Vertebrate diversity and biomass along a recovery gradient in a lowland tropical forest
Data files
Dec 13, 2024 version files 83.54 KB
-
Grella_et_al._vertebrate_assessment_R-script.R
25.64 KB
-
README.md
5.03 KB
-
S4_environment.ods
23.51 KB
-
S5_Observed_for_57_plots_and_predicted_values_for_all_65_plots.csv
16.70 KB
-
S6_community.csv
5.05 KB
-
S7_biomass_diversity.csv
3.04 KB
-
S8_community_indicspecies.csv
4.56 KB
Abstract
Deforestation of tropical forests have resulted in extensive areas of secondary forests with the potential to restore biodiversity to former old-growth forest levels. The recovery of vertebrate communities is an essential component of biodiversity and ecosystem restoration, as vertebrates provide key ecosystem functions. However, little is known about the recovery trajectories and habitat preferences of vertebrates in tropical landscapes with differing land-use legacies. We used camera traps covering three weeks to study the activity of ground-based mammals and birds in the understory of 57 sites along a forest recovery gradient, ranging from active agriculture, such as pastures and cacao plantations, to naturally recovering forests and old-growth forests in the Chocó rainforest in north-western Ecuador. Our results show that diversity and biomass of wild vertebrates are highest in old-growth forests and late recovery stages, while for domestic vertebrates, these indices are highest in agricultural land. Additionally, while species-habitat networks showed low habitat specificity for vertebrate species, an indicator species analysis found no species to indicate old-growth forests, Dasyprocta punctata and Tayassu pecari to indicate all forest types, and Aramides wolfi and Pecari tajacu to indicate late regeneration forests. We suggest that these patterns are caused by a high habitat connectivity and large amounts of remaining old-growth forest in our study area. Our findings indicate that secondary forests have a high potential for the recovery of vertebrate species diversity and biomass to old-growth level in lowland tropical forests with short regeneration times.
README: Vertebrate diversity and biomass along a recovery gradient in a lowland tropical forest
https://doi.org/10.5061/dryad.bnzs7h4mj
Description of the data and file structure
These data set contain data derived by camera traps on 65 plots in the reserves Reserva Río Canandé and Reserva Tesoro Escondido in northwest Ecuador.
Files and variables
File: Grella_et_al._vertebrate_assessment_R-script.R
Description: R code for data analysis
- Includes the following analysis:
- Analysis of species diversity and biomass using a generalized linear model
- Network analysis
- Indicator species analysis
- Predictions of species diversity and biomass using a generalized additive model
File: S4_environment.ods
Description: Environmental information of plot location, land-use legacy, surrounding forest cover and forest distance of each plot
- Project3: Information about plot selection
- Reassembly: Plots that are in the final plot selection of the REASSEMBLY Research Unit
- Wuerzburg: Additional plots that were sampled for this study
- Camera Year: Sampling year in which the camera traps were installed
- Plot_ID: Plot name
- PREX: Information if this plot was part of the perturbation-recruitment experiment of the REASSEMBLY Research Unit. For this experiment parts of the plots were either fenced for preventing vertebrates access or disturbed by removing vegetation and leaf litter.
- 1 means that there was a PREX treatment on the plot
- NA means that there was no PREX treatment on the plot
- Latitude and Longitude: GPS coordinates of plot
- Elevation: Altitude of plot (meters above sea level)
- Treatment1, Treatment2, Category4, Category7: Different variations of the land-use categories depending on regeneration age and land-use legacy (pasture or cacao plantation)
- Regeneration: Year in which the plot was abandoned for natural regeneration
- Succession: Years since abandonment
- Forest_1km: Forest cover in a 1 km radius. See methods in Escobar et al. (2024).
- Distance_forest_m: Distance from plot to the next old-growth forest when plot was agricultural land. See methods in Escobar et al. (2024).
- Distance_edge_m: Distance from plot to the next forest edge when plot was forested. See methods in Escobar et al. (2024).
File: S5_Observed_for_57_plots_and_predicted_values_for_all_65_plots.csv
Description: Evironmental information (S4) and biomass and diversity information (S7) complemented with predictions of diversity and biomass on each plot. Predictions were calculated using a generalized additive model.
- Bird_pred_bio: Prediction of bird biomass
- Dom_all_pred_bio: Prediction of domestic vertebrate biomass
- Mammal_pred_bio: Prediction of mammal biomass
- Birdsq0_pred: Prediction of bird number of observed species (q0)
- Domesticsq0_pred: Prediction of domestic vertebrate number of observed species (q0)
- Mammalsq0_pred: Prediction of mammal number of observed species (q0)
- Birdsq1_pred: Prediction of bird effective species number (q1)
- Domesticsq1_pred: Prediction of domestic vertebrate effective species number (q1)
- Mammalsq1_pred: Prediction of mammal effective species number (q1)
File: S6_community.csv
Description: List of captured species on each plot
- Latin Name: Latin species name
- Group: Species were grouped into
- Mammal
- Bird
- Domestic Bird (DomBird)
- Domestic Mammal (Dom_Mammal)
- Human
File: S7_biomass_diversity.csv
Description: Diversity and biomass on each plot
- Plot: plot name
- Category7: Land-use categories used for the analysis of species diversity and biomass
- Bird_Bio: Bird biomass
- Mammal_Bio: Mammal biomass
- Dom_all_Bio: Domestic vertebrate biomass
- Birdsq1: Bird effective species number (q1)
- Birdsq0: Bird number of observed species (q0)
- Domesticsq1: Domestic vertebrate effective species number (q1)
- Domesticsq0: Domestic vertebrate number of observed species (q0)
- Mammalsq1: Mammal effective species number (q1)
- Mammalsq0: Mammal number of observed species (q0)
File: S8_community_indicspecies.csv
Description: List of captured species on each plot (S6) formated for indicator species analysis
Code/software
The provided code is written in the programming language R. For analysis we used R version 4.2 and the following R packages: ggplot2, ggpubr, stringr, bipartite, tidyr, indicspecies, mgcv, nlme, readODS, ncf, MASS.
Literature
Escobar, S., Newell, F. L., Endara, M.-J., Guevara-Andino, J. E., Landim, A. R., Neuschulz, E. L., Nußer, R., Müller, J., Pedersen, K. M., Schleuning, M., Tremlett, C. J., Villa-Galaviz, E., Schaefer, H. M., Donoso, D. A., & Blüthgen, N. (2024). Reassembly of a tropical rainforest ecosystem: A new chronosequence in the Ecuadorian Chocó tested with the recovery of tree attributes. bioRxiv [Preprint]. https://doi.org/10.1101/2024.03.21.586145