Distribution of body sizes in arthropod taxa and communities

Body size distributions of 16 arthropod taxa and 7 communities (species assemblages) were analysed. Body size distributions in taxa and in communities were close to lognormal (10 distributions out of 23 were not significantly different from lognormal, P>0.05). Their skewness was mostly positive (in 17 out of 23 data sets), but small (-0.26 to 0.43; in 15 out of 23 data sets it was not significantly different from zero, P>0.05). No evidence in support of the hypothesis that the decrease in the number of species of animals (S) from medium to the largest body length (L) can be described as S similar to L(xd) with x(d) close to -2 (May 1978, 1986) was found, as x(d) ranged from -9.1 to -1.8 in taxa and from -6.1 to -2.1 in communities. Alternative patterns were observed: there was a close correlation between x(d), i.e., the rate of decrease in the number of species from the modal to the largest body sizes, and x(a), i.e., the rate of increase in the number of species from the smallest to the modal body sizes. The most important result is that standard deviation and, consequently, x(a) and x(d), were correlated with the number of species within a taxon. These results indicate that both x(a) and x(d) values are determined by evolutionary dynamics idiosyncratic for each taxon. A large variability and no consistent pattern in the x(a) and x(d) values in communities indicate that there may be no general assembly rules determining overall body size distribution of species in insect taxocoenoses.