Biodiversity Evolution Through the Permian—Triassic Boundary Event: Ostracods from the Bükk Mountains, Hungary

One of the most complete Permian-Triassic boundary sections located in the Bükk Mountains (Hungary) was sampled for ostracod study. Seventy-six species are recognized, belonging to twenty genera. Fifteen new species are described and figured: Acratia? jeanvannieri Forel sp. nov., Acratia nagyvisnyoensis Forel sp. nov., Bairdia anisongae Forel sp. nov., Bairdia davehornei Forel sp. nov., Callicythere? balvanyseptentrioensis Forel sp. nov., Cytherellina? magyarorszagensis Forel sp. nov., Eumiraculum desmaresae Forel sp. nov., Hollinella fengqinglaii Crasquin sp. nov., Hungarella gerennavarensis Crasquin sp. nov., Langdaia bullabalvanyensis Crasquin sp. nov., Liuzhinia venninae Forel sp. nov., Liuzhinia bankutensis Forel sp. nov., Microcheilinella egerensis Forel sp. nov., Reviya praecurukensis Forel sp. nov., Shemonaella? olempskaella Forel sp. nov. One species is renamed: Bairdia baudini Crasquin nom. nov. Comparison of the Bálvány North section with the Meishan section (Zhejiang Province, South China), Global Boundary Stratotype Section and Point (GSSP) of the Permian-Triassic Boundary (PTB), reveals discrepancies linked to the environmental setting and particularly to bathymetry. The stratigraphical distribution of all the species is given and diversity variations are discussed. The Bálvány North section exhibits the lowest extinction rate of all PTB sections studied for ostracods analysis associated with a high level of endemism.


Introduction
The Permian-Triassic Boundary (PTB) was recognized in several sections of the Bükk Mountains (Northern Hungary) in continuous marine successions. The Bálvány North section (48°06'179N, 20°28'491E) is located on the northern slope of Mount Bálvány, 1 km NW of Bánkút (northern Bükk Moun− tains) (Fig. 1) and is one of the best known localities for the Permian-Triassic transition. This outcrop was studied in great detail by Haas et al. (2004Haas et al. ( , 2006Haas et al. ( , 2007 and Haas (2006, 2009). These authors described the facies development along the section. Numerous ostracods were observed in thin sections. We sampled this section for ostracod analysis in the framework of a large−scale study on the extinction and recov− ery of this group through the PTB interval. This research was conducted in South China: Meishan GSSP, Zhejiang Province Forel and Crasquin 2011b); Guizhou Province (Forel et al. 2009); Sichuan Province (Crasquin− Soleau and Kershaw 2005); Guangxi Province (Crasquin− Soleau et al. 2006); in Tibet Forel and Crasquin 2011a;Forel et al. 2011); in Northern Italy, Bulla section (Crasquin et al. 2008); in SW Taurus, Turkey, Çürük Da section (Crasquin−Soleau et al. 2002, 2004a. In Hun− gary, Kozur published a systematic (1985b) and a biostrati− graphic evaluation (1985a) of Upper Palaeozoic ostracods from the Bükk Mountains. Here, seventy−six species belong− ing to twenty genera are recognized and their potential bio− stratigraphic utility evaluated. Fig. 1. Location map of the northern part of the Bükk Mountains (after Less et al. 2002) with Bálvány North P-T boundary section. Fm., Formation.   limestone  sandy marlstone   340  320  300  280  260  240  220  200  180  160  140  120  100  80  60  40  20  0  Permian bivalve and brachiopod assemblage (Posenato et al. 2005). In addition, carbon isotope values from the bioclastic limestone display a negative shift with a prominent peak in the upper two−third of the BSB , which is considered as an isotope chemostratigraphic marker (e.g., Erwin et al. 2002). Numerous small cavate spores were en− countered from the uppermost part of the BSB suggesting that this interval may be of Triassic age (Haas et al. 2004). The conodont Hindeodus parvus (Kozur and Pjatakova, 1976), which is recognised as the marker for the base of the Triassic (Yin et al. 1996(Yin et al. , 2001, was found in the lower part of the overlying platy mudstone unit (0.5 m; Gerennavár Limestone Formation). Thus, the first documented occur− rence of the conodont H. parvus is 0.55 m above the negative peak of the d 13 C curve, and 0.2 m above the Formation boundary (Haas et al. 2007;Sudar et al. 2008). Upsection, planar stromatolites occur (0.6 m in the studied section).

Material and methods
Twenty−five samples have been processed for ostracod anal− ysis ( Fig. 2A). The extraction of calcareous ostracod cara− paces from calcareous rocks is achieved by hot acetolysis (Lethiers and Crasquin 1988;. Nineteen samples yielded ostracods (Fig. 2B). Seventy−six species belonging to twenty genera are identified and figured (Figs. 4,6,10,12,15,18,(21)(22)(23). Fifteen species are new and here described. Nearly all specimens are represented by complete carapaces; this testifies to absence or limitation of post−mortem transportation with low wave energy and/or rapid burial by high sedimentation ratio (Oertli 1971). Some of the described species are presented in open nomenclature because of poor preservation and/or too low number of speci− mens to differentiate between intraspecific variations and different species. However, we decided to figure all the ma− terial recovered because these new data exemplify an unex− pected survival phenomenon of microbial origin in a refuge area and are important for the understanding of events fol− lowing the end−Permian extinction.
Diagnosis.-A species of Hollinella with an elongated cara− pace, a small H/L ratio, a L 1 vertical and thin, bulbous L 2 and L 3 , and a well−expressed L 4 , which is connected ventrally with L 1 .
Description.-Carapace elongated (H/L 0.47-0.53); DB long and straight; ACA 110-120°; AB with large radius of curva− ture, maximum of curvature located at mid−H; VB slightly convex to nearly straight; PB with small radius of curvature, maximum of curvature located high, near to the DB; PCA quite rectangular; carapace flattened along free margins; L 1 thin and quite vertical connected ventrally to L 4 which is al− ways expressed; L 2 bulbous, located at mid−H, it could be sub− divided vertically into two nodes connected ventrally; L 3 bul− bous and rounded, generally not overpassing DB; maximum of H located behind anterior third of L; ornamentation not ob− served. Internal features unknown.
Discussion.-The preservation is not very good. All the specimens are represented by isolated valves. Many authors have attributed Lower Triassic specimens of Hollinella from South China (Zheng 1976;Wang 1978;Wei 1981;Hao 1992Hao , 1994 and from the Bükk Mountains (Kozur 1985b) to the species Hollinella tingi (Patte, 1935). Although these specimens without doubt belong to the genus Hollinella they are unlikely H. tingi (see discussion in Crasquin−Soleau et al. 2004a). In most cases, the specimens from the Lower Trias− sic are too corroded to allow a precise attribution. H. feng− qinglaii sp. nov. described here differs from "Hollinella tingi" sensu Wang (1978) and sensu Wei (1981) (Gerry et al. 1987). This last species has a more elongated L 2 and L 1 is not clearly connected to ventral lobe. The specimen of identified by Kozur (1985a: pl. 13: 3) as Hollinella tingi is incomplete and has a larger posterior border and a more bulbous L 3 .  Description.-Carapace with long straight DB; ACA 140-150°; AB with large radius of curvature, maximum located at mid−H; VB long, straight in the youngest specimens, a little convex in the adults; PB with quite small radius of curvature, maximum located at or a little below mid−H; DB and VB converge towards the posterior end with an angle of 15°at the instars and 25°at the adults; PCA 150°; S 2 very faint, vis− ible on adult specimens located at mid−H and mid−L; RV overlaps LV all along free margins; surface smooth. Internal features unknown. Discussion.-Specimens attributed to Langdaia suboblonga Wang, 1978 were found in the Lower Triassic of Northern It− aly (Bulla section; Crasquin et al. 2008). This specific attri− bution is wrong. Langdaia suboblonga Wang, 1978 Guan et al. 1978Guan et al. : 152, pl. 39 :10. 1987 Silenites subsymmetrica Shi, 1987;Chen 1987: 63, pl. 15: 4-11. 2004 Bairdia subsymmetrica (Shi, 1987) Discussion.-This species was described by Shi (Shi and Chen 1987) from the uppermost Permian (Changhsingian) of Meishan. It was attributed to the genus Bairdia by Crasquin− Soleau et al. (2004a, b (Wang 1978), Guangxi (Shi and Chen 2002), and Meishan GSSP section, Zhejiang ). The PB are very similar in the two species. In A. nagyvisnyoensis sp. nov., the maximum of curvature is located lower, the Hmax is located more posteriorly and the VB is more convex. A. nagyvisnyoensis sp. nov. is also close to Acratia zhongyin− gensis Wang, 1978 from Wuchiapingian and Changhsingian of Southern China: Guizhou and Yunnan (Wang 1978), Guangxi (Shi and Chen 2002), and Meishan GSSP section, Zhejiang , and from the Changhsin− gian of Bulla section, Italy (Crasquin et al. 2008 Material.-5 complete and 1 broken carapaces, samples 08BAN62, 63, 64, and 67 (see Fig. 2B), level 8, Gerennavár Limestone Formation, Bálvány North section, Hungary, Griesbachian, Lower Triassic.
Diagnosis.-A species of Hungarella with a high vertical PB, Hmax located behind mid−L.

Diagnosis.-A species of Cytherellina? with a high and ver− tical PB and Hmax located in front of mid−L.
Dimensions.-L 200-550 μm; H 110-320 μm (see Fig. 14). Description.-Carapace short (H/L 0.60); PB almost vertical with angular contact with VB; VB straight to gently concave at both valves; dorsal parts of LV more or less regularly rounded, DB and PDB in continuity, ADB nearly straight; dorsal parts of RV clearly distinct, PDB, DB and ADB long and straight; AB regularly arched with maximum of curva− ture located at mid−H; at adult specimens AB could be later− ally flattened; Hmax located in front of mid−L; LV slightly overlaps RV all around the carapace with maximum at PDB. Internal features unknown. Discussion.-Cytherellina? magyarorszagensis sp. nov. is comparable to Pseudobythocypris guiqianensis Yuan, 2009 from the Upper Changhsingian of South China (Yuan et al. 2007(Yuan et al. , 2009. The latter species is longer and the maximum of convexity of the AB is located lower. Great confusion exists in the systematics of the smooth shelled bairdiid genera of the Upper Permian-Lower Triassic. A full revision of all these forms is necessary. Family Pachydomellidae Berdan and Sohn, 1961Genus Microcheilinella Geis, 1933 Fig. 16).
Description.-Carapace moderately elongated (H/L 0.60); LV: DB gently convex to straight; AB with quite large radius of curvature (particularly in larval stages) and ventral part quite vertical; VB slightly convex to straight; PB regularly rounded with maximum of curvature located close to mid−H; RV: DB straight and inclined backwards; anterior part of the carapace in shape of a quadrant, quite vertical in ventral part; VB straight; PB equivalent to PB of LV; LV overlaps RV all around the carapace; relative weak overlap with a maximum located at posterior part of DB; hinge line invaginated; larval stages show a larger AB than adults. Internal features unknown. Discussion.-Microcheilinella egerensis sp. nov. is quite close to two species described from the Upper Permian of the Meishan section , Microcheilinella shicheni Crasquin, 2010 andMicrocheilinella rectodorsata Forel, 2010, both from the Changhsingian. M. shicheni has a PB with a smaller radius of curvature, a convex DB in both valves and more−or−less parallel and horizontal DB and VB. M. rectodorsata has quite the same shape at the AB but PB shows a smaller radius of curvature and the carapace is more elongated. M. hungarica Kozur, 1985 from the Lower Wu− chiapingian of Bükk Mountains (Kozur 1985b) is longer and has a more slender PB particularly in RV.
Very few species of Microcheilinella are recorded from the Triassic: M. sp. from the Spathian of Pakistan (Sohn 1970), M. cf. venusta Chen, 1958 and M.? sp. 1 sensu Crasquin−Soleau et al. (2006)  Family Bythocytheridae Sars, 1926Genus Callicythere Wei, 1981 Type species: Callicythere emeiensis Wei, 1981;Lower Triassic, Sichuan Province, China Callicythere? balvanyseptentrioensis Forel sp. nov. Description.-Carapace quite equivalve with surface smooth without S 2 nor narrow furrow; sexual dimorphism is well−expressed. Heteromorphs: DB convex; AB laterally compressed with large radius of curvature and maximum of convexity located at or below mid−H; VB straight and over− reached by posteroventral wing−like inflation; PB laterally flattened with quite large radius of curvature and maximum of convexity located at or below mid−H; Hmax located at anterior part of DB. Tecnomorphs: DB straight to gently convex; AB laterally compressed with large radius of cur− vature and maximum of convexity located at or below mid−H; VB as in females; PB as in females but with smaller radius of curvature; Hmax located at anterior part of DB; carapace in dorsal view thinner than females. Internal fea− tures unknown. Discussion. -Wei (1981)  the Family Cytherissinellidae Kashevarova, 1958. It seems that this attribution is due to the similarity between Calli− cythere and Lutkevichinella Schneider, 1956 (for discussion see Wei 1981: 506). The description of the Family Cytheris− sinellidae is: "elongate suboblong, dorsal margin straight, an− terior and posterior ends rounded, with faint to distinct narrow sulcus extending straight downward from mid−dorsal region; surface reticulated and may bear inconspicuous longitudinal ribs." (Moore 1961: Q290-Q292). In the description of Calli− cythere, Wei (1981) pointed out "… shallow V−shaped de− pression, on mid−dorsal region". Because of this "V−shaped depression", close to a true S 2 , and the presence of latero−ven− tral structures we attribute this genus to the Bythocytheridae Sars, 1926. The doubt about the generic attribution of our new species is due to the absence (or non−preservation?) of a dorso−median depression or sulcus.
Callicythere? balvanyseptentrioensis sp. nov. is close to Callicythere postiangusta Wei, 1981 from the Lower Trias− sic of Emei (Sichuan, South China; Wei 1981). The latter species has a PB with a larger radius of curvature and less distinct ACA.
Order and suborder indet. Genus Eumiraculum     Material.-6 complete and 1 broken carapaces, sample 08BAN47 (see Fig. 2B Remarks on biostratigraphic value of ostracods in this section Kozur (1985a, b) proposed a zonation based on ostracods for the Upper Palaeozoic of Bükk Mountains. However, in these papers, there is no section with ostracod distribution available. He defined seven ostracod assemblage zones from the Mosco− vian (middle Upper Carboniferous) to Lower Triassic. Kozur (1985a) defined an "Indivisia buekkensis Assemblage Zone" as index of Upper Permian and a "Hollinella tingi Assemblage Zone" as index of index of lowermost Triassic. This last assem− blage is associated with Isarcicella isarcica (Huckriede, 1958), conodont index of lowermost Triassic. The species occurring in the two last assemblage zones, VI "Indivisia buekkensis As− semblage Zone" and VII "Hollinella tingi Assemblage Zone", which correspond to our data ( Table 1). The zones are defined from borehole sediments from Bükk Mountains (see Kozur 1985a). The two species Indivisia buekkensis and "Hollinella tingi" (Hollinella tingi [Patte, 1935] is an Lower Permian spe− cies from South China; see the discussion about this last spe− cies above and in Crasquin−Soleau et al. 2004a) are not recog− nized in Bálvány section. Indivisia symmetrica Kozur, 1985, which is present in Kozur's (1985a) "Indivisia buekkensis As− semblage Zone" was recognized in the upper part of Meishan GSSP section ). This species could be a good marker for the uppermost Changhsingian. Callicythere mazurensis (Styk, 1972) which is present in "Hollinella tingi Assemblage Zone" was recognized in the Upper Changhsin− gian of Bulla section (North Italy, Crasquin et al. 2008 Kozur (1985a). Asterisked is a species which actually is present before its "assemblage zone".

Indivisia symmetrica
Kozur, 1985 *Judahella bogschi bogschi Kozur, 1985 Praepilatina sp.   Nineteen of the twenty−five samples collected from this interval yielded ostracods ( Fig. 2A). The species distribu− tion is given in Fig. 2B; abundance and species richness are presented in Fig. 20A. In productive samples, specific rich− ness varies from 1 (08BAN66) to 28 (08BAN47) species and abundance from 4 (08BAN66) to 2135 (08BAN63) specimens. In most assemblages the variation of species richness correlates with the abundance. Only the samples 08BAN62 and 08BAN63 record a strong disparity between the abundance (increase from 1017 to 2135) and the number of species (decrease from 20 to 8). Assemblages of the Gerennavár Formation (08BAN61 to 08BAN69) show a higher abundance than the samples of the Nagyvisnyó For− mation (08BAN47 to 08BAN60). Based on the iridium con− tent, a change of sedimentation rate seems to occur at the transition between the uppermost beds of the Nagyvisnyó Formation and the lower part of the Boundary Shale Beds (Haas et al. 2007). Therefore we consider that changes in the sedimentation rate do not influence the diversity of our assemblages. However, we did not sample the BSB because of the siliciclastic lithology. We observed carefully the rocks with lens on field and we haven't seen any ostracods. The extinction rates from the Permian to the Triassic at the Bálvány North section show particularities. The number of superfamilies/ families does not change from the Permian to the Triassic, although a decrease is recorded for the number of genera (from 17 to 14 genera) and species (from 54 to 36 species). The specific extinction rate is 74% (40 species ex− tinct at PTB). Thus, at the Bálvány North section, the loss of ostracod diversity is confined to the genus and species lev− els. The species extinction rate at Bálvány North is signifi− cantly lower than the 99% recorded at Meishan section, GSSP of the PTB (Zhejiang Province, South China; Cras− quin et al. 2010a ;Forel et al. 2011;Crasquin and Forel in press). Fourteen species cross the PTB. On 36 Griesbachian species, 22 are new and document a specific turnover rate of about 61% in Bálvány North. This specific turnover is 67% in Meishan (but only 3 species are recognized) and 90% in Bulla (Italian Dolomites) (Crasquin and Forel in press). Two successive patterns can be distinguished here among the surviving species (Fig. 2B): the first one is documented by species which cross the PTB but disappear a few centi− metres above the PTB; the second one is recorded by 7 (?9) species with a longer range within the Lower Triassic. It re− flects 2 successive phases of survival for the ostracods, which are associated with the microbialites at the Bálvány North section: · 24 ostracod species are recorded in the first 30 centimetres of the Triassic Formation, including 13 species, which are not present in Permian beds. In this portion of the Geren− navár Formation, the faunal renewal is close to 54%. · 26 species are found in the upper part of the microbialites, including 17, which are not present in the Permian beds. This indicates a relative faunal renewal of about 47% for the upper part of the section. The curve illustrating the variations of ostracod diversity (both species richness and abundance) can be divided into successive peaks (P) and drops (D) (Fig. 20A): · The maximum of species diversity is recorded at the base of the section by the assemblages 08BAN47 and 08BAN48 (P1: 28 species in 08BAN47). · Ostracods are absent from 08BAN49.1, 49.2, 49.3 (D1). · Assemblages from 08BAN50 to 08BAN60 show a slight rediversification, reaching medium values of species rich− ness (P2). · Above the BSB, assemblages from the base of the Geren− navár Formation (08BAN61 to 08BAN63) are character− ised by an increase of the species richness and abundance (P3). The maximal abundance is reached in 08BAN63 (2135 specimens). · The assemblages 08BAN64 and 08BAN65 record a re− duction of diversity (D3). · A slight rediversification is recorded at the top of the sam− pled part of the Gerennavár Formation (P4: 08BAN66 to 08BAN69). The samples 08BAN49.1, 49.2, 49.3, 57, 58, 68 yielded no ostracods. This absence of ostracods can be explained by (i) their initial absence at the time of deposition, implying a harsh environment, (ii) the displacement of the carapaces by current activity, (iii) the non−preservation or dissolution of carapaces. However, because most of the specimens recovered are closed carapaces, we assume the transportation to be limited (Oertli 1971). Note that instars (larval stages) are present for many species which also indicates, more strongly, that these taxa are in−situ (e.g., Boomer et al. 2003). Moreover, according to re− cent comparative analysis of ostracod extraction protocols, the method used in our study has limited dissolution effects on microfossils (Rodrigues et al. 2012). To conclude, the non− productive samples come from levels of homogeneous litho− logies, where surrounding samples are productive. The ab− sence of ostracods seems not to be linked to facies particulari− ties. Therefore, we consider the assemblages observed as auto− chthonous and relatively representative of the biocoenosis.
A study of the Permian-Triassic interval at the Bálvány North section ) shows a decreasing trend of abundances throughout the sampled part of the Nagyvisnyó Formation on the basis of bioclasts in thin sections (Fig. 20B). This trend and the concurrent disappearance of taxa was considered by Haas et al. (2006) to be directly related to the end−Permian events and not to local facies changes. This hypothesis can be tested by the present ostracod diversity re− cord.
The highest species richness of ostracod assemblages is recorded at the base of the section (sample 08BAN47, P1) and is correlated with the maximum of bioclast abundance. The ostracod data following upsection show lower species richness and a very low abundance, in accordance with the bioclast abundance pattern. However, this trend of diversity reduction is gradual for bioclast abundance whereas it is sharp for the ostracod diversity and followed upsection by relatively stable values. We therefore conclude that the bio− clast abundance pattern does not reflect the real dynamics of extinction of different organisms. Another significant discrepancy is recorded at the base of the Gerennavár Formation where bioclasts are rare but ostra− cod assemblages more abundant (highest abundance in P3, sample 08BAN63) and diversified than those of the Nagy− visnyó Formation. Haas et al. (2006) indicated that packstone laminae with fine fragments of calcareous algae and other bioclasts occur in the stromatolites and interpreted them as storm deposits. In addition to the fact that ostracods are found as closed carapaces, they are found in the entire sampled part of the Gerennavár Formation (except in 08BAN68, Fig. 2A,  B). We therefore deduce that their presence does not result from storm current activity but reflects more suitable environ− mental conditions. An analysis of the composition of assem− blages documents their relative homogeneity with respect to superfamilies and families in the entire microbialite succes− sion (Forel et al. 2013). Thus, the abundance pattern at the base of the stromatolites of the Bálvány North section is sug− gested to be due to special environmental conditions. The ostracods we describe here through the Permian-Tri− assic interval in the Bálvány North section show 2 unusual features: (i) high abundance and diversity, in contrast with ear− lier studies of ostracod faunas at the very base of the Gries− bachian which all document impoverished faunas (e.g., South China, Crasquin and Kershaw 2005;Crasquin−Soleau et al. 2006;Crasquin et al. 2010a;Forel and Crasquin 2011b;Tibet: Forel and Crasquin 2011a;Forel et al. 2011;Iran: Mette 2008); (ii) a lower extinction rate than the ostracods at Mei− shan section Forel et al. 2011 (Zhang et al. 2007) of poorly oxygenated to anoxic setting (Wignall and Hallam 1993). This oxygen defi− ciency in a basinal environment is supported by the absence of ostracods in the Induan of the West Pingdingshan section (Chaohu, Anhui Province; Tong et al. 2005). The two first au− thors of this paper processed more than one hundred and fifty samples from the Griesbachian-Dienerian part of the Yinkeng Formation and no ostracods were discovered. Lower Triassic beds in Meishan were accumulated at greater water depth than the Lower Triassic beds at the the Bálvány North section (see below for details). These discrepancies between ostracod re− cords are related to more hospitable conditions in Bálvány North provided by the microbial ecosystem from which the microbialites originate. These refuges are documented from several localities and may have prevented the oxygen deple− tion reaching levels lethal for ostracods (Forel et al. 2013). Crasquin et al. (2010a) also suggested that the ostracod diversity variations in the Upper Permian of Meishan may be due to the changes in stability/instability of environments as result of sea level changes. Accordingly, increase is linked to stable conditions during high sea−level. Forel and Crasquin (2011b) refined this one−parameter model with data from the Lower Triassic of Meishan. According to Haas et al. (2004), the complete succession at the Bálvány North section was ac− cumulated in an outer ramp setting with a deepening upward trend (Hips and Haas 2009). The important drop in ostracod diversity at the transition from interval P1 to D1 (level 1) could reflect the onset of instability, but the lithofacies do not record such environmental modifications. Stable environ− mental conditions due to great water depth are recorded at level 4 of the Nagyvisnyó Formation (Hips and Pelikan 2002;Haas et al. 2004Haas et al. , 2006. This environmental stability is indicated by the lack of distinct ostracod diversity changes, which agrees with the suggestion of Crasquin et al. (2010a). The diversity increase at the base of the Gerennavár Forma− tion could be related in first approximation to the stability of the subtidal environment. However, because of the refuge conditions provided by Lower Triassic microbial mats, it seems unlikely to relate ostracod diversity variations within the microbialites to stability conditions. Indeed, the onset of microbial growth marks the transition between: (i) ostracod faunas mainly forced by external parameters (stability/insta− bility related to hydrodynamism and oxygenation) in the Up− per Permian Nagyvisnyó Formation; and (ii) faunas whose dynamic is entirely linked to internal parameters associated with the microbial ecosystem functioning (food and oxygen supply by microbes) in the Lower Triassic Gerennavar For− mation.
The ostracods of Bükk Mountains (Hungary) present a special behaviour in the general landscape of the PTB inter− val. Indeed, they have here the lowest specific extinction rate (74%) of all the reference sections studied for ostracod anal− ysis (Crasquin and Forel in press). Furthermore, the fauna exhibits an important endemism. Only 13 species are com− mon with other areas. These palaeobiogeographical particu− larities of the Hungarian section will be considered in a forth− coming work. improve our manuscript. This work is part of the IGCP 572 "Restora− tion of Marine Ecosystems following the Permian-Triassic Mass Ex− tinction: lessons for the present". The field work was carried out with the support of 2011 MNHN ATM "Biodiversité actuelle et fossile. Cri− ses, stress, restaurations et panchronisme: le message systématique".