Contamination in Denta Sweetclover During Successive Generations of Seed Increase

Contamination levels observed in the breeder, foundation, and certified seed generations of 'Denta' sweetdover (Melilotus alba Desr.) were 0.71, 1.69, and 3.62%, respectively, as measured by the frequency of plants high in content of o-hydroxycinnamic acid. Based on a sample of plants tested simultaneously for o-hydroxycinnamic acid content and ,B-glucosidase activity, a small but consistent decrease in tl;J.e frequency of cu plants occurred with succeeding seed generations, while shifts in the frequency of b plants were not consistent in magnitude or direction. Contamination through cross-pollination with CuCuBB plants may be the major factor responsible for the observed decreases in the frequency of cu plants. Additional index words: Seed contamination, Genetic shift, o-hydroxycinnamic acid. S RIFTS in gene frequencies during the seed produc. tion sequence can drastically alter the genetic purIty of crop varieties. In improved varieties of forage crops, such shifts are most common when the varieties are grown in specialized seed producing areas outside the normal areas of adaptation. Marked changes in l?cation of seed production or in management practIces, when superimposed on the high degree of genetic variability and cross-pollinating habit of most forage crops, encourage maximum genetic shifts. Changes in varietal characteristics have been found in a.Hal.fa (Medicago sativa L.) (3, 14, 15), red clover (Trzfolzum pratense L.) (2, 18, 19), Ladino clover (T. repens L.) (10, 11, 16, 17), and sweetclover (Melilotus a!ba Desr.) (4) when different seed sources or generatlOns were compared. Most of these studies have shown that varietal changes are due primarily to differential con.tributions by the clones or genotypes making up a v.anety to the total seed produced at a specified locatlOn or under a specific management system. Many of the stu~i~s have relied upon quantitatively inherited characte!lstlcs such as yield, height, maturity, and winter surviVal for detecting genetic shifts. Useful data have been obtained, but precise measures of genetic shifts are difficult in such studies because effects of in~ividual genes cannot be observed in quantitative traits. In most forage crops, a scarcity of reliable genetic markers has made precise measurements of genetic shifts impossible. However, simply inherited and easily identifiable marker genes have been used for study of gene frequency changes in Ladino clover and sweetclover. Portz and Jackobs (13) studied the frequency of cyanophoric plants from 235 seedlots of cer1 Contribution from the Crops Research Division, Agricultural Research Ser,:ice, U. S. Department of Agriculture, and the Nebraska AgrIcultural Experiment Station, Lincoln 68503. Supported in part by the National Science Foundation (Grant No. GB-1l48). Published wi.th the approval of the Director as Paper No. 2475, Journal SerIes, Nebraska Agr. Exp. Sta. Received Dec. 20, 1968. 2 Research Geneticist, Crops Research Division, Agricultural Research Service, U. S. Department of Agriculture, and Bert Rodgers Professor of Agronomy, University of Nebraska. 367 tified Ladino clover produced in four western states. The frequency of plants containing both the cyanoglucoside and enzyme varied from 0 to 43% between seed lots obtained from one state, and the occurrence of plants with the cyanoglucoside alone varied from 0 to 67%. Three possible reasons for these wide differences were suggested: (a) differences in the original seed lots, (b) genetic shift due to natural selection, and (c) outcrossing to other white clovers. Alternatives (a) and (c) were eliminated by Stanford et al. (16) in an investigation of gene frequency changes in Pilgrim Ladino clover. Differences in date of first flower and intensity of flowering were attributed to genetic shift, but a study of the genes producing the cyanoglucoside, its hydrolyzing enzyme, red flecking on the leaf, and y-leaf marking showed no strongly consistent changes m gene frequency with advance in generation. Goplen and Weber (4), using a recessive genetic marker for low-coumarin content, determined contamination levels in Cumino sweetclover through four generations of seed increase. Mean contamination levels of 0.27%,1.11%,4.21% and 19.80% were found for breeder, foundation, registered, and certified seed, respectively. Cross-pollination with pollen from highcoumarin plants in nearby areas was suspected as the major factor in the rapid increase in contamination, although the lower inherent seedling vigor and seed yield of Cumino compared to the contaminants may have been responsible. Contamination by volunteer plants, admixtures of seed, or preferential fertilization was not significant. . T~is I?aper deals wit~ changes in the phenotypic dIstnbutlOn of plants dIffering with respect to the C u / cu ~nd B / b . allelic pairs in successive generations of seed mcrease m Denta sweetclover, a variety low in ?-hydroxycinnamic acid content. The Cu/cu alleles mf~uence the co?tent of o-hydroxycinnamic acid glucosides and are mdependent of the B / b alleles which det~r.mine the presence or absence of ,G-glucosidase actIvIty. Plants of the CuCu genotype are high in content ?f o-hydroxycinnamic acid glucosides, and prepar.at.lOns of the BB genotype possess ,G-glucosidase actIVIty. The Cu and B genes are both lacking in dominance (7, 9). MATERIALS AND METHODS Breeder seed of 'Denta', produced at the Wisconsin Agricultural Experiment Station in Madison, Wisconsin, constituted the initial stock in these investigations. The five lots each of foundation and certified seed were produced in North and South Dakota, where sweetdover is a well-adapted legume. A 250to 500-g sample was obtained from each lot of seed.' Each ~a~ple of seed. was thor~ughly mixed, and a portion was .scarIfIed for plantu;g. Plantmgs were made in rows approxImately 5 cm apart m 35X 50-cm wooden flats containing • The authors express their appreciation to J. D. Colburn of the South Dakota Crop Improvement Association and R. L. Harrison and D. C. Ebeltoft of North Dakota State University for providing the foundation and certified seed samples used in this study. 368 CROP SCIENCE, VOL. 9, MAY-JUNE 1969 a mixture of composted soil and sand. Approximately 500 to 700 seeds were planted in each flat; six flats were planted from each seed lot. Plants were grown in a greenhouse for approximately 2 to 3 months under natural illumination supplemented with incandescent light to provide an 18-hour photoperiod. When most plants were at an early bud stage, a portion of a leaflet from each plant was tested qualitatively for a-hydroxycinnamic acid. The paper test employed was similar to that previously described for unheated samples (6), except that a drop of [:I-glucosidase preparation was added to each sample spot as soon as an entire sheet of samples had been taken. This procedure insured the hydrolysis of cis-a·hydroxycinnamic acid glucoside in those samples lacking endogenous [:I-glucosidase activity, and thus permitted the accurate classification of Cuand cucu plants regardless of constitution with respect to the B/b alleles. To promote the presence of a high proportion of cis-a-hydroxycinnamic acid glucoside in Cuplants, and thereby assure a clear distinction between Cuand cucu individuals, testing was done only on sunny days preceded by a full day of sunlight (8). In a portion of the plants from most seed lots, a qualitative assay for [:I-glucosidase activity was done simultaneously with the qualitative test for o-hydroxycinnamic acid, by means of a second sample spot taken from the same leaflet as the first sample. This procedure also was similar to that previously described for unheated samples (6), except that a drop of sweetclover leaf extract containing cis-o-hydroxycinnamic acid glucoside (approximately I mg/ml) was added to the sample spot. By this means, Bplants were readily distinguished from bb plants. RESULTS AND DISCUSSION There was a small but steady increase in contamination of Denta seed lots in succeeding generations as measured by the frequency of Cu plants (Table 1). The increase was slightly more than two-fold for both the foundation and certified seed classes, compared to the breeder and foundation classes, respectively. The range of contamination in the lots of certified seed is well within the reported tolerance limits of 10% for cattle (5) and 25% for sheep (12). However, the values observed for the foundation and certified seed classes are considerably above the maximum levels allowed by the International Crop Improvement Association (I). Simultaneous testing of plants for o-hydroxycinnamic acid content and {i-glucosidase activity produced the data on phenotypic numbers presented in Table 2. These data permitted calculation of the phenotypic frequencies shown in Table 3. The chi-square test was used to determine whether or not the observed changes in frequency were significant. In applying this test the assumption was made that no shifts in phenotypic frequency should occur with suceeding generations. Thus, the frequencies observed in the breeder and foundation generations were used, before rounding to the extent shown in Table 3, to calculate expected numbers in the foundation and certified generations, respectively. These expected numbers were then compared with observed numbers in the foundation and certified generations. In this comparison, observed and expected numbers differed at the 1 % level of probability in every instance except that involving the observed numbers of Band b individuals in the certified generation (Table 4). Accordingly, the decrease in frequency of cu plants and the corresponding increase in the Cu phenotype, although small, appear to be significant. By comparison, changes in the frequencies of the band B phenotypes were less consistent, both in direction and magnitude. By calculating the products of appropriate pairs of the phenotypic frequencies in Table 3, the expected Table 1. Size of sample and degree of contamination in breed· er, foundation, and certified seed lots of Denta sweetclover.