Aquaculture nutrition, tập 19, số 1, 2013

Fish fed with the basal diet showed the lowest final bodyweight, feed intake, feed efficiency, SGR and PER signifi-cantly increased with the levels of dietary valine increased showed sig

1 1,2,3 1,4 1,3 1,3 1,3 1,3

1

Production University Key Laboratory of Sichuan Province, Sichuan Agricultural University, Sichuan, Ya’an, China;

4

Animal Nutrition Institute, Sichuan Academy of Animal Science, Chengdu, China

This study investigated the effects of valine on growth,

intes-tinal enzyme activities and microflora in juvenile Jian carp

(Cyprinus carpio var Jian) A total of 1200 fish with an

5.3 (unsupplemented control), 8.7, 11.8, 14.9, 18.7 and

the specific growth rate, feed efficiency, body protein and

lipid content of fish were significantly improved by the

activ-ities of trypsin, amylase, lipase, chymotrypsin, glutamate

oxaloacetate transaminase (GOT) and glutamate pyruvate

transaminase (GPT) took the similar trends Similarly, the

optimum levels of dietary valine induced increases in the

intestinal length, weight, folds height and activities of

alka-line phosphatase, gamma-glutamyl transpeptidase and

crea-tine kinase In contrast, the trends of muscle GOT activity

and plasma ammonia content were opposite Intestinal

Aero-monas, Escherichia coli, Lactobacillus and Bacillus were

changed by dietary valine supplementations The dietary

Together, these results indicated that valine improved fish

growth, digestive and absorptive ability

microflora, trypsin, valine

Received 16 April 2011; accepted 9 December 2011

Correspondence: Xiao-Qiu Zhou, Animal Nutrition Institute, Sichuan

Agricultural University, Ya’an 625014, China E-mails: zhouxq@sicau.

edu.cn; xqzhouqq@tom.com

Fish, like other animals, do not have a true protein ment but require a well-balanced mixture of essential andnon-essential amino acids (Wilson 2002) The essentialamino acids for common carp (Cyprinus carpio) and chi-nook salmon (Oncorhynchus tshawytscha) are the same 10amino acids that are required by the rat (Santiago & Lovell1988) However, the complete 10 essential amino acidrequirement have been established for only a limited num-ber of cultured fish species (Abidi & Khan 2004) Valinewas known as an essential amino acid for fish (Nose 1979)

require-It has been demonstrated that dietary valine deficiencydecreased the weight gain of common carp (Nose 1979)and Indian major carp (Cirrhinus mrigala) (Ahmed & Khan2006) The weight gain of fish is associated with the accre-tion of protein (Bureau et al 2000) Sveier et al (2000)reported that protein deposition in fish is mainly associatedwith amino acid metabolism Fish have a remarkablecapacity to utilize amino acids both as metabolic fuel and

as precursors for protein, lipid and carbohydrate synthesis(Wood 1993) Glutamate oxaloacetate transaminase (GOT)and glutamate pyruvate transaminase (GPT) play animportant role in ammonia detoxification in freshwater tel-eost (Hochachka & Somero 1973) Furthermore, fishexcrete the majority of their nitrogenous waste as ammonia(Wood 2001) Plasma ammonia levels are the varied widelymeasurements of excretory processes for hagfish (Eptatretusstoutii) (Arillo et al 1991) The GOT and GPT activitiesand plasma ammonia content were affected by the levels ofdietary protein in juvenile Jian carp (C carpio var Jian)(Liu et al 2009) However, the relationships between die-tary valine and activities of GOT and GPT as well as

.

Aquaculture Nutrition

plasma ammonia content (PAC) of fish are unclear.

Accordingly, further studies are required to address the

effects of valine on amino acid metabolism in fish

The intestine and pancreas are important for nutrient

digestion and absorption of fish (Garcı´a-Gasca et al 2006)

Pedersen & Sissons (1984) reported that the growth and

development of intestine and pancreas play an important

role in digestion ability and absorption function of calf To

our knowledge, few studies have been conducted to

investi-gate the effects of valine on the growth of intestine and

hepatopancreas of fish It has been demonstrated that the

jejunal crypt depth in rat was affected by valine (Takada

decreased in rat fed the valine-devoid diet (Sidransky et al

1960) Moreover, BCAA (valine, leucine and isoleucine)

can provide nitrogen for the synthesis of glutamate in rat

(Bixel et al 1997) and pig (Chen et al 2009) Lin & Zhou

(2006) showed that glutamine can improve intestine protein

content (IPC) in Jian carp These results indicated that

valine may affect the growth of the intestine and pancreas

of fish, which needs to be investigated

Digestion and absorption of nutrients depend on the

activities of the digestive and brush border enzymes (Klein

digestive enzymes into the intestinal lumen, including

lipase, amylase, trypsin and chymotrypsin (Hitoshi et al

gradient that drives amino acid and vitamin transport into

cells and is critical for absorption of fluid from the

intes-tine (Lingrel 2010) Gamma-glutamyl transpeptidase plays

an essential role in the final hydrolysis and assimilation

of dietary proteins (Douglas et al 1999) Furthermore,

Villanueva et al (1997) reported that the absorption of

nutrients such as lipid, glucose, calcium and inorganic

phosphate was depended on the activity of alkaline

phos-phatase Additionally, creatine kinase plays a key role in

the energy homeostasis (Wallimann et al 1998) Sidransky

showed moderate loss of zymogen granules in rat fed with

valine-devoid diet In human, valine can stimulate

pancre-atic enzyme secretion (Go et al 1970) Bixel et al (1997)

reported that valine presented as vehicle molecule of

nitrogen in the glutamate/glutamine cycle in rat

Gluta-mine can increase the activities of intestinal lipase,

juvenile Jian carp (Lin & Zhou 2006) These data

indi-cated that valine may influence the activities of digestive

and brush border enzymes in fish, which needs to be

investigated

Microflora is associated with the gastrointestinal tract offish and is sensitive to dietary changes (Ringø & Birkbeck1999) Human intestinal microorganisms provide instructivesignals for several aspects of intestinal development, includ-ing epithelial cell maturation (Hooper et al 2003) andangiogenesis (Stappenbeck et al 2002) Sugita et al (1996)indicated that the bacteria in marine crab digestive tractsecreted amylase Amino acids can be used by microorgan-isms for construction of specific cell proteins or canundergo transformations leading to produce different meta-bolic substances (Ringø & Birkbeck 1999) Valine can beused as an effective source of nitrogen in the process ofmultiplication of the bacteria Desulfotomaculum ruminis(Szyman´ska et al 2002) These results indicated that valinemay be related to the growth of fish intestinal microflora,which needs to be addressed

Therefore, we hypothesize that valine could increasefish growth through increasing digestive and absorptivecapacity and influencing the balance of intestinal microfl-ora Hence, the purpose of this study was to investigatethe effects of dietary valine on the growth, body compo-sition, enzyme activities and intestinal microflora of fish,which could provide a part of theoretical evidence forthe effect of valine on fish growth The optimum dietaryvaline requirement for the juvenile Jian carp was alsoevaluated

The basal diet was formulated in Table 1 Fish meal andgelatin were used as the main protein sources because theyare limited in valine The dietary protein level was fixed at

(Liu et al 2009) Crystalline amino acids (Jiangsu NantongEastern Amino Acid Co Ltd., Nantong, China) were used

whole chicken egg protein, excluding the test amino acidvaline, according to the method described by Abidi &

to provide graded concentrations of 6 (unsupplemented

made isonitrogenous by supplementation of glycine Zinc,iron, pyridoxine, pantothenic acid, inositol, thiamin andriboflavin were formulated to meet the nutrient require-ments of Jian carp according to previous studies conducted

in our laboratory (He et al 2009; Jiang et al 2009; Wen

.

2011; Tan et al 2011) The levels of other nutrients met

the requirements for common carp according to the NRC

(National Research Council) (1993) The pH of each diet

was adjusted to 7.0 by gradually adding 6.0 M NaOH

(Li et al 2009) After prepared completely, the diets were

used (Shiau & Lo 2000) The valine concentrations in

experimental diets were determined according to the

method described by Ahmed & Khan (2006) to be 5.5(unsupplemented control), 9.7, 12.5, 15.6, 18.7 and 20.5 g

All experimental protocols were approved by the sity of Sichuan Agricultural Animal Care Advisory Com-mittee The juvenile Jian carp were obtained from theTong Wei Hatchery (Sichuan, China) Fish were accli-mated to the experimental conditions for one month Atotal of 1200 Jian carp, with a mean initial weight of

closed recirculating water system with continuous aeration.Each aquarium was randomly assigned to one of four rep-licates of the six dietary treatments The water change rate

was drained through biofilters so as to reduce ammoniaconcentration and remove solid substances in the water.The experimental units were maintained under a naturallight and dark cycle Dissolved oxygen was higher than

their respective diets to apparent satiation six times perday for the first 30 days and four times per day from 31st

to 60th days This feeding rhythm was established in vious study performed by our laboratory (Xiao et al.2011) Uneaten feed was removed by siphoning after eachmeal

pre-Fish from each aquarium were weighed at the beginningand at the end of the feeding trial At the beginning ofthe experiment, 30 fish from the same population wererandomly collected to determine the initial body composi-tion At the end of the feeding trial, five fish from eachaquarium were frozen for estimating the final protein,lipid and ash composition (AOAC (Association of Offi-cial Analytical Chemists) 1998) After 12 h of fasting, 15fish from each aquarium were killed and the muscle,intestine and hepatopancreas were quickly collected and

five fish from each aquarium were sampled to measurethe height of intestinal folds according to Lin & Zhou(2006)

Six hours after the last feeding, five fish from eachaquarium were collected for obtaining blood samples

Table 1 Composition and nutrients content of basal diet

1.3

phosphorus, n-3 and n-6 contents were calculated according to

NRC (National Research Council) (1993) and Bell (1984).

leucine 22.57, lysine 18.19, methionine 8.35, cysteine 0.51,

pheyl-alanine 15.03, tyrosine 12.17, threonine 12.90, tryptophan 3.95,

0.76 g All ingredients were diluted with corn starch to 1 kg.

Mn), 4.09 g;

L-gly-cin were reduced to compensate, each mixture was made

and 755.4, 724.5, 693.6, 662.7, 631.8, 600.9 g corn starch,

respec-tively.

from the caudal vein with heparinised syringes Prior to

sampling, fish in each aquarium were fed solely and blood

was withdrawn in order The blood samples were

centri-fuged at 4000 g for 15 min (Liu et al 2009) Plasma was

collected for ammonia determination using the method of

Tantikitti & Chimsung (2001)

The hepatopancreas, intestine and muscle samples were

homogenized in 10 volumes (w/v) of ice-cold physiological

saline solution and centrifuged at 6000 g for 15 min at

activities of trypsin (Hummel 1959), chymotrypsin

(Hum-mel 1959), amylase (Furne´ et al 2005), lipase (Furne´ et al

2005), alkaline phosphatase (Bessey et al 1946),

creatine-kinase (Tanzer & Gilvarg 1959), gamma glutamyl

hepatopancreas was determined by the method of Bradford

(1976) Three fish from each aquarium were collected

for estimating the counts of intestinal microflora The

activities of GOT and GPT in muscle and hepatopancreas

were determined by the methods of Bergmeyer & Bernt

(1974a,b)

Data on initial body weight (IBW), final body weight

(FBW), feed intake (FI), proximate composition of feed

and carcass, hepatopancreas and intestinal weight,

intesti-nal and body length, and hepatopancreas and intestiintesti-nal

protein were used to calculate the following parameters:

 Percent weight gain (PWG) = 100 9 (g weight gain/g

initial body weight)

 Feed efficiency (FE) = (g weight gain/g feed intake)

weight) - ln(initial weight))/number of days]

 Protein efficiency ratio (PER) = g weight gain/g

pro-tein intake

 Protein retention value (PRV) = 100 9 (g fish protein

gain/g protein intake)

 Relative gut length (RGL) = 100 9 (cm intestine

length/cm total body length)

 Hepatosomatic index (HSI) = 100 9 (g wet

hepato-pancreas weight/g wet body weight)

 Intestosomatic index (ISI) = 100 9 (g wet intestine

weight/g wet body weight)

 Intestine protein content (IPC) = 100 9 (g intestine

protein/g wet intestine weight)

 Hepatopancreas protein content (HPC) = 100 9 (g

hepatopancreas protein/g wet hepatopancreas weight)

Differences among the dietary treatments were

level of significance through SPSS 17.0 for windows Aquadratic regression model was used to determine theoptimal level of dietary valine (Zeitoun et al 1976) Therelationship between dietary valine and the growth per-formance, whole body composition, activities of digestiveand absorptive enzymes were subjected to a quadraticregression model

Table 2 shows the final body weight, feed intake, FE, SGRand PER of juvenile Jian carp fed graded levels of valine

Fish fed with the basal diet showed the lowest final bodyweight, feed intake, feed efficiency, SGR and PER

signifi-cantly increased with the levels of dietary valine increased

showed significantly quadratic responses to the increasinglevels of dietary valine as well as significant response to themean valine intake per fish Feed efficiency significantlyincreased with increasing levels of dietary valine up to

addi-tion, FE and PER were significantly quadratic responses toincreasing dietary valine levels The dietary valine require-

As shown in Table 3, fish body moisture, protein and lipidcontent and PRV were significantly affected by dietaryvaline The body protein content and PRV were the lowest

However, the moisture of fish carcasses was not influenced

by the levels of dietary valine Moreover, regression analysisshowed that fish body protein content, lipid and PRV were .

quadratic responses to the increasing levels of dietary

valine The ash of fish carcasses was the lowest in Jian carp

content was quadratic response to the increased levels of

dietary valine

Plasma ammonia content (PAC) and the activities of GOTand GPT of juvenile Jian carp are shown in Table 4 GOTand GPT activities in the hepatopancreas significantly

Regres-sion analysis showed that GOT and GPT activities in thehepatopancreas were quadratic responses to increasing lev-els of dietary valine In addition, GPT activities in musclesignificantly increased with levels of dietary valine up to

decreased with the increment levels of dietary valine

activities in muscle were quadratic responses to the ing levels of dietary valine, respectively Plasma ammoniacontent significantly decreased and the minimum value

regres-sion analysis showed that plasma ammonia content (PAC)was significantly quadratic response to increasing levels ofdietary valine

Hepatopancreas weight, hepatosomatic index, creas protein content, intestinal length and weight, relativegut length, intestosomatic index and intestinal proteincontent of juvenile Jian carp fed with diets containing

hepatopan-Figure 1 Quadratic regression analysis of specific growth rate

(SGR) according to dietary valine levels (a) and against mean

groups of Jian carp with 50 fish per group The dietary valine

graded levels of valine are presented in Table 5

Hepatopan-creas weight significantly inHepatopan-creased and the maximum

value occurred in fish fed with the diet containing

similar trend was observed in the intestinal weight

Hepato-somatic index (HSI) and intestoHepato-somatic index (ISI) of fish

Hepatopancreas and intestinal protein content were not

Regression analysis showed that the HW, IW and ISI

were quadratic responses to the levels of dietary valine

Intestinal length significantly increased and the maximum

thereafter A similar trend was observed in RGL more, regression analysis showed that the IL and RGL werequadratic response to the levels of dietary valine

Further-The folds height in all intestinal segments is given inTable 6 Intestinal folds height in the proximal intestine(PI) was the highest in fish fed with a diet containing

in the mid (MI) and distal intestines (DI) showed the lar trends with that in PI The folds height in the PI washigher than that in the MI and DI Regression analysissuggested that the folds height in the PI, MI and DI werequadratic response to the levels of dietary valine

Values are mean ± SD of four replicate tanks, with five fish in each replicate.

) of juvenile Jian carp fed diets containing graded levels

Mean values with the different superscripts in the same row are significantly different (P < 0.05).

.

The activities of trypsin, lipase, chymotrypsin and amylase

in the intestine were enhanced with the increment levels of

dietary valine (Table 7) Trypsin and lipase activities were

the highest in Jian carp fed with a diet containing

in amylase activity The chymotrypsin activity in the

intes-tine was significantly lower in fish fed with diets containing

analysis showed that the trypsin, chymotrypsin, lipase and

amylase in the intestine were quadratic response to the els of dietary valine Trypsin and lipase activities in thehepatopancreas were the highest in Jian carp fed with the

and chymotrypsin activities in the hepatopancreas werethe highest in Jian carp fed with the diet containing

that the trypsin, lipase and amylase activities in the pancreas were quadratic response to the levels of dietaryvaline (Table 8)

c-GT and CK in the intestine are presented in Table 9

hepatosomatic index (HSI), relative gut length (RGL), hepatopancreas protein content (HPC) and intestinal protein content (IPC) of juvenile Jian carp fed diets containing graded levels of valine for 60 days

Table 6 Folds height (lm) in proximal intestine (PI), mid intestine (MI), distal intestine (DI) of juvenile Jian carp fed diets containing

Mean values with the different superscripts in the same row are significantly different (P < 0.05).

The activities of alkaline phosphatase in the PI, MI and

DI were the highest in Jian carp fed with the diet

the PI, MI and DI were quadratic response to the levels

activity was the highest in fish fed with a diet containing

to increasing levels of dietary valine The activity of CK

in the PI, MI and DI was the highest in fish fed with a

analy-sis showed that the CK activity in the MI was

signifi-cantly quadratic response to the increasing levels of

dietary valine

The counts of Lactobacillus, Bacillus, Escherichia coli and

intesti-nal Aeromonas was significantly higher than that of fish fed

Regression analysis showed that the populations of nal Lactobacillus, E coli and Aeromonas were quadraticresponses to increasing levels of dietary valine

intesti-The importance of dietary valine for normal growth of Jiancarp was demonstrated in the present study In this study, a

Mean values with the different superscripts in the same column are significantly different (P < 0.05).

Mean values with the different superscripts in the same column are significantly different (P < 0.05).

.

reduced weight gain was observed in Jian carp fed with the

valine-insufficient diet Similar observations have been

reported in Indian major carp (Ahmed & Khan 2006) A

reduction in feed intake was regarded as the primary factorresponsible for the depressed growth observed in Atlanticsalmon fry (Rollin et al 2006) Ahmed & Khan (2006)

Table 9 Activities of alkaline phosphatase (AKP, mmol of nitrophenol released per gram tissue per hour), Na+,K+-ATPase (lmol of

demonstrated that a depressed growth rate in Indian major

carp (C mrigala) fed a diet containing less than the optimum

amount of valine was due to loss of appetite and poor feed

efficiency This study also showed that the deficiency of

die-tary valine caused a decrease in FI Moreover, regression

analysis revealed that FI was significantly quadratic response

to the increasing levels of dietary valine and significant

response to the mean valine intake per fish (Fig 2)

Additionally, correlation analysis indicated that the FI was

(Fig 3) These results indicated that the increased SGR,

PER and FE in fish fed diet with valine supplementation

may be partly associated with the increased FI Bureau et al

(2000) found that fish weight gain is associated with the

accretion of protein, fat etc In the present study, fish fed

body protein and PRV These data indicated that valine

enhanced the protein utilization of fish Similar observations

were reported in Indian major carp (Ahmed & Khan 2006)

Yoshizawa (2004) reported that BCAA (valine, leucine and

isoleucine) played an important role in protein synthesis in

mammals In addition, fish body lipid content were

signifi-cantly lower when dietary valine exceeding 14.9 g valine

major carp, Labeo rohita (Hamilton) fry (Abidi & Khan

2004) Based on the quadratic regression analysis for SGR,

the valine requirement of juvenile Jian carp was estimated to

was higher than that for common carp (NRC 1993) It mayattribute to the difference in fish species Jian carp is the newvariety of C carpio in China (Sun et al 1995) and grows30% faster than the common carp (Dong & Yuan 2002) It isconsistent with the studies that Jian carp had higher nutrientrequirements than common carp, such as lysine (Zhou

2009), thiamine (Huang et al 2011) and zinc (Tan et al

2011)

Lim et al (2001) reported that reduction in the rate ofproteolysis and amino acid catabolism resulted in adecrease in ammonia production of mudskippers (perioph-

study, the plasma ammonia content was lower for fish fedwith optimum levels of dietary valine It suggested that theoptimal valine supplementation reduced the production ofammonia, supporting a higher protein efficiency ratio inthis group GOT and GPT play an important role in pro-tein and amino acid catabolism (Balogun & Fetuga 1980;

Segner & Verreth 1995) In the present study, GOT and

Figure 2 Quadratic regression analysis of feed intake (FI)

accord-ing to dietary valine levels (a) and against mean valine intake per

carp with 50 fish per group.

Figure 3 Linear regression analysis between feed intake (FI) and SGR (a), feed efficiency (FE, b) and protein efficiency ratio (PER, c) of juvenile Jian carp fed diets containing graded levels of valine

Jian carp with 50 fish per group.

.

GPT activities in the hepatopancreas and GPT activity in

muscle were significantly increased with higher dietary

suggested that optimum valine might improve the

utiliza-tion of amino acid and protein However, the GOT activity

in muscle was decreased with the increment levels of

die-tary valine up to a certain point In mammals, valine is

metabolized primarily in muscle, whereas most other amino

acids are metabolized mainly in liver (Young 1970) The

catabolism of valine is initiated by transamination resulting

2009) NADP-dependent glutamate dehydrogenases (GDHs)

have a role in reversible conversion of glutamate and

a-ketoglutarate (Boles et al 1993) Glutamate

dehydro-genase present will compete with GOT and GPT for

(Fahienl et al 1977) But there was no more information

about the mechanism by which dietary valine decreased the

GOT activities in muscle of fish, and the possible

mecha-nism needs to be further studied

The growth and development of the intestine and pancreas

play an important role in the digestion ability and absorption

function of calf (Pedersen & Sissons 1984) In this study, the

weight of hepatopancreas and intestine were significantly

improved in fish fed diets with valine supplementation

How-ever, few studies have evaluated the effects of valine on the

growth of the intestine and hepatopancreas in fish It was

reported that valine could influence the pancreatic protein

and jejunal crypt depth in rat (Sidransky et al 1960; Takada

isoleu-cine) can provide nitrogen for the synthesis of glutamate in

rat (Bixel et al 1997) and pig (Chen et al 2009) These data

supported the present result that valine could improve the

growth of intestine and hepatopancreas Lin & Zhou (2006)

showed that amino acids, such as glutamine, also could

improve the growth and development of the digestive tract

and associated organs in fish

The gastrointestinal tract enzyme profile is an indicator

of nutrient digestibility and utilization (Deng et al 2010)

In this study, the activities of trypsin, lipase, chymotrypsin

and amylase in the hepatopancreas and intestine enhanced

with the increment levels of dietary valine, which indicated

that valine improved the digestive ability of fish, supporting

the protein retention value results The pancreas can secrete

a large number of digestive enzymes into the intestinal

lumen, including lipase, amylase, trypsin and chymotrypsin

(Hitoshi et al 2007) In this study, there are significant

cor-relation between trypsin, lipase and amylase activities in

rlipase= +0.914, P < 0.05; ramylase= +0.983, P < 0.01) Thesedata indicated that the increase of intestine digestive enzymeactivities may be partly due to the promotion of hepatopan-creatic enzyme activities by valine However, in fish, fewstudies have evaluated the effects of valine on the activities

of digestive enzymes In human, valine could stimulate creatic enzyme secretions (Go et al 1970) Moreover, Si-dransky et al (1960) reported that the acinar cells of thepancreas showed moderate loss of zymogen granules in ratsfed valine-devoid diet These data supported that valinecould improve the digestive ability of fish

pan-Nutrient absorption in fish intestine is related to gutmicrovilli and brush border enzyme activities (Kjørsvik

the absorption ability in aquatic animals (Farhangi et al.2001) In the present study, intestinal folds height in PI,

MI and DI were improved with increasing dietary valinelevels, which indicated that valine improved the absorptionability of intestine However, no studies have evaluated the

and vitamin transport into cells and is critical for tion of fluid from the intestine (Lingrel 2010) Gamma glut-amyl transpeptidase play an essential role in the finalhydrolysis and assimilation of dietary protein (Douglas

absorption of nutrients such as lipid, glucose, calcium andinorganic phosphate is depended on alkaline phosphatase.Creatine kinase plays a key role in the energy homeostasis(Wallimann et al 1998) Furthermore, Sweeney & Klip(1998) reported that the absorption of most nutrients weredependent on active transport which needs energy expendi-ture In the present study, dietary valine significantly

gamma-glutamyl transpeptidase activities in fish However,

at present, it has scarcely reported that dietary valineaffects brush border enzyme activities A similar tendencywas reported in a previous study from our laboratory, inwhich the glutamine (Lin & Zhou 2006) and lysine (Zhou

activities of juvenile Jian carp

Microflora is associated with the healthy of nal tract of fish (Ringø & Birkbeck 1999) Lactobacilli arerequired to maintain a healthy intestine (Aguirre & Collins1993) It produces bacteriocin-like substances controllingovergrowth of potentially pathogenic bacteria (Boris &Barbes 2000) Adversely, Aeromonas species are regarded

gastrointesti-as enteropathogens and in some cgastrointesti-ases are linked to ggastrointesti-astro-enteritis (Merino et al 1995) Salinas et al (2008) reported

gastro-that Lactobacillus delbrueckii subsp lactis was able to

pre-vent foregut tissue damage caused by Aeromonas

in vitro Additionally, E coli in the intestine are pathogens

with the particular potential of causing enteric infection

(Rio-Rodriguez et al 1997) These intestinal microflora are

sensitive to dietary changes (Ringø & Birkbeck 1999)

Szy-man´ska et al (2002) reported that valine can be a source

of nitrogen in the process of multiplication of the bacteria

of valine could promote the growth of Lactobacillus and

indicated that the optimum level of valine could influence

the balance of intestinal microflora by promoting the

growth of beneficial bacterium and depressing the growth

valine improved the growth of Lactobacillus, E coli and

Aeromonas At present, it has scarcely reported that dietary

valine affects microbial population However, a similar

ten-dency was reported in a previous study from our

labora-tory, in which the effect of methionine on fish intestinal

microflora has been considered (Tang et al 2009)

How-ever, the mechanism by which the dietary valine affects the

intestinal microflora of fish is still unknown

Therefore, we conclude that dietary valine could improve

fish growth and increase body protein and lipid

composi-tion and enhance hepatopancreatic and intestinal enzyme

activities as well as influence the balance of intestinal

mi-croflora of juvenile Jian carp The mechanism by which

valine enhanced activities of digestive and absorptive

enzymes and improved intestinal microbial population

should be further studied The dietary valine requirement

This research was financially supported by the National

Department Public Benefit Research Foundation

(Agricul-ture) of China (201003020) and Program for New Century

authors would like to thank the personnel of these teams

for their kind assistance

Abidi, S.F & Khan, M.A (2004) Dietary valine requirement of

Indian major carp, Labeo rohita (Hamilton) fry J Appl

Aguirre, M & Collins, M.D (1993) Lactic acid bacteria and

Ahmed, I & Khan, M.A (2006) Dietary branched-chain amino acid valine, isoleucine and leucine requirements of fingerling Indian major carp, Cirrhinus mrigala (Hamilton) Br J Nutr.,

96, 450–460.

AOAC (Association of Official Analytical Chemists) (1998) cial Methods of Analysis, 16th edn AOAC International, Wash- ington, DC.

Offi-Arillo, A., Margiocco, C., Melodia, F., Mensi, P & Schenone, G.

(1981) Ammonia toxicity mechanism in fish: studies on Rainbow

Balogun, O.O & Fetuga, B.L (1980) Liver glutamate-oxalate transaminase and glutamate-pyruvate transaminase activity in pigs as influenced by dietary methionine and lysine levels Bio-

Bell, J.M (1984) Nutrients and toxicants in rapeseed meal: a

Bergmeyer, H.U & Bernt, E (1974a) Glutamate-oxaloacetate transaminase, UV-assay, manual method In: Methods of Enzy-

Press, New York.

Bergmeyer, H.U & Bernt, E (1974b) Glutamate-pyruvate aminase, UV-assay, manual method In: Methods of Enzymatic

New York.

Bessey, O.A., Lowry, O.H & Brock, M.J (1946) A method for the rapid determination of alkaline phosphatase with five cubic milli-

Bixel, M.G., Hutson, S.M & Hamprecht, B (1997) Cellular bution of branched-chain amino acid aminotransferase isoen- zymes among rat brain glial cells in culture J Histochem.

Boles, E., Lehnert, W & Zimmermann, W (1993) The role of the NAD-dependent glutamate dehydrogenase in restoring growth

on glucose of a Saccharomyces cerevisiae phosphoglucose

Boris, S & Barbes, C (2000) Role played by lactobacilli in

Bradford, M (1976) A rapid and sensitive method for the cation of microgram quantities of protein utilizing the principle

Bureau, B.P., Azevedo, P.A., Tapia-Salazar, M., Cuzon, G (2000) Pattern and cost of growth and nutrient deposition in fish and shrimp: potential implications and applications In: Avances en Nutricio´n Acuı´cola V Memorias del V Simposium Internacional

Me´rida, Yucata´n, Mexico.

Chen, L.X., Li, P., Wang, J.J., Li, X.L., Gao, H.J., Yin, Y.L., Hou, Y.Q & Wu, G.Y (2009) Catabolism of nutritionally essen- tial amino acids in developing porcine enterocytes Amino Acids,

37, 143–152.

Deng, J.M., Mai, K.S., Ai, Q.H., Zhang, W.B., Tan, B.P., Xu, W.

& Liufu, Z.G (2010) Alternative protein sources in diets for anese flounder Paralichthys olivaceus (Temminck and Schlegel):

Jap-II Effects on nutrient digestibility and digestive enzyme activity.

Dong, Z.J & Yuan, X.H (2002) The utilizations of heterosis in

Douglas, S.E., Gallant, J.W & Bullerwell, C.E (1999) Molecular investigation of aminopeptidase N expression in the winter

.

Fahienl, A., Mow, S.L & Kmiotek, E (1977) Effect of aspartate

on complexes between glutamate dehydrogenase and various

Farhangi, M., Carter, C.G & Hardy, R.W (2001) Growth,

(Oncorhynchus mykiss) to different dietary inclusion levels of

340.

Fountoulaki, E., Alexis, M.N., Nengas, I & Venou, B (2005)

Effect of diet composition on nutrient digestibility and digestive

enzyme levels of gilthead sea bream (Sparus aurata L.) Aquacult.

Furne´, M., Hidalgo, M.C., Lopez, A., Garcı-Gallego, M., Morales,

A.E., Domezain, A., Domezaine, J & Sanz, A (2005) Digestive

enzyme activities in Adriatic sturgeon Acipenser naccarii and

rainbow trout Oncorhynchus mykiss A comparative study

Garcı´a-Gasca, A., Galaviz, M.A., Gutie´rrez, J.N &

Garcı´a-Orte-ga, A (2006) Development of the digestive tract, trypsin activity

and gene expression in eggs and larvae of the bullseye puffer fish

Go, V.L.W., Hoffman, A.F & Summerskill, W.H.J (1970)

Pancre-ozymin bioassay in man based on pancreatic enzyme secretion:

potency of specific amino acids and other digestive products.

He, W., Feng, L., Jiang, J., Liu, Y & Zhou, X.-Q (2009) Dietary

pyridoxine requirement of juvenile Jian carp (Cyprinus carpio

Hooper, L.V., Stappenbeck, T.S., Hong, C.V & Gordon, J.I.

(2003) Angiogenins: a new class of microbicidal proteins

Huang, H.-H., Feng, L., Liu, Y., Jiang, J., Jiang, W.-D., Hu, K.,

Li, S.-H & Zhou, X.-Q (2011) Effects of dietary thiamin

sup-plement on growth, body composition and intestinal enzyme

activities of juvenile Jian carp (Cyprinus carpio var Jian)

Hummel, B.C.W (1959) A modified spectrophotometric

determina-tion of chymotrypsin, trypsin, and thrombin Can J Physiol.

Jiang, W.-D., Feng, L., Liu, Y., Jiang, J & Zhou, X.-Q (2009)

Growth, digestive capacity and intestinal microflora of juvenile

Jian carp (Cyprinus carpio var Jian) fed graded levels of dietary

Kjørsvik, E., Meeren, T van der, Kryvi, H., Arnfinnson, J &

Kvenseth, P.G (1991) Early development of the digestive tract

of cod larvae, Gadus morhua L., during and starvation J Fish

Klein, S., Cohn, S.M., Alpers, D.H (1998) The alimentary tract in

nutrition In: Modern Nutrition in Health and Disease (Shils,

MD.

Li, P., Burr, G.S., Wen, Q., Goff, J.B., Murthy, H.S & Gatlin, D.

M III (2009) Dietary sufficiency of sulfur amino acid

com-pounds influences plasma ascorbic acid concentrations and liver

peroxidation of juvenile hybrid striped bass (Morone

Li, W., Zhou, X.-Q., Feng, L., Liu, Y & Jiang, J (2010) Effect of

dietary riboflavin on growth, feed utilization, body composition

and intestinal enzyme activities of juvenile Jian carp (Cyprinus

Lim, C.B., Chew, S.F., Anderson, P.M & IP, Y.K (2001)

Reduc-tion in the rates of protein and amino acid catabolism to slow

down the accumulation of endogenousammonia: a strategy

potentially adopted by mudskippers (periophthalmodon seri and boleophthalmus boddaerti) during aerial exposure in

Lin, Y & Zhou, X.-Q (2006) Dietary glutamine supplementation improves structure and function of intestine of juvenile Jian carp

Ling, J., Feng, L., Liu, Y., Jiang, J., Jiang, W.-D., Hu, K., Li, S.-H & Zhou, X.-Q (2010) Effect of dietary iron levels on growth, body composition and intestinal enzyme activities of juvenile Jian carp (Cyprinus carpio var Jian) Aquacult Nutr.,

16, 616–624.

Lingrel, J.B (2010) The physiological significance of the tonic steroid/ouabain-binding site of the Na,K-ATPase Annu.

Liu, Y., Feng, L., Jiang, J., Liu, Y & Zhou, X.-Q (2009) Effects

of dietary protein levels on the growth performance, digestive capacity and amino acid metabolism of juvenile Jian carp (Cyp-

Merino, S., Rubires, X., Knøchel, S & Tomas, J.M (1995)

Murı´n, R., Mohammadi, G., Leibfritz, D & Hamprecht, B (2009)

Nose, T (1979) Summary report on the requirements of essential amino acids for carp In: Finfish Nutrition and Fishfeed Tech-

Berlin.

NRC (National Research Council) (1993) Nutrient Requirements

of Fish National Academy Press, Washington, DC, USA Pedersen, H.E & Sissons, J.W (1984) Effect of antigenic soyabean protein on the physiology and morphology of the gut in the

Ringø, E & Birkbeck, T.H (1999) Intestinal microflora of fish

Rio-Rodriguez, R.E.D., Inglis, V & Millar, S.D (1997) Survival

Rollin, X., Wauters, J.B., Bodin, N., Larondelle, Y., Ooghe, W., Wathelet, B & Abboudi, T (2006) Maintenance threonine requirement and efficiency of its use for accretion of whole-body threonine and protein in Atlantic salmon (Salmo salar L.) fry.

Rosalki, S.B., Rau, D., Lehmann, D & Prentice, M (1970) Gamma-glutamyl transpeptidase in chronic alcoholism Lancet,

2, 1139.

Salinas, I., Myklebust, R., Esteban, M.A., Olsen, R.E., Meseguer,

J & Ringø, E (2008) In vitro studies of Lactobacillus delbrueckii subsp lactis in Atlantic salmon (Salmo salar L.) foregut: tissue responses and evidence of protection against Aeromonas salmoni-

Santiago, C.B & Lovell, R.T (1988) Amino acid requirements for

Segner, H & Verreth, J (1995) Metabolic enzyme activities in vae of the African catfish, Clarias gariepinus: changes in relation

Shiau, S.Y & Lo, P.S (2000) Dietary choline requirements of

Sidransky, H & Baba, T (1960) Chemical pathology of acute

amino acid deficiencies III Morphologic and biochemical

changes in young rats fed valine- or lysine-devoid diets J Nutr.,

70, 463–483.

Stappenbeck, T.S., Hooper, L.V & Gordon, J.I (2002)

Develop-mental regulation of intestinal angiogenesis by indigenous

Sugita, H., Kawasaki, J., Kumazawa, J & Deguchi, Y (1996)

Pro-duction of amylase by the intestinal bacteria of Japanese coastal

Sun, X.Y., Zhang, J.S., Shi, Y., Wang, J & Gong, Y (1995)

Stud-ies on the genetic characteristic of Jian carp (Cyprinus carpio

Sveier, H., Raae, A.J & Lied, E (2000) Growth and protein

turn-over in Atlantic salmo (Salmo salar L.): the effect of dietary

Szyman´ska, K., Rydzewska, A & Domka, F (2002) Catabolic

activity of desulfotomaculum rumini bacteria in media containing

Takada, N., Ogita, K., Taguchi, T., Masumoto, K & Suita, S.

(2005) Effect of a valine-rich diet on a rat model of short bowel

Tan, L.-N., Feng, L., Liu, Y., Jiang, J., Jiang, W.-D., Hu, K., Li,

S.-H & Zhou, X.-Q (2011) Growth, body composition and

intestinal enzyme activities of juvenile Jian carp (Cyprinus carpio

Tang, L., Wang, G.-H., Jiang, J., Feng, L., Liu, Y., Li, S.-H.,

Kuang, S.-Y & Zhou, X.-Q (2009) Effect of methionine on

intestinal enzymes activities, microflora and humoral immune of

Tantikitti, C & Chimsung, N (2001) Dietary lysine requirement of

freshwater catfish (Mystus nemurus Cuv & Val.) Aquac Res.,

32, 135–141.

Tanzer, M.L & Gilvarg, C (1959) Creatine and creatine kinase

Villanueva, J., Vanacore, R., Goicoehea, O & Amethauer, R.

(1997) Intestinal alkaline phosphatase of the fish Cyprinus carpio:

regional distribution and membrane association J Exp Zool.

Wallimann, T., Dolder, M., Schlattner, U., Eder, M., mann, T., O’Gorman, E., Ru¨ck, A & Brdiczka, D (1998) Some new aspects of creatine kinase (CK): compartmentation, structure, function and regulation for cellular and mitochon-

234.

Wen, Z.-P., Zhou, X.-Q., Feng, L., Jiang, J & Liu, Y (2009) Effect of dietary pantothenic acid supplement on growth, body composition and intestinal enzyme activities of juvenile

476.

Weng, C.F., Chiang, C.C., Gong, H.Y., Chen, M.H.C., Lin, C.

J.F., Huang, W.T., Cheng, C.Y., Hwang, P.P & Wu, J.L.

-ATPase and creatine kinase in response to salinity changes in the euryhaline teleost,

Wilson, R.P (2002) Amino acids and protein In: Fish Nutrition,

Aca-demic Press, San Diego, CA.

Wood, C.M (1993) Ammonia and urea metabolism and excretion.

Press, Boca Raton, FL, USA.

Wood, C.M (2001) Influence of feeding, exercise, and

Xiao, W.-W., Feng, L., Liu, Y., Jiang, J., Hu, K., Jiang, W.-D.,

Li, S.-H & Zhou, X.-Q (2011) Effects of dietary methionine hydroxy analogue supplement on growth, protein deposition and intestinal enzymes activities of juvenile Jian carp (Cyprinus car-

Yoshizawa, F (2004) Regulation of protein synthesis by

Young, V.R (1970) The role of skeletal and cardiac muscle in the regulation of protein metabolism In: Mammalian Protein

Aca-demic Press, New York, USA.

Zeitoun, I.H., Ullrey, D.E., Magee, W.T., Gill, J.L & Bergen, W.

G (1976) Quantifying nutrient requirements of fish J Fish Res.,

1 1 1 2 2

1

Growth performance, muscle cellularity, flesh quality, and

plasma ghrelin were investigated in 0+ and 1+ farmed

Atlantic salmon (Salmo salar L.) from 40 g to 4.3 kg

Reduced meal frequency was introduced in both smolt

groups from ~1.5 kg; one meal per second day (<5 °C) to

one daily meal (>5 °C), while control groups were fed one

to three daily meals Results show that 0+ salmon had

higher final fibre number and density, pigment content, red

colour intensity, firmer flesh, and lower fillet fat content

than 1+ salmon at 4.3 kg, affected by season and smolt

type Muscle fibre recruitment was an important

determi-nant of fillet firmness and colour, possibly influenced by

the prenatal temperature regime Fish fed reduced meal

fre-quency showed temporal reduced feeding ration, but

growth performance was not compromised in any smolt

groups at harvest However, fillet fat, gaping, and colour

decreased by less frequent feeding, with permanent effects

in 1+ salmon for gaping and fat Reduced meal frequency

is therefore considered to be a promising tool for managing

important flesh quality attributes in salmon without

com-promising growth performance It is also suggested that

ghrelin stimulates short-term appetite, and perhaps also in

the longer term

performance, muscle cellularity, seasonal variation, smolt

type

Received 11 April 2011; accepted 9 December 2011

Correspondence: C.A Johnsen, Faculty of Biosciences and Aquaculture,

University of Nordland, N-8049 Bodø, Norway E-mail: chris.andre.

johnsen@uin.no

In modern, high-intensity Atlantic salmon (Salmo salar L.)farming breeding programs, high-energy feeds, large seacages, light manipulation, and out-of-season production ofunderyearling (0+) smolts have primarily been developed tomaximize production volume However, over the last dec-ade, the salmon industry has become more focused onqualitative goals due to several factors such as environmen-tal impact of feed production, sustainability of marine fish-eries, alternative feed ingredients, nutritional requirements,and unforeseen flesh quality problems The production of0+ smolts, transferred to sea during autumn, 8–9 monthsafter hatching, has become a commercially important strat-egy for spreading the availability of market-sized salmonthroughout the year (Duncan et al 1998) Increased incu-bation temperature and altered photoperiod prior to smol-tification is a commonly used method in the production of0+ smolts (Duncan et al 1998, 2002; Lysfjord et al 2004).Traditional 1+ smolts follow the ‘in-season’ growth pattern,and are usually transferred to sea during the following

the large seasonal changes in temperature and light areboth being well documented to affect growth patterns in 0+and 1+ salmon (Mørkøre & Rørvik 2001; Nordgarden

characteris-tics, which are important for the consumer perception ofthe final product are also strongly affected by season(Shearer et al 1994; Mørkøre & Rørvik 2001; Nordgarden

age, i.e smolt type, have been ambiguous, and severalstudies have concluded that the smolt type has no impact

on growth, muscle cellularity or flesh quality (Duncan et al.1998; Mørkøre & Rørvik 2001; Roth et al 2005; Vieira

.

Aquaculture Nutrition

et al 2005) However, Ytrestøyl et al (2004) has showed

that different smolt production strategies lead to higher

carotenoid content in 0+ salmon Possibly, an altered

mus-cle structure could explain the differences in colouration

and mechanical properties between the two smolt types

(Ytrestøyl et al 2004) However, muscle cellularity during

the seawater stage is currently unknown for 0+ and 1+

salmon reared at northern latitudes

Firm texture is a valued sensory characteristic for

con-sumers (Hillestad et al 1998; Einen et al 1999) and an

important attribute for the processing industry (Johnston

2001a), and farmed salmon tend to have a softer flesh than

their wild counterparts (Johnston et al 2006) Together

with gaping, occurrence of too soft flesh have been

reported by the salmon industry during spring and summer

in the southwest of Norway (Mørkøre & Rørvik 2001),

and during summer and autumn in the northern Norway

Feeding regime, diet composition and environment can

influence both muscle cellularity and flesh quality in salmon

(Johnston 1999), and muscle cellularity is an important

determinant for flesh texture and colour in fish (Dunajski

1979; Hatae et al 1990; Hurling et al 1996; Johnston et al

2000a) Reduced feed rations can improve product quality

in salmon by giving lower fillet fat content, less gaping and

firmer texture, and increased carcass-yield (Einen et al

1999) However, this may also decrease growth rate, and

hence economical liquidity A recent study suggests that

starvation for five weeks prior to harvest improves the

resistance of Atlantic salmon to acute stress, and thus is

positive for several flesh quality attributes (Mørkøre et al

2008) This is also shown for Atlantic cod (Gadus morhua),

were fasting increased fillet texture with as much as twofold

over a four week period (Hagen & Solberg 2010) In

com-parison, long-term starvation (86 days) of salmon was

sug-gested to be a rather weak tool to change flesh quality

(Einen & Thomassen 1998), and maybe not suitable for

commercial use due to substantial loss of body weight

(Einen et al 1998) Feed utilisation is known to improve

with reduced feed ration (Jobling 1994), thus lowering

pro-duction costs Currently, there are no available commercial

feeding regimes suitable for manipulating flesh quality

during season or across smolt types without jeopardizing

growth rate A pilot study on 1+ salmon has shown

prom-ising results from reduced meal frequency with regard to

muscle structure, flesh quality and economy, without

com-promising growth performance (Johnsen 2006)

The hormone ghrelin has recently been identified in

rain-bow trout (Oncorhynchus mykiss) (Kaiya et al 2003),

Atlantic salmon (Murashita et al 2009), and Arctic charr

(Salvelinus alpinus) (Frøiland et al 2010) The role of lin in fish is suggested being of a multifunctional characterand species-specific, such as being related to food intake,growth hormone (GH) release, and metabolism, e.g

ghre-growth, adiposity and reproduction (Kaiya et al 2008) Atleast in some fish species and/or physiological states, ghre-lin function in fish appears different to that in mammals(Jo¨nsson et al 2007, 2010; Murashita et al 2009; Frøiland

be an appetite suppressing (anorexigenic) hormone, with nodirect link to fat deposits (Jo¨nsson et al 2010), whereas inAtlantic salmon, an appetite stimulating (orexigenic) rolehas been suggested (Murashita et al 2009) In Arctic charr,

a role in long-term regulation of energy homeostasis andsexual maturation are suggested (Frøiland et al 2010)

Ghrelin might therefore be an indicator for certain fleshquality characteristics in salmonids, e.g fillet fat, textureand/or gaping, but this needs further investigation

Two experiments were undertaken to examine growth,feed utilisation, muscle cellularity, flesh quality, and plasmaghrelin levels in underyearling (0+) and yearling (1+) Atlan-tic salmon smolts throughout the subsequent seawatergrow-out period For improved management practices, theeffects of reduced meal frequency were investigated in bothsmolt types A feeding practice applicable for commercialuse was prioritised in the alternative feeding regime Thefunction of ghrelin in salmon, and the contribution of mus-cle cellularity to fillet firmness, gaping and colour were alsoinvestigated

Atlantic salmon used in the present experiment were ferred to seawater as underyearling (0+, experiment 1) andyearling (1+, experiment 2) smolt, both produced from thesame egg batch (SalmoBreed stock, Salten Havbruk AS,67°N, Norway) The 0+ group was incubated at 8 °C andhatched in January 2007 (week 2), while the 1+ group wasincubated at 2 °C and hatched in April 2007 (week 14)

trans-Subsequent transfer from the hatchery to the smolt tion site (at 10 g), pre-smolt groups were held indoor in

) supplied with a mixture of water and seawater All fish were size-graded twice, fedcommercial feeds from BioMar AS (Bio-Optimal Start,CPK, CPK Svev, CPK+) and Skretting AS (Nutra ST,Olympic, Parr LB, Response LB), and vaccinated with

.

The parr projected to give a 0+ smolt were reared in water

of 10.8 °C (range 8.0–17.0 °C) All 0+ fish were exposed to

a 24 : 0 LD (hours light : hours dark) light regime, and

to the tanks The parr projected to give a 1+ smolt were

reared in water with average salinity of 8& (range 4–10&)

and temperature of 4.2 °C (range 0.9–16.0 °C) Increased

salinity (>10&) and photoperiod manipulation (24 : 0 LD

8 weeks, and 24 : 0 LD during the final period) were used

to advance the smoltification process Both smolt types

were tested for hypo-osmoregulatory ability (n = 50/smolt

previously described method (Lysfjord et al 2004)

The 0+ and 1+ smolt were transferred to Gildeska˚l Research

Ol-dervika and Stivika (1.5 km between) on 20 October 2007

(54 g) and 30 July 2008 (46 g), respectively The fish were

randomly distributed in four sea cages per smolt group

smolt types were divided into two sub groups; one control

group (CG) and one treatment group (TG), each with four

were acclimatised for minimum one month prior to

introduc-tion of reduced meal frequency at a weight of 1668 ± 107 g

(408 days) and 1325 ± 99 g (344 days) for the 0+ and 1+

group, respectively The control groups followed a

commer-cial feeding regime, while treatment groups were fed with

reduced meal frequency from 1 December 2008 to 1 July

2009 (0+) and from 9 July 2009 to 10 December 2009 (1+)

Meal frequency was adjusted to water temperatures

accord-ing to the followaccord-ing procedure; one daily meal (CG) or one

meal per second day (TG) were provided at low temperature

(<5 °C), and 2–3 daily meals (CG) or one daily meal (TG) at

temperatures higher than 5 °C Both smolt groups were fed

commercial high-energy feeds from BioMar AS (Myre,

Nor-way) during the experiment (50–4300 g) Feeds were

pro-vided in meals by automatic feeders (Betten T1A, Betten

Maskinstasjon AS, Norway; volume 1000 L) until apparent

satiation Pellet size and dietary composition changed in

accordance with commercial practice (dietary protein; 470 –

the fish grew larger, resulting in a concomitant increase in

fish were 2.72 kg and 2.78 kg, respectively

Ambient conditions were similar between the rearingsites at Oldervika (0+ fish) and Stivika (1+ fish) for thesame time period (Fig 1) However, light and temperatureprofile differed substantially between the smolt types ofsimilar size due to the difference in smolt production proto-col and transfer dates The 0+ fish were exposed to artificiallight (24 h, 400 W) from seawater transfer until 13 May

2008, followed by natural photoperiod for 409 days untiltermination (Fig 1) The 1+ fish were exposed to naturallight during the first 141 days after seawater transfer, andartificial light (24 h, 400 W) was added from 18 December

to 15 May 2009 followed by natural light until harvest.Average water temperatures (daily registered at 1, 3 and

5 m) during experiments were 7.3 °C (range 2.6–13.3 °C)and 8.6 °C (range 2.8–13.3 °C) for 0+ and 1+ fish, respec-tively (Fig 1)

Feed provision was registered daily, and individual fishweight and biomass were estimated in all cages using theStorvik biomass estimation system (Storvik AS, Sunn-dalsøra, Norway) at predefined dates (0+; 1 December 2008,

Figure 1 Seawater temperature (solid line; —) and light regime for the 0+ group (short dashes; - - -) and the 1+ group (dotted; ····) from seawater transfer (ST) to termination of the experiment (End) The 0+ salmon were transferred to seawater 20 October

2007, exposed to artificial light until 13 May 2008, followed by natural photoperiod until termination; 1 July 2009 The 1+ salmon were transferred to seawater 30 July 2008, with artificial light added from 18 December 2008 to 15 May 2009, followed by natu- ral photoperiod until termination; 10 December 2009.

28 March 2009, 20 April 2009, 21 May 2009, and 1 July

2009, 1+; 9 July 2009, 30 August 2009, 14 September 2009,

16 October 2009, 17 November 2009, and 10 December

2009) throughout the experiment Between 916 and 1439 fish

the data from biomass estimations were used for

calcula-tion of growth and feed/gain ratio Thermal growth

,

mortality was low during the treatment period with no

significantly higher for 0+ salmon compared to 0+

(Table 1)

Salmon from both smolt groups were sampled at ~40, 100,

190, 500, 1100, 2000, 3000 and 4300 g during the experiments

for evaluation of flesh quality characteristics, plasma ghrelin

levels, and muscle cellularity (muscle cellularity not analysed

at 3000 g) Initial sampling of 0+ (33& salinity) and 1+

and 23 July 2008, respectively (one tank per smolt group;

December 2007 (0+) and 8 October 2008 (1+) from rearing

reduced meal frequency, sampling at 190, 500 and 1100 g

out 11 March 2008, 21 August 2008 and 27 October 2008 for

the 0+ group, and 28 October 2008, 8 December 2008 and 29

May 2009 for the 1+ group, respectively Last three sampling

fre-quency, were carried out 9 January 2009, 26 March 2009, and

1 July 2009 for the 0+ group, and 28 August 2009, 9 October

2009, and 10 December 2009 for the 1+ group, respectively

In the final sampling, 40 similar sized fish (4283 ± 351 g;

mean ± sd) from each of the four experimental groups, i.e 0+

CG, 0+ TG, 1+ CG, 1+ TG, were harvested No maturationwas observed in either the 0+ or 1+ group Harvested salmonwere netted at random near the surface during the morningmeal Fish that deviated substantially from the cage meanweight (i.e outliers), as indicated by the Storvik measure-ments, were avoided during sampling All fish were killedwith a sharp blow to the head, and blood was immediately

was obtained through centrifugation at 3000 g at 4 °C for

5 min Approximately 1 mL was collected for later ghrelinanalysis and immediately acidified with HCl to a final concen-tration of 0.1 M (Jo¨nsson et al 2007) Plasma samples were

water for minimum 30 min before being transported to versity of Nordland (UiN) Upon arrival round body weight

] Gutted fish were stored on ice inpolystyrene boxes for four days prior to analyses of fleshquality characteristics (e.g fillet firmness, gaping, colour,water, fat, protein) at UiN Samples taken for SalmoFan

for 14 days prior to analysis at AquaLab AS (Eidsva˚g,Norway)

Table 1 Fish performance and amount of feed provided in control groups (CGs) and treatment groups (TGs; reduced meal frequency) for

0+ or 1+ salmon over the treatment period

Round fish weight at start (Wtstart) and end (Wtend), and thermal growth coefficient (TGC3) Values represent mean ± SD (n = 4 per

group), and different superscript letters within a row indicate significant differences between groups (ANOVA, P < 0.05).

.

Principles are adapted and modified according to a

previ-ously described method (Johnston et al 1999, 2000b;

Hagen et al 2006) Briefly, a 0.5 cm thick steak was

prepared anterior to the first dorsal fin ray and the left

Total cross-sectional area of the fast muscle was calculated

using Sigma ScanPro 5.0 software (SPSS Inc., USA) from

a calibrated digital image of the steak Blocks were

mounted on cork sheets using cryomatrix (Shandon

Cryomatrix, Thermo Electron Corporation, UK), frozen in

2-methylbutane cooled to near its freezing point ( 159 °C)

stained with Harris haematoxylin (Sigma–Aldrich), and

of 1000 muscle fibres per fish (equally distributed between

blocks) were digitised using a light microscope equipped

with a digital camera (Axioskop 2 mot plus, Carl Zeiss,

Germany) and image analysing software (Axio Vision Rel

4.2, Carl Zeiss) Fibre area and diameter were computed by

the software, while the total fibre number was calculated as

The 2.5th and 97.5th percentile of the fast fibrediameter size-distribution were used as an estimate of

the minimum and maximum muscle fibre diameter in

salmon, respectively Active muscle fibre recruitment was

Fillet firmness was determined on post-rigor muscle

sam-ples according to Johnston et al (2004), using a

TA-XTPLUS texture analyzer with Texture Expert Exceed 2.52

software (Stable Micro Systems, Surrey, UK) The

instru-ment was equipped with a load cell of 50 kg, a 60° knife

edge steel blade, and a slotted insert platform for muscle

measurements were performed in duplicate for each fish

the epaxial myotomes in the left NQC (3000 and 4300 g),

or as measurements directly on the fillet from the rightNQC (500, 1100 and 2000 g, Fig 2) The area under thecurve during shearing was calculated as the total workdone (mJ), either for cutting through the muscle blocks orcutting 90% into the fillets Gaping over the total length ofthe left fillet (post-rigor) was scored semi-quantitativelyaccording to a previously described method (Johnsen et al.2011) Gaping was evaluated for fish sampled from 500 to

4300 g in all experimental groups

Flesh colour was measured in triplicate on post-rigor filletsfrom the left Norwegian Quality Cut (NQC, Fig 2) fromfish sampled between 100 and 4300 g using a MinoltaChromameter 300 (Minolta Camera Co Ltd, Japan) in theCIE L*a*b* measuring mode (NS-9402E 1994) In addi-

lineal (DSM, Switzerland) under standardised conditions in

a light cabinet (Ra >90, colour temperature >5000 K)according to Norwegian standards (NS-9402E 1994) Sal-moFan evaluation was carried out on the right NQC fillet(Fig 2) from fish sampled at 4300 g Total carotenoid con-centrations were determined on the same fillet cuts andfrom the same sampling point used for the SalmoFan eval-uation (Fig 2) Duplicate samples (10 g) were homogenisedwith 20 g sodium sulphate, transferred to a glass containercontaining 50 mL ethyl acetate and mixed for 60 min (NS-9402E 1994) The extract was filtered and added to aquartz cuvette, and the optical density was measured at

472 nm using a spectrophotometer (Shimadzu, UV 1700),with ethyl acetate used as the blank

Chemical composition of post-rigor fast muscle from theleft NQC fillet (Fig 2) was determined in duplicate sam-

(NS-9401E 1994) Fat percentage was determined

(NS-9402E 1994), water as weight loss after drying at

104 °C for 20 h and protein was analysed as

Foss Tecator AB, Sweden) Salmon up to ~190 g were onlyanalysed for water content Near infrared transmissionanalysis (NIT, Infratec 1225 Food and Feed Analyzer;Foss Tecator AB, Sweden) was performed on homogenisedmuscle samples for fish larger than 190 g (n = 200 per

smolt group) Chemical analyses of fat, water and protein

were performed on a representative selection of 80 fish per

smolt group (equally distributed between sampling dates)

and further used for calibration of NIT data Prediction of

individual percentage fat, water and protein content of all

400 samples were performed using partial least squares

(PLS) regression as described by previous studies (Solberg

1992, 1997)

Plasma levels of ghrelin were analysed with the

radioimmu-noassay (RIA) previously established for rat (Hosoda et al

2000), slightly modified for rainbow trout by Jo¨nsson et al

as label The RIA was validated for Atlantic salmon with a

test of parallelism showing that the slopes of the ghrelin

standard curve and Atlantic salmon plasma were similar

(P = 0.4) The hormone analysis was carried out on five

exper-imental group) at 190, 500, 1100, 2000, 3000 and 4300 g

sampling points

Statistica v 9.1 (StatSoft Inc., Tulsa, OK, USA) was used

regression analysis Assumptions of statistical methodswere tested prior to analysis, and in case of non-homoge-

Wilk’s W test and visual assessment of histograms) data

test) Differences between smolt groups (0+ vs 1+) andbetween feeding regimes (CG vs TG) with regard togrowth and feed utilisation were analysed by one-way

den-sity, number, 2.5th and 97.5th percentile) and flesh quality

with muscle fibre and flesh characteristics as dependentvariables, smolt type as categorical variable and season(week no.) as covariate Distribution of muscle fibre diame-ter was evaluated using smooth nonparametric distributionswhere 1000 measurements of fast fibre diameter were fittedusing a kernel function (Johnston et al 1999) Experimen-tal groups (0+ CG, 0+ TG, 1+ CG, 1+ TG) that werecompared had a similar body mass and length (n = 20 each

to test the null hypothesis that the probability densityfunctions (PDFs) of groups were equal over all diameters

To supplement this test, density curves for each treatmentwere compared graphically by constructing a variabilityband around the density estimate for the combined popula-tions using the mean smoothing parameter h, varying

Figure 2 Sampling sites for muscle fibres (anterior to the first dorsal fin ray) and flesh quality analyses (Norwegian Quality Cut, NQC)

performed at regular intervals during the experiment (~40, 100, 190, 500, 1100, 2000, 3000 and 4300 g).

.

between 0.076 and 0.151 for the different groups (Bowman

& Azzalini 1997) Density curves were used to distinguish

underlying structure in the distributions from random

vari-ation providing an indicator of which part(s) of the

differences Data represent mean ± SD (n = number of

samples) Significance was accepted at P-value of <0.05

Average body weight of underyearling (0+) salmon

incre-ased with 4136 g (620 days; 20.7 months) during the

2.72 In the initial phase, from seawater transfer to

intro-duction of reduced meal frequency (408 days), the fish

weight increased from 54 g to 1668 ± 107 g, respectively,

fol-lowing treatment period (212 days), body weight increased

feed, with no differences between feeding regimes (Table 1;

Fig 3a,c) Feed/gain was calculated to 1.00 ± 0.05 at final

sampling, and was neither affected by reduced meal

fre-quency nor season (Table 1 and Fig 3d) However, the

amount of feed provided in the treatment group wasreduced during certain periods of the winter (early Decem-ber to late January, Fig 3b)

Body weight of yearling (1+) salmon increased with

4318 g (498 days; 16.6 months) during the experiment,

(344 days), weight increased from 46 g to 1325 ± 99 g, with

period (154 days), weight increased to 4364 ± 176 g, with

no differences between feeding regimes (Table 1 andFig 3a) The group provided reduced meal frequency had

(P = 0.022, Fig 3c) than the control group, as a quence of significant less feed provided during weeks 33and 36 (Fig 3b) However, growth in the treatment groupwas improved towards harvest (10 December), and was notsignificantly different from the control (Table 1) Totalamount of feed provided and feed/gain during the sameperiod were 549 558 kg and 1.08 ± 0.04, respectively, with

conse-no difference between feeding regimes for the 1+ salmon(Table 1)

The accelerated 0+ group smoltified ~7 months posthatching, which was half the time of the 1+ production(~14 months) However, the 1+ salmon had higher growthrate during the seawater phase, but did not manage to

per-iod with reduced meal frequency Growth and feed utilisation were estimated using a biomass estimation system.

catch-up with the 0+ group At harvest, the growth

advan-tage of the 0+ salmon was reduced to ~4 months

Number of fast muscle fibres per trunk cross-section was

similar between smolt groups at start (40 g, ~162 000

fibres, Fig 4a), but the number of fibres recruited per

day from hatching were almost twice as high in the 0+

compared to the 1+ group During the first growth period

was highest in the 0+ fish (144 000) compared to the

1+ fish (110 000), corresponding to ~1500 and 1400

sam-pling, the total fibre number was highest in the 1+ fish

(446 000 and 661 000) compared to 0+ fish (373 000 and

winter-sum-mer, respectively At 1100 g, fibre number had increased to

886 000 (0+) and 796 000 (1+), corresponding to 3900 and

respec-tively The recruitment of fast fibres towards 2000 and

At the final sampling, the total muscle fibre number was

significantly higher in the 0+ group (1 039 000) compared

to the 1+ group (914 000) However, the average number

of muscle fibres recruited per day during the seawater

per-iod was highest in the 1+ group (1500) compared to the 0+

group (1300) when data were corrected for time differences,

a curvilinear correlation between fibre number and body

mass of individual fish; i.e fibre number = 138 147 +

With the exception of some fluctuations in fast muscle

diameter increased gradually throughout the experiment

respec-tively (Fig 4b) At the end of the experiment, the average

fibre diameter in the 1+ group was about 8% larger than in

the 0+ group Change in fast muscle fibre diameter among

smolt groups was concurrent with the change in fibre density

until 190 g, when they were stabilising (Table 2)

Subse-quently, fibre densities decreased towards the final sampling,

where the highest level was found in the 0+ group

The 2.5th percentile was lowest in the 0+ fish at 100,

1100, 2000 and 4300 g, reflecting a higher proportion of

the smallest muscle fibres compared to the 1+ group(Tables 2 & 3) On the other hand, the 97.5th percentile(maximum diameter) of the 0+ group was highest in the 40and 190 g sampling point, but no difference was detected

97.5th percentile increased during the experiment indicatingthat hypertrophic growth was dominant Presence of mus-

recruited fast fibres in salmon, which presently ceased at

~2 kg in the 1+ group and somewhere between 2 and 4 kg

in the 0+ group (Tables 2 & 3) At 40 g, the percentage ofnew recruited muscle fibres was ~3.5%, decreasing to 1.0%

(0+) and 0.2% (1+) in the 2000 g sampling point, and 0%

at 4300 g Reduced meal frequency had no effect on either

97.5th percentile in the 0+ or the 1+ group (Table 3)

Size distributions of fast muscle fibres in the 0+ and 1+

the 190 g sampling, significant differences between smoltgroups were present in all sampling points At the 40 gsampling (Fig 5a), the main difference in muscle fibre dis-tribution among smolt types was between ~40–70 lm,being lowest in the 0+ group After seawater transfer(100 g, Fig 5b) the 0+ group responded by increasing thefast fibre recruitment compared to the 1+ group At 500 g

most abundant in the 1+ group, but shifted towards

1100 g, being higher for the 0+ group (Fig 5e) During the

~1.3 kg (1+) Significant differences between smolt groups (0+ CG

vs 1+ CG) are indicated with asterisks; *0.01 < P < 0.05 or

***P < 0.001.

.

last two sampling points (2000 and 4300 g, Fig 5f,g) the

muscle fibre distribution of 0+ had a bimodal distribution,

reflecting the higher number of small fibres (15–40 lm)

compared to the 1+ This supports the previous results on

recruitment rates, 2.5th percentiles, and overall fibre

num-bers No significant influence of reduced meal frequency on

fibre size distribution in the 0+ or 1+ group was found at

Fillet firmness, i.e shear work, increased for both the 0+

and 1+ group between the 500 and 3000 g sampling, but

was constant towards the final sampling (Fig 7) At the

3000 and 4300 g sampling points, fillets were firmest in the0+ group (P = 0.033 and 0.015, respectively) and not influ-enced by body weight when analysis was performed on

n = 160) when measured directly on the fillet No weightdifferences were detected between smolt groups at any timeduring the experiments, but the 0+ group had significantlysofter flesh at 500 g (P < 0.001) and 1100 g (P < 0.001,Fig 7) compared to the 1+ group At 2000 g, fillet firmnesswas similar between smolt groups (P = 0.440), but the 0+treatment group had softer flesh than the 0+ control group

(P < 0.001), probably due to lower k-factor and thinner

fillets (Tables 2 & 3) However, reduced meal frequency did

not influence fillet firmness in either the 0+ or the 1+ group

at 3000 or 4300 g Fillet firmness was positively correlated

to fibre number and density after weight corrections,

explaining 10.4% and 16.3% of the total variation

respec-tively (P < 0.001, n = 240)

Both smolt groups had low (<0.5) gaping scores during

the experiments (<0.5, Tables 2 & 3) Gaping scores

indicat-ing that larger and/or faster growindicat-ing fish were more

proned to gaping The 0+ group had significantly less

gap-ing than the 1+ group prior to 4300 g, but no differences

between smolt groups were seen at the final sampling

(P = 0.071) However, the 1+ group provided reduced meal

frequency had a lower gaping score at 4300 g compared to

the control (P < 0.001) No such effect was found in the 0+

treatment group, e.g reduced meal frequency Muscle

cellu-larity did not affect gaping in the present study

The a*-value (intensity of red colour) correlated positively

to the fibre number and fibre density after weight

Figure 6 Distribution of fast muscle fibre diameter in control groups; CG (solid line; —) and treatment groups; TG (long dashes; – –) for

0+ salmon at (a) 2000 g, and (b) 4300 g, and for 1+ salmon at (c) 2000 g and (d) 4300 g The dotted line (····) represents the average

probability density function (PDF) of the combined population, and the gray shaded area represents 100 bootstrap samples of the PDF.

Statistical differences are indicated where the solid and dotted lines fall outside the reference band (shaded area).

Figure 7 Change in fillet firmness, i.e shear work, for each group during the experiment Measurements were performed directly on the right NQC fillet (500, 1100 and 2000 g) or in standardised muscle blocks from the left NQC fillet (3000 and 4300 g) Sam- pling dates are indicated with italic text (0+) or bold text (1+).

Control (CG) and treatment groups (TG; reduced meal frequency) for each smolt type were created from ~1.7 (0+) and ~1.3 kg (1+), and significant differences between feeding regimes (CG versus

smolt groups (0+ CG vs 1+ CG) are indicated with asterisks;

*0.01 < P < 0.05 or ***P < 0.001.

.

tions, explaining 13.5% and 22.6% of the total variation

respectively (linear, P < 0.001, n = 320) The a*-value

increased in fish from both smolt groups during the

experi-ment (Tables 2 & 3), but dropped in the 1+ group during

and during the second fall in the 0+ group

with the substantial shift in environmental conditions from

summer to autumn when daylight and temperature were

reduced At the initial sampling (100 g) the a*-value was

highest in the 1+ group, but changed towards 500 g being

highest in the 0+ group, a result that was maintained until

harvest (6.2 vs 5.8, P = 0.005) a*-value and total pigment

0.01, n = 160), reflecting a highest concentration in the 0+

deter-mined by the SalmoFan did not differ between the smolt

groups (Table 3), indicating reduced sensitivity compared

to the Minolta measurements Reduced meal frequency had

no impact on either flesh colour or pigment concentration

at the final sampling (4300 g, Table 3)

In general, the b*-value (intensity of yellow colour) was

highest in the 0+ group in most sampling points until

har-vest (4300 g, 15.1 vs 14.1, P < 0.001), except at 100 g

when 1+ group had a higher intensity (Tables 2 & 3) 0+

salmon fed reduced meal frequency showed a temporal

lower b*-value compared to the control (3000 g, P <

0.001), but no differences were found at final sampling

(Table 3) b*-value was not affected by reduced meal

frequency in the 1+ group (Table 3)

salmon (<190 g) had a paler fillet colour than larger fish

(>190 g, Tables 2 & 3) However, 0+ salmon was in general

paler than the 1+ fish (Tables 2 & 3) Reduced meal frequency

had no impact on L*-value in any smolt group (Table 3)

Highest a*-, b*-, and L*-values in the 0+ group at

har-vest (Table 3) indicate that colour differences among smolt

groups were related to both the intensity of the present

colour and the colour tone (i.e hue) Colour tone in large

salmon (>3 kg) was characterised by a high intensity of

red/orange compared to smaller fish (<500 g) characterised

by a low intensity of orange/yellow

Water percentage differed between the smolt groups at the

initial (40 g) sampling point, being higher in the 0+

2000 g sampling points, water remained highest in the 0+group Reduced meal frequency as an alternative feedingregime had a significant effect in both smolt groups, with watercontent being reduced in the 1+ treatment group at 3000 and

0+ treatment group, reduced meal frequency had significant

A significant, inverse linear correlation was found between

P < 0.001, n = 400) Although fillet fat was not analysed infish <500 g, it is reasonable to assume that fat contentbetween 40 and 500 g was highest in the 1+ salmon (Fig 8).Opposite to the water content, 0+ fish had a lower fat per-centage in all sampling points, except in the 1100 and 2000 gsamplings (P < 0.05, Tables 2 & 3) The subsequent monthstowards the final sampling were characterised by a largerincrease in the 1+ group compared to 0+ (Table 3), reflecting

a lower fat content in the 0+ group Reduced meal frequencylowered the fat content in the 0+ salmon at 2000 and 3000 gcompared to control, but there were no differences betweenfeeding regimes in the final sampling (Table 3) The 1+ sal-mon provided reduced meal frequency had significantlylower fat content at 3000 g, a difference that persisted until

4300 g (P = 0.011, Table 3)

Figure 8 Change in fillet water content for each group during the experiment Sampling dates are indicated with italic text (0+) or bold text (1+) Control (CG) and treatment groups (TG; reduced meal frequency) for each smolt type were created from ~1.7 (0+) and ~1.3 kg (1+), and significant differences between feeding

Signifi-cant differences between smolt groups (0+ CG vs 1+ CG) are indicated with asterisks; *0.01 < P < 0.05, **0.001 < P < 0.01, or

***P < 0.001.

A linear, inverse correlation was also found for protein

The changes in protein content were relatively small for

both of the 0+ (20.9–19.9%) and 1+ group (21.9–19.8%)

during the experiments (Table 3) However, protein content

was higher in the 1+ salmon sampled at 500 g (P < 0.001)

and 1100 g (P < 0.001), and lower at 2000 g (P = 0.019)

and 3000 g (P = 0.007), compared to the 0+ salmon, and

with no difference between smolt groups at the final

sampling Reduced meal frequency did not influence the

protein content in either the 0+ or the 1+ group (Table 3)

Plasma ghrelin levels were highest in the 1+ group at 190 g

(2.6 ± 0.7 pM, P < 0.001) and in the 0+ group at 500 g

(2.2 ± 2.1 pM, P = 0.022) (Fig 9) Treatment groups

pro-vided reduced meal frequency had 4–6 fold higher plasma

ghrelin levels at 2000 g (1+; P < 0.001) and 3000 g (0+;

P = 0.004) compared to controls From 3000 g, levels were

reduced towards the final sampling, except in the 1+ CG

Thus, there were no differences in plasma ghrelin levels

among the 1+ CG and 1+ TG (~4.0 ± 5.8 pM), while the 0+

CG had higher levels compared to the 0+ TG (2.1 ± 2.9 pM

vs 0.7 ± 0.7 pM, P = 0.005) at the 4300 g sampling point

Ghrelin was slightly negative correlated to fillet fat content

(P > 0.05)

Season had significant impact on k-factor, muscle fibrenumber (P < 0.001), diameter (P = 0.046) and density(P = 0.046), 2.5th and 97.5th percentile (P < 0.009), waterand fat content (P < 0.001), a*-value (P < 0.002), and filletfirmness (P < 0.001) On the other hand, season had noinfluence on gaping, fillet protein or plasma ghrelin levels

Smolt type had significant impact on muscle fibre ter (P < 0.009), 2.5th (P < 0.009) and 97.5th percentile(P = 0.049), fillet firmness (P < 0.001), gaping (P = 0.006),a*-value (P < 0.001), water and fat content (P < 0.001), andghrelin levels (P < 0.001), while k-factor was not affected

diame-Despite a higher growth rate during the seawater period,the 1+ salmon did not catch-up with the larger 0+ fish, but

~4 months This is in contrast to other studies showingsimilar growth performance across smolt types for theoverall rearing period (Duncan et al 1998, 2002; Mørkøre

& Rørvik 2001) In addition, 0+ salmon have been found

to grow faster during the first year in sea while 1+ fishgrow fastest during the second year (Roth et al 2005)

However, the current study differs with regard to strain,experimental design and environmental conditions, andmay explain the differences in the results A growth advan-tage for the 0+ fish during early seawater stages is associ-ated with better hypo-osmoregulatory ability than the 1+

salmon transferred to sea in the spring, mainly due tohigher temperatures (Lysfjord et al 2004) Photoperiodmanipulation is known to boost somatic growth by alteringthe endogenous rhythms and the metabolism in Atlanticsalmon (Duncan et al 2002; Nordgarden et al 2003a,b)

The 1+ salmon in the present study were transferred to sea

in July at temperatures >10 °C, favouring better growthduring the initial seawater phase (Bendiksen et al 2002) Inagreement with others, the concurrent patterns of tempera-ture and light with feed provision and growth suggest amajor influence by season (Duncan et al 1998, 2002; Mør-køre & Rørvik 2001; Nordgarden et al 2003a) Eventhough season, in particular water temperature, appears to

be the dominant driving force of growth differentiationbetween smolt types in the present study, there is reason toassume that smolt type also contributes substantially (John-ston et al 2003; Macqueen et al 2008) It is also known

Figure 9 Change in plasma ghrelin levels for each group during

the experiment Sampling dates are indicated with italic text (0+)

or bold text (1+) Control (CG) and treatment groups (TG;

reduced meal frequency) for each smolt type were created from

~1.7 (0+) and ~1.3 kg (1+), and significant differences between

Sig-nificant differences between smolt groups (0+ CG vs 1+ CG) are

indicated with asterisks; *0.01 < P < 0.05 or ***P < 0.001.

.

that hatching and start-feeding temperatures affect the

expression of different trypsin isozymes (digestive

prote-ases) in the pyloric caeca of Atlantic salmon, where 8 °C is

considered as the lower limit for expression of the common

trypsin isozyme TRP-2*100 (Rungruangsak-Torrissen et al

1998) Proteases with lower temperature optimum might

therefore be presented in the 1+ salmon (incubated at 2 °C)

consequently favouring improved digestion at low

tempera-tures, compared to 0+ salmon incubated at 8 °C which

pos-sibly are better adapted to higher temperatures This might

explain some of the differences in both growth performance

and pattern between smolt types during the seawater

per-iod Reduced water temperature, growth restriction and

continuous light treatment have all been demonstrated to

improve feed utilisation in salmon (Jobling 1994; Bendiksen

data indicate that the feed utilisation was lowest between

March and May when also temperatures were at the lowest

(0+ salmon), in line with others (Mørkøre & Rørvik 2001)

Hence, low feed utilisation during certain periods of the

year seems to be a problem in commercial salmon farming

However, it can be challenging to prevent overfeeding in

large open sea cages with fluctuating weather conditions,

but reduced nutrient digestibility at lower temperatures and

vertical distribution of fish might have contributed

(Opped-al et (Opped-al 2001; Bendiksen et (Opped-al 2003; Ytrestøyl et (Opped-al 2005)

Seasonal changes in accumulated feed utilisation were

pres-ently observed, but there were no differences between smolt

types at the final sampling This suggests that the time of

harvest has an impact on the final results Even though the

0+ fish grew slower during the seawater phase, this is an

important strategy to spread the availability of market-size

salmon over the year from a commercial point of view

(Duncan et al 1998) Thus, the out-of-season smolt

strat-egy can potentially improve the production capacity and

contribute to balance the supply and demand of the

market

most salmonids to maintain maximum feed intake and

growth, a commonly used feeding strategy in the Norwegian

salmon industry, but the optimal frequency decreases with

increased body size and lower temperatures (Alana¨ra¨ et al

2001; Storebakken 2002) Although the energy content of

commercial salmon feed has increased during the recent

years, feeding regimes have not changed substantially

Therefore, optimisation of feeding practices in relation to

season and size are thought to be of significant importance

in control of flesh quality The extensive fat accumulation

seen in salmon (Jobling et al 2002b), might indicate that

the energy obtained through the feed is above the ment adequate for maximum growth during most of theseason, especially in large fish In the present experiment,the capacity to feed in the 0+ treatment group was reducedduring the first weeks, but then improved However,reduced meal frequency did not affect growth or feedutilisation (Fig 3c,d) On the other hand, slightly reduced

1+ group This indicates that salmon have the ability toadapt to reduced meal frequency (within ~3 months), possi-bly by increasing stomach capacity to maintain a highgrowth or by inducing compensatory catch-up growth.Feeding frequency has been shown to strongly correlatewith the gastric evacuation time in juvenile Korean rockfish(Sebastes schlegeli), resulting in improved feed utilisationwith increased time for digestion (Lee et al 2000) Linne´r &Bra¨nna¨s (2001) have also showed that rainbow trout fedreduced meal frequency grows better due to reduced meta-bolic rate and swimming activity, e.g low energy consump-tion, in agreement with others (Jobling 1994; Carter et al.2001)

This study shows that there are major differences in musclecellularity; i.e fibre number, density, diameter and size-distribution between the 0+ out-of-season and the 1+ in-season salmon Season and smolt type contribute signifi-cantly, both having an effect on hyperplastic (new recruit-

diameter) myotomal growth This is in agreement with vious studies indicating that thermal experience during theembryogenesis determines the myogenic phenotype duringthe subsequent seawater period for adult salmon (Johnston

2008) Macqueen et al (2008) have demonstrated that vest sized salmon (~4 kg) incubated at 5 or 8 °C have ahigher final fibre number than groups incubated at 2 or

har-10 °C In general, growth rates increase with temperature,thus incubation at 2 and 5 °C leads to significant smallerfish at smoltification than those incubated at higher tem-peratures (Macqueen et al 2008) However, the low incu-

growth during the seawater period (Macqueen et al 2008),also seen in the present experiment The 0+ salmon incu-bated at 8 °C have a higher final fibre number compared

to the 1+ group incubated at 2 °C It is therefore ably to assume that higher incubation temperature (i.e.smolt type) in the present study has influenced muscle

reason-hyperplasia through increased number of myogenic

progen-itor cells developed during embryogenesis (Johnston 2006)

Nevertheless, temperature and light may also have

contrib-uted during the seawater growth period as described by

Johnston et al (2003b, 2004) In line with Johnston et al

(2004), the present data therefore suggest that immediate

application of continuous light enhances fibre recruitment

in the 0+ fish, but not permanently Periods of low

temper-ature were concurrent with periods of reduced fibre

recruit-ment, and may indicate a greater influence of temperature

than light during the late seawater period However, the

bimodal fibre distribution in the 0+ salmon, e.g increased

fibre recruitment might be caused by the prenatal

tempera-ture regime as previously suggested An alternative

out-of-season smolt production based on light manipulation

(Vieira et al 2005), did not have any effect on muscle

cellu-larity at harvest These results illustrate that the choice of

smolt production strategy can influence muscle cellularity

in commercial harvest size salmon

Muscle cellularity is known to affect flesh texture in

sev-eral fish species (Hatae et al 1990; Hurling et al 1996;

Johnston 1999, 2001b; Hagen et al 2007), confirmed in

explained ~10% and ~16% of the total variation in fillet

firmness, respectively Thus, textural differences between

smolt types are partly explained by the difference in

mus-cle cellularity, influenced by both season and smolt type

At the family level, a positive correlation between fibre

density and raw flesh texture has previously been

demon-strated, explaining 49% of the total variation (Johnston

panel (Johnston et al 2000a) Other factors important for

raw muscle texture are collagen (Hatae et al 1986; Sato

2005) The softer texture in the 0+ treatment than control

group at 2000 g is likely caused by thinner fillets due to

reduced k-factor Gaping results in downgrading during

processing of salmon fillets (Michie 2001), and is caused

by breakage of the myoseptum (Johnston et al 2002)

Johnston et al (2002) demonstrated that individual

varia-tion in fibre density is important in the development of

gaping, and almost non-existing with fibre density in

muscle cellularity on gaping was hypothesised through the

amount and distribution of connective tissue and

cytoskel-etal proteins binding the contractile filaments to the commata, possibly with contribution of muscle proteinasesduring post-mortem storage (Johnston et al 2002) How-ever, no such correlation was seen in the present study,probably affected by the low levels, sensitivity of themethod, or covered by other factors Nevertheless, signifi-cant differences in gaping were present between the 0+

myo-and 1+ group, except at the final sampling Smolt typewas the major contributor, with no impact of season,while reduced meal frequency had a significant effect inthe 1+ group Other experiments have suggested that mus-cle remodelation after periods of rapid growth may causesoft flesh in fish (Christiansen et al 1992; Martinez et al

2011), as well as reduced concentration of mature collagencross-links (Li et al 2005; Johnston et al 2006; Hagen

slaughter stress from fish loadings, transport or holdingmight be of greater importance for major gaping prob-lems in modern fish farming than feed or feeding duringthe grow out period Future studies should investigate theinfluence of hydroxylysyl pyridinoline (PYD), e.g maturecollagen cross-links, in fish having high to extreme gaping

at the time of harvest

In the present study, fillet fat content had no impact onfillet firmness or other quality characteristics However, arecent study identified fillet fat as the physical flesh charac-teristic having the largest impact on organoleptic properties

in cooked salmon muscle (Johnsen et al 2011) The presentalterations of fillet fat content, due to impact of smolt type,season, and/or feeding regime, might therefore be of signifi-cant importance for the consumer’s acceptance of pro-cessed products In the 0+ group, it is suggested thatpermanent effects are obtainable through adjustments ofthe alternative feeding regime, for instance by extendingthe winter-regime period (e.g one meal per second day)towards the end of the experiment Feed energy is also welldocumented to influence fat deposition in salmon (Jobling

content prior to slaughter are e.g starvation and reducedfeed rations, but lost or reduced growth has been consid-ered a major drawback of such alternatives (Einen et al

1998, 1999) The present study demonstrates that reducedmeal frequency is an effective tool to control fat deposition

in salmon muscle without compromising growth mance, and thus suitable for commercial use It is also ofparticular importance to notice that 1+ salmon accumulatemore fillet fat than 0+ fish, and that most fillet fat is depos-ited during autumn, both critical factors for optimaltailoring

perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor- perfor-.

Colour is an important quality criterion for consumer

salmon flesh is due to absorption of carotenoid pigments

(mainly astaxanthin) from the diet (Davies 2008) In the

present study, a*-value and pigment concentration were

highly correlated, thus both are good indicators for flesh

pigmentation Other factors such as muscle cellularity

(Johnston et al 2000a, 2004), connective tissue, cytoskeletal

proteins, proteolytic enzymes and pH are likely to influence

the degree of sarcomere miss-alignment and hence the

internal organisation of pigment molecules within the flesh,

and therefore colour perception (Johnston et al 2006) The

present study supports that other factors than carotenoid

pigments affect colour perception in salmon, demonstrated

by the significant correlation between the muscle cellularity

and a*-value Nevertheless, flesh astaxanthin seems to be

the major contributor to colour, explaining ~40% of the

total variation Previous studies have demonstrated that

both feed intake and temperature affect digestibility of

as-taxanthin in salmon, being reduced at lower temperatures

and with higher feed intake and growth (Ytrestøyl et al

2005, 2006) In line with Ytrestøyl et al (2004), the 0+

sal-mon in the current study had higher pigment concentration

and colour intensity than 1+ fish at harvest In addition to

smolt type and season, reduced feeding and growth during

the seawater period appears to have contributed to

improved pigment utilisation in the 0+ group However,

fibre recruitment did also contribute, strongly affected by

the incubation temperature (i.e smolt type) The colour

drop seen during autumn is concurrent with shorter days in

both smolt groups, but the reason for this is unknown A

recently published study has shown reduced carotenoid

digestibility with increased feed rations, possible explaining

poorer pigmentation in rapidly growing salmon (Rørvik

nega-tive effect on colour, probably due to temporal reduced

feed provision (affecting feed intake) and hence lower

intake of pigments However, colour was restored towards

harvest, possibly due to improved pigment utilisation, a

reasonable assumption as there were no differences in feed

provision or fibre recruitment

An important finding of the present study was that the

plasma ghrelin levels were substantially raised in the 1+ and

0+ treatment groups at 2000 and 3000 g respectively A

delayed raise was seen in the 1+ control group compared to

the treatment group, whereas the levels were relatively stable

in the 0+ control group (~1.8 pM) during the study It istherefore evident that the feeding regime had a significanteffect on the plasma ghrelin levels, possibly through short-term appetite control influenced by the inter-meal timeperiod This may also explain the elevated ghrelin levels inthe 1+ control group in October 2009, influenced by shorterdays affecting time for digestion Other studies have shownthat ghrelin has an orexigenic effect during short-term star-

2009) A similar orexigenic effect was seen in the presentstudy with reduced meal frequency Plasma ghrelin levelsalso differed between smolt types, with no impact of season

In contrast, a previous study suggested that ghrelin is related

to long-term regulation of energy homeostasis in Arcticcharr through seasonal changes (Frøiland et al 2010) Thissupports the notion that ghrelin may have species-specificroles, even among salmonids or among life stages Smolttype was found to contribute significantly to ghrelin levels inthe present study, but the mechanism of this long-term effect

is unclear However, different growth potential establishedduring the embryogenesis might have contributed A weakcorrelation was found between plasma ghrelin levels andmuscle fat content, but this was highly influenced by thefeeding regime Thus, a direct causal relationship appearsnot to exist in the present study The present data thereforesuggest that ghrelin may have a role in the short-term stimu-lation of appetite and food intake in salmon, similar to itsrole in mammals Ghrelin may also have a long-term regula-tory effect through the impact of smolt type, possibly alsorelated to feed intake and growth control However, ghrelindoes not appear to act as an adiposity signal

The present study demonstrates substantial differences ingrowth performance, muscle structure, flesh quality, andplasma ghrelin levels among the 0+ and 1+ salmon Seasonhad a significant impact on most flesh characteristics,which also are highly influenced by smolt type Musclefibre recruitment and final fibre number were importantdeterminants of fillet firmness and colour, possibly influ-enced by the prenatal temperature regime Reduced mealfrequency gave initial reduced feeding and growth in the 1+salmon, also influencing gaping (1+), fillet fat (0+ and 1+),and flesh colour (0+ and 1+), while muscle cellularity wasunaffected Ghrelin levels seemed to stimulate short-termappetite, and were influenced by smolt type Reduced mealfrequency is therefore considered to be a promising tool for

managing important flesh characteristics in salmon without

compromising growth performance

We thank Gildeska˚l Research Station AS for fish

hus-bandry, in particular Mr Roy Arne Eilertsen and Mr

Ed-vard Kjørsvik The assistance during processing of samples

from the technical staff in the Seafood Quality Research

Group at the Faculty of Biosciences and Aquaculture,

Uni-versity of Nordland, is also appreciated This study was

funded by BioMar AS, the Norwegian Research Council

(project 180007), and the University of Nordland

Alana¨ra¨, A., Kadri, S., Paspatis, M (2001) Feeding management.

In: Food Intake in Fish (Houlihan, D., Boujard, T & Jobling,

Bendiksen, E.A˚., Jobling, M & Arnesen, A.M (2002) Feed intake

of Atlantic salmon parr Salmo salar L in relation to

Bendiksen, E.A˚., Berg, O.K., Jobling, M., Arnesen, A.M &

Ma˚søval, K (2003) Digestibility, growth and nutrient

utilisa-tion of Atlantic salmon parr (Salmo salar L.) in relautilisa-tion to

Bowman, A.W & Azzalini, A (1997) Applied Smoothing

Tech-niques for Data Analysis: The Kernel Approach with S-Plus

Illustration Oxford Science Publications, Oxford University

Press, Oxford.

Carter, C., Houlihan, D., Kiessling, A., Me´dale, F., Jobling, M.

(2001) Physological effects of feeding In: Food Intake in Fish

Blackwell Science Ltd, Oxford.

Christiansen, J.S., Martinez, I., Jobling, M & Amin, A.B (1992)

Rapid somatic growth and muscle damage in a salmonid fish.

Davies, S.J (2008) Colouration and flesh quality in farmed salmon

and trout In: Improving Farmed Fish Quality and Safety (Lie,

Duncan, N.J., Auchinachie, N., Robertson, D., Murray, R &

Bromage, N (1998) Growth, maturation and survival of

out-of-season 0+ and 1+ Atlantic salmon (Salmo salar) smolts.

Duncan, N.J., Thrush, M.A., Elliott, J.A.K & Bromage, N.R.

(2002) Seawater growth and maturation of Atlantic salmon

(Salmo salar) transferred to sea at different times during the

Einen, O & Thomassen, M.S (1998) Starvation prior to slaughter

in Atlantic salmon (Salmo salar) II White muscle composition

and evaluation of freshness, texture and colour characteristics in

Einen, O., Waagan, B & Thomassen, M.S (1998) Starvation prior

weight loss, body shape, slaughter- and fillet-yield, proximate

Einen, O., Mørkøre, T., Røra˚, A.M.B & Thomassen, M.S (1999)

product quality of Atlantic salmon (Salmo salar) Aquaculture,

178, 149–169.

Frøiland, E., Murashita, K., Jørgensen, E.H & Kurokawa, T.

(2010) Leptin and ghrelin in anadromous Arctic charr: Cloning and change in expressions during a seasonal feeding cycle Gen.

Hagen, Ø & Solberg, C (2010) Fasting of farmed Atlantic cod (Gadus morhua L.) used as tool to improve fillet texture during

Hagen, Ø., Solberg, C & Johnston, I.A (2006) Sexual dimorphism

of fast muscle fibre recruitment in farmed Atlantic halibut

Hagen, Ø., Solberg, C., Sirnes, E & Johnston, I.A (2007) chemical and structural factors contributing to seasonal varia- tion in the texture of farmed Atlantic halibut (Hippoglossus

Hatae, K., Tobimatsu, A., Takeyama, M & Matsumoto, J.J.

(1986) Contribution of the connective tissues on the texture

Hatae, K., Yoshimatsu, F & Matsumoto, J.J (1990) Role of

Hevrøy, E., Azpeleta, C., Shimizu, M., Lanze´n, A., Kaiya, H., Espe, M & Olsvik, P (2010) Effects of short-term starvation on ghrelin, GH-IGF system, and IGF-binding proteins in Atlantic

Hillestad, M., Johnsen, F., Austreng, E & A˚sga˚rd, T (1998) Long-term effects of dietary fat level and feeding rate on growth, feed utilization and carcass quality of Atlantic salmon Aquacult.

Hosoda, H., Kojima, M., Matsuo, H & Kangawa, K (2000) Ghrelin and des-acyl ghrelin: two major forms of rat ghrelin peptide in gastrointestinal tissue Biochem Biophys Res Com-

Hurling, R., Rodell, J.B & Hunt, H.D (1996) Fiber diameter and

Jobling, M (1994) Fish Bioenergetics Chapman and Hall, London.

Jobling, M., Larsen, A.V., Andreassen, B & Olsen, R.L (2002a) Adiposity and growth of post-smolt Atlantic salmon Salmo salar

Jobling, M., Larsen, A.V., Andreassen, B., Sigholt, T & Olsen, R.

L (2002b) Influence of a dietary shift on temporal changes in fat deposition and fatty acid composition of Atlantic salmon post-smolt during the early phase of seawater rearing Aquacult.

Johansen, S.J.S (2003) Compensatory Growth and the Lipostat:

Patterns of Growth and Lipid Deposition in Atlantic Salmon, Salmo salar L University of Tromsø, Norway.

Johnsen, C.A (2006) Fôringsregimets innvirkning på tilvekst og kvalitet i Atlantisk laks, Salmo salar L (Effects of feeding regime on growth and flesh quality in Atlantic salmon, Salmo

University of Nordland, Bodø, Norway.

Johnsen, C.A., Hagen, Ø & Bendiksen, E.A˚ (2011) Long-term effects of high-energy, low-fishmeal feeds on growth and flesh characteristics of Atlantic salmon (Salmo salar L.) Aquaculture,

312, 109–116.

Johnston, I.A (1999) Muscle development and growth: potential

.

Johnston, I.A (2001a) Genetic and environmental determinants of

muscle growth pattern In: Muscle Development and Growth

Johnston, I.A (2001b) Implications of muscle growth patterns for

the colour and texture of fish flesh In: Farmed Fish Quality

Oxford.

Johnston, I.A (2006) Environment and plasticity of myogenesis in

Johnston, I.A., Straugnell, G., McCraken, M.L & Johnstone, R.

(1999) Muscle growth in normal sex ratio and all-female diploid

Johnston, I.A., Alderson, R., Sandham, C et al (2000a) Muscle

fibre density in relation to the colour and texture of smoked

Johnston, I.A., Alderson, R., Sandham, C et al (2000b) Patterns

of muscle growth in early and late maturing populations of

Johnston, I.A., McLay, H.A., Abercromby, M & Robins, D.

(2000c) Early thermal experience has different effects on growth

and muscle fibre recruitment in spring- and autumn-running

Johnston, I.A., Manthri, S., Alderson, R., Campbell, P., Mitchell,

D., Whyte, D., Dingwall, A., Nickell, D., Selkirk, C &

Robert-son, B (2002) Effects of dietary protein level on muscle

cellular-ity and flesh qualcellular-ity in Atlantic salmon with particular reference

Johnston, I.A., Manthri, S., Alderson, R., Smart, A., Campbell,

P., Nickell, D., Robertson, B., Paxton, C.G.M & Burt, M.L.

(2003a) Freshwater environment affects growth rate and muscle

fibre recruitment in seawater stages of Atlantic salmon (Salmo

Johnston, I.A., Manthri, S., Smart, A., Campbell, P., Nickell, D.

& Alderson, R (2003b) Plasticity of muscle fibre number in

sea-water stages of Atlantic salmon in response to photoperiod

Johnston, I.A., Manthri, S., Bickerdike, R., Dingwall, A., Luijkx,

R., Campbell, P., Nickell, D & Alderson, R (2004) Growth

per-formance, muscle structure and flesh quality in out-of-season

Atlantic salmon (Salmo salar) smolts reared under two different

Johnston, I.A., Li, X., Vieira, V.L.A., Nickell, D., Dingwall, A.,

Alderson, R., Campbell, P & Bickerdike, R (2006) Muscle and

flesh quality traits in wild and farmed Atlantic salmon

Jo¨nsson, E., Forsman, A., Einarsdottir, I.E., Kaiya, H.,

Ruoho-nen, K & Bjo¨rnsson, B.Th (2007) Plasma ghrelin levels in

rain-bow trout in response to fasting, feeding and food composition,

and effects of ghrelin on voluntary food intake Comp Biochem.

Jo¨nsson, E., Kaiya, H & Bjo¨rnsson, B.Th (2010) Ghrelin

decreases food intake in juvenile rainbow trout (Oncorhynchus

mykiss) through the central anorexigenic corticotropin-releasing

Kaiya, H., Kojima, M., Hosoda, H., Moriyama, S., Takahashi, A.,

Kawauchi, H & Kangawa, K (2003) Peptide purification,

com-plementary deoxyribonucleic acid (DNA) and genomic DNA

cloning, and functional characterization of ghrelin in rainbow

Kaiya, H., Miyazato, M., Kangawa, K., Peter, R.E & Unniappan,

S (2008) Ghrelin: A multifunctional hormone in

non-mamma-lian vertebrates Comp Biochem Physiol A: Mol Integr

Lee, S.-M., Hwang, U.-G & Cho, S.H (2000) Effects of feeding frequency and dietary moisture content on growth, body compo- sition and gastric evacuation of juvenile Korean rockfish

Li, X.J., Bickerdike, R., Lindsay, E., Campbell, P., Nickell, D., Dingwall, A & Johnston, I.A (2005) Hydroxylysyl pyridinoline cross-link concentration affects the textural properties of fresh and smoked Atlantic salmon (Salmo salar L.) flesh J Agric.

Linne´r, J & Bra¨nna¨s, E (2001) Growth in Arctic charr and bow trout fed temporally concentrated or spaced daily meals.

Lysfjord, G., Jobling, M & Solberg, C (2004) Atlantic salmon,

Macqueen, D.J., Robb, D.H.F., Olsen, T., Melstveit, L., Paxton, C.G.M & Johnston, I.A (2008) Temperature until the ‘eyed stage’ of embryogenesis programmes the growth trajectory and

298.

Martinez, I., Wang, P.A., Slizyte´, R., Jorge, A., Dahle, S.W., Can˜as, B., Yamashita, M., Olsen, R.L & Erikson, U (2011) Protein expression and enzymatic activities in normal and soft

Michie, I (2001) Causes of downgrading in the salmon farming industry In: Farmed Fish Quality (Kestin, S & Warris, P eds),

Mørkøre, T & Rørvik, K.A (2001) Seasonal variations in growth, feed utilisation and product quality of farmed Atlantic salmon (Salmo salar) transferred to seawater as 0+ smolts or 1+ smolts.

Mørkøre, T., Mazo T, P.I., Tahirovic, V & Einen, O (2008) Impact of starvation and handling stress on rigor development and quality of Atlantic salmon (Salmon salar L) Aquaculture,

277, 231–238.

Murashita, K., Kurokawa, T., Nilsen, T.O & Rønnestad, I (2009) Ghrelin, cholecystokinin, and peptide YY in Atlantic salmon (Salmo salar): molecular cloning and tissue expression Gen.

Qual-ity Measurements, p 4 Standard Norge, Oslo.

Fat, p 2 Standard Norge, Oslo.

Oppedal, F., Juell, J.E., Tarranger, G.L & Hansen, T (2001) ficial light and season affects vertical distribution and swimming behaviour of post-smolt Atlantic salmon in sea cages J Fish

Rørvik, K.A., Ytrestøyl, T., Lundberg, E., Jakobsen, F.A., sen, A.A & Bjerkeng, B (2010) How apparent digestibility of carotenoids, macronutrients, and minerals are differently affected

Jakob-by ration level in Atlantic salmon, Salmo salar J Appl

Roth, B., Johansen, S.J.S., Suontama, J., Kiessling, A., Leknes, O., Guldberg, B & Handeland, S (2005) Seasonal variation in

flesh quality, comparison between large and small Atlantic

sal-mon (Salmo salar) transferred into seawater as 0+ or 1+ smolts.

Rungruangsak-Torrissen, K., Pringle, G.M., Moss, R & Houlihan,

D.F (1998) Effects of varying rearing temperatures on

expres-sion of different trypsin isozymes, feed converexpres-sion efficiency and

growth in Atlantic salmon (Salmo salar L.) Fish Physiol

Sato, K., Yoshinaka, R., Sato, M & Shimizu, Y (1986) Collagen

con-tent in the muscle of fishes in association with their swimming

Shearer, K.D., Asgard, T., Andorsdottir, G & Aas, G.H (1994)

Whole-body elemental and proximate composition of Atlantic

Solberg, C (1992) Near infrared spectroscopy of fish samples In:

Near Infra-Red Spectroscopy Bridging the Gap Between Data

Analysis and NIR Application (Hildrum, K.I et al eds),

Seafood From Producer to Consumer Integrated Approach to

Amterdam.

Storebakken, T (2002) Part I: Marine fish Atlantic salmon, Salmo salar In: Nutrient Requirement and Feeding of Finfish for

CABI Publishing, New York.

Vieira, V.L.A., Johansen, S.J.S., Bickerdike, R & Johnston, I.A.

(2005) Impact of accelerated smoltification on muscle structure and fillet firmness at harvest in Atlantic salmon (Salmo salar).

Ytrestøyl, T., Coral-Hinostroza, G., Hatlen, B., Robb, D.H.F &

Bjerkeng, B (2004) Carotenoid and lipid content in muscle of Atlantic salmon, Salmo salar, transferred to seawater as 0+ or 1+ smolts Comp Biochem Physiol B: Biochem Mol Biol., 138,

Ytrestøyl, T., Struksnæs, G., Koppe, W & Bjerkeng, B (2005) Effects of temperature and feed intake on astaxanthin digestibil- ity and metabolism in Atlantic salmon, Salmo salar Comp Bio-

Ytrestøyl, T., Struksnæs, G., Rørvik, K.A., Koppe, W & eng, B (2006) Astaxanthin digestibility as affected by ration

224.

.

1 2, 2 2 1

1 1

Animal Nutrition, Faculty of Fisheries, Kagoshima University, Kagoshima, Japan

A 50-day feeding trial was conducted to determine the

effects of dietary oxidized fish oil (OFO) and vitamin C

(VC) on growth and oxidative stress in juvenile red sea

bream Test diets were formulated with 2 degrees of

VC (0, 400 and 800 ppm) No significant difference was

found on growth performance between fish fed OFO with

400 or 800 mg VC and the control group that fed a diet

with fresh fish oil after 50 days However, fish fed OFO

without VC supplement indicated significantly poor growth

than the control group Liver and muscle thiobarbituric

acid reactive substances were reduced by increased VC

intake of fish Fish fed diets containing low OFO with 400

and 800 mg VC, high OFO with 800 mg VC, and fresh fish

oil are allocated in the zone of high resistance against

oxi-dative stress together with low oxioxi-dative stress condition

On the other hand, no VC supplemented group was under

the highest oxidative stress condition In conclusion,

die-tary oxidized lipid increased the oxidative stress condition

supple-ment improved growth and health of juvenile red sea

bream

oxidative stress, red sea bream, vitamin C

Received 16 September 2011; accepted 5 December 2011

Correspondence: S Koshio, Laboratory of Aquatic Animal Nutrition,

Faculty of Fisheries, Kagoshima University, Shimoarata 4-50-20,

Kago-shima 890-0056, Japan E-mail: koshio@fish.kagoKago-shima-u.ac.jp

Fish oil containing large amount of n-3 highly unsaturatedfatty acids is susceptible to lipid peroxidation, whereoxygen attacks the double bond in fatty acids to form lipidhydroperoxides (Frankel 1998) Lipid hydroperoxides arethe primary products of lipid oxidation, and they areunstable and readily decompose fatty acid radicals, and reactwith cellular biomembranes such as protein and DNA,causing severe lesions and subsequently high mortality

of animals (Cortesi & Privett 1972) Secondary products oflipids oxidation include alcohols, ketones, and aldehydes,many of which not only produce special flavour, colour andtexture, but also decrease the nutritive values and safety ofproducts Undesirable influences on fish metabolism byconsuming oxidized dietary oil have been observed in anumber of studies (Sakai et al 1992; Koshio et al 1994;Kosutarak et al 1995; Baker & Davies 1997; Hamre et al.2001; Mourente et al 2002; Gao et al 2011) On the otherhand, to protect cells and tissues from oxidative damage, fishhave endogenous antioxidant defense system such assuperoxide dismutase and catalase to help counteract theactivity of free radical They also have different abilities toprotect body against oxidative damage The other defensivemethod is to use exogenous antioxidative compounds such

as vitamins C, E and A, which scavenge free radicals.Vitamin C (VC) is an essential nutrient for physiologicalfunction of most animal species (Lim & Lovell 1978), and

it plays an important role as water-soluble antioxidant.The first and major line of antioxidative defense systemagainst radicals by VC is to prevent lipid peroxidation inplasma (Nordberg & Arner 2001) The VC also promotesbeneficial effects on serum bactericidal activity, phagocyticactivity, antibody levels and lysozyme activity (Hardie

.

Aquaculture Nutrition

there is still limited information on the relationship

between dietary VC and its antioxidative effects with

differ-ent degrees of lipid oxidation for fish species

Generally, measurement of hematological parameters

may indicate oxidative status as well as general health

con-dition of cultured fish Ren et al (2005) reported that

Hematocrit value was enhanced by dietary VC levels for

Japanese eel On the other hand, base on our unpublished

data (Komilus 2008), red sea bream fed dietary oxidized

lipid generated high level of reactive oxygen species and

decreased efficacy of antioxidant system in blood plasma

Therefore, this study was carried out to determine the

effects of dietary oxidized fish oil (OFO) and VC

supple-mentation on growth performance, tissues vitamins C and

E contents, lipid peroxidation, fatty acid composition and

blood parameters of juvenile red sea bream, Pagrus major,

which is one of most important aquaculture species in

Japan

Cod liver oil without any antioxidant was obtained from

Nippon Suisan Kaisha Ltd (Tokyo, Japan) and oxidized

(POV) was monitored every 8 h intervals until the values

high oxidation, respectively, were reached The basal diet

-Ascorbyl-2-monophosphate-Na/Ca (AMP-Na/Ca) (0, 500

two degrees of OFO were supplemented to the basal diet

sub-stances and fatty acid composition of test diets are

until feeding

Juvenile red sea bream were obtained from a local

hatch-ery, in Miyazaki prefecture, Japan, and transported alive to

the Kamoike Marine Production Laboratory, Faculty of

Fisheries, Kagoshima University, Japan Prior to the

feed-ing trial, all fish were acclimated to the indoor rearfeed-ing

conditions for 2 weeks and fed a commercial pelleted fishfeed (Higashimaru, Kagoshima, Japan) A total of 252 fish

hundred round tanks (100 L) with through ambient perature filtered seawater at the same density (12 fish

regime was applied in the trial The pH and salinity of the

test diets at 08:00 and 16:00 h to apparent satiation levelfor 50 days Uneaten diets were collected every day, driedand weight to determine feed intake (FI) At the end offeeding trial, all fish were individually weighed and mea-surements of fork length were taken to obtain the data ofgrowth performances of fish Three fish were randomlysampled from each tank for whole-body proximate analy-

tank and pooled into 1 tube as a blood sample And fromthe other fish, liver was dissected out and weighed individu-

Table 1 Formulation and proximate composition of basal diet g kg Ingredients

1 Nippon Suisan, Ltd., Japan.

2 Cod liver oil without antioxidants, Nippon Suisan Ltd., Japan.

3

.

3230 mg; K2HPO4, 8870 mg; Fe Citrate, 1100 mg; Ca Lactate,

12090 mg; Al (OH)3, 10 mg; ZnSO4, 130 mg; CuSO4, 4 mg; MnSO4,

30 mg; Ca (IO3)2, 10 mg; CoSO4, 40 mg.

a-Toch-opherol Acetate (E), 31.7 mg; Thiamine-Nitrate (B1), 48.1 mg;

Riboflavin (B2), 160.3 mg; Pyridoxine-HCl (B6), 38.2 mg; balamine (B12), 0.1 mg; d-Biotin, 4.8 mg; Inositol, 3207.6 mg; Nia- cine (Nicotic acid), 641.4 mg; Ca Panthothenate, 224.6 mg; Folic

319.4 mg.

K.K.)

.

ally to calculate hepatosomatic index Brain, liver, heart,

eye and muscle (without skin) samples were collected,

contents Only muscle and liver used for the analysis of

α-tocopherol content and TBARS while some parts of the

muscle were also freeze-dried for fatty acid composition

Crude protein was determined by the Kjeldahl method

(Association of Official Analysis Chemists (AOAC) 1995),

total lipid was extracted by the method of Bligh & Dyer

(1959), and crude ash was analyzed by combustion in

constant weight according to Association of Official

Analy-sis Chemists (AOAC) (1995) POV of oil samples was

determined by the method of IUPAC (1987) Measurement

of TBARS was carried out using a method adapted from

Yagi (1987)

The AsA contents of fish samples were quantified by the

method of Sakakura et al (1998) AsA contents of test

diets were determined based on the method of Ai et al

(2006) with some modifications Powdered diet samples

distilled water, shaked for 25 min, and let it stand for 25 min,

and 30 ml of upper water solution layer was collected and

centrifuged at 3000 rpm for 5 min 0.2 M acetic acid buffer

water-bath for 2.0 h, then cooled on an ice water-bath 0.15 mL of

phosphoric acid was added to it, and centrifuged at 3000 rpmfor 6 min Then 0.4 mL supernatant was mixed with 0.04 mL0.2% 2,6-sodium dichloroindophenol solution, 0.04 mL 3%thiourea solution (50% ethanol), 0.4 mL 50% sodiumacetate and 0.4 mL 0.1% o-phenylene diamine solution

high-performance liquid chromatography (HPLC) tions of HPLC analysis for AsA contents in fish tissues ordiets were those previously described in Sakakura et al.(1998) For whole blood AsA concentration, we followed themethod as described by Sato et al (1991)

determined as described previously (Gao et al 2011) Fattyacid compositions of test diets, liver and muscle sampleswere analyzed according to Querijero et al (1997) Condi-tions for HPLC analysis were the same as those previouslydescribed by Komilus et al (2008)

The micro hematocrit method was used for the nation of hematocrit level (Ren et al 2005) Hemoglobinwas determined by a SPOTCHEM (SPOTCHEMtm EZmodel SP-4430, Arkray, Inc Kyoto, Japan) Reactive oxy-

potential (BAP) were also measured spectrophotometricallyfrom blood plasma with an automated analyzer FRAS(Diacron International s.r.l., Grosseto, Italy) by following(Morganti et al 2002; Kader et al 2011)

Statistical analysis was performed with analysis of

Fish oil status

∑ Saturates includes 14:0, 16:0 and 18:0.

∑ Monoences includes 16:1n-7, 18:1n-9, 18:1n-7, 20:1n-9, 22:1n-11 and 22:1n-9.

∑ n-6 includes 18:2n-6, 20:2n-6 and 20:4n-6.

∑n-3 includes 18:3n-3, 18:4n-3, 20:4n-3, 20:5n-3, 22:5n-3 and 22:6n-3.

Abacus 19 Concepts, Berkeley, CA, USA) The Data

from each group were compared using Turkey Kramer test

used to test the effects of dietary VC level and degree of

die-tary oxidized lipid, and their interactions excluding the

con-trol diet

Growth performances of the fish are given in Table 3 VC

supplementation was a significant factor to affect growth

parameters such as final weight (FW), body weight gain(BWG), specific growth rate (SGR) and feed conversionratio (FCR) On the other hand, the degree of oxidationapplied in this study was not a significant factor on thegrowth Fish fed OFO without dietary VC supplementation(Diets 2 and 5) showed significantly lower FW, BWG,

no significant difference was found in survival rate and FI

good growth to that of control diet was found in fish fed

Initial wt (g)

Final wt (g)

3

4

3

.

Only oxidation degree was a significant factor on crude

protein and fat of fish whole body, but both factors were

significant on crude ash of fish body (Table 4) Whole body

protein contents significantly decreased with increased

dietary oxidation, and those of fish fed higher oxidized oil

were significantly lower than those of control group On the

other hand, lipid contents significantly increased with

increased dietary oxidation, and those of fish fed higher

oxidized oil were significantly higher than those of control

group Whole body ash contents significantly increased with

increased dietary VC, and decreased with increased dietary

oxidation

AsA contents of several fish organs and whole blood are

shown in Table 5 Significant effect was detected for only

dietary VC levels on AsA contents in organs and whole

blood measured AsA contents in those organs and whole

blood increased with increased dietary VC In comparisonwith the control, fish fed diet without VC had significantlylower AsA contents in the organs and plasma after 50 days

supplementa-tion and oxidasupplementa-tion degree were significant factors, but thiseffect was not on muscle (Table 6) When fish fed the

increased with in creased dietary VC levels With increased

oxidation degree was an only significant factor on TBARS

of liver In other wards, the values increased with increasedoxidation degree in the diets Although the statistical signif-icance was not detected, muscle TBARS decreased withincreased dietary VC and increased with increased dietaryoxidation degrees

Fatty acid compositions of muscle and liver of fish were

shown in Tables 7 and 8, respectively Incremental dietary

VC increased percentages of EPA, DHA, 22:5n-3 and

total n-3 fatty acids in both liver and muscle of fish fedOFO diets On the contrary, the percentages of 14:0, 18:0and 18:1n-9 decreased with increased dietary VC levels

Table 7 Fatty acid composition (%) of muscle of red sea bream fed test diets for 50 days

Fatty acids

Values are expressed as mean ± SE from triplicate groups Data with an asterisk in a row are significantly different from those of

Fatty acids

.