No abstract

The extensive Quaternary loess-paleosol deposits of South Island, New Zealand, represent one of the major proxy records of paleoclimatic changes in the Southern Hemisphere. We attempted to produce the first numeric chronology of these subaerial sequences in the Canterbury region by using thermoluminescence and infrared-stimulated luminescence dating methods. We examined five exposures: a 6 m thick section at Cust, north of Christchurch; two thicker (c. 14 m) sequences on Banks Peninsula (Barrys Bay and Onawe sites); farther south, a c. 12 m sequence in Timaru; and a c. 7 m sequence on the coast at the Normanby site near Timaru. Our results are largely based on single experiments per sample, and therefore provide imprecise ages for several of the older samples. The most satisfactory results are those from the youngest site (Cust), for which three samples were dated. Here, phases of maximum loess deposition are dated at 73 ± 13 ka (basal loess-paleosol unit L3), 41 ± 5 ka (basal L2), and 27 ± 3 ka (basal LI). At Barrys Bay an age of 70 ± 15 ka was obtained in the basal LI, and at Timaru two separate samples in the base of LI also yielded ages of c. 70 ka, thus correlating the entire Cust loess sequence with the LI loess unit at these two other study sites. Only at Barrys Bay were ages (c. 130-250 ka) in stratigraphic order obtained for older samples (units L2 and deeper). At the other sites, some samples in the sub-Ll units gave age reversals, and some (including the oldest sample at Barrys Bay) yielded poor precision (e.g., 20%). Units L2 at Timaru and Barrys Bay may correlate to all or part of MIS 6; however, the poor precision and some age reversals in other units at these sites and at Normanby and Onawe preclude any unambiguous correlations between sites or to the MIS time-scale. Nevertheless, in the absence of any prior numeric ages, these first results serve as a basis for more precise future dating of these units. Although these reconnaissance dating results illustrate some of the problems for luminescence dating of such sequences in South Island, they do provide a beginning for a more accurate correlation of terrestrial and terrestrial-marine sedimentary sequences in this part of the Southern Hemisphere.

723

The loess deposits in the southern part of the South Island represent at least four major periods of accumulation. The deposits include the present soil and three well-defined paleosols separating individual loess members. A fourth poorly defined paleosol within the uppermost loess member is considered to represent a minor period of soil formation.The name Stewarts Claim Formation is proposed for the loess deposits of the region. The formation comprises four members: Yellow loess member (the youngest); Brown A loess member; Brown B loess member; and Brown C loess member (the oldest).The distribution pattern and morphology of the loess deposits is described. It is considered that in most localities, paleosols do not have a complete paleo-solum. The break between a paleosol and the overlying loess of the succeeding deposit is marked by a sharp boundary, in many places associated with a discontinuous stone line. The morphology of paleosols is comparable to that of the lower part of the solum or sub-solum horizons in present day soils developed on loess, particularly the yellow-grey earths. Thus it is considered that a period of "pedosphere stripping" has decapitated the soil prior to the deposition of the succeeding loess layer.Wood fragments at the base of the second youngest loess member immediately above the second major paleosol give a radiocarbon date (NZ1457) of 21 500 ± 1100 years B.P.The loess deposits are considered to span the period from the Waimean Glacial Stage to the present day and are correlated with advances and retreats of the ice during the Pleistocene.

No abstract