A workshop conference entitled “Late Cenozoic Magnetostratigraphy: Comparisons with Bio-, Climato-, and Lithozones” took place in Tokyo and Otsu, Japan, between October 28 and November 1, 1974. It was organized by G. J. Kukla and H. Nakagawa as an outcome of the PA-70-17 Project of the International Geological Correlation Programme, launched in 1970 by the International Union of Geological Sciences. The workshop was supported by the Japanese Society for the Promotion of Science, by IGCP, and the National Science Foundation (of the United States). Out of 70 participating geophysicists, geologists, and paleontologists, 37 attended in person, while the remaining 33 contributed by mail. Prior to the conference a questionnaire was distributed in order to collect opinions on several issues. The workshop's objective was to tackle problems in the recognition of depositional polarity and to review the relationship of magneto-stratigraphic units with radiometric data and with regional bio-, litho-, and climatostratigraphic systems of the Pliocene and Pleistocene. The principal conclusions of the conference were summarized as follows:

(1) The only practical (though not infallible) way of demonstrating the validity of interpreted depositional polarity and magnetostratigraphic zonation is to reproduce the results in parallel, widely separated sections with different lithology and sedimentation rates.

(2) The hazard of unrecognized postdepositional normal overprints must be recognized and considered in all attempted correlations. Multiple sampling of parallel sections and consistency checks of magnetic data are a prerequisite for correct results.

(3) Continuous sequences with high sedimentation rates, such as those which originated in subsiding basins, are viewed as prospective candidates for international stratotypes of the Pliocene and Pleistocene. The investigation of such deposits should be accelerated.

(4) Magnetostratigraphy must follow basic stratigraphic principles. Magnetic zones should be clearly defined and locally labeled, whether or not their correlation with paleomagnetic chronology appears possible.

(5) Improved data on absolute age of reversals are needed. For that purpose, multiple K/Ar and fission-track analyses of key volcanic layers should be performed on a continuing basis.

Marine Pliocene and Pleistocene beds of the Boso Peninsula are more than 3 km thick and contain abundant fauna and flora. Almost continuous deposition occurred at a high sedimentation rate from the early part of the last reversed geomagnetic polarity epoch to the middle part of the last normal polarity epoch. Several microbiostratigraphic marker horizons enable correlation with deep-sea sediments fully independent of magnetostratigraphic zonation. Globorotalia truncatulinoides first occurs below the normal polarity event BO-B-2, which is correlated with the Olduvai Event in deep-sea sediments, whereas the disappearance of discoasters and the first occurrence of Gephyrocapsa caribbeanica are recognized at the top of BO-B-2, and the first occurrence of Gephyrocapsa oceanica is recognized at the top of BO-B-1. Based on the correlation of the Boso section with Mediterranean stage stratotype sections, the transition from the Pliocene to Pleistocene is estimated to be in the lower part of the Kazusa Group.

The Uonuma group, located in the Niigata oil field, represents the upper part of a thick and mostly continuous series of late Cenozoic sediments, in which only the Uonuma is of nonmarine origin. Two main magnetozones and four subzones are recognized and correlated with the Matuyama Reversed and Brunhes Normal polarity epochs. Pollen spectra indicate a climatic deterioration between the Jaramillo Event and the Brunhes/Matuyama boundary, possibly commencing during the Jaramillo Event.

We sampled 150 sites in fine-grained Plio-Pleistocene sediments of the Palm Springs and Imperial Formations. Sampling was confined to 3000 m of stratigraphically continuous section containing abundant vertebrate fossil remains of the Vallecito Creek, Arroyo Seco, and Layer Cake local faunas of the Irvingtonian and Blancan Land Mammal Ages. The magnetic stability of these sediments was sufficient to delineate the magnetic stratigraphy, which ranges from below the Cochiti event at the base, to the Matuyama reversed magnetic epoch at the top of the section. Eight faunal events are placed relative to the magnetic polarity sequence. They are cf. Pliohippus extinction, Geomys appearance, cf. Equus appearance, Tremarctos appearance, Hypolagus extinction, cf. Odocoileus appearance, Smilodon and ? Euceratherium appearance. The latter two faunal events characterize Irvingtonian Land Mammal Age. The transition from the Blancan to Irvingtonian Land Mammal Age occurs in the lower Matuyama magnetic epoch close to the Pliocene-Pleistocene boundary. The appearance of European migrants during the lower Matuyama magnetic epoch indicates that the Bering land bridge was exposed for animal migration between Europe and North America at this time.

Eight diatom and two silicoflagellate datum levels from the Pliocene and Pleistocene of the equatorial Pacific are ranked and related to the paleomagnetic reversal record, calcium carbonate fluctuations, and the oxygen isotope record. The datum levels and their position with respect to the paleomagnetic reversal record are as follows: (1) last occurrence of Thalassiosira convexa midway between the top of the Gauss and the base of the Olduvai; (2) transition from Rhizosolenia praebergonii to Rhizosolenia praebergonii var. A just below the base of the Olduvai; (3) first appearance of Pseudoeunotia doliolus in the lower part of the Olduvai; (4) last occurrence of Rhizosolenia praebergonii just above the Olduvai; (5) first appearance of Asteromphalus hiltonianus midway between the top of the Olduvai and the base of the Jaramillo; (6) first appearance of Mesocena elliptica just below the base of the Jaramillo; (7) last occurrence of Mesocena elliptica midway between the top of the Olduvai and the base of the Brunhes; (8) upward increase in abundance of Thalassiosira oestrupii just below the base of the Olduvai; (9) last occurrence of Nitzschia reinholdii just above the base of the Olduvai; (10) peak occurrence of Roperia tesselata var. ovata just above Datum Level 9. By second-order correlation it is possible to relate these datum levels to the oxygen isotope record. Such a correlation suggests that some datum levels may be climate related.

The magnetostratigraphic subdivision of the Plio-Pleistocene series in the Kinki district around Osaka in southwest Japan, based on NRM measurements of volcanic ash beds, is correlated with biostratigraphic and paleoclimatic marker beds as well as past sea level oscillations. At least 10 high sea level stands occurred in the study area between about 1.5 and 0.27 million years ago, all associated with relative warming. A significant change in the composition of flora, from Metasequoia to Fagus dominated assemblages, was observed around the Jaramillo event.

At least 17 times during the past 1.7 million years, the deposition of loess containing characteristic cold-resistant gastropods was interrupted by the development of temperate interglacial forests. This conclusion was reached in a study of paleomagnetically dated fossiliferous loess sequences in Krems, Austria and Brno, Czechoslovakia. Sequences of windblown loess interlayered with hillwash loams and steppe and forest soils exposed in brickyards around Brno and Praha, Czechoslovakia, revealed eight major depositional cycles within the Brunhes paleomagnetic epoch. We now report nine additional cycles of late and middle Matuyama age bringing the total number of glacial-interglacial cycles to 17, which occurred after the end of the Olduvai. The cycles are separated by marklines, levels of abrupt environmental change correlative with the terminations in deep-sea sediments. They are the boundaries between the windblown loess containing cold-resistant snail assemblages and between the clayey originally decalcified soils, accompanied by warmth loving Helix and Banatica snail faunas of hardwood forests. Because the presence of temperate forests in northwestern and central Europe is instrumental in the definition of an interglacial, each markline represents a glacial-interglacial boundary and each cycle is a glacial-interglacial cycle.

Paleomagnetic measurements have been carried out on 103 specimens taken at about 15 cm intervals in a sea cliff exposing the marine terrace formation. Secondary components were removed by partial demagnetization in a peak field of 200 Oe. Two short reversed magnetozones are recognized. Geological and biostratigraphical evidence indicates that they are equivalent to the Blake event, which was first reported by Smith and Foster (1969) (Science 163 , 565–567) from the North Atlantic deep-sea cores. The present result suggests a split nature for the Blake Event.

At the San Joaquin plant site near Bakersfield, California, the Brunhes/Matuyama reversal was identified 4 m below an ash probably correlative with the Bishop Tuff. Deposition of coarse detritus in a high-energy environment preceded the reversal, whereas a low-energy lake and back-swamp type sedimentation persisted through most of the interval, separating the reversal and the ash. Sediments below the reversal are strongly remagnetized in a normal field. Identification of the original polarity is possible only with some of the specimens, so that accurate positioning of the reversal plane required dense and multiple sampling. Similar precaution at all sites where reversals are used for time-stratigraphic correlations is recommended. Paleomagnetic investigation was particularly useful in proving the presence of unfaulted strata at least 0.5 my old, as required by safety regulations of the U.S. Nuclear Regulatory Commission (NRC). This is due to the fact that any thick layer of reversely magnetized sediment is with high probability older than 0.7 my.

Postglacial and lacustrine sediments from Lake Michigan have been paleomagnetically investigated. The primary goal was to test the feasibility of correlation between several cores in terms of their magnetic signatures. After alternating field and thermal demagnetizations, inclination and relative declination values were obtained that show characteristic short-term fluctuations. Major features, including changes in the magnetic intensity of the samples, correlate well in three of the longer cores, in good agreement with independent lithostratigraphic determinations. In addition, we have been able to detect local disconformities that appear as interruptions of the magnetic record. A continuous record of normal polarity has been found, extending from the present day to 13,000 14C vears BP. No reversals were recorded, though at two stratigraphic levels shallow inclinations, together with declination swings and low intensities of magnetization, may indicate rapid geomagnetic excursions. The magnetic record was attained at the time of deposition or shortly thereafter, and can be used for a wide range of applications.

The Gothenburg Magnetic Excursion in a broad sense ranges from 13,750 to 12,350 years BP and ends with the Gothenburg Magnetic Flip at 12,400−12,350 years BP (= the Fjärås Stadial in southern Scandinavia) with an equatorial VGP position in the central Pacific. The Gothenburg Magnetic Flip is recorded in five closely dated and mutually correlated cores in Sweden. In all five cores, the inclination is completely reversed in the layer representing the Fjärås Stadial dated at 12,400−12,350 years BP. The cores were taken 160 km apart and represent both marine and lacustrine environments. The Gothenburg Magnetic Flip represents the shortest excursion and the most rapid polar change known at present. It is also hitherto the far best-dated paleomagnetic event. The Gothenburg Magnetic Excursion and Flip are proposed as a standard magnetostatigraphic unit.