Human presence shifts the landscape of fear for a free-living mammal
Data files
Nov 04, 2024 version files 172.49 KB
-
GUD_Summary_dryad.xlsx
168.64 KB
-
README.md
3.85 KB
Abstract
Humans may play a key role in providing small prey mammals spatial and temporal refuge from predators, but few studies have captured the heterogeneity of these effects across space and time. Global COVID-19 lockdown restrictions offered a unique opportunity to explore how a sudden change in human presence in a semi-urban park impacted wildlife. Here we quantify how changes in the spatial distributions of humans and natural predators influenced the landscape of fear for the California ground squirrel (Otospermophilus beecheyi) in a COVID (2020) and non-COVID (2019) year. We used a structural equation modeling approach to explore the direct and indirect effects of human presence, predator presence, and habitat features on fear responses [e.g., giving-up densities (GUDs), number of foragers, and average food intake rate while at food patches]. In 2019, humans and dogs had moderate effects on GUDs; squirrels were less fearful (lower GUDs) in areas frequently visited by humans and dogs, but the effects of raptors were weak. In contrast, in 2020, the effects of humans and dogs on GUDs were weak; squirrels were more fearful of high raptor activity, open sky, and ground cover. In both years, squirrels farthest from refuge were the most risk-averse. Overall, our analyses revealed an increase in perceived risk from natural predators in 2020, associated with a change in the concentration of human presence. Thus, risk-sensitive foraging was dynamic across space and time, depending on a complex interplay between human, dog, and natural predator activity, and microhabitat features. Our findings elucidate the myriad ways humans, directly and indirectly, influence animal perception of safety and danger.
https://doi.org/10.5061/dryad.pk0p2ngv4
This dataset contains data collected during the summers of 2019 and 2020 for a giving-up density (GUD) experiment (GUD_Summary_dryad.xlsx). Along with it is the R code for analysis and graphs (SEM_2019-2020.R). Additional data including information about the activity calculations and squirrel density were collected throughout each season and located in a separate files. Please contact the authors for access to these files.
Description of the data and file structure
Data specific information for GUD_Summay_dryad.xlsx:
- Number of variables: 35
- Number of cases/rows: 612
- Variable list:
- Year: year the trial occurred
- Location: location within our study site that the plate was closest to, calculated by shortest Euclidean distance.
- Date: date the trial was conducted on
- Plate_id: number of the giving-up density plate
- Day: day of the year the trial was conducted on
- Trial_num: number of trial within each year. 3 trials were conducted each year.
- Start_time: time food was added to plate
- End_time: time food was removed from plate
- Millet_start_amount_g: amount of millet in grams that started on the plate
- Millet amount_ed_g: amount of millet in grams that was weighed at conclusion of trial
- Amount_eaten: total amount of millet in grams eaten during trial. Calculated from millet start amount minus millet end amount.
- Percent_remain: percentage of millet remaining at the end of the trial.
- Seed_remaining: number of seeds remaining on plate at the end of the trial
- Dist_nearest_burrow_m: distance in meters plate is away from a burrow
- SquirrelDens: density of squirrels in area calculated as the number of unique squirrels in a location per observation from observational data recorded two weeks prior to trial
- Non_target_visitor: binary; whether a non-target visitor foraged on the plate (0 = non-target visitor did not forage on plate, 1 = non-target forager did forage on plate)
- Ground_cover: percentage of total grass cover on ground surrounding plate (see methods)
- Ground_cover_bin: categorical based on number in ‘ground_cover’
- Sky_cover: percentage of tree and bush cover above plate (see methods)
- Sky_cover_bin: categorical based on number in ‘sky_cover’
- X_utm & Y_utm: Plate locations
- Multiple_forager: binary; whether more than one squirrel visited the plate during the trial (0 = only 1 squirrel listed plate during trial, 1 = more than 1 squirrel during trial)
- Num_of_unique_foragers: Total number of unique squirrels who visited the plate during the trial
- Total_time_foraging: Combined total time (minutes) all squirrels spent on plate during the trial
- Foraging_mean: average time (minutes) squirrels spent on plate during trial
- Foraging_mean_rd: whole integer of average time (minutes) squirrels spent on plate during trial
For all activity measures below: Seasonal measures for each visitor type at each benchmark location. Calculated as the number of disturbances per hour of observation throughout the entire season. See methods.
rattlesnake_activity = Activity of rattlesnakes
coyote_activity = Activity of coyotes
bop_activity = Activity of raptors
nat_pred_activity = Activity of rattlesnakes, coyotes, & raptors combined
human_dog_activity = Activity of humans & dogs combined
human_activity = Activity of humans
dog_w_human_activity = Activity of dogs on-leash with at least one human
dog_alone_activity = Activity of off-leash dogs, human presence not required
Sharing/Access information
Please contact Jennifer E. Smith or Chelsea A. Ortiz-Jimenez for any questions or collaboration.