Water is an essential element for human and animal survival. Its availability in sufficient quantity and quality contributes to the maintenance of human health (Makoutode et al., 1999). Unevenly distributed over the earth, this natural resource constitutes a major environmental issue (Dovonon et al., 2011). Yet, a combination of factors including agricultural exploitation, poor management of water resources, and population growth significantly reduces its availability and accessibility (Ouattara et al., 2018). In addition to these anthropogenic factors, climatic factors related to global changes considerably disrupt the availability of water resources. Moreover, rapid urbanization, industrialization and the non-rational use of fertilizers and pesticides generate various pollutants which cause numerous disturbances to aquatic environments (McKinney, 2002). This can affect the physico-chemical and biological quality of the receiving aquatic environments (Mulliss et al., 1997). Water quality is therefore becoming one of the permanent concerns of mankind, which has devoted an entire legislation and ratified many international protocols and conventions to ensure its protection and preservation (Kahoul et al., 2014).

 

The Bandama River is the only river which has its watershed located entirely in Côte d'Ivoire (Iltis and Lévêque, 1982; Kouamé et al., 2011). Due to its north-south orientation, this river covers different climatic and biogeographic zones. Its lower course plays a major role in the conservation of aquatic biodiversity and is therefore classified as the only freshwater Key Biodiversity Area (KBA) of Côte d'Ivoire (Starnes and Darwall, 2021). Its mouth at Grand-Lahou is part of a wetland classified as a Ramsar site and contributes to the maintenance of an impressive wealth of fauna. Unfortunately, the main course of the Bandama River is under the influence of two large hydroelectric dams (Kossou, Taabo) and several hydro-agricultural and agro-pastoral dams (Traoré, 1996). The main riverbed is constantly subject to illegal gold panning. In its watershed, there are intense industrial and agricultural activities with massive use of chemical fertilizers (Lévêque et al., 1983). In the Grand-Lahou region, there are agro-industrial blocks of oil palm and rubber trees as well as a manganese mining industry. All these anthropic activities constitute real threats that could jeopardize the conservation of aquatic biodiversity. Thus, in order to assess the impact of this anthropic pressure on the ichthyological fauna, the present study was initiated in the estuarine zone of the Bandama River. It focuses on the analysis of the physico-chemical environment of the river, its level of contamination in metallic trace elements and the level of ecological stress of fish species. The results of this investigation could contribute to the establishment of a management plan for aquatic resources in the peripheral area of the Azagny National Park, but more importantly, these results could be used as references or management models for future work.

Study area

 

The estuarine area of the Bandama River is located in Grand-Lahou, between 5°17 - 5°90 North latitudes and 4°47 - 4°57 West longitudes. At the outlet, the Bandama River joins to the east with the Ebrié lagoon, through the Azagny canal. It is connected in the west to the Grand Lahou Lagoon. The Grand-Lahou pass is the only sea outlet for the lagoon system, the Bandama and the Boubo Rivers. This estuarine zone of the Bandama River constitutes the natural limit with the Azagny National Park and the urban zone of Grand-Lahou. Surveys occurred at 3 sampling sites (Ba1, Ba2 and Ba3) following a longitudinal pattern (Figure 1).

 

 

Data collection

 

The physic-chemical parameters (water temperature, pH, salinity, dissolved oxygen, conductivity, total dissolved solids) were determined monthly from March 2019 to February 2020 by using a multiparameter AQUA_Red Aquameter. The sediment samples were collected using the Van-Veen grab sampler. Once collected, the sediment was immediately placed in stomacher-type polyethylene bags, carefully sealed, labelled and placed in an electric cooler before being transported to the laboratory for cold drying and analysis. The dried samples were ground until all particles passed through a 63 µm nylon sieve after removal of coarse debris. The dried sediment samples were digested in a closed Teflon container with aqua regal (concentrated hydrochloric acid 4/1 v/v to concentrated nitric acid) according to McGrath and Cunlife (1985). All total heavy metal concentrations in the sediment samples were determined by a flame atomic absorption spectrophotometer (type VARIAN SpectrAA 110 or AAS). All measurements of the aforementioned parameters were performed in triplicate and the average of the three independent measurements was reported. The detection limits of AAS were 0.035, 0.013, 0.046 and 0.051 mg.kg^-1, for Mn, Fe, Pb and Hg, respectively in the sediment samples. Fish samples were captured monthly from March 2019 to February 2020 at gill nets (10, 25, 30, 35, 40, and 50 mm mesh size). Captured fish were identified from Paugy et al. (2003^a,b) and names were updated in Fishbase (Froese and Pauly, 2019). Identified fish were weighed to the nearest gram, measured in cm and then preserved in 10% formaldehyde.

 

Data analysis

 

Descriptive analysis was applied to data in order to highlight the central tendency (mean) and variation (standard deviation) of physicochemical variables. Kolmogorov–Smirnov test for normality at α = 0.05 showed that water variables data were normally distributed, and these were subjected to one-way ANOVA.

 

The geo-accumulation index (Igeo) initially described by Müller (1969) to assess the degree of contamination of sediments was calculated. It is calculated according to the relationship:

 

Igéo = log₂ (Cs / 1,5 Cref)

 

Where Cs is the concentration of metal n in the sediment and Cref is the background value of the metal. The constant 1.5 is the correction factor for the background due to lithogenic effects. Due to the lack of background values for trace metals in the study area, the geochemical background concentrations determined by Wedepohl (1995) in the upper continental crust were used. The levels of contamination according to Igeo values are shown in Table 1 (Müller, 1969).

 

 

The enrichment factor (EF) defines the number of times an element is enriched relative to its abundance in a reference material abundance in a reference material (earth crust). It makes it possible to discriminate between natural sources (Kinimo et al., 2018; Hakanson, 1980). According to Ghrefat (2006), metal concentration in sediments was normalized to metal concentrations of average shale. Fe and Al are widely used as normalizer. In this study, Fe is used as normalizer element. The EF index for a metal is defined as follows equation (Hakanson, 1980):

 

EF= (Cs/[Fe]s )/(Cref/[Fe]ref)

 

 

and Fe in the sampled sediment, respectively and Cref and [Fe]ref are background concentration in the UCC. Different classes of sediment enrichment in trace metals are shown in Table 2.

 

 

The Kruskall-Wallis K-test was applied to the dataset to assess the spatial variation of metal pollution levels in the Bandama River estuary.

 

The application of the ecological stress index such as the ABC curve allows the use of fish fauna to determine the environmental conditions existing in the aquatic environments (Hay et al., 1996). The ABC curve is defined as the average of the difference between the cumulative proportions in terms of biomass and abundance:

 

 

Where ABC = Abundance-Biomass Comparison Index; Bi = proportion in biomass of species i (ranked in descending order of proportion); Ai = proportion in abundance (number of individuals) of species i (ranked in descending order of proportion); N = total number of species observed.

 

The graphs obtained are curves that make it possible to distinguish schematically three phases in the evolution of a stand (Warwick et al., 1987): a non-stress phase, when the biomass curve is above that of abundance; a light stress phase, when the two curves almost overlap; and a high stress phase, when the abundance curve is above that of biomass.

 

The log-transformed linear model expressed by the following equation (Lévêque, 2006) was used to determine the length-weight relationship:

 

Log W = log a + b log TL

 

Where W: the weight of the fish in g and TL: the total length of the fish in cm. The constant "a" represents the intercept of the regression line and b the slope of the relationship. Student?s t-tests (ts) were used to test whether the slope "b" was significantly different from the theoretical value of 3 (p < 0.05). Thus, the ts value for each species was calculated according to the following expression (Zar, 1984):

 

ts = (b-3)/sb

 

Where b the slope and sb the standard error of the slope.

 

 

Where SW: the variance of the body weight, STL: the variance of the total length and n: the sample size.

The results of the measurements of the physico-chemical variables of the water are recorded in Table 3. For all these variables, the ANOVA test showed a significant spatial variation (p < 0.05) only for the electrical conductivity and salinity of the water, which are higher in station Ba3 (400.6 µS/cm; 16.76‰) and lower in station Ba1 (70.94 µS/cm; 11.86‰). The absence of significant spatial variation for the other variables indicates that this area has physico-chemical homogeneity. Electrical conductivity and salinity which follow an increasing gradient towards Grand-Lahou channel which connects the Atlantic Ocean and Tagba lagoon. This spatial variation of these parameters would be linked to the rise of marine waters from the Atlantic Ocean in the estuarine zone of the Bandama River. Salinity is a characteristic parameter of the Bandama River estuary. It varies between 11.86 and 16.76‰, whereas it is zero in the upper part of the river. Wognin et al. (2007) indicated that this salinity is linked to the upwelling of marine waters at the mouth of the river which can reach up to 70 km upstream. For these authors, the physico-chemical environment of the Bandama River estuary is influenced by the lagoon systems (Ebrié and Grand Lahou) and the marine seasons. The average range of water temperatures (27.86-29.39°C) measured in the estuarine zone of the Bandama River was found to be relatively close to the average values obtained on the upper part of the river by Aboua (2012) (27.60°C) and Lozo (2016) (28.05) by Kouassi et al. (2005) in the Ebrié Lagoon. In the Bandama River estuarine zone, the waters are acidic (6.20-6.41) while the upper part of this river indicates a more alkaline zone (pH = 7.59; 7.06) (Aboua, 2012; Lozo, 2016). The relatively lower pH values in the estuarine zone compared to the upper Bandama River could be explained by the acidification of the marine waters that seasonally ascend into the Bandama River estuarine zone through the Grand-Lahou pass (Koné et al., 2021). The dissolved oxygen level (5.70-6.62 mg/L) indicates a fairly oxygenated estuarine zone compared to the values obtained (2.06 - 7.5 mg/L) in the upper river by Aboua (2012). This oxygenation of the environment could be associated with the weak water flow in the estuarine zone. This would favour the development of phytoplankton, the main source of oxygen in the aquatic environment. High values of electrical conductivity were measured in the estuarine zone with a significant variation between stations B1 and B3.

 

 

The trace metal concentrations in the sediments of the different stations are summarised in Table 4. This result indicated an order of importance in the following sense: Fe > Mn > Pb > Hg > Cd for all the stations visited. The Kruskall-Wallis test applied to the data set did not indicate significant spatial variation. However, the results revealed very high concentrations of total mercury and manganese in the Bandama River estuarine zone. Analysis of the geo-accumulation values (Figure 2) for these heavy metals shows that the Bandama River sediments could be considered uncontaminated for Cd and Pb (Igeo < 0). The Igeo values for Hg (about 3) suggest that all sediments are moderately to highly contaminated with Hg and those of Mn (Igeo ≥ 4) reveal that all sediments are highly to extremely contaminated with Mn. According to Mir et al. (2021), heavy metals may come from natural and anthropogenic processes and end up in various environmental compartments (soil, water, air and their interface). The natural emissions of heavy metals occur under numerous and certain environmental conditions. Volcanic eruptions, sea-salt sprays, forest fires, rock weathering, biogenic sources and particles of wind-borne soil are included in these pollutants. The release of metals from their endemic spheres to different environmental compartments will lead to natural weathering processes. However, Cheikh et al. (2018) indicate that the high metal levels in the sediments could be due to human activities. The enrichment factor (EF) values show that the sediments are low in Pb enrichment, with EF values less than 1 while it is moderately enriched with cadmium with EF ranging between 1 and 4 (Figure 3). In contrast, sediments of Bandama River estuary are severely to extremely enriched in Hg and Mn with EF values more than 25. This trend of Hg and Mn enrichment in the Bandama River estuarine zone suggests an anthropogenic source while the presence of Pb and Cd in the sediments could be of natural origin. According to Mir et al. (2021), industries, irrigation, drainage, mining and metallurgical processes, as well as runoff contribute to the release of pollutants into various compartments of the ecosystem.

 

 

From the north of Côte d'Ivoire, the Bandama River drains its basin with important mining industries, vast agricultural exploitation and urban areas for which the estuarine zone becomes a real outlet. This important agricultural, mining and urban focus could explain the severe enrichment of Hg and Mn. These heavy metals may reach the body of the fish directly from water or sediments through the gills/skin of the fish or from the food/prey of the fish through its food canal.

 

The level of ecological stress of fish in the Bandama estuary was assessed using ABC curves (Figure 4). This result shows that the species abundance curve is above the biomass curve with a Clark index value lower than 0 (W= - 0.046). This result highlights the existence of high stress at the fish species level. In agreement with Whitfield and Elliot (2002), environmental degradation and fishing pressure, heavily practiced in the Bandama estuary, may be the cause of the ecological stress observed on the fish fauna.

 

 

The length-weight relationship of the 20 species caught in the Bandama estuary was studied in the present study (Table 5). These results showed that the coefficient of determination (r²) ranged from 0.552 (Schilbe mandibularis) to 0.985 (Liza falcipinnis). Besides, 55% of the LWRs had r2 values higher than 0.80, 20% had r2 values between 0.80 and 0.70, while 25% had r2 values lower than 0.70. The estimates of b ranged from 1.530 for Chrysichthys maurus to 3.970 for Pellonula vorax with a mean value of 2.590 (SD = 0.576). The kind of growth, determined by Student's t-test, revealed that three species, Polydactylus quadrifilis, Awaous lateristriga and Sarotherodon melanotheron showed isometric growth (b=3). For the other species, b was significantly (Student t-test: p < 0.05) different from 3. Two species (P. vorax and Heseptus odoe) showed positive allometric growth (b>3) and the 15 last ones a negative allometric growth (b<3) (Table 2). It is true that several reasons can explain the negative allometric growth observed in these 15 fish species. However, it is noted that most of these fish are fish of economic interest to the local communities. Therefore, it is likely that the negative allometric growth observed in most of the fish in the Bandama estuarine zone is the cumulative effect of environmental degradation and artisanal fishing, which is the main source of income for local communities living along the Bandama River mouth of the Grand-Lahou lagoons.

 

The authors have not declared any conflict of interests.

This research was funded by the PASRES « Programme d'Appui Stratégique à la Recherche Scientifique ». They thank the OIPR « Office Ivoirien des Parcs et Réserves », the Azagny National Park management and the University of Félix Houphouët-Boigny for their technical support.