Canonical variate analysis and the generalized distance Dz statistic are useful ways of analyzing within and between variation in species. Little attempt has been made previously to quantify and assess the variation in calyptrate Diptera other than in descriptive terms. In this study, the variation was assessed for 11 species of muscoid Diptera and populations of the five calliphorid species were analyzed in greater detail. Nine morphometric characters of the wing were used. Most of the variance (937:) fell in the first three canonical variates for the species comparison. and emphasised shape atrributes. although size was important. In particular. the position of vein R , . humeral vein 11. and the discal cell suggested that the sarcophagid flies formed an intermediate group between the calliphorid and muscid flies examined. Size components were more important than shape components within species. and these differences between populations could not be related to a geographic cline.
IntroductionIn this study, the similarities between several species of muscoid flies are examined with respect to wing characters. The examination of character variation within a single functional "organ" system can provide much information in closely related taxa. It could be argued that the wing reflects much of the evolutionary history of the group, the shape being relatively stable unlike size which can be influenced by nongenetic factors. It is from studying patterns of character variation in populations, including geographic variation, that an insight into relationships between those populations may be gained. Many workers have previously assessed single character variation over geographic range by univariate methods, but increasingly, groups of characters are being assessed simultaneously using multivariate techniques (Sokal 1965;Gould and Johnston 1972;Thorpe 1976). Shape factors especially, because they reflect more complex interaction, can only be adequately assessed by using multivariate methods that simultaneously consider the variation of many characters (Atchley 1971 ; Atchley and Hensleigh 1974). Further, as adaptations of populations are multidimensional, variation within and between taxa may be better resolved by using such multivariate methods as canonical variate analysis (= multiple discriminate function analysis). This type of intra-specific variation analysis is carried out with diverse aims, for it can be used to describe variation patterns of a species, or to subdivide a variable species into subspecific units, thereby gaining an insight into the evolutionary forces acting upon a species (Sokal 1965). The multivariate analysis allows for the efficient summarization of large data sets so that populations or species may be compared or contrasted using fewer metrics. Effectively data sets representing the morphology of taxa can be expressed as a single data point in multivariate hyperspace. Biological interpretation is made from the relative positions and distances between these data points.In wings of di...