The upstream/downstream movement paradox of lotic invertebrates: quantitative evidence from a Welsh mountain stream

1. This study attempts to determine the mechanisms by which the downstream displacement of lotic invertebrates is compensated by the upstream movements of immatures and adults. To this end, submerged and aerial traps were set up at three sites, each 100 m apart, on a small mountain stream (Yr Ogof) in North Wales and operated for 12 months. 2. At Site 1 (the source), 19.3% of the benthic invertebrates lost as drift were replaced by the upstream movement of aquatic stages. At Site 2, the average upstream compensation was 51.1% (although there was a marked difference between the two traps at the site). At downstream Site 3 the average compensation was 55.4%. The net loss of benthos downstream over 1 year was estimated to be 261 920 individuals from Site 1, 41 891 from Site 2, and 40470 from Site 3. 3. At Sites 1 and 2, the numbers of invertebrates drifting were positively correlated with both current speed and water depth. At Sites 2 and 3, the numbers moving upstream were positively correlated with water temperature. At all three sites, the numbers of animals drifting were positively correlated with the numbers moving upstream, under water. 4. In none of the eight common species of stonefly did females show a strong overall upstream flight preference - however, two species, Leuctra nigra and Nemoura erratica, showed a significant downstream preference. 5. As with the stonefly nymphs, most of the caddisfly larvae showed a longitudinal distribution. Agapetus fuscipes larvae were most abundant at Site 1, yet females showed no preferred flight direction - this would seem to be inappropriate for maintaining this larval distribution pattern. However, females caught moving upstream at Site 2 were bigger and contained twice as many eggs, which were also larger, than females moving upstream at Site 1. The flight patterns of the other Trichoptera also seemed inappropriate for maintaining their respective larval distributions. Immature insects moved upstream at certain sites and times, but trends were not consistent. There was no correlation between female flight direction and local wind speed and direction. 6. Downstream displacement of the most abundant non-insect species, Gammarus pulex, was quite considerable at all three sites. However, at several times during the year the numbers moving upstream approached or exceeded those drifting. Maintenance of benthic population densities in this species is thought to be through a combination of upstream movement (which was highly correlated with benthic density at Sites 1 and 3) and production of young over a long (10 month) reproductive period. Maintenance of the benthic populations of the immature stages of the majority of insect species in Yr Ogof is thought to be through oviposition by females in the vicinity of their emergence sites. 200 D. D. Williams and N. E. Williams 7. Differences in density, size of individuals and life history were found among sites for several benthic species. The concept of optimal larval habitat is discussed in the context of longitudinal zonation and the flight behaviour of ovipositing females