Skip to main content

Two decades of annual landscape-scale tree growth and dynamics in old-growth tropical rainforest in the CARBONO Project, La Selva Biological Station, 1997-2018

Cite this dataset

Clark, Deborah; Clark, David (2021). Two decades of annual landscape-scale tree growth and dynamics in old-growth tropical rainforest in the CARBONO Project, La Selva Biological Station, 1997-2018 [Dataset]. Dryad.


Here we present the complete data series from a 21-yr study of the annual growth and dynamics of trees, palms and lianas in the old-growth tropical rainforest at the La Selva Biological Station in Costa Rica. These observations were part of the CARBONO Project, a multidisciplinary team study of forest carbon cycling.  The project was designed to assess forest processes at the landscape scale by sampling with replication across the within-landscape edaphic heterogeneity typical of tropical forests.  Through more than two decades, forest growth and dynamics were assessed annually.  The annual time-step used in the CARBONO Project effectively captured forest responses to major disturbances and to interannual and climatic variation.  Annual measurements also enhanced the accuracy and long-term consistency of the data.  To our knowledge, the resulting records are unique for tropical forests, where the dominant approach to studying the dynamics of a given forest has been to use a single plot and multi-year inter-census intervals. To date these CARBONO Project data have revealed: multi-decadal forest stability in spite of the short-term changes in forest structure resulting from major natural disturbances (e.g., the 1997-1998 Strong El Niño, and the extreme windstorm of May 2018); the dynamics and importance of large trees; and the responses of a major component of ecosystem productivity,  aboveground wood production, to interannual and long-term climatic and atmospheric change.  These data have also contributed to many remote-sensing studies.

The data set consists of annual observations through the period 1997–2018 of the floristics, survival, recruitment, and growth of all woody stems (diameter > 10 cm) in a landscape-scale plot network.  At completion of the study, the data spanned 6705 individuals and 21 years. The data set is complete and has been through extensive internal checks for quality assurance.   Detailed data documentation and an emphasis on measurement repeatability were prioritized through the study.  The metadata include an extensive README file describing the data files and the methods, a document detailing the data management and qa/qc, and the scanned original field data-sheets for the 22 annual censuses.

We gratefully acknowledge the careful long-term field work and data entry and checking by paraforesters Leonel Campos Otoya and William Miranda Conejo.  Logistical support and the long-term protection of the La Selva reserve were provided by the Organization for Tropical Studies. The Ministerio de Ambiente y Energía of Costa Rica granted permits to carry out this study through the years of the study (most recently: Resolución No. 037-2018-ACCVC-PI).


A network of 18 0.5-ha forest inventory plots was sited across the upland old-growth landscape using the soil and topographic coverages of La Selva’s Geographic Information System.  The plots were sited using a random-block design.  The plot network sampled the three dominant edaphic conditions (younger-oxisol alluvial terrace, older-oxisol plateau, and older-oxisol slope), with replication (six plots in each edaphic type).

During plot establishment, all live individuals were mapped, tagged and identified.  In Sept.-Dec. 1997 and annually thereafter, all trees were censused for survival and their diameter measured using standardized protocols, and the recruits to all plots (stems newly reaching 10 cm diameter) were identified, tagged, and measured.  At the outset of each year’s census, a large sample of trees were measured and then re-measured after multiple days to quantify the repeatability of the measurements (made by the same two paraforesters over the entire study).  The two project PIs supervised each annual census on a daily basis and, after data digitization by the paraforesters, carried out all data management, with an emphasis on data qa/qc throughout the study. 


Usage notes

The field methods and all variables and calculated data in the data files are explained in detail in the included README file.


National Science Foundation, Award: DEB 9629245

United States Department of Energy, Award: TECO program, 1996-2003

National Science Foundation, Award: LTREB 0841872

National Science Foundation, Award: LTREB 1357177

Andrew W. Mellon Foundation

Conservation International, Award: Award to D.B. Clark