Concordance among stream assemblages and spatial autocorrelation along a fragmented gradient

Patterns of spatial autocorrelation of biota and distributional similarity (concordance) between assemblages of different organism groups have important implications in both theoretical ecology and biodiversity conservation. Here we report environmental gradients and spatial distribution patterns of taxonomic composition among stream fish, benthic macroinvertebrate, and diatom assemblages along a fragmented stream in south‐western France. We quantified spatial patterns of lotic assemblage structure along this stream, and we tested for concordance in distribution patterns among the three taxonomic groups. Our results showed that both environmental characteristics and stream assemblages were spatially autocorrelated. For stream fish and diatom assemblages, these patterns reflected assemblage changes along the longitudinal stream gradient, whereas environmental variables and benthic macroinvertebrates exhibited a more patchy spatial pattern. Cross‐taxa concordance was significant between stream fish and diatoms, and between stream fish and benthic macroinvertebrates. The assemblage concordance between stream fish and diatoms could be attributed to similar responses along the longitudinal gradient, whereas those between stream fish and benthic macroinvertebrates may result from biotic interactions. Based on potential dispersal capacities of taxa, our results validated the hypotheses that weakly dispersing taxa exhibit greater concordance than highly dispersing ones and that dispersal capacities affect how taxonomic groups respond to their local environment. Both diatoms and highly dispersing stream fish were affected by stream fragmentation (i.e. the number of dams between sites), while the effect of fragmentation was not significant for invertebrates that fly well in their adult stage, thus emphasizing the importance of the way of dispersal. These results suggest that addressing the effects of dispersal capacity on stream assemblage patterns is crucial to identifying mechanisms behind patterns and to better understanding the determinants of stream biodiversity.     

Biological quality metrics: their variability and appropriate scale for assessing streams

The concept of spatial scale is at the research frontier in ecology, and although focus has been placed on trying to determine the role of spatial scale in structuring communities, there still is a further need to standardize which organism groups are to be used at which scale and under which circumstances in environmental assessment. This paper contributes to the understanding of the variability at different spatial scales (reach, stream, river basin) of metrics characterizing communities of different biological quality elements (macrophytes, fishes, macroinvertebrates and benthic diatoms) as defined by the Water Framework Directive. For this purpose, high-quality reaches from medium-sized lowland streams of Latvia, Ecoregion 15 (Baltic) were sampled using a nested hierarchical sampling design: (river basin → stream → reach). The variability of metrics within the different groups of biological quality elements confirmed that large-bodied organisms (macrophytes and fish) were less variable than small-bodied organisms (macroinvertebrates and benthic diatoms) at reach, stream and river basin scales. Single metrics of biological quality elements had the largest variation at the reach scale compared with stream and basin scales. There were no significant correlations between biodiversity indices of the different organism groups. The correlation between diversity indices (Shannon’s and Simpson’s) of the biological quality elements (macrophytes, fish, benthic macroinvertebrates and benthic diatoms) and a number of measured environmental variables varied among the different organism groups. Relationships between diversity indices and environmental factors were established for all groups of biological quality elements. Our results showed that metrics of macrophytes and fish could be used for assessing ecological quality at the river basin scale, whereas metrics of macroinvertebrates and benthic diatoms were most appropriate at a smaller scale.     

Are biological classifications of headwater streams concordant across multiple taxonomic groups?

1. Studies assessing human impacts on freshwater ecosystems are typically based on a single taxonomic group, often macroinvertebrates or fish. Unfortunately, the degree to which such macroinvertebrate or fish‐based surveys can be generalised across other taxonomic groups remains largely unknown. A prerequisite for useful generalisations is that different taxonomic groups exhibit concordant patterns of community structure across sites. 2. We examined the concordance among fish, benthic macroinvertebrates and bryophytes in 32 streams in a boreal catchment in Finland. Our goal was to test how consistently different taxonomic groups classify stream sites; for example, can site groupings based on macroinvertebrates be used as a surrogate for bryophyte or fish assemblage classification? 3. Our results show that community classifications in headwater streams are not concordant across taxonomic groups, at least not at the within‐river system scale. The lack of concordance reflected the fact that all three groups responded to different environmental factors. Macroinvertebrate community structure was mainly correlated with stream size and pH, whereas bryophytes were related to water colour, nutrient content and in‐stream habitat variability. Fish community structure was best described by stream depth, substrate size and water oxygen concentration. 4. Our results suggest that great care should be taken when typologies based on benthic macroinvertebrates, or any other taxonomic group, are extrapolated to other groups in creating typologies of lotic environments.     

Multi-scale effects of resource patchiness on foraging behaviour and habitat use by longnose dace, Rhinichthys cataractae

1. We examined the response of a predatory benthic fish, the longnose dace ( Rhinichthys cataractae ), to patchiness in the distribution of benthic macroinvertebrates on cobbles at three hierarchical spatial scales during summer and autumn 1996, and spring 1997 in a southern Appalachian stream. 2. At the primary scale (four to five individual cobbles separated by <1 m), the intensity of foraging was not correlated with the biomass of benthic macroinvertebrates/cobble, regardless of season. 3. At the secondary scale (i.e. foraging patches <5 m in diameter) we found that benthic macroinvertebrates were patchily distributed in summer, but not in autumn or spring. Concomitantly, in summer, longnose dace foraged on cobbles with a significantly higher biomass of benthic macronvertebrates than nearby, randomly selected cobbles with similar physical conditions (i.e. longnose dace tended to avoid low-prey foraging patches). In contrast, when benthic macroinvertebrates were distributed homogeneously (spring and autumn), dace did not select patches with a significantly higher biomass of benthic macroinvertebrates than that available on randomly selected cobbles. 4. At the tertiary scale (i.e. stream reaches 11–19 m long), the biomass of benthic macroinvertebrates (per cobble per reach) was patchily distributed (i.e. differed significantly among reaches) in all seasons. Among reaches with physical characteristics preferred by longnose dace, (i.e. erosional reaches dominated by cobble/boulder substratum and high current velocity), we detected a significant, positive correlation between the biomass of benthic macroinvertebrates/cobble and longnose dace density in all seasons. 5. Our results demonstrated that both spatial and temporal patchiness in resource availability influenced significantly the use of both foraging patches and stream reaches by longnose dace.     

Effects of stream predator richness on the prey community and ecosystem attributes

It is important to understand the role that different predators can have to be able to predict how changes in the predator assemblage may affect the prey community and ecosystem attributes. We tested the effects of different stream predators on macroinvertebrates and ecosystem attributes, in terms of benthic algal biomass and accumulation of detritus, in artificial stream channels. Predator richness was manipulated from zero to three predators, using two fish and one crayfish species, while density was kept equal (n = 6) in all treatments with predators. Predators differed in their foraging strategies (benthic vs. drift feeding fish and omnivorous crayfish) but had overlapping food preferences. We found effects of both predator species richness and identity, but the direction of effects differed depending on the response variable. While there was no effect on macroinvertebrate biomass, diversity of predatory macroinvertebrates decreased with increasing predator species richness, which suggests complementarity between predators for this functional feeding group. Moreover, the accumulation of detritus was affected by both predator species richness and predator identity. Increasing predator species richness decreased detritus accumulation and presence of the benthic fish resulted in the lowest amounts of detritus. Predator identity (the benthic fish), but not predator species richness had a positive effect on benthic algal biomass. Furthermore, the results indicate indirect negative effects between the two ecosystem attributes, with a negative correlation between the amount of detritus and algal biomass. Hence, interactions between different predators directly affected stream community structure, while predator identity had the strongest impact on ecosystem attributes.     

Two Newly Recorded Species of tribe Eucosmini (Lepidoptera,Tortricidae) from North Korea

Stream restoration affects stream biodiversity by improving water quality and habitat environments. Benthic macroinvertebrates are used as indicators of changes in stream environment, and, therefore, the effects of restoration can be evaluated by a long‐term monitoring of benthic macroinvertebrates. Cheonggye stream is an urban stream in Seoul, Korea, and a restoration project was conducted from July 2003 to September 2005. The purpose of this study was to investigate the long‐term changes in the benthic macroinvertebrate communities in Cheonggye stream after the restoration project in 2005. A 6‐year field study was seasonally conducted at five study sites in Cheonggye stream from November 2005 to October 2011. Benthic macroinvertebrates were sampled quantitatively using a Surber sampler (50 × 50 cm; mesh 0.25 mm, two replications per site). As a result, species richness of benthic macroinvertebrates rapidly increased after the restoration project and peaked (35 species per sample) in 2006 but gradually dropped to approximately 20 species up to the recent year. As a consequence, community indices changed gradually according to species richness and abundance: Dominance indices peaked in 2010 and species diversity indices (H') peaked in 2007. According to a functional feeding group analysis, the composition of collector–filterers increased at first, but the decreased gradually to the recent year. In contrast, collector–gatherers showed an opposite tendency. The composition of clingers increased during the initial 4 years (2005–2008), whereas burrowers rapidly increased after 2008.     

Patch based predation in a southern Appalachian stream

Streams are characterized by high degrees of patchiness that could influence the role of predators in these systems. Here we assess the impact of predatory benthic fishes on benthic macroinvertebrate density, biomass, and community structure at the patch scale in a fourth order stream in the southern Appalachians. We tested the role of predation in two different patch types: patches inhabited by adult mottled sculpin ( Cottus bairdi ) and random patches. We placed 30 basket pairs (one open to fish predation, and one from which fish predators were excluded) in the streambed at each patch type. We also tested for potential basket effects by setting up a basket control area. Although there was some evidence of basket artifacts on macroinvertebrate density in sculpin patches, these artifacts were not consistent and we do not feel that they affected our results because predators did not affect macroinvertebrate density. In random patches, predation did not significantly affect macroinvertebrate density or biomass. Predators significantly reduced macroinvertebrate biomass in sculpin patches but did not affect prey density. When the data-set was size-limited to exclude macroinvertebrates too large for consumption by sculpin, macroinvertebrate biomass did not differ significantly between exclusion and open baskets. This suggests that sculpin can reduce macroinvertebrate biomass through a combination of consumption and by predator-induced emigration of large macroinvertebrates into areas that are protected from sculpin. In addition, invertebrate predator biomass was higher in predator exclusion baskets in sculpin patches indicating that predation pressure remained high in the exclusion baskets despite fish exclusion. These results illustrate the heterogeneity of streams and the effect of small-scale differences (e.g. location of predators' territories) on local processes. Experiments that utilize these differences can provide insights into these stream processes.     

The response of stream fish to local and reach‐scale variation in the occurrence of a benthic aquatic macrophyte

1. The aquatic macrophyte Podostemum ceratophyllum has been shown to increase stream productivity, abundance and biomass of benthic invertebrates, and local occurrences of some stream fishes. However, experimental evidence that fishes preferentially associate with Podostemum is lacking, and the value of Podostemum as a predictor of stream fish assemblage composition has not been studied. 2. We conducted two short‐term (2 week), small‐scale (36 m²) experimental manipulations of Podostemum cover in the Conasauga River (Georgia and Tennessee, U.S.), and found higher abundances of benthic insectivorous fishes in patches with augmented (>80%) compared to reduced (7%) Podostemum cover. In an observational study, we quantified associations among percent cover of Podostemum, fish species richness, land cover, shoal length and base‐flow turbidity at 20 randomly selected shoals from a 39‐km reach that spanned a gradient of decreasing forest land cover. 3. Richness of all fish species and of lotic fishes peaked in the centre of the study reach, and richness was weakly correlated with predictor variables. Occupancy models for individual species also indicated that longitudinal position was a strong covariate for 13 of 19 species examined, with little support that Podostemum cover influenced occupancy. 4. Local associations may reflect choices by benthic fishes to utilise Podostemum, whereas downstream decline in fish species richness and Podostemum cover may reflect altered capacity of the system to support native species.     

Implications of a Hierarchical Relationship Among Channel Form,Instream Habitat,and Stream Communities for Restoration of Channelized Streams

Hierarchy theory provides a conceptual framework for understanding the influence of differently scaled processes on the structure of stream communities. Channel form, instream habitat, and stream communities appear to be hierarchically related, but the strength of the relationships among all components of this hypothesized hierarchy have not been examined. We sampled channel form, instream habitat, fishes, and macroinvertebrates in a channelized stream in Mississippi and Alabama to examine the hypothesis that a hierarchical relationship exists among channel form, instream habitat, and stream communities. Instream habitat, fishes, and macroinvertebrates were sampled in May, July, and September 2000. Measurements of channel form were obtained in July 2000. Mantel tests, multiple regressions, and correlation analyses were used to assess strength of the relationships among channel form, instream habitat, and stream communities. Positive correlations were observed between channel form and instream habitat, and correlations observed between these factors were the greatest observed in our study. Overall, fish and macroinvertebrate communities exhibited stronger relationships with instream habitat than with channel form. Species richness, evenness, and abundance tended to exhibit greater correlations with instream habitat, while species composition had greater correlations with channel form. We concluded that channel form, instream habitat, and stream communities were hierarchically related.     

Separating physical disturbance and nutrient enrichment caused by Pacific salmon in stream ecosystems

1. Pacific salmon (Oncorhynchus spp.) deliver marine‐derived nutrients to the streams in which they spawn and die, and these resource subsidies can increase the abundance of stream biota. In strong contrast, physical disturbance from salmon spawning activity can reduce the abundance of benthic organisms. Previous experimental designs have not established the relative effects of these two contrasting processes on stream organisms during a salmon run. 2. We combined manipulative and observational field studies to assess the degree of nutrient enrichment, physical disturbance, and the net effect of salmon on the abundance of benthic periphyton. Related salmon‐mediated processes were also evaluated for benthic macroinvertebrates. Mesh exclosures (2 × 2 m plots) prevented salmon from disturbing areas of the stream channel, which were compared with areas to which salmon had access. Sampling was conducted both before and during the late‐summer spawning run of pink (O. gorbushca) and chum (O. keta) salmon. 3. Streamwater nitrogen and phosphorus concentrations increased sharply with the onset of the salmon run, and highly significant positive relationships were observed between the numbers of salmon present in the stream and these dissolved nutrients. Before the salmon run, periphyton biomass (as chlorophyll a) and total macroinvertebrate abundance were very similar between control and exclosure plots. During the salmon run, exclosures departed substantially from controls, suggesting significant disturbance imparted on benthic biota. 4. Comparing exclosures before and during the salmon run enabled us to estimate the effects of salmon in the absence of direct salmon disturbance. This ‘nutrient enrichment potential’ was significant for periphyton biomass, as was a related index for macroinvertebrate abundance (although enhanced invertebrate drift into exclosures during the salmon run could also have been important). Interestingly, however, the net effect of salmon, evaluated by comparing control plots before and during the salmon run, was relatively modest for both periphyton and macroinvertebrates, suggesting that nutrient enrichment effects were largely offset by disturbance. 5. Our results illustrate the importance of isolating the specific mechanisms via which organisms affect ecosystems, and indicate that the relative magnitude of salmon nutrient enrichment and benthic disturbance determines the net effect that these ecologically important fish have on stream ecosystems.     

Environmental Variability, Historical Contingency, and the Structure of Regional Fish and Macroinvertebrate Faunas in Ouachita Mountain Stream Systems

In 1990–1992, the United States Forest Service sampled six hydrologically variable streams paired in three different drainage basins in the Ouachita Mountains, Arkansas, U.S.A. Fishes, macroinvertebrates, and stream environmental variables were quantified for each stream. We used these data to examine the relationship between regional faunas (based on taxonomy and trophic affiliation of fishes and macroinvertebrates) and measured environmental variables. Because fishes are constrained to their historically defined drainage basins and many insect taxa are able to cross basin barriers, we anticipated that both groups would respond differently to environmental variability. Fishes were influenced more by environmental variability that was unique to their historical drainage basins, but macroinvertebrates were associated more strongly with environmental variability that was independent of drainage basins. Thus, the individual drainage basins represented a historical constraint on regional patterns of fish assembly. For both fishes and macroinvertebrates, groupings based on taxonomy and trophic affiliation showed a similar response to environmental variability and there was a high degree of association between taxonomic and trophic correlation matrices. Thus, trophic group structure was highly dependent on the taxonomic make-up of a given assemblage. At the basin-level, fish and macroinvertebrate taxa were associated more strongly with environmental variability than the trophic groups, and these results have implications for basin-level studies that use trophic groupings as a metric to assess ecological patterns. Trophic categories may not be a useful ecological measure for studies at large spatial scales.     

Diversity assessment of benthic macroinvertebrate communities in Banco National Park (Banco Stream,Côte d'Ivoire)

In the present study, a first inventory of benthic macroinvertebrates in the Banco Stream, Côte d'Ivoire, and the correlations between environmental variables and taxonomic richness were analysed. Seven stations were sampled monthly over a 1‐year period, using a hand net (10 × 10 cm, 250 μm mesh, 50 cm length). One hundred and thirty‐two macroinvertebrate taxa were recorded. These taxa were distributed among 74 families and 15 orders belonging to Insecta (118 taxa; 89% of total richness), Oligochaeta (seven taxa), Crustacea (five taxa) and Mollusca (two taxa). Kruskal–Wallis test revealed significant difference (at least P < 0.05) in macroinvertebrate richness between upstream stations (S1 and S2) and stations S4, S5 and S6. Chironominae and Tanypodinae (Insecta) were the two very frequent taxa in all the stations. Lumbriculidae (Oligochaeta), Desmocaris trispinosa (Crustacea) and Eurymetra sp. (Insecta) were frequently found in the samples. Hierarchical cluster analysis revealed three groups of sampling stations according to taxonomic similarity. Taxonomic richness was significantly and negatively correlated with conductivity, while it was significantly and positively correlated with substrate types (woody debris and gravel). Due to the fact that Banco stream is the locality type of an endemic shrimp species (Macrobrachium thysi), this basin is of high conservation priority.     

Taxonomic and regional patterns in benthic macroinvertebrate elemental composition in streams

1. Ecological stoichiometry has been used to better understand dynamics in consumer growth and the role of consumer‐recycled nutrients because it focuses on more than one element. Most research has focused on pelagic rather than benthic consumers. Variation in elemental composition among benthic consumer taxa would suggest that taxa differ in their susceptibility to nutrient limitation or in their role in recycling nutrients. 2. We collected benthic macroinvertebrates from streams in two regions (Indiana–Michigan and Wisconsin, U.S.A.) to examine taxonomic and regional variation in benthic macroinvertebrate body carbon (C), nitrogen (N), and phosphorus (P) concentrations and ratios. 3. Elemental composition varied little within taxa common to both regions. In contrast, elemental composition differed greatly among taxa and appeared to be related to phylogeny. The elemental composition of macroinvertebrates clustered into three distinct groups: insects, mollusks, and crustaceans. To a lesser extent, insects and mollusks also differed in elemental composition among genera. 4. Functional feeding groups (FFGs) differed in elemental composition, with predators having a higher N content than other groups. Substantial elemental imbalances between C and N were found between most primary consumers and their likely food sources, and the magnitude of the imbalance depended in part on the FFG. 5. Our results support an assumption of most ecological stoichiometry models that, within a species, the elemental composition of aquatic invertebrates is relatively constant. Variation in elemental composition among taxa at various higher taxonomic levels suggests that susceptibility of stream invertebrates to nutrient limitation and their role in nutrient cycling will strongly depend on phylogeny.     

Food sources and trophic structure of fishes and benthic macroinvertebrates in a tropical seagrass meadow revealed by stable isotope analysis

Stable carbon and nitrogen isotope analysis was used to examine the food sources and trophic structure of 17 fish species and six groups of benthic macroinvertebrates in a seagrass meadow in North Sulawesi, Indonesia. The seagrass, their associated epiphytes, sediment organic matter (SOM) and particulate organic matter (POM) were identified to be the food sources, with δ¹³C values ranging from ?19.49 (POM) to ?9.66‰ (seagrass). The δ¹³C of the 23 fauna taxa were between ?18.57 (Arothron manilensis) and ?11.62‰ (Protoreaster sp.). For five of the six groups of benthic macroinvertebrates, seagrass and their epiphytes contributed more than 69.4%. For 14 of the 17 fish species, seagrass and their epiphytes are the main contributors. For 15 of the 17 fishes, the trophic levels inferred from SIA are lower than those from the previously reported diet composition analysis. These findings show that seagrass and their epiphytes are consumed by most of the fish and benthic macroinvertebrates, and are important for a large portion of the food web in seagrass meadows in the Coral Triangle area.     

Restoration of rapids habitat in a Great Lakes connecting channel,the St. Marys River,Michigan

Aquatic habitat has been extensively altered throughout the Laurentian Great Lakes to increase navigation connectivity. In particular, the St. Marys River, a Great Lakes connecting channel, lost >50% of its historic rapids habitat over the past century. In 2016, the natural flow was restored to the Little Rapids area of the St. Marys River. The goal of our study was to evaluate physical and ecological responses to the restoration of the Little Rapids area. Extensive habitat and biological data were collected prior to restoration (2013 and 2014), and after restoration (2017 and 2018). Measured parameters included total suspended solids, current velocity, benthic macroinvertebrates, and larval, juvenile, and adult fishes. Total suspended solids stayed low (<4 mg/L) following restoration, with the exception of a single construction‐related event. Pre‐restoration data indicated that all measured velocities were below the target flow rate of 0.24 m/s, whereas 70% of the measured habitat was above the target flow post‐restoration. Abundance and richness of benthic macroinvertebrates were reduced following restoration (>90% reduction). We observed a 45% increase in richness of larval fish 2 years after restoration and a 131% increase in catch per unit effort. For adult fishes, the proportion of individuals with a preference for fast‐moving waters increased from 1.5 to 45% in the restored area, and from 7 to 15% upstream of the restored area; a similar response was observed for lithophilic spawners. The physical and biological conditions of the Little Rapids improved and resembled conditions typical of rapids habitat extent in other areas of the river and other systems.     

Indirect effects of the algivorous fish Plecoglossus altivelis altivelis on the growth of two insectivorous benthic fish

1. Animals exploiting different resources may nevertheless interact if one species indirectly alters the abundance and distribution of the food of the other. To analyse this indirect effect, we conducted experiments in artificial pools and in the field to investigate the influence of the algivorous fish Plecoglossus altivelis altivelis (known as the ayu) on two species of insectivorous benthic fish, Pseudogobio esocinus esocinus and the goby Gymnogobius petschiliensis .
2. In the pool experiments, algal biomass was not correlated with the number of ayu, but the percentage of blue-green bacteria rose as the number increased. The number of aquatic macroinvertebrates on the upper surface of ceramic tiles placed in the pool bed decreased as the number of ayu increased.
3. Although ayu and the benthic species did not interact directly, the reduction in invertebrate abundance on the upper surface of tiles in the pool reduced the growth rate of the benthic insectivores.
4. In field experiments, the introduction of ayu into habitats with P. esocinus esocinus or G. petschiliensis reduced the growth rate of these benthic fish. In the field experiment that was carried out over 5 years in the G. petschiliensis habitat, the population density of the goby decreased when ayu were stocked.
5. The ayu is a strong interactor or bioengineer in streams, affecting not only benthic algae but also aquatic invertebrates and fishes. We conclude that to predict the outcome of interspecific interactions amongst fishes in streams with high algal production, possible indirect effects must be considered alongside better known direct effects.     

Incorporating invertebrate predators into theory regarding the timing of invertebrate drift

Theory concerning the timing of lotic invertebrate drift suggests that daytime-feeding fish cause invertebrates to restrict their drift behavior to the nighttime. However, there is growing evidence that the nighttime foraging of invertebrate predators also contributes to the nocturnal timing of drift, though it is unclear whether the nocturnal behavior of invertebrate predators is innate or proximately caused by fish. In two experiments, one conducted in a fish-bearing stream and a second in a fishless stream, we compared the drift patterns of Baetidae (Ephemeroptera) from channels with and without benthic invertebrate predators. We tested whether invertebrate predators affect the timing of drift, either as a proximate cause of nocturnal drift in the fishless stream (diel periodicity) or as a proximate cause of a pre-dawn peak in drift in the fish-bearing stream (nocturnal periodicity). In the fish-bearing stream experiment, a pre-dawn increase of baetid drift occurred independently of invertebrate predators, indicating that invertebrate predators were not the proximate cause of nocturnal periodicity in the stream. In the fishless stream experiment, invertebrate predators caused more baetid drift at night than during the day, indicating that invertebrate predators caused the nocturnal drift pattern we observed in the stream, and that invertebrate predators can influence drift timing independently of fish. Therefore, we suggest that both visually feeding fish and nocturnally foraging benthic predators, when present, affect the timing of invertebrate drift; visually feeding fish by reducing daytime drift, and benthic predators by increasing nighttime drift.     

Environmental and spatial controls of taxonomic versus trait composition of stream biota

相似文献   

Changes in habitat structure, benthic invertebrate diversity, trout populations and ecosystem processes in restored forest streams: a boreal perspective

1. Most Finnish streams were channelised during the 19th and 20th century to facilitate timber floating. By the late 1970s, extensive programmes were initiated to restore these degraded streams. The responses of fish populations to restoration have been little studied, however, and monitoring of other stream biota has been negligible. In this paper, we review results from a set of studies on the effects of stream restoration on habitat structure, brown trout populations, benthic macroinvertebrates and leaf retention. 2. In general, restoration greatly increased stream bed heterogeneity. The cover of mosses in channelised streams was close to that of unmodified reference sites, but after restoration moss cover declined to one‐tenth of the pre‐restoration value. 3. In one stream, densities of age‐0 trout were slightly lower after restoration, but the difference to an unmodified reference stream was non‐significant, indicating no effect of restoration. In another stream, trout density increased after restoration, indicating a weakly positive response. The overall weak response of trout to habitat manipulations probably relates to the fact that restoration did not increase the amount of pools, a key winter habitat for salmonids. 4. Benthic invertebrate community composition was more variable in streams restored 4–6 years before sampling than in unmodified reference streams or streams restored 8 years before sampling. Channelised streams supported a distinctive set of indicator species, most of which were filter‐feeders or scrapers, while most of the indicators in streams restored 8 years before sampling were shredders. 5. Leaf retentiveness in reference streams was high, with 60–70% of experimentally released leaves being retained within 50 m. Channelised streams were poorly retentive (c. 10% of leaves retained), and the increase in retention following restoration was modest (+14% on average). Aquatic mosses were a key retentive feature in both channelised and natural streams, but their cover was drastically reduced through restoration. 6. Mitigation of the detrimental impacts of forestry (e.g. removal of mature riparian forests) is a major challenge to the management of boreal streams. This goal cannot be achieved by focusing efforts only on restoration of physical structures in stream channels, but also requires conservation and ecologically sound management of riparian forests.