Micro-evolution in the weevil genus Larinus: the formation of host biotypes and speciation

Authors

  • D. T. BRIESE,

    Corresponding author
    1. †CSIRO Biological Control Unit, Campus International de Baillarguet, 34982 Montferrier-sur-Lez Cedex, France
      CSIRO Entomology, GPO Box 1700, Canberra 2601, Australia. Fax: + 61–6246-4000.
    Search for more papers by this author
  • C. ESPIAU,

    1. †CSIRO Biological Control Unit, Campus International de Baillarguet, 34982 Montferrier-sur-Lez Cedex, France
    Search for more papers by this author
  • A. POUCHOT-LERMANS

    1. ‡School of Pure and Applied Biology, University of Wales, College of Cardiff, Cardiff, UK
    Search for more papers by this author

  • Biological control is a very applied discipline, but occasionally we can look at more fundamental aspects of ecology that affect it - the evolution of host choice is one of these. David Briese is a Principal Research Scientist with CSIRO (Canberra, Australia) and is interested in population ecology in relation to biological control. He recently spent 5 years at Montpellier studying potential control agents of Onopordum thistles, where he collaborated to produce this paper with Christophe Espiau, who supplied the technical expertise on gel electrophoresis, and Annick Pouchot-Lermans, who helped analyse some of the data for her Diploma in Applied Insect Taxonomy at the University of Wales, Cardiff.

CSIRO Entomology, GPO Box 1700, Canberra 2601, Australia. Fax: + 61–6246-4000.

Abstract

Data are presented on allozyme variation between 15 populations of the stenophagous capitulum weevil, Larinus cynarae, and three populations of its congener, L. latus, that had been collected throughout the northern mediterranean range of these species. A phenetic analysis of these data revealed no direct relationship between genetic variation and host-plant association within L. cynarae, but there was a strong geographical structuring of allozyme patterns. Most of the genetic variation was due to differences between geographical regions and variation within these was small. Wright's FST values showed that Italian and Greek populations of L. cynarae were most distinct from L. latus, with southern Iberian, northern Spanish and French populations increasingly less so. This pattern was associated with a cline in the frequencies of certain alleles along this geographical arc from France to Greece. A phenogram of Nei's genetic distances indicated the close genetic relationship between the two species of Larinus and separated the populations of L. cynarae into three allopatric groups. These groups have different host-plant spectra — dominated by Cynara cardunculus in Italy and Greece, Cynara humilis/Onopordum in southern Iberia and Onopordum spp. in France/Northern Spain — and can be considered to be host biotypes of L. cynarae. L. latus, which occurs in Greece and further east is also an Onopordum specialist. An analysis of the phylogeny of this group of Larinus indicates a primary separation into eastern (L. latus) and western (L. cynarae) taxa, with further branching of the L. cynarae lineage into the putative host-biotypes. An hypothesis for the evolution of these taxa is given, based on the evolutionary history of host-plant taxa and geographical constraints.

Ancillary