The influence of predation danger on the distribution of non-breeding shorebirds in a tropical estuary system

I studied non-breeding shorebirds in the extensive mangrove-mudflat system of Northern Nariño, Colombia. I asked how the non-breeding distributions of 18 species are influenced by functional traits, the interplay between food and danger attributes of landscapes, and interactions with other species. I found that almost all the area’s ~8000 Whimbrels (Numenius phaeopus) roost together on just one of the hundreds of available mangrove islands. Much smaller numbers occasionally roost on a few other islands. The larger roost site is distinguished by its location, having a larger amount of Whimbrel feeding habitat (mudflat) within a 12 km radius than almost any other mangrove island, and being more isolated from the mainland and thus from terrestrial predators, but not more isolated from villages or shipping channels than other islands. Within a subset of nine shorebird species, an increase in body mass predicted an increase in wing load both within and between species. Contrary to expectations, wing load did not correlate strongly with escape performance (take off speed), but as expected, heavier wing loads did correlate with stronger escape responses (flight initiation distances) across species. Species with higher escape performance use habitats that are more productive, but also more dangerous, while species with lower escape performance reacted sooner to predator stimulus. Tactile and gregarious species show stronger responses to safety gradients. An analysis of co-occurrence of species pairs demonstrated that non-random patterns were prevalent within communities of non-breeding shorebirds. Species pairs tracking same or opposing environmental gradients explain some positive and negative associations, but a large proportion of the associations was due to residual variation linked to the species themselves. Positive associations could be explained by heterospecific attraction associated with reducing predation danger and public information about resources. The fewer negative associations could indicate competitive interference. Alternatively, other sources of environmental variation not captured in this study could explain these “species only” associations. Our results contrast with previous studies of avian communities for which shared environmental responses play a larger role and suggest that social interactions are as important in structuring shorebird communities. This thesis demonstrates how using distribution models informed by species’ morphology, behavior, and interactions with other species, we will be better equipped to understand the effects of habitat conversion on the conservation of migratory shorebirds.