The TREES Project:
Long-term annual performance of tropical rain forest trees

Deborah A. Clark and Susan G. Letcher
(November 2015)


The TREES Project is the longest-running annual census of tree performance and microsite condition anywhere in the tropics. Because some tree species in tropical wet forests do not produce reliable annual rings, the only way to monitor their growth is to carefully measure it every year. The TREES Project was launched in 1983 by Deborah A. Clark and David B. Clark during their tenure as station directors at La Selva Biological Station. Initially the project focused on the demography of six tree species, documenting their growth and estimating how long they take to pass from one size class to the next and eventually reach the canopy. After a decade of research, emergent patterns suggested that these six species, despite their diversity, tended to show similar interannual growth fluctuations: the trees shared good growth years and bad growth years. This observation sparked a shift in the project’s focus from population ecology toward ecosystem ecology, documenting how interannual changes in climate affect tree growth and demography, which in turn impacts landscape-scale carbon storage. The list of species surveyed annually was expanded from six to nine in 1988 in order to incorporate a greater range of life history traits, and a tenth species was added in 1998. In 1996, the Clarks and Steven F. Oberbauer launched the CARNONO project , an associated research initiative investigating landscape-scale carbon storage in this forest.

Results from the TREES project and CARBONO have revealed a strong link between climate and forest performance. Trees tend to perform poorly in years with high nighttime temperatures, suggesting that temperature-driven increases in respiration rates are negatively affecting their carbon balance. High daytime vapor pressure deficit in the dry season (i.e., low relative humidity in the air relative to the levels inside leaf tissues) is an additional stressor that can inhibit tree growth. Climate change in recent decades has brought warmer and drier conditions to La Selva, and as these trends continue, this extremely well-studied tree population will provide deeper insights into the long-term prognosis for tropical forests. The TREES Project reveals the importance of careful, repeated, long-term measurements in elucidating individual-level and ecosystem-level responses of forests to global change.

The initial funding for the TREES Project came from the NSF-DEB program, with additional funding from the Andrew W. Mellon Foundation and the Organization for Tropical Studies. Since 2001, the project has been continuously funded by the NSF LTREB program. In 2012, Susan G. Letcher joined the project as co-PI.

Go to top


The TREES Project is an annual census of approximately 2500 individual trees of ten species, ranging in size from 50 cm height to nearly 2 m diameter. The trees are distributed across c. 800 ha of old-growth tropical forest in the eastern half of La Selva Biological Station. New individuals are added to the census each year to make up for mortality, giving a total sample size of approximately 4300 trees (as of February 2015).

Each year, the trees are measured according to a standard protocol. Calibrations at the beginning of each census and periodically throughout the survey ensure data quality. Since the beginning of the project, all of the fieldwork has been conducted by the same five individuals (paraforesters Leonel Campos, William Miranda, and Gerardo Vega or project PIs Deborah A. Clark and David B. Clark). This highly curated dataset provides an extremely reliable record of tropical tree growth and demography.

TREES project PIs also work closely with the La Selva laboratory manager to generate a consistent, cross-checked record of meteorological data from the site. In 2007, Deborah Clark generated a cross-checked record of daily rainfall since 1957 and daily temperature since 1980, based on La Selva data and regional meteorological stations. Since 2007, she has provided annual cross-validated meteorological records for the La Selva web site (temperature, rainfall, and solar radiation). In 2015, Susan Letcher automated the cross-validation in an R script.

Core annual-subannual measurements

Trees in the dataset range in size from small saplings 50 cm height to canopy emergent trees more than 1 m in diameter. Six species have been measured since 1983 ( Balizia elegans, Dipteryx panamensis, Hieronyma alchorneoides, Hymenolobium mesoamericanum, Lecythis ampla, and Minquartia guianensis). Three fast-growing species were added in 1988 (Cecropia insignis, Cecropia obtusifolia, and Simarouba amara), and the most abundant species at La Selva (Pentaclethra macroloba) was added in 1998. For each tree, the following observations are recorded:

  • Canopy position
    The position of the tree’s crown is assessed on a categorical scale based on two independent measurements, with values as follows: no light (1.0, NLD - ninguna luz directa) = no direct light from the side nor from above. Indirect light (but no direct overhead light), with three levels of intensity: 1.5 (ALL-B - alguna luz del lado, bajo) = low indirect light; 2.0 (ALL-M - alguna luz del lado, mediano) = medium indirect light; 2.5 (ALL-A - alguna luz del lado, alto) = high indirect light (no direct overhead illumination, but strong light from the side). Some overhead light (3.0, ALA - alguna luz de arriba) = 10-90% of the canopy with direct overhead light. Full overhead light (4.0, PLA - plena luz de arriba) = more than 90% of the canopy with direct vertical light. Fully exposed canopy (5.0, CCE - copa completamente expuesta) = more than 90% of the canopy completely exposed to overhead light; i.e., completely free of obstacles in a 45° cone around the crown.
  • Habitat
    The area surrounding the plant is classified based on forest structure: D (dosel) = at the plant, the forest canopy is at the level of mature forest in the surrounding area. For plants that are not yet in the canopy, this habitat category indicates that they are surrounded by mature forest; for trees that are of canopy height, this category indicates that they are at the level of the surrounding canopy. SD (subdosel) = canopy height reduced by perturbation (but greater than 2 m; see CL and SD/CL, below). CL (claro) = an area of at least 9 m2 that is free of vegetation from 2 m height up to the open sky. SD/CL refers to a plant that is >2 m height but not at the level of the canopy, found in an area of vegetation <2 m tall – that is, an isolated plant in a recent clearing. BS (bosque secundario) = secondary forest; a handful of trees from the original census are in secondary forest, though no new trees are marked in this habitat. CM (camino) = trees that are affected by a trail (again, no new trees are marked in this habitat).
  • Stem Condition
    The condition of each stem is assessed among the following categories: D (dañado) = with one main stem that presents an abrupt change in diameter of ≥25% of the stem diameter at that point. M (multiple) = with two or more main stems of nearly the same size. H (horizontal) = horizontal, parallel to the soil, without resprouts. HR (horizontal con rebrotes) = horizontal with vertical resprouts. RA (roto arriba) = one main stem that has lost its top, with resprouts emerging from below the damaged point. N (normal) = not fitting any of these other categories. Hollow stems are noted in the field sheets. .
  • Height
    All trees ≤ 16 m height are measured to the nearest cm. Height is measured vertically from the highest side of the tree’s base to the highest meristem (up to 1.5 m) or to the highest leaf (1.5 – 16 m), using a rigid tape for small trees and a telescoping pole for larger trees..
  • Diameter
    With the exception of very large trees with irregular stems or extremely high buttresses, all trees are measured for diameter at precisely specified stem locations. Very small stems are measured at 20 to 40 cm above the base, at a point marked with paint. Larger stems are measured at 1.3 m above the base, and very large stems are measured above the buttresses with up to two 3 m ladders on each side of the tree. When the measurement point is changed, the diameter is taken at the old and new measurement points for continuity unless the original measurement point has been lost. .
  • Topographic position 
    When a new tree is added to the dataset, it is referenced to existing trees, to the La Selva grid posts, or to a trail marker. In addition, the slope and aspect are measured at the base of the tree.

Go to Top

Results to Date

The TREES project is a globally unique long-term record of tree growth and survivorship of diverse species in a lowland wet forest. The high quality and long duration of the measurements make this data set ideal for investigating tree demography and individual trees’ responses to changes in microsite conditions and climate. As demonstrated by an analysis of different-length segments of the TREES record (Clark and Clark 2011), a long time series is required to accurately detect the sensitivities of tree growth and mortality to climatic factors.

Initially established to study community ecology, the TREES project evolved into an important bellwether for the impacts of climate change on tropical forests. In the first decade of the project, the TREES project showed that tropical forest trees may take centuries to reach the canopy. Saplings of shade-tolerant species can persist for at least a decade in low-light conditions and may go for years without any appreciable growth. Small increases in light availability lead to strong growth responses. Physical damage is a pervasive and important stressor. The first decade of study also revealed a coupling of tree growth and climate, with lower wood production in years with hotter nighttime temperatures. Models and subsequent observations have confirmed the trend in these species, and the CARNONO project has confirmed that wood production in many species follows the same pattern.

Other findings from the TREES project have involved the distribution of species with respect to edaphic conditions; the abundance of lianas and hemiepiphytes and their impacts on tree growth and survival; the landscape-scale distribution of very large trees (≥60 cm diameter); the distribution of epiphytes in large tree crowns; the dynamics of fine root biomass; and other aspects of these species’ contributions to the ecosystem.

Go to top

Current Status

From 1983 - 2010 D.A. Clark and D.B. Clark lived on-site and managed the project through daily interactions with the three long-term Costa Rican field technicians participating in the project. In 2010 the Clarks moved to the United States. Since then the daily meetings with technicians have been conducted with Skype video calls. At the end of each field day the technicians scan the day's field sheets and email them to the Clarks, who review them before the next day's meeting at the start of the work day. In addition the Clarks make 2-3 trips a year to La Selva to continue technician training, update equipment inventories, and address any issues that can't easily be worked on in video calls. The combination of daily contact supplemented by regular physical visits is working smoothly.

In 2012, Susan G. Letcher joined the project. She is managing the 2015 field census and forging links with modelers and researchers from other disciplines to amplify the value of this unique data set.

Go to top


[under construction]

Go to top

Photographs and Maps

[under construction]

Go to top

Project Staff

  • Debora A. Deborah A. Clark, Principal Investigator
    Contact information:
    Mailing Address: 1384 Lindenwood Grove, Colorado Springs, CO 80907 USA
    Telephone: (719) 310-4168
    E-mail: deborahanneclark @
    Web page:
    Profile: Deborah Clark is one of the original PIs of the TREES Project and has been studying the ecology of Costa Rican forests since 1980. In 2010 she moved from La Selva to Colorado and now manages the TREES and CARBONO Project activities through daily skype conferences and with regular visits to Costa Rica. When not doing tropical biology, she can be found hiking the backcountry of the Colorado Front Range with CO-PI David Clark.

    Press to see Deborah A. Clark´s CV

    Press to see Deborah A. Clark´s Photo
  • Susan G. Letcher, Principal Investigator
    Contact information:
    Mailing Address: 735 Anderson Hill Road, Purchase, NY 10577 USA
    Telephone: (914)-251-6646
    E-mail:  This e-mail address is being protected from spam bots, you need JavaScript enabled to view it
    Web page:
    Profile:  Susan Letcher is a terrestrial ecologist and environmental scientist whose research centers on anthropogenic impacts in tropical wet forests. She has been working in Costa Rica since 2003. She teaches environmental studies at Purchase College (SUNY). She has been involved in the TREES project since 2012.

    Press to see Susan G. Letcher´s CV

    Press to see Susan G. Letcher´s Photo
  • David B. Clark, founding Principal Investigator
    Contact information:
    Mailing Address: 1384 Lindenwood Grove, Colorado Springs, CO 80907 USA
    Telephone: (719) 310-4180
    E-mail: dbclark50 @
    Web page:
    Profile:  David Clark is one of the original PIs of the TREES project and the CARBONO Project. He has been working in Costa Rican forests since 1980. His areas of research include tree demography, forest productivity and dynamics, the effects of global climate change on tropical forests, and the use of remotely-sensed data to scale results from plots to larger landscapes. In 2010 he moved from La Selva to Colorado, where he continues his research in CARBONO using daily video-calls to project staff combined with regular visits to Costa Rica. When not doing tropical biology, he can be found playing Celtic music on mandolin and fiddle or hiking the backcountry of the Colorado Mountains with his wife and CO-PI Deborah Clark.

    Press to see David B. Clark´s CV

    Press to see David B. Clark´s Photo

Go to top

Data and Metadata

  • Data use policy: under construction
  • Scanned field sheets


































  • Data tables:

    Final Database

  • Meteorological data and metadata: under construction

Go to top

web counter
web counter
Last Updated ( 01/31/17 )
Organization for Tropical Studies
Site powered by Joomla!