Based on phylogenetic analyses using rDNA and plastid sequence data, and the examination of morphological characters, we infer that Pleurostylia, as currently delimited, is a polyphyletic group. Pleurostylia serrulata and two newly described species from Africa are part of the New World Crossopetalum lineage. By contrast, Pleurostylia s. s. consists of all remaining species, which form a clade nested within a primarily Malagasy lineage. We present preliminary evidence that Pleurostylia opposita, the most geographically widespread species in the entire Celastraceae family, may include cryptic species. Although molecular evidence supports the monophyly of Crossopetalum, this genus does not appear to be well defined by a single morphological synapomorphy, rather the following combination of character states is diagnostic: 4-merous flowers; stamens inserted in the sinuses of a ± 4-lobed intrastaminal floral disk; pistil (2- or) 4-locular with one atropous ovule per locule; stigma conspicuously (2-) 4-branched; drupe usually single seeded and asymmetric with an excentric style remnant; and seeds exarillate with ramified postchalazal bundles visible on the surface and with endosperm present. A taxonomic treatment of Crossopetalum in Africa is presented, in which the new combination Crossopetalum serrulatum is proposed and two new species, C. bokdamii and C. mossambicense , are described and their conservation status assessed.
Supplementary Figure 1: Parsimony-based strict consensus of the 26S rDNA data matrix for the Pleurostylia analysis.
Parsimony-based strict consensus of the 26S rDNA data matrix for the Pleurostylia analysis. Jackknife values ≥ 50% are shown above each branch
Darbyshire16Jan2016_FIGS1_26S rDNA Pleurostylia.jpg
Supplementary Figure 2: Parsimony-based strict consensus of the ITS rDNA data matrix for the Pleurostylia analysis.
Parsimony-based strict consensus of the ITS rDNA data matrix for the Pleurostylia analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS2_ITS rDNA Pleurostylia.jpg
Supplementary Figure 3: Parsimony-based strict consensus of the rDNA (ITS + 26S) data matrix for the Pleurostylia analysis.
Parsimony-based strict consensus of the rDNA (ITS + 26S) data matrix for the Pleurostylia analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS3_rDNA ITS+26S Pleurostylia.jpg
Supplementary Figure 4: Parsimony-based strict consensus of the matK data matrix for the Pleurostylia analysis. Jackknife values ≥ 50% are shown above each branch.
Parsimony-based strict consensus of the matK data matrix for the Pleurostylia analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS4_matK Pleurostylia.jpg
Supplementary Figure 5: Parsimony-based strict consensus of the trnL-F data matrix for the Pleurostylia analysis.
Parsimony-based strict consensus of the trnL-F data matrix for the Pleurostylia analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS5_trnL-F Pleurostylia.jpg
Supplementary Figure 6: Parsimony-based strict consensus of the plastid (matK + trnL-F) data matrix for the Pleurostylia analysis.
Parsimony-based strict consensus of the plastid (matK + trnL-F) data matrix for the Pleurostylia analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS6_matK+trnL-F Pleurostylia.jpg
Supplementary Figure 7: Parsimony-based strict consensus of the molecular data matrix for the Pleurostylia analysis.
Parsimony-based strict consensus of the molecular data matrix for the Pleurostylia analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS7_molecular data matrix Pleurostylia.jpg
Supplementary Figure 8: Parsimony-based strict consensus of the 26S rDNA data matrix for the Crossopetalum analysis.
Parsimony-based strict consensus of the 26S rDNA data matrix for the Crossopetalum analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS8_26S rDNA Crossopetalum.jpg
Supplementary Figure 9: Parsimony-based strict consensus of the ITS rDNA data matrix for the Crossopetalum analysis.
Parsimony-based strict consensus of the ITS rDNA data matrix for the Crossopetalum analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS9_ITS rDNA Crossopetalum.jpg
Supplementary Figure 10: Parsimony-based strict consensus of the rDNA (ITS + 26S) data matrix for the Crossopetalum analysis.
Parsimony-based strict consensus of the rDNA (ITS + 26S) data matrix for the Crossopetalum analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS10_rDNA ITS+26S Crossopetalum.jpg
Supplementary Figure 11: Parsimony-based strict consensus of the matK data matrix for the Crossopetalum analysis.
Parsimony-based strict consensus of the matK data matrix for the Crossopetalum analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS11_matK Crossopetalum.jpg
Supplementary Figure 12: Parsimony-based strict consensus of the trnL-F data matrix for the Crossopetalum analysis.
Parsimony-based strict consensus of the trnL-F data matrix for the Crossopetalum analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS12_trnL-F Crossopetalum.jpg
Supplementary Figure 13: Parsimony-based strict consensus of the plastid (matK + trnL-F) data matrix for the Crossopetalum analysis.
Parsimony-based strict consensus of the plastid (matK + trnL-F) data matrix for the Crossopetalum analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS13_plastid matK+trnL-F Crossopetalum.jpg
Supplementary Figure 14: Parsimony-based strict consensus of the molecular data matrix for the Crossopetalum analysis.
Parsimony-based strict consensus of the molecular data matrix for the Crossopetalum analysis. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS14_molecular data matrix Crossopetalum.jpg
Supplementary Figure 15: Parsimony-based strict consensus of the trnL-F data matrix for the Crossopetalum analysis that includes C. bokdamii.
Parsimony-based strict consensus of the trnL-F data matrix for the Crossopetalum analysis that includes C. bokdamii. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS15_trnL-F Crossopetalum incl bokdamii.jpg
Supplementary Figure 16: Parsimony-based strict consensus of the plastid (matK + trnL-F) data matrix for the Crossopetalum analysis that includes C. bokdamii.
Parsimony-based strict consensus of the plastid (matK + trnL-F) data matrix for the Crossopetalum analysis that includes C. bokdamii. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS16_matK+trnL-F Crossopetalum incl bokdamii.jpg
Supplementary Figure 17: Parsimony-based strict consensus of the molecular data matrix for the Crossopetalum analysis that includes C. bokdamii.
Parsimony-based strict consensus of the molecular data matrix for the Crossopetalum analysis that includes C. bokdamii. Jackknife values ≥ 50% are shown above each branch.
Darbyshire16Jan2016_FIGS17_data matrix Crossopetalum incl bokdamii.jpg
Table 1: Data-matrix and tree statistics for each of the phylogenetic analyses.
Data-matrix and tree statistics for each of the phylogenetic analyses. “CI” = ensemble consistency index (Kluge and Farris 1969) on the most parsimonious tree(s) for the parsimony-informative characters. “RI” = ensemble retention index (Farris 1989).
Table 1_data matrix.docx