The Campanian-Maastrichtian Boundary G. S. Odin (editor) 2001 Elsevier Science B.V.

CHAPTER F2

The Campanian-Maastrichtian boundary: correlation from Tercis (Landes, SW France) to Europe and other continents

G. S. Odin

Sommaire

Des marqueurs de correlation ont ete signales a maintes reprises dans les pages precedentes; ils sont utilises dans ce chapitre pour un tour d'horizon sur les sections qui, dans le monde, renferment un enregistrement sedimentaire de la limite entre les Etages Campanien et Maastrichtien.

Les correlations entre Tercis et les stratotypes historiques des Etages sont possibles grace aux nannofossiles calcaires et, surtout, aux kystes de dinoflagelles pour les microfossiles, aux ammo­nites, aux echinides et aux asterides pour les macrofossiles. Les niveaux du stratotype du Cam­panien (Coquand, Champagne charentaise) correspondent a 1'Unite d'Avezac. Le stratotype historique de Dumont pour le Maastrichtien se situe au-dessus de la partie exploitee de la carriere de Tercis. La limite moderne se situe dans la sous unite a silex clair de I'Unite Les Vignes, precise-ment dans I'intervalle assez large correspondant a la lacune de definition entre les stratotypes his­toriques des Etages.

Afin de renforcer 1'information stratigraphique autour de la limite, des sections auxiliaires sont examinees sans que ces sections n'aient un role a jouer dans la definition elle-meme. Elles n'ont pas ete soumises a un agrement formel mais les stratigraphes pourront y puiser les informations qui leur servent habituellement et qu'ils ne trouveraient pas a Tercis. Une correlation precise avec la reference globale de Tercis est necessaire. Avec la section des Apennins, remarquable pour sa magne-tostratigraphie, les nannofossiles calcaires et les

foraminiferes planctoniques fournissent des lignes de correlations coherentes (emergences et extinc­tions de Rd. calcarata, de Q. trifidum et de Q. gothicum; extinctions de A. parens parens et de A. parens eonstrietns). Elles se combinent harmo-nieusement avec la limite entre les magnetozones 33 et 32 pour permettre, entre les cotes 40 et 160 notamment, de transposer avec une incertitude de quelques metres seulement les informations ma-gnetostratigraphiques depuis les Apennins vers Tercis et les informations des macrofossiles, notamment, depuis Tercis vers les Apennins.

La succession boreale de Lagerdorf-Kronsmoor-Hemmoor (pres de Hambourg, Allemagne) fut proposee pour servir de reference pour le Cretace Superieur boreal. La limite Campanien-Maastrich-tien est dans I'affleurement de Kronsmoor qui ne pent servir comme reference globale du fait du petit nombre d'outils biostratigraphiques utilisables hors du Domaine Boreal voire simplement regio-nalement (belemnites a apparition diachrone). La correlation avec la succession de Tercis est delicate; la presence d'un enregistrement digne de confiance des compositions isotopiques du strontium en Allemagne n'est d'aucun secours a cet effet. Les ammonites sont absentes ou tres rares, de meme que les foraminiferes planctoniques de large repar­tition; les nannofossiles calcaires sont insuffisamment decrits ou peu abondants. Des possibilites existent du cote des dinokystes, des foraminiferes benthiques et, peut-etre, des micro-brachiopodes.

Pour I'instant, nous n'avons pas reuni d'argu-ments de correlation directe suffisants mais des

806

moyens de correlation indirecte sont envisages tels que la sequence magnetostratigraphique (une inter­pretation peu assuree en Allemagne), Texamen des taux de sedimentation (une approche risquee) et un point cle biostratigraphique adopte depuis 1'obser­vation, en 1992, de la disparition de A . hyatti contemporaine de 1'apparition de B. lanceolata en Pologne. Ni cette coincidence, ni meme le caractere instantane de 1'apparition de B. lanceolata en differentes regions du Domaine Boreal ne sont acceptes par tous les auteurs. Ainsi, Atabekian (1995) soutient qu'il y aurait (ample) coexistence de ces deux taxons cle ce qui ferait de A . hyatti un indice de Campanien eleve mais aussi de Maas-trichtien (au sens boreal classique). Des problemes de taxonomie ne sont peut-etre pas etrangers a ces divergences de vue. Pour le moment, I'utilisation des correlations indirectes notees ci-dessus permet d'aboutir a un schema de correlation tel que la limite modeme a Tercis serait sub-contemporaine de la limite boreale usuelle; ce schema n'est pas contredit par les quelques informations sur les ammonites communes a Tercis et a 1'Allemagne du Nord.

On notera que le taxon cle P. neubergicus n'a ete implique dans aucune des correlations realisees entre les sections citees jusqu'ici.

II existe d'autres sections comprenant la limite Campanien-Maastrichtien dans le monde. L'une d'entre elles est situee dans la vallee de la Vistule (Piotrawin, Pologne), une autre en Irlande (Antrim). D'autres existent dans les massifs entou-rant la Mer Noire et la Mer Caspienne (Crimee, Daghestan, Mangyshlak, Kopetdag). Dans la Tethys, les sections en bord de mer du Pays Basque et, a I'interieur, en Navarre (Espagne) ainsi que les affleurements au Sud du Kef (Tunisie) qui con-tiennent a la fois ammonites, foraminiferes planctoniques et nannofossiles calcaires peuvent foumir d'interessantes comparaisons. II existe une variete de sections qui couvrent la limite d'Etage si Ton s'en refere aux microfossiles: Plate-forme Arabe, Amerique Centrale etc. Les sections de la Gulf-Coast (Sud des Etats-Unis) et du Western Interior (Centre de 1'Amerique du Nord) sont particulierement interessantes puisque, par de rares ammonites cosmopolites et par les foraminiferes planctoniques, des liens peuvent etre etablis

directement avec Tercis tandis que des liens indirects peuvent aussi etre proposes entre les ages numeriques mesures en Amerique du Nord et estimes a Tercis. Des sections plus eloignees sont disponibles dans la presqu'ile de Sakhaline et au Nord du Japon, en Inde, en Antarctique.

En bref, les efforts du Groupe de Travail Maastrichtien pour caracteriser la succession de Tercis par une variete d'outils stratigraphiques ont ete couronnes de succes dans la mesure oii les connaissances ainsi rassemblees permettent de tracer des lignes de correlation avec des bassins de depot diversifies, ce qui fait de Tercis un site particulierement adapte pour abriter une section de reference.

1. Introduction

Correlations between Tercis and other sections are summarised here. The first question of interest is the comparison between the newly proposed boundary level and the historical stratotypes of the Campanian and the Maastrichtian stages. This question is a fundamental one for ensuring con­sistency between the former and the future definition.

The second question of interest concerns the correlations between Tercis and possible auxiliary sections. The designation of such sections is not encouraged by the Commission on Stratigraphy: at least from a formal point of view, the stratotype Point (GSSP) must be unique. The suggestion of several "type sections" even with some sort of hierarchy between them (holo-stratotype, auxiliary . . .) may lead to confusion if the correlation is not perfectly established, and it is difficult to arrive at a satisfactory convention in this matter. Correlation lines can only be expressed with some uncertainty which depends on the stratigraphical tool used and on local circumstances. This is why auxiliary sections are commented on below without formal statute; no formal vote was taken on them although the principle of searching for such auxiliary sections had been favourably considered during several meetings. Auxiliary sections increase the quality of the definition of the stage boundary by considering stratigraphical tools absent from the type section.

807

The third question of interest is the correlation Hnes between Tercis and other sections known to document the same interval of time in a variety of palaeogeographical and environmental places.

Finally, potential areas for correlation are pre­sented briefly in terms of palaeogeographical distribution of several key fossils.

2. Tercis with regard to historical stratotypes

2.1. Correlation between Tercis and the Campanian historical stratotype of Charente

Ammonites of the Charente area (figure 17 in Neumann & Odin, chap El) can be used to correlate the succession at Tercis with the Campan­ian, first defined by Coquand. For example, the genus Hoplitoplacenticeras is present in the Char­ente area up to the foraminiferal biozone C IVb (with the species H. dolbergense also present at Tercis and with another species possibly up to biozone C V). The fauna of Unit Q (Campanien 4, C VI foraminiferal biozone) would contain Pseu-doxybeloceras cf. interruptum, Nostoceras (B.) polyplocum, and Trachyscaphites pulcherrimus. This fauna is consistent with that found in the interval between levels 5 and 40 at Tercis.

Nostoceras archiacianum is quoted from Unit R according to Lambert (1981). Unit R corresponds to the foraminiferal biozone C VII. Ward & Orr (1993) quoted the presence of A . cf. archiacianum in the inoceramid-rich bed while Kiichler (in Kiichler & Odin, chap D4e) locates a A . archiacia­num zone between levels 30 and 65 at Tercis. Finally, one specimen of Nostoceras hyatti has been quoted from the "Couches de Maurens" which would correspond to the Dordonian of Coquand and that Platel (appendix 2 in chap. El) now locates in a new "Campanien 6" unit.

The information about the historical stratotype of the Campanian of Charente is clear: the levels with Bostrychoceras polyplocum are Campanian; levels above C VIII are automatically imputed to the Dordonian later considered equivalent to the Maas-trichtian. Since the work by Blaszkiewicz (1980) it has become apparent that there were two ammonite biozones between the polyplocum zone and the neubergicus zone. Blaszkiewicz proposed to intro­

duce the Didymoceras donezianum zone and the Nostoceras pozaryskii zone in the Polish succes­sion of the Vistula Valley near Piotrawin. The A . pozaryskii zone was considered Campanian in 1980 because the regional practise was to locate the base of the Maastrichtian at the FO of Belemnella lanceolata in Poland, and the latter signal was found immediately above the A . pozaryskii zone. Two questions can be commented on. The first question concerns the correlation between Poland and Charente. The Campanian age commonly accepted since Blaszkiewicz for the A . hyatti zone seems mistaken in that its real age was simply "pre-FO of B. lanceolata". This corresponds to the Dordonian of the stratotype area of Charente (excluded from the type Campanian), which means that the A . hyatti zone should have been considered Maastrichtian in the acceptance of the Coquand's stage boundary. The second question is that if the Polish succession is continuous, like in Tercis, and if the A . pozaryskii zone is actually equivalent and contemporaneous with the A . hyatti zone of Tercis, then the FO of B. lanceolata can be located at the level of the LO of A . hyatti at Tercis, i.e., about 1 metre below the stage boundary recommended in this volume.

In summary, and considering the ammonite information alone, the historical definition of the Campanian stage and the definition in the Boreal Domain were not consistent because the "Campan­ian" pozaryskii zone of Poland was "Dordonian" in age. The definition was more consistent in the Tethyan Domain considering that the LO of Rd. calcarata might well be contemporaneous with the top of the Campanian of Coquand. It is worthwhile noting here that, in agreement with Hancock & Kennedy (1993), one of the major interests of the succession at Tercis is that the interval which comprises the whole A . hyatti biozone of ques­tionable chronostratigraphic age is well documented.

The echinoid fauna gives information with the occurrences of Offaster pilula, from the upper Campanian in northern Aquitaine (biozones C IVb-C V); these would correspond to the lowest occurrences of this taxon at Tercis (levels 1 to 15). However, the same taxon has also been identified in levels above: the upper d'Avezac and lower Les

Vignes Units. In England, O. pilula is associated to Micraster coranguinum as at Tercis and this pair of taxa could be used as a correlation tool. One must note, finally, an O. pilula zone in Germany (lower lower Campanian) but this zone does not corre­spond to the levels quoted above. A second echinoid correlation line is suggested between the occurrence of Echinocorys ovata of Charente (C V-C Via) and that of the morphologically closely-related Echinocorys conoidea of Tercis between levels 52 and 66 at Tercis (Neraudeau & Odin, this volume, chap. D6).

The asteroid faunal assemblage at Tercis (figure 4 in chap. D5a) shows biogeographic affinities with the northern Aquitaine and boreal domains. Corre­lation lines include the evolutionary appearances of Metopaster tercensis nov. sp. (at or below level 78.5 at Tercis; Campanien 5 of Charente), and

possibly also the regional first occurrence in Aquitaine of Nymphaster tethysiensis nov. sp. (level 46.4 at Tercis, Campanien 3 of Charente?). Additional knowledge would make these correla­tions more precise.

Calcareous nannofossil data are reviewed by Neumann & Odin (chap. El); the co-occurrence of Q. gothicum and Q. trifidum has been documented in foraminiferal biozones C V (top) to C VIII. Nannoconids are quoted from the same biozones (figure 14 B in chap. El) and if their LO is contemporaneous in Charente and at Tercis, this locates the top of the Campanian of Charente below level 90 at Tercis. The LOs of E. eximius and R. anthophorus from Charente do not have an evolu­tionary meaning.

Dinoflagellate cysts are present and diagnostic both at Tercis and in northern Aquitaine (Archiac,

Tercis level

120-

100-

80-

60-

-

40-

-20-

-

0--

Ammonites

? Mussidan

7

•B $5

z:

? Maurens

Unit 7

E 13

.eg

.5

En 3

8 Q.

8.

CO

z

evil

cvi T evil A to

C V § ! CIVa Q .

cdi ^

_ ^ d "5. o I

CIVb - CV?

Nannofossils

E

'sz

z D )

d + E

E 3

1 o

CVIII

to

.CV CO 3

o o o o c c CO z

Echinoids

Ijcvia l\oM

co~r ^ • Q .

d .

CV

CIVb

Dinocysts

CO 0) T3 O

O CO 3

f^CVIII CO

'o CO

8 X o —1

S^VIa CO

CO

8 d q —1

North Aquitaine

biozones or lithological units

z < z |

CO

c 3 CO

2

o o

"Campanien 6" ^

^

' * , ' ' '-.X

,_

5 o 5 o

> o

> o

^ > o

cc

O

CO Q .

Campanien 5

^ c CO 3

1 Campanien 4 ^

< Z < Q-

< o

Campanien 3

'GSO'200o\

Fig. 1. Correlation between Tercis and the historical stratotype of Charente. Additional information has also been obtained from asteroids. Vertical lines show the location of the fossils at Tercis; the corresponding biozones of Charente are noted beside.

809

Aubeterre). The taxon Palaeohystrichophora infu-sorioides becomes very rare from level 38 or so at Tercis; it is present up to the top of the for-aminiferal biozone C V of N Aquitaine. Similarly, Odontochitina costata disappears at about level 39 at Tercis and in biozone C Via at Aubeterre. Finally, Xenascus ceratioides is present up to the top of biozone C VIII in northern Aquitaine (figure 15 in chap. El) and disappears between levels 86.5 and 95.6 at Tercis. This suggests that the Campanian of Coquand lies below level 90 at Tercis in agreement with the information provided by the nannoconids.

22. Correlation between Tercis and the Maastrichtian historical stratotype ofLimburg

The question of the base of the Maastrichtian stage in Limburg is made complex by the subsequent changes in the original definition. It can be simplified by strictly considering the historical definition. The "Systeme Maestrichtien" of Dumont corresponds to the Maastricht Formation of Limburg and to the underlying Lanaye Member of the Gulpen Formation (figure 2 in Jagt, this volume, chap. E2) . These sediments are in the mid to upper portion of our modern Maastrichtian stage. They were deposited several Ma after the LO of the ammonite A . hyatti or the acceptably sub-con­temporaneous FO of Belemnella lanceolata. The original Maastrichtian levels thus correspond to levels located above the ones studied in the main section at Tercis that biostratigraphical studies restrict to the lower Maastrichtian. Therefore, there is a hiatus between the historical stratotype of the Campanian of Coquand and the historical strato­type of the Maastrichtian of Dumont; it corresponds to the succession between levels 90 and 174 at Tercis and still to a portion of the levels above. There is thus room to choose a level in between these two stratotypes and the definition presented in this volume does not contradict the historical definitions.

Incidentally, the successive usages of the stage name Maastrichtian in NW Europe are discussed by Jagt (figure 2 in chap. E2) who concludes that the most recently used boundary level in the Boreal Domain (FO of B. lanceolata) is poorly recorded in the Beutenaken Member of the Gulpen Formation (see also figure 4 in chap. E5a). A biostratigraphical

connection can be suggested between the Campan-ian-Maastrichtian boundary newly defined at Tercis and the Limburg area, through the distribution of the dinoflagellate cysts. Antonescu et al. (this volume, chap. C2d) note several bio-horizons of interest. Some of them (the LO of the genus Odontochitina; the LO of Raetiaedinium trunci-gerum; the FO of T. evittii; the FO of A. varium) are found "between the top of the B. langei Zone and the base of the pseudobtusa to obtusa Zones of Limburg". This is essentially within the B. lanceo­lata zone which is of short duration (about 0.5 Ma, see figure 2 below). A direct correlation is thus possible from Tercis to the belemnite-documented succession of Limburg thanks to the powerful dinocySt biostratigraphical tool.

3. Tercis with regard to auxiliary sections

3.1. Correlation from Tercis to the section of the Bottaccione

The direct correlation between the section of the Bottaccione and Tercis can be done using both calcareous nannofossils and planktonic foramini-fera. Five taxa are useful: Radotruncana calcarata, Quadrum trifidum and Q. gothicum, Aspidolithus parcus parens and A. p. constrictus. In addition, the magnetostratigraphic limit between magnetozones 33 and 32 is useful. Combination of these data led Lewy & Odin (figure 2 in chap. B2d) to propose a continuous correlation between levels 40 to 160 at Tercis and metres 288 to 332 in the section of the Bottaccione. The uncertainty on these correlations is about 0.1 Ma as an order of magnitude (figure 2 and §4 in chap. E5c). This is not precise in terms of high-resolution stratigraphy; however, the magne­tostratigraphic signal is correlated with better precision. The four consistent correlation lines for a duration of deposition of about 5 Ma bracketing the newly defined stage boundary precisely constrains the stage boundary itself which is concluded to be at metre level 320 in Italy with an uncertainty of ± 3 metres.

The duration between the key-horizons docu­mented in both successions and the stage boundary can be estimated using the mean rates of deposition (Tercis: 25 m/Ma; Bottaccione: 9 m/Ma). The metre levels shown by Gardin, del Planta et al.

810

Table 1. Estimates of the time elapsed between six key horizons and the Campanian-Maastrichtian boundary as documented at Tercis and at the Bottaccione (stage boundary at the Bottaccione as previously suggested by Premoli-Silva & Sliter, 1994).

Correlation line

LO A. p. constrictus LO A. parcus parcus LO Q. trifidum LO Q. gothicum Ca-Maa boundary boundary 33/32 LO Rd. calcarata

Position

Tercis level

160.1 142.8 ? 134.2 129.1 115.2 81 62.5

Bottaccione

332 330 329 329 330 (327.5) 307 297

Time/stage boundary (Ma)

Tercis

+ 1.8 +1.1 +0.8 +0.5

-1.4 -2.1 (to-2.6)

Bottaccione

+0.2 (+0.5) +0.0 (+0.3) -0.1 (+0.2) -0.1 (+0.2)

-2.6 (-2.3) -3.7 (-3.4)

Difference

+1.6 +L1 +0.9 +0.6

+ 1.2 <1.6

(figure 2 in chap. E4) are used in table 1 for the nannofossil bio-horizons together with the stage boundary previously suggested for the Bottaccione section by Premoli-Silva & Slitter (1994, their table 9: level 330; but possibly 327.5 after their table 6). The difference in age (Difference in table 1) always shows the same trend: the boundary is younger in the section of the Bottaccione; this documents a good consistency in the relative location of the bio-horizons between the two sections. This also indicates that the time-equivalent Tercis boundary should be located below metre level 330 in the section of the Bottaccione. A mean difference of LI Ma can be suggested with an uncertainty of ±0.2 Ma corresponding to the sum of the uncertainty on the mean rate (optimistically 10% of the value) and of the precision about the location of the horizons (±0.1 Ma or so). The location near metre level 320 is thus obtained.

3.2. Correlation from Tercis to Kronsmoor

The Kronsmoor section (50 km North of Hamburg, northwest Germany) has been studied in many publications (Kennedy et al., 1995). Schulz (1978) published a first paper on the lithostratigraphy and biostratigraphy of the Campanian and Maastrich-tian of Lagerdorf and Kronsmoor. Ernst & Schulz (1980) gave a description of the quarries at Lagerdorf and Kronsmoor for an Excursion of the International Geological Congress in Paris. Surlyk (1982) studied the brachiopods from the Campan­ian-Maastrichtian profile near Kronsmoor and proposed 12 brachiopod zones for late Campanian and Maastrichtian deposits of NW Europe (Surlyk, 1984). There is a report by Ernst (1984) on a large

sample series from the white chalk section of Lagerdorf and Kronsmoor (Campanian through Maastrichtian). Schultz et al. (1984) studied the Coniacian to Maastrichtian Stage boundaries in the section which was proposed as a standard one during the International symposium at Copenhagen (1986); an abstract was published the same year in "Geologic mediterraneenne" (1984). Willems (1988) documented the calcareous dinoflagellate cysts from the Upper Cretaceous white chalk facies from the Coniacian to the Maastrichtian. Schonfeld et al. (1991) documented the oxygen isotope composition of Upper Cretaceous chalk at Lager­dorf. Mac Arthur et al. (1992, 1993) proposed a correlation of the Campanian-Maastrichtian bound­ary between Kronsmoor and the Norfolk area (UK) based on Sr isotopic ratios related to cephalopod biostratigraphy. The belemnite control and the strontium isotopic record are probably the major interests of the Kronsmoor section. Therefore, Kronsmoor would be appropriate for locating an auxiliary section.

Belemnites are the major group used in many outcrops of the Boreal Domain. The Kronsmoor succession is poor in ammonites. Amongst them, Hoploscaphites constrictus is the most quoted one at the base of the lanceolata zone but also the most rarely documented one and this occurrence is questioned (Schulz et al., 1984). N. polyplocum is another ammonite quoted from the underlying beds at Lagerdorf. Schonfeld & Schulz (1996) noted that the base of the section at Kronsmoor is located at the base of the langei belemnite zone which is equated to the top of the polyplocum ammonite zone. The same paper showed the key ammonite

ni

present only at the bottom of the homonym zone and the fossil is generally considered rare in NW Germany. If one postulates a similar age for B. polyplocum at Tercis and Kronsmoor, then the polyplocum-bearing levels at Tercis (between lev­els 20 and 45) must be located much far below the lowest beds exposed at Kronsmoor (figure 2). No other ammonites are quoted from the boundary

interval at Kronsmoor or in the area, which makes the richness at Tercis of no use for direct correla­tion. Indirect correlation can be suggested via the last occurrence of Nostoceras hyatti in westernmost outcrops (figure 3 below). Brachiopods similar to those from Kronsmoor are present at Tercis but their distribution is not yet firmly established (Gaspard & Odin, this volume, chap. Dl).

Fig. 2. A correlation between the relative time-intervals covered by the sections at Kronsmoor and Tercis. Numerical ages are estimated Ma (est. Ma).

812

Concerning microfossils, the report on calcar­eous nannofossils in Schonfeld & Schulz (1996) was short and did not give the location of the bio-horizons. The Sissingh's scheme (CC zones) applied in Germany is not appropriate at Tercis. It is thus not possible to utiHse these published data but correlation between Tercis and Kronsmoor could presumably be obtained in the future by means of calcareous nannofossils. The FO of Belemnella lanceolata is located at Kronsmoor in the CC 23A nannofossil biozone and could be contemporaneous with the stage boundary recom­mended at Tercis which probably lies in this zone as well (figure 3 in chap. C3e).

Benthic foraminifera were considered to be useful by Schonfeld & Schulz (1996). Unfortu­nately, the bio-horizons quoted near the regional Campanian-Maastrichtian boundary are not similar to those recorded at Tercis. The same genera or species are present in the two successions, such as Bolivinoides, Gavelinella clementiana, Bolivina incrassata. The FO of the latter is contempora­neous with the occurrence of A . polyplocum in Poland, NW Germany, and England. The signal is present at Tercis and might be consistent but problems exist in the identification of this Bolivina (Tronchetti et al., this volume, chap. C4c). Addi­tional data are needed for this fossil group. A German expert was asked to study the material from Tercis, but time was lacking for the study to be achieved.

The physico-chemical tools can only give indi­rect correlation lines since these tools are equivocal and need to be supported by dating evidence. The magnetostratigraphic information discussed in Schonfeld & Schulz (1996) was pessimistic; the authors suggested that the previously published magnetic record was not reliable in view of their new data. An optimistic view of the German record is proposed by Lewy & Odin (figure 1 in chap. B2d). If this interpretation is correct, the limit between the grimmensis/granulosus and the lanceo­lata macrofossil zones of Germany would lie in a normal magnetozone identified as 32N2. This is consistent with the location obtained from the correlation between Tercis and the section of the Bottaccione (mid portion of magnetozone 32N2). According to the interpretation of Lewy and Odin,

the limit between magnetozones 33 and 32 would not be far from the base of the langei macrofossil zone located in a 4 m-thick sedimentological gap below the Kronsmoor succession (Schultz et al., 1984). At Tercis, the same limit is documented between levels 80 and 81.

A correlation between the Tercis and Kronsmoor successions has been proposed by Odin & Turpin (figure 2 in chap. B2b) for comparison of the Sr isotopic record in the two areas. For this compar­ison, the stage boundary has been taken at the same level (level 115.2 at Tercis coinciding with the limit grimmensis-lanceolata). If one uses the mean rate of deposition of 25 m/Ma at Tercis (which is well documented) and 24 m/Ma at Kronsmoor (which is the mean rate calculated for the 270 m-thick Campanian deposits accumulated at Lagerdorf and Kronsmoor between -83 Ma and -72 Ma) the result obtained is shown in figure 2. The limit between magnetozones 33 and 32 interpreted at Kronsmoor falls at level 70 at Tercis in figure 2; this is not so far from the measured position (level 80-81) and the relative depositional rates accepted in the two sections are not far from reality. This exercise is thus a realistic, though imprecise, way of correlat­ing the two areas until more correlation lines are documented.

In summary, the biostratigraphy at Kronsmoor is established on regional faunal grounds (NW Ger­man or NW European domain) and cannot yet be correlated to other areas with precision. In fact, the use of regional chronostratigraphical units has often been considered sufficient within NW Europe and no attention has been paid to a comparison with other areas. In addition, and in spite of repeated efforts in this way, the Working Group could not benefit from revised information on this succession for this volume. It is believed that additional microfossil investigation undertaken in parallel in the two areas will improve the situation in such a way that Kronsmoor could be used with confidence as an auxiliary section.

4. Successions documenting the Campanian-Maastrichtian boundary

Correlation lines have been established above from Tercis to historical stratotypes on the one hand and

813

to two well known successions on the other hand. The result of this work is summarised in figure 3. The Maastrichtian historical stratotype succession is not shown because the levels designated by Dumont's original definition are located above the levels illustrated in the figure (see chap. E2). Major key bio-horizons are shown and the diversity of the fossil groups quoted in the succession at Tercis (ammonites, foraminifera, calcareous nannofossils, dinocysts, asteroids) proves the correlation poten­tial of this succession. Incidentally, the occurrence of P. neubergicus at Tercis is not used for correlation toward any of the successions con­cerned in the paragraphs above.

4.1. North and East European successions

The Campanian quarry, south of Piotrawin village in the Vistula River valley (Poland), is a section covering a portion of the succession at Tercis (Blaszkiewicz, 1980; M. Machalski, personal com­munication). The most quoted opinion is that the Nostoceras hyatti zone is documented there while B. lanceolata is found in strata slightly higher in small pits in the village; thus, the usually accepted regional boundary is somewhere in between. The evolutionary appearance of Pachydiscus neubergi­cus is not recorded (taxon absent); index species of planktonic foraminifera are absent (Peryt, 1995). Benthic foraminifera are diversified and the taxa used for the foraminiferal zonation recorded at Kronsmoor by Schonfeld & Schulz (1996) are present including Bolivina incrassata, Globor-otalites hiltermanni, Bolivina decurrens, Bolivinoides paleocenicus. This fossil group sup­ports the opinion that the stage boundary would be in the quarry itself (Peryt, 1995). Other authors in agreement with W. J. Kennedy are of the opinion that the ammonite fauna documents the hyatti zone with the occurrence of N. hyatti and A . helicinum\ but the original identification of these taxa as a distinct species: Nostoceras pozaryskii is still supported by A. Blaszkiewicz (personal commu­nication to M. Machalski, 1998). Belemnites collected there were submitted to identification in 1998 but there were not enough individuals for statistic treatment and identification. Other experts believe that the hyatti zone would be partly

Maastrichtian because this fossil occurs above the FO of B. lanceolata in some regions including European Russia (Atabekian, 1995). The same could be concluded from the presence of A . helicinum "13 m above" "the entrance date of P. neubergicus" in N Spain (Ktichler, 1995). Addi­tional information is thus required from the Polish succession where magnetostratigraphic signals have been observed (Hambach et al., 1995). A first conclusion is that individual belemnites are difficult to use for a firm location of a stage boundary.

Concerning other countries, J. Hancock (per­sonal communication, V-2000) notes that the section in the Bally castle area of County Antrim in Northern Ireland straddles the Campanian-Maas-trichtian boundary but there are too few dating indicators.

J. Hancock (op. cit.) also recalls the presence of fossiliferous (macrofossils) sections in the Man­gyshlak hills (north-west Kazakhstan, Naidin et al., 1984), including two he saw at Sulukapy and Aksyirtau. In the same area, Naidin & Benjamovski (1995) quoted the Aktulagaj section (Ural-Emba region of the Caspian Syneclise); this latter section shows the transition between the Campanian and the Maastrichtian stages with Belemnitella gr. langei and Pseudokossmaticeras galicianum\ Menuites ambiguus, Pachydiscus oldhami, P. sto-baei, and P. subrobustus in the uppermost Campanian and Belemnella gr. lanceolata, Acan-thoscaphites tridens, H. gr. constrictus in the basal Maastrichtian deposits. Again, there seems to be some difficulty in identifying belemnites and Hoploscaphites constrictus (abbreviation: gr.). Ata­bekian (1995) quoted the sections of Mangyshlak and nearby outcrops of "West Kopetdag, North Caucasus (Daghestan), and other regions from the European part of Russia" to document the fact that A . hyatti is Maastrichtian in age.

In the same area, the FO of Contusotruncana contusa would be contemporaneous with the FO of Belemnella lanceolata in basal Maastrichtian deposits from Crimea (Alekseev & Kopaevich, 1997); the latter authors published a spectacular picture of the Beshkosh section (Crimea), a cliff which covers the whole Maastrichtian (121 m thick), a portion of the Campanian, and a portion of the Danian.

814

CAMPANIAN (Coquand)

Campanien 3

P3

Campan 4

C IVb C V C VI

Campanien 5

: S . DORDONIAN (Coquand)

^:"Campan 6"

evil C VIM Maurens Mussidan

^ LO Hoplitoplacenticeras -I

B. polyplocum

LO Odontochitina costata

FO Metopaster tercensis

Quadrum trifidum + Q. gothicum

LO Nannoconids -<-J

N. hyatti

O

> m z m > m >

CAMPANIAN MAASTRICHTIAN

D'AVEZAC UNIT pp

grey I glaucony-poor glaucony -rich

0 |

gl. -poor gl.-rich/poor

LES VIGNES UNIT pp

pale flint dark flint

00 | 0

o

LOO E. eximius ^\"

LO O. costata ^ ' —:?

LO E. eximius

-T^ I Rd. calcarata

ro lO

TRZ Quadrum trifidum LO nannoconids

R. anthophorus

B. polyplocum

^ LO Hoplitoplacenticeras

Nostoceras hyatti & allied forms

FO Metopaster tercensis

I W ^ I

LO Q. gothicum O LO A. p. constrictus >

2

5' o c

R. levis

ro 2

» <D

m o CO CO

m I go m CO

Pachydiscus neuberglcus

.4^

o - • P LOQ. goth. w j ^ FO Reinhardtites levis o '

GO

O

ammonite / belemnite assumed intra-boreal connection

FO Belemnella lanceolata

LO Nostoceras hyatti

815

4.2. Other sections comprising the Campanian-Maastrichtian boundary

The nearest successions comparable to the Tercis one can be found in the French and Spanish Basque Region (Zumaya, Bidart: beach, Hendaye: Baie de Loya). These sea-side outcrops contain micro-fossils; macrofauna is rare near the Campanian-Maastrichtian boundary at Zumaya. Tectonic fea­tures are common at Bidart and they disturb the stratigraphical continuity (Bilotte et al., 1999). Sections from another area of N Spain (Navarra) have been studied (Kiichler & Kutz, 1989; Kutz, 1995; Kuchler, 1998, and this volume, chap. E3). Contemporaneous macrofossils (ammonites, echi-noids, brachiopods) and microfossils (calcareous nannofossils) are useful for comparison of the distribution between Navarra and Tercis located 80 km to the North.

In Tunisia (South of El Kef), the Campanian-Maastrichtian sediments contain a good microfauna (Bellier et al., 1983, Salaj & Maamouri, 1984; Salaj, 1988; Salaj & Wiedmann, 1989; Li & Keller, 1998); macrofauna is relatively rare but ammonites are present. The study of this region has not been sufficiently diversified so far, but Tunisia could be used for comparison using different correlative tools. More to the East, spectacular outcrops of upper Cretaceous marls bracketing the Campanian-Maastrichtian are available in Turkey (Cros et al., 1991) and Syria (E. Fourcade, personal commu­nication, V-2000). These basinal deposits are rich in planktonic foraminifera (including Rd. calcar-ata) and calcareous nannofossils.

Finally, it is worth mentioning the situation in the Arabian Peninsula where Platel studied the geology of Oman (Philip & Platel, 1998; Platel et al. 1992; 1994a; 1994b). In central-eastern Oman (Huqf area), the Campanian-Maastrichtian boundary (chalk facies) crops out in the Nafun section. The section comprises lower Campanian to lower Maastrichtian (Gansseri zone) deposits. The basal

Samhan Formation contains ammonites, echinoids, rudists, foraminifera; the chalk facies corresponds to the overlying 400 m-thick Fiqa Formation where planktonic and deep benthic foraminifera and calcareous nannofossils are present together with crinoids. Rd. calcarata is not quoted among the few diagnostic forms identified.

To the West of Europe, the uppermost Creta­ceous Andean deposits of South America are generally of continental origin but Campanian-Maastrichtian shallow marine deposits are described from Colombia (J. Bourgois, personal communication, 2000). They are known as the Guadalupe Formation, or the Colon, Mito Juan, Catatumbo Formation; they are clastic in nature, may reach 1 km in thickness and have been dated by a palynological approach (Boinet et al., 1985).

From Central America, the Haitian Campanian-Maastrichtian deposits display correlation between calcareous nannofossils, planktonic foraminifera and volcaniclastic datable levels (see chap. E5b). To the South-East of Mexico (Chiapas Province) and in Guatemala, marine deposits corresponding to the Campanian-Maastrichtian boundary repre­sent two kinds of environments (Fourcade et al., 1999). The basinal succession is several hundred m thick (the Sepur Formation) and has been dated using planktonic foraminifera (including Rd. cal­carata). There are also platform deposits (the Angostura Formation, more than 1000 m-thick, upper Campanian and Maastrichtian in age) with rudists and large foraminifera (Michaud & Four­cade, 1989).

Data from American sections on land of major interest have been quoted in this volume. The Gulf Coastal Plain series of the southern United States (Puckett & Mancini, 1998) contains both ammon­ites and planktonic foraminifera for correlation. The Western Interior succession outcropping in the Walsenburg section of Colorado (Kennedy et al., 1992) or in the Red Bird section of Wyoming (Hicks et al., 1995; 1999) contains ammonites

Fig. 3. Correlation between historical stratotypes, the Tercis main section, and informal auxiliary successions. The Tercis succession is locally expanded in order to show probable (moderate) condensation (black bars in column levels). The key correlation shown in the rightmost column was introduced by Kennedy et al. (1992) whose study of Polish material led them to believe that their A. hyatti zone contained the belemnite B. langei and was directly overlain by the B. lanceolata Zone.

816

(sometimes), and magnetostratigraphic informa­tion, and volcaniclastic levels which were dated during the past 25 years.

More "exotic" sections were also quoted by J. Hancock (op. cit.) such as the Tancem Quarry at Kallankurichi in Tamil Nadu, southern India; the western side of the Sakhalin Peninsula, far eastern Russia; river valleys in Hokkaido (northern Island of Japan, e.g. the Hobetsu Valley); on Seymour Island in the Antarctic Peninsula. Crame (1999) noted that the fossiliferous Maastrichtian stage is well represented in NE Antarctic Peninsula by a 1 km-thick clastic succession and that the base of the stage had been dated by means of strontium isotope stratigraphy.

Deposits lying in the present Oceans have not been discussed above but it is important to recall that the Campanian-Maastrichtian is also recorded in series present in a wide oceanic domain and that these deposits can only be correlated to Tercis using dinoflagellate cysts, calcareous nannofossils and planktonic foraminifera (we presently have no information on how to characterise the stage boundary at Tercis in terms of radiolarians). Due to this large domain of application of Cretaceous stratigraphy, the type section for the stage boundary needs to be characterised with those planktonic microfossils.

5. Palaeogeographical extension of some fossils

Another aspect of the correlation potential of a succession is the palaeogeographical extension of the stratigraphical tools recorded in the sediment. The palaeogeographical distribution of fossils has been considered in each palaeontological chapter and only a few examples are discussed below. Microfossils are generally widely distributed; the calcareous nannofossils of Tercis are probably the most widely useful for correlation since both strictly Tethyan forms (Q. trifidum and Q. gothicum of low-latitudes) and cosmopolitan forms (Aspido-lithus parens constrictus can be found both in Boreal and Tethyan deposits) are present. Dino­flagellate cysts have also shown a good potential

for correlation between high-latitude and low-latitude deposits (see above correlation from Tercis to Charente and to Limburg).

Among the planktonic foraminifera, the distribu­tion map of Radotruncana calcarata is of high interest. The information in figure 4 is taken from Masters (1977), Puckett & Mancini (1998), and additional unpublished observations from col­leagues who provided us with the information on the Campanian-Maastrichtian outcrops (§4.2.). In addition, the taxon is quoted with other globo-truncanids in tectonical studies from the Himalaya (Bassoullet et al., 1978) or from India and Pakistan (Moullade & Nairn, 1983, p. 363, who quote Kureshy, 1976). The taxon was probably present everywhere in the oceanic basins between 35° North and 40 to 50° South. Three points in the W Pacific and five points in the East-central Pacific Ocean suggest that the taxon was probably present in the whole Ocean at least North of the Equator between latitudes 0 and 30°. Similarly, the presence of the taxon on the East and West borders of the Atlantic Ocean suggests that the taxon was also present everywhere in between when Masters showed a single point to the East of Florida.

Among the macrofossils present both below and above the stage boundary, the ammonite Diplomo-ceras cylindraceum seems to be the most widely distributed (Hancock & Kennedy, 1993; Ward & Kennedy, 1993). With the exception of the Bra­zilian find (the precise location is not known to us), this taxon lived preferentially in the palaeolatitudes higher than 30°; there is an interesting com­plementarity in the distribution of this taxon and the previous one's. Figure 4 shows (stars in circle) the approximate location of the finds; they are restricted to outcrops presently on Earth because we have no information about the presently-oceanic areas. Note that the star in the East of Australia stands for a questionable find in New Zealand.

In addition, we should consider an unpublished circular written by Madeleine Neumann a few months before her death, where she expressed the opinion that Nostoceras hyatti would be the

Fig. 4. Palaeogeographical distribution of three key fossils. Maps simplified after Camoin et al. (1993, see Bilotte et al. this volume, chap. A4). Continental area: light grey; stars: approximate location of published observation; extrapolated extension: dark grey areas.

817

Radotruncana calcarata

DIplomoceras cylindraceum

Nostoceras hyatti

8. Vrielynck / C. Abrial 2000

818

appropriate key fossil for characterisation of the Campanian-Maastrichtian boundary both because the taxon was constantly quoted in connection with that stage boundary and because its distribution was rather large (figure 4, after Hancock & Kennedy, 1993; supplemented with Ward SL Orr, 1997 for Baja California; Atabekian, 1995 for Daghestan, N Caucasus; Robaszynski et al., 1998 for Tunisia). This distribution does not reach eastern countries (India, Australasia) or South America but appears wider than the distribution of Pachydiscus neubergicus which is known from Europe, Africa including Madagascar (Hancock & Kennedy, 1993) and SE India (Pondicherry, Ken­nedy & Summesberger, 1986); Jagt & Felder (1999) also quote Oman (Arabian Peninsula) and ?Sakhalin(E of Russia).

Finally, the palaeogeographic distribution of the inoceramid genus Trochoceramus is under revision (Walaszczyk personal communication July 2000 after Walaszczyk, Cobban & Harries, in prep.) and the resulting distribution is larger than previously assumed due to increasing knowledge of Maas-trichtian deposits. Trochoceramus is present in both the Boreal and Tethyan Domains in Europe (Coten-tin to Poland and Ukraine, and Northern Germany to Aquitaine), North America (Atlantic Seaboard, Gulf Coast, Western Interior from Texas to Mon­tana), Africa (Algeria to Egypt and Senegal to Angola and Madagascar), South America (Colum­bia). The Pacific area is the single one where it is still unknown.

6. Conclusion

With regard to the historical stratotype, the Tercis section can be correlated using dinocysts, nanno-fossils, ammonites, echinoids, and asteroids; they indicate that the Campanian of Coquand would be located below level 90-100 at Tercis. Dumont's Maastrichtian would correspond to levels higher than the quarried portion of the geological site. As a result, the location of the newly defined Campan­ian-Maastrichtian boundary in the middle of the pale-flint-bearing sub-unit of the Les Vignes Unit which corresponds roughly with the appearance of P. neubergicus at Tercis is consistent with the original definitions of the bracketing stages.

The magnetostratigraphically well-documented section in the Apennines can be correlated to the succession at Tercis. Between levels 40 and 160 of Tercis, the calcareous nannofossil and planktonic foraminiferal records provide us with a control of good quality and suggest correlation lines with a precision of a few metres. In this situation, the magnetostratigraphic record of the Apennines can be transferred to the Tercis section with confidence and used for correlation.

Correlation with the Boreal succession of Kronsmoor which straddles the Campanian-Maastrichtian boundary is difficult to establish partly because we could not collect recent informa­tion on the German section, partly because the same fossil groups from the two areas could not be studied by the same experts, and partly because the Kronsmoor section is poor in biostratigraphical control. Some progress could be achieved by a direct comparison of the micro-brachiopod and the benthic foraminiferal faunas of the two areas. In the absence of this information, an indirect correlation has been proposed using a poorly established magnetic control, depositional rates, and a bio­stratigraphical key observed in Poland: the contemporaneous LO of A . hyatti and FO of B. lanceolata. The resulting scheme does not contra­dict the ammonite evidence available from the two areas.

Other sections either correlated to the Boreal succession (Ireland), or to the Tethyan succession at Tercis and in the Apennines could provide us with additional information on the Campanian-Maastrichtian stratigraphy. In addition, some North American successions containing the stage bound­ary could be correlated with the Tercis one using either the planktonic foraminiferal content or the ammonites.

Thanks to the diversity of the stratigraphical tools used at Tercis it is possible to suggest correlation lines with a variety of depositional basins and, thus, the geological site can fulfil its role as a reference section.

Acknowledgments

The knowledge gathered in this chapter results of the considerable amount of information acquired

819

by the Maastrichtian Working Group members for the last six years. This chapter took advantage of the useful information provided by A. Atabekian, J.-P. Bassoullet, J. Bourgois, A. Dhondt, E. Four-cade, J. M. Hancock, M. Machalski, E. Molina, F. Robaszynski on sections previously unknown to us. I. Walaszczyk was kind enough to send a summary

of his future paper on the distribution of the inoceramid genus Trochoceramus. Careful review by F. Robaszynski significantly enriched the pre­sent chapter. Final reading and improvement of the form by M. A. Lamaurelle are deeply acknowl­edged.

(Prepared: May 2000; revised: July 2000)