Nutritive evaluation of fish acid silage in Labeo rohita fingerlings feed

Abstract

Viscera of various fish species was collected during local harvest at the Department of Fisheries and Aquaculture, Ravi Campus, University of Veterinary and Animal Sciences, Lahore pond facility. Acid silage was prepared by systematic applications of formic acid. The proximate analysis of the product revealed 5.16 ± 0.65% moisture, 32.17 ± 0.12% crude protein, 9.56 ± 0.14% lipids and 6.50 ± 0.32% ash contents. Total microbial aerobic plate count was 1.69 × 10⁴± 0.06 × 10³ cfu g⁻¹ while the coliforms were recorded as 0.97 × 10⁴± 0.02 × 10³ cfu g⁻¹. The pure silage was free of aflatoxins B₁, B₂ and G₁ and G₂; however, microbial load and aflatoxin values vary in different feed ratios. The feeding trial showed significant change in all three feeds prepared from different ratios of silage; nevertheless, feed containing 75% acid silage showed better growth in Labeo rohita fingerling diet when compared with its counterparts. Our studies suggest that the fish silage can be a cheaper and effective alternative to fishmeal in fish feeds, if carefully handled and properly processed. This is because fish silage is prepared from fish waste body viscera, which is utilized neither in human nor in animal feeds in the raw form. On the other hand, fishmeal is a main and expensive ingredient used in livestock and fish feeds. Utilization of fishmeal is on the rise while its production is on decline, which is continuously escalating its price. Furthermore, the manufacturing of fish silage is simple and requires relatively lesser inputs as compared to fishmeal manufacturing.

Keywords:

• acid silage
• aflatoxin
• microbial load
• Labeo rohita
• fish growth

1. Introduction

Fisheries sector is playing an important role in the improvement of socio-economic conditions in many countries of the world (Chiarini et al. 2009). Coastline belt in Pakistan is 1050 km long and provides job opportunities to one million people directly or indirectly. But the production of fish from this natural resource is not consistent and is declining due to over and indiscriminate fishing and unchecked flow of pollutants (Hassan 1996). Aquaculture is the only alternate that can compensate these deficiencies and cater the emerging demands of quality protein. Expensive feed stuffs are among the major limitations in the expansion of inland aquaculture as more than 50% of the total operational cost is the cost of feed (Llanes et al. 2008). Protein requirements of fish are satisfied by fishmeal, which is one of the most common feed ingredients in fish feeds. But fishmeal is expensive and its supplies are inconsistent. Therefore, the nutritionists are seeking cost-effective, easily available and quality fish feed ingredients as a substitute of fishmeal (Goddard 1996; Hardy 2008; Santana-Delgado et al. 2008).

The body viscera of fish, viz., intestine, liver and swim bladder, etc., constitute more than 30% of its total body weight. These are typical by-products of fish processing and are not utilized for human food and disposed-off as ‘wastes’ that pollute the environment when discarded (Pillay 1991; Goddard & Perret 2005; Westerman & Bicudo 2005). According to FAO (2008) estimates, more than 20 million tons of body viscera are globally produced per annum, which can be used as a good source of protein. One method of minimizing this waste disposal problem and effectively utilizing fish body wastes is converting it to a safe and utilizable product such as in the preparation of silage (Arruda et al. 2007). Fish silage is brown liquid and is a good source of protein with high digestibility. Silage is produced by a variety of methods and has nutritional value (Tatterson & Windsor 1974). In general, silage prepared by acidic method is preferred over other methods because it has more nutritional value and is comparable to fishmeal (Fagbenro et al. 1994). It can be used as a sole ingredient or can be included in feed formulations with plant by-products as evident in tilapia, Oreochromis niloticus (omnivorous) and the North African catfish, Clarias gariepinus (carnivorous), with productions similar to fishmeal (Hussain & Offer 1987; Fagbenro et al. 1994; Fagbenro & Jauncey 1998). Therefore, the objectives of the present study were to prepare fish silage and then test its efficacy in rohu (Labeo rohita) fingerlings growth.

2. Materials and methods

2.1. Experimental site

The present study was conducted at the Fish Farm Facilities, Department of Fisheries and Aquaculture, Ravi Campus, University of Veterinary and Animal Sciences, Lahore.

2.2. Preparation of silage

Fish body viscera were collected during harvesting season. Fish waste was minced by electric mincer and then 3% (weight by volume) formic acid was added in the minced fish waste to lower the pH up to 3.5 and 250 mgL^–1 Butylated hydroxytoluene (BHT) antioxidant was added to inhibit lipid oxidation. The mixture was stored at 30°C for a period of 30 days with daily thorough mixing and its pH was also checked and maintained at 3.5 to avoid petrification (Oetterer 2002). Initially, the mixture was in semi-solid form but it started to liquefy on the fourth day. Silage was ready to use after 30 days.

2.3. Proximate composition

Proximate composition of prepared fish acid silage was determined by standard methods of Association of Official Analytical Chemists (2000) no 950-46, 992-15, 991-36 and 920-163, for moisture, protein, fat and ash contents, respectively.

2.3.1. Moisture determinations

Wet silage sample was weighed, placed in Petri dish and then dried in oven overnight at 105°C for overnight. Petri dish was taken out the next day and weighed again. The loss in weight represented the moisture contents and was determined. The percentage is determined by the following formula: where W₁ = initial weight of the sample

W ₂ = final weight of the sample

2.3.2. Protein determination

Five grams of dried fish silage sample was taken in a flask and mixed with digestion mixture (potassium sulphate + copper) and transferred to a flask containing 200 mL of concentrated H₂SO₄. This flask was placed on a heating block, the heaters were turned on and the flask was kept there until white fumes stopped appearing and the solution became clear, indicating completion of the digestion process. The solution was removed away from the heater and then cooled. The solution was diluted with the addition of 60 mL of distilled water and its pH was raised to 6.5–7 by adding 45% NaOH solution. Then five to six drops of indicator solution was added and the flask was connected with a condenser with the tip immersed in standard acid and heated until NH₃ was evaporated. The final solution mixture was then titrated against NaOH. Protein contents were then determined applying the following mathematical formula:

A = volume of 0.2 N HCl used for sample titration

B = volume of 0.2 N HCl used in blank titration

N = normality of HCl

W = weight of sample

14 = atomic weight of nitrogen

6.25 = constant for nitrogen calculation

2.3.3. Ash determination

Ten grams of sample was taken in a crucible and weighed. Crucible with sample was placed in muffle furnace at a temperature of 550°C for 5–6 hours. When the sample turned white, it was taken out and weighed again. White-coloured contents remaining at the bottom of the crucible represented ash, which was carefully weighed and its percentage present in the feed was calculated by the following formula.

2.3.4. Lipid extraction

The soxhelt apparatus was set and 5 g of sample was placed in the extraction thimble and transferred to the condenser. Petroleum ether was filled in a flask and the apparatus was switched on. This process was continued for 16 hours. Then turned the heaters were switched off, and the flask was removed and gently dried on the same heater. When the contents of the flask smelled oily, they were removed and weighed and the fat content in the test sample was calculated using the following formula.

2.4. Microbial load

Total microbial load and coliform count was determined in the dry acid silage following the procedures of USP (2005) and USP (28-NF23).

2.5. Detection of aflatoxins

Aflatoxins were detected following the procedure described by Pitt and Hocking (1997). Twenty-five grams of dry silage was taken in a 500-mL flask and then 90 mL chloroform (stabilized with 1% ethanol), 10 mL methanol and 10 mL water were added to it. The mixture was stirred for 15 min at 1300 rpm and then it was filtered. Five microlitres of the filtrate was loaded and run on a thin-layer chromatography (TLC) plate until it reached one third of the plate. The TLC plate was then taken out, dried and examined under TLC UV lamp.

2.6. Feed formulation

After proximate analysis, three diets having 30% protein were prepared containing (1) 100% fish silage, (2) 75% fish silage and (3) 50% fish silage and were labelled as T₁, T₂, T₃. Diet without silage served as control T₄, which contained fishmeal instead of silage (Table 1). Soybean meal and rice bran were added with silage to maintain the protein level at 32% with different ratios (Table 1). The pH of silage was increased by the addition of calcium carbonate to obviate any damage to fish during intake of feed and subsequent digestion.

Table 1. Percentages of various ingredients and their protein level used in four diet formulations based on variable levels of fish silage.

	100% silage (T1)		75% silage (T2)		50% silage (T3)		Control (T4)
Ingredients	Weight (g)	CP%	Weight (g)	CP%	Weight (g)	CP%	Weight (g)	CP%
Fishmeal	–		–				32	9.8
Fish silage	100	32.17	75	24	50	16.08	–
Soybean meal	–		10	5.7	22	13.11	29	16.5
Rice bran	–		15	2.5	23	3.83	39	6.5
Total	100	32.17	100	32.2	100	33.02	100	32.8

2.7. Feeding trial

All the diets were prepared and offered to fish in powder form. Labeo rohita fingerlings showing average weight 2.34 ± 0.05 g and average length 7.6 ± 0.03 cm were collected from the fish nursery at Manawan, Lahore, and were acclimatized in fibre glass tanks (70 × 32 × 45 cm) for 5 days then were divided into four groups; three groups received experimental diets and the fourth group was kept on control diet (without silage). All the dietary treatments including control had three replicates. Fish weight and length were taken at the time of initiation of the experiment, and thereafter, the increase in fish weight and length were noted on a fortnightly basis. All the fish were fed at 4% of their body weight twice a day at 8 am and 4 pm. Before feeding, the leftover feed and faecal matter were removed from water. At the end of the trial, all the fish were harvested, weighed and measured, and some fishes were dried for the determination of body composition. Water quality parameters like temperature, pH and dissolved oxygen (DO) of aquarium water were monitored on a daily basis.

2.8. Growth and feed intake

Average weight gain (AWG), average length gain (ALG), specific growth rate (SGR%), feed conversion ratio (FCR) and survival rate were computed using the following formulae:

AWG = final weight (W _f) − initial weight (W _i)

ALG = final length (mm) − initial length (mm) FCR = total dry feed fed/total wet weight gain. where N₀ = initial number of fish in each replicate

N _t = final number of fish in each replicate.

2.9. Fish body proximate composition

At the end of the trial, proximate composition of under trial fish were determined following the standard methods of Association of Official Analytical Chemists (2000) no. 950-46; 992-15; 991-36 and 920-163, for moisture, protein, fat and ash contents, respectively.

2.10. Statistical analysis

The data obtained were analysed by SAS (1999; statistical package; version 8.0 for Windows). Data obtained from studies based on completely randomized experimental design were subjected to one-way analysis of variance. Results were considered significant at P < .05. Means of each treatment including control then were compared using Duncan’s multiple range test for level of statistical significance among treatments.

3. Results and discussion

3.1. Silage preparation and proximate analysis

Acid silage prepared during the present study was a brown, viscous liquid with strong fishy smell having pH values between 3.5 and 4.5. Acid silage can be satisfactorily preserved at pH values ranging from 3.5 to 4.5 (Vidotti et al. 2002). Similarly, Toledo et al. (2007) reported that acid silage pH values <4.5 represent good quality silage.

Silage prepared during the current study contained 5.16 ± 0.65% moisture, 32.22 ± 0.17% protein, 9.56 ± 0.14% fats and 6.50 ± 0.32% ash (Table 2). However, proximate composition of silages is not consistent as values vary from case to case because they depend on several factors like storage time, raw material and local environmental conditions (González & Marín 2005; Albrecht & Torpoco 2008). According to Bureau et al. (2000), the fat content of fish silage depends on fat content of the raw materials as well as its source and season of harvesting.

Table 2. Nutrient and anti-nutrient evaluation of formulated diets based on different ratios of acid silage.

Parameters	100% silage (T1)	75% silage (T2)	50% silage (T3)	Control (T4)
Moisture, %	5.16 ± 0.65a	5.38 ± 0.66a	5.67 ± 0.66a	5.13 ± 0.71a
Protein, %	32.17 ± 0.12a	32.20 ± 0.73a	32.02 ± 0.97a	32.83 ± 0.35a
Fats, %	9.56 ± 0.14a	8.41 ± 1.43ab	7.62 ± 1.33b	8.14 ± 1.03b
Ash, %	6.50 ± 0.32b	8.40 ± 0.45a	7.63 ± 0.41ab	4.90 ± 0.91c
Microbial load, cfu g^−1	1.69 × 10^4 ± 0.6 × 10^3a	1.29 × 10^4 ± 0.4 × 10^3b	1.35 × 10^4 ± 0.7 × 10^3b	1.27 × 10^4 ± 1.3 × 10^3b
Coliform, cfu g^−1	0.97 × 10^4 ± 0.02 × 10^3a	0.81 × 10^4 ± 0.03 × 10^3b	0.77 × 10^4 ± 0.07 × 10^3bc	1.09 × 10^4 ± 0.08 × 10^3a
Aflatoxin B1, ppb	0 ± 0.00c	0 ± 0.00c	2.43 ± 0.00a	0.97 ± 0.43b
Aflatoxin B2, ppb	0 ± 0.00c	9.21 ± 1.76b	15.00 ± 0.76a	7.87 ± 0.70b
Aflatoxin G1, ppb	0 ± 0.00c	7.23 ± 1.76a	7.34 ± 0.76a	5.87 ± 0.70b
Aflatoxin G2, ppb	0 ± 0.00c	13.67 ± 0.94a	14.43 ± 0.36a	6.78 ± 0.41b

Note: Values given are mean ±standard error. Values with the same letter in the same row are not statistically significantly (P < .05) different.

3.2. Microbial load and aflatoxins

The prepared silage showed a total microbial load of 1.69 × 10⁴ ± 0.06 × 10³ cfu g^–1 while the coliform count was 0.97 × 10⁴ ± 0.02 × 10³ cfu g^–1. However, the acceptable limit for total aerobic plate count of food product is 10⁷ cfu g^–1 wet weight (ICMSF 1978). The values for total microbial load and coliform were within acceptable limits rather in lower concentration. When the prepared silage was tested for aflatoxin AFB₁, AFB₂, AFG₁ and AFG₂, none of these was detected and silage was found to be totally aflatoxin free, which might be due to lower pH values of silage, which hindered the growth of aflatoxins (Avantaggiato et al. 2005), whereas the aflotoxin values in the other three feed treatments T₂, T₃ and T₄ were within limits, i.e. less than 20 ppb. The acceptable limits of aflatoxin in feed are less than 20 ppb (Lovell 1992). The silage prepared in the current study was aflatoxin free and, hence, can be safely used in fish feeds.

3.3. Water quality parameters

In the present study, water temperature (°C), DO (mg L⁻¹), pH and alkalinity (ppm) remained uniform among all the treatments (Table 3) and were within acceptable limits for optimum growth of L. rohita fingerlings as described by Boyd (1982) and Abid and Ahmed (2009). Consistency and uniformity of water quality parameters among treatments revealed that the presence of acid in silage did not have any bearing on water quality.

Table 3. Water quality parameters during the 90-day feeding trial.

Treatments	pH	Temperature (°C)	DO ppm	Alkalinity mg L^−1	TDS mg L^−1
T1	7.1 ± 0.93a	31 ± 0.43a	5.5 ± 0.45a	130 ± 1.34a	450 ± 0.31bc
T2	7.5 ± 1.23a	31.5 ± 0.54a	6.5 ± 0.54a	130 ± 0.73a	473 ± 0.43bc
T3	7.3 ± 0.86a	31 ± 0.98a	6.0 ± 0.69a	131 ± 0.45a	480 ± 0.42b
T4	7.5 ± 0.56a	30 ± 0.88a	6.1 ± 0.73a	130 ± 0.98a	574 ± 0.21a

Note: Values given are mean ± standard error. Means with the same letter in the same column are not statistically significantly (P < .05) different.

3.4. Feeding trial

According to Inayat and Salim (2005), growth and FCR are the important indicators of the acceptability of feed in fish-feeding experiments. Data regarding the growth of fish fed with100%, 75% and 50% acid silage and control diets, viz., weight gain, length increase, FCR, SGR and survival rate, are shown in Table 4. In the present study, the prepared silage was fed to L. rohita fingerlings with an average weight of 2.67 ± 0.89 g. Statistically significant variations (p < 0.05) were recorded in weight gain among all the four treatments containing acid silage percentages, viz., in T₁, T₂, T₃ and T₄ and maximum weight gain was recorded in T₂ diets (75% acid silage). Our findings are in line with those by Srinivasan et al. (1985) who observed better weight gain and growth rates in Cyprinus carpio fed with silage-based experimental diets as compared to control. Better growth rate of fish fed with 75% acid silage may be due to the presence of comparatively higher amount of free amino acids and active hydrolytic enzymes (Gallagher 1993).

Table 4. Weight gain, length gain, FCR, SGR and survival rate of Labeo rohita fingerlings.

Treatments	Total weight gain	Total length gained	FCR	SGR	Survival, %
T1	6.70 ± 2.38ab	7.96 ± 0.81a	3.14 ± 1.36a	0.85 ± 0.38a	100 ± 0.00a
T2	8.23 ± 2.73a	7.91 ± 0.96a	1.98 ± 1.44b	0.92 ± 0.12a	100 ± 0.00a
T3	5.94 ± 2.75b	7.83 ± 1.14a	3.21 ± 1.42a	0.73 ± 0.22b	100 ± 0.00a
T4	6.41 ± 3.35b	7.36 ± 1.29a	2.44 ± 1.08b	0.83 ± 0.09ab	100 ± 0.00a

Note: Values given are mean ± standard error. Means with the same letter in the same column are not statistically significantly (P < .05) different.

During present study, all the fish were fed at 4% of net fish body weight. Salim and Sheri (1999) also reported that 4% feeding of L. rohita showed better growth performance. Lower FCR values were observed for T₂ diet (75% acid silage), while the same was highest for T₃ diet (containing 50% acid silage). The FCR values recorded in the present study for T₂ was more economic (1.98 ± 0.44) as compared to other diets. FCR value for our prepared diets were lower than the values observed by Ali and Salim (2004), who recorded 5.27 FCR for rice polish and 3.02 FCR for fishmeal; however, Jabeen et al. (2004) reported similar values for cotton seed meal diets. The difference in FCR may be due to different fish used and different feed type. The FCR values for all the other treatment diets are in line with the findings of Shabbir et al. (2003).

During the present study, the SGR varied significantly between the control and the treatment diets; however, 100% survival was observed in all the treatment and control diets. At the termination of experiment, proximate analysis of fish body showed that acid silage treatment diets have no adverse effects on body composition of fish and fish body composition did not vary significantly.

3.4.1. Fish body composition

The proximate composition of fish body is shown in (Table 5). The results indicate that there is a significant difference among the three treatments in some values (P < .05). Protein contents were recorded as 57.24 ± 0.67, 63.01 ± 0.66, 58.53 ± 0.29 and 55.16 ± 0.89 in T₁, T₂, T₃ and T_4, respectively. According to Shearer (1994), the external factors like feed, season and other environmental conditions and internal factors of fish like age, sex and reproduction period mainly affect the final composition of the fish body. The lipid contents in the fish body vary by 16.43 ± 0.92, 17.47 ± 1.49, 11.83 ± 0.29 and 18.77 ± 0.27 in T₁, T₂, T₃ and T_4, respectively. According to Hanley (1991), the fish lipid contents are important in fish muscle and carcass, and according to Oduor-Odote and Kazungu (2008), the lipid contents in fish vary according to variation in feed intake. So the present variation in lipid contents may be due to a variation in feed intake and other factors described earlier. The ash contents in fish body were observed as 15.41 ± 1.67, 10.97 ± 0.34, 15.05 ± 0.47 and 7.52 ± 0.39 in T₁, T₂, T₃ and T_4, respectively. Many biologists have confirmed that fish body composition may vary due to a difference in the environmental condition, age, sex and water quality and state of maturity of fish (Brett et al. 1969; Craig et al. 1989).

Table 5. Proximate analysis of Labeo rohita fingerlings after the 90-day feeding trail.

Parameters	100% silage (T1)	75% silage (T2)	50% silage (T3)	Control (T4)
Moisture	8.13 ± 0.78a	9.71 ± 1.23a	8.53 ± 0.45a	4.09 ± 1.11b
Protein	57.24 ± 0.67b	63.01 ± 0.66a	58.53 ± 0.29b	55.16 ± 0.89b
Fats	16.43 ± 0.92b	17.47 ± 1.49ab	11.83 ± 0.29c	18.77 ± 0.27a
Ash	15.41 ± 1.67a	10.97 ± 0.34b	15.05 ± 0.47a	7.52 ± 0.39c

Note: Values given are mean ± standard error. Values with the same letter in the same row are not statistically significantly (P < .05) different.

4. Conclusion

It can be concluded from the present experiment that fish silage can be successfully used in aqua feeds to replace fishmeal to reduce feed cost.

Acknowledgements

The authors wish to express their gratitude to the Pakistan Higher Education Commission, Islamabad, for providing funds for the completion of this project.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Pakistan Higher Education Commission Islamabad [grant number 1415-Bm4-250].