Upstream migration is an important part of the lifecycle of aquatic invertebrates, which essentially serves the purpose of distribution and drift compensation. The aim of the present study was to investigate the migration performance of different species at different flow velocities in the waters of the young moraine landscape of Mecklenburg-Western Pomerania. Results allow drawing conclusions on the performance of aquatic invertebrates in fish migration facilities.

Aquatic communities across the Eurasian steppe face increasing anthropogenic pressures due to rapid population growth, catchment-wide land-use changes, and climate change. The particular type, intensity, overlay, and legacy of impacts along longitudinal gradients of Eurasian river networks provide a unique setting to investigate ecological responses in identifiable multiple stressor environments. We studied macroinvertebrate communities along the Kharaa River, Mongolia, which display a distinct, downstream gradient of moderate nutrient enrichment, disturbed bank morphology, reduced riparian vegetation, elevated turbidity, increased fine sediment substrate proportions, and fine sediment intrusion into the hyporheic zone. Within the encountered ranges of physical and chemical environmental factors (TP 0.02-0.09 mg/L, TN 0.33-0.96 mg/L, conductivity 167-322 mu S/cm, formazin nephelometric units 0.62-5.43) and hyporheic fine sediment intrusion (0.9-1.6 g dry weight [DW]center dot L-1 center dot day(-1)) the population densities and biomass of macroinvertebrates were high (5,313 +/- 410 individuals/m(2) and 2,656 +/- 152 mg DW/m(2)) and notably stable. In contrast, macroinvertebrate community structure showed strong and statistically significant negative linear relationships (Pearson's r) with turbidity, that is, for taxa richness (r = -.83), Shannon index of diversity (r = -.89), Evenness (r = -.86), the relative abundance of Ephemeroptera, Plecoptera, and Trichoptera (EPT) individuals (r = -.93) and relative biomass of hard substrate colonizers (r = -.86). The relative biomasses of fine substrate colonizers, as well as Chironomidae and Oligochaeta (both r = .76), were positively correlated with mean turbidity values. In addition, the Proportion of Sediment-sensitive Invertebrates (PSI) methodology was adjusted for local application and the resulting index scores followed a similar pattern, with PSI also being significantly correlated (r = .66) with the relative abundance of EPT individuals, the latter being the most sensitive macroinvertebrate community index. We conclude that fine sediment load is the key factor for shaping macroinvertebrate community structure in the multistressor setting of the Kharaa River followed by hydromorphological habitat complexity determined by shear stress, substrate, and grain size distributions. We suggest that the implementation of effective regional management strategies aiming at the reduction of fine sediment pollution should be given the highest priority.

The aim of this study was to support the development of ecological stream quality assessment tools in order to provide a method for sustainable water management in Turkey. Therefore, we present two new or adapted indices based on benthic invertebrates. To develop and adapt the indices, 17 streams were studied and separated into three quality classes, which were supported by four community indices (EPT [%], EPTCBO [%], number of Individuals, evenness), and 23 taxa were identified as indicators for these three quality classes. As a first biological index, we adapted the Hindu Kush-Himalaya biotic score (HKHbios) to the Euphrates catchment by establishing a new and ecoregion-specific score list (Euph-Scores) by scoring 93 taxa depending on their distribution between the quality classes. Based on these scores, several average score per taxon values (ASPT value) were calculated. All ASPT values of the Euph-Scores separated the quality classes significantly. After a comparison of the different ASPT values we recommend to use the weighted ASPT, because the weighting enabled a sharper differentiation between the quality classes and named it Euphrates Biotic Score (EUPHbios). As a second biological index, we propose the proportion of habitat specialists. To calculate this index, a habitat score was developed by analysing the habitat preferences of several benthic invertebrates. Habitat score values were assigned to the 20 most common taxa from the streams in the best quality class (natural streams). The proportion of habitat specialists, identified using the new habitat score, differed significantly between the three quality classes, with higher values in natural streams than in polluted streams. In the light of the results, the presented methods appear to be suitable for developing a multi-metric index for assessment programs for the mountainous regions of the Middle East.

The invasion of the Ponto–Caspian amphipod Dikerogammarus villosus in European rivers is assumed to reduce macroinvertebrate diversity and to alter ecosystem functions. D. villosus shows an extraordinarily flexible feeding behavior including the ability to use various food sources. On the other hand, its response to predation risk seems to depend on environmental factors. To evaluate the ecological function of D. villosus, we estimated the daily food consumption for different food sources and analyzed potential effects of predator avoidance behavior on feeding. D. villosus consumption of willow leaves or chironomid larvae was quantified in 24-h laboratory experiments with and without kairomones of the European bullhead (Cottus gobio). Consumption rates were estimated based on gut content and gut evacuation rate under semi-natural laboratory conditions enabling the animals to feed over the whole time of the evacuation rate experiment. We observed very high evacuation rates and consequently high consumption rates up to 89% of body weight per day. Consumption rates differed significantly between food sources: D. villosus ingested more leaves than chironomid larvae. In contrast, predator cues did not affect the feeding of D. villosus. This might be explained by its strong refuge affinity and probably benefits its successful invasion. A comparison of the estimated consumption rates with results of an own consumption experiment (and other studies) under more artificial conditions indicated that more natural conditions result in higher consumption rates. Consequently, feeding rates from highly artificial experiments should be used with great caution to assess the ecosystem function of D. villosus.

Invasive species can affect native communities by replacing competitors, overexploiting prey species or altering ecosystem structure. One example is the Ponto-Caspian amphipod Dikerogammarus villosus which has established large populations in European rivers and is widely considered as the main cause for the decline of native benthic invertebrates. This effect has been mainly associated with direct predation, whereas the indirect effects via competition for primary resources are poorly understood and possibly underestimated. To assess the probability of these indirect effects, we performed five outdoor flow-through mesocosm experiments in three European rivers, manipulating the density of D. villosus. We quantified its in-situ food consumption during three 24-h gut content surveys in the mesocosms. Gut evacuation rates for correction were measured in the laboratory for different food sources and under continuous feeding. We analysed the invader's effects on primary resources by quantifying periphyton biomass and community leaf litter decomposition in the mesocosms at different D. villosus densities. The remarkably high food consumption rates (0.38-1.27 mg mg(-1) d(-1), in dry mass/dry body mass) of D. villosus can be attributed to its high gut evacuation rates. The leaf litter decomposition rates indicate that D. villosus is an efficient shredder; however, there was no effect on periphyton biomass. Our results indicate that D. villosus may be a strong competitor with primary consumers in benthic food webs of invaded rivers, with both direct and indirect negative effects on benthic communities. High consumption rates together with opportunistic feeding behaviour probably promote the invasion success of this amphipod.

Dikerogammarus villosus, one of the most successful invaders in European river systems, is commonly regarded as a threat to native biodiversity and a main factor structuring the benthic community of invaded systems. The impact of D. villosus has been intensively studied in small-scale experiments and field observations, but its impact on natural communities on a larger scale remains unclear. Here, we investigated the benthic community structure at ten sites covering a broad range of habitats along the River Rhine (Central Europe) and its tributaries, to determine whether D. villosus is one of the main factors structuring the benthic community. Community composition was analysed using non-metric multidimensional scaling, distance-based redundancy analysis, and correlation analyses. D. villosus was one of nine relevant taxa present that altogether reflected a large part of the variation in the benthic samples, but further analyses indicated that the species might be less important for the community structure than other relevant taxa. Moreover, all nine relevant taxa together can explain only a similar amount of variation in our samples than the five relevant non-faunal environmental factors (water temperature, pH, conductivity, percentage of medium-sized gravel and macrophytes). Overall, our results suggested that rather a combination of non-faunal environmental factors than D. villosus mainly structure the benthic community composition on this larger spatial scale.

The invasive amphipod Dikerogammarus villosus is assumed to threaten native biodiversity in rivers. In spite of small-scale experiments and field observations, its impact on natural communities is largely unknown because it seems to be variable and long-term analyses are rare. We analysed long-term data from the Upper Elbe and Middle Rhine (Germany) for invasion patterns and changes in the community structure. In addition, mesocosm experiments were performed in both rivers to identify density effects of D. villosus on the communities. We assumed that D. villosus is a driver of changes in the macroinvertebrate community and that effects are river-specific due to differing benthic communities. We found two invasion patterns for D. villosus with fast invasion in the River Elbe and slower invasion in the River Rhine. The impact of D. villosus on the species composition was weak in both river communities. Invasion seems to have reduced taxa number and individuals and increased Shannon diversity in the River Rhine, but not in the River Elbe. The correlations between the densities of the invader and other taxa in the long-term data were mostly positive with the exception of two native taxa in the River Rhine, indicating a lack of strong negative species interactions. Also in the mesocosm experiments, the biomass gradient of D. villosus adults did not cause significant changes in the communities. The community in the River Rhine seemed to be more vulnerable to the D. villosus invasion than that in the River Elbe. This might be caused by a dominance of invasive species interacting positively with one another, as suggested by the 'invasional meltdown' theory. The study suggests that community-level effects of invasion may differ between rivers, probably due to differences in the community composition.

The characterisation of natural stream conditions is the first important step to analyse ecological quality of streams in the Euphrates basin. We found that the community indices correspond to very good ecological conditions in five natural streams of that region. The macroinvertebrates composition differed significantly between September and May. Number of taxa and Shannon index were significantly higher in autumn than in spring. FPOM and biofilm were the most relevant basal resources of benthic invertebrates.

The estuarine maximum turbidity zone (MTZ) can be assumed to be a stressful environment featuring special conditions of great biological importance with an excess of organic matter, high-deposition rates, large variations in salinity, and dredging activities. Under such harsh conditions, populations may remain below the carrying capacity and competition is assumed to be of little importance, as predicted by the stress-gradient hypothesis. Therefore, we hypothesized that invertebrates of similar feeding types may utilize the same resources. To test our hypothesis, we chose the three most abundant taxa classified in literature as deposit feeders (Bathyporeia pilosa, Boccardiella ligerica, Marenzelleria sp.) and two taxa classified as predominately predacious (Palaemon longirostris, Crangon crangon) and determined their isotopic niches based on a stable isotope analysis for the MTZ of the Elbe Estuary (Germany). We expected the isotopic niches of similar feeding types to show a clear overlap if our hypothesis was true. Our results showed that the isotopic niches of no two taxa overlapped within each feeding group, indicating different resource use and the absence of competition. The sediment analysis revealed that two of the deposit feeders inhabited significantly different mean grain sizes. The lack of overlap of isotopic niches within each feeding group may be due to differences in habitat and feeding behavior in the case of the deposit feeders and due to different migration behavior in the case of the predominately predacious species. However, competition may have occurred in the past, resulting in a divergence of feeding niches during evolution.

The invasive amphipod Dikerogammarus villosus has become increasingly dominant in German river ecosystems since the River Rhine invasion in the mid-1990s. Because it is preying on other invertebrate taxa, its appearance is often assumed to be associated with a drastic decrease in species diversity and changes in natural benthic communities. Despite this, the trophic function of D. villosus and its predation potential are rarely studied in natural river communities. Here, we assess the trophic function of D. villosus in two invaded systems, the Elbe River and the River Rhine, using stable carbon and nitrogen isotope analyses. In the two studied river food webs, D. villosus had a generally low trophic position (TP), indicating a likely function as primary consumer or at most as an omnivore. The significantly higher TP in the River Rhine (TP = 2.6) than in the Elbe River (TP = 1.9) suggested a partial use of animal prey in the River Rhine. This was supported by the results of the isotope mixing model SIAR which predicted in addition to a high importance of herbivory in both rivers a likely feeding on other invasive amphipod species in the Rhine community. We conclude that D. villosus has a variable but mostly low predacious behaviour, probably depending on the benthic community structure. To allow more realistic trophic analyses for D. villosus in the field, we determined the trophic enrichment factors (TEFs) for nitrogen and carbon isotopes of two diet types in the tissue of D. villosus in an eight-week laboratory experiment. The carbon isotope enrichment ( 1 3 C) from the chironomid diet was negative in the consumer (-1.75 ‰), whereas the carbon from leaf litter was enriched positively in the tissue (3.27 ‰). The TEFs for nitrogen ( 1 5 N) were more similar for plant and animal diet (leaf litter: 3.56 ‰, chironomids: 2.29 ‰). Because of the high specificity of the enrichment factors with respect to species and food source, we recommend their use in further trophic analyses.

The prey selectivity of fish depends largely on traits of the prey and the predator. Preferable prey traits might be different for visual predators (such as drift-feeding salmonids) and rather non-visual predators (such as benthic feeders). We evaluated the explanatory power of five prey traits and prey long-term abundance for the prey selection of small benthivorous fish by analysing the macroinvertebrate community and the diet of gudgeon (Gobio gobio) and stone loach (Barbatula barbatula) in two small submontane streams. Fuzzy principal component analyses, as well as electivity indices, revealed that the fish fed selectively. Prey size and feeding type were the most descriptive variables for the fish diet, followed by mean abundance, whereas microhabitat preference, locomotion mode and current velocity preference were less important. The fish preferred prey that was both small and consistently abundant, grazers and sediment feeders. Larger prey and shredders were avoided. The selection patterns of both fish species differed from those of visual fish predators but strongly resembled each other. Supporting this, in gudgeon which feeds slightly more visually than the strictly nocturnal stone loach, selectivity concerning prey traits as well as prey mean abundance was slightly more pronounced. We analyzed also selectivity for prey clusters based on the three most important variables. The observed selectivity patterns concerning these clusters were less pronounced but supported the other results. The maximum (neutral) electivity index was that of gudgeon for small, abundant grazers or sediment feeders, including chironomids.We conclude that prey selection of benthivorous fish that forage mainly non-visually can largely be explained by a small number of prey traits which probably work in combination. The prey preferences of these predators seem to be closely connected to their active foraging mode and to depend partly on the ability to detect prey visually.

Diverse benthic communities in streams include a wide variety of predators with different habitat preferences, e.g. for pools or riffles. We hypothesised that these preferences result in mesohabitat-specific predator community structures with quantitative differences concerning predation intensity by vertebrate and invertebrate predators, importance of intraguild predation, or top–down pressure. This hypothesis was evaluated for a small submontane stream by means of mesohabitat-specific quantification of prey consumption by two benthivorous fish species (Gobio gobio and Barbatula barbatula) and several invertebrate predators. The estimation was based on daily food rations and diet composition of predators and mesohabitat-specific predator biomass. We found clear differences between the two mesohabitat types. Predator food webs were less complex in pools than in riffles. Fish predation was more important than invertebrate predation in pools, and intraguild predation had a higher relative importance in these mesohabitats. These differences were probably caused by the mesohabitat use of G. gobio, the largest top predator, which preferred pools. Consequently, the predator food webs were more similar between the mesohabitats when fish were absent. Top–down pressure on primary consumers by all predators together was lowest in pools without fish, but the effect was not significant. Omnivory (including cannibalism) was intense, but its potentially destabilising effects were probably counterbalanced by mesohabitat connectivity. From the results of our experimental study, we conclude that even in small stream ecosystems, food web structures and predation pathways can differ between mesohabitats and that a mesohabitat-specific consideration will help to explain the variety of top–down effects on benthic communities.

Summary We conducted a paired large‐scale predation experiment over 32 months in two streams being seasonally shaded by deciduous riparian trees, using the benthivorous fish species gudgeon (Gobio gobio) and stone loach (Barbatula barbatula) as top predators. The biomass of benthic grazers and periphyton in the presence/absence of fish was measured and the periphyton production was compared with the consumption rates using a model‐based approach. A three‐level trophic cascade from benthivorous fish via benthic grazers to periphyton was evident from the field experiment. Integrated over the whole study period, fish reduced the biomass of benthic grazers and indirectly increased the periphyton biomass. Scenario analyses, using a simple dynamic model, indicated top‐down control of periphyton to be strongest during autumn, when periphyton growth was light‐limited, and weaker in the spring, when periphyton growth was not light‐limited. The seasonal light supply variation was caused by shading due to deciduous riparian trees during the vegetation period. This asymmetry in temporal processes weakened the top‐down control in a natural benthic community. Even though grazer biomass is naturally reduced in summer, due to the emergence of the most abundant species (mayflies), a grazer biomasses high enough to reduce the spring periphyton peak could not be sustained by the low summer periphyton growth. We suppose that the temporal decoupling of grazer biomass from periphyton biomass might be caused by the very short generation time of the primary producers (days) compared with the long generation time of the primary consumers (mostly 1 year).

Flexibility is an important adaptive attribute of the feeding periodicity of grazing mayfly larvae because most natural environments offer a wide variation in local predation risk in terms of space, time or predator species. In this study any changes of diel feeding periodicity and consumption rates of Baetis rhodani (Ephemeroptera) were analysed in response to different densities of benthivorous fish (Barbatula barbatula, Gobio gobio) by quantifying gut fullness using the fluorescence of algal pigments. Laboratory experiments with the grazer species B. rhodani were conducted by using different concentrations of chemical fish cues. In order to assess the transferability of the results to a larger scale, the experimental results were compared with field observations in two second order streams using different densities of freely foraging benthivorous fish. During the presence of chemical fish cues in the laboratory experiments the feeding periodicity of the B. rhodani larvae were mostly diurnal while in the absence of fish chemicals nocturnal feeding was observed. The same patterns could be detected in the field during the experiments with the different fish densities. These findings indicate that the larvae were able to assess variations in the predation risk and to alter their feeding habits by making flexible behavioural adjustments. The results from the laboratory experiments further suggested that the behavioural response is controlled by fish density. Behavioural changes were observed for medium and high concentrations of the fish cues but not for a very low concentration. In the field however, the mere presence of fish seemed to be sufficient to induce the observed behavioural shifts. Although the presence of benthivorous fish seemed to cause a lower consumption rate of the B. rhodani larvae in the field, such a reduction could not be found in the laboratory experiments. A conclusion from this study is that the identification of behavioural modifications is an essential component needed for a better understanding of complex trophic interactions in benthic communities. Accurate evaluation and detailed observation of direct and indirect effects cannot be made without the consideration of such behavioural mechanisms.

Regulating mobility by actively entering the drift under imminent predation risk is an avoidance strategy employed by aquatic macroinvertebrate species that is widely accepted within the scientific community. This response was most evident with respect to diurnal predators that feed in the water column, such as many salmonids. We investigated the role of the nocturnal benthivorous gudgeon [Gobio gobio (L.)] on the drift activity of two macroinvertebrate species known to display this behaviour: Baetis rhodani (PICTET) and Gammarus pulex (L.). Laboratory drift experiments using gudgeon kairomones were conducted with the results determining significant altered activity in the presence of gudgeon kairomones for both macroinvertebrate species. B. rhodani showed reduced drift activity in the kairomone treatment compared to the kairomone-free control, with a distinct nocturnal pattern being observed for both. G. pulex shifted from a similar day/night movement pattern to a nocturnal movement pattern with decreased activity during the day. Reduce activity during the day, whilst maintaining normal activity at night would not reduce the probability of encountering a nocturnal predator under natural conditions and therefore appears to not be a meaningful anti-predator response. To assess the relevance of these findings under natural conditions, we compared the experimental results with drift measurements from field observations. These show a significant reduction in drift activity for G. pulex and slight tendencies for reduced night-time drift for B. rhodani, under seasonal variations. We conclude that the behaviour in response to the physical contact or the hydrodynamic stimuli of nocturnal predators is the most likely explanation for the differences between the results from our laboratory experiment and the field observation. We further discuss that the observed migration patterns might have different species specific consequences for density stabilisation on a population level.

Omnivory is prevalent in terrestrial and aquatic food webs. However, the extent and seasonality of predatory feeding by omnivores in stream food webs is largely unknown. To understand better these aspects of omnivory in stream food webs, we investigated seasonal changes in the trophic positions of 2 omnivores in a small forested stream. We selected the amphipod, Gammarus pulex, and the caddisfly larvae, Hydropsyche spp., as key organisms because both taxa are common, reach high biomasses in many stream ecosystems, and have broad food spectra. We used stable-isotope analysis of the most prevalent taxa in the benthic macroinvertebrate community to assess the trophic positions of the 2 omnivorous taxa in different seasons. We estimated the degree of predatory feeding by quantifying the importance of different food resources with a stable-isotope mixing model (IsoSource). The predation capacity of omnivores, defined as the fraction of omnivore biomass associated with predation, was compared to the biomass of strictly predatory invertebrates. Our analysis indicated a predation capacity of Hydropsyche spp. similar to that of strict invertebrate predators, whereas the predatory biomass of G. pulex was even higher than those of the other invertebrate predators because of a combination of high biomass and large proportions of animal prey (50-90%) in the diets of both omnivores. The trophic position of omnivores in winter was comparable to that of invertebrate predators (trophic level 3), whereas in summer, their trophic position was comparable to primary consumers (level 2). These shifts might be caused by seasonally varying prey availability for Hydropsyche spp. and life-cycle patterns for G. pulex. Predation by omnivores was generally high in our study stream and probably has been underestimated in other stream ecosystems where omnivores, such as Hydropsyche spp. and G. pulex, are common.

Summary1. The spatial heterogeneity of ecosystems as well as temporal activity patterns of organisms can have far‐reaching effects on predator–prey relationships. We hypothesised that spatiotemporal constraints in mesohabitat use by benthic fish predators would reduce habitat overlap with benthic invertebrates and lead to mesohabitat‐specific predation risks.2. We analysed the spatiotemporal activity patterns of two small‐bodied benthivorous fishes, gudgeon (Gobio gobio) and stone loach (Barbatula barbatula), and of benthic invertebrates in a small temperate stream during three 24‐h field experiments. By applying a novel method of field video observation, we monitored the spatiotemporal foraging behaviour of the fish in their natural environment. A parallel analysis of invertebrate mesohabitat use by means of small area Hess sampling allowed a direct estimation of habitat overlap at a pool–riffle scale.3. Gudgeon showed a dominant spatial activity pattern preferring pools at all times of day, whereas stone loach used both mesohabitats but with a distinct temporal (nocturnal) activity pattern. The patterns of residence were not identical with those of active foraging. Invertebrate community composition differed significantly between mesohabitats but not between times of day. More than half of the total dissimilarity between pools and riffles was accounted for by six invertebrate taxa. Five of these were subject to higher fish predation in pools than in riffles. The total prey consumption of the two fish species together in pools was about three times as high as in riffles. Trophic niche breadth of stone loach and thus its predation range was broader than that of gudgeon.4. These results indicate that the potential predation risk for stream invertebrates depends on the combination of spatial and temporal patterns of both predator and prey. Given the distinct differences in predation risk found between pools and riffles, we conclude that spatial heterogeneity at the mesohabitat scale can influence mechanisms and consequences of selective predation. We also suggest that the analysis of spatiotemporal predator–prey relationships should not be based on the premise that the main residence habitat and active foraging habitat of a predator are identical.

Freshwater ecosystems are polluted with various environmental chemicals. For example, pesticides enter the aquatic environment via spray drift or surface runoff from catchment areas used for industrialised agriculture. In the present study, we investigated the response of the grazeraufwuchs interaction when exposed to the herbicide terbutryn in a microcosm experiment. Terbutryn induced a trophic cascade with negative effects on grazers by inhibiting growth of pri- mary producers. We determined a no-observed-effect concentration (measured as particulate organic carbon, NOECPOC) of 0.21 µg l �1 for aufwuchs biomass and a lowest-observed-effect con- centration (LOECPOC) of 2.01 µg l �1 . Furthermore, a shift in the aufwuchs community was detected in all terbutryn treatments nearly 4 wk after exposure, and this shift persisted until the end of the experiment in the treatment with the highest terbutryn concentration. In addition, reduced energy stores (triglycerides, TG) of the grazer Rhithrogena semicolorata (Heptageniidae: Ephemeroptera) were found (NOECTG = 0.03 µg l �1 , LOECTG = 0.21 µg l �1 ), which was interpreted as an indication of moderate starvation during the experiment. While starvation did not result in reduced larval growth as might have been expected, the ob served reduction of TG content in larvae (24%) is consistent with results from a separate experiment in which starvation alone induced a 15% reduction of TG content as well as reduced emergence, reduced size at emergence, and reduced egg production.

P>1. We conducted an experimental study of predation by benthivorous fish on a natural community of stream invertebrates using a reach-scale approach. Over a 2-year period (experimental phase), the benthic invertebrate community of a stretch containing two species of benthivorous fish was compared with a fishless stretch. Thereafter, all fish were removed and benthic community structure was analysed again to account for natural differences between the two stretches (reference phase).2. Benthivorous fish at the moderate densities investigated did not affect total benthic biomass or density, but did alter species composition. In addition, the fish effect differed between pool and riffle habitats, with larger effects in the pools indicating a habitat-specific predation effect. In the reference phase, when all fish were removed from the stream, the difference between the two stretches was reduced.3. The benthivorous fish reduced the densities of four taxa (Pisidium sp., Dugesia gonocephala, Gammarus pulex, Limoniidae), representing 29% of total biomass. It is possible that density reductions of other species were masked by prey migration despite the relatively large spatial scale. Indeed, higher drift activity in the upstream fishless stretch could have increased the density of Baetis rhodani in the fish stretch, as indicated by the results of a drift model.4. Our results provide insights into stream food web ecology because fish predation showed effects even in a natural system where habitat complexity was high, environmental factors were highly variable and many predator and prey species interacted and because benthivorous fish were the focus, whereas the majority of previous predation experiments in streams have used drift-feeding trout.