Search

Dispersals have been shown to be critical to the evolution of the long isolated but megadiverse flora of Madagascar and the surrounding islands of the western Indian Ocean, but we are just beginning to understand the directionality of these dispersals. With more than half of its species occurring in the western Indian Ocean region (WIOR), the paleotropical subfamily Dombeyoideae provides a particularly useful case study through which to better understand the biogeography of the WIOR, and yet its biogeography is poorly understood. Here we sampled six molecular markers from all 20 genera in the Dombeyoideae to reconstruct the most complete phylogeny to date for the subfamily. From this, divergence times, calibrated with three fossils (two dombeyoid, one malvoid), and ancestral range estimations were hypothesized. Biogeographic stochastic mapping (BSM) analyses on the maximum clade credibility tree were completed and compared to BSM analyses on 1,000 trees randomly sampled from the posterior distribution of trees resulting from the dating analysis. We found the Dombeyoideae crown node diverged ca. 53 million years ago out of a broad ancestral range involving all three major areas of its distribution: Madagascar, Africa, and Asia. The majority of diversification and dispersals in the subfamily occurred within the last ca. 10 million years, mostly from the Pliocene onwards. There were roughly five dispersals from Madagascar to Africa (and only one in reverse), at least six from Madagascar to surrounding islands of the WIOR (Mascarenes and Comoros), and one dispersal from Madagascar to Asia (and ca. 1 in reverse). Other long-distance dispersals included one from Africa to St. Helena and one from Africa to Australasia, both from within the most widespread clade, the Cheirolaena & allies clade, and one dispersal from Asia to Africa. Critically, the Dombeyoideae provide strong evidence for considering the island of Madagascar as a source for the colonization of continents, as well as the surrounding islands of the WIOR. Furthermore, narrow sympatry was a key process in the evolution of the subfamily, particularly in Madagascar and the Mascarenes., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2022. Published by Elsevier Inc.)

With about 100 species distributed worldwide, Hymenophyllum subg. Hymenophyllum is the largest subgenus of filmy ferns. It also displays morphological disparity and extreme chromosome numbers variation, with n=11, 12, 13, 14, 18, 21, 22, 26, 28, and 34. We use DNA sequences from the genes rbcL, part of accD, rps4, and the intergenic spacers rbcL-accD, rps4-trnS, and trnG-trnR to infer relationships within Hymenophyllum, with a focus on this subgenus. In the subgenus Hymenophyllum, two main clades are retrieved together with several minor clades whose placement within the subgenus remains ambiguous. We then investigate the evolution of chromosome numbers in the genus and the subgenus, using a maximum likelihood approach taking into account phylogenetic uncertainty. We provide evidence that Hymenophyllum experienced descending aneuploidy during its evolutionary history, especially within the subgenus Hymenophyllum. Reduction in chromosome numbers is particularly extreme in one clade of the subgenus, with the lowest number reported for homosporous ferns. In addition, this group of species displays a high instability in its chromosome number. Both the reduction and the instability in chromosome number may coincide with the distribution of these species in either temperate areas or at high elevations., ((c) 2010 Elsevier Inc. All rights reserved.)

We performed a phylogenetic analysis of the fern genus Elaphoglossum using two non-coding chloroplast spacers: trnL-trnF and rps4-trnS. The sampling includes 123 species, of which 80 have not been previously sequenced, and for the first time includes species from Africa and the Indian Ocean area. The results of this expanded study largely agree with an earlier molecular study based on a smaller group of neotropical species and with the morphology-based classification of Mickel and Atehortua. We found, however, that some infrageneric groups such as section Elaphoglossum are not monophyletic. Besides section Elaphoglossum pro parte, we recognize six sections: two new monospecific, unnamed sections, and the previously established sections Lepidoglossa, Squamipedia, Amygdalifolia, and "Subulate-scaled clade." We divide the subulate-scaled clade into subsection Setosa (hydathodes present) and Polytrichia (hydathodes absent), and section Elaphoglossum is divided into subsections Platyglossa and Pachyglossa, two groups that do not appear to be supported by any single morphological character. In general, however, the main clades are supported by morphology. Finally, we discuss the species of the Indian Ocean region and their affinities with the neotropical ones. Out of the 11 species pairs postulated by Moran and Smith on the basis of morphology, two are well supported (E. eximium-E. aubertii; E. piloselloides-E. spatulatum) and three are not supported (E. ciliatum-E. humbertii; E. muscosum-E. poolii; E. paleaceum-E. deckenii), and two remain unresolved (E. erinaceum-E. hybridum; E. glabellum-E. acrostichoides) because our molecular markers were not variable enough. Four species pairs could not be tested because specimens were lacking. Unsupported species pairs are best interpreted as morphological convergences. Two additional species pairs are proposed: E. cuspidatum-E. succisaefolium; E. doanense-E. hornei. Placement of the species from the Indian Ocean suggests that at least 13 long-distance dispersal events occurred between the Neotropics and the Indian Ocean-Africa.

We investigated the phylogenetic relationships of the filmy fern genus Hymenophyllum s.l. using the rbcL and rps4 genes and the intergenic spacer rps4-trnS. Because of variation in length of the noncoding marker, we tested and compared three methods for integrating indels. They proved to be useful for estimating a phylogeny of the genus. The rps4-trnS marker, with coded indels integrated, produced better resolution than analysis of either rps4 or rbcL, and combining the three data sets allowed us to obtain a well resolved and strongly supported topology. We interpret our data as showing support for the classical bigeneric system for the family, and call into question several classifications proposed in the past century. The segregate genera Cardiomanes, Hymenoglossum, Serpyllopsis, and Rosenstockia are embedded within Hymenophyllum s.l. Although the deepest relationships within the genus remain uncertain, two subgenera described by Morton do have some support: (1) Sphaerocionium, in which the problematic section Microtrichomanes is embedded; and (2) a diverse Hymenophyllum, including species that were placed originally in Serpyllopsis, Rosenstockia, Hemicyatheon, and Craspedophyllum by Copeland. Subgenus Mecodium appears to be polyphyletic; nevertheless, a subgroup within Mecodium is strongly supported. Several unexpected associations gain support from cytological data and certain morphological characters not previously used to distinguish species groups within Hymenophyllum s.l.

The interaction between three independent data sets (anatomy/morphology, cytology, molecules) has been evaluated within the controversial genus Trichomanes (Hymenophyllaceae). Anatomy/morphology, cytology, and rbcL sequences, despite their high and significant level of incongruence, were thus empirically combined with differential weighting in a cladistic analysis within Trichomanes in order to give an appreciation of the contribution of each data set in the resulting topologies and to study more precisely the nature of potential conflicts. Results show that any standard statistics values (such as bootstrap) do not appear to be objectively useful for the choice of the "best" topology or the "good" clades provided by the combination. This weighting approach reveals three cases: (i) some clades (such as subgenus Didymoglossum) are always retrieved and correspond to the absence of conflicts between the different data, (ii) some new clades (such as subgenus Achomanes) are either provided or reenforced as a "synergetic" result of the combination of the data and (iii) that remaining conflicting clades reflect the persistence of incongruence between data whatever the weighting., (Copyright 1998 Academic Press.)