Chapter 8: Broadening the Genetic Basis

8.1 Induced mutations

The variation that is displayed between all living plants and animals, including all crop species, is the result of natural mutations at the DNA level, with subsequent recombination and selection occurring, much of it over millions of years. But this has also been accompanied by changes at a structural level of the genetic material, such as rearrangement within and between chromosomes.

Mutations result in the generation of additional genetic variation within plant species. It has been estimated that mutations occur naturally with a frequency of 1 in 1,000,000 (one in a million, $c08-math-0001$ ) per generation per locus. Most of these mutations are recessive and deleterious and so these new alleles usually do not survive at anything other than a very low frequency in nature. However, if the mutation results in an advantageous effect, the genotype possessing the mutation may thus be more adapted to the environment compared with the non-mutant types, and hence will tend to leave more offspring. Over generations this therefore leads to an increase in the frequency of the new allele within the population. Unlike other more recent developments described in this chapter, mutations have been used explicitly as an additional source of genetic variation during the last 90 years or so.

Obviously, the extremely low rate of natural mutation and even lower frequency of desirable mutation events are such that natural mutation has had little impact on modern plant breeding strategies. However, natural occurring mutations are still utilized and commercialized in clonal crops like apple. One such example are “spur” apple cultivars developed from previous existing varieties, which appear as spontaneous mutations in the field and which can be propagated by apple breeders. Another, more famous example would be the potato cultivar ‘Russet Burbank’.

In the mid-1920s it was discovered that X-rays could be used to induce high mutation rates, first shown in the fruit fly and later in barley. Plant breeders were quick to realize the potential of induced mutation and, with the inevitable over-enthusiasm for a new idea, mutation breeding became common practice in almost all crop species and in many ornamental flower breeding programmes.

The aim of mutation breeding is to stimulate an increase in the frequency of mutation events within a crop species and then to select desirable new alleles from amongst the mutants produced. More basically, mutation breeding has been utilized to make minor, advantageous genetic changes in already established and adapted cultivars through induced mutation treatments, for example, by inducing mutations in a highly adapted crop cultivar and then screening the resulting mutated lines for a specific character of interest. By doing this it is hoped to retain the existing cultivar's adaptability while adding the mutated advantageous trait.

Mutagenesis derived lines in a plant breeding scheme are labelled according to the number of generations after mutagenesis has taken place. For example, the mutated generation is called $c08-math-0002$ , while the generation immediately after the mutation treatment is termed the $c08-math-0003$ . In appropriate species, these plants can be self-pollinated to produce an $c08-math-0004$ generation, and so on (in line with $c08-math-0005$ etc., for the more usual sexual generations).

8.1.1 Methods of increasing the frequency of mutation

In general terms there are two methods that have been used to produce an increased frequency of mutations in plant species: radiation and chemical induction, with the highest frequency of mutation derived cultivars being from radiation-induced mutants.

Mutations following exposure to radiation are produced by a variety of effects from physical damage through to disturbing chemical bonds. Two main types of radiation have been utilized to induce mutation in crop breeding schemes.

Gamma rays are the most favoured radiation source in plant breeding and have been used to develop 64% of the radiation-induced mutant derived cultivars. Gamma rays are electromagnetic radiation with a high energy level and are produced by the disintegration of radioisotopes. The two main sources of gamma radiation for induced mutation are from cobalt-60 and caesium-137. Plant breeding programmes have used gamma ray radiation treatments applied in a single dose, or have treated whole plants with long exposure to gamma radiation.
X-rays were the original radiation mutagen, yet have been responsible for only 22% of the cultivars released worldwide from mutation-induced breeding programmes. X-rays are produced when high-speed electrons strike a metallic target. X-rays are high-energy ionizing radiation which have wavelengths ranging from ultraviolet to gamma radiation. Mutations are induced by exposing seeds, whole plants, plant organs, or plant parts to X-ray radiation of a required frequency for a specific time. X-rays have to be handled by trained radiologists, and so they are not always easily accessible to plant breeders, who often have to rely upon medical facilities (e.g. hospitals) for mutagenic treatment of plants.

Other forms of radiation that have been used to induce mutation include neutrons (an electrically neutral elementary nuclear particle produced by nuclear fission by uranium-235 in an atomic reactor), beta radiation (negatively charged particles that are emitted from radioisotopes such as phosphorus-32 and carbon-14) and ultraviolet radiation (used primarily for inducing mutations in pollen grains).

Many of the mutagenic chemicals are alkylating agents that bind to cellular DNA and interfere with chromosome division, such as: sulphur mustards; nitrogen mustards; epoxides; ethylene-imines; sulphates and sulphones; diazoalkanes; and nitroso-compounds. The most commonly used chemical mutagens have been ethyl-methane-sulphonate (EMS) and ethylene-imine (EI). It should be noted that all mutagenic chemicals are highly toxic, as well as being highly carcinogenic, so at all times they must be exclusively handed by qualified personnel in appropriate facilities.

The frequency of mutation can also be influenced by the oxygen level in plant material (higher oxygen related to increased plant injury and chromosomal abnormalities), water content (also related to oxygen content) and temperature.

8.1.2 Types of mutation

Mutations can be conveniently classified into four types:

Genome mutations where there are changes in chromosome number due to either addition or loss of whole chromosomes or sets of chromosomes.
Structural changes in the chromosomes involving translocation (a chromosomal aberration involving an interchange between different non-homologous chromosomes), inversions (changes in the arrangement of the loci, but not in their number), deficiencies, deletions, duplications and fusions (reduction or increase in the number of loci borne by the chromosome).
Gene mutations, often termed point mutations, where the change is in a single gene, and often the result of a single base-pair change at the DNA level. Because of the redundant nature of the genetic code, some point mutations do not affect the amino acid encoded by the triplet of nucleotides where the mutation took place. These are silent mutations.
Extranuclear mutations where the mutational event occurs in one of the cytoplasmic organelles. The DNA involved includes plastids and mitochondria and means that the mutation will usually be transmitted from one generation to the next through just one of the sexes, usually via the egg cells (i.e. maternally inherited). One example of this form of mutation is cytoplasmic male sterility, common in many crop species.

8.1.3 Plant parts to be treated

Mutagenic agents can be applied to different parts of plants and still produce effects. Seeds are the most obvious vehicle for treatment, and can be treated with either chemical or radiation mutagens. Seed have been preferred by many breeders because seeds are more tolerant to a wide range of physical conditions, such as being desiccated, soaked in liquid, heated, frozen or maintained under varying oxygen levels, which allows their exposure to the various mutagens to be rather easily carried out. Seed treated with an induced mutagen will result in plants that are:

Non-mutants, the same as the parent plant.
True mutants, where a mutation event has occurred throughout the whole plant.
Chimera, where only a portion of the resulting plant has been mutated.

Another attractive possibility in mutation breeding is to treat pollen grains with radiation or chemical mutagens. The major advantage of treating pollen grains is that they are easily collected in large numbers and can easily be presented to a radiation source. Pollen grains are effectively single cells, so induced mutation of pollen avoids the occurrence of chimeras. Pollen grains are also generally haploid, in terms of their genetic composition, and so this opens the possibility, in an increasing number of species, of using tissue culture treatment to lead to their direct development into plantlets – which, with suitable treatment, can be induced to double their chromosome number and give true breeding, homozygous lines that will express both recessive and dominant mutated alleles.

Treatment of whole plants is less common but can be achieved using X-rays or gamma rays. This is often carried out using small plants or plantlets but, for example in Japan, they have built a large facility (resembling a sports arena) with a large gamma source at its centre, and a large number of fully grown plants are exposed over varying periods.

The treatment of cuttings and apical buds with radiation or chemical mutagens can be effective in developing mutant types in new shoots and plantlets. An important factor is whether the meristematic region forms mutations, since this is the region from which the new propagules develop. Treating cuttings and apical buds has been particularly important in developing mutants in clonally propagated cultivars.

It is now becoming more popular to combine mutation with in vitro cell and plant growth. The idea centres upon mimicking the possibilities developed with micro-organisms and so often involves treating single cells with a chemical or radiation mutagen and screening regenerating plantlets. One very desirable approach is the use of selective media, which only allows the growth of specific mutant types. This has been useful in developing herbicide-resistant cultivars where a low concentration of the selected herbicide is added to the media, and so only herbicide-resistant cell lines multiply.

8.1.4 Dose rates

It is apparent that all mutagenic treatments are basically damaging to plants. When too high a dose rate is applied, all the plant cells may be killed. Conversely, if the applied mutagen dose rate is too low, then very few mutant types will be induced. It is therefore first necessary to determine an appropriate dose rate to use. The optimal dose rate will change according to crop species, plant part exposed to the mutagen and its physiological state. Indeed, the first stage of most mutagenesis-based breeding is to determine the most appropriate dose rate to minimize adverse effects, yet still produce sufficiently high levels of mutation.

In simple terms, dose rate is equal to mutagenic $c08-math-0006$ applied. In chemical mutagenesis this involves the concentration of mutagenic chemical and time that plant cells were exposed to the chemical solution. Intensity of radiation can be altered by varying the distance from the radiation source or by varying the radiation form. The dose received can also be adjusted by changing mutagenic intensity or exposure time (or both). It is common to investigate different dose treatments until one is found that allows about 50% of plants to survive the treatment. These tests are called lethal dose 50 or $c08-math-0007$ tests.

8.1.5 Dangers of using mutagens

Mutagenic chemicals and radiation are effective because they alter the genetic makeup of plants and create variation. They will, of course, similarly affect the DNA of plant breeders who are exposed to them! It is not therefore possible to overemphasise the importance of using appropriate safety procedures when handling any mutagen. As already mentioned, the facilities for applying mutagenic treatments (in this case mainly radiation) are not always directly available to the average plant breeder; specialized operators or personnel (e.g. hospital radiologists) usually carry out the actual exposure to the radiation.

To use chemical mutagenic agents safely requires a number of safety features, spelt out in many countries (and by most suppliers in safety/hazard assessments) by specific safety protocols. Staff using these chemicals should be aware of the advised risks and safety procedures. Minimum safety will probably require suitable gloves, protective clothing and safety glasses combined with compulsory ‘Good Laboratory Management Practice’. It is also important that procedures and equipment are in place to deal with appropriate disposal of chemicals, and to contain and clean up any accidental spills of mutagenic chemicals.

8.1.6 Impact of mutation breeding

Mutation-derived cultivars have been released as a direct result of mutagenesis, or have mutant genotypes as parents in traditional breeding programmes. Since the inception of mutation breeding, over 2,250 cultivars have been released worldwide (FAO/IAEA [Food and Agricultural Organization/International Atomic Energy Agency] Mutant Varieties Database). It should be noted, however, that a high proportion of mutation-derived cultivars that have been released have been ornamental plants and flowers rather than agricultural crops. Over 70% of these cultivar releases were developed directly from mutant breeding lines and many of them were developed and released in Asia, with 27% being developed in China and 11% developed in India. Mutation-induced cultivars are not quite as common in other countries, although over 125 mutant-induced cultivars have been released in the US and 32 in the UK (31 of which were barley cultivars) in the past 70 years.

The highest numbers of mutant cultivar releases were in rice (433), followed by barley (269), wheat (220), soybean (89), groundnut (47), maize (32), pea (32), cotton (24) and millet (24). Only 46 mutant fruit cultivars were released over the same period. Mutant genes were developed mainly for dwarf stature, improved disease resistance, stress resistance, herbicide resistance, and improved grain or oil quality.

Although many have argued that these released cultivars have made little impact on our agricultural crops, some major positive impacts cannot be denied. Semi-dwarf rice derived from mutation breeding has been cultivated on millions of hectares. The barley cultivar ‘Diamant’, developed as a gamma-ray mutant of ‘Valticky’, was selected to have the ert dwarfing gene. It has been estimated that over 150 cultivars have been released in Europe that have Diamant in their pedigree. In addition, the Scottish barley cultivar ‘Golden Promise’, which also has this mutant dwarfing gene, is arguably the best cultivar ever released in the country and has been a major contributor to the Scottish brewing industry. Similarly, mutant durum wheat occupied over 25% of the Italian wheat acreage in the mid-1980s. Finally, health concerns about ‘trans fats’ in our diets has prompted many food processors and others in the food industry to use non-hydrogenated vegetable fats, which are low in polyunsaturated fats. The first canola cultivar with low linolenic acid, ‘Stellar’, inherited the fatty acid desaturase gene from a German EMS-induced mutant line coded as M47. Other low linoleic acid mutants have subsequently been developed using microspore mutagenesis, and now ultra-low polyunsaturated canola oil cultivars with highly elevated oleic acid contents are commercially available.

8.1.7 Practical applications

Having decided which mutagenic agent to use and a suitable mutagenic treatment strategy (i.e. rate, time, and which plant part to treat), in reality the physical treatment of plant cells by a mutagenic agent to induce mutation is the easy part of mutation breeding. By far the most difficult aspect of mutation breeding relates to selecting desirable mutants while avoiding the subsequent detrimental effects of mutagenesis.

Mutagens are non-discriminatory and inevitably produce a complex mixture of mutations. Mutants selected as having the trait of interest may also have undergone some chromosomal rearrangements and structure changes as well as exhibiting non-genetic (or at least non-nuclear) aberrations. It is also rare that selected mutants have been genetically altered for the single gene of interest, and there may be multiple mutation events, most of which have a negative impact on the normal growth of the plant. In seed-propagated crops, sterile segregants need to be sorted out and discarded in the first round of selection. But even having done this, selected mutants will have been altered in a number of different ways – most of them bad!

Consider the following example of using mutation breeding techniques in yellow mustard. The aim of the breeding programme was to screen a large number of mutants (derived by chemical mutagenesis using EMS) to produce lines that were low in seed meal glucosinolate content. After screening several thousand lines over a four-year period (using glucose-sensitive test tape), eight lines were identified that showed lower glucosinolates compared with the non-mutated parent genotype (‘Tilney’).

However, when these eight lines were grown in replicated field trials it was found that all mutant lines were considerably lower in yield, produced smaller and less vigorous plants and matured later than the parental genotype (Table 8.1). So although exposure to the mutagen had produced mutations that had affected glucosinolate content in the manner hoped for, it had also affected other aspects of the genotype such that the selected lines all appeared to have mutated for other important traits. Further crossing and selection was clearly necessary before an adapted cultivar could be developed.

c08-math-0008 — **Table 8.1** Seed yield, oil content, ground cover, days from planting to flowering and days from planting to flower ending of the yellow mustard (*Sinapis alba*) cultivar ‘Tilney’, and eight EMS mutants selected from Tilney with modified seed quality.

It has to be pointed out that experience suggests that, apart from loss of function alleles, it is usually easier to find a new allele than to create one. It may be significant to note that several years of intense mutagenesis breeding effort to develop yellow mustard lines with very low glucosinolate content in the seed meal failed to achieve this objective. Interestingly, the year after the mutagenesis programme was stopped, a gene that almost eliminated seed glucosinolates in yellow mustard was identified within a wild Sinapis alba population from Poland.

Initially many plant breeders believed that mutation breeding would have a revolutionary effect on cultivar development. Although there are multiple examples, like those above, of success in mutation breeding, for many this revolution has not happened and it is unlikely it will do so. Indeed, many believe that mutation breeding is too unpredictable and should only be considered as a last resort technique when all other avenues have tried, because of its limitations.

The question must therefore be asked as to what are the circumstances in which mutation techniques can be useful? To address this question the following points should be considered.

Mutagenesis is an indiscriminate process and generates large numbers of undesirable variants along with those that are wanted.
Most successful mutant cultivars and mutation-derived cultivars have resulted in selection for characters controlled by single (or at best a few) genes. Quantitatively inherited traits are more challenging in mutation breeding and will require large efforts to achieve success, if indeed success is ever achieved.
The frequency of desirable mutants is likely to be low, and it is essential that there are suitable selection techniques to screen the thousands of mutants generated to identify the few that have the required mutated trait. The most common example of mass selection in mutation breeding relates to herbicide tolerance. Most herbicide tolerance in crop species is qualitatively controlled, and many thousands of mutated breeding lines can easily be screened for tolerance by simply spraying the mutants with the herbicide of choice, on the premise that all those that survive carry a mutated form of the gene conferring tolerance. Other examples would include selecting mutants that exhibit morphological changes in plant structure (e.g. dwarfs), maturity, flower colour, qualitative disease resistance, or enhanced end-use quality (e.g. fatty acid profile or starch types).
When desirable mutants are selected they are usually adversely affected by the mutagenic treatment, and it will be necessary to ‘clean up’ the mutant genotype either by recurrent selfing and selection, or more likely by recurrent back-crossing and selection. Most mutations are recessive, and identification of recessive mutations is difficult due to them being masked by their dominant counterpart alleles. In diploid seed-propagated crops, recessive mutations can be identified by selfing or inter-mating mutated lines, albeit the frequency of lines that are homozygous recessive for the desirable gene mutation will be very low. For obvious reasons, selection of desirable recessive mutations in clonally propagated crops is considerably more difficult, and breeders require excessive effort in selfing mutant lines, or hope for the extremely rare event where both (all) alleles at a given loci have the same recessive mutated gene.

The unpredictable nature of mutagenesis raises the question as to whether it is possible to ‘direct’ the mutational effects towards changing only characters of interest and to affect them in rather particular ways. The first possibility for ‘direction’ of the effects that can be exploited is that different mutagens have different forms of action, as noted earlier. Some induce point mutations and so are more likely to produce a particular array of effects, while others are likely to induce grosser structural chromosomal changes. An even more specific array of possibilities arises from the potential to induce site-specific mutagenesis. Consider our increasing ability to identify particular genes, to clone or synthesize these and introduce them back into the plant species (or, of course, to another) – this clearly raises the potential to ‘mutate’ these DNA sequences and reintroduce them.

One lesson that was learned from the early efforts of mutation breeders is that it is necessary to have clear objectives, which are biologically reasonable, if success is to be achieved. However, even in cases that have been well organized with realistic objectives, the effort that was required to sort out the desired products in a useable form was often greater than would have been required to achieve the same results from a more traditional hybridization breeding scheme.

In plant breeding the need for new sources of variation and mutations (natural or induced) will feature as part of future breeding efforts. Therefore mutation breeding has a very real place in cultivar development, but it would be unwise to base a complete variety development programme on mutagenesis.

In summary, a mutagenesis breeding programme must deal with large numbers of mutated lines so that the low frequency of desirable mutations, in an acceptable genetic background, can be selected. Similarly, a mutation scheme must offer a quick, cheap and effective selection screen to identify the few desirable mutants.

8.2 Interspecific and intergeneric hybridization

Another method of increasing the genetic diversity of a crop species is by interspecific or intergeneric hybridization. When sources of variation for a character of interest (e.g. disease or pest resistance) cannot be found within existing genotypes in a species, it is logical to look at related species or genera and examine the possibility to introgress traits from them into the one of interest.

Interspecific hybridization refers to crosses between species within the same genus (i.e. $c08-math-0009$ ), while intergeneric hybrids are crosses between different genera $c08-math-0010$ .

The probability of developing a successful new cultivar is related to the frequency of desirable (or undesirable) characteristics in the parents used in hybridization. The most commonly used parental lines will be adapted cultivars or highly desirable genotypes from within breeding programmes. When a character of interest is not available within this gene pool, then the obvious next step is to screen other lines that are not as adapted, or indeed in wild genotypes, in an attempt to identify expression of the desired trait in them. If the character cannot be identified within this wider germplasm source, then breeders spread their search wider and will screen related species in an attempt to find a natural genetic source.

Successful interspecific or intergeneric hybridization should therefore be considered when:

the desired expression of a character of interest is not available within the gene pool of adapted genotypes, or their unadapted counterparts from the same species;
acceptable expression for this trait has been shown to exist within a related species or genera;
it is possible to introgress alleles from the related species into the cultivated species.

Successful interspecific crossing depends on two factors: obtaining viable seeds from plants in the first generation of the cross between the species (and later generations), and eliminating undesirable characters also introduced from the donor species. One, or both, of these factors may be the major determining factor in the actual success of gene transfer between species by this approach (see later for the possibilities using genetic transformation).

8.2.1 Characters introduced to crops from wild related species

A high proportion (over 80%) of genes introduced to our crop species through interspecific or intergeneric hybridization relate to pest and disease resistance. This trend continues today, whereby wild species related to our crop species are continually being screened and evaluated to identify new genes for resistance to crop pests and diseases. Resistance to grassy stunt virus was introgressed from Oryza nivara to cultivated rice. A number of late blight resistance genes have been transferred from Solanum demissum into potato cultivars. In addition, most new potato cultivars released in the European Union contain the $c08-math-0011$ gene conferring resistance to potato cyst nematode (Globodera rostochiensis) transferred from S. verni. Cabbage seedpod weevil resistance has been transferred to rapeseed through intergeneric hybridization between Brassica napus and Sinapis alba. More recently, genes conferring resistance to Hessian fly, a major insect pest of wheat in the US, have been transferred from Aegilops tauschii. Several rust-tolerant genes have been introgressed into wheat bread from related Triticum species.

In addition, significant efforts to develop genetic tolerance to wheat stem rust race Ug99 through the introgression of gene Sr39 existing in goatgrass (Aegilops speltoids) have been carried out. Sr39 provides tolerance to multiple stem rust races. The use of molecular markers helped to eliminate over 95% of the goatgrass DNA from $c08-math-0012$ genotypes, as well as to monitor the introgression of this gene into elite wheat breeding lines.

Other traits that have been transferred through interspecific or intergeneric hybridization include abiotic stresses, drought tolerance, heat tolerance and salinity tolerance. Examples of enhanced yield or quality characters from wild relatives into crop plants, not surprisingly, are very rare because many different loci would need to be introgressed, which increases the probability of also passing undesirable loci. Nevertheless, there are successful stories about the use of wild relatives to enhance the pool of genetic variation existing in crops; using wild Brassica oleracea resources, scientists at the John Innes Centre (Norwich, UK) have been able to significantly enhance in broccoli (Brassica oleracea) the accumulation of glucosinolates – secondary metabolites that play a cancer-preventive role in humans. The first commercial cultivars with these enhanced levels of glucosinolates have already been released to the market in several countries.

8.2.2 Factors involved in interspecific or intergeneric hybridization

In order to make a successful hybrid involving two different species, there needs to be some degree of compatibility between the parents used. A number of factors need to be addressed to ensure successful gene introgression.

The first stage, which must be achieved, is that the male and female gametes from the different genotypes must unite to form a zygote. Failure at this stage can result from:

inability of pollen grains to germinate on the receptive stigma of the female parent;
failure of pollen tubes to develop successfully and grow down the style, or non-attraction of the pollen tube towards the ovary;
inability of male gametes that do reach the embryo sac to actually fuse with the egg cell;
inability of the nuclei from pollen and egg to fuse.

All of these aspects are related to fertilization barriers, and a number of techniques can be used to overcome the problems at each stage. In vitro fertilization (i.e. using excised organs) can sometimes be used to overcome problems in the first two barriers listed above.

Success in interspecific hybridization may be unidirectional (i.e. style length differences that mean that pollen tubes fail to reach the ovary) and these can be overcome by attempting the reciprocal cross. Therefore successful hybrids might be possible from the mating $c08-math-0013$ , but difficult or unsuccessful when tried as $c08-math-0014$ . A good example of this is seen in the cross $c08-math-0015$ , which will produce viable hybrid seed if the cross is carried out in this direction (i.e. B. napus as female). If, however, B. oleracea is the female, then very few or no seed is produced (without using tissue culture techniques).

Cross incompatibility resulting from failure of pollen grains to germinate and develop pollen tubes is associated with proteins on the pistil that interact unfavourably with proteins in the pollen. In some cases this reaction has been overcome by mixing pollen from the donor species with compatible pollen from the female species (so called “mentor effect”).

Pollen tubes often fail to reach the ovary (or ‘miss’ the ovary) due to the physical differences in style lengths between the different species. This can sometimes be overcome by mechanically reducing the style length of the longer style parent, although this will only be successful if the shortened pistils remain receptive (i.e. as in maize). In the extreme case the complete pistil can be removed and pollen applied directly into the ovary, but this usually requires in vitro techniques.

Once fertilization is achieved the problems are not necessarily over. When two species differ in ploidy level the resultant offspring are usually sterile, and it may be necessary to first reduce or increase the ploidy of the parents prior to crossing. In potato, potato cyst nematode resistance was identified in Solanum verni (a close relative to cultivated potato, S. tuberosum). S. verni is diploid while S. tuberosum is tetraploid. Two methods of successful hybridization have been achieved:

Doubling the ploidy level of S. verni, using colchicine, and then carrying out interspecific hybridization at the tetraploid level.
Producing dihaploids from S. tuberosum and crossing the two species at the diploid level. Progeny from the hybrid cross are then doubled to the tetraploid level using colchicine or by spontaneous doubling arising during growth in vitro.

A similar manipulation of ploidy in interspecific crosses in potato was used to introgress late blight resistance (Phytophthora infestans) into cultivated potato cultivars. The source of blight resistance was found in a wild relative of the cultivated potato (S. demissum), which is a hexaploid. A small proportion of tetraploid progeny can be obtained by crossing dihaploid S. tuberosum (see above and in haploid section) with the hexaploid S. demissum.

If attempts to obtain hybrid seed by means of sexual crossing fail, then somatic fusion (protoplast fusion) can provide a realistic possibility. Isolated wall-less cells (protoplasts) can be induced to fuse and affect the production of a somatic hybrid, and hence facilitate genetic transfer between two species if it can be followed by regeneration of whole plants. The resulting somatic hybrids will have the combined chromosome number of both parents (e.g. as in allotetraploids), so it may be necessary to first reduce the ploidy of parental lines or reduce the ploidy of hybrid combinations produced. However, the most difficult aspects of this technology are:

being able to regenerate plants from protoplasts, even without fusion;
selecting fused heterokaryons from unfused or self-fused parental protoplast.

8.2.3 Post-fertilization

After fertilization, failure of seeds to develop and/or to reach maturity can result from embryo and/or endosperm abortion, or failure in the stages of embryo or fruit development to complete their necessary development to give mature seeds.

Successful fruit and flower retention after fertilization can be a simple function of a dependency on having a sufficient number of developing embryos. In some interspecific or intergeneric hybrids the number of fertilized ovules is too low to stimulate mature fruit development, and growth regulators (e.g. gibberellic acid) have been used as a means to encourage fruit retention. It has also been suggested that increasing the frequency of developing seeds in fruits (by applying a mixture of compatible and incompatible (mentor) pollen) can be used to avoid flower or fruit abscission.

Many interspecific or intergeneric hybridizations fail as a result of post-fertilization factors, which cause embryo or endosperm abortion. It may, however, still be possible to obtain hybrid plants despite abortion by using in vitro techniques such as:

ovule culture, where the complete ovary is removed from the plant and aseptically transferred to a growth media chosen where the nutrients are provided and thus seed development can proceed;
embryo rescue, where immature embryos are excised from the ovary and transferred to growth media; the chosen media therefore replaces the natural endosperm (which may not have developed or has aborted).

Sometimes a combination of both techniques is necessary to successfully achieve interspecific hybrid seed production. Early in the embryo development the ovary is removed and cultured in vitro to achieve embryos, which are then transferred when of a suitable size in order to allow successful rescue and culture.

Finally, it is not uncommon in hybrid crosses that rather than resulting in hybrid combinations the resulting seeds develop as matromorphic plants, which are thus derived exclusively from the maternal genotype. This characteristic has been developed to advantage in producing homozygous lines (doubled haploid lines) (e.g. $c08-math-0016$ ). The seeds from interspecific crosses should thus be checked to ensure that the matromorphs are discarded if the desire is to produce hybrids – but retained if this feature is being used to produce haploids of the maternal genotype!

8.2.4 Hybrid sterility

In many cases the $c08-math-0017$ plants resulting from interspecific crosses are completely (or partially) sterile. A common technique used to overcome sterility, caused by lack of chromosome pairing, is to induce chromosome doubling in the hybrid, and hence develop alloploids. When doubled, it allows each chromosome to have a homologue with which to pair at meiosis, and thus reduce the infertility problem.

8.2.5 Backcrossing

After interspecific hybridization the resulting progeny will generally contain a large proportion of undesirable characters from the donor species, along with the character it was wished to introduce. In such circumstances it is necessary to carry out several rounds of backcrossing to the host species, with selection for the new character, to obtain genotypes that will have commercial value. Any programme involving interspecific or intergeneric hybridization is therefore likely to be long term, but, of course, once you have a well-adapted genotype you can use it as a parent for further variety production.

8.2.6 Increasing genetic diversity

Many crop species have a relatively narrow genetic base, and it is often advantageous to broaden genetic diversity by introgressing traits from related weedy species. Several crop species (i.e. rapeseed and wheat) have evolved as allopolyploids, whereby they contain complete chromosome sets from two or more diploid ancestors. Greater genetic diversity and variation can be achieved in breeding by resynthesizing the crop species from its ancient ancestors. One such example has been the use of alleles from Brassica oleracea wild accessions to increase the levels of glucosinolates, in broccoli.

8.2.7 Creating new species

It is even possible to create new crop species by intergeneric hybridization, but despite the possible attraction of this there are very few instances where successful new crops have resulted. However, three notable examples are:

Triticale, which resulted from intergeneric hybridization between wheat (Triticum) and rye (Secale).
Tritordeum, which resulted from intergeneric hybridization between wheat (Triticum) and barley (Hordeum).
Raphanobrassica, which resulted from the intergeneric cross between kale (Brassica oleracea) and radish (Raphanus sativus).

c08f001 — Figure 8.1 Winter biennial forms of yellow mustard (*Sinapis alba*) (left) and Indian mustard (*Brassica juncea*) (right) produced through intergeneric hybridization.

When each of these new species was created there was great hope that they would almost immediately have high potential commercial value. However, in none of the cases has this full commercialization at a global scale occurred – at least not yet. Only in the case of triticale have a limited number of cultivars been released; nevertheless, on a worldwide scale its impact, in terms of food/feed production, has been negligible thus far.

8.3 Plant genetic transformation

The stable, heritable introduction of foreign genes into plants represents one of the most significant developments affecting the production of crop species in a continuum of advances in agricultural technology relating to plant breeding. The progress in this area has depended largely on two main components: the tissue culture systems having been developed that provide an amenable vehicle for the transformation induction; and the use of molecular biology to isolate, modify and clone agronomically relevant genes and the transfer of these into plant species.

The term ‘genetic transformation’ comes from that used for a much longer period, bacterial transformation in which DNA has been successfully transferred from one isolate to another or between species of bacteria, and integrated into the genome. It was shown that the stably transformed bacteria then expressed the new genes and had appropriately altered phenotypes. In eukaryotes, transformation has a further complicating dimension, at least in many plant breeding contexts. The transforming DNA must not only be integrated into a chromosome, but it must be a chromosome of a cell or cells that will develop into germ-line cells. Otherwise the ‘transformation’ will not be passed on to any sexual progeny. Transient transformation refers to the event where the introduced gene is not fully integrated into the host genome.

Using plant transformation techniques has made feasible the transfer of (mostly) single genes (i.e. simply inherited traits) into plants, to have such transgenes properly expressed and to function successfully. Theoretically, at least, specific genes can be transformed from any source into developed cultivars or advanced breeding lines in a single step. Plant transformation, therefore, allows plant breeders to bypass barriers that limit sexual gene transfer, and exchange genes (and traits) from unrelated species where incompatibility does not allow sexual hybridization. These recombinant DNA techniques therefore apparently allow breeders to transfer genes between completely unrelated organisms. For example, bacterial genes can be transferred and expressed in plants. Nevertheless, up to this point the genes or gene components used to develop transgenic plants only include plant, bacterial and/or virus origins or genes modified in the laboratory. None of the transgenic plants currently commercially available to farmers in many parts of the world contain genes, or components thereof, of human and/or animal origin.

However, as we learn more and more about the DNA and hence the genes involved, the perspective of the picture somewhat changes. Increasing direct evidence of the presence in different species of the same basic gene, or clear variants of it, demonstrate the greater conservation of genetic material during evolution than we expected. Also, we are being reminded of the existence of parallel natural processes for much of what we regard as novel. For example, bacteria, viruses and phages have already successfully evolved mechanisms to transfer genes, in just the same way that is regarded by some as being so alien! But clearly, the new techniques enable modern plant breeders to create new variability beyond that existing in the currently available germplasm on a different scale and in a different time-frame from that which was possible previously. Each plant carrying, at a specific chromosomal location, a given DNA construct introduced through genetic transformation is referred to as arising from a specific transformation event. Events with unique, stable chromosomal insertions are commonly selected for commercial development. Many such events have been developed, such as DAS-1507-1, MON-89034-3 and SYN-IR162-4 created by Dow AgroSciences/Pioneer, Monsanto and Syngenta, respectively; they all encode tolerance to Lepidopteran insects in maize. There are also an increasing number of commercial events developed in the academic/public sector, such as EMBRAPA in Brazil and USDA/Cornell University in the US. For instance, EMBRAPA (Brazil's National Institute of Agronomic and Livestock Research) has just developed event EMB-PV051-1 in common bean in Brazil, which renders this key crop tolerant to the bean golden mosaic virus.

In addition, a myriad of universities and research institutions in most countries now develop transgenic plants on a more or less routine basis for diverse research purposes. Although plant transformation has added (and some say dramatically) to the tools available to the breeder for genetic manipulation, it does, as with all techniques, have its limitations. Some of the limitations will reduce with increased development of methodologies, while others are those that are inherent to the basic approach.

At present, recombinant DNA techniques tend to be restricted to the transfer of a single or a few gene(s). This means that they are very effective where the trait is determined by one, or a few, gene(s). However, some agronomic traits showing continuous variation are actually controlled by a few loci showing rather large effects, but others by a myriad of genes with much lesser effects. So, for example, it is not clear how much yield itself, which could be argued is one of the most important characters of interest to farmers, can be manipulated by discrete steps of individual transformation events. Interestingly, however, recent reports do indicate the potential to transform with a number of genes (constructs) in one go with a reasonably high level of co-transformation. In addition, it is feasible to pile up, or stack, several transformation events through breeding or molecular approaches. In the case of maize, at the time of writing there are hybrids developed for several countries that carry up to eight genes, and soybean varieties are being developed that carry up to three stacked events.

It may seem obvious, but another restriction currently imposed is that the techniques are only readily applied to genes that have been identified and cloned. Despite the development of genomics, bioinformatic and high throughput sequencing approaches, the number of such desirable genes is still relatively modest, although increasing rapidly. What is becoming clear is that there is a deficiency in the knowledge of the underlying biochemistry or physiology of most agronomically relevant traits. Another feature is the rather limited availability of suitable promoters for the genes that are to be introduced. Optimizing the expression of transgenes for a particular developmental stage or in specific tissues has been fully recognized, and so the search for promoters now equals that for the genes themselves. In addition, it has been recognized that because of the random nature of the incorporation of transgenes into the host's genome, a large number of transformed plants need to be produced in order to allow the selection of the few that have the desired expression of the transgene without any detrimental alteration of all the characters of the host. Breeders have played a critical role in the development of novel breeding schemes and selection approaches, so enabling the selection of the best events which have then been used in breeding programmes.

8.3.1 A glimpse at the genetic transformation of plants

Before plant transformation can be used successfully in a plant breeding programme and cultivars are developed using recombinant DNA techniques, the following have to be in place.

A desirable, agronomically relevant gene must be available for insertion into the target host plant. Therefore a DNA construct containing the gene that will confer a particular trait of interest must be developed and introduced to a plant, which then could be used in breeding programmes.
There must be a suitable mechanism to transfer the gene to the target plant. In dicots (and also now increasingly cereals and grasses) the most commonly used vector has been A. tumefaciens. Agrobacterium tumefaciens is the causal agent of crown gall disease and produces tumorous crown galls on infected plants. The utility of this bacterium as a gene transfer system was first recognized when it was demonstrated that crown galls were actually produced as a result of the transfer and introgression of genes from the bacterium into the genome of the host plant cells. This natural mechanism for the introduction of genetic material that Agrobacterium has developed is an exquisite example of a naturally evolved mechanism to transfer genes from one species into another, also known as horizontal gene transfer. The crown galls produced by Agrobacterium induce in its natural hosts the synthesis of plant growth regulators and other compounds like opines which are used by Agrobacterium as a source of nitrogen and carbon. Vectors based on Agrobacterium used to develop transgenic plants are known as disarmed, since they lack the set of genes inducing crown galls. Therefore they retain the ability to transfer DNA, yet lack any genetic component that might represent a pathogenic, disease-causing factor.

Several physical or mechanical DNA delivery systems have also been developed, and have been particularly popular for monocots. The most common, at least initially, of these systems involved electroporation of protoplasts, although now particle bombardment is regarded as the method with the widest applicability. Particle bombardment involves the use of gold or tungsten microparticles $c08-math-0018$ , carrying the DNA to be introduced on their surface, which are accelerated to speeds of up to several hundred metres per second. Particle bombardment has been used for gene transfer into a variety of target tissues including pollen cells, apices and reproductive organs.

Because transformation events created through particle bombardment tend to produce less precise or multiple insertions, Agrobacterium tumefaciens-based approaches are generally the methods of choice in current research efforts aimed at creating transgenic variation.

A suitable construct has to be created that includes:
- a promoter region that is recognized by the host, which drives the expression of an introduced gene and therefore its pattern of expression. These may be:
  - Constitutive promoters such as the 35S of cauliflower mosaic virus or nos from the nopaline synthase gene, which switches the gene on in all plant tissues.
  - Tissue specific promoters such as those from $c08-math-0019$ -amylase (specific to the aleurone in some cereal species), patatin (specific to tubers) and phaseolin (specific to cotyledons).
  - Inducible promoters such as those from alcohol dehydrogenase I (induced by anaerobiosis); and chlorophyll a/b binding protein (induced by light).
- a structural gene that encodes a protein involved in a trait of importance to farmers, such as insect resistance, herbicide tolerance, drought tolerance, virus resistance or quality.
- a transcription terminating sequence at the $c08-math-0020$ end of the gene which signals the end of the transcription of the structural gene.

c08f002 — Figure 8.2 Regeneration of Indian mustard (*Brassica juncea*) plantlets from callus tissues.

It should also be noted that molecular biologists might have to redesign the gene of interest, for optimal codon usage (i.e. codons most frequently used by the protein synthetic machinery in the host plants). This is because the codon usage varies among species, and it might be the case that a codon frequently used in the donor species to encode a given amino acid is less frequently used in the host species. In addition, it is sometimes necessary to include a ‘transit’ peptide in the genetic construction, when the expressed protein functions in a particular cell compartment (for instance the chloroplast). In addition, optimization of gene expression might be achieved by including short DNA sequences (‘enhancers’) in the promoter region.

Regardless of the delivery (vector) system used to transform plants, foreign DNA will be inserted into relatively few cells. A means, therefore, must be available to select, or at least significantly enrich, the cells that have been transformed. This is achieved through the use of a selectable gene marker along with the gene of agricultural, nutritional and/or industrial relevance that needs to be introduced into the host plant. In the early years of genetic transformation, the most usual selection agents used were antibiotics (mainly kanamycin or hygromycin) or herbicides (i.e. glufosinate or glyphosate). As the few cells effectively transformed also carried a selectable gene marker rendering them tolerant to such selection agents, they were able to thrive in selective culture media and give rise to adult, fertile plants. A growing number of transgenic plants lack selection markers, since depending on the genetic transformation strategy employed, they could be segregated out from the structural gene using routine breeding procedures.

Thus, genetic transformation enables the insertion of foreign DNA into the genome of a few selectable cells. A method must, of course, be available to develop intact mature, fertile plants from these single transformed cells. One of the major early barriers in transformation of a number of crop species was the inability to regenerate whole plants from single cells in vitro. However, over the years efficient tissue culture systems have been developed, enabling the generation of full, fertile plants from genetically transformed tissues in many species. In many dicots, leaf disks are transformed by infection with A. tumefaciens. Plantlets are then regenerated by tissue culture methods from the leaf disks. In many monocots, cultured cells or embryos are transformed by a suitable DNA delivery system (e.g. the particle gun) and intact plants are then regenerated from transformed cells, again in tissue culture. Other methods that have had success are the transformation of embryogenic cell cultures or protoplasts, followed by regeneration of whole plants.

Once whole, fertile transformed plants have been produced they need to be thoroughly characterized. This may be achieved by some or all of the following, often in series:

Polymerase chain reaction (PCR) techniques can be used to detect the presence of transgenes, although these techniques alone cannot fully characterize insertions into the host genome.
Southern blots are used to assess how the gene is integrated into the plant genome, and to estimate the number of gene copies that have been inserted. It is likely that current Southern blots will be replaced by electronic Southern blots in the years to come.
Northern blots to assess gene expression through RNA transcribed from transgenes. Northern blots have been increasingly replaced by high throughput approaches such as quantitative reverse transcriptase-PCR.
Western blots and enzyme immunoassays to characterize protein expression of the transgene in plant tissues

However, this is just the start, since after such tests have shown that transformed plants have the gene of interest, that it is integrated and that it is functional, this simply means it is worth proceeding further in the development of the GM crop. It then needs to be demonstrated that the gene (and expression of the trait it encodes) is stable; this first means demonstrating Mendelian transmission of the transgene through clonal or sexual generations in order to show that any progeny will inherit and express the gene as expected. Subsequently, extensive field and molecular testing begins to ascertain whether the expression of the gene has the desired effect on the phenotype, whether any other characters have been affected directly or by the transformation process and, of course, what the actual field performance is. Once field performance has been demonstrated, plant transformation events must undergo the thorough, extensive and detailed characterization analyses requested by regulatory agencies before their commercialization is allowed.

Genetic transformation does not constitute a breeding approach by itself. Instead, it is a very powerful source of novel genetic variation which, like any other source, must be channelled through an effective breeding programme, if successful cultivars carrying an agronomically relevant transformation event are to be developed and commercialized.

8.3.2 Some applications of genetic engineering to plant breeding

Weeds represent a significant constraint to crop production in most countries. Weed control is based on the use of herbicides, although there are still countries where it is based on human labour. If a plant was able to withstand herbicides while its associated weeds do not, it would address the needs of many farmers. Engineering herbicide tolerance into crops represents a novel alternative for conferring selectivity of specific herbicides. Two general approaches have been taken in engineering herbicide tolerance:

altering the level and sensitivity of the target enzyme to the herbicide;
incorporating a gene that will detoxify the herbicide.

As an example of the first approach, glyphosate, the active ingredient of herbicides such as ‘Roundup’, acts by specifically inhibiting the enzyme 5-enolpyruvylshimate-3-phosphate synthase (EPSPS). Tolerance to glyphosate has been engineered into various crops such as maize, soybean, canola, cotton and wheat by introducing genetic constructs for the production of a tolerant variant of the EPSPS from Agrobacterium strain CP4.

Genetic transformation has also been used to develop plants tolerant to other herbicides such as gluphosinate. In addition, novel systems of herbicide tolerance are being developed which would enable the use of herbicides such as 2,4 D and dicamba in weed control programmes based on the use of herbicide tolerant plants developed through biotechnology.

The production of plants that are tolerant to insect damage has been another application of genetic engineering with important implications for crop production. Transgenic plants with insect tolerance are widely used in maize and cotton in several countries already, and there are already transgenic soybeans able to withstand the attack of several Lepidopteran species.

Insect damage not only can create significant yield reduction by itself, but additionally the wounds it creates can open the way for opportunistic fungi to attack plants already damaged by insects. One route by which progress in engineering insect resistance in transgenic plants has been achieved is by using the genes of Bacillus thuringiensis (B.t.) which produces insect-specific endotoxins (so-called B.t. toxins). B.t. is a bacterium that produces a crystalline protein (also called Cry proteins) during sporulation which, when cleaved to the mature toxin peptide, produces paralysis of the intestine of specific target insects and so leads to their death. Thus it provides a useful and selective means of insect control. It is interesting to note that diverse microbial formulations of B.t. have been used for many years to control insects in organic food production systems. There are other sets of B.t. proteins that have also enabled the development of commercial insect resistance transformation events, such as VIPs (vegetative insecticidal proteins).

However, with these resistance genes, an overall strategy is needed to prevent (or at least significantly reduce) evolution of resistance to the toxin in the insects that are being controlled. Insect resistance management programmes involve the use of refuges – field sections cultivated with an insect-susceptible version of the same cultivar – in order to reduce the likelihood of selection of insects resistant to B.t. toxins.

Another trait of interest is virus resistance, where significant resistance to tobacco mosaic virus (TMV) infection, termed ‘coat protein-mediated protection’, has been achieved by expressing only the coat protein gene of virus in transgenic plants. This approach has produced similar results in transgenic tomato and potato, although in some other cases a similar approach seems not to be as effective. It has been suggested that other genes, such as viral replicase genes among others, might provide an effective mechanism to control virus infection in plants. Recent work has shed light on the mechanism involved in such examples, and it has been demonstrated that these transgenic phenotypes are the result of RNA interference, an extremely specific DNA-homology based, naturally occurring mechanism, that enables the switching-off of the expression of genes.

There is a wide list of dicot crop species that have proved successful hosts for transformation, including alfalfa, apple, carrot, cauliflower, celery, cotton, cucumber, flax, horseradish, lettuce, potato, rapeseed, rice, rye, sugarbeet, soybean, sunflower, tomato, tobacco and walnut. In monocots, maize is leading the way, but is being followed by wheat, barley and rice.

Initial cultivar development using recombinant DNA techniques has focused upon modifying or enhancing traits that relate directly to the traditional role of farming (i.e. agronomic traits). These, as noted above, have included the control of insects and weeds, and mostly have relied on the expression of single bacterial genes in plants. What can be termed the ‘first generation’ of genetically engineered crops have been released into large-scale agriculture (including maize, tomato, canola, squash, potato, soybean and cotton) during the last 15 years, and other crops are already ‘in the pipeline’ as again mentioned above. More recent work has focused on modifying end-use quality (especially fatty acid, starch and vitamin precursors). In addition, transgenic maize plants aimed at industrial applications are already commercially available in the US. For instance, Syngenta developed transgenic corn hybrids carrying an enhanced form of the enzyme alpha amylase, which increases the efficiency of the corn-to-ethanol production process. Farmers growing such corn hybrids receive a premium price when delivering the corn to ethanol-producing facilities. The first transgenic maize plants able to successfully withstand certain levels of drought at the farmer's field level have been made available to farmers in the US by Monsanto.

It is expected that the availability of novel genes and genetic transformation methods will allow the development of transgenic plants carrying traits of increasing complexity and agronomic relevance. Nevertheless, much research is still needed before the availability of transgenic plants able to carry complex traits like grain yield or disease tolerance becomes a reality.

The genetic transformation of plants has had a significant and positive impact on global agriculture. The first transgenic cultivars were made available to farmers in 1994, and since then, over 1,500 million hectares have been grown with transgenic cultivars, an area almost 50% larger than the total land mass of Europe. Not only farmers running large farming operations have benefited from these, but also millions of small farmers around the world.

8.3.3 Cautions and related issues

There have been a number of concerns that have arisen over the past few years as the application of plant transformation technology has expanded, and particularly as new transgenic crops have been released into commercial cultivation. Plant breeders need to be aware of the concerns, as well as the local and international regulations, that apply to plants derived using recombinant DNA techniques, as well as to other forms of gene manipulation (e.g. induced mutations). In addition to the general social and environmental concerns, the breeder must consider the following:

Is the level of expression of the genetically engineered crop plant sufficiently useful to agriculture to merit the time and resources that have gone into its development, and what will need to be done in terms of any further development?
Is there a gene dose effect that will further optimize the results? In most cases what breeders require are single copy insertions which are inherited like any other Mendelian loci in their breeding programmes.
What would be the best way to deploy the technology? Plant breeders are aware of the constant evolution occurring in organisms causing diseases and pests, which can often overcome single gene resistance when introduced into commercial cultivars. In this regard, transgenic crops must be managed within schemes encouraging good agricultural practices, such as the use of refuges already described in the case of crops resistant to insects, and the rational use of herbicides and agronomic practices such as crop rotation in the case of plants resistant to herbicides.
It was at first, naïvely believed that plant breeders would be able to take an adapted cultivar and simply transform it with a specific gene to give an ‘instant’ new cultivar – one that had all the previous desirable characteristics but also with the transformed trait. It is now known that this is not in fact the case. New cultivars derived through plant transformation require the same rigorous field-testing prior to release that traditionally developed cultivars do. Multiple transformation events are necessary to ensure that one such event would give a transformed plant that has the desired level of expression for the altered trait, plus no deleterious epistasis interaction with the transformed gene or background changes. Therefore, genetic transformation represents a previously unforeseen source of jobs, professional development opportunities and research objectives to plant breeders.

Think questions

Under what circumstances would a plant breeder choose to use mutation breeding?

List two methods that have been used by plant breeders to induce increased frequency of mutagenesis in plants.

List three problems that a plant breeder may encounter when using mutagenesis in a breeding programme.

What is possible with plant transformation (recombinant DNA) techniques that is not possible with more traditional breeding techniques.

You have been asked to use recombinant DNA techniques in your alfalfa breeding programme to increase profitability. Describe all the processes you will have to complete to achieve this goal.

List three vectors that have been used to transform plants with genes from other organisms.

Briefly describe three problems that plant breeders may encounter in developing transgenic cultivars.

Name three types of mutation that can occur in plants, and briefly describe the features of each type. Outline two difficulties that might be problematic when using mutagenesis in a plant breeding programme.

Interspecific and intergeneric hybridization can sometimes be useful techniques in plant breeding by introgression of characters and genes from different species. In an interspecific cross between Brassica napus and B. rapa, there was no evidence that any B. napus egg cells had been fertilized. List three factors that could have caused this non-fertilization.

Identifier	Seed yield $c08-math-0008$	Oil content (%)	Percentage ground cover, 1–9 scale	Days to flowering (days)	Days to flower end (days)
Tilney	2,166	30.1	8.7	54	80
Til.M3.A	555	30.5	3.0	59	86
Til.M3.B	1,797	29.2	4.7	58	85
Til.M3.C	1,004	30.4	2.0	60	86
Til.M3.I	1,347	29.7	4.7	59	86
Til.M3.II	1,242	29.7	3.0	60	87
Til.M3.III	766	30.1	2.3	59	86
Til.M3.IV	1,480	31.0	3.3	58	86
Til.M3.V	1,215	31.0	5.0	58	86