Data from: Geography and site-specific factors, rather than recent climate, dominated Spanish wintering bird communities
Data files
Dec 02, 2025 version files 2.73 GB
-
Community_analysis.zip
416.25 MB
-
README.md
8.59 KB
-
Spatiotemporalanalysis.zip
2.31 GB
-
Table_S6_Supplementary_Material.txt
24.36 KB
Abstract
Here, we analyse a dataset on the abundance of 97197 ringed birds collected from December to February at 78 sites across eastern Spain during the last 25 years (1997-2022). To fully understand the drivers of species composition, we analysed the components of β-diversity (turnover and nestedness) in relation to geographic and environmental gradients, including long-term changes in precipitation and temperature. Additionally, we examined the association patterns among species, species richness, and the interactions with geographical and environmental variables, and the temporal trends of these species and environmental variables. Mixed Graphical Models (Graphical lasso) revealed that dissimilarity in elevation, habitat, and sampling effort primarily drove turnover and nestedness, and that the assemblages remained largely stable. GLMMs showed that only six species exhibited significant trends driven solely by temporal effects, whereas most species were driven by longitude alone or combined spatiotemporal factors, suggesting strong site-specific responses and no consistent general pattern in the region. Although not statistically significant, the maximum temperature variables exhibited a positive temporal trend, suggesting gradual warming on the condition preceding each survey during the study period.
Dataset DOI: 10.5061/dryad.wwpzgmsxd
Description of the data and file structure
The dataset consists of a comprehensive compilation of bird-ringing records collected at 78 sites across the eastern Iberian Peninsula (Spain). It covers the period from December 1997 to February 2022 and includes a total of 97,197 ringed birds.
Files and variables
File: Spatiotemporalanalysis.zip
This folder contains the raw and formatted data to reproduce the analyses. Additionally, it contains the script to perform GLMMs to estimate the trends of the abundance, and presence/absence data of richness and 83 bird species, and the climatic variables. It also contains the workspaces with the model results.
DB.species.enviromental.txt: Contains Richness, the number of taxa on a given survey (column 1), and species abundance of the 83 species (from column 2 to column 84). The species names are abbreviated; for the full name, please see Table S4 from the Supplementary Material.
From column 85 onwards are the geographical, climatic, and site-specific variables.
Time: Julian day of the survey
Longitude: Longitude of the ringing site
Latitude: Latitude of the ringing site
Elevation: Elevation in meters above sea level of the ringing site
Rainfall: Rainfall on the day of the sampling day
Rainfall.max.90: Maximum rainfall/day during 90 days before the sampling
Rainfall.sum.7: Accumulated rainfall the week before the sampling
Rainfall.sum.30: Accumulated rainfall in the 30 days before the sampling
Rainfall.sum.360: Accumulated rainfall over the 360 days before the sampling (equivalent to the preceding year)
Tmax: Maximum temperature on the sampling day
Tmax.max.90: Maximum temperature of the maximum on the 90 days before sampling.
Tmax.max.360: Maximum temperature of the maximum on the 360 days before sampling (equivalent to the preceding year).
NDWI: Normalized Difference Water Index
Net.length: Mist-net length (meters)
Hours: Mist-net operation time (hours)
Site: Bird ringing station name
Habitat: Dominant habitat from six possible habitats: Riparian, Scrubland, Wetland, Woodland, Urban, and Farmland
model_results.RData: It contains two files, results_list with the results of the best model for abundance data of the 83 species, and Richness. The secondfiles significant_species provides information about the significance of the variables included in the mode,. which are Time, Longitu,de and the interaction of Time*Longitude.
models_results_presence.RData: It contains two files, results_presence with the results of the best model for presence-absence data of the 83 species. The second file, significant_species_presence, provides information about the significance of the variables included in the mod,l. which are Time, Longitude, and the interaction of Time*Longitude.
Dredge_climatic_variables_workspaces: It contains the model results performed by the dredge function from the MuMIn package for the climatic data.
Dredge_presence_absence_workspaces: It contains the model results performed by the dredge function from the MuMIn package for the presence/absence bird data.
Dredge_climatic_workspaces: It contains the model results performed by the dredge function from the MuMIn package for the abundance bird data.
File: Community_analysis.zip
Description: It contains two scripts to perform Mixed Graphical Models, one for Richness and the other for Beta diversity (turnover and nestedness), and 3 text files.
Coordinates and habitat.UFT8 contains:
Longitude: Longitude of the ringing site
Latitude: Latitude of the ringing site
Altitud: Elevation in meters above sea level of the ringing site
Hours: Mist-net operation time (hours)
Site: Bird ringing station name
Habitat: Dominant habitat from six possible habitats: Riparian, Scrubland, Wetland, Woodland, Urban, and Farmland
Entitat: Ringing scheme
Counts.species.site.day.month.year.long.no.restrictions.txt: it contains species abundance (full name) and the other variables explained below.
Training.dataset.compleate.no.restrictions.NVDI.txt: It contains all the variables before selecting the variables with a correlation < 0.5:
Time: Julian day
Site: Bird ringing station name
Longitude: Longitude
Latitude: Latitude
Elevation: Elevation in m.a.s.l.
Habitat: Six habitats: Riparian, Scrubland, Wetland, Woodland, Urban, and Farmland
Rainfall: Rainfall on the day of the sampling day
Rainfall.max.7: Maximum rainfall/day during the week before the sampling
Rainfall.max.30: Maximum rainfall/day during 30 days before the sampling
Rainfall.max.90: Maximum rainfall/day during 90 days before the sampling
Rainfall.max.360: Maximum rainfall/day during 360 days before the sampling
Rainfall.sum.7: Accumulated rainfall the week before the sampling
Rainfall.sum.30: Accumulated rainfall in the 30 days before the sampling
Rainfall.sum.90: Accumulated rainfall in the 90 days before the sampling
Rainfall.sum.360: Accumulated rainfall over the 360 days before the sampling (equivalent to the preceding year)
Tmax: Maximum temperature on the sampling day
Tmax.max.7: Maximum temperature of the maximum on the week before sampling
Tmax.max.30: Maximum temperature of the maximum on the 30 days before sampling
Tmax.max.90: Maximum temperature of the maximum on the 90 days before sampling
Tmax.max.360: Maximum temperature of the maximum on the 360 days before sampling (equivalent to the preceding year)
Tmax. Mean.7: Mean of the maximum temperature in the week before sampling
Tmax. Mean.30: Mean of the maximum temperature on the 30 days before sampling
Tmax .mean.90: Mean of the maximum temperature on the 90 days before sampling
Tma x.mean.360: Mean of the maximum temperature on the 360 days before sampling (equivalent to the preceding year)
Tmin: Minimum temperature on the sampling day
Tm in.min.7: Minimum of the minimum temperature in the week before sampling
Tmin.min.30: Minimum of the minimum temperature on the 30 days before sampling
Tmin.min.90: Minimum of the minimum temperature on the 90 days before sampling
Tmin.min.360: Minimum of the minimum temperature on the 360 days before sampling (equivalent to the preceding year)
Tmin.mean.7: Mean of the minimum temperature in the week before sampling
Tmin.mean.30: Mean of the minimum temperature on the 30 days before sampling
Tmin.mean.90: Mean of the minimum temperature on the 90 days before sampling
Tmin.mean.360: Mean of the minimum temperature on the 360 days before sampling (equivalent to the preceding year)
NVDI: Normalized Difference Vegetation Index
NDWI: Normalized Difference Water Index
Net.length: Mist-net length (meters)
Hours: Mist-net operation time (hours)
MGM.Fit.log.p.1.R1.RData: working space to load for the Richness analysis in the Mix Graphical Models
MGM.Fit.beta.diversity.R1.planar.dist. RData: working space to load for the Beta diversity analysis in the Mix Graphical Models
File: Table_S6_Supplementary_Material.txt
Description: This table contains the covariation between all the species and variables included in the community analyses.
Code/software
The analysis was performed with R version 4.3 and packages betapart, mgm, qgraph, igraph, and lme4.
The folder Community analysis contains two scripts:
-
Mixed Graphical Models + Richness llog+1- final for a Mixed Graphical Lasso model between the species Richness, species an, and selected variables.
-
Mixed Graphical Models beta diversity R1.planar.dist, where the two Beta diversity components (turnover and nestedness) are extracted and plotted with the covariated variables.
The folder Spatiotemporal analysis includes one script, SpatiotemporalanalysisR,1 to perform the GLMMs for abundance, presence/absence, and climatic data.
Access information
Other publicly accessible locations of the data:
- None
Data was derived from the following sources:
- Part of the data was derived from the winter bird ringing protocol of the SYLVIA program, which is run by the Catalan Ornithological Institute (ICO)