Question

What evidence did Marzluff find about how ravens communicate while they are roosting at night?

Answer 1

In the winter, ravens in the northern New England forests feed from carcasses, a highly patchy rich food resource. As argued elsewhere (Heinrich 1988b, 1989, Marzluff et al. !994), the use of these carcasses by large groups that accumulate is best explained by re- cruitment from communal roosts. These roosts serve as overnight resting places near food (Caccamise and Morrison !986, Morrison and Caccamise !990). How- ever, they are formed there (apparently because they reduce commuting costs) only after recruitment from a distant location. That is, without the information exchange occurring at a distant roost, there would be no large build-up of birds at the food and, subse- quently, no roost near it. Communal roosting may be a conditional strategy, used primarily when food is temporarily abundant in patches (Davies 1986). Pre- sumably, if food were less patchy, birds could always remain in the same roost and feed at the closest feed- ing site(s).

Like Rabenold (!987), I argue that recruitment could be demonstrated by large numbers of birds arriving in groups shortly after the bait was discovered. Fur- thermore, the data suggest a mechanism: following the early bird to leave the roost.  

Almost all of the large groups of ravens came early in the feeding cycle. Furthermore, they only came prior to sunrise. The data are consistent with the idea that birds anxious to feed, yet not knowledgeable of food, follow those leaving early. Those leaving early likely have a predetermined destination. However, the meat at a very large carcass lasts a number of days and, after the birds know of its location, they no longer need to arrive before dawn. They can now leave on their own at any time. In accordance with this prediction I found that although the size of the feeding group soon stabilizes, the times of arrival of the individuals get spread out, and birds begin to arrive as individuals, pairs, or in small groups.  

Counter to the above expectation, it could also be reasoned that as more birds become knowledgeable of a carcass, the more the distribution of arrivals be- comes skewed to earlier in the morning. However, satiation likely mitigates against this. As already in- dicated, late in the feeding cycle birds often pass by in the morning or loiter nearby for hours without feeding at all. The more well-fed a bird is (i.e. the more it has knowledge of food bonanzas), the less likely it may be inclined to head directly to a food source 0.5 h before daylight. Adult ravens under the weather conditions of my study (concurrently examined) have an average mass loss of only 45 g (or 3.5% body mass) per day under food deprivation, and they regain all of this mass in 5 h or less of feeding (Heinrich 1994a). Thus, ravens that have fed all day or until late in the evening are presumably not as highly motivated to feed the next morning as those that have not fed for a week or more. One unanswered question is how it is possible that several hundred birds can learn of the bait, whereas roost size in the study area is almost always less than 100, and usually less than 50 (pets. obs.; pets. comm. from Frank Oatman, George Lisi, Brett Engstrom and Clarence Nutting, Marzluff et al. !994). Therefore, over 300 birds feeding at a carcass cannot be accounted for by recruitment from the local roost alone. More than one roost may then be consolidated, or several roosts may partake of the same bait.