Reproductive success, food supply, and the evolution of clutch-size in the Glaucous-winged gull

Lack (1947) proposed that, in nidicolous birds, clutch-size has evolved to correspond, on average, to the most productive brood size. The limit is normally set by the maximum number of young the adults can adequately feed to fledging. Recent studies using gulls to test Lack's hypothesis have shown that the most common and most productive clutch-size do not coincide. Recent increases in human refuse may have been a factor in these results. In this study, the Glaucous-winged Gull Larus glancescens was used to test Lack's hypothesis. Experiments were designed to test the possible effects of refuse on the birds' capabilities of raising extra young. Both normal (1-3 chicks) and supernormal (4-6 chicks) broods were set up on both a colony (Mandarte Island) where refuse was used by the gulls and on colonies (Cleland Island and islands (QCI) in the Queen Charlotte Islands, B.C.) where refuse was not used. The results did not support Lack's hypothesis. Chicks grew better on Cleland and QCI, where only natural food was used, than on Mandarte where both refuse and natural food was used. On Cleland, chicks in all brood sizes reached an average weight of 1000 g (adult weight) before fledging. On Mandarte, the maximum weight was significantly below 1000 g for most brood sizes. Numbers of chicks fledged for each brood size increased with increasing brood size on all the colonies. Post-fledging survival rates indicated that on Cleland, chick survival was similar for normal and supernormal broods. On Mandarte, chick survival was better for the normal broods than for supernormal broods. The contribution to future breeding populations by individuals from different brood sizes was highest for a brood of six on Cleland, but was highest for a brood of three on Mandarte. The results are contrary to what Lack predicted both because a supernormal brood on Cleland was the most productive and because on Mandarte refuse did not give the birds an advantage over those colonies where refuse was not available. On Cleland, Pacific Sandlance was the predominant food. On Mandarte Pacific Herring was the predominant food. On Mandarte in 1971, I found that refuse formed up to 25 percent of the chick diet even though significantly more time was required by the adults when foraging for refuse as opposed to natural foods. The duration of the average foraging trip increased with chick age but this was due to an increased use of refuse by the adults as the chicks got older. Reasons for the high success on Cleland and QCI are discussed including both the possiblility of a recent change in the abundance of sandlance and the possible influence of reproductive effort on adult mortality. A winter study was carried out in south-western British Columbia in order to assess the use made of refuse sites during the winter by the Glaucous-winged Gull. I found that up to 65,000 glaucous-winged gulls wintered in the lower mainland region of British Columbia and that between 70 and 90 percent of these birds were using refuse sites. Relatively few birds were using the intertidal zone, possibly because it is not exposed to any extent during daylight hours. The numbers of gulls in this area are discussed in relation to known information on the total population along the west coast of North America.Science, Faculty ofZoology, Department ofGraduat