Understanding the Carbon Balance of Amazon Forests


Understanding the Carbon Balance of Amazon Forests


Amazonia has the world’s most extensive forests, with the largest reservoir of above-ground organic carbon - around 90 Pg C in live trees alone - and the most species. They are under strong human pressure through logging, conversion and exploitation of resources. They face a warming climate and a changing atmosphere. Because of the vast scale of Amazonia, these factors have the potential to significantly modify the global atmospheric greenhouse gas burden (CO2, CH4), the earth’s climate, and the overall biodiversity of the planet. Some scenarios suggest catastrophic release of carbon from Amazon soil and vegetation this century, accelerating climate change globally even before accounting for the impacts of deforestation.

While the role of Amazonia in the carbon cycle is clearly of global importance, its behavior is contentious, even for that portion not undergoing rapid land use change. Understanding the current carbon balance of the whole system is critical to determine its role in either slowing or accelerating climate change through the 21st century. In particular, there is a need to:

  1. quantify the dominant fluxes into and out of main carbon pools in biomass and soil,
  2. the environmental controls of these fluxes,
  3. establish the infrastructure - plots, methods, expertise - to allow long-term monitoring across the vast region, and
  4. make these data freely available after sufficient quality control.


This proposal will help the only group currently active in integrating on-the-ground primary forest monitoring across the whole Amazon basin (RAINFOR) to tackle these urgent needs. Our overall goal is to determine the current carbon balance of Amazon forests, together with the associated fluxes and their sensitivity to soil and climate variability. A subsidiary goal is to improve the scientific infrastructure for future monitoring of Amazon biomass and soil carbon. The project is led by the University of Leeds and also has INPA (Brazilian Institute for Amazonian Research) as a major partner, and 30 other institutional partners world-wide. The Ecosystem Dynamics group at OUCE is specifically responsible for establishing a network of 16 intensive RAINFOR monitoring plots at 6 sites across the Amazon (in Brazil, Bolivia and Peru). At these plots, comprehensive monthly surveys of all key carbon stocks and fluxes will (1) Provide definitive baseline estimates of current forest carbon storage, and (2) Track ongoing changes in forest carbon cycling.

Project Progress & Outputs

Selected Sites & Research Fellows

See the RAINFOR Partners and Map pages for a quick summary of where and with whom we work

  1. Caxiuanã, Brazil (4 plots, 1 on Oxisol soil, 1 on a Terra Preta Anthrosol soil, and 2 plots constituting the Drought Experiment and Control both on Oxisol soils). Intensive monitoring of the Drought Experiment has continued, with a few gaps, since 2000, while the other plots were established in 2004. The Oxisol soil typifies the infertile, slow-growing but high biomass forests of eastern Amazonia, whereas the Terra Preta and drought experiment plots give us important information on how soil management and future climate change respectively could affect soil and forest C dynamics. Mauricio da Costa is coordinating carbon cycle measurements at the site, as part of his doctorate at UFPA (Partners: Antonio Lola Da Costa at UFPA & Samuel Soares Almeida at MPEG).
  2. Tambopata, Peru (2 plots, 1 each on Ultisol and Oxisol soils). High-intensity measurements started in 2005. Together with collaborators we may also duplicate aspects of the methodology in the nearby station of Los Amigos. These sites typify the high dynamism, low biomass sites of western Amazonia. Felix Filio Farfan Amezquita is responsible for data collection and analysis at the site (Partners: Norma Salinas at UNSAAC, Nigel Pitman at WFU).
  3. Forests
  4. Andean Montane Forest Transect in the Kosñipata Valley, Peru (4 plots, 2 each at 3000m and 1500m altitude). A plot at each altitude was established in 2006. As part of the current project, new plots have been installed to assess spatial variation in carbon cycling within forests at the same altitude. These sites allow exploration of the different allocation of biomass and soil carbon in montane systems, and assessment of vulnerability to warming and drying. In addition to the 4 main plots we will occasionally recensus 9 others along an elevational transect Given the exceptional physical demands of the site we have hired two researchers to coordinate work at this site- Liliana Durand Baca and Walter Huaraca Huasco (Partners: Norma Salinas at UNSAAC, Miles Silman at WFU).
  5. Tanguro, Brazil (2 plots, 1 burnt every three years, 1 undisturbed). Set within a well established large-scale burning experiment running above-ground measurements since 2003. These plots provide crucial insights into natural forest dynamics at the dry southern extremity of the Amazon forest and interactions between carbon cycling and fire. Wanderley Rocha da Silva is the selected researcher running our program of measurements at this site (Partners: Oswaldo Carvalho at IPAM, Mike Coe at WHRC)
  6. Kenia, Bolivia (2 plots). Completely new plots were established in late 2008 as part of this project. This site occurs at the dry ecotone between forest and savanna, and likely to be sensitive to seasonal and interannual variations in climate. Alejandro Araujo Murakami is in charge of our measurements at this site. (Partners: Luzmila Arroyo at MHNNKM, Tim Killeen at CI).
  7. Allpahuayo, Peru (2 plots, 1 each on white sand and Ultisol soil). This site is located close to the wettest, most aseasonal, and biodiverse core of Amazonia, providing a suitable contrast to other locations. Gilberto Aguilar is running all our work at this site. (Partners: Abel Monteagudo and Rodolfo Vásquez at JBM, Fredy Ramirez at UNAP)



In November 2008, we ran a 10-day workshop for the selected Research Fellows at the University of Oxford, U.K. Participants received extensive training on the methods and equipments to be implemented at the fieldsites. Given the fieldwork focus of the previous workshop at Los Amigos (run by RAINFOR partners at Leeds), the Oxford workshop concentrated instead on developing a sound understanding of equipment functioning, data organization and data entry requirements. In addition, basic data analysis procedures were described and participants were encouraged to develop independent research projects which they could potentially conduct as part of their ongoing work at the field sites.

Fieldwork & Data Analysis

An intensive six-month period in the latter half of 2008 saw successful establishment of the infrastructure necessary to measure all key above- and below-ground carbon stocks and fluxes at all six sites. All hired Research Fellows are now successfully conducting monthly measurements. Already, some of the data received represent the most detailed and extensive datasets of their kind yet available for the tropics (folivory, necromass, partitioned soil respiration). In general, the data will enhance understanding of the sensitivity of Amazon forest carbon storage to variation in soil, climate and fire. In the process we are building up a cadre of highly skilled local young scientists to ensure that the long-term capacity for ecological monitoring exists in the region. Together with these researchers, we intend to inform policy makers and climate models with accurate, integrated, on-the-ground information. As a first step towards this goal we have made a variety of documents available online (see below). After a short period of data organization and error checking we plan to make all of our datasets accessible online, with certain restrictions (see the RAINFOR ethical data and publication policy). Meanwhile we have been developing the necessary collaborative links to consider expanding the network to include other plots in South America (Brazil, French Guyana), Africa (Ghana, Gabon) and Asia (Borneo).

Online Materials

These documents are still in a state of revision. If you spot any mistakes or have any general queries please contact .

  1. RAINFOR GEM Intensive Plots Manual (Version 2.2, Jun 2012)
  2. Project Methods Manual (Version 1, 2009, now superseded by above v2.2). All measurements at the sites follow a standardized protocol detailed in this manual. Portions of the manual have been adapted from existing RAINFOR manuals.

Moore Foundation Project Datasets Summary

SiteDataPlotLatitude, Longitude
Caxiuaña, BrazilRespiration, BiomassCaxiuaña Plot A-1.736797, -51.46184
Caxiuaña Plot B-1.735681, -51.46198
Caxiuaña Torre-1.719797, -51.45818
Caxiuaña Terra Preta-1.856006, -51.43519
Kenia, BoliviaRespiration, BiomassKenia A-16.01583, -62.73019
Kenia B-16.00903, -62.74258
Kosnipata, PeruRespirationKosnipata ESP-13.1903, -71.597
Kosnipata SPI-13.0472, -71.5425
Kosnipata SPII-13.0491, -71.5369
Kosnipata WAY-13.1903, -71.587
Tambopata, PeruBiomassTambopata Plot 3-12.82987, -69.27099
Tambopata Plot 4-12.83871, -69.29660
Tanguro, BrazilRespirationTanguro-13.0765, -52.3858

Data Summary

  1. Respiration data - All spreadsheets contain information on plot, date and time, CO2 reference value and calculated respiration flux (g CO2 m-2 hr-1). Regression relationship and slope values have also been calculated for some of the plots.
    • Caxiuaña, Plots A, B, Torre & Terra Preta: stem respiration August & October 2009. Soil partitioning experiment, August 2009-November 2009.
    • Tanguro, Plots A & C: stem respiration, June 2009. Total soil, soil partitioning experiment, January 2009-August 2009. Soil control, August 2009.
    • Kenia, Plots A & B: stem respiration, February 2009. Soil control, soil partitioning experiment, total soil data, November 2008-March 2009.
    • Kosnipata, 4 plots, Esperenza (ESP), San Pedro 1750 m (SPI), San Pedro 1500 m (SPII), Wayquecha (WAY): soil control & soil partitioning experiment, January 2009-February 2010. Deadwood, October 2009-January 2010. Total soil, January-February 2010.
  2. Biomass data - Spreadsheets contain live stem identification information, diameter measurements and growth increment (dendrometer) data for trees and lianas. Mortality and recruitment data are included for Tambopata.
    • Caxiuaña, Plots A, B, Torre & Terra Preta: stem growth, January 2009-November 2009.
    • Kenia, Plots A & B: stem growth, November 2008-February 2010.
    • Tambopata, Plots 3 & 4: diameter measurements & stem growth (dendrometer data), July 2007-January 2009.

Project details