J Plant Res (2002) 115:381–392 © The Botanical Society of Japan and Springer-Verlag Tokyo 2002 Digital Object Identifier (DOI) 10.1007/s10265-002-0049-3 ORIGINAL ARTICLE Trevor R. Hodkinson • Mark W. Chase • M. Dolores Lledó • Nicolas Salamin • Stephen A. Renvoize Phylogenetics of Miscanthus, Saccharum and related genera (Saccharinae, Andropogoneae, Poaceae) based on DNA sequences from ITS nuclear ribosomal DNA and plastid trnL intron and trnL-F intergenic spacers Received: February 4, 2002 / Accepted: June 19, 2002 / Published online: August 28, 2002 Abstract DNA sequences were used to assess the mono- phyly and inter-relationships of Miscanthus, Saccharum Introduction and related genera in the Saccharum complex. Three DNA regions were sequenced, including the trnL intron and the Tribe Andropogoneae (Poaceae) includes many species trnL-F intergenic spacer of the plastid genome and the ITS with high economic value, including the C4 grasses Saccha- region of nuclear ribosomal DNA (nrDNA). Because it was rum officinarum L. (sugarcane), Sorghum bicolor (L.) more variable, the ITS region proved most suitable for phy- Moench (sorghum) and Zea mays L. (maize). Subtribe Sac- logenetic reconstruction at this level, and the results indi- charinae Griseb. includes Saccharum L. and Miscanthus cate that Miscanthus s.l. and Saccharum s.l. are polyphyletic. Anderss., the latter having considerable potential as a bio- A set of species from Saccharum section Ripidium (clade a) mass crop for renewable energy production and raw mate- do not group closely with any members of Saccharum s.l.. A rial for the cellulose and paper industries (Bullard et al. number of Miscanthus species from eastern or south- 1995; Clifton-Brown and Lewandowski 2000). Saccharinae eastern Asia represent a monophyletic group with a basic according to Clayton and Renvoize (1986), also include chromosome number of 19 (clade b), but the other species Eriochrysis P. Beauv., Eulalia Kunth, Eulaliopsis Honda, from Africa and the Himalayas are clearly excluded. There Homozeugus Stapf., Imperata Cyr., Lophopogon Hack., is support for a monophyletic Saccharum s.s. clade including Microstegium Nees, Pogonatherum P. Beauv., Polytrias S. officinarum and S. spontaneum that is sister to Miscanthus Hack. and Spodiopogon Trin. The subtribe is morphologi- s.s. (clade c). There is no evidence to support the division cally defined by their terminal inflorescence (except Eulali- of some Saccharum s.l. into the genera currently known as opsis and Pogonatherum) of solitary or digitate racemes and Erianthus and Narenga. Saccharum contortum (= Erianthus paired similar spikelets. The paired spikelets are often plu- contortus), S. narenga (= Narenga porphyrocoma) and mose; the callus is rounded or truncate. The lower glume is Erianthus rockii, group more closely with Miscanthus fus- mostly thin, and the lower floret is usually reduced to a cus, a species from the Himalayas and also with the African sterile lemma. The upper lemma is entire or bilobed and Miscanthus s.l. species (= Miscanthidium, clade d). can have a glabrous awn. Despite these characteristics the only known morphological synapomorphy for Saccharinae Key words Erianthus • Miscanthus • Molecular • would be their bisexual paired spikelets. Other Andropogo- Saccharum • Sugarcane • Systematics neae have paired spikelets, but one of these is usually either male or sterile. However, many exceptions exist, such as some species of Ischaemum (Ischaemninae) and Schizach- rium (Andropogoninae), which have bisexual paired spike- lets. Saccharinae, therefore, are poorly defined, and their monophyly remains insufficiently evaluated. Systematists have used the term “Saccharum complex” to describe a subset of the Saccharinae (Erianthus, Miscanthus, Narenga, Saccharum and Sclerostachya) implicated in the origin of • • • T.R. Hodkinson M.W. Chase M.D. Lledó S.A. Renvoize sugarcane and in which the taxonomy is particularly con- Royal Botanic Gardens, Kew, Richmond, Surrey, UK fused (Daniels and Roach 1987). There is a need to charac- T.R. Hodkinson (*) • N. Salamin terise this complex more comprehensively. Department of Botany, Trinity College, University of Dublin, Dublin 2, The monophyletic status of many genera within Saccha- Ireland Tel. +353-16081128; Fax +353-16081147 rinae is also in doubt. The most widely debated is Saccharum e-mail: [email protected] itself. Saccharum s.l. (Clayton and Renvoize 1986) has been Springer-VerlagTokyoJournal of Plant ResearchJ Plant Res102650918-94401618-0860s10265-002-0049-30049Bot Soc Jpn and Springer-VerlagOriginal Article 382 divided into a number of other genera including Erianthus However, they also recognised that Eriochrysis, Eulalia, Michaux, Narenga Bor and Ripidium Trin., but Clayton and Imperata, Miscanthus, Saccharum and Spodiopogon form a Renvoize (1986) chose to combine all of these genera under closely knit group in which the phylogenetic relationships Saccharum because the characters used to define them were are unclear. Saccharum is considered by many as the closest thought to be more suited to infra-generic categorisation. relative of Miscanthus, and these two genera frequently They argued that the division of awned (Erianthus) and hybridise (Sobral et al. 1994). awnless species is artificial and the separation of Narenga, The species of Miscanthus can be distinguished from with its coriaceous glumes, is trivial because this is no more those of Saccharum by their tough inflorescence rachis and than an extreme expression of a trend found elsewhere in both spikelets of a pair being pedicellate, although the the genus. Saccharum s.l. has approximately 40 species pedicels are of different lengths. Most members of Miscant- (Clayton and Renvoize 1986), Erianthus approximately 20 hus s.l. are native to eastern or south-eastern Asia (China, and Narenga only two (Adati and Shiotani 1962). Saccha- Japan and neighbouring regions); two species are known rum s.s. (Price 1963; Daniels and Roach 1987) is distributed from the Himalayas and four from southern Africa (Fig. 1). throughout the tropics and subtropics due to cultivation, but In the past, the African species have been placed in a sep- the species are native to south-eastern Asia. Erianthus is cos- arate genus, Miscanthidium Stapf, mainly on the basis of mopolitan and, according to Celarier (1956), can be divided their elongate inflorescence axis and short racemes, but into two subgenera, one with two anthers and an American these differences were not considered sufficient to warrant distribution; the other with three anthers and an Old World separation from Miscanthus by Clayton and Renvoize distribution. The Old World distribution of Saccharum s.l. is (1986) because M. floridulus (Labill.) Warb. Ex. K. Schum. given in Fig. 1. The New World species are predominantly & Lauterb. and M. fuscus (Roxb.) Benth. [= Sclerostachya found in North America. fusca (Roxb.) A. Camus] in Asia also have elongate inflo- The taxonomic status of Miscanthus is also in a state of rescence axes. flux, and little is known about the identity and inter- Groups of species at sectional rank within Miscanthus relationships of its species. According to Clayton and have been recognised, and a key to Miscanthus species was Renvoize (1986), Miscanthus s.l. comprises approximately given in Hodkinson et al. (1997). The most comprehensive 20 species and appears well-defined morphologically. effort to subdivide the genus was made by Lee (1964b, c, d), 1 Fig. 1. Distribution of Miscanthus and Saccharum sensu lato species in the Old World. The major areas of distribution are shown by rings, but exclude occasional records from elsewhere. Saccharum officinarum (sugarcane) is not included because it has a widespread distribution due to cultivation 383 who separated the Asian species into four sections that Using DNA sequence data of nuclear and plastid DNA broadly agreed with the treatments of Honda (1930) and regions, this study aimed to assess the monophyly of Mis- Adati and Shiotani (1962). There are four African Miscant- canthus and Saccharum and their phylogenetic relationships hus species not included in the genus by Lee (1964b, c, d) to other Saccharinae. DNA sequencing is particularly well namely: M. ecklonii (Nees) Mabb., M. junceus (Stapf) suited for phylogenetic studies and has been used exten- Pilger, M. sorghum (Nees) Pilger and M. violaceus (K. sively for such purposes at many different taxonomic levels Schum.) Pilger. Himalayan M. fuscus was also excluded (Chase et al. 1993; Hsiao et al. 1994; Soltis et al. 1999; from Miscanthus by Lee (and recognised as Sclerostachya Salamin et al. 2002). The internal transcribed spacer (ITS) fusca), and has been included by Clayton and Renvoize regions of nuclear ribosomal DNA (for a review see Baldwin (1986). Miscanthus brevipilus Hand.-Mazz., M. changii, Y.N. et al. 1995), and the trnL intron and trnL-F intergenic spacer Lee, and M. eulaliodes Keng ex. Hand.-Mazz., which were of plastid DNA (hereafter trnL-F; Taberlet et al. 1991; Gielly listed by Lee (1964b, c, d), have not been evaluated here and Taberlet 1994; Hopper et al. 1999; Molvray et al. 1999; because they were not available for study. Miscanthus trans- Chase et al. 2000; Lledó et al. 2000) were used to construct morrisonensis appears to intergrade with M. sinensis on a phylogenetic hypotheses with parsimony methods. The ITS morphological level. Another taxon known as M. conden- region has proven useful for phylogenetic studies at this tax- satus Hackel has been given species rank by previous onomic level in various plant groups, including grasses authors, but falls within the normal range of morphological (Baldwin et al. 1995; Hsiao et al. 1995a, b, 1999; Hodkinson variation found in M. sinensis (Koyama 1987). Lee (1964a) et al. 2000; Grass Phylogeny Working Group 2001; recognised M. condensatus on the basis of leaf anatomy Hodkinson et al. 2002a) and so has the trnL-F region (Briggs because it differs in having unclosed bundles in the midrib. and Johnson 2000; Briggs et al. 2000). A number of grass In addition, M. condensatus only grows at low elevation in ITS sequences have been published and/or deposited in the coastal zone of Japan.