Annual survival rates of migratory shore and upland game birds

Band recoveries from hunted populations are widely used to estimate survival and harvest rates. We used long-term (1950–2010) band-recovery data from the United States and Canada to estimate average survival and recovery rates for juvenile and adult sandhill cranes (Grus canadensis), American coots (Fulica americana), clapper rails (Rallus longirostris), and Wilson's snipe (Gallinago delicata). Banding data were typically sparse and Brownie recovery rates (f) were <1% for most species and age classes, so we also utilized encounters from birds banded and recovered throughout the year, including nonhunting recoveries. Estimated juvenile and adult annual survival rates based on combined dead recoveries were 0.917 (SE = 0.080) and 0.884 (SE = 0.006) for sandhill cranes; 0.325 (SE = 0.025) and 0.583 (SE = 0.008) for American coots; 0.368 (SE = 0.070) and 0.500 (SE = 0.025) for clapper rails; and 0.372 (SE = 0.077) and 0.653 (SE = 0.020) for Wilson's snipe. American coots had sufficient band-recovery data during 1951–1987 to estimate annual process variation in survival for juveniles ((Formula presented.) = 0.017) and adults ((Formula presented.) = 0.056). Based on average estimated juvenile and adult survival rates for each species, stable populations could be achieved with average annual recruitment rates (expressed as fledged young per adult in the preseason population) of 0.15 (SD = 0.02) for sandhill cranes, 1.26 (SD = 0.19) for American coots, 1.41 (SD = 0.31) for clapper rails, and 0.97 (SD = 0.23) for Wilson's snipe, with uncertainty reflecting measurement error in survival rates. Although band-recovery models can be successfully applied to small data sets (e.g., 300–3,000 total recoveries in our analyses), juvenile survival estimates lacked precision and annual variation was inestimable for most species, leading to considerable uncertainty for population modeling efforts.