Gene . Sel. E ol. 39 (2007) 431–446 A ailable online a :
c
INRA, EDP Sciences, 2007 www.gse-jou nal.o g
DOI: 10.1051/gse:2007013 O iginal a icle
B eeding salmonids o eed efficiency
in cu en ishmeal and u u e plan -based
die en i onmen s
Che yl D. Qa∗, An i Ka, Juha Kb,
Ossi Rc
aMTT Ag i ood Resea ch Finland, Bio echnology and Food Resea ch, Biome ical Gene ics,
31600 Jokioinen, Finland
bFinnish Game and Fishe ies Resea ch Ins i u e Jy äskylä, Su on ie 9, 40500 Jy äskylä,
Finland
cFinnish Game and Fishe ies Resea ch Ins i u e, Te o Fishe ies Resea ch and Aquacul u e,
72210 Te o, Finland
(Recei ed 9 No embe 2006; accep ed 20 Janua y 2007)
Abs ac – The aquacul u e indus y is inc easingly eplacing ishmeal in eeds o ca ni o ous
ish wi h soybean meal (SBM). This die change p esen s a po en ial o geno ype-en i onmen
(G ×E) in e ac ions. We es ed whe he cu en salmonid b eeding p og ammes ha e alua e
and selec wi hin ishmeal die s also imp o e g ow h and efficiency on po en ial u u e SBM
die s. A o al o 1680 Eu opean whi e ish om 70 amilies we e ea ed wi h ei he ishmeal- o
SBM-based die s in a spli - amily design. Indi idual daily gain (DG), daily eed in ake (DFI)
and eed efficiency (FE) we e eco ded. T ai s displayed only weak G ×E in e ac ions as a i-
ances and he i abili ies did no diffe subs an ially be ween he die s, and c oss-die gene ic
co ela ions we e nea uni y. In bo h die s, DFI exhibi ed mode a e he i abili y and had e y
high gene ic co ela ion wi h DG whe eas FE had low he i abili y. P edic ed gene ic esponses
demons a ed ha selec ion o inc ease DG and FE on he ishmeal die lead o a ou able e-
sponses on he SBM die . Selec ion o FE based on an index including DG and DFI achie ed
a leas double FE gain e sus selec ion on DG alone. The e o e, cu en b eeding p og ammes
a e imp o ing he biological abili y o salmonids o use no el plan -based die s, and aiding he
aquacul u e indus y o educe ishmeal use.
eed efficiency /geno ype-en i onmen in e ac ion /selec ion /aquacul u e /Co egonus
la a e us
∗Co esponding au ho : che yl.quin on@m . i
A icle published by EDP Sciences and a ailable a h p://www.gse-jou nal.o g
o h p://dx.doi.o g/10.1051/gse:2007013
432 C.D. Quin on e al.
1. INTRODUCTION
The use o ishmeal in aquacul u e eeds has become a majo issue as he
global indus y con inues o g ow [10]. Fishmeal is mos ly p oduced om
wild-caugh small pelagic ish species, and is conside ed o be he supe io
p o ein sou ce o eleos ishes [18]. I is cu en ly he majo componen o
die s o a med ca ni o ous salmonids such as ainbow ou (Onco hynchus
mykiss) and A lan ic salmon (Salmo sala ) whose g owe die s ypically con-
ain 30–50% ishmeal (2000 s a is ics [35,45]). Howe e , se e al en i onmen-
al and economic easons exis o educing ishmeal use. These include he
decline and luc ua ion o wild s ocks ha es ed o ishmeal p oduc ion due
o o e ishing and na u al en i onmen al ac o s, ins abili y and p edic ed in-
c eases in ishmeal p ices, and nega i e consume pe cep ion o he sa e y o
ishmeal- ed p oduc s [10,34]. Acco dingly, eed manu ac u e s aim o eplace
ishmeal wi h al e na i es such as plan p oduc s. Soybean meal is one sou ce
o p o ein ha can be success ully subs i u ed in pa o ishmeal in ca ni o-
ous ishes’ eeds [18,23,36,41,43]. Some majo eed companies ha e pledged
o educe ishmeal use by a leas 50% by 2010, and i has been p edic ed ha
wi hin 10 yea s, ish die s will be close o 100% ege a ian [39].
Aquacul u e gene ic imp o emen p og ammes aim o complemen such
p oduc ion sys em changes. B eeding p og ammes exis o all majo a med
ca ni o ous ish species, and ha e achie ed imp o emen s in g ow h, eed e -
iciency, disease esis ance and p oduc quali y ai s [12]. Imp o ed eed effi-
ciency in pa icula is a goal o mos ish b eeding p og ammes, bu gene ic pa-
ame e es ima es o eed consump ion and efficiency a e a e [21,22,24,42].
This is because eco ding me hods o indi idual ish eed in ake ha e only
ecen ly been implemen ed in la ge-scale gene ics esea ch [21, 22, 42]; hus
no cu en b eeding p og ammes selec di ec ly o eed efficiency. Fu he -
mo e, all b eeding p og ammes e alua e and selec b oods ock based on pe -
o mance wi h cu en ishmeal-based die s. Fu u e plan -based die s may im-
pac ish b eeding p og ammes i geno ype-en i onmen (G ×E) in e ac ions
occu . G ×E in e ac ions may occu in he o m o geno ype e- anking ac oss
en i onmen s, o scaling effec s as indica ed by en i onmen al diffe ences in
ai gene ic a ia ion [9,30]. In a wo s -case scena io, supe io geno ypes on
a ishmeal die would ac ually be in e io on a plan -based die . This would be
shown by a nega i e gene ic co ela ion be ween die s [8]. In his case, cu en
selec ion on ishmeal die s would comp omise pe o mance on u u e plan -
based die s.
In his s udy, he impac o a no el soybean p o ein-based die on selec i e
imp o emen o g ow h, eed in ake and eed efficiency was in es iga ed in
B eeding ish o no el die efficiency 433
Eu opean whi e ish (Co egonus la a e us L.), a salmonid now a med com-
me cially in Finland. A b eeding p og amme has ecen ly been es ablished o
imp o e g ow h and eed efficiency in his species. To assess he deg ee o
which hese ai s a e affec ed by G ×E in e ac ions when ea ed wi h ei he
a adi ional ishmeal die o a po en ial u u e soybean meal-based die , we
es ed whe he hese ai s exp ess die -speci ic pheno ypic and gene ic a i-
a ion, and es ima ed ai gene ic co ela ions be ween he die s. To quan i y
he impac o G ×E in e ac ion on selec ion esponse, we p edic ed gene ic
changes occu ing on bo h die s in esponse o al e na i e s a egies o selec-
ion o g ow h o eed efficiency on ei he ishmeal o soybean meal-based
die s.
2. MATERIALS AND METHODS
2.1. Die o mula ions
Two p ac ical isoni ogenous and isocalo ic die s we e o mula ed (Tab. I).
In he ishmeal (FM) die , ishmeal supplied 100% o he die a y p o ein. This
die ep esen s a ypical comme cial die used in whi e ish a ming. In he soy-
bean meal (SBM) die , 50% o he die a y p o ein was eplaced wi h SBM-
de i ed p o ein. This die ep esen s a ealis ic u u e die . Me hionine, lysine
and phospho us supplemen s we e added o he SBM die o balance hese le -
els wi h he FM die . Ing edien s we e mixed in a Hoba - ype mixe , ex uded
(Clex al BC 45, FR) o 3.5 mm pelle s, e-d ied a 40–45 ◦C, op d essed wi h
ish oil and s o ed in a eeze un il use.
2.2. Popula ion and expe imen al design
Whi e ish in he expe imen o igina ed om he b eeding p og amme based
a he Te o s a ion o he Finnish Game and Fishe ies Resea ch Ins i u e
(FGFRI). The o iginal b oods ock was es ablished in 1998 by ma ing 50 wild
males and 150 wild emales o igina ing om he Kokemäki Ri e , Finland.
In Oc obe 2003, he gene a ion o he cu en expe imen was es ablished:
45 si es we e ma ed wi h 52 dams in a pa ial ac o ial design o c ea e 70 ull-
and hal -sib amilies. Each si e was ma ed o an a e age o 1.6 dams and each
dam o an a e age o 1.3 si es (bo h anges 1–2). Gene ic ela ionships among
si es and dams a e unknown.
Families we e kep sepa a e du ing incuba ion and ea ly ea ing. A he
eyed-egg s age (Janua y 2004), he amilies we e anspo ed o he FGFRI
434 C.D. Quin on e al.
Table I. Fo mula ion and analysed nu ien composi ion o he expe imen al ishmeal
and soybean meal die s.
Fishmeal (g·kg−1) Soybean meal (g·kg−1)
Ing edien
Fishmeal, capelina480 255
Soybean meal, de a ed, hulls 0 411
emo ed by sie ea
Fish oil, sandeela160 180
Whea glu ena20 20
Whea meala267.5 34.0
Whea s a ch, p egela inizeda52.5 52.5
Mono calcium phospha eb016
L-lysinec07
DL-me hionined0.0 4.5
Vi amin and mine al mixe20 20
Analysed composi ion
Wa e 23 27
C ude p o ein 398 404
C ude lipid 210 210
C ude ib e 10 22
N- ee ex ac s 359 337
G oss ene gy (MJ·kg−1) 22.6 22.3
Sou ces: aRaisio, FI; bKemi a, FI; cBASF, DE; dWelding GmbH, DE.
eAdded o supply (pe kg die ): e inol ace a e 8000 IU, cholecalci e ol 3000 IU, all-
ace-α- ocophe yl ace a e 300 IU, menadione sodium bisul i e 10 mg, hiamine HCl
21 mg, ibo la in 30 mg, calcium d-pan o hena e 92 mg, bio in 0.3 mg, olic acid
6 mg, i amin B12 0.04 mg, niacin 120 mg, py idoxineHCl 20 mg, asco bic acid (35%
S ay C) 900 mg, inosi ol 200 mg, manganese oxide (62% Mn) 100 mg, zinc oxide
(74% Zn) 200 mg, po assium iodide (76% I) 6 mg.
Calcula ed as 1000 – (wa e +p o ein +lipid + ib e).
Laukaa Resea ch S a ion and incuba ed (wa e 4–6 ◦C). A e ha ching (Feb u-
a y 2004), 100 o 150 ish pe amily we e held in indoo 150 L ib eglass anks
(wa e 13–15 ◦C) and ed comme cial d y die s (la ae: AgloNo se, EWOS
L d., NO; ju eniles: Nu a Pa and Royal Sil e , Raisio L d., FI). In June
2004, app oxima ely 40 andom indi iduals om each amily we e agged by
injec ing a passi e in eg a ed ansponde in o he body ca i y. To gi e all he
ish he same ini ial nu i ional en i onmen and o iden i y i agging ha med
B eeding ish o no el die efficiency 435
any ish, all ish we e ed wi h a 1:1 mix u e o he wo expe imen al die s o
ou weeks p io o he ial.
Twen y- ou agged ish pe amily we e andomly sampled o he die ial.
To cons uc a spli - amily design, each amily was i s andomly spli in o
wo g oups o be ea ed wi h he al e na i e die s. Each g oup was e enly dis-
ibu ed o e 6 ound 0.6 m3 eplica e anks (12 expe imen al anks in o al).
Consequen ly, he die ial began wi h a o al o 1680 ish, each ank con ain-
ing 140 ish ( wo ish om each amily).
The ial was conduc ed om July 29 o Oc obe 21, 2004, du ing which ish
ipled in weigh . Fish we e ed 6 h·d−1using bel eede s. Fish we e coun ed
and bulk-weighed biweekly and eeding was adjus ed acco ding o a e age
weigh and ank biomass. To ensu e excess eeding, he eed amoun s supplied
we e adjus ed o be 1.3 imes highe han p edic ed by Koskela [27]. Tanks
we e supplied wi h esh wa e (14.8–15.1 ◦C; low a e 8–16 L·min−1; ou -
le wa e O2le el >80% sa u a ion) and 24 h ligh was p o ided wi h ceiling
luo escen ubes. The expe imen al condi ions we e s anda dized o pe mi ac-
cu a e compa ison o amilies ac oss die s. This es en i onmen de ia es om
comme cial ci cums ances whe e ish a e ea ed in la ge ou doo ne cages un-
de na u ally a ying en i onmen al condi ions.
2.3. T ai s eco ded
Indi idual body weigh s we e eco ded o he nea es g a he beginning and
end o he ial (Tab. II). Daily eed in ake was measu ed by X- adiog aphy [17]
5 imes pe indi idual, wi h 2-week in e als be ween measu emen s. On eed
in ake measu emen days, ish we e ed wi h he same me hods and quan i y as
du ing no mal days, bu eed pelle s included lead glass beads (Ballo ini size
8.5; Jencons L d., UK) isible in X- ay. A e wa d, ish we e anaes he ized,
iden i ied, and X- ayed (Benne HFQ 3000P X- ay machine, US).
To ans o m he numbe o glass beads ed o he amoun o eed inges ed,
p edic i e eg ession models we e es ablished in a sepa a e s udy. Fo each
die , 16 samples o known weigh s we e aken om he bead-labelled pel-
le s and X- ayed. The numbe o beads p esen in each sample was coun ed,
and die -speci ic eg ession equa ions we e ob ained (R2=0.97−0.98). Feed
in ake (g) o each indi idual was p edic ed using hese equa ions om he
numbe o beads obse ed unde X- ay [17].
T ai s analysed we e indi idual daily weigh gain (DG), a e age daily eed
in ake (DFI) and eed efficiency (FE). Daily gain was calcula ed as he di -
e ence be ween he ini ial and he inal body weigh s, di ided by he numbe

436 C.D. Quin on e al.
o days in he ial (77–80 d, depending on he ank). Indi idual DFI was cal-
cula ed by i ing epea ed measu es analysis o a iance wi h measu emen
ime (1–5) as he andom epea ed ac o , and hen calcula ing leas squa es
means o each indi idual (MIXED p ocedu e, SAS
9.1; SAS
Ins . Inc.,
US). This was done sepa a ely wi hin each expe imen al ank. Feed efficiency
was calcula ed as he a io o DG o DFI.
2.4. S a is ical analysis o die effec s
Die effec s on he means o body weigh s, DG, DFI and FE we e es ed wi h
analysis o a iance (MIXED p ocedu e). S a is ical models included die as a
ixed effec , and eplica e ank nes ed wi hin die , amily, and die - amily in-
e ac ion as andom effec s. Fo all ai s, a iance due o andom expe imen al
ank- amily in e ac ion was ze o and hus was excluded. S anda d e o s and
deg ees o eedom o he F- es s o he ixed effec s we e calcula ed using
he Kenwa d and Roge op ion. Addi ional analyses we e pe o med o s an-
da dize DG and DFI o a common body weigh by adding ini ial weigh as a
eg ession co a ia e o he abo e model.
2.5. Gene ic analysis
In o de o examine G ×E in e ac ions, obse a ions eco ded unde each
die ea men we e ea ed as sepa a e ai s. Fo ins ance, DG eco ded on FM
(DGFM) and SBM die s (DGSBM) we e de ined as diffe en ai s.
Pheno ypic and gene ic pa ame e s o DG, DFI and FE we e es ima ed us-
ing mul iple- ai animal models wi h DMU so wa e, applying es ic ed max-
imum likelihood and a e age in o ma ion me hods [31]. Models con ained ex-
pe imen al ank as a ixed effec , and ull-sib amily, indi idual gene ic and
esidual e o as andom effec s. The indi idual gene ic effec included addi i e
gene ic effec s and pa s o po en ial dominance effec s. The andom ull-sib
amily effec con ained (co) a iance due o common incuba ion and ea ly ea -
ing o ull sibs, as well as pa s o po en ial dominance (co) a iances. Residual
co a iances be ween ai s measu ed in diffe en die s we e se o ze o. S an-
da d e o s o (co) a iances we e ob ained by a i s -o de Taylo se ies expan-
sion o he a e age in o ma ion ma ix o he es ima ed (co) a iances. He i-
abili y (h2) was calcula ed as he a io o gene ic a iance o o al pheno ypic
a iance. Full-sib amily effec (c2) was calcula ed as he a io o ull-sib am-
ily a iance o o al pheno ypic a iance. When calcula ing ai co ela ions
wi hin die s, we did no es ima e co ela ions be ween FE and i s componen
B eeding ish o no el die efficiency 437
ai s because his p ac ice can be conside ed s a is ically ague due o au o-
co ela ion effec s. To aid he eade in pe cei ing he esul s, we do howe e
p esen die means, a iances and he i abili ies o FE.
Daily gain and eed in ake a e commonly exp essed ela i e o body weigh .
Thus, he analysis o DG and DFI was also ca ied ou using a model ha
included ini ial body weigh as a co a ia e. Fo his s udy, we use he e ms
“absolu e” and “ ela i e” o e e o ai s analysed wi hou and wi h ini ial
body weigh as a co a ia e, espec i ely.
2.6. P edic ion o gene ic esponses o selec ion
Selec ion index calcula ions we e used o p edic esponses in DG, DFI and
FE on bo h FM and SBM die s o pheno ypic selec ion [15]. Selec ion was no
p ac iced di ec ly o FE because when selec ing on a a io, gene ic changes
in he indi idual componen ai s a e e y difficul o con ol [14]. Fou se-
lec ion s a egies we e compa ed: (a) selec ion o DGFM; (b) selec ion o
maximum FEFM whe e inc eased DGFM and dec eased DFIFM we e selec ed
o simul aneously and selec ion index weigh s we e se o ob ain maximum
gene ic change in FEFM; (c) selec ion o DGSBM; and (d) selec ion o maxi-
mum FESBM whe e inc eased DGSBM and dec eased DFISBM we e selec ed o
ob ain maximum gene ic change in FESBM. S a egy (a) is compa able o cu -
en aquacul u e b eeding p og ammes ha only selec o g ow h a e. S a e-
gies (c) and (d) p edic he effec s o selec ion pe o med on he po en ial u u e
soybean die s.
Di ec and co ela ed gene ic esponses o one gene a ion o selec ion we e
calcula ed by R=i(bG)(bPb)−1
2,whe eRis he ec o o esponses, i is
in ensi y o selec ion (se o 1), bis he ec o o ela i e index weigh s which
sum o 1, Gis he gene ic co a iance ma ix and Pis he pheno ypic co a i-
ance ma ix. Gand Pwe e esul s om he 4- ai gene ic pa ame e es ima ion
model o DGFM,DFI
FM,DG
SBM and DFISBM desc ibed abo e. The pa ame-
e s o FE we e no needed because all esul s o his a io can be p edic ed
om i s componen ai s DG and DFI [14].
To gene a e he al e na i e selec ion s a egies, ela i e index weigh s we e
modi ied as ollows. To ob ain s a egy (a), he index weigh o DGFM was
se o 1, while he weigh s o he h ee emaining ai s we e ze o. Simila ly,
o s a egy (c), he index weigh o DGSBM was se o one. Fo he selec ion
s a egies o maximise FE (b and d), he maximum gene ic esponse in FE on
bo h die s was ob ained when hal o he index weigh was on DG and hal
agains DFI.
438 C.D. Quin on e al.
Table II. Leas squa es means (±s.e.), and s a is ical es s o he die effec o ai s
eco ded on ishmeal and soybean meal die s. Denomina o deg ees o eedom (dd )
shown; all nume a o d =1.
Fishmeal Soybean meal Tes s a is ics
T ai nmean ±s.e. nmean ±s.e. dd FP
Ini ial body weigh (g) 818 40.9 ±0.85 829 40.6 ±0.87 10.0 0.78 0.3984
Final body weigh (g) 765 131.2 ±2.38 768 125.8 ±2.24 9.7 9.36 0.0125
Daily gain (g·d−1) 765 1.145 ±0.0212 768 1.091 ±0.0201 9.5 20.0 0.0014
Daily eed in ake (g·d−1) 817 0.948 ±0.0236 829 1.027 ±0.0285 9.6 7.23 0.0235
Feed efficiency 765 1.208 ±0.0150 768 1.078 ±0.0240 8.5 23.2 0.0011
To calcula e gene ic esponse in FE, mean FE was i s calcula ed o each
die om he da a, i.e., be o e selec ion. Then, gene ic esponses o selec ion
we e calcula ed o DG and DFI, and he new mean FE was calcula ed om
hese [22].
3. RESULTS
3.1. Die diffe ences
As expec ed, whi e ish pe o med be e on he FM die . Fish ed he FM die
had signi ican ly highe inal body weigh and DG, lowe DFI, and be e FE
han hose ed he SBM die (Tab. II). These die diffe ences emained ( esul s
no shown) when DG and DFI we e s anda dised by including ini ial body
weigh as a co a ia e (P<0.0001) in he s a is ical models.
3.2. Pheno ypic and gene ic (co) a ia ion
The e was no e idence o die diffe ences in ai pheno ypic o gene ic
a ia ion. In bo h die s, DG and DFI showed mode a e he i abili y whe eas
FE showed e y low he i abili y ha did no diffe om ze o (Tab. III). Fo a
gi en ai eco ded on bo h die s, he diffe ences in he i abili ies be ween die s
we e ela i ely small (mean absolu e diffe ence =0.07) and con idence in e -
als o e lapped conside ably. Because he i abili ies may emain cons an e en
i he unde lying gene ic and esidual a ia ions change, die -speci ic coeffi-
cien s o a ia ion we e also calcula ed o each ai . Die diffe ences be ween
coefficien s o pheno ypic a ia ion o pai s o ai s we e small (mean abso-
lu e diffe ence =1.1%) as well. Coefficien s o gene ic a ia ion o absolu e
and ela i e DG, DFI and FE we e only sligh ly highe on he FM die (mean
diffe ence =2.9%).
B eeding ish o no el die efficiency 439
Table III. Pheno ypic a iance (VP), coefficien s o gene ic (CVG) and pheno ypic
a ia ion (CVP), he i abili y (h2±s.e.) and ull-sib amily effec (c2±s.e.) o absolu e
and ela i e (Rel.) ai s measu ed on ish ea ed wi h ishmeal and soybean meal die s.
Die ,T ai VPCVGCVPh2±s.e.c2±s.e.
Fishmeal
Daily gain 0.134 16.2 31.9 0.26 ±0.18 0.13 ±0.09
Daily eed in ake 0.095 15.6 32.5 0.23 ±0.15 0.08 ±0.07
Feed efficiency 0.044 4.3 17.4 0.06 ±0.10 0.04 ±0.05
Rel. daily gain 0.090 17.0 26.1 0.42 ±0.17 0.04 ±0.07
Rel. daily eed in ake 0.068 15.6 27.6 0.32 ±0.14 0.03 ±0.06
Soybean meal
Daily gain 0.100 13.1 29.0 0.20 ±0.15 0.08 ±0.08
Daily eed in ake 0.104 12.8 31.3 0.17 ±0.15 0.09 ±0.08
Feed efficiency 0.033 4.3 16.9 0.07 ±0.11 0.07 ±0.06
Rel. daily gain 0.076 14.3 25.2 0.32 ±0.14 0.04 ±0.06
Rel. daily eed in ake 0.081 12.6 27.8 0.21 ±0.15 0.10 ±0.08
Gene ic co ela ions be ween he same ai s eco ded in each die indica ed
e y li le e- anking o amilies ac oss he die s. Gene ic co ela ions (±s.e.)
be ween he die s o DG (0.97±0.21), DFI (0.93±0.28) and FE (1.00±0.95)
we e all close o uni y. The la ge s anda d e o o he FE co ela ion was
likely caused by he low he i abili y o his ai . Gene ic co ela ions be ween
he die s o ela i e DG (0.99 ±0.13) and ela i e DFI (0.97 ±0.22) we e also
e y high.
Wi hin-die ai co ela ions we e simila in bo h die s. Pheno ypic ( P)and
gene ic co ela ions ( G±s.e.) be ween DG and DFI we e high and posi i e
on bo h FM ( P=0.88, G=0.97 ±0.05) and SBM die s ( P=0.86, G=
0.93 ±0.10). Co ela ions be ween ela i e DG and DFI (FM P=0.82, G=
0.97 ±0.05; SBM P=0.82, G=0.96 ±0.10) we e simila o hose o
absolu e ai s.
3.3. P edic ion o selec ion esponses
The selec ion index calcula ions showed ha cu en selec ion on ishmeal
die s will lead o s ong a ou able co ela ed gene ic changes o pe o -
mance on SBM die s (Fig. 1). Li le diffe ence was obse ed be ween he di-
e s o gene ic esponse o selec ion. We emphasize ha al hough he gene ic
and pheno ypic pa ame e poin es ima es used in he selec ion index diffe ed
446 C.D. Quin on e al.
[32] McDonagh M.B., He d R.M., Richa dson E.C., Oddy V.H., A che J.A., A hu
P.F., Mea quali y and he calpain sys em o eedlo s ee s ollowing a single
gene a ion o di e gen selec ion o esidual eed in ake, Aus . J. Exp. Ag ic. 41
(2001) 1013–1021.
[33] McPhee C.P., Ke J.C., Came on N.D., Pe i-pa um pos u e and beha iou o
gil s and he loca ion o hei pigle s in lines selec ed o componen s o efficien
lean g ow h, Appl. Anim. Beha . Sci. 71 (2001) 1–12.
[34] Naylo R.L., Goldbu g R.J., P ima e a J.H., Kau sky N., Be e idge M.C.M.,
Clay J., Folke C., Lubchenco J., Mooney H., T oel M., Effec o aquacul u e on
wo ld ish supplies, Na u e 405 (2000) 1017–1024.
[35] New M.B., Wijks öm U.N., Use o ishmeal and ish oil in aqua eeds: u -
he hough s on he ishmeal ap, FAO Fishe ies Ci cula C975, Food and
Ag icul u e O ganiza ion o he Uni ed Na ions, Rome, 2002.
[36] Olli I.J., K ogdahl Å., Nu i i e alue o ou soybean p oduc s in die s o ain-
bow ou (Onco hynchus mykiss, Walbaum) ea ed in esh wa e , Ac a Ag ic.
Scand. A 44 (1994) 185–192.
[37] Pal i Y., Sil e s ein J.T., Wieman H., Phillips J.G., Ba ows F.T., Pa sons J. E.,
E alua ion o amily g ow h esponse o ishmeal and glu en-based die s in ain-
bow ou (Onco hynchus mykiss), Aquacul u e 255 (2006) 548–556.
[38] Pi ch o d W.S., Gene ic imp o emen o eed efficiency o bee ca le: wha
lessons can be lea n om o he species? Aus . J. Exp. Ag icul . 44 (2004)
371–382.
[39] Powell K., Fish a ming: Ea you eg, Na u e 426 (2003) 378–379.
[40] Re s ie S., S o ebakken T., Roem A.J., Feed consump ion and con e sion in
A lan ic salmon (Salmo sala ) ed die s wi h ish meal, ex ac ed soybean meal
o soybean meal wi h educed con en o oligosaccha ides, ypsin inhibi o s,
lec ins and soya an igens, Aquacul u e 162 (1998) 301–312.
[41] Re s ie S., S o ebakken T., Bae e jo d G., Roem A.J., Long- e m p o ein and
lipid g ow h o A lan ic salmon (Salmo sala ) ed die s wi h pa ial eplacemen
o ish meal by soy p o ein p oduc s a medium o high lipid le el, Aquacul u e
193 (2001) 91–106.
[42] Sil e s ein J.T., Boswo h B.G., Waldbiese G.C., Wol e s W.R., Feed in ake
in channel ca ish: is he e a gene ic componen ? Aquacul . Res. 32 s1 (2001)
199–205.
[43] S o ebakken T., Re s ie S., Ruy e B., Soy p oduc s as a and p o ein sou ces
in ish eeds o in ensi e aquacul u e, in: D ackley J.K. (Ed.), Soy in Animal
Nu i ion, Fede a ion o Animal Science Socie ies, Sa oy, 2000, pp. 127–170.
[44] Syl én S., Rye M., Simiane H., In e ac ion o geno ype wi h p oduc ion sys em
o slaugh e weigh in ainbow ou (Onco hynchus mykiss), Li es . P od. Sci.
28 (1991) 253–263.
[45] Tacon A.G.J., Use o ish meal and ish oil in aquacul u e: a global pe spec i e,
Aqua . Resou . Cul . De . 1 (2004) 3–14.