Spatial scale and sampling strategy in paleoecological studies of vegetation patterns in mountainous terrain

Consideration of spatial scale is fundamental to the successful design of paleoecological studies of previous vegetation patterns. For pollen data to record a particular pattern, the paleoecological sampling sites should be spaced appropriately to allow for: (1) the size of the pollen source areas of the individual sites, i.e. the area from which most of the pollen reaching a basin originates; and (2) the spatial scale of the vegetation patterns of interest, e.g. regional vegetation zones and landscape-level mosaics within the zones. These considerations are particularly important in mountainous landscapes where vegetation gradients are often steep. Sampling sites must be closely spaced and consequently pollen source areas of individual sites may include several vegetation zones. Analytical procedures are then required to highlight vegetation patterns recorded by the pollen data. In this paper, we review the advantages and difficulties of paleoecological studies in mountainous regions, discuss the spatial-scale considerations required when interpreting the data from such studies, and illustrate these issues using our recent studies in the mountains of eastern North America.