Melastomataceae is a large plant family with approximately 5000 species in about 160 genera. Members of the family are easily recognizable by their typical leaves with sub-parallel secondary veins (acrodomal venation), and porose and flexed anthers (geniculate). Over 40% of the species in the family belong to the tribe Miconieae, a group of 30 genera with fleshy fruits found exclusively in the New World tropics. In Neotropical forests the Melastomataceae are among the five most diverse families and representatives of the tribe Miconieae are a dominant element in the forest understory. Species of Miconieae play an important role in the diet of frugivorous birds and some mammals in these environments. Moreover, two of the most notable weeds in Hawaii and other Pacific islands belong to this group: Koster's curse (Clidemia hirta) and velvet tree (Miconia calvescens). In spite of all of this, the relationships among species of the Miconieae, and the limits and relationships within genera in this group are poorly understood. This project will assess the evolutionary relationships within the Miconieae with the use of DNA sequences from all three genomes (nucleus, chloroplast and mitochondrion) and morphological data. The morphological data will be gathered using an exemplar species approach coding directly from herbarium, liquid-preserved and live specimens. In total, we expect to include about one fifth of all species from all genera in the tribe. The resulting phylogenetic analyses will be the foundation of character evolution studies (e.g. a-origin of ant-domatia; b-inflorescence position) and biogeographical studies (e.g. origin of Caribbean flora and biogeography). Morphological data (along with other diagnostic characters) will be used to produce interactive web-based keys to the genera of Miconieae, and to species in selected genera. Lastly, our results will be used as the basis of generic realignments, leading to a new intratribal classification system. The present study will have broad impacts in education, domestic and international collaborations, enhancement of participation of underrepresented groups, and dissemination of knowledge. Two doctoral candidates will be intimately involved in the research here proposed, and direct interaction with local scientists at the different countries to be visited will help to increase their knowledge of systematic theory and practices, especially in herbaria and universities where graduate training in these disciplines is not available. The PIs and collaborators come from a variety of institutions, (museums, large universities, and liberal art colleges) from the US, Europe and Latin America. Because of a lack of monographs for the larger genera in the tribe and local experts to identify specimens, the interactive keys will be extremely useful for ecologists, conservation biologists and local scientists. Collections made during the course of this study will help to increase the collections of local herbaria, the NYBG, and FMNH. Collection data, photographs and other information will be available through the internet. Finally, the PI, Fabian Michelangeli, is a young Hispanic scientist (who will collaborate with Walter Judd, co-PI, and Renato Goldenberg co-PI from Brazil) and knowledge gained through this study will enhance the participation of underrepresented minorities in science.
