Breeding Methods: Self Pollinated Crops

1

• Breeding Methods Self Pollinated Crops
• Pure Line (Recount Johannsen. 1903)
• usually no hybridization
• IPs selected from a heterogenous popul (i.e.
  genetically variable)
• PRs are grown from IP and selections made again
  if variability exist within the row
• procedure continues until homogeneity is achieved
• last phase is field testing

2

• Breeding Methods Self Pollinated Crops
• Mass Selection
• may or may not include hybridization
• make IP selections based on single, ideal or
  desirable phenotype and BULK seed
• may repeat or go directly to performance testing
• Mass Selection has 2 important functions
• rapid improvement in land-race or mixed cultivars
• maintenance of existing cultivars (sometimes
  purification)
• Many pbers of self pollinated crops believe
  that combining closely related pure lines imparts
  genetic flexibility or buffering capacity and
  so are careful to eliminate only obvious off types

3

• Breeding Methods Self Pollinated Crops
• Bulk Method
• plant a genetically variable (say F2) in a
  relatively large block
• allow natural forces/stresses to be effective or
  apply artificial stress (e.g. apply a disease
  pathogen)
• harvest and bulk
• repeat
• low cost low effort few records carry for many
  cycles
• Key Issue the selective pressures are strictly
  those associated with survival. It Does not Or
  May not follow that the traits associated with
  survival those of economic interest
• i.e. Does survival agricultural fittness????

4

• Breeding Methods Self Pollinated Crops
• Points to consider in Bulk Method (contd)
• natural selection changes gene freq. via natural
  survival
• breeder may assist nature and discard obviously
  poor types
• relieves breeder of most record keeping
• most of us treat bulks with extremely low inputs
  and low expectations BUT?????
• Breeder Hopes
• survival of competing alleles is non-random
• inferior competitors are inferior agriculturally
• morphological uniformity increases with
  generations
• steady improvement in yield (i.e. poor yielders
  contribute less to each succeeding generation

5

• Breeding Methods Self Pollinated Crops
• Pedigree Method
• most popular
• essentially a plant to row system to develop near
  pure lines
• followed by performance testing of resulting
  strains
• this method and its variants require a lot of
  record keeping
• Advantages
• if selection is effective, inferior types are
  discarded in the IP phase and before strain
  testing
• selection each season involves a different
  environment which provides for the effective
  selection of stable genotypes
• genetic relationships are known and can be used
  to maximize genetic variability among retained
  strains

6

• Breeding Methods Self Pollinated Crops
• Pedigree Method (contd)
• Disadvantages
• cant be used in environments where ?2G is not
  expressed, thus eliminating off-season nurseries
  (not a major obstacle)
• amount of record keeping
• subjective nature of IP selections, thus
  experience of the one making the IP selections is
  important
• requires more land and labor than, say, the bulk
  or mass methods

7

• Breeding Methods Self Pollinated Crops
• Pedigree Method (contd)
• Genetic Considerations
• additive genetic variability decreases within
  lines and increases among lines, assuming no
  selection
• recall the movement toward homozygosity
  following the hybridization of unlike and
  homozygous parents
• dominant genetic variability complicates pedigree
  selection
• homozygous and heterozygous individuals look
  alike and therefore you may continually select
  the heterozygote
• THUS, selection can be discontinued with
  phenotypic uniformity within a line is obtained

8
Breeding Methods Self Pollinated Crops Pedigree
Method (contd) General Outline
9

• Breeding Methods Self Pollinated Crops
• Pedigree Method (contd) Other considerations
• the number of PRs can increase relative to the
  number of initial F3 plants selected from the F2
  population
• family origin is the initial F2 plant selected,
  i.e. the initial F3 PR
• F3 F4 F5 F6 gen.
• x x x x x x
• x x x x x x x 1 row
  strain
• x x x x x x
• x x x
  x x x x Bulk within family
• 
  x x x x x x to produce strain

10
Breeding Methods Self Pollinated Crops Pedigree
Method (contd) Naming lines and
strains 1995 P1 x P2 95001 (first parental
combination of 1995) 1996 F1 95001
(no need to change identifier) 1997 F2
95001-1 through 95001-n (IP selections) 1998
F3 95001-1-1 through 95001-n-n (IP
selections) 1998 F4 95001-1-1-r1
(i.e. row 1) through 95001-n-n-rn or we
could use any number of other identifiers name
of children series and row number (i.e.,
field location) shorten to 9501 through 95300
(confounded) 0195 through 30095 (reverse
order) point is that the strain is identified
as such until Cv release
11

• Breeding Methods Self Pollinated Crops
• Single Seed Descent (modified pedigree)
• two phases of pedigree method
• development of pure lines via IPS
• selection among pure lines for desired traits
• SSD is a system to rapidly develop pure lines
  followed by selection among those pure lines (2
  generations/yr of seg. gen.)
• SSD is especially useful in developing inbred or
  pure lines for environments different that those
  in which the segregating populations, e.g., F2
  through F4, would be grown. A good example would
  be in the development of inbreds for hybrids.
• there are 3 basic types of SSD
• 1. single seed 2. single hill 3. multiple
  seed

12

• Breeding Methods Self Pollinated Crops
• Single Seed Descent (modified pedigree)
• single seed (SSD)
• from each plant in a segregating population, e.g.
  F2, one selfed seed is obtained and bulked
  within parental population
• 2. single hill (SHD)
• several segregating plants, e.g. F2, are grown in
  a hill. Selfed seed are harvested from each
  plant in the hill and used to establish a hill
  the following generation THUS hill identity must
  be maintained
• 3. multiple seed (MSD)
• simply means that the breeder collects gt 1 seed
  per plant, e.g.
• soybean breeders may collect a single pod rather
  than a single seed
• cotton breeders may collect a single boll rather
  than a single seed

13
Breeding Methods Self Pollinated Crops Single
Hill Descent (modified pedigree) General Outline
Note that SHD INSURES that each F2 plant is
represented each generation and for final PR
selection A hill originates from a single F2
selection and multiple seed are harvested. Part
is kept in reserve and part are planted in a
hill. At maturity, a number of seed are
harvested from each hill (F23 plants) for
planting the F24 hill and for reserve. thus
germ. and plant loss will not result in an F2
plant being lost from the population.
14
Breeding Methods Self Pollinated Crops Single
Seed Descent (modified pedigree) General Outline
Note that the outline if for one parental
combination and that selection can be practiced
in any generation!
15
Breeding Methods Self Pollinated Crops Single
Hill Descent (modified pedigree) General Outline
Note that the outline if for one parental
combination and that selection can be practiced
in any generation!
16

• Breeding Methods Self Pollinated Crops
• Single Seed Descent (modified pedigree)
• Genetic Considerations
• expected genotypic frequencies are those of an
  idealized diploid population without selection
• i.e., for a given locus, heterozygosity decreases
  at (1/2)n
• additive genetic variability among plants, i.e.
  plant to plant variability, increases at the rate
  of (1F)(additive genetic variance), where F is
  the inbreeding coefficient and is equal to 0 in
  the F2, 0.5 in the F3, 0.75 in the F4, etc.
  (Recall that the inbreeding coefficient (F) is
  the probability that the two alleles at a given
  locus are identical by descent, i.e both
  inherited from the same parent
• no natural selection (unless environment affects
  germination)
• yield potential doesnt affect SSD since 1 seed
  (or a few) represents a plant regardless of its
  yield potential

17

• Breeding Methods Self Pollinated Crops
• Single Seed Descent (modified pedigree)
• Genetic Considerations
• Recall that the inbreeding coefficient (F) is the
  probability that the two alleles at a given locus
  are identical by descent, i.e both inherited from
  the same parent
• The F2 S0 thus the inbreeding coefficient 0
• I-1 A1 A1 (cross with A2 A2)
• II-1 A1 A2 II-2 A1 A2 II-3 A1 A2
• III-1 A1 A1 III-2 A1 A1
• (inbred) (not inbred)

A1A2 x a1a2 A1a1 A1a2
A2a1 A2a2
18

• Breeding Methods Self Pollinated Crops
• Single Seed Descent (modified pedigree)
• General advantages of SSD, SHD, MSD
• easy to maintain populations
• no natural selection pressure
• well suited to greenhouse or winter nursery
  advancement of gen.
• General disadvantages of SSD, SHD, and MSD
• artificial selection is based on pure line or
  individual plants and not on the progeny
  performance. THUS there is no accumulation of
  desirable progeny.
• natural selection CAN NOT influence the
  population in a positive way

19

• Breeding Methods Self Pollinated Crops
• Single Seed Descent (modified pedigree)
• Advantages of SSD per se
• requires less time and land than SHD or MSD
• max. genetic variability within population since
  every plant traces to a different F2 plant
• Disadvantages of SSD per se
• every F2 plant may not be advanced due to
  germination failure
• must adjust size of populations for germination
  percent
• requires more time at harvest than MSD because
  you must obtain one sample to plant the following
  generation and one reserve

20

• Breeding Methods Self Pollinated Crops
• Single Seed Descent (modified pedigree)
• Advantages of SHD per se
• every plant traces to a different F2 and thus
  variability is maximized
• Disadvantages of SHD per se
• requires more time at planting and harvest
• requires more land than SSD or MSD
• requires more record keeping (e.g.
  identification of individual hills rather than
  simply a bulk within each parental combination)

21
Breeding Methods Self Pollinated Crops Single
Seed Descent (modified pedigree) Epilogue Must
be sure and harvest seed to plant the following
generation AND seed to to hold in reserve in case
of crop failure. This is an issue in SSD and SHD
but usually less so in MSD.
22

• Breeding Methods Self Pollinated Crops
• Pure line
• Mass
• Bulk
• Pedigree
• Modified pedigree
• Backcross
• same form whether self or cross pollinated
  species
• only difference is pollination control
• with backcross we approach homozygosity at the
  same rate as with selfing
• goal is to move 1 to a few traits from a donor
  parent (deficient in other traits) to a recurrent
  parent (deficient in the trait of interest)
• ISSUE Genetic Gain

23
Backcross in dominant trait (B1) P1 (B0B0)
x P2 (B1B1) (RP??) F1 (B0B1) (phenotype
??) backcross to P1 (B0B0) x B0B1
BC1F1 5050 B0B0 x B0B1 B1 is dominant thus
we can see it and so BC P1 (B0B0) x B0B1
BC2F1 5050 B0B0 x B0B1 continue
backcrossing B0B1 to RP until RP phenotype
recovered and THEN ?? Simple!!!!
24
Backcross in recessive trait (B0) P1
(B0B0) x P2 (B1B1) (RP??) F1 (B0B1)
(phenotype ??) (can I make a backcross
here?) backcross to P2 (B1B1) x B0B1
BC1F1 5050 B1B1 x B0B1 B1 is dominant
thus we CAN NOT see the B0 (phenotype) so what
next ??? Self the BC1F1 ? BC1F2 .25 B0B0
.5 B0B1 .25 B1B1 Identify the recessive
phenotype, BC1F2 (B0B0) x P2 (B1B1)
BC2F1 (B0B1) (can I backcross this gen?)
25
Backcross in recessive trait (B0) Time of
Expression P1 (B0B0) x P2 (B1B1)
(RP??) F1 (B0B1) (phenotype ??) (can I make
a backcross here?) backcross to P2 (B1B1) x B0B1
BC1F1 5050 B1B1 x B0B1 B1 is dominant
thus we CAN NOT see the B0 (phenotype) so
what next ??? Self the BC1F1 ? BC1F2 .25
B0B0 .5 B0B1 .25 B1B1 Identify the
recessive phenotype, BUT What if expression of
the B0 allele occurs after pollination?? Must
wait until the BC1F3 to make the next
backcross BC1F3 (B0B0) x P2 (B1B1) BC2F1
(B0B1) Equivalent to the F1 Thus Repeat cycle
26
Backcross How Many Goal backcross in a dominant
allele (A) AA x aa F1 Aa backcross to
aa BC1F1 Aa aa identify Aa and backcross
to aa BC2F1 Aa aa repeat
BCnF1 self progeny row and select for AA So
in every generation, the A allele occurs at a
freq. of .5 or .25 until selfing and selection
But how rapidly to we approach the recurrent
parent for all other genes??? (next slide)
27
Backcross How Many Goal backcross in a dominant
allele (A) AA (donor parent) x aa (recurrent
parent) thus the goal is to have a recurrent
parent that is AA but unchanged for ALL other
alleles e.g. consider that the RP is aaBB and
the donor is AAbb so relative to the B locus bb
x BB F1 Bb and backcrossed to RP
(BB) BC1F1 5050 BBBb or 75 of its alleles
B and 50 of the progeny are homoz BB in
making the next cross there is an equal
probability that either genotype will be used,
therefore
28
Backcross How Many Goal BC in a dominant allele
(A) and considering B gene (contd) BC2F1
population will have the following
distribution .75 BB .25 Bb empirical
derivation BB x bb F1 Bb x RP (BB)
BC1F1 BB Bb x RP (BB) BC2F1 BB
BB BB BB BB BB Bb Bb 87.5
of the alleles are B AND 75 of the plants are
homoz. BB (2m 1) / 2mn where mgen. of
selfing/BC and n of genes (22 1) / 221 ¾
0.75 prob. of B being homozygous BB
29
Backcross How Many Movement toward homozygosity
(2m 1) / 2mn
30
Look at the effects of selfing without selection
another way Consider the expansion of 1 - (2m
1)n where mgen. of selfing5 n
of genes5 1/32 - 31/325 and drop
denominator and change sign 1 315
(remember that we are reducing heteroz so heteroz
1) (15) (5)(14)(31) 10(13)(312)
10(12)(313) 5(1)(314) (315) No. of
plants No. heteroz. loci No. homoz loci
1 5 0 155 4 1
9,610 3 2 297,910 2 3
4,617,605 1 4 28,629,151 0 5 Thus the
homozygosity 28,629,151 / 33,554,432 85.3
31
Backcross Effect of linkage What if an
undesirable gene is linked to the target
gene? e.g. Q is dominant for resistance to
reallybad bug but is linked to a recessive gene,
r, for susceptibility to nasty blight such
that Recurrent parent is qqRR (RP) and our donor
parent is QQrr (DR) qqRR x QQrr and our
desirable type is QQRR) The only way to get QQRR
is from a crossover in the heterozygous
QqRr. qqRR x QQrr F1 QqRr x qqRR QqRr
gametes QR, Qr, qR, and qr but since Q and R
genes are linked lets use at 10 map units or
10 cM such that qR and and Qr occur at a
freq. of .45 each and QR and qr at a freq. of
.05 each.
32
Backcross Effect of linkage F1 QqRr x
qqRR QqRr gametes QR, Qr, qR, and qr qqRR
gamete qR REMEMBER qqRR RP AND we
want to eliminate q (replace with Q from the
DP) without eliminating R but the QR gamete
occurs at only 5 probability in each backcross
generation. So, backcrossing actually
enhances the probability of accomplishing this
because we get the heterozygous QqRr in multiple
generations and the opportunity to select for
QQRR WHEREAS in a single cross with selfing we
get only one (1) chance OR at least fewer chances
because selfing reduces the percentage of
heterozygous individuals each generation THUS
reducing the possibility of crossovers.
33
Backcross Effect of linkage F1 QqRr x
qqRR Probability of eliminating an undesirable
gene linked to a desirable 1 (1- p)m1
(Allard, 1960) Where p map units (or cM or
recombination fraction) and m number of
backcrosses Probability the the
Undesirable Recombination Gene will be
eliminated fraction 5 Backcrosses
Selfing 0.50 0.98 0.50
0.20 0.74 0.20 0.10 0.47
0.10 0.02 0.11 0.02
0.01 0.06 0.01 0.001 0.006
0.001 Selection is practiced for Q and not r
34

• Backcross Epilogue
• Limited use of BC to create a population for
  selection that fosters wider genetic variance
  and modest introgression is a separate issue
  than a repeated BC to derive a new cultivar
• Jensen suggested that a 3-way (a backcross to
  another recurrent or superior parent following
  the single cross of a desirable and an
  undesirable parent) was superior to single cross
  followed by pedigree or other selection
  methodology
• Carpenter Fehr. 1986. crossed cultivated sb
  with wild species and scored resulting
  generations (F2 and F3 (or BCF equivalent)) and
  found the following degrees of recovery
• BC0 0 BC3 22
• BC1 0 BC4 51
• BC2 2 BC5 65

These were ISH The suggestion in the article is
that recovery was lt theoretical COULD suggest
that traits measured were determined by about 10
genes each
35

• Backcross Epilogue II
• BC must be used with other, more exploratory
  procedures otherwise Gs0
• Must have a suitable recurrent parent
• of BCs to make? usually 4
• Use several RP plants! WHY?
• To incorporate gt 1 trait, use parallel programs
  and then converge
• Evaluation phase can be less stringent because
  you should already know the utility of the
  recurrent parent!