Apparent Digestibility and Utilization of Protein and Phosphorus in diets of incorporated with Sprouted Sorghum, Phytase and Protease Enzymes for African Catfish (Clarias gariepinus)

Apparent digestibility, deposition and retention of crude protein and phosphorus were determined for 30%, 35%, 50% and 55% CP diets incorporated with sprouted sorghum, phytase and protease enzymes ,fed to African catfi sh fi ngerlings. All digestibility coeffi cients signifi cantly differed (p<0.05) among diets with 92.66% observed for phosphorus in 30% and 35%CP diets with1250 units of phytase, and 81.49% for protein in 50%CP and 55% CP diets with 1250 units of protease. Crude protein deposition and retention signifi cantly differed (p<0.05) with the highest (705.10 and 10.03 respectively) recorded in diets with 1250 protease while that of phosphorus did not in both cases (p>0.05). This demonstrated that protease was more useful in starter (high protein) diets and phytase enzyme in grower (low protein) diets and forms a basis for effi cient use of phytase and protease enzymes in catfi sh diet formulations for different growth stages. Research Article


Introduction
The North African catfi sh (Clarias gariepinus) is the main farmed fi sh species in Africa due to its fast growth and tolerance to handling and stress [1,2]. its culture is dependent on supply of quality diets which comprise about 60-70% of the operational costs. Raising catfi sh from fry/post hatch to table/ market size requires diets with 55% -30%CP for normal growth (0.5-1kg in 6 -8 months). Such quality diets have historically been made from fi sh meal. However human consumption of small pelagic fi sh species like silver fi sh used in production of fi sh meal reduced its supply and use in fi sh diets due to its high cost .This led to adoption of plant protein due to wide distribution and relatively low cost. Plant ingredients however contain antinutrients including the heat stable phytic acid that bind and reduce digestion of nutrients to support normal fi sh growth. Poorly digested feed pollute the environment through nutrient enrichment reducing the economic and social viability of fi sh farming. Unlike other antinutrients, phytic acid is found in almost all plant material [3,4]. It is not economically destroyed by a number of conventional food processing techniques including heat treatments like roasting [5]. Among techniques that unlock nutrients in plant material, sprouting is an ancient practice especially in legume and cereals.
According to [6,7], sprouting activates a number of naturally occurring enzymes in plant material including phytase that exist in insoluble, low activity (210-670Ukg -1 ). The increase in enzymatic activity breaks down antinutrients into simple, easily absorbed nutrients. Sprouting reduced phytic acid subsequently increasing phosphorus and protein in plant fi sh diets and hypoallergic baby weaning diets respectively [8]. Incorporation of sprouted sorghum into carbohydrate rich foods is also widely practiced to induce fermentation during which complex starch is reduced to simple readily assimilated glucose. To reduce processing time, most enzymes observed during sprouting have been extracted and packaged for a number of industrial processes. This not only made them expensive but also altered their effi ciency because a single plant ingredient can have different antinutrients thus requiring different enzymes whose ratio in nature have not yet documented. For fi sh feed, enzyme sources that cope with high pelleting temperatures (>70oC), variations in diet composition and physiology of the digestive tract among species are required. Due to scanty information on digestibility of diets incorporated with sprouted grain and purifi ed enzymes, effects of sprouted sorghum, phytase and protease enzymes on apparent digestibility of organic matter, crude protein and phosphorus was evaluated to guide effi cient enzyme use.

Experimental design
Two cascading experiments of completely randomized factorial design were conducted. The fi rst assessed the effect of dietary enzyme types and the second assessed the effect of phytase and protease levels on digestibility of catfi sh diets.
The fi rst experiment was of a 4 by 4 randomized factorial design in which diets of 30%, 35%, 50% and 55% CP were each incorporated with sprouted sorghum, 750 units of phytase and protease with feed samples without enzyme as negative controls. This resulted into 16 experimental diets that were randomly assigned to 48 glass tanks in three replicates and used to assess the effect of enzyme type on digestibility.
After results of the fi rst experiment showed that apparent digestibility of diets incorporated with phytase and protease were higher than for the control and sorghum, the second experiment started to determine the optimal level of phytase and protease inclusion.
The second experiment was of a 4 by 6 randomized factorial design. Diets of 30%, 35%, 50% and 55% CP were each incorporated with 750, 1000 and 1250 units kg -1 of phytase and protease enzymes. This gave rise to 24 enzyme /diet treatments which were randomly assigned to 72 glass tanks in triplicates and used to assess the effect of enzyme level on digestibility.

Formulation of experimental diets
Diets of 30% and 35%, 50% and 55% CP were formulated using Feed Win Least Cost computer software, developed by Practical Training Center (PTC+), 1999 (currently AERES Training Centre International), Barneveld, Netherlands. For the fi rst study, each diets was incorporated with 0.15gkg -1 of phytase enzyme (Bacillus lincheniformis 5000AUg -1 (® KOFFOZYME Israel) and 0.00125gkg -1 of protease enzyme; (Trichoderma reesei, 600,000 AUg -1 (®Cibenza, Novus international, USA) all worth 750FTUkg -1 and 10gkg -1 of sprouted sorghum fl our. A portion in each case was not incorporated with any of the enzymes and served as a negative control (Table 1a). For the second experiment, phytase was incorporated at 0.2g for 1000 units and 0.25g for 1250 and protease, 0.00167 were for 1000 units and 0.00208 g for 1250 units/kg-1 (Table 1b). All the diets were incorporated with chromium (iii) oxide (Cr 2 O 3 ) as inert marker at a rate of 0.05% (500g/kg1) (grade (Table 1c) [9].
The diets and enzymes were measured using a digital scale (Sartorius AG Germany, CPA52025 DS), to the nearest 0.01, mixed into water and then mixed into each diet category. The diet-enzyme mixture was left to stand for 40 minutes as incubation time and pelleted sinking 2mm diameter pellets at 56 -70 o C (MUSA BODY, pelleting machine Kampala, Uganda). The pelleted diets were dried under a well-ventilated Iron roofed shade for two days. They were after packed in 2kg water proof laminated plastic bags and kept on wooden shelves at room conditions.

Stocking and management of experimental glass tanks
Experimental glass tanks of 20 litre capacity were placed in doors on wooden shelves. The room was made dark by sealing off all light spaces along windows and doors using a black dump proof course (DPC) polythene paper to reduce stress associated with light sensitivity of catfi sh [10]. Each tank was fi lled with about 15ltr of dechlorinated water at room temperature (average 22 o C). Water was constantly aerated using rubber air diffusers connected to an air compressor by rubber tubings. Each tank was stocked with fi fteen (15) catfi sh fry of 10.55±2.06 gm (total length 8.25±1.34 cm). Fish was acclimatized to glass tank feeding and faecal collection by siphoning by feeding them on 2mm diameter 30% CP diet for a week. After which, experimental diets were randomly assigned to the stocked glass tanks in triplicates. The catfi sh fry were fed at 5% of their body  weight with the daily ration divided into two portions that were fed from 10: 00 -11:00hrs and 05:00-06:00hrs daily for 60 days (2 months). Fish tanks were cleaned of leftover feed an hour after every feeding and faeces collected daily by siphoning with rubber tubing. Each time feacal matter was collected, it was placed in labeled feacal containers and kept frozen at -20 o C.
At the end of the experiment, all faecal samples from all the triplicate treatments were thawed and then pooled together into bigger containers. Water was constantly aerated using air blowers attached to an air compressor (HAILE ACO-328) and replaced at a rate of 50% once daily during the morning cleaning to maintain dissolved oxygen concentration between 3-4mgl -1 .

Data collection
Water quality parameter: Dissolved oxygen; pH, water temperature, ammonia, reactive phosphorus (orthophosphate) and alkalinity were determined using a Sera Aqua-test kit with reagents (Sera GmbH, Heinsberg, Germany). About four (4)  and 1700-1800hours at a rate of 5% of their body weight.
Satiation was judged to have been reached when un eaten pellets were retained at the tank bottom. An hour after feeding ( at 10: 00hrs and 18:00hrs) the uneaten food was siphoned off the tank bottom according to [11], onto a fi ne sieve and gently transferred into 10ml plastic faecal containers. Care was taken to count the number of whole uneaten pellets whose weight was estimated from taking the weight of a similar number of unfed dry pellets and then subtracted from the weight of total administered feed to estimate the feed intake (Kg).

Weight gain
At the beginning and end of the experiment, the fi sh were fasted for 24 hours. Seven (7) samples of fi sh were randomly picked from each experimental treatment and their live weight measured to the nearest 0.01 g using a digital weighing scale.
The difference between initial and fi nal live weight was used as the weight gain of the fed catfi sh pellets from the supplied and estimating the weight of the same number of dry pellets from each respective experimental diet.

Proximate nutrient composition
At the start and end of the experiment, three fi sh samples from each treatment were collected and humanly sacrifi ced (100 mg/1 benzocaine).They were dried to constant weight and ground into fi ne powder (MUSA Body, Kampala Uganda).
They were assessed for whole body proximate nutrient composition following standard methods by [12]. Dry matter was determined by the oven drying to constant weight (Gallencamp) SANYO,OMT OVEN, ash by muffl e furnace, crude fat by ether extraction according to [13] and crude protein based on Kjeldahl's method using a selenium catalyst, (N, 6.25) and energy (bomb calorimeter: Gallenkamp Autobomb, calibrated with benzoic acid). Chromic oxide concentration in the diets and faecal samples were determined according to [14,15].

Determination of apparent digestibility
The indirect method of determining digestibility according to (De Silva and Anderson, 1995; Guillaume et al., 2001;Jobling, 1994) was used with Chromium (iii) oxide as an inert marker at a rate of 0.5%. In all cases apparent digestibility coeffi cients of dry matter was calculated according to [16,17].

Determination of nutrient utilisation
Data in percentages/proportions was fi rst transformed into arcsine as (ASIN (SQRT (number)) and later back-transformed by squaring the sine of the arcsine. Data that was not normality distributed even after transformations was analyzed using the non-parametric Kruskal-Wallis one-way analysis of variance (ANOVA). Data with normal distribution and homogeneity of variance was analyzed using the parametric two way analysis of variance (ANOVA). While considering interaction. Data with signifi cant variance of the mean was subjected to pair wise comparisons using Tukey's test. In all cases signifi cant differences were declared at 95% confi dence interval (p<0.05).

Fish survival and weight gain
Percentage survival, the weight gain and fi nal weight (in   (Table 4b). On the other hand, apparent    digestibility of organic matter was observed to be high in diets with 750 protease and least in 750 phytase (Table 4b).

Protein and phosphorus retention
Equally, crude protein retention differed among enzyme types (H=13.22, df = 3, p = 0.004) as well as among phytase and protease levels (H=47.5, df=5, p<0.001) while phosphorus retention did not (H = 4.752, df = 3, p = 0.191. and H=1.35, df=5, p=0.93). Retained crude protein varied from 4.12 -6.48 among enzyme types and from 4.74-10.03 among phytase and protease levels. It was in all cases higher in diets incorporated with protease than in those with phytase and the control ( Table 5). Phosphorus retention varied from 1.20 -3.16 among enzyme types and from 0.69-3.91 among phytase and protease inclusion levels ( Table 6).

Water quality
The high ammonia (3.33-11.30) and phosphorus recorded in diets incorporated with protease in diets phytase (5.29-10.55 and 7.43-11.28) respectively was attributed to liberation of crude protein and phosphorus by these enzymes The ammonia and phosphorus levels were however more than torarable levels for catfi sh (0.2mg/l, 2mg and 0.3 mg/l (20) and allowable discharge/effl uent quality of 10mgl -1 in Uganda [21]. This was attributed to accumulation of organic waste in form of disintegrated left over feed and feacal matter. However no fi sh deaths were registered as portrayed by high fi sh survival rates (88.89 to 100%). This was attributed to the high levels dissolved oxygen (3.78 -4.32 mgl -1 ) that were provided by constant aeration and the morelese constant and neutral water pH (6.76 -7.01), optimal for catfi sh growth (Tables  3a,3b). The observed pH stability was thought to be related to the relatively high carbonate hardness (alkalinity) of the water used (Table 3b). The high phosphorus concentrations (5.29-11.28 mgl -1 ) were attributed to accumulation of wastes especially uneaten feed and feaces that could not be siphoned before it disintegrated in water. Unlike in pond systems were the soil and some phytoplankton can absorb some ammonia and phosphates, no organisms (other than fi sh) would help in glass tanks. These water quality parameters could have negatively infl uenced fi sh growth among other factors and the actual effect was not established.
None signifi cant differences were recorded for survival, fi nal weight and weight gain and FCR among enzyme types or their inclusion levels. This was attributed to the infl uence of aeration and apparent digestibility.
The high apparent digestibility (AD) of organic matter (63.01-96.48%), crude protein (66.98-97.28) and phosphorus (74.44-98.50%) observed in diets incorporated with phytase and protease among enzyme types implied better effi ciency of purifi ed enzymes than sprouted sorghum. This could be attributed to enhanced specifi city in commercial phytase  and protease than in the enzyme cocktail (mix) in sprouted sorghum as reported by [6]. Effi ciency of the enzyme cock tail is thought to have been degraded by proteolytic enzymes such as proteases activated by sprouting in sorghum as previously observed by [22]. was attributed to breakdown of insoluble proteins by protease enzyme.
The apparent digestibility of organic matter and crude protein obtained were higher than 32.10-35.30% and 70.44-77.92% respectively observed by [23], when 32%CP with and without fi sh meal respectively were fed to Labeo rohita.
Apparent digestibility of organic matter was again higher than 66.6 % for fi sh meal and 57.8% for soybean used as ingredients for Nile tilapia extruded diets by [24]. The high digestibility of organic matter in these diets was attributed to low crude fi bre (<4) (

Protein and phosphorus deposition and retention
High crude protein and phosphorus deposition and retention recorded in diets incorporated with protease and phytase particularly with 1250 units was ascribed to better apparent digestibility of these nutrients. Deposited crude protein to a 200-270 fold than the phosphorus deposited could have been related to higher content of crude protein in diets with 1250 protease units than in other diets. As long as energy is not limiting, increasing diet protein is known to increase protein deposition. However, retention of only 1.42-1.96% out of the crude protein deposited (242-705.10gkg-1) compared to 13.29-100% of the little phosphorus deposited indicated that the catfi sh was in a condition of protein balance. This implied that almost a similar amount of crude protein deposited was used up in body/metabolism and not contributing to building body tissue. This was attributed to protein breakdown to overcome effects of stress due to over handling during feacal collection. The high rate of protein breakdown was also refl ected in extremely high ammonia levels, a product of protein metabolism in the culture water (Table 3a). It is in line with observations of [25], in patients with head trauma and [26], and is in agreement with [25]. On the other hand the high phosphorus retention (100%) at 750 phytase and protease inclusion levels than at 1250 units of enzymes (13.29-29.35% implied better utilisation in body tissue.

Conclusions
Incorporating 1250 units of phytase and protease enzymes in catfi sh diets improved digestibility and uptake of crude protein and phosphorus. We therefore recommend use of purifi ed enzymes over sprouted sorghum until its phytase and protease activity essays and feasibility are determined.