Abstract Detail

Systematics Section

Rothfels, Carl [1], Larsson, Anders [2], Kuo, Li-Yaung [3], Korall, Petra [4], Chiou, Wen-Liang [5], Pryer, Kathleen [6].

Ancient, rapid, yet resolvable: The backbone phylogeny of the eupolypod II radiation.

The eupolypod II clade comprises nearly a third of extant fern species, including such well-known groups as the spleenworts (Asplenium), the ladyferns (Athyrium), and the marshferns (Thelypteris). While this clade is frequently recovered with strong support in phylogenetic analyses, the relationships within it have resisted elucidation by both molecular and morphological data, and classifications have correspondingly varied wildly. Earlier results suggest that the perennial difficulty in resolving the backbone relationships in the Eupolypods II is due to three factors: 1) the ancient rapid radiation of this group (the phylogeny is characterized by long ingroup branches among a series of very short backbone internodes); 2) the absence of a closely related outgroup; and 3) the presence of strongly heterogeneous lineage-specific rates of evolution within the ingroup. We assembled a five-locus dataset (atpA, atpB, matK,rbcL, trnG-R) from 81 accessions selected to capture the deepest divergences in Eupolypods II, and evaluated the resulting phylogenetic hypothesis against potential artifacts including those induced by the ancient rapid radiation model, highly divergent outgroups, lineage-specific rate heterogeneity, and the failure of some implementations of Bayesian inference to account for the possibility of polytomies in the option set (the "star-tree paradox" artifact). The backbone internodes (seven out of eight of which were highly supported under both maximum likelihood and Bayesian inference) proved robust to our evaluations, which nonetheless revealed unexpected behavior of our methods on these data, including model-mediated impacts of outgroup composition and strongly divergent effects of two methods intended to avoid the Bayesian star-tree paradox artifact. This study is one of the few investigations of these methods with empirical data, and provides a template for those wishing to resolve ancient rapid radiations (or other challenging phylogenetic problems) with moderate amounts of data. Among the significant phylogenetic results are the strongly supported eupolypod II backbone, the demonstrated paraphyly of Woodsiaceae as currently circumscribed, and the well-supported placement of the enigmatic genera Homalosorus, Diplaziopsis and Woodsia.

Broader Impacts:

1 - DUKE UNIVERSITY, Department Of Biology, BOX 90338, DURHAM, NC, 27708, USA
2 - Uppsala University, Systematic Biology, Evolutionary Biology Centre, Norbyv. 18D, Uppsala, SE-752 36, Sweden
3 - National Taiwan University, Institute of Ecology and Evolutionary Biology, No. 1, Sec 4, Roosevelt Road, Taipei, 10617, Taiwan
4 - Uppsala University, PHANEROGAMIC BOTANY, Norbyv. 18D, Uppsala, SE-752 36, Sweden
5 - TAIWAN FORESTRY RESEARCH INST, 53 NAN-HAI RD, TAIPEI, 100, Taiwan
6 - Duke University, Science Drive, Durham, NC, 27708-0338, USA

Keywords:
Bayesian estimation
moderate data
molecular phylogeny
Phycas
phylogeny evaluation
rapid radiation
rate heterogeneity
reduced consensus
star-tree paradox
Woodsiaceae.

Presentation Type: Oral Paper:Papers for Sections
Session: 04
Location: Lindell B/Chase Park Plaza
Date: Monday, July 11th, 2011
Time: 9:00 AM
Number: 04003
Abstract ID:470