REPRODUCTIVE ECOLOGY OF THE WESTERN GULL: THE IMPORTANCE OF NEST SPACING GEORGE L . H UNT, JR., AND MOLLY w. H UNT GULL colonies usually exhibit considerable variation between pairs in both territory size and timing of breeding. This variation is subject to natural selection, and timing of breeding and territory size can be cor- related with reproductive success ( Paynter 1949; Paludan 195 1; Ver- meer 1963, 1970; Brown 1967; Kadlec and D rury 1968 ; Kadlec et al. Patterson ( 1965) and Kruuk ( 1964) showed that for the small Black- headed Gull (Larus ridibundus ) the principal source of egg and chick mortality was predation. For these birds, optimum predator defense re- sulted from nesting during that period when the maximum number of birds were breeding. Pairs breeding either earlier or later suffered greater loss of chicks to predation. Patterson ( 1965) also postulated that there should be an optimal spacing of nests, based on a balance between clump- ing of nests for enhancement of group defense and scattering of nests for effectiveness of the cryptic coloring of eggs and chicks (Tinbergen et al. For the larger Larus gulls neighboring pairs may be a more important source of chick loss than predators (Paynter 1949, Emlen 1956, Parsons 1971). For these species large territory size may reduce the chance that a chick will trespass into a neighbor's territory. It has also been found that adults defend their territories more actively when they have chicks than when they are incubating (Hunt and Hunt MS, also this study, see below). Thus in these species early nesting may be advantageous as a means of minimizing chick loss to neighboring pairs. In light of the conflicting selective pressures on territory size and tim- ing of breeding, it is reasonable to expect that the optimum territory size and time for breeding will vary from one colony to another, depending upon the relative importance of predators or neighbors for chick loss. The colony of Western Gulls (Larus occidentalis wymani) nesting on Santa Barbara I sland, California, provided an opportunity to test this hypothesis, as the only potential chick predators there are adult Western Gulls. In the present study we identified the causes of chick mortality and related the reproductive success of pairs to aspects of spacing of nests and timing. STUDY SITE AND METHODS We conducted our study on Santa Barbara Island, Channel Islands National Monu- ment, Santa Barbara County, California. Santa Barbara Island is a 260-ha (1 square mile) island 61 km (38 miles) from the nearest mainland and 39 km (24 miles) from The Auk 92: 270-279. April 1975