Evaluating the influence of environmental variables on fish abundance and distribution in the Singhiya River of Morang District, eastern Nepal

: Monitoring the impact of fishing pressure on the Singhiya River is critical for resource development and sustainability, and the present situation is alarming and causing critical concern among the public. This study aimed to identify fish community trends over time and space in the river, and to investigate the impact of environmental variables on fish abundance and dispersion. Monthly fish sampling was performed from October 2020 to September 2021 from the 5 th to 10 th of each month. We used three cast nets of various mesh sizes (0.5, 2, & 4 cm) and monofilament gill nets with mesh sizes of 6, 8, & 10 cm. A total of 7,593 fish were collected, representing 61 species from seven orders, 20 families, and 37 genera. Similarity percentage (SIMPER) analysis revealed 78.8% similarity among six stations, with the primary contributing species: Puntius chola (28.2%), Puntius sophore (13.5%), Pethia ticto (5.33%), Chagunius chagunio (3.76%), Barbonymus gonionotus (3.69%), Puntius terio (3.46%), Opsarius shacra (2.2%), and Opsarius bendelisis (2.1%). Analysis of variance (ANOVA) on canonical correspondence analysis revealed that four of the seven selected environmental variables had significant relationship with the fish assemblage such as water parameters velocity, temperature, pH, and hardness. Overfishing and direct discharge of industrial waste into water resources may be the primary causes for the decline in fish diversity in Singhiya River.


INTRODUCTION
Freshwater bodies are vulnerable to habitat fragmentation, human encroachment, climate change, pollution, and biological invasions (Radinger et al. 2019). The combined effects of environmental pollution, unprecedented rates of biodiversity change, hydrological alteration, dam construction, and disconnection between the rivers and their lakes are possibly the largest threats to freshwater fish biodiversity (Huang & Li 2016). The diversity of the natural population is partially dependent on the environmental variables which always affect the competing populations (Chowdhury et al. 2011;Hossain et al. 2012). The factors influencing fish assemblages involve the environmental variables which are spatially heterogeneous and temporally variable and biotic interactions such as competition and predation (Harvey & Stewart 1991;Grossman et al. 1998). The environmental variables such as water velocity (Li et al. 2012;Adhikari et al. 2021;Limbu et al. 2021b), water depth (Kadye et al. 2008;Li et al. 2012;Mia et al. 2019;Chaudhary & Limbu 2021), substrate (Vlach et al. 2005;Yan et al. 2010), water temperature (Hossain et al. 2012;Nsor & Obodai 2016), and dissolved oxygen (Guo et al. 2018) all have been found to affect fish abundance and distribution in the rivers and streams.
In Nepal, few studies have looked at fish diversity and its link with environmental factors (Mishra & Baniya 2016;Limbu et al. 2020). Information on the relationship between fish community structure and environmental variables can aid in the preservation and management of aquatic biodiversity in the face of human-caused problems such as pollution and global climate change (Li et al. 2012). The Singhiya River has been altered due to several human encroachments such as settlements, factories, embankments, sand mining, electrofishing, damping and agriculture. To date, the space and time pattern of low-land, Terai region remains relatively unknown. Moreover, the details on fish community structure relating to their anthropogenic activities is also scanty. Facts about the relationship between fish community structure and environmental conditions can help us retain and lead aquatic biodiversity away from human-caused challenges like pollution and global climate change (Li et al. 2012).
The present study aimed to detect fish community patterns in the Singhiya River through time and space, as well as to evaluate the impact of environmental variables on fish abundance and dispersion. The current study expected that during the annual dry season, when water current and volume are reduced, fish abundance in the Singhiya River would be increased. We also hypothesized that the structure of fish assemblages will vary according to seasonal fluctuation defined by environmental variables.

Study area
Singhiya River is situated in the Morang district of Eastern Nepal (Figure 1). It is a perennial river that originates from the periphery of Hattimuda, Dulary, and Sundar Haraicha and surges through the Budiganga Municipality and Biratnagar Sub-metropolitan, and from the Buddhanagar it crosses the border of India. It lies in the latitude and longitude coordinates of 26.913 o N & 87.157 o E, respectively. The water of this river is mainly used for irrigation. The vegetation bordering the river is mixed, mostly consisting of bamboo and coniferous forest and the dominant river substrata consist of cobbles, pebbles, gravel, and sand. In total, six sampling stations were set up to gather fish. Residents settled along the entire river in the catchment, and numerous small and large factories were established in stations 1, 2, 3, & 4 whereas, stations 6 & 7 were set up close to the city.
The Singhiya River region experiences mostly sunny weather, with occasional clouds, and the water is muddy due to increased anthropogenic activities near the area of human settlement but crystal clear in its origin parts. The study area for this research includes 22 km of river basin starting from Hattimuda to Buddhanagar of Morang District.

Data collection, Identification, and Preservation
From October 2020 to September 2021, fish samples were taken every month. Sample collection started on the 5 th and continued to the 10 th of the selected month, i.e., October, November, December (2020), January, February, March, April, May, June, July, August, and September (2021). We made 72 samples at six stations, namely, (S1) Hattimudha, (S2) Puspalal Chowk, (S3) near Hanuman Mandir, (S4) Hatkhola, (S5) Jahda Bridge, and (S6) Buddhanagar, with fish sampling carried out between 0070 h and 0090 h. We employed three cast nets of various sizes, one with a mesh size of 0.5 cm, diameter of 5 m, and a weight of 2 kg, and another with a mesh size of 2 cm, diameter of 5 m, and a weight of 4 kg. A cast net with a diameter of 4 cm, a length of 7 meters, and a weight of 7 kg was also utilized. Cast netting was used to cover 150 m to 200 m across each J TT station, ensuring that all conceivable habitats were covered (Limbu et al. 2021b). For each cast net, a total of 10 throws were made over one hour. The fish were also caught using monofilament gill nets with mesh sizes of 6, 8, & 10 cm. Nine-gill nets were left late in the evening (1700h-1800 h) and pulled out early in the morning (0600h-0700 h) at a sample distance of 150-200 m at each station Fish sampled were photographed and identified in the field, and unidentified specimens were preserved in 10% formalin for later identification. The remaining samples were released to their original habitat after the photography. Standard fish taxonomy literatures (Talwar & Jhingran 1991;Jayaram 2010;Shrestha 2019;Fricke et al. 2021) and other available standard literature were used to identify the fish. During field visits environmental variables such as water temperature, dissolved Oxygen (DO), pH, total hardness, water velocity, alkalinity, and free carbon dioxide (CO) were investigated using the American Public Health Association's standard methodology (APHA 2012). A digital thermometer was placed in the water at a depth of 1 foot to measure the water temperature (°C). The Winkler titrimetric method was used to determine DO (mg/l). A pH meter was used to determine the pH (HI 98107, HANNA Instrument). The EDTA titrimetric technique was used to evaluate total hardness (mg/l). With the help of a stopwatch, a small ball and a measuring tape, water velocity (m/s) was determined using the float method. The alkalinity (mg/l) was measured using the titration method. The titrimetric method was used to detect free carbon dioxide (mg/l) using phenolphthalein as an indicator.

Data analysis
To examine potential variation over space and time a one-way analysis of variance (ANOVA) was used for temperature, pH, dissolved oxygen, hardness, and water velocity. To determine which means were significantly different at the 0.05 level of probability, a posthoc Tukey HSD test was used (Spjøtvoll & Stoline 1973). In the first step of data processing, the diversity of the fish assemblage was quantified, and then a statistical comparison was performed. Data on fish abundance were subjected to various diversity indices (Shannon, Simpson dominance, evenness, and species richness). All of the diversity indices were created using data from 12 months (each month, six samples were taken, S1-S6) and the data were used directly in the analysis, according to Magurran (1988) for each fish community sample. The Shannon diversity index (Shannon & Weaver 1963) takes into account both the number of species and the distribution of individuals within species. The Shannon diversity was calculated using the following formula: (1) Where S is the total number of species and Pi is the relative proportion of i th of species.
The Simpson index (Harper 1999) is a dominance index which gives more weight to common or dominant species.
The Simpson dominance index was calculated by using following formula: (2) Where n i is number of individuals of species i. The Evenness index (Pieleu 1966) measures how evenly or uniformly the relative abundances Pi (i=1..,S) are distributed across the S different species, irrespective of the value of S and the Evenness index was determined by the following equation: E= H'/ log S (3) Where, H' = Shannon diversity index S = Total number of species in the sample. In the multivariate analysis, rare species (<1%) were excluded in the analysis as they tend to affect multivariate analyses (Gauch 1982). Samples by species and environmental variables were analyzed through a multivariate analysis tool. Detrended correspondence analysis (DCA) (Hill & Gauch 1983) was performed to determine whether redundancy correspondence analysis (RDA) or canonical correspondence analysis (CCA) would be the most appropriate model to describe the association between species and environmental variables. The value of first axis length (3.14) and eigen value (0.53) obtained from DCA suggested that the uni-model associated with canonical correspondence analysis (CCA) (Ter Braak 1986) was more applicable. Therefore, a direct multivariate ordination method (Legendre & Legendre 1998) based on a linear response of species to environmental gradients was applied. Collected fish abundance and determined environmental variables were used directly in the multivariate analysis (Yan et al. 2010;Hossain et al. 2012;Vieira & Tejerina-Garro 2020;Tumbahangfe et al. 2021).
The one-way permutational multivariate analysis of variance (perMANOVA) (Clarke 1993) was used to determine whether there was a significant difference between the spatial and temporal scales of the collected fish data. A similarity percentage (SIMPER) (Clarke 1993) analysis was used to visualize the major contributing species in both space and time. Furthermore, Individual rarefaction analyses (Colwell et al. 2012), was performed across stations and months. All the statistical analysis were performed in R software (R Core Team 2019), 2.5-6 version.

J TT
Opsarius bendelisis (2.1%); 77.5% similarity was found between months, and the top contributing species were as listed above. Tables 3 & 4 show the results of diversity indices. The highest Shannon diversity index (2.79) was found at station 2 (S2) and in the month of August (2.94) whereas the lowest (1.76) was found at station 1 (S1) and in June (1.51). Analysis of variance (ANOVA) testing for both time and space revealed a significant (P <0.05) difference across six stations, but no significant (P >0.05) difference for the Shannon diversity index over twelve months. The highest Simpson dominance index (0.91) was found at station 2 (S2) and in the month of August (0.93) while the lowest Simpson index value (0.67) was found at station 6 (S6) and in the month of June (0.61). There was no significant (P >0.05) difference in the Simpson dominance index across the six sampling points and months. Similarly, the highest Evenness index (0.59) was at stations 1 & 2 and the month of August (0.59) whereas the lowest value (0.44) was found at stations 5 & 6 respectively, and the month of June (0.42). There was also no significant (P >0.05) difference in the Evenness index between the six stations and months. On the other hand, the highest Species richness value was observed at station 6 (S6) and in the month of October (36) while the lowest value was found at station 1 (S1) and the month of June (21). The species richness index differed significantly (P <0.05) between the six sampling

J TT
locations and months.

DISCUSSION
This is the first study to describe the spatial and temporal fluctuation of fish community structure in a Nepalese low-land river. The outcomes of this study will improve our understanding of the variance in fish communities for the benefit of Nepalese low-land river conservation, which recorded a total of 7,593 individuals, represented by 61 species belonging to seven orders, 20 families, and 37 genera. This suggests that the Singhiya River provides a significant source of livelihood and food to local fisherman and communities. The representation of Cypriniformes, Siluriformes, and Anabantiformes orders found in this study is consistent with the information reported in the other river systems of Nepal such as the Mechi River (Adhikari et al. 2021), Ratuwa River Phewa Khola (Limbu et al. 2021b).
The present findings revealed that the maximum number and diversity of fish species were collected in October, September, and November. During June and July, water velocity was found to be low and the water temperature was found to be high in the current study. Because of the low water velocity, the fishermen could do the most of the fishes. River discharge and water temperature have a much greater impact on the amount and diversity of fish (Kriauciuniene et al. 2019). Overfishing, industrial discharges, and sand mining may have impacted the amount and diversity of fish in the Singhiya River. Furthermore, essential aquatic ecosystem measurements such as species richness and diversity indices are influenced by changes in abiotic parameters such as river discharge and water temperature (Crane & Kapuscinski 2018;Parker et al. 2018 et al. 2021). River output appears to be influenced by water level variations caused by climate change and water management, as well as fishing pressure (Halls 2015). Monitoring the impact of fishing pressure on the Singhiya River's exploited fish population is critical for resource development and sustainability. The present situation in the Singhiya River is still sounding the alarm and causing critical concern among the public. As the biodiversity of freshwater fish keeps on decreasing mainly due to anthropogenic impacts, it is apparent that there has been a serious lack of scientific basis and truly ecological action for sound river basin management (Li et al. 2012 The Shannon diversity index takes into account the richness and proportion of each species, while the evenness and dominance indices reflect the relative number of individuals and the proportion of common species, respectively (Hossain et al. 2012;Yang et al. 2021). The highest Shannon diversity index (2.79) was identified at station 2 and in August (2.94), while the lowest (1.76) was discovered at station 1 and in June (1.51). A high Shannon diversity index is linked to a small number of individuals, whereas a low Simpson's diversity index is linked to a large number of individuals (Hossain et al. 2012;Temesgen et al. 2021). A biodiversity index seeks to categorize a sample's diversity (Magurran 1988) and is easily influenced by the number of specimens, sample size, and environmental factors (Leonard et al. 2006). The highest Simpson dominance index (0.91) was found at station 2 and the month of August (0.93), while the lowest Simpson index value (0.67) was obtained at station 6 and the month of June (0.61). Similarly, the highest evenness index (0.59) was observed at stations 1 and 2 and in August (0.59), while the lowest value (0.44) was recorded at stations 5 and 6 and in June (0.42). The maximum species richness (35.31) value was found at station 6 and the months of October (35.97), while the lowest (18.44) value was recorded at station 1 and the month of June (20.66). The species richness index varies considerably (P <0.05) between the six sampling locations and months. Overall, stations 2, 3, 4, 5, & 6 and the months of October, January, February, May, August, & September were likely to be rich with richness and diversity, because these sections were deeper and larger in terms of water depth and surface cover than station 1 section within the system. The river width and depth The information on the interaction between environmental variables and fish community structure can assist us in maintaining and managing aquatic biodiversity in the face of human-caused problems such as pollution, global climate change, and so on (Li et al. 2012). The influence of environmental variables on fish abundance, diversity, and distribution was checked by canonical correspondence analysis. In the current study, water velocity, water temperature, total alkalinity, pH, and total hardness are the major influencing factors (P <0.05) to determine the fish diversity, abundance, and distribution of the Singhiya River. Water velocity (Yu & Lee 2002;Yan et al. 2010;Adhikari et al. 2021), water temperature (Kadye et al. 2008;Temesgen et al. 2021), total alkalinity (Edds 1993Pokharel et al. 2018), pH (Pokharel et al. 2018Limbu et al. 2021b;Rajbanshi et al. 2021), and total hardness Shrestha et al. 2021) have also been found to be influencing factors to shape the fish assemblage structure.

CONCLUSION
The Singhiya River exhibits a good ichthyofaunal diversity, represented by 61 species of fish belonging to seven orders, 20 families, and 37 genera. Of 61 species, Puntius chola, Puntius sophore, Pethia ticto, and Barbonymus gonionotus were the dominant fish species recorded in Singhiya River. However, commercially important species such as Labeo catla, Bagarius spp., Chitala chitala, Sisor spp., and Notopterus notopterus were not recorded during the study period. Thus, conservation of these species has become urgent in Singhiya River. Overfishing and direct discharge of industrial waste into water resources may be the primary causes for the decline in fish diversity in Singhiya River. Therefore, practices like dumping of industrial waste, overfishing, and sand mining should be minimized, monitored, and if required, prohibited to protect the Singhiya River's aquatic flora and fauna and natural ecology. The canonical correspondence analysis suggested that an important environmental variables in structuring the fish community in the Singhiya River were water velocity, temperature, pH, and hardness. Lastly, the current study, in conjunction with the preceding examination, could serve as a baseline scenario for future analysis of the Singhiya River and other connected water bodies in the coming decades.