|dc.description.abstract||In New Zealand, prairie grass has attracted attention because of its ability to grow well into the autumn and winter, and a number of physiological studies on the growth of this plant have been reported (karim, 1961; Langer, 1962; Dougherty, 1962). However, strains available in this country are largely unimproved and display some agriculturally undesirable properties, including lack of persistency under heavy grazing and a tendency to early flowering (Syme, 1943; Crawford, 1960).
Two improved strains, Pergamino “Martin Fierro” and Angel Gallardo, have been produced by mass selection techniques in Argentina, while a strain, Lamont, has been selected in the U.S.A. as resistant to the head smut Ustilago bullats (Johnson, 1957). In Australasia, the only attempt at improvement of prairie grass was that of Donald(1939) who selected five distinct lines, from which it was proposed to build up a strain for commercial use. However, there are no further reports on the continuance of this project.
Prairie grass presents a number of special problems to the plant breeder. The cleistogamous flowers are extremely small and early maturing, a factor likely to lead to complications in any crossing programme. There are conflicting reports on the chromosome number of the species and, even where this has not been in doubt, inter-strain hybrids have, in some cases, shown considerable steritlity (Stebbins, 1949). Little is known of the variability present in agronomically important characters, between and within strains, or of the likely response to selection for any of such characters. In this respect, the amount of cross-pollination in a population will have an important bearing on the degree of heterozygosity present in single plants, and hence on the number of generations over which response to selection could be expected to continue.
In the present investigation, a study is made of the environmental factors affecting the occurrence of cleistogamy and chasmogamy in prairie grass, with a view to obtaining a means of inducing the production of chaasmogamous flowers for crossing purposes. The numbers and gross features of the chromosomes in several strains are examined. A detailed analysis is made of phenotypic variation in agronomically important characters between and within strains, and response to selection for two characters in two strains is assessed.
The term ‘strain’ is used, throughout this thesis, in the widest sense, to denote groups of plants of distinctive origin. Some of these are ‘cultivars’ in the true sense of the word, and some are not.||en