Ž . Aquaculture 188 2000 363–382 www.elsevier.nlrlocateraqua-online Potential of plant-protein sources as fish meal ž / substitutes in diets for turbot Psetta maxima : growth, nutrient utilisation and thyroid status Christine Burel a , Thierry Boujard a, ) , Sadasivam J. Kaushik a , Gilles Boeuf b , Serge Van Der Geyten c , Koen A. Mol c , Eduard R. Kuhn c , Alain Quinsac d , Michel Krouti d , ¨ Daniel Ribaillier d a Laboratoire de Nutrition des Poissons, Unite mixte INRA-IFREMER, Station d’Hydrobiologie, BP 3, ´ 64310 Saint Pee-sur-NiÕelle, France ´ b Laboratoire de Physiologie des Poissons, IFREMER, Station de Brest, BP 70, 29280 Plouzane, France ´ c Laboratory of ComparatiÕe Endocrinology, Catholic UniÕersity of LeuÕen, Naamsestraat 59, 3000 LeuÕen, Belgium d Laboratoire d’Analyses CETIOM, 270, aÕenue de la Pomme de Pin, 45160 Ardon, France Received 4 October 1999; received in revised form 1 February 2000; accepted 1 February 2000 Abstract An experiment was conducted in order to assess the incorporation in diets for juvenile turbot of Ž . Ž . Ž . extruded lupin Lupinus albus and heat-treated RM1 or untreated RM2 rapeseed meals Ž .Ž . Brassica napus 26 and 40 mmol glucosinolaterg DM, respectively . The level of incorporation of 30% for each plant-protein, as well as 46% for RM1 and 50% for lupin was tested and Ž compared with a fish meal based control diet. Triplicate groups of turbot initial body weight of 66 . g were fed by hand with isonitrogenous and isoenergetic experimental diets, twice daily and to visual satiety, during 63 days. Extruded lupin can be incorporated in diets of turbot up to a level of 50% without adverse effects on growth performance and body composition. Rapeseed meal can only be incorporated at levels up to 30%, but a preliminary heat treatment of RM is necessary in order to improve its nutritional quality. In turbot-fed the RM-based diets, plasma T levels were 4 Ž . reduced with low dietary content in glucosinolate breakdown products 3.6 mmolrg , but no Ž decrease in plasma T levels was observed with the higher level of toxic compounds 4.4 3 . mmolrg . A significant deiodinase type II compensatory effect, leading to an increase of the ) Corresponding author. Tel.: q 33-5-59-51-59-96; fax: q 33-5-59-54-51-52. Ž . E-mail address: [email protected]T. Boujard . 0044-8486r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. Ž . PII: S0044-8486 00 00342-2
Potential of plant-protein sources as fish mealž /substitutes in diets for turbot Psetta maxima :
growth, nutrient utilisation and thyroid status
Christine Burel a, Thierry Boujard a,), Sadasivam J. Kaushik a,Gilles Boeuf b, Serge Van Der Geyten c, Koen A. Mol c,
Eduard R. Kuhn c, Alain Quinsac d, Michel Krouti d,¨Daniel Ribaillier d
a Laboratoire de Nutrition des Poissons, Unite mixte INRA-IFREMER, Station d’Hydrobiologie, BP 3,´64310 Saint Pee-sur-NiÕelle, France´
b Laboratoire de Physiologie des Poissons, IFREMER, Station de Brest, BP 70, 29280 Plouzane, France´c Laboratory of ComparatiÕe Endocrinology, Catholic UniÕersity of LeuÕen, Naamsestraat 59,
3000 LeuÕen, Belgiumd Laboratoire d’Analyses CETIOM, 270, aÕenue de la Pomme de Pin, 45160 Ardon, France
Received 4 October 1999; received in revised form 1 February 2000; accepted 1 February 2000
Abstract
An experiment was conducted in order to assess the incorporation in diets for juvenile turbot ofŽ . Ž . Ž .extruded lupin Lupinus albus and heat-treated RM1 or untreated RM2 rapeseed meals
Ž . Ž .Brassica napus 26 and 40 mmol glucosinolaterg DM, respectively . The level of incorporationof 30% for each plant-protein, as well as 46% for RM1 and 50% for lupin was tested and
Žcompared with a fish meal based control diet. Triplicate groups of turbot initial body weight of 66.g were fed by hand with isonitrogenous and isoenergetic experimental diets, twice daily and to
visual satiety, during 63 days. Extruded lupin can be incorporated in diets of turbot up to a level of50% without adverse effects on growth performance and body composition. Rapeseed meal canonly be incorporated at levels up to 30%, but a preliminary heat treatment of RM is necessary inorder to improve its nutritional quality. In turbot-fed the RM-based diets, plasma T levels were4
Ž .reduced with low dietary content in glucosinolate breakdown products 3.6 mmolrg , but noŽdecrease in plasma T levels was observed with the higher level of toxic compounds 4.43
.mmolrg . A significant deiodinase type II compensatory effect, leading to an increase of the
0044-8486r00r$ - see front matter q2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0044-8486 00 00342-2
( )C. Burel et al.rAquaculture 188 2000 363–382364
conversion of T to T , was observed in vitro in the liver of turbot fed RM1-based diets. The4 3
intake of lupin-based diets also had an effect on thyroid status with an increase of plasma T3
levels and of deiodinase type I activity in liver and kidney, suggesting an increase in thedegradation of rT and in the conversion of T to T . q 2000 Elsevier Science B.V. All rights3 4 3
ŽInformation on the nutritional requirements of turbot are rather limited Guillaume et. Žal., 1991 . Nevertheless, several experiments Adron et al., 1976; Andersen and Alsted,
.1993; Danielssen and Hjernes, 1993 indicate that this fish species requires a high levelŽ .of dietary protein 50% to 60% of the diet . It is therefore of interest to replace some of
the fish meal, which represents between 50% and 70% of the diet, with protein-richplant ingredients. Nevertheless, fish meal and plant-protein differ in a number of waysincluding protein and energy content, amino acid profile, mineral composition, as wellas nutrient and energy digestibility depending on animal species. Plant materials also
Ž .contain a high proportion of fibre and anti-nutritional factors ANFs that may haveŽ .adverse effects on both their nutritional value and palatability Kaushik, 1989 . Among
the plant ingredients of potential value as protein sources for fish, lupin and rapeseed areoften cited.
Several workers have demonstrated the value of lupin seed meals in diets for rainbowŽtrout De la Higuera et al., 1988; Hughes, 1988, 1991; Gomes and Kaushik, 1989;
.Moyano et al., 1992; Bangoula et al., 1993; Gouveia et al., 1993; Burel et al., 1998 .The maximal dietary level of incorporation of lupin has been found to vary between30% and 70% without affecting growth performance of fish, depending upon the variety
Ž .of lupin and the technological process lupin seed meal or extruded lupin involved.Ž .Plant breeders have developed some strains of white lupin L. albus with low content
Ž .of alkaloids Roemer, 1993 , components which affect the palatability of the diet. ForŽinclusion in fish diets, lupin does not seem to require heat treatment De la Higuera et
. Ž .al., 1988 probably because it does not contain haemagglutinins lectins or trypsininhibitors. Another advantage of lupin is its low phosphorus content which can lead to awaste reduction of this compound without any adverse effect on growth performance of
Ž .the fish Burel et al., 1998 .Rapeseed meal is also a potential substitute for fish meal and has been the object of
Žnumerous studies in rainbow trout Yurkowski et al., 1978; Hardy and Sullivan, 1983;Hilton and Slinger, 1986; Leatherland et al., 1987; Gomes and Kaushik, 1989; AbdouDade et al., 1990; McCurdy and March, 1992; Gomes et al., 1993; Teskeredzic et al.,
. Ž . Ž .1995 , Atlantic salmon Higgs et al., 1982 and channel catfish Webster et al., 1997 .Ž .Rapeseeds are known to contain glucosinolates GLS whose deleterious effects limit theŽ . Ždietary level of incorporation of rapeseed meal RM in fish diet to 20–30% Hardy and
Sullivan, 1983; Hilton and Slinger, 1986; Leatherland et al., 1987; Gomes and Kaushik,.1989; Gomes et al., 1993; Burel, 1999 . The GLS metabolites, such as isothiocyanates,
thiocyanate anions, and vinyloxazolidinethiones, have a goitrogenic activity in animals,
( )C. Burel et al.rAquaculture 188 2000 363–382 365
Žincluding fish Yurkowski et al., 1978; Higgs et al., 1982, Hardy and Sullivan, 1983;Leatherland et al., 1987; Hossain and Jauncey, 1988; Teskeredzic et al., 1995; Burel,
.1999 . A hyperactivity of the thyroid follicles, accompanied by low plasma T and T3 4
levels have been observed, as well as an in vitro deiodinase compensatory effect, i.e. anincrease of the conversion of T in T and a decrease of the degradation of T . The4 3 3
mechanism of action of such GLS breakdown products has been demonstrated inŽ . Žmammals reviewed by Mawson et al., 1994 , and appears to be similar in fish Burel,
.1999 .The present study was undertaken in order to determine the effect of a dietary
incorporation of extruded lupin and RM on growth performance, feed intake and nutrientutilisation of juvenile turbot. In addition, the thyroidal status, i.e. plasma T and T3 4
Ž .levels and the deiodinase activities D1, D2 and D3 in different tissues, was alsoevaluated in order to study the potential goitrogenic effect of the toxic compoundscontained in the RMs in comparison with rainbow trout. Attention was also paid to therelationship between growth and thyroid status in turbot.
2. Material and methods
2.1. Characteristics of plant ingredients tested, feed preparation and distribution
Ž . Ž . ŽThe extruded lupin L Lupinus albus was provided by UNIP Union Nationale.Interprofessionelle des Proteagineux, Paris, France . The processing of lupin was´
Žconducted by CETIOM Centre Technique Interprofessionel des Oleagineux Metropoli-´ ´.tains, Bordeaux, France . After crushing, the lupin seeds were separated from the hulls
Ž .by exhaustion cleaner-separating system Denis D50 . The raw flour obtained wasŽ .extruded extruding system France extrusion, 1278C, 31 bar, 89.8 Wrkg .
Ž .Rape seeds Brassica napus, provided by CETIOM were submitted to very strongdehulling treatment, in order to reduce their fibre content. Then, two different methods
Ž .of fat extraction were used. The first, producing rapeseed meal 1 RM1 , consisted of aŽ .pressure-cooking 978C, 50 bar under wet conditions in order to extract oil. The
residual oil was extracted by six consecutive hexane washings in percolation, then aŽ .desolventation was realised by steam injection 1058C, 6 bar . The second method,
Ž .producing rapeseed meal 2 RM2 , consisted of a direct oil extraction: rape seeds weresubmitted to a double pressing, then oil was extracted by eight consecutive hexane
Ž .washings. The desolventation was realised by steam injection 808C, 3 bar . The heattreatment of the RM1 decreased the total glucosinolate level to 26 mmolrg DM,compared to the RM2 that contained 40 mmol GLSrg DM.
Ž .The plant-ingredients were ground F400 mm before incorporation into diets. Theirchemical composition is shown in Table 1, accompanied by data on the fish meal andextruded peas used, and their amino acid profiles are given in Table 2. As the results of
Ž .our previous studies Burel et al., 1998 showed that the incorporation of 30% ofextruded peas in the diet of rainbow trout had deleterious effect neither on the growthperformance, nor on nutrient and energy utilisation and on thyroid status, this ingredienthas been used in the present study as a source of carbohydrate.
( )C. Burel et al.rAquaculture 188 2000 363–382366
Table 1Chemical composition of the dietary ingredients
Ž .Six experimental diets were formulated Table 3 with the fish meal content varyingŽ .between 34.5% and 59%. Two levels of incorporation 30% and 50% were made for
extruded lupin. The levels of incorporation of 30% and 46% were tested for RM1, butŽonly the level of 30% for RM2 which, according to the results obtained in trout Burel,
.1999 , was likely to induce growth depression and thyroid disturbance, because of itshigher level in GLS. All diets were formulated to be isonitrogenous and isoenergetic
Table 2Ž .Level of total essential amino acids gr16 g nitrogen in protein-sources as compared to the requirements of
turbotData in italic are the amino acids which could be deficient in the ingredient for turbot.
) Ž .Data on the requirements in amino acids of turbot are from the estimations of Kaushik 1998 .
( )C. Burel et al.rAquaculture 188 2000 363–382 367
Table 3Formulation and chemical composition of diets used in the growth experiment
Ž .sodium selenite 30% Se , 1.5 mg; KCl, 0.45 g; NaCl, 0.2 g.6 Ž .The amount of N-free extracts lignin, no-starch polysaccharides, oligosaccharides was estimated as
Ž . Ž .follows: N-free extracts % DM s100y AshqCrude proteinqCrude fatqStarch .7The total dietary iodine levels were estimated from the iodine contents of ingredients proposed by
Ž .Ferrando and Blum 1989 .
Ž .Table 3 and their amino acid profile met or exceeded the requirements of turbot,Ž .according to the estimation of Kaushik 1998 . The dietary content in GLS and
( )C. Burel et al.rAquaculture 188 2000 363–382368
Table 4Ž .GLS and GLS breakdown product levels mmolrg DM in the RM-based diets used in the experiment
0.05 mmolrg DM is the threshold of assay sensibility.
1Alkenyl glucosinolates.2Aromatic and indol glucosinolates.3Measured in the diets.4Calculated from the rapeseed meals contents.5Estimated as describing in Section 2.
Ž .hydrolytic toxic end products of GLS estimated values are shown in Table 4. For theR2-30 diet, the concentration of total GLS measured in the diet was higher than thetheoretical concentration calculated. Disregarding the potential errors of estimation or ofmeasurement, we concluded that the concentration of GLS derivatives was nil in thisdiet.
2.2. Experimental conditions and design
2.2.1. Preliminary experimentIn order to determine the best feeding strategy in terms of number and timing of
meals, a preliminary experiment was performed under the same experimental conditionsŽ . Ž .as in the present study see below with 960 turbot initial mean weight around 65 g fedŽ .a fish meal-based diet 51% DM protein and 12% DM lipid . Four feeding schedules
were tested in triplicate: one meal in the morning, one meal in the evening, two mealsŽ . Ž .morning and evening , and three meals morning, midday and evening . The best resultsŽ . Ž .Table 5 were obtained with two meals per day morning and evening when turbotwere fed by hand to apparent satiety.
Since no data were available in the literature on the daily variation in thyroid status,plasma thyroid hormone levels and deiodinase activities were measured in fish from the
Žpreliminary experiment on feeding schedules, at five times in the day 3:00, 8:00, 13:00,.18:00 and 23:00 . Results showed that these parameters were neither significantly
affected by feeding schedules nor by the hour of sampling.
( )C. Burel et al.rAquaculture 188 2000 363–382 369
Table 5Growth, feeding parameters and final whole body composition of turbot fed a fish meal-based diet to satiation
Ž . Ž . Žby one meal per day morning or evening , two meals morning and evening or three meals morning, midday. Ž .and evening during 9 weeks IBW: 65.7"0.5 g
ŽSuperscript letters indicate intergroup statistical differences One way ANOVA and Tukey’s multiple range. Ž . Ž .test between dietary conditions. Means ns3 with no common letter are significantly different P -0.05 .
For each row, the MSE between groups and within groups is given.IBW and FBW are initial and final body weight respectively. The initial whole body sample contained: 24.1%of dry matter, 71.1% DM of protein, 8.8% DM of fat, 20.8 kJrg DM of energy, 15.5% DM of ash and 2.6%DM of protein.
Morning Evening Two Three MSE between MSE withinmeal meal meals meals groups groups
Growth performancea ab b abFinal body 129.9 133.4 142.7 134.2 88.47 15.31
Ž . Ž .weight FBW ga a b abDaily growth 1.64 1.69 1.87 1.74 0.030 0.004
1Ž .index DGIab a b bVoluntary feed 1.29 1.25 1.41 1.39 0.018 0.003
2Ž .intake VFIab b ab aFeed efficiency 0.92 0.97 0.93 0.90 0.0028 0.0004
3Ž .FE
( )Final whole body composition DM basisb a b bŽ . Ž .Dry matter DM % 25.2 23.9 25.2 25.0 1.15 0.09
b b a bŽ .Ash % 14.6 14.7 13.6 14.8 0.88 0.31Ž .Gross energy kJrg 20.7 21.1 20.9 20.8 0.06 0.19
Ž .Phosphorus % 2.6 2.6 2.3 2.5 0.04 0.02
1 ŽŽ .1r3 Ž .1r3.DGIs100= FBW y IBW rduration.2 Ž . ŽŽ . .VFIs100=dry feed intake g r IBWq FBW r2 =duration.3 Ž . Ž .FEsWet weight gain g rdry feed intake g .
2.2.2. ExperimentŽ .The experiment was performed with juvenile turbot initial mean weight 50 g which
Ž .were reared at the IFREMER facilities in Brest France , according to the methodsŽ .described by Person-Le-Ruyet et al. 1991 . Turbot were randomly allotted into 12
Ž .Swedish-type tanks 1 m=1 m; 500 l; 65 fishrtank supplied with filtered, gas-de-Žsaturated and thermoregulated fresh seawater. Water temperature was 16"0.58C mean
."S.D. , flow rate was maintained at 10 lrmin, salinity was 34–35‰ and the oxygenŽlevel was maintained above 5–6 mg O rl. An artificial photoperiod of 12Lr12D light2
.on at 08:00, 160 lx at the water surface was maintained using 18 W neon tubes, and anartificial dawn and dusk was created during the first and last half hours of the
Ž .photophase, using 25 W bulbs 20 lx at the water surface .ŽAfter a 1-week adaptation period, the fish were fed the experimental diets twice a
. Ž .day, by hand, to visual satiety for 9 weeks two groups per treatment . Fish biomasswas determined for each replicate every 3 weeks, and the total feed intake was recorded.Four days after termination of the growth measurements, tissues and blood were
( )C. Burel et al.rAquaculture 188 2000 363–382370
Ž .sampled from 24 fish per dietary treatment 12 per duplicate after the morning meal.Blood samples were obtained from vessel near the caudal peduncle using a heparinisedsyringe. Blood was immediately centrifuged and plasma was stored at y208C. Liver,brain, heart and kidney were removed immediately after blood sampling, frozen in liquidnitrogen and stored at y808C.
2.3. Analytical methods
2.3.1. Chemical analysis of ingredients, diets, and whole bodyŽWhole fish bodies were pooled 3=10 fish at the start of the experiment, then 5 fish
.per treatment at the end , ground and freeze-dried before chemical analyses. Ingredients,Ždiets, and whole body samples were analysed following standard procedures AOAC,
.1995 : dry matter after drying at 1058C for 24 h; ash by combustion at 5508C for 12 h;Ž . Žprotein N=6.25 by the Dumas method with an Elementary Analyser NA 2000 Fison
. Ž .Instrument ; gross energy in an adiabatic bomb calorimeter Parr ; fat after extractionwith petroleum ether by the Soxhlet method, total phosphorus by spectrophotometricanalysis of the phosphovanadomolybdate complex after mineralization and acid diges-
Ž . Žtion AFNOR, 1980 . Starch was measured by an enzymatic method Thivend et al.,.1972 using glucoamylase and glucose oxidase. For amino acids, the ingredients were
hydrolysed with 6 N HCl at 1108C for 24 h and the chromatographic separation andŽanalysis of the amino acids was performed after orthophthaldehyde OPA: Sigma P
.1378 derivatization of amino acids using High Performance Liquid ChromatographŽ .HPLC-Varian Model 5000, Aminotag reversed phase column following modified
Ž .procedure of Gardner and Miller 1980 .Ž .GLS in RMs and diets were analysed according to Quinsac et al. 1991 . GLS
breakdown products in diets were estimated considering that the GLS content measuredin RM2 is the initial amount of intact GLS without hydrolysis, and that the final amountmeasured in the RM1- and RM2-based diets resulted from hydrolysis occurring duringthe oil extraction process in the case of the RM2-based diets or the granulation of thediets. Then, the following calculation was used for the RM1- and RM2-based diets:
ŽAmount of breakdown productss Amount of GLS in dry matter of RM2=% dietary. Žincorporation of RM1 or RM2=% dry matter of diet y amount of GLS measured in
.diet=% dry matter of dietThe amount of GLS derivatives, expressed as equivalent mmol of GLS hydrolysed,
can be over-estimated because of the volatile breakdown products which may be lostduring oil extraction or granulation. To convert values into mgrg, the followingapproximate relationships can be used: 0.432 and 0.123 mmols1 mg for GLS and GLSbreakdown products, respectively.
2.3.2. Measurement of the hormonal status
Ž .2.3.2.1. Assay for thyroid hormones in plasma. Plasma thyroid hormone T and T3 4Ž .levels were measured with a radioimmunoassay RIA described by Boeuf and Prunet
Ž . Ž .1985 , and modified by Martinez et al. 1995 . The percentage of specific bindingobtained was 37 for T and 25 for T . The percentage of unspecific binding was 10 for4 3
T and 9 for T . The intra-assay coefficient of variation was 10% for both T and T .4 3 4 3
( )C. Burel et al.rAquaculture 188 2000 363–382 371
Ž .2.3.2.2. Deiodinase assay. Mol et al. 1998 characterised three types of deiodinaseŽ .activities in turbot D1, D2 and D3 and determined that liver, brain, heart and kidney
contained the highest level of enzymatic activity. Deiodinase activity was determined inŽ .microsomal fractions from pooled tissue samples =4 , prepared as described in Burel
Ž . Ž .et al. 1998 . Following the method described earlier by Mol et al. 1998 , the D1activity was measured in liver, kidney and heart, D2 in liver, and D3 activity in brainŽ . Ž .ns3 . The amount of microsomal Mx protein used was 1 mgrml for D1 in kidneyand heart, 0.25 mgrml for D1 in liver and D3 in brain, and 0.5 mgrml for D2 liver. Thesubstrate-cofactor solution contained 0.1 mM of rT for D1, 1 nM of T for D2 and 103 4
Ž125 .nM of T for D3, plus 0.1 mCi of the preferred radioactive substrate I and DTT as3Žcofactor 20 mM for D1 in heart and kidney; 50 mM for D1 in liver; 40 mM for D2 in
.liver; 30 mM for D3 in brain . The incubation was conducted as follows: 1 h at 308C forliver and kidney D1 and brain D3, 1 h at 378C for heart D1 and 1 h at 258C for liver D2.
Table 6Growth, feeding and feed utilisation, and final body composition of turbot fed to satiation the experimentaldiets during 9 weeks
Ž .Superscript letters indicate intergroup statistical differences. Means ns2 with no common letter areŽ .significantly different One way ANOVA and Tukey’s multiple range test; P -0.05 . For each row, the MSE
between groups and within groups is given. The growth and feeding parameters were calculated as describedin Table 5. The initial whole body sample contained: 21.7% of dry matter, 69.3% DM of protein, 6.1% DM offat, 20.6 kJrg DM of energy, 17.5% DM of ash and 2.9% DM of phosphorus.
Control L-30 L-50 R1–30 R1–46 R2–30 MSE between MSE withingroups groups
Growth and feeding parametersInitial body 48.6 47.2 50.2 49.6 48.7 47.1 3.12 0.83
Ž . Ž .weight IBW gbc c bc abc a abFinal body 122.9 127.5 122.5 113.5 102.2 106.2 206 23
Ž . Ž .weight FBW gcd d bcd abc a abDaily growth 2.13 2.29 2.06 1.88 1.65 1.81 0.107 0.013
Ž .index DGIb ab a a a aVoluntary feed 1.37 1.27 1.22 1.25 1.23 1.23 0.009 0.003
Ž .intake VFIab c bc ab a abFeed efficiency 1.02 1.17 1.10 1.01 0.93 1.01 0.013 0.001
( )C. Burel et al.rAquaculture 188 2000 363–382372
The deiodinase activity is expressed as femtomol of substrate converted per mg proteinper minute.
2.4. Analysis of data
Data are presented as means and for each parameter studied, the mean squared errorŽ .MSE between groups and within groups is given. To test the effect of dietarytreatments, data were subjected to a one-way analysis of variance, and when appropriate,means were compared by the Tukey’s multiple range test. Pearson’s correlation coeffi-
Ž .cients r ; P-0.05 were used to determine the relationship between the differentS
results obtained.
3. Results
3.1. Growth performance, feed intake and feed utilisation
Ž . Ž .The use of extruded lupin L , irrespective of its level of incorporation 30% or 50% ,Žled to growth performance similar to that obtained with the control diet Table 6, Fig.
.1A . A transient decrease in feed intake with the increase in incorporation of lupin was
Ž . Ž . Ž .Fig. 1. Changes over time of the daily growth index A , voluntary feed intake B and feed efficiency C inŽ .juvenile turbot fed the experimental diets over 9 weeks. Means "standard error, ns2 with no common
Ž .letter for each period are significantly different P -0.05, One way ANOVA, Tukey’s multiple range test .ABW means average body weight.
( )C. Burel et al.rAquaculture 188 2000 363–382 373
Ž .observed 7% and 11% lower for 30% and 50% of lupin, respectively , but turbotŽ .became accustomed after 3 weeks Fig. 1B . The incorporation of lupin into the diet,
Žsignificantly improved feed efficiency 15% and 8% higher for 30% and 50% of lupin,.respectively over the first 6 weeks. Subsequently, the feed efficiency in turbot fed the
lupin-based diets became similar to the control, and the value found with 50% of lupinŽ .was significantly lower than with 30% Fig. 1C . Final body composition of turbot was
Ž .not affected by incorporation of lupin Table 6 . However, protein, energy and phospho-rus retention were significantly higher than the control in turbot fed the diet containing30% of lupin, and energy and phosphorus retention were also significantly higher in
Ž .those fish fed the diet containing 50% of lupin Table 7 . A reduction of nitrogen andphosphorus intakes was accompanied by a reduction of nitrogen and phosphorus losses
Ž .for each level of lupin incorporation Table 7 .Incorporation of the low GLS RM1 at the level of 30% did not lead to any significant
decrease in growth performance of turbot. Nevertheless, at the level of incorporation of46%, the final body weight of turbot was 17% lower compared to the control group.When high GLS RM2 was used at the level of 30%, the decrease in growth wasintermediate compared to the two other cases with a final body weight 13.5% lower than
Table 7Ž . Ž .Nutrient retention and nitrogen N and phosphorus P balances of turbot fed to satiation the experimental
diets during 9 weeksŽ .Superscript letters indicate intergroup statistical differences. Means ns2 with no common letter are
Ž .significantly different One way ANOVA and Tukey’s multiple range test, P -0.05 . For each row, the MSEbetween groups and within groups is given.IBW and FBW are initial and final body weight, respectively.
Control L-30 L-50 R1–30 R1–46 R2–30 MSE between MSE withingroups groups
1( )Retention % of intakeab c b ab a aProtein 30.4 35.7 32.8 30.9 28.3 29.4 14.15 1.26
retentiona b b a a aEnergy 25.4 31.8 31.6 26.7 23.9 25.1 23.42 1.74
retentiona b b a a aPhosphorus 32.9 45.2 44.4 31.9 35.7 33.0 73.39 12.07
retention
( )N and P balance mgrkg per dayb a a a a2 aN intake 1110 1026 986 996 970 999 5035 1839bc bc ab ab3 c aN gain 338 367 323 308 274 293 2167 382
4 b a a a a abN total loss 772 660 662 688 696 705 3332 871b a a b b b2P intake 232 184 162 226 212 220 1521 81
3P gain 77 83 72 72 76 73 35 29b a a b b b4P total loss 156 101 90 154 137 147 1630 129
1 Ž . ŽRetention % s100= FBW=final carcass nutrient or energy contentyIBW=initial carcass nutrient or.energy content rnutrient or energy intake.
2 ŽŽ . .Daily intakesnutrient intaker IBWqFBW r2 rduration.3 Ž . ŽŽDaily gains Final whole body nutrient contenty initial whole body nutrient .content r IBWq. .FBW r2 rduration.
4 Total losssDaily nutrient intakeyDaily nutrient gain.
( )C. Burel et al.rAquaculture 188 2000 363–382374
Ž .that of the control group. Feed intake was significantly reduced around 10% in eachŽ . Ž .case Table 6 , but considering changes over time Fig. 1B , feed intake of turbot fed the
RM-based diet became similar to the control group after almost 3 weeks of feeding.Feed efficiency was not significantly affected. Final body composition remained un-
Ž . Ž .changed Table 6 , as well as protein, phosphorus and energy retention Table 7 .
Ž . Ž . Ž .Fig. 2. Plasma concentrations of triiodothyronine T and thyroxin T A , and deiodinase activities, i.e. D13 4Ž . Ž . Ž .activity degrading the rT in liver left axis , in kidney and heart right axis B , D2 activity converting the T3 4
Ž . Ž .into T in liver C , and D3 activity degrading the T in brain D , measured in blood and tissues samples of3 3Žturbot after 9 weeks of feeding for each dietary treatment. Means are given with standard deviation ns24 for
.plasma hormone levels and ns6 for deiodinase activities, tissues pooled=4 . Means with no common letterŽ .are significantly different P -0.05, One-Way ANOVA and Tukey’s multiple range analysis .
( )C. Burel et al.rAquaculture 188 2000 363–382 375
Phosphorus balance was not affected, but a lower nitrogen intake was accompanied by alower nitrogen gain and a lower nitrogen loss, and this was especially significant with46% of RM.
3.2. Hormonal status
Ž .The ingestion of extruded lupin L by turbot, irrespective of its dietary incorporationlevel had no significant effect on plasma T levels after 67 days of feeding, whereas4
plasma T levels were 50% and 25% higher than the control with 30% and 50% of lupin3Ž . Ž .incorporation, respectively Fig. 2A . The in vitro degradation of rT D1 activity3
increased significantly by twofold in liver of fish fed the diet containing 30% and 50%Ž .of lupin and by 1.5 in kidney of fish fed the diet containing 50% of lupin Fig. 2B .
On the contrary, when turbot were fed diets containing heat-treated RM1, theirplasma T levels were significantly reduced: a drop of 40% and 60% with 30% and 46%4
of RM1, respectively, compared to the control, but the plasma T levels were not3Ž . Ž .affected Fig. 2A . The in vitro conversion of T into T D2 activity was strongly4 3
stimulated in liver; an increase of 3.5- and 9-fold in turbot fed the diet containing 30%Ž .and 46% of RM1, respectively, compared to the control Fig. 2C . The in vitro
Ž .degradation of rT D1 activity was also stimulated in liver, being two- and fivefold3Ž . Ž .higher for 30% and 46% of RM1 Fig. 2B . The in vitro degradation of T D3 activity3
in brain was not significantly affected by RM1 incorporation compared to the control,Ž .irrespective of its dietary level Fig. 2D . However, this activity was significantly lower
in fish fed the diet with 46% of RM1 compared to those fed the diet with only 30%.When 30% of RM2 was incorporated in the diet, neither T nor T levels in plasma of4 3
Ž .turbot were affected Fig. 2A , but the in vitro D1 activity was significantly affected; anincrease of threefold of the degradation of rT in liver, and a decrease of twofold in the3
Ž .heart Fig. 2B .
Ž . Ž . Ž .Fig. 3. Daily growth index DGI measured in fish over 9 weeks, related to plasma levels in T A and T B4 3
measured in fish after 9 weeks of feeding for each dietary treatment. Each symbol represents one replicate.The symbol represented under brackets in B is excluded from the correlation analysis. The DGI was calculatedfrom the biomass data of each replicate and the plasma T and T data correspond to the mean of measures in4 3
Ž .12 fish per replicate. The Pearson’s correlation coefficients r are significant with P -0.05.S
( )C. Burel et al.rAquaculture 188 2000 363–382376
A significant correlation was found between plasma T levels measured after 9 weeks4Ž . Ž .of feeding and the daily growth index DGI of turbot Fig. 3A . A significant
correlation between plasma T levels and DGI was found only if one group of turbot fed3Ž .the control diet was excluded from the analysis Fig. 3B .
4. Discussion
4.1. Extruded lupin
The findings of this study indicate that extruded lupin can be successfully incorpo-rated up to a level of 50% in the diets for turbot, despite its apparent deficiency in lysineand methionine. This high level of incorporation allowed for a 25% reduction of the fishmeal content in the diet, without any deleterious effects on the growth and the bodycomposition of turbot. The use of extruded lupin could be economically beneficial interms of fish production because of the high nutritional value of lupin based diets,demonstrated by high feed efficiency and high protein and energy retention. Moreover,our results showed that the nitrogen and phosphorus losses per unit production werereduced by 14 and 40% respectively, compared to those of the control fish. Thisreduction was caused by both a reduction of intake of these compounds and animprovement of their retention. The practical use of extruded lupin in feeds for farmed
Žfish may reduce effluent loads. The low phosphorus content of lupin 0.6 compared to.2.1% DM total P in fish meal did not lead to a deficiency in turbot when incorporated
up to the 50% level.The maximal level of incorporation of lupin in the diet of juvenile turbot found here
is higher than in previous studies in rainbow trout where the maximum was determinedŽbetween 20% and 40% De la Higuera et al., 1988; Hughes, 1988, 1991; Gomes and
.Kaushik, 1989; Moyano et al., 1992; Bangoula et al., 1993; Gouveia et al., 1993 , but inŽ .accordance with the results of our earlier study in trout Burel et al., 1998 , using the
Ž .same extruded lupin. According to our supplier UNIP , the lupin used was a «sweet»Ž .variety with a low content in alkaloids -0.02%DM , antinutritional factor affecting the
palatability of the diet. In addition, the favourable effect of the extrusion process on theŽdigestibility of dry matter has been shown with other vegetal ingredients Kaushik,
.1989 . These two facts may explain the high feed efficiency found here and the goodacceptability of this ingredient, in contrast to the earlier results of De la Higuera et al.Ž . Ž .1988 and Gomes and Kaushik 1989 . Indeed, turbot may be more susceptible to the
Ž .palatability problem of lupin than rainbow trout Burel et al., 1998 , but they seemed tobecome accustomed to the diets after 3–4 weeks.
The increase in plasma T level and of in vitro deiodinase D1 activity in liver and3
kidney of turbot fed the lupin-based diets indicates that the production of T was3
probably increased in vivo. A stimulation of T formation has already been demon-3Ž .strated in rainbow trout fed similar diets Burel et al., 1998 . It has been proposed that
carbohydrates are an important regulator of T secretion, and protein serves as the signal4Ž .to activate T formation Eales et al., 1993; MacKenzie et al., 1998 . In our case, diets3
were isonitrogenous, but the different dietary amino acid profile in the lupin-based diet
( )C. Burel et al.rAquaculture 188 2000 363–382 377
may have been responsible for the stimulation of the T production. The high plasma T3 3
levels were accompanied by high growth rates in turbot fed lupin-based diets especiallyat the 30% incorporation level. It has been known for some time that thyroid hormones
Žin conjunction with other hormones may enhance growth of fish Narayansingh and.Eales, 1975; Donaldson et al., 1978 and some authors have shown a close relation
Žbetween their growth rate and plasma thyroid hormone levels Eales and Shostak, 1985;.Boeuf and Gaignon, 1989; McCormick and Saunders, 1990; Gomez et al., 1997 .
However, in most of these studies as in the present study, a causal relationship betweenplasma thyroid hormone levels and growth rates is difficult to establish.
4.2. RMs
The results of the present study indicate that RM has a lower potential than extrudedlupin as a fish meal substitute in the feed of juvenile turbot. However, the heat-treatedRM, can be incorporated in the diets of turbot up to a level of 30%, without significanteffect on growth performance or body composition, despite a lower feed intakecompared to the control fish. At the 46% incorporation level, or at 30% with theuntreated RM, growth was depressed. This depression was due mainly to a decrease infeed intake. A lower feed intake of the RM-based diets, attributed to the presence ofprogoitrine, or sinapine or tannins, which all affect the palatability of the diet, have been
Ž .previously found in terrestrial animals reviewed by Mawson et al., 1993 , as well as inŽ .rainbow trout Hilton and Slinger, 1986 . However, it is in disagreement with the results
Ž .obtained in our previous study in rainbow trout using the same ingredients Burel, 1999as shown in Table 8. The use of new cultivars obtained by genetic selection andror thenew technological treatments of RM can improve the acceptability of this ingredient, butturbot appear more susceptible than rainbow trout to the organoleptic quality of diets.However, they seemed to become accustomed to RM-based diets after 3–4 weeks. Noeffect on feed efficiency was found in turbot, contrary to the findings of earlier studies
Žin rainbow trout Yurkowski et al., 1978; Leatherland et al., 1987; McCurdy and March,. Ž .1992; Gomes et al., 1993; Burel, 1999 , Atlantic salmon Higgs et al., 1982 , channel
Ž . Ž . Žcatfish Webster et al., 1997 , tilapia Jackson et al., 1982 and sea bream Haiqing and.Xiqin, 1994 . This may be attributed to the higher proportion of fish meal in the
RM-based diets used with turbot compared to the diets used for the other fish species.In turbot fed the diets containing the heat-treated RM, lower plasma T levels were4
observed, as well as changes in the deiodinase activities, compared to the control fish,which were more pronounced at the 46% incorporation level than at the 30% level.However, T levels remained unchanged in the two cases. Although the plasma T3 3
levels did not decline, it would be worth assessing if these thyroid disturbances do notaffect growth performance over a long term in the case of the lower level of incorpora-tion. When the RM without heat treatment was incorporated at the level of 30%, neitherT nor T concentrations in plasma were affected, and only an effect on the deiodinase4 3
activities was observed. The present study showed a strong dose–response relationshipbetween the amount of toxic compounds and plasma T levels as previously shown in4
Ž .rainbow trout Yurkowski et al., 1978; Burel, 1999 and the apparent lack of goitrogenic
( )C. Burel et al.rAquaculture 188 2000 363–382378
effect of the untreated RM may be attributed to the absence of GLS breakdownproducts. However, in the absence of denaturation of myrosinase in untreated RM, ahydrolysis of the GLS into toxic compounds seems to be possible in the intestinal tractof turbot. In general, when the amount of GLSrGLS breakdown products are taken intoaccount, the thyroidal disturbances observed in turbot were less marked than those in
Žfish reared in freshwater Yurkowski et al., 1978; Higgs et al., 1982; Hardy and.Sullivan, 1983; Hossain and Jauncey, 1988; Teskeredzic et al., 1995; Burel, 1999
Ž .Table 8 . These findings suggest that the iodine from seawater possibly reduces thegoitrogenic effect of RM in the same way as dietary iodine supplementation in mammalsŽ .Anke et al., 1980; Ludke et al., 1985; Vermorel and Evrard, 1987 and rainbow troutŽ .Burel, 1999 . Indeed, thiocyanate anions are competitors for iodine in thyroidmetabolism, but their effect can be reduced by iodine supply. In rainbow trout, the effectof a dietary iodine supplementation or of the iodine from seawater was higher with
Ž .untreated RM than with the heat-treated one Burel, 1999 . This difference wasŽattributed to a difference in the profile of GLS breakdown products a higher proportion
.of thiocyanate anions , and could also be an explanation for the lack of goitrogeniceffect of this RM meal in turbot.
The in vitro response of the deiodinase activities to the lower release of T in turbot4
fed the heat-treated RM based diets was mainly a stimulation of the D1 and D2 activities
Table 8Ž .Comparison of the effect of the dietary incorporation of RM RM1 or RM2 in turbot and rainbow trout
Data about turbot originate in the present study and data about rainbow trout originate in our previous studiesŽ .Burel, 1999 , using the same RMs.IBW and FBW are initial and final body weight, respectively.
Turbot Rainbow trout1Plasma T levels affected already at 3.6 mmolrg . . . 1.5 mmolrg4
Plasma T levels not affected at least up to 4.4 mmolrg affected already at3
3.7 mmolrgDeiodinase efficient up to at least 4.4 mmolrg . . . under 3.7 mmolrgcompensatory activities:
2D1 activity affected already at a concentration nil not affected up to atleast 17.4 mmolrg
4Growth performance affected already at 46% of . . . 30% of RM1 orRM1 and at 30% of RM2 RM2
Feed intake affected already at 30% of RM not affected at leastup to 50% of RM
Feed efficiency not affected up to at least 46% of affected already atRM1 and 30% of RM2 30% of RM
1mmol of GLS derivativesrg of diet.
2 In the case of the untreated RM, with the myrosinase not inactivated. A hydrolysis of the dietary GLS,and then a production of goitrogenic derivatives, is possible in the intestinal tract of fish.
3Not significant compared to in control fish but significant comparing to the other RM-based diets.4 Percent of incorporation of RM into the diets.
( )C. Burel et al.rAquaculture 188 2000 363–382 379
in liver, suggesting an in vivo increased production of T . Similar changes in deiodinase3
activities succeeding in effectively adjusting the plasma T levels were previously3Ž .observed in rainbow trout Burel, 1999 , but only up to 20% of incorporation of RM,
Ž .corresponding to a lower dietary content of toxic compounds Table 8 . It is interestingto note that in turbot, this adjustment of the deiodinase activities was not observed in alltissues tested, since the activity of D1 in kidney and heart were not affected in fish fedthe heat-treated RM-based diet. Moreover, the changes observed in vitro in the activityof D1 in the case of turbot fed the untreated one do not seem logical with a stimulationin liver and an inhibition in heart.
Taking into account all the dietary treatments, including the fish fed the lupin-baseddiets, a significant correlation between plasma T levels and growth of turbot was4
observed. In the case of T , the biologically active hormone, a significant correlation3
was found only if one group of turbot fed the control diet was excluded from theanalysis. As average, these fish expressed low plasma T levels and no clear explanation3
Ž .can be proposed about that fact. Because of the few number of replicates ns2 , it isnot possible to claim that this value was an outliner and that growth depression wasrelated to low plasma T levels in turbot fed the RM-based diets. Moreover, in the case3
of fish fed the untreated RM-based diet, growth was significantly depressed, but neitherthe T nor the T plasma levels were significantly lower compared to the control, even if4 3
the ‘‘outliner control group’’ was excluded from the comparison. These findings suggestthat in turbot, it was the presence of ANFs altering the palatability of the RM diet thatwas responsible for the lower growth performance observed, rather than the thyroiddisturbances. Nevertheless it seems that if the heat treatment favoured involuntarily GLSdegradation into toxic breakdown products, it had a beneficial effect on the quality ofRM.
In conclusion, the present study indicates that extruded lupin can be incorporated inthe diets of turbot up to a level of 50% without affecting growth performance and fishquality. In addition, its use could be beneficial in terms of abatement of the pollutiongenerated by turbot farms, owing to a decrease of nitrogen and phosphorus losses.Heat-treated RM must be limited to a dietary level of 30%. Turbot seems to be lesssusceptible to the goitrogenic compounds contained in RM than rainbow trout. Themarine environment of turbot, richer in iodine, probably favours the reduction of thedeleterious effect of thiocyanate anions. The plasma T levels were reduced at a minimal4
dietary content of toxic compounds of 3.6 mmolrg. However, up to at least 4.4mmolrg, the plasma T levels were not affected, due to the deiodinase compensatory3
effect.
Acknowledgements
This work was partially supported by CETIOM and by UNIP. The authors gratefullyacknowledge the skillful technical assistance of D. Blanc, J. Breque, H. Le Delliou, A.`Le Roux, L. Noterdaeme, P. Peyrotte, L. Quemeneur, A. Severe, C. Vachot, W. Van´ `Ham, and F. Voets.
( )C. Burel et al.rAquaculture 188 2000 363–382380
References
Abdou Dade, B., Aguirre, P., Blanc, D., Kaushik, S.J., 1990. Incorporation du colza 00 sous forme de tourteauŽ .ou d’amande dans les aliments de la truite arc-en-ciel Oncorhynchus mykiss : performance zootechnique
et digestibilite. Bull. Fr. Peche Piscic. 317, 50–57.´ ˆAdron, J.W., Blair, A., Cowey, C.B., Shanks, A.M., 1976. Effects of dietary energy level and dietary energy
Ž .source on growth, feed conversion and body composition of turbot Scophthalmus maximus L. . Aquacul-ture 7, 125–132.
AFNOR, 1980. Association Franßaise de Normalisation. Vol. 18–106. Aliments des Animaux-Dosage duŽphosphore total-Methode spectrophotometrique Animal feeding stuffs-Determination of total phosphorus´ ´
Ž .Andersen, N.G., Alsted, N.S., 1993. Growth and body composition of turbot Scophthalmus maximus L. inŽ .relation to different lipidrprotein ratios in the diet. In: Kaushik, S., Luquet, P. Eds. , Fish Nutrition in
Ž .Practice, Biarritz France , 1991, June 24–27. INRA, Paris, pp. 479–491.Anke, M., Schwarz, S., Hennig, A., Groppel, B., Grun, M., Zenker, G., Glos, S., 1980. Der einfluss
zusachtlicher zink-und jodgaben auf rapsextraktionsschrotbedingte schaden beim schwein. Monatsh.¨ ¨Veterinarmed 35, 90–94.¨
AOAC, 1995. Official Methods of Analysis of AOAC International. Vol. I. Agriculture Chemicals; Contami-nants, Drugs. 16th edn. AOAC International, Arlington, VA.
Ž .Bangoula, D., Parent, J.P., Vellas, F., 1993. Nutritive value of white lupin Lupinus albus var Lutop fed toŽ .rainbow trout Oncorhynchus mykiss . Effects of extrusion cooking. Reprod. Nutr. Dev. 33, 325–334.
Boeuf, G., Gaignon, J.L., 1989. Effects of rearing conditions on growth and thyroid hormones during smoltingof atlantic salmon Salmo salar L. Aquaculture 45, 29–38.
Ž q q .Boeuf, G., Prunet, P., 1985. Measurements of gill Na –K ATPase activity and plasma thyroid hormonesŽ .during smoltification in Atlantic salmon Salmo salar L. . Aquaculture 45, 111–119.
Burel, C., 1999. Utilisation de proteines d’origine vegetale dans l’alimentation de la truite arc-en-ciel´ ´ ´Ž . Ž . ŽOncorhynchus mykiss et du turbot Psetta maxima : valeur nutritionelle et effets sur l’axe thyroıdien Use¨
.of plant-protein in the feed of rainbow trout and turbot: nutritional value and effects on thyroid axis . PhDthesis, University of Rennes 1, France.
Burel, C., Boujard, T., Corraze, G., Kaushik, S.J., Boeuf, G., Mol, K.A., Van Der Geyten, S., Kuhn, E.R.,¨Ž .1998. Incorporation of high levels of extruded lupin in diets for rainbow trout Oncorhynchus mykiss :
nutritional value and effect on thyroid status. Aquaculture 163, 325–345.ŽDanielssen, D.S., Hjertnes, T., 1993. Effect of dietary protein levels in diets for turbot Scophthalmus
. Ž . Ž .maximus to market size. In: Kaushik, S., Luquet, P. Eds. , Fish Nutrition in Practice, Biarritz France ,1991, June 24–27. INRA, Paris, pp. 89–96.
De la Higuera, M., Garcia-Gallego, M., Sanz, A., Cardenete, G., Suarez, M.D., Moyano, F.J., 1988.ŽEvaluation of lupin seed meal as an alternative protein source in feeding of rainbow trout Salmo
.gairdneri . Aquaculture 71, 37–50.Donaldson, E.M., Fagerlund, U.H.M., Higgs, D.A., McBride, J.R., 1978. Hormonal enhancement of growth in
Ž .fish. In: Hoar, W.S., Randall, D.J. Eds. , Fish Physiology, Bioenergetics and Growth Vol. 8 AcademicPress, London, pp. 455–580.
Eales, J.G., MacLatchy, D.L., Sweeting, R.M., 1993. Thyroid hormone deiodinase systems in salmonids, andtheir involvement in the regulation of thyroid status. Fish Physiol. Biochem. 11, 313–321.
Eales, J.G., Shostak, S., 1985. Correlations between food ration, somatic growth parameters and thyroidfunction in Artic charr, SalÕelinus alpinus L. Comp. Biochem. Physiol. 80A, 553–558.
Ž .Ferrando, R., Blum, J.C. Eds. , L’Alimentation des Animaux Monogastriques: Porc, Lapin, Volaille.w xNutrition of Monogastric Animals: Pig, Rabbit, Poultry INRA, Paris, France, 293 pp.
Gardner, W.S., Miller, W.H., 1980. Reverse-phase liquid chromatography of amino acids after reaction withOphthalaldehyde. Anal. Biochem. 101, 61–70.
Gomes, E.F., Corraze, G., Kaushik, S., 1993. Effects of dietary incorporation of a co-extruded plant proteinŽ .rapeseed and peas on growth, nutrient utilization and muscle fatty acid composition of rainbow troutŽ .Oncorhynchus mykiss . Aquaculture 113, 339–353.
( )C. Burel et al.rAquaculture 188 2000 363–382 381
Gomes, E.F., Kaushik, S.J., 1989. Incorporation of lupin seed meal, colzapro or triticale as proteinrenergyŽ .substitutes in rainbow trout diets. In: Takeda, M., Watanabe, T. Eds. , The Current Status of Fish
Nutrition in Aquaculture, Proceedings of the Third International Symposium on Feeding and Nutrition inFish, 28 August–1 September 1989, Toba, Japan. pp. 315–324.
Gomez, J.M., Boujard, T., Boeuf, G., Solari, A., Le Bail, P.-Y., 1997. Individual diurnal plasma profiles ofŽ .thyroid hormones in rainbow trout Oncorhynchus mykiss in relation to cortisol, growth hormone, and
growth rate. Gen. Comp. Endocrinol. 107, 74–83.Gouveia, A., Oliva-Teles, A., Gomes, E., Rema, P., 1993. Effect of cookingrexpansion of three legume seeds
Ž .on growth and food utilization by rainbow trout. In: Kaushik, S.J., Luquet, P. Eds. , Fish Nutrition inPractice, June 24–27 1991, Biarritz, France. INRA, Paris, pp. 24–27.
Guillaume, J., Coustans, M.-F., Metailler, R., Person-Le Ruyet, J., Robin, J., 1991. Flatfish, turbot, sole, and´Ž .plaice. In: Wilson, R.P. Ed. , Handbook of Nutrient Requirements of Finfish. CRC Press, Boston, pp.
77–82.Haiqing, S., Xiqin, H., 1994. Effects of dietary animal and plant protein ratios and energy levels on growth
Ž .and body composition of bream Megalobrama skolkoÕii Dybowski fingerlings. Aquaculture 127,189–196.
Higgs, D.A., McBride, J.R., Markert, J.R., Dosanjh, B.S., Plotnikoff, M.D., Clarke, W.C., 1982. Evaluation ofŽ .Tower and Candle rapeseed canola meal and Bronowski rapeseed protein concentrate as protein
Ž .supplements in practical dry diets for juvenile chinook salmon Oncorhynchus tshawytscha . Aquaculture29, 1–31.
Hilton, J.W., Slinger, S.J., 1986. Digestibility and utilization of canola meal in practical-type diets for rainbowŽ .trout Salmo gairdneri . Can. J. Fish. Aquat. Sci. 43, 1149–1155.
Ž . ŽHossain, M.A., Jauncey, K., 1988. Toxic effects of glucosinolate allyl isothiocyanate synthetic and from.mustard oil cake on growth and food utilization in common carp. Indian J. Fish. 35, 186–196.
Ž .Hughes, S.G., 1988. Assesment of lupin flour as diet ingredient for rainbow trout Salmo gairdneri .Aquaculture 71, 379–385.
ŽHughes, S.G., 1991. Use of lupin as a replacement for full-fat soy in diets for rainbow trout Oncorhynchus.mykiss . Aquaculture 93, 57–62.
Jackson, A.J., Capper, B.S., Matty, A.J., 1982. Evaluation of some plant proteins in complete diets for thetilapia Sarotherodon mossambicus. Aquaculture 27, 97–109.
Kaushik, S.J., 1989. Use of alternative protein sources for intensive rearing of carnivorous fishes. In: Flos, R.,Ž .Tort, L., Torres, P. Eds. , Mediterranean Aquaculture, 1989. Ellis Horwood, Chichester, pp. 125–138.
Ž .Kaushik, S.J., 1998. Whole body amino acid composition of European seabass Dicentrarchus labrax ,Ž . Ž .gilthead seabream Sparus aurata and turbot Psetta maxima with an estimation of their IAA needs.
Aquat. Living Resour. 11, 351–358.Leatherland, J.F., Hilton, J.W., Slinger, S.J., 1987. Effects of thyroid hormone supplementation of canola
Žmeal-based diets on growth, and interrenal and thyroid gland physiology of rainbow trout Salmo.gairdneri . Fish Physiol. Biochem. 3, 73–82.
Ludke, H., Schone, F., Hennig, A., 1985. Der Einfluss von Jodkupfer-und Zink-Zulagen zu Rationen mithohem Rapsextraktionsschrotanteil auf Wachstum und Schildrussenfunktion des Mastschweines. 1. Ein-¨
wfluss auf die Mastleistung. Influence of iodine–copper and zinc supplementations to rations with a hightxlevel of rapeseed meal on growth and thyroid function of growing pigs Arch. Tierernahr., Berlin 35,¨
835–845.MacKenzie, D.S., Moore VanPutte, C., Leiner, K.A., 1998. Nutrient regulation of endocrine function in fish.
Aquaculture 161, 3–25.Martinez, I., Dreyer, B., Argersborg, A., Le Roux, A., Boeuf, G., 1995. Effect of triiodothyronine and rearing
temperature on growth and skeletal myosin heavy chain isoform transition during early development inŽ .salmonid SalÕelinus alpinus L. . Comp. Biochem. Physiol. 112B, 717–725.
Mawson, R., Heaney, R.K., Zdunczyk, Z., Kozlowska, H., 1993. Rapeseed meal — glucosinolates and theirantinutritional effects: Part 1. Rapeseed production and chemistry of glucosinolates. Die Nahrung 37,131–140.
Mawson, R., Heaney, R.K., Zdunczyk, Z., Kozlowska, H., 1994. Rapeseed meal — glucosinolates and their
( )C. Burel et al.rAquaculture 188 2000 363–382382
antinutritional effects: Part 4. Goitrogenicity and internal organs abnormalities in animals. Die Nahrung 38,178–191.
McCormick, S.D., Saunders, R.L., 1990. Influence of ration level and salinity on circulating thyroid hormonesŽ .in juvenile Atlantic salmon Salmo salar . Gen. Comp. Endocrinol. 78, 224–230.
McCurdy, S.M., March, B.E., 1992. Processing of canola meal for incorporation in trout and salmon diets.JAOCS 69, 213–220.
Mol, K., Van Der Geyten, S., Burel, C., Darras, V.M., Boujard, T., Kuhn, E.R., 1998. Comparative study of¨iodothyronine outer ring and inner ring deiodinases activities in five teleostean fishes. Fish Physiol.Biochem. 18, 253–266.
Moyano, F.J., Cardenete, G., De la Higuera, M., 1992. Nutritive values of diets containing a high percentageof vegetable proteins for trout, Oncorhynchus mykiss. Aquat. Living Resour. 5, 23–29.
Narayansingh, T., Eales, J.G., 1975. The influence of physiological doses of thyroxine on the lipid reserves ofŽ .starved and fed brook trout, SalÕelinus fontinalis Mitchill . Comp. Biochem. Physiol. 52B, 407–412.
Ž .National Research Council NRC , 1993. Nutrient Requirements of Fish. National Academy Press, WA, DC.Person-Le Ruyet, J., Baudin-Laurencin, F., Devauchelle, N., Metailler, R., Nicolas, J.-L., Robin, J., Guil-´
Ž . Ž .laume, J., 1991. Culture of turbot Scophthalmus maximus . In: McVey, J.P. Ed. , CRC Handbook ofMariculture Vol. 2 CRC Press, Boston, pp. 21–40.
Quinsac, A., Ribaillier, D., Elfakir, C., Lafosse, M., Dreux, M., 1991. A new approach to the study ofglucosinolates by isocratic liquid chromatography: Part I. Rapid determination of desulfated derivatives ofrapeseed glucosinolates. J. Assoc. Off. Anal. Chem. 74, 932–939.
M., Vaara, T., Bridges, R.B., 1995. Assessment of undephytinized and dephytinized rapeseed proteinconcentrate as sources of dietary protein concentrate as sources of dietary protein for juvenile rainbow
Ž .trout Oncorhynchus mykiss . Aquaculture 131, 261–277.Thivend, P., Mercier, C., Guilbot, A., 1972. Determination of starch with glucoamylase. In: Whistler, R.L.,
Ž .Bemiller, J.N. Eds. , Methods in Carbohydrate Chemistry Vol. IV Academic Press, New York, pp.100–105.
Vermorel, M., Evrard, J., 1987. Valorization of rapeseed meal: 4. Effects of iodine, copper and ferrous saltsupplementation in growing rats. Reprod. Nutr. Develop. 27, 769–779.´
Webster, C.D., Tiu, L.G., Tidwell, J.H., Grizzle, J.M., 1997. Growth and body composition of channel catfishŽ .Ictalurus punctatus fed diets containing various percentages of canola meal. Aquaculture 150, 103–112.
Yurkowski, M., Bailey, J.K., Evans, R.E., Tabachek, J.-A.L., Burton Ayles, G., 1978. Acceptability ofŽ .rapeseed proteins in diets of rainbow trout Salmo gairdneri . J. Fish. Res. Board Can. 35, 951–962.
