RESEARCH ARTICLE
Gene ics o G ow h Reac ion No ms in
Fa med Rainbow T ou
Panya Sae-Lim
1
*, Han Mulde
2
, Bja ne Gje de
1
, Heikki Koskinen
3
, Ma ie Lillehamme
1
,
An i Kause
4
1Aquacul u e and Gene ics, No ima, Oslo eien 1, Ås, No way, 2Animal B eeding and Genomics Cen e,
Wageningen Uni e si y, Wageningen, he Ne he lands, 3Aquacul u e Uni , Na u al Resou ces Ins i u e
Finland, Te o, Finland, 4Biome ical Gene ics, Na u al Resou ces Ins i u e Finland, Jokioinen, Finland
*panya.sae-lim@no ima.no
Abs ac
Rainbow ou is a med globally unde di e se uncon ollable en i onmen s. Fish wi h low
mac oen i onmen al sensi i i y (ES) o g ow h is impo an o h i e and g ow unde hese
uncon ollable en i onmen s. The ES may e ol e as a co ela ed esponse o selec ion o
g ow h in one en i onmen when he gene ic co ela ion be ween ES and g ow h is nonze o.
The aims o his s udy we e o quan i y addi i e gene ic a iance o ES o body weigh
(BW), de ined as he slope o eac ion no m ac oss b eeding en i onmen (BE) and p oduc-
ion en i onmen (PE), and o es ima e he gene ic co ela ion (
g(in , sl)
) be ween BW and
ES. To es ima e he i able a iance o ES, he cohe i abili y o ES was de i ed using selec-
ion index heo y. The BW eco ds om 43,040 ainbow ou pe o ming ei he in eshwa e
o seawa e we e analysed using a eac ion no m model. High addi i e gene ic a iance
o ES (9584) was obse ed, in e ing ha gene ic changes in ES can be expec ed. The
cohe i abili y o ES was ei he -0.06 (in e cep a PE) o -0.08 (in e cep a BE), sugges ing
ha BW obse a ion in ei he PE o BE esul s in low accu acy o selec ion o ES. Ye , he
g(in , sl)
was nega i e (-0.41 o -0.33) indica ing ha selec ion o BW in one en i onmen is
expec ed o esul in mo e sensi i e ish. To a oid an inc ease o ES while selec ing o BW,
i is possible o ha e equal gene ic gain in BW in bo h en i onmen s so ha ES is main ained
s able.
In oduc ion
The pe o mance o o ganisms is in luenced by he su ounding en i onmen al condi ions,
leading o pheno ypically plas ic esponses o en i onmen al changes. Such plas ic esponses
ha e been obse ed, o example, as adap i e plas ici y in he neck ee h o Daphnia (wa e
leas) which de elops as a p o ec i e esponse o he chemical cues o a p eda o y Chaobo us
p esen in he wa e [1]. In ish species, pheno ypic plas ici y has been explo ed especially om
ecological and e olu iona y poin s o iew. The e is e idence o gene ic basis o pheno ypic
plas ici y, o example in salmonids [2], T inidadian guppies (Poecilia e icula a)[3,4] and
pup ishes (Cyp inodon ne adensis)[5].
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 1/17
OPEN ACCESS
Ci a ion: Sae-Lim P, Mulde H, Gje de B, Koskinen
H, Lillehamme M, Kause A (2015) Gene ics o
G ow h Reac ion No ms in Fa med Rainbow T ou .
PLoS ONE 10(8): e0135133. doi:10.1371/jou nal.
pone.0135133
Edi o : Gen Hua Yue, Temasek Li e Sciences
Labo a o y, SINGAPORE
Recei ed: Feb ua y 6, 2015
Accep ed: July 18, 2015
Published: Augus 12, 2015
Copy igh : © 2015 Sae-Lim e al. This is an open
access a icle dis ibu ed unde he e ms o he
C ea i e Commons A ibu ion License, which pe mi s
un es ic ed use, dis ibu ion, and ep oduc ion in any
medium, p o ided he o iginal au ho and sou ce a e
c edi ed.
Da a A ailabili y S a emen : Na u al Resou ces
Ins i u e Finland (Luke) own he da a unde lying his
pape . Please send eques s o he da a o pe i.
heinimaa@luke. i.
Funding: This s udy is unded by No wegian
Resea ch Council (NRC: 234144/E40), h p://www.
o sknings ade .no/en/Home_page/1177315753906.
The unde s had no ole in s udy design, da a
collec ion and analysis, decision o publish, o
p epa a ion o he manusc ip .
Compe ing In e es s: The au ho s ha e decla ed
ha no compe ing in e es s exis .
Among animal b eede s, pheno ypic plas ici y is e med mac oen i onmen al sensi i i y
(ES) [6]. I has been o in e es o animal b eede s because o i s connec ion wi h animal’s pe -
o mance ac oss en i onmen s [7,8] and o he obus ness and wel a e o animals [9]. Fo a
geno ype, such as a clone, amily, popula ion, o a species, mac oen i onmen al sensi i i y
can be de ined by i s slope o eac ion no m ac oss en i onmen s. Assuming a linea eac ion
no m, he deg ee o mac oen i onmen al sensi i i y can be quan i ied by he eg ession slope
o a geno ype's pe o mance, such as g ow h, agains an en i onmen al g adien [10–12].
Rainbow ou Onco hynchus mykiss (Walbaum 1792) is one o he main ish species a med
unde di e se en i onmen al condi ions ac oss con inen s. Rapid g ow h is one o he mos
impo an ai s o p o i able ou a ming. Howe e , ish may no be able o main ain high
g ow h when ea ing condi ions a e subop imal. The e o e, a mo e obus ish wi h high s abil-
i y o g ow h is impo an o h i e unde a iable en i onmen al condi ions. To quan i y
he po en ial o changing mac oen i onmen al sensi i i y h ough selec ion, an es ima e o
gene ic a iance in mac oen i onmen al sensi i i y is equi ed. The gene ic a ia ion in he
mac oen i onmen al sensi i i y is known as non-pa allel eac ion no ms, causing geno ype-
by-en i onmen in e ac ion (GxE) [13]. E idences o GxE in g ow h o ainbow ou ha e
been epo ed [14–20]. Howe e , so a mos s udies use mul i- ai model in which GxE is
quan i ied as he gene ic co ela ion be ween he eco ds o he same ai measu ed in di e en
en i onmen s. Such gene ic co ela ion exp esses he magni ude o e- anking o amilies wi h
espec o hei b eeding alue, bu i does no p o ide an explana ion on how mac oen i on-
men al sensi i i y can e ol e ac oss en i onmen s. The concep o mac oen i onmen al sensi-
i i y has ne e been applied o b eeding in aquacul u e be o e.
In Finland, he na ional b eeding p og amme o ainbow ou b eeds especially o
imp o ed g ow h pe o mance in comme cial p oduc ion en i onmen a he Bal ic Sea [21].
Howe e , he s ock is also ea ed in inland eshwa e p oduc ion en i onmen s, and expo ed
o Russia and Asia whe e he p oduc ion en i onmen di e s om Finland subs an ially.
Hence, he mac oen i onmen al sensi i i y is conside ed as an impo an ai .
The aims o his s udy we e wo- old. Fi s ly, we quan i y he gene ic a iance o ES,
de ined as he slope o eac ion no m ac oss seawa e and eshwa e p oduc ion en i onmen s
in Finland, using a eac ion no m model. Secondly, o s udy whe he selec ion o as g ow h
in one en i onmen will change ES, we es ima e he gene ic co ela ion be ween ES and body
weigh in one en i onmen , de ined as he in e cep o eac ion no m. In addi ion, we de i ed
he cohe i abili y o ES. Al hough, he gene ic co a iance ma ix om eac ion no m and
mul i- ai models is in e changeable [7,22–23], he eac ion no m model is chosen as he
me hod in his s udy because i p o ides he gene ic pa ame e s o ES and body weigh di ec ly
wi hou in e changing. To be able o compa e ou esul s wi h p e ious s udies, we exploi his
in e changeable p ope y o calcula e gene ic a iance in ES and i s gene ic co ela ion wi h
in e cep in aquacul u e GxE s udies ha all ha e used a mul i- ai model.
Ma e ials and Me hods
E hics S a emen
All p ocedu es in ol ing animals we e app o ed by he animal ca e commi ee o he Na u al
Resou ces Ins i u e Finland. To enhance animal wel a e and amelio a e su e ing du ing all ish
handling, he ish we e always i s anaes he ized using MS-222.
Da a sou ce
All ish used in his s udy we e ob ained om he Finnish na ional b eeding p og amme.
B eeding candida es a e held a he Te o ish a m in cen al Finland ( eshwa e nucleus
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 2/17
s a ion) and he sibs o he b eeding candida es a e es ed a comme cial sea s a ions loca ed a
he Bal ic Sea. The pheno ypic da a had 53,638 eco ds o body weigh a agging om ou
yea classes and belonged o wo subpopula ions, one wi h yea classes o 1996 and 1999 and
he o he wi h 1997 and 2000. Bo h o hese subpopula ions we e es ablished om he pa en s
o yea class 1993. Si es we e ma ed o dams using ei he pa e nal nes ed ma ing o pa ial ac-
o ial ma ing designs. Each yea class consis ed o 94 o 197 ull-sib amilies es ablished om
he ma ing o 37 o 95 si es wi h 79 o 129 dams. A e ha ching, inge lings om he same
ull-sib amily we e held in one o mo e amily anks un il he inge lings eached agging size
(mean body weigh o app oxima ely 50 g).
Du ing he agging, ull-sibs om each amily we e andomly sampled and di ided in o wo
o h ee ba ches ha we e ea ed ei he a he eshwa e nucleus s a ion (de ined as “b eeding
en i onmen ”o BE) o a one o wo seawa e s a ions (de ined as “p oduc ion en i onmen ”
o PE) a he Bal ic Sea. When he ish we e 2-yea -old, hey we e indi idually weighed a BE
( ai : BW
BE
, in g) and PE ( ai : BW
PE
, in g) s a ions. The o al numbe o eco ds analysed
was 22,175 indi iduals o BW
BE
and 20,865 indi iduals o BW
PE
(Table 1). The a e age
BW
BE
(SD) and BW
PE
(SD) we e 1094 (363.9) g. and 1050.0 (334.5) g, espec i ely. The pedi-
g ee was aced back o he pa en s ( he ounde s) o he 1990 yea class. The ances o s back o
he ounde popula ion o he 1990 yea class we e included in he pedig ee.
Gene ic Analysis
Reac ion no m model. A eac ion no m model was used o es ima e gene ic (co) a iance
o ES ( eg ession slope) o body weigh s eco ded on 2-yea -old ish. (Co) a iance compo-
nen s o all analyses we e es ima ed using es ic ed maximum likelihood in ASReml e sion
3.0 [24]. App oxima e s anda d e o s we e calcula ed wi h ASReml ollowing Fishe e al.
[25].
In addi ion o he analysis o obse ed body weigh s, he analysis was also pe o med wi h
log- ans o med body weigh s. This was o es he hypo hesis ha gene ic a iance in ES may
be in luenced by a scale e ec , ypically obse ed o body weigh in ish species, i.e., inc easing
a iance o BW wi h inc easing mean o BW. Fo ins ance, pa allel eac ion no ms o geno-
ypes (no gene ic a iance o slopes) wi h di e en in e cep s a e in ac ansla ed in o di e -
en magni udes o sensi i i y i change in body weigh is calcula ed as a pe cen age change in
he ai mean. The log- ans o ma ion educes such scale e ec [26].
Table 1. Popula ion s uc u e.
Subpopula ion I Subpopula ion II
1996 1999 1997 2000
Popula ion s uc u e
Si es, dams 57, 129 37, 94 65, 79 95, 121
Si es pe dam, mean ( ange) 1.00 (1–1) 1.00 (1–1) 2.41 (1–3) 1.63 (1–3)
Dams pe si e, mean ( ange) 2.26 (1–4) 2.54 (1–4) 2.93 (1–5) 2.06 (1–5)
Full-sib amilies, amily anks 129, 129 94, 135 191, 259 197, 197
Numbe o fish wi h eco ds
F eshwa e nucleus s a ion 4994 3084 8099 5998
Fish pe ull-sib amily 38.7 32.8 42.4 30.4
Seawa e s a ion 2573 2442 8351 7499
Fish pe ull-sib amily 19.9 26.0 43.7 38.1
doi:10.1371/jou nal.pone.0135133. 001
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 3/17
The eac ion no m model was:
yhijklmn ¼bin þbslXhþYC SITE SEX MATijklþ
am;in þam;slXhþcn;in þcn;slXhþehijklmn;ð1Þ
whe e yis he obse a ion (body weigh o log body weigh ) o he m
h
indi idual. The β
in
and
β
sl
a e he fixed eg ession coe ficien s o he popula ion in e cep (in ) and slope (sl), espec-
i ely. The X
h
is he eg esso o he en i onmen s (X
h
= 0 and 1) in which he in e cep was
placed a X
h
= 0. The fixed e ec YC×SITE×SEX×MAT was included in he model o co ec
o he in e ac ion o he i
h
yea class (YC,i= 1996, 1997, 1999, 2000), he j
h
es s a ion
(SITE,j= 1: BE, 2 o 4: sea- es s a ions), he k
h
sex (SEX,k= 1: male, 2: emale, o 9:
unknown), and he l
h
ma u i y (MAT,l= 2: ma u e a 2-yea -old, 3: ma u e a 3-yea -old, 9:
unknown). The ais he andom addi i e gene ic e ec o in e cep (in ) and slope (sl) o eac-
ion no m,
ain
asl
"#
~ MVN[0,AG
RN
], whe e Ais he addi i e gene ic ela ionship ma ix,
G
RN
is gene ic co a iance ma ix om he eac ion no m model, and MVN is mul i a ia e no -
mal dis ibu ion. The c
n
is he andom ull-sib ank e ec (unique numbe s in di e en yea
classes), explaining an e ec common o ull-sibs o he han addi i e gene ics ( ank e ec due
o he sepa a e ea ing o he amilies p io o agging and non-addi i e gene ic e ec ),
cin
csl
"#
~ MVN[0,IC
RN
], whe e C
RN
is common en i onmen al co a iance ma ix and Iis he iden-
i y ma ix. The e~N(0,
Is2
e1
0
0Is2
e2
"#
) is he andom esidual e ec o an animal min en i-
onmen hwi h o each en i onmen a di e en esidual a iance. The si e’s and o sp ing’s
es ima ed b eeding alues (EBVs) o slope ob ained om model (1) we e used o illus a e he
ange o addi i e gene ic alues o slope a ailable o selec ion.
The magni ude and he sign o a gene ic co ela ion be ween he slope and in e cep , and
gene ic a iance o he in e cep , can change depending on which en i onmen he in e cep
is de ined. Hence, he model was un wice, ei he wi h PE (Xh1=0)o BE(Xh2= 0) as he
in e cep en i onmen . To illus a e he co a iance be ween EBVs o slope and in e cep , si e’s
EBVs o he slope when he in e cep we e placed a PE we e anked and a o al o fi een si es
wi h he highes , close o ze o, and he lowes EBVs o he slope we e chosen o plo ing he
eac ion no m.
Gene ic cha ac e is ics o mac oen i onmen al sensi i i y. The s ic sense o he i abil-
i y o ES is he a io be ween addi i e gene ic a iance o a slope o pheno ypic a iance o he
slope. Due o he lack o pheno ypic a iance o he slope, i is no possible o calcula e he he i-
abili y o ES. Th ee al e na i e pa ame e s we e used he e o desc ibe gene ic cha ac e is ics
o ES. Following Scheine [27], he i abili y o ES (h2
ES) was calcula ed as:
h2
ES ¼s2
GxE
s2
P;ANOVA
;ð2Þ
whe e s2
GxE is geno ype by en i onmen in e ac ion a iance. The s2
GxE is equal o he s anda dized
addi i e gene ic a iance o he slope (^
s2
a;sl ^
s2
X) ha is independen om di e en scales o an
en i onmen al a iable (X). The ^
s2
Xis he a iance o X[28], i.e., ^
s2
Xis 0.5 in his s udy, as he pos-
sible alues o Xin his s udy a e 0 and 1. The s2
P;ANOVA is o al pheno ypic a iance ac oss en i on-
men s, in Scheine 's app oach calcula ed om an analysis o a iance (ANOVA) [27]. Because
s2
P;ANOVA may no be a ailable om he eac ion no m model, we adop ed Scheine ’s he i abili y by
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 4/17
eplacing he s2
P;ANOVA by ^
s2
P;To al ¼ðnBE 1Þ^
s2
PBW;BE þðnPE 1Þ^
s2
PBW;PE þnBE nPEðWBE WPE Þ2=ðnBE þnPE Þ
ðnBE þnPE 1Þ

,whe enis
he numbe o animals wi h a eco d o an animal ai , ^
s2
PBW is pheno ypic a iance o he ai
and Wis he mean. No e ha nei he s2
P;ANOVA no ^
s2
P;To al is he pheno ypic a iance o he en i-
onmen al sensi i i y. Thus, h2
ES is mo e desc ip i e a he han a p edic i e pa ame e [28]. Fu -
he mo e, he defini ion o he i abili y in Eq (2) does no coincide wi h he he i abili y being he
eg ession o b eeding alue on pheno ype.
Because he exp ession in Eq (1) is no p edic i e o esponse o selec ion, we de ined a sec-
ond measu e called cohe i abili y ollowing selec ion index p inciples. The pheno ype (P)o an
indi idual ha includes eac ion no m pa ame e s can be de ined as (1). We assume no co a i-
ances among a,c, and e, because he e is no ela ionship among a,c, and e. The pheno ypic a -
iance (s2
P) o a ai is:
s2
P¼s2
a;in þ2Xsa;in ;sl þX2s2
a;sl þs2
c;in þ2Xsc;in ;sl þX2s2
c;sl þs2
eð3Þ
Es ima ed addi i e gene ic e ec o slope (^
asl) is equal o he eg ession on Pde ia ed om
he popula ion mean, o ^
asl ¼bðPmÞ. The eg ession coe ficien (b) o he b eeding alue o
slope on pheno ype is:
b¼co ðasl;PÞ
s2
P¼sa;in ;sl þXs2
a;sl
s2
Pð4Þ
The bin Eq (4)is“cohe i abili y” o ES. The e m cohe i abili y is used ins ead o he i abil-
i y because cohe i abili y de ines he inhe i ance o associa ion be ween ES and BW in one
en i onmen . The addi i e gene ic co a iance be ween in e cep and slope changes along he
le els o he en i onmen al ac o (X). Hence, he magni ude and sign o cohe i abili y is
dependen on he alue o X. A nega i e cohe i abili y is possible i he absolu e o −σ
a,in , sl
is
g ea e han Xs2
a;sl and/o Xis nega i e and absolu e Xs2
a;sl is g ea e han σ
a,in , sl
. The sign o
he cohe i abili y explains he change in co ela ed esponse o ES when mass selec ion o
highe pheno ypic alues is pe o med. When in e cep is placed o he en i onmen , in which
selec ion is p ac ised on P,Xbecomes ze o, leading o:
b¼sa;in ;sl
s2
a;in þs2
c;in þs2
e;in ¼sa;in ;sl
s2
Ph
;ð5Þ
whe e s2
Phis he pheno ypic a iance o BW in he selec ion en i onmen h. This alue is equal
o s2
P;in desc ibed abo e.
Finally, o unde s and he po en ial gene ic esponse in ES, he accu acy (
IH
) o p edic ing
b eeding alue o ES when a selec ion c i e ion is BW in one o he en i onmen s is equal o:
IH ¼ffiffiffiffiffiffiffi
bg
s2
a;sl
s¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðsa;in ;sl þXs2
a;slÞ2
s2
Ps2
a;sl
s¼sa;in ;sl þXs2
a;sl
sPsa;sl
;ð6Þ
whe e g is sa;in ;sl þXs2
a;sl. The Eq (6) is equi alen o he equa ion de i ed by Kolmodin and
Bijma [29].
Cohe i abili y o ES changes depending on a deg ee and o ms o GxE. To demons a e he
ela ionship be ween cohe i abili y and GxE in bo h o ms, i.e. geno ype e- anking and
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 5/17

he e ogenei y o a iances, Eq (5) is ea anged as (see S1 Appendix):
b¼
sa;ðE1;E2Þs2
a;in
s2
P;in ¼ gsa;E1sa;E2s2
a;in
s2
P;in
;ð7Þ
whe e
g
is he gene ic co ela ion be ween ai s measu ed in wo di e en en i onmen s
(E
1
and E
2
). The
g
di e en om uni y indica es a p esen o geno ype e- anking. The
g¼sa;ðE1;E2Þ
sa;E1sa;E2
, whe e sa;ðE1;E2Þ,sa;E1and sa;E2a e addi i e gene ic co a iance and addi i e gene ic
s anda d de ia ion in E
1
and E
2
, espec i ely.
Assume ha he e is no he e ogenei y o addi i e gene ic a iances (s2
a;E1=s2
a;E2=s2
a;in )
and s2
P;E1=s2
P;E2=s2
P;in ,Eq(7) is equal o:
b¼h2 gh2¼h2ð g1Þð8Þ
Eq (8) shows eg ession o cohe i abili y on geno ype e- anking, whe e he slope and in e -
cep is equal o he h
2
o a ai . I he h
2
= 0.3 and
g
a ies om -1 o 1, he magni ude o
cohe i abili y, ega dless o he sign inc eases when he gene ic co ela ion di e s om he
uni y and he cohe i abili y is a maximum when he gene ic co ela ion equals -1. Placing he
in e cep (X= 0) in ei he E
1
o E
2
does no in luence he magni ude o he cohe i abili y
(Fig 1).
Assume he e is he e ogenei y o addi i e gene ic a iances (s2
a;E16¼ s2
a;E2), Eq (7) is equal
o:
b¼ðhE1hE2Þ gh2
in ð9Þ
In con as o Eq (8), Eq (9) shows ha he e ogenei y o addi i e gene ic a iances esul s in
di e en alues o cohe i abili y because h
2in
changes, depending on he in e cep (X=0)
which is placed in ei he E
1
o E
2
as shown in Fig 2.
Calcula ion o eac ion no m pa ame e s
Fo he in e cep o eac ion no ms (body weigh a he in e cep en i onmen ), he i abili y
(h2
in ) and common en i onmen al e ec (c2
in ) we e calcula ed as: h2
in ¼^
s2
a;in =^
s2
P;in ,
c2
in ¼^
s2
c;in =^
s2
P;in ,whe e^
s2
P;in is equal o ^
s2
a;in þ^
s2
c;in þ^
s2
e;in , which is he pheno ypic a iance
o BW in he in e cep en i onmen when X= 0. Fo he slope o eac ion no ms, he h2
ES was
calcula ed using Eq (1) while he cohe i abili y was calcula ed using Eq (5), assuming ha
in e cep is placed in he selec ion en i onmen (X=0)asi eflec s ac ual si ua ion o selec-
i e b eeding in aquacul u e. The cohe i abili y was calcula ed wice, ei he ha ing BE o PE
as he selec ion en i onmen . The gene ic co ela ion be ween in e cep and slope (
g(in , sl)
)
was calcula ed as: gðin ;slÞ¼^
sain ;sl
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
^
s2
a;in ^
s2
a;sl
p.
Compa ison o p e ious s udies. The e a e no p e ious s udies using eac ion no m
model o s udy en i onmen al sensi i i y in aquacul u e. Hence, o compa e he eac ion no m
pa ame e s o he p esen s udy o he p e ious GxE s udies, (co) a iance componen s o he
p e ious s udies calcula ed using mul i- ai model we e used o calcula e he (co) a iance
componen s o eac ion no m pa ame e s (see S1 and S2 Appendixes). The h2
ES, cohe i abili y
and
g,(in ,sl)
we e calcula ed. The choice o GxE pape s in aquacul u e species was based on he
ollowing: g ow h ai s as he s udied ai , a leas 30 ull-sib amilies, and p o iding all he
pa ame e s needed o he calcula ions. In o al, 17 s udies we e ound, he species co e ing
A c ic cha (Sal elinus alpinus)[30], A lan ic cod (Gadus mo hua)[31], Common ca p
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 6/17
(Cyp inus ca pio)[32], Eu opean whi efish (Co egonus la a e us)[33], Eu opean sea bass
(Dicen a chus lab ax)[34], Nile ilapia (O eoch omis nilo icus)[35,36], Pacific whi e sh imp
(Li openaeus annamei)[37], Rainbow ou [14,16–21,38], Shi anus ilapia (O eoch omis shi -
anus)[39].
Resul s
In he Finnish da a, GxE o BW exis ed in bo h o ms; e- anking as indica ing by
g
o BW
be ween BE and PE was 0.73, and he e ogenei y o gene ic a iances (Table 2). Bo h phenom-
ena induce gene ic a ia ion o ES.
Gene ic a iance o mac oen i onmen al sensi i i y
The addi i e gene ic a iance o slope o BW (9584) was conside able and he h2
in was mode a e
in bo h en i onmen s (0.23 o PE and 0.25 o BE), he h2
ES was low (0.07) implying he addi-
i e gene ic a iance o ES explains only a small p opo ion ela i e o o al pheno ypic a i-
ance o BW ac oss en i onmen s. Simila ly, he cohe i abili y o ES was low and nega i e, i.e.,
-0.06 o PE and -0.08 o BE. Thus he accu acy o selec ion o ES o BW is e y low when
applying indi idual selec ion o BW in one o he en i onmen s.
Fig 1. Rela ionship be ween cohe i abili y and he gene ic co ela ion be ween en i onmen s. The inpu pa ame e s a e a ai wi h pheno ypic
a iance o 1 and he i abili y o 0.3 which a e he same ac oss wo en i onmen s. The gene ic co ela ion anges om -1 o 1.
doi:10.1371/jou nal.pone.0135133.g001
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 7/17
The magni ude o he i abili y o ES o log- ans o med BW was simila o he he i abili y
o ES o obse ed BW. The addi i e gene ic a iance o slope o log- ans o med BW was 69%
in PE and 65% in BE o he addi i e gene ic a iance o in e cep , ela i ely sligh ly highe han
on he obse ed scale (57% in PE and 53% in BE). This indica es ha simple scale e ec s did
no gene a e gene ic a ia ion o mac oen i onmen al sensi i i y.
When PE was he in e cep , he slope EBVs o si es (-234.5 o 228.8) and animals (-210.6
o 199.8) anged om s ongly nega i e o posi i e (Fig 3). A posi i e slope implies ha EBVs
o BW a e ele a ed in BE as compa ed o EBVs o BW in he in e cep en i onmen PE. I BE
was he in e cep en i onmen , he EBVs o slope would change sign.
Gene ic co ela ion be ween in e cep and slope
The signi ican
g(in , sl)
(SE) be ween BW in a gi en en i onmen and ES anged om -0.33
(0.10) o -0.41 (0.10), depending on he en i onmen used as he in e cep en i onmen
(Table 2). The si es wi h s eep slope EBVs had high in e cep (a PE o BE) (Fig 4). The si es
wi h la slope EBV had low in e cep EBV. The nega i e co ela ions om log- ans o med
Fig 2. Rela ionship be ween cohe i abili y, he e ogenei y o addi i e gene ic a iances and he gene ic co ela ion be ween en i onmen s. The
inpu pa ame e s a e a ai ha has di e en magni udes o he i abili y; 0.1 (line wi h ci cles) and 0.5 (line wi h squa es) in wo di e en en i onmen s (E
1
o
E
2
) and pheno ypic a iances a e equal o 1. The gene ic co ela ion anged om -1 o 1. Line g aphs show ha he e ogenei y o addi i e gene ic a iances
esul s in di e en alues o cohe i abili y because h
2in
(0.1 o 0.5) changes, depending on he in e cep which is placed in ei he E
1
o E
2
.
doi:10.1371/jou nal.pone.0135133.g002
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 8/17
da a (-0.40 o -0.42) emained simila o un ans o med da a. This shows ha apid g ow h in
one en i onmen is gene ically ela ed o ele a ed sensi i i y ac oss en i onmen s.
Gene ic pa ame e s calcula ed om he p e ious GxE s udies
In he p e ious aquacul u e s udies on GxE in g ow h, he h2
ES anged om 0.010 o 0.207
(median = 0.110) while he cohe i abili y anged om -0.600 o 0.500 (median = -0.011; in e -
cep a E
1
and = -0.078; in e cep a E
2
)(Fig 5). The
g(in , sl)
be ween g ow h ai s and ES a -
ied om -1.00 o 0.94 (median = -0.386) as shown in Fig 6.
Discussion
Gene ic a ia ion o mac oen i onmen al sensi i i y
Subs an ial addi i e gene ic a iance o mac oen i onmen al sensi i i y (ES) o bo h obse ed
and log- ans o med body weigh (BW) indica es po en ial o gene ic esponse o selec ion on
ES. A e he log- ans o ma ion o BW, he a iance componen s o ES we e educed bu h2
ES
emained simila o he one es ima ed om he un ans o med da a. This indica es ha scale
e ec s (high a iance depending on high mean) do no explain he gene ic e ec s o ES.
Table 2. Va iance componen s and gene ic co ela ions be ween in e cep and slope om he eac-
ion no m (RN) models.
Pa ame e In e cep
P oduc ion B eeding
Body weigh
s
^
a;in
216754.9 18040.0
s
^
a;sl
29584.3 9584.3
s
^
c;in
23041.1 3227.3
s
^
c;sl
23822.7 3822.7
s
^
e;in
253197.8 51092.8
s
^
P;To al
273168.5 73168.5
h2
in 0.23 (0.03) 0.25 (0.03)
h2
ES 0.07 (0.03) 0.07 (0.03)
c2
in 0.04 (0.01) 0.04 (0.01)
Cohe i abili y -0.06 (0.02) -0.08 (0.02)
g(in , sl)
-0.33 (0.10) -0.41 (0.10)
Log(body weigh )
s
^
a;in
20.016 0.017
s
^
a;sl
20.011 0.011
s
^
c;in
20.003 0.003
s
^
c;sl
20.004 0.004
s
^
e;in
20.092 0.071
s
^
P;To al
20.102 0.102
h2
in 0.15 (0.02) 0.18 (0.03)
h2
ES 0.06 (0.02) 0.06 (0.02)
c2
in 0.03 (0.01) 0.04 (0.01)
g(in , sl)
-0.40 (0.10) -0.42 (0.10)
doi:10.1371/jou nal.pone.0135133. 002
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 9/17
9. Ellen E, S a L, Ui dehaag K, B om F, KlopčičM, Reen s R, e al. (2009) Robus ness as a b eeding goal
and i s ela ion wi h heal h, wel a e and in eg i y. In: Klopcic M, Reen s R, Philipsson J, Kuipe s A, edi-
o s. B eeding o obus ness in ca le. Wageningen Academic Publishe s. pp. 45–53.
10. De Jong G (1990) Quan i a i e gene ics o eac ion no ms. J E olu ion Biol 3: 447–468.
11. Schmalhausen II (1949) Fac o s o e olu ion: he heo y o s abilizing selec ion. Ox o d, English: Bla-
kis on. 327 p.
12. Finlay K, Wilkinson G (1963) The analysis o adap a ion in a plan -b eeding p og amme. C op Pas u e
Sci 14: 742–754.
13. Falcone DS, Mackay TFC (1996) In oduc ion o quan i a i e gene ics. 4 h Edi ion ed. UK: Longman,
Essex. pp. 464 pp.
14. Fishback AG, Danzmann RG, Fe guson MM, Gibson JP (2002) Es ima es o gene ic pa ame e s and
geno ype by en i onmen in e ac ions o g ow h ai s o ainbow ou (Onco hynchus mykiss)as
in e ed using molecula pedig ees. Aquacul u e 206: 137–150.
15. Kause A, Ri ola O, Paananen T (2004) B eeding o imp o ed appea ance o la ge ainbow ou in wo
p oduc ion en i onmen s. Aquac Res 35: 924–930.
16. Kause A, Ri ola O, Paananen T, Män ysaa i E, Eskelinen U (2003) Selec ion agains ea ly ma u i y in
la ge ainbow ou Onco hynchus mykiss: he quan i a i e gene ics o sexual dimo phism and geno-
ype-by-en i onmen in e ac ions. Aquacul u e 228: 53–68.
17. Le Bouche R, Quille E, Vandepu e M, Lecal ez JM, Goa don L, Cha ain B, e al. (2011) Plan -based
die in ainbow ou (Onco hynchus mykiss Walbaum): A e he e geno ype-die in e ac ions o main
p oduc ion ai s when ish a e ed ma ine s. plan -based die s om he i s meal? Aquacul u e 321:
41–48.
18. Pie ce LR, Pal i Y, Sil e s ein JT, Ba ows FT, Halle man EM, Pa sons J (2008) Family g ow h
esponse o ishmeal and plan -based die s shows geno ype× die in e ac ion in ainbow ou (Onco -
hynchus mykiss). Aquacul u e 278: 37–42.
19. Sae-Lim P, Kause A, Mulde HA, Ma in KE, Ba oo AJ, Pa sons J, e al. (2013) Geno ype-by-en i on-
men in e ac ion o g ow h ai s in ainbow ou (Onco hynchus mykiss): A con inen al scale s udy. J
Anim Sci 91: 5572–5581. doi: 10.2527/jas.2012-5949 PMID: 24085417
20. Syl én S, Rye M, Simiane H (1991) 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: 253–263.
21. Kause A, Ri ola O, Paananen T, Wahl oos H, Män ysaa i EA (2005) Gene ic ends in g ow h, sexual
ma u i y and skele al de o ma ions, and a e o inb eeding in a b eeding p og amme o ainbow ou
(Onco hynchus mykiss). Aquacul u e 247: 177–187.
22. Via S, Gomulkiewicz R, De Jong G, Scheine SM, Schlich ing CD, Van Tiende en PH (1995) Adap i e
pheno ypic plas ici y: consensus and con o e sy. T ends Ecol E ol 10: 212–217. PMID: 21237012
23. Van Tiende en PH, Koelewijn HP (1994) Selec ion on eac ion no ms, gene ic co ela ions and con-
s ain s. Gene Res 64: 115–125. PMID: 7813902
24. Gilmou AR, Gogel BJ, Cullis BR, Thompson R (2009) ASReml Use Guide Release 3.0. NSW Depa -
men o Indus y and In es men .
25. Fische T, Gilmou A, We J (2004) Compu ing app oxima e s anda d e o s o gene ic pa ame e s
de i ed om andom eg ession models i ed by a e age in o ma ion REML. Gene Sel E ol 36: 363–
369. PMID: 15107271
26. Lande R (1979) Quan i a i e gene ic analysis o mul i a ia e e olu ion, applied o b ain: body size
allome y. E olu ion: 402–416.
27. Scheine SM, Lyman RF (1989) The gene ics o pheno ypic plas ici y I. He i abili y. J E olu ion Biol 2:
95–107.
28. Scheine SM (1993) Gene ics and e olu ion o pheno ypic plas ici y. Annu Re Ecol Sys 24: 35–68.
29. Kolmodin R, Bijma P (2004) Response o mass selec ion when he geno ype by en i onmen in e ac-
ion is modelled as a linea eac ion no m. Gene Sel E ol 36: 435–454. PMID: 15231233
30. Nilsson J (1990) He i abili y es ima es o g ow h- ela ed ai s in A c ic cha (Sal elinus alpinus). Aqua-
cul u e 84: 211–217.
31. Kols ad K, Tho land I, Re s ie T, Gje de B (2006) Gene ic a ia ion and geno ype by loca ion in e ac ion
in body weigh , spinal de o mi y and sexual ma u i y in A lan ic cod (Gadus mo hua) ea ed a di e en
loca ions o No way. Aquacul u e 259: 66–73.
32. Ninh NH, Ponzoni RW, Nguyen NH, Woolliams JA, Tagga JB, McAnd ew BJ, e al. (2011) A compa i-
son o communal and sepa a e ea ing o amilies in selec i e b eeding o common ca p (Cyp inus ca -
pio): Es ima ion o gene ic pa ame e s. Aquacul u e 322: 39–46.
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 16 / 17

33. Che yl D, An i K, Juha K (2007) B eeding salmonids o eed e iciency in cu en ishmeal and u u e
plan -based die en i onmen s. Gene Sel E ol 39: 431–446. PMID: 17612482
34. Dupon -Ni e M, Vandepu e M, Ve gne A, Me dy O, Ha ay P, Cha anne H, e al. (2008) He i abili ies
and GxE in e ac ions o g ow h in he Eu opean sea bass (Dicen a chus lab ax L.) using a ma ke -
based pedig ee. Aquacul u e 275: 81–87.
35. Cha o-Ka isa H, Komen H, Reynolds S, Rezk MA, Ponzoni RW, Bo enhuis H (2006) Gene ic and en i-
onmen al ac o s a ec ing g ow h o Nile ilapia (O eoch omis nilo icus) ju eniles: Modelling spa ial
co ela ions be ween hapas. Aquacul u e 255: 586–596.
36. Khaw HL, Ponzoni RW, Hamzah A, Abu-Baka KR, Bijma P (2012) Geno ype by p oduc ion en i on-
men in e ac ion in he GIFT s ain o Nile ilapia (O eoch omis nilo icus). Aquacul u e 326: 53–60.
37. Cas illo-Juá ez H, Casa es JCQ, Campos-Mon es G, Villela CC, O ega AM, Mon aldo HH (2007) He i-
abili y o body weigh a ha es size in he Paci ic whi e sh imp, Penaeus (Li openaeus) annamei,
om a mul i-en i onmen expe imen using uni a ia e and mul i a ia e animal models. Aquacul u e
273: 42–49.
38. Tobin D, Kause A, Män ysaa i EA, Ma in SA, Houlihan DF, Dobly A, e al. (2006) Fa o lean? The
quan i a i e gene ic basis o selec ion s a egies o muscle and body composi ion ai s in b eeding
schemes o ainbow ou (Onco hynchus mykiss). Aquacul u e 261: 510–521.
39. Maluwa AO, Gje de B, Ponzoni RW (2006) Gene ic pa ame e s and geno ype by en i onmen in e ac-
ion o body weigh o O eoch omis shi anus. Aquacul u e 259: 47–55.
40. Janssens M (1979) Cohe i abili y: i s ela ion o co ela ed esponse, linkage, and pleio opy in cases o
polygenic inhe i ance. Euphy ica 28: 601–608.
41. Hill W, Mulde H (2010) Gene ic analysis o en i onmen al a ia ion. Gene Res 92: 381–395.
42. Janhunen M, Kause A, Veh ilainen H, Ja isalo O (2012) Gene ics o mic oen i onmen al sensi i i y o
body weigh in ainbow ou (Onco hynchus mykiss) selec ed o imp o ed g ow h. PLoS One 7:
e38766. doi: 10.1371/jou nal.pone.0038766 PMID: 22701708
43. Mulde H, Bijma P, Hill W (2007) P edic ion o b eeding alues and selec ion esponse wi h gene ic he -
e ogenei y o en i onmen al a iance. Gene ics 175: 1895–1910. PMID: 17277375
44. Sae-Lim P, Kause A, Janhunen M, Veh iläinen H, Koskinen H, Gje de B, e al. (2015) Gene ic (co) a i-
ance o ainbow ou (Onco hynchus mykiss) body weigh and i s uni o mi y ac oss p oduc ion en i on-
men s. Gene Sel E ol 47: 46. doi: 10.1186/s12711-015-0122-8 PMID: 25986847
45. Sae-Lim P, Gje de B, Nielsen HM, Mulde H, Kause A (2015) A e iew o geno ype-by-en i onmen
in e ac ion and mic o-en i onmen al sensi i i y in aquacul u e species. Re Aquacul u e: In p ess.
46. Jinks J, Connolly V (1973) Selec ion o speci ic and gene al esponse o en i onmen al di e ences.
He edi y 30: 33–40.
47. Walsh B, Lynch M (2013) E olu ion and selec ion o quan i a i e ai s: II. Ad anced opics in b eeding
and e olu ion. A ailable: h p://ni o.biosci.a izona.edu/zbook/NewVolume_2/new ol2.h ml. Accessed
12 July 2013.
48. Rosielle A, Hamblin J (1981) Theo e ical aspec s o selec ion o yield in s ess and non-s ess en i on-
men . C op Sci 21: 943–946.
49. B ascamp E. Selec ion indices wi h cons ain s; 1984. pp. 645–654.
50. Sae-Lim P, Komen H, Kause A, Mulde H (2014) Iden i ying en i onmen al a iables explaining geno-
ype-by-en i onmen in e ac ion o body weigh o ainbow ou (Oncho ynchus mykiss): eac ion no m
and ac o analy ic models. Gene Sel E ol 46: 16. doi: 10.1186/1297-9686-46-16 PMID: 24571451
51. Flahe y M, Szus e B, Mille P (2000) Low salini y inland sh imp a ming in Thailand. Ambio 29: 174–
179.
Gene ics o Mac oen i onmen al Sensi i i y
PLOS ONE | DOI:10.1371/jou nal.pone.0135133 Augus 12, 2015 17 / 17