Environmental flows assessment integrating snow trout habitat requirements in the Shakhimardan basin, Central Asia

Vol.:(0123456789) International Journal of Environmental Science and Technology (2025) 22:16935–16946 https://doi.org/10.1007/s13762-025-06761-2 ORIGINAL PAPER Environmental flows assessment integrating snow trout habitat requirements intheShakhimardan basin, Central Asia D.S.Hayes1 · T.Hägele2· I.Kopecki2· B.Zeiringer1· E.Karimov3· B.K.Karimov4· J.Coeck5· P.Verhelst5· J.DeKeyser6· O.Omonov4· M.Schneider2 Received: 7 October 2024 / Revised: 27 July 2025 / Accepted: 28 August 2025 / Published online: 15 September 2025 © The Author(s) 2025 Abstract The increasing demand for hydropower development in Central Asia threatens the ecological integrity of high-mountain rivers. This study applies an ecohydraulics-based approach to assess environmental flow requirements for a diversion hydropower project on the Koksu River, Uzbekistan. We integrated high-resolution 2D hydrodynamic modeling with novel habitat data for snow trout (Schizothorax eurycephalus), incorporating fuzzy logic sets and rules to model habitat preferences across three life stages. Our findings recommend an environmental flow regime that maintains seasonal variability, with base flows ranging from 10.6% to 17.7% of the mean annual flow, ensuring habitat availability during critical life cycleperiods. We contextualize the results within adaptive management frameworks by highlighting preliminary results from an ongoing biotelemetry study. This ecohydraulics-based approach provides a replicable model for environmental flow assessments in Central Asia and beyond. Keywords Abstraction· Eflows· E-flows· Snowtrout· Schizothoracinae· Kyrgyzstan· Shohimardon· Water management· Syr Darya River Introduction Central Asia has a long history of river regulation, particularly water abstraction. During the time when Central Asia was part of the Soviet Union (1918–1991), people pursued the goal of turning the steppes of Central Asia into fertile cropland. The Soviet times led to the creation of weirs and canals that divert and transfer water from the rivers to fields for crop production (Pérez Martín, 2017; Zonn etal. 2020). Unsustainably large water demand led to diminishing water resources for aquatic ecosystems. The resulting shrinkage of the Aral Sea from the 1960s onwards (Graham etal. 2017; Micklin 2010) is one of the best-known environmental disasters, entailing also consequences for regional economics and health (Gungoren and Regallet 1998; Severskiy 2004). River regulation has so far focused on the large rivers in the lowlands and piedmont areas (Karimov and Talskikh 2022). High-mountain Central Asia still features a largely pristine environment when compared to the rivers that flow through the valleys (Gozlan etal. 2019). Nowadays, however, there is an increasing pressure on rivers and streams flowing through this high-mountain area for hydropower Editorial responsibility: Samareh Mirkia. * D. S. Hayes daniel.ha[email protected] 1 BOKU University, Department ofEcosystem Management, Climate andBiodiversity, Institute ofHydrobiology andAquatic Ecosystem Management, 1180Wien, Austria 2 SJE Ecohydraulic Engineering GmbH, 71522Backnang, Germany 3 Department ofZootechnics andVeterinary, Tashkent State Agrarian University, 100140Tashkent, Uzbekistan 4 Tashkent Institute ofIrrigation andAgricultural Mechanization Engineers, National Research University, 100000Tashkent, Uzbekistan 5 Research Institute forNature andForest, 1000Brussels, Belgium 6 BOKU University, Department ofLandscape, Water andInfrastructure, Institute ofHydraulic Engineering andRiver Research, 1200Wien, Austria 16936 International Journal of Environmental Science and Technology (2025) 22:16935–16946 development due to the increasing demand for renewable energy resources (Radovanović etal. 2021). Hydropower development in sensitive areas can put aquatic biota and their ecosystems at risk (Benejam etal. 2016; Geist 2021). Around one-fourth of Central Asia’s 120 fish species are currently listed on the Red List (Gozlan etal. 2019). However, the situation may be even more critical, as still many of the region’s fish species are categorized as data-deficient (Mamilov etal. 2021). There is also insufficient knowledge regarding fish ecology, especially on topics essential for the sustainable development of hydropower, including the determination of environmental flows (Graham etal. 2017; Schmutz etal. 2025). As part of the European ‘Hydro4U’ project, a small diversion hydropower plant is being built at the Koksu River, a mountain stream located in the Uzbek exclave of Shakhimardan, surrounded by Kyrgyzstan. The project aims to ensure sustainable hydropower development by applying international and European standards to prevent loss of habitats and maintain migration corridors (Alapfy etal. 2025; Jorde etal. 2022; Reisenbüchler etal. 2021). River water diversion is one of the key environmental impacts, necessitating mitigation through the release of adequate environmental flows (Schmutz etal. 2025), i.e., river flows “capable of maintaining the natural functions and processes regarding quality, quantity, and temporal cycles, to retain the integrity and resilience of riverine ecosystems […] as well as associated ecosystem services” (Hayes etal. 2018). Different methods exists to determine environmental flows (hereafter “e-flows”) (Acreman and Dunbar 2004; Poff and Matthews 2013), and the method choice may depend upon ecosystem size and complexity, conservation needs, as well as number, type, and importance of users, conflicts among them, or percentage of ecosystem affected (Nale etal. 2020). Methodologies based on habitat modeling exhibit a strong linkage to ecological requirements and allow a good understanding of what flows are required to reach certain protection levels. This is done by linking data about the physical properties of a studied river (e.g., water depths, flow velocities, substrate sizes) with information on the physical requirements of a target species and their development stage. Once these functional relationships are defined, they can be applied to different river flow scenarios (Acreman and Dunbar 2004; Melcher etal. 2018). The use of habitat modeling based on the outcome of hydrodynamic calculations allows for a detailed analysis of habitats and fish passage as crucial elements of sustainability as, for example, listed in the hydropower sustainability assessment protocol of the International Hydropower Association (IHA 2020). We evaluate habitat quality at different flow magnitudes via fuzzy rule-based modeling (Noack etal. 2013; Schneider 2001), allowing to integrate expert knowledge and data from habitat studies. These fuzzy rule systems provide a spectrum of potential physical criteria combinations, allowing the determination of habitat quality and, subsequently, deduction of environmental flow recommendations (Melcher etal. 2018). The Koksu River hosts Schizothorax eurycephalus, a snow trout species native to Central Asia’s Syr Darya River drainage (Rozimov etal., in prep.), but for which limited ecological knowledge is available to date. First population status assessments in Koksu River revealed the presence of all life cycle stages and various age classes (labeled as S. eurystomus in last publications; Karimov etal. 2024, 2023), showing a good natural reproduction in the catchment. However, river reaches further downstream in the basin are heavily fragmented, underlining the ecological significance of the Koksu River for snow trout. Environmental flow releases that sustain habitats for all life stages are, therefore, key for fish ecological sustainability. This study assesses e-flow requirements for a small diversion hydropower project on the Koksu River. Using habitat modeling, we incorporated novel microhabitat preference data for three life stages – juvenile, sub-adult, and adult – of snow trout (Schizothorax eurycephalus), collected on-site in April 2022. These habitat preferences were translated into fuzzy logic rules and applied within the CASiMiR habitat modeling software (Noack etal. 2013; Schneider 2001). We derived habitat suitability criteria to determine e-flow magnitude, proposing a regime that maintains seasonal flow variability in alignment with the natural flow regime to sustain adequate snow trout habitat quality and quantity. This study provides a case study of an ecohydraulics-based e-flows assessment for a hydropower project in Uzbekistan. The snow trout preference data can be applied throughout Central Asia and beyond, and our approach offers a replicable framework for future environmental assessments. Study site The study area is located in the Uzbek exclave of Shakhimardan in the Alai Mountain Range south of the Fergana Valley (Fig.1). The river system of the exclave consists of the Aksu and the Koksu Rivers. At their confluence, they form the Shakhimardan River (Fig.1). Together, the Koksu and the Aksu Rivers drain an area of about 800 km2, which is characterized by high mountains ranging from 1,388 to 5,235m in elevation (mean = 3,280m). The Koksu River, the smaller tributary, contributes about 150 km2 to this total catchment area. The hydrology and sediment transport of the Koksu River within the exclave are strongly influenced by a natural earth dam around two kilometers upstream of the study site (Fig.1). The dam was formed by the sediments of a massive compound wedge failure triggered by an earthquake in 1766 16937International Journal of Environmental Science and Technology (2025) 22:16935–16946 (Strom and Abdrakhmatov 2018). This event led to the formation of two lakes: Lake Kurbankul and the smaller Lake Yashilkul. Both lakes are fed by snow and glacier melt and have been steadily increasing in size. The dam acts as a sediment barrier and a water filter (Khikmatov and Pirnazarov 2020). Water infiltrates through the landslide deposits and re-emerges through springs at the valley bottom (Pak 2019). The Koksu River’s flow regime can be characterized as nivo-glacial with the lowest flows occurring in April and May, and highest ones in July and August (Fig.2a). Median Fig. 1 Location of the study site in Central Asia, including details on a the Shakhimardan exclave and b the Koksu River and the four study reaches (R1-R4) for environmental flow modeling Fig. 2 Monthly a river discharge and b water temperature of Koksu River. The plots depict the median ± 95% confidence interval (2011–2020). Data source: Hydromet, Uzbekistan 16938 International Journal of Environmental Science and Technology (2025) 22:16935–16946 river discharges during January to May (2.1 m3/s) are half of the flow magnitude of those occurring from July to September (4.2 m3/s). Maximum discharges are usually triggered by rainfall events. This relatively small variability in discharge between high and low flows throughout the year is related to the natural sediment dam upstream of the study area, which buffers the flow (mean annual flow = 2.8 m3/s). Similar to river flows, also water temperature remains fairly constant, ranging from 8.0 to 10.6°C monthly average over a ten-year observational period. Median monthly water temperatures exceed 10°C from June to September. From December to February, they are below 9°C (Fig.2b). The Koksu River’s morphology is in a near-natural state with only small-scale bank protection measures (e.g., concrete walls, gabions) and bridges. An artificial waterfall of ca. 3.5m in height (39°58′10.23" N, 71°49′47.32" E), created by road construction in the 1970s, blocks upstream fish migration in the mid-section (Karimov etal. 2023). The river has a mean gradient of 4% between the natural dam and the confluence with the Aksu River, which is typical for a mountain river (Comiti etal. 2007). The upstream section is characterized by step-pool sequences, whereas the lower parts features riffle-pool habitats. The river habitats feature a high diversity of river width and depth, as well as flow velocities and substrate conditions. A 2MW diversion hydropower with the novel Francis Container hydropower type is being constructed in the Koksu River (Jorde etal. 2022). The weir (39°57′54.73" N, 71°49′59.09" E) will create an impoundment, from where up to 3 m3/s of water is diverted through a 2.3km long pressure pipeline leading to the powerhouse (39°58′54.10" N, 71°48′42.56" E). To ensure adequate environmental water releases in the river stretch affected by water diversion, we performed eco-hydraulic surveys in four study reaches, representative for the different morphological characteristics of the river (Fig.1), as a basis for determining an e-flow regime. Reach#1 is positioned just downstream of the diversion weir, whereas reach#2 is located another 265m further downstream. Reaches#3 and #4 are located in the downstream part of diversion stretch: Reach#4 is situated next to the powerhouse and reach#3 starts 365m upstream of reach#4. The length of the investigated detail reaches ranged from 80 to 220m. Reach#3 is wider than the other reaches and provides larger portions of shallow areas during low flows. Reach#2 is the narrowest one, with a deep central channel and a comparatively small bank zone. Materials andmethods The following steps were performed to study the sustainability of the small diversion hydropower at Koksu River with regards to e-flow magnitudes. (i) Fish surveys were conducted through electrofishing, incl. habitat assessment of the target species, followed by (ii) setup of hydrodynamic numerical 2D models, and(iii) habitat simulations to predict the quality and availability of aquatic habitat under different flow magnitudes to (iv) deduct e-flow recommendations. Habitat assessment oftarget species: snow trout Two native fish species inhabit the Shakhimardan River basin: snow trout (Schizothorax eurycephalus) and stone loach (Triplophysa ferganaensis) (Karimov etal. 2023; Rozimov etal., in prep.; Sheraliev and Peng 2021). However, only snow trout occurs in large parts of the Koksu River. It is also of commercial value, being an object of recreational fisheries and a species that is poached for food. Even though S.eurycephalus is listed as ‘least concern’ (LC) in the Red List, it is threatened by hydropower development, water abstraction, and pollution, which can cause population declines throughout its range (Karimov 2015; Karimov etal. 2019). Therefore, snow trout was selected as the target species for this study to indicate environmental water requirements. S.eurycephalus is a rheophilic species native to many Central Asian rivers, including the upper sections of the Syr Darya River and its headwaters (Berg 1948; Nikolsky 1938; Rozimov etal., in prep.). The snow trout hides under stones during the day. At night, it leads a more active lifestyle. Its diet varies, feeding on, e.g., plankton, aquatic vegetation, terrestrial insects that fall into the water, chironomid larvae, and mollusks. Large individuals also feed on small fish and fry (Karimov etal., accepted; Nikolsky 1938). S. eurycephalus reaches maturity in the third or fourth year of life with a body length of around 20cm for females and 16cm for males but there is also an earlier maturation with a length of 10–12cm, indicating dwarfism, perhaps mainly in small rivers (Berg 1948; Nikolsky 1938) like Koksu River. Some species of the genus Schizothorax are resident while others are potamodromous, performing spawning migrations to the upper reaches of the rivers and tributaries. Fish spawning in valley and foothill reservoirs is extended and lasts from April to September but in the high-altitude conditions of the Pamirs, spawning takes place in a short time in June-July (Grishchenko 1976; Timirkhanov 2001). However, in our field studies in such areas During 2021–2023 we have found much earlier running males. Point electrofishing sampling was conducted in April of 2022 to study snow trout habitat use in the Shakhimardan basin, i.e., Koksu, Aksu, and Shakhimardan Rivers (Fig.1). We used backpack electrofishing gear (Honda FEG 1500, 1.8kW) to sample the full array of mesohabitats (i.e., pool, riffle, run, ruffle, pocket pools, cascade, or combinations thereof) in the study area. For each microhabitat where a snow trout was caught (n = 240), total fish length [mm] 16939International Journal of Environmental Science and Technology (2025) 22:16935–16946 and the following (hydraulic) parameters were recorded: (i) water depth [cm], (ii) mean flow velocity [cm/s] (averaged from two-point measurements at 80% (v80) and 20% (v20) of water depth), (iii) and dominant substrate sizes (choriotope). Habitat preferences for each of these three abiotic parameters were univariately assessed for three life cycle stages delineated by total fish length: juvenile (< 120mm), subadult (120–200mm), and adult (≥ 200mm) (Zeiringer etal., in prep.). These habitat assessment data were then converted into fuzzy sets and corresponding fuzzy rules for each of the three age classes, forming a multivariate habitat preference function (Zeiringer etal. 2025). Hydrodynamic numerical modeling We selected four reaches, representing the diversity of morphological conditions (see above), within the diversion stretch for e-flows assessment (Fig.1). For each of these reaches, the bathymetry was created using terrestrial survey points and a digital elevation model (DEM) obtained from drone flight pictures processed with the Structure from Motion (SfM) approach (Westoby etal. 2012). The comparison with the terrestrial survey confirmed a high accuracy of the SfM data. Hydrodynamic modeling was performed using the twodimensional hydrodynamic model SRH-2D (Lai 2010). The hydrodynamic numerical 2D model was calibrated by comparing modelled water surface elevations with corresponding water surface elevations measured during the survey campaign. Modeling was performed with a high resolution in the four detail reaches (see R1-R4 in Fig.1), as well as for the whole diverted Koksu stretch in a slightly lower resolution. The model for the whole stretch was used for the assessment of minimum (migration) depth whereas the detail stretches were considered not only for the evaluation of minimum depth but also for habitat evaluation. Model roughness was set according to Kopecki etal. (2017) depending on water depth. The model’s boundary conditions were defined as inflow discharge at the upstream model end and as critical outflow at the outflow (steep gradient river). Habitat simulations E-flow magnitudes must ensure (1) a minimum water depth in the migration corridor to allow fish movements and (2) best possible availability of aquatic habitat under the constraints of water diversion. The assessment of flow-dependent changes of instream passability and habitat based on the results of habitat suitability modeling is internationally acknowledged state-of-the-art for e-flows determination (Melcher etal. 2018; Zeiringer etal. 2018). Based on the results of the 2D hydrodynamic-numerical modeling, we generated maps for water depths at different flow magnitudes to evaluate snow trout passability. Water depths < 10cm were considered non-passable. Depths > 20cm were considered fully passable. Short river sections with water depths between 10 and 15cm were considered as pessimal sections where migration is restricted but still possible if its length is short (Zeiringer etal., in prep.). Furthermore, we used the habitat modelling system CASiMiR (Noack etal. 2013; Schneider 2001) to predict the quality and availability of snow trout habitat. CASiMiR intersects the habitat requirements of target species with the results of a hydraulic model (local water depth, flow velocity) and morphological information (bottom substratum, cover for fish). The model uses a fuzzy rule-based approach for the description and processing of habitat requirements. We used the originally univariate preference functions of snow trout habitat requirements (Zeiringer etal., in prep.) to set up a multivariate description (see Zeiringer etal. 2025). In a first step, we define categories of suitability as e.g. “optimal” or “rather too high” or “unsuitable high” for the three different habitat parameters considered: (i) flow velocity, (ii) water depth, and (iii) size of dominant substrate in the river bottom’s surface layer. These categories are then described by fuzzy sets that can be illustrated through membership functions (see Fig.3). In this model, the considered habitat parameters flow velocity and depth, derived from hydrodynamic data, along with substrate size from mapping, are not confined to single, Fig. 3 Fuzzy sets for five categories of flow velocity described by membership functions (red = unsuitable low, orange = rather too low, green = optimal, turquoise = rather too high, brown = unsuitable high) 16940 International Journal of Environmental Science and Technology (2025) 22:16935–16946 discrete categories. Instead, according to the fuzzy set theory, they are assigned to multiple categories (fuzzy sets) simultaneously, reflecting the inherent variability and overlap in ecological systems. For example, a flow velocity of 0.4m/s might be considered 50% within the “optimal” flow velocity category (indicated in green in Fig.3) and 50% within the “rather too low” flow velocity category (indicated in orange in Fig.3). This dual categorization allows for a more gradual transition between categories, capturing the inherent imprecision that typifies ecological relationships. The categories in this model are crafted with specific attention to the preferences of the species and life stages for which they are defined, as suggested by the linguistic terms used. This approach enables the description of habitat requirements in a comprehensive, multivariate manner. For instance, the model incorporates rules such as: If flow velocity is “rather too high” and water depth is “optimal” and substratum size is “rather too high”, then the overall habitat suitability is considered “high.” This rule reflects the ecological knowledge that, for species like the bottomoriented snow trout, a higher-than-preferred flow velocity can be offset by a similarly high substratum size, assuming the water depth is not a limiting factor. This compensatory mechanism allows the transformation of expert knowledge into numerical parameters for habitat suitability (Zeiringer etal. 2025). To aid in understanding and visualizing these complex interactions, the habitat requirements are presented as habitat suitability squares. As an example, Fig.4 illustrates the requirements for sub-adult snow trout, where each square on the grid corresponds to a unique combination of flow velocity (horizontal axis) and water depth (vertical axis), with the substrate variability represented by different squares (10 substrate choriotopes = 10 squares). The color of each square indicates the level of habitat suitability according to the legend, providing a clear, visual summary of the data. The complete list of the fuzzy rules behind the illustrated habitat suitability squares for all three snow trout life stages can be found in Zeiringer etal. (2025). The simulation results in a habitat suitability index SIbetween 0 (i.e., unsuitable) and 1 (i.e., perfectly suitable). The calculated suitabilities are appointed to each mesh element of the hydraulic model. That way, habitat suitability maps showing the spatial habitat distribution are created. To this aim, five suitability classes are defined by splitting the SI into equal sizes (Fig.4). Environmental flows assessment We consider natural streamflow patterns (Fig.2a) and snow trout life cycle habitat and migration requirements (Zhang etal. 2021) to determine e-flow rules for distinct seasons – winter (December to March), spring (April to May), summer (June to September), and fall (October to November) – to ensure seasonal flow variability in accordance to the natural flow regime. Results anddiscussion To determine e-flow regimes for the Koksu River, we assessed instream passability (i.e., the available migration corridor) and the quality and availability of snow trout habitat at different flow magnitudes. Instream migration corridor The first assessment focused on defining flow magnitudes needed to guarantee instream passability. Based on the minimum water depth criteria of 20cm or 10–15cm in shorter sections, we defined minimum flows necessary to ensure a year-round migration corridor at 0.3 m3/s. Figure5 illustrates this result for reach#3, which reflects the shallowest reaches within the whole diverted stretch. Fig. 4 Habitat suitability squares illustrating the habitat requirements of sub-adult snow trout 16941International Journal of Environmental Science and Technology (2025) 22:16935–16946 Snow trout habitat In the second step, we assessed habitat changes with regards to flow magnitude alterations by analyzing the distribution of suitability index (SI) classes. The SI class distribution illustrates habitat availability and gives more detailed information on the available wetted areas with high, medium and low-quality habitats in five classes for all investigated discharges. Figure6 exemplifies the spatial changes in sub-adult fish habitat at five different flow magnitudes. In a further step, the total area of habitats per SI class is plotted for the entire range of assessed discharge conditions (0.3–2.4 m3/s) to allow for an integrated habitat assessment (Fig.7, FiguresS1-S2). Environmental flow recommendations Based on the fish sampling data, the critical life cycle stage (size class) for determining e-flows in Koksu River are sub-adult snow trout (120–200mmTL), as adult fish were hardly caught in Koksu River during the sampling campaign. The habitat data show that juveniles are less impacted by flow changes due to water abstraction than the other two life stages. Hence, if sub-adult habitats are sustained, juvenile snow trout will find adequate habitats, too (see also Fig.7 and FiguresS1-S2, reflecting the availability of highly suitable habitat over flow). Following, the recommended e-flows regime is described for each of the four seasons (Table1). A winter base flow of 0.3 m3/s is recommended, meaning that a mean weighted usable area (WUA) of 73% is sustained when compared to the natural flow of 2 m3/s. Regarding the highest SI groups (SI > 0.6; see Fig.7 and FiguresS1-S2), this winter baseflow would ensure 41% of high-quality habitats. Overall fish passability is guaranteed, exhibiting only small sections with passage water depths 10–15cm. April and May are likely critical times of the year as snow trout become active and therefore require higher discharges (e.g., in preparation for spawning activities), but spill flows are not yet available (see below). Therefore, we recommend releasing a base flow of 0.5 m3/s with the aim of improved migration for medium-sized fish with depths ≥ 0.15m. For this flow rate, 78% of the WUA that is available forthe reference discharge are provided; aside from that,63%of high quality habitats remain available. According to the hydrological data,weir overflow is likely to occur from July to September (but not yet in June). Thus,the required e-flows, including regular spill Fig. 5 Water depth map for reach#3, the shallowest of the four sections, at a flow magnitude of 0.3 m3/s Fig. 6 Habitat suitability maps for sub-adult snow trout at reach#1 at flow magnitudes between 0.3 and 2.4 m3/s 16942 International Journal of Environmental Science and Technology (2025) 22:16935–16946 flows, can bemost probably achieved (Fig.8). The summertime baseflows are 0.5 m3/s (Table1). In fall, a minimum baseflow requirement of 0.4 m3/s is proposed. These flows are slightly higher than winter discharges to allow for unhindered migration. Discussion Schizothoracinae fish are widespread throughout Asia (Du etal. 2024). Many species belonging to this cyprinid group, such as S. eurycephalus, are ecological indicators that are Fig. 7 Integrated habitat results SI areas chart subadult snow trout. See the Supplementary Material for results on juvenile and adult fish Table 1 Monthly baseflows [m3/s] for the recommended environmental flows release for Koksu River Month Q [m3/s] January 0.30 February 0.30 March 0.30 April 0.50 May 0.50 June 0.50 July 0.50 August 0.50 September 0.50 October 0.40 November 0.40 December 0.30 Mean 0.42 Fig. 8 Monthly environmental flow baseflows and average spill flows for the Koksu River The curve including spill flows is based on the mean monthly natural flows subtracted by a pressure pipe discharge of 3 m3/s 16943International Journal of Environmental Science and Technology (2025) 22:16935–16946 threatened by hydropower development (Bhatt and Manish 2023). A hydropower dam on the upper Mekong River caused recruitment failure of S.lissolabiatus by disrupting the natural river flow regime. This altered flow lead to a narrower window for reproduction, causing the young-ofthe-year fish to disappear from river sections downstream of the dam (Ding etal. 2023). In the Koksu River, a new hydropower project will lead to the diversion of water along a three-kilometer long river stretch (Jorde etal. 2022). This article assessed environmental water requirements for this stream section by performing fuzzy-logic habitat modeling for snow trout using novel microhabitat preference data (Zeiringer etal. 2025). Our e-flows recommendation aims at sustaining the snow trout population with a mean WUA of around 80% in critical seasons. The high-quality habitats (SI > 0.6), however, are reduced to around 60% year-round. With regards to high-quality habitat, it shall be noted that there is a steeper gradient from 0.3 to 0.5 m3/s than from 0.5 m3/s to higher discharges (see also Fig.7, FiguresS1-S2). Therefore, minimum baseflows of 0.5 m3/s are recommended for spring and summer. In the case that water flows over the weir, an even larger share of high-quality habitats remains (Fig.8). Population status assessments revealed the presence of all life cycle stages and various age classes (Karimov etal. 2024), showing a good natural reproduction in the Shakhimardan River basin. However, fish were completely absent upstream of the artificial barrier in the residual flow stretch of Koksu River (Fig.1), presenting an insurmountable migration barrier in the river reach affected by water abstraction. If adequate e-flows with seasonal flow variability are combined with measures to re-establish connectivity at the waterfall and the diversion weir, this project could enlarge the usable river length for snow trout by around two river kilometers compared to the status quo. To this aim, the construction of an asymmetric rough ramp (Mühlbauer etal. 2020) and a vertical slot fish pass is currently underway at the waterfall and the weir, respectively. The current legislation of the Republic of Uzbekistan does not provide specific norms or guidelines for e-flow releases. In light of this, an e-flows regime releasing seasonal baseflows 10.6–17.7% of the mean annual flow is within the range determined by methods globally (Tharme 2003). Our assessment suggests releasing seasonally staggered baseflows, similar to the outcomes of the building block methodology (King and Louw 1998). Such staggered e-flow releases have been proven successfully in a variety of case studies, such as improving habitats for Thymallus thymallus in a diversion stretch of an Austrian river (Zeiringer etal. 2015). Aside from such seasonal changes in baseflow magnitude, also short-term flow dynamics may constitute an important feature of e-flow regimes, including peak flows supporting the resetting of river morphology (Hayes etal. 2018; Yarnell etal. 2015), and promoting the recovery of native fish (Baruch etal. 2024). In the Koksu River, river flows are naturally controlled by the rockslide, limiting seasonal dynamics (Fig.2a). Reductions in future water availability due to climate change are indirectly integrated by recommending a fixed seasonal discharge with regards to the present mean seasonal flow patterns. This study’s e-flows assessment was done using microhabitat data from spring 2022. At that time, the fish density was rather low. Under the assumption that this represents a good ecological status for a Central Asian mountain river, the fish are resident in the river stretch year-round, and the site is not impacted by poaching, the recommended e-flows release sustains highly suitable habitats, ensuring an adequate habitat supply for the present fish. A regular electrofishing campaign in the fall of 2022 revealed even higher densities of adult and sub-adult snow trout in Koksu River than in spring (unpubl. data). The fish survey in fall showed that larger fish use more or less the same habitat as sub-adults that were found in spring, justifying relating the assessment to the habitat results for the sub-adult life stage. The adult life stage is indirectly integrated in the assessment, assuming that a habitat overlap is given in this case. Considering that the impacts fish species and aquatic habitats are rarely considered in Uzbekistan and Central Asia more generally as part of hydropower development, these data present and assessment approach presents a much needed step forward to ensure the sustainability of water infrastructure projects (Jorde etal. 2022). Nonetheless, it was surprising to catch and observe much more snow trout in the Koksu River during fall, with numbers in the hundreds, than during the spring. In contrast, adult snow trout were less prominent in the much larger Shakhimardan River in fall than they were in spring. Notably, the waterflow conditions in the Shakhimardan River made it hard to perform electrofishing on the majority of the river due to high discharges by melting water coming from the mountains. Some snow trout species tend to migrate downstream to winter habitats (Rai etal. 2002). In the study area, a reverse pattern can be observed: first results of a radio telemetry study with 29 tagged snow trout indicate that, in spring, fish leave the Koksu River to go downstream into the Shakhimardan River, even exiting the exclave and migrating into the Kyrgyz part of the Shakhimardan River. In fall, they move upstream to areas near their initial catching and tagging location. This may indicate high site fidelity. On the other hand, the Koksu River may provide key overwintering habitats as the water temperature of the Koksu River is warmer than in the Shakhimardan River during fall and winter. The ongoing telemetry study will shed light on the spatiotemporal movement behavior of snow trout and how they use different habitats throughout the year. In this regard, the