RESEARCH ARTICLE
Complemen a y me hods assessing sho and
long- e m p ey o a ma ine op p eda o –
Applica ion o he g ey seal- ishe y con lic in
he Bal ic Sea
Malin T e in
1
, Rod igo Espa za-Salas
2¤a
, Annika S o
¨mbe g
3
, Pa ik Tang
1¤b
,
Ii is Kokkonen
1
, Annika He e o
4
, Kaa ina Kauhala
5
, Olle Ka lsson
3
, Raisa Tiilikainen
6
,
Ma kus Ve emaa
7
, Tuula Sinisalo
8
, Reijo Ka
¨kela
¨
1
, Ka l Lunds o
¨mID
9
*
1Molecula and In eg a i e Biosciences Resea ch P og amme, Facul y o Biological and En i onmen al
Sciences, Uni e si y o Helsinki, Helsinki, Finland, 2Depa men o Biology, Uni e si y o Oulu, Oulu, Finland,
3Depa men o En i onmen al Resea ch and Moni o ing, Swedish Museum o Na u al His o y, S ockholm,
Sweden, 4Na u al Resou ces Ins i u e, Luke, Helsinki, Finland, 5Na u al Resou ces Ins i u e, Luke, Tu ku,
Finland, 6Me sa¨halli us Pa ks & Wildli e, Sa onlinna, Finland, 7Es onian Ma ine Ins i u e, Uni e si y o Ta u,
Ta u, Es onia, 8Depa men o Biological and En i onmen al Science, Uni e si y o Jy a
¨skyla
¨, Jy a
¨skyla
¨,
Finland, 9Depa men o Aqua ic Resou ces, Swedish Uni e si y o Ag icul u al Sciences, Lysekil, Sweden
¤a Cu en add ess: Depa men o Bioin o ma ics and Gene ics, Swedish Museum o Na u al His o y,
S ockholm, Sweden
¤b Cu en add ess: Depa men o Biological Sciences, Uni e si y o Be gen, Be gen, No way
*ka l.lunds [email protected]
Abs ac
The g owing g ey seal (Halichoe us g ypus) popula ion in he Bal ic Sea has c ea ed con-
lic s wi h local ishe ies, compa able o simila eme ging p oblems wo ldwide. Adequa e
in o ma ion on he o aging habi s is a equi emen o esponsible managemen o he seal
popula ion. We in es iga ed he applicabili y o a ailable die a y assessmen me hods by
compa ing mo phological analysis and DNA me aba coding o gu con en s (sho - e m die ;
n = 129/125 seals, espec i ely), and issue chemical ma ke s i.e. a y acid (FA) p o iles o
blubbe and s able iso opes (SIs) o li e and muscle (mid- o long- e m die ; n = 108 seals
o he FA and SI ma ke s). The me hods p o ided complemen a y in o ma ion. Sho - e m
me hods indica ed p ey species and e ealed die a y di e ences be ween age g oups and
a eas bu o limi ed ime pe iod. In he cen al Bal ic, he ing was he main p ey, while in he
Gul o Finland pe cid and cyp inid species oge he comp ised he la ges pa o he die .
Pe ch was also an impo an p ey in he wes e n Bal ic P ope . The DNA analysis p o ided
i m iden i ica ion o many p ey species, which we e neglec ed o iden i ied only a species
g oup le el by mo phological analysis. Li e SIs dis inguished spa ial o aging pa e ns and
iden i ied po en ially mig a ed indi iduals, whe eas blubbe FAs dis inguished indi iduals
equen ly u ilizing ce ain ypes o p ey. Tissue chemical ma ke s o adul males sugges ed
specialized eeding o ce ain a eas and p ey, which sugges ha hese indi iduals a e espe-
cially p one o cause economic losses o ishe ies. We ecommend combined analyses o
gu con en s and issue chemical ma ke s as die a y moni o ing me hodology o aqua ic op
p eda o s o suppo an op imal ecosys em-based managemen .
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 1 / 26
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Ci a ion: T e in M, Espa za-Salas R, S o¨mbe g A,
Tang P, Kokkonen I, He e o A, e al. (2019)
Complemen a y me hods assessing sho and
long- e m p ey o a ma ine op p eda o –
Applica ion o he g ey seal- ishe y con lic in he
Bal ic Sea. PLoS ONE 14(1): e0208694. h ps://doi.
o g/10.1371/jou nal.pone.0208694
Edi o : Leszek Ka czma ski, Uni e si y o Hong
Kong, HONG KONG
Recei ed: Janua y 19, 2018
Accep ed: No embe 22, 2018
Published: Janua y 2, 2019
Copy igh : ©2019 T e in 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 : All ele an da a a e
wi hin he pape and i s Suppo ing In o ma ion
iles.
Funding: The collec ion and analysis o seal and
ish samples was suppo ed by he EU unded
Ecoseal p ojec in 2012-13. Addi ional unding was
ob ained om Finnish Cul u al Founda ion
(00150994), O o A. Malms dona ions ond (7-
5077-44) and Oska O
¨ lunds s i else (2-3496-18).
In oduc ion
Inc easing seal popula ions wo ldwide ha e c ea ed esou ce compe i ion and con lic s
be ween he seals and local comme cial ishe ies, leading o culling p og ams wi h unce ain
bene i s [1,2]. Thus, eliable scien i ic da a on seal eeding habi s and esou ce exploi a ion is
equi ed. The Bal ic g ey seal (Halichoe us g ypus) popula ion has eco e ed om he low
numbe s in he 1980s, caused by ex ensi e hun ing and en i onmen al oxins, o abou 30 000
coun ed animals [3,4]. Consequen ly, con lic s wi h coas al ishe ies ha e inc eased, mainly
due o damage o ca ch and ishing gea [5] bu also because o possible esou ce compe i ion
and bycaugh seals [6]. Selec i e emo al o specialized p oblem seals has been sugges ed as a
me hod o mi iga e damage o ishe ies and a he same ime a oid o e hun ing [5,7].
Ta ge ed hun o p oblem seals is easible i indi idual p e e ences o ce ain eeding a eas
and p ey species exis . Bal ic g ey seals, as a popula ion, ha e been conside ed oppo unis ic
p eda o s, an in e p e a ion based on analysis o gu con en s [8–10]. Acco ding o hese s ud-
ies, he ing (Clupea ha engus), is he mos impo an p ey, ollowed by cod (Gadus mo hua)
and sp a (Sp a us sp a us) in he Bal ic P ope , and common whi e ish (Co egonus la a e us)
and endace (Co egonus albula) in he Gul o Bo hnia. Howe e , i is no known how ep e-
sen a i e his in o ma ion is. The adi ional die es ima ion me hod, based on mo phological
iden i ica ion o he p ey emains in he gu , only ep esen s he mos ecen die , and migh
be biased owa ds p ey wi h long- e ained ha d pa s (HP). Cu en ly, he HP analysis could
be complemen ed wi h DNA analysis o he gu con en s which may educe bias caused by
diges i e e osion [11] and e eal p ey wi h no ecognizable ha d pa s [12]. ICES geog aphical
egions (subdi isions, SD) a e commonly used o assess and manage ish s ocks in he Bal ic
Sea, and hese egions co espond o spa ial di e ences in hyd og aphy and ecology [13].
Since p e ious s udies u ilizing HP analysis ha e iden i ied ICES geog aphical egions, sam-
pling gea ype and age g oup as he mos impo an explana o y ac o s o Bal ic g ey seal
die a ia ion [8], he seals sampled o his s udy we e g ouped acco dingly. In addi ion, pos-
sible ecological di e ences be ween he wes e n and eas e n coas o he same ICES SD we e
also aken in o accoun when g ouping he indi iduals. The e ec o gende on he die has
been ega ded as less impo an ac o [8], al hough die a y di e ences be ween male and
emale g ey seals ha e been documen ed in o he a eas [14].
Recen s udies ha e sugges ed ha indi idual g ey seals, ins ead o being oppo unis ic,
ha e specialized eeding a eas and beha iou s [5,15,16]. Al hough Bal ic g ey seals a e capable
o long-dis ance mo emen s, e en be ween ICES subdi isions, a ailable in o ma ion sugges s
ha hey o age on a mo e local spa ial scale in he icini y o p e e ed haul-ou a eas, how-
e e wi h subs an ial indi idual di e ences [16]. The possible ideli y o he indi iduals o ce -
ain o aging a ea was add essed by eco ding he loca ions and gea ypes he indi iduals
we e ound in. P o ided ha he indi iduals om he same a ea and gea - ype sys ema ically
show simila die a y ma ke p o iles, which howe e a e di e en om he ma ke p o iles o
he indi iduals om o he gea ypes wi hin possible daily ange o swimming, he seals likely
ha e se led eeding a eas and habi s. Me hods p o iding es ima es on long- e m die may help
o e eal such indi idual specializa ion in ce ain eeding a eas and ypes o p ey consumed
he ein. This long- e m die a y in o ma ion can be ob ained om chemical ma ke s in p eda-
o issues, such as blubbe , li e and muscle. T ans e o die a y a y acids (FAs) in o ma ine
mammal blubbe is assumed o occu wi h li le me abolic emodeling, which makes he FAs
sui able o die moni o ing [17,18]. Howe e , when using blubbe FAs o s udy seal eeding
ecology i should be no ed ha seal blubbe is e ically laye ed and he composi ion o he
ou e mos laye is ai ly s able due o i s he mo egula o y ole [19–21]. The middle and inne -
mos laye s a e ega ded me abolically ac i e, wi h he inne laye assumingly e lec ing mid-
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 2 / 26
Compe ing in e es s: The au ho s ha e decla ed
ha no compe ing in e es s exis .
e m (a ew weeks) die , whe eas he middle laye in eg a es long- e m (se e al mon hs) die-
a y in o ma ion [20,22]. Di e ing ac iona ion o hea y (e.g.δ
13
C, δ
15
N o δ
34
S) and ligh ele-
men iso opes in p ey leads o p edic able changes in he s able iso ope (SI) alues in p eda o
issues and di e en issues can p o ide SI-based die a y in o ma ion in di e en ime scales,
e.g. weeks o li e samples and mon hs o muscle samples [23,24]. To be success ully accom-
plished, he FA and SI analyses equi e ex ensi e p ey FA and SI lib a ies.
By using da a om a a ie y o me hods i is possible o ge es ima es on sho -, mid- and
long- e m die s o indi idual seals. The i s aim o he s udy was o compa e he sho - e m
die es ima es ob ained om HP and DNA analysis o g ey seal gu con en s, and o in es iga e
he complemen a i y o hese wo me hods. Second, we compa ed he powe o issue FA and
SI p o iles in assessing mid- and long- e m eeding habi s and examined whe he hese me h-
ods a e able o e eal indi idual, o age- and sex-g oup ela ed specializa ion. We hypo hesized
ha he esul s om HP and DNA analyses would di e om each o he . Fu he , we hypo he-
sized ha signi ican ly di e en chemical ma ke p o iles e e o indi iduals specialized in a
ce ain o aging a ea and/o die , whe eas simila ma ke p o iles would mean no p e e en ial
use o habi a o p ey. G owing seal popula ions may adop new o aging a eas and esou ces,
and adequa e in o ma ion on he spa ial and empo al die a y a iabili y cla i ies he ecological
ole o ma ine mammals and may o e means o mi iga ing con lic s be ween seals and ish-
e ies. In addi ion, die a y shi s and issue chemis y o op p eda o s a e in eg a ed p oxies o
ood web changes, hus indica ing he dynamics and heal h o he ecosys em [25,26].
Me hods
Sample collec ion
The seal and ish samples we e collec ed in collabo a ion wi h ongoing na ional and in e na-
ional moni o ing p og ammes o ish and seals: in Sweden p omo ed by he En i onmen al
P o ec ion Agency (h p://www.swedishepa.se) and he Agency o Ma ine and Wa e Man-
agemen (h p://www.ha och a en.se) and ca ied ou by he Uni e si y o Ag icul u al Sci-
ences (h p://www.slu.se) and Museum o Na u al His o y (h p://www.n m.se); in Finland
conduc ed by he Na u al Resou ces Ins i u e (h p://www.luke. i). The samples we e collec ed
du ing 2011 and 2012 and co e ed he ICES SDs 27, 29, 30 and 32 o he Bal ic Sea (Fig 1).
Seal samples
Blubbe , muscle, li e (n = 108 o each) and gu samples (n = 129 and 125 o HP and DNA,
espec i ely) om g ey seals (all sample ypes we e aken om 67 indi iduals) we e collec ed
du ing 2011 and 2012. The SD29 and 30 include pelagic and coas al a eas, and he wes and
eas coas ecosys ems could p o ide he seals wi h di e en die ha ing dis inc chemical ma k-
e s. Howe e , all he indi iduals om SD29 we e collec ed in he a chipelago be ween Åland
Islands and Tu ku, and hus o med an ecologically uni o m sample. The seals collec ed in SD
30 we e mainly om he wes coas (n = 22, excep o DNA n = 20) bu specimens o he eas
coas (n = 11 HP/DNA, n = 6 o FA/SI) we e included as well, and hus he seals we e sub-
g ouped in o wes e n SD30 and eas e n SD30. Seal sex and age (numbe o cemen um zones
in canine ee h longi udinal sec ions [27]) we e eco ded, as well as in o ma ion on sampling
loca ion, da e and cause o dea h: sho ei he close o ishing gea (C) o ish a m (F) o in
o he a eas (O) o bycaug h wi h di e en gea s. The ype o ishing gea was documen ed:
awl (T), su ace yke (S) o bo om yke (B). The su ace ykes had he loa ing push-up
design and we e mean o ca ch he la ge pelagic species, A lan ic salmon (Salmo sala ), sea
ou (Salmo u a) and common whi e ish, whe eas he bo om ykes we e placed a he bo -
om, had a ious adi ional s uc u es, and we e mean o ca ch pe ch (Pe ca lu ia ilis),
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 3 / 26
Fig 1. Si es o collec ion o he g ey seals s udied o he die a y p oxies. ICES a eas o he Bal ic Sea and numbe o
di e en g ey seal indi iduals collec ed om he subdi isions 27, 29 (only he no h-eas e n pa i.e. he a chipelago
be ween Åland Islands and Tu ku, was included), 30 (di ided in o SD30 wes and SD30 eas g oups) and 32 (eas e n
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 4 / 26
pikepe ch (Sande luciope ca), Eu opean eel (Anguilla Anguilla) and cyp inids. The gea ype
in o ma ion was used in his s udy o de ine he speci ic habi a o he a ea whe e he seal was
collec ed. Since 5 yea s is he mos likely age o he i s bi h o he g ey seal [28], he 0–4 yea -
old seals we e classi ied as subadul s and he 5+ yea -old seals as adul s. The hun ed seal indi-
iduals (n = 75, 74, 64 o HP, DNA and FA/SI) we e sampled in he ield, and he diges i e
ac and o iginal la ge-size issue samples we e s o ed in eeze (-25˚C) be o e subsampled in
he labo a o y ( o HP, DNA, FA and SI analysis). The bycaugh seals (n = 54, 51, 44 o HP,
DNA and FA/SI) we e collec ed whole and s o ed in eeze (-25˚C) be o e sampled and sub-
sampled in he labo a o y du ing au opsy. The sample s o age ime be o e he analyses o he
ma e ial was less han 6 mon hs o all ypes o analyses.
Re e ence lib a y o p ey ish issue
Whole ish we e s o ed in eeze (-25˚C) be o e homogenized and sampled in he labo a o y
o he 4 ypes o analyses. The ish issue lib a y c ea ed consis ed o 26 species bu he p o-
ound species-le el analyses o he whole da a wi h egional compa isons emain ou o he
scope o his s udy and will be published sepa a ely. This ull ish ma e ial o 433 indi iduals
we e a i s used o add ess he chemical ma ke a iabili y o he Bal ic ishes, and subse-
quen ly he 11 mos p obable p ey species ( o FA n = 233 and o SI n = 216) we e chosen o
he compa a i e analyses o his s udy (Tables A-D in S1 Table). The ull da a, howe e , we e
u ilized o iden i y he indi idual FAs esponsible o he la ges in e species a ia ion and hus
b inging wi h hem die a y in o ma ion in o p eda o issues. Fo p ey-p eda o compa isons
o he s udy, he FAs and SIs o 11 key p ey ish species (mo e han 200 ishes), caugh om he
main habi a s o he s udy a ea and epo ed o o m he base o he g ey seal die [8], we e ana-
lysed and he powe o FAs and SIs o dis inguish hese pelagic (he ing, sp a , A lan ic salmon
and sea ou ), coas al p eda o y (pikepe ch, pike Esox lucius and pe ch) and deme sal (com-
mon whi e ish, eelpou Zoa ces i ipa us and oach Ru ilus u ilus) ish was demons a ed.
Eu opean eel, being a mig a ing species was no ca ego ized in o any a o emen ioned habi a .
Gu con en mo phological analysis
The mo phological HP analysis ollowed he me hodology desc ibed by Lunds o¨m e al.
[8,29]. B ie ly, con en s om s omachs and in es ines we e placed on a 0.5 mm sie e and a
small po ion o he p oduced liquid sample was collec ed and s o ed a -20˚C o subsequen
DNA analysis. P ese ed p ey specimens we e iden i ied and measu ed, ollowed by iden i ica-
ion o sie ed o oli hs and o he HPs by using e e ence collec ions (5 specimens o a ying
size o each species) and li e a u e [30,31]. Sizes, numbe s and biomass o p ey i ems (mos ly
ish wi h only a ew in e eb a e Sadu ia en omon specimens) inges ed pe indi idual seal
we e es ima ed by conside ing all p ey HPs, known ela ionships be ween o oli h size and ish
size, and compensa ing o diges i e e osion o o oli hs [29].
DNA me aba coding o gu con en s
DNA was ex ac ed indi idually o e e y s omach, in es ine and colon con en sample using a
QIAmp DNA s ool miniki (Qiagen N. V. Venlo, Ne he lands) ollowing he manu ac u e ’s
pa , he Russian sea a ea decluded), and s udied o A) gu con en s (HP and DNA, n = 129 and 125, espec i ely) and
B) issue chemical ma ke s (blubbe FAs and issue SIs, n = 108). Seal indi iduals (M = male, F = emale) we e u he
ca ego ized by he way/place o collec ion: T = awl, S = su ace yke, B = bo om yke, C = close o ishing gea , F = by
ish a m, O = open wa e , UD = unde ined ishing gea and UK = unknown hun ing a ea (C/F/O inside he SD).
Bycaugh seals = T, S, B, UD; hun ed seals C, F, O, UK.
h ps://doi.o g/10.1371/jou nal.pone.0208694.g001
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 5 / 26

“p o ocol o Human DNA”. An app oxima ely 270 base-pai long agmen o he 16s DNA
gene (he eby 16s) was ampli ied by polyme ase chain eac ion (PCR) o be used as a “DNA
ba coding” ma ke o p ey species iden i ica ion. PCR p ime s, o wa d p ime 16sP eyF
(5’-CGTGCRAAGGTAGCG-3’) and e e se p ime 16sP eyR (5’-CCTYGGGCGCCCCA
AC-3’) we e designed by aligning and iden i ying a iable sec ions o 16s sequences om
a ious ma ine e eb a es p esen in he Bal ic Sea, including seals and aqua ic bi ds. The 3’
nucleo ide o he o wa d p ime misma ches he 16s sequence o seals, which inhibi s he
ampli ica ion o seal DNA, maximizing he p ey DNA ampli ica ion.
The p ime pai was es ed ini ially using e e ence DNA empla e om 47 di e en ish
species and eigh bi d species om he Bal ic Sea egion. Wi h he excep ion o Agna ha spe-
cies (Lampe a lu ia ilis and Pe omyzon ma inus), all samples p oduced equally s ong PCR
p oduc s as isualized in aga ose gels (da a no shown). Eigh o wa d and eigh e e se p im-
e s we e syn hesized con aining unique combina ions o six nucleo ides a he 5’ end. Such
p ime s we e used o p oduce 64 unique “ba code” iden i ie combina ions o acili a e mul i-
plexing o indi iduals in pa allel sequencing and subsequen de-mul iplexing o he ou pu
da a, as desc ibed by [32].
PCR eac ions we e ca ied ou in olumes o 25 μL con aining 12.5 μL Ho S a Taq mas e
mix (Qiagen), 1 μL o each PCR p ime (10 μM concen a ion), and 2 μL o DNA ex ac .
Cycling condi ions included an ini ial 5 minu e (min) dena u ing s ep a 95˚C; 40 cycles o
dena u ing a 94˚C o 30 seconds (s), 54˚C o 30 s and 68˚C o 60 s; and ending wi h a inal
ex ension s ep o 72˚C o 10 min.
PCR p oduc s we e pooled in g oups o 64 ba coded indi iduals. Pooled eac ions we e
hen used o cons uc DNA lib a ies o sequencing ollowing he “Rapid lib a y p epa a ion
me hod manual” o GS junio Ti anium se ies (Roche, Ma ch 2012) wi h he ollowing modi-
ica ions: he nebuliza ion s ep was omi ed, he RLdNTP, RL T4 polyme ase and RL Taq poly-
me ase we e no included in he agmen end- epai eac ion, and he small agmen
emo al was ca ied ou by aga ose-gel size selec ion and excision. Each o he pooled 64 indi-
idual eac ion lib a ies was p epa ed using a di e en molecula iden i ie adap e (MID).
DNA lib a ies we e sequenced in wo di e en uns in a GS-Junio ins umen (Roche), ol-
lowing he emPCR ampli ica ion manual- Lib-L” and he “Sequencing me hod manual GS
junio Ti anium Se ies” p o ocols (Roche).
The DNA sequence da a ou pu in Fas A o ma and i s espec i e quali y sco es we e com-
bined in o a Fas Q ile using Galaxy [33]. Sequence eads wi h ei he a <60 bp leng h, a quali y
sco e o <15 o a non-de ined base call (N-bases) o >2% we e il e ed ou om he da ase
using PRINSEQ [34]. The so ing o he sequencing ou pu ile in o indi idual lib a ies and
indi iduals wi hin lib a ies, espec i ely, was ca ied ou using he p og am 454 ag so ing by
Johan Nylande (h ps://gi hub.com/nylande /454_ ag_so ing). Compa isons we e pe -
o med using he BLASTn algo i hm and species iden i ica ion om he ou pu sequences was
ca ied ou using he BLAST+ p og am [35], wi h he indi idually agged DNA sequences (in
Fas A o ma ) as a que y da abase, and he nucleo ide collec ion (n ) as a e e ence subjec
da abase. Only he highes sco e o each compa ison was kep . Ma ching da abase eco ds
we e hen compa ed indi idually using BLASTn in o de o iden i y and co ec ambiguous
ma ches. Indi idual samples ha p oduced less han 100 alid p ey sequence ma ched we e
disca ded om u he analyses. Finally, he die a y da a om he DNA analysis we e
exp essed as ela i e p opo ions o axon speci ic sequences wi hin a sample. The p opo ion
o DNA sequences om a p ey species indica es i s con ibu ion o he die bu i is no equal
o he ela i e biomass consumed [36,37].
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 6 / 26
Seal blubbe and ish a y acid analysis
Blubbe samples we e consis en ly collec ed om abo e s e num. In addi ion, he accu a e
sampling loca ion has been epo ed o ha e negligible in luence on he FA composi ion o
pinniped blubbe [20]. The blubbe and e e ence ish samples we e s o ed a —20˚C un il
analysis. FA me hyl es e s we e p epa ed om subsamples o blubbe (dissec ed wi h skin and
muscle) acco ding o published p o ocols [20] and homogena es o whole ish (2 g subsample).
Upon sampling, he blubbe was ozen in liquid ni ogen, and e ically adjacen subsamples
we e aken by 3 mm in e als om skin o muscle, whe e he 3–6 mm abo e muscle ep e-
sen ed he inne laye and he 6 mm abo e muscle o 18 mm below skin ep esen ed he mid-
dle laye . These bounda ies o he middle laye we e con i med by s udying he comple e
e ical p o iles o each FA mol% in he blubbe column o he adul males. Simila laye s
wi h same FA cha ac e is ics we e ound in he g ey seals o his s udy as epo ed p e iously
o inged seals [20,21].
The FA composi ion in he seal and ish issue samples was analyzed by gas ch oma og a-
phy acco ding o p e iously published p ocedu es [20,38] using a Shimadzu GC-2010 Plus
equipmen (Shimadzu Scien i ic Ins umen s, Kyo o, Japan) wi h lame-ioniza ion de ec o
(FID) o quan i ica ion o he FAs. Iden i ica ion o he FA s uc u es was pe o med by Shi-
madzu GCMS-QP2010 Ul a (Shimadzu) wi h mass selec i e de ec o (MSD). Bo h sys ems
we e equipped wi h Zeb on ZB-wax capilla y columns (30 m, 0.25 mm ID and ilm hickness
0.25 μm; Phenomenex, To ence CA, USA). The FA composi ions we e exp essed as mol%
p o iles, and he FAs we e abb e ia ed: [ca bon numbe ]:[numbe o double bonds] n-[posi-
ion o he i s double bond calcula ed om he me hyl end] (e.g. 20:5n-3). When s udying he
ish homogena es (o 26 Bal ic species), 9 FAs (14:0, 16:1n-7, 18:1n-9, 18:2n-6, 18:3n-3, 18:4n-
3, 20:1n-7, 20:4n-6, 22:6n-3) explained he mos pa o he in e speci ic a ia ion and hus
hese we e used as die a y ma ke s o he seals. These FAs showed he la ges ela i e s anda d
de ia ions among he FAs p esen wi h le els no a ec ed by me hodological a ia ion (only
he FAs wi h signals exceeding 10x he eplica e a ia ion le el we e accep ed o ma ke can-
dida es), and hey also we e esponsible o he main pa o he da a a ia ion in he P incipal
Componen Analysis (PCA; see S a is ics) using as loadings s anda dized mol% da a o ei he
he ull 26 species o he selec ed 11 main p ey species.
Seal issue and ish s able iso ope analysis
Seal muscle and li e samples, and he e e ence ish homogena es we e ozen, eeze-d ied
and powde ed o δ
13
C, δ
15
N and δ
34
S analyses. A maximum o 0.6 mg o each sample was
loaded in o a 4x6 mm in capsule and combus ed in Elemen a Va io Py ocube elemen al ana-
lyse (Elemen a , Ge many) connec ed o Isop ime 100 CF-IRMS (Isop ime UK) mass spec-
ome e . Di e ences in he iso ope alues we e measu ed ela i e o s anda ds and exp essed
as pe mil (‰) de ia ion om Vienna PeeDee belemni e (VPDB) o ca bon, om a mo-
sphe ic N
2
(AIR) o ni ogen, and Vienna Canon Diablo Me eo i e T oili e (V-CDT) o sul-
phu [39,40]. Mo e p ecisely δ
13
C, δ
15
N, δ
34
S (‰) = (R sample / R s anda d– 1) ×103, whe e
R sample is he a io be ween he hea y iso ope and i s ligh e coun e pa o he sample, and
R s anda d is he a io o he in e na ional s anda d [41].
Fo s anda d e e ence ma e ials pike muscle (FSS) s anda d was used as he in e nal labo-
a o y s anda d, calib a ed agains iso opic s anda ds (e.g. CH
6
, N
2
and sphale i e NBS 123)
p o ided by he In e na ional A omic Ene gy Agency (IAEA, Vienna). FSS has known alues
o δ
13
Cs d = -26.39 ‰, δ
15
Ns d = 13.08 ‰ and δ
34
Ss d = 12.45 ‰ and was used as wo king
s anda d o examine iso opic d i wi hin and h oughou he un. Elemen al analysis s anda d
e e ence ma e ial, sul anilamide (IVA Analysen echnike. K.) was used o co ec he % C, %
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 7 / 26
N and % S da a. As lipids a e known o be
13
C-deple ed (ha ing lowe δ
13
C ela i e o o he
majo issue cons i uen s as p o eins) [42,43], ca bon iso ope alues (δ
13
C) we e ‘lipid no mal-
ized’ o bo h seal issues and all ish samples using he C/N a io acco ding o [44].
Tu no e a es o issue chemical ma ke s
In o ma ion on he u no e a es o SIs and FAs in he s udied seal issues (li e , muscle, blub-
be ) [20,45,46] sugges s ha he long- e m die a y ma ke s may u he all in o wo imescale
ca ego ies. Rega ding s able iso opes, seal li e has ela i ely as u no e o biomolecules
while he u no e o muscle is ela i ely slow [47–50]. In mammals, he ca bon u no e ime
in li e has a hal -li e o 6.4 days and muscle 27.6 days [47,48]. Some signal o pas die may
s ill be de ec able a e a pe iod oughly 2 o 3 imes ha o he iso opic hal -li e o he issue
[49,50]. The e o e, we assumed he δ
13
C, δ
15
N and δ
34
S alues om seal li e e lec he die a y
elemen s 2–3 weeks p io o sampling, and he iso ope alues om muscle issue should ep e-
sen he elemen s assimila ed up o 2–3 mon hs p io o sampling.
In mammals, clea ance o ci cula ing chylomic ons and abso p ion o FAs begins in min
scale and he lipids no immedia ely needed o ene gy me abolism a e s o ed in he adipose
issue [51,52]. This sugges ha he FA composi ion in he inne mos laye s o blubbe is likely
a ec ed a e a b ie pos p andial pe iod. Un o una ely, he sys ema ic wo ks de ining he FA
u no e a es a di e en dep hs o seal blubbe a e missing. Since die a y polyunsa u a ed
FAs (PUFAs) a e p e e en ially ound in he inne laye s o blubbe , his laye is likely he me -
abolically mos ac i e laye o blubbe [20,45]. This iew is u he suppo ed by he ac s ha
in inged seals he inne mos blubbe shows he la ges composi ional simila i ies wi h he
po en ial p ey ish FAs, and ha in acciden ally caugh indi iduals om he same a ea o Lake
Saimaa he FA composi ion o he inne mos blubbe laye s showed he la ges indi idual
composi ional a iabili y [20,45]. By using ull laye biopsies in a ha bou seal eeding expe i-
men , No ds om e al. [53] es ima ed he o e all blubbe FA u no e a e being 2–3 mon hs,
hus gi ing a jus i ied es ima e o he middle blubbe . The u no e a e o he FAs in he
inne mos blubbe is likely much sho e .
S a is ics
FA and SI da a we e subjec ed o mul i a ia e PCA (Si ius 8.5 so wa e, Pa e n Recogni ion
Sys ems, Be gen, No way) o assess composi ional di e ences be ween he samples and high-
ligh he ma ke FAs and SIs mainly esponsible o he a ia ion in he da a. P io o he anal-
ysis, FA da a we e a csine (o he squa e oo ) ans o med o imp o e da a no mali y, and all
FA and SI a iables we e s anda dized o p e en la ge componen s om domina ing he anal-
ysis. Since sys ema ic small di e ences in he ela i e concen a ions o die - e lec ing small
componen s o he FA p o ile may ca y equally impo an die a y in o ma ion as he di e -
ences in la ge componen s [54], he s anda diza ion p ocedu e, despi e losing he o iginal
a ios o he di e en FAs, was ega ded as a sound choice. In PCA, he ela i e posi ions o
he samples and a iables we e plo ed using he i s wo p incipal componen s and sepa a-
ions be ween sample g oups we e es ed o s a is ical signi icance by using So Independen
Modelling o Class Analogy (SIMCA) [55,56] and ega ding P<0.05 signi ican . SIMCA is a
supe ised classi ica ion me hod building mul iple PCA-based class models, and as a “so ”
me hod i can classi y a sample in o se e al o e lapping classes. SIMCA uses F- es o e alua e
he sample Euclidean dis ances om he models, and i is ega ded as a obus me hod, which
can be applied o da a ha ing non-no mal dis ibu ion, al hough i pe o ms ideally wi h
da a ha ing no mal dis ibu ion o ans o med o be e no mali y [57]. Howe e , when
he limi ed da a o 11 adul g ey seal males ( he ones wi h accu a ely eco ded backg ound
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 8 / 26
in o ma ion) we e analyzed o chemical ma ke s, we in pa allel ca ied ou he non-me ic
mul idimensional scaling, nMDS (P ime 6, PRIMER-E, Auckland, New Zealand) and analysis
o simila i ies, ANOSIM wi h non- ans o med da a in o de o examine whe he he esul s
o hese s a is ical analyses we e sensi i e o he ype o mul i a ia e me hod chosen.
Resul s
Indica o s o sho - e m die
In he die o subadul males, mo phological HP analysis (n = 37) and DNA me aba coding
(n = 42) iden i ied simila numbe o p ey axa (15 species + 7 axa e sus 18 species + 5 axa,
espec i ely) (Table 1). Bo h me hods iden i ied he ing (46.7% o he consumed mass as
assessed om he HP analysis s 39.8% o he DNA sequences), pe ch (13.6 s 11.0%) and eel-
pou (10.3 s 6.1%) as he mos impo an p ey species. DNA analysis indica ed a ma kedly
highe con ibu ion o sp a , h ee-spined s ickleback Gas e os eus aculea us and cod o he
die compa ed o HP analysis (2.6 s 9.2%, <0.1 s 5.3% and 0.2 s 3.3%, espec i ely). DNA
me aba coding also de ec ed he p esence o impo an die a y species no iden i ied by HP
analysis: b eam Ab amis b ama (6.0%) and bu bo Lo a lo a (3.4%), and o he 5 species wi h a
DNA sequence p opo ion <1% (sand goby Poma ochis us minu us, A lan ic salmon, whi e
b eam Blicca bjoe kna, black goby Gobius nige and ainbow ou Oncho ynchus mykiss, in he
o de o descending p opo ion). The isopod c uc acean Sadu ia en omon was de ec ed by HP
analysis (2.9%) bu no by DNA analysis.
In he die o adul males, mo phological analysis (n = 51) dis inguished highe numbe o
p ey axa han he DNA analysis (n = 44) (23 species + 6 axa s 18 species + 4 axa) (Table 1).
Bo h me hods iden i ied he ing (HP 29.0% s DNA 24.6%) as an impo an p ey. Cyp inids
we e also among he main i ems and he HP analysis es ima ed he p opo ion o oach and
b eam o 5.0 and 3.4%, espec i ely, and o he unde ined cyp inids o 14.4%. DNA me aba -
coding inc eased he axonomic esolu ion o cyp inids in he die , showing a sha e o 10.8%
o b eam, 5.1% o oach, 1.0% o daces Leuciscus sp. and 0.2% o whi e b eam. In addi ion,
HP analysis epo ed a mass p opo ion o 9.2% o common whi e ish while 13.2% o he
DNA sequences belonged o Co egonus species, i.e. common whi e ish o endace. The con i-
bu ion o A lan ic salmon, pikepe ch, sp a , sea ou and u e Gymnocephalus ce nua di e ed
ma kedly be ween he me hods wi h la ge p opo ions indica ed by he DNA analysis (1.5 s
6.2%, 1.5 s 5.8%, 0.5 s. 2.9%, <0.1 s. 1.2%, and <0.1% s. 0.8%, espec i ely). In he adul
males, u bo Scoph halmus maximus, ou -ho ned sculpin Myoxocephalus quad ico nis, com-
mon dab Limanda limanda, ench Tinca inca, black goby and he ben hic isopod Sadu ia
en omon we e only de ec ed by he mo phological analysis (lis ed in he o de o descending
p opo ion).
The HP analysis o emale subadul s (n = 26) dis inguished a sligh ly lowe numbe o p ey
axa (11 species + 4 axa) han DNA analysis (n = 25; 14 species + 4 axa) (Table 2). Bo h me h-
ods iden i ied he ing as he mos impo an die a y species (54.0 s 38.3%). The species o ally
missed in HP analyses we e sand goby, A lan ic salmon and oach wi h DNA sequence p o-
po ions o 4.1, 3.3 and 1.3%, espec i ely. In addi ion, b eam, pikepe ch, ainbow ou and
Co idae species, i.e. sculpins we e also only de ec ed in he DNA analysis bu wi h DNA
sequence p opo ions <1%.
Also in he adul emales, he e we e a iabili y in he axa iden i ied by he HP (n = 15; 8
species + 4 axa) and DNA analysis (n = 14; 9 species + 3 axa) bu s ill he o al numbe o axa
was he same (Table 2). Bo h me hods iden i ied he ing (42.3 HP% s 59.3 DNA%) as he
mos impo an die a y species, wi h eelpou (23.2 s 20.0%), and common whi e ish/Co ego-
nus spp. (12.8 s 7.6%) as o he majo i ems. The small con ibu ions o oach, sp a , b eam,
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 9 / 26
C18 polyunsa u a ed FAs (PUFAs), especially 18:4n-3. A he same ime hese indi iduals had
low pe cen ages o 16:1n-7, 20:1n-7 and 20:4n-6, plen i ul in deme sal ish. Howe e , he SIs
alone showed signi ican di e ence only be ween he awl (T, wi h high δ
34
S) e sus coas al
yke g oups (S and B, wi h high δ
13
C and δ
15
N) when using muscle SIs as a iables. When
using Li e SIs, PCA ollowed by SIMCA showed no s a is ically signi ican di e ences. The
indi iduals ha we e bycaugh in he pelagic su ace o coas al bo om ykes (S s B) did no
Fig 3. PCA sco es plo s o he mid- e m ma ke s A) inne blubbe FA, B) li e SI, and he long- e m ma ke s C)
middle blubbe FA and D) muscle SI da a o g ey seal indi iduals (n = 108) collec ed om ICES-subdi isions 27, 29,
30 and 32. Symbol key is p esen ed below igu es. Loadings plo s o he a iables we e added as inse s. Wi hou any
subg ouping o he indi iduals, acco ding o age o sex, he pai ed SIMCA es s (P<0.05) showed no signi icances in
any compa isons be ween ICES a eas.
h ps://doi.o g/10.1371/jou nal.pone.0208694.g003
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 16 / 26

di e om each o he in any o he compa isons by FAs and SIs, alone o combined. Fu he
compa isons on he sepa a ion powe o FAs and SIs using a la ge se o adul male indi idu-
als a e p esen ed in Fig B in S1 Fig.
Fo consis ency, all he mul i a ia e compa isons we e pe o med by using me ic PCA and
SIMCA. To ensu e ha he esul s o he compa isons using he limi ed da a o 11 adul males
we e no biased, hese PCA and SIMCA compa isons (Fig 4) we e epea ed by non-me ic
nMDS and ANOSIM (S2 Fig). The esul s we e essen ially he same, he nMDS/ANOSIM con-
i ming he s a is ically signi ican sepa a ions indica ed by PCA/SIMCA. The only ma ked
di e ence was ha he non-me ic app oach de ec ed a signi ican di e ence in li e SI p o-
iles o he adul males collec ed om su ace yke and awl (S s T), which acco ding o he
me ic analysis was no signi ican .
Discussion
In he cu en s udy, he powe o di e en me hods o de e mine ma ine mammal die o
eeding ecology was compa ed. S udies using se e al complemen a y me hods, allowing o
con i m he esul s o one die a y p oxy wi h ano he , ha e no been conduc ed on any seal
species in he Bal ic Sea p e iously, and he e a e no s udies applying all he p oxies included
in his wo k om any o he ma ine mammal popula ion ei he . Hence, in he absence o e e -
ence s udies, ou s udy is unique. The i s published su eys o Bal ic g ey seal die used gu
p ey emains and we e based on ma e ial om he 1960-70s [58,59]. Lunds o¨m e al. [8,29]
and Kauhala e al. [9] used ma e ial om he ea ly 2000s. The ea e , Bal ic g ey seal die has
no been s udied in a decade. Howe e , he la ge spa ial and empo al a iabili y o Bal ic ood
webs and ish s ocks [60–63] makes his g ey seal popula ion an excellen model o s udying
he powe o di e en die moni o ing me hods. Pa allel o he ecosys em and ood web change
he g ey seal popula ion is g owing and may adop new o aging habi s.
Sho - e m me hods add ess die a y ish species
Mo phological iden i ica ion o undiges ed p ey emains elies on expe ise and e e ence col-
lec ions bu enables es ima es o inges ed p ey sizes and biomass. DNA me aba coding o gu
Table 5. The accu a e backg ound in o ma ion on 11 adul male g ey seal indi iduals ha we e g ouped acco ding o he ca ching gea ype and used o es he
abili y o he issue chemical ma ke s o indica e di e ences in eeding a ea o p ey ype.
G oup ID Age (y s) Blubbe dep h (mm) ICES a ea Bycaugh in Ta ge ish species
T 1588 16 33 30 awl he ing
T 1613 28 42 30 awl he ing
T 1610 10 60 30 awl he ing
T 1629 19 45 30 awl he ing
S 1606 7 48 32 su ace yke salmonids, common whi e ish
S 1593 10 30 32 su ace yke salmonids, common whi e ish
S 1553 15 21 32 su ace yke salmonids
S 1526 15 27 32 su ace yke salmonids
B 1574 17 36 32 bo om yke pe ch, pikepe ch, cyp inids
B 1598 11 42 29/32�bo om yke pikepe ch
B 1624 10 42 32 bo om yke pe ch, pikepe ch, cyp inids
Indi idual iden i y code (ID), age, blubbe dep h on s e num, ICES-a ea and he gea ype (T = awl, S = su ace yke, B = bo om yke) whe e he seal indi idual was
collec ed om, and he ish species a ge ed wi h he gea a e p esen ed.
�Indi idual 1598 was caugh a he bo de o SD29 and 32.
h ps://doi.o g/10.1371/jou nal.pone.0208694. 005
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 17 / 26
Fig 4. PCA biplo s o A) inne blubbe FA, B) middle blubbe FA, C) li e SI, D) muscle SI, E) combined inne
blubbe FA and li e SI da a, and F) combined middle blubbe FA and muscle SI da a o 11 Finnish adul male g ey
seal indi iduals. Resul s o pai ed SIMCA es s o he s a is ical signi icance o he composi ional di e ences (P<0.05)
a e lis ed in he uppe -le co ne o he plo s (no in panel C). T = bycaugh in awl, S = bycaugh in su ace yke,
B = bycaugh in bo om yke. On he plo , he numbe s combined wi h he gea speci ica ion le e indica e indi idual
seal iden i ica ion code.
h ps://doi.o g/10.1371/jou nal.pone.0208694.g004
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 18 / 26
con en s p o ides mo e exac p ey iden i ica ion bu does no p o ide in o ma ion abou p ey
size, and expe ience in con e ing DNA sequence p opo ions o biomass does no ye exis .
Thus, he es ima ed mass p opo ions om he HP analyses and he DNA sequence p opo -
ions a e no ully compa able o each o he . Despi e hese limi a ions, he DNA me aba cod-
ing clea ly demons a ed p ey axa ha a e unde es ima ed by he HP analyses.
In acco dance wi h he s udy conduc ed by Lunds o¨m e al. [8], he ing was in gene al he
mos equen ly consumed p ey. The diges i e ac ma e ial o Lunds o¨m and co-wo ke s
was collec ed du ing 2001–2005, all seasons and seal age classes included, and i co e ed an
a ea om he Gul o Bo hnia ill he sou he n Bal ic, bu wi hou SD32, included in he cu -
en s udy. Acco ding o he sho - e m me hods applied in his wo k, o he impo an p ey
we e pe ch, eelpou , cyp inids, common whi e ish, sp a and pikepe ch, howe e wi h ma ked
di e ences be ween a eas. Compa ed o he mo phological s udy o Lunds o¨m e al. [8], he
cu en s udy ga e highe die a y sha es o pe ch, pikepe ch and cyp inids likely explained by
he inclusion o samples om SD32, Gul o Finland in his s udy (Tables 3and 4). When com-
pa ing he mo phological HP and DNA da a, Salmo species we e ound o be unde ep e-
sen ed by he HP analysis in male seals (Tables 1,3and 4).
The sho - e m me hods sugges ed die a y di e ences be ween age g oups and a eas, also
ound in he s udy by Lunds o¨m e al. [8]. Since hese me hods a e limi ed in gi ing snapsho
es ima es on he die , no i m conclusions could be made on he die a y di e ences be ween
males and emales due o he low numbe o emale samples. The di e ences in die be ween
a eas desc ibe a shi om he ing dominance in he cen al pa s o pe cid and cyp inid (espe-
cially b eam) dominance in he Gul o Finland, and impo ance o pe ch in he wes e n Bal ic
P ope . The p ey o he Bal ic g ey seals is dis inguished om he die s o g ey seals o he
A lan ic, whe e he ing is a mino die a y componen bu gadoids, la ish and sandeels a e e-
quen ly consumed [64,65].
Chemical issue ma ke s equi e e e ence p ey lib a y
Rega dless o accu a e p ey iden i ica ion, no sho - e m me hod e eals he in eg a ed a e age
p ey o a ee- anging ma ine mammal, which may mig a e and hus a di e en imes exploi
di e en habi a s and p ey. A emp ing o a ain da a o empo al ep esen a i eness om
sho - e m gu samples would equi e equen ly epea ed hun . Die assessmen using FAs and
SIs o e long- e m die a y es ima es bu wi h he d awback o ailing o each i m p ey species
iden i ica ion. Bal ic g ey seals ha e so a been s udied li le o issue FAs [66,67], and issue
SIs ha e p e iously only been s udied o Bal ic inged seals (Phoca hispida bo hnica) [68].
Success ul ood web s udies equi e a ep esen a i e e e ence lib a y o p ey FAs and SIs
and ha he p ey species ha e cha ac e is ically di e en chemical ma ke s. Used oge he ,
FAs and SIs a e complemen a y since hey a e p oxies o di e en die a y componen s. While
he FAs a e de i ed om die a y lipids, he δ
15
N alues and in his wo k also he δ
13
C alues,
analysed by using delipida ed samples [69], ep esen p o eins. FA p o iling has p e iously
p o en o be an e ec i e me hod o s udy p eda o o aging ecology and may e en indica e
speci ic p ey species [17,18] bu he powe o he me hod in such species-le el die de e mina-
ion has also been ques ioned [70]. The habi a -speci ic FA p o iles o ish ha e hei o igin in
plank on o mic oo ganisms media ed by in e eb a es o he issues o he ish [71,72]. The
pelagic ish species he ing, sp a , A lan ic salmon and sea ou ob ained especially high p o-
po ions o C18 PUFAs, he 18:4n-3 being especially abundan in he ing and sp a . The
deme sal species we e dis inguished by he FAs 16:1n-7, 20:1n-7 and 20:4n-6.
Compa isons o SIs o p ey and p eda o p o ides a leas in o ma ion on he ophic le els
o he die a y ish and he a ea hey o igina e om, bu mixing models can also iden i y
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 19 / 26
be ween limi ed numbe o es ablished p incipal p ey axa [73,74]. In gene al, ma ine biomes
end o ha e high δ
15
N and δ
13
C bu high δ
34
S may come om eshwa e inpu , sedimen s,
p ecipi a ions and a ious an h opogenic sou ces and hus may o e in o ma ion on o aging
a ea [46]. Among he Bal ic ish, he δ
34
S was en iched in all s udied Salmo species. To ou
knowledge, he δ
34
S has no been used in he s udies o Bal ic ood webs be o e bu should
become a equen ly analyzed elemen since aqua ic p ima y p oduce s ha e been epo ed o
ha e la ge a ia ion o δ
34
S alues, and hei le el o ophic-s ep ac iona ion is low, which
makes he elemen use ul in e ealing ood web ela ions [75,76]. The ea ly yea s he Salmo
species spen in i e o hei es ua ine die a y sou ces, wi h sulphu supply di e en om he
pelagic sea a eas, may ha e in luence on his alue (Table D in S1 Table).
Al hough he SI a ios in gene al had weake powe han he FAs in g ouping he s udied
ish species acco ding o hei habi a , δ
13
C managed o subg oup pelagic species (all high in
δ
34
S) by sepa a ing he p eda o s salmon and ou (high δ
13
C, high δ
34
S) om he plank i o -
ous he ing and sp a (low δ
13
C, high δ
34
S, Fig 2). This u he sepa a ion likely s ems om he
δ
13
C alue inc easing owa ds highe ophic le els. Acco ding o li e a u e, pelagic species in
gene al ha e highe δ
13
C alues han coas al species, which a e a ec ed by eshwa e un-o
wi h soil o ganic ma e ha ing low δ
13
C alues [46]. In line wi h his, he A lan ic salmons
analyzed o his s udy we e caugh in he Bal ic P ope . The Bal ic deme sal ishes we e
sligh ly en iched in δ
15
N likely due o he ac ha he ophic le el o sedimen bio a is highe
han ha o pelagic plank on [77].
Mid- and long- e m me hods e eal indi idual specializa ion o a ea and
ype o p ey
P o ided ha no p io subg ouping o he seal indi iduals was made o diminish biological
a ia ion, he only chemical issue ma ke ha g ouped he seals acco ding o he s udied
pa ame e s (ICES a ea, sex, age o cause o dea h (bycaugh /hun ed)) was he li e SIs (wi h
he ime window o weeks and e lec ing die a y p o ein componen ), which indica ed ha he
indi iduals om SD27 and wes e n SD30 had simila signa u es di e ing om hose in he
SD29 and SD32 samples. The SD29 seals we e all om he a ea be ween Åland Islands and
Tu ku. This wes -eas pa e n was b oken by 7 seals (6.5%) caugh on he wes coas o SD27
and SD30 bu wi h li e SI signa u es simila o mos o he indi iduals caugh in he eas ,
which sugges s wes wa d mig a ion. These possibly mig a ed indi iduals we e o di e en age
and sex, and had blubbe FA composi ions simila o hose o he o he SD27 and SD30 wes -
e n indi iduals. This lea es simila o aging a ea in he pas , be e indica ed by SIs han FAs
[46,75], as a emp ing uni ying ac o . Since his inding is pu ely chemical ma ke da a-d i en,
we un o una ely ha e no di ec eleme ic o o he p oo ha hese indi iduals would ha e
been mig a ed. In heo y, speci ic loca ions wi h elemen SI cha ac e is ics di e en om he
su oundings may exis . Howe e , he p opo ion o mig a ing seals sugges ed by his s udy is
in he same ange wi h a ecen GPS acking s udy o male g ey seal o he Bal ic, epo ing
ha 12% o he indi iduals mig a ed (mo ed ou om a 60x120 km
2
a ea) du ing a couple o
mon hs’ ime [16].
In he case o blubbe FAs, he lack o dis inc sample g oups in he PCA o all 108 indi idu-
als made i di icul o ecognize he die a y o igin o he a ia ion in blubbe FA p o iles.
Howe e , educing biological a ia ion among he indi iduals s udied may help in ela ing is-
sue FA p o ile di e ences o eeding a ea and die , and indeed among he subadul males he
inne blubbe FA and li e SI p o iles g ouped indi iduals acco ding o he SD a ea hey we e
collec ed in (Fig A in S1 Fig). When s udying ee- anging wild specimens, ideally, he in lu-
ence o die on he chemical ma ke s can be s udied by compa ing he ma ke s be ween g oups
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 20 / 26
o indi iduals o he same gende and age g oup, and collec ed in he same a ea and om same
ype o ishing gea loca ed in simila habi a s. In his s udy, such g oups o adul males, p e-
sumably ha ing simila o aging ecology, had simila issue chemical ma ke s and we e suc-
cess ully g ouped by me ic PCA/SIMCA and non-me ic nMDS/ANOSIM. Despi e he
sample was small, his sugges s indi idual die a y specializa ion no de ec able i he s udied
indi iduals had had oppo unis ic o aging habi s. Thus, he die a y e ec on his ma ke was
no ques ionable. In addi ion, he blubbe FA p o iles o he males om he a ea/gea sub-
g oups we e en iched by he FAs cha ac e is ic o he ish usually caugh in ha a ea by ha
speci ic gea ype. This die a y e ec was sus ained also when using a la ge da ase (Fig B in
S1 Fig). Recen s udies on g ey seal males’ o aging beha iou in he Bal ic ha e shown clea
si e ideli y o he same a ea wi h abou 100 km ange [16]. In SD30, Ko¨nigson e al. [15]
de ec ed specializa ion o adul g ey seal males o salmon ca ching su ace ykes. In addi ion,
based on byca ch s a is ics o he p esen da a, Kauhala e al. [5] classi ied males om all age
g oups as po en ial “p oblem seals”. Conside ing all he die a y p oxies employed in his wo k,
di e en comme cially exploi ed ish species, e.g. he ing, a e consumed by bo h young and
adul seal indi iduals. The issue chemical ma ke s, howe e , e ealed indi idual long- e m
specializa ion o a ce ain habi a and ype o p ey o he adul males, and he DNA analyses,
especially, con i med ha A lan ic salmon was included in hei die . These indings imply
ha he adul males a e he mos likely indi iduals o cause local and sus ained loss o ca ch o
he mos aluable ish, and gea damage.
A ailable in o ma ion on he a es o abso p ion and u no e o FAs and SIs in seal issues
[20,21,75] sugges ha he FA and SI alues o he inne blubbe and li e e lec he die con-
sumed 2–3 weeks p io o sampling, and he chemical signa u es o he middle blubbe and
muscle issue should ep esen die assimila ed up o 2–3 mon hs p io o sampling. Hence, we
p e e he naming “mid- e m” die when inne blubbe FAs and li e SIs a e used in die
assessmen , while “long- e m” die would be desc ibed by middle blubbe FAs and muscle SIs.
Compa ison o hese ma ke p o iles o he adul males om he well-de ined subg oups
e ealed ha he bes sepa a ion powe in hese week–mon h ime scales is ob ained wi h FAs.
Combined use o FA and SI da a as loadings did no imp o e he sepa a ion powe . The FAs
and SI a ios o he adul male g ey seals complied wi h he cha ac e is ic FAs and SI a ios
ound in he key ish o hei ca ching habi a . Fo example, PCA showed a su p isingly s ong
co ela ion be ween he high li e δ
34
S and blubbe 18:4n-3 in he adul male seals bycaugh in
awls. Thus, by his i s use o he δ
34
S o s udy o aging ecology o he seals we could de ine
δ
34
S as a ma ke o consuming SD30 pelagic ish. The ac ha bo h mid- e m and long- e m
ma ke s sepa a ed he seal indi iduals in simila ways sugges s indi idual long- e m specializa-
ion o a ce ain ype o p ey and habi a .
Conclusions
Analysis o gu con en s was equi ed o iden i y p ey species, and bo h mo phological analysis
o p ey ha d pa s and DNA me aba coding showed clea die a y di e ences be ween age
g oups and a eas. Conce ning he o aging ecology o he seals, hese p oxies o he e y ecen
die canno e eal po en ial specializa ion o indi iduals o ce ain eeding a ea o p ey ype.
Fo his pu pose he mid- and long- e m ma ke s can be used. The bycaugh adul males
o med dis inc i e g oups ha ing simila FA and SI ma ke s, which esembled he ma ke pa -
e ns o he ish caugh in he a ea by he gea ype in which he seals we e ound (Fig 4,S2
Fig). A p obable in e p e a ion was ha hese adul males had been using he same o aging
a eas o long, and pe haps aiding he gea s he e epea edly. Since hese g oupings by he
mid- and long- e m die a y ma ke s we e he mos ob ious in he adul males, his is likely a
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 21 / 26

consequence o specialized o aging o male e i o ial beha iou . Selec i e emo al o p ob-
lem seals has been sugges ed o mi iga e he con lic s be ween seals and coas al ishe ies. In he
ligh o he cu en s udy, i implemen ed, such selec i e culling should be di ec ed owa ds he
adul males ha we e ound o be he mos specialized in hei o aging ac ics, and may locally
cause signi ican economic losses o ishe ies in he o m o gea damage and loss o ca ch.
Di e ences in mid- e m die , e lec ing o aging a eas, we e also seen in he li e SIs, which
also may ha e dis inguished a ew mig a ed seals.
This s udy sugges s a combina ion o mul iple die es ima ion me hods as he op imal p o-
ocol o assess as de ailed in o ma ion as possible abou eeding habi s o aqua ic op p eda o s.
E icien use o he die a y me hods, howe e , se s high equi emen s o eco ding de ailed
backg ound in o ma ion on he s udied indi iduals, which is a p e equisi e o disco e ing
die a y subg oups in la ge di e se da a se s.
Suppo ing in o ma ion
S1 Table. Fish a y acid (FA) and s able iso ope (SI) sample numbe s, SIMCA esul s o
ish FA and SI, and ish SI o iginal alues.
(PDF)
S1 Fig. PCA and SIMCA o issue FAs and SIs using da a subse s o subadul and adul
male g ey seals.
(PDF)
S2 Fig. MDS and ANOSIM analyses o he issue chemical ma ke s in 11 adul males wi h
accu a e backg ound in o ma ion.
(PDF)
S1 Da a. O iginal ish and seal da a.
(XLSX)
Acknowledgmen s
We hank he Na ional Resou ces Ins i u e o Finland (LUKE), Swedish Uni e si y o Ag icul u al
Sciences (SLU), Swedish Museum o Na u al His o y (SMNH) and Es onian Ma ine Ins i u e o
collec ion o ish and seal samples. Cha lo a Mo aeus a SMNH p o ided aluable suppo in col-
lec ing and au opsying he Swedish seals. Sonja Myllyla¨is hanked o he skilled assis ance in he
SI labo a o y, and Ma kus Ahola (LUKE) o c i ically e iewing he manusc ip .
Au ho Con ibu ions
Concep ualiza ion: Malin T e in, Rod igo Espa za-Salas, Annika S o¨mbe g, Pa ik Tang,
Ii is Kokkonen, Annika He e o, Kaa ina Kauhala, Olle Ka lsson, Raisa Tiilikainen, Ma -
kus Ve emaa, Tuula Sinisalo, Reijo Ka¨kela¨, Ka l Lunds o¨m.
Da a cu a ion: Malin T e in, Rod igo Espa za-Salas, Annika S o¨mbe g, Pa ik Tang, Ii is
Kokkonen, Annika He e o, Tuula Sinisalo, Reijo Ka¨kela¨, Ka l Lunds o¨m.
Fo mal analysis: Malin T e in, Rod igo Espa za-Salas, Tuula Sinisalo, Reijo Ka¨kela¨, Ka l
Lunds o¨m.
In es iga ion: Malin T e in, Rod igo Espa za-Salas, Annika S o¨mbe g, Pa ik Tang, Ii is
Kokkonen, Annika He e o, Kaa ina Kauhala, Ma kus Ve emaa, Tuula Sinisalo, Reijo
Ka¨kela¨, Ka l Lunds o¨m.
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 22 / 26
Me hodology: Malin T e in, Rod igo Espa za-Salas, Tuula Sinisalo, Reijo Ka¨kela¨, Ka l
Lunds o¨m.
P ojec adminis a ion: Malin T e in, Raisa Tiilikainen, Reijo Ka¨kela¨, Ka l Lunds o¨m.
Visualiza ion: Malin T e in, Reijo Ka¨kela¨, Ka l Lunds o¨m.
W i ing – o iginal d a : Malin T e in, Rod igo Espa za-Salas, Tuula Sinisalo, Reijo Ka¨kela¨,
Ka l Lunds o¨m.
W i ing – e iew & edi ing: Malin T e in, Rod igo Espa za-Salas, Annika S o¨mbe g, Pa ik
Tang, Ii is Kokkonen, Annika He e o, Kaa ina Kauhala, Olle Ka lsson, Raisa Tiilikainen,
Ma kus Ve emaa, Tuula Sinisalo, Reijo Ka¨kela¨, Ka l Lunds o¨m.
Re e ences
1. Bu le JRA, Middlemas SJ, G aham IM, Ha is RN. Pe cep ions and cos s o seal impac s on A lan ic
salmon ishe ies in he Mo ay Fi h, Sco land: Implica ions o he adap i e co-managemen o seal- ish-
e y con lic . Ma Policy. 2011; 35:317–323. h ps://doi.o g/10.1016/j.ma pol.2010.10.011.
2. Bowen WD, Lidga d D. Ma ine mammal culling p og ams: e iew o e ec s on p eda o and p ey popu-
la ions. Mamm Re . 2013; 43:207–220. h ps://doi.o g/10.1111/j.1365-2907.2012.00217.x.
3. Hå ding KC, Ha
¨ ko
¨nen TJ. De elopmen in he Bal ic g ey seal (Halichoe us g ypus) and inged seal
(Phoca hispida) popula ions du ing he 20 h cen u y. Ambio. 1999; 28:619–625.
4. Helcom. h p://www.helcom. i/bal ic-sea- ends/da a-maps/biodi e si y/seals21.6.2017. 2017.
5. Kauhala K, Ku kilah i M, Ahola MP, He e o A, Ka lsson O, Kunnas an a M, e al. Age, sex and body
condi ion o Bal ic g ey seals: A e p oblem seals a andom sample o he popula ion? Ann Zool Fennici.
2015; 52:103–114. h ps://doi.o g/10.5735/086.052.0209.
6. Vanha alo J, Ve emaa M, He e o A, Aho T, Tiilikainen R. By-ca ch o g ey seals (Halichoe us g ypus)
in Bal ic ishe ies – A Bayesian analysis o in e iew su ey. PloS ONE. 2014; 9. h ps://doi.o g/10.
1371/jou nal.pone.0113836.
7. G aham IM, Ha is RN, Ma ejuso a I, Middlemas SJ. Do ’ ogue’ seals exis ? Implica ions o seal
conse a ion in he UK. Anim Conse . 2011; 14:587–98. h ps://doi.o g/10.1111/j.1469-1795.2011.
00469.x.
8. Lunds o
¨m K, Hje ne O, Lunne yd SG, Ka lsson O. Unde s anding he die composi ion o ma ine mam-
mals: g ey seals (Halichoe us g ypus) in he Bal ic Sea. ICES J Ma Sci. 2010; 67:1230–1239. h ps://
doi.o g/10.1093/icesjms/ sq022.
9. Kauhala K, Kuunas an a M, Val onen M. Die o g ey seals in Finland in 2001–2007: a p elimina y
s udy. Hallien a in o Suomen me ialueella 2001–2007 –alus a a sel i ys. In Finnish wi h English sum-
ma y. Suomen Riis a. 2011; 57:73–83.
10. Suu onen P, Leh onen E. The ole o salmonids in he die o g ey and inged seals in he Bo hnian Bay,
no he n Bal ic Sea. Fish Res. 2012; 125:283–288. h ps://doi.o g/10.1016/j. ish es.2012.03.007.
11. Jobling M, B eiby A. The use and abuse o ish o oli hs in s udies o eeding habi s o ma ine pisci o es.
Sa sia. 1986; 71:265–274. h ps://doi.o g/10.1080/00364827.1986.10419696.
12. Me
´heus E, Al onsi E, Le Me
´nec P, Hassani S, Jung J-L. DNA ba coding o he iden i ica ion o so
emains o p ey in he s omach con en s o g ey seals (Halichoe us g ypus) and ha bou po poises
(Phocoena phocoena). Ma Biol Res. 2015; 11:385–395. h ps://doi.o g/10.1080/17451000.2014.
943240.
13. Oja ee H, Jaanus A, MacKenzie BR, Ma in G, Olenin S, Radziejewska T, e al. S a us o Biodi e si y
in he Bal ic Sea. PLoS ONE. 2010; 5:e12467. h ps://doi.o g/10.1371/jou nal.pone.0012467. PMID:
20824189
14. Beck CA, I e son SJ, Bowen WD, Blancha d W. Sex di e ences in g ey seal die e lec seasonal a ia-
ion in o aging beha iou and ep oduc i e expendi u e: e idence om quan i a i e a y acid signa u e
analysis. J Anim Ecol. 2007; 76:490–502. h ps://doi.o g/10.1111/j.1365-2656.2007.01215.x. PMID:
17439466
15. Ko
¨nigson S, Fja
¨lling A, Be glind M, Lunne yd SG. Male g ay seals specialize in aiding salmon aps.
Fish Res. 2013; 148:117–123. h ps://doi.o g/10.1016/j. ish es.2013.07.014.
16. Oksanen SM, Ahola MP, Leh onen E, Kunnas an a M. Using mo emen da a o Bal ic g ey seals o
examine o aging-si e ideli y: implica ions o seal- ishe y con lic mi iga ion. Ma Ecol P og Se . 2014;
507:297–308. h ps://doi.o g/10.3354/meps10846.
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 23 / 26
17. B adshaw CJA, Hindell MA, Bes NJ, Phillips KL, Wilson G, Nichols PD. You a e wha you ea : desc ib-
ing he o aging ecology o sou he n elephan seals (Mi ounga leonina) using blubbe a y acids. P oc
Biol Sci. 2003; 270:1283–1292. h ps://doi.o g/10.1098/ spb.2003.2371. PMID: 12816642
18. Budge SM, I e son SJ, Koopman HN. S udying ophic ecology in ma ine ecosys ems using a y acids:
A p ime on analysis and in e p e a ion. Ma Mamm Sci. 2006; 22:759–801. h ps://doi.o g/10.1111/j.
1748-7692.2006.00079.x.
19. I ing L, Ha JS. The me abolism and insula ion o seals as ba e-skinned mammals in cold wa e . Can
J Zool. 1957; 35:497–511. h ps://doi.o g/10.1139/z57-041.
20. S andbe g U, Ka
¨kela
¨A, Lyde sen C, Ko acs KM, G ahl-Nielsen O, Hy a
¨ inen H, e al. S a i ica ion,
composi ion, and unc ion o ma ine mammal blubbe : he ecology o a y acids in ma ine mammals.
Physiol Biochem Zool. 2008; 81:473–485. h ps://doi.o g/10.1086/589108. PMID: 18532882
21. S andbe g U, Sipila
¨T, Koskela J, Kunnas an a M, Ka
¨kela
¨R. Ve ical a y acid p o iles in blubbe o a
eshwa e inged seal – Compa ison o a ma ine ela i e. J Exp Ma Bio Ecol. 2011; 407:256–265.
h ps://doi.o g/10.1016/j.jembe.2011.06.021.
22. Budge SM, Coope MH, I e son SJ. Demons a ion o he deposi ion and modi ica on o die a y a y
acids in pinniped blubbe using adiolabelled p ecu so s. Physiol Biochem Zool. 2004; 77:682–687.
h ps://doi.o g/10.1086/420945. PMID: 15449239
23. Tieszen LL, Bou on TW, Tesdahl KG, Slade NA. F ac iona ion and u no e o s able ca bon iso opes
in animal issues—implica ions o a δ13C analysis o die . Oecologia. 1983; 57:32–37. h ps://doi.o g/
10.1007/b 00379558. PMID: 28310153
24. Dale um F, Ange bjo
¨ n A. Resol ing empo al a ia ion in e eb a e die s using na u ally occu ing s a-
ble iso opes. Oecologia. 2005; 144:647–658. h ps://doi.o g/10.1007/s00442-005-0118-0. PMID:
16041545
25. I e son SJ, Sp inge AM, Ki aysky AS. Seabi ds as indica o s o ood web s uc u e and ecosys em a i-
abili y: quali a i e and quan i a i e die analyses using a y acids. Ma Ecol P og Se . 2007; 352:235–
244. h ps://doi.o g/10.3354/meps07073.
26. Engelha d GH, Peck MA, Rindo A, Smou S, an Deu s M, Raab K, e al. Fo age ish, hei ishe ies,
and hei p eda o s: who d i es whom? ICES J Ma Sci. 2014; 71:90–104. h ps://doi.o g/10.1093/
icesjms/ s 087.
27. Hewe HR. The de e mina ion o age, in he g ey seal (Halichoe us g ypus) sexual ma u i y, longe i y
and a li e- able. P oc Zool Soc Lond. 1964; 142:593–623. h ps://doi.o g/10.1111/j.1469-7998.1964.
b04631.x.
28. Ha ding KC, Ha
¨ ko
¨nen T, Helande B, Ka lsson O. S a us o Bal ic g ey seals: Popula ion assessmen
and ex inc ion isk. NAMMCO Sci Publ. 2007; 6:33–56. h ps://doi.o g/10.7557/3.2720.
29. Lunds o
¨m K, Hje ne O, Alexande sson K, Ka lsson O. Es ima ion o g ey seal (Halichoe us g ypus)
die composi ion in he Bal ic sea. NAMMCO Sci Publ. 2007; 6:177–196. h ps://doi.o g/10.7557/3.
2733.
30. Ha
¨ ko
¨nen T. Guide o he o oli hs o he bony ishes o he no heas A lan ic. Danibu ApS, Helle up,
Denma k. 256 pp. 1986.
31. Leopold MF, an Damme CJG, Philippa CJM, Win e CJN. O oli hs o No h Sea ish: ish iden i ica ion
key by means o o oli hs and o he ha d pa s. Wo ld Biodi e si y Da abase CD-ROM Se ies. Expe
Cen e o Taxonomic Iden i ica ion (ETI), Ams e dam. 2001.
32. Binladen J, Gilbe MTP, Bollback JP, Pani z F, Bendixen C, Nielsen R, e al. The use o coded PCR
p ime s enables high- h oughpu sequencing o mul iple homolog ampli ica ion p oduc s by 454 pa allel
sequencing. PloS ONE. 2007; 2:e197. h ps://doi.o g/10.1371/jou nal.pone.0000197. PMID: 17299583
33. Blankenbe g D, Kus e GV, Co ao N, Ananda G, Laza us R, Mangan M, e al. Galaxy: a web-based
genome analysis ool o expe imen alis s. Cu en P o ocols in Molecula Biology: John Wiley & Sons,
Inc; 2001.
34. Schmiede R, Edwa ds R. Quali y con ol and p ep ocessing o me agenomic da ase s. Bioin o ma ics.
2011; 27:863–864. h ps://doi.o g/10.1093/bioin o ma ics/b 026. PMID: 21278185
35. Camacho C, Coulou is G, A agyan V, Ma N, Papadopoulos J, Beale K, e al. BLAST+: a chi ec u e
and applica ions. BMC Bioin o ma ics. 2009; 10:421. h ps://doi.o g/10.1186/1471-2105-10-421.
PMID: 20003500
36. Deagle BE, Chia adia A, McInnes J, Ja man SN. Py osequencing aecal DNA o de e mine die o li le
penguins: is wha goes in wha comes ou ? Conse Gene . 2010; 11:2039–2048. h ps://doi.o g/10.
1007/s10592-010-0096-6.
37. Thomas AC, Ja man SN, Haman KH, T i es AW, Deagle BE. Imp o ing accu acy o DNA die es ima es
using ood issue con ol ma e ials and an e alua ion o p oxies o diges ion bias. Mol Ecol. 2014;
23:3706–3718. h ps://doi.o g/10.1111/mec.12523. PMID: 24102760
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 24 / 26
38. Ka
¨kela
¨R, Ka
¨kela
¨A, Kahle S, Becke PH, Kelly A, Fu ness RW. Fa y acid signa u es in plasma o cap-
i e he ing gulls as indica o s o deme sal o pelagic ish die . Ma Ecol P og Se 2005; 293:191–200.
h ps://doi.o g/10.3354/meps293191.
39. F y B. S able iso ope ecology: Sp inge , New Yo k, NY, USA; 2006.
40. C aw o d K, McDonald RA, Bea hop S. Applica ions o s able iso ope echniques o he ecology o
mammals. Mamm Re . 2008; 38:87–107. h ps://doi.o g/10.1111/j.1365-2907.2008.00120.x.
41. Bond AL, Hobson KA. Repo ing s able-iso ope a ios in ecology: ecommended e minology, guidelines
and bes p ac ices. Wa e bi ds. 2012; 35:324–331. h ps://doi.o g/10.1675/063.035.0213.
42. Focken U, Becke K. Me abolic ac iona ion o s able ca bon iso opes: implica ions o di e en p oxi-
ma e composi ions o s udies o he aqua ic ood webs using δ
13
C da a. Oecologia. 1998; 115:337–
343. h ps://doi.o g/10.1007/s004420050525. PMID: 28308424
43. DeNi o MJ, Eps ein S. In luence o die on he dis ibu ion o ca bon iso opes in animals. Geochim Cos-
mochim Ac a. 1978; 42:495–506. h ps://doi.o g/10.1016/0016-7037(78)90199-0.
44. Kiljunen M, G ey J, Sinisalo T, Ha od C, Immonen H, Jones RI. A e ised model o lipid-no malizing
δ13C alues om aqua ic o ganisms, wi h implica ions o iso ope mixing models. J Appl Ecol. 2006;
43:1213–1222. h ps://doi.o g/10.1111/j.1365-2664.2006.01224.x.
45. S andbe g U, Sipila
¨T, Koskela J, Kunnas an a M, Ka
¨kela
¨R. Ve ical a y acid p o iles in blubbe o a
eshwa e inged seal – Compa ison o a ma ine ela i e. J Exp Ma Bio Ecol. 2011; 407:256–265.
h ps://doi.o g/10.1016/j.jembe.2011.06.021.
46. Rubens ein DR, Hobson KA. F om bi ds o bu e lies: animal mo emen pa e ns and s able iso opes.
T ends Ecol E ol 2004; 19:256–263. h ps://doi.o g/10.1016/j. ee.2004.03.017. PMID: 16701265
47. Tieszen LL, Bou on TW, Tesdahl KG, Slade NA. F ac iona ion and u no e o s able ca bon iso opes
in animal issues—impica ions o a δ
13
C analysis o die . Oecologia. 1983; 57:32–37. h ps://doi.o g/
10.1007/b 00379558. PMID: 28310153
48. Hilde b and GV, Fa ley SD, Robbins CT, Hanley TA, Ti us K, Se heen C. Use o s able iso opes o
de e mine die s o li ing and ex inc bea s. Can J Zool. 1996; 74:2080–2088. h ps://doi.o g/10.1139/
z96-236.
49. Ku le CM, Wo hy GAJ. S able ni ogen and ca bon iso ope a ios in mul iple issues o he no he n u
seal Callo hinus u sinus: implica ions o die a y and mig a o y econs uc ions. Ma Ecol P og Se .
2002; 236:289–300. h ps://doi.o g/10.3354/meps236289.
50. Phillips DL, Eld idge PM. Es ima ing he iming o die shi s using s able iso opes. Oecologia. 2006;
147:195–203. h ps://doi.o g/10.1007/s00442-005-0292-0. PMID: 16341714
51. Coope MH, I e son SJ, He as H. Dynamics o blood chylomic on a y acids in a ma ine ca ni o e:
implica ions o lipid me abolism and quan i a i e es ima ion o p eda o die s. J Comp Physiol B. 2005;
175:133–145. h ps://doi.o g/10.1007/s00360-004-0469-6. PMID: 15657738
52. F ayn KN, A ne P, Yki-Ja
¨ inen H. Fa y acid me abolism in adipose issue, muscle and li e in heal h
and disease. Essays Biochem. 2006; 42:89–103. h ps://doi.o g/10.1042/bse0420089. PMID:
17144882
53. No ds om CA, Wilson LJ, I e son SJ, Tolli DJ. E alua ing quan i a i e a y acid signa u e analysis
(QFASA) using ha bou seals Phoca i ulina icha dsi in cap i e eeding s udies. Ma Ecol P og Se .
2008; 360:245–263. h ps://doi.o g/10.3354/meps07378.
54. Ka
¨kela
¨R, Fu ness RW, Kahle S, Becke PH, Ka
¨kela
¨A. Fa y acid signa u es in seabi d plasma a e a
complex unc ion o die composi ion: a cap i e eeding ial wi h he ing gulls. Func Ecol. 2009;
23:141–149. h ps://doi.o g/10.1111/j.1365-2435.2008.01475.x.
55. Wold S, Sjo
¨s o
¨m M. SIMCA: A me hod o analyzing chemical da a in e ms o simila i y and analogy.
In: Kowalski B, edi o . Chemome ics: Theo y and Applica ion. ACS Symposium Se ies. 52. Washing-
on DC: Ame ican Chemical Socie y; 1977. p. 243–82.
56. Dunn WJ III, Wold S. Rela ionships be ween chemical s uc u e and biological ac i i y modeled by
SIMCA pa e n ecogni ion. Bioo ganic Chemis y. 1980; 9:505–523.
57. Ch is ie OHJ, Wold S. The in luence o skewed measu emen dis ibu ions commonly ound in geo-
chemis y. Anal Le 1979; 12: 979–993. h ps://doi.o g/10.1080/00032717908059777.
58. To moso DD, Rez o GV. In o ma ion on he dis ibu ion, numbe and eeding habi s o inged and
g ey seals in he gul s o Finland and Riga in he Bal ic Sea. Finnish Game Resea ch. 1978; 37:14–7.
59. So
¨de be g S. Feeding habi s and comme cial damage o seals in he Bal ic. In: P oceedings o he Sym-
posium on he seal in he Bal ic, June 4–6, 1974, Lidingo
¨, Sweden Na ional Swedish En i onmen P o-
ec ion Agency, S ockholm. 1975; PM 591:66–78.
60. Wasmund N, Uhlig S. Phy oplank on ends in he Bal ic Sea. ICES J Ma Sci. 2003; 60:177–186.
h ps://doi.o g/10.1016/S1054-3139(02)00280-1.
Using mul iple die a y me hods
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0208694 Janua y 2, 2019 25 / 26