Effect of Radiotransmitters on Northern Bobwhite Annual Survival

Abstract: Radiotelemetry has been widely used in northern bobwhite (Colinus virginianus) research to estimate survival rates and other demographic parameters. Biologists have used this knowledge to study bobwhite ecology, develop management theory, and base management actions. We tested the assumption that radiotransmitters do not bias survival rates of bobwhites by comparing survival rates of banded bobwhites with and without radiotransmitters on Tall Timbers Research Station (TTRS) from 1999 to 2004. We used Burnham's model in Program MARK and model-selection procedures to determine relative importance of year, gender, and radiotagged status on annual survival rates and recovery processes. Three plausible models (relative quasi-likelihood Akaike's Information Criterion ΔQAIC_c] < 3) included year dependence in survival and an additive effect of gender but no radiotransmitter effect. Models including a radiotransmitter effect in survival were >8 ΔQAIC_c from the top models, had low Akaike model weights (w_i < 0.007), and low importance weight (S̀w_i(radio) = 0.01). We also compared band—recapture survival estimates from the QAIC_c minimizing model to staggered entry Kaplan—Meier (KM) survival estimates from 2000 to 2003. Annual KM survival estimates of male and female bobwhites were within the 95% confidence interval of band—recapture estimates in 7 of 8 comparisons. We conclude that radiotelemetry is a reliable technique for determining bobwhite survival. Managers should view information from properly conducted telemetry research as reliable and useful for management.