Phylogeny of the Paniceae (Poaceae: Panicoideae): integrating plastid DNA sequences and morphology into a new classification.

Included in the PACMAD clade of the family Poaceae (Panicoideae, Arundinoideae, Chloridoideae, Micrairoideae, Aristidoideae, Danthonioideae), the tribe Paniceae s.l. is one of the largest tribes of the subfamily Panicoideae, with more than 2000 species. This tribe comprises a huge morphological, cytological and physiological diversity represented by different inflorescence types, several basic chromosome numbers, and at least four major photosynthetic pathways. The tribe Paniceae has been the subject of molecular studies that have confirmed its paraphyly: two major clades were recognized based on their basic chromosome numbers (x = 9, x = 10). The x = 10 Paniceae clade is sister to the Andropogoneae-Arundinelleae s.s. clade (x = 10), while the combined x = 10 clade is sister to the x = 9 clade that contains the remaining genera of Paniceae. As a result of a recent realignment within the tribe in terms of the phylogenetic position of minor and major Paniceae genera, a reanalysis of the whole sampling is performed and new underrepresented taxa are discussed. A total of 155 genera, currently considered within subfamily Panicoideae, are represented here by almost all genera of Paniceae s.l., representatives of Andropogoneae and Arundinelleae s.s., and the endemic and small tribe Steyermarkochloeae; we also included specimens of subfamily Micrairoideae, tribes Isachneae and Eriachneae. The sampling includes as outgroups 18 genera of the PACMAD clade (excluding Panicoideae) and four genera from the BEP clade (Bambusoideae, Ehrhartoideae, Pooideae), rooting with Bromus inermis. A matrix with 265 taxa based on the combined evidence from ndhF plastid sequences (2074 bp) and 57 morphological characters was subjected to parsimony analyses. Jackknife resampling was used to calculate group support. Most clades are characterized by morphological, cytological, anatomical, and/or physiological characters. Major tribal changes are based on the basic chromosome number; the pantropical x = 9 clade is here recognized as Paniceae s.s., while the American x = 10 Paniceae s.l. is restricted to the reinstated tribe Paspaleae. The optimization of the photosynthetic pathway for the Paspaleae-Andropogoneae-Arundinelleae s.s. clade, including the monotypic Reynaudia, shows a plesiomorphic C4 state while the ancestral state for Paniceae s.s. is ambiguous. If Reynaudia were not included or placed elsewhere, the ancestral photosynthetic pathway for both the Paspaleae-Andropogoneae-Arundinelleae s.s. clade and the Paniceae s.s. would be unambiguously C3 . In order to explore character evolution further, the morphological characters were mapped onto one of the most parsimonious trees. A relationship between photosynthetic pathways and inflorescence morphology is suggested here for the first time. Based on the optimization of morphological characters and additional data, we propose names for almost all inner clades at the rank of subtribe with a few groups as incertae sedis. With this extensive sampling, we resolved the phylogenetic relationships and the assignation of synapomorphies, and improved the support in subtribe sorting; consequently a robust circumscription of the tribe Paniceae s.l. is proposed.

Phylogenetic studies favour the unification of Pennisetum, Cenchrus and Odontelytrum (Poaceae): a combined nuclear, plastid and morphological analysis, and nomenclatural combinations in Cenchrus.

BACKGROUNDS AND AIMS: Twenty-five genera having sterile inflorescence branches were recognized as the bristle clade within the x = 9 Paniceae (Panicoideae). Within the bristle clade, taxonomic circumscription of Cenchrus (20-25 species), Pennisetum (80-140) and the monotypic Odontelytrum is still unclear. Several criteria have been applied to characterize Cenchrus and Pennisetum, but none of these has proved satisfactory as the diagnostic characters, such as fusion of bristles in the inflorescences, show continuous variation. METHODS: A phylogenetic analysis based on morphological, plastid (trnL-F, ndhF) and nuclear (knotted) data is presented for a representative species sampling of the genera. All analyses were conducted under parsimony, using heuristic searches with TBR branch swapping. Branch support was assessed with parsimony jackknifing. KEY RESULTS: Based on plastid and morphological data, Pennisetum, Cenchrus and Odontelytrum were supported as a monophyletic group: the PCO clade. Only one section of Pennisetum (Brevivalvula) was supported as monophyletic. The position of P. lanatum differed among data partitions, although the combined plastid and morphology and nuclear analyses showed this species to be a member of the PCO clade. The basic chromosome number x = 9 was found to be plesiomorphic, and x = 5, 7, 8, 10 and 17 were derived states. The nuclear phylogenetic analysis revealed a reticulate pattern of relationships among Pennisetum and Cenchrus, suggesting that there are at least three different genomes. Because apomixis can be transferred among species through hybridization, its history most likely reflects crossing relationships, rather than multiple independent appearances. CONCLUSIONS: Due to the consistency between the present results and different phylogenetic hypotheses (including morphological, developmental and multilocus approaches), and the high support found for the PCO clade, also including the type species of the three genera, we propose unification of Pennisetum, Cenchrus and Odontelytrum. Species of Pennisetum and Odontelytrum are here transferred into Cenchrus, which has priority. Sixty-six new combinations are made here.

Diversity of boll weevil populations in South America: a phylogeographic approach.

A phylogeographic approach was conducted to assess the geographic structure and genetic variation in populations of the boll weevil Anthonomus grandis, which is the most harmful insect pest of cotton in the Americas. COI and COII mitochondrial gene sequences were analyzed to test a former hypothesis on the origin of the boll weevil in Argentina, Brazil and Paraguay, using samples from Mexico and USA as putative source populations. The analysis of variability suggests that populations from South American cotton fields and nearby disturbed areas form a phylogroup with a central haplotype herein called A, which is the most common and widespread in USA and South America. The population from Texas has the A haplotype as the most frequent and gathers in the same group as the South American populations associated with cotton. The sample from Tecomán (México) shows high values of within-nucleotide divergence, shares no haplotype in common with the South American samples, and forms a phylogroup separated by several mutational steps. The sample from Iguazú National Park (Misiones Province, Argentina) has similar characteristics, with highly divergent haplotypes forming a phylogroup closer to the samples from cotton fields, than to the Mexican group. We propose that in South America there are: populations with characteristics of recent invaders, which would be remnants of "bottlenecks" that occurred after single or multiple colonization events, probably from the United States, and ancient populations associated with native forests, partially isolated by events of historical fragmentation.

Phylogeny of the Pantomorus-Naupactus complex based on morphological and molecular data (Coleoptera: Curculionidae).

The Pantomorus-Naupactus complex is a Neotropical group of broad-nosed weevils (Coleoptera: Curculionidae) including several parthenogenetic species usually assigned to the genera Naupactus Dejean, Pantomorus Schoenherr, Asynonychus Crotch, Aramigus Horn, Eurymetopus Schoenherr and Graphognathus Buchanan. Sixteen species were studied to test hypotheses on the monophyly of these genera, and on the origin of the parthenogenetic lineages. A matrix of 30 morphological characters and 999 positions of the Cytochrome Oxidase I gene, was analyzed with separate partitions and simultaneously, under equal and implied weights, and with different transversion/transitions costs. The ILD test indicates that the incongruence between the molecular and morphological data is not significant. Under equal weights, the molecular data resulted in a single tree and morphology in 34 trees; under implied weights morphology gave a different tree, and under TV:TS ≥ 4:1 molecular and combined analyses resulted in the same optimal tree. According to the latter, Naupactus includes Graphognathus, and is thus paraphyletic and basal regarding remaining genera, Pantomorus is polyphyletic and includes Aramigus and Asynonychus, and Eurymetopus is monophyletic. The species in which apomictic parthenogenesis has been verified (Aramigus tessellatus, Asynonychus cervinus and Graphognathus lecuoloma), belong to different clades of the Pantomorus-Naupactus complex, with basal sexual relatives.