Maternal Behavior Indicates Survival and Cause‐Specific Mortality of Moose Calves

Abstract

ABSTRACTContinuing research on cause‐specific mortality and annual survival of moose (Alces alces) calves in northeastern Minnesota, USA, is important to understanding the long‐term trajectory of the population. In 2013 and 2014, we observed global positioning system (GPS)‐collared, female moose exhibit a specific behavior (i.e., mortality movement) associated with the death of their GPS‐collared neonate. The females made a rapid, long‐distance movement (flee), followed by a return to the calf mortality site. We used characteristics of this movement in 2013–2014 (n = 46) to develop models for assessing calf survival, and then evaluated these models using female movement rates (n = 49) in 2015−2016. Using this behavior as an indicator of calf mortality in 2016, we conducted field investigations, leading to evidence of 15 mortalities at a mean age of 30.6 ± 15.5 (SE) days (range = 3–243 days). We launched 21 investigations in response to a mortality movement and they resulted in confirmation of 11 of the 15 calf mortalities. Specific causes of mortality included 9 wolf (Canis lupus)‐kills, 3 black bear (Ursus americanus)‐kills, 1 unknown predator‐kill, and 2 deaths following vehicle collisions. The mean distance females fled after a mortality was 1,873 ± 412 m (range = 126–5,805 m, n = 14). Females that made return visits returned a mean 2.8 ± 0.5 times (range = 1–5, n = 8) to within a mean 106 ± 22 m (range = 34–230 m, n = 8) of the mortality site. Calf survival to 30 days of age was 67 ± 8% (95% CI = 53–84%, n = 36) but declined to 53 ± 8% (95% CI = 39–72%, n = 36) by 3 months of age. We developed 2 population‐level movement models to improve the efficacy of using the mortality movement to identify and locate calf mortalities in real time via field investigations. The first approach, a temporal‐based model, used a 3‐day average movement velocity threshold (118 m/hr) for all females to indicate calf mortality and accurately predicted survival status in 51% (n = 105) of the cases. The second approach, an age‐specific model using different thresholds (28–135 m/hr) for females relative to calf age, was 80% (n = 231) accurate. Using movement behavior of females to assess calf mortality yielded important insights into mechanisms influencing the population decline that will inform future management decisions. © 2019 The Wildlife Society