EFFECTS OF A BENTHIVOROUS AND A DRIFT-FEEDING FISH ON A BENTHIC STREAM ASSEMBLAGE

I assessed the impact of both drift-feeding (Salmo trutta. brown trout ) and benthic-feeding (Cottus gobio, bullhead) fish on a benthic assem blage during a I-month-long field experiment. I used enclosures contai ning cobble/gravel substrata with 6-mm mesh net that allowed invertebr ates to drift freely in and out of enclosures. Four treatments: arrang ed. in a factorial design, were tested: a predator-free control, bullh eads only (2.67 bullheads/m(2), two per enclosure), brown trout only ( 2.67 brown trout/m(2), two per enclosure) and bullheads and brown trou t together (2.67 fish/m(2), one of each). Bullheads reduced the densit ies of seven invertebrate taxa (Gammarus pulex amphipods, Baetis rhoda ni mayfly nymphs, Leuctra spp, stonefly nymphs, Polycentropus spp, cad dis larvae, Pacifastacus leniusculus signal crayfishes, Simuliidae, bl ackfly larvae, and Limnephilidae, caddis larvae) whereas brown trout o nly affected one taxon (B. rhodani). The weaker effect of brown trout on benthic prey was probably related to its heavy reliance on terrestr ial prey. Dietary analyses showed that more than 80% of prey consumed by brown trout were terrestrial animals, whereas bullhead only consume d benthic prey. Neither bullhead nor brown trout affected the absolute number of immigrating or emigrating invertebrates in enclosures, but bullhead affected the per capita emigration rates of G. pulex. Direct predation by bullhead was more important than avoidance behaviour (dri ft) in determining densities of six of the seven taxa: only G. pulex d ensities were equally affected by avoidance behaviour and direct preda tion. Direct predation by brown trout was also more important in deter mining densities of B. rhodani. The presence of bullhead raised periph yton biomass. presumably mediated via their consumption of grazers, Br own trout had no effect on periphyton biomass and these results indica te that the presence of alternative prey in this case terrestrial anim als, may have repercussions for fish-benthic macroinvertebrate-periphy ton interactions and may potentially explain some of the variable impa cts of fish on benthic macroinvertebrates that have been reported in t he literature.