Phylogenetic relationships in brown argus butterflies (Lepidoptera: Lycaenidae: Aricia) from north-western Europe

Lycaenid butterflies of the Aricia agestis-artaxerxes complex pose an unresolved taxonomic and conservation problem in north-western Europe. Two key issues require resolution: (i) how many species of Aricia occur in north-western Europe and what are their distributions?; (ii) how is the morphological variation observed in north-western Europe best explained? We investigated phylogenetic relationships and phylogeographic patterns in this species group using mitochondrial and nuclear markers in comparison with morphological variation. A 325 bp fragment of the mitochondrial cytochrome-b gene was sequenced from 179 individuals representing 18 populations from the UK and Scandinavia. Seventeen enzyme-coding loci were analysed from 538 individuals from the same populations. Highly congruent phylogenies between mitochondrial and allozyme markers demonstrate that the sample is composed of two closely related species, A. agestis and A. artaxerxes. Both marker types also suggest that Scottish and northern Scandinavian A. artaxerxes populations are conspecific, and consequently do not support the endemic status of A. artaxerxes in the UK. The subspecies division of British populations of A. artaxerxes is also not supported by phylogenetic analyses. Allozyme and mitochondrial analyses cluster two populations from the Peak District, UK, differently. The former suggests that they are A. artaxerxes whilst the latter suggests they are A. agestis. Further research is required to find the reason for this disagreement, which could be associated with the different dynamics of nuclear and mitochondrial genes across a hybrid zone between the two species. (C) 2002 The Linnean Society of London.