Functional connectivity in a continuously distributed, migratory species as revealed by landscape genomics

Maintaining functional connectivity is critical for the long-term
conservation of wildlife populations. Landscape genomics provides an
opportunity to assess long-term functional connectivity by relating
environmental variables to spatial patterns of genomic variation resulting
from generations of movement, dispersal, and mating behaviors. Identifying
landscape features associated with gene flow at large geographic scales
for highly mobile species is becoming increasingly possible due to more
accessible genomic approaches, improved analytical methods, and enhanced
computational power. We characterized the genetic structure and diversity
of migratory mule deer (Odocoileus hemionus) using 4,051 single nucleotide
polymorphisms in 406 individuals sampled across multiple habitats
throughout Wyoming, USA. We then identified environmental variables
associated with genomic variation within genetic groups and statewide
using a stepwise approach to first evaluate nonlinear relationships of
landscape resistance with genetic distances and then use mixed-effects
modeling to choose top landscape genomic models. We identified three
admixed genetic groups of mule deer and found that environmental variables
associated with gene flow varied among genetic groups, revealing
scale-dependent and regional variation in functional connectivity. At the
statewide scale, more gene flow occurred in areas with low elevation and
mixed habitat. In the southern genetic group, more gene flow occurred in
areas with low elevation. In the northern genetic group, more gene flow
occurred in grassland and forest habitats, while highways and energy
infrastructure reduced gene flow. In the western genetic group, the null
model of isolation by distance best represented genetic patterns. Overall,
our findings highlight the role of different seasonal ranges on mule deer
genetic connectivity, and show that anthropogenic features hinder
connectivity. This study demonstrates the value of combining a large,
genome-wide marker set with recent advances in landscape genomics to
evaluate functional connectivity in a wide-ranging migratory species.