Pelagic habitat and offspring survival in the eastern stock of Atlantic bluefin tuna

Pelagic habi a and o sp ing su i al in he eas e n s ock o A lan ic bluefin una Pa icia Regle o 1, *, Rosa Balbı´n 1 , F anciso Ja ie Abascal 2 , An onio Medina 3 , Diego Al a ez-Be as egui 4 , Lei Rasmuson 5,6 , Bap is e Mou e 4 ,Sa´ma Sabe 7 , Au elio O ega 8 , Edu ne Blanco 1 , Fe nando de la Ga´nda a 8 , F anciso Ja ie Alemany 1 , G. Wal e Ing am J 9 , and Manuel Hidalgo 1 1 Ins i u o Espa~ nol de Oceanog a ı´a, Cen e Oceanog a`fic de les Balea s, Ecosys em Oceanog aphy G oup (GRECO), Moll de Ponen s/n, 07015 Palma de Mallo ca, Spain 2 Cen o Oceanog a´fico de Cana ias, Ins i u o Espa~ nol de Oceanog a ı´a, 38180 San a C uz de Tene i e, Spain 3 Depa amen o de Biologı´a, Facul ad de Ciencias del Ma y Ambien ales, A da. Repu´blica Saha aui, s/n, 11510 Pue o Real, Ca´diz, Spain 4 Balea ic Islands Coas al Obse ing and Fo ecas ing Sys em, Pa c Bi , Nao e, Bloc A 28p. p a 3, Palma de Mallo ca, Balea ic Islands, Spain 5 Ma ine Resou ces P og am, O egon Depa men o Fish and Wildli e, 2040 SE Ma ine Science D i e, Newpo , OR, USA 6 NOAA, Na ional Ma ine Fishe ies Se ice, Sou heas Fishe ies Science Cen e , Miami Labo a o ies, 75 Vi ginia Beach D i e, Miami, FL, USA 7 Ins i u o Espa~ nol de Oceanog a ı´a, Cen o Oceanog a´fico de Ma´laga, Pue o Pesque o, s/n, 29640 Fuengi ola, Spain 8 Ins i u o Espa~ nol de Oceanog a ı´a, Cen o Oceanog a´fico de Mu cia, 30860 Pue o de Maza o´n, Mu cia, Spain 9 NOAA, Na ional Ma ine Fishe ies Se ice, Sou heas Fishe ies Science Cen e , Mississippi Labo a o ies, 3209 F ede ic S ee , Pascagoula, MS 39567, USA *Co esponding au ho : el: þ34 971 133 720; e-mail: [email p o ec ed]. Regle o, P., Balbı´n, R., Abascal, F. J., Medina, A., Al a ez-Be as egui, D., Rasmuson, L., Mou e, B., Sabe , S., O ega, A., Blanco, E., de la Ga´nda a, F., Alemany, F. J., Ing am, G. W., Hidalgo, M. Pelagic habi a and o sp ing su i al in he eas e n s ock o A lan ic bluefin una. – ICES Jou nal o Ma ine Science, doi:10.1093/icesjms/ sy135. Recei ed 14 Ma ch 2018; e ised 16 Augus 2018; accep ed 20 Augus 2018. In his manusc ip , we es how an unde s anding o geog aphical a ia ion in la al fi ness in ela ion o empe a u e and habi a use could be a use ul me hod o imp o e ou unde s anding o ec ui men and de elop be e indices o annual ec ui men . On he basis o he assump ion ha g ow h and su i al o una la ae a e influenced by empe a u e, we ha e de eloped a po en ial la al su i al index o A lan ic bluefin una (Thunnus hyn- nus) by combining empi ical da a om egg and la al ea ing expe imen s wi h empe a u e da a om hyd odynamic models. The expe imen s we e designed o es he ull ange o empe a u e a iabili y ha bluefin la ae would expe ience in he field and p o ide a mechanis ic unde s anding o he p ocesses d i ing egg and la al su i al. We hen de eloped a biological model using he empe a u e- ela ed g ow h exp essions and a size- dependen su i al unc ion o he la ae. The biological model was applied o a ime-se ies o spa ially explici empe a u e da a o he wes e n Medi e anean om he S ai o Gib al a o 6E, which includes he majo ecognized bluefin una eas e n s ock spawning a ea, he Balea ic Sea. Ou esul s show ha a eas wi h high p obabili ies o la al su i al coincide wi h hose ha would be conside ed as op imal based on o he da a sou ces (ich hyoplank on su eys, spawning emale loca ions om comme cial fishe ies da a, and adul acking da a). Howe e , e idence o spawning has been ound in a eas wi h subop imal he mal habi a s, as p edic ed by he model, which we discuss ega ding sampling e o and salini y on s. The e was a good ma ch be ween he su i al index and ec ui men indices om s anda dized CPUE fishe ies da a. These esul s ha e implica ions o ou unde s anding o he ec ui men p ocess o he eas e n s ock o A lan ic bluefin una, since hey sugges ha he combined e ec s o empo al and spa ial a iabili y o he en i onmen d i e ec ui men success, which has impo an implica ions o he managemen o he species. Keywo ds: A lan ic bluefin una, fish la ae ec ui men , Medi e anean Sea, spawning s ock, empe a u e V CIn e na ional Council o he Explo a ion o he Sea 2018. All igh s ese ed. Fo pe missions, please email: [email p o ec ed] ICES Jou nal o Ma ine Science (2018), doi:10.1093/icesjms/ sy135 Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018 In oduc ion The en i onmen s in which ish ep oduce usually p o ide a ou able ood condi ions o he la ae, sui able empe a u es o egg and la al de elopmen , and/o sui able e en ion a eas (Mullon e al., 2002). Gi en he di icul ies in di ec ly s udying bo h spawning and ec ui men p ocesses o many species, pa - icula ly la ge pelagic ish, a knowledge o he e ec s o en i on- men al a iabili y on egg and la al g ow h and su i al a es can be used al e na i ely o o ecas spawning loca ions, ep oduc i e ou pu , and/o ec ui men s eng h. Mos ishe ies assessmen models and much o ishe ies heo y is p edica ed on a ela ion- ship be ween he numbe o spawne s and he e en ual ec ui s ha hey p oduce. Al hough he e is s ong e idence o en i on- men ally d i en ec ui men (Ve -P e e al., 2013;Szuwalski e al., 2015), en i onmen al in o ma ion is a ely included in s ock assessmen s o la ge pelagic ish. Unde s anding and moni- o ing he physical and biological p ocesses ha con ol ec ui - men , he ecological co ne s one o de e mining ish popula ion dynamics, is essen ial o imp o ing cu en ishe ies s ock assess- men s. Pa icula ly, he e is a need o ake in o accoun geog aph- ical a iabili y in o sp ing i ness i spa ial managemen decisions a e o be aken in he u u e (Be ge e al., 2017). A lan ic blue in una (Thunnus hynnus) a e managed by he In e na ional Commission o he Conse a ion o A lan ic Tunas (ICCAT) as wo sepa a e s ocks: an eas e n and a wes e n s ock. As wi h mos assessed ish, es ima es o ec ui men and he unc- ional shape o he s ock– ec ui ela ionship a e majo sou ces o unce ain y. The e is no consensus among expe s on which ype o s ock– ec ui ela ionship o use, and ends in ec ui men o he wo s ocks ha e been sugges ed o be co ela ed ei he wi h spawning biomass o en i onmen al a iabili y (e.g. see e e ences in Po ch and Lau e a, 2016). Fo he eas e n s ock, he unc ional o m o he s ock– ec ui ela ionship emains elusi e, as e i- denced by he clea asynch onous luc ua ion be ween spawning biomass and ec ui men a a mul idecadal scale (Anonymous, 2017;Figu e 1), wi h high spawning biomass being associa ed wi h low ec ui men e en s un il he 1990s, low spawning bio- mass associa ed wi h high ec ui men e en s du ing he 1990– 2000s, and again high spawning biomass associa ed wi h low e- c ui men e en s since 2008. On he o he hand, he inc easing end in annual ec ui men om he 1970s o he ea ly 2000s has been ela ed o inc easing summe empe a u es in he Balea ic Sea (Ha o d e al., 2017), which sugges s a signi ican en i onmen – ec ui men ela ionship. Howe e , he scena io seems o be mo e complex since es ima es om he la es assess- men show ha ec ui men is dec easing despi e he con inual wa ming end (Anonymous, 2017;Figu e 1). This scena io sug- ges s a po en ial densi y-dependen con ol a la ge empo al scales ha , in combina ion wi h la ge-scale inc easing ends in wa e empe a u e due o clima e wa ming in he Medi e anean, may explain long- e m luc ua ions in ec ui men . Howe e , sho - e m a ia ions, pa icula ly he in e annual a ia ion ha has occu ed in he las decade, challenge he hypo hesis o a di- ec he mal in luence on in e annual a iabili y in ec ui men (Figu e 1). The cu en challenge o de eloping a model o he ec ui - men dynamics o A lan ic blue in una is o unde s and he mechanis ic in luence o egional en i onmen al d i e s ha link spa ial pa e ns o la al pe o mance o in e annual a iabili y in ec ui men . We applied labo a o y-de i ed, empe a u e- dependen g ow h and su i al ela ionships o A lan ic blue in una eggs and la ae o empe a u e eco ds om he wes e n Medi e anean du ing he spawning season. We hen used hese da a o de elop an annual index o po en ial la al su i al in one o he p ima y spawning a eas o he eas e n s ock o A lan ic blue in una. We es ed he hypo hesis ha he mal condi ions which p omo e g ow h and su i al o A lan ic blue in eggs and la ae coincide wi h he loca ion o a majo spawning g ound a ound he Balea ic Islands in he wes e n Medi e anean basin. Fu he , we hypo hesize ha hese he mal condi ions a e associ- a ed wi h empo al a ia ion in annual ec ui men . We compa ed ou index o a ailable indices o ec ui men om ishe ies-dependen da a. Ou inal goal was o p oduce mechanis ic-based knowledge ha will allow scien is s o include en i onmen al a iabili y and i s spa ial ends in o u u e s ock assessmen s o A lan ic blue in una. Ma e ial and me hods S udy si e and ime-se ies A lan ic blue in una a e managed sepa a ely as wo s ocks. The eas e n s ock is de ined as being eas o he 45W me idian (Figu e 2a). This s udy ocuses on he wes e n Medi e anean po ion o he eas e n s ock (Figu e 2b). The majo spawning and b eeding a ea o he eas e n s ock o A lan ic blue in una (Alemany e al., 2010;Ha o d e al., 2017) is he Balea ic Sea, lo- ca ed in he cen e o he wes e n hal o he wes e n Medi e anean Sea (Figu e 2c). The da a analysed in his s udy we e om 2000 o 2014, he same yea s o which de ailed moni- o ing o he spa ial dis ibu ion o la ae, spawne s, and acking da a o A lan ic blue in una in he wes e n Medi e anean we e a ailable o compa ison o he model esul s. This ime-se ies includes yea s wi h a iabili y in he mal condi ions, ec ui men , and spawning-s ock biomass (Figu e 1). Time-se ies o spa ially explici en i onmen al da a The selec ion o en i onmen al da a o use in his s udy was based on he abili y o he a ailable oceanog aphic models o ep oduce he CTD obse a ions collec ed du ing 2000–2014. Using da a om summe 2004, 72 CTD obse a ions (Alemany e al., 2010) we e compa ed wi h empe a u e and salini y ields a 5-m dep h om ou di e en oceanog aphic models o es how well he models ep esen ed he spa ial a iabili y o hese a iables. The ou di e en models include wo e sions o he Medi e anean model a ailable om he Cope nicus Ma ine Se ice (CMEMS- MED e sion 02 and s 03, h p://ma ine.cope nicus.eu), he GLORYS2V1 e sion o MERCATOR (h ps://www.me ca o - ocean. ), and he MEDAR-MEDATLAS summe clima ological alues. Model da a we e ex ac ed om he g id-poin ha con- ained he CTD s a ion. In his way, i was possible o cons uc a ime-se ies o each model. The ou model ime-se ies we e com- pa ed wi h he CTD obse a ions using a Taylo diag am (Taylo , 2001). Dynamic heigh (DH) was calcula ed by e ically in eg a - ing he speci ic olume ob ained om he CTD da a, using 600 m as he le el o no mo ion, as desc ibed by Balbı´n e al. (2014). The same me hodology was applied o he model and clima ological da a. 2P. Regle o e al. Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018 Compa ison o empe a u es om he model da a o CTD obse a ions indica es ha bo h o he CMEMS-MED model e - sions we e he bes a eplica ing he spa ial pa e n o he CTD obse a ions (wi h a co ela ion coe icien close o 0.9 and a cen ed Roo -Mean-Squa e di e ence a ound 0.4) (Supplemen a y Figu e S1). This indica es ha he ho izon al s uc u e o empe a u e a iabili y was well eplica ed. In he case o salini y, he MEDAR-MEDATLAS clima ological alues we e he mos s ongly co ela ed wi h he obse a ions (wi h a co ela ion coe icien o 0.6), while he CMEMS-MED models p esen alues o only a ound 0.4 and a cen ed Roo -Mean- Squa e di e ence a ound 0.3(Supplemen a y Figu e S2). The e o e, salini y was no well ep esen ed in he models and could no be used o calcula ions o he loca ion o on al a eas. Taylo analysis o DH showed ha obse a ions om he MEDAR-MEDATLAS we e co ela ed wi h obse a ions, wi h co ela ion coe icien s o 0.65, and o all he o he models wi h co ela ions coe icien s below 0.35 ( esul s no shown). The e o e, he DH o mesoscale s uc u es was no well ep e- sen ed in he models, demons a ing ha geos ophic cu en s de i ed om he models we e no usable o de ailed calcula ions o possible d i e ec s. As such, only empe a u e da a we e well ep esen ed by he a ailable models o he pe iod o s udy (2000–2014) and, he e o e, only he empe a u e model om CMEMS-MED was used. We compa ed he a e age empe a u e be ween 1 and 15 June each yea om 2002 o 2014 in an a ea cen ed in he Balea ic Islands (0.5–4.5E 38.2–40.5N) using he CMEMS-MED model and sa elli e da a (L4 JPL MUR p oduc , only a ailable since 2002). This compa ison shows ha he CMEMS-MED model eplica es e y well he in e annual a iabili y o empe a u e in he a ea (Supplemen a y Figu e S3). The e o e, he model can be used o model he empo al and spa ial a ia ions in empe a u e. Gi en he low capabili y o he model o ep esen salini y, we used maps o salini y de i ed om ield c uises o in es iga e he dis ibu ion o la ae and adul s ela i e o his a iable wi hin he spa ial co e age o he c uises cen ed in he a ea a ound he Balea ic Islands ins ead. Summe su eys ca ied ou du ing 2001–2005 and 2012–2014 had enough spa ial esolu ion o esol e he ypical egional mesoscale s uc u es, he e o e allow- ing he iden i ica ion o he loca ion o salini y on s, while he a ea co e ed in 2006 and 2011 was oo small (Supplemen a y Figu e S4). No CTD salini y da a we e a ailable in 2007–2010. Tempe a u e da a om SeaBi d 911þand SeaBi d 25 CTDs we e p ocessed using he Sea-Bi d Elec onics Da a P ocessing ou- ines, and salini y was calib a ed using IAPSO s anda d seawa e and a Guildline 8400A salinome e (see de ails in Balbı´n e al., 2014). Geog aphical fi ness o la al su i al and annual su i al index Simula ions we e conduc ed o analyse whe he he spa io- empo al a iabili y o en i onmen al a iables in luenced he geog aphical i ness o la al su i al. Speci ically, we es ed he in luence o he spa ial and empo al empe a u e a iabili y on he su i al o eggs spawned each day o lexion s age h oughou he spawning season. These simula ions we e conduc ed wi hou explici ly inco po a ing any spa ial spawning s a egy. Eggs we e eleased daily h oughou he wes e n Medi e anean ( om 6W o6 E). Releases we e scaled using a gi en p obabili y o being eleased (P) es ima ed om i ing a polynomial unc ion o he gonadosoma ic index o 528 emales sampled om he comme cial ishe y om he S ai o Gib al a o he Balea ic Sea in he wes e n Medi e anean du ing sp ing- summe 2003–2014 (Regle o e al., 2018a): P¼28:3156 þ0:0024609 d27:4867e06 d34:5819 d0:5 (1) whe e Pis he p obabili y o egg elease and dis he day o he yea . Daily p obabili y o egg elease was no malized so ha elease p obabili ies summed o 1 a he end o each spawning season. We made wo assump ions: (i) he same spawning window applies o all yea s and (ii) he du a ion o an indi idual ish’s spawning pe iod is independen o he spawne ’s body size. Fo each day and posi ion, he eggs ha ched in o la ae ollowing a Figu e 1. Es ima es o spawning-s ock biomass in housands o onnes (whi e do s) and ec ui men a age 1 in millions o fish (black do s) om he 2017 assessmen o he eas e n s ock o A lan ic bluefin una (Anonymous, 2017) and inc easing a e age empe a u es in July o he Balea ic Sea o e he yea s [c osses and fi ( empe a u e ¼0.0288 yea – 36.842; 2 ¼0.16; p<0.01) as dash line]. Habi a and su i al o blue in una 3 Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018 su i al unc ion ela ing ha ching success, H, o empe a u e T (in C) in a empe a u e ange be ween 17 and 34C(Regle o e al., 2018a) o mula ed as: H¼–1:27T2þ63:78T–727:98; 2¼0:92;p<0:001 (2) The eggs ha ched in o la ae wi h an ini ial a e age d y weigh (W 0 ) based on measu emen s o ecen ly ha ched la ae in he labo a o y (W 0 ¼0.018 mg DW, 60.007 s.d., n¼27; Regle o e al., 2018a). G ow h om ha ching o he pos - lexion la al s age was modelled o each daily coho in each spa ial uni as a unc ion o empe a u e. These models we e de eloped using he maximum po en ial speci ic g ow h a es a each empe a u e ob ained using age–weigh ela ionships om labo a o y expe i- men s o A lan ic blue in una la ae (Regle o e al., 2018a). The maximum po en ial speci ic g ow h a es SGR (mg mg 1 d 1 ) in- c eased wi h empe a u e (T) and be e i a linea ela ionship exp essed as: SGR ¼0:0418T–0:8355 ð 2¼0:84;p<0:001Þ(3) The applicable ange o empe a u e on which he ela ionship is based is 22–29C(Regle o e al., 2018a). Mo ali y (M) was modelled as a unc ion o la al size (W) ollowing he exp ession (McGu k, 1986 ci ed in Regle o e al., 2018a): M¼0:00022W0:85 (4) Equa ions (1)–(4) a e used o compu e daily la al g ow h and mo ali y o he ime pe iod om ha ching un il he la ae each he lexion s age. This assigns a su i al p obabili y o each posi- ion and ini ial day. To isualize he p ocess, he accumula ed su i al p obabili y o each posi ion was calcula ed as he ime in eg al o he su i al p obabili ies o each ini ial day. Using di - e en a eas, mean alues om hese maps we e calcula ed o in e annual compa ison. La al and adul spa ial da a Annual geog aphical la al i ness om ou model was compa ed wi h he obse ed spa ial dis ibu ion o la ae and adul ish in he wes e n Medi e anean. A lan ic blue in una la ae we e sampled on c uises conduc ed du ing sp ing and summe du ing 2001–2014 in an a ea ha co e ed 180 220 miles a ound he Balea ic Islands wi h a 10-nau ical mile sepa a ion be ween s a- ions (Alemany e al., 2010). Du ing 2001–2005, la ae we e col- lec ed using Bongo ne s wi h a mou h diame e o 60 cm equipped wi h 333-mm meshes down o 70 m dep h, whe eas om 2006 onwa d, la ae we e collec ed using Bongo ne s wi h a mou h diame e o 90 cm equipped wi h 500-mm meshes down o 30 m dep h (see Alemany e al., 2010;Ing am e al., 2017;Regle o e al., 2018a o mo e de ails). Fish la ae om one eplica e, p e- se ed in 4% bu e ed o malin in seawa e , we e so ed using a s e eoscopic mic oscope, and he o al numbe o A lan ic blue in una la ae was enume a ed. We selec ed only da a om s a ions whe e p esence was posi i e and da a on da e, la i ude, and longi- ude we e a ailable. Fo he loca ion o A lan ic blue in una spawne s, posi ion da a we e ob ained om comme cial ishing ac i i ies (pu se- seine and longline) and elec onic agging in o ma ion. In o ma ion on da e, la i ude, and longi ude o he se s o blue in una schools by he pu se-seine essel “La F au II” was p o ided by G up Bal ego´ o he yea s 2000–2014. On he basis o p e ious his ological s udies (A anda e al., 2013a), all pu se-seine ope a- ions in he a ea we e assumed o a ge schools o ac i ely spawning ish. In con as , longlines a e se o e a b oade ex en- sion in he wes e n Medi e anean and a ge a mo e di e se pop- ula ion o A lan ic blue in una in e ms o ep oduc i e s a e (Medina e al., 2007). Hence, GPS posi ions o longline ca ches we e used only when hey cap u ed spawning emales. Spawning condi ion was con i med by his ological analysis when he o a y con ained pos o ula o y ollicles and/o mig a o y-nucleus oocy es and/o hyd a ed oocy es. A o al o 47 blue in una adul s we e agged unde wa e wi hin he pu se-seine ne s du ing egula comme cial ishing a ound he Balea ic Islands ea ly in he spawning seasons o 2009 (14 June), 2010 (8 June), and 2011 (9 June) (A anda e al., 2013b). Ano he 24 blue in una we e agged, hal o hem unde - wa e and hal onboa d he suppo essels, in he aps loca ed in he S ai o Gib al a a ea ea ly in he 2011 spawning season (Abascal e al., 2016). Fish we e agged wi h pop-up sa elli e ags (MK10 and miniPAT; Wildli e Compu e s, Redmond, WA, USA) a ached by a mono ilamen e he o a da , which was inse ed Figu e 2. S udy a ea. (a) The dis ibu ion o he eas e n s ock o A lan ic bluefin una encompasses he egion o he A lan ic eas wa d om he 45me idian and he Medi e anean Sea. (b) The wes e n Medi e anean is he a ea loca ed mos wes wi hin he Medi e anean Sea. (c) The Balea ic Sea is loca ed in he wes e n Medi e anean Sea. 4P. Regle o e al. Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018 in o he do sal muscula u e a he base o he second do sal in. T ansmi ed in o ma ion was i s p ocessed using he manu ac- u e so wa e (DAP p ocesso ; Wildli e Compu e s). T acks we e es ima ed by Collec e Localisa ion Sa elli e (CLS) using a Kalman il e /smoo he app oach cons ained by ligh le el, sea su ace empe a u e, and bo om opog aphy as desc ibed by Roye and Lu ca age (2008). We only included pa hs o June and July, which included pa o he ish mig a ion o he b eeding g ounds (only in he case o ish agged in he una aps), he spawning pe iod, and he e u n o he o aging a eas in he A lan ic Ocean. Time-se ies o ec ui men Time-se ies o ec ui men we e es ima ed using a ishe y- dependen ec ui men index employed in he la es assessmen o he eas e n s ock by ICCAT conduc ed in 2017 (Anonymous, 2017). We spli his ime-se ies in o wo ime-se ies due o changes in ishing gea . The i s , 2000–2006, used an age- agg ega ed index o 2- and 3-yea -old ish ha es ed by he Spanish bai boa ishe y, whe eas he index a e 2006 used an age-agg ega ed index o 5–6 yea olds. We did no conside es i- ma es o ec ui men da a om VPA ou pu s, gi en he unce - ain y in he VPA es ima es ha esul in un eliable alues o ecen da es (Anonymous, 2017). Resul s Geog aphic pa e ns o spawning and la al su i al may be explained by he ole ha empe a u e plays on he ea ly li e s ages o A lan ic blue in una (Figu e 3;Supplemen a y Figu e S5). Po en ial ha ching p obabili ies we e 40–60% e e y- whe e in he wes e n Medi e anean excep in he no he n and sou he n limi s (Gul o Lions and Albo an Sea, espec i ely) whe e empe a u es we e usually lowe han 20C and, he e o e, ha ching p obabili ies we e e y low (Figu e 3a–c; Supplemen a y Figu e S5). Howe e , po en ial la al su i al a eas we e much mo e es ic ed han po en ial ha ching a eas (Figu e 3d– ; Supplemen a y Figu e S5). The a ea a ound he Balea ic Sea emained op imal o la al su i al despi e annual di e ences in empe a u e, sugges ing i o be he bes a ea in he wes e n Medi e anean o blue in una la al su i al (Figu e 3d– , Supplemen a y Figu e S5). La i udes abo e 41N and below 36.5N always had he lowes la al su i al; in some cases, no la ae su i ed (Figu e 3d– ,Supplemen a y Figu e S5). Wa me summe s enla ged he longi udinal ex ension o po en ial a eas wi h enhanced la al su i al, whe eas colde yea s sh unk po en- ial la al habi a s (Figu e 3d and ). As a consequence o op imal he mal condi ions o he eggs and la ae, 2003 and 2006 esul ed in good yea classes om he model, 2005 and 2009– 2010 in mode a e yea classes, and he es o he yea s in poo yea classes (Figu es 3d– and 4, and Supplemen a y Figu e S5). The geog aphic dis ibu ion o a eas ha p omo e inc eased la al i ness a e consis en wi h he spa ial ep oduc i e s a egy o blue in una, as shown by he spa ial o e lap o la ae and spawning emales as well as o modelled da a (Figu e 3d– , Supplemen a y Figu e S5). The loca ions o spawne s, as obse ed om ishe ies da a, we e spa ially es ic ed o he a ea o he ba- sin a ound he Balea ic Sea as we e po en ial la al habi a s p e- dic ed by he model and he p esence o la ae om ich hyoplank on su eys (Figu e 3d– ,Supplemen a y Figu e S5). T acks o adul s indica ed mo emen s om he S ai o Gib al a owa ds he a eas wi h he highes la al su i al (Figu e 3 , Supplemen a y Figu e S5). Spawning was also obse ed in subop- imal he mal habi a s, as p edic ed by he model, sou h o he islands despi e sui able he mal habi a s nea by (Supplemen a y Figu e S5). We could no es hese habi a s wi h he da a a ail- able om obse a ions since ich hyoplank on su eys we e only a ailable a local scales (10 s o km), and sampling was conduc ed mos ly sou h o he islands and a ely no h (see sampled s a ions as do s in Supplemen a y Figu e S4). No did indus ial and long- line ishe ies o elec onic ags co e he o e all he mal sui able habi a s (Figu e 3,Supplemen a y Figu e S5). Spa ial limi a ions in sampling e o could be impo an in explaining dis ibu ions a he local scale, such as in 2004 and 2007, when maps sugges ha la ae and spawne s we e dis ibu ed in low-quali y he mal habi a s, al hough highe -quali y habi a s we e loca ed nea by, o in 2013, wi h ela i ely high-quali y habi a no h o he islands, bu wi h la ae being dis ibu ed mos ly sou h (Supplemen a y Figu e S5). The p esence o spawning in subop imal he mal habi a s, ye in ela i ely s able loca ions, (in e e y yea , i is obse ed mainly sou h o he islands, and seldom o a ely o he no h), could also be an indica ion ha he adul s use o he cues o de e mine spawning loca ions a a smalle spa ial scale. Besides empe a u e, cha ac e iza ion o habi a equi emen s and p e e ences a he local scale could be in luenced by he o ma ion o salini y g a- dien s due o he con e gence o less-saline ecen A lan ic wa e s and mo e-saline- esiden A lan ic wa e s (Supplemen a y Figu e S4). The no he nmos a ea o he Balea ic a chipelago is usually occupied by esiden A lan ic wa e , whe eas he con lu- ence o new and esiden A lan ic wa e usually occu s a he sou he n bo de o he su eyed a ea, sou h o he islands, as o example in 2002, 2003, 2004, and 2012 o c ossing he Mallo ca Channel owa ds Cab e a Island, as in 2001 and 2005 (Supplemen a y Figu e S4). The po en ial la al su i al index om he model and he ec ui men index om he bai boa ishe y ollowed a simila end (Figu e 5), sugges ing a di ec he mal in luence on in e annual a iabili y in ec ui men a he sho - e m scale in ecen yea s. The s ong signal o he 2003 yea class bo h in he model and in he bai boa ishe y da a is pa icula ly ema kable. Discussion Ou wo k co obo a es p e ious hypo heses ha he Balea ic Sea is an op imal a ea o A lan ic blue in una egg and la al su i al, ega dless o in e annual a iabili y in summe empe a u e. O sp ing always ha e a highe chance o su i ing o he lexion s age he e han in any o he egion/a ea o he analyzed wes e n Medi e anean. We ha e moni o ed he occu ence o A lan ic blue in una la ae and spawning emales in he a ea since 2000 and ha e con i med ha blue in una epea edly spawn in he Balea ic Sea, a spa ial ep oduc i e s a egy ha op imizes he chances o su i al o he o sp ing in e ms o empe a u e. The geog aphic ange o bes la al su i al expands sou hwes wa d in wa me yea s, whe eas i is es ic ed o he Balea ic Sea in colde yea s. This con ac ion–expansion o po en ial la al habi- a is e lec ed in he annual su i al index, highes in he wa m- es yea s and lowes in he coldes yea s. The index complemen s p e ious empe a u e indices (e.g. Ha o d e al., 2017), down- scaling he p ocess o a egional index ha mechanis ically e lec s he in luence o empe a u e on eggs and la ae, bu also changes in he spa ial dis ibu ion o sui able spawning a eas. Habi a and su i al o blue in una 5 Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018 Figu e 3. Spa ial dis ibu ion o he cumula i e p opo ion o eggs success ully ha ching o (a) 2003, (b) 2005, and (c) 2011. Spa ial dis ibu ion o he cumula i e p opo ion o la al su i al o he pos -flexion s age o (d) 2003 (good la al su i al index), (e) 2005 (mode a e la al su i al index), and ( ) 2011 (poo la al su i al index). No e he di e en scales in he figu e o each yea . Posi i e loca ions o he p esence o A lan ic bluefin una la ae ( ed do s), spawning emales cap u ed by pu se-seine (g een do s), and by longline (pink do s) a e shown on op o he la al su i al. Mo emen s o bluefin una adul s du ing June–July a e shown in g ey om agged fish in 2011. 6P. Regle o e al. Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018 The basis o de i ing he habi a sui abili y index in ou s udy was labo a o y-de i ed la al g ow h and weigh -a -age models a di e en empe a u es unde he assump ion ha his ela ion- ship e lec s g ow h a es in na u e. Resul s om o he labo a o y expe imen s on cul u ed la al Paci ic blue in (Thunnus o ien a- lis) and yellow in (T. albaca es) una ha e also shown a clea in- c ease in g ow h a es wi h empe a u e when he e ec o empe a u e was isola ed om o he a iables (Tanaka e al., 2008;Kimu a e al., 2010;Wexle e al., 2011), sugges ing a po- en ial ela ionship be ween g ow h and empe a u e. The e a e e y ew ield-based g ow h es ima es o A lan ic blue in una, and hose a ailable ha e been es ima ed o a na ow he mal ange o age in e al ei he in he Medi e anean (Ga cı´a e al., 2006) o in he Gul o Mexico (Malca e al., 2017), o indica e how simila labo a o y g ow h a es a e o hose om ield- cap u ed la ae a compa able empe a u es. The e a e indica- ions ha g ow h a es may inc ease wi h empe a u e, al hough mo e s udies on g ow h a es a e needed be o e we can use a ield-de i ed ela ionship o es ima e speci ic g ow h a es in ela- ion o empe a u e (see e iew in Muhling e al., 2017). Po en ial di e ences be ween labo a o y es ima es and ield a es could a - ec in e p e a ions o ime–space a iabili y in he mal habi a su i al indices. On he o he hand, he e could be a si ua ion whe e he mal condi ions o habi a a e good o maximal, bu o he condi ions, e.g. poo ood a ailabili y, in which he habi a index would o e es ima e su i al p obabili y. Timing o spawn- ing in A lan ic blue in una has been shown o be ela ed o bo h empe a u e and ood abundance o la ae, while compe i ion and p eda ion among una la ae, ia cannibalism o pisci o y, can a ec la al su i al in a e y oligo ophic a ea (Regle o e al., 2011,2018a). Nowadays, dis ibu ion o ood a he basin spa ial scale can only be app oached om chlo ophyll measu e- men s, bu since he e is no co ela ion be ween spa ial pa e ns in chlo ophyll and ood a ailabili y o una la ae (Regle o e al., 2017), we canno ye cha ac e ize po en ial su i al habi a s e- ga ding ood a ailabili y. Long- e m a e age inc easing empe a u es in he Medi e anean may explain he posi i e end in ec ui men o e long ime-se ies (decadal scale), bu may no wo k in he sho - e m (in e annual scale), pa icula ly in he las decade (Figu e 1). Ou index shows ha he mal condi ions in 2003 and 2006 we e op imal o A lan ic blue in una egg and la ae su i al compa ed wi h o he yea s. The 2003 yea class is ecog- nized as one o he s onges , domina ing ca ches om he Japanese longline and he Spanish bai boa ishe ies (Rod ı´guez- Ma ı´n e al., 2013;Suzuki e al., 2013;Kimo o and I oh, 2017), which i s well wi h op imal en i onmen al condi ions in he spawning g ounds. Yea 2003 was cha ac e ized by anomalous wa m wa e ha has been conside ed as a majo hea wa e in Medi e anean wa e s. E idence in he ishe ies da a o a s ong 2006 yea class is no as clea . Mos o he 2004–2007 yea classes we e no conside ed by he ICCAT Blue in Tuna Wo king G oup o be ealis ic (Anonymous, 2017). In o ma ion on ecen ec ui - men is cu en ly highly unce ain, because no ishe y cap u es eas e n blue in una un il abou age 7 o 8 (Anonymous, 2017), and ends in cap u e pe uni e o (CPUE) o hese age classes can be subjec o o he cumula i e sou ces o a ia ion, o he han ec ui men . Compa ing he model o he Spanish bai boa ishe - ies index, based on an index o 2- and 3-yea -olds in 2000–2004, o he wise 5- and 6-yea -olds, sugges an e ec o empe a u e on ec ui men . In he case o ages 2–3, he i be ween he su i al index and he ishe ies index is e y good, al hough sho , and in he case o ages 5 and 6, he e is a ma ch in spi e o all he sou ces o unce ain y in he CPUE s anda diza ion and all he ac o s ha a e ac ing be ween ec ui men and he ime ish each ha age. The e a e some di icul da a issues wi h he ec ui men da a om i ual popula ion analysis (VPA) models used in assess- men s ha educes he likelihood o a meaning ul link be ween e- c ui men es ima ed om he assessmen and he new indices we ha e de eloped. The e o e, we ha e no made quan i a i e use o VPA da a (B ooks and De oba, 2015). Cu en da a a ailabili y o his s ock limi s ou unde s anding o ec ui men p ocesses and he abili y o esol e which ac o s d i e a ia ions in his species. The e a e s ill some issues ha may limi he applica ion o he su i al index de eloped in his s udy o he en i e Medi e anean (A lan ic blue in una eas e n s ock). The habi a index shown in ou s udy is only o he wes e n Medi e anean, whe eas some a eas wi hin he eas e n and cen al Medi e anean a e also spawning a eas o his species (Regle o e al., 2018a). Con ibu ions o o he spawning a eas o he su i al index, as i Figu e 4. Time-se ies o he po en ial la al su i al index ob ained om he model o 2000–2014. Figu e 5. Va ia ion in ec ui men and he po en ial la al su i al index o he eas e n s ock o A lan ic bluefin una. Annual a ia ion in he ec ui men co esponds o ages 2–3 (black do s) and ages 5–6 (whi e do s) es ima ed om he bai boa index (Anonymous, 2017) (co ec ed o ime lag o 2 and 5 yea s, espec i ely) and su i al index (c osses). No e da a a e only a ailable o 2000–2010 wi hin he ime-se ies conside ed in his s udy. Habi a and su i al o blue in una 7 Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018 is es ima ed now, a e assumed o ha e cons an p opo ions in all yea s, which is an un es ed hypo hesis, since egg p oduc ion ha occu s in he di e en a eas has no ye been documen ed. I he con ibu ion o each egion only depends on empe a u e, hen simila empe a u e egimes ac oss a eas would esul in simila egg p oduc ion o a gi en a ea. Howe e , spa ial con ibu ions o o al s ock egg p oduc ion in he di e en a eas ha e no been es ed, and p opo ions migh change om yea o yea . I has been shown ha salini y and salini y on s a e impo - an in de ining blue in una spawning places in he wes e n Medi e anean (Regle o e al., 2012;Al a ez-Be as egui e al., 2014,2016), and bo h empe a u e and salini y should ideally be used o de eloping ec ui men models, so ha bo h e ec s a e inco po a ed. One p ocedu e could be o use salini y o de ine he loca ion and size o he mo e p obable spawning a ea in each yea and hen, gi en ha egion, de ine i s he mal sui abili y o g ow h and su i al (i.e. a measu e o habi a quali y) using he ela ionships de i ed om he labo a o y. In his way, habi a size, loca ion, and quali y migh be con igu ed mo e ep esen a- i ely han hey a e in exis ing indices and could pe haps explain mo e a iabili y in ec ui men . Un o una ely, hyd og aphic models we e unable o desc ibe salini y a ia ions, and he many CTD da a om ou ex ensi e ieldwo k we e only a ailable o a small a ea wi hin he en i e wes e n Medi e anean. The e o e, we cu en ly do no ha e ools o de elop a salini y-de i ed habi a index. We ha e combined di e en samplings o p o ide he mos ac- cu a e spa ial spawning dis ibu ion in ela ion o empe a u e, al hough some bias occu s due o di e en sampling e o s. In some yea s, he only da a a ailable we e om he ishe ies since no la al su ey could be conduc ed (e.g. 2007). Sampling e o du ing la al su eys is mo e concen a ed sou hwa d han no h o he islands. Pu se-seine s ha e ecen ly been aking hei quo a in jus a ew days o less, so he spa ially de i ed da a om he pu se-seine lee p o ides ela i ely ew da a poin s. Also, in he ishe ies da a, we ha e included only hose cap u es whe e we we e assu ed ha he ish we e ac ually spawning, as in e ed om o a ian his ological analyses. The e is qui e a demand o link egg–la al ecology o ec ui men o his s ock ha equi es con inui y in sampling, as he ones used in ou s udy, bu also new sampling (e.g. seasonal- epea ed su eys, o oli h analyses o he su i o s) co e ing a wide spa ial and empo al scale. P io i ies o his species should be iden i ied in he ICCAT wo k- ing g oups oge he wi h s akeholde s and scien is s. A c i ical ool o imp o e ou unde s anding o geog aphical i ness in e ms o la al g ow h and su i al is he use o ealis ic hyd odynamic models ha simula e ealis ic d i ajec o ies coupled o indi idual-based models ha can include beha iou (Fiksen e al., 2007). Limi a ions o his app oach a ise om he lack o well-calib a ed and alida ed ci cula ion models a e- gional scales, he lack o species-speci ic o mula ions o g ow h and su i al h ough on ogeny, he lack o in e annual spa ial dis ibu ion da a o bo h adul s and o sp ing, and he lack o ime-se ies alida ion o ci cula ion models ha gene ally a e upda ed wi hou main aining con inui y o ime-se ies, which is c i ical o ishe ies s udies. To minimize he me hodological bias and suppo he po en ial ope a ionaliza ion and u u e imple- men a ion o an indi idual-based model, in he p esen s udy, we ha e ensu ed: (i) a ca e ul alida ion o empe a u e da a om he hyd odynamic model in he s udy a ea, (ii) he use o g ow h exp essions ob ained om labo a o y expe imen s speci ically o A lan ic blue in una and applied o obse ed empe a u e eco ds du ing he spawning season, and (iii) a compa ison o model esul s wi h ield obse a ion da a o geog aphic dis ibu- ions o la ae and adul s. While ci cula ion models mimic cu en s well a la ge spa ial and empo al scales, hey may pe o m poo ly a he egional spa- ial and sho empo al scales. Ou alida ion shows ha a ailable models, wi h a su icien ly long ime-span o be o use o ou e- sea ch, a e no capable o accu a ely ep oducing he spa ial dis- ibu ion o di e en wa e masses, cha ac e ized by hei salini y, and mesoscale ac i i y in he a ea. This limi ed ou abili y o (i) ep oduce on al s uc u es and (ii) include d i ajec o ies o es ima e he la al su i al index. Blue in una spawning is ela ed o pa e ns in wa e masses ha explain in e annual di e - ences in he spa ial dis ibu ion o la ae in he a ea (Al a ez- Be as egui e al., 2016). The ci cula ion models es ed did no p o ide accu a e salini y da a and, he e o e, spa ial pa e ns in salini y and on s could no be ep oduced p ope ly. As such, d i ajec o ies could no be es ima ed accu a ely and, he e o e, we e no included in he es ima ion o he po en ial la al su i al index. Howe e , loca ion o he densi y on in he Balea ic Sea, iden i ied om salini y g adien s, would signal he bes a eas o o sp ing su i al. While ci cula ion model ou pu s may ep o- duce well gene al pa e ns o e en ion-dispe sal in la ge a eas o he Medi e anean, hey ail o ep oduce de ailed in e annual pa e ns in he s udy a ea when hey s ongly depend on meso- scale ci cula ion. No el mul ipla o m obse ing sys ems, linking modelling capabili ies and in si u da a, can p o ide alida ed high- esolu ion models o ecen yea s o assess he oceano- g aphic mesoscale scena io in he wes e n Medi e anean (Tin o e´ e al., 2013;Juza´ e al., 2016). Ou s udy ep esen s a ade-o be ween empo al a ailabili y o modelling p oduc s and he capabili y o models o ep oduce di e en en i onmen- al a iables and key ecological p ocesses. One common p ocedu e in he li e a u e is o use o mulae e- la ing empe a u e and g ow h o di e en species han he p i- ma y s udy species unde he assump ion ha he unc ions a e simila . Howe e , o mulae and adap a ions o local en i on- men s ha e been shown o be species-speci ic, and we should be cau ious when using exp essions ha a e no speci ic o he popu- la ion unde s udy. In ou case, we we e able o design a se o expe imen s using A lan ic blue in una and co e he ull ange o empe a u e a iabili y obse ed in he ield (Regle o e al., 2018a). On he o he hand, we know ha su ace empe a u es a e good desc ip o s o egg and la al habi a o his species, which always occupies he i s me e s o he wa e column and show no clea e ical mig a ion pa e ns (Regle o e al., 2018b). We ha e assumed he same spawning window o all he yea s ha may esul in sligh ly highe mo ali ies in yea s when he wa e wa ms la e since eggs will be p oduced, bu no su i e. Despi e he ac ha we de eloped speci ic g ow h unc ions o A lan ic blue in una, we used a size-dependen mo ali y cu e de i ed om McGu k (1986). One ecommenda ion is o es i- ma e species-speci ic mo ali y cu es, which will need o be con- side ed in he u u e. Di ec ly including en i onmen al a iabili y in s ock assess- men models o including ishe ies-independen ec ui men in- dices a e wo ways o imp o e cu en s ock assessmen models o he eas e n spawning s ock o A lan ic blue in una. The index we p oposed may be mo e use ul, no as a measu e o absolu e e- c ui men , bu as a ela i e index in he s ock assessmen . 8P. Regle o e al. Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018 New assessmen wo king g oups o his species may need o dis- cuss he bes me hodology o inco po a ing obse ed en i on- men al e ec s on blue in una g ow h and su i al in cu en assessmen s. Su i al maps combined wi h ha ching a eas would allow he implemen a ion o spa ial egula ions ha would mini- mize byca ch o blue in una spawne s and maximize o sp ing su i al in he wes e n Medi e anean. These could be use ul ools i spa ial decisions a e o be aken o he managemen o he eas e n s ock o A lan ic blue in una o i clima e change e ec s in spa ial dis ibu ions a e o be aken in o accoun a di - e en empo al scales (Tommasi e al., 2017). Supplemen a y da a Supplemen a y ma e ial is a ailable a he ICESJMS online e - sion o he manusc ip . Acknowledgemen s We hank he people in ol ed in he ex ensi e ield sampling, in- ol ed in he expe imen al wo k and da a p ocessing, and in pa - icula Melissa Ma in o he help wi h he la al sampling. We a e g a e ul o Bal ego´ S.L. o p o iding loca ions o ishing ac- i i ies o he yea s o s udy, Calade os del Medi e a´neo SL o p o iding blue in una eggs o he expe imen s. We hank ICCAT o p o iding he assessmen da a. We hank J. Co o use ul discussions on he assessmen da a and blue in una ecol- ogy. This esea ch has ecei ed pa ial unding om he Eu opean Union’s Ho izon 2020 esea ch and inno a ion p o- g amme unde g an ag eemen No. 678193 (CERES) and he Blue in una P ojec (IEO-SOCIB ag eemen ). E.B. and M.H. ac- knowledge he p edoc o al FPI Fellowship and pos doc o al g an suppo , espec i ely, om he egional go e nmen o he Balea ic Islands, Conselle ia d’Educaccio´, Cul u a i Uni e si a s, selec ed as pa o an ope a ional p og am co inanced by he Eu opean Social Fund. All expe imen s we e ca ied ou in acco - dance wi h ele an guidelines on animal expe imen a ion on ish. The me hods used in he cu en s udy we e accep ed by he Minis y o Economy, Indus y and Compe i i eness o Spain and he S ee ing Commi ee o he p ojec CTM2011-29525-C04- 02. We a e e y g a e ul o wo anonymous e e ees and he edi o who p o ided e y use ul commen s o imp o e he pape . Re e ences Abascal, F. J., Medina, A., de La Se na, J. M., Godoy, D., and A anda, G. 2016. T acking blue in una ep oduc i e mig a ion in o he Medi e anean Sea wi h elec onic pop-up sa elli e a chi al ags using wo agging p ocedu es. Fishe ies Oceanog aphy, 25: 54–66. Alemany, F., Quin anilla, L., Velez-Belchi, P., Ga cia, A., Co e´s, D., Rod iguez, J. M., Fe na´ndez de Puelles, M. L., e al. 2010. Cha ac e iza ion o he spawning habi a o A lan ic blue in una and ela ed species in he Balea ic Sea (wes e n Medi e anean). P og ess in Oceanog aphy, 86: 21–38. Al a ez-Be as egui, D., Ciannelli, L., Apa icio-Gonzalez, A., Regle o, P., Hidalgo, M., Lo´ pez-Ju ado, J. L., Tin o e´, J., e al. 2014. Spa ial scale, means and g adien s o hyd og aphic a iables de ine pe- lagic seascapes o blue in and bulle una spawning dis ibu ion. PLoS ONE, 9: e109338. Al a ez-Be as egui, D., Hidalgo, M., Tugo es, M. P., Regle o, P., Apa icio-Gonza´lez, A., Ciannelli, L., Juza, M., e al. 2016. Pelagic seascape ecology o ope a ional ishe ies oceanog aphy: model- ling and p edic ing spawning dis ibu ion o A lan ic blue in una in Wes e n Medi e anean. ICES Jou nal o Ma ine Science, 73: 1851–1862. Anonymous. 2017. Repo o he 2017 ICCAT blue in una da a p e- pa a o y mee ing. Collec i e Volume o Scien i ic Pape s ICCAT, 74: 2268–2371. A anda, G., Abascal, F. J., Va ela, J. L., and Medina, A. 2013b. Spawning beha io and pos -spawning mig a ion pa e ns o A lan ic blue in una (Thunnus hynnus) asce ained om sa elli e a chi al ags. PLoS ONE, 8: e76445. A anda, G., Medina, A., San os, A., Abascal, F. J., and Galaz, T. 2013a. E alua ion o A lan ic blue in una ep oduc i e po en ial om o a ian his ology and i s use in s ock assessmen s. Jou nal o Sea Resea ch, 76: 154–160. Balbı´n, R., Lo´pez-Ju ado, J. L., Flexa, M. M., Regle o, P., Ve´lez- Velchı´, P., Gonza´lez-Pola, C., Rod ı´guez, J. M., e al. 2014. In e annual a iabili y o he ea ly summe ci cula ion a ound he Balea ic Islands: d i ing ac o s and po en ial e ec s on he ma- ine ecosys em. Jou nal o Ma ine Sys ems, 138: 70–81. Be ge , A. M., Goe hel, D. R., Lynch, P. D., Quinn, II T., Mo mede, S., McKenzie, J., and Dunn, A. 2017. Space oddi y: he mission o spa ial in eg a ion. Canadian Jou nal o Fishe ies and Aqua ic Sciences, 74: 1698–1716. B ooks, E. N., and De oba, J. J. 2015. When “da a” a e no da a: he pi alls o pos hoc analyses ha use s ock assessmen model ou - pu . Canadian Jou nal o Fishe ies and Aqua ic Sciences, 72: 634–641. Fiksen, Ø., Jo gensen, C., K is iansen, T., Vikebo, F., and Huse, G. 2007. Linking beha iou al ecology and oceanog aphy: la al be- ha io de e mines g ow h, mo ali y and dispe sal. Ma ine Ecology P og ess Se ies, 347: 195–205. Ga cı´a, A., Co e´s, D., Ramı´ ez, T., Feh i-Bedoui, R., Alemany, F., Rod ı´guez, J. M., Ca pena, A., e al. 2006. Fi s da a on g ow h and nucleic acid and p o ein con en o ield-cap u ed Medi e anean blue in (T. hynnus) and albaco e (T. alalunga) la ae: a compa a i e s udy. Scien ia Ma ina, 70: 67–78. Ha o d, W. J., Ka nauskas, M., Wal e , J. F., and Liu, H. 2017. Non-pa ame ic modeling e eals en i onmen al e ec s on blue- in una ec ui men in A lan ic, Paci ic, and Sou he n Oceans. Fishe ies Oceanog aphy, 26: 396–412. Ing am, J , G. W., Al a ez-Be as egui, D., Regle o, P., Balbı´n, R., Ga cı´a, A., and Alemany, F. 2017. Inco po a ion o habi a in o - ma ion in he de elopmen o indices o la al blue in una (Thunnus hynnus) in he Wes e n Medi e anean Sea (2001–2005 and 2012–2013). Deep Sea Resea ch Pa II: Topical S udies in Oceanog aphy, 140: 203–211. Juza´, M., Mou e, B., Renaul , L., Goma a, S., Sebas ia´n, K., Lo a, S., Bel an, J. P., e al. 2016. SOCIB ope a ional ocean o ecas ing sys- em and mul i-pla o m alida ion in he Wes e n Medi e anean Sea. Jou nal o Ope a ional Oceanog aphy, 9: s155–s166. Kimo o, A., and I oh, T. 2017. Simple upda e o he s anda dized blue in cpue o Japanese longline ishe y in he A lan ic up o 2016 ishing yea . Collec i e Volume o Scien i ic Pape s ICCAT, 73: 1957–1976. Kimu a, S., Ka o, Y., Ki agawa, T., and Yamaoka, N. 2010. Impac s o en i onmen al a iabili y and global wa ming scena io on Paci ic blue in una (Thunnus o ien alis) spawning g ounds and ec ui - men habi a . P og ess in Oceanog aphy, 86: 39–44. Malca, E. M., Muhling, B. A., F anks, J., Ga cı´a, A., Tilley, J., Ge a d, T., Ing am, G. W. J , e al. 2017. Age and g ow h o la al blue in una (Thunnus hynnus): e idence om he Gul o Mexico com- pa ed o he S ai s o Flo ida and Balea ic Sea (Medi e anean). Fishe ies Resea ch, 190: 24–33. McGu k, M. D. 1986. Na u al mo ali y o ma ine pelagic ish eggs and la ae: ole o spa ial pa chiness. Ma ine Ecology P og ess Se ies, 34: 227–242. Medina, A., Abascal, F. J., A ago´n, L., Mou en e, G., A anda, G., Galaz, T., Belmon e, A., e al. 2007. In luence o sampling gea in assessmen o ep oduc i e pa ame e s o blue in una in he wes e n Medi e anean. Ma ine Ecology P og ess Se ies, 337: 221–230. Habi a and su i al o blue in una 9 Downloaded om h ps://academic.oup.com/icesjms/ad ance-a icle-abs ac /doi/10.1093/icesjms/ sy135/5115365 by gues on 07 Oc obe 2018