Get access

Random amplified polymorphic DNA (RAPD) and quantitative trait analyses across a major phylogeographical break in the Mediterranean ragwort Senecio gallicus Vill. (Asteraceae)

Authors

  • Hans Peter Comes,

    Corresponding author
    1. School of Environmental and Evolutionary Biology, University of St Andrews, St Andrews, Fife KY16 9TH, UK
    Search for more papers by this author
    • *

      Present address: Institut für Spezielle Botanik und Botanischer Garten, Johannes Gutenberg-Universität Mainz, D-55099 Mainz, Germany.

  • Richard J. Abbott

    1. School of Environmental and Evolutionary Biology, University of St Andrews, St Andrews, Fife KY16 9TH, UK
    Search for more papers by this author

H. P. Comes. Fax: + 49-6131393524; E-mail:comes@mail.uni-mainz.de

Abstract

Random amplified polymorphic DNA (RAPD) and quantitative trait variation of the widespread and ephemeral Senecio gallicus were surveyed in 11 populations sampled from the Iberian Peninsula and southern France. The aim of the study was to compare population relationships and levels of geographical differentiation with chloroplast (cp) DNA and allozyme variation assessed previously in the same populations. Employing multivariate statistics, a moderate level of intraspecific differentiation was observed among populations from Iberian coastal and inland regions for both RAPDs and quantitative traits. However, RAPDs provided greater resolution in identifying additional population structure within the hypothesized, Pleistocene refugial source area of the species in coastal Iberia. A major part of the geographical subdivision in RAPD and quantitative traits was concordant with the coastal vs. inland divergence as previously inferred from cpDNA haplotype frequencies, but strongly contrasted with the geographical uniformity of the species for allozymes. This concordance across various nuclear and cytoplasmic markers (RAPDs/quantitative traits, cpDNA) suggests that geographical uniformity for allozymes is more attributable to low rates of evolution and/or small genome sampling rather than high rates of pollen dispersal, slow rates of nuclear lineage sorting, or indirect balancing selection. The present study underscores the value of using additional classes of nuclear markers for narrowing the numbers of competing causal hypotheses about intraspecific cpDNA–allozyme discrepancies and their underlying evolutionary processes.

Ancillary