We used the topographical data sets GTOPO30 at 30 arc-sec resolution (about 1km at the equator) from the US Geographical Survey as the basis for our altitudinal contours. All the individual tiles, less the ones from Antarctica, were merged together using the tools available in the ArcView extension 'Spatial Tools' (Hooge 1999). The global data set was then resampled to a lower resolution of 10km, by using the standard ArcView resampling tools. This allowed an easier handling of the data. A simple reclassification of this data set allows us to highlight the desired altitude range, which is then vectorized by creating polygons by hand in ArcView. To ensure that this lower resolution did not significantly affect the borders of the vegetation zones, we made several tests using the same manipulation with the original resolution. These tests showed that a resolution of 10km gives similar results, which was acceptable, especially when recalling the very subjective nature of these maps, and the uncertainties of border locations.
Altitudinal vegetation zones were substantially lower at the LGM than at present, mainly due to the cooling relative to the present, which is evident in all parts of world. These differences will be described in detail for each region, but in general it seems that altitudinal zones were approximately 500-1000m lower than at present. The cooling, and the lowering of altitudinal zones, was most dramatic in the mid- and high-latitudes, where most areas beyond 50 degrees north would probably have had no vegetation above 500m. In the tropical latitudes, the shift in vegetation zones was less drastic, but still involved a change of the order of hundreds of metres. In effect, the lower montane zone of tropical mountains spread out into the tropical lowlands, as indicated by the abundance of montane trees in lowland vegetation at that time (e.g. Hooghiemstra 1989; Behling 1998).
The following altitudinal zones were used everywhere in the tropics between 22 degrees north and south of the Equator, on the basis of various reviews of LGM vegetation in the tropics (e.g. Behling 1998): In areas where tropical forest of some sort is reconstructed from palaeoevidence as being present in the lowlands, the upper limit of tropical lowland forest is designated as 500m above present sea level. Montane forest if present is in the interval 500-2000m. Tundra is in the interval from 2000m to 3500m. Alpine desert, perennial snow or ice is present above 3500m. All these vegetation zones are lowered some 500m (Hope 1987; Hooghiemstra 1989; Behling 1998) from the present-day vegetation zonation.
a) For the zones 22 degrees-30 degrees north or south of the Equator: upper forest limit (if forest is present in lowlands) is at 1200m. Regardless of lowland vegetation, tundra is present in the interval from 1200m to 2000m. Montane/polar desert is present beyond 2000m, unless specific evidence suggests the presence of an ice cap (Hooghiemstra 1989; Hope 1987).
b) We placed the LGM upper forest line in the latitudinal band 30 degrees north to 45 degrees north at 800m. We put 'tundra' vegetation above 800m (e.g. Yu et al. 2000).
c) All areas above 500m altitude and north and south of 50 degrees were labelled as alpine/polar desert, unless regional evidence suggests an ice sheet. This is based on the general assumption that, in the high latitudes, greater cooling superimposed upon the background cooling due to ice age conditions would eliminate montane vegetation. However, the use of the 500m upper vegetation limit is preliminary, covering a very broad latitudinal band in which temperature certainly varied considerably, and may need to be revised on the basis of reconstructed summer temperatures on a more regionally specific basis. Since no forested areas are recorded from the LGM beyond 50 degrees north and south, no upper forest limit is ascribed. In specific regions (e.g. in the western USA) where a more complex altitudinal mosaic was clearly present, refinements to the scheme were added. These are described in the regional sections.
© Internet Archaeology
URL: http://intarch.ac.uk/journal/issue11/2/3.3.html
Last updated: Mon Dec 3 2001