Half of my PhD is understanding where the bears are, the other half is to understand where the people are. When I break my PhD into that simple sentence, it makes me wonder why things are so hard… In the past month, I’ve been focusing on analyzing the mountain of data that will help me understand where the bears are and how they are using their habitat. And I’m finally at a point to share some of what I’ve found, but that isn’t to say that I’ve figured out what it all means yet. (Step by step).
A Bear’s Home
All animals have what’s called a home range, which is basically where a critter lives and all the space it needs to access it’s food, shelter, and mates. My first step in data analysis was to look at where bears lived in the Rocky Mountain National Park landscape. I wanted to know not only exactly where their home ranges were, but how big they were, how close to people they were, and how that changed with the seasons.
The first thing I learned is that a bear’s home changes a lot throughout the year. Have a look at this map of one bear in Banff National Park.
The first thing I noticed looking at this map is the spring home range (pink) is much smaller than either the summer or the fall home range; the spring home range is also more road-side and in the valley bottom. This is largely due to the fact that most available habitat in the spring is in the low elevations and road-side since the snow hasn’t fully melted yet. The average home range size of a female grizzly bears in the spring is 22 km2, and a male’s is 119 km2 – that’s a big difference mostly due to the fact that males wander far and wide looking for females during mating season. In the summer, bears move into the higher elevations, a little farther from roads but they are still accessing some habitat in the lower elevations. In area, their home ranges grow quite a lot (123 km2 for females, and 184 km2 for males). In the fall, most this bear’s home range is in the higher elevations, away from roads. Home ranges appear to shrink a little in the fall (95 km2 for females, and 122 km2 for males). Caveat: I’ve put some numbers on here so you can get an idea, but these are just the averages – any scientist knows that putting an average out there without talking about the variation around that average isn’t statistically meaningful.
While the details change from bear to bear, the basic pattern holds true. None of that was really surprising, but it’s always nice to see a map. (ArcGIS is turning me into a big mapping geek).
Moving House… every season
A bear’s home range changes every season, but my question was is that change real, significant, and meaningful. I know the map above looks like it is, but that’s only one bear and maybe she’s exceptional. The whole purpose for stats is to see if a change we perceive or observe is actually real, and not just due to chance. So I ran some tests to see how seasonal home ranges change for bears in our National Parks. Because these home ranges change so much seasonally, all of these analyses are done on seasonal home ranges. To do this, I created a series of random points in the seasonal home range and compared those to the used points (the GPS locations) in the bear’s seasonal home range. What I found was a little surprising.
I thought bear’s would avoid trails and roads, but what I found was that bears used habitat closer to trails and roads than random in all seasons. Within that context, the relationship to trails was stronger than roads, so bears were closer to trails than random. That was a little surprising as I expected bears to avoid roads and trails. Comparing the seasons, I could see that the relationship between spring and distance to trails was the strongest, meaning that bears are using habitat closer to trails in the spring than in the other seasons. In the fall, this relationship is the weakest meaning that they are using habitat farther from trails during that season.
This makes it look like bears are ok with using habitat close to people, which likely some of them are. In the next test, what I really found was that distance to trails and roads wasn’t the most important factor in a bear’s decision making process.
Get in my belly – a bear’s living mantra
I’ve talked a lot (as have most bear researchers) about how habitat quality is the most important factor determining where a bear will live. My results support that too. While they may select habitat closer to trails and roads than expected, it’s really habitat quality (where the food is) that determines where bears are. When I put the Use vs. Available (used vs. random points) results into this context, I can see that regardless of season bears select habitat of high quality and that factor is 100x more influential than whether or not that habitat is close to a road or trail.
Basically, if there’s food there and its plentiful, a bear is more than happy to enjoy it and fill his or her belly.
When I looked at distance to road/trail and habitat quality in the same test, I did find that bears are closer to roads and trails than expected in summer and spring. Again this relationship was stronger with trails than roads, which means bears are happy to use habitat closer to trails than roads. That last one seems a little intuitive but now I have some numbers to show it (and that is so exciting)!
Incorporating a measure of human use
Throughout my data analysis, I am using roads and trails as a measure of human use. The above tests talk about how close a bear is to trails/roads when it’s using habitat in its home range. But obviously how many trails/roads exist in a bear’s home range is an important factor to consider. If a bear has a high trail or road density in its home range, then it will be closer to trails and roads than another bear. Statistical tests do show a difference between individual bears, which makes sense because some bears live mostly around towns and they will inherently have higher trail and road densities in their home ranges.
I more wanted to see if males were different from females. Some research from British Columbia has shown that females with cubs will select for habitat close to people as a refuge from dominant males, who may want to kill their cubs. I’m interested to see if that is happening here in the Rockies too.
When I looked at trail and road density between age/sex classes, I found that adult males have a lower trail density than females with cubs and adult females have a lower trail density than females with cubs. This suggests females with cubs are using habitat with a higher level of trails, which could mean a higher level of human activity. At this point I can’t say whether that is because of refuging or just because males get first choice of habitat and they like habitat with less people. I’ve got a whole array of other analyses to do before I can figure that out.
And here’s the graph that shows it all. It’s pretty obvious that spring trail and road density is the highest, but look at the size of those error bars! That means that there’s also a lot of variation in the spring, particularly for Females with Cubs (FwC) and Subadults. One of the reasons why there are significant differences between males and females is that there is also less variation between those bears, so differences are actual differences and not just due to chance.
What does it all mean?
Well at this point, I don’t really know so I’m not going to say. This first step in analysis has just confirmed things that other studies have shown to be true. Males have less trail density than females, home ranges expand over the seasons, and bears use habitat closer to trails in the spring.
My next step will be get more into the GPS data and see how bears are moving across this landscape, not just where they are. I’ll also start working with the camera data to see how people are using the landscape. And eventually it will all come together and be awesome.
Right now, it’s all in pieces but those pieces are starting to look pretty neat!