Abstract

Avena breviaristata, collected only once (1882) in Algeria and never re-collected since, is a very mysterious grass species because unfortunately even the type specimen got lost 60--80 years ago. Morphological information on this species was thus based on a few published descriptions, which made it difficult, however, to correctly infer the genus affiliation of this morphologically odd species. Avena breviaristata became affiliated in the past with various oat-like genera (Avenula, Helictotrichon, Tricholemma). Due to the recent rediscovery of the type specimen at the P herbarium and the opportunity to study this specimen, we report here on the morphological characters of A. breviaristata underpinned by meaningful illustrations. They are discussed in comparison with the morphology of representative species of the above-mentioned genera. Uncommon characters of the spikelets (type of disarticulation of the rachilla, lemma structure, lodicules) and to some extent of the inflorescences, leaves and leaf sheaths support the inclusion of A. breviaristata in the North African genus Tricholemma. Considering biogeography, T. breviaristatum from the arid Hauts Plateaux in Algeria is a highly xeromorphic counterpart of the mesomorphic species T. jahandiezii, which is confined to higher altitudes of the rather humid Moyen Atlas in Morocco. This underlines the status of Tricholemma as a relic endemic. Our morphological survey supports the classification of Avenula (only A. pubescens) as separate from Helictotrichon s.str. and Helictochloa. Moreover, morphological evidence does not support an origin of A. pubescens by intergeneric hybridization between the latter genera as hypothesized in some prior studies. Especially the glabrous palea, the special shape of the lodicules and the structure of the awn show no intermediacy. The complicated history of the type collection of T. breviaristatum and the role of botanical authors are given.