Connected headwaters: Indelible field evidence of dispersal by a diverse caddisfly assemblage up stream valleys to dry catchment boundaries

1. Integral to many ecological models is the notion that dispersal connects populations and communities in disparate locations. For stream insects, however, there is very little empirical information about which species and what proportion of the species in a community are likely to disperse, in what direction and how far they travel, and whether dispersal is successful in connecting populations and communities. This study examines whether species of caddisfly can disperse between catchments and potentially connect communities.2. We used a novel sampling design comprising five pairs of streams. Each stream in a pair flowed from the same boundary, but in opposite directions and into isolated catchments. We surveyed abundances of adult caddisflies on the catchment boundary (boundary species) where there was no water (Up location), and on two locations per stream (Middle, Down locations) where there was permanently flowing water (resident species). If caddisflies fly along the topographic features of stream valleys, then we expected to find adults on the ridge between catchments and, if dispersal is successful, stream communities in adjacent catchments should have similar species composition.3. Over 130 species of caddisfly (from 18 families) were collected. Approximately half of them occurred on at least one catchment boundary, and Up locations averaged almost 20 species. Boundary species included both the smallest- and largest-bodied taxa and they were a near perfect subset of species in resident assemblages. However, boundary species were not simply the most common in resident assemblages. There was no evidence of sex-biased dispersal. Unsurprisingly, assemblage composition varied across the landscape, within and among catchments. However, resident assemblages within stream pairs were significantly more similar to one another than unpaired streams in either the same or in different catchments.4. Our results suggest that a surprisingly diverse set of caddisflies disperse up to catchment boundaries and potentially connect streams on either side of catchment divides. Nevertheless, many other caddisfly species were never recorded at Up sites and hence may never or only rarely reach catchment boundaries.5. These outcomes suggest, firstly, that some caddisflies are capable of trans-catchment dispersal but it is incorrect to assume that all caddisflies are good dispersers, as is often stated in the literature. Secondly, caddisfly assemblages in headwater streams may be less isolated than is often considered and the dendritic structure of stream networks may be less important for their distribution patterns than the proximity of tributaries in neighbouring catchments, in many landscapes. Trans-catchment dispersal could ensure community connectivity and facilitate recovery after catchment-wide extinctions.

Automated summary

This study examines the dispersal ability of caddisfly species between catchments and their potential to connect communities. It was found that some caddisflies can disperse up to catchment boundaries, but not all species have good dispersal ability. The proximity of catchments may play a more important role in caddisfly distribution patterns than the dendritic structure of stream networks.