As Burkina Faso is arid and not provided with important natural waters, a policy of water resources development led to the creation of 1500 reservoirs which size varies from 0 to 25,000 ha depending on the season. The development of water resources resulted in the development of commercial capture fisheries [1], which employs 32,000 fishermen, 3400 sellers of fresh fish, 2300 sellers of processed fish and 3000 women fish processors [2]. The archives of the General Directorate for Fish Resources reveal that in tonnes, the national production of fish was 28,365 (capture fisheries: 27,803, aquaculture: 562) in 2019 and 29,752 (capture fisheries 29,105, aquaculture 647) in 2020. This source adds that in 2020, the country imported 146,995 tonnes, that is five times more than the national production; in 2020, the consumption of fish per capita was 8 kg. Increasing the national production of fish is therefore a requirement mentioned in many national development policies, such as the national plan for social and economic development and national strategy for sustainable development of capture fisheries and aquaculture by 2025.

The feasibility study of the large size reservoir of Ouessa on the most important river that is the Mouhoun is already done. Important roles are assigned to fishing and fish farming, but scientific knowledge lacks suddenly, which can compromise the management of the future reservoir fishery. Though a few works like [3] [4] [5] and [6] were recently done on the upper part of Mouhoun, they do not provide a comprehensive understanding of the state of the fish community of the upcoming reservoir. Moreover, in Burkina Faso, the fish communities are seldom studied before the creation of reservoirs, including the large size ones that are Toece (5000 ha large; created in 1995), Ziga (8000 ha; 2000), Sourou (10,000 ha; 1984), Kompienga (20,000 ha; 1988) and Bagré (25,000 ha; 1992). In Kompienga, such study was done by a foreign company, but the results were said to be for internal use only. However, we need reliable information on the fish community for conservation and sustainable management [7] [8] [9].

Therefore, the general objective of this piece of work is to provide reliable information on the ichthyofauna of the area of Ouessa on the downstream part of River Mouhoun. The specific ones are to describe the fish assemblages, the length-weight relationship, some indices of diversity of species, the condition factor, the yield per fish effort and some physico-chemical factors.

The future dam of Ouessa will be built on the Mouhoun River, formerly known as Black Volta, at a site that is close to the frontier with the Republic of Ghana. The site is located in the downstream section of the Burkinabe part of the most important river, an area that probably hosts a large number of fish species. According to [10], the reservoir will be 40000 ha large and will yearly produce 4500 t of fish and that currently the area hosts 1265 fishermen (of which 5% are women) who, together annually land 585.5 tonnes of fish.

The fish were sampled from 06 to 15 March 2020 with cast nets of 15 mm of mesh size and gill nests of 15 to 65 mm. Two professional fishermen were recruited for this purpose. In the water the fishermen were moving from downstream to upstream generally by canoe and exceptionally by foot. In total we covered about 35 km of reach that was divided into 7 sections (Bridge, Konkiri, Memère, Dianlé, Dike, Bon and Laro,) of about 5.8 km-long each (Table 1 and Figure 1). Each section was covered by a sampling session and the number of

throws of cast net were counted. Additionally, we sampled in the small size reservoir of Boura (32 km far from the Mouhoun) which water flows into the area that will be flooded by the future reservoir.

The fishes were sorted into species thanks to [11] and each fish was weighed at the nearest 0.1 g and its total length measured at the nearest 0.1 cm. Specimens of species difficult to identify were conserved in alcohol 90˚ for deeper examination in the Laboratory of Animal and Fish Biology and Ecology of the University Joseph KI-ZERBO of Ouagadougou. We also observed the landings of local commercial fishermen to look for species and size of specimen that we could not catch. We questioned local fishermen, about the names of species in dagara and gourounsi the two local languages in the area and their signification in dagara the most spoken language.

Between 07 h and 09 h that is the start of each sampling session, we measured the pH, the conductivity, the temperature and the dissolved oxygen thanks to a portable multimeter HANNA, Model COMBO and the transparency with a Secchi disc.

We calculated the frequencies of families, genera and species, the index of diversity of Shannon-Weaver ( H ′ ) and the index of eveness of Pielou (E).

H ′ = − ∑ P i ∗ ln P i

where the quantity p_i is the proportion of individuals found in the i^th species [12]. The values of H ′ usually range between 1.5 and 3.5 and only rarely exceeds 4.5 [13].

We calculated the index of evenness of Pielou (E) E = H ′ / log 2 ( S ) where H ′ is the value of Shannon index and S the taxonomic richness. This index denotes a balance relationship between species and its individual richness of a sample. It ranged between 0 and 1 which expresses the absolute distribution of relative abundance of species [14].

The length - weight relationship was appreciated thanks to [15]: P t = a L t b , where Pt and Lt are respectively the weight (g) and the total length (cm). The constants a et b are derived from the linearisation of the relationship by logarithmic transformation in the form of log 2 ( P t ) = log 2 ( a ) × b ∗ log 2 ( L t ) [16]. The growth of the species is isomeric when b = 3, allometric positive when b > 3 and allometric negative when b < 3.

The condition factor K of the most common 24 species was calculated: K = P ∗ 100 / L t 3 , [17] where Pt is the total weight in (g) and Lt the total length of the fish in (cm).

Finally, the yield per unit of fishing effort was also calculated, the unit of fishing effort being the throw of cast net.

In total we observed 1059 specimens sorted into 16 families, 27 generas and 42 species. At family level, the Mochokidae was dominant (44.3%), followed by the Alestidae (16.9%), the Mormyridae (12.90%) and the Cichlidae (6.7%). None of the remaining ones exceeded 5% (Figure 2).

The species were dominated by Synodontis nigrita (26.4%), followed by Synodontis schall (12.7%), Marcusenius senegalensis (9.7%), Brycinus nurse (7.8%) and Alestes baremoze (6.4%). None of the remaining ones exceeded 5% (Figure 3).

Many of the species did not have their names indicated by the local fishermen in the local languages (Table 2). The names are given with reference to the morphology, the colour or the ethology of the species. For instance, B. nurse is named Zourzier in dagara language, meaning that it has a red tail. Also Zu saala, the name of C. anguillaris in dagara means the fish slips easily through the fingers. For some species we were able to find the names, but not the meaning.

The values of index of Shannon-Wiener H' was 3.97 and the equitability E of Pielou 0.74.

For the 24 most frequent species the mean length and weight and the parameters of their relationship are given in Table 3. We estimated that the number of specimens of the 18 remaining species is too low to allow a reliable calculation of the LW relationship. We found a positive allometric growth for 13 species (b > 3; p < 0.05), a negative allometric growth for 10 species (b < 3; p < 0.05) and an isometric growth (b = 3; p > 0.05) for 1 species.

The values of the condition factor for the 24 species vary from 0.140 to 4.320, the lowest value being for H. forskali and the highest one for H. bebe (Table 4).

In terms of biomass, on average a cast net throw yielded 21.91 g, with the highest record at Konkiri (39.72 g) and the lowest one at Mémère (15.06 g). As for the number of fishes, the mean yield of throw was 0.45, with the maximum in Dianlé (0.76 fish) and the minimum at Boura and at the bridge (0.58 fish). The details are given in Table 5.

The mean content of dissolved oxygen was 6.70 mg/l with the minimum of 5.64 at Dianlé and the maximum of 7.9 at the dike, the mean value of pH 7.86 with the minimum of 7.45 at Boura and the maximum of 8.06 at Laro, the mean transparency 29.64 cm with the minimum of 17.1 at Laro and the maximum of 107 at Boura and the mean temperature in ˚C was 28 with the minimum of 27 at

Bon and the maximum 29.7 at Dianlé, the mean conductivity in µS/cm was 113.75 with the minimum of 110 at Bon, at the Dike, at Konkiri and at Memère and the maximum of 120 at Boura, Dianlé, Laro. The details are given in Table 6.