The diet of otters was studied over one season on the Mendocino National Forest in northern California. Undigested prey remains were analyzed from 94 otter scats collected between 18 April and 29 August 1982. Fish were found to be the predominant food, occurring in 89.4% of the scats with a mean volume estimate of 63%. Fishes consumed were primarily catostomids and salmonids. Crayfish and snakes were common prey items as well. Insects were often found in otter scats, but repre- sented little volume. Bird, turtle, frog, and snail remains occasionally were found in the scats. Feeding habits of the North American river otter (Lutra canadensis) have received considerable attention in the literature of wildlife biology, with special emphasis often placed upon the relative proportions of game and non-game fish species taken. One approach has been to study the remains of prey in stomachs or intestines of otters taken by commercial trappers (Lagler and Ostenson 1942; Ryder 1955; Hamilton 1961; Toweill 1974, Lauhachinda 1977). While this method permits relatively accurate esti- mations of prey size and numbers (Lagler and Ostenson 1942), it requires grave com- mitment on the part of the otter. In areas where otter are fully protected, analysis of scats is a workable alternative to digestive tract analysis (Trautman 1936; Wilson 1954; Greer 1955; Sheldon and Toll 1964; Knudsen and Hale 1968; Modafferi and Yocom 1980; Melquist and Hornocker 1983). Scat analysis has certain inherent shortcomings. First, the digestive process renders even some hard parts less identifiable. Second, hard indigestible portions of prey are over- represented compared to soft parts. Mech ( 1970) hypothesized that because of this, smaller prey with greater surface area:volume ratios would be over-represented in scats of carnivores. Floyd et al. (1978) tested this idea with captive wolves and found that smaller animals were indeed over-represented in frequencies of remains in scats. Last, assessment of feeding habits by frequency of occurrence in scats has been shown to be inherently biased, over-estimating food items occurring in low proportions and under-estimating at high proportions (Lockie 1959). Because of this last problem, it has been recommended (Korschgen 1980) that some estimate of volume or weight be included with frequency of occurrence data in all feeding habits studies. Lagler and Ostenson (1942) estimated volumes of otter stomach contents, and Modafferi and Yocom (1980) measured volumes of food items found in otter scats. This communication presents feeding habits information obtained from river otter scats collected in 1982 on the Mendocino National Forest in northwestern California. Data are expressed as percent frequency of occurrence and mean percent estimated volume.
Read full abstract