Chromosome counts for 27 species plus one intersubsectional hybrid are presented from all seven subsections of section Vireya in the genus Rhododendron. Twenty-three are new, and five confirm previously published counts. The plants originate from a wide geographical and altitudinal range; all have a uniform somatic number of 2n=26.

Alpinia galanga is an important species cultivated as a culinary spice and is the type species of the genus. It is hence a member of sect. Alpinia subsect. Alpinia (with non-tubular bracteoles). However, molecular phylogenetic analyses suggest that A. galanga is closely related to A. nigra in sect. Allughas (with tubular bracteoles). This clade, which includes A. conchigera, is strongly supported with a bootstrap value (BS) of 100% and a decay index (DI) of >+6. These results are based on the internal transcribed spacer (ITS) region of the 18S–25S nuclear ribosomal DNA. The region (405–423bp) was sequenced from 17 accessions representing 16 taxa of Zingiberaceae, including 15 species of Alpinia and one outgroup. The sequence divergence ranged from 0.5 to 15.6% among the ingroup and from 10.1 to 13.3% between the ingroup and the outgroup. The results also strongly support the sister relationship of A. rafflesiana and A. javanica in section Allughas (BS=100%, DI=>+6), thus the whole section is paraphyletic. Section Alpinia subsect. Catimbium is monophyletic (BS=100%, DI=+5). On the other hand, sect. Alpinia subsect. Alpinia is paraphyletic (BS=100%, DI=>+6) with respect to sect. Alpinia subsect. Catimbium. The results from a phylogenetic analysis of a subset of the taxa using the spacer between trnL (UAA) 3′ exon and trnF (GAA) of chloroplast DNA confirmed the position of A. galanga in sect. Allughas. It appears that A. galanga has evolved within sect. Allughas and the absence of tubular bracteoles is a convergence with sect. Alpinia, possibly as a result of evolution under domestication.

A phylogenetic study of Roscoea (Zingiberaceae), a high-altitude genus of an otherwise tropical plant family, was undertaken using sequence data from the internal transcribed spacers (ITS) of the nuclear ribosomal DNA (nrDNA). Two species of Cautleya and two species of Curcuma were used as outgroups. This resulted in an aligned matrix of 436bp (ITS1, 203bp; ITS2, 233bp). Sequence divergence of ITS1 and ITS2 within the ingroup ranged from 0–13.9% and 0–7.6% respectively. The results suggest that Roscoea is monophyletic (BS=99%; DI=>3) with the genus Cautleya as sister group. Roscoea itself is divided into two sister clades which correlate with geography: a ‘Chinese’ clade (BS=67%; DI=+2) and a ‘Himalayan’ clade (BS=59%; DI=+1). These two groups are disjunct across the ‘Brahmaputra gap’, a region in which no Roscoea spp. have been recorded. The only species which occurs on both sides of the Brahmaputra gap is Roscoea tibetica. However, the western populations of Roscoea tibetica (from Bhutan) show numerous morphological differences. It is therefore possible that Bhutanese R. tibetica represents a distinct taxon, possibly more closely allied to Himalayan species.

A re-examination of the status of the monotypic genus Saniella Hilliard & Burtt (Hypoxidaceae) was inspired by the rediscovery of Empodium occidentale on Hantam Mountain in Calvinia Division, SW Cape. This plant does not belong to Empodium, and has critical features that show that it is better placed as a second species of Saniella, but at the same time it lessens the gap between Saniella and Spiloxene. It is concluded that the whole question of the interrelations of Saniella, Spiloxene and Pauridia needs to be studied in depth. In the interim these three genera should all be maintained. Empodium occidentale is accordingly transferred to Saniella.

Cystopteris dickieana R.Sim is a rare fern protected in Britain under the 1981 Wildlife and Countryside Act. Most current floras treat it as a distinct species but ever since it was first discovered in Scotland in the 1830s, there has been considerable debate about its taxonomic status within the C. fragilis complex. This debate centres on the relative importance of two characters, the architecture of the fronds and the surface sculpturing of the spores, in delimiting C. dickieana from other taxa in the C. fragilis complex. The type specimens of C. dickieana have distinctive fronds. Plants with similar frond architecture have, to date, been recorded growing naturally only at the site in Scotland from which the type specimens were collected and at one other site nearby. The type specimens of C. dickieana also have mature spores with surface sculpturing often described as ‘rugose’. These are distinctive and unusual in the genus Cystopteris, in which most taxa have ‘echinate’ spores. However, rugose-spored plants have been recorded not only at, and near, the type locality in Scotland but also at many other sites in the northern hemisphere in populations of plants defined largely on the basis of frond architecture as C. fragilis or C. baenitzii. This indicates that spore sculpturing should not be used alone to delimit C. dickieana from other taxa within the C. fragilis complex but, despite this, the literature on ‘C. dickieana’ contains many reports of studies on material identified as C. dickieana solely on the basis of spore sculpturing. This, combined with the fact that most comparative studies have also failed to include material known to have come from the type locality, has resulted in considerable and continuing uncertainty over the taxonomic status and distribution of C. dickieana.

Cystopteris dickieana R. Sim is a fern species that currently receives legal protection in Britain on the basis that it is endemic to Scotland and extremely rare. However, ever since it was first discovered in the 1830s, there has been considerable debate about its taxonomic status within the Cystopteris fragilis complex. This debate centres on the relative importance of two characters, the architecture of the fronds and the surface sculpturing of the spores, in delimiting C. dickieana from other taxa in the C. fragilis complex and is complicated by the fact that most comparative studies reputed to have included ‘C. dickieana’ have failed to include either the type material of C. dickieana or material from the ’type population‘ (i.e. the population of plants, of uniform and distinctive morphology, that still grows at the type locality for C. dickieana). The principal aim of this investigation was to generate some of the critical data and information currently missing from the debate concerning the taxonomic status of C. dickieana by including material from the ’type population‘ in an allozyme study of several Scottish Cystopteris populations. The frond architecture, spore sculpturing, and allozyme banding patterns were investigated in five natural and two cultivated populations of the C. fragilis complex. All plants from the ‘type population’ of C. dickieana had the distinctive frond architecture and rugose spores of the type material. Two populations comprised plants that had the frond architecture and echinate spores typical of C. fragilis. Also examined were two populations containing plants that had frond architecture typical of C. fragilis but rugose spores similar to those of the type material of C. dickieana. These plants occurred with typical C. fragilis in one of the populations and with typical C. dickieana in the other. In the latter population, situated close to the ‘type population’ for C. dickieana, a third, ‘smooth’ spore type occurred in nine out of 55 fertile plants. Consistent and analysable allozyme results were obtained for seven enzymes at ten loci: eight loci were polymorphic and all but the smallest population showed variation at one or more loci. In total, 22 allozyme phenotypes were recorded in 226 individuals. No consistent correlation was found between frond architecture, spore sculpturing and allozyme banding pattern. Our results provide no support for the recognition of C. dickieana as a distinct species related to C. fragilis.

The floristic and ecological aspects of the tree vegetation of a remnant forest in the Godoy Forest State Park were studied. Two hundred and six native species of the region occur in the park. Comments on ecological aspects, such as dispersion, structure of vegetation and succession of these species, are provided. An analysis of the ecological parameters indicated that the occurrence of late secondary and climax species (55.8%) was greater than that of pioneer (6.8%) and early secondary species (34.5%). This seems to indicate a predominant dependence of biotic factors on the abiotic factors and an advanced stage of succession.

The forest across a sharp boundary between greywacke and serpentinized peridotite is described from a site with seasonal rainfall on Palawan, Philippines. The forest on greywacke was of much larger stature (trees up to 26m) than that on the serpentinized peridotite (trees up to 18m). The tree (>10cm dbh) species richness was the same on both substrata with 38 species in one 0.16ha plot on each side of the boundary. There were many more individuals in the greywacke plot (149) than on the serpentinized peridotite (114). Floristically the plots were very different, with only one tree species, an unidentified Madhuca, occurring on both sides of the boundary. The soil over the greywacke was notably more acid, had lower Mg/Ca quotients, and much lower nickel concentrations than the soil over serpentinized peridotite.