Identifying elk and mule deer parturition events.
We inferred elk parturition events from GPS-collared elk between 2016 and 2019 using the rolling minimum convex polygon (MCP) method described by Nicholson et al. (2019) to estimate large herbivore parturition events based on localized movements. We assigned a parturition event for elk as the first day a rolling MCP (based on a 24-hour window) decreased to <30 hectares or less for a minimum of 120 hours. If these conditions were not met, we assumed the elk did not give birth. We validated the method using independent elk GPS locations with known birth dates and locations (N = 30) from a previous study in the same area (Long et al. 2016). The mean discrepancy between birth dates determined from field investigations by Long et al. (2016) and birth dates predicted by the rolling MCP method was 18 hours.
Parturition events of mule deer were determined either by 1) monitoring GPS-collared adult females for localized movement and then searching the area for neonates where a cluster of GPS locations had formed, 2) using vaginal implant transmitters (details in Jackson et al. (2021)), or 3) using the rolling MCP inference method (Nicholson et al.2019). For the rolling MCP method, we assigned a parturition event for deer as the first day a rolling MCP (based on a 24-hour window) was <15 hectares or less for a minimum of 120 hours. We validated the rolling MCP method for 14 deer parturition events determined from field investigations and found a mean discrepancy of 33 hours.