Im pact of cli ma tic changes in the Late Pleis to cene on mi gra tions and ex tinc tions of mam mals in Eu rope: four case stud ies
Mateusz BACA1, Adam NADACHOWSKI2, *, Grzegorz LIPECKI2, Pawe³ MACKIEWICZ3, Adrian MARCISZAK4, Danijela POPOVIÆ5, Pawe³ SOCHA4, Krzysztof STEFANIAK4 and Piotr WOJTAL2
1 Uni ver sity of War saw, Cen ter of Precolumbian Stud ies, Krakowskie Przedmieœcie 26/28, 00-927 Warszawa, Po land
2 In sti tute of Sys tem at ics and Evo lu tion of An i mals, Pol ish Acad emy of Sci ences, S³awkowska 17, 31-016 Kraków, Po land
3 Uni ver sity of Wroc³aw, De part ment of Genomics, Fac ulty of Bio tech nol ogy, Joliot-Cu rie 14a, 50-383 Wroclaw, Po land
4 Uni ver sity of Wroc³aw, De part ment of Paleozoology, In sti tute of En vi ron men tal Bi ol ogy, Sienkiewicza 21, 50-335 Wroc³aw, Po land
5 Uni ver sity of War saw, Cen tre of New Tech nol o gies, Banacha 2C, 02-097 Warszawa, Po land
Baca, M., Nadachowski, A., Lipecki, G., Mackiewicz, P., Marciszak, A., Popoviæ, D., Socha, P., Stefaniak, K., Wojtal, P., 2017. Im pact of cli ma tic changes in the Late Pleis to cene on mi gra tions and ex tinc tion of mam mals in Eu rope: four case stud - ies. Geo log i cal Quar terly, 61 (2): 291–304, doi: 10.7306/gq.1319
Cli mate changes that oc curred dur ing the Late Pleis to cene had pro found ef fects on the dis tri bu tion of many plant and an i mal spe cies and in flu enced the for ma tion of con tem po rary fau nas and flo ras of Eu rope. The course and mech a nisms of re - sponses of spe cies to past cli mate changes are now be ing in tensely stud ied by the use of di rect ra dio car bon dat ing and ge - netic anal y ses of fos sil re mains. Here, we re view the ad vances in un der stand ing these pro cesses by the ex am ple of four mam mal spe cies: woolly mam moth (Mammuthus primigenius), cave bear (Ursus spelaeus s.l.), saiga an te lope (Saiga tatarica) and col lared lem mings (Dicrostonyx ssp.). The cases dis cussed here as well as oth ers show that mi gra tions, range shifts and lo cal ex tinc tions were the main re sponses to cli mate changes and that the dy nam ics of these cli mate-driven pro - cesses were much more pro found than was pre vi ously thought. Each spe cies re acted in its in di vid ual man ner, which de - pended on its bi ol ogy and ad ap ta tion abil i ties to chang ing en vi ron men tal and cli ma tic con di tions. The most se vere changes in Eu ro pean eco sys tems that af fected the larg est num ber of spe cies took place around 33–31 ka BP, dur ing the Last Gla cial Max i mum 22–19 ka BP and the Late Gla cial warm ing 15–13 ka BP.
Key words: an cient DNA, ra dio car bon dat ing, mi gra tions, cli mate changes.
INTRODUCTION
The Late Pleis to cene was an in ter val marked by mul ti ple cli - mate changes of a mag ni tude greater than those ob served to - day (Ras mus sen et al., 2014). The im pact of those fluc tu a tions on Late Pleis to cene mam mal pop u la tions has been in tensely stud ied by ex am i na tion of changes in spa tial dis tri bu tion of fos - sil mam ma lian fau nas in time (e.g., Gra ham et al., 1996;
Sommer and Nadachowski, 2006; Lis ter and Stu art, 2008;
Markova and van Kolfschoten, 2008). This pe riod wit nessed also mass ex tinc tion on an un prec e dented scale not ob served in fos sil re cords for mil lions of years (Sandom et al., 2014; Stu - art, 2015). By the end of the Pleis to cene, most ter res trial megafaunal spe cies (heavier than 44 kg) be came ex tinct or their pop u la tion sizes de creased sub stan tially (Barnosky et al., 2004; Koch and Barnosky, 2006). The course and the tim ing of
these events dif fer from re gion to re gion and the causes of these ex tinc tions are the sub ject of on go ing de bate. Some re - search ers point only to cli mate changes, some blame hu man ac tiv i ties such as overhunting and hab i tat al ter na tion (Alroy, 2001; Sandom et al., 2014) whereas oth ers sug gest a com bi na - tion of these two causes as the main trig ger (Barnosky et al., 2004; Prescott et al., 2012; Stu art, 2015).
In re cent years two main ad vances ac cel er ated re search in this field. The first one was an in crease in the ac ces si bil ity and pop u lar ity of AMS ra dio car bon dat ing. The grow ing num ber of di rectly dated fos sils al lowed for pre cise track ing of changes in spe cies’ geo graphic dis tri bu tion, mi gra tions and dat ing ex tinc - tion events (e.g., MacPhee et al., 2002; Stu art, 2005; Sommer et al., 2008, 2014; Pacher and Stu art, 2009; Stu art and Lis ter, 2012, 2014). It en abled di rect cor re la tion be tween these events and cli ma tic and en vi ron men tal changes. The sec ond ad vance was the emer gence of an cient DNA stud ies. The ex am i na tion of ge netic di ver sity has added an other level of com plex ity to re - search into the Late Pleis to cene pop u la tions. An cient DNA fa - cil i tated in ves ti ga tions of intraspecific pro cesses, such as pop u - la tion re place ments or changes in their di ver sity; events that usu ally do not man i fest in the fos sil re cord. These two meth ods
* Corresponding author, e-mail: nadachowski@isez.pan.krakow.pl Received: April 18, 2016; accepted: June 13, 2016; first published online: March 6, 2017
com bined proved to be the most ver sa tile ap proach in re con - struc tion of past pop u la tion his to ries (e.g., Cam pos et al., 2010a, b; Lorenzen et al., 2011; Horn et al., 2014; Palkopoulou et al., 2016).
The main aim of our study is to re view the re cent prog ress that has been made in re search into four spe cies: woolly mam - moth (Mammuthus primigenius), cave bear (Ursus spelaeus s.l.), saiga an te lope (Saiga tatarica) and col lared lem ming (Dicrostonyx ssp.) and to dis cuss its im pli ca tions for our un der - stand ing of the im pact of cli mate changes on mi gra tions and ex tinc tions of mam mals in the Late Pleis to cene.
WOOLLY MAMMOTH (MAMMUTHUS PRIMIGENIUS)
El e phants (Elephantidae) are the larg est ter res trial mam - mals, with a stout body, char ac ter is tic long, highly pro nounced proboscid or trunk, a com bi na tion of nose and up per lip, large ears, col umn-like limbs, and small tail. The woolly mam moth rep re sents el e phants well-adapted to the cold and arid steppe-tun dra en vi ron ment (Maschenko, 2002; Lis ter et al., 2005). This spe cies was wide spread dur ing the Late Pleis to - cene from West ern Eu rope through the whole of north ern Asia
to the north ern part of North Amer ica (Kahlke, 2015). Mor pho - log i cal data has re vealed that woolly mam moths were pres ent in Eu rope from around 200 ka BP un til the end of the Pleis to - cene (Lis ter and Sher, 2001; Lis ter et al., 2005). How ever, ge - netic stud ies in di cated that the Late Pleis to cene his tory of Mammuthus primigenius was char ac ter ized by a com plex se - ries of range ex pan sions and con trac tions, de mo graphic changes and clade re place ments (Palkopoulou et al., 2013).
Phylo gen etic anal y ses re vealed that Holarctic mam moths be longed to three dis tinct mi to chon drial (mtDNA) lin eages (Fig.
1). The most wide spread lin eage I had a nearly Holarctic dis tri - bu tion, lin eage II was con fined to Cen tral-East Asia while spec i - mens from lin eage III are known from Eu rope (Debruyne et al., 2008; Palkopoulou et al., 2013). The di ver gence of lin eages I and II was pre vi ously es ti mated to ca. 1 Ma (Debruyne et al., 2008; Gilbert and Drautz, 2008), how ever, most re cent es ti ma - tions sug gest a much youn ger date, about 300 ka BP (Palkopoulou et al., 2013). Co ales cent sim u la tions sug gested that a split of three mam moth pop u la tions took place around 200 ka BP and was fol lowed by a de mo graphic ex pan sion that started around 121 ka BP (Palkopoulou et al., 2013). This ex - pan sion co in cides broadly with the end of Eemian Interstadial, which sug gests that mam moths sur vived this warm pe riod con - fined to refugial ar eas and ex panded as cli mate be came cooler at the be gin ning of the Weichselian Gla ci ation (Palkopoulou et
Fig. 1. Woolly mam moth (Mammuthus primigenius)
A – Bayesian phy log eny of Holarctic woolly mam moths based on mtDNA cytochrome b se quences. The tree is a chronogram where branch lengths de note time elapsed since di ver gence and the po si tions of tips cor re spond to cal i brated ra dio car bon ages of sam ples; B – dis tri bu - tion of palaeontological sites with woolly mam moth re mains ra dio car bon-dated to the pe ri ods in di cated. Colours in di cate mi to chon drial DNA lin eages (mod i fied af ter Palkopoulou et al., 2013)
al., 2013). Sur pris ingly, this was not sup ported by the anal y ses of whole palaeogenomes, which in di cated a much ear lier ex - pan sion ca. 280 ka years ago and a max i mum ef fec tive pop u la - tion size dur ing the Eemian (Palkopoulou et al., 2015).
De spite these am bi gu ities in the early his tory of mam moth pop u la tions, an cient DNA re vealed also two more re cent pop u - la tion turn overs. In the Eemian Inter gla cial and Early Weichse - lian, woolly mam moths that be longed to clade I were most prob - a bly con fined to North Amer ica. It was es ti mated that ca. 66 ka BP they started to ex pand west wards through Beringia and reached the area oc cu pied by mam moths of clade II. Both pop - u la tions lived in sympatry for 20 ka, when clade II sud denly dis - ap peared from the fos sil re cord ca. 40 ka BP (Palkopoulou et al., 2013). Woolly mam moths be long ing to clade I ex panded far ther west and reached Eu rope. The ear li est spec i men in Eu - rope orig i nat ing from this clade came from the site in Vologda Oblast in Rus sia and was dated to 32 ka BP. The ap pear ance of clade I in Eu rope co in cides with the dis ap pear ance of the en - demic Eu ro pean pop u la tion (clade III), with the lat est spec i men dated to ca. 34 ka BP. There is, how ever, no ev i dence for any over lap be tween these pop u la tions and it seems that the ex tinc - tion of clade III was not driven by the ap pear ance of new com ers from the east. This sce nario is sup ported by the lack of ra dio - car bon dat ing of mam moth re mains in Cen tral Eu rope be tween ca. 34 and 33 ka BP, e.g. in Po land (Nadachowski et al., 2011;
Fig. 2). An im por tant in crease of mam moth pop u la tion size in Eu rope took place be tween 31 and 29 ka BP, which is con - firmed by dense ra dio car bon dat ing of mam moth re mains col - lected from al most the whole North Eu ro pean Plain (Nada -
chowski et al., 2011; Ukkonen et al., 2011; Markova et al., 2013). In ter est ingly, dur ing al most the whole of the Last Gla cial Max i mum (LGM), be tween ca. 22 and 18 ka BP (Fig. 2), there are no dated mam moth re cords in north west ern, north ern and cen tral Eu rope sites (Stu art et al., 2004; Nadachowski et al., 2011; Ukkonen et al., 2011), which sug gests a long-term con - trac tion. Mam moths re-im mi grated to Eu rope for the next 3–4 mil len nia at the end of GS-2 (Green land Stadial 2) with the re - turn of se vere cli ma tic con di tions and open veg e ta tion (Lis ter and Stu art, 2008; Nadachowski et al., 2011; Fig. 2). The next break in the mam moth re cord co in cides roughly with the BÝlling -AllerÝd warm ing (Green land Interstadial 1; GI-1) fol - lowed by one more recolonization of Lat via and Es to nia, dated to the Youn ger Dryas (GS-1) (Láugas et al., 2002; Stu art et al., 2002; Ukkonen et al., 2011; see Fig. 7). The last pop u la tions of Mammuthus primigenius in Eu rope lived in north west ern Rus - sia and dis ap peared around the Youn ger Dryas-Ho lo cene boun d ary ca. 11.8–11.4 ka BP (Láugas et al., 2002; Ukkonen et al., 2011).
CAVE BEAR (URSUS SPELAEUS S.L.)
The bear fam ily (Ursidae) are large mam mals with a big head and thick neck, small eyes and short tail, mus cu lar bod ies with stout legs and large paws. Most bears are om ni vores, al - though the di ets of po lar bear (Ursus maritimus) and gi ant
Fig. 2. Woolly mam moth (Mammuthus primigenius)
Ra dio car bon dates of woolly mam moth re mains from Po land along side the NGRIP GICC05 ice core d18O re cord (Svens son et al., 2008).
Dates were com piled from Nadachowski et al. (2011), Paw³owska (2015), Wilczyñski et al. (2015), Wojtal and Wilczyñski (2015) and recalibrated in OxCal v. 4.2 (Ramsey, 2009) us ing the Intcal13 cal i bra tion curve (Reimer et al., 2013). MIS – Ma rine Iso to pic Stages, GS – Green land Stadials, GI – Green land Interstadials; LGM – Last Gla cial Max i mum (af ter Mix et al., 2001). Grey stripes il lus trate in ferred pe ri - ods of the ab sence of mam moths from Po land
panda (Ailuropoda melanoleuca) are very nar row and spe cial - ized. The for mer is a car ni vore and the lat ter is an ob li gate con - sumer of bam boo. It is as sumed that the cave bear (Ursus spelaeus s.l.), one of the most wide spread mam mals in the Late Pleis to cene in Eu rope, was an ob li gate veg e tar ian (Baca et al., 2016). It evolved from the Mid dle Pleis to cene Ursus deningeri and dif fer en ti ated into sev eral forms rec og nized at mor pho log i - cal or ge netic lev els (Hofreiter et al., 2002; Rabeder et al., 2004a, b). In Eu rope, two main forms, U. spelaeus and U.
ingressus ex isted. They dif fer en ti ated prob a bly be tween 414,000 and 173,000 years ago (Knapp et al., 2009). U. spela - eus lived mainly in West ern Eu rope and its re mains have been found in Spain, France, Ger many, Bel gium, It aly and Aus tria, al though it has also been re corded in Altai (Rabeder et al., 2004b; Knapp et al., 2009). U. ingressus in hab ited south east - ern and Cen tral Eu rope and its re mains have been dis cov ered mainly in Ro ma nia, Slovenia, Ukraine, Czech Re pub lic, Po - land, Slovakia and Greece but also in Aus tria, Ger many and Swit zer land (Rabeder et al., 2004b; Baca et al., 2014). More - over, two small, dwarf cave bear forms that are con sid ered as sub spe cies, U. spelaeus eremus and U. spelaeus ladinicus have been re ported in high al pine caves in Aus tria and It aly (Rabeder and Hofreiter, 2004; Rabeder et al., 2004a; Fig. 3A).
An other ma jor group of large bears, named U. deningeri kudarensis, was dis cov ered in the Cau ca sus (Baryshnikov, 1998; Knapp et al., 2009). Re cent anal y sis of mtDNA of the Mid dle Pleis to cene U. deningeri from Sima de los Huesos, Atauperca, Spain, re vealed that U. deningeri kudarensis con sti - tutes the most di ver gent cave bear lin eage (Dabney et al., 2013). It was sug gested that it be longed to a sep a rate branch of cave bear evo lu tion and hence its tax o nomic sta tus was changed to U. kudarensis (Dabney et al., 2013; Stiller et al., 2014).
The geo graphic or i gin of U. ingressus is not known, but the basal po si tion of haplotypes from the Ro ma nian site Peêtera cu Oase in the phy log eny of this spe cies points to south east ern Eu rope (Baca et al., 2012). The phylogeographic pic ture of U.
ingressus is un clear as the mtDNA phy log eny lacks sig nif i cant sup port and clear phylogeographic lin eages are not dif fer en ti - ated (Fig. 3A). It seems, how ever, that the spread of U.
ingressus in Eu rope may have pro ceeded in de pend ently along the main Eu ro pean moun tain ranges, the Alps and Carpathians (Baca et al., 2014).
It has been pro posed that be tween 60 and 50 ka BP U.
ingressus started west ward mi gra tion along the Alps (Hofreiter et al., 2004; Rabeder and Hofreiter, 2004; Münzel et al., 2011).
The ear li est re mains of this spe cies from the Aus trian and Swiss Alps, dated to ca. 50 ka BP, are known from Gamssulzen Cave. The wes tern most re cord of U. ingressus came from the Schnurenloch site near Bern, Swit zer land, but spe cies at tri bu - tion was based solely on mor phol ogy with out ge netic con fir ma - tion (Rabeder and Hofreiter, 2004). Re cently, sev eral cave bear spec i mens from the Kraków-Czêstochowa Up land, Po land yielded mtDNA haplotypes iden ti cal to those found in Alpean U.
ingressus, which sug gests that this pop u la tion spread also north wards be yond the Carpathian Arc (Popoviæ et al., 2015).
In some cases, the ap pear ance of U. ingressus was as so ci - ated with the dis ap pear ance of other cave bear forms that in - hab ited the area. The most in ter est ing ev i dence of such re - place ment came from the three cave sites in the Ach Val ley in Swabian Jura, Ger many: Hohle Fels, Geißenklösterle, and Sirgenstein (Hofreiter et al., 2007; Münzel et al., 2011). The lat - est oc cur rence of the na tive in hab it ant in this area, U. spelaeus, was dated to ca. 31.5 ka BP while the ear li est re cord of U.
ingressus from Geißenklösterle was dated to 36.3 ka BP. Only a sin gle U. ingressus in di vid ual yielded such a date while most of oth ers were dated to ca. 32 ka BP, which sug gests that the main im mi gra tion took place just be fore the lo cal ex tinc tion of U.
spelaeus. A sim i lar re place ment was re corded in Herdengel Cave in Aus tria, where U. ingressus re placed one of the small cave bear forms, U. s. eremus. All of the eremus cave bears were dated by strati graphic con text to more than 60 ka BP, whereas all U. ingressus were youn ger than 37 ka BP (Stiller et al., 2014). The ap pear ance of U. ingressus did not al ways re sult in the re place ment of other cave bear forms. In the two Aus trian caves, Ramesch and Gamssulzen, lo cated about 10 km from each other, U. ingressus lived side by side with U. s. eremus for at least 15,000 years (Hofreiter et al., 2004).
Less is known about the cave bear pop u la tion that in hab ited sur round ings of the West ern Carpathians and Sudetes. Ra dio -
Fig. 3. Cave bear (Ursus spelaeus s.l.)
A – Bayesian phy log eny of Eu ro pean cave bears based on mtDNA con trol re gion se quences. The tree is a phylogram where branch lengths are pro por tional to amounts of se quence dif fer ences; B – spa tial dis tri bu tion of cave bear re mains clas si fied as U. s. spelaeus and U.
ingressus (af ter Popoviæ et al., 2015). Colours in di cate mi to chon drial DNA lin eages of cave bears
car bon dates sug gest more or less con tin u ous pres ence of cave bears in this area (Nadachowski et al., 2008; Wojtal et al., 2015) dur ing the Late Pleis to cene. Pre lim i nary re sults of ge - netic in ves ti ga tion of spec i mens from mul ti ple sites in Po land, the Czech Re pub lic, Slovakia and Ukraine re vealed that U.
ingressus was the only form of cave bear pres ent in this part of Eu rope (Popoviæ et al., 2015; Fig. 3B). It ap peared in the Carpathians and Sudetes much ear lier than in the Alps. The ear li est re mains con firmed with mor pho log i cal and an cient DNA anal y ses are from NiedŸwiedzia Cave in the Sudetes.
Bone col la gen from one spec i men from this cave was dated with the U-Th method to about 80 ka BP (with an ac com pa ny ing
14C date of >50 ka 14C BP) and the other was stratigraphically dated to more than ca. 70 ka BP (Baca et al., 2014). The mtDNA haplotypes of spec i mens from NiedŸwiedzia Cave formed a di ver gent clus ter in phylo gen etic trees, which con firms the early sep a ra tion and ex pan sion of this pop u la tion (Baca et al., 2012, 2014). Re cently, two cave bears with sim i lar mtDNA haplotypes were re corded far ther west in Zoolithen Cave (Up - per Franconia, Ger many) (Stiller et al., 2014). In this cave, re - mains of U. spelaeus were also dis cov ered and in ter est ingly, an op po site re place ment was sug gested as the es ti mated age of both U. ingressus spec i mens was older than that of nine U.
spelaeus in di vid u als (Stiller et al., 2014). These re sults are, how ever, based on mo lec u lar dat ing of the re mains and should be fur ther con firmed with di rect ra dio car bon dat ing.
An other de bated is sue in cave bear pop u la tion his tory is the tim ing and causes of their ex tinc tion. Di rect ra dio car bon dates in di cate that last cave bears went ex tinct prior to the LGM. Un til re cently, it was thought that they dis ap peared from the fos sil re - cord quite syn chro nously in dif fer ent parts of Eu rope around 2414C ka BP (28 ka BP) at the end of GI-3 (Hofreiter et al., 2002;
Pacher and Stu art, 2009; Bocherens et al., 2014). Palaeo - genetic anal y ses showed, how ever, that the de mise of cave bears started ca. 50 ka 14C BP (Stiller et al., 2010), thus about 25,000 years be fore their fi nal ex tinc tion. It has been pro posed that the chang ing cli mate was one of the main causes of the cave bear ex tinc tion (Baca et al., 2016).
Cave bears are gen er ally con sid ered her biv o rous based on their craniodental ad ap ta tions (Kurtén, 1976; Mackiewicz et al., 2010; Wiszniowska et al., 2010; van Heteren et al., 2014) and most of the stud ies of sta ble iso topes (d13C, d15N) from bone and tooth col la gen con firm that they were strict veg e tar i ans (Bocherens et al., 1994, 1997; Taboada et al., 1999; Fernán - dez -Mosquera et al., 2001; Münzel et al., 2011; Krajcarz et al., 2016). The cli mate change which be gan af ter GS-3 caused se - vere trans for ma tions in plant com mu ni ties all around Eu rope (Helmens, 2014). Veg e ta tion sea sons short ened and the avail - abil ity of high qual ity plant ma te rial, which seems cru cial for the sur vival of cave bears, de creased. Di etary hab its of cave bears did not change dur ing their pres ence for the last 10,000 years in Eu rope and this eco log i cal niche con ser va tism may have led to the de cline of their pop u la tions (Bocherens et al., 2014).
Be sides the en vi ron men tal changes sev eral other fac tors might have in flu enced the cave bear pop u la tions. There is a sub stan tial ev i dence of hunt ing of cave bears by hu mans (Münzel et al., 2011; Wojtal et al., 2015), as well as com pe ti tion by them for caves as a shel ter (Grayson and Delpech, 2003).
Pos si bly also large car ni vores such as cave lion (Panthera spelaea) and cave hy ena (Crocuta crocuta spelaea) hunted hi - ber nat ing cave bears (Diedrich, 2014). Anal y sis of whole mi to - chon drial genomes of bears from sev eral sites in north west ern Spain re vealed re cently that cave bears, in con trast to the brown bears, ex hib ited strong fi del ity to the caves were they were born. It was sug gested that fac ing grow ing com pe ti tion with mod ern hu mans and Neanderthals such hom ing be hav -
iour might have con trib uted to their ex tinc tion. Ten dency to come back to the same cave for dor mancy may pre vented col o - ni za tion of new caves and made cave bears a pre dict able prey for hunt ers (Fortes et al., 2016).
Re cently sev eral young cave bear spec i mens were re - ported in West ern Eu rope. One spec i men from the Rochedane site (French Jura), con firmed with an cient DNA as U. spelaeus, yielded an AMS date of about 28.5 ka BP (Bocherens et al., 2014). Two spec i mens from Chiostraccio Cave (Siena, It aly) were dated to ca. 28 and 27 ka BP, re spec tively (Mar tini et al., 2014). Stiller et al. (2014) sug gested that cave bear pop u la tions might have de clined from east to west as most of the sam ples youn ger than 30 ka BP were found in West ern Eu rope. How - ever, young spec i mens have also been re ported in East ern Eu - rope (Fig. 4). Two spec i mens from the Kraków-Czêstochowa Up land, from Deszczowa and Komarowa caves, were dated to 28.6 ka BP and con firmed as U. ingressus (Popoviæ et al., 2015;
Wojtal et al., 2015). An other one from Isabela Textorisova Cave (Velka Fatra Mts., Slovakia) (Sabol et al., 2014) was sim i larly dated to 28.7 ka BP. The youn gest ge net i cally con firmed cave bear spec i men so far come from Stajnia Cave in Po land and was dated to around 26.1 ka BP. Baca et al. (2016) gath ered 206 ra dio car bon dated spec i mens and fol low ing eight ap - proaches es ti mated the ex tinc tion time of U. spelaeus sensu lato to be tween 27.0 and 24.3 ka BP. These find ings clearly in - di cate that the pat tern rec og nized by Stiller et al. (2014) was a re sult of sam pling bias and that the late cave bear sur vived in - de pend ently in iso lated pop u la tions in dif fer ent parts of Eu rope, even into the mid dle of GS-3 stadial. Es pe cially the karst re - gions may have pro vided suit able mi cro cli mate for long sur vival of this spe cies (Baca et al., 2016).
The large amount of di rectly dated and ge net i cally ana lysed cave bear re mains pro vide in sight into cave bear pop u la tion dy - nam ics in Eu rope. However, some as pects of the mode of life of cave bears and de tails of their ex tinc tion, evo lu tion and phylo - geography still await explanation.
SAIGA ANTELOPE (SAIGA TATARICA)
Hol low-horned ru mi nants (Bovidae) are the most di verse fam ily in the or der Artiodactyla. More del i cately-built spe cies, in - clud ing the ge nus Saiga, be long to the Antilopini to gether with ga zelles and other an te lopes (Groves and Leslie, 2011). The bi - zarre-look ing saigas are not re lated to sheep and goats, as used to be thought, but are the sis ter group of the ga zelles.
They are non-ter ri to rial and no madic an te lopes, gath er ing in mas sive herds of many thou sands of in di vid u als be fore mi gra - tions. The saiga is a spe cial ized steppe her bi vore adapted to flat plains and avoid ing rug ged ter rains. At pres ent it in hab its dry steppes and semi-deserts but dur ing the Late Pleis to cene it was wide spread in vast ar eas of Eur asia and North Amer ica, be long ing to so-called “Mammuthus–Coelodonta” fau nal com - plex (Kahlke, 1999, 2014). The num ber of dis persal events to Eu rope dur ing the Late Pleis to cene was lim ited by the eco log i - cal re quire ments of this spe cies and by weather con di tions. The large thick ness of snow cover in win ter was the larg est re stric - tion on long-dis tance mi gra tions of saiga.
There is still hot de bate whether the saiga rep re sents two or one spe cies. Some palaeontologists dis tin guish two fos sil spe - cies: Saiga bo re alis and S. tatarica (Baryshnikov and Tikhonov, 1994), oth ers only one spe cies with two sub spe cies, S. tatarica bo re alis and S. tatatrica tatarica in spite of some dis tinct dif fer - ences found in the skull mor phol ogy be tween both forms
(Kahlke, 1991; Ratajczak et al., 2016). The cur rent ge netic stud ies also sug gest that both fos sil and re cent saiga should be clas si fied as one spe cies, Saiga tatarica (Kholodova et al., 2006; Cam pos et al., 2010b).
The per ma nent oc cur rence of this an te lope in steppe ar - eas in East ern Eu rope north of the Black Sea, in the Cri mea and Dobruja (Dobrogea) in Ro ma nia dur ing all of MIS 3 can not be ques tioned (Markova et al., 1995; Péan et al., 2013; Ridush et al., 2013). Pre vi ously pub lished over views sug gest that the
saiga was also pres ent in West ern and Cen tral Eu rope in MIS 3 (Stew art, 2007; Markova et al., 2010). How ever, this ev i - dence was not re cently con firmed by di rect ra dio car bon dat ing (Nada chowski et al., 2016). Dur ing MIS 2 and the early part of MIS 1 (the Late Gla cial), saiga ap peared in Eu rope in three im - mi gra tion waves (Fig. 5). The old est one was prob a bly re - stricted to Cen tral Eu rope (e.g., Po land and Czech Re pub lic;
Nada cho wski et al., 2016) and was tem po rally lim ited to the pe riod just be fore GI-2 (Fig. 5A), i.e. a short warm ing be tween
Fig. 5. Saiga an te lope (Saiga tatarica)
A – a plot of di rect ra dio car bon dates of saiga an te lope in Eu rope (Langlais et al., 2015; Nadachowski et al., 2016; Barshay-Szmidt et al., 2016) along side the NGRIP GICC05 d18O ice core re cord, cal i brated us ing pro gram OxCal v. 4.2 (Ramsey, 2009); GS – Green land Stadials, GI – Green land Interstadials; B – spa tial dis tri bu tion of sites with saiga re mains in Eu rope; white cir cles de note sites with re mains dated by ar - chae o log i cal con text, while col oured cir cles in di cate sites with ra dio car bon-dated re mains; colours cor re spond to three pu ta tive waves of mi - gra tion of saiga into Eu rope (sup ple mented and mod i fied af ter Nadachowski et al., 2016)
Fig. 4. Cave bear (Ursus spelaeus s.l.)
A – dis tri bu tion of palaeontological sites with cave bear re mains youn ger than 32 ka BP; B – cave bear re mains youn ger than 32 ka BP along side the NGRIP GICC05 d18O ice core re cord; spec i mens ge net i cally de ter mined as U. spelaeus are col oured in blue,
and U. ingressus in green
23.3 and 22.9 ka BP (Ras mus sen et al., 2014). The sec ond, geo graph i cally much wider mi gra tion started just af ter the LGM, ca. 19.5 ka BP, and lasted for ca. 3,500–4,000 years to ca. 15.5 ka BP (Nadachowski et al., 2016; Fig. 5). Dur ing this time, Saiga tatarica was pres ent in the whole of Eu rope north of the Carpathians, Alps and Pyr e nees and reached the Aquitaine Ba sin and Gascony in SW France. The com par i son of a still lim ited num ber of di rect dates sug gests rather mi gra - tions of saiga from west ern Eu rope to its east ern refugial ar - eas be cause the dates from France (Langlais et al., 2015;
Barshay-Szmidt et al., 2016) are older than those from Ger - many and Po land (Nadachowski et al., 2016; Fig. 5). This in - ter est ing ob ser va tion should be, how ever, con firmed by more di rect ra dio car bon dat ing. The re duc tion in range of the saiga in Eu rope was con tin ued in the Late Gla cial (Fig. 5) due to the de vel op ment of veg e ta tion cover not suit able for this her bi - vore. How ever, dur ing cooler ep i sodes of the Late Gla cial, be - tween the BÝlling and AllerÝd (GI-1d, for mer Older Dryas) and within the AllerÝd (GI-1c2) (Ras mus sen et al., 2014) Saiga tatarica prob a bly re turned to Cen tral Eu rope, but this re-im mi - gra tion was re stricted only to the area north of the Carpathians (Nadachowski et al., 2016; Fig. 5). It seems that this spe cies was un able to ex tend its range to the west of Eu rope dur ing the Youn ger Dryas (GS-1), the last cold phase of the Pleis to - cene, even though the cli ma tic and en vi ron men tal con di tions were suit able for this her bi vore.
The his tory of saiga in the Eu ro pean Late Pleis to cene, es - pe cially the di rec tions of mi gra tions (west ward or east ward), needs fur ther in ves ti ga tion bas ing on new ra dio car bon dates and on palaeogenetic stud ies to ex plain the com plex mi gra tion events of this endangered ruminant.
COLLARED LEMMING (DICROSTONYX SSP.)
Mor pho log i cal and mo lec u lar ev i dence shows that voles and lem mings cre ate a monophyletic group of ro dents placed by most au thors within the Arvicolinae subfamily (Chaline et al., 1999). Col lared lem mings (Dicrostonyx ssp.) are cold-adapted an i mals that are re stricted to dry and tree less Arc tic tun dra en - vi ron ments (Kowalski, 1995). They are a key spe cies in trophic net works in Arc tic eco sys tems as they are prey for pred a tors such as arc tic fox, snow owl and stoat. Four con tem po rary spe - cies are rec og nized within the ge nus: D. torquatus that in hab its Eur asia from White Sea to Be ring Strait, as well as D. groenlan - dicus, D. richardsoni and D. hudsonius that in habit North Amer - ica. These spe cies have been rec og nized on the ba sis of phylo - gen etic anal y sis of mi to chon drial DNA, karyotype di ver sity and hy bridi sa tion ex per i ments (Jarrell and Fregda, 1993; Fedorov and Goropashnaya, 1999). This clas si fi ca tion is cur rently the most com monly ac cepted one.
At pres ent col lared lem mings show nearly cir cum po lar dis - tri bu tion and do not ex ceed lat i tude 65°N, but in the Pleis to - cene their range was much wider and their re mains are known from palaeontological sites through out Eu rope and Asia (e.g., Mar kova et al., 2010; Ponomarev and Puzachenko, 2015).
Pleis to cene col lared lem mings are clas si fied ac cord ing to the grow ing com plex ity of occlusal sur face of mo lar teeth. The old - est ones are known from the Early Pleis to cene and are clas si - fied as Predicrostonyx (Nadachowski, 1992), which was fol - lowed by D. renidens, D. meridionalis and D. simplicior in the Mid Pleis to cene (Zazhigin, 1980; Smirnov et al., 1986) and D.
gulielmi in the Late Pleis to cene. It was sug gested that the tran - si tion be tween D. gulielmi and the re cent D. torquatus took place in the Late Gla cial about 15–13 ka BP. Morphometric anal y ses of mo lar teeth re vealed that dur ing this time, a sub - stan tial change in the struc ture of col lared lem ming pop u la - tions took place (Smir nov, 2002). This was fur ther sup ported by ge netic anal y ses of con tem po rary col lared lem ming pop u - la tions. Phylogeographic anal y ses re vealed the ex is tence of five allo patric pop u la tions within the pres ent dis tri bu tion of the spe cies (Fedorov et al., 1999). Each of the pop u la tions was char ac ter ized by low ge netic di ver sity, which was in ter preted as an ef fect of the re gional pop u la tion re duc tions that might have taken place dur ing the warm in ter vals of the Ho lo cene.
Other ev i dence came from the palaeogenetic anal y ses of col - lared lem ming re mains from the Pymva Shor site in the north - ern Pre- Urals that re vealed the sig na ture of a se vere pop u la - tion bot tle neck around 14.5 ka BP (Prost et al., 2010). The cli - mate warm ing dur ing the BÝlling- AllerÝd (GI-1a-c) pe riod in the Late Gla cial was ac com pa nied by a north ward ex pan sion of for ests and sup pos edly this forced the con trac tion and iso la - tion of col lared lem ming pop u la tions that re sulted in their de - cline (Fedorov, 1999).
Re cently, Palkopoulou et al. (2016) used an cient DNA and ra dio car bon dat ing to re con struct the Late Pleis to cene evo lu - tion ary his tory of col lared lem mings. They ob tained mtDNA se quences from more than 300 col lared lem ming spec i mens from mul ti ple palaeontological sites, and 48 di rect ra dio car bon AMS dates. Phylo gen etic anal y ses re vealed that five dis tinct pop u la tions of col lared lem mings, rep re sented by five mi to - chon drial lin eages (EA1–EA5; Fig. 6A), ex isted in Eu rope.
Each of the pop u la tions was widely dis trib uted, rang ing from West ern France to the Ural Moun tains. The ra dio car bon dat - ing re vealed that sub se quent pop u la tions suc ceeded each other through time dur ing the last ca. 50 ka. This was in ter - preted as a se ries of col lared lem ming pop u la tion ex tinc tions and recolonizations that took place across the whole Eu ro - pean con ti nent. The phylo gen etic pat tern found in di cated that sub se quent recoloni zations pro ceeded in an east to west di - rec tion from a hy po thet i cal refugial area in north east ern Si be - ria. Dates for the two ear li est pop u la tions (EA1 and EA2) are close to ra dio car bon dat ing lim its but it seems that both were al ready pres ent in Eu rope prior to 50 ka BP. The EA1 pop u la - tion van ished around 50 ka BP and the EA2 pop u la tion around 42.3 ka BP. Pop u la tion EA3 ap peared in Eu rope ca. 32 ka BP and dis ap peared 22.8 ka BP and was fol lowed by pop u la tion EA4, dated to a short pe riod be tween 22.2 and 20.5 ka BP.
Pop u la tion EA5 emerged ca. 20.3 ka BP and dis ap peared from Eu rope ca. 14.5 ka BP. All Ho lo cene and mod ern col - lared lem mings from Si be ria be long to this EA5 mtDNA lin - eage (Fig. 6B).
It has been pos tu lated that the pop u la tion turn overs that took place across such a large geo graph ical range had to be driven by en vi ron men tal or eco log i cal changes in the steppe - -tun dra eco sys tem. The dat ing of the two ear li est pop u la tions (EA1 and EA2) is not very pre cise and the con nec tion be tween their de mise and cli mate changes re mains un clear. The spread of pop u la tion EA3, which seems re li ably dated, started just af ter GI-5 and might be as so ci ated with cli mate cool ing that be gun with the on set of GS-5. Sur pris ingly, the two fol low - ing pop u la tion turn overs, i.e. the re place ment of pop u la tion EA3 by EA4 and EA4 by EA5, took place within the LGM. Dur - ing this pe riod, arc tic tun dra en vi ron ments pre ferred by the col lared lem mings spread over large ar eas of Eur asia (Tarasov et al., 2000). Thus, the causes of these re place -
ments re main un clear; how ever, it has been shown that they co in cide with brief warm ing pe ri ods re corded in the high res o - lu tion palynological re cord from Lago Grande di Montichio, It - aly (Palkopoulou et al., 2016).
DISCUSSION AND CONCLUSIONS
High-res o lu tion re con struc tions of mi gra tion and ex tinc tion events dur ing the Late Pleis to cene ob tained in re cent years show that spe cies re sponded to cli mate changes ac cord ing to their in di vid ual ad ap ta tions and there are no grounds for con - sid er ing spe cies as fau nal com plexes re spond ing si mul ta - neously and in the same man ner (Hofreiter and Stew art, 2009;
Stew art et al., 2010). This is well-il lus trated by the com par i son of three spe cies re viewed here: woolly mam moth, saiga an te - lope and col lared lem ming. These are all con sid ered as mem - bers of the “Mammuthus–Coleodonta” fau nal com plex (Kahlke, 1999, 2014), but their Late Pleis to cene his to ries in Eu rope dif fer sub stan tially. Dur ing the last 50 ka mam moths were pres ent in Eu rope more or less con tin u ously, with a pos - si ble range con trac tion only dur ing the LGM. The first con - firmed ap pear ance of the saiga an te lope was most prob a bly re lated with the brief warm ing GI-2 dated to ca. 23 ka BP. How - ever, only af ter the LGM was im por tant ex ten sion of its range in Eu rope ob served. Col lared lem mings most prob a bly oc - curred, as did mam moths, con tin u ously since 50 ka, but their pop u la tion turn over was much more in tense and hap pened more of ten than in the case of pre vi ous spe cies. In ter est ingly, Fig. 6. Col lared lem ming (Dicrostonyx ssp.)
A – Bayesian phy log eny of Eur asian col lared lem mings based on mtDNA cytochrome b se quences. The tree is a chronogram where branch lengths de note time elapsed since di ver gence and the po si tions of tips cor re spond to cal i brated ra dio car bon sam ple age. Colours in di cate dif fer ent mtDNA lin eages; B – spa tial dis tri bu tion of palaeontological sites with col lared lem ming in dif fer ent pe ri ods. Colours in di cate dif fer - ent mtDNA lin eages (mod i fied af ter Palkopoulou et al., 2016) and white cir cles de note sites with re mains dated by ar chae o log i cal con text
along with these ap par ent dif fer ences, some sim i lar i ties are vis i ble. It seems that most of the ob served events are grouped around three pe ri ods. The first one cov ers the GI-5 warm ing and the on set of the GS-5 cool ing around 33–31 ka BP, the sec ond one partly over laps the max i mum ex tent of the Scan di - na vian Ice Sheet dur ing the LGM at ca. 22–19 ka BP and the last one is the abrupt cli mate warm ing dur ing the Late Gla cial ca. at 15–13 ka BP (Fig. 7). In agree ment with that, the cli mate dur ing the LGM (as de fined by Mix et al., 2001) and the warm - ing dur ing the Late Gla cial have been pre vi ously rec og nized as ma jor fac tors that sub stan tially in flu enced the dis tri bu tion of many mam mal spe cies in Eu rope.
Dur ing the LGM, nu mer ous spe cies with drew from Eu rope or re stricted their ranges into so-called south ern gla cial refugia (Hew itt, 2004; Stew art and Coo per, 2008). This con cerns most of tem per ate spe cies such as red deer (Cervus elaphus;
Sommer et al., 2008; Skog et al., 2009), roe deer (Capreolus
capreolus; Sommer et al., 2009; Sommer and Zachos, 2009), brown hare (Lepus europaeus; Stamatis et al., 2009) as well as cold-adapted ones, such as mam moths and saiga an te lopes (Fig. 7). Dur ing the max i mum ex tent of the Scan di na vian Ice Sheet, en vi ron men tal con di tions in Cen tral and north west ern Eu rope were ex tremely harsh, with con tin u ous per ma frost cov - er ing most of Po land, Ger many, Bel gium, Neth er lands and north ern France (Huijzer and Vandenberghe, 1998). Ukkonen et al. (2011) sug gested that dur ing that time pro duc tiv ity of en vi - ron ments was ex tremely low, which forced the con trac tion of even cold-adapted spe cies. How ever, the harsh con di tions dur - ing the LGM do not seem to have led to many ex tinc tion events.
Coo per et al. (2015) pointed out that the grad ual cli mate de te ri - o ra tion prior to and the rel a tive sta bil ity dur ing the LGM al lowed mam ma lian pop u la tions to re treat or to adapt to the chang ing en vi ron ment. Two Eu rope-wide re place ments of Dicrostonyx pop u la tions ob served be tween 23 and 20 ka BP sug gest, how - Fig. 7. Pop u la tion turn overs and ex tinc tions in Eu rope based on ra dio car bon dat ing
A – woolly mam moth, colours in di cate dif fer ent mtDNA clades (only spec i mens with ge netic in for ma tion are shown); B – woolly mam moth from Eu rope, with out dif fer en ti a tion into the clades, source of the dates is as in Fig ure 2 sup ple mented with data from Nadachowski et al.
(2011) and Ukkonen et al. (2011); C – cave bear, blue – U. s. spelaeus, green – U. ingressus; D – saiga an te lope, colours in di cate three pu - ta tive ex pan sion waves; E – col lared lem mings, colours in di cate dif fer ent pop u la tions (EA1–EA5). MIS – Ma rine Iso tope Stages, GI – Green land Interstadials; GS – Green land Stadials; LGM – Last Gla cial Max i mum (af ter Mix et al., 2001); ver ti cal stripes in di cate tim ing of pop u la tion turn overs or ex tinc tions
ever, that some short-term en vi ron men tal changes must have taken place at that time.
The Late Gla cial pe riod was marked by sev eral cli mate os - cil la tions, with the warm BÝlling-AllerÝd Interstadial (GI-1) in ter - rupted by the Older Dryas (GI-1d) and fol lowed by the Youn ger Dryas (GS-1) cold phases (Steffensen et al., 2008). These rapid changes are linked to the ex tinc tion of the cave lion (Panthera spelaea; Stu art and Lis ter, 2011) and woolly rhi noc - eros (Coelodonta antiquitatis; Stu art and Lis ter, 2012).
The pe riod 33–31 ka BP has been sug gested only re cently from the anal y ses of col lared lem mings. Palkopoulou et al.
(2016) no ticed that the ap pear ance of one of the Dicrostonyx pop u la tions in Eu rope at the on set of GS-5, ca. 31 ka BP, co in - cides broadly with the re place ment of the en demic Eu ro pean mam moth pop u la tion by an other one from Asia and with the re - place ment of U. s. spelaeus by U. ingressus in Cen tral Eu rope (Fig. 7). Fur ther more, re cent new ra dio car bon AMS data sho - wed that the last pop u la tions of spot ted hy ena (Crocuta crocuta) also be came ex tinct ca. 31 ka BP (Stu art and Lis ter, 2014). The re cent world wide sur vey of ex tinc tion and pop u la -
tion re place ments re vealed that the tim ing of these events was sig nif i cantly cor re lated with rapid and high am pli tude cli mate changes dur ing interstadials (Coo per et al., 2015). The con cen - tra tion of such events around GI-5 fits well to this hy poth e sis as this interstadial rep re sents one of the most in stan ta neous cli - mate changes dur ing the last gla ci ation.
All four case stud ies pre sented here un veil the Late Pleis to - cene com mu ni ties as com plex and highly dy namic, char ac ter - ized by geo graphic range shifts, mi gra tions, re place ments and lo cal ex tinc tions, which seem to be com mon phe nom ena rather than ex cep tions. Coo per et al. (2015) note that such pro cesses might be es sen tial for main tain ing eco sys tem sta bil ity in pe ri ods of high cli mate vari abil ity.
Ac knowl edge ments. The work re lated to cave bears was sup ported by the Pol ish Na tional Sci ence Cen tre grant no.
2012/07/B/NZ8/02845. We are grate ful to the re view ers L. Láugas and T. van Kolfschoten for their con struc tive com - ments and to M. Molak for her sug ges tions and help with proof-read ing and cor rec tion of the manu script.
REFERENCES
Alroy, J., 2001. A multispecies over kill sim u la tion of the end-Pleis to - cene megafaunal mass ex tinc tion. Sci ence, 292: 1893–1896.
Baca, M., Stankovic, A., Stefaniak, K., Marciszak, A., Nada - chowski, A., Wêgleñski, P., Mackiewicz, P., 2012. Ge netic anal y sis of cave bear spec i mens from NiedŸwiedzia cave.
Palaeontologica Electronica, 15: 16.
Baca, M., Mackiewicz, P., Stankovic, A., Popoviæ, D., Stefaniak, K., Czarnogórska, K., Nadachowski, A., G¹siorowski, M., Hercman, H., Weglenski, P., 2014. An cient DNA and dat ing of cave bear re mains from NiedŸwiedzia Cave sug gest early ap - pear ance of Ursus ingressus in Sudetes. Qua ter nary In ter na - tional, 339–340: 217–223.
Baca, M., Popoviæ, D., Stefaniak, K., Marciszak, A., Urbanowski, M., Nadachowski, A., Mackiewicz, P., 2016. Re treat and ex - tinc tion of the Late Pleis to cene cave bear (Ursus spelaeus sensu lato). The Sci ence of Na ture, 103: 92.
Barnosky, A.D., Koch, P.L., Feranec, R.S., Wing, S.L., Shabel, A.B., 2004. As sess ing the causes of late Pleis to cene ex tinc - tions on the con ti nents. Sci ence, 306: 70–75.
Barshay-Szmidt, C., Costamagno, S., Henry-Gam bier, D., Laroula ndie, V., Pétillon J.-M., Boudadi-Maligne, M., Kuntz, D., Langlais, M., Mallye, J.-B., 2016. New ex ten sive fo cused AMS 14C dat ing of the Mid dle and Up per Mag da lenian of the west ern Aquitaine/Pyr e nean re gion of France (ca. 19–14 ka cal BP): Pro pos ing a new model for its chro no log i cal phases and for the tim ing of accupation. Qua ter nary In ter na tional, 414:
62–91.
Baryshnikov, G., 1998. Cave bears from the Paleolithic of the Greater Cau ca sus. Il li nois State Mu seum Sci en tific Pa pers, 27:
69–118.
Baryshnikov, G., Tikhonov, A., 1994. Notes on skulls of Pleis to - cene saiga of north ern Eur asia. His tor i cal Bi ol ogy, 8: 209–234.
Bocherens, H., Fizet, M., Mariotti, A., 1994. Diet, phys i ol ogy and ecol ogy of fos sil mam mals as in ferred from sta ble car bon and ni tro gen iso tope biogeochemistry: im pli ca tions for Pleis to cene bears. Palaeo ge ogra phy, Palaeoclimatology, Palaeo ec ol ogy, 107: 213–225.
Bocherens, H., Billiou, D., Patou-Mathis, M., Bonjean, D., Otte, M., Mariotti, A., 1997. Paleobiological im pli ca tions of the iso - to pic sig na tures (13C,15N) of fos sil mam mal col la gen in
Scladina Cave (Sclayn, Bel gium). Qua ter nary Re search, 48:
370–380.
Bocherens, H., Bridault, A., Drucker, D.G., Hofreiter, M., Münzel, S.C., Stiller, M., van der Plicht J., 2014. The last of its kind? Ra dio car bon, an cient DNA and sta ble iso tope ev i - dence from a late cave bear (Ursus spelaeus ROSEN - MÜLLER, 1794) from Roche dane (France). Qua ter nary In ter - na tional, 339–340: 179–188.
Cam pos, P.F., Willerslev, E., Sher, A., Or lando, L., Axelsson, E., Tikhonov, A., Aaris-SÝrensen, K., Green wood, A.D., Kahlke R.D., Kosintsev, P., Krakhmalnaya, T., Kuznetsova, T., Lemey, P., MacPhee, R., Norris, C.A., Shep herd, K., Suchard, M.A., Zazula, G.D., Shapiro, B., Gilbert, T.P., 2010a. An cient DNA anal y ses ex clude hu mans as the driv ing force be hind late Pleis to cene musk ox (Ovibos moschatus) pop u la tion dy nam ics.
Pro ceed ings of the Na tional Acad emy of Sci ences, 107:
5675–5680.
Cam pos, P.F., Kristensen, T., Or lando, L., Sher, A., Kholodova, M.V., Götherström, A., Hofreiter, M., Drucker, D.G., Kosin - tsev, P., Tikhonov, A., Baryshnikov, G.F., Willerslev, E., Gilbert, M.T.P., 2010b. An cient DNA se quences point to large loss of mi to chon drial ge netic di ver sity in the saiga an te lope (Saiga tatarica) since the Pleis to cene. Mo lec u lar Ecol ogy, 19:
4863–4875.
Chaline, J., Bru net-Lecomte, P., Montuire, S., Viriot, L., Cou rant, F., 1999. Anat omy of the arvicoline ra di a tion (Rodentia): palaeo - geo graphi cal, palaeo eco logi cal his tory and evo lu tion ary data.
Annales Zoologici Fennici, 36: 239–267.
Coo per, A., Turney, C., Hughen, K.A., Barry, W., Mc Don ald, H.G., Bradshaw, C.J.A., 2015. Abrupt warm ing events drove Late Pleis to cene Holarctic megafaunal turn over. Sci ence Ex press, 349: 1–8.
Dabney, J., Knapp, M., Glocke, I., Gansauge, M.T., Weihmann, A., Nickel, B., Valdiosera, C., García, N., Pääbo, S., Arsuaga, J.L., Meyer, M., 2013. Com plete mi to chon drial ge nome se - quence of a Mid dle Pleis to cene cave bear re con structed from ultrashort DNA frag ments. Pro ceed ings of the Na tional Acad - emy of Sci ences, 110: 15758–15763.
Debruyne, R., Chu, G., King, C.E., Bos, K., Kuch, M., Schwarz, C., Szpak, P., Gröcke, D.R., Matheus, P., Zazula, G., Guthrie, D., Froese, D., Buigues, B., de Marliave, C., Flemming, C.,
Poinar, D., Fisher, D., Southon, J., Tikhonov, A.N., MacPhee, R.D.E., Poinar, H.N., 2008. Out of Amer ica: an cient DNA ev i - dence for a new world or i gin of Late Qua ter nary woolly mam - moths. Cur rent Bi ol ogy, 18: 1320–1326.
Diedrich, C.G., 2014. Palaeopopulations of Late Pleis to cene top pred a tors in Eu rope: Ice Age spot ted hy e nas and steppe lions in bat tle and com pe ti tion about prey. Pa le on tol ogy Jour nal, 2014:
1–34.
Fedorov, V.B., 1999. Con trast ing mi to chon drial DNA di ver sity es ti - mates in two sympatric gen era of Arc tic lem mings (Dicrostonyx:
Lemmus) in di cate dif fer ent re sponses to Qua ter nary en vi ron - men tal fluc tu a tions. Pro ceed ings of the Royal So ci ety of Lon don B: Bi o log i cal Sci ences, 266: 621–626.
Fedorov, V., Goropashnaya, A., 1999. The im por tance of ice ages in di ver si fi ca tion of arc tic col lared lem mings (Dicrostonyx): ev i - dence from the mi to chon drial cytochrome b re gion. Hereditas, 130: 301–307.
Fedorov, V.B., Fredga, K., Jarrell´, G.H., 1999. Mi to chon drial DNA vari a tion and the evo lu tion ary his tory of chro mo some races of col lared lem mings (Dicrostonyx) in the Eur asian Arc tic. Jour nal of Evo lu tion ary Bi ol ogy, 12: 134–145.
Fernández-Mosquera, D., Vila-Taboada, M., Grandal-d’Anglade, A., 2001. Sta ble iso topes data (d13C, d15N) from the cave bear (Ursus spelaeus): a new ap proach to its palaeo environment and dor mancy. Pro ceed ings of the Royal So ci ety B: Bi o log i cal Sci - ences, 268: 1159–1164.
Fortes, G.G., Grandal-d’Anglade, A., Kolbe, B., Fernandes, D., Meleg, I.N., García-Vázquez, A., Pinto-Llona, A.C., Consta - ntin, S., de Torres, T.J., Ortiz, J.E., Frischauf, C., Rabeder, G., Hofreiter, M., Barlow, A., 2016. An cient DNA re veals dif fer - ences in be hav iour and sociality be tween brown bears and ex - tinct cave bears. Mo lec u lar Ecol ogy, 25: 4907–4918.
Gilbert, M., Drautz, D., 2008. Intraspecific phylo gen etic anal y sis of Si be rian woolly mam moths us ing com plete mi to chon drial genomes. Pro ceed ings of the Na tional Acad emy of Sci ences, 105: 8327–8332.
Gra ham, R.W., Lundelius, E.L., Gra ham, M.A., Schroeder, E.K., Toomey, R.S., An der son, E., Barnosky, A.D., Burns, J.A., Churcher, C.S., Grayson, D.K., Guthrie, D., Harington, C.R., Jef fer son, G.T., Mar tin, L.D., Mc Don ald, G., Morlan, R.E., Semken, Jr., H.A., Webb, S.D., Werdelin, L., Wil son, M.C., 1996. Spa tial re sponse of mam mals to Late Qua ter nary en vi - ron men tal fluc tu a tions. Sci ence, 272: 1601–1606.
Grayson, D.K., Delpech, F., 2003. Ungulates and the Mid dle-to-Up - per Paleolithic tran si tion at Grotte XVI (Dordogne, France).
Jour nal of Ar chae o log i cal Sci ence, 30: 1633–1648.
Groves, C.P., Leslie, D.E.Jr., 2011. Fam ily Bovidae (hol low -horned ru mi nants). In: Hand book of the Mam mals of the World, 2 Hoofed Mam mals (eds. D.E. Wil son and R.A. Mittermeier):
444–779. Lynx Edicions, Bar ce lona.
Hew itt, G.M., 2004. Ge netic con se quences of cli ma tic os cil la - tions in the Qua ter nary. Philo soph i cal trans ac tions of the Royal So ci ety of Lon don. Se ries B, Bi o log i cal Sci ences, 359:
183–195.
Helmens, K.F., 2014. The last inter gla cial-gla cial cy cle (MIS 5-2) re-ex am ined based on long proxy re cords from cen tral and north ern Eu rope. Qua ter nary Sci ence Re views, 86: 115–123.
Hofreiter, M., Stew art, J., 2009. Eco log i cal change, range fluc tu a - tions and pop u la tion dy nam ics dur ing the Pleis to cene. Cur rent Bi ol ogy, 19: R584–R594.
Hofreiter, M., Capelli, C., Krings, M., Waits, L., Conard, N., Münzel, S., Rabeder, G., Nagel, D., Paunovic, M., Jambresiæ, G., Meyer, S., Weiss, G., Pääbo, S., 2002. An cient DNA anal y - ses re veal high mi to chon drial DNA se quence di ver sity and par - al lel mor pho log i cal evo lu tion of Late Pleis to cene cave bears.
Mo lec u lar Bi ol ogy and Evo lu tion, 19: 1244–1250.
Hofreiter, M., Rabeder, G., Jaenicke-Després, V., Withalm, G., Nagel, D., Paunovic, M., Jambresiæ, G., Pääbo, S., 2004. Ev i - dence for re pro duc tive iso la tion be tween Cave Bear pop u la - tions. Cur rent Bi ol ogy, 14: 40–43.
Hofreiter, M., Münzel, S., Conard, N.J., Pol lack, J., Slatkin, M., Weiss, G., Pääbo, S., 2007. Sud den re place ment of cave bear mi to chon drial DNA in the late Pleis to cene. Cur rent Bi ol ogy, 17:
R122–3.
Horn, S., Prost, S., Stiller, M., Makowiecki, D., Kuznetsova, T., Benecke, N., Pucher, E., Hufthammer, A.K., Schouwenburg, C., Shapiro, B., Hofreiter, M., 2014. An cient mi to chon drial DNA and the ge netic his tory of Eur asian bea ver (Cas tor fi ber) in Eu - rope. Mo lec u lar Ecol ogy, 23: 1717–1729.
Huijzer, B., Vandenberghe, J., 1998. Cli ma tic re con struc tion of the Weichselian Pleniglacial in north west ern and cen tral Eu rope.
Jour nal of Qua ter nary Sci ence, 13: 391–417.
Jarrell, G., Fregda, K., 1993. How many kinds of lem mings? A tax o - nomic over view. In: The Bi ol ogy of Lem mings (eds. N. Stenseth and R. Ims): 45–57. Ac a demic Press, Lon don.
Kahlke, R.D., 1991. Pleis to cene dis tri bu tional and evo lu tion ary his - tory of the ge nus Saiga Gray, 1843 (Mammalia, Artiodactyla, Bovidae) in the Palaearctic. Vertebrata PalAsiatica, 10: 314–322.
Kahlke, R.D., 1999. The His tory of the Or i gin, Evo lu tion and Dis - persal of the Late Pleis to cene Mammuthus-Coelodonta Fau nal Com plex in Eur asia (Large Mam mals). Fenske Com pa nies:
Rapid City, SD.
Kahlke, R.D., 2014. The or i gin of Eur asian Mam moth fau nas (Mammuthus-Coelodonta Fau nal Com plex). Qua ter nary Sci - ence Re views, 96: 32–49.
Kahlke, R.D., 2015. The max i mum geo graphic ex ten sion of Late Pleis to cene Mammuthus primigenius (Proboscidea, Mamma - lia) and its lim it ing fac tors. Qua ter nary In ter na tional, 379:
147–154.
Kholodova, M.V., Milner-Gulland, E.J., Easton, A.J., Amgalan, L., Arylov, Iu.A., Bekenov, A., Grachev, Iu.A., Lushchekina, A.A., Ryder, O., 2006. Mi to chon drial DNA vari a tion and pop u la - tion struc ture of the crit i cally en dan gered saiga an te lope Saiga tatarica. Oryx, 40: 103–107.
Knapp, M., Rohland, N., Weinstock, J., Baryshnikov, G.F., Sher, A., Nagel, D., Rabeder, G., Pinhasi, R., Schmidt, H.A., Hofreiter, M., 2009. First DNA se quences from Asian cave bear fos sils re veal deep di ver gences and com plex phylogeographic pat terns. Mo lec u lar Ecol ogy, 18: 1225–1238.
Koch, P.L., Barnosky, A.D., 2006. Late Qua ter nary ex tinc tions:
state of the de bate. An nual Re view of Ecol ogy, Evo lu tion, and Sys tem at ics, 37: 215–250.
Kowalski, K., 1995. Lem mings (Mammalia, Rodentia) as in di ca tors of tem per a ture and hu mid ity in the Eu ro pean Qua ter nary. Acta Zoologica Cracoviensia, 38: 85–94.
Krajcarz, M., Pacher, M., Krajcarz, M.T., Laughlan, L., Rabeder, G., Sabol, M., Wojtal, P., Bocherens, H., 2016. Iso to pic vari - abil ity of cave bears (d15N, d13C) across Eu rope dur ing MIS 3.
Qua ter nary Sci ence Re view, 131: 51–72.
Kurtén, B., 1976. The cave bear story: life and death of a van ished an i mal. Co lum bia Uni ver sity Press.
Langlais, M., Laroulandie, V., Costamagno, S., Pétillon, J.-M., Mallye, J.-B., Lacrampe-CuyaubÀre, F., Boudadi-Maligne, M., Barshay-Szmidt, C., Masset, C., Pubert, E., Rendu, W., Lenoir, M., 2015. Pre miers temps du Magdalénien en Gironde Réévaluation des fouilles Trécolle ´ Saint-Germain-le-RiviÀre (France). Bul le tin de la Société Préhistorique Française, 112:
5–58.
Lis ter, A.M., Sher, A.V., 2001. The or i gin and evo lu tion of the woolly mam moth. Sci ence, 294: 1094–1097.
Lis ter, A.M., Sher, A.V., van Essen, H., Wei, G., 2005. The pat tern and pro cess of mam moth evo lu tion in Eur asia. Qua ter nary In ter - na tional, 126–128: 49–64.
Lis ter, A.M., Stu art, A.J., 2008. The im pact of cli mate change on large mam mal dis tri bu tion and ex tinc tion: ev i dence from the last gla cial/inter gla cial tran si tion. Comptes Rendus Geoscience, 340: 615–620.
Lorenzen, E.D., Nogués-Bravo, D., Or lando, L., Weinstock, J., Binladen, J., Marske, K.A., Ugan, A., Borregaard, M.K., Gilbert, M.T.P., Niel sen, R., Ho, S.Y.W., Goebel, T., Graf, K.E.,
Byers, D., Stenderup, J.T., Ras mus sen, M., Cam pos, P.F., Leon ard, J.A., Koepfli, K.P., Froese, D., Zazula, G., Stafford, T.W.Jr., Aaris-SÝrensen, K., Batra, P., Haywood, A.M., Singarayer, J.S., Valdes, P.J., Boeskorov, G., Burns, J.A., Davydov, S.P., Haile, J., Jenkins, D.L., Kosintsev, P., Kuznetsova, T., Lai, X., Mar tin, L.D., Mc Don ald, G.H., Mol, D., Meldgaard, M., Munch, K., Stephan, E., Sablin, M., Sommer, R.S., Sipko, T., Scott, E., Suchard, M.A., Tikhonov, A., Willerslev, R., Wayne, R.K., Coo per, A., Hofreiter, M., Sher, A., Shapiro, B., Rahbek, C., Willerslev, E., 2011. Spe cies-spe - cific re sponses of Late Qua ter nary megafauna to cli mate and hu mans. Na ture, 479: 359–364.
Láugas, L., Ukkonen, P., Jungner, H., 2002. Dat ing the ex tinc tion of Eu ro pean mam moths: new ev i dence from Es to nia. Qua ter - nary Sci ence Re views, 21: 1347–1354.
Mackiewicz, P., Wiszniowska, T., Olejniczak, A.J., Stefaniak, K., Socha, P., Nadachowski, A., 2010. Anal y sis of den tal enamel thick ness in bears with spe cial at ten tion to Ursus spelaeus and U. wenzensis (=minimus) in com par i son to se lected rep re sen ta - tives of mam mals. In: Mor phol ogy and Sys tem at ics of Fos sil Ver te brates (ed. D. Nowakowski): 60–77: DN Pub lisher, Wroc³aw.
MacPhee, R.D.E., Tikhonov, A.N., Mol, D., de Marliave, C., van der Plicht, H., Green wood, A.D., Flemming, C., Agenbroad, L., 2002. Ra dio car bon chro nol o gies and ex tinc tion dy nam ics of the late Qua ter nary mam ma lian megafauna of the Taimyr Pen in - sula, Rus sian Fed er a tion. Jour nal of Ar chae o log i cal Sci ence, 29: 1017–1042.
Markova, A.K., van Kolfschoten, T. eds., 2008. Evo lu tion of Eu ro - pean Eco sys tems Dur ing Pleis to cene–Ho lo cene Tran si tion 24–8 kyr BP (in Rus sian with Eng lish sum ma ry). KMK Sci en tific Press, Mos cow.
Markova, A.K., Smirnov, N.G., Kozharinov, A.V., Kazantseva, N.E., Simakova, A.N., Kitaev L.M., 1995. Late Pleis to cene dis - tri bu tion and di ver sity of mam mals in North ern Eur asia.
Paleontologia i Evolució, 28–29: 5–143.
Markova, A.K., Puzachenko, A.Y., van Kolfschoten, T., 2010. The North Eur asian mam mal as sem blages dur ing the end of MIS 3 (Brianskian–Late Karginian–Denekamp Interstadial). Qua ter - nary In ter na tional, 212: 149–158.
Markova, A.K., Puzachenko, A.Y., van Kolfschoten, T., van der Plicht, J., Ponomarev, D.V., 2013. New data on changes in the Eu ro pean dis tri bu tion of the mam moth and the woolly rhi noc - eros dur ing the sec ond half of the Late Pleis to cene and the early Ho lo cene. Qua ter nary In ter na tional, 292: 4–14.
Mar tini, I., Coltorti, M., Mazza, P.P., Rustioni, M., Sandrelli, F., 2014. The lat est Ursus spelaeus in It aly, a new con tri bu tion to the ex tinc tion chro nol ogy of the cave bear. Qua ter nary Re - search, 81: 117–124.
Maschenko, E., 2002. In di vid ual de vel op ment, bi ol ogy and evo lu - tion of the woolly mam moth. Cra nium, 19: 4–120.
Mix, A.C., Bard, E., Schnei der, R., 2001. En vi ron men tal pro cesses of the ice age: land, oceans, gla ciers (EPILOG). Qua ter nary Sci - ence Re views, 20: 627–657.
Münzel, S.C., Stiller, M., Hofreiter, M., Mittnik, A., Conard, N.J., Bocherens, H., 2011. Pleis to cene bears in the Swabian Jura (Ger many): ge netic re place ment, eco log i cal dis place ment, ex - tinc tions and sur vival. Qua ter nary In ter na tional, 245: 1–13.
Nadachowski, A., 1992. Early Pleis to cene Predicrostonyx (Roden - tia, Mammalia) from Po land. Acta Zoologica Cracoviensia, 35:
203–216.
Nadachowski, A., Lipecki, G., Stefaniak, K., Lorenc, M., Wojtal, P., 2008. Ex tinc tion of the cave bear (Ursus spelaeus) in Cen tral Eu rope (in Pol ish and Eng lish sum mary). In: Archaeozoological re search in Po land and Mid dle-East Eu rope, data – meth ods – in ter pre ta tions (eds. D. Makowiecki, R. Ab³amowicz, D. Ab³amo - wicz, K. Smiarowski and M. Makohonienko): 76–78. Bogucki Wydawnictwo Naukowe, Poznañ.
Nadachowski, A., Lipecki, G., Wojtal, P., Miêkina, B., 2011. Ra - dio car bon chro nol ogy of woolly mam moth (Mammuthus primi - genius) from Po land. Qua ter nary In ter na tional, 245: 186–192.
Nadachowski, A., Lipecki, G., Ratajczak, U., Stefaniak, K., Wojtal, P., 2016. Dis persal events of the saiga an te lope (Saiga tatarica) in Cen tral Eu rope in re sponse to the cli ma tic fluc tu a - tions in MIS 2 and the early part of MIS 1. Qua ter nary In ter na - tional, 420: 357–362.
Pacher, M., Stu art, A.J., 2009. Ex tinc tion chro nol ogy and palaeo bi - ol ogy of the cave bear (Ursus spelaeus). Boreas, 38: 189–206.
Palkopoulou, E., Dalen, L., Lis ter, A.M., Vartanyan, S., Sablin, M., Sher, A., Nystrom Edmark, V., Barnes, I., Germonpre, M., Barnes, I., Thomas, J.A., 2013. Holarctic ge netic struc ture and range dy nam ics in the woolly mam moth. Pro ceed ings of the Royal So ci ety of Lon don B: Bi o log i cal Sci ences, 280:
20131910.
Palkopoulou, E., Mallick, S., Skoglund, P., Enk, J., Rohland, N., Li, H., Omrak, A., Vartanyan, S., Poinar, H., Götherström, A., Reich, D., Dalén, L., 2015. Com plete genomes re veal sig na - tures of de mo graphic and ge netic de clines in the Woolly Mam - moth. Cur rent Bi ol ogy, 25: 1–6.
Palkopoulou, E., Baca, M., Abramson, N.I, Sablin, M., Socha, P., Nadachowski, A., Prost, S., Germonpre, M., Kosintsev, P., Smirnov, N.G., Vartanyan, S., Ponomarev, D., Nystrom, J., Nikolskiy, P., Jass, C.N., Litvinov, Y.N., Kalthoff, D.C., Grigo - riev, S., Fadeeva, T., Douka, A., Higham, T.F.G., Ersmark, E., Pitulko, V., Pav lo va, E., Stew art, J.R., Weglenski, P., Stan - kovic, A., Dalen, L., 2016. Syn chro nous ge netic turn overs across West ern Eur asia in Late Pleis to cene col lared lem mings.
Global Change Bi ol ogy, 22: 1710–1721.
Paw³owska, K., 2015. Elephantids from Pleis to cene Po land: state of knowl edge. Qua ter nary In ter na tional, 379: 89–105.
Péan, S., Puaud, S., Crépin, L., Prat, S., Quiles, A., van der Plicht, J., Valladas, H., Stu art, A.J., Drucker, D.G., Patou - -Mathis, M., Lanoë, F., Yanevich, A., 2013. The Mid dle to Up - per Paleolithic se quence of Bu ran-Kaya III (Cri mea, Ukraine):
new strati graphic, paleoenvironmental, and chro no log i cal re - sults. Ra dio car bon, 55: 1454–1469.
Ponomarev, D., Puzachenko, A., 2015. Evo lu tion of occlusal shape of the first and sec ond up per mo lars of Mid dle-Pleis to - cene col lared lem mings (Dicrostonyx, Arvicolinae, Rodentia) in north east Eu ro pean Rus sia. Boreas, 44: 741–759.
Popoviæ, D., Baca, M., Stefaniak, K., Lipecki, G., Nadachowski, A., Marciszak, A., Sabol, M., Ridush, B., Roblíèková, M., Káòa, V., Obada, T., Stankovic, A., Mackiewicz, P., 2015. Ge - netic stud ies of cave bear (Ursus ingressus) pop u la tions from Cen tral and East ern Eu rope. In: 21st In ter na tional Cave Bear Sym po sium, Leiden, Neatherlands. Pro gram and Ab stracts pp.
28, WPZ, Leiden.
Prescott, G.W., Wil liams, D.R., Balmford, A., Green, R.E., Manica, A., 2012. Quan ti ta tive global anal y sis of the role of cli - mate and peo ple in ex plain ing late Qua ter nary megafaunal ex - tinc tions. Pro ceed ings of the Na tional Acad emy of Sci ences, 109: 4527–4531.
Prost, S., Smirnov, N., Fedorov, V.B., Sommer, R.S., Stiller, M., Nagel, D., Knapp, M., Hofreiter, M., 2010. In flu ence of cli mate warm ing on arc tic mam mals? New in sights from an cient DNA stud ies of the col lared lem ming Dicrostonyx torquatus. PLoS One, 5: e10447.
Rabeder, G., Hofreiter, M., 2004. Der neue Stammbaum der alpinen Höhlenbären. Die Höhle, 55: 1–19.
Rabeder, G., Hofreiter, M., Nagel, D., Withalm, G., 2004a. New taxa of Al pine cave bears (Ursidae, Carnivora). Ca hiers Scienti - fiques-Muséum d’Histoire Naturelle de Lyon, HS 2: 49–67.
Rabeder, G., Hofreiter, M., Withalm, G., 2004b. The sys tem atic po - si tion of the Cave Bear from Potoèka zijalka (Slovenia).
Mitteilungen der Kommission für Quartärforschung der Österre i - chi schen Akademie der Wissenschaften, 13: 197–200.
