Do pools impede drift dispersal by stream insects?

Abstract

Drift of stream insects is one of the most ubiquitous forms of downstream dispersal and thought to be a key factor influencing the persistence of local populations. Identifying the factors that limit drift dispersal between habitat patches is needed to understand the connectivity of insect populations along river channels. We determined whether insects drifting between riffle habitats (i.e. patches of suitable habitat) were impeded by natural, slow‐moving pools (i.e. unsuitable habitat), limiting dispersal to the next downstream riffle, by estimating drift rates entering and exiting pools. We also investigated whether the frequency of drift dispersal between riffle habitats decreased with increasing pool size (length, width and depth), resulting from increasing areas of low or zero current velocity and/or distance between habitat patches. We found that for the majority of study taxa (7 of the 8 taxa), drift dispersal between riffles was significantly hindered by the intervening pool habitat, supporting our prediction that natural, large slow‐moving pools impede the number of invertebrates drifting between riffle habitats. There were three taxa whose drift rates were significantly reduced by increasing pool depth or width—Offadens hickmani, Austrophlebioides spp. and Austrosimulium spp., and weak relationships (p < .10) were also found for Coloburiscoides munionga and Asmicridea edwardsii. Drift was not associated with pool length for any species, suggesting that hydraulic conditions within a pool limit drift dispersal and not the total distance between riffle habitat patches. Overall, our results suggest drift dispersal seems unlikely to be a major mechanism of long distance dispersal, for at least some common insects. These findings indicate that drift may play a more limited role in stream insect dispersal than is commonly considered and that flying adults may therefore have much greater influence on connectivity among populations.

Barriers to dispersal: The effect of a weir on stream insect drift

“…As reported in the study of Brooks et al (), a significant proportion of drifters of all four focal taxa failed to disperse from the upstream riffle through natural pools to the next riffle (see Brooks et al, for results of t tests; Table , Figure ), but the weir pool presented an even greater barrier. On average, the numbers of drifters emigrating from a natural pool were between 24 and 41% less than immigrating into a pool.…”

“…and Austrosimulium spp. have previously been found to be related to increases in pool depth and width (Brooks et al, ), indicating that their contrasting drifting methods are equally diminished by slow moving water. The greater areas and multiple zones of zero and low‐flow areas in the weir pool appear to have exacerbated this effect, particularly on larger Austrosimulium spp.…”

“…Mean source/sink index for drifting study taxa in natural pools and Mowamba weir. Natural pools are solid black circles and error bars are 95% confidence intervals (Brooks et al, ). Mowamba weir SS values are in grey circles (December) and open circles (April).…”

“…and Cheumatopsyche spp. through natural pools are unrelated to pool depth and width (Brooks et al, ). Therefore, the higher drift reductions through the weir pool cannot be explained by the slower flow in the weir but possibly caused by the weir wall at the downstream end of the weir pool blocking downstream movement by walking and drifting.…”

Experimentally disentangling the influence of dispersal and habitat filtering on benthic invertebrate community structure

“…Even on smaller spatial scales like the stream reach scale, natural complexity and habitat variability can directly affect the process of macroinvertebrate dispersal (Brooks, Wolfenden, Downes, & Lancaster, 2017). The influence of dendritic network structure also depends heavily on the traits of organisms, particularly traits that are related to dispersal (Bogan & Boersma, 2012;Brown & Swan, 2010;Finn, Theobald, Black, & Poff, 2006;Razeng et al, 2016;Schmera et al 2017).…”