River morphology of mountain streams influenced by an extreme windstorm in the High Tatra Mountains (northern Slovakia)

In this study we describe how the river morphology of seven High Tatras streams has been influenced in different ways by a severe windstorm that occurred in November 2004. A control site situated in an undisturbed area is compared with six sites in windstorm damaged areas. In 2009–2010 River Habitat Survey (RHS) was used to assess the character of 500 m lengths of stream and to derive associated morphological indices (HQA, HMS sensu Raven 1998, RHS indices sensu Vaughan 2010). Large amounts of large woody debris in the channel influenced the streams despite most of it having been removed after the windstorm. Streams situated in deforested valley slopes were probably more affected by erosion; they had higher amounts of transported inorganic matter and also depositional bar features. Some changes in land-use had occurred with streams situated on the south-east side of the High Tatras having a more uniform banktop vegetation structure than the other streams. Distinct secondary succession of bank top vegetation was observed along windstorm influenced streams. The highest percentage of shrubs occurred where impacted streams remained untouched. Morphologically, the streams most affected by windstorm are those situated in deforested steep valley slopes that are affected by erosion and siltation.     

Caddisflies (Trichoptera) as good indicators of environmental stress in mountain lotic ecosystems

We analysed the influence of environmental stress (mainly deforestation of catchment due to windstorm) on taxonomic structure as well as composition of functional groups (FG) based on zonation preferences and feeding types of caddisfly species in 11 streams (12 sites) with various environmental conditions situated in the High Tatra Mts, Slovakia. Using Spearman correlation, we confirmed an expected positive association between taxonomic and functional groups richness (Ri), diversity (Shannon — Wiener Index) (Di) and equitability (Eq), but not with habitat diversity (HQA indices). By testing measured physiographical, physico-chemical and hydromorphological factors using CCA analysis we found that stream order and temperature best explained the functional and taxonomic structure of caddisfly assemblages. The occurrence of Rhyacophila glareosa decreased with increasing daily mean water temperature positively correlated with deforestation, whereas R. tristis (dominating at sites most affected by erosion) showed the opposite pattern. Composition of functional groups based on zonation preferences and feeding types distinguished well near natural sites from impacted ones, dominated by crenal/rhithral predators, and explained more of the overall variance of the species-environment relationships than taxonomic composition of caddisfly assemblages.     

Biodiversity loss in Appalachian headwater streams (Kentucky,USA): Plecoptera and Trichoptera communities

Government and academic studies indicate that many streams in the Appalachian Mountains have degraded biological communities stemming from a variety of regional landuses. Headwater stonefly (Plecoptera) and caddisfly (Trichoptera) assemblages were assessed between 1999 and 2004 in relation to pervasive landuse disturbances (coal mining and residential) in mountainous areas of eastern Kentucky, USA. Indicator metrics (richness, abundance, tolerance, and an observed/expected (O/E) null model) were compared among 94 sites with different land use pressures including least disturbed reference, residential, mining, and mixed mining and residential categories. Thirty-three stonefly species from 26 genera and 9 families were identified; Leuctra, Acroneuria, Haploperla, and Isoperla comprised the core genera that commonly decreased with disturbance. Caddisflies were represented by 48 species, 32 genera, and 14 families. Core caddisfly genera (Neophylax, Pycnopsyche, Rhyacophila, Lepidostoma, and Wormaldia) were extirpated from most disturbed sites. Species richness was significantly higher at reference sites and reference site mean tolerance value was lowest compared to all other categories; relative abundance of both orders was variable between disturbance groups. Non-metric multidimensional scaling (for riffle-dwelling stonefly and caddisfly genera) clustered reference sites distinctly from most other sites. The O/E index was highly correlated with individual habitat and chemical stressors (pH, conductivity) and on average, it estimated ~ 70% loss of common stoneflies and caddisflies across all disturbed landuse categories. Expected plecopteran and trichopteran communities were radically altered in streams draining mining and residential disturbance. Long-term impacts incurred by both landuses will continue to depress these vulnerable indigenous fauna.     

The influence of environmental variables on larval growth of stoneflies (Plecoptera) in natural and deforested streams

A total 14 streams located in the Tatra Mts (Slovakia) were investigated. The studied river basins have different deforestation range (0–45.5%) influenced by windstorm. Daily and annual water temperature suggested that in streams overflowing through the area deforested by windstorm the temperature was higher than in selected non-disturbed streams. Higher average nutrient concentrations in samples from these streams are connected with nutrient mobilization from disturbed forest soils after the windstorm. Most of the damaged streams have higher trophic status, including POM and biofilms. The growth of the most abundant larvae increases with deforestation. Growth of species Leuctra rauscheri, Protonemura Montana, and Perlodes intricatus correlates with the sum of daily temperatures over a given time interval. Growth of species Brachyptera seticornis, Leuctra autumnalis and Isoperla sudetica are affected by stream trophy.     

Fish foraging effects on benthic assemblages along a warm-temperate stream: differences among drift feeders, benthic predators and grazers

In streams, physical and biotic conditions change from the headwaters to the mouth, shaping longitudinal patterns in community structure. We examined how fish foraging effects on periphyton and benthic invertebrates changed along a longitudinal gradient of a warm-temperate stream in southwestern Japan. We established three study sites according to changes in the composition of fish feeding guilds (upper site characterized by drifting-invertebrate feeders, Oncorhynchus masou ; middle site by benthic invertebrate feeders, Rhinogobius spp.; lower site by the presence of periphyton grazers, Sicyopterus japonicus ), and performed two manipulative experiments to examine effects of different fish assemblages on periphyton and benthic invertebrate abundances. Results of an exclosure experiment suggested that fishes had no effect on the benthic assemblages at the upper and middle sites whereas fishes reduced the abundances of both periphyton and invertebrates on stone surfaces at the lower site, where both benthic invertebrate feeders and grazers inhabited. A subsequent enclosure experiment showed that the reduction of invertebrate densities at the lower site was caused by the grazers rather than benthic invertebrate feeders. These experimental results suggested that effects of fishes on benthic assemblages are intensified downstream, owing to the occurrence of the grazing fish. Furthermore, observational data based on field sampling suggested that such grazing effects were reflected in longitudinal patterns in periphyton and invertebrate abundances. Overall results emphasize an important role of the grazing fish ( S. japonicus ) in shaping longitudinal patterns in benthic assemblage structure.     

A comparison of benthic macroinvertebrate assemblages among different types of alpine streams

1. Benthic macroinvertebrate assemblages were compared among a diverse array of first‐order alpine tundra streams of the Swiss Alps. 2. A principal components analysis separated sites into three main groups: rhithral streams, rhithral lake outlets, and kryal sites including outlets and streams. Rhithral streams contained the most diverse and taxon rich assemblages, being colonised by both non‐insect taxa and Ephemeroptera, Plecoptera, Trichoptera and Diptera. 3. Rhithral lake outlets supported high densities of non‐insect taxa such as Oligochaeta, Nemathelminthes and crustaceans. Despite low taxon richness, kryal sites had high Ephemeroptera and Plecoptera abundances. Chironomidae were most common at all sites. 4. Collector‐gatherers were dominant at all sites, whereas filter‐feeders were rare. Scrapers and shredders were more common in streams than lake outlets. 5. Water temperature and algal standing crops were higher at rhithral lake outlets than rhithral streams, perhaps providing more favourable habitat for non‐insect taxa. Glacial runoff was the dominant factor influencing macroinvertebrate assemblages of kryal streams and kryal lake outlets. Alpine lakes influenced the environmental conditions of their outlets and, consequently, their macroinvertebrate assemblages unless being constrained by a glacial influence.     

Structural and functional responses of stream insects to copper pollution

Benthic invertebrate assemblages and leaf decomposition rates were usedto assess the effects of copper on community structure and function in EastProng Creek, VA. Copper concentrations ranged from 12–32 µgl^–1 at three downstream disturbed sites and 0.1–2.0µg l^–1 at an upstream reference site. Benthic samplingshowed reduced taxonomic richness and numbers of individuals at thedisturbed sites and leaf decomposition rates at the disturbed sites were1.4–2.7 times slower than at the upstream reference site. Partialrecovery of community structure occurred after a passive treatment systemreduced copper to levels below chronic effect concentrations. Post-treatmentbenthic sampling indicated a recovery in taxonomic richness in the threedisturbed sites. During the 6-week period of treatment, themacroinvertebrate community in the disturbed sites recovered to a similarcomposition to that found in the reference site. However, no significantdifference in leaf decomposition rate was observed, suggesting a residualeffect of copper on our measure of functional response. Our results indicatethat copper pollution significantly interrupts the action of shreddersprocessing leaf material in polluted streams.     

Top-down and bottom-up influences on populations of a stream detritivore

1. Knowledge of the influence of predatory fish in detritus‐based stream food webs is poor. We tested whether larval abundance of the New Zealand leaf‐shredding caddisfly, Zelandopsyche ingens (family Oeconesidae), was affected by the presence of predatory brown trout, Salmo trutta and the abundance of their primary detrital resource (Nothofagus leaves). 2. The density of Z. ingens and the biomass of leaves were determined in seven fishless streams and four trout streams in the Cass region, central South Island, on four occasions spanning 5 years. 3. Physicochemical conditions were similar in trout and fishless streams, but ancova indicated that Z. ingens numbers were positively related to leaf biomass and that caddisfly numbers were significantly greater in fishless streams than trout streams for any given biomass of leaf. The cases of trout stream larvae were also heavier per unit length than those in fishless streams. 4. Our results provide evidence for both top‐down and bottom‐up influences on a detritus‐based stream food web. Although stream detritivores may benefit from a habitat that provides both food and a degree of protection from predators, top‐down effects of predators on detritivore population abundance were still important. Thus, detrital resource availability may determine maximum attainable population size, whereas predation is likely to reduce the population to a level below that.     

Tufa Barriers from a Caddisfly's Point of View: Streams or Lake Outlets?

The caddisfly community composition in different microhabitats at tufa barriers was studied in the Plitvice Lakes NP, Croatia. Three tufa barriers were investigated and six emergence traps were installed at each site covering various microhabitats. Sampling was conducted monthly during the year 2008. Tufa barriers are lake outlet habitats, but they feature a variety of microhabitats similar to streams, which is reflected in the composition of caddisfly communities regarding longitudinal distribution preferences. The caddisfly communities at all three sites were dominated by species typical for the rhithral zone, but there was a shift in species composition along a longitudinal gradient, from the epirhithral to the metarhithral zone. Analysis of functional feeding guilds showed considerable differences between the caddisfly community at the Labudovac barrier and both downstream barriers, shifting from one with a quite diverse composition, to one completely dominated by passive filter‐feeders. Passive filter feeders were not represented by the same taxa at up‐ and downstream barriers (i.e., by Hydropsyche species and Wormaldia species, at the Labudovac barrier and at both downstream barriers, respectively). Due to high complexity and habitat diversity, the highest diversity and equitability of caddisfly communities were recorded at microhabitats with particulate tufa and medium current velocity (10–20 cm/s). The lowest diversity and species richness were recorded for silt with low current velocity (0–10 cm/s). Abundance of caddisflies was positively correlated with current velocity due to a very high proportion of rheophilic passive filter feeders in the communities. However, community composition and structure is only to some degree influenced by microhabitat characteristics, but rather by their general position within the barrage‐lake system, i.e., longitudinal distribution and thereby availability of different food resources. (© 2012 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Response of bird assemblages to windstorm and salvage logging — Insights from analyses of functional guild and indicator species

Natural disturbances, such as fire, windstorms and insect outbreaks, are important drivers of biodiversity in forest ecosystems, but at the same time cause large economic losses. Among the natural disturbances in Europe, windstorms cause the highest economic loss. After such storms, damaged forest stands are commonly salvage logged to restore economic value. However, such interventions could affect species assemblages of various taxonomic groups, including breeding birds. Despite these potential effects, investigations of the impacts of post-storm logging are largely lacking. We thus investigated assemblages of breeding birds in 21 logged and 21 unlogged windstorm-disturbed forest plots and 18 undisturbed, control forest plots using fixed-radius point-stop counts three, five and seven years after a windstorm within the Bavarian Forest National Park as part of the European Long-Term Ecosystem Research Network. We recorded 2100 bird individuals of 55 bird species. Bird assemblages were predominantly altered by the consequences of the windstorm and affected only to a minor degree by subsequent logging of storm-felled trees. Nevertheless, bird species richness was significantly reduced by post-storm logging within the first season. In general, the windstorm led to a shift in bird assemblage composition from typical forest species towards open- and shrub-land species. Assemblages of logged and unlogged disturbed plots consisted mainly of long-distance migrants and ground-foraging bird species, whereas assemblages of undisturbed control plots consisted of resident species that forage within vegetation. Both unlogged and logged storm-felled coniferous forest stands were inhabited by endangered or declining bird species, such as Water Pipit (Anthus spinoletta) on logged plots and Eurasian Redstart (Phoenicurus phoenicurus) on unlogged plots. Indicator species analyses suggested that species of unlogged storm disturbed plots depended on storm-created legacies, such as pits and snags, for foraging and nesting. Hence, we recommend reducing post-storm logging of these legacies to support species restricted to unlogged disturbed forest. To increase the diversity of breeding birds on the landscape scale, diverse logged and unlogged post-disturbance stands should be provided, which could be gained by a partial benign neglect strategy of storm disturbed forest stands.     

Characterization of benthic bacterial assemblages in a polluted stream using denaturing gradient gel electrophoresis

To study differences in bacterial assemblages among sites with different environmental conditions, sediment samples were collected from three sites along a South Carolina (U.S.A.) coastalplain stream with varying levels of anthropogenic perturbation. The objective of this study was to compare the bacterial assemblages among these sites to detect possible impacts from the disturbance. To accomplish this comparison, DNA was extracted from the samples and subjected to the polymerase chain reaction using primers designed to amplify bacterial 16S rRNA genes. Relative measures of bacterial genetic diversity, assessed using denaturing gradient gel electrophoresis (DGGE), revealed greater numbers of unique sequences at the disturbed sites. The number of bands, which is analogous to species richness, did not vary predictably among sites. Similarity indices revealed a high level of similarity among replicate samples from each site and low similarity between samples from different sites. This study demonstrated that bacterial assemblages differed among sites and that the presence or absence of certain species, represented by unique DGGE bands, differed among sites; unique bands were most commonly encountered at the disturbed sites. Based on the evidence gathered, we conclude that benthic bacterial assemblages vary longitudinally and that anthropogenic disturbance may alter the bacterial component of streams.     

Effects of natural hydrological variability on fish assemblages in small Mediterranean streams: Implications for ecological assessment

Small Mediterranean streams are shaped by predictable seasonal events of flooding and drying over an annual cycle, and present a strong inter and intra-annual variation in flow regime. Native fish assemblages in these streams are adapted to this natural environmental variability. The distinction of human-induced disturbances from the natural ones is thus a crucial step before assessing the ecological status of these streams. In this aim, the present study evaluates the effects of natural hydrological variability on fish assemblages from disturbed and least disturbed sites in small intermittent streams of south Portugal. Data were collected over the last two decades (1996–2011) in 14 sites located in the Guadiana and Sado river basins. High variability of fish assemblages was strongly dependent on human-induced disturbances, particularly nutrient/organic load and sediment load, and on natural hydrological variability. Natural hydrological variability can act jointly with anthropogenic disturbances, producing changes on fish assemblages structure of small intermittent streams. In least disturbed sites, despite the natural disturbances caused by inter-annual rainfall variations (including drought and flood events), fish assemblages maintained a long-term stability and revealed a high resilience. On the contrary, disturbed sites presented significantly higher variability on fish assemblages and a short and long-term instability, reflecting a decrease on the resistance and resilience of fish assemblages. Under these conditions, fish fauna integrity is particularly vulnerable and the ecological assessment may be influenced by natural hydrological variations. High hydrological variability (especially if it entails high frequency of dryer years and meaningful cumulative water deficit) may affect the impact of the human pressures with significant and consistent consequences on fish assemblage composition and integrity. In this study, fish metrics that maximize the detection of human degradation and minimize the response to natural variability were based on the relative abundance of native species (insectivorous species, eurytopic species, water column species, native lithophilic species), relative abundance of species with intermediate tolerance and relative number of exotic species. Results highlight the importance of assessing temporal variability on stream biomonitoring programs and emphasize the need to improve the assessment tools, accounting for long-term changes in fish assemblages, namely by selecting the most appropriate fish metrics that respond to anthropogenic disturbances but exhibit low natural temporal variability, essential both in the characterization of the biological reference conditions and in the development of fish indexes in intermittent streams.     

Assessing stream dwelling caddisfly assemblages (Insecta: Trichoptera) collected by light traps in Hungary

In the present study the conservation status of light trap-collected stream dwelling caddisfly assemblages was evaluated on the basis of original data and also from the literature representing different regions in Hungary. Altogether the catches of eight complete seasons were compared. Species richness, diversity and endangerment indices as well as a newly introduced rarity index (RI) expressing the average rarity of the collected species in Hungary were calculated to evaluate the conservation status of the streams in the Bakony, Börzsöny and Bükk Mountains and in the Zala Hills. The results highlight the importance of streams in maintaining endangered species of caddisfly in Hungary and bring to attention the fact that species richness or diversity are not necessarily the best indicators of conservation status of caddisfly assemblages. Indicators, for instance RI, which take into consideration the general rarity of the species seem to reflect more sharply conservation status and thus are more appropriate for the assessment of caddisfly assemblages.     

Site effects outweigh riparian influences on ground-dwelling beetles adjacent to first order streams in wet eucalypt forest

In wet eucalypt forest with a rainforest understorey the vegetation adjacent to first order streams does not form a distinct riparian strip. This study investigated the riparian response of terrestrial ground-dwelling beetles adjacent to four such streams in Tasmania, Australia. Beetle assemblages varied more between the four sites than they did with distance from stream within sites, where they exhibited a measurable but subtle riparian response. The extent of the riparian zone varied between the four study sites, with a 1–5 m riparian zone at three sites and a gradually changing community up to 50–100 m upslope at one site. There was a trend for greater between plot variability immediately adjacent to the streams, possibly because this is a more highly disturbed environment. None of the habitat variables measured were consistently associated with riparian or upslope assemblages of beetles, probably explaining the subtlety of the beetles’ riparian response. Forest conservation efforts for terrestrial species should not necessarily be focused on the riparian zone in preference to upslope areas.     

The influence of microhabitat on availability of drifting invertebrate prey to a net-spinning caddisfly

SUMMARY. 1. Invertebrate drift was sampled at both a rockface and a deep pebble riffle site in streams draining both a clearcut and a forested catchment.
2. A sampler was designed to separate the bottom 2 cm of flow, encompassing the effective range of caddisfly (Hydropsychidae: Trichoptera) catchnets. from upper flow.
3. No significant difference in drift density (numbers per cubie metre) was seen between sites within each stream. However, numbers per square centimetre intake area per day at the rockface sites were 4 times higher in the clearcut and 10 times higher in the forested stream than at the pebble-riffle site.
4. Rockface habitat which had highest drift availability was also the site of maximum secondary production of the predaceous collector- filterer Parapsyche cardis in both streams studied.
5. Increased sediment load in the clearcut stream may influence the efficiency of utilization of invertebrate drift by collector-filterers.     

Characteristics of caddis larvae assemblages from shallow lakes in the Bükk Mountains,North Hungary

Seventeen caddisfly species of four families (Limnephilidae, Sericostomatidae, Phryganeidae, and Leptoceridae) with a total of 1047 individuals were collected from four sampling sites in the littoral zone of Disznóskút Lakes in the Bükk Mountains. Three of these four families of Trichoptera were also recorded in the Central European lakes. From among the four substrates (sand, silt, large stones, and aquatic plants) of the littoral zone, silt and aquatic plants were dominant. Besides species of Limnephilidae found in higher than the average abundance in the silt, species of Phryganeidae and Leptoceridae also occurred. The trophic relationships of the species of these families can be considered as follows: shredders- predators- scrapers, shredders- detritivores- predators, and predators- shredders- scrapers. From the aquatic plants the species of Limnephilidae, Phryganeidae and Leptoceridae, belonging to the trophic relationships of detritivores- shredders, scrapers- predators, shredders- detritivores- predators, and predators- shredders- scrapers- detritivores were collected. The analysis (PCA) of the spatial and seasonal distribution of Trichopteran larvae at the four littoral zone sites shows that the assemblages collected near the inflow of the stream into the lake in late summer and near the inflow of the small channel into the lake in summer are very different from the others, for which Limnephilus elegans, Micropterna lateralis, and Stenophylax permistus are responsible at the previous site, while Halesus digitatus, Halesus tessellatus, and Mystacides nigra are responsible at the latter one. The other assemblages (mostly those in winter) are very similar to each other, which can be explained by the decreasing number of detectable species and their presence in low abundance. For the variance, Limnephilus rhombicus (57%) and Potamophylax nigricornis (16%) are essentially responsible. Considering the species composition of the larval assemblages, the site at the inflow of the stream into the lake was the most different (38%) from the others, and was 62% similar to them. This dissimilarity can possibly be explained by the presence of a mixed assemblage with typical stream and lake inhabitants of species of Limnephilidae present in high abundance at this silt dominated site and other caddis species that are typical stream inhabitants or common in streams and small rivers, and may have drifted into the lake.

     

Predictable changes in trophic community structure along a spatial disturbance gradient in streams

1. We assessed insect and resource standing stocks along a spatial gradient of flood disturbance in 19 sub-alpine Swedish streams to test the prediction that change in trophic structure arises from the joint action of disturbance, which affect basal resources, and resource-control, which ties the response of the consumers to the response of the resources.
2. Trophic structure, quantified as scores of non-metric multidimensional scaling based on the biomass of insect trophic groups, changed predictably along the disturbance gradient. In early summer, predators and algae feeders decreased relative to suspension feeders with increasing disturbance; in autumn, algae feeders decreased relative to leaf feeders with increasing disturbance.
3. Across the disturbance gradient, the biomass of algae-, deposit- and leaf-feeders was principally controlled by the availability of the respective resource (algae, fine detritus and coarse detritus), while disturbance only had subsidiary effects on algae feeders in early summer.
4. Overall, patterns in trophic-group biomass along the disturbance gradient were more likely to reflect indirect effects of disturbance via impact on the resources, which reverberated to the consumers because of resource-control, rather than direct effects. In contrast with the view that stream communities are the result of stochastic colonization following disturbance events, in the study streams the trophic structure of insect assemblages is predictable and partly organized by resource-control across a broad range of disturbance conditions.     

Diversity,occurrence and feeding traits of caddisfly larvae as indicators for ecological integrity of river-floodplain habitats along a connectivity gradient

In order to assess ecological values of Lower Rhine and Meuse floodplain habitats we studied the spatial and seasonal variation in diversity, species assemblages and feeding traits of caddisfly larvae in water bodies over the lateral connectivity gradient: eupotamon: main and secondary channels; parapotamon: channels connected permanently with the main channel only at their downstream ends; plesiopotamon: disconnected channels close to the main channel; paleopotamon: abandoned meanders at a greater distance from the main channel.Spatial variety was studied by analyzing the summer species composition in 70 Lower Rhine and Meuse water bodies which were categorized in connectivity habitats, whereas seasonal variety was studied in Lower Rhine water bodies along a connectivity gradient by monthly sampling over a whole year. Physico-chemical data and environmental parameters were recorded for each water body during sampling. Diversity and species assemblages of caddisfly larvae varied in relation to connectivity, macrophyte diversity and abundance and stream velocity. A comparison with historical records and species lists from less disturbed rivers showed that diversity in the main channel was very low.Caddisfly larvae species assemblages varied over the connectivity gradient. Lotic habitats (eupotamon) were separated from the lentic ones, and the well vegetated paleopotamon from the sparsely vegetated parapotamon and pleisopotamon habitats, indicating the overall importance of vegetation and current velocity for the species assemblages. Hydropsychidae have been found in the eupotamon exclusively, whereas Limnephilidae, Hydroptilidae and Polycentropodidae have been found predominantly in the paleopotamon water bodies. Leptoceridae were found in all floodplain water body categories. A similar pattern of distribution of families along the lateral connectivity gradient was found in more natural rivers.Caddisfly larvae species feeding traits showed a clear differentiation over the lateral connectivity gradient with filter-feeders and scrapers most important in the eupotamon and parapotamon, and shredders, piercers and predators most dominant in the paleopotamon habitats, indicating the importance of nutritional resources in relation to hydrological connectivity for the structure and functioning of caddisfly larvae species assemblages. The analysis of the species feeding traits allows generalizations towards the entire aquatic community and general prognoses for other floodplain ecosystems.     

Spatially variable response of native fish assemblages to discharge,predators and habitat characteristics in an arid‐land river

1. Fish assemblages and habitats were sampled annually at fixed sites in three tributaries of the Gila River catchment over a 21‐year span that included prolonged low‐ and high‐flow periods. Model selection was used to evaluate responses of seven native fishes with variable ecological traits (four small‐bodied cyprinids, one large‐bodied cyprinid, and two large‐bodied catostomids) to mean annual discharge and predacious non‐native fishes across the three sites. We also compared habitat use and overlap of native and non‐native fishes to identify potential for negative interactions among species. 2. Assemblage structure (species abundance and richness) and recruitment of native species was strongly and primarily affected by mean annual discharge and secondarily by location and densities of non‐native predators (mainly the centrarchid Micropterus dolomieui). 3. Densities of age‐0 catostomids and small‐bodied cyprinids were positively associated with discharge, and this pattern was strongest in the tributary with the lowest densities of non‐native predators. Absence or extreme low abundance of natives during low‐flow years was most pronounced at the sites where non‐native predators were comparatively common. Densities of adults of large‐bodied native species also varied by site, but often were positively associated with densities of non‐native predators. 4. Spatially variable responses of native fish assemblages indicated that the persistence of native fishes could be jeopardized if key habitats were lost or flow regimes unnaturally altered, particularly during low‐flow conditions when recruitment of native fishes is low and predation by non‐natives is high. Large‐bodied species may be less vulnerable to multiple years of poor conditions because adults are able to avoid predation by non‐natives and thus can rely on occasional high discharge years for successful recruitment. 5. As in other arid‐land streams, native fish assemblages of the Gila River Basin continue to decline. Our results indicate that conservation requires specific knowledge and consideration of physical influences as well as life‐history attributes of native and non‐native fishes.