Molecular phylogeny of Persicaria (Persicarieae, Polygonaceae)

Relationships within the Polygonaceae have been recently examined using rbcL sequences, with an emphasis on Polygonum and its segregates. Here we test these results with respect to Polygonum (sensu lato) with an expanded dataset, including additional species and gene regions. Specifically, we focus on inferring the relationships of Eupersicaria (Polygonum sect. Persicaria in many prior treatments), using the chloroplast genes rbcL, trnL–F, trnK intron–matK, and psbA–trnH IGS, and nuclear ribosomal ITS sequences. We conclude that Eupersicaria is monophyletic and most closely related to Tovara and Echinocaulon. In turn, this clade is most closely related to Cephalophilon. The sister group of this entire Persicaria clade contains Bistorta and a clade including Aconogonon and Koenigia, which supports the monophyly of the Persicarieae. Within Eupersicaria there appears to be a deep split between P. amphibia and the remaining species, and there is strong conflict regarding the placement of P. punctata. These results set the stage for a more detailed phylogenetic analysis of Eupersicaria. Keywords—chloroplast DNA, Eupersicaria, ITS, Polygonum, Templeton test. Polygonum, in the broad Linnaean sense, has presented a great taxonomic challenge. A wide variety of classification systems have been proposed over the years and the same names have been applied to quite different groups and at different taxonomic ranks (Haraldson 1978). Communication has become difficult to the point that most botanists are content to sidestep the underlying issues by reference to Polygonum sensu lato. In attempting to sort out phylogenetic relationships in this group, we have adopted a set of names and circumscriptions that we hope will cause the least confusion. In making these choices we have been guided principally by the desire to conform to standard usage as much as possible, to have names for the smallest (least inclusive) taxa that have been recognized, and to have a unique name for each taxon under discussion. Among modern systems, the classification of Haraldson (1978) comes the closest to achieving these objectives. Along with most modern workers, Haraldson recognized a genus Persicaria, comprised of some 150 species of annual or perennial herbs or vines distributed around the northern temperate and into tropical regions (Brandbyge 1993). These plants are characterized by many-flowered spicate or capitate panicles (Stanford 1925a); usually entire, ciliate or pectinate ochrea; 4–5 tepals with trifid venation; and 4–8 stamens (Haraldson 1978; Ronse Decraene and Akeroyd 1988). Within Persicaria Haraldson (1978) recognized four sections: Cephalophilon (ca. 16 species), Echinocaulon (ca. 21 species), Persicaria (ca. 60 species), and Tovara (ca. 3 species). She considered Aconogonon (ca. 25 species), Bistorta (ca. 50 species), and Koenigia (ca. 9 species) to be closely related genera in her Persicarieae. Persicarieae is distinguished from the other major tribe Polygoneae (including Polygonum sensu stricto, Polygonella, Oxygonum, Fallopia, Pteropyrum, Atraphaxis, and Calligonum) by having nondilated filamentous stamens with the number reduced mostly in the inner whorl, the presence of nectaries, trifid tepal venation, and mostly rectangular to elongated tepal epidermis cells (Ronse Decraene and Akeroyd 1988; Ronse Decraene et al. 2000). There is a difference of opinion regarding the placement of Fagopyrum (ca. 16 species). Haraldson (1978) excluded it from her Persicarieae on the basis of petiole anatomy and life-form while Ronse Decraene and Akeroyd (1988) included it in Persicarieae based on floral characters. Although most of these taxa have been recognized since the major treatments by Meisner (1826, 1856), there has been great disparity in the taxonomic ranks assigned to them (Haraldson 1978). For our present purposes these rank assignments are irrelevant, and we will use these names only as designations for putative clades. Thus, for example, we will simply refer to Tovara or to Bistorta, rather than to “section Tovara” or to “genus Bistorta”. This approach is unproblematic except in the case of the name Persicaria itself, since this name has been applied both to a genus-level taxon and to a section-level taxon. We will use Persicaria for the more inclusive, genus-level taxon, in keeping with standard usage. To avoid any confusion, for purposes of this paper we have adopted the name Eupersicaria for the less inclusive, sectionlevel taxon. The choice of the name Eupersicaria is based on Gross (1913a, b), who used it for what most authors have referred to as section Persicaria (although exact circumscriptions have varied somewhat from worker to worker). With these names in hand, the purpose of our paper becomes easy to describe. Our aim is to test the monophyly of each of the aforementioned taxa and to infer their relationships to one another using DNA sequence data. Our primary focus is on relationships within Persicaria, and on the identification of its closest relatives. Most especially we are interested in assessing the monophyly and relationships of Eupersicaria as a prelude to more detailed phylogenetic and evolutionary studies within this group. The data used here are from the chloroplast genes rbcL, trnL–F, partial matK with trnK intron, and psbA–trnH IGS, and from the nuclear ribosomal ITS region. Our analyses build upon and test the results of a study of Polygonaceae by Lamb Fyre and Kron (2003) based on rbcL sequences alone. MATERIALS AND METHODS Taxon Sampling—Appendix 1 provides voucher specimen information on the 26 accessions we used to represent 24 species: eight accessions represent six species of Eupersicaria (two accessions each were included of P. amphibia and P. hydropiper); four species of Echinocaulon were included, three of Cephalophilon, two each of Tovara and Bistorta, and one of Aconogonon and each of the remaining six groups. A Rumex species was included for rooting purposes. This sample was chosen to represent all of the major clades within Persicaria sensu Haraldson (1978) and its presumed close relatives; it also largely corresponds to the sample in Lamb Frye and Kron (2003). Most of our samples were collected in the field by STK between 2002–2004 in North America, Crete, China, and Korea, but Systematic Botany (2008), 33(1): pp. 77–86 © Copyright 2008 by the American Society of Plant Taxonomists