Independent and interactive effects of anthropogenic disturbance and habitat on small mammals

Ecosystems are experiencing anthropogenic disturbances at a global scale, resulting in widespread habitat loss, fragmentation, and alteration. Yet, we know little about how habitat attributes may interact with landscape-scale human disturbance to influence local wildlife communities. Sagebrush habitats range-wide have been particularly altered. In the past two decades, energy development has increased in sagebrush habitats in the Intermountain west of North America. While responses to energy development have been documented for game species such as the greater sage-grouse (Centrocercus urophasianus) and mule deer (Odocoileus hemionus), studies documenting responses of non-game mammals are lacking. We examined the effects of structural habitat characteristics in areas with and without energy development on the abundance and diversity of small mammals in sagebrush steppe. Small mammals were live-trapped across gradients of sagebrush cover and height in 2009 and 2010 within 2 natural gas fields and adjacent control areas in the Upper Green River Basin, WY, USA. Small mammal density varied marginally across gradients of sagebrush cover and height with species-specific patterns. The density of deer mice (Peromyscus maniculatus), western harvest mice (Reithrodontomys megalotis), and reproductive individuals increased with sagebrush cover and height. Conversely, grasshopper mouse (Onychomys leucogaster) and sagebrush vole (Lemmiscus curtatus) density was inversely related to sagebrush cover and height. In addition, the density of deer mice, western harvest mice, northern grasshopper mice, juvenile individuals, and species richness were higher at energy development sites. Population estimates of deer mice showed a significant interaction between sagebrush habitat treatment and energy development. In summary, our results suggest both independent and interactive effects of habitat and disturbance on the small mammal community. Therefore both must be considered in management actions related to human disturbance. Additionally, we performed a methodological assessment of live trap types. The ability of researchers to sample small mammal populations is affected by bias introduced by trapping methods. Havahart live traps captured significantly more small mammals than expected (> 25%), while Sherman live traps captured significantly fewer small mammals than expected (< 75%) in sagebrush steppe habitats. Havahart traps captured more than expected of most species and age classes of small mammals regardless of local habitat or the presence of natural gas development. Sherman live traps captured ten species and Havahart traps captured six species across three levels of sagebrush cover and height and between sites with and without natural gas development. Use of multiple trap types in small mammal studies may reduce bias associated with sampling methods.