Behavioral changes and nutritional consequences to elk (Cervus canadensis) avoiding perceived risk from human hunters

Abstract

Abstract The life‐and‐death stakes of predator–prey encounters justify the high price of many anti‐predator behaviors. In adopting these behaviors, prey incur substantial non‐consumptive costs that can have population‐level consequences. Because prey knowledge of risk is imperfect, individuals may even adopt these costly behaviors in the absence of a real threat. For example, rather than only avoid hunters, many species categorically avoid all anthropogenic activity. Although hunting seasons only increase risk for specific individuals (e.g., males), non‐target individuals may still perceive human hunters as a source of risk and respond accordingly. Here, we used a large‐scale experiment including 89 animal‐years of location data from 62 unique individuals over 6 yr to quantify the duration, magnitude, and energetic consequences of changes to movement and resource selection behavior adopted by non‐target female elk (Cervus canadensis) in response to human hunters during three separate experimental 5‐d hunts (elk archery, deer rifle (Odocoileus hemionus or Odocoileus virginianus), and elk rifle). We predicted that elk response to hunters would be brief, but that strong behavioral responses to hunters (e.g., strengthened avoidance of roads and trails) would carry nutritional costs. We measured the duration of hunt‐related changes in elk speed using quantile regression, further quantified the strength of elk behavioral responses to hunters using population‐level resource selection functions, and evaluated whether anti‐predator resource selection behavior translated to measurable metabolic costs in the form of reduced body fat heading into winter. Elk responses to human hunters were stronger in the day than at night and were generally more pronounced during the elk hunts than during deer hunts. During hunts, elk shifted their diurnal behavior to avoid forage and intensified their avoidance of roads and trails. The combination of these changes in behavior led to a predicted pattern of distribution during the hunt that differed substantially from the distribution prior to the hunt. Lactating females that more strongly avoided roads entered winter in poorer nutritional condition, suggesting that the changes in resource selection we describe carry corresponding nutritional costs that have the potential to impact subsequent population performance.

Keywords

• anti‐predator behavior
• Cervus canadensis
• non‐consumptive effects
• non‐lethal effects
• predation risk
• predator–prey interaction