Proximate and ultimate causes of adoption in ring-billed gulls

From 1987 to 1994, the annual frequency of adoption by breeding pairs at a Lake Erie ring-billed gull, Larus delawarensis, colony ranged from 3 to 37% (average 8%/year, N=7 years) and, on average, foster parents raised 0.5 fewer of their own chicks to fledging than pairs that did not adopt. The key evolutionary question is: why do some individuals apparently suffer the costs associated with a behaviour that appears to enhance the fitness of others? From 1992 to 1995, I investigated the proximate and ultimate causes of adoption in ring-billed gulls from the perspectives of both the foster parents and adopted chicks, and tested predictions that differentiated between various competing adaptive and nonadaptive hypotheses that have been proposed to explain it. While I was able to demonstrate a breeding cost, I failed to identify any benefits to foster parents. Thus, the adaptive hypotheses that rely on the foster parents benefiting were not supported (e.g. kin selection, reciprocal altruism, acquisition of parenting experience). From the foster parent's perspective, adoption was mediated through errors in parent-offspring recognition. Under natural conditions, most fostering pairs were tending small chicks (<6 days old) at the time of adoptions; in chick-transfer experiments, resident parents did not discriminate against foreign chicks until their own chicks were 7-9 days old. Chicks (N=25) that subsequently abandoned their natal nests were lighter, and grew at a slower rate, than chicks that survived to fledging in their home broods. Thus, departing chicks were at a survival disadvantage in their home broods. Chicks that gained acceptance into foreign broods where they were older/larger than the resident chicks realized high survival at the expense of their foster siblings and parents. Based upon individual growth rates and the corresponding survival probabilities, disadvantaged chicks approximately doubled their survival chances through foster care. Why has selection not eliminated adoption? I argue that adoption is an evolutionary arms race between the two principle actor groups; disadvantaged chicks, which benefit through foster care, and host parents, which avoid providing foster care (e.g. infanticide). In ring-billed gulls, selection has failed to eliminate adoption because the long-term reproductive cost (estimated at 4%, this study) of an occasional adoption is probably offset by the relatively higher costs associated with stricter kin discrimination mechanisms (e.g. parental infanticide). (c) 1998 The Association for the Study of Animal Behaviour.