In January 2022, we carried out a count of floating cages in the bays. The results showed 218 cages. This number seems to be less than satisfactory in relation to the demand for aquaculture products in Bukavu. An increase in the number of cages would still be favorable to meet the demand of the urban population, estimated at 1,625,785 inhabitants as of 2019 (Ulengabo 2020). It was observed that of all cages, 156 are being exploited (i.e. 72%) and 62 not being exploited (i.e. 28%), which shows that farmers are abandoning cage cultivation despite the support of IITA. The farmers, which would make another analysis in the following days on the causes of abandonment of cage culture in Lake Kivu are not listed many farmers mention the high cost of the investment compared to the cost of production, the water quality and feed nutrition. However, Hambrey (2008) remarks that in Africa, failures have been frequent, especially in the case of support by the government or aid projects. This is partly due to the haphazard nature of cage culture, which requires encouragement, motivation and understanding of local conditions to succeed.
Even though the physical conditions of the studied bays are favorable for cage culture (CMFRI 2012), 62 cages are already not in use. These are cages that float on the surface of water without fish, which means a loss of potential fish yield, are also a source of chemical pollution from the materials from which the cages are made. Eliminating unused cages would be important to reduce their effects on the aquatic environment of the bays of Lake Kivu.
Our results show that only Ndendere Bay has the highest number of cages compared to the other bays (Fig. 2). This bay is much appreciated by the operators in view of the number of anglers who frequent it compared to the other bays. The security of their cage would be guaranteed by the involvement of the fishermen in the culture. The physical parameters collected inform us that few of the cages are less than 5 m wide (Fig. 3). The cages that are wider than 5 m are more numerous. The study by Zaghoudi (2013) on the impact of floating cage fish farming on sensitive coastal sites in the Khniss lagoon and the Kuriates islands in Tunisia shows that the location of the cages should not coincide with traditional fishing areas. In the bays of Bukavu, the cages can be kept at a distance to the coast of 21–50 m. These areas are less threatened by the concentration of plastic waste near the shoreline and would avoid the destruction of macrophytes. The international standards of the FAO (2004), placing the cages at a distance of 20–50 m from the coastal zone, are to be strictly observed in the bays of Lake Kivu. This strengthens the cohesion with anglers who are concentrated in the 2–10 m wide zone. The bays are not the same in terms of morphology explain the significate difference (p < 5%).
Cage culture is practiced in areas with a benthic depth of 5–20 m (CMFRI 2012, https://www.fao.org/3/x5744e/x5744e0f.htm). These authors report that this depth maximizes water exchange avoids wave formation and reduces the risk of contamination in order to avoid oxygen depletion, accumulation of debris and the build-up of certain harmful gases generated by the decomposition of waste deposited at the shore. All the studied cages in the bays of the Bukavu sub-basin are at a depth favorable for cultivation, ranging from more than 5 m from the benthic zone. This is because Lake Kivu has a rather shoreline than the other Albertine Rift lakes. Whatever the local cages when the submerged depth of the net is 4 m, the difference of the cage to the bottom zone is acceptable. The nature of the benthic bottom observed in the studied bays consists of mud, sand and plastic waste. This does not meet the criteria of an ideal site for cage culture, including a firm substrate with a combination of fine gravel, sand and clay (CMFRI 2012).
The cages used in the Bukavu bays are made of High Density Polyethylene (HDPE) but the type differs between partners. The local cages have a surface area of 20 m² and are rectangular in shape, generally 3 m wide and 6 m long, or a volume about 80 m3. These cages are made up of 8 plastic drums with a capacity of 250 litres and nylon nets for a platform of two cages, one of which is a cage for fingerlings with a 4-mesh net and the other a pre-growth net (6 meshes). This net is also used for grow-out in local cages. The latter are mounted with soft metal pipes x40x40. Our observations show that in some cages mosquito nets are used and in others cloths for the fingerling cages. These practices can reduce the circulation of oxygen in the cage. The farmers report that these cages are capable of holding 10,000 fingerlings.
The FPI/IITA cages have a surface area of 36 m², i.e. a volume of 216 m3, and are made of HDPE plastic pipes with a diameter of 200 mm. In the four studied bays of the Bukavu sub-basin, 4 cages make up a platform of which 3 are used: the first cage for fry rearing (4 meshes net), the second for pre-fattening (6 meshes net) and the last for fattening (7 mesh net). All cages with fish are covered by a protective net to avoid predators. They can hold 30,000 fry per platform. Compared to Lake Victoria, a study confirms that the viability of cage culture for Nile Tilapia is a function of large volume (Musa et al. 2021). This suggests that production would be low in Lake Kivu.
It is easily understood that cage culture in the bays of Lake Kivu lacks institutional arrangements and coordination between public sector institutions to establish the necessary regulations that can accommodate cage aquaculture development. The 2004 FAO recommendations on cage culture in Africa are guiding principles that would assist institutions responsible for managing marine and inland water resources to regulate aquaculture.
In the studied Bukavu sub-basin bays, the species Oreochromis niloticus is used in culture. Interview surveys with farmers show that 54% of fingerlings come from Rwanda and 46% from D.R. Congo. Some mentioned that the Rwanda fingerlings are preferred because they are mono-sex types compared to those from D.R. Congo. Others report of delays in the supply of fry from the Congolese causing the net to be left in the waters. Despite this preference, dependence is not significant. Observers continue to wonder about the real strain used to better monitor this culture. However, a recent study in the Bukavu basin revealed that the imported strains of Oreochromis niloticus, currently being cultured in cages in Lake Kivu, have six times more monogenic (new to the lake) parasites than the native strains of the lake (Muhindo 2020). Leakage is likely to occur, which could be a vector for the introduction of species into Lake Kivu (Rana & Telfer 2008) and the risk of an invasion of these new parasites that would decimate Oreochromis niloticus species and other fish native to the lake, if these groups proliferate (Muhindo 2020). Control fisheries should be encouraged in habitat monitoring and in the use of a strain native to Lake Kivu. This strain would have little impact on the aquatic environment.