RAI, HARDEEP S.1*, HEATH O'BRIEN1, RICHARD G. OLMSTEAD2, and SEAN W. GRAHAM1. 1Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada, T6G 2E9; 2Department of Botany, University of Washington, Seattle, Washington, USA 98195. - Phylogeny of the cycads and their placement in the seed plants, as inferred from a large chloroplast data set.
The cycads have a central place in our understanding of the
phylogenetic relationships of the seed plants. Their circumscription
into two suborders (Zamiineae and the monofamilial Cycadineae) is
supported by various molecular studies and is further confirmed here.
We investigated higher-order relationships by sampling a large portion
of the chloroplast genome for Bowenia, Cycas, Ceratozamia, Dioon,
Encephalartos, Stangeria and Zamia, representing all the families,
subfamilies and tribes in Stevenson’s (1992) treatment of the cycads.
We obtained ~13.5 kb (unaligned) of DNA sequence data per taxon,
spanning a diverse range of coding sequences, introns and intergenic
spacers dispersed throughout the chloroplast genome. Our results are
largely congruent with published molecular studies, and provide
substantial support for most of the inferred backbone of cycad
phylogeny. Dioon is strongly supported as the sistergroup of the
remaining members of Zamiineae. Our findings also support the novel
arrangement of Stangeria with subfamily Zamioideae (represented here
by Zamia and Ceratozamia), as was recently noted by Bogner and
Francisco-Ortega. The precise placement of Bowenia was not fully
resolved, but its membership in Stangeriaceae was strongly rejected.
In contrast to the other seed plants, cycad chloroplast genomes share
two features of their molecular evolution with Ginkgo – a slower rate
of evolution and an apparently elevated transition:
transversion ratio. The question of cycad placement within the seed
plants remains unresolved, and analyses based on different gene
samplings, taxon samplings and phylogenetic optimality criteria
provide different arrangements of the five living spermatophyte
groups. Split-decomposition analysis suggests that there is
conflicting signal in the seed-plant data, possibly a function of
long-branch distortion. This conflict appears to be reduced by
improving the taxon sampling. None of our analyses support a
sistergroup relationship between Ginkgo and the cycads, but this
possibility could not be rejected using parsimony-based parametric
bootstrapping.
Key words: Bowenia, chloroplast phylogeny, Cycads, Ginkgo, SOWH test, Stangeria