Scale-dependent effects of density and habitat on foal survival

Identifying the most appropriate scale to study factors influencing life history is important to evolutionary ecology and wildlife management. For example, the scale at which density is assessed and explains variation in survival can affect how biologists observe and interpret population dynamics, which can influence plans for managing populations. Feral horses (Equus ferus caballus) contrast with most ungulates by exhibiting a mating system characterized by female-defense polygyny with persistent, non-territorial breeding groups (bands) and female-biased initial (natal) and subsequent (breeding) dispersal. We predicted that for horses, offspring movements coupled with female-biased breeding dispersal would increase the scale at which density best related to juvenile survival compared to species with greater female philopatry. From 2008 to 2013, we censused the population of feral horses on Sable Island, Canada. We annually computed individual-specific local densities for 442 foals (horses/km2 in radii of2,000 m, 4,000 m, and 8,000 m fixed to a band's centroid of movements) and whole-island (total) population density, group (band) size, and local access to surface freshwater, which affected movement patterns and selection of vegetation by females. The population of feral horses increased from 380 in 2008 to 559 in 2013. Overwinter survival of foals averaged 82.8%. Island-wide density was the most important predictor of foal mortality and was negatively associated with survival, with a lesser negative effect from local density. Increased access to surface freshwater (ponds) was an important predictor of foal survival but only at certain scales. Our study emphasizes the relevance of a multi-scale approach when analyzing the response of fitness components to changes in habitat and population processes, which may be influenced by the particular social organization of the species.