Geo log i cal Quar terly, 2005, 49 (2): 185–194

Lateglacial-Mid dle Holo ce ne stable iso tope recor ds in two coeval stalag mite s from the Bihor Mountains, NW Romania

Tu dor TÃMAª, Bogdan P. ONAC and Ana-Voica BOJAR

Tãmaº T., Onac B. P. and Bojar A.-V. (2005) — Lateglacial-Mid dle Ho lo cene sta ble iso tope re cords in two co eval sta lag mites from the Bihor Mountains , NW Rom ania. Geol. Quart., 49 (2): 185–194. Warszawa.

Oxy gen and carbon stable iso tope records of two sta lagm ites from NW Ro mania pro vide a contin u ous cli matic record be tween 14.8 and 5.6 ka BP. The chronol ogy is estab lishe d by 21 TIMS uranium serie s ages. Uncer tainties in the isotope chro nology range be tween ±20 18 13 and ±100 yr. The d O val ues are pos i tively cor re lated with tem per a ture, whereas d C fluctu a tions suggest changes in soil CO2 produc - tion. Lateglacial depo si tion of both stalag m ites started at ~14.8 ka BP. The d18O re cords subse quently show a slow decline in tem pera - tures until 12.6 ka BP. Three warmer pe riods with incre ased soil produc tiv ity occurr ed at 14.5–13.9 ka BP, 13.6–13.2 ka BP, and 12.9–12.6 ka BP. Lower d18O and high d13C values betwee n 12.6 and 11.4 (11.7) ka BP indi cate a cold and dry clim ate dur ing the Youn - ger Dryas (GS-1). During the Early Ho lo cene, three short cold inter vals are marked on the d18O profiles at 11.0–10.6, 10.5–10.2 and 9.4–9.1 ka BP. For the rem ainder of the Holo cene sequence , the d18O records show less vari a tion be tween 9 and 7.8 ka BP and gradual warm ing from 7.6–5.6 ka BP. The speleothem records corre late with the Greenland ice core re cords and with other proxies throughout Europe and the North Atlan tic region.

Tu dor Tãmaº and Bogdan P. Onac, Quater nary Re search Group, “Babeº-Bolyai” Uni ver sity, Kogãlniceanu 1 and “Emil Racoviþã” In - sti tute of Spe le ol ogy, Clinicilor 5, 400006 Cluj-Napoca, Ro ma nia, e-mails: [email protected], [email protected]; Ana-Voica Bojar, Insti tute of Earth Sci ences, Karl-Franzens Uni ver sity, Heinrichstrasse 26, A-8010 Graz, Austria, e-mail: [email protected] (receiv ed: Decem ber 28, 2004; accepte d: April 15, 2005).

Key words: NW Ro ma nia, Lateglacial-Mid dle Ho lo cene, sta lag mites, sta ble iso topes, ura nium-tho rium TIMS dating.

INTRODUCTION tope profil es on two sta lagm ites from northwest ern Rom ania. The d18O and d13C speleothem re cords show sim i lar i ties with the Green land ice core re cords (Dansgaard et al., 1989; Calcium carbon ate de pos its from caves, such as sta lag mites Johnsen et al., 1997) and are corre lat ed with pol len records, or flowstones, provide infor m ati on on the na ture of past cli - lake sed i ments and veg e ta tion dy nam ics (Wohlfarth et al., mates and en vi ron ments. Changes of cli ma tic con di tions were 2001; Björkman et al., 2002; Bodnariuc et al., 2002; Feurdean shown to in flu ence sta ble iso tope geo chem is try dur ing and Bennike, 2004; Magny, 2004). speleothem de po si tion (Hendy and Wil son, 1968; McDermott, 2004). The timing of past clima tic shifts in such con tinen tal re - cords is im portant for com pari sons with marine or ice core re- SITE DESCRIPTION cords and for reconstructions of climate variability. U-Th TIMS dating of speleothems provides good preci sion of result s in terms of radio m etric ages and sam ple stra tigra phy The V11 Cave is locat ed in the southeas tern part of (Li et al., 1989; Richards and Dorale, 2003). Speleothems Vãrãºoaia Glade (Bihor Mountai ns, NW Rom ania) , at 1254 m dated by TIMS provide valuable palaeoclimatic in for ma tion a.s.l. (Fig. 1A). The area is a typ ical karst pla teau with sink - when as soci ate d with other proxies such as d18O and d13C iso - holes and ponors and the vege ta ti on consis ts mainly of spruce to pic com po si tion of cal cite (Dorale et al., 1998; McDermott et stands and alpine herbs (Bodnariuc et al., 2002). The mean an - al., 1999; McDermott, 2004; Wil liams et al., 2005). nual tem per a ture is 5°C and the mean annual precip i ta ti on ex- In this study we present a recon struc tion of Lateglacial- ceeds 1200 mm at the Stâna de Vale winter resort (1102 m Mid-Ho lo cene cli mate changes based on con tin u ous sta ble iso- a.s.l), 11 km NW from the cave. The pres ent-day cli mate of the 186 Tudor Tãmaº, Bogdan P. Onac and Ana-Voica Bojar

Fig. 1. A — geograph ical lo cation of the V11 Cave; karst areas are marked in black, the gray circle marks the Bihor Moun tains; B — plan and lon gi tudi nal sec tion of the V11 cave show ing the sampling site of sta lag mites S22 and S117 area is pre dom i nantly in flu enced by west-north west oce anic air Sam ple S117 (Fig. 2B) is 48 cm in height (measured along masses, with the South ern Carpathians act ing as a bar rier the growth axes) and is very differ ent from S22 in terms of against Med i ter ra nean in flu ences. shape and stra tig ra phy. The re peated dis place ment of its The V11 cave formed in Anisian car bon ate rocks: black growth axis was proba bly produced by sliding over the clayey lime stones (Guttenstein fa cies) and grey dolostones. The cave sub stra tum. On the ba sis of mac ro scopic ob ser va tions, the sam - has 1166 m of survey ed pas sages, a verti cal exten sion of 67 m ple could be di vided into 4 zones with 10 differ ent growth axes (–37 m; +30 m), and it devel oped as a dendrit ic maze on four (Tãmaº and Causse, 2001): main karstification lev els (Damm, 1993; Fig. 1B). The I. 0–9 cm: from the base to a frac ture line covered by subse - speleothems are re stricted to a few sec tors and con sist of cal cite quent de posi ti on (growth axes A10–A8); thin growth levels , sta lag mites, sta lac tites, old (fos sil) flowstones and some ar- slightly porous dark brown calcit e; agon ite crystal ag gre gates and crusts. Pres ent-day cal cite de - II. 9–29.5 cm (A7–A5); light brown to white com pact cal - pos its oc cur as soda straws and small active stalag mites. cite lamina, with a small growth hia tus at 25 cm, marked by a The two sta lagm ites (S22 and S117), both inac ti ve when thin film of clay. This sequence ended when the stalag m ite sampled, were col lected 200 m away from the cave en trance, at broke (not shown in Fig. 2B); about 60 m below the surface (Fig. 1B). The first one grew on a III. 29.6–36.4 cm (A4–A2); de posi ti on resum ed and youn - breakdown block and the sec ond on clay de posit ed on the pas- ger growth layers cov ered the frac ture sep arat ing zones I and II; sage floor. Sev eral ac tive soda straws from their close vi cin ity yel low ish white, trans par ent, al ter nat ing com pact and po rous were sampled for re cent cal cite and drip wa ter iso to pic anal y- cal cite lev els; ses. The mean an nual temper ature and the rel ative hu mid ity of IV. 32.5–48 cm (A1). The sta lagm ite slipped once more the cave air, measured be tween 2003 and 2004, are 6.5°C and over the substra tum and the younger growth lay ers partly cov - 97–98%, re spec tively. ered zone III, induc ing a hi atus of ~0.6 ka.

U-Th DATING SAMPLE DESCRIPTION AND ANALYTICAL PROCEDURES The U and Th rati os were measured in peak jumping mode on a Finnigan-MAT 262 ther mal ion iza tion mass spec trom e ter SAMPLE DESCRIPTION at the Laboratoire des Sci ences du Climat et de l’Environ- nement, Gif sur Yvette, France. De tails of the re search and Sample S22 is a 34 cm-tall sta lag mite con sist ing mostly of chem i cal pro ce dure are given by Tãmaº and Causse (2001). A dense, yellow-brown, low-Mg calcit e laminae, com posed of total of 21 sam ples (12 from S22 and 9 from S117) from the up- large prism atic crystal s, conti nu ously grown from one level to perm ost parts of the two sta lagm ites (15 to 5.6 ka) were se - another and ori ented oblique to the outer surface (Fig. 2A). The lected for this study (Ta ble 1). About 1 to 2.5 g of calcit e was first 8 cm-thick in ter val from the base con sists of al ter nat ing ex tracted from ca. 3 mm thick lev els for each anal ysis. The ura - thin transpar ent and opaque levels , which ends with a nium content ranged be tween 0.23 and 0.68 ppm. The U-Th well-marked corrosional hia tus, after which growth was con- ages cover the time inter vals of 14.5–5.6 ka BP for S22 and tinu ous over the last 26 cm to the top. Betwe en 10.5 and 5.8 cm 13.5–6.1 ka for S117 (Ta ble 1 and Fig. 2A, B). A basal age of from the top, there is a highly porous white opaque level made 14.8 ka BP for the begin ning of the Lateglacial was obtai ned on up of cal cite microcrystals. S117 by lin ear ex trap o la tion. Lateglacial-Mid-Holocene stable isotope records in two coeval stalagmites from the Bihor Mountains, NW Romania 187

Fig. 2. Sampling pro file along the two sta lag mites

A — S22; B — S117; A1–A4 on S117 show direc tions of growth axes; white ellypses show the posi tion of sam ples taken for U-Th dating; white dots along the growth levels indi cate posi tions of sam ples checked for ‘Hendy’ tests (D, E on S22; A4-1, A1-4 on S117); scale bars are 2 cm

STABLE ISOTOPE ANALYSES 52±3 yr. BP (Tãmaº, 2003), were analysed as well. Drip wa ter from two soda straws was collec ted from July 2003 to June Ox y gen and car bon sta ble iso tope anal y ses were per formed 2004. However, the high amount of snow dur ing the winter of on the upper part of the sta lagm ites S22 and S117 (the upper 2003 made the access to the cave im pos sible from De cem ber 25 cm for S22 and the upper 13 cm for S117), corre spond ing to 2003 to May 2004 so the water sam ple collec ted in May repre - the inter val be tween 15 and 5.6 ka BP. A total of 306 sam ples sents an av er age of these five months. Drip rates are much (147 from S22, and 159 from S117) were drilled out at 2 mm lower during the winter months. The isoto pic com posi ti on of inter vals for S22 and 1 mm inter vals for S117, along the centra l the dis solved in or ganic car bon (DIC) as well as the iso to pic axes of the sta lagm ites, us ing 0.5 and 0.3 mm diam eter drills. com posi ti on of oxy gen in water were done at Joanneum Re - These inter vals were consid ered conve nient , as the length and search Graz, De part ment of Hydrogeology, us ing au to matic aver age growth rates of S117 gener all y repre sent half the val- prep a ra tion lines and a Finnigan MAT Delta Plus mass spec- ues re corded for S22. The ages were as signed to each sta ble trome ter. The iso to pic val ues of wa ter samples are given in ‰ 1 iso tope sam pling point by lin ear in ter po la tion be tween neigh- rel a tive to SMOW for d18O and to PDB for d 3C. bour ing TIMS dates. Two tests for iso to pic equi lib rium were done on “indi vid ual growth layers” for each sta lagm ite (Fig. 3). A mini m um of 7 samples at 5 mm apart was analysed for each RESULTS growth level. The isoto pic analy ses were perform ed on whole-rock sam - U-Th DATING ples using an auto m atic Kiel II prepa ra ti on line and a Finnigan MAT Delta Plus mass spec trome ter at the In stitute of Earth Sci - The ages and isotope rati os obtai ned by U-Th dat ing are pre- ences, Karl-Franzens Uni ver sity (Aus tria). Car bon di ox ide was sented in Ta ble 1. In S22, an age of 14.5 ka at 8.4 cm from the relea sed by reac ti on with phospho ric acid at 70°C. The isoto pic base follows a level dated at 46.5 ka, from which it is sepa rat ed val ues are pre sented in per mil rel ative to PDB. NBS-19 and an by a hia tus (Fig. 2A). Consid er ing its shape, we assum ed this hi- in ter nal lab o ra tory stan dard were ana lysed con tin u ously for ac- a tus was corrosional, formed by undersaturated drip water that curacy control , and the standard devi a ti on was 0.1‰ for d18O dissolve d a part of the cal cite form erly de posit ed (Tãmaº and and 0.06‰ for d13C. Causse, 2001). In S117, an age of 23.4 ka was obtai ned for the Two pres ent-day cal cite sam ples from active soda straw level be low the hia tus (Fig. 2B) (Tãmaº and Causse, 2001). The tips and two from a stalag mite top from the same cave, dated at age obtai ned for the level just above the hi atus in S117 (Fig. 2B) 188 Tudor Tãmaº, Bogdan P. Onac and Ana-Voica Bojar

Ta ble 1

U-Th ages and iso tope ra tios ob tained on the two sta lag mites (details in text and in Tãmaº and Causse, 2001)

S22

d* [cm] U [ppm] ± 2s 234U/238Ua ± 2s 234U/238Ui ± 2s 230Th/234U ± 2s 230Th/232Th Age [ka BP] ± 2s [ka] 33.7 0.407 0.993±0.021 0.992±0.018 0.050±0.001 27.5 5.570 0.118 30.2 0.446 0.970±0.003 0.969±0.003 0.068±0.001 189 7.628 0.106 29.4 0.417 0.981±0.003 0.980±0.002 0.069±0.001 505 7.810 0.039 28.5 0.371 1.001±0.002 1.000±0.002 0.077±0.001 282 8.724 0.041 23.0 0.450 0.927±0.003 0.925±0.002 0.082±0.001 851 9.349 0.056 21.8 0.440 0.927±0.003 0.925±0.002 0.083±0.002 371 9.480 0.157 20.8 0.411 0.929±0.005 0.927±0.004 0.089±0.001 283 10.118 0.111 19.8 0.344 0.930±0.002 0.928±0.002 0.092±0.001 779 10.536 0.141 18.9 0.403 0.921±0.002 0.918±0.002 0.091±0.001 775 10.339 0.074 15.9 0.493 0.921±0.003 0.918±0.002 0.096±0.001 193 10.983 0.044 12.0 0.549 0.905±0.002 0.901±0.002 0.102±0.001 263 11.753 0.046 10.0 0.520 0.919±0.003 0.916±0.002 0.110±0.001 137 12.747 0.035 8.4 0.609 0.885±0.002 0.880±0.001 0.125±0.001 100 14.526 0.090

S117 47.8 0.283 0.953±0.003 0.952±0.002 0.055±0.001 86 6.141 0.038 47.1 0.447 0.923±0.008 0.922±0.006 0.068±0.001 311 7.699 0.067 44.5 0.446 0.904±0.002 0.901±0.002 0.082±0.001 429 9.268 0.069 41.4 0.333 0.930±0.003 0.928±0.003 0.091±0.001 838 10.347 0.070 40.8 0.345 0.881±0.002 0.877±0.002 0.095±0.001 490 10.909 0.048 39.3 0.378 0.870±0.002 0.866±0.002 0.098±0.001 943 11.223 0.059 38.0 0.655 0.882±0.002 0.878±0.002 0.105±0.001 1376 12.092 0.129 35.9 0.519 0.863±0.003 0.858±0.002 0.114±0.001 874 13.198 0.135 35.5 0.467 0.867±0.002 0.862±0.002 0.117±0.001 650 13.525 0.097 34.6 0.513 0.819±0.002 0.810±0.001 0.1365±0.001 29 16045** 0.089

*d — distance from base; ** — re jected due to de tri tal contam ina tion

was 16 ka, but it was reject ed due to its low 230Th/232Th ra tio (Ta- The mini m um growth rates are re lated to the warm ing ble 1). By extra po la ti on from the neighbour ing age of 13.5 ka to phase at the begin ning of the Lateglacial (14.5–12.7 ka in S22) the hi atus, an age of 14.8 ka was obtai ned for the basal level of and during the fi nal growth inter val, from ca. 7.7 ka unti l this growth inter val. Two other dates were obtai ned with low growth cessa ti on (17.2 mm/ka for S22 and 7.1 mm/ka for 230Th/232Th ra tios, lo cated at the top of the two sta lag mites (Ta- S117). Apart from that, S117 has fairly consta nt growth rates, ble 1). Correc ti ons for detri tal content were not appli ed in this except for a maxi m um of 82.8 mm/ky at the begin ning of the case be cause the cor rected ages would still fall within the un cer - Ho lo cene (ca. 200 years) fol lowed by a hia tus of ca. 560 years, tainty limits of the iso to pic re cords. due to a change in the growth axis posi ti on. S22 has a more The aver age growth rates of the two sta lagm ites for the compli cated growth re cord, with faster rates within the GS-1 whole Lateglacial-Mid Ho lo cene growth in ter val are (Younger Dryas) and during the first part of the Boreal 28.2 mm/ka for S22 and 14.3 mm/ka for S117. The growth (46.6–90.5 mm/ka) and around the Boreal -Atlan ti c transi ti on rates be tween any two suc ces sive U-Th ages vary be tween 11.2 (88.0–91.6 mm/ka), sepa rat ed by slow-growth inter vals and 91.6 mm/ka for S22, and be tween 7.0 and 82.8 mm/ka for (15.7 mm/ka, betwe en 10.1 and 9.5 ka, and 11.0 mm/ka, be- S117. Tem poral reso lu ti ons for the isotope records are thus be- tween 8.7 and 7.8 ka). A short, fast growth inter val of less than tween 20 and 190 yr/mm for S22 (for isotope sam pling points 200 years is recorded be tween 7.8 and 7.6 ka. This latter inter - at 2 mm) and 12 and 142 yr/mm for S117. These give a maxi - val and the one betwe en 9.3–8.7 ka are also marked by changes mum shift of ±100 yr in the stable isotope chronol ogy. in calcit e petrog ra phy, white porous zones made up by cal cite Lateglacial-Mid-Holocene stable isotope records in two coeval stalagmites from the Bihor Mountains, NW Romania 189

Fig. 3. d18O and d13C profil es along growth layers of S22 and S117 microcrystals re plac ing the com pact palisadic fab ric. Such fab - a tions are in ter preted to re flect dif fer ent sources of DIC be tween rics can also be observe d during the BÝling-AllerÝd in S117, the winter -sum mer months, with a higher con tri bu tion of bed- but they are not asso ci ate d with changes in growth rates. rock and atm ospheri c carbon in the cold sea son. Using the frac - tion ation fac tors be tween HCO3–CO2 gas (Mook et al., 1974) d13 STABLE ISOTOPE ANALYSES and CO2 gas–cal cite (Bottinga, 1968), the cal cu lated mean C of cal cite is –7.5‰, only with 0.2‰ lower than the measured Speleothems are pre cip i tated in iso to pic equi lib rium with car bon iso to pic com po si tion of the pres ent-day cal cite their drip wa ter when the CO2 from wa ter degasses slowly, at (–7.3‰). The measured and the calcu lat ed val ues support the sta ble cave tem per a tures with out air cur rents. Two con di tions view that pre cip ita tion of the cal cite at the samplin g site is near have to be tested to es tab lish the iso to pic equi lib rium: first, that iso to pic equi lib rium with the drip wa ter. the d18O val ues of a single growth layer show no signif i cant For speleothems pre cip i tated in iso to pic equi lib rium with the vari a tion (Hendy, 1971), and second, that the d18O and d13C drip water, the d18O iso to pic com po si tion is de pend ent on com - values measured along the growth axis are not corre lat ed. The po si tion of pre cip i ta tion, and the tem per a ture-de pend ent frac- tests for isoto pic equili brium done on the two stalag m ites show tionati on betwe en drip water and carbon ate . The aver age d18O lit tle or no cor re la tion be tween the ox y gen and carbon records. The two lay ers tested from S117 show d18O max i mum vari a tions of 0.26‰, which is close to twice the an a lytic er ror; in the case of S22, the maxi m um vari ati on is of 0.53‰ (Fig. 3). The ‘Hendy’ tests sug gest that ki netic ef fects were not sig nif i cant dur ing de po si tion. Also, no cor re la tion ex ists be tween d18O and d13C val ues along the growth axes of the two sta- lag mites (Fig. 4). There is, however, a good cor- re la tion be tween the two d18O records (Fig. 6 B, C), proving they are suit able for palaeoclimatic in ter pre ta tions (Dorale et al., 1998, 2002). The sea sonal d18O vari ati ons in drip water samples col lected from the cave are up to 1.4‰ (Fig. 5A), with slightly lower val ues during the sum mer months. The am pli tude of sea sonal tem - pera ture vari ati on inside the cave does not ex - ceed 1°C. Us ing the equa tion of Friedmann and O`Neil (1977), the calcu lated d18O val ues of modern calcit e have a mean of 22.5‰, which is ~0.8‰ lower than the isoto pic com posi ti on of present- day cal cite (23.3‰). Dis solved inor - ganic carbon (DIC) val ues for the col lected drip wa ter show pro nounced sea sonal vari a tion (Fig. Fig. 4. d18O vs. d13C values of sam ples taken along the growth axes of the stalag m ites 5B), with higher values during winter . The vari- R2 — squared cor re la tion co ef fi cient 190 Tudor Tãmaº, Bogdan P. Onac and Ana-Voica Bojar

the pres ent day cal cite pre cip i tates in equi lib rium with the DIC. Using the frac tionati on factor of Mook et al. 13 (1974) and the measured d CDIC for May, when the veg e ta tion starts to grow, the cal cu lated com po si tion of 13 soil d CCO2 is ~ –22‰. The value is close to the one mea sured by Rightmire (1978), and in the range of soil 13 d CCO2 in C3 dom inate d landsca pes (Bender, 1968). 13 Sim i larly, us ing the mea sured d CDIC of ~ –7‰ for No- 13 vember, the calcu lated value of d CCO2 is higher (~ –17‰) showing a larger parti ci pa ti on of at mospheri c 13 CO2 with d C ~ –7‰. Accord ingl y, colder/ dryer cli- mate condi ti ons in the past would im ply increa sed d13C values, due to a higher parti ci pa ti on of at mospheri c CO2. During warmer peri ods, the higher propor ti on of soil CO2 from the seepage wa ter would induce lower d13C val ues. Figures 6 B and C present the d18O and d13C pro files of the two stalag m ites. The general patter ns of both ox- ygen and carbon isoto pic profil es show higher am pli - tudes for the Lateglacial and Early Holo cene . d18O val- ues range betwe en –7.0‰ and –9.3‰ dur ing the Lateglacial–Early Holo cene (14.8–10.2 ka BP) and be - tween –7.5‰ and –8.5‰ for the rest of the Holo cene record. Higher aver age values, of –7 to –8‰, occur be- tween 14.7 and 13.2 ka, at 11.2 ka and be tween 8 and 5.6 ka BP. The mini m um d18O val ues oc cur between 13.2 and 12.8 ka and betwe en 12.5 and 10.2 ka. The d13C val ues vary be tween –4.1‰ and –9.2‰. The high- Fig. 5. Monthly vari ation of d18O and dis solved in or ganic car bon (DIC) 18 13 est values (–6 to –4.1‰) occur betwe en 12.6 and 11.7 val ues for drip wa ter plotted with the cal culated d O and d C iso to pic com po si tion of cal cite (con tin u ous line) and the mea sured d18O, and d13C ka and be tween 11.2 and 10.9 ka. Rela ti vely lower val - of pres ent-day cal cite (pdc) ues, of –6.5 to –9‰ charac ter ize the inter val 14.7–12.7 ka BP. Holo cene d13C varia ti ons are betwe en –6.5 and value of the drip wa ter collec ted in the cave during 2003–2004 is –8.7‰, with lower am pli tudes after 8.8 ka BP. –10.6‰, close to the –10.5 ‰ value measured for pre cipi ta ti on at the Urºilor Cave, 12 km SW from our site, during 1997–1998 (six months miss ing). In Europe, the d18O com po si tion of rain - LOCAL AND REGIONAL PALAEOCLIMATE wa ter shows a pos i tive cor re la tion with tem per a ture, with a co ef - CORRELATIONS fi cient of ~0.6‰°C–1 (Rozanski et al., 1992). Given that the speleothems studied were de posit ed close to isoto pic equili brium with the drip wa ter, a pos itive cor re lation is as sessed between the Com paring the isotope profil es from the V11 sta lagm ites d18O re cord and tem per a ture vari a tions. with the d18O records of the Greenland ice cores, a good overall The carbon from speleothems has three main sources: at - corre la ti on is observe d for the studied sequence (Fig. 6). The mo spheric CO2, soil CO2 and carbon from limestone s differ ence s in chronol ogy, in the or der of 50–100 years, are (Schwarcz, 1986; Genty et al., 2001). The d13C val ues in linked to the reduced growth rates and conse quentl y, to the speleothems may be in flu enced by sev eral fac tors as: type of lower tem po ral res o lu tions of the sta lag mites. pho to syn the sis (C3 or C4), soil biogenic ac tiv ity, car bon ate dis- After Term ina ti on I, the warm ing phase from 14.8 to 14.4 ka so lu tion and pre cip i ta tion, and drip rate (Schwarcz, 1986; is re corded in the S117 oxy gen isoto pic com posi ti on as a posi - Lauritzen and Lundberg, 1999). At the V11 site the vege ta ti on tive shift of ~1.2‰. The age esti mated for the begin ning of de po - consis ted only of C3 plants at least since 15 ka BP (Bodnariuc et siti on in S117 at 14.8 ka BP corre sponds with the age of 14.7 ka al., 2002). For the tem perate regions, soil CO2 rep re sents the BP de ter mined for the postglacial warm ing in the Greenland ice most im portant source for car bon in speleothems (Genty et al., cores (Björck et al., 1998) and with the high-reso lu ti on record of 2001). The present day d13C from V11 cave drip wa ter shows Wohlfarth et al. (2001) from Preluca Þiganului, NW Rom ania, heavier val ues during winter , when biogenic acti vit y is low, who obtai ned an age of 14.7–14.6 cali brat ed ka BP for the begin - and lower val ues in sum mer, when more soil CO2 is produced. ning of lake sedi m enta ti on. For this stage, the study of Wohlfarth

Rightmire (1978) studied seasonal vari ati ons of soil PCO2 and et al. (2001), as well as our isotope record, sus tain the iso to pic com po si tion, show ing that dur ing win ter, the soil synchronicity be tween the cli matic tran si tions in the North At - 13 13 d CCO2 is close to the at mospheri c val ues. In contra st, during lanti c and East ern Europe. The d C re cord of S117 shows a pro- 13 the sum mer period, soil d CCO2 decrea ses to values of –21.5‰, gressive decrea se (–6.1 to –7.4‰) betwe en 14.8–14.2 ka, indi - par allel to an in crease in PCO2 . In this study we have shown that cat ing the devel op m ent of soil and vege tal cover. Lateglacial-Mid-Holocene stable isotope records in two coeval stalagmites from the Bihor Mountains, NW Romania 191

From 14.4 ka on, the decrea s- ing trend in the d18O record of S117 in di cates grad ual cli mate cooling. A sim ilar trend is ob- served for S22 from 13.6 ka. While the S22 profil e has too low res o lu tion to reg is ter shorter-term cli ma tic os cil la tions be tween 14.5 and 12.7 ka, several maxim a on the S117 oxy gen profil e are inter - preted as warmer pe ri ods with higher soil produc ti vit y. These in- ter vals, dated at 14.5–13.9 ka BP, 13.6–13.2 ka BP, and 12.9–12.6 ka BP, corre spond ing to minim a on the d13C re cord, are as signed to BÝlling (GI-1e in the ice core event stra tig ra phy of Björck et al., 1998), and to GI-1c–GI-1a (AllerÝd). In NW Ro mania, simi - lar cli mate con di tions for this first part of the Lateglacial were deter - mined by Wohlfarth et al. (2001) and Feurdean and Bennike (2004) from pollen, macrofossil and lake sed i ment stud ies. A cold/wet event, recorded be tween 13.2 and 12.8 ka BP in the speleothem d18O pro files is cor re lated with an intra-AllerÝd cold pe riod or GI-1b and has a du ra tion com pa ra ble with the one in the GRIP record. The oxy gen profil es of the two stalag m ites show a drop of 0.6–1‰ at 12.6 ka BP, inter - preted as a change to colder climate. This in ter val, with a du - rati on of ~1.2 ka, approx i m ates well the Younger Dryas, or GS-1 in the GRIP stra tig ra phy. The mean d18O value for the GS-1 is –8.7‰ PDB at this site, whereas the measured pres ent-day cal cite has a d18O value of 23.1‰ (SMOW) or –7.3‰ (PDB), showing a ~1.4‰ enric hm ent with respec t to GS-1. The rapid 18 in crease of d13C to values of –4.5 Fig. 6. A — d O ice record from GRIP at a 20-year res o lu tion (orig i nal data from the Greenland Summ it Ice Cores CD-ROM, 1997); B and C— d18O and d13C profil es along the growth axes to –5‰ at 12.7–12.6 ka BP of the two stalag m ites; D — thick ness-age model for the two stalag m ites points to a dras tic de crease of soil lines — S22; dashed lines — S117; dashed grey lines in di cate corre la tive events in GRIP and the sta lag mite re- ac tiv ity, as a con se quence of cords; O.D. — Older Dryas lower tem per a tures and pre cip i - ta tion. It is in ter est ing that from these val ues, us ing the frac tionati on factors of Mook et al. Rom ania show reduc ti ons of the forest cover (Fãrcaº et al., 13 (1974) and Bottinga (1968), the cal culat ed d CDIC value of ~ 1999; Tanþãu, 2003) and gradual replac em ent of forest by herb –8‰ is simi lar to the DIC value measured in Novem ber for and shrub-dom i nated veg e ta tion co ex ist ing with bar ren ar eas drip wa ter in the cave. This would suggest that the d13C com po- (Björkman et al., 2002, 2003; Feurdean and Bennike, 2004). d18 d13 si tion of soil CO2 during GS-1 was sim ilar to the one of modern A somewhat differ ent trend of the O and C re cords is no- win ters. Pol len and palaeovegetation data from other sites in ticed be tween 11.7 and 11.4 ka BP. The carbon values de crease 192 Tudor Tãmaº, Bogdan P. Onac and Ana-Voica Bojar rapidl y at 11.7–11.6 ka BP, suggest ing increa sing wetness and soil Fraxinus) around 9.3 ka BP, along with a re-expan sion of pro duc tiv ity, whereas superposed d18O shifts towards heavier val - Corylus (Björkman et al., 2003). We may only infer that the ues (warmer clim ate) on both sta lag mites are re corded some 200 higher growth rates during this pe riod were driven by local years later. At sites from NW Ro mania, the re place ment of open causes. As for most of the Ho lo cene, no clear cli matic signal can veg e ta tion by an open Betula, Pinus and Picea-dom i nated for est be in ferred from the sta lag mite d13C re cord for this pe riod, but around 11.5 ka BP indi cat e warm and moist cli mati c condi ti ons at other studies in NW Rom ania indi cat e drier condi ti ons at least the be ginning of the Holo cene (Björkman et al., 2002; Bodnariuc unti l 8.2 ka (Tanþãu et al., 2003; Feurdean and Bennike, 2004). et al., 2002). Our data for the onset of GS-1 and for the GS-1/Ho- Betwe en 9 and 7.8 ka, fairly con stant d18O val ues (–7.5 to locene transi ti on is also consis tent with other proxies from west - –8‰) in di cate rel a tively sta ble av er age tem per a tures and the ern/cen tral Eu rope (Von Grafenstein et al., 1999; Walker et al., low reso lu ti on of the isoto pic record does not al low a more de- 1999; Ammann et al., 2000). tailed cli mate re con struc tion. Speleothem depo si ti on in the V11 cave was conti nu ous The in ter val with de creased d18O val ues (–8.5‰) be tween during the whole GS-1 inter val, a less com mon sit ua ti on within 7.8 and 7.6 ka BP in S22 is asso ci ate d with a porous depo si ti on tem per ate karst re gions (Denniston et al., 2001; Niggemann et fa cies, where the ox ygen and car bon re cords are covariant, and al., 2003; Wurth et al., 2004). Although this cold inter val had a this prob a bly rep re sents an other short ep i sode of de po si tion in ma jor in flu ence on the veg e ta tion, our dataset may point to a iso to pic dis equi lib rium. The S117 iso to pic re cords, which are less pronounced cooling than in Western Europe (Walker, more re li able, sug gest that no sig nif i cant tem per a ture change 1995; Renssen et al., 2000; Shotyk et al., 2002). occurred dur ing this pe riod. Another cooling is marked by lower d18O during the After 7.8 ka BP, the oxy gen profil es of S22 and S117 show Preboreal, with two short-term inter vals (11.0–10.7 and a slight increa sing trend to the level of the modern value. This 10.5–10.2 ka) when d18O values reached –9‰, close to the final growth inter val of the two speleothems cor re sponds with ones re corded throughout GS-1. Both these inter vals corre lat e an inter val of reduced growth frequency for the Rom anian with lower d18O in the Green land ice core re cords (Dansgaard speleothems (Onac and Lauritzen, 1996) and with a Mid-Euro - et al., 1989; Johnsen et al., 1997; Björck et al., 2001). The lat ter pean lower lake level phase (Magny, 2004). A drier period in in ter val is as so ci ated with the first re-ex pan sion of Corylus, an the region of the V11 cave might be indi cat ed by the pol len data indi ca tor of cooler and drier clim ate around 10.3 ka in several and by the strong de crease of sedi m enta ti on rates at two peat Ro ma nian pol len re cords (Fãrcaº et al., 1999; Rösch and sites in the Bihor Mountai ns (Bodnariuc et al., 2002). Fischer, 2000; Fãrcaº, 2001; Bodnariuc et al., 2002; Tanþãu et Betwe en 7.8 and 5.7 ka, during the period known as the al., 2003; Feurdean, 2005)) and in other proxies around the “Ho lo cene cli ma tic op ti mum” (Bell and Walker, 1992; North At lan tic re gion (Nesje et al., 2001; Seppä et al., 2002). Huntley and Prentice, 1993), the cli mate of northeas tern Eu- The be ginning of the Boreal is marked by an as cending rope was char acter ized by temper atures and amounts of pre cip - trend of the d18O profil es (10.2–9.5 ka BP) and by a pro- ita ti on higher than present ones. With the uncer tai nti es im - nounced de crease of growth rates, but in this pe riod, as well as posed by the reduced sam pling reso lu ti ons of the two stalag - from 9.5 to 7.6 ka BP, the d13C re cords are less well cor re lated. mites in this growth inter val, we may only infer that the oxy gen The d18O trend suggest s a warm ing phase at the be ginning of profil es of V11-22 and V11-117 im ply a gradual rise of the av- the Bo real. Clim ate warming is also doc u mented from central er age temper atures to the level of the pres ent-day values. Since Eu ro pean lake sed i ments by Stockhausen and Zolitschka ~8 ka BP, the d13C has alm ost consta nt val ues around –7‰, (1999) and from Rom anian peat de posit s by Feurdean and sim ilar to the ones measured for the pres ent-day cal cite. The Bennike (2004), which also indi cat e a drier period in NW Ro- growth cessa ti on of the two stalag m ites, at 6.1 and respec tively mania. This phase approx i m ately coin cide s with the appear - 5.6 ka BP, was most proba bly driven by local causes, such as ance and expan sion of the decid u ous forest s at several Rom a- changes in the paths of the drip wa ter or another change in posi - nian sites (Fãrcaº et al., 1999; Rösch and Fischer, 2000; tion, as pre viously seen in the strati gra phy of S117. Björkman et al., 2002, 2003; Bodnariuc et al., 2002). At 9.5– 8.9 ka (10.5–5.8 cm from top), the S22 stalag m ite has a zone with a porous fabric , com posed of small fast-growth CONCLUSIONS calcite crystals, for which the growth rate reaches the max i- mum val ues re corded. Car bon os cil la tions have an am pli tude of 2‰, with a mini m um of –9‰, while the oxy gen profil es Two stalag m ites from the V11 Cave, NW Rom ania show show oscil lati ons around –8‰. This sequence proba bly was consis tent overlap ping stable isotope records for most of their de pos ited in iso to pic dis equi lib rium, trig gered by the higher growth pe riod be tween the Lateglacial and Middle Holo cene . drip rates, which may have af fected the d13C signal, but the av- The d18O iso to pic re cord is in ter preted to be pos i tively cor re lated er age d18O val ues are con sis tent in the two sta lag mites. d18O re- with tem per a ture vari a tions, while the d13C re cord appears to be cords may indi cat e a short cooling cen tered at 9.3 ka BP, corre - mainly con trolled by varia tions in soil CO2 pro duc tion at least lated with lower d18O in the Green land ice core re cords. This dur ing the Lateglacial. Within the limi ta tions impo sed by the low short-term event is present at around 9.5 ka BP in centra l Euro - growth rates and sam pling reso lu ti on of the two stalag m ites, the pean lake sedi m ents (von Grafenstein et al., 1999) and North Lateglacial–Early Holo cene part of the d18O and d13C pro files is At lan tic ma rine sed i ments (Bond et al., 2001). In Rom ania, con sis tent with the Green land ice core re cords. The GI-1e palaeovegetation data also suggest a colder period, marked by (BÝlling) and GI-1c–GI-1a (AllerÝd) in ter vals are well defined the de cline of thermophilous tree taxa (Tilia, Quercus and on the d18O and d13C re cords of S117. The GS-1 (Younger Lateglacial-Mid-Holocene stable isotope records in two coeval stalagmites from the Bihor Mountains, NW Romania 193

Dryas) cold event is re flected in both oxy gen and carbon records, im ents and other speleothem re cords, sug gest ing the in flu ence but it is better marked on the d13C pro files, which in di cate an of the North At lanti c clim ate sys tem over southeastern Europe. abrupt ter mina ti on at 11.7–11.6 ka, well corre lat ed with ice-core data and other proxies . The conti nu ous speleothem de po si tion in Aknowledgements. The authors thank Augusto Mangini, the V11 cave during the GS-1 may point to a less pronounced Jeff Dorale and Victor Polyak for their construc tive com ments that cli mate cooling than in Western Europe. helped to im prove the manuscri pt. This work has bene fited from In the early part of the Holo cene , three short-term shifts to- many dis cus sions with Angi Feurdean on pollen and wards lower d18O in the sta lag mite record s, dated at 11.2–10.6, palaeovegetation data. Thanks are due to Mugur Bogãtean for 10.5–10.2 and 9.4–9.1 ka BP, corre spond with cold events help dur ing cave water sam pling. Techni cal support from marked in the ice-core d18O records. Betwe en 9 and 7.6 ka BP, Albrecht Leiss, Joanneum Research on iso tope anal yses of water the d18O val ues in di cate stable mean tem per atures, but the re - sam ple are greatly ap preci ate d. U-Th dating was supporte d by the cord is less detai led due to re duced growth rates. From 7.6 ka Contra ct 4778836 A of the MinistÀre Français de l’Education 18 BP unti l the end of de posi ti on, d O grad u ally increases to the Nationale, de la Re cher che et de la Technologie to Christiane pres ent-day value. Over all, the Lategla cial-Mid dle Ho lo cene Causse, and partly by IGCP-379 to Yuan Daoxian. CEEPUS Net- cli mate deduced from the isoto pic profil es of the two sta lag- work A105 is acknowl edged for faci li tat ing coop er a ti on and FWF mites may be cor re lated with the ice core re cords and with Project 16258-N06 for parti al finan cia l support. palaeoclimatic data from other proxies such as pollen, lake sed-

REFERENCES

AMMANN B., BIRKS H. J. B., BROOKS S. J., EICHER U., VON DANSGAARD W., WHITE J. W. C. and JOHNSEN S. J. (1989) — The GRAFENSTEIN U., HOFMANN W., LEMHDAL G., SCHWANDER abrupt ter mina tion of the Younger Dryas clim ate event. Nature , 339: J., TOBOLSKI K. and WICK L. (2000) — Quanti fi ca tion of biotic re- 532–533. sponse to rapid clim atic changes around the Younger Dryas — a syn- DENNISTON R. F., GONZALEZ L. A., ASMEROM Y., POLYAK V., the sis. Palaeogeogr. Palaeoclimat. Palaeoecol., 159: 313–347. REAGAN M. K. and SALTZMAN M. R. (2001) — A high-reso lution BELL M. and WALKER M. J. C. (1992) — Late Qua ter nary en vi ron men tal speleothem record of clim atic variabil ity at the Alleröd-Youn ger change: phys i cal and hu man per spec tives. Longman. London. Dryas transi tion in Mis souri, cen tral United States. Palaeogeogr. BENDER M. M. (1968) — Vari ations in the 13C/12C ratios of plants in re la- Palaeoclimat. Palaeoecol., 176: 147–155. tion to the pathway of photosynthetic car bon fix a tion. Photochemistry, DORALE J. A., EDWARDS R. L., ITO E. and GONZÁLEZ L. A. (1998) 10: 1339–1344. — Cli mate and veg eta tion his tory of the midcontinent from 75 to 25 BJÖRCK S., MUSCHELER R., KROMER B., ANDERSEN C. S., ka: a speleothem record from Crev ice Cave, Mis souri, USA. Sci ence, HEINEMEIER J., JOHNSEN S., CONLEY D., KOC N., SPURK M. 282: 1871–1874. and VESKI S. (2001) — High-res o lu tion anal ysis of an early Ho lo - DORALE J. A., EDWARDS R. L. and ONAC B. P. (2002) — Stable iso - cene cli mate event may im ply de creased so lar forc ing as an im por tant topes as en vi ron men tal in di ca tors in speleothems. In: Karst Pro cesses cli mate trig ger. Ge ol ogy, 12: 1107–1110. and the Carbon Cycle (ed. S. Yuan). Final Re port of IGCP, 379: BJÖRCK S., WALKER M. J. C., CWYNAR L. C., JOHNSEN S., 107–120. KNUDSEN K.-L., LOWE J. J., WOHLFARTH B. and INTIMATE FÃRCAª S. (2001) — Pollen data on peat bogs in the Caliman Mountains . MEMBERS (1998) — An event stra tigra phy for the Last Term ina tion Un pub lished Ph.D. the sis. Babes-Bolyai Univ. Cluj-Napoca. in the North At lan tic region based on the Green land ice-core re cord: a FÃRCAª S., DE BEAULIEU J. L., REILLE M., COLDEA G., pro posal by the INTIMATE group. J. Quatern. Sc., 13: 283–292. DIACONEASA B., GOEURY C., GOSLAR T. and JULL T. (1999) — BJÖRKMAN L., FEURDEAN A., CINTHIO K., WOHLFARTH B. and First 14C datings of Late Gla cial and Ho lo cene pol len se quences from POSSNERT G. (2002) — Lateglacial and early Ho lo cene veg e ta tion the Ro manian Carpathians. Compt. Rend. Acad. Sc. Paris, Sc. Vie, de vel op ment in the Gutâiului Mountains , NW Rom ania. Quatern. Sc. 322: 799–807. Rev., 21: 1039–1059. FEURDEAN A. (2005) — Ho locene for est dy nam ics in northwest ern Ro- BJÖRKMAN L., FEURDEAN A. and WOHLFARTH B. (2003) — ma nia. The Ho lo cene, 15 (3): 435–446. Lateglacial and Ho lo cene for est dy nam ics at Steregoiu in the Gutaiului FEURDEAN A. and BENNIKE O. (2004) — Late Qua ter nary palaeo eco - Moun tains, NW Ro ma nia. Rev. Palaeobot. Palynol., 25 (14): 1–34. logi cal and palaeoclimatological re con struc tion in the Gutaiului BODNARIUC A., BOUCHETTE A., DEDOUBAT J. J., OTTO T., Mountains , northwest Rom ania. J. Quatern. Sc., 19: 809–827. FOUNTUGNE M. and JALUT G. (2002) — Ho lo cene vegetational FRIEDMAN I. and O`NEIL J. (1977) — Com pila tion of sta ble isotope history of the Apuseni moun tains, cen tral Ro mania. Quatern. Sc. Rev., frac tion ation fac tors of geo chem i cal in ter est. In: Data of Geo chem is- 21: 1465–1488. try (ed. M. Fleischer). U. S. Geol. Surv. Prof. Pa per, 440-KK. BOND G., KROMER B., BEER J., MUSCHELER R., EVANS M. N., GENTY D., BAKER A., MASSAULT M., PROCTOR C., GILMOUR M., SHOWER W., HOFFMAN S., LOTTI-BOND R., HAJDAS I. and PONS-BRANCHU E. and HAMELIN B. (2001) — Dead carbon in BONANI G. (2001) — Per sistent so lar influ ence on North Atlan tic cli- sta lag mites: car bon ate bed rock palaeodissolution vs. age ing of soil or - mate dur ing the Ho lo cene. Sci ence, 294: 2130–2136. ganic mat ter. Im pli ca tions for 13C vari a tions in speleothems. Geochim. BOTTINGA Y. (1968) — Cal culate d frac tionation factors for car bon and Cosmochim. Acta, 65 (20): 3443–3457. ox y gen iso to pic ex change in the sys tem cal cite-car bon di ox ide-wa ter. HENDY C. H. (1971) — The isoto pic geochem istry of speleothems -I. The J. Physic. Chem., 72: 800–808. calcu la tion of the ef fects of dif ferent modes of form ation on the isoto - DAMM P. (1993) — Consideraþii asupra sistemului carstic Vãrãºoaia–Boga. pic com po si tion of speleothems and their ap pli ca bil ity as Anuarul ºtiinþific al Societãþii Ardelene de Speologie: 30–49. palaeoclimatic in di ca tors. Geochim. Cosmochim. Acta, 35: 801–824. 194 Tudor Tãmaº, Bogdan P. Onac and Ana-Voica Bojar

HENDY C. H. and WILSON A. T. (1968) — Palaeoclimatic data from RÖSCH M. and FISCHER E. (2000) — A ra dio car bon dated Ho lo cene pol- speleothems. Nature, 219: 48–51. len pro file from the Banat Mountains (southwest ern Carpathians, Ro- HUNTLEY B. and PRENTICE C. (1993) — Holo cene veg eta tion and cli- ma nia). Flora, 195: 277–286. mates of Eu rope. In: Global Cli mates since the Last Gla cial Max i mum SCHWARCZ H. P. (1986) — Geo chron ol ogy and iso tope geo chem is try in (eds. H. E. Wrights et al.): 136–169. Min nesota Univ. Press. speleothems. In: Hand book of En vi ron men tal Iso tope Geo chem is try JOHNSEN S. J., CLAUSEN H. B., DANSGAARD W., GUNDESTRUP N. (eds. P. Fritz and J.-Ch. Fontes): 271–303. Elsevier. Am ster dam. S., HAMMER C. U., ANDERSEN U., ANDERSEN K. K., SEPPÄ H., BIRKS H. H. and BIRKS H. J. B. (2002) — Rapid cli ma tic HVIDBERG C. S., DAHL-JENSEN D., STEFFENSEN J. P., SHOJI changes during the Green land stadial 1 (Younger Dryas) to early Ho lo- H., SVEINBJÖRNSDÓTTIR A. E., WHITE J. W. C., JOUZEL J. and cene tran si tion on the Nor we gian Barents Sea coast. Boreas, 31: FISHER D. (1997) — The d18O record along the Green land Ice Core 215–225. Project deep ice core and the problem of possi ble Eemian clim atic in - SHOTYK W., KRACHLER M., MARTINEZ-CORTIZAS A., stabil ity . J. Geophys. Res., 102: 26397–26410. CHEBURKIN A. K. and EMONS H. (2002) — A peat bog record of LAURITZEN S.-E. and LUNDBERG J. (1999) — Speleothems and cli - natu ral , pre-anthropogenic enric h ments of trace ele m ents in atm o - mate: a spe cial is sue of the Ho lo cene. The Ho lo cene, 9 (6): 643–647. spheric aero sols since 12 370 14C yr BP, and their vari ation with Holo - LI W.-X., LUNDBERG J., DICKIN A. P., FORD D. C., SCHWARTZ H. P., cene cli mate change. Earth Planet. Sc. Lett., 199: 21–37. MCNUTT R. and WILLIAMS D. (1989) — High-preci sion STOCKHAUSEN H. and ZOLITSCHKA B. (1999) — En vi ron men tal mass-spec tro met ric ura nium-se ries dat ing of cave de pos its and im pli- changes since 13,000 cal. BP reflected in mag netic and ca tions for palaeoclimate stud ies. Na ture, 339: 534–536. sedimentological proper ties of sedi m ents from Lake Holzmaar (Ger- MAGNY M. (2004) — Ho lo cene cli mate vari abil ity as reflected by many). Quatern. Sc. Rev., 18: 913–925. mid-Eu ro pean lake-level fluc tu a tions and its prob a ble im pact on pre- TÃMAª T. (2003) — Miner al ogy and geo chem istry of speleothems from his toric hu man set tle ments. Quatern. Internat., 113: 65–79. some caves in Bihor Mountains . Ph.D. thesis abstra ct. Babes-Bolyai McDERMOTT F. (2004) — Palaeoclimate re con struc tion from sta ble iso- Univ. Cluj-Napoca. tope vari a tions in speleothems: a review . Quatern. Sc. Rev., 23: TÃMAª T. and CAUSSE C. (2001) — U–Th TIMS chronol ogy of two sta- 901–918. lag mites from V11 Cave (Bihor Moun tains, Ro ma nia). The o retic. McDERMOTT F., FRISIA S., HUANG Y., LONGINELLI A., SPIRO B., App. Karstol., 13–14: 25–32. HEATON T. H. E., HAWKESWORTH C., BORSATO A., KEPPENS TANÞÃU I. (2003) — Recherches pollenanalytiques dans les Carpates E., FAIRCHILD I., VAN BORGH C., VERHEYDEN S. and SELMO orientales (Roumanie). Historie de la veg eta tion et de l’action E. (1999) — Ho lo cene cli mate vari abil ity in Eu rope: ev i dence from humaine. Ph.D. the sis ab stract. Babes-Bolyai Univ. Cluj-Napoca. d18O, tex tural and ex ten sion-rate vari a tions in speleothems. Quatern. TANÞÃU I., REILLE M., DE BEAULIEU J.-L., FÃRCAª S., GOSLAR T. Sc. Rev., 18: 1021–1038. and PATERNE M. (2003) — Veg e ta tion his tory in the East ern Ro ma- MOOK W. G., BOMMERSON J. C. and STAVERMAN W. H. (1974) — nian Carpathians: pollen analy sis of two sequences from the mohoº Car bon iso tope frac tion ation be tween dis solved bi car bon ate and gas- crater. Vegetat. Hist. Archaeobot., 12: 113–125. eous carbon diox ide. Earth Planet. Sc. Lett., 22: 169–176. VON GRAFENSTEIN U., ERLENKEUSER H., MULER J., BRAUER A., NESJE A., MATTHEWS J. A., DAHL S. O., BERRISFORD M. S. and JOUZEL J. and JONSEN S. (1999) — A mid-Euro pean decad al iso - ANDERSSON C. (2001) — Ho lo cene gla cier fluc tu a tions of tope-clim ate record from 15,500 to 5000 years BP. Science, 284: Flatebreen and winter precip i ta tion changes in the Jostedalsbreen re - 1654–1657. gion, western Norway , based on glaciolacustrine sed i ment re cords. WALKER M. J. C. (1995) — Clim atic change in Eu rope dur ing the last The Ho lo cene, 11: 267–280. Gla cial-Inter gla cial tran si tion. Quatern. Internat., 28: 63–76. NIGGEMANN S., MANGINI A., RICHTER D. K. and WURTH G. (2003) WALKER M. J. C., BJÖRCK S., LOWE J. J., CWYNAR L. C., JOHNSEN — A palaeoclimate record of the last 17,600 years in sta lagm ites from S., KNUDSEN K.-L., WOHLFARTH B. and INTIMATE GROUP the B7 cave, Sauerland, Germ any. Quatern. Sc. Rev., 22: 555–567. (1999) — Isoto pic ‘events’ in the GRIP ice core: a stratotype for the ONAC B. P. and LAURITZEN S.-E. (1996) — The clim ate of the last Late Pleis to cene. Quatern. Sc. Rev., 18: 1143–1150. 150,000 years recorde d in speleothems: pre lim i nary re sults from WILLIAMS P. W., KING D. N. T., ZHAO J.-X. and COLLERSON K. D. 18 north-west ern Ro ma nia. The o retic. App. Karstol., 9: 9–21. (2005) — Late Pleis to cene to Ho lo cene com pos ite speleothem O and 13 RENSSEN H., ISARIN R. F. B., VANDENBERGHE J., C chronol o gies from South Island, New Zealand — did a global LAUTENSCHLAGER M. and SCHLESE U. (2000) — Per mafrost as Youn ger Dryas really ex ist? Earth Planet. Sc. Lett., 230 (3–4): a crit i cal fac tor in paleoclimate model ling: the Younger Dryas case in 301–317. Eu rope. Earth Planet. Sc. Lett., 176: 1–5. WOHLFARTH B., HANNON G., FEURDEAN A., GHERGARI L., RICHARDS D. A. and DORALE J. A. (2003) — Ura nium-serie s chronol - ONAC B. P. and POSSNERT G. (2001) — Recon struc tion of clim atic ogy and en vi ron men tal ap pli ca tions of speleothems. In: Ura nium — and envi ron m ental changes in NW Rom ania dur ing the early part of Se ries Geo chem is try (eds. B. Bourdon et al.). Rev. Min eral. the last deglaciation (15,000–13,600 cal years BP). Quatern. Sc. Rev., Geochem., 52: 407–450. 20: 1897–1914. 13 WURTH G., NIGGEMANN S., RICHTER D. K. and MANGINI A. (2004) RIGHTMIRE C. T. (1978) — Sea sonal varia tion in PCO2 and C content of soil at mo sphere. Wa ter Resour. Res., 14: 691–692. — The Youn ger Dryas and Ho lo cene cli mate record of a sta lag mite ROZANSKI K., ARAGUAS-ARAGUAS L. and GONFIANTINI R. from Hölloch Cave (Ba var ian Alps, Ger many). J. Quatern. Sc., 19 (3): (1992) — Rela tion betwee n long-term trends of oxy gen-18 isotope 291–298. com po si tion of pre cip i ta tion and cli mate. Sci ence, 258: 981–985.