Apparent Digestibility of Selected Feed Ingredients for Humpback Grouper, Cromileptes Altivelis

Aquaculture 218 (2003) 529–538 www.elsevier.com/locate/aqua-online

Apparent digestibility of selected feed ingredients for humpback grouper, Cromileptes altivelis

Asda Laininga,*, Rachmansyaha, Taufik Ahmada, Kevin Williamsb

a Research Institute for Coastal Fisheries, Jl. Makmur Dg. Sitakka, Maros, South Sulawesi 90511, Indonesia b CSIRO Marine Research, PO Box 120, Cleveland, Qld. 4163, Australia Received 2 April 2002; received in revised form 5 September 2002; accepted 5 September 2002

Abstract

The dry matter (DM), crude protein (CP) and gross energy (GE) apparent digestibility of nine feed ingredients locally available in South Sulawesi were determined with juvenile (f20 g) humpback grouper. In each of the three 4 4 latin square experiments, humpback grouper were fed a reference diet and three other diets in which test feed ingredients were substituted at rates of either 40% or 30% for animal or plant sources, respectively. Chromic oxide was used as the indigestible marker. Faeces were collected by sedimentation at 3-h intervals. Each experiment entailed four collection periods with each diet being randomly changed between the tanks. The apparent digestibility of GE could not be determined for three feed ingredients as the amount of collected faeces was insufficient for energy analysis. The derived DM, CP and GE apparent digestibility coefficients (ADC; means F S.D.) for the test feed ingredients were, respectively: shrimp head meal—58.5 F 3.33, 78.0 F 1.32 and 63.6 F 0.89; soybean meal—54.8 F 2.72, 67.2 F 1.29 and 51.1 F 0.89; palm oil cake meal—45.3 F 2.37, 80.5 F 1.30 and 40.4 F 3.74; oven-dried blood meal—48.1 F 0.85, 55.2 F 1.35 and not determined (nd); formic acid-fermented blood meal— 67.9 F 1.63, 87.5 F 0.55 and nd; propionic acid-fermented blood meal—61.7 F 2.60, 84.2 F 0.69 and nd; local sardine meal—87.2 F 2.53, 92.5 F 1.40 and 85.2 F 0.90; local fish meal—59.1 F 1.23, 82.4 F 1.99 and 77.2 F 1.91; and rice bran meal—22.2 F 1.52; 59.5 F 1.41 and 44.3 F 0.97. The protein of both plant and animal feed ingredients was well digested by humpback grouper, while the DM and GE of the protein-rich animal feed ingredients were more digestible than the carbohydrate- rich plant feed ingredients. D 2003 Elsevier Science B.V. All rights reserved.

Keywords: Barramundi cod; Polka dot grouper; Blood; Soybean; Local fish meal; Palm oil cake; Acid preservation

* Corresponding author. Tel.: +62-411-371544; fax: +62-411-371545. E-mail address: [email protected] (A. Laining).

1. Introduction

Humpback grouper, also called barramundi cod in Australia, is an important species of grouper in the Asia Pacific region, particularly in Indonesia, because of its high value and demand for export. Premium price of up to US$100/kg wholesale is achieved in the Chinese live-fish markets of Hong Kong and Southern China. At present, digestibility coefficients of common feed ingredients have been reported for only a few species of warmwater marine carnivorous fishes and none for humpback grouper (NRC, 1993). Feedstuff digestibility assessment in fish is an essential prerequisite in determining nutrient requirements, for screening the potential nutritive value of alternative feed ingredients and in the development of nutritionally adequate diets at least cost (Hajen et al., 1993). Apparent digestibility can be determined by direct or indirect methods. However, complete recovery of faeces as required for direct apparent digestibility determination is virtually impossible for aquatic animals. Indirect methods use the difference in the concentration of a digestibility marker in the feed and in the faeces to calculate apparent digestibility. Inherent in this approach is the need to obtain a representative sample of the voided faeces that is unaffected by leaching or urine contamination (Ellis et al., 1987). Various methods for obtaining a representative faecal sample have been suggested including dissection procedures to recover digesta from the intestine (Henken et al., 1985; Brown, 1991) manual stripping (Sullivan and Reigh, 1995), anal suction (Spyridakis et al., 1989) and novel tank and equipment designs to facilitate the collection of faeces while minimizing nutrient lose into the surrounding water (Satoh et al., 1992). Regardless of the method used to determine apparent digestibility, the derived estimate refers to the whole diet that was fed. Substitution procedures as described by Cho et al. (1982) and refined by Forster (1999) enable the apparent digestibility of a single ingredient in a multi-ingredient diet to be determined. This study was aimed at determining the apparent digestibility of several feed ingredients locally available in South Sulawesi for juveniles of humpback grouper.

2. Materials and methods

2.1. Experimental diets

A reference diet and nine test diets in which animal ingredients (shrimp head meal, three types of blood meal and two types of fish meal) substituted at 40% and plant ingredients (soybean meal, palm oil cake meal and rice bran) substituted at 30% of the reference diet were formulated (Table 1). All test feed ingredients were obtained locally from commercial sources other than the blood meals. Fresh blood was collected from a cattle abattoir, dried in a laboratory oven at 60 jC and finely ground in a laboratory grinder (dried blood). Preservation was made by adding either formic acid (formic acid-fermented blood meal) or propionic acid (propionic acid-fermented blood meal) to blood meal at a rate of 3% w/w. Batches for feeding trials (5 kg/treatment) were held in covered polyethylene containers at room temperature (27–28 jC) for 5 days, harvested and oven-dried at 60 jC for 24 h. Chromic oxide was added to the diets at an inclusion rate of A. Laining et al. / Aquaculture 218 (2003) 529–538 531

Table 1 Formulation (g/kg, air dry) of experimental diets Ingredient and international feed number (IFN) Diet Ingredient IFN Reference Test diet 1 Test diet 2 Fish meal 5-01-985 570 342 399 Soybean meal (roasted, full-fat) 5-14-005 80 48 56 Wheat gluten – 100 60 70 Wheat flour 4-05-199 60 36 42 Rice bran 4-03-928 80 48 56 Fish oil 7-08-049 40 24 28 Squid oil – 30 18 21 Vitamin mixa –301821 Mineral mixb –106 7 Test ingredient 1 (animal origin)c – 0 400 0 Test ingredient 2 (plant origin)d – 0 0 300 Chromic oxide – 10 10 10 a At 30 g/kg inclusion level, provided in 1 kg of final diet: retinol, 540 mg; cholecalciferol, 9.125 mg; a- tocopherol, 212.4 mg; menadione, 375 mg; thiamin, 300 mg; riboflavin, 750 mg; pyridoxine, 300 mg; cyanocobalamin, 3.5 mg; ascorbic acid, 4500 mg; folic acid, 150 mg; nicotinic acid, 1800 mg; D-pantothenic acid, 1500 mg; biotin, 3.75 mg; and D/L-methionine, 1500 mg. b At 10 g/kg inclusion level, provided in 1 kg of final diet: Ca, 3.25 g; P, 1.0 g; Fe, 60 mg; Mn, 40 mg; I, 0.75 mg; Cu, 3 mg; and Zn, 37.5 mg. c Animal origin: shrimp head meal (5-04-226) and local mixed-fish meal (5-01-974) manufactured by TAS Coy, Makassar, South Sulawesi; sardine meal (5-02-015): dried sardine from fish landing site and extruded at RICF feed mill; blood meal (5-00-380): oven-dried (60 jC) and ground-fresh bovine blood from local abattoir, Makassar, South Sulawesi. d Plant origin: soybean meal (5-14-005): whole soybean seed supplied by PT Inti Tani and heat-extruded at RICF feed mill; palm oil cake (5-04-487) manufactured by PT Pertani, Luwu, South Sulawesi; rice bran (4-03- 928) supplied by PT Pertani, Sidrap, South Sulawesi.

1% as the digestibility marker. Diets were prepared by blending all ingredients into a moist dough using a planetary mixer. Pellets were formed by extruding the dough twice through a 3-mm diameter die plate fitted to a screw mincer attached to an A200 Hobart dough mixer. The extruded strands were steamed for 10 min, dried overnight at 60 jC and reduced to pellets of approximately 3 mm length, then stored at 15 jC before they were used for the test fish.

2.2. Fish and culture conditions

Hatchery-reared juvenile humpback grouper (Cromileptes altivelis) were transported from the Research Station for Coastal Fisheries, Gondol, Bali, to the floating net cages, Barru Regency, South Sulawesi, and adapted for 1 month before they were used for the experiment. From this batch, 200 juveniles of 20-g initial weight were selected according to size uniformity and freedom of physical defects and distributed equally to four 1 1 1.2-m cages. Cages were suspended from a floating platform in seawater of approximately 16 m depth and strong tidal exchange. Three experiments were carried out sequentially, enabling the apparent digestibility of three individual feed ingredients to be determined each time. Each experiment constituted a 4 4 latin square design in which 532 A. Laining et al. / Aquaculture 218 (2003) 529–538 four diets (one reference diet and three test ingredient diets) were examined over four collection periods. In each experiment, 20 fishes from each cage were transferred to a 200- l cylindro-conical faecal collection tank (conical base sloped at 35j) that was fitted with a faeces collection chamber (Allan et al., 1999). Following a 5-day acclimatization period during which the fishes were fed their prescribed diet twice daily to satiety at 0700 and 1600 h, faeces were collected at 3-h intervals between the morning and afternoon feeds. Collected faeces were gently rinsed with freshwater to remove contaminating seawater, oven-dried (40 jC) and stored in a sealed bottle at 40 jC until analyzed. Faecal collection continued for 5–7 days when it was judged that a sufficient sample had been collected for chemical analysis. After each faecal collection period, diets were reallocated to the collection tanks in accordance with the latin square design, and faecal collection recommenced after a further 5-day acclimatization period. This process was repeated for each of the four collection periods and sequentially for each of the three experiments.

2.3. Digestibility calculation and analyses

A representative sample of feed or oven-dried faeces was homogenized using a mortar and pestle and analyzed essentially by AOAC (1990) procedures: dry matter (DM) by oven-drying at 105 jC for 16 h; ash by ignition in a muffle furnace at 550 jC for 16 h; crude protein (CP) by micro-Kjeldahl analysis and energy by bomb calorimetry. Total lipid was determined following a chloroform: methanol extraction (Bligh and Dyer, 1959) and chromium by using a Shimadzu UV-VIS 2401 PC spectrophotometer after acid digestion. The chemical composition of the test feed ingredients is detailed in Table 2. The apparent digestibility coefficient (ADC) of the diet was derived from the equation:

ADCð%Þ¼100½1 ðMi=Mf ÞðCf =CiÞ; ð1Þ where Ci and Cf are the concentrations (%DM) of nutrient in the ingested diet and faecal output, respectively, and Mi and Mf are the concentrations (%DM) of the marker in the ingested diet and faecal output, respectively. The ADC of a nutrient in an ingredient (ADCIngr) added to the reference diet was calculated by difference, assuming no

Table 2 The proximate composition and gross energy (GE) of the air-dried test feed ingredients Test feed ingredients DM CP C. lipid C. fibre Ash GE (%) (%) (%) (%) (%) (cal/g) Shrimp head meal 93.7 49.8 3.8 2.0 25.1 3257 Soybean meal (roasted, full-fat) 90.7 41.0 18.4 7.2 6.0 4310 Palm oil cake 91.7 10.6 12.7 31.8 4.4 3767 Dried blood meala 96.8 84.3 0.2 6.2 3.6 4844 Formic acid-preserved blood meala 96.5 87.2 0.2 2.7 3.4 5333 Propionic acid-preserved blood meala 96.5 84.7 0.2 4.5 3.1 5101 Local sardine meal 94.1 65.2 4.8 1.6 12.2 4146 Local mixed-fish meal 93.0 44.7 9.9 3.4 24.7 4242 Rice bran 92.3 13.8 11.9 14.3 10.5 3141 a See Section 2.2 for description of processing of the blood meals. A. Laining et al. / Aquaculture 218 (2003) 529–538 533 associative effects between the added ingredient and the reference diet. In calculating the apparent digestibility of the test ingredient, the procedure of Forster (1999) used the nutrient contribution of the test ingredient rather than its weight contribution.

ADCIngrð%Þ¼½ADNCom fADNRef ð1 SRNutÞg=SRNut; ð2Þ where ADNCom is the ADC (%) of the nutrient in the combined diet, ADNRef is the ADC (%) of the nutrient in the reference diet and SRNut is the substitution rate (as decimal) for the nutrient in question. Calculation of SRNut is:

SRNut ¼ðNTestSRWtÞ=½ðNTestSRWtÞþfNRef ð1 SRWtÞg; ð3Þ where NTest is the concentration (%) of the nutrient in the test ingredient, NRef is the concentration (%) of the nutrient in the reference diet and SRWt is the substitution rate of the ingredient in the reference diet on a weight basis (in decimal; either 0.4 for animal feed ingredient or 0.3 for plant feed ingredient). ADC data for the diets and those derived for the test feed ingredients in each experiment were subjected to a least squares analysis of variance for a latin square design, isolating effects due to column (fish) and row (collection period). Differences in ADC among diets were examined a posteriorily using Fischer’s protected t-test (Snedecor and Cochran, 1989) at a 5% level of probability.

3. Results

The apparent digestibility of diets and those derived for the test feed ingredients in experiment 1 are given in Table 3. DM digestibility was highest for the shrimp head

Table 3 The dry matter (DM), crude protein (CP) and gross energy (GE) apparent digestibility coefficients (%) of diets and of substituted test feed ingredients examined in experiment 1 Diet and ingredient designation Dry matter Crude protein Gross energy Diets Reference 56.9a 81.6a 81.5a Shrimp head diet 61.8a 80.1a 76.6b Soybean diet 56.3b 77.7b 70.9c Palm oil cake diet 53.6c 81.5a 69.2d F S.E.M. 0.77 0.44 0.45

Test feed ingredients Shrimp head meal 58.5 F 3.33a 78.0 F 1.32b 63.6 F 0.89a Soybean meal (full-fat) 54.8 F 2.72b 67.2 F 1.29c 51.1 F 0.89b Palm oil cake meal 45.3 F 2.37c 80.5 F 1.30a 40.4 F 3.74c Each coefficient is the mean of four replicates. a, b, c, d: Within diets and test feed ingredients comparisons; means in a column without a common letter differ ( P < 0.05). 534 A. Laining et al. / Aquaculture 218 (2003) 529–538

Table 4 The dry matter (DM), crude protein (CP) and gross energy (GE) apparent digestibility coefficients (%) of diets and of substituted test feed ingredients examined in experiment 2 Diet and ingredient designation Dry matter Crude protein Gross energy Diets Reference 56.3b 86.2a nd Dried blood diet 53.0c 72.3b nd Formic blood diet 59.0a 86.5a nd Propionic blood diet 60.1a 85.3a nd F S.E.M. 0.66 1.75

Test feed ingredients Dried blood meal 48.1 F 0.85c 55.2 F 1.35c nd Formic blood meal 67.9 F 1.63a 87.5 F 0.55a nd Propionic blood meal 61.7 F 2.60b 84.2 F 0.69b nd Each coefficient is the mean of four replicates. a, b, c: Within diets and test feed ingredients comparisons; means in a column without a common letter differ ( P < 0.05). nd: Not determined as insufficient faecal sample for energy analysis. substituted diet and significantly higher than for substituted diets containing soybean and palm oil cake, with the latter being significantly lower than the former. CP digestibility was similar for the shrimp head and palm oil cake substituted diets and significantly higher than for the soybean substituted diet. For gross energy (GE) digestibility, each of the three substituted diets differed significantly from each other, with shrimp head being the highest and palm oil cake the lowest. The derived apparent digestibility of each test feed ingredient differed significantly from each other: DM and GE digestibility were highest for shrimp head meal and lowest for palm oil cake meal; and CP digestibility was highest for palm oil cake meal and lowest for full-fat soybean meal. The apparent

Table 5 The dry matter (DM), crude protein (CP) and gross energy (GE) apparent digestibility coefficients (%) of diets and of substituted test feed ingredients examined in experiment 3 Diet and ingredient designation Dry matter Crude protein Gross energy Diets Reference 58.7c 89.7b 86.9a Local sardine diet 70.4a 91.3a 86.2a Local mixed-fish meal diet 65.9b 86.8c 83.3b Rice bran diet 48.3d 86.1c 71.9c F S.E.M. 0.45 0.30 0.45

Test feed ingredients Local sardine meal 87.2 F 2.53a 92.5 F 1.40a 85.2 F 0.90a Local mixed-fish meal 59.1 F 1.23b 82.4 F 1.99b 77.2 F 1.91b Rice bran meal 22.2 F 1.52c 59.5 F 1.41c 44.3 F 0.97c Each coefficient is the mean of four replicates. a, b, c, d: Within diets and test feed ingredients comparisons; means in a column without a common letter differ ( P < 0.05). A. Laining et al. / Aquaculture 218 (2003) 529–538 535 digestibility of CP was generally high for all ingredients (67–81%), while that of GE was low (40–64%). In experiment 2, it was not possible to determine the apparent digestibility of GE, as the amount of faeces collected was insufficient for energy analysis. Substitution of the reference diet with dried blood meal significantly reduced the apparent digestibility of both DM and CP (Table 4). The DM and CP digestibility of diets containing the blood meal subjected to propionic acid or formic acid treatment was better than for the diet containing dried blood meal. The derived DM and CP digestibility for each of the test ingredients differed significantly from each other, being highest for formic acid-treated blood and lowest for dried blood. In experiment 3, substitution of the reference diet with local sardine meal significantly increased the apparent digestibility of DM and CP but not that of GE. Including the local mixed-fish meal also increased the apparent digestibility of DM, but CP and GE digestibility were both significantly lowered. In the case of the rice bran, the derived DM, CP and GE apparent digestibility was very low and significantly inferior to all other test ingredients (Table 5). The derived DM, CP and GE apparent digestibility of the local sardine meal was significantly higher than all other test ingredients.

4. Discussion

Humpback grouper showed generally a high capacity to digest protein. Excluding the poorly digested oven-dried blood meal and rice bran meal, the apparent digestibility of plant protein (67.2–80.5%) was almost as good as that of animal protein (78.0–92.5%). However, the apparent digestibility of DM was generally poor, especially for plant feed ingredients where values ranged from 22.2% for rice bran meal to 54.8% for soybean meal. By comparison, the DM apparent digestibility of animal feed ingredients was higher, especially for the local sardine meal (87.2%). The local mixed-fish meal and the shrimp head meal were far less digestible with a DM apparent digestibility of only 59.1% and 58.5%, respectively, and lower than either of the two fermented blood meals (61.7% and 67.9%). The apparent digestibility of GE was comparatively high for the two fish meals with the local sardine meal being significantly more digestible than the local mixed-fish meal (85.2% and 77.7%, respectively). Although, the apparent energy digestibility of the shrimp head meal was low (63.6%), it appears to be higher than for each of the three plant feed ingredients—full-fat soybean meal (51.1%), palm oil cake meal (40.4%) and rice bran (44.3%). The low-energy digestibility of these plant feed ingredients can be attributed to their high-carbohydrate content and poor digestibility by carnivorous fish (Lupatsch et al., 1997). We have not been able to find any other grouper studies with which our present apparent digestibility values can be compared. Other warmwater carnivorous marine fish such as Asian sea bass and red drum appear to have a higher capacity to digest plant ingredients than humpback grouper. In Asian sea bass, the apparent energy digestibility of plant protein meals ranged from 98.8% for wheat gluten to 56.1% for canola meal, with values of 69.4% and 75.9% for solvent extracted soybean meal and microwave-roasted full-fat soybean meal, respectively (McMeniman and Williams, personal communication). 536 A. Laining et al. / Aquaculture 218 (2003) 529–538

In red drum, the apparent energy digestibility of wheat, dehulled soybean meal and cottonseed meal was reported by Gaylord and Gatlin (1996) to be 61.6%, 63.3% and 70.4%, respectively, which compare with very low values of 33.7%, 37.8%, 22.1% and 12.0% that were reported by McGoogan and Reigh (1996) for wheat middling, soybean meal, cottonseed meal and rice bran, respectively. Although direct comparisons of derived digestibility values between experiments (and between different species) are fraught with danger, the overwhelming evidence is that plant feed ingredients are less digestible for humpback grouper than for other warmwater marine carnivorous fish such as Asian sea bass or red drum. Generally, freshwater and warmwater fish appear to digest carbohydrates more effectively than marine fish and cold water fish (Wilson, 1994). However, all three species appear capable of efficiently digesting the protein of both plant and animal sources with digestibility coefficients typically higher than 70%. The observation in the present humpback grouper work that preservation of blood by treatment with either propionic acid or formic acid significantly increased the apparent digestibility of DM and CP was an interesting finding. Fermentation of fish and fishery wastes with organic acids, with or without addition of lactic acid bacteria or filamentous fungi, is a traditional way of preserving or enhancing their nutritional value as animal feed ingredients (Jones, 1975; Hassan and Heath, 1986). Although fermentation is most usually applied to fishery wastes, the practice has been successfully used to increase the apparent protein digestibility and overall nutritive value of sesame seed in diets for rohu fingerlings (Mukhopadhyay and Ray, 1999). Moreover, acidification of diets with formic acid was reported by Vielma and Lall (1997) to significantly increase the apparent digestibility of phosphorus in rainbow trout. As shown in this study, preservation can also be successfully used to improve the apparent digestibility of blood meal. We have no satisfactory explanation as to why preserving with formic acid was more effective than with propionic acid in improving the apparent digestibility of the blood. The analyzed higher CP content of the formic acid-preserved blood suggests that the preservation process resulted in a net gain in protein (possibly through hydrolysis and loss of other organic constituents), which was highly digestible. This is supported by amino acid analysis that showed the total amino-N content to be highest for the formic acid-preserved blood (55.0%), followed by the propionic acid-preserved blood (45.6%) and lowest for the oven-dried blood (31.9%). In conclusion, the study has shown that humpback grouper are able to efficiently digest the protein of both plant and animal feed ingredients. However, they have a very limited capacity to digest carbohydrate-rich products such as many plant feedstuffs and shrimp head meal, which additionally has a high-ash content (25.1%). Because of this poor digestive capacity, the potential to use carbohydrate-rich meals as even partial substitutes for fish meal in diets for humpback grouper appears to be quite low. Blood meal was better digested than plant meals and thus has a greater potential to be used as a dietary replacement of fish meal in humpback grouper diets and particularly if its nutritive value is enhanced through preservation. The two types of local sources of fish meal were well digested by humpback grouper and also have a great potential to substitute fish meal, although their utilization has relatively high competition with human consumption. A. Laining et al. / Aquaculture 218 (2003) 529–538 537

Acknowledgements

This experiment was supported by a grant from the Australian Centre for International Agricultural Research as part of an international collaborative project to develop improved hatchery and grow-out technology for grouper aquaculture (FIS 97/73).

References

Allan, G.L., Rowland, S.J., Parkinson, S., Stone, D.A.J., Jantrarotai, W., 1999. Nutrient digestibility for juvenile silver perch Bidyanus bidyanus: development of methods. Aquaculture 170, 131–145. AOAC, 1990. Official Methods of Analysis, 15th ed. Association of Official Analytical Chemists, Washington, DC, USA. 1094 pp. Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37, 911–917. Brown, P.B., 1991. Comparison of fecal collection methods for determining phosphorus absorption in rainbow trout. Fourth International Symposium Fish Nutrition and Feed, Biarritz (France), 24–27 June, Institut National de la Recherche Agronomique, Paris, pp. 443–447. Cho, C.Y., Slinger, S.J., Bayley, H.S., 1982. Bioenergetic of salmonid fishes: energy intake, expenditure and productivity. Comp. Biochem. Physiol. 73B, 25–41. Ellis, R.W., Long, J., Leitner, L., Parsons, J., 1987. Estimation of crude protein, energy, and amino acid digestibilities in freshwater prawns (Macrobrachium rosenbergii) and crayfish (Pacifastacus leniusculus) with a fecal collection system. Prog. Fish-Cult. 49, 303–305. Forster, I., 1999. A note on the method of calculating digestibility coefficients of nutrients provided by single ingredients to feeds of aquatic animals. Aquac. Nutr. 5, 143–145. Gaylord, T.G., Gatlin III, D.M., 1996. Determination of digestibility coefficients of various feedstuffs for red drum (Sciaenops ocellatus). Aquaculture 139, 303–314. Hajen, W.E., Beames, R.M., Higgs, D.A., Dosanjh, B.S., 1993. Digestibility of various feedstuffs by post- juvenile chinook salmon (Oncorhynchus tshawytscha) in sea water: 1. Validation of technique. Aquaculture 112, 321–332. Hassan, T.E., Heath, J.L., 1986. Biological fermentation of fish waste for potential use in animal and poultry feeds. Agric. Wastes 15, 1–15. Henken, A.M., Kleingeld, D.W., Tijssen, P.A.T., 1985. The effect of feeding level on apparent digestibility of dietary dry matter, crude protein and gross energy in the African catfish, Clarias gariepinus (Burchell 1822). Aquaculture 51, 1–11. Jones, I.D., 1975. Effect of processing by fermentation of nutrients. In: Harris, R.S., Karmas, E. (Eds.), Nutri- tional Evaluation of Food Processing. Avi Publication Coy, Westport, CN, USA, pp. 324–354. Lupatsch, I., Kissil, G.W., Sklan, D., Pfeffer, E., 1997. Apparent digestibility coefficients of feed ingredients and their predictability in compound diets for gilthead seabream, Sparus aurata L. Aquac. Nutr. 3, 81–89. McGoogan, B.B., Reigh, R.C., 1996. Apparent digestibility of selected ingredients in red drum (Sciaenops ocellatus) diets. Aquaculture 141, 233–244. Mukhopadhyay, N., Ray, A.K., 1999. Effect of fermentation on the nutritive value of sesame seed meal in the diets for rohu, Labeo rohita (Hamilton), fingerlings. Aquac. Nutr. 5, 229–236. NRC (National Research Council), 1993. Nutrient Requirements of Fish. National Academy Press, Washington, DC. 114 pp. Satoh, S., Cho, C.Y., Watanabe, T., 1992. Effect of fecal retrieval timing on digestibility of nutrients in rainbow trout diet with the Guelph and TUF faeces collection system. Nippon Suisan Gakkaishi 58, 1123–1127. Snedecor, G.W., Cochran, W.G., 1989. Statistical Methods, 8th ed. Iowa State Univ. Press, Ames, IA, USA. 503 pp. Spyridakis, P., Metailler, R., Gabaudan, J., Riaza, A., 1989. Studies on nutrient digestibility in European sea bass (Dicentrarchus labrax): I. Methodological aspects concerning faeces collection. Aquaculture 77, 61–70. 538 A. Laining et al. / Aquaculture 218 (2003) 529–538

Sullivan, J.A., Reigh, R.C., 1995. Apparent digestibility of selected feedstuffs in diets for hybrid striped bass (Morone saxatilis Morone chrysops). Aquaculture 138, 313–322. Vielma, J., Lall, S.P., 1997. Dietary formic acid enhances apparent digestibility of minerals in rainbow trout, Oncorhynchus mykiss (Walbaum). Aquac. Nutr. 3, 265–268. Wilson, R.P., 1994. Review: utilization of dietary carbohydrate by fish. Aquaculture 124, 67–80.