Seed Weight Influence on Seedling Hydrocyanic Acid Potential in Sorghum

Grain sorghum [Sorghum bicolor (L.) Moench) typically produces larger seeds than sudangrass [S. bicolor, formerly S. sudanense (Piper) Stapf]; and grain sorghum seedlings are higher in hydrocyanic acid potential (HCN-p) than sudangrass seedlings. Previous studies have shown a seed-parent effect on seed weight and HCN-p in reciprocal F, hybrids of sorghum X sudangrass. This study was conducted to determine whether the seed-parent effect on HCN-p could be attributed primarily to the difference in seed size between reciprocal sorghum X sudangrass hybrids. Large and small seeds of lowHCN-p sudangrass, high-HCN-p sorghum, and their reciprocal hybrids, were visually selected, individually weighed, and planted in growth chambers. Height, fresh weight, and HCN-p of the resulting 7-d-old shoots were measured. On average, the selected large seeds weighed ~1.6 times as much as the small seeds. Shoots from the large seeds were ~1.2 times as tall and 1.4 times as heavy as those from small seeds, and they contained =1.3 times as much HCN per shoot as did shoots from small seeds. However, the HCN-p (mg kg~' fresh wt.) of the shoots from large seeds was only slightly greater than that of shoots from small seeds (709 and 701 for first leaves of shoots, and 278 and 267 for shoot remainders from large and small seeds, respectively). Shoots from 5-d-old light-grown and etiolated seedlings also failed to show appreciable differences in HCNp due to seed size. It was concluded that the seed-parent effect on shoot HCN-p in crosses of sorghum and sudangrass was not caused primarily by the seed-parent effect on seed weight. I A RECENT study of the inheritance of seedling hydrocyanic acid potential in crosses between highHCN-p sorghum and low-HCN-p sudangrass, results indicated a maternal or seed-parent effect for seed weight and seedling HCN-p in the Fj and backcross generations (Lamb et al., 1987); however, no evidence of this reciprocal effect was found in the F2, suggesting that cytoplasmic inheritance is not involved. High, positive correlations (P < 0.01) between seed weight and seedling HCN-p were found for all entries taken together (r = 0.85**; significant at p < 0.01), for all parental lines (r = 0.82**), for all F, populations (r = 0.69**), and for all backcross populations (r = J.F.S. Lamb and F.A. Haskins, Dep. of Agronomy; H.J. Gorz and K.P. Vogel, USDA-ARS and Dep. of Agronomy, Nebraska Agric. Res. Div., Lincoln, NE 68583. Published as Paper no. 9219, Journal Series, Nebraska Agric. Res. Div. Research was conducted under Project no. 12-114. Received 9 July 1990. *Corresponding author. Published in Crop Sci. 31:1014-1016 (1991). Abbreviation: HCN-p, hydrocyanic acid potential. 0.85**) (Lamb et al., 1987); however, correlations between seed weight and seedling HCN-p for individual entries, the pooled F2's, or within types of seed parent in the F! or backcross generations were generally nonsignificant. Thus, seed weight per se appeared not to have a large effect on seedling HCN-p. The foregoing report (Lamb et al., 1987) dealt only with seed weight and seedling HCN-p, and values for seed weight were means calculated from the number and total weight of seeds planted in each replicate. Weights of individual seeds giving rise to the sampled seedlings were not known. The primary objective of the present study was to investigate in greater detail the possible effect of seed weight on seedling HCN-p and other seedling traits within sorghum and Sudangrass parental lines and reciprocal F! sorghum X sudangrass hybrids. Backcross and F2 populations were excluded because the genetic diversity in such populations was undesirable for this study. Individual seeds were weighed, and measurements were made of the fresh weight, height, and HCN-p of individual shoots arising from identified seeds. A brief study also was made that used seedlings grown in water-saturated vermiculite. The rationale for this study was that in the absence of exogenous nutrients, seed-size dependent differences in seedling traits might be especially distinct. MATERIALS AND METHODS Seven-Day, Light-Grown Seedlings Two low-HCN-p sudangrass lines, 1901 and 1904; and four high-HCN-p sorghum lines, ACK60, BCK60, ARedlan, and BRedlan, were used as parental lines in this study. Reciprocal sorghum X sudangrass hybrids included the following: 1901 X BCK60, ACK60 X 1901, 1901 X BRedlan, ARedlan X 1901, 1904 X BCK60, ACK60 X 1904, 1904 X BRedlan, and ARedlan X 1904. Further descriptions of the parental lines and the crosses used to produce the F! hybrids are given by Lamb et al. (1987). Large and small seeds of each entry were selected visually. For each replicate, 10 large and 10 small seeds of each entry were individually weighed. The 10 large seeds of a given entry were planted in identified positions in a single row in a small plastic pan, and 10 small seeds of the same entry were planted in an adjacent row. The 14 entries (14 pairs of rows) were randomly arranged in each replicate of the experiment, and eight separate plantings provided a total of eight replicates. Seedlings were grown in a soil mixture under continuous cool-white fluorescent light (« 150 jimol m~ s-) at 27° C, essentially as described by Gorz et al. (1977). LAMB ET AL.: SEED WEIGHT INFLUENCE ON HYDROCYANIC ACID IN SORGHUM 1015 At 7 d after planting, three representative seedlings were chosen for assay from each of the 28 rows. Seedling height was measured, and the first leaf of each sampled seedling was then harvested separately from the shoot remainder, which was excised at the soil surface. Both shoot portions were weighed, extracted, and assayed as described by Gorz et al. (1977). Analysis of variance for a randomized complete-block design was used for all traits measured. The error cross products and sums of squares matrix from this analysis were used to calculate partial correlation coei~cients for seedlings within entries for all traits. Phenotypic correlations between traits across all entries were determined by using mean squares and mean products from the analyses of variance and covariance as described by Falconer (1981). Five-Day, Lightand Dark-Grown Seedlings Parental lines 1904 and BRedlan and the crosses 1904 X BRedlan and ARedlan X 1904 were used. Two sets of 20 small and 20 large seeds (individually weighed) of each parent and F1 hybrid were planted in water-saturated vermiculite. One set was placed in continuous light at 27° C as described above, and the other in a light-excluding container in the same growth chamber. At 5 d, seedlings were harvested and each was separated into shoot, root, and seed remnant. The portions were bulked within entry, light treatment, seed size, and type of portion. The bulked samples were weighed, extracted, and assayed for HCN-p by the spectrophotometric procedure of Gorz et al. (1977). Also, the absorbance spectrum of each base-diluted extract was scanned from 400 to 220 rim. RESULTS AND DISCUSSION Seven-Day, Light-Grown Seedlings Results were similar within each of the parental types and also within each type of F~ hybrid; therefore, data were pooled within each of these four types of entry as shown in Table 1. Means for both first-leaf and shoot-remainder HCN-p showed the same sort of seed-parent influence as that previously observed by Lamb et al. (1987); that is, values for sorghum × sudangrass hybrids were higher than those for sudangrass × sorghum hybrids. This was true of seedlings from both small and large seeds. Results were consistent within entries and across all entries for seed size differences (Table 1). Seedlings from large seeds were significantly taller and had significantly greater first-leaf and shoot-remainder weights than those from small seeds. There were no significant differences due to seed size within entries or across all entries for firstleaf HCN-p or shoot-remainder I-ICN-p, except for shoot-remainder I-ICN-p of sudangrass × sorghum seedlings, and that difference was small (12 mg kg -~ fresh wt.). Hydrocyanic acid per shoot is a product of shoot weight and I-ICN-p, and the apparent effect of seed size on I-ICN per shoot was primarily a result of the effect on shoot weight rather than on HCN-p. Within-entry and overall phenotypic correlations for all pairs of traits are Shown in Table 2. The withinentry values indicated significant association of seed Table 1. Influence of seed size on various traits of 7-d-old Hght-grown seedlings of Iow-HCN-p (hydrocyanic acid-potential) sudangrasses, high-HCN-p sorghums, and their reciprocal Fz hybrids.1. Sudangrass Sorghum Sudangrass × Sorghum × Trait Seed size parents parents sorghum sudangrass Over all entries Seed weight, Small 7.1 + 0.1"* 18.9 :t: 0.4** 8.2 + 0.1"* 17.7 + 0.3** 13.8 + 0.3** mg seed -I Large 12.5 + 0.1 27.6 + 0.5 14.7 + 0.1 27.9 _+ 0.2 21.8 + 0.4 Seedling heisht, Small 111 + 2** 87 ± 1"* 113 ± 1"* 134 + 2** 112 + 1"* ram Large 133 + 4 97 + 1 141 + I 150 ± 2 130 + 1 First-leaf weight, Small 13.5 ± 0.4** 12.5 _+ 0.3** 14.8 + 0.2** 16.5 + 0.3** 14.5 + 0.2** ms fr wt. Large 16.6 + 0.5 13.6 + 0.2 19.0 + 0.3 19.2 + 0.3 17.2 + 0.2 First-leafHCN-p, Small 221 + 1