ABSTRACT
Ameroglossum is a rare plant genus endemic to northeastern of Brazil, initially monospecific (A. pernambucense) and recently expanded by the description of eight new species and two related genera. The genus was initially placed in the family Scrophulariaceae, but this has never been phylogenetically tested. This group is ecologically restricted to rocky inselberg habitats that function as island-like systems (ILS) with spatial fragmentation, limited area, environmental heterogeneity, temporal isolation and low connectivity. Here we use a phylogenetic perspective to test the hypothesis that Ameroglossum diversification was related to island-like radiation in inselbergs. Our results support that Ameroglossum is monophyletic only with the inclusion of Catimbaua and Isabelcristinia (named here as Ameroglossum sensu lato) and this group was well-supported in the family Linderniaceae. Biogeographic analyses suggest that the ancestral of Ameroglossum and related genus arrived in South America c.a. 15 million years ago by long-distance dispersal, given the ancestral distribution of Linderniaceae in Africa. In rocky outcrop habitats, Ameroglossum s.l. developed floral morphological specialization associated with pollinating hummingbirds, compatible with an island-like model. However, no increase in speciation rate was detected, which may be related to high extinction rates and/or slow diversification rate in this ecologically restrictive environment. Altogether, in Ameroglossum key innovations involving flowers seem to have offered opportunities for evolution of greater phenotypic diversity and occupation of new niches in rocky outcrop environments.
Subject(s)
Ecosystem , Lamiales , Phylogeny , Flowers/genetics , BrazilABSTRACT
PREMISE: The Lower Cretaceous Crato Konservat-Lagerstätte (CKL) preserves a rich flora that includes early angiosperms from northern Gondwana. From this area, the recently described fossil genus Santaniella was interpreted as a ranunculid (presumably Ranunculaceae). However, based on our examination of an additional specimen and a new phylogenetic analysis, we offer an alternative interpretation. METHODS: The new fossil was collected from an active quarry for paving stones in the state of Ceará, northeastern Brazil. We assessed support for alternative phylogenetic hypotheses using a combined analysis of morphological data and DNA sequence data using Bayesian inference. We used a consensus network to visualize the posterior distribution of trees, and we used RoguePlot to illustrate the support for alternative positions on a scaffold tree. RESULTS: The new material includes a flower-like structure not present in the original material and also includes follicles preserved at early stages of development. The flower-like structure is a compact terminal cluster of elliptical sterile laminar organs surrounding internal filamentous structures that occur on flexuous axes. Phylogenetic analyses did not support the fossil placement among eudicots. Instead, Santaniella appears to belong in the magnoliid clade. CONCLUSIONS: The presence of seeds in a marginal-linear placentation and enclosed in a follicle supports the fossil as an angiosperm. However, even though most characters are clearly recognizable, its combination of characters does not provide strong support for a close relationship to any extant order of flowering plants. Its position in the magnoliid clade is intriguing and, based on plicate carpels, it is definitely a mesangiosperm.