Shifting baselines increase the risk of misinterpreting biodiversity trends
Data files
Feb 11, 2026 version files 2.76 MB
-
_Shifted_baselines_-_Analyses_.zip
21.12 KB
-
_Shifted_baselines_-_Data_processing_.zip
45.41 KB
-
_Shifted_baselines_-_Figures_.zip
671.37 KB
-
Data_S1.xlsx
2.01 MB
-
README.md
8.69 KB
Abstract
Ecological studies quantifying the impact of land-use change on biodiversity may be sensitive to the choice of reference points – or baselines – particularly when sampling across human land-use gradients and other space-for-time comparisons. Much depends on whether the chosen baseline has already undergone shifts in species composition because of hunting, habitat loss, and degradation. However, few studies have assessed the influence of shifting baselines on estimates of anthropogenic impacts. Using new survey data from five West African land-use gradients, we examine how habitat patch size and structure influence the estimated impact of land-use change on bird species richness and functional diversity. We show that smaller forests have already lost many forest-dependent birds, particularly those with large body size or specialised ecological niches, leading to reduced estimates of biodiversity loss after deforestation. The steepest biodiversity loss was found in mid-sized forests, whereas relatively shallow declines were estimated for the most extensive forests, despite their richer taxonomic and functional diversity. In these larger forest blocks, accurate estimates of biodiversity loss may require longer transects extending beyond the biodiversity ‘shadow’ caused by the more extensive spillover of forest species into the surrounding landscape, potentially linked to source-sink dynamics. These findings suggest that biodiversity assessments are highly sensitive to baseline selection and transect design, highlighting the risk of underestimating land-use impacts unless shifting baselines are carefully considered.
This DRYAD repository contains the code and data associated with the manuscript: Shifting baselines increase the risk of misinterpreting biodiversity trends published in the journal Ecography.
Before you start
To reconstruct our analyses, all our scripts are run through R and can be executed on a standard PC.
We have depended on the following R packages to produce analyses and figures:
glmmTMB
DHARMa
multcomp
psych
emmeans
scales
sf
geosphere
lwgeom
rgeos
sp
dplyr
ggplot2
ggeffects
tidyverse
patchwork
ggpubr
These should be installed before rerunning code (using "install.packages(name_of_packages)" in R). After installation, the required packages to run each script will be loaded automatically when the script is run.
Data
The Excel file Data S1.xlsx contains the raw survey data used in our analyses.
The Excel file is organised into the following sheets:
Data_S1.xlsx
├── Metadata
├── Survey metadata
├── Survey results
├── Traits
├── Total diversity
├── Literature
├── Sources
The sheets Metadata, Literature and Sources provide contextual information but do not contain data necessary to rerun analyses.
The sheets Survey metadata, Survey results, Traits and Total diversity have to be converted to .csv files and stored in a folder called data before rerunning scripts.
Descriptions of all variables can be found in Metadata.
Data processing
Description of the data and file structure
The folder structure of _Shifted_baselines_-Data_processing.zip is as follows:
Shifted baselines - Data processing.zip
├── [Shifted baselines - Data processing] Calculating SR, RA and FDis.R
├── [Shifted baselines - Data processing] Calculating distance to forest edge.R
├── GEE canopy cover.txt
├── GEE canopy height.txt
├── Forest shapefiles
├──ahokwa.cpg
├──ahokwa.dbf
├──ahokwa.prj
├──ahokwa.qmd
├──ahokwa.shp
├──ahokwa.shx
├──ankasa.cpg
├──ankasa.dbf
├──ankasa.prj
├──ankasa.qmd
├──ankasa.shp
├──ankasa.shx
├──dompim.cpg
├──dompim.dbf
├──dompim.prj
├──dompim.qmd
├──dompim.shp
├──dompim.shx
├──gola.cpg
├──gola.dbf
├──gola.prj
├──gola.qmd
├──gola.shp
├──gola.shx
├──kakum.cpg
├──kakum.dbf
├──kakum.prj
├──kakum.qmd
├──kakum.shp
└──kakum.shx
Calculating diversity metrics
The script [Shifted baselines - Data processing] Calculating SR, RA and FDis.R calculates the diversity metrics used in all downstream analyses from the raw bird point-count data in Survey_results.
It formats the data so that each row has the landscape variables (taken from 'Survey_metadata') and the diversity metrics associated to one assemblage survey.
The script outputs a formatted .csv called data/Output script 1.csv which is used for analyses. The script has to be run before scripts in Shifted baselines - Analyses.zip can be run.
NOTE: the only script in Shifted baselines - Data processing.zip that has to be run ahead of rerunning analyses and reproducing figures is [Shifted baselines - Data processing] Calculating SR, RA and FDis.R. Other scripts in this folder are provided for reference only, to be able to reproduce methods used to calculate landscape variables (i.e. canopy cover etc.) in Survey_metadata.csv.
Calculating landscape variables
Using the shapefiles provided in the folder Forest shapefiles and the script Shifted baselines - Data processing] Calculating distance to forest edge.R the variable Distance.to.Forest.km provided in Survey_metadata.csv can be calculated.
The text files GEE canopy cover.txt and GEE canopy height.txt are java code used to calculate the Forest.Cover.200m and Canopy.Height.200m.mean variables provided in Survey_metadata.csv. This code was run in Google Earth Engine using the Hansen et al. 2013 and Potapov et al. 2022 canopy cover and height layers. The code is provided as an example and does not need to be run to reproduce analyses.
Analyses
Description of the data and file structure
The folder structure of _Shifted_baselines_-Analyses.zip is as follows:
Shifted baselines - Analyses.zip
├── [Shifted baselines - Analysis] FDis ~ Canopy cover.R
├── [Shifted baselines - Analysis] FDis ~ Distance to edge.R
├── [Shifted baselines - Analysis] FDis ~ Forest size.R
├── [Shifted baselines - Analysis] RA ~ Forest size.R
├── [Shifted baselines - Analysis] Sensitivity analysis slopes.R
├── [Shifted baselines - Analysis] Sensitivity analysis points.R
├── [Shifted baselines - Analysis] SR ~ Canopy cover.R
├── [Shifted baselines - Analysis] SR ~ Distance to forest.R
├── [Shifted baselines - Analysis] SR ~ Forest size.R
Modelling diversity in reponse to forest size
The script [Shifted baselines - Analysis] SR ~ Forest size.R uses generalised linear mixed models to model bird assemblage species richness across forest size, as presented in Fig. 3. The script uses Output script 1.csv (generated using the script [Shifted baselines - Data processing] Calculating SR, RA and FDis.R) as input and generates Fig. 3 as output.
Similarly, the scripts [Shifted baselines - Analysis] RA ~ Forest size.R and [Shifted baselines - Analysis] FDis ~ Forest size.R model bird assemblage relative abundance and functional dispersion across forest size, as presented in Fig. S6 and Fig. S7. The script uses Output script 1.csv (generated using the script [Shifted baselines - Data processing] Calculating SR, RA and FDis.R) as input.
Modelling diversity slopes in reponse to land-use change
The scripts [Shifted baselines - Analysis] SR ~ Canopy cover.R and [Shifted baselines - Analysis] SR ~ Distance to forest.R use generalised linear mixed models to test whether there are differences in species richness slopes across land-use change gradients between study areas. Land-use change is measured as canopy cover percentage and distance to forest edge respectively. The scripts use Output script 1.csv (generated using the script [Shifted baselines - Data processing] Calculating SR, RA and FDis.R) as input. Results are presented in Fig. 4 and Fig. 5.
The scripts [Shifted baselines - Analysis] FDis ~ Canopy cover.R and [Shifted baselines - Analysis] FDis ~ Distance to forest.R repeat the same analysis but for bird assemblage functional dispersion. The scripts use Output script 1.csv (generated using the script [Shifted baselines - Data processing] Calculating SR, RA and FDis.R) as input. Results are presented in Fig. S9 and Fig. S10.
Sensitivity analyses
The script [Shifted baselines - Analysis] Sensitivity analysis points uses t-tests to determine whether there are significant difference between the bird assemblage species richness detected in 3-point versus 4-point assemblages. Results are reported in Fig. S4.
The script [Shifted baselines - Analysis] Sensitivity analysis slopes repeats the analyses presented in [Shifted baselines - Analysis] SR ~ Canopy cover.R and [Shifted baselines - Analysis] SR ~ Distance to forest.R but excluding 4-point assemblages to test whether unevenness in assemblage sampling may have biased results. Outputs are reported in Table S10.
Figures
Description of the data and file structure
The folder structure of _Shifted_baselines_-Figures.zip is as follows:
Shifted baselines - Figures.zip
├── [Shifted baselines - Figures] Plotting Fig. 4.R
├── [Shifted baselines - Figures] Plotting Fig. 5.R
├── [Shifted baselines - Figures] Plotting Fig. S1.R
├── [Shifted baselines - Figures] Plotting Fig. S2 and Fig. S3.R
├── [Shifted baselines - Figures] Plotting Fig. S8.R
├── [Shifted baselines - Figures] Plotting Fig. S9.R
├── [Shifted baselines - Figures] Plotting Fig. S10.R
├── [Shifted baselines - Figures] Plotting phylogeny Fig. 2.R
├── Tree output.nex
Phylogenetic trees used to visualise the phylogenetic diversity of surveyed species were downloaded from https://birdtree.org/.
For reproducibility, copies of the exact trees used are provided in Shifted baselines - Figures.zip as Tree output.nex.
