A Mid-Atlantic and a National Population Model of Northern Bobwhite Demographic Sensitivity

Abstract

Numerous field studies have estimated fecundity and survival rates for northern bobwhites (Colinus virginianus), but a synthetic population model based on life-stage simulation analysis (LSA) was only recently developed to examine demographic sensitivity of the finite rate of population change. We compare local demographic parameters of bobwhite versus a national compilation to identify limiting demographic factors for improved regional habitat planning and management. The national compilation provided a useful overview but combined parameters across populations at different latitudes and under different management regimes. We parameterized our LSA model to examine the sensitivity of the finite rate of growth (k) to simulated variation in 9 demographic parameters primarily estimated from field studies for one population in regional decline in New Jersey. Our model results predicted population declines in New Jersey (k 1⁄4 0.55) comparable to the national estimate (k 1⁄4 0.54), but notable differences occurred in sensitivity of demographic variables. The national model predicted winter survival of adults made the greatest contribution to variance of k (r2 1⁄4 0.42) followed by summer survival of adults (r2 1⁄4 0.13), and survival of chicks (r2 1⁄4 0.11). Our regional model for New Jersey also predicted winter and summer survival of adults would make the greatest contribution to variance of k (r2 1⁄4 0.33 and r2 1⁄4 0.13). The New Jersey model, in contrast to the national model, showed that annual variation in components of fecundity had a large effect on Var(k): including clutch size (r2 1⁄4 0.18 vs. national r2 1⁄4 0.01), nest success (r2 1⁄4 0.20 vs. national r2 1⁄4 0.06), and the number of young produced per nest that survived 30 days (r2 1⁄4 0.53 vs. national r2 1⁄4 0.16). Slopes of linear regression between simulated variation in each demographic variable against k were similar between the national and regional models. The slope for number of young produced per nest that survived 30 days with one exception was lower in the New Jersey data indicating more young are required to realize a stationary population. Our simulation results suggest management practices that improve winter survival or the number of young surviving 30 days will have the greatest potential to increase bobwhite population growth rate in New Jersey. Future linkage of models of demographic performance to experimental habitat manipulations will aid regional scientific planning to improve necessary habitat management.