Do ecological analogues assemble their common features in the same order? An investigation of regularities in evolution, using sand-dwelling lizards as examples

Abstract

Ecological analogues provide an opportunity to assess the regularity of the evolutionary process and some of the factors that control this. By using cladistic methods of phylogeny reconstruction, it is possible to see if two or more stocks have responded in the same way to similar environments, or whether factors such as accident, contingency and previous history have restricted such orderliness. A method is described for comparing the sequences in which ecological analogues assemble their common traits, including ways of handling incomplete data and of testing the statistical significance of the results. The method is applied to situations that mainly involve lizard groups independently associated with aeolian sand habitats. A case where the lineages concerned are closely related is contrasted with two where this is not so. The first instance involves an intrafamilial comparison of three members of the lacertid genera Meroles and Acanthodactylus , which show strong concordance in trait order and great similarity in their independently acquired traits. The other comparisons are of members of different families: Meroles anchietae (Lacertidae), Uma (Phrynosomatidae), Phrynocephalus arabicus (Agamidae) and Pristurus carteri (Gekkonidae). Here concordance in trait order is much lower and independently developed traits often show substantial differences in the various groups. The principal reason for the disparity in results appears to be the much longer and more varied separate histories of the lineages involved in the interfamilial comparisons before they finally entered aeolian sand. These historical differences result in particular independently acquired features developing much earlier in some lineages than others and in the development of phylogenetic constraints and proclivities that influence the detailed ways some environmental problems are solved. Finally, no evidence could be found that traits which evolve in similar sequences in different lineages are developmentally interconnected.