Molecular Phylogeny & Evolution of Polystachya (Orchidaceae)

The genus Polystachya (subfamily Epidendroideae, tribe Vandeae, and subtribe Polystachynae) with around 150 species is distributed in Madagascar, and tropcial Africa, Asia and America (Polystachya is one of only three orchid genera with such a broad distribution). The species of Polystachya are epiphytic, occasionally lithophytic or terrestrial. Molecular phylogenetic analyses of this genus are planned in this project as a backbone to answer many interesting questions. Molecular analyses will be done using both plastid (matK gene, spacer region trnH/psbA and intron rps16) and nuclear markers; in addition to the multicopy rDNA, internal transcribed spacer (ITS), two low- copy nuclear gene (PRK and LFY/FLO) is planned. Chromosome work done up to now shows that the genus has diploids as well as polyploids with the same basic chromosome number. Since the chromosomes are small, flurochome banding will be done to study the karyotype evolution. In addition to this, in situ hybridization is planned to find the relationships between the diploids and the polyploids and the evolutionary changes in the number of rDNA loci in polyploids with special reference to section Polystachya. The habitat evolution from terrestrial to epiphytic will be tested using the molecular phylogeny. The evolution of genome size in ephiphytes versus terrestrials and diploids versus polyploids will be investigated.

The orchid genus Polystachya comprises approximately 250 species with a centre of distribution in tropical Africa, and also found in Madagascar, tropical Asia and the Neotropics. It was selected as model for phylogenetic studies because of its pantropical distribution, variation in ecology, morphology and ploidy, and a lack of evolutionary information on which to base ongoing taxonomic work. We aimed to: 1) Create a strong, genus-wide phylogenetic hypothesis using plastid DNA sequence data, with which to examine chromosome evolution, biogeography and the infrageneric taxonomy; 2) Look for reticulate evolution in the genus using nuclear and plastid data, and determine the origins of tetraploid Madagascan, Asian, and Neotropical species in more detail; 3) Use AFLP markers to examine the relationships of some closely related species in more detail, at the scale of the pantropical dispersal of the genus and at population level in two Costa Rican species. We used 5.3 kb of DNA from two plastid regions to make a well-supported phylogeny for the 89 species for which material was available for molecular work. The plastid tree could be divided into five main clades, plus a number of species-poor, early diverging lineages. Separate centres of diversity are found in eastern and western Africa, and species found in these areas group together at deep phylogenetic levels, with largely separate diversification in eastern and western Africa for long periods of evolutionary history. There is a directional bias in dispersal from western to eastern Africa, with species originating in western Africa more likely to have subsequently become widespread whereas species originating in eastern Africa are more likely to have remained endemic to the region. This pattern could be due to the tendency of eastern African species tending to be more restricted to montane and sub-montane habitats, and the refugial status of these habitats during Pleistocene climatic fluctuations. Analysis using finer-scale distribution showed that some subclades have diversified within smaller geographical areas but did not give clear results at deep phylogenetic levels. Genome size measurements (but not guard cell size) could be used to infer ploidy from 112 Polystachya accessions. Diploid genome size (1C) ranges from 0.58-1.03 pg whereas polyploids range from 1.10-1.80 pg. Polyploidy has occurred several times during the course of Polystachya evolution and in some cases has been followed by significant diversification; one group of tetraploids has spread rapidly and recently throughout the tropics. Plastid DNA from this group is uninformative about relationships within this group due to a low level of sequence divergence. Analysis of sequences from three low-copy nuclear genes as well as plastid and nuclear ribosomal DNA provided more information on the origins of tetraploid clades. The pantropical tetraploid species group had multiple origins, with the Neotropical members of the group arising independently from the Paleotropical members. Another group of tetraploids endemic to the Malagasy Islands also has hybrid origins, but from more genetically dissimilar parents and this is reflected in the greater range of morphological variation now found within the group, compared to the pantropical tetraploids. Although the gene trees from individual loci were incongruent, comparing them using methods that filtered out incongruent clades unique to a single dataset showed little evidence for hybridisation and introgression among diploid accessions. Incongruences were more likely due to limitations in the phylogenetic information available from individual loci. AFLP data linked Sri Lanka and South East Asia populations with those from Africa and the Malagasy Islands. More detailed, population-level sampling of Polystachya foliosa and P. masayensis, two closely-related species of Polystachya in Costa Rica showed evidence of introgression in only one population from 17 sampled. In general the species appear genetically distinct, despite occurring in the same localities and having recent hybrid origins. Evidence from both nuclear low copy DNA sequences and AFLP markers supported the view that species concepts among the pantropical tetraploids should be redefined, especially within P. concreta.