It's been a while since I've written a science blog post, and not because I haven't been reading papers. On the contrary, I've had the exact opposite problem! I'd like to revisit some work done on moose on Isle Royal, Michigan, one of the best studied ecosystems in the US. Reading about the Isle Royal studies was part of what got me into biology to begin with.
Isle Royal is an island in Lake Superior, north of the tip of the Upper peninsula, and not far from the coast of Canada. It's around 70x15 km in size, and almost all of it is encompassed in Isle Royale National Park. The island is the result of geologic tilting, which results in long ridges and islets that all run in the same north-east sort of direction. It's a wonderfully beautiful area, and the site of much study between wolves and moose. There used to be caribou there as well, but they were expatriated. I would love to see them reintroduced there, but I'm a tad biased towards Caribou, aren't I?
Joan Edwards, now at Williams Collage, did regular surveys of the location of moose sightings. They patrolled moose trails, and did coastal surveys using a boat, and recording whether it was a bull, cow, whether it was alone, whether it had a calf, and so forth. Additionally, she surveyed the diet through observation, recording bouts greater than 10 minutes.
What she'd found is that there were quite the change in locations between cows with calves and all other moose. Bulls and cows without calves tended towards the ridges on the main island for the beginning of the growing season, before meandering their way towards the shoreline and the aforementioned small islets in July-Sept. This is to contrast with cows with calves, whom were very strongly associated with shoreline, or islets, from May through the remainder of the season. Moreover, they showed that the cows with calves had a very poor quality diet when compared to those without. I've reproduced a figure showing their distribution in various seasons, and you can really see that the cows with calves really had a strong association with those non-interior sites.
Now cow moose have a large investment in calves. It's not the pregnancy so much, but the lactation afterwards that really kicks them in their behinds. Whereas a male's fitness tends to come from its ability to cover multiple females, a females fitness comes from its ability to successfully bring the offspring to independence. Thus, a male is limited by its ability to snarf down forage and become large enough to be dominant, and a female is limited by her ability to trade off predation risk and her basic needs in terms of body condition. Additionally, you could infer that the variance in reproductive success in bulls (either they have a ton of success, or very very little) would lead them to very different risk-benefit calculations as to whether they want to eat in food-rich but predator heavy areas.
Put into this context, Joan Edwards' research makes quite a bit of sense. Each class appears to be trying to maximize their fitness. She throws out Calves are likely to imitate their mothers as an alternate explanation, but I think she's right to dismiss this alternative fairly easily, since they could just as easily find the same assemblage of forage elsewhere, and at greater quantities. Addationally, I haven't seen the behavioural patterning work go very far with moose, and it's quite possible that that hypothesis petered out in the intervening decades.
However, I do have some general criticisms. First, the study is dependant on sightings, but I saw no estimation of sightability. Supposedly the long duration of observation could rectify that, but in such a dense, closed forest, opportunities for missing animals are rife. Secondly, the sightings could have been systematically biased by one of two ways. First, the lowest predation risk areas for cows with calves on the mainland could have been the hardest to get to, due to terrain, vegetation cover, and other factors. Human bias in these studies can be sizeable, as no one likes to hang around in a hard to reach boggy area looking for moose. Secondly, the mere act of moving about could have scared the cows with calves off in the mainland, because of their heightened concern as to predators. Really, the way to do this sort of study the best would be through radio telemetry, aka collaring and fallering. All that said, I think the effect is generally a real one, albeit not as strong as Dr Edwards suggests.
Edwards, J. (1983). Diet shifts in moose due to predator avoidance Oecologia, 60 (2), 185-189 DOI: 10.1007/BF00379520