Data from: Reassembly of a tropical rainforest: A new chronosequence in the Chocó tested with the recovery of tree attributes
Data files
Dec 12, 2024 version files 816.84 KB
-
analysis_reassembly.R
53.13 KB
-
cultivated_trees.csv
369 B
-
plots_reassembly.csv
51.31 KB
-
plots_reassembly.R
17.56 KB
-
README.md
5.63 KB
-
tree_data_measures.csv
156.35 KB
-
tree_data.csv
532.50 KB
Abstract
From hunting and foraging to clearing land for agriculture, humans modify forest biodiversity, landscapes, and climate. Forests constantly undergo disturbance–recovery dynamics and understanding them is a major objective of ecologists and conservationists. Chronosequences are a useful tool for understanding global restoration efforts. They represent a space-for-time substitution approach suited for the quantification of the resistance of ecosystem properties to withstand disturbance and the resilience of these properties until reaching pre-disturbance levels. Here we introduce a newly established chronosequence with 62 plots (50 ✕ 50 m) in active cacao plantations and pastures, early and late regeneration, and old-growth forests in the extremely wet Chocó rainforest. Plots were located across a 200 km2 area, with a total area of 95 km2 within a 1-km radius. Our chronosequence covers the largest total area of plots compared to others in the Neotropics with 15.5 ha. Plots ranged from 159–615 masl in a forested landscape with 74 ± 2.8 % forest cover within a 1-km radius including substantial old-growth forest cover. Land-use legacy and regeneration time were not confounded by elevation. We tested how six forest structure variables (maximum tree height and DBH, basal area, number of stems, vertical vegetation heterogeneity, and light availability), aboveground biomass (AGB), and rarefied tree species richness change along our chronosequence. Forest structure variables, AGB, and tree species richness increased with regeneration time and are predicted to reach similar levels to those in old-growth forests after ca. 30–116, 206, and 76 yrs, respectively. Compared to previous work in the Neotropics, old-growth forests in Canandé accumulate high AGB that takes one of the largest time spans reported until total recovery. Our chronosequence comprises one of the largest tree species pools, covers the largest total area of regenerating and old-growth forests, and has higher forest cover than other Neotropical chronosequences. Hence, our chronosequence can be used to determine the time for recovery and stability (resistance and resilience) of different taxa and ecosystem functions, including species interaction networks. This integrative effort will ultimately help to understand how one of the most diverse forests on the planet recovers from large-scale disturbances.
README: Reassembly of a tropical rainforest: A new chronosequence in the Chocó tested with the recovery of tree attributes
https://doi.org/10.5061/dryad.c59zw3rgb
Description of the data and file structure
This dataset is associated with Sebastián Escobar, Felicity L. Newell, María-José Endara, Juan E. Guevara-Andino, Anna R. Landim, Eike Lena Neuschulz, Ronja Hausmann, Jörg Müller, Karen Pedersen, Matthias Schleuning, Constance J. Tremlett, Edith Villa-Galaviz, H. Martin Schaefer, David A Donoso, Nico Blüthgen. Ecosphere. In press.
In this study, we introduce a newly established chronosequence with 62 plots (50 ✕ 50 m) in active cacao plantations and pastures, early and late regeneration, and old-growth forests in the extremely wet Chocó rainforest. We surveyed all trees with a circumference at breast height of 25 cm (DBH: 7.95 cm) within the 62 plots, including wild and cultivated trees, and took measurements of trees. We then tested how six forest structure variables (maximum tree height and DBH, basal area, number of stems, vertical vegetation heterogeneity, and light availability), aboveground biomass (AGB), and rarefied tree species richness change along our chronosequence. We showed that all tree attributes increase with regeneration time, showing a recovery of tree attributes between 30 and 206 years.
The repository contains two R scripts and several documents with raw data to replicate the analyses and figures in the manuscript.
Files and variables
File: analysis_reassembly.R
Description: This R script is used to replicate the statistical analyses performed in the manuscript. The analyses include calculations of maximum tree height and DBH, basal area, number of stems, vertical vegetation heterogeneity, and light availability, aboveground biomass, rarefied tree species richness. It also works to plot the recovery of these variables, and to filter cultivated species. This script is used along with the files tree_data_measures.csv, tree_data.csv, cultivated_trees.csv, and plots_reassembly.csv.
File: cultivated_trees.csv
Description: This file contains a list of 11 cultivated tree species and it is used to filter these species in case only wild trees want to be analyzed. This file is used with the script analysis_reassembly.R.
File: plots_reassembly.csv
Description: This files contains baseline and environmental data collected in the plots. This file is used with the scripts analysis_reassembly.R and plots_reassembly.R. Most of the variables are self explanatory or are referred and explained in the manuscript, except the following:
elev_30m_dem Elevation from Digital Elevation Model (m)
solar_annual Annual solar radiation (kWh/m2) calculated in ArcGIS.
TWI Topographic Wetness Index calculated calculated in ArcGIS.
TPI Topograhic Position Index calculated using raster:terrain.
TRI Topograhic Ruggedness Index calculated using raster:terrain.
Conv.ind Convergence Index calculated using the starsExtra package.
Forest_1km Proportion forest in 1-km radius buffer.
Forest_500m Proportion forest in 500-m radius buffer.
Forest_100m Proportion forest in 100-m radius buffer.
Forest_1km_JRC_TMF Proportion forest in 1-km radius buffer from JRC_TMF.
Distance_forest Distance nearest forest (m) - NOTE: all regeneration classified as non-forest
Distance_edge Distance nearest forest edge (m) - NOTE: includes pasture, cacao, all known regeneration Patch_ha Size of patch around plot (ha) - NOTE: not relevant for Matrix plots -km radius buffer. Cacao_1km Proportion cacao in 1
Cacao_Reg1_1km Proportion cacao regen I in 1-km radius buffer.
Pasture_1km Proportion pasture in 1-km radius buffer.
Pasture_Reg1_1km Proportion pasture regen I in 1-km radius buffer.
Pasture_Reg2_1km Proportion pasture regen II in 1-km radius buffer.
Oil_Palm_1km Proportion oil palm in 1-km radius buffer.
File: plots_reassembly.R
Description: This R script is used to replicate the comparisons performed between plots within each category (active plots, regenerating plots, and old-growth forest plots; or cacao plots, and pasture plots). These comparisons include differences in regeneration time, elevation, location, duration use, patch size, distance to nearest forest edge, and percentage of forest around plots. Regressions between two variables are also included to discard their correlation. This script is used along with the file plots_reassembly.csv.
File: tree_data_measures.csv
Description: This file contains circumference and height data of the surveyed trees. In addition it presents the column "Stem" which is used when more than one stem was found for an individual. This file is used with the script analysis_reassembly.R.
File: tree_data.csv
Description: This file contains data on the tree survey, including plot number, individual number, individual code (plot number + individual number), Family, Genus, and Species. This file is used with the script analysis_reassembly.R.
Date of data collection
February 2022 - July 2023
Geographic location of data collection
Reserves Canandé and Tesoro Escondido, Esmeraldas, Western Ecuador
Funding sources that supported the collection of the data
This work was supported by the Deutsche Forschungsgemeinschaft (DFG) funded Research Unit REASSEMBLY (FOR 5207). REASSEMBLY is a collaborative project between German and Ecuadorian institutions.
Missing data/not determined: NA/ND
Methods
Our assessment of trees on each plot occurred from February 2022 to July 2023 and comprised tree species identification and labeling of all individuals ≥ 25 cm of circumference at 1.3 m above the ground (≥ 7.95 cm diameter at breast height, DBH), including palms and lianas. We also measured the height of each tree using a laser rangefinder/hypsometer (Forestry Pro II, Nikon). , Nikon). Given that the terrain in the area is irregular and some height measures could be overestimated, any tree > 50 m was limited to this height. If a tree had more than one stem ≥ 25 cm circumference, we counted and measured up to the four thickest stems.