Browsing by Subject "Resource Selection Function"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    An Assessment of Recreational Harvest Impacts on Wolves (Canis Lupus) in the Western Great Lakes, USA: Do Hunters Selectively Harvest Wolves and How Might Wolf Movement Be Affected By Deer Hunting?
    (2020-03) Pruszenski, Jordan
    Conservation efforts in the Western Great Lakes have helped the wolf population in this region surpass federal mandates for a recovered population. If this distinct population segment (DPS) is removed for the Endangered Species List, the state management agencies are required to maintain stable populations within recovery stipulations. This thesis explores how recreational harvest potentially impacts the Western Great Lakes (WGL) wolf population through (1) a review of recent wolf harvest seasons in the contiguous United States and the biological concerns surrounding harvest seasons, (2) an analysis of the first wolf harvest seasons in the Western Great lakes to assess evidence of hunter selection or wolf vulnerability, and (3) a review of habitat selection and movement functions that can be used to better understand how wolves respond to recreational deer hunting. In the first chapter, I synthesize the structure of wolf harvest seasons in both the Northern Rocky Mountain region (NRM) and the WGL and discuss the wolf population trends that state wildlife managers have reported within a larger discussion of the research on wolf social behavior and life history that could be disrupted by recreational wolf harvests. Wolves in the NRM have been hunted intermittently since 2009, with Idaho and Montana consistently conducting wolf seasons starting in 2011 and Wyoming since 2017. Wolves in Minnesota, Michigan, and Wisconsin, which make up the WGL DPS, were delisted from 2012 to 2015 and these three states held wolf seasons during this time until the wolf population was again federally protected. Wolf harvest seasons were proposed to decrease human-wolf conflict by limiting the population size in these states and to allow recreational hunting and trapping. State managers are continuing to learn and adapt the percentage of the population that should be harvested to meet the state’s wolf population size goals and have acknowledged the need for better population estimates to understand how the wolf population growth rate is impacted by the harvest rate. Although states monitor the number of wolves harvested and take biological measurements from wolves that are killed such as sex, age, and female breeding status, there has been little research into if a specific type of wolf that is vulnerable to harvest and how the removal of these individuals could impact the population stability. For wolves, pack persistence is important for reproductive success. If either the female or male breeding individual (“breeder”) are killed, packs are more likely to not have pups that year and have a higher chance of disbanding completely especially during wolf mating and breeding season, which the wolf harvest seasons overlap. The few studies of pack stability in Idaho have found that harvest did not increase breeder loss and pack persistence rates did not decrease largely because packs in this long-term study were large with an average of nine members. In the event of a breeder loss, packs with six individuals or more are more likely to replace the breeding individual and produce pups that year or continue raising pups if they were already born prior to breeder loss. However, all states in the NRM have reported that average pack sizes have decreased to six members or less since the start of wolf harvest and average pack sizes in the WGL have been between five and three individuals. Pack persistence and other aspects of wolf life history need to be monitored not only to assess population size and potential impacts of hunting, but also to monitor genetic diversity that is important for a healthy wildlife population. The second chapter is an investigation into the WGL harvest seasons to explore hunter selection or wolf vulnerability. As mentioned above, disproportionate take of individuals that are more likely to be a breeder could lead to unforeseen decreases in the wolf population due to destabilizing wolf pack structure. For this analysis, I compiled estimates of the WGL wolf population sex ratio, adult female breeder to non-breeder ratio, and age distribution from WGL state and federal necropsied wolf reports. I also compared the harvest data to estimates derived from the scientific literature, including constructing a Leslie projection matrix model of the age distribution to more accurately estimate the WGL population age distribution. The total WGL harvest data was compared to population estimates and also assessed state level to investigate varying state management impacts on the harvest results. Additionally, at each of these scales, the harvest data was separated into hunting types (i.e., rifle, bow) or trapping to explore hunter selection or wolf vulnerability by harvest method The sex ratio of harvested wolves and the estimates did not indicate a biologically significant disproportionate take of one sex over the other. At all scales, there was a higher proportion of breeding females killed than would be expected compared to the estimates. However, the biological significance of this result needs to be further investigated since sexually mature females constituted a small proportion of the overall harvest. The largest proportion of wolves harvested at all scales and harvest methods were sexually immature wolves. These findings suggest that the early wolf harvest seasons in the WGL did not result in disproportionate harvest of wolves that are more likely to be important for maintaining pack stability. However, this analysis also highlights the need for accurate population estimates to understand which wolves are vulnerable to harvesting and predict how the wolf harvest will impact population dynamics. The third chapter explores how individual wolf behavior could change due to recreational harvest seasons, as assessed through with Resource Selection Functions (RSF) and Step Selection Functions (SSF). The advancement in Global Positioning Systems (GPS) has increased the frequency and accuracy of animal location data and has made robust statistical analysis of habitat selection and movement possible. RSF compare locations where the animal has occurred (“used locations”) to alternative locations. Since it is difficult to know locations that the animal has not used, RSF commonly compare used locations to random location within the study area or animal’s home range. In a case study, I show how RSF can be used to identify changes in wolf habitat selection of forested areas, roads, and trails before, during, and after deer hunting season near Voyageurs National Park in northern Minnesota, USA. I discuss the benefits and challenges of using a RSF design, one of which is that the static nature of the used locations in this analysis ignores the temporal connection between these locations as the animal is moving about the landscape. SSF incorporates both the spatial and temporal aspects of animal GPS locations to more realistically constrain alternative locations to what is biologically possible for the animal and to characterize animal movement patterns. While RSF compare used point locations, SSF use two consecutive used point locations to construct a “step”. This used step is compared to alternative steps originating from the same starting location but varying in distance (“step length”) and turning angle. SSF are powerful tool that can combine a variety of habitat selection and movement pattern variables depending on the research question. I demonstrate how SSF can be used to gain insight into changes in wolf habitat selection and movement speed at varying times of day in response to the deer hunting season. Although limited by the sample size of GPS collared wolves, the RSF and SSF preliminary studies demonstrate how these tests can be used to further our understanding of indirect human impacts on wolf behavior and access to resources.