1. INTRODUCTION Permo-Triassic (P/Tr) boundary events, which took place approximately 250 Ma ago, led to the most exten- sive mass extinction in the history of life. A number of possible explanations for this profound break in the evolution of life have been proposed, such as volcanic activity, sea-level fluctuation, changes in sea-water chemistry, an extra-terrestrial impact event and various related factors (YOICHI, 1994). The most recently pro- posed cause of the mass extinction at the end of the Per- mian is a combination of these more or less co-occur- ring events operating in three phases (ERVIN, 1996). The first began with the marine regression during the Late Permian and resulted in the destruction of many marine basins, reduction in the habitat area of many organisms and increased climatic instability. The sec- ond phase involved the eruption of the Siberian Traps and further environmental degradation. The final phase may have started immediately prior to the boundary when the Late Permian regression ended and the earli- est Triassic transgression begun. The global events out- lined above coincide with isotope and elemental anom- alies recorded in several P/ Tr boundary sections all Stable Isotope Event Markers Near the Permo-Triassic Boundary in the Karavanke Mountains (Slovenia) Tadej DOLENEC 1, 2, Sonja LOJEN 2, Stanko BUSER 1 and Matej DOLENEC 3 over the world. One of the most remarkable anomalies is the worldwide negative shift of δ13C of inorganic and organic carbon across the P/Tr boundary (MAGARITZ et al., 1992; WANG et al., 1994; WOLBACH et al., 1994; FAURE et al., 1995). A corresponding oxygen isotopic shift is more or less parallel, but less prono- unced. Furthermore, significant shifts in sulphur (KAJI- WARA et al., 1994) and strontium isotopes (KRAMM & WEDEPOHL, 1991) have also been recorded. In this study we present the results of stable isotope analysis of the P/Tr boundary section in the Karavanke Mountains (Fig. 1), and discuss their implications with respect to a better understanding of the nature and caus- es of the P/Tr boundary events in this part of Western Palaeotethys. 2. GEOLOGICAL SETTING AND STRATIGRAPHY In the southern Karavanke Mountains, the Middle Permian Val Gardena Formation of mostly fluvial ori- gin, is overlain by a 270 m thick Upper Permian carbon- ate sequence, which was named the Karavanke Formati- on (BUSER, 1974). The boundary between the two for- mations is transitional and is characterized by thin san- dy red dolomite layers alternating with the topmost Val Gardena shales and sandstones. The thickness of the transitional unit which grades upward into the Karava- nke Formation is about 5 m (DOLENEC et al., 1981). The evaporitic sequence, up to 70 m thick, represents the basal unit of the Karavanke Formation composed of cellular dolomite, which alternates with rare black bitu- minous shales, and grey vuggy dolomites. The evapori- tic sequence is overlain by a 200 m thick succession of fossiliferous biomicritic dolomites. The Upper Permian age of these beds is indicated by calcareous algal asse- mblages, as well as by foraminifera (RAMOV©, 1986). The lithostratigraphic boundary between the Upper Per- mian Karavanke Formation and the Lower Triassic (Scythian) beds is placed at the end of the sedimentation of the well-bedded grey dolomicrite. It is followed by a red coloured partly terrigenous sequence predominantly composed of thin-bedded siltstones, mudstones and sandstones alternating with micritic dolomites, that con- tain no characteristic fossils. These earliest Triassic beds were deposited in an extremely shallow sea, which GEOL. CROAT. 52/1 77 - 81 2 Figs. ZAGREB 1999 Key words: Permo-Triassic boundary, Karavanke Mountains, Slovenia, Oxygen, Carbonate carbon, Organic carbon stable isotopes. 1 Department of Geology, University of Ljubljana, AπkerËeva 12, SLO-1000 Ljubljana, Slovenia. 2 Department of Environmental Sciences, Joæef Stefan Institute, Jamova 39, SLO-1000 Ljubljana, Slovenia. 3 Geoexp d.o.o., Slap 21, SLO-4290 TræiË, Slovenia. Abstract Stable isotope analyses of carbonates and organic matter from the Permo-Triassic boundary section in the Karavanke Mountains, Slove- nia, indicate a further example of the “light carbon” event across the boundary. In this section the changes in carbon isotope values were a direct result of the culmination of the marine regression and associat- ed events at the end of the Permian, which caused a drop in primary productivity, as well as related local environmental changes, with no evidence of any considerable diagenetic overprint. 78 Geologia Croatica 52/1 gradually became a wide, extensive mud flat (ASSE- RETO et al., 1973). Their thickness is about 25 m. In the investigated area these beds are mostly overlain by Lower Triassic dark grey and brown micritic and sparit- ic limestones intercalated with oolitic limestone, marls and shales. 3. MATERIALS AND METHODS The boundary profile in the Karavanke Mountains was sampled at 10 m intervals, except in the vicinity of the lithostratigraphically defined P/Tr boundary where sampling intervals were reduced to 20 cm. The relative stratigraphic position of the samples and the analytical results are presented in Fig. 2. The isotopic measure- ments were carried out on un-dolomitized limestone and un-calcitized dolomite samples. The mineralogy of the carbonate phases was determined by X-ray diffrac- tometry and by examination of thin sections by stan- dard optical methods, including staining with Alizarin- red. All samples were also evaluated by petrographic methods to assess their diagenetic history. Only unre- crystallized or insignificantly recrystallized samples were used for isotopic measurements. Samples were obtained as a split of powder prepared from rock chips remaining after thin section preparation. In order to speed up the reaction time and to ensure complete reac- tion of carbonates, powdered rock samples for δ18O and δ1 3C analysis were prepared by overnight digestion in >100% phosphoric acid at 50°C. CO2 gas released dur- ing acid treatment was cryogenically cleaned and ana- lyzed for O and C isotopic composition on a Varian MAT 250 mass spectrometer. Data were corrected for kinetic fractionation between phosphoric acid and car- bonates using fractionation factors of 1.00925 for cal- cite and 1.01038 for dolomite (BEUKES et al., 1990). The δ18O and δ13C values were normalized by assuming δ1 8O and δ1 3C values of -2 . 44‰ and +2 . 48‰ for IAEA-CO-1 standard on the PDB scale. For preparation of the total organic carbon, pow- dered whole rock samples were treated with heated 3M hydrochloric acid at 50°C to react with the carbonates. Upon cessation of CO2 evolution, excess acid was removed by repeated washing (three to four times) with doubly distilled water until a neutral pH was deter- mined. After the final decanting of water, the carbonate free residues were oven-dried at 50°C. Organic carbon isotope ratios were measured in the carbonate-free resi- dues in the Europa 20-20 Stable Isotope Analyser (Europa Scientific Ltd.) with the ANCA-NT prepa- ration module for on-line combustion of bulk solid samples and chromatographic separation of the gases. Organic carbon isotope values were calibrated using the IAEA-CH-7 standard with a δ1 3C value of -3 1 .8‰ on the PDB scale. All bulk rock and carbonate-free residue samples were measured two or three times. The results are reported in the conventional delta notation as ‰ devia- tions from the PDB standard (CRAIG, 1957) for oxy- gen, carbonate and organic carbon. The analytical pre- cision based on multiple analysis of internal laboratory standards was ±0 . 02‰ for δ1 8O, ±0 . 01‰ for δ1 3Cc a r b . a n d ±0 . 0 08‰ for δ1 3Co r g ., respectively. Overall analyti- cal reproducibility of the isotopic data was ±0.15‰ for oxygen, ±0.1‰ for carbonate carbon and ±0.095‰ for organic carbon. 4. RESULTS AND DISCUSSION The transition from Middle Permian to Upper Per- mian is characterized by a considerable enrichment of dolomite with 1 3C (from -2.50 to +3 . 83‰) and 1 8O (from -8.67 to -3 . 34‰) (DOLENEC et al., 1981), as well as by a drop in δ1 3C of total organic carbon (from -21.90 to -2 4 . 49‰). Positive δ1 3C and δ1 8O shifts in dolomite (Fig. 2) may reflect the transgression of the Palaeotethys Sea on the vast alluvial Middle Permian landscape. Documentation of this transgression exists not only in the south of Tethys, but also to the north in the Zechstein basin (ASSERETO et al., 1973). The gen- eral hypothesis, suggested to explain positive δ1 3C shifts of carbonate carbon, is that the expansion of shal- low shelf areas increased the organic carbon burial rate and enriched the ocean in 13C (COMPTON et al., 1990; FAURE et al., 1995). The corresponding oxygen iso- tope excursion is similar to that in δ1 3C, and also sug- gests a change from terrestrial to marine-evaporitic con- ditions. The δ1 3C of terrestrial plant remains in the Val Gardena Formation varies between -22.13 and -21.65‰ (DOLENEC, 1984). These values are up to 2.8‰ more positive than those of the evaporitic sequence. A nega- tive shift at the Middle Permian-Upper Permian transi- tion most likely suggests that the ratio of terrestrial to marine organic carbon changed. Fig. 1 Reconstruction of the supercontinent Pangea in the Late Per- mian after SUN et al. (1989). The point indicates the approximate position of the studied area in the Karavanke Mountains. The dolomites of the basal evaporitic sequence show a variation of δ1 3Cc a r b . in the range of +0.69 to +3 . 83‰, and of δ1 8O between -5.88 to -3 . 34‰. The oxygen isotopic composition of the evaporitic sequence is not as high as expected from recent evaporitic envi- ronments (TUCKER, 1990). The observed reduction in δ18O indicate an influx of fresh water into the evaporitic basin, less evaporation and/or a variety of different post-depositional processes, that may affect either oxy- gen or carbon or both stable isotopic records (BRAND & VEIZER, 1981; MAGARITZ & HOLSER, 1991). By selecting the least visibly weathered and recrystalli- zed samples from the investigated sections we attempt- ed to minimise the possible post-depositional effects. A weak positive correlation (r = 0.40) between δ1 8O and δ1 3C of the dolomite samples and a weak negative cor- relation (r = -0.22) between δ1 8O and δ1 3C of the lime- stone samples most probably suggests that the isotopi- cally light meteoric water and dolomitizing fluids dur- ing diagenesis and burial to some extent reset the origi- nal whole rock oxygen and carbon isotopic composi- tion. The low positive correlation (r = 0.51) between δ1 3Cc a r b . and δ1 3Co r g . indicates that the whole rock iso- topic composition was not modified by the oxidation of organic matter during diagenesis. Based on these obser- vations, we tend to believe that the isotopic composi- tion of the investigated Karavanke Mountain carbonate rocks has not been seriously altered after their forma- tion and that the primary palaeoceanographic signal was not completely overprinted. The transition from Permian to Triassic is character- ized by a prominent negative shift of carbonate δ1 3C as well as total organic carbon and a similar, but less pro- nounced δ1 8O decrease. The major drop of δ1 3Cc a r b . a n d δ1 3Co r g ., similar to that seen globally, begins approxi- mately 15 m below the boundary. The δ1 3Cc a r b . c u r v e reaches a minimum peak value of -1 . 86‰ about 8 m below the boundary and after that there is another mini- mum of -1 . 57‰ at the end of the Permian and a posi- tive excursion of +0 . 55‰ in the lowermost Scythian, before settling to values which are 1 to 2‰ lower rela- tive to those in the Upper Permian. The position of the negative δ1 3Cc a r b . peak anomaly may indicate that the P/Tr boundary in the Karavanke Mountains should be placed a little further downsection, since in the Carnic Alps a dramatic δ1 3Ccarb. drop occurs right after the stratigraphic P/Tr boundary which is placed within the lowermost 0.5 m of the 4 m thick oolitic Tessero Hori- zon or at its base (HOLSER et al., 1991). In the Kara- vanke Mountains this oolitic unit at the base of the Scythian was not recognized. The corresponding total organic carbon isotope curve is essentially parallel. The first (-2 9 . 89‰) and the second (-2 6 . 65‰) negative anomaly for δ1 3Co r g . are coeval with the first two nega- 79Dolenec, Lojen, Buser & Dolenec: Stable Isotope Event Markers Near the Permo-Triassic Boundary... Fig. 2 Stable isotope composition of carbonates (δ13Ccarb., δ18O), total organic carbon (δ13Corg.) and carbon isotope fractionation between carbon- ate and total organic carbon (δ13Ccarb.-org.) across the Permo-Triassic boundary in the Karavanke Mountains. a) Dark grey and brown micritic and sparitic limestone intercalated with oolitic limestone, marls and shales; b) a red partly terrigenous sequence composed of siltstones, mudstones and sandstones alternating with micritic dolomites; c) light grey fossiliferous biomicritic dolomite; d) an evaporitic sequence composed of cellular dolomite intercalated with black bituminous shales and grey vuggy dolomites; e) a transitional unit composed of sandy red dolomite alternating with shales and sandstones; 1-3) carbonate and organic carbon anomalies. 80 Geologia Croatica 52/1 tive shifts observed for δ13C of carbonate carbon, while the third decrease of δ1 3Corg. (-2 6 . 92‰) occurs in the Triassic, 7 m below the third negative shift of δ1 3Cc a r b . values. During the Permian-Triassic transition the Kara- vanke Mountains constituted a semi-restricted marine basin, surrounded by vast areas of an extremely shallow epicontinental sea into which a considerable amount of terrigenous material was transported (DOLENEC et al., 1981). The environmental stress conditions caused decre- ased bioproductivity in the upper water column and an associated decrease in dissolved carbonate (ZHENG et al., 1993). A drop of δ1 3C values in carbonate precipi- tates reflects this decrease. The incorporation of light carbon from eroded and oxidized organic matter into carbonates and photosynthetic marine organisms would also cause both to become isotopically lighter (WOL- BACH et al., 1994). The variation between δ13Ccarb. and δ1 3Corg. values, which can be used to evaluate major changes in the carbon cycle by removing the effect of changes in the δ13C of the surface water dissolved inor- ganic carbon reservoir (HOLLANDER et al., 1993), also indicates changes in bioproduction. For the P/T r transition in the Karavanke Mountains the decreasing isotopic composition of total organic carbon of 6.5‰ (from -23.19 to -2 9 . 69‰) and increasing ∆δ1 3Cc a r b . - o r g . of 3.78‰ (from 24.66 to 28.44‰) about 8 m below the boundary can be explained by a dramatic decrease in primary production. This breakdown in the photosyn- thesis/ respiration cycle appears to coincide with events related to the accelerated fall in sea level and/or wide- spread volcanic activity in Siberia. These events began substantially before the end of the Permian and affected the carbon cycle long enough to cause a clear perturba- tion in both δ13C records. The δ13C shift of 3.5‰ (from -8.89 to -25.39‰) in total organic carbon toward high- er values and the reduction in ∆δ1 3Cc a r b . - o r g . by 4.1‰ (from 28.23 to 24.13‰) at the P/Tr boundary could be the consequence of a subsequent recovery period in the biological system before the sea gradually became a wide, extensive mud flat. This situation was changed after the sedimentation of the red terrigenous sequence, due to the earliest Triassic marine transgression which spread anoxic bottom water over the entire region of the Karavanke Mountains, and marked the end of the peri- od of extreme variability in δ1 3C of both carbonate and organic carbon. During this phase deposition of the Lower Scythian dark grey and brown limestones began. The variability of δ1 8O in the interval straddling the P/Tr boundary shows slightly different trends with respect to those of δ1 3C in carbonate and organic car- bon, roughly indicating environmental changes from marine to desultory evaporitic conditions affected by an excessive input of terrigenous material, and by local freshening of waters, due to the influx of more or less isotopically modified meteoric waters into the sedimen- tary environment. A significant drop of δ1 8O from - 4 . 19‰ to -9 . 51‰ approximately 25 m above the bou- ndary is stratigraphically coincident with abrupt chan- ges in the lithology, as well as carbonate mineralogy of the rocks. The Scythian limestones are considerably depleted in 1 8O (from 3 to 8‰) relative to both the underlying lowermost Scythian dolomite and the Upper Permian dolomite of the Karavanke Formation. Such depletion cannot be interpreted only in terms of dolomite-calcite fractionation which is thought to be between +3 and +4‰ (DICKSON, 1990). It may also be caused by a change in the δ1 8O of seawater, a decrease of salinity, changes in the depositional envi- ronment, diagenetic processes and burial, or some com- bination of all of these factors. Due to the muddled oxy- gen isotopic pattern, it is difficult to give a definite explanation of the observed δ1 8O variations at the P/T r transition. However, it is important to note that the δ13C signals of inorganic and organic carbon are independent of lithology and are preserved in both Scythian dolo- mite and limestone. 5. CONCLUSIONS The results we have presented in this study indicate that the transition from the Permian to the Triassic in the Karavanke Mountains is characterized by a strongly negative complex pattern of carbonate and organic car- bon anomalies, accompanied by changes in the δ1 8O values of the boundary carbonate rocks. These anom- alies are supposed to record extremely adverse changes of the environment due to global events which marked the terminal Permian productivity crash and the P/ T r boundary. We suggest a causal connection between the global carbon isotopic anomalies and the marine regres- sion at the end of the Permian, but the exact controls remain complex and enigmatic. Acknowledgements The Ministry of Science and Technology, Republic of Slovenia, and Geoexp d.o.o., TræiË, Slovenia financi- ally supported this study. To both these institutions we express our sincere thanks. 6. REFERENCES ASSERETO, R., BOSSELINI, A., FANTINI SESTINI, N. & SWEET, W.C. (1973): The Permian-Triassic boundary in the Southern Alps (Italy).- In: LOGAN, A. & HILLS, L.V. (eds.): The Permian and Triassic Systems and Their Mutual Boundary. Albertal. Soc. Petrol. Geol. Mem., 2, 176-199. BEUKES, N.J., KLEIN, C., KAUFMAN, A.J. & HAY- ES, J.M. (1990): Carbonate petrography, kerogen distribution, and carbon and oxygen isotope variati- ons in an Early Proterozoic transition from limesto- ne to iron-formation deposition, Traansvaal Super- group, South Africa.- Econ. Geology, 85, 663-690. 81Dolenec, Lojen, Buser & Dolenec: Stable Isotope Event Markers Near the Permo-Triassic Boundary... BRAND, U. & VEIZER, J. (1981): Chemical diagene- sis of a multicomponent carbonate system - 2: stable isotopes.- J. Sediment. Petrol., 51, 987-997. BUSER, S. (1974): Neue Feststellungen im Perm der westlichen Karavanken.- Carinthia II, 164/84, 27- 37. COMPTON, J.S., SNYDER, S.W. & HODELL, D.A. (1990): Phosphogenesis and weathering of shelf sediments from the southern United States: Implica- tions for Miocene δ1 3C excursions and global cool- ing.- Geology, 18, 1227-1230. CRAIG, H. (1957): Isotopic standards for carbon and oxygen and correction factors for mass spectromet- ric analysis of carbon dioxide.- Geochim. Cosmo- chim. Acta, 12, 133-149. DICKSON, T. (1990): Carbonate mineralogy and che- mistry.- In: TUCKER, M.E. & WRIGHT, V.P.: Carbonate Sedimentology. Blackwell Scientific Publications, Oxford, 482 p. DOLENEC, T. (1984): The origin of the uranium mine Æirovski vrh.- Unpublished PhD Thesis, University of Ljubljana, 187 p. DOLENEC, T., OGORELEC, B. & PEZDI», J. (1981): Upper Permian and Scythian beds in the TræiË area.- Geologija, 24/2, 217-238, Ljubljana. ERVIN, D.H. (1996): Permian global bio-events.- In: WALISER, O.H. (ed.): Global Events and Event Stratigraphy in the Phanerozoic. Springer-Verlag, Berlin, 251-264. FAURE, K., DE WIT, M.J. & WILLIS, J.P. (1995): Late Permian global coal hiatus linked to 1 3C - depleted CO2 flux into the atmosphere during the final consolidation of Pangea.- Geology, 23, 507- 510. HOLLANDER, D.J., McKENZIE, J.A. & KENNETH, J.H. (1993): Carbon isotope evidence for unusual plankton blooms and fluctuation of surface water C O2 in “Strangelove Ocean” after terminal Creta- ceous event.- Palaegeogr. Palaeoclim. Palaeoecol., 104, 229-237. HOLSER, W.T., SCHÖNLAUB, H.P., BOECKEL- MANN, K. & MAGARITZ, M. (1991): The Permi- an-Triassic of the Gartnerkofel-1 core (Carnic Alps, Austria): synthesis and conclusions.- Abh. Geol. B., A 45, 213-232. KAJIWARA, Y., YAKAMITA, S., ISHIDA, K., ISHI- GA, H. & IMAI, A. (1994): Development of a lar- gely anoxic stratified ocean and its temporary mas- sive mixing at the Permian/Triassic boundary sup- ported by the sulfur isotopic record.- Palaeogeogr. Palaeoclim. Palaeoecol., 111, 367-379. KRAMM, U. & WEDEPOHL, K.H. (1991): The isoto- pic composition of strontium and sulfur in seawater of Late Permian (Zechstein) age.- Chem. Geol., 90, 253-262. MAGARITZ, M. & HOLSER, W.T. (1991): The Per- mian-Triassic of the Gartnerkofel-1 Core (Carnic Alps, Austria): carbon and oxygen isotope varia- tion.- Abh. Geol. B., A 45, 149-163. MAGARITZ, M., KRISHNAMURTHY, R.V. & HOL- SER, W.T. (1992): Parallel trends in organic and inorganic carbon isotopes across the Permian/Trias- sic boundary.- Am. Jour. Sci., 292, 727-739. RAMOV©, A. (1986): Marine development of the uppermost Æaæar beds and the lowermost Scythian beds.- In: Permian and Permian-Triassic Boundary in the South Alpine Segment of the Western Tethys. Excursion Guidebook (ed. by the Italian IGCP 203 Group), 39-42. SUN, S., LI, J., CHEN, H., PENG, W., HSU, K.J. & SHELTON, J.W. (1989): Mesozoic and Cenozoic sedimentary history of South China.- Am. Assoc. Pet. Geol. Bull., 73, 1247-1269. TUCKER, M.E. (1990): Geological background to car- bonate sedimentation.- In: TUCKER, M.E. & WRI- GHT, V.P.: Carbonate Sedimentology. Blackwell Scientific Publications, Oxford, 28-69. WANG, E., GELDSETZE, H.H.J. & KRUOSE, H.R. (1994): Permian-Triassic extinction: Organic δ1 3C evidence from British Columbia, Canada.- Geology, 22, 580-584. WOLBACH, W.S., ROEGGE, D.R. & GILMOUR, I. (1994): The Permian-Triassic of the Gartnerkofel-1 core (Carnic Alps, Austria): Organic carbon isotope variation.- In: New Developments Regarding the KT Event and Other Catastrophes in Earth History. LP, Contribution No. 825, Lunar and Planetary Institute, Houston, 138 p. YOICHI, E. (1994): Patterns and paleoenvironmental implications of end-Permian extinction of Rugosa in South China.- Palaeogeogr. Palaeoclimatol. Palaeo- ecol., 107, 165-177. ZHENG, Y., HONG-FEI, H. & LIAN-FENG, Y. (1993): Carbon and oxygen isotope event markers near the Frasnian Famenian boundary, Luxin sec- tion, South China.- Palaeogeogr. Palaeoclimatol. Palaeoecol., 104, 97-104. Manuscript received January 18, 1999. Revised manuscript accepted May 28, 1999. 82 Geologia Croatica 52/1