Cytological Investigation of Parents, Offspring and Backcross Derivatives Involved in the Interspecific Cross Phaseolus lunatus L. X P. polystachyus (L.) B.S.P. By BIRENDRA SINGH FOZDAR A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA August, 1962 ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation for the constant advice and guidance given by Dr. A. P. Lorz, Horticulturist, during the preparation of this dissertation. To Dr. V. F. Nettles, Dr. J. R. Edwardson, Dr. Yoneo Sagawa and Mr. L. H. Halsey for their kindly advice and helpful criticisms, his deep appreciation is extended. He wishes to extend his thanks to Dr. W. 0. Ash for his suggestions regarding statistical analysis. He appreciates the cooperation and help received from all the members of the Department of Vegetable Crops, especially for the financial help and encouragement received from Dr. F. S. Jamison, Head of the Department of Vegetable Crops. The assistance of Mrs. W. F. Johnson in the preparation of the final manuscript is appreciated. To my wife Vidya, my thanks are due for her constant encouragement . -ii- t ?? 70 TABLE OF CONTENTS ACKNOWIEDGEMENTS . . LIST OF TABLES . . . LIST OF ILLUSTRATIONS INTRODUCTION .... REVIEW OF LITERATURE materials and methods Plant Material. . , Technique . . . . , Page ii iv vi 1 5 17 17 20 RESULTS AND OBSERVATIONS 29 Phaseolus lunatus var. Fordhook progeny Ehaseolus. polys tachvu.s F]_ Hybrid ! ! ! ! The Amphidiploid and Progeny # . # — • var. Fordhook x P. rolystachyus "sterile," diploid ?2 (category 5) Backcrosses and Progenies 29 35 43 52 77 78 DISCUSSION 86 Affinities of Parental Species 86 Origin of the Anphidiploid 88 Amphidiploidy with Respect to Speciation and Breeding 91 SUMMARY LITERATURE CITED . APPENDIX BIOGRAPHICAL SKETCH -iii- LIST OF TABLES Table Page 1. Comparison of unit characters of P. nolystachyus . amphidiploid derivatives (categories 4, 10 and 11) and P. lunatus var. Fordhook 30 2. Frequencies of chromosome configurations in P>€'s of P. lunatus var. Fordhook at diakinesis and Mj. . 34 3. Pollen grain diameters of P. lunatus var. Fordhook in microns 38 4. Frequencies of chromosome configurations of PMC's of ?. nolystachvus at diakinesis and Mj 42 5. Pollen grain diameters of P. nolystachyus in microns 45 6. Analysis of variance of pollen grain diameters of P. lunatus var. Fordhook, P. ooly^tachyus and the amphidiploids (categories 1, 2, 4, 10 and 11) . . . 4 6 7. "t" test for comparison of mean diameters of pollen grains of P. lunatus and P. nolystachvus 46 8. Frequencies of chromosome configurations in PIC's of the F]_ (category 3) at diakinesis and Mj . . . . 48 9. Frequencies of chromosome configurations in the PIC's of ?2 (category 4) at diakinesis and metaphase I 60 10. Pollen grain diameter of F2 (category 4) 6l 11. "t" test for comparison of mean diameter of pollen grains of F^ (category 4) and P, lunatus (category 4) 63 12. Frequencies of chromosome configurations in PMC's of Fj (categorjr 10) at diakinesis and Mj 67 -iv- LIST OF TABLES — Continued Table Page 13. Pollen grain diameters of F-, (category 10) plant 1 in microns ^ 70 14. "t" test for means of diameter of F~ (category 10) plant 1 and F2 (category k) ... . 72 15. Frequencies of chromosome configurations in PMC's of F^ (category 11) at diakinesis and Mj. 74 16. Frequencies of chromosome configurations of "sterile," diploid F^ (category 5) 77 17. Frequencies of chromosome configurations in PMC's of first backcross and progeny at diakinesis and metaphase I 79 18. Frequencies of different chromosome configuration types at diakinesis and of PMC's of 20 plants from first generation of Backcross 2 (lunatus recurrent) (category 8) 8k 19. Frequencies of different chromosome configuration types at diakinesis and Kj of PMC's of progeny of 2 first generation plants of backcross 2 (lunatus recurrent) (category 9) 85 20. Diameter of pollen grains of F~ (category 10) plant 3 in microns 104 21. Diameter of pollen grains of F- (category 10) plant 4 in microns 105 22. Pollen grain diameters of F^ (category 11) plant 1 in microns 1063 23. Pollen grain diameter of F^ (category 11) in plant 2 in microns 107 -v- LIST OF ILLUSTRATIONS Figure Page 1. Pedigree of material investigated. Numbers in parenthesis refer to text description (page 17 and 18) and will be used henceforth for the purpose of identification 19 2. Method for excluding anther debris 24 3. Development of roots grown by wick device .... 25- 4. Germinating seedlings of P, lunatus var. Fordhook, the Fj and P. polys tachvus 31 5. Leaves, flowers and pods of _P. lunatus var. Fordhook, F^ and P. nolystachyus 32 6. Seeds and flowers of P. lunatus . F.. and P. polystachyus 1. . 32 7. Metaphase plate from a root tip cell (above) and an idiogram (below) of P, lunatus var. Fordhook (Fuelgen; 3.500X) . . 7 33 8. Photomicrograph of a PMC of P. lunatus var. Fordhook showing pachytene chromosomes (1,875X). . 36 9. A PMC of P. lunatus. var. Fordhook showing 10IT and 2X atMj(l,700X) 36 10. A PMC showing diakinesis with 11TT in P. lunatus. var. Fordhook (1,875X) . .” 37 11. Frequency distribution of pollen grain diameters of P. lunatus var. Fordhook 39 12. Metaphase plate from a root- tip cell (above) and an idiogram (below) of P. polys tachvus 41 13. A PI SC showing M- in P. polystachvus with 10TT and 2X (1.700X) ^4 -vi- LIST OF ILLUSTRATIONS— Continued Figure Page 14. Frequency distribution of pollen grain diameters of P. nolyrtachvus 47 15. PIC of F1 showing 6 and 10]. at Mj (1.700X) . . 50 16. PMC of F, showing 2 lagging chromosomes at AT (i.toox)1 ; . . 50 17. PMC of F^ showing a short broken bridge and fragments (1,700X) • 51 18. PIC of F, showing a non-movement of chromosomes to the poles at Aj (1,700X) 51 19. PMC showing an unreduced nucleus following M- in F1 (1,700X) T . . 53 20. Pedigree of araphidiploids (F2, F^ and F^) (categories 4, 10 and 11) . 55 21. The amphidiploid F2 plant (category 4) showing flower color, leaf and pod shape 56 22. PMC of F? (category 4) showing a chain quadrivalent and 8J (2,675X) 58 23. PMC of Fp (category 4) showing 19 jj and 6j (2,675X) 59 24. Frequency distribution of pollen grain diameters of Fp (category 4) 62 25. Germinating seedlings of F- (category 10) showing near hypogeal germination 64 26. Plant 4 of F~ (category 10) showing leaf, pod and flower characters and general habit of growth . . 64 27. Photomicrograph of PMC of F~ (category 10) plant 4 showing 18 jj and 2jy at diakinesis (1,000X) ... 66 28. PMC of Fo (category 10) plant 4 showing ]9jj and several secondary associations between the1 bivalents (2,675X) 66 29. Separating chromosomes in 22/22 at AJ in PIC of F^ (category 10) plant 4 (1,200X) . 69 -vii- LIST OF ILLUSTRATIONS— Continued Figure Page 30. Frequency distribution of diameter of pollen grains of F^ (category 10) plant 1 71 31. Flowers of a diploid, Fordhook like derivative and an amphidiploid , F^ (category 11) plant 2 . . 73 32. PNC of Fr_ (category 11) plant 2 showing 20 and at metaphase I 75 33. Pollen grains of F^ (category 11) plant 2 (168X) 76 34. A root tip cell from first backcross (category 7) showing 22 chromosomes at metaphase (3»500X). . . 80 35. A PIC from first backcross (category 7) showing 7n, 8X at Mj. (3,500X) 80 36. Vine and bush segregates from backcross 2 (lunatus recurrent) second generation (category 9) . . . . 82 37. A PMC from backcross 2 (lunatus recurrent) first generation plant (category 8) showing 10TT and 2~ (3,500X) .1 83 38. A photomicrograph showing pollen grains of backcross 2 (lunatus recurrent) second generation plant (category 9) (168X) 83 39. Frequency distribution of pollen grain diameters of Tj (category 10) plant 3 108 40. Frequency distribution of pollen grain diameters of F^ (category 10) plant 4 109 41. Frequency distribution of pollen grain diameters of F^ (category 11) plant 1 110 42. Frequency distribution of pollen grain diameters of (category 11) plant 2 Ill -viii- INTRODUCTION EhaseoIUfi lmiatus L.» the lima bean, is an important vegetable legune of the United States. Florida once used to be one of the leading states in the production of the lima bean but over the past several years there has been a considerable decline in its acreage in this state. Still nearly two thousand acres are planted to this crop annually (12), There are the large- seeded (macrospermus ) and the small- seeded (microspermus) varieties of lima beans. The large, thick- seeded type represented by the commercial variety Fordhook is regarded as having the best table quality (51). One of the problems encountered in the production of Fordhook type lima beans in encrusted or impacted soil is the mechanical injury to the seedling at the time of emergence. Due to epigeal germination of the lima bean, the cotyledons have to be pushed out of the soil as the hypocotyl elongates (Figure 4). During this germination process the massive cotyledons are often broken away from the axis, resulting in "bald heads" breakage of hypocotyl, and reduced stand and vigor of seedlings. The damage at germination might be avoided if it were possible to breed a Fordhook-type lima bean with a hypogeal character of germination. With this as the primary object, -1- attempts have been made to produce interspecific crosses between the lima beans and other closely related species having the hypogeal character of germination (16, 26). It is conceivable that the lima bean has affinities with the other New World species of Phaseolus. This view is supported by the recent evidence presented by Mackie (31) to the effect that lima bean is a native of Guatemala rather than of Africa as was believed by Linneaus. Most of the crosses so far attempted between the lima bean and the other American species of Fhaseolus have not been successful. There was one, however, the primary success of which was reported by Lorz (26). It involved Phaseolus lunj^us L. var. Fordhook x Phaseolus polystachvus (L. ) B. S. P. The latter species is the wild thicket bean which is a native of the eastern United States. According to Small (44) its geographical distribution extends from Florida northwards to Maine and westward from the Atlantic seaboard to Texas and Nebraska. The cotyledons of this species remain entirely under- ground during germination and all the growth of the shoot takes place by the elongation of epicotyl and the plumule (Figure 4). Lorz (26) also considered the possibility that in addition to the valuable germination character other desirable characters might be transferred in such a cross: e.g., possible unrecognized disease and pest resistance, vegetative vigor, perennial cold- hardy rootstock and other physiological attributes of survival value. The hybrid between these two species expressed nearly completely hypogeal germination (Figure 4). It also showed a high degree of self- and cross— sterility. Following the primary success with the cross reported by Lorz (26), a cholchicine- induced chimera was obtained by Lorz (29) in F1 vegetative material in which the amount of apparently viable pollen was increased from approximately 5 per cent in the untreated to approximately 90 per cent in the chimera and the average volume of the individual, apparently viable pollen grains was doubled. Lorz (27) also described fifteen individuals resulting from Fordhook pollinated by the F^ and observed what he regarded as an evidence of linkage involving an association of the character of hypogeal germination and that of vine habit, both characters derived from P. polvstachvus. All this material was subsequently lost without any further work (29). Following a repeat of the original cross two ?2 plants were, however, derived as a result of open pollination. One seed was also produced in pollinating the F^ with Fordhook type pollen, thus establishing a backcross with Fordhook lima ancestry. From one of the two Fg plants and from the single backcross individual further progenies and backcrosses were derived. The need for cytological study of this material was indicated by the lack of any marked homology between the chromosomes of the two species, as exhibited by cytological examination of Fl (10), also by the high fertility of F^ and later generations. The determination of the chromosomal mechanisms involved in this material should be of considerable value in designing a breeding program for the utilization of this material. It is now unquestionably recognized by biologists that cytology is as great an ally of taxonomy as it is of genetics. At present there is little information on the chromosome morphology of the species involved. Therefore, besides the above mentioned practical considerations, cytological analyses of these derivatives and karyological studies, per se. would be highly valuable from the point of view of taxonomy. This study is thus primarily concerned with the cytological investigation of two Phaseolus species, namely, P. lunatus and — * the F^ to F^ generations from this cross, and the backcross generations from the hybrid with P. lunatus (recurrent). An attempt has been made to establish the karyotype of the two parents, the chromosome numbers of the derived material and the homologies and association affinities as indicated by raeiotic behavior. These studies were also supplemented by pollen counts and size measurements. REVIEW OF LITERATURE Interspecific hybridization is not a new phenomenon in plants. Cyto taxonomical studies have revealed that hybridization is of widespread occurrence in nature. Several instances have also been found where a cross between two species has been followed by a spontaneous reduplication of the chromosome complement of the diploid hybrid giving rise to an allopolyploid or an amphidiploid. According to some authors this phenomenon has played a significant role in the process of evolution (11, 47). There is now ample evidence to show that some of our important crop plants, wheat, cotton and tobacco have originated in this way (11). Interspecific hybrids have also been produced art ific ally. Plant breeders often resort to interspecific hybridization in order to incorporate "new" genes in the cultivated plants. Such hybrids are also produced experimentally to augment cyto taxonomical studies for tracing relationships between species, genera or other taxonomic groups. In this section the cyto logical aspects of hybridization and amphidiploidy will be surveyed briefly. A brief review will also be made of the available literature on interspecific hybridization in the genus Phaseolus . -5- Allopolyploids Though there is a series of intermediates, allopolyploids at the tetraploid level have been classified into two general types by Stebbins (k6). The first of these types is called the typical allopolyploid or amp hi diploid and the second as segmental allopolyploid or segmental amphidiploid. A typical allopolyploid is one in which there is no intergenomic pairing of the chromosomes and the identity of parental genomes is preserved. Most representatives of this type are highly constant because of alio syndesis or pairing between homologous chromosomes. The second type of anphidiploid or allopolyploid classified by Stebbins involves a partial or console te in te rgenomal pairing of chromosomes. lanital allopolyploids Karpechenko ' s (22) Raphanobrassica derived from the intergeneric hybrid Raphanus sativus L. (fl = 9) x Brassica oleracea L. (n = 9) is a classical and perhaps the earliest recognized example of an experimental arphi diploid. The ^ from this cross had 18 univalents, 9 from one parent and 9 from the other. During metaphase of the first division the chromosomes were distributed at random to the two poles. The plants were highly sterile and only a few seeds were produced. Cytological examination showed that there were 36 chromosomes in F2 somatic cells. The F^ was assumed to have arisen from the union of unreduced gametes. Meiosis in the tetraploid plant was found to be regular. Eighteen bivalents appeared at me tap ha se I, 9 originating from the pairing of Rap harms homologs and the other 9 from Brassica homologs, and disjunction was normal. The gametes produced by the tetraploid carried 9 Rapt&nuq and 9 Brassica chromosomes. The tetraploid plants were fully fertile. All the progeny from this cross were uniform in appearance and no segregation was observed. A similar case of spontaneous amphidiploidy was reported by Buxton and Newton (4) while investigating the cytology of F2 from the species cross Digitalis ambjqua x Digitalis purpurea. The n and 2n number of both the species was 28 and 56* respectively. It was discovered that while the F1 was a diploid the had 56 and 112 chromosomes in gametic and somatic cells, respectively. It was concluded that the doubling of the chromosomes had occurred (hoe to the union of unreduced F^ gametes. The unreduced gametes were assumed to result from failure of reduction division and the formation of restitution nuclei. The ?2 generation differed from Fx principally in size and fertility. There was no segregation of parental characters. Segmental allopolyploids A great deal of work has been done on Primul^ kewensis . a diploid hybrid of P. flprlbunda x P. verticillata (aj = 18) and its derivatives. Newton and Pel lew (34) have studied the cytology and genetics of the fertile tetraploid progeny of tewensis with 2n = 36 chromosomes. Their investigation shows that doubling took place in the somatic tissues of the hybrid. -8- Meiosis, both in the diploid P. kewensis and its tetraploid derivative was found to be regular but only the tetraploid was fertile. The high degree of fertility and constancy of the tetraploid has been explained by Winge’s hypothesis of inter- specific pairing of the chromosomes. A sporadic variation of the tetraploid was observed and has been explained as the possibility of some intraspecific pairing of chromosomes. It has been found to be associated with the loss or gain of a chromosome. There was also some variability in the tetraploid which was not so associated and was probably allelic. They also observed a plant with 26 somatic chromosomes. It was derived from a presumed triploid by self-fertilization. In meiosis it formed 10 bivalents and 6 univalents. It was used as a male parent for a backcross with floribunda . and produced descendants with combined characters of vertlci j lata and floribunda . Artificial triploids were also obtained by crossing diploid varieties of floribunda with the tetraploid. Ichijima (19) investigated the cytology of tetraploids originating from selfed progeny of an F^ plant of the cross Fra^rl^ bracteata Heller (n = 7) x Fragaria vesca rosea Rost, (n = 7) . In the tetraploids 14 bivalent chromosomes were found at metaphase I. Root tip counts showed 28 somatic chromosomes. It was believed by Yamell (53) that this F1 plant originated as a chromosomal chimera. Meiosis of F^ tetraploid was found to be irregular. Four chromosomes were occasionally found to be associated at -9- metaphase I. It was deduced from genetical evidence that a chromosome of F. ves^a rosea seems to pair as readily with the corresponding one from F. bracteata as with the £. vesca rosea homolog. Anaphase I showed occasional lagging. The second anaphase sometimes showed an early disjunction of a chromosome pair. Meiosis in some of the F^ plants showed an increased amount of irregularity, especially at anaphase I. Jones and Clarke (20) reported the discovery of a natural, fertile amphidiploid from a cross between Allium cepa L. var. Australian Brown x jllium fjstulosa L. type Nebuka. It shows greater vegetative vigor than either parent and is perennial like type Nebuka. Meiotic behavior is fairly regular. Sixteen bivalents (2n = 16) were generally formed at metaphase I. Fragments, micronuclei and chromatin bridges were observed but were not as common as in the diploid hybrid. Another example of natural segmental amphidiploid is that derived from the hybrid Nicotiana glutinosa (n = 12) x (a = 24) reported by Clausen and Goodspeed (6). In the diploid hybrid there were bivalents as well as univalents at metaphase I. During anaphase I the bivalent partners approach the poles while the univalents remain in the equatorial zone either dividing or preparing to divide. In the tetraploid, on the other hand, there appeared to be no univalent chromosomes and the bivalent partners (36 IT ) moved in regular fashion to $ the poles. The amphidiploid was uniform and constant. The authors believed that the F^ must have arisen from a doubling of -10- the chromosome number immediately or soon after fertilization by which a tetraploid hybrid with pairs of chromosomes was produced. This plant might be represented by the chromosomal formula 12 GG + 24 TT. Skirm (43) experimentally produced a polyploid by thermal treatment of the natural hybrid Trade scan tia canal iculata x T. iijunil^s. One of the 4n derivatives from this treatment showed primarily bivalent chromosomes pairing during meiosis. The origin of this plant was attributed to chromosome doubling following fertilization. The argument presented in favor of this view is that most Tradescantias are heterozygous in respect to chromosome structure. The production of a tetraploid from such a stock by meiotic irregularities resulting in diploid gametes would produce a plant in which no 2 of the 4 genomes would be identical. Thus a lack of preferential pairing would favor quadrivalent formation. In this particular plant, however, the meiotic behavior suggested that chromosome doubling occurred at the time of fertilization or during embryonic development. In other words it had been a case of somatic doubling. Kostoff (23) made extensive studies on the cytological behavior of (jj^q^ana glauca (_n = 12) x Ntcotiana 1 angsdorffii (n = 9 ) amphidiploid. He reported differences in the chromosome morphology of the two parental species. The F hybrids usually had 21 somatic chromosomes. It was reported that N. glauca chromosomes had small segments homologous with portions of N* langsdorffii chromosomes. -11- Another important observation made by the author was the formation of chromosomes with new genetic contents. This was supposed to be due to exchange of parts in bivalent and tri valent groups during the mieosis of F hybrids following allosyndesis and, in exceptional cases, following autosyndesis between homologous segments of the partially homologous chromosomes and probably between heterochromatic regions of nonhomologous or partially homologous chromosomes. The Fj hybrids were self-sterile. Most of the backcrosses ii* langsdorffli had 30 somatic chromosomes. The meiosis of these backcrosses varied from plant to plant in irregularity. These plants also showed variation in morphology and fertility. One amphidiploid was also obtained in the backcross. The original amphidiploid type was studied through six generations and it was discovered that the viability of the pollen increased with each successive generation till it was 99.5 per cent in Fg. This was correlated with a corresponding decrease in the number of multivalent and univalent associations. Euploid chromosome alterations led ultimately to changes in the size of nuclei and cells. The addition of each genora led to a significant increase in size. It was also observed that aneuploidy conditioned changes in the size of nuclei and cells but such changes were not always significant. Also, the amphidiploids originating from the asyndetic F1 had normal syndesis and were highly constant, while those originating from hybrid with complete or partial allosyndesis were not -12- constant. It was postulated that changed mutation rate and chromosome rearrangements in the amphidiploid were responsible for the gradual increase in its fertility. The amphidiploids and their derivatives were physiologically isolated. They were reported to cross either with difficulty or not at all with other species and species hybrids or with the parental species. Interspecific Hybrids in the Genus Phaseolus A number of interspecific crosses have been attempted in the genus Phaseolus. The earliest record of such a cross is that made by Mendel in 1866 (15), between Phaseolus vulgaris L. and Phaseolus multiflorus. Lem., a runner bean. This cross was later repeated by other workers. Lamprecht (24) made extensive genetical and cytological study of the cross P. vulgaris x P. pnxltiflorus. He discovered that though meiosis of the F^ ran a normal course, there was a high degree of pollen sterility. The degeneration of the pollen grains occurred following telophase. As a result of this, the showed a loss of multiflorus characters and was consequently matroclinous to a great extent. This was explained by the fact that the new gene combinations were not viable in the existing plasma. In the later generations it was found that a combination of vulgaris and multiflorus properties met with greater resistance as attempts were made to introduce more genes into the combination. Eventually, from some of the later generations could be selected -13- entirely constant and fertile lines containing combinations of taxonomic characters of both species. Thus a bridge was thrown across the species boundary between P. vulgaris and P. multiflorua . A cross between P. vulgaris L. and P. mungo L. , the Urd bean, an oriental species, has been reported by Strand (48). The main objective in this breeding program was to secure an offspring that was resistant to the Mexican bean beetle. The cross was successful and an generation was raised. Three different interspecific crosses in Phaseolus have been made by Honma and Honma and Heeckt. The first of these crosses made in 1956 (15) involved P. vulgaris x P. acutifolius. Chromosome smears of the root tips showed no morphological differences between the complements of the hybrids and the parents. The second cross reported in 1958 involved P. igOpgi-peus x P. lunatus (16). This was an effort to transmit the hypogeal germination habit into _P, lunatus . the lima bean. The seedlings exhibited the hypogeal germination habit and, in general, the vegetative characters appeared more like — • coccineus parent. The segregation for germination habit ranged from epigeal to hypogeal suggesting a polygenic inheritance. The third cross attested in 1959 (17) was between P. vulgaris and P. l.una^u? . The object was to combine a darker seed coat color with green cotyledons in snap beans. Several interspecific crosses were attempted by Lorz (28) in the genus Pjaagoluq from time to time. They are enumerated below P. vulgaris x P. coccineus x P. polyanthus x P. xanthotrichus x P. glabellus x P. acutifolius x P. lathyroides x P. atropuroureus x P. aureus x P. mungo * P- angularis — . x P. calcaratus x P. helvolus _ x P. speciosus x P. lunatus (and reciprocal) P. lunatus x P. polys tachvus — x Phaseolus sp. (PI No. 201202) Many of the above noted crosses were not successful. In several cases sterility was encountered in the hybrid. A preliminary report on the success of P. lunatus x P. £0lystaiw. f,. stenie. This F individual, was diploid afld was a highly sterile plant. P 6. 7. 8. 9. 10. 11. i- (J°* 3 above it £, lunatus var. Fordhook, the first backcross was made in 1958 using Fordhook parent* was a single individual representing this backcross. It was a perennial which produced seeds sparingly. The progeny from presumed selfing of the first ^Ih^SS-i N°* 6 abov®)* There were six individuals in this progeny. 6 x IfflHtus var. Fordhook is the first fTT *5® 8econd ^ckcross (Fordhook * Twenty- seven individuals from this generation were available for study. Progeny of No. 8 is the second generation from the second backcross (Fordhook recurrent). This ™K7^73riSed °f 14 plants d®**ived from several different BC2 individuals. |* var. Fordhook x P. polvstachvue . the FJ P^geny consisting of 7 plants derivS from the amphidiploid F2 plant. lana^as var. Fordhook x P. Dolysiachyus . the F4 progeny derived from the F_ Sjhid^Sid. This generation is contained Vy 10 plants. ■19- o Q Si. U $ o O CM ■P CD • A! O O 55 o o fS r as° o S§ x - CO w ■P 0) X w 5 g. U 0> E p o CO

T3 5 S •gs P a) q a c P L *H Q) P O CO p TJ I I O P bo •3 <£ T) «H O rN -20- Technique The material is rather difficult to handle cytologically. No standard technique has been evolved so far. Several worker's have employed and proposed different schedules for fixing, staining and mordanting or for otherwise treating difficult plant materials. Since different materials do not respond alike to different treatments and reagents, a preliminary trial was conducted to determine the most satisfactory combination of treatments both for the pollen-mother-cells and the root tips. The procedures used for these 2 kinds of materials are discussed separately as follows. Pollen-mother-cells Preliminary examination showed that the anthers of Phased us are quite small, and that each anther contains a relatively small number of pollen-mother-cells. Micro sporo genesis takes place when the buds are quite small (< 2 mm long). The buds were fixed at 2-hour intervals during the day from 6:00 A.M. to 4:00 P.M. in order to determine whether a particular time was most favorable for optimum production of raeiotic division. It was found that the divisions were not restricted to any particular time and no single period was conspicuously more favorable than any other. However, there seemed to be a relatively larger number of dividing cells at about 6:00 A.M. during the months of June and July. There was an apparent shift in this time towards 8:00 A.M. with the approach of -21- colder weather. Consequently an attempt was made to fix as much of the bud material as possible close to the 8:00 A.M. hour. Ukrattyeg The most common fixative recommended for pollen-mother-cells intended for ace to carmine squash preparations consists of a mixture of three parts of absolute ethyl alcohol to one part of glacial acetic acid (Belling 2). The original fonnula of Belling has since been modified and adapted for more intense staining of small chromosomes. It was thought desirable to make a preliminary trial with 3 different fixatives, namely. Belling and the modifications proposed by Swaminathan si &1. (49) and by Hyde and Gardella (18). Hyde and Gardella's modification was found to give the most intense staining as well as increased differentiation of details. It was, therefore, used with slight changes as a basis for the following preparation of the fixative most widely used in this investigation. The procedure used was as follows: 1. An excess of 5 gms of ferric chloride was dissolved in 300 ml of distilled water until a thick precipitate of Fe(0H)3 resulted. 2. The precipitate was thoroughly washed in a Buchner funnel, dried in a desiccator and stored in a refrigerator in a sealed container. 3. Two grams of the Fe(OH)- thus prepared was dissolved in 50 ml of propionic acid \ y gently heating. 4. This was filtered and 50 ml absolute ethyl alcohol was then added to the cooled filtrate. The addition of a few drops of acetocarmine to the above mixture, as recommended by Hyde and Gardella (18), was omitted as it invariably precipitated iron. -22- The whole inflorescence containing buds in a graded series of stages of maturity was placed in the fixing fluid, so that it was easily possible to determine which size was near optimum for meiosis. Consequently, it was possible to select for squashing only those buds which were near the optimum. The fixation was quite satisfactory with such small buds and it was not necessary to remove the floral coverings. The optimum period of fixation was found to be from 12 to 24 hours. If the buds were not inane diately used they were placed in 70 per cent ethyl alcohol and stored in the refrigerator. Two different stains were tried; a 1.5 per cent orcein solution in 45 per cent glacial acetic acid prepared according to Vosa (52) and ace to carmine in which iron was introduced as a mordant, Hie latter was found to give decidedly superior preparations with respect to intensity as well as differentiation. Acetocannine was prepared by dissolving 1 gm of carmine in 100 ml of boiling 45 per cent glacial acetic acid and refluxing for one hour. To half of this was added a few drops of Fe(OH)^ solution in 45 per cent acetic acid until the liquid became bluish red without any visible precipitate. This was then combined with the rest of the untreated ace to carmine. Acetocarmine staining was adopted for all subsequent work with the pollen-mo then- cells . Except for a few camera lucida sketches, all other drawings and photomicrographs were prepared using this procedure. SfflaaMflg -23- All the buds were detached from the axis and placed on a flat bottomed watch glass containing 70 per cent ethyl alcohol. Those of near optimum size were sorted out for squashing. In order to obtain as many as possible of the pollen- mother-cells , all the anthers of a bud were squashed under a single coverslip. The bud was dissected under a binocular microscope in a drop of the stain fixative on a slide and all the anthers taken out with the help of a needle. The anthers were then mashed thoroughly with the blunt end of a scalpel in as small a quantity of the stain fixative as practicable because an excess of the stain would not only allow the anthers to float and escape squashing but would dilute the suspension of pollen-mo the r- cells and permit many to escape from under the coverslip. All anther debris was excluded from the film beneath the coverslip by touching one edge of the coverslip to one edge of the drop containing the squashed anthers (Figure 2). Then, as the coverslip was allowed to rest flat against the slide, all the supernatant liquid along with the suspended pollen- mother- cells was drawn underneath by capillarity while the coarser debris remained behind. The slide was then heated gently over the flame of a spirit lamp, pressed lightly between bibulous papers and sealed with a lil mixture of paraffin and Canada balsam. These preparations were made permanent by the quick freezing method of Conger and Fairchild (7). For studying the prophase chromosomes the postmordanting method suggested by MacDonald (30) was employed. - 24- Cove ralip Slide Anther debris Figure 2.— Method for excluding anther debris. Preparation of root. ^r. Bie root tips were obtained from tin sources, germinated seedlings and rooted cuttings. B» eeeds were germinated between -oist paper towels and the actively growing root tips were taken as they appeared. Later the seedlings were transferred to Plastic pots about 4 inches square, 4 inches deep with a centre! hole in the bottom with a nylon wick inserted through the hole. The pots were filled with a well-prepared sterilised soil and the seedling planted in the center of the surface of the pot. Each pot was placed over a jar partially full of water with the nylon wick dipping in it. Ihe bottom of the pot acted as a lid and a waterevapor saturated atmosphere was built up in the lower Jar. This device caused the water to move up to the root zone ty capillarity. Ihe growing roots soon paralleled the course of the wick and grew into the water inside the jar (Figure 3). Thus more material for root tips was available and these were obtained before the seedlings were finally transplanted in the ground. -25- Figure 3.— Development of roots grown by wick device. -26- Cuttlngs Cuttings were employed in order to obtain root tip material from adult plants. An effort was made to grow cuttings in the ordinary way and some success was achieved if the cuttings were grown in a well aerated medium. However, this was a slow and an unsatisfactory method. Some investigators (42) have used growth regulators to induce rooting of cuttings for cytological use. It was, therefore, decided to try a similar procedure. Satisfactory results were obtained by treating the cuttings with a 50 ppm aqueous solution of indolebutyric acid for 24 hours prior to planting. The cuttings were also planted in pots with the wick device. Treatment with indolebutyric acid resulted in an early and a more profuse rooting as compared to the untreated cuttings. Pa-.tratqent. fixation and staining The somatic chromosomes are quite small and it was difficult to get a well stained preparation. Several schedules of treatment and stains were tried as recommended (3, 32, 33, 35, 39). The most satisfactory results were obtained with the schedule used by Sagawa (38) with some modifications. This was as follows: Fresh and actively growing root tips were pre- treated in a .003 M aqueous solution of 8- hydroxyquinoline for 3 1/2 hours at 6°C. Fixed in a 3*2:1 mixture of absolute ethyl alcohol, glacial acetic acid and chloroform for 30 minutes at 60°C. 3. fjydrolized in 1 N HC1 for 10 minutes at 60°C. 4. Washed with distilled water. 5. 6. 27- Stained with leuco-basic-fuchsin for 1 hour. Bristly stained tips were cut and treated with 5 per cent aqueous solution of pectinase prepared according to Setterfield et al. (41) for 4 hours. 7. Six-seven root tips were squashed on a slide in a drop of 45 per cent acetic acid the coverslip was applied and the preparation was pressed hard between bibulous papers. 8. Sealed with lsl mixture of paraffin and Canada balsam. 9. Made permanent by the quick freezing method of Conger and Fairchild (7). Pollen studies Study of pollen grains was made with two main objectives in view. First, to learn whether their chromosome content is reflected in their volume, enabling the researcher to devise a rapid method of predicting changes in chromosome number by pollen grain measurement. Second, to use it as an indication of polyploidy when more suitable material was not available. A heirarchial sailing design was used. Observations were recorded on the diameter of 20 randomly selected pollen grains from 5 randomly selected, fully mature buds on each plant. Variance was analyzed and comparisons between means were made using student* s "t" test. Equipment Observations were made with a Zeis L, research type microscope equipped with acromatic objectives 8 N.A. 20, 20 N.A. 40, apochromatic oil immersion 90 N. A. 1.4 condenser. Drawings were made with an Abbe camera lucida. Some of the -28- photomicrographs were taken with a box micro- camera on Kodak panchro-press film sheets. Later it was possible to use Contaflex 35 inn single lens reflex camera in conjunction with a Lafayette "micro-dap tor. " The other photographs were taken with either a Contaflex or other 35 nn camera using a variety of film types. RESULTS AND OBSERVATIONS Phaseolus lunatus var. Fordhook progeny (Category l) Morphological characters Since Fordhook is a well-known commercial variety of lima bean, its detailed morphological description is omitted here. Only those characters are mentioned which mainly contrast with those of P. polystachvus . These are: white flowers, forward projection of the wings, absence of pigmentation in hypocotyl and internodes, large lens— shaped seeds with white seed coat, large flat pods, epigeal germination, bush habit and an absence of any marked response to photoperiod (Table l). Some of these characters are represented in Figures 4, 5, and 6. Cytology Cytological examination of the root tips showed that the somatic chromosome number was 22. This confirms the number reported by Ka rpe chenko in 1925 (21) and later corroborated by Dhaliwal e£ (10). A critical analysis of the karyotype revealed that there are the following types of chromosomes in this species (Figure 27). Type A - A pair of comparatively long chromosomes. No primary or secondary constructions were visible in this pair. - - -29- Comparison of unit characters of J?. polys tachvusf amphidiploid derivatives (categories 4, 10 and 11) and P. lunatus var. Fordhook -30- 3 8 u ni 6 a i i co > r—i O O G,*| £ V co c 0 & R tC h 1 rt £ & 73 x: a) o to a a> a> P •H p 73 £ 5 to §e V rH & 3 a & S E T5 rt 8 & tt) to •P i-i O O rH c rH O I rH VO ON 5 TS S t 3 CO o S3 p 8 rH a o & *? 8) 5 Eh * * -31- Figure 4._Geminating seedlings of P. lunatus van. Fordhook, the and P. polys tachvas . -32- Figure 5.— leaves, flowers and pods of P, lunatus var. Fordhook, F^ and L- P_olystachyus . Figure 6.— Seeds and flowers of P. lunatus. Fx and P. polystachvus. -33- Figure 7 ft * . ji * A ♦ U as ( <« i 0 « n i c Types A B C D E F G —Metaphase plate from a root tip cell (above) and an idiogran (below) of P. Ivina tus var. Fordhook (Fuel gen; 3,500X) -3*- Type B - A pair of comparatively long chromosomes with a submedian primary constriction and a submedian secondary constriction in the long arm. type C - Two pairs of comparatively medium sized “ chromosomes with a median or a near median centro- mere constriction. ■type D - A pair of comparatively medium sized chromosomes with a submedian primary constriction. ^ j,^° Pairs °f comparatively small chromosomes with median primary constriction. pP®. f • Two comparatively small chromosomes in which constrictions were not apparent. Type G - A pair of hookshaped chromosomes, the smallest in the complement. An attempt was made to study the pachytene behavior of the chromosomes but it was found to be difficult to trace the chromosomes throughout their length as they were much tangled (Figure 8). Most of the studies of the pollen-mother-cells were, therefore, confined to diakinesis, metaphase I and anaphase I to determine the pairing relationship and disjunction of chromosomes. Table 2 shows the pairing relationship as observed at diakinesis and metaphase I (Mj). Table 2 frequencies of chromosome configurations in PMC’s of P. lunatus var* Fordhook at diakinesis and Mj — Configurations “n °i 10n 2I 9n *1 Total Frequencies 12 5 1 18 11 can be seen that 2/3 of the observed PICs showed regular pairing of chromosomes to form 11 bivalents. A few cells with -35- l°n and 2l (Figure 9) and a rare cell with 9jj and ^ were also observed. A critical comparison of all the observed figures revealed that most of the univalents appeared due to an early disjunction of 1 or 2 bivalents. None of the cells observed in diakinesis showed any univalents. Figure 10 shows such a diakinesis stage. It can also be seen here that there is a single, spherical nucleolus attached to one pair of chromosomes. Though most of the anaphase I figures showed a regular separation of 11 chromosomes to each pole, a rare instance of irregularity with lO/ll separation and loss of 1 chromosome was also observed. Over 90 per cent of the pollen grains were apparently normal. The mean diameter of the pollen grains was 38.87 microns (Table 2). Figure 11 shows an almost normal distribution of the diameters of the pollen grains. Phaseolus polystachvus (category 2) Morphological characters Morphologically P. polys tachvus differs from P. lunatus var. Fordhook in some essential respects. These characters are: purple flowers, somewhat lateral extension of wings, purple pigmentation of hypocotyl and intemodes, small seeds with mottled fawn seed coat, slender glabrous pods, hypogeal germination, vine habit and a marked flowering response to short days. It has a cold hardy rootstock by means of which -36- Figure 8. — Photomicrograph of a PMC of P. lunatus var. Fordhook showing pachytene chronosomes (1,875X). Figure 9. — A PIC of P. lunatus . var. Fordhook showing 10XI and 2j at (1,70QX). -37- Figure 10. —A PMC showing dia kinesis with 11 -p- in P, lunatus. var. Fordhook (l,8?5X) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 -38- Table 3 Pollen grain diameters of P. lunatus var. Fordhook in microns Buds 1 2 3 4 5 40.80 35.38 40.60 44.08 33.06 41.76 35.96 38.86 45.24 3*+.80 40.02 37.70 41.18 46.40 43.50 40.02 38.28 39.44 40.60 35.96 38.86 35.96 38.28 45.82 34.22 39.44 37.12 41.76 44.08 34.22 42.34 41.18 33.64 45.82 34.80 42.34 37.12 41.18 *+3.50 35.96 38.86 38.28 41.18 *+5.98 45.80 38.86 38.28 38.86 47.56 36.5*+ 41.76 39.44 37.70 42.34 32.48 40.02 40.02 44.66 49.88 34.22 41.18 35.38 36.54 43.50 34.22 38.86 37.12 40.02 41.18 31.90 39.44 37.12 41.18 44.08 33.64 42.92 38.28 40.02 46.40 34.80 44.66 36.54 40.02 44.66 35.38 49.30 40.02 *0.50 44.08 33.06 41.76 40.60 39.44 *+3.50 36.54 JtZilZ -JZ>70 40.60 .a.y -.33.06 826.12 757.48 798.66 807.36 697.16 = x = 38.87 -39- suifBaS uaxiod jo aaqutnu Figure 11. — Frequency distribution of pollen grain diameters of P. lunatus var. Fordhooi. it overwinters (Table 1). Some of these characters can be seen in Figures 4, 5 and 6. Cytology Cytological examination of the root tips showed that the somatic chromosome number is 22. This confirms the number previously reported by Allard and Allard in 1940 (1). A critical analysis of the karyotype showed the following types of chromosomes in this species (Figure 12). Type A - A pair of comparatively long chromosomes which did not reveal any centromere or sedondary constrictions. l^Pe B - Four pairs of comparatively long or median chromosomes with a submedian primary constriction and a submedian secondary constriction in the long arm. T^rpe D - A pair of comparatively medium sized chromosomes with a submedian primary constriction. Type F — Two pairs of comparatively small chromo- somes in which constrictions were not apparent. Type H - A pair of comparatively long chromosomes with a submedian primary constriction and a sub- terminal secondard constriction in the short arm. ?yPe 1 ~ A pair of comparatively long chromosomes with a median primary constriction and a submedian secondary constriction in one arm. Type J - A pair of small chromosomes showing no constrictions of any kind. On conparing the types of chromosomes in the two parent species it is apparent that the type H, I and J were possessed by P. golygtapiiyys and not by P. lunatus. Similarly types C, E and G were found in _P. lunatus and not in _P. polys tachyus . Types A, B, D and F occurred in both species. Thus on a visual -41- l i > * C ) . Types A B D F H I j Figure 12.— Metaphase plate from a root^tip cell (above) and an idiogram (below) of _P, polystachvus. -42- basis it might be possible to conclude that complete homology- may exist between one or more of the chromosome pairs belonging to the types last mentioned. However, this does not preclude the possibility of homology between small segments of the types which are visually unrelated. Another important observation concerns the total length of the chromosomes in each genora. Since the preparations were made with different treatments, it was considered that any absolute size comparison between the two genomes would not be valid. However, it can be seen that P. polys tachyus had at least 3* possibly 4, pairs of comparatively long chromosomes within its complement. On the other hand, P. lunatus had only 2 pairs of comparatively long chromosomes. This might be taken as an indication of a greater total length of the P. polystachvus genom as compared to that of P. lunatus. Table 4 shows the chromosome configurations as seen in diakinesis and metaphase I of the pollen-mother-cells. It can be seen that pairing was usually quite regular. Occasionally, there were 2 or 4 univalents, as were also observed in P. lunatus . Their origin was traced to an early disjunction. Table 4 Frequencies of chromosome configurations of PMS's of P. polys tachyus at diakinesis and M_ Configurations un 10n 9II Total °I 2I “x Frequencies 18 5 2 25 -43- Figure 13 shows Mj of a pollen- mother^ cell . There were 2 univalents and 10 bivalents observable in this cell. Over 95 per cent of the pollen grains appeared viable. The mean diameter of the pollen grains was found to be 41.89 (Table 5). Analysis of variance (Table 6) shows that significant differences exist between the pollen grain diameters of the parents and F^ and F^. A significant difference determined by "t" test was found between the mean diameters of P. lunatus and P. polys tachyus pollen grains, the latter being larger (Table 7). Statistical evidence regarding the greater volume of P. polys tachyus pollen grains supports the cytological observation of a greater total length of chromosomes which could be responsible for the larger volume. The frequency distribution of the diameter shows a comparatively greater uniformity of diameter of the pollen grains of this species (Figure 14). F-]_ Hybrid (Category 3) Morphological characters The hybrid showed a combination of characters of the two parents. In several respects it was intermediate between the two while in others it showed a stronger resemblance to P. polystachvus. These are: vine habit, colored flowers, lateral spread of wings and hypogeal germination. It also showed a tendency to short day flowering response. There was an apparent indication Figure 13. A PMC showing M_ in P IOjj and 2j (1,700X)_ . polys taohvus with 1 2 3 ■ 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 -45- Table 5 Pollen grain diameters of P. polystachyus in microns Buds 1 2 3 4 5 41.18 41.76 40.60 44.08 42.92 40.60 45.29 43.50 46.98 43.50 37.70 39.44 45.82 41.18 42.34 38.28 40.02 37.12 45.82 42.34 42.34 40.02 41.18 44.08 42.34 42.34 40.02 39.44 42.92 42.34 38.28 41.18 38.28 48.72 43.50 41.18 37.70 38.28 48.72 43.50 42.92 42.34 40.60 43.50 42.92 41.76 40.60 38.28 45.24 42.34 40.60 41.18 39.44 44.08 39.44 36.54 42.92 40.02 46.40 35.38 41.18 41.76 36.54- 50.46 42.34 40.60 41.18 38.44 34.22 44.08 46.98 47.56 41.18 47.56 49.50 38.28 40.02 40.02 44.08 40.60 41.76 40.60 38.28 42.34 42.92 41.18 45.24 38.28 43.50 43.50 44.08 46.98 39.44 46.40 38.28 40.02 40.60 40.02 46.40 -J8x86 817.80 836.41 794.76 896.68 842.94 x = 41.89 -46- Table 6 Analysis of variance of pollen grain diameters of P. lunatus var. Fordhook, ?. polystachvus and the amphidiploids (categories 1, 2, 4, 10 and 11) Source d/f S. Square M. Square F F Plants 7 9,161.79 1,308.83 16.90** Buds in plants 32 2,478.26 77.45 2.33** Pollen in buds 760 25,306.49 33,30 Total 799 ^^Significant at .01 level, » Table 7 t test for^conparison of mean diameters of pollen grains of lunatus and _P. nolystachyus "t" (.01) = *2 = ^1.89 - 38.87 = 3.02 = . m ** 2(n-l)^ 2S2 * 0.752 4‘013 - pooled sum of squares. ** Significant at .01 level. Number of Pollen Grains -47- -48- of hybrid vigor as shown by a more vigorous vine development than in the P. polystachyus parent. It has not been possible to determine yet whether it can overwinter by means of a cold hardy rootstock. The seed coat color, unlike either parent, was a uniform purplish black. It was inferred by Lorz (29) that there are interacting factors responsible for seed coat color, because solid or self coloring does not exist in JP. polystachyus and the dark pigmentation is not present in ^P. lunatus var. Fordhook. These characters can be seen in Figures 4, 5 and 6. .OrtqiQfiy The present observations basically substantiated the findings of Dhaliwal al. (10) regarding the meiotic behavior of the F^. A large number of irregularities of various kinds were observed. Table 8 summarizes these observations. Table 8 Frequencies of chromosome configurations in PMC’s of the F, (category 3) at diakinesis and Mj- 1 Configurations 11 10 9876532 Total n n n n n n 5ii jii n 1 ^ °°I 21 4i 6I 10! 12x 16j 1^ Frequencies 1 1 28524-24 29 There were rare cells in which there was a regular n airing of chromosomes at Mj or diakinesis. Hie abnormal pollers mother- cells had 2 or more univalents. The largest number of univalents observed was 18 with only 2jj. The most frequent configuration -49- consisted of 8^. and 6^ which can be seen in Figure 15. Some of the univalents undoubtedly appeared due to an early disjunction as was also the case in both the parents. A critical examination of several diakineses showed that probably the most frequent configuration was 8^ and 6j. The separation of chromosomes at anaphase I was also frequently irregular. Pollen-mother-cells with 10/12 were quite common. Several cells with one or more chromosomes forming micronuclei were also observed. In one pollen-mother- cell the distribution of chromosomes at anaphase I was 8/9/3 forming 2 macro-nuclei and 1 micro-nucleus. Another commonly observed irregularity was the presence of laggards which were excluded from being incorporated in the principal daughter nuclei after anaphase I (Figure 16). Chromosome bridges and fragments could also be observed. Figure 17 shows a cell with 2 fragments and a short broken bridge. In those pollen- mother-cells where the number of univalents is high there was a tendency for all the chromosomes to aggregate in the center of the cell after anaphase I. This condition probably arose by non-attachment of the univalents to the spindle fibers and consequent lack of movement to the poles at anaphase I. Figure 18 shows this condition in a pollen- mother-cell . It was probable that the preceding irregularity gave rise to still another kind of irregularity observed in an exceptional pollen-mother-cell . Normally, there are two macro-nuclei in -50- Figure 16. PIC of F1 showing 2 lagging chromosomes at (l,700X) -51- Figure 17. — PMC of showing a short broken bridge and fragments (1,?00X). $ 4 • : «a r* 0 Figure 18. — PJC of F. showing a non-movement of chromosomes to the poles at Aj (1,700X). -52- meiotic interphase at the end of the first division. They nay be more or less separated within the cell. However, their separate identities are maintained. Unlike the mitotic interphase, the meiotic interphase was characterized by the presence of a nucleolar-like body attached to each condensed chromosome. These nucleolar-like bodies were countable in this stage and gave a fairly accurate indication of the chromosome content of the meiotic interphase nucleus. After the first division each daughter nucleus, during interphase, was seen to possess approxi- mately 11 of these nucleolar-like bodies. The exceptional nucleus shown in Figure 19, however, had approximately 22 of these bodies. There was this single nucleus in the center of the pollen-mother- cell. There appears to be little doubt, therefore, that this was either a restitution nucleus or a nucleus in which the anaphase was abortive. In either case this nucleus seemed to possess the unreduced chromosome number. Pollen grain studies showed that more than 90 per cent of the pollen was apparently abnormal. The pollen grains were devoid of contents and did not take stain. The Amphidiploid and Progeny (Categories 4, 10 and 11) The fertile F2 (category 4) and its progeny, F^ (category 10) and F^ (category 11) are treated together as being araphidiploids. They are basically alike cytologically as well as morphologically, -53- Figure 19 .—PMC showing an unreduced nucleus following in F1 (1.70QX). ”1 -54- and constitute a group different from the diploids. As some of the individuals in these progenies exhibited segregation of parental characters it was thought desirable to keep track of different individuals . To facilitate this a pedigree has been constructed and is shown in Figure 20. ?2* fertile, 4n (No. 4) jbrphological characters In almost all respects this single plant was essentially like its progenitor, the F^ hybrid. It combined the characters of both the original parents. Like the F, it showed a strong resemblance to P. polys tachvus in vine habit, colored flowers, shape of wings, pigmentation of the hypocotyl and the in te modes and the near hypogeal germination. The seed coat color, like ^-s uniform purplish black. Figure 21 shows some of these characters and Table 1 coup ares the characteristics of the arnp hi diploids with those of the parent species . There were, however, a few important respects in which it showed a significant difference from the F^. A striking difference was observed in the fertility and seed set of the two generations. V/hile the F1 plant was nearly completely sterile, the ?2 was fairly fertile. There also appeared to be larger flowers in the latter. otology Cytological examination of the pollen- mother-cells and root tips revealed the 2n chromosome number to be 44. It was -55- F, (Fertile, kru) 1 (No. 4) F (fer+ile, *m) J (No. 10) “ F. (Fertile, lm ) 4 (Ho. 11) “ ?lant~l 2 Plant — 1 2 3 Single Plant 5 6 7 Figure 20. — Pedigree oF anphidioloids (F7, Fq and Fi) (categories 4, 10 and 11.) J -56- Figure 21. The amphidiploid F~ plant (category U) showing flower color, and leaf and pod shape . -57- apparent that a reduplication of the entire chromosome complement of the plant had occurred, in some way, to give rise to this amp hi diploid individual, A detailed analysis of the chromosome configurations at diakinesis and metaphase I showed that bivalent association of chromosomes had much increased as compared to that in F^, This is probably due to auto syndetic pairing giving rise 22 jj. All the pollen-mother-cells, however, did not show regular raeiosis. Irregularity was observed in the form of quadrivalents , univalents and an occasional tri valent. Thais it would appear that there were homologies in whole or part of some of the chromosomes thereby giving rise to multivalent associations resulting from the simultaneous occurrence of alio- and autosyndesis. About 50 per cent of the cells showed one or more quadrivalents. As many as 3 p with 14.^, and 1^ were observed. Figure 22 shows a pollen— mo the cell with a chain quadrivalent and 8 . The origin of most of the univalents appeared to be an early disjunction, which also characterized the parental species and the F^. Figure 23 shows another metaphase I with 6 univalents and 19 bivalents. Another characteristic feature of metaphase I was the presence of secondary associations between the bivalents, a possible indication alio syndetic homology. This was more clearly demonstrated in F^. A con^jarison of the mean diameter of this plant* s pollen grains (Table 10) and that of either parent (Tables 3 and 5) indicates a big difference in the volume of their pollen grains. If the mean diameter (49.48 ) of the F^ and a mean value between the two parents (40.38 ) be converted into volume the respective -58- Figure 22. — PMC of Y (category 4) showing a chain quadrivalent and 8j (2,675X). -59- Figure 23.— PIC of F2 (category 4) showing 19^ and 6 (2,675X). -60- Table 9 Frequencies of chromosome configurations in the PIC's of F2 (category 4) at diakinesis and metaphase I Configuration 'types Total I 0 2 4 0 6 2 0 8 1 II 22 21 20 20 19 19 18 16 14 m 0 0 0 0 0 0 0 0 1 IV 0 0 0 1 0 1 2 1 3 Frequencies 12 4 5 7 1 6 2 1 1 39 values are 62, 993.01^a 3 and 34,237.43^3. Thus the pollen grains of F have almost doubled in volume as compared to the intermediate value between the two parents. Table 11 compares the mean diameter of the pollen grains of F^ with that of P. lunatus . diploid species. It can be seen that there is a highly significant difference between the two. The frequency distribution of the pollen grain diameters (Figure 24) shows a greater range of dispersion as compared to the diploid ancestors. This might be an indication of some variation of chromosome contents of the microspores. Approximately 85 per cent of the pollen grains appear to be normal. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 -61- Table 10 Pollen grain diameter of F (category 4) Buds 1 2 3 4 5 60.90 55.10 52.78 56.84 49.30 49.88 53.36 44.66 55.10 ^♦6.40 58.00 55.68 45.24 40.60 51.62 53.94 55.10 48.14 49.30 44.66 49.30 53.94 46.98 46.98 52.20 50.46 49.88 55.10 37.70 51.62 49.30 40.60 52.20 56.26 52.20 46.98 55.68 53.94 42.92 41.18 53.94 63.80 49.30 48.72 52.20 40.60 46.40 47.56 48.72 46.98 53.94 37.70 50.46 44.66 46.40 42.92 55.10 41.76 53.36 ^6.40 50.46 41.18 51.62 44.08 39.44 45.24 43.50 48.14 40.02 51.04 64.38 52.78 51.62 47.56 46.40 59.74 59.16 35.38 44.66 49.88 45.24 56.2 6 41.18 49.30 52.20 46.40 59.74 57.42 49.30 40.02 49.30 56.84 46.98 42.34 40.02 -&.S4 _ 49.88 -55*10 41.18 ,027.76 1 ,041.68 975.56 939.61 962.94 = X = 4,947.55 49.48 Number of Pollen Grains -62- Figure 24. —Frequency distribution of pollen grain diameters of F2 (category 4). -63- Table 11 f0(categoryS4) °f F2 t(.TJl) 2(n-l) n - ~-9‘46 - lp -R7 .Z3H n 10.61 0.906 = 11.71 ** ** pooled sura of squares . Significant at .01 level. fertile, (category 10) fbrpho logical characters This progeny of the araphidiploid plant conprised 7 surviving individuals. Basically all these plants were similar in morphological features in which they also resemble their parents, the F2. The flower characters of the first four were observed and they resemble the F2 in this respect. The germination was hypogeal as seen in Figure 25. A comparison of the plants showed that they were not entirely uniform. There appeared to be segregation of some of the parental characters with perceptible variations in vigor and flowering response observed from plant to plant. Plants 2 and 3 were conparatively insensitive in photo- periodic response while plants 1 and U to 7 appeared to be responsive to short photoperiod. Plant 4 (Figure 26) was planted in uhe field in the fall of 1961 and the aerial parts were killed by the frost in December and January, 1962. It regenerated in 64- Figure 26. — Plant 4 of F_ (category 10) showing leaf, pod and flower characters and general habit of growth. Figure 25. — Germinating seedlings of F_ (category 10) showing near hypogeal germination. -65- the spring from the cold hardy rootstock, a P. nn1v«^.^y.r derived character. It was not determined whether this property was also possessed by its sibs. Cytology Chromosome number was determined to be 2jj = 44 by the examination of the pollen-raotherocells and in several eases, of the root tip squashes. However, one of the plants (No. 4) showed a deviation in chromosome number. The best estimate of the number that could be obtained from the examination of the pollen-motherocells of this plant was 2n = 42. Confirmation of this could not be made by the study of root tips as this plant had died. Another attempt was made to obtain verification of the above count. Root Up smears from the seedling of one of its progeny were examined. The number in this plant was estimated to be aj = 40. Bud material was available from several plants (Nos. 1, 2, 3 and 4). The examination of the pollen-motherocells from these plants showed that raeiosis was more or less irregular. Table 12 records the chromosome configuraUons as observed at diakinesis and metaphase I of the poUen-mother-cells. It can be seen that a total of about 50 per cent of the pollen-roother-cells had a regular meiosis with the formaUon of 22 bivalents. F3 thus showed comparatively greater regularity of meiosis than F2. Most of the irregulariUes here consist of varying numbers of quadrivalent formations as was the case in F Figure 27 shows a diakinesis with 18-r-r and ? . , H d 2jy. As many as 6 quadrivalents -66- f Figure 27.— Photomicrograph of PMC of F~ (category 10) plant 4 showing 18-q. and 2-^ at diafcLnesis (1,000X) Figure 28.— PIC of F^ (category 10) plant 4 showing 19,, and 6_ and several secondary associations between^the 1 bivalents (2f675X) -67- Total 22 37 80 o O o VO rH rH rH CM rv o ■4- i — i -3- -3- o r\ rH rH -tf- rH rv « — 1 O O -3- rH CM O rv rH rH VO rH (Vi rH O VO O rv rH l\l rv t — 1 CM r- O (VI r — i 00 O rH i — i rH oo rH i — 1 rH O 00 O CM Ov rv rH rH VO On O O rH rH rH -3- O O o CM CM (VI fH o (Vi rH o O o O rH VO CM VO (Vi rH CM CM rH O O W o (Vi O O O Cv- On VO (VI rH H rv CM rv M H III ft C o 3 g 3 -*-> C bo 3 ro E c +5 6 tT* -68- were observed in rare instances but the most commonly occurring number of quadrivalents was 1 with 20^. Thus the formation of quadrivalents as well as univalents was reduced in as compared to F2. Presence of secondary associations between the bivalents which was noticed for the first time in F^ was more clearly noticeable in F^, Figure 28 shows such associations between several of the bivalents at metaphase I in the pollen-mother- cells of one of the F^ plants (No. 4). Separation of the chromosomes at anaphase I was mostly regular, 22 chromosomes going to each pole (Figure 29). However, cases of 21/23 and 20/24 were also observed. Pollen grain measurements of diameter was recorded for 1 of the F«j plants in Table 13. The pollen grain mean diameters of the other two plants were not significantly different from each other or from the F 2 plant at .01 level. The diameter of the third one was significantly different from the F^ as shown in Table 14. This particular F^ plant (No. l) was the same which showed an aberrant chromosome number. Apparently an aneuploid loss conditioned a significant change in the volume of the pollen grains as indicated by the reduction in its diameter. The frequency distribution of the diameters does not appear to yield any significant information. The distribution shown by F^ plant 1 was, however, very much like that of P. polys tachyus (Figure 30). -69- ♦ # Figure 29. — Separating chromosomes in 22/22 at A_ in PMC of Tj (category 10) plant 4 (1,200X). 1 -70- Table 13 Pollen grain diameters of F (category 10) plant 1 in microns Buds Observation Numbers 1 2 3 4 5 1 49.30 45.98 49.30 46.40 49.30 2 37.70 44.66 48.14 46.98 46.40 3 44.66 46.98 44.08 44.08 45.82 4 48.72 49.30 50.36 38.86 49.30 5 44.66 35.38 43.50 44.08 43.50 6 31.90 48.18 50.46 44.08 47.56 7 37.12 47.56 46.98 42.92 38.86 8 40.02 50.46 51.62 40.60 47.56 9 46.40 52.78 46.40 46.40 41.18 10 44.66 42.34 43.50 46.40 52.78 11 41.76 37.70 46.40 41.18 47.56 12 51.62 41.18 41.18 45.82 49.30 13 40.60 40.60 44.66 39.44 39.44 14 52.78 42.34 43.50 47.56 45.24 15 63.22 37.12 46.98 46.98 39.44 16 42.34 44.66 43.50 40.60 43.50 17 43.50 41.18 47.56 43.50 46.40 18 52.20 40.02 43.50 44.08 41.18 19 45.24 48.14 42.34 35.96 42.34 20 Vt.66 38.86 34.80 ■ Aagfe 41.18 Total 903.06 876.38 911.76 877.16 897.84 = 4,466.20 x = 44.66 35 30 25 20 15 10 5 -71- diameter in microns igure 30. —Frequency distribution of diameter of pollen grains of (category 10) plant 1 -72- Table 14 "t" test for means of diameter of F (category 10) plant 1 and (category 4) t(.0l) = *1 ~ X2 _ 49.48 - 44.66 2(n-l) 2S2 / 2S? —L. n V n - 4.82 _ , ** - " 6-33 S| = pooled sum of squares. **Signif leant at .01 level. F^, fertile, (category 11) I'brohological characters This generation of the araphidiploid showed a continued resemblance to the original araphidiploid, the F 2 in most of the characters. However, it was by no means constant, at least, as far as morphological characters were concerned. One of the F^ plants (No. 2 ) showed white flowers (Figure 31) through in other respects it was not much different than its sibs. Figure 31 also shows a marked difference in flower size of a diploid and an araphidiploid. Cytology Cytological examination of the root tips and the pollen- mother— cells revealed that the 2n chromosome number was 44 in -73- Figure 31. —Flowers of a diploid, Fordhook like derivative and an amphidiploid, F^ (category 11) plant 2. -74- plants 1 and 2 the only plants available for study. Table 15 shows the chromosome configurations observed at diakinesis and metaphase I. There appeared to be no essential difference between the configurations observed for and F^. There were several trl valent and quadrivalent associations in these 2 F^ plants. Secondary associations were also observed (Figure 32). Table 15 Frequencies of chromosome configurations in FTC’s of Fr (category 11) at diakinesis and K_ Configurations Total I 0 2 0 4 0 l 1 n 22 21 20 20 18 18 14 m 0 0 0 0 0 1 1 IV 0 0 1 0 2 1 3 Frequencies Plant 1 8 4 2 3 1 18 Plant 2 _6 -2 — - JL _6 . _ Total 14 7 2 7 3 6 1 40 The amount of apparently viable pollen produced by plant ? is about 90 per cent. Figure 33 shows some of the pollen grains. There were no significant differences in the mean pollen grain diameters of F^ and F^ and Fg (Appendix, Tables 23 and 24), nor are there any significant differences in the frequency distribution of the pollen grain diameters of these progenies. -75- Figure 32. — PMC of F^ (category 11) plant 2 showing 20 and at metaphase I. -76- Figure 33. — Pollen grains of F^ (category 11) plant 2 (168X). -77- P. lmm tus var. Fordhook x P. polystachvus "sterile," diploid (category 5) Morphological characters The single plant belonging to this category was also like the F^ in several morphological characters including flower color and seed coat color. The germination of this plant was completely hypogeal. It was weak and was highly sterile. It set only two seeds so far even with manual pollination. The seedlings produced by this plant did not survive. The weak growth of the plant prevented collection of enough bud and root tip material for a thorough cytological examination and whatever buds were available did not fix well. A few pollen-mother-cells were, however, observed at metaphase I. Observation showed that this was a diploid individual with 2n = 22. Table 16 is a record of the chromosome configurations observed. Table 16 Frequencies of chromosome configurations of "sterile," diploid 7^ (category 5) Configurations 11^ 10^ 8JI 6 Total Frequencies 02417 -78- It can be seen that all the cells showed 2 or more univalents. A majority of cells had as many as 8 bivalents and 6 univalents. At anaphase I many fragments were observed. Thus, this plant had a very irregular meiosis. Over 95 P©r cent of the pollen was apparently abnormal. The irregular meiosis was probably responsible for this condition. Backcrosses and Progenies (Categories 6, 7, 8 and 9) x P, lunatus var. Fordhook (recurrent), and progeny (categories 6 and 7) Morphological characters The first backcross individual (category 6) was a perennial. In many morphological characters it showed a shift to lunatus form while in others it retained the characters of the F . In vine habit and seed coat color it was like the F^. It had, however, white flowers and did not show any response to day length. In these latter respects it was, therefore, like the Fordhook parent. It was relatively infertile. The progeny of this plant (category 7) consisted of 6 individuals. They showed a segregation in some of the parental characters including vine and bush habit. The germination character was also segregating in this progeny. Cytology Cytological examination showed that all these plants were -79- diploids with 2xi-22 chromosomes . Figure 34 shows a root tip cell of the first backcross individual. The frequencies of chromosome configurations observed at diakinesis and metaphase I of the pollen— mother-cells of the backcross and 2 of the progeny are recorded in Table 17. One of these two plants (plant 2) was a Fordhook bush type derivative and was interesting from the standpoint of breeding. Table 17 Frequencies of chromosome configurations in PfC's of first backcross and progeny at diakinesis and metaphase I Configurations un 10 n 9n 8 n 7 U 6 n 9 II Total 0I 2i 6i 8 I 10 i 1 IV BC^ 3 6 8 5 1 0 0 23 Progeny Plant 1 8 6 5 0 0 0 0 19 2 0 4 6 0 1 5 l 17 Total 11 16 19 5 2 5 l 59 It can be seen that meiosis was more or less irregular in all these plants. Only about l/3 of the total pollen-mother- cells showed a regular formation of 11 . Mostly there are 9ji and An occasional quadrivalent was also observed. Figure 35 shows a typical pollen-mother-cell from one of the first backcross progeny (plant 2). The amount of apparently good pollen was approximately 80 per cent in these plants. -80- Figure 3^.— A root tip cell from first backcross (category 7) showing 22 chromosomes at metaphase (3»5(XK ) . Figure 35.— A PMC from first backcross (category 7) showing 7ii» 8i at lli (3,50ox). -81- Second Backcross Progenies (Categories 8 and 9) Morphological characters These two progenies of the second backcross were closer to Fordhook in appearance than that of the first backcross. However, they contain several P. polys tachyus characters in combination. Both these progenies showed segregation of some of the parental characters which include bush and vine habit, and seed coat color. Figure 36 shows a vine and a bush segregate from the second generation. Individuals of the first generation also exhibited marked variations in fertility. Some were completely sterile while others were fertile and vigorous. Cytology An analysis of metaphase I figures of the pollen- mother- cells showed that all these derivatives were diploids with 32 = 22 number of chromosome and that meiosis was almost as regular as in the parental species. Tables 18 and 19 show the chromosome configurations in first and second generations respectively. Figure 37 shows metaphase I of a typical pollen- mother-cell from the first generation. In most of these plants approximately 90 per cent of the pollen grains were apparently viable (Figure 38). -82- Figure 36.— Vine and bush segregates from backcross 2 (lunatus recurrent) second generation (category 9). -83- * 4* $ Figure 37.— A PMC from backcross 2 (lunatus recurrent) first generation nlant (category 8) showing 10 and I ^ 3 1 jOO/. ) # i-I Figure 3e.-A nhotonicrograph showing pollen grains of i cto-oss ? <16®)“ reCUrrent) seoond €ene ration plant (category 9) -84- Table 18 Treqiiencles of different chromosome configuration types at diakinesis and^ of PMC’s of 20 plants from first generation of backcross 2 (lunatus recurrent) (category 8) Configuration T^pes Chromosome Complement * °24 0168 II 11 10 9 9 9 8 7 III o 0 0 0300 17 0 00 1000 6 Total 5 No. of 3 PMC 0 examined 1 16 2 15 3 18 4 23 5 12 6 21 7 6 8 8 10 16 12 26 13 6 14 10 15 4 16 7 17 24 19 21 20 6 21 5 25 18 26 16 Total 20 278 1 0 1 0 0 0 0 1 0 4 0 0 0 0 6 0 2 0 1 0 0 2 0 0 0 0 0 0 3 0 0 0 0 0 0 1 0 0 0 0 0 0 2 0 3 0 0 0 0 5 0 1 0 0 0 0 2 0 0 0 0 0 0 2 0 0 0 0 0 0 ^ 0 1 0 0 0 0 ^ 0 00000 3 4 00001 0 0 0 0 0 0 0 10 00000 30 0 0 0 0 0 20 0 0 0 0 0 00 00000 0 0 00000 50 00000 49 4 12 1 1 18 20 27 25 15 22 11 14 18 28 11 14 14 7 25 24 8 5 18 21 345 Table 19 Frequencies of different chromosome configuration types at diakinesis and Mj. of PMC’s of progeny of 2 first generation plants of backcross 2 (lunatus recurrent) (category 9) Configuration Types I 0 2 4 0 1 6 3 Total Chromosome II li 10 9 9 9 8 8 Complement hi 0 0 0 0 0 0 1 IV 0 0 0 1 0 0 0 Progeny of Plant 3 Offspring No. 1 5 12 3 0 0 1 0 21 5 3 14 1 0 0 2 0 20 Total 2 8 26 4 0 0 3 0 41 Progeny of Plant 6 Offspring No. 1 20 0 0 1 0 0 0 21 2 18 2 0 0 0 0 0 20 3 19 3 0 0 0 0 0 22 4 17 0 0 0 0 0 0 17 6 20 3 0 0 0 0 0 23 Total 6 94 8 0 1 0 0 0 103 Grand Total 102 34 4 1 0 3 0 144 DISCUSSION Affinities of Parental Species Evidence demonstrating the presence of taxonomic affinities between P. lunatus and P, nolystachvus has come to light from several different sources during the present investigation. First, there is evidence from the karyotype that not only the general size of the chromosomes was within the same range but also that several pairs of chromosomes of the two species agreed in morphological features to an extent that they may very well be homologous to each other. Evidence of taxonomic affinity between P. lunatus and _P. polystachyus is also derived from meiotic behavior of F^ hybrid. The pollen-mother-cells of the hybrid showed a variation in the number of bivalents. In rare instances 11 bivalents were observed. According to Stebbins (45) the pairing of chromosomes in species hybrids is affected by environmental conditions. However, the observations described here agree with those of Dhaliwal et _al, (10), with respect to the hybrid. It appears, therefore, that the pairing behavior was approximately the same under Gainesville and Utah conditions. There may, however, be genetic or other unknown factors affecting the pairing -86- -87- relationship in the hybrid. Nevertheless, there were indications that homologies occur in several pairs of chromosomes between the two genomes. Additional evidence is presented in the pairing relationship of the chromosomes during the meiosis in the amphidiploid. All the amphidiploid generations observed so far show the occurrence of quadrivalent associations at metaphase I. This indicates that homologies of sufficient strength exist in these chromosomes to allow a strong attraction between interspecific homologs to occur simultaneously with attractive forces between intraspecific homologs. The presence of secondary associations is an extension of the same phenomenon but indicates a lesser degree of attraction between interspecific chromosomes apparently only partly homologous. According to Lawrence (25) the phenomenon of secondary association is intimately connected with polyploidy and implies a certain relationship between the species. Genetic evidence of relationship was provided by the occurrence of segregation of parental characters in the amphidiploid generations. The appearance of a white- flowered plant in the amphidiploid can only be explained in either of two ways: (1) the occurrence of crossing over at one or more loci between segments of chromosomes from the two species, (2) the separation, without crossing over, from the quadrivalent of pairs of homologs derived from the same parental source. Nevertheless, the differentiation of the genomes during the process of speciation has proceeded to such an extent that a -88- hybrid between these two species was completely sterile. The presence of chromosome bridges and fragments at the meiotic anaphase I of the hybrid, indicates that part of the genomic differentiation has developed through inversions and translocations. According to Darlington (9) the sterility of F1 hybrids between closely related species is often due to these small chromosomal rearrangements. As the result of these rearrangements small chromosomal deficiencies and duplications result in non-viable gametes. Dobzhansky (11) also describes a similar situation in connection with the differentiation of races of Drosophila pseudo-obscura . Origin of the Amphidiploid The origin of the amphidiploid takes place as a result of natural or artificial hybridization followed by doubling of the chromosomes at some stage during the life cycle of the organism. There are several possible ways in which this can happen. The mode of origin can be basically classified in two types: 1 • The union of unreduced gametes During this process some gametes may have the combined haploid chromosomal complements of the two species as a result of meiotic irregularities in the species hybrid. The meiotic irregularity may be accompanied by a failure of an inadequately developed spindle to cause normal dipolar separation of the chromosomes at anaphase I. Consequently the chromosomes remain in a single group and coalesce to form a restitution nucleus -89- inst*ad of tw0 nuclei. This condition has been found in Nicotiana j&k&ccwn X It. 2yJ.y?strj,P (37), -Crepis rubra x £. foe-H Ha (36) and in several other species crosses, and is supposed to be responsible for the origin of their amphidiploids. Non- disjunction of chromosomes can also be brought about by the reunion of anaphase complements into a restitution nucleus. This kind of non- reduction has been reported by Tanaka and Kamemoto (50) in Vanda species. 2. Somatic doubling The other basic mode of origin of amphidiploid is due to the doubling of chromosome number in the somatic tissues of the hybrid. This doubling could occur at almost any stage in the life cycle of the plant. If it occurs in the hybrid zygote the entire F-j^ is polyploid. This kind of doubling or a doubling soon after fertilization probably accounts for formation of amphidiploids in £raasica x sameatrin (13), Hicotiana glutinosa X 2*. ■tabaccuffl (6) ^ other cases. Doubling may also occur in vegetative parts of an Fj^ hybrid. This would result in the formation of a chromosomal chimera with the tetraploid number in an otherwise diploid hybrid. Thus the hybrid may show both the 2n and % number in the pollen-mother-cells depending on the location of the buds. Such Fj hybrid would contain sterile and fertile sectors. This kind of origin of amphidiploids has been reported in Primula kfiWenfilfl (3^). Similar conditions were also suspected to have occurred following colchicine treatment of the tips of the vine cuttings of the original £. lamias x £. polustaoW hybrl(i reported by Lori (29). Amphidiploids my also arise through -90- the doubling of chromosomes in somatic tissues at a late stage of development. During this process an archesporial nucleus begins. to divide but the division is arrested after the chromosomes are split; thus a tetraploid pollen-mother-cell originates. Unlike the rest of the pollen-mother-cells of the interspecific hybrid, this type of cell undergoes normal meiosis and produces diploid microspores. This has been reported for the amphidiploid of Fragaria bracteata x F. helleri (19). Whether diploid megaspores can be produced in this way remains to be determined. Two additional modes of origin which are based on doubling of chromosomes in somatic cells, or production of unreduced gametes involve slightly different processes (14). An amphidiploid can originate by hybridization of two autopolynloids. Another indirect way of origin when there is non-reduction in only one type of gamete, concerns backcrossing in two successive generations. According to Goodspeed (14) the fertility of a newly discovered amphidiploid may point to the type of its origin especially in hybrids of distantly related species. In most of these cases the amphidiploids have been discovered as F2 plants produced by a rarely developed seed from the otherwise highly sterile plant as in Raohanobrassica (22). Thus, the amphidiploid in F2 of the hybrid P. lunatus x _P. polystachvus as found in the present investigation probably originated in this way. Somatic doubling as a means of origin is not supported by circumstantial evidence. It has been observed cytologically that in rare cases all the chromosomes aggregate in the center of the pollen- mother-cell -91- at anaphase I. This condition results from a failure of the chromosomes to move to the poles following metaphase I. A pollen-mother-cell has been observed with a nucleus containing an unreduced chromosome number following metaphase I. Such a pollen-raothei^cell would eventually produce two micro spores each with the diploid chromosome content. Assuming that a similar condition exists in megasporogenesis, chance fertilization of an unreduced egg by an unreduced sperm would give rise to a perfectly viable seed carrying the amphidiploid embryo. In the investigation unreduced pollen-mother-cells have been observed prior to the first division. Therefore, somatic duplication of chromosomes in the archesporial tissues is unlikely. Armhidiploidy with Respect to Speciation and Breeding The production of an amphidiploid in the genus Phaseolus is not only of academic importance with respect to speciation but may have valuable practical implications. According to Senn (40) Leguminosae is deficient in examples of polyploid species as compared to other plant families. In the literature concerning amphidiploidy in the Leguminosae only one other similar case has been reported. Dana (8) has obtained evidence of spontaneous amphidiploid in the hybrid Phaseolus aureus x _P. trilobus both of which are oriental species. According to the classification proposed by Stebbins (46) the P. lunatus x ?. nolystachyus amphidiploid falls in the category of segmental allopolyploids. The criterion for this classification -92- is the presence of partial intergenomal pairing of chromosomes. As a result of this condition a segregation of interspecific differences occurs so that in the later generations the arphidiploids produce segregates resembling more or less closely one or the other of the original parents. Clausen et al. in 1945 (5) advanced the hypothesis that segmental allopolyploids are likely to be less successful and, therefore, infrequent in nature. Two possible reasons are mentioned for this condition by the above authors. First, it is presumed that the incompatibility and sterility barriers which had developed during the differentiation of the diploid species would result in the appearance of weak and sterile types in the progeny. This condition has been noticed in a number of segmental polyploids, for example, Allium ceoa x A. fistulosura (20) and Tradescantia canaliculata x T. humilis (43) which are either themselves weak or completely sterile, or they produce sterile offspring. The second reason advanced by Clausen _ej, ^1. (5)» Tor the failure of the segmental allopolyploids is the occurrence of segregation and hence the lack of true breeding quality. These individuals may be thrown out of proper physiological balance with the environment. This condition has also been met within the above mentioned amphidiploids . The amphidiploids of Primula kewensis (34) and Triticum durum x T. Timopheevi (54) have also shown segregation in later generations in soite of their high degree of fertility and regular pairing of chromosomes at the start. The authors have, however, recognized the possibility that such types might be successful if they become -93- stable in later generations through the elimination of the weaker types. There are, however, some examples of segmental allopolyploids which have been carried through several generations and from which highly fertile and constant lines have been obtained. Kostoff (23) has shown that the fertility and constancy of the amphidiploid Nicotiana glauca x N. langsdorffii increased in the advanced generations. In this connection Stebbine (^6) has made the following significant statement, "There is good reason to believe that ... the segregating allopolyploids are a valuable source of new variants for the plant breeder and well worth his attention, although the production of useful types from them obviously requires considerable time." The occurrence of the white flowered F4 amphidiploid (Plant No. 2, Figure 20) lends support to this contention. Another valuable feature of segmental allopolyploidy from the plant breeders point of view is the possibility that they can be crossed to autopolyploids of their parental species and thus make possible the transfer of genes from one species to another on the tetraploid level when it is impossible on diploid level. From the standpoint of natural selection also it seems that the P. lunatus x _P. polys tachyus amphidiploids stand a better chance of survival as compared to amphidiploids which are annual in habit. It has already been demonstrated that this amphidiploid is a perennial and possesses efficient means of vegetative propagation in the form of a cold hardy rootstock. It is also capable of producing adventitious roots on vine branches in contact with the ground and some of these branches subsequently became established as independent segments of the clone. Such perennating amphidiploids may hybridize with a parental auto- tetraploid, which may exist in nature, thereby losing their well-known handicaps and increasing their variability and adaptability to new habitats. Thus, such amphidiploids can play an important role in speciation and evolution. Another possible use to which this amphidiploid can be experimentally employed is the derivation of aneuploid forms with the parental genom of either parent plus aneuploid numbers from the other. In a backcrossing program, for example, if the amphidiploid is backcrossed with P . lunatus . the resulting offspring will possess three genomes, two of which will be from the P. lunatus species and the third one from P, polys tachyus . During the meiosis of the backcross, 11 lunatus bivalents and 11 polystachyus univalents would be ideally present at metaphase I. The separation of bivalents at anaphase I is expected to be normal with 11 chromosomes going to each pole but the univalents will become distributed at random resulting in the formation of aneuploid gametes with a full haploid complement of the lunatus chromosomes with one or more chromosomes of polystachyus . A chance union with similar gamete will give rise to an aneuploid series consisting of 2n lunatus plus 1 to 10 pairs of polystachyus chromosomes. Theoretically, a large number of these combinations are possible. A plant breeder can thus have a large number of variants from which a selection of favorable segregates can be made. -95- Under natural conditions, though, the aneuploids which possess only favorable combination of genes will survive. The above conditions are possible on the behavior of whole chromosomes. The combinations are further augmented if crossing over occurs. In nature it is conceivable that crosses could occur between the amphidiploid and P. oolvstachyus with lunatus genes and/or chromosomes. SUMMARY Cytological investigations of the backcrosSjF^, "fertile" and "sterile" F2, and amphidiploid derivatives of the inter- specific cross Phaseolus lunatus x P. polys tachyus were conducted in the Department of Vegetable Crops, University of Florida as a supplement to the interspecific hybridization program in the genus Phaseolus . The plant material consisted of P. lunatus var. Fordhook, P. oolystachyus . their F-^ hybrids and backcross (P. lunatus recurrent) and amphidiploid derivatives. Acetocarmine squashes of the nollen-mother-cells were prepared according to a modified schedule suggested by Hyde and C-ardella (18), using Fe(OH)^ as a mordant. Root tips were obtained from germinating seedlings and indolebutyric acid treated cuttings. They were then prepared according to Fuelgen technique following pretreatment with 8-hydroxiquiniline as suggested by Sagawa (38) and employing pectinase after staining to facilitate dispersion of cells and spreading of the chromosomes. Karyotypes of the species P. lunatus and P. nolystachyus were studied and idiograms prepared. The two species appear to agree in morphological characters of several chromosome pairs while others are distinctly different. There is a significant -96- -97- difference in the pollen grain diameter of the two species, suggesting a possible correlation of a greater total length of the polystachyus chromosomes with increased pollen grain size. The chromosome number of 2n = 22, as previously reported, for both species, was confirmed. Dhaliwal's (10) observations regarding the irregularity of meiosis in the hybrid was also confirmed by observance of univalents at metaphase I, and of laggards, chromosome bridges and apparent fragments. Stages interpreted as concerned with the formation of restitution nuclei were observed immediately following metaphase I in the almost completely sterile hybrid. The chromosome number was established as 2n = 44 for the single fertile amphidiploid F2 plant and most of its derived Fj and F^ progenies with the exception of a single aneuploid plant. One F^ individual apparently had 2n = 42 chromosomes, but in view of the comparatively large number of chromosomes involved, this number is not positively established. Some meiotic irregularity in the amphidiploids was associated with the formation of quadrivalents which involves allosyndetic homologies between chromosomes as suggested by a comparison of the karyotype of the two species. The occurrence of one essentially amphidiploid F^ individual with white flowers is genetic evidence that amphidiploidy is not absolute and that the recombination of genes from the two species is possible for at least some homologous units in the -98- respective lunatus and oolystachvus genomes. Individual characteristics of both species and their amphidiploid derivatives are compared. All changes in the chromosome numbers were accompanied by a corresponding change in the size of pollen grains. All backcross derivatives studied had 2n = 22 number of chromosomes. Segregation of characters was observed with respect to vine habit and seed coat color. Some of the plants were fertile while the others were sterile. Essentially P. lunatus type? have been obtained but have some ^P. polys tachvus derived characters concerned principally with seed coat color and plant habit determination. The fertility of the amphidiploids was found to be high enough to permit easy maintenance of several lines and it would appear likely that later generations could result in the development of fully fertile lines capable of competing in the wild with the native _P. polys tachyus . Frequency distribution studies of pollen grain diameters indicated a fairly constant pollen size in _P. p o ly ? ta chyus . nearly the same degree of uniformity in P. lunatus but exhibited a greater increase in the range of variability in the amphidiploids. LITERATURE CITED 1. Allard, H. A. and H. F. Allard. 1940. The wild bean ■EhasfiQlus EQlystachvus (L.) E. S. P. Its chromosome number. J. Wash. Acad. Sci. 30:335-337. 2. Belling, J. 1926. The iron aceto-carmine method of fixing and staining chromosomes. Biol. Bull. 50:160-162. 3. Bhaduri, P. N. and P. N. Ghosh. 1954. Chromosome squashes in cereals. Stain Technology 29:269-276. 4. Buxton, B. H. and W. C. F. Newton. 1928. Hybrids of Tftritabs their fertility and cytology. Jour. Genetics 19:269-277. 5. Clausen, J., D. D. Keck and W. M. Hiesey. 1945. Experimental studies on the nature of species II. Plant evolution through amphidiploidy and autoploidy with examples from Madinae. Publ. Camagie Instn. No. 564. 6. Clausen, R. E, and T, H. Goodspeed. 1925. Interspecific hybridization in Nicotians. II A tetraploid glutlnosa- Tabaccum hybrid, an experimental varification of Wlnge's hypothesis. Genetics 10:279-284. 7. Conger, A. D. and L. Fairchild. 1953. A quick-freezing method for making smear slides permanent. Stain Technology 28:281-283. 8. Dana, S. 1961-62 Personal communication. 9. Darlington, C. D. 1937. Recent Advances in Cytology. 2nd. ed. Blakiston, Philadelphia. 10. Dhaliwal, A. S., L. H. Pollard and A. P. Lorz. I96I. Cytological behavior of an species cross ( Phaspnl up lanatus L. var. Fordhook fbaseolus polys tachvus L.) In press . 11. Dobzhansky, T. 1941. Genetics and the Origin of Species, Columbia University Press, New York. -99- -100- 12. Florida State Marketing Bureau, Annual Agricultural Statistics Summary. Nov. 1959. Annual Report No. 41, pp. 70-71. 13. Frandsen, H. N. and 0. Winge. 1932. Brassica napocampestris, a new constant amphidiploid species hybrid. Hereditas 16:212-218. 14. Goodspeed, T. H. and M. V. Bradley. 1942. Amphidiploidy. Bot. Rev. 8:272-316. 15. Honma, S. 1956. A bean interspecific hybrid. Jour. Heredity 47:217-220. 16. and 0. Heeckt. 1958. Bean interspecific hybrid involving Phaseolus coccineus x P. lunatus. Proc. Amer. Soc. Hort. Sci. 72:3^0-3^4. 17. . 1959. Interspecific hybrid between Phaseolus vulgaris and P. lunatus . Jour. Heredity 50:233-237. 18. Hyde, B. B. and C. A. Gardella. 1953- A mordanting fixation for intense staining of small chromosomes. Stain Technology 28:305-308. 19. Ichijiraa, K. 1926. Cytological and genetic studies on Fragaria. Genetics 11:590-604. 20. Jones, H. A. and A. E. Clarke. 1942. Natural amphidiploid from an onion hybrid Allium ceoa L. x Allium fistulosa L. Jour. Heredity 33:25-32. 21. Karpechenko, G. D. 1925. On the chromosomes of Phaseolinae. Bull. Appl. Bot. and FI. Breeding 14(2) :143-148. 22. Karpechenko, G. D. 1928. Polyploid hybrid of Raphanus sativus L. x Brassica oleracea L. Bull. Appl. Botany 17:365-408. 23. Kostoff, D. 1939. Studies on polyploid plants. XXI Cytogetic behavior of the allopolyploid hybrid Nicotiana glauca Grah. x Nicotiana Langsdorffii Weinm. and their evolutionary significance. Jour. Genetics 37:129-209. 24. Lamprecht, H. 1941. Die Artgrenze zwischen Phaseolus vulgaris L. und multiflorus Lam. Hereditas 27:51-175. 25. Lawrence, W. J. C. 1931. Hie secondary association of chromosomes. Cytologia 2:352-384. 26. Lorz, A. P. 1952. Interspecific cross in Phaseolus . Science 115:702-703. -101- 27. Lorz, A. P. 1953. Preservation of Phaseolus polystachyus gem plasm in crosses with P. lunatus . Unpublished report. 28. . 1957. Interspecific hybridization in Phaseolus at the University of Florida, Ann. Rept. Bean Improvement Cooperative 1957 (I):l6-17. 29. . 1962 Personal communication. 30. MacDonald, M. D. 1961. Barley pachytene chromosome and the localisation of Erectoides translocation point on chromosome 5. Canadian J. Genetics and Cytology 3:13-17. 31. Mackie, W. W. 1943. Origin, dispersal and variability of the lima bean, Phaseolus lunatus Hilgardia 15(1). 32. Metzger, R. J. and E. R. Leng. 1955. A smear technique for some species of Leguminosae. Stain Technology 305^1-^2. 33. Meyers, J. R. 1945. Prefixing with para-dichlorobenzene to facilitate chromosome study. Stain Technology 2-:121-l25. 3^. Newton, W. C. F. and C. Pellew. 1929. Primula kewensls and its derivatives. Jour. Genetics 20:40 5- 4^7. 35* O'Mara, J, G. 1958. Acetic acid method for chromosome studies in prophase and metaphase in meristems. Stain Technology 201-204. 36. Poole, C. F. 1931. The interspecific hybrid Creols rubra x £. foetida and some of its derivatives I. Univ. Calif. Publ. 37. Rybin, V. A. 1929. Cytological features of the alio te trap lo id Nicotians Tabacum x Njcotiana sylvestris. Proc. USSR Congr. Genet. Plant and Animal Breed. 2:437-445. 38. Sagawa, Y. 1961. Personal communication. 39. Sen, N. K. and J. G. Bhowal. I960. Cytotaxonomic studies in Vjgna. Cytologia 25:195-207. 40. Senn, H. A. 1938. Chromosome number relationship in the Leguminosae. Bibliographies Genetics XII. 41. Setterfield, G., R. Schrieber and J. Woodward. 1954. Meiotic frequency determinations and microphotometric fuelgen dye measurement in root tins. Stain Technology 29:113-120. 42. Singh, D. N. 1961. Use of growth induced rooting in cytological studies. Nature (London) 189:420-421. -102- 43. Skirra, G. W. 1942. Bivalent pairing in an induced tetraploid of Trandescantia. Genetics 27:635-639. 44. Small, J. K. 1933. Manual of the Southeastern Flora. Hie Science Press Printing Co., Lancaster, Pennsylvania (1st. ed. ) 723-725. 45. Stebbins, G. L., Jr. 1945. The cytologlcal analysis of species hybrids II. Bot. Rev. 11:463-486. . 1947. Types of polyploids: their classification and significance, Adv. Genet. I 403-429. 47. - 1957. Variation and Evolution in Plants. Columbia University Press, New York. 48. Strand, A. B. 1943. Species crosses in genus Phaseolus . Proc. Amer. Soc. Hort. Sci. 42:569-573. 49. Swaminathan, M. 5., M. L. Magoon and K. L. Mehra. 1954. A simple propiono-carmine PMC smear method for plants with small chromosomes. Indian Jour. Genet. Plant Breeding 14:87-88. 50. Tanaka, R. and H. Kamenoto. I960. Meiotic chromosome behavior in diploid and polyploid Vanda Orchid hybrids. Cytologia 25:405-418. 51. Thompson, H. C. and W. C. Kelly, Vegetable Crops. McGraw Hill Book Co., Inc., New York. 52. Vosa, C. G. 1961. A modified aceto-orcein method for pollen mother cells. Caryologia 14:107-110. 53. Yamell, S. H. 1931. A study of certain polyploid and amphiploid forms in Fragaria . Genetics 16:455-489. 54. Zhebrak, A. R. 1946. New amphiploid series of wheat and their significance for selection and evolution. Amer. Nat. 80:271-279. APPENDIX -104- Table 20 Diameter of pollen grains of F~ plant 3 in microns ^ (category 10) Observation Numbers 1 2 3 4 5 1 51.04 ^3.50 48.16 41.18 45.24 2 51.62 47.56 53.36 36.54 48.70 3 51.62 41.76 44.66 52.20 46.98 4 51.62 52.20 37.70 42.92 44.08 5 51.04 44.08 36.5^ 53.36 47.56 6 57.42 50.46 49.30 45.82 47.56 7 46.40 44.66 43.50 53.94 43.50 8 51.62 42.34 52.20 36.54 52.78 9 56.84 38.86 52.78 5^.52 42.39 10 49.88 41.76 49.88 44.08 49.88 11 51.04 41.18 49.88 47.56 55.10 12 50.46 39.44 40.60 45.82 50.46 13 51.62 49.30 55.10 51.62 50.46 14 52.20 50.46 40.60 53.94 46.40 15 45.82 49.30 37.70 37.70 56.84 16 44.08 45.24 48.14 37.70 44.08 17 51.62 45.24 52.78 50.46 52.78 18 53.36 44.08 55.10 53.94 46.98 19 47.56 43.50 52.78 46.98 48.14 20 2ZA? 45.82 46.40 55.68 46.40 Sun 1,003.98 900.74 947.14 942.50 966.26 x = 47.61 -105- Table 21 Diameter of pollen grains of (category 10) plant 4 in microns Observation Numbers 1 2 3 4 5 1 52.78 39.44 40.60 41.18 52.20 2 44.08 56.84 48.72 38.28 46.40 3 43.50 52.78 53.94 46.98 40.02 4 62.64 40.60 40.60 43.50 50.46 5 51.04 62.64 38.28 42.92 51.04 6 54.52 49.30 37.70 47.56 41.18 7 40.60 52.20 31.90 68.72 51.04 8 40.46 53.36 34.22 47.56 46.98 9 50.46 52.20 51.04 43.50 48.72 10 42.34 53.36 49.30 43.50 46.40 11 53.94 54.52 51.62 49.30 47.56 12 44.08 53.94 48.72 39 M 49.88 13 42.34 59.16 44.08 45.82 51.04 14 58.00 52.20 52.20 52.20 48.14 15 42.92 55.10 42.34 49.30 38.86 16 55.10 44.66 49.88 53.36 40.02 17 54.52 44.66 41.18 48.72 40.46 18 5?.20 51.04 42.92 36.34 45.24 19 34.96 47.56 47.56 51.62 49.88 20 44.66 55.10 49.88 49.88 38.28 am 976.14 1,030.66 896.68 917.68 933.80 4,754.96 x = 47.55 -106- Table 22 Pollen grain diameters of (category 11) plant 1 in microns Observation Number Sum 1 53.94 44.08 58.00 46.30 53.90 2 49.88 54.52 40.60 49.88 55.68 3 55.10 34. 80 42.34 47.56 52.20 4 51.04 45.24 41.18 58.58 37.70 5 51.04 46.40 48.72 52.78 53.36 6 49.88 56.58 52.78 54.52 52.20 7 56.84 54.52 40.60 47.56 49.30 8 46.26 51.62 47.56 41.76 52.20 9 46.40 40.60 46.40 43.50 52.20 10 44.66 34.80 48.14 34.22 44.66 11 54.52 50.46 49.88 46.40 52.20 2 40.02 44.08 45.24 48.72 52.78 13 42.92 40.60 42.92 37.12 50.46 14 44.08 51.04 40.60 55.68 46.98 15 44.08 55.10 46.40 48.72 44.08 16 40.60 52.20 40.02 47.50 52.20 17 44.66 47.56 51.04 39.44 46.98 18 46.40 45.82 51.04 34.22 44.66 19 51.62 51.62 52.94 45.24 45.24 20 48.72 49.88 40.60 , jt2>5Q 37.12 972.66 953.52 927.00 923.20 976.10 = x = 47.52 -107- Table 23 Pollen grain diameter of (category 11) in plant 2 in microns Observation Numbers 1 49.88 42.92 2 38.28 42.92 3 48.72 40.60 4 47.56 38.28 5 59.16 42.92 6 42.92 37.12 7 38.28 45.24 8 48.72 51.04 9 46.40 44.08 10 53.36 52.20 11 49.88 49.88 12 47.56 47.56 13 45.24 42.92 14 63.80 46.40 15 44.08 60.32 16 52.20 61.48 17 59.16 59.16 18 58.00 58.00 19 48.72 37.12 20 55.68 40.60 987.60 930.76 46.40 53.36 48.72 51.04 56.84 41.76 51.04 54.52 49.88 44.08 51.04 46.40 44.08 56.84 52.20 39.44 53.36 46.40 38.28 44.08 48.72 54.52 52.20 41.76 42.92 48.72 54.52 48.72 44.08 53.36 41.76 27.12 49.88 44.08 35.96 48.72 32.48 41.76 48. 72 48.72 54.52 47.56 52.20 47.56 44.08 49.88 39.44 54.52 51.04 53.36 52.20 45.24 49.88 54.52 54.52 39.44 46.40 46.40 -52^6 _ JU6 926.84 957.42 983.68 = 4,786.30 y = 47.86 Number of Pollen Grains -108- Figure 39. --Frequency distribution of pollen grain diameters of F^ (category 10) plant 3. Nunber of Pollen Grains -109- Figure 40.— Frequency distribution of pollen grain diameters of (category 10) plant 4. Number of Pollen Grains -110- -111- -3- vO O VO VO VPv CM 3 5 O -4" VO rv CVl rv suf^UQ uaxioj jo aaqurifj diameter in microns Figure ^2.— Frequency distribution of pollen grain diameters of Fr (category 11) plant 2. BIOGRAPHICAL SKETCH Eirendra Singh Fozdar was born July 6, 1919, at Hasanpur, India. In 193^, he graduated from Government Intermediate College, Etawah. In 1938, he received the degree of Bachelor of Science in Agriculture from Agra University and the degree of Master of Science in Botany from the same University in 1940. From 1942 to 1953, he served in the Department of Agriculture, Government of Uttar Prasdesh, as Lecturer in Botany at the Agricultural College, Kannur. At the same time he also served in the National Cadet Corps of India. From 1953 to August, 1959, he served as Lecturer in the Central College of Agriculture, Indian Agricultural Research Institute, New Delhi. In September, 1959, he enrolled in the Graduate School of the University of Florida. He worked as a graduate assistant in the Department of Vegetable Crops while pursuing his studies towards the degree of Doctor of Philosophy. Eirendra Singh Fozdar is married to the former Vidya Wati Devi and is the father of two children. He is a member of the Indian Society of Genetics and Plant Breeding, Indian Horticultural Society, American Society for Horticultural Science and Gamma Sigma Delta. -112- This dissertation was prepared under the direction of the chairman of the candidate’s supervisory committee and has been approved by all members of that committee. It was submitted to the Dean of the College of Agriculture and to the Graduate Council, and was approved as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August 11, 1962 Dean, College of Agriculture