Aquaculture 184 Ž2000. 115–132 www.elsevier.nlrlocateraqua-online

Digestibility of macronutrients, energy and amino acids, absorption of elements and absence of intestinal enteritis in Atlantic salmon, Salmo salar, fed diets with wheat gluten T. Storebakken a,) , K.D. Shearer b, G. Baeverfjord a , ˚ ˚ a, T. Scott b, A. De Laporte d B.G. Nielsen c , T. Asgard a

b

AKVAFORSK (Institute of Aquaculture Research A.S.), Sunndalsøra N-6600, Norway Northwest Fisheries Science Center, NOAAr NMFS, 2725 Montlake BouleÕard E., Seattle, WA 98112, USA c BioMar, Myre N-8430, Norway d Amycor, Amylum Group, Burchstraat 10, Aalst B-9300, Belgium Accepted 3 September 1999

Abstract Apparent digestibility coefficients ŽADCs. of macronutrients and energy, and apparent absorption coefficients ŽAACs. of amino acids and elements were assessed in an experiment with 0.9 kg Atlantic salmon reared in saltwater tanks. Duplicate groups of fish were fed five diets, where 0, 6.25, 12.5, 25 and 50% of crude protein ŽCP. from fish meal ŽFM. was replaced with CP from wheat gluten ŽWG.. In Experiment 2, triplicate groups of 0.9 kg salmon were fed a FM diet, a diet with 15% of CP from FM replaced with extracted, toasted soybean meal ŽSBM., and a diet with 35% of CP from WG, for 18 weeks. Experiment 2 was designed to determine whether WG caused pathological changes in the intestinal epithelium, and if a diet with 35% of CP from WG could support rapid growth. There was a trend toward increased ADC of fat and energy in the diets with WG, and the diet with 25% WG was ranked significantly higher than the FM control. The ADC of CP and AACs of all amino acids except alanine and lysine increased significantly with increasing proportion of dietary protein from WG, and the results indicate that absorption of individual amino acids from WG was between 94% and 100%. WG is low in lysine, but the results indicate that the requirement for lysine was nearly met and the requirement for other essential amino acids was met even with the highest WG inclusion level, due to high dietary protein concentration and the supplementary amino acid profile of FM. There was no reduced absorption of Ca, P or Mg in the

)

Corresponding author. Tel.: q47-71-69-53-14; fax: q47-71-69-53-01; e-mail: [email protected] 0044-8486r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 0 4 4 - 8 4 8 6 Ž 9 9 . 0 0 3 1 6 - 6

116

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

salmon fed diets with WG. The absorption of Zn was higher in all the diets with WG than in the FM diet. The fish doubled their weight during Experiment 2, and there were no significant differences in growth among dietary treatments in salmon fed the WG, FM or SBM diets. No indication of intestinal pathology was seen in the salmon fed the FM or WG diets, while 60% of the examined fish fed SBM had SBM-induced changes in the mucosa of the posterior intestine. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Wheat gluten–soybean meal–LT fish meal; Digestibility–fat–protein–energy; Absorption–amino acid–elements–minerals–phosphorus; Soybean-induced enteritis; Atlantic salmon — Salmo salar

1. Introduction Wheat gluten ŽWG. may act both as a source of protein and a pellet binder. Traditionally, starches are used to provide the necessary binding of extruded diets. However, Atlantic salmon has a limited ability to hydrolyse starch in the intestine, and to regulate the blood glucose concentration when the level of digestible carbohydrate is high ŽHemre et al., 1995a,b.. Thus, the amount of starch in diets for salmon is kept low. The use of indigestible hydrocolloid binders used in moist feeds results in reduced digestibility of both protein and fat ŽStorebakken, 1985; Storebakken and Austreng, 1987; Fagbenro and Jauncey, 1995; Yamamoto and Akiyama, 1995.. Such problems are not seen with WG, since it is digestible ŽFagbenro and Jauncey, 1995; Yamamoto and Akiyama, 1995.. Several experiments have shown that WG can successfully replace a large proportion of the fish meal ŽFM. in diets for rainbow trout, provided that the diets are supplemented with lysine, the first limiting amino acid in gluten ŽPfeffer et al., 1992; Davies et al., 1997; Schumacher and Wax, 1997.. The literature is contradictory with respect to the need for supplementation with arginine ŽPfeffer et al., 1992; Davies et al., 1997., which is the second limiting amino acid in gluten. There does not seem to be any need to supplement diets for rainbow trout with threonine, which is the proposed third limiting amino acid ŽPfeffer et al., 1992.. WG is a highly digestible source of protein, as demonstrated by Pfeffer et al. Ž1995. who found an apparent digestibility of 99% for crude protein ŽCP. when feeding a diet with 92.7% gluten and 1.45% lysine to rainbow trout. The digestibility of protein from gluten was higher than those obtained for various poultry by-products, or hydrothermally treated soy- or field- beans. Similarly, Sugiura et al. Ž1997. found that WG had the numerically highest value with respect to apparent digestibility of dry matter and protein, when compared with several FMs, poultry by-products, and vegetable protein sources, both in coho salmon and rainbow trout. Skonberg et al. Ž1998. have shown that incorporation of WG into a diet for rainbow trout did not adversely affect flavour or pigmentation of the fillets. Sugiura et al. Ž1997. also studied apparent absorption of various elements in WG meal, and found that it was ranked among the best of the ingredients tested both with respect to availability of Ca, Fe, K, Mg, P, Sr and Zn. Other vegetable protein-rich feed ingredients like soybean meal ŽSBM. and corn gluten resulted in lower, and partly negative estimates for absorption of several elements in both fish species. The first aim of the present study was to find the digestibility of macronutrients, energy and amino acids, and the

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

117

absorption of ash and individual elements in diets where up to half of the protein from LT FM was replaced by WG in diets for Atlantic salmon. Morphological changes are observed in mucosa of the distal intestine of salmonids when using SBM in the diet Žvan den Ingh et al., 1991; Baeverfjord and Krogdahl, 1996.. These changes have been suggested to be of a similar nature to gluten intolerance Žceliac disease. in man ŽBaeverfjord and Krogdahl, 1996.. The changes can briefly be described as a combination of four features ŽBaeverfjord and Krogdahl, 1996.: Ž1. a shortening of heights of the mucosal foldings; Ž2. a loss of the normal supranuclear vascuolisation of the absorptive cells in the intestinal epithelium; Ž3. a widening of the central stroma within the mucosal foldings, with increased amounts of connective tissue; and Ž4. a profound infiltration of inflammatory cells in the lamina propria. Intestinal changes are moderate to small in chinook salmon and rainbow trout when they are fed diets with soy protein concentrate ŽBureau et al., 1998.. The changes, however, occur when salmonids are fed diets with the alcohol extract obtained by producing the protein concentrate from soybeans ŽKrogdahl et al., 1995; van den Ingh et al., 1996; Bureau et al., 1998.. Soy-protein-concentrate diets supplemented with Quillaja bark sapponines caused significant intestinal damage in coho salmon and rainbow trout ŽBureau et al., 1998., but purified soy sapponine did not cause soybean-induced enteritis in Atlantic salmon ŽKrogdahl et al., 1995.. Published information concerning the occurrence of such reactions in the intestine of salmonids fed other vegetable protein-rich feed ingredients other than soybean products does not, to our knowledge, exist. The second aim of the study was to find out if similar reactions were seen in the mucosa of the distal intestine of Atlantic salmon fed a diet with WG.

2. Materials and methods 2.1. Experiment 1 — digestibility 2.1.1. Feed ingredients and diets The macronutrient and amino acid compositions of the feed ingredients containing protein are presented in Table 1, while the element composition of the same ingredients and the calcium phosphate are presented in Table 2. The WG was produced by physical extraction from wheat, dried and finely ground Ž) 98% through 160 mm screens.. The FM was from herring Ž Clupea harengus ., caught and dried less than 2.5 months prior to the experiment. According to the producer, this batch of FM contained 20.5% watersoluble CP Žof total CP.; 0.10 g putrescine kgy1 , 0.45 g cadaverine kgy1 ; - 0.1 g histamine kgy1 , and the protein digestibility assessed with mink was 92.7%. The wheat meal was pre-cooked, dried and ground Ž95% through 160 mm, 99% through 300 mm screens.. Five diets ŽTable 3. were produced with WG replacing 0, 6.25, 12.5, 25 and 50% of the CP from FM. The diets were formulated to contain 440–450 g protein, 320 g fat, 130 g starch, 77–84 g ash and 15 g phosphorus kgy1 dry matter ŽTable 4., based on a preliminary analyses of the feed ingredients. The diets were extruded by T. Skretting, Stavanger, Norway, with a diameter of 6 mm.

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

118

Table 1 Macronutrient and amino acid composition of the protein-containing feed ingredients WG a

FM b

Wheat meal c

Dry matter ŽDM., g kg

950

928

900

In meal, kgy1 CP, g Fat, g Starch, g Ash, g Gross energy, MJ Phytic acid, g

810 55.6 70 10.1 22.4 2.1

748 122.1 0 119 21.0 –

118 11.7 700 7.1 15.8 0.5

3.5 3.7 5.2 38.3 11.4 3.0 2.7 3.9 2.2 1.6 3.5 7.0 3.2 5.0 1.5 1.9 3.6 0.9

9.1 4.4 4.3 13.0 3.7 5.1 6.2 5.0 0.8 2.6 4.0 7.3 2.9 3.6 7.2 1.9 6.7 1.0

4.4 2.9 5.1 37.8 12.2 3.4 3.1 4.2 2.2 1.5 3.6 7.3 2.7 5.1 2.0 2.1 3.7 0.9

y1

Amino acids, g Ž100 g CP.y1 Asp Thr Ser Glu Pro Gly Ala Val Cys Met Ile Leu Tyr Phe Lys His Arg Trp a

Amylgluten, Amylum Group, Aalst, Belgium. Norse LT-94, Norwegian Fish Meal Industries, Bergen, Norway. Used in Experiment 1. c Suprex Wheat, Codrico, Rotterdam, the Netherlands. b

The buoyancy of the diets decreased with increasing concentration of WG as the diets with 0–12.5% WG were floating while 25% and 50% WG were sinking pellets. 2.1.2. Fish and facilities Each of the five diets was fed to two groups of Atlantic salmon Ž Salmo salar L.. with a mean initial weight of 0.94 kg for 2 weeks ŽExperiment 1.. Each group of 20 fish was kept in a 1 m2 circular fiberglass tank with a water depth of 45 cm. Each tank was supplied with saltwater Ž32 g ly1 salinity., with a temperature of 7–88C. The salmon had been fed a commercial diet ŽT. Skretting, Stavanger, Norway. prior to the experiment. During the experiment, the fish were fed in excess, with electrically driven disc feeders ŽAkvaprodukter, Sunndalsøra, Norway., every 10 min, 7 h dayy1 . Excess feeding was verified by uneaten feed trapped by strainers in water outlet from the tanks, but feed intake was not quantified. At the termination of working hours, feed floating in the tanks was removed and the fish were left without feed until next morning

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

119

Table 2 Mineral composition of the protein-containing feed ingredients and the calcium phosphate WG a

FM b

Wheat meal c

Dicalcium phosphated

23.9 3.26 155 12.2 2.64 5.0 14.5 22.1 43.3 121.4

0.1 0.82 9.6 1.0 0.15 3.6 0.10 1.2 0.9 10.0

305 12.0 3347 2.3 6.13 457 2.6 205 198 242

.y1

Elements, Žkg DM Ca, g 1.0 Cu, mg 6.26 Fe, mg 77.7 K, g 1.3 Mg, g 0.70 Mn, mg 35.4 Na, g 0.26 P, g 2.5 Sr, mg 2.3 Zn, mg 59.6 a–c

See footnotes to Table 1. Feed grade, Windmill Dicalpos, The Windmill Feed Phosphates, Tessenderjo Chemie, Rotterdam, the Netherlands. d

in order to avoid ingestion of feed that may have leached nutrients. When floating feed was present in one or more tanks, the removal procedure was done in all tanks, in order to standardise the stress on the fish. 2.1.3. Sampling and analyses Faeces were collected by careful stripping at 10 and 14 days of feeding. Stripping of faeces was carried out as described by Austreng Ž1978.. Prior to stripping, the fish were anesthetised with metocain, 60 mg ly1 , dissolved in saltwater.

Table 3 Formulation of the diets for Experiment 1 Protein from gluten, percentage of total protein

0

6.25

12.5

25

50

0 567.9 174.0 239.5 0.1 13.4 5.2

36.8 530.1 169.9 241.1 0.1 16.9 5.2

73.4 492.6 165.7 242.9 0.1 20.2 5.2

146.3 418.3 157.4 246.3 0.1 26.5 5.2

291.3 269.9 141.0 253.2 0.1 39.4 5.2

y1

Formulation, g kg diet WG a FM a Wheat meal a Fish oil b Inert marker, Y2 O 3c Dicalcium phosphatea Vitamin and mineral premix d a

See footnotes to Tables 1 and 2. Denofa, Fredrikstad, Norway. c Sigma, St. Louis, MO, USA. d Active ingredient supplied per kilogram feed. Vitamins: retinol acetate, 5 mg; vitamin D 3 , 4.8 mg; a-tocopheryl acetate, 400 mg; thiamin–Cl, 15.3 mg; riboflavin, 26 mg; pyridoxine–Cl, 15.3 mg; Ca-panthotenate, 88.9 mg; niacin, 153.1 mg; folic acid, 5.2 mg; vitamin B12 , 20.0 mg; biotin, 150.0 mg; myo-inositol, 408.2 mg; vitamin K 3 , 40.0 mg; ascorbate polyphosphate, 1 g Žall vitamins were from F. Hoffmann La Roche, Basel, Switzerland.; choline chloride, 2.4 g. Minerals: 48.0 mg CuSO4 ; 2.0 mg KI; 109.4 mg MnSO4 ; 0.4 mg Na 2 SeO 3 ; 257.1 mg ZnSO4 Žall mineral salts were from Sigma.. b

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

120

Table 4 Chemical composition of the diets in Experiment 1 Protein from gluten, % Dry matter ŽDM., g kgy1 In diet, kgy1 CP, g Fat, g Starch, g Ash, g Gross energy, MJ Phytic acid, g Amino acids, g Ž100 g CP.y1 Asp Thr Ser Glu Pro Gly Ala Val Cys Met Ile Leu Tyr Phe Lys His Arg Trp Elements, Žkg DM.y1 Ca, g Cu, mg Fe, mg K, g Mg, g Mn, mg Na, g P, g Sr, mg Zn, mg a

0

6.25

12.5

25

50

942

956

947

952

951

431 319 116 80.5 24.1 ND a

438 323 106 78.5 24.4 ND a

437 322 114 76.1 24.3 0.1

450 322 114 75.6 24.4 0.3

458 315 114 71.5 24.6 0.5

7.8 4.0 3.8 13.3 4.1 4.7 5.5 4.7 0.8 2.3 3.9 6.8 2.7 3.5 6.4 1.8 6.2 0.9

6.8 3.5 3.6 13.7 4.3 4.1 4.7 4.3 0.9 2.1 3.5 6.2 2.4 3.2 5.4 1.7 5.6 0.9

7.2 3.8 4.0 16.4 5.2 4.3 5.0 4.6 1.0 2.2 3.8 6.8 2.7 3.7 5.6 1.8 5.8 0.9

6.5 3.5 4.1 19.4 6.2 4.1 4.5 4.4 1.2 2.0 3.7 6.7 2.8 3.9 4.7 1.7 5.3 0.9

5.4 3.2 4.1 23.7 7.8 3.7 3.8 4.1 1.4 1.8 3.6 6.5 2.8 4.1 3.6 1.8 4.6 0.8

17.7 16.9 143 7.81 1.66 52.2 7.89 15.5 27.2 169

17.5 17.1 148 7.37 1.58 59.1 7.32 15.3 25.8 160

17.6 17.4 159 6.63 1.51 58.7 6.95 15.1 24.7 157

18.2 17.7 176 5.66 1.41 63.2 6.09 15.2 23.4 159

19.0 16.6 206 4.41 1.23 67.7 4.21 15.1 21.1 153

Below detection limit.

The faeces were pooled from each tank, freeze-dried, and ground with a pestle and mortar. Feed ingredients, diets and faeces were analysed for nutrient composition Ždry matter, CP, crude fat, starch, energy, ash. and concentration of yttrium oxide as described by Refstie et al. Ž1997.. Elements in faeces and diets were analysed according

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

121

to Shearer Ž1984.. Phytic acid was analysed according to Thomson and Erdman Ž1982., following the experimental protocol described by Stone et al. Ž1984.. Amino acids were analysed according to Anonymous Ž1998.. 2.1.4. Calculations and statistical analyses Apparent digestibility coefficients ŽADC. of macronutrients and energy, and apparent absorption coefficient ŽAAC. of amino acids, ash and elements were calculated as described by Austreng Ž1978., except that Y2 O 3 was used as inert marker. The apparent AACs for several elements, such as Ca, Mg, Na, and Sr, are underestimated due to the uptake of elements from the water and excretion by the intestine ŽStorebakken et al., 1998.. The results were analysed statistically by one-way analysis of variance with tank mean as the unit of observation. Significant Ž P - 0.05. differences were ranked with Duncan’s multiple range test. Results concerning the ADC of protein and AACs of amino acids were analysed by regression according to the following model: ADC or AAC s a q bG,

G is protein Ž % of total CP . from WG.

2.2. Experiment 2 — histology in the distal intestine and growth 2.2.1. Feed ingredients and diets Three diets FM control; SBM control diet with 15% CP from SBM; WG diet with 35% of CP from WG Žthe same type as used in Experiment 1. ŽTable 5. were

Table 5 Formulation and chemical composition of the diets for Experiment 2 Diet name

FM

SBM

WG

Formulation, g kg diet WG a FM b SBM c Wheat meal Fish oil d Constant ingredientse

0 507 0 193 295 5

0 434 127 134 300 5

167 321 0 197 310 5

Chemical composition, kg y 1 diet Dry matter ŽDM., g CP, g Fat, g Ash, g Gross energy, MJ

932 413 341 60 25.3

932 396 330 61 24.9

942 395 347 41 25.8

y1

a

See footnote to Table 1. Norse LT-94 from Herring, Norsildmel. c DenoSoy, extracted, toasted, with hulls, Denofa, Fredrikstad, Norway. d NorSalmOil, Nordsildmel. e Vitamin premix, 2 g kgy1 ; mineral premix, 2 g kgy1 Žformulations proprietary to BioMar.; Carophyll Pink, 8% astaxanthin, Hoffmann-La Roche, Basel, Switzerland, 1 g kgy1 . b

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

122

formulated and extruded to 9 mm pellets by BioMar, Myre, Norway. The level of SBM inclusion was selected to facilitate rapid growth, but still enabling the detection of histological changes in the intestine. The FM, SBM and WG were from the same producers as in Experiment 1. The FM was from herring, from a different batch than in Experiment 1 Ž74.5% CP; 21.9% water-soluble CP; 0.66 g cadaverine kgy1 , 0.16 g NH 3-N kgy1 ; 0.29 g histamine kgy1 ; 92.1% mink protein digestibility.. 2.2.2. Fish, facilities and statistical analyses The experiment was carried out at AKVAFORSK, Averøy, Norway, with Atlantic salmon with an initial weight of 0.95 kg, for 18 weeks. Each diet was fed to apparent satiation to salmon in three floating net-pens with a volume of 125 m3, initially with 350 salmon in each pen. During the first month, the fish were fed by hand, one meal a day. During the rest of the experiment, the fish were fed one meal by hand and one by electric feeders. The salinity of the water was 33 g ly1 . The temperature at 3 m was 58C at the start, rapidly declined to 3.0–3.58C, and then increased to 12.58C at the end of the experiment, with an average temperature of 7.48C. The oxygen concentration in the pens, measured daily, was never below 8 mg ly1 . The growth results were statistically analysed by one-way ANOVA as described for Experiment 1. 2.2.3. Histology Five fish from each of the two control diets and 15 salmon fed the diet with WG were obtained for examination of intestinal morphology. The lower number of samples from the control diets was chosen because the lack of reaction to FM and the development of intestinal pathologies in fish fed SBM have been described previously ŽBaeverfjord and Krogdahl, 1996.. The fish were killed individually by an overdose of metacain. The fish were not fasted before sampling. The abdominal cavity was opened, the distal intestine was identified and dissected, and a 1-cm long piece was cut from the middle of the distal intestine. The ring of intestinal tissue was cut open, and the sample was rinsed in saline Ž9 g ly1 . to remove the gut contents. The sample was fixed by immersion in phosphate-buffered formalin Ž4%, pH 7. and stored in a refrigerator. After fixation, the samples were dehydrated in ethanol, equilibrated in xylene, and embedded in paraffin according to routine techniques. Sections of approximately 5 mm were cut and stained with haematoxylin and eosin ŽH & E.. Intestinal mucosa were classified into one of three categories based on a combination of four features ŽBaeverfjord and Krogdahl, 1996.: Ž1. a shortening of heights of the

Table 6 ADCs for fat, starch and energy Žmean"S.E.M.., ns 2 tanks per feed Protein from wheat gluten, %

0

6.25

12.5

25

50

ADC, % Fat Starch Energy

91.6"0.6 a,b 72.2"3.9 87.0"0.4 a,b

92.1"0.3 a,b,c 63.5"1.1 87.3"0.3 b

90.9"0.2 a 64.0"0.5 86.1"0.4 a

93.1"0.4 c 66.0"1.4 88.5"0.2 c

92.6"0.2 b,c 61.2"1.4 88.5"0.2 c

a,b,c

Different superscripts indicate significant Ž P - 0.05. differences among means.

Table 7 ADCs for CP and AAC of individual amino acids Žmean"S.E.M..1 , ns 2 tanks per feed 6.25

12.5

25

50

Regression2

r

P Žslope) 0.

Estimated 3 ADC or AAC in gluten, %

ADC, % Protein

88.6"0.4 a

89.7"0.1a

89.7"0.4 a

92.2"0.6 b

93.6"0.1b

88.8q0.118G

0.94

0.0002

100.6

AAC, % Asp Thr Ser Glu Pro Gly Ala Val Cys Met Ile Leu Tyr Phe Lys His Arg Trp

87.9"1.4 92.9"0.5a 92.1"0.7 a 95.5"0.3 b 92.9"0.1a 88.3"0.4 a 92.6"1.5 94.9"0.5 80.4"0.6 a 92.8"0.4 95.6"0.5a 96.1"0.4 a 95.3"0.7 95.2"0.6 a 96.1"0.3 93.8"0.8 a 96.3"0.1a 87.8"0.7 a

88.7"1.4 93.8"0.5a,b 93.3"0.2 a 96.6"0.1c 94.5"0.3 b 88.7"0.5a 94.1"0.1 95.6"0.3 83.4"0.7 b 93.7"0.9 96.1"0.1a 96.6"0.1a 95.7"0.1 95.7"0.0 a,b 96.3"0.1 94.7"0.1a 96.6"0.2 a,b 87.7"0.5a

88.3"0.2 93.3"0.1a 93.4"0.1a 96.3"0.0 c 95.2"0.2 c 89.6"0.2 a,b 93.8"0.3 95.1"0.1 84.6"0.4 b 93.3"0.2 95.6"0.2 a 96.3"0.2 a 95.8"0.2 95.8"0.2 a,b 96.0"0.1 94.7"0.1a 96.8"0.1b 88.4"0.1a

89.2"0.8 93.3"0.3 a 93.1"1.0 a 94.2"0.3 a 96.6"0.0 d 90.8"0.6 b 93.9"0.6 95.5"0.3 89.2"0.1c 93.9"0.4 96.1"0.1a,b 96.7"0.1a 95.9"0.2 96.3"0.2 b 95.8"0.1 94.9"0.1a 96.9"0.1b 91.2"1.1b

90.0"0.4 95.0"0.2 b 95.9"0.2 b 98.3"0.1d 97.8"0.2 e 92.8"0.4 c 94.7"0.2 96.6"0.1 92.3"0.2 d 95.6"0.3 97.0"0.2 b 97.5"0.1b 97.2"0.2 97.4"0.1c 95.8"0.3 96.3"0.3 b 97.4"0.1c 92.1"0.2 b

87.9q0.058G 93.0q0.035G 92.3q0.065G 95.8q0.053G 93.7q0.090G 88.3q0.092G 93.2q0.030G 95.0q0.029G 81.6q0.233G 92.9q0.052G 95.6q0.026G 96.1q0.027G 95.3q0.035G 95.3q0.041G 96.1–0.008G 94.0q0.043G 96.4q0.021G 87.6q0.099G

0.72 0.76 0.84 0.94 0.95 0.97 0.53 0.79 0.96 0.86 0.83 0.89 0.86 0.93 0.69 0.86 0.93 0.89

0.018 0.010 0.0024 0.0001 0.0001 0.0001 0.11 0.0060 0.0001 0.0016 0.0031 0.0005 0.0013 0.0001 0.069 0.0088 0.0001 0.0005

93.7 96.5 98.8 99.4 102.7 97.5 96.2 97.9 104.9 98.1 97.8 98.8 98.8 99.4 95.3 98.3 98.5 97.5

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

0

Dietary protein from WG, %

1

Different superscriptsa,b,c,d,e indicate significant Ž P - 0.05. differences among diet means. ADC of protein or AAC of amino acids aq bG, G is protein Ž% of total crude protein. from WG. 3 Estimated by extrapolating dietary WG to 100% of CP. 2

123

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

124

mucosal foldings; Ž2. a loss of the normal supranuclear vascuolisation of the absorptive cells in the intestinal epithelium; Ž3. a widening of the central stroma within the mucosal foldings, with increased amounts of connective tissue; and Ž4. a profound infiltration of inflammatory cells in the lamina propria. The results were categorised as follows: 0: No deviations from normal morphology; A: Intestinal tissue shows changes which are typical for diets containing solvent-extracted SBM, as described above; and B: Morphological changes observed, but to a lesser extent than in A andror not all criteria present.

3. Results The WG contained 2.1 g phytic acid kgy1 ŽTable 1.. Thus, phytate-P only accounted for 26% of total P in the gluten ŽTable 2.. Relative to the herring meal, gluten contained more glutamic acid, phenylalanine, proline, tyrosine and serine, while the content of aspartate, glycine, alanine, arginine, valine, methionine and lysine was lower ŽTable 1.. The amino acid composition of the wheat matched that of the WG. Compared with the FM, WG contained considerably less Ca, K, Na, P and Sr, moderately less Fe, Mg and Zn, and more Cu and Mn ŽTable 2.. The chemical composition of the diets ŽTable 4. was characterised by a slight increase of protein, from 431 to 458 g kgy1 , with increasing proportion of protein from WG. The amino acid composition of the diets varied according to the composition in the feed ingredients. Dietary fat, starch and energy were similar in all diets. Phytic acid was not detectable in the FM control diet or in the diet with 6.25% gluten, and it increased with increasing gluten inclusion up to 50%. The levels of Ca, Fe and Mn increased in the diets with increasing gluten content, Mg, Sr and Zn decreased slightly, while the decrease in K and Na was of a higher magnitude. The concentrations of Cu and P were similar in all five diets.

Table 8 Apparent absorption of elements Žmean"S.E.M.. Protein from wheat gluten, %

0

6.25

12.5

25

50

Apparent absorption, % Ca Cu Fe K Mg Mn Na P Sr Zn

y35"5 37"2 11.6"1.0 96.0"0.1 y500"76 y1"8 41.0"1 35.7"0.7 y382"88 33.5"3.3 a

y39"1 41"2 10.6"0.8 93.3"0.6 y546"15 y1"2 29.6"7 32.6"0.0 y434"1 45.4"1.6 b

y31"8 47"4 11.0"0.2 93.3"1.2 y537"44 5"3 5.7"25 34.4"0.8 y444"9 46.9"4.7 b

y33"2 47"2 13.6"6.2 93.8"1.3 y517"31 7"1 20.3"16 34.1"4.9 y442"5 48.2"0.2 b

y29"4 47"5 14.2"0.6 94.6"1.2 y451"6 10"1 26.6"20 39.3"1.3 y397"8 45.3"1.6 b

a,b

Different superscripts indicate significant Ž P - 0.05. differences among diet means.

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

125

Table 9 Weight and growth of the salmon in Experiment 2 Žmean"S.E.M.. a , ns 3 pens per feed Diet name

FM

SBM

WG

Fish weight, g 0 weeks 18 weeks SGRb

952"14 2186"28 0.66"0.02

953"10 2073"27 0.62"0.02

956"11 2176"49 0.65"0.01

a

No significant Ž P ) 0.05. differences were present. Specific growth rate, 100 dy1 wlnŽWF .ylnŽWI .x; d, feeding days; W, mean fish weight in a pen; subscripts I and F, initial and final values. b

The ADC for fat ŽTable 6. was significantly higher in the diet with 25% gluten than in the FM control diet and the diet with 12.5% gluten, while it did not differ from the other two diets. There also was a tendency Ž P s 0.077. that the ADC of starch was higher in the FM control diet than in the diets with gluten. The ADC of energy was highest for the two highest inclusion levels of gluten, and it was higher in the diet with 6.25% gluten than in the diet with 12.5% gluten. The ADC of CP and AACs of all amino acids ŽTable 7., except alanine and lysine, showed a significant increase with increasing proportion of protein from WG. The AAC of lysine was negatively correlated Ž r s 0.72. to dietary concentration of WG. The ADCs of protein for the diets with 25% and 50% WG were significantly higher than the values obtained for the diets with less WG. The only significant difference in apparent absorption of elements ŽTable 8. was that Zn was higher in all the diets with WG than in the FM control diet. The absorption estimates for Ca, Mg and Sr were negative. The estimate for absorption of Na was biased by a large non-systematic variation. The absorption of P was not significantly different for the various diets, and the mean absorption of P was 35%. The fish doubled their weight during Experiment 2 ŽTable 9., and there were no significant differences among the various dietary treatments in fish weights or growth. Results from the histological evaluation of the distal intestine of the salmon are presented in Table 10. Fish fed the diet with FM as the only source of protein displayed no changes in intestinal tissue. In the fish fed the diet with 15% of protein from SBM, three of the five salmon displayed changes which were typical for diets containing

Table 10 Pathological changes of the distal intestine of salmon fed a FM control diet, a diet with 15% of the protein from extracted, toasted SBM, and a diet with 35% protein from WG Diet

Number of samples

Observations

Aa

Ba

FM SBM WG

5 5 15

No changes from normal morphology Variable changes No changes in 14 of 15 evaluated fish. Minor changes in 1 of 15 fish

0r5 1r5 0r15

0r5 2r5 1r15

a

A: Typical changes as described for solvent-extracted SBM. All criteria fulfilled. B: Pathological changes indicated, but less grave or not all criteria present.

126

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

solvent extracted SBM. For this group as a whole, the changes were somewhat less severe than observed previously ŽBaeverfjord and Krogdahl, 1996., but the effect was nevertheless clearly identified. In salmon fed the diet with 35% protein from WG, one fish of the 15 which were examined showed very moderate reactions. The changes consisted of inflammatory cell infiltration of submucosal connective tissue structures, and were of unspecific nature.

4. Discussion The low amount of phytic acid and the low proportion of phytic P in the gluten diets indicate that endogenous phytases were active during processing of the feed ingredient, since about two thirds of the P in cereal grains is in the form of phytate ŽPointillard, 1993.. Phytase from wheat bran has previously been used to hydrolyse phytic acid from cannola meal in blends of fish silage, cannola meal and wheat bran ŽStone et al., 1984. and phytase from wheat middlings has been shown to increase P bioavailability in diets for pigs ŽLei et al., 1997.. The slight increase in dietary CP content, from 430 to 458 g kgy1 diet, with increasing content of WG is a direct result of the diet formulation, aimed at keeping the diets iso-energetic, and with a constant concentration of starch. Mundheim and Opstvedt Ž1990. did not find significant effects on ADC of protein by replacing protein with fat, while dietary starch affects ADC of other nutrients in Atlantic salmon ŽAksnes, 1995; Hemre et al., 1995a; Grisdale-Helland and Helland, 1997.. The increased Ca concentration in the diets with increasing gluten inclusion is a result of keeping the concentration of P constant, by supplementation with calcium phosphate, while the variation in the concentration of the other elements is a result mainly of differences between the FM and gluten. The differences in elemental concentration, however, were small to moderate, and should not affect uptake of the actual elements or interact with the uptake of other elements. The growth rate obtained in Experiment 2 by feeding the diet with 35% protein from WG and the FM diet was similar, on a thermal growth coefficient ŽIwamata and Tautz, 1981. basis, to that observed when feeding a diet similar to the FM control to 1 kg salmon at higher water temperatures ŽEinen and Roem, 1997.. This high growth rate illustrates that WG may successfully replace one third of the FM protein in high-energy diets for salmon. The digestibility of fat and energy was comparable to, or higher than, that previously reported for Atlantic salmon fed extruded high energy diets in saltwater ŽJohnsen et al., 1993; Aksnes, 1995; Thodesen and Storebakken, 1998.. In contrast to what has been reported when replacing FM with SBM ŽRefstie et al., 1998., replacement of LT FM with WG did not result in reduced ADC of fat or energy. On the contrary, ADC of fat was higher in the diet with 25% of the protein from gluten than that of the FM diet. The higher ADC of energy in the diets with 25% and 50% gluten than the FM control diet is the sum of effects of ADC of fat and protein, modified by the tendency toward lower ADC of starch in the diets with WG. This tendency toward reduced ADC of starch may

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

127

be ascribed to the a-amylase inhibitor found in wheat ŽHofer and Sturmbauer, 1985.. This inhibitor is a protein, which may be concentrated in the protein fraction when producing the gluten. The ADC of protein in the FM control diet agrees with previous findings with LT-94 FM in extruded high-energy diets fed to Atlantic salmon in saltwater ŽJohnsen et al., 1993; Aksnes, 1995.. The AAC of amino acids was generally 2–3% units higher than the results reported for a cold-pelleted LT-70 FM-based diet with approximately 180 g fat kgy1 ŽSkrede et al., 1998.. This illustrates that the quality of the herring meal used in the diets was high. In spite of this, the results showed that the majority of amino acids were better digested in gluten than in LT herring meal. Consequently, the ADC of CP increased with increasing inclusion of WG in the feed. The regression equation in Table 7 indicates that when WG is 100% of dietary CP, it is predicted that several of the amino acids will be completely absorbed from the diet. However, such extrapolations should be made with care, since the highest level of inclusion used in the experiment was 50% of the protein from WG, and that 12% Ž100 = Ž1 y r 2 .. of the variation in ADC of protein was due to other factors than the diet, or interactions among dietary components. The estimated ADC of CP, using the regression in Table 7, for a diet with WG as the only source of protein was 100%, in keeping with the value of 99% obtained by Pfeffer et al. Ž1995. in rainbow trout. The ADC of protein for the diet without gluten was 88.8%. The digestabilities presented in this paper were apparent values, not corrected for endogenous losses. True digestabilities of protein and amino acids are higher than the corresponding apparent values, because of the endogenous cost of protein digestion ŽBatterham, 1994.. It is probable that the tendency of lowered AAC of lysine with increasing dietary gluten level is explained by the endogenous lysine excretion representing a relatively higher proportion of faecal amino acids with increasing dietary gluten levels. Availability of cysteine in feed ingredients for salmonids may be low ŽSkrede et al., 1981, 1998. due to the negative effects of heat when the feed ingredient is dried, such as denaturation, oxidation and cross-linkage ŽOpstvedt et al., 1984.. Furthermore, sulphurcontaining amino acids are limiting in SBM ŽLovell, 1989.. Thus, the combination of high concentration and availability of cysteine in WG may contribute to an improved balance of absorbed amino acids in salmonids fed diets with limited availability of cysteine or ingredients which are low in sulphur-containing amino acids. There is limited information available on the amino acid requirement of Atlantic salmon reared in saltwater, but the requirement of lysine has recently been established to 16–18 g kgy1 dietary dry matter or 0.79–0.89 g MJy1 digestible energy in slowly growing salmon with an initial weight of 0.4 kg ŽBerge et al., 1998.. These workers observed mortality and poor growth in the beginning of the study, and SGR was only 0.69 at 88C in the best group at the end of the experiment. When compared to whole-body amino acid profiles in Atlantic salmon ŽWilson and Cowey, 1985., lysine is the first limiting amino acid in WG. Wilson and Cowey Ž1985. reported that whole salmon contained as much as 9.3 g lysiner100 g amino acids, while the lysine concentration in WG is only 1.5 gr100 g CP. This illustrates the need for lysine supplementation to the diets when a major amount of the protein originates from WG.

128

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

The optimal supplementation of lysine and eventually other essential amino acids must be resolved in future experiments with Atlantic salmon. In spite of the imbalance in amino acid composition of the gluten, the good growth obtained by replacing 35% of FM protein by WG in Experiment 2 indicates that the essential amino acid supply for protein synthesis may have been adequate. This seems to also be the case for the diet with 50% of CP from gluten. The lysine supply from this diet may be estimated based on the following assumptions: 1 kg of gain in a 1-kg Atlantic salmon constitutes of 180 g CP ŽShearer et al., 1994.. Whole-body protein of Atlantic salmon contains 9.28 g lysiner100 g amino acids ŽWilson and Cowey, 1985.. If the amino acid-N accounts for 92% of CP ŽStorebakken, unpublished., the net lysine requirement for growth, thus, should be approximately 15.4 g kgy1 gain. The lysine content of the diet with 50% protein from WG was 3.6 gr100 g CP, the dietary CP concentration was 458 g kgy1 , and AAC of lysine was 95.8. If assuming a feed conversion ratio of 1 kg dry feed intake per kilogram gain, the digestible lysine intake would be 15.8 g kgy1 gain. The gross lysine content in the diet with 50% of CP from WG, not corrected for digestibility, was 16.5 g kgy1 dry matter, or 0.76 g MJy1 digestible energy, both close to the lower range of the estimated requirements for lysine in salmon ŽBerge et al., 1998.. A consequence of the high protein level used in practical salmon diets, thus, is that the lysine requirement for growth was nearly or fully met in salmon fed the diet with 50% of CP from gluten. Some of absorbed lysine probably is deaminated and metabolised, and it is possible that supplementation with lysine would be needed in order to facilitate rapid growth and efficient feed conversion in salmon fed a diet with as much as 50% of CP from WG. Based on the same assumptions used for indicating lysine requirement based on composition of growth and digestible amino acid intake, the requirements for all other essential amino acids seemed to have been met with solid margins in all diets. In the diet with 50% CP from WG, histidine Ž60% excess. and threonine Ž70% excess. appeared to be the 2nd and 3rd limiting amino acids. The first limiting essential amino acid in the FM control diet appeared to be histidine Ž50% in excess.. The negative apparent absorption estimates for Ca, Mg, and Sr illustrates that those elements were taken up from the water and excreted in the gut. There are previous examples of negative absorption values for Ca, even when the fish are reared in freshwater ŽStorebakken, 1985.. In spite of this negative absorption estimate, salmon is able to maintain homeostatic regulation of Ca ŽStorebakken et al., 1998.. Ca is efficiently taken up from the water, and water-borne Ca, from drinking or excretion by the intestine, is the source of the negative absorption. Uptake of Sr closely resembles that of Ca ŽYamada and Mulligan, 1987.. Salmonids are also able to take up Mg ˚ ˚ 1992. and Zn ŽSpry et al., 1988. from the water. In contrast to ŽShearer and Asgard, what has been found for diets with 75% of the protein from soy protein concentrate ŽStorebakken et al., 1998., the diets with WG did not result in any drastic increase in the faecal output of Na. The P absorption at 35% in the diets was slightly lower than the 41% previously obtained with an extruded LT-94 FM control diet fed to Atlantic salmon in saltwater ŽStorebakken et al., 1998.. This previous diet contained 19.5 g P kgy1 , and the sources of P were 595 g FM and 13.5 g dicalcium phosphate kgy1 . The present control diet

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

129

contained 15.5 g P kgy1 , from 568 g FM and 13.5 g dicalcium phosphate kgy1 . The difference most probably is rationalised by differences in availability of P from the FM source, which has been documented in experiments with rainbow trout ŽRiche and Brown, 1996., and higher availability of P from the added dicalcium phosphate ŽNordrum et al., 1997.. The availability of P in the present experiment was equal to that obtained by Hillestad et al. Ž1999. using a diet with a solvent-extracted LT-94 FM to salmon in saltwater. This diet contained 15.3 g kgy1 , mainly originating from the FM, and the diet had no inorganic P added. The lack of reduced absorption of Ca, P or Mg in the salmon fed the diets with WG is logical, based on the low concentration of phytic acid present in feed ingredients and diets. The absorption of Zn in the FM control diet was lower than found previously in Atlantic salmon ŽStorebakken et al., 1998., while the values for the diets with gluten were slightly lower than those obtained with phytase-treated soy protein concentrate, and more than twice the values obtained with 75% of the protein from untreated soy concentrate. A direct comparison is, however, limited by the concentration of Zn being almost 100 mg kgy1 lower in the present than in the previous study. Sugiura et al. Ž1997. also found that Zn from WG was highly absorbable in rainbow trout. The present study does not give any explanation for the improved absorption of Zn when WG is included in the diets. Absorption of Zn is mediated by low-molecular-weight substances Žpeptides or amino acids. originated from protein digestion or feedback-regulated pancreatic secretions in rats ŽEvans et al., 1975; Wapnir and Stiel, 1986. and further regulated by metallothionin in the epithelial cells in lower vertebrates ŽHogstrand and Haux, 1991.. Zn is mainly excreted through the gills in rainbow trout ŽHardy et al., 1987.. Increased absorption and a homeostatic plateau for intestinal uptake appear to have been reached at the lowest level of gluten. The inclusion of WG affected the physical quality of the pellets. All pellets absorbed fat well, but the density of the diets increased with increasing inclusion of WG. Adjustments were made in the extrusion parameters to get a good physical quality, and it may be that the deviations from the general pattern in ADC of macronutrients was the result of a modification in the extrusion process rather than depending on the nutritional qualities of the feed ingredients. In this study, the difference in intestinal morphology between groups fed a FM diet and a diet with extracted SBM was clearly established, although the SBM-induced changes were less severe than previously observed by Baeverfjord and Krogdahl Ž1996.. There was, however, a difference between the soybean inclusion levels in this Ž15%. and the previous Ž40%. study. For the diet with WG, the small changes observed in one of the 15 individuals were non-specific and moderate, and were considered to be of little relevance to the question in issue.

5. Conclusion WG seems to be a useful feed ingredient in grower diets for Atlantic salmon. WG supported rapid growth and was highly digestible, did not reduce digestibility of fat or

130

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

energy or availability of essential elements, but a slight reduction of digestibility of starch was seen. The results also indicate that WG at a level up to 35% of dietary protein may be used with no adverse effect on the mucosa of the distal intestine.

Acknowledgements The experiments were supported by grants from Amylum, Aalst, Belgium and BioMar, Myre, Norway.

References Aksnes, A., 1995. Growth, feed efficiency and slaughter quality of Atlantic salmon, Salmo salar, L., given feeds with different ratios of carbohydrate and protein. Aquacult. Nutr. 1, 241–248. Anonymous, 1998. Commission directive 98r64rEC of September 1998 establishing community methods of analysis for the determination of amino-acids, crude oils and fats, and olaquindox in feedingstuffs ammending Directive 71r393rEEC. Off. J. Eur. Commun. L257, pp. 14–28. Austreng, E., 1978. Digestibility determination in fish using chromic oxide marking and analysis of contents from different segments of the gastrointestinal tract. Aquaculture 13, 265–272. ˚ 1996. Development and regression of soybean meal induced enteritis in Baeverfjord, G., Krogdahl, A., Atlantic salmon, Salmo salar L., distal intestine: a comparison with the intestines of fasted fish. J. Fish. Dis. 19, 375–387. Batterham, E.S., 1994. Ileal digestabilities of amino acids in feedstuffs for pigs. In: D’Mello, J.P.F ŽEd.., Amino Acids in Farmed Animal Nutrition. CAB International, Wallingford, UK, pp. 113–131. Berge, G.E., Sveier, H., Lied, E., 1998. Nutrition of Atlantic salmon Ž Salmo salar .; the requirement and metabolic effect of lysine. Comp. Biochem. Physiol. 120A, 477–485. Bureau, D.P., Harris, A.M., Cho, C.Y., 1998. The effect of purified alcohol extracts from soy products on feed intake and growth of chinook salmon Ž Oncorhynchus tsawytsscha. and rainbow trout Ž Oncorhynchus mykiss .. Aquaculture 161, 27–43. Davies, S.J., Morris, P.C., Baker, R.T.M., 1997. Partial substitution of fish meal and full-fat soya bean meal with wheat gluten and influence of lysine supplementation in diets for rainbow trout, Oncorhynchus mykiss ŽWalbaum.. Aquacult. Res. 28, 317–328. Einen, O., Roem, A.J., 1997. Dietary proteinrenergy ratios for Atlantic salmon in relation to fish size, growth, feed utilization and slaughter quality. Aquacult. Nutr. 3, 115–126. Evans, G.W., Grace, C.I., Votava, H.J., 1975. A proposed mechanism for zinc absorption in the rat. Am. J. Physiol. 228, 501–505. Fagbenro, O., Jauncey, K., 1995. Water stability, nutrient leaching and nutritional properties of moist fermented fish silage diets. Aquacult. Eng. 14, 143–153. Grisdale-Helland, B., Helland, S.J., 1997. Replacement of protein by fat and carbohydrate in diets for Atlantic salmon Ž Salmo salar . at the end of the freshwater stage. Aquaculture 152, 167–180. Hardy, R.W., Sullivan, C.V., Koziol, A.M., 1987. Absorption, body distribution and excretion of dietary zinc by rainbow trout Ž Salmo gairdneri .. Fish Physiol. Biochem. 3, 133–143. Hemre, G.-I., Sandnes, K., Torrissen, O., Waagbø, R., 1995a. Carbohydrate nutrition in Atlantic salmon Ž Salmo salar L... Aquacult. Fish. Manage. 26, 149–154. Hemre, G.I., Sandnes, K., Lie, Ø., Waagbø, R., 1995b. Blood chemistry and organ nutrient composition in Atlantic salmon, Salmo salar L., fed graded amounts of wheat starch. Aquacult. Nutr. 1, 37–42. ˚ ˚ T., Berge, G.M., 1999. Determination of digestibility of commercial salmon feeds. Hillestad, M., Asgard, Aquaculture 179, 81–94. Hofer, R., Sturmbauer, C., 1985. Inhibition of trout and carp a-amylase by wheat. Aquaculture 48, 277–283.

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

131

Hogstrand, C., Haux, C., 1991. Mini-review. Binding and detoxification of heavy metals in lower vertebrates with reference to metallothionein. Comp. Biochem. Physiol. 100C, 137–141. Iwamata, G.K., Tautz, A.F., 1981. A simple growth model for salmonids in hatcheries. Can. J. Fish. Aquat. Sci. 38, 649–656. Johnsen, F., Hillestad, M., Austreng, E., 1993. High energy diets for Atlantic salmon. Effects on pollution. In: Kaushik, S.J., Luquet, P. ŽEds.., Fish Nutrition in Practice, Biarritz ŽFrance., June 24–27, 1991. Ed. INRA, Paris ŽLes Colloques, no. 61., pp. 391–401. ˚ Roem, A., Bæverfjord, G., 1995. Effects of soybean sapponine, raffinose and soybean alcohol Krogdahl, A, extract on nutrient digestibilities, growth and intestinal morphology in Atlantic salmon. In: Svennevig, N., ˚ ŽEds.., Quality in Aquaculture. Proc. Int. Conf. Aquaculture ’95 and the Satellite Meeting Krogdahl, A. Nutrition and Feeding of Cold Water Species. Trondheim, Norway, August 9–12, 1995. Eur. Aquacult. Soc. Spec. Publ. no. 23, Gent, Belgium, pp. 118–119. Lei, X.G., Han, Y.M., Rokener, K.R., 1997. Cereal phytases may improve phosphorus bioavailability. Feedstuffs 69 Ž52. 12, 15. Lovell, R.T., 1989. Use of soybean products in diets for aquaculture species: revised. Proc. Soybean Utilization Alternatives. The Center for Alternative Crops and Products, Univ. Minnesota, Feb. 16–18, 1988, pp. 235–265. Mundheim, H., Opstvedt, J., 1990. Effect of dietary level of protein and fiber on apparent protein digestibility in the rainbow trout Ž Oncorhynchus mykiss . and salmon Ž Salmo salar . and comparison of protein digestibility in mink Ž Mustela Õison., rainbow trout and salmon. In: Takeda, M, Watanabe, T. ŽEds.., The Current Status of Fish Nutrition in Aquaculture. Proc. 3rd Int. Symp. on Feeding and Nutrition of Fish, Toba, Japan, August 28–September 1, 1989, pp. 195–200. ˚ ˚ T., Shearer, K.D., Arnesen, P., 1997. Availability of phosphorus in fish bone meal and Nordrum, S., Asgard, inorganic salts to Atlantic salmon Ž Salmo salar . as determined by retention. Aquaculture 157, 51–61. Opstvedt, J., Miller, F., Hardy, R.W., Spinelli, J., 1984. Heat induced changes in sulfhydryl groups and disulfide bonds in fish proteins and their effects on protein and amino acid digestibility in rainbow trout Ž Salmo gairdneri .. J. Agric. Food Chem. 32, 929–935. Pfeffer, E., Al-Sabty, H., Haverkamp, R., 1992. Studies on lysine requirements of rainbow trout Ž Oncorhynchus mykiss . fed wheat gluten as only source of dietary protein. J. Anim. Physiol. Anim. Nutr. 67, 74–82. Pfeffer, E., Kinzinger, S., Rodenhutscord, M., 1995. Influence of the proportion of poultry slaughter by-products and of untreated or hydrothermically treated legume seeds in diets for rainbow trout, Oncorhynchus mykiss ŽWalbaum., on apparent digestibilities of their energy and organic components. Aquacult. Nutr. 1, 111–117. Pointillard, A., 1993. Importance of phytates and cereal phytases in the feeding of pigs. In: Wenk, C., Boessinger, M. ŽEds.., Enzymes in Animal Nutrition. Kartause Ittingen, Switzerland, pp. 192–198. Refstie, S., Helland, S.J., Storebakken, T., 1997. Adaptation to soybean meal in diets for rainbow trout, Oncorhynchus mykiss. Aquaculture 153, 263–272. Refstie, S., Storebakken, T., Roem, A.J., 1998. Feed consumption and conversion in Atlantic salmon Ž Salmo salar . fed diets with fish meal, extracted soybean meal, or soybean meal with reduced content of oligosaccharides, trypsin inhibitors, lectins and soya antigens. Aquaculture 162, 301–312. Riche, M., Brown, P.B., 1996. Availability of phosphorus from feedstuffs fed to rainbow trout, Oncorhynchus mykiss. Aquaculture 142, 269–282. Schumacher, A., Wax, C., Gropp, J.M., 1997. Plasma amino acids in rainbow trout Ž Oncorhynchus mykiss . fed intact protein or a crystalline amino acid diet. Aquaculture 151, 15–28. Shearer, K.D., 1984. Changes in the elemental composition of hatchery-reared rainbow trout, Salmo gairdneri, associated with growth and reproduction. Can. J. Fish. Aquat. Sci. 41, 1592–1600. ˚ ˚ T., 1992. The effect of water-borne magnesium on the dietary magnesium requirement Shearer, K.D., Asgard, of the rainbow trout Ž Oncorhynchus mykiss .. Fish Physiol. Biochem. 9, 387–392. ˚ ˚ Shearer, K.D., Asgard, T., Andorsdottir, G., Aas, G.H., 1994. Whole body elemental and proximate ´ composition of Atlantic salmon Ž Salmo salar . during the life cycle. J. Fish. Biol. 44, 785–7997. Skonberg, D., Hardy, R.W., Barrows, R.T., Dong, F.M., 1998. Color and flavour analyses from farm-raised rainbow trout Ž Oncorhynchus mykiss . fed low-phosphorus feeds containing corn or wheat gluten. Aquaculture 166, 269–277.

132

T. Storebakken et al.r Aquaculture 184 (2000) 115–132

˚ Austreng, E., 1981. Digestibility of amino acids in raw fish flesh and meat-andSkrede, A., Krogdahl, A., bone-meal for the chicken, fox, mink and rainbow trout. Z. Tierphysiol., Tierernaehr. Futtermittelkd. 25, 35–49. Skrede, A., Berge, G.M., Storebakken, T., Herstad, O., Aarstad, K.G., Sundstøl, F., 1998. Digestibility of bacterial protein grown on natural gas in mink, chicken and Atlantic salmon. Anim. Feed Sci. Technol. 76, 103–116. Spry, D.J., Hodson, P.V., Wood, C.M., 1988. Relative contribution of dietary and waterborne zinc in the rainbow trout, Salmo gairdneri. Can. J. Fish. Aquat. Sci. 45, 32–41. Stone, F.E., Hardy, R.W., Spinelli, J., 1984. Autolysis of phytic acid and protein in cannola meal Ž Brassica spp.., wheat bran ŽTriticum spp.. and fish silage blends. J. Sci. Food Agric. 35, 513–519. Storebakken, T., 1985. Binders in fish feeds: I. Effect of alginate and guar gum on growth, digestibility, feed intake and passage through the gastrointestinal tract of rainbow trout. Aquaculture 47, 11–26. Storebakken, T., Austreng, E., 1987. Binders in fish feeds: II. Effect of different alginates on the digestibility of macronutrients in rainbow trout. Aquaculture 60, 121–131. Storebakken, T., Shearer, K.D., Roem, A.J., 1998. Availability of protein, phosphorus and other elements in fish meal, soy protein concentrate and phytate-treated soy protein concentrate based diets to Atlantic salmon, Salmo salar. Aquaculture 161, 365–379. Sugiura, S.H., Dong, F.M., Rathbone, C.K., Hardy, R.W., 1997. Apparent protein digestibility and mineral availabilities in various feed ingredients for salmonid feeds. Aquaculture 159, 177–202. Thodesen, J., Storebakken, T., 1998. Digestibility of diets with precooked rye or wheat by Atlantic salmon, Salmo salar L. Aquacult. Nutr. 4, 123–126. Thomson, D.B., Erdman, J.W., 1982. Phytic acid determination in soybeans. J. Food Sci. 47, 513–517. ˚ Olli, J.J., Hendriks, H.G.C.J.M., Koninkx, J.G.J.F., 1991. Effects of van den Ingh, T.S.G.A.M., Krogdahl, A., soybean-containing diets on the proximal and distal intestine in Atlantic salmon Ž Salmo salar .: a morphological study. Aquaculture 94, 297–305. ˚ 1996. Alcohol-soluble components in soybeans cause van den Ingh, T.S.G.A.M., Olli, J.J., Krogdahl, A., morphological changes in the distal intestine of Atlantic salmon, Salmo salar L. J. Fish. Dis. 19, 47–53. Wapnir, R.A., Stiel, L., 1986. Zinc intestinal absorption in rats: specificity of amino acids as ligands. J. Nutr. 116, 2171–2179. Wilson, R.P., Cowey, C.B., 1985. Amino acid composition of whole body tissue of rainbow trout and Atlantic salmon. Aquaculture 48, 373–376. Yamada, S.B., Mulligan, T.J., 1987. Marking unfed salmonid fry with dissolved strontium. Can. J. Fish. Aquat. Sci. 44, 1502–1506. Yamamoto, T., Akiyama, T., 1995. Effect of carboxymethylcellulose, alpha-starch, and wheat gluten incorporated in diets as binders on growth, feed efficiency, and digestive enzyme activity of Japanese flounder. Fish. Sci. 61, 309–313.