SUMMARY A mixed nutrient supplement was added to stream‐side experimental channels in a closed canopy rainforest in tropical Queensland over a 7‐month period, raising nitrogen and phosphorus concentrations well above ambient. Nutrient supplements enhanced decomposition of some leaf species, but levels of chlorophyll a and fine particulate organic matter did not differ between treatment and control channels. Shredding by Anisocentropus kirramus (Trichoptera) had a significant effect on leaf litter breakdown, but nutrient supplements did not enable A.kirramus to shred more material; however, biochemical analysis of A. kirramus showed that nutrient supplements enhanced the nutritional quality of the litter for this species. Treatment channels contained 75% more invertebrates than control channels, but significant differences in abundance (all positive) were detected in only five of the 109 invertebrate species present. There was no change in species richness or evenness between treatment and control channels. The limited community response indicates that (a) variables other than nutrients (e.g. light) restricted primary productivity, (b) low nutrient concentrations limit invertebrate production, (c) invertebrate community composition in this closed canopy, forest stream is resistant to nutrient enhancement, and (d) there was no direct relationship between productivity and diversity in this tropical stream.

No abstract

In-situ artificial stream channels and ex-situ laboratory mesocosms were used to measure the responses of macroinvertebrate assemblages, from upland and lowland tropical streams, to high loads of fine clay sediment. Significant responses were observed mainly in the in-situ channels in the upland stream, where densities and the number of taxa were lower in the treatment channels than in controls. There was no evidence of any taxon being particularly sensitive to sedimentation, with a general decrease in densities across several taxa and differences only detectable for abundant taxa. Animals moved downstream in response to the treatment, but only a short distance within the channels. However, further colonization after the treatment was impeded in the treatment channels. In the mesocosm experiments, the upland macroinvertebrate assemblage demonstrated some negative effects; however, the lowland fauna was very tolerant to sedimentation, even when exposure was extended to 15 days. Together, the mesocosm and in-situ experiments indicate that there is a behavioural response to sedimentation because of a change in the habitat, and that the lowland macroinvertebrate assemblage is more tolerant of sedimentation, at least in the short term.