IMPROVEMENT OF SUGAR BEET BY MEANS OF INDUCED TRIPLOIDY 1) SEIJI MATSUMURA National Institute of Genetics, Misima

Transcription

1 IMPROVEMENT OF SUGAR BEET BY MEANS OF INDUCED TRIPLOIDY 1) SEIJI MATSUMURA National Institute of Genetics, Misima and AKIRA MOCHIZUKI Laboratory of Genetics, Ky8'o University Received November 10, 1952 UDC Improvement work with sugar beet by means of induced polyploidy has been carried out since 1941 in the Kihara Institute for Biological Research under the guidance of Prof. H. Kihara. Recently, this work has gained the cooperation of the National Institute of Genetics, the Laboratory of Genetics of Kyoto University, the Laboratory of Plant Breeding of Hokkaido University, the Hokkaido Agricultural Experiment Station and the Japan Beet-Sugar Manufacturing Tetraploid Company. I. Induced polyploidy plant of several varieties of sugar beet have been obtained by colchicine treatment of seeds, young seedlings and roots (Matsumura and Yamashita 1941). In the colchicine treatment, better results were obtained by applying a 0.4% solution on seedlings by drops. Matsumura, Mochizuki and Akemine (1942) succeeded in inducing tetraploids in many varieties of sugar, fodder, table, leaf and wild beets. Fig. 1. Meiotic chromosomes in the 1st metaphase of PMMC's in sugar beets. Xca. 2,000. a, 2x (9u). b, 3x (9m). C, 4x (2iv+14u). In sugar beets triploid plants were obtained by crossing tetraploids with diploids ; also an octoploid shoot was found in a tetraploid plant. Somatic and meiotic chromosomes in 2x, 3x, and 4x beets have been compared (Fig. 1). The results 1> Contributions from the National Institute of Genetics, Japan, No. 43. Contributions from the Laboratory of Genetics, Biological Institute, Kyoto University, No. 240.

2 48 JAPANESE JOURNAL OF GENETICS Vol. 28 No. 2 were generally in accord with those of Rasmusson and Levan (1939). Comparative studies of pollen grains and flowers of the polyploid plants have been carried out (Figs. 2 and 3). - Fig. 2. Comparison of flowers of polyploid sugar beets. ~ ca From left to right : 2x, 3x, 4x and 8x. Fig. 3. Comparison of pollen grains of polyploid sugar beets. ~ca From left to right II. Yield of tetraploid sugar beet From the results of Matsumura, Mochizuki and Suzuka (1950) it appears that there is no significant difference in number and length of leaves between 2x and 4x plants, but width and thickness of leaf blades and stalks are markedly greater in 4x than in 2x (Figs. 4 and a b Fig. 4. Hon-iku No. 398 (b) and its tetraploid (a). ~ ca. 1/22. Fig. 5. Roots and leaves of Hon-iku No. 398 (b) and its tetraploid (a). ~ ca. 1/25. 5). Since 4x plants had a later maturity and a higher resistance to diseases, their leaves looked fresher and were of a lighter green even at the harvesting time (end of October in Hokkaido) than 2x leaves. Table I shows the preliminary results of comparative studies on the yield of 2x and 4x plants of 2 common sugar beet varieties in Hokkaido, Hon-iku No. 398 and No. 48. Weight of roots and raw sugar production in 4x were remarkably higher than

3 June 1953 S. MATSUMURA and A. MOCHIZUKI: IMPROVEMENT OF SUGAR BEET 49 those of 2x. From experiments repeated several times on a little larger scale it was also Table I. Comparison of yield in 2x and 4x beets ascertained that 4x beets had higher resistance to diseases, insects and putrefaction in storage, smaller root-heads and higher sugar purity than 2x. On the other hand, the maximum yielding capacity of 4x beets was hard to estimate accurately, because of their late maturity, especially since unfavorable weather conditions during the late period of growth or the advent of early cold weather in the fall greatly influence their yield. Thus, sometimes the yield of 4x is lower than that of 2x (comp. Tab. II, results in 1945). III. Yield of triploid sugar beet Because of selfsterility of sugar beet, triploid seeds have been obtained by surrounding single 4x plant by several 2x plants. The results of investigations of 3x, 2x and 4x beets, Table II. Comparison of yield and sugar content in 2x, 3x and 4x beets conducted in Hokkaido Agricultural Experiment Station in 1945, in unfavorable weather conditions, are shown in the upper part of Table II. Varieties used in this study were Hon-iku No. 398 and its polyploids. Triploid beets were more vigorous, grew better and always showed a higher yield than the others. Brix (refractmeter %), sugar purity and resistance to diseases and insects of 3x beets were intermediate between those of 2x and 4x beets. These experiments were repeated in 1946 on the same scale and proved to be generally in accord with the earlier ones (Mochizuki and Matsumura 1950). If 2x and 4x beets are planted in alternate rows, 3x seeds can be obtained through natural pollination either from the 2x or 4x plants. Offspring from 4x mother plants were called, for convenience' sake, 3x-A, while those from 2x plants were designated as 3x-B. In 1947 similar comparative studies on 2x, 3x and 4x beets were carried out in Hokkaido Agricultural Experiment Station (Tab. II, lower part). 3x-A showed the best result, while the yield of 3x-B

4 50 JAPANESE JOURNAL OF GENETICS Vol. 28 No. 2 was intermediate between 2x and 4x. It can be safely assumed that the lower yield of 2*8 than that of 3x-A is due to the appearance of many 2x beets in 3x-B. For the large stak production of 3x seeds it was, therefore, found advisable to plant 4x and 2x beets in the ratio of 3 : 1. In this way relatively many 3x plants were obtained (comp. next chapt.). In 1949 comparative studies on polyploids of Hon-iku No. 162 were carried out on a large scale in various districts of Hokkaido. The results showed that, as expected, the 3x-A beets were superior to either 2x or 4x. It could be further predicted that intervarietal 3x hybrids, for instance, between a high yielding 4x variety and a disease-resistant 2x, would bring still better results. Therefore, in x hybrid beets between the tetraploid of the high sugar percentage Hon-iku No. 48 (called No. 4048) and the most widely grown variety in Hokkaido 2x Hon-iku No. 192, were studied in various districts. From these results an increase of about 10-15% in yield in 3x roots was ascertained, compared with 2x Hon-iku No. 192, in spite of the inferior germination rate of 3x seeds (about 10% decrease in standing numbers) (Mochizuki and Matsumura 1950; Matsumura and Mochizuki 1951). In order to solve the problem of inferior germination rate of 3x, various intervarietal triploid hybrids have been produced and compared with each other in 1951 and On the other hand, an embryological study of various polyploid varieties is being carried out, chiefly by the junior author. Also, taking into consideration that 3x seeds are markedly larger and have a smaller number of rootlets (radicles) per seed-ball than 2x, the practice in larger amounts (with a 20-30% increase in volume) was introduced. Table Ill. Germination test of various triploid hybrids (investigated with 100 seed-balls) to sow them Tables III-VI show comparative results of experiments conducted on the same scale -in 1951 in two experimental fields of the Japan Beet-Sugar Manufacturing. Company, Obihiroo at*i

5 June 1953 S. MATSUMURA and A. MOCHIZUKI : IMPROVEMENT OF SUGAR BEET 51 Shibetsu in Hokkaido. In these experiments various 3x seeds, obtained by planting of 4x and 2x beets in the ratio of 3 : 1, were used. The seeds were sown in the beginning of May and harvested in the middle of October. Table W. Measuring of several characters in various 3x hybrids in the middle of growth period The germination rate of the 4x seed-balls was variable ; it was very low in Hon-iku No. 48-4x (No. 4048), but relatively high in Hon-iku No x (No. 4398) and No x (No. 4402). Also in various 3x combinations the seed-balls obtained from No mother ;plants showed an obviously lower germination rate and a lower number of good germs with endosperm per seed-ball (Tab. III). In the middle of the growth period (middle of July), plant height and number of leaves were investigated (Tab. W). There was no significant difference in these characters between 3x hybrids and 2x beets (Hon-iku No. 192), while 3x beets had clearly a larger girth at the end of August than the 2x. At the same time susceptibility to diseases was investigated and denoted by an index Fig. 6. Comparison of sugar beets in 2x and 3x. Top : Hon-iku No x, the most widely grown variety in Hokkaido Bottom : 3x combination, Hon-iku No x ~No x. Xca. 1'8.5.

6 52 JAPANESE JOURNAL OF GENETICS Vol. 28-No. 2 (Tab. V). The larger the index, the more susceptible were the leaves. One triploid com. bination and its reciprccal one, No x 162-A and,-b, were most resistant of all, while the combinations No x 399 and No x 48 were most susceptible. Table VI shows results of the comparative studies in yield and sugar content of these 3x hybrids. In this case the combination No x 162 showed the best results (Fig. 6). Therefore, it is the most promising one. These results were also ascertained in experiments repeated on the same scale in Finally this combination was recommended for promotion as one of the best and called 3r-No. 1 by Hokkaido Agricultural Experiment Station and Hokkaido Seed Association. IV. Production of triploid seeds Triploid seeds of sugar beet can be obtained from reciprocal crosses between 4x and 2x plants. 4x plants have, in general, lower fertility and rather lower number of seed-balls than 2x. But the volume of seed-balls obtained from one plant is almost similar in 2x and 4x, because the seed-balls of 4x are larger than those of 2x. Offspring from only one 4x plant, surrounded by several 2x plants, were almost all triploid (Tab. VII) (Matsumura, Mochizuki and Akemine 1942). Similar results were gained from a circular arrangement of one 2x surrounded by several 4x plants (Mochizuki, unpubl.). Production of triploid -seeds in this way is practically difficult, unconvenient and uneconomic for large scale seed production by farmers. In x and 4x beets were planted in alternate rows. As mentioned above, seeds, called 3x-A and 3x-B, were obtained through natural pollination from the 4x and 2x mother plants, respectively. Somatic chromosomes of these offspring were observed and the, number of triploids in 3x-A was compared with that in 3x-B (Tab. VIII), with the result that beside 3x, many 2x (70%) were found in 3x-B, while only a few 4x (10%) in 3x-A. Therefore, the rate of 3x becomes 60% in a mixture of 3x-A and 3x-B seeds. The best way how to plant 4x and 2x beets was studied, in order to get as many 3x as possible. Fig. 7. Plan of planting 4x (0) and 2x beets (x) for production of 3x sends. A, in alternate rows (in the ratio of 1 : 1). B, in the ratio of 3 : 1. C, in the ratio of 8 : 1. Figure 7 brings three arrangements of the planted 4x and 2x beets. The somatic chromosome number of the offspring in 3x-A and 3x-B has been determined chiefly by the junior author. 4x and 2x beets are planted in schema B in the ratio of 3 : 1 and in schema C in the ratio of 8 : 1. In both cases the frequencies of 2x, 3x and 4x found in the offspring of 3x-A and 3x-B were similar to those found in the arrangement with alternate rows (in the ratio of 1 : 1). Therefore, 3x plants are contained in about 75% in the 3 : 1 ratio, if the seeds are collected from 4x and 2x mother plants together. Commercially, we can speak of them as of triploid seeds. Seeds in both 3x-A and 3x-B, obtained from the 4x and 2x mother plants planted in the ratio of 8: 1, had only a very low germination capacity 3x-A. These seeds could not be used in practice, especially in. For the large scale production of 3x seeds by farmers, it was found advisable to plant 4x and 2x beets in the ratio of 3: 1. Why do so many 2x plants appear in the offspring of 3x-B? This is a question of great

7 June 1953 S. MATSUMURA an9 A. MOCHIZUKI: IMPROVEMENT OF SUGAR BEET 53 Table V. Investigation of susceptibility to diseases in various 3x hybrids Table VI a. Comparison of yield and sugar content in various 3x hybrids (1951) Table VI b. Comparison of yield and sugar content in various 3x hybrids (1952)

8 .54 JAPANESE JOURNAL OF GENETICS Vol. 28 No. 2 importance which we must try to solve as soon as possible. Table IX. Comparison of x and 2x pollen-tube growth on 2x and 4x stigmata (unit : /-A) First the blooming season and the daily cycle of flowering were compared in 2x and 4x plants by the staffs of the Manufacturing Fig. 8. Number of opened flowers per day in 2x and 4x beets. Company in Obihiro and Shibetsu and by those of the Laboratory of Plant breeding, Hokkaido University, in Sapporo. Active blooming lasts almost 2 weeks in both, 2x and 4x plants. But 4x plants begin to bloom one week later than 2x. Therefore, in the first week of blooming 2x plants have to be pollinated by x pollen grains which results in the occurrence of many 2x plants in the offspring from the 2x mother plants. On the other hand, in the last week of blooming mainly 4x plants are bloom ing and only a few 2x. In spite of this fact, there were only a few 4x offspring from 3x-A, which of pollination were the results by 2x pollens. It is assumed that the x pollens elonagte their tube faster than most of 2x pollens (Tab. IX). Figure 8 shows the daily numbers of opened flowers in the 2x and 4x plants of Hon-iku No. 162, in an experiment which was arranged so that they' ibegan to bloom almost at the same time Fig. 9. Daily cycle in flowering of 2x and 4x. (July 9th). This investigation was carried out from July 14 th to August 7th in Obihiro. Relation between

9 June 1953 S. MATSUMURA and A, MOCHSZUK : IMPROVEMENT OF SUGAR BEET 55 time of blooming and number of flowers was generally similar in 2x and 4x plants. Figure 9 shows this relation for Hon-iku No. 162 and its tetraploid. From the graph it is clear that flowers open mostly before noon. opening flowers decrease gradually. Secondly, the germination There is a maximum at 8 o'clock and later the number of and tube growth rate of pollen grains were compared in 2x and' 4x plants. This study was mainly carried out by the junior author. 4x plants had mostly about 60% good pollen grains. But they had only about 60% germination capacity on an adequate culture medium, compared with those of 2x. A study on an artificial medium for pollen grains of sugar beets was being carried out in the Laboratory of Plant Breeding and the Hokkaido Agricultural Experiment Station. Table IX shows the tube growth of x and 2x pollen grains on the 2x and 4x stigmata. (Mochizuki 1950). It is doubtless that x pollen has a remarkably faster growth rate than 2x pollen, on both 2x and 4x stigmata. This could be partly a cause of the appearance of relatively many 2x plants from 2x mothers and of relatively few 4x plants from 4x mothers in the production of 3x seeds. author. Detailed reports on 3x seed production will be given on a later occasion by the junior V. Conclusion Triploid beets are more vigorous, grow better and always show a higher yield, a higher sugar purity, a higher sugar content and a higher resistance to diseases and attack by insects than the diploid Hon-iku No. 192, the most widely grown variety, according to our comparative studies carried out on a large scale in various districts of Hokkaido. Triploid seeds, however, are inferior in germination rate. In order to solve this problem, various intervarietal triploid hybrids were obtained and compared with one another. The germination rate of the tetraploid seeds was variable ; it was very low in Hon-iku No. 48-4x, but relatively high in Hon-iku No x and No x. On the other hand, Hon-iku No x and its triploid hybrids were remarkably susceptible to diseases. Further, in the triploid combination Hon-iku No x x Hon-iku No x root yield and sugar content showed about 10-15% increase, compared with those of Hon-iku No Therefore, this combination was most promising and was recommended as the best variety, called 3n-No. 1 by the Hokkaido Agricultural Experiment Station and the Hokkaido Seed Association. It is now planned to replace in several years the common varieties with this combination on about 1/4 (ca. 5, 000 ha) of the Hokkaido beet area. For the large scale production of triploid seeds, the self-incompatibility of this plant could be utilized. By planting of rows with 4x and 2x mother plants alternately, 3x seeds could be obtained through natural pollination from 4x or 2x mother plants. The frequency of 3x plants, however, varies to some extent under the influence of some factors, such as : ratio between the 2x and 4x plants, difference in blooming season and difference in pollen-tube growth rate of the parents. It was found advisable to plant 4x and 2x beets in the ratio of 3 : 1. In this way relatively many 3x plants (ca. 75%) were obtained. In this case it will be surely necessary to synchronize the blooming season of 4x and 2x plants.

10 56 JAPANESE JOURNAL OF GENETICS Vol. 28 No. 2, Acknowledgement The writers beg to express their indebtedness to Prof. H. Kihara, Laboratory of Genetics, Kyoto University, for his guidance and encouragement. Sincere appreciation is extended to Prof. S. Nagao, Laboratory of Plant Breeding, Hokkaido University, for his valuable suggestions. We also wish to tha:ik the staffs of the Laboratory of Plant Breeding, the Hokkaido Agricultural Experiment Station and the Japan Sugar-Beet Manufacturing Company for the help in field experiments carried out in various districts of Hokkaido. The prese zt study was supported by grants from the Ministry of Education (Aid to Developmental Scientific Research) and from the Research Fund of Hokkaido Sugar-Beet Prod'action Prc moting Ass: ciation. Literature.Matsumura, S. and Mochizuki, A Improvement of sugar beet by means of triploidy. (Japanese) Rep. of Hokkaid8 Sugar-Beet Prod. Prom. Ass. No. 3. and Akemine, T Genetische and zytologische Untersuchungen bei Beta-Arten. II. (Japanisch mit Deut. Zusammenf.) Rep. Kihara Inst. Biol. Res. No and Suzuka, Idem. III. (Japanisch mit Deut. Zusammenf.) Ibid. No 4. and Yamashita, K Idem. I. (Japanisch mit Deut. Zusammenf.) Journ. Manchurian Agron. 3. imochizuki, A Pollen tube growth in polyploid sugar beet. (Japanese). Jap. Journ. Gen and Matsumura, S Triploid sugar beet and its applications. (Japanese) Ibid. 24. and Genetic and cytological studies on the genus Beta. IV. (Japanese with Engl. sum.) Rep. Kihara Inst. Biol. Res. No. 4. Rasmusson, J. and Levan, A Tetraploid sugar beets from colchicine treatments. Hereditas 25.