North American Lily Society Yearbook. 1975, p. 78.
Interspecific Incompatibility and the Improbable Cross
R. M. Withers
Hawthorn, Victoria, Australia

In this article it is proposed to survey attempts to achieve success in the production of difficult lily hybrids, to outline briefly possible reasons for interspecific incompatibility, and finally to recommend methods by which such crosses can be made with every hope that they will be successful. It is assumed that the seed and pollen parents are of similar chromosome status, and genetically capable of being crossed. The use of colchicine to create polyploid lilies and the make sterile lilies fertile, is quite a separate subject.

The germination of lily pollen on a stigma, the production of the pollen tube and its long journey from the stigma through the style to the ovules in the ovary, and subsequent fertilization, is fully described by Mrs. Sophie C. Ducker in an article entitles "The Life History of the Genus Lilium."1

The lily literature contains many references to attempts being made to accomplish difficult crosses, but it is not very often that successful results are reported. Some successful crosses have taken place by pure chance. Little had been written about the reasons for success or failure, and also little has been written of methods used in making difficult crosses. Less still has been written about methods which may be used in the hope of achieving success.

L. candidum and L. chalcedonicum have always been difficult to cross. The first hybrid between these was L. x testaceum, which probably appeared as a chance hybrid seedling in a field of L. candidum in Germany or Holland about 1810. The cross has been repeated since. In 1932 Jones raised L. x testaceum Jones variety.

For many years L. auratum and L. speciosum were difficult to cross, and it was not until the 1930's when Jillian Wallace and L. x parkmannii Beecher's Variety were raised in Australia, that hybrids between these two species appeared in any number. Later, in both the United States and in New Zealand, L. rubellum and L. japonicum were introduced into the Parkman Hybrid group or Oriental Hybrid group as it then became known; and more recently, L. alexandrae and L. nobilissimum have been introduced into the group.

L. Black Beauty was raised by Mr. Leslie Woodriff in the United States in 1957 as a chance seedling in a field of lilies. It was later found to be a hybrid between L. speciosum rubrum and L. henryi. Attempts to recreate this hybrid have proved difficult.

*This hybrid has flowered in Oregon and appeared to be simply a good L. longiflorum. (Editor's note)
Dr. Robert Withers is a hybrid that has recently created a lot of interest in Australia, being the winner of a number of Australian Lilium Society awards in 1974. It was grown from seed originating from a cross made by Mr. Fritz Seifert in Germany of L. longiflorum pollinated by the red banded L. auratum.* It was a controlled cross, but interference from insects could not be excluded. The flowers are L. longiflorum in type. It will be of interest to see if it is receptive to either L. longiflorum or Oriental Hybrid pollen.

More recently, reports have been received from Japan of the successful crossing of L. longiflorum and L. auratum. This cross will be released for distribution after it has been registered.

In past years, various methods have been tried in attempts to achieve success with difficult crosses. It has been suggested that the stigma of the seed parent should be smothered with pollen from the intended incompatible pollen parent, and the process repeated daily for several days, in the hope that fertilization of a few ovules may occur in the ovary with the resultant production of a few seeds.

In the Netherlands, Y. O. Kho and J. Baer carried out a microscopic investigation to find the causes of the failure of interspecific crosses between scaled and unscaled rhododendron species. It was found that pollen germination on the stigma, and pollen tube growth in the style proceed normally, but that difficulties occur when the pollen tubes reach the ovary. A small number of pollen tubes enter an ovule, but only in a few cases is seed setting effected (2-4 seeds per fruit). The majority of the tubes do not reach an ovule; the tubes often have thickened tips or, when they continue growth, they start to coil without penetrating an ovule.2 Maybe a similar position exists in some cases of interspecific incompatibility in lilies.

In past years some lily hybridizers have used naphthalene acetamide which was applied to the ovary in the hope that the seed parent would become more receptive to incompatible pollen. It does not appear that much success has been achieved by this method.

In Australia, several lily hybridizers have maintained that they have had more success by applying mixed pollen to the stigma, the theory being that by using mixed pollen, some of the pollen excites the stigma and makes it more receptive to other pollen. However, using mixed pollen in this way, it has not been demonstrated that an improbable cross has been obtained. Certainly, a seed capsule full of good viable seed can be obtained by this method, but it is more likely that the full production of seed has resulted from the maximum use being made of the available compatible pollen in the mixture. The first lily hybrid registered with the Registrar of the Australian Lilium Society was Rose Hall, raised by the late Mr. Gilbert Errey. It is recorded that it was bred from a L. davidii seedling pollinated with mixed pollen of L. cernuum and L. rubellum. The L. rubellum pollen had no apparent effect in the cross.

In recent years important research work has been progressing at the Australian National University in Canberra on interspecific incompatibility, mainly in relation to poplars but to a lesser extent to eucalypts and petunias.

R. B. Knox, R. R. Willing and Anne E. Ashford found that "proteins in pollen grain walls are essential for germination in poplars: they also affect the incompatibility systems on which pollen tube growth is dependent. One reason for the failure of interspecific hybridization in flowering plants is that pollen from one species will not germinate on a foreign stigma, even when the second species appears to be closely related. This is why Michurin and his successors in the U.S.S.R. tried to overcome this obstacle to hybridization by mixing incompatible with compatible (mentor) pollen, in the hope that the latter would provide the necessary stimulus for germination and tube growth. Although the first attempts were not conspicuously successful, Stettler has more recently obtained hybrids between incompatible species of poplar by using mentor pollen killed by gamma radiation. We have been able to repeat Stettler's experiments with other species of poplar and now demonstrate that the function of the killed compatible pollen is to provide what we have previously called recognition substances. These constitute a large part of the protein located in the inner cellulose layer of the pollen grain wall."3

In another paper, R. B. Knox, R. R. Willing and L. D. Pryor describe the recognition pollen method which enabled successful hybridization of the black popular, Populus deltoides, and the white poplar, P. alba. In these species, hybridization is normally prevented by incompatibility barriers. For use, the recognition pollen (killed pollen of the compatible type produced by freeze-thawing, gamma irradiation or by treatment with methanol or ether) is mixed with fresh incompatible pollen and the mix dusted on the stigmas. All the progeny are hybrids. The determinative role played by pollen grain recognition substances in the control of interspecific incompatibility reactions on the stigma is demonstrated by these experiments. The recognition pollen method appears to have widespread applications in plant improvement.4

Similar works is also being carried out at the Royal Botanic Gardens, Kew. J. Heslop-Harrison, R. B. Knox and Y. Heslop-Harrison have shown that during maturation of the pollen in species of Iberis (Cruciferae), materials from the anther tapetum are transferred to cavities of the exine. These sporophytically-derived materials are responsible for the rejection response induced in the stigmatic papillae within 4 to 6 hours following incompatible pollination, a response which leads to the occlusion of the stigma papillae by the polysaccharide, callose. Callose occlusion has also been found to extend into the pollen tubes.5

With this knowledge of a mechanism of incompatibility, although it has mainly involved research work in poplars, one is able to proceed with similar work in lilies, and adopt methods which have every hope of overcoming incompatibility in breeding.

Mr R. R. Willing, working with poplars, eucalypts, petunias, and crosses between evergreen rhododendrons and deciduous azaleas, has found that both the pollen and the stigma have two components concerned with compatible and incompatible crosses.

In a compatible cross, the pollen germinates on the stigma, a pollen tube is produced, and fertilization and the production of viable seed follows.

In the incompatible cross, the pollen wall protein combines with a component of the stigma to form callose which obliterates the pollen tube and prevents fertilization. He has found that removal of one or other of these two factors prevents the formation of callose and makes the cross compatible.

In making crosses between incompatible parents with copious pollen, the use of mixed pollen has advantages. The compatible pollen is sterilised with X-rays, gamma rays using a dose of 100,000 Rad's; or by soaking the pollen in ethanol or methanol for 4 minutes, warming with a blower or hairdrier until the pollen is dry, when it is ready for use. It has been found that 95% of the sterilised compatible pollen mixed with 5% of incompatible pollen applied to the stigma has resulted in up to 100% seed production. The sterilised compatible pollen has stopped the formation of callose or reaction body in the incompatible pollen by a reaction between their pollen wall proteins with the consequent removal of one of the factors essential for the production of callose.

In making crosses between incompatible parents with less copious pollen the use of mixed pollen is not necessary. All that need be done is to remove one or other of the factors concerned in the production of callose and the cross will become compatible.

It is easy to remove the pollen factor. The incompatible pollen should be washed with hexane or ethyl acetate but not the alcohols. Wash with the fluid until the supernatant fluid is clear, then dry gently, and the pollen is ready for use.

To remove the stigma factor is the simplest method of all, and the best to use. Over 30 solvents have been tried on the stigma, and all work with the exception of the alcohols. Chloroform and ether have a burning effect on the stigma surface. Hexane has the least detrimental effect and is the best solvent to use. A camel hair brush should be dipped in the hexane, the excess wiped off on a piece of wood, and then it should be brushed a few times on the stigma. The stigma must not be wet with the solvent, the camel hair brush being barely damp. The probable mode of action is to remove or inactivate a waxy-type surface on the stigma.

Two other techniques are worthy of mention and trial. Dr. Peter Valder of Sydney University has successfully crossed evergreen with deciduous rhododendrons by cutting off the style of the seed parent just in front of the ovary and then applying pollen from the pollen parent to the cut surface of the style. The pollen has germinated, and has had a much shorter journey to the ovules in the ovary, and successful fertilization and seed production has taken place.

Finally, stigma grafts may overcome interspecific incompatibility in lilies, having as they do such large stigmas which are easy to operate upon. A thin section is excised from the surface of the stigma of the seed parent, and a section or graft of corresponding thickness and shape is cut from the surface of the stigma of the incompatible pollen parent and applied to the cut surface of the stigma of the seed parent. Germination of the pollen should then take place on its own stigma surface, pollen tubes should be produced without the formation of callose and fertilization and seed production should follow. The graft has only to live for about 2 hours to give time for the germination of the pollen and production of pollen tubes. Precautions should be taken that the instruments used in the operation are as clean as possible. This method has proved to be successful in crossing petunias.

Techniques for making incompatible crosses have been outlined; and although little work of this nature has been carried out with lilies, it is hoped that breeders will use the techniques to create new and more interesting lilies.

Grateful thanks are due to Professor Bruce Knox and Mr. Rudolf Willing for their generous help and advice.


  1. Sophie C. Ducker. "The Life History of the Genus Lilium." Page 17, "Liliums in Australia" by R. M. Withers.
  2. Y. O. Kho and J. Baer. "A Microscopical Research on the Incompatibility in the Cross Rhododendron impeditum X Rhododendron williamsianum" Euphytica 19 (1970), 303-309.
  3. R. B. Knox, R. R. Willing and Anne E. Ashford. "Role of Pollen-Wall Proteins as Recognition Substances in Interspecific Incompatibility in Poplars." Nature. Vol. 237, No. 5355, pp. 381-383, June 16, 1972.
  4. R. B. Knox, R. R. Willing and L. D. Pryor. "Interspecific Hybridization in Poplars using Recognition Pollen." Silvae Genetica 21, Heft 3-4, 65-148, 1972.
  5. J. Heslop-Harrison, R. B. Knox and Y. Heslop-Harrison, "Pollen-wall Proteins: Exine-held fractions associated with the Incompatibility Response in Cruciferae." Theoretical and Applied Genetics 44, 133-137 (1974).