Data from: Evidence for ecotone speciation across an African rainforest-savanna gradient
Data files
Feb 16, 2026 version files 91.53 MB
Abstract
Accelerating climate change and habitat loss make it imperative that plans to conserve biodiversity consider species' ability to adapt to changing environments. However, in biomes where biodiversity is highest, the evolutionary mechanisms responsible for generating adaptative variation and, ultimately, new species are frequently poorly understood. African rainforests represent one such biome, as decadal debates continue concerning the mechanisms generating African rainforest biodiversity. These debates hinge on the relative importance of geographic isolation versus divergent natural selection across environmental gradients. Hindering progress is a lack of robust tests of these competing hypotheses. Because African rainforests are severely at-risk due to climate change and other anthropogenic activities, addressing this long-standing debate is critical for making informed conservation decisions. We use demographic inference and allele frequency-environment relationships to investigate mechanisms of diversification in an African rainforest skink, Trachylepis affinis, a species inhabiting the gradient between rainforest and rainforest-savanna mosaic (ecotone). We provide compelling evidence of ecotone speciation, in which gene flow has all but ceased between rainforest and ecotone populations, at a level consistent with infrequent hybridization between sister species. Parallel patterns of genomic, morphological, and physiological divergence across this environmental gradient and pronounced allele frequency-environment correlation indicate speciation is mostly likely driven by ecological divergence, supporting a central role for divergent natural selection. Our results provide strong evidence for the importance of ecological gradients in African rainforest speciation and inform conservation strategies that preserve the processes that produce and maintain biodiversity.
https://doi.org/10.5061/dryad.xwdbrv1f4
Description of the data and file structure
This repository contains data used in analyses for our manuscript Ecotone Speciation Across an African Rainforest-Savanna Gradient, that was published in Molecular Ecology in 2023. Through a comprehensive demographic analysis of RAD-seq data from 14 populations in Cameroon and Gabon, we present genomic evidence for ecotone speciation in equatorial Africa in a rainforest skink, Tracylepis affinis. Specifically, we document the recent termination of gene flow between rainforest and ecotone populations, between which there is both morphological and physiological divergence. This is the first documented case of ecologically driven rainforest-ecotone speciation in an African vertebrate species.
Files and variables
File: PopulationIds.xlsx
Description: This file contains the geographic population ids for all samples used in the analysis.
File: PRJNA763817_TrachylepisAffinis_RADseq-STACKS_OnePerRadLocus.vcf.gz
Description: This file contains a restricted set of the original called genotypes such that one randomly sampled SNP per RADseq locus is retained.
File: PRJNA763817_TrachylepisAffinis_RADseq-STACKS.vcf.gz
Description: This file contains the full set of genotype calls derived from our RADseq data.
File: PRJNA763817_TrachylepisAffinis_StudySitesEnvData.xlsx
Description: This file contains latitude, longitude and environmental data for sampling sites from which skink samples were collected. For a full description of variables and how they were obtained see Freedman et al, 2023, Molecular Ecology. Variables included in the file are below. Unless noted otherwise, temperature and precipitation variables that start with "BIO" are in degrees Celsius x 10 and mm, respectively.
BIO1: mean annual temperature, from WorldClim
BIO2: mean diurnal range (i.e. mean of monthly (max temp - min temp)), from WorldClim
BIO4: temperature seasonality (standard deviation of monthly temperature x 100), from WorldClim
BIO5: maximum temperature of the warmest month, from WorldClim
BIO6: minimum temperature of the coldest month, from WorldClim
BIO12: annual precipitation, from WorldClim
BIO15: precipitation seasonality (coefficient of variation), from WorldClim
BIO16: precipitation of the wettest quarter, from WorldClim
BIO17: precipitation of the driest quarter, from WorldClim
quickscat: QuikSCAT backscatterometer (mean) – measure of surface moisture from radar remote sensing (see Long et al. 2001)
quickscatstd: QuikSCAT backscatterometer (standard deviation of quickscat) – measure of moisture variation (see Long et al. 2001)
srtm: mean elevation in meteres obtained from the Shuttle Radar Topography Mission, from WorldClim
treecover: percent tree cover, from Hansen et al. (2002)
ndvimax: maximum Normalized Difference Vegetation Index, from Zhang et al. (2005)
ndvigreen: vegetation greenness during greenest season derived from NDVI, from Zhang et al. (2005)
Code/software
All software and versions, and specific command lines can be found in either Freedman et al., 2023, Molecular Ecology, or in the associated GitHub repository: https://github.com/adamfreedman/TrachylepisAffinisSpeciation.
Access information
Demultiplexed fastq files are available at the NCBI Short Read Archive (SRA) bioproject PRJNA763817. Whole-body skink specimens and ethanol-preserved tissues are accessioned at the Museum of Vertebrate Zoology (MVZ) at the University of California, Berkeley.
RAD-seq data were collected from 208 individuals from 14 geographic populations in Cameroon and Gabon. Genotypes were generated with STACKS. Remotely sensed environmental variables for study sites were collected from a variety of sources to be detailed in the manuscript in preparation.
