Impact of environmental features on the distribution of Marcusenius species (Pisces, Mormyridae) in Malebo Pool of Kinshasa in the Congo River

The current anarchic exploitation of Marcusenius fish in the Malebo Pool located in the Congo River stands as a threat of extinction to certain species, while the ecology of most of them are hardly known. This study aims at evaluating the impact of environmental factors on the distribution of Marcusenius fish caught in six fishing stations in the Malebo Pool. Canonical correspondence analysis of environmental factors classified the stations into two groups, which differ in vegetation formation, bottom substrates, water current velocity, pH, suspended particles and in dissolved oxygen. Thus, the five stations located upstream and in the middle of the pool are characterized by low water current velocity, slightly acidic pH, bottom substrates composed of silt, plant debris, clay, silt and sand. The predominant Echinocloa and Eichhornia plant formations found in these sites provide shelter, food sources and impact on the distribution and abundance of seven Marcusenius species in Malebo pool. The rocky substrate of Kinsuka station associated with the plant formation dominated by Ledermanniella and Pennisetum, and a high-water current constitute a habitat that is not favourable for Marcusenius species.


INTRODUCTION
Currently, ichthyological resources are subject to anarchic exploitation in Malebo Pool by fishermen from several fishing camps along the River in Kinshasa, such as Ngamanzo, Kimpoko, Kinkole, Kingabwa and Kinsuka pêcheur. These camps are crowded areas exerting great pressure on the aquatic environment (Tembeni et al., 2019), and constitute a threat of extinction to most of the species whose ecology is little known. Knowledge of the ecology of species is essential to ensure proper management of fisheries resources. It is the basis for the understanding of their dynamics population and life history strategies (Tembeni et al., 2019). The rational exploitation and conservation of this fishery resource in Malebo Pool can be properly planned (Zango et al., 2016). Fishes of the Mormyridae family are not exempted from this pressure. Mormyridae are nocturnal with small eyes covered by a thick skin and generally live in murky, dark and muddy waters. They have electrical organs of muscular origin located in their caudal peduncles that emit electrical signals to locate prey, objects and communicate with other species (Stiassny et al., 2007, Decru et al., 2019. This family, which is particularly diversified in Malebo Pool, represents 20 to 22% of the catches using passive nets after 12 h a day of fishing effort (Mbadu, 2010). These fishes are divided into 12 genera and 47 species, 8 of which belong to the genus Marcusenius (Kramer et al., 2016;Manda et al., 2018;Tembeni et al., 2019).
Most of the studies done on fishes of the genus Marcusenius have been mainly devoted to taxonomic and systematic aspects (Adjibade et al., 2020;Stiassny et al., 2007;Maake et al., 2014;Decru et al., 2016;Kisekelwa et al., 2016;Decru et al., 2019) and few to biological aspects (Lévêque and Paugy, 2006;Amira et al., 2019;Ndinga et al., 2021;Seraphin et al., 2022). However, there have been no in-depth studies on ecology on these species of the genus Marcusenius living in Malebo Pool. Suggestions were made therefore, to study and classify the habitats of the seven species of the genus Marcusenius (M. macrolepidotus, M. monteiri, M. schilthuisiae, M. stanleyanus, M. greshoffii, M. moorii and M. fuscus) out of the eight cited by Hanssens et al. (2008), living together in Malebo Pool, which would contribute to a better understanding of this high diversity in this large section of the Congo River.

Study sites
The study was conducted in Malebo Pool which is located on the lower middle reaches of the Congo River between Kinshasa and Brazzaville at 4°20' South latitude and 15°30' East longitude ( Figure  1). It belongs to the ichthyo-geographic province of Congo (formerly Zaire) which contains the entire Congo River basin, Lake Tanganyika and Lake Kivu (Lévêque and Paugy, 2006).

Fish sampling and identification
The fish samples were collected using the passive gillnet fishing technique organized once a month from July 2020 to June 2021, during the dry and rainy seasons in six sites: Kabongo, Mipongo, Molondo, Japon, Kingabwa and Kinsuka (Figure 1). To limit the biasness of the selectivity of gillnets  a battery of nets with mesh sizes of 15, 25, 35, 50, 80 and 100 mm between nodes, 50 to 100 m long and 1 to 2 m high was used. After every fishing campaign, Mormyridae fish were separated from the other species and grouped by genus and identified using the key of Stiassny et al. (2007). Species of the genus Marcusenius were identified using the key of Boden (Decru et al., 2016).

Measurement of environmental variables
For every field trip, environmental factors were measured at each station. The fishing stations were georeferenced using a Garmin map 62 st GPS. Surface water temperature, pH, conductivity, salinity and turbidity were measured in the field using a multiparameter Oyster series, pH/Conductivity/TDS/Salinity/ORP Meter 341350 A. Turbidity was measured using the Secchi disk. Water depth was measured using the Math Dip device. Water current velocity was measured by timing the movement of a 0.5 L plastic bottle half full of water over a 7 m distance (Djidohokpin et al., 2017). Dissolved oxygen was measured using the Dissolved oxygen meter AZ 8403. Water samples were collected for the laboratory analysis of Ca +2 , Mg +2 , K + , SO 4 -2 , PO 4 -3 and NO 3 using a spectrometer Hach DR 2400.

Substrate analysis and characterisation of aquatic vegetation
Substrates collected from each station were sent to the University of Kinshasa, Faculty of Agricultural Sciences' soil laboratory for particle size analysis based on Pauwels method (Mulaji et al., 2016). 100 g of each dried sample was mechanically sieved from different trunks. The quantity retained on every sieve was weighted on a 0.0001 g precision analytical balance (Orhus®). The characteristic plant species of every fishing site were collected and identified based on the APG system (Angiosperm Phylogeny Group) in collaboration with the botanists at the UNIKIN Herbarium. Observation of the plant formations made it possible to establish the list of species present in every site and quantitative evaluation based on the scale of abundance-dominance of Braun-Blanquet (Mucina et al., 2016). Subsequently, the abundance-dominance coefficients of the dominant species (0; 1; 2; 3; 4 and 5) corresponding respectively with a partial or total cover (< 1%, 1-5%, 5-25%, 25 -50%, 50 -75%, 75 -100%) were recorded using Excel® file along with the other environmental parameters. The resulting table was imported into Past Version 2.17c Software for analysis.

Canonical correspondence analysis
Canonical Correspondence Analysis was performed using Past version 2.17c software in order to identify the stations that have similar environmental features. These features can explain the distribution of fish species of the genus Marcusenius in different stations (Tessier et al., 2016).

Species diversity indices of Mormyridae and Marcusenius
At the end of every fishing campaign, the list of species of genus Marcusenius caught including the total number of fish was established. To categorize this set of Marcusenius in Malebo Pool, several diversity indices were evaluated. Shannon diversity index "H'" (Tembeni et al., 2019) is expressed by the following relation: It is maximal when the species have identical abundances in the population and minimal when a single species dominates the entire population.
Regardless of the species richness, it is very helpful to compare potential dominance between the stations and the sampling dates. For this, the Simpson's diversity index is used: Whereby fi = ni/N and N the total number of ni individuals of the given species. 1 stands for maximum diversity and 0 for minimum diversity.
For the purpose of this study, these indices were calculated using Past version 2.17c and Microsoft Excel 2010 software.

RESULTS
The results summarized in Table 1 show that the average pH of the water varies between 6.2 and 6.8, and temperature between 27.6 and 28.  Table 3.
A total of 403 Marcusenius belonging to 7 species were caught with gillnets in Molondo and Mipongo stations, which are respectively the first and second most

Physico-chemical features of the water
Variation in mean water temperature can impact the abundance and temporal distribution of fish species (Mahamba et al., 2018). Mean temperatures measured ranged from 27.6 to 28.2°C. Low temperatures were recorded during the dry season and high temperatures during the rainy season but do not justify a heterogeneous distribution of Marcusenius species. Generally, it is a small fluctuation of the water temperature under the influence of the air temperature. The pH measured in all the stations is slightly acidic and close to neutral. It varies from 6.2 to 6.8. The lowest values were recorded in Molondo station, while the highest values were recorded in Kinsuka. This is probably due to the proximity between Molondo station and Mbamu Island with a forest flap and the dead leaves could lower the pH and produce macro invertebrates which some Marcusenius species consume. On the contrary, the slightly higher values observed in Kinsuka could be related to the enormous wastes from NDjili River, Funa River and the industries close to the Kingabwa fishing station presented in Figure 1. It is also the reason electrical conductivity of these waters in Kingabwa is slightly higher (26.7 µS.cm -1 ) compared to Molondo and Mipongo stations (22.1 µS.cm -1 ). Given that dissolved oxygen is an ecological factor that can impact the distribution of Mormyridae fish, its variation impacts the species distribution (Tessier et al., 2016). Dissolved O 2 values measured at all stations varied between 5.9 -8 mg.l -1 , but the highest values were observed in Kinsuka. This station is characterized by high water velocity (0.45 m.s -1 ), favouring water mixing and very good oxygenation. The greatest depths were recorded in Kabongo, Molondo and Mipongo. Shallow depths were recorded in Kinkole, Kingabwa and Kinsuka. These depths impacted the distribution of specimens in the different fishing stations. Small specimens were caught at shallow depths while large specimens at high depths. Water current speed is undoubtedly one of the essential features underlying the plant and animal population distribution (Mahamba et al., 2017).

Habitat features
The five stations located upstream and in the middle of the pool are characterized by weak water current, plant formations with predominantly Echinocloa pyramidalis and Eichhornia crassipes; and substrates consisting of silt, plant debris, clay and silt. Distinctively, Kinsuka station has high-water current, plant formation with predominantly Penisetum and Ledermanniella; a rocky, gravelly and sandy substrate. The nature of this substrate was also observed by Pwema et al. (2011). Vegetation plays an important ecological role in the formation of micro-habitats, substrate modification and maintenance of physico-chemical features. These micro-habitats are resources for invertebrate development, the main food for Mormyridae fishes (Tessier et al., 2016) and Mahamba et al., 2017).

Distribution of Marcusenius in Malebo Pool habitats
Concerning the distribution of the seven Marcusenius spp in the pool, results showed that they are not uniform. This  seems to be influenced by environmental features that characterize each site. Hence, according to these features, Malebo Pool ecosystem is subdivided into two sub-systems: 1. The lentic environment which includes Kabongo, Kinkole, Kingabwa, Mipongo and Molondo stations. This sub-system is characterized by weak water currents, muddy substrate, less sand and clay, and vegetation dominated by Echinocloa and Eichhornia. The nature of the bottom substrates impacts the distribution of plant formations which characterizes each station; it depends on these habitats and constitutes essential food resources for Marcusenius spp. Marcusenius prefer and live in this lentic subsystem of Malebo Pool. These stations are also characterised by dark waters with low transparency and mud with plant debris. These results agree with those of Stiassny et al. (2007) and Decru et al. (2019) that Mormyridae are nocturnal and generally live in dark, muddy, turbid waters. 2. The lotic environment located downstream of Malebo Pool includes Kinsuka station. It is characterized by highwater velocity, rocky substrate associated with poor flora. These features are probably the basis for the scarcity of Marcusenius species in the lotic subsystem of Kinsuka.
Marcusenius can thrive in a variety of environments (lakes, rivers, streams) with a preference for the lotic environment from which they originate (Hopkins, 1986); these environments provide the necessary habitat and food resources for their survival. In Malebo Pool, the lotic environment (Kinsuka site) is made up of habitats where the substrate does not favour the development of vegetation under which prey can grow. That is why these fishes live in specific types of environment.

Conclusion
The analysis of the correlation between the characteristic environmental features of the six fishing stations in Malebo Pool and the seven species of fish of the genus Marcusenius investigated showed that the six fishing stations are divided into two groups based on the similarity of the environmental features analysed. These two groups differ from one another by the nature of substrate, characteristics of plant formations and speed of water current. These distinctive features affect the distribution and abundance of the seven Marcusenius species found in Malebo Pool.