The Palaeogeography of Lake Pannon During Deposition of the Congeria rhomboidea Beds

Lake Pannon covered the area of the Pannonian Basin during the Late Miocene. Because of its wide-range distribution, economic importance and unique, mostly “caspi-brackish” endemic biota, its lacustrine deposits have attracted much scientific attention in the past hundred years. According to the latest stratigraphic division accepted for the area and bioprovince of the Western or Central Paratethys these deposits belong to the PannonThe Palaeogeography of Lake Pannon During Deposition of the Congeria rhomboidea Beds


INTRODUCTION
Lake Pannon covered the area of the Pannonian Basin during the Late Miocene.Because of its wide-range distribution, economic importance and unique, mostly "caspi-brackish" endemic biota, its lacustrine deposits have attracted much scientific attention in the past hundred years.According to the latest stratigraphic division accepted for the area and bioprovince of the Western or Central Paratethys these deposits belong to the Pannon- The Palaeogeography of Lake Pannon During Deposition of the Congeria rhomboidea Beds Sándor GULYÁS Jr.
Palaeogeographic and palaeoecologic studies started with the analysis of superficial and shallow borehole lacustrine deposits in the Transdanubian region of Hungary at the beginning of the century (H A L A V Á T S , 1902;LO ˝RENTHEY, 1905).From the 1930's onwards the inner parts of the basin were also studied because of the initiation and development of hydrocarbon prospecting in the areas of the Great Hungarian Plain.
The reliability of any palaeogeographic reconstruction largely depends on the accuracy of the applied stratigraphy.The older palaeogeographical reconstructions were only undertaken on small parts of the lake on the basis of lithostratigraphic and biostratigraphic data.These however, as well as the synthetic reconstructions for larger areas were by no means accurate, as the stratigraphic concept and the sedimentological model underlying them were not without certain ambiguities.According to the traditional Hungarian model based on lithostratigraphic and biostratigraphic division of the lacustrine deposits, the slow deposition of the suspended load from the water of the lake was responsible for the gradual infilling of the lacustrine basin.
Based on abrupt changes in the lithology and the fossil record the age of the lower finer-grained, deepwater deposits was determined to be "Lower Pannonian", while the age of the upper shallow-water, sandy deposits were defined to be "Upper Pannonian" (STRAUSZ, 1941;SZÉLES, 1971) (Figs. 1 and 2).Thus the palaeogeographic maps based on this model were in fact facies maps.
Halaváts (in H A L A V Á T S, 1892) referred to an acceptable sedimentation model by noticing the following as early as the end of the last century: "The Congeria rhomboidea horizon indispensable for the accurate determination of the Pontian stage has been described so far only from the southern parts of Hungary (Austro-Hungary).According to data presented in the collection of the Hungarian Royal Geological Institute and from the literature, localities of this horizon are known only from the Krassó-Szörény Mountains (Muntii Semencului Transylvania) and the hills surrounding the Mecsek Mountains and the city of

Zagreb. This horizon is totally missing in other parts of the country."
Since the younger Pontian deposits are restricted to the southern parts of former Hungary, the lake must have covered only this part of the country and not the whole Pannonian Basin during the Pontian as the traditional model assumed.However, the recognition of the importance of delta sedimentation in the south-eastern part of the Great Hungarian Plain only brought changes in the traditional sedimentological, stratigraphical and palaeogeographical thinking in the 1970's (MUCSI & RÉVÉSZ, 1975;MAGYAR & RÉVÉSZ, 1976;RÉ-VÉSZ, 1980;GAJDOS et al., 1983;BÉRCZI & PHIL-LIPS, 1985).The geometry of progradational delta bodies with the particular facies and subfacies could be clearly identified on the seismic profiles from the inner basin areas (POGÁCSÁS, 1984;POGÁCSÁS & RÉ-VÉSZ, 1987;MATTICK et al., 1988;JUHÁSZ, 1991) (Figs. 3 and 4).
In the past few decades the Hungarian magnetostratigraphic research has undergone significant development (RÓNAY & SZEMETHY, 1979;COOKE et al. , 1979;ELSTON et al., 1990), and today a large number of data are readily available for purposes of integrated r e s e a r c h (BALOGH & RAKOVITS, 1976;BALOGH & JÁMBOR, 1987;BALÁZS & NUSSZER, 1987).The application of these data onto the seismic profiles shows that Lake Pannon formations are diachronous ( POGÁCSÁS et al., 1988POGÁCSÁS et al., , 1989POGÁCSÁS et al., , 1990;;VAKARCS e t al., 1994).Korpásné Hódi (in KORPASNÉ HÓDI et al., 1992) was the first to analyze correlations between the newly defined lithofacies and the occurring biofacies for the area of the Danube-Tisza Interfluve.JUHÁSZ & MAG-YAR (1992) extended this type of study to the Tiszántúl area, and suggest that it is possible to characterize certain depositional and deltaic environments by given molluscan assemblages marked by dominant, wellidentifiable species (Fig. 5).In this way the biostratigraphic horizons were distinguishable from the biofacies existing in the lake.It also facilitated the chronostratigraphic division of lacustrine deposits in a certain environment or facies by means of analysis of evolu-tionary lineages of molluscan species (MAGYAR et al., 1999; see Fig. 6).
With the help of these research results it became more and more apparent, that despite the frequent fluctuations in the water level, the infilling of the basin was mainly a uni-directional process.Prograding deltas from the margins resulted in an approximately southsoutheast migration of the northern shorelines, and therefore a gradual southward migration of certain facies or depositional environments.As individual environments can be characterized by given molluscan assemblages, the geographical distribution of some molluscan species found in these environments will give a more or less accurate view of the changes in the lake's extension through time.As with the gradual S-  SE shoaling of the lake and the migration of the facies, the younger forms of the molluscan species were also restricted to the southern areas of the lacustrine basin., 1994).The aim of this study was to map the palaeoshoreline of Lake Pannon for a given moment as accurately as possible using palaeobiogeographic data, and to test the underlying new sedimentary model for the basin.
The bivalve species Congeria rhomboidea M .H Ö R N E S occurs widely in the Upper Pannonian (Pontian sensu STEVANOVI∆) deposits both in Hungary and in the neighbouring countries of Croatia, Bosnia-Herzegovina, Serbia and Romania as well, which makes it suitable for such investigation (Fig. 8).HALAVÁTS (1892) acknowledged its primary function for age determination when in his own words he "nominated this well-traceable geological horizon after one of its permanently occuring molluscan forms C o ngeria rhomboidea h o r i z o n " .He placed the age of the horizon into the upper parts of the then accepted Upper Pontian stage.
Several theories concerning the origin and the evolutionary lineage of this species have come to light since the second half of the 19th century.Most authors were looking for relative forms of this species outside the Pannonian Basin (ANDRUSOV, 1897(ANDRUSOV, , 1909;;HOERNES, 1901;WENZ, 1942).StevanoviÊ was the first to point to a very important factor in the palaeogeography of Lake Pannon and the Paratethys to underlie his phylogenetic concept of Congeria rhomboidea M. HÖRNES (STEVANOVI∆, 1961(STEVANOVI∆, , 1978)).
The Pannonian basin evolved during the Miocene period.The Late Miocene belongs to the post-rift phase of basin evolution accompanied by isostatic thermal subsidence due to the thinning and cooling of the lithosphere following the rifting process (K Á Z M É R, 1990; R O Y D E N et al., 1983).During the Miocene, connection of the Central Paratethys with the Mediterranean and the Indo-Pacific Ocean was repeatedly opened and closed (RÖGL & STEININGER, 1983).The final isolation of the Central Paratethys commenced in the beginning of the Late Miocene (STEININGER et al., 1988).The Attic phase of the Alpine orogeny in the Southern Carpathian Mountains isolated the water masses covering the Pannonian Basin from the rest of Parathetys from as early as the Bessarabian until the beginning of the Pontian (13.5-8.5 my), i.e. during the whole Pannonian.In the Sarmatian this resulted in the creation of miohaline, "brackish" and later on in the Pannonian and Pontian oligohaline, "caspi-brackish" conditions in the lake.These factors induced the evolution of an endemic molluscan biota bearing unique features.During the Early Pontian, the connection between the Pannonian and Dacian Basins was restored by means of a waterway formed near the present Iron Gate: the -erdap or Porta Ferra Strait.Across this strait the eurytropic species, which had evolved in the Pannonian Basin could have migrated eastward into the Dacian-Euxinic Basins.Consequently, this species emerged and developed in the Pannonian Basin and migrated eastward following the restoration of the connections of waterways at a later stage (STEVANOVI∆, 1961(STEVANOVI∆, , 1978)).

The evolution of Congeria rhomboidea M. HÖR-N E S is now relatively well-understood. The gradual evolution of Congeria rhomboidea M. HÖRNES f r o m Congeria zsigmondyi and a relation with C o n g e r i a p a r t s c h i seems to be well-traceable in the Pannonian
Basin through a number of transitional forms with gradual changes in the shell morphology.From the bivalve species Congeria zsigmondyi or a certain type of it ("semiptera") a strongly convex form with a high shell develops ("praerhomboidea").Then on this form a strong second rib appears ("dubocaensis"), and finally the size of the shell increases ("rhomboidea") reaching an extreme rate in the final stage of development ("dilatata") (Fig. 7).This gradual change in the shell's morphology occurs in such a way that there is an extraordinary variety at all levels, so the names represent only certain stages of the evolution.Because Congeria rhomboidea M. HÖRNES exhibits an almost continuous development with hardly distinguishable stages or forms, it is very hard to define the area of its distribution in the lake without knowing how to define the species itself.
In this study the form "dubocaensis" is considered to be the very first representative of the species Congeria rhomboidea M. HÖRNES on the basis of the morphological feature noticed even by LO ˝R E N T H E Y (1905) while he was tracing the lineage between C. triangularis and C. rhomboidea; i.e. the appearance of the second strong rib on the shell to define the species.
According to magnetostratigraphic data this bivalve species appeared in the lake approximately 8.5 million years ago (Fig. 6).Thus its geographical distribution will give us a more or less accurate view about the extension of the lake at this time.
A distributional database has been set up from 77 pieces of literature dealing with the Pannonian and the Pontian.Additional data comes from collections at the University of Szeged (JATE), University of Budapest (ELTE), the Hungarian Geological Institute and the University of Debrecen (KLTE) as well as from the Hungarian Oil and Gas Company (MOL Rt.) (see GULYÁS, 1998 for details).Data had been applied to maps with a scale of 1:600,000, which were reduced afterwards.
The geographical distributions of the bivalve species Congeria praerhomboidea S T E V A N O V I ∆ and the representatives of the Prosodacnomya genera have also been applied onto maps as a complementary and verificatory check, as Congeria praerhomboidea S T E V A-N O V I ∆ is the direct ancestor of Congeria rhomboidea M. HÖRNES.According to the delta progradation model the geographical distribution of this species was expected to be restricted northward of that of C o n g e r i a r h o m b o i d e a M. HÖRNES.Congeria rhomboidea M .H Ö R N E S comes primarily from sublittoral depositsclay and silt.Thus its distribution will mark the borderline of the sublittoral facies in the lake.In addition the distribution of the representatives of the P r o s o d a c n om y a genus of the same age coming from littoral deposits will show the incidence of the littoral facies and the gradually freshening lagoons at the same time.

RESULTS AND DISCUSSION
The constructed palaeobiogeographic map (Fig. 8) displays the geographical distributions of Congeria rhomb o i d e a M. HÖRNES and Congeria praerhomboidea STEVANOVI∆ with 192 data points as well as the distribution of localities with representatives of the Prosodacnomya genus (310 data points).The larger signs indicate general localities like "the northern slope of Majevica Mt." for example.
The scarcity of dots on the map in certain areas is due to lack of information because of an insufficient number of boreholes in the area, or the lack of a continuous data record on sublittoral facies deposits, rather than real rarity of the species.Conversely, in other areas where many dots are displayed on the map, this does not necessarily imply that more animals must have existed.Rather these dots indicate the distribution of localities with surface outcrops, where there is easy access to these fossil-bearing layers.Thus lots of data can be found in these areas.
The created palaeobiogeographic map records the outcome of complex processes, which had been going on for millions of years in the lake, and do not represent one single moment in the lake's history.As the northern shoreline migrated to the S-SE with the continuos shoaling of the lake, the littoral and sublittoral facies shifted with the passage of time.Consequently, for example the specimen of the Prosodacnomya g e n u s found at Lake Balaton might be even millions of years older than the one from Battonya, South-east Hungary.In order to draw the boundaries of a given facies marked by the geographical distributions of specimens of Congeria rhomboidea M. HÖRNES or the P r o s od a c n o m y a genus, the maximal area of their distribution had to be considered by taking into account their marginal occurrences.These will more or less accurately give the boundary of the facies at that time, when this species or genus first appeared in the lake.the combinations of these two maps a new palaeogeographic map has been set up with the palaeoshorelines and the roughly estimated boundaries of the two facies in the lake for a given moment (Fig. 11).
On the basis of data presented on the map a water coverage of roughly 75,000 km 2 has been calculated.The path of the lake's northern shoreline seems to correspond to the line of the 8.2 million year chronoboundary on the map prepared by VAKARCS et al. (1994) on the basis of seismic and magnetostratigraphic data.Therefore the results of the two methods seem to correspond well, underlying the accuracy of the new delta progradational model.However, the accuracy of the age determination of VAKARCS et (1994) on their map remain questionable (8.2 mya vs. 8.5 mya).and 9).These imply that there were no significant and permanent changes in the position of the shoreline in these south- The water coverage assumed by the presence of littoral P r o s o d a c n o m y a s in the alluvial plain areas near Sopron, and the Kisalföld, Northwestern Hungary are highly questionable (Fig. 10 -dotted line area), and this region was not considered to be a part of the lake's area on the final map (Fig. 11).Fossils coming from these areas are very hard to determine due to poor preservation, and the majority of them are lost.Specimens collected by István Vitális can be found in the Hungarian Natural History Museum, though the single P r o s o d a c -n o m y a from Sopron is missing.There are three major ways of interpreting these data.First it could be assumed that all the fossil descriptions are incorrect, so there were no P r o s o d a c n o m y a s existing in this area.Alternately, the lake managed to occupy the alluvial plain area of the Kisalföld for a very short time, so it seems to be acceptable to have some localities with Prosodacnomya there.Or perhaps the P r o s o d a c n o m y a was capable of adapting to fresh-water alluvial conditions, so its areal distribution did not correspond to the changes in the lake's water coverage.Further investigation is required to clarify this problem.

CONCLUSION
Results of this palaeobiogeographic study seem to corroborate sedimentation hypotheses based on seismic, magnetostratigraphic and sedimentological data for Lake Pannon, implying a clear S-SE migration of the northern palaeoshorelines and depositional environments.This indicates two major applications of palaeontologic and palaeobiogeographic methods.First, the utilization of an integrated multidisciplinary approach, embedding palaeontological data, gives more reliable and accurate results.It also seems to be the only relevant tool in settling the controversies surrounding the stratigraphic subdivison and correlation of Lake Pannon deposits.
Secondly, the majority of petroleum displays are connected to certain depositional environments.Studies such as this may help petroleum exploration by tracing the migration and geographic distribution of depositional environments significant from the point of reservoir and source potential through time in a basin on a purely palaeontological, palaeoecological basis.
The bivalve species Congeria rhomboidea M. HÖRNES o c c u r s widely in the Upper Pannonian (Pontian sensu StevanoviÊ) deposits of Hungary and the neighbouring countries.Its evolution is relatively well understood.According to magnetostratigraphic data this species appeared in the lake 8.5 mya.According to the maximal geographical distributions of C. rhomboidea and its ancestor Congeria praerhomb o i d e a ST E V A N O V I∆ occurring in sublittoral clay and silt along with the representatives of Prosodacnomya coming from littoral and lagoon deposits of the same age the estimated water coverage was around 75,000 km 2 at the time of first emergence of C. rhomboidea in the lake.In the north the distribution of C. praerhomboidea is strictly restricted to the north of that of C. rhomboidea, its descendent, implying a clear S-SE trend in the migration of the lake's northern palaeoshoreline.Distributions of the littoral P r o s o d a c n o m y a s in relation to the sublittoral C. rhomboidea of the same age display a similar pattern.Meanwhile the western and southern palaeoshorelines underwent only minor fluctuations.

Fig. 1
Fig. 1 Chronostratigraphic chart for the areas of the Central and Eastern Paratethys (after S T E I-NINGER et al., 1990).

Fig. 2
Fig.2The sedimentation models for Lake Pannon(GULYÁS, 1998).Deep-water deposits are marked with black and dark grey, while the shallow-water deposits with light grey and white.The differences in the hue indicate changes in the ages of these deposits (lighter ones are younger).It is clearly observable that the isochronous surfaces marking the changes in the surface with time are inclining towards the center of the basin (2 -delta progradational model).In contrast with the traditional model (1), according to which the abrupt changes in the fossil record and the lithological features mark chronohorizons this type of aggradation results in a gradual movement of the facies from the margins towards the center of the basin (2).

Fig. 3
Fig.3General configuration of seismic sequences and supersequences in the Pannonian Basin of Hungary(after MATTICK et al., 1988).
2. MATERIALS AND METHODPalaeogeographic maps based on the new sedimentary model have been based only on magnetostratigraphic and seismic data so far, and only for the area of presentday Hungary ( PO G Á C S Á S et al., 1993; VAKARCS e t al.

Fig. 4
Fig. 4 Idealized view of prograding delta sedimentation and the disribution of sedimentary environments from the SE part of the Pannonian Basin (after B É R C Z I & PHILLIPS, 1985).

Fig. 6
Fig. 6 Integrated correlation chart for the Lake Pannon deposits (after MAGYAR et al., 1999).Note that according to the chart the base of the Pontian (8.5 mya) is 1-1.5 my.older in the area of Lake Pannon than in the Eastern Parathetys.
The littoral and lagoonal zones marked by the P r os o d a c n o m y a are mainly restricted to the northern areas of the lake situated roughly north of the line of Hrvatsko Zagorje (Northern Croatia) -Mecsek Mts.(Southwest-Hungary) -Paks (on the bank of the Danube in South-Hungary) -Egyek (North-eastern Hungary) (Fig. 11).At the same time in the southern Slovenian, Croat-ian, Bosnian and Serbian parts sublittoral and deepwater environments existed in the lake.In some parts in the centre of the lake and on the south-eastern and south-western marginal areas Congeria rhomboidea M .H Ö R N E S occurs together with P r o s o d a c n o m y a in the localities of BeoËin, Serbia (STEVANOVI∆ et al., 1990), an area between the rivers of Sava and Majevica, Northern Bosnia (STEVANOVI∆ et al., 1990), Hrvatsko zagorje, Croatia (BASCH & AEAGAR-SAKA», 1992) (Figs. 9 and 10).Congeria rhomboidea M. HÖR-N E S appears in a lower position in sublittoral deposits in the sequences, while the representatives of P r o s odacnomya occur in the overlying younger littoral deposits.Similarly in the southern areas -Posavo-Tamnava, Kolubara Basin, Mislodjin, Bodarevac, BuËje and Beo-Ëin, Southern Serbia (S T E V A N O V I ∆, 1951; S T E V A-NOVI∆ et al., 1990) -the older Congeria praerhomboidea S T E V A N O V I ∆ and the younger Congeria rhomboidea M. HÖRNES occur in the same localities, but in different stratigraphic levels (Figs. 8

Fig. 9
Fig. 9 Maximum areas of distribution for Congeria rhomboidea M. HÖRNES and Congeria praerhomboidea STEVANOVI∆.Note that on the northern part of the Pannonian basin the distribution of Congeria praerhomboidea STEVANOVI∆ is restricted to the north of the maximum distribution area of Congeria rhomboidea M. HÖRNES, indicating a continuous unidirectional S-SE migration of the palaeoshoreline there.Meanwhile on the southern part the two areas tend to overlap at some places indicating only minor fluctuations in the position of the shoreline.

Fig. 10
Fig. 10 Maximum area of distribution of the P r o s o d a c n o m y a genus.Dotted line indicates uncertainty in the water coverage for the area of Kisalfld.