The Upper Jurassic of Europe : its subdivision and correlation

In the last 40 years, the stratigraphy of the Upper Jurassic of Europe has received much attention and considerable revision; much of the impetus behind this endeavour has stemmed from the work of the International Subcommission on Jurassic Stratigraphy. The Upper Jurassic Series consists of three stages, the Oxfordian, Kimmeridgian and Tithonian which are further subdivided into substages, zones and subzones, primarily on the basis of ammonites. Regional variations between the Mediterranean, Submediterranean and Subboreal provinces are discussed and correlation possibilities indicated. The durations of the Oxfordian, Kimmeridgian and Tithonian Stages are reported to have been 5.3, 3.4 and 6.5 Ma, respectively. This review of the present status of Upper Jurassic stratigraphy aids identification of a number of problems of subdivision and definition of Upper Jurassic stages; in particular these include correlation of the base of the Kimmeridgian and the top of the Tithonian between Submediterranean and Subboreal Europe. Although still primarily based on ammonite stratigraphy, subdivision of the Upper Jurassic is increasingly being refined by the incorporation of other fossil groups; these include both megafossils, such as aptychi, belemnites, bivalves, gastropods, brachiopods, echinoderms, corals, sponges and vertebrates, and microfossils such as foraminifera, radiolaria, ciliata, ostracodes, dinoflagellates, calcareous nannofossils, charophyaceae, dasycladaceae, spores and pollen. Important future developments will depend on the detailed integration of these disparate biostratigraphic data and their precise combination with the abundant new data from sequence stratigraphy, utilising the high degree of stratigraphic resolution offered by certain groups of fossils. This article also contains some notes on the recent results of magnetostratigraphy and sequence chronostratigraphy.

The term 'Upper Jurassic' ('Oberer Jura') was introduced by von Buch (1839).Arkell (1956) revived this name with only minor changes in its chronostratigraphic content.The term 'Upper Jurassic' in the sense of Arkell (1956) was accepted by the First and Second 'Colloque du Jurassique ' at Luxembourg in 1962 and1967; only the stage name 'Purbeckian' was eliminated, as it was considered to characterise merely a distinct lithofacies.This usage was followed by the five subsequent International Symposia on Jurassic Stratigraphy at Erlangen in 1984, Lisbon in 1987, Poitiers in 1991, Mendoza in 1994and Vancouver in 1998.Focus on the formal stratigraphic subdivision of the Jurassic, and the Upper Jurassic in particular, is reflected in the series of key meetings since the early 1960's (Table 1).
The term 'Malm' was included in the recommendations of the First Luxembourg Colloquium in 1962 as an alternative term for the 'Upper Jurassic' (Maubeuge 1964).Although this term, like the term 'Tithonian' (see below), is not based on a geographical site, it has been widely used since its introduction by Oppel (1858Oppel ( , 1865)).Referring to the Tithonian stage, Arkell (1956, p. 8) wrote: "it is too late to abolish it after a hundred years of continuous use"; this also applies to the term 'Malm'.It is important to note that both 'Upper Jurassic' and 'Malm' are chronostratigraphic terms; the latter, in particular, has frequently been used in a lithostratigraphic sense by some authors.
At the First Colloquium in Luxembourg in 1962, a subdivision of the Jurassic System into stages was proposed, the basic framework of which has survived to the present day.The stages were defined by their lower and upper ammonite zones.The recommendations of the First Colloquium (Maubeuge 1964) were thus a landmark in the history of international agreements concerning the subdivision of the Jurassic System into series and stages.
After a period of discussion following the publication of the resolutions of the Luxembourg Colloquia (Maubeuge 1964(Maubeuge , 1970)), these proposals have been accepted worldwide, the only exception being that in the former Soviet Union the Callovian has been considered to belong to the Upper Jurassic (see Krymholts et al. 1988), while in the rest of the world the Callovian is included in the Middle Jurassic.However, following a decision by the Interdepartmental Stratigraphic Commitee in 1989, the Callovian is also now considered in Russia to belong to the Middle Jurassic (Zhamojda 1991).
According to the most recent estimates, the Late Jurassic had a duration of a little more than 15 million years according to Gradstein et al. (1994Gradstein et al. ( , 1995;;Ogg 1995), or 12 million years (+5.6/-7.3)according to Pálfy et al. (1998).The data of Gradstein et al. (1995) have been used in Figures 2, 4 and 5 of this paper; on this basis each of the three Upper Jurassic stages has an average length of 5 million years, while zones and subzones have approximate durations of 700 000 and 300 000 years, respectively.Each subzone comprises at least three horizons, each of which has an approximate duration of 100 000 years.

Subdivision and definition of stages: status and unsolved problems
On the basis of the recommendations of the First Luxembourg Colloquium (Maubeuge 1964), the Upper Jurassic Series was subdivided into four stages for the Boreal and Subboreal regions: Oxfordian, Kimmeridgian (sensu anglico), Portlandian (sensu anglico) and Volgian, and three for the Submediterranean and Mediterranean regions: Oxfordian, Kimmeridgian (sensu gallico equivalent to 'Crussolian') and Tithonian (equivalent to 'Danubian' and 'Ardescian').In the following decades, there has been much confusion as the Kimmeridgian and Portlandian Stages have often been used differently in different parts of Europe.
In 1990, a formal vote of the International Subcommission on Jurassic Stratigraphy (ISJS) led to the decision to use stages with approximately the same vertical age ranges and uniform names in both regions: Kimmeridgian (sensu gallico) and Tithonian (Zeiss 1991a).With regard to the still unresolved correlation problems between the Boreal and Mediterranean provinces, it was agreed that the Volgian can be used as an alternative stage for the Tithonian in Subboreal and Boreal regions (Fig. 1).
The main problem which remained to be solved was the definition of the lower boundary of each stage.To date, no stage has a type locality and a defined lower boundary (global stratotype section and point, or GSSP) formally accepted by the International Commission on Stratigraphy (ICS).There are of course a lot of proposals, but they have not been validated according to the guidelines and rules of the ICS (Cowie et al. 1986;Remane et al. 1996).
The most intractable problems are to find isochronous levels in Submediterranean (and Mediterranean) and Subboreal (and Boreal) Europe for the lower boundary of the Kimmeridgian Stage and for the upper boundary of the Tithonian (Volgian) Stage.As the former is of particular significance for Upper Jurassic subdivision and correlation, it will be treated here in some detail (see below).
Other problems are the unification of the differing subdivisions of stages into substages, correlation of the zones of each stage between the different areas of Europe and the development of better correlation charts from the Boreal regions to the Mediterranean areas.Provisional correlation charts on a zonal and subzonal level for each stage of the European Upper Jurassic are presented here (see Figs 2-5).Zones and subzones are used here as chronozones following the International Stratigraphic Guide (Salvador 1994); originally, many of them were defined as biozones whereas others were used as standard zones, standard chronozones or biochronological standard zones, i.e.only the base is defined while the top is defined by the base of the next overlying unit (Callomon 1965(Callomon , 1984a(Callomon , 1994)).Problems arise, however, due to inconsistent usage of the term 'standard zone'.In Northwest Europe, standard zones are mostly used following the concept of Callomon (1994), whereas in central and southern Europe, standard zones are often synonymous with biostratigraphic zones for use in biochronology (Cariou & Hantzperque 1997).It is often difficult, therefore, to determine in which meaning 'standard zones' are used.The problems encountered in moving from biostratigraphic field data to biochronological interpretations have been discussed recently by Remane (1991).
The term biochronological zone is now used by many authors instead of chronostratigraphic zone, if the zone is based on fossil data.As the ultimate subdivision of biochronology, French authors use the term 'biohorizon' (e.g.Enay 1997); their concept is therefore 'sensiblement different' from the pure biostratigraphic horizon concept of J.H. Callomon (Dommergues 1997).
No attempt has been made here to correlate ammonite faunal horizons due to the variable nature of the published research on the Upper Jurassic in the various sedimentary basins of Europe.Where necessary, however, correlations of horizons are discussed in the text.
For an example of such horizon correlations, the reader is referred to the work of Callomon (1984c) on the Upper Jurassic of North America (this study also covers the European Amoeboceras subdivision).Although attempts have been made to generalise horizons for the whole 'domaine tethysien' and 'domaine boréal' (Cariou et al. 1997;Hantzperque et al. 1997), these appear premature and of little practical use, given the present state of knowledge.
The 'faunal horizon' approach clearly represents a method for increasing precision in correlation and dating in the future, when the data from the various sedimentary basins reach the necessary standard.It is already proving useful in deciphering the history of basin deposits at a resolution that was hitherto impossible; such data, in particular, allow us to date events more precisely and to determine the 'completeness' of the sedimentary record i.e. to identify accurately the position and duration of hiatuses.A prerequisite is, however, that it is possible to reconstruct the complete succession of faunal horizons by correlating individual local successions.
The Upper Jurassic (Malm) Series (Fig. 1) In this paper, subdivision and correlation of the Upper Jurassic Series have been carried out mainly using ammonites.Other fossil groups are reviewed briefly, however, with request to their biochronologic resolution and correlation potential.Many papers have been published on Upper Jurassic ammonites and their chronostratigraphic resolution (see detailed discussion below).For a broad overview, the reader is referred to the papers of Cariou et al. (1997), Geyssant (1997) and Hantzpergue et al. (1997) for western Europe and the Mediterranean.Other important, partly regional compilations and revisions have been published by Sapunov (1979), Donovan et al. (1981), Krymholts et al. (1988), Malinowska et al. (1988), Enay et al. (1994) and Schlegelmilch (1994).

Lower boundary (Middle-Upper Jurassic Series boundary)
The lower boundary of the Oxfordian Stage is rather well-defined by ammonite zones and subzones and only requires more precise definition with respect to the lowermost faunal horizon, which then would characterise the beginning of the lowermost subzone (and zone) of the stage.Furthermore, it appears that the lower boundary is approximately (on a subzonal level) the same in Boreal and Mediterranean areas.Once the type faunal horizon has been chosen, then the problem of the type locality for the boundary will also have been solved.At present, this boundary lies in France between the uppermost horizon of the Quenstedtoceras lamberti Zone of the Upper Callovian Substage (the Cardioceras paucicostatum horizon) and the lowermost horizon of the Q. mariae Zone; this was first named in France after Peltoceratoides elisabethae (Fortwengler & Marchand 1994a), but afterwards was changed to Hecticoceras (Brightia) thuouxense (Fortwengler & Marchand 1994b, c), a species described only recently (Fortwengler et al. 1997).In Dorset, however, Cardioceras cf.woodhamense and C. woodhamense are found in the lowermost levels of the Q. mariae Zone (Callomon & Cope 1996), whereas in north-west France, C. woodhamense has been collected only in the third horizon of the Q. mariae Zone (Vidier et al. 1993).In south-east France, this horizon is only recognised tentatively.These different faunal horizons all lie in the Cardioceras scarburgense Subzone, the lower subzone of the Q. mariae Zone, so that the age difference of these horizons (if any) should not be too large.A vote by the Callovian/Oxfordian Boundary Working Group in 1995 resulted in a preference for a type locality in south-east France, with the consequence that the Oxfordian would begin with the H.(B.) thuouxense horizon (see above), but a final decision was not taken (Meléndez 1995;Meléndez et al. 1998).

Upper boundary (Jurassic-Cretaceous System boundary)
In accordance with the decision of the ISJS (see above), there are two alternative stages for the uppermost part of the Jurassic System: Tithonian and Volgian.As they differ in duration, the boundary may be drawn at two different levels, i.e. there are two variants of the Jurassic-Cretaceous boundary.Accordingly, the mem-bers of the former Jurassic-Cretaceous Boundary Working Group agreed to work provisionally with two boundaries (Remane 1986;Remane et al. 1986;Zeiss 1986 (Rawson et al. 1978;Kejsi et al. 1988;Sey & Kalacheva 1993a).In the first case, the type locality should be best selected in south-eastern France, where the Ardescian Substage (Upper Tithonian) and the Berriasian Stage were originally described.Subsequent studies have revealed that the sequences are not complete at the base, however, so that it has been suggested that the best sections illustrating the Jurassic-Cretaceous boundary beds and their fauna are situated in southern Spain (Enay & Geyssant 1975;Tavera 1985;Tavera et al. 1994;Enay et al. 1998a, b).
In a recent review of the Berriasian Stage, Hoedemaker (1994) stated that the Jurassic-Cretaceous boundary is typically placed at one of two different levels, either at the base or at the top of the Jacobi Chronozone: "Investigators of Jurassic stratigraphy prefer the lower of these two boundaries, investigators of the Cretaceous stratigraphy the upper" (Hoedemaker 1994, p. 12).
At the same time, there has also been an attempt to trace the Jurassic-Cretaceous boundary based on geomagnetic anomalies from the Tethys to southern England (Ogg et al. 1994).In the Tethyan-Atlantic faunal realm, the top of magnetic polarity reversal M19r approximately coincides with the Tithonian-Berriasian boundary in the Mediterranean area.This reversal is difficult to place precisely in England, but it seems to be situated in the lowermost Purbeck beds.If so, it would demonstrate once again that the 'Upper Volgian' (Casey 1973) or 'Upper Portlandian' of England (Wimbledon 1980), i.e. the zonal sequence Subcraspedites primitivus -Subcraspedites lamplughi, overlaps with the Lower Berriasian.Wimbledon (1980) also included the 'Upper Volgian' zones of Casey (1973) in the 'Portlandian' of Britain, thus extending the stage upwards by three further zones (termed here 'Upper Portlandian').In a recent compilation chart, W.A. Wimbledon (in: Callomon & Cope 1996) correlated these 'Upper Portlandian' zones and the Upper Volgian zones of the Russian platform with parts of the Lower Berriasian.
In Poland, the Jurassic-Cretaceous boundary has been traced by joint studies of ostracodes and ammonites (Marek et al. 1989) whereby the Upper Tithonian and the lower part of the Lower Berriasian could be recognised as well as the Upper Berriasian (= 'Ryazanian').
In a recent paper (Marek & Shulgina 1996), the ammonites of the Berriasian (Ryazanian) were considered to belong to the interval upper occitanica -lower boissieri Zones.
In a recent development, the Interdepartmental Stratigraphic Committee of Russia (ISC) approved the following resolutions of its commissions on the Jurassic and Cretaceous Systems (Rostovtsev & Prozorovskiy 1997, p. 48).
"1.To draw the Jurassic-Cretaceous boundary in the Boreal Realm between the middle and upper substages of the Volgian, and not …… as …… earlier adopted in Russia (1978).This boundary mainly corresponds to the Tithonian/Berriasian boundary in Tethyan realm (Colloque Lyon-Neuchâtel, 1975).Correspondingly, the Lower Volgian in the whole correlated with the Lower and Middle Tithonian; the Middle Volgian, with the Upper Tithonian; the Upper Volgian, with two lower zones of the Berriasian (Jacobi/Grandis and Occitanica).
2. To transfer the Volgian Stage in its former range to the category of regional stratigraphic units (regional stage).To distinguish as chronostratigraphic units in the boundary part of the Jurassic and Cretaceous scale of Russia only Tithonian and Berriasian." These resolutions, which were precipitated by the work of Sey & Kalacheva (1993a), confirmed the earlier opinions of many authors concerning Upper Jurassic/Lower Cretaceous correlations.
It is clear that general consensus has not yet been reached; it is assumed, however, that the present Tithonian-Berriasian boundary is not suitable for global correlation.It may be preferable, therefore, to return to an old proposal: to define the Jurassic-Cretaceous boundary at the base of the B. boissieri Zone, where many guide fossils of different groups are available for correlation.Recent studies in the Caucasus area by Remane (1997) are supportive of this proposal.

Lower boundary
See discussion above.
this paper, following common usage.An example of the ongoing debate is the inclusion of the D. bifurcatus Zone in the Middle or Upper Oxfordian; this zone was introduced by Enay (1966) as the upper subzone of the G. transversarium Zone but was later considered as the lowermost zone of the Upper Oxfordian (Cariou et al. 1971).Preference is given here to a subdivision in which the D. bifurcatus Zone is included in the Middle Oxfordian (Fig. 1) as has also been proposed by Meléndez (1989), Cariou & Meléndez (1990), Cariou et al. (1991a) and Gygi (2000a) although not followed by Cariou et al. (1991bCariou et al. ( , 1997)).
While the lower and middle substages have the same lower boundaries in Submediterranean and Subboreal Europe, the position of the lower boundary of the upper substage differs.In Boreal Europe, it has been drawn at three different levels (Wright 1996a, fig. 6).The solution to draw it at the base of the A. glosense Zone is well-known (Sykes & Callomon 1979;Wright 1980); it would correspond to the base of the P. luciaeformis Subzone in the G. transversarium Zone, i.e. the boundary would be drawn around one and a half zones deeper than in the Submediterranean subdivision.It seems preferable to draw the boundary at the lower boundary of the A. rosenkrantzi Zone, corresponding approximately to the lower boundary of the Upper Oxfordian both in Submediterranean Europe (base of E. bimammatum Zone) and in Subboreal Europe (base of R. pseudocordata Zone), although the latter lies somewhat deeper (Matyja & Wierzbowski 1997, fig. 4).

Zones
The zonal and subzonal subdivision of the Lower Oxfordian Substage was established by Arkell (1941) using Quenstedtoceras mariae and Cardioceras cordatum as index species; it can be used over large areas of northern and central Europe (Fig. 1) and is also applicable in the Dauphinois basin of south-east France as recently demonstrated by Fortwengler & Marchand (1994a;Fortwengler et al. 1995).
From the base of the Middle Oxfordian, perisphinctids and peltoceratids become the dominant ammonite groups with respect to index fossils at the substage level in the Submediterranean and Subboreal provinces.The first aulacostephanids (Decipia) also appear at this level.The Perisphinctes plicatilis, Gregoryceras transversarium and Dichotomoceras bifurcatum Zones make up the Middle Oxfordian Substage in the Submediterranean area, the Perisphinctes plicatilis, P. pumilum and P. cautisnigrae Zones are representative of the Subboreal province.Boreal indexes are Cardioceras tenuicostatum and C. tenuiserratum, Amoeboceras glosense and A. serratum.The correlation between Subboreal perisphinctid and amoeboceratid zones was well demonstrated by Wright (1996b).
There is a difference in the usage of the P. plicatilis and G. transversarium Zones in Submediterranean Europe.Although Gygi & Marchand (1982) replaced the basal C. vertebrale Subzone with the C. densiplicatum Zone and included the P. antecedens Subzone in the G. transversarium Zone, subsequent authors have not followed the arguments of these authors and have continued to use the P. plicatilis Zone in the sense of Cariou et al. (1991a, b), i.e. with C. vertebrale and P. antecedens Subzones (e.g.Meléndez & Fontana 1993, fig. 4;Cariou et al. 1997).Cariou et al. (1991a) defined the G. transversarium Zone to contain the P. parandieri, P. luciae-formis, L. schilli and P. rotoides Subzones.In a more recent publication, Gygi (1995) included in the lower part of the G. transversarium Zone not only the P. antecedens Subzone but also the C. densiplicatum Subzone i.e. the whole P. plicatilis Zone (following the original usage of Oppel & Waagen 1866; R.A. Gygi, personal communication 1997).In further contributions to the Upper Jurassic of Switzerland (Gygi 2000b, c), the G. transversarium Zone is subdivided into the C. densiplicatum, P. antecedens, and P. luciaeformis Subzones; the overlying D. bifurcatus Zone contains in its lower part the L. schilli Subzone, which is considered in Spain and France to represent the upper part of the G. transversarium Zone (see above).The main reason for these differences is the occurrence of L. schilli in Switzerland above the vertical range of G. transversarium.
The Amoeboceras serratum Zone of Malinowska (1991) contains Epipeltoceras (uhligi group) and Ringsteadia salfeldi thus indicating, at least partly, equivalence with the lower E. bimammatum Zone (E.hypselum Subzone); this demonstrates that the A. serratum Zone of this author is younger in age than the A. serratum Zone of Sykes & Callomon (1979).The A. regulare Subzone of Malinowska (1991) seems to represent the upper E. bimammatum and perhaps the lowermost I. planula Zones, while the A. lineatum Subzone apparently corresponds to the rest of the I. planula Zone and the S. galar Zone.
The most difficult problems associated with these Upper Oxfordian zones concern their correlation in the Subboreal and Submediterranean schemes; this aspect is discussed in detail below.Some minor problems may be caused by the different hierarchical status of zones and subzones in the Subboreal and Boreal provinces.For example, Atrops et al. (1993b) recognised the A. regulare, A. rosenkrantzi and A. bauhini Zones, whereas Malinowska (1991) subdivided the R. pseudocordata Zone into the A. regulare and A. lineatum Subzones or, in Boreal Europe, into the A. regulare and A. rosenkrantzi Subzones.However, comparing the correlation chart of Malinowska (1991, table 3) with that of Matyja & Wierzbowski (1997, fig.3), it becomes evi-dent that the P. pseudocordata Zone of Malinowska corresponds only to the upper part of the A. regulare, the A. rosenkrantzi and the P. baylei Zones.
Another example is the variable status of A. bauhini as an index species.There is the A. bauhini horizon in the upper E. bimammatum zone equivalent to the P. densicostata horizon (Schweigert & Callomon 1997), the A. bauhini Subzone of the A. rosenkrantzi Zone (Sykes & Callomon 1979;Cariou et al. 1997), equivalent to the P. baylei Zone of Birkelund & Callomon (1985), and the A. bauhini Zone.Although initially equivalent to the P. densicostata horizon (Wierzbowski & Smelror 1993), the A. bauhini Zone was expanded by Matyja & Wierzbowski (1997, 1998) to correlate with the uppermost P. pseudocordata Zone and nearly the whole P. baylei Zone on the one hand and with the whole I. planula Zone and uppermost E. bimammatum Zone on the other; a little more restricted was the A. bauhini Zone of Schweigert & Callomon (1997), who excluded the S. galar Subzone of the I. planula Zone (see below).
At the base of the Middle Oxfordian, ammonites of the Perisphinctes plicatilis Zone provide the last possibility for long-distance correlation.Higher up in the Middle Oxfordian, zonal correlations become more and more difficult, best illustrated by the charts of J.H. Callomon (in: Wright 1980), Enay & Meléndez (1984) and Cariou et al. (1991b; see also Fig. 2).The divergent views are also well-documented by the tables of Malinowska (1991), Wright (1996a), Cariou et al. (1997) and Matyja & Wierzbowski (1997).

Kimmeridgian (Fig. 4)
Following the Luxembourg recommendations of 1962 and 1967 (Maubeuge 1964(Maubeuge , 1970)), two possibilities existed with respect to usage of the Kimmeridgian Stage, namely either a long version ('sensu anglico') or a short version ('sensu gallico'), both with differing zonal content and boundaries (see below).Use of two different versions of the Kimmeridgian evoked much confusion in following years and led to endless discussion.Therefore a vote of the International Subcommission on Jurassic Stratigraphy (ISJS) on this question was arranged in 1990, simultaneously with the vote on the Tithonian Stage (see below); the members of the ISJS voted for a 'short' version of the Kimmeridgian Stage (i.e.'sensu gallico').This meant that in future the upper boundary of the Kimmeridgian Stage should be coincident with the lower boundary of the Tithonian Stage and its Boreal equivalent, the Volgian (Zeiss 1991a).The lower boundary of the stage, however, remained ambiguous (see below).
Because of the still unresolved problems at the Oxfordian-Kimmeridgian boundary, the lower boundary of the Kimmeridgian Stage is drawn in this paper at the base of the Sutneria platynota Zone, following the above-mentioned adoption of a short Kimmeridgian Stage (i.e.'sensu gallico' or according to the 'continental' concept; Enay 1980b).The Working Group of the Oxfordian-Kimmeridgian Boundary is mandated to finally define the boundary at a level which allows farreaching correlations and corresponds to the resolutions of the International Commission on Stratigraphy (ICS); see also the discussions by Wierzbowski (1999Wierzbowski ( , 2001)).
Until such a definition has been taken by the Oxfordian-Kimmeridgian Boundary Working Group, voted on by the ISJS and approved by ISC, it seems useful to maintain the traditional boundaries in both biogeographic provinces, and it is premature to draw the Oxfordian-Kimmeridgian boundary in the Submediterranean area in the upper part of the E. bimammatum Zone (cf.Gygi 2000a, b).

Lower boundary
As the Luxembourg recommendations made it possible to select between two distinct versions of the Kimmeridgian Stage, the lower boundary was also defined twofold.In Subboreal regions of Europe, the boundary was drawn at the lower boundary of the Pictonia baylei Zone, whereas in Submediterranean regions it was placed at the base of the Sutneria platynota Zone (Maubeuge 1964, p. 85-86).At that time, it was supposed that both boundaries were more or less isochronous (Ziegler 1964), although doubts remained (e.g.Zeiss 1965;Cariou et al. 1971).
With the publication of Sykes & Callomon (1979), new impetus was given to further studies, which have suggested that the assumed time equivalence is erroneous or, at best, only partially true (Matyja & Wierzbowski 1988;Wierzbowski 1991;Atrops et al. 1993b;Schweigert 1995a, b).The main reasons for this view were the discovery of new Amoeboceras faunas by these authors Fig. 4. A tentative correlation chart for the Kimmeridgian Stage in Europe (thick lines as in Fig. 2).Modified after Zeiss (1965), Atrops (1982), Sarti (1988), Hantzpergue et al. (1991), Wierzbowski & Smelror (1993), Kutek & Zeiss (1994), Schweigert & Zeiss (1994) and Matyja & Wierzbowski (1997, 1998).and a re-evaluation of Salfelds (1915) Cardioceras paper as well as that of Koerner (1963), particularly with respect to their remarks concerning the type locality and possible type horizon of Cardioceras (= Amoeboceras) bauhini.The discussion of Wierzbowski (1991) concerning the range of the genus Ringsteadia in Poland is also important in this context.It soon became evident that Amoeboceras bauhini has its type horizon just below the upper boundary of the E. bimammatum Zone, (see A. Zeiss in: Enay & Meléndez 1984).The studies of Schweigert (1995a, b;Schweigert & Callomon 1997) resulted in similar conclusions, but led to a more precise faunal horizon subdivision of the Upper Oxfordian in Württemberg, SW Germany and to better correlation possibilities with England, with respect to the A. bauhini and the A. bayi (?= A. subtilicaelatum) horizon.
Some problems remained unsolved, however: 1. Does the A. bauhini horizon of southern Germany represent the same time interval as the beds bearing A. bauhini in England, Scotland and the Barents Sea?
Or is there a difference, and the vertical range of this species is different in these two areas?What is the situation in Poland, representing an intermediate region?
2. Does the A. subtilicaelatum horizon of southern Germany represent the same time interval as the A. bayi horizon in England?Or is there also a difference in the vertical range of these species in different parts of Europe?
3. Which units in the Subboreal realm correspond to the succession from the base of the I. planula Zone (with three or four faunal horizons) and the top of the lower S. galar Zone, which in Submediterranean Europe occurs between the A. bauhini and the A. subtilicaelatum (?= A. bayi) horizon?
It is not easy to answer these questions given the present state of knowledge; the following points are pertinent prior to discussion of these problem areas.The usage of A. bauhini as an index ammonite began with its introduction by Sykes & Callomon (1979) as a subzone of the A. rosenkrantzi Zone (uppermost Oxfordian); its stratigraphic position was subsequently revised by Birkelund & Callomon (1985), who regarded the A. bauhini Subzone and the P. baylei Zone (Lower Kimmeridgian) as approximate equivalents.One year prior to this latter publication, A. Zeiss (in: Enay & Meléndez 1984, fig. 6) had used A. bauhini informally as a zonal index in a correlation chart to show its approximate correspondence with the I. planula Zone sensu lato; this view was also held by Atrops et al. (1993b) and Matyja & Wierzbowski (1997, 1998).Wierzbowski & Smelror (1993) established the A. bauhini Zone formally and suggested that it was equivalent to only the lower part of the P. baylei Zone (the P. densicostata horizon); in more recent papers, Matyja & Wierzbowski (1994, 1995, 1997, 1998) provided charts showing the correlation between the A. bauhini Zone and the I. planula Zone sensu lato as well as with the P. baylei Zone (with the exception of the uppermost part).Finally, in southern Germany, an A. bauhini horizon was described by Schweigert (1995b;Schweigert & Callomon 1997) in the upper part of the T. hauffianum Subzone (uppermost E. bimammatum Zone); the latter authors correlated the Boreal A. bauhini Zone with the I. planula Zone sensu stricto, whereas the S. galar Zone was correlated with the Amoeboceras kitchini Zone.
The Amoeboceras bayi horizon was introduced by Birkelund & Callomon (1985) in the upper part of the P. baylei Zone, whereas Wierzbowski & Smelror (1993) reported the species at the base of their A. subkitchini Subzone.Atrops et al. (1993b) found the species, or closely related forms, in the Sutneria platynota Zone of the Submediterranean area.Schweigert (1995b) established an A. subtilicaelatum horizon in the uppermost part of the Sutneria galar Zone, assuming that A. bayi is only a variant of A. subtilicaelatum, which would then have priority.This conflicts with the opinion of Salfeld (1915), that A. lineatum and A. subtilicaelatum are very close and perhaps synonymous.Schweigert (1995b) also assumed that many specimens determined earlier as 'A.bauhini' belong in reality to A. bayi.To verify these assumptions, a comprehensive re-evaluation of the Upper Oxfordian -Lower Kimmeridgian Amoeboceras species complex (A.bauhini -A.bayi -A.subtilicaelatum -A.lineatum) would be necessary.Such a study should also illustrate the variation within each species in time and space (see, for example, Klieber 1981;Birkelund & Callomon 1985;Matyja & Wierzbowski 1988, 1994;Schweigert & Callomon 1997).
For the time interval of the I. planula Zone, Malinowska (1991) established the A. lineatum Subzone in Poland.It was introduced as the upper subzone of the R. pseudocordata Zone, but the precise correlation with other areas is not clear; from the list of fossils one would conclude that the S. galar Zone is not present.However, as a Sutneria sp.(of the galar/praecursor group?) is mentioned in the text but not figured, a deci-sion is difficult; its low stratigraphic level in the Goldap section would favour the S. praecursor Zone.In addition, Wierzbowski (1978) has described A. lineatum and A. bauhini together from the lower part of the I. planula Zone; thus, the A. lineatum Subzone seems to correspond to the lower part of the P. baylei Zone rather than to the upper part of the R. pseudocordata Zone.Malinowska (1988) reported specimens of A. bauhini only from the Lower Kimmeridgian, but these forms belong to other species such as A. bayi or A. cf.cricki.
In the Subboreal province, the Pictonia baylei Zone consists of two or three horizons.The lowermost horizon in Great Britain and the Boulonnais area is the Pictonia densicostata horizon; as mentioned above, this probably corresponds to the A. bauhini horizon.In the Boulonnais and Normandy areas, this is followed by the Pictonia baylei horizon sensu stricto, and, more widespread in France, the P. baylei and P. thurmanni horizon.In Dorset, the second horizon is apparently missing (Hantzpergue 1989), while the third one is represented by the P. baylei and P. normandiana horizon, which can also be observed in East Greenland (P.aff.normandiana horizon, Birkelund & Callomon 1985).P. normandiana is regarded as a synonym of P. thurmanni by Hantzpergue (1989).This third horizon also contains A. bayi.
What conclusions can be made from all these observations?
1.It seems likely that A. bauhini has a longer range in south Germany, as suggested by the many records of this ammonite species from the E. hypselum Subzone of the E. bimammatum Zone to the I. planula and S. galar Zones and even from the S. platynota Zone; a number of these determinations, although probably not all, may however be erroneous (Schweigert 1995b).Data from Poland also demonstrate that the range of A. bauhini is not restricted to the upper T. hauffianum Zone (= A. bauhini horizon), but extends as in southern Germany from the upper E. hypselum Subzone of the E. bimammatum Zone to the top of the I. planula Zone sensu lato (Matyja & Wierzbowski 1997, 1998).It is likely, therefore, that the A. bauhini Zone is of longer duration in the Submediterranean area, because it comprises not only the A. bauhini horizon of the upper T. hauffianum Subzone, but also three or four horizons of the I. planula Zone sensu stricto and at least one horizon of the lower S. galar Zone.As mentioned above, Malinowska (1991, p.16-17) apparently introduced the term A. lineatum Subzone for such an extended A. bauhini Zone.Approximately the same time interval has been called the A. bauhini Subzone (of an unnamed zone) by Matyja & Wierzbowski (1994, 1995) and subsequently elevated to the A. bauhini Zone (Matyja & Wierzbowski 1997, 1998); this zone is now correlated with the upper E. bimammatum Zone and the I. planula Zone sensu lato.It should also be noted that there is some evidence, at least in Scotland, that above the P. densicostatum bed follows another, younger bed with A. bauhini and Pictonia sp.(Wright 1989).This could be a hint that there are some more beds with A. bauhini, but without P. densicostata, which could correspond to the higher horizons of the P. baylei Zone.
In England, in contrast, Cox & Richardson (1982) observed A. bauhini in the uppermost part of the A. rosenkrantzi (= R. pseudocordata) Zone.If these determinations are correct, A. bauhini may occur a little earlier than the P. densicostata horizon.One can conclude from these observations that the range of A. bauhini, even in the Subboreal regions, is not restricted to the P. densicostata horizon or the 'A.bauhini Zone' sensu Wierzbowski & Smelror (1993).

If it can be confirmed that Amoeboceras bayi and
Amoeboceras subtilicaelatum are synonymous, as assumed by Schweigert (1995b), then the upper horizon of the Sutneria galar Zone (A. subtilicaelatum horizon) may correspond to the Amoeboceras bayi horizon of the lowermost Kimmeridgian Amoeboceras kitchini Zone.It should be noted, however, that A. bayi has also been reported from the lower ('Orthosphinctes') horizon of the S. platynota Zone (Atrops et al. 1993b).(c) The inclusion of this part of the Submediterranean subdivision in an A. lineatum Subzone (Malinowska 1991) with its unprecise limits (in southern Germany, the species is known to occur in the Upper Oxfordian and Lower Kimmeridgian) will not help significantly; this subzone can be replaced by the A. bauhini Zone, as used by Matyja & Wierzbowski (1997, 1998).

(a)
(d) There are apparently different possibilities of correlation and further research is necessary to clarify the situation.
(e) The Subboreal Oxfordian-Kimmeridgian boundary (R. pseudocordata/P.baylei Zone) can, with a high degree of probability, be positioned within the Submediterranean and Mediterranean scheme in the uppermost part of the E. bimammatum Zone on the basis of the correlation of the A. bauhini horizon with the P. densicostata horizon.The Submediterranean Oxfordian-Kimmeridgian boundary remains at the base of the S. platynota Zone.

Additional remarks on Lower Kimmeridgian correlation
As mentioned above, the upper S. galar Zone (A. subtilicaelatum horizon) is probably an equivalent of the Amoeboceras bayi horizon (Schweigert 1995a), which extends into the lower part of the S. platynota Zone (Amoeboceras horizon with A. bayi, see Atrops et al. 1993b).This contrasts somewhat with the correlation of Birkelund et al. (1983, table 1), who considered the Pictonia baylei Zone and the Paraspidoceras rupellense Zone of Hantzpergue (1979) to be equivalent.Hantzpergue (1989), too, correlated the P. baylei Zone with the P. rupellense Zone (horizons R1 and R2); horizons P1-3 of the I. planula Zone sensu lato are considered to be equivalent to the R. pseudocordata Zone sensu lato (Hantzpergue 1989, tables E, F).In his sections, he found the Upper Oxfordian Sutneria galar in the Lithacosphinctes gigantoplex horizon (P3), immediately below his P. rupellense Zone (see Fig. 2).The P. rupellense Zone itself is situated between the gigantoplex horizon (P3) of the uppermost Idoceras planula Zone sensu lato and the Rasenia cymodoce Zone (Fig. 4); it is therefore considered to be equivalent to the lowermost Submediterranean Kimmeridgian (S. platynota Zone; Schairer 1970;Atrops 1982;Olóriz & Rodríguez-Tovar 1996); its lower horizon (R1) seems to correspond to the upper part of the lower ('Orthosphinctes') subzone of the S. platynota Zone, whereas the lower part (Amoeboceras horizon) of this zone is not represented; its upper horizon (R2) contains the index 'Ardescia virgatoides', which is similar to forms of the Ardescia desmoides horizon of the Ardescia desmoides Subzone of the middle Sutneria platynota Zone and is therefore very important for correlation to the Submediterranean region.
Above the P. ruppelense Zone, Hantzpergue (1989) subdivided the Rasenia cymodoce Zone into nine horizons (C1-9); the R. cymodoce horizon (C2) could be traced from western France to Normandy and the Subboreal regions.In northern Europe, the R. cymodoce horizon is rather widespread (Wierzbowski 1989) and in Spitsbergen it represents the only rasenoid horizon within the Amoeboceras succession.In East Greenland, Birkelund & Callomon (1985, fig. 5) recognised two other horizons below the horizon of Rasenia cymodoce ('17'), namely the 'Pachypictonia' horizon ('16') and the Rasenia inconstans horizon ('15').These horizons of the lower R. cymodoce Zone were considered to be equivalent to the P. altenense horizon (C1; Hantzperque 1989); they are probably equivalent to the lower Ataxioceras hippolytense Subzone of the lower Ataxioceras hypselocyclum Zone of south-east France, whereas the R. cymodoce horizon perhaps has its equivalents in the upper part of this subzone.
In the middle and upper part of the R. cymodoce Zone, only a few possibilities remain for far-reaching correlations in Europe, such as the Eurasenia aulnisa horizon (C5), which contains the highly characteristic Submediterranean subzonal index A. lothari, and the Semirasenia askepta horizon (C7), which has been found in Scotland, England, Normandy, western France (Birkelund & Callomon 1985;Hantzpergue 1989) and southern Germany (Heller 1964;Doben & Heller 1968).In northern Germany, Submediterranean ammonites of Early Kimmeridgian age have been found in sediments which had earlier been attributed to the Upper Oxfordian (Fischer 1991).

Substages
The Kimmeridgian Stage has been subdivided into two or three substages; here a subdivision into three sub-stages is preferred.If the Middle Kimmeridgian is not recognised, then the middle and the upper part are united as Upper Kimmeridgian (Fig. 4).

Zones
In the Mediterranean and Submediterranean provinces, the Lower Kimmeridgian consists of three zones, which can be correlated approximately as follows: (1) Sowerbyceras silenum -Sutneria platynota, (2) Ataxioceras hypselocyclum -Taramelliceras strombecki and (3) Crussoliceras divisum -Mesosimoceras herbichi (Fig. 4).Their further subdivision into subzones is different in both areas (Fig. 4); precise correlation of these units is thus difficult (Pavia et al. 1987;Sarti 1993).Detailed subdivisions into subzones and faunal horizons have been proposed in south-east and western France (Atrops 1982;Hantzpergue 1989); that of south-east France can also be used with some minor changes in southern Germany.The Submediterranean zonal subdivision as established by Geyer (1961) can be used from the Iberian Peninsula to Bulgaria and Turkey (Sapunov 1977a;Lopez Marques 1983;Alkaya 1992).In Poland, the Submediterranean zonal subdivision has been adopted by Malinowska (1988) and Matyja & Wierzbowski (1998).In Subboreal and Boreal regions, subdivision into two zones is typical (see above): (1) Pictonia baylei and (2) Rasenia cymodoce.These zones can be replaced by the Amoeboceras kitchini Zone in areas where no perisphinctids occur (e.g.Wierzbowski & Smelror 1993); this zone may extend into the lower part of the Aulacostephanus mutabilis Zone (see below).
The Middle and Upper Kimmeridgian Substages together consist of three zones in all parts of Europe (Fig. 4).
In regions where no perisphinctids are present, these latter zones can be replaced in the lowermost parts by the Amoeboceras kitchini Zone (see above) followed by the A. kochi, A. elegans and Suboxydiscites taimyren-sis Zones (Fig. 4).The latter index has been taken from northern Siberia charts (Birkelund & Callomon 1985), but there is no mention of this species in more western regions, with the exception of a determination from the Middle Kimmeridgian of Greenland.Therefore, for these Boreal regions too, Aulacostephanus autissiodorensis seems to represent the more appropriate index species.
The Middle Kimmeridgian zonal and subzonal subdivisions can be applied without great difficulty in Boreal, Subboreal and Submediterranean Europe, as there are large regions with overlapping guide fossils, whereas in the Mediterranean province, only a zonal subdivision is possible.Hantzpergue (1989) established a detailed subdivision in western France, which can also be used in northern France (Geyssant et al. 1993;Proust et al. 1993) and traced as far as Germany (Zeiss 1991b;Schweigert 1993aSchweigert , 1996a)), England, Norway and East Greenland (Hantzpergue 1989).
An unresolved problem is the lower boundary of the A. mutabilis Zone; it is drawn at the base of the A. lineatum horizon in western and northern France (Hantzpergue 1989;Hantzpergue et al. 1997), but in England, following the revisions of Birkelund et al. (1983), it is placed four horizons deeper, at the base of the S. askepta horizon.Recent investigations in central Poland came to similar results (Matyja & Wierzbowski 1998); these workers traced the boundary to a slightly deeper level in the upper A. hypselocylum Zone.In Germany and the Submediterranean region, the usage from south-east France has been followed (Hantzpergue 1989;Hantzpergue et al. 1991;Zeiss 1991b), which facilitates correlation with the base of the A. acanthicum Zone; the lower boundary of this zone in Germany is traditionally drawn at the incoming of the first representatives of the genus Aulacostephanus (lineatum group).In a recent publication by Hantzpergue et al. (1997), the problems of this boundary are well illustrated by their table 12; in the 'Biome Franco-germanique', the lower boundary of the A. mutabilis Zone is drawn below its lowermost horizon (linealis horizon), whereas the base of the A. mutabilis Subzone, curiously, is placed two horizons higher (attenuatus horizon).It is evident that the new data from Poland (Matyja & Wierzbowski 1998), which place the base of the A. mutabilis Zone much deeper, will probably necessitate revision of all these correlations.
In the Upper Kimmeridgian (upper A. autissiodorensis Zone) of Poland and the Russian platform, a Sarmatisphinctes fallax Subzone has been established (Mesezhnikov 1984(Mesezhnikov , 1988;;Kutek & Zeiss 1994, 1997).For the lower part (lower A. autissiodorensis Zone), the Discosphinctiodes subborealis Subzone is proposed; D. subborealis is a significant index fossil.In Poland, Aulacostephanus autissiodorensis has been found only in the lower and middle parts of the S. fallax Subzone.In western Siberia, a zone of Virgataxioceras dividuum is the equivalent of the S. fallax Subzone (Mesezhnikov 1988).In northern Germany, Schweigert (1996a) stated, based on re-study of previous collections, that the A. autissiodorensis Zone is probably present.
For the Upper Kimmeridgian of western France, a useful subdivision has been proposed by Hantzpergue (1989), who subdivided the A. autissiodorensis Zone into two subzones, the A. autissiodorensis and the Gravesia irius Subzones, each with two faunal horizons.The succession in the Boulonnais area and farther north has been worked out in detail by Geyssant et al. (1993) and Geyssant (1994); the succession in southern England was reported by Cox & Gallois (1981), Birkelund et al. (1983) and Callomon & Cope (1996).

Correlation
Many difficulties are encountered in correlating zones (and subzones) of the Lower Kimmeridgian in Europe, mainly between the Submediterranean and Subboreal regions, but also between the Submediterranean and Mediterranean areas (Fig. 4).Many correlations are arbitrary and well-constrained correlation is only possible at certain levels.Such correlation possibilities in the Lower Kimmeridgian Substage have already been explained in connection with the problems of the Oxfordian-Kimmeridgian boundary.Some problems exist around the Lower-Middle Kimmeridgian boundary, as the base of the A. mutabilis Zone is variably defined in different parts of Europe (see above).Considering the most recent results from Poland (Matyja & Wierzbowski 1998), the lower boundary of the Subboreal A. mutabilis Zone lies within the uppermost part of the Submediterranean A. hypselocyclum Zone, i.e. one zone deeper than previously assumed.
In the Middle Kimmeridgian Substage, correlations within the A. acanthicum/A.mutabilis Zones and the A. eudoxus Zone pose no great problems although the uppermost part of the A. eudoxus Zone of western France (A. contejeani Subzone) seems to correspond to the lower part of the H. beckeri Zone in south Germany (Schweigert 1993b).Correlation of the A. kochi Zone with the upper part of the A. mutabilis and/or the lower part of the A. eudoxus Zone (Wierzbowski & Smelror 1993) is still tentative, as is the correlation of the A. elegans Zone with most of the A. eudoxus Zone.
Correlation of the Upper Kimmeridgian Substage (Submediterranean H. beckeri Zone with the Subboreal A. autissiodorensis Zone) was hitherto only possible by indirect arguments.The elaboration of a new zonal and subzonal subdivision in western France by Hantzpergue (1989) and the new discoveries by Schweigert (1993aSchweigert ( , b, 1994) ) in Germany and by Kutek & Zeiss (1997) in Poland now permit correlation of parts of the Upper Kimmeridgian of western, central and eastern Europe and perhaps also western Siberia.

Chronometric data
The duration of the Kimmeridgian Stage has been estimated to be 3.4 Ma (Gradstein et al. 1995;Ogg 1995;Ogg & Gutowski 1996); for precise data, see Figure 4.
Tithonian and Volgian (Fig. 5) The Tithonian, and its Boreal equivalent the Volgian, have been confirmed as stage names by a vote of the International Subcommission on Jurassic stratigraphy in 1990 (Zeiss 1991a).A further stage name 'Bononien' (for the 'Upper Kimmeridgian sensu anglico', proposed by Cope 1993) seems unnecessary and could result in each region with a differing zonal subdivision claiming its own stage name, leading only to more confusion rather than to international agreement concerning uniform nomenclature.Furthermore, due to the different meanings of the stage 'Portlandian' in different countries, it was voted in 1990 that usage of this name should be discontinued.The most recent review of the Tithonian Stage and its ammonites is that provided by Geyssant (1997); for the Volgian Stage and ammonite biostratigraphy, see Gerasimov et al. (1995), Callomon & Cope (1996) and Kutek & Zeiss (1997).

Lower boundary
The base of the Tithonian Stage is defined by the base of the Hybonoticeras hybonotum Zone.It is generally supposed that the base of the coeval Gravesia gigas, Virgatosphinctoides elegans and Ilowaiskya klimovi Zones are drawn at approximately the same time level (see also below).

Substages
The Tithonian is subdivided into two or three substages; here preference is given to a tripartite Tithonian Stage (Fig. 5).If only two substages are used, then the lower and middle part are united as the lower substage ('Danubian'), the upper substage corresponds to the 'Ardescian' substage.Type regions for the Lower and Middle Tithonian Substages have been proposed by Barthel (1975) and Zeiss (1975).The type region for the Upper Tithonian Substage, the Ardescian, has been revised by Cecca et al. (1989a, b).

Zones and Subzones
Whereas the two lower stages of the Upper Jurassic have two main zonal subdivisions, at least four subdivisions are necessary in the upper stage (Fig. 5).This is due to the extreme provincialism of ammonites caused by the increasing isolation of late Jurassic marine basins, which seem to have only rarely been directly connected; interbasinal migration was apparently only favoured during the lowermost zone of the stage.
The most important lower zone is that of Hybonoticeras hybonotum, which can be followed over long distances in Mediterranean and Submediterranean Europe (Zeiss 1968;Olóriz 1978;Sapunov 1979;Sarti 1988); in southern Germany it is possible to recognise three subzones and seven horizons in the H. hybonotum Zone (Schweigert & Zeiss 1999).In central Europe, the latter overlaps with the Gravesia gigas Zone, which has a rather wide distribution regionally in central and western Europe.During the last decades, many new discoveries have been reported and the genus Gravesia and the stratigraphy of the beds with Gravesia have been revised (Hahn 1963;Zeiss 1974;Hantzpergue 1989;Schweigert 1994Schweigert , 1996a, b;, b;Schweigert et al. 1996;Zeiss et al. 1996;Dimke & Zeiss 1997).In the Subboreal subprovince, the genus Gravesia is also present, but less numerous, so that other index fossils have been given priority, such as Virgatosphinctoides elegans in northwestern and Ilowaiskya klimovi in eastern Europe (Cope 1967;Cope et al. 1980;Kutek & Zeiss 1974, 1994, 1997;Callomon & Birkelund 1982;Mesezhnikov 1988).According to Callomon & Cope (1996), Gravesia cf.gravesiana occurs in the lower part of the Virgatosphinctoides scitulus Zone, thus demonstrating the correlation with the upper H. hybonotum Zone (containing G. gravesiana).In northern Germany, beds with Gravesia gigas intermedia are apparently the youngest beds containing Jurassic ammonites (Schweigert 1996a) and are succeeded by brackish and freshwater sediments up to the Jurassic-Cretaceous boundary.In these beds, ostracodes have proved to be the best guide fossil (Bischoff & Wolburg 1963;Schudack 1994, fig. 24), permitting subdivision of the Tithonian Stage in northwest Germany into four zones.In other areas, such as eastern England and Denmark, subdivision into nine zones is possible using ostracodes (Christensen 1988;Schudack 1994, fig. 24).
The upper zone of the Lower Tithonian in Mediterranean Europe, the zone of Semiformiceras darwini (or of Virgatosimoceras albertinum), is apparently equivalent to the Neochetoceras mucronatum and Franconites vimineus Zones (each of them with two subzones and some horizons) of Submediterranean Europe, as they have numerous faunal elements in common (Enay & Geyssant 1975;Olóriz 1978;Cecca et al. 1986;Sarti 1984Sarti , 1988;;Cecca 1990a, b).Precise correlations have still to be worked out, however, and at present this is difficult as no subzones or even horizons have been recognised in the Tethyan realm.The Submediterranean zones have been traced from south-east France via southern Germany to Hungary as well as in Bulgaria and perhaps also Turkey (Zeiss 1968;Sapunov 1977bSapunov , 1979;;Vigh 1984;Fözy 1988Fözy , 1993;;Alkaya 1989;Atrops 1994;Fözy et al. 1994).Correlation with the Subboreal regions is only tentative and different proposals have been published (Fig. 5;Zeiss 1977;Mesezhnikov 1988;Kutek & Zeiss 1997).
In Subboreal Europe, the situation is not much better and correlations between the different subprovinces of Northwest and eastern Europe are only approximate.Consequently, different zonal subdivisions are also applied in these subprovinces.In eastern Europe, for example, species of the genus Ilowaiskya are used (e.g. the Ilowaiskya sokolovi and I. pseudoscythica Zones; Mesezhnikov 1988;Kutek & Zeiss 1997), whereas in Northwest Europe, representatives of the genera Virgatosphinctoides, Arkellites and Pectinatites have been selected (e.g. the Virgatosphinctoides scitulus, W. wheatleyensis, Arkellites hudlestoni and Pectinatites pectinatus Zones); each of these latter zones can be subdivided into two subzones (Cope et al. 1980;Callomon & Birkelund 1982;Geyssant 1997).
In eastern Europe, the equivalents of the Middle Tithonian Substage are probably the upper part of the Lower Volgian (upper Ilowaiskya pseudoscythica and Ilowaiskya tenuicostata Zones).The latter unit is discernable in Poland but has not been recognised in Russia to date (Kutek & Zeiss 1974, 1988, 1994, 1997;Mesezhnikov 1988;Kutek 1994).In its upper part, the Pseudovirgatites puschi horizon is important due to its mixed fauna (Kutek & Zeiss 1974, 1988, 1997).A local time equivalent in north-eastern Austria is probably the Isterites austriacus Zone with Buchia rugosa as an important guide fossil (Fig. 5).A quite different zonal subdivision exists in Great Britain and the adjoining Subboreal and Boreal regions as far as Greenland (Cope 1978(Cope , 1980;;Wimbledon 1980;Callomon & Birkelund 1982;Kejsi et al. 1988); the Middle Tithonian perhaps corresponds to the main part of the Pectinatites pectinatus Zone and perhaps to the Pavlovia pallasioides Zone of England or to the Dorsoplanites primus and Pavlovia iatrensis Zone of East Greenland.
The Upper Tithonian Substage consists of two or three zones in the Mediterranean area.In southern Spain, the lowermost zone has been identified as the Simplisphinctes Zone (Tavera 1985).This unit has not been identified in northern Italy (Sarti 1988), but could be recognised as far as north-eastern Austria, where the same ammonite fauna (containing the genus Oloriziceras) occurs (Zeiss & Bachmayer 1989).In the absence of the rather peculiar index genus Simplisphinctes, this zone was called the Oloriziceras magnum Zone for this region (Zeiss 2001).Above the Simplisphinctes (or S. abnormis or O. magnum) Zone, the Paraulacosphinctes transitorius Zone (with the first Crassicollaria) occurs.A Micracanthoceras micracanthum Zone is sometimes adopted instead of the P. transitorius Zone; this zone apparently also contains the equivalents of the Simplisphinctes (better S. abnormis) Zone (Enay & Geyssant 1975;Sarti 1988;Geyssant 1997).Some authors consider the Simplisphinctes and P. transitorius Zones as subzones of the M. micracanthum Zone (Benzaggagh & Atrops 1997;Geyssant 1997) although the former authors, based on Moroccan data, only partially substituted the Simplisphinctes Subzone, replacing its upper part and the P. transitorius Subzone by two new subzones, that of 'Micracanthoceras (Corongoceras) spp.' and that of 'Moravisphinctes spp.'.It is very important that these new subzones can be correlated rather precisely with the calpionellid subdivision; the Chitinoidella boneti Subzone (of the Chitinoidella spp.Zone) corresponds to the first two subzones.The base of the Crassicollaria spp.Zone (Zone A) approximately coincides with the base of the Moravisphinctes spp.Subzone, which corresponds to the lower part of this zone (= Subzone A1).
The Durangites Zone follows above the P. transitorius Zone.In northern Italy, this zone was named the Durangites vulgaris Zone by Sarti (1988); this term has also been adopted by other authors.In some countries, this zone has not been recognised; the equivalents of this zone are then apparently included in the P. transitorius Zone, which sometimes even includes parts of the Lower Cretaceous (e.g.Sapunov 1977b).The fauna of this zone has been mainly described by Tavera (1985), Tavera et al. (1994) and Enay et al. (1998a, b).
During the Middle Volgian, central Poland belonged to the eastern Subboreal subprovince, but only the lowermost unit, the Zaraiskites scythicus Zone (with the lower Z. scythicus and upper Z. zarajskensis Subzones) is represented (Kutek 1994).Brackish sediments prevail higher in the Polish section and yield ostracodes; the Cypridea dunkeri and the Cypridea granulosa Zones can be recognised.On the Russian Platform, the lowermost horizon of the Z. scythicus Subzone (Z.quenstedti horizon in Poland) is probably represented by beds containing Zaraiskites disprosopa and Isterites(?) contradictionis (Ilovaiskij & Florenskij 1941).
On the Russian Platform, a Dorsoplanites panderi Zone is now used instead of the Z. scythicus Zone (Mesezhnikov 1988;Kutek 1994); above follows the Virgatites virgatus Zone (with three subzones: V. gerasimovi, V. virgatus and C. ivanovi; Gerasimov et al. 1995).The V. virgatus Zone is succeeded by the Epivirgatites nikitini and Lomonossovella blakei Zone (separated by Callomon & Birkelund (1982), and, in reverse order, by Mesezhnikov (1988) but adopted as a single zone by other Russian authors (e.g.Gerasimov et al. 1995)).The uppermost Middle Volgian is represented by the Paracraspedites oppressus Zone (Mesezhnikov 1988).In the Baltic area, Middle Volgian ammonites are rare although a few specimens from Lithuania were mentioned by Rotkyte .(1976,1987).In Scandinavia, Middle Volgian ammonites have been found in Denmark (Birkelund & Pedersen 1980) and in Norway (Birkelund et al. 1978).In England and East Greenland, Dorsoplanitidae are prevalent, but in both these regions, the subdivisions are distinct; in England, Pavlovia pallasioides, Pavlovia rotunda and Virgatopavlovia fittoni characterise the lower part of the Middle Volgian whereas Progalbanites albani and the giants Glaucolithites glaucolithus, Galbanites okusensis, Kerberites kerberus and Titanites anguiformis characterise the upper part (Cope 1978;Wimbledon & Cope 1978).As in Russia, the uppermost zone is the Paracraspedites oppressus Zone (Casey 1973;Kejsi & Mesezhnikov 1986;Kejsi et al. 1988), but not all authors adopt this zone.In East Greenland, there are some similarities with Siberian ammonite successions, but in general the subdivision there has its own character and, with three exceptions, its distinct index species (Callomon & Birkelund 1982;Mesezhnikov 1988)
Due to problems of provinciality, such correlation schemes are necessarily speculative and ultimately unsatisfactory.A tentative summary correlation scheme is given in Figure 5, based on developments since earlier attempts by the author (Zeiss 1983(Zeiss , 1986)).A similar, although in detail somewhat different, correlation chart has recently been published by Hantzpergue et al. (1998).
Concerning the Middle and Upper Tithonian (upper Lower and Middle Volgian) Substages, a number of observations are pertinent.Although correlation between the Mediterranean and Submediterranean area is quite possible in the lowermost Middle Tithonian Substage (S. semiforme/R.richteri -V.rothpletzi/S.pennicilatum Zones), a number of different proposals have been made for the higher zones (Enay & Geyssant 1975;Olóriz 1978;Jeletzky 1984Jeletzky , 1989;;Cecca et al. 1986;Kutek 1994).A satisfactory answer to this problem requires complete revision of the famous Submediterranean Neuburg fauna and sections, in which some levels with distinct ammonite faunas have already been recognised by Barthel (1964Barthel ( , 1975)).
In eastern central Europe (north-eastern Austria, Moravia, central and southern Poland), some Submediterranean and Mediterranean ammonites genera of Middle and Late Tithonian age are represented by characteristic forms.They sometimes interfinger with Subboreal elements, thus providing good potential for correlation (Kutek & Wierzbowski 1986;Kutek & Zeiss 1988, 1997;Kutek 1994).The I. tenuicostata and Z. scythicus Zones of central and southern Poland, for example, display interesting forms with affinities to both the Submediterranean and Subboreal provinces.Combined with observations from other localities, this facilitates better correlation between these two regions: (1) the Pseuvirgatites puschi horizon of the uppermost Ilowaiskya tenuicostata Zone contains Isterites species described from the higher parts of the Neuburg beds, i.e. of late Middle Tithonian age, and (2) the Z. regularis horizon of the lower Z. zarajskensis Subzone (upper Z. scythicus Zone) contains Pseudovirgatites scruposus and calpionellids indicative of the calpionellid Zone A, such that correlation is possible with the lower part of the Paraulacosphinctes transitorius Zone.
In the Boreal and Subboreal provinces, quite different zonal subdivisions exist, mainly based on different perisphinctid groups, such as the Pectinatitinae and Dorsoplanitinae in England, Denmark, Norway and Greenland and the Ilowaiskyinae, Virgatitinae and Dorsoplanitinae in Poland and Russia.The correlation of these zones is rather arbitrary, as demonstrated by Callomon & Birkelund (1982), Mesezhnikov (1988) and W.A. Wimbledon (in: Callomon & Cope 1996), and is based mainly on similar, but non-identical species of Dorsoplanitinae.

Chronometric data
The approximate duration of the Tithonian has been estimated to be 6.7 Ma (Gradstein et al. 1995;Ogg 1995); for precise data, see Figure 5.

Biochronological importance of nonammonite fossil groups: a review
The Jurassic System is the classic one for subdivision by ammonites.This fossil group has been used with much success since the pioneering work in the last century by workers such as L. von Buch, A. d'Orbigny, A. Oppel, F.A. Quenstedt, K.A. von Zittel and S. Buckman.Indeed, this contribution on the chronological subdivision of the Upper Jurassic of Europe has been compiled primarily using ammonites (see above).However, Upper Jurassic marine sediments of epicontinental shelves, the habitat of ammonites, are not present everywhere in Europe, so that ammonites are not always available.It is often necessary, therefore, to utilise other fossil groups with proven stratigraphic value such as bivalves, brachiopods, foraminifera, ostracodes and distinct plant mega-, micro-and nannofossil groups.Radiolarians, calpionellids, conchostracans, insects and vertebrates should also be added to this list; the first two groups are very useful in pelagic sedimentary basins whereas the last ones are used with much success in the stratigraphic subdivision of continental sediments, such as those of central and eastern Asia and of North America.
The challenging task of correlating between the different fossil subdivision schemes has been addressed for individual groups (e.g.Le Hégarat & Remane 1968;Surlyk & Zakharov 1982).Multidisciplinary correlation charts, typically for microfossil groups, have only been successfully developed within the last two decades.Useful though incomplete examples of such schemes, including micro-and macrofossils, have recently been published by Tavera et al. (1994), R. Enay (in: Cariou & Hantzpergue 1997), Gramann et al. (1997) and Remane (1997).It remains as one of the more important tasks, however, to establish European multidisciplinary correlation charts that incorporate all fossil groups important for biochronology and also include radiometric ages and palaeomagnetic reversal data.During the editorial work, it was brought to the attention of the author that charts fulfilling many of these expectations have recently been published by Hardenbol et al. (1998, charts 6-7); of special interest are the chronometric data for most of the biochronostratigraphic units (see below).

Aptychi
In the Tethyan regions, aptychi have proven to be a useful addition to ammonites for the subdivision of Upper Jurassic sediments.Following the studies of Durand & Gąsiorowski (1970) and Gąsiorowski (1962Gąsiorowski ( , 1985)), it is possible to differentiate eleven zones of aptychi using four larger groups of aptychi, the lamellaptychi and laevaptychi and to a lesser degree the laevilamellaptychi and punctaptychi.Correlation between aptychi and ammonite zones still poses problems (A.Wierzbowski, personal communication 1998).Eliás et al. (1996) also used aptychi ranges for biostratigraphy, but without a zonal subdivision; they preferred a multidisciplinary correlation method using the calpionellid subdivision as reference.

Belemnites
The most recent review of this fossil group is that of Doyle & Bennett (1995) which includes a section on Middle and Upper Jurassic belemnite groups, including those of Europe.This publication presents a comprehensive review of the subject, including the work of Saks & Nalnyaeva (1964, 1966), Riegraf (1980Riegraf ( , 1981)), Combémorel & Mariotti (1986), and Doyle & Kelly (1988); a range chart of the most useful taxa for biostratigraphy of the Middle and Upper Jurassic is included by Doyle & Bennett (1995).The stratigraphic ranges of some more important Polish species have been published by Pugazewska (1988) and Malinowska (1997) and those of Sicily by Combémorel & Mariotti (1990).Recently, Combémorel (1997) compiled all data available for the Tethys and the Boreal region of Europe and for each of them presented a correlation scheme with the subdivisions based on ammonites and belemnites; see also Hardenbol et al. (1998, chart 7).

Bivalves
The most important group of bivalves for biostratigraphic purposes in the Upper Jurassic of Europe is the genus Buchia.In the Boreal regions of Eurasia and North America, it is of particular importance as a supplement to ammonites.The genus has been the focus of many papers in the last decades such as Zakharov (1981Zakharov ( , 1987Zakharov ( , 1990)), Surlyk & Zakharov (1982), Jeletzky (1984), Kelli (1990), Sey & Kalacheva (1993b) and Sha & Fürsich (1994).An interesting interpretation of the different ranges of Buchia species in America and Eurasia has been presented by Hoedemaker (1987).Stratigraphic range lists of selected bivalve species from Poland have been published by Karczewski & Pugaczewska (1988) and Malinowska (1997).
A correlation chart that is mainly based on buchiid bivalves but also includes other bivalve genera (e.g.Retroceramus) has been compiled for northern Russia and the circum-Pacific regions by Damborenea et al. (1992).In the Upper Jurassic, the stratigraphic resolution of bivalve taxa, with the exception of buchiids, seems to be rather limited and/or needs further research (Damborena et al. 1992).For some regions, stratigraphic range lists of selected bivalve species have been published, for example for Poland (Malinowska 1997;Karczewski & Pugaczewska 1998) and for northern Germany by Kaever et al. (1976).

Gastropods
The biostratigraphic resolution of this group in the Jurassic is not very high, but in special cases, when other guide fossils are not present, some representatives of the group may be used.An example from the Upper Jurassic of France (Nerineaceae) has been published recently by Barker (1994).Range lists of selected species from Poland have been published by Karczewski (1988) and Malinowska (1997).

Brachiopods
The most recent reviews of this group with respect to Upper Jurassic brachiopods are those of Ager (1994) and Alméras et al. (1991Alméras et al. ( , 1994)), especially for France and Britain, and Boullier & Laurin (1997) for the Tethys and the 'Domaine NW européen français'.Ager (1994) considered the group within a global context.Alméras et al. (1994) discussed the facies dependence of brach-iopods, concluding that distinct zonal species of brachiopods are often necessary for different facies.For biostratigraphical purposes, it is possible to subdivide the Upper Jurassic of England and north-west France into nine zones and some subunits.The Polish species have been figured and described by Barczyk (1988); range charts are given in Malinowska (1997).Prozorovskaja (1993) presented an overview of the brachiopod subdivision of the Upper Jurassic of the southern part of the former USSR.

Echinoderms
To date, there is no subdivision scheme of the Upper Jurassic with respect to echinoderms.Some genera have biostratigraphic value; Saccocoma, for example, has been used in some multidisciplinary schemes.Thierry et al. (1997) presented range charts of the Upper Jurassic regular and irregular echinoid genera and species of France, with the expectation that with detailed research it would be possible to create a subdivision scheme comparable to that developed for the brachiopods of France.

Corals (scleractinians)
This group has poor biostratigraphic resolution.Its usefulness for stratigraphic purposes is therefore rather limited, also because of the close dependence of corals on ecological factors (Rosendahl 1988).Nevertheless, Beauvais (1988) subdivided the Upper Jurassic Series (except the Lower Oxfordian) into six zones based on madreporians (scleractinians).Polish species with range charts have been presented by Roniewicz & Morycowa (1988) and range charts were published by Malinowska (1997).

Sponges
This fossil group is poorly suited to regional correlation, but some species may be useful for local subdivision; examples from France have been presented by Gaillard (1997).

Vertebrate megafossils
Jurassic vertebrate fossils are too scarce to be used as guide fossils.Nevertheless, if vertebrate remains are studied thoroughly, they frequently provide valuable biostratigraphic information (e.g.elasmobranchian teeth, Gramann et al. 1997).It should be mentioned that the Jurassic Period in Europe saw the early evolution of mammals, the flourishing of the first true birds and the first wave of the acme of the dinosaurs.
In other parts of the globe, vertebrates have been used for stratigraphy; in North America, for example, Turner & Peterson (1998) subdivided the Upper Jurassic Morrison Formation into four biozones on the basis of dinosaurs, whereas in China, fish are used for subdivision (Chen 1990).

Foraminifera
In the 1950-60s, foraminifera were one of the most important microfossil groups, together with ostracodes, for relative age determinations of marine sediments in boreholes; their importance has decreased in more recent times.Studies of foraminifera faunas from outcrops in southern Germany were reviewed by Groiss (1984).An account of epistominian zonation was given by Ascoli (1988), who also presented zonations and correlations between east Canadian offshore wells and the East European Platform (Grigelis & Ascoli 1995).Foraminifera from northern Germany were presented by Klingler et al. (1962) and Gramann et al. (1997).The guide fossils and characteristic species of the Upper Jurassic foraminifera of Poland have been published by Bielecka (1988) and Styk (1997), those of the Russian Platform by A.Y. Azbel (in: Mesezhnikov 1989).Foraminifera of Sweden were studied by Norling (1972) and Guy-Ohlson & Norling (1988).A short compilation of Upper Jurassic foraminifera in Britain has been published by Shipp & Murray (1981), together with a range chart and figures of index species.The most recent reviews of foraminifera of Europe have been compiled by Ruget & Nicollin (1997) on the small benthic forms, and by Bassoulet (1997a) on the large forms; see also Hardenbol et al. (1998, chart 7).

Radiolaria
This microfossil group, which has been the subject of much scientific research in recent years in Europe, is of particular importance in the Tethyan region.A comprehensive monograph was recently published by Baumgartner et al. (1995) on the radiolarians of the Tethys, including a catalogue of all Tethyan species.The biochronological potential for subdividing the Upper Jurassic Series into 'Unitary Association Zones' (U.A.Z.) is well-demonstrated; there are six such zones covering the whole Upper Jurassic.They have a duration of between 2-6 Ma.This monograph demonstrates the significant advances in research into this group, especially if new quantitative concepts, such as the 'Unitary Association Zones', are applied to the biochronological subdivision of the Upper Jurassic.
Research into radiolarians and their stratigraphic potential has also been on the increase outside the Tethys, as demonstrated by recent publications concerning the Submediterranean province (Riegraf 1987;Kießling 1997;Zügel 1997;Zügel et al. 1998), and even the Subboreal and Boreal provinces, including the North Sea (Dyer & Copestake 1989), the Russian Platform and the Barents Sea (Vishnevskaya 1993(Vishnevskaya , 1997(Vishnevskaya , 1998;;Kozlova 1994).Dyer & Copestake (1989) introduced a biozonation based on a succession of ten radiolarian events in the Kimmeridgian and Tithonian.Important attempts are also underway to correlate the new peri-Tethyan radiolarian assemblages with different micro-and macrofossil biozonations (Vishnevskaya & De Wever 1997); owing to strong provincialism, direct correlation between the peri-Tethyan and Tethyan zonations is still very difficult, but has been undertaken recently (Hardenbol et al. 1998, chart 7).

Ciliata
This group is important only in the Tethyan region and the surrounding shelf deposits; the most comprehensive studies of the ciliata in recent years have been published as a result of the Sümeg meeting (Fülöp 1986;Remane et al. 1986).Polish forms have been reported by Nowak (1988) and those of Spain by Tavera et al. (1994) and Olóriz et al. (1995).Remane (1997Remane ( , 1998) ) recently published informative reviews of the state-ofthe-art of the group, providing tables which include the stratigraphic succession of calpionellid species and the correlation of calpionellid, nannofossil and ammonite subdivisions with magnetostratigraphic events.Nearly simultaneously, Blau & Grün (1997a, b) and Grün & Blau (1996, 1997) proposed a revision of the calpionellid zonal and subzonal division.For the Tithonian Stage, they introduced and formally defined two zones and seven subzones; the duration of zones in the Jurassic is less than one million years, that of subzones about 300 000 years.
Important results from the southern Tethyan margin have been contributed by Benzaggagh & Atrops (1995, 1997).These workers provided precise correlation and species range charts for ammonites and calpionellids for the lower part of the calpionellid succession, which previously was poorly known, and clarified the succession of zones and subzones from the Middle Tithonian Semiformiceras fallauxi/Chitinoidella dobeni Subzone to the Upper Tithonian Durangites vulgaris/ Crassicollaria A3 Subzone.An important contribution on the calpionellid faunas of the southern and eastern Tethyan region of Europe was presented by Reháková & Michalík (1997); the western Carpathians and their foreland in Moravia were treated by Řehánek (1990) and Reháková (1995Reháková ( , 2000)).In all these last-mentioned publications, the Middle/Upper Tithonian boundary has apparently been drawn a little too high.Following the results of Benzaggagh & Atrops (1995, 1997), this boundary lies between the Dobeni and Boneti Subzones of the Chitinoidella Zone and not above this zone.

Ostracodes
This group has a rather high stratigraphic resolution and has therefore been used frequently and successfully for the subdivision of sediments in northern Germany, Poland, England, the Netherlands, the North Sea Basin, France and Russia.In a recent monograph, Schudack (1994) revised the ostracodes of the Upper Jurassic in north-west Germany, documenting the correlation possibilities of this group in western, central and northern Europe.The Upper Jurassic of north-west Germany was subdivided into nineteen ostracode zones, representing variable durations (0.25-2.5 Ma ;Schudack 1996a;Gramann et al. 1997).This study also presents a comprehensive list of all important publications on ostracodes.In northern Europe, the papers of Herngreen et al. (1988), Herngreen & Wang (1989) and Guy-Ohlson & Norling (1994) deal with this group in the Netherlands and Sweden, respectively.In Poland, Bielecka et al. (1988) treated the group, and range charts have been published by J. Szteijn (in: Marek & Pajchlova 1997); Danish faunas were described by Christensen (1988).The most recent reviews of European ostracodes are those of Bodergat (1997) on marine ostracodes and Colin (1997) on non-marine ostracodes; see also Hardenbol et al. (1998, chart 7).

Dinoflagellata
Dinoflagellate cysts have become a widely used supplement to ammonites and are of particular importance in the subsurface.In a recent study, Poulsen (1996) emphasised the important role of dinoflagellates in Jurassic stratigraphy while comparing the Upper Jurassic of Denmark and Poland.The marine Upper Jurassic of Denmark was divided into seven zones and fifteen subzones whereas that of Poland was divided into four zones and twelve subzones (Poulsen 1996); the dinoflagellate cyst zonation of the Jurassic of Subboreal Europe is reviewed in Poulsen & Riding (2003, this volume).In Great Britain, Riding & Thomas (1992) have delivered the most recent compilation of dinoflagellates.Other important papers are those of Sarjeant (1979), Riley (1980), Riley & Fenton (1982) and Riding & Sarjeant (1984); one concerning Russia is that of Lentin & Vozzhennikova (1990).In the Netherlands, Herngreen et al. (1988; see also Herngreen & Wang 1989) presented a report on the stratigraphic bioevents based on the first and last appearance of dinoflagellate cyst species which made possible a subdivision into nine zones.In north-west Germany, the Oxfordian and Kimmeridgian has been subdivided into three dinoflagellate zones and eight subzones (Gramann et al. 1997).Detailed subdivisions for the Boreal and Tethyan regions have recently been published by Hardenbol et al. (1998, chart 7).

Calcareous nannofossils (coccoliths, nannolith groups)
Recent advances in Jurassic calcareous nannofossil research have been reviewed by Bowen (1996), who dealt with several general aspects of this group, such as evolutionary succession, species diversity and longevity, distribution and provincialism, which are all important when regarding the utility of the group for biostratigraphic purposes.If conditions are favourable, then it is possible to subdivide the Upper Jurassic into five Boreal nannofossil zones (with six subzones) or three Submediterranean nannofossil zones (with seven subzones); correlation between these two regions is thus still problematic.The calcareous nannofossil bioevents were recently reviewed by Gardin (1997).Subdivisions and correlations for the Boreal/Subboreal and the Tethyan/Submediterranean Provinces can be found in Hardenbol et al. (1998, chart 7).

Charophyaceae
This group of calcareous algae has received new impetus with respect to its potential for biostratigraphy.In a recent publication, the results of a local zonal subdivision based on charophytes in the Lower Saxony Basin of north-west Germany (Schudack 1996b) has been correlated firstly with the new European Mesozoic charophyte biozonation (Riveline et al. 1996), secondly with the subdivisions of other microfossil groups in north-west Germany, such as ostracodes and dinocysts, and thirdly with the old micropalaeontological subdivisions for the Upper Jurassic (Malm) of north-west Germany (e.g.Klingler et al. 1962;Wick & Wolburg 1962).In north-west Germany, from the Upper Oxfordian to the top of the Tithonian, five charophyte zones are now recognised, whereas in other parts of western Europe there are only three (Schudack 1991(Schudack , 1993)).The stratigraphic resolution of this group is not very high in the Upper Jurassic.Each biozone represents a duration of between 0.5 and over 2 million years.The Charophyaceae of western Europe have been revised in detail by Schudack (1993), and a useful compilation of all new data in Europe has been compiled by Riveline et al. (1996); see also Hardenbol et al. (1998, chart 7).

Dasycladaceae
This group seems to be only locally important for biostratigraphy (e.g.Portugal, Italy, Dinarids); a short review was presented by Bassoulet (1997b).

Spores and pollen
The value of pollen and spore grains for stratigraphic subdivision is not very high in the Upper Jurassic.The palynostratigraphy of Sweden (north-west Skåne) was discussed by Guy-Ohlson & Norling (1988) in connection with a study of the microflora of some boreholes.It was revealed that "detailed correlation without the presence of dinoflagellates or other significant taxa appears difficult if not impossible" (Guy- Ohlson & Norling 1988, p. 15).In the Central Graben of the southern North Sea, Upper Jurassic sediments were subdivided into four zones on the basis of sporomorphs (Herngreen et al. 1988;Herngreen & Wang 1989).In north-west Germany, the Upper Jurassic was divided into four zones using spores and pollen (Gramann et al. 1997).
The group apparently has its greatest importance at the system boundaries; it has been used successfully at the Triassic-Jurassic boundary and, to a lesser degree, at the Jurassic-Cretaceous boundary.

Sequence chronostratigraphy
Sequence stratigraphy is gaining in importance in chronostratigraphic correlation, as illustrated recently by the presentation of a framework for the European Mesozoic and Cenozoic basins (Hardenbol et al. 1998).Many data have been used and compiled in charts, two of which are important for the Upper Jurassic.They demonstrate the sequence chronostratigraphy (Sequences, T-R Facies Cycles, Major Transgressive-Regressive Cycles) for the Boreal and Tethyan realms combined with the ammonite biochronostratigraphy and magnetostratigraphy, plotted against the time scale.
The Upper Jurassic of Europe starts in the upper half of the transgressive part of the Second Major T-R Cycle (1st order cycle, named North Sea Cycle) in the Jurassic and ends within the regressive phase of this cycle.A total of 21 sequences (3rd order cycles) have been recognised (Ox 0-8, Ki 1-7, Ti 1-6) and three 2nd order T-R cycles (T8b-R10b) in the Boreal area, whereas the number in the Tethyan area is somewhat lower.A detailed overview of the North Sea Cycle in Europe (from the North Sea to south-east France) has been presented by Jacquin et al. (1998); marginal areas have been studied as follows: East Greenland (Surlyk 1991;2003, this volume), Portugal, Lusitanian Basin (Leinfelder & Wilson 1998), Portugal and Spain, South Iberian Margin (Olóriz et al. 1991), south-east France (Jan du Chêne et al. 2000), Switzerland (Gygi et al. 1998), West Carpathians (Reháková 2000) and Russia (Sahagian et al. 1996).
Fig. 3.A tentative correlation chart for some alternative subdivisions of parts of the Oxfordian Stage in Mediterranean and Submediterranean Europe.
It can be concluded from the above that correlation of the A. bauhini and P. densicostata horizon with the A. subtilicaelatum and A. bayi horizon is possible, but the vertical ranges of the former species may be longer and the correlation may thus be only partial.Consequently, the position of the upper boundary of the A. bauhini Zone and the lower boundary of the A. bayi horizon require more precise definition.(b)In the sequence between the A. bauhini and the A. subtilicaelatum horizons, equivalent to the middle part of the Pictonia baylei Zone, the P. baylei horizon of Normandy and the upper A. bauhini-bearing beds in Scotland (e.g.bed 38 with Pictonia sp.,Wright 1989) could be expected.They may have their equivalents anywhere in this succession, whereas other parts of the Submediterranean succession are not represented in the Subboreal sections or only by gaps.
Subdivision of the Upper Jurassic Series of Europe into stages, substages and zones.Substage usage varies in the literature, dependent on author; those indicated are only examples.

:
Dorsoplanites primus, Pavlovia iatrensis, Pavlovia rugosa, Pavlovia communis and Dorsoplanites liostracus characterise the lower part of the Middle Volgian, whereas Dorsoplanites gracilis, Epipallasiceras pseudapertum, Crendonites anguinus, Laugeites groenlandicus and Epilaugeites vogulicus are represented in the upper part.The lower part of the Upper Volgian Praechetaites tenuicostatus Zone of East Greenland may correspond to the uppermost part of the Middle Volgian, the upper Paracraspedites oppressus Zone of England and the lower Praechetaites exoticus Zone (= lowermost Craspedites okensis Zone sensu lato) of northern Siberia.