No abstract

A survey of 20 km'^ of species-rich kwongan (sclerophyllous shrubland or sandplain vegetation) is reported, A total of 429 vascular plant species were found, of which 338 were recorded at the eighty-seven systematically located 0,1 ha releves, Eive vegetation units were deflned following analysis ofthe site-floristics data. Small but distinctive suites of species (totalling 26% of recorded species) characterized a gradient in soils from the sands and gravels ofthe lateritic uplands, through the deep sands of the dune sequence, to the clays ofthe winter-wet depressions. However the soils-related axis in the ordinations accounted for only ca. 6% of the variability in the data, indicating the complexity of the vegetationenvironment interactions. Species richness was greatest in a zone of apparent overlap between the lateritic group of species and those of the deep sand areas. The winter-wet areas were poorest in species. The distributions of the floristicallydeflned vegetation units did not closely parallel those of physiognomic units deflned by interpretation of air photos. Structural dominance was negligible in all but the winter-wet areas and the few dominants appeared poorly correlated with other species and with the physical environment. Comparisons are made with areas of similar vegetation elsewhere in Australia and in South Africa,

This paper reports on the kinds of geographic patterns encountered in the distribution of Australian species of Acacia and on some climatic correlates of these patterns. The analyses were based on distribution data of 837 species mapped on a 1° x 1.5° grid. The area of highest density of species was the south-west corner of the continent, especially adjacent to the major boundary separating the Arid Zone from the more humid South West Botanical Province. The second major centre of richness occurred in eastern Australia south of the Tropic of Capricorn along the topographically heterogeneous Great Dividing Range. Secondary centres of species-richness occurred in northern and north-eastern Australia, a number of rocky tablelands of the Arid Zone and in western Victoria. The principal species-poor areas were located in sandy and some riverine areas of the Arid Zone, in temperate forests of Tasmania and in coastal areas of the north of the continent. The geographic patterns of each section of Acacia, when combined with those of species density, highlighted the tropical (section Juliflorae) v. temperate areas (sections Phyllodineae, Pulchellae, Botrycephalae and Alatae). The numerical classification of grids resulted in the recognition of eight major Acacia areas, arranged under four Acacia regions: (1) South-west; (2) Eastern, comprising a southern and a northern area; (3) Northern, comprising an eastern and a western area; (4) Central, comprising a south-east, a central and a north-west area. A discriminant function analysis indicated that precipitation was more important than temperature in distinguishing between areas. Discussion of the potential evolutionary significance of these findings and brief comparison with other biogeographic studies are given.

Kwongan vegetation (sclerophyllous shrublands) growing on a single soil type was surveyed over a 5 000 km 2 area centered about Eneabba, Western Australia. All samples were on laterite, a deeply weathered, infertile soil developed since the Tertiary over much of the ancient, subdued landscape of south-western Australia. The vegetation was rich in angiosperm plant species and, although quite uniform in structure, highly variable floristically, suggesting the existence of a complex mosaic of species distributions. Of the four floristic groups defined using both ordination and classification techniques, the group including the Mt Lesueur area was the most distinctive. Site groupings were predominantly regional although the influence of climatic factors, particularly potential evaporation, was significant. Biogeographic patterns of each of the major plant families were not coincident. The geographic arrangement of the floristic groups does not correspond with existing vegetation maps compiled primarily on the basis of physiognomy.

No abstract