Original Articles

A preliminary assessment of association‐analysis in the Kroonstad area


Vegetation in the Kroonstad area was sampled by means of a stratified random distribution of 235 4mx4m quadrats within physiographic‐physiognomic categories as determined and delineated on aerial photographs. “Normal” and “inverse” association‐analyses were carried out “Normal” association‐analysis where low‐frequency species were, respectively, included in and excluded from the computation, gave broadly similar results in so far as the major groupings are concerned. Local frequency and constancy estimates were computed for various groupings. Association tables were drawn up for major groupings and Poore's “successive approximation” approach was adopted in reducing species associations to noda. There is a strong subjective element in this determination of noda, and an effort should be made to reduce it to a minimum.

Get new issue alerts for Proceedings of the Annual Congresses of the Grassland Society of Southern Africa