Carbon Storage and Exchange 122 Soil CO2 Efflux in Tropical Forest in Central Amazon Eleneide Doff Sotta 1 Antônio D. Nobre 2 Patrick Meir 3 Yadvinder Malhi 4 This study investigated the spatial and temporal variation of the soil CO2 efflux in a forest near Manaus, Amazonas, Brazil, as well as its relationship with soil temperature and rainfall. After normalization of the soil CO2 efflux data for the mean soil temperature (25º C), we found CO2 the lowest value was 2.18 micromol CO2 m-2s-1 and the highest 10.22 micromol CO2 m-2s-1; the average value was 5.74 micromol CO2 m-2s-1. The diurnal cycle in efflux was most strongly correlated with soil temperature, which peaked between 13 and 14 hours. The temperature was responsible for nearly 80% of the daily soil CO2 efflux variation when soil water content was not a limiting factor. Soil CO2 efflux decreased by nearly 1.4 micromol CO2 m-2s-1 during rainfall. Soil CO2 concentration gradient measurements showed that there was a build up in the concentration of CO2 within the soil after the rainfall, when the soil pores were filled with water, creating a barrier for the gas exchange between soil and atmosphere. Soil CO2 efflux rate also decreased when the soil was dry. The soil water proved to be an important controlling factor to the efflux during rain events and dry season. 1 Rod. do Coqueiro, Conj. Pau D'arco, casa 9, Tel: (091) 235 0068, Fax: (091) 235 0122, E-mail: [email protected]; 2 INPA; 3 University of Edinburgh; 4 University of Edinburgh. Carbon Storage and Exchange 123 Estimating above ground biomass and carbon in young secondary vegetation in Eastern Amazonia J. C. Randolph 1 Eduardo S. Brondizio 2 Emilio F. Moran 3 Using field data collected across 38 secondary succession sites, this poster presents allometric equations to estimate biomass in areas of young secondary vegetation in eastern Amazonia. Data derived from destructive measurements of biomass are used in conjunction to vegetation inventory estimates for the same sites. Laboratory analysis provides data on wood density and total carbon for sampled individuals. Research sites are located around the towns of Igarape-Acu and Tome-Acu, Para State. Above ground biomass direct estimations are compared to allometric equations derived from literature data. The poster discusses issues of extrapolation and accuracy in above ground biomass and carbon estimation in Eastern Amazonia. 1 SPEA 441, Bloomington, IN, USA, 47405, Tel: 812-8556181, Fax: 812-8553000, E-mail: [email protected]; 2 Indiana University, E-mail: [email protected]; 3 Indiana University, E-mail: [email protected]. Carbon Storage and Exchange 124 Above-ground biomass and leaf area index of pastures in Rondônia, Brazil Maarten J. Waterloo 1 Pavel Kabat 2 Humberto R. da Rocha 3 Helber C. Freitas 4 Amaury Carruzo 5 Niek Jan Bink 6 Friso Holwerda 7 Rosivaldo Lelis da Silva 8 Fabrício B. Zanchi 9 Beatriz Machado Gomes 10 Paulo Jorge Oliveira 11 The leaf area index of man-made pastures is an important parameter for modelling the effects of deforestation in the Amazon region on climate change. Yet, little is known about its spatial and temporal variation. To partly fill this gap in our knowledge on these pasture ecosystems, aboveground biomass, litter mass, specific leaf area and leaf area index were measured at two pasture sites in Rondônia, Brazil. At the Rolim de Moura mixed grass pasture (71% Brachiaria sp., 13% Panicum maximum, 16% other species) measurements were made during the wet season of 1999. These indicated an above-ground biomass, litter mass, specific leaf area and leaf area index of 0.317 kg m-2, 0.047 kg m-2, 18 m2 kg-1, and 1.5, respectively. At Fazenda Nossa Senhora Aparecida, near Ouro Pretu, monthly measurements were made over a year in a rather homogenous (99%), grazed Brachiaria pasture. Biomass and litter mass averaged 0.317 kg m-2 and 0.272 kg m-2 respectively. Both biomass and litter mass showed distinct seasonal patterns with biomass being higher during the wet season and litter mass being higher during the dry season. Monthly values of the specific leaf area ranged between 14 and 19 m2 kg-1, with somewhat higher values measured during the dry season. The leaf area index showed a distinct seasonal variation, ranging from about 1 in the dry season to about 2 in the wet season. Wet season LAI results, obtained from destructive sampling at Fazenda Nossa Senhora, showed fair agreement (2.1 versus 2.7) with those obtained using optical techniques. 1 P.O. Box 47, 6700-AA, Wageningen, The Netherlands, Tel: +31 317 474778, Fax: +31 317 424812, E-mail: [email protected]; 2 Alterra Green World Research, Wageningen, The Netherlands, Email: p.kabat@alterra,wag-ur.nl; 3 USP, E-mail: [email protected]; 4 USP, E-mail: [email protected]; 5 USP, E-mail: [email protected]; 6 Vrije Universiteit Amsterdam, Email: [email protected]; 7 Vrije Universiteit Amsterdam, E-mail: [email protected]; 8 UNIR, E-mail: [email protected]; 9 UNIR, E-mail: [email protected], 10 UNIR, E-mail: [email protected]; 11 CPTEC - INPE, E-mail: [email protected]. Carbon Storage and Exchange 125 Increasing Carbon Sequestration on Deforested Land Through the Use of Legume Trees and Fertilizer Inputs Karen A McCaffery 1 Erick C M Fernandes 2 The predominant soil constraints in the Amazon are P deficiency, Al toxicity, and low nutrient reserves (Sanchez 1987). This has significant implications for carbon sequestration in recovering ecosystems. Oxisols occupy an estimated 220 m ha or 46% of the land base in the Amazon (Cochrane and Sanchez 1982), and have poor chemical characteristics, including low levels of base-forming cations and limited nutrient reserves (Cochrane and Sanchez 1982). On upland soils, Ca and P may be the primary limiting nutrients to plant productivity (Fernandes et al. 1997). A study is underway to evaluate the potential of P, Ca and gypsum combined with legume trees for increasing carbon sequestration on deforested, degraded land. Soil fertility management that encourages deep rooting may lead to C deposition at deeper horizons where it may be less likely to be lost via oxidation (Lugo and Brown 1993). Gypsum can promote movement of cations into the subsoil making nutrients available to developing roots at depth. The study was established on an upland clayey Oxisol near Manaus, Brazil in March 2000 in a RCBD with six fertilizer treatments (+P; +Ca; +P and Ca; +Ca and gypsum; +P, Ca and gypsum; and an untreated control) x four tree species (Gliricidia sepium ‘14/84’; G. sepium ‘vertical’; Inga edulis and Senna reticulata). This study will assess whether additions of P, Ca combined with gypsum improve the exploitable rooting volume, NPP, and BNF, and thereby increase contributions to ecosystem carbon and nutrients. 1 Cornell University/Embrapa, CD Bougainville, Bloco Sândalo, Apto. 505, Rua Rei Arthur, PQ Dez, Manaus, AM, Brasil, Tel: +31 317 474778, Fax: +31 317 424812, Email: [email protected]; 2 Cornell University, Email: [email protected] Carbon Storage and Exchange 126 Biomass dynamics in anthropogenic landscapes of the Central Amazon William F. Laurance 1 Rita C. G. Mesquita 2 We summarize results from a large-scale study of biomass dynamics in fragmented, intact, and secondary forests in the central Amazon. Repeated censuses of over 64,000 trees (>=10 cm diameter) in permanent forest-mensuration plots reveal a substantial loss of aboveground biomass in fragmented forests, the result of sharply elevated tree mortality near fragment margins. This loss of tree biomass is not offset by increased growth of lianas and small trees in fragments. In contrast, intact forests in this region and elsewhere in the Amazon may be functioning as a globally significant carbon sink, possibly because of faster plant growth resulting from increasing CO2 fertilization. Secondary forests in this region rapidly accumulate above-ground biomass, accruing up to 30% (90-110 tons/ha) of that in primary forests (360+/-50 tons/ha) in the first ten years of growth. However, landuse history and especially repeated burning have dramatic effects on soil seedbanks, regenerating forests, and rates of carbon sequestration. Some implications of these findings for Amazon biomass dynamics and conservation will be highlighted. 1 Smithsonian Institution and INPA, C.P. 478, Manaus, AM, 69011-970, Brazil, Email: [email protected], Tel.: 92 642 1148, Fax: 92 642 2050; 2 Smithsonian Institution and INPA, Email: [email protected] Carbon Storage and Exchange 127 Carbon Dioxide Measurements in the Bolivian Amazonia René Gutiérrez 1 The environmental CO2 mixing ratio was measured in Cobija, Bolivia (11º3´S, 68º31´W) during one year. The location is a small town surrounded by humid tropical forest ans is typical of the Bolivian Amazonia. The air samples were taken near one meter below the canopy height, and this is reflected in the large variations inside the daily cycle values: a difference which usually is gretar than 100 ppmv between maxima and minima. The mean value for the steady period in the afternoon hours is 406 ppm. The fires incremented this value in 28 ppm during the 1999 "burning season". Also, some special events are reported, whith short-duration peaks of extremely high values. Most of these events had been identified with large forest firesand strong winds. 1 Laboratorio de Física de la Atmósfera, Instituto de Investigaciones Físicas, Universidad Mayor de San Andrés, La Paz, BOLIVIA. Postal Address: P.O. Box 3164, La Paz, Bolivia, Tel - fax: (591 - 2) – 799155, E-mail: [email protected]. Carbon Storage and Exchange 128 Mudanças globais e taxa de crescimento das espécies arbóreas da Amazônia Simone Aparecida Vieira 1 Plínio B. de Camargo 2 Susan Trumbore 3 Irving F. Brown 4 Elsa R. H. Mendoza 5 Diogo Selhorst 6 Jeffrey Chambers 7 Joaquim dos Santos 8 Para compreender o papel da região Amazônica no balanço global de carbono e em particular o potencial das florestas primárias de atuarem como provedor (“source”) ou sorvedor (“sink”) de carbono devido à mudanças climáticas, é essencial que se conheça a dinâmica de importantes reservatórios de carbono que operam em escalas de tempo maiores que as anuais, em especial os processos de crescimento e mortalidade das árvores. A determinação da idade das árvores da floresta tropical úmida pode fornecer informações cruciais para o entendimento da dinâmica das populações arbóreas e das taxas de ciclagem de carbono assim como estimar o tempo médio de permanência do carbono na vegetação. Os objetivos deste trabalho são: (1) determinar as taxas médias anuais de crescimento das árvores, (2) o crescimento diamétrico médio dos últimos 30 anos, (3) a distribuição etária (4) o estoque e o tempo de permanência do carbono na vegetação e como estes parâmetros variam em decorrência do gradiente climático existente na Amazônia utilizando-se análise do 14C da celulose do tronco das árvores. Serão feitos estudos dendrométricos, dendrocronológicos e isotópicos. Os resultados da atividade isotópica obtida nos anéis de crescimento das árvores serão utilizados no ajuste de uma curva de calibração de atividade do 14 C. As amostras serão coletadas em plots implantados próximos a Manaus, Santarém e Rio Branco. 1 CENA/USP, Av. Centenário 303, 13416-000, Piracicaba, SP, Telefone: 19-4294788. Fax: 19-4294610, Email: [email protected]; 2 CENA/USP, E-mail: [email protected]; 3 UCI-Dept Earth System Science, E-mail: [email protected]; 4 UFAC, E-mail: [email protected]; 5 UFAC: E-mail: [email protected]; 6 Harvard: [email protected], [email protected]; 7 INPA, E-mail: [email protected]; 8 INPE, E-mail: [email protected]. Carbon Storage and Exchange 129 Photosynthetic gas exchange among ecosystem compartments in forest and pasture ecosystems Tomas Domingues 1 Jean Ometto 2 Luiz Martinelli 3 Larry Flanagan 4 Jim Ehleringer 5 The biosphere exerts a strong influence on the atmosphere via gas exchange processes, such as photosynthesis. The objective of this study is to collect photosynthetic response curves of dominant life forms in both forest and pasture ecosystems at Santarém and Manaus. Data collected with a LiCor 6400 are providing an integrated picture of photosynthetic gas exchange and will be of considerable use in interpreting ecosystem gas exchange measurements and isotopic data collected by other projects. Measurements are being made at regular intervals so that phenological dynamics in the structural and physiological properties of the canopy can be assessed. In particular, measurements are being made during both the rainy and the dry season to better understand and represent in models the effects of moisture stress on canopy processes. Over a 2-year period, the photosynthetic gas exchange measurements that will characterize plant responses during wet and dry seasons are, photosynthetic dependence on light (30 °C, 70 % r.h., 360 ppm CO2 ), internal CO2 (30 °C, 70 % r.h., 1800 µmol m -2 s-1 PFD), and humidity (30 °C, 360 ppm CO2 , 1800 µmol m -2 s-1 PFD). Response curves are being measured from trees (up, mid, and lower canopy) and lianas (upper canopy) within reach of towers in the forest ecosystems and from grasses and shrubs at pasture ecosystems. Measurements taken from different heights will relate photosynthetic response curves to CO2 , humidity, and light with foliar nitrogen and water potential for both wet and dry seasons. 1 University of Utah, E-mail: [email protected]; University of Utah. 2 CENA/USP; 3 Lethbridge University; 4 Carbon Storage and Exchange 130 Stable Isotope Analyses of Air, Water, and Organic Materials in Forest and Pasture Sites Jean Ometto 1 Larry Flanagan 2 Luiz Martinelli 3 Tomas Domingues 4 Jim Ehleringer 5 The concentrations and isotopes ratio of CO2 atmospheric in plant canopies have been measured to characterize isotope fluxes in forest and pasture sites in Manaus (ZF2 Tower and Fz. Esteio), Santarem (Flona Tapajos wooden tower and Km 77 pasture flux tower) and Rondonia (Rebio Jaru flux tower and Fz. Nossa Senhora) during the past year. In addition, we have measured (a) the oxygen isotope ratios of leaf water, stem water, and water vapor to help interpret the oxygen isotopes of carbon dioxide, and (b) carbon and nitrogen isotope ratios in leaves and soils within these ecosystems to help interpret the carbon isotopes of carbon dioxide. The results encompass four field campaign on March, May and October, 1999 and February 2000, for Manaus and Santarem, and from May 1999 and February 2000 for Rondonia. The carbon isotope ratios of atmospheric CO2 at the forest sites ranged from -16.73 to -7.47 ‰, with most of the differences accounted for by canopy height. Oxygen isotope ratios of carbon dioxide exhibited a similar variance, but the highest values were observed at the top of the canopy. At pastures the carbon isotope ratios of atmospheric CO2 were more enriched than at forest sites. Data are presented for oxygen isotope ratios of various water sources as a mechanism for interpreting variation in the oxygen isotopes of carbon dioxide. Carbon and nitrogen isotope ratios have been analyzed on the dominant plants in forest and pasture systems in all 3 regions. In Manaus, the average carbon isotope values were - 32.41 ‰, -12.60 ‰, and -30.34 ‰ for forest trees, pasture grasses and pasture shrubs, respectively. In Santarem the average carbon isotope values were -33.58 ‰, -12.80 ‰, and -30.35 ‰ for forest trees, pasture grasses and pasture shrubs, respectively. 1 CENA-USP; 2 University of Utah, Lethbridge University; Utah; 5 University of Utah. 3 CENA-USP; 4 CENA-USP, University of Carbon Storage and Exchange 131 Light and water capture by vegetation communities on abandoned degraded pastures in the Manaus region Steven A. Welch 1 Susan J. Riha 2 John M. Duxbury 3 Marco A. Rondon 4 Erick C.M. Fernandes 5 Abandoned, degraded pastures, when compared to mature forest or healthy pastures, are believed to have less photosynthetic canopy and roots. Reduced ability to capture light and water resources decreases plant productivity, thereby impeding forest recovery or development of other desirable plant communities. Also, fewer leaves and roots could affect critical interactions between the biosphere and atmosphere, such as evapotransporation, light absorption and albedo. If occurring on a large scale, the sensible and latent heat balance could be affected. Monitoring the capture of light and water resources offers a means to evaluate the recuperation of abandoned, degraded pastureland while generating information essential to evaluate impacts of land use change in Amazonia. This research proposes to use repeated measurements of leaf area indices (LAI) and soil water stocks as indices of resource capture in areas of different land use. Resource capture dynamics are being monitored in primary forest, pastures, and secondary forests of various ages (4, 8 and 12 years). Agroforestry management areas are also being monitored. Soil water stocks are being monitored weekly at 0-3 m depth using a neutron probe and at 0-2 m using a TDR probe. LAI is being measured using hemispheric images on a bimonthly basis. Results from this work are expected to offer insight about development of resource capture systems by recovering vegetation, and how human intervention might accelerate that development. 1 2 Cornell/Embrapa-CPAA, Manaus, E-mail: [email protected]; Cornell University, E-mail: 3 [email protected]; Cornell University, E-mail: [email protected]; 4 EMBRAPA-CPAA, E-mail: [email protected]; 5 Cornell University, E-mail: [email protected]. Carbon Storage and Exchange 132 Modeling Seasonal and Interannual Variability in Terrestrial Ecosystem Hydrologic Fluxes for the Brazilian Amazon Region Christopher Potter 1 Alicia Torregrosa 2 Steven Klooster 3 Joseph Coughlan 4 Vanessa Brooks Genovese 5 Jennifer Dungan 6 Matthew Bobo 7 Previous field site measurements have implied that regional hydrology and productivity in the Amazon basin has varied significantly during the past decades of El Nino events in both managed agricultural and natural forest areas. We investigated this hypothesis using a regional ecosystem model for coupled water and biogeochemical cycling. Seasonal and interannual controls on evapotranspiration and net ecosystem production (NEP) were studied with integration of multi-year satellite data to characterize Amazon land surface properties over time. Background analysis of temporal and spatial relationships between regional rainfall patterns and satellite observations for vegetation land cover reveals several notable patterns in the model driver data. Autocorrelation analysis for monthly vegetation "greenness" index (NDVI) from the Advanced Very High Resolution Radiometer (AVHRR) and monthly rainfall (both at 8-km resolution) indicates a significant lag time correlation of up to 12 months. At lag times of 24 and 36 months, autocorrelation values did not exceed the 95 percent confidence interval at locations west of about 47 deg. W, which is near the transition zone of seasonal tropical forest and other (non-forest) vegetation types. At lag times of 12 months or less, the location near Manaus (ca. 60 deg. W) represents the farthest western point in the Amazon region were seasonality of rainfall accounts significantly for monthly variations in forest phenology, as observed using NDVI. Ecosystem model results using this newly generated combination of regional forcing data from satellites suggest that (1) evapotranspiration fluxes from deeply rooted Amazon forests are driven primarily by seasonal and interannual variability in surface radiation fluxes, and (2) Amazon primary forests can be relatively strong net sinks for atmospheric carbon dioxide, particularly during wet (non El Nino) years. However, drought effects during El Nino years can reduce net primary production in forests of the eastern Amazon by 10-20 percent, compared to long-term average estimates of regional productivity. 1 NASA Ames Research Center, Ecosystem Science and Technology Branch, Moffett Field, CA 94035; 2 Earth System Science and Policy, California State University, Monterey Bay, Seaside, CA 93955; 3 CA SU; 4 NASA Ames; 5 Johnson Controls World Services, NASA; 6 CA SU; 7 Johnson Controls World Services, NASA. Carbon Storage and Exchange 133 Fluxes of CO2 , water and energy over several savannas John Grace 1 Savanna-type vegetation is the most abundant land surface, constituting almost 17 % of the terrestrial cover. We report new data on the biophysical characteristics of savanna vegetation in Venezuela and Brazil, from the project SAVAFLUX. Data from eddy covariance sensors, investigating intact and disturbed ecosystems, reveal seasonal and geographical variation. We show the effect of burning campo sujo. Comparison will be made with other forms of tropical vegetation including tropical forests and pastures. 1 University of Edinburgh, United Kingdom, E-mail: [email protected]