Mycoheterotrophy and plastid genome evolution in the early-diverging epidendroid orchid tribe Nervilieae: independent transitions in <i>Epipogium</i> and <i>Stereosandra</i>

Parasitic organisms are of interest in evolutionary biology, often displaying drastic modifications in morphology, physiology, genomes, and ecology. These properties, however, make them challenging from a systematics perspective. Mycoheterotrophy, in which plants become non-photosynthetic parasites on fungi, is an excellent example, and this unique life history has evolved numerous times in the orchid family. Here, we focused on <i>Stereosandra</i>, a genus of mycoheterotrophic orchid comprising a single species, <i>S. javanica</i>, about which little is known. <i>Stereosandra</i> has been placed in the orchid tribe Nervilieae, along with the leafy, autotrophic <i>Nervilia</i>, and the leafless, mycoheterotrophic <i>Epipogium</i>. We characterized the first complete plastid genome for <i>Stereosandra</i> and used nuclear sequence capture to determine its relationships within Nervilieae. This study presents the first genetic data ever produced for <i>Stereosandra</i>. The plastid genome exhibits rampant gene losses, pseudogenes, and reduced size relative to <i>Nervilia</i> but not to the extent seen in <i>Epipogium</i>. There is evidence of relaxed negative selection in six genes in <i>Stereosandra</i>, including <i>matK</i>, which functions in Group IIA intron removal of seven plastid genes, four of which have been lost or pseudogenized in this species. Applying mixture models, plastid genomes provided weak support for a sister position of <i>Stereosandra</i> to a clade of <i>Epipogium</i> + <i>Nervilia</i>. Nuclear phylogenomic analyses provided strong support for the same relationships. Ancestral state reconstruction revealed clear evidence that mycoheterotrophy evolved multiple times in the tribe from leafy ancestors. This study provides a previously unidentified, convergent instance of the evolution of full mycoheterotrophy in plants. We discuss the results in the context of proposed models of reductive plastid genome evolution and the genomic and evolutionary consequences of radical life history shifts in heterotrophic plants.