Référence 300: Montanari A., Odin G.S. & R. Coccioni, (rédacteurs) (1997). Miocene Stratigraphy; An integrated approach. Developments in Stratigraphy Series, Elsevier , 15: 694 +XVII pp. (cf. http://earth.elsevier.com; + book + full list) (En vente chez l'éditeur) Copyright: La permission de l'éditeur est nécessaire pour utiliser électroniquement le matériel contenu dans ce travail. permission@elsevier.co.uk

Ce fichier propose: - Avant propos (en anglais) par G. S. Odin (voir) - Préface (en anglais) par A. Montanari (voir) - Liste des présentations (voir) - Texte des présentations des Parties A, C, D, E (voir) - Remerciements des rédacteurs (en anglais) (voir)

FOREWORD

This volume results from the enthusiastic involvement of Alessandro Montanari and Rodolfo Coccioni in the geology of the Italian Apennine area. This region is exceptional in the presence of pelagic sediments deposit in quiet environments together with volcanic explosive activity which is the source of frequent interbedded volcaniclastic layers. These features make the area very suitable for documenting the geological history through the application of a multi-faceted approach that is the base of modern integrated stratigraphy.
The "Miocene Project" has a great deal with stratigraphic geochronology, a subject in which I am particularly involved as the leader of the Subcommission on Geochronology of the International Commission on Stratigraphy. For this reason and for having collaborated on a similar research project in the past: the Eocene/Oligocene boundary, Alessandro and Rodolfo asked for my collaboration in this project.
In this volume, I have been careful to follow certain rules. These rules result from my involvement in the search for establishing a unified "calendar" of the historical geology, which is the main object of chronostratigraphy. This "calendar" is a succession of Stages and groups of Stages; the latter are conventional units which combine empirical observations together with subjective choices both of which need rigor if they are to be meaningful and long-lived.

The importance of historical stratotypes
In my view, the use of the Global Stratotype Section and Point (GSSP) concept supported by the International Commission on Stratigraphy for siting Stage-boundaries is not an attempt to provide different definitions of units; these units usually exist and have been defined… by their historical stratotypes. These are bodies of rocks containing stratigraphical signals of different type. The role of GSSP's is to fix more precisely the fuzzy or multiple definitions of the Stage boundaries. The precision required, and attainable is on the scale of about tens of ka in the Miocene.
Hence, chronostratigraphic units can be regarded as conventional units based on real bodies of rocks and the boundaries between them may be denoted by an error bar which corresponds to the "thickness" of the boundary. Today, it is necessary that this thickness is reduced to the resolution obtainable by the best stratigraphical tools.

How to attain precision at the boundaries ?
GSSPs are designed for achieving unique and precise boundary definitions. Essentially, as the historical stratotypes, GSSPs are points which provide a local time horizon that by extrapolation, may allow the recognition of an instant of geological history. Their position has to be selected by correlation, in reasonable consistency with the historical stratotypes. Their characterization must involve two factors in this order of priority, the easiest correlation potential from a particular section and the best possible obtainable precision. The first factor is mostly determined by biostratigraphical constraints of unequivocal character; the second by lithostratigraphical and physico-chemical constraints usually of more equivocal character. In this volume, the problem we have dealt with is to document the first point using both biostratigraphy and geochronology, in order to propose useful sections for future siting of GSSP's.
Physico-chemical data have also been considered for more precise characterization, though with less success. Additional search would still be necessary to achieve a better resolution. This resolution has ultimately to become better than the potential diachronism of the biostratigraphic markers and the current uncertainty linked to the analytical calibration of the geochronologically derived numerical ages.

Chronostratigraphy and biostratigraphy
I asked the volume contributors to pay particular attention to distinguish between the conventional and the real nature of the chronostratigraphic units, and the correlation criteria used for the recognition of the boundaries of those units. It is difficult to explain the need to avoid the confusion between any particular criterion useful for correlation and the actual, fundamentally "lithological", definition of a boundary. One solution is to avoid the temptation to make a boundary correspond to any particular criterion which then tends to become "the" magic criterion. The preference for multiple approaches at GSSP's is recommended by the International Commission on Stratigraphy. Multiple overlapping criteria also become necessary if a single magic criterion is rejected in favor of a set of bracketing signals that characterize a GSSP.
The problem mostly concerns biological signals. I have tried to convince colleagues that biological signals are not accurate time markers at the scale that we now deem to be necessary (high resolution = 20 ka). This requires that chronostratigraphic units should avoid a definition at, or approximation with, the known location of a single biosignal in a section. In terms of high resolution, the location of biological signals in the field is a characteristic of a section which may or may not be widespread to other sections.
The discussion may be extended to biozones, especially those defined by first or last occurrences. It is not uncommon to see in textbooks and even in research papers that, following Eras, Periods, Epochs and Ages (Stages) the next finer time subdivision will be the biozone. This is fundamentally incorrect: the nature of units is different, the coincidence of limits between units in the two systems of subdivision is not necessary, and I dare say, not desirable. Let us suggest a comparison. Chronostratigraphic units are like centuries in the occidental calendar; the biozones are like the reigns of sovereigns or dynasties. The former are conventional and useful for a number of geologists (chronostratigraphic units) / peoples (centuries). The later are particular to a section or a basin (biozones) / country (reigns). Centuries cannot be subdivided in reigns, and the limits of these two kinds of intervals do not coincide.

In short, avoiding confusion between biostratigraphic and chronostratigraphic subdivisions is mandatory for a precise definition and correct use of the Stage names. The best way to avoid such a confusion is to disconnect the limits of the two kinds of unit. The data reported in this volume show that the historical definitions of Stages do not coincide with later established biozonal subdivisions useful for long distance correlation.

The global character of our search
I have encouraged the enlargement of the initial number of studies (from the Northern Apennines), to strive for a global extent that chronostratigraphic subdivisions must approach. My colleagues and I were luck to meet interested geologists working in other basins, countries, and continents; but application of our approach to other sections and basins still has to be considered even though that this volume has reached its last stage toward publication.

Geochronology in stratigraphy
Proposals have been made during the last decade to circumvent the great difficulty of documentation of the geological time scale by focusing on physically measured ages. Mathematical treatment of data (use of chronograms), magneto-kilometric interpolation (hypothesis of constant sea-floor spreading rate), and astrochronologic approaches have been used to propose "reliable" age models. The experience shows that: 1- in absence of enough data, a common occurrence, mathematical treatment is not always useful to derive meaningful numbers; 2- extrapolation-interpolation procedures may lead to incorrect numbers, 3- real progress for such efforts are better based on data acquisition rather than on mathematical or modeling criteria.
Let us have a look at two examples. As early as in 1975, I emphasized (Odin, ref. 50) that an age of 37-38 Ma (Harland et al., 1964, Palmer, 1983, and later magneto-kilometrically derived products) was too old for the Eocene/Oligocene boundary: an age at 34 Ma based on geochronological data (Odin, 1982, ref. 100, Odin and Odin, 1990, ref. 226) was proposed. This was independently supported by Glass and colleagues who dated Late Eocene North American tektites (Glass et al., 1986). But the old age was maintained in some popular syntheses favoring the interpolation approach. An extremely pernicious effect resulted from this dogmatic old age. A significant number of ages including excellent K-Ar and Rb-Sr ones at about 34 Ma for Late Eocene biotite separates was published by Montanari et al. (1985). Instead of using this set of data to document the young age favored by Glass or myself, the authors initially preferred to question the stratigraphy of the dated section at Massignano and suggested an intermediate age at 35.4 Ma based on a single level (later shown to be problematic). I convinced Alessandro to reconsider the question and we undertook a new geochronological study. This led us to propose the section at Massignano for locating the GSSP of the Eocene-Oligocene boundary with a newly documented age at 33.7 ± 0.5 Ma (Odin et al., 1991, ref. 236).
The second example concerns the Tortonian/Messinian boundary. All time scales published before 1992 have considered an age of about 6.5 Ma for this boundary. Mathematical calculation, interpolation-extrapolation, and astrochronologic approaches did not question this number (poorly constrained with direct measurements). The pioneer work by Vai and colleagues (1993) has shown that this age was more than 10% too young; several contributions in this volume confirm this fact.
The indirect approaches have failed to show their potential of predictability.
It may appear that the real age of this or that boundary is not important: we need only to refer to a common scale. However, the incorrect estimate of the age of the two boundaries discussed above, particularly in the case of the Eocene/Oligocene boundary, has lead to an important problem for continental geologists. In North America, the stratigraphy of the Land Mammal Ages was documented with geochronologic data, but in Europe the continental fauna were correlated to marine Stages the age of which was taken from current and biased numerical time scales. In this situation, the so called "Grande Coupure" at the Eocene/Oligocene boundary was supposed at a different time in the two palaeogeographic domains: intercontinental correlations resulted in artificial "diachroneity" of the evolution of mammals. The problem is only resolved when the 3 to 4 Ma younger age has finally been agreed for the boundary. However, for the past 15 years, intercontinental correlation remained incorrect partly because priority for geochronological ages has not been rationally applied at early stage.
In this forezword, I wished to demonstrate that geochronology is a necessary and integral discipline of stratigraphy. Search for application of this approach must be a priority because indirect approaches are simply subjective estimates, their use being realized only after they have been tied to a numerical time scale.
Concerning the question of time scale calibration, the main problem in the correlation between the two complementary approaches (geochronological documentation and modeling for "continuous" scales) is commonly based on the misunderstanding of the actual meaning of the basic numerical age data.
The most difficult exercise in geochronology is not to obtain an age but to correctly estimate and express a realistic geological uncertainty on the calculated analytical age.
These were some points which have been debated during the putting together of this volume. We intended to prepare more than just a simple collection of contributions with some sort of unified direction, and to influence the funding of this basic knowledge of historical geology, i.e. stratigraphy.
After many months spent in this pursuit, I hope that we have achieved at least a part of the goal of presenting an "Integrated Approach to Miocene Stratigraphy". We have tried.

Paris, 20 June 1996
G.S. Odin
Chairman Subcommission on Geochronology

Acknowledgements
This text has benefited from interesting discussions with and significant formal improvement by Silvia Gardin and Bilal Haq.

Additional references:
Glass, B. P., Hall, C. M. and York, D., 1986. 40Ar/39Ar laser probe dating of North American tektite fragments from Barbados and the age of the Eocene-Oligocene boundary. Chem. Geol. (Isotope Geos. sect.), Sp. Iss.: Calibration of the Phanerozoic Time Scale, (G. S. Odin, ed.), 59, 181-186.
Odin G. S., 1975. De glauconiarum constitutione origine aetateque. Thèse Doct. d'État, Univ. P. & M. Curie, 250 p.
Odin G. S. (Edit), 1982. Numerical dating in Stratigraphy. John Wiley Publishers, Chichester, 1040 p.

 

 

PREFACE

At the end of the Miocene Columbus Project (MICOP) in Portonovo (Ancona, Italy, November 11-15, 1992), there was much satisfaction and excitement, among the 100 or so participants who attended the meeting, for the positive outcome of this international scientific workshop. The MICOP was promoted under the aegis of the Subcommission on Geochronology (S.O.G. - Gilles Serge Odin, Chairman) of the International Union of Geological Sciences (I.U.G.S.), and organized by the editors of this volume. It was prompted when new data and insight about the geology and palaeontology of the Miocene Epoch started to emerge from the studies by numerous European, American, and Japanese researchers. In fact, the preliminary results presented in 50 original abstracts [Montanari, Coccioni, and Odin (Eds.), 1992] shed some light on the many intriguing aspects of this Epoch which witnessed such extraordinary events as the upheaval of the Alpine-Himalayan orogenesis, global climatic and environmental changes, the progressive closure of the western Tethys Ocean and consequent isolation of the Mediterranean Sea from the rest of the world's oceans, and ultimately, the appearance of hominids in Africa.

In a few words, the Miocene represents the prelude to modern geologic times, a Renaissance in the history of the Earth, when the biologic and physical environments of the planet reached the familiar profiles of today's world.
The positive outcome of the MICOP encouraged the organizers to propose the publication of two special issues, one that would have collected the scientific contributions relevant to the regional stratigraphy of the Miocene, and the other, in the form of a book, mainly focused on the general stratigraphic attributions of this Epoch. Our optimism, backed by the success of the MICOP workshop, lead us to believe that in less than a couple of years we would have accomplished these tasks.
The publication of a special issue of the Giornale di Geologia entitled "Miocene Stratigraphy of Italy and adjacent Areas" (Coccioni, Montanari, and Odin, 1994) was punctually accomplished about two years after the MICOP, and succeeded in documenting various aspects of Miocene regional stratigraphy of Italy and surrounding areas.

For the second issue, we engaged in the ambitious task of introducing, in the most exemplar and exhaustive way as possible, the concept of integrated stratigraphy in general, and the Miocene Integrated Stratigraphy in particular. By integrated stratigraphy, we concordly intended the direct application of geochronology to biostratigraphically defined sedimentary sequences which could be characterized also with other time-relative tools (i.e., lithostratigraphy, magnetostratigraphy, and chemostratigraphy). The works presented at the MICOP conference showed that in Italy, Spain, and Japan there are, in fact, Miocene sedimentary sequences suitable for the application of integrated stratigraphy (i.e., marine, fossiliferous, continuous sections containing datable volcaniclastic layers).
In short, integrated stratigraphy is an interdisciplinary team effort which can lead to a more reliable and cross checked recognition of the time relations between samples collected vertically and horizontally, and provides the means for confident correlation of geological and palaeobiological events at a global scale.
Unfortunately, longer time than expected was required to accomplish the publication of this volume. As senior editor, and therefore responsible for this delay, I wish to address my personal apologies to those contributors who punctually submitted their completed works before the proposed deadline of 1994, having waited another two years to see them materialized into a publication. On the other hand, this delay has been inevitable for two main reasons. One is that many of the results presented at the MICOP conference were preliminary, and required further (time-consuming and pains-taking) analytical work by interdisciplinary and international research teams before being submitted in their definitive and acceptable forms. The second reason is that the book could not be launched as an unfinished ship, but required a non trivial editorial effort to ensure internal consistency (both formal and substantial), and cross-checked verification among the innumerable and interdisciplinary data documented by 80 authors in 37 original contributions, and corroborated by 777 cited references.

Finally, I would like to stress that this volume is not simply a black on white documentation of the work done by individual participants in an international scientific workshop, but it represents a genuine team achievement of the kind that modern science needs for it's fundamental process of dynamic progress. Nor it is an issue that is meant to be more or less absorbed and then archived in a bookshelf as a closed case file, but rather an honest starting point, and at the same time a stimulus, for further studies on the still unresolved mysteries of the fascinating Miocene Epoch.

Acknowledgements
I would like to thank my colleagues, friends, and co-editors Gilles Serge Odin and Rodolfo Coccioni for their firm commitment and faithful dedication in working as a team on the complex, costly, and sometime nerve-racking editorial work for this volume.
A special thanks goes to Walter Alvarez, and Bernard Beaudoin for having financially supported the editorial work which was mostly carried out at the Geological Observatory of Coldigioco. I also would like to thank Tanya Atwater, Kevin Stewart, Paul Myrow, July Maxon, Paul and Debbie Kopsick, the geology students of the Keck Program, and the students of the Carleton College who in the summer 1995 helped revising, often with useful critical remarks, most of the manuscripts of this volume.

Coldigioco, 19 May, 1996
Alessandro Montanari

MIOCENE INTEGRATED STRATIGRAPHY
rédacteurs : Alessandro Montanari, Gilles S. Odin, & Rodolfo Coccioni

Un sommaire en français est annexé à la fin de chaque contribution de ce volume.

 

 

LISTE DES CONTRIBUTIONS

Avant-propos par G. S. Odin	v
Préface par A. Montanari	ix
Liste des auteurs	xi

Part A - The historical stratotypes

Part B - GEOLOGY OF THE TWO MAIN STUDY AREAS

Part C - STUDIES RELEVANT TO THE LOWER MIOCENE SUBSERIES

Part D - STUDIES RELEVANT TO THE MIDDLE MIOCENE SUBSERIES

Part E - INTEGRATED STUDIES RELEVANT TO THE UPPER MIOCENE SUBSERIES

Part F. SYNTHESES

MIOCENE INTEGRATED STRATIGRAPHY
edited by Alessandro Montanari, Gilles S. Odin, and Rodolfo Coccioni
Elsevier Publishers

Introduction à la partie A: THE HISTORICAL STRATOTYPES
CHRONOSTRATIGRAPHIC UNITS: HISTORICAL STRATOTYPES AND GLOBAL STRATOTYPES
G. S. Odin

In agreement with common usage, the word Stage is used in this presentation both for the actual rocks deposited during an interval of time, and for the interval of time itself (sometimes called Age). Designation of Stages (and Ages) is the purpose of chronostratigraphy. Stages are the fundamental and smallest subdivisions used for most of the Phanerozoic (i.e. from Cambrian to Pliocene). Definition of these units is thus of primary interest for stratigraphers. The main purpose of this volume is to improve the knowledge and definition of the Stages for the Miocene Epoch.
Stages have been defined referring to a lithologic unit with particular palaeontologic content usually deposited in a shallow water shelf or platform environment usually rich in macrofossils. Therefore, a Stage is usually defined by a body of rocks with a characterizing fossil content. Where palaeontology and lithology were different at a given time over a broad area, different Stages were defined. The resulting chronostratigraphy has long been dependent on basinal or broader palaeogeographic areas with difficulties when correlations had to be drawn over greater distances.
In Europe, the widely used subdivision for the Miocene Epoch comprises three sub-Epochs (Early, Middle, and Late), each divided into two Stages; the sequence is from older to younger: Aquitanian, Burdigalian, Langhian, Serravallian, Tortonian and Messinian. These six Stages were originally defined in three areas: the oldest two in Aquitaine (western France), the next three in the Piedmont region of northwestern Italy, and the youngest in the island of Sicily (southern Italy). Later, the Burdigalian Stage of Aquitaine (the lower portion of the Stage) has been supplemented with a type section in a fourth area: the Rhône Valley (higher portion of the Stage).
These historical definitions use bodies of rock for the definition of the Stages; in modern terms the Stages are defined by the various characteristics of a body of rock which can be used for correlation purpose: petrographical, chemical, physical and biological. This means that Stages are defined by a content. The main definition of a Stage is a "type section" (sometimes supplemented with auxiliary sections) comprising rocks deposited during the main portion of time represented by the Stage itself.
In the last few decades, geologists have included consideration of a global practice when recommending chronostratigraphic conventions. In addition, a more precise definition than in previous time may now be useful -if not necessary- for these conventional units because modern approaches sometimes allow a greatly improved resolution of time. With this purpose in mind, the Commission on Stratigraphy has recommended that historical stratotype sections be supplemented by a new concept aimed at defining precise boundaries between Stages. This concept is the Global Stratotype Section and Point (later abbreviated GSSP).
Two aspects must be emphasized: in the peferred view supported here, 1) GSSP is aimed at giving additional information for supplementing historical stratotypes, and 2) it practically defines a boundary between two Stages and not particularly the base of the next Stage.
1) The Stage definition must primarily -and always- be defined in terms of actual content since the matter of earth scientists is primarily the study of actual rocks. In this situation, the instant defined by a GSSP (or the interval of time defined between two GSSP) is not self-sufficient for defining a Stage: the latter must correspond -at least in part- to a geological object. In other words, one may consider that Stage boundaries are already defined by historical stratotypes but that they must be improved with the use of the GSSP concept.
The Stage boundary itself must be defined using changes in characteristics observed within a section. The former priority given to the base of a Stage comes from an ancient concept of Stages still supported by some stratigraphers. They used the principle of historical stratotype and, in order to eliminate the real problem of imprecise Stage boundaries, they recommended an additional convention which was to select the Stage boundary at the base of the younger Stage in the historical stratotype. In this convention, everything located below the base of the historical stratotype is part of the preceding Stage. With this solution, it is clear that the object "next Stage" is defined by actual rocks with a given geologic content but the "former Stage" is only theoretical, and concretely undefined due to the (common) sedimentary break located below the first rocks of the historical stratotype. Here again, the GSSP concept emphasizes the presence of concrete material (for both the older and younger Stages) for defining the basic units of chronostratigraphy.
The aim of a study for proposal of a GSSP is not to change the definitions of Stages. GSSP must be consistent with the previous definitions of the bracketing historical stratotypes.
In other words, the aim of this study is to elaborate Stages of global use defined by a combination of historical stratotypes and GSSP. In this situation, improvement of Stage definition using GSSP necessarily requires a good knowledge of the historical stratotypes, the use of the most recent stratigraphic tools in these classic sections, as well as the sections where GSSP are proposed to be located. I have therefore encouraged experts to summarize the data on "historical stratotypes", and to try to obtain additional up to date information on these sections.
Part A of this volume presents the data available from the four historical areas quoted in the beginning of this introduction: Aquitaine, Rhône Valley, Italian Piedmont, and Sicily, as well as from some nearby areas where additional data have been recently considered for Stage definitions. The authors were requested to present their data following an organization and content similar to the one suggested for submission of GSSPs. These comprise ten points, two of which concern access and study, four on the quality of the geological record, and four other points on the availability of correlation tools (adapted from Odin, 1992).

1. Accessibility: The GSSP has to be accessible to investigators without politic or social restrictions, excessive effort or expense.
2. Permanent artificial marker: A permanent artificial marker should highlight the GSSP in the outcrop to insure its easy identification, uniqueness, and constancy. Designation of an official responsible for preservation of the site is suggested.
3. Marine character: Within the Miocene, GSSPs should be selected within a marine, pelagic depositional environment allowing correlations over large areas. Non-pelagic deposits may be more useful for some intervals of time, and the potential for correlation into continental deposits must be generally considered.
4. Continuous deposition and thickness (applicability of sequence stratigraphy): The GSSP has to be chosen within a section with continuous deposition across the boundary. The sedimentation rate should be sufficiently high (5 to 50 m per Ma) in order to register environmental variations in detail. The thickness of the section below and above the GSSP should be great enough to record the eustatic cycles preceding and following the boundary (i.e. thicknesses corresponding to approximately 1 to 2 Ma on each side of the boundary).
5. Quiet history (no tectonics, strong diagenesis or bioturbation): The section should be free of structural complication, strong diagenetic alteration, and redeposition. Tectonic activity may change the original succession of the strata whereas strong diagenetic changes may destroy or alter the magnetic, geochemical, and biological record. The record and resolution of palaeontologic, magnetic, and geochemical events have to be undisturbed, without excessive bioturbation, reworking or vertical mixing.
6. No facies change: The absence of a facies change across the boundary should reduce the presence of depositional hiati (see 4.) and insure the representativity of the variations characterised by the stratigraphic tools (see 7, 8, 9, and 10).
7. Applicability of biostratigraphy (the first order condition): A good palaeontologic record should enable biostratigraphic correlations based on several groups of fossils. For the Miocene, significant calcareous nannofossils and planktonic Foraminifera must be present.
8. Applicability of magnetostratigraphy: The applicability of magnetostratigraphy is particularly important for the last 150 Ma. For the Miocene Epoch, the complexity of the magnetic polarity record requires thick sections to allow analysis of significantly long records.
9. Applicability of chemostratigraphy: The composition of the marine water masses (trace elements, stable and unstable isotopes) changes with time and chemostratigraphic signals have an excellent correlation potential. Sr isotopes are of special interest for the Miocene Epoch, and pelagic carbonates seem to be most favourable for the application of this tool.
10. Applicability of geochronology: The possibility of direct numerical dating of the GSSP by the presence of geochronometers is important for the definition of the numerical time scale. This is particularly feasible for the Miocene Epoch.

More general conditions/recommendations may be added. The most important is that a GSSP should be consistent with previous historical stratotypes - i.e. when possible, the point should not be located at a place clearly identified as or suspected to be within the previously defined Stages. This means that, when data area collected for the establishment of an appropriate GSSP, the main question to be answered will be: where are the layers equivalent to those of the historical stratotypes in the candidate section?
A second recommendation concerns the criteria used for selecting a particular point in the section for location of the global stratotype point (GSP). A number of previous proposals for GSSP have been based upon only one (palaeontological) criterion. In some cases, the criterion of correlation is taken itself as the "criterion of definition"; for example, the last occurrence of a fossil -coinciding with the level where the GSSP has been selected- is the only criterion considered for characterizing the limit. This is a source of considerable error and arises from a lack of understanding of the concept. A GSSP must strictly remain a point the lateral extension of which separates a content below from a content above. The sum of the characteristics of the rocks below and above give the means for correlating the point. In this view, a fundamental rule for the precise location of a GSSP within the interval not defined by the historical stratotypes will be: "the best location of a GSSP is with respect to a series (a maximum number) of potential correlation criteria". This excludes, at least formally, the use of -or dependance upon- a single criterion.
In the search for and description of sections able to host GSSP, data to be gathered should address the ten recommendations described above.

SOMMAIRE : Présentation de la partie A
Unités chronostratigraphiques: stratotypes historiques et stratotypes globaux

L'élaboration d'une échelle chronostratigraphique de validité internationale nécessite une définition précise des unités fondamentales que sont les Étages. Ceux-ci peuvent être conçus, aujourd'hui, comme la combinaison
1- de leur définition classique par le stratotype historique qui détermine l'essentiel de leur contenu concret et
2- de leurs limites précisées par le concept de point stratotype global (les PSG). Dans cette vue, une bonne connaissance des stratotypes historiques est une condition sine qua non pour le progrès des conventions. En effet, l'amélioration de la définition des Étages ne passe pas par un changement mais par un complément de cette définition par l'ajout de points stratotypes compatibles avec les conventions déjà en usage.
La contribution à l'élaboration de ces PSG -pour les temps miocènes- est l'objectif de ce travail; il importait donc de débuter cet ouvrage par un rappel et, lorsque cela fut possible, des résultats nouveaux sur les stratotypes historiques. Ces exposés sont suivis par les propositions de PSG en cours d'élaboration pour les limites inférieure et supérieure de la Série. Ces synthèses, conçues pour aider à la présentation de coupes susceptibles d' accueillir un PSG sont l'occasion de rappeler ou de présenter les caractères qui permettront d'utiliser les PSG pour des corrélations aussi larges que possibles.

 

Introduction à la partie C : STUDIES RELEVANT TO THE LOWER MIOCENE SUBSERIES
A. Montanari and R. Coccioni

In the western Tethys domain, the beginning of the Miocene is characterized by widespread explosive volcanism accompanying the uplifting of the Apennine and Alpine orogens. Volcanism was actually active in Sardinia and in the southeastern Venetian Alps (Monti Lessini, Monti Berici, Colli Euganei) since the late Eocene, but it was mainly characterized by mafic magmatism represented by thick andesite sequences in western Sardinia, and basaltic to trachytic sequences in the Venetian area.
Numerous biotite-rich volcanosedimentary layers interbedded with Upper Eocene and Oligocene pelagic marls throughout the northeastern Apennines represent the distal air-fall products of this Mediterranean mafic volcanism, although their provenance (i.e. Sardinia vs. Alpine or others volcanic sources) remains unknown. Nevertheless, these volcanosedimentary layers provided excellent material for a direct radioisotopic age calibration of the Late Eocene and Oligocene with unprecedented accuracy and precision.
During the Miocene, the western Mediterranean volcanism evolves to a progressively more sialic, explosive magmatism. In western Sardinian, Palaeogene andesites underlay pyroclastic deposits including dacite and rhyolite tuffs which prelude the Early Miocene rifting of the Corso-Sardinian microplate from the southwestern European margin, and its counterclock-wise rotation with consequent opening of the Balear Sea. Once again, the pelagic and hemipelagic basins of the Apennines functioned as relatively tranquil depositional sites for the distal wind-blown tephra produced by the western Mediterranean syntectonic volcanism.

Plagioclase-rich volcanosedimentary layers were found and dated by Odin et al. (Chapter C1) in the Carrosio section (northwestern Apennines), and yielded 40Ar/39Ar ages around 22.4 Ma, therefore consistent with a lowermost Miocene age inferred from the planktonic foraminiferal record.
Several lower Miocene, marine, fossiliferous sections in the northeastern Apennines of the Emilian region contain volcanosedimentary layers with plagioclase and biotite suitable for radioisotopic dating (Odin et al., Chapter C2), and yielded preliminary 40Ar/39Ar ages ranging between 22 and 19 Ma. Thus, these sections bear a strong potential for the calibration of the Early Miocene chronostratigraphic scale although they require further studies for improving the precision of both radioisotopic ages and biostratigraphic determinations.

In the northeastern Apennines of the Umbria and Marche regions, the lower Miocene is represented by the Bisciaro formation, and the lower part of the overlying Schlier formation. The general mafic to sialic evolutive trend of the Miocene Mediterranean volcanism is reflected by the mineralogy of the volcanosedimentary layers interbedded with these pelagic units. In the Bisciaro formation, sanidine is totally absent and the felsic fraction is represented mostly by andesine (plagioclase: An36-52), quartz, and traces of anorthoclase. Amphiboles and heavy minerals are rare, and biotite is present only in a very few ash layers. Rare sanidine first occurs in a biotite-rich layer which marks the boundary between the Bisciaro and the Schlier formations (mid Burdigalian). Sanidine is than found as an accessory mineral in practically all the volcanosedimentary layers in the upper Burdigalian to Tortonian Schlier formation.
The Bisciaro in the Contessa Valley, near Gubbio, is exposed in continuity with the underlying palaeogene and Cretaceous pelagic carbonate sequence which has been the object of various stratigraphic and geochronologic studies. A first integrated stratigraphic study of the Bisciaro section at Contessa was carried out by Montanari et al. (1991) who produced a detailed biostratigraphy based on planktonic Foraminifera and calcareous nannofossils, a magnetostratigraphy, a strontium isotopic profile throughout the Oligocene and lower Miocene, and 40Ar/39Ar single crystal laser fusion dates on plagioclase from two volcaniclastic layers bracketing the the Aquitanian/Burdigalian boundary.

In Chapter C3 of this volume, Montanari et al. present new data from Bisciaro exposures at Contessa improving the magnetostratigraphy, biostratigraphy (planktonic Foraminifera, calcareous nannofossils, and dinoflagellates), chemostratigraphy (Sr, O, and C isotopes, and trace elements), and geochronology of the upper Chattian to mid Burdigalian interval previously studied.
New 40Ar/39Ar step heating analyses on plagioclase originally dated with the single crystal laser fusion technique permit a tightly interpolated age estimate of 20.5 Ma for the first occurrence of Globigerinoides trilobus which constitutes a useful biomarker for the identification of the Aquitanian/Burdigalian boundary. However, the lower part of the Bisciaro and the uppermost part of the underlying Scaglia Cinerea in the Contessa section are strongly condensed and may contain one or more hiatuses compromising the accuracy for the age estimate of biostratigraphic boundaries. For the same reason, the recognition of the Oligocene/Miocene boundary in this incomplete interval remains uncertain. Nevertheless, new palaeomagnetic data, which permit the identification of Chron 6C, and the refinement of the planktonic foraminiferal biostratigraphy, restrict the location of this within a geochronogically calibrated interval ranging from 25 Ma to 23.5 Ma.

Combined biostratigraphic and magnetostratigraphic analyses of a new stretch of Bisciaro section exposed along the Contessa highway permit the identification of the N6/N7 planktonic foraminiferal zonal boundary in the lower part of Chron 5Cr. This allows the correlation of the upper part of the Contessa section with the lower part of the Moria section described by Deino et al. in Chapter D1 of this book, and provides stratigraphic continuity and completeness for the lower Miocene sequence in the Umbria-Marche region.
The most continuous and complete section of Bisciaro known at present in the Umbria-Marche Apennines is located at Santa Croce di Arcevia, about 35 km northwest of Gubbio. This section bears a great potential for high-resolution integrated stratigraphy of the lower Miocene. The only drawback is that it is located in a private backyard and the access for sampling requires permission from the land lord.

Nevertheless, Coccioni et al. in Chapter C4 produce a detailed lithostratigraphy and biostratigraphy based on planktonic Foraminifera, calcareous nannofossils, and dinoflagellate cysts. These results help to implement the condensed and incomplete Contessa section, and permits correlation with the upper Burdigalian portion of the Moria section described in Chapter D4.
In the Ebro basin of northern Spain, the Early Miocene is mostly represented by continental deposits. In Chapter C5, Odin and co-workers describe the floral and mammal biostratigraphy of this vast basin reporting the presence of a conspicuous volcaniclastic layer which can be recognized in several distant sections. Fresh sanidine contained in this volcaniclastic marker yielded a 40Ar/39Ar plateau age of 19.4 Ma which constitutes a strong geochronologic tight point for correlating lower Miocene (Ramblian) continental biostratigraphy with Tethyan marine biostratigraphy which is well calibrated in the Italian sections.

 

Introduction à la partie D : STUDIES RELEVANT TO THE MIDDLE MIOCENE SUBSERIES
A. Montanari and R. Coccioni

During the Middle Miocene (Langhian and Serravallian), the Alpine-Himalayan orogenesis reached its paroxysmic phase while the Tethyan Ocean approached its definitive closure. This major tectonic event deeply affected the world palaeoclimate and ecology with important repercussions on regional biological rearrangements in both marine and terrestrial environments. On the other hand, this widespread tectonism caused interruption of pelagic sedimentation in most of the epeiric and oceanic basins throughout the Tethyan domain, and the onset of copious synorogenic detrital sedimentation (i.e., flysch and molasse) in tectonically active foredeep basins. For these reasons, continuous and complete pelagic sequences covering the Middle Miocene, and potentially suitable for detailed integrated stratigraphic studies, are rare throughout the Tethyan domain.
Rare cases of open sea, essentially hemipelagic mid Miocene basins are found in the external (eastern) foothills of the northern Apennines which were involved in the northeasterly-migrating orogenic deformation only in the Pliocene. A well-exposed, continuous section located near the village of Moria, in the northwestern Marche region, contains all the essential qualities for high-resolution integrated stratigraphic studies throughout the upper Burdigalian and lower Langhian (Deino et al., Chapter D1). Three volcanic ashes have been dated in this section yielding very precise 40Ar/39Ar plateau ages of 17.1 Ma (biotite), 16.2 Ma (sanidine), and 15.5 Ma (sanidine), respectively. The upper two dates bracket the Burdigalian/Langhian boundary identified on the basis of the first occurrence (FO) of Praeorbulina glomerosa sicana, here recognized within magnetic Chron 5Br. The tightly interpolated radioisotopic age for this event is 15.9 Ma. The planktonic foraminiferal biostratigraphy is also integrated with a detailed calcareous nannofossil biozonation. Trace elements chemostratigraphic analyses exhibit significant excursions of Mg, Fe, Sr and Mn just prior to, or at the Burdigalian/Langhian boundary, coinciding also with a considerable negative shift of the d18O. Strontium isotope analyses on fish teeth indicate a monotonic increase of the 87Sr/86Sr values throughout the Moria section. However, the isotopic ratios in this section are significantly lower than ratios measured in coeval oceanic carbonates outside the Mediterranean domain. This may indicate that during the Middle Miocene, the proto-Mediterranean had already a limited water exchange with the rest of the world's oceans. If confirmed in other sections, this differentiation will command caution in using geochemical criteria for correlating mid Miocene Tethyan sequences with oceanic ones.
In the northeastern Apennines, Langhian sections potentially suitable for detailed integrated stratigraphic analyses are found near Apiro (about 50 km SE of Moria; see Montanari et al. in Chapter D2), and in the Cònero Riviera near Ancona. The latter exhibits a continuous hemipelagic sequence from the upper Langhian to the lower Messinian, and is described by Montanari et al. in Chapter E1. A strong potential for integrated stratigraphy is found in central Sicily where a Serravallian section is well calibrated with planktonic foraminiferal biostratigraphy, and contains a biotite-rich volcaniclastic layer which yielded a 40Ar/39Ar age of 13.5 Ma (Odin et al., Chapter D3).
Widespread volcanism in the Betic Cordilleras produced abundant volcaniclastic material which is preserved in Langhian to upper Tortonian marine sequences throughout southeastern Spain (Montenat et al., Chapter D4). This Spanish situation certainly bears a strong potential for high-resolution integrated stratigraphy of the Miocene Epoch and encourages further detailed interdisciplinary studies.
A particularly promising situation for detailed integrated stratigraphic studies outside the Tethyan domain is found in Japan. Serravallian sections containing well preserved calcareous (Foraminifera and nannofossils), and siliceous (radiolaria and diatoms) planktonic microfossils are located in central Japan, and promote detailed studies for biostratigraphic intercorrelation among these fossils groups (Takahashi and Oda, Chapter D5). Moreover, the mid Miocene Japanese sections contains volcanic tuffs which yielded preliminary fission track ages from zircon ranging from 13 to 12 Ma, and a biotite K/Ar age of 11.6 Ma, with analytical uncertainties as large as 1.5 Ma. More precise K-Ar and 40Ar/39Ar ages were obtained by Takahashi and Saito (Chapter D6) from two key tuff beds which bracket the N13/N14 foraminiferal zonal boundary marked by the FO of Globigerina nepenthes. Additional 40Ar/39Ar dates from biotite and sanidine in the upper tuff key bed were obtained by Odin et al. (Chapter D7) allowing a precise age estimate of 11.76 ± 0.1 Ma for the FO of G. nepenthes. An equally precise 40Ar/39Ar plateau age of 11.48 Ma from volcaniclastic biotite in the Cònero Riviera section lead to an age estimate of 11.3 Ma for the same foraminiferal event. Thus, the integrated stratigraphic study by Odin and co-workers reveals a slight diachronism for the FO of G. nepenthes between Japan and Italy. Nevertheless, on the basis of these precise geochronologic data, an age between 11.0 and 11.5 Ma can be estimated for the Serravallian/Tortonian boundary which shortly postdates the FO of G. nepenthes.

 

Introduction à la partie E : STUDIES RELEVANT TO THE UPPER MIOCENE SUBSERIES
A. Montanari and R. Coccioni

During the Late Miocene (Tortonian and Messinian), the Tethyan Ocean was definitively closed as a result of syn-orogenic collisional tectonism, and its Mesozoic and Cenozoic sedimentary sequences were deformed and uplifted along the emerging Alpine-Himalayan orogenic system. Here and there in the proto-Mediterranean basin, deep water, open sea environments persisted through the Late Miocene, and were involved in the orogenic deformation only later on, in the Pliocene. However, the dominant Late Miocene sedimentary deposits throughout the Alpine-Himalayan domain are represented by thick flysch and molasse sequences which quickly filled tectonic foredip troughs immediately before being deformed and obducted on the emerging orogens.
Tectonic upheaval accompanied by violent and widespread volcanism affected also areas outside the Alpine-Himalayan domain, as well as marine basins which were still open to the rest of the world's oceans. Finally, the Messinian witnessed two other extraordinary events which are currently at the focus of heated debates and fascinating scientific research: the temporary desiccation of the Mediterranean basin, and the appearance of the first hominids in Africa.
From this flash overview of the Late Miocene scenario, it appears clear that this exceptionally active time of geologic history offers subjects for scientific research in the most disparate fields of earth sciences which will be absorbed by generations of geo-scientists for many years to come. However, if in one hand the Late Miocene is generous of rocks, phenomena, and tectonic structures, on the other hand it is parsimonious of those marine sedimentary deposits which allow an integrated stratigraphic approach for a much needed improvement of the geologic time scale. Moreover, the actual chronostratigraphy of this sub-Epoch is currently based on Mediterranean stratigraphic sections which not necessarily bear fossil records reliable for time-correlations among distant sequences throughout the world.
The first seven chapters in Part D of this book report the results of integrated stratigraphic studies in several sections located along the eastern side of the northern Apennine foreland palaeobasin, namely the Apennine foothills near Faenza (Emilia region), and the coastal cliffs of the Cònero Riviera near Ancona (Marche region). The latter constitute a rare, perhaps unique case in which the entire Miocene Epoch is represented by a nearly complete and continuous sequence of pelagic and hemipelagic sediments.
In Chapter E1, Montanari and co-workers document the results of their interdisciplinary stratigraphic studies on the Cònero Riviera sequence covering the time interval from the mid Langhian to the lower Messinian. These include detailed bio-event determinations based on planktonic Foraminifera and calcareous nannofossils which allow precise correlations with well established biostratigraphic models. Unfortunately, a detailed palaeomagnetic study of a rich sample collection from the Serravallian and lower Tortonian portion of this sequence was not inducive to magnetostratigraphic determinations due to the magnetic weakness and/or altered conditions of these marly sediments.
The interval comprising the upper Serravallian and the lowermost Tortonian in the Cònero sequence was analyzed in detail for geochemical determinations of trace elements. Significant shifts of the total iron, strontium, and magnesium values found near the Serravallian/Tortonian boundary may constitute a viable tool for the recognition of this interstage boundary among western Tethyan sections. Stable isotopes were also measured in this interval but they show fairly uniform trends. On the other hand, both oxygen and carbon isotope curves exhibit significant trend shifts just above the Tortonian/Messinian boundary, which may very likely be related to the initiation of the Messinian salinity crisis.
87Sr/86Sr ratios in biogenic carbonate and fish teeth throughout the Serravallian, Tortonian, and lower Messinian interval of the Cònero Riviera sequence show considerable scattering and are overall deviant from isotopic ratios measured in coeval oceanic sequences. A quick negative trend inversion of the 87Sr/86Sr values is recorded in the upper Tortonian and it is followed by an equally rapid trend inversion to higher values in the lower Messinian. Again, these geochemical signatures are indicative of the particular palaeoenvironmental conditions which developed in the Apennine palaeobasin at a time when the western Tethyan was no longer exchanging deep marine waters with the rest of the world's oceans (see also Chapter B1). An organic geochemistry study was also carried out on black shales from the Cònero sequence indicating, in fact, that the organic matter preserved in Serravallian shales is predominantly of open marine origin, whereas the Messinian black shales carry a clear terrestrial organic signature. In summary, geochemical trends shifts in the Cònero Riviera sequence may be useful tools for correlating sections within the Tethyan domain but also indicate that this palaeobasin was developing its own geochemical equilibria already in the Serravallian and had restricted water mass exchange with the open ocean at least since then. This would prevent the use of chemostratigraphic tools for correlating mid and upper Miocene Tethyan sections with oceanic ones.
Unaltered biotite from two volcaniclastic layers located in the lower, and uppermost Serravallian respectively, were dated with the 40Ar/39Ar incremental heating technique yielding very precise ages. From these, Montanari and co-workers derived a tightly interpolated age of 11.0 ± 0.2 Ma (2s) for the Serravallian/Tortonian boundary.
The next five chapters of this book (Chapters E2 to E6) report detailed integrated stratigraphic data from several sections in the Faenza area of the Romagna Apennines. These sections contain the Tortonian/Messinian boundary, and exhibit a sedimentary facies characterized by open marine marls interbedded with black shales, practically devoid of turbiditic and/or siliciclastic deposits. Moreover, these sections contain several biotite-rich volcaniclastic layers which permit precise and accurate geochronologic dating of the Tortonian/Messinian boundary. In Chapter E2, Odin and co-workers present new litho- and biostratigraphic data from the Pieve di Gesso section which can be easily correlated with more complete and continuous sections exposed in the nearby Monte del Casino and Monte Tondo areas (see Chapters E3 to E5). A number of new radioisotopic dates from biotite and plagioclase obtained by Odin and co-workers are consistent with a suite of dates from volcaniclastic layers presented in following chapters, and confirm an age slightly older than 7 Ma for the Tortonian/Messinian boundary. Using the rich geochronologic documentation of these northern Apennine sections, Vai in Chapter E3 examines the cyclostratigraphy of the Messinian finding a good consistency between geochronologic data and cyclostratigraphic sequence. It appears that orbital precession cycles (about 20 ka) are dominant throughout this Stage, whereas obliquity cycles (about 40 ka) are prevalent in the upper Messinian. The cyclostratigraphic approach appears, in this case, a viable means for calibrating Messinian stratigraphic and geologic events with extreme precision and accuracy, and leads to an estimate of about 2 Ma for the duration of this Stage. Details about nannofossil biostratigraphic and palaeomagnetic analyses of the Monte del Casino and Monte Tondo sections are presented by Negri and Vigliotti in Chapter E4, whereas a complete account on the integrated stratigraphy of these sections are given by Laurenzi and co-workers in Chapter E5, and Odin and co-workers in Chapter E6. From these studies, and age of 7.2 Ma is derived fro the Tortonian/Messinian boundary.
In Chapter E7, Odin and co-workers present new data about the integrated stratigraphy of the Maccarone section, located in the eastern Marche region of Italy. This is a rare, perhaps unique situation in which the interval across the Messinian/Zanclean (i.e., Miocene/Pliocene) boundary is represented by a continuous sequence post-evaporitic, fossiliferous, marine pelitic sediments which contain, in the upper part of the Messinian, a prominent volcanic tuff layer. Two 40Ar/39Ar incremental heating dates from pristine biotite from this tuff yield an age of 5.5 Ma. Thus, Odin and co-workers estimate an age younger than 5.5 Ma for the Miocene/Pliocene boundary, in good agreement with previous estimates slightly older than 5 Ma.
An interesting stratigraphic situation, very promising for integrated stratigraphic studies, is represented by the sedimentary basin of southeastern Spain, and described by Montenat and Serrano in Chapter E8. In this region, the Tortonian is genuinely marine while the Messinian records the evolution of the basin from open marine, to evaporitic, and finally to terrestrial environments. Both marine and terrestrial sediments are fossiliferous and contain volcanic rocks such as rhyodacites and lamproites which may be suitable for radioisotopic dating.
Detailed calcareous nannofossil data from several mid to upper Miocene sections in the Miura Peninsula of Central Japan are presented by Saito and co-workers in Chapter E9. These sections contain several key tuff beds which yielded reliable mid Tortonian and Messinian K-Ar ages. The stratigraphy, mineralogy and petrography of these tuffs are described in detail by Saito and co-workers, and constitute a great potential for a numerical calibration of the Late Miocene calcareous nannofossil biostratigraphy in an oceanic setting. This permit to test the consistency of correlation, on the basis of robust geochronologic dates, among stratigrafic sections representing open ocean environments, and isolated Tethyan environments described in the preceding chapters of this book. Similarly, a great potential for integrated stratigraphic study of the entire Miocene Epoch is found in the fossiliferous, volcanic tuff-bearing sequence of the Boso Peninsula of Central Japan. A geologic description and discussion of preliminary biostratigraphic data from this sequence are presented by Takahashi and co-workers in Chapter E10.

 

 

ACKNOWLEDGMENTS

We would like to express our most sincere gratitude to the experts who presented the state of the knowledge of the historical stratotype sections in this volume. Following their knowledgeable elaborations made it possible to progress toward a better undestanding of the foundation of the stratigraphy of the Miocene Epoch. Our synthesis would not have been possible without the help of field geologists who worked in areas other than Italy. Their contribution has been critical for a broader view of the available stratigraphic record in the best known Miocene sections throughout the world, and this will be necessary for the proper selection and evaluation of candidate GSSPs. This matter remains to be considered more thoroughly in the near future, and we would like to encourage additional stratigraphic work in countries around the Mediterranean basin as well as in other distant areas around the world.
Although this work has been first undertaken under the aegis of the Subcommission of Geochronology, the biostratigraphic input has been essential, and the experts who contributed in this are acknowledged; we can only recognize that biostratigraphy is still a fundamental discipline in earth sciences and, despite a recent tendency in underestimating and minimizing the painstaking work of paleontologists and biostratigraphers, we want to stress that it constitutes the major basis of Phanerozoic stratigraphy, and hope that it will be properly supported in the future.
With the recent development and application of elaborated analytical techniques, experts in organic matter, trace elements, and isotope geochemistry, have brought to this work a significant amount of information investing considerable time, technical, and financial resources. To them goes our sincere gratitude.
Our main purpose has been to combine and, in some cases, coordinate the information provided by the many contributors to this volume, and come up with a synthesis for a comprehensive understanding of the Miocene time sequence recorded in sedimentary rocks. We may have partially reached this purpose but we also recognize that we may have come short in emphasizing the high value of each contribution, properly exploiting their conclusions in this synthesis.

Our strongest hope is that this work as a whole will be considered as a generous group effort, and understood not just for its specific contents and particular information, but as a suggestive encouragment for further developing the approach of Integrated Stratigraphy in other portions of the Earth's geologic record.

on April, 1996
G.S. Odin, A. Montanari, R. Coccioni