Vegetations structure and substrate of the northern part of the Great Dyke,Rhodesia: Gradient analysis and dominance-diversity relationships

Summary Contrasts in physiognomy and floristic composition of the vegetation of serpentine and non-serpentine substrates are very strong in the northern part of the Rhodesian Great Dyke, and the boundaries between these types are very sharp.Principal components analysis and reciprocal averaging are used to carry out a gradient analysis of the vegetation. The variation in the vegetation of the study area is interpreted in terms of one complex gradient representing a, series from relatively favourable to unfavourable for plant growth. Toxic effects of the substrates and water conditions are the main variables in this interpretation; the latter factor has a reversed effect on serpentine as compared to the non-serpentine substrates, and is less important than the toxicity factor.Comments are made on the table arrangement suggested by reciprocal averaging.The relatively favourable substrates were generally richer in species, but highest and lowest species numbers did not occur at the extremes of the interpreted gradient. The trend in the number of families per stand is similar to that in species numbers for the non-serpentine substrates, but is lowest in the badly-drained serpentine habitat interpreted as most unfavourable to plant growth. The species/family ratio per stand does not show much variation between any of the communities.The percentage of monocotyledonous species is strikingly higher in the communities on serpentine as compared to those in non-serpentine habitats. The total number of dominant species is not markedly different in the communities, but this means that dominant species are relatively more common in the floristically poorer communities of the serpentine, especially those with relatively high total cover values. Values for eveness and Simpson's index of heterogeneity calculated per stand largely confirm this. Again the number of monocotyledonous species amongst the dominants is surprisingly high in the communities on serpentine. Several suggestions explaining the relative success of monocotyledons in serpentine habitats are discussed.Nomenclature is according to the present (1.1.1977) usage at the National Herbarium, Salisbury, Rhodesia.One of us (M.J.A. Werger) wishes to gratefully acknowledge the financial support for this study by the Netherlands Foundation for the Advancement of Tropical Research (WOTRO) and by the Faculty of Science, University of Nijmegen. We are very much indebed to Jo Louppen and Mike Dale for their help with the computations and their comments on the results.