Ancient Gondwana break‐up explains the distribution of the mycoheterotrophic family Corsiaceae (Liliales)

Abstract

Aim Many plant families have a disjunct distribution across the southern Pacific Ocean, including the mycoheterotrophic family Corsiaceae, which provides a prime example of this biogeographical pattern. A better grasp of the family's evolutionary relationships is needed to understand its historical biogeography. We therefore aimed to (1) test the uncertain monophyly of Corsiaceae, (2) define its phylogenetic position, and (3) estimate divergence times for the family, allowing us to assess whether the distribution of the family is the result of vicariance.Location Southern South America and Australasia.Methods We analysed various combinations of mitochondrial and nuclear data to address the monophyly, phylogenetic position and age of Corsiaceae. To test its monophyly, we used a three-locus data set including most monocot orders, and to infer its exact phylogenetic position, we used a five-locus extended data set. We corroborated these findings using an independent plastome dataset. We then used a two-locus dataset with taxa from all monocot orders, and a three-locus dataset containing only taxa of Liliales, to estimate divergence times using a fossil-calibrated uncorrelated lognormal relaxed-clock approach.Results Corsiaceae is a monophyletic family and the sister group of Campynemataceae. This clade is the sister group of all other Liliales. The crown age of Corsiaceae is estimated to be 53 Ma (95% confidence interval 30-76 Ma).Main conclusions Corsiaceae is an ancient family of mycoheterotrophic plants, whose crown age overlaps with the plate-tectonic split of Gondwana, consistent with a vicariance-based explanation for its current distribution.

Pollen Structure and Diversity in Liliales

“…Reticulate or microreticulate sculpturing is the most common type; it occurs in all three major clades of Liliales and also in Corsia. Despite earlier suggested similarities in pollen structure between Corsiaceae and Campynemataceae, especially in pollen sculpturing (Erdtman 1952), our results show similar reticulate or microreticulate sculpturing in both families, a feature that is probably plesiomorphic for Liliales (Mennes et al 2015).…”

“…Sometimes endexine is present, though it may be absent. In some taxa, the exine is reduced to granules or echini resting on a lamellated Kim et al (2013), Mennes et al (2013Mennes et al ( , 2015, and Petersen et al (2013). Campynemataceae and Corsiaceae are shown as forming a clade that is sister to all other Liliales.…”

“…Phylogenetic studies of Liliales differ in the genera represented; they also conflict in some details of relationships (Kim et al 2013;Petersen et al 2013;Mennes et al 2015). As a preliminary estimation of possible character evolution within the order, we used these existing phylogenies to construct a composite tree, with conflicting tologies presented as polytomies ( fig.…”

“…Molecular analyses indicate a close relationship between Campynemataceae (two genera) and the mycoheterotrophic family Corsiaceae (three genera). Corsiaceae are sometimes placed as sister to all other Liliales (Petersen et al 2013); recent data indicate that Campynemataceae and Corsiaceae form a sister pair (Mennes et al 2013(Mennes et al , 2015. The remaining members of Liliales are placed in three clades ( fig.…”

Evolution of mycoheterotrophy in Polygalaceae: The case of Epirixanthes

“…Epirixanthes appears to have evolved relatively recently toward mycoheterotrophy, as many groups of mycoheterotrophic plants are much older; e.g., Triuridaceae have an estimated mean stem age of 84 Ma ( Mennes et al, 2013 ), and Corsiaceae have an estimated mean stem age of 70 Ma ( Mennes et al, 2015 ). The relatively wide range of AMF the genus is symbiotically linked to possibly represents a relic of a more generalistic ancestral fungal symbiosis, similar to that of Polygala ( Rath et al, 2013 ) or Salomonia .…”