Morphological phylogenetic analysis of Ophrys (Orchidaceae): insights from morpho‐anatomical floral traits into the interspecific relationships in an unresolved clade. (17th September 2015)

• Record Type: Journal Article
• Authors: Francisco, Ana; Porto, Miguel; Ascensão, Lia
• Is Part Of: Botanical journal of the Linnean Society. Volume 179:Number 3(2015:Nov.)
• Page Start: 454
• Page End: 476
• Publication Date: 2015-09-17
• Subjects: Botany -- Periodicals; 580
• DOI: 10.1111/boj.12332
• Languages: English
• ISSNs: 0024-4074

Abstract

<abstract abstract-type="main"> <title> <x xml:space="preserve">Abstract</x> </title> <p>Reconstructing the phylogeny of the sexually deceptive orchid genus <italic>Ophrys</italic> is crucial to our understanding of the evolution of its complex floral morphology. Molecular phylogenetic analyses showed that section <italic>Pseudophrys</italic> forms a well supported clade with <italic>Ophrys bombyliflora</italic>, <italic>O. tenthredinifera</italic> and <italic>O. speculum</italic>, but were unable to elucidate the relationships between these four groups of taxa. Here we conduct a morphological phylogenetic analysis of this unresolved clade of <italic>Ophrys</italic> based on a data matrix of 45 macro‐ and micromorphological and anatomical floral characters, using maximum parsimony and Bayesian inference. Our cladistic analysis yielded a single most parsimonious tree and a Bayesian 50% majority‐rule consensus tree which differed in their overall topology but agreed that <italic>O. tenthredinifera</italic> and <italic>O. bombyliflora</italic> are not sister groups. The phylogenetic placement of <italic>O. tenthredinifera</italic> was ambiguous since it shares six valid synapomorphies each with the cluster of <italic>O. speculum</italic>–<italic>O. bombyliflora</italic> and with section <italic>Pseudophrys</italic>. In contrast, <italic>O. bombyliflora</italic> is most likely the sister group to <italic>O. speculum</italic>, a finding that rejects an earlier morphological phylogenetic hypothesis and favours the existing molecular trees based on nuclear ITS rather than plastid data. © 2015 The Linnean Society of London, <italic>Botanical Journal of the Linnean Society</italic>, 2015, <bold>179</bold>, 454–476.</p> </abstract>