Carbon and nitrogen dynamics in a maritime Antarctic stream

• 1 The carbon and nitrogen dynamics in a maritime Antarctic lake outflow stream were investigated. The stream and the algal communities could be split into two zones: a semi-aquatic margin consisting of a perennial cyanobacteria/diatom mat and a flowing channel with a similar perennial mat that was overgrown by annual filamentous chlorophytes during the course of the summer.
• 2 Neither algal community was limited by nutrient availability. Major nutrients were always available in the stream water. There were slight differences in the atomic ratios of the mats, the N:P ratios in the channel mat being lower than those in the marginal mat. However, both these and the total dissolved N:P ratio in the stream water were all close to those that indicate a balanced supply.
• 3 There was no net carbon or nitrogen accumulation by the marginal mat suggesting that uptake processes were balanced by loss processes.
• 4 Maximum rates of carbon fixation (0.1–0.5mgCg^?1 dry weight h^?1) were similar to those of other perennial Antarctic algal mats. Productivity appeared to be limited by physical factors, but the effects of irradiance and temperature could not be separated.
• 5 There were no heterocystous cyanobacteria in the mat communities and rates of atmospheric nitrogen fixation were very low (0–10ngNmg^?1 mat Nh^?1). Fixation accounted for only 0.3% of the nitrogen accumulation of the channel mats, but was higher in the marginal mat where uptake of other sources of nitrogen was also low.
• 6 Nitrogen accumulation by the channel mat averaged 0.34gNm^?2 day^?1. Only 0.05gNm^?2 day^?1 was accounted for by the uptake of dissolved inorganic nitrogen (nitrate plus ammonium). The major (80%) source of nitrogen appeared to be dissolved organic nitrogen. Recycling of nitrogen within the stream ecosystem may also be important.

     

Metabolism of a desert stream

SUMMARY. Rates of photosynthesis and community respiration were determined for benthic assemblages in Sycamore Creek, a Sonoran Desert stream in Arizona. Benthos in this stream can be separated into (1) mats of Cladophora glomerata and associated epiphytes and (2) assemblages of epipelic diatoms and blue-green algae. Community respiration and net photosynthesis were measured for these assemblages using submerged light-dark chambers in situ . Multiple regression analysis was used to predict (1) gross photosynthesis as a function of photosynthetically active radiation, temperature and chlorophyll-α concentration; and (2) community respiration as a function of temperature and biomass.
Calculations suggest that Sycamore Creek is autotrophic during the summer ( P/R = 1.7) and that the rates of gross photosynthesis ( P =8.5 g O₂ m⁻² day⁻¹) and community respiration ( R = 5.1 g O₂ m⁻² day⁻¹) are high for a small stream. Considerable difference exists between the Cladophora mat assemblages, in which mean P is 12.5gO₂m⁻² day⁻¹and the P/R ratio is 2.3, and the epipelic assemblages in which mean P is 4.4 g O₂m⁻² day⁻¹ and P/R is 0.96. The high rate of gross photosynthesis, low litter inputs, high biomass of algae and the intermittent but severe floods that characterize Sycamore Creek indicate that this stream and other similar desert streams are net exporters of organic matter and are, thereby, truly autotrophic stream ecosystems.     

新疆塔里木河干流荒漠河岸林植被带变化

基于近40年5个不同时期的Landsat遥感数据和样地调查资料,对塔里木河干流以荒漠河岸林为主体的植被带的时空变化进行了反演,结果表明:(1)塔里木河干流平均植被覆盖度为15%,空间差异明显,总体上呈上游高(20%)、下游低(9%),北岸高(17%)、南岸低(12%)的特征。植被带宽度也有同样的分布特点;(2)塔里木河干流植被覆盖度总体呈上升趋势,平均上升速率为0.0038·10年-1,但不同河段差异明显,上游增加速度最快,平均增加速率为0.0083·10年-1,中游增加速度为0.0013·10年-1,而下游呈现下降趋势,下降速率为-0.0003·10年-1;(3)从不同植被覆盖等级来看,塔里木河干流以中低植被覆盖度所反映的荒漠河岸林呈减少和退化趋势,以高植被覆盖度所反映的耕地和人工植被面积呈增加趋势;(4)塔里木河干流上游植被带宽度和覆盖度的增加是由研究区内人工绿洲扩大所致,是以牺牲下游生态环境为代价的。     

Influence of base flow stream bank seepage on riparian zone nitrogen biogeochemistry

We examined the effect of sustained stream bank seepage during base flow conditions on the pore water nitrogen biogeochemistry of two riparian zones in lowland agricultural areas in southern Ontario, Canada. Nitrate, ammonium and dissolved oxygen concentrations in riparian subsurface water over a two-year period showed well-organized spatial patterns along stream bank seepage flow paths that extended seasonally up to 25 m inland. High levels of dissolved oxygen and NO₃⁻ in stream inflow were depleted rapidly at the stream bank interface suggesting the occurrence of aerobic microbial respiration followed by denitrification. A zone of NH₄⁺ accumulation persisted in more anaerobic sediments inland from the bank margin, although the magnitude and intensity of the pattern varied seasonally. A bromide tracer and NO₃⁻ co-injection at the stream bank interface indicated that bank seepage occurred along preferential flow paths in a poorly sorted gravel layer in the two riparian zones. Depletion of NO₃⁻ in relation to co-injected bromide confirmed that the bank margin was a hot spot of biogeochemical activity within the riparian zone. Conceptual models of humid temperate riparian zones have focused on nitrogen biogeochemistry in relation to hillslope to stream hydrologic flow paths. However, our results suggest that sustained stream bank inflow during low flow conditions can exert a dominant control on riparian nitrogen cycling in lowland landscapes where level riparian zones bounded by perennial streams receive limited subsurface inflows from adjacent slopes.     

The effects of riparian restoration following saltcedar (Tamarix spp.) biocontrol on habitat and herpetofauna along a desert stream

Amphibians and reptiles (herpetofauna) have been linked to specific microhabitat characteristics, microclimates, and water resources in riparian forests. Our objective was to relate variation in herpetofauna abundance to changes in habitat caused by a beetle used for Tamarix biocontrol (Diorhabda carinulata; Coleoptera: Chrysomelidae) and riparian restoration. During 2013 and 2014, we measured vegetation and monitored herpetofauna via trapping and visual encounter surveys (VES) at locations affected by biocontrol along the Virgin River in the Mojave Desert of the southwestern United States. Twenty‐one sites were divided into four riparian stand types based on density and percent cover of dominant trees (Tamarix, Prosopis, Populus, and Salix) and presence or absence of restoration. Restoration activities consisted of mechanically removing non‐native trees, transplanting native trees, and restoring hydrologic flows. Restored sites had three times more total lizard and eight times more yellow‐backed spiny lizard (Sceloporus uniformis) captures than other stand types. Woodhouse's toad (Anaxyrus woodhousii) captures were greatest in unrestored and restored Tam‐Pop/Sal sites. Results from VES indicated that herpetofauna abundance was greatest in the restored Tam‐Pop/Sal site compared with the adjacent unrestored Tam‐Pop/Sal site. Tam sites were characterized by having high Tamarix cover, percent canopy cover, and shade. Restored Tam‐Pop/Sal sites were most similar in habitat to Tam‐Pop/Sal sites. Two species of herpetofauna (spiny lizard and toad) were found to prefer habitat components characteristic of restored Tam‐Pop/Sal sites. Restored sites likely supported higher abundances of these species because restoration activities reduced canopy cover, increased native tree density, and restored surface water.     

Long-term nitrate removal in a stream riparian zone

The long-term capacity of riparian zones in regulating groundwater nitrate fluxes is not well understood. This study analyses patterns of nitrate removal for the period 1994–2012 at two sites in a river floodplain that have received high groundwater nitrate loading from a large upland aquifer for over 32 years. During the study, mean NO₃ ^?–N concentrations entering the riparian zone varied between 20–30 and 30–42 mg/L at the upstream and downstream sites respectively, but did not show any clear inter-annual trend. A permeable sand layer in the riparian zone is underlain by a regional aquitard at a depth of 5–6 m and 4 m at the upstream site and downstream site respectively. Denitrification resulted in a decline in nitrate concentrations as lateral groundwater flow in the sand layer interacted with buried peat and channel bar deposits that range up to 3 m in depth at both riparian sites. This interaction was greater at the downstream site where the organic deposits extend down to <1 m from the aquitard in some locations. At the upstream site nitrate removal efficiency in the sand layer, at depths of 3–4 m ~20 m from the river bank, declined from 68 % in 1996–1998 to 42 % in 2009–2012. A smaller decline from 92 to 82 % occurred in the sand layer 10 m from the river bank during the study. In contrast, no clear pattern of change was evident at the downstream site where a nitrate removal efficiency of 98–100 % occurred at the river bank in most years between 1994 and 2012. These data suggest that the long-term nitrate removal performance of some riparian zones may decline if carbon availability for denitrification becomes limited as a result of variations in the quantity, quality and location of subsurface organic deposits that interact with deeper groundwater flowpaths.     

Nitrogen fixation in a desert stream ecosystem

Few measurements of nitrogen fixation exist for streams. Desertstreams are warm, well lighted, and often supportabundant cyanobacterial populations; thus N₂ fixationmay be significant in these N-poor ecosystems. N₂fixation was measured in situ by acetylene reductionfor two patch types (Anabaena mat and anepilithic assemblage). Patch-specific rates were highcompared with published values (maximum 775 µgN₂ [83 µmol C₂H₄]mg chl a ^-1 h^-1or 51 mg N₂ [5.4 mmol C₂H₄] m_-2 h_-1).Daytime fixation was higher than nighttimefixation, and temperature, light and inorganic Nconcentration explained 52% of variance in hourlyrates over all dates. Diel input-output budgets wereconstructed on five dates when cyanobacteria werepresent in the stream. Diel N₂ fixation rates weremeasured for comparison with reach-scale diel nitrogenretention, to assess the importance of this vector to Neconomy of the stream. Fixation accounted for up to85% of net N flux to the benthos, but its importancevaried seasonally. Finally, we applied biomass-specificfixation rates to 1992 and 1993 biomass data to obtainseasonal and annual N₂ fixation estimates.Cyanobacteria were absent or rare during winter andspring, thus most of the annual N₂ fixation occurredduring summer and autumn. Annual rates of nitrogenfixation for 1992 and 1993 (8.0 g/m₂ and 12.5g/m₂) were very high compared to other streams,and moderately high compared to other ecosystems.Like other phenomena in this disturbance-proneecosystem, nitrogen fixation is strongly influenced bythe number and temporal distribution of flood events.     

Carbon and nitrogen limitations of soil microbial biomass in desert ecosystems

Microbial biomass nitrogen was measured in unamended (dry) and wetted soils in ten shrubland and grassland communities of the Chihuahuan desert, southern New Mexico, by the fumigation-extraction method. Microbial biomass-N in dry soils was undetectable. Average microbial biomass-N in wetted soils among all plant communities was 15.3 μg g^-1 soil. Highest values were found in the communities with the lowest topographic positions, and the minimum values were detected in the spaces between shrubs. Microbial biomass was positively and significantly correlated to soil organic carbon and extractable nitrogen (NH₄ ⁺ + NO₃ ^-). In a stepwise multiple regression, organic carbon and extractable nitrogen accounted for 40.9 and 5.6%, respectively, of the variance in microbial biomass-N among all the samples. Among communities, the soil microbial biomass was affected by the ratio of carbon to extractable nitrogen. Our results suggest a succession in the control of microbial biomass from nitrogen to carbon when the ratio of carbon to nitrogen decreases during desertification.     

The linkage between riparian predators and aquatic insects across a stream-resource spectrum

1. Spatial subsidies, defined as the flow of energy, nutrients, organisms or pollutants from one habitat to another, have been shown to affect the food–web dynamics in a wide range of ecosystems. An important subsidy to riparian communities is the contribution of adult stream insects to terrestrial predators such as birds, bats and lizards, but also invertebrates including ground and web‐building spiders. 2. We surveyed 37 first‐ and second‐order forest streams across differing environmental gradients in the Central South Island, New Zealand, to investigate the relationship between potential aquatic prey subsidies and predatory riparian arachnids. We anticipated that stream‐insect biomass would be positively associated with riparian arachnids, as a result of emergent adult aquatic insect subsidies to the adjacent habitat. 3. We confirmed positive associations between stream‐insect biomass as a predictor variable and riparian arachnid biomass (R² = 0.42, F_1,34 = 25.2, P < 0.001) and web densities (R² = 0.45, F_1,14 = 11.5, P < 0.01) respectively as dependent variables after adjusting for the confounding effects of environmental variables. Hierarchical partitioning confirmed the importance of stream insect biomass as a statistically significant contributor to the total explained variance in analyses calculated for arachnid biomass, abundance and web density. 4. A concurrent survey of spider‐web density along 20‐m transects from the stream edge into the forest indicated a strong decline in web‐building spider density moving away from the stream (R² = 0.41, F_1,158 = 109, P < 0.001), with stream‐insect biomass as a significant covariate (F_1,149 = 17.7, P < 0.001). 5. Our results suggest that productivity gradients present in the donor system affect the magnitude of the interaction between adjacent habitats. Productivity gradients may lead to increased reciprocal subsidies through a positive feedback loop involving the predation of spiders and other predatory terrestrial invertebrates by aquatic predators. However, terrestrial insectivores such as birds, bats and lizards that are not readily used as prey by aquatic predators may circumvent the feedback cycle by consuming a large proportion of emergent aquatic‐insect biomass. This may lead to asymmetry in the strength of food–web linkages between aquatic and terrestrial habitats.     

Dietary changes in predators and scavengers in a nocturnally illuminated riparian ecosystem

Aquatic and terrestrial ecosystems are linked by fluxes of carbon and nutrients in riparian areas. Processes that alter these fluxes may therefore change the diet and composition of consumer communities. We used stable carbon isotope (δ¹³C) analyses to test whether the increased abundance of aquatic prey observed in a previous study led to a dietary shift in riparian consumers in areas illuminated by artificial light at night (ALAN). We measured the contribution of aquatic‐derived carbon to diets in riparian arthropods in experimentally lit and unlit sites along an agricultural drainage ditch in northern Germany. The δ¹³C signature of the spider Pachygnatha clercki (Tetragnathidae) was 0.7‰ lower in the ALAN‐illuminated site in summer, indicating a greater assimilation of aquatic prey. Bayesian mixing models also supported higher intake of aquatic prey under ALAN in spring (34% versus 21%). In contrast, isotopic signatures for P. clercki (0.3‰) and Pardosa prativaga (0.7‰) indicated a preference for terrestrial prey in the illuminated site in spring. Terrestrial prey intake increased in spring for P. clercki under ALAN (from 70% to 74%) and in spring and autumn for P. prativaga (from 68% to 77% and from 67% to 72%) and Opiliones (from 68% to 72%; 68% to 75%). This was despite most of the available prey (up to 80%) being aquatic in origin. We conclude that ALAN changed the diet of riparian secondary consumers by increasing the density of both aquatic and terrestrial prey. Dietary changes were species‐ and season‐specific, indicating that the effects of ALAN may interact with phenology and feeding strategy. Because streetlights can occur in high density near freshwaters, ALAN may have widespread effects on aquatic–terrestrial ecosystem linkages.     

Microcrustacean prey and macroinvertebrate predators in a stream food web

1. The consumption of microcrustacea by two polyphagous predators, larvae of the caddisfly Plectrocnemia conspersa (Curtis) and the alderfly Sialis fuliginosa Pictet, was investigated in an English stream with a well-known macro- and microinvertebrate fauna. Benthic samples were collected in August, November, December and April, and the gut contents of all individuals of both predators were examined. 2. All the microcrustacean groups (Cyclopoida, Harpacticoida, Chydoridae and Ostracoda) were identified in gut contents. Of the ten taxa present in the benthos, all occurred in the diet of P. conspersa; nine were found in S. fuliginosa. 3. Ontogenetic shifts in the diets of both predators were found, and microcrustacea were consumed more frequently by small than large instars. 4. There was little evidence of selective feeding by P. conspersa, whereas ostracods were over-represented in the diet of S. fuliginosa, compared with benthic relative densities. The Chydoridae were under-represented in the diet of both predators. 5. The food web of Broadstone Stream is perhaps the most detailed web available for any running water habitat. Increased taxonomic resolution produced marked changes in values of connectance and predator-prey ratios. Linkage density remained fairly constant at different levels of resolution and were high, indicative of a web of generalist species. Omnivory was pronounced and may be characteristic of donor-controlled systems where organic detritus is the primary energy base.     

Ecology of invertebrate predators in a Coastal Plain stream

1. The spatial and temporal abundance patterns, production and feeding habits of invertebrate predators were determined in a sand-bottomed, headwater stream in Virginia, U.S.A. 2. Annual mean density and biomass of predators in debris dams were 3897 individuals m^?2 and 2.5 g dry mass m^?2, respectively, but only 711 individuals m^?2 and 0.2 gm^?2 on the sediment. Predator production was 8.36 gm^?2yr^?1 in debris dams compared to 1.52 gm^?2yr^?1 on the sediment. Annual predator production, weighted by habitat availability, was 1.73gm^?2yr^?1. 3. The predominant taxa in terms of production were the chironomids Thienemannimyia spp, complex, Ablabesmyia parajanta, Zavrelimyia sp., and the odonate Cordulegaster maculata. Chironomidae and Odonata together comprised 77% of the production of the predator guild. 4. Based on gut content analysis and calculations of the trophic basis of production, estimated predator production was supported mostly by Chironomidae (38%), detritus (20%), unidentifiable insects (14%), and Ephemeroptera (11%). Total food ingestion by predators was 9.8 gm^?2yr^?1, 63% of which was detritus and 37% of which was animal material on an areal basis. The predator guild consumed an estimated 94% of primary invertebrate consumer production on the channel surface of the stream.     

Factors controlling hyporheic respiration in a desert stream

1. Experimental manipulations were performed to determine the biological, chemical and physical attributes that govern sediment respiration in the hyporheic zone of Sycamore Creek, a Sonoran Desert stream. 2. Hyporheic respiration per unit volume of sediment was inversely related to diameter of sediment particles, indicating that respiration is affected by availability of substrate for microbial colonization (i.e. sediment surfaces). Respiration rate per unit surface area on sediments was positively correlated with particle diameter, indicating greater metabolic activity of microbes on larger sediments. 3. Hyporheic respiration was more than twice as high in water collected from the surface flow than from subsurface flow. Further, hyporheic respiration was highest immediately following exposure of sediments to surface water and declined over time, presumably due to exhaustion of labile organic matter. 4. Microbial activity was stimulated by addition of algal leachate; however, amendments of leaf leachate had little effect. Respiration was also elevated with dextrose and leucine amendments, but not with inorganic nitrogen additions, indicating hyporheic respiration is carbon limited. 5. Water from the stream surface is probably enriched in labile organic matter derived from algae and stimulates respiration at points of hydrologic downwelling where surface water enters hyporheic sediments. The physical structure of sediments further affects metabolism by affecting the area available for microbial attachment.     

The lateral extent of the subsidy from an upland stream to riparian lycosid spiders

Adult aquatic insects emerging from streams can subsidize riparian food webs, but little is known of the spatial extent of these subsidies. Stable isotope (¹⁵N) enrichment of aquatic insects, principally a species of stonefly (Plecoptera: Leuctridae), emerging from an upland stream was used to trace the subsidy from the stream ecosystem to riparian spiders (Lycosidae). The downstream profile of spider δ¹⁵N correlated closely with that of adult stoneflies, indicating that they were deriving nutrition from aquatic sources. The contribution of adult aquatic insects to spider diets was determined using a two-source mixing model. Adult aquatic insects made up over 40% of spider diets adjacent to the stream, but <1% at 20 m from the stream. Enrichment of riparian spiders declined exponentially with distance from the stream channel. Aquatic-terrestrial subsidies were spatially restricted, but locally important, to riparian lycosid spiders at the study site.     

Denitrification and nitrous oxide emissions from riparian forests soils exposed to prolonged nitrogen runoff

Compared to upland forests, riparian forest soils have greater potential to remove nitrate (NO₃) from agricultural runoff through denitrification. It is unclear, however, whether prolonged exposure of riparian soils to nitrogen (N) loading will affect the rate of denitrification and its end products. This research assesses the rate of denitrification and nitrous oxide (N₂O) emissions from riparian forest soils exposed to prolonged nutrient runoff from plant nurseries and compares these to similar forest soils not exposed to nutrient runoff. Nursery runoff also contains high levels of phosphate (PO₄). Since there are conflicting reports on the impact of PO₄ on the activity of denitrifying microbes, the impact of PO₄ on such activity was also investigated. Bulk and intact soil cores were collected from N-exposed and non-exposed forests to determine denitrification and N₂O emission rates, whereas denitrification potential was determined using soil slurries. Compared to the non-amended treatment, denitrification rate increased 2.7- and 3.4-fold when soil cores collected from both N-exposed and non-exposed sites were amended with 30 and 60 μg NO₃-N g⁻¹ soil, respectively. Net N₂O emissions were 1.5 and 1.7 times higher from the N-exposed sites compared to the non-exposed sites at 30 and 60 μg NO₃-N g⁻¹ soil amendment rates, respectively. Similarly, denitrification potential increased 17 times in response to addition of 15 μg NO₃-N g⁻¹ in soil slurries. The addition of PO₄ (5 μg PO₄-P g⁻¹) to soil slurries and intact cores did not affect denitrification rates. These observations suggest that prolonged N loading did not affect the denitrification potential of the riparian forest soils; however, it did result in higher N₂O emissions compared to emission rates from non-exposed forest soils.     

The impact of vertebrate and invertebrate predators on a stream benthic community

I assessed the impact of both vertebrate and invertebrate predators on a lotic benthic community in a 1-month-long experiment, using enclosures containing cobble/gravel bottoms, with large-mesh netting that allowed invertebrates to drift freely. Brown trout (Salmo trutta) and leeches (Erpobdella octoculata) were used as predators and four treatments were tested: a predator-free control, leeches only, trout only, and leeches and trout together. A density of 26.7 leeches/m² (20 leeches/enclosure) and 1.3 trout/m² (one trout per enclosure) was stocked into the enclosures. The total biomass of invertebrate prey was significantly lower in the trout and trout plus leech treatments than in the leech and control treatments, which were due to strong negative effects of trout on Gammarus. On the individual prey taxon level, both trout and leeches affected the abundance of Asellus , Baetis and Ephemerella, whereas the abundance of Gammarus was only affected by trout, and the abundance of Orthocladiinae and Limnephilidae was only affected by leeches. In the treatment with trout and leeches together, the abundance of Ephemerella and Baetis was higher than when trout or leeches were alone, which was probably due to predator interactions. Leeches and trout had no effects on prey immigration but did affect per capita emigration rates. Both trout and leeches indirectly increased periphyton biomass in enclosures, probably due to their strong effects on grazers. Both trout and leeches were size-selective predators, with trout selecting large prey, and leeches selecting small prey. Size-selective predation by trout and leeches affected the size structure of five commonly consumed prey taxa. Trout produced prey populations of small sizes owing to consumption of large prey as well as increased emigration out of enclosures by these large prey. Leech predation produced prey assemblages of larger size owing to consumption and increased emigration of small prey. These results suggest that in lotic habits, predatory invertebrates can be as strong interactors as vertebrate predators. Received: 23 June 1997 / Accepted: 4 May 1998     

Fish habitat associations in a spatially variable desert stream

White sands pupfish Cyprinodon tularosa were sampled with minnow traps along the length of Lost River, a highly variable desert stream in New Mexico, to evaluate if their abundance, median standard length ( L _S) and sex ratio values differed by microhabitat variables. Habitat measurements were made in a 4 m² area around the minnow trap. White Sands pupfish populations were primarily structured by L _S. As specific conductance and surface area increased, median size of fish decreased. Shallow areas with higher specific conductance were correlated with an increased number of fish.     

Responses of riparian trees and shrubs to flow regulation along a boreal stream in northern Sweden

1. Flow dynamics is a major determinant of riparian plant communities. Therefore, flow regulation may heavily affect riparian ecosystems. Despite the large number of dams worldwide, little specific information is available on the longitudinal impacts of dams on vegetation, for example how far downstream and at what degree of regulation a dam on a river can influence riparian woodlands. 2. We quantified the long‐term responses of riparian trees and shrubs to flow regulation by identifying their lateral distribution and habitat conditions along a boreal river in northern Sweden that has been regulated by a single dam since 1948. The regulation has reduced annual flow fluctuations, this effect being largest at the dam, downstream from which it progressively decreases following the entrance of free‐flowing tributaries. 3. We related changes in the distribution patterns, composition, abundance and richness of tree and shrub species to the degree of regulation along the river downstream from the dam. Regulation has triggered establishment of trees and shrubs closer to the channel, making it possible to measure ecological impacts of flow regulation as differences in vegetation attributes relative to the positions of tree and shrub communities established before and after regulation. 4. Trees and shrubs had migrated towards the mid‐channel along the entire study reach, but the changes were largest immediately downstream of the dam. Shrubs were most impacted by flow regulation in terms of lateral movement, but the effect on trees extended furthest downstream. 5. The species composition of trees progressively returned to its pre‐regulation state with distance downstream, but entrance of free‐flowing tributaries and variation in channel morphology and substratum caused local deviations. Species richness after regulation increased for trees but decreased for shrubs. The changes in species composition and richness of trees and shrubs showed no clear downstream patterns, suggesting that other factors than the degree of regulation were more important in governing life form.