Cytotogia 37: 317-326, 1972 Breeding System, Crossability Relationships and Isolating Mechanisms in the Solanum nigrum Complex J. Venkateswarlu and M. Krishna Rao Department of Botany, Andhra University, Waltair, India Received October 29, 1970 Solanum nigrum L. and its related species, which go by the name black night shades or deadly nightshades, constitute a taxonomically difficult species complex of very variable forms with world-wide distribution. This species complex belongs to the section Morella of the genus Solanum. Although Bhaduri (1933, 1945 and 1951), Chennaveeraiah and Patil (1968) and Tandon and Rao (1964, 1966a and b) studied the Indian representatives of the group they remain to be understood fully compared to the European and American representatives (Stebbins and Paddock 1949, Soria and Heiser 1961, Heiser, Soria and Burton 1965). With this aim in view we have studied the cytogenetics of the S. nigrum complex utilizing 16 acces sions including seven diploids, five tetraploids and four hexaploids from Indian as also from other geographic regions. This paper embodies results relating to breed ing system, crossability relationships and isolating mechanisms. Materials and methods Source of the material, chromosome number, name of the collector and acces sion number are given in Table 1. Hybridization: For purposes of hybridization flower buds were emasculated two days prior to their opening by removing the anthers only. All the plants used in crosses were grown in insect proof green house and the flowers were protected with butter paper bags before and after pollination. Pollen was collected on to a clean sterilised slide by tapping the flowers and it was applied by gently rubbing the stigma against the pollen surface of the slide. Fixation and staining: In the cross Indian hexaploid S. nigrum ~Indian diploid S. nigrum, for the study of pollen germination and pollen tubes flowers were fixed in 1:3 acetic-alcohol for 24 hours; then styles were cleared in chloral hydrate, hydrolysed in 1 N HCl for a few minutes and washed well in tap water. Styles were subsequently squashed in a drop of acetocarmine placing a cover-slip over the material and gently pressing under several folds of filter paper to flatten the styles. For study of events following fertilization, material was embedded in parafin and microtome sections were cut 12 to 15 p and stained in Delafield's hematoxylin following the conventional methods.
318 J. Venkateswarlu and M. Krishna Rao Cytologia 37 Table 1. Source of the material, chromosome number, name of the collector and accession number of S. nigrum complex species and races Observations Breeding system: Though all the species and races of the present study are usually self-pollinated, cross-pollination occasionally occurs. One definite case of natural hybridization between Indian hexaploid and one of the other two hexa ploid races of S. nigrum, New Zealand and Denmark, was encountered during this study. In three plants, which were on the basis of morphological features suspected to be hybrids of this origin, pollen sterility was high and meiotic irregularities were
1972 Breeding System, Crossability Relationships and Isolating Mechanisms 319 frequent. Subsequent comparison with experimentally produced hybrids between natural hexaploids left no doubt as to the hybrid origin of the three plants. Crossability relationships: Crosses were made in 48 combinations utilising representative forms of the three naturally occurring ploidy levels. Crosses were studied in ten other combinations using hybrids or colchicine induced autotetra ploids or amphiploids. The results are diagrammatically presented in the crossing diagram (Fig. 1). Fig. 1. Crossability and sterility barriers between 16 accessions of the Solanum nigrum complex. Accession numbers are as given in Table 1. S314n: Auto-tetraploid of race S31; H16 C. A.: Col chicine amphiploid of the hybrid S16 ~S33. Arrowhead points towards female parent in a cross. Arrow heads at both ends of a line indicate reciprocal crosses. ----- Cross failed completely. _??_ Hybrid highly sterile (0-6% good pollen). _??_ Hybrid semisterile (12 to 28% good pollen). _??_ Hybrid highly fertile.... Result uncertain (In the crosses S14 ~S144 and S33 ~S18 healthy seeds were obtained; but they were not germinated. In crosses S14 ~S35 and S18 ~S1l healthy fruits with seeds were set, but the plants died before the fruits were ripe.)
320 J. Venkateswarlu and M. Krishna Rao Cytologia 37 Though diploids could never be crossed with hexaploids, they could be crossed with each other or with the tetraploids fairly easily. Crosses between two tetraploids or two hexaploids readily produced hybrids. Hybrids between tetraploids and hexaploids could not be obtained. Hexaploids were produced by applying colchicine to triploid hybrids obtained by crossing diploids and tetraploids. Natural hexa ploids could be crossed to these artificially produced hexaploids quite successfully. In all cases of successful hybridization the resulting hybrids were healthy and vig orous and usually partly fertile, except the triploids which were completely male and female sterile. Two probable cases of cytoplasmic barrier to hybridization were recorded, one among the diploids and another among the tetraploids. Diploid S. nigrum, S 30 as female parent could not be crossed to the diploids races S 14 and S 31, but in one reciprocal cross attempted (S 14 ~S 30), hybrids were obtained. Among the tetraploids, S. nigrum of India S 33, could be readily crossed with S. villosum Dun. S 18, but its reciprocal cross failed. Fifteen flowers of the reciprocal cross produced a single fruit with empty shrivelled seeds. In all other cases results of reciprocal crosses were similar and reciprocal hybrids were indistinguishable. The occurrence of triploids in a cross between two diploids was recorded for the first time in the S. nigrum complex. S. nodiflorum Jacq. and S. americanum Mill. had regular meiosis and their Fl hybrid was partly pollen sterile and had slightly irregular meiosis. This hybrid, ameircanum ~nodiflorum, as maternal parent, was backcrossed to S. nodiflorum and in the progeny of fifteen plants three were found to be triploids. In view of the regular meiosis in the paternal parent and the morphological appearance of the triploids, it seems probable that the triploids had arisen by the fertilization of an unreduced egg of the maternal parent by the normal gamete of S. nodiflorum. Thus the triploids have two sets of S. nodiflorum and one set of S. americanum chromosomes. Crossability between diploids and tetraploids of S. nigrum of India was studied utilizing two diploid races, S 14 and S 31, and two tetraploid races, S 15 and S 33, in the following five combinations: i) S 14 ~S 33; ii) S 33 ~S 14; iii) S 14 ~S 15; iv) S 31 ~S 33 and v) S 31 ~S 15. Crosses were successful in the combinations ii and iv but failed in the other three, thus showing that the two diploid races exhibit a difference in their crossability to the same tetraploid. The cross between diploid and tetraploid was successful with S 14 as male parent whereas it was successful with S 31 as seed parent. Cytological examination revealed that the genomes of S 14 and S 31 seem to differ by an inversion. It is probable that the genomes of the two races also differ in genes which control crossability with tetraploids. Another significant point revealed by these crosses is the success of a cross when the lower ploidy parent is used as the seed parent (cross iv). Similarly crossability relation ships were studied utilising one diploid Indian race, S 14, one European diploid, S. nodiflorum S 16, and two European tetraploids, S. villosum, S 18 and S. luteum Mill. S, 6, in the following six combinations: i) S 16Z ~S 18; ii) S 18 ~S 16; iii) S 16 ~ S 6; iv) S 6 ~S 16; v) S 18 ~S 14 and vi) S 6 ~S 14. Triploid hybrids were obtained in crosses i to iv but there was no seed set in the rest two crosses, revealing cros sability relationships similar to those described above. In this case also, cytological
1972 Breeding System, Crossability Relationships and Isolating Mechanisms 321 examination of hybrids between the two diploids showed that their genomes differ by one translocation and some additional small structural alterations as revealed by several unpaired segments at pachytene. Crossability between diploids and hexaploids was studied utilising two hexa ploids, S. nigrum of India, S 11, and S. memphiticum Mart. S 26, and three diploids, S. nigrum of India, S 14 and S 31, and S. nodiflorum, S 16, in the following six com combinations: i) S I 1 ~S 14; ii) S 14 ~S 11; iii) S 31 ~S 11; iv) S 11 ~S 31; v) S 11 ~S 16; and vi) S 26 ~S 14. None of the crosses were successful. The cross Indian hexaploid ~Indian diploid (cross i) was studied in some detail to under stand the nature of crossability barrier between the two sympatric races. One hundred and twenty flowers of the hexaploid were pollinated with pollen from diploid. All fruits resulting in this cross resembled those formed after self-polli nation of the hexaploid except for their somewhat samller size and lack of seeds. A high per cent (80 to 90%) of pollen grains of the diploid were found to germinate on the hexaploid stigmas within twenty four hours after pollination. Pollen tubes were observed to descend down the stylar tissue and a few of them were observed at the base of the style also. Globular stage of the embryo was observed in the stain ed microtome sections of the hexaploid S. nigrum four days after pollination. Material collected eight days after pollination contained a little larger but only a globular embryo. In these ovules endosperm also was developed. Therefore, it could be presumed that fertilization took place and failure of embryo differentia tion and seed abortion occurred at a stage later than the globular embryo stage. Reversal of crossability: In diploid S. nigrum S 31, one plant treated with colchicine had diploid and tetraploid branches. Flowers of the diploid branches were pollinated with pollen of S 33 (Indian natural tetraploid) and all the result ing seeds produced triploid plants on germination, thus showing easy crossability between them. Flowers on the tetraploid branches of the same plant were also pollinated with pollen from S 33 and its reciprocal cross was also made; both these crosses, however, failed revealing that the autotetraploid of S 31 could not be successfully crossed to the natural tetraploid either as seed or pollen parent. Eventhough the same diploid, S 31, could not be crossed to the natural hexa ploid S. nigrum of India, S 11, either as male or female parent, autotetraploid of S 31 crossed with S I 1 as pollen parent readily producing pentaploid hybrids. These results clearly demonstrate that autopolyploidy reverses the crossability of relationships in the S. nigrum complex. Diploid S. nigrum, as seed parent, crosses with natural tetraploid but fails to hybridize with natural hexaploid, while its autotetraploid behaves in the reverse way. Isolating mechanisms: The various isolating mechanisms observed in the pre sent study are designated hybrid inviability, hybrid sterility and hybrid breakdown following Stebbins (1958). The species and races isolated by the different mechani sms are listed below with brief descriptions. Hybrid inviability: The hybrid zygotes are inviable or adaptively inferior to those of the parental species. S. nigrum diploid and S. nigrum hexaploid of India are isolated by this mechanism. Hybrid sterility: Failure of hybrids to produce a normal complement of
322 J. Venkateswarlu and M. Krishna Rao Cytologia 37 functional sex cells. i) S. nigrum races, diploids, S 14 and S 30; and S 14 and S 31; ii) S. nigrum diploid, S 14, and S. nodiflorum; iii) S. nigrum diploid, S 14 and S. americanum; iv) S. nigrum diploid, S 14 and S. gracile Dun.; v) S. nigrum diploid, S 14 and S. nigrum, S 33; vi) S. nodiflorum and S. villosum; vii) S. nigrum hexaploid races, S 11 and S 8, and S 11 and S 19; and viii) S. nigrum hexaploid races, S 11 and S 8 and S. memphiticum. In these various combinations partly or completely pollen sterile hybrids were produced and hence the parents of the various hybrids are to be regarded as isolated by hybrid sterility. Of special interest here are the ob servations on combinations v and vi of this list. Taking combination v, S. nigrum Indian diploid S 14 ~S. nigrum Indian tetraploid S 33, the F, hybrids were totally sterile triploids. Amphiploids, produced by application of colchicine to these triploids, were fertile but were partly pollen sterile. C, generation was raised from selfed seed of C, hexaploids; of the 25 seeds sown only six germinated and three alone survived to maturity. One of these was totally male and female sterile. This failed to set fruit either on selfing or on crossing to the natural hexaploids. The other two were partly pollen sterile, but self-fertile. This indicates sterility of the triploid hybrids is due in part to disharmonious gene combination and this genic sterility is passed on to the hexaploids also on chromosome doubling. These genes segregate on selfing of the hexaploids producing fertile and totally sterile plants. In combination vi, the F, hybrids of the cross S. nodiflorum ~S. villosum, were also totally male and female sterile triploids. In addition, in these hybrids the stamens were narrow and petaloid with indistinctly developed anther lobes containing few pollen mother cells. On colchicine treatment hexaploids were obtained, but com plete fertility was not restored. Stamens were better developed in the hexaploid compared to the triploid but were still deformed. Anthers being indehiscent, the plants were male sterile. Pollen sterility was high, but plants were female fertile producing fruits with viable seeds on crossing to any of the natural hexaploids. The resulting hybrids were also partly sterile and the petaloid stamens were present in some of the plants. Furthermore, certain hybrids exhibit chromosome instability leading to chromosome mosaicism (Venkateswarlu and Krishna Rao 1969). Hybrid breakdown: Hybrid breakdown expresses itself as inviability or adap tive inferiority of all of part of the F2 or backcross hybrids. i) S. nodi forum and S. americanum are to be considered as isolated by this mechanism. F, hybrids produced by reciprocal crosses were healthy and vigorous with 70 to 80% pollen sterility. Fifteen plants were observed in the F2 generation of each of the reciprocal hybrids. In the F2 of S. nodiflorum ~S. americanum one plant was stunted and weak; flowers were very small on this plant with dwarf petals and stamens. This failed to set any fruit eventhough its sibs were highly fertile. It died rather very early compared to its sibs. In the F2 generation of the reciprocal cross one plant had high pollen sterility, almost approaching 100%. This was found to be desy naptic with a mean frequency, of 13.8 univalents per cell at metaphase I. In the progeny of the backcross, (S. americanum ~S. nodiflorum) ~S. nodiflorum, three of the 15 plants were triploids with total pollen and seed sterility. These two species are, threfore, isolated not only by hybrid sterility but also by hybrid breakdown. ii) S. nigrum S 11 ~S. memphiticum; F, hybrids were healthy and vigorous and partly
1972 Breeding System, Crossability Relationships and Isolating Mechanisms 323 sterile. In Fa generation 20 plants were raised from each of the reciprocal hybrids. Seven plants in the total 40 were weak and stunted in growth with deformed leaves and flowers. They died early without setting fruit. Genetic barriers: Discussion One of the obvious isolating mechanisms which restricts gene exchange in the Solanum nigrum complex is autogamy. Geographical isola tion is operative in races which are allopatric. During the present study in the sympatric as also in the allopatric races and species, more specific types of genetic barriers have been uncovered viz., hybrid sterility, hybrid inviability and hybrid breakdown. Moreover it has been observed that even in closely related forms, the isolation of species or races is accomplished by often quite distinct means and the interbreeding of a given pair of species or races is, as a rule, prevented not by the operation of a single mechanism but by the operation several mechanisms at the same time. This appears to be the case with any group of closely related species (Dobzhansky 1951). Under the heading hybrid inviability are included all those mechanisms which prevent or retard the development of hybrids from the first division of the zygote to the final differentiation of the reproductive organs or gametes. The diploid and hexaploid races of S. nigrum of India, which are sympatric, are isolated by this mechanism. Stebbins (1958) proposed three immediate causes for hybrid invia bility: i) disharmony in the chromosomes and genes of the two parental species as they are combined in the hybrid nuclei; ii) disharmonious interaction between the chromosomes or genes of one species and the cytoplasm of the other; and iii) in hibition of the development of the hybrid embryo by the endosperm nourishing it. In the cross of the two races under discussion, endosperm formation and development of the hybrid embryo have been observed up to only the globular stage; therefore inhibition of the development of the hybrid embryo by the endo sperm nourishing it (cause iii above) seems to be the more probable cause of hybrid inviability. Hybrid sterility as an isolation mechanism has been encountered in the present study in many cases. Completely male and female sterile hybrids have been ob tained in the crosses between diploids and tetraploids. Hybrids with varying de grees of pollen sterility have been obtained in a number of crosses at all levels of ploidy. According to Stebbins (1958) the exact basic causes of hybrid sterility are only partly known and are probably numerous, but all may be considered as special examples of the same kind of genetic imbalance which is responsible for hybrid inviability. Sterility in the various hybrids in the present study seems to be the result of both disharmonious gene combination and chromosomal structural differences of the parents. In the various hybrids lack of structural homology between the parental chromosomes is indicated not only by reduced amount of pairing of chromosomes in the hybrids compared to the parents, but also by the occasional presence of rings or chains, indicating heterozygosity for translocations, and of bridge-fragment configurations suggesting heterozygosity for inversions.
324 J. Venkateswarlu and M. Krishna Rao Cytologia 37 In addition, in the hybrids S. nidiflorum and S. nigrum S 14 ~S. americanum, cryptic structural hybridity also seems to be responsible for sterility as can be inferred by the presence of differential unpaired segments at pachytene observed cytologically during this study of these hybrids. Diplontic sterility which is due to deformed stamens in the hybrid S. nodiflorum ~S. villosum which is passed on to the colchicine amphiploids and its hybrids with natural hexaploids, seems to be genic in origin. So also the sterility in the hybrid S. nigrum S 14 ~S. nigrum S 33 is in part genic; in this case chromosome doubling partly restores fertility but in the progeny of the amphiploids completely sterile segregants are found. In both these cases irregular meiosis due to lack of chromosome homology also contributes to hybrid sterility. S. nigrum Indian hexaploid and S. memphiticum among the hexaploids and S. nodiflorum and S. americanum among the diploids are isolated by hybrid break down as can be concluded from the occurrence of weak and completely sterile plants in the F2 progenies of the hybrids involving the species. Stebbins (1958) thought it likely that hybrid breakdown may be due to differences between the parents in chromosome structure as well as in gene content. In the present study, chromosome differentiation between the parents is revealed by meiotic irregularities in the hybrids which seems to be responsible for hybrid breakdown. Crossability: Tandon and Rao (1966a, b) had the success of a cross only when the higher chromosomal form was used as the pistillate parent, when the cross attempted involved races of different ploidy levels. In general this relation ship was found to hold good in our studies too; but it was also found that the use of a higher ploidy parent as female is not an essential criterion for a cross to be successful. Triploid hybrids were produced in the cross S. nigrum diploid S 31 as seed parent and S. nigrum natural tetraploid, S 33 as pollen parent. Similarly S. nodiflorum, a diploid, could be crossed as seed parent either to S. villosum, S 18 or S. luteum S 16, both of which are tetraploids. Thus all diploid races are not isolated from the tetraploid races to the same extent; some races seemed to have diverged enough as to be unable to cross as seed parents while others still remain at a lower level of differentiation retaining the ability to cross with tetraploids either as seed or pollen parents. Cytogenetic data obtained by us are consistent with the conclusion that the diploid and tetraploid races are closely related and diploids seem to have contributed some chromosomes to the genome of the tetraploids. Then the loss of crossability between the races even to the extent that the diploids cannot be pistil parents, is a decisive step in the evolution and divergence of the complex. However, this divergence appears to be just one of the mechanisms of the genetic isolation of the diploids and tetraploids and seems to be relatively of minor importance at the present stage of evolution of the S. nigrum complex. Hybrid sterility (cent percent pollen and ovule sterility) seems to be the major hurdle for gene exchange between the diploid and tetraploid races. European diploid and tetraploid seem to be isolated to a greater extent compared to the Indian races. In the former, triploid hybrids between them possess deformed petaloid stamens and this deformity is carried on to the hexaploids which are func tionally male sterile for the reason that their anthers do not dehisce and shed pollen. Such a deformity was not observed in the triploid hybrid between the Indian races.
1972 Breeding System, Crossability Relationships and Isolating Mechanisms 325 Reversal of crossability was observed during the present study concerning the diploid, tetraploid and hexaploid races of India. Diploid S. nigrum as female crosses with natural tetraploid but fails to hybridize with natural hexaploid, while its autotetraploid crosses with the natural hexaploid, but fails to hybridize with the natural tetraploid. This seems to be a phenomenon of rather common occurrence in the Solanum nigrum complex, since Westergaard (1958) observed that, while autotetraploid S. nodiflorum produces hybrids with hexaploid S. nigrum both in nature as well as when artificially pollinated with a mixture of its pollen (2n=24) and that of S. nigrum (n=36), the diploid never crosses with the hexaploid. While the mechanism is not understood, an analogy can perhaps be drawn to the behaviour of S alleles in Solanum. It seems probable that the genic balance present in the diploid is upset on chromosome doubling resulting in reversal of crossability. Acknowledgements The work was done during the tenure of a scheme to study the cytognetics of the Solanum nigrum complex. We are thankful to the Council of Scientific and Industrial Research, New Delhi, for financing the Scheme. Summary Breeding system, crossability relationships and isolation mechanisms have been studied in the Solanum nigrum L. complex, using 16 accessions including seven diploids, five tetraploids and four hexaploids. Of these six are Indian forms and special attention has been paid to them. It was found that the diploids could be crossed with each other and also with natural tetraploids fairly easily, but never with the hexaploids. Crosses between two tetraploids or two hexaploids readily produced hybrids. In general, in crosses involving two ploidy levels, a cross suc ceeds when the higher ploidy plant was crossed as the seed parent; but exceptions were found wherein the lower ploidy plant could be successfully used as the seed parent. In this respect all diploid races are not isolated from the tetraploid races to the same extent; some races were divergent enough to be crossed only as male parents while others could be crossed in either direction. Compared to this diver gence in crossability relationships hybrids sterility seems to be the major hurdle for gene exchange between the diploids and tetraploids. The Indian diploid and tetraploid races are isolated to a lesser extent as compared to the European diploid and tetraploid races. Autopolyploidy reverses crossability relationships. The diploid S. nigrum of India as seed parent crosses with natural tetraploid S. nigrum of India but fails to cross with natural hexaploid S. nigrum of India. However, the autotetraploid of diploid S. nigrum crosses with natural hexaploid and fails to hybridize with natural tetraploid. In addition to self-pollination and geographic isolation as factors restricting gene exchange between the different forms, hybrid sterility, hybrid breakdown and hybrid inviability were found to be operative in the complex. Diploid and hexaploid races of S. nigrum of India are isolated by hybrid
326 J. Venkateswarlu and M. Krishna Rao Cytologia 37 inviability, while the diploid and tetraploid races of S. nigrum of India are isolated by hybrid sterility. References Bhaduri, P. N. 1933. Chromosome numbers of some Solanaceous plants of Bengal. J. Indian bot. Soc. 12: 56-64.- 1945. Artificially raised auto-tetraploid S. nigrum L., and the species problem in the genus Solarium. Proc. Indian Sci. Cong. Section B: 77-78. - 1951. Interrelationships of non-tuberiferous species of Solanum with consideration on the origin of Brinjal (S. melongena L.). Indian J. Genet. 2: 75-82. Chennaveeraiah, M. S., and Patil, S. R. 1968. Some studies in the Solanum nigrum L. complex. Genet. Iberica 20: 23-36. Dobzhansky, Th. 1951. The Genetics of the Origin of Species. 3rd Edition. Columbia University Press, N. Y. Heiser, C. B., Soria, J., and Burton, D. L. 1965. A numerical taxonomic study of Solanum species and hybrids. Amer. Naturalist 99: 471-488. Soria, J. V., and Heiser, C. B. 1961. A statistical study of relationships of certain species of the Solanum nigrum complex. Econ. Bot. 15: 245-255. Stebbins, G. L. 1958. The inviability, weakness and sterility of interspecific hybrids. Adv. Genet. 9: 147-215.- and Paddock, E. F. 1949. The Solanum nigrum complex in Pacific North America. Madrono, 10: 70-81. Tandon, S. L. and Rao, G. R. 1964. Cytogenetical investigations in relation to the mechanism of evolution in hexaploid Solanum nigrum L. Nature (Lond.) 201: 1348-1349.- and - 1966a. Interrelationships in the Solanum nigrum complex. Indian J.Genet. 26: 130-141.- and - 1966b. Genome analysis in Solanum nigrum L. J. Cytol. Genet. 1: 41-45. Venkateswarlu, J. and Krishna Rao, M. 1969. Chromosome numerical mosaicism in some hybrids of the Solanum nigrum complex. Genetica 40: 400-406. Westergaard, M. 1958. In the discussion on a paper by Stebbins, G. L. "Longevity, habitat and release of genetic variability in higher plants". Cold Spring Harbor Symposia Quant. Biol. 23.