Animal populations may exhibit regional hierarchical dynamics including regional common trends and site-specific variability. Landscape permeability and habitat resistance may influence animal movements and subsequently regional population dynamics and genetic structure. Long-distance dispersal or movements may synchronize the dynamics and generic structure of many geographically separated populations. Consequently, understanding population dynamics and genetic structure across large spatial extents has implications for regional management and conservation of wildlife populations. We aimed to assess landscape-abundance relationships, spatial synchrony, regional hierarchical population dynamics, and spatial genetic structure of eastern wild turkey (Meleagris gallopavo silvestris) populations in Mississippi, USA. We used ensemble machine learning methods to generate a statewide habitat suitability map of wild turkeys with data on occurrence locations and land cover and land use in Mississippi. We used a nonparametric spatial correlogram to detect spatial synchrony and used Bayesian spatiotemporal dynamic models to estimate regional hierarchical dynamics of 27 wild turkey populations from 1992 to 2005. We genotyped 224 wild turkeys captured across Mississippi with 10 microsatellite markers, and estimated effective dispersal distance by spatial genetic analysis and population genetic divisions using multivariate clustering.
Results/Conclusions
Diverse land covers are required for wild turkey habitat, with forests being indispensable. Wild turkey relative abundance exhibited a nonlinear, threshold response to variation in amount of forest from 1992 to 2005. Twenty seven wild turkey populations were synchronized within about 260 km with an average regional synchrony of 0.17. However, spatial synchrony was unlikely to be caused by climate variability, as predicted by the Moran effect, nor by dispersal. Maximum dispersal distance was about 56 km. Spatial variations in average abundance and temporal trends explained more (60%) spatiotemporal variability of wild turkey populations than climate variability (<1%). Regional habitat degradation may have resulted in regional population declines and consequent synchronization of wild turkey populations, which we term the habitat degradation hypothesis. We detected three distinct genetic divisions of the 224 genotyped wild turkeys; however, these genetic divisions were not spatially distinct. Wild turkey populations exhibited more local variability than regional variability. Scale-dependent landscape-abundance relationships demonstrated that habitats need to be evaluated at a spatial scale larger than the stand level, such as the annual dispersal range of the species. The lack of spatial population structure and genetic structure suggests site-specific wild turkey management is warranted.