Skip to main content

Data from: Using large-scale tropical dry forest restoration to test successional theory

Cite this dataset

Werden, Leland et al. (2020). Data from: Using large-scale tropical dry forest restoration to test successional theory [Dataset]. Dryad.


Microclimatic conditions change dramatically as forests age and impose strong filters on community assembly during succession. Light availability is the most limiting environmental factor in tropical wet forest succession; by contrast, water availability is predicted to strongly influence tropical dry forest (TDF) successional dynamics. While mechanisms underlying TDF successional trajectories are not well understood, observational studies have demonstrated that TDF communities transition from being dominated by species with conservative traits to species with acquisitive traits, the opposite of tropical wet forest. Determining how functional traits predict TDF tree species’ responses to changing environmental conditions could elucidate mechanisms underlying tree performance during TDF succession. We implemented a 6-hectare restoration experiment on a degraded Vertisol in Costa Rica to determine (1) how TDF tree species with different resource-use strategies performed along a successional gradient; and (2) how ecophysiological functional traits correlated with tree performance in simulated successional stages. We used two management treatments to simulate distinct successional stages including: clearing all remnant vegetation (early-succession), or interplanting seedlings with no clearing (mid-succession). We crossed these two management treatments (cleared/interplanted) with two species mixes with different resource-use strategies (acquisitive/conservative) to examine their interaction. Overall seedling survival after two years was low, 15.1%-26.4% in the four resource-use strategy × management treatment combinations, and did not differ between the management treatments or resource-use strategy groups. However, seedling growth rates were dramatically higher for all species in the cleared treatment (year 1: 69.1% higher; year 2: 143.3% higher) and defined resource-use strategies had some capacity to explain seedling performance. Overall, ecophysiological traits were better predictors of species’ growth and survival than resource-use strategies defined by leaf and stem traits such as specific leaf area. Moreover, ecophysiological traits related to water-use had a stronger influence on seedling performance in the cleared, early-successional treatment, indicating that the influence of microclimatic conditions on tree survival and growth shifts predictably during TDF succession. Our findings suggest that ecophysiological traits should be explicitly considered to understand shifts in TDF functional composition during succession and that using these traits to design species mixes could greatly improve TDF restoration outcomes.


This archive contains all raw data files used to produce results in Werden et al. 2020. Using large-scale tropical dry forest restoration to test successional theory. Ecological Applications. This manuscript details the results from a 6-hectare tropical dry forest restoration experiment that was implemented at Estación Experimental Forestal Horizontes in Área de Conservación Guanacaste in northwestern Costa Rica (10.712N, 85.594W). Please see included readme file for detailed descriptions of data files. 

Usage notes

This archive contains: 

1. A ReadMe file for the data set which details what is contained in each of the following datasets: README_WerdenEtAl2020_EcologicalApplications.rtf

1. Data for all seedling surveys conducted in 2015-2017 for a 6-hectare restoration experiment at Estación Experimental Forestal Horizontes in Área de Conservación Guanacaste in northwestern Costa Rica. Every surviving seedling was measured during each of the four surveys: seedlingSurveys_WerdenEtAl2020_EcolApps.csv

2. Air and soil temperature data collected in August and September of 2015 in six restoration experiment plots. Measurements were taken at the center of each plot with iButtons (DS1921G-F5# Thermochron; Maxim Integrated, San Jose, CA, USA) simultaneously measuring air temperature (0.5 m off the ground), and soil temperature (5 cm depth): tempAirSoil_WerdenEtAl2020_EcolApps.csv

3. Volumetric soil moisture measurements made in the six restoration experiment plots in August 2015, September 2015, and July 2016, . Measurements were made in the top 0-5 cm of the mineral layer with a DeltaSM150 soil moisture sensor (Delta-T Devices, Burwell, UK): soilMoisture_WerdenEtAl2020_EcolApps.csv

4. Measurements of fine root stocks in the six restoration experiment plots collected in February 2018. Six root samples were collected a root corer (8 cm diameter, 15 cm depth) along two parallel transects (50 m apart) in each plot (6 plots × 2 transects per plot × 6 samples per transect = 72 root samples). We washed soil from roots, dried roots to constant mass at 60 C, and determined root dry mass (g) for each sample: rootStocks_WerdenEtAl2020_EcolApps.csv

5. Full-sun photosynthetically active radiation (PAR) measurements taken in the restoration experiment. Measurements were taken 1.5 m above the ground with an AccuPAR LP-80 Ceptometer (Decagon Devices, Washington, USA) in November, 2017: PAR_WerdenEtAl2020_EcolApps.csv

6. Forest inventory surveys of remnant vegetation in the three interplanted treatment plots. Trees >=10 cm diameter were measured and identified  For multi-stemmed trees we measured all stems >=10 cm in DBH for each individual: remnantVegetation_WerdenEtAl2020_EcolApps.csv


National Science Foundation, Award: 1600710

National Science Foundation, Award: 11-582

National Science Foundation, Award: DEB-1053237

Garden Club of America, Award: Award in Tropical Botany