Environmental analysis
To evaluate environmental variation between El Pardo and Navacerrada, we computed the score of each cell that an individual would have to cross in a theoretical straight-line migration event following valley bottoms on a principal component analysis (PCA) that combined all Bioclim environmental variables (2.0 dataset; cell resolution = 1x1 km; Booth et al. 2014) using R core (R Core Team, 2013). This PCA has been previously used to summarize environmental variation (temperature and humidity) across all the species’ range, yielding a single principal axis that opposed dry and warm cells to wet and cold ones (Llanos-Garrido et al., 2021). To assess the environmental gradient between these populations, we also considered the scores at the cells occupied by the populations themselves.