Study site and subjects
We studied free-ranging ground squirrels at Briones Regional Park in Contra Costa County, California, USA (Latitude: 37.93 North, Longitude: 122.13 West, Elevation: 319 m above mean sea level). For this study, we focused on a 0.96-hectare site that is a well-known picnic and resting area near a main entrance to the otherwise less human-disturbed 6117-acre Briones Regional park (Fig. S1). Since 2013, we live trapped, marked, and released individuals of known age, sex, reproductive state, and mass (see Smith et al. 2018 for details). Further, we recorded the proportion of days an individual was trapped on available trapping days (henceforth ‘trappability’). Squirrels were individually fur marked, and on non-trapping days, we identified individuals from a distance using binoculars to record spatial locations and, in 2018 and 2019, their flight initiation distances (FIDs, see below). We combined all spatial data from May to July of these two years to determine each squirrel’s home site, or the locations each individual squirrel spent the majority of their time. Locations of trapped and free-living squirrels were noted daily based on a standard set of natural (e.g., burrows, trees) and artificial (e.g., picnic tables, outhouse) landmarks at the study site. We quantified human activity by counting the number of humans present within 15 m of each landmark (Hammond et al. 2019). Following van der Marel et al. (2019), we calculated human activity at each location by dividing the number of humans at a location by its total observation time each summer (henceforth ‘home site human activity score’). Within our site, human activity ranged over a spatial gradient going eastward from moderately high to low.