Data from: Species-specific responses to paleoclimatic changes and landscape barriers drive contrasting phylogeography of co-distributed lemur species in northeastern Madagascar
Data files
Dec 10, 2025 version files 13.70 GB
-
README.md
6.15 KB
-
vanElstetal_2026_MolEcol.tar.gz
13.70 GB
Abstract
River barriers have long played a central role in diversification models of tropical regions, including the exceptionally biodiverse island of Madagascar. Although their role is best understood by integrating additional factors such as elevation and the ecological niche of a species, empirical studies integrating these variables remain rare. We used restriction site-associated DNA sequencing to assess the combined effect of rivers, topography, climate, and forest cover on the distributions and diversity of four Microcebus and two Avahi species (Primates, Lemuriformes) in northeastern Madagascar. We inferred population structure, gene flow, and genetic diversity, and assessed the association of these ecogeographic variables and genetic differentiation using isolation-by-resistance models. Our results show that significant differences in genetic diversity and connectivity among species can be explained by species-specific responses to landscape features and phylogeographic histories. Specifically, rivers present general barriers to gene flow, but dispersal between inter-river systems is possible via high-elevation headwater regions. While this led to high connectivity and genetic diversity in M. lehilahytsara and A. laniger, gene flow among M. jonahi populations is limited by low climatic niche suitability at higher elevations. Moreover, the more restricted distributions of M. macarthurii, M. simmonsi and A. mooreorum likely resulted from refugial dynamics and sea level fluctuations leading to allopatric divergence and microendemism. Together, the findings illustrate how ecological differences among species and temporal landscape dynamics mediate the role of rivers as dispersal barriers. They also highlight the importance of prioritizing river headwaters and topographically complex regions, which were shown to promote connectivity, in conservation efforts.
https://doi.org/10.5061/dryad.4xgxd25nt
The archive vanElstetal_2026_MolEcol.tar.gz holds input, configuration, and output files for the following analyses conducted as part of the publication van Elst et al. (2026), Mol. Ecol.:
- RAD genotyping
- Phylogenetic inference
- Population structure and gene flow
- Drivers of population structure
- Genetic diversity
1 Description of the data and file structure
1.1 RAD genotyping
For the genus Avahi, ./RAD_genotyping contains catalog output files (directories stacks.m*.M*/catalog.m*.M*.n*/populations.r*) and summary statistics (count_fixed_variant.tsv and n_snps_per_locus.tsv) of the de novo parameter tuning pipeline with varying combinations of the Stacks v2.53 parameters m, M, n and r (see Paris et al., 2017) for the full and species-specific (A. laniger) sample sets. Output files of the chosen best catalogs are given in the directories final (raw VCF files are populations.snp.vcf), together with the final filtered VCF files (populations.snps.filtered06.mac3.vcf), which were used in downstream analyses, and missingness statistics (populations.snps.filtered06.mac3.vcf.imiss).
For the genus Microcebus, ./RAD_genotyping contains output files of the reference-guided Stacks v2.53 genotype calling pipeline (directories populations; raw VCF files are populations.snp.vcf) and the final filtered VCF files (populations.snps.filtered06.mac3.vcf) with missingness statistics (populations.snps.filtered06.mac3.vcf.imiss) for the full and species-specific (M. jonahi+M.macarthurii, M. lehilahytsara, M. simmonsi) samples sets, which were used in downstream analyses.
1.2 Phylogenetic inference
./phylogenetic_inference contains output files of maximum likelihood phylogenetic inference with IQ-TREE v2.2.0, conducted on the full Avahi and Microcebus genotype call sets.
1.3 Population structure and gene flow
./clustering_analysis contains output files (*.P and *.Q across different number of clusters K and cross validation errors crossval.txt) of clustering analysis conducted on the species-specific genotype call sets (A. laniger, M. jonahi+M.macarthurii, M. lehilahytsara, M. simmonsi) with ADMIXTURE v1.3.0.
./gene_flow contains geographic sample coordinates and genetic distances (D_PS) between samples of the species-specific genotype call sets (A. laniger, M. jonahi, M. lehilahytsara, M. simmonsi), which were used as input to estimation of gene flow rates under a coalescent model with the R package 'gene.flow.inference' v0.0.0.9. In addition, it contains R workspace images (*.RData) of the analyses. Results are also given in Supplementary Tables S8-S11 of the original publication.
./excess_allele_sharing contains output (out_*) and configuration (SETS.txt and tree.nwk) files for tests of excess allele sharing between specific Avahi and Microcebus populations with Dsuite v0.4 r38 as described in the original publication.
1.4 Drivers of population structure
./isolation_by_distance contains geographic and genetic (D_PS) distances between samples (*_ind.txt) and populations (*_pop.txt) of the species-specific genotype call sets (A. laniger, M. jonahi, M. lehilahytsara, M. simmonsi), which were used as input to isolation-by-distance tests with the R package 'vegan' v2.5-7. Results are given in Supplementary Figs. S26 and S27 of the original publication.
./estimated_effective_migration_surfaces contains input, configuration (*.ini) and output files (directories results) for visualizations of spatial deviations from a model of isolation-by-distance based on the species-specific genotype call sets (A. laniger, M. jonahi, M. lehilahytsara, M. simmonsi) with EEMS.
./isolation_by_resistance contains raster files (*.tif) as well as geographic sample coordinates and genetic distances (D_PS) between samples of the species-specific genotype call sets (A. laniger, M. jonahi, M. lehilahytsara, M. simmonsi), which were used as input to isolation-by-resistance modeling with the R package 'radish' v0.0.2. Common raster files used in all analyses are found in the directory common_rasters. Raster files were created with the R packages 'terra' v1.6.47, 'rasterdiv' v0.3.4, 'whitebox' v2.2.0, 'ENMtools' v1.1.2 and 'ENMeval' v2.0.4. Modeling results are given in Supplementary Tables S4-S7 of the original publication.
1.5 Genetic diversity
./genetic_diversity contains individual observed heterozygosities estimated from the full Avahi and Microcebus genotype call sets with VCFtools v0.1.17. Results are also given in Supplementary Tables S12 and S13 of the original publication.
2 Sharing/access information
All new sequencing data have been made available through NCBI BioProject PRJNA807164. Individual BioSample accessions are given in Supplementary Table S1 of the original publication.
3 Code/Software
Associated scripts are found at GitHub. For detailed information on used software, please see the GitHub repository and the original publication.
