Predation and scavenging of salmon carcasses along spawning streams in the Scottish Highlands

Salmon carcasses were recorded during October December along spawning streams in 3 study areas: 1 upper Shee Water, R. Tay system (2000 & 2001); 2. Upper Clunie Water, Aberdeenshire Dee system (2000 [part of area] and 2001) and 3. Kinlochewe River, R. Ewe system (2001 only). Estimates of numbers of salmon within each study area are based on fish observations, redd counts and counts at fish traps. Interpretations of "predated" vs. "scavenged dead" salmon carcasses were based on field observations of carcass condition, carcass location; amounts and parts of carcasses consumed; and from foot prints, trails, and other signs. Additional data for Kinlochewe relates to the movements and fate of 12 radio-tagged salmon within the study area. Otters were found to be present in all study areas. Estimated mean densities of spawning salmon were 50 fish/km for Shee (both years), 20 fish/km for Clunie (both years) and 10 fish/km fr Kinlochewe (2001). The majority of carcasses were of males, and many were taken dead or near dead from the water towrds the end of the spawning period. Remains of females containing eggs were recorded from all three study areas. Contemporaneous otter spraints containing large salmonid eggs were recorded locations within 400m of over 75% of these carcasses in Clunie and Kinlochewe, and over 20% of these carcasses in Shee (where spraints were collected from only one bank, once per week). Crude minimum estimates for predated preor partspawned female salmon are: 10% (2000) and 6% (2001) for Shee, 10% (2001) for Clunie and 10% (2001) for Kinlochewe of respective total adult female salmon populations. Salmonid remains were the most frequently recorded item in otter spraints collected during the study period in all study areas. Both large and small (trout or salmon parr) salmonid remains were found. Otters were the principle scavengers of carcasses, although fox, mink and pine marten were also recorded. The vulnerability of salmon to otter predation is likely to vary according to stream channel morphology, predator and scavenger densities, and salmon and alternative prey densities. Levels of predation by otters may be elevated near natural or artificial obstructions to fish migration, and where there is little cover for adult fish. Salmon may become increasingly vulnerable to predation by otters as streams become shallower toward their upper reaches. These factors should be considered when estimating salmon spawning targets for stream systems and when developing recommendations for the management of freshwater habitat for salmon.