UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA EFFECT OF STORAGE, BULB SIZE, SPACING, AND TIME OF PLANTING ON PRODUCTION OF ONION SEED H. A. JONES and S. L. EMSWELLER BULLETIN 628 April, 1939 UNIVERSITY OF CALIFORNIA BERKELEY, CALIFORNIA CONTENTS PAGE Storage studies with the variety Ebenezer 3 Storage studies with other varieties and strains 6 Influence of size of mother bulb on production of seed heads and seed 8 Spacing studies 9 Influence of time of planting on seed yield 10 Discussion 12 Summary 14 Literature cited 14 EFFECT OF STORAGE, BULB SIZE, SPACING, AND TIME OF PLANTING ON PRODUCTION OF ONION SEED 1 H. A. JONES 2 and S. L, EMSWELLER* Onion-seed growing is confined to certain districts of the West, chiefly because the semiarid weather of summer usually provides favorable con- ditions for pollination of the flowers and for harvesting, curing, and threshing of the seed. Many factors are known to influence the number of seed stems formed and the yield of seed. Those that have been studied by the authors and are discussed in this paper are storage, size of mother bulb, spacing, and time of planting. The experiments were conducted at Davis, California. Various strains, especially of the variety Ebenezer, are known to vary a great deal in their capacity to produce seed. During certain seasons in California well-known strains of this variety have failed to produce even a fair crop. Instead of forming seed stems the mother plants may develop several vegetative shoots which later form large bulb^united at the base ; or, the plant may develop a mixture of bulbs and seed stems. Because of results of some earlier work done at the Maryland Agricul- tural Experiment Station by the senior author, it was thought that this tendency to produce vegetative shoots might be due at least in part to the temperature at which the mother bulbs are stored previous to planting. STORAGE STUDIES WITH THE VARIETY EBENEZER To determine if possible the effect that certain storage conditions might have on seed production, a strain of Ebenezer was secured that was known to seed poorly in some years. Bulbs of this strain were grown over a period of five years for the storage studies. Shortly after the bulbs were harvested each year, usually late in July or early in August, they were placed under six different storage condi- tions. Four of the lots were held in constant temperature rooms at about 38.5°, 46.5°, 53.5° and 86° Fahrenheit respectively. One lot was held in a basement where the temperature ranged from about 61° to 71.5°. When the bulbs were placed in storage the temperature of this basement was usually about 71.5°. During autumn the temperature gradually dropped and was usually at about 61° at the time the bulbs were removed for planting. There was very little fluctuation between day and night. High 1 Received for publication November 7, 1938. 2 Professor of Truck Crops and Olericulturist in the Experiment Station. Resigned. 3 Assistant Professor of Truck Crops and Assistant Olericulturist in the Experi- ment Station. Resigned. [3] 4 University of California — Experiment Station humidity conditions in the basement favored rotting and some years only a few bulbs survived. Therefore, data for basement storage are not in- cluded in table 1. Another lot was stored in the onion warehouse where the temperature and humidity fluctuated with the weather. The bulbs were removed from storage and planted in the field in early December except in the season 1930-31 when planting was done on Jan- uary 18. In any one year bulbs of comparable size from all storage con- ditions were planted in all plots. An attempt was made to plant bulbs of about the same size every year but this was not always possible. In the years 1926-27 to 1929-30 the average weight of bulb planted was about TABLE 1 Yields of Seed Heads and Seed from Mother Bulbs of the Variety Ebenezer Stored Under Different Conditions Storage condition Average number of seed heads per plot Yield of seed, pounds per acre 1927 1928 1929 1930 1931 Mean 38.5° F 190.4 213.7 238.3 53.6 213.1 561 1,259 1,386 276 861 290 196 381 16 234 495 505 620 63 519 772 768 804 353 948 1,248 1,486 1,200 7 1,159 673.0 46.5° F 842.4 53.5° F 877.8 86.0° F 142.5 Warehouse 744.1 2XS.E. of the difference between means of 10 15.1 70.3 54, 50, 54, and 108 grams for the respective years. In 1930-31 the bulbs from 86° F storage averaged 66 grams ; all others averaged 76. The bulbs were set in the field in shallow furrows and covered imme- diately after planting. They were spaced 8 inches apart in the row from center to center of the bulb. A plot consisted of one row of 50 plants. These were systematically replicated from two to five times according to the number of bulbs available. Harvesting was done every few days by cutting the seed heads that had about 10 to 25 per cent of the capsules dehisced. After curing, threshing, and fanning, the seed was washed, dried, and weighed. Plants from bulbs stored at 46.5° and 53.5° F were always the first to bloom and consequently seed matured first in these plots. The time of flowering and seed maturity was delayed more in plants from bulbs held at 86° than at any other storage temperature. Storage conditions had a marked effect on the total number of seed heads produced (table 1). The bulbs held at 53.5° F gave the largest number of seed heads ; those held at 86° the lowest (fig. 1) . There is a sig- Bul. 628] Production of Onion Seed nificant difference between the number of heads produced by the bulbs stored at 53.5° and those held at 46.5°, (only 7° lower) if twice the standard error of a difference is taken as the limits of experimental error. 4 There was no significant difference in results between storage at 46.5° and the warehouse temperatures, both of these being more favorable for the production of seed heads than was 38.5°. The effect of storage temperature on yield of seed was also very pro- nounced. The mean yield from the 53.5° F plots was higher than that from ^f*^ff^^^ * t^f JU i -%r h Fig. 1. — The variety Ebenezer. The center row was grown from mother bulbs stored at 86° F; note the scarcity of seed stalks. The row to the right of the center was grown from bulbs stored in the warehouse and the row to the left of the center from bulbs stored at 53.5°. Photographed June 18, 1931. any other, although not significantly higher than the 46.5° plots, a differ- ence between means of 70.3 pounds of seed per acre being required for odds of 19 to 1. On the 86° plots the yields were very low. Both high and low temperatures had a depressing effect on seed yield but high tempera- ture was most effective. When the variance of the test is separated into its component parts the parts due to years and to storage treatment and their interaction are all highly significant. The high variance due to years is to be expected be- cause of differences in climatic conditions and prevalence of disease and severity of thrips attack. The high value for the interaction between years and storage can no doubt be partially explained by the influence of climatic conditions between time of planting and the time of differen- * In this paper minimum values required for significance of differences have been calculated as 2 x S.E. of a difference, assuming that odds of 19:1 against chance alone represent significance. 6 University of California — Experiment Station tiation of the floral axes. Except in 1930-31 the mother bulbs in these tests were planted in early December and as reported by Jones and Emsweller (3) 5 , the primordia of the floral axes are not differentiated until some time in February. The environmental conditions that prevail during this period no doubt tend to exaggerate or smooth out the differ- ences due to storage. Variance 6 due to storage treatment is significantly greater than the interaction between years and storage by odds of more than 99 to 1. Thus the relative differences shown by storage treatment for the mean of five years occurred rather consistently. In 1926-27 the yield of bulbs formed in the seed field as well as the yield of seed was recorded for each of the plots. As only the first two repli- cates were used for the calculation of seed yield, these were used to de- termine bulb yield. The yields of bulbs in pounds per acre from the mother bulbs stored at the several conditions were as follows: 53.5° F, 2,320; 46.5°, 4,405; 38.5°, 4,987 ; warehouse, 7,573; 61°-71.5°, 11,237; and 86°, 14,646. The data show that the storage conditions most favor- able for bulb formation were most unfavorable for the production of seed. STORAGE STUDIES WITH OTHER VARIETIES AND STRAINS In 1927-28, in addition to the poor-seeding strain of Ebenezer, the effect of storage temperature on seed production was determined on a more free-seeding strain of Ebenezer, Yellow Globe Danvers, and a strain of the latter (pedigree No. 15-8), inbred for one generation. The methods of handling both in storage and in the field were the same as described above. The average weight of bulb planted was 50 grams for all varieties. In all cases the first seed stems to appear were in the 53.5° F plots and the last to appear were in the 86° plots. Figure 2 shows how much farther advanced the 53.5° plots are than the 86° plots on May 13. Storage at 86° F, at 61°-71.5°, and in the warehouse caused a marked retardation in time of seed harvest. The effect was present in all varieties tested. The influence of storage temperature on the number of seed heads produced was also very marked as shown in table 2. In all varieties the 53.5° F storage produced the most seed heads but not significantly higher than the 61°-71.5° or the warehouse temperatures. In every case the few- est heads were harvested from the highest temperature, 86°, but there was little difference between the 38.5° and the 46.5° lots. In both Eben- 5 Italic numbers in parentheses refer to "Literature Cited" at the end of this bulletin. 6 Variance analysis has been applied to the data of all the tables; only the results of this analysis are indicated throughout the paper. Bul. 628] Production of Onion Seed ezer and strain 15-8 the seed yield of the 86° plots was less than any of the others, a difference of 82 pounds being required for significance. Variance analysis shows highly significant values for number of seed heads harvested and for seed yield in regard to varieties, storage, and the ' i ii ■botH ■ ■":: :% ■»' ^A ,;J-\ *** \ / * ■-■.. 3fe*c8k 'IhHI Fig. 2. — The variety Yellow Globe Danvers. Plants from bulbs stored at 53.5°F are in full bloom; those from bulbs stored at 86° have flower buds still enclosed within the bract. Photographed May 13, 1928. interaction between varieties and storage. The interaction shows that not all varieties used responded the same to the several storage treatments. Variance in seed yield due to varieties is significantly greater than the interaction between varieties and storage by odds of more than 99 to 1. TABLE 2 Effect of Bulb Storage Conditions on Yield of Seed Heads and Seed in 1927-28 Yellow Globe Danvers Ebenezer Strain No. 15-8 Storage conditions Seed heads per plot Pounds seed per acre Seed heads per plot Pounds seed per acre Seed heads per plot Pounds seed per acre 38.5° F 175 179 239 174 230 237 613 470 635 469 761 789 272 267 292 101 265 277 761 787 788 203 677 791 157 176 215 132 207 212 370 46.5° F 389 53.5° F 457 86.0° F 61°-71.5°F 184 416 Warehouse 435 2XS.E. of the difference between means of 5 plots 22 82 22 82 22 82 8 University of California — Experiment Station INFLUENCE OF SIZE OF MOTHER BULB ON PRODUCTION OF SEED HEADS AND SEED The Yellow Globe Danvers variety of onion was used to determine the influence of size of mother bulb on the number of seed heads formed per plant and on seed yield, the same strain being used during the four years the tests were run. Bulbs were stored in the warehouse from harvest to planting time. The plots each consisted of 2 rows 27 feet long and they were systematically replicated five times each year. Spacing was 3 feet TABLE 3 Average Yield of Seed Heads and Seed from Mother Bulbs of Different Size in Seasons of 1927-28 to 1930-31; Variety, Yellow Globe Danvers Weight of mother bulbs, grams Number of seed heads per plant Pounds seed per acre Grams seed per plant 11-20 2.41 3.02 3.46 3.74 4.14 4.42 4.64 4.75 407 551 648 668 744 759 786 796 9.16 21-30 11.98 31-40 13.98 41-50 51-60 14.65 16.08 61-70 16.50 71-80 17.13 81 , and over 17.30 2XS.E. of the difference between means of 20 plots 0.14 40 0.84 between rows, and 8 inches from center to center of adjacent bulbs in the row. The mean numbers of seed heads and weights of seed harvested per plant and the mean seed yields per acre for the four years are given in table 3. There was an increase in number of seed heads harvested with each increase in size of mother bulb. All differences are significant except be- tween the two largest sizes. There was also an increase in yield of seed per acre and in seed yield per plant for each increase in size of mother bulb. In seed heads harvested per plant, seed yield per acre, and seed yield per plant, the greatest differences were between the two smallest sizes and in general the differences became less with each increase in size of mother bulb, with the least difference exhibited between the two largest sizes. Variance analysis shows highly significant values for years, weights, and plots. Variability due to years is caused by variation in environmen- tal conditions. A high value for weights in comparison with interactions is obtained and means that the differences due to weights would hold Bul. 628] Production of Onion Seed under a wide range of conditions. A high value of interaction between years and weights indicates that the various sizes do not respond the same each year — this is no doubt being caused by varying climatic con- ditions. SPACING STUDIES Studies were conducted also to determine the influence of various spac- ings upon the production of seed stems and seed yield. The same strain of Yellow Globe Danvers was used and the bulbs were handled in the same manner as for the studies on the effect of size of mother bulbs. The TABLE 4 Average Yield of Seed Heads and Seed from Plants Grown at Various Spacings in Seasons of 1928-29 to 1930-31 ; Variety , Yellow Globe Danvers Inches apart in the row Number of seed heads per plot Number of seed heads per plant Pounds seed per acre Grams seed per plant 3 315 256 188 146 104 3.32 3.47 3.63 3.74 3.88 1,337 1,145 884 736 563 12.05 4 13.20 6 8 14.50 15.90 12 17.77 2XS.E. of the difference be- tween means of 15 plots .... 13 0.25 66 1.45 bulbs varied in weight from 50 to 120 grams but were planted with simi- lar sizes in all plots so that results are comparable. They were set 3, 4, 6, 8, and 12 inches from center to center of the bulbs in the row. Rows were 3 feet apart. Plots were single rows 27 feet long. The plots were system- atically replicated five times each year. The average number of seed heads harvested per plot and per plant and the yield of seed per acre and per plant are given in table 4. The number of seed heads harvested per plot was significantly reduced with each increase in spacing. In no case did the increase in space pro- vided per plant compensate for the decrease in number of plants ; this holds for seed yield as well as for the number of seed heads harvested. There was also an increase in the number of seed heads harvested per plant with each increase in spacing. There was a significant decrease in yield of seed per acre for each increase in spacing. For each increase in distance between bulbs there was an increase in seed yield per plant but the increase was not always statistically significant, and was not large enough to compensate for the reduced number of bulbs planted per acre. Variance analysis of seed yield per acre and per plant and seed heads 10 University of California — Experiment Station harvested per plant and per plot gives highly significant values for both years and spacings. Variance in seed yield per acre and seed heads per plot due to spacing is significantly greater than the interaction between years and spacing by odds of more than 99 to 1, indicating that the relative differences shown by spacing during the three years occurred rather consistently. INFLUENCE OF TIME OF PLANTING ON SEED YIELD The same strain of Yellow Globe Danvers as was used for studies on the effect of size of mother bulb and spacing was used to determine the effect of time of planting mother bulbs in the field on seed yield, on the number TABLE 5 Effect of Time of Planting Mother Bulbs on Time of Seed Maturity and on Yield of Seed Heads and Seed in Season of 1928-29 ; Variety, Yellow Globe Danvers Date bulbs planted December 7 December 27 January 1 January 22 January 30 February 15 March 2 2XS.E. of the difference between means <>f 5 plots Heads harvested July 5 to July 23 Heads harvested July 24 to August 9 Total seed heads per plot 108 96 90 102 98 95 99 Pounds seed per acre 693 618 614 539 496 in of seed heads harvested, and on the time of maturity. There was consid- erable variation in dates of planting and harvesting for the different years so the data for each year are presented separately (tables 5-7). The average weight of bulb used each year was about 40 grams. The spac- ing was 8 inches between bulbs and 3 feet between rows. Plots were single rows 27 feet long. Quintuplicate plantings were made each year. The dates of planting used in this experiment did not significantly influence the total number of seed heads harvested. They did, however, have considerable effect on the date of maturity. In general as the plant- ing elate was delayed there was a decrease in the number of seed heads harvested during the first harvesting period and an increase in the sec- ond period. Variance analysis shows a value for planting date in 1930-31 that is probably not significant. The interaction between planting date and har- Bul. 628 Production of Onion Seed 11 vest time gives a value for 1928-29 that exceeds that for odds of 99 to 1, and for 1929-30' and 1930-31 one that exceeds the value for odds of 19 to 1. Although there was no significant difference in the total number of seed heads for the different dates of planting, the time of planting did sig- TABLE 6 Effect of Time of Planting Mother Bulbs on Time of Seed Maturity and on Yield of Seed Heads and Seed in Season of 1929-30 ; Variety, Yellow Globe Danvers Date bulbs planted Heads harvested July 3 to July 16 Heads harvested July 17 to August 5 Total seed heads per plot Pounds seed per acre October 17 November 8 November 23 December 20 January 1 January 23 2XS.E. of the difference between means of 5 plots 176 180 169 164 149 158 22 948 1,026 947 766 751 739 96 nificantly influence the seed yield per acre in both 1928-29 and 1929-30, there being in a general a decrease in yield as the planting date was de- layed. This decrease in yield of seed per acre was not due to the decrease TABLE 7 Effect of Time of Planting Mother Bulbs on Time of Seed Maturity and on Yield of Seed Heads and Seed in Season of 1930-31 ; Variety, Yellow Globe Danvers Date bulbs planted Heads harvested June 20 to June 29 Heads harvested June 30 to July 13 Total seed heads per plot Pounds seed per acre 77 55 60 59 79 77 88 88 156 132 148 147 941 782 884 823 2XS.E. of the difference between means 40 40 57 105 in number of seed stems produced but to other factors such as high tem- perature and increase in thrips population which have a deleterious effect on the late-maturing heads. In 1930-31 the planting range was too narrow to bring out any effect of time of planting. Variance analysis 12 University of California — Experiment Station gives a value for planting dates that approaches the value for odds of 99 to 1 in 1928-29 and exceeds it in 1929-30. DISCUSSION There has been only a limited amount of experimental work reported on factors influencing seed production in the onion and this has been chiefly in connection with studies of premature seeding in the growing of bulbs from small sets. Boswell (1), working in Maryland in 1923, found that not only did the size of the bulb influence the percentage of seeders but also within any size range the percentage was influenced by storage con- TABLE 8 Influence of Size of Sets and of Storage Temperature on Percentage of Seeders 32° F for 8 months 50° F for 8 months 32° F for 6.5 months and 50° for 7 weeks Average weight of mother bulb, grams Per cent seeders Average weight of mother bulb, grams Per cent seeders Average weight of mother bulb, grams Per cent seeders 3.8 7.7 15.3 34.3 0.0 0.9 12.3 25.0 4.2 7.7 14.9 27.5 12.5 30.2 68.3 81.0 6.7 13.2 32.3 9.6 38.3 65.2 Source of data: Boswell (1). ditions. For convenience of reference and because of the importance of these data, some of BoswelPs results are presented in table 8. The data show that bulbs stored at 50° F produced the highest percentage of seed- ers and at 32° it was markedly reduced. The inhibiting effect on seeding by storage at 32° was partially overcome by storing at 50° for a time previous to planting. More recently Thompson and Smith (3) working in New York State, have presented data showing the influence of storage temperatures and size of set upon percentage of seeders. While the studies had to do with set onions there seems to be no reason why the results cannot be applied to the larger-sized onions used for mother bulbs. The largest size of set used by Thompson and Smith in their studies was approximately 1% inches in diameter, which is somewhat smaller than the size commonly used for seed production. These sets were stored, at temperatures shown in table 9, from early October until late April at which time they were set in the field. The data show that the storage temperatures markedly influenced the percentage of plants going to seed. For all varieties, stor- Bul. 628] Production of Onion Seed 13 age at 40° and 50° F was most favorable for bolting; while storage at both higher and lower temperatures had an inhibiting influence. These authors also determined the effect of keeping the sets during the last month of storage at temperatures differing from those used for the main storage period. Southport Yellow Globe sets of the size mentioned above were used. Transferring from a temperature of 30° or 32° F to one TABLE 9 Percentage of Seed Stalks Developed from Large Onion Sets* Stored at Different Temperatures Storage temperature, degrees Fahrenheit Ebenezer Yellow Globe Red Wethersfield 30f 2.20 9.14 40.98 35.02 1.58 33.37 42.70 79.98 68.64 18.51 40.36 32 68.10 40 84.38 50 78.86 60-70 24.58 * Size: 13 /(6 to V/% inches in diameter. t Average of 1933-1935; all others are averages of five years, 1932-1936. Source of data: Thompson and Smith (3). of 40° or 50° or from 60° or 70° to 32° or 50° gave an increase in the per- centage of seeders. Shifting from 40° and 50° to 60°-70° resulted in a decrease. The temperature conditions to which the growing plants are subjected are also important in determining the percentage of bolters as shown by results secured when onion sets of the Ebenezer and Red Wethersfield va- rieties were planted in the greenhouse (Thompson and Smith) and grown under three temperature ranges : 50°-60°, 60°-70°, and 70°-80° F. Usually 100 per cent of the plants seeded when grown at 50°-60°, while none seeded at 70°-80°. Only a few bolted at the 60°-70° range. These data indicate the important part played by temperature during both the storage and growth periods in determining whether the onion plant shall bulb or form seed stalks. The best temperature for mother-bulb storage appears to be about 45° to 55° F. Both high and low storage temperatures inhibit produc- tion of seed stems and seed yield. In districts where it is necessary to hold the bulbs at temperatures of 30°-32° during the winter, a shift to about 45°-55° for a few weeks during the latter part of the storage period should stimulate development of seed stems. Early planting of mother bulbs is important for good seed yields. De- layed planting causes the seed heads to bloom late when the temperatures are high and thrips injury more severe. 14 University of California — Experiment Station SUMMARY The best temperature at which to store mother bulbs of a highly non- bolting strain of Ebenezer was about 53.5° F. Plants from bulbs stored at this temperature bloomed and ripened their seed earlier, produced more seed stems and yielded more seed per acre than plants from mother bulbs stored at higher or lower temperatures. In a comparison of eight different sizes of mother bulbs of the Yellow Globe Danvers variety ranging in weight from about 15 to 90 grams, there was found to be an increase in number of seed heads per plant, in seed yield per plant, and in seed yield per acre with each increase in size of bulb. In spacing tests with the Yellow Globe Danvers variety, bulbs were set 3, 4, 6, 8, and 12 inches apart in a row. A significant increase in seed yield per plant was secured with each increase in spacing but a decrease in yield of seed per acre. When bulbs of Yellow Globe Danvers were planted at different dates throughout the winter there was a delay in time of seed maturity of the late-planted bulbs ; and, accompanying this, a decrease in yield of seed. LITERATURE CITED 1. Boswell, V. R. 1923. Influence of the time of maturity of onions on the behavior during storage and the effect of storage temperature on subsequent vegetative and repro- ductive development. Proc. Ainer. Soc. Hort. Sci. 20:234-39. 2. Jones, H. A., and S. L. Emsweller. 193G. Development of the flower and inacrogametophyte of Allium cepa. Hil- gardia 10:415-28. 3. Thompson, H. C, and Ora Smith. 1938. Seedstalk and bulb development in the onion (Allium cepa L.) New York (Cornell) Agr. Exp. Sta. Bui. 708:1-21. IOim-5, '39(7602)