RAPESEED PROTEIN PRODUCTS AS FISH MEAL REPLACEMENT … · protein in fish feeds, is a limited...

Aus dem Institut für Tierzucht und Tierhaltung

der Agrar- und Ernährungswissenschaftlichen Fakultät

der Christian-Albrechts-Universität zu Kiel

___________________________________________________________________________

RAPESEED PROTEIN PRODUCTS AS FISH MEAL

REPLACEMENT IN FISH NUTRITION

Dissertation

zur Erlangung des Doktorgrades

der Agrar- und Ernährungswissenschaftlichen Fakultät

der Christian-Albrechts-Universität zu Kiel

vorgelegt von

Master of Science

HANNO SLAWSKI

aus Neustadt in Holstein

Kiel, 2011

___________________________________________________________________________

Dekanin: Prof. Dr. K. Schwarz

Erster Berichterstatter: Prof. Dr. C. Schulz

Zweiter Berichterstatter: Prof. Dr. A. Susenbeth

Tag der mündlichen Prüfung: 14.07.2011

___________________________________________________________________________

Die Dissertation wurde mit dankenswerter finanzieller Unterstützung aus dem Europäischen Fischereifond und

dem Zukunftsprogramm Fischerei des Landes Schleswig-Holsteins angefertigt

II

Gedruckt mit Genehmigung der Agrar- und Ernährungswissenschaftlichen Fakultät der

Christian-Albrechts-Universität zu Kiel

III

Table of Contents

General Introduction ................................................................................................................... 1 Chapter 1: Replacement of fish meal with rapeseed protein concentrate in diets fed to common carp (Cyprinus carpio L.) ............................................................................................ 4 Chapter 2: Replacement of fish meal with rapeseed protein concentrate in diets fed to wels catfish (Silurus glanis L.) ......................................................................................................... 16 Chapter 3: Austausch von Fischmehl durch Rapsproteinkonzentrat in Futtermitteln für Steinbutt (Psetta maxima L.) .................................................................................................... 34 Chapter 4: Total fish meal replacement with rapeseed protein concentrate in diets fed to rainbow trout (Oncorhynchus mykiss W.) ................................................................................ 47 Chapter 5: Replacement of fish meal with albumin and globulin rapeseed protein fractions in diets fed to rainbow trout (Oncorhynchus mykiss W.) ............................................................. 69 Chapter 6: Total fish meal replacement with canola protein isolate in diets fed to rainbow trout (Oncorhynchus mykiss W.) .............................................................................................. 85 General Discussion ................................................................................................................. 102 General Summary ................................................................................................................... 111 Zusammenfassung .................................................................................................................. 114 Danksagung ............................................................................................................................ 117 Lebenslauf .............................................................................................................................. 118

IV

List of Tables

Table 1.1 Proximate composition and amino acid profiles of fish meal and

rapeseed protein concentrate and concentration of antinutritional factors detected in RPC

9

Table 1.2 Formulation, amino acid profiles and proximate composition of experimental diets for common carp

10

Table 1.3 Growth response, feed efficiencies and survival of carp fed experimental diets

11

Table 1.4 Proximate whole body composition of carp fed the experimental diets

11

Table 2.1 Nutrient composition and essential amino acid profiles of fish meal and rapeseed protein concentrate and concentration of antinutritional factors detected in RPC

20

Table 2.2 Formulation, essential amino acids composition and proximate composition of experimental diets

21

Table 2.3 Growth response, feed intake, feed efficiencies, condition factor and survival of wels catfish fed experimental diets

23

Table 2.4 Proximate whole body composition of wels catfish fed the experimental diets

24

Table 2.5 Blood haematocrit content and blood serum values of wels catfish fed experimental diets

25

Table 3.1 Nährstoff- und Aminosäurenzusammensetzung von Fischmehl und Rapsproteinkonzentrat

37

Table 3.2 Formulierung der Versuchsfuttermittel

38

Table 3.3 Nährstoff- und Aminosäurenzusammensetzung der Versuchsfuttermittel

39

Table 3.4 Ganzkörperzusammensetzung der Steinbutt nach der Fütterungsperiode

41

Table 3.5 Wachstumsparameter und Futterverwertung der Steinbutt nach dem Fütterungsversuch

41

Table 4.1 Proximate and amino acid composition of fish meal and rapeseed protein concentrate and concentration of antinutritional factors determined

51

V

Table 4.2 Formulation of experimental diets

52

Table 4.3 Proximate and amino acid composition of experimental diets

53

Table 4.4 Growth response, feed efficiencies and survival of rainbow trout fed experimental diets

56

Table 4.5 Proximate whole body composition of rainbow trout fed experimental diets

56

Table 4.6 Blood parameters of trout fed experimental diets

57

Table 5.1 Nutrient composition and amino acid profiles of fish meal, albumin concentrate and globulin concentrate

68

Table 5.2 Formulation and nutrient composition and amino acid profiles of experimental diets used in the digestibility trial

69

Table 5.3 Formulation, proximate nutrient composition and amino acid composition of experimental diets for rainbow trout

72

Table 5.4 Apparent digestibility coefficients

73

Table 5.5 Growth performance, feed intake and feed efficiencies of rainbow trout fed experimental diets

74


75

Table 6.1 Nutrient composition and essential amino acid profiles of fish meal and canola protein isolate

83

Table 6.2 Formulation, nutrient composition and essential amino acid profiles of experimental diets used in the digestibility trial

84

Table 6.3 Formulation, proximate composition and essential amino acid profiles of experimental diets

86

Table 6.4 Apparent digestibility coefficients

88

Table 6.5 Growth response, feed intake and feed efficiencies of rainbow trout fed experimental diets

89


89

1

General Introduction

In 2009 aquaculture production hit a landmark: half of all fish and shellfish destined for

human consumption were cultured, and production of farmed fish eclipsed that of wild caught

fish. But, the increased aquaculture production also accounted for 68 % of the worldwide fish

meal consumption (Naylor et al. 2009). However, fish meal, the most important source of

protein in fish feeds, is a limited resource with an annual production volume between 5 to 6.5

Mio t (FAO 2004). Tremendous price increases for fish meal together with environmental

concerns therefore force the aquaculture sector to find alternative protein sources to be

included in fish feeds. Presently, most relevant alternatives are protein concentrates derived

from vegetables. Among them, soybean protein concentrates have become a commonly

accepted fish feed ingredient and fish meal alternative (Gatlin et al. 2007). While soybean

ranks as number one oilseed worldwide (222.2 Mio t/a), protein products derived from

rapeseed, which ranks as number three oilseed worldwide (61.6 Mio t/a) (FAO 2010), are less

commonly used as fish feed ingredients. However, simple oilcakes or rapeseed meals with

increased protein content produced from oilcakes that were de-oiled with organic solvents

have been widely tested as protein sources in feeding trials with several fish species.

Experiments with rainbow trout (Burel et al. 2000a,c; Shafaeipour et al. 2008), Nile tilapia

(Davies et al. 1990), common carp (Dabrowski and Kozlowska 1981) and turbot (Burel et al.

2000a,b) have shown, that the nutritional quality of simple rapeseed products is below that of

fish meal although they contained a well balanced amino acid profile. Particularly

antinutritional factors (ANF) determine the quality of rapeseed products for fish nutrition. The

most prominent ANF in rapeseed products are glucosinolates, phytic acid, phenolic

constituents and indigestible carbohydrates (Francis et al. 2001). By several processing

techniques the level of antinutrients in rapeseed products can be reduced and their value for

fish nutrition can be improved. Dehulling of seeds and utilisation of high temperatures and

organic solvents during oil extraction as well as sieving of meal decrease content of

glucosinolates, phytate, fibre, cellulose, hemicellulose, sinapin and tannins (Fenwick et al.

1986; Anderson-Haferman et al. 1993). Protein extraction from meals by methanol-ammonia-

treatment or ethanol-treatment will increase protein level and effectively remove

glucosinolates, phenolic compounds, soluble sugars, such as sucrose, and some

oligosaccharides (Naczk and Shahidi 1990; Chabanon et al. 2007). In different countries,

rapeseed protein products of high quality were produced for application in animal nutrition.

2

However, these products were made for test purposes in small volumes until their potential as

protein source in animal nutrition is clarified. Besides nutritive quality, their costs of

production must decrease to make rapeseed protein products available at a competitive price

compared to other protein sources, especially fish meal.

In the present study, different protein protein products derived from rapeseed (including

canola) were tested as fish meal replacement in diets for several fish species. A high quality

rapeseed protein concentrate (RPC) with a protein content of 71 % was evaluated as fish meal

replacement in diets for common carp (chapter 1), wels catfish (chapter 2), turbot (chapter 3)

and rainbow trout (chapter 4). Based on the results presented in chapter 4, in chapter 5 the

potential of two rapeseed protein concentrates partitioned in albumin and globulin fractions as

fish meal alternatives was evaluated in a digestibility study and a consecutive growth trial

with rainbow trout. Compared to the RPC, the fractionized protein concentrates were

produced under lower cost and time effort. In chapter 6 a canola protein isolate with a crude

protein content of 81 % was evaluated as fish meal alternative in diets for rainbow trout. The

nutritional quality of the raw material was determined in a digestibility experiment followed

by a growth trial.

References

Anderson-Hafermann, J.C., Zhang, Y., Parsons, C.M., 1993. Effects of processing on the

nutritional quality of canola meal. Poultry Science 72, 326-333.

Burel, C., Boujard, T., Tulli, F., Kaushik, S.J., 2000a. Digestibility of extruded peas, extruded

lupin, and rapeseed meal in rainbow trout (Oncorhynchus mykiss) and turbot (Psetta

maxima). Aquaculture 188, 285–298.

Burel, C., Boujard, T., Kaushik, S.J., Boeuf, G., van der Geyten, S., Mol, K.A., Kühn, E.R.,

Quinsac, A., Krouti, M., Ribaillier, D., 2000b. Potential of plant-protein sources as fish

meal substitutes in diets for turbot (Psetta maxima): growth, nutrient utilisation and

thyroid status. Aquaculture 188, 363-382.

Burel, C., Boujard, T., Escaffre, A.M., Kaushik, S.J., Boeuf, G., Mol, K., Van Der Geyten, S.,

Kühn, E.R., 2000c. Dietary low glucosinolate rapeseed meal affects thyroid status and

nutrient utilization in rainbow trout (Oncorhynchus mykiss). Brit. J. Nutr. 83, 653–664.

Chabanon, G., Chevalot, I., Framboisier, X., Chenu, S., Marc, I., 2007. Hydrolysis of

rapeseed protein isolates: Kinetics, characterization and functional properties of

hydrolysates. Process Biochem. 42, 1419–1428.

3

Dabrowski, K., Kozlowska, H., 1981. Rapeseed meal in the diet of common carp reared in

heated waters. I. Growth of fish and utilization of the diet. In: K. Tiews (ed.).

Aquaculture in Heated Effluents and Recirculation Systems. Heenemann, Hamburg, pp.

263-274.

Davies, S.J., McConnel, S., Bateson, R.I., 1990. Potential of rapeseed meal as an alternative

protein source in complete diets for tilapia (Oreochromis mossambicus Peters).

Aquaculture 87, 145-154.

Food and Agriculture Organization (FAO), United Nations (2004). FAO Fisheries

Department, Fishery Information, Data and Statistics Unit. Fishstat Plus: Universal

software for Fishery Statistical Time series, version 2.30 (www.fao.org).

FAO (2010): http://faostat.fao.org/site/567/default.aspx#ancor

Fenwick, G.R., Spinks, E.A., Wilkinson, A.P., Henry, R.K., Legoy, M.A., 1986. Effect of

processing on the antinutrient content of rapeseed. J. Sci. Food Agr. 37, 735-741.

Francis, G., Makkar, H.P.S., Becker, K., 2001. Antinutritional factors present in plant-derived

alternate fish feed ingredients and their effects in fish. Aquaculture 199, 197–227.

Gatlin III, D.M., Barrows, F.T., Brown, P., Dabrowski, K., Gaylord, T.G., Hardy, R.W.,

Herman, E., Hu, G., Krogdahl, Å., Nelson, R., Overturf, K., Rust, M., Sealey, W.,

Skonberg, D., Souza, E.J., Stone, D., Wilson, R., Wurtele, E., 2007. Expanding the

utilization of sustainable plant products in aquafeeds: a review. Aquaculture Research

38, 551-579.

Naczk, M., Shahidi, F., 1990. Carbohydrates of canola and rapeseed. In: F. Shahidi (ed.).

Canola, Rapeseed: Production, Chemistry, Nutrition & Processing Technology. Van

Nostrand Reinhold, New York, pp211-220.

Naylor, R.L., Hardy, R.W., Bureau, D.P., Chiu, A., Elliott, M., Farrell, A.P., Forster, I., Gatlin

III., D.M., Goldburg, R.J., Hua, K., Nichols, P.D., 2009. Feeding aquaculture in an era

of finite resources. PNAS 106, 15103-15110.

Shafaeipour, A., Yavari, V., Falahatkar, B., Maremmazi, J.G.H., Gorjipour, E., 2008. Effects

of canola meal on physiological and biochemical parameters in rainbow trout

(Oncorhynchus mykiss). Aquaculture Nutrition 14, 110–119.

4

Chapter 1: Replacement of fish meal with rapeseed protein concentrate in

diets fed to common carp (Cyprinus carpio L.)

H. Slawski1,2, H. Adem3, R.-P. Tressel3, K. Wysujack4, U. Koops4 and C. Schulz1,2

1Gesellschaft für Marine Aquakultur mbH, Hafentörn 3, D-25761 Büsum

2Institute of Animal Breeding and Husbandry, Christian-Albrechts-Universität zu Kiel,

D-24098 Kiel

3Pilot Pflanzenöltechnologie Magdeburg e.V., Berliner Chaussee 66, D-39114 Magdeburg

4Johann Heinrich von Thünen-Institut, Federal Research Institute of Rural Areas, Forestry and

Fisheries; Institute of Fisheries Ecology, Wulfsdorfer Weg 204, D-22926 Ahrensburg

Published in: The Israeli Journal of Aquaculture (2011) 63, 605-611.

5

Abstract

The potential of rapeseed protein concentrate (RPC) as fish meal alternative in diets for

common carp (initial average weight 26.7 ± 0.8 g) was evaluated. Triplicate groups of fish

were fed isonitrogenous (40.4 ± 0.2 % crude protein) and isocaloric (21.4 ± 0.1 kJ g-1)

experimental diets with 0% (R0), 33% (R33), 66% (R66) or 100 % (R100) of fish meal

replaced with rapeseed protein concentrate. At the end of the 56 days feeding period, growth

parameters and feed efficiencies were not significantly different between fish fed on diet R0

and R33. Diets R66 and R100 led to reduced diet intake and feed efficiencies resulting in

lower growth performances. It appears that diet taste and amino acid profiles were negatively

affected by high inclusion levels of rapeseed protein concentrate resulting in reduced diet

acceptance and protein value. It is concluded, that the used rapeseed protein concentrate can

effectively replace 33 % of fish meal in diets for carp without using palatability enhancers or

amino acid supplements.

6

Introduction

Wide availability, high protein content and a desirable amino acid profile have caused an

interest in rapeseed products as fish meal alternative in fish feeds. Rapeseed and canola

products have been tested as protein sources in diets for several fish species, including

rainbow trout (Thiessen et al. 2004), Coho salmon (Higgs et al. 1979), Chinook salmon

(Satoh et al. 1998), tilapia (Yigit and Olmez 2009), channel catfish (Lim et al. 1998), silver

perch (Booth and Allan 2003), carp (Dabrowski and Kozlowska 1981), red sea bream

(Glencross et al. 2004), and turbot (Burel et al. 2000ac). It was found, that the nutritional

quality of rapeseed products largely depends on their levels of antinutritional factors.

Prominent antinutritional factors in rapeseed are glucosinolates, phytic acid, phenolic

constituents (e.g. tannins), and indigestible carbohydrates (Francis et al. 2001). Several

processing techniques have been adapted to reduce the level of antinutrients in rapeseed in

order to improve its value for fish nutrition. Dehulling of seeds and utilisation of high

temperatures and organic solvents (hexane) during oil extraction as well as sieving of meal

decrease content of glucosinolates, phytate, fibre, cellulose, hemicellulose, sinapin and

tannins (Anderson-Haferman et al. 1993; Mawson et al. 1993, 1994ab, 1995; Leming et al.

2004) and increase protein level in meals (Mwachireya et al. 1999). In addition, protein

extraction from meals by methanol-ammonia-treatment or ethanol-treatment will further

increase protein level and effectively remove glucosinolates, phenolic compounds, soluble

sugars, such as sucrose, and some oligosaccharides (Naczk and Shahidi 1990; McCurdy and

March 1992; Chabanon et al. 2007) but will also increase levels of non-digestible fibre

(Mwachireya et al. 1999). In the present study liquid water extractions combined with

ultrafiltration were used to further increase the protein concentration of the final product and

at the same time deposit non-digestible fibres. The resulting rapeseed protein concentrate

(RPC) contained 71 % crude protein. Momentarily, rapeseed and canola protein products of

similar quality are being produced in different countries for application in animal nutrition.

However, these products are produced for test purpose until their potential as protein source in

animal nutrition is clarified. Besides nutritive quality, their costs of production will have to

become low enough to make rapeseed and canola protein concentrates available at a

competitive price compared to other protein sources, e.g. fish meal. As basic trial in a series

of consecutive feeding trials in order to optimize the produced RPC for application in fish

feeds, the product was tested as fish meal replacement in pelleted diets, using juvenile

7

common carp as model species. By this, we intended to determine the fundamental native

limitations of our RPC as fish feed ingredient.

Materials and methods

Diet preparation and experimental procedures

Four experimental diets were formulated to replace fish meal with rapeseed protein

concentrate (RPC) at 0, 33, 66, or 100 % (designated as R0, R33, R66 and R100,

respectively). Vitamins and minerals were added to diets to meet the dietary requirements of

carp (NRC, 1993). Diets were manufactured to give pellets 4 mm in diameter (L 14-175,

AMANDUS KAHL, Reinbek, Germany). The diets were formulated to be isonitrogenous

(40.4 ± 0.2 % crude protein) and isocaloric (21.4 ± 0.1 kJ g−1). Since this is the first trial in a

series of consecutive feeding trials investigating our RPC as fish meal replacement, we

intended to highlight direct effects on feed quality resulting from dietary RPC incorporation.

Therefore diets were formulated without palatability enhancers or crystalline amino acids.

Diet formulations, nutritional compositions and amino acid profiles are given in Table 1.1.

Solvent extracted RPC was obtained from PPM, Magdeburg, Germany. For oil extraction,

rapeseed was cold pressed and residual oil removed by a hexane treatment. Glucosinolates

were extracted with an ethanol solution. Liquid water extraction as well as dia- and

ultrafiltration of proteins followed by spray drying provided a protein concentrate with 71 %

crude protein content (Table 2.1).

The growth trial was conducted at the Johann Heinrich von Thünen Institute of Fisheries

Ecology, Ahrensburg, Germany. In the growth trial, common carp (Cyprinus carpio L.) was

used as model fish. In its juvenile stage, common carp has a high dietary protein requirement

(Fine et al. 1996) making this relatively modest fish an ideal model species for fish meal

replacement studies. Juvenile common carp that had been hatched in the institute were used.

One week before the experiment started 12 fish were stocked in each of twelve experimental

tanks (70 L; bottom surface 480 cm2), being part of a freshwater recirculation system. Tanks

were provided with water at 1 L min-1 (temperature: 23.8 ± 0.7 °C; O2: 6.5 ± 0.7 mg L-1; pH:

7.0 ± 0.7; NH4+: <0.1 mg L-1; NO2

-: <0.2 mg L-1). Photoperiod was in accordance to natural

rhythmic from February to April at our latitude (53° 41' 0" N). For a one week adaptation

period fish were fed the control diet (Table 1.1) in 4 daily meals until apparent satiation. After

the adaptation period, fish were fasted for one day and initial average weight was determined

(26.7 ± 0.8 g). Triplicate groups of 10 fish were fed the experimental diets in four daily meals

8

(at 8.00 a.m., 11.00 a.m., 2.00 p.m., 5.00 p.m.) to apparent satiation for 56 days. At the

beginning and at end of the experiment, 2 fish per tank were removed and analyzed for

proximate body composition.

Chemical analysis and laboratory procedures

Experimental diets and homogenized fish bodies were analysed for dry matter (DM) (105°C,

until constant weight), crude ash (550°C, 2 hours), crude fat (Soxtec HT6, Tecator, Höganäs,

Sweden) and crude protein content (N x 6.25; Kjeltec Auto System, Tecator, Höganäs,

Sweden). Raw material and dietary amino acid concentrations were analysed as described by

Tzovenis et al. (2009).

Calculations and statistical analysis

Weight gain (WG): (final weight – initial weight) / initial weight × 100; Specific growth rate

(SGR) (% per day): (ln final body weight – ln initial body weight) × 100 / days fed; Feed

conversion ratio (FCR): g dry feed intake / g wet body weight gain; Protein efficiency ratio

(PER): g wet body weight gain / g protein intake; Gross energy intake (GEI): gross energy

content in diet × g dry feed intake; Survival (%): initial fish count – dead fish count × 100 /

initial fish count.

All diets were assigned by a completely randomized design. The data were checked for

normal distribution using Kolmogoroff Smirnov Test and eventually subjected to

transformation. Data were analyzed by one-way analysis of variance (ANOVA) with “R“

statistical software. When differences among groups were found, the differences in means

were made with Tukey’s Honestly Significant Difference test. Statistical significance was

determined by setting the aggregate type I error at 5 % (P<0.05) for each set of comparisons.

Data are presented as means±SD.

9

Table 1.1: Formulation (g kg-1), amino acid profiles (% of dietary protein) and proximate composition (% of dry

matter) of experimental diets for common carp

Ingredients R0 R33 R66 R100

Herring meala 240 160 80 0

Soyprotein concentrateb 160 160 160 160

Rapeseed protein concentratec 0 75 155 235

Wheat glutend 150 150 150 150

Sunflower oil 53 57 58 60

Rapeseed oil 27 28 29 30

Dextrose 150 150 148 145

Maize starch 200 200 200 200

Vit/MinMixe 20 20 20 20

Amino acids

Aspartic acid 7.40 7.28 7.33 7.29

Threonine 3.38 3.37 3.56 3.68

Serine 4.43 4.29 4.67 4.76

Glutamic acid 21.45 22.01 22.92 23.80

Prolin 7.40 7.34 7.57 7.76

Glycin 5.57 5.12 4.72 4.39

Alanine 4.59 4.31 4.14 3.95

Cystine 1.35 1.51 1.71 1.92

Valine 4.38 4.53 4.38 4.52

Methionine 1.84 1.75 1.69 1.63

Isoleucine 3.89 4.02 3.85 4.02

Leucine 6.94 7.07 7.25 7.52

Phenylalanine 4.51 4.56 4.67 4.87

Histidine 2.27 2.37 2.48 2.63

Lysine 4.67 4.34 3.80 3.24

Arginine 5.38 5.53 5.75 6.00

Proximate composition

Dry matter (%) 91.7 92.4 93.4 94.0

Crude protein 40.4 40.1 40.6 40.5

Crude fat 11.0 11.1 10.4 9.4

Ash 7.6 6.3 5.3 4.1

NfEf 41.0 42.5 43.7 46.0

Gross energyg (kJ g-1) 21.2 21.5 21.5 21.5 aVFC GmbH, Cuxhaven, Germany; bIMCOSOY 60 Piglet, IMCOPA, Araucaria, Brasil; cPPM, Magdeburg, Germany; dEuroduna-Technologies GmbH, Barmstedt, Germany; eAA-Mix 517158 & 508240, Vitfoss, Gråsten, Denmark; fNitrogen

free extract = 100 – (%crude protein + %crude fat + %ash + %fibre); gCalculated by: crude protein = 23.9 kJ g-1; crude fat =

39.8 kJ g-1; NfE, fibre: 17.6 kJ g-1

10

Table 1.2: Proximate composition (% dry matter) and amino acid profiles (% of dietary protein) of fish meal

(FM) and rapeseed protein concentrate (RPC) and concentration of antinutritional factors detected in RPC

FM RPC

Dry matter (%) 91.6 94.2

Crude protein 69.0 71.0

Crude fat 7.0 2.2

Ash 20.7 9.2

Crude fibre 0.5 4.8

NfEa 2.8 12.8

Gross energyb (kJ g-1) 19.9 25.2

Amino acids

Aspartic acid 8.29 7.60

Threonine 3.90 4.44

Tryptophane 0.84 1.42

Serine 4.06 4.40

Glutamic acid 12.47 17.80

Prolin 4.69 5.89

Glycine 8.13 5.29

Alanine 6.41 4.70

Cystine 0.80 2.18

Valine 4.45 5.43

Methionine 2.37 2.03

Isoleucine 3.63 4.29

Leucine 6.46 7.81

Tyrosine 2.62 3.28

Phenylalanine 3.52 4.28

Histidine 2.00 2.99

Lysine 6.55 5.70

Arginine 5.84 7.49

Glucosinolates (µmol g-1) 0.2

Phytic acid (mg kg-1) < 500

Tannins (g 100g-1) < 0.005 aNitrogen free extract = 100 – (%crude protein + %crude fat + %ash + %fibre) bCalculated by: crude protein = 23.9 kJ g-1; crude fat = 39.8 kJ g-1; NfE, fibre: 17.6 kJ g-1

11

Results

Final weight, weight gain and standard growth rate were not significantly different between

fish fed on the control diet and diet R33. In addition, feed and gross energy intake as well as

feed conversion and protein efficiency ratio were not significantly different between control

diet and diet R33 fed fish. On the contrary, fish fed on diets R66 and R100 showed

significantly reduced growth performance values, lower feed intake and decreased feed

efficiencies (Table 1.3). No significant differences in whole body composition were detected

between fish fed control diet and fish receiving test diets (Table 1.4).

Table 1.3: Growth response, feed efficiencies and survival of carp fed experimental diets


Initial weight (g) 26.4 ± 0.6 26.7 ± 0.6 26.5 ± 0.8 27.2 ± 1.3

Final weight (g) 73.5a ± 4.1 71.2a ± 4.3 60.6b ± 3.2 49.7c ± 2.3

Weight gain (%) 178.4a ± 20.3 167.0a ± 17.3 128.7b ± 18.3 83.3c ± 16.3

SGR (% day-1) 1.83a ± 0.13 1.75a ± 0.12 1.47b ± 0.15 1.08c ± 0.16

DM Feed intake 51.5a ± 4.11 49.43a ± 3.82 42.20b ± 2.47 34.80c ± 3.80

FCR 1.09a ± 0.04 1.11a ± 0.02 1.24b ± 0.07 1.56c ± 0.14

PER 36.96a ± 1.23 36.09a ± 0.69 32.72b ± 1.95 26.18c ± 2.36

GEI (kJ) 109.2a ± 10.7 106.3ab ± 10.1 90.7b ± 6.5 74.8bc ± 10.0

Survival (%) 96.7 ± 5.8 96.7 ± 5.8 93.3 ± 5.8 96.7 ± 5.8

Means in the same row with different superscript letters are significantly different determined by Tukey's test (P<0.05)

Table 1.4: Proximate whole body composition of carp fed the experimental diets


Moisture (%) 76.3 ± 0.3 75.9 ± 0.3 75.5 ± 0.4 75.3 ± 0.4

Crude protein (% w.w.a) 16.7 ± 0.3 16.9 ± 0.8 17.2 ± 0.5 17.3 ± 0.3

Crude fat (% w.w.) 4.4 ± 0.4 4.3 ± 0.5 4.4 ± 0.3 4.2 ± 0.3

Ash (% w.w.) 2.1 ± 0.2 2.3 ± 0.1 2.4 ± 0.1 2.5 ± 0.1

Initial body composition: moisture 75.7 %, crude protein 12.6 %, crude fat 2.6 %, ash 1.3 % aw.w.: wet weight

12

Discussion

Rapeseed protein concentrate has been found to be a viable alternative to fish meal in fish

feeds. Different replacement levels of fish meal have been achieved by inclusion of rapeseed

and canola protein concentrate in feeding trials with several fish species (Dabrowski and

Kozlowska 1981; Lim et al. 1998; Burel et al. 2000; Booth and Allan 2003; Glencross et al.

2004; Thiessen et al. 2004; Yigit and Olmez 2009). It was found, that antinutritional factors

present in rapeseed and canola particularly determine its value for fish nutrition. It was

therefore recommended to greatly reduce antinutritional factors in rapeseed protein

concentrates in order to achieve higher fish meal replacement levels in fish diets. In the

present study a rapeseed protein concentrate (RPC) with 71 % crude protein content and

particularly low levels of glucosinolates, phytic acid and tannins (Table 1.2) was tested as fish

meal alternative in diets for common carp. The RPC successfully replaced 33 % of fish meal

protein from a control diet without retarding fish growth performance, feed intake or feed

efficiencies. At 66 % and 100 % fish meal replacement with RPC, however, fish growth

performance, feed intake and feed efficiency decreased compared to the control group. This is

in contrast to findings by Dabrowski and Kozlowska (1981) who successfully replaced 100 %

of fish meal protein from diets for common carp with rapeseed meal protein without reducing

fish weight gain or standard growth rate. We believe that a 66 and 100 % fish meal

replacement with our RPC negatively affected diet palatability and protein quality and

therefore limited fish growth performance.

We observed, that increasing dietary levels of RPC as fish meal replacement diminished diet

acceptance by common carp. This is documented in lower feed intake in fish fed diet R66 and

R100 compared to diet R0 and R33 (Table 1.3). It is known that the bitter taste exuded by

glucosinolates and their breakdown products present in rapeseed and canola meals can

potentially retard diet acceptance by fish. This was found in rainbow trout and turbot at

dietary glucosinolate levels of 7.3 µmol g-1 or 18.7 µmol g-1, respectively (Burel et al.

2000b,c). Because the RPC used in our study contained about 0.2 µmol g-1 (Table 1.2) the

highest dietary glucosinolate concentration was therefore 0.05 µmol g-1 in R100.

Nevertheless, we noticed a mustard smell in diets R66 and R100 resulting from high RPC

inclusion. This smell clearly influenced diet acceptance by carp as RPC was the dominant

protein source in diets R66 and R100. In Dabrowski and Kozlowska (1981) diet acceptance

by carp was probably equalized by the usage of several protein sources (blood meal, yeast,

soyabean meal, barley meal and rapeseed meal) giving a more versatile diet taste.

13

Beside palatability problems, high dietary inclusion of RPC led to insufficient dietary amino

acid concentrations for common carp. Table 1.1 shows the dietary amino acid profiles. Diets

high in RPC are low in lysine, because of low fish meal inclusion and no inclusion of other

protein sources of animal origin. As reported by Ogino (1980) the lysine requirement of

common carp is 5.7 % of dietary protein. But, although diets R0 and R33 contained 4.67 %

and 4.34 % lysine respectively, good growth results obtained with these diets counteract the

observation by Ogino (1980). However, the fast drop in growth rates in fish fed on diet R66

and R100 confirms an inappropriate quality of their dietary protein.

In conclusion, 33 % of fish meal can be replaced by rapeseed protein concentrate from diets

for common carp. Consecutive feeding trials will clarify if higher fish meal replacement with

our RPC can be achieved by using palatability enhancers and amino acid supplements.

References



Booth, M.A., Allan, G.L., 2003. Utilization of digestible nitrogen and energy from four

agricultural ingredients by juvenile silver perch Bidyanus bidyanus. Aquaculture

Nutrition 9, 317-326.




Burel, C., Boujard, T., Escaffre, A.M., Kaushik, S.J., Boeuf, G., Mol, K., van der Geyten, S.,

Kühn, E.R., 2000b. Dietary low glucosinolate rapeseed meal affects thyroid status and

nutrient utilization in rainbow trout (Oncorhynchus mykiss). British Journal of Nutrition

83, 653–664.


Quinsac A., Krouti M., Ribaillier, D., 2000c. Potential of plant-protein sources as fish





hydrolysates. Process. Biochem. 42, 1419–1428.

14




263-274.

Fine, M., Zilberg, D., Cohen, Z., Degani, G., Moav, B., Gertler, A., 1996. The Effect of

Dietary Protein Level, Water Temperature and Growth Hormone Administration on

Growth and Metabolism in the Common Carp (Cyprinus carpio). Comparative

Biochemistry and Physiology A 114, 35-42.



Glencross, B., Hawkins, W., Curnow, J., 2004. Nutritional assessment of Australian canola

meals. I. Evaluation of canola oil extraction method and meal processing conditions on

the digestible value of canola meals fed to the red seabream (Pagrus auratus, Paulin).

Aquaculture Research 35, 15-24.

Higgs, D.A., Fagerlund, U.H.M., McBride, J.R., Plotnikoff, M.D., Dosanjh, B.S., Markert,

J.R., Davidson, J., 1983. Protein quality of Altex canola meal for juvenile chinook

salmon (Oncorhynchus tshawytscha) considering dietary protein and 3,5,3’-triiodo-L-

thyronine content. Aquaculture 34, 213-238.

Higgs, D.A., Markert, J.R., MacQuarrie, D.W., McBride, J.R., Dosanjh, B.S., Nichols, C.,

Hoskins, G., 1979. Development of practical dry diets for coho salmon, Oncorhynchus

kisutch, using poultry-by-product meal, feather meal, soybean meal and rapeseed meal

as major protein sources. In: J.E. Halver, K. Tiews (eds.). Proceedings of the World

Symposium on Fish Nutrition and Fishfeed Technology. Vol. 11, Hamburg, 1978.

Leming, R., Lember, A., Kukk, T., 2004. The content of individual glucosinolates in rapeseed

and rapeseed cake produced in Estonia. Agraaeteadus 15, 21–27.

Lim, C., Klesius, H., Higgs, D.A., 1998. Substitution of canola meal for soybean meal in diets

for channel catfish Ictalurus punctatus. J. of the World Aquacult. Society 29, 161-168.

Mawson, R., Heaney, R.K., Zdunczyk, Z., Kozlowska, H., 1993. Rapeseed meal -

glucosinolates and their antinutritional effects: Part 1. Rapeseed production and

chemistry of glucosinolates. Die Nahrung 37, 131–140.

Mawson, R., Heaney, R.K., Zdunczyk, Z., Kozlowska, H., 1994a. Rapeseed meal -

glucosinolates and their antinutritional effects. Part 3. Animal growth and performance.

Die Nahrung 38, 167–177.

15

Mawson, R., Heaney, R.K., Zdunczyk, Z., Kozlowska, H., 1994b. Rapeseed meal -

glucosinolates and their antinutritional effects. Part 4. Goitrogenicity and internal organs

abnormalities in animals. Die Nahrung 38, 178–191.


glucosinolates and their antinutritional effects: 7. Processing. Die Nahrung 39, 32–41.

McCurdy, C.M., March, B.E., 1992. Processing of canola meal for incorporation in trout and

salmon diets. J. Am. Oil Chem. Soc. 69, 213-220.

Mwachireya, S.A., Beames, R.M., Higgs, D.A., Dosanjh, B.S., 1999. Digestibility of canola

protein products derived from the physical, enzymatic and chemical processing of

commercial canola meal in rainbow trout, Oncorhynchus mykiss (Walbaum) held in

fresh water. Aquaculture Nutrition 5, 73-82.

Naczk, M., Shahidi, F., 1990. Carbohydrates of canola and rapeseed. In: F. Shahidi (ed.).

Canola, Rapeseed: Production, Chemistry, Nutrition & Processing Technology. Van

Nostrand Reinhold, New York, pp211-220.

NRC (National Research Council), 1993. Nutrient Requirements of Fish. National Academy

Press, Washington, DC.

Ogino C., 1980. Protein requirement of carp and rainbow trout. B. Jpn. Soc.Sci. Fish. 46, 171.

Satoh, S., Higgs, D.A., Dosanjh, B.S., Hardy, R.W., Eales, J.G., Deacon, G., 1998. Effect of

extrusion processing on the nutritive value of canola meal for chinook salmon

(Oncorhynchus tshawytscha) in seawater. Aquaculture Nutrition 4, 115-122.

Thiessen, D.L., Maenz, D.D., Newkirk, R.W., 2004. Replacement of fish meal by canola

protein concentrate in diets fed to rainbow trout (Oncorhynchus mykiss). Aquaculture

Nutrition 10, 379–388.

Tzovenis, I., Fountoulaki, E., Dolapsakis, N., Kotzamanis, I., Nengas, I., Bitis, I., Cladas, Y.,

Economou-Amilli, A., 2009. Screening for marine nanoplanktic microalgae from Greek

coastal lagoons (Ionian Sea) for use in mariculture. J. Appl. Phycol. 21, 457–469.

van Barneveld, R.J., 1998. Influence of oil extraction method on the nutritional value of

canola meal for growing pigs. Project DAS38-1188. Pig Research and Development

Corporation, Canberra, p 46.

Yigit, N.O., Olmez, M., 2009. Canola Meal as an Alternative Protein Source in Diets for Fry

of Tilapia (Oreochromis niloticus). Bamidgeh 61(1), 35-41.

16

Chapter 2: Replacement of fish meal with rapeseed protein concentrate in

diets fed to wels catfish (Silurus glanis L.)

H. Slawski1,2, H. Adem3, R.-P. Tressel3, K. Wysujack4, U. Koops4,

S. Wuertz1 and C. Schulz1,2



D-24098 Kiel




Published in: Aquaculture Nutrition (2011) DOI: 10.1111/j.1365-2095.2011.00857.x

17

Abstract

The potential of rapeseed protein concentrate as fish meal alternative in diets for wels catfish

(initial average weight 86.5 ± 1.9 g) was evaluated. Sixteen fish were stocked into each of

twelve experimental tanks being part of a freshwater recirculation system. Fish were

organized in triplicate groups and received isonitrogenous (603 ± 3 g CP kg-1) and isocaloric

(23.0 ± 0.3 kJ g-1) experimental diets with 0, 25, 50, and 75 % of fish meal replaced with

rapeseed protein concentrate (710 g CP kg-1). At the end of the 63 days feeding period, weight

gain, standard growth rate, feed intake, feed conversion ratio and protein efficiency showed

no significant difference between control group and fish fed on diets with 25 % reduced fish

meal content by inclusion of rapeseed protein concentrate. Higher dietary fish meal

replacement negatively affected diet quality and palatability resulting in reduced feed intake,

feed efficiencies and fish performance. However, blood serum values of triglycerides, glucose

and protein were not significantly different between treatment groups, still indicating a

favourable nutrient supply from all experimental diets.

18

Introduction

In 2009, worldwide production of rapeseed (including canola) was 61.6 Mio t. Thus,

rapeseed, commonly produced in temperate regions, ranked as number three oilseed

worldwide, only surpassed by soybean (222.2 Mio t) and cotton seed (64.0 Mio t) (FAO

2010). For soybean, a crop mainly cultivated in warm regions, efforts and research have been

undertaken to make it a commonly accepted fish feed ingredient and fish meal alternative

(Gatlin et al. 2007). The usage of rapeseed products as fish feed ingredients, however, is

limited. Either simple oilcakes or rapeseed meals with increased protein content produced

from oilcakes that were de-oiled with organic solvents have been tested as protein sources in

feeding trials with several fish species, among them Oncorhynchus mykiss (Burel et al.

2000ac, 2001; Thiessen et al. 2003, 2004; Shafaeipour et al. 2008), Oreochromis

mossambicus (Davies et al. 1990), Ictalurus punctatus (Webster et al. 1997), Cyprinus carpio

(Dabrowski and Kozlowska 1981), Pagrus auratus (Glencross et al. 2004) and Psetta maxima

(Burel et al. 2000ab). In general, experimental results showed that the nutritional quality of

simple rapeseed products is below that of fish meal although they contained a well balanced

amino acid profile. Particularly antinutritional factors (ANF) determine the quality of

rapeseed products for fish nutrition. Prominent ANF in rapeseed products are glucosinolates,

phytic acid, phenolic constituents (e.g. tannins) and indigestible carbohydrates (Mawson et al.

1995; Francis et al. 2001). Several processing techniques can be adapted to reduce the level of

antinutrients in rapeseed products and improve their value for fish nutrition. Dehulling of

seeds and utilisation of high temperatures and organic solvents (hexane) during oil extraction

as well as sieving of meal decrease content of glucosinolates, phytate, fibre, cellulose,

hemicellulose, sinapin and tannins (Fenwick et al. 1986; Anderson-Haferman et al. 1993;

Tripathi et al. 2000) and increase protein level in meals (Mwachireya et al. 1999). Protein

extraction from meals by methanol-ammonia-treatment or ethanol-treatment will increase

protein level and effectively remove glucosinolates, phenolic compounds, soluble sugars, such

as sucrose, and some oligosaccharides (e.g. raffinose and stachyose) (Naczk and Shahidi

1990; Chabanon et al. 2007). Last but not least water treatment appears to be a cost effective

method for removing glucosinolates from rapeseed meals (Tyagi 2002). Sporadically,

rapeseed protein products of high quality are being produced in different countries for

application in animal nutrition. However, these products are produced for test purposes in

small volumes until their potential as protein source in animal nutrition is clarified. Besides

nutritive quality, their costs of production will have to become low enough to make rapeseed

19

protein products available at a competitive price compared to other protein sources, especially

fish meal. In the present study, a high quality rapeseed protein concentrate containing 710 g

CP kg-1 was tested as fish meal replacement in fish diets. Diets were fed to wels catfish

(Silurus glanis L.), a carnivorous species that is believed to have potential for indoor

recirculation farming in Europe as a high value product for local markets (Mazurkiewicz et al.

2008). Fish performance and blood serum parameters were investigated to evaluate rapeseed

protein concentrate as fish meal alternative in diets for wels catfish.


Diet preparation and experimental procedures

Four experimental diets were formulated in which fish meal was replaced with rapeseed

protein concentrate (RPC) at 0, 25, 50, and 75 % level (designated as R0, R25, R50, or R75,

respectively). Solvent extracted RPC was obtained from the PPM e.V., Magdeburg, Germany.

For the production of RPC a batch of rapeseed (variety Lorenz, Norddeutsche Pflanzenzucht,

Germany) was conditioned in a vacuum dryer for 15 minutes at 60-70 °C to inactivate the

enzyme myrosinase. Then rapeseed was cold pressed. To remove residual oil from the oilcake

(129 g kg-1 oil, 313 g kg-1 protein) it was crushed into 1-5 mm particle size followed by a

hexane treatment. The treatment lasted for two hours and the incubation temperature was 60

°C. Hexane treated rapeseed meal extract was desolventised under pressure to remove hexane

(< 300 ppm), then rapeseed meal extract was further crushed to a particle size of 0.2 to 0.1

mm. A four step treatment using a 75 % ethanol solution (35 min at 60 °C) aimed to remove

glucosinolates from rapeseed meal extract. This resulted in a residual oil content of the nearly

glucosinolate free rapeseed meal extract of 11 g kg-1 and a protein content of 398 g kg-1. In the

following, protein was gained through liquid water extraction (rapeseed meal extract 1:15

water). For this, the suspension was heated to 40-45 °C followed by two hours of constant

agitation. Afterwards the suspension was decanted. Following decantation the solvent was

collected and residue material was secondly extracted (residue 1:10 water, 5 % NaCl) at 40-45

°C and one hour contact time under constant agitation. Following extraction the suspension

was decanted. Solvent was collected and residue prepared for a third extraction. Then solvents

of extraction 1, 2 and 3 were collected to remove low-molecular compounds and to

concentrate dissolved proteins by dia- and ultrafiltration. During filtration conductivity was

checked. Protein washing ended, when conductivity was 5-6 mS cm-1, corresponding to a

protein content of 600 g kg-1. The gained material was spray dried at 70-80 °C, which led to a

20

rapeseed protein concentrate with 710 g kg-1 protein content (Table 2.1). Vitamins and

minerals were added to diets to meet the dietary requirements of freshwater fish (NRC 1993).

The diets were formulated to be isonitrogenous (603 ± 3 g CP kg-1) and isocaloric (23.0 ± 0.3

MJ kg-1). Essential amino acid concentrations did not differ considerably between

experimental diets. The diets were manufactured to give pellets 4 mm in diameter (L 14-175,

AMANDUS KAHL, Reinbek, Germany). Diet formulations, proximate compositions and

amino acid profiles are given in Table 2.2.

Table 2.1: Nutrient composition (g kg-1 dry matter) and essential amino acid profiles (g kg-1 protein) of fish meal

(FM) and rapeseed protein concentrate (RPC) and concentration of antinutritional factors detected in RPC

FM RPC

Dry matter 916 942

Crude protein 690 710

Crude fat 70 22

Ash 207 92

Phosphorus 29 21

Crude fibre 5 48

NfEa 28 128

Gross energyb (kJ g-1) 19.9 25.2

Essential amino acids

Arginine 58.4 74.9

Histidine 20.0 29.9


Leucine 64.6 78.1

Lysine 65.5 57.0

Methionine 23.7 20.3


Threonine 39.0 44.4

Tryptophane 8.4 14.2

Valine 44.5 54.3

Glucosinolates (µmol g-1) 0.2

Phytic acid (g kg-1) < 0.5

Tannins (g 100g-1) < 0.005 aNitrogen free extract = 100 – (%crude protein + %crude fat + %ash + %fibre). bCalculated by: crude protein = 23.9 MJ kg-1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1.

21

Table 2.2: Formulation (g kg-1), essential amino acids composition (g kg-1 crude protein) and proximate

composition (g kg-1 dry matter) of experimental diets


Herring meala 500 375 250 125

Rapeseed protein concentrate 0 122 243 362

Soyprotein concentrateb 110 110 110 110

Blood mealc 105 105 105 105

Wheat glutend 65 65 65 65

Fish oila 120 126 132 138

Dextrosec 80 77 75 75

Vit/MinMixe 20 20 20 20


Arginine 52.9 55.6 58.3 59.1

Histidine 30.9 33.0 33.9 34.6

Isoleucine 30.2 31.7 33.2 32.1

Leucine 75.9 79.1 81.3 81.7

Lysine 59.3 59.2 57.7 54.7

Methionine + Cysteine 27.5 28.9 30.4 31.1

Phenylalanine 42.9 44.9 46.4 47.1

Threonine 34.1 35.6 36.6 37.1

Valine 49.9 52.0 53.5 52.6


Dry matter (g kg-1) 922 931 939 944

Crude protein 607 603 600 600

Crude fat 158 165 158 148

Ash 131 114 99 83

Phosphorus 14.8 14.1 13.0 12.7

NfE + fibref 104 118 143 169

calculated Glucosinolates

(µmol g-1)

0 0.02 0.05 0.07

Gross energyg (MJ kg-1) 22.6 23.1 23.1 23.2 aVFC GmbH, Cuxhaven, Germany; bIMCOSOY 60 Piglet, IMCOPA, Araucaria, Brasil.; cEuroduna-Technologies GmbH,

Barmstedt, Germany; dCargill Deutschland GmbH, Krefeld, Germany; eAA-Mix 517158 & 508240, Vitfoss, Gråsten,

Denmark; fNitrogen free extract = 100 – (%crude protein + %crude fat + %ash).; gCalculated by: crude protein = 23.9 MJ kg-

1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1.

Tryptophane was not analyzed.

22

The growth trial was conducted at the Johann Heinrich von Thünen Institute of Fisheries

Ecology, Ahrensburg, Germany. Juvenile wels catfish (Silurus glanis L.) were obtained from

the Ahrenhorster Edelfisch GmbH & CO KG (Ahrenhorst, Germany). Two weeks before the

experiment started 17 fish were stocked in each of nine experimental tanks (96 L; bottom

surface 480 cm2), being part of a recirculation system. Tanks were provided with freshwater

at 2 L min-1 (temperature: 26.9 ± 0.7 °C; O2: 6.8 ± 0.5 mg L-1; pH: 7.3 ± 0.5; NH4+: <0.6 mg

L-1; NO2-: <0.2 mg L-1). Photoperiod was in accordance to natural rhythmic. In respect of the

fishes’ light sensitivity, tanks were half covered with translucent plastic lids. For a two week

adaptation period fish were fed the control diet in four daily meals until apparent satiation.

After the adaptation period, initial average fish weight was determined (86.5 ± 1.9 g). For an

experimental period of 63 days, triplicate groups of fish were fed the experimental diets in

four daily meals (8:00 and 11:00 a.m., 2:00 and 5:00 p.m.) until apparent satiation. At the

beginning and at end of the experiment, two fish per tank were removed and stored at -23 ºC

for the determination of initial and final body composition.

Sampling

At the end of the feeding period, blood samples from the caudal vein and artery of eight fish

per experimental treatment were taken with a heparinized syringe (1 ml). Blood haematocrit

percentage was determined after centrifugation (3500 rpm, 6 min) of glass tubes filled with

fresh blood in a haematocrit centrifuge (Haematokrit 210, Andreas Hettich GmbH & Co.KG,

Tuttlingen, Germany). Remaining fresh blood was filled in Eppendorf tubes and centrifuged

(4000 rpm, 5 min). Supernatant blood plasma was separated into Eppendorf tubes and stored

at -84 °C in a freezer.


Diets and homogenised fish bodies were analysed in duplicate for proximate composition.

Dry matter was calculated from weight loss after drying in an oven at 105 °C until constant

weight. Fat content was determined after HCl hydrolysis (Soxtec HT6, Tecator, Höganäs,

Sweden) and total nitrogen content by the Kjeldahl technique (protein = N × 6.25; Kjeltec

Auto System, Tecator, Höganäs, Sweden). Ash content was calculated from weight loss after

incineration of samples in a muffle furnace for 2 hours at 550 °C. Raw material and dietary

amino acid concentrations were analysed as described by Tzovenis et al. (2009). Blood

plasma concentrations of triglycerides, glucose and protein were determined using a

23

microplate reader (Infinite 200®, TECAN Group Ltd., Männedorf, CH) and commercial kits

(Triglycerides GPO and Glucose GOD-PAP, Greiner Diagnostic GmbH, Bahlingen,

Germany; Roti®-Quant, CARL ROTH GmbH + Co.KG, Karlsruhe, Germany).


Fish performance was determined, using the following formulae:

Specific growth rate (SGR, % day-1) = (ln final body weight – ln initial body weight) × 100 /

days fed

Feed intake as % body weight day-1 = the mean feed consumption per fish per day as a

percentage of the daily fish body weight for the experimental period. The daily fish body

weight was calculated using daily SGR values equal to the final SGR of each tank.

Feed conversion ratio (FCR) = g dry feed intake / g wet body weight gain

Protein efficiency ratio (PER) = g wet body weight gain / g protein intake

Condition factor (CF) = g body weight / cm total length3 × 100

Survival (%) = (initial fish count - dead fish count) / initial fish count × 100

All diets were assigned by a completely randomized design. Biological and analytical data

were checked for normal distribution using the Kolmogoroff Smirnov Test and eventually

subjected to transformation. Data were subjected to regression and one-way analysis of

variance (ANOVA) using SPSS 17.0 for Windows. When differences among groups were

identified, multiple comparisons among means were made using Tukey’s HSD test. Statistical

significance was determined by setting the aggregate type I error at 5% (P<0.05) for each set

of comparisons.

Results

No significant differences in growth performance parameters and feed efficiencies were

detected between control diet and R25 diet fed fish. Compared to the control group, fish

growth performance, voluntary feed intake and feed efficiencies declined at fish meal

replacement levels above 25 %. Feed intake as % body weight was not affected up to 50 %

fish meal replacement level (Table 2.3). While fish growth performance, voluntary feed intake

and feed efficiencies significantly correlated with the dietary inclusion level of RPC (Table

2.3) no correlation was found between feed intake as % body weight and dietary inclusion of

RPC. No significant differences in whole body composition were detected between fish fed

24

on the control diet and fish receiving RPC diets (Table 2.4). Significant correlations were

found between dietary RPC and phosphorus level and whole-body moisture (R2=0.50 and

0.48, P<0.05), fat (R2=0.54 and 0.53, P<0.01) and ash (R2=0.56 and 0.58, P<0.01) content.

Table 2.3: Growth response, feed intake, feed efficiencies, condition factor and survival of wels catfish fed

experimental diets

R0 R25 R50 R75 *R2 P

Initial weight (g) 87.8 ± 2.7 85.5 ± 1.2 86.1 ± 1.5 86.7 ± 2.0

Final weight (g) 279.8a ± 9.1 264.0a ± 7.7 218.4b ± 15.7 159.4c ± 4.4 0.92 <0.01

SGR (%) 1.84a ± 0.10 1.79a ± 0.06 1.48b ± 0.10 0.97c ± 0.04 0.86 <0.01

Feed intake (g DM) 102.9a ± 3.8 97.1a ± 3.2 85.7b ± 5.3 60.9c ± 3.1 0.88 <0.01

Feed intake as %

body weight day-1

0.96a ± 0.04 1.02a ± 0.03 0.97a ± 0.02 0.81b ± 0.01 0.36 ns

FCR 0.53a ± 0.03 0.56a ± 0.02 0.68b ± 0.04 0.86c ± 0.04 0.84 <0.01

PER 1.13a ± 0.06 1.11a ± 0.02 0.92b ± 0.05 0.72c ± 0.02 0.88 <0.01

GEI (MJ) 2.33a ± 0.09 2.24a ± 0.07 1.98b ± 0.12 1.41c ± 0.07 0.85 <0.01

CF 0.57 ± 0.04 0.62 ± 0.04 0.61 ± 0.05 0.61 ± 0.04 0.08 ns *R2: parameter values are regressed to the dietary level of rapeseed protein concentrate.

Values are given as mean ± standard deviation. Values in the same row with common superscript letters are not significantly

different (P<0.05).

Table 2.4: Proximate whole body composition (g kg-1 wet weight) of wels catfish fed the experimental diets

Parameter R0 R25 R50 R75

Moisture 740 ± 12 746 ± 10 755 ± 10 763 ± 11

Crude protein 153 ± 6 152 ± 7 150 ± 8 144 ± 5

Crude fat 85 ± 5 82 ± 7 76 ± 6 72 ± 6

Ash 24 ± 2 23 ± 3 21 ± 1 20 ± 2

Initial body composition: moisture 823 g kg-1, crude protein 131 g kg-1, crude fat 28 g kg-1, ash 18 g kg-1.

25

Haematocrit values as well as blood serum values determined showed no significant

difference between treatment groups and were not correlated to the dietary inclusion level of

RPC (Table 2.5).

Table 2.5: Blood haematocrit content and blood serum values of wels catfish fed experimental diets

Parameter R0 R25 R50 R75

Haematocrit (Proportion of 1) 0.23 ± 0.04 0.24 ± 0.04 0.25 ± 0.02 0.24 ± 0.02

Triglycerides (mM L-1) 4.72 ± 2.21 6.55 ± 4.89 7.20 ± 4.52 8.47 ± 1.46

Glucose (mM L-1) 6.54 ± 1.33 7.10 ± 1.27 6.66 ± 1.11 6.43 ± 1.27

Protein (g L-1) 36.0 ± 2.8 35.5 ± 2.1 36.1 ± 2.8 37.1 ± 1.9

Discussion

While usability and limitations of simple rapeseed products as fish feed ingredients have been

widely investigated (Dabrowski and Kozlowska 1981; Davies et al. 1990; Webster et al. 1997;

Burel et al. 2000a,b,c, 2001; Thiessen et al. 2003, 2004; Glencross et al. 2004; Shafaeipour et

al. 2008), lack of information exists about the benefits of high quality products originating

from rapeseed oilcakes with protein contents comparable to or above that of fish meal. Higgs

et al. (1982) successfully replaced 25 % of dietary protein from a fish meal control diet for

juvenile Oncorhynchus tshawytscha with rapeseed protein concentrate (613 g CP kg-1)

without reducing growth rate and food (protein) utilization. In the study, however, higher fish

meal replacement levels with rapeseed protein concentrate were not evaluated.

The results of our study demonstrate that 25 % of dietary fish meal can be replaced with RPC

in diets fed to wels catfish without negative effects on feed efficiencies and fish growth.

When 50 % of dietary fish meal was replaced with RPC the feed intake as % of fish body

weight was not significantly different from the control group but feed efficiencies and fish

growth were reduced. At 75 % fish meal replacement level fish showed reduced diet

acceptance and reluctant feed intake as a result of unfavourable diet taste. It appears therefore,

that the level of blood meal incorporated into diets as feed attractant did not effectively

counteract the negative effects on diet taste resulting from rapeseed protein concentrate. It is

known that the bitter taste exuded by glucosinolate metabolites, such as isothiocyanates and

26

vinyloxazolidinethiones, present in rapeseed meals can potentially retard diet acceptance by

fish. This was found in O. mykiss and P. maxima at dietary glucosinolate levels of 7.3 µmol g-

1 or 18.7 µmol g-1, respectively (Burel et al. 2000bc). Because the RPC used in our study

contained 0.2 µmol glucosinolates g-1 (Table 2.1) the highest calculated dietary glucosinolate

concentration was 0.07 µmol g-1 in diet R75 (Table 2.2). This value is far below the level

when glucosinolates become detrimental on food intake of O. mykiss and P. maxima (Burel et

al. 2000bc) but to our observation the typical mustard smell of glucosinolates was still

noticeable in diets R50 and R75. It seems, therefore, that wels catfish is more sensitive

towards a bitter diet taste than other carnivorous fish species. This goes together with the

fish’s excellent developed olfactory organ (Jakubowski and Kunysz 1979). Reduced feed

intake in fish fed on diets R50 and R75 resulted in lower growth rates and reduced feed

conversion compared to the control group (Table 2.3). For prospective feeding trials with

rapeseed protein products in wels catfish it appears recommendable to use other feed

attractants than blood meal. Fish protein hydrolysate, squid hydrolysate, stick water or krill

meal at dietary levels from 30 to 50 g kg-1 have shown to be effective feed attractants and

sources of amino acids and minerals when diets low in fish meal were fed to carnivorous fish

(Espe et al. 2006, 2007; Torstensen et al. 2008; Kousoulaki et al. 2009). Since fish behaved

calm in all treatment groups increased energy expenditure due to feed searching activity in

high RPC groups did not deplete feed conversion. Thus, lower feed efficiency might be a

result of reduced diet digestibility due to RPC inclusion.

In the present study we did not determine the digestibility of nutrients and minerals from RPC

in wels catfish. We discovered that faeces collection from wels catfish in order to determine

nutrient and mineral digestibility appears hardly possible. On the one hand, faeces of wels

catfish are slimy and rapidly dilute in water. This precludes faeces collection with an

automatic collector. On the other hand, faeces stripping, even when fish are anaesthetized,

will stress the sensitive fish. As a result, wels catfish will stop feed intake for days. Killing

fish, as a last alternative to gain faeces, requires a high number of individuals to collect

enough faeces for laboratory analysis. In the present study, fish count was not sufficient to

gain required amounts of faeces for laboratory analysis. Therefore, assumptions regarding

nutrient and mineral digestibility of RPC in wels catfish are based on studies conducted with

rapeseed protein products in other fish species. Mwachireya et al. (1999) stated that fibre

levels, either alone or together with phytate, can have greatest adverse effects on the

digestibility of canola protein products for O. mykiss. The authors reported that among

27

different canola products tested only canola protein isolate (908 g CP kg-1) met nutrient

digestibility coefficients corresponding to fish meal. In our study, the applied processing

techniques to produce RPC from rapeseed oilcake led to relatively low levels of phytic acid in

the final RPC (0.5 g kg-1). Accordingly, calculated dietary phytic acid concentrations

originating from RPC were 0.1 and 0.2 g kg-1 in diets R50 and R75, respectively. In fish

nutrition studies, phytic acid concentrations that negatively influence mineral and nutrient

availability are commonly higher. Spinelli et al. (1983) observed decreased growth rates in

rainbow trout fed a diet containing 5 g kg-1 synthetic phytic acid. Synthetic phytic acid at

concentrations of 5 and 10 g kg-1 feed resulted in lower growth performance in common carp

(Hossain and Jauncey 1993). Due to insignificant phytic acid concentrations in diets R50 and

R75, we assume, that diet digestibility was mainly reduced by fibre and other complex

carbohydrates. However, lack of information exists about the influence of complex

carbohydrates on nutrient digestibility in wels catfish. But it is known from other carnivorous

fish that complex carbohydrates can greatly reduce mineral and nutrient availability from

aquafeeds, thereby reducing feed efficiencies as observed in Salmo salar and P. maxima

(Storebakken et al. 1998; Burel et al. 2000). The RPC used in our study contained 48 g kg-1

fibre and calculated NfE was 128 g kg-1. This resulted in a dietary fibre + NfE content of 143

or 169 g kg-1 compared to 102 g kg-1 in the control group. We therefore assume, that diets

R50 and R75 were less digestible than the control diet for wels catfish, due to increased

dietary fibre and NfE contents originating from RPC. In contrast, Hayen et al. (1993) found

that rapeseed protein concentrate (600 g CP kg-1) was as digestible as LT herring/capelin meal

when fed to O. tshawytscha. According to our findings, it seems that wels catfish is highly

sensitive towards dietary fibre and NfE.

The amino acid requirement of wels catfish, to our knowledge, is not known. We therefore

assume that it is comparable to other carnivorous fish such as rainbow trout. Accordingly,

experimental diets were formulated to contain amino acid concentrations above established

requirement levels (NRC 1993). However, due to antinutritional factors present in RPC,

digestibility of amino acids could have been negatively affected as it is known from other

protein sources of vegetable origin (Francis et al. 2001). In particular lower dietary levels of

lysine in diets R50 and R75 compared to the control diet together with reduced lysine

digestibility might have negatively influenced feed efficiencies and fish growth in the present

study.

28

Fish body composition was not significantly different between treatments (Table 2.4).

Regression analysis, however, revealed a correlation between the dietary level of RPC and/or

phosphorus and the moisture, fat and ash content in fish bodies. Sinking ash levels in fish

body indicate that the levels of available phosphorus in diets were not sufficient to meet the

dietary requirement of wels catfish. It is known that whole-body ash can be reduced when

carnivorous fish are fed a diet deficient in available phosphorus (Skonberg et al. 1997; Shao et

al. 2008). Although dietary levels of phosphorus were 12.7 to 14.8 g kg-1 and therefore above

established requirement levels for many fish species (NRC 1993) it seems possible that

phosphorus availability from RPC was lower than from fish meal. Antinutritional factors such

as phytic acid, fibre and other complex carbohydrates present in RPC are known to influence

phosphorus availability in fish (Francis et al. 2001). However, as shown above, phytic acid

concentrations in diets R50 and R75 were insignificant. We assume therefore, that phosphorus

availability was mainly reduced by fibre and other complex carbohydrates. The often

increased whole-body lipid content with high dietary levels of vegetable proteins that has

been reported in several fish species (Adelizi et al. 1998; Kaushik et al. 2004) was not

observed here. Instead, the whole-body lipid level tendentially decreased with increasing

dietary level of vegetable protein as reported by Espe et al. (2006) when feeding Atlantic

salmon a diet devoid of fish meal. According to Espe et al. (2006), it seems possible therefore,

that the substitution of fish meal with plant proteins may not give the same results in different

fish species.

In prospective feeding trials with wels catfish and rapeseed protein concentrate it appears

advisory to supplement diets with a phosphorus source such as dicalcium phosphate in order

to overcome problems regarding phosphorus availability. This has been shown to positively

affect dietary phosphorus supply, feed efficiencies and fish growth when diets rich in plant

based proteins are applied (Lee et al. 2010).

In the present study, blood serum parameters were not significantly different between

treatment groups and no correlation with dietary RPC levels was found (Table 2.5). As

suggested by Caruso and Schlumberger (2002) blood serum parameters can be used to

estimate the health status of fish. Our results indicate that fish did not suffer from malnutrition

and that dietary nutrient supply was sufficient to support growth and maintain average body

development in all feeding groups. However, found individual blood values of wels are highly

variable as attested by a high standard deviation. This was also found by Caruso and

Schlumberger (2002) who established a baseline blood haematocrit value of 0.25 ± 0.01 for

29

wels catfish. The fish (individual weight 55 to 250 g) were reared in a closed warm water

system (26-27°C) and were fed at 1 % of their biomass per day. The haematocrit baseline

value corresponds to the value determined in our study (Table 2.5). Detected blood serum

values of triglycerides, glucose and protein, however, differ from values cited by Jirásek et al.

(1998). They monitored blood serum values of one-year old wels catfish (individual weight

752 and 1288 g) held in heated effluent water of a power station. Fish were fed on different

pelleted diets that contained 400 to 490 g CP kg-1 and up to 220 g CF kg-1. Average blood

serum values determined were total lipids 9.95 mM L-1, glucose 49.39 mM L-1 and total

protein 29.89 g L-1. We assume that differences between values published by Jirásek et al.

(1998) and in the present study reflect the different compositions of used diets and possible

differences in starvation time before sampling. It is known that blood serum values generally

represent the nutrient composition of a diet (Jirásek et al. 1998) and that starvation time

before sampling can have a significant influence on plasma glucose, triglycerides and protein

concentrations (Shi et al. 2010).

In conclusion, wels catfish accept diets formulated to contain 122 g kg-1 of the used RPC as

fish meal replacement. At higher dietary RPC inclusion diet taste became undesirable for wels

catfish, thereby reducing feed intake and fish growth. Antinutritional factors present in RPC

might also have reduced dietary phosphorus and amino acid availability with negative effects

on feed efficiencies and fish growth. To overcome difficulties with diet taste and nutrient

availability we suggest the use of ANF free rapeseed protein isolates in prospective feeding

trials.

References

Adelizi, P.D., Rosati, R.R., Warner, K., Wu, Y.V., Muench, T.R., White, M.R., Brown, P.B.,

1998. Evaluation of fish-meal free diets for rainbow trout, Oncorhynchus mykiss.

Aquaculture Nutrition 4, 255–262.








30






83, 653–664.

Burel, C., Boujard, T., Kaushik, S.J., Boeuf, G., Mol, K.A., Van der Geyten, S., Darras, V.M.,

Kühn, E.R., Pradet-Balade, B., B. Quérat, B., Quinsac, A., Krouti, M., Ribaillier, D.,

2001. Effects of rapeseed meal glucosinolates on thyroid metabolism and feed

utilization in rainbow trout. Gen. Comp. Endocr. 124, 343–358.

Caruso, D., Schlumberger, O., 2002. Haematological baseline values of juvenile Silurus

glanis (Siluridae) reared in intensive conditions. Cybium 26, 65-70.







263-274.




Espe, M., Lemme, A., Petri, A., El-Mowafi, A., 2006. Can Atlantic salmon (Salmo salar)

grow on diets devoid of fish meal? Aquaculture 255, 255–262.

Espe, M., Lemme, A., Petri, A., El-Mowafi, A., 2007. Assessment of lysine requirement for

maximal protein accretion in Atlantic salmon using plant protein diets. Aquaculture 263,

168–178.






31





38, 551-579.

Glencross, B., Hawkins, W., Curnow, J., 2004. Nutritional assessment of Australian canola




Hajen, W.E., Higgs, D.A., R.M. Beames, R.M., Dosanjh, B.S., 1993. Digestibility of various

feedstuffs by post-juvenile chinook salmon (Oncorhynchus tshawytscha) in sea water. 2.

Measurement of digestibility. Aquaculture 112, 333-348.

Higgs, D.A., McBride, J.R., Markert, J.R., Dosanjh, B.S., Plotnikoff, M.D., Clarke, W.C.,

1982. Evaluation of Tower and Candle rapeseed (canola) meal and Bronowski rapeseed

protein concentrate as protein supplements in practical dry diets for juvenile Chinook

salmon (Oncorhynchus tshawytscha). Aquaculture 29, 1-31.

Hossain, M.A., Jauncey, K., 1993. The effect of varying dietary phytic acid, calcium and

magnesium levels on the nutrition of common carp, Cyprinus carpio. In: S.J. Kaushik &

Luquent, P. (eds.) Fish Nutrition in Practice, pp. 705-715. Proceedings of International

Conference, Biarritz, France, June 24–27, 1991.

Jakubowski, M., Kunysz, E., 1979. Anatomy and morphometry of the olfactory organ of the

wels Silurus glanis L. (Siluridae, Pisces). Z. Mikrosk. Anat. Forsc. – Abt. 2, 93, 728-

735.

Jirásek, J., Mareš, J., Palíková, M., 1998. Haematological and biochemical indices of blood in

wels (Silurus glanis L.) from intensive aquaculture. Acta Vet. Brno 67, 227-233.

Kaushik, S.J., Covès, D., Dutto, G., Blanc, D., 2004. Almost total replacement of fish meal by

plant protein sources in the diet of a marine teleost, the European seabass,

Dicentrarchus labrax. Aquaculture 230, 391–404.

Kousoulaki, K., Albrektsen, S., Langmyhr, E., Olsen, H.J., Campbell, P., Aksnes, A., 2009.

The water soluble fraction in fish meal (stickwater) stimulates growth in Atlantic salmon

(Salmo salar L.) given high plant protein diets. Aquaculture 289, 74–83.

Lee, K.-J., Powell, M.S., Barrows, F.T., Smiley, S., Bechtel, P., Hardy, R.W., 2010.

Evaluation of supplemental fish bone meal made from Alaska seafood processing

32

byproducts and dicalcium phosphate in plant protein based diets for rainbow trout

(Oncorhynchus mykiss). Aquaculture 302, 248–255.


glucosinolates and their antinutritional effects: Part 7. Processing. Die Nahrung 39, 32–

41.

Mazurkiewicz, J., Przybyl, A., Golski, J., 2008. Evaluation of selected feeds differing in

dietary lipids levels in feeding juveniles of wels catfish, Silurus glanis L. Acta Ichthyo.

Pisca., 38, 91-96.





NRC (National Research Council) (1993) Nutrient Requirements of Fish. National Academy

Press, Washington, DC, USA.




Shao, Q., Ma, J., Xu, Z., Hu, W., Xu, J., Xie, S., 2008. Dietary phosphorus requirement of

juvenile black seabream, Sparus macrocephalus. Aquaculture 277, 92–100.

Shi, X. Zhuang, P., Zhang, L., Chen, L., Xu, B., Feng, G., Huang, X., 2010. Optimal

starvation time before blood sampling to get baseline data on several blood biochemical

parameters in Amur sturgeon, Acipenser schrenckii. Aquaculture Nutrition 16, 544-548.

Skonberg, D.I., Yogev, L., Hardy, R.W., Dong, F.M., 1997. Metabolic response to dietary

phosphorus intake in rainbow trout (Oncorhynchus mykiss). Aquaculture 157, 11–24.

Spinelli, J., Houle, C.R., Wekell, J.C., 1983. The effect of phytates on the growth of rainbow

trout (Salmo gairdneri) fed purified diets containing varying quantities of calcium and

magnesium. Aquaculture 30, 71–83.

Storebakken, T., Shearer, K.D., Roem, A.J., 1998. Availability of protein, phosphorus and

other elements in fishmeal, soy protein concentrate and phytase-treated soy protein-

concentrate-based diets to Atlantic salmon, Salmo salar. Aquaculture 161, 365–379.

Thiessen, D.L., Campbell, G.L., Adelizi, P.D., 2003. Digestibility and growth performance of

juvenile rainbow trout (Oncorhynchus mykiss) fed with pea and canola products.


33

Thiessen, D.L., Maenz, D.D., Newkirk, R.W., Classen, H.L., Drew, M.D., 2004. Replacement

of fish meal by canola protein concentrate in diets fed to rainbow trout (Oncorhynchus

mykiss). Aquaculture Nutrition 10, 379–388.

Torstensen, B.E., Espe, M., Sanden, M., Stubhaug, I., Waagbø, R., Hemre, G.-I., Fontanillas,

R., Nordgarden, U., Hevrøy, E.M., Olsvik, P., Berntssen, M.H.G., 2008. Novel

production of Atlantic salmon (Salmo salar) protein based on combined replacement of

fish meal and fish oil with plant meal and vegetable oil blends. Aquaculture 285, 193-

200.

Tripathi, M.K., Agrawal, I.S., Sharma, S.D., 2000. Effect of physio-chemical treatments on

glucosinolates content of various rapeseed–mustard meals. Indian J. Anim. Nutr. 17,

211–216.

Tyagi, A.K., 2002. Influence of water soaking of mustard cake on glucosinolate hydrolysis.

Animal Feed Science and Technology 99, 215–219.




Webster, C.D., Tiu, L.G., Tidwell, J.H., Grizzle, J.M., 1997. Growth and body composition of

channel catfish (Ictalurus punctatus) fed diets containing various percentages of canola

meal. Aquaculture 150, 103-112.

34

Chapter 3: Austausch von Fischmehl durch Rapsproteinkonzentrat in

Futtermitteln für Steinbutt (Psetta maxima L.)

H. Slawski1,2, H. Adem3, R.-P. Tressel3, K. Wysujack4,

Y. Kotzamanis5 und C. Schulz1,2


2Institut für Tierzucht und Tierhaltung, Christian-Albrechts-Universität zu Kiel, D-24098 Kiel




5Hellenic Centre for Marine Research, Institute of Aquaculture, Ag.Kosmas, Helleniko,

Athens, Hellas, Greece

Published in: Züchtungskunde (2011) 83, (6), 451-460.

35

Zusammenfassung

Rapsproteinkonzentrat wurde als Alternative zu Fischmehl in Futtermitteln für juvenile

Steinbutt getestet. Triplikate Fischgruppen erhielten isonitrogene (58,1 ± 0,9 % RP)

Futtermischungen mit gleichem Gehalt an Bruttoenergie (21,5 ± 0,3 MJ kg−1), in denen 0, 33

und 66 % (bezeichnet als R0, R33 oder R66) Fischmehl durch Rapsproteinkonzentrat ersetzt

wurde. Am Ende des Fütterungszeitraumes (84 Tage) zeigten Fische, welche die Mischung

R0 oder R33 erhalten hatten, signifikant bessere Zuwachsraten und Futteraufnahmen als

Fische, die Mischung R66 erhielten. Verringerte Futteraufnahme und folglich schlechteres

Fischwachstum in der R66 Gruppe beruhen vermutlich auf geschmacklichen

Beeinträchtigungen der Futtermischung hervorgerufen durch Glucosinolate im

Rapsproteinkonzentrat. Für die Kontrollfuttermischung wurde gegenüber den -mischungen

R33 und R66 eine signifikant bessere Futterverwertung festgestellt. Dies geht vermutlich auf

geringere Protein-, Aminosäuren- und Phosphorverfügbarkeit in den Mischungen R33 und

R66 zurück.

36

Einleitung

Der stagnierenden Produktionsmenge an Fischmehl, der wichtigsten Proteinquelle in

Fischfuttermitteln, steht eine steigende Nachfrage des Aquakultursektors nach diesem

Rohstoff gegenüber. Um der dadurch eintretenden Verteuerung des Fischmehls und folglich

der Fischfuttermittel entgegenzuwirken, benötigt die Industrie für die Herstellung von

Futtermitteln hochwertige Fischmehlalternativen. Hierbei sind pflanzliche Proteinquellen von

höchster Relevanz (Gatlin et al. 2007). Mit einer weltweiten Produktionsmenge von 61,6 Mio

t in 2009 ist Raps (Brassica napus L., B. campestris L.) eine der bedeutendsten Ölsaaten

(FAO 2010). Entsprechend ist der bei der Rapsölgewinnung entstehende Rapsölkuchen ein

weit verfügbares Nebenprodukt. Dieser wurde sowohl in unbehandelter Form als auch nach

Bearbeitung mit organischen Lösungsmitteln in Futtermitteln für verschiedene Fischarten

getestet, darunter Regenbogenforelle (Burel et al. 2000ab; Thiessen et al. 2004; Shafaeipour et

al. 2008), Tilapia (Davies et al. 1990), Karpfen (Dabrowski und Kozlowska 1981) und

Steinbutt (Burel et al. 2000ac). Es zeigte sich, dass insbesondere antinutritive Faktoren (ANF)

wie Glucosinolate, Phytinsäure, und unverdauliche Kohlenhydrate die Eignung einfacher

Rapsprodukte für die Fischernährung beeinträchtigen (Francis et al. 2001). Der Gehalt an

ANF in einfachen Rapsprodukten kann allerdings durch verschiedene Bearbeitungsverfahren

verbessert werden. Hierzu zählen eine Wärmebehandlung der Saat, das Schälen der Saat, die

Nutzung organischer Lösungsmittel (z.B. Hexan) nach dem Ölpressen,

Dekantierungsverfahren sowie Ultrafiltration in Wasser gelöster Proteinbestandteile (Fenwick

et al. 1986; Anderson-Hafermann et al. 1993; Mawson et al. 1995; Mwachireya et al. 1999;

Tyagi 2002). In der vorliegenden Untersuchung wurde ein eigens angefertigtes

Rapsproteinkonzentrat mit einem Proteingehalt von 71 % als Fischmehlersatz in Futtermitteln

für Steinbutt (Psetta maxima L.) getestet. Anhand eines Fütterungsversuches sollte untersucht

werden, ob das hochaufgereinigte Rapsproteinkonzentrat einen höheren Fischmehlaustausch

im Futtermittel als bisher untersuchte, relativ einfach verarbeitete Rapsprodukte ermöglicht.

Material und Methoden

Herstellung der Futtermischungen

Es wurden drei Versuchsfuttermischungen formuliert. In diesen wurde Fischmehl zu 0, 33

und 66 % (bezeichnet als R0, R33 oder R66) durch Rapsproteinkonzentrat (RPK)

ausgetauscht. Das RPK wurde im PPM e.V., Deutschland, hergestellt (Slawski et al. im

Druck). Die Nährstoffzusammensetzung des RPK ist in Tab. 3.1 aufgeführt. Basierend auf

37

den Aminosäurengehalten der verwendeten Rohstoffe wurden die Mischungen so konzipiert,

dass der Aminosäurenbedarf mariner Fische gedeckt war (NRC 1993). Vitamine und

Mineralstoffe wurden den Mischungen dem Bedarf mariner Fische entsprechend beigefügt

(NRC 1993). Die drei Futtermischungen waren isonitrogen (58,1 ± 0,9 % RP) und wiesen

gleiche Gehalte an Bruttoenergie auf (21,5 ± 0,3 MJ kg−1). Die Mischungen wurden zu Pellets

mit 4 mm Durchmesser gepresst (L 14-175, AMANDUS KAHL, Reinbek, D).

Futtermittelformulierung und -zusammensetzung sind in Tab. 3.2 und 3.3 dargestellt.

Tabelle 3.1: Nährstoff- und Aminosäurenzusammensetzung von Fischmehl und Rapsproteinkonzentrat

Fischmehl Rapsproteinkonzentrat

Rohwasser

(% der lufttrockenen Substanz)

8,4 5,6

Nährstoffzusammensetzung (% der Trockensubstanz)

Rohprotein 69,0 71,2

Rohfett 7,0 0,6

Rohasche 20,7 16,1

Phosphor 2,9 0,9

Rohfaser 0,5 0,5

NfEa 2,9 11,6

Bruttoenergieb (MJ kg-1) 19,9 19,4

Aminosäurenzusammensetzung (% der Trockensubstanz)

Arginin 4,03 4,83

Histidin 1,38 2,75

Isoleucin 2,50 2,70

Leucin 4,45 5,39

Lysin 4,52 5,60

Methionin + Cystein 2,19 4,64

Phenylalanin 2,43 2,56

Threonin 2,69 2,97

Valin 3,07 3,67

Antinutritive Faktoren

Glucosinolate (µmol g-1) 1,32

Phytinsäure (g 100g-1) 1,77 aN-freie Extraktstoffe = 100 – (%Rohprotein + %Rohfett + %Asche + %Rohfaser) bBerechnung: Rohprotein = 23,9 MJ kg-1; Rohfett = 39,8 MJ kg-1; Rohfaser, NfE: 17,6 MJ kg-1

38

Aufbau des Fütterungsversuchs

Der Fütterungsversuch wurde am Institut für Fischereiökologie (Außenstelle Ahrensburg) des

Johann Heinrich von Thünen Institutes, Bundesinstitut für Ländliche Räume, Wald und

Fischerei, durchgeführt. Juvenile Steinbutt aus der Ejsing Seafarm (Vinderup, DK) wurden

verwendet. Jeweils 14 Fische wurden in eines der neun Versuchsbecken (96 L; Grundfläche

4800 cm2) gesetzt. Angeschlossen an eine Salzwasserkreislaufanlage wurden die Becken mit

2 L min-1 künstlichem Meerwasser versorgt (Temperatur: 17,3 ± 1,1 °C; O2: 7,8 ± 0,2 mg L-1;

Salinität: 25,8 ± 0,5 ‰; pH-Wert: >7,8; NH4+: <0,07 mg L-1; NO2

-: <0,13 mg L-1). Die

Photoperiode entsprach der für den Standort (53° 41' 0" N) natürlichen Rhythmik von Juli bis

Oktober. Für eine zweiwöchige Anpassungsphase erhielten die Fische die Mischung R0 in

einer täglichen Fütterung bis zur scheinbaren Sättigung. Nach der Anpassungsphase wurden

die Fische für zwei Tage genüchtert und ihr Anfangsgewicht wurde bestimmt (73,5 ± 0,7 g).

Für einen Zeitraum von 84 Tagen wurden dann in drei Wiederholungen je Behandlung

Gruppen der Fische mit den Versuchsfuttermischungen einmal täglich (8:30 Uhr) bis zur

scheinbaren Sättigung gefüttert. Zu Beginn und am Ende des Versuches wurden zwei Fische

pro Becken für die Bestimmung der Ganzkörperzusammensetzung entnommen.

Tabelle 3.2: Formulierung der Versuchsfuttermittel (g kg-1)

Futterkomponenten R0 R33 R66

Heringsmehl 450 300 150

Rapsproteinkonzentrat 0 145 295

Sojaproteinkonzentrat 100 100 100

Garnelenmehl 85 85 85

Blutmehl 95 95 95

Kartoffelstärke 90 90 90

Dextrose 75 70 60

Fischöl 55 65 75

Vitamin/Mineralmixa 20 20 20

Weizengluten 30 30 30 aAA-Mix 507101, Vitfoss, Gråsten, Dänemark

39

Tabelle 3.3: Nährstoff- und Aminosäurenzusammensetzung der Versuchsfuttermittel

Inhaltsstoffe R0 R33 R66

Nährstoffzusammensetzung (% der Trockensubstanz)

Rohprotein 57,5 57,7 59,1

Rohfett 10,8 11,1 10,8

Rohasche 13,6 12,3 10,8

Phosphor 1,7 1,4 1,1

Rohfaser 0,8 0,7 0,5

NfEa 17,4 18,2 18,7

GEb (MJ kg-1) 21,2 21,6 21,8

Aminosäurenzusammensetzung (g 100g Rohprotein-1)

Arginin 5,97 6,13 5,91

Histidin 2,47 3,15 3,16

Isoleucin 3,15 3,23 3,07

Leucin 7,94 8,25 7,93

Lysin 5,81 6,21 5,99

Methionin + Cystein 2,39 2,66 2,84

Phenylalanin 4,37 4,49 4,23

Threonin 3,75 3,85 3,66

Valin 5,19 5,34 5,05

Antinutritive Faktorenc

Glucosinolate (µmol g-1) - 0,20 0,40

Phytinsäure (g 100g-1) - 0,25 0,52 aN-freie Extraktstoffe = 100 – (%Rohprotein + %Rohfett + %Asche + %Rohfaser) bBerechnung: Rohprotein = 23,9 MJ kg-1; Rohfett = 39,8 MJ kg-1; Rohfaser, NfE: 17,6 MJ kg-1

ckalkuliert anhand der Konzentrationen in den Rohstoffen (Tabelle 1)

Durchführung der Nährstoffanalysen

Futtermittel und homogenisierte Fischkörper wurden in Duplikaten auf ihren Nährstoffgehalt

untersucht. Der Feuchtegehalt wurde über Trocknung der Materialien in einem Ofen bei 105

°C bis zu einem konstanten Gewicht ermittelt. Der Aschegehalt wurde mittels

Gewichtsverlust nach zweistündiger Veraschung im Muffelofen bei 550 °C bestimmt. Der

Fettgehalt wurde über Etherextraktion (Soxtec HT6, Tecator, Höganäs, S) und der

Gesamtstickstoffgehalt mit dem Kjeldahl-Verfahren (Rohprotein = N × 6,25; Kjeltec Auto

System, Tecator, Höganäs, S) erfasst. Die Aminosäurenzusammensetzung wurde nach dem

bei Tzovenis et al. (2009) geschilderten Verfahren ermittelt. Dabei wurde der

40

Tryptophangehalt nicht bestimmt. Der Glucosinolatgehalt im Rapsproteinkonzentrat wurde

anhand der Methodik in der EC-Gazette (1864/90 Nr. L 170/28, 3.7.90) bestimmt. Der Gehalt

an Phytinsäure wurde anhand des Verfahrens von Harland und Oberleas (1986) ermittelt.

Berechnungen und statistische Verfahren

Fischwachstum, Futterverwertung und physiologische Parameter wurden anhand folgender

Formeln bestimmt:

Spezifische Wachstumsrate (SWR, % Tag-1) = (ln Endgewicht – ln Anfangsgewicht) × 100 /

Fütterungstage

Futterquotient (FQ) = g Trockenfutteraufnahme / g Gewichtszunahme

Proteinwirkungsverhältnis (PER) = g Gewichtszunahme / g Proteinaufnahme

Fultonscher Konditionsfaktor (K) = g Körpergewicht / cm Gesamtlänge3 × 100

Überlebensrate (%) = (Anzahl Fische Versuchsbeginn – Anzahl Totfische) / Anzahl Fische

Versuchsbeginn × 100

Erhobene Daten wurden mithilfe der Statistiksoftware SPSS 17.0 ausgewertet. Die Daten

wurden mittels Kolmogoroff Smirnov Test auf Normalverteilung untersucht und falls nötig

transformiert. Nach der Varianzanalyse der Mittelwerte in einer ANOVA wurden signifikant

unterschiedliche Mittelwerte mithilfe des Tukey Tests identifiziert (P<0,05).

Ergebnisse

Wachstumsleistung, Futterverwertung und Körperzusammensetzung

Während Wachstumsleitungen und Futteraufnahme in den Gruppen R0 und R33 keine

signifikanten Unterschiede zeigten, waren Wachstum und Futteraufnahme in der Gruppe R66

signifikant schlechter. Die Futterverwertung war in den Gruppen R33 und R66 signifikant

schlechter als bei der Kontrollfuttermischung. Allerdings war die Verwertung der Mischungen

R33 und R66 einheitlich. Der Konditionsfaktor war signifikant geringer in Gruppe R66

gegenüber den Gruppen R0 und R33. Die Überlebensrate zeigte genauso wie die

Körperzusammensetzung der Tiere keine gruppenübergreifenden signifikanten Unterschiede

(Tab. 3.4 und 3.5).

41

Tabelle 3.4: Ganzkörperzusammensetzung der Steinbutt nach der Fütterungsperiode

R0 R33 R66

Rohwasser 79,1 ± 1,2 78,4 ± 1,5 79,5 ± 3,0

Rohprotein 14,5 ± 0,8 15,0 ± 0,9 13,9 ± 1,3

Rohfett 2,6 ± 0,2 2,8 ± 0,5 3,4 ± 0,8

Rohasche 4,1 ± 0,9 3,9 ± 0,5 3,7 ± 0,7

Ganzkörperzusammensetzung zu Versuchsbeginn: Wasser 76,8 %, Rohprotein 15,%, Rohfett 3,6%, Rohasche 3,5%

Tabelle 3.5: Wachstumsparameter und Futterverwertung der Steinbutt nach dem Fütterungsversuch

R0 R33 R66

Anfangsgewicht (g/Fisch) 73,1 ± 1,0 74,1 ± 0,4 73,3 ± 0,4

Endgewicht (g/Fisch) 147,5a ± 10,3 145,0a ± 7,5 122,2b ± 4,5

SWR (%/Tag) 0,83a ± 0,07 0,80a ± 0,07 0,61b ± 0,05

Futteraufnahme (g/Fisch) 73,9a ± 8,2 82,6a ± 10,8 58,2b ± 1,1

FQ (g/g) 1,00a ± 0,06 1,16b ± 0,03 1,20b ± 0,09

PER 1,75a ± 0,11 1,49b ± 0,04 1,42b ± 0,11

K 1,80a ± 0,16 1,76a ± 0,12 1,56b ± 0,13

Überlebensrate (%) 94,4 ± 9,6 100 ± 0,0 100 ± 0,0

Werte in denselben Zeilen mit unterschiedlichen Indices sind signifikant verschieden (Tukey Test; P<0,05)

Diskussion

Während die Einsetzbarkeit einfacher Rapsprodukte als Inhaltsstoff in Fischfuttermitteln

hinlänglich untersucht wurde (Dabrowski und Kozlowska 1981; Davies et al. 1990; Burel et

al. 2000a,b,c; Thiessen et al. 2004; Shafaeipour et al. 2008), ist über die Eignung qualitativ

hochwertiger Rapsprodukte, die mit Fischmehl vergleichbare Proteingehalte besitzen, wenig

bekannt. Higgs et al. (1982) ersetzten erfolgreich 25 % des Futtermittelproteins aus einer

Fischmehlkontrollmischung für juvenile Königslachse durch Rapsproteinkonzentrat (61,3 %

RP) ohne die Wachstumsraten und Proteinverwertung der Fische gegenüber einer

Kontrollgruppe zu beeinträchtigen. Die Ergebnisse der vorliegenden Untersuchung zeigen,

dass 33 % des Fischmehls im Futtermittel für Steinbutt durch RPK ersetzt werden kann, ohne

die Futteraufnahme oder das Fischwachstum zu verringern. Bei einem 66 %igen

42

Fischmehlaustausch durch RPK zeigten die Fische geringere Futterakzeptanz und reduzierte

Futteraufnahme. Dies geht vermutlich auf geschmackliche Beeinträchtigung des Futters durch

RPK zurück. Es scheint, dass das in den Mischungen als Geschmacksverbesserer eingesetzte

Blutmehl und Garnelenmehl nicht ausreichte, um durch RPK auftretende negative Effekte auf

den Futtermittelgeschmack auszugleichen. Der bittere Geschmack von

Glucosinolatmetaboliten wie Isothiocyanaten und Vinyloxazolidinethionen, die in

Rapsmehlen enthalten sind, kann die Futtermittelakzeptanz von Fischen verschlechtern. Dies

wurde für Regenbogenforelle und Steinbutt bei Futterglucosinolatgehalten von 7,3 bzw. 18,7

µmol g-1 festgestellt (Burel et al. 2000bc). Das RPK in unserer Studie enthielt 1,3 µmol

Glucosinolate g-1 (Tab. 3.1). Entsprechend betrug der höchste Futterglucosinolatgehalt 0,4

µmol g-1 in Mischung R66 (Tab. 3.3). Dieser Wert unterschreitet die laut Burel et al.

(2000b,c) problematische Futterglucosinolatmenge für Regenbogenforelle und Steinbutt.

Allerdings ging trotz der niedrigen Glucosinolatkonzentrationen in den Mischungen von

Futtermittel R66 ein für Glucosinolate typisch senfiger Geruch aus. Für weitere

Fütterungsversuche mit RPK erscheint es sinnvoll, Futtermischungen mit zusätzlichen

Geschmacksverbesserern anzureichern. Beispielsweise wurden Fischproteinhydrolysat,

Tintenfischhydrolysat oder Krillmehl als effektive Geschmacksverbesserer und Quellen von

Aminosäuren und Mineralien identifiziert, wenn Futtermittel mit niedrigem Fischmehlanteil

an carnivoren Fischarten verfüttert wurden (Espe et al. 2006; Torstensen et al. 2008).

Während die verringerte Akzeptanz gegenüber der Mischung R66 vermutlich auf

geschmackliche Beeinträchtigungen zurückging, dürfte für die schlechtere Verwertung der

Mischungen R33 und R66 eine Beeinträchtigung der Nährstoff- und

Mineralstoffverfügbarkeit ursächlich sein.

Rohfaser und komplexe Kohlenhydrate können die Verdaulichkeit und Energieverfügbarkeit

von Futtermitteln für Steinbutt verringern (Burel et al. 2000a,c). Der Anteil an Rohfaser +

NfE des in der vorliegenden Untersuchung verwendeten RPK war mit 12,1 % deutlich höher

als bei dem verwendeten Fischmehl (3,4 %). Es erscheint daher möglich, dass mit steigendem

RPK-Anteil in den Futtermischungen R33 und R66 der Gehalt an verdaulicher Energie

zurückging, was zu schlechterer Futterverwertung gegenüber der Kontrollmischung führte.

Neben dem Gehalt an verdaulicher Energie könnte auch das geringere Phosphorangebot durch

den Austausch von Fischmehl mit RPK ursächlich für verringerte Futterverwertung sein.

Obgleich der Futtermittel-Phosphorbedarf des Steinbutts unserem Wissen nach bisher nicht

ermittelt wurde, lagen die Phosphorgehalte in den Futtermitteln mit über einem Prozent

43

oberhalb des bekannten Bedarfswertes für die meisten Fische (NRC 1993). Es ist aber

bekannt, dass erhöhte Phosphorgehalte im Futtermittel die Futterverwertung bei Lachs,

Dorsch und Wolfsbarsch verbessern können (Vielma und Lall 1998; Roy und Lall 2003;

Oliva-Teles und Pimentel-Rodrigues 2004). Dagegen können antinutritive Faktoren wie

Phytinsäure, Rohfaser und andere komplexe Kohlenhydrate in Rapsproteinprodukten die P-

Verfügbarkeit in Fischfuttermitteln verringern (Mwachireya et al. 1999; Francis et al. 2001).

Das in der vorliegenden Studie getestete RPK enthielt 1,77 g Phytinsäure 100g-1, was 0,25

und 0,52 g Phytinsäure 100g-1 in den Mischungen R33 bzw. R66 entspricht (Tab. 3.3).

Spinelli et al. (1983) beobachteten verringertes Wachstum bei Regenbogenforellen, die

Futtermittel mit 0,5 g Phytinsäure 100g-1 erhielten. In Futtermitteln für Karpfen führte

synthetische Phytinsäure bei Konzentrationen von 0,5 bis 1,0 g 100g-1 zu geringeren

Wachstumsleistungen (Hossain und Jauncey 1993).

Es scheint daher möglich, dass die Phytinsäurekonzentrationen in den Mischungen R33 und

R66 negativen Einfluss auf die P-Verfügbarkeit hatten und somit die Futterverwertung und

das Wachstum verschlechterten. Der signifikant geringere Konditionsfaktor in Gruppe R66

veranschaulicht zusätzlich die schlechtere körperliche Entwicklung der Steinbutt. Tendenziell

variierende Asche- und Fettgehalte in der Körperzusammensetzung bekräftigen die

Vermutung einer geringen P-Verfügbarkeit bei hohem Fischmehlaustausch (Tab. 3.4).

Erhalten carnivore Fische ein Futtermittel, das reich an pflanzlichen Proteinen ist und relativ

geringe Mengen an verfügbarem Phosphor enthält, können sinkende Körperaschegehalte und

steigende –fettgehalte die Folge sein (Skonberg et al. 1997; Adelizi et al. 1998; Kaushik et al.

2004). Der für diesen Effekt verantwortliche Mechanismus ist unklar. Es wird angenommen,

dass eine Akkumulation von Fettsäuren durch eine gestörte β-Oxidation (Takeuchi und

Nakazoe 1981) oder oxidative Phosphorylierung durch P-Mangel bewirkt wird. Dies könne zu

einer Beeinträchtigung des Citratzyklus und einer Anreicherung von Acetyl-CoA und

schliesslich zu einer gesteigerten Fettsäuresynthese führen (Skonberg et al. 1997). Um die

getroffenen Annahmen zum Einfluss der Phosphorverfügbarkeit auf die Futterverwertung

beim Steinbutt zu bekräftigen, wären Informationen zum tatsächlichen Phosphorbedarf dieser

Fischart vonnöten.

Zusammenfassend ist festzustellen, dass Steinbutt bei Aufnahme eines Futtermittels, bei dem

33 % des Fischmehls durch Rapsproteinkonzentrat ersetzt wurden, keine Verschlechterung

der Wachstumsleistungen zeigten. Bei höherem Einsatz von RPK als Fischmehlersatz wurde

die Futtermittelakzeptanz für Steinbutt stark beeinträchtigt und folglich das Wachstum

44

negativ beeinflusst. Zudem dürften Phytinsäuregehalte im RPK zu einer Beeinträchtigung der

P-Verfügbarkeit geführt haben, weshalb zusätzlich Futterverwertung und Fischwachstum

verringert wurden. In folgenden Fütterungsversuchen mit RPK sollten daher zusätzliche

Geschmacksverbesserer eingesetzt werden, um eine bessere Futteraufnahme zu erzielen.

Literatur










Kühn, E.R., 2000b. Dietary low glucosinolate rapeseed meal affects thyroid status and


83, 653–664.

Burel, C., Boujard, T., Kaushik, S.J., Boeuf, G., Van Der Geyten, S., Mol, K.A., Kühn, E.R.,

Quinsac, A., Krouti, M., Ribaillier, D., 2000c. Potential of plant-protein sources as fish




heated waters. I. Growth of fish and utilization of the diet. In: K. Tiews (ed.),


263-274.







45


processing on the antinutrient content of rapeseed. Journal of the Science of Food and

Agriculture 37, 735-741.



Gatlin III, D.M., Barrows, F.T., Brown, P., Dabrowski, K., Gaylord, G.T., Hardy, R.W.,


Skonberg, D., Souza, J.E., Stone, D., Wilson, R., Wurtele, E., 2007. Expanding the


38, 551-579.

Harland, B.F., Oberleas, D., 1986. Anion-exchange method for determination of phytate in

foods: Collaborative study. J. Assoc. Off. Anal. Chem. 69, 667.






magnesium levels on the nutrition of common carp, Cyprinus carpio. In: S.J. Kaushik

and P. Luquet (eds.) Fish Nutrition in Practice, pp. 705-715. Proceedings of

International Conference, 24–27, 1991, Biarritz.

Kaushik, S.J., Covés, D., Dutto, G., Blanc, D., 2004. Almost total replacement of fish meal by





41.







46

Oliva-Teles, A., Pimentel-Rodrigues, A.M., 2004. Phosphorus requirement of European sea

bass (Dicentrarchus labrax L.) juveniles. Aquaculture Research 35, 636–642.

Roy, P.K., Lall, S.P., 2003. Dietary phosphorus requirement of juvenile haddock

(Melanogrammus aeglefinus L.). Aquaculture 221, 451-468.






Slawski, H., Adem, H., Tressel, R.-P., Wysujack, K., Koops, U., Wuertz, S., Schulz, C. (xxx):

Replacement of fish meal with rapeseed protein concentrate in diets fed to wels catfish

(Silurus glanis L.). Aquaculture Nutrition (article in press).




Takeuchi, M., Nakazoe, J., 1981. Effect of dietary phosphorus on lipid content and its

composition in carp. Bulletin of the Japanese Society of Scientific Fish. 47, 347–352.








200.


Animal Feed Science and Technology 99, 215–219.



coastal lagoons for use in mariculture. Journal of Applied Phycology 21, 457–469.

Vielma, J., Lall, S.P., 1998. Phosphorus utilisation by Atlantic salmon (Salmo salar) reared in

freshwater is not influenced by higher dietary calcium intake. Aquaculture 160, 117-

128.

47

Chapter 4: Total fish meal replacement with rapeseed protein concentrate

in diets fed to rainbow trout (Oncorhynchus mykiss W.)

H. Slawski1, 2, H. Adem3, R.-P. Tressel3, K. Wysujack4, U. Koops4, Y. Kotzamanis5,

S. Wuertz1 and C. Schulz1, 2



D-24098 Kiel




5Hellenic Centre for Marine Research, Institute of Aquaculture, Ag.Kosmas, Helleniko,

Athens, Hellas, Greece

Published in: Aquaculture International (2011), http://www.doi 10.1007/s10499-011-9476-2

48

Abstract

The potential of rapeseed protein concentrate as fish meal alternative in diets for rainbow

trout (initial average weight 37.8 ± 1.4 g) was evaluated. Nine experimental tanks of a

freshwater flow through system were stocked with 12 fish each. Triplicate groups of fish

received isonitrogenous (47.9 ± 0.5 % CP) and isoenergetic (22.4 ± 0.2 kJ g-1) experimental

diets with 0, 66 and 100 % of fish meal substituted with rapeseed protein concentrate (71.2 %

CP). As the amino acid profile of rapeseed protein concentrate was comparable to fish meal

there was no need to supplement experimental diets with synthetic amino acids. At the end of

the 84 days feeding period, fish growth performance, feed intake and feed efficiencies were

not compromised when 100 % of fish meal in the control diet was replaced with rapeseed

protein concentrate, revealing a SGR of 1.19 or 1.10, a FCR of 1.09 or 1.18 and a feed intake

of 78.5 g or 74.7 g in fish fed on the control diet or fed the diet devoid of fish meal,

respectively. Intestinal morphology did not reveal any histological abnormalities in all dietary

groups. Blood parameters including haematocrit, haemoglobin as well as glucose,

triglycerides and total protein in the plasma were not different between treatment groups.

Thus, the rapeseed protein concentrate tested here has great potential as an alternative to fish

meal in rainbow trout diets.

49

Introduction

In search of fish meal alternatives in aqua feeds, rapeseed (including canola) products have

been widely tested as protein sources in diets for several fish species, among them rainbow

trout (Stickney et al. 1996; Burel et al. 2001; Thiessen et al. 2003; Shafaeipour et al. 2008),

tilapia (Davies et al. 1990), wels catfish (Slawski et al. in press), gilthead sea bream (Kissel et

al. 2000) and turbot (Burel et al. 2000ac). In general, experimental results demonstrated that

the nutritional quality of simple rapeseed products such as oilcakes or rapeseed meals is

below that of fish meal despite a well balanced amino acid profile. It was reported, that the

nutritional quality of these simple rapeseed products largely depends on the level of

antinutritional factors. Glucosinolates, phytic acid, phenolic constituents and indigestible

carbohydrates have been recognized as predominant antinutritional factors in rapeseed

products (Francis et al. 2001). The contents of these antinutritional factors in rapeseed

products can be reduced by several processing techniques thereby improving its value for fish

nutrition. Dehulling of seeds and application of high temperatures and organic solvents (e.g.

hexane) during oil extraction as well as sieving of meal decrease content of glucosinolates,

phytate, fibre, cellulose, hemicellulose, sinapin and tannins (Fenwick et al. 1986; Anderson-

Haferman et al. 1993; Mawson et al. 1995; Tripathi and Agrawal 2000). Extraction processes

from meals with methanol-ammonia or ethanol will increase protein level and effectively

remove glucosinolates, phenolic compounds, soluble sugars, such as sucrose and some

oligosaccharides (Naczk and Shahidi 1990; Chabanon et al. 2007).

Sporadically, rapeseed protein products of high quality are being produced in different

countries for use in animal nutrition. However, these products are only produced for test

purposes in small volumes until their potential as alternative protein source is documented.

Besides nutritive quality, costs of production need to be to sufficiently reduced to turn

rapeseed protein products an economically feasible alternative to other protein sources, in

particular fish meal. In the present study, a high quality rapeseed protein concentrate (71 %

CP) gained after hexane and ethanol extraction as well as ultrafiltration was tested as fish

meal replacement in diets of rainbow trout, thereby investigating fish growth performance,

feed efficiencies, and the potential impacts on blood parameters and histological intestinal

morphology.

50


Preparation of experimental diets

Three experimental diets were formulated to replace fish meal with rapeseed protein

concentrate (RPC) at 0, 66, or 100 % (designated as R0, R66, or R100, respectively). Solvent

extracted RPC was obtained from PPM, Magdeburg, Germany. For the production of RPC a

batch of rapeseed (variety Lorenz, Norddeutsche Pflanzenzucht, Hohenlieth, Germany) was

conditioned in a vacuum dryer for 15 minutes at 60-70 °C to inactivate the enzyme

myrosinase. Then rapeseed was cold pressed. To remove residual oil from the oilcake (12.9 %

crude lipid, 31.3 % crude protein) it was crushed into 1-5 mm particle size followed by a

hexane treatment. The treatment lasted for two hours and the incubation temperature was 60

°C. Hexane treated rapeseed meal extract was desolventised under pressure to remove hexane

(< 300 ppm), then rapeseed meal extract was further crushed to a particle size of 0.2 to 0.1

mm. A four step treatment using a 75 % ethanol solution (35 min at 60 °C) aimed to remove

glucosinolates from rapeseed meal extract. This resulted in a residual oil content of the nearly

glucosinolate free rapeseed meal extract of 1.1 % and a protein content of 39.8 %. In the

following, protein was gained through liquid water extraction (rapeseed meal extract 1:15

water). For this, the suspension was heated to 40-45 °C followed by two hours of constant

agitation. Afterwards the suspension was decanted. Following decantation the solvent was

collected and residue material was secondly extracted (residue 1:10 water, 5 % NaCl) at 40-45

°C and one hour contact time under constant agitation. Following extraction the suspension

was decanted. Solvent was collected and residue prepared for a third extraction. Then solvents

of extraction 1, 2 and 3 were collected to remove low-molecular compounds and to

concentrate dissolved proteins by dia- and ultrafiltration. During filtration conductivity was

checked. Protein washing ended, when conductivity was 5-6 mS cm-1, corresponding to a

protein content of 60 %. The gained material was spray dried at 70-80 °C, which led to a

rapeseed protein concentrate with 71 % crude protein content (Tab. 4.1).

51

Table 4.1: Proximate and amino acid composition of fish meal and rapeseed protein concentrate and

concentration of antinutritional factors determined

Fish meal Rapeseed protein concentrate

Proximate composition (% of dry weight)

Crude protein 69.0 71.2

Crude fat 7.0 0.6

Ash 20.7 16.1

Phosphorus 2.89 0.94

Crude fibre 0.5 0.5

NfEa 2.8 11.6

Gross energyb (MJ kg-1) 19.9 19.4

Essential amino acids (g 100g-1 crude protein)

Arginine 5.84 6.78

Histidine 2.00 3.86


Leucine 6.45 7.57

Lysine 6.55 7.87

Methionine

(+ Cysteine)

2.36

3.17

2.36

6.52


Threonine 3.90 4.17

Valine 4.45 5.15

Antinutritional factors

Glucosinolates (µmol g-1) - 1.32

Phytic acid (g 100g-1) - 1.77 aNitrogen free extract = 100 – (%crude protein + %crude fat + %ash + %fibre) bCalculated by: crude protein = 23.9 MJ kg-1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1

Tryptophane was not analyzed

52

Since the available amount of RPC was limited due to high cost of production, not more than

two test diets containing RPC were applied in the feeding trial. Diets were formulated to be

isonitrogenous (47.9 ± 0.5 % CP) and isocaloric (22.4 ± 0.2 MJ g-1). Vitamins and minerals

were added to diet mixes to meet the dietary requirements of rainbow trout (NRC 1993).

Because the amino acid profile of RPC was similar to fish meal, the concentrations of

essential amino acids did not differ considerably between experimental diets. Therefore,

supplementation of synthetic amino acids in experimental diets was not required. The diets

were manufactured to give pellets 4 mm in diameter (L 14-175, AMANDUS KAHL,

Reinbek, Germany).

Based on raw material analysis, dietary concentrations of the prominent antinutritional factors

were calculated. Accordingly, concentrations of glucosinolates were 0.26 or 0.39 µmol g-1 and

concentrations of phytic acid were 0.35 or 0.52 g 100g-1 in diets R66 or R100, respectively.

Dietary formulations, proximate and amino acid compositions are given in Tab. 4.2 and 4.3.

Table 4.2: Formulation of experimental diets (%)

Ingredients R0 R66 R100

Herring meal 30 10 0

Rapeseed protein concentrate 0 20 29.5

Soyprotein concentrate 10 10 10

Blood meal 10 10 10

Crustacean meal 9.5 9 9

Potato starch 12 12 12

Dextrose 12.5 11.5 11.5

Fish oil 11 12.5 13

Vit/MinMixa 2 2 2

Wheat gluten 3 3 3 aAA-Mix 517158 & 508240, Vitfoss, Gråsten, Denmark

53

Table 4.3: Proximate and amino acid composition of experimental diets (% of dry weight)

R0 R66 R100

Proximate composition (% of dry weight)

Crude protein 48.1 48.2 47.3

Crude fat 15.8 16.0 15.7

Ash 11.2 9.7 9.3

Phosphorus 1.27 0.97 0.82

Crude fibre 0.9 0.7 0.7

NfEa 24.1 25.4 27.1

Gross energyb (MJ kg-1) 22.2 22.5 22.4


Requirementc

Arginine 2.61 2.73 2.62 1.15

Histidine 1.31 1.51 1.42 0.58

Isoleucine 1.38 1.39 1.38 1.37

Leucine 3.72 3.77 3.80 1.36

Lysine 2.78 2.80 2.84 2.77

Methionine

(+ Cysteine)

0.86

1.05

0.78

1.15

0.78

1.13

0.80

Phenylalanine 2.06 2.03 2.08 1.20

Threonine 1.65 1.66 1.70 1.03

Valine 2.41 2.42 2.41 1.57

200 295

Antinutritional factorsd

Glucosinolates (µmol g-1) 0.26 0.39

Phytic acid (g 100g-1) 0.35 0.52 aNitrogen free extract = 100 – (%crude protein + %crude fat + %ash + %fibre) bCalculated by: crude protein = 23.9 MJ kg-1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1 cRodehutscord et al. (1997) dcalculations based on ANF levels in raw materials (Tab. 1)

Design of the feeding trial

The feeding trial was conducted at the Institute of Fisheries Ecology of the Johann Heinrich

von Thünen Institute, Federal Research Institute of Rural Areas, Forestry and Fisheries,

Germany, using rainbow trout (Oncorhynchus mykiss W.) that had been hatched in the

institute. Fish were randomly distributed into nine 40 L tanks arranged as flow through system

(water exchange 0.5 % min-1; 10.9 ± 0.5 °C), providing a triplicate (n=12 fish) per diet. For a

54

two week acclimatisation period fish were fed the control diet until apparent satiation once a

day. After the adaptation period, fish were fasted for two days and initial individual weight

was determined (37.8 ± 1.4 g). At the beginning and at the end of the experiment, 2 fish per

tank were bulk sampled for the determination of the initial and final body composition and

stored at -23 ºC. During the feeding trial, fish were fed to apparent satiation (at 8.00 a.m.) for

84 days. Uneaten feed was siphoned from the bottom of the tank, separated from faeces,

weighed back and frozen. Based on the dry matter quantification of waste feed the daily feed

intake was calculated. All diets were assigned by a completely randomized design.

Photoperiod was in accordance to the local natural photoperiod from July to October (53° 41'

0" N). Oxygen concentration (>8 mg L-1) and pH (6.5-7) in the effluent water of the flow

through system were measured daily, ammonia (<0.1 ppm), nitrite (<0.2 ppm) and nitrate

(<50 ppm) were determined photometrically twice a week.

Sampling

At the end of the feeding trial, 8 fish per treatment were randomly sampled: Blood was taken

from the caudal vein with a heparinized syringe (1 ml). The haematocrit was determined in

heparinised micro capillaries upon centrifugation (10000 g, 6 min) in a Haematokrit 210

centrifuge (Hettich, Tuttlingen, Germany). Blood haemoglobin content was determined

photometrically (540 nm) within 48 h from whole blood stored 4°C as haemoglobincyanide

with the haemoglobin FS kit (DiaSys, Germany). Plasma was separated, (5000 g, 5 min)

shock frozen and stored at -80 °C until analysed.

Seven fish from each treatment were killed by intersection of the spinal cord. After careful

dissection, 0.5 cm pieces of intestinal tissue were taken from each of four locations: (1)

stomach; (2) central part of the pyloric caecae; (3) mid-intestine; and (4) distal-intestine.

Histological samples were fixed in Histofix®, dehydrated and embedded in paraffin. Sections

of approximately 5 µm were cut and stained with haematoxylin and eosin (H&E) using

standard histological procedures (Sheehan and Hrapchak 1980). Sections of gastrointestinal

tract were evaluated qualitatively by light microscopy by comparing with experimental

controls for the presence of inflammation and alterations in the architecture of the mucosa,

eosinophilic granular layer and goblet cell count and size (Adelizi et al. 1998).

55

Analysis

Feeds and homogenized fish body samples were analysed for dry matter (105°C, until

constant weight), crude lipid (Soxtec HT6, Tecator, Höganäs, Sweden), crude protein

(N×6.25; Kjeltec Auto System, Tecator, Höganäs, Sweden) and ash (550°C, 2h). Amino acids

were determined after acid hydrolysis (6 N, 110°C, 24 h) and derivatisation by AccQ-Tag

according to the amino acid analysis application solution (Waters, Eschborn, Germany). DL-

Norvaline (Sigma) 2.5 mM was used as standard. UPLC was performed on an Acquity system

(Waters, Eschborn, Germany) equipped with PDA detector set at 260 nm. Column used was

BEH C18 column (100 mm × 2.1 mm i.d., 1.7 µm) from Waters. Flow rate was 0.7 mL min−1

and column temperature was kept at 55°C. Peak identification and integration was performed

with software Empower 2 (Waters, Eschborn, Germany) using an Amino Acid Standard H

(Pierce, USA). Tryptophan was not quantified due to its susceptibility to acid hydrolysis.

Plasma glucose and triglycerides were determined by the enzymatic colorimetric GPO-PAP

method with commercial kits (Greiner, Bahlingen, Germany). Plasma protein was quantified

according to Bradford with the Roti-Quant kit and a BSA standard dilution (CARL ROTH,

Karlsruhe, Germany) series. All colorimetric assays were measured with a microplate reader

Tecan® Infinite 200 (Crailsheim, Germany) and calculated from a standard dilution series.


Fish performance was determined using the following formulae:

Specific growth rate (SGR, % day-1) = (ln final body weight – ln initial bw) × 100 / days fed




All diets were assigned by a completely randomized design. Data are presented as mean ±

standard deviation (SD) of n samples. Biological and analytical data were checked for normal

distribution using the Kolmogoroff Smirnov Test and eventually subjected to transformation.

Data were subjected to one-way analysis of variance (ANOVA) using SPSS 17.0 for

Windows (SPSS Inc., Chicago, US). When differences among groups were identified,

multiple comparisons among means were made using Tukey’s HSD test. Statistical


of comparisons.

56

Results

Growth performance, feed efficiencies and body composition

Final growth, SGR, feed intake and survival of fish were not compromised when RPC

replaced up to 100 % of dietary fish meal in the control diet (Tab. 4.4). The FCR and PER

were significantly lower in fish fed on diet R66 (1.22±0.03 and 1.70±0.04) compared to the

control group (1.09±0.02 and 1.90±0.04) but similar to fish fed diet R100 (1.18±0.06 and

1.80±0.08).

Table 4.4: Growth response, feed efficiencies and survival of rainbow trout fed experimental diets

R0 R66 R100

Initial weight 37.6 ± 1.9 37.4 ± 1.6 38.5 ± 1.0

Final weight 102.8 ± 11.5 94.0 ± 2.4 97.0 ± 6.6

SGR 1.19 ± 0.08 1.10 ± 0.06 1.10 ± 0.08

Feed intake (DM) 78.5 ± 10.98 75.8 ± 3.39 74.7 ± 5.12

FCR 1.09a ± 0.02 1.22b ± 0.03 1.18ab ± 0.06

PER 1.90a ± 0.04 1.70b ± 0.04 1.80ab ± 0.08

Survival (%) 100 ± 0.0 97.2 ± 4.8 91.7 ± 8.3

Means in the same row with different superscript letters are significantly different (Tukey's test, P<0.05)

Dietary treatment did not influence whole body moisture or protein content. Although not

significantly different from the control group, whole body fat content revealed a slight

increase with increasing dietary RPC inclusion. Percentages of body ash were significantly

lower in trout feeding diets R66 and R100 compared to those fed the control diet (Tab. 4.5).

Table 4.5: Proximate whole body composition (% of original substance) of rainbow trout fed experimental diets

R0 R66 R100

Moisture 68.8 ± 1.1 69.5 ± 0.4 67.9 ± 0.2

Crude protein 15.7 ± 0.4 16.0 ± 0.2 16.1 ± 0.1

Crude fat 13.2 ± 1.3 13.6 ± 0.4 14.9 ± 0.5

Ash 2.0a ± 0.1 1.7b ± 0.2 1.4b ± 0.2

Initial body composition: moisture 72.3%, crude protein 16.1%, crude fat 9.5%, ash 2.6%


57

Blood features

Neither blood haematocrit values nor haemoglobin concentrations displayed significant

differences between the treatment groups. Additionally, investigated blood serum values were

not significantly different between treatment groups (Tab. 4.6).

Table 4.6: Blood parameters of trout fed experimental diets

R0 R66 R100

Haemoglobin (mM L-1) 3.78 ± 0.49 3.22 ± 0.68 3.72 ± 0.62

Haematocrit (Proportion of 1) 0.38 ± 0.06 0.34 ± 0.04 0.36 ± 0.04

Triglycerides (mM L-1) 7.32 ± 2.70 5.74 ± 1.68 5.44 ± 2.13

Glucose (mM L-1) 10.98 ± 3.82 9.82 ± 2.55 9.82 ± 2.94

Protein (g L-1) 33.6 ± 1.9 34.1 ± 1.4 33.7 ± 1.7

Histology features

No changes in histological morphology were observed between dietary treatments for any of

the intestinal regions examined and no pathological reactions were identified.

Discussion

In several fish feeding trials rapeseed products have been found to be viable alternatives to

fish meal (Davies et al. 1990; Burel et al. 2000, 2001; Kissel et al. 2000; Thiessen et al. 2003;

Shafaeipour et al. 2008). However, when replacing relatively high percentages of dietary fish

meal, it was found, that antinutritional factors present in rapeseed products such as

glucosinolates, phytic acid or polysaccharides can negatively affect feed taste, feed intake and

nutrient absorption and consequently deplete fish growth (Francis et al. 2001). Despite a

favourable amino acid profile, the nutritional quality of rapeseed products tested so far was

below that of fish meal. The rapeseed protein concentrate tested in the present study was

produced to contain a protein level comparable to fish meal. Processing techniques applied

(Slawski et al. in press) led to a RPC with relatively low levels of glucosinolates (1.32 µmol

g-1), phytic acid (1.77 g 100g-1), polysaccharides and other antinutritional factors. In

comparison, rapeseed meals tested by Burel et al. (2000a,b,c; 2001) in fish meal replacement

studies with rainbow were either pressure cooked or directly oil extracted. These meals

contained 26 µmol g-1 or 40 µmol g-1 glucosinolates and 4.43 or 4.15 g 100g-1 phytic acid,

58

respectively, and led to reduced growth performance when replacing 33 % of dietary fish

meal. With our rapeseed protein product, it was possible to replace 100 % of fish meal from a

rainbow trout diet without negative effects on fish growth or feed efficiencies.

It has been reported (Burel et al. 2000; Slawski et al. in press) that fish meal replacement with

rapeseed products reduces feed intake even at low dietary concentrations, most probably due

to the bitter taste of glucosinolates. In the present study, feed intake did not vary significantly

between treatment groups thereby indicating the elimination of bitter flavour as well as a

suitable diet taste at high dietary RPC inclusion. Blood meal and particularly crustacean meal

incorporated into the diets potentially contributed to maintaining appetence and diet taste as

well as nutritional quality as reported in studies on alternative plant protein sources fed to

carnivorous fish (Espe et al. 2006, 2007; Torstensen et al. 2008).

While diet taste and feed intake were not compromised by RPC inclusion, slightly negative

tendencies in fish growth and feed efficiencies in treatment groups R66 and R100 still

indicate some unfavourable effects, though not significant here. Although nutrient

digestibility was not determined in the current study, it has been reported that even products at

low replacement, high protein content and subsequent low levels of antinutrients have adverse

effects on nutrient digestibility. Mwachireya et al. (1999) found that fibre levels - either alone

or together with phytates - negatively affects digestibility of canola protein products in

rainbow trout. The authors reported that among different canola products tested only a protein

isolate (90.8 % CP) met nutrient digestibility coefficients corresponding to fish meal.

Therefore, it appears possible, that remaining levels of antinutritional factors present in the

RPC might have a slightly negative impact on nutrient digestibility, thereby depressing feed

efficiencies in treatment groups R66 and R100 compared to the control group.

Besides reduced diet digestibility due to antinutritional factors, experimental result suggested,

that dietary phosphorus content and phosphorus availability were reduced in treatment groups

R66 and R100. This is indicated by decreasing body ash levels in respective groups as it is

known that body ash level can be reduced when fish are fed a diet deficient in available

phosphorus (Skonberg et al. 1997; Shao et al. 2008). As shown in Tab. 4.3, dietary

phosphorus levels decreased from 1.27 to 0.82 % with increasing dietary level of RPC.

Although these values are above established requirement levels for rainbow trout (Ogino and

Takeda 1978) it seems possible that better phosphorus availability in diet R0 positively

affected feed efficiencies. Antinutritional factors such as phytic acid, fibre and other complex

carbohydrates present in RPC might have furthermore reduced phosphorus availability in fish

59

(Francis et al. 2001). The assumption, that phosphorus availability was lower in diets R66 and

R100 is supported by slight differences in fish whole-body lipid content (Tab. 4.5). Although

not significantly different between treatment groups a tendency for increased body lipid

content with higher dietary RPC content can be detected. Increased whole-body lipid content

with high dietary levels of vegetable proteins has been reported in several fish species

(Adelizi et al. 1998; Kaushik et al. 2004). It is believed that this is caused by the accumulation

of fatty acids due to impaired β-oxidation (Takeuchi and Nakazoe 1981) or oxidative

phosphorylation due to phosphorus deficiency, thereby inhibiting the TCA cycle and leading

to an accumulation of acetyl-CoA and an increased fatty acid synthesis (Skonberg et al.

1997). For prospective feeding trials dietary phosphorus supplementation might support

phosphorus availability as shown by Lee et al. (2010).

With regard to the plasma parameters, no significant differences between the control group

and fish fed diets containing RPC were found. Consistent blood haemoglobin, haematocrit

and serum values therefore indicate an equal nutritional status among feeding groups

(Congleton and Wagner 2006).

In this study, no histopathological abnormalities compared to the control were observed. The

use of RPC did not cause any detectable histological alterations as found in Atlantic salmon

when receiving diets that contained certain amounts of solvent extracted soybean meals,

which have been identified to potentially induce enteritis (van den Ingh et al. 1991;

Baeverfjord and Krogdahl 1996).

In conclusion, the rapeseed protein concentrate processed and tested here showed great

potential as an alternative to fish meal in rainbow trout diets. With fish meal prices on the rise

it remains a matter of time until high quality rapeseed protein products can be produced in

large quantities at a competitive price and become a common fish feed ingredient.

References

Adelizi, P.D., Rosati, R.R., Warner, K., WuÆ� &vR� � �V çF—F–

W �2 � B� � � �6ö× WF—F—fR� � &– �6R � æB� &V6ö � �ÖR � �6öÖÖöâ f—

� �6‚ fVVB –æw&VF– �VçBàÐÕ&VfW&Væ6W0Ô FVÆ—

� �¦’Â äBâÂ� � � � �&÷6 F’Â "å"âÂ v &æ � �W"Â ²âÂ wR �Â ’ �åbâÂ ×VVæ6‚Â� Bå"â

�Â v†—FR � � � �Â Òå"âÂ '&÷vâÂ ä"âÂ� � ““ � �‚â Wf ÇV� F– � �öâ öb f—

6‚ÖÖV� Â� g&VR� F– �WG2 f÷"

60

arge quantities at a competitive price and become a common fish feed ingredient.

References

Adelizi, P.D., Rosati, R.R., Warner, K., Wu, Y.V., Muench, T.R., White, M.R., Brown,

P.B., 1998. Evaluation of fish-meal f �ree diets for Æ &vR� � �V çF—F–

� � � � � �W2 B 6ö× WF—F—fR� � &– � � � � � �6R æB &V6öÖR 6öÖÖ �öâ f—

6‚� fV �VB –æw&VF–VçBàÐÕ&VfW&Væ6W0Ô� FVÆ—

¦’Â� � �äBâÂ &÷6� F’Â� �"å"âÂ v� &æ � �W"Â ²âÂ wR �Â ’åbâ �Â ×VVæ6 �‚Â Bå"â

�Â v†— � � � � � � � � �FRÂ Òå"âÂ '&÷vâÂ ä"âÂ ““‚â Wf ÇV F– � �öâ öb f—

� � �6‚ÖÖV Â g&VR F– �WG2 f÷"

61


RefepC., Boujard, T., Escaffre, A.M., Kaushik, S.J., Boeuf, G., Mol, K., Van Der

Geyten, S., Kühn, E.R., 2000b. Dietary low glucosinolate rapeseed meal affects thyroid

status and nutrient utilization in rainbow trout (Oncorhynchus mykiss). British Journal of


� � � �RÂ Òå"âÂ '&÷vâÂ ä"âÂ� � “ � �“‚â Wf Ç �V F– � �öâ öb f—6‚ÖÖV� Â� g&VR� F–

�WG2 f÷"

62


References







� � �6‚ÖÖV Â g&VR F– �WG2 f÷"

63






erences







� � �6‚ÖÖV Â g&VR F– �WG2 f÷"

64






� FVÆ—



� � �6‚ÖÖV Â g&VR F– �WG2 f÷"

65










Burel, C., Boujard, T., Kaushik, S.J., Boeuf, G., Van Der Geyten, S., Mol, K.A., Kühn, E.R.,

Quinsac, A., Krouti, M., Ribaillier, D., 2000c. Potential of plant-protein sources as fish



Burel, C., Boujard, T., Kaushik, S.J., Boeuf, G., Mol, K.A., Van Der Geyten, S., Darras,

V.M., Kühn, E.R., Pradet-Balade, B., Quérat, B., Quinsac, A., Krouti, M., Ribaillier, D.,






Congleton, J.L., Wagner, T., 2006. Blood-chemistry indicators of nutritional status in juvenile

salmonids. Journal of Fish Biology 69, 473-490.








168–178.

66


processing on the antinutrient content of rapeseed. J. Sci. Food. Agr. 37, 735-741.







38, 551-579.










41.





Naczk, M., Shahidi, F., 1990. Carbohydrates of canola and rapeseed. In: Shahidi, F. (ed.)

Canola, Rapeseed: Production, Chemistry, Nutrition & Processing Technology, Van

Nostrand Reinhold, New York.


Press, Washington, DC.

Ogino, C., Takeda, H., 1978. Requirements of rainbow trout for dietary calcium and

phosphorus. B. Jpn. Soc. Sci. Fish. 44, 1015-1018.

Rodehutscord, M., Becker, A., Pack, M., Pfeffer, E., 1997. Response of rainbow trout

(Oncorhynchus mykiss) to supplements of individual essential amino acids in a

67

semipurified diet, including an estimate of the maintenance requirement for essential

amino acids. J. Nutr. 126, 1166–1175.






Sheehan, D.C., Hrapchak, B.B., 1980. Theory and Practice of Histotechnology, 2nd edn.

Mosby Publishing Co., St Louis, USA.



Slawski, H., Adem, H., Tressel, R.-P., Wysujack, K., Koops, U., Schulz, C., 201x.

Replacement of fish meal with rapeseed protein concentrate in diets fed to common carp

(Cyprinus carpio L.). Israeli Journal of Aquaculture (in press).

Slawski, H., Adem, H., Tressel, R.-P., Wysujack, K., Koops, U., Wuertz, S., Schulz, C., 201x.

Replacement of fish meal with rapeseed protein concentrate in diets fed to wels catfish

(Silurus glanis L.). Aquaculture Nutrition (in press).

Stickney, R.R., Hardy, R.W., Koch, K., Harrold, R., Seawright, D., Massee, K.C., 1996. The

effects of substituting selected oilseed protein concentrates for fish meal in rainbow

trout Oncorhynchus mykiss diets. Journal of the World Aquaculture Society 27, 57–63.


composition in carp. B. Jpn. Soc. Sci. Fish. 47, 347–352.











200.

68


glucosinolates content of various rapeseed–mustard meals. Indian Journal of Animal


Van Den Ingh, T.S.G.A.M., Krogdahl, Å., Hendriks, H.G.C.J.M., Koninkx, J.G.J.F., 1991.

Effects of soybean-containing diets on the proximal and distal intestine in Atlantic

salmon: a morphological study. Aquaculture 94, 297–305.

69

Chapter 5: Replacement of fish meal with albumin and globulin rapeseed

protein fractions in diets fed to rainbow trout (Oncorhynchus mykiss W.)

H. Slawski1, 2, H. Adem3, R.-P. Tressel3, K. Wysujack4,

F. Nagel1 and C. Schulz1, 2



D-24098 Kiel




…submitted to!

70

Abstract

The potential of two rapeseed protein concentrates partitioned in albumin and globulin

fractions as fish meal alternative was evaluated. In a digestibility experiment with juvenile

rainbow trout apparent digestibility coefficients were determined by indirect marker method.

ADCs of protein from fish meal (89.2±1.1 %) and globulin concentrate (88.8±0.6 %) were

significantly higher than from albumin concentrate (77.7±1.4 %). ADCs of dietary dry matter

were similar between the control diet (62.5±4.7 %) and the globulin concentrate diet

(62.3±0.5 %), but significantly lower in the albumin concentrate diet (56.2±1.5 %). In a

consecutive growth trial, ten rainbow trout (initial average weight 31.5±0.5 g) were stocked

into each of 21 experimental tanks of a freshwater flow-through system. Fish were organized

in triplicate groups and received experimental diets with 0, 50, 75, or 100 % of fish meal

replaced with albumin or globulin concentrate on the basis of digestible protein. At the end of

a 70 day feeding period it was found that only in treatment group A50, fish growth

performance and feed intake were not negatively affected by dietary treatment. However, feed

efficienies were not significantly different compared to the control group at 100 % or 75 %

fish meal replacement level with albumin or globulin concentrate, respectively. Significant

lower fish survival rates were observed when fish received diets A75, A100, G50, G75, or

G100 compared to the control diet or diet A50. For the whole body composition, the crude

protein content was significantly lower in fish fed diet G75 or G100 compared to the control

diet, while fish fed on diet A50, A75, or A100 were lower in body fat content than fish fed on

the control diet. It is concluded, that the used albumin concentrate can effectively replace 50

% of dietary fish meal in rainbow trout diets. The used globulin concentrate negatively

influenced diet palatability, thereby reducing diet intake and subsequently fish growth.

71

Introduction

Fish meal, the most important source of animal protein for fish diets, is a limited resource.

Due to increasing prices for fish meal together with environmental concerns the aquaculture

sector is forced to find alternative protein sources to be included in fish feeds. Wide

availability, relatively high protein contents and a desirable amino acid profile have caused an

interest in rapeseed products as ingredients for fish feed production. But, the nutritional

quality of rapeseed products largely depends on their levels of antinutritional factors,

particularly glucosinolates, phytic acid, phenolic constituents and indigestible carbohydrates,

as it was found in feeding trials with several fish species (Webster et al. 1997; Burel et al.

2000a,b,c, 2001; Thiessen et al. 2003, 2004; Glencross et al. 2004; Shafaeipour et al. 2008). It

was generally observed, that the nutritional quality of simple rapeseed products was below

that of fish meal, mainly due to antinutritional factors present in rapeseed products (Mawson

et al. 1995; Francis et al. 2001). By several processing techniques the level of antinutrients in

rapeseed products can be decreased and their value for fish nutrition improved (Fenwick et al.

1986; Anderson-Haferman et al. 1993; Tripathi et al. 2000). Protein extraction from meals by

ethanol-treatment will increase protein level and effectively remove glucosinolates, phenolic

compounds, soluble sugars and some oligosaccharides (Naczk and Shahidi 1990; Chabanon et

al. 2007). However, particularly ethanol-treatments are costly and time consuming (Slawski et

al. in press a,b).

In the present study, two rapeseed protein products either consisting mainly of albumin or

globulin fractions were derived from rapeseed after a hexane treatment and ultrafiltration. The

aim of the study was to clarify, if fractionized protein concentrates produced with less effort

than high quality protein concentrates or isolates are valuable fish meal alternatives in the

nutrition of rainbow trout. For this, nutrients digestibility was determined by indirect marker

method and compared to fish meal and the concentrates were evaluated as fish meal

replacement in rainbow trout diets.


Digestibility trial

Three diets were produced for the digestibility trial. A quantity of 10g kg-1 of titanium oxide

was added to a control diet mixture as inert marker for determination of apparent digestibility

coefficients. For the determination of apparent digestibility coefficients of rapeseed albumin

and rapeseed globulin concentrate diets were formulated that consisted of 700 g kg-1 of the

72

control diet and of 300 g kg-1 rapeseed albumin or globulin concentrate, respectively (on as is

basis). The rapeseed albumin and globuline concentrates were produced by the PPM,

Magdeburg, Germany. For the production of the concentrates a batch of rapeseed (variety

Lorenz, Norddeutsche Pflanzenzucht, Hohenlieth, Germany) was conditioned in a vacuum

dryer for 15 minutes at 70-80 °C to inactivate the enzyme myrosinase. Then rapeseed was

cold pressed. To remove residual oil from the oilcake (12.9 % crude lipid, 31.3 % crude

protein) it was crushed into 1-5 mm particle size followed by a hexane treatment. The

treatment lasted for two hours and the incubation temperature was 60 °C. Hexane was

removed from rapeseed meal extract by ventilation until the material contained not more than

1.5 % hexane. Then rapeseed meal extract was further crushed to a particle size of 0.2 to 0.1

mm. In the following, protein was gained through liquid water extraction (rapeseed meal

extract 1:10 water). For this, the suspension was heated to 40-45 °C followed by one hour of

constant agitation. Afterwards the suspension was decanted. Following decantation the

solvent was collected and residue material was secondly extracted (residue 1:10 water, 5 %

NaCl) at 40-45 °C and one hour contact time under constant agitation. Following extraction

the suspension was decanted. Then solvents of extraction 1 and 2 were separately treated in

order to receive mainly rapeseed globulin or albumin molecules after dia- and ultrafiltration

(membrane size: 10 kDa). During filtration conductivity was checked. Protein washing ended,

when conductivity was 5-6 mS cm-1. The gained materials were spray dried at 70-80 °C,

which led to a globulin or albumin concentrate with a crude protein content of 56.3 or 70.1 %,

respectively (Table 5.1). Diet mixes were manufactured to give pellets 4 mm in diameter (L

14-175, AMANDUS KAHL, Reinbek, Germany). Diet formulations, nutritional compositions

and amino acid profiles are presented in Table 5.2.

73

Table 5.1: Nutrient composition (% dry matter) and amino acid profiles (g 100g-1 CP) of fish

meal, albumin concentrate and globulin concentrate

Fish meal Albumin Globulin

Dry matter (%) 91.6 94.6 94.8

Crude protein 69.0 70.1 56.3

Crude fat 7.0 0.38 0.37

Ash 20.7 20.8 8.6

NfEa 3.3 8.7 29.5

Gross energyb (MJ kg-1) 19.9 18.4 18.8

Amino acids

Arginine 5.84 6.50 6.24

Histidine 2.00 3.28 2.62

Isoleucine 3.62 4.00 3.87

Leucine 6.45 7.55 7.04

Lysine 6.55 6.95 5.07

Methionine

(+ Cysteine)

2.37

3.17

1.86

5.22

2.02

4.18

Phenylalanine 3.52 3.89 4.16

Threonine 3.90 4.24 4.20

Valine 4.45 5.31 5.00

Phytic acid (g 100g-1) 2.04 1.53

Glucobrassicanapin 0.29

Glucobrassicin

Gluconapin 0.86

Gluconapoleiferin

Progoitrin 1.16

4-Hydroxyglucobrassicin

∑ Glucosinolates (µmol g-1) <0.10 2.31 aNitrogen free extract = 100 – (%crude protein + %crude fat + %ash). bCalculated by: crude protein = 23.9 MJ kg-1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1.

74

Table 5.2: Formulation and nutrient composition (g kg-1) and amino acid profiles (g kg crude protein-1) of

experimental diets used in the digestibility trial

Control Albumin Globulin

Herring meal 685.6 480 480

Albumin 300

Globulin 300

Maizestarch 144.9 101 101

Fishoil 99.5 70 70

Vit/MinMixa 20.0 14 14

Wheat Gluten 40.0 28 28

Titaniumdioxide 10.0 7 7

Nutrient composition

Moisture (% wet weight) 66.1 69.1 84.8

Crude protein 525.1 559.1 542.5

Crude fat 162.7 135.2 127.8

Ash 139.7 185.8 129.8

NfEb 172.5 119.9 199.9

Gross energyc (MJ kg-1) 22.1 20.8 21.6

Amino acids

Arginine 57.2 59.8 58.2

Histidine 20.7 25.3 22.6

Isoleucine 37.2 37.9 37.0

Leucine 66.0 68.5 66.4

Lysine 62.0 64.4 58.0

Methionine

(+ Cysteine)

24.4

32.7

21.3

39.0

22.4

35.2

Phenylalanine 36.4 36.6 37.4

Threonine 37.4 39.0 37.8

Valine 45.3 47.6 46.2 aAA-Mix 507101, Vitfoss, Gråsten, Denmark. bNitrogen free extract = 100 – (%crude protein + %crude fat + %ash).

cCalculated by: crude protein = 23.9 MJ kg-1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1.

75

Digestibility trial

The digestibility trial was conducted in the experimental facilities of the Gesellschaft für

Marine Aquakultur (GMA) in Büsum, Germany. Fifty rainbow trout (average weight

41.9±3.3 g), obtained from Fischzucht Reese (Sarlhusen, Germany) were stocked into each of

9 cylindrical tanks (350 L) of a freshwater recirculation system. Photoperiod was artificially

controlled (6.00 a.m. to 6.00 p.m.). Tanks were provided with filtered water at 9 L min-1

(temperature: 15.5±0.7 °C; O2: 9.2±0.5 mg L-1; pH: 7.2±0.5; NH4+: <1.0 mg L-1; NO2

-: <0.5

mg L-1). Fish were fed at 2 % of their body weight per day in three portions to assure effective

nutrient utilization. Beside an overflow, tanks had a funnel shaped bottom were faeces

accumulated. The funnel was connected to a pipe outlet. Faeces were obtained by continuous

sieving of pipe outlet water. After a one week adaptive period to the experimental diets

readily excreted faeces were collected for 7 days. Faecal samples were freeze dried before

analysis. Following dietary and faeces nutrient and marker analysis apparent digestibility

coefficients (ADCs) of dry matter and protein of the reference diet were calculated according

to Maynard & Loosly (1969):

ADC of dry matter of diet (%) = 100 × [1 – (dietary TiO2/faecal TiO2)]

ADC of protein of diet (%) = 100 × [1 – (dietary TiO2/ faecal TiO2) × (faecal protein

concentration / dietary protein concentration)]

ADCs of dry matter and protein in albumin or globulin concentrate were calculated as follows

(Sugiura et al. 1998):

ADC of dry matter of CPI (%) = (ADC of the test diet – 0.7 × ADC of the reference diet) / 0.3

ADC of protein of albumin or globulin concentrate (%) = [(protein concentration in test diet ×

protein ADC of the test diet) – (0.7 × protein concentration in reference diet × protein ADC of

the reference diet)] / (0.3 × protein concentration in ingredient)

Data are presented as means with standard deviation (Table 5.4). Means were compared with

Student’s t-test (P<0.05).

76

Growth trial

Based on the results from the digestibility trial five experimental diets were formulated in

which 0, 50, 75 or 100 % of digestible fish meal protein was replaced with digestible protein

from albumin or globulin concentrate (designated as Control, A50, A75, A100, G50, G75,

G100, respectively). Diets were supplemented with vitamins and minerals to meet the dietary

requirements of freshwater fish (NRC 1993). Since essential amino acid concentrations did

not differ considerably between experimental diets supplementation of synthetic amino acids

appeared unnecessary. The diet mixtures were manufactured to give pellets 4 mm in diameter

(L 14-175, AMANDUS KAHL, Reinbek, Germany). Diet formulations, proximate

compositions and amino acid profiles are presented in Table 5.3.

The growth trial was conducted in the experimental facilities of Johann Heinrich von Thünen-

Institut, Institute of Fisheries Ecology (Ahrensburg, Germany). Juvenile rainbow trout that

had been hatched in the institute were used. One week before the experiment started ten fish

were stocked in each of fifteen experimental tanks (40 L), being part of a flow-through

system. Photoperiod was in accordance to natural rhythmic from August to October at our

latitude (53° 41' 0" N). Tanks were provided with well freshwater at 1 L min-1 (temperature:

12.1±0.5 °C; O2: 7.8±0.2 mg L-1; pH: 7.2±0.5; NH4+: <0.1 mg L-1; NO2

-: <0.2 mg L-1). For a

one week adaptation period fish were fed the control diet in two daily meals until apparent

satiation. After the adaptation period, fish were fasted for two days and initial average weight

was determined (31.5 ± 0.5 g). Triplicate groups of fish were fed the experimental diets twice

daily (at 8.30 a.m. and 16.00 p.m.) to apparent satiation for 70 days. At the beginning and at

end of the experiment, samples of the initial and final fish population (21 x 2 fish) were taken

and stored at -23 ºC to determine initial and final body composition.


Diets and homogenised fish bodies were analysed in duplicate for proximate composition.

Dry matter was calculated from weight loss after drying in an oven at 105 °C until constant

weight. Fat content was determined after HCl hydrolysis (Soxtec HT6, Tecator, Höganäs,

Sweden) and total nitrogen content by the Kjeldahl technique (protein = N × 6.25; Kjeltec

Auto System, Tecator, Höganäs, Sweden). Ash content was calculated from weight loss after

incineration of samples in a muffle furnace for 2 hours at 550 °C. Raw material and dietary

amino acid concentrations were analysed as described by Tzovenis et al. (2009). Tryptophane

was not analyzed.

77

Table 5.3: Formulation, proximate nutrient composition (g kg-1) and amino acid composition

(g 100g-1 dietary protein) of experimental diets for rainbow trout

Control A50 A75 A100 G50 G75 G100

Herring meal 325 162.5 81.3 0 162.5 81.3 0

Albumin concentrate 0 183.6 275.4 367.2

Globulin concentrate 0 200 300.2 400.2

Blood meal 150 150 150 150 150 150 150

Soyprotein concentrate 135 135 135 135 135 135 135

Potato starch 232 199.9 183.3 167.8 183.5 158.5 134.8

Fish oil 118 129 135 140 129 135 140

Vit/MinMixa 20 20 20 20 20 20 20

Wheat gluten 20 20 20 20 20 20 20

Nutrient composition

Moisture (% wet weight) 71 69 68 71 75 70 79

Crude protein 489 526 545 551 507 518 536

Crude fat 159 165 162 167 157 156 154

Ash 80 84 82 69 65 54 47

NfEb 272 225 212 213 271 272 263

Gross energyc (MJ kg-1) 22.8 23.1 23.2 23.6 23.1 23.4 23.6

Phytic acid (g kg-1) 3.74 5.62 7.50 3.06 4.59 6.12

Glucobrassicanapin

Glucobrassicin

Gluconapin 0.17 0.21

Gluconapoleiferin

Progoitrin 0.23 0.30 0.37

4-Hydroxyglucobrassicin

∑ Glucosinolates (µmol g-1) 0.23 0.47 0.58

Amino acids

Arginine 5.59 5.74 5.73 5.87 5.59 5.70 5.68

Histidine 3.98 4.27 4.26 4.30 4.04 4.01 4.13

Isoleucine 2.88 2.95 2.97 3.03 2.89 2.95 2.89

Leucine 8.89 9.09 8.96 9.08 8.92 8.78 8.99

Lysine 7.00 7.12 7.03 6.98 6.54 6.30 6.21

Methionine

(+ Cysteine)

1.72

2.56

1.59

3.14

1.52

3.33

1.51

3.49

1.56

2.73

1.54

2.87

1.47

2.94

Phenylalanine 4.91 4.92 4.79 4.99 4.97 4.97 5.08

Threonine 3.50 3.69 3.64 3.77 3.54 3.56 3.57

Valine 6.16 6.28 6.15 6.22 6.18 6.16 6.21 aAA-Mix 507101, Vitfoss, Gråsten, Denmark bNitrogen free extract = 100 – (%crude protein + %crude fat + %ash) cCalculated by: crude protein = 23.9 MJ kg-1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1

78



Specific growth rate (SGR, % day-1) = (ln final body weight – ln initial bw) × 100 / days fed






subjected to transformation. Data were subjected to a one-way analysis of variance (ANOVA)

using SPSS 17.0 for Windows (SPSS Inc., Chicago, US). When differences among groups

were identified, multiple comparisons among means were made using Tukey’s HSD test.

Statistical significance was determined by setting the aggregate type I error at 5% (P<0.05)

for each set of comparisons.

Results

Digestibility coefficients

As shown in Table 5.4, the ADC of dry matter in the control diet (62.5±4.7 %) was not

significantly different towards the globulin concentrate diet (62.3±0.5 %), but significantly

higher compared to the albumin concentrate diet (56.2±1.5 %). The ADC of protein from fish

meal (89.2±1.1 %) – the predominant protein source in the control diet - was similar to the

ADC of protein in the globulin concentrate diet (88.8±0.6 %) but significantly lower in the

albumin concentrate diet (77.7±1.4 %) (P<0.05).

Table 5.4: Apparent digestibility coefficients

Fish meal diet Albumin

concentrate diet

Globulin

concentrate diet

ADC of dry matter 62.5a ± 4.7 56.2b ± 1.5 62.3a ± 0.5

ADC of test ingredient 41.6b ± 4.9 62.0a ± 1.6

Crude protein 89.2a ± 1.1 77.7b ± 1.4 88.8a ± 0.6

Values in the same row with different superscript letters are significantly different (P<0.05).

79


Fish growth response, feed intake and feed efficiencies were not negatively affected when 50

% of digestible fish meal protein was replaced with protein from albumin concentrate (Table

5.5). At higher fish meal replacement with albumin concentrate, final weight, specific growth

rate and feed intake were significantly lower compared to the control group and fish receiving

diet A50. Feed conversion ratio and protein efficiency ratio, however, were not significantly

different towards the control group at 100 % fish meal replacement level with albumin

concentrate. When fish meal was replaced with globulin concentrate, growth performance and

feed intake were significantly reduced at 50 % fish meal replacement level (Table 5.5). Feed

efficiencies were similar to the control group up to 75 % fish meal replacement with globulin

concentrate (Table 5.5). Significant lower fish survival rates were observed when fish

received diets A75, A100, G50, G75, or G100 compared to the control diet or diet A50 (Table

5.5). For the whole body composition, significant higher body moisture contents were

determined for fish fed diet A75 or A100 compared to the control diet. The whole body crude

protein content was significantly lower in fish fed diet G75 or G100 compared to the control

diet, while fish fed on diet A50, A75, or A100 were lower in whole body fat content than fish

fed on the control diet (Table 5.6). Body ash content was not significantly different among

dietary treatments.

Table 5.5: Growth performance, feed intake and feed efficiencies of rainbow trout fed

experimental diets

Control A50 A75 A100 G50 G75 G100

Initial weight 31.4±0.3 31.3±0.9 31.0±0.6 31.9±0.2 31.8±0.1 31.4±0.2 31.4±0.2

Final weight 91.3a±7.6 90.6a±2.8 79.7b±5.7 79.9b±4.6 77.7b±5.4 66.3c±3.9 60.8c±4.0

SGR 1.57a±0.12 1.57a±0.02 1.39b±0.08 1.35b±0.07 1.31b±0.11 1.09c±0.10 0.97c±0.11

Feed intake (DM) 67.6a±3.9 60.3ab±3.7 52.7bc±3.9 54.3bc±8.0 57.1bc±3.7 49.5c±6.9 44.0cd±2.1

FCR 1.14ab±0.08 1.02a±0.04 1.09a±0.04 1.04a±0.08 1.26b±0.07 1.41c±0.05 1.39c±0.02

PER 1.80ab±0.13 1.92a±0.07 1.72b±0.07 1.65b±0.11 1.61b±0.09 1.38c±0.04 1.44c±0.03

Survival (%) 96.6a±5.7 96.6a±5.7 76.7b±11.5 50.0b±26.4 46.6b±31.1 50.0b±36.0 46.6b±28.9

Values are given as mean ± standard deviation. Values in the same row with different superscript letters are

significantly different (P<0.05). SGR, Specific growth rate; FCR, Feed conversion ratio; PER, Protein efficiency ratio.

80

Table 5.6: Proximate whole body composition (g kg-1 wet weight) of rainbow trout fed

experimental diets

Control A50 A75 A100 G50 G75 G100

Moisture 696a ± 5 708ab ± 8 711b ± 6 717b ± 6 702a ± 1 707ab ± 9 708ab ± 7

Crude protein 164a ± 2 164a ± 4 165a ± 4 163ab ± 5 158ab ± 6 155b ± 4 155b ± 5

Crude fat 117a ± 7 96b ± 9 95b ± 8 93b ± 8 109ab ± 8 110ab ± 9 110ab ± 9

Ash 33 ± 3 32 ± 4 31 ± 2 29 ± 3 29 ± 3 27 ± 3 28 ± 2



Discussion

In the conducted feeding trial 50 % of dietary fish meal was successfully replaced with

albumin concentrate. Higher fish meal replacement levels with albumin concentrate or fish

meal replacement with globulin concentrate failed and resulted in reduced feed intake,

decreased growth performance and high mortalities.

The digestibilty of protein from albumin and globulin concentrate was determined and

compared to fish meal protein. While protein from globulin concentrate (88.8±0.6 %) was as

efficiently digested as fish meal protein (89.2±1.1 %), the digestibility of protein from

albumin concentrate was lower (77.7±1.4 %). The ADC of protein from albumin concentrate

is comparable to that of canola meal protein (74.0 %) determined in Atlantic salmon

(Anderson et al. 1992). Furthermore, Mwachireya et al. (1999) observed ADCs of protein

between 77.4 to 88.1 % from differently processed canola meals. Different ADCs for protein

can result from the raw material NfE content, which can negatively influence protein

digestibility in carnivorous fish (Storebakken et al. 1998; Burel et al. 2000a; Francis et al.

2001). Mwachireya et al. (1999) stated that fibre levels, either alone or together with phytate,

can have greatest adverse effects on the digestibility of canola protein products for rainbow

trout. The albumin concentrate (CP: 70.1 %) used in the present study contained 20.8 % crude

ash, 8.7 % NfE and 2.04 g 100g-1 phytic acid while the globulin concentrate (CP: 56.3 %)

contained 8.6 % crude ash, 29.5 % NfE and 1.53 g 100g-1 phytic acid. It appears possible

therefore, that the negative effect on protein digestion from the combination of NfE and

phytic acid in albumin concentrate was more severe than in globulin concentrate. Simple

81

rapeseed products have been widely investigated as feed ingredients in fish growth studies

(Webster et al. 1997; Burel et al. 2000a,b,c, 2001; Thiessen et al. 2003, 2004; Glencross et al.

2004; Shafaeipour et al. 2008). However, the usability of high quality rapeseed protein

products originating from rapeseed oilcakes with protein contents comparable to or above that

of fish meal was seldomly evaluated. Higgs et al. (1982) successfully replaced 25 % of dietary

protein from a fish meal control diet for juvenile Chinook salmon with rapeseed protein

concentrate (CP: 61.3 %) without reducing growth rate and protein utilization. In the present

study, feed intake and fish growth were similar between fish fed on the control diet and fish

receiving diet A50. In all other treatments, feed intake and growth performance were

significantly reduced compared to fish fed on the control diet (Table 5.6). Results indicate

therefore, that diet palatability was negatively influenced at high fish meal replacement with

albumin (diets A75 and A100) or globulin concentrate (diets G50, G75 and G100). Reduced

diet palatability and decreased feed intake when using rapeseed protein concentrate at high

inclusion levels in diets for common carp or wels catfish has also been reported by Slawski et

al. (in press, a,b). This was referred to glucosinolates present in rapeseed, which are known to

negatively influence diet taste (Burel et al. 2000a,b,c). In the present experiment, low diet

intake in respective treatment groups also led to aggressive fish behaviour, which resulted in

further reduction of feed intake, consequently low growth performance and high mortalities.

Although applied processing techniques led to levels of glucosinolates in albumin and

globulin concentrate which were until now assumed to be to low to have detrimental effects

on food intake in rainbow trout (Burel et al. 2000c), it appears that neither albumin nor

globulin concentrate can effectively replace fish meal. The poor fish development in treatment

groups with low feed intake is further indicated in different whole body composition. As

presented in Table 5.6, in respective treatment groups fish bodies were relatively low in

protein (G75 and G100) or body fat (A50, A75 and A100) compared to normal developed fish

from the control group, pointing out to nutritional imbalances (Jobling 1994).

In conclusion, albumin concentrate appeared to be more suitable as fish meal replacement in

rainbow trout diets. However, feed intake of diets containing globulin concentrate or high

levels of albumin concentrate was significantly reduced possibly by reduced diet palatability.

This resulted in aggressive fish behaviour and consequently poor growth and high mortalities.

82

References

Anderson, J.S., Lall, S.P., Anderson, D.M., Chandrasoma, J., 1992. Apparent and true

availability of amino acids from common feed ingredients for Atlantic salmon (Salmo

salar) reared in sea water. Aquaculture 108, 111-124.













83, 653–664.












Glencross B, Hawkins W, Curnow J (2004) Nutritional assessment of Australian canola




83





Jobling, M., 1994. Fish bioenergetics. Chapman and Hall, London, UK.

Maynard, L.A., Loosly, J.K., 1969. Animal Nutrition. 6th edn. McGraw-Hill, New York,

USA.



41.





Naczk, M., Shahidi, F., 1990. Carbohydrates of canola and rapeseed. In: Canola, Rapeseed:

Production, Chemistry, Nutrition & Processing Technology. (Shahidi, F. ed.), pp. 211-

220. Van Nostrand Reinhold, New York, NY.






Slawski, H., Adem, H., Tressel, R.-P., Wysujack, K., Koops, U., Schulz, C., 201xa.


(Cyprinus carpio L.). Israeli Journal of Aquaculture (in press).

Slawski, H., Adem, H., Tressel, R.-P., Wysujack, K., Koops, U., Wuertz, S., Schulz, C.,

201xb. Replacement of fish meal with rapeseed protein concentrate in diets fed to wels

catfish (Silurus glanis L.). Aquaculture Nutrition (in press).




84

Sugiura, S.H., Dong, F.M., Rathbone, C.K., Hardy, R.W., 1998. Apparent protein digestibility

and mineral availabilities in various feed ingredients for salmonid feeds. Aquaculture

159, 177–202.









211–216.







85

Chapter 6: Total fish meal replacement with canola protein isolate in diets

fed to rainbow trout (Oncorhynchus mykiss W.)

H. Slawski1, 2, F. Nagel1, K. Wysujack3, D. T. Balke4, P. Franz5 and C. Schulz1, 2



D-24098 Kiel



4BioExx Speciality Proteins Ltd., 219 (North) Dufferin Street, Toronto, Ontario, Canada

5HELM AG, Business Unit Animal Nutrition, Nordkanalstrasse 28, D-20097 Hamburg

…submitted!

86

Abstract

The potential of canola protein isolate as fish meal alternative in diets for rainbow trout was

evaluated. Apparent digestibility coefficients for protein from fish meal (89.2±1.1 %) and

canola protein isolate (84.6±1.8 %) were determined by indirect marker method in a

digestibility experiment with juvenile rainbow trout. ADC of dietary dry matter was slightly

lower for the control diet (62.5±4.7 %), but not significantly different to the test diet

(65.9±3.1 %). In a consecutive growth trial, twenty fish (initial average weight 31.5±0.5 g)

were stocked into each of fifteen experimental tanks of a freshwater flow-through system.

Fish were organized in triplicate groups and received experimental diets with 0, 25, 50, 75

and 100 % of fish meal replaced with canola protein isolate on the basis of digestible protein.

At the end of a 70 days feeding period it was found that growth performance (individual final

weight: 91.3±7.6 g to 107.9±5.8 g), feed intake (65.4±4.4 g to 69.8±2.5 g) and feed

efficiencies (FCR: 0.92±0.04 to 1.14±0.08) in treatment groups receiving diets devoid of fish

meal were not negatively affected compared to the control group. The tested canola protein

isolate was therefore identified to be a highly valuable fish meal alternative, not affecting diet

taste, feed intake and feed efficiencies.

87

Introduction

Worldwide, 61.6 Mt of rapeseed/canola (Brassica napus L., B. campestris L.) were farmed as

sources of vegetable oil (FAO 2010). Thus, following oil extraction, enormous amounts of

oilcake become available. Canola Meal is widely used in livestock feed systems and Canola

concentrates have been developed also for use in feed system. Recently Canola isolates have

been developed for the food industry with the first producer nearing commercial release.

Although the amino acid profile of canola is suitable for fish nutrition (Higgs et al. 1996), the

oilcake or processed products that were de-oiled with organic solvents retain a variety of

antinutritional factors namely glucosinolates, phytic acid, phenolic constituents and

indigestible carbohydrates (Mawson et al. 1995; Francis et al. 2001). These antinutritional

factors potentially limit the suitability of simple canola products as a protein source and fish

meal alternative in finfish diets at relatively high inclusion levels as shown in experiments

with Oncorhynchus mykiss (Burel et al. 2000a,c 2001; Thiessen et al. 2003, 2004; Shafaeipour

et al. 2008), Oreochromis mossambicus (Davies et al. 1990), Ictalurus punctatus (Webster et

al. 1997), Cyprinus carpio (Dabrowski and Kozlowska 1981), Pagrus auratus (Glencross et

al. 2004) and Psetta maxima (Burel et al. 2000a,b). While several processing techniques such

as dehulling of seeds, heat and water treatments, utilisation of organic solvents and

ultrafiltration will increase protein levels and reduce levels of antinutrients in canola products

(Fenwick et al. 1986; Anderson-Hafermann et al. 1993; Tripathi et al. 2000; Tyagi 2002;

Chabanon et al. 2007) the benefits for fish nutrition are variable.

In previous work, a rapeseed protein concentrate with a crude protein content of 710 g kg-1

and extremely low levels of glucosinolates was tested as fish meal replacement in diets for

Cyprinus carpio and Silurus glanis (Slawski et al. in press, a,b). Replacement of fish meal

with rapeseed protein concenrate at levels above 33 % in carp and 25 % in wels catfish with

rapeseed protein concentrate negatively affected diet taste and feed intake leading to reduced

feed efficiencies.

In the present study, a canola protein isolate with a crude protein content of 812 g kg-1 was

tested as fish meal replacement in the nutrition of rainbow trout. Consecutive to a digestibility

trial the isolate was evaluated as fish meal replacement on the basis of digestible protein in a

growth trial. By this, we aimed to demonstrate the limitless potential of canola protein isolate

as protein source in fish diets.

88


Digestibility trial

A quantity of 10 g kg-1 of titanium oxide was added to the control diet mixture as inert marker

for estimation of apparent digestibility coefficients. To estimate apparent digestibility

coefficients of canola protein isolate (CPI) a second diet was formulated. This consisted of

700 g kg-1 of the control diet and 300 g kg-1 of CPI. The isolate was produced by BioExx

Specialty Proteins Ltd. (Saskatoon, Canada) (Tab. 6.1) using a novel cold processing

sequence. It consisted of low temperature conditioning, cold oil pressing, low temperature

solvent extraction and desolventization followed by aqueous processing for isolation of the

soluble proteins. The resulting purified protein solution was spray dried to limit thermal

damage. Diet mixes were manufactured to give pellets 4 mm in diameter (L 14-175,

AMANDUS KAHL, Reinbek, Germany). Diet formulations, nutritional compositions and

amino acid profiles are given in Table 6.2.

Table 6.1: Nutrient composition (g kg-1 dry matter) and essential amino acid profiles (g kg-1 protein) of fish meal

and canola protein isolate

Fish meal Canola protein isolate Dry matter (g kg-1) 916 946 Crude protein 690 812 Crude fat 70 11 Crude ash 207 27

Phosphorus 24 6 NfEa 34 150 Gross energyb (MJ kg-1) 19.9 22.5 Essential amino acids Arginine 58.4 70.9 Histidine 20.0 27.6 Isoleucine 36.2 41.8 Leucine 64.5 75.4 Lysine 65.5 51.0 Methionine (+ Cysteine)

23.7 31.7

19.4 40.5

Phenylalanine 35.2 41.9 Threonine 39.0 41.8 Valine 44.5 51.7 Glucosinolates (µmol g-1) 0.13 Phytate (g kg-1) 6.90 aNitrogen free extract = 100 – (%crude protein + %crude fat + %ash + %fibre). bCalculated by: crude protein = 23.9 MJ kg-1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1

89

Table 6.2: Formulation, nutrient composition (g kg-1) and essential amino acid profiles (g kg-1 crude protein) of

experimental diets used in the digestibility trial

Ingredients Fish meal control

diet

Canola protein

isolate test diet

Herring meala 685.6 480

Canola protein isolate 300

Maizestarchb 144.9 101

Fish oila 99.5 70

Vit/MinMixc 20.0 14

Wheat glutend 40.0 28

Titaniumdioxidee 10.0 7


Dry matter (g kg-1) 934 933

Crude protein 525 594

Crude fat 163 114

Crude ash 140 128

NfEf 172 164

Gross energyg (MJ kg-1) 22.0 21.6


Arginine 57.2 62.3

Histidine 20.7 23.6


Leucine 66.0 68.9

Lysine 62.0 56.9

Methionine

(+ Cysteine)

24.4

32.7

22.0

35.4


Threonine 37.4 38.6

Valine 45.3 47.1 aVFC GmbH, Cuxhaven, Germany; bEuroduna-Technologies GmbH, Barmstedt, Germany; cAA-Mix 517158 & 508240,

Vitfoss, Gråsten, Denmark; dCargill Deutschland GmbH, Krefeld, Germany; enaturhaus, Neckarwestheim, Germany; fNitrogen free extract = 100 – (%crude protein + %crude fat + %ash).; gCalculated by: crude protein = 23.9 MJ kg-1; crude fat

= 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1. Tryptophane was not analyzed.

The digestibility trial was conducted in the experimental facilities of the Gesellschaft für

Marine Aquakultur (Büsum, Germany). Fifty rainbow trout (average weight 41.9±3.3 g),

obtained from Fischzucht Reese (Sarlhusen, Germany) were stocked into each of 6 cylindrical

90

tanks (350 L) of a freshwater recirculation system. Photoperiod was artificially controlled

(6.00 a.m. to 6.00 p.m.). Tanks were provided with filtered water at 9 L min-1 (15.5±0.7 °C;

O2: 9.2±0.5 mg L-1; pH: 7.2±0.5; NH4+: <1.0 mg L-1; NO2

-: <0.5 mg L-1). Fish were fed at 2 %

of their body weight per day in three portions to assure effective nutrient utilization. Beside an

overflow, tanks had a funnel shaped bottom where faeces accumulated. The funnel was

connected to a pipe outlet. Faeces were obtained by continuous sieving of pipe outlet water.

After a one week adaptive period to the experimental diets readily excreted faeces were

collected for 7 days. Faecal samples were freeze dried before analysis. Following dietary and

faeces nutrient and marker analysis apparent digestibility coefficients (ADCs) of dry matter

and protein of the reference diet were calculated according to Maynard and Loosly (1969):

ADC of dry matter of diet (%) = 100 × [1 – (dietary TiO2/faecal TiO2)]

ADC of protein of diet (%) = 100 × [1 – (dietary TiO2/ faecal TiO2) × (faecal protein

concentration / dietary protein concentration)]

ADCs of dry matter and protein in CPI were calculated as follows (Sugiura et al. 1998):

ADC of dry matter of CPI (%) = (ADC of the test diet – 0.7 × ADC of the reference diet) / 0.3

ADC of protein of CPI (%) = [(protein concentration in test diet × protein ADC of the test

diet) – (0.7 × protein concentration in reference diet × protein ADC of the reference diet)] /

(0.3 × protein concentration in ingredient)

Data are presented as means with standard deviation (Tab. 6.4). Means were compared with

Student’s t-test (P<0.05).

Growth trial

Based on the results from the digestibility trial five experimental diets were formulated in

which 0, 25, 50, 75 or 100 % of digestible fish meal protein was replaced with digestible

protein from canola protein isolate (designated as I0, I25, I50, I75, or I100, respectively).

Diets were supplemented with vitamins and minerals to meet the dietary requirements of

freshwater fish (NRC 1993). Since essential amino acid concentrations did not differ

considerably between experimental diets supplementation of synthetic amino acids appeared

unnecessary. The diet mixtures were manufactured to give pellets 4 mm in diameter (L 14-

175, AMANDUS KAHL, Reinbek, Germany). Diet formulations, proximate compositions

and amino acid profiles are presented in Table 6.3.

91

Table 6.3: Formulation, proximate composition (g kg-1) and essential amino acid profiles (g kg-1 crude protein)

of experimental diets

Ingredients Control I25 I50 I75 I100

Herring meala 325 243.8 162.5 81.3 0

Canola protein isolate 0 72.8 145.5 218.3 291.1

Blood mealb 150 150 150 150 150

Soyprotein concentratec 135 135 135 135 135

Potato starchb 232 235.5 239 242.4 245.9

Fish oila 118 123 128 133 138

Vit/MinMixd 20 20 20 20 20

Wheat glutene 20 20 20 20 20


Dry matter (g kg-1) 929 935 933 934 938

Crude protein 489 488 481 512 506

Crude fat 159 157 156 157 159

Crude ash 80 72 47 57 37

Phosphorus 10.3 8.0 5.0 6.0 4.0

NfEf 272 283 316 274 298

Glucosinolatesg

(µmol g-1)

0.01 0.02 0.03 0.04

Phytateg (g kg-1) 0.50 1.00 1.50 2.00

Gross energyh (MJ kg-1) 22.8 22.9 23.3 23.3 23.7


Arginine 55.9 58.0 57.5 60.4 60.4

Histidine 39.8 40.8 43.0 40.0 42.0

Isoleucine 28.8 29.8 27.6 30.8 30.4

Leucine 88.9 90.9 92.9 90.1 91.3

Lysine 70.0 68.4 65.1 65.0 62.1

Methionine

(+ Cysteine)

17.2

25.6

17.0

27.1

14.3

26.6

16.3

29.1

14.5

29.2

Phenylalanine 49.1 50.5 50.8 49.7 50.1

Threonine 35.0 36.0 35.2 37.1 36.3

Valine 61.6 62.9 63.1 60.8 62.5 aVFC GmbH, Cuxhaven, Germany; bEuroduna-Technologies GmbH, Barmstedt, Germany; cIMCOSOY 60 Piglet, IMCOPA,

Araucaria, Brasil.; dAA-Mix 517158 & 508240, Vitfoss, Gråsten, Denmark; eCargill Deutschland GmbH, Krefeld, Germany; fNitrogen free extract = 100 – (%crude protein + %crude fat + %ash).; gCalculated according to concentration in raw

material: 0.13 µmol g-1; hCalculated by: crude protein = 23.9 MJ kg-1; crude fat = 39.8 MJ kg-1; NfE, fibre: 17.6 MJ kg-1.

Tryptophane was not analyzed.

92

The growth trial was conducted in the experimental facilities of Johann Heinrich von Thünen-

Institut, Federal Research Institute of Rural Areas, Forestry and Fisheries; Institute of

Fisheries Ecology (Ahrensburg, Germany). Juvenile rainbow trout that had been hatched in

the institute were used. One week before the experiment started twenty fish were stocked in

each of fifteen experimental tanks (40 L), being part of a flow-through system. Photoperiod

was in accordance to natural rhythmic from August to October at our latitude (53° 41' 0" N).

Tanks were provided with well freshwater at 1 L min-1 (temperature: 12.1±0.5 °C; O2: 7.8±0.2

mg L-1; pH: 7.2±0.5; NH4+: <0.1 mg L-1; NO2

-: <0.2 mg L-1). For a one week adaptation

period fish were fed the control diet (Table 6.4) in two daily meals until apparent satiation.

After the adaptation period, fish were fasted for two days and initial average weight was

determined (31.5 ± 0.5 g). Triplicate groups of fish were fed the experimental diets twice

daily (at 8.30 a.m. and 16.00 p.m.) to apparent satiation for 70 days. At the beginning and at

end of the experiment, a sample of 15 x 2 fish of the initial fish population was taken and

stored at -23 ºC to determine initial and final body composition.


Diets and homogenised fish bodies from each tank were analysed in duplicate for proximate

composition. Dry matter was calculated from weight loss after drying in an oven at 105 °C

until constant weight. Fat content was determined after HCl hydrolysis (Soxtec HT6, Tecator,

Höganäs, Sweden) and total nitrogen content by the Kjeldahl technique (protein = N × 6.25;

Kjeltec Auto System, Tecator, Höganäs, Sweden). Ash content was calculated from weight

loss after incineration of samples in a muffle furnace for 2 hours at 550 °C. Raw material and

dietary amino acid concentrations were analysed as described by Tzovenis et al. (2009).



Specific growth rate (SGR, % day-1) = (ln final body weight – ln initial body weight) × 100 /

days fed




93



subjected to transformation. Data were subjected to linear regression analysis in order to

detect correlations between diet formulation and fish performance and/or fish body

composition. Data were also subjected to a one-way analysis of variance (ANOVA) using

SPSS 17.0 for Windows (SPSS Inc., Chicago, US). When differences among groups were

identified, multiple comparisons among means were made using Tukey’s HSD test. Statistical


of comparisons.

Results

Digestibility coefficients

As shown in Table 6.4, ADC of dry matter in the control diet (62.5±4.7 %) was slightly lower

than in the test diet (65.9±3.1 %). The ADC of protein from fish meal (89.2±1.1 %) – the

single protein source in the control diet - was significantly higher than from canola protein

isolate (84.6±1.8 %) (P<0.05). Accordingly, amino acid digestibility followed this trend.

Table 6.4: Apparent digestibility coefficients

Fish meal control diet Canola protein isolate test diet

ADC of dry matter 62.5 ± 4.7 65.9 ± 3.1

ADC of test ingredient 73.9 ± 10.3

Crude protein 89.2a ± 1.1 84.6b ± 1.8

Amino acids

Arginine 94.4 88.6

Histidine 91.8 89.9

Isoleucine 92.1a 81.8b

Leucine 93.2a 83.6b

Lysine 94.4a 83.0b

Methionine

(+ Cysteine)

92.3a

80.7a

79.0b

89.2b

Phenylalanine 89.2a 79.3b

Threonine 90.8a 80.4b

Valine 92.2a 83.1b

Values in the same row with different superscript letters are significantly different (P<0.05).

94


Fish growth response, feed intake and feed efficiencies were not negatively affected when 100

% of digestible fish meal protein was replaced with protein from CPI (Table 6.5). In contrast,

with a final weight of 107.9±5.8 g and a SGR of 1.80±0.10, fish receiving diet I75 grew

significantly better than fish fed the control diet (final weight: 91.3±7.6 g; SGR: 1.57±0.12).

In addition, diet I75 gave a better feed conversion ratio (0.92±0.04) and protein efficiency

ratio (2.13±0.09) than the control diet (FCR: 1.14±0.08; PER: 1.80±0.13). Regression

analysis revealed significant positive correlations between dietary level of CPI and FCR

(R2=0.53, P<0.05) as well as PER (R2=0.37, P<0.05) indicating a relation between dietary

level of CPI and feed efficiencies. Survival of fish was not negatively affected by any

treatment. No significant differences in whole body composition were detected between fish

fed control diet and fish receiving CPI diets (Table 6.6). Furthermore, no correlations between

dietary level of CPI/ash/phosphorus and fish body parameters were identified.

Table 6.5: Growth response, feed intake and feed efficiencies of rainbow trout fed experimental diets

Control I25 I50 I75 I100 *R2 P

Initial weight (g) 31.4 ± 0.3 31.6 ± 0.2 31.3 ± 0.4 31.7 ± 0.4 31.7 ± 0.1

Final weight (g) 91.3a ± 7.6 101.2ab ± 6.2 100.9ab ± 3.3 107.9b ± 5.8 99.5ab ± 3.5 0.27 ns

SGR (%) 1.57a ± 0.12 1.71ab ± 0.10 1.72ab ± 0.03 1.80b ± 0.10 1.68ab ± 0.04 0.27 ns

Feed intake (g DM) 67.6 ± 3.9 69.4 ± 5.7 68.4 ± 3.5 69.8 ± 2.5 65.4 ± 4.4 0.02 ns

FCR 1.14a ± 0.08 1.00ab ± 0.09 0.98ab ± 0.03 0.92b ± 0.04 0.97ab ± 0.02 0.53 <0.05

PER 1.80a ± 0.13 2.06ab ± 0.17 2.11ab ± 0.07 2.13b ± 0.09 2.05ab ± 0.04 0.37 <0.05

R2: parameter values are regressed to the dietary level of canola protein isolate.

Values are given as mean ± standard deviation. Values in the same row with different superscript

letters are significantly different (P<0.05).

SGR, Specific growth rate; FCR, Feed conversion ratio; PER, Protein efficiency ratio.

Table 6.6: Proximate whole body composition (g kg-1 wet weight) of rainbow trout fed experimental diets

Control I25 I50 I75 I100

Moisture 696 ± 5 700 ± 5 694 ± 11 704 ± 1 707 ± 6

Crude protein 164 ± 2 162 ± 9 162 ± 7 161 ± 7 156 ± 3

Crude fat 117 ± 7 117 ± 6 119 ± 8 109 ± 4 111 ± 5

Crude ash 33 ± 3 32 ± 2 30 ± 3 32 ± 2 28 ± 2


95

Discussion

Results show that canola protein isolate can be used as a valuable source of protein in diets for

rainbow trout. In the present feeding trial it was possible to replace all fish meal without

negative effects on feed intake, feed efficiencies and fish growth. In general, specific growth

rates ranged from 1.57 to 1.80 and feed conversion ratios varied between 0.92 and 1.14. This

is in line with results from other experiments investigating fish meal alternatives in rainbow

trout diets (Adelizi et al. 1998; Drew et al. 2007).

In several studies, the ADCs of canola protein products in fish diets have been determined. In

experiments with Atlantic salmon an ADC of protein from canola meal of 74.0 % has been

found (Anderson et al. 1992). However, the canola meal tested had a protein content of 390 g

kg-1 and therefore contained significant amounts of NfE. These are known to negatively

influence protein digestibility in carnivorous fish (Storebakken et al. 1998; Burel et al. 2000a;

Francis et al. 2001). Mwachireya et al. (1999) evaluated the protein digestibility of a canola

protein isolate in rainbow trout diets. The digestibility trial was conducted as described by

Hayen et al. (1993), using a settling column for the collection of fish faeces. For the canola

protein isolate tested Mwachireya et al. (1999) determined an ADC of protein of 97.6 %. This

value was regarded as one of the highest ever reported in fish nutrition studies. The authors

ascribed this high protein digestibility to the high level of protein (908 g kg-1) and low levels

of all antinutritional factors and indigestible carbohydrates present in the canola protein

isolate compared to canola concentrates. Mwachireya et al. (1999) concluded that fibre levels,

either alone or together with phytate, can have greatest adverse effects on the digestibility of

canola protein products for rainbow trout. The CPI used in our study contained 812 g kg-1 of

crude protein and 150 g kg-1 of NfE. It appears possible, that NfE negatively affected protein

digestibility of our CPI. Accordingly, the amino acid digestibility was also lower in CPI than

in the fish meal control diet. It has to be noted, however, that the ADC of protein (84.6±1.8

%) was in a range with other rapeseed/canola protein products tested in rainbow trout. In

example, ADCs of protein of 90.9±2.3 % from solvent extracted and 88.5±1.5 % of protein

from heat treated rapeseed meal were reported by Burel et al. (2000a) using an automatic

faecal collector as described by Choubert et al. (1982) being similar to the system applied in

the present study. Furthermore, Mwachireya et al. (1999) observed ADCs for protein between

77.4 to 88.1 % for differently processed canola meals.

The limitations of simple rapeseed products as feed ingredients in fish growth studies have

been widely investigated (Dabrowski and Kozlowska 1981; Davies et al. 1990; Webster et al.

96

1997; Burel et al. 2000a,b,c, 2001; Thiessen et al. 2003, 2004; Glencross et al. 2004;

Shafaeipour et al. 2008). However, lack of information exists about the benefits of high

quality products originating from rapeseed/canola oilcakes with protein contents comparable

to or above that of fish meal. Higgs et al. (1982) successfully replaced 25 % of dietary protein

from a fish meal control diet for juvenile Chinook salmon with rapeseed protein concentrate

(613 g CP kg-1) without reducing growth rate and food (protein) utilization. In that study,

however, higher fish meal replacement levels with rapeseed protein concentrate were not

evaluated. Rapeseed protein concentrate with a protein content of 710 g kg-1 was evaluated as

fish meal replacement in diets for wels catfish and common carp (Slawski et al. in press, a,b).

It was found, that diet taste and feed intake were negatively influenced by high dietary

inclusion of rapeseed protein concentrate probably due to glucosinolates present in rapeseed.

In addition, feed efficiencies and consequently fish growth decreased when wels catfish or

carp received diets with more than 25 % or 33 % of fish meal replaced with rapeseed protein

concentrate. This was referred to dietary levels of NfE and insufficient phosphorus

availability (Slawski et al. in press, a,b).

In the present study, feed intake was similar in all feeding groups. This indicates that the taste

of canola protein isolate was well accepted by rainbow trout. In addition, we found no

negative influence from dietary inclusion of CPI on fish growth performance and feed

efficiencies. Interestingly, fish receiving diet I75 grew significantly better than fish receiving

the control diet. A reason for this might be an increased dry matter digestibility with

increasing dietary incorporation of CPI for fish meal. As shown in Table 6.4, dietary

incorporation of CPI led to slightly higher dry matter digestibility (65.9±3.1 %) than in a fish

meal control diet (62.5±4.7). The slightly higher but not significantly different growth

performance and feed efficiencies of fish in treatment group I75 compared to other groups

could be attributed to varying dietary levels of phosphorus and NfE. Diet I75 (6.0 g P kg-1)

contained more phosphorus than diet I50 (5.0 g P kg-1) or I100 (4.0 g P kg-1). Dietary

phosphorus requirements ranging from 5.0 to 8.0 g kg-1 have been reported for rainbow trout

(Ogino and Takeda 1978). Antinutritional factors such as phytic acid, fibre and other complex

carbohydrates present in CPI may have contributed to reduced phosphorus availability in fish

(Francis et al. 2001). Severe phosphorus deficiencies from any dietary treatment, however,

appear unlikely. It is known that whole-body ash is reduced when carnivorous fish are fed a

diet deficient in available phosphorus (Skonberg et al. 1997; Shao et al. 2008) and that whole-

body lipid content can be increased due to high dietary levels of vegetable protein (Adelizi et

97

al. 1998; Kaushik et al. 2004). Latter is believed to be caused by the accumulation of fatty

acids due to impaired β-oxidation (Takeuchi and Nakazoe 1981) or oxidative phosphorylation

due to phosphorus deficiency, thereby inhibiting the TCA cycle and leading to an

accumulation of acetyl-CoA and an increased fatty acid synthesis (Skonberg et al. 1997).

In the present study, neither differences in whole body composition nor correlations between

dietary phosphorus content and ash levels in fish body indicating insufficient dietary

phosphorus supply were detected. It has been reported, however, that increased dietary

phosphorus levels can improve feed efficiencies and consequently growth in Atlantic salmon,

cod and sea bass (Vielma and Lall 1998; Roy and Lall 2003; Oliva-Teles and Pimentel-

Rodrigues 2004). Accordingly, a slightly higher phosphorus level in diet I75 compared to diet

I100 may have resulted in tendentially better feed efficiencies and fish growth.

Besides different dietary phosphorus supply, varying dietary levels of NfE might also have

contributed to slight differences in growth performances and feed efficiencies among

treatment groups. Since dietary NfE can potentially reduce nutrient and mineral digestibility

in fish (Storebakken et al. 1998; Burel et al. 2000a; Mwachireya et al. 1999; Francis et al.

2001) lower dietary levels of NfE in diet I75 (274 g kg-1) compared to diet I50 (316 g kg-1) or

diet I100 (298 g kg-1) might have resulted in slightly better growth performance and feed

efficiencies.

In conclusion, the canola protein isolate tested has shown great potential as fish meal

replacement in diets for rainbow trout. High dry matter and protein digestibility together with

unaffected palatability make canola protein isolate a promising candidate as protein source in

fish diets.

Acknowledgements

We gratefully acknowledge the financial assistance provided by HELM AG and BioExx for

this project.

References



Aquaculture Nutrition 4, 255-262.

98
















83, 653–664.








Choubert, G., De la Noue, J., Luquet, P., 1982. Digestibility in fish: improved device for the

automatic collection of faeces. Aquaculture 29, 185–189.


heated waters. I. Growth of fish and utilization of the diet. In: Tiews, K. (ed.)

Aquaculture in Heated Effluents and Recirculation Systems, pp. 263-274. Heenemann,

Hamburg, Germany.




99

Drew, M.D., Ogunkoya, A.E., Janz, D.M., Van Kessel, A.G., 2007. Dietary influence of

replacing fish meal and oil with canola protein concentrate and vegetable oils on growth

performance, fatty acid composition and organochlorine residues in rainbow trout









Glencross, B., Hawkins, W., Curnow, J., 2004. Nutritional assessment of Australian

canola meals. I. Evaluation of canola oil extraction method and meal processing

conditions on the digestible value of canola meals fed to the red seabream (Pagrus

auratus, Paulin). Aquaculture Research 35, 15-24.

Hajen, W.E., Higgs, D.A., Beames, R.M., Dosanjh, B.S., 1993. Digestibility of various

feedstuffs by post juvenile Chinook salmon (Oncorhynchus tshawytscha) in sea water.

2. Measurement of digestibility. Aquaculture 112, 333-348.





Higgs, D.A., Dosanjh, B.S., Beames, R.M., Prendergast, A.F., Mwachireya, S.A., Deacon, G.,

1996. Nutritive value of rapeseed/canola protein products for salmonids. In: CFIA Proc.,

May 15-17, 1996, Dartmouth/Halifax (Fischer, N. ed.), pp. 187-196. Canadian Feed

Industry Association (CFIA), Ottawa, Canada.






41.

100

Maynard, L.A., Loosly, J.K., 1969. Animal Nutrition. 6th edn. McGraw-Hill, New York,

USA.










bass (Dicentrarchus labrax L.) juveniles. Aquaculture Research 35, 636-642.










Slawski, H., Adem, H., Tressel, R.-P., Wysujack, K., Koops, U., Schulz, C., 201xa.


(Cyprinus carpio L.). Israeli Journal of Aquaculture (in press)

Slawski, H., Adem, H., Tressel, R.-P., Wysujack, K., Koops, U., Wuertz, S., Schulz, C.,

201xb. Replacement of fish meal with rapeseed protein concentrate in diets fed to wels

catfish (Silurus glanis L.). Aquaculture Nutrition (in press)




Sugiura, S.H., Dong, F.M., Rathbone, C.K., Hardy, R.W., 1998. Apparent protein digestibility

and mineral availabilities in various feed ingredients for salmonid feeds. Aquaculture

159, 177–202.

101


composition in carp. B. Jpn. Soc. Sci. Fish. 47, 347–352.









211–216.


Anim. Feed Sci. Tech. 99, 215–219.






128.




102

General Discussion

Simple rapeseed products have been widely tested as protein source in feeds for several fish

species (Dabrowski and Kozlowska 1981; Davies et al. 1990; Burel et al. 2000; Shafaeipour

et al. 2008). In general it was found that the nutritional quality of simple rapeseed products is

below that of fish meal. In particular, antinutritional factors (ANF) determine the quality of

rapeseed products for fish nutrition, among them glucosinolates, phytic acid, phenolic

constituents and indigestible carbohydrates (Francis et al. 2001). Different processing

techniques have been identified to reduce the level of antinutrients in rapeseed products

(Fenwick et al. 1986; Naczk and Shahidi 1990; Anderson-Haferman et al. 1993; Chabanon et

al. 2007) and potentially increase their value for fish nutrition. But, lack of information exists

about the benefits of high quality rapeseed products with protein contents comparable to or

above that of fish meal. However, results obtained by Higgs et al. (1982) indicate the

applicability of rapeseed protein concentrate (RPC) in fish nutrition. The authors successfully

replaced 25 % of dietary protein from a fish meal control diet for juvenile Oncorhynchus

tshawytscha with RPC (CP: 61 %) without reducing growth rate and food (protein) utilization.

In the present study, different protein products derived from rapeseed (including canola) were

tested as fish meal replacement in fish diets. A high quality rapeseed protein concentrate

(RPC) with a protein content of 71 % was evaluated as fish meal replacement in diets for

common carp (chapter 1), wels catfish (chapter 2), turbot (chapter 3) and rainbow trout

(chapter 4). Advanced processing techniques applied led to a RPC with relatively low levels

of glucosinolates (1.32 µmol g-1), phytic acid (1.77 g 100g-1), polysaccharides and other

antinutritional factors. In comparison, rapeseed meals tested by Burel et al. (2000a,b,c; 2001)

in fish meal replacement studies with rainbow trout were either pressure cooked or directly oil

extracted. These meals contained 26 or 40 µmol g-1 glucosinolates and 4.43 or 4.15 g 100g-1

phytic acid, respectively, and led to reduced growth performance when replacing 33 % of

dietary fish meal.

In common carp, the RPC successfully replaced 33 % of fish meal protein from a control diet

without retarding fish growth performance, feed intake or feed efficiencies. At 66 % and 100

% fish meal replacement with RPC, however, fish growth performance, feed intake and feed

efficiencies decreased compared to the control group. In wels catfish, 25 % of dietary fish

meal was successfully replaced with RPC without negative effects on feed efficiencies and

fish growth. When 50 % of dietary fish meal was replaced with RPC the feed intake as % of

103

fish body weight was not significantly different from the control group but feed efficiencies

and fish growth were reduced. In turbot, 33 % of dietary fish meal was replaced with RPC. At

66 % fish meal replacement level, feed intake and fish growth were significantly reduced. In

rainbow trout, feed intake and fish growth were not compromised when RPC replaced up to

100 % of dietary fish meal in the control diet.

Based on the results presented in chapter 4, in chapter 5 the potential of two rapeseed protein

concentrates partitioned in albumin and globulin fractions as fish meal alternatives was

evaluated in a digestibility study and a consecutive growth trial with rainbow trout. Compared

to the RPC, production of the fractionized protein concentrates was simplified. At 75 % fish

meal replacement with albumin concentrate, feed intake and fish growth were significantly

lower compared to the control group and fish receiving diet A50. Feed conversion ratio and

protein efficiency ratio, however, were not significantly different towards the control group at

100 % fish meal replacement level with albumin concentrate. When fish meal was replaced

with globulin concentrate, growth performance and feed intake were significantly reduced at

50 % fish meal replacement level. Feed efficiencies were similar to the control group up to 75

% fish meal replacement with globulin concentrate.

In chapter 6 a canola protein isolate (CPI) with a crude protein content of 81 % was evaluated

as fish meal alternative in diets for rainbow trout. The CPI was produced using a novel cold

processing sequence. It consisted of low temperature conditioning, cold oil pressing, low

temperature solvent extraction and desolventization followed by aqueous processing for

isolation of the soluble proteins. The resulting purified protein solution was spray dried to

limit thermal damage. Fish growth response, feed intake and feed efficiencies were not

negatively affected when 100 % of digestible fish meal protein was replaced with protein

from CPI.

In general, results obtained in the course of the present study demonstrate the high potential of

rapeseed protein products as fish meal alternative in fish nutrition. In the following, results of

the different trials are compared and possible limitations when using rapeseed protein

products are discussed.

104

Diet palatability

We observed a diminished diet acceptance at 66 % (and 100 %) fish meal replacement level

in carp and turbot and at 75 % in wels catfish, which is documented in lower feed intake in

respective fish groups. In trout, diet palatability appeared to be negatively affected at 75 %

(and 100 %) fish meal replacement level with albumin and at 50 % (as well as 75 and 100 %)

with globulin concentrate, also resulting in reduced feed intake. It is known that the bitter

taste exuded by glucosinolate metabolites, such as isothiocyanates and

vinyloxazolidinethiones, present in rapeseed meals can potentially retard diet acceptance by

fish. This was found in rainbow trout and turbot at dietary glucosinolate levels of 7.3 µmol g-

1 or 18.7 µmol g-1, respectively (Burel et al. 2000bc). Because the rapeseed protein products

used in our study contained 0.2-2.3 µmol glucosinolates g-1, dietary glucosinolate

concentrations were 0.07-0.6 µmol g-1. These values are far below the level when

glucosinolates were assumed to become detrimental on food intake of rainbow trout and

turbot (Burel et al. 2000bc). However, to our observation the typical mustard smell of

glucosinolates was still noticeable in respective experimental diets with high inclusion of

rapeseed protein products. When feeding rainbow trout with RPC and CPI however, feed

intake did not vary significantly between treatment groups thereby indicating the elimination

of bitter flavour as well as a suitable diet taste at high dietary RPC inclusion. It can only be

speculated, that carp, turbot and wels catfish are more sensitive towards a bitter diet taste than

rainbow trout. For prospective feeding trials with carp, turbot and wels catfish it appears

recommendable to use strong feed attractants, such as fish protein hydrolysate, squid

hydrolysate, stick water or krill meal to maintain high feed intake (Espe et al. 2006, 2007;

Torstensen et al. 2008; Kousoulaki et al. 2009).

Nutrient availability

Lower feed efficiencies observed at high RPC inclusion levels particularly in wels catfish,

turbot and trout might be a result of reduced dietary nutrient availability. Because of

insignificant phytic acid concentrations in respective diets, it is assumed that nutrient

availability was mainly reduced by fibre and other complex carbohydrates. It is known from

carnivorous fish that complex carbohydrates can greatly reduce mineral and nutrient

availability from aquafeeds, thereby reducing feed efficiencies as observed in Atlantic salmon

and turbot (Storebakken et al. 1998; Burel et al. 2000). The RPC used in our study contained

4.8 % fibre and calculated NfE was 12.8 %. In comparison, the NfE content of the used fish

105

meal was 3.4 %. High dietary inclusion of RPC therefore lead to relatively higher dietary fibre

+ NfE contents than in respective experimental control feeds. It can be assumed, that nutrient

availability from diets with high content of RPC was reduced.

Protein digestibility

Protein digestibility of rapeseed products has been determined in several studies. In Atlantic

salmon, the ADC of canola meal protein was 74.0 % (Anderson et al. 1992). ADCs of protein

of 90.9±2.3 % from solvent extracted and 88.5±1.5 % of protein from heat treated rapeseed

meal were reported by Burel et al. (2000a) in studies with rainbow trout. In addition,

Mwachireya et al. (1999) observed ADCs of protein between 77.4 to 97.6 % for differently

processed canola products. Digestibility experiments undertaken in the present study revealed

that protein from globulin concentrate (88.8±0.6 %) was as efficiently digested as fish meal

protein (89.2±1.1 %), while the ADCs of protein from albumin concentrate (77.7±1.4 %) and

canola protein isolate (84.6±1.8 %) were significantly lower. Different ADCs for protein can

result from the raw material NfE content, which can negatively influence protein digestibility

in carnivorous fish (Storebakken et al. 1998; Burel et al. 2000a; Francis et al. 2001). The

albumin concentrate contained less NfE (8.7 %) than the globulin concentrate (NfE: 29.5 %)

and the CPI (15.0 %). Accordingly, Mwachireya et al. (1999) stated, that high protein

digestibility demands lowest levels of all antinutritional factors and indigestible carbohydrates

present in canola products.

Body composition

Tendentially sinking body ash levels suggest reduced phosphorus availability from diets high

in RPC when fed to wels catfish, turbot and rainbow trout. It is known that body ash levels

can be reduced when fish are fed a diet deficient in available phosphorus and rich in vegetable

protein (Skonberg et al. 1997; Adelizi et al. 1998; Kaushik et al. 2004; Shao et al. 2008). In

example, the phosphorus levels of diets for rainbow trout decreased from 1.27 to 0.82 % with

increasing dietary level of RPC. Although these values are above established requirement

levels for rainbow trout and other fish species (Ogino and Takeda 1978; NRC 1993) it can not

be excluded that better phosphorus availability in diets devoid of RPC positively affected feed

efficiencies. It is known that excessive dietary phosphorus content can improve the feed

efficiency of diets for Atlantic salmon, cod and European sea bass (Vielma und Lall 1998;

Roy und Lall 2003; Oliva-Teles und Pimentel-Rodrigues 2004). Antinutritional factors such

106

as phytic acid, fibre and other complex carbohydrates present in RPC can potentially reduce

the phosphorus availability in fish (Mwachireya et al. 1999; Francis et al. 2001). In the

present studies, dietary phytic acid concentrations originating from RPC were up to 0.2 g kg-1

in diets for wels catfish and up to 5.2 g 100g-1 phytic acid in diets for turbot. In comparison,

Spinelli et al. (1983) observed decreased growth rates in rainbow trout fed a diet containing 5

g kg-1 synthetic phytic acid. Synthetic phytic acid at concentrations of 5 and 10 g kg-1 feed

resulted in lower growth performance in common carp (Hossain and Jauncey 1993). While

negative effects resulting from phytic acid concentrations in diets for wels catfish appear

marginal, for turbot, however, dietary phytic acid concentrations have probably reduced

phosphorus availability and finally reduced feed efficiencies and growth. In prospective

feeding trials with rapeseed protein products it appears advisory to supplement diets with a

phosphorus source such as dicalcium phosphate in order to overcome problems regarding

phosphorus availability (Lee et al. 2010). In comparison, no significant differences in whole

body composition were detected between carp or rainbow trout fed control diet and fish

receiving RPC or CPI diets, respectively. Furthermore, no correlations between dietary level

of CPI/ash/phosphorus and fish body parameters were identified.

Amino acid content

In the experiments with carp and wels catfish, high dietary inclusion of RPC may have led to

insufficient dietary supply of lysine. This might have negatively influenced feed efficiencies

and fish growth. Diets high in RPC were low in lysine, because of low fish meal inclusion and

no inclusion of other protein sources of animal origin. Additionally, ANF in RPC could have

negatively affected amino acid digestibility as it is known from other protein sources of

vegetable origin (Francis et al. 2001).

Blood features and histopathology

Certain blood values were determined in wels catfish and in rainbow trout fed the RPC.

Investigated blood values were not significantly different between treatment groups.

Consistent blood haemoglobin, haematocrit and serum values therefore indicate an equal

nutritional status among feeding groups (Congleton and Wagner 2006). In addition, a

histological investigation of the digestive tract of rainbow trout fed the RPC did not cause any

detectable histological alterations as found in Atlantic salmon when receiving diets that

107

contain high amounts of solvent extracted soybean meal (van den Ingh et al. 1991;

Baeverfjord and Krogdahl 1996).

References









Baeverfjord, G., Krogdahl, Å., 1996. Development and regression of soybean meal induced

enteritis in Atlantic salmon, Salmo salar L. distal intestine: a comparison with the

intestines of fasted fish. Journal of Fish Diseases 19, 375–387.











83, 653–664.








108

Congleton, J.L., Wagner, T., 2006. Blood-chemistry indicators of nutritional status in juvenile

salmonids. Journal of Fish Biology 69, 473-490.




263-274.








168–178.










magnesium levels on the nutrition of common carp, Cyprinus carpio. In: S.J. Kaushik &

Luquent, P. (eds.) Fish Nutrition in Practice, pp. 705-715. Proceedings of International

Conference, Biarritz, France, June 24–27, 1991.

Jobling, M., 1994. Fish bioenergetics. Chapman and Hall, London, UK.




Kousoulaki, K., Albrektsen, S., Langmyhr, E., Olsen, H.J., Campbell, P., Aksnes, A., 2009.

The water soluble fraction in fish meal (stickwater) stimulates growth in Atlantic salmon

(Salmo salar L.) given high plant protein diets. Aquaculture 289, 74–83.

109









Naczk, M., Shahidi, F., 1990. Carbohydrates of canola and rapeseed. In: Canola, Rapeseed:

Production, Chemistry, Nutrition & Processing Technology. (Shahidi, F. ed.), pp. 211-

220. Van Nostrand Reinhold, New York, NY.

NRC (National Research Council) (1993) Nutrient Requirements of Fish. National Academy





bass (Dicentrarchus labrax L.) juveniles. Aquaculture Research 35, 636-642.
















110





200.

Van Den Ingh, T.S.G.A.M., Krogdahl, Å., Hendriks, H.G.C.J.M., Koninkx, J.G.J.F., 1991.

Effects of soybean-containing diets on the proximal and distal intestine in Atlantic

salmon: a morphological study. Aquaculture 94, 297–305.



128.

111

General Summary

Fish meal, the most important source of marine protein for fish diets, is a limited resource.

Increasing prices for fish meal together with environmental concerns force the aquaculture

sector to find alternative protein sources to be included in fish feeds. Wide availability,

relatively high protein contents and a desirable amino acid profile have caused an interest in

rapeseed (and canola) products as ingredients for fish feed production. However, the

nutritional quality of rapeseed products largely depends on their levels of antinutritional

factors, particularly glucosinolates, phytic acid, phenolic constituents and indigestible

carbohydrates. Several processing techniques can be adapted to reduce the level of

antinutrients in rapeseed in order to improve its value for fish nutrition. In the present study a

high quality rapeseed protein concentrate (RPC) with a protein content of 71 % was evaluated

as fish meal replacement in diets of different fish species.

In chapter 1 the RPC was tested as fish meal replacement in diets for juvenile common carp

(Cyprinus carpio L.). Triplicate groups of fish were fed isonitrogenous (40.4 ± 0.2 % CP) and

isocaloric (21.4 ± 0.1 kJ g-1) experimental diets with 0 %, 33 %, 66 % or 100 % of fish meal

replaced with RPC. Results from a 56 day feeding trial showed, that growth parameters and

feed efficiencies were not significantly different between fish fed on the control diet and the

diet with 33 % of fish meal replaced with RPC. At higher RPC inclusion levels, diet intake

and feed efficiencies were reduced resulting in lower growth performances. It appeared that

diet taste and amino acid profiles were negatively affected by high dietary inclusion levels of

RPC.

In chapter 2 the potential of RPC as fish meal alternative in diets for juvenile wels catfish

(Silurus glanis L.) was evaluated. Fish were organized in triplicate groups and received

isonitrogenous (60.3 ± 0.3 % CP) and isocaloric (23.0 ± 0.3 kJ g-1) experimental diets with 0

%, 25 %, 50 % and 75 % of fish meal replaced with RPC. At the end of the 63 day feeding

period, growth performance, feed intake and feed efficiencies were not significantly different

between the control group and fish fed on diets with 25 % of fish meal replaced with RPC.

Higher dietary RPC inclusion negatively affected diet quality and palatability resulting in

reduced feed intake, feed efficiencies and fish performance.

In chapter 3 RPC was tested as fish meal alternative in diets for juvenile turbot (Psetta

maxima L.). Triplicate groups of fish were fed isonitrogenous (58.1 ± 0.9 % CP) experimental

diets with equal gross energy content (21.5 ± 0.3 MJ kg−1) where 0 %, 33 % and 66 % of fish

112

meal were replaced with RPC. At the end of the feeding period (84 days), fish fed the control

diet or the diet with 33 % of fish meal replaced with RPC showed significantly higher weight

gain and feed intake than fish fed diet with the highest RPC inclusion. This was mostly

attributed to negative effects on diet taste resulting from the glucosinolate content of RPC.

Furthermore, the control diet gave significantly better feed efficiencies than diets containing

RPC, probably due to lower protein, amino acid and phosphorus availability in these diets.

In chapter 4 RPC was tested as fish meal alternative in diets for juvenile rainbow trout.

Triplicate groups of fish received isonitrogenous (47.9 ± 0.5 % CP) and isoenergetic (22.4 ±

0.2 kJ g-1) experimental diets with 0, 66 and 100 % of fish meal substituted with RPC. At the

end of the 84 day feeding period, fish growth performance, feed intake and feed efficiencies

were not compromised when 100 % of fish meal in the control diet was replaced with RPC. In

addition, intestinal morphology did not reveal any histological abnormalities in all dietary

groups. Blood parameters including haematocrit, haemoglobin as well as glucose,

triglycerides and total protein in the plasma were not different between treatment groups.

In chapter 5 the potential of two rapeseed protein concentrates partitioned in albumin and

globulin fractions as fish meal alternatives was evaluated. These fractionized protein

concentrates were produced under lower cost and time effort compared to the rapeseed protein

concentrate in the experiments presented above. In a digestibility experiment with juvenile

rainbow trout apparent digestibility coefficients were determined by indirect marker method.

ADCs of protein from fish meal (89.2±1.1 %) and globulin concentrate (88.8±0.6 %) were

significantly higher than from albumin concentrate (77.7±1.4 %). In a consecutive growth

trial, juvenile rainbow trout were organized in triplicate groups and received experimental

diets with 0, 50, 75 or 100 % of fish meal replaced with albumin or globulin concentrate on

the basis of digestible protein. It was found that only in treatment group A50, fish growth

performance and feed intake were not negatively affected by dietary treatment. However, feed

efficienies were not significantly different compared to the control group at 100 % or 75 %

fish meal replacement level with albumin or globulin concentrate, respectively. Significant

lower fish survival rates were observed when fish received diets A75, A100, G50, G75, or

G100 compared to the control diet or diet A50. The experiment showed that the quality of the

fractionized protein concentrates was below that of the rapeseed protein concentrate used in

chapter 4.

In chapter 6 the potential of a canola protein isolate (CP: 81 %) as fish meal alternative in

diets for juvenile rainbow trout was evaluated. Apparent digestibility coefficients for protein

113

from fish meal (89.2±1.1 %) and canola protein isolate (84.6±1.8 %) were determined by

indirect marker method in a digestibility experiment with juvenile rainbow trout. In a

consecutive growth trial, fish organized in triplicate groups received experimental diets with 0

%, 25 %, 50 %, 75 % and 100 % of fish meal replaced with canola protein isolate on the basis

of digestible protein. At the end of a 70 day feeding period it was found that growth

performance, feed intake and feed efficiencies in treatment groups receiving diets devoid of

fish meal were not negatively affected compared to the control group.

Experimental results from all feeding trials conducted demonstrate an enormous potential of

high quality rapeseed protein products as protein source in fish feeds. Particularly for the

nutrition of rainbow trout, rapeseed protein concentrate and canola protein isolate appear to be

highly valuable fish meal alternatives.

114

Zusammenfassung

Fischmehl ist die wichtigste marine Proteinquelle für Fischfuttermittel und zugleich ein

limitierter Rohstoff. Steigende Fischmehlpreise sowie ökologische Bedenken bei dessen

Nutzung veranlassen die Aquakulturindustrie nach alternativen Proteinquellen für die

Nutzung in Fischfuttermitteln zu suchen. Weite Verfügbarkeit, ein relativ hoher Proteingehalt

sowie ein wünschenswertes Aminosäurenprofil führen zu wachsendem Interesse an

Rapsprodukten (einschliesslich Canola) als Rohstoff für die Fischfutterproduktion. Allerdings

hängt die nutritive Qualität von Rapsprodukten für die Fischernährung stark von deren Gehalt

an antinutritiven Faktoren ab, insbesondere Glucosinolaten, Phytinsäure, phenolischen

Verbindungen und unverdauliche Kohlenhydraten sind hierfür von Bedeutung. Verschiedene

Verarbeitungstechniken erlauben eine Senkung des Gehaltes der Antinutritiva in

Rapsprodukten und erhöhen deren Wert für die Fischernährung.

In der vorliegenden Studie wurde ein qualitativ hochwertiges Rapsproteinkonzentrat (RPK)

mit einem Rohproteingehalt von 71 % als Fischmehlersatz in Futtermitteln für Karpfen

(Cyprinus carpio L.), Wels (Silurus glanis L.), Steinbutt (Psetta maxima L.) und

Regenbogenforelle (Oncorhynchus mykiss W.) getestet.

In Kapitel 1 wurde das RPK als Fischmehlersatz in Futtermitteln für juvenile Karpfen

eingesetzt. Triplikate Fischgruppen erhielten isonitrogene (RP: 40.4 ± 0.2 %) Futtermittel, in

denen 0 %, 33 %, 66 % oder 100 % des Fischmehls durch RPK ausgetauscht worden war.

Nach einer 56-tägigen Fütterungsperiode waren Gewichtszunahme und Futterverwertung bis

zu einem Fischmehlaustausch von 33 % durch RPK nicht signifikant verschieden gegenüber

der Kontrollgruppe. Bei höherem Einsatz von RPK waren Futteraufnahme und

Futterverwertung signifikant verringert, was zu geringeren Gewichtszunahmen in den

betreffenden Gruppen führte. Dies geht vermutlich auf eine negative Beeinflussung des

Futtermittelgeschmacks sowie verschlechterter Aminosäuremuster in den Futtermitteln mit

sehr hohem RPK-Anteil zurück.

In Kapitel 2 wurde die Eignung von RPK als Fischmehlalternative in Futtermitteln für

juvenile Welse untersucht. Die Fische waren in triplikate Versuchsgruppen eingeteilt und

erhielten isonitrogene (RP: 60.3 ± 0.3 %) Futtermittel, in denen 0 %, 25 %, 50 % oder 75 %

des Fischmehls durch RPK ausgetauscht worden war. Am Ende der 63-tägigen

Fütterungsperiode waren Wachstumsleistungen, Futteraufnahme und Futterverwertung bis zu

einem Fischmehlaustausch von 25 % nicht signifikant verschieden zur Kontrollgruppe.

115

Fischmehlaustausch mit RPK von 50 % und 75 % hatte offensichtlich negative Effekte auf

Futtermittelqualität und –geschmack, wodurch es zu geringerer Futteraufnahme, verringerter

Futterverwertung und abnehmenden Wachstumsleistungen kam.

In Kapitel 3 wurde RPK als Fischmehlersatz in Futtermitteln für juvenile Steinbutt getestet.

Triplikate Fischgruppen wurden mit isonitrogenen (RP: 58.1 ± 0.9 %) Futtermitteln gefüttert,

bei denen 0 %, 33 % oder 66 % des Fischmehls durch Rapsproteinkonzentrat ausgetauscht

worden war. Am Ende des 84-tägigen Fütterungsszeitraums, zeigten Fische, die das

Kontrollfutter oder das Futter mit 33 %igem Fischmehlaustausch erhalten hatten, signifikant

höhere Gewichtszunahmen und Futteraufnahme als Fische, die das Futter mit 66 %igem

Fischmehlaustausch gefressen hatten. Die Unterschiede in der Futteraufnahme gehen

vermutlich auf negative Effekte auf den Futtermittelgeschmack durch Glucosinolate im RPK

zurück. Desweiteren erzielte das Kontrollfuttermittel bessere Futterverwertung gegenüber den

Futtermitteln, die RPK enthielten, was mit geringerer Protein-, Aminosäuren- und

Phosphorverfügbarkeit zusammenhängen dürfte.

In Kapitel 4 wurde RPK als Fischmelalternative in Futtermitteln für juvenile

Regenbogenforellen eingesetzt. Triplikate Fischgruppen erhielten isonitrogene (RP: 47.9 ±

0.5 %) Futtermittel, in denen 0 %, 66 % oder 100 % des Fischmehls durch RPK ersetzt

worden war. Am Ende des 84-tägigen Fütterungszeitraums zeigten die Wachstumsleistungen,

die Futteraufnahme und die Futterverwertung zwischen den Versuchsgruppen keine

signifikanten Unterschiede. Auch der lichtmikroskopisch untersuchte Verdauungstrakt zeigte

zwischen den Gruppen keine Unterschiede. Desweiteren waren Blutparameter wie Hämatokrit

und Hämoglobin sowie Glucose, Triglyceride und Gesamtprotein gruppenübergreifend

einheitlich.

In Kapitel 5 kamen zwei Rapsproteinkonzentrate zur Anwendung. Die in eine Albumin- und

eine Globulinfraktion aufgeteilten Proteinkonzentrate wurden unter geringerem Kosten- und

Zeitaufwand als das in vorangegangenen Experimenten verwendete RPK hergestellt. In einer

Verdaulichkeitsuntersuchung an juvenilen Regenbogenforellen wurde die

Nährstoffverdaulichkeit der Proteinkonzentrate mittels der indirekten Markermethode erfasst.

Die Verdaulichkeit des Proteins aus Fischmehl (89.2±1.1 %) und Globulinkonzentrat

(88.8±0.6 %) war signifikant höher als bei dem Albuminkonzentrat (77.7±1.4 %). In einem

folgenden Wachstumsversuch erhielten in triplikate Gruppen eingeteilte Forellen Futtermittel,

bei denen 0 %, 50 %, 75 % oder 100 % des Fischmehls auf der Basis verdaulichen Proteins

durch Albumin- oder Globulinkonzentrat ausgetauscht worden war. Lediglich in der

116

Versuchsgruppe mit 50 % Fischmehlaustausch durch Albuminkonzentrat waren die

Wachstumsleistungen, die Futteraufnahme und die Überlebensrate einheitlich mit den

Ergebnissen aus der Kontrollgruppe. Allerdings war die Futterverwertung nicht signifikant

schlechter gegenüber der Kontrollgruppe bis zu 100 % bzw. 75 % Fischmehlaustausch durch

Albumin- oder Globulinkonzentrat. Die Untersuchung zeigte, dass die nutritive Qualität der

kostengünstiger hergestellten Proteinkonzentrate unter der des RPK aus Kapitel 4 liegt.

In Kapitel 6 wurde Canolaproteinisolat (RP: 81 %) als Fischmehlalternative in Futtermitteln

für juvenile Regenbogenforellen eingesetzt. Die Verdaulichkeit von Protein aus Fischmehl

(89.2±1.1 %) und Canolaproteinisolat (84.6±1.8 %) wurde über die indirekte Markermethode

in einem Verdaulichkeitsexperiment erfasst. In einem darauf aufbauenden Wachstumsversuch

wurden triplikate Fischgruppen mit Futtermitteln gefüttert, bei denen 0 %, 25 %, 50 %, 75 %

oder 100 % des Fischmehls auf der Basis verdaulichen Proteins durch Canolaproteinisolat

ersetzt worden war. Am Ende der 70-tägigen Fütterungsperiode zeigten die

Wachstumsleistungen, Futteraufnahme und Futterverwertung keine negative Beeinträchtigung

trotz 100 %igem Fischmehlaustauschs.

Die Ergebnisse aus allen Fütterungsversuchen demonstrieren das große Potenzial von

qualitativ hochwertigen Rapsproteinprodukten für die Fischernährung. Besonders für die

Ernährung von Regenbogenforellen wurden RPK und Canolaproteinisolat als hervorragende

Fischmehlalternativen identifiziert.

117

Danksagung

An dieser Stelle möchte ich den Menschen Dank entgegenbringen, die zum Gelingen der

vorliegenden Arbeit beigetragen haben.

Ich bedanke mich bei meinem Betreuer Herrn Prof. Dr. Carsten Schulz für die Überlassung

des interessanten Themas und das mir geschenkte Vertrauen bei der Projektbearbeitung. Die

gute fachliche Betreuung und die ermöglichte Teilnahme an nationalen und internationalen

Konferenzen förderten ein positives Arbeitsklima.

Herrn Simon Kreft und Frau Nina Bajdura danke ich für die verwaltungstechnische

Bearbeitung des Projektes.

Dank gilt meinen lieben Kolleginnen und Kollegen vom vTI für die Unterstützung bei der

Versuchsdurchführung und die überaus schöne Zeit in Ahrensburg. Besonders bei Herrn Prof.

Dr. Hilge bedanke ich mich, dass ich als Gast in Ahrensburg arbeiten durfte.

Den Kollegen vom PPM danke ich für die gute Zusammenarbeit.

Ich bedanke mich bei Herrn Dr. Florian Nagel für die Hilfsbereitschaft bei der

Versuchsdurchführung und die angenehme Büronachbarschaft.

Meinen Eltern danke ich für die wie immer selbstlose Unterstützung jeglicher Art.

Und ganz besonders danke ich Dir, Sophie, weil Du mir allzeit Kraft gibst.

118

Lebenslauf

Name Hanno Slawski

Geburtstag 28.09.1981

Geburtsort Neustadt in Holstein, Schleswig-Holstein

Staatsangehörigkeit Deutsch

Familienstand ledig

Schulbildung

1988 – 1992

1992 - 2001

Hochtorgrundschule Neustadt in Holstein

Kreisgymnasium Neustadt in Holstein

Zivildienst

09/2001 – 07/2002

Karl-Schütze-Heim für Menschen mit geistiger Behinderung,

Merkendorf

Studium

10/2002 – 08/2004

03/2005 – 08/2006

09/2006 – 10/2008

Bachelorstudiengang Agrarwissenschaften an der

Christian-Albrechts-Universität zu Kiel

Bachelorstudiengang Agrarwissenschaften an der

Humboldt-Universität zu Berlin

Masterstudiengang Fishery Science and Aquaculture an der

Humboldt-Universität zu Berlin

Berufliche Tätigkeiten

11/2008 – 03/2011

01/2011 – 03/2011

seit 04/2011

Wissenschaftlicher Mitarbeiter bei der Gesellschaft für Marine

Aquakultur mbH, Büsum, bei Herrn Prof. Dr. C. Schulz

Wissenschaftlicher Mitarbeiter am Institut für Tierzucht und

Tierhaltung der Christian-Albrechts-Universität zu Kiel bei

Herrn Prof. Dr. C. Schulz

R&D Manager bei Aller Aqua A/S

RAPESEED PROTEIN PRODUCTS AS FISH MEAL REPLACEMENT … · protein in fish feeds, is a limited...

Documents

Transcript of RAPESEED PROTEIN PRODUCTS AS FISH MEAL REPLACEMENT … · protein in fish feeds, is a limited...