Inferring speciation modes in a clade of Iberian chafers from rates of morphological evolution in different character systems
1 Zoologische Staatssammlung München, Münchhausenstr. 21, 81247 Munich, Germany
2 Department of Entomology, Natural History Museum, Cromwell Road, London SW7 5BD, UK
3 Museo Nacional de Ciencias Naturales, José Gutiérrez Abascal 2, 28006 Madrid, Spain
4 Institute of Evolutionary Biology (CSIC-UPF), Passeig Maritim de la Barceloneta 37-49, 08003 Barcelona, Spain
BMC Evolutionary Biology 2009, 9:234 doi:10.1186/1471-2148-9-234Published: 15 September 2009
Studies of speciation mode based on phylogenies usually test the predicted effect on diversification patterns or on geographical distribution of closely related species. Here we outline an approach to infer the prevalent speciation mode in Iberian Hymenoplia chafers through the comparison of the evolutionary rates of morphological character systems likely to be related to sexual or ecological selection. Assuming that mitochondrial evolution is neutral and not related to measured phenotypic differences among the species, we contrast hypothetic outcomes of three speciation modes: 1) geographic isolation with subsequent random morphological divergence, resulting in overall change proportional to the mtDNA rate; 2) sexual selection on size and shape of the male intromittent organs, resulting in an evolutionary rate decoupled to that of the mtDNA; and 3) ecological segregation, reflected in character systems presumably related to ecological or biological adaptations, with rates decoupled from that of the mtDNA.
The evolutionary rate of qualitative external body characters was significantly correlated to that of the mtDNA both for the overall root-to-tip patristic distances and the individual inter-node branches, as measured with standard statistics and the randomization of a global comparison metric (the z-score). The rate of the body morphospace was significantly correlated to that of the mtDNA only for the individual branches, but not for the patristic distances, while that of the paramere outline was significantly correlated with mtDNA rates only for the patristic distances but not for the individual branches.
Structural morphological characters, often used for species recognition, have evolved at a rate proportional to that of the mtDNA, with no evidence of directional or stabilising selection according to our measures. The change in body morphospace seems to have evolved randomly at short term, but the overall change is different from that expected under a pure random drift or randomly fluctuating selection, reflecting either directional or stabilising selection or developmental constraints. Short term changes in paramere shape possibly reflect sexual selection, but their overall amount of change was unconstrained, possibly reflecting their lack of functionality. Our approach may be useful to provide indirect insights into the prevalence of different speciation modes in entire lineages when direct evidence is lacking.