Pergamon Quaremary Inremational, Vol. 40, pp. 43-52,1997. 0 1997 INQLJA/ Elsevier Science Ltd PII: 5104@-6182(%)00%&2 All rightsreserved. Printed in Great Britain. 1040-6182/97 $32.00 LONG-PERIODIC VARIATIONS IN THE EARTH’S OBLIQUITY AND THEIR RELATION TO THIRD-ORDER EUSTATIC CYCLES AND LATE NEOGENE GLACIATIONS L.J. Lourens and F.J. Hilgen Department of Geology, lnstitute of Earth Sciences, Budapestlaan 4, 3584 CD Utrecht, The Netherlands Evolutive spectra of four high-resolutionclimatic proxy records reveal that the spectral power in the obliquity frequency band increases during five well-defined intervalsin the last 5.3 million years. These intervalscormspond to periods of high-amplitudevariationsin the tilt (obliquity) of the Earth’s axis, connected with a 1.2 myr cycle, and to sea-level lowstands of third-order eustatic cycles in the Haq et al. (1987) curve. This implies that the development of majorPlio-Pleistocene glaciations proceededepisodically and that third-order cycles are glacio-eustatically controlled.In (at least) two of these intervals sinistrally-coiled neogloboquadrinids enteredthe Medikrmnean at times of glacial periods. One intervalstartsaround2.8 Ma and is characterized by the fmt occm’rence of N. atlantica (s). Anotherintervalbegins at the Plio-Pleistocene boundaryaround 1.8 Ma and is characterized by the common occurrence of sinistrally-coiled N. puchydenna. Also during the late Miocene thixd-order cycles correlatewell with the 1.2 myr cycle of obliquity. A significant drop in the third-order eustatic curve starting at the Tortonian-Messinian boundary,for instance, coincides with a period of high-amplitudevariationsin the obliquity time series connected with the 1.2 myr cycle. Q 1997 INQLJN Elsevier Science Ltd. All rights reserved. INTRODUCTION What caused the onset of major northern hemisphere glaciations around 2.8 million years ago? This question has gained considerable scientific interest during the last decade, because the history of the onset is now being documented in unprecedented detail in continuous deep marine sequences. But despite the high-resolution records and the application of increasingly sophisticated climate models, this question has not yet been answered satisfactorily. Some researchers focus on the continuously changing positions of the Earth’s continents and oceans and argue that the opening and closing of isthmuses could create critical ‘gateways’ for altering oceanic circulation and hence climate. The final closure of the Panama Isthmus occurred shortly before the glacial era began (Keigwin, 1978; Marshall, 1988; Keller et al., 1989) and may have led to an increase of North Atlantic Deep Water production, due to enhanced flux of relatively warm and high saline surface waters to the high latitude eastern Atlantic. The oceanic heat loss from NADW formation was released to the atmosphere over the Norwegian- Greenland Sea resulting in mild winter temperatures and large volumes of precipitation required for the develop- ment of continental ice caps. Recent investigations of benthic foraminiferal 613C from several ODP sites, however, provide no evidence for a progressively increasing ventilation of NADW as would be expected, but rather point to a decrease in NADW and/or NAIW production during this transition (Tiedemann, 1991; Raymo et al., 1992). Alternatively, the onset of major northern hemisphere glaciations has been explained by a long-term decline in the concentration of atmospheric carbon dioxide. This would reduce the amount of heat trapped in the atmosphere and, as a consequence, lead to ‘greenhouse’ cooling (Chamberlin, 1899; Vincent and Berger, 1985). Following this line of reasoning, an increase in produc- tivity between 3.2 and 2.4 Ma resulted in a decrease of atmospheric CO;?, thus, creating a planetary balance in favour of glaciations (Samthein and Fenner, 1988). On even longer time scales CO;! is removed from the atmosphere by burial of organic carbon and by weath- ering of silicate rocks to form carbonates. Acceleration in seafloor spreading rates and associated mountain uplift may have caused the increase in chemical weathering of silicate-bearing (volcanic and basement) rocks needed to reduce the atmospheric COZ concentration and, hence, global cooling (Raymo et al., 1988). In addition, uplift of the Himalayan-Tibetan plateau and southwestern North America may have created favourable atmospheric conditions for the development of ice caps over eastern Canada, Scandinavia and Siberia (Ruddiman and Kutz- bath, 1989). On the other hand, subduction releases CO* into the atmosphere through volcanic activity. A decel- aration in seafloor spreading would, therefore, reduce the volcanic CO* release and lead to favourable conditions for glaciations as well. Application of the astronomical calibrated polarity time scale (APTS) (Shackleton et al., 1990, 1995a; Hilgen, 1991a, b), however, revealed no significant changes in seafloor spreading rates in four out of five plate pairs during the critical interval between 4 and 2 million years ago (Wilson, 1993). Moreover, it has been questioned (Molnar and England, 1990) whether the apparant accelerations in late Cenozoic uplifts are real or whether the geological and palynological evidence is merely a product of the climate change itself. Raymo (1991) cites the sharp rise in *‘Sr/8%r during the Neogene 43