Efficient tissue culture and regeneration methods were established using immature wheat embryos as expiants. Genotype differences in culturability were evident, and from the ten accessions most amenable to culture, a total of 2,846 plants were regenerated. Extensive somaclonal variation for morphological and biochemical traits was observed among 142 regenerants of a Mexican breeding line, 'Yaqui 50E', and their progeny. Variant characters included height, awns, tiller number, grain colour, heading date, waxiness, glume colour, gliadin proteins and α-amylase regulation. The variant characters were heritable through two seed generations and included traits under both simple and quantitative genetic control. Segregation data suggested that mutations both from dominance to recessiveness, and from recessiveness to dominance, had occurred. Most mutations in the primary regenerants were in the heterozygous state but some were true-breeding and presumed to be homozygous. Chromosome loss or addition did not account for the variation and none of the variant phenotypes was observed in over 400 plants from the parental seed source. The distinctive parental gliadin pattern was maintained in the somaclones thus excluding seed contamination or cross-pollination as a source of the variation.

The objectives of this study were to select and initially characterize mutants of soybean (Gleine max L. Meff. cv Williams) with decreased ability to reduce nitrate. Selection involved a chlorate screen of approximately 12,000 seedlings (progeny of mutagenized seed) and subsequent analyses for low nitrate reductase (LNR) activity. Three lines, designated LNR-2, LNR-3, and LNR-4, were selected by this procedure.In growth chamber studies, the fully expanded first trifoliolate leaf from N03-grown LNR-2, LNR-3, and LNR-4 plants had approximately 50% of the wild-type NR activity. Leaves from urea-grown LNR-2, LNR-3, and LNR4 plants had no NR activity while leaves from comparable wild-type plants had considerable activity, the latter activity does not require the presence of NOs-in the nutrient solution for induction and on this basis is tentatively considered as a constitutive enzyme. Summation of constitutive (urea-grown wild-type plants) and inducible (N03-grown LNR-2, LNR-3, or LNR-4 plants) leaf NR activities approximated activity in leaves of NO3--grown wild-type plants. Root -(17, 25). Chlorate also competes with N03 for uptake in Hordeum vulgare and Arabidopsis thaliana roots (5, 6). Thus, C103-can act as an analog of NO3 during uptake, as well as during reduction by NR. The analog characteristic of C103 and the resultant toxicity symptoms associated with C103-reduction suggested that this ion could be used as a screening agent to identify organisms with decreased N03 uptake and metabolism. Although mechanisms of C103 toxicity other than that involving NR are possible (14,19,27), Cl03 resistant mutants which are NR deficient have been found in higher plants (7, 22,26) and plant cell lines (14,18,19). Soybeans (Glycine max L. Merr.) have been shown to display characteristic C103 toxicity symptoms (browning and stunting of leaves) (9, 28), and this was the basis for using a C103 screen to identify plants defective in N03 metabolism. This paper reports on the selection, from progeny of mutagenized seed, and initial characterization of three soybean lines which have decreased leaf NR activity compared with wild-type plants. MATERIALS AND METHODS

A total of 256 selected lines derived from tissue culture of three hexaploid wheat cultivars were grown in a replicated hill plot experiment to examine somaclonal variation in agronomic characters. The lines were derived by single plant selection for various characters from a total of 100 regenerants, and were either SC3 or SC4 generation in the test. Significant variation was found in all the characters measured: height, grain number per spike, kernel weight, yield, total dry weight and harvest index. In most cases, variation could be identified which was both less than and greater than the parental controls. However, there was also an apparent effect of the parent cultivar on the total amount and direction of the variation. For two cultivars, lines could be traced back through the culture phase to individual explant embryos. Many of the original embryos contributed significant variation, and most characters showed significant variation arising from more than one embryo. In the following year, 32 lines selected from the hill plot experiment were grown in larger replicated plots and yield, harvest index and a number of grain and baking quality characters were measured. No lines selected for high yield or harvest index maintained significant improvements over their parental controls. However, significant variation was displayed for many of the quality characters examined. Significant increases in kernel weight, hardness and protein content, and a significant reduction in yellow pigmentation represented potentially useful improvements. Only unfavourable variation was seen in flour yield and in mixograph height, time and breakdown.