Nesting density and reproductive success in a colonial seabird, the magellanic penguin

A central tenet of density-dependent habitat selection theory, that individ ual fitness generally declines with increasing population density, has led to the hypothesis that, for colonial species, living at high density produc es advantages that outweigh the costs. Seabirds have provided evidence for this hypothesis, and positive effects of high density have been offered as explanations for coloniality in this overwhelmingly colonial group. We inve stigated effects of nesting density on reproductive success of a colonial s eabird that nests at variable densities, the Magellanic Penguin (Spheniscus magellanicus). Using breeding success data collected from 1984 to 1991 as part of a long-term study of this species at a large colony at Punta Tombo, Argentina, and experiments that included manipulation of nesting density, we addressed effects of nesting density at two spatial scales: the within-c olony area scale (hundreds to thousands of square meters) and the very smal l scale of a penguin's immediate surroundings (distance to nearest neighbor ). Nest density at the colony averaged 8.7 active nests/100 m(2) and ranged fr om <1 to >50 active nests/100 m(2) (n = 873 plots). In 100-m(2) plots sampl ed from 1987 to 1990 over large areas of the colony of similar average nest quality and distance inland, the number of fledglings per active nest was inversely correlated with nest density. Experienced breeders using high-qua lity nests in a high-density study area (>20 nests/100 m(2)) had 0.25 fewer fledglings per active nest than similar birds nesting in a low-density are a (<6 nests/100 m(2)). Penguin fights (during which eggs and chicks may be lost), nonbreeding "prospectors" (who can start fights and kill chicks), an d nest predation were all significantly more prevalent in the high-density area. Within a low-density area, birds nesting within touching distance of neighbors (<less than or equal to>80 cm) had significantly lower success th an those with more distant neighbors. Nests where fighting occurred were si gnificantly closer to their neighbors than were nests where no evidence of fighting was observed. We found no association between predation and distan ce to nearest neighbor. Experimental data were consistent with correlative results. In a plot with reduced density, fledging success was higher (P < 0.10) and fewer agonistic behaviors and predators were observed than in higher density plots. Chicke n eggs placed in vacant nests remained untouched by predators longer in low -density areas of the colony than in high-density areas. All results indicated that yearly fledging success declines with increasing density at both the local and area scale, and that rates of predation and intraspecific aggression increase with density. The consistently negative e ffect of increased density begs the question of why Magellanic Penguins are colonial. Possible explanations include benefits to lifetime reproductive success not reflected in yearly hedging success (e.g., mate and nest acquis ition or retention) and benefits of aggregation at spatial scales larger th an those investigated in this study.