Few comprehensive studies on stream assessment and biomonitoring have been conducted in tropical, freshwater watersheds. Currently under threat from climate change, urbanization and increasing freshwater demands, there is a need for innovative approaches to tropical watershed assessment and management. This study investigated cascade habitat macroinvertebrate communities among four tropical mountain streams with the goal of enhancing future efforts to identify flow biocriteria for watersheds of Polynesia. Cascade macroinvertebrate communities were compared between streams of differing size and magnitude of flow removal to evaluate the biological effects of water withdrawal on benthic communities. Two cascade microhabitats, identified as torrenticolous and amphibious, were evaluated for macroinvertebrate community differences and presence of native taxa among watersheds. Cascade habitat in general was reduced, by as much as 98%, in downstream reaches, having a significant impact on the stream ecosystem physical template important for native stream communities. In addition, two‐way ANOVA results revealed no main effects, but significant interactions of watershed size and flow removal on mean macroinvertebrate density for torrenticolous microhabitats; however, the opposite was true for the amphibious microhabitat. Diversity was significantly higher under undiverted flow conditions (t = 4.21, df = 272, p = 0.0004) and in torrenticolous microhabitats (t = 3.86, df = 272, p < 0.0001) over the entire study period. The amphibious microhabitat was composed of 39% native taxa, while the torrenticolous microhabitat contained <7%. This study provides new options for biomonitoring of native populations in Polynesian watersheds. Further studies that support the development of in‐stream flow criteria to preserve cascades are important to understanding the role of this habitat in tropical stream ecosystem function. Copyright © 2010 John Wiley & Sons, Ltd.

Summary 1. The amphidromous life cycle of several species of neritid snails, shrimp and gobies throughout the tropics includes juveniles that migrate from the ocean to breed in fresh water. In many Hawaiian streams, the decline of Neritina granosa, an endemic gastropod, has been associated with habitat degradation and water withdrawal, which are common factors affecting tropical rivers around the world. 2. We investigated the effects of water withdrawal and density on dispersal and upstream migration of N. granosa using three experimental treatments: (i) reduced flow (RF) owing to a stream diversion, (ii) natural flow (NF) and (iii) natural flow with artificially increased snail density. For each treatment, snails were differentially tagged and released in a stream without a natural, extant population of N. granosa. 3. Capture rates ranged from 17 to 65% over a 63‐day period following release. Captures on 2–6 days after release measured initial dispersal and migration, whereas longer‐term migration rates were calculated from snails captured 16–63 days after release. Snails under NF displayed positive rheotactic behaviour, with only 3–12% demonstrating initial downstream movement. Under RF, 22–77% of snails moved downstream or showed no bias either way. 4. Initial mean upstream migration rate (UMR) was 0.25, 0.66 and 1.16 m day−1 under RF, NF and natural flow with increased snail density, respectively. Longer‐term migration rates did not differ significantly between treatments, and the overall mean was 0.62 m day−1. 5. Principal component analysis and generalised linear models were used to identify habitat characteristics important to UMR, with habitat and reach‐scale hydraulics as the most important factors. The relationship between discharge and UMR suggested it would take 11–35 years for snails to migrate past the most upstream water diversion. However, rates from published studies of neritid snail species migrating en masse or in long lines under natural situations suggested that N. granosa could migrate above stream diversions within 72 days–2.5 years (when in an aggregation) and 29 days–1.1 years (when following in long lines of individuals). 6. An understanding of upstream neritid snail migration can be used for the management and conservation of this and other migratory species in tropical streams.

Mountain streams that originally supported Hawaiian cultural practices have been diverted for development, agriculture and tourism for over 150 years. Habitat characteristics and benthic macroinvertebrate community responses to water withdrawal were studied in four West Maui Mountain watersheds. We compared riffle and cascade habitats upstream and downstream of the highest‐elevation diversion in each stream and further compared streams to understand variation among watersheds. Riffles were shallow areas with moderate flow, whereas cascades had high‐velocity water flowing over boulders and were separated into torrenticolous (submerged) and amphibious (splash zones on adjacent exposed rock) microhabitats. Among streams, downstream discharge was reduced by 84–99%, flow velocity was four times greater upstream, and depth was 50% lower downstream. There was a significant 44% reduction in downstream macroinvertebrate density (t = 3.261, df = 136, p = 0.0014); however, density did not significantly differ among streams (F = 1.95, df = 3, p = 0.125). Habitat‐corrected density, based on total available habitat area, indicated significantly greater proportions of native taxa in amphibious microhabitats compared with riffle and torrenticolous habitats. Non‐native Chironomidae and Trichoptera (Cheumatopsyche sp. and Hydroptila sp.) were dominant (>95%) and ubiquitous in riffles, whereas native Limonia sp. dominated (30%) amphibious microhabitats. Macroinvertebrate community structure varied among streams, sites and microhabitats, indicating inconsistent response to water withdrawal, dependent upon watershed size and microhabitat conditions. Our findings contribute to water management and restoration efforts focused on conservation of native species and habitat integrity in tropical streams worldwide. Copyright © 2013 John Wiley & Sons, Ltd.