Cephalantholejeunea temnanthoides (R.M.Schust.) R.M.Schust., the only known member of the genus Cephalantholejeunea R.M.Schust. (Lejeuneaceae, Hepaticae), is a rare Amazonian rheophyte with entire underleaves and unique, bisexual branches consisting of a terminal androecium with 2–6 female branches at its base, the gynoecial branches with only one gynoecium and lacking innovations. Sporophytes described here for the first time indicate that the genus is a member of the subfamily Ptychanthoideae, tribe Ptychantheae, not of the subfamily Lejeuneoideae as was uniformly believed.

Communicating Editor: Alan Whittemore

A combination of macro- and micromorphological, anatomical, and phytogeographical data supports the recognition of three species in the Carex backii complex (within sect. Phyllostachyae). Carex backii is widely distributed throughout temperate and boreal regions of North America and is easily recognized by its long, smooth surfaced perigynia (4.8–6.6 mm) with long beaks (1.9–2.9 mm). The leaves are dull green to yellow-green with green, crenate margins. Carex saximontana is confined to the Great Plains region of North America and is distinguished by its shorter, papillose perigynia (3.2–4.9 mm) with much shorter beaks (0.6–1.2 mm). In addition, the leaves are greenish-glaucous to dark herbage-green with distinctive hyaline margins. Carex cordillerana occurs from central British Columbia to central Utah and is described as a widespread new species. Carex cordillerana can be recognized by its dull green to yellow-green leaves with green margins that are densely covered in papillae. Perigynia are intermediate in size between C. saximontana and C. backii (3.9–5.4 mm), but have a unique epidermal cell morphology. Perigynium beaks are similar in length to those found in C. saximontana (0.5–1.6 mm). The North American Carex flora has proven to be a rich source of taxonomic novelties over the past 20 years. This is nowhere better exemplified than in Carex sect. Phyllostachyae.

Communicating Editor: Jeff H. Rettig

The Bromeliaceae is characterized by a 2n chromosome number of 50, but exceptions have been reported. These include scattered polyploids and presumed aneuploid reductions in isolated species from subfamilies Tillandsioideae and Bromelioideae, and 2n = 34 or 36 for all species of Cryptanthus (Bromelioideae) thus far examined. Two hypotheses have been proposed that address the origin of the low chromosome number (2n = 34 or 36, compared to 50) in Cryptanthus. We have attempted to test these hypotheses using quantification of nuclear DNA by flow cytometry. DNA quantifications for examined Cryptanthus species fell into two groups. In one there was no significant difference between Cryptanthus and selected bromeliad species known to have 2n = 50. In the second, Cryptanthus species had approximately twice as much 2C nuclear DNA. Results partially support the hypothesis that the Cryptanthus chromosome number (2n = 34) originated by descending aneuploidy, a derived condition which, with other apomorphies, supports the hypothesis that Cryptanthus is derived within the subfamily Bromelioideae. First chromosome number reports are presented for two species of Cryptanthus and one species of Orthophytum.

Communicating Editor: Alan Whittemore

During preparations for a revision of the genus Ribes L. for the entire neotropical region, three taxa in Ecuador, Colombia, and Costa Rica were recognized as new and undescribed species. These are formally named here. Ribes costaricensis sp. nov. from Costa Rica has been traditionally misidentified as R. leptostachyum Benth., from which it differs in angular (not round) ovaries and glabrous (not setose) stems. Ribes luteynii sp. nov. has been confused with R. hirtum Willd. in the past, from which it differs in leaf size and outline and details of the indument. Ribes caldasiensis sp. nov., from Colombia, is also a segregate of the widespread R. hirtum and differs in having a flaring hypanthium (vs. cylindrical) and lax (vs. dense racemes) and also in details of the indument. All three species are endemic to narrow regions from which their respective closest relatives are absent.

Communicating Editor: Thomas G. Lammers

Australian species of Uromyrtus are revised and separated on morphological differences under the general lineage species concept using the criterion of character fixation. Two new species are described. The first, Uromyrtus lamingtonensis, is endemic to southeastern Queensland and its mountainous southeastern boundary with New South Wales. It is distinguished from congeneric Australian species by dense oil glands on both surfaces of the relatively short, acute leaves. The second, Uromyrtus tenellus, occurs in northeastern Queensland. It differs from other Australian members of the genus by the combination of rounded leaf bases, cuspidate to apiculate leaf apices, absence of prominent oil glands on either laminar surface, and relatively long petals. Detailed descriptions, illustrations, and distribution maps are presented for the four Australian taxa. Keys are provided to separate the Australian species of Uromyrtus and to separate Uromyrtus from the New World genus Ugni, which it resembles in several aspects of floral morphology. Transfers are made from Myrtus for two New Caledonian species, Uromyrtus baumannii and Uromyrtus sunshinensis.

Communicating Editor: Matt Lavin

Members of Solanum series Conicibaccata in Mexico and Central America are very similar. All are tetraploids (2n = 4x = 48). Recent authors have recognized three or four species: S. agrimonifolium, S. woodsonii, S. longiconicum (sometimes included in the next), and S. oxycarpum. We had difficulty distinguishing these species in the herbarium, and needed to resolve species boundaries for ongoing floristic studies. We studied this group in the field throughout Mexico, Guatemala, Costa Rica, and Panama, grew collections in the greenhouse, studied herbarium specimens, determined ploidy levels through flow cytometry, and generated molecular data using Random Amplified Polymorphic DNA. Molecular data distinguish S. agrimonifolium, S. longiconicum, and S. oxycarpum. Solanum woodsonii was not available for molecular analysis. All four species can be distinguished morphologically, but only by leaf character states that overlap in range, by pubescence differences that are best observed in living specimens, and by a seed spot character that is only evident on living or recently gathered specimens.

Communicating Editor: Jeff H. Rettig

The current circumscription of Elymus based on cytogenetic analyses includes all allopolyploid Triticeae species containing the St (Pseudoroegneria) genome. In North American Elymus, the St genome is combined with H (from Hordeum) in an allotetraploid (StStHH) configuration. The goal of this study is to determine whether molecular phylogenetic analyses support existing cytogenetic data with regard to the evolutionary origin of North American Elymus. Analyses were performed using sequences from the nuclear starch synthase gene, and include multiple species of Elymus, Pseudoroegneria, and Hordeum, along with representatives of most of the other monogenomic genera in the Triticeae. The results support the hypothesis that Pseudoroegneria and Hordeum are the genome donors to the North American Elymus tetraploids. One species currently placed in Elymus (an octoploid, Elymus californicus) appears to be unrelated to the rest. The close relationships among Elymus H-genome sequences, with one exception, are consistent with a single origin of the group. The St-genome group consists of two well-defined clades, but support for the monophyly of the entire St group is weak. There are shortcomings associated with the current dependence on genome pairing data for grouping and ranking in the Triticeae, but molecular phylogenetic data suggest that, in many cases, groups delimited by cytogenetic data do in fact correspond to evolutionary lineages.

Communicating Editor: John V. Freudenstein

Hybridization has long been considered a source of taxonomic complexity in Rubus. This study uses molecular data to examine natural hybridization between R. caesius, a facultatively agamospermous tetraploid of subgenus Rubus (blackberries) and R. idaeus, a sexual diploid of subgenus Idaeobatus (raspberries). Both species are widespread in Europe, and hybridization between them is postulated to have given rise to numerous species, including R. maximiformis and R. picticaulis. We sequenced the nuclear ribosomal DNA internal transcribed spacer (ITS) region of R. caesius, R. idaeus, five putative hybrids from Sweden, and one individual each of R. maximiformis and R. picticaulis from Germany. Rubus caesius differs from R. idaeus at 20 nucleotide sites and one insertion/deletion event. The putative hybrids exhibit nucleotide polymorphism additivity at all or most of these sites indicating that gene flow does occur naturally between R. caesius and R. idaeus. For each of five putative hybrids we analyzed two to 21 ITS region clones by sequencing and using two restriction enzymes that distinguish the parental species at three sites. Individual 729 has only intact parental ITS sequences and may therefore be an F₁, while the other four show chimeric (mixtures of parental) ITS sequences, suggesting that they may be later-generation hybrids.

Communicating Editor: James R. Manhart

Pelagodoxa (Arecaceae, Arecoideae, Areceae) is a Pacific genus of uncertain geographic origin. Two species have been described. The type species, Pelagodoxa henryana, from the Marquesas Islands, has large fruits, 8.5 cm in diameter, while a species with smaller fruits (6 cm in diameter) was described as P. mesocarpa and later considered as a synonym of P. henryana. The smaller-fruited entity is known from the Melanesian archipelagos of the southwest Pacific, but the type of P. mesocarpa is of doubtful origin. No truly wild population of any Pelagodoxa is known either in the Marquesas or in the southwest Pacific. The structure and histology of the pericarp and seed of the two fruit morphs were investigated and proved to be similar, except for a color change after ripening in the mesocarp of the smaller fruit morph. Distinctive anatomical features of the fruit of Pelagodoxa include an outer layer of cork tissue, an expanded soft parenchymatous mesocarp lacking raphides and sclereids, numerous vascular bundles with thick fibrous sheaths running radially through the pericarp, conspicuous stegmata on the surface of bundle fibers, an endocarp consisting of a thickened locular epidermis, with conspicuous crystalline structures located at the distal end of each radially elongated cell, and an additional layer of sclerenchyma derived from the seed coat. These features suggest a possible relationship with Sommieria within the subtribe Iguanurinae, in which Pelagodoxa is currently included, but they are also reminiscent of the pericarp of Orania (Oraniinae). A distinct operculum on the endocarp, which characterizes most genera of Iguanurinae, is lacking in Pelagodoxa.

Communicating Editor: Kathleen A. Kron

A phylogenetic parsimony analysis based on 76 morphological characters of 43 Mallotus, 3 Macaranga, 1 Claoxylon, 1 Cleidion, 1 Sampantaea, and 1 Wetria species, resulted in 314 trees of length 602 (CI = 0.600, RI = 0.685), which could be summarized in a well resolved consensus cladogram. This consensus cladogram indicates that the genus Mallotus is possibly polyphyletic and that sections Hancea and Oliganthae should probably be excluded from Mallotus. It is proposed to refer to Mallotus as either Mallotus sensu lato (with the sections Hancea and Oliganthae) or Mallotus sensu stricto (excluding the sections Hancea and Oligantae). The genus Macaranga forms a monophyletic group within Mallotus s.s. The traditional section delimitations within Mallotus s.s. do not circumscribe monophyletic clades of species with the exception of section Polyadenii. It is suggested that sections Stylanthus, Rottlera, Mallotus, and the genus Macaranga should form one monophyletic clade. Sections Axenfeldia and Rottleropsis remain unresolved, but are probably closely related. The position of Macaranga within Mallotus suggests that Macaranga evolved from pioneer ancestors that lived in open, scrub-like vegetation. It is hypothesised that Macaranga originated relatively recently in South East Asia during the Oligocene or early Miocene (between 38 and 15 million years ago).

Communicating Editor: Paul Wilson

Phylogenetic relationships among 48 species of Prunus were investigated by means of maximum parsimony analysis of sequence polymorphism from nuclear ITS and chloroplast trnL-trnF spacer DNA. Several genera that have been implicated as close relatives of Prunus in previous studies were utilized as outgroups. The objective of this study was to reconstruct the phylogeny of Prunus with the purpose of reviewing previously described taxonomic relationships and providing a basis for studies of morphological evolution in the genus. Variability and homoplasy are higher in ITS than in trnL-trnF. Parsimony analysis of both markers, separately and in combination, supported a close relationship among Exochorda, Oemleria, and Prinsepia but did not support a sister relationship of this clade with Prunus. Prunus is monophyletic and divided into two groups. The first is a strongly supported clade composed of subgenera Padus, Laurocerasus, and Cerasus. None of these subgenera is monophyletic and the trait used to characterize Padus and Laurocerasus, flowers arranged in racemes, can be more broadly interpreted as a synapomorphy for the clade. The second group included the subgenera Amygdalus and Prunus, and sections Microcerasus (subgenus Cerasus), and Penarmeniaca, which has never been assigned to a subgenus. The exclusively North American subgenus Emplectocladus may be an early diverging taxon sister to the second group. Sections Prunus and Armeniaca are monophyletic with low bootstrap support. Section Penarmeniaca, from southwestern US, and P. besseyi (section Microcerasus) are sister taxa. A Eurasian origin of Prunus is supported.

Communicating Editor: Richard Jensen

Evolutionary relationships within the wintergreen group (Gaultheria, Pernettya, Diplycosia, and Tepuia) are poorly understood and little studied. This study analyzes molecular data (matK, nrITS, and atpB-rbcL spacer) from 34 representatives of the wintergreen group and 19 outgroups in order to determine the phylogenetic position of the wintergreen group within the “Andromedeae” and examine phylogenetic relationships among taxa currently placed in Gaultheria, Pernettya, and Diplycosia. Results show that the wintergreen group is monophyletic and is sister to a clade that contains Chamadaphne and Eubotrys. Gaultheria is likely paraphyletic, with both Pernettya and Diplycosia derived from within it. Pernettya, as sampled, is not monophyletic, P. tasmanica does not form a clade with other species of Pernettya. Diplycosia is monophyletic and sister to G. cumingiana. Although the matK analysis suggests that Tepuia may be sister to the clade containing Gaultheria, Pernettya, and Diplycosia, bootstrap support for this relationship is low and the position of Tepuia is unresolved in the combined analysis. Increased taxon sampling, especially in South East Asia, and more variable DNA regions are recommended for future phylogenetic studies in the wintergreen group.

Communicating Editor: Matt Lavin