Density-dependent dispersal in birds and mammals

Density-dependent dispersal can be caused by various mechanisms, from competition inducing individuals to emigrate (positive density-dependence) to social crowding effects impeding free movement (negative density-dependence). Various spatial population models have incorporated positively density-dependent dispersal algorithms, and recent theoretical models have explored the conditions for densitydependent dispersal (DD) to evolve. However, while the existence of DD is well documented in some taxa such as insects, there is no clear picture on its generality in vertebrates. Here I review the available empirical data on DD in birds and mammals, focusing mainly on variation in dispersal between years and on experimental density manipulations. Surprisingly few studies have explicitly focused on DD, and interpretation of the available data is often hampered by differences in approach, small sample sizes and/or statistical shortcomings. Positive DD was reported in 50 and 33% of the selected mammal and bird studies, respectively, while two studies on mammals (out of eight) reported negative DD. Among bird studies, DD was more often reported for emigration rates or long-distance recoveries than for average distances within finite study areas. Experimental studies manipulating densities (mainly on mammals) have consistently generated positive DD, typically showing reduced emigration in response to partial population removal. Studies that examined dispersal in relation to seasonal changes in density (small mammals only) have more often reported negative DD. Studies that compared dispersal between sites differing in density, also show a mixture of positive and negative DD. This suggests that dispersal changes in a more complex way with seasonal and spatial density variation than with annual densities, and/or that these results are confounded by other factors differing between seasons and sites, such as habitat quality. I conclude that both correlational and experimental studies support the existence of positive, rather than negative, density-dependent dispersal in birds and mammals.