Population dynamics across geographical ranges: time-series analysis of three small game species

Abstract

Little is known about how density-independent and density-dependent processes affecting population dynamics vary geographically across species' ranges. To address this problem for three vertebrate species (Northern Bobwhite [Colinus virginianus], Ring-necked Pheasants [Phasianus colchicus], and eastern cottontails [Sylvilagus floridanus]), we examined spatially subdivided, long-term (1966–2001), seasonal (January, April, July, and October), time-series data from the Kansas Rural Mail Carrier Survey, USA. All three species have range boundaries in Kansas, with population abundances declining toward the periphery of their ranges. We quantified the strengths of density-dependent and density-independent processes affecting the dynamics of 10 populations of each species ranging from low (peripheral) to high (central) mean abundance using first-order autoregressive models that incorporate measurement error.

For all three species, peripheral populations with lower mean abundance tended to have greater population variability. This pattern could potentially be explained by peripheral regions experiencing either weaker density dependence or greater environmentally driven density-independent fluctuations in per capita population growth rates. In general, density dependence did not vary among geographic regions, although there was a trend for smaller, peripheral populations to exhibit stronger density dependence. Density-independent variability in per capita population growth rates was higher in peripheral populations. Furthermore, density-independent fluctuations in per capita population growth rates were weakly correlated with temperature and precipitation and were highest for the period October through January, identifying fall as the period of greatest environmentally driven variability in population dynamics. Per capita population growth rates fluctuated in moderate synchrony among regions, especially for more abundant, nonperipheral populations in close geographical proximity.

The strong density-dependent and stronger density-independent processes in smaller, peripheral populations suggest that the greater variability in peripheral populations' densities is caused by greater population sensitivity to environmental fluctuations. This may make peripheral populations more likely to go extinct and leads to the prediction that, if these species decline to the point of becoming endangered, this decline will be accompanied by a contraction in their geographical ranges.