Spatial and temporal variation in the influence of grazing macroinvertebrates on epilithic algae in a montane stream

A caddisfly-exclusion experiment, employing four trials over owmelt magnitude and summer baseflow nutrient regime, was used to assess effects of variation in macroinvertebrate grazing intensity and nutrient supply on epilithic algae in a montane headwater stream. Ceramic tiles, either inaccessible to caddis larvae or situated to allow access to all grazers, served as substrata for algal accrual. Trials were run during or just after peak melt discharge each year, and again at summer baseflow in two reaches that differed in patterns of ground-water/surface water exchange. In 1995, a year of heavy snowmelt, stream-water nitrogen to phosphorus ratios (N : P) suggested N limitation on all but one sampling date. By the start of trial 1, mayfly and chironomid densities had recovered from melt-induced reductions, but caddis populations had not and exclusion had no effect. Caddis densities had recovered by trial 2 and exclusion effects were noted, particularly in the upstream reach where caddis larvae were abundant Snowmelt in 1996 occurred during trial 3, but was of low intensity and had no apparent effect on stream biota. Grazers were abundant during trial 3, exclusion effects were evident in both reaches, grazerresistant taxa dominated algal assemblages, and live-diatom percentages were uniformly low. Trial 4, run at baseflow in 1996 under high N : P (and high N), was characterized by high algal densities, taxonomic diversity, and diatom live-cell percentages, relatively low grazer densities, significant exclusion effects in both reaches, and reach x treatment interactions indicating between-reach differences in nutrient regime. Negative correlations were noted between mayfly density and algal cell numbers and diatom live-cell percentages on natural substrata. Caddis densities correlated negatively with live-diatom percentage, but not algal density. Of 23 common algal taxa on caddis-accessible tiles, 28-d accrual rates of only 5 correlated negatively with caddis densities; eleven (14 if exclusion substrata were included) correlated negatively with mayfly densities. Changes in accrual rates varied with algal growth form; accrual rates of grazer-sensitive taxa fell sharply when grazer densities exceeded 1000/m². Our data illustrate that grazing effects can vary between spatially proximate and physically similar stream reaches, and that the nature of these effects may differ among years and seasons.