Summary
Top‐down control of benthic communities by fishes in temperate and Neotropical streams has received considerable attention, yet relatively little is known about the occurrence or strength of such top‐down control in tropical Asian streams. We undertook pool‐scale removals of an entire macroconsumer assemblage of fishes and predatory shrimp, as well as its component guilds (algivores versus predators), in three Hong Kong streams during the dry season and monitored the responses of periphyton, benthic macroinvertebrates (predominantly insects) and leaf‐litter breakdown.
Pools supported high densities and biomass of predatory macroconsumers. Algivorous balitorid fishes had a mean density of 30.1 indiv m−2 (range: 1.5–58.2 indiv m−2) and a wet‐weight biomass of 20.9 g m−2 (range: 1.6–34.2 g m−2); values for predatory fishes (mainly a nemacheilid and a goby) were 9.4 indiv m−2 (range: 7.3–11.1 indiv m−2) and 29.7 g m−2 (range: 24.1–38.0 g m−2). Densities of predatory palaemonid shrimp were likewise high (mean: 14.0 indiv m−2; range: 8.9–19.0 indiv m−2).
Despite high macroconsumer standing stocks, there was little evidence of top‐down control of benthic communities. The removal of algivorous fishes had no effects on algal biomass, periphyton accumulation, silt accrual or abundance of benthic insects. The removal of all predators (fishes and shrimp), as well as the removal of all macroconsumers, was associated with a small increase in the density of insect scrapers (especially baetid mayflies), although no such effect was evident when predatory fishes (but not shrimp) were removed. Apparently, shrimp substituted for the role of predatory fishes by cropping any excess prey. However, the removal of the entire predatory guild or all macroconsumers produced no additional changes in macroinvertebrate assemblages, nor any influence on periphyton or leaf‐litter breakdown.
The pool‐scale approach used in this investigation, combined with the inclusion of three study streams, should have provided a realistic picture of the significance of top‐down control of benthic communities by macroconsumers. Statistical power may have been limited by logistical constraints on the number of replicate pools. Localised nutrient limitation caused by the removal of macroconsumers might have contributed to the apparent weakness of top‐down control of benthic communities. However, given the high standing stocks of fishes and shrimps, it is remarkable that experimental manipulations resulted in generally small or negligible effect sizes.