Mark Elbroch-Ecological impacts of recolonizing wolves on a hunted cougar population

Mark Elbroch-Ecological impacts of recolonizing wolves on a hunted cougar population

Posted by


– There is tremendous
interest in the relationships between mountain lions and wolves. We in the southern Yellowstone ecosystem have been working on these
questions for many, many years. This thing I’m gonna present on right now is a collaborative effort
between three organizations. Howard Quigley and myself from Panthera. Lucille Marescot and Heiko Wittmer from Victoria University in Wellington. And Derek Craighead from Beringia South. This work in particular, we feel, really contributes to that ever-growing body of research on the ecological impacts of wolves on natural systems. Here in the GYE and further afield, that work has really emphasized
the wolf as a predator, and their direct and
indirect effects on prey, which, of course, result
in the controversies around trophic cascades. But, in recent years, there’s also been an expansion of focus. And it’s looking at the
wolf not only as a predator, but as a dominant competitor. The wolf is just one carnivore,
as Dan just emphasized, in a system of many carnivores. The influence of wolves
on other carnivores have results that echo
through the ecosystem. It’s one thing to look at these effects, both direct and indirect, of wolves within the
boundaries of national parks, where the influence of
people has been minimized. But it’s quite another
to look at those effects as they start to intermingle with other management strategies. Wolves have successfully
recolonized and expanded beyond Yellowstone, and
now they’re intermingling, for instance, with hunting quotas. Some research has shown
that as wolves re-colonize areas where they were formally removed, that the combined effects of these wolves and human hunters can sometimes result in dramatic and unanticipated
declines in that prey. And our question, really, was, would this also be true
for mountain lions? Now, mountain lions
aren’t prey for wolves, but they’re subordinate competitors in the presence of wolves. And not only that, they’re
a trophy game species. Throughout Wyoming, where we were, but, in fact, throughout much of the west, where they can be legally harvested. Our hypotheses were quite simple, that the dual effects of new competition with re-colonizing wolves, in combination with existing
pressures from human hunters, would result in decline of the local mountain lion population. More specifically, we expected that wolves would have the most impact
on young mountain lions, or smaller mountain lion survivorship, whereas humans would, in
contrast, have the greatest impact on adult
survivorship probabilities. Our work is conducted in the south. We are east of Grand Teton,
south of Yellowstone. Importantly, it is
predominantly national forest. We work in the Bridger-Teton, but we also overlap into Grand Teton, and on the National Elk Refuge. That rectangle is sort
of where we capture cats, and of course, they go further afield. This analysis has many, many moving parts, and I’m gonna try to move through quickly and highlight the key points
so you can follow along. Capture and Monitoring. Like everybody else that
captures and puts colors on mountain lions, we
capture adults and juveniles. But one of the things that’s
distinctive that we’ve done, and Tony Ruth did here, was to really look at kitten survivorship. So, we began to capture
kittens at 5-7 weeks, and to follow them through time as well. When you look at, when you
put a collar on an animal, and you follow them through time, what’s really important is that you often end with their fate. And we look at the causes of mortality. So, we go into the field,
we do gross necropsies, we work with veterinarians, we send tissue and blood
samples off for disease. And then we can use all of this data, the monitoring, how long
they’re in the project, how they end up, their fate. And we can place that in
capture-mark-recapture models. We used a program called E-Surge, and I’m just gonna mention that because it’s fairly new in the
mammal research world. For those who would like
to follow up and learn how this is different than
Program MARK or others, I listed this right here. It’s just about to come
out in population biology, it’s not my work, it’s Lucille Marescot and Heiko Wittmer and several others, but it’s on black-tailed deer. And there you can sort
of compare and contrast with Program MARK and other programs. But importantly, what do these models do? Two things. First, they came up with the survival probabilities
for mountain lions. Second, they come up with
cause-specific mortality rates. How often does a certain
cause of mortality occur within the population? We categorized mortality
into five big groups. Human-caused, that’s
legal hunting, poaching, wildlife-management action, roadkill, predation by any predator, starvation, other natural causes, which for us was predominantly disease. And then, when we couldn’t figure it out. We took all of that, we took
our survival probabilities, and now we go to the next
stage of the analysis, and we want to know what explains it. Why is it changing over
time? Things like that. We have numerous
co-variants that we look at. For instance, are survivorship for males different than females? Is it different for different
age classes of mountain lions? So, we get three. Kittens,
juveniles and adults. Rather than four seasons,
we used two seasons. We used the “hunt” season, which ranges from September
1st to March 31st, that’s the legal hunting
season in Wyoming, and the non-hunt season. And realized that fall and winters, the typical attributes of these seasons, deep snow, aggregated
prey, things like that, are all part of that hunt season, and the non-hunt season would be the spring and summer characteristics
of prey in mountain lions. Wolves, we looked at this
in two different ways. First, we created two temporal categories, and we used a cutoff date
of January 1st, 2006. And the reason is, from
2005-2006, in our study system, the number of wolves doubled. We called before this time
period, “low wolf density”, and on the other side,
“high wolf density”. The other thing we did
is we added an annual minimum counts reported
by US Fish and Wildlife. And we used that as a co-variant, a continuous co-variant in the model. Last, we included cougar densities, annual change for those numbers. And that’s to assess the possibility that in high-density years,
there’d be more strife between mountain lions, and of course there’s the
possibility of infanticide, which, in mountain lions, is typically males killing kittens. Then, we needed a measure of fecundity, which essentially is just, what’s the mean litter
size in our study area? How often does a female give birth? We take all of those
numbers and we place them in the next stage of our analysis, the Leslie matricies, and
they serve two purposes. First, we want it to quantify, or calculate population growth. The big question, has
the population changed? Is it flat, no change over the many years? Has it declined? Has it grown? Second, we can do what is
called “perturbation analyses”. This is a way of manipulating
those cause-specific mortality rates to assess the impact of different forms of mortality
on population growth. We can also play with those numbers if we wanted to change population growth. For instance, for management, sort of forecasting how we might want to bring up mountain lion numbers, or decrease mountain lion numbers. And we can play with those numbers in a perturbation analysis. Once more around, over 13 years, we monitored 124 different mountain lions. 74 kittens, 35 adults. We identified the cause of death for 67 of the 91 mountain lions that died during that study period. We threw that all into these models, we came up with two top models, meaning they performed equally well, and so we stuck with both of them. We referred to one as the “hunt model”, and the other as the “wolf model”. And you can see that neither of them included sex as a driver in
variation and survivorship. There was no different between male and female survivorship. However, there was a big difference in kittens vs. juveniles vs. adults. Kittens have horrible,
horrible survivorship, and so that was part of
both of these models. The first one emphasized the hunt season, so it said that survivorship was radically lower in the hunt season
vs. the non-hunt season. And the second one, was
that temporal category of low and high wolf density. You can see that, and
in fact, I can tell you that survivorship was
worse when wolves were up. Some of the other models did well, and I thought I’d just share one. This one, on the bottom there, is number of wolves in the system. 25, 50, 75. You can see that as wolves
increased in the system, that in every age category,
survivorship went down. Across the board, every
age of mountain lions decreased in survivorship. We took our mean litter size of three, our interbirth rate, meaning
how often females give birth. That’s 26.5, we actually
changed that to 24 months, to meet the needs of our next matrice. And, we came up with two growth rates, based on our two different top models. Based on the hunt model,
we came up with a lambda, that’s that funny little symbol, of 0.86. Doing the wolf model,
we divided it from when it was low density of 0.94, and at high density it was at 0.86. For those unfamiliar with lambda, if it equals 1, that means the population has stabled, no change. If it’s above 1, it means it’s growing. And in the case of all
three of these examples, when it’s below 1, it
means it’s in decline. Here’s what it looks like on the ground. In red, these are, it’s the log, but these are annual cougar densities each year, of the study. And in blue, these are
minimum wolf counts. Again, the log, but these
are the annual wolf counts for the study system
during those same years. You can see, this big “X” means that the two numbers are correlated. What are our conclusions? Well, we felt that, just like when wolves were turned to an area,
and then their effects on prey are combined with
the effects of human hunters, that you can see a decline
in the local prey population. We believe we saw the exact
same thing with mountain lions. Yes, the dual impacts of new competition with wolves, when added with the current and existing effects of human hunters on local mountain lions, has resulted in their decline. Yes, wolves had the highest impact on young or small mountain lion survivorship. Just in the last 3-4 years, we’ve lost six kittens to wolves. Yes, humans have the highest impact on adult survivorship probabilities. 49% of the adults in the 13
years were killed by people. Based on these results, we then can work with our perturbation analysis to come up with management strategies. This is a big assumption, but assuming that we want to stabilize the local mountain lion population,
we want to stop the decline, what do we need to do? If we want to bring that
population to a lambda of 1, we can use the hunt model, and, say we can decrease human hunting by 60%, and that’s because the hunt model was most sensitive to adult survivorship during the hunting season. Alternatively, we could
use the other model and decrease predation on
mountain lion kittens by 80%, essentially remove 80% of the competitors from mountain lions in the system. The wolf model was most
sensitive to kitten survivorship. I’m gonna leave you with
one last ecological thought. I’m gonna compare what we’ve learned there with other cats around the world, and, in fact, mountain
lions in other places. Mountain lions here, in
fact, in the northern range when Tony Ruth was here. African lions, leopards, all of these cats are really characterized
by two forms of mortality. One is humans. Unsurprising there, a large carnivore. But the other is intraspecific,
meaning cat vs. cat, mountain lion vs. mountain lion, African lion vs. African lion, etc., etc. Whereas, in contrast, what we’ve seen in the southern Yellowstone ecosystem, yes, we share the fact that humans are one of the major drivers of mortality for mountain lions in the
southern Yellowstone ecosystem. But, different than everywhere else, it’s interspecific, meaning
cat vs. other predator. In our case, it’s mountain lion vs. wolf. Thanks to numerous
funders and collaborators, Game and Fish, Forest
Service, Park Service, etc. Thanks for listening. (applause)

Leave a Reply

Your email address will not be published. Required fields are marked *