I think I just witnessed a gang war between crows. It was a little before ten in the morning, on a day with intermittent flurries. At first I thought the crows were mobbing a predator – that’s what it sounded like – but then I realized they were diving at each other in the sky over Sapsucker Ridge to the west. More birds kept streaming overhead from the east all the while I stood watching, about ten minutes. Most of them landed in the treetops and added their voices to the raucous cheering section for the aerial battles, which included at least a couple dozen combats at a time; the ridge blocked my view of a lot of the action. Brisk winds aloft made for an exciting display of maneuvers: diving, chasing, feinting – Top Gun stuff, for sure.
As is usual with crows, it’s difficult to know how to interpret what I was watching. Maybe it was all just play behavior, occasioned by the wind conditions. Ten minutes later, when I stick my head out the door, there’s no sign of a crow anywhere.
Most non-specialists would probably tend to assume that scientists know a lot about the behavior and life histories of the commoner birds and mammals, especially here in the northeastern U.S., but such is rarely the case. A recent study of white-tailed deer, for example, made a couple surprising discoveries simply by fastening digital cameras to bucks’ antlers: deer touch muzzles constantly, they found. And one deer kept returning to the same spot to drink, despite the availability of closer water sources. What’s surprising about this is that white-tailed deer are probably one of the most-studied animals in the world, apart from human beings, laboratory rats, and fruit flies. How could previous researchers have failed to notice such apparently common behaviors?
So I wasn’t sanguine about throwing much light on what I’d just seen by a quick glance through the literature on the American crow (Corvus brachyrhynchos). The Stokes Guide to Bird Behavior, Volume 1, confirms that “There has been amazingly little study of Crows by researchers, so most of the best questions about the birds still cannot be answered.” Its summary of research to date (1979) included this:
Outside of the few months of the breeding season, Crows are extremely gregarious. After the last young have fledged, the family group usually joins other groups of Crows, and these begin to form a large flock that divides up for feeding during the day, but gathers again each night to roost. The roost becomes an important focal point in the birds’ life outside the breeding season. Each morning the roost breaks up into smaller flocks that disperse across the countryside to feed. Some flocks may fly up to fifty miles from the roost each day. In midafternoon these smaller flocks start back toward the communal roost. They fly along flight lines used each day and are joined by other flocks as they go. Often there are preroosting sites, where flight lines coincide and Crows stop to feed before making the final trip to the roost. At these spots there may be much chasing and other spectacular dives as the returning Crows join the others at the preroosting spot. Then just before dusk all the Crows in the area enter the roost site together.
What I saw would seem to have occurred at the wrong time of day for preroosting behavior. But this description does make it clear that not all chasing and diving is antagonistic.
In The American Crow and the Common Raven, Lawrence Kilham describes territorial antagonism across a boundary between two large groups of crows in Lyme, New Hampshire.
I watched forty-five territorial encounters beween 1981 and 1983. While no two were alike, certain behaviors were observed repeatedly. Cawing, heard in thirty-eight of the territorial encounters consisted of sharp caws corresponding to what Good (1957) refers to as warning calls and Camberlain and Cornwell (1971) as simple scolding calls. The crows of both groups cawed when flying toward each other at a distance. This was especially so on early mornings when most encounters took place. If one group alighted in the fields and another in trees, the latter did the most cawing….Other forms of behavior [aside from walking and bluffing displays] included aerial melees (n = 17), bunching (n = 5), and pursuits (n = 5). The melees were spectacular when all members of both groups swirled into the air for three or four seconds, with some swooping on others….
Territorial behaviors also included circular flights (n = 11), which carried the crows of one group a short distance over the boundary into their neighbors’ territory as if to demonstrate where the boundary lay, and treetop sitting (n = 17). The latter was particularly striking on November 26  when the crows of both groups, after a series of melees, perched on the very tops of their dead elms in full view of each other.
Kilham found that groups facing off over common boundaries tended to be “of the same or about the same size.” None of the conflicts he observed seem to have resulted in serious injury, let alone death. And he speculates that most of the actual combats are left to the dominant or nuclear males, based on his observations of American crows in Florida, and on another scientist’s observations of white-winged choughs, which have similar social patterns.
A more recent source of information is the American Crow monograph for the authoritative Birds of North America series: No. 647, by N. A. M. Verbeek and C. Caffrey, 2002. Over twenty years after Stokes, not much has changed: “Although much has been published about this [species of] crow, we still know relatively little about it.” They add that observations made in one locale may not describe the behavior of crows from somewhere else in the species’ range.
One of the things that appears to vary from place to place is whether territories are maintained throughout the year. Crows on Cape Cod and in New Jersey, California, Oklahoma, Florida and New York do maintain year-round territories, but crows in Ohio and in the northern part of their range in Canada do not. So goodness knows where central Pennsylvania fits. On the topic of aerial melees, Verbeek and Caffrey simply quote Kilham. If there haven’t been many other observations in the literature, that may be because few observers are as patient as he was (me, I went inside when my hands began to freeze). Also, given that American crows favor mixed and open habitat, territorial boundaries might tend to pass through forested areas or follow wooded ridgetops, making interactions along them harder to observe.
As habitat generalists and omnivores with well-developed learning abilities and complex social structures, crows offer many parallels to the behavior of humans and other primates. To a non-scientist like me, it seems natural to characterize group antagonistic behavior as a gang war, but I have to be careful not to let my judgements be too colored by prejudices which are not merely anthropocentric, but ethnocentric as well. For example, we Westerners tend to associate wars with struggles for dominance, leading ideally to the conquest of one group by another. But among the crows that Kilham observed, territorial conflicts seemed to work more to maintain a balance of power – not unlike the low-intensity conflicts that are thought to have been the norm for warfare in much of native North America before 1492. Kilham watched wandering flocks numbering as many as forty crows trespass into group territories without much conflict beyond a lot of cawing. These flocks apparently consisted of nonbreeders – juveniles with no interest in setting up a territory of their own.
It’s also important to remember that territories are maintained cooperatively, for the shared benefit of the crows that use it: “Territories provide improved protection through greater familiarity with safe areas; reduced interference with nest building and copulations; greater protection of stored food; and increased assurance of a good food supply,” Kilham notes. Nor does cooperative behavior end when the breeding season is over. Fall and winter roost sites can only get to be as large as they are by bringing together crows from diverse, non-overlapping daytime feeding territories, and mixing local birds with migrants – foreigners from the north. I wonder if the diving and chasing behavior commonly observed at preroosting sites doesn’t represent a playful, ritualistic form of inter-group conflict? After all, even among humans, the line between play and warfare can get awfully fuzzy – think of the World Cup. In any case, whether we choose to focus on the relatively rare, spectacular outbreaks of crow-to-crow combat, or on the cooperative nature of crow sociability as a whole, probably says a lot more about us than it does about the crows.
The most promising research on American crows since Kilham comes from Kevin McGowan of Cornell University, who has been banding crows and following their progress from year to year since 1988. His crow website includes helpful information on roosting behavior in the F.A.Q. page, and an overview of his own crow study, with links to publications. He writes, “I frequently see crows locked together tumbling out of trees in the spring. Although I have never witnessed an actual killing, I would not be at all surprised to see crows kill another crow from outside the family group that was trespassing.”