A description for genotypic stability was developed based on the concept of a genotypic stability space. Environments were considered the measurement criteria. The coordinate position for a genotype was determined by the deviations of expected yield of the genotype from its yield if stable. An expression for the response of a stable genotype was developed. The relative genotypic stability measure was defined as the distance of a genotype from the center of the arrangement and presented as a measure of homeostasis. The comparative genotypic stability measure was defined as the distance between the positions of two genotypes and designed to measure the similarity of stability responses. The development permitted the testing of hypotheses with respect to stability concepts. A set of regional soybean test data was evaluated. The results for relative stability and the regression approach were comparable. Limitations for the regression approach were noted. Genotypes analyzed had different homeostatic properties. The genotypic responses to environmental stimuli appeared to be unique to genotypes and difficult to predict.

No abstract

Laboratory techniques for screening soybean (Glycine max (L.) Merrill) genotypes for tolerance to Al would facilitate breeding programs for areas where AI toxicity limits soybean production. Genotypes from a broad‐base soybean population were screened For Al tolerance based on root growth rate of 5‐day‐old seedlings grown in solutions containing 0.4 ppm Al and 10 ppm Ca (pH 4.6). The selections were intermated to produce a second cycle. Three cycles of divergent selection were completed. The changes associated with selection were evaluated in laboratory studies and in greenhouse studies using Lynchburg sandy loam soil (Aeric Paleaquult). Tolerance to Al as measured by root growth rates proved to be heritable (0.67) with responses in each cycle of selection; however selections completely tolerant to 0.4 ppm Al in solution were not obtained. When high quality seed was used, genetic differences for tolerance were not affected by the environment used for seed production. Selection based on seedlings from individual plants was feasible. Genotypes selected for Al tolerance had less reduction in growth in soils with high exchangeable Al saturation than non‐tolerant selections; however, the differences were not as distinct as that found for the seedling root‐growth studies. Divergent selections had similar levels of metabolic intermediates in the root tips when seedlings were grown in 0 ppm Al, but pyruvate and ATP levels increased significantly under Al stress for the tolerant but not the non‐tolerant selections. Selection for Al tolerance based on growth of 5‐day‐old soybean seedlings in 0.4 ppm Al apparently identified genetic differences for tolerance in established plants as well as tolerance which appeared unique to the seedling stage.

No abstract

U NDER natural conditions soybeans are largely selfpollinated with pollination occurring in the advanced bud stage before the flower opens. Natural hybridization between plants in adjacent rows was observed to be 0.04% by Woodworth (6) and 0.07 and 0.18%, respectively, in successive years by Garber and Odland (4). Takagi (5) obtained 0.62% natural hybrids when different genotypes were grown in close contact. With alternate 3-to 4-inch spacing of different varieties within the row, Cutler (2) observed 1.10% hybrids with a range of 0.38 to 2.43% crossed seed in individual genotypes. In general, natural crossing between soybean plants in adjacent rows appears to be something less than 0.5% and about 1.0% for plants grown in close contact. The percentages of natural crossing in soybeans reported in the literature cited is that actually observed. A number of the authors stated that the percentage of natural hybridization would be doubled if reciprocal crossing occurred. Percent hybridization would be doubled only when the probabilities were equal that an outcross gamete carried 1 Joint contribution from the Crops Research Division, ARS, USDA, as No.