Divergent Selection for Hydrocyanic Acid Potential in Sudangrass

Alteration in hydrocyanic acid potential (HCN-p) of the forage is one of the important objectives in sudangrass [Sorghum sudanense (Piper) Stapf] breeding programs. The effectiveness of recurrent phenotypic selection for increasing or decreasing HCN-p in sudangrass was evaluated in two cycles of individual plant selection in the cultivar 'Greenleaf. In cycle 1, HCNp means of the high and low populations -were higher and lower, respectively, than for Greenleaf, but only the low-HCN-p population was significantly different from Greenleaf. In cycle 2, mean HCN-p values of both populations differed significantly from Greenleaf. The average realized heritability for the two cycles was 0.40 while broad-sense heritability estimates averaged 0.86. After two cycles of selection, the low and highHCN-p populations differed from Greenleaf by about 17 and 30%, respectively. Additional index words: Prussic acid, Dhurrin, Forage quality, Heritability, Sorghum sudanense. S [Sorghum sudanense (Piper) Stapf] and sorghum-sudangrass hybrids are used extensively to provide supplementary feed to animals as pasture or greenchop. Precautions in managing the crop are necessary to prevent animal losses due to prussic acid (hydrocyanic acid) poisoning. All known sudangrasses and sorghums [S. bicolor (L.) Moench] contain dhurrin [(S)-phydroxymandelonitrile /3-D-glucopyranoside] which yields hydrocyanic acid when hydrolyzed enzymatically in disrupted plant tissues or in the rumen of consuming animals. Breeding sudangrasses with lowered hydrocyanic acid potential (HCN-p) would reduce the danger of hydrocyanic acid poisoning, and permit greater flexibility in the management of this crop. Also, sudangrasses with reduced HCN-p would be useful in the development of sorghum-sudangrass hybrids with lowered HCN-p. 'Contribution from USDA-ARS and the Nebraska Agric. Exp. Stn., Lincoln. Published as Paper No. 6173, Journal Series, Nebraska Agric. Exp. Stn. Received 18 Feb. 1981. The work reported was conducted under Nebraska Agric. Exp. Stn. Projects 12-088 and 12-114. 'Supervisory research geneticist, USDA-ARS; George Holmes professor of Agronomy, Univ. of Nebraska; and research agronomist, USDA-ARS, Lincoln, NE 68583, respectively. GORZ ET AL.: HYDROCYANIC ACID POTENTIAL IN SUDANGRASS 323 Breeding for lower HCN-p in sudangrass was facilitated by the recent development of a simple, rapid, nondestructive spectrophotometric procedure (6). This procedure involves the assay of individual first leaves from 7-day-old seedlings grown under controlled conditions. Following completion of the assay, selected seedlings can be transplanted to the field or greenhouse for the production of self or cross-pollinated seed for use in the next cycle of selection. In a review of studies of the inheritance of cyanogenesis in sudangrass and sorghum, Nass (9) reported that dominant or partially dominant factors were involved in the genetic control of both high and low HCN-p. Most studies suggested multigenic inheritance although one or two major genes also were hypothesized. Hogg and Ahlgren (8) evaluated 175 inbred lines ranging from low to high HCN content during a 3-year period and reported that HCN content of the lines was stable over years. They also reported that low-HCN strains could be developed by crossing low-HCN inbred lines. Barnett and Caviness (1) reported broad-sense heritability estimates of 0.41 and 0.68 for HCN production in populations derived from two sorghum × sudangrass crosses. Sorghum is classified as a predominately self-pollinated crop with outcrossing averaging 6% (10). The outcrossing percentage of sudangrass is usually higher, but reported values vary quite widely. Garber and Atwood (5) observed 76.4, 18.2, and 34.4% cross-pollination in Pennsylvania for the years 1941, 1942, and 1943, respectively, while Hogg and Ahlgren (8) reported in 1943 that cross-pollination in Wisconsin ranged from 4.5 to 10%. Thus, extensive cross-pollination is possible, but the percentage that occurs in a specific seed field is apparently dependent upon the location, environmental factors, and the type of sudangrass being grown. Recent work by Foster et al. (4) demonstrated positive results for bi-directional mass selection in a grain sorghum population. Thus, selection based on population improvement techniques, developed for use with cross-pollinated crops, may be successfully used with crops such as sorghum that have a Very low percentage of cross pollination. ’Greenleaf’ sudangrass is used extensively as the male parent in commercial production of sorghum-sudangrass hybrids (7). In 1977, a program of recurrent selection for low and high HCN-p in Greenleaf sudangrass was initiated at the Nebraska Station. The objectives of this research were to evaluate the effectiveness of recurrent selection for increasing or decreasing HCN-p in sudangrass populations, and to develop a reselected Greenleaf strain with lower HCN-p than the parent. MATERIALS AND METHODS All HCN-p values in this study were obtained by use of the spectrophotometric procedure (6). In this procedure, first-leaf samples from week-old seedlings were weighed, and dhurrin was extracted and hydrolyzed by autoclaving the samples in water. Aliquots of the extracts were then diluted in base, and absorbance was read at 330 nm, the absorption maximum ofp-hydroxybenzaldehyde. HCN-p values were derived from the 330 nm absorbance (A3~o) readings by simple calculations (6). The initial population consisted of 184 seedlings of the cultivar Greenleaf grown from Kansas foundation seed. From this initial population, 19, 17, and 17 seedlings were selected to represent the lowest, intermediate, and highest HCN-p levels in further studies. The selected seedlings were transplanted to the field, heads were bagged prior to anthesis, and selfed seed was harvested. Five plants from each HCN-p group with adequate supplies of selfed seed were selected for further study. Seven replications of the seed from the 15 selected plants were planted for assay of HCN-p. A total of 1405 S, seedlings were assayed with a range of 81 to 110 per line. This phase of the study will be referred to as cycle 0. For the next step, cycle 1, only those seedlings with the highest and lowest levels of HCN-p were selected from the total of 1405 S~ seedlings obtained from the 15 original plants. Selections were made across all replications and without regard to the original HCN-p group (high, medium, or low) from which the seedling originated. Thus, the 26 seedlings selected for lowHCN-p included 14, 9, and 3 seedlings from the low, medium, and high-HCN-p groups, respectively. Similarly, the 23 seedlings selected for high-HCN-p included 9, 5, and 9 seedlings from the low, medium, and high-HCN-p groups, respectively. The low and high-HCN-p populations were transplanted to separate field isolations in 1978, with plants on 61-cm centers. Open-pollinated seed was harvested from individual plants. Three replications of 10 seedlings each from each line (i. e., from each parent plant) were assayed for HCN-p. Greenleaf was included in each replication as a control. Plants for the next cycle of selection (cycle 2) were selected from individual seedlings of the 8 lines whose mean HCN-p was lowest among the 26 low-HCN-p lines, and from the highest 8 lines among the 23 high-HCN-p lines. The selected low and high-HCN-p populations included 54 and 56 plants, respectively. Each population was transplanted to a separate field isolation in 1979, with plants randomized on 107-cm centers. A severe infestation of chinch bugs [Blissus leucopterus leucopterus (Say)] destroyed some plants in the low-HCN-p isolation; seed was produced on only 33 of these plants. Open-pollinated seed was harvested from individual plants, and three replications of 10 seedlings from each plant that produced sufficient seed (31 plants from the low-HCN-p isolation and 55 from the highHCN-p isolation) were assayed for HCN-p. Greenleaf was included in each replication as a check. The breeding procedures used and the time period covered by each cycle are summarized in Table 1. A narrow-sense heritability estimate was obtained by the regression of offspring produced by selfing on their parents in cycle 0. In this situation, bop, the regression of offspring on parents, was used as the heritability estimate (Hb). Realized heritability estimates (Hr) (3) were also determine__d for c_.ycles0, and 2 using the following equation: Hr = (Xoh Xo~)/(Xph Xp~) where the X’s are means and the subscripts, o, p, h, and are offspring, parent, high and low, respectively. Variance components from the analyses of the replicated tests of the progeny of each cycle were used to calculate the ratio: Hv = a2g/a~g+ a~e)where a2gand a~e are the genetic and environmental components of variance, respectively. This ratio was difficult to interpret for this study because it was unknown whether the parents were homozygous or heterozygous individuals or whether the progenies were the products of crossing, selfing, or both. It is an estimate, however, of the variance among lines that is due to genotypic differences and as such can be used as a broad-sense heritability estimate. Gain from selection (Gs) was calculated both as a deviation from Greenleaf (Gsd) and as a deviation percentage from Greenleaf (G,p) for cycles 1 and 2. The equations used were as follows: (a) G~ for low HCN-p (b) G,~ for high HCN-p = (Xo~ Xg). (c) Gsp for low HCN-p = (X_~ Xo_0/X~ × 100. (d) G,p for high HCN-p = (Xoh Xs)/Xg × 100. 324 CROP SCIENCE, VOL. 22, MARCH-APRIL 1982 Table 1. Means, standard deviations, and ranges for HCN-p of parents and their progenies for two cycles of divergent selection for HCN-p in Greenleaf sudangrass.