The dangers of leaving home: dispersal and mortality in snakes

For animal populations in many parts of the world, direct (albeit often accidental) killing by humans may be a signi®cant source of mortality. Many snakes are killed by people (especially by automobiles) every year, but the determinants of a snake's vulnerability to anthropogenic mortality (and thus, patterns of mortality with respect to sex, age and season) are poorly known. We present data on 652 French snakes of six species (Coluber viridi ̄avus, Elaphe longissima, Natrix maura, N. natrix, Vipera aspis, V. berus) killed either by natural predators, domestic animals or humans (including roadkills). We used information on seasonal patterns of mortality (plus information on population structure from 338 captures of live snakes) to test the hypothesis that snakes are killed mostly when they disperse from their usual home ranges. This hypothesis generates several falsi®able predictions on the expected correlates of mortality rates; most of these predictions are supported by our data. For example, young-of-the-year snakes are killed primarily in the period immediately after hatching (while they disperse); subadults (which are sedentary) generally experience low mortality rates; adult males are killed mainly during the mating season (especially in species where mate-searching males travel widely); and adult females in oviparous species are killed during their egg-laying migrations. Relative to population density, species that use frequent long-distance movements in foraging experience higher mortality than sedentary ambush foragers. In one species (E. longissima), larger males are more at risk. The success of these predictions suggests that movement patterns of snakes may o€er valuable indices of their vulnerability to direct anthropogenic mortality. # 1999 Elsevier Science Ltd. All rights