Seasonal and year to year variation in the macroinvertebrate communities of New Zealand forest streams : a thesis submitted in partial fulfilment of the requirements for the degree of Master of Science in Ecology at Massey University, Palmerston North, New Zealand

Abstract

The bed movement of 42 streams in the Ruahine Forest Park, Urewera National Park, and Cass-Craigiebum region was predicted from each stream's channel and catchment characteristics. While a stepwise regression was relatively unsuccessful in predicting tracer particle movement, an artificial neural network analysis achieved strong correlations with measured tracer particle data. Forty-three streams in the Ruahine and Tararua Forest Parks were sampled in the summers of 1996 and 2001, and the macroinvertebrate communities compared. Changes in community structure between the two surveys did not correlate with any measured environmental characteristics including stream bed movement and change in periphyton biomass. MCI scores changed by a mean of 12.8 points between the two surveys, and the number of sites attaining an MCI score indicative of a 'pristine' stream dropped from 40 to 29. This appears to be related to a change in stream temperature, with streams that were cooler in 2001 than in 1996 showing an increase in MCI, while those which were warmer showed a decrease. Changes such as these could have a marked effect on biomonitoring programmes that use reference sites similar to these streams. In both 1996 and 2001, a greater number of taxa were collected from sites with more periphyton - taxon richness appears to asymptote at chlorophyll a concentrations greater than 5 μg/cm² Twelve streams within the Ruahine Forest Park were sampled every three months between June 2000 and May 2001. Both periphyton biomass and macroinvertebrate taxon richness tended to decrease with bed movement. While macroinvertebrate community structure showed marked changes over the study period, these changes were not linked to bed movement or variation in periphyton level. The seasonal changes observed in these streams are not significantly different to the changes seen between the summers of 1996 and 2001 - community structure was no more stable between two summers separated by five years than it was between the seasons of a single year. Eight artificial channels were laid on the bed of the Turitea Stream. At the onset of the experiment, half of the channels contained neither invertebrates nor periphyton cover, while the other half had no invertebrates but an initial periphyton layer. Drift samples indicate that approximately one in four drifting invertebrates colonised the channels during the 14 day study period, with benthic taxon richness reaching a peak after only four days. Colonisation was not affected by periphyton biomass. Some of the less common taxa that were present in the water column did not colonise the channels within 14 days.