Inbreeding

Introduction

Inbreeding for the development of homozygous and homogeneous lines is only important for the development of pure-line and hybrid cultivars. It is not used for the development of clonal or synthetic cultivars because self-pollination results in severe inbreeding depression.

It is possible to carry out inbreeding by half-sib or full-sib mating, instead of self-pollination. However, the number of generations required to reach homozygosity is more than through the use of self-pollination. As a result, half-sib or full-sib matings rarely are used to develop inbred lines for cultivar development. The chart in figure 1 illustrates the number of generations it may take to attain complete homozygosity by different mating systems.

Factors influencing the choice of methods for obtaining pure lines/inbred lines

Environments available

One of the key factors that influences the method of advancing a population from the F₁ to later generations is the environments available to the breeder. The environment for which a breeder is developing new cultivars will be referred to as the local environment. In most breeding programs today, a breeder will use off-season nurseries for generation advance, which include greenhouses, growth chambers, and locations where the environment is not the same as the local environment. For example, the wheat cultivar Select was developed by breeders in South Dakota for their local environment [JPR 5:196-201 (2011)]. For generation advancement of the population from which the cultivar was developed, the breeders used Yuma, Arizona, for the off-season nursery. The possibility of effectively selecting among plants or lines in Yuma for traits of interest would influence how a population would be planted and harvested in Yuma.

The effectiveness of selection in an off-season nursery is highly dependent on the trait. Although seed yield commonly cannot be evaluated in off-season nurseries, selection for other traits may be possible. For example, the cranberry bean cultivar Bellagio was developed for Michigan [JPR 4:171-174 (2010)]. In the off-season nursery at Isabela, Puerto Rico, the breeders were able to select for upright vine, lodging resistance, good pod load, cranberry bean seed size and color, and freedom from diseases.

The two methods that are most influenced by the environment are the bulk and pedigree. In the bulk method, the seed planted in one generation is obtained by sampling from a bulk of seed harvested the previous generation. The likelihood that a plant will have one of it progeny represented in the sample used for planting is directly associated with the number of seeds the plant produced. An F₂ plant that produced 40 seeds would be less likely to be represented in the next generation than an F₂ plant that produced 200 seeds. If the number of seeds produced by plants in off-season nurseries was not representative of their production in the local environment, they could be eliminated from the population by use of the bulk method, even though they may have the potential to be a superior new cultivar or inbred. Use of the pedigree method is also highly dependent on the environment of the off-season nursery. If a trait is not adequately expressed in the off-season nursery, selection would be ineffective.

It is common for breeders to use a combination of methods during generation advance. For example, a breeder may select among F₂ plants in the local environment, harvest individual F₃ plants in an off-season nursery without selection, and resume selection among F_3:4 lines the following season in the local environment.

Cost

Half of the genetic variation among individuals in a population will be expressed in the F₂ generation. To maximize genetic variation for seed yield and other traits for which selection is not feasible during generation advance, it is ideal to have as many different F₂ plants and their progenies represented in the population when lines are selected for evaluation. The methods that can assure that every line derived from a population traces to a different F₂ individual are the pedigree, the single-seed procedure of single-seed descent, and the single-hill procedure of single-seed descent. Doubled haploid lines normally trace to individual gametes from an F1 plant, each of which are genetically different. These methods require more labor than the bulk or the multiple-seed procedure of single seed descent. The breeder must decide how to balance cost versus genetic variability in the lines derived from a population.

Generation in which lines are derived

Regardless of the method of generation advance used for a population, the breeder must decide at which generation lines will be derived for extensive testing as a potential cultivar or as an inbred in a hybrid. In selecting parents for mating, genetic differences in their parentage or in their molecular makeup are emphasized; therefore, there would be many heterozygous loci in the F₁ hybrid. As a result, complete homozygosity is rarely achieved in plants from which pure-line cultivars or inbreds are derived, with the exception of the doubled haploid method. Instead, the breeder derives lines at a generation when there is an adequate frequency of plants whose progeny will be sufficiently uniform in appearance and homogeneous for major genes of interest.

With multiple segregating loci, the proportion of homozygotes in various selfing generations can be estimated using the following formula: [(2^m – 1)2^m]ⁿ: where m is the number of selfing generations (F₂ = 1; F₃ = 2) and n is the number of segregating loci. For example, an F₁ plant with four independent segregating loci will result in the following frequency of homozygous plants in F₂, [(2¹ – 1)/2¹]⁴ 1/16 = 6.25%. The expected proportion of completely homozygous plants in F₂ and later generations of selfing for different numbers of segregating loci are indicated in table 1. The proportion of homozygotes decreases sharply with increasing heterozygosity in the F₁.

Table 1. Frequency of completely homozygous individuals in various selfing generations in relation to the number of segregating loci in F_1.

Number of segregating loci in the F1	F2	F4	F6	F8
1	50.00	87.5	96.87	99.22
2	25.00	76.56	93.85	98.44
3	12.50	66.99	90.91	97.67
4	6.25	58.62	88.07	96.91
5	3.13	51.29	85.32	96.15
10	0.01	26.31	72.79	92.45
100	7.89 x 10-31	6.1 x 10-4	4.18	45.64
1000	9.33 x 10-302	1.02 x 10-58	1.63 x 10-14	0.04

The generation in which a cultivar or inbred line is derived for commercial use is influenced by the degree of homozygosity and homogeneity required and the cost and time involved in each generation of selfing. The number of heterozygous loci is an F₁ hybrid generally is very large, which means that only a very small percentage of plants are completely homozygous when lines are derived (Table 1). As a result, pure-line cultivars and inbreds of hybrids rarely are completely homozygous and homogeneous, unless they were derived from doubled haploids. The breeder attempts to achieve the level of homogeneity that is required for commercial acceptance of the cultivar or inbred line. In general, breeders of pure-line cultivars commonly derive cultivars in an earlier generation than breeders responsible for developing inbreds of hybrids. For example, the hard red winter wheat cultivar Camelot originated from an F₃-derived line [JPR 3:256-263 (2009)]. In contrast, the maize inbred line GT603 was developed through seven generations of self-pollination [JPR 5:211-214 (2011)]. One reason for seeking more homozygosity and homogeneity in an inbred is to be able to recognize and remove off-type plants in a hybrid seed production field before crossing occurs. A second reason is that if two inbreds are crossed that are not adequately homozygous and homogeneous, undesirable segregation may occur in a commercial hybrid field.

Examples of how the inbreeding methods described in the text have been applied in development of cultivars and germplasm can be found in the Journal of Plant Registrations. In the following examples, the underlined sentences are comments that did not appear in the text.

Example 1: Bulk method combined with mass selection

Information found in JPR 5:151-155 (2011) describes the development of ‘Gadsby’ barley by use of the bulk method and mass selection. The breeders developed Gadsby as a general-purpose barley with resistance to scald.

• Step 1. In 1997, the F₂ plants from a cross between the parents H92066001 and TR248 were grown in the field at Lacombe, Alberta, Canada. The plants were inoculated with scald by spreading straw that had been collected from a field infested with the disease the previous year. The F₃ seed of the population was harvested in bulk.
• Step 2. For mass selection, the F₃ seed was screened over a gravity table and the heavier seed was saved for planting. It was assumed by the breeders that the heavier seed was from F₂ plants with resistance to scald and the lighter seed was from susceptible genotypes.
• Step 3. In 1998, the selected F₃ seed was planted in the field and the first two steps were repeated again. The same two steps were repeated with the F₄ bulk in 1999 and the F5 bulk in 2000.
• Step 4. In 2001, the F₆ bulk was grown in 2001 at two locations and a head from each of multiple individual plants was harvested and threshed separately. The progeny of each plant was evaluated in subsequent generations as F₆-derived lines; therefore, Gadsby was an F₆-derived cultivar.

The purpose of mass selection is to increase the frequency of individuals with a desired trait in a population. By infecting the plants each generation and selecting the heavier ones for planting, the breeders of Gadsby assumed that the frequency of F₆ plants with resistance to scald would be greater than if the bulk method had been used without infection or mass selection.

Mass selection also can be used with single-seed descent. For example, if a disease caused resistant plants to look different than susceptible ones, one or a few seeds could be harvested from the resistant ones to plant the next generation.

Mass selection is not expected to be perfect in selecting desirable seeds or plants. In the breeding of Gadsby, some of the susceptible plants may have escaped infection and produced seed as heavy as the resistant ones. As a result, some of the seed planted the next generation may have been from susceptible plants. Even though it is not perfect, it can potentially increase the frequency of desirable individuals in a population.

It is important to recognize that the term “mass selection” when used in connection with inbreeding a population has a very different meaning than when used in connection with recurrent selection (Lessons 13 and 14 of this course). When a population is self-pollinated, as in the case of Gadsby barley, the frequency of homozygous individuals increases each generation of mass selection. When mass selection is carried out for recurrent selection, selected plants are intercrossed, which maintains the heterozygosity of individuals.

Example 2: Single-seed procedure of the single-seed descent

The hard red spring wheat cultivar Duclair was developed by use of the single-seed procedure of single-seed descent [JPR 5:349-352 (2011)]. Some of the information provided below is not in the article and was obtained directly from the senior author.

• Step 1. In the greenhouse, the breeders crossed the cultivar Choteau with the line MT0249 to form a single-cross population.
• Step 2. The F₁ plants were grown in the field and F₂ seed was harvested.
• Step 3. The F₂ seed were planted in the greenhouse in September. One head from each plant was removed. The plants in the greenhouse are small and tillering generally does not occur; therefore, the plants can be distinguished from each other. One head that usually contains four to eight seeds is removed from each plant. One healthy seed from each head is bulked together for planting the next generation. The heads are stored as a reserve in case something destroys the subsequent planting.
  • Step 3. The F₃ seeds were planted in the greenhouse. Each plant was harvested individually to form an F_3:4 line.
• Step 4. The F_3:4 lines were planted in Bozeman, MT, during May. Individual lines were selected and single plants were harvested from the selected rows. The authors refer to the F_3:4 headrows as F₄ head rows. In their system of nomenclature, the row is described by the generation of the plants that are growing and does not describe when the line was derived.
  • Step 4. The F_4:5 lines (F₅ headrows) were planted for evaluation. The F₄-derived lines were evaluated in subsequent generations. One of the F₄-derived lines became the cultivar Duclair.

Example 3: Multiple-seed procedure of single-seed descent

The breeders of ‘UA 4910’ soybean referred to their method of generation advance as modified single-pod descent [JPR 5:49-53 (2011)]. In the text for this course, their procedure is called the multiple-seed procedure of single-seed descent. When the term “modified” is used for a method of generation advance, it means that the breeders modified what they considered the standard way that a method would be carried out. The only way to understand what was modified is to carefully ready what they did in comparison to what would be the standard procedure.

• Step 1. In 1997, the cross of Asgrow A4715 to DP 3478 was made in the field at Fayettevile, AR, to obtain a single-cross population.
• Step 2. The F₁ plants were grown in the 1997-1998 winter nursery in Costa Rica and F₂ seed was harvested.
• Step 3. In 1998, the F₂ population was grown in the field and two to three pods each containing two or three seeds were harvested from about 600 plants and the pods were threshed in bulk. Part of the seed was planted in Costa Rica (season 2).
• Step 4. The F₃ plants were grown in the 1998-199 winter nursery in Costa Rica. Two to three pods each containing two or three seeds were harvested from the plants and threshed in bulk.
• Step 5. The F₄ seed was planted in Keiser, AR, during 1999, and 60 plants from the population were harvested and threshed individually. The progeny of the plants were evaluated as F₄-derived lines in subsequent generations; therefore, UA 4910 was an F₄-derived cultivar.

Example 4: Pedigree method

Georgia-09B peanut cultivar was developed by use of the pedigree method with a backcross population [JPR 4:175-178 (2010)].

• Step 1: The cultivar Georgia Green was crossed to the line GA 942004. The F₁ generation was grown to obtain F₂ seed. F₂ plants were grown and harvested individually. A sample of seed from each F₂ plant was evaluated for oleic acid concentration in the oil and one of them was chosen for crossing as a F_2:3 line.
• Step 2: In 1998, the F_2:3 line was backcrossed to the recurrent parent Georgia Green to obtain BC₁F₁ seed. The goal of the backcross was to develop a population that had individuals with the desirable agronomic traits of Georgia Green and the high oleic and low linolenic acid traits of the donor parent GA 942004.
• Step 3: The BC₁F₁ seed was planted at the Univ. of Georgia Coastal Plain Experiment Station and the BC₁F₂ seed was harvested.
• Step 4: The F₂ seed was space planted and individual plants were selected based on plant and seed characteristics. The BC₁ designation used by the authors to precede the F designation of every generation will be left off in the following description so that the generation designations will be the same as used for other population types, such as a single cross or three-way cross.
• Step 5: The selected F₂ plants were grown as F_2:3 lines in two-row plots. Lines with desirable traits were selected and one or more individual F₃ plants with desirable traits were selected from within the selected lines.
• Step 6: The selected F₃ plants were grown as F_3:4 lines in two-row plots and one or more desirable F₄ plants were harvested individually from selected lines. The F_3:4 lines that traced to single plant selections from the same F_2:3 row the previous season would be grown in adjacent rows and referred to as a family. In the standard pedigree method, the breeder would first select the best families. Second, the best F_3:4 line(s) within the selected families would be identified. Third, the best individual F₄ plants within the selected lines would be harvested and threshed individually.
• Step 7: The selected F₄ plants were grown as F_4:5 lines in two-row plots. The best lines were selected and threshed in bulk to obtain F_4:6 seed. As in the step 6, the breeder would select the best families, followed by selection of the best line(s) within the selected families.
• Step 8: The F₄-derived lines were evaluated in replicated tests for three years. Georgia 09B was a F₄-derived cultivar.

Example 5: Early-generation testing of selfed lines

The developers of ‘Barlow’ hard red spring wheat described their inbreeding procedure as a combination of the modified bulk and pedigree methods. The process also involved early generation testing of selfed lines [JPR 5:62-67 (2011)]. This example is a good illustration of how breeders use combinations of methods to advance a population.

• Step 1: The cross of the line ND 744 to the line ND 721 was made in the fall greenhouse at North Dakota State University in Fargo, ND.
• Step 2: The F₁ seeds of the single-cross population were grown in the greenhouse in the spring of 2000.
• Step 3: The F₂ population was planted in the field during the summer of 2000. Spikes were harvested from 200 plants selected for their reaction to leaf rust, plant vigor, height, and earliness. Twenty of the spikes were selected and threshed individually. Selection among individual F₂ plants is the first step in the pedigree method.
• Step 4: The 20 F₃ headrows (F_2:3 lines) were grown in Christchurch, New Zealand, during the winter of 2000-2001. Desirable lines were selected and each line was threshed in bulk to obtain F_2:4 seed. In the typical pedigree method, one or more individual F₃ plants would have been harvested from the selected lines. The breeders used the term “modified bulk” because they harvested each F_2:3 line in bulk, instead of harvesting individual plants.
• Step 5: The selected F_2:4 lines were planted in a yield test during the summer of 2001. A spike from 10 individual F₄ plants were harvested from the F_2:4 lines with desirable yield and other agronomic traits. The evaluation of the F_2:4 lines in replicated tests for quantitatively inherited traits represented an early generation test. It is considered an early generation test because the F_2:4 lines were not adequately homozygous and homogeneous to merit release as cultivars per se. The purpose of the test was to select lines for quantitative traits based on replicated tests whose progeny had the potential to become cultivars.
• Step 6: The seed from each of the ten spikes was planted in a headrow (F_4:5 line) in New Zealand during the winter of 2001-2002. One F_4:5 line was selected and threshed in bulk. This F₄-derived line was tested in subsequent generations and became the cultivar Barlow. In the standard pedigree method, the 10 F4:5 lines that were derived from the 10 individual spikes of each selected F2:4 line the previous season would be considered a family. The breeder would select the best families, followed by selection of the best line(s) within the best families.

Example 6: Development of lines by the soybean breeding program of Iowa State University.

The following pictures taken by Walter Fehr illustrate one method used by the soybean breeding program at Iowa State University to develop lines for evaluation as potential cultivars.