PROJECT SUMMARY

Tropical rain forests (TRFs) contribute substantially to the global carbon cycle, accounting for ~40% of terrestrial net primary production (NPP), ~60% of forest biomass, and ~27% of carbon stored in forest soils. Contrary to the general expectation of a constant climate in TRFs, annual variability in climate is significant. Recent studies show that interannual variation in carbon cycling in TRFs is large enough to be detected as anomalies in atmospheric CO2 concentration that are correlated with annual variation in tree growth. However, knowledge of TRF carbon cycling is poor; sizes of pools and fluxes are uncertain, and environmental controls over fluxes are very poorly known. While the few eddy covariance studies in TRFs suggest that they act as significant carbon sinks, this method may underestimate ecosystem respiration overestimating the sink strength. Further, a warmer, drier climate may yield decreased NPP and increased respiration. Resolving these issues will require unbiased, landscape-scale characterization of the structure and function of TRF canopies, assessment of carbon budgets and annual fluxes using several different methods, and an understanding of the causes of interannual variability.
This project will address key questions about carbon cycling in a TRF that will reduce uncertainties about carbon fluxes and their controls: 1) How is photosynthetic and respiratory capacity distributed within the canopy? 2) How are canopy structure and function distributed across the landscape and how does this variability contribute to ecosystem photosynthesis? 3) How do ecosystem photosynthesis and respiration vary with the environment? 4) Do independent estimates of annual net carbon exchange agree? 5) How does environment affect photosynthesis and plant respiration to alter tree growth? 6) How will carbon cycling change in a TRF if climate changes? We will answer these questions for a wet TRF at La Selva, Costa Rica, by combining field measurements of canopy structure and function, existing ecosystem flux data, a long term record of annual tree growth and aboveground net primary production, and canopy and ecosystem models. We will accomplish this by 1) measurement of forest structure, photosynthesis, and respiration on vertical transects (forest floor to the canopy surface) stratified across the landscape; 2) extrapolation of these chamber samples (and ongoing soil respiration data) to the landscape; 3) comparison of the extrapolated fluxes and their environmental responses with three years of existing measurements of whole ecosystem fluxes as estimated by eddy covariance in this forest; 4) comparison of three independent estimates of annual net carbon exchange; 5) using the best available data to calibrate the MBL-GEM model of ecosystem biogeochemistry; and 6) linking the short-term flux data with 20 years of annual tree growth and 8 years of annual NPP by using MBL-GEM to test hypotheses about the causes of interannual variability in growth and NPP. We hypothesize that changes in photosynthesis and the ratio of autotrophic respiration to photosynthesis underlie the annual variability in aboveground NPP, and that drought or storm-caused disturbance will decouple NPP and annual net carbon exchange by increasing heterotrophic respiration relative to photosynthesis.
This research will significantly improve current understanding of the structure of tropical forest canopies, the linkages between canopy structure and function, the size and controls over annual fluxes, and the causes of interannual variability in the carbon cycle in this biome. Our study will link physiologically based models with the longest extant record of annual variability in tree growth for TRFs, and will provide valuable data and methodology for predicting the effect of changes in climate on the carbon cycle in TRFs. Finally, our study will provide data and methodology to allow the assessment of the structure and function of TRFs over the landscape using remote sensing. The La Selva Biological Station is the best-characterized tropical wet forest landscape in the world in terms of GIS-based information (elevation, soil carbon and nutrients, vegetation height), annually measured ANPP, several-year eddy covariance data, a long-term record of tree growth, and remote-sensed data. The P.I.'s are an experienced multidisciplinary team with complementary expertise in forest ecology, ecophysiology, ecological modeling, and remote sensing.

Project Components

1) Quantification of biomass-weighted photosynthetic and respiration rates in an old-growth tropical rain forest, based on highly replicated direct measurements; assessment of the impacts of within-landscape variation in canopy structure and leaf physiology on these processes.

2) Scaled-up estimates of total canopy photosynthesis (GPP) and total ecosystem respiration (Re) and their response to short-term climate variation for a tropical rain forest.

3) Assessment of the adequacy of modeled responses of forest C uptake to climate variability by comparing MAESTRO estimates to GPP estimates derived from eddy covariance measurements and estimates of soil and wood respiration.

4) Inter-comparison of three independent approaches to estimating GPP, Re, and NEE for tropical rain forest.

5) Exploration of the climatic drivers underlying our 16-yr record of interannual variation in tree growth using the MBL-GEM ecosystem process model.

Unique Canopy Access Opportunity in Lowland Forest in Costa Rica (note that the field campaign is nearing completion - finishing March 2005).
The CARBONO research group (www.Carbono.org), a multidisciplinary team studying the stocks and flows of carbon in tropical rainforest wishes to announce a unique opportunity for collaboration in canopy research. Funding from the National Science foundation will support a field campaign of approximately 45 vertical canopy transects at the La Selva Biological Station in Costa Rica (www.ots.ac.cr/en/laselva/) beginning in June 2003. The vertical transects will be permitted by erecting a walk-up scaffolding tower at each site for approximately 2 weeks. The primary project goal of the Towers project (www.carbono.fiu.edu/tower) will be to conduct vertical measurements of forest structure and physiology. However, we recognize the unique opportunity presented by access to vertical transects from understory to above-canopy in tropical rain forest to researchers outside of our field of interest. We will welcome collaborators with their own funding to conduct research off of the tower that does not interfere with our research. Examples might include study of canopy epiphytes, secondary compounds along vertical transects, mammal behavior (bats). Unfortunately, due to CARBONO personnel limitations, we will be unable to collect samples for collaborators. The field campaign is scheduled to last 20 months. The vertical transect sites will be randomly located in old-growth forest stratified for canopy height.

Interested parties should email Steve Oberbauer at Florida International University (oberbaue@fiu.edu). Research and collecting permits are required to conduct research at La Selva and can be obtained with help from OTS (www.ots.ac.cr). For samples to be taken out of county, export permits are also required.

Project Personnel

Elizabeth Braker Ph.D.- Senior Investigator-Occidental College

Molly Cavaleri, M.S. - -Ph.D. student- Colorado State University - cavaleri@cnr.colostate.edu

Deborah A. Clark Ph.D.- - Co-Pi- University of Missouri-St. Louis - daclark@sloth.ots.ac.cr

David B. Clark Ph.D.- - Co-Pi-University of Missouri-St. Louis - dbclark@sloth.ots.ac.cr - Co-Pi

Christian J. Caruso- REU student -Montana State University

Andrea Garcia - M.S. student -Florida International University -andrea.garcia@fiu.edu

Carlos Herra - Tower field crewmember- La Selva Biological Station

Elias Herra - Tower field crewmember-La Selva Biological Station

Enoch Astua Herra - Tower field crewmember- La Selva Biological Station

Dan Hoffman- REU student- Occidental College

Damaris Jarquin - Sample processing technician -La Selva Biological Station

Bonnie L. Kwiatkowski - Ecosystem modeler - Ecosystem Center, Marine Biological Center

Juan Miranda - Tower field crewmember- La Selva Biological Station

Steven F. Oberbauer Ph.D. - P.I. - Florida International University - oberbaue@fiu.edu

Harlyn Ordonez, M.S. - Tower ecophysiologist -La Selva Biological Station - harlyn@sloth.ots.ac.cr

Paulo Olivas - Tower program coordinator - La Selva Biological Station - polivas@sloth.ots.ac.cr

Vinicio Paniagua - Tower field crewmember- La Selva Biological Station

Michael G. Ryan Ph.D.- Co-Pi-U.S. Forest Service Rocky Mountain Research Station - mgryan@fs.fed.us

Ed Rastetter Ph.D. - Senior investigator - Ecosystem Center, Marine Biological Center - erastett@mbl.edu

Roy Blanco Salazer - Tower field crewmember- La Selva Biological Station

Sampling Plan

Initial site selection: We stratified the La Selva landscape into 9 types of cells with ca.
equal numbers using 0-10 cm soil depth P concentration and DEM-derived slope
coverages, 10 x 10 m cells (Figure A). We used a random number generator to select 8 cells
in each of the 9 conditions. We buffered trails by 25 m, avoided active permanent
plots and the La Selva Ecological Reserve (but not the futureTEAM plot), and points that fell on
grid posts or on swamp or recent alluvial soil. We picked the 6 cells in each
condition closest to trails, in order of closeness.

Figure A. Classification of forest into soil phosphorus and slope categories for sampling plan. High (pink), medium(green), and low (blue) phosphorus levels. Different shades within a color indicate high, medium, and low slopes. White areas are swamps or second growth. Black lines are improved trails. (total area of image is approximately 700 ha.

For each potential site: We located each plot by its NW corner in LS
coordinates. We measured from the nearest grid post to this point. We staked this
corner. From there, we used a tape measure and clinometer to locate in the S.
quadrant a point 3 m away and a 0* slope. We staked this SW corner, and
measured the azimuth from the NW to the SW corner. From the NW and SW corners
we measured 3 m at (azimuth-90*) and staked the NE and SE corners. We then measured the
slopes in degrees from the NE to NW and SE to SW corners. In the plot,
we count and measure stem diameter of all stems >=5 cm, including all stems
that have any part at all of their trunk within the plot. We counted and
measured (or estimate, if buttressed) stem diameter of all stems >=10 cm
within 1 m of the plot, and measure the perpendicular distance to the plot
boundary for each of these. We used a laser rangefinder and clinometer to
establish the position of vegetation 90* above the plot, and measured the height
of the highest vegetation above the plot. We used the clinometer to locate all
branches >=10 cm diameter that cross the vertical projection of the plot,
and estimated their diameter. We then noted unusual conditions within the plot (lots
of CWD, tip-up, Atta nest, water seep or seasonal stream, swampy soil.) and
noted if chainsaw will be necessary to clear CWD.

Of the nearly 100 potential sites, we selected approximately 54 suitable sites (Figure D).

Figure D. Tower sample locations.

This material is based upon work supported by the National Science Foundation. Any opinions, findings, conclusions, or recommendations expressed in the material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.