Phylogenetic Relationships of the Wardiaceae (Musci); Evidence from 18s rRNA and rps4 Gene Sequences

Abstract

Wardiaceae are a monotypic family of aquatic mosses endemic to the Western Cape province of South Africa. The family is often placed in the Isobryales, usually in or near the Fontinalaceae, although some classifications place it near Rhacocarpus in the Leucodontales. A number of features, including the fused columella-operculum complex and the sporophyte shape, seem at odds with such a placement. We use a molecular data set comprising nuclear-encoded 18s rRNA and chloroplast-encoded rps4 gene sequences to address this problem. Parsimony analyses provide very strong support for the inclusion of Wardia among the haplolepideae, and in the rps4 and combined analyses, there is some support for a sister group relationship with Blindia (Seligeriaceae). This placement is in agreement with features of gametophyte and sporophyte morphology as well as with habitat. An s.e.m. study of the rudimentary peristome of Wardia indicates that peristome morphology is also consistent with such a placement. Wardia is a monotypic genus (containing only Wardia hygrometrica Harvey & Hooker) endemic to the Western Cape Province of South Africa (Hooker 1837). This genus was placed in the Fontinalaceae by Jaeger and Sauerbeck (1879) (although in a supplement to the same publication these authors assigned the genus to the Crypheaceae), and most later authors, including Cardot (1892) and Brotherus (1925), accepted this position. However, Welch (1943) drew attention to a number of features showing marked divergence from the Fontinalaceae and removed Wardia from that family, erecting the monotypic Wardiaceae to accommodate it. Subsequently, the family has often been placed in the Isobryales, an order of diplolepideous pleurocarps, usually near the Fontinalaceae (e.g., Vitt 1984) or even submerged within it. Buck and Vitt (1986), drawing attention to the enlarged alar cells and furrowed capsules, suggested an alternative placement within the Leucodontales, near Rhacocarpus. A number of features of Wardia seem at odds with either placement. For example, the upright, pyriform-turbinate capsules with long-rostrate opercula are very reminiscent of those found in members of certain haplolepideous families (e.g., Pottiaceae, Seligeriaceae). Furthermore, the capsules are systylious (i.e., the columella is elongate and fused to the operculum), a feature not known in the diplolepideous groups, but one that does occur sporadically among the haplolepideae. Indeed, Hooker took this trait to indicate relationships to Scouleria, usually treated as a member of the Grimmiales (Churchill 1985; Vitt 1984). More recently, Buck (1995) pointed out that Wardia is probably not a pleurocarp and, on the basis of its systylious capsules and aquatic habitat, suggested a possible relationship to the Seligeriaceae. Much of the higher classification of mosses is based currently on peristome morphology and development (Edwards 1984; Vitt 1984; Vitt et al. 1998). As noted by Dixon (1931), the peristome of Wardia is rudimentary, and hence provides little evidence at maturity (at least as seen by conventional microscopy) for systematics. As far as we are aware, no developmental studies are available that might shed light on the nature of the peristome in this genus. As part of our ongoing studies of the molecular systematics of bryophytes (Cox & Hedderson 1999; Hedderson et al. 1996), we have sequenced the nuclear-encoded small subunit rRNA gene and the chloroplast-encoded rps4 gene for representatives of most of the major groups of mosses. In this paper we use a subset of these data to examine the phylogenetic relationships of Wardia. We also present an s.e.m. study of its reduced peristome. MATERIALS AND METHODS Tax n sampling.-The 13 taxa used in this study are listed in Table I with GENBANK accession numbers for the sequences derived for each species. These taxa represent all the major groups of haplolepideous mosses currently available to us (including Blindia of the Seligeriaceae, and Scouleria), as well as the Fontinalaceae, the Hypnobryales, and several major families of the Bryales. Funaria was included as an outgroup (Hedderson et al. 1996). DNA extraction, sequencing, and alignment.-DNA was extracted and target genes amplified and sequenced 0007-2745/99/26-31$0.75/0 This content downloaded from 157.55.39.189 on Fri, 09 Dec 2016 05:30:54 UTC All use subject to http://about.jstor.org/terms 1999] HEDDERSON ET AL.: PHYLOGENETIC RELATIONS OF WARDIACEAE 27 TABLE 1. Taxa used in analyses with Genbank Accession Numbers. Voucher information is also provided for the Wardia specimen (deposited at RNG) used to generate molecular data reported here.