Historical biogeography and the origin of stomatal distributions inBanksiaandDryandra(Proteaceae) based on their cpDNA phylogeny

Abstract

Banksia and Dryandra have undergone extensive speciation and adaptive radiation, especially in Australia's isolated Southwest Botanical Province. We derive a phylogeny for these groups based on cpDNA sequences and use it to reconstruct their historical biogeography and evolution of leaf traits thought to be adapted to drought and/or nutrient poverty. Slowly evolving regions (trnL intron, trnL/trnF spacer) are used to resolve large-scale relationships; faster evolving regions (rp116 intron, psbA/trnH and trnT/trnL spacers) are used to resolve relationships among closely related species. Banksia is paraphyletic with respect to Dryandra. The lineage underwent a basal split into two clades (here named /Cryptostomata and /Phanerostomata), and four infrageneric taxa supported by morphological cladistic analyses (series Spicigerae, Abietinae, Tetragonae, and Banksia) are not monophyletic. Dispersal-vicariance analysis resolves a southwestern Australian origin for the lineage, with two later expansions to the east followed by vicariance events. Stomatal crypts arose with the /Cryptostomata, which is characterized by tough, long-lived leaves and common in southwestern Australia. Sequestering of stomata also arose multiple times in /Phanerostomata, which is characterized by softer, short-lived leaves and common in moister coastal areas, via inrolling of the margins of narrow leaves and restricting stomata to shallow pits. The hypothesis that sclerophylly preadapted the plants to xeromorphy is supported in the case of shallow stomatal pits and deep stomatal crypts, but not narrow, needle-like leaves.

A new suprageneric classification of the Proteaceae, with an annotated checklist of genera

“…We have recognised four tribes in the subfamily Grevilleoidae, all of which are resolved in some molecular analyses but only one of which, the Banksieae, is strongly supported in all analyses in which it has been tested (Hoot & Douglas 1998, Mast & Givnish 2002. We have recognised these groups at tribal rank in part to maintain nomenclatural stability of a number of better supported subtribes already recognised by Johnson and Briggs (1975) and Douglas (1995) and in the expectation that phylogenetic research currently in progress (by Austin Mast and co-workers) will resolve clades at the tribal level with greater confidence in the near future.…”

“…A supertree analysis using matrix representation and parsimony ('MRP' -BinindaEmonds et al 2002, 2004 was conducted that synthesised the results of published and unpublished molecular phylogenetic analyses of the Proteaceae (Hoot & Douglas 1998, Mast & Givnish 2002, Barker et al 2002 (Analysis 1), WB&D ITS and WB&D rbcL -see above). The terminal taxa were currently recognised genera of Proteaceae.…”

“…Although monophyly of a number of genera has been minimally tested using molecular data, only Banksia has been subject to detailed sampling at the species level (Mast 1998, Mast & Givnish 2002 and was shown to be non-monophyletic by inclusion of Dryandra, the genus that was previously thought to be its sister group (Thiele & Ladiges 1996). Barker et al (2002) also showed that several of the African genera (Mimetes, Sorocephalus, Spatalla and Paranomus) could potentially be non-monophyletic, but with less comprehensive sampling of species.…”

“…Hoot and Douglas (1998) is an analysis of representatives of most genera of Proteaceae, based on an alignment of DNA sequences for the chloroplast loci atpβ and the atpβ-rbcL intergenic spacer. Mast and Givnish (2002: fig. 2) is an analysis of 28 species of Banksia plus five species of Dryandra and representatives of 12 outgroup genera, based on an alignment of cpDNA sequences for the rpl16 intron, trnL intron and trnL-trnF intergenic spacer.…”

Potential use of DNA barcoding for the identification of Salvia based on cpDNA and nrDNA sequences

“…Although psbA-trnH has highly variable interspecific distances for Salvia, it also has the highest intraspecific distances. This discrepancy directly causes a large overlap and greatly reduces its species identification rate because insertions and deletions are common in this genetic region, even in closely related species (Mast and Givnish, 2002;Miller et al, 2003;Winkworth and Donoghue, 2005). This limitation is supported by the wide variation in sequence length, ranging from 422 to 455.…”