Camera traps were deployed at 320 sites across the Mojave Desert ecoregion and 265 sites across the Great Valley ecoregion between March and July of 2016 and 2017. At each survey location, a Reconyx PC900 camera trap was cable-locked onto a T-post that was securely placed in the ground and baited with a 1-kg salt lick, 500 ml of oatmeal-peanut butter mixture, and 150 g of fishy cat food. Cameras were programmed to take three photos at each trigger event with a delay of one second between trigger events. Each camera was deployed for an average of 34 days at sites in the Mojave Desert ecoregion and 29 days at sites in the Great Valley ecoregion.

Photos collected were reviewed to identify observed animals to the species-level, and multi-species hierarchical occupancy models were used to estimate 1) species-specific probabilities of occupancy, 2) the richness of terrestrial mammal species weighing >0.5kg, and 3) community and species-specific responses to the different ecological covariates. Each 24-hr period was treated as a repeat survey at a particular camera, where a '1' indicated species x was photographed at camera y on trap day z and a '0' indicated that it was not. A species' probability of occupancy (i.e., the probability a species used the area surrounding a camera trap during our survey) was assumed to be influenced by ecological covariates, such as the presence of an artificial water catchment, precipitation, and temperature in the Mojave Desert ecoregion, or water availability, temperature, and natural vegetation cover in the Great Valley ecoregion. A species' probability of detection (i.e., the probability a species was photographed if present) was assumed to be influenced by ecological covariates, such as human disturbance and precipitation.