Survival rates of a neotropical parrot: Implications for latitudinal comparisons of avian demography

Latitudinal variation in avian demography played an important early role in the development of life history theory, especially in the idea of a cost o f reproduction. Recent attempts to determine the survivorship of tropical b irds with mark-recapture statistics have proved controversial. Here, we use a small neotropical bird, the Green-rumped parrotlet (Forpus passerinus), as a model system for investigating sources of heterogeneity that might bia s interspecific comparisons. Mark-resighting data were collected on 1334 ad ult parrotlets over a decade. We expected adult survival to be low because this parrot lays a large clutch (mean = 7 eggs), is a cavity nester, and br eeds in a highly seasonal environment. A two-age-class term in local surviv al was nonsignificant, indicating that an age or transience effect was unim portant. Local survival of males did not vary annually, but 19.3% of the ye arly variation in female survival was explained by rates of nest loss durin g stages when females were incubating or brooding young. The overall local survival rate of parrotlets (phi = 0.565) was identical to temperate hole-n esting species of the same body size but was lower than that of tropical bi rds that lay smaller clutches. However, we also detected considerable heter ogeneity in parrotlet survival. Females and males that were sighted but did not breed comprised a mean 23.5% and 52.9% of our population, respectively . Using multistate models, we found that breeders had significantly higher probabilities of local survival (phi = 0.678 vs. 0.486), of retaining their status as breeders (psi = 0.719 vs. 0.279), and of detection (p = 0.997 vs . 0.375) than did nonbreeders. Overall, males and females had comparable lo cal survival rates (breeders phi = 0.698 vs. 0.658, nonbreeders phi = 0.536 vs. 0.436). Our estimates of local survival could be affected by breeding dispersal, but site fidelity of parrotlets was strong: 95% of adults moved <500 m in consecutive years. A literature review for tropical birds showed that mark-resighting studies usually report return rates based on resightin gs of breeding or territorial adults, whereas mist net studies rely on reca ptures and pool birds of different age and social status in their calculati ons of local survival. Future studies should attempt to compare subsets of avian populations that are similar in demography. Because rates of site fid elity and social system may differ among species, these factors must also b e considered in interspecific comparisons of avian life histories.