Timber harvesting operations have significant effects on both water quantity and water quality. The effects on water quantity have been well documented both in Australia and elsewhere. The effects on water quality are less widely appreciated, and include elevated concentrations of dissolved salts, suspended solids and nutrients, especially during peak flow periods. Several Australian studies have failed to measure peak flow transport of suspended solids, or have measured it inadequately, thus severely underestimating transport. The major short-term effects of timber harvesting on the aquatic biota result from increased sediment input into streams or increased light through damage to, or removal of, the riparian vegetation. Sediment which settles on, or penetrates into, the stream bed is of more concern than suspended sediment, and can lead to long-term deleterious changes to fish and invertebrate populations. Increased light causes an increase in stream primary production which may increase invertebrate densities, and alter community composition. These biological consequences have not yet been adequately investigated in Australia. Longer-term effects, as yet not investigated in Australia, include changes to stream structure as the regrowth forest has fewer large logs to fall into the stream. These large logs play a major role as habitat and retention structures in streams. There has been no attempt to evaluate the effects of timber production activities, including pesticide use and fuel reduction burning, on the Australian stream biota. Likewise, although buffer zones are widely advocated as a protection measure for streams in Australia, there have been no studies to evaluate their effectiveness.

Twenty-five patches (1 m^) of natural stream substratum in the Acheron River. Victoria, were physically disturbed by kicking and raking during winter 1986 and summer 1987. The macroinvertebrate composition of these disturbed patches was examined at various times over the following 71 days, and compared with adjacent undisturbed control patches sampled concurrently.The disturbance did not alter the particle-size distribution (>150 nm) of the disturbed patches. Organic material was reduced in the disturbed patches by about 70% in each season, but returned to control levels within 21 days in winter and 8 days in summer.The total number of species, and the density of species and individuals were all significantly reduced by the disturbance. Recovery of species density was complete after 21 days during winter and 8 days during summer, and the density of individuals recovered after 71 days during winter and 8 days during summer. The differences were due to the slower colonization rate of Chironomidae in winter, either because of a lower drift rate, or a slower recovery of detritus in winter.Individual species showed variations in colonization patterns, most increasing steadily at various rates, with some declining after an initial rapid increase (e.g. Baetis spp.). In the latter case, the density changes were mirrored in the control patches, emphasizing the need to take control samples concurrently with experimental samples.In each season, the species remaining immediately following the disturbance, and those subsequently colonizing the disturbed patches were in the same rank order (Spearman Rank correlation) as their occurrence in the control patches, suggesting that no taxa were differentially affected by the treatment.No evidence was found to allow the application of the Intermediate Disturbance Hypothesis to explain species diversity at the scale of this study. It appears that current hypotheses developed to explain the relationship between diversity and disturbance in sessile communities do not apply to highly mobile communities in streams.

The effect of artificially elevating concentrations of suspended sediment on macroinvertebrate drift was studied in the Acheron River, 100 km north-east of Melbourne. Two experimental channels were established in the stream, and suspended sediment was introduced into one channel over a period of 1.5 h. The second channel was left undisturbed as a control. The concentration of suspended sediment was altered every 15 min, rising and falling to imitate concentrations reported during natural flood events. Drift was collected from two nets at the downstream end of each channel during each 15-min period. Collections were made for three 15-min periods before the introduction of the sediment and for three periods after the release. The addition of suspended sediment at a mean concentration of 133.4 mg L-1 over a 15-min period (compared with around 20 mg L-1 in the control channel) resulted in a sevenfold increase in the total number of drifting invertebrates. At lower concentrations (both before and after this peak concentration), drift densities were more similar to prerelease conditions. The number of drifting taxa also showed an increase during the period of high release. Although the experiment did not conform strictly to a full experimental design, the results indicated that there may be a threshold level of suspended sediment that initiates macroinvertebrate drift, and this experiment represents an appropriate starting point for future investigations.

Keywords : sampling device for macroinvertebrates of large stones in streams, surface area of stones Abstract A device for quantitatively sampling the macroinvertebrates of large stones in streams is described . In comparison to the usual method for sampling the fauna of large stones, (the lifting of stones upstream of a hand net), the present method accurately samples stonedwelling animals that are good swimmers . Details are given of a reliable method to measure the surface area of stream stones using thin plastic sheeting .

In the Acheron River, southern Victoria, patches of riffle substratum (ca 1 m-) were disturbed every 10 days by kicking and raking. .After 20 days, i.e. three disturbances, a further set of patches was disturbed once. Eor the next 70 days macroinvertebrate dynamics were monitored in the two sets of disturbed patches and also in contiguous control patches. There were no differences in the temporal changes in total species richness, number of species per sample, densities of individuals, or species diversity (H') between the two disturbance regimes. The composition of the fauna colonizing each disturbance regime was similar, and after 33 days the number of species per sample was similar in disturbed and control patches. The fauna appears to be well adapted to physical disturbance and current ideas linking species richness and disturbance cannot be readily applied to stream communities at the temporal and spatial scales of this experiment.