Nutrient chemistry and eutrophication risk assessment of the Ghaghara river, India Open Access

Runoff from watershed areas having different land-use types is received into rivers, which act as a conduit transporting nutrients in dissolved and suspended form from land to a receiving water body, which may be a river, lake, or ocean. Nutrients in an aquatic system can be contributed from allochthonous inputs (weathering materials from its floodplain, atmospheric precipitation, domestic sewage, and runoff from the agriculture areas), and autochthonous input (instream production of organic matter and its mineralization) (Adamczuk et al. 2019). Excess loading of nutrients in an aquatic ecosystem can result in eutrophication conditions, which may promote excessive growth of unwanted phytoplankton species, ultimately leading to the destruction of the aquatic ecosystem (Howarth & Marino 2006).

Human-induced change in the biogeochemical cycling of nutrients by processes like land use and land cover changes, fossil fuel burning, and runoff from residential and agricultural areas results in a many-fold increase in nutrient loading of the aquatic ecosystem (Meybeck 1982; Pizarroa et al. 2010; Weigelhofer et al. 2018). The increased loading of the nutrients may alter the nutrient balance ratio (C: N:P: Si) required for primary productivity and may lead to changes in species composition and function (Danielsson et al. 2008). The occurrence of coastal hypoxia zones is one of the major problems associated with nutrient loading, as reported from the Gulf of Mexico (Rabalais & Turner 2019), Baltic Sea (Caballero-Alfonso et al. 2015), East China Sea (Chen et al. 2007), and Pacific Ocean (Higashi et al. 2012). Therefore, it is important to measure nutrient concentration and its variation in the aquatic system for proper management of nutrients and to mitigate harmful impacts arising due to excess loading of nutrients (Lima et al. 2010).

The Ghaghara river originates from the Tibetan plateau and is a transboundary river. Out of the total catchment area of ∼127,950 km², ∼55% lies in Nepal, and the remaining part lies in India. The Ghaghara river drains through 5 districts of Nepal with low population density and 21 districts in India having a high population density (Mohan 2018). The objective of the present study was to (1) assess the spatial and seasonal variation in the dissolved nutrient concentration in the Ghaghara river; (2) to identify factors controlling nutrient chemistry in the Ghaghara river by using multivariate statistical analysis; (3) to determine the potential of the Ghaghara river to cause eutrophication in the River Ganga through the use of dissolved elemental ratios and indicators for coastal eutrophication potential (ICEP) values.

The Ghaghara river is the largest tributary of the Ganga river in terms of annual flow contribution and its major tributaries are Sharda, Kwano, Rapti, Chhoti Gandak, Seti, and Daha. The river originates from the Matsatung glacier in the Himalaya near Tibet at an elevation of about 4,800 m and lies between 79°29′ to 84°49′ E longitude and 25°47′ to 30°55′ N latitudes. It joins the Ganga river near Maharajganj, in Bihar state, after a journey of about 1,080 km (Singh & Awasthi 2011). The total catchment basin area of the Ghaghara river is 127,950 km² and the annual average discharge of the River Ghaghara is about 94,400 m³ yr⁻¹ (Singh et al. 2017). The climate of the Ghaghara basin is sub-tropical, monsoonal with annual mean temperature varying from 10.9 °C to 31.71 °C, and the average annual rainfall in the Ghaghara Basin is 1,000.30 mm with 87% of rainfall occurring during the monsoon (CWC 2020). The Ghaghara basin mainly consists of Pleistocene older alluvium (Varanasi older alluvium and T2 terrace surface) and Holocene newer alluvium (Bangar and Khadar). The older alluvium is characterized by different lithologies such as kankar, gravel, sand, silt, clay, etc (Singh et al. 2020). The river basin mainly consists of agricultural land and forest area followed by water bodies and built-up area as a major land use pattern. The major forest types that exist in the area are deciduous forest, evergreen forest, and mixed forest (Singh et al. 2017).

Water samples (n = 54) were collected in acid-cleaned polyethylene bottles from 18 locations in the Ghaghara river and its tributaries during the post-monsoon (October (4th to 7th) 2018), pre-monsoon (February (5th to 8th), 2019), and monsoon season (July (15th to 18th) 2019) (Figure 1). Sampling point coordinates were taken by using the Trimble Juno-3B GPS. Samples for Biochemical Oxygen Demand (BOD) analysis were collected in BOD glass bottles. The sampling location details, along with the major land use patterns in the sampling areas, are given in Table 1. The sampling event was carried out over a 4-day span covering a river stretch of 722 km during each season. Water samples were collected from the mid-section of the river channel by using either a bridge crossing over the river or a boat. The sample collection, preservation, and transportation to the laboratory were carried out as per the protocol given in APHA (2005).

The pH, Total Dissolved Solids (TDS), Electrical Conductivity (EC), and Dissolved Oxygen (DO) were measured at the time of sample collection by using a water and soil analysis kit (Labtronics, model- LT68). Samples were then transported in an icebox to the laboratory, except the sample collected for BOD analysis, as this required to be incubated at a higher temperature. The BOD concentration was estimated by a change in DO concentrations over a given period of 5 days in water samples at 20 °C (Delzer & McKenzie 2003). Water samples were analyzed for NH₄⁺ (Phenate method), NO₃⁻ (Manual Hydrazine Reduction method), PO₄³⁻(Ascorbic acid method), and H₄SiO₄ (Molybdosilicate method) after filtration with 0.45-micron nylon filter paper by the photometric method using a UV-Vis Spectrophotometer (Systronics) as given in APHA (2005).

The One-Way Analysis of Variance (ANOVA) and Pearson's correlation coefficient analysis were carried out using MS Excel 2019. The ANOVA analysis was used to determine significant spatial and seasonal variation in the dissolved nutrient concentration in the Ghaghara river system. In the case of ANOVA analysis, if the sum of squares for groups (SSC) value is higher than the sum of squares for error (SSE), the null hypothesis is rejected (Ostertagová & Ostertag 2013). The results obtained from ANOVA analysis were reported as (F_critical = F_calculated, significant level P) and if the F_calculated value is greater than the F_critical value, and P < 0.05 then the null hypothesis will be rejected. Pearson's correlation coefficient analysis was carried out to determine the linear relationship between water quality parameters based on the value of the correlation coefficient (r). The correlation can be positive, where an increase in one parameter leads to an increase in another variable also, and in the case of negative correlation, it will result in the decrease of another variable. A value above 0.5 represents a strong correlation and indicates a common source of origin or common pathway of transport in the watershed area. Factor analysis was carried out by using ‘Statistical Package for Social Sciences (SPSS), version-10.0’. Factor analysis utilizes five aspects of data analysis including data standardization by using correlation matrix, applicability evaluation by using the magnitude of the correlation coefficient, principal factor extraction based on the eigenvalue, principal factor interpretation by using factor rotation, and factor scores by utilizing regression equations and play important role in the determination of natural and anthropogenic factor responsible for the variation in the water quality (Jha et al. 2009; Gao et al. 2011). The study area map was prepared through the Quantum Geographical Information System (Q-GIS) software.

The chemical composition of Ghaghara river water is given in Table 2. To assess the impact of anthropogenic activities taking place in the catchment area, the water quality parameters were compared with surface water quality standards given for best-designated use (drinking) by the Central Pollution and Control Board (CBCB 2013) and Bureau of Indian Standards (BIS 2012), along with value reported for unpolluted rivers (in which the natural process of rock weathering, input from dry and wet deposition contribute to the water chemistry) (Table 2) (Meybeck 1979). The Ghaghara river water was generally alkaline in nature and increases in the downstream section of the river, but spatial variation was not statistically significant (p > 0.05) (Tables 2 and 3). The pH value falls within the range given by BIS for drinking water (Table 2). The water samples collected from Himalayan rivers are usually alkaline (Singh et al. 2005; Sharma et al. 2017). The pH value showed significant seasonal variation (p < 0.05) with a lower pH value during the monsoon season due to the addition of low pH water from precipitation (Table 3 and Figure 2(a)) (Saarinen & Kløve 2012). The EC and TDS measurements give the idea of the ionic concentration of a given water sample (Hem 1970). The Ghaghara river water was in the low salinity type class (1,500 μS/cm) during the study period (Table 2) (Kumar et al. 2016). The EC values showed a relatively higher value in the samples located in the upstream and downstream section of the river indicating the contribution from agriculture and built-up areas present (Table 1 and Figure 2(a)). The EC value was relatively lower in the monsoon season, indicating the dilution effect due to high discharge (Table 2 and Figure 2(a)). EC value doesn't show significant spatial and seasonal variation (p > 0.05) (Table 3). The TDS value observed during this study was in the range of 201–248 mg l⁻¹. Similar values had been reported by the earlier studies for the Ghaghara river (Sarin et al. 1989; Singh et al. 2017). The TDS load in the river system is attributed to the natural process of weathering along with the addition of sewage waste and runoff from agriculture and urban areas located in the watershed area (Table 1). TDS value doesn't show significant spatial and seasonal variation (p > 0.05) in the Ghaghara river but relatively higher TDS values were observed in the sample located in the downstream section, though this was well within the permissible limit for drinking water given by BIS (Tables 1, 3 and Figure 2(a)). The DO level in any aquatic system indicates the balance between oxygen-consuming processes and oxygen-producing processes at the time of sample collection and concentration above 4 mg l⁻¹ is required for maintaining the productivity of the aquatic ecosystem (Hem 1970; Shah & Joshi 2017). The DO values observed in this study were well above the permissible limit given by CPCB for surface water and were in the range (7.4–8.06 mg l⁻¹) reported from the earlier studies (Table 2) (Singh et al. 2017). The DO value showed significant spatial and seasonal variation (p < 0.05) in the Ghaghara river (Table 3 and Figure 2(a)). The oxygen concentration showed a declining trend along the river stretch, with the lowest value observed in the samples collected from the downstream region. The reason behind this low concentration can be attributed to high organic matter load from the high urban density areas and agriculture areas, which required oxygen for its decomposition (Table 1 and Figure 2(a)). BOD represents the amount of oxygen consumed by the microorganisms to degrade the organic matter present in the aquatic system and can be used as a proxy for the organic load in the river (Hem 1970). The BOD values observed in this study were in the range observed for the Ganga river basin (0.0 to 7 mg l⁻¹) and Ghaghara river (2.47 to 3.53 mg l⁻¹) (Table 2) (NEERI 2017; Singh et al. 2017). Based on the BOD values observed, Ghaghara can be placed in the clean to moderately polluted category (Table 2) (Hocking 2005). About 74% of the collected samples showed a BOD value higher than the permissible limit given by CPCB for surface water (Table 2). The increasing BOD load in the downstream section of the river is related to the increasing population density in the watershed area along with a contribution from agriculture runoff, which promotes instream phytoplankton growth and organic matter load (Table 1 and Figure 2(a)). The Uttar Pradesh Pollution Control Board (UPPCB 2019) had identified 50 km river stretches between Barhalganj, Gorakhpur, to Bhagalpur, Deoria, as critically polluted in the Ghaghara river under Priority V (3–6 mg l⁻¹) category, based on values of BOD due to discharge of 14 MLD untreated sewage dumped through 9 drains in this stretch. The BOD value showed significant spatial and seasonal variation (p < 0.05) in the Ghaghara river (Table 3 and Figure 2(a)).

Nitrogen and phosphorus are limiting macronutrients that are required for the growth of plant and animal cells. NH₄⁺ and NO₃⁻ are the most dominant species of dissolved inorganic nitrogen in the aquatic ecosystem. The average concentration of NH₄-N ion in the Ghaghara river was 3 to 8 times more than the unpolluted river, indicating the contribution from the addition of sewage and industrial waste, atmospheric wet and dry deposition, and runoff from agriculture and animal feedlot areas; however, concentration was within the permissible limit given by BIS (2012) (Table 2) (Miller et al. 2011; Du et al. 2017). The NH₄-N ion concentration showed significant spatial and seasonal variation (p < 0.05) in the Ghaghara river (Table 3). The upstream section of the river system had less population density and lower NH₄-N ion concentration, and as the population density in the watershed area increases the NH₄-N ion concentration also increases in the river (Table 1 and Figure 2(b)). The NH₄-N ion concentration showed the highest concentration during the pre-monsoon period, indicating the dominance of point source emission during the dry period, and as the discharge increased, the relative contribution from non-point sources started dominating in the study area (Yadav & Pandey 2017). Nitrate on average constitutes 77% of the total nitrogen load from the world's major rivers (Turner et al. 2003). Nitrate was the major contributor of dissolved inorganic nitrogen (DIN) load in the Ghaghara river and its concentration was 3 to 10 times higher than the concentration reported in the unpolluted river (Table 2). The observed concentration of NO₃-N was relatively lower than the average value reported for most of the South Asian rivers (Subramanian 2008). Peierls et al. (1991) showed that nitrate concentration strongly correlates with the watershed population density and a similar trend was observed in the Ghaghara river system, with a higher concentration observed in the downstream section having higher population density and agriculture area (Table 1 and Figure 2(b)). The seasonal behavior of the NO₃-N ion was opposite to the ammonium ion, indicating the role of the instream nitrification process, in which the NH₄-N ion gets oxidized into the NO₃-N ion along with runoff from agriculture and urban land during high discharge periods (Bernhardt et al. 2002). Nitrate concentration usually follows the seasonal variation pattern of river discharge with relatively higher concentration during the onset of monsoon and a low concentration of nitrate was reported during the low discharge period (Dinnel & Bratkovich 1993). The NO₃-N ion concentration showed significant seasonal variation (p < 0.05); however, the spatial variation was not statistically significant (p > 0.05) (Table 3).

The phosphate acts as a limiting nutrient in surface water bodies due to its tendency to precipitate along with the uptake for biological activity (Correll 1990). The concentration of PO₄-P showed a marked increase during the last decade that can be attributed to the increasing urban and agricultural growth in the study area during this period (Singh et al. 2017). The observed concentration of PO₄-P was relatively low compared to other Himalayan rivers like the Indus (150 μgl⁻¹), Ganges and Brahmaputra river systems (120 μg l⁻¹) (Ramesh et al. 2015). The PO₄-P ion concentration was relatively lower in the upstream region mainly due to the presence of forest as a major land-use type and in the downstream region cropland and urban are present, which contribute the PO₄-P ion to the river system (Table 1) (Fadiran et al. 2008). The relatively higher concentration of PO₄-P ion during the monsoon period indicated the contribution of phosphate from soil flushing and suspended particle transport from the catchment area (Table 2 and Figure 2(b)). The presence of agricultural areas in the river catchment is one of the major reasons behind the increases in PO₄-P ion concentrations during high precipitation months (Ide et al. 2019). The PO₄-P ion concentration showed significant seasonal variation (p < 0.05); however, the spatial variation was not statistically significant (p > 0.05) (Table 3).

Dissolved silica is a micronutrient that is essential for the growth of Diatoms, Radiolarians, and sponges in the aquatic ecosystem. Silica is mainly of geogenic origin and produced from the chemical weathering of rocks and soil present over the earth's surface; however, biogenic silica cycling can also be a major contributor to silica flux in the riverine system (Durr et al. 2011). The average silica concentration observed in the Ghaghara river during this study was lower than the value reported for the unpolluted river and can be attributed to the change in the land use land cover pattern in the study area along with the construction of dams and barrages in the river course, which reduce the transport of reactive silica through the river system (Chen et al. 2014) (Table 2). The concentration of SiO₂-Si in the Ghaghara river was in the range of 784–6,692 μgl⁻¹ for the headwaters of the Ganges and its tributaries (Bickle et al. 2003). The silica concentration doesn't show significant spatial and seasonal variation (p > 0.05); however, the concentration was relatively higher during the monsoon period, associated with the increased transportation of suspended material (Table 3 and Figure 2(b)) (Yadav & Pandey 2017).

Statistical analysis, including Pearson's correlation coefficient and Factor analysis, was utilized to delineate factors controlling the chemistry of the Ghaghara river. The result of Pearson's correlation coefficient analysis is given in Table 4. The population density showed a positive correlation with the BOD and NH₄-N concentration and a negative correlation with DO concentration, indicating the contribution of sewage waste from the catchment area (Table 4). The pH value and DO concentration showed a negative correlation during the pre-monsoon season and monsoon season, which can be attributed to the reduction in the algal population responsible for photosynthesis with an increasing pH value of the water (Zang et al. 2011). EC and TDS showed a significant positive correlation in the Ghaghara river (Table 4). The DO and BOD value showed a negative correlation during the pre-monsoon and monsoon seasons in the Ghaghara river, indicating the effect of organic load and its subsequent mineralization (Table 4). The BOD value showed a positive correlation with the NH₄-N and PO₄-P concentration in the pre-monsoon period, indicating the effect of the addition of sewage waste along with runoff from the agriculture area (Table 4).

R-mode factor analysis by a factor having an eigenvalue of more than one was carried out to understand the factor responsible for the nutrient chemistry in the Ghaghara river, and results are given in Figure 3. Three factors were identified during the pre-monsoon season, which explains 74% variation in the observed data set in the Ghaghara river. Factor 1 explained 39% variation in the observed data set and showed positive loading for the pH, BOD, NH₄-N and PO₄-P, and negative loading for DO concentration, indicating the contribution of organic matter from sewage waste in the study area (Figure 3(a)). Factor 2 explained 22% of the variability observed in the data set and showed positive loading for the EC and TDS, and can be attributed to the contribution from the chemical weathering in the catchment area. Factor 3 showed positive loading for SiO₂-Si and negative loading for the NO₃-N, indicating the instream production and subsequent decomposition of diatoms (Figure 3(a)). Three factors explaining 68% of the total variance in the data set were identified during the monsoon season (Figure 3(b)). Factor 1 explain 27% variation in the observed data set and showed positive loading for the BOD, NH₄-N, NO₃-N, and PO₄-P and negative loading for DO concentration, indicating the contribution from point and non-point sources present in the study area (Figure 3(b)). Factor 2 contributes 23% variability observed in the data set and showed positive loading for the EC and TDS values, and can be attributed to the weathering processes taking place in the catchment area along with the input from precipitation (Figure 3(b)). Factor 3 showed negative loading for pH and positive loading for the SiO₂-Si and DO concentrations in the study area (Figure 3(b)). Three factors accounting for 64% variation in the data set were identified during the post-monsoon season. Factor 1 accounts for 26% variation in the observed data set and showed positive loading for the TDS EC, BOD, and NO₃-N concentrations, indicating the contribution from the chemical weathering along with the organic matter production and its subsequent mineralization in the stream channel. Factor 2 explains the 19% variability observed in the data set and showed positive loading for the pH and DO value in the study area. Factor 3 showed positive loading for the PO₄-P and SiO₂-Si, indicating contribution from the leaching of soil and sediment present in the catchment area and river channel (Figure 3(c)).

The molar ratios (N:Si:P) have been widely used to estimate the nutrient limitation of primary productivity in the aquatic ecosystem. If other macro and micronutrients are present in sufficient amount, the N: Si: P ratio of 16:15:1 is required for optimum growth of phytoplankton and diatoms (Redfield et al. 1963; Turner et al. 2003). In the Ghaghara river, the DIN: DIP molar ratio ranged from about 3 to 184, with an annual average value of 40 ± 42. The DIN: DIP ratio calculated for the Ghaghara River was 2.5 times higher than the Redfield ratio, and this indicated the P limitation in the Ghaghara River (Figure 4(a)). DIN:DIP ratio greater than 16:1 has been reported from 77% of the world's larger river indicating increased loading of nitrogen due to anthropogenic sources (Turner et al. 2003). The downstream section of the Ghaghara river showed a relatively higher loading of nitrogen relative to phosphorus, as indicated by a higher DIN: DIP ratio in the downstream section (Figure 4(a)). The observed value of the DIN:DIP ratio was relatively higher than the value reported for the Yamuna river, which is another important tributary of the Ganga river and is mainly due to the lower phosphate loading in the Ghaghara river system (Sharma et al. 2017). Li et al. (2017) reported a DIN: DIP ratio in the range of 15 to 330, with an average value of 101 in the Pearl river estuary system, which is drained by the Pearl river, the third-longest river and having relatively higher loading of nitrogen in the watershed area.

The DSi:DIN molar ratio ranges from 0.3 to 26 with an annual average value of 4.6 ± 4.4 in the Ghaghara River. The average DSi:DIN molar ratio in the Ghaghara river was 4 times higher than the 1:1 ratio required for the conducive growth of diatoms, which are important components of the aquatic food chain system (Turner et al. 2003). The average DSi:DIN ratio showed a declining trend in the midstream and downstream section of the Ghaghara river, which indicates the increased loading of nitrogen and low silica input. It may be due to increased biological uptake and retention from the construction of dams and barrages on the river (Figure 4(b)) (Turner et al. 2003). The DSi:DIN ratio is showing a declining trend in most of the receiving coastal areas of the world due to increasing nitrogen loading from the rivers (Turner et al. 2003). Sharma et al. (2017) and Li et al. (2017) reported DSi:DIN ratios of 0.1 to 16 with an average value of 2 and 0.16 to 3.25, with an average value of 0.66 in the Yamuna river and Pearl river estuary system respectively.

Rivers act as a major transporter of dissolved nutrients to the ocean and contribute about 16.2 Tg yr⁻¹ NO₃-N, 2.6 Tg yr⁻¹ PO₄-P, and 194 Tg yr⁻¹ SiO₂-Si to the world's oceans (Turner et al. 2003). The specific yields (t km⁻²yr⁻¹) of nutrients were calculated using the nutrient concentrations, annual discharge, and drainage area and its comparison with other major river systems (Table 5). The annual average discharge and drainage area of the Ghaghara river at Revilganj, Bihar, near its confluence with the Ganga river was 94.4 km³ year⁻¹ and 128,000 km² respectively (Rai et al. 2010). The total annual transport of dissolved inorganic nitrogen (DIN) from the Ghaghara river was 7.33 × 10⁴ t yr⁻¹. NO₃ constituted about 86% of DIN load transported by the Ghaghara River, which is contrary to the River Yamuna where NH₄-N contributes 84% of DIN load (Sharma et al. 2017). The difference in the dominant inorganic nitrogen species is mainly due to the dominance of agricultural land use in the Ghaghara river basin, and runoff from agricultural land contributes nitrate to the river system (Singh et al. 2017). The specific yield of nitrogen from the Ghaghara river was lower than the Yamuna river system, mainly due to the lower urban population percentage in the catchment area but higher than the rivers located in the South and Central part of India including the Godavari, Cauvery, and Narmada rivers (Table 5). The watershed areas of the Lena and Yukon rivers draining to the Arctic ocean were 23 and 6.67 times larger than the Ghaghara river. However, the DIN flux was more in Ghaghara river owing to the higher population density in the catchment area (Table 5) (Holmes et al. 2000). The annual flux of PO₄-P and SiO₂-Si was 0.42 × 10⁴ and 12.34 × 10⁴ t yr⁻¹respectively from the Ghaghara river. The specific yield of phosphate and silica was lower than the value reported for the Yamuna river system, but nitrate loading in the Ghaghara river catchment was relatively higher (2.7 times) and can be a major factor behind the eutrophication problem in the Ghaghara river (Table 5).

The N-ICEP value was 1.71 kg C·km⁻²day⁻¹ and P-ICEP was −2.93 kg C·km⁻²day⁻¹ for the Ghaghara river. The increase in N-ICEP is mostly associated with anthropogenic activities like agriculture and sewage discharge from urban areas (Garnier et al. 2010). The positive value of N-ICEP indicated the presence of excess nitrogen over silica in Ghaghara river water, which may favor the growth of non-diatoms phytoplankton species and eutrophication in the receiving water of River Ganga. The excess loading of nitrogen from the watershed area should be reduced by judicious use of nitrogen fertilizer in the agriculture field and provision of natural or constructed wetlands between the farmland and receiving water bodies. The negative value of P-ICEP indicated the limitation of phosphate for the phytoplankton growth in the Ghaghara river. Pedde et al. (2017) reported positive ICEP values for the year 2000 in the Bay of Bengal and projected a further increase in the ICEP value for the year 2050. Trivedi & Trivedi (2014) reported an increase in nutrient loading and eutrophication condition in the stretches of the Ganga and its tributaries. Prajapati et al. (2020) showed the presence of the eutrophic condition in the 520 km long middle stretch of Ganga River between Kannauj to Varanasi due to input of dissolved nutrients from the point- and non-point sources. Santy et al. (2020) reported eutrophic conditions in a 238 km long stretch of the River Ganga in changing river flow and land-use scenarios. Siddiqui & Pandey (2019) reported eutrophic conditions in the middle and lower section of the Ganga River due to the presence of high population density and input from point and non-point sources. In a river system, even if the nutrient concentration is high, the water velocity during the high flow condition will wash away the algae produced to the downstream regions and minimize the risk of eutrophication (Shen et al. 2013). Remmal et al. (2017) reported planktonic blooms in the Yamaska River during the low river discharge period. Climate change further aggravates the risk of eutrophication due to a reduction in the river discharge condition (Charlton et al. 2018). Santy et al. (2020) reported a 13.7% reduction in the river flow in the middle section of the Ganga River from 1968–1990 to 1991–2012, indicating the increased risk of eutrophication. All these studies suggest that the River Ganga is prone to eutrophication risk associated with high nutrient discharge from anthropogenic sources during the low discharge period and input of nutrients from the Ghaghara river can further aggravate the risk of eutrophication.

Water samples were collected and analyzed to understand the nutrient dynamics in the Ghaghara river. Nitrate was the major component of the dissolved inorganic nitrogen (DIN) load in the Ghaghara river, contrary to ammonium, which was the major contributor of DIN load in the Yamuna river. Nutrient concentrations showed significant seasonal variation in the Ghaghara river during the study period. Pearson's correlation coefficient and Factor analysis indicated the contribution from the allochthonous sources (domestic sewage and runoff from the agriculture areas) along with autochthonous sources (instream production of organic matter and its mineralization) as the major controlling factor of nutrient chemistry in the Ghaghara river. The higher DIN:DIP molar ratio indicated the phosphate limitation for primary productivity in the Ghaghara river. The declining ratio of DSi:DIN along the Ghaghara river stretch indicated the increased loading of nitrogen in the river system. The nitrogen transport from the Ghaghara river has the potential to create a eutrophication problem in the receiving water bodies, as indicated by the positive value of N-ICEP, and remediation measures are required to reduce the loading of nitrogen from the non-point sources.

Pawan Kumar Jha thanks the University Grant Commission (UGC), India for providing funding (No.F.30-373/2017(BSR) for research work presented in this paper. The authors thank UGC for providing Rajiv Gandhi National Fellowship (RGNF) to Nirdesh Kumar Ravi and CRET fellowship to Atul Srivastava. The authors thank the University of Allahabad, Prayagraj, and Banaras Hindu University, Varanasi, for providing the necessary facilities. Kirpa Ram thanks the Department of Science and Technology and SERB, Govt. of India, for providing financial support under the ECRA (#ECR000490) program. Authors thank the editor and anonymous reviewers for their valuable suggestions.

All relevant data are included in the paper or its Supplementary Information.