Phylogenetic relationships of Santalales with insights into the origins of holoparasitic Balanophoraceae

Abstract

To date molecular data have not revealed the exact phylogenetic position of Balanophoraceae in relation to hemiparasitic Santalales. To elucidate the phylogeny of Santalales and the position of Balanophoraceae, three plastid genes (matK, rbcL, accD), three nuclear genes (SSU and LSU rDNA and RPB2) and one mitochondrial gene (matR) from 197 Santalales samples (including 11 Balanophoraceae species) were analyzed with parsimony, maximum likelihood and Bayesian inference methods. Our results demonstrate that Balanophoraceae is composed of two well‐supported clades: a relatively slow‐evolving one including Dactylantus, Hachettea, and Mystropetalon (Mystropetalaceae) and an extremely fast‐evolving one composed of the remaining Balanophoraceae s.str. Support for monophyly of the two clades was low, thus it appears holoparasitism has arisen twice independently in Santalales. These two clades appeared during a time of great change in the order (ca. 100 Ma) when several major evolutionary innovations emerged, e.g., the root hemiparasites of Santalaceae s.l., the first aerial parasites (Misodendraceae), herbaceous root parasites (Schoepfiaceae), root parasitic Loranthaceae (the ancestors of aerial parasitic mistletoes), as well as the holoparasites in Balanophoraceae and Mystropetalaceae.

Evolution of pollen morphology in Loranthaceae

“…Except for one genus ( Elytranthe ), all genera treated by Liu and Qiu (1993) and Han et al (2004) are covered by new material in our study.

Figure 2.Molecular phylogram and phylogeographic-systematic framework for the Loranthaceae. Shown is the ‘best-known’ maximum likelihood tree based on a concatenated genus-consensus sequence matrix including data from two nuclear-encoded ribosomal DNA regions (18S, 25S rDNA) and three plastid regions ( mat K gene, trn L intron and trn L- trn F intergenic spacer, and rbc L gene) rooted with Nuytsia as the first diverging lineage (following Vidal-Russell & Nickrent 2008b; Su et al 2015). Stippled lines indicate branches that have been reduced by factor 2.

…”

“…Shown is the ‘best-known’ maximum likelihood tree based on a concatenated genus-consensus sequence matrix including data from two nuclear-encoded ribosomal DNA regions (18S, 25S rDNA) and three plastid regions ( mat K gene, trn L intron and trn L- trn F intergenic spacer, and rbc L gene) rooted with Nuytsia as the first diverging lineage (following Vidal-Russell & Nickrent 2008b; Su et al 2015). Stippled lines indicate branches that have been reduced by factor 2.…”

“…Nickrent et al (2010) recognised one possibly para- or polyphyletic tribe (Psittacantheae) and two monophyletic tribes (Elytrantheae, type genus not sequenced yet, and Lorantheae) within the aerial parasite clade. Although four of Nickrent et al’s (2010) 11 Psittacantheae and Lorantheae subtribes do not conform with molecular clades reconstructed by Vidal-Russell and Nickrent (2008b) and Su et al (2015), they are also not rejected by molecular data with strong support. Many branches in the molecular phylograms produced so far, simply lack high support.…”

“…A recent study using molecular data clarified some phylogenetic relationships within the family (Vidal-Russell & Nickrent 2008b); that study forms – to some degree – the basis of a revised systematic framework (Nickrent et al 2010; summarised in Table I). The root parasite Nuytsia (monotypic Nuytsieae) has been suggested to represent the first diverging lineage within the Loranthaceae, forming a ‘basal’ grade with the Gaiadendreae comprising the only other two root parasites, Atkinsonia and Gaiadendron (Su et al 2015). The remaining aerial parasitic genera of the family are considered to be monophyletic (Vidal-Russell & Nickrent 2008b; Nickrent et al 2010; Su et al 2015); the corresponding clade in molecular phylograms is, however, poorly supported by bootstrapping (BS < 60) and moderately to high using Bayesian probabilities (PP > 0.8; see online Supplemental Material [OSM] File S1).…”

“…The root parasite Nuytsia (monotypic Nuytsieae) has been suggested to represent the first diverging lineage within the Loranthaceae, forming a ‘basal’ grade with the Gaiadendreae comprising the only other two root parasites, Atkinsonia and Gaiadendron (Su et al 2015). The remaining aerial parasitic genera of the family are considered to be monophyletic (Vidal-Russell & Nickrent 2008b; Nickrent et al 2010; Su et al 2015); the corresponding clade in molecular phylograms is, however, poorly supported by bootstrapping (BS < 60) and moderately to high using Bayesian probabilities (PP > 0.8; see online Supplemental Material [OSM] File S1). Nickrent et al (2010) recognised one possibly para- or polyphyletic tribe (Psittacantheae) and two monophyletic tribes (Elytrantheae, type genus not sequenced yet, and Lorantheae) within the aerial parasite clade.…”