Plastid atpA data provide improved support for deep relationships among ferns

Over the past decade, as a consequence of the application of DNA sequence data and phylogenetic approaches to systematic studies of ferns, unprecedented progress has been made toward a full understanding of the fern tree of life. Analyses of single-gene (Hasebe & al., 1993, 1994, 1995; Manhart, 1995; Pryer & al., 1995; Kranz & Huss, 1996; Wolf, 1997; Vangerow & al., 1999; Wolf & al., 1999; Gastony & Johnson, 2001) and subsequently multiple-gene (Wolf, 1996; Wolf & al., 1998; Pryer & al., 2001a, 2004; Schneider & al., 2004; Wikström & Pryer, 2005) datasets have helped to answer many long-standing questions in fern systematics (see Smith, 1995, for an overview of morphology-based hypotheses and a list of then-unanswered questions), and have greatly clarified our understanding of higher-level fern relationships (see Pryer & al., 2004, for a current synopsis). Nonetheless, the molecular datasets assembled and analyzed to date have not been sufficient to definitively resolve all parts of the global fern phylogeny; additional data and more extensive sampling are necessary. As existing molecular datasets are expanded, it is essential to identify and sequence the most useful molecular markers. In ferns (and other plants), the nuclear, mitochondrial, and plastid genomes—and their inherent multitude of coding and non-coding regions—evolve at different rates (Manhart, 1995; Kranz & Huss, 1996; Vangerow & al., 1999; Soltis & al., 2002; Small & al., 2005; Wikström & Pryer, 2005). Because of this rate heterogeneity, not all regions are suitable for simultaneously reconstructing deep and finer-scale fern relationships; appropriate markers must evolve fast enough to provide substantial phylogenetic signal, but slow enough to allow for an accurate assessment of homology. If the ultimate goal is to reconstruct a comprehensive fern phylogeny from an analysis of DNA sequence data, then there is little sense in the extensive sequencing of markers that are essentially invariable or simply unalignable across ferns. In this study, we explore the phylogenetic potential of the plastid atpA gene (coding for the alpha subunit of ATP synthase). We evaluate the utility of this marker relative to four previously sequenced markers (three plastid, one nuclear), and provide a revised phylogenetic hypothesis for ferns based on the analysis of a combined 5-gene (>6 kb) dataset.