A comparative analysis of mountain lion (Puma concolor) predation site properties between Yellowstone
National Park and Snoqualmie National Forest
School of Environmental and Forest Sciences, College of the Environment, University of Washington
Connor J. Meyer
Introduction:
Mountain lions (Puma concolor) are found widely dispersed throughout the western hemisphere,
boasting a range from the Canadian Yukon to the southern tip of South America (Hornocker et al. 2009).
Their generalist nature and extreme adaptability allows them to survive in many different landscapes with
other top predators, including wolves (Canis lupus), bears (Ursus spp.) and jaguars (Panthera onca). With
wolves re-colonizing the west, cougars and wolves will once again compete for niche space in ecosystems
in which wolves have been absent for ~70 years. Both species are able to subsist in the same ecosystem,
however, in Washington State, wolves will likely displace cougars as they re-colonize, leading to a potential
increase in human-cougar interactions and perhaps an decline in cougar population numbers.
For my project I set out to compare predation site properties between two different ecosystems in
which cougars are found, with the hopes of predicting how cougars in the west may react to wolves. The
western study site is located in parts of Snoqualmie National Forest, King County and Snohomish County.
Yellowstone National Park will serve as the wolf-present study site. These two locations differ in more than
just wolf presence and absence, as Snoqualmie is characterized by large swaths of forest with patches of
clear cuts and an extensive road network. Yellowstone is quite the contrary as it is a large open landscape
with occasional patches of forest. By comparing slope, canopy cover at two different radii and sorting the
kill habitat into one of three (open, edge, closed) categories I hope to determine if cougars are using similar
attributes of the landscape. This would help inform a prediction as to how much re-colonizing wolves are
going to affect already existing cougar populations.
Results: Part I: Snoqualmie
Figure 1: Three Box and whisker plots, each showing Individual cougar ID (n=6) (characterized by letter informing gender F, and a unique number) plotted against a 95% confidence interval of slope (left), canopy cover (%) within a 15 meter radius of each kill site (middle) and canopy cover (%) within a 200 meter radius of each kill site (right).
Primary ANOVA testing shows that slope (p=0.037), percent canopy cover at 15 meters (p-value= 0.053) and percent canopy cover at 200 meters (0.038) all vary when comparing individual cougars in Snoqualmie
National Forest. Although all are statistically significant, in actuality percent canopy cover does not look significantly different (i.e. at a 15 meter radius, four of the six cougars have made nearly all kills within 100%
cover). MANOVA testing that paired slope and canopy cover also showed that there is variability between individual cougar predation sites in Snoqualmie. With a p-value of 0.012 for the 200 meter radius and a
p-value of 0.0091 for the 15 meter radius. The chi-squared test on the number of kills in either open, edge or closed cover areas, there is little variation as cougars tend to perform the most kills in areas deemed
closed (p-value=0.16)
Part II: Snoqualmie vs. Yellowstone
Google Earth imaging showing the two study sites , the distance between them as well as the visible difference in land-cover between the two points. Courtesy of Google Earth
Figure 2: Three Box and whisker plots, each showing data collection location plotted against a 95% confidence interval of slope (left), canopy cover (%) within a 15 meter radius of each kill site (middle) and canopy cover (%) within a 200 meter radius of each kill site (right).
Methods:
ANOVA testing showed that slope (p-value= 2.51x10-10), percent canopy cover at 15 meters (p-value= 6.11x10-7) and percent canopy cover at 200 meters (p-value= 6.65x10-14) were varied drastically when comparing
the two different habitats. Perhaps showing exactly how versatile of a predator cougars are, the percent canopy cover at the 15 meter range varied from 0% all the way up to kills made in full cover (100%). Almost every
Data was obtained in two different ways depending on study site.
-13
-16
recorded
kill
in
Snoqualmie
occurred
at
a
slope
greater
than
those
in
Yellowstone.
MANOVA
testing
showed
similar
results,
both
the
15
meter
(p-value=
2.47
x10
)
and
the
200
meter
(p-value=
2.2x10
) were response
• In Snoqualmie cougars are captured, fitted with a GPS collar and then tracked to determine kill sites.
Two or more GPS pings in one area is indicative of a predation (Anderson et al. 2003)and after observing variables along with slope. The categorization of kill sites also proved to vary between the two sites as the p-value was 0.006.
a “cluster” of GPS points a team was sent in to collect data.
• In Yellowstone, researchers would take advantage of heavy snow to track cougar movement until disPredation site properties were shown to vary both within a single study site and between both study sites. Cougars in Snoqualmie make kills in areas
covering a predation site where data would be recorded.
where slope varied significantly, although all were shown to have a high percentage of canopy cover. This could be due to niche differentiation or the
solitary nature of cougars that is not conducive to sharing space between individuals. Based on previous studies, I would have expected cougars in YelOnce all data was collected methods were as followed:
lowstone to have made kills in areas that had more canopy cover and an increased slope (Elbroch et al. 2015; Ruth et al. 2009) compared to Snoqualmie
• Overlay predation sites onto slope and vegetation GIS layers to determine slope and percent canopy
due to wolf presence. However, the Yellowstone killsites proved to be on flatter ground with an average incline of 14.82 degrees while Snoqualmie had
cover for each site.
a mean incline of 61.32 degrees. Yellowstone kills were also made in areas with less overall percent canopy cover. This is most likely due a lack of avail• Use Google Earth imaging closest to date of predation to categorize kill site habitat with either “open”,
able canopy cover and the fact that this data was collected in the winter, when snow forces cougars to make kills closer to edges (Laudre et al. 2000). The
“edge” or “closed”.
habitat in which the Yellowstone kills were made is also conducive to wolf presence , therefore cougar predation site characteristics may be more closely
• Analyze data using MANOVA, ANOVA and Chi-squared tests. First comparing individual cougars in
linked to climate and landscape features than the presence of a competitor.
Snoqualmie, and then compare Snoqualmie to Yellowstone.
Although cougars in Snoqualmie and Yellowstone use the landscape in different ways, predation site properties at the two study sites shows just how
versatile and adaptable cougars can be. With the ability to make a kill near a suburban backyard (Snoqualmie; Kertson et al. 2011) or the middle of an
open plain (Yellowstone) cougars clearly demonstrate their ability to remain a part of thelandscape. As wolves re-colonize Washington, I believe that
they will displace cougars to a degree, but the Yellowstone data shows that cougars are very adept at continuing to make kills in areas conducive to wolf
presence, therefore cougars should be able to adequately compete with wolves and remain within their current range.
Discussion:
Future Research:
A satellite image of F17 predation sites location in Snoqualmie National Forest. Courtesy of
Google Earth.
YNP cougar predation sites overlayed upon a slope layer.
Typical predation site in Snoqualmie National Forest. Photo courtesy of Connor Meyer
• Under-story canopy cover may be more indicative of how cougars are using the landscape, a study looking at the under-story cover
at each kill site may give more insight into cougar hunting strategy.
• Re-colonization of wolves gives researchers a rare opportunity
to study how the two predators interact and how one may displace
the other. A study site looking at both species’ movement and killsites would be ideal for determining how each affect the others niche
differentiation.
• GPS collaring of cougars in Yellowstone will provide more data
as well as allow for nesting of individual cougar kills.
A cougar (F17) captured and processed in Snoqualmie National Forest. Photo courtesy of Brian Kertson.
Acknowledgments:
First and foremost I would like to thank my advisor, Dr. John Marzluff, for his guidance and advice on my senior thesis. For the
cougar data and GIS layers related to Snoqualmie National Forest as well as advice throughout I would like to thank Dr. Brian
Kertson, M.S. student Clint Robins, and the entire volunteer research team. For supplying and allowing me to use the Yellowstone
National Park cougar data I would like to thank Dr. Daniel Stahler and the entire Yellowstone Cougar Project research team. For the
Yellowstone GIS layers I would like to thank Dr. Monika Moskall, Intermap, Precision Forestry Cooperative, Ania Jakola, Caileigh
Shoot and Jonny Dunne.
References:
Anderson Jr., C. R., & Lindzey, F. G. (2003). Estimating cougar predation rates from GPS location clusters. The Journal of Wildlife
Management, 307-316.
Elbrock, M. L., Lendrum, P. E., Newby, J., Quigley, H., & Thompson, D. J. (2015). Recolonizing wolves influence the realized niche
of resident cougars. Zoological Studies, 41.
Hornocker, M., & Negri, S. (2009). Cougar: Ecology and Conservation. Chicago: The University of Chicago Press.
Kertson, B. N., Spencer, R. D., Marzluff, J. M., Hepinstall-Cymerman, J., & Grue, C. E. (2011). Cougar space use and movements in
the wildland-urban landscapes of Western Washington. Ecological Applications, 2866-2881.
Laundre, J., & Hernandez, L. (2000). Habitat composition of successful kill sites for lions in southeastern Idaho and northwestern
Utah. In proceedings of the sixth mountain lion workshop.
Kill site investigation team led by M.S. student Clint Robins (right) after success- Ruth, T. K., & Murphy, K. (2009). Competition with Other Carnivores for Prey. In M. Hornocker, & S. Negri, Cougar: Ecology and
Conservation (pp. 164-172). Chicago: The University of Chicago Press.
fully processing a kill. Photo courtesy of Sasha Hendricks