Temporal dynamics override spatial gradients in Afrotropical reservoir water quality

Temporal dynamics override spatial gradients in Afrotropical reservoir water quality | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Temporal dynamics override spatial gradients in Afrotropical reservoir water quality Abdullahi Ibrahim, Suleiman Omeiza Eku Sadiku, Deborah Vivienne Robertson-Andersson, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9118764/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 11 You are reading this latest preprint version Abstract Freshwater systems in tropical Africa face growing pressures from dams, climate change, and land-use expansion, yet we still do not fully understand which factors most significantly affect water quality. Knowing this is crucial for designing effective monitoring and management, especially when resources are limited. We conducted monthly water-quality surveys at six sites across the Kainji–Jebba reservoir cascade in Nigeria from April 2024 to March 2025, encompassing two dams and three distinct habitat types (riverine, ecotonal, and lacustrine). Using a modified NSF Water Quality Index (WQI), variance partitioning (RDA), Granger causality, and six changepoint detection methods, we untangled the effects of space, time, land use, and climate, and identified early-warning indicators of water quality change. Our findings show that when changes happen matters more than where. Seasonal and monthly variations explained 28.3% of water quality variance, while spatial differences accounted for only 1.8%, and direct land-use effects were minimal (0.4%). Agricultural impacts were expressed primarily through short, wet-season runoff pulses that affected all sites simultaneously, rather than through persistent spatial differences. After false discovery rate correction, no relationships between climate variables and water quality remained significant. WQI, electrical conductivity (EC), and relative humidity (RH) emerged as reliable early-warning sentinel indicators, detecting regime shifts in 71.4% of cases. These findings suggest that monitoring programmes in tropical reservoir systems should prioritise targeted seasonal windows: the wet-season runoff peak (June–July) and the arrival of sediment-rich black floods (January–February), rather than expanding spatial coverage. Reservoir water quality Temporal variability Sentinel indicators Land-use impacts Niger River Basin Tropical Africa Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Introduction Freshwater ecosystems harbour disproportionate biodiversity relative to their spatial extent. Covering less than 1% of Earth’s surface, they support over 10% of all animal species and more than half of all fish species (Albert et al., 2021 ; Reid et al., 2019 ). However, these systems are in rapid decline. Freshwater vertebrate populations have crashed by 84% since 1970, outpacing losses in forests and oceans, with nearly a third of freshwater fish now threatened with extinction (Tickner et al., 2020 ; WWF, 2020 ). Dams are a major culprit: 58,000 large dams now fragment 60% of the world’s rivers, disrupting the flow patterns on which aquatic life depends (Barbarossa et al., 2020 ; Grill et al., 2019 ). Understanding how dams, climate, and land-use change interact to degrade water quality is therefore crucial, especially in tropical regions where biodiversity is highest, and data are scarcest (Faye et al., 2023 zéquel et al., 2020 ). Most of what we know about dam impacts comes from studies of temperate rivers. The River Continuum Concept predicts that river chemistry and biology change gradually downstream, with dams creating distinct zones of disruption (Vannote et al., 1980 ; Ward & Stanford, 1995 ). Temperature and rainfall drive seasonal patterns by influencing oxygen levels, nutrient processing, and thermal stratification (Jane et al., 2021 ; Poff et al., 1997 ). Tropical rivers, however, work differently. The Flood Pulse Concept posits that seasonal flooding, rather than downstream gradients, is the primary organising force (Junk et al., 1989 ). Annual floods connect rivers to floodplains, driving nutrient cycling and fish reproduction in ways dams supposedly interrupt (Tockner & Stanford, 2002 ; Wantzen et al., 2008 ). Even so, dammed tropical rivers often maintain surprising connectivity through spill events and fish movement, potentially buffering the effects of fragmentation (O’Mara et al., 2024 ; Walsh et al., 2022 ). Whether temperate or tropical models better predict water quality in regulated African rivers remains largely untested. A third set of drivers, direct anthropogenic modification of watersheds, may supersede both. Tropical deforestation removes riparian buffers that filter runoff (FAO, 2020 ; Oester et al., 2025 ). Fertiliser use has nearly doubled global riverine nitrogen transport to the ocean over recent decades (Beusen et al., 2016 ), while rapid urban growth generates large volumes of largely untreated wastewater (Jones et al., 2021 ). If these anthropogenic pressures dominate, then dam location and climate may explain only minor spatial variation in water quality, but this remains poorly understood because few studies have quantified the relative importance of competing drivers in African systems (Mbaka & Mwaniki, 2017 ). Sub-Saharan Africa exemplifies this knowledge gap. Despite rich freshwater biodiversity and high endemism across major river basins, the region is severely underrepresented in the global freshwater literature (Chapman et al., 2022 ; Pototsky & Cresswell, 2021 ). Nigeria, with over 300 documented fish species (Froese & Pauly, 2025 ), remains markedly underrepresented in the freshwater/water-quality literature. In our recent bibliometric assessment of Scopus-indexed literature, we identified only 17 relevant publications from Nigeria since 1996, representing about 1% of global output. The Upper Niger Basin has seen little systematic research on water quality, despite its two major dams having operated for more than 40 years and having widespread impacts on downstream communities and environments (Awojobi & Tetteh, 2017 ; Oyebande, 2001 ). Following impoundment, artisanal fishery catches declined sharply from roughly 28,000 to about 8,000 tons per year, alongside significant changes in sediment dynamics, reflecting the broader impacts of river regulation on tropical river fisheries and ecosystems (Arantes et al., 2019 ; Arthington et al., 2006 ; Chong et al., 2021 ; Ita et al., 1985 ; Olaosebikan & Raji, 2013 ; Oyebande & Odunuga, 2010 ). No systematic water quality assessment has been conducted since impoundment. Ongoing pressures are compounding past impacts. Satellite data indicate that agriculture now occupies 42–58% of the land draining into these reservoirs, with riparian forests disappearing at approximately 2.3% per year (Hansen et al., 2013 ; Zanaga et al., 2022 ). Urban areas have expanded by 340% since 2000, yet fewer than 15% of the population has access to proper wastewater treatment (Eteh et al., 2024 ; Lanrewaju, 2012 ). Invasive water hyacinth now covers 20–35% of reservoir margins during low-water periods, causing dissolved oxygen to plummet (Uwadiae et al., 2021 ). Climate projections warn of 15–25% greater rainfall variability and 1.5–2.5°C warming by 2050, likely intensifying both flood pollution pulses and dry-season concentration effects (Intergovernmental Panel on Climate Change (IPCC), 2022 ; Sylla et al., 2018 ). Fortunately, new tools make rigorous assessment feasible even with limited resources. Satellite imagery at 10-m resolution now maps land use across entire watersheds (Zanaga et al., 2022 ). Granger causality testing can distinguish whether climate drives water quality or merely correlates with it, using relatively short time series (Zolghadr-Asli et al., 2021 ). Changepoint detection identifies when systems shift between states, flagging degradation before it becomes irreversible (Lund et al., 2023 ; Scheffer et al., 2009 ). Water quality indices condense multiple parameters into single management-relevant scores (Chidiac et al., 2023 ; Sutadian et al., 2015 ; Uddin et al., 2021 ). Together, these tools can partition variance among dams, climate, and land use, and identify which parameters provide the earliest, most reliable warnings. We applied this analytical toolkit to monthly water-quality surveys conducted from April 2024 through March 2025 across the Kainji–Jebba cascade, encompassing one complete wet–dry cycle. Three questions guided the work. First, do dams create distinct water quality zones (as temperate theory predicts), or does hydrological connectivity maintain basin-wide uniformity? Second, what drives temporal patterns, climate, flood pulses, or land use? Third, which parameters detect problems earliest and most reliably? The answers have direct implications for how monitoring and management should be designed not only in Nigeria, but across the many data-poor tropical river systems where urgent conservation needs collide with limited research capacity (Chapman et al., 2022 zéquel et al., 2020 ). Beyond resolving ecological theory, this analytical framework identifies when and where monitoring effort yields the greatest return, thereby supporting more efficient allocation of limited monitoring resources in tropical reservoir systems. Materials and Methods Study Area We conducted this study in Nigeria’s Upper Niger River Basin, focusing on the Kainji–Jebba reservoir cascade (Fig. 1 ). The two hydroelectric dams, Kainji (10.5°N, 4.6°E; commissioned 1968) and Jebba (9.1°N, 4.8°E; commissioned 1985), form a cascade along the Niger River and have a combined installed capacity of approximately 1,338 MW (International Hydropower Association, n.d.). The region has a tropical climate with clearly defined wet (April–October) and dry (November–March) seasons, receiving mean annual rainfall of 1,100–1,400 mm (Nigerian Meteorological Agency, 2023 ). At each dam, we sampled three habitat types that collectively capture the full environmental gradient created by impoundment: riverine zones (upstream free-flowing sections, current 0.3–0.8 m/s); ecotonal zones (transitional areas between flowing river sections and reservoir backwaters); and lacustrine zones (deep reservoir cores, > 8 m depth, with negligible current; Wetzel, 2001 ). The six sampling sites were: Kainji Riverine (Garafini: 10.01555°N, 4.58584°E), Kainji Ecotonal (Awuru: 9.78447°N, 4.63231°E), Kainji Lacustrine (Yuna: 9.92093°N, 4.58925°E), Jebba Riverine (Gbajibo: 9.38147°N, 4.61988°E), Jebba Ecotonal (Juju Rock: 9.14158°N, 4.80761°E), and Jebba Lacustrine (Kokodi: 9.20615°N, 4.75372°E). Sites were georeferenced using a handheld GPS (Garmin eTrex 30x) and mapped using the sf package in R 4.5.1. Study Design and Temporal Scope We adopted a nested sampling design with monthly surveys from April 2024 through March 2025, providing 12 temporal replicates at each of six sites (n = 72 total samples). This design captures the full hydrological cycle while balancing detection of seasonal variability against logistical feasibility. Each site was sampled once per month between 06:00 and 10:00 hours to minimise diel variation. Sampling dates were synchronised across all sites within each month (all six sites sampled within 5–7 consecutive days) to isolate spatial from temporal effects. Water Quality Sampling and Laboratory Analysis Surface water (0.5 m depth) was collected using a Van Dorn horizontal water sampler (Wildco, 1.2 L) at lacustrine sites, accessed via motorised boat, and by grab sampling at riverine and ecotonal sites. Water temperature, pH, and electrical conductivity (EC) were measured in situ using a calibrated multi-parameter probe (Bluelab Combo Meter). Dissolved oxygen was measured using a portable meter (Milwaukee MW600 PRO). Water clarity was recorded as Secchi disk depth (m) and converted to a relative turbidity proxy (SD⁻¹) for index scoring, following Davies-Colley & Smith ( 2001 ); derived values are treated as a comparative proxy rather than instrument-measured NTU. Nitrate-nitrogen and phosphate-phosphorus were analysed by UV-Vis spectrophotometry at the National Institute for Freshwater Fisheries Research (NIFFR) Water Quality Laboratory, following APHA standard colourimetric procedures (APHA, 2017 ). Biochemical oxygen demand (BOD₅) was determined using a five-day incubation at 20°C. Water samples were collected in acid-washed polyethene bottles, stored at 4–6°C, and transported to the laboratory within six hours. All analyses were conducted in duplicate; coefficients of variation ranged from 2.1% to 8.7%. Meteorological Data and Land-Use Characterisation Meteorological data (air temperature, rainfall, relative humidity (RH), and atmospheric pressure) were obtained from validated stations operated by the Nigerian Meteorological Agency: the NIFFR meteorological station at New Bussa for Kainji sites, and the Jebba Mainstream Energy station for Jebba sites. Daily mean values corresponding to each sampling date were extracted. Land-cover composition was extracted from ESA WorldCover 2021 (10-m resolution; Zanaga et al., 2022 ) using 5-km radius buffers around each sampling point. We calculated the percentage cover of agricultural land, urban/built-up areas, and forest. Moran’s I indicated moderate spatial autocorrelation in agricultural cover (I = 0.42, p = 0.031), but not in urban extent or forest cover. Water Quality Index ( WQI ) Calculation We calculated WQI using a modified National Sanitation Foundation method (Brown et al., 1970 ), aggregating eight parameters: dissolved oxygen, pH, BOD₅, water temperature, NO₃-N, PO₄-P, turbidity, and electrical conductivity. Faecal coliforms were excluded because they primarily reflect public health risk rather than ecological integrity for fish. Each parameter was assigned a rescaled weight summing to 1.0 after redistributing the excluded faecal coliform weight (Table 1 ). Raw measurements were converted to dimensionless sub-indices (Q i ; 0–100) using NSF-based piecewise-linear transformation functions (Brown et al., 1970 ), and the composite WQI was calculated as WQI = Σ(Q i × W i ). A sensitivity analysis testing alternative weighting schemes (Primary, Proportional, and Equal) confirmed that spatial and temporal WQI patterns were insensitive to moderate variations in weights (Figures S1 a, S1b). WQI classes: Excellent (80–100), Good (65–79), Medium (50–64), Bad (25–49), Very Bad (0–24). Table 1 Rescaled parameter weights for the modified NSF Water Quality Index, with faecal coliform excluded and its weight redistributed proportionally across the eight retained parameters. Parameter Original Weight Rescaled Weight Rationale for Inclusion Dissolved Oxygen 0.170 0.200 Critical for fish respiration; hypoxia (< 4 mg/L) causes mortality pH 0.120 0.140 Regulates metabolic processes; extreme values (pH 9) are lethal BOD₅ 0.100 0.120 Indicates organic pollution; high BOD depletes oxygen Water Temperature 0.100 0.120 Tropical species are sensitive to thermal stress above 32°C. Nitrates (NO₃-N) 0.100 0.120 Eutrophication indicator; >5 mg/L triggers algal blooms Phosphate (PO₄-P) 0.080 0.100 Limiting nutrient; >0.1 mg/L promotes harmful algae Turbidity 0.080 0.100 Affects light penetration; high turbidity reduces prey detection and promotes the growth of harmful algae. Conductivity 0.100 0.120 Proxy for dissolved solids; >1000 µS/cm stresses freshwater fish. Total 0.85 1.00 Faecal coliform excluded (0.15 weight redistributed) Statistical Analyses Differences in WQI across dams, habitat types, and seasons were assessed using Kruskal–Wallis rank-sum tests, with post-hoc pairwise comparisons by Dunn’s test (Benjamini–Hochberg FDR correction; Benjamini & Hochberg, 1995 ). Multivariate water quality patterns were characterised using Bray–Curtis dissimilarity (Bray & Curtis, 1957 ) on log(x + 1)-transformed values, PERMANOVA (999 permutations; Anderson, 2017 ), and NMDS ordination (stress < 0.2), with 95% confidence ellipses showing group clustering. Variance partitioning was performed using partial redundancy analysis (RDA) with adjusted R² to quantify unique contributions of spatial (dam identity, habitat type), temporal (season, month), and land-use (agricultural, urban, and forest cover within 5-km buffers) predictors to water quality variation. Reported values (28.3%, 1.8%, 0.4%) represent unique adjusted R² fractions after partitioning shared variance among predictor sets (Legendre & Legendre, 2012 ). Analyses used the vegan package (v2.6-4) in R 4.5.1. Granger causality testing (Granger, 1969 ) assessed whether lagged climate variables (air temperature, rainfall, relative humidity, atmospheric pressure) linearly predicted water quality parameters beyond their own autoregressive dynamics. For each stratification level (2 dams × 3 habitats × 2 seasons = 12 groups), directional tests were conducted for each of the 4 climate drivers and 7 water quality parameters. Lags 1–3 were evaluated, with the optimal lag selected by AIC minimisation, yielding 336 directional tests in total. All series were assessed for stationarity using Augmented Dickey–Fuller tests (Dickey & Fuller, 1979 ); non-stationary series were first-differenced. P-values were adjusted for multiple comparisons using the Benjamini–Hochberg false discovery rate (FDR) procedure (α = 0.05; Benjamini & Hochberg, 1995 ). Analyses employed the lmtest and vars packages in R. Granger causality reflects predictive precedence in time series, not mechanistic causation. Six complementary changepoint procedures were used to identify abrupt shifts in the water-quality time series: Pruned Exact Linear Time (PELT), Binary Segmentation (BinSeg), variance-based detection, visual inspection of structural breaks, CUSUM screening, and Bayesian changepoint analysis. Consensus changepoints were defined as those detected by two or more methods within plus/minus one month. Detection rates were used to identify which parameters were most sensitive to perturbation and therefore best suited for sentinel monitoring. In addition, independent structural-change validation for key dissolved oxygen and WQI series was undertaken using CUSUM and fluctuation tests implemented via the strucchange framework, drawing on established cumulative-sums approaches for environmental time series and structural-break analysis (Castillo-Mateo, 2022 ; Regier et al., 2019 ; Killick & Eckley, 2014 ; Zeileis et al., 2003 ). Results Water Quality Characterisation Across 72 samples from April 2024 to March 2025, the overall mean WQI was 65.35 (median 65.20; SD 4.17), placing the system in the Medium/Fair quality class across multiple established classification frameworks (Brown et al., 1970 ; Balamurugan et al., 2020 ; Canadian Council of Ministers of the Environment, 2001 ). This consistent classification — neither pristine nor severely polluted — suggests a system under moderate but meaningful environmental stress (Table 2 ). WQI values ranged from a site mean low of 64.64 (Kainji) to a site mean high of 66.06 (Jebba), and from a monthly low of 59.95 (February) to a monthly high of 71.54 (May). Table 2 Descriptive statistics of the Water Quality Index (WQI) overall and by dam, habitat, and season. N = sample size; SD = standard deviation. Values for dam and habitat represent aggregated monthly means; seasonal values aggregate across all sites. Category Sub-Category N Mean WQI Median WQI SD WQI Overall 72 65.35 65.20 4.17 By Dam Jebba 36 66.06 65.66 3.84 Kainji 36 64.64 64.68 4.41 By Habitat Ecotonal 24 65.17 64.42 3.81 Lacustrine 24 65.78 65.82 4.19 Riverine 24 65.11 65.58 4.60 By Season Dry 48 65.60 65.66 4.41 Wet 24 64.85 64.36 3.67 WQI Variation by Dam and Habitat Type The mean WQI at Jebba Dam (66.06) was slightly higher than at Kainji (64.64). Although this difference is modest in absolute terms, it suggests that local factors beyond regional hydrology influence water quality. Kainji, located approximately 100 km upstream, is the first reservoir to receive the sediment-laden white flood, which carries high concentrations of pollutants and suspended material from upstream agricultural catchments (Adeogun et al., 2018 ). This seasonal influx likely depresses water quality disproportionately at Kainji, while Jebba may benefit from partial settling and dilution as water passes through the upper reservoir (Adegbehin et al., 2016 ; Adeogun et al., 2020 ; Nwobi-Okoye & Igboanugo, 2013 ). Habitat-level differences were smaller still but followed a consistent spatial pattern (Fig. 2 ). Lacustrine zones showed the highest mean WQI (65.78), followed by ecotonal (65.17) and riverine (65.11) habitats. This pattern reflects the primary modes of pollution transport: riverine and ecotonal zones receive the most concentrated upstream pollutant loads, while the deeper, low-flow lacustrine zones allow suspended material to settle and some soluble pollutants to disperse (Irenosen et al., 2012 ; Okafor et al., 2025 ). Taken together, spatial variation amounted to less than one WQI unit across habitats and 1.4 units across dams, small relative to the 11.6-unit range observed across months. Together, these patterns indicate that upstream runoff and terrestrial activities are the dominant sources of water quality degradation in the system. Seasonal and Monthly Variation The seasonal pattern was counterintuitive: mean WQI was higher in the dry season (65.60) than in the wet season (64.85), contrary to the expectation that increased rainfall and discharge would dilute pollutants and improve water quality. This reversal of the expected dilution effect reflects the dominant role of runoff-driven pollution. During the wet season, rather than diluting pollutants, increased rainfall mobilises sediments, nutrients, and organic matter from agricultural and urban catchments into the reservoirs, with the runoff effect outweighing any dilution benefit (Medupe & Letshwenyo, 2024 ; US EPA, 2025b). The monthly WQI trajectory provides a more detailed temporal signature of this pattern (Fig. 3 ). Values peaked in April and May (71.28 and 71.54, respectively), corresponding to the end of the dry season when terrestrial pollutant loads were minimal. A sharp decline occurred in June and July (68.71 and 63.11), coinciding with the onset of the rainy season and the first flush of accumulated pollutants from land surfaces (Wiranegara et al., 2023 ). Values remained depressed throughout the wet season, reflecting continued non-point source pollution inputs, before partially recovering in the early dry season. A second notable drop occurred in January–February (61.11 and 59.95), coinciding with the arrival of the black flood from the Fouta Djallon highlands via Guinea, Mali, and Niger (Nwobi-Okoye & Igboanugo, 2013 ). Although black-flood waters pass through extensive upstream floodplains where some sediment is deposited, the prolonged flow appears to transport an accumulation of dissolved solutes and organic pollutants from distant catchments, producing a secondary deterioration of water quality. February’s mean WQI of 59.95 represents the annual low point and falls below the Good/Medium threshold, making this period a critical risk window for aquatic health. A direct comparison of temporal and spatial variation reveals that temporal fluctuations in WQI far exceeded spatial differences. Monthly values ranged by 11.59 units, from 71.54 in May to 59.95 in February, whereas the difference between dams was only 1.42 units and the difference between the highest- and lowest-WQI habitats was just 0.67 units. These patterns indicate that annual hydrological dynamics dominate WQI variability, while dam- and habitat-level effects remain comparatively small. Bimodal Distribution of Water Quality The histogram of WQI values (Fig. 4 ) shows an apparent deviation from a single normal distribution, with two distinct peaks that define a bimodal distribution. The primary distribution, represented by the broad peak, is centred around a WQI of approximately 64–65 and reflects the system’s prevailing moderate water quality state. In contrast, a second, much narrower peak is centred around a WQI of approximately 71–72. This bimodal pattern is a critical finding that indicates the presence of two distinct seasonal water-quality regimes. The exceptionally high WQI values in the narrow peak correspond to the mean monthly WQI values for April (71.28) and May (71.54). The low standard deviations for these months, specifically 1.39 in April and 0.61 in May, explain the tight and concentrated nature of this distribution, indicating a period of high stability and homogeneity in water quality. The broader distribution corresponds to the WQI values for the remaining ten months of the year, which are consistently lower and more variable. This shift is likely governed by the annual hydrological cycle, with the onset of the wet season triggering a rapid transition from the brief high-quality regime to the dominant lower-quality state (Medupe & Letshwenyo, 2024 ). Causal Modelling The Granger causality analysis revealed no statistically significant linear relationships between climate variables and water quality parameters after FDR correction across all 336 stratified tests (Fig. 5 ). This null result is not a limitation of the analysis; it is a substantive finding about the nature of these systems. The stratified design was important because it assessed climate–water quality relationships within distinct dam, habitat, and seasonal contexts, reducing the risk of masking local dynamics through aggregation and misleading inference from pooled data (Chen et al., 2020 ; Wakefield & Lyons, 2010 ). Several clusters of unadjusted near-significant relationships (Table 3 ) were identified, most of which were concentrated in the Kainji Riverine habitat during the dry season. The most ecologically suggestive patterns included relative humidity predicting both pH and EC, and air temperature, windspeed, and pressure predicting NO₃-N. These dry-season signals are consistent with evaporation–concentration processes, whereby higher humidity slows evaporative water loss and increases dissolved ion concentrations, as well as with wind-driven resuspension of mineralised sediment nutrients during low-water conditions (Duong et al., 2023 ; Khurram et al., 2025 ; Hicks & McMahon, 2003 ; Pelikán & Marková, 2013 ; Rolls et al., 2012 ). However, none survived multiple-comparison correction, and these patterns should therefore be treated as hypotheses for future testing rather than established relationships. Table 3 Near-significant Granger causality relationships (unadjusted p < 0.05) from climate drivers to water quality parameters, ordered by evidence strength. No relationships remained significant after FDR correction (adjusted p < 0.05). Site Habitat Season Cause Effect P_Value Adj_P Kainji Riverine Dry RH pH 0.010 0.089 Kainji Riverine Dry Temp_Air NO3_N 0.004 0.081 Kainji Riverine Dry Windspeed NO3_N 0.007 0.081 Kainji Riverine Dry Pressure NO3_N 0.007 0.081 Kainji Riverine Dry RH EC 0.004 0.081 Jebba Riverine Wet Pressure Turbidity 0.023 0.473 Note : Although these relationships did not reach formal statistical significance after FDR correction (adjusted p-value < 0.05), they represent the strongest potential climate–water quality links identified in the analysis and are discussed as ecologically suggestive trends for future hypothesis testing. Variance Partitioning Variance partitioning revealed that temporal variation (season and month) was the dominant driver of the multivariate water quality structure, accounting for 28.3% of the variance (F = 14.32, p = 0.001) (Fig. 6 , Table S1 ). Spatial factors (dam location and habitat type) explained only 1.8% of variance but remained statistically significant (F = 1.85, p = 0.038), whereas anthropogenic land use showed no significant independent effect (0.4%, p = 0.303) (Fig. 6 , Table S1 ). NMDS ordination (stress = 0.059) visually confirmed this strong seasonal structuring, with wet-season samples distinctly separated from dry-season samples along axis 2, while dam and habitat identity were secondary organising factors (Fig. 7 ). Changepoint Detection and Early-Warning Indicators The comparative analysis of changepoint detection algorithms revealed substantial variation in their performance (Fig. 8 ). PELT and Binary Segmentation (BinSeg) achieved the highest detection rates (95.2% each), reflecting their strength in identifying multiple changepoints within complex environmental time series (Killick & Eckley, 2014 ; Killick et al., 2012 ). Notably, BinSeg matched the performance of the exact PELT algorithm despite being computationally less intensive, suggesting its suitability for large-scale future monitoring applications. In contrast, CUSUM achieved only 11.9% detection, and the Bayesian approach failed to detect any changepoints (0.0%), likely due to mismatches between their statistical assumptions and the non-stationary, multi-scaled nature of the data (Aminikhanghahi & Cook, 2017 ; Lund et al., 2023 ), as CUSUM tests are often optimised for detecting abrupt, single-point shifts in the mean from a known value (Killick & Eckley ( 2014 ); Lund et al. ( 2023 ). Across parameters, WQI, electrical conductivity, and relative humidity showed the highest consensus detection rates (71.4% each; Table 4 ), highlighting their value as early-warning indicators. WQI, in particular, offers an integrated measure of overall water quality (Aljanabi et al., 2021 ). Because it combines multiple variables, it is especially sensitive to environmental change: shifts in any component, such as DO, pH, or turbidity, can be reflected in detectable changepoints in the index (Chidiac et al., 2023 ; Gonçalves & Costa, 2011 ). EC is a reliable integrative signal of dissolved solids and responds quickly to pollution events and shifts in hydrological regime (US EPA, 2025a). The role of RH as a sentinel indicator is less direct, but it likely reflects its link to evaporation dynamics and the concentration of dissolved ions during low-flow dry periods. Water temperature (64.3%), air temperature (57.1%), and DO (54.8%) showed moderate detection rates and remain useful as supporting context indicators. Rainfall had the lowest detection rate (35.7%), which is consistent with its highly variable, event-driven nature and its poor fit with regime-shift detection methods designed to identify sustained changes in the mean (Lund et al., 2023 ; Reeves et al., 2007 ). Consensus changepoint analysis confirmed significant spatial heterogeneity in changepoint timing and strength across ecotonal, lacustrine, and riverine systems (Fig. 9 ). Riverine environments showed earlier and more frequent changepoints, reflecting their greater exposure to upstream disturbances; lacustrine systems showed distinct changepoints associated with internal biogeochemical cycles and stratification dynamics (Wetzel, 2001 ; Zhao et al., 2025 ). Structural change validation tests (CUSUM and Fluctuation tests) provided independent corroboration for DO and WQI changepoints, with CUSUM showing particularly strong significance for DO in ecotonal and riverine habitats, and both methods showing higher significance for lacustrine WQI at Kainji (Fig. 10 ). Table 4 Changepoint detection rates for water quality and environmental parameters across all sites and methods. Consensus is defined as detection by ≥ 2 methods within ± 1 month. Parameter Total Analyses Detections Detection Rate (%) Water Quality Index (WQI) 42 30 71.4 Electrical Conductivity (EC) 42 30 71.4 Relative Humidity (RH) 42 30 71.4 Water Temperature 42 27 64.3 Air Temperature 42 24 57.1 Dissolved Oxygen (DO) 42 23 54.8 Rainfall 42 15 35.7 Discussion This study demonstrates that the dynamics of water quality in the Kainji–Jebba reservoir cascade are shaped primarily by timing rather than location. Seasonal and monthly processes explained 28.3% of water quality variation, far outweighing spatial factors (1.8%) and direct land-use effects (0.4%). These results are consistent with flood pulse theory, in which seasonal hydrological cycles drive ecosystem dynamics more powerfully than the longitudinal gradients central to temperate models such as the River Continuum Concept (Junk et al., 1989 ; Tockner & Stanford, 2002 ; Vannote et al., 1980 ). High wet-season flows homogenise conditions across dams and habitat types, while dry-season concentration effects produce system-wide changes that override site-specific differences. This temporal structuring mirrors patterns reported in other tropical reservoirs where seasonal stratification and hydrological cycles overwhelm longitudinal gradients (Wiranegara et al., 2023 ; Zhang et al., 2024 ). Spatial Patterns: Weak Zonation in a Connected System The dominance of temporal over spatial factors indicates that, in this system, when you monitor matters more than where. The small but statistically significant spatial differences — slightly higher WQI at Jebba and marginally better quality in lacustrine habitats — are interpretable in terms of local hydrology. Kainji receives the first impact of the white flood from upstream agricultural areas, while Jebba benefits from partial settling through the upper reservoir (Adeogun et al., 2018 ; Adegbehin et al., 2016 ). Lacustrine zones’ marginally higher WQI reflects water residence time and settling of suspended material, consistent with the wider limnological literature on reservoir self-purification (Irenosen et al., 2012 ). However, these spatial differences are minor in magnitude and operationally less important than the seasonal signal for monitoring design. Land-Use Effects: Temporal Rather Than Spatial The minimal land-use variance (0.4%, p = 0.303) might seem to contradict the clear evidence throughout the data that agricultural runoff degrades water quality; the sharp WQI drops at the wet-season onset, the lowest values in riverine areas, and the bimodal WQI distribution all point to a land-use influence. The resolution to this apparent paradox lies in scale. Land use does not generate persistent spatial gradients where some sites are consistently worse than others; rather, its effects manifest as synchronous, system-wide runoff pulses during the wet season. When all sites deteriorate simultaneously, the variance attributable to land use appears in the temporal rather than the spatial component (Carvalho et al., 2019 ; Saturday et al., 2025 ; Shakeri Bostanabad et al., 2025 ). Agricultural pollution is therefore a temporal driver, not a spatial one, and it is captured within the 28.3% temporal fraction. Absence of Climate Signals: Nonlinearity and Buffering The failure of any climate–water quality relationship to survive FDR correction across 336 stratified Granger tests is a substantive finding rather than a methodological limitation. It suggests that at least three interacting factors override linear, direct climate forcing. First, internal biogeochemical feedbacks, microbial respiration, nutrient cycling, and phytoplankton dynamics can amplify, dampen, or redirect initial meteorological forcing, masking the original climate signal (Farrell et al., 2020 ; Kondowe et al., 2022 ). Second, dam operations and regulated releases impose an anthropogenically controlled hydrological regime that decouples water quality from meteorological drivers (Capon et al., 2021 ; Vivan et al., 2014 ). In such systems, direct linear links between climate trends and water-quality responses may be difficult to detect, particularly when causal pathways are nonlinear or mediated through complex hydrological interactions (Bonotto et al., 2022 ; Ye et al., 2022 ). Third, the relationships between climate and water quality may be nonlinear or threshold-governed in ways that linear Granger tests are not designed to detect (Shakeri Bostanabad et al., 2025 ; Ye et al., 2022 ). The cluster of near-significant dry-season signals at Kainji Riverine is suggestive. During low-flow periods when dilution capacity is reduced, the system may become temporarily more sensitive to meteorological forcing, providing a hypothesis worth testing with longer time series and nonlinear methods. Sentinel Indicators and Early-Warning Framework The high consensus detection rates for WQI, EC, and RH (71.4% each) provide a practical and evidence-based foundation for a tiered early-warning monitoring system. WQI’s composite nature makes it inherently sensitive to multi-parameter deterioration, translating complex data into a single actionable score that policymakers can act on (Mamat et al., 2023 ; Shaaban & Stevens, 2025 ). EC responds rapidly to influxes of dissolved solids from pollution events or altered hydrology (US EPA, 2025a), while RH’s sentinel value reflects its role in governing evaporation-concentration dynamics that amplify dissolved ion loads during dry periods. Together, these three parameters detect incipient regime shifts more reliably than any single conventional parameter. Temperature and DO retain importance as ecological context indicators, but are less sensitive to the specific stressors operating in this system. This interpretation is reinforced by the supplementary weighting sensitivity analysis, which showed that the temporal and spatial conclusions were unchanged across primary, proportional, and equal WQI weighting schemes. The consensus changepoint framework adds a confidence dimension to this early-warning system. High multi-method agreement (4–5 methods) indicates a high-confidence structural break warranting immediate investigation and management response. Lower agreement (2–3 methods) provides a preliminary alert that justifies intensified monitoring before a full response is triggered. This tiered structure transforms changepoint detection from a research tool into a practical risk-based management instrument (Lund et al., 2023 ). Management Implications Our results support three clear management priorities. First, monitoring programmes should intensify sampling during the two highest-risk windows: June–July at the onset of the rains, when WQI drops from 71.54 to 63.11, and January–February during the black flood, when WQI reaches its annual minimum of 59.95. Bi-weekly sampling during these windows, combined with monthly surveillance otherwise, would concentrate effort when it matters most (Carvalho et al., 2019 ; Shaaban & Stevens, 2025 ). Second, sentinel indicators (WQI, EC, RH) should be prioritised, with temperature and DO retained as core ecological context metrics. This configuration delivers early detection at a lower cost than continuously monitoring the full parameter set. Third, watershed management must address the sources of first-flush pollution loads. Interventions such as riparian buffer restoration, erosion control, and stormwater retention would directly target the wet-season runoff pulses responsible for the most significant water-quality deterioration. Given that dam release schedules also influence mixing and dilution, coordination between reservoir operators and water quality monitoring programmes during risk windows offers additional scope for mitigation (Chiromo et al., 2016 ). Limitations Two limitations deserve explicit acknowledgement. First, with only six sampling sites across two reservoirs, formal spatial hotspot detection using Getis–Ord Gi* statistics was not feasible, as this method requires at least ten spatial units for reliable results (Bivand et al., 2013 ). We therefore relied on descriptive summaries and visual mapping to characterise spatial patterns, which, while informative, lack the inferential rigour of formal spatial statistics. Second, one annual cycle is sufficient to characterise seasonal patterns, the dominant signal in tropical systems (Junk et al., 1989 ; Sharma et al., 2015 ), but insufficient to distinguish long-term regime shifts from interannual variability. Whether the observed patterns represent stable features of this system or reflect conditions specific to the 2024–2025 hydrological year cannot be determined from the present dataset. Longer-term monitoring with vertical profiling to capture stratification dynamics, and the application of nonlinear causal modelling, are priorities for future work. Conclusions This study reveals that water quality in the Upper Niger River Basin is primarily governed by seasonal hydrological cycles and watershed pollution, rather than by spatial differences between dams or habitats or by short-term climate variability. Monthly WQI values declined from a dry-season peak of 71.54 (May) to an annual minimum of 59.95 (February), driven by two distinct runoff-pulse mechanisms: the white flood (wet-season onset, June–July) and the black flood (January–February). No direct links between climate variables and water quality survived rigorous FDR correction, suggesting that human-modified hydrology and watershed pollution are the proximate drivers of change. The bimodal WQI distribution and the recurring seasonal pollution pulses suggest that human pressures are beginning to overwhelm the system’s natural recovery capacity. Variance partitioning confirms the primacy of temporal dynamics: when you monitor matters more than where. Changepoint analysis provides both the evidence base and the practical tools for a tiered early-warning system centred on three sentinel indicators: WQI, EC, and RH, deployed strategically during the two seasonal risk windows. Beyond the Niger Basin, this integrated approach, combining WQI, variance partitioning, Granger causality, and multi-method changepoint detection, offers a transferable template for other data-poor tropical river systems where urgent conservation needs and limited monitoring resources must be reconciled. Declarations Ethics approval: This research received ethical clearance from the University of KwaZulu-Natal (AREC/00003628/2021) and Federal University of Technology, Minna (Assignment No. 0000014EAU). Field collection was authorised by the Niger State Ministry of Agriculture (Ref: MEAR/FISH/4/VOL.III/690). Competing interests: The authors declare that they have no competing interests. Funding: This work was funded by a Tertiary Education Trust Fund (TETFund) doctoral fellowship to Mr A. Ibrahim, with additional support from the University of KwaZulu-Natal through a three-year tuition fee remission. Prof. M. Okpeku’s research allocation supports publication costs. Author Contribution Conceptualisation: AI, SOES, MO; Methodology: AI, SOES, MO; Data acquisition: AI; Formal analysis and investigation: AI; Writing — original draft: AI; Writing — review and editing: AI, SOES, DVRA, MO; Funding acquisition: AI, MO; Resources: MO; Supervision: MO, DVRA, SOES. All authors read and approved the final manuscript. Acknowledgement The first author acknowledges TETFund, Nigeria, for the doctoral fellowship awarded through the Federal University of Technology, Minna. We are grateful to the Federal University of Technology, Minna, for granting study leave and to the University of KwaZulu-Natal for institutional support. We thank Mainstream Energy Solutions Limited and NIFFR for providing meteorological data and facilitating field access. We are particularly grateful to Mallam Muhammad Jafar Baba for facilitating linkage with the Meteorological Department of Mainstream Energy Solutions, Jebba. We thank Elsevier for access to the Scopus database. Analyses were performed using R version 4.5.1. Reference management was conducted using EndNote. Data Availability The datasets generated during this study will be deposited in the Dryad Digital Repository upon manuscript acceptance. The DOI will be provided during the revision process. References Adegbehin, A., Yusuf, Y., Iguisi, E., & Zubairu, I. (2016). Reservoir inflow pattern and its effects on hydroelectric power generation at the Kainji Dam. WIT Transactions on Ecology and the Environment, 203, 233–244. https://doi.org/10.2495/EID160211 Adeogun, A. G., Ibitoye, B. A., Salami, A. W., & Ihagh, G. T. (2020). Sustainable management of erosion prone areas of upper watershed of Kainji hydropower dam, Nigeria. Journal of King Saud University—Engineering Sciences , 32(1), 5–10. https://doi.org/10.1016/j.jksues.2018.05.001 Adeogun, A., Sule, B., & Salami, W. (2018). Simulation of sediment yield at the upstream watershed of Jebba lake using SWAT. Malaysian Journal of Civil Engineering , 27. https://doi.org/10.11113/mjce.v27.15906 Albert, J. S., Destouni, G., Duke-Sylvester, S. M., Magurran, A. E., Oberdorff, T., Reis, R. E., Winemiller, K. O., & Ripple, W. J. (2021). Scientists’ warning to humanity on the freshwater biodiversity crisis. Ambio , 50(1), 85–94. https://doi.org/10.1007/s13280-020-01318-8 Aljanabi, Z. Z., Al-Obaidy, A. H. M. J., & Hassan, F. M. (2021). A brief review of water quality indices and their applications. IOP Conference Series: Earth and Environmental Science , 779(1), 012088. https://doi.org/10.1088/1755-1315/779/1/012088 Aminikhanghahi, S., & Cook, D. J. (2017). A survey of methods for time series change point detection. Knowledge and Information Systems , 51(2), 339–367. https://doi.org/10.1007/s10115-016-0987-z Anderson, M. J. (2017). Permutational multivariate analysis of variance (PERMANOVA). In StatsRef: Statistics Reference Online. Wiley. https://doi.org/10.1002/9781118445112.stat07841 APHA (2017). Standard Methods for the Examination of Water and Wastewater (23rd ed.). American Public Health Association. Arantes, C. C., Fitzgerald, D. B., Hoeinghaus, D. J., & Winemiller, K. O. (2019). Impacts of hydroelectric dams on fishes and fisheries in tropical rivers through the lens of functional traits. Current Opinion in Environmental Sustainability , 37, 28–40. https://doi.org/10.1016/j.cosust.2019.04.009 Arthington, A. H., Bunn, S. E., Poff, N. L., & Naiman, R. J. (2006). The challenge of providing environmental flow rules to sustain river ecosystems. Ecological Applications , 16(4), 1311–1318. https://doi.org/10.1890/1051-0761(2006)016[1311:TCOPEF]2.0.CO;2 Awojobi, O. N., & Tetteh, J. (2017). The impacts of climate change in Africa: A review of the scientific literature. Journal of International Academic Research for Multidisciplinary , 5(11), 39–52. Balamurugan, P., Kumar, P., Shankar, K., Nagavinothini, R., & Vijayasurya, K. (2020). Non-carcinogenic risk assessment of groundwater in southern part of Salem district in Tamilnadu, India. Journal of the Chilean Chemical Society , 65(1), 4697–4707. http://dx.doi.org/10.4067/S0717-97072020000104697 Barbarossa, V., Schmitt, R. J. P., Huijbregts, M. A. J., Zarfl, C., King, H., & Schipper, A. M. (2020). Impacts of current and future large dams on the geographic range connectivity of freshwater fish worldwide. Proceedings of the National Academy of Sciences of the United States of America , 117(7), 3648–3655. https://doi.org/10.1073/pnas.1912776117 Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society : Series B, 57(1), 289–300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x Beusen, A. H. W., Bouwman, A. F., Van Beek, L. P. H., Mogollón, J. M., & Middelburg, J. J. (2016). Global riverine N and P transport to ocean increased during the 20th century despite increased retention. Biogeosciences , 13(8), 2441–2451. https://doi.org/10.5194/bg-13-2441-2016 Bivand, R. S., Pebesma, E., & Gómez-Rubio, V. (2013). Applied Spatial Data Analysis with R (2nd ed.). Springer. https://doi.org/10.1007/978-1-4614-7618-4 Bonotto, G., Peterson, T., Fowler, K., & Western, A. (2022). Identifying causal interactions between groundwater and streamflow using convergent cross-mapping. Water Resources Research , 58. https://doi.org/10.1029/2021WR030231 Bray, J. R., & Curtis, J. T. (1957). An ordination of the upland forest communities of southern Wisconsin. Ecological Monographs , 27(4), 326–349. https://doi.org/10.2307/1942268 Brown, R. M., McClelland, N. I., Deininger, R. A., & O’Connor, M. F. (1970). A water quality index—do we dare? Water & Sewage Works, 117(10), 339–343. Canadian Council of Ministers of the Environment. (2001). CCME Water Quality Index 1.0, Technical Report. Available at: https://ccme.ca/en/res/wqimanualen.pdf (Accessed 23 October 2025). Capon, S. J., Stewart-Koster, B., & Bunn, S. E. (2021). Future of freshwater ecosystems in a 1.5°C warmer world. Frontiers in Environmental Science , 9, 784642. https://doi.org/10.3389/fenvs.2021.784642 Carvalho, L., Mackay, E. B., Cardoso, A. C., Baattrup Pedersen, A., Birk, S., Blackstock, K. L., Borics, G., Borja, A., Feld, C. K., Ferreira, M. T., Globevnik, L., Grizzetti, B., Hendry, S., Hering, D., Kelly, M., Langaas, S., Meissner, K., Panagopoulos, Y., Penning, E., … Solheim, A. L. (2019). Protecting and restoring Europe’s waters: An analysis of the future development needs of the Water Framework Directive. Science of the Total Environment , 658, 1228–1238. https://doi.org/10.1016/j.scitotenv.2018.12.255 Castillo-Mateo, J. (2022). Distribution-free changepoint detection tests based on the breaking of records. Environmental and Ecological Statistics , 29(3), 655–676. https://doi.org/10.1007/s10651-022-00539-2 Chapman, C. A., Abernathy, K., Chapman, L. J., Downs, C., Effiom, E. O., Gogarten, J. F., Golooba, M., Kalbitzer, U., Lawes, M. J., & Mekonnen, A. (2022). The future of sub-Saharan Africa’s biodiversity in the face of climate and societal change. Frontiers in Ecology and Evolution , 10, 790552. https://doi.org/10.3389/fevo.2022.790552 Chen, J., Wang, D., Li, G., Sun, Z., Wang, X., Zhang, X., & Zhang, W. (2020). Spatial and temporal heterogeneity analysis of water conservation in Beijing-Tianjin-Hebei urban agglomeration. GeoHealth , 4(8), e2020GH000248. https://doi.org/10.1029/2020GH000248 Chidiac, S., El Najjar, P., Ouaini, N., El Rayess, Y., & El Azzi, D. (2023). A comprehensive review of water quality indices: history, models, attempts and perspectives. Reviews in Environmental Science and Bio/Technology , 22(2), 349–395. https://doi.org/10.1007/s11157-023-09650-7 Chiromo, A., Amu, O., Otonglo, C., & Balderamma, F. (2016). Kainji and Jebba hydro power plant rehabilitation project. ESIA Summary, Private Sector Department (OPSD3), 1–28. Chong, X. Y., Vericat, D., Batalla, R. J., Teo, F. Y., Lee, K. S. P., & Gibbins, C. N. (2021). A review of the impacts of dams on the hydromorphology of tropical rivers. Science of the Total Environment , 794, 148686. https://doi.org/10.1016/j.scitotenv.2021.148686 Davies-Colley, R. J., & Smith, D. G. (2001). Turbidity, suspended sediment, and water clarity: A review. Journal of the American Water Resources Association , 37(5), 1085–1101. https://doi.org/10.1111/j.1752-1688.2001.tb03624.x Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association , 74(366), 427–431. https://doi.org/10.2307/2286348 Duong, Q. M., Garcia Vidales, D., & Schwartz, B. J. (2023). The effects of humidity on the electrical properties of semiconducting polymers. Applied Physics Letters , 123(20). https://doi.org/10.1063/5.0169905 Eteh, D. R., Egobueze, F. E., Paaru, M., Otutu, A., & Osondu, I. (2024). The impact of dam management and rainfall patterns on flooding in the Niger Delta. Discover Water , 4(1), 1–22. https://doi.org/10.1007/s43832-024-00185-8 FAO. (2020). Fishery and Aquaculture Country Profiles: Nigeria. https://www.fao.org/fishery/facp/NGA/en Farrell, K. J., Ward, N. K., Krinos, A. I., Hanson, P. C., Daneshmand, V., Figueiredo, R. J., & Carey, C. C. (2020). Ecosystem-scale nutrient cycling responses to increasing air temperatures vary with lake trophic state. Ecological Modelling , 430, 109134. https://doi.org/10.1016/j.ecolmodel.2020.109134 Faye, J. M., Ndao, S., & Sambou, L. E. V. (2023). Evolution of the physico-chemical quality of the water in the Manantali Dam Reservoir from 1989 to 2022. Journal of Geoscience and Environment Protection, 11(9), 351–366. https://doi.org/10.4236/gep.2023.119022 Froese, R., & Pauly, D. (2025). FishBase. World Wide Web electronic publication. https://www.fishbase.org/ (Accessed April 2025). Gonçalves, A. M., & Costa, M. (2011). Application of change-point detection to structural components of water quality variables. AIP Conference Proceedings , 1389, 1565–1568. https://doi.org/10.1063/1.3637928 Granger, C. W. J. (1969). Investigating causal relations by econometric models and cross-spectral methods. Econometrica , 37(3), 424–438. https://doi.org/10.2307/1912791 Grill, G., Lehner, B., Thieme, M., Geenen, B., Tickner, D., Antonelli, F., Babu, S., Borrelli, P., Cheng, L., Crochetiere, H., Ehalt Macedo, H., Filgueiras, R., Goichot, M., Higgins, J., Hogan, Z., Lip, B., Mcclain, M. E., Meng, J., Mulligan, M.,…Zarfl, C. (2019). Mapping the world’s free-flowing rivers. Nature , 569(7755), 215–221. https://doi.org/10.1038/s41586-019-1111-9 Hansen, M. C., Potapov, P. V., Moore, R., Hancher, M., Turubanova, S. A., Tyukavina, A., Thau, D., Stehman, S. V., Goetz, S. J., Loveland, T. R., Kommareddy, A., Egorov, A., Chini, L., Justice, C. O., & Townshend, J. R. G. (2013). High-resolution global maps of 21st-century forest cover change. Science , 342(6160), 850–853. https://doi.org/10.1126/science.1244693 Hicks, D., & McMahon, R. (2003). Temperature and relative humidity effects on water loss in Perna perna. Bulletin of Marine Science , 72, 135–150. International Hydropower Association. (n.d.). Mainstream Energy Solutions Limited. Available at: https://www.hydropower.org/our-members/mainstream-energy-solutions-limited (Accessed 13 May 2025). Intergovernmental Panel on Climate Change (IPCC) (2022). Climate Change 2022: Impacts, Adaptation and Vulnerability. Cambridge University Press. Available at: https://www.ipcc.ch/report/ar6/syr/ (Accessed: 10 July 2025). Irenosen, O. G., Festus, A. A., & Coolborn, A. F. (2012). Water quality assessment of the Owena multi-purpose dam, Ondo State. Journal of Environmental Protection , 3, 14–25. https://doi.org/10.4236/jep.2012.31003 Ita, E. O., Imevbore, A. M. A., & Odjugo, P. A. O. (1985). Impact of the Kainji Dam on fisheries in Nigeria. Nigerian Journal of Aquatic Sciences , 1(1), 15–27. Jane, S. F., Hansen, G. J. A., Kraemer, B. M., Leavitt, P. R., North, R. L., Pilla, R. M., Williamson, C. E., Woolway, R. I., Zwart, J. A., & Rose, K. C. (2021). Widespread deoxygenation of temperate lakes. Nature , 594, 66–70. https://doi.org/10.1038/s41586-021-03550-y Jézéquel, C., Tedesco, P. A., Darwall, W., Sversut Dias, M., Frederico, R. G., Hidalgo, M., Hugueny, B., Maldonado Ocampo, J., Martens, K., Ortega, H., Torrente Vilara, G., Zuanon, J., & Oberdorff, T. (2020). Freshwater fish diversity hotspots for conservation priorities in the Amazon Basin. Conservation Biology , 34, 956–965. https://doi.org/10.1111/cobi.13466 Jones, E. R., Van Vliet, M. T., Qadir, M., & Bierkens, M. F. (2021). Country-level and gridded estimates of wastewater production, collection, treatment and reuse. Earth System Science Data , 13(2), 237–254. https://doi.org/10.5194/essd-13-237-2021 Junk, W. J., Bayley, P. B., & Sparks, R. E. (1989). The flood pulse concept in river-floodplain systems. Canadian Special Publication of Fisheries and Aquatic Sciences , 106(1), 110–127. Khurram, M., Neuber, S., Sill, A., & Helm, C. A. (2025). Influence of relative humidities on highly electrically conductive polyelectrolyte multilayer films. Langmuir , 41(14), 9361–9368. https://doi.org/10.1021/acs.langmuir.4c05382 Killick, R., & Eckley, I. A. (2014). changepoint: An R package for changepoint analysis. Journal of Statistical Software , 58(3), 1–19. Killick, R., Fearnhead, P., & Eckley, I. A. (2012). Efficient detection of multiple changepoints within an oceanographic time series. Proceedings of ISI World Statistics Congress , 4137–4142. Kondowe, B. N., Masese, F. O., Raburu, P. O., Singini, W., Sitati, A., & Walumona, R. J. (2022). Seasonality in environmental conditions drive variation in plankton communities in a shallow tropical lake. Frontiers in Water , 4, 883767. https://doi.org/10.3389/frwa.2022.883767 Lanrewaju, F. (2012). Urbanisation, housing quality and environmental degeneration in Nigeria. Journal of Geography and Regional Planning , 5, 422–429. https://doi.org/10.5897/JGRP12.060 Legendre, P., & Legendre, L. (2012). Numerical Ecology (3rd ed.). Elsevier. Lund, R. B., Beaulieu, C., Killick, R., Lu, Q., & Shi, X. (2023). Good practices and common pitfalls in climate time series changepoint techniques. Journal of Climate , 36(23), 8041–8057. https://doi.org/10.1175/JCLI-D-22-0954.1 Mamat, N., Mohd Razali, S. F., & Hamzah, F. B. (2023). Enhancement of water quality index prediction using support vector machine with sensitivity analysis. Frontiers in Environmental Science , 10, 1061835. https://doi.org/10.3389/fenvs.2022.1061835 Mbaka, J. G., & Mwaniki, M. W. (2017). A critical review of the effect of water storage reservoirs on organic matter decomposition in rivers. Environmental Reviews , 25(2), 193–198. https://doi.org/10.1139/er-2016-0041 Medupe, M., & Letshwenyo, M. W. (2024). Investigation of self-purification and water quality index during dry and rainy seasons in the Khurumela Stream (Botswana). Journal of Ecohydraulics , 10(1), 1–18. https://doi.org/10.1080/24705357.2024.2363755 Nigerian Meteorological Agency. (2023). 2023 State of the Climate in Nigeria. https://www.nimet.gov.ng Nwobi-Okoye, C., & Igboanugo, A. (2013). Predicting water levels at Kainji Dam using artificial neural networks. Nigerian Journal of Technology , 32(1), 129–136. O’Mara, K., Venarsky, M., Stewart-Koster, B., McGregor, C., Schulz, C., Marshall, J., & Bunn, S. E. (2024). Hydrological connectivity and environment characteristics explain spatial variation in fish assemblages in a wet–dry tropical river. Hydrobiologia , 851(21), 5207–5221. https://doi.org/10.1007/s10750-024-05676-2 Oester, R., Keck, F., Moretti, M. S., Altermatt, F., Bruder, A., & Ferreira, V. (2025). A global synthesis on land-cover changes in watersheds shaping freshwater detrital food webs. Global Change Biology , 31(8), e70380. https://doi.org/10.1111/gcb.70380 Okafor, C. P., Nduka, J. K., Okafor, P. O., & Offor, C. C. (2025). Behavioural dynamics of water quality parameters in Agulu Lake Water, South-East Nigeria. Cleaner Water , 4, 100127. https://doi.org/10.1016/j.clwat.2025.100127 Olaosebikan, B. D., & Raji, A. (2013). Field Guide to Nigerian Freshwater Fishes (2nd ed.). Federal College of Freshwater Fisheries Technology, New Bussa. Oyebande, L. (2001). Streamflow regime change and ecological response in the Lake Chad Basin in Nigeria. IAHS Publications , 266, 101–112. Oyebande, L., & Odunuga, S. (2010). Climate change impact on water resources at the transboundary level in West Africa. Open Hydrology Journal, 4(1), 163–172. https://doi.org/10.2174/1874378101004010163 Pelikán, P., & Marková, J. (2013). Wind effect on water surface of water reservoirs. Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis , 61, 1823–1828. https://doi.org/10.11118/actaun201361061823 Poff, N. L., Allan, J. D., Bain, M. B., Karr, J. R., Prestegaard, K. L., Richter, B. D., … Stromberg, J. C. (1997). The natural flow regime: A paradigm for river conservation and restoration. BioScience , 47(11), 769–784. Pototsky, P. C., & Cresswell, W. (2021). Conservation research output in sub-Saharan Africa is increasing, but only in a few countries. Oryx , 55(6), 924–933. https://doi.org/10.1017/S0030605320000046 Reeves, J., Chen, J., Wang, X. L., Lund, R., & Lu, Q. Q. (2007). A review and comparison of changepoint detection techniques for climate data. Journal of Applied Meteorology and Climatology , 46(6), 900–915. https://doi.org/10.1175/JAM2493.1 Regier, P., Briceño, H., & Boyer, J. N. (2019). Analysing and comparing complex environmental time series using a cumulative sums approach. MethodsX , 6, 779–787. https://doi.org/10.1016/j.mex.2019.03.014 Reid, A. J., Carlson, A. K., Creed, I. F., Eliason, E. J., Gell, P. A., Johnson, P. T. J., Kidd, K. A., Maccormack, T. J., Olden, J. D., Ormerod, S. J., Smol, J. P., Taylor, W. W., Tockner, K., Vermaire, J. C., Dudgeon, D., & Cooke, S. J. (2019). Emerging threats and persistent conservation challenges for freshwater biodiversity. Biological Reviews , 94(3), 849–873. https://doi.org/10.1111/brv.12480 Rolls, R. J., Leigh, C., & Sheldon, F. (2012). Mechanistic effects of low-flow hydrology on riverine ecosystems. Freshwater Science , 31(4), 1163–1186. https://doi.org/10.1899/12-002.1 Saturday, A., Herrnegger, M., Kangume, S., & Stecher, G. (2025). Spatiotemporal variability of surface water quality in tropical agriculture-dominated catchments. Scientific Reports , 15(1), 42971. https://doi.org/10.1038/s41598-025-27066-x Scheffer, M., Bascompte, J., Brock, W. A., Brovkin, V., Carpenter, S. R., Dakos, V., Held, H., Van Nes, E. H., Rietkerk, M., & Sugihara, G. (2009). Early-warning signals for critical transitions. Nature , 461(7260), 53–59. https://doi.org/10.1038/nature08227 Shaaban, N. A., & Stevens, D. K. (2025). Transforming complex water quality monitoring data into water quality indices. Water Resources Management , 39(8), 3883–3899. https://doi.org/10.1007/s11269-025-04135-4 Shakeri Bostanabad, R., Amini, H., Fakheri, F., Lam, M., & Zahraie, B. (2025). Comparative analysis of correlation and causality inference in water quality problems. Scientific Reports , 15. https://doi.org/10.1038/s41598-025-85908-0 Sharma, S., Gray, D. K., Read, J. S., O’reilly, C. M., Schneider, P., Qudrat, A., Gries, C., Stefanoff, S., Hampton, S. E., & Hook, S. (2015). A global database of lake surface temperatures collected by in situ and satellite methods from 1985–2009. Scientific Data , 2, 1–19. https://doi.org/10.1038/sdata.2015.8 Sutadian, A. D., Muttil, N., Yilmaz, A. G., & Perera, B. J. C. (2015). Development of river water quality indices — a review. Environmental Monitoring and Assessment, 188(1), 58. https://doi.org/10.1007/s10661-015-5050-0 Sylla, M. B., Faye, A., Giorgi, F., Diedhiou, A., & Kunstmann, H. (2018). Projected heat stress under 1.5°C and 2°C global warming in West Africa. Earth’s Future , 6(7), 1029–1044. https://doi.org/10.1029/2018EF000873 Tickner, D., Opperman, J. J., Abell, R., Acreman, M., Arthington, A. H., Bunn, S. E., Cooke, S. J., Dalton, J., Darwall, W., Edwards, G., Harrison, I., Hughes, K., Jones, T., Leclère, D., Lynch, A. J., Leonard, P., Mcclain, M. E., Muruven, D., Olden, J. D., … Young, L. (2020). Bending the curve of global freshwater biodiversity loss. BioScience , 70, 330–342. https://doi.org/10.1093/biosci/biaa002 Tockner, K., & Stanford, J. A. (2002). Riverine flood plains: present state and future trends. Environmental Conservation , 29(3), 308–330. Uddin, M. G., Nash, S., & Olbert, A. I. (2021). A review of water quality index models and their use for assessing surface water quality. Ecological Indicators , 122, 107218. https://doi.org/10.1016/j.ecolind.2020.107218 United States Environmental Protection Agency. (2025a). Indicators: Conductivity. https://www.epa.gov/national-aquatic-resource-surveys/indicators-conductivity United States Environmental Protection Agency. (2025b). Soak Up the Rain: What’s the problem? https://www.epa.gov/soakuptherain/soak-rain-whats-problem Uwadiae, R., Daudu, E., & Lawal, M. (2021). Impact of Water Hyacinth (Eichhornia crassipes) Infestation on Water Quality and Growth of a Benthic Mollusc Pachymelania aurita Müller (Gastropoda: Melaniidae): Experimental Evaluation. Anchor University Journal of Science and Technology , 2(1). https://doi.org/10.4314/aujst.v2i1 Vannote, R. L., Minshall, G. W., Cummins, K. W., Sedell, J. R., & Cushing, C. E. (1980). The river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences , 37(1), 130–137. https://doi.org/10.1139/f80-017 Vivan, E., Yakubu, A., & Caleb, A. (2014). Effects of human activities on water quality of Lamingo Dam. International Journal of Interdisciplinary Research and Innovations , 2, 97–104. Wakefield, J., & Lyons, H. (2010). Spatial aggregation and the ecological fallacy. In Handbook of Spatial Statistics (pp. 541–558). CRC Press. Walsh, G., Pease, A. A., Woodford, D. J., Stiassny, M. L., Gaugris, J. Y., & South, J. (2022). Functional diversity of Afrotropical fish communities across river gradients in the Republic of Congo. Frontiers in Environmental Science , 10, 981960. https://doi.org/10.3389/fenvs.2022.981960 Wantzen, K. M., Junk, W. J., & Rothhaupt, K.-O. (2008). An extension of the flood pulse concept (FPC) for lakes. Hydrobiologia , 613(1), 151–170. https://doi.org/10.1007/s10750-008-9584-2 Ward, J., & Stanford, J. (1995). The serial discontinuity concept: extending the model to floodplain rivers. Regulated Rivers: Research & Management, 10(2–4), 159–168. https://doi.org/10.1002/RRR.3450100211 Wetzel, R. G. (2001). Limnology: Lake and River Ecosystems (3rd ed.). Academic Press. Wiranegara, P., Sunardi, S., Sumiarsa, D., & Juahir, H. (2023). Characteristics and changes in water quality based on climate and hydrology effects in the Cirata Reservoir. Water , 15(17), 3132. https://doi.org/10.3390/w15173132 WWF. (2020). Living Planet Report 2020 — Bending the Curve of Biodiversity Loss. WWF, Gland, Switzerland. Ye, L., Tan, L., Wu, X., Cai, Q., & Li, B. L. (2022). Nonlinear causal analysis reveals an effective water level regulation approach for phytoplankton blooms controlling in reservoirs. Science of the Total Environment , 806(Pt 4), 150948. https://doi.org/10.1016/j.scitotenv.2021.150948 Zanaga, D., Van De Kerchove, R., Daems, D., De Keersmaecker, W., Brockmann, C., Kirches, G., Wevers, J., Cartus, O., Santoro, M., Fritz, S., Lesiv, M., Herold, M., Tsendbazar, N.-E., Xu, P., Ramoino, F., & Arino, O. (2022). ESA WorldCover 10 m 2021 v200. Zenodo. https://doi.org/10.5281/zenodo.7254221 Zeileis, A., Leisch, F., Hornik, K., & Kleiber, C. (2003). strucchange: Testing for structural change in linear regression relationships. R News, 3(2), 7–13. Zhang, Y., Gao, X., Sun, B., & Liu, X. (2024). Hydrodynamics, diagenesis and hypoxia variably drive benthic oxygen flux in a river–reservoir system. Water Resources Research , 60(1), e2023WR035449. https://doi.org/10.1029/2023WR035449 Zhao, S., Hermans, M., Niemistö, J., Vesterinen, J., & Jilbert, T. (2025). Stratification controls the magnitude of in-lake phosphorus cycling. Hydrobiologia , 852(2), 359–376. https://doi.org/10.1007/s10750-024-05701-4 Zolghadr-Asli, B., Enayati, M., Pourghasemi, H. R., Naghdyzadegan Jahromi, M., & Tiefenbacher, J. P. (2021). Application of Granger-causality to study climate change impacts on depletion patterns of inland water bodies. Hydrological Sciences Journal , 66(12), 1767–1776. https://doi.org/10.1080/02626667.2021.1944633 Additional Declarations No competing interests reported. Supplementary Files GraphicalabstractFinal.tif SupplementaryFigureS1aCaption.docx SupplementaryFigureS1bvariancepartitioning.png SupplementaryFigureS1acorrelationheatmap.png SupplementaryFigureS1bCaption.docx SupplementaryTableS1.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 20 Apr, 2026 Reviews received at journal 20 Apr, 2026 Reviews received at journal 13 Apr, 2026 Reviewers agreed at journal 02 Apr, 2026 Reviewers agreed at journal 02 Apr, 2026 Reviews received at journal 02 Apr, 2026 Reviewers agreed at journal 01 Apr, 2026 Reviewers invited by journal 31 Mar, 2026 Editor assigned by journal 23 Mar, 2026 Submission checks completed at journal 23 Mar, 2026 First submitted to journal 13 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9118764","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":616519445,"identity":"85a4c641-6f51-41f0-99af-b57fe84d5333","order_by":0,"name":"Abdullahi Ibrahim","email":"","orcid":"","institution":"University of KwaZulu-Natal","correspondingAuthor":false,"prefix":"","firstName":"Abdullahi","middleName":"","lastName":"Ibrahim","suffix":""},{"id":616519446,"identity":"8004973c-b06e-449c-9565-92539845b413","order_by":1,"name":"Suleiman Omeiza Eku Sadiku","email":"","orcid":"","institution":"Federal University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Suleiman","middleName":"Omeiza Eku","lastName":"Sadiku","suffix":""},{"id":616519447,"identity":"cc1eadd9-0429-40a2-b29a-901b9e7eaae1","order_by":2,"name":"Deborah Vivienne Robertson-Andersson","email":"","orcid":"","institution":"Chrysalis Training and Skills Development","correspondingAuthor":false,"prefix":"","firstName":"Deborah","middleName":"Vivienne","lastName":"Robertson-Andersson","suffix":""},{"id":616519448,"identity":"521e6df5-a906-432b-bf31-98af2b8dbcbc","order_by":3,"name":"Moses Okpeku","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2ElEQVRIie3RPwuCQBzG8ecInI5cz0XfghHY1mtRHJxamxqE4FoOW+td1DtQDmzxzwto0cXWWoKWitqazrag++4ffj94AJ3uBzMpQGJIG7QvscSbnMb9iVu8ySOIvyBl1m75MdoVh67BYgpnlSpImYSjPT/OdmUcuchDEOErSE09q32RGjmDkWIANZncWi4jtyac4Z7CMBvVY8Ijey59t1wajPAUlCmuWKIKrU0lR9siN1iQhJQxxRWTzrKLmEtnWEQdO1+ntrNWXMHHhD56Ldp7Qp1Op/vbnsvBR9yDaqf2AAAAAElFTkSuQmCC","orcid":"","institution":"University of KwaZulu-Natal","correspondingAuthor":true,"prefix":"","firstName":"Moses","middleName":"","lastName":"Okpeku","suffix":""}],"badges":[],"createdAt":"2026-03-14 01:23:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9118764/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9118764/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106189292,"identity":"fd98444c-35a2-463c-9928-87648e6fae62","added_by":"auto","created_at":"2026-04-05 17:09:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":286036,"visible":true,"origin":"","legend":"\u003cp\u003eStudy area and sampling sites at Kainji and Jebba dams along the Upper Niger River, Nigeria. Site codes: K/J (Kainji/Jebba); R/E/L (Riverine/Ecotonal/Lacustrine). Monthly sampling conducted April 2024–March 2025 (n = 72 samples). Coordinate system: WGS 84.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/6a82a2463783d59964471ce5.png"},{"id":106189307,"identity":"db1f4c91-aea7-4acf-9cf8-ba953e61caee","added_by":"auto","created_at":"2026-04-05 17:09:40","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":273962,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eInteraction effects of dam, habitat, and season on WQI. \u003c/strong\u003eError bars represent the standard deviation of the mean.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/c91302e6d4b3b00fff442be3.png"},{"id":106189418,"identity":"0b9f7694-cdbc-415b-bbe8-5d43bcc064eb","added_by":"auto","created_at":"2026-04-05 17:09:54","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":447791,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMonthly mean WQI from April 2024 to March 2025\u003c/strong\u003e. Red error bars represent standard deviation. Peak quality (May, 71.54) and annual minimum (February, 59.95) are labelled.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/27d7713c82c16d33e1e4ae90.png"},{"id":106189306,"identity":"8c71bf09-a191-43c1-9ca3-16e2c2b3150c","added_by":"auto","created_at":"2026-04-05 17:09:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":206628,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistribution of WQI values across all 72 samples\u003c/strong\u003e. \u003cstrong\u003eThe bimodal structure (a narrow, high-quality peak in green and a broader distribution in red) reflects two distinct seasonal water-quality regimes.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/003731ffc91065bcbdcbe848.png"},{"id":106189303,"identity":"d079859c-57b4-40b4-97ac-c1ab148f4a97","added_by":"auto","created_at":"2026-04-05 17:09:38","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":361049,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHeatmap of FDR-adjusted p-values for Granger causality tests from climate drivers to water quality parameters at Jebba and Kainji\u003c/strong\u003e. All adjusted p-values exceed the 0.05 significance threshold.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/494b37ecae6124cb641e1e45.png"},{"id":106189340,"identity":"62684d62-f115-477b-932d-9b4cd72f9224","added_by":"auto","created_at":"2026-04-05 17:09:46","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":220722,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVariance partitioning of water quality (adjusted R² from conditional RDA, n = 72). Temporal factors dominate. *** p \u0026lt; 0.001; * p \u0026lt; 0.05; ns = not significant.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/aeff0146bc26ade7b9502b83.png"},{"id":106189313,"identity":"f031892a-6351-4589-b61b-d0c2ad1f3616","added_by":"auto","created_at":"2026-04-05 17:09:41","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":78683,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNMDS ordination of water quality parameters by season and reservoir (stress = 0.059). Points represent individual samples, coloured by season and shaped by reservoir. Ellipses represent 95% confidence intervals.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/11e86e3414498f5983c4f8be.png"},{"id":106189338,"identity":"69ecf553-e181-426c-9de3-63868ed93080","added_by":"auto","created_at":"2026-04-05 17:09:45","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":210321,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChange-point detection rates by (A) statistical method and (B) environmental parameter\u003c/strong\u003e. PELT and BinSeg achieved the highest detection rates (95%); WQI, EC, and RH were most frequently identified with changepoints (71.4% each). Numbers in parentheses indicate total analyses per category (n = 42)\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/ee3b54de82b58c4b4ecf7a70.png"},{"id":106189293,"identity":"b4c92319-0353-46a7-a71a-92264e34cdaa","added_by":"auto","created_at":"2026-04-05 17:09:37","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":284111,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConsensus changepoint analysis showing locations and confidence levels of statistically significant regime shifts across habitats. Colour intensity indicates consensus strength (2–5 methods agreeing).\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/d59ccea0b75e8dddc561bfc1.png"},{"id":106189299,"identity":"5babd957-78d8-4b95-86ea-29571786d722","added_by":"auto","created_at":"2026-04-05 17:09:37","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":265315,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStructural change test results (CUSUM and Fluctuation tests) for dissolved oxygen and WQI at Jebba and Kainji across three habitats. Bars above the dashed line indicate significance at p \u0026lt; 0.05 (-log10 scale).\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/b08ec02e9f39cffcdc3edbbb.png"},{"id":106405953,"identity":"e63ebf5b-8e57-404f-8bb5-abce1564fca3","added_by":"auto","created_at":"2026-04-08 09:29:05","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4315776,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/3613a98a-b0cb-421d-b13f-cf980af6f448.pdf"},{"id":106189339,"identity":"e9f5f78f-5893-4a02-afb7-fa6759d58360","added_by":"auto","created_at":"2026-04-05 17:09:45","extension":"tif","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1402732,"visible":true,"origin":"","legend":"","description":"","filename":"GraphicalabstractFinal.tif","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/d8b749feb0d65332bbc4dfc9.tif"},{"id":106189337,"identity":"fe5a7d3f-02f9-4bff-996f-ae73b1f831de","added_by":"auto","created_at":"2026-04-05 17:09:45","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":13862,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigureS1aCaption.docx","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/154509c7a36c6000b072da65.docx"},{"id":106189403,"identity":"8114872b-7d27-4511-803a-d14e6f1d67e5","added_by":"auto","created_at":"2026-04-05 17:09:53","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":36901,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigureS1bvariancepartitioning.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/8d0e598868943a5c09ab21ce.png"},{"id":106403753,"identity":"9a623ac3-be4d-4fb6-a3de-da4cd3e218c2","added_by":"auto","created_at":"2026-04-08 09:14:55","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":63680,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigureS1acorrelationheatmap.png","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/272feea066569a62f51843f6.png"},{"id":106189336,"identity":"580b061e-f697-4c02-b91a-d8a891c6a806","added_by":"auto","created_at":"2026-04-05 17:09:45","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":13968,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigureS1bCaption.docx","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/c48d1bfba81d9e0feeb73ac2.docx"},{"id":106189335,"identity":"590089a8-805d-4174-9aa8-8e1becc3e3dc","added_by":"auto","created_at":"2026-04-05 17:09:45","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":15028,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9118764/v1/26016ca94faa1f8ee702a890.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Temporal dynamics override spatial gradients in Afrotropical reservoir water quality","fulltext":[{"header":"Introduction","content":"\u003cp\u003eFreshwater ecosystems harbour disproportionate biodiversity relative to their spatial extent. Covering less than 1% of Earth\u0026rsquo;s surface, they support over 10% of all animal species and more than half of all fish species (Albert et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Reid et al., \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). However, these systems are in rapid decline. Freshwater vertebrate populations have crashed by 84% since 1970, outpacing losses in forests and oceans, with nearly a third of freshwater fish now threatened with extinction (Tickner et al., \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; WWF, \u003cspan citationid=\"CR119\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Dams are a major culprit: 58,000 large dams now fragment 60% of the world\u0026rsquo;s rivers, disrupting the flow patterns on which aquatic life depends (Barbarossa et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Grill et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Understanding how dams, climate, and land-use change interact to degrade water quality is therefore crucial, especially in tropical regions where biodiversity is highest, and data are scarcest (Faye et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2023\u003c/span\u003ez\u0026eacute;quel et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eMost of what we know about dam impacts comes from studies of temperate rivers. The River Continuum Concept predicts that river chemistry and biology change gradually downstream, with dams creating distinct zones of disruption (Vannote et al., \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e1980\u003c/span\u003e; Ward \u0026amp; Stanford, \u003cspan citationid=\"CR116\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Temperature and rainfall drive seasonal patterns by influencing oxygen levels, nutrient processing, and thermal stratification (Jane et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Poff et al., \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e1997\u003c/span\u003e). Tropical rivers, however, work differently. The Flood Pulse Concept posits that seasonal flooding, rather than downstream gradients, is the primary organising force (Junk et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e1989\u003c/span\u003e). Annual floods connect rivers to floodplains, driving nutrient cycling and fish reproduction in ways dams supposedly interrupt (Tockner \u0026amp; Stanford, \u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Wantzen et al., \u003cspan citationid=\"CR115\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Even so, dammed tropical rivers often maintain surprising connectivity through spill events and fish movement, potentially buffering the effects of fragmentation (O\u0026rsquo;Mara et al., \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Walsh et al., \u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Whether temperate or tropical models better predict water quality in regulated African rivers remains largely untested.\u003c/p\u003e \u003cp\u003eA third set of drivers, direct anthropogenic modification of watersheds, may supersede both. Tropical deforestation removes riparian buffers that filter runoff (FAO, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Oester et al., \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Fertiliser use has nearly doubled global riverine nitrogen transport to the ocean over recent decades (Beusen et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), while rapid urban growth generates large volumes of largely untreated wastewater (Jones et al., \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). If these anthropogenic pressures dominate, then dam location and climate may explain only minor spatial variation in water quality, but this remains poorly understood because few studies have quantified the relative importance of competing drivers in African systems (Mbaka \u0026amp; Mwaniki, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSub-Saharan Africa exemplifies this knowledge gap. Despite rich freshwater biodiversity and high endemism across major river basins, the region is severely underrepresented in the global freshwater literature (Chapman et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Pototsky \u0026amp; Cresswell, \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Nigeria, with over 300 documented fish species (Froese \u0026amp; Pauly, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), remains markedly underrepresented in the freshwater/water-quality literature. In our recent bibliometric assessment of Scopus-indexed literature, we identified only 17 relevant publications from Nigeria since 1996, representing about 1% of global output. The Upper Niger Basin has seen little systematic research on water quality, despite its two major dams having operated for more than 40 years and having widespread impacts on downstream communities and environments (Awojobi \u0026amp; Tetteh, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Oyebande, \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). Following impoundment, artisanal fishery catches declined sharply from roughly 28,000 to about 8,000 tons per year, alongside significant changes in sediment dynamics, reflecting the broader impacts of river regulation on tropical river fisheries and ecosystems (Arantes et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Arthington et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Chong et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Ita et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e1985\u003c/span\u003e; Olaosebikan \u0026amp; Raji, \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Oyebande \u0026amp; Odunuga, \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). No systematic water quality assessment has been conducted since impoundment.\u003c/p\u003e \u003cp\u003eOngoing pressures are compounding past impacts. Satellite data indicate that agriculture now occupies 42\u0026ndash;58% of the land draining into these reservoirs, with riparian forests disappearing at approximately 2.3% per year (Hansen et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Zanaga et al., \u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Urban areas have expanded by 340% since 2000, yet fewer than 15% of the population has access to proper wastewater treatment (Eteh et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Lanrewaju, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Invasive water hyacinth now covers 20\u0026ndash;35% of reservoir margins during low-water periods, causing dissolved oxygen to plummet (Uwadiae et al., \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Climate projections warn of 15\u0026ndash;25% greater rainfall variability and 1.5\u0026ndash;2.5\u0026deg;C warming by 2050, likely intensifying both flood pollution pulses and dry-season concentration effects (Intergovernmental Panel on Climate Change (IPCC), \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Sylla et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFortunately, new tools make rigorous assessment feasible even with limited resources. Satellite imagery at 10-m resolution now maps land use across entire watersheds (Zanaga et al., \u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Granger causality testing can distinguish whether climate drives water quality or merely correlates with it, using relatively short time series (Zolghadr-Asli et al., \u003cspan citationid=\"CR127\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Changepoint detection identifies when systems shift between states, flagging degradation before it becomes irreversible (Lund et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Scheffer et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Water quality indices condense multiple parameters into single management-relevant scores (Chidiac et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Sutadian et al., \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Uddin et al., \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Together, these tools can partition variance among dams, climate, and land use, and identify which parameters provide the earliest, most reliable warnings.\u003c/p\u003e \u003cp\u003eWe applied this analytical toolkit to monthly water-quality surveys conducted from April 2024 through March 2025 across the Kainji\u0026ndash;Jebba cascade, encompassing one complete wet\u0026ndash;dry cycle. Three questions guided the work. First, do dams create distinct water quality zones (as temperate theory predicts), or does hydrological connectivity maintain basin-wide uniformity? Second, what drives temporal patterns, climate, flood pulses, or land use? Third, which parameters detect problems earliest and most reliably? The answers have direct implications for how monitoring and management should be designed not only in Nigeria, but across the many data-poor tropical river systems where urgent conservation needs collide with limited research capacity (Chapman et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003ez\u0026eacute;quel et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Beyond resolving ecological theory, this analytical framework identifies when and where monitoring effort yields the greatest return, thereby supporting more efficient allocation of limited monitoring resources in tropical reservoir systems.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Area\u003c/h2\u003e \u003cp\u003eWe conducted this study in Nigeria\u0026rsquo;s Upper Niger River Basin, focusing on the Kainji\u0026ndash;Jebba reservoir cascade (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The two hydroelectric dams, Kainji (10.5\u0026deg;N, 4.6\u0026deg;E; commissioned 1968) and Jebba (9.1\u0026deg;N, 4.8\u0026deg;E; commissioned 1985), form a cascade along the Niger River and have a combined installed capacity of approximately 1,338 MW (International Hydropower Association, n.d.). The region has a tropical climate with clearly defined wet (April\u0026ndash;October) and dry (November\u0026ndash;March) seasons, receiving mean annual rainfall of 1,100\u0026ndash;1,400 mm (Nigerian Meteorological Agency, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAt each dam, we sampled three habitat types that collectively capture the full environmental gradient created by impoundment: riverine zones (upstream free-flowing sections, current 0.3\u0026ndash;0.8 m/s); ecotonal zones (transitional areas between flowing river sections and reservoir backwaters); and lacustrine zones (deep reservoir cores, \u0026gt;\u0026thinsp;8 m depth, with negligible current; Wetzel, \u003cspan citationid=\"CR117\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). The six sampling sites were: Kainji Riverine (Garafini: 10.01555\u0026deg;N, 4.58584\u0026deg;E), Kainji Ecotonal (Awuru: 9.78447\u0026deg;N, 4.63231\u0026deg;E), Kainji Lacustrine (Yuna: 9.92093\u0026deg;N, 4.58925\u0026deg;E), Jebba Riverine (Gbajibo: 9.38147\u0026deg;N, 4.61988\u0026deg;E), Jebba Ecotonal (Juju Rock: 9.14158\u0026deg;N, 4.80761\u0026deg;E), and Jebba Lacustrine (Kokodi: 9.20615\u0026deg;N, 4.75372\u0026deg;E). Sites were georeferenced using a handheld GPS (Garmin eTrex 30x) and mapped using the sf package in R 4.5.1.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStudy Design and Temporal Scope\u003c/h3\u003e\n\u003cp\u003eWe adopted a nested sampling design with monthly surveys from April 2024 through March 2025, providing 12 temporal replicates at each of six sites (n\u0026thinsp;=\u0026thinsp;72 total samples). This design captures the full hydrological cycle while balancing detection of seasonal variability against logistical feasibility. Each site was sampled once per month between 06:00 and 10:00 hours to minimise diel variation. Sampling dates were synchronised across all sites within each month (all six sites sampled within 5\u0026ndash;7 consecutive days) to isolate spatial from temporal effects.\u003c/p\u003e\n\u003ch3\u003eWater Quality Sampling and Laboratory Analysis\u003c/h3\u003e\n\u003cp\u003eSurface water (0.5 m depth) was collected using a Van Dorn horizontal water sampler (Wildco, 1.2 L) at lacustrine sites, accessed via motorised boat, and by grab sampling at riverine and ecotonal sites. Water temperature, pH, and electrical conductivity (EC) were measured in situ using a calibrated multi-parameter probe (Bluelab Combo Meter). Dissolved oxygen was measured using a portable meter (Milwaukee MW600 PRO). Water clarity was recorded as Secchi disk depth (m) and converted to a relative turbidity proxy (SD⁻\u0026sup1;) for index scoring, following Davies-Colley \u0026amp; Smith (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2001\u003c/span\u003e); derived values are treated as a comparative proxy rather than instrument-measured NTU. Nitrate-nitrogen and phosphate-phosphorus were analysed by UV-Vis spectrophotometry at the National Institute for Freshwater Fisheries Research (NIFFR) Water Quality Laboratory, following APHA standard colourimetric procedures (APHA, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Biochemical oxygen demand (BOD₅) was determined using a five-day incubation at 20\u0026deg;C. Water samples were collected in acid-washed polyethene bottles, stored at 4\u0026ndash;6\u0026deg;C, and transported to the laboratory within six hours. All analyses were conducted in duplicate; coefficients of variation ranged from 2.1% to 8.7%.\u003c/p\u003e\n\u003ch3\u003eMeteorological Data and Land-Use Characterisation\u003c/h3\u003e\n\u003cp\u003eMeteorological data (air temperature, rainfall, relative humidity (RH), and atmospheric pressure) were obtained from validated stations operated by the Nigerian Meteorological Agency: the NIFFR meteorological station at New Bussa for Kainji sites, and the Jebba Mainstream Energy station for Jebba sites. Daily mean values corresponding to each sampling date were extracted. Land-cover composition was extracted from ESA WorldCover 2021 (10-m resolution; Zanaga et al., \u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) using 5-km radius buffers around each sampling point. We calculated the percentage cover of agricultural land, urban/built-up areas, and forest. Moran\u0026rsquo;s I indicated moderate spatial autocorrelation in agricultural cover (I\u0026thinsp;=\u0026thinsp;0.42, p\u0026thinsp;=\u0026thinsp;0.031), but not in urban extent or forest cover.\u003c/p\u003e \u003cp\u003e \u003cb\u003eWater Quality Index (\u003c/b\u003eWQI\u003cb\u003e) Calculation\u003c/b\u003e\u003c/p\u003e \u003cp\u003eWe calculated WQI using a modified National Sanitation Foundation method (Brown et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e1970\u003c/span\u003e), aggregating eight parameters: dissolved oxygen, pH, BOD₅, water temperature, NO₃-N, PO₄-P, turbidity, and electrical conductivity. Faecal coliforms were excluded because they primarily reflect public health risk rather than ecological integrity for fish. Each parameter was assigned a rescaled weight summing to 1.0 after redistributing the excluded faecal coliform weight (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Raw measurements were converted to dimensionless sub-indices (Q\u003csub\u003ei\u003c/sub\u003e; 0\u0026ndash;100) using NSF-based piecewise-linear transformation functions (Brown et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e1970\u003c/span\u003e), and the composite WQI was calculated as WQI\u0026thinsp;=\u0026thinsp;Σ(Q\u003csub\u003ei\u003c/sub\u003e \u0026times; W\u003csub\u003ei\u003c/sub\u003e). A sensitivity analysis testing alternative weighting schemes (Primary, Proportional, and Equal) confirmed that spatial and temporal WQI patterns were insensitive to moderate variations in weights (Figures \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea, S1b). WQI classes: Excellent (80\u0026ndash;100), Good (65\u0026ndash;79), Medium (50\u0026ndash;64), Bad (25\u0026ndash;49), Very Bad (0\u0026ndash;24).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRescaled parameter weights for the modified NSF Water Quality Index, with faecal coliform excluded and its weight redistributed proportionally across the eight retained parameters.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParameter\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOriginal Weight\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRescaled Weight\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRationale for Inclusion\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDissolved\u003c/p\u003e \u003cp\u003eOxygen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.170\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCritical for fish respiration; hypoxia (\u0026lt;\u0026thinsp;4 mg/L) causes mortality\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.140\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRegulates metabolic processes; extreme values (pH\u0026thinsp;\u0026lt;\u0026thinsp;6 or \u0026gt;\u0026thinsp;9) are lethal\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBOD₅\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eIndicates organic pollution; high BOD depletes oxygen\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater\u003c/p\u003e \u003cp\u003eTemperature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTropical species are sensitive to thermal stress above 32\u0026deg;C.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNitrates\u003c/p\u003e \u003cp\u003e(NO₃-N)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEutrophication indicator; \u0026gt;5 mg/L triggers algal blooms\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhosphate\u003c/p\u003e \u003cp\u003e(PO₄-P)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.080\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLimiting nutrient; \u0026gt;0.1 mg/L promotes harmful algae\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTurbidity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.080\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAffects light penetration; high turbidity reduces prey detection and promotes the growth of harmful algae.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eConductivity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.120\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eProxy for dissolved solids; \u0026gt;1000 \u0026micro;S/cm stresses freshwater fish.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eTotal\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e0.85\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e1.00\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eFaecal coliform excluded (0.15 weight redistributed)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e\n\u003ch3\u003eStatistical Analyses\u003c/h3\u003e\n\u003cp\u003eDifferences in WQI across dams, habitat types, and seasons were assessed using Kruskal\u0026ndash;Wallis rank-sum tests, with post-hoc pairwise comparisons by Dunn\u0026rsquo;s test (Benjamini\u0026ndash;Hochberg FDR correction; Benjamini \u0026amp; Hochberg, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Multivariate water quality patterns were characterised using Bray\u0026ndash;Curtis dissimilarity (Bray \u0026amp; Curtis, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1957\u003c/span\u003e) on log(x\u0026thinsp;+\u0026thinsp;1)-transformed values, PERMANOVA (999 permutations; Anderson, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), and NMDS ordination (stress\u0026thinsp;\u0026lt;\u0026thinsp;0.2), with 95% confidence ellipses showing group clustering.\u003c/p\u003e \u003cp\u003eVariance partitioning was performed using partial redundancy analysis (RDA) with adjusted R\u0026sup2; to quantify unique contributions of spatial (dam identity, habitat type), temporal (season, month), and land-use (agricultural, urban, and forest cover within 5-km buffers) predictors to water quality variation. Reported values (28.3%, 1.8%, 0.4%) represent unique adjusted R\u0026sup2; fractions after partitioning shared variance among predictor sets (Legendre \u0026amp; Legendre, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Analyses used the vegan package (v2.6-4) in R 4.5.1.\u003c/p\u003e \u003cp\u003eGranger causality testing (Granger, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e1969\u003c/span\u003e) assessed whether lagged climate variables (air temperature, rainfall, relative humidity, atmospheric pressure) linearly predicted water quality parameters beyond their own autoregressive dynamics. For each stratification level (2 dams \u0026times; 3 habitats \u0026times; 2 seasons\u0026thinsp;=\u0026thinsp;12 groups), directional tests were conducted for each of the 4 climate drivers and 7 water quality parameters. Lags 1\u0026ndash;3 were evaluated, with the optimal lag selected by AIC minimisation, yielding 336 directional tests in total. All series were assessed for stationarity using Augmented Dickey\u0026ndash;Fuller tests (Dickey \u0026amp; Fuller, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e1979\u003c/span\u003e); non-stationary series were first-differenced. P-values were adjusted for multiple comparisons using the Benjamini\u0026ndash;Hochberg false discovery rate (FDR) procedure (α\u0026thinsp;=\u0026thinsp;0.05; Benjamini \u0026amp; Hochberg, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). Analyses employed the lmtest and vars packages in R. Granger causality reflects predictive precedence in time series, not mechanistic causation.\u003c/p\u003e \u003cp\u003eSix complementary changepoint procedures were used to identify abrupt shifts in the water-quality time series: Pruned Exact Linear Time (PELT), Binary Segmentation (BinSeg), variance-based detection, visual inspection of structural breaks, CUSUM screening, and Bayesian changepoint analysis. Consensus changepoints were defined as those detected by two or more methods within plus/minus one month. Detection rates were used to identify which parameters were most sensitive to perturbation and therefore best suited for sentinel monitoring. In addition, independent structural-change validation for key dissolved oxygen and WQI series was undertaken using CUSUM and fluctuation tests implemented via the strucchange framework, drawing on established cumulative-sums approaches for environmental time series and structural-break analysis (Castillo-Mateo, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Regier et al., \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Killick \u0026amp; Eckley, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Zeileis et al., \u003cspan citationid=\"CR123\" class=\"CitationRef\"\u003e2003\u003c/span\u003e).\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eWater Quality Characterisation\u003c/h2\u003e \u003cp\u003eAcross 72 samples from April 2024 to March 2025, the overall mean WQI was 65.35 (median 65.20; SD 4.17), placing the system in the Medium/Fair quality class across multiple established classification frameworks (Brown et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e1970\u003c/span\u003e; Balamurugan et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Canadian Council of Ministers of the Environment, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). This consistent classification \u0026mdash; neither pristine nor severely polluted \u0026mdash; suggests a system under moderate but meaningful environmental stress (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). WQI values ranged from a site mean low of 64.64 (Kainji) to a site mean high of 66.06 (Jebba), and from a monthly low of 59.95 (February) to a monthly high of 71.54 (May).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescriptive statistics of the Water Quality Index (WQI) overall and by dam, habitat, and season. N\u0026thinsp;=\u0026thinsp;sample size; SD\u0026thinsp;=\u0026thinsp;standard deviation. Values for dam and habitat represent aggregated monthly means; seasonal values aggregate across all sites.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCategory\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSub-Category\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean WQI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMedian WQI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSD WQI\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOverall\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e65.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e65.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4.17\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBy Dam\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eJebba\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e66.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e65.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.84\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eKainji\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e64.64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e64.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4.41\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBy Habitat\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEcotonal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e65.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e64.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.81\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLacustrine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e65.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e65.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4.19\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRiverine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e65.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e65.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4.60\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBy Season\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e65.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e65.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4.41\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e64.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e64.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.67\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eWQI Variation by Dam and Habitat Type\u003c/h3\u003e\n\u003cp\u003eThe mean WQI at Jebba Dam (66.06) was slightly higher than at Kainji (64.64). Although this difference is modest in absolute terms, it suggests that local factors beyond regional hydrology influence water quality. Kainji, located approximately 100 km upstream, is the first reservoir to receive the sediment-laden white flood, which carries high concentrations of pollutants and suspended material from upstream agricultural catchments (Adeogun et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). This seasonal influx likely depresses water quality disproportionately at Kainji, while Jebba may benefit from partial settling and dilution as water passes through the upper reservoir (Adegbehin et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Adeogun et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Nwobi-Okoye \u0026amp; Igboanugo, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHabitat-level differences were smaller still but followed a consistent spatial pattern (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Lacustrine zones showed the highest mean WQI (65.78), followed by ecotonal (65.17) and riverine (65.11) habitats. This pattern reflects the primary modes of pollution transport: riverine and ecotonal zones receive the most concentrated upstream pollutant loads, while the deeper, low-flow lacustrine zones allow suspended material to settle and some soluble pollutants to disperse (Irenosen et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Okafor et al., \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Taken together, spatial variation amounted to less than one WQI unit across habitats and 1.4 units across dams, small relative to the 11.6-unit range observed across months. Together, these patterns indicate that upstream runoff and terrestrial activities are the dominant sources of water quality degradation in the system.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eSeasonal and Monthly Variation\u003c/h2\u003e \u003cp\u003eThe seasonal pattern was counterintuitive: mean WQI was higher in the dry season (65.60) than in the wet season (64.85), contrary to the expectation that increased rainfall and discharge would dilute pollutants and improve water quality. This reversal of the expected dilution effect reflects the dominant role of runoff-driven pollution. During the wet season, rather than diluting pollutants, increased rainfall mobilises sediments, nutrients, and organic matter from agricultural and urban catchments into the reservoirs, with the runoff effect outweighing any dilution benefit (Medupe \u0026amp; Letshwenyo, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; US EPA, 2025b).\u003c/p\u003e \u003cp\u003eThe monthly WQI trajectory provides a more detailed temporal signature of this pattern (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Values peaked in April and May (71.28 and 71.54, respectively), corresponding to the end of the dry season when terrestrial pollutant loads were minimal. A sharp decline occurred in June and July (68.71 and 63.11), coinciding with the onset of the rainy season and the first flush of accumulated pollutants from land surfaces (Wiranegara et al., \u003cspan citationid=\"CR118\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Values remained\u003c/p\u003e \u003cp\u003edepressed throughout the wet season, reflecting continued non-point source pollution inputs, before partially recovering in the early dry season.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eA second notable drop occurred in January\u0026ndash;February (61.11 and 59.95), coinciding with the arrival of the black flood from the Fouta Djallon highlands via Guinea, Mali, and Niger (Nwobi-Okoye \u0026amp; Igboanugo, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Although black-flood waters pass through extensive upstream floodplains where some sediment is deposited, the prolonged flow appears to transport an accumulation of dissolved solutes and organic pollutants from distant catchments, producing a secondary deterioration of water quality. February\u0026rsquo;s mean WQI of 59.95 represents the annual low point and falls below the Good/Medium threshold, making this period a critical risk window for aquatic health.\u003c/p\u003e \u003cp\u003eA direct comparison of temporal and spatial variation reveals that temporal fluctuations in WQI far exceeded spatial differences. Monthly values ranged by 11.59 units, from 71.54 in May to 59.95 in February, whereas the difference between dams was only 1.42 units and the difference between the highest- and lowest-WQI habitats was just 0.67 units. These patterns indicate that annual hydrological dynamics dominate WQI variability, while dam- and habitat-level effects remain comparatively small.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eBimodal Distribution of Water Quality\u003c/h2\u003e \u003cp\u003eThe histogram of WQI values (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) shows an apparent deviation from a single normal distribution, with two distinct peaks that define a bimodal distribution. The primary distribution, represented by the broad peak, is centred around a WQI of approximately 64\u0026ndash;65 and reflects the system\u0026rsquo;s prevailing moderate water quality state. In contrast, a second, much narrower peak is centred around a WQI of approximately 71\u0026ndash;72.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis bimodal pattern is a critical finding that indicates the presence of two distinct seasonal water-quality regimes. The exceptionally high WQI values in the narrow peak correspond to the mean monthly WQI values for April (71.28) and May (71.54). The low standard deviations for these months, specifically 1.39 in April and 0.61 in May, explain the tight and concentrated nature of this distribution, indicating a period of high stability and homogeneity in water quality. The broader distribution corresponds to the WQI values for the remaining ten months of the year, which are consistently lower and more variable. This shift is likely governed by the annual hydrological cycle, with the onset of the wet season triggering a rapid transition from the brief high-quality regime to the dominant lower-quality state (Medupe \u0026amp; Letshwenyo, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eCausal Modelling\u003c/h2\u003e \u003cp\u003eThe Granger causality analysis revealed no statistically significant linear relationships between climate variables and water quality parameters after FDR correction across all 336 stratified tests (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This null result is not a limitation of the analysis; it is a substantive finding about the nature of these systems. The stratified design was important because it assessed climate\u0026ndash;water quality relationships within distinct dam, habitat, and seasonal contexts, reducing the risk of masking local dynamics through aggregation and misleading inference from pooled data (Chen et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Wakefield \u0026amp; Lyons, \u003cspan citationid=\"CR113\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Several clusters of unadjusted near-significant relationships (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) were identified, most of which were concentrated in the Kainji Riverine habitat during the dry season. The most ecologically suggestive patterns included relative humidity predicting both pH and EC, and air temperature, windspeed, and pressure predicting NO₃-N. These dry-season signals are consistent with evaporation\u0026ndash;concentration processes, whereby higher humidity slows evaporative water loss and increases dissolved ion concentrations, as well as with wind-driven resuspension of mineralised sediment nutrients during low-water conditions (Duong et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Khurram et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Hicks \u0026amp; McMahon, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Pelik\u0026aacute;n \u0026amp; Markov\u0026aacute;, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Rolls et al., \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). However, none survived multiple-comparison correction, and these patterns should therefore be treated as hypotheses for future testing rather than established relationships.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eNear-significant Granger causality relationships (unadjusted p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) from climate drivers to water quality parameters, ordered by evidence strength. No relationships remained significant after FDR correction (adjusted p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSite\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHabitat\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSeason\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCause\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eEffect\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eP_Value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAdj_P\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKainji\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRiverine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003epH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.010\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.089\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKainji\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRiverine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTemp_Air\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eNO3_N\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.081\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKainji\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRiverine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eWindspeed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eNO3_N\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.081\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKainji\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRiverine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePressure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eNO3_N\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.081\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKainji\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRiverine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDry\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eRH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eEC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.081\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJebba\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRiverine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eWet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePressure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eTurbidity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.473\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"7\"\u003e\u003cb\u003eNote\u003c/b\u003e: Although these relationships did not reach formal statistical significance after FDR correction (adjusted p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05), they represent the strongest potential climate\u0026ndash;water quality links identified in the analysis and are discussed as ecologically suggestive trends for future hypothesis testing.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eVariance Partitioning\u003c/h2\u003e \u003cp\u003eVariance partitioning revealed that temporal variation (season and month) was the dominant driver of the multivariate water quality structure, accounting for 28.3% of the variance (F\u0026thinsp;=\u0026thinsp;14.32, p\u0026thinsp;=\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Spatial factors (dam location and habitat type) explained only 1.8% of variance but remained statistically significant (F\u0026thinsp;=\u0026thinsp;1.85, p\u0026thinsp;=\u0026thinsp;0.038), whereas anthropogenic land use showed no significant independent effect (0.4%, p\u0026thinsp;=\u0026thinsp;0.303) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). NMDS ordination (stress\u0026thinsp;=\u0026thinsp;0.059) visually confirmed this strong seasonal structuring, with wet-season samples distinctly separated from dry-season samples along axis 2, while dam and habitat identity were secondary organising factors (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eChangepoint Detection and Early-Warning Indicators\u003c/h2\u003e \u003cp\u003eThe comparative analysis of changepoint detection algorithms revealed substantial variation in their performance (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). PELT and Binary Segmentation (BinSeg) achieved the highest detection rates (95.2% each), reflecting their strength in identifying multiple changepoints within complex environmental time series (Killick \u0026amp; Eckley, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Killick et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Notably, BinSeg matched the performance of the exact PELT algorithm despite being computationally less intensive, suggesting its suitability for large-scale future monitoring applications. In contrast, CUSUM achieved only 11.9% detection, and the Bayesian approach failed to detect any changepoints (0.0%), likely due to mismatches between their statistical assumptions and the non-stationary, multi-scaled nature of the data (Aminikhanghahi \u0026amp; Cook, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Lund et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), as CUSUM tests are often optimised for detecting abrupt, single-point shifts in the mean from a known value (Killick \u0026amp; Eckley (\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2014\u003c/span\u003e); Lund et al. (\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAcross parameters, WQI, electrical conductivity, and relative humidity showed the highest consensus detection rates (71.4% each; Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), highlighting their value as early-warning indicators. WQI, in particular, offers an integrated measure of overall water quality (Aljanabi et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Because it combines multiple variables, it is especially sensitive to environmental change: shifts in any component, such as DO, pH, or turbidity, can be reflected in detectable changepoints in the index (Chidiac et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Gon\u0026ccedil;alves \u0026amp; Costa, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). EC is a reliable integrative signal of dissolved solids and responds quickly to pollution events and shifts in hydrological regime (US EPA, 2025a). The role of RH as a sentinel indicator is less direct, but it likely reflects its link to evaporation dynamics and the concentration of dissolved ions during low-flow dry periods. Water temperature (64.3%), air temperature (57.1%), and DO (54.8%) showed moderate detection rates and remain useful as supporting context indicators. Rainfall had the lowest detection rate (35.7%), which is consistent with its highly variable, event-driven nature and its poor fit with regime-shift detection methods designed to identify sustained changes in the mean (Lund et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Reeves et al., \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2007\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eConsensus changepoint analysis confirmed significant spatial heterogeneity in changepoint timing and strength across ecotonal, lacustrine, and riverine systems (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). Riverine environments showed earlier and more frequent changepoints, reflecting their greater exposure to upstream disturbances; lacustrine systems showed distinct changepoints associated with internal biogeochemical cycles and stratification dynamics (Wetzel, \u003cspan citationid=\"CR117\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Zhao et al., \u003cspan citationid=\"CR125\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Structural change validation tests (CUSUM and Fluctuation tests) provided independent corroboration for DO and WQI changepoints, with CUSUM showing particularly strong significance for DO in ecotonal and riverine habitats, and both methods showing higher significance for lacustrine WQI at Kainji (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eChangepoint detection rates for water quality and environmental parameters across all sites and methods. Consensus is defined as detection by \u0026ge;\u0026thinsp;2 methods within \u0026plusmn;\u0026thinsp;1 month.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParameter\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal Analyses\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDetections\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDetection Rate (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater Quality Index (WQI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e71.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eElectrical Conductivity (EC)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e71.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRelative Humidity (RH)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e71.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater Temperature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e64.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAir Temperature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e57.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDissolved Oxygen (DO)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e54.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRainfall\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e35.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study demonstrates that the dynamics of water quality in the Kainji\u0026ndash;Jebba reservoir cascade are shaped primarily by timing rather than location. Seasonal and monthly processes explained 28.3% of water quality variation, far outweighing spatial factors (1.8%) and direct land-use effects (0.4%). These results are consistent with flood pulse theory, in which seasonal hydrological cycles drive ecosystem dynamics more powerfully than the longitudinal gradients central to temperate models such as the River Continuum Concept (Junk et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e1989\u003c/span\u003e; Tockner \u0026amp; Stanford, \u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Vannote et al., \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e1980\u003c/span\u003e). High wet-season flows homogenise conditions across dams and habitat types, while dry-season concentration effects produce system-wide changes that override site-specific differences. This temporal structuring mirrors patterns reported in other tropical reservoirs where seasonal stratification and hydrological cycles overwhelm longitudinal gradients (Wiranegara et al., \u003cspan citationid=\"CR118\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR124\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eSpatial Patterns: Weak Zonation in a Connected System\u003c/h2\u003e \u003cp\u003eThe dominance of temporal over spatial factors indicates that, in this system, when you monitor matters more than where. The small but statistically significant spatial differences \u0026mdash; slightly higher WQI at Jebba and marginally better quality in lacustrine habitats \u0026mdash; are interpretable in terms of local hydrology. Kainji receives the first impact of the white flood from upstream agricultural areas, while Jebba benefits from partial settling through the upper reservoir (Adeogun et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Adegbehin et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Lacustrine zones\u0026rsquo; marginally higher WQI reflects water residence time and settling of suspended material, consistent with the wider limnological literature on reservoir self-purification (Irenosen et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). However, these spatial differences are minor in magnitude and operationally less important than the seasonal signal for monitoring design.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eLand-Use Effects: Temporal Rather Than Spatial\u003c/h2\u003e \u003cp\u003eThe minimal land-use variance (0.4%, p\u0026thinsp;=\u0026thinsp;0.303) might seem to contradict the clear evidence throughout the data that agricultural runoff degrades water quality; the sharp WQI drops at the wet-season onset, the lowest values in riverine areas, and the bimodal WQI distribution all point to a land-use influence. The resolution to this apparent paradox lies in scale. Land use does not generate persistent spatial gradients where some sites are consistently worse than others; rather, its effects manifest as synchronous, system-wide runoff pulses during the wet season. When all sites deteriorate simultaneously, the variance attributable to land use appears in the temporal rather than the spatial component (Carvalho et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Saturday et al., \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Shakeri Bostanabad et al., \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Agricultural pollution is therefore a temporal driver, not a spatial one, and it is captured within the 28.3% temporal fraction.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eAbsence of Climate Signals: Nonlinearity and Buffering\u003c/h2\u003e \u003cp\u003eThe failure of any climate\u0026ndash;water quality relationship to survive FDR correction across 336 stratified Granger tests is a substantive finding rather than a methodological limitation. It suggests that at least three interacting factors override linear, direct climate forcing. First, internal biogeochemical feedbacks, microbial respiration, nutrient cycling, and phytoplankton dynamics can amplify, dampen, or redirect initial meteorological forcing, masking the original climate signal (Farrell et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Kondowe et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Second, dam operations and regulated releases impose an anthropogenically controlled hydrological regime that decouples water quality from meteorological drivers (Capon et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Vivan et al., \u003cspan citationid=\"CR112\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). In such systems, direct linear links between climate trends and water-quality responses may be difficult to detect, particularly when causal pathways are nonlinear or mediated through complex hydrological interactions (Bonotto et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ye et al., \u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Third, the relationships between climate and water quality may be nonlinear or threshold-governed in ways that linear Granger tests are not designed to detect (Shakeri Bostanabad et al., \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Ye et al., \u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The cluster of near-significant dry-season signals at Kainji Riverine is suggestive. During low-flow periods when dilution capacity is reduced, the system may become temporarily more sensitive to meteorological forcing, providing a hypothesis worth testing with longer time series and nonlinear methods.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eSentinel Indicators and Early-Warning Framework\u003c/h2\u003e \u003cp\u003eThe high consensus detection rates for WQI, EC, and RH (71.4% each) provide a practical and evidence-based foundation for a tiered early-warning monitoring system. WQI\u0026rsquo;s composite nature makes it inherently sensitive to multi-parameter deterioration, translating complex data into a single actionable score that policymakers can act on (Mamat et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Shaaban \u0026amp; Stevens, \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). EC responds rapidly to influxes of dissolved solids from pollution events or altered hydrology (US EPA, 2025a), while RH\u0026rsquo;s sentinel value reflects its role in governing evaporation-concentration dynamics that amplify dissolved ion loads during dry periods. Together, these three parameters detect incipient regime shifts more reliably than any single conventional parameter. Temperature and DO retain importance as ecological context indicators, but are less sensitive to the specific stressors operating in this system. This interpretation is reinforced by the supplementary weighting sensitivity analysis, which showed that the temporal and spatial conclusions were unchanged across primary, proportional, and equal WQI weighting schemes.\u003c/p\u003e \u003cp\u003eThe consensus changepoint framework adds a confidence dimension to this early-warning system. High multi-method agreement (4\u0026ndash;5 methods) indicates a high-confidence structural break warranting immediate investigation and management response. Lower agreement (2\u0026ndash;3 methods) provides a preliminary alert that justifies intensified monitoring before a full response is triggered. This tiered structure transforms changepoint detection from a research tool into a practical risk-based management instrument (Lund et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eManagement Implications\u003c/h2\u003e \u003cp\u003eOur results support three clear management priorities. First, monitoring programmes should intensify sampling during the two highest-risk windows: June\u0026ndash;July at the onset of the rains, when WQI drops from 71.54 to 63.11, and January\u0026ndash;February during the black flood, when WQI reaches its annual minimum of 59.95. Bi-weekly sampling during these windows, combined with monthly surveillance otherwise, would concentrate effort when it matters most (Carvalho et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Shaaban \u0026amp; Stevens, \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Second, sentinel indicators (WQI, EC, RH) should be prioritised, with temperature and DO retained as core ecological context metrics. This configuration delivers early detection at a lower cost than continuously monitoring the full parameter set. Third, watershed management must address the sources of first-flush pollution loads. Interventions such as riparian buffer restoration, erosion control, and stormwater retention would directly target the wet-season runoff pulses responsible for the most significant water-quality deterioration. Given that dam release schedules also influence mixing and dilution, coordination between reservoir operators and water quality monitoring programmes during risk windows offers additional scope for mitigation (Chiromo et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eLimitations\u003c/h2\u003e \u003cp\u003eTwo limitations deserve explicit acknowledgement. First, with only six sampling sites across two reservoirs, formal spatial hotspot detection using Getis\u0026ndash;Ord Gi* statistics was not feasible, as this method requires at least ten spatial units for reliable results (Bivand et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). We therefore relied on descriptive summaries and visual mapping to characterise spatial patterns, which, while informative, lack the inferential rigour of formal spatial statistics. Second, one annual cycle is sufficient to characterise seasonal patterns, the dominant signal in tropical systems (Junk et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e1989\u003c/span\u003e; Sharma et al., \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), but insufficient to distinguish long-term regime shifts from interannual variability. Whether the observed patterns represent stable features of this system or reflect conditions specific to the 2024\u0026ndash;2025 hydrological year cannot be determined from the present dataset. Longer-term monitoring with vertical profiling to capture stratification dynamics, and the application of nonlinear causal modelling, are priorities for future work.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThis study reveals that water quality in the Upper Niger River Basin is primarily governed by seasonal hydrological cycles and watershed pollution, rather than by spatial differences between dams or habitats or by short-term climate variability. Monthly WQI values declined from a dry-season peak of 71.54 (May) to an annual minimum of 59.95 (February), driven by two distinct runoff-pulse mechanisms: the white flood (wet-season onset, June\u0026ndash;July) and the black flood (January\u0026ndash;February). No direct links between climate variables and water quality survived rigorous FDR correction, suggesting that human-modified hydrology and watershed pollution are the proximate drivers of change.\u003c/p\u003e \u003cp\u003eThe bimodal WQI distribution and the recurring seasonal pollution pulses suggest that human pressures are beginning to overwhelm the system\u0026rsquo;s natural recovery capacity. Variance partitioning confirms the primacy of temporal dynamics: when you monitor matters more than where. Changepoint analysis provides both the evidence base and the practical tools for a tiered early-warning system centred on three sentinel indicators: WQI, EC, and RH, deployed strategically during the two seasonal risk windows.\u003c/p\u003e \u003cp\u003eBeyond the Niger Basin, this integrated approach, combining WQI, variance partitioning, Granger causality, and multi-method changepoint detection, offers a transferable template for other data-poor tropical river systems where urgent conservation needs and limited monitoring resources must be reconciled.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eEthics approval:\u003c/h2\u003e \u003cp\u003eThis research received ethical clearance from the University of KwaZulu-Natal (AREC/00003628/2021) and Federal University of Technology, Minna (Assignment No. 0000014EAU). Field collection was authorised by the Niger State Ministry of Agriculture (Ref: MEAR/FISH/4/VOL.III/690).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eCompeting interests:\u003c/strong\u003e \u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding:\u003c/h2\u003e \u003cp\u003eThis work was funded by a Tertiary Education Trust Fund (TETFund) doctoral fellowship to Mr A. Ibrahim, with additional support from the University of KwaZulu-Natal through a three-year tuition fee remission. Prof. M. Okpeku\u0026rsquo;s research allocation supports publication costs.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eConceptualisation: AI, SOES, MO; Methodology: AI, SOES, MO; Data acquisition: AI; Formal analysis and investigation: AI; Writing \u0026mdash; original draft: AI; Writing \u0026mdash; review and editing: AI, SOES, DVRA, MO; Funding acquisition: AI, MO; Resources: MO; Supervision: MO, DVRA, SOES. All authors read and approved the final manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe first author acknowledges TETFund, Nigeria, for the doctoral fellowship awarded through the Federal University of Technology, Minna. We are grateful to the Federal University of Technology, Minna, for granting study leave and to the University of KwaZulu-Natal for institutional support. We thank Mainstream Energy Solutions Limited and NIFFR for providing meteorological data and facilitating field access. We are particularly grateful to Mallam Muhammad Jafar Baba for facilitating linkage with the Meteorological Department of Mainstream Energy Solutions, Jebba. We thank Elsevier for access to the Scopus database. Analyses were performed using R version 4.5.1. Reference management was conducted using EndNote.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets generated during this study will be deposited in the Dryad Digital Repository upon manuscript acceptance. The DOI will be provided during the revision process.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAdegbehin, A., Yusuf, Y., Iguisi, E., \u0026amp; Zubairu, I. (2016). Reservoir inflow pattern and its effects on hydroelectric power generation at the Kainji Dam. \u003cem\u003eWIT Transactions on Ecology and the Environment,\u003c/em\u003e 203, 233\u0026ndash;244. https://doi.org/10.2495/EID160211\u003c/li\u003e\n\u003cli\u003eAdeogun, A. G., Ibitoye, B. A., Salami, A. W., \u0026amp; Ihagh, G. T. (2020). Sustainable management of erosion prone areas of upper watershed of Kainji hydropower dam, Nigeria. \u003cem\u003eJournal of King Saud University\u0026mdash;Engineering Sciences\u003c/em\u003e, 32(1), 5\u0026ndash;10. https://doi.org/10.1016/j.jksues.2018.05.001\u003c/li\u003e\n\u003cli\u003eAdeogun, A., Sule, B., \u0026amp; Salami, W. (2018). Simulation of sediment yield at the upstream watershed of Jebba lake using SWAT. \u003cem\u003eMalaysian Journal of Civil Engineering\u003c/em\u003e, 27. https://doi.org/10.11113/mjce.v27.15906\u003c/li\u003e\n\u003cli\u003eAlbert, J. S., Destouni, G., Duke-Sylvester, S. M., Magurran, A. E., Oberdorff, T., Reis, R. E., Winemiller, K. O., \u0026amp; Ripple, W. J. (2021). Scientists\u0026rsquo; warning to humanity on the freshwater biodiversity crisis. \u003cem\u003eAmbio\u003c/em\u003e, 50(1), 85\u0026ndash;94. https://doi.org/10.1007/s13280-020-01318-8\u003c/li\u003e\n\u003cli\u003eAljanabi, Z. Z., Al-Obaidy, A. H. M. J., \u0026amp; Hassan, F. M. (2021). A brief review of water quality indices and their applications. \u003cem\u003eIOP Conference Series: Earth and Environmental Science\u003c/em\u003e, 779(1), 012088. https://doi.org/10.1088/1755-1315/779/1/012088\u003c/li\u003e\n\u003cli\u003eAminikhanghahi, S., \u0026amp; Cook, D. J. (2017). A survey of methods for time series change point detection. \u003cem\u003eKnowledge and Information Systems\u003c/em\u003e, 51(2), 339\u0026ndash;367. https://doi.org/10.1007/s10115-016-0987-z\u003c/li\u003e\n\u003cli\u003eAnderson, M. J. (2017). Permutational multivariate analysis of variance (PERMANOVA). In StatsRef: Statistics Reference Online. Wiley. https://doi.org/10.1002/9781118445112.stat07841\u003c/li\u003e\n\u003cli\u003eAPHA (2017). Standard Methods for the Examination of Water and Wastewater (23rd ed.). American Public Health Association.\u003c/li\u003e\n\u003cli\u003eArantes, C. C., Fitzgerald, D. B., Hoeinghaus, D. J., \u0026amp; Winemiller, K. O. (2019). Impacts of hydroelectric dams on fishes and fisheries in tropical rivers through the lens of functional traits. \u003cem\u003eCurrent Opinion in Environmental Sustainability\u003c/em\u003e, 37, 28\u0026ndash;40. https://doi.org/10.1016/j.cosust.2019.04.009\u003c/li\u003e\n\u003cli\u003eArthington, A. H., Bunn, S. E., Poff, N. L., \u0026amp; Naiman, R. J. (2006). The challenge of providing environmental flow rules to sustain river ecosystems. \u003cem\u003eEcological Applications\u003c/em\u003e, 16(4), 1311\u0026ndash;1318. https://doi.org/10.1890/1051-0761(2006)016[1311:TCOPEF]2.0.CO;2\u003c/li\u003e\n\u003cli\u003eAwojobi, O. N., \u0026amp; Tetteh, J. (2017). The impacts of climate change in Africa: A review of the scientific literature. \u003cem\u003eJournal of International Academic Research for Multidisciplinary\u003c/em\u003e, 5(11), 39\u0026ndash;52.\u003c/li\u003e\n\u003cli\u003eBalamurugan, P., Kumar, P., Shankar, K., Nagavinothini, R., \u0026amp; Vijayasurya, K. (2020). Non-carcinogenic risk assessment of groundwater in southern part of Salem district in Tamilnadu, India. \u003cem\u003eJournal of the Chilean Chemical Society\u003c/em\u003e, 65(1), 4697\u0026ndash;4707. http://dx.doi.org/10.4067/S0717-97072020000104697\u003c/li\u003e\n\u003cli\u003eBarbarossa, V., Schmitt, R. J. P., Huijbregts, M. A. J., Zarfl, C., King, H., \u0026amp; Schipper, A. M. (2020). Impacts of current and future large dams on the geographic range connectivity of freshwater fish worldwide. \u003cem\u003eProceedings of the National Academy of Sciences of the United States of America\u003c/em\u003e, 117(7), 3648\u0026ndash;3655. https://doi.org/10.1073/pnas.1912776117\u003c/li\u003e\n\u003cli\u003eBenjamini, Y., \u0026amp; Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. \u003cem\u003eJournal of the Royal Statistical Society\u003c/em\u003e: Series B, 57(1), 289\u0026ndash;300. https://doi.org/10.1111/j.2517-6161.1995.tb02031.x\u003c/li\u003e\n\u003cli\u003eBeusen, A. H. W., Bouwman, A. F., Van Beek, L. P. H., Mogoll\u0026oacute;n, J. M., \u0026amp; Middelburg, J. J. (2016). Global riverine N and P transport to ocean increased during the 20th century despite increased retention. \u003cem\u003eBiogeosciences\u003c/em\u003e, 13(8), 2441\u0026ndash;2451. https://doi.org/10.5194/bg-13-2441-2016 \u003c/li\u003e\n\u003cli\u003eBivand, R. S., Pebesma, E., \u0026amp; G\u0026oacute;mez-Rubio, V. (2013). Applied Spatial Data Analysis with R (2nd ed.). Springer. https://doi.org/10.1007/978-1-4614-7618-4\u003c/li\u003e\n\u003cli\u003eBonotto, G., Peterson, T., Fowler, K., \u0026amp; Western, A. (2022). Identifying causal interactions between groundwater and streamflow using convergent cross-mapping. \u003cem\u003eWater Resources Research\u003c/em\u003e, 58. https://doi.org/10.1029/2021WR030231\u003c/li\u003e\n\u003cli\u003eBray, J. R., \u0026amp; Curtis, J. T. (1957). An ordination of the upland forest communities of southern Wisconsin. \u003cem\u003eEcological Monographs\u003c/em\u003e, 27(4), 326\u0026ndash;349. https://doi.org/10.2307/1942268\u003c/li\u003e\n\u003cli\u003eBrown, R. M., McClelland, N. I., Deininger, R. A., \u0026amp; O\u0026rsquo;Connor, M. F. (1970). A water quality index\u0026mdash;do we dare? Water \u0026amp; Sewage Works, 117(10), 339\u0026ndash;343.\u003c/li\u003e\n\u003cli\u003eCanadian Council of Ministers of the Environment. (2001). CCME Water Quality Index 1.0, Technical Report. Available at: https://ccme.ca/en/res/wqimanualen.pdf (Accessed 23 October 2025).\u003c/li\u003e\n\u003cli\u003eCapon, S. J., Stewart-Koster, B., \u0026amp; Bunn, S. E. (2021). Future of freshwater ecosystems in a 1.5\u0026deg;C warmer world. \u003cem\u003eFrontiers in Environmental Science\u003c/em\u003e, 9, 784642. https://doi.org/10.3389/fenvs.2021.784642\u003c/li\u003e\n\u003cli\u003eCarvalho, L., Mackay, E. B., Cardoso, A. C., Baattrup Pedersen, A., Birk, S., Blackstock, K. L., Borics, G., Borja, A., Feld, C. K., Ferreira, M. T., Globevnik, L., Grizzetti, B., Hendry, S., Hering, D., Kelly, M., Langaas, S., Meissner, K., Panagopoulos, Y., Penning, E., \u0026hellip; Solheim, A. L. (2019). Protecting and restoring Europe\u0026rsquo;s waters: An analysis of the future development needs of the Water Framework Directive. \u003cem\u003eScience of the Total Environment\u003c/em\u003e, 658, 1228\u0026ndash;1238. https://doi.org/10.1016/j.scitotenv.2018.12.255\u003c/li\u003e\n\u003cli\u003eCastillo-Mateo, J. (2022). Distribution-free changepoint detection tests based on the breaking of records. \u003cem\u003eEnvironmental and Ecological Statistics\u003c/em\u003e, 29(3), 655\u0026ndash;676. https://doi.org/10.1007/s10651-022-00539-2\u003c/li\u003e\n\u003cli\u003eChapman, C. A., Abernathy, K., Chapman, L. J., Downs, C., Effiom, E. O., Gogarten, J. F., Golooba, M., Kalbitzer, U., Lawes, M. J., \u0026amp; Mekonnen, A. (2022). The future of sub-Saharan Africa\u0026rsquo;s biodiversity in the face of climate and societal change. \u003cem\u003eFrontiers in Ecology and Evolution\u003c/em\u003e, 10, 790552. https://doi.org/10.3389/fevo.2022.790552\u003c/li\u003e\n\u003cli\u003eChen, J., Wang, D., Li, G., Sun, Z., Wang, X., Zhang, X., \u0026amp; Zhang, W. (2020). Spatial and temporal heterogeneity analysis of water conservation in Beijing-Tianjin-Hebei urban agglomeration. \u003cem\u003eGeoHealth\u003c/em\u003e, 4(8), e2020GH000248. https://doi.org/10.1029/2020GH000248\u003c/li\u003e\n\u003cli\u003eChidiac, S., El Najjar, P., Ouaini, N., El Rayess, Y., \u0026amp; El Azzi, D. (2023). A comprehensive review of water quality indices: history, models, attempts and perspectives. Reviews in Environmental \u003cem\u003eScience and Bio/Technology\u003c/em\u003e, 22(2), 349\u0026ndash;395. https://doi.org/10.1007/s11157-023-09650-7\u003c/li\u003e\n\u003cli\u003eChiromo, A., Amu, O., Otonglo, C., \u0026amp; Balderamma, F. (2016). Kainji and Jebba hydro power plant rehabilitation project. ESIA Summary, Private Sector Department (OPSD3), 1\u0026ndash;28.\u003c/li\u003e\n\u003cli\u003eChong, X. Y., Vericat, D., Batalla, R. J., Teo, F. Y., Lee, K. S. P., \u0026amp; Gibbins, C. N. (2021). A review of the impacts of dams on the hydromorphology of tropical rivers. \u003cem\u003eScience of the Total Environment\u003c/em\u003e, 794, 148686. https://doi.org/10.1016/j.scitotenv.2021.148686\u003c/li\u003e\n\u003cli\u003eDavies-Colley, R. J., \u0026amp; Smith, D. G. (2001). Turbidity, suspended sediment, and water clarity: A review. \u003cem\u003eJournal of the American Water Resources Association\u003c/em\u003e, 37(5), 1085\u0026ndash;1101. https://doi.org/10.1111/j.1752-1688.2001.tb03624.x\u003c/li\u003e\n\u003cli\u003eDickey, D. A., \u0026amp; Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. \u003cem\u003eJournal of the American Statistical Association\u003c/em\u003e, 74(366), 427\u0026ndash;431. https://doi.org/10.2307/2286348\u003c/li\u003e\n\u003cli\u003eDuong, Q. M., Garcia Vidales, D., \u0026amp; Schwartz, B. J. (2023). The effects of humidity on the electrical properties of semiconducting polymers. \u003cem\u003eApplied Physics Letters\u003c/em\u003e, 123(20). https://doi.org/10.1063/5.0169905\u003c/li\u003e\n\u003cli\u003eEteh, D. R., Egobueze, F. E., Paaru, M., Otutu, A., \u0026amp; Osondu, I. (2024). The impact of dam management and rainfall patterns on flooding in the Niger Delta. \u003cem\u003eDiscover Water\u003c/em\u003e, 4(1), 1\u0026ndash;22. https://doi.org/10.1007/s43832-024-00185-8\u003c/li\u003e\n\u003cli\u003eFAO. (2020). Fishery and Aquaculture Country Profiles: Nigeria. https://www.fao.org/fishery/facp/NGA/en\u003c/li\u003e\n\u003cli\u003eFarrell, K. J., Ward, N. K., Krinos, A. I., Hanson, P. C., Daneshmand, V., Figueiredo, R. J., \u0026amp; Carey, C. C. (2020). Ecosystem-scale nutrient cycling responses to increasing air temperatures vary with lake trophic state. \u003cem\u003eEcological Modelling\u003c/em\u003e, 430, 109134. https://doi.org/10.1016/j.ecolmodel.2020.109134\u003c/li\u003e\n\u003cli\u003eFaye, J. M., Ndao, S., \u0026amp; Sambou, L. E. V. (2023). Evolution of the physico-chemical quality of the water in the Manantali Dam Reservoir from 1989 to 2022. Journal of Geoscience and Environment Protection, 11(9), 351\u0026ndash;366. https://doi.org/10.4236/gep.2023.119022\u003c/li\u003e\n\u003cli\u003eFroese, R., \u0026amp; Pauly, D. (2025). FishBase. World Wide Web electronic publication. https://www.fishbase.org/ (Accessed April 2025).\u003c/li\u003e\n\u003cli\u003eGon\u0026ccedil;alves, A. M., \u0026amp; Costa, M. (2011). Application of change-point detection to structural components of water quality variables. \u003cem\u003eAIP Conference Proceedings\u003c/em\u003e, 1389, 1565\u0026ndash;1568. https://doi.org/10.1063/1.3637928\u003c/li\u003e\n\u003cli\u003eGranger, C. W. J. (1969). Investigating causal relations by econometric models and cross-spectral methods. \u003cem\u003eEconometrica\u003c/em\u003e, 37(3), 424\u0026ndash;438. https://doi.org/10.2307/1912791\u003c/li\u003e\n\u003cli\u003eGrill, G., Lehner, B., Thieme, M., Geenen, B., Tickner, D., Antonelli, F., Babu, S., Borrelli, P., Cheng, L., Crochetiere, H., Ehalt Macedo, H., Filgueiras, R., Goichot, M., Higgins, J., Hogan, Z., Lip, B., Mcclain, M. E., Meng, J., Mulligan, M.,\u0026hellip;Zarfl, C. (2019). Mapping the world\u0026rsquo;s free-flowing rivers. \u003cem\u003eNature\u003c/em\u003e, 569(7755), 215\u0026ndash;221. https://doi.org/10.1038/s41586-019-1111-9\u003c/li\u003e\n\u003cli\u003eHansen, M. C., Potapov, P. V., Moore, R., Hancher, M., Turubanova, S. A., Tyukavina, A., Thau, D., Stehman, S. V., Goetz, S. J., Loveland, T. R., Kommareddy, A., Egorov, A., Chini, L., Justice, C. O., \u0026amp; Townshend, J. R. G. (2013). High-resolution global maps of 21st-century forest cover change. \u003cem\u003eScience\u003c/em\u003e, 342(6160), 850\u0026ndash;853. https://doi.org/10.1126/science.1244693\u003c/li\u003e\n\u003cli\u003eHicks, D., \u0026amp; McMahon, R. (2003). Temperature and relative humidity effects on water loss in Perna perna. \u003cem\u003eBulletin of Marine Science\u003c/em\u003e, 72, 135\u0026ndash;150. \u003c/li\u003e\n\u003cli\u003eInternational Hydropower Association. (n.d.). Mainstream Energy Solutions Limited. Available at: https://www.hydropower.org/our-members/mainstream-energy-solutions-limited (Accessed 13 May 2025).\u003c/li\u003e\n\u003cli\u003eIntergovernmental Panel on Climate Change (IPCC) (2022). Climate Change 2022: Impacts, Adaptation and Vulnerability. Cambridge University Press. Available at: https://www.ipcc.ch/report/ar6/syr/ (Accessed: 10 July 2025).\u003c/li\u003e\n\u003cli\u003eIrenosen, O. G., Festus, A. A., \u0026amp; Coolborn, A. F. (2012). Water quality assessment of the Owena multi-purpose dam, Ondo State. \u003cem\u003eJournal of Environmental Protection\u003c/em\u003e, 3, 14\u0026ndash;25. https://doi.org/10.4236/jep.2012.31003\u003c/li\u003e\n\u003cli\u003eIta, E. O., Imevbore, A. M. A., \u0026amp; Odjugo, P. A. O. (1985). Impact of the Kainji Dam on fisheries in Nigeria. \u003cem\u003eNigerian Journal of Aquatic Sciences\u003c/em\u003e, 1(1), 15\u0026ndash;27.\u003c/li\u003e\n\u003cli\u003eJane, S. F., Hansen, G. J. A., Kraemer, B. M., Leavitt, P. R., North, R. L., Pilla, R. M., Williamson, C. E., Woolway, R. I., Zwart, J. A., \u0026amp; Rose, K. C. (2021). Widespread deoxygenation of temperate lakes. \u003cem\u003eNature\u003c/em\u003e, 594, 66\u0026ndash;70. https://doi.org/10.1038/s41586-021-03550-y\u003c/li\u003e\n\u003cli\u003eJ\u0026eacute;z\u0026eacute;quel, C., Tedesco, P. A., Darwall, W., Sversut Dias, M., Frederico, R. G., Hidalgo, M., Hugueny, B., Maldonado Ocampo, J., Martens, K., Ortega, H., Torrente Vilara, G., Zuanon, J., \u0026amp; Oberdorff, T. (2020). Freshwater fish diversity hotspots for conservation priorities in the Amazon Basin. \u003cem\u003eConservation Biology\u003c/em\u003e, 34, 956\u0026ndash;965. https://doi.org/10.1111/cobi.13466\u003c/li\u003e\n\u003cli\u003eJones, E. R., Van Vliet, M. T., Qadir, M., \u0026amp; Bierkens, M. F. (2021). Country-level and gridded estimates of wastewater production, collection, treatment and reuse. \u003cem\u003eEarth System Science Data\u003c/em\u003e, 13(2), 237\u0026ndash;254. https://doi.org/10.5194/essd-13-237-2021\u003c/li\u003e\n\u003cli\u003eJunk, W. J., Bayley, P. B., \u0026amp; Sparks, R. E. (1989). The flood pulse concept in river-floodplain systems. \u003cem\u003eCanadian Special Publication of Fisheries and Aquatic Sciences\u003c/em\u003e, 106(1), 110\u0026ndash;127.\u003c/li\u003e\n\u003cli\u003eKhurram, M., Neuber, S., Sill, A., \u0026amp; Helm, C. A. (2025). Influence of relative humidities on highly electrically conductive polyelectrolyte multilayer films. \u003cem\u003eLangmuir\u003c/em\u003e, 41(14), 9361\u0026ndash;9368. https://doi.org/10.1021/acs.langmuir.4c05382\u003c/li\u003e\n\u003cli\u003eKillick, R., \u0026amp; Eckley, I. A. (2014). changepoint: An R package for changepoint analysis. \u003cem\u003eJournal of Statistical Software\u003c/em\u003e, 58(3), 1\u0026ndash;19.\u003c/li\u003e\n\u003cli\u003eKillick, R., Fearnhead, P., \u0026amp; Eckley, I. A. (2012). Efficient detection of multiple changepoints within an oceanographic time series. \u003cem\u003eProceedings of ISI World Statistics Congress\u003c/em\u003e, 4137\u0026ndash;4142. \u003c/li\u003e\n\u003cli\u003eKondowe, B. N., Masese, F. O., Raburu, P. O., Singini, W., Sitati, A., \u0026amp; Walumona, R. J. (2022). Seasonality in environmental conditions drive variation in plankton communities in a shallow tropical lake. \u003cem\u003eFrontiers in Water\u003c/em\u003e, 4, 883767. https://doi.org/10.3389/frwa.2022.883767\u003c/li\u003e\n\u003cli\u003eLanrewaju, F. (2012). Urbanisation, housing quality and environmental degeneration in Nigeria. \u003cem\u003eJournal of Geography and Regional Planning\u003c/em\u003e, 5, 422\u0026ndash;429. https://doi.org/10.5897/JGRP12.060\u003c/li\u003e\n\u003cli\u003eLegendre, P., \u0026amp; Legendre, L. (2012). Numerical Ecology (3rd ed.). Elsevier.\u003c/li\u003e\n\u003cli\u003eLund, R. B., Beaulieu, C., Killick, R., Lu, Q., \u0026amp; Shi, X. (2023). Good practices and common pitfalls in climate time series changepoint techniques. \u003cem\u003eJournal of Climate\u003c/em\u003e, 36(23), 8041\u0026ndash;8057. https://doi.org/10.1175/JCLI-D-22-0954.1\u003c/li\u003e\n\u003cli\u003eMamat, N., Mohd Razali, S. F., \u0026amp; Hamzah, F. B. (2023). Enhancement of water quality index prediction using support vector machine with sensitivity analysis. \u003cem\u003eFrontiers in Environmental Science\u003c/em\u003e, 10, 1061835. https://doi.org/10.3389/fenvs.2022.1061835 \u003c/li\u003e\n\u003cli\u003eMbaka, J. G., \u0026amp; Mwaniki, M. W. (2017). A critical review of the effect of water storage reservoirs on organic matter decomposition in rivers. \u003cem\u003eEnvironmental Reviews\u003c/em\u003e, 25(2), 193\u0026ndash;198. https://doi.org/10.1139/er-2016-0041 \u003c/li\u003e\n\u003cli\u003eMedupe, M., \u0026amp; Letshwenyo, M. W. (2024). Investigation of self-purification and water quality index during dry and rainy seasons in the Khurumela Stream (Botswana). \u003cem\u003eJournal of Ecohydraulics\u003c/em\u003e, 10(1), 1\u0026ndash;18. https://doi.org/10.1080/24705357.2024.2363755\u003c/li\u003e\n\u003cli\u003eNigerian Meteorological Agency. (2023). 2023 State of the Climate in Nigeria. https://www.nimet.gov.ng \u003c/li\u003e\n\u003cli\u003eNwobi-Okoye, C., \u0026amp; Igboanugo, A. (2013). Predicting water levels at Kainji Dam using artificial neural networks. \u003cem\u003eNigerian Journal of Technology\u003c/em\u003e, 32(1), 129\u0026ndash;136.\u003c/li\u003e\n\u003cli\u003eO\u0026rsquo;Mara, K., Venarsky, M., Stewart-Koster, B., McGregor, C., Schulz, C., Marshall, J., \u0026amp; Bunn, S. E. (2024). Hydrological connectivity and environment characteristics explain spatial variation in fish assemblages in a wet\u0026ndash;dry tropical river. \u003cem\u003eHydrobiologia\u003c/em\u003e, 851(21), 5207\u0026ndash;5221. https://doi.org/10.1007/s10750-024-05676-2\u003c/li\u003e\n\u003cli\u003eOester, R., Keck, F., Moretti, M. S., Altermatt, F., Bruder, A., \u0026amp; Ferreira, V. (2025). A global synthesis on land-cover changes in watersheds shaping freshwater detrital food webs. \u003cem\u003eGlobal Change Biology\u003c/em\u003e, 31(8), e70380. https://doi.org/10.1111/gcb.70380 \u003c/li\u003e\n\u003cli\u003eOkafor, C. P., Nduka, J. K., Okafor, P. O., \u0026amp; Offor, C. C. (2025). Behavioural dynamics of water quality parameters in Agulu Lake Water, South-East Nigeria. \u003cem\u003eCleaner Water\u003c/em\u003e, 4, 100127. https://doi.org/10.1016/j.clwat.2025.100127 \u003c/li\u003e\n\u003cli\u003eOlaosebikan, B. D., \u0026amp; Raji, A. (2013). Field Guide to Nigerian Freshwater Fishes (2nd ed.). Federal College of Freshwater Fisheries Technology, New Bussa.\u003c/li\u003e\n\u003cli\u003eOyebande, L. (2001). Streamflow regime change and ecological response in the Lake Chad Basin in Nigeria. \u003cem\u003eIAHS Publications\u003c/em\u003e, 266, 101\u0026ndash;112.\u003c/li\u003e\n\u003cli\u003eOyebande, L., \u0026amp; Odunuga, S. (2010). Climate change impact on water resources at the transboundary level in West Africa. Open Hydrology Journal, 4(1), 163\u0026ndash;172. https://doi.org/10.2174/1874378101004010163\u003c/li\u003e\n\u003cli\u003ePelik\u0026aacute;n, P., \u0026amp; Markov\u0026aacute;, J. (2013). Wind effect on water surface of water reservoirs. \u003cem\u003eActa Universitatis Agriculturae et Silviculturae Mendelianae Brunensis\u003c/em\u003e, 61, 1823\u0026ndash;1828. https://doi.org/10.11118/actaun201361061823\u003c/li\u003e\n\u003cli\u003ePoff, N. L., Allan, J. D., Bain, M. B., Karr, J. R., Prestegaard, K. L., Richter, B. D., \u0026hellip; Stromberg, J. C. (1997). The natural flow regime: A paradigm for river conservation and restoration. \u003cem\u003eBioScience\u003c/em\u003e, 47(11), 769\u0026ndash;784.\u003c/li\u003e\n\u003cli\u003ePototsky, P. C., \u0026amp; Cresswell, W. (2021). Conservation research output in sub-Saharan Africa is increasing, but only in a few countries. \u003cem\u003eOryx\u003c/em\u003e, 55(6), 924\u0026ndash;933. https://doi.org/10.1017/S0030605320000046\u003c/li\u003e\n\u003cli\u003eReeves, J., Chen, J., Wang, X. L., Lund, R., \u0026amp; Lu, Q. Q. (2007). A review and comparison of changepoint detection techniques for climate data. \u003cem\u003eJournal of Applied Meteorology and Climatology\u003c/em\u003e, 46(6), 900\u0026ndash;915. https://doi.org/10.1175/JAM2493.1\u003c/li\u003e\n\u003cli\u003eRegier, P., Brice\u0026ntilde;o, H., \u0026amp; Boyer, J. N. (2019). Analysing and comparing complex environmental time series using a cumulative sums approach. \u003cem\u003eMethodsX\u003c/em\u003e, 6, 779\u0026ndash;787. https://doi.org/10.1016/j.mex.2019.03.014\u003c/li\u003e\n\u003cli\u003eReid, A. J., Carlson, A. K., Creed, I. F., Eliason, E. J., Gell, P. A., Johnson, P. T. J., Kidd, K. A., Maccormack, T. J., Olden, J. D., Ormerod, S. J., Smol, J. P., Taylor, W. W., Tockner, K., Vermaire, J. C., Dudgeon, D., \u0026amp; Cooke, S. J. (2019). Emerging threats and persistent conservation challenges for freshwater biodiversity. \u003cem\u003eBiological Reviews\u003c/em\u003e, 94(3), 849\u0026ndash;873. https://doi.org/10.1111/brv.12480\u003c/li\u003e\n\u003cli\u003eRolls, R. J., Leigh, C., \u0026amp; Sheldon, F. (2012). Mechanistic effects of low-flow hydrology on riverine ecosystems. \u003cem\u003eFreshwater Science\u003c/em\u003e, 31(4), 1163\u0026ndash;1186. https://doi.org/10.1899/12-002.1\u003c/li\u003e\n\u003cli\u003eSaturday, A., Herrnegger, M., Kangume, S., \u0026amp; Stecher, G. (2025). Spatiotemporal variability of surface water quality in tropical agriculture-dominated catchments. \u003cem\u003eScientific Reports\u003c/em\u003e, 15(1), 42971. https://doi.org/10.1038/s41598-025-27066-x\u003c/li\u003e\n\u003cli\u003eScheffer, M., Bascompte, J., Brock, W. A., Brovkin, V., Carpenter, S. R., Dakos, V., Held, H., Van Nes, E. H., Rietkerk, M., \u0026amp; Sugihara, G. (2009). Early-warning signals for critical transitions. \u003cem\u003eNature\u003c/em\u003e, 461(7260), 53\u0026ndash;59. https://doi.org/10.1038/nature08227\u003c/li\u003e\n\u003cli\u003eShaaban, N. A., \u0026amp; Stevens, D. K. (2025). Transforming complex water quality monitoring data into water quality indices. \u003cem\u003eWater Resources Management\u003c/em\u003e, 39(8), 3883\u0026ndash;3899. https://doi.org/10.1007/s11269-025-04135-4\u003c/li\u003e\n\u003cli\u003eShakeri Bostanabad, R., Amini, H., Fakheri, F., Lam, M., \u0026amp; Zahraie, B. (2025). Comparative analysis of correlation and causality inference in water quality problems. \u003cem\u003eScientific Reports\u003c/em\u003e, 15. https://doi.org/10.1038/s41598-025-85908-0\u003c/li\u003e\n\u003cli\u003eSharma, S., Gray, D. K., Read, J. S., O\u0026rsquo;reilly, C. M., Schneider, P., Qudrat, A., Gries, C., Stefanoff, S., Hampton, S. E., \u0026amp; Hook, S. (2015). A global database of lake surface temperatures collected by in situ and satellite methods from 1985\u0026ndash;2009. \u003cem\u003eScientific Data\u003c/em\u003e, 2, 1\u0026ndash;19. https://doi.org/10.1038/sdata.2015.8\u003c/li\u003e\n\u003cli\u003eSutadian, A. D., Muttil, N., Yilmaz, A. G., \u0026amp; Perera, B. J. C. (2015). Development of river water quality indices \u0026mdash; a review. Environmental Monitoring and Assessment, 188(1), 58. https://doi.org/10.1007/s10661-015-5050-0\u003c/li\u003e\n\u003cli\u003eSylla, M. B., Faye, A., Giorgi, F., Diedhiou, A., \u0026amp; Kunstmann, H. (2018). Projected heat stress under 1.5\u0026deg;C and 2\u0026deg;C global warming in West Africa. \u003cem\u003eEarth\u0026rsquo;s Future\u003c/em\u003e, 6(7), 1029\u0026ndash;1044. https://doi.org/10.1029/2018EF000873\u003c/li\u003e\n\u003cli\u003eTickner, D., Opperman, J. J., Abell, R., Acreman, M., Arthington, A. H., Bunn, S. E., Cooke, S. J., Dalton, J., Darwall, W., Edwards, G., Harrison, I., Hughes, K., Jones, T., Lecl\u0026egrave;re, D., Lynch, A. J., Leonard, P., Mcclain, M. E., Muruven, D., Olden, J. D., \u0026hellip; Young, L. (2020). Bending the curve of global freshwater biodiversity loss. \u003cem\u003eBioScience\u003c/em\u003e, 70, 330\u0026ndash;342. https://doi.org/10.1093/biosci/biaa002\u003c/li\u003e\n\u003cli\u003eTockner, K., \u0026amp; Stanford, J. A. (2002). Riverine flood plains: present state and future trends. \u003cem\u003eEnvironmental Conservation\u003c/em\u003e, 29(3), 308\u0026ndash;330.\u003c/li\u003e\n\u003cli\u003eUddin, M. G., Nash, S., \u0026amp; Olbert, A. I. (2021). A review of water quality index models and their use for assessing surface water quality. \u003cem\u003eEcological Indicators\u003c/em\u003e, 122, 107218. https://doi.org/10.1016/j.ecolind.2020.107218\u003c/li\u003e\n\u003cli\u003eUnited States Environmental Protection Agency. (2025a). Indicators: Conductivity. https://www.epa.gov/national-aquatic-resource-surveys/indicators-conductivity\u003c/li\u003e\n\u003cli\u003eUnited States Environmental Protection Agency. (2025b). Soak Up the Rain: What\u0026rsquo;s the problem? https://www.epa.gov/soakuptherain/soak-rain-whats-problem\u003c/li\u003e\n\u003cli\u003eUwadiae, R., Daudu, E., \u0026amp; Lawal, M. (2021). Impact of Water Hyacinth (Eichhornia crassipes) Infestation on Water Quality and Growth of a Benthic Mollusc Pachymelania aurita M\u0026uuml;ller (Gastropoda: Melaniidae): Experimental Evaluation. \u003cem\u003eAnchor University Journal of Science and Technology\u003c/em\u003e, 2(1). https://doi.org/10.4314/aujst.v2i1\u003c/li\u003e\n\u003cli\u003eVannote, R. L., Minshall, G. W., Cummins, K. W., Sedell, J. R., \u0026amp; Cushing, C. E. (1980). The river continuum concept. \u003cem\u003eCanadian Journal of Fisheries and Aquatic Sciences\u003c/em\u003e, 37(1), 130\u0026ndash;137. https://doi.org/10.1139/f80-017\u003c/li\u003e\n\u003cli\u003eVivan, E., Yakubu, A., \u0026amp; Caleb, A. (2014). Effects of human activities on water quality of Lamingo Dam. \u003cem\u003eInternational Journal of Interdisciplinary Research and Innovations\u003c/em\u003e, 2, 97\u0026ndash;104. \u003c/li\u003e\n\u003cli\u003eWakefield, J., \u0026amp; Lyons, H. (2010). Spatial aggregation and the ecological fallacy. In Handbook of Spatial Statistics (pp. 541\u0026ndash;558). CRC Press.\u003c/li\u003e\n\u003cli\u003eWalsh, G., Pease, A. A., Woodford, D. J., Stiassny, M. L., Gaugris, J. Y., \u0026amp; South, J. (2022). Functional diversity of Afrotropical fish communities across river gradients in the Republic of Congo. \u003cem\u003eFrontiers in Environmental Science\u003c/em\u003e, 10, 981960. https://doi.org/10.3389/fenvs.2022.981960\u003c/li\u003e\n\u003cli\u003eWantzen, K. M., Junk, W. J., \u0026amp; Rothhaupt, K.-O. (2008). An extension of the flood pulse concept (FPC) for lakes. \u003cem\u003eHydrobiologia\u003c/em\u003e, 613(1), 151\u0026ndash;170. https://doi.org/10.1007/s10750-008-9584-2\u003c/li\u003e\n\u003cli\u003eWard, J., \u0026amp; Stanford, J. (1995). The serial discontinuity concept: extending the model to floodplain rivers. Regulated Rivers: Research \u0026amp; Management, 10(2\u0026ndash;4), 159\u0026ndash;168. https://doi.org/10.1002/RRR.3450100211\u003c/li\u003e\n\u003cli\u003eWetzel, R. G. (2001). Limnology: Lake and River Ecosystems (3rd ed.). Academic Press.\u003c/li\u003e\n\u003cli\u003eWiranegara, P., Sunardi, S., Sumiarsa, D., \u0026amp; Juahir, H. (2023). Characteristics and changes in water quality based on climate and hydrology effects in the Cirata Reservoir. \u003cem\u003eWater\u003c/em\u003e, 15(17), 3132. https://doi.org/10.3390/w15173132\u003c/li\u003e\n\u003cli\u003eWWF. (2020). Living Planet Report 2020 \u0026mdash; Bending the Curve of Biodiversity Loss. WWF, Gland, Switzerland.\u003c/li\u003e\n\u003cli\u003eYe, L., Tan, L., Wu, X., Cai, Q., \u0026amp; Li, B. L. (2022). Nonlinear causal analysis reveals an effective water level regulation approach for phytoplankton blooms controlling in reservoirs. \u003cem\u003eScience of the Total Environment\u003c/em\u003e, 806(Pt 4), 150948. https://doi.org/10.1016/j.scitotenv.2021.150948\u003c/li\u003e\n\u003cli\u003eZanaga, D., Van De Kerchove, R., Daems, D., De Keersmaecker, W., Brockmann, C., Kirches, G., Wevers, J., Cartus, O., Santoro, M., Fritz, S., Lesiv, M., Herold, M., Tsendbazar, N.-E., Xu, P., Ramoino, F., \u0026amp; Arino, O. (2022). ESA WorldCover 10 m 2021 v200. Zenodo. https://doi.org/10.5281/zenodo.7254221\u003c/li\u003e\n\u003cli\u003eZeileis, A., Leisch, F., Hornik, K., \u0026amp; Kleiber, C. (2003). strucchange: Testing for structural change in linear regression relationships. R News, 3(2), 7\u0026ndash;13.\u003c/li\u003e\n\u003cli\u003eZhang, Y., Gao, X., Sun, B., \u0026amp; Liu, X. (2024). Hydrodynamics, diagenesis and hypoxia variably drive benthic oxygen flux in a river\u0026ndash;reservoir system. \u003cem\u003eWater Resources Research\u003c/em\u003e, 60(1), e2023WR035449. https://doi.org/10.1029/2023WR035449\u003c/li\u003e\n\u003cli\u003eZhao, S., Hermans, M., Niemist\u0026ouml;, J., Vesterinen, J., \u0026amp; Jilbert, T. (2025). Stratification controls the magnitude of in-lake phosphorus cycling. \u003cem\u003eHydrobiologia\u003c/em\u003e, 852(2), 359\u0026ndash;376. https://doi.org/10.1007/s10750-024-05701-4\u003c/li\u003e\n\u003cli\u003eZolghadr-Asli, B., Enayati, M., Pourghasemi, H. R., Naghdyzadegan Jahromi, M., \u0026amp; Tiefenbacher, J. P. (2021). Application of Granger-causality to study climate change impacts on depletion patterns of inland water bodies. \u003cem\u003eHydrological Sciences Journal\u003c/em\u003e, 66(12), 1767\u0026ndash;1776. https://doi.org/10.1080/02626667.2021.1944633\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"environmental-monitoring-and-assessment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"emas","sideBox":"Learn more about [Environmental Monitoring and Assessment](http://link.springer.com/journal/10661)","snPcode":"10661","submissionUrl":"https://submission.nature.com/new-submission/10661/3","title":"Environmental Monitoring and Assessment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Reservoir water quality, Temporal variability, Sentinel indicators, Land-use impacts, Niger River Basin, Tropical Africa","lastPublishedDoi":"10.21203/rs.3.rs-9118764/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9118764/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFreshwater systems in tropical Africa face growing pressures from dams, climate change, and land-use expansion, yet we still do not fully understand which factors most significantly affect water quality. Knowing this is crucial for designing effective monitoring and management, especially when resources are limited. We conducted monthly water-quality surveys at six sites across the Kainji\u0026ndash;Jebba reservoir cascade in Nigeria from April 2024 to March 2025, encompassing two dams and three distinct habitat types (riverine, ecotonal, and lacustrine). Using a modified NSF Water Quality Index (WQI), variance partitioning (RDA), Granger causality, and six changepoint detection methods, we untangled the effects of space, time, land use, and climate, and identified early-warning indicators of water quality change. Our findings show that when changes happen matters more than where. Seasonal and monthly variations explained 28.3% of water quality variance, while spatial differences accounted for only 1.8%, and direct land-use effects were minimal (0.4%). Agricultural impacts were expressed primarily through short, wet-season runoff pulses that affected all sites simultaneously, rather than through persistent spatial differences. After false discovery rate correction, no relationships between climate variables and water quality remained significant. WQI, electrical conductivity (EC), and relative humidity (RH) emerged as reliable early-warning sentinel indicators, detecting regime shifts in 71.4% of cases. These findings suggest that monitoring programmes in tropical reservoir systems should prioritise targeted seasonal windows: the wet-season runoff peak (June\u0026ndash;July) and the arrival of sediment-rich black floods (January\u0026ndash;February), rather than expanding spatial coverage.\u003c/p\u003e","manuscriptTitle":"Temporal dynamics override spatial gradients in Afrotropical reservoir water quality","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-05 17:09:17","doi":"10.21203/rs.3.rs-9118764/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-20T10:10:46+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-20T08:37:22+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-13T13:29:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"82026155520591861122024205746190029198","date":"2026-04-02T10:18:37+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"315194905428098583388334530315903759344","date":"2026-04-02T08:40:49+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-02T06:33:57+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"84348329218545758387272698369746214871","date":"2026-04-01T08:25:52+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-31T08:29:20+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-23T14:50:17+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-23T14:49:31+00:00","index":"","fulltext":""},{"type":"submitted","content":"Environmental Monitoring and Assessment","date":"2026-03-14T01:20:12+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"environmental-monitoring-and-assessment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"emas","sideBox":"Learn more about [Environmental Monitoring and Assessment](http://link.springer.com/journal/10661)","snPcode":"10661","submissionUrl":"https://submission.nature.com/new-submission/10661/3","title":"Environmental Monitoring and Assessment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"3cadbbb5-2f99-4a8a-b093-d71251b2da12","owner":[],"postedDate":"April 5th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-04-20T10:26:44+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-05 17:09:17","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9118764","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9118764","identity":"rs-9118764","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}