At Marion Island a project was initiated by Steve Holness, Western Cape University, South Africa and Jan Boelhouwers, Uppsala University, Sweden, continuing an earlier 5-year project in the Maritime Sub-Antarctic.

The 290-km large Marion Island (46o 54’S, 37o 45’E) rises 1230 m asl, in the southern Indian Ocean, north of the Antarctic Polar Convergence. The climate is dominated by strong westerly winds, high relative humidity, a small temperature range, relative low temperature, and high precipitation. The study area is the peak of a shield volcano. Marion Island comprises typical sub-Antarctic biota with a scarcity of species. Coastal areas are characterised by bogs and mires in tussock grassland. The objectives of the project are to assess responses of geomorphic processes to climate changes, to analyse and describe Holocene sedimentological records, to explore and generate proxy-climate data and to integrate this with data and modelling from other researchers in palaeolimnology and climate modeling, to examine possible Northern and Southern Hemisphere teleconnections/ interhemispheric relationships between climate change, and to do detailed morphological and sedimentological analysis of key landforms (e.g. patterned ground, slope failures, blockfi elds) in order to assess palaeoenvironmental implications of these landforms. Investigation techniques include: organic sedimentary records from peat bogs, testate amoebae activity, peat humifi cation, ground microclimates, surface sediment activity, morphology and relative ages of relict glacial and periglacial features (including evidence of rapidly degrading permafrost for high altitude areas of the island) and weathering processes and rates. A Ph.D. thesis and a M.Sc. thesis and at least six peer-reviewed publications from recent periglacial research on Marion Island have either been published or are in press.

A key issue in the southern African periglacial research is the nature of Quaternary palaeoenvironments. Recent publications continue to suggest evidence for localised glaciation, while the counter arguments suggest a relatively inactive geomorphological environment during the Last Glacial Maximum, with somewhat drier conditions than present with deep, seasonal ground freezing. It is clear that landforms used to argue for niche glaciation during the Quaternary have been misinterpreted and that considerable research is required to clarify and resolve the discrepancies. Contemporary climatic data are scarce and extrapolating Quaternary palaeoenvironments is, thus, exceptionally diffi cult. Current research by Paul Sumner and Werner Nel, Pretoria, South Africa, using automated logging equipment is aimed at gathering data regarding contemporary climatic and ground microclimate in the High Drakensberg/Lesotho area. Further investigations are being conducted on openwork block accumulations, sorted patterned ground and colluvial mantles to establish both current and past environmental conditions.

A session of the International Geographical Union (IGU) Commission on Climate Change and Periglacial Environments was held at the IGU Regional Conference in Durban from 4-7 August 2002. Paper topics ranged from periglacial climates during the Last Glacial in Europe to climate change issues in the maritime Sub-Antarctic and a review of current debates on Quaternary palaeoenvironments of the High Drakensberg in southern Africa. It was clear from discussion that quantitative approaches using modern analytical techniques and modelling are needed to resolve the periglacial and glacial questions for southern Africa.

Ian Meiklejohn (