Behav Ecol Sociobiol (1990) 26:181-190 Behavioral Ecology and Sociobiology © Springer-Verlag 1990 The effects of nesting stage, sex, and type of predator on parental defense by killdeer (Charadrius vociferous): testing models of avian parental defense Dianne H. Brunton* School of Natural Resources, University of Michigan, Ann Arbor, Michigan, MI 48109-1115, USA Received July 19, 1988 / Accepted September 16, 1989 Summary. Two models predicting the temporal patterns of parental investment in offspring de- fense over the nesting cycle were tested. The first is based on offspring age, the other on the vulnera- bility of offspring to predation. Both models make very similar predictions for altricial species after eggs have hatched, i.e., increases in intensity of parental defense until fledging. For precocial spe- cies, however, the post-hatching predictions of each model are different: the offspring age model predicts a continued increase in defense intensity, while the vulnerability model predicts a decline. I examined the temporal patterns of parental de- fense of a precocial shorebird, the killdeer (Charad- rius vociferus), and determined which model was supported. Killdeer responses to human and natu- ral predators were observed. Killdeer were less will- ing to leave the nest, responded most intensely, and displayed closest to a potential predator around hatching. Defense intensity increased from early to late incubation as predicted by the off- spring age model. However, after hatching killdeer parental defense declined for both males and fe- males, thus supporting the vulnerability model for this stage. Males and females responded signifi- cantly differently to all types of predators. Males took greater risks, remained on the nest longer, defended offspring more intensely, and displayed closer to the predator than females at the approach of a potential predator. Responses to natural pre- dators depended on the type of predator and the approach made by the predator; a greater range of defense behavior was used for predators ap- proaching on the ground compared to aerial pre- dators. In general, killdeer responses to humans were more intense and less variable than their re- * Present address: Department of Biology, Yale University, OML, P.O. Box 6666, New Haven, CT 06511-7444, USA sponses to natural predators. This was most likely because the human intruder approached nests and chicks more directly and closer than natural preda- tors. Introduction Parental defense and its associated risks are an im- portant component of parental investment (Trivets 1972). Predation is a powerful selective pressure and in species with parental care, parents often enhance the survival of their offspring by actively defending them from predators (Gottfried 1979; Greig-Smith 1980; Gochfeld 1984). Defense behav- ior is assumed to be performed at some risk to the defending bird (Simmons 1955; Armstrong 1956; Kruuk 1964; Barash 1975; Andersson et al. 1980; Greig-Smith 1980; Waiters 1982; Weather- head 1982; Brunton 1986). Although distraction displays, in particular injury-feigning, appear to place the performing bird at a greater risk of preda- tion, the actual risks are difficult to assess and may be small. Few observations of mortality of displaying adults have been made (Jourdain 1936; Myers 1978; Brunton 1986). The alertness of the displaying bird is considered evidence of the small probability of mortality (Gochfeld 1984), i.e., an alert bird is unlikely to be captured by a potential predator. At the very least, defending parents are expending time and energy that cannot be ex- pended on somatic activities. Thus, selection should result in a level of defense that is a com- promise between the benefit to the current brood and the survival of the parent and future broods (Trivers 1972). The hypotheses developed within the cost/ben- efit framework of parental investment have been