Wilson Bull., 109(4), 1997, pp. 748-749

The significance of fishing by Northwestern Crows.-Fishing by crows has been reported anecdotally in a few cases, specifically in the Carrion Crow (Corn,us corone) (Dunn 1985) and Northwestern Crow (C. caurinus) (Shepherd 1988). We describe here the behavior in greater detail and suggest that intertidal zone fishing provides a significant source of nutrition to Northwestern Crows during limited, but predictable foraging periods. On two occasions we have observed large groups of Northwestern Crows fishing systematically for Pacific sand lances (Ammodytes hexapterus) buried in sandy areas at very low tides.

The first observations were made on 25 August 1996 between 08:08 and 08:45 PST at Meadowdale Park Beach, north of Seattle, Washington on Puget Sound. The tide height ranged from .03 m above to -0.2 m below mean low tide during this period. The birds demonstrated a distinctive behavior of flipping sand to each side with their beak while digging a 5-8 cm hole in the sand. This usually resulted in the bird pulling out a silvery slender fish approximately 150 mm in length. The birds held the fish sideways in their bills initially, then set the fish on the sand, reoriented the fish lengthwise and swallowed it whole. This behavior was observed in the majority of birds on the beach at that time, numbering from 36 to 41. Three specific birds were observed for 10 minute focal samples, documenting their behavior both prior and subsequent to the capture of one of these fish. These observations indicated that a crow could find and handle a fish in less than five minutes. One of those three birds lost its prey to a conspecific: scrounging (intraspecific theft [Vickery et al. 1991) was observed at high rates during periods of fishing. Gulls also took sand lances from crows and were never observed digging for the fish themselves.

The second observation period was on 28 August 1996 between 10:20 and 11:23 PST at Meadowdale Park Beach. The tide height during this period ranged from 0.1 m above to -0.3 m below mean low tide. Again, a large majority of the 28 birds on the beach were observed to perform this behavior, though it only occurred in this narrow window of time and tide height. We observed 12 "scrounges" of sand lances during this period. Again, gulls took fish from crows and were not observed searching or digging for fish.

We eventually obtained one of the fish by running up to birds who were digging but had not yet obtained their fish. This sample was confirmed as a Pacific sand lance (Ammodytes hexapterus). The high rate of both intra- and interspecific theft suggested to us that the sand lances were highly valued by both crows and gulls. We were particularly interested in the value of a fish in satisfying a crow's daily energy requirements. The average daily energy requirement of a 285g crow (Pyle et al. 1987) can be estimated at 403.7 kJ/day (Walsberg 1983). The average length of a sand lance is 15.5 cm with a caloric content of 22.52 kJ/gdry weight (Vermeer & Devito 1986). We converted the average length into wet weight (Vermeer & Devito 1986) and converted wet weight to dry weight (0.29 g-dry/g-wet; Cooper 1978). This resulted in a value of 120.1 kJ/fish. In a similar fashion, the energy content of intertidal worms, a frequent intertidal food item (pers. obs.), (average dry-weight: 25mg) was calculated to be 0.42 kJ/worm (Goss-Custard 1977a, 1977b). Thus, the average crow needs only 3.4 sand lances/day to satisfy their daily energy requirements, whereas they would need 961.2 average-sized worms. Clearly these fish provide a valuable energetic resource. However, the fish are only available bimonthly at the lowest tides (Shepherd 1988). In our study area, these lowest tides (<0 m) occur for 4-8 successive days and last approximately 30-60 minutes each day.

From a more general point-of-view, this resource is unusual in that it is highly valuable and limited, but predictable in its temporal distribution. Most food resources that are highly rewarding and of limited duration are unpredictable in their occurrence (e.g., insect molting swarms). Perhaps this unusual combination of characteristics explains why the resource produces such intense intra- and interspecific competition and has been exploited by crows (and through the foraging abilities of crows, other predators) in both the Kodiak Archipelago (Shepherd 1988) and the Puget Sound region.

Acknowledgements-We thank Doug Dailer and Bill Karras of the Snohomish County Parks & Recreation Dept. for their support and encouragement of this project.