At the end of the day, I love it. All of it, even the hard bits.
I’ve been waiting a long time to write this blog post… over 3 years in fact. The entire premise of my PhD has been to answer this central question – how do grizzly bears use habitat around hiking trails of various levels of human use in the Rocky Mountain National Parks? I can now share the answers I’ve found.
Attempts to answer the question
To investigate this question, I used 4 years of GPS collar data from 27 individual grizzly bears and data generated from 159 different remote camera sites on hiking trails in Banff, Kootenay, and Yoho National Parks. The dataset is huge involving many spreadsheets and more of my brain than has ever been active. I have literally invested blood, sweat, and tears (and a lot of each of these things) into answering this question.
I used the remote camera data to do a presence-absence analysis predicting when and where grizzly bears would be caught on camera using human use trails. This kind of analysis basically looks at all the times I captured a grizzly bear on camera and all the times I didn’t capture a grizzly bear on camera and tries to predict what exactly was happening on all the occasions I caught a bear (on camera). I specifically wanted to know how season, trail human use level, trail type (hiking vs biking etc), the cameras’s distance to road, and habitat quality would predict the likelihood of catching a bear. I also used the remote camera data to examine if there was a threshold number of people using trails that related to when bears would use those same trails.
I used the GPS collar data to conduct a use-availability analysis looking at how bears use habitat in their home ranges. This is pretty similar to a presence-absence analysis in that I basically compared where bears were (I had a GPS location) and where they weren’t (through random locations in their home ranges where I didn’t have locations). I then compared the attributes of the used locations to the random locations; specifically I wanted to know how habitat quality, distance to road or trail, and season impacted use.
I also ran a Step Selection Function (SSF) to look at grizzly bear movement. This analysis is kind of like a use-availability analysis except it looks at grizzly bear steps, which is the process of a grizzly bear moving from one location to another. A SSF creates random steps and compares them to used steps – it looks at where a grizzly bear ended up vs. all the options it had. I wanted to know how distance to high and low human use trails, habitat quality, time of day, and distance to roads impacted a grizzly bear’s steps.
From the remote cameras
The remote cameras showed that grizzly bears were more likely to be captured on trails in the spring and summer than the fall. Habitat quality, distance to road, and the number of human events did not influence whether or not a bear used a trail. Bears use trails as movement corridors, so habitat attributes didn’t necessarily relate to how they move through the landscape.
The threshold analyses showed that most grizzly bear events on trails take place in the night or wee hours of the morning, before 8 groups of people used the trail. They were also more likely to use trails that had less than 17 groups of people in 24 hours. So while the number of human events over 3 weeks did not influence if bears were captured on camera, the patterns on a finer time scale suggest they were more likely to be captured on trails before people used them.
A grizzly bear on the trail at night. I have a few of these night time pics!
From the GPS collar data – Use-availability
Grizzly bears continually selected for higher quality habitat throughout the seasons. This is no surprise, bears go where the food is. Within their seasonal home ranges, grizzly bears selected for habitat farther from roads but closer to trails that were close to roads in the spring. In the fall, grizzly bears selected habitat closer to roads and trails than expected. This pattern may be a reflection of where their home ranges are. In the spring, grizzly bear home ranges are small and in the valley bottoms, where the green vegetation is starting to appear. In these valley bottom home ranges, they are selecting for habitat far from roads but much of the early green-up is near trails so they are somehow striking a balance between being far from roads while using habitat near trails that are closer to roads. In the fall, grizzly bear home ranges are larger and in the higher elevations. Even though they are selecting habitat closer to trails and roads than random, their whole home range is farther from roads and trails so bears aren’t necessarily closer to trails in the fall. This is supported by the remote camera data that found bears less likely to be captured on trails in the fall.
Overall, grizzly bears used habitat closer to roads in the spring than the fall, and used habitat closer to roads and trails in the summer than the fall.
You might start to see a trend here… spring is a big deal. Bears use habitat closer to trails in the spring, they are more likely to be captured on trails in the spring, and they don’t have anywhere else to go in the spring because the higher elevations are covered in snow.
From the GPS collar data – Step Selection Functions (SSF)
The one obvious thing from the step selection function is that no bear is like other bears. They are all individuals. Some bears selected steps closer to trails, some closer to roads, some farther from trails and roads, and other selected steps close to low human use trails and far from high human use trails. This individual variation made this analysis incredibly challenging. Bears basically do what they want when they want. Most bears continually selected for high quality habitat, wherever it was in their home range. This means that some bears are willing to accept the increased risk associated with being close to roads and trails to access high quality habitat.
A little subadult that I photographed next to the Icefields Parkway. Right next to the road.
The most striking result from the SSF is that bear movement rates increased when they were close to high use trails, particularly during the spring. This could be because bears are being displaced by people using the trail and are moving in and out of high quality close to the trail, or it could be that they aren’t hanging out around high use trails and just pass through those areas. Either way, their use of high quality habitat near high use trails could be impacted.
What’s the big deal
There are risks for bears that select habitat or steps close to roads and trails.
Increased mortality – Bears that use habitat close to roads have a higher chance of being struck by a vehicle.
Increased risk of habituation – Bears that use habitat closer to roads and trails have a higher chance of becoming habituated to people, which can lead to increased mortality when and if they leave the National Parks and enter into an non-protected landscape.
Increased movement = increased expended energy = less food consumed = unhappy bear. Bears who move more around trails might not be eating as much, which means they run the risk of being less healthy. There are all kinds of implications from this ranging from bear size to reproductive success.
Bears take steps to reduce their chance of an encounter with people by using trails in the early morning or at night, exhibiting some ability to learn and to use habitat when people aren’t there.
The most obvious thing in my remote camera data was that people are everywhere in the parks all the time. Grizzly bears living in Banff, Kootenay, and Yoho are continually sharing their habitat with people and they’re doing a pretty good job at it. But there are risks to these bears using this habitat and we need to consider that when we are playing in their home ranges.
In my next post, I’ll try and piece this all together – bears and people and coexistence.
For now, I’ll finish off this PhD and look forward to another summer on the trails in the wilderness!