Quaternary vegetation and climate changes on Banks Peninsula, South Island, New Zealand

Late Quaternary terrestrial and marine pollen records from the Canterbury Plains and Banks Peninsula suggest that climates during the peak of Marine Isotope Stage 7 (MIS 7) were similar to those prevailing during stage 5e and the Holocene. Podocarp forest (notably Prumnopitys taxifolia—matai) characterises each interglaciation. In contrast, marine records from DSDP 594 cores, off the east coast of Canterbury, indicate that stage 7 was dominated by montane forest (Libocedrus sp. and Phyllocladus). This suggests temperatures as much as 3 °C colder than indicated by the Banks Peninsula assemblage. Age control from both sites appears to be robust. Some of the differences may be related to the taphonomy of the pollen at both sites. DSDP 594 may reflect a more southerly catchment of fluvially and aeolian-derived pollen than does the Banks Peninsula site.Banks Peninsula was alternately separated from, and joined to, the mainland as Quaternary sea levels fell and rose. Assuming modern ocean current patterns, during interglacials the south–north Southland Current would have swept through the seaway separating the island from the mainland, diverting the flow of rivers embouching on the Canterbury coast, and moving sediments and fluvially transported pollen northwards. Little of this material would have reached DSDP 594, nor, if wind patterns were similar to those of today, would wind-borne pollen from Banks Peninsula have reached the site. It is probable that vegetation on the Peninsula was consistently distinct from that recorded at DSDP 594, which has a more southerly derivation.In contrast to the high mountain areas of the South Island, the low levels of grass pollen in the available record suggest that the Peninsula retained a woody vegetation over much of its area during glacial periods. This was favoured by the physiography of the area, with a variety of micro-climates, and by the extensive areas available for colonisation at times of low sea level. The podocarp forest of MIS 7 was replaced by an open shrubby vegetation in which Leptospermum and Kunzea (Leptospermum-type pollen) was locally dominant, and in which Plagianthus, Phyllocladus, Coprosma and Myrsine were prominent. Charcoal is associated with this change. Most of the recorded taxa, with the exception of Phyllocladus, are present on the Peninsula today. A gap in the pollen record coincides with the Last Interglacial and Last Glaciation, but a return of forest vegetation is documented in the later Holocene.The reconstructions do not exclude the possibility of a cooler stage 7. They do highlight the importance of excluding local/regional non-climatic effects before interpreting climate change from data sets, and reinforce the necessity of testing marine records against compatible terrestrial ones.