Drastic climatic shift at ∼2.8 Ma as recorded in eolian deposits of China and its implications for redefining the Pliocene-Pleistocene boundary

Abstract

There has been a long-running debate over the position of the Pliocene-Pleistocene boundary, i.e. 1.8 Ma versus 2.6 Ma. This study presents litho- and pedo-stratigraphic characteristics, grain size, magnetic susceptibility, and dust sedimentation rates for ‘red clay’-loess sequences over northern China, in order to assess the drastic climatic event(s) recorded in the quasi-continuous eolian deposits, and to readdress the position of the Pliocene-Pleistocene boundary. In complete Chinese loess sequences, 33 soil-loess couplets have been identified and labeled with S i-L i system. Traditionally, the base of Chinese loess has been designated at the bottom of loess unit L33. However, a typical loess layer within the uppermost ‘red clay’ is identified and formally designated here as L34 and the soil unit between L33 and L34 as S33, which lowers the basal age of Chinese loess from 2.6 Ma to ∼2.8 Ma. Although the Neogene ‘red clay’ and the overlying Quaternary loess are both of windblown origin, they have quite different pedogenic characteristics. The loess consists of alternating reddish soil and yellowish loess layers, while the ‘red clay’ is composed of soils with much redder colour and carbonate nodule layers. In general, the “red clay” has been subjected to stronger pedogenic processes than the overlying loess, and can be regarded as an extremely thick soil complex. The alternation of loess and soils reflects large-scale oscillations between glacial and interglacial conditions, whereas the well-developed ‘red clay’ deposition indicates a relatively stable climate condition. At ∼2.8 Ma, shortly before the Gauss/Matuyama boundary, the eolian sequences show a sedimentary transition from the well-weathered ‘red clay’ to the alternating loess and paleosol deposits, accompanied by coarser particle size, increased dust accumulation rates, and change of magnetic susceptibility signature. These indicate a large climatic shift from long-lasting warm-humid conditions to large-amplitude cold-dry and warm-humid fluctuations around 2.8 Ma, and a dramatic increase in aridity over the dust source region. In contrast, such a drastic event is not identified around the top of the Olduvai event or at any other time of the past 7.7 Ma. As the Gauss/Matuyama boundary offers a sound foundation for global time correlation and chronostratigraphic classification, these results still support the proposal of ICS and INQUA, that the conventional Pliocene-Pleistocene boundary should be lowered from 1.8 Ma to 2.6 Ma.