A chronology of climatic downturns through the mid- and late- Holocene: tracing the distant effects of explosive eruptions from palaeoclimatic and historical evidence in northern Europe

RESEARCH/REVIEW ARTICLE A ch onology o clima ic down u ns h ough he mid- and la e- Holocene: acing he dis an e ec s o explosi e e up ions om palaeoclima ic and his o ical e idence in no he n Eu ope Samuli Helama, 1 Ja i Holopainen, 2 Ma c Macias-Fau ia, 3 Mau i Timonen 1 & Ka i Mielika ¨inen 4 1 Finnish Fo es Resea ch Ins i u e, No he n Uni , PO Box 16, FI-96301 Ro aniemi, Finland 2 Depa men o Geosciences and Geog aphy, Uni e si y o Helsinki, PO Box 64, FI-00014 Helsinki, Finland 3 Ox o d Long- e m Ecology Labo a o y, Biodi e si y Ins i u e, Depa men o Zoology, Uni e si y o Ox o d, Tinbe gen Building, Sou h Pa ks Road, Ox o d OX1 3PS, UK 4 Finnish Fo es Resea ch Ins i u e, Sou he n Uni , PO Box 18, FI-01301 Van aa, Finland Keywo ds Geoch onology; dend oclima ology; his o ical ag icul u e; palaeoclima e olcanism. Co espondence Samuli Helama, Finnish Fo es Resea ch Ins i u e, Ro aniemi Uni , PO Box 16, FI-96301 Ro aniemi, Finland. E-mail: [email p o ec ed] Abs ac Geoch onological da a o he coni e ee ings in a egion sensi i e o clima ic e ec s o explosi e e up ions we e analysed o sudden g ow h educ ions in associa ion wi h ex ao dina ily cool econs uc ed summe empe a u es since 5500 B.C. T ee- ing da a came om he s ems o li ing ees and sub ossil ee emains collec ed as inc emen co es and discs, espec i ely, om an a ea o no he nmos Finnish Lapland (70688N o30208E). Calenda yea da es when he ee- ing signa u es (i.e., g ow h educ ions and econs uc ed empe a u es) we e concu en we e compa ed wi h sulpha e da a om G eenland ice co es. P e ious new e idence a e in ag eemen in demons a ing olcanism behind la e-Holocene e en s in 1601 A.D. and 536 A.D., sugges ing ha he same causal ela ionship can be implied u he back in ime. Ou da a show ha ea lie e en s we e ound o ha e occu ed in he yea s 330 B.C., 874 B.C., 1464 B.C., 1584 B.C., 2564 B.C. and 2850 B.C. In e es ingly, e en s o lesse magni ude ollowed he h ee majo e en s in 542 A.D., 1453 B.C. and 1579 B.C. by a ew yea s. Na u al disas e s, and g ain c op ailu es, occu ed as a esul o hese e en s, as has been documen ed o he summe o 1601 A.D. h ough Finnish his o ical da a and b oadly in he No he n Hemisphe e. Clima e has su p ised humans du ing his o ic and likely p e- his o ic imes, causing sudden al e a ions in ag icul u e, ecology and economy, and may do so in he u u e. We a gue ha he clima e change wi h he mos magni ied impac s on socie y may be a nega i e empe a u e anomaly ha ab up ly dec eases esou ce a ailabili y o e wide spa ial scales. Clima e change may su p ise us wi h unexpec ed phy- sical changes in he Ea h’s a mosphe e (S ee s & Glan z 2000). Su p ising e en s ha cause he mos damage o human heal h and p ope y and impac he na u al en i onmen s a e sudden, discon inuous changes lead- ing o ex eme wea he and clima e e en s (Eas e ling e al. 2000; S ee s & Glan z 2000). Obse a ions and models a e in ag eemen in showing ha s ong clima e anomalies can esul om he a mosphe ic e ec s o la ge explosi e olcanic e up ions (Ches e 1988; Robock 2000; Miles e al. 2004). Indeed, a special cha ac e is ic o olcanism is ha unusually iolen e up ions a e known o ollow unusually long pe iods o quie and ha he e up ion o olcanoes wi h no his o ic ac i i y may lead o he wo s disas e s (Simkin 1993). A mul idisciplina y e alua ion is needed o complemen he his o ical eco d o olcanism. Di ec and seconda y e idence o he e up ions can be aced om deep ice co es and long ee- ing se ies, whe e impu i ies by olcanic ae osols and g ow h (page numbe no o ci a ion pu pose) Pola Resea ch 2013. #2013 S. Helama 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-Noncomme cial 3.0 Unpo ed License (h p://c ea i ecommons.o g/licenses/by-nc/3.0/), pe mi ing all non-comme cial 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. 1 Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 anomalies, espec i ely, a e known o ollow such e up- ions (Zielinski 2000). A ecen s udy elucida ed he clima ic esponse in Eu ope ollowing 15 majo opical e up ions o e he las hal millennium (Fische e al. 2007) and con i med he clea pa e n o summe empe a u e cooling du ing he i s and second pos -e up ion yea s (B adley 1988; Robock 2000). O hese, he s onges signal o cooling is ound du ing he yea a e he e up ion. No ably, Finland lies in he a ea o he mos p onounced summe cooling (see igu e 1 in Fische e al. 2007). In he same egion, ee ings and hei summe empe a u e e- cons uc ions ha e been sugges ed o exhibi olcanic signa u es as dis an e ec s o explosi e e up ions (Helama, Lindholm e al. 2005; Salze & Hughes 2007; Helama, La ¨a ¨nelaid e al. 2010). Simila indica ions ha e been de ec ed in adjacen a eas o no he n Eu ope (Ge ais & MacDonald 2001; Salze & Hughes 2007; Shumilo e al. 2011). By i ue o i s geog aphical loca ion and he sensi i i y o ee ings as a p oxy o clima e, ee- ing ma e ial om he egion o no he n Finland known as Lapland would appea p i ileged o acing he clima ic e ec s o pas olcanism. The ongoing ask o explo ing sub ossil wood in Finnish Lapland o dend oclima ic pu poses has so a esul ed in a ch onology spanning 7.6 housand yea s (E onen e al. 1999; E onen e al. 2002; Helama e al. 2008). The ee- ing ch onology o Finnish Lapland has been used o econs uc he egion’s pas summe empe a u e a iabili y since 5500 B.C. on annual o millennial scales (Helama, Macias Fau ia e al. 2010). This ime ame p o ides an in e glacial window excluding he ea ly Holocene, du ing which pe u bed olcanism ol- lowed he deglacia ion isos a ic adjus men o magma chambe s o he associa ed changes in c us al s esses, as indica ed by chemical da a yielded by deep ice co es (Zielinski e al. 1994; Zielinski e al. 1997). Compa ed o he G eenland ice-co e chemis y (see Table 1), ee- ing da ing allows o g ea e p ecision in de e mining when e en s occu ed (Baillie 2008, 2010). The signa u e yea s o he new palaeoclima ic econ- s uc ion (Helama, Macias Fau ia e al. 2010) ha e no so a been sys ema ically examined. To analyse he da a, we employ obus me hodology based on mul iple lines o e idence. Fi s , ou analyses a e cons ained by he knowledge o he e ec s in he egion o explosi e e up ions ha ake place a away (Fische e al. 2007). Ensu ing he obus ness o ou dend och onological analyses, wo di e en ee- ing me hods we e used o examine he ini ial dend oclima ic da a p oducing he ch onologies. In addi ion, we combined a p ede e mined de ini ion o he olcanic signa u e in he dend och o- nological da a (Ge ais & MacDonald 2001) wi h a compa ison o he ch onologically cons ained e idence o he G eenland ice-co e chemis y (Table 1). Fo he la e-Holocene, i was possible o supe pose he palaeo- clima e esul s upon he his o ical da a and e iew he consequences in con ex o human ecology (Holopainen & Helama 2009). These analyses pa ed he way o ou new eco d o clima ic down u ns, wi h linkages o he dis an e ec s o olcanism on no he n Eu ope h ough he p esen in e glacial. Ma e ial and me hods Ini ial ee- ing ch onology Da a comp ising he se ies o ee ings we e de i ed om he s ems o li ing ees and unks o sub ossil ee emains collec ed as inc emen co es and discs, espec- i ely, om an a ea o no he nmos Finnish Lapland loca ed a 70688N o30208E (E onen e al. 1999; E onen e al. 2002; Helama e al. 2008). The wid hs o he ings we e measu ed unde he mic oscope and he esul ing ee- ing se ies c oss-da ed using es ablished dend och onological echniques (Aniol 1983; Holmes 1983; Van Deusen 1990). In his ou ine me hod, he se ies a e examined o e o s and he in e -co ela i e sample se ies a e p og essi ely added in o he mean ch onology (F i s 1976). In addi ion o c oss-da ing wi hin he ee- ing da a o Finnish Lapland, he ee- ing ch onologies o o he pa s o Finland (Helama, Lindholm e al. 2005) and he adjacen a ea in no he n Sweden (G udd e al. 2002) ha e been success ully cons uc ed (E onen e al. 2002). The Finnish Lapland ch onology comp ises 69 li ing and 1249 sub ossil ee- ing se ies o Sco s pine (Pinus syl es is L.) and he combined ch onology co e s he in e al om 5633 B.C. un il he p esen -day (Helama, Macias Fau ia e al. 2010). T ee- ing sensi i i y The non-clima ic ends in he ee- ing g ow h we e emo ed p io o he a ge ed analyses using conse a i e g ow h end modelling. A modi ied nega i e exponen- ial cu e (F i s e al. 1969) o linea eg ession wi h nega i e o ze o slope was i ed o each indi idual ing wid h se ies. The ee- ing indices we e compu ed as a ios be ween he measu ed and modelled wid hs. This me hod o s anda diza ion (F i s 1976) is known o be well sui ed o Sco s pine ee- ing wid h da a (Lindholm A ch onology o mid- and la e-Holocene clima ic down u ns S. Helama e al. 2 (page numbe no o ci a ion pu pose) Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 1996; Helama, Lindholm e al. 2004). The mean ch on- ology was calcula ed using a biweigh obus es ima ion (Cook e al. 1990). The a iance o he ch onology was s abilized using he me hod o Osbo n e al. (1997), using a ime-independen es ima e o mean in e se ies co ela ion ( 0.344; Helama, Lindholm e al. 2004). Following he p e iously applied p o ocol (Ge ais & MacDonald 2001), he ch onology was Fig. 1 A mid- and la e-Holocene ch onology o clima ic down u ns. (a) T ee- ing sensi i i y (i.e., sudden change in g ow h condi ions). Please no e ha only nega i e depa u es a e gi en, he alues he e o e indica ing g ow h educ ions. (b) Recons uc ed summe (July) empe a u e a iabili y (black line) wi h he g een and blue a eas indica ing he 95% and 99% con idence in e als o he econs uc ion. The s udy pe iod was 5500 B.C. h ough 2005 A.D. The yea s discussed in he ex a e shown as ee- ing da ed calenda yea s B.C. and A.D. S. Helama e al. A ch onology o mid- and la e-Holocene clima ic down u ns Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 3 (page numbe no o ci a ion pu pose) cha ac e ized o he ampli ude o he in e -annual g ow h a ia ions, calcula ed using he sensi i i y (F i s 1976). The sensi i i y was de e mined as (x x 1 )/ (x x 1 ), whe e x is he mean index o yea (Ge ais & MacDonald 2001). The ee- ing signa u e o la ge- scale clima ic e ec s om explosi e e up ions can lag he ac ual olcanic e en (Jones e al. 1995; B i a e al. 1998). Tha is, he excessi ely nega i e alues in yea could be expec ed om he explosi e e up ions om he same yea o he p e ious yea -n, ha is, du ing he i s and second pos -e up ion yea s (B adley 1988; Fische e al. 2007). Indica ions o clima ic cooling A palaeoclima ic econs uc ion (Helama, Macias Fau ia e al. 2010) was p oduced using a eg ession o he egional summe empe a u e on he ee- ing wid h ch onology. Fi s , an empi ically designed p ocedu e esembling egional cu e s anda diza ion (B i a e al. 1996) was applied, wi h he excep ion ha he conca i y o he g ow h end was adjus ed o pas changes in pine popula ion densi y in he s udy egion (Helama, Timonen e al. 2005; Helama, Macias Fau ia e al. 2010). Tha is, he non-clima ic a ia ions we e emo ed om he indi i- dual ee- ing se ies wi h a modi ied nega i e exponen ial cu e (F i s e al. 1969) uned o accoun o palaeoeco- logical in o ma ion abou he non-clima ic changes in he end shape (Helama, Macias Fau ia e al. 2010). The ee- ing indices we e de i ed as a ios be ween he obse ed and e e ence alues, and he ch onology was a e aged using a biweigh obus mean o p oduce annual index alues. The a iance o he ch onology was s abilized using he me hod o Osbo n e al. (1997), using a ime- independen es ima e o mean in e se ies co ela ion (  0.330) (Helama, Lindholm e al. 2004). Second, he ee- ing ch onology was eg essed agains ins umen al clima e da a om a wea he s a ion in Ka asjok, no he n No way (69828?N; 25831?E) o depic empe a u e a ia ions on a Celsius scale (Helama, Macias Fau ia e al. 2010). This me hod was used he e as i was p e iously e alua ed as a obus palaeoclima ic model o simila da a (Helama e al. 2009). The econs uc ion accoun s o mo e han 40% o he o al obse ed a iabili y (R 2 0.43; R 2 ADJ 0.41) o e he ull calib a- ion pe iod (18772004 A.D.). Mo eo e , ou p e ious analysis showed ha no only ee ings o li ing pines bu also he ee- ing ch onology o pu ely sub ossil logs co ela e posi i ely wi h summe empe a u es (Helama, Holopainen e al. 2004). Applying he ob ained ans e unc ion o he ee- ing da a o e he p e-calib a ion pe iod, he ch onology was ans o med in o yea ly es ima es o summe empe a u es o 5500 B.C. h ough 2005 A.D. Con idence in e als o he inal econs uc- ion we e compu ed om he au o eg essi e s uc u e o he esiduals o he linea associa ion be ween he econs uc ion and ins umen al da a and he yea s 1877 A.D. and 2004 A.D., based on 1000 Mon e Ca lo simula- ions (Macias Fau ia e al. 2010; Macias-Fau ia e al. 2012). The ex emely cool summe s ollowing explosi e olcanic e en s wi h la ge-scale clima ic e ec s we e iden i ied om he econs uc ion as he mos nega i e empe a u e depa u es om he long- e m mean o he econs uc ion. Resul s A look a he plo o he ee- ing a iabili y e eals a cha ac e is ic pauci y o anomalously poo g ow h in e ms o dend och onological sensi i i y (Fig. 1a). The mos nega i e yea s o g ow h a e no clus e ed wi hin a limi ed pe iod bu a e sp ead o e se e al millennia. Conside ing he la e-Holocene (he e, 12005 A.D.), he ee- ing eco d shows ex eme d ops in g ow h as ha ing occu ed in 1601 A.D. and 536 A.D. These yea s we e econs uc ed as ha ing been excep ionally cool (Fig. 1b). Fo bo h o hese yea s, he summe empe a- u es we e econs uc ed o ha e been coole han 108C on a e age, which is mo e han h ee s anda d de ia ions om he econs uc ed mean o 138C. The yea 536 A.D. was ollowed by ano he yea o educed g ow h, in 542 A.D., du ing which he empe a u es a e econs uc ed o ha e been nea ly as cool as six yea s be o e (Fig. 2). T ee- ing da a o he B.C. e a indica e ha he e we e wo yea s o educed g ow h*in 330 B.C. and 2850 B.C.* ha we e compa able o he wo la e-Holocene e en s desc ibed abo e (Fig. 1a). The yea s 874 B.C., 1584 B.C. and 2564 B.C. we e also econs uc ed o be ex emely cool yea s. Howe e , he yea s 874 B.C. and 1584 B.C. did no show compa able sudden g ow h supp ession, implying ha he econs uc ed coolness du ing hese e en s was caused by a p og essi e cooling Table 1 The G eenland ice co es e e ed o in his s udy. Ice co e Re e ences Dye 3 (pa o he G eenland Ice Shee P ojec ) Clausen e al. (1997), Vin he e al. (2006), La sen e al. (2008) G eenland Ice Co e P ojec (GRIP) Clausen e al. (1997), Sou hon (2002), Vin he e al. (2006), La sen e al. (2008) No h G eenland Ice Co e P ojec (NGRIP) Rasmussen e al. (2006), Vin he e al. (2006), La sen e al. (2008) G eenland Ice Shee P ojec 2 (GISP2) Alley e al. (1993, 1997), Meese e al. (1997), Zielinski e al. (1994), Sou hon (2002) A ch onology o mid- and la e-Holocene clima ic down u ns S. Helama e al. 4 (page numbe no o ci a ion pu pose) Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 o e wo o mo e consecu i e yea s. A quiescence o he ea lies i s wo-and-a-hal millennia o he eco d was con as ed wi h an excep ion o g ow h educ ion in 4866 B.C. (Fig. 1a). O no e, he e en s o 1464 B.C. and 1584 B.C. we e ol- lowed by cool yea s 11 and i e yea s la e , espec i ely* in 1453 B.C. and 1579 B.C. These anomalies b ing o mind he e en s in 536 A.D. and 542 A.D. (Fig. 2). Discussion La e-Holocene e en s The signa u e yea s o he la e-Holocene occu ed in 1601 A.D. and 536 A.D. A close look a hese e en s is essen ial as hey allow o he na u al p oxy e idence o be com- pa ed o his o ical documen a ion. His o ical desc ip ions om Sweden, No way and Iceland indica e ha he sun was obscu ed in 1601 A.D. (Kalela-B undin 1997). The G eenland Ice Co e P ojec (GRIP; Clausen e al. 1997) and G eenland Ice Shee P ojec 2 (GISP2; Zielinski e al. 1994) ice-co e da a sugges olcanic signa u es in 1601 A.D. and 1604 A.D., espec i ely (Table 2). La e s udies ha e assigned he coolness o he yea 1601 A.D. o he e up ion o Huaynapu ina, in Pe u, ha had occu ed he p e ious yea (B i a e al. 1998; de Sil a & Zielinski 1998). The geog aphical loca ion o he s udy egion poses challenges o ag icul u e because o he no he n clima e (Pa y 1975; Pa y & Ca e 1985), and one would expec o see clima ic e ec s on pas ag icul u al p oduc ion in he egion, as p oposed by Solan ie (1997). Documen a y e idence o ag icul u al success (To nbe g 1989) e eals ma ked a iabili y o ye and ba ley c ops o e he in e al o ime in which we a e in e es ed, and seems o indica e yea s o c op ailu e in sou he n Finland (ca. 608N; 228E) in 1554 A.D., 1577 A.D. and 1601 A.D. (Fig. 3). His o ical ha es s o ye and ba ley ha e been analysed on he basis o con inuous eco ds be ween 1551 A.D. and 1609 A.D. (Holopainen & Helama 2009). Recalcula ing hei da a, we ind ha he a e age g ain- igu e ( a io be ween sown and ha es ed g ain) o he pe iod o 15921600 A.D. was 5.5, whe eas o he yea o 1601 A.D., i was as low as 2.8 (Fig. 3). A much b oade spa ial scales, A well (2001) and Ve osub & Lippman (2008) e iewed e idence om documen a y and na u al a chi es ha show agg ega ions o anomalously cool wea he and amine in se e al coun ies and con inen s. Fig. 2 Recu ing e en s in g ow h and clima e condi ions. The due s occu ed in 536/542 A.D., 1464/1453 B.C. and 1584/1579 B.C., shown he e as sensi i i ies and econs uc ed empe a u es. Please no e ha only nega i e depa u es a e gi en o sensi i i y. The econs uc ed summe (July) empe a u e a iabili y (black line) is shown wi h he g een and blue a eas indica ing he 95% and 99% con idence in e als o he econs uc ion. Table 2 Likely co espondences be ween he ee- ing da es p esen ed in his s udy and p e iously published e idence om he G eenland ice co es. Yea Ice-co e e idence 1601 A.D. 1604 (Zielinski e al. 1994); 1601 (Clausen e al. 1997) 536 A.D. 53354392 (La sen e al. 2008) 330 B.C. 365 (Zielinski e al. 1994) 874 B.C. 864 (Zielinski e al. 1994), 888 (Clausen e al. 1997) 1464 B.C. 1442, 1454, 1459 (Zielinski e al. 1994); 1457, 1463 (Clausen e al. 1997) 1584 B.C. 1577, 1594, 1600 (Zielinski e al. 1994) 2564 B.C. No acidi y peaks a ound his dend och onological da e 2850 B.C. 2815 (Zielinski e al. 1994) 4866 B.C. 4893 (Zielinski e al. 1994) S. Helama e al. A ch onology o mid- and la e-Holocene clima ic down u ns Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 5 (page numbe no o ci a ion pu pose) The 536 A.D. e en has long in igued scien is s o a ious disciplines (S o he s 1984, 1999; B i a e al. 1990; Ze e be g e al. 1994; A ja a 2005; Baillie 2008; La sen e al. 2008; Woods 2010). The e en wi h he associa ed e idence in e ed om ee ings has been p e iously associa ed wi h he Rabaul olcano on he island o New B i ain, Papua New Guinea (S o he s & Rampino 1983a, 1983b; S o he s 1984). New e idence ob ained by adioca bon da ing o cha coal collec ed om he basal uni s o he local e up ion deposi s has sugges ed, howe e , ha he Rabaul e up ion occu ed du ing he 7 h cen u y A.D. (McKee e al. 2011). The lack o olcanic deposi s in ice co es (Clausen e al. 1997) has also led o he sugges ion ha a come exploding in he uppe a mosphe e could esul in he deposi- ion o deb is in he uppe a mosphe e, which would educe he pene a ion o sunligh and cause a clima ic down u n (Rigby e al. 2004). Howe e , he compa ison o sulpha e om mul iple ice co es*Dye 3 (pa o he G eenland Ice Shee P ojec ), GRIP and No h G eenland Ice Co e P ojec (NGRIP)*un eiled a signal consis en wi h he olcanic hypo hesis o he e en (La sen e al. 2008). The empo al pa e n o sulpha e a ia ions in he ice co es u he ancho ed he ice-co e and ee- ing indica ions o olcanic dus and clima ic down- u ns in he yea s 522 A.D., 532 A.D., 536 A.D., 542 A.D. and 574 A.D. (Baillie 2008). The da a shown he e suppo ou unde s anding o 536 A.D. and 542 A.D. as ha ing been among he cooles summe s in he Holocene (Fig. 2). Compa isons o e he B.C. e a E o s o link ee- ing and ice-co e da a become mo e complex as one looks u he back in ime. Se e al s udies examining indi idual and mul iple ice co es ha e indi- ca ed inc easing e o s in laye coun ing wi h olde ice ma e ials (Clausen e al. 1997; Meese e al. 1997; Sou hon 2002; Rasmussen e al. 2006; Vin he e al. 2006). The es ima ions ha e shown a maximum da ing e o o 2% in he Holocene segmen s o he G eenland ice ch onologies (Meese e al. 1997; Rasmussen e al. 2006). An addi ional 80-yea o se was ound close o 53005400 yea s ago when compa ing he GRIP and GISP2 ch onolo- gies (Sou hon 2002). Mo eo e , ou ee- ing esul s du ing he B.C. e a did no always compa e well wi h o he ee- ing es ima es (Salze & Hughes 2007). Wi h espec o di e en ee species and geog aphical loca ions, he de ia ions can o igina e om co espond- ingly dissimila ecological ci cums ances. Di e ences can also esul om disc epancies in me hods (Baillie & Mun o 1988; Salze & Hughes 2007) o de i e ee- ing in o ma ion. An indica ion o he 330 B.C. e en in ou ee- ing eco ds could be linked wi h a GISP2 sulpha e peak in 365 B.C. (Zielinski e al. 1994) (Table 2). Though he appa en di e ence o 35 yea s is la ge, e o s in he ice ch onology (Meese e al. 1997) may accommoda e i . B oadly coinciding wi h hese da es, he e up ion o A acazo occu ed in he Wes e n Co dille a o Ecuado (Siebe & Simkin 2002). Desc ibed om a geological pe spec i e by Hidalgo e al. (2008), he e up ions a e es ima ed o ha e had a la ge explosi e magni ude. The calib a ed adioca bon da es a e age o 390260 (1-sigma con idence le el [CI]) and 40023 calib a ed yea s B.C. (2-sigma CI) (Hidalgo e al. 2008). In addi ion, ou indica ion o he yea 874 B.C. may ma ch he sulpha e signal o 864 B.C. shown in he GRIP2 ice co e (Zielinski e al. 1994) and ha o 888 B.C. in Fig. 3 Ch onologies o ha es s and clima e a iabili y. (a) Rye and ba ley g ain- igu es in sou h-wes e n Finland o he pe iod 1549 A.D. h ough 1619 A.D. shown wi h (b) he econs uc ed summe empe a- u e a iabili y o his s udy (black line) wi h he g een and blue a eas indica ing he 95% and 99% con idence in e als o he econs uc ion. The g ain- igu e da a we e ecalcula ed om he o iginal s udy (Holopainen & Helama 2009) s abilizing he a iance o he g ain- igu es using a me hod iden ical o ee- ing analyses (Osbo n e al. 1997). A ch onology o mid- and la e-Holocene clima ic down u ns S. Helama e al. 6 (page numbe no o ci a ion pu pose) Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 he GRIP co e (Clausen e al. 1997). The la e could poin o a majo olcanic e up ion in he No he n Hemisphe e as he deposi ed amoun o olcanic acid was es ima ed o be h ee imes mo e han he mean annual backg ound acid deposi ion a e (Clausen e al. 1997). The ad an ages o ancho ing he ee- ing and ice-co e e idence o olcanism by mul iple ecu ing e en s has p e iously been demons a ed (Baillie 2008, 2010). In his ega d, he e is ee- ing e idence o a wo-s age e en in ol ing he yea s 1464 B.C. (s onge ) and 1453 B.C., wi h an 11-yea in e al. A simila pai o e en s was obse ed in ee ings in 1584 B.C. (s onge ) and 1579 B.C. The 1584/1579 B.C. e en s would show an in e es ing co espondence wi h he GISP2 sulpha e anomalies da ed o 1600 B.C. and 1594 B.C., wi h a di e ence o six yea s (Zielinski e al. 1994). Ye a single sulpha e anomaly nea es o he ee- ing da es would sugges an ice-co e da e o 1577 B.C. (Zielinski e al. 1994). Rega ding he 1464/1453 B.C. e en s, he ee- ing da a co esponded wi h he GISP2 sulpha e peaks da ed o 1454 B.C. and 1442 B.C., wi h a 12-yea in e al (Zielinski e al. 1994). Mo eo e , he Dye 3 ice co e indica ed chemical signa u es o he yea s o 1463 B.C. and 1457 B.C. (Clausen e al. 1997). The molumines- cence da ing o olcanic samples om he island o Yiali in G eece (dis inguishable om eph as om he mo e ex ensi ely s udied Medi e anean e up ion o he San- o ini [The a] olcano) de i ed a mean age o 1460 B.C. (Li i zis e al. 1996). Dend och onological da es o 1628 B.C. om Eu opean oak ee- ing wid hs (Baillie & Mun o 1988), 1627 B.C. as a os -damaged ing (LaMa che & Hi schboeck 1984), a g ow h minimum in 1626 B.C. (Salze & Hughes 2007) om he No h Ame ican b is lecone pine (Pinus longae a and Pinus a is a a) da a, and he sulpha e e en in 1623 B.C. om GISP2 ice-co e da a (Zielinski e al. 1994) ha e been sugges ed as aces o he San o ini e up ion. Ou da a did no indica e anomalous condi ions du ing he pe iod o 16271623 B.C. The e o e, he pai ed e en s in 1584 and 1579 B.C. ound in his s udy ep esen ye a new line o e idence o he highly uns able clima e in he mid-second millennium B.C. The new e idence o he e en s o po en ially olcanic o igin in 2564 B.C. did no ha e a po en ial ma ch in he ice-co e esul s (Table 2). The ee- ing da e o 2850 B.C. appea ed wi hin he unce ain y bounda ies o he ice-co e e en in 2815 B.C. (Zielinski e al. 1994). The ea lies ee- ing indica ion o olcanism in 4866 B.C. could ma ch he GISP2 e idence o high sulpha e alues in 4893 B.C. (Zielinski e al. 1994). Geoch onology wi h ecological implica ions T ee- ing da a sensi i e o summe empe a u e luc ua- ions we e examined o anomalously nega i e p oxy depa u es in a egion whe e dis an e ec s o explosi e e up ions ha e been shown o p oduce a pa e n o cooling o he i s and second yea s a e he olcanic e en s (Fische e al. 2007). This was shown o he yea 1601 A.D., wi h s ong impac s on he ag icul u al p oduc ion as econs uc ed om his o ical documen s om sou he n Finland (Fig. 3). The in e annual a ia- bili y in c op yields in his egion is among he lowes in Finland (Mukula & Ran anen 1989a, b), which sugges s ha la ge yea - o-yea a ia ions may ha e been expe ienced elsewhe e in Finland. The associa ion is plausible as bo h he ha es s o mode n (Mukula & Ran anen 1987, 1989a,b) and his o ical (Holopainen & Helama 2009) ag icul u e a e known o depend on he g owing season clima e. Inc eased human mo ali y in he egion in asso- cia ion wi h he c op ailu es has been documen ed o he i s yea s o he 17 h cen u y A.D. (Ju ikkala 1958, 2003a, 2003b). We acknowledge ha linking clima e phenomena wi h pas e up ions in ol es sup- posi ions (Sadle & G a an 1999), bu we ne e heless sugges his connec ion o he e en in 1601 A.D. in no he n Finland. The clima ec op associa ion may ha e played a ole in he ag icul u e-dependen economic changes ha occu ed a ound ha ime (Holopainen e al. 2012) and a ec ed he ep oduc i e success among he poo amilies li ing in he egion (Ricka d e al. 2010). The empo al associa ion o ee- ing anomalies wi h la e-Holocene e up ions is sugges i e o a ela ionship ha can be ex ended back in ime. In his ega d, pollen da a indica e spo adic cul i a ion ac i i y in pa s o Finland since he B onze Age and hus du ing a leas he pas h ee millennia (Taa i sainen e al. 1998). The summe empe a u e d i en en i onmen al p oduc i - i y has also been shown o in luence hun e ga he e popula ion size luc ua ions in he egion (Talla aa a e al. 2010; Talla aa a & Seppa ¨2011). A clea pic u e o his ype was de i ed ia p oxy compa ison be ween he summed p obabili y dis ibu ion o a chaeological adioca bon da es (da es-as-da a) and a e hicknesses o o ganic sedimen laye s om a lake wi hin he s udy egion (Talla aa a & Seppa ¨2011). Conclusions This pape has p esen ed new e idence o ab up educ ions in clima ic cooling du ing he mid- and la e- Holocene. The backbone o his s udy was he 7.6 S. Helama e al. A ch onology o mid- and la e-Holocene clima ic down u ns Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 7 (page numbe no o ci a ion pu pose) housand yea -long ee- ing ch onology om Finnish Lapland. P e ious ee- ing and ice-co e s udies, as well as his o ical documen a y e idence, we e also d awn upon. The mos se e e o he la e-Holocene e en s (1601 A.D. and 536 A.D.) clea ly illus a ed he dis an e ec s o explosi e e up ions in he s udy egion o no he n Eu ope. In pa icula , he e en in 1601 A.D. exempli ied educed su i al po en ial o his o ic humans when subjec ed o a olcanic-induced clima ic su p ise. Lack o cohe ence be ween e en s in e ed om ee- ing da a and ice-co e sulpha e da a became e iden o e he B.C. e a. I may be ai o s a e ha hese inconsis encies esul om inc easing ice-laye coun ing e o s and p oblems ela ed o annual laye iden i ica ion in he ice co es. On he o he hand, he ee- ing da a yield mo e accu- a e da es o cooling e en s bu do no p o e whe he a olcanic e up ion was he cause. None heless, he majo i y o he ee- ing da es o e he B.C. e a can be linked wi h sulpha e peaks in mul iple G eenland ice co es. O e all, he compa isons p oduced added e idence o linking he ee- ing and ice-co e signals o olcanism o e he p esen in e glacial. This pape also suppo s he idea o analysing and in e p e ing he human ecology da a in he con ex o geoch onological e idence. Such analyses will help us o ou line he e ec s ha ou socie y will encoun e wi h he nex sudden e en o olcanic- induced la ge-scale cooling. La ge and explosi e olcanic e up ions ha e occu ed se e al imes in he pas and i is highly likely he geological ac i i y o ou plane will p oduce simila e en s in he u u e, hough he iming o e up ions and he in ensi y o hei clima ic and en i onmen al e ec s may again su p ise people (S ee s & Glan z 2000). Examining iolen clima e changes in he Holocene shows ha hey ha e low-p obabili y o occu ing in he sho - e m bu ha he ecu ence o simila new e en s is una oidable in he long- e m. Socie y is now supposed o be p epa ing o g adual wa ming as he cu en clima ic ends a e p edic ed o con inue o e he nex cen u y (Be gs o ¨m e al. 2011), which could educe ou abili y o cope when a sudden e en o olcanic-induced cooling akes place. In he pe spec i e o human ecology, disas e s occu when he bu e ing capaci y o socie y is exceeded by na u al e en s (Ka es 1971; McLaughlin & Die z 2008). Clea ly, he clima e phenomena desc ibed in his pape ul il he c i e ia o an h opologically-de ined disas e s, in ol ing a combina ion o na u al and echno- logical ea u es and a popula ion in a condi ion o ulne - abili y (Oli e -Smi h 1996). Ou esea ch has indica ed he ad e se e ec s o ab up nega i e empe a u e anoma- lies on ood p oduc ion, which has ob ious ecological and economic implica ions (Holopainen & Helama 2009; Ricka d e al. 2010; Holopainen e al. 2012). Ou indings may be seen pa allel wi h a chaeological (Gamble e al. 2005) and palaeon ological (Foley 1994) esea ch which shows ha clima e changes s ongly dis u b human (and hominin) popula ions, which espond by con ac ing. Acknowledgemen s We hank P o esso Mike Baillie and an anonymous e e ee o hei c i ical e iew o he manusc ip . This s udy was suppo ed by he Academy o Finland (g an no. 251441). Re e ences Alley R.B., Meese D.A., Shuman C.A., Gow A.J., Taylo K.C., G oo es P.M., Whi e J.W.C., Ram M., Wadding on E.D., Mayewski P.A. & Zielinski G.A. 1993. Ab up inc ease in G eenland snow accumula ion a he end o he Younge D yas e en . Na u e 362, 527529. Alley R.B., Shuman C.A., Meese D.A., Gow A.J., Taylo K.C., Cu ey K.M., Fi zpa ick J.J., G oo es P.M., Zielinski G.A., Ram M., Spinelli G. & Elde B. 1997. Visuals a ig aphic da ing o he GISP2 ice co e: basis, ep oducibili y, and applica ion. Jou nal o Geophysical Resea ch*Oceans 102, 2636726382. Aniol R.W. 1983. T ee- ing analysis using CATRAS. Dend och onologia 1,4553. A ja a A. 2005. The Mys e y Cloud o 536 CE in he Medi e anean sou ces. Dumba on Oaks Pape s 59,7394. A well W.S. 2001. Volcanism and sho - e m clima ic change in Eas Asian and wo ld his o y. Jou nal o Wo ld His o y 12, 2998. Baillie M.G.L. 2008. P oposed e-da ing o he Eu opean ice co e ch onology by se en yea s p io o he 7 h cen u y AD. Geophysical Resea ch Le e s 35, L15813, doi: 10.1029/ 2008GL034755. Baillie M.G.L. 2010. Volcanoes, ice-co es and ee- ings: one s o y o wo? An iqui y 84, 202215. Baillie M.G.L. & Mun o M.A.R. 1988. I ish ee ings, San o ini and olcanic dus eils. Na u e 332, 344346. Be gs o ¨m I., Ma sson T., Niemela ¨E., Vuo enmaa J. & Fo sius M. 2011. Ecosys em se ices and li elihoods* ulne abili y and adap a ion o a changing clima e. VACCIA Syn hesis Repo . Helsinki: Finnish En i onmen Ins i u e. B adley R.S. 1988. The explosi e olcanic e up ion signal in No he n Hemisphe e con inen al empe a u e eco ds. Clima ic Change 12, 221243. B i a K.R., Ba holin T.S., Ecks ein D, Jones P.D., Ka le ´nW., Schweing ube F.H. & Ze e be g P. 1990. A 1,400-yea ee- ing eco d o summe empe a u es in Fennoscandia. Na u e 346, 434439. B i a K.R., Jones P.D., Schweing ube F.H., Ka le ´nW.& Shiya o S.G. 1996. T ee- ing a iables as p oxy-clima e indica o s: p oblems wi h low- equency signals. In P.D. A ch onology o mid- and la e-Holocene clima ic down u ns S. Helama e al. 8 (page numbe no o ci a ion pu pose) Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 Jones e al. (eds.): Clima ic a ia ions and o cing mechanisms o he las 2000 yea s. Pp. 941. Be lin: Sp inge -Ve lag. B i a K.R., Jones P.D., Schweing ube F.H. & Osbo n T.J. 1998. In luence o olcanic e up ions on No he n Hemi- sphe e summe empe a u e o e he pas 600 yea s. Na u e 393, 450455. Ches e D.K. 1988. Volcanoes and clima e: ecen olcanolo- gical pe spec i es. P og ess in Physical Geog aphy 12,135. Clausen H.B., Hamme C.U., H idbe g C.S., Dahl-Jensen D., S e ensen J.P., Kip s uhl J. & Leg and M. 1997. A com- pa ison o he olcanic eco ds o e he pas 4000 yea s om he G eenland Ice Co e P ojec and Dye 3 G eenland ice co es. Jou nal o Geophysical Resea ch*Oceans 102, 2670726723. Cook E., Shiya o S. & Mazepa V. 1990. Es ima ion o he mean ch onology. In E.R. Cook & L.A. Kai iuks is (eds.): Me hods o dend och onology: applica ions in he en i on- men al sciences. Pp. 123132. Do d ech : Kluwe Academic Publishe s. de Sil a S.L. & Zielinski G.A. 1998. Global in luence o he AD 1600 e up ion o Huaynapu ina, Pe u. Na u e 393, 455458. Eas e ling D.R., Meehl G.A., Pa mesan C., Changnon S.A., Ka l T.R. & Mea ns L.O. 2000. Clima e ex emes: obse a- ions, modeling, and impac s. Science 286, 20682074. E onen M., Hy a ¨ inen H. & Ze e be g P. 1999. Holocene humidi y changes in no he n Finnish Lapland in e ed om lake sedimen s and subme ged Sco s pines da ed by ee ings. The Holocene 9, 569580. E onen M., Ze e be g P., B i a K.R., Lindholm M., Me ila ¨inen J. & Timonen M. 2002. The sup along Sco s pine ee ing eco d o Finnish Lapland: pa 1, ch onology cons uc ion and ini ial e e ences. The Holocene 12, 673680. Fische E.M., Lu e bache J., Zo i a E., Te S.F.B., Cas y C. & Wanne H. 2007. Eu opean clima e esponse o opical olcanic e up ions o e he las hal millennium. Geophysical Resea ch Le e s 34, L05707, doi: 10.1029/2006GL027992. Foley R.A. 1994. Specia ion, ex inc ion and clima ic change in hominid e olu ion. Jou nal o Human E olu ion 26, 275289. F i s H.C. 1976. T ee ings and clima e. London: Academic P ess. F i s H.C., Mosimann J.E. & Bo o C.P. 1969. A e ised compu e p og am o s anda dizing ee- ing se ies. T ee- Ring Bulle in 29,1520. Gamble C., Da ies W., Pe i P., Hazelwood L. & Richa ds M. 2005. The a chaeological and gene ic ounda ions o he Eu opean popula ion du ing he La e Glacial: implica ions o ‘ag icul u al hinking’. Camb idge A chaeological Jou nal 15, 193223. Ge ais B.R. & MacDonald G.M. 2001. T ee- ing and summe - empe a u e esponse o olcanic ae osol o cing a he no he n ee-line, Kola Peninsula, Russia. The Holocene 11, 499505. G udd H., B i a K.R., Ka le ´n W., Ba holin T.S., Jones P.D. & K ome B. 2002. A 7400-yea ee- ing ch onology in no he n Swedish Lapland: na u al clima ic a iabili y exp essed on annual o millennial imescales. The Holocene 12, 657665. Helama S., Holopainen J., Timonen M., Ogu so M.G., Lindholm M., Me ila ¨inen J. & E onen M. 2004. Compa ison o li ing- ee and sub ossil ingwid hs wi h summe em- pe a u es om 18 h, 19 h and 20 h cen u ies in no he n Finland. Dend och onologia 21, 147154. Helama S., La ¨a ¨nelaid A., Tie a ¨ a ¨inen H., Macias Fau ia M., Kukkonen I.T., Holopainen J., Nielsen J.K. & Valo i a I. 2010. La e Holocene clima ic a iabili y econs uc ed om inc emen al da a om pines and pea l mussels*a mul i- p oxy compa ison o ai and subsu ace empe a u es. Bo eas 39, 734748. Helama S., Lindholm M., Me ila ¨inen J., Timonen M. & E onen M. 2005. Mul icen ennial ing-wid h ch onologies o Sco s pine along no h-sou h g adien ac oss Finland. T ee-Ring Resea ch 61,2132. Helama S., Lindholm M., Timonen M. & E onen M. 2004. De ec ion o clima e signal in dend och onological da a analysis: a compa ison o ee- ing s anda diza ion me hods. Theo e ical and Applied Clima ology 79, 239254. Helama S., Macias Fau ia M., Mielika ¨inen K., Timonen M. & E onen M. 2010. Sub-Milanko i ch sola o cing o pas clima es: mid and la e Holocene pe spec i es. Geological Socie y o Ame ica Bulle in 122, 19811988. Helama S., Mielika ¨inen K., Timonen M. & E onen M. 2008. Finnish sup a-long ee- ing ch onology ex ended o 5634 BC. No wegian Jou nal o Geog aphy 62, 271277. Helama S., Timonen M., Holopainen J., Ogu so M.G., Mielika ¨inen K., E onen M., Lindholm M. & Me ila ¨inen J. 2009. Summe empe a u e a ia ions in Lapland du ing he Medie al Wa m Pe iod and he Li le Ice Age ela i e o na u al ins abili y o he mohaline ci cula ion on mul i- decadal and mul i-cen ennial scales. Jou nal o Qua e na y Science 24, 450456. Helama S., Timonen M., Lindholm M., Me ila ¨inen J. & E onen M. 2005. Ex ac ing long-pe iod clima e luc ua ions om ee- ing ch onologies o e imescales o cen u ies o millennia. In e na ional Jou nal o Clima ology 25, 17671779. Hidalgo S., Monzie M., Almeida E., Chazo G., Eissen J.-P., an de Plich J. & Hall M.L. 2008. La e Pleis ocene and Holocene ac i i y o he A acazo Ninahuilca Volcanic Complex (Ecuado ). Jou nal o Volcanology and Geo he mal Resea ch 176,1626. Holmes R.L. 1983. Compu e -assis ed quali y con ol in ee- ing da ing and measu emen . T ee-Ring Bulle in 43,6975. Holopainen J. & Helama S. 2009. Li le Ice Age a ming in Finland: p eindus ial ag icul u e on he edge o he g im eape ’s scy he. Human Ecology 37, 213225. Holopainen J., Ricka d I.J. & Helama S. 2012. Clima ic signa u es in c ops and g ain p ices in 19 h-cen u y Sweden. The Holocene 22, 939945. Jones P.D., B i a K.R. & Schweing ube F.H. 1995. T ee- ing e idence o he widesp ead e ec s o explosi e olcanic e up ions. Geophysical Resea ch Le e s 22, 13331336. Ju ikkala E. 1958. Suomen alonpojan his o ia. (The his o y o he Finnish peasan .) Suomalaisen Ki jallisuuden Seu an oimi uksia 257. Helsinki: Suomalaisen Ki jallisuuden Seu a. S. Helama e al. A ch onology o mid- and la e-Holocene clima ic down u ns Ci a ion: Pola Resea ch 2013, 32, 15866, h p://dx.doi.o g/10.3402/pola . 32i0.15866 9 (page numbe no o ci a ion pu pose)