Fishmeal Replacement with Animal Protein Source (Crocodylus niloticus Meat Meal) in Diets of Mozambique Tilapia (Oreochromis mossambicus) of Different Size Groups

Abstract

Fish are generally known to change their nutritional requirements depending on their life stage and formulating feeds for different size groups to meet their dietary needs is essential. This study aimed to assess the potential of Crocodylus niloticus meat meal as an animal protein source replacing fishmeal in Oreochromis mossambicus diets. Ten fry (0.07 g fish⁻¹) were randomly assigned to three formulated diets (0% (D1), 50% (D2), and 100% (D3)), and each diet had three replicates. The fry were fed 10% body weight per day (BWd⁻¹) for 30 days. New diets (0% (D4), 50% (D5), and 100% (D6)) were introduced, and the feeding rate was reduced to 5% BWd⁻¹ for 48 days. After that, the fish were fed 2% BWd⁻¹ for 78 days, the same diets used for fingerlings. All size groups were fed two portions of their daily ration at 10:30 h and 15:30 h. Our results point to the suggestion that Crocodylus niloticus meat meal may replace fishmeal for Oreochromis mossambicus, as there were no significant differences in weight gain (G), specific growth rates (SGR), gross feed conversion ratios (GFCR), or protein efficiency ratios (PER) for fry fed different diets. Furthermore, there were similarities in Gs, SGRs, GFCRs, and PER in fingerlings and sub adult to adult fish fed D4 and D5. The cost analysis of ingredients used in diets with 50% and 100% Crocodylus niloticus meat meal indicated that it was profitable to use this meat meal in diets of O. mossambicus of all groups. The profit index of 0.3 for fry, 0.8 for fingerlings, and 1.9 for subadults to adults for 100% fishmeal diets were lower than 0.4 and 0.5 for fry, 0.9 and 1.1 for fingerlings, and 2.3 and 2.9 for sub adult to adult fish fed diets with 50% and 100% crocodile meat meal, respectively.

1. Introduction

Oreochromis mossambicus is a freshwater species that belongs to the family Cichlidae [1]. This species is currently vulnerable in Southern Africa [1,2]. The vulnerability is due to the introduction of invasive species, such as Oreochromis niloticus, hybridization, habitat competition, and diseases as the main reasons for decreasing O. mossambicus in their natural habitat [3,4,5,6,7,8,9]. Therefore, it is essential to conserve this species through aquaculture. Oreochromis mossambicus is regarded as a good candidate for aquaculture, because of high fecundity, its ability to utilize both plant and animal nutrients for growth efficiently, increased meat quality and good consumer acceptance, potential to develop value-added fish products, and resistance to diseases [10,11,12]. However, the success of any aquaculture species depends on the supply of adequate nutrients, both in quantity and quality [13]. Due to the rapid growth of aquaculture, some resources, such as fishmeal, are limited.

Fishmeal production is mainly sourced from the forage fish species [14]. Forage fish are described as the prey for other animals to eat [14]. They play an essential role in marine ecosystems because they transfer energy from primary producers (e.g., plankton) to higher trophic-level species, including large fish, marine mammals, and sea birds [14]. Fish caught for fishmeal production potentially represent a loss in production of higher trophic level species in the ecosystem because low stock abundance reduces ecosystem services, such as food provisioning, to other elements of the ecosystem [15]. According to [16], numerous studies have shown that animal by-product meals arising from the processing of slaughtered farm livestock offer great potential for use as dietary fishmeal replacements within aquaculture feed. A review of animal protein sources used in aquaculture diets showed that some by-products, such as crocodile meat, have not been assessed as a fishmeal replacement [13], except for the recent study on juvenile dusky kob (Argyrosomus japonicus) by [17]. Although fish are known to change their nutritional requirement depending on their life stage [18], different size groups are not considered when these animal protein sources are evaluated in aquaculture. Furthermore, the nutritional values of the crocodile meat meal are comparable to other by-products, such as meat bone meal, meat meal, feather meal, blood meal, and poultry by-products meal used in aquaculture [19,20], and different grades of fishmeal [21]. The world’s demand for proteins of animal origin is expected to double by 2050 [22].

Furthermore, fishmeal is the most expensive component of aquaculture feeds because of its competing use as a feed ingredient for other livestock species [23]. Seventy-five percent of the world’s fish stocks used for fish meal production are currently considered fully exploited or overexploited, including many small pelagic fish [24]. Increasing demand, unstable supply, and the high price of the fishmeal with an expansion of aquaculture have resulted in the need to search for alternative protein sources.

Crocodiles are cultured mainly for producing skins used to create high-quality fashion accessories [25]. As in fish farming, the increase in production costs in this industry forced the farmers to look at alternative means of increasing profitability [26]. The major components of source of profitability are skin production and meat production and all these are related to tourism. However, the demand for crocodile meat, especially in South Africa, is low and strict regulations are imposed on the industry regarding the use and disposal of crocodile carcasses. According to [26], crocodile meat is re-used on farms as unprocessed feed for other crocodiles and public health regulations are enforced in processing crocodilian meat for human consumption. This involves the design, construction, operation of abattoirs, food safety standards, and procedures explicitly established for the processing of crocodilians, making the whole process less cost effective [27]. Furthermore, managing of the abattoirs come with additional responsibilities related to packaging, labelling, shipping, and record-keeping [28]. Hence, this study aimed to assess the potential of Crocodylus niloticus meat meal as an animal protein source, replacing fishmeal in the diet of Oreochromis mossambicus. To our knowledge, to date there are no studies that have been conducted targeting at replacing fishmeal with Crocodylus niloticus meat meal in diets of Oreochromis mossambicus. If using Crocodylus niloticus meat meal becomes suitable, the aquaculture industry will benefit by reducing supply constraints imposed by high cost and competitive uses for fishmeal. This could also translate into less dependence on marine-derived protein sources currently being over-exploited. The study’s findings could be used to produce more fish as a source of protein for poor communities and contribute to poverty alleviation and food security. Furthermore, the findings and recommendations on crocodile meat meal could be beneficial to crocodile farmers who are finding it costly to dispose of crocodile meat as a by-product for the skin in South Africa.

2. Materials and Methods

2.1. Processing Crocodile Meat into Meal

Crocodile meat was purchased from Shallow Drift Crocodile Abattoir, South Africa. The meat was cut into small thin pieces, dried in an oven for three hours, ground using the laboratory blender (Model HQBTWTS3, Waring Commercial, Torrington Connecticut, United State of America), and then sieved through a 1.0 micron mesh. The resulting crocodile meal samples were taken to the Animal Science Departments, University of KwaZulu Natal for proximate analysis, while some samples were sent to Stellenbosch University for amino acid analysis and the rest were stored in a heavy-duty clear plastic bag at room temperature until used for fish feed preparation. Nitrogen (N) content was determined on a Leco TruMac ® Carbon Nitrogen Sulfur elemental analyzer using Dumas’s combustion. Crude protein was calculated as N × 6.25. Crude fat content was determined using the Soxhlet method, as described in the AOAC Official Method 920.39 [29]. Crude fiber was determined as the loss of ignition of dried lipid-free residues with 1.25% H₂SO₄ and 1.25% NAOH solutions, using the filter bag technique with the ANKOM Fibre Analyzer 200. Moisture content was determined using an air-circulated oven at 95 °C for 72 h. Ash content was determined by burning pre-weighed samples in a muffle furnace at 550 °C overnight, as described in the AOAC Official Method 942.05. Minerals were determined using the fast sequential atomic absorption spectrometer (AA280FS) [30], after ashing the samples. The results are shown in

Table 1. Composition of raw mixture Crocodylus niloticus meat meal.

Nutrient (%)	Values for Raw Mixture Meal of Crocodylus niloticus
Crude protein	83.04
Moisture	9.78
Crude fat	4.48
Ash	2.41
Crude fiber	0.04
Selected Minerals (%)
Potassium	32.90
Sodium	11.17
Calcium	1.93
Magnesium	1.49
Zinc	0.22
Iron	0.21
Aluminum	0.16
Copper	0.04
Amino Acids (g/100 g dry matter)
Arginine	7.55
Histidine	4.88
Isoleucine	4.10
Leucine	8.06
Lysine	7.27
Methionine	4.53
Phenylalanine	8.37
Threonine	8.37
Valine	4.32
Non-Essential Amino acids (g/100 g dry matter)
Alanine	5.88
Asparagine	8.61
Glutamic acid	14.34
Glycine	5.82
Proline	3.06
Serine	4.27
Tyrosine	6.28

.

2.2. Diets

Six experimental diets (

Table 2. Main ingredients and proximate composition of the experimental diets.

	Fry Diets (D1–D3)			Fingerlings, Sub-Adult, and Adult Fish (D4–D6)
Ingredients (g/kg)	D1 (0% CM)	D2 (50%/50% FM/CM)	D3 (100% CM)	D4 (0% CM)	D5 (50%/50%FM/CM)	D6 (100% CM)
Fishmeal 1	25.000	12.500	-	15.000	7.500	-
Maize 2	20.000	20.139	20.279	30.000	30.000	30.000
Crocodile meal 3	-	9.032	18.063	-	8.406	16.812
Soybean meal 46 4	15.000	15.000	15.000	15.000	15.000	15.000
Canola seed meal 5	15.000	15.000	15.000	15.000	15.000	15.000
Maize gluten 60 6	10.000	10.000	10.000	8.324	6.436	4.548
Wheat bran 7	8.601	9.174	9.746	4.936	4.264	3.592
Canola oil 8	3.531	4.309	5.087	4.356	4.774	5.191
Monocalcium phosphate 9	1.719	2.940	4.161	1.031	4.565	8.099
Vitamin premix 10	0.800	0.800	0.800	0.800	0.800	0.800
Limestone 11	-	-	-	3.528	2.244	0.959
L-lysine HCL 12	0.349	1.107	1.864	2.024	1.012
Total	100	100	100	100	100	100
Proximate Composition (%)
Moisture **	8.995	8.895	8.994	8.627	8.517	8.746
Crude protein *	38	38	38	32	32	32
Crude fat **	8.842	8.090	6.408	8.344	8.935	5.446
Ash **	7.536	7.632	7.413	9.185	9.674	10.949
DE (MJ/kg) *	13.426	11.707	9.988	12.840	11.320	9.800
Essential Amino Acids (g/100 g dry matter)
Arginine ***	2.82	2.87	3.05	2.36	3.31	2.83
Histidine ***	2.03	1.94	1.90	1.63	2.27	1.61
Isoleucine ***	1.53	1.56	1.63	1.30	1.45	1.45
Leucine ***	3.36	3.54	3.79	3.00	3.13	2.99
Lysine ***	2.04	3.24	4.13	3.02	1.52	2.11
Methionine ***	1.26	1.22	1.32	0.97	1.29	1.12
Phenylalanine ***	3.82	3.87	4.04	3.11	3.88	2.89
Threonine ***	2.35	2.39	2.50	2.00	2.64	2.24
Valine ***	1.93	1.93	1.94	1.67	1.82	1.73
Non-Essential Amino Acids (g/100 g dry matter)
Alanine ***	2.33	2.54	2.65	2.12	2.02	2.05
Asparagine ***	3.24	3.85	4.14	3.08	3.34	3.21
Glutamic acid ***	6.34	7.62	8.45	6.23	6.86	6.32
Glycine ***	2.65	2.59	2.62	2.04	2.66	2.43
Proline ***	2.16	2.28	2.47	1.95	1.99	1.90
Serine ***	2.12	2.16	2.24	1.88	2.39	1.95
Tyrosine ***	2.85	2.92	3.27	2.32	3.15	2.33

FM—fishmeal, CM—crocodile meat meal, ¹ Pioneer Fishing, ^{2 + 8} Spar, ³ Shallow Drift Abattoir, ⁴ Irwing Soya,⁵ Southern Oil (Pty)ltd, ⁶ Tongaat-Hulett, ⁷ Milmac, ⁹ Bragan Chemicals, ¹⁰ SA Premix, ¹¹ Idwala Industrial Holdings (Pty) Ltd., ¹² Protea Chemicals, all from South Africa. * Taken from feed formulations. ** Analyzed at Animal Science department, University of KwaZulu Natal, South Africa. *** Analyzed at Stellenbosch University, South Africa.

) were formulated for different size groups of O. mossambicus. D1 and D4 are diets with 0% of crocodile meat meal, D2 and D5 are diets with 50% fishmeal and 50% crocodile meat meal, D3 and D6 are diets with 100% crocodile meat meal. All diets are for fry and fingerlings, respectively. D1, D2, and D3 for fry had 38% crude protein, D4, D5, and D6 for fingerlings had 32% crude protein. All diets were prepared by pre-weighing all dry ingredients separately. The combined ingredients in a bowl were mixed for fifteen minutes. Eight ml of water was added per kg of dry ingredients to make a dough. The dough was pelleted using a hand-operated meat mincer, and the pellets were dried in the sun [31]. All diets were used as treatments. Samples of all diets were analyzed for moisture, crude fat, and ash at the Animal Science Department at the University of KwaZulu Natal using procedures explained for crocodile meat meal, while some samples were sent to the University of Stellenbosch for amino acids analysis. The results of proximate analysis for experimental diets are presented in Table 2.

2.3. Experimental Fish

Newly hatched fry were purchased from the University of Zululand, Department of Zoology, South Africa and transported to the University of Kwa-Zulu Natal. They were allowed to acclimatize in the experimental tanks for 14 days before the feeding trial. The fry were fed a commercial diet purchased at the local supplier Avi-Products (Pty) Ltd., Pietermaritzburg, South Africa, during the acclimation period. It was not possible to separate sexes into males or females to avoid fish breeding, as small-size animals were used. Furthermore, the formulated diets tested are to be used in commercial farming and using both sexes is essential.

2.4. Feeding Experiment

The three parts study was conducted at Animal House, University of KwaZulu-Natal, Pietermaritzburg campus, South Africa. Ninety fry with an initial weight of 0.07 ± 1 g fish⁻¹ were randomly assigned to three treatments as follows: D1 (0% crocodile meat meal with 100% fish meal, D2 (50% crocodile meal and 50% fishmeal, and D3 (100% crocodile meat meal and 0% fishmeal, in nine glass tanks with a volume of 40 L each. The study was conducted in triplicates, and aeration was supplied to fish tanks by air stones connected to air pumps to maintain the oxygen supply to the fish, and water was circulated using 400 L/h submersible water pumps that were in containers, with gravel and sand as filters. Ten individuals were used in each of the three replicates, totaling thirty O. mossambicus for each treatment. The feeding trial conditions were maintained at a temperature of 28 ± 2 °C, and the light was set at the 12 h dawn and 12 h dark cycle.

In part 1, the fry were fed two portions of their daily feed ration (10%) [32], two times a day (at 10:30 h and 15:30 h) for thirty days. The planned feeding frequency for fry was four times a day, as recommended by [33] but changed to two times per day due to COVID-19 restrictions.

After 30 days of feeding, in part 2, all fish were fed new three diets, which were D4, D5, D6 diets with the same crocodile meal percentage as D1, D2, and D3, respectively. The feeding rate was reduced to 5% of their body weight per day in each replicate and fed 2 times per day (at 10:30 h and 15:30 h) for 48 days.

After 48 days, in part 3, all groups were considered as sub-adults to adults, and the same fingerling diets (D4, D5, and D6) were fed at a feeding rate of 2% body weight per day to fish twice a day (at 10:30 h and 15:30 h) [34] for 84 days. Water quality parameters and mortality were monitored throughout the experimental period and were within tilapia tolerable ranges.

2.5. Growth and Feed Utilization Measurements and Calculations

For all three parts, the fish were weighed individually on a weekly interval from each tank, and the mean wet weight for fish per tank was calculated. Fish weight measurements were carried out to determine the correct amount of feed fed to experimental tanks per week as the fish grow. The weight gain (G), specific growth rate (SGR), growth feed conversion ratio (GFCR), protein efficiency ratio (PER), and survival rate (SR) were calculated using the following formulae, as reported by [33,34,35].

Weight gain (G) (g) = final weight (FW) − initial weight (IW)

Specific growth rate (SGR) (%/day) = Ln (final weight) − Ln (initial weight)/experimental duration (days) × 100

Gross feed conversion ratio (GFCR) = weight of food fed/weight gain

Protein efficiency ratio (PER) = weight gain/protein fed

Survival rate (SR) (%) = number of surviving fish/number of fish stocked × 100

It was not possible to measure the total quantity of food not consumed, therefore, the gross feed conversion ratio (GFCR) was calculated instead of the feed conversion ratio (FCR), which is the ratio between feed intake and weight gain measured during the experimental period. [36]. The results were reported according to different size groups (fry, fingerlings, sub-adult to adult) as per feeding rate adjustment.

2.6. Feed Costs Calculations

Incidence cost and profit index were calculated as an economic indicator using the following formulae as described by [37], assuming that only the ingredients costs are the only variable costs and the operating costs are constant.

$$\mathrm{Incidence}\mathrm{cost}=\frac{\mathrm{Cost}\mathrm{of}\mathrm{feed}}{\mathrm{Quantity}\mathrm{of}\mathrm{fish}\mathrm{produced}\left(\mathrm{kg}\right)}$$

$$\mathrm{Profit}\mathrm{index}=\frac{\mathrm{local}\mathrm{market}\mathrm{value}\mathrm{of}\mathrm{fish}}{\mathrm{Cost}\mathrm{of}\mathrm{feed}}$$

The estimated market prices were ZAR 3.00 for fry, ZAR 6.00 for fingerlings and ZAR 15.00 for sub adult to adult fish per fish, based on the University of Zululand sale prices.

2.7. Statistical Analysis

The SPSS software v27 (IBM Corp., Armonk, NY, USA) [38] was used to analyze the data for G, SGR, GFCR, PER, SR, and fish body composition parameters. We checked the data for normality using the Shapiro–Wilk test. The means of the treatments were tested for significant differences using one-way analysis of variance (ANOVA) at a significance level of α = 0.05. The results were considered significantly different at a probability of p < 0.05. Tukey’s multiple comparison test was used to compare the variance among the means.

3. Results

3.1. Weight Gain

The weight gains of O. mossambicus fry were not significantly different among all the diets fed (

Table 3. Mean initial weight (IW), final weight (FW), weight gain (G), specific growth rate (SGR), gross food conversion ratio (GFCR), protein efficiency ratio (PER) and survival rate (SR) of Oreochromis mossambicus fry fed for 30 days, fingerlings fed for 48 days, and sub-adult to adult fed for 84 days different diets, with crocodile meal as fishmeal replacement. Values are means (±SD) of three replicates for each treatment. The results were significantly different at p < 0.05. Degree of freedom between groups = 2, within groups = 7, F = F statistic, and P = probability.

Fry Perfomance
Variables	D1	D2	D3	F	P
IW (g)	0.0837 ± 0.007 a	0.0743 ± 0.005 a	0.0700 ± 0.011 a	2.078	0.220
FW (g)	0.8090 ± 0.138 a	0.8487 ± 0.516 a	0.6840 ± 0.495 a	1.850	0.250
G (g)	0.7253 ± 0.144 a	0.7743 ± 0.052 a	0.8140 ± 0.061 a	1.5513	0.299
SGR (%/day)	7.5400 ± 0.838 a	8.1187 ± 0.302 a	7.6155 ± 0.783 a	0.648	0.560
GFCR	2.1933 ± 0.814 a	2.2833 ± 0.238 a	2.1650 ± 0.064 a	0.035	0.960
PER	0.0191 ± 0.0038 a	0.0204 ± 0.0013 a	0.0162 ± 0.0016 a	1.878	0.246
SR (%)	90.00 ± <0.001 a	96.70 ± 5.773 b	90.00 ± <0.001 a	144.40	<0.001
Incidence cost	11.36	8.75	8.27	-	-
Profit index	0.3	0.4	0.5	-	-
Fingerlings	D4	D5	D6	F	P
IW (g)	0.8090 ± 0.138 a	0.8487 ± 0.516 a	0.6840 ± 0.495 a	1.850	0.250
FW (g)	3.9740 ± 0.224 a	3.4367 ± 0.616 a	2.0910 ± 0.169 b	12.220	0.011
G (g)	3.1650 ± 0.100 a	2.5880 ± 0.577 a	1.4075 ± 0.119 b	13.336	0.009
SGR (%/day)	3.3340 ± 0.243 a	2.8947 ± 0.281 b	2.3275 ± 0.0168 b	11.008	0.015
GFCR	2.0467 ± 0.085 a	3.1867 ± 0.293 b	5.2450 ± 0.431 c	81.809	<0.001
PER	0.0989 ± 0.0031 a	0.0809 ± 0.0180 a	0.0439 ± 0.0037 b	13.334	0.009
SR (%)	90.00 ± <0.001 a	93.33 ± 5.774 b	90.00 ± <0.001 a	77.45	<0.001
Incidence cost	1.96	1.90	2.50	-	-
Profit index	0.8	0.9	1.1	-	-
Subadult to Adult	D4	D5	D6	F	P
IW (g)	3.9740 ± 0.224 a	3.4367 ± 0.616 a	2.0915 ± 0.169 b	12.220	0.011
FW (g)	10.7100 ± 0.629 a	8.4400 ± 2.268 a,b	4.6250 ± 0.502 b	9.809	0.018
G (g)	6.7360 ± 0.53 a	5.0033 ± 1.757 a,b	2.5335 ± 0.333 b	7.724	0.029
SGR (%/day)	1.1767 ± 0.061 a	1.0467 ± 0.198 a	0.9450 ± 0.035 a	2.032	0.225
GFCR	2.0367 ± 0.200 a	3.2967 ± 0.545 b	3.640 ± 0.665 b	8.578	0.024
PER	0.2102 ± 0.0167 a	0.1564 ± 0.05492 ab	0.0797 ± 0.0104 b	7.849	0.028
SR (%)	77.3333 ± 2.0817 a	89.6333 ± 4.7035 b	70.8500 ± 4.0501 a	19.084	0.0025
Incidence cost	0.73	0.77	1.13	-	-
Profit index	1.9	2.3	2.9	-	-

Mean values ± standard deviation within same row with different superscripts are significantly different (p < 0.05).

). The weight gains for fingerlings, and sub-adult to adult fish were significantly different (p < 0.05, ANOVA) among the experimental diets fed (Table 3). The fingerlings fed D4 and D5 had significantly higher (p < 0.05, Tukey test) weight gain than those fed the D7 diet. For the sub-adult to adult fish, the weight gain of those fed D4 was significantly higher than those fed D6, but similar to those fed D5. There were no significant differences (p < 0.05, Tukey test) in weight gains of sub adult to adult fish fed D5 and D6 (Table 3).

3.2. Specific Growth Rate

There were no significant differences in the specific growth rate of the fry fed different diets. There were significant differences (p < 0.05, ANOVA) in SGR of O. mossambicus fingerlings and sub-adult to adult fish. Significant differences (p < 0.05, Tukey test) were observed in fingerling fed D4 and D6. Those fed D5 had specific growth rates similar to the fish fed both D4 and D6 diets. For the sub-adult to adult size group, the differences were among the O. mossambicus fed D6, which was significantly lower than those fed D4 and D5 (p < 0.05, Tukey test).

3.3. Gross Feed Conversion Ratio

There were no significant differences (p > 0.05; ANOVA) in GFCRs of O. mossambicus fry among all diets fed (Table 3). Gross feed conversion ratios of 2.19 (D1), 2.28 (D2), and 2.17 (D3) were recorded. However, GFCRs for fingerlings and sub-adult to adult size groups were significantly different (p < 0.05, ANOVA) among the diets fed. The GFCR of 2.0 (D4) was significantly better (p < 0.05, Tukey test) than 3.18 (D5) and 5.2 (D6) for fingerlings of O. mossambicus. The GFCR of O. mossambicus sub-adult to adult fish fed the D5 diet was similar to those fed the D6 diet but significantly different (p < 0.05, Tukey test) from those provided the D4 diet.

3.4. Protein Efficiency Ratio

The protein efficiency ratios for O. mossambicus fry were not significantly different among all the diets fed. There were significant differences (p < 0.05, ANOVA) in PER for fingerlings and subadult to adult fish size groups. The PER was significantly lower for fingerlings and sub-adult to adult fish (p < 0.05, Tukey test) fed D6 than those fed D4, and D5.

3.5. Survival Rate

There were significant differences (p < 0.05, ANOVA) in the survival rate of all size groups fed different diets. The survival rate for O. mossambicus ranged between 90% and 96.7% for fry, 90% and 93.3% for fingerlings; 68.57 and 89.63% for sub-adult to adult fish.

3.6. Weekly Mean Weights of Experimental Groups

The results showed that the weekly mean body weights of O. mossambicus fed with different diets were significantly different (p < 0.05; ANOVA) from the second week of feeding to week four. However, there were no significant differences in weight gain among all the diets from week five to week nine. Then, from week ten, differences in weight gains among the diets continued until week twenty-five when the experiment was terminated (

Table 4. Weekly mean weights (±SD) for Oreochromis mossambicus fed different diets for 25 weeks. Degree of freedom between groups = 3, within groups = 7, F = F statistic, and P = probability.

Weeks	D1	D2	D3	F	P
Initial weight	0.0837 ± 0.007	0.0743 ± 0.005	0.0700 ± 0.011	2.078	0.220
Week 1	0.1247 ± 0.172	0.1167 ± 0.016	0.0920 ± 0.008	2.895	0.146
Week 2	0.2480 ± 0.033 a	0.2113 ± 0.041 a,b	0.1355 ± 0.002 b	6.849	0.037
Week 3	0.4817 ± 0.021 a	0.2923 ± 0.102 b	0.2370 ± 0.014 b	9.867	0.018
Week 4	0.8487 ± 0.155 a	0.4387 ± 0.027 b	0.3450 ± 0.027 b	13.928	0.009
Week 5	0.8090 ± 0.137 a	0.8487 ± 0.052 a	0.6840 ± 0.049 a	1.850	0.250
Week 6	1.0063 ± 0.060 a	1.0307 ± 0.200 a	0.8370 ± 0.068 a	1.362	0.337
	D4	D5	D6
Week 7	1.3160 ± 0.042 a	1.2587 ± 0.212 a	1.0670 ± 0.048 a	2.029	0.226
Week 8	1.6290 ± 0.092 a	1.5953 ± 0.216 a	1.3505 ± 0.445 a	2.324	0.193
Week 9	2.0443 ± 0.096 a	1.9787 ± 0.223 a	1.6190 ± 0.072 a	5.623	0.053
Week 10	2.6067 ± 0.139 a	2.3803 ± 0.306 a,b	1.8005 ± 0.094 b	8.485	0.025
Week 11	3.2800 ± 0.235 a	2.8560 ± 0.441 a	1.8875 ± 0.121 b	11.501	0.013
Week 12	3.9740 ± 0.224 a	3.4367 ± 0.616 a	2.0915 ± 0.169 b	12.222	0.012
Week 13	4.8200 ± 0.164 a	4.1400 ± 0.466 a	2.3350 ± 0.125 b	12.583	0.011
Week 14	5.5600 ± 0.403 a	4.3967 ± 0.819 a	2.5250 ± 0.219 b	16.260	0.006
Week 15	6.3567 ± 0.580 a	4.7067 ± 0.929 a,b	2.7500 ± 0.240 b	16.054	0.007
Week 16	6.7033 ± 0.280 a	5.0500 ± 1.058 a,b	3.0900 ± 0.184 b	15.970	0.007
Week 17	7.4100 ± 0.302 a eggs	5.5133 ± 1.210 a,b	3.0650 ± 0.177 b	16.548	0.006
Week 18	7.7867 ± 0.448 a	5.9000 ± 1.272 a	3.3300 ± 0.226 b	16.177	0.007
Week 19	8.7133 ± 0.520 a	6.4867 ± 1.677 a,b eggs	3.2800 ± 0.269 b	13.994	0.009
Week 20	9.1267 ± 0.615 a eggs	6.7733 ± 1.708 a,b	3.7500 ± 0.382 b	12.882	0.012
Week 21	9.5233± 0.527 a	7.4000± 2.053 a,b	3.9300± 0.339 b	10.322	0.017
Week 22	9.5733 ± 0.665 a	7.8067 ± 2.128 a,b	4.2750 ± 0.559 b	8.727	0.023
Week 23	9.6800 ± 1.773 a	8.0333 ± 2.073 a,b	4.4900 ± 0.424 b	8.755	0.023
Week 24	10.3433 ± 0.885 a	8.3133 ± 2.234 a,b	4.5750 ± 0.459 b	6.959	0.036
Week 25	10.7100 ± 0.629 a	8.4400 ± 2.268 a,b	4.6250 ± 0.502 b	9.737	0.018

Mean values ± standard deviation within same row with different superscripts are significantly different.

). Regardless of all the differences, all fish in all groups increased their body weight. Initial mean weights were similar in all diets. Egg production was not part of the parameters measured. However, eggs were observed during the weekly mean weight measurements. In fish fed D4 (at week 17 and week 20), D5 (at week 19), and there were no eggs nor fry observed in D6 during the experimental period.

3.7. Economic Analysis

All size groups fed diets with 50% fishmeal and 50% crocodile meal had consumed more feed than other groups (Figure 1a–c). The costs for known ingredients show that using 100% fish meal (D1 and D4) was more costly than using 50% (D2 and D5) and 100% (D3 and D6) crocodile meal (Table 3). The profit index indicates that it is profitable to use 100% and 50% crocodile meal in diets of O. mossambicus of all size groups.

4. Discussion

Fish are generally known to change their nutritional requirements depending on their life stage [18] and formulating feeds for different size groups to meet their dietary needs is essential. Studies conducted on fishmeal replacement by animal protein sources, as reviewed by [13], did not consider different size groups of fish when testing various ingredients in the diets. Furthermore, the control diets used in the reviewed studies were not commercial feeds and the diet formulations were used without the protein source tested [39,40,41,42].

Producing high quality fish, reducing the cost of feed, and minimizing the use of forage fish in fishmeal production (pressure in the marine ecosystem) are the main reasons for the need to replace fishmeal in animal feeds [43,44]. To our knowledge, studies conducted on the evaluation of crocodile meal as a fishmeal replacement in aquaculture for different size groups of O. mossambicus diets are scanty. From our previous study on the nutritional value of the Nile crocodile meal, we recorded values comparable to other by-products from poultry, such as feather meal, blood meal, and bone meal used in aquaculture diets [20].

The weight gain for fry was not significantly different among all diets. This may be because all the diets had similar crude protein level, moisture, and ash. However, fingerlings fed D6 had significantly lower weight gain than those fed D4 and D5. Sub adult to adult fish fed D4 had weight gain similar to those fed D5; the fish fed D6 had significantly lower weight gain than those fed D4 but similar to those fed D5. Diets with high-fat contents are known to have high concentrations of saturated fats and are characterized by high 18:2 n-6 polyunsaturated fatty acids (PUFA), which reduces the palatability of fish diets [17,45]. Our results are not in agreement with the findings of [46], who reported reduced growth performance, due to the high fat content in silver sea bream Rhabdosargus sarba fed a poultry by-product meal diet above the 25% substitution level. The D6 group had lower crude fat content and lower weight gain compared to those fed D4 and D5.

According to [47], fish growth rate decreases as they grow older. Our study results agree with [47]’s statement, as the SGRs of O. mossambicus fry ranged from 7.54% to 8.11%, and it was significantly higher than 2.89% to 3.33% for fingerlings and 0.94% to 1.17% for the sub-adults and adults. Similar results were reported by [48] for O. niloticus fry and young tilapia.

The feed conversion ratio determined by other authors is similar to GFCR in the current study, and it is defined as the amount of feed consumed to produce one unit of animal biomass gain [49]. Higher efficiency is indicated by lower GFCR values [50]. Gross feed conversion ratios of 2.16 to 2.28 in fry, fingerlings, sub-adult, and adult were observed, except for D5 (3.18) and D6 (5.24) for fingerlings, and sub adult to adult fish. Regardless of fish species, initial weight, duration of the experiment, and feeding frequency used, our results are similar to those reported for the recommended levels of other animal protein sources, such as mopane worms (Imbrasia belina), grasshopper (Zonocerus variegate), field cricket (Gryllus bimaculatus), blowfly maggot (Chrysomya megacephala), super worm (Zophobas morio), fermented feather meal, feather meal, poultry by-products, fish silage, and shrimp head meal [40,51,52,53,54,55,56] used in aquaculture. Our GFCR values were also within the range of 1.0 to 2.4 reported for farmed fish species, such as Atlantic salmon (Salmo salar), Chanel catfish (Ictalurus punctatus), common carp (Cyprinus carpio), grass carp (Ctenopharyngodon idella), rainbow trout (Oncorhynchus mykiss), and shrimp species, such as giant tiger prawn (Penaeus monodon) and whiteleg shrimp (Litopenaeus vannamei) [49]. Lower efficiency was observed in higher GFCR values reported for D5 and D6 for fingerlings, sub-adults, and adults, possibly because the diets were changed from fry to fingerling diets, and the feeding rate was reduced from 10% to 5%, regardless of the size of the fish. Therefore, maybe 5% was just enough for fingerlings fed D5 and D6 diets to maintain their weight and not enough for them to grow.

The protein efficiency ratio measures how the protein source in a diet can provide the essential amino acids required by fish [37]. Regardless of their sizes, O. mossambicus requires 0.99 g/100 g methionine, 3.79 g/100 g lysine, 0.43 g/100 g tryptophan, 2.83 g/100 g arginine, 1.80 g/100 g tyrosine, 1.05 g/100 g histidine, 2.93 g/100 g threonine, 2.01 g/100 g isoleucine, 3.40 g/100 g leucine, and 2.50 g/100 g phenylalanine [37]. All our diets had less methionine than that required. This could be the main reason for the lower PER in D6, as it was lower in methionine than those required by O. mossambicus, which resulted in lower energy than the other diets.

The survival rates in all size groups from the study were higher than 50%. There were significant differences in all size groups. Fish fed D2 and D5, diets with 50% fishmeal and 50% crocodile meal had significantly higher survival rates than those fed D1, D3, D4, and D6. The mortality recorded was mainly due to fish jumping from tanks, even though nets were used to cover the tanks. The lower survival rates in sub adults and adults may be because the fish stocked as fry were used throughout the experimental period, which was 25 weeks.

Even though the early sexual maturity of O. mossambicus is generally known [57], our study did not use monosex fish, which are preferred in aquaculture projects [58]. We used both males and females O. mossambicus because it was difficult to differentiate them as males or females at the start of the experiment, as smaller fish were used. Furthermore, the tested diets were formulated to be used in commercial production. Therefore, it is essential to use both sexes. Using both males and females in one tank resulted in fish breeding when they were still small. During the sub adult and adult stage, eggs were first observed at the 100% fish meal without crocodile meal, and then at 50%/50% fishmeal and crocodile meal. There were no eggs nor fry observed at the 100% crocodile meal-based diet. These results could mean that crocodile meal is a good source of protein that will delay sexual maturity in O. mossambicus. Tilapia females have a lower growth rate than males in general [59]. Even though it is a common feature of fish in a population to have variation in individual growth, in commercial fish culture, it is a drawback, as size determines the price [60].

The costs for ingredients used in feed preparations were higher in diets with fishmeal than those with crocodile meal in all size groups in the current study.

5. Conclusions

There were no significant differences in weight gain, specific growth rates, gross feed conversion ratios or protein efficiency ratios of O. mossambicus fry fed different diets. Considering the similarities in Gs, SGRs, GFCRs, and PER in fingerlings and sub-adult to adult fish fed D4 and D5, the Crocodylus niloticus meat meal has the potential to substitute fishmeal for all size groups of O. mossambicus. The costs of ingredients used in the diets with 50% and 100% Crocodylus niloticus meat meal indicated that it was profitable to use this meal in diets of O. mossambicus of all size groups. There is a potential for obtaining crocodile meat for free, as it is considered waste to some crocodile farmers; therefore, it could be an economically viable alternative source of protein. Future studies are recommended in growth experiments considering monosex fish prioritized in aquaculture projects, testing other crocodile meal levels and fish health status by determining the hematological parameters.

6. Limitations

Using the same feeding frequency of two times per day for all size groups is regarded as a limiting factor for maximum growth, especially for fry and fingerlings, as they need to be fed more frequently.

7. Recommendations

To minimize costs and maximize production efficiency, different size groups should be fed (feeding rates, feeding frequencies, and composition of diets) according to their sizes.