Orue-Etxebarria, X.

The Global Stratotype Section and Point (GSSP) for the base of the Bartonian (middle Eocene) stage is as yet undefined. Herein we assess the GSSP potential of the successions found in the Basque–Cantabrian and Aquitanian basins (western Pyrenees). On the basis of the available data, no outcrop in the Biarritz and Pamplona areas fulfilled the requirements outlined by the International Commission on Stratigraphy. However, the succession exposed on the eastern side of the Cape of Oyambre (San Vicente de la Barquera, province of Cantabria, northern Spain) did so, and yielded positive preliminary results.

The Global Stratotype Section and Point (GSSP) for the base of the Bartonian (middle Eocene) stage is as yet undefined. Here, we assess the GSSP potential of the La Acebosa Formation exposed on the eastern side of the Cape of Oyambre (San Vicente de la Barquera, province of Cantabria, northern Spain). Sedimentological and benthic foraminiferal data indicate that the La Acebosa Formation represents a bathyal environment related to a deepening, eastwardfacing slope. Preliminary planktic foraminiferal and calcareous nannofossil results show that the upper part of the section corresponds to zones E11 and CP14a, respectively, and thus could include the Lutetian–Bartonian boundary. Although the quality of demagnetization is quite poor, preliminary magnetostratigraphic data suggest that Chron C19n, which is most likely to define the Bartonian GSSP, could be recorded in the upper part of the La Acebosa Formation. A denser magnetostratigraphic and biostratigraphic sampling is required in the upper part of this unit in order to accurately identify Chron C19n and to establish the sequence of events across the Lutetian–Bartonian transition.

The global stratotype sections and points for the bases of the Selandian (Middle Paleocene) and Thanetian (Upper Paleocene) stages have been defined in the coastal cliff along the Itzurun Beach at the town of Zumaia in the Basque Country, northern Spain. In the hemipelagic section exposed at Zumaia the base of the Selandian Stage has been placed at the base of the Itzurun Formation, ca. 49 m above the Cretaceous/ Paleogene boundary. At the base of the Selandian, marls replace the succession of Danian red limestone and limestone-marl couplets. The best marine, global correlation criterion for the basal Selandian is the second radiation of the important calcareous nannofossil group, the fasciculiths. Species such as Fasciculithus ulii, F. billii, F. janii, F. involutus, F.pileatus and F. tympaniformis have their first appearance in the interval from a few decimetres below up to 1.1 m above the base of the Selandian. The marker species for nannofossil Zone NP5, F. tympaniformis, first occurs 1.1 m above the base. Excellent cyclostratigraphy and magnetostratigraphy in the section creates further correlation potential, with the base of the Selandian occuring 30 precession cycles (630 kyr) above the top of magnetochron C27n. Profound changes in sedimentology related to a major sea-level fall characterize the Danian-Selandian transition in sections along the margins of the North Atlantic. The base of the Thanetian Stage is placed in the same section ca. 78 m above the Cretaceous/Paleogene boundary. It is defined at a level 2.8 m or eight precession cycles above the base of the core of the distinct clay-rich interval associated with the Mid-Paleocene Biotic Event, and it corresponds to the base of magnetochron C26n in the section. The base of the Thanetian is not associated with any significant change in marine micro-fauna or flora. The calcareous nannofossil Zone NP6, marked by the first occurrence of Heliolithus kleinpelli starts ca. 6.5 m below the base of the Thanetian. The definitions of the global stratotype points for the bases of the Selandian and Thanetian stages are in good agreements with the definitions in the historical stratotype sections in Denmark and England, respectively.

An integrated bio-, magneto- and cyclostratigraphic study of the Ypresian/Lutetian (Early/Middle Eocene) transition along the Pyrenean Otsakar section (Payros et al., 2011) resulted in the identification of the C22n/C21r chron boundary and of the calcareous nannofossil CP12a/b zonal boundary; the latter is the main correlation criterion of the Lutetian Global Stratotype Section and Point (GSSP) recently defined at Gorrondatxe (Basque Country). By counting precession-related mudstone-marl couplets of 21 ka, the time lapse between both events was calculated to be of 819 ka. This suggests that the age of the CP12a/b boundary, and hence that of the Early/Middle Eocene boundary, is 47.76 Ma, 250 ka younger than previously thought. This age agrees with, and is supported by, estimates from Gorrondatxe based on the time lapse between the Lutetian GSSP and the C21r/C21n boundary. The duration of Chron C21r is estimated at 1.326 Ma. Given that the base of the Eocene is dated at 55.8 Ma, the duration of the Early Eocene is of 8 Ma, 0.8 Ma longer than in current time scales. The Otsakar results further show that the bases of planktic foraminiferal zones E8 and P10 are younger than the CP12a/b boundary. The first occurrence of Turborotalia frontosa, being approximately 550 ka older that the CP12a/b boundary, is the planktic foraminiferal event that lies closest to the Early/Middle Eocene boundary. The larger foraminiferal SBZ12/13 boundary is located close to the CP12a/b boundary and correlates with Chron C21r, not with the C22n/C21r boundary.

The GSSP for the base of the Lutetian Stage (early/ middle Eocene boundary) is defined at 167.85 metres in the Gorrondatxe sea-cliff section (NW of Bilbao city, Basque Country, northern Spain; 43º22'46.47" N, 3º 00' 51.61" W). This dark marly level coincides with the lowest occurrence of the calcareous nannofossil Blackites inflatus (CP12a/b boundary), is in the middle of polarity Chron C21r, and has been interpreted as the maximumflooding surface of a depositional sequence that may be global in extent. The GSSP age is approximately 800 kyr (39 precession cycles) younger than the beginning of polarity Chron C21r, or ~47.8 Ma in the GTS04 time scale. The proposal was approved by the International Subcommission on Paleogene Stratigraphy in February 2010, approved by the International Commission of Stratigraphy in January 2011, and ratified by the International Union of Geological Sciences in April 2011.

An integrated bio-, magneto- and cyclostratigraphic study of the Ypresian/Lutetian (Early/Middle Eocene) transition along the Otsakar section resulted in the identification of the C22n/C21r chron boundary and of the calcareous nannofossil CP12a/b zonal boundary; the latter is the main correlation criterion of the Lutetian Global Stratotype Section and Point (GSSP) recently defined at Gorrondatxe (Basque Country). By counting precession-related mudstone–marl couplets of 21 ka, the time lapse between both events was calculated to be 819 ka. This suggests that the age of the CP12a/b boundary, and hence that of the Early/Middle Eocene boundary, is 47.76 Ma, 250 ka younger than previously thought. This age agrees with, and is supported by, estimates from Gorrondatxe based on the time lapse between the Lutetian GSSP and the C21r/C21n boundary. The duration of Chron C21r is estimated at 1.326 Ma. Given that the base of the Eocene is dated at 55.8 Ma, the duration of the Early Eocene is 8 Ma, 0.8 Ma longer than in current time scales. The Otsakar results further show that the bases of planktonic foraminiferal zones E8 and P10 are younger than the CP12a/b boundary. The first occurrence of Turborotalia frontosa, being approximately 550 ka older that the CP12a/b boundary, is the planktonic foraminiferal event that lies closest to the Early/Middle Eocene boundary. The larger foraminiferal SBZ12/13 boundary is located close to the CP12a/b boundary and correlates with Chron C21r, not with the C22n/C21r boundary.

The Early/Middle Eocene (Ypresian/Lutetian) transition is represented by a hiatus in many North European sections, including those in which the classic stratotypes were originally defined. However, the Global Stratotype Section and Point of the Lutetian Stage, which is still pending definition, should be placed at a globally correlatable event included within that unrepresented interval. The Pyrenean Eocene outcrops display sedimentary successions that offer the rare opportunity to analyse the Ypresian/Lutetian boundary interval in almost continuous sections and in very different settings. Seven reference stratigraphic sections were selected on the basis of their quality and correlated by means of biomagnetostratigraphic data. This correlation framework casts light on the sequence of chronostratigraphic events that characterize the Ypresian/Lutetian boundary interval, which may prove useful in defining the main correlation criterion of the base of the Lutetian. All of the Pyrenean sections show a similar sedimentary evolution, despite being up to 350 km apart from each other, containing deposits of different origins (intrabasinal carbonate sediments, siliciclastic sediments sourced from the Iberian plate, and terrigenous sediments sourced from the uplifting Pyrenees) and despite having been accumulated in different sedimentary environments (from continental to deep marine) and in different geodynamic settings (piggy-back basin, foreland basin and cratonic margin). This common evolution can be readily interpreted in terms of a sea-level driven depositional sequence whose lowstand and transgressive systems tracts are included within the Ypresian/Lutetian boundary interval. The Pyrenean Ypresian/Lutetian depositional sequence can reasonably be correlated with depositional sequences from classic North European areas, shedding light on the palaeoenvironmental history which in those areas has not been recorded. Furthermore, these depositional sequences may possibly correlate with others from the Antarctic Ocean and from New Jersey, as well as with oceanic temperature variations, suggesting that they might be the result of climatically-driven glacioeustatic sea-level changes. Should this hypothesis prove correct, it would confirm previous suggestions that the onset of Antarctic glaciations needs to be backshifted to the late Ypresian at least.

It has long been known that a major larger foraminifera turnover (LFT) occurred at the boundary between the Thanetian and Ilerdian stages, but its possible correlation with the Paleocene-Eocene thermal maximum (PETM) was unsuspected until the work of Baceta (1996), and has been controversial ever since. After summarizing the history of this controversy, we present information from three new sections that conclusively resolve the issue, all of them placed less than 2 km to the east of the classical Campo section in the southern Pyrenees. In these three sections, an up to 7 meter-thick intercalation of continental deposits rich in pedogenic carbonate nodules is sandwiched between uppermost Thanetian and lowermost Ilerdian shallow marine carbonates. The δ13C composition of 42 pedogenic nodules collected from two of these sections (San Martín and La Cinglera) ranges between –11.4 and -14.3‰ and averages –12.9‰, values that conclusively represent the PETM and for the first time are recorded in sections where the LFT is clearly represented. Further, a high-resolution lithological correlation between Campo and the three new sections across the P-E interval unquestionably demonstrates that the lowermost marine beds with autochthonous specimens of Alveolina vredenburgi (a tell-tale of the LFT) are laterally interfingered –and are therefore coeval- with the nodule-bearing PETM continental deposits. On the basis of the new evidence, the temporal coincidence of the PETM and the LFT can no longer be doubted.

The Gorrondatxe section, a prospective Lutetian Global Stratotype Section and Point (GSSP), has recently been used as the master reference section to reassess the correlation between Eocene magnetostratigraphic and calcareous planktonic biostratigraphic scales. However, the exact calibration of some events remained ill defined, as they were thought to be missing in Gorrondatxe due to a fault. The most important missing events were the first occurrence of the planktonic foraminifera Turborotalia frontosa and the C22n/C21r chron boundary. Either might be a reliable correlation criterion for the Ypresian/Lutetian boundary, as both approach the age of the original Lutetian Stratotype. New studies allowed the identification of the former event 9 m above the Gorrondatxe fault, within magnetic polarity Chron C21r and calcareous nannofossil Zone CP12a. Distinctive test features that characterize the most primitive morphotype of T. frontosa are described. Despite the high turbidite content, recurrent pelagic limestone–marl couplets and bundles occur, whose formation was driven by precession and eccentricity astronomical cycles. The first occurrence of T. frontosa was found 27 couplets and 5.5 bundles (60 m) below the first occurrence of the calcareous nannofossil Blackites inflatus, which is dated at 48 Ma. Hence, the age of the first occurrence of T. frontosa is estimated at 48.55 Ma, confirming that it is the most suitable planktonic foraminiferal correlation criterion for the Ypresian/Lutetian boundary. These results show that the stratigraphic interval missing due to the Gorrondatxe fault cannot be greater than a few metres and reinforce the value of this section as a prospective Lutetian GSSP.

An abrupt environmental disruption occurred in the photic zone and at the seafl oor during the mid-Paleocene biotic event (MPBE). Calcareous nannoplankton, planktic foraminifer, and benthic foraminifer assemblages at Zumaia section (western Pyrenees) underwent a rapid and remarkable transformation. The major calcareous plankton assemblage changes suggest a shift from relatively cooler mesotrophic to warmer, more oligotrophic conditions, indicating a disturbed environment due to the warming of the ocean. Benthic foraminifer assemblages were also signifi cantly affected by the MPBE; diversity of the assemblages and buliminids show net decline and the low food and opportunistic taxa increase in abundance. The reorganization of the planktic ecosystem possibly involved changes in the food fl ux (type and quantity) to the seafl oor, thus triggering changes in the benthic communities. A 1‰ negative δ13C shift and a 30% carbonate content decrease are recorded in connection with the biotic event. This suggests that during the MPBE, as in the Paleocene- Eocene Thermal Maximum (PETM), an input of a large mass of isotopically depleted carbon into the ocean and atmosphere could have lowered the deep-sea pH, triggering a rapid shoaling of the lysocline and contributing to greenhouse warming. The MPBE was short lived: according to the counting of limestone-marl couplets, the stratigraphic expression of precession cycles throughout the Zumaia section, the MPBE lasted for ~52–53 k.y., with the core of the event representing ~10–11 k.y. The Zumaia section is the fi rst land-based locality in which the MPBE is recognized and described in detail. Due to its expanded character and excellent paleontological record, this section may prove to be a global reference section for the study of this shortlived event.