Effects of lethal management on gray wolf pack persistence and reproduction in Wisconsin, USA.

Direct human-caused mortality accounts for about half of all large mammal mortality in North America. For social species like gray wolves (Canis lupus), the death of pack members can disrupt pack structure and cause pack dissolution, and mortality of breeding adults or wolves during reproduction and pup-rearing can decrease pup recruitment. We estimated minimum and maximum probability of wolf pack persistence in Wisconsin, USA, during biological years (15 April-14 April) 2011-2019 and evaluated the influence of pack size and legal harvest mortality on pack persistence during 2012-2014. Harvests comprised 75-161 mortalities within 194 monitored packs during 2012-2014, with 56-74% of packs having no wolves harvested each year. As an index of reproduction during 2013-2019, we also estimated the proportion of packs where pups responded to howl surveys. We evaluated the influence of pack size, legal harvest, and agency removal on reproduction during 2013-2015. Annual maximum pack persistence probability was uniformly high (0.95-1.00), and annual minimum pack persistence probability ranged from 0.86-0.98 with a possible decline during years of harvest. Reproduction was similar in years following harvest and agency removal (2013-2015, pup response = 0.27-0.40), and years without harvest or agency removal the year prior (2016-2019, pup response = 0.28-0.66). Pack size had a positive effect on pack persistence and reproduction. Total number of wolf mortalities and number of adult male and females removed did not influence pack persistence or reproduction. We suggest that low per-pack mortality, timing of harvest and agency removal, and harvest characteristics during 2012-2014 supported stable pack persistence and reproduction.

A landscape of overlapping risks for wolf-human conflict in Wisconsin, USA.

Managing risk requires an adequate understanding of risk-factors that influence the likelihood of a particular event occurring in time and space. Risk maps can be valuable tools for natural resource managers, allowing them to better understand spatial characteristics of risk. Risk maps can also support risk-avoidance efforts by identifying which areas are relatively riskier than others. However, risks, such as human-carnivore conflict, can be diverse, multi-faceted, and overlapping in space. Yet, efforts to describe risk typically focus on only one aspect of risk. We examined wolf complaints investigated in Wisconsin, USA for the period of 1999-2011. We described the spatial patterns of four types of wolf-human conflict: livestock depredation, depredation on hunting hounds, depredation on non-hound dogs, and human health and safety concerns (HHSC). Using predictive landscape models and discriminant functions analysis, we visualized the landscape of risk as a continuous surface of overlapping risks. Each type of conflict had its own unique landscape signature; however, the probability of any type of conflict increased closer to the center of wolf pack territories and with increased forest cover. Hunting hound depredations tended to occur in areas considered to be highly suitable wolf habitat, while livestock depredations occurred more regularly in marginal wolf habitat. HHSC and non-hound dog depredations were less predictable spatially but tended to occur in areas with low housing density adjacent to large wildland areas. Similar to other research evaluating the risk of human-carnivore conflict, our data suggests that human-carnivore conflict is most likely to occur where humans or human property and large carnivores co-occur. However, identifying areas of co-occurrence is only marginally valuable from a conservation standpoint and could be described using spatially-explicit human and carnivore data without complex analytical approaches. These results challenge our traditional understanding of risk and the standard approach used in describing risk. We suggest that a more comprehensive understanding of the risk of human-carnivore conflict can be achieved by examining the spatial and non-spatial factors influencing risk within areas of co-occurrence and by describing the landscape of risk as a continuous surface of multiple overlapping risks.