Phytogeny and behavioural evolution in ants, with a discussion of the role of behaviour in evolutionary processes

Publisher Website
Google Scholar
Abstract
The family Formicidae is redefined on the basis of all common characters shared by its members, which allows a hypothetical sketch of its unknown presumed ancestor. The first cladistic analysis is presented of internal ant phylogeny at the subfamily level, based on the 10 extant subfamilies commonly recognized in the literature. Although some weaknesses remain, this phylogenetic reconstruction shows several concrete improvements. None of the fossil subfamilies allow comparison with the recent subfamilies by neontological criteria. Among recent subfamilies no autapomorphic characters have been found for the Aneuretinae which clearly show only primitive character states easy to identify by comparison with the Dolichoderinae. The former subfamily has been regarded as a straightforward synonym of the latter. The Ponerinae have been still characterized tentatively on the basis of two weak synapomorphies but the real monophyly of its members appears to be questionable. The subfamilial phylogeny, as a whole, appears relatively weakly documented on several branchings with exception of the Nothomyrmeciinae, Dolichoderinae, and Formicinae on one side which can be easily opposed to all remainder recent subfamilies on the other side, by having a constriction after abdominal segments I and II only instead of after segments I, II, and III. Another important conclusion reached in this paper concerns the Formicinae, which can no longer be considered as the most advanced extant ants. A variety of complex and simple behaviours, chosen as examples of general evolutionary trends within the ants, all show a mosaic picture of parallelisms and convergences. Two reasons are offered to explain these findings. The first, particular to ants, is that eusociality represents an evolutionary constraining influence allowing further selection only for a few behaviours like nesting behaviour, communication modes, social parasitism, etc. It is not surprising, therefore, that the few rewarding options appear convergently in different stems within the ants, and among other, unrelated, social insects as well. The second, more general, reason is the understandable tendency of taxonomists to produce phylogenies based on purely morphological characters. An analysis of the major morphological trends among ants shows that all correspond to behavioural evolutionary sequences. Such sequences are, of course, much more difficult to describe and to quantify than is morphology. Only one systematically «pure» (i.e. without morphological correlates and virtually free from important selection pressures) ethological character has been found to be widely consistent with the subfamily classification proposed here and increasing its information contents. This is the manner in which ants transport other adult ants. This is probably the highest rank phylogenetically relevant character detected so far which is based on behaviour only.