Institut Francois de Recherche Scientifique pour le Developement en Cooperation (ORSTOM), Centre de Montpellier, BP 5045, 34032 Montpellier Cedex, France. Present address: Department of Plant Breeding and Biometry, Cornell University, Ithaca, N.Y. 14853 U.S.A.
During my stay in the United States, I visited several herbaria: those of the Botanical Gardens in New York (Bronx) and Saint Louis (Missouri), of the Smithsonian Institution in Washington D.C. and of the National University in Asuncion and the National Museum in Rio de Janeiro. During the time available, I paid particular attention to the specimens of the Oryza latifolia complex. I also had the opportunity to visit briefly and collect specimens of the 0. latifolia complex in the Guanacaste Province of Costa Rica (28-29 April 1989) and in the humid Chaco of Paraguay (2-3 May 1989).
The distribution of the 0. latifolia complex extends from Southern Mexico to Northern Argentina. At least three forms are distinguishable: 0. grandiglumis, 0. alta and 0. latifolia, although they are often considered as conspecific. Oryza grandiglumis is clearly distinguishable by its long glumes. Oryza alta has larger spikelets which tend to resemble those of the Sativa group (A genome) more than those of the Latifolia group of species (B, C, BC and CD genome).
As measured in the herbarium of the Smithsonian Institution in Washington D.C., the size of the body of spikelets in 0. alta varied between 8.0 and 8.5 mm in length and 2.0 to 3.2 mm in width. The respective measurements for 0. latifolia were 4.3 to 7.7mm and 2.0 to 2.6mm. All specimens of 0. alta observed in the herbaria visited were collected in the Amazon Basin, except for one specimen from Belize and one from the Lake Ipacaray in Paraguay.
Forms of the 0. latifolia complex tend generally to be tall and vigorous ("gigantic"). Numerous herbarium vouchers indicate however that a much smaller ecotype is commonly encountered in Central America. The 0. latifolia complex is tetraploid with a CD genome. The genome D is not known at diploid level but a diploid form was reported, although not properly documented, from the humid Chaco of Paraguay (Bruchner 1977).
Comparing the working collection that we gathered from the collections held at the National Institute of Genetics (Mishima, Japan) and the International Rice Research Institute (Philippines) and studying isozyme (Second 1984) and chloroplast DNA polymorphism (Dally 1988), we lacked several of the above mentioned forms, although we had samples of 0. latifolia, 0. alta and 0. gratidiglumis: We lacked the 0. alta forms with the largest spikelets resembling those of the A genome species, the small ecotype of 0. latifolia from Central America, and the form from Paraguay that could be a diploid resembling 0. latifolia.
I collected the two last fornis in Costa Rica and Paraguay respectively, as the following accessions:
AA1 - 0. latifolia. Costa-Rica, Guanacaste Prov., 3.5km past Canas on the way to Liberia. Road side ditch and adjacent irrigated paddy field. Plants small, 60 to 80cm high.
AA2 - 0. latifolia. Costa-Rica. Guanacaste Prov. On the way from Guajinikil to Murcielago part of Santa Rosa National park. Gallery forest remnant, at the entrance of the park. A few scattered plants, grazed.
AA3 - 0. latifolia. Costa-Rica. Guanacaste Prov. In paddy fields and irrigation canals of the Tempisque valley.
AA4- O. latifolia. Paraguay. Near Tacuara, along the humid Chaco main highway, Km 112-114. A dense population in a pool surrounded by small trees and shrubs, 80 to 180cm high.
AA5 - 0. latifolia. Same as AA4 but scattered under trees and shrubs along the highway in the same area.
AA6 - 0. latifolia. Same as AA5, Km 98-100. Along the highway.
The time of my visit in Costa Rica, at the end of the dry season, was not proper but 0. latifolia could still be collected in two different types of habitats: a) road-side ditch adjoining a rice field, irrigation canal and rice fields; b) forest gallery along river. In the first environment, the plants were not larger than diploid 0. punctata in Africa. It was the first time to my knowledge that 0. latifolia was reported to occur as a weed in rice fields. In the second environment, the plants were damaged by grazing and did not appear to be in the peak of their growing season, more likely to be during the rainy season.
In Paraguay, the populations documented with herbarium vouchers around the lake Ipacaray (in particular one 0. alta specimen, with big spikelets, 8.5 X 2.2mm, not dated but obviously an historically early collection) could not be found. The environment around the lake Ipacaray has been profoundly modified since the early collections, made near Asuncion since the mid 19th century. 0. latifolia was found however in the humid Chaco as described by Bruchner (1977). The main population I found was in a small pool surrounded by shrubs. I saw other populations in the same area along the road. The plants were about 1.5 to 2m high, with smaller leaves than their counterparts in the Amazon Basin. They flowered abundantly and were normally fertile. No well developed rhizomes could be found. It is, however, very likely that I collected the form described by Bruchner (1977).
It should be stressed that the habitat of 0. latifolia in the humid Chaco cannot be considered, as Bruchner (1977) did, genuinely wild. In particular, cattle were introduced to the area during the European colonization and account for the main cash production in the area. This has doubtless had a profound impact on the local vegetation.
The chromosome numbers of a few representatives of each population collected were observed in aceto-carmine stain under 1250X magnification. As judged from comparison with known diploid and tetraploid checks, they were all found to be tetraploids. This tetraploid nature was confirmed by the observation of fixed heterozygosity for isozymes of shikimate dehydrogenase and endopeptidase.
Most botanists and rice specialists consider the 0. latifolia complex to be part of the natural ecosystem in tropical America. However, earlier isozyme data indicated a close relationship of this complex with its Asian counterparts. As the geographic pattern of variation of the Oryza at molecular level in the Old World strongly suggest that very long distance dispersal of rice does not occur naturally, I concluded contrary to the common belief that no Oryza genus were present in America before the human (hence probably European) colonization (Second 1984). in the absence of fossil evidence, it is very difficult to definitively resolve this dilemma although the molecular data accumulating in various laboratories as well as results of wide crosses should allow in the near future a reevaluation of the available evidence.
While providing new material for these studies, the present collection and observations add new elements to the discussion: 1) The identification of a diploid form of the 0. latifolia complex (possible D genome) appears very unlikely; 2) the 0. latifolia complex also occurs as a weed in rice fields; 3) not all forms in the 0. latifolia complex are gigantic (character which distinguish them most at morphological level from their Old World relatives); 4) wild species of rice in Latin America occur in habitats that cannot be characterized as necessarily representative of the original ecosystems before the European colonization (cattle grazed areas, rice fields) or in naturally disturbed environments like along rivers with moving shores; 5) there is no strong argument against a hypothesis of a naturalization of these species within the last three centuries, and 6) particular attention should be given to the 0. alta forms from the Amazon Basin which bridge the Sativa and the Latifolia group of species at morphological level.
Considering the high diversity of this species complex, the hypothesis of its rapid speciation and differentiation would represent an interesting biological phenomenon with potential practical outconies for breeding and thus a valuable line of research.
I thank Ing. Agr. Jorge E. Rodas for kindly guiding me in Paraguay and Dr. Nora Lapitan for her help in cytogenetic observations.
References
Brucher, E., 1977. Morpho-genetic study of a perennial wild rice (Oryza sp.) from the Paraguayan Chaco. SABRAO J. 9(2): 86-90.
Second, G., 1984. A new insight into the genome differentiation in Oryza L. through isozymic studies. In A. K. Sharma and A. Sharma (eds.), Advances in chromosomes and cell genetics, p. 45-78. Oxford and IBH Pub]. Co., New Delhi.