RELATIVE EFFECTS OF SURVIVAL AND REPRODUCTION ON THE POPULATION-DYNAMICS OF EMPEROR GEESE

Populations of emperor geese (Chen canagica) in Alaska declined someti me between the mid-1960s and the mid-1980s and have increased little s ince. To promote recovery of this species to former levels, managers n eed to know how much their perturbations of survival and/or reproducti on would affect population growth rate (lambda). We constructed an ind ividual-based population model to evaluate the relative effect of alte ring mean values of various survival and reproductive parameters on la mbda and fall age structure (AS, defined as the proportion of juv), as suming additive rather than compensatory relations among parameters. A ltering survival of adults had markedly greater relative effects on la mbda than did equally proportionate changes in either juvenile surviva l or reproductive parameters. We found the opposite pattern for relati ve effects on AS. Due to concerns about bias in the initial parameter estimates used in our model, we used 5 additional sets of parameter es timates with this model structure. We found that estimates of survival based on aerial survey data gathered each fall resulted in models tha t corresponded more closely to independent estimates of lambda than di d models that used mark-recapture estimates of survival. This disparit y suggests that mark-recapture estimates of survival are biased low. T o further explore how parameter estimates affected estimates of lambda , we used values of survival and reproduction found in other goose spe cies, and we examined the effect of an hypothesized correlation betwee n an individual's clutch size and the subsequent survival of her young . The rank order of parameters in their relative effects on lambda was consistent for all 6 parameter sets we examined. The observed variati on in relative effects on lambda among the 6 parameter sets is indicat ive of how relative effects on lambda may vary among goose populations . With this knowledge of the relative effects of survival and reproduc tive parameters on lambda, managers can make more informed decisions a bout which parameters to influence through management or to target for future study.