Estudos biossistemáticos no gênero habenaria willd. (ORCHIDACEAE)

Abstract

Habenaria is one of the richest orchid genera in Brazil, with ca. 170 species found in all areas of the country. A molecular phylogeny with neotropical species of the genus was carried out and from this phylogeny it was possible to infer many evolutionary scenarios in the genus. One of these could be a correlation between chromosome numbers and the recovered clades, demonstrating that this type of analysis would be potentially fitted in unraveling some taxonomic and phylogenetic aspects of the genus, and this was the first goal of this work. Two of the largest clades recovered in this same phylogeny (the Habenaria sect. Nudae clade and the Mexican clade) were not well resolved. Concerning the Mexican clade, the performed phylogeny included a small sample, with species morphologically divergent clustered in the same clade, a fact that led the authors of this paper to suggest that increasing the sample would be crucial for a better clade resolution. In turn, Habenaria sect. Nudae was well sampled in the performed phylogeny, but in this phylogeny only two molecular markers were used, carrying us to the third goal of this work: the usage of a larger set of molecular markers in the phylogenetic analysis in order to improve the resolution between taxa, and also the implementation of other tools such as the analysis of the genetic and morphological variation of the involved specimens. In this work it was possible to observe that the chromosome numbers did not present a direct and uniform correspondence to the clades recovered in the phylogeny and that chromosome variations observed in the species of the genus possibly arose independently. The trees obtained after increasing the sampling of Mexican Habenaria species revealed a strong geographic structuring and the splitting of the Mexican species in two subclades non directly related. Morphological differences suggest that adaptations to distinct pollinators or pollination strategies can be responsible for the observed splitting, and in this context we propose a recircumscription of the sections Habenaria and Clypeatae and the establishment of the new section Novemfidae. We carried out, with the studied species in the Habenaria nuda complex, a phylogenetic analysis using three markers. The larger number of markers was insufficient in resolving affinities between the studied morphotypes. With this scenario, we made analyses of genetic variability using microsatellite markers and morphometric multivariate analyses involving 293 specimens pertaining to 10 morphotypes; The results indicate the existence of new species akin to the H. nuda complex and the status change of a variety to species. After the analyses were completed and using them as a guidance, we performed a more detailed study of characteres non approached by morphometry and it was possible to better evaluate the identity of many taxa, to circumscribe more acurately the species H. nasuta, H. nuda, and H. rodriguesii and to describe siw new species. Taking into account all the results obtained in this work, we conclude that the genus Habenaria is in an active process of speciation and can be considered as an important evolutionary model to be studied in greater details and with diversified tools.