Abstract

This study was carried out with the aim of reducing the cost of aquaculture feeds in Côte d'Ivoire. Therefore, three (3) fish feeds for the reproduction stage in the three main agro-ecological zones of the country were formulated (SG1, SG2 and G). Male and female Oreochromis niloticus broodstock with mean initial masses of 62.63 ± 1.23 g and 53.88 ± 4.54 g respectively were distributed in nine 9 m³ floating cages at a density (1 ind/m²) of one male for every three females. The different batches of fish were fed at 3-2.5% of biomass with feed containing 35% protein. Feed was distributed twice a day at 9 am and 5 pm for two months. At the end of the rearing period, all the fish were weighed individually and samples were dissected, the gonads removed and weighed. A sample of the ovaries was taken so that the oocytes could be counted under an electron microscope. In hapas, after two months of rearing, the condition indices for the males were 1.20 ± 0.00, 1.42 ± 0.10 and 1.63 ± 0.06. Gonadal weight ranged from 1.26 ± 0.08 to 1.53 ± 0.9 g. In females, absolute fecundity ranged from 572 ± 51 to 644 ± 44 oocytes/female. Oocyte diameter ranged from 3.02 ± 0.07 to 3.28 ± 0.02 mm. Relative fertility values ranged from 7683.2 ± 756.91 to 8257.28 ± 455.6.

1. Introduction

Oreochromis niloticus is the most widely farmed species in fish farms in Côte d'Ivoire. However, one of the obstacles to the expansion of tilapia farming in Côte d'Ivoire is the lack of sufficient quantities of high-quality fingerlings ¹. Furthermore, another significant limiting factor is the high cost of commercial aquaculture feed ². This is the main expense in fish production and can account for over 60% of production costs in semi-intensive systems ³. The high price of feed limits fish farmers’ ability to feed fish properly, particularly broodstock. This situation can negatively affect reproductive performance and the availability of fry. To address this, improving broodstock management through feed control is an essential tool ⁴. In fish farming, broodstock feeding is the most critical element of sustainable aquaculture production, as it provides the nutrients necessary for gonadal development, egg yield and fry quality ^{5, 6}. Furthermore, to ensure the sustainable growth and expansion of tilapia farming, effective measures such as improved nutrition and feeding, with a particular focus on broodstock, are essential for producing large, high-quality eggs ^{7, 8}. Indeed, the importance of protein content on gonadal maturation and reproductive success has been highlighted by several authors ^{9, 10}. The nutritional requirements for optimal reproductive performance in tilapia broodstock have already been reported by several authors ^{11, 12, 13}. One potential solution is for fish farmers to produce their own feed using cheaper ingredients available in the agro-ecological zones (Guinean, Sudano-Guinean 1 and 2) of Côte d’Ivoire. The aim of this study is to formulate low-cost, isoproteic diets containing 35% protein, adapted to the local agro-ecological conditions of Côte d’Ivoire, using available raw materials in order to improve the reproductive success of O. niloticus broodstock.

2. Materials and Methods

The experimental site was the Green fish farm, located in southern Côte d'Ivoire, in the Agboville department, specifically within the sub-prefecture of Azaguié. The fishpond's water supply was provided by a dam fed by a stream. The pond selected for the experiment was previously drained, cleaned, and then refilled. Its water supply was provided by gravity from the dam. Furthermore, protective devices were installed at the entry and exit points to prevent the intrusion of unwanted species and to avoid any escape of the farmed fish.

Three (3) hapas were installed in triplicate in the same pond, in accordance with the method described in ¹⁴, as follows: each 9 m³ net (length × width × height; 4 m × 3 m × 0.75 m) was secured to four (4) stakes. The depth of the hapas in the water was 0.6 m. The ropes were attached 20 cm above the bottom of the pond and the upper ropes 70 cm above the water level. These hapas were arranged in three rows across the full width of the pond. They were made from mosquito nets, sewn together with rope to form rectangular enclosures. These stakes were placed in the pond filled with water. As for the pond, it was supplied with water by gravity from the dam. Protective barriers were installed at the entrance and exit of the pond to prevent, on the one hand, the intrusion of unwanted fish and, on the other hand, the loss of experimental fish. Landing nets, buckets, barrels, sieves, rulers and various scales were used to measure and weigh the fish during the control fishing. A CAMRY scale with a capacity of 5 kg was used to weigh the daily rations, and an EKS scale with a capacity of 5 kg and an accuracy of 1 g was used to determine the weight of the experimental fish.

The biological material consisted of formulated feed from three agro-ecological zones. These were the Sudano-Guinean zone 1 (south-central), the Sudano-Guinean zone 2 (central and eastern) and the Guinean zone (southern and western) of Côte d'Ivoire, as well as Nile tilapia (Oreochromis niloticus) at the reproductive stage. Broodstock of the Bouaké strain of O. niloticus were purchased from a fish farm to establish the various experimental groups.

Using the Excel linear programming method, three feed formulas were developed at 35% protein, with ash (4-8%), carbohydrate (25-40%), fibre (˂ 10%) and energy (18-25 kJ/g) contents within the ranges recommended for rearing Oreochromis niloticus broodstock ^{11, 12, 13}. Fishmeal with 55% protein, cotton and soya cake and wheat bran available locally in the fish farming agro-ecological zones were used for the formulation. The different feed formulations are presented in Table 1.

Proximate composition of fish and feeds was determined according to ¹⁵ and ¹⁶. Dry matter (DM) was determined after drying the samples in an oven (80°C) until a constant mass. Crude protein was measured using Kjeldhal method (N% × 6.25), Ash was measured by incineration at 550°C in a muffle furnace for 24 hours, and crude fat was determined using Soxhlet extraction with hexane as solvent. The gross energy content of feeds and fish samples were calculated based on energy equivalents of crude protein, crude fat, and total carbohydrates; 23.7, 39.5, and 17.2 kJ/g, respectively ¹¹. The mineral contents of the samples were determined by atomic absorption spectrophotometer. All analyses were made in triplicate for each sample.

Male and female broodstock of Oreochromis niloticus had average initial weights of 62.63 ± 1.23 g and 53.88 ± 4.54 g, respectively. The males had a higher average weight than the females in order to minimise aggression during reproductive interactions. These broodstock were distributed across nine 9 m³ floating cages at a stocking density (1 ind./m²) of one male to three females. This experimental phase took place from 12 September to 12 November. The fish were fed isoproteic experimental diets (SG1, SG2 and G) containing 35% protein respectively in the different cages. The daily ration was divided into two meals, at 9 am and 5 pm, from September to November. The different groups of fish were fed at a rate of 3% of biomass per day at the start of the experiments, then adjusted to 2.5% based on samples taken during control catches. At the end of the rearing period, all fish were weighed individually and samples were dissected; the gonads were removed and weighed (Figure 1). Ovarian samples were taken and the oocytes were counted using an electron microscope.

The condition factor K, gonado-somatic indices, absolute and relative fecundities were calculated for males and/or females according to the following formulae:

(1)

(2)

(3)

(4)

Statistical analysis of data collected was performed on STATISTICA 7.1 software. Data were expressed as means ± standard deviation. One-way ANOVA analysis of variance was used to compare the values. Then, Duncan multiple range tests were used to compare differences among means. Differences were considered significant at p < 0.05.

3. Results

Table 2 presents the biochemical and mineral compositions of the three experimental diets (SG1, SG2, and G). The analysis reveals several significant differences likely to influence the zootechnical performance and physiology of the fish. In general, the diets have similar protein content (~35%), but differ in their energy and mineral composition. Diet SG1 is richer in carbohydrates and ash, but it has the best protein/energy ratio. Conversely, diets SG2 and G are richer in lipids. They are therefore more energy-dense, with higher gross energy but a lower protein/energy ratio. In terms of minerals, diet SG1 is distinguished by high levels of phosphorus, magnesium, and potassium, while diet G is richer in calcium and sodium.

The nine cages were placed in the same pond to ensure that the experiment was conducted under identical conditions. To this end, the temperature was around 30.58 ± 0.89°C. The average pH was estimated at 9.22 ± 1.08. The dissolved oxygen level recorded in the rearing pond was 7.96 ± 0.61 mg/L. As for the average conductivity, its value was 15.67 ± 3.58 µS/cm (Table 3).

The Table 4 shows the effect of three experimental diets (SG1, SG2 and G) on various reproductive parameters in males and females. The results are expressed as mean ± standard deviation, and different letters indicate the presence of significant differences (p < 0.05) between treatments.In males, the eviscerated carcass weight varied significantly between treatments. Fish fed diet SG1 (131.43 ± 3.17 g) and SG2 (127.64 ± 2.92 g) had significantly higher values than those fed diet G (119.14 ± 2.8 g). The condition factor (K) also showed significant variation between treatments. The values increased from SG1 (1.20 ± 0) to G (1.63 ± 0.06). Regarding gonadal mass, males fed SG1 (1.53 ± 0.9 g) had a significantly higher gonadal mass than those receiving SG2 (1.31 ± 0.01 g) and G (1.26 ± 0.08 g). By contrast, the gonadosomatic index (GSI) remained unchanged (0.01 ± 0) across the three treatments. In females, the eviscerated carcass weight varied significantly between treatment groups. Fish fed SG1 (82.41 ± 0.83 g) had the highest values, followed by SG2 (76.41 ± 1.49 g) and G (69.35 ± 3.32 g). The condition factor (K) showed a gradual and significant increase across the treatments, with the highest values observed in fish fed G (2.81 ± 0.05), followed by SG2 (2.38 ± 0.03) and SG1 (2.17 ± 0.03). Ovarian mass was significantly higher in females fed SG1 (3.49 ± 0.25 g) compared with the SG2 (2.69 ± 0.23 g) and G (2.48 ± 0.25 g) treatments. In contrast, the gonadosomatic index showed no significant difference between the treatments (0.03–0.04). Similarly, absolute fertility, relative fertility and oocyte diameter showed no significant differences between the diets (p > 0.05).

4. Discussion

The biochemical composition of experimental feeds is a key factor in zootechnical performance and the metabolic utilisation of nutrients in fish. In the present study, the crude protein, fibre, energy, ash and carbohydrate contents of the three formulated feeds fall within the range generally recommended for the feeding of broodstock ^{11, 12, 13}. The fat contents in diets SG2 (10.87%) and G (11.66 %) exceed the standards recommended by these same authors. However, ¹⁷ suggests fat levels in tilapia feed of between 5 % and 12 %. With regard to mineral composition, the variations observed between the diets likely reflect differences in the raw materials used.

The physicochemical parameters of the water play a decisive role in the growth, survival and productivity of fish reared in ponds, particularly in tilapia such as Oreochromis niloticus. The average water temperature was 30.58 ± 0.89°C, which generally corresponds to the favourable range for tilapia farming. Indeed, the optimum temperature for the growth of this species is generally between 26 and 30°C, with a wider tolerance extending up to approximately 32°C ^{17, 18, 19}. The average pH observed in the pond was 9.22 ± 1.08, indicating an alkaline environment. For tilapia, the optimal pH is generally between 6.5 and 8.5, although this species is relatively tolerant to pH variations ¹⁷. An increase in pH in fish ponds may be linked to high photosynthetic activity of phytoplankton, which consumes carbon dioxide dissolved in the water. Values slightly above the optimal range may nevertheless be observed in highly productive systems without necessarily causing severe stress in tilapia, a species renowned for its high environmental tolerance. Although the pH value recorded is slightly high (pH above 9) in this pond, this is not a cause for concern as Oreochromis niloticus can survive successfully and tolerate a pH range of 4 to 11 ²⁰. The recorded dissolved oxygen concentration (7.96 ± 0.61 mg/L) is favourable for fish production. Indeed, studies indicate that tilapia generally require a minimum dissolved oxygen concentration of around 5 mg/L ²¹. High concentrations of dissolved oxygen in the water are often associated with high photosynthetic activity and good water circulation in the ponds. The electrical conductivity measured in this study (15.67 ± 3.58 µS/cm) indicates a low concentration of dissolved mineral salts in the water. This value is below the lethal range, which is between 3.8 and 10 µS/cm ²².

With regard to males, the results show that eviscerated weight is significantly higher in fish fed diets SG1 and SG2 compared to diet G. This suggests that diets SG1 and SG2 provide better energy and protein utilisation, promoting somatic growth. However, the condition factor (K) is higher with the G diet, indicating a greater accumulation of body reserves despite a lower weight. Gonadal mass was significantly higher in the SG1 group (1.53 g), indicating a more efficient allocation of nutrients to reproduction. However, the gonadosomatic index (GSI) remained low and was identical across treatments (0.01). Indeed, in O. niloticus, a gonadosomatic index value between 0.05 and 0.15 is said to characterise very different stages of ovarian maturation, such as a gonad at the start of maturation or a post-spawning gonad (after the release of male or female gametes) ²³. In females, eviscerated weight follows the same trend as in males (SG1 > SG2 > G), indicating better growth with the SG1 diet. In contrast, the condition factor (K) increases significantly from SG1 to G, reflecting increased lipid accumulation, which is often associated with preparation for reproduction or energy storage. Ovarian mass is significantly higher under SG1, confirming that this diet promotes gonadal maturation. However, as in males, the IGS remains low (0.03-0.04) and does not differ significantly from the other treatments. This suggests that the females are also in the post-laying phase, with their ovaries in the process of recovery. Furthermore, the results obtained were higher than the values for the gonadosomatic index of mature females (2.6 %), relative fertility (3,720 oocytes/kg of female body weight) and oocyte diameters (2.17 and 2.55 mm) reported by ²⁴ and ²⁵ for the same species in a farming environment. Similarly, the absolute fertility values (572 to 644 oocytes/female) were higher than those reported by ²⁶, which ranged from 348 to 467 oocytes/female for females weighing 64.5 to 80.95 g fed a diet containing 38 % protein. These results showed that the values for ovarian mass and gonadosomatic indices in females were higher than those for testicular mass and gonadosomatic indices in males. These findings are consistent with the work ²⁷. According to this author, in the African intertropical zone, the testes of O. niloticus always weigh less than the ovaries.

5. Conclusion

This study highlights the influence of diets on the zootechnical and reproductive performance of breeding stock. The results show that the experimental diets, particularly diet SG1, significantly improve eviscerated weight and gonadal mass in both males and females, reflecting a more efficient allocation of nutrients towards the growth and development of the gonads. However, the absence of significant variation in the gonadosomatic index (GSI), absolute and relative fertility, and oocyte diameter between treatments suggests that the various diets tested generally meet the essential nutritional requirements for reproduction. These feeds should therefore be made available to Ivorian fish farmers to contribute to the mass production of high-quality fry on Ivorian farms.

ACKNOWLEDGEMENTS

This work is a part of the framework of the PAFARCI (project to support the recovery of agricultural sectors in Côte d'Ivoire) project funded by the FIRCA (national inter-professional fund for research and agricultural council). Authors express their gratitude to the FIRCA and also thank feed sellers feeds and fish farmers for their help in this work. Authors also accord thanks to the Ivorian Association of Agronomic Sciences (AISA) for the good management of the funds for this project.

References