Rodent, snake and raptor use of restored native perennial grasslands is lower than use of unrestored exotic annual grasslands

1. In California's Central Valley, most native grasslands have been destroyed or degraded due to invasion, farming and development. Grassland restoration is often assumed to provide improved wildlife habitat, ostensibly increasing the abundance and diversity of at least some native wildlife species relative to unrestored, invaded annual grasslands. 2. We compared rodent, snake and raptor activity and species richness at paired unrestored and restored grasslands across four blocked locations in the Central Valley using trapping and observational surveys in up to four seasons per guild from 2014 to 2015. Restored treatments were planted with native perennial grasses 13-24 years prior to study initiation but were partially re-invaded by Mediterranean annual grasses and forbs. Unrestored treatments contained similar non-native plant species assemblages as restored treatments, but did not contain any native grass. 3. Rodent, snake and raptor activity was generally higher in unrestored relative to restored treatments. For rodents, the non-native Mus musculus (house mouse) showed the greatest disparity in abundance, while greater raptors and snakes likely responded to greater rodent abundance. 4. Within treatments, species-specific rodent responses were related to structure of physical vegetation. In particular, Peromyscus maniculatus (native deer mouse) was associated with more bare ground and shorter vegetation, while the house mouse was associated with less bare ground and taller vegetation, regardless of treatment type. Substantial changes in rodent species composition were observed over short periods of time (