Factors Affecting Fertility in Daylilies
Joseph C. Halinar
The Daylily Journal, vol 43, No. 4, 1988, pages 401-405

Fertility problems can be frustrating to the daylily hybridizer. Certain desirable parents will sometimes be completely infertile, producing very few seeds, or certain specific crosses may not produce seeds even though both parents are fertile. Particular culture practices or breeding methods can be used to reduce some of these fertility problems. An understanding of the various factors that can affect fertility will help hybridizers to understand the specific problems they face and possible help overcome, reduce, or circumvent the fertility problems.

The various fertility problems can be generally classified as either genetic, chromosomal or cytoplasmic. Both diploids and tetraploids are affected by these various factors, but the tetraploids are probably more sensitive to fertility problems because of the greater complexity associated with the double set of chromosomes.

The most important genetic factors affecting fertility are self-incompatibility, cross-incompatibility and male sterility.


Daylilies have a gametophytic self-incompatibility system. The incompatibility reaction is initiated by the genotype of the microgametophyte (pollen) interacting in the style with the genotype of the pod parent. Self-incompatibility genes are usually represented by a lower case "s" with subscripts to indicate the various alleles at the self-incompatibility locus.

A pod parent of genotype s1s2 will inhibit the growth of pollen tubes carrying either the s1 or s2 alleles, and no seed will be produced. Thus, a cultivar of genotype s1s2 will not self pollinate or cross with another cultivar with genotype s1s2. A cultivar of genotype s1s3 will produce pollen grains of genotypes s1 and s3. Thus, in a cross of s1s2 x s1s3, the s3 pollen grains of the pollen parent will grow and fertilize the egg cells, but not the s1 pollen. The resulting progenies will be of genotypes s1s3 and s2s3. In a cross of s1s2 x s3s4, the genotypes of the progenies will be s1s2, s1s4, s2s3 and s2s4.

There are large numbers of different self-incompatibility alleles in natural populations, even in small populations. The number of different self-incompatibility alleles in the modern daylily germplasm pool is unknown. The fewer the number of different alleles, the greater the chance of fertility problems because the greater the change that different cultivars may have the same self-incompatibility genotype. The modern daylily germplasm pool is based on only a small number of species and probably has a limited number of different clones of each species. Thus, we can speculate that there is a relatively low number of self-incompatibility alleles in the modern daylily germplasm pool.


Cross-incompatibility occurs when pollen of one species fails to germinate on the stigma or grow in the style of another species. The mechanism by which the cross-incompatibility reaction occurs is not well understood, but does involve a very complex genetic system. Many daylily species do not cross or do so reluctantly. Modern daylilies are a complex mixture of many species and can be expected to be a complex mixture of the various components of the cross-incompatibility system. Thus, a pod parent may contain enough of the cross-incompatibility system to recognize a particular pollen as foreign and, hence, initiate the cross-incompatibility reaction; a different pollen may not be recognized as foreign because it lacks the necessary recognization component while a third pollen might be inhibited but not enough to completely prevent a few seeds from being produced. This extreme mixing and segregation of the cross-incompatibility system from many different species can help explain much of the inconsistency in fertility patterns encountered in hybridizing.


Male or pollen sterility, both genetic and cytoplasmic-genetic, is well know in the plant world. Cytoplasmic male sterility is very important in the commercial production of F1 hybrid seed. It involves the interaction of a specific genotype with a specific cytoplasm. Currently, there does not appear to be any good examples of cytoplasmic male sterility in daylilies.

Genetic male sterility depends only on the presence of a specific genotype, usually a recessive gene, and does not depend on the cytoplasm of the plant. The development of a functional pollen grain depends on a large number of genes working properly. The malfunction of any one gene can result in a non-functional pollen grain or a pollen with a low level of viability. Also, any gene that affects the morphological development of the anther can result in functional male sterility because of the lack of an anther or the development of a malformed anther.

The presence of female or pod sterility genes is less well understood or documented because of the difficulty of studying the female gametophyte compared to the male side. Their presence would help explain pod sterile but pollen fertile plants. Currently, there is no good evidence to support the presence of any specific female sterility genes in the daylily germplasm pool. However, since male sterility genes exist, it seems reasonable to surmise that they can also exist on the female side. A detailed genetic study could detect their presence but would require a complex, time consuming study.

A problem can also arise when there is a large discrepancy in the style lengths of the parents. Some of the daylily species have short styles while others have very long styles. The genes that control pollen tube length may have segregated independently of style length with the result that some modern hybrids may produce pollen that can not grow long enough to reach the egg cell, even though they have average to long styles.

Chromosomal factors probably account for the vast majority of the fertility problems in daylilies. Three main causes affecting fertility are genome incompatibility, aneuploidy and chromosomal abnormalities.


Each haploid set of chromosomes contributed by each parent is referred to as a genome. Genomes are generally represented by a capital letter, often the first letter of the species name if the species source of the cytoplasm is known. For example, Hemerocallis fulva could be represented by the symbol FF, with each genome representing 11 chromosomes. H. minor could be represented by the symbol MM and a H. fulva x H. minor hybrid would then be genome FM.

Different genomes are brought together when species are crossed. If these genomes are closely related, they will pair during meiosis and be fertile. If the two genomes are not closely related, then the chromosomes of each genome will not pair during meiosis, no functional gametes will be produced and, hence, the hybrids will be infertile. There can be any degree of pairing from complete pairing to non-pairing, and, hence, a whole range of fertility from fully fertile to completely infertile.

Genome incompatibility becomes much more complex at the tetraploid level. Basically, tetraploids can be classified as either auto- or allotetraploids. If we take a fertile species or hybrid and double it to the tetraploid level, we then have four similar genomes pairing during meiosis. The four chromosomes pairing during meiosis can not properly separate and, thus, generally do not produce viable gametes. Autotetraploids are generally infertile or have low fertility.

If we cross two species or hybrids that have dissimilar genomes, then the chromosomes of each genome will not pair and the hybrid will be infertile. However, if we double the number of chromosomes to the tetraploid level, each chromosome of each genome will now have a similar chromosome to pair with. If we look at these tetraploids during meiosis we would see 22 pairs of chromosomes as if they were a diploid, and, thus, fertile. These are generally referred to as allotetraploids. Even though these allotetraploids have a double set of chromosomes, they look like diploids during meiosis and are, thus, also referred to as amphidiploids.

Modern day tetraploids are a complex mixture of genomes, and often exhibit a mixture of autotetraploid and allotetraploid behavior.


Aneuploidy occurs when there is an unbalanced chromosome structure that is not a full multiple of the basic chromosome number (11 in daylilies). The most common form of aneuploidy is either one less chromosome (21) or one more chromosome (23) than normal. There are many other complex forms of aneuploidy. The unbalanced chromosome structure generally results in a sever loss of fertility, although not always. The formation of pollen is much more sensitive to aneuploidy than the female side. Thus, aneuploids are often more fertile as pod parents than as pollen parents.


Chromosome abnormalities are minor changes in a plant's chromosome structure. These include deletions, the loss of a small piece of chromosome; duplications, the addition of a small piece of chromosome; inversions, the interchange of gene order and translocations, the exchange of a piece of one chromosome to another chromosome. All these chromosome abnormalities affect fertility to various degrees, from slight to severe reduction in fertility.


Cytoplasm has its own identity and each species is considered to have its own separate cytoplasm. There is a delicate balance between the cytoplasm and the genome. A disruption in this balance can result in embryo or endosperm abortion.

Cytoplasm is inherited as a maternal trait. The pollen contributes a haploid set of chromosomes but no cytoplasm. This unidirectional inheritance of cytoplasm can change the cytoplasm-genome relationship. In the cross A x B, the resulting hybrids have genome AB in A cytoplasm while the reciprocal cross, B x A, has the same genome but now in B cytoplasm. By repeatedly backcrossing, it is possible to transfer the genome of one species into the cytoplasm of another species. For example, ((A x B) x B) results in progenies that have 75% of the B genome but in A cytoplasm. Five generations of backcrossing will result in progenies with 97% of the B genome in the A cytoplasm. Genomes in alien cytoplasm may exhibit a range of behavior from normal to non-viable; fertility problems are not uncommon.

Modern daylily hybrids are a complex of different genomes. Particular genome combinations may not be compatible with their cytoplasm and may result in embryo or endosperm abortion.

The endosperm provides nutrients for the growth of the embryo. Failure of the endosperm to develop will result in embryo abortion. Endosperm is sensitive to the cytoplasmic-genome interaction that is the result of unidirectional inheritance of cytoplasm, but is also sensitive to the genome composition of the endosperm. The endosperm is formed by the fusion of two special nuclei in the female gametophyte with a nucleus from the pollen. A triploid endosperm is the most common but many of the genera of the lily family have a pentaploid endosperm. For the cross A x B a triploid endosperm has genome AAB, a pentaploid endosperm has genome AAAAB; the cross B x A will have endosperms BBA and BBBBA, respectively. The different genome constitution of the endosperm with possible cytoplasmic interaction, resulting from reciprocal crosses could result in endosperm abortion in one direction but not in the other direction.

Diagnosis of particular fertility problems is not easy. Sometimes it is possible to rule out certain causes or to make educated guesses such that it may be possible to alleviate or circumvent the fertility problem. Diagnosing and getting around fertility problems will be discussed in a future article.


Notes: This copy of this article is slightly different from the published manuscript. There were a few minor grammatical errors that I corrected. I also changed the word "sterile" to infertile in a few cases. Often the word "sterile" is used in place of infertile when the more correct word is infertile.

When discussing fertility problems it is always considered that the difference of the length of the styles in parents that have a very large difference in style length could be a factor in affecting fertility. It is often suggested to use the pollen from a long styled parent on the short style rather than the other way around. In daylilies there are plants with very short styles and others with very long styles. I have made crosses in both directions using parents with different style lengths and have not found any problems. Differences in style length can be a factor in some cases, but I don't think this is a serious problem in daylilies. Still, if you are having a problem with pollinating a long style plant, such as a spider, with a plant with a short style, try using the short style plant as a pod parent.