LBA JiParana

The Large Scale Biosphere-Atmosphere Experiment in Amazonia (LBA) is an international research initiative led by Brazil. LBA is centered on two key questions:

  1. How does Amazônia currently function as a regional entity?
  2. How will changes in land use and climate affect the biological, chemical, and physical functions of Amazônia, including the sustainability of development in the region and the influence of Amazônia on Global climate?

As a joint endeavor with the Centro de Energia Nuclear na Agricultura (CENA, Piracicaba Brazil), we are conducting the "Biogeochemical Dynamics in River Corridors of the Amazon Basin and their Response to Anthropogenic Change" project. Our perspective is that the periodically or permanently flooded areas play important roles in the hydrology and biogeochemistry of the Amazon basin, and are particularly subject to changes in land cover and land use. As land cover changes, with concomitant alterations in hydrology and vegetation, significant changes in the partitioning between aqueous and gaseous species are be expected in river corridors. Accordingly, our research objectives for river corridors are to: 1) Establish the geographic and geochemical sources of C, N, and P species, 2) Define the biogeochemical consequences for the gas emissions and water chemistry associated with these species of anthropogenic perturbations against the background of natural environmental variability, and 3) Determine the distance that sediments, nutrients, and organic matter travel downstream before they are taken up, decomposed, temporarily stored, permanently buried or degassed, as mediated by the transport properties of the river system and the reactivity of the materials themselves. Our focus is the mesoscale Ji-Parana River Basin, in the State of Rondônia, Western Amazônia. The basin is important as both a focus of landuse activities in the Amazon, and as a "model" for regional dynamics.

Research Elements

  1. Development of Spatial Model of the Ji-parana Basin. We are developing an integrated analysis of the landscape characteristics of the Ji- Paraná river, based on the spatial analysis of various spatial patterns, including soil properties, river network, topography, and land use/cover (Ballester et al in review). This physical template is a GIS-based comprehensive tool to support the understanding of the biogeochemistry of surface waters. We divided the Ji-Paraná river basin into 14 drainage units, organized according to the river network morphology and degree of land use impact. Each sector corresponds to a sampling point where our group is determining river biogeochemistry. To delineate sub basin boundaries of the sampling sites, a 50 meter vertical resolution Digital Elevation Model (DEM) and river network were derived from 1:100000 maps from the Brazilian Institute of Geography and Statistics. Soil units were derived from a 1:500.000 map that was manually digitized. A series of soil profiles derived from the SIGTERON data set are under processing to derive soil property maps. A land use/cover map for 1999 was produced by the digital classification of eight Landsat 7 ETM+ scenes already geocorrected and georeferenced, acquired from the Tropical Rain Forest Information Center (TRFIC) at Michigan State University. A time series of land use and land cover change is under development. To access the spatial distribution of human interference we are also working on the following data sets: roads, settlement projects and legal reserves (national parks and Indian reservations). These data sets were obtained from the Planaflora Project. The geology and geomorphology maps were also obtained from the same source. It is important to point out that these data are available only for this project due to restrictions imposed by the data provider. We are working on the possibility to make these data public domain at least for the Ji-Paraná river basin. We are also working on the demographic and agricultural census data from IBGE. We have already compiled a 20 year data time series. A series of socioeconomic and agricultural indicators were chosen and made spatially-explicit using the county as a study unit. These data are under analysis to develop a socioeconomic and agricultural diagnosis of the study area.

  2. Ji-parana Field Transect Measurements. Our fieldwork is based on two sets of transects through the Ji- Paraná basin. The first set involved sampling the main river channel and tributaries, covering the whole basin in expeditions that lasted 20-30 days. The second approach was designed to compare two sub-basins with contrasting land use/cover. For this, we sampled the upper Ji-Parana river at the city of Cacoal and the Urupa river at the city of Ji-Parana both weekly during the rainy season, and monthly during the dry season. A total of 7 expeditions were conducted along the whole basin, providing detailed spatial information. The sub-basin dataset consists of 25 expeditions, focusing on acquiring valuable temporal coverage to facilitate in the fast response to storm runoff events. During the last two years, a total of around 200 days were spent in the field.

    In the field, we measured discharge, pH, dissolved oxygen, electrical conductivity and temperature. Respiration rates were calculated from the consumption of dissolved oxygen in bottles incubated in the dark. Within a month from the period of sampling, the concentrations of the anions Cl-, SO42-, NO3-, NO2- and PO43- and cations Na+, K+, NH4+, Ca2+ and Mg2+ were determined, along with the concentrations of dissolved inorganic and organic carbon, and suspended sediments were determined gravimetrically. Large volume (100 L) samples were ultrafiltered in the lab to separate the coarse (>63 µm) fine (0.1-63 µm) and ultradissolved (1-100 nm) organic matter. These fractions were analyzed for C and N concentrations and isotopic signatures, and organic matter fractions were further analyzed with the alkaline cupric oxide reaction procedure to geochemically characterize lignin and other biopolymers. Resulting interpretations include Bernandes et al. (in review), Krusche et al (in review), Ballester et al. (in review)

  3. Analytical. We have conducted a number of experimental studies designed to aid in our interpretation of field observations. Although over 90% of organic matter (OM) transported by rivers of the Amazon is either sorbed to fine minerals or has remained dissolved, the compositions of these two fractions are quite different. These biochemical and isotopic differences have long been used to infer differences in the sources, degradation and transport of dissolved versus fine particulate OM, which in turn define the carbon cycling characteristics of the watershed. However, because it is known that sorptive exchanges of OM are important between dissolved and particulate phases (Hedges et al. 1997), we asked whether sorptive processes might be masking or enhancing these compositional signatures that are used to infer other processes (diagenesis, transport or source). Our findings were startling. Preferential sorption of certain biomolecules appeared alone to be responsible for all biochemical signatures related to nitrogen (Aufdenkampe et al. 2001) and for stable carbon and nitrogen isotopic signatures (Aufdenkampe et al., in prep). Specific compositional differences between dissolved and fine particulate OM - even in the absence of differences in diagenesis, transport or source - has had a substantial impact on our interpretations of carbon cycling processes. For instance, with this knowledge, we can now confidently conclude that conversion of forest to pasture in the Ji-Parana watershed has already resulted in measurable changes in the composition of organic materials reaching aquatic systems (Bernardes et al., in review).

    In an effort to determine the possible influence of burning on the composition of organic matter in rivers, we analyzed samples for the presence of black carbon (BC), in the form of soot and char residues of combustion. We worked to adapt the commonly used cupric oxide oxidation method to the quantification of BC. This technique separates and quantifies certain molecules that may be used as molecular tracers. Using a set of charcoals prepared in the laboratory, we identified a suite of 21 molecules that are yielded by charcoals, but absent from the source wood and a set of potentially interfering samples. We also found these molecules in samples of natural charcoals and several sediment samples. Recently, however, we found another significant source of these molecules that may interfere with their use as molecular tracers for charcoal. Work is continuing to determine the feasibility of this methodology.

    In another field-based study, we employed radiocarbon dating techniques to constrain the characteristic turnover times of organic matter fractions within rivers of the Amazon. Samples collected as part of the LBA-funded field campaigns were analyzed at the Center for Accelerator Mass Spectrometry (CAMS), Lawrence Livermore National Laboratories through a CAMS mini-grant. By showing that the average age of riverine organic and inorganic carbon decreases downstream (Aufdenkampe et al., in prep; Mayorga et al. in prep.), these results have provided firm evidence of the magnitude of carbon flows through rivers to the atmosphere as hypothesized by Richey et al. (in press). These samples (over 100) will dramatically increase the number of D14C samples available, with concomitant increase in understanding.

  4. In situ Measurement and Analysis of Trace Gas Fluxes. In our original proposal we hypothesized that the role of wetlands in the overall carbon cycle of the Amazon may be very important, but is very underdetermined. Hence we launched two sets of complementary activities. The first was direct trace gas measurements using floating and embedded chambers at a study site on the Urupá River near the confluence with the Ji-Paraná River (K. Balster, thesis in progress). The farm at this site has a wide range of landcover types as well as different land uses. A transect from the river bank to the edge of the water was sampled from late March to early June 2001 during the receding water stage. The 500 gas samples from the 5 flooded forest sites, 2 flooded pasture sites and 7 dry sites are being analyzed for methane and carbon dioxide concentrations. Flasks for isotope analysis were also taken from the sampling sites. Preliminary results suggest methane fluxes are positively correlated with water depth.

  5. Analysis of Existing pCO2 Data and Floodplain Inundation Patterns. Our second activity related to gas fluxes was to assess what the potential role of periodically-inundated environments may be in the overall basin C cycle by computing the evasion(outgassing) of CO2 as a function of the seasonality of inundation of both large-scale (>100 m in width, detectable by radar) and small scale (100 m in width) hydrographic environments (Richey et al, 2002). This work has the startling conclusion that outgassing is much greater than previously thought, and may be of comparable magnitude to the lower estimates of sequestration. If so, this will change how we think about land-water coupling in the humid tropics.

  6. Model Integration. To provide overall coherence and integration of the sources, transport and fate of carbon, we started the development of the "River Basin Organic Matter and Biogeochemistry Synthesis" model (ROMBUS, Aufdenkampe and Mayorga, in prep). The overall goal of this model development is to predict the concentration and flux of dissolved and particulate carbon in rivers throughout a tropical river basin over an annual cycle, when coupled to the hydrologic model. The model partitions the carbon reservoirs within a river into measurable dissolved and particulate organic carbon pools (DOC and POC, respectively; each exhibit distinct dynamics and compositional traits that hold over a very broad range of geological, hydrological and climatic conditions). In contrast, most land-surface carbon models divide OM into unmeasurable (conceptual) pools with characteristic turnover times that are difficult to verify (e.g., the CENTURY). The ROMBUS design is more consistent with newer initiatives in the soil carbon modeling community. Unique to ROMBUS is the explicit parameterization of organo-mineral interactions, which are increasingly acknowledged as fundamental to OM stabilization.

Training and Education

  1. University Cooperation in Rondonia. We have established cooperation with the Lutheran University (ULBRA) and the Federal University of Rondonia (UNIR) at the city of Ji-Paran·. Groups of students and professors from these universities are being trained in the techniques of field sampling and analysis. Additionally, the students receive the necessary literature and explanations about the biogeochemistry of river basins, and are prepared to enroll in graduation programs. Nei Kavaguichi Leite (UNIR) is an example of this setup. He was trained in the field for a year, and is now a MSc student at CENA. Upon finishing his MSc, he will return to UNIR, where he will be responsible for the new chemistry laboratory, which we are installing there.
  2. Students Supported
    • Aufdenkampe, Anthony. Univ. of Washington. Ph.D. USA. The Role of Sorptive Processes in the Organic Carbon and Nitrogen Cycles of the Amazon River Basin.
    • Ballester, Maria Victoria Ramos. CENA. Post-doctoral. Chilean (Permanent resident in Brazil). From patterns to process: land use and land cover changes and its effects on the biogeochemistry of surface waters of tropical rivers in western Amazônia (Ji-Paran·, Rondônia).
    • Balster, Kellie. Univ. of Washington. Masters. USA. Trace gas emissions from a periodically inundated reach of the Urupa River.
    • Bernandes, Marcelo. CENA. Post-doctoral. Brazilian. Organic matter composition of rivers of the Ji-Paraná basin.
    • Bolson, Marcos Alexandre. UNIR-JP. Undergraduate Intern (iniciação científica). Brazilian. Carbon dynamics in two rivers of Rondônia with different land use/cover.
    • Bonelle, Nilton. ULBRA-JP, Brazilian, PhD. The biogeochemistry of the Urupá River (Rondônia, Brazil).
    • Coburn, Rebekkah. Univ. of Washington and CENA, undergraduate intern at CENA-USP from UW. Canadian. Land use and Land cover mapping of the Ji-Paraná river basin.
    • Gouveia Neto, Sergio Candido de. UNIR-JP. Undergraduate Intern (iniciação científica). Brazilian. Temporal variability of the water quality of two rivers of Rondônia with different land use/cover.
    • Hanada, Lais. CENA (Ecologia de Agroecossitemas). Masters. Brazilian. Land use and land cover changes at the agricultural frontier in western Amazônia, the Ji-Paraná River Basin, Rondônia.
    • Leite, Nei Kavaguichi. CENA (Centro de Energia Nuclear na Agricultura). MSc. Brazilian. Land use/cover changes in the Ji-Paraná basin and its effects on the river biogeochemistry.
    • Macedo, Gelson de, ULBRA-JP. PhD. Brazilian. Biogeochemical effects of land use/cover changes at the interface land-water in the Urupá river, Ji-Paraná, RO.
    • Marcondes, Renata, CENA (Economics). Undergraduate Thesis. Brazilian. Occupation and land use in the Ji-Paraná River basin (Rondônia): socio-economic and agricultural survey.
    • Remington, Sonya. Univ. of Washington. Masters. USA. Carbon transport through the Ji-parana River system, using the VIC hydrology model.
    • Toledo, Andre Marcondes, CENA. Ph.D. Brazilian. Environmental diagnosis of the Ji-Paraná river basin, Rondônia.
    • Victoria, Daniel de Castro, CENA (Agronomy). Undergraduate thesis. Brazilian. Using GIS and the Thornthwaite hydric balance to estimate the potential evapotranspiration in the Ji-Paraná river basin (RO).