Pedigree Method
Fehr, Chapter 25.
The pedigree method describes a selection protocol utilized during the inbreeding of populations of self- and cross-pollinated species in the development of homozygous lines. Superior single plants are selected in successive generations and a record is maintained of the parent-progeny relationships (which may extend to grand-parent, great-grandparent, or more ancestral generations). Selections are most frequently based on visual evaluations of high heritability traits, but any selection protocol, such as laboratory evaluation of plant chemical composition or marker assisted selection can be incorporated, if desired. It is noteworthy that most all breeders use the pedigree method at some point in their programs and it dovetails readily with other breeding protocols such as Mass, Bulk, or Topcross selection.
The term was first utilized to describe the protocol used at Svalof, Sweden in the late 1890’s whereby desirable single plants were selected from heterogeneous landrace populations. Seed from a selected plant was grown in its own row the following generation (‘plant-to-row”, “head-to-row”, “ear-to-row”). It was termed the “system of pedigree”. The resulting rows in Svalof were predominantly homogeneous because they were selected from landraces of autogamous species, whereas today the pedigree method is generally implemented in segregating generations where variation will be observed both within and among progeny rows. The Vilmorin Company in France developed the system independently, and it was termed the “Vilmorin method of selection”.
Implementation:
A very general outline is presented below; most programs have there own subtle, and not-so-subtle, alterations. In the NC State small grains program, the actual design is influenced by field equipment and the need for a simple system that can be followed by part-time high school and undergraduate workers.
The pedigree system can be implemented in any generation and continued until relatively homozygous lines are developed. It can begin in later generations of inbreeding particularly if combined with the Bulk or Mass selection methods in earlier generations of inbreeding. The Pedigree method is more labor intensive than the latter, thus breeders will advance populations using Bulk or Mass selection in early generations and implement the Pedigree after having culled obviously inferior populations. When implemented in later generations there will be less variation within rows and more variation between rows which should enhance selection efficiency. This is where the tedious work we labored through concerning variance changes upon inbreeding becomes relevant, and in our program we initiate the pedigree method by growing F3:4 lines because the additive variance among lines is one and one-half what it was in the original F2 population.
GENERALIZED PEDIGREE METHOD PROTOCOL
Season Procedure
1 Plant 500 – 5000 F2 plants.
Select individual plants (250)
2 Grow F2:3 rows (250)
Select best rows (50)
Select best plants in rows (2-3)
3 Grow F3:4 lines (123)
Select best families (20)
Select best rows in families (2)
Select best plants in rows (2)
4 Grow F4:5 lines (80)
Select best families (15)
Select best rows in families (1-2)
Select best plants in rows (1)
5 Grow F5:6 lines (20)
Select best families (15)
Harvest best rows in bulk
6 Testing of F5:7 lines in two
locations, one rep per loc.
7 Extensive multi-environment
testing begins (F5:8).
Pull typical heads from yield plots in one replication to initiate pure seed development.
8 Continue multi-environment testing.
Grow 50-100 F8:9 head-rows of lines still in test to increase pure seed.
Season 1. The size of the F2 or S0 population can vary anywhere from 50 to 500 to 5,000 to 5 million plants! They can be planted at a low density to permit identification of individual plants. Larger population sizes would be utilized when unadapted germplasm is included in the immediate pedigree.
The ratio of individuals selected : propagated varies from about 1 : 10 to zero depending upon the number of traits being selected, the particular environment experienced during the season, and environmental influences on trait expression. Heritability is on ‘a single plant basis’, thus only highly heritable traits can be improved in this generation. Plant height and maturity are typical of highly heritable traits that would undergo selection at this juncture. Leaf pathogen development may or may not occur in this rather open canopy structure, but selection for soil-borne pathogen resistance can be effective if the pathogen is uniformly distributed throughout the nursery. Selection for grain yielding ability would likely be a waste of effort (save for elimination of obvious deformed types) due to low heritability and because plant productivity under commercial planting density may be unrelated to vigor in low density plots. (However, there was extensive work conducted by Fasoulas in Greece that proved the contrary in small grains, but the technique was too labor intensive to be utilized to any extent).
In our example, 250 individual plants are harvested and seed is kept separate. Part of the seed would be put in cold storage and the remainder utilized to plant F2:3 rows in season 2.
Season 2: The F2:3 lines are grown in single rows that are large enough to accommodate enough plants to provide an indication of the general features of the line. Selection continues on a single plant basis in this generation, but the primary emphasis is upon selection of superior rows. Heritability has been increased to ‘a single plot basis’, although the plots are smaller than one would typically use for a yield trial.
In our example, 50 rows are selected and two to three plants are selected from each superior row to give, say, 123 F3:4 lines. Part of the seed would be put in cold storage and the remainder utilized to plant rows in season 2.
In our program we have a much higher culling rate in this first ‘head-row’ generation. We generally select less than 10 percent of rows, and we select nothing from perhaps 5 percent of the populations.
Season 3: One hundred twenty three F3:4 lines are grown in rows of the same dimensions as in season 2. However, there is a family structure overlaid on the nursery in this generation. Progenies of plants that traced to the same row in season 2 are planted in adjacent rows as a family. One of the main advantages of the pedigree method results from the keeping of pedigree records. Relationships between lines are known and this permits the selection of many lines from good families or the sampling of many lines from different families, if the breeder wishes to reduce potential redundancy. In this example we have 50 families with two to three F3:4 lines per family.
Selection continues on a single plant basis in this generation, but the primary emphasis is upon selection of superior families. Heritability increases because selection is now based on two to three adjacent plots, which is closer to the size of a yield plot than the previous generation. I am not sure of the strict definition of this heritability, but one can understand that the features of the family based on a two to three row sample is more accurate than one based on a single row, or single plant.
Once the superior families are identified, then plants are selected from the best row(s) within these superior families. In our example 20 of the 50 families were selected and two plants were selected from each of the two best rows within each family to provide seed for 80 F4:5 lines. Part of the seed would be put in cold storage and the remainder utilized to plant rows in season 4.
Throughout these segregating generations, selections are generally made visually and relatively quickly without precise measurements, save visual comparisons with check plots of standard varieties and breeding lines strategically placed throughout the nursery. Breeders prefer to view as much material as possible in these early, highly variable generations, believing that their knowledge of the crop will result in significant, and relatively cheap, gains from selection. Nevertheless, much of this selection is against obviously inferior genotypes, rather than selection for the very best. The pedigree method requires the breeder be acquainted with the crop over the range of target environments and be fully aware of morphological characteristics of superior cultivars. This can only be achieved by spending lots, and lots, and lots, and lots, and lots of time in your nurseries, in nurseries belonging to other breeders and in discussions with growers and end-users.
Breeders typically do not have a set number of selections to make in each population after the F2 or S0. Some populations will be excellent and some a disappointment, despite the fact that good parents were utilized to develop both. It would be inefficient to set strict guidelines on numbers of selections to take from each population. ‘Go with the flow’ is a better rule, but one needs to remain cognizant of the resources the program has to evaluate material in succeeding generations.
We evaluate approximately 45,000 head-rows per season. Head-rows are planted one foot apart, so one trip through this part of the nursery is 8.5 miles. We conduct most of our selection during the 75 day period between March 15 and May 30. If we devoted one minute to each head-row it would take 63 eight-hour days to work this material. In reality we probably spend 15 person days in the head-row nursery during this period, so we devote less than 15 seconds to each head-row per year.
Any and all special techniques that enhance selection pressure should be emplo