The history of the human landscapes of New Guinea
Kosipe is a large swamp 100 km north of Port Moresby which provides a record of more than 40,000 years from 2,000 m altitude. Although subalpine plants were more common during the glacial times the area was always forested. Records of human occupation are preserved in the cool climate organic soils of the slope above the swamp, from ca 28,000 years BP (White et al. 1970). Fire is evident from about 30,000 BP in the swamp sediments while hearths have been identified in the slope mantle. It seems likely that people occupied or visited this high altitude site to collect the large heads of Pandanus nuts. They left behind massive stone blades, possibly adzes. The usage to which these were put is still not known.
The manipulation of plants and landscapes leading to agriculture Period II 20,000–5000 years ago
Evidence for the extension of human landscapes is apparent on sensitive ecological boundaries such as the savannah-rainforest in the lowlands and the alpine treeline during the period of climate transition about 12–10,000 years ago. In these locations the encroachment on areas by forest has been resisted by fire and perhaps active clearance. The highlands have always supported forest and here the timing of clearance is quite variable, although the large basins so far looked at seem to have substantial clearances by 7000 years ago (Hope & Golson 1995). Similarly the long history of occupation along trade routes far from modern settlements demonstrates that the linkages have an extensive past. The highlanders are separated today by wet lower montane forests from the coastal resources. Although this ecological zone has not been well investigated, available data suggest that clearance is relatively recent (Gorecki & Gillieson 1989). It may be that this zone has contributed to the isolation of the highland peoples, despite the obvious passage of people and trade goods through it.
The highland valleys
In the montane zone there are relatively few pollen records that are continuous through this time period (Figure 2). Reasons for this may be the cooler and possibly drier climate that prevailed during the late glacial transition which altered conditions for deposition and preservation of organic material in sedimentary basins. Alternatively, increased burning and manipulation of forested environments by people may have caused erosion or deflation of sediments resulting in breaks in sedimentary records. However, it remains true that the separation of human activity from climate change as driving forces behind the sediment records we study is problematic in the absence of independent archaeological or palaeoclimatological data. This problem is exacerbated in the last glacial period when human activity in the landscape may have been strongly influenced by climate change rather than outpacing or overriding the climate signal (Haberle & Chepstow-Lusty 2000).
During the last glacial maximum at around 18,000 BP the highland valleys between 1500 and 1700 m asl were subject to a much cooler climate, perhaps as much as 7°C cooler than present mean annual temperatures, and frequent frosts and droughts (Haberle 1998). At Haeapugua in the Tari Basin (1630 m asl), where Nothofagus forests and open grasslands formed a mosaic vegetation pattern, the appearance of cold-adapted herbs such as Astelia (only found above 2700 m asl today) reflect the influence of cold mean annual temperatures on the basin floor. At Kuk Swamp in the Wahgi Valley (1580 m asl; Powell