Biological Materials and Products

The Clarias gariepinusfish (males and females) were sourced from Lake Kivu. A 0.9% w/v NaCl physiological saline solution (500 ml) from Shijiazhuang No. 4 Pharmaceutical was used. Cloves, used as an anesthetic prior to surgery, were purchased at the local market. The feed used for the fish (broodstock and larvae) was the Koudijs brand (44% crude protein) with a particle size of 4.0 mm (floating extrudates) for optimal ingestion (Koudijs Animal Nutrition, (2022)) [13] and Skretting Perla Larva Pro (62% protein), with the particle size (0.2 mm then 0.3 mm) adapted to the mouth opening of the Skretting larvae, (2023) [14]. Following the partial orchiectomy and suturing of the abdominal wall, a methylene blue solution was applied topically to the incision site. This local antiseptic treatment was administered to prevent opportunistic fungal infections, primarily caused by Saprolegniasp., and to promote skin healing in male broodstock in an aquatic environment [15][16].

The fertilized eggs were incubated in circular polyvinyl chloride (PVC) tanks with a capacity of 50 liters. These incubators were equipped with a continuous water-renewal system (a closed system with physical and biological filters) to maintain an optimal level of dissolved oxygen. The bottoms of the tanks were fitted with drainage pipes and overflow control valves. A UV lamp was installed on these incubators to ensure that the water was free of microorganisms. The plastic tanks held 200 liters of water for rearing broodstock. The rearing tanks for Clarias gariepinuslarvae were made of polyvinyl chloride (PVC) and had a capacity of 20 liters. Equipment included a mini oxygenator for aeration pumps, a siphon hose for removing impurities and dead larvae, nets for collecting broodstock and larvae, microtest tubes for measuring ammonia and nitrite levels, and an OHAUS electronic scale, model: Pioneer PX4201 with an accuracy of 0.1 g for weighing broodstock, feed, and Clarias gariepinuslarvae; a 5-ml syringe for collecting sperm; latex gloves to ensure hygiene when handling fish; a workbench to ensure comfort during handling; and a surgical kit (scissors, forceps, disposable pipettes, and a usable scalpel) for performing orchiectomy on male broodstock. The incubation rack was used to arrange the eggs in the incubator. A clean towel was used to restrain the female Clarias gariepinusprior to stripping to prevent the mucous membrane from slipping and to avoid handling accidents. Small basins were available for collecting eggs prior to fertilization.

The experimental trials took place over a three-month period, from January to April 2026, which coincided with part of the rainy season in South Kivu—a time favorable for the natural reproduction of Clariasin tropical environments and for the availability of the facilities that would serve as a hatchery.

Experimental Design and Broodstock Distribution

A total of 30 mature Clarias gariepinus broodstock were mobilized for this study. This experimental stock included 12 reproductive females and 18 mature males. Distribution and use of these 18 males were as follows: 12 males were directly integrated into the factorial comparison plan of sperm collection methods, with 6 males subjected to partial orchiectomy (surgery group) and 6 males subjected to conventional sacrifice for direct milt collection. To optimize biological material use, pituitary glands (hypophyses) of the 6 sacrificed males were immediately collected and preserved for 7 days to evaluate their hormonal induction capacity after storage in 70% ethanol for 6 females. The remaining 6 males (surplus broodstock) were sacrificed on day 8 of the trial for fresh pituitary collection, allowing preparation of crude hormonal extract necessary and sufficient for ovulatory induction of the other 6 reproductive females.

Surgical Sperm Collection

To ensure a continuous supply of sperm without having to slaughter the males, the researchers tested surgical techniques that allow for sperm extraction while ensuring the fish’s survival. According to Diyaware et al. (2010) [17], non-lethal sperm collection in C. gariepinuscan be performed via a minor abdominal incision under anesthesia, followed by partial extraction of the testicular lobes. This approach preserves individuals for future use and reduces broodstock stress and mortality [18].

The surgical procedure was performed in three successive steps on six males. The first step involved anesthetizing the fish by immersing them in a clove solution at a concentration of 0.5 g of powder per liter of water. The cloves, purchased at the local market, were finely ground before being completely dissolved in water. The active ingredient in this solution is eugenol, a natural anesthetic widely used in fish farming. The immersion lasted five minutes [18].

The second step consisted of the surgical procedure and removal of the testes. After anesthesia, the male was removed from the solution and placed in a supine position on a damp cloth laid out on the operating table. The abdomen was disinfected with 70% technical ethanol, then incised with a scalpel using a fine longitudinal incision along the peritoneal cavity [18]. The incision, ranging from 3 to 8 cm in length depending on the fish’s size, extended from the posterior to the anterior part of the urogenital papilla and was made carefully to avoid any damage to the internal organs. The organs within the peritoneal cavity were then gently retracted using forceps to gain access to the testes. One of the testes was carefully removed using forceps and scissors. After removal, the peritoneal cavity was closed and sutured using absorbable suture material (2/0 chromic gut) with a simple discontinuous suture consisting of 5 to 9 stitches. Postoperative disinfection was performed with methylene blue [18].

The third step is post-surgical recovery. After suturing, the six fish were placed in recovery tanks for the convalescence period. The first feeding took place 48 hours after suturing to prevent complications related to food ingestion and to minimize stress on the sutures. Wound healing was monitored once a week, with each fish briefly removed from its tank using a net for examination [18].

Postoperative monitoring involved daily checks of the physicochemical parameters of the rearing water during the ten days required for the broodstock to heal completely. This monitoring focused on temperature, dissolved oxygen, pH, and electrical conductivity. At the same time, ammonia and nitrite concentrations were assessed by visual colorimetric analysis, using a color chart for comparison. These analyses were conducted in accordance with standard protocols for monitoring water quality in aquaculture [19]. At the same time, the post-surgery survival rate of the fish was assessed throughout the observation period [18]. With regard to the physicochemical parameters, changes in the values were monitored over a ten-day period for the entire experimental group.

Sperm Collection via Euthanasia

Sperm collection is a crucial step in the artificial reproduction process for Clarias gariepinus, as it determines the quality of fertilization and the long-term availability of male broodstock. Traditionally, sperm collection in this species involves the sacrifice of the breeding male, followed by testicular removal and maceration to obtain usable sperm [20]. Although this method is commonly used in Africa, it has a major drawback: the systematic loss of the male broodstock, which limits the sustainability of broodstock populations in hatcheries [21][22].

Sperm collection and recovery were performed with particular care to avoid any premature activation of the spermatozoa. After removal using forceps, the testes were placed on a pre-dried dish. Extraction then involved making an incision in the testes with fine scissors, followed by gentle manual pressure to release the seminal fluid into a beaker. Finally, the seminal fluid was aspirated using a 5-ml syringe, allowing for precise measurement of the total volume collected from each male, in accordance with the procedure described by Sayah (2016) [23].

The volume of semen (ml) collected from each Clarias gariepinusmale, both surgically and by euthanasia, was recorded individually.

Hormonal Induction and Female Reproduction Protocol

Hormonal induction in African catfish Clarias gariepinus relies on the use of pituitary extracts, a technique widely employed in aquaculture to stimulate ovulation in mature females. The induction product was prepared from homologous pituitary glands collected from mature males. The crude hormonal extract was prepared at a concentration of 1 g/mL in 0.9% NaCl physiological saline, produced by Shijiazhuang No. 4 Pharmaceutical. This pituitary solution was then administered by a single intramuscular injection at a dose of 1 g/mL of female body weight, according to the protocol described by Okomoda et al. (2015) [24]. Following injection, females were kept in individual holding tanks under controlled hatchery thermal conditions, respecting a strict ovulatory latency time of 9 h 30. This experimental protocol was inspired by the work of Otoh et al. (2025) [25], which demonstrated that pituitary extracts yield fertilization and hatching rates comparable to those obtained with commercial hormones.

Egg Collection by Stripping and Fertilization Steps

Artificial reproduction was carried out using a non-invasive protocol to preserve the integrity and survival of aquaculture broodstock females, following the approaches of Esa et al. (2023) [26]. At the end of the required thermal latency period, mature oocytes were manually extracted by gentle abdominal pressing (stripping). The expelled ovarian fluid was directly collected in small clean, dry collection basins pre-weighed on a precision electronic balance (OHAUS Pioneer PX4201, accuracy 0.1 g), allowing recording of the total spawning mass (Wf).

Immediately after stripping, artificial fertilization was performed dry according to the conventional method described for Siluriformes by Viveen et al. (1985) [20]. Milt (sperm), precisely measured using a sterile 5 ml syringe, was added directly to the oocytes in the dry container. The oocyte-sperm mixture was gently homogenized dry for a few seconds. Sperm activation and fertilization were triggered by adding a small volume of clean hatchery water, followed by gentle continuous stirring for about 1-2 minutes. Fertilized eggs were then immediately rinsed with clean water to remove excess seminal fluid, then spread evenly on incubation frames. These frames were placed in circular PVC incubators with a capacity of 50 liters, equipped with a closed continuous water renewal system to maintain optimal oxygenation conditions, in accordance with hatchery standards for Siluriformes by Biegniewska et al. (2010) [27].

Environmental Conditions and Larval Rearing Parameters

Intensive larval rearing was conducted for 30 days in experimental units consisting of small PVC tanks with a useful capacity of 20 liters. Stocking density per tank was uniformly set at 50 larvae per unit (i.e., an initial density of 2.5 larvae∙L⁻¹). Water renewal and removal of organic waste (feces and food residues) were ensured daily using appropriate siphon tubes, in accordance with water quality management recommendations in inland aquaculture [19][28].

To rigorously interpret survival and growth dynamics, physicochemical variables of the environment were strictly monitored twice daily (07:00 and 18:00). During the entire fry rearing phase, from Day 3 to Day 30 post-yolk sac resorption, average water physicochemical conditions were maintained stable and homogeneous among experimental tanks (ANOVA, p > 0.90) within the biological comfort ranges described for Clarias gariepinus by Haylor (1991): temperature increased progressively and controlled from 26.58˚C ± 0.14˚C to 28.2˚C ± 0.09˚C at the end of rearing; average dissolved oxygen saturation was maintained between 6.79 ± 0.1 mg/L and 6.47 ± 0.09 mg∙L⁻¹ thanks to a continuous mini-pump aeration system; and pH stabilized between 7.44 ± 0.04 and 6.92 ± 0.03. Toxic nitrogenous compounds were rigorously controlled: total ammonia (NH3) remained null or below detection threshold (<0.2 mg/L), while nitrite concentration (NO2) evolved from 0 ± 0 mg/L at the start of incubation to very low stable traces of 0.03 ± 0.01 mg/L at the end of larval rearing.

Feeding

In this study, a single diet was used to feed all larvae from the different treatment groups. This diet is Perla, produced by Skretting; it is a high-performance starter diet specifically formulated for fish larvae. It comes in the form of extruded micro-pellets with an extremely fine particle size (0.2 mm from 0 to 10 days, then 0.3 mm from 11 to 30 days), which is suited to the small size of the larvae’s mouth openings [14]. The diet was tested in 20-liter tanks containing 50 larvae per rearing tank. The larvae were fed 6 times a day for the first 15 days, then 3 times a day until weaning at 30 days post-hatching [29].

Comparison ofFertilization Rates,Egg Hatching Rates,Survival, andLarval Growthaccording toSperm Collection Methods

The reproductive performance of Clariasgariepinus was evaluated by the fertilization rate—based on the count of transparent eggs—and the hatching rate—calculated as the ratio of viable larvae to fertilized eggs—according to the methodology of Tilahun et al. (2016) [30]. These parameters allow for the quantification of the effectiveness of artificial reproduction, with the fertilization rate calculated from an initial sample and the hatching rate determined at the end of incubation.

the fertilization rate (%), calculated according to Tilahun et al. (2016) [30].Weight in gLength in cm TF = Number   of   transluscent   fertilized   eggs Total   number   of   eggs   ( incubation ) × 100

the hatching rate (%), calculated according to Tilahun et al. (2016) [30].

The measured parameters included: Survival rate (%): proportion of surviving larvae on Day 10, Day 20, and Day 30; survival rate (%) on Day 10, Day 20, and Day 30, according to the formula by Tilahun et al. (2016) [30];

Total number of individuals initially counted at stage d4

Growth in length (cm) and weight (g): measured by sampling 10 larvae per tank every 10 days up to 30 days of rearing, using a digital caliper and a digital scale; Condition factor (K): calculated according to Fulton K = ( Weight   in   g Length 3   in   cm ) × 100 .

Experimental Design and Statistical Analyses

The experimental unit was strictly defined for each level of protocol analysis. For evaluation of morpho-ponderal characteristics, growth, and broodstock condition, the experimental unit was the individual fish (n = 12 females and n = 6 males per treatment group). For monitoring water physicochemical parameters (temperature, pH, dissolved oxygen, electrical conductivity, nitrites, and ammonia), the experimental unit was the individual rearing tank (n = 12 tanks in total). For zootechnical evaluation of fertilization, hatching, and overall larval survival at 30 days, the experimental unit was also the tank, with (n = 3) replicate tanks per treatment combination (12 tanks in total).

For monitoring the offspring weight and length growth, the 10 larvae randomly sampled from each tank at D10, D20, and D30 constituted subsampling units (technical replicates). To avoid pseudo-replication bias, individual measurements of these 10 larvae were averaged per tank. It is the mean value per experimental unit (tank, n = 3 per combination) that was subjected to two-factor ANOVA, Student’s t-test, and Principal Component Analysis (PCA), thus ensuring strict independence of observations. Data normality was verified beforehand by the Shapiro-Wilk test. The sketch of the experimental design is shown in Figure 1.

Figure 1.Sketch of the experimental design. SU = Surgery; SA = Sacrifice.

The effectiveness of surgical semen collection was evaluated by direct comparison with the conventional method involving sacrifice. The results inTable 1 detail the maturity parameters (IGS) and quantitative semen yields obtained for each treatment.

Table 1.Comparison of surgical and sacrificial methods based on sperm volume and IGS.

Table 1 shows encouraging results for the management of male broodstock of Clarias gariepinus.

First, the sexual maturity of the two groups was equivalent, with a nearly identical Gonado-Somatic Index (GSI) of approximately 0.93 ± 0.26 for the operated individuals and 0.96 ± 0.29 for the sacrificed individuals. This similarity in GSI validates the rigor of the sampling and allows for a fair comparison of collection performance.

Quantitatively, the average volume of semen collected per sacrifice is similar to that obtained surgically (p = 0.29). This demonstrates that the surgical method allows for the extraction of a semen volume that is similar to that of the conventional lethal method.

3.1.1. Relationship between Sperm Volume Produced, Male Body Size, and Testicular Weight

To determine whether male stature is an effective selection criterion, the relationship between body weight and the volume of semen collected was analyzed in Figure 2 and Table 2.

Figure 2. Relationship between sperm volume produced and testicular weight.

Table 2.Correlation between body size, testicle weight, and sperm production.

Figure 2 and Table 2 show a strong positive linear relationship between testicle weight and collected semen volume (r= 0.985; p< 0.0001), proving that semen production is strictly determined by gonadal mass. There is a near-perfect alignment of the data points where the trend lines for “Surgery” and “Sacrifice” completely overlap. This overlap provides clear biological evidence that partial removal from a live fish yields exactly the same volume of semen as traditional sacrifice for the same gonadal mass. The non-lethal method guarantees maximum collection yield while sparing the lives of the broodstock.

Conversely, the fish’s total body weight proves to be a misleading indicator, showing no significant correlation with testicular weight (p= 0.383) or sperm volume (p= 0.186). This finding shows that a heavy broodstock is not necessarily more fertile and scientifically validates the exclusive use of the Gonadosomatic Index (GSI), rather than visual gross weighing, to select high-performing males in hatcheries.

3.1.2. Survival of Male Breeding Stock after Orchiectomy

Assessing the survival of broodstock following surgery is the primary indicator of the protocol’s success. Table 3 summarizes the survival rate observed over a ten-day period (D10) following orchiectomy.

Table 3.Summary of post-surgical survival of males.

Table 3 illustrates the daily monitoring of the six male broodstock during the ten days following the surgical procedure and reveals a survival rate of 100%. No individual died as a result of the surgery or postoperative handling. This success, combined with the optimal water quality maintained during the stabilization phase (O₂ > 7 mg/L), confirms the safety of the orchiectomy-based semen collection protocol for Clarias gariepinus. This result confirms the technical feasibility of this non-lethal method, which preserves the entire broodstock for future breeding cycles, thereby optimizing the hatchery’s profitability.

3.1.3. Impact of the Sperm Collection Technique on Male Gamete Viability

To evaluate the impact of the sperm collection technique on the viability of male gametes, a comparison of fertilization, hatching, and larval survival rates was conducted between the two groups. Table 4 and Figure 3 present the reproductive performance data and the results of the associated Student’s t-test, which allows assessment of the biological efficacy of partial orchiectomy compared to traditional sacrifice.

Table 4 and Figure 3 illustrate the reproductive performance and provide biological evidence of the protocol’s success. The results show that semen quality is preserved, regardless of whether the male undergoes surgery or is euthanized using the traditional method. In fact, the fertilization rate obtained through surgery (80% ± 8.05%) is virtually identical to that obtained through sacrifice (79% ± 5.62%), a similarity confirmed by a very high p(0.8086). This trend stabilizes in subsequent stages, particularly for hatching and larval survival, where statistical tests reveal no significant differences (p> 0.05). These rates are consistent with the biological standards observed in Clarias gariepinusand demonstrate that the fry produced is equally vigorous, regardless of the sperm origin.

Table 4.Reproductive performance (surgery and culling).

Figure 3. Comparison of reproductive performance by sperm source (surgery or sacrifice).

The distribution of the reproductive performance for each collection method is illustrated by the box plots in Figure 3. This graphical representation allows for visualization of the data dispersion and the overlap of results between surgical collection and sacrifice. Visual inspection reveals near-complete overlap of the boxes for the three indicators (fertilization, hatching, and survival). This closeness of the medians confirms the absence of any significant difference between the two methods.

Although there is slightly greater variability in the hatching rate in the “Sacrifice” group, the overall performance of the surgical method remains highly stable. This graph provides definitive visual evidence that non-lethal collection guarantees fry production identical to that of the traditional method, thereby validating its adoption in hatcheries.

3.1.4. Egg-Laying and Fertility Performance by Sperm Collection Method

The reproductive performance data obtained during the experiment—specifically, clutch weight and absolute fertility as a function of the type of male parent—are summarized in Table 5.

Table 5.Egg-laying and fertility performance.

Table 5 shows that semen collection method (surgery vs. euthanasia) does not significantly affect clutch weight and absolute fertility (p > 0.05). Regarding egg weight, the mean values recorded were 94.47 ± 23.28 g for the Surgery group and 95.74 ± 23.86 g for the Sacrifice group. Statistical analysis confirms the absence of a significant difference between these two methods (p = 0.927). The coefficients of variation, at 24.64% and 24.92% respectively, indicate similar dispersion and homogeneity of the data between the two groups.

Similarly, absolute fecundity does not vary significantly depending on the treatment applied to the males (p= 0.772). Females in the Surgery group had an average fecundity of 62650.00 ± 14260.11 eggs, while those in the Sacrifice group had an average of 65362.67 ± 17560.94 eggs.

The associated coefficients of variation (22.76% for surgery and 26.87% for sacrifice) reflect broadly comparable intra-group variability with high stability in both cases.

Partial orchiectomy does not affect the females’ physiological response to induction or the final oocyte yield.

Larval survival rates over time shows disparities during the 30-day rearing period, a critical phase of development, as shown in Figures 4(a)-(c).

Figure 4 presents the results of the longitudinal study of larval viability conducted during the first thirty days of rearing. These results illustrate larval survival and allow for a comparison of the robustness of the offspring produced using the two semen collection methods. Survival rates on the 20th and 30th days varied significantly depending on the surgical method (p20d = 0.07; p30d = 0.03), unlike on the 10th day, when no notable difference was observed (p10d = 0.41). It thus appears that the euthanasia method resulted in lower larval survival compared to that achieved through surgery.

Figure 4.Larval survival after 10 days (a), 20 days (c), and 30 days of rearing (c).

Figure 5, which shows larval survival over 30 days, reveals a gradual decline in the population, with no peaks of mass mortality. It can be observed that larvae from the surgical group (blue curve blue curve) consistently maintained a higher survival rate than those from the sacrifice group (red curve), ending with approximately 88.6% survivors versus 82%.

Figure 5. Survival rate dynamics of larvae over 30 days.

This result is crucial because it proves that semen collection via partial orchiectomy in no way impairs the vigor of the offspring. Finally, the stability of the slopes starting on the fourth day confirms that the use of Perla feed (Skretting) successfully secured the critical phase of dietary transition, thereby ensuring the production of vigorous and high-performing fry for the hatchery. Furthermore, the surgical method and the K index on the ^10thday are negatively correlated. This indicates that the sacrifice resulted in larvae with a lower K index on the ^10thday.

The growth dynamics of the larvae, monitored from the 10th to the 30th day of rearing, are illustrated in Figure 4. This figure allows us to visualize the evolution of average weight and to compare the metabolic efficiency of the offspring produced using the two sperm collection methods.

Figure 6. Weight growth kinetics of larvae fed Perla (Skretting) according to the method of male semen collection.

Figure 6 illustrates the trend in average individual larval weight between the 10th and 30th days of rearing. It highlights a quasi-exponential growth, which is particularly marked after the Day-20. At the start of the study on Day 10, weights are very low and identical for both groups, then increase gradually until Day 20, when the two curves converge at around 0.4 g. The final phase between D20 and D30 shows a sharp acceleration in weight gain, confirming the complete assimilation of Skretting’s Perla feed during the transition to exogenous feeding. In comparison, the blue curve representing the larvae from the surgery group stands out and remains above the red curve of the sacrifice group at the end of the experiment. By Day 30, the surgery group clearly exceeded the 1.5 g threshold, while the control group barely reached this value.

3.2.1. Body Condition of Fry at the End of Rearing

The health and vigor of the fry at the end of the 30-day rearing period were assessed by calculating the Fulton condition index (K). Table 6presents a comparison of this index among the offspring fed Perla (Skretting).

Table 6.Body condition of fry at 30 days.

Student’s t-test (homogeneity): p = 0.363.

Table 6 summarizes the results of the K Condition Index analysis on Day-30 and confirms the good health and body condition of the fry. It can be observed that the index remains stable and consistent around an average value of 0.60, with no significant difference between larvae from the surgical group and those from the traditional sacrifice group (p= 0.363) and it indicates that the larvae did not suffer from nutritional deficiencies and that the Perla feed (Skretting) was fully assimilated in all batches.

3.2.2. Growth Performance

The efficiency of feed assimilation and the rate of larval development were evaluated by calculating the Specific Growth Rate (SGR). Table 7 summarizes the daily growth performance for the two experimental groups during the critical period from the 10th to the 30th day.

Table 7.The specific growth rate from the first to the 30th day.

Student’s t-test (Surgery vs. Sacrifice): p= 0.537.

Table 7 summarizes the results of the Specific Growth Rate (SGR) analysis and provides definitive evidence of the success of the rearing protocol. With average values of 22.42%/day for the surgery group and 21.97%/day for the sacrifice group, the larvae exhibited exceptional growth between the 10th and 30th days. Mathematically, this means that the individual biomass of the fry increased by nearly one-quarter of their own weight every day, confirming the exponential growth phase and the perfect assimilation of the Perla feed (Skretting). In fact, there is no significant difference between the two sperm collection methods.