Changing diets over time: knock-on effects of marine megafauna overexploitation on their competitors in the southwestern Atlantic Ocean

A icle Changing die s o e ime: knock-on e ec s o ma ine mega auna o e exploi a ion on hei compe i o s in he sou hwes e n A lan ic Ocean Ma ia Bas*, Angélica M. Ti oli, I an B iz i Godino, Mónica Salemme, Fe nando San iago, Juan Bau is a Bela di, Flo encia Bo ella, Damián G. Vales, En ique A. C espo, and Luis Ca dona Abs ac .—This s udy compa es he δ 15 N alues and he ophic posi ion o wo seabi d species h oughou he la e Holocene in h ee egions in he sou hwes e n A lan ic Ocean o assess he hypo hesis ha he deci- ma ion o mega auna led o changes in he ophic posi ion o mesop eda o s. Mode n and ancien mol- lusk shells we e also analyzed o accoun o changes in he iso opic baseline h ough ime. Resul s e ealed ha mode n Magellanic penguins ha e highe δ 15 N alues han hei ancien conspecifics in he h ee egions, a e con olling o changes in he iso opic baseline. This was also ue o mode n Impe ial shags compa ed wi h ancien uniden ified co mo an s/shags om he wo a eas whe e ancien specimens we e eco e ed (sou he n Pa agonia and he Beagle Channel). Such empo al a iabili y migh be caused by h ee non–mu uallyexclusi e p ocesses: dec eased a ailabili y o pelagic squa lobs e esul - ing om dec easing p ima y p oduc i i y h ough he la e Holocene, inc eased a ailabili y o small fishes esul ing om he sequen ial deple ion o o he pisci o es (Sou h Ame ican u seal and sea lion and A gen ine hake) since he la e eigh een h cen u y, and modifica ion o he mig a o y pa e ns o Magel- lanic penguins. Al hough disen angling he ela i e con ibu ion o all hose p ocesses is impossible a his ime, he esul s epo ed he e demons a e ha he ecology o Magellanic penguins and Impe ial shags has unde gone majo changes since he la e Holocene. Ma ia Bas†. Depa men o E olu iona y Biology, Ecology and En i onmen al Sciences, Biodi e si y Resea ch Ins i u e (I BIO), Uni e si y o Ba celona, A inguda Diagonal 643, 08028 Ba celona, Spain. †P esen add ess: Renewable Ma ine Resou ces Depa men , Ins i u e o Ma ine Sciences (ICM-CSIC), Passeig Ma í im de la Ba celone a, no. 37-49, 08003 Ba celona, Spain. E-mail: [email p o ec ed] Angélica M. Ti oli and Fe nando San iago. Cen o Aus al de In es igaciones Cien íficas–CONICET, Be na do Houssay 200, 9410 Ushuaia, A gen ina. E-mail: am i [email protected],[email p o ec ed] I an B iz i Godino. Equip de Rece ca A queològica i A queomè ica de la Uni e si a de Ba celona (ERAAUB), Ins i u d’A queologia de la Uni e si a de Ba celona (IAUB), Uni e si y o Ba celona, Mon aleg e 6-8, 08001 Ba celona, Spain; Cen o Aus al de In es igaciones Cien íficas–CONICET, Be na do Houssay 200, 9410 Ushuaia, A gen ina and Depa men o A chaeology, Uni e si y o Yo k, King’s Mano , YO1 7EP Yo k, U.K. E-mail: [email p o ec ed]om Mónica Salemme. Cen o Aus al de In es igaciones Cien íficas–CONICET, Be na do Houssay 200, 9410 Ushuaia, A gen ina and ICSE, Uni e sidad Nacional de Tie a del Fuego, Fuegia Baske 251, 9410 Ushuaia, A gen ina. E-mail: [email protected] Juan Bau is a Bela di. Uni e sidad Nacional de la Pa agonia Aus al, Unidad Académica Río Gallegos (ICASUR)- CIT San a C uz (CONICET), A enida Lisand o de la To e 860, 9400 San a C uz, A gen ina. E-mail: juanbau is[email p o ec ed] Flo encia Bo ella. CONICET-INCUAPA, UNCPBA, A enida Del Valle 5737, B7400JWI Ola a ía, A gen ina. E-mail: flo bo [email p o ec ed] Damián G. Vales and En ique A. C espo. Cen e o he S udy o Ma ine Sys ems (CESIMAR, CONICET), Boule a d B own 2915, U9120ACD Pue o Mad yn, A gen ina. E-mail: ales@cenpa -conice .gob.a , kike@cenpa -conice .gob.a Luis Ca dona. Depa men o E olu iona y Biology, Ecology and En i onmen al Sciences, Biodi e si y Resea ch Ins i u e (I BIO), Uni e si y o Ba celona, A inguda Diagonal 643, 08028 Ba celona, Spain. E-mail: [email p o ec ed] Accep ed: 9 May 2022 *Co esponding au ho . © The Au ho (s), 2022. Published by Camb idge Uni e si y P ess on behal o The Paleon ological Socie y. This is an Open Access a icle, dis ibu ed unde he e ms o he C ea i e Commons A ibu ion licence (h ps://c ea i ecommons.o g/ licenses/by/4.0/), which pe mi s un es ic ed e-use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. 0094-8373/23 Paleobiology, 49(1), 2023, pp. 176–190 DOI: 10.1017/pab.2022.19 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess In oduc ion Dwindling numbe s o la ge ma ine p eda- o s a e one o he mos pe asi e signa u es o ecosys em o e fishing (Jackson e al. 2001), and he emo al o op p eda o s o en sp eads h ough ood webs as ophic cascades (F ank e al. 2005; Mumby e al. 2006) and compe i o elease (Laws 1977; Aal o and Baske 2013; Su ma e al. 2014). The sou hwes e n A lan ic Ocean is no excep ion, and he sequen ial de elopmen o indus ial whaling, sealing, and fishing la gely educed he popula ions o many ma ine mammals and p eda o y fishes be ween he eigh een h and wen ie h cen u ies (Vales e al. 2020 and e e ences he ein). In pa - allel, mos coas al p eda o s in he sou h- wes e n A lan ic Ocean shi ed hei die s ollowing he decima ion o hei own popula- ions, which esul ed in a significan inc ease in hei ophic posi ions and a educ ion in he deg ee o indi idual ophic specializa ion (D ago e al. 2009,2017; Zen eno e al. 2015; Vales e al. 2017; Bas e al. 2019,2020b). In con as o ma ine mammals and la ge fishes, seabi ds ha e no been in ensely exploi ed by humans in he sou hwes e n A lan ic Ocean, al hough hey we e consumed egula ly by hun e -fishe -ga he e people inhabi ing he egion since he middle Holo- cene (Ti oli and Zang ando 2011; Bo ella and C uz 2012; Zang ando and Ti oli 2015), and Eu opean sailo s and se le s hun ed hem o oil and collec ed hei eggs (A ms ong 1994; C uz e al. 2010; G osso 2016 and e e ences he ein). Magellanic penguins (Spheniscus magellani- cus) and Impe ial shags (Leucoca bo a iceps) a e cu en ly he mos abundan coas al seabi ds o Pa agonia and nes all he way om la i ude 42°S o la i ude 54°S (F e e e al. 2005; Schia ini e al. 2005). Magellanic pen- guins ha e d ama ically inc eased bo h hei popula ion size and geog aphic ange du ing he wen ie h cen u y (Boe sma e al. 1990; Bouza e al. 2009; Raya Rey e al. 2014). In o - ma ion abou he Impe ial shags is sca ce and does no e eal any consis en end in he egion, al hough mos colonies inc eased since he 1990s (F e e e al. 2005;RayaRey e al. 2014). Magellanic penguins o aging o Pa agonia eed mainly on small pelagic fish and ju enile A gen ine hake, as well as squid and c us a- ceans (F e e e al. 1996; Scola o e al. 1999; Schia ini e al. 2005; Scioscia e al. 2014). This pa e n is e e sed in Tie a del Fuego, whe e pelagic c us aceans p e ail in he die o Magel- lanic penguins, al hough his has been a ecen change (Dodino e al. 2020). In con as , Impe - ial shags eed p ima ily on ben hic fishes, al hough small pelagic fishes and cephalopods a e also consumed (Gosz onyi and Kuba 1998; Pun a e al. 2003; Ha is e al. 2016). In e es ingly, he expansion o Magellanic penguins in A lan ic Pa agonia has pa alleled he decline o o a iids and hake (Boe sma e al. 1990; Koen-Alonso and Yodzis 2005; Vales e al. 2015), and all h ee a e p eying la gely on small pelagic fishes and squid (Angelescu and P ensky 1987; Koen Alonso e al. 2000; Baylis e al. 2014; Vales e al. 2015). P eda o decima ion has esul ed in he inc ease in biomass o small pelagic fishes and squid (Koen-Alonso and Yodzis 2005; Sán- chez e al. 2012), hus sugges ing ha he expansion o Magellanic penguins migh ha e esul ed om a educ ion o compe i ion and inc eased ood a ailabili y (Boe sma e al. 1990). S able iso ope analysis is a aluable ech- nique o econs uc his o ical changes in he die o p eda o s o e long pe iods, because he s able iso ope a ios in hei issues in eg a e hose o hei p ey (Bea hop e al. 2004), al hough u no e a es a y ac oss issues and in eg a e die a y in o ma ion a di e en empo al scales (Bea hop e al. 2004). Bone has a slow u no e a e, and s able iso ope a ios in he bone o ganic ma ix a e age he iso opic signa u es o p ey du ing se e al yea s (Tieszen e al. 1983; Hobson and Cla k 1992) and hence o e a p oxy o indi idual a iabili y compa able o epea ed measu e- men s o o he issues (Ca dona e al. 2017). This is pa icula ly use ul in bi ds, as die s may a y la gely du ing b eeding and non- b eeding seasons (Hobson and Cla k 1992; Sil a e al. 2014). This s udy aims o es he hypo hesis ha Magellanic penguins and Impe ial shags cu - en ly ha e inc eased hei ophic posi ion CHANGING DIETS OVER TIME 177 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess compa ed wi h hei conspecifics li ing du ing he middle and la e Holocene. To do his, we compa e he s able iso ope a ios o ni ogen in he bone issue o mode n Magellanic pen- guins and Impe ial shags wi h hose o ancien conspecifics eco e ed om a chaeological si es h oughou A lan ic Pa agonia and he Beagle Channel coas s. Fu he mo e, we com- pa e esou ce pa i ioning be ween ancien King penguins (Ap enody es pa agonicus) and Magellanic penguins om he la e Holocene o sou he n Pa agonia wi h ha o hei mode n conspecifics in he nea by Mal inas/Falkland Islands, as hey a e no sympa ic anymo e on he mainland. Ma e ials and Me hods S udy A ea and Sample Collec ion.—Bo h mode n and a chaeological samples we e col- lec ed along he A gen ine coas and g ouped in h ee egions: no he n Pa agonia (Río Neg o and Chubu P o inces), sou he n Pa a- gonia (San a C uz P o ince and he A lan ic coas o Tie a del Fuego egion), and he Bea- gle Channel (Tie a del Fuego P o ince). These h ee egions p esen di e en oceano- g aphic ea u es and dis inc iso opic baselines (Sapo i i e al. 2015). Addi ionally, wi h he pu pose o compa ing s able iso ope a ios o ancien seabi ds ( his s udy) wi h hose in hei mode n coun e pa s (Che el e al. 2002; Weiss e al. 2009) in sou he n Pa agonia, a ou h a ea was defined: he Mal inas/Falk- land Islands (Fig. 1). Bone samples o ancien seabi ds (Magel- lanic penguins, King penguins, and uniden i- fied co mo an s/shags) and shells o ancien limpe s (Nacella magellanica), Chilean mussels (My ilus chilensis), and ibbed mussels (Aulaco- mya a a) we e collec ed om a chaeological si es da ing back o he middle and la e Holocene (Fig. 1 and Supplemen a y Table 1, espec i ely). Mollusk shells we e used o econs uc he iso opic baseline o each egion and pe iod. The s able iso ope a ios o a ew ancien fish species we e ob ained om he li e a u e (Supplemen a y Table 1). In o de o a oid pseudo eplica ion, bones om he neu oc anium (fishes) o long limb bones (bi ds) wi h he same la e ali y we e used o bo h ancien and mode n specimens. The skele al elemen s analyzed o each species a ied ac oss a chaeological si es, because o une en occu ence, bu in a-indi idual a i- abili y in he s able iso ope a ios o skele al elemen s o fish and bi ds is usually much smalle han in e indi idual a iabili y (Bas and Ca dona 2018; Hyland e al. 2021). S able Iso ope Analysis.—Mode n samples we e s o ed in a eeze a −20°C un il analysis. So issues we e emo ed om he seabi d bone and mollusk shells, insed wi h wa e , and allowed o d y a oom empe a u e. Fishes we e hawed a oom empe a u e, boiled be ween 5 and 10 minu es, and dissec ed o emo e he selec ed bones. Shells and bones we e la e d ied in a s o e a 60°C o 24 hou s. Once d y, each sample was g ound o fine powde using a mo a and pes le. Powde ed shell samples we e fi s demine alized by soak- ing in 1 N HCl un il no mo e CO 2 was eleased FIGURE 1. Map o he loca ions unde s udy in A gen ina (Sou h Ame ica) showing he a chaeological si es and sampled species. Dashed squa es show he ou la ge a eas: no he n Pa agonia (A), sou he n Pa agonia (B), Bea- gle Channel (C), and Mal inas/Falkland Islands (D). Black do s deno e he ac onym o each a chaeological si e (see Supplemen a y Table S1) wi h an ex ension showing he species sampled he e (black and whi e animals). G ay do s deno e he mode n sampling loca ions wi h an ex en- sion showing he species sampled he e (g ayscale animals). MARIA BAS ET AL.178 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess (Sapo i i e al. 2014a,b), insed wi h dis illed wa e o 24 hou s, and d ied again o 24 hou s a 50°C, and hen lipids we e emo ed h ough sequen ial inses wi h a 2:1 chlo o- o m:me hanol solu ion un il he solu ion was anspa en (Folch e al. 1957). Samples we e hen d ied again o 24 hou s a 50°C, and 0.5 mg o each sample was weighed in o a in cup. Lipids we e emo ed om d y, powde ed bone samples as desc ibed abo e, bu hey we e demine alized wi h 0.5 N HCl (Newsome e al. 2006; Bas and Ca dona 2018). Then, sam- ples we e d ied again o 24 hou s a 50°C, and 0.3 mg o each sample was weighed in o a in cup. Acidifica ion has no significan e ec on he δ 15 N alue o mollusk shells (Ca michael e al. 2008) o bone collagen (Tu oss e al. 1988). Tin cups we e combus ed a 900°C and analyzed in a con inuous-flow iso ope a io mass spec ome e (Flash 1112 IRMS Del a C Se ies EA, The mo Finnigan). Abundance o s able iso opes is exp essed using he δno a ion, whe e he ela i e a ia- ions o s able iso ope a ios a e exp essed as pe mil (‰) de ia ions om a p edefined e e - ence scale: a mosphe ic ni ogen o δ 15 N. S able iso opic e e ence ma e ials o known 15 N/ 14 N a ios, as gi en by he In e na ional A omic Ene gy Agency (Vienna, Aus ia), we e used o calib a ion. Iso opic e e ence ma e ials we e employed o ecalib a e he sys- em once e e y 12 samples and we e analyzed in o de o compensa e o any measu emen d i o e ime. The aw da a we e ecalcula ed aking in o accoun a linea eg ession p e i- ously calcula ed o iso opic e e ence ma e ials (Sk zypek 2013). S a is ical Analysis.—The s able iso ope a ios o mode n and ancien o ganisms canno be compa ed di ec ly, because he iso opic base- line may a y empo ally (Casey and Pos 2011). None heless, he p o eins ha make up he o ganic ma ix o mollusk shells a e p e- se ed and o e sui able ma e ial o econ- s uc he changes in he iso opic baseline (Casey and Pos 2011; D ago e al. 2017; Misa i e al. 2017). Fi s , he o se be ween he a e age δ 15 N o ancien limpe s and mussels and ha o hei mode n conspecifics was calcula ed when di e ences we e s a is ically significan , and ha amoun was la e sub ac ed om he δ 15 N alues o ancien e eb a e samples o allow compa ison wi h mode n alues (called he “co ec ion ac o ”;Table 1). Second, a e co ec ion o any baseline shi acco ding o molluscan s able iso ope a ios o each a ea and pe iod, ancien and mode n alues o δ 15 N o Magellanic penguins and co - mo an s/shags we e compa ed (see “Resul s”). I should also be no ed ha he published δ 15 N alues o mode n g enadie and eelpou we e ob ained om demine alized and delipided bone (Zang ando e al. 2016) and muscle sam- ples (Riccialdelli e al. 2017), espec i ely. Acco ding o Ankjæ ø e al. (2012), δ 15 N alues om muscle in fishes did no di e om hose o bone collagen, and hence we used δ 15 N alues om muscle o eelpou o compa e i wi h he s able iso ope a ios o bones om ancien conspecifics. In addi ion, published δ 15 N alues o mode n King and Magellanic penguins om he Mal inas/Falkland Islands we e ob ained om ea he s (Weiss e al. 2009) and blood cell samples (Che el e al. 2002), espec i ely. The e o e, he δ 15 N alues o blood cells om Magellanic penguins we e con e ed o hose expec ed o ea he s acco d- ing o he o se be ween hese wo issues o King penguins o allow compa ison (Che el e al. 2005a,b). Thi d, he ophic posi ion o each p eda o (TP p ) was calcula ed as: TPp=[( d 15Np− d 15Nm)/3] +2(1) whe e δ 15 N p is he δ 15 N a e age alues o each p eda o ; δ 15 N m is he δ 15 N a e age alue o mollusks; “3”co esponds o he ophic dis- c imina ion ac o ; and mussels and limpe s we e conside ed he bi o es a TP = 2 (Cau e al. 2009). Then, ancien and mode n alues o he ophic posi ion (TP) o Magellanic pen- guins and co mo an s/shags we e compa ed. All ancien and mode n δ 15 N and TP alues we e compa ed independen ly using gene al linea models (GLM) as un in IBM SPSS S a is- ics ( . 23.0.0.2 o Mac), wi h wo fixed ac o s (species and pe iod) o in e eb a es and one fixed ac o (pe iod) o he seabi d species, unless o he wise s a ed. Then, Tukey’s (HSD) pos hoc es s we e un o assess he empo al CHANGING DIETS OVER TIME 179 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess a ia ion o he δ 15 N and TP alues in each a ea. The Bon e oni co ec ion was used o adjus αle els pe es depending on compa ed pe iods pe a ea, espec i ely. Resul s The δ 15 N alues o all g oups (species × pe - iod) we e no mally dis ibu ed and ulfilled he homoscedas ici y equi emen . The shells o ancien mollusks we e always en iched in 15 N compa ed wi h hose o hei mode n conspeci- fics, and di e ences in he δ 15 N alueso mol- lusks o all pe iods we e s a is ically significan (Table 1, Supplemen a y Fig. 1, Supplemen a y Table 3), hus e ealing he exis ence o a pa allel d op in he δ 15 N baseline in he h ee egions du - ing he pas 2000 yea s. Acco dingly, he s able iso ope a ios om ancien penguins and uniden- ified co mo an s/shags we e ans o med using he co ec ion ac o s o allow compa ison wi h hose o mode n conspecifics om he same egion (Tables 1,2). The δ 15 N alues and he mean ophic posi ion o mode n Magellanic penguins we e signifi- can ly highe han hose o hei ancien conspe- cifics in he h ee egions (Table 2,Fig. 2). In addi ion, s a is ically significan di e ences we e also obse ed be ween ancien samples om di e en pe iods in he Beagle Channel (Table 2,Fig. 2). Likewise, he δ 15 N aluesand he mean ophic posi ion o mode n Impe ial shags we e significan ly highe han hose o ancien uniden ified co mo an s/shags in sou h- e n Pa agonia and he Beagle Channel (Table 2, Fig. 2). Ancien Magellanic and King penguins om sou he n Pa agonia di e ed in hei a e - age δ 15 N and hence o aged a a di e en ophic posi ion (S uden ’s - es ; Table 3), bu he di e - ence was no as la ge as cu en ly exis s in he Mal inas/Falkland Islands. We lack bone samples om ancien fishes om no he n Pa agonia, bu he δ 15 N alues o ancien Magellanic penguins we e only 4‰ highe han hose o con empo a y ibbed mus- sels (Fig. 3A), hus sugges ing a lowe ophic TABLE 1. Resul s o gene al linea model (GLM) wi h wo fixed ac o s (species and pe iod) pe o med o assess he empo al a ia ion o he δ 15 N alues in shells and, when necessa y, compensa e o any iso opic baseline shi be ween he pe iods conside ed. Nis sample size; δ 15 N(‰) is epo edas mean ± SD. Co ec ion ac o (CF) was calcula ed by di e ence be ween mean iso ope alues o mollusks o mode n and ancien samples. *S a is ically significan di e ences ( p< 0.05) be ween ancien and mode n samples. † S able iso ope da a om Bas e al. (2020b). ‡ S able iso ope da a om Bas e al. (2019). A ea Species Loca ion Pe iod (cal y BP) Pe iod Nδ 15 NCF 15 N mollusks No he n Pa agonia Aulacomya a a Playa Las Lisas 2 (p.1) 2620–2081 Ancien 5 13.7 ± 1.4* 3.02 Nacella magellanica Playa Las Lisas 2 (c.2) Ancien 5 15.7 ± 0.5* Aulacomya a a Península Valdés Mode n Mode n 5 11.5 ± 0.3 — Nacella magellanica Península Valdés Mode n 5 11.8 ± 0.5 — Sou he n Pa agonia My ilus chilensis † Río Chico 1 6585–5868 Ancien 5 13.7 ± 0.4* 0.94 Nacella magellanica † Río Chico 1 Ancien 5 12.2 ± 0.8* My ilus chilensis † La A cillosa 2 5864–5776 Ancien 5 14.0 ± 0.2* 2.61 Nacella magellanica † La A cillosa 2 Ancien 5 15.6 ± 0.3* My ilus chilensis Ma gen Su 885–838 Ancien 5 13.9 ± 0.6* 2.20 Nacella magellanica Ma gen Su Ancien 5 14.5 ± 0.4* My ilus chilensis Teis XI 442 Ancien 5 12.0 ± 0.6* 1.33 Nacella magellanica Teis XI Ancien 5 12.8 ± 0.9* My ilus chilensis † Pun a Ma ía Mode n Mode n 5 12.1 ± 0.6 — Nacella magellanica † Pun a Ma ía Mode n 5 11.9 ± 0.3 — My ilus chilensis Ma ía Luisa Mode n Mode n 5 11.6 ± 0.3 — Nacella magellanica Ma ía Luisa Mode n 5 10.5 ± 0.4 — Beagle Channel, Tie a del Fuego My ilus chilensis Imiwaia I (M/K) 6723–6520 Ancien 5 15.4 ± 1.2* 2.84 Nacella magellanica Imiwaia I (M/K) Ancien 5 13.4 ± 0.5* My ilus chilensis ‡ Lanashuaia II 1278–1013 Ancien 5 12.3 ± 0.6* 0.93 Nacella magellanica ‡ Lanashuaia II Ancien 5 12.7 ± 0.6* My ilus chilensis Shamakush X 487 Ancien 5 13.4 ± 0.4* 1.19 Nacella magellanica Shamakush X Ancien 4 12.0 ± 0.6* My ilus chilensis ‡ Bahía Cambace es Mode n Mode n 5 12.1 ± 0.8 — Nacella magellanica ‡ Bahía Cambace es Mode n 5 10.9 ± 0.2 — MARIA BAS ET AL.180 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess TABLE 2. A chaeological and mode n samples used in he cu en s udy. Supe sc ip le e s deno e s a is ically significan di e ences ( p< 0.05) be ween a chaeological and mode n samples; howe e , he Bon e oni co ec ion was used o adjus he αle els pe es depending on he numbe o pe iods compa ed. Nis sample size; δ 15 N, δ 15 N co (‰), and ophic posi ion (TP) a e epo ed as mean ± SD. Co mo an s/shags a e epo ed as Phalac oco ax/Leucoca bo spp., because ancien co mo an s a e uniden ified co mo an / shag species. *S able iso ope da a om Bas e al. (2019). A ea Species Loca ion Pe iod (cal y BP) Nδ 15 Nδ 15 N co TP No he n Pa agonia Magellanic penguin (Spheniscus magellanicus)Fa o San Ma ías (Sondeo 2) 3004 2 18.2 ± 1.2 a 15.2 ± 1.2 2.2 ± 0.2 a Bajada de los Pescado es 2 2157 3 19.2 ± 2.7 a 16.2 ± 2.7 2.2 ± 0.8 a Península Valdés Mode n 5 19.0 ± 1.1 b —4.5 ± 0.3 b Sou he n Pa agonia Magellanic penguin (Spheniscus magellanicus)Río Chico 1 + La A cillosa 2 6585–5776 2 15.1 ± 1.6 a 13.3 ± 0.4 1.8 ± 0.1 a Ma gen Su 885–838 2 15.5 ± 1.1 a 13.3 ± 1.1 1.6 ± 0.2 a Cabo Ví genes Mode n 5 19.0 ± 0.5 b - 4.3 ± 0.2 b Co mo an /shag (Phalac oco ax/Leucoca bo spp.) Río Chico 1 + La A cillosa 2 6585–5776 3 16.1 ± 1.9 a 14.6 ± 1.6 2.2 ± 0.5 a Cabo Ví genes 20 1131 5 17.1 ± 2.6 a 14.9 ± 2.6 2.2 ± 0.9 a Ma gen Su 885–838 5 17.6 ± 0.9 a 15.4 ± 0.9 2.4 ± 0.3 a Teis XI 442 3 17.9 ± 2.3 a,b 16.5 ± 2.3 3.5 ± 0.6 a,b Deseada Island Mode n 5 19.7 ± 0.4 b —4.6 ± 0.1 b Beagle Channel, Tie a del Fuego Magellanic penguin (Spheniscus magellanicus)Imiwaia I (M/K) + Mischiúen (F) 6723–5014 3 17.7 ± 0.6 a 14.9 ± 0.6 2.2 ± 0.2 a Mischiúen (C) + Shamakush I (D) 918–772 5 17.1 ± 0.9 a 16.2 ± 0.9 3.2 ± 0.3 b Shamakush X (E) + Kaiawo eha III (K) 545–487 3 16.6 ± 1.1 a 15.4 ± 1.1 2.9 ± 0.4 b Ma illo Island Mode n 5 18.7 ± 1.2 b —4.4 ± 0.4 b Co mo an /shag (Phalac oco ax/Leucoca bo spp.) Lanashuaia II* 1278–1013 4 9.4 ± 1.8 a 8.5 ± 1.8 0.8 ± 0.5 a Alicia Island Mode n 5 19.7 ± 0.4 b —4.7 ± 0.1 b CHANGING DIETS OVER TIME 181 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess posi ion han ha o hei mode n conspecifics. Cu en ly, he δ 15 N alues o Magellanic pen- guins om no he n Pa agonia a e simila o hose o banded cusk eel, pink cusk eel, and cod icefish; 1‰highe han hose o hake; 3‰ highe han hose o A gen ine ancho y; and 6‰highe han he a e age alue o limpe s and ibbed mussels (Fig. 3B). The δ 15 N alues o ancien uniden ified co - mo an s/shags om sou he n Pa agonia we e highly a iable, bu hey we e always in be ween hose o p eda o y fishes such as hake, eelpou , snoek, and pink cusk eel and only 3‰–4‰highe han he a e age o limpe s and mussels (Fig. 4A–C). Howe e , ancien Magellanic penguins om sou he n Pa agonia we e always ex emely deple ed in 15 N, and hei δ 15 N alues we e so low ha hey did no di e om hose o con empo a y limpe s in Teis XI, Cabo Ví genes, and Ma gen Su (Fig. 4A–C), hus sugges ing ha hey we e no o aging locally mos o ha ime. The same was ue o ancien King penguins (Fig. 4B). Cu en Magellanic penguins and Impe ial shags om sou he n Pa agonia ha e simila δ 15 N alues and highe alues han hose o fishes and mollusks: 1‰abo e hose o eelpou ; 3‰highe han hose o hake, pink cusk eel, and Pa agonian blenny; 6‰ abo e hose o Pa agonian g enadie ; and 7‰ highe han he a e age o limpe s and mussels (Fig. 4D). Finally, he δ 15 N alues o ancien Magel- lanic penguins om he Beagle Channel we e only 3‰–5‰highe han hose o mollusks (Fig. 5A–C), and ancien uniden ified co mo - an s/shags had ex emely low alues o δ 15 N (Fig. 5B). Con e sely, bo h species o mode n TABLE 3. A chaeological and mode n samples om Magellanic and King penguins used in he cu en s udy. Supe sc ip le e s deno e s a is ically significan di e ences ( p< 0.05) be ween con empo a y samples. Nis sample size; δ 15 N, δ 15 N co (‰), and ophic posi ion (TP) a e epo ed as mean ± SD. *S able iso ope a ios o mode n specimens om Weiss e al. (2009). † S able iso ope a ios o mode n specimens om Che el e al. (2002). A ea Species Loca ion Sample Nδ 15 Nδ 15 N co TP Sou he n Pa agonia ancien Magellanic penguin (Spheniscus magellanicus) Ma gen Su (885–838 cal y BP) Bone 2 15.5 ± 1.1 13.3 ± 1.1 a 1.6 ± 0.2 a King penguin (Ap endoy es pa agonicus) Cabo Ví genes 20 (1131 cal y BP) Bone 3 13.5 ± 0.7 11.3 ± 0.7 b 1.0 ± 0.2 b Mal inas/Falkland Islands mode n Magellanic penguin (Spheniscus magellanicus)* New Island (Wes Mal inas) Fea he s 5 15.7 ± 0.5 a —3.2 ± 0.2 a King penguin (Ap endoy es pa agonicus) † Volun ee Beach (Eas Mal inas) Blood cells co 5 11.3 ± 0.3 b —1.8 ± 0.1 b FIGURE 2. δ 15 N alues (mean ± SD) o ancien and mode n Magellanic penguins and co mo an s/shags o each la ge a ea. Black ci cles: mode n samples; whi e ci cles: ancien samples. See Supplemen a y Table 1 o ac onyms. MARIA BAS ET AL.182 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess seabi ds ha e δ 15 N alues simila o hose o eelpou and cod icefish and 8‰highe han he a e age o limpe s and mussels (Fig. 5D). Discussion The p o ocol used he e would allow ob ain- ing unbiased δ 13 C and δ 15 N alues o bone FIGURE 4. δ 15 N alues o in e eb a es, pelagic fishes, ben hic fishes, and seabi ds o (A) Río Chico 1 and La A cillosa 2 (6585–5776 cal y BP); (B) Cabo Ví genes 20 and Ma gen Su (1131–885 cal y BP); (C) Teis XI (442 cal y BP); and (D) mod- e n om he sou he n Pa agonia. A i hme ic mean and s anda d de ia ion (mean ± SD) a e shown o each species. Key: squa es, in e eb a es (INV); diamonds, pelagic fishes (PF); iangles, ben hic fishes (BF); ci cles, ai -b ea hing p eda o s (ABP). See Supplemen a y Table 1 o ac onyms. FIGURE 3. δ 15 N alues o in e eb a es, pelagic fishes, ben hic fishes, and Magellanic penguins o (A) Fa o San Ma ías (Sondeo 2) and Bajada de los escado es 2 (3004–2197 cal y BP) and (B) mode n om no he n Pa agonia. A i hme ic mean and s anda d de ia ion (mean ± SD) a e shown o each species. Key: squa es, in e eb a es (INV); diamonds, pelagic fishes (PF); iangles, ben hic fishes (BF); ci cles, ai -b ea hing p eda o s (ABP). See Supplemen a y Table 1 o ac onyms. CHANGING DIETS OVER TIME 183 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess collagen, as hey a e measu ed using he s and- a d p ocedu e o his ype o sample (News- ome e al. 2006; Gui y e al. 2016; Bas and Ca dona 2018; Gui y and Hun 2020). DeNi o (1985) epo ed ca bon o ni ogen (C:N) a omic a ios o bone collagen o ange om 2.9 o 3.6, and his has been he s anda d equi emen o decades bo h in ecology and a chaeology. Mos o he samples analyzed he e sa isfied his equi emen (Supplemen a y Table 2), bu ecen ly Gui y and Szpak (2020,2021)ha e epo ed a much na owe accep able ange (3.0–3.3). Acco dingly, many samples in his s udy wi h C:N a omic a ios anging om 3.4 o 3.6, may s ill con ain some aces o lipid o humic acid, in mode n and ancien samples, espec i ely, and hence migh yield sligh ly biased δ 13 C alues. Fo his eason, we discuss he e only hei δ 15 N alues, as nei he lipids no humic acid con ain ni ogen, and hence colla- gen is he only sou ce o ni ogen in acidified bone samples (Bas and Ca dona e al. 2018; Bas e al. 2020a; Gui y and Hun 2020; Gui y and Szpak 2020,2021). I should also be no ed ha he o ganic ma ix o mollusk shells is a mix u e o p o eins and chi in, a polysaccha ide con aining ni ogen (Fu uhashi e al. 2009). As a esul , he C:N a io o he o ganic ma ix o mollusk shells including equal amoun s o p o- ein and chi in is close o 5.5 and hence di e s om ha expec ed o collagen. The esul s epo ed he e e eal majo changes in he δ 15 N o Magellanic penguins and co mo an s/shags in he sou hwes e n A lan ic Ocean since he middle Holocene. Ce ainly, sample size o some species and a chaeological si es is small, bu di e ences be ween ancien and mode n conspecifics a e so huge and consis en ac oss a eas, pa icu- la ly o Magellanic penguins, ha we belie e ha ou conclusions a e obus . I should be no ed ha male and emale mode n Magellanic penguins do no di e in hei a e age δ 15 N alues (Scioscia e al. 2014; Sil a e al. 2014; Ba ionue o e al. 2020; Rosciano e al. 2020; Dodino e al. 2021), and he compa ison o δ 15 N in ancien and mode n Magellanic pen- guins is he e o e unlikely o be a ec ed by FIGURE 5. δ 15 N alues o in e eb a es, pelagic fishes, ben hic fishes, and seabi ds o (A) Imiwaia I (M/K) and Mischiúen (F) (6723–5014 cal y BP); (B) Lanashuaia II, Mischiúen (C) and Shamakush I (D) (1278–772 cal y BP); (C) Shamakush X (E) and Kaiawo eha III (K) (487 and 545 cal y BP, espec i ely); and (D) mode n om he Beagle Channel (Tie a del Fuego). A i hme ic mean and s anda d de ia ion (mean ± SD) a e shown o each species. Key: squa es, in e eb a es (INV); dia- monds, pelagic fishes (PF); iangles, ben hic fishes (BF); ci cles, ai -b ea hing p eda o s (ABP). See Supplemen a y Table 1 o ac onyms. MARIA BAS ET AL.184 h ps://doi.o g/10.1017/pab.2022.19 Published online by Camb idge Uni e si y P ess