Matthew A. Mumma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Matthew A. Mumma has a total of 161 co-authors. Here are the top co-authors by number of shared publications:
Shane P. Mahoney
S. E. Gullage and S. P. Mahoney, Conservation Visions, LLC, St. John's, NL, Canada
9 shared publications- Lisette P. Waits
Department of Fish and Wildlife Sciences, University of Idaho, Moscow, ID 83844-1136, USA
5 shared publications Chris J. Johnson
University of Northern British Columbia
4 shared publicationsDennis L. Murray
Environmental and Life Sciences Graduate Program, Trent University, Peterborough, Ontario, Canada
4 shared publications- Guillaume Bastille‐Rousseau
Center for Wildlife Sustainability Research, Southern Illinois University Carbondale Illinois USA
4 shared publications Katherine L. Parker
University of Northern British Columbia
4 shared publications
Latest Publications
- Open Access
Further development of camera‐based estimates of abundance for unmarked animals
Estimating abundance is a cornerstone of wildlife management, conservation, and research, and a variety of approaches have been developed to meet this need. Camera traps have been used to estimate abundance for decades, with earlier deployments using a mark–recapture framework and requiring uniquely identifiable individuals. Recently, there has been a surge in methods using camera trap data that do not require individual identification, including instantaneous sampling (IS) and space‐to‐event (STE) methods. The IS and STE methods involve randomly deploying cameras to simultaneously take photos at fixed intervals and have the advantages of not requiring estimates of animal movement rate and detection probability and are generally robust to variation in animal movement patterns and space use in contrast to nonrandom camera deployments. We describe three alternative estimators of animal density based on IS and STE methods and demonstrate how each are derived. We examined the performance of IS, STE, and the three alternatives using simulated data for 90 different scenarios representing five population densities, three different grouping behaviors, three assumptions of animal movements and habitat selection, and two levels of sampling effort (i.e., 50 and 130 cameras). We evaluated the accuracy of abundance estimates and assessed bias in four alternative approaches for estimating the SE and CIs. We found all estimators were unbiased when there was no habitat selection or when habitat selection was uncorrelated with viewshed area. However, IS was the only estimator that remained unbiased when habitat selection was related to viewshed area. Nonparametric bootstrapping across cameras was the least biased approach for estimating SEs, and unbiased CIs were obtained via a semi‐parametric approach. Across our scenarios, we found that when densities were >2 individuals/km 2 and when >50 photos with animals were observed, a CV in the abundance estimate <20% was generally achieved with the IS method. Our results emphasize the importance of random placement of cameras, as we show that nonrandom deployments such as placing cameras along roads or game trails, without understanding and accounting for the potential habitat selection, can significantly bias abundance estimates.
Open AccessFurther development of camera‐based estimates of abundance for unmarked animals
Estimating abundance is a cornerstone of wildlife management, conservation, and research, and a variety of approaches have been developed to meet this need. Camera traps have been used to estimate abundance for decades, with earlier deployments using a mark–recapture framework and requiring uniquely identifiable individuals. Recently, there has been a surge in methods using camera trap data that do not require individual identification, including instantaneous sampling (IS) and space‐to‐event (STE) methods. The IS and STE methods involve randomly deploying cameras to simultaneously take photos at fixed intervals and have the advantages of not requiring estimates of animal movement rate and detection probability and are generally robust to variation in animal movement patterns and space use in contrast to nonrandom camera deployments. We describe three alternative estimators of animal density based on IS and STE methods and demonstrate how each are derived. We examined the performance of IS, STE, and the three alternatives using simulated data for 90 different scenarios representing five population densities, three different grouping behaviors, three assumptions of animal movements and habitat selection, and two levels of sampling effort (i.e., 50 and 130 cameras). We evaluated the accuracy of abundance estimates and assessed bias in four alternative approaches for estimating the SE and CIs. We found all estimators were unbiased when there was no habitat selection or when habitat selection was uncorrelated with viewshed area. However, IS was the only estimator that remained unbiased when habitat selection was related to viewshed area. Nonparametric bootstrapping across cameras was the least biased approach for estimating SEs, and unbiased CIs were obtained via a semi‐parametric approach. Across our scenarios, we found that when densities were >2 individuals/km 2 and when >50 photos with animals were observed, a CV in the abundance estimate <20% was generally achieved with the IS method. Our results emphasize the importance of random placement of cameras, as we show that nonrandom deployments such as placing cameras along roads or game trails, without understanding and accounting for the potential habitat selection, can significantly bias abundance estimates.
- 30 January 2025
- Published by Wiley in Ecological Monographs
- Vol. 95 (1)
Mammalian predator co‐occurrence affected by prey and habitat more than competitor presence at multiple time scales
The behavior and abundance of sympatric predators can be affected by a complex dominance hierarchy. The strength of antagonistic interactions in predator communities is difficult to study and remains poorly understood for many predator assemblages. Predators directly and indirectly influence the broader ecosystem, so identifying the relative importance of competition, prey, and habitat in shaping predator interactions has broad conservation and management implications. We investigated space use among five predator species (black bear [Ursus americanus], bobcat [Lynx rufus], coyote [Canis latrans], mountain lion [Puma concolor], and gray wolf [Canis lupus]) across three temporal scales in northern Idaho, USA. We used camera trap data to test whether potentially subordinate predators spatially avoided dominant predators and how prey availability influenced those relationships. We found few instances of subordinate predators spatially avoiding dominant predators and only at the finest temporal scale of our analyses. Instead, habitat features generally influenced predator space use patterns at coarser scales whereas prey and competitor presence influenced space use patterns at finer scales. Co‐occurrence was positively associated between coyotes and bobcats at coarser timescales and between mesopredators and apex predators at finer timescales. Bobcats and mountain lions temporarily delayed the use of sites recently visited by coyotes and black bears, respectively. And all predator species used sites sooner following the detection of a competitor in areas with higher relative abundances of prey (primarily white‐tailed deer [Odocoileus virginianus]). Our results suggest attraction to shared habitats and prey resources influenced space use in the predator community more than avoidance of competitors. We propose that the effects of interspecific interactions on predator distributions were most evident for mesopredators because their trophic position requires balancing risks and rewards associated with prey, apex predators, and other mesopredators. In addition, relatively high densities of a common prey source likely facilitated the spatial coexistence in this predator community. Our study demonstrates the value of simultaneously assessing multiple interspecific interactions across different spatiotemporal scales to discern relationships within the predator guild.
- 30 January 2025
- Published by Wiley in Ecological Monographs
- Vol. 95 (1)
Mammalian predator co‐occurrence affected by prey and habitat more than competitor presence at multiple time scales
The behavior and abundance of sympatric predators can be affected by a complex dominance hierarchy. The strength of antagonistic interactions in predator communities is difficult to study and remains poorly understood for many predator assemblages. Predators directly and indirectly influence the broader ecosystem, so identifying the relative importance of competition, prey, and habitat in shaping predator interactions has broad conservation and management implications. We investigated space use among five predator species (black bear [Ursus americanus], bobcat [Lynx rufus], coyote [Canis latrans], mountain lion [Puma concolor], and gray wolf [Canis lupus]) across three temporal scales in northern Idaho, USA. We used camera trap data to test whether potentially subordinate predators spatially avoided dominant predators and how prey availability influenced those relationships. We found few instances of subordinate predators spatially avoiding dominant predators and only at the finest temporal scale of our analyses. Instead, habitat features generally influenced predator space use patterns at coarser scales whereas prey and competitor presence influenced space use patterns at finer scales. Co‐occurrence was positively associated between coyotes and bobcats at coarser timescales and between mesopredators and apex predators at finer timescales. Bobcats and mountain lions temporarily delayed the use of sites recently visited by coyotes and black bears, respectively. And all predator species used sites sooner following the detection of a competitor in areas with higher relative abundances of prey (primarily white‐tailed deer [Odocoileus virginianus]). Our results suggest attraction to shared habitats and prey resources influenced space use in the predator community more than avoidance of competitors. We propose that the effects of interspecific interactions on predator distributions were most evident for mesopredators because their trophic position requires balancing risks and rewards associated with prey, apex predators, and other mesopredators. In addition, relatively high densities of a common prey source likely facilitated the spatial coexistence in this predator community. Our study demonstrates the value of simultaneously assessing multiple interspecific interactions across different spatiotemporal scales to discern relationships within the predator guild.
- Open Access
Intrinsic and environmental drivers of pairwise cohesion in wild Canis social groups
Animals within social groups respond to costs and benefits of sociality by adjusting the proportion of time they spend in close proximity to other individuals in the group (cohesion). Variation in cohesion between individuals, in turn, shapes important group‐level processes such as subgroup formation and fission–fusion dynamics. Although critical to animal sociality, a comprehensive understanding of the factors influencing cohesion remains a gap in our knowledge of cooperative behavior in animals. We tracked 574 individuals from six species within the genus Canis in 15 countries on four continents with GPS telemetry to estimate the time that pairs of individuals within social groups spent in close proximity and test hypotheses regarding drivers of cohesion. Pairs of social canids (Canis spp.) varied widely in the proportion of time they spent together (5%–100%) during seasonal monitoring periods relative to both intrinsic characteristics and environmental conditions. The majority of our data came from three species of wolves (gray wolves, eastern wolves, and red wolves) and coyotes. For these species, cohesion within social groups was greatest between breeding pairs and varied seasonally as the nature of cooperative activities changed relative to annual life history patterns. Across species, wolves were more cohesive than coyotes. For wolves, pairs were less cohesive in larger groups, and when suitable, small prey was present reflecting the constraints of food resources and intragroup competition on social associations. Pair cohesion in wolves declined with increased anthropogenic modification of the landscape and greater climatic variability, underscoring challenges for conserving social top predators in a changing world. We show that pairwise cohesion in social groups varies strongly both within and across Canis species, as individuals respond to changing ecological context defined by resources, competition, and anthropogenic disturbance. Our work highlights that cohesion is a highly plastic component of animal sociality that holds significant promise for elucidating ecological and evolutionary mechanisms underlying cooperative behavior.
Open AccessIntrinsic and environmental drivers of pairwise cohesion in wild Canis social groups
Animals within social groups respond to costs and benefits of sociality by adjusting the proportion of time they spend in close proximity to other individuals in the group (cohesion). Variation in cohesion between individuals, in turn, shapes important group‐level processes such as subgroup formation and fission–fusion dynamics. Although critical to animal sociality, a comprehensive understanding of the factors influencing cohesion remains a gap in our knowledge of cooperative behavior in animals. We tracked 574 individuals from six species within the genus Canis in 15 countries on four continents with GPS telemetry to estimate the time that pairs of individuals within social groups spent in close proximity and test hypotheses regarding drivers of cohesion. Pairs of social canids (Canis spp.) varied widely in the proportion of time they spent together (5%–100%) during seasonal monitoring periods relative to both intrinsic characteristics and environmental conditions. The majority of our data came from three species of wolves (gray wolves, eastern wolves, and red wolves) and coyotes. For these species, cohesion within social groups was greatest between breeding pairs and varied seasonally as the nature of cooperative activities changed relative to annual life history patterns. Across species, wolves were more cohesive than coyotes. For wolves, pairs were less cohesive in larger groups, and when suitable, small prey was present reflecting the constraints of food resources and intragroup competition on social associations. Pair cohesion in wolves declined with increased anthropogenic modification of the landscape and greater climatic variability, underscoring challenges for conserving social top predators in a changing world. We show that pairwise cohesion in social groups varies strongly both within and across Canis species, as individuals respond to changing ecological context defined by resources, competition, and anthropogenic disturbance. Our work highlights that cohesion is a highly plastic component of animal sociality that holds significant promise for elucidating ecological and evolutionary mechanisms underlying cooperative behavior.
- Open Access
Delineating wildfire burns and regrowth using satellite imagery to assess moose (Alces alces) spatial responses to burns
Climate change is predicted to increase the frequency and size of wildfires. Wildfire burns can influence wildlife space use, but prior to understanding the relationship between wildlife and burns, the extent of wildfires and the regrowth of burned areas must be determined. We used remotely sensed, multispectral satellite imagery to calculate differenced normalized burn ratios (NBRs) to define areas burned by wildfire and assess temporal trends in burn frequency and extent in central British Columbia. We then used NBRs as an index of vegetation regrowth to model regrowth trajectories within burns. Next, we examined spatial responses to burns by a moose (Alces alces) population that recently declined in abundance. We utilized locations used by and available to collared moose to determine spatial responses as a function of years since wildfire and vegetation regrowth. We also assessed functional responses to burns by moose, dependent upon the proportion of burned areas available to each individual. We hypothesized that (h1) wildfire burns within our study area increased from 1985 to 2017, (h2) that moose increase their use of burns in comparison with availability as burns age as a result of vegetation regrowth, and (h3) that moose demonstrate functional responses to burns (i.e., selection of burns by individual moose declines as the amount of burned areas increases). We observed that the frequency and extent of burns are increasing within our study area. In contrast to h2, moose in summer and fall decreased the selection of burns as burns aged. We, however, observed that moose responded most positively to locations with intermediate vegetation regrowth (i.e., NBR values), which according to burn regrowth trajectories were most likely to be achieved 13 or more years post‐wildfire; given that most burns within our study area were ≤13 years of age, we predict that the use of burns will likely increase as vegetation regrows. Moose only selected for burns in fall, and we did not find support for the presence of a functional response. Our research demonstrated the utility of remotely sensed imagery and NBRs to define burn locations and to reveal current and probable future spatial responses to burns by a wildlife population of concern.
Open AccessDelineating wildfire burns and regrowth using satellite imagery to assess moose (Alces alces) spatial responses to burns
Climate change is predicted to increase the frequency and size of wildfires. Wildfire burns can influence wildlife space use, but prior to understanding the relationship between wildlife and burns, the extent of wildfires and the regrowth of burned areas must be determined. We used remotely sensed, multispectral satellite imagery to calculate differenced normalized burn ratios (NBRs) to define areas burned by wildfire and assess temporal trends in burn frequency and extent in central British Columbia. We then used NBRs as an index of vegetation regrowth to model regrowth trajectories within burns. Next, we examined spatial responses to burns by a moose (Alces alces) population that recently declined in abundance. We utilized locations used by and available to collared moose to determine spatial responses as a function of years since wildfire and vegetation regrowth. We also assessed functional responses to burns by moose, dependent upon the proportion of burned areas available to each individual. We hypothesized that (h1) wildfire burns within our study area increased from 1985 to 2017, (h2) that moose increase their use of burns in comparison with availability as burns age as a result of vegetation regrowth, and (h3) that moose demonstrate functional responses to burns (i.e., selection of burns by individual moose declines as the amount of burned areas increases). We observed that the frequency and extent of burns are increasing within our study area. In contrast to h2, moose in summer and fall decreased the selection of burns as burns aged. We, however, observed that moose responded most positively to locations with intermediate vegetation regrowth (i.e., NBR values), which according to burn regrowth trajectories were most likely to be achieved 13 or more years post‐wildfire; given that most burns within our study area were ≤13 years of age, we predict that the use of burns will likely increase as vegetation regrows. Moose only selected for burns in fall, and we did not find support for the presence of a functional response. Our research demonstrated the utility of remotely sensed imagery and NBRs to define burn locations and to reveal current and probable future spatial responses to burns by a wildlife population of concern.
- Top Cited Papers
- 8 July 2022
- Published by Elsevier in Trends in Ecology & Evolution
- Vol. 37 (10) , 911-925
Dynamic landscapes of fear: understanding spatiotemporal risk
- 8 July 2022
- Published by Elsevier in Trends in Ecology & Evolution
- Vol. 37 (10) , 911-925
Top Cited PapersDynamic landscapes of fear: understanding spatiotemporal risk
- 1 October 2019
- Published by Canadian Science Publishing in Canadian Journal of Zoology
- Vol. 97 (10) , 866-879
Moose, caribou, and fire: have we got it right yet?
Natural disturbance plays a key role in shaping community dynamics. Within Canadian boreal forests, the dominant form of natural disturbance is fire, and its effects are thought to influence the dynamics between moose (Alces alces (Linnaeus, 1758)) and the boreal ecotype of woodland caribou (Rangifer tarandus caribou (Gmelin, 1788)). Boreal caribou are considered “threatened” and population declines are attributed, at least in part, to disturbance-mediated apparent competition (DMAC) with moose. Here, we tested a primary prediction of the DMAC hypothesis: that moose respond positively to burns within and adjacent to the caribou range. We assessed moose selection for ≤25-year-old burns (when selection is predicted to be strongest) at multiple spatial scales and evaluated whether moose density was correlated with the extent of ≤40-year-old burns (a time frame predicted to negatively affect caribou). Against expectation, moose showed avoidance and low use of ≤25-year-old burns at all scales, regardless of burn age, season, and type of land cover burned. These findings mirrored the demographic response, as we found no correlation between ≤40-year-old burns and moose density. By contradicting the prevailing hypothesis linking fires to caribou population declines, our results highlight the need to understand regional variation in disturbance impacts on caribou populations.
- 1 October 2019
- Published by Canadian Science Publishing in Canadian Journal of Zoology
- Vol. 97 (10) , 866-879
Moose, caribou, and fire: have we got it right yet?
Natural disturbance plays a key role in shaping community dynamics. Within Canadian boreal forests, the dominant form of natural disturbance is fire, and its effects are thought to influence the dynamics between moose (Alces alces (Linnaeus, 1758)) and the boreal ecotype of woodland caribou (Rangifer tarandus caribou (Gmelin, 1788)). Boreal caribou are considered “threatened” and population declines are attributed, at least in part, to disturbance-mediated apparent competition (DMAC) with moose. Here, we tested a primary prediction of the DMAC hypothesis: that moose respond positively to burns within and adjacent to the caribou range. We assessed moose selection for ≤25-year-old burns (when selection is predicted to be strongest) at multiple spatial scales and evaluated whether moose density was correlated with the extent of ≤40-year-old burns (a time frame predicted to negatively affect caribou). Against expectation, moose showed avoidance and low use of ≤25-year-old burns at all scales, regardless of burn age, season, and type of land cover burned. These findings mirrored the demographic response, as we found no correlation between ≤40-year-old burns and moose density. By contradicting the prevailing hypothesis linking fires to caribou population declines, our results highlight the need to understand regional variation in disturbance impacts on caribou populations.
- 18 September 2019
- Published by Springer Nature in Landscape Ecology
- Vol. 34 (11) , 2575-2597
Functional responses to anthropogenic linear features in a complex predator-multi-prey system
- 18 September 2019
- Published by Springer Nature in Landscape Ecology
- Vol. 34 (11) , 2575-2597
Functional responses to anthropogenic linear features in a complex predator-multi-prey system
- Open Access
Modeling multispecies predator–prey dynamics: predicting the outcomes of conservation actions for woodland caribou
Woodland caribou is a Threatened or Endangered subspecies across much of Canada. In many cases, these caribou are declining because of human‐mediated predation in the form of apparent competition. Provincial and federal agencies have employed a number of conservation actions to arrest the decline, but insufficient time and limited replication make it difficult to test the efficacy of those activities. We built a three‐species population model that served as a tool to integrate current knowledge, explore uncertainties, and assess activities focused on the recovery of caribou. We applied the model to interconnected populations of caribou, moose, and wolves at two locations in Canada (Chinchaga [British Columbia], Charlevoix [Québec]) and explored the efficacy of conservation actions unique to each population: predator control, restoration of linear features, reduction of habitat for moose, supplementation, maternal pens, and predator exclosures. Results confirmed that caribou at these two locations faced different recovery challenges, dictated by specific population dynamics and threats. Wolf predation, enhanced by seismic lines and resource roads, was responsible for the decline of the simulated Chinchaga population. The most cost‐effective recovery actions for that population were long‐term lethal wolf control (λ1–50 = 1.014, $25,665/caribou), a large‐scale predator exclosure (λ1–50 = 1.015, $170,767/caribou), and the aggressive restoration of linear features (λ1–50 = 1.002, $531,675/caribou). In contrast, simulations suggested that calf predation by black bears, a fixed source of mortality in the model, limited the growth of the Charlevoix population. Assuming high survival of adult caribou and poor recruitment of juveniles, a maternal pen was the most effective action for the recovery of those caribou (λ1–50 = 1.004, $148,473/caribou). Short‐term population supplementation provided only a temporary increase in abundance (λ1–50 = 0.994, $62,143/caribou). The model was limited by ecological and data uncertainties, but served as an effective platform for representing and testing our understanding of the complex interspecific interactions that underlie the recovery of woodland caribou.
Open AccessModeling multispecies predator–prey dynamics: predicting the outcomes of conservation actions for woodland caribou
Woodland caribou is a Threatened or Endangered subspecies across much of Canada. In many cases, these caribou are declining because of human‐mediated predation in the form of apparent competition. Provincial and federal agencies have employed a number of conservation actions to arrest the decline, but insufficient time and limited replication make it difficult to test the efficacy of those activities. We built a three‐species population model that served as a tool to integrate current knowledge, explore uncertainties, and assess activities focused on the recovery of caribou. We applied the model to interconnected populations of caribou, moose, and wolves at two locations in Canada (Chinchaga [British Columbia], Charlevoix [Québec]) and explored the efficacy of conservation actions unique to each population: predator control, restoration of linear features, reduction of habitat for moose, supplementation, maternal pens, and predator exclosures. Results confirmed that caribou at these two locations faced different recovery challenges, dictated by specific population dynamics and threats. Wolf predation, enhanced by seismic lines and resource roads, was responsible for the decline of the simulated Chinchaga population. The most cost‐effective recovery actions for that population were long‐term lethal wolf control (λ1–50 = 1.014, $25,665/caribou), a large‐scale predator exclosure (λ1–50 = 1.015, $170,767/caribou), and the aggressive restoration of linear features (λ1–50 = 1.002, $531,675/caribou). In contrast, simulations suggested that calf predation by black bears, a fixed source of mortality in the model, limited the growth of the Charlevoix population. Assuming high survival of adult caribou and poor recruitment of juveniles, a maternal pen was the most effective action for the recovery of those caribou (λ1–50 = 1.004, $148,473/caribou). Short‐term population supplementation provided only a temporary increase in abundance (λ1–50 = 0.994, $62,143/caribou). The model was limited by ecological and data uncertainties, but served as an effective platform for representing and testing our understanding of the complex interspecific interactions that underlie the recovery of woodland caribou.
- Open Access
- 1 February 2019
- Published by Wiley in Wildlife Biology
- Vol. 2019 (1)
Intrinsic traits of woodland caribou Rangifer tarandus caribou calves depredated by black bears Ursus americanus and coyotes Canis latrans
For prey species, individuals in substandard physical condition are assumed to be more vulnerable to predation (Errington 1946) potentially resulting from behaviours that increase risk or via a lesser capacity to escape predators (Lima and Dill 1990, Curio 1993). Although numerous studies demonstrate increased predation on individuals that are perceived to be in worse physical condition (Kunkel et al. 1999, Wright et al. 2006, Krumm et al. 2009), others fail to support this trend (Karanth and Sundquist 1995, Anderson et al. 2007). The life history characteristics of prey and predator explain some of the variability underlying the distribution of depredated prey with regards to individual physical condition. For example, disproportionate predation on substandard individuals is theorized to be more likely for prey species that are difficult to catch (Temple 1987, Wirsing 2002). Predator hunting strategies, however, are also thought to play a role in determining which individuals are depredated from a prey population (Michalko and Pekár 2016).
- 1 February 2019
- Published by Wiley in Wildlife Biology
- Vol. 2019 (1)
Open AccessIntrinsic traits of woodland caribou Rangifer tarandus caribou calves depredated by black bears Ursus americanus and coyotes Canis latrans
For prey species, individuals in substandard physical condition are assumed to be more vulnerable to predation (Errington 1946) potentially resulting from behaviours that increase risk or via a lesser capacity to escape predators (Lima and Dill 1990, Curio 1993). Although numerous studies demonstrate increased predation on individuals that are perceived to be in worse physical condition (Kunkel et al. 1999, Wright et al. 2006, Krumm et al. 2009), others fail to support this trend (Karanth and Sundquist 1995, Anderson et al. 2007). The life history characteristics of prey and predator explain some of the variability underlying the distribution of depredated prey with regards to individual physical condition. For example, disproportionate predation on substandard individuals is theorized to be more likely for prey species that are difficult to catch (Temple 1987, Wirsing 2002). Predator hunting strategies, however, are also thought to play a role in determining which individuals are depredated from a prey population (Michalko and Pekár 2016).
- 1 December 2018
- Published by Elsevier in Biological Conservation
- Vol. 228, 215-223
Predation risk for boreal woodland caribou in human-modified landscapes: Evidence of wolf spatial responses independent of apparent competition
- 1 December 2018
- Published by Elsevier in Biological Conservation
- Vol. 228, 215-223
Predation risk for boreal woodland caribou in human-modified landscapes: Evidence of wolf spatial responses independent of apparent competition