Search

Gilliesieae are a South American tribe of Amaryllidaceae characterized by high floral diversity. Given different taxonomic interpretations and proposals for generic and specific relationships, a representative phylogenetic analysis is required to clarify the systematics of this group. The present study provides a framework for understanding phylogenetic relationships and contributing to the development of an appropriate taxonomic treatment of Gilliesieae. Molecular analyses, based on nuclear (ITS) and plastid DNA sequences (trnL-F and rbcL), resolve with strong support the monophyly of the tribe and the differentiation of two major clades. Clade I comprises the genera Gilliesia, Gethyum and Solaria and Clade II includes Miersia and Speea. These well-supported clades are mostly congruent with vegetative and karyotype characters rather than, e.g., floral symmetry. At the generic level, all molecular analyses reveal the paraphyly of Gilliesia and Miersia. Gethyum was found to be paraphyletic, resulting in the confirmation of Ancrumia as a distinct genus. Several instances of incongruent phylogenetic signals were found among data sets. The calibrated tree suggests a recent diversification of the tribe (Pliocene–Pleistocene), a contemporary process of speciation in which instances of hybridization and incomplete lineage sorting could explain patterns of paraphyly and incongruence of floral morphology.

To accompany the 12 manuscripts on Bromeliaceae in this issue, we here present a summary of evolutionary and phylogenetic studies at different levels in the family. We discuss biogeography, speciation and hybridization in the light of current knowledge.

Post-print del articulo publicado en Botanical Journal of the Linnean Society. Version revisada y corregida

Tulipa (tulips; Liliaceae) is a genus of geophytes comprising c. 76 species, occurring from southwestern Europe and North Africa to Central Asia. The taxonomy and classification of the genus have been contentious in the past. We investigated the phylogenetic relationships in the genus using DNA sequences from five plastid regions (trnL intron and trnL–trnF spacer, rpl16 intron, rps12–rpl20 intergenic spacer and matK) and the internal transcribed spacer (ITS) region of nuclear ribosomal DNA. Amana and Erythronium were used as outgroups. Sequences were obtained from 25 Tulipa taxa representing all major lineages previously identified as distinct and four outgroups (two Amana spp. and two Erythronium spp.). In the combined maximum parsimony analysis, Tulipa was strongly supported as monophyletic and four clearly defined clades in the genus were obtained, although the relationships between them were unclear. In support of previous molecular studies, the results suggest that section Clusianae should be excluded from subgenus Tulipa and accepted at subgeneric rank. Subgenus Eriostemones and subgenus Tulipa (excluding Clusianae) were both strongly supported. Tulipa sprengeri, traditionally placed in subgenus Tulipa, was shown to be a member of Eriostemones. Orithyia, in this study represented by T. uniflora, formed a fourth lineage, also to be treated at subgeneric level. In the Bayesian analysis, the genus Tulipa was strongly supported and the same four lineages (subgenera) were identified. In this case, Orithyia was sister to the rest of the genus (with moderate support) and subgenera Clusianae and Eriostemones together formed a clade with strong support. Original species descriptions and type specimens of as many names as possible were reviewed and, on this basis, a revised checklist with full synonymy, typification and distribution is provided. The status of T. ×gesneriana and its synonyms is discussed. All accepted species are classified into the four subgenera supported by our phylogenetic study. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 280–328.

Many molecular studies have shown the monocot order Liliales to be well supported; morphologically, it is defined by synapomorphies of tepalar nectaries and extrorse anthers, in contrast with septal nectaries and introrse anthers commonly found in other monocots, especially Asparagales, with which it was often confused in the past. It comprises c. 1500 species, 67 genera and 9–11 families. Although monophyly is clear, the phylogenetic relationships among some of the families are still unclear. In this study, we examine the inter- and infrafamilial relationships among Liliales in phylogenetic analyses based on four plastid loci (matK, rbcL, atpB and atpF-H). We performed phylogenetic analyses and constructed maximum parsimony and Bayesian trees for 49 genera and 148 taxa in ten families of Liliales sensu Angiosperm Phylogeny Group (APG) III using the combined DNA data. The monophyly of Liliales, except for Corsiaceae (Arachnitis), was strongly supported by both analyses. Campynemataceae were sister to the rest of the order, excluding Corsiaceae. The other families formed two well-defined clades, (Colchicaceae + Alstroemeriaceae) and (Liliaceae, Smilacaceae, (Rhipogonaceae + Philesiaceae)), and one weakly supported clade with Melanthiaceae and Petermanniaceae. Subfamilial and tribal circumscriptions for the three larger families, Colchicaceae, Melanthiaceae and Liliaceae, agreed well with the results of this study, except for the subfamily Calochortoideae of Liliaceae, which was split into two separate clades of Calochortus and Tricyrtis. In addition, we found several taxa with a 10-bp inversion in matK, which could contribute additional homoplasy to these analyses if included without re-coding. Phylogenetic relationships among families of Liliales were better defined here than in a previous molecular analysis, although the placement of Corsiaceae with plastid data remains problematic. Based on these results, reconsideration of the circumscriptions of Rhipogonaceae + Philesiaceae and the subfamilial circumscription for Calochortoideae of Liliaceae is suggested. © 2013 The Linnean Society of London

Phylogenetic relationships in the genus Paphiopedilum were studied using nuclear ribosomal internal transcribed spacer (ITS) and plastid sequence data. The results confirm that the genus Paphiopedilum is monophyletic, and the division of the genus into three subgenera Parvisepalum, Brachypetalum and Paphiopedilum is well supported. Four sections of subgenus Paphiopedilum (Pardalopetalum, Cochlopetalum, Paphiopedilum and Barbata) are recovered as in a recent infrageneric treatment, with strong support. Section Coryopedilum is also recovered, with low bootstrap but high posterior probability values for support of monophyly. Relationships in section Barbata remain unresolved, and short branch lengths and the narrow geographical distribution of many species in the section suggest that it possibly underwent rapid radiation. Mapping chromosome and genome size data (including some new genome size measurements) onto the phylogenetic framework shows that there is no clear trend in increase in chromosome number in the genus. However, the diploid chromosome number of 2n = 26 in subgenera Parvisepalum and Brachypetalum suggests that this is the ancestral condition, and higher chromosome numbers in sections Cochlopetalum and Barbata suggest that centric fission has possibly occurred in parallel in these sections. The trend for genome size evolution is also unclear, although species in section Barbata have larger genome sizes than those in other sections. © 2012 The Linnean Society of London, Botanical Journal of the Linnean Society, 2012, 170, 176–196.

Many endemic plant species belong to taxonomically complex groups. These endemics have often arisen as a consequence of recent and rapid evolutionary divergence facilitated by processes such as hybridization, polyploidy and/or breeding system transitions. The rapid and dynamic nature of divergence in taxonomically complex groups leads to problems in the implementation of traditional species-based approaches for the conservation of the biodiversity that they contain. Firstly, the taxa of interest can be difficult to define and identify, leading to practical difficulties in implementing conservation measures. Secondly, a species-based approach often fails to capture the complexity of diversity present in the taxonomically complex group. To accommodate these challenges, we have developed a Process-Based Species Action Plan approach. This is designed to conserve the processes leading to the generation of biodiversity, rather than focusing on the preservation of individual named taxa. We illustrate the approach using a group of endemic tree species (Sorbus) on the Scottish island of Arran that have originated via a combination of multiple recent hybridization events and apomixis. The plan focuses on the optimization of habitat management to ensure the reproduction and regeneration of Sorbus in the zone in which these evolutionary processes operate, and to facilitate hybridization that will ensure the continued generation of diversity in this group. © 2011 The Linnean Society of London, Botanical Journal of the Linnean Society, 2012, 168, 194–203.

Morphological and molecular phylogenetic studies show that there is a close relationship between Coffea and Psilanthus. In this study we reassess species relationships based on improved species sampling for Psilanthus, including P. melanocarpus, a species that shares morpho-taxonomic characters of both genera. Analyses are performed using parsimony and Bayesian inference, on sequence data from four plastid regions [trnL–F intron, trnL–F IGS, rpl16 intron and accD–psa1 intergenic spacer (IGS)] and the internal transcribed spacer (ITS) region of nuclear ribosomal DNA (ITS 1/5.8S/ITS 2). Several major lineages with geographical coherence, as identified in previous studies based on smaller and larger data sets, are supported. Our results also confirm previous studies showing that the level of sequence divergence between Coffea and Psilanthus species is negligible, particularly given the much longer branch lengths separating other genera of tribe Coffeeae. There are strong indications that neither Psilanthus nor Coffea is monophyletic. Psilanthus melanocarpus is nested with the Coffea–Psilanthus clade, which means that there is only one critical difference between Coffea and Psilanthus; the former has a long-emergent style and the latter a short, included style. Based on these new data, in addition to other systematically informative evidence from a broad range of studies, and especially morphology, Psilanthus is subsumed into Coffea. This decision increases the number of species in Coffea from 104 to 124, extends the distribution to tropical Asia and Australasia and broadens the morphological characterization of the genus. The implications for understanding the evolutionary history of Coffea are discussed. A group of closely related species is informally named the ‘Coffea liberica alliance’. © 2011 The Linnean Society of London, Botanical Journal of the Linnean Society, 2011, 167, 357–377.

The amplified fragment length polymorphism (AFLP) technique was used to examine genetic variation among old and newly emerged individuals of Hyophorbe lagenicaulis (the Round Island bottle palm) on Round Island to assess surviving levels of diversity in the wild population and to evaluate the suitability of old cultivated stocks on Mauritius as a source of seed for reintroduction. The analysis of AFLP data for 48 individuals of H. lagenicaulis (individuals from Round Island and elsewhere), two H. verschaffeltii, two H. vaughanii, one H. amaricaulis and one H. indica yielded 81 variable and six monomorphic bands. Analysis of molecular variance (AMOVA) and Shannon's indices showed a high level of genetic variation within the wild population on Round Island and a smaller amount of genetic variation among cultivated individuals. A neighbor joining analysis resulted in an unrooted network of genetic distances in which the five Hyophorbe spp. were separated and much variation within H. lagenicaulis was recovered. The Round Island populations of H. lagenicaulis contain representatives of the genetic variation found within the species as a whole. However, a few individuals, both wild and cultivated, represent apparently rare AFLP profiles, and thus, if a more representative distribution of genotypes is wanted for the wild population, cultivated sources could be introduced to Round Island from Mauritian gardens and plantings. © 2011 The Linnean Society of London, Botanical Journal of the Linnean Society, 2011, 167, 301–310.

The Global Strategy for Plant Conservation (GSPC), adopted by the Conference of the Parties to the Convention on Biological Diversity (CBD) in 2002, came to the end of its first phase in 2010. Thus we felt that this was an appropriate theme for a discussion meeting jointly sponsored by the Royal Botanic Gardens, Kew, and the Royal Society of London (in June 2010). Adoption of the GSPC – by the parties to the CBD and implemented through international organizations – as a framework to address the importance of plants in sustaining biodiversity of all organisms is unique; it was also much more ambitious than any other conservation measure ever adopted. It consisted of 16 outcome-oriented targets across five major objectives and tried to describe the targets in ways that could be measured, to one degree or another. Like many international agreements, it was subject to varying interpretations and retained flexibility for implementation at national and local levels, so that interpretations were uniform in neither content nor intent. The GSPC nonetheless stands as a remarkable document and at least a limited success. The problem of bringing together scientists (from botanical gardens, museums and universities), conservation non-governmental organizations (NGOs) and politicians to develop policies that then need to be implemented by national governments was a major challenge. The limitations imposed by this process are reflected in the ambiguous wording of some of the targets in the GSPC. The flexibility of the targets also made the Strategy easier to promote and more adaptable to local conditions, so they were not necessarily a disadvantage. However, if the second phase (2011– 2020) is to succeed, there needs to be improved dialogue between all parties in both planning and implementation. Again, this process is pushing back the barriers that have previously prevented such concerted international activity, and so these too are worth considering further to determine how they can be minimized in the second phase. The successes of the GSPC were partly due to the formation of networks that promoted the Strategy and guided its implementation and the identification of a facilitating organisation that would be responsible for progress towards each target. Little or no progress was often a symptom of lack of engagement between, or availability of, appropriate actors and passive facilitation. Many botanical gardens and conservation NGOs accepted the challenges of the GSPC and played major roles in promoting it internationally and within their own respective spheres of influence, and without this involvement, successes would have been limited. One could argue that a document such as the GSPC really was not necessary for action to occur, but just having a formal, internationally agreed strategy provided a common focus for activity and made it more likely that governments would accept it and put resources into local implementation, so it acted as a nucleating agent, precipitating action when there would have been little or none without it. Having realized the enormous challenges facing the plant conservation community, scientists in the movement worked hard to translate their priorities into targets that governments could understand and believe to be tractable, thus making compromises but increasing the likelihood of broad adoption and implementation. It was an opportunity seized, and a timely evaluation of the GSPC in terms of the general process of combining science and policy could benefit not only phase II of the GSPC, but also the general process with the possible outcome of increased incorporation of science into policy-making. Botanical Journal of the Linnean Society, 2011, 166, 213–216.

Petenaea cordata (from northern Central America) was first described in Elaeocarpaceae and later placed in Tiliaceae, but most authors have been uncertain about its familial affinities. It was considered a taxon incertae sedis in the Angiosperm Phylogeny Group classification (APG III). A recent collection was made in Guatemala, and analysis of both rbcL and atpB in the large Soltis et al. angiosperm matrix, the most completely sampled study published to date, indicated a moderately supported relationship to Tapiscia (Tapisciaceae, the only member of the newly recognized order Huerteales in this matrix; 81% bootstrap support). We then conducted a more restricted analysis using the Bayer et al. rbcL/atpB matrix for Malvales supplemented with the other genera of Huerteales from published studies. Our results indicate a distant, weakly supported sister-group relationship to the African genus Gerrardina (Gerrardinaceae; Huerteales). After comparison of the characters cited in the literature and an examination of herbarium material of both genera, we could find no obvious synapomorphies for Gerrardina and Petenaea or any other relationship of the latter, and we therefore propose the new monogeneric family, Petenaeaceae. The polymorphic order Huerteales now comprises four small families: Dipentodontaceae, Gerrardinaceae, Petenaeaceae and Tapisciaceae. © 2010 The Linnean Society of London, Botanical Journal of the Linnean Society, 2010, 164, 16–25.

We report the largest eukaryotic genome to date in the monocot Paris japonica (Melanthiaceae, 1C = 152.23 pg), measured using flow cytometry. This value is 15% larger than any previous estimate and extends the range of genome sizes to c. 2400-fold across angiosperms and c. 66 000-fold across eukaryotes. © 2010 The Linnean Society of London, Botanical Journal of the Linnean Society, 2010, 164, 10–15.

Plastid microsatellite loci developed for Cephalanthera longifolia were used to examine the level of genetic variation within and between populations of the three widespread Cephalanthera species (C. damasonium, C. longifolia and C. rubra). The most detailed sampling was in C. longifolia (42 localities from Ireland to China; 147 individuals). Eight haplotypes were detected. One was detected in the vast majority of individuals and occurred from Ireland to Iran. Three others were only found in Europe (Ireland to Italy, England to Italy and Austria to Croatia). Two were only found in the Middle East and two only in Asia. In C. damasonium, 21 individuals from 10 populations (England to Turkey) were sampled. Only one haplotype was detected. In C. rubra, 34 individuals from eight populations (England to Turkey) were sampled. Although it was not possible to amplify all loci for all samples of this species, nine haplotypes were detected. Short alleles for the trnS-trnG region found in two populations of C. rubra were characterized by sequencing and were caused by deletions of 26 and 30 base pairs. At this level of sampling, it appears that C. rubra shows the greatest genetic variability. Cephalanthera longifolia, C. rubra and C. damasonium have previously been characterized as outbreeding, outbreeding with facultative vegetative reproduction and inbreeding, respectively. Patterns of genetic variation here are discussed in the light of these reproductive system differences. The primers used in these three species of Cephalanthera were also demonstrated to amplify these loci in another five species (C. austiniae, C. calcarata, C. epipactoides, C. falcata and C. yunnanensis). Although it is sometimes treated as a synonym of C. damasonium, the single sample of C. yunnanensis from China had a markedly different haplotype from that found in C. damasonium. All three loci were successfully amplified in two achlorophyllous, myco-heterotrophic species, C. austinae and C. calcarata. © 2010 The Linnean Society of London, Botanical Journal of the Linnean Society, 2010, 163, 181–193.

Strong geographical isolation within the distribution of a species may result in differentiated lineages exhibiting conspicuous phenotypic differences. In the present paper, we investigate whether plastid and phenotypic variation is geographically structured within the Olea europaea complex in Macaronesia, which comprises three subspecies separated by oceanic barriers: maroccana (south-west Morocco), guanchica (Canary Islands) and cerasiformis (Madeira archipelago). Plastid variation showed a significant pattern of geographical structure (NST > GST = 0.56), because of the lack of shared haplotypes among subspecies and the presence of a single and private haplotype in the eastern Canary Islands. Such a clear molecular structure, however, was not reflected in a congruent pattern of phenotypic differentiation among taxa in leaf morpho-functional traits. Despite the substantial genetic differentiation observed between the subspecies from Madeira and the Canary Islands, they displayed both higher leaf size (leaf area) and specific leaf area (leaf surface area-to-mass ratio) than their continental counterparts, probably as a result of oceanic conditions in subtropical environments. Unlike most of the plant groups previously studied in the Macaronesian region, the lineages of Olea illustrate how low phenotypic differentiation can be also related to a clear molecular differentiation in oceanic island enclaves. © 2010 The Linnean Society of London, Botanical Journal of the Linnean Society, 2010, 162, 54–63.

Darwin's interest in carnivorous plants was in keeping with the Victorian fascination with Gothic horrors, and his experiments on them were many and varied, ranging from what appears to be idle curiosity (e.g. what will happen if I place a human hair on a Drosera leaf?) to detailed investigations of mechanisms. Mechanisms for capture and digestion of prey vary greatly among the six (or more) lineages of flowering plants that have well-developed carnivory, and some are much more active than others. Passive carnivory is common in some groups, and one, Roridula (Roridulaceae) from southern Africa, is so passively carnivorous that it requires the presence of an insect intermediate to derive any benefit from prey trapped on its leaves. Other groups not generally considered to be carnivores, such as Stylidium (Stylidiaceae), some species of Potentilla (Rosaceae), Proboscidea (Martyniaceae) and Geranium (Geraniaceae), that have been demonstrated to both produce digestive enzymes on their epidermal surfaces and be capable of absorbing the products, are putatively just as ‘carnivorous’ as Roridula. There is no clear way to discriminate between cases of passive and active carnivory and between non-carnivorous and carnivorous plants – all intermediates exist. Here, we document the various angiosperm clades in which carnivory has evolved and the degree to which these plants have become ‘complete carnivores’. We also discuss the problems with definition of the terms used to describe carnivorous plants. © 2009 The Linnean Society of London, Botanical Journal of the Linnean Society, 2009, 161, 329–356.

We provide here a subfamilial scheme for the expanded asparagalean families Amaryllidaceae, Asparagaceae and Xanthorrhoeaceae. Our recommendation is that the first family has three subfamilies (Agapanthoideae, Allioideae and Amaryllidoideae), the second has seven (Agavoideae, Aphyllanthoideae, Asparagoideae, Brodiaeoideae, Lomandroideae, Nolinoideae and Scilloideae) and the last has three (Asphodeloideae, Hemerocallidoideae and Xanthorrhoeoideae). Tribal names are provided for the large subfamilies Allioideae, Amaryllidoideae and Scilloideae. The use of these subfamily names permits easier descriptions of characters by specialists for these well-supported subclades, but the use of the broader family limits greatly simplifies the taxonomy of Asparagales and thus makes the teaching of these families much easier. A new subfamilial name, Xanthorrhoeoideae, and a new tribal name, Oziroeeae, are proposed. © 2009 The Linnean Society of London, Botanical Journal of the Linnean Society, 2009, 161, 132–136.

We address the generic limits of Gnidia (Thymelaeaceae) through a phylogenetic analysis of nuclear ribosomal DNA internal transcribed spacer (ITS) and plastid rbcL, trnL intron and trnL-F intergenic spacer regions. Maximum parsimony and Bayesian inference were used to produce trees and assess internal support. The most significant conclusion drawn from the molecular analysis is that Gnidia is polyphyletic as currently circumscribed, comprising at least four distinct lineages that are each related to other genera within Thymelaeoideae. Gnidia pinifolia and G. racemosa are members of a clade within which Struthiola is embedded; a second group of species allies with Drapetes as sister to Passerina; and a third lineage corresponds to the previously recognized genus Lasiosiphon. The remaining species of Gnidia included in this study are allied with the Australian genus Pimelea. The taxonomic implications of these findings are discussed in relation to the principle of monophyly. © 2009 The Linnean Society of London, Botanical Journal of the Linnean Society, 2009, 160, 402–417.

Variation at five polymorphic microsatellite loci was used to investigate genetic diversity and differentiation of two tetraploid Canarian endemics, Bencomia exstipulata and B. caudata. Data were analysed and are discussed in terms of tetrasomic (autotetraploid) and disomic (allotetraploid) inheritance. In both cases, genetic diversity values were similar to those described in other tetraploid plant species. High genetic differentiation between the only two described natural populations of B. exstipulata was detected (FST = 0.411). Bayesian cluster analysis revealed a geographical structure with distinct genetic groups from each island. High genetic differentiation and low genetic diversity of the B. exstipulata population from Tenerife suggest a recent population bottleneck, perhaps caused by the most recent major volcanic eruption, for this natural locality. This may be heightened by possible inbreeding depression and the monoecy of these species. Polymorphic microsatellite loci were also tested across all species in the Bencomia alliance. These reliably amplified the target sequence, suggesting a high degree of conservation of the sequences flanking the microsatellites. © 2009 The Linnean Society of London, Botanical Journal of the Linnean Society, 2009, 160, 429–441.

The biota of the isolated, humid brejo forests ofnorth-east Brazil have interested biologists for manyyears and have played a part in the continuing effortto understand the historical biogeography andpossible migration routes between the major forestbiomes of Amazonia and Atlantic Brazil (for example,Andrade-Lima, 1953, 1966, 1982; Ducke, 1959;Prance, 1973; Vanzolini, 1973; Ab’Saber, 1982;Bigarella & Andrade-Lima, 1982; Lourenco, 1988;Giulietti & Forero, 1990; Ledru, 1991, 1993; Oliveira-Filho & Ratter, 1995; Behling, 1998; Oliveira, Barreto& Suguio, 1999; Barreto, Pessenda & Suguio, 2002;Borges-Nojosa & Caramaschi, 2003; Costa, 2003;Cavalcanti & Tabarelli, 2004; Porto, Cabral T Vasconcelos, Almeida & Costa, 2004).Despite the destructive impact of human activity,these two large forest regions are dominating features

We report the first confirmed occurrence in Britain of Orchis × angusticruris Franch. ex Rouy, a hybrid between two closely related orchid species of anthropomorphic Orchis (O. purpurea Huds. × O. simia Lam.) that hybridize frequently in Continental Europe. Seven individual hybrids, most likely F1 plants representing a single interspecific pollination event, first flowered with both parents in May 2006 at a nature reserve in the Chiltern Hills near Goring, Oxfordshire. Univariate and multivariate morphometric analyses (43 characters plus 12 indices), internal transcribed spacer sequencing, plastid microsatellites and amplified fragment length polymorphism (AFLP) analyses together readily separate the parents and confirm that O. purpurea was the ovule parent and O. simia the pollen parent, presumably reflecting the greater frequency and/or later flowering period of the latter at the site. This study reinforces a more general observation that, in most orchids, the ovule parent contributes substantially more to the hybrid phenotype than does the pollen parent, perhaps reflecting cytoplasmic inheritance. In contrast, the hybrids are placed closer to O. simia than to O. purpurea in the AFLP tree. Apparently recent arrivals, the few O. purpurea plants at Goring contrast genetically with the two other small populations of this species known in the Chilterns, but rather are consistent with relatively uncommon Continental populations. This suggests that the plants may have been deliberately introduced at Goring by man, although transport from the Continent in high-level air currents cannot be ruled out. The Goring population of O. simia is likely to have become genetically impoverished through (1) preferential removal of many relatively fit plants to herbaria in the 19th century and/or (2) a catastrophic population crash in the first half of the 20th century. However, both our re-examination of herbarium specimens and our population genetic data indicate past hybridization among anthropomorphic Orchis species occurring naturally in the Chilterns. Thus, we tentatively recommend retention of the hybrid plants at Goring, despite their likely anthropogenic origin from Continental material and the partial viability of their pollen and seeds, which offers opportunities for future introgression. Although the Goring hybrids broadly resemble morphologically O. militaris, another anthropomorphic Orchis still found at two Chiltern localities, sufficient morphological and molecular differences were observed to strongly refute our initial hypothesis that O. militaris could have originated through hybridization between ancestors that resembled O. purpurea and O. simia. The comparatively complex genetic properties evident in both O. simia and O. purpurea merit further study. © 2008 The Linnean Society of London, Botanical Journal of the Linnean Society, 2008, 157, 687–711.

The parentage of polyploid Sorbus species in the British Isles was investigated using plastid DNA microsatellites. Four hundred and fifty-three samples from 30 taxa were screened using six microsatellite fragments, which gave 28 haplotypes. The haplotypes formed groups clearly related to the ancestral diploids Sorbus aria, Sorbus aucuparia, and Sorbus torminalis. Species in the Sorbus aria group all had Aria haplotypes (with the exception of one English S. aria), species in the Sorbus anglica group had an Aucuparia haplotype, and species in the Sorbus latifolia group had a Torminalis haplotype. Sorbus intermedia had an Aucuparia haplotype. This indicated that the hybridization events that led to the formation of species in the S. anglica and S. latifolia groups usually did so with S. aria s.l. as the pollen-donating (paternal) parent. The polyploids S. anglica, Sorbus bristoliensis, Sorbus croceocarpa, Sorbus decipiens, Sorbus devoniensis, Sorbus hibernica, Sorbus lancastriensis, Sorbus leptophylla, Sorbus leyana, Sorbus minima, Sorbus rupicola, Sorbus subcuneata, Sorbus vexans, Sorbus whiteana, Sorbus wilmottiana, and three unnamed taxa may each be derived from a single maternal lineage. The polyploids Sorbus eminens, Sorbus porrigentiformis, and S. latifolia have multiple maternal lineages. The two primary diploid hybrids S. × thuringiaca and S. × vagensis have arisen many times independently. © 2007 The Linnean Society of London, Botanical Journal of the Linnean Society, 2007, 154, 291–304.

Four plastid markers, four nuclear markers and 14 morphometric characters were used in this study to investigate the evolution of Dactylorhiza baltica (Orchidaceae) in European Russia. In total, 98, 214 and 775 samples from 85, 112 and 121 populations were involved in the combined and separate molecular and morphometric analyses, respectively. In most cases, morphometric measures were done on exactly the same plants that were used for DNA studies. Dactylorhiza baltica plants from European Russia are most probably the products of several recent and mostly local hybridization events between the diploids D. fuchsii and D. incarnata, which have each been the maternal parent on different occasions. Considerable introgression into the parental diploids via the allopolyploid D. baltica is also hypothesized. Several morphological characters, such as length of the lip lateral lobe and the length of longest leaf, were found to be robust and could be useful in identification of D. baltica. This study demonstrates the advantage of ‘combined’ techniques, especially in the case of taxonomically complex taxa. © 2005 The Linnean Society of London, Botanical Journal of the Linnean Society, 2005, 147, 257–274.