<^^^ "K, Division of Agricultural Sciences UNIVERSITY OF CALIFORNIA ".ri;i^ ALTER DATES to Eacilitate Hybrid Seed Production JOHN D. ATKIN GLEN N. DAVIS Ofornia agricultural ■"~ % STATION BULLETV A GUIDE TO NEW ONION HYBRIDS Poor pollination, resulting when parental lines fail to flower at the same time, has been an obstacle to the production of new and desir- able onion crosses. The studies reported here show that flowering dates can be adjusted to coincide, thus insuring successful pollination and satisfactory seed yield. Flowering dates are influenced by both bulb storage temperatures and planting dates. If comparative flowering dates of different onion varieties are known, storage temperatures and planting dates can be adjusted to bring two parental lines into flower at approximately the same time. This bulletin reports on the comparative flowering dates of a number of different onion varieties. It also describes three experiments, con- ducted in 1950, 1951, and 1952, showing the effect of various bulb storage temperatures and planting dates on date of flowering, number of seed stalks produced, number of flowers per umbel, and seed yield. Early planting dates and storage at 50° F brought about an early flowering date and greatest seed yield for all varieties studied. Late plantings meant not only later flowering dates — they also resulted in fewer seed stalks, fewer flowers per umbel, and a smaller seed yield. If parental lines cannot be made to coincide, it is recommended that the later-flowering line be used as the female parent. THE AUTHORS: John D. Atkin, formerly Research Assistant, Department of Vegetable Crops, University of California, Davis, is Assistant Professor of Vegetable Crops, New York Agricultural Experiment Station, Geneva, Nev^ York. Glen N. Davis is Professor of Vegetable Crops and Olericulturist In the Experiment Station, Davis. NOVEMBER, 1954 [2] ALTERING ONION FLOWERING DATES TO FACILITATE HYBRID SEED PRODUCTION* JOHN D. ATKIN GLEN N. DAVIS The search for new hybrids It has been demonstrated that F^ hy- brid onions, superior to open-pollinated types, can be developed. Hybrids out- standing for size, color, uniformity of shape, and time of maturity are now available. F^ hybrids resistant to disease and insects that annually cause excessive losses are in the not too distant future. Many of the difficulties of producing hybrid onion seed have been overcome. The male-sterile character originally re- ported in the breeding line Italian Red 13-53 by Jones and Emsweller (1937) can be transferred to any desired line by a system of back-crossing to the male- sterile parent (Jones and Clarke, 1943). A system of breeding using the male- sterile character has made it possible to produce hybrid onion seed with very little costly hand labor. While large yields of seed can be ob- tained from certain parental combina- tions, the seed yield of many desirable crosses is so low that the production of seed in large commercial plantings is economically hazardous. Jones and Davis (1944) found that inbreeding onions was usually accompanied by a loss in vigor. This loss in vigor in the female parent is one reason for reduced seed yields. An- other is poor pollination, which results * Submitted for publication February 9, 1954. when the parental lines do not flower at the same time. If the flowering dates of one or both parents could be altered so flowering would coincide, a large in- crease in yield of seed in certain combi- nations might be expected. Earlier studies pointed the way Earlier studies indicated that flower- ing dates might be altered by different bulb storage temperatures and different dates of planting of the mother bulbs. They gave little information, however, concerning how much flowering dates can be altered. Heath and Holdsworth (1948) ad- vanced the theory that there are two hor- mone systems controlling bulbing and flowering in the onion plant. The two systems are diametrically opposed to each other so that if the environmental conditions are favorable for flowering, a mature onion bulb will send up a seed stalk and flowering will follow. However, if conditions are very favorable for bulb- ing, a mature bulb will grow vegetatively and several bulbs will be produced, while flowering will be suppressed. If conditions are not too unfavorable for flowering, flowering may only be de- layed. High temperatures during the growing season or during storage will destroy part or all of the flowering hor- mone so that flowering may be delayed or [3] entirely inhibited, depending upon the temperature and duration of the storage treatment. Medium bulb storage tempera- tures (50° F) are most favorable for flower formation, while low temperatures (32° F) inhibit flower bud formation. When short-day varieties of onions that are matured in Texas in late March or early April are planted at Greeley, Colorado, about July 15, each bulb will form a number of shoots and each shoot will produce a mature onion bulb in the fall (Jones, Perry, and Edmundson, 1949). The combination of high tem- perature and long days favors quick bulb- ing, and flowering is suppressed. Miller (1932) reported that mature onion bulbs stored at room temperature and planted under high temperature failed to produce seed stalks. They formed vegetative plants which later formed mature bulbs. Jones (1927), working with the variety Ebenezer, found that bulbs stored at 7.5° C (45.5° F) or at 11° to 12° C (51.8° to 53.6° F) produced plants that flowered earlier and gave larger yields of seed than plants grown from bulbs stored at higher or lower temperatures. Jones and Emsweller (1939) discov- ered that plants grown from bulbs stored at 46.5° or 53.5° F flowered earlier than plants from bulbs stored at higher or lower temperatures. The 53.5° F treat- ment resulted in a larger number of seed heads than any other treatment. When planting was delayed until February 15 or March 2, flowering was delayed and seed yields were lower. Reduced seed yields in late-flowering plants may be due in part to thrips injury and high summer temperatures. Thompson and Smith (1938) observed that relatively low temperatures are necessary for seed stalk development. Onion plants kept at 70° to 80° F did not form seed stalks. They also found that onion sets stored at 50° F gave a higher percentage of seed stalks in the field than did sets stored at higher or lower temperatures. Hartsema (1947), Blaauw, Hartsema and Luyten (1944), and Holdsworth (1927), working with sets, noted that storage temperatures above 20° C (68° F) are best for preventing subsequent flowering in the field. Flowering in com- mercial bulb crops grown from sets can also be largely prevented by delaying planting until late in the spring. REPORT FOR 1950: Excel, Southport Yellow Globe, Colorado Sweet Spanish, and Stockton G-36 varieties were planted to determine comparative flowering dates. Excel bulbs were studied to learn the effects of storage temperature and planting date on date of flowering. Only a limited supply of bulbs of the varieties Excel, Southport Yellow Globe, Colorado Sweet Spanish, and Stockton G-36 was available in 1950. Before the experiment began, the bulbs had all been held in common storage — that is, they were stored in the bulb storage house at the Davis Experiment Station. This house is insulated, but there is no artificial refrigeration or heat. Consequently, dur- ing hot weather the inside temperature is quite high ; as cooler weather approaches in the fall, the inside temperature also declines. With the Excel variety four storage treatments — common storage, 50°, 32°, and 68° F — and two dates of planting were used. Because of the limited num- ber of bulbs of the other varieties, only one storage treatment — common stor- [4] age and three dates of planting were used. Fifty bulbs were planted of each storage treatment and planting date. The limited supply of bulbs made it impos- sible to replicate the plantings or to include the desired number of treatments. The data were analyzed by comparing the means of two groups (Snedecor, 1946). Each umbel or each flowering date was used as a replication. While this is not the most desirable form of analysis, it did provide a measure of the significance of the difference. The average number of flowers per umbel was computed as follows: ten plants of each treatment were selected according to a chart of random numbers (Snedecor, 1946). If a plant to be used had more than one umbel, the umbels were numbered, and the umbel used was selected by use of the chart mentioned above. Actual flower counts were made after the umbels were in full bloom. The umbels of the Southport Yellow Globe plants were badly deformed, and many of the buds died so early that accurate counts were impossible. The mean dates of flowering for all experiments reported here were com- puted by using May 1 as the number one date. For example, a mean flowering date of 46 would be June 15. An umbel was considered to be in flower on the date the first flower opened. Table 1 . Effect of Storage Temperature and Planting Date on Flowering Date and Number of Flowers per Umbel in Four Varieties of Onions, 1950 Variety Planting date Bulb storage tem- perature (deg. F) Av. no, flowers per umbel Mean differ- ence L.S.D. at 5 per cent level Mean date of flower- ing* 33.0 41.0 27.0 32.5 32.5 43.8 23.4 32.8 28.3 30.2 38.8 24.4 28.9 34.7 Mean differ- ence L.S.D. at 5 per cent level Excel 2/3/50 3/4/50 2/3/50 3/4/50 2/3/50 3/4/50 2/3/50 3/4/50 32 32 50 50 68 68 c.s.f c.s. 210.9 94.8 166.8 85.3 187.9 96.6 166.9 99.2 116.1 81.5 91.3 67.7 82.2 26.0 65.1 54.0 8.0 5.5 11.3 9.4 3.5 1.9 2.4 1.9 Colorado Sweet Spanish . . 11/16/49 2/3/50 3/4/50 c.s. c.s. c.s. 331.7 187.4 115.9 144.3 71.5 141.0 n.s.t 1.9 8.6 1.9 3.2 Stockton G-36 11/16/49 2/3/50 3/4/50 c.s. c.s. c.s. c.s. c.s. c.s. c.s. 551.6 245.0 209.8 306.6 35.2 194.1 n.s.t 4.5 5.8 2.0 2.2 Southport Yellow Globe 11/16/49 2/3/50 3/4/50 4/14/50 .... .... 33.0 36.6 43.0 51.9 3.6 6.4 8.9 1.9 1.8 2.4 * Days after May 1. t Common storage. t Not significant. [5] Table 1 shows how storage treatments and dates of planting affected flowering dates in the 1950 experiment. While the Excel variety showed some differences in the number of flowers per umbel accord- ing to different storage treatments, the differences were not significant. On the other hand, within varieties, a delay in planting date resulted in a significant decrease in the number of flowers per umbel, except in the latest plantings of Stockton G-36 and Colorado Sweet Span- ish. In all cases within varieties, regard- less of storage temperature, a delay in the date of planting resulted in a signifi- cantly later date of flowering. Table 2 gives a detailed record of the flowering date of Excel bulbs from four storage treatments and two planting dates. Bulbs stored at 32° and 68° F pro- duced plants that flowered later than plants from bulbs stored either in com- mon storage or at 50° F. The results show also that the later planting resulted in a later flowering date. REPORT FOR 1951: Comparative flowering date studies were continued with the varieties Texas Early Grano, Excel, Colorado Sweet Spanish, and Southport Yellow Globe. Stockton G-36 bulbs were used in the ex- periments with varying bulb storage temperatures and planting dates. Direct seeding experiments were conducted with Texas Early Grano, Excel, and Yellow Bermuda. Comparative flowering dates A limited number of bulbs of the vari- eties Texas Early Grano, Excel, Colorado Sweet Spanish, and Southport Yellow Globe was available. These were all planted on the same date, following com- mon storage, to check their flowering dates. The mean dates of their flowering are given in table 3. When the bulbs were planted on the same date, there was a spread of nine days between the earliest and latest flowering varieties. If the com- parative flowering time of the different varieties is known, a proper adjustment of planting dates should bring two pa- rental lines into flower simultaneously. The effect of storage temperature on Stocicton G-36 In the fall of 1950 enough bulbs of the variety Stockton G-36 were available to plant three replications (50 bulbs per replication) of four different bulb stor- age treatments and four planting dates. These bulbs were placed in the different storages on August 3, 1950. Excessive rain during the winter months delayed planting operations until there was exces- sive spoilage of the bulbs being held in storage for the later plantings. As a re- sult, only a single planting, on Novem- ber 4, was made of G-36 bulbs. The results shown in table 4 indicate that the 50° F storage gives a much earlier date of flowering and a greater number of seed stalks per plant than the other treatments used. Seed-to-seed plantings Three replications of seed of the vari- eties Texas Early Grano, Excel, and Yel- low Bermuda were planted at 10-day in- tervals, beginning August 1 and ending September 30. This direct seeding ex- periment had a twofold objective: 1) to find how early seed should be planted in the fall to insure bolting; and 2) to [6] Table 2. Daily Record of Number of Excel Umbels Beginning to Flower, 1950 Date Feb. 3 planting March 4 planting Common En storage = F 32° F 68° F Common storage 50° F 32° F 68° F May 14. . . 1 15. . 16... 1 17. . . 3 18. . 4 19. . . 1 20 . 3 21. . 4 5 22. . 8 6 23... 2 1 24. .. 3 4 2 25... 1 4 2 26. . 9 7 4 2 27... 6 5 3 3 2 28... 2 4 2 3 4 2 29. . 5 3 3 2 2 6 30 . 4 3 4 6 4 31. . 2 4 3 10 8 3 June 1 .. . 3 2 2 6 5 1 2. .. 3 2 3 4 3. . 3 2 4 1 2 4... 1 2 2 2 5 3 5. . . 1 1 1 1 6 . 1 2 3 1 1 2 7. .. 1 1 3 8... 2 1 2 1 2 9. . 1 1 6 2 10. . 1 1 6 11... 3 2 1 12... 1 1 1 3 13. . 1 2 5 14. . . 1 2 15. .. 1 1 3 16... 1 2 17... 1 1 1 18. . . 1 19... 1 1 1 20 . 1 21... 22. . 1 1 23 . 1 24. .. 1 2 25. .. , . , . 26. . , . 27. . 1 1 [7] Table 3. Date of Flowering of Four Varieties of Onions (Bulbs Planted Nov. 4, 1950*) Variety Number of replications Mean date of flowering t Excel 3 2 2 1 24.3 28.6 29.6 33.5 Texas Early Grano Colorado Sweet Spanish Southport Yellow Globe * Bulbs placed in storage Aug. 3, 1950. t Days after May 1. Table 4. Effect of Storage Treatment on Stockton G-36 Onion Bulbs Planted Nov. 4, 1950"^ 32° F 50° F 77° F Common storage L.S.D. at 5 per cent level Mean date of flowering f No. plants per plot 28.9 33.0 3.7 14.2 46.3 5.6 28.9 44.0 4.0 24.4 45.0 4.5 3.0 8.8 0.6 No. seedstalks per plant * Bulbs placed in storage Aug. 3, 1950. t Days after May 1. Table 5. Percentage of Plants Bolting and Mean Date of Flowering of Three Varieties Planted Seed-to-seed Variety Planting date 1950 8/1 8/11 8/21 8/31 9/10 9/20 9/30 Texas Early Grano : Per cent flowering 91.2 90.1 62.4 76.7 29.5 65.7 6.8 Date of flowering 27.2* 26.6 28.3 31.0 33.9 32.9 33.9 Yellow Bermuda : Per cent flowering 99.2 96.9 82.6 80.9 68.2 52.6 35.9 Date of flowering 25.0 25.9 24.6 27.6 29.4 32.0 31.0 Excel: Per cent flowering . ....t ....f 99.5 96.0 75.1 77.9 49.2 Date of flowering . ....t ....t 21.8 23.9 28.3 29.6 33.8 * Days after May 1, 1951. t Poor lot of seed, did not germinate. [8] determine if the date of flowering in seed- to-seed plantings could be altered by the date of planting. Percentages of plants that bolted and mean dates of flowering for the varieties that were planted for seed-to-seed are shown in table 5. Here again, later plant- ings resulted in later flowering dates, and a smaller percentage of plants formed seed stalks. In the latest plantings so few of the plants produced seed stalks that the validity of the mean flowering date is doubtful. When the seed-to-seed method is used, it seems reasonable to assume that at least 90 per cent of the plants should form seed stalks to prevent an excessive selec- tion for easy bolters. However, this may not be so important in the production of Fi hybrid seed where the seed is used only to grow a commercial bulb crop and not as stock seed or breeding material, H we assume that when 90 per cent of the plants form seed stalks we are safe from excessive selection for easy bolters, the evidence of the three above varieties sug- gests that at Davis the varieties Bermuda and Texas Grano should be planted August 1. With the variety Excel plant- ing could be delayed until about August 20. Planting dates would have to be ad- justed to different areas of production. REPORT FOR 1952: Australian Brow^n, Yellow^ Bermuda, Red Wethersfield. Southport Red Globe, and Brigham Yellow Globe w^ere added to the comparative flowering date studies. Storage temperature and planting date experiments w^ere conducted wdth Australian Browns. Flow^ering trends were analyzed. Comparative flowering dates As a continuation of the date of flower- ing studies in 1952, mother bulbs of six varieties — Australian Brown, Yellow Bermuda, Red Wethersfield, Southport Red Globe, Stockton G-36, and Brigham Yellow Globe — were planted on Novem- ber 16, 1951. In addition, bulbs of the variety Australian Brown were separated into large and small to determine the effect, if any, of size of mother bulbs on date of flowering. The experiment was set up in a randomized block. Each range contained one 60-bulb plot of each vari- ety and there were four ranges, thus four replications of each variety. The flowering dates of certain varieties differed significantly (table 6). More important, however, than the fact that actual differences exist is the number of days that elapse between the earliest and latest flowering varieties. At Davis there was a maximum difference of eight days. An even greater spread might be found in other areas. A spread of as much as eight days between flowering dates of two lines is of great importance in the production of hybrid seed, as will be shown later. No significant difference in flowering date was noted between large and small Aus- tralian Brown bulbs. How different storages and planting dates affect Australian Browns To study the effect of storage treatment and date of planting on flowering date, number of seed stalks produced, number of flowers per umbel, and seed yield, only bulbs of the variety Australian Brown were planted. These were all placed in the different storages on Sep- tember 27, 1951. For the data on flowering date, num- [9] Table 6. Relative Date of Flowering of Six Onion Varieties in 1952 (Planted Nov. 16, 1951) Variety Days from May 1 to first flowering Rep. 1 Rep. 2 Rep. 3 Rep. 4 Mean Australian Brown (small) * . . . . Australian Brown (large) f. . . . Yellow Bermuda 32.5 32.4 23.9 31.8 29.7 32.1 27.3 35.2 31.1 23.4 29.2 30.4 33.2 26.8 30.0 32.4 24.8 30.1 28.2 30.5 24.4 30.7 32.5 25.4 31.3 28.7 30.3 26.1 32.1 32.1 24 4 Brigham Yellow Globe Red Wethersfield 30.6 29 2 Southport Red Globe Stockton G-36 31.5 26 2 Least significant difference at 5 per cent level for variety means, 1.9. * Average weight 1 oz. t Average weight 4.5 oz. ber of seed stalks produced, and number of flowers per umbel, four replications of 60 bulbs each (for each storage treatment and planting date) were planted, but data were taken only from the center 50 plants in each replication. The field was laid out in a split plot design with the dates of planting arranged in a Latin square as the main plots and the bulb storage treatments as subplots random- ized within the main plots. Guard rows were planted along the edges of the field and between the date-of-planting plots. A separate field, different from the above only in that each replication contained 120 bulbs, was planted for the seed-yield data, and seed was harvested from all plants. To facilitate the taking of data, all bulbs were planted 8 inches apart in rows 42 inches apart. To prevent possible differences due to different nitrogen levels during the win- Table 7. Effect of Bulb Storage Treatment and Date of Planting on Flowering Date, 1952 Planting date Bulb storage at Date means 50° F Common storage 32° F 68° F Number of days after May 1=" 1= Nov. 16, 1951 Dec. 15, 1951 Feb. 12, 1952 March 24, 1952t 29.6 30.4 34.6 42.4 31.5 31.3 33.4 35.7 42.9 33.5 31.1 35.5 44.5 54.7 32.6 37.6 49.2 31.2 34.2 41.0 Storage means 37.0 39.8 Least significant difference at 5 per cent level for date means, 1.4. Least significant difference at 5 per cent level for storage means, 0.9. Interaction (storage x date) significant at the 1 per cent level. * Each figure is the mean of four replications. t Not included in statistical analysis. [10] ter, 16-20-0 fertilizer (500 pounds per acre) was applied before planting. Dur- ing the spring and early summer the plots harvested for seed received four applications of DDT to prevent thrips damage. To insure adequate pollination, two hives of bees were placed in the seed- yield field. The fields received normal irrigation and cultivation. Seed heads were harvested at several different dates to prevent excessive shattering. Since less than 15 per cent of the plants from bulbs stored at 68° F and planted on March 24 bolted (fig. 1), no data from this treatment were taken on date of flowering, number of seed stalks, or number of flowers. Seed yields, how- ever, were taken for this treatment. Date of flowering. Table 7 shows the effect of storage treatments and plant- ing dates on the date of flowering of Aus- tralian Brown plants. Significant differ- ences were obtained from both storage treatment and date of planting. Regard- less of the storage temperature, as the date of planting was delayed (or the storage period lengthened), the number of days from May 1 to flowering was extended. In every case the earliest plant- ing date resulted in the earliest flowering. Similarly, storage treatment, regardless of planting date, had a significant effect on the flowering date — 50° F and com- mon storage resulting in the earliest flowering. The November 16 planting date gave the earliest flowering, 29.6 days from May 1. It may be noted that bulbs planted in November, 1950, responded more strik- ingly to storage treatment than bulbs planted in November, 1951 (tables 4 and 7). Two facts may account for this: 1) the variety used in the 1950-1951 experi- ments was the Stockton G-36, an early, soft summer onion, while the 1951-1952 studv involved the Australian Brown, a Fig. 1. Comparative amount of bolting of Australian Brown onions, mother bulbs planted on March 24, 1952. Storage treatments of the mother bulbs before planting are indicated on the signs. [11] Table 8. Effect of Bulb Storage Treatment and Date of Planting on Total Number of Umbels per Plot, 1952 Planting date Bulb storage at Date means 50° F Common storage 32° F 68° F Number of umbels per plot '= Nov. 16, 1951 Dec. 15, 1951 Feb. 12, 1952 March 24, 1952 1 301 320 266 210 268 278 260 213 249 225 158 42 261 217 154 270 260 210 Storage means 296 269 211 211 Least significant difference at 5 per cent level for date means, 22. Least significant difference at 5 per cent level for storage means, 17. Interaction (storage x date) significant at the 1 per cent level. * Each figure is the mean of four replications. t Not included in statistical analysis. late storage onion; 2) in the fall of 1950 the bulbs were placed in storage on August 3, while in 1951 the bulbs were not placed in storage until September 27. Number of seed stalks. The num- ber of seed stalks or umbels produced per plot (table 8) was also influenced by both the planting date and the storage treatment of mother bulbs before plant- ing. The earlier planting dates, regard- less of storage treatment, resulted in the greatest number of umbels per plot; stor- age at 50° F and common storage, re- gardless of date of planting, gave the greatest number of umbels. The 50° F or common storage treatments, together Table 9. Effect of Bulb Storage Treatment and Date of Planting on Number of Flowers per Umbel, 1952 Planting date Bulb storage at Date means 50° F Common storage 32° F 68° F Number of flowers per umbel* Nov. 16, 1951 Dec. 15, 1951 Feb. 12, 1952 March 24, 1952t 878 850 605 392 896 790 625 440 839 701 538 425 945 736 677 889 769 611 Storage means 778 770 693 786 Least significant difference at 5 per cent level for date means, 125. Least significant difference at 5 per cent level for storage means, 54. Interaction (storage x time) not significant. * Each figure is the mean of four replications, t Not included in statistical analysis. [12] with early planting, gave the greatest total number of umbels per plot. Number of flowers. The number of flowers per umbel was computed by se- lecting the first umbel to flower on each of the first 12 plants to flower in each replication. The open flowers on each umbel were counted three times a week Table 10. Effect of Bulb Storage Treatment and Date of Planting on Yield of Seed, 1952 Planting date Bulb storage at Date means 50° F Common storage 32° F 68° F Yield of seed (pounds per plot) * Nov. 16, 1951 Dec. 15, 1951 Feb. 12, 1952 March 24, 1952 7.9 7.2 6.2 4.4 7.6 6.9 6.6 4.5 6.4 7.3 6.6 4.8 1.4 7.6 6.4 4.2 0.6 7.6 6.8 5.4 2.7 Storage means 6.4 5.0 4.7 Least significant difference at 5 per cent level for date means, 0.7. Least significant difference at 5 per cent level for storage means, 0.3. Interaction (storage x date) significant at the 1 per cent level. * Each figure is the mean of four replications. Table 1 1. Summary of the Effect of Storage Treatments, 1952 50° F Common storage 32° F 37.0 211 693 5.0 68° F L.S.D. at 5 per cent level Date of flowering* 31.5 296 778 6.4 33.5 269 770 6.4 39.8 211 786 4.7 0.9 17 54 0.3 No. umbels per plot No. flowers per umbel Seed yields per plot (lbs.) * Days after May 1. Table 12. Summary of the Effect of Date of Planting, 1952 Date of flowering* No. umbels per plot No. flowers per umbel. . . Yield of seed per plot (lbs. j * Days after May 1. Nov. 16 31.2 270 889 7.6 Dec. 15 34.2 260 769 6.8 Feb. 12 41.0 210 611 5.4 March 24 2.7 L.S.D. at 5 per cent level 1.4 22 125 0.7 [13] Table 13. Flowering Trend of individual Umbels, 1952 Date 5/20 5/22 5/24 5/27 5/29 6/1 6/4 Percentage of flowers open* 1.7 4.8 13.9 36.2 52.6 67.1 84.0 Date 6/6 6/7 6/9 6/11 6/13 6/16 6/18 Percentage of flowers open* 92.0 94.1 97.6 98.8 99.4 100.0 100.0 Means of 63 umbels which started to flower on May 20, 1952. and were discarded at the time of count- ing. The data are presented in table 9. Regardless of the storage treatment, a delay in the date of planting resulted in a decrease in the number of flowers per umbel. Storage treatments seemed to have less influence on the number of flowers per umbel, the 32° F treatment being significantly poorer than the other three. Seed yield. Since the seed-bearing potential (number of flowers per umbel) of the plants in question was influenced by the planting date and storage treat- ment of mother bulbs, it is to be expected that seed yield would be affected by the same factors. Table 10 shows that as the planting date was delayed, regardless of storage treatment of mother bulbs, seed yields decreased. Storage treatments of mother bulbs resulted in significant dif- ferences in the yield of seed per plot, the 50° F and common storage treatments giving significantly higher seed yields cr UJ O 100 - l^ 90- Z — 80- LU CD ^60- "=> 50- u. O 40- \— 30- S '0- Q. I ^ 1 L • • • ® ® ® SI®® ® ® ® 17 20 22 24 27 29 31 MAY II 13 16 18 20 23 25 27 JUNE Fig. DATE 2. Flowering trends of plots which started to flower on May 17 and 27, 1952. (Based on data in table 14.) [14] than storage at either higher or lower temperatures. The earliest planting date with previous storage of mother bulbs at 50° F or common storage gave the high- est yield of seed per plot. Flowering trends The flowering trend of individual um- bels should be of major interest to onion breeders and producers of hybrid seed. Table 13 show^s the mean trend of flower- ing on 63 umbels that started to flower on May 20, 1953. It should be noted that once an umbel starts to flower, the open- ing of the individual flowers proceeds very rapidly; more than half the flowers per umbel were open nine days after the first flower opened. The rate of flower opening will un- doubtedly be influenced by the tempera- ture ; the purpose of these data is to show that although an umbel may be in flower for 30 days, the actual effective flowering period is much shorter. Thus, any dis- crepancy of flowering dates between in- bred lines planted for hybrid seed will result in lower seed production. Table 14 shows flowering trends by Table 14. Flowering Trend Within Plots, 1952 No. of plots 13 17 3 4 4 6 Total no. of umbels Cumulative percentage of umbels with open flowers on successive dates 5/17 1.7 5/20 5/22 5/24 5/27 5/29 5/31 6/3 3,757 10.4 2.5 18.6 6.1 1.0 .... 30.9 13.6 2.8 1.2 46.6 28.7 10.0 6.6 3.4 55.3 40.9 19.2 16.5 11.3 .9 64.2 53.3 33.8 31.8 26.7 4.3 72.6 65.8 49.9 54.8 48.1 12.7 4,382 709 934 929 1,088 No. of plots 13 17 3 4 4 6 Total no. of umbels Cumulative percentage of umbels with open flowers on successive dates 6/5 6/7 6/9 6/11 6/13 6/16 6/18 6/20 3,757 77.1 73.5 63.6 67.7 60.7 23.5 80.0 76.6 69.4 71.8 65.1 30.4 85.8 82.1 77.6 80.5 74.1 43.4 89.5 86.6 81.4 84.7 79.8 52.7 93.4 90.3 87.2 89.9 84.9 64.1 96.7 93.8 91.3 94.4 90.0 74.8 98.7 96.7 95.4 97.9 94.1 85.2 99.4 97.7 97.3 98.4 95.9 88.4 4,382 709 934 929 1,088 No. of plots 13 17 3 4 4 6 Total no. of umbels Cumulative percentage of umbels with open flowers on successive dates 6/23 6/25 6/27 6/30 7/3 7/5 7/7 7/9 3,757 99.8 99.0 99.0 98.8 97.4 93.0 99.9 99.2 99.3 99.1 98.6 95.6 100.0 99.7 99.6 99.4 99.1 97.4 99.9 100.0 99.6 99.7 98.8 99.9 99.7 100.0 99.4 100.0 100.0 99.6 99.8 100.0 4,382 709 934 929 1,088 [15] plots, which are more or less comparable to seed fields. In the experiments with Australian Browns in 1952, 64 plots (60 plants per plot) were involved. Different treatments caused various plots to start flowering on different dates. It was deemed of interest to know what percent- age of the umbels in the plots had open flowers on successive dates once flower- ing started. The first line in Table 14 be- low the dates gives the data from 13 plots that started to flower on May 17. The figures below the dates are percentages of the total number of umbels in the plots that had open flowers by that date. The successive lines are data from the plots that started to flower on later dates. The flowering trends of the plots that started to flower on May 17 and May 27 (fig. 2) show that a 10-day difference in flowering date means that most of the umbels on the early flowering line will be in flower while relatively few on the late line are in flower. Another factor important in obtain- ing maximum seed yields is that early- flowering umbels within a line have many more flowers than later-flowering umbels within the same line. Four replications (60 plants each) came from bulbs stored at 50° F and planted on November 16. Within these plots, 45 umbels starting to flower on May 17 and 20 averaged 876.1 ±31.1 flowers per umbel, while the average for 47 umbels that did not start to flower until June 5 and 7 was 374.7 ±10.5. In other words, within a line, late-flowering umbels have a much lower seed-bearing potential than early- flowering umbels. For maximum yields of hybrid seed, the two inbred parental lines must come into flower on approximately the same date. The data suggest that if the flower- ing dates of two inbred lines cannot be adjusted to approximately coincide, the later-flowering line should be the female parent. LITERATURE CITED Blaauw, a. H., Annie M. Hartsema, and Ida Luyten 1944. Bloemen of boUen bij Allium cepa L. Ill en IV. K. Akad. van Wetensch. te Amsterdam, Verslag. van Wis en Natuurk. Afd. 53 (5) :274-291. Hartsema, Annie M, 1947. De periodieke ontwikkeling van Allium cepa L. var. zittauer riesen. Wageningen Land- bouwhoogesch. Meded. 48 (6) : 263-300. Heath, O. V. S., and M. Holdsworth 1948. Morphogenic factors as exemplified by the onion plant. Soc. Exp. Biol. Symposia 2:326- 350. Holdsworth, M. 1927. A comparative study of onion varieties in relation to bolting and yield when grown from sets. Ann. Appl. Biol. 32 (1) :22-34. Jones, H. A. 1927. The influence of storage temperature on seed production in the Ebenezer onion. Amer. Soc. Hort. Sci. Proc. 24:61-63. Jones, H. A., and A. E. Clarke 1943. Story of hybrid onions. U.S. Dept. Agr. Yearbook 1943-1947. p. 320-326. Jones, H. A., and G. N. Davis 1944. Inbreeding and heterosis and their relation to the development of new varieties of onions. U.S. Dept. Agr. Tech. Bui. 874:1-28. Jones, H. A., and S. L. Emsweller 1937. A male-sterile onion. Amer. Soc. Hort. Sci. Proc. 34:582-585. 1939. Effect of storage, bulb size, spacing and time of planting on production of onion seed. California Agr. Exp. Sta. Bui. 628:1-14. Jones, H. A., B. A. Perry, and W. C. Edmundson 1949. Vegetative propagation of short-day varieties of onions as an aid in a breeding program. Amer. Soc. Hort. Sci. Proc. 53:367-370. Miller, Julian C. 1932. Effect of temperature upon the character of growth of mature onion bulbs. Amer. Soc. Hort. Sci. Proc. 29:514-516. Snedecor, George W. 1946. Statistical methods. 4th ed. 485 p. Iowa State College Press, Ames, Iowa. Thompson, H. C, and 0. Smith 1938. Seedstalk and bulb development in the onion (Allium cepa L.). Cornell Agr. Exp. Sta. Bui. 708:1-21. 7im-ll,'54(5680)BEB [16]