Multi‐model application informs prey composition of mountain lions Puma concolor

Abstract

Global positioning system (GPS) receivers allow researchers to collect location data that provide information about fine-scale animal movements. For large carnivores, these data are routinely processed to identify clusters of GPS locations which are investigated to validate feeding sites, estimate prey species composition and model the likelihood of predation events based on characteristics of GPS location data within clusters. Although developing predation models entails a high level of field effort, researchers are apprehensive in applying system-specific models to other systems both spatially and temporally. Our objectives were to apply and compare multiple predation models to predict and identify feeding events outside of the geographic areas where models were developed. Using our multi-model approach, we identified feeding sites and estimated kill rates and prey composition of mountain lions Puma concolor in southwest Wyoming, USA from 2017 to 2019. Our approach increased our field efficiency by reducing potential field site investigations by 63%. We provide results of estimated diet composition in our study area where mountain lion prey included relatively high proportions of pronghorn (12.3%) and smaller mammals, particularly coyotes (7.6%). We also provide a comparison of the predation models developed across unique ecoregions of North America, and how they performed when applied to our area. We believe a similar approach could be adopted for other large carnivore populations where multiple models have been developed to characterize feeding events.

A field test of R package GPSeqClus: For establishing animal location clusters

“…GPSeqClus (Clapp et al, 2021) proved to be a package with consistent rules (Webb et al, 2008) that performed well on our data with a location‐acquisition rate one quarter that of the rate used by Clapp et al (2021, 2022) and hundredths of the rate used by Bischof et al (2022). This result (1) greatly broadens the location‐acquisition rates at which GPSeqClus has been shown to be useful, (2) adds new carnivore and prey species with which it can be used and (3) documents that with wolves, the relatively low location‐acquisition rate is sufficient to detect remains of the ungulates they kill or scavenge, thus maximizing the life of the radio‐collar and extending the period for which it is useful.…”

“…Clapp et al (2021:787) recently developed “a sequential‐clustering algorithm package ( GPSeqClus ) for Program R (R Development Core Team, 2021) to process location datasets based on user‐defined parameters. GPSeqClus also calculates an array of movement attributes commonly applied as covariates to develop cluster‐based models.” The basic function of GPSeqClus is to discern, from an animal‐location dataset, clusters of locations that bear field checking to determine ecological or behavioural information about the animal, such as predation or denning.…”

Cats and dogs: A mesopredator navigating risk and reward provisioned by an apex predator

“…Mule deer ( Odocoileus hemionus ), elk ( Cervus canadensis ), and pronghorn ( Antilocapra americana ) were the most abundant ungulates, followed by moose ( Alces alces ) and a small population of bighorn sheep ( Ovis canadensis ) on the southern edge of the study area. Mountain lion diets consisted primarily of mule deer (63.4%), elk (12.2%), and pronghorn (11.5%; Clapp et al, in press). Coyotes accounted for the next largest percentage of the total 262 mountain lion kills (8.4%), indicating the risk of mortality to coyotes from mountain lions was higher compared with other systems where coyotes were absent or present at lower levels in mountain lion diets (Anderson & Lindzey, 2003; Elbroch et al, 2013; Knopff et al, 2010; Wilckens et al, 2016).…”

“…We excluded investigated kill sites when prey were smaller than a juvenile mule deer (<1 year) to increase the likelihood that there were scavengeable resources available to coyotes. To increase our sample size of mountain lion kills with large prey present for our direct assessment, we used the methods described in Clapp et al (in press) to generate predictions of kills made by mountain lions based on clusters of mountain lion locations. We used a receiver operating characteristic (ROC) curve and the subset of investigated kill sites containing large prey relative to investigated kill sites where we found small or no prey to develop a new cutoff for identifying kills of large prey that maximized sensitivity, while allowing for a 5% false positive rate.…”

“…We used a receiver operating characteristic (ROC) curve and the subset of investigated kill sites containing large prey relative to investigated kill sites where we found small or no prey to develop a new cutoff for identifying kills of large prey that maximized sensitivity, while allowing for a 5% false positive rate. We compared the sensitivity metric across the four models implemented in Clapp et al (in press) and found the Wilckens et al (2016) model performed well for our application.…”