Detecting the earliest signs of human settlement in the pollen record
Fire as an indicator of human presence
(from McGlone & Wilmshurst 1999b)
Sporadic traces of charcoal in a core accompanied by short-term forest disturbance cannot be used as a reliable indicator of human presence. However, major deforestation by fire, followed by substantial burning, is a reliable indicator. Dating these major events is a key to determining the timing of first Mäori environmental impact.
Mäori lacked metal implements or forage animals, but made habitual and widespread use of fire to modify and manage the landscape. Fire was used to keep tracks clear, to maintain their dwelling areas free of tall vegetation, to clear sites for cultivation, and to encourage bracken (Pteridium esculentum) which provided a major source of carbohydrate for their diet. Even tall wet evergreen forest could be burnt in a dry season, or could be prepared for fire through felling of understorey saplings and small trees to provide dry fuel.
The landscapes that the first settlers encountered were for the most part heavily forested. By 1000 years BC, between 85% and 90% of the New Zealand landscape was forested, with grassland and scrubland highly restricted below treeline. While forest was more open, low and scrubby in the dry areas of the eastern South Island, it still appears to have been continuous. Only in the semi-arid heartland of Otago and South Canterbury did a low conifer-dominated scrub, or scrub-grassland prevail. Fernland appears to have been equally restricted except following major volcanic eruptions in the central North Island when brief peaks of bracken dominance - probably lasting no more than several decades - occurred.
Significant natural fires seem to have been largely confined to the dry eastern coasts of the North and South islands, although soil charcoal can be found throughout New Zealand. In the inland South Island, limited fires may have played a role in helping to maintain an open vegetation throughout the Holocene. There is some indication that fire frequency increased in this region after about 1000 years BC. It is from this background of infrequent but significant natural fire that the impact of humans must be distinguished.
Some people have argued that smaller fires preceding deforestation in fossil pollen records (indicated by increases of bracken spores and charcoal) are highly suggestive of human activity, and that any fire episodes should be considered as possibly caused by early settlers. There is abundant evidence of short-term (<50 years) forest disturbances caused by small fires during the last 10 000 years in some pollen records from New Zealand. On this basis, are we to believe that people were in New Zealand thousands of years ago? It is a weak argument, especially when there is no supporting archaeological or palaeoecological evidence for the presence of humans in New Zealand before about AD 1280, despite years of excavations and research on this subject.
Two factors may help decide if a fire is of natural origin or not. First, natural fire was rare in the moist, densely forested regions in the west of both the North and South islands. Most soil charcoal indications of natural fire come from drier eastern regions (<1000 mm/year). Therefore, fires in moist dense forest are highly likely to have been lit by people. Second, human-lit fires usually reoccur frequently - often with rapid return periods of less than a decade. Only in exceptional sites with annually laminated sediments can estimates of fire frequency be determined, and this has yet to be done in New Zealand. However, the persistence of short-lived seral vegetation that can only be maintained by disturbance (e.g. bracken, tutu (Coriaria arborea) and some grasslands) is a good indicator of regularly and frequently repeated fire. We can be confident that the widespread destruction of rainforest by fire is highly unlikely without human involvement. If destruction of any forest type is accompanied by repeated fire and the spread and persistence of short-lived seral vegetation, human agency is virtually a certainty, particularly when backed up by alternative evidence for human presence for example, from archaeological materials.
In New Zealand, the sheer change in magnitude of fire occurrence at a regional scale is strong evidence for it being anthropogenic. Records of soil charcoal and the presence of microscopic charcoal fragments in lake and bog sediments unmistakeably indicate a massive increase in fire from the 13th century onwards. For instance, in the dry inland southeastern Canterbury district there is an order of magnitude increase in soil charcoal around the 12th century. Such an increase in fire is unparalleled in the previous 8000 years and follows the appearance and rapid spread of archaeological sites.
The beginning of human influence my be difficult to establish with any certainty at an individual site. In non-problematic cases there is an abrupt increase in charcoal particle influx, from sporadic or low levels, to continuous high levels of input. At the same time, there is a sharp decrease in tree pollen, and pollen and spores of seral scrub, bracken or grassland communities dominate. Swamp and lake sediments often record inwash of silt and sand from fire-disturbed catchments, and major changes in the swamp or lake edge flora from oligotrophic sedges and rushes to nutrient-demanding wetland plants such as raupo (Typha orientalis). Not all sites yield such a clear picture. Sometimes there is a prolonged phase prior to the massive increase of charcoal, where there are slightly elevated or sporadic charcoal levels and traces of seral vegetation. Do these earlier traces represent a coincidental increase in the natural fire background, perhaps through drought or volcanic eruption? Do they result from reworking of small amounts of pollen and charcoal in the core, or do they indicate early limited burning by humans? There is no way of knowing which interpretation is correct. Nevertheless, exactly how these traces are interpreted is crucial in estimating the date of first human environmental impact.
A further complication is that fire frequency and magnitude are not directly related to human population size. Vast areas of virtually uninhabitated back country were burnt in both the North and South islands. It seems likely that fire was an intended consequence of human presence. Vast areas of forest and other vegetation were burnt and cleared, and repeated firing prevented forest regeneration. Burning in districts remote from areas of main population concentrations was almost certainly to facilitate movement and to improve local food resources. As firing of dryland vegetation is not labour intensive, there is no way of estimating how many people were involved. During a drought, a single individual could conceivably be responsible for the destruction of hundreds of square kilometres of dryland forest. Very different sized groups would cause approximately the same amount of vegetation burning if they ranged over the same area. Only in dense moist forest, where clearances required labour-intensive preparatory work, should we expect a relationship between human density and burning intensity. Small groups ranging widely would therefore leave as obvious a signature as large groups, provided their return period was sufficiently frequent to prevent forest regeneration.
 |
 |
Top