Journal of Heredity 32: 418-422 (1941)
CHROMOSOME NUMBER AND HYBRIDIZATION IN GLADIOLUS
*
RONALD BAMFORD
*Scientific Paper No. A5. Contribution No. 1772 of the
Maryland Agricultural Experiment Station (Department of Botany).

IN a previous paper2 a list of the chromosome numbers of many species and varieties of Gladiolus was presented and the earlier works on this subject reviewed. At that time particular emphasis was placed on the heteroploid nature of the genus and since then other reports23,33 have confirmed these results. When the many historical records of the development of this group, 6,7,12,13,17,26,34 particularly in relation to the development of the summer-flowering types, are compared to such chromosome lists it appears that species and varieties with many different chromosome numbers28 have been involved. This has again been emphasized in some recent works on hybridization5,14,15,19,20,21,29 concerning the South African species. The presence of triploid (3n=45) and pentaploid types (5n=75) in the chromosome lists and the infrequent reports of sterility,18 clearly indicated that at least diploid-tetraploid and tetraploid-hexaploid crosses have been made. It has been known for over a century12 that the South African species could be intercrossed but these are all diploids.

For the past six years our attention has been devoted not only to checking many, of the reports indicated above but we have extended our crosses to the widest possible extremes in so far as chromosome number was concerned. Although the study is not complete, many interesting hybrids have been obtained and it seemed wise to summarize at this point. This report presents the results thus far and extends a previous summary.3

Materials and Methods

The seeds and corms were secured from a variety of sources. In addition to those mentioned previously,2 Mr. W. M. James of Santa Barbara. California, has generously provided some other species, and even some of his own hybrids, which were used in this study.

The majority of the crosses were made in the greenhouse between January and May. The flowers or buds were emasculated in the early morning or on the previous afternoon and the anthers were placed in glass vials and stored in a cool place. Preliminary experiments showed that fresh pollen was best and after two days it was generally discarded. Pfeiffer25 has recently shown that the viability of Gladiolus pollen decreases with age. Bagging was resorted to only when pollinations were made outside during the summer.

The seeds were collected as they matured and, after a summer storage period, planted the next fail. Seedling counts were taken at intervals throughout the winter and spring months. A successful hybridization was arbitrarily considered to have been made when the seedlings appeared. This was arrived at because in many cases the plants died during the first year. On the other hand, many groups of seeds, which appeared to he good, did not germinate and it is possible with better cultural conditions the results might have been altered to some extent.

During the past six years over twenty-five thousand pollinations have been made involving about four thousand different crosses. The majority involved diploid species because these are very abundant and flower abundantly during the winter.

Chromosome Number

Since the publication of the last list certain acquisitions have been made and the chromosome number of these is reported below:

TABLE I.

Chromosome Numbers of Gladiolus Species.

NAME

n

2n

G. communis L. variety

90

180

G. ———— var. Golden Orange

45

90

G. illyricus Koch

 

90

G. watsonius Thunb.

 

30

G. maculatus Sweet

 

30

G. watermeyeri L. Bolus

 

30

G. martleyi L. Bolus

 

30

G. namaquensis Ker. Gawl.

 

30

It is apparent that our previous report2 of the heteroploid nature of the genus is extended. The form of G. communis has the highest number thus far reported for this genus.

Hybridization

The results of the extensive series of pollinations resulted in nearly sixteen hundred successful hybrids, many of which were reciprocals. The comparative success in relation to chromosome number appears in Figure 1

The problem of interspeci ic hybridization generally centers around two points (1) Whether or not the cross can be made and (2) whether the resultant hybrids are fertile or sterile. Both have some relation to chromosome number.

In Gladiolus it seems certain (Figure 1) that, except under wide differences. the range of possibilities for hybridization is great. altli( )ugh it is evident that the chances of successful hybridization diminish as the chromosome numbers of the Species or varieties become farther apart. Sharp31 points out what we have observed, that "as a general rule, one may expect the difficulty of obtaining a hybrid to increase with numerical and qualitative dissimilarities in the chromosomal complements of the parental types.

It appears then, as though the general report from many sources, as expressed by Emsweller et al8 is true, namely that the ''original species have been combined and hybrids have been crossed and recrossed until the resulting multiple hybrids possess characters derived from many species." The reported origin of the horticultural type as a cross between G. psittacinus and G. cardinalis or G. oppositiflorus represents a hexaploid-diploid cross. We have not had an opportunity to make this actual cross as yet, due chiefly to the difference in flowering dates, but a similar cross involving some of the diploid S. African species (Frontispiece A) and G. byzantinus (Frontispiece N), a hexaploid, has been made. This same type of cross was reported nearly 50 years ago by Beal6

GLADIOLUS CROSSES MADE
Figure 1
 

30

45

60

75

90

120

138±

180

30

+++

+++

++

++

+

+-

+-

+-

45

++

+

++

+

+-

-

0

+-

60

++

+

+++

++

++

++

+++

++

75

+-

0

+

-

++

-

-

-

90

+-

-

+

+-

++

++

-

-

120

-

0

-

0

-

-

0

+++

138±

+-

0

0

-

0

0

0

0

180

0

0

0

0

0

0

0

0

The chromosome constitution of gladiolus varies from 30 (diploid) to 180 (dodecaploid). This chart shows the 64 possible crosses between the various types. +++ = 50-100%; ++ = 25-50%: + = 5-25%: +- = Rare: - = Negative: 0 = Not made.

Over one hundred years ago Dean Herbert12 was successful in combining many diploid species into one hybrid. This is clearly indicated by our results (Frontispiece B, C, D, F, G) and we now have some hybrids with seven species involved in their parentage. Time is the only factor which has prevented us from increasing that number. A similar hybrid with only five species is presented in Frontispiece C.

One of the critical points in the history of Gladiolus has been the reported crosses involving tetraploids and diploids. This seemed to be the hereditary bridge over which many of the diploid characters have entered into certain present tetraploids. That such hybrid triploids (Frontispiece H, I, J, L, P) can be formed is evident and we have made over two hundred different ones. The most interesting point (Figure 1) is that many, used mostly as seed parents, are partially fertile and, when backcrossed with either their parents or similar types, some of them form progeny (Frontispiece A). We now have the chromosome numbers of many, of these second generation crosses and the results will be presented separately. Most of them are aneuploids. It might be well to mention that some of these aneuploids are also fertile (Figure 2).

A TRIPLE HYBRID
Figure 2
This showy hybrid has a complex ancestry as follows: [Professor Donders (60) x G. tristis (30)] x G. tristis (30). It is an aneuploid, and fertile.

The widest cross involving a diploid species was a diploid-dodecaploid cross made in 1940. This represents the widest possible combination. Another wide cross is presented in Frontispiece E.

The bulk of the commercial types of Gladiolus are tetraploids and it is apparent that, once they are formed. they represent an extremely plastic group with reference to hybridization. They are not only easily crossed with diploids, triploids, and other tetraploids, but also with the higher polyploids (Frontispiece A). The pentaploid types previously reported2 undoubtedly represented hybrids between the tetraploid varieties and such species as G. psittacinus and G. dracocephalus which are hexaploids. Like the triploids, the pentaploids which we have produced are partially fertile when backcrossed with their parents or similar types. The progeny of such crosses are aneuploids. This again may account for the presence of characters in the tetraploid summer-flowering types which are truly of hexaploid origin. Our information concerning hybrids between the higher polyploids is not exhaustive enough at this time to warrant any conclusions. They represent two distinct types of species: the so-called Eurasian group and the tropical African group. The Eurasian group crosses easily with each other, even though their chromosome numbers are different, but we have had considerable difficulty with germinating apparently good seeds.

The information presented above is the usual result experienced by the students of many genera. Either experimentally or by the observation of chromosome conditions in suspected hybrids it has been shown repeatedly that interspecific hybridization is possible and more than likely it is one of the chief factors in the phylogenetic development of any group of plants. Hybridity is the basis of allopolyploidy as an evolutionary principle. The work on Viola,4,10 Chrysanthemum,32 Triticum, Aegilops, etc.,1,30 Nicotiana,11 Narcissus,24 Rosa,9 Tulipa35 and many others have shown that interspecific hybridization must have been a factor in their development. The presence of many aneuplolds in Iris,16,27 suggests that a great deal of interspecific hybridization has occurred there, as it appears to have in Gladiolus.

Summary

    *The author wishes to acknowledge the valuable assistance of Mr. R. E. Jones in making, recording and growing many of these hybrids.
  1. The heteroploid nature of the genus is confirmed and extended.
  2. Hybrids can be made between species representing the limits of the genus, as measured by chromosome number, but the comparative success is less as such limits are reached. This confirms the recorded history of the development of the commercial gladiolus.
  3. Some of the triploids and pentaploids are fertile and aneuplolds are produced. Some of these aneuploids are now producing progeny. This may show the path by which characters of diploid and hexaploid species have been transmitted to commercial tetraploids.*

Literature Cited

  1. AASE, H. C. Bot. Rev. 1: 467-496. 1933
  2. BAMFORD, RONALD. Jour. Agr. Res. 51: 945-950. 1935.
  3. ΡΡΡΡ Gladiolus 40-43. 1937
  4. ΡΡΡΡ and GERSHOY, A. Vt. Agr. Exp. Sta. Bull. 325. 53 p. 1930.
  5. BARXHART, P. D. Flor. Exch. 83-3: 23. 1934.
  6. BEAL, A. C. N. Y. Agr. (Cornell) Ext. Bull. No. 9:93-188. 1916.
  7. CRAWFORD, M. The Gladiolus. New York, N. Y. 100 pp. 1921.
  8. EMSWELLER, S. L., ET AL. U.S.D.A Yearbook. 890-998. 1937.
  9. ERLANSON, E. W. Gen. 16: 75-96. 1931.
  10. GERSHOY, A. III. Vt. Agr. Exp. Sta. Bull. 367. 91 p. 1934.
  11. GOODSPEED, T. H. Univ. Cal. PubI. Bot. 17: 369-398. 1934.
  12. HERBERT, WM. Herbertia 4: 29-62. 1937.
  13. HOTTES, A. C. Jour. Hered. 6: 499-504. 1915.
  14. JAMES, W. M. Glad. Bull. (N. Y.) 2(10): 7-9. 1935.
  15. JAMES, W. M. Ibid. 3(5): 5. 1936.
  16. LONGLEY, A. E. Amer. Iris Soc. Bull. 29: 43-49. 1928.
  17. MCLEAN, F. T. Glad. Rev. 4: (3)30-31. 4: (4)23-28. 1927.
  18. ΡΡΡΡ Ibid. 5: 303-304. 1928.
  19. ΡΡΡΡ Jour. N. Y. Bot. Gard, 34: 73-80. 1933.
  20. ΡΡΡΡ Jour. Hered. 29: 115-121. 1938.
  21. ΡΡΡΡ Bull. Torr. Bot. Club. 65: 181-197.  1938.
  22. ΡΡΡΡ The Gladiolus. 197 pp. New York, N. Y. 1941.
  1. MENSINKAI, S. W. Cytologia 10: 59-72. 1939.
  2. NAGAO, S. Mem. Coll. Sci. (Kyoto) 8: 81-199. 1933.
  3. PFEIFFER, N. E. Contrib. Boyce Thomp. Inst. 10 (4): 427-440. 1939.
  4. PRIDHAM, A. M. S. N. Y. Agr. (Cornell) Ext. Bull. No. 231. 1-65. 1932.
  5. RANDOLPH, L. F. Amer. Iris Soc. Bull. 52: 61-66. 1934.
  6. ΡΡΡΡ Glad. Bull. (N. Y.) 3(4): 10-13. 1936.
  7. ROGERS, D. M. Gladiolus 44-47. 1936.
  8. SANDO, W. J. Jour. Hered. 26: 229-232. 1935.
  9. SHARP, L. W. Introduction to cytology. New York, N. Y. 567 p. 1934.
  10. SHIMOTOMAI. N. Jour, of Sci. (Hiroshima) 2:1-100. 1933.
  11. SUGUIRA, T. I. Cytol. 7: 544-595. 1936.
  12. VAN FLEET, W. Mem. Hort. Soc. N. Y. 1: 143-149. 1902.
  13. WOODS, M. W. and BAMFORD, R. Amer. Jour. Bot. 24: 175-184. 1937.

Hybrids Between Gladiolus of Varied Chromosome Constitution

The genus Gladiolus contains species and varieties differing greatly in chromosome number. The basic haploid chromosome number is 15, and many 30-chromosome diploids are known. Most of the commercial varieties of gladiolus are tetraploids with 60 chromosomes. This composite photograph shows species and hybrids with the chromosome number of the parent forms in parenthesis.

 
A G. tristis v. concolor (30)
B G. tristis (30) x G. hirsutus (30)
C G. tristis (30) x G. watsonius (30)
D [G. callistus (30) x G. undulatus (30)] x G. blandus (30)
E G. tristis (30) x G. communis (138±)
 
F [G. cuspidatus (30) x G. tristis (30)] x [G. blandus (30) x G. anqustus (30)]
G [G. callistus (30) x G. undulatus (30)] x [G. tristis (30) x (G. tristis (30) x G. hirsutus (30))]
H Professor Donders (60) x G. tristis (30)
I Joost v. d. Vondel (60) x G. hirsutus (30)
 
J Prof. Donders (60) x G. hirsutus (30)
K Miss Bloomington (60) x G. pappei (30)
L Miss Bloomington (60) x G. angustus (30)
M Prof. Donders (60) x G. communis (138±)
 
N G. byzantinus (90)
O Prof. Donders (60) x G. byzantinus (90)
P Edith Mason (60) x G. cuspidatus (30)
Q Nymph (45) x Miss Bloomington (60)
R Dillenberg (60) x G. communis (138±)