Skip to main content

Major population splits coincide with episodes of rapid climate change in a forest-dependent bird

Cite this dataset

Warmuth, Vera-Maria et al. (2021). Major population splits coincide with episodes of rapid climate change in a forest-dependent bird [Dataset]. Dryad.


Climate change influences population demography by altering patterns of gene flow and reproductive isolation. Direct mutation rates offer the possibility for accurate dating on the within-species level but are currently only available for a handful of vertebrate species. Here, we use the first directly estimated mutation rate in birds to study the evolutionary history of pied flycatchers (Ficedula hypoleuca). Using a combination of demographic inference and environmental niche modelling, we show that all major population splits in this forest-dependent system occurred during periods of increased climate instability and rapid global temperature change. We show that the divergent Spanish subspecies originated during the Eemian-Weichselian transition 115 – 104 thousand years ago (kya), and not during the last glacial maximum (26.5 - 19 kya), as previously suggested. The magnitude and rates of climate change during the glacial-interglacial transitions that preceded population splits in pied flycatchers were similar to, or exceeded, those predicted to occur in the course of the current, human-induced climate crisis. As such, our results provide a timely reminder of the strong impact that episodes of climate instability and rapid temperature changes can have on species’ evolutionary trajectories, with important implications for the natural world in the Anthropocene.


Species distribution modeling. We used the ensemble approach implemented in the R package biomod2 to model the breeding distribution of pied flycatchers for the past 200,000 years using paleoclimate data published in Beyer et al. 2020 (Scientific Data). Each of the four chosen climate variables (minimum temperature, temperature variability, minimum precipitation, precipitation variability) represent averages across the breeding season (May, June, and July). Pied flycatcher occurrence data consist of breeding evidence records from the European Breeding Bird Atlas 2 (EBBA2) project for Europe, and of gbif occurrence data for the region east of the Ural mountains.

Demographic modeling: We used DADI (Gutenkunst et al. 2010, Nat Prec) to fit a model of ancestral size change, followed by a split-with-migration, to two-dimensional site frequency spectra of pied flycatchers. This model was previously found to provide the best fit to pied flycatcher genetic data (Warmuth & Ellegren, 2019, Mol Ecol Res). Site frequency spectra were derived from ddRADSeq data and generated using ANGSD. The populations represent a subset of populations in Lehtonen et al 2009 (Mol Ecol).

For further details, please refer to the publication associated with this entry.

Usage notes

Occurrence data (file: sdmdata.csv) are a combination of the EBBA2 breeding evidence dataset (50x50km) and gbif data for the region east of the Ural mountains (not covered by the EBBA2 dataset). Gbif occurrence data were filtered to include only occurences from the months of may, june and july. The combined dataset (EBBA2 and gbif) was filtered for duplicates and represents one occurrence point per grid cell (50x50km).

Genetic data (folder: bam_files). Fastq files were mapped to a version of the Ficedula hypoleuca reference genome that no longer exists in the public domain. We have therefore included the mapped reads (bam files) rather than the raw fastq files. For details on the mapping parameters, see Warmuth & Ellegren, 2019, Mol Ecol Res.


Knut and Alice Wallenberg Foundation, Award: 2014/0044

Swedish Research Council, Award: 2013‐8271

Swedish Research Council, Award: 2020-04686

Birgit and Hellmuth Hertz Foundation of The Royal Physiographic Society of Lund, Award: 2017-39034

Academy of Finland, Award: 314254

Estonian Research Council, Award: IUT34-8