Gradient analysis of forest vegetation in the Maungataniwha Range, western Northland, New Zealand

Abstract Phytosociological data from 51 forest study plots in the Maungataniwha Range, New Zealand, were analysed by numerical techniques including ordination, classification, and direct gradient analysis to identify vegetation/ environmental patterns. Nine community types segregated by specific ranges of environmental conditions were identified. Interpretations of ordination axes suggest temperature and moisture are prime factors affecting community composition. Gradient analysis is seen as an important step in evaluating the representativeness of existing and potential nature reserves. Previously, the techniques have been shown to be useful in an area where vegetation/environmental patterns are obvious. The present study demonstrates the utility where patterns are not obvious.