Phylogenetic relationships and biogeography of Fuchsia (Onagraceae) based on noncoding nuclear and chloroplast DNA data

Abstract

To examine relationships and test previous sectional delimitations within Fuchsia, this study used parsimony and maximum likelihood analyses with nuclear ITS and chloroplast trnL-F and rpl16 sequence data for 37 taxa representing all sections of Fuchsia and four outgroup taxa. Results support previous sectional delimitations, except for F. verrucosa, which is related to a Central American clade rather than to section Fuchsia and is described here as a new section Verrucosa. The basal relationships within Fuchsia are poorly resolved, suggesting an initial rapid diversification of the genus. Among the species sampled, there is strong support for a single South Pacific lineage, a southern South American/southern Brazilian lineage, a tropical Andean lineage, and one or two Central American and Mexican lineages. There is no clear support for an austral origin of the genus, as previously proposed, which is more consistent with Fuchsia's sister group relationship with the boreal Circaea. An ultrametric molecular clock analysis (all minimal dates) places the split between Fuchsia and Circaea at 41 million years ago (mya), with the diversification of the modern-day lineages of Fuchsia beginning at 31 mya. The South Pacific Fuchsia lineage branches off around 30 mya, consistent with fossil records from Australia and New Zealand. The large Andean section Fuchsia began to diversify around 22 mya, preceded by the divergence of the Caribbean F. triphylla at 25 mya. The Brazilian members of section Quelusia separated from the southern Andean F. magellanica around 13 mya, and the ancestor of the Tahitian F. cyrtandroides split off from the New Zealand species of section Skinnera approximately 8 mya.

Molecular phylogeny, divergence time estimates, and historical biogeography of Circaea (Onagraceae) in the Northern Hemisphere

“…The three fossils of Myrtaceae, Melastomataceae, and Combretaceae produced similar estimates for the Circaea-Fuchsia split (42.7, 41.2 and 40.7 mya, respectively) (Berry et al, 2004). Because these times were consistent with the earliest fossil records of Circaea and Fuchsia (pollen from the Oligocene in Australia, Berry et al, 1990), Berry et al (2004) used them to estimate divergence times of major clades of Fuchsia. In this paper, we followed Berry et al (2004) and constrained the Circaea-Fuchsia node with the age of 41.5 mya (the average age from different fossil constraints).…”

“…Therefore we combined the two data sets to explore whether resolution and support would be improved by increasing the amount of sequencing data. This approach was performed by direct inspection of nodes with bootstrap scores above 70% in the separate analyses (Kluge, 1989;Nixon and Carpenter, 1996;Muellner et al, 2003;Acevedo-Rosas et al, 2004;Berry et al, 2004;Inda et al, 2008). Phylogenetic analyses were undertaken for the nrITS, for the concatenated plastid data set, and for all molecular markers combined using maximum parsimony (MP) and Bayesian inference (BI) methods (Rannala and Yang, 1996).…”

“…Pollen fossils are ample for Fuchsia, Ludwigia and their relatives in the Tertiary (Martin, 2003), suggesting an early diversification of these genera. To estimate the divergence of Fuchsia, Berry et al (2004) used the Circaea-Fuchsia split estimated by Sytsma et al (2004), which was based on calibrated rbcL and ndhF sequence data for the Myrtales using three fossil constraints. The three fossils of Myrtaceae, Melastomataceae, and Combretaceae produced similar estimates for the Circaea-Fuchsia split (42.7, 41.2 and 40.7 mya, respectively) (Berry et al, 2004).…”

“…To estimate the divergence of Fuchsia, Berry et al (2004) used the Circaea-Fuchsia split estimated by Sytsma et al (2004), which was based on calibrated rbcL and ndhF sequence data for the Myrtales using three fossil constraints. The three fossils of Myrtaceae, Melastomataceae, and Combretaceae produced similar estimates for the Circaea-Fuchsia split (42.7, 41.2 and 40.7 mya, respectively) (Berry et al, 2004). Because these times were consistent with the earliest fossil records of Circaea and Fuchsia (pollen from the Oligocene in Australia, Berry et al, 1990), Berry et al (2004) used them to estimate divergence times of major clades of Fuchsia.…”

“…In contrast to the herbaceous, circumboreal Circaea, the 107 species of Fuchsia are shrubs, trees, lianas, or epiphytes distributed primarily in Central and South America, especially in the tropical Andes, with two species in New Zealand and one on the Pacific island of Tahiti (Berry et al, 2004;Wagner et al, 2007). Despite the lack of unambiguous morphological synapomorphies, molecular data support the combination of Circaea and Fuchsia into a single tribe Circaeeae (Wagner et al, 2007).…”