Migration is thought to have evolved in large mammalian herbivores as a strategy to exploit favorable spatial gradients in plant phenology and predation. Migrating individuals obtain higher forage quality for longer than their nonmigratory counterparts, and because their predators are bound to denning areas on the prey's wintering or dry-season range, migrants may also reduce their rates of predation. Historically these factors promoted the abundance of migratory ungulates in many ecosystems, but ungulate migration is now globally threatened. Concern over declining migration has focused primarily on seasonal habitat loss and the physical obstruction of migration routes, both of which can reduce carrying capacity and are associated with particularly dramatic declines. Less attention has been devoted to evaluating changes to the classic benefits of migration, a potentially important mechanism of migratory decline. We evaluated the benefits of migration for a partially-migratory elk population in the Greater Yellowstone Ecosystem (GYE) that has recently experienced declining performance of the migratory subpopulation despite intact seasonal ranges and migration routes. We first characterized the herd's transition with a series of comparative demographic and fitness indices for migratory and nonmigratory subpopulations. Because these comparisons suggested no contemporary benefit of migration, we next evaluated patterns of phenology, climate, and predator density that could account for the migratory decline.
Migratory calf recruitment declined by more than 70% from 1989-2009, alongside relatively stable nonmigratory recruitment. Over three years from 2007-2009, pregnancy of migratory elk averaged 68%, versus 89% for nonmigratory elk. The depression in migratory pregnancy was largely driven by the low pregnancy of prior-year lactators, indicating alternate-year reproduction in this subpopulation. This finding was accompanied by a 7% difference in the body fat levels of migratory lactators versus nonlactators and related to considerable phenological change in recent years. From 1989-2009, the summer range of migratory elk showed an increasing rate and shorter duration of vegetation green-up, but there were no significant phenological trends for nonmigratory elk. Migratory elk were also exposed to higher large carnivore densities associated with the restoration and expansion of grizzly bear (Ursus arctos) and gray wolf (Canis lupus) populations in the wilderness core of the GYE. We suggest that anthropogenic reversal of key ecological gradients that favor migration are responsible for the declining performance of migratory elk in this population. Such changes to the costs and benefits of migration may influence temperate montane ungulates with elevational migrations in other landscapes.