26 Comparative studies on the nutrition of two species of abalone, Haliotis tuberculata L. and Haliotis discus hannai Ino. IV. Optimum dietary protein level for growth

The weight gain, protein gain, soft body to shell ratio, and carcass levels of protein and lipid of both abalone species were significantly ANOVA, P < 0.05 affected by the dietary protei

Aquaculture

Comparative studies on the nutrition of two species

level for growth Kangsen Mai a**, John P Mercer ‘, John Donlon b

a She&h Research Laboratory, University College Galway, Galway, Ireland

b Deparmenr of Biochemisrry, University College Galway, Galway, Ireland

Accepted 30 May 1995

Abstract

A 100 day growth experiment was conducted to identify the optimum dietary protein level for the juveniles of two species of abalone, Haliotis tuberculata and Haliotis discus hannai A mixture of vitamin-free casein and gelatin (4.34: 1) supplemented with crystalline amino acids was used as the protein source to simulate the amino acid profile of abalone body Eight purified diets were formulated

to provide graded protein levels ranging approximately from 0 to 50% The weight gain, protein gain, soft body to shell ratio, and carcass levels of protein and lipid of both abalone species were significantly (ANOVA, P < 0.05) affected by the dietary protein level The protein requirements of these abalone were evaluated from weight gain and protein gain respectively, by using the second-order polynomial regression analysis On the basis of weight gain, the optimum protein levels were estimated to be 22.3-32.3%, and 23.3-35.6% for H tuberculata and H discus hannai, respectively According to the protein gain, the statistical analysis indicated that the optimum ranges of protein requirements were 24.0-34.5% and 25.2-36.6% for H tuberculata and H discus hannai, respectively Based on these results, about 35% dietary protein from good quality sources is recommended for the maximum growth of both abalone species; and, if dietary protein is reduced from 35 to 25%, the growth of these abalone may be depressed with 5% likelihood

Keywords: Hakotis ruberculata; Halioris discus hnnai; Comparative nutrition; Protein requirements, molluscs

1 Introduction

Animals do not have an absolute protein requirement but require a well-balanced mixture

of essential and non-essential amino acids (NRC, 1983; Wilson, 1989) The most common

* Corresponding author at: College of Fisheries, Ocean University of Qingdao, Peoples Republic of China Tel

166 K Mai et al /Aquaculture 136 (1995) 165-180

and economical source of amino acid mixture is from natural proteins in feedstuffs Infor- mation on the amino acid requirements of an animal provides a basis to evaluate the nutritional value of protein sources and then to select proper protein sources for formulating feeds for the animal Because it is the principal diet component for animal growth, and has the highest cost consideration for commercial feeds, protein has been given priority in nutritional requirement studies (Lim et al., 1979) The minimum amount of dietary protein needed to supply adequate amino acids and produce maximum growth has been determined with semi-purified and purified test diets in about 30 species of fish and crustaceans (NRC, 1983; Tacon and Cowey, 1985; Wilson, 1989) Molluscan nutrition lags far behind that of fish and crustaceans and much less is known of the protein requirements of molluscs largely due to their lower commercial importance in aquaculture (Carefoot, 1982) There have been some publications concerning the protein nutrition of molluscs (e.g Duncan et al., 1985; Nell, 1985; Uki et al., 1985; Utting, 1986; Hawkins et al., 1989; Kreeger and Langdon, 1992; Kreeger, 1993; Mercer et al., 1993) From the viewpoint of quantitative requirements, however, there have been only four reports on the protein requirements of abalone, H discus

(Ogino and Kato, 1964)) H discus hunnai (Uki et al., 1986)) H kumtschutkuna (Taylor,

1992) and H midue (Britz et al., 1994), It has been demonstrated that intensive culture of abalone using nutritionally complete diets is feasible; however, information on the nutrient requirements of this animal is far from sufficient to support industrial scale feed manufacture Investigations into the protein requirement of abalone have used either white fish meal (Ogino and Kato, 1964; Uki et al., 1986; Britz et al., 1994) or casein (Uki et al., 1986; Taylor, 1992) as the sole protein source in the test diets Following the results of feeding trials, casein has been considered as the best protein source for abalone (Uki et al., 1985)

At present, no information is available on the quantitative requirements of abalone for essential amino acids (EAA), however, the analyses of amino acid profiles of abalone body indicate that casein is probably limiting in some EAA, especially arginine (Mai et al., 1994) Therefore, the protein requirement of abalone could be overestimated if casein is used as the sole protein source in experimental diets The protein requirements of fish are usually estimated by feeding a balanced diet containing graded levels of high quality protein, which is generally a casein/gelatin mixture supplemented with crystalline amino acids to simulate the amino acid profiles of either fish body or whole hen’s egg protein (Tacon and Cowey, 1985) Hence, this study was designed to identify the optimum level of dietary protein for the growth of the two commercially important abalone species, H tubercuhtu and H discus hunnui by employing the protein source, casein/gelatin mixture supplemented with crystalline amino acids to simulate the EAA pattern of abalone body

2 Materials and methods

2.1 Experimental diets

Formulations of the experimental diets and their proximate composition are shown in Table 1 On the basis of essential amino acid ratios (A/E ratios) (Arai, 198 1) , the EAA pattern in the soft-body of these abalone (Mai et al., 1994)) expressed as mean A/E ratios, was used as a reference The mixture of casein and gelatin (4.34: 1) was supplemented with

1

168 K Mai et al /Aquaculture 136 (1995) 165-180

a crystalline amino acid mix including arginine, threonine and methionine to simulate the reference protein The A/E ratios and the degree of similarity (DS) in EAA pattern (Mai

et al., 1994) of the protein sources and the experimental diets to that of the reference are also given in Table 1 A mixture of corn oil and menhaden fish oil ( 1: 1) was used as the lipid source Dextrin, the major carbohydrate source, was used to adjust the protein level Eight experimental diets were formulated from the purified ingredients to provide graded protein levels ranging approximately from 0 to 50% Gross energy of experimental diets correspondingly ranged from 15.8 to 19.7 kJ g-’ as determined by bomb calorimetry, and energy/protein ratios (E/P ratios) were from 39.1 to 5653.6 kJ gg’ protein

Procedures for food preparation were modified from the method described by Uki and Watanabe ( 1992) Casein, gelatin and some minerals that were in the form of small grains were ground individually using a Pascal Mill and then passed through a mesh with 125 pm pore size After adding water (about 120%, w/w) to the mechanically mixed ingredients containing 18% of sodium alginate, a paste was made by using an electronic mixer The paste was shaped into 0.5 mm thick sheets, which were cut into 1 cm* flakes The flakes were dipped in an aqueous solution of CaCl, (5%, w/v) for one minute The surplus solution was drained naturally, then the flakes were sealed in a sample bag and stored at

- 20°C until use

2.2 Animal rearing

To maintain the water temperature, a re-circulating system was used This system com- prised a glass fibre reservoir tank (3 m3), a high position tank ( 150 1) and two flush trays (220 X 120 X 30 cm) Each flush tray held 24 rearing units constructed from 10 1 PVC flowerpots with covers, and the bases replaced with 1.0 mm mesh Each rearing unit was stocked with 25 abalone juveniles The seawater in the reservoir tank was pumped into the high position tank, where it was aerated, then delivered to each rearing unit through 1 O cm (i.d.) PVC tubes The flow rate of water through each unit was about 1 1 min- ‘ The water depth in the flush trays was maintained at about 25 cm and the excess water was returned

to the reservoir through the outlets Water temperature during the experiment was maintained

at 13-15°C Similar size juveniles of H discus hannai (378.3 f 16.7 mg) and H tubercuzata

( 182.7 + 7.8 mg) were selected from the hatchery produced population, then assigned to the rearing system using a completely randomised design with eight treatments and three replicates per treatment At the same time, 100 juveniles of each species were randomly sampled from the same cohort and stored at - 20°C until subsequent analysis for carcass composition Abalone were fed the appropriate diet every third day at a satiation level with

a little leftover The whole system was thoroughly cleaned just before each feeding and refilled with pre-heated and filtered ( 10 pm cartridge) seawater Under such management, good water quality in the system can be maintained (K Mai, unpublished observation) The feeding experiment was run for 100 days

2.3 Sample collection and analysis

At the termination of the experiment, 15 abalone from each replicate were frozen for subsequent chemical analysis Growth is expressed as mean weight gain and protein gain

K Mai et al /Aquaculture I36 (1995) 165-180 169

(mg per abalone) The initial and final samples were slightly thawed, and shell and soft- body were separated The soft-body to shell ratio (SB/S ratio, w/w) was computed to provide an index of nutritional status for abalone

An aliquot of soft-body tissue from each sample was lyophilised to a constant weight to determine the moisture content The rest of each soft-body sample was homogenised with

an equivalent volume of distilled water for 10 min in an Omni-mixer (Sorvall, New Town, CT), then the homogenate was freeze-dried and ground into fine powder for analyses of protein and total lipid Protein was estimated by a modification of the Lowry procedure (Hartree, 1972) with bovine serum albumin as the calibration standard Extraction of lipid was carried out according to the method of Bligh and Dyer ( 1959) and the lipid levels were determined gravimetrically The levels of protein and lipid in the experimental diets were analysed as described above The gross energy in the diets was estimated with a Gallenkamp ballistic bomb calorimeter The mean gains in body weight and protein were calculated as follows:

MWG (mg/abalone) = Wt- Wi

MPG(mg/abalone)=SBt.(l-Mr).Pt-SBE’.(l-Mi) Pi

where MWG is mean weight gain; Wi, Wt is initial or final mean body weight (mg); MPG

is mean protein gain; SE, SBt is initial or final soft-body weight (mg); SBi, t= Ri, t Wi, t/

( 1 + Ri, t); Ri, Rt is initial or final soft-body to shell ratio (SB/S ratio); Mi, Mt is initial or

final moisture level in soft body (%); Pi, Pt is initial or final protein level in soft body (%)

2.4 Statistical analysis

All percentage data were square-root arcsine transformed prior to analysis Data from each treatment were subjected to one way ANOVA When overall differences were signif- icant at less than 5%, Tukey test was used to compare the mean values between individual treatments Protein requirements of the juvenile abalone were estimated from weight gain and protein gain using both the broken-line model (Robbins et al., 1979) and the second- order polynomial regression analysis model (Lovell, 1989) All statistics were calculated using Systat@ package (SYSTAT, 1992)

3 Results

3.1 Survival and weight gain

The data of survival and weight gain of abalone fed the experimental diets are shown in Table 2 During the 100 day experimental period the mean survival, ranging from 92.0 to

98.7% for H discus hannai and from 82.7 to 96.0% for H tuberculata, was generally high

and not related to dietary protein level, even though both species fed the approximately protein-free diet (Diet 1) showed relatively lower survival The weight gain of both abalone species was significantly affected by dietary protein levels (ANOVA, P < 0.001) The

progressive increase in mean weight gain reached a maximum value at a protein level of

2 of

tuberculata in the

y = 6.902 +30.482x - 0.428x"Z RV = 0.902 0 H dircrcs hami

y = 35.569 + 19.712x - 0.305r'Z R"2 = 0.919 A H.m!xrculata

Xmax=32.3% (H I)

I I I

Dietary protein (%)

but within the 95% confidence interval; X,,,, a protein level required for Y,,,,,; X,, a protein level associated with Y,

25% for H discus hannai and 40% for H tuberculata The results of Tukey test, however,

showed that weight gains of both species at or beyond 20% dietary protein concentration were not significantly different; but 50% dietary protein appeared to significantly depress

gain of H tuberculata (Table 2) On the basis of weight gain, the results estimated by the broken-line model showed that the protein requirements of H discus hannai and H tuber-

culata were 17.3% and 17.4%, respectively (Table 4) However, the second-order poly- nomial regression analysis revealed that the ranges of dietary protein levels that produced the maximum weight gains (Y,,) and lower gains within the 95% confidence limits of Y,,,,, ( Y, ) were 23.3-35.6% (X,-X,,,) for H discus hannai and 22.3-32.3% for H tub-

erculata (Fig 1, Table 4)

The protein requirements of both abalone species seem to be similar when they were estimated with the same analysis model (Table 4) One remarkable difference in growth performance between these abalone was that feeding on Diet 1 (approximately 0.3% pro-

tein), H discus hannai exhibited a negative weight gain, while H tuberculata showed a

positive one (Table 2 and Fig 1)

3.2 Carcass composition and protein gain

Except for moisture content, the body composition was significantly affected by the dietary treatments (Table 3) Soft body to shell ratios (SB/S ratio, w/w) could be regarded

3 of

Table 4

Comparison of the protein requirements of abalone estimated by the broken-line model and the quadratic regression analysis model

‘Weight gain

2Protein gain

‘Numbers in parentheses are residual mean squares

as a good condition index of nutritional status for abalone because it showed a very similar pattern to that of weight gain Statistically, SB/S ratio did not reflect the effect of dietary protein levels on growth performance as sensitively as weight gain The lowest SB/S ratio was always observed for the abalone fed diet 1 (0.3% protein) and higher SB/S values were observed for H discus hannai fed 20-35% protein and for H tuberculata fed 20-40% protein Whole soft-body lipid levels, ranging from 6.6 to 8.0% for H discus hannai and from 6.8 to 8.6% for H tuberculata, steadily decreased with increasing dietary protein Conversely, carcass protein levels, ranging from 40.6 to 49.0% for H discus hannai and from 41.5 to 51.9% for H tuberculata, showed a linear increase when dietary protein increased from 0.3 to 50.2%

Based on F values of ANOVA (Table 2 and Table 3), protein gains appeared to be more sensitively affected by dietary protein levels than did weight gains Protein gains ranged from - 3.3 to 40.9 mg per abalone for H discus hannai and from 1.6 to 35.4 mg per abalone for H tuberculata Similarly to weight gains, maximum protein gain was recorded for H discus hannai fed 25% dietary protein and for H tuberculata fed 40% dietary protein Based on protein gain, the protein requirements assessed by the broken-line model were 21.6% and 27.3% for H discus hannai and H tuberculata, respectively (Table 4) The second-order polynomial regression demonstrated that the minimum dietary protein levels giving the maximum protein gains were 36.6% for H discus hannai and 34.5% for H tuberculata, and the lower dietary protein levels, i.e 25.2% and 24.0% for H discus hannai and H tuberculata respectively, could reduce their protein gains with 5% likelihood

4 Discussion

Estimates of quantitative nutrient requirements are influenced not only by the criteria used but by the statistical methods chosen to evaluate criterion response to differing dietary nutrient concentrations (Zeitoun et al., 1976) In quantifying protein requirements of fish, the most common criterion used by researchers is growth (Tacon and Cowey, 1985; Lovell, 1989) Some other criteria, such as nitrogen balance, protein efficiency ratio (PER), net

y = 2.6139 + 2.343 -0.032~A2 ~"2 =0.912 0 H.discwhm~i

y = 0.171 + 1.865x 0.027x”Z RY = 0.924 A f/ ~,&rcu,aara

Xmax=36.6% (H d)

Dietary protein (%I

Fig 2 Relationship between protein gain and dietary protein level for H discus hctnnui and H tuberdata as

but within 95% confidence interval; X,,,,, a protein level required for Y,,,,,; X, , a protein level associated with Y,

protein utilisation (NPU) and feed conversion efficiency (FCE) , are sometimes employed The reliability of the results based on the latter criteria depends strongly upon the accuracy

in determining the feed consumed The aquatic environment makes it difficult to obtain accurate data on feed intake of aquatic animals Therefore, growth in terms of weight gain

is generally sensitive to the dietary protein content, and sometimes shows a more clear-cut protein requirement level than PER, FCE, etc (Moore et al., 1988) Because abalone eat very slowly, using radula to scrape off particles of feed and send them into the mouth rather than swallowing feed pellets directly as most fish do, great difficulties are encountered in obtaining reliable data on their feed intake, particularly on artificial feeds As a result, the protein requirements of the juveniles of H discus hannai and H tuberculata were evaluated

by their growth data in the present study Weight gain is a reliable indicator for growth as long as the experimental variable is not expected to affect composition of gain in the animal (Lovell, 1989) In this study, dietary protein significantly affected the carcass composition

of both abalone species (Table 3) Hence, it is generally believed that protein gain could

be a more reliable measure for true growth than weight gain, and thus more reliable estimates

of protein requirements may be yielded from the protein gain data In the current study, however, protein gain was still highly correlated with weight gain in both abalone species

(r = 0.99)) and very similar estimates of protein requirements were obtained from these two types of data by the second-order polynomial regression analysis (Fig 1 and Fig 2)