IMPROVEMENT OF GRASSES THROUGH INDUCED CHROMQSO’MAL RECOMBINATIONS Thesis for the Degree of DB. D. MICHIGAN STATE UNIVERSITY Don J. Heinz 1961 This is to certify that the thesis entitled .L ILIPROVELCEIIT OP GRASSES THROUGH IZTDUCED CHROZIOSGI-IAL itECOi-{BIETATIOLIS presented by Don J. Heinz has been accepted towards fulfillment of the requirements for Ph. D. degree in Farm Crops / ‘\ I 1 Q“ (C 61 i @ng Major professor / Date Ifovem‘oer 3, 1901 0-169 LIBRARY Michigan State University ‘wva—v—v—v—w “\ ABSTRACT IMPROVEMENT OF GRASSES THROUGH INDUCED CHROMOSOMAL RECOMBINATIONS by Don J. Heinz Polyploid and structural hybrid systems have evolved naturally giving certain genera greater survival value over that of ancestral types. Consideration should be given to the use of these systems in applied breeding. The purpose of this study was to determine 1) the effects of irradiation (source of irradiation was neutrons) induced translocations in tetraploid Dactylis glomerata and 2) the results of in- duced polyploidy in Lolium species. Irradiation in Dactylis resulted in a wide range of translocations, as measured by the number of multivalent chromosome associations of more than four at diakinesis. One-half-hour of irradiation was more efficient than the one- fourth or one-hour treatments in the induction and survival of translocations. The morphological and fertility charac- teristics of the K2 population (one-half-hour treatment) were not affected. Three plants heterozygous for a large number of translocations and more vigorous than any found in the control population were isolated. There was a sig- nificant difference at the 20% level between the three largest plants heterozygous for translocations and the three Don J. Heinz largest control plants. A breeding system for polyploid forage grasses, based on structural hybridity has been proposed. Seed from many inter- and intraspeeific crosses of Lolium multiflorum and L. oerenne were treated with colchicine. Also, part of the seed was given an irradiation treatment in addition to the colchicine treatment. Tetraploid sectors were obtained from mature plants derived from treated seed. Progeny obtained from corresponding diploid and tetraploid clones were studied in the field. Greater rust resistance was found in induced tetra- ploid Lolium multiflorum populations over corresponding dip- loid populations. Some tetraploid plants were more vigorous than any found in the diploid populations. The vigor of Lolium multiflorum was combined with the rust resistance of L. perenne in diploid and tetraploid hybrids. Tetraploid hybrids in the first generation were not superior to corresponding diploid hybrid populations. There were more differences between the original clonal lines than between the corresponding diploid and tetraploid populations. However, further crossing and selection might result in progress at the tetraploid level. Close agreement was found between autotetraploid Lolium multiflorum and the hybrids of L. multiflorum X p. perenne in various meiotic configurations, suggesting that cytogenetically the two species do not have many chromo- somal differences. IMPROVEXEHT OF GRASSES THROUGH INDUCED CHROMOSOHAL RECOHBIEATIOES By Don J. Heinz A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Farm Crops 1961 ACKNOWLEDGEMENT The author wishes to express his sincere gratitude to Doctor Fred C. Elliott for his guidance throughout the course of this study and for his helpful advice in the prep- aration of the manuscript. The author is grateful to Doctor Carter H.Harrison for his advice and assistance during the preparation of the manuscript. Also, the author is very grateful to his wife Marsha for her encouragement during the course of this study and for typing this manuscript. ii TABLE OF CONTESTS LIST OF TABLES . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . IITTRODUCTIOISI O C C O O O O O O O O O O 0 REVIEW OF LITERATURE . . . . . . . . . . PART ONE - IRRADIATIO] STUDIES ID DACTYLIS Materials and Methods Results PART TWO - POLYPLOID STUDIES IR LOLIUI . Materials and Kethods Results DISCUSSION . . . . . . . . . . . . . . . COl‘u—‘CLUSIOZTS o o o o o o o o o o o o o o 0 LITERATURE CITED . . . . . . . . . . . . iii 12 12 18 41 41 46 55 67 69 LIST OF TABLES Table Page 1. Neutron source for P-11 UT—AEC: West animal tunnel, graphite reactor, ORIL, with U235 plate . . . . . . . . . . . . . . . 13 2. Ranked means of the percentage of stainable pollen in the F1 and R1 . . . . . . . . . 19 3. Ranked means of percentage of chromosomes involved in bivalent associations and average bivalent association per micro- SpOI‘OCyte in the F1 and 1‘31 0 o o o o o o o 20 4. Ranked means of percentage of chromosomes involved in quadrivalent associations and average quadrivalent associations per microsporocyte in the F1 and N1 . . . 21 5. Ranked means of percentage of chromosomes involved in multivalent associations over four in the F1 and R1 . . . . . . . . 21 6. Ranked means of percentage of fertility in the F1 and I‘I‘l o o o o o o o o o o o o o 28 7. Percentage of plants with under ten culms, range in culms per population, for plants from P.I. 237174, P.I. 251112, FrBde, Brage, and a cross 5 X 35-38 . . . . . . . 29 8. Ranked means of percentage of germination in the F2 and N2 . . . . . . . . . . . . . 3O 9. Ranked means of percentage of survival in the F2 and 151.2 0 o o o o o o o o o o o o o 30 10. Ranked means of percentage of vigor in the F2 and N2 0 a o o o o o o o o o o o o o o 31 11. Number of plants analysed in the F2 and N2 populations with the number of plants heterozygous for translocations . . . . . 32 12. Means of percentage of fertility in the F2 and N2 0 o o o o o o o o o o o o o o o 33 iv Table 13. 14. 15. 16. 17. 13. :19. Page Ranked means of the average green weights in grams of individual plants in the F2 and N2 0 o o o o o o o o o o o o o o o o 33 Plant weights in grams and percentage of fertility for the three heaviest single plants from the irradiated (% hr.) and con- trol populations in the F2 and R2 . . . . 34 Ranked means of percentage of plants free of rust lesions in diploid and tetraploid Lolium multiflorum populations . . . . . 47 Types of pairing in diploid and tetraploid Lolium multiflorum . . . . . . . . . . . 48 Ranked means of individual plant dry weights in grams in diploid and tetraploid Lolium multiflorum populations . . . . . . . . . SO Ranked means of dry weight in grams of indi- vidual plants from diploid and tetraploid hybrid Lolium multiflorum X L. perenne populations . . . . . . . . . . . . . . . 52 Types of pairing in diploid and tetraploid hybrids of Lolium multiflorum X L. 2222222 . . . . . . . . . . . . . . . . . 53 LIST OF FIGURES Page Microsporocytes from the control popula- tion showing range in bivalent and quad- rivalent formations observed: a. 14 II; b. 12 II, 1 IV; c. 10 II, 2 IV; d. 8 II, 3 IV; e. 6 II, 4 IV; f. 4 II, 5 IV; g. 2 II, 6 IV; h. 7 IV . . . . . . . . . . 22 MicrOSporocytes from plants of the N1 pop- ulation showing range in multivalent formations over four: a. 5 II, 1 chain of 18; b. 4 II, 1 IV, 1 chain of 16; c. 1 II, 3 IV, 1 chain of 14; d. 6 I, 5 II, 3 IV; e. 6 II, 2 IV, 1 chain of 8; f. 6 II, 1 IV, 1 ring of 12; g. 7 II, 1 IV ("Frying pan" formation), 1 chain of 10 . . . . . . . . . . . . . . . . . 25 MicrOSporocytes from plants of the N1 pop- ulation showing irregularities at anaphase I: a. unequal separation 13:15; b. unequal separation 12:16; 0. chromatid separation . . . . . . . . . 27 Normal anaphase I showing equal separa- tion to both poles . . . . . . . . . . . 27 Plant number one with a general view of the F2 and N2 populations . . . . . . . 35 Microsporocyte from plant number one, show- ing 7 II, 2 IV, and 1 ring of 6 . . . . 35 Plant number two heterozygous for trans- locations! 0 O 0 O O O O O O O O O O O O 36 Kicrosporocytes from plant number two: a. 8 II, 1 IV, 1 chain of 8; b. 7 II, 1 ring of 10; c. 2 I, 2 II, 1 IV, 1 ring of 8, 1 chain of 10; d. meta- phase showing separation of chain of 8 . 36 Plant number three which had mierosporo- cytes with a ring of 6 and a chain of 8 O O O O O O O O O O I O O O O O O O O 39 vi Figure Page 10. Plant showing tetraploid sector. This was one of the methods used to separate tetraploid from diploid tissue, prior to cytogenetic analysis . . . . . . . . 43 11. MicrOSporocytes from tetraploid Lolium multiflorum plants: a. 5 I, 3 II, 1 III, 2 IV, 1 ring of 6; b. 2 I, 9 II, 2 IV; c. anaphase showing chromatid segregation and laggards . . . 49 12. MicrOSporocytes from a monosomic (2n:27) hybrid plant: a. 1 I, 7 II, 3 IV; b. chromatid segregation at anaphase I . 49 13. Iicrosporocytes from a triploid hybrid plant: a. 3 I, 7 II, 1 IV; b. anaphase . 54 14. Hicrosporocytes from diploid hybrid plants: a. 4 I, 5 II; b. 7 II . . . . . 54 15. Normal metaphase from a tetraploid hybrid plain-t: 6 II, 4 IV 0 o o o o o o o o o o 54 vii INTRODUCTION There are many grass species in use on farms, ranches, and rangelands in the United States which have been derived directly from introductions or native Species. Some species have been improved through breeding, but progress has been slow. Detailed genetic and cytogenetic information is limited and only a few have been studied to any great extent. Also, there is a general lack of agreement concerning the evaluation and definition of desirable forage characteristics. Some gains have been made through breeding in disease resistance, leafiness, seed yields, and date of maturity, while total dry weight yields remain constant. Host new varieties being released are not significantly higher in yield from those already in use. A better knowledge of the genetic system of the grasSes is needed. This associated with new methods and techniques should give greater progress through breeding. Polyploid and structural hybrid systems have evolved naturally giving certain genera greater survival value over that of ancestral types. Consideration should be given to the use of these systems in applied breeding. Efforts to apply these techniques through artifical means have failed in many cases. Lack of success was most likely due to short range programs, with immediate improvement the goal. When 2 immediate success was not obtained, the technique was written off as not being of value. New techniques should be used in combination with other methods to give superior plant types. The purpose of this study was to determine 1) the effects of radiation-induced translocations in tetraploid Dactylis glomerata, and 2) the results of induced polyploidy in Lolium species. REVIEW OF LITERATURE Cytogenetics and Breeding_of Fogage Grasses. Several reviews have been published on the subjects of genetics, cyto- genetics, and breeding problems in perennial forage grasses. The divergence of opinion held by forage breeders concerning the objectives to be attained and breeding methods used in the improvement of forage grasses have been reviewed by Hanson and Carnahan (1956). Cytogenetic and genetic studies made on forage grasses have been reviewed by Myers (1947) and Carnahan and Hill (1961). Myers lists two reasons for the importance of cytogenetical studies in forage grasses: "(a) to serve as an adjunct to morphological data in studies of taxonomy and phylogeny and (b) to provide fundamental information for the improvement of species by breeding." Effects of Igradiation on Plant Yields. A large num- ber of studies have been made concerning the effects of ion- izing radiations on plants and their use in plant breeding (Mac Key, 1956; Konzak, 1957; Sax, 1957; Sparrow, 1957; Elliott, 1958; Gaul, 1958; and Smith, 1958). Increased yields through selection within irradiated populations have been reported by several workers. Gregory (1955, 1956) was able to demonstrate a measurable gain in 4 peanut yields as a result of selection within x-irradiated populations for more productive types. He attributed this gain in yield to the induction of valuable mutations. Humphrey (1954), studying the effects of thermal neutrons on soybeans, noted a number of lines with increased vigor, greater shattering resistance or both, and one line with significantly higher yields. Papa §t_gl. (1961) re- ported no substantial yield increase from selection within irradiated soybean lines, but did find several high yielding progenies from irradiated plants that combined high yield with additional desirable qualities. Mac Key (1954) reported improved agronomic character- istics in wheat as a result of radiation induced deficiency duplications. He also reported increased yields in wheat and oat lines selected from irradiated populations. Gelin (1954) released a new higher yielding pea vari- ety selected from irradiated lines. He associated the increased yields with induced mutations. Andersson and Olsson (1954) reported the release of a higher yielding white mustard variety as a result of selec- tion within irradiated lines. Evolution of Structural Hybridity. Structural hybrid- ity in the form of translocations has played an important evolutionary role in certain plant genera. The most studied of these groups has been Oenothera, in which structural hy- bridity is a dominant and important feature. Stebbins (1950) 5 and Burnham (1956) list a number of other plant genera possess- ing similar structural features, but not as extensive or com- plex as found in Oenothera. Swanson (1957) has presented the main features of the Oenothera system in a short and concise manner. Basically the system evolves around the formation of a ring of 14 chromo- somes at meiosis due to the accumulation of translocations. Alternate disjunction occurs giving two complexes (called gaudens and valens in Q. lamarckiana and regins and curvans in Q. muricata) of seven chromosomes which function as sep- arate linkage groups. The translocation heterozygote has been maintained through the evolution of a balanced lethal system and self fertilization. In Oenothera lamarckiana the balanced system acts as a zygotic lethal, as contrasted to a gametic lethal system in Q. muricata. Only translocation heterozygotes are formed. Cleland (1950) has summarized the evolutionary fea- tures of Oenothera. On the basis of extensive cytogenetic studies he has proposed the following steps in the evolu- tion of structural hybridity in Oenothera. (1). The appearance and incorporation of translo- cations until all 14 chromosomes were in a single ring at meiosis. (2). The appearance of lethals, both gametophytic and zygotic. When these were coupled with large rings the population tended to increase in the number of heterozygous individuals, until the presence of only heterozygotes was 6 assured through balanced lethals. (3). The establishment of self fertilization. This may have helped overcome sterility, which probably was asso- ciated with balanced lethals and large rings of chromosomes. (4). The reduction in flower size, making the plant less efficient for insect pollination. The occurrence of large rings, balanced lethals, and self fertilization in Oenothera gives a high survival rate, but severely restricts chromosome or genetic recombination. Several cases of the artificial formation of struc- tural hybrids have been reported as in Campanula persicifolia (Darlington and Gairdner, 1937) and corn (Burnham, 1956), but only a portion of the total chromosome complement was involved in large rings. Nishimura and Kurakami (1953) reported the production of large rings in barley by successive x-ray treatments in an effort to apply the Oenothera system to barley. Yamashita (1951) produced a structural heterozygote in einkorn wheat involving all 14 chromosomes in one large ring. Qytoggnetic Studies in Dactylis. A number of cyto- genetic studies have been made on Dactylis. Quadrivalent formation at diakinesis and metaphase I have been reported by a number of workers. Muntzing (1933 and 1937) reported an average of 3.48 and 3.5 quadrivalents per microsporocyte. He reported no pollen mother cells with greater than quadri- valent associations. Myers and Hill (1940) reported an average of 3.3, 3.8, and 4.2 quadrivalents per micrOSporocyte for three plants. From a study of twenty plants, Myers and Hill (1942) reported the average number of quadrivalents per sporocyte ranged from 2.42 to 4.39. Myers (1943) found a range of 3.0 to 4.09 quad- rivalents in nine plants. McCollum (1958) reported a mean number of 3.44 quad- rivalents at metaphase versus 3.5 at diakinesis for artifi— cially induced tetraploids from crosses between diploid sub- species of Dactylis. Mfintzing (1937) reported uneven distribution of chromosomes in 15.3% of the microsporocytes examined at ana- phase I. Kyers (1943) observed unequal distribution of chromosomes in 13 out of 520 microsporocytes examined at anaphase I. Myers and Hill (1942) found that the number of micro- sporocytes with univalents varied from 3.4 to 13.8% per plant at metaphase I, and from 1.7 to 34.0% of microsporocytes of different plants at anaphase I had lagging chromosomes. Myers (1943) found no correlation between the number of quad- rivalents found and number of univalents at anaphase, but re- ported a correlation between univalents at metaphase I and laggards at anaphase I. The number of bivalents may range from 0 to 14 and the frequency of quadrivalents from 7 to 0 per microsporocyte. There have been no reports of multivalents found with greater than four chromosomes. Irradiation Studies in Dactylis. There are only a few reports on the effect of irradiation in orchardgrass. Stebbins and McCollum (1955) reported the induction of trans- locations by X—irradiation on diploid DactVliS. 213? (1960) reported the possible induction of a mutant giving a plant free of silica deposits on the leaves. Mac Key (1956) mentions that mutation studies are being conducted on orchardgrass by workers in Sweden. Osborne (1957) lists several workers who have had orchard- grass seed irradiated with neutrons, but no reports have been published on their work to date. Polyploidy Studies in Lolium. Polyploidy has been an important factor in the formation of new races and species in nature. Many of our most important crop plants are of polyploid origin. Carnahan and Hill (1961) estimated that of some 2,300 grass Species for which the chromosome numbers are known, approximately 80% are polyploid. Lolium is one of few genera in the Gramineae in which all known Species are diploid (2n:14). In recent years interest has been shown in polyploidization of Lolium Species for the improvement of cer- tain agronomic characteristics. Mehta and Swaminathan (1957) have reviewed studies on the induction of polyploidy in forage crops, including work with Lolium. Therefore, this review will be limited to recent work in polyploidization of ryegrass. 9 Hit (1958) and Hertzsch (1959) reported the produc- tion of large numbers of tetraploid Italian ryegrass plants. They found higher green weights, a lower total percentage of dry weight, lower numbers of leaves and of culms in the tetra- ploids when compared to their diploid counterparts. They re- ported a potential for greater winter hardiness in selected tetraploid plants over that observed in diploids. A reduc- tion in fertility was observed in most newly established tetra- ploid lines. De R00 (1959) reported that progress could be made for fertility and other important agronomic characteristics by selection in tetraploid lines of Italian ryegrass. Hit and Speckmann (1954) found greater rust resistance in tetraploid Italian ryegrass after one generation of selec- tion, while in perennial ryegrass rust resistance was im- proved over the diploid counterpart only after repeated selec- tion. Wit (1959) reported that tetraploids of Italian rye- grass surpassed the diploids in winter hardiness, forage yield, and seed production; while in perennial ryegrass, cold resistance, forage yield, and rust resistance could be raised to or above the diploid level in as little as one cycle of selection. The development of induced tetraploid ryegrass varie- ties haS been enhanced by crossing doubled parents (induced tetraploids) from several Sources of germplasm. In fact, Hit (1959) has submitted the first tetraploid Italian rye- 10 grass varieties for registration in the Netherlands. The production of hybrid varieties from interSpecific crosses of Lolium multiflorum X p. perenne has been of inter- est in recent years. Through selection and stringent seed certification programs, the two Species have developed into distinct morphological types. ‘2. multiflorum is characterized as an annual or Short lived species producing a palatable, high yielding growth, while p. perenne is a perennial Species, less palatable, and with less herbage. The two Species are completely interfertile, and from natural or controlled crosses superior plants can be obtained combining the desirable qualities of both species. However, upon further cross pollination among the hybrids, segregation occurs giving parental types and inferior intermediate types. Several hybrid varieties have been released at the diploid level resulting from selection within large progenies, in com- bination with strict standards for seed multiplication. Corkill (1945) was successful in producing a hybrid variety between the two Species combining their desirable characteristics. He mentioned that strict controls must be maintained during seed multiplication, as there was a tendency for early flowering, less desirable types to become predomi- nant in the variety. Kuchar (1957), Desroches (1958), and the Dutch seed establishment of H. Mommerstug (1957) have reported the suc- cessful production and selection of desirable hybrids that may be useful as varieties. 11 Although several workers have tried to combine the desirable characteristics of Lolium multiflorum and L. perenne, there has been no report of utilizing polyploidy for this purpose. PART ONE IRRADIATION STUDIES IE DACTYLIS KATERIALS AND METHODS Source material for this study was obtained from Six superior clones selected from individually Spaced plants of orchardgrass from P.I. 237174. Open pollinated seed was harvested from all six plants. Equal numbers of seed were then selected from each plant for control and irradiated groups. The seed was irradiated at the Oak Ridge, Tennessee, facilities of the Atomic Energy Commission, courtesy of T. S. Osborne of the UT-AEC Agricultural Research Laboratory. A total flux of 7.7 X 108 N/ch/sec was administered for one- fourth, one-half, and one-hour treatments. Table 1 gives a breakdown of the neutron source. Of the "effective" flux, 93.9% was between 0.025 Kev (epithermal) and 8.6 Kev (fast). The amount of gamma contamination was unknown. Neutrons were used as the irradiation source, Since it has been Shown previously by several workers (Caldecott, 1955; D'A‘nato 2353;” 1958; and Larter and Elliott, 1956) that a higher frequency of translocations resulted per irradi- ated plant surviving than from other sources of ionizing rad- iation. 13 AQOfiprHssssoo Hemomwomv mswopmo .m masons .Hn nopmd Ao.oo—v m.m~¢_ Ao.oo—v oo_ N >.N "manpoa m.m w.w¢ m.o mo? M ¢.m 00— H ©.w ad o.o_ ©.>¢F ~.F NOF N m.F mop N m.m m P.5F o.mmm m.¢ sop N m.m mo? N m.— D m.se o.mos m.el mop a l._ mo. w m.» as o.m_ w.om >.¢m mo, x m.F ror N m.m 5m gruesome v m.m m.m¢ m.¢m @0— N m.¢ mnowlm m.m 54 >o mom WMPOH ma\mmm. NsHm WMpoa oom\mao\msOHp:oz camgmmsge Haom pqoonm anoommm .Ndam was» so HHom popaSoa opoae mmm: near . HMO .Hopowmh mpaflmmnm .HmSSSp HmsHsm 9mm: «OM41HD Fwnm Mom oohfiom noHPSoz .— mamas 14 Control and irradiated seed (herein referred to by the letter N) were germinated under controlled conditions (300 C. for 8 hours and 200 C. for 16 hours, with an eight- hour light period) for 21 days. The seedlings were individ- ually tranSplanted into two-inch peat pots filled with a 1:1 mixture of peat and sand containing essential growth nutrients. When a majority of the seedlings had approximately 15 to 20 leaves, they were vernalized for three months in a refrigera- tor maintained at 35 to 400 F., with a nine-hour light period every 24 hours. The progeny (F1 and N1) obtained from plant number two was used to carry out this study. Three hundred seed were designated control, with 120 in each of the one-fourth, one-half, and one-hour N treatments. The seedlings were transplanted into Six-inch pots after vernalization. Early in the Spring of 1960 the plants were hand tranSplanted to the field. Each treatment was placed in an isolated block to insure against contamination from outside sources of pollen. Two panicles on each F1 and N1 plant were tagged dur- ing the period of heaviest anthesis. Seed for the F2 and N2 pOpulationS were harvested from these tagged panicles. These same panicles were also used to determine fertility. Sixty seed from the following number of F1 and N1 plants were germinated and the seedlings were used to make up the F2 and H2 populations: Control: 75 plants Irradiated % hour: 54 plants Irradiated % hour: 75 Plants Irradiated 1 hour: 54 plants Germination conditions were the Same as for the F1 and N1 populations. The seedlings were tranSplanted into two-inch peat pots and allowed to develop in the greenhouse under a 24 hour light regime prior to mechanical tranSplant- ing to the nursery on September 10, 1960, water being applied in the process. Fertilizer was applied at the rate of 350 pounds of 15-15-15 per acre in the Spring of 1961. The field design consisted of three completely ran- domized blocks. Each block had 258 plots with eight plants per plot from one of each of the F1 and N1 parents. The plants were individually space planted on two-foot centers in three-foot rows. Thus, there were 24 progeny represented in the design from each of the F1 and N1 plants included in this study. The percentage of stainable pollen was ascertained in the F1 and N1 populations at the time of anthesis. Pollen from one panicle branch was Shaken into aceto-carmine, and a cover slip placed over the preparation. An immediate count was made of the stainable and nonstainable pollen from five different positions on the Slide. Percentages of fertility were determined for the F1, H1, F2, and N2 populations by counting the number of florets and seed set on the three center branches of one pan- 16 icle per plant. Survival and vigor ratings were taken in the Spring of 1961. All plants failing to survive up to October 1, 1960, were replaced. Vigor was scored on the basis of 0 - 5, with 0 : dead and 5 = most vigorous. For the statistical analysis of Spring vigor the ratings were changed to a per- centage of vigor. In the fall of 1961 individual plants were cut approx- imately 1.5 inches above the crown, weighed, and average plant weights determined. Statistical analyses of field data from the F2 and N2 populations were made on averages for the 24 progeny represented from each original parental clone. Cytogenetic analysis was made on microsporocytes from florets of individual panicles collected when they were com- pletely out of the boot, between 1:15 and 2:30 P.M. on sunny days. Whole panicles were fixed and stored for up to ten months in Newcomer's solution which consisted of the follow- ing: 6 parts 98% isopropyl alcohol 3 parts propionic acid 1 part petroleum ether 1 part acetone 1 part dioxane Cytogenetic analyses were made from temporary propionic carmine smears of pollen mother cells from plants of the F1, N1, F2, and N2 populations. Photomicrographs were 17 made from these preparations. Duncan's Multiple Range Tests for means with equal (1955) and unequal (1957) replications were used to test differences between means for significance. In these tests, differences required for Significance between means vary with the number of means in the comparison. Furthermore, the differences required increase as means further apart in rank are compared for Significance. Means in the same range are not Significantly different; however, a significant difference exists between means found in different ranges. Critical values for the Duncan's Multiple Range Tests were taken from tables developed by Harter (1960). All results analysed as percentages were transformed by the arcsin transformations for proportions as worked out by C. I. Bliss and tabled by Snedecor (1956). When the ob- servations in an analysis of variance problem are proportions, homogeneity of variance cannot be assumed because 0’; varies with P (proportion) and with N (number). Homogeneity of vari- ance is necessary for a valid analysis of variance. Both the transformed and actual percentages are presented in the tables. RESULTS There were no Significant differences in germina- tion between treatments for the progeny grown from the Six original plants. The percentage of germination for the progeny of plant number two which made up the F1 and N1 populations was as follows: Control: 90.7 Irradiated & hour: 85.0 Irradiated é hour: 93.3 Irradiated 1 hour: 87.5 The percentage of F1 and H1 progeny from plant num- ber two surviving to maturity was as follows: Control: 81.7 Irradiated % hour: 80.0 Irradiated % hour: 85.0 Irradiated 1 hour: 85.8 In general there were very few morphological indica- tions of induced mutations in either generation of the irradi- ated populations. One plant in one of the N1 populations Showed a chlorophyll deficiency. It was visible early in the Spring with a gradual disappearance of the characteristic during the summer. This is what would be expected from cross pollinated plants, Since most mutations are recessive and 18 19 would not be expressed in the phenotype unless they were homozygous for the new mutant. The only way for homozygosity to occur in these populations would be through selfing or close inbreeding. The percentage of stainable pollen for the F1 and N1 is presented in Table 2. Although Significant differences were obtained between populations, the range in percentages of stainable pollen from individual plants of the populations was similar. The control had a range of 65.3 to 98.9%, the one-half-hour treatment 63.8 to 100%, and the one-hour treat- ment 56.2 to 97.4% stainable pollen. The use of the percent- ages of stainable pollen as an indicator of induced transloca- tions in individual plants was negated by the Similarity found in the range for stainability in the F1 and N1 populations. TABLE 2. Ranked means of the percentage of stainable pollen in the F1 and N1 Control % hr. 1 hr. T t t rea men (F1) (N1) (N1) Means: Transformed 74.7 71.5 63.5 Actual 93.0 89.2 78.8 Lsn 1% level1 r = 69.89 P value = 24.31** SE = 1.18 c.v.% = 10.23 I ** Significant at the 1 percent level. 1All means underlined by the same line are not Significantly different from each other. All means not underlined by the same line are significantly different from each other. 20 The cytogenetic analysis of the F1 and N1 population for induced translocations was based on the number of chromo- somes associated in multivalents over four. Analysis was made at diakinesis on microsporocytes from 25 plants each of the control, one-half, and one-hour treatments. A total of 2,637 microsporocytes were analysed at diakinesis and ana- phase. Tables 3, 4, and 5 give the ranked means for the per- centage of chromosomes associated as bivalents, quadrivalents, and multivalents over four per microsporocyte in F1 and N1 populations as well as the number of bivalent and quadrivalent associations per pollen mother cell. TABLE 3. Ranked means of percentage of chromosomes involved in bivalent associations and average bivalent association per microsporocyte in the F1 and N1 V' Control 1 hr. hr. Treatment (F1) (N1) (N1) Means: Transformed 53.4 45.4 45.2 Actual 64.2 50.7 51.1 BivalentS/micrOSporocyte 8.98 7.08 7.06 LSR 1% level r = 48.0 F value = 20.56** SE = 1.03 c.v.% H O \1 .q an Significant at the 1 percent level. The number of bivalents and quadrivalents found in microsporocytes from control plants were in agreement with those reported in the literature (see Mfintzing, 1933 and 1937; Myers and Hill, 1942; Myers, 1943; and McCollum, 1958). 21 Figure 1 shows the range in the number of bivalent and quad- rivalent associations per microsporocyte observed at diakine- sis in the control population. Several plants were observed having microsporocytes with a range of 14 bivalents through seven quadrivalents. Very few univalents or trivalent associa- tions were observed. TABLE 4. Ranked means of percentage of chromosomes involved in quadrivalent associations and average quadrivalent associa- tions per microsporocyte in the F1 and N1 hr. Control' 1 hr. Treatment N1) (F1) (N1) Means: Transformed 39.7 36.1 34.8 Actual 40.6 35.0 32.9 QuadrivalentS/microSporocyte 2.88 2.47 2.28 1% level LSR 5% level Y = 36.85 F value = 7.69** 52 = .93 c.v.% 12.57 ** Significant at the 1 percent level. TABLE 5. Ranked means of percentage of chromosomes involved in multivalent associations over four in the F1 and N1 1 hr. % hr. Control Treatment (N1) N1) (F1) Means: Transformed 22.3 14.3 1-3 Actual 16.0 7.4 .3 LSR 1% level X'z 12.63 F value : 54.64** Si : 1.44 C.V.% : 56.81 ** Significant at the 1 percent level. 22 Fig. 1. Microsporocytes from the control pop- ulation Showing range in bivalent and quadrivalent formations observed: a. 14 II; b. 12 II, 1 IV; c. 10 II, 2 IV; d. 8 II, 3 IV; e. 6 II, 4 IV; f. 4 II, 5 IV; g. 2 II, 6 IV; h. 7 IV. 24 The percentage of chromosomes associated as bivalents at diakinesis was reduced in N1 populations. There was a wide range in multivalent associations over four in the N1 populations (Fig. 2a - g). The largest association observed was a chain of 18 (Fig. 2a). The most common associations were those of six, eight, and ten in order of decreasing frequency. Cytogenetic analysis of micrOSporocyteS at anaphase I Showed a higher frequency of irregular anaphase divisions in the N1 populations as compared to the F1. The F1 had 88.5% regular divisions, the one-half-hour treatment 65.8%, and the one-hour treatment 63.2%. The majority of the irreg- ularities involved unequal distribution (Fig. 3a and b) and laggards showing chromatid separation (Fig. 30). Some bridg- ing or stickiness was observed at anaphase I, but there was no evidence of accompanying fragments. The percentage of fertility for the F1 and N1 is presented in Table 6. Fertility counts represent seed set from cross fertilization between plants in isolated treat- ment blocks. In this study there was little difference in survival and general vigor between F1 and M1 populations, although a reduction in stainable pollen and percentage of fertility was found in the one-hour treatment of the N1 population. The number of culms per plant from the F1 and N1 progeny of P.I. 237174, P.I. 251112, and the Swedish varie- ties Frade and Brage were counted to obtain a measure of Fig. 2. Microsporocytes from plants of the 21 population Showing range in multivalent formations over four: a. 5 II, 1 chain of 18; b. 4 II, 1 IV, 1 chain of 16; c. 1 II, 3 IV, 1 chain of 14; d. 6 I, 5 II, 3 IV; e. 6 II, 2 IV, 1 chain of 8; f. 6 II, 1 IV, 1 ring of 12; g. 7 II, 1 IV ("Frying pan" formation), 1 chain of 10. - \