Odd/even length differences of Alaska sockeye, Chinook, chum and coho salmon (1962–2016) did not increase with the odd/even difference in North Pacific pink salmon catch

Pink salmon are much more abundant in odd years in most of the North Pacific, and authors use this biennial cycle as evidence that pink salmon competition reduces the size of other salmon. We did a test of the dose-response that this hypothesis predicts. We used 1,951 length-at-age series of sockeye, Chinook, chum and coho salmon from 12 Alaska regions (ADF&G data compiled on KNB, 18,607 three-year contrasts, 1963–2015) and NPAFC commercial pink catch of all countries. For each series and year we calculated the odd/even half difference of log length around a local linear trend, and the same quantity for pink catch, which changed from near 0 to about 100 million fish between years. Odd/even length differences were present but small, with a regional sign: Bering Sea sockeye +0.44% (95% year-block bootstrap CI 0.32 to 0.64) and Chinook +0.24% longer in odd years; Gulf of Alaska coho −1.11% (−1.51 to −0.85) and chum −0.77% shorter. In region fixed-effect models, these four signals did not change with the pink odd/even catch difference: slopes were near zero (for example Gulf coho −0.03% per 10 million fish, CI −0.15 to +0.10), a dose-proportional effect was outside each 95% interval, and the intercepts were different from zero. Log catch, Asian or North American catch and a one-year lag gave the same result; in sample subsets, only Chinook escapement samples gave a slope different from zero. The only clear slope (Bering coho) had the sign opposite to competition. If the slope measures a one-season effect, the rise of mean pink catch from 112 to 296 million fish (1962–1976 to 2002–2016) explains at most 1.1% of the 4.4% decrease of Bering sockeye length-at-age and at most 0.6% of the 6.1% decrease of Gulf Chinook. Limits: catch is only a proxy for pink abundance at sea; the odd/even contrast does not measure effects that add over several ocean years; sexes are pooled and sampling dates are not known; data end in 2016.

Pacific salmon · pink salmon · competition · body size decline · length-at-age · biennial cycle · natural experiment · dose-response · Alaska · Oncorhynchus · density dependence · fisheries