River and stream rehabilitation projects are increasing in number, but the success or failure of these projects has rarely been evaluated, and the extent to which buffers can restore riparian and stream function and species composition is not well understood. In New Zealand the widespread conversion of forest to agricultural land has caused degradation of streams and riparian ecosystems. We assessed nine riparian buffer zone schemes in North Island, New Zealand that had been fenced and planted (age range from 2 to 24 years) and compared them with unbuffered control reaches upstream or nearby. Macroinvertebrate community composition was our prime indicator of water and habitat quality and ecological functioning, but we also assessed a range of physical and water quality variables within the stream and in the riparian zone. Generally, streams within buffer zones showed rapid improvements in visual water clarity and channel stability, but nutrient and fecal contamination responses were variable. Significant changes in macroinvertebrate communities toward “clean water” or native forest communities did not occur at most of the study sites. Improvement in invertebrate communities appeared to be most strongly linked to decreases in water temperature, suggesting that restoration of in‐stream communities would only be achieved after canopy closure, with long buffer lengths, and protection of headwater tributaries. Expectations of riparian restoration efforts should be tempered by (1) time scales and (2) spatial arrangement of planted reaches, either within a catchment or with consideration of their proximity to source areas of recolonists.

No abstract

Human perception of two important visual aspects of the appearance of river and lake water, clarity and colour, was investigated from the perspective of bathing and aesthetics in New Zealand. Bankside interviews were conducted at the same time as measurements were made for clarity and colour. We found that water appearance, perceived suitability of the water for bathing, and bathing activity are closely linked, as is perception of water clarity and colour. Water is perceived as just suitable for bathing at a horizontal black disc visibility of 1.2 m (equivalent Secchi depth c. 1.5 m) with a 90 percentile of 2.2 m (equivalent Secchi depth c. 2.75 m), confirming earlier work. As might be expected, blue waters are preferred to yellow waters, and water is perceived as just suitable for bathing if the Munsell colour is greater than around 30 units (i.e., at the low end of the green-yellow range). For bathing waters, consideration of personal safety is very important; for aesthetics, surroundings are important. People's perception of visual aspects (i.e., clarity and colour) appears to have a strong influence on their rating of overall site suitability.

We investigated the short-term effects on stream flow, habitat, water temperature, periphyton biomass, and macroinvertebrate communities of applying integrated catchment management (iCM) to improve the poor economic and environmental performance of a hill farm used for sheep and cattle grazing. Stream responses to a range of changes in land use and riparian management were compared at four reaches in sub-catchments within a pastoral hill farm area and nearby pasture and native forest reference reaches over 6 years before and 6 years after iCM implementation. at the whole catchment scale, annual runoff from the pasture treatment catchment before ICM was 7% higher than from the native forest reference catchment, and 7-day lowflow was 11% lower, but runoff and low flow both declined by c. 6%/yr following afforestation (mainly Pinus radiata) of 62% of the pasture catchment. At sub-catchment level, pine afforestation with 10 m wide riparian setbacks (site A) was associated with declining channel width, water width, water temperature and periphyton biomass, and increasing periphyton organic content and changes in macroinvertebrate community indices and composition indicating recovery towards reference native forest stream conditions. Site M, involving a mix of partial pine afforestation (36% of catchment) and riparian cattle exclusion with riparian Populus deltoides planting, was associated with generally similar trends to site a. Riparian reforestation with natives and exclusion of all livestock was associated with more rapid reduction in water temperature and streambed cover by fines and macrophytes in a reach with a 1 m channel (site RS) than a 2 m wide channel (site RM). Reaches with active riparian tree planting had increased streambed cover by small wood and roots. In general, the macroinvertebrates indicated a "rubber band" response to ICM actions, with relatively rapid recovery towards native forest stream macroinvertebrate metric values and faunal composition, although restoration of composition was incomplete. Water temperature reduction was identified as a key factor in this improvement along with the 1-2 km proximity of colonists in undisturbed native forest streams.

We characterised water chemistry, aquatic habitat, macrophytes, and invertebrate assemblages in eight lowland streams-five in Westland, South Island, and three in the Waikato, North Island, New Zealand. Factors influencing invertebrate community structure over large (between ecoregions) and small (within an ecoregion) spatial scales were investigated. The Westland sites had generally lower nutrient concentrations, conductivity, and water clarity, and coarser substrates than the Waikato sites, reflecting differences in geological history, hydrology, and land-use intensity. The macrophyte communities in each region were very different in species composition and structure, but seasonal abundance patterns in both regions were fairly typical of New Zealand streams in general. Alien obligate submerged macrophyte species were absent from the Westland lowland stream sites, and may partly reflect the isolated nature of this region where colonisation by asexual macrophytes would be difficult. Waikato stream invertebrate faunas were *Author for correspondence. M02006; Published 14 November Received 23 January 2002; accepted 19 July 2002dominated by molluscs (mainly Potamopyrgus) and crustaceans (mainly amphipods), whereas Ephemeroptera, Trichoptera, and Coleoptera taxa dominated the Westland stream faunas. The overall structure of invertebrate assemblages appeared to be influenced by a combination of regional differences in substrate type, nutrient concentrations, water clarity, and macrophyte cover. Westland streams with more upstream pastoral development had higher proportions of molluscs than those with predominantly forested or scrub catchments. Our study suggests that large-scale ecoregional differences may override smaller-scale land-use effects on lowland stream invertebrate communities, and that management strategies should be developed on an ecoregional basis for lowland stream ecosystems.