FEED Business Worldwide - October / November, 2011
Effect of feeding sweet potato peels on growth, feed utilisation and performance of nile tilapia (Oreochromis niloticus)
 The population of the world is growing at an exponential rate [1]. This situation calls for quick action and an aggressive approach tailored at food production to feed the already high human population in order to ameliorate inadequate food supplies and the consequential malnutrition. One of the promising solutions to the shortage of animal protein intake in developing countries is the proper development of aquaculture [1]. Fish feed is the most expensive input in aquaculture operations [2]. Much of the high cost of feed arises from extensive reliance on protein sources, such as fishmeal and shrimp meal [3, 4]. The shortage and high cost of pelleted feed severely constrained the development of low cost aquaculture systems suitable for small-scale farmers in the developing countries; hence, the need to assess the potential of non-conventional raw ingredients such as the sweet potato peels. It would therefore be more economical to utilize plant protein in fish feeding than high cost animal protein materials [5].
The use of plant-derived materials as fish feed ingredients is limited by the presence of a wide variety of anti-nutritional substances [6]. Among these are protease inhibitors, phytates, glucosinolates, saponins, tannins, lectins, oligosaccharides and non-starch polysaccharides, phytoestrogens, alkaloids, antigenic compounds, gossypols, cynogens, mimosine, cyclopropenoid, fatty acids, canaranine, antivitamins and phorbol esters.
These authors noted that protease inhibitors, phytates, antigenic compounds and alkaloids, at levels usually present in fish diets containing commercially available plants-derived protein sources, are unlikely to affect fish growth performance. In contrast, glucosinolates, saponins, tannins, soluble non-starch polysaccharides, gossypol, and phorbol esters are more important from a practical point of view. However, as noted by Kays [7], the sweet potato peel is devoid of most of these agents as the sweet potato plant usually stores these chemicals in its tubers.
The test fish, Oreochromis niloticus are principally herbivorous, although occasionally omnivorous. This is an efficient converter of waste foodstuff and appears to thrive well on artificial supplemental feed [8].
The main aim of this research is to reduce the cost of conventional fish diets while the objective is to investigate the effect of sweet potato peels inclusion in fish diet on growth responses of Oreochromis niloticus, food utilisation and glucose and protein content of the blood plasma.
Collection and acclimatization of experimental fish
Four hundred and twenty mixed-sex fingerlings of the cichlid, Oreochromis niloticus of the same broodstock (mean weight 0.47 ± 0.01 g) obtained from the hatchery of the Rockwater Fish Farm, Jos, Nigeria, were used for this investigation. The fingerlings were acclimatized to laboratory conditions for  14 days.
Formulation of experimental diets
The peels used in diet formulation were obtained from fresh sweet potatoes, washed properly and sun-dried for ten days under hygienic conditions (placed on pre-sterilized white tray and covered with wire-mesh) , after which they were winnowed and sieved to get rid of any foreign materials. The peels were then milled into a fine powder and sieved through a 0.5 mm mesh screen. Proximate analysis of the processed peel is presented in Table 1.
Other raw ingredients of the experimental diets were fishmeal, groundnut cake, corn meal, cassava flour, corn oil, a-cellulose, chromic oxide, and vitamin and mineral premix. These diets were designated SPP0 (control), SPP5, SPP10, SPP15, SPP20 and SPP25, respectively. Table 2 shows the ingredient and proximate compositions of the experimental diets. All experimental diets were iso-nitrogenous (31.23 ± 0.22% crude protein) and iso-caloric (26.35 MJ/kg). Chromic oxide (Cr2O3) was incorporated in the diets to serve as the faecal indicator [9] for the determination of nutrient digestibility.
Experimental set-up
The design utilized eighteen plastic tanks of 20-litre capacity and a constant water volume of 17 litres maintained in each tank. Prior to feeding of experimental diets, the fish were starved overnight to empty their gut and increase their appetite and reception for the new diets. A total of 24 fish were randomly selected, sacrificed and used in determining initial carcass proximate composition.
For the experimental trials, 20 fish were weighed using Mettler Toledo PB602 top-loading balance and introduced into each of the experimental tanks. Each tank was assigned to one of the six experimental diets (triplicate tanks per diets). The fish were fed (3% body weight) twice daily at 0800 and 1800 h. Faecal samples were siphoned out and the water replaced daily before any subsequent feedings. The collected faecal samples were dried and stored for subsequent digestibility determination.
Fish in each experimental tank were collectively weighed weekly for the determination of growth rate. The experimental period lasted 10 weeks. Mean water quality parameters during the experimental period were 5.62 ± 0.09 mg/l, 2.04 ± 0.05 mg/l, 22.60 ± 0.18 mg/l, 20.67 ± 0.18OC and 6.19 ± 0.02 for dissolved oxygen, free carbon dioxide, alkalinity, temperature and pH, respectively.
Indices of growth, feed utilisation and biochemical parameters
At the end of the experimental period, the following growth and feed utilisation indices were calculated; percentage weight gain, specific growth rate (SGR), food conversion efficiency (FCE), food conversion ratio (FCR) and apparent net protein utilisation (ANPU). These factors were calculated using the following equations as described by Brown [10], Winberg [11], Castell and Teiws [12] and Miller and Bender [13], respectively:
At the end of the feeding trials, fish from each tank was subjected to plasma glucose and plasma total protein analyses using basic anthrone and electrophoresis methods, respectively as described by Wedemeyer and Yasutake [14].
Statistical analysis
Analysis of Variance (ANOVA) for completely randomized design was used at 95%  significant level to test for significant differences between the various treatment means obtained for the growth, feed utilisation parameters, carcass composition and biochemical parameters (plasma glucose and protein). The least significance difference (LSD) test was used to determine which pairs of the treatment means differed significantly. The chi-square test was employed to analyse the digestibility values.
Food utilisation indices
The groups of fish fed on diet SPP0 utilized the experimental diets better than the other groups of fish. On the other hand the groups of fish fed diet SPP25 had the least feed utilisation indices (Table 4). Values of ANPU obtained for the six experimental diets were not significantly different (P > 0.05) for all the groups of fish.
The mean weekly cumulative weight gain of the experimental fish Oreochromis niloticus fed varying levels of sweet potato peel meal for 10 weeks is presented in Figure 1. Results of mean weight gain and SGR of the experimental fish at the end of the feeding trials are presented in Table 3. The initial weight of the fish was near constant (0.47 ± 0.01 g) in all the experimental diets as there was no significant difference (P > 0.05) among the six treatments.
The groups of fish fed diet SPP0 (control diet) had the highest percentage weight gain (284.81 ± 17.42%), and this was significantly different (P < 0.05) from the percentage weight gain of the groups of fish fed on diets SPP5, SPP10, SPP15 and SPP20, while the groups of fish fed diet SPP25 recorded the least percentage weight gain (96.01 ± 8.25%). Similarly, the groups of fish fed diet SPP0, had the highest SGR value (1.92 ± 0.06%/day), which was significantly higher than (P < 0.05) all the other treatments, while the least SGR value was obtained from fish fed diet SPP25 (0.96 ± 0.06 %/day). 
The optical readings for six experimental diets, the corresponding faecal samples and the calculated apparent digestibility values are shown in Table 5. Statistical analysis indicates no significant differences (P > 0.05) amongst the various dietary treatments. In absolute terms, the nutrient digestibility of the experimental diets by the fish was greatest (30.50%) in the control diet (SPP0) and lowest in diet SPP25 (28.00%).
Proximate composition of fish carcass
The proximate composition of experimental fish carcass at the beginning and end of the feeding trials are presented in Table 6. Carcass of fish fed diet SPP25 had the least protein content, which was significantly different from the other diets except SPP15. The ash content was near constant, as there was no significant difference amongst the fish fed the six experimental diets. Fish fed diet SPP0 recorded the highest level of lipid deposit which was significantly different from the other dietary treatments.
Plasma glucose and protein
Mean values of the plasma glucose and protein of the experimental fish after the feeding trials are presented in Table 7. The groups of fish fed control diet had the least mean values of plasma glucose, which was significantly different (P < 0.05) from the mean values obtained for the groups of fish. The groups of fish fed control diet recorded the highest mean values of plasma protein, which was significantly different (P < 0.05) from the mean values obtained for the other groups of fish.
Discussion & implications
The nutritional quality of sweet potato peel meal as determined by growth indices in this study was adequate, and in terms of survival the sweet potato peel can be successfully incorporated in the diets of cultured fish species, as no mortality was recorded in the experimental tanks during the feeding period. Although final body weight and growth rate were higher in fish fed the control diet. However, no deleterious depression in growth was observed in fish fed diet incorporated with 5, 10, 15 and 20%, respectively. However, at higher inclusion level (25%) reduced performance by the fish was observed. Secondly, one of the most common difficulties observed when alternative sources of feedstuffs are used in fish diets is acceptance and palatability by the fish [15]. However, in this present study, the fish avidly consumed the experimental diets.
The above are excerpts, full versions are only available in FEED Business Worldwide. For subscriptions enquiries, e-mail membership@efeedlink.com