Introduction

Phylogenetic studies have greatly improved our understanding of the large family Acanthaceae (McDade et al., 2000 and references herein) and were discussed in the last comprehensive review of the generic level taxonomy of the family by Scotland & Vollesen (2000). Several major well-supported phylogenetic lineages have been revealed and these have been discussed on the basis of morphological and molecular evidence. Expanded investigations have subsequently permitted the delimitation and generic circumscriptions within certain tribes to be further refined, such as in the case of the largest of these, the Ruellieae Nees (c. 1200 spp.) (Tripp, 2007; Tripp et al., 2009; Tripp et al., 2013). The Ruellieae is well-represented in Madagascar, and while some of the necessary nomenclatural changes following the revised generic circumscription have already been effected (Callmander et al., 2014; Tripp, 2007; Tripp et al., 2013), others require further investigation.

Clarke (1899) described the genus Acanthopale C. B. Clarke in his treatment of the Acanthaceae for “Flora of Tropical Africa” to accommodate all the African species formerly included in Strobilanthes Blume, from which it differed notably in pollen morphology. Benoist (1939) retained in the latter genus the only species of this group previously known from Madagascar, S. madagascariensis Baker, and described two new species. However, S. madagascariensis (but not the other species) was later transferred to Acanthopale by Bremekamp (1943), who described additional species from Continental Africa and argued that the two genera are distinct. Benoist (1945, 1946, 1947) nevertheless continued to regard the two genera as synonyms and described six additional species and one variety from Madagascar in Strobilanthes. One of these species, S. gracilicaulis Benoist, was later transferred to the genus Dyschoriste Nees (Benoist, 1967).

A pragmatic approach to the systematics of Strobilanthes has been adopted recently and provides geographical, morphological and molecular phylogenetic coherence (Carine & Scotland, 2002; Moylan et al., 2004).This approach has been adopted in all recent taxonomic revisions (e.g., Wood & Scootland, 2003, 2006,2009; Deng et al., 2006,2010). The genus Strobilanthes as currently defined includes more than 450 spp. and is distributed from South Asia to Northern Australia through Southeast Asia and Malesia (Moylan et al., 2004), whereas Acanthopale is restricted to Africa and Madagascar (Bremekamp, 1943, 1944; Scotland & Vollesen, 2000; Tripp et al., 2013).

In the course of reviewing the genera of Acanthaceae for the “Catalogue of the Vascular Plants of Madagascar” (Madagascar Catalogue, 2015), we noticed that only a single species of Strobilanthes from Madagascar had an available combination in Acanthopale, despite the general acceptance that the genus Strobilanthes is absent from Africa and Madagascar. For the preparation of this note we examined all of the material cited by Benoist (1967) as well as additional collections known from Madagascar and deposited at G, K, MO and P, mostly more recent, and reviewed the eight species of Strobilanthes that he recognised in the Flora. In this note we establish the five necessary new combinations for the Malagasy species recognized in Strobilanthes by Benoist (1967) and place two species and one variety in synonymy. We also provide brief explanatory observations for certain species, and a full list of specimens examined and distribution maps for all the accepted species of Acanthopale can be found in the Madagascar Catalogue (2015).

Recent taxonomic treatment for Continental Africa accepts 6 to 8 species (Vollesen, 2008), and the 6 species accepted here for Madagascar therefore nearly double the number of species in the genus. A complete revision of the genus may reveal that certain African and Malagasy species should be regarded as synonyms, but will also likely reveal new species endemic to Madagascar. Tripp (2007) has suggested that two Malagasy species currently placed in Ruellia L. with echinate pollen grains (typically coarsely reticulate in Ruellia) may also be better included in Acanthopale: Ruellia cyanea Benoist and R. dissidens Benoist. Further taxonomic and molecular work is needed to determine the systematic position of these two species and indeed to fully understand the diversity of the genus in Madagascar as a whole and its relation with the African species. A complete revision of Acanthopale is clearly needed (see also Clarke, 1899; Vollesen, 2008).

Nomenclature