Seasonal survival of Greater Snow Geese and effect of hunting under dependence in sighting probability

Although much of life-history theory assumes increased mortality at certain stages (e.g., migration), survival rates are rarely estimated on a seasona l basis within the annual cycle of migratory species. We estimated variatio ns in seasonal survival rates in a long-lived, hunted species in the presen ce of short-term (between consecutive seasons) and long-term (between years in the same season) dependence in sighting probabilities. We also tested t he two contrasting hypotheses that hunting mortality is either compensatory or additive to natural mortality. This study was conducted on adult female Greater Snow Geese (Chen caeruldescens atlantica) from 1990 to 1998, and i s based on 3890 neck-banded birds and 13 657 resightings on the northern br eeding grounds in summer, and southern staging areas in spring and autumn. Birds were 10-20% more likely to be seen in autumn and spring if they were seen on the previous occasion (summer and autumn, respectively). Birds were 30-40% more likely to be seen in autumn and spring if they were last seen in the same season in the previous year. Differences in behavior according to family status (presence or absence of young) and heterogeneity in site f idelity may explain the dependence in sighting probabilities; failure to ac count for this may lead to biased survival estimates. Monthly survival rate s from spring to summer (3-mo period) and summer to autumn (2.5-mo period) were equal (0.989 +/- 0.003, mean +/- SE) and showed little variation over the years., even though the two lengthy migratory flights (3000 krn each) a nd breeding occurred during these periods. In contrast, monthly winter surv ival (from autumn to spring, 6.5 mo) was lower than during the other season s and varied significantly over the years (range 0.936 +/- 0.021 to 0.993 /- 0.008). Mean survival of adult females (corrected for neck band loss) wa s 0.96 from spring to autumn, 0.86 during winter, and 0.83 +/- 0.05 for the whole year. Natural mortality (i.e., excluding hunting) was equal among se asons and did not vary over the years, which suggests that mortality risk i s not increased during migration or reproduction. There was a significant i nverse relationship between winter survival and the kill rate (b = - 1.21 /- 0.56), which suggests that hunting mortality was additive to natural mor tality. This is probably a general feature of long-lived species because th eir low and relatively constant natural mortality rate does not allow them to compensate for an additional source of mortality such as hunting. Contra ry to life-history theory, we did not find evidence that migration or repro duction entailed a survival cost in this long-distance migrant bird.