Notes on the Taxonomy of the Genus Oryza

An enumeration of the species of Oryza as well as a key to the genus has been prepared by Chatterjeei).He recognizes 23 valid species from amongst the 30-40 specific names published previously.Since this publication, three new species have been described, bringing the total to 26.Though the genus is small, it is economically important and its taxonomy is of interest to workers tracing the interrelationships, origin, and evolution of cultivated rice.Kihara2) has drawn attention to some probable errors in nomenclature.The present author has committed some errors in the past and the notes given below are intended to rectify them, and also to assist others in this field.This note is based on cytogenetical studies on Oryza collection at the Central Rice Research Institute, Cuttack, and observations on the Oryza species and hybrids being grown at the National Institute of Agricultural Sciences, Hiratsuka 1) Oryza perennis Moench.Chatterjeei) includes under this specific name Asian, American, as well as African perennial wild rices.A group of African wild rices can be conveniently separated from O. perennis and renamed O. barthii A. Cheval.This species is characterized by erect habit, and strongly spreading rhizomes, as contrasted to the residual O. perennis, having weak rhizome development, and spreading, prostrate or floating habit.Both the species are adapted for outcrossing, but seed setting under selfing bags is partial in O. perennis, and nil in O. barthii.The failure to set seed on selfing, in O. barthii suggests self incompatibility, but, the observation needs confirmation.If self incompatibility exists, it is a distinctive feature, and a new specific name is justifiable.As far as is known, O. barthii and O. perennis have allopatric distribution, and hybridization between the two is difficult, and artificial cross pollinations have been hitherto unsuccessful.2) O. cubensis.The variety of wild rice occurring in Cuba must be included under O. perennis and Chatterjeei) does not record O, cubensis as a validly published name.The plants hitherto collected and grown have a semi-erect habit as contrasted to the floating or prostrate habit of Asian O. perennis, and since they are geographically isolated from the latter, can be given a varietal status, namely O. perennis var.cubensis. 3) O. paraguaensis.There is no record of this specific name being published according to the rules of nomenclature.The South American plants given this specific name, are taxonomically identical with O. alta Swallen, and according to the rules this latter would be the valid name.4) O. latifolia Desv.This name has been applied to two different species, one distributed in Tropical America and the other in Asia and possibly in Africa.The specific name O. latifolia is validly applicable to the American species which is a tetraploid (2n=48).The Asian species is diploid (2n=24) and is classifiable as O .officinalis Wall.In the standard book on Flora of British India, by Hooker, this species is listed as O. latifolia and the error has been copied by others.Chatterjeei) uses the character, ligule fringes, to discriminate between the two species, the tetraploid O. latifolia having fringed ligules, the diploid O. officinalis not having the fringes.This distinction is not absolute, because new observations have shown