CONDITION AND VULNERABILITY TO HUNTING AMONG MALLARDS STAGING AT LAKEST-CLAIR, ONTARIO

A basic assumption of band-recovery analysis is that recovered samples are representative of the banded population. This assumption may be v iolated if individuals in poor body condition are disproportionately v ulnerable to hunting. Thus, we used body mass and band-recovery data f rom 1,412 mallards (Anas platyrhynchos) trapped at Lake St. Clair, Ont ario, during August-September of 1989 and 1990 to test the hypothesis that individuals in relatively poor condition before the hunting seaso n subsequently are at a disproportionately high risk of being shot. Su pporting this hypothesis, binary regression analysis revealed negative relationships (P less-than-or-equal-to 0.05) between preseason condit ion (body mass corrected for structural size) and direct recovery prob ability in 3 of 8 separate analyses (by year, sex, and age), and overa ll. Our data support the position that condition biases may be widespr ead in hunter-shot samples of waterfowl, indicating a need for caution when using band-recovery data to estimate hunting mortality and other recovery-based parameters.