Population-level consequences of risky dispersal

Achieving sufficient connectivity between populations is essential for persistence, but costs of dispersal may select against individual traits or behaviours that, if present, would improve connectivity. Existing dispersal models tend to ignore the multitude of risks to individuals: while many assess the effect of mortality costs, there is also a risk of failing to find new habitat, especially when the entire inhabitable area remains both small and fragmented. There are few known rules governing whether individuals evolve to disperse more, or less, than what is ideal for population connectivity and persistence. Here we aim to fill this gap, while also noting that evolution might not only produce suboptimal dispersal behaviour: it also influences individual heterogeneity in dispersal. Intuitively, we might expect heterogeneity to improve connectivity, as some individuals will travel far. However, we show that this is only true if dispersal distances on average are quite short; heterogeneity can also lead to reduced connectivity because it can reduce the proportion of the most profitable ('safest') intermediate dispersal distances. In general, our results show that conditions typically associated with conservation concerns (small and fragmented habitats inhabited by a species with a low birth rate) are also ones that are most likely to lead to suboptimal dispersal traits. This prompts the question of assisted dispersal in cases of urgent conservation concern.