Back

Fine-scale spatio-temporal variation in natural selection on reproductive traits across two wild bird species

Soraker, J. S.; Araya-Ajoy, Y.; Cole, E. F.; Sheldon, B. C.

2026-02-27 ecology
10.64898/2026.02.27.708424 bioRxiv
Show abstract

Natural selection drives adaptive evolution and underpins eco-evolutionary processes. However, the causes of variation of selection (strength and direction), and its spatio-temporally scaling, remain poorly understood. Most selection estimates are population-wide, ignoring the variation in fine-scale selection regimes within populations. We quantify variation in selection within sub-populations of a long-term study of great tits (Parus major) and blue tits (Cyanistes caeruleus) in Wytham Woods, UK. Using mixed-effect modelling, we decomposed temporal, spatial and spatio-temporal variation in selection on laying date, clutch size and their correlational selection. Across species, we detected non-linear selection for larger clutch size and earlier laying, and negative correlational selection. Spatio-temporal effects explained most variation in selection across traits and species, suggesting that the causes of selection are controlled by interacting temporal (e.g. climate) and spatial (e.g. habitat) mechanisms. Laying date had strong temporal (among-year) variation for both species, whereas spatial components generally explained little variation. Selection estimates were highly correlated between species for all traits except for correlational selection, with higher agreement when analyzing fledgling versus recruitment numbers. Our results suggest common factors underpinning small-scale spatio-temporal variation in selection in both species and that such analyses can identify the sources causing natural selection.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.