Australian Journal of Ecology (1982) 7, 415^19 Short Note Simpson Desert dunecrest assemblages R. BUCKLEY Department of Biogeography and Geomorphology, Research School of Pacific Studies, Australian National University, PO Box 4, Canberra, Australia 2600 though Zygochloa and Chamaesyce may also be present at lower densities (Buckley 1981b). Why? The first hypothesis to be tested is that there are consistent differences in texture or fertility between the soils occupied by the species concerned. Abstract There are two distinct plant assemblages on the dunecrests of the Simpson Desert. The first is characterized by Zygochloa paradoxa and Cham- aesyce wheeled. The second is marked by the pre- sence of Calotis erinacea and Helichrysum am- biguum. There are no consistent differences in tex- ture or fertility between the soils occupied by these four species on the sandridges though there are some differences in other habitats. Hence static soil differences cannot account for the distribution of the two assemblages. It is suggested that the Calotis patches represent populations spread from focal plants established after the 1960s drought, but this remains untested. Introduction Though dunecrest vegetation on the central Aus- tralian sandridges is more consistent than that of flanks or swales, there are nevertheless variations. There are, for example, differences between the Gibson and Simpson Deserts (Buckley 1981 a,b), and from north-west to south-east across the Simp- son Desert (Fatchen & Barker 1979; Buckley 1981 b). During 1975-6 there were also two distinct plant assemblages on the Simpson dunecrests. The first was characterized by Zygochloa paradoxa and Chamaesyce wheeleri. The second was marked particularly by the presence of Calotis erinacea, often accompanied by Helichrysum ambiguum. Present address; AMDEL, PO Box 114, Eastwood, Australia 5063. Methods Soils supporting each of the four species were sampled at 10-30 cm and 2 m depth and analysed for particle-size distribution, total C and N and extract- able P and Ca as described by Buckley (1982). Compaction and mobility were assessed in the field with the aid of a spike penetrometer and sand traps respectively. These were described by Buckley (1979; pp. 70-1): the measures are only relative, and have therefore been converted to indices. The soil nutrient status of the four species were compared by discriminant analysis. I compared various dis- criminant algorithms by using the derived dis- criminant functions to classify the original data (testing equality of the covariance matrices using Box's M and its associated F-test: Cooley & Lohnes 1971), and chose the method which classified the greatest proportion of cases correctly. Results and discussion Before discussing soil hardness and mobility, it is important to note that the four species have different growth forms, as summarized in Table 1. This affects their responses to sand mobility. Pene- trometer readings on a scale of 1 to 4 are summarized in Table 2. I took readings from a selection of individual plants on dunecrests and slopes and also from anomalous habitats such as grader banks. I tried to choose representative habitats in appropriate proportions, but the figures in Table 2 cannot be described as a random sample. In particular, the anomalous habitats are over-represented in view of their small area relative to the overall area of the 0307-692X/82/120(M)415 $02.C 1982 Blackwell Scientific Publications