RECURRENT SELEC'flON AND iNBREEDNG
FOR MATURITY {N MAEZE

Thai: for flu 009m 9? Ph. D.
MECHIGAN STATE UNIVERSITY
Alain Francois Corcos

1960

W7"li7/Sllls/7/7f/7li77/7/Z/f7777M

’
i
This is to certifg that the ‘ r
!

thesis entitled

RECURRENT SELECTION AND INBREEDING
FOR MATURITY IN MAlZE

presented by

ALAIN FRANCOIS CORCOS

has been accepted towards fulfillment
of the requirements for

Ph

.D. degreein Farm Creps

Major professor

0-169

 

 

 

LIBRARY

Michigan State
University

 

 

'LASRJ

 

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from

 

LIBRARIES
w your record. FINES Will
be charged if book is
returned after the date
.§§r§_ stamped below.
22 '5/ :3;
37 K203
fittflehw

NOV 0 4 2009
ljl 2 3 "f?

 

 

 

 

 

 

"‘7'! p: ' W
H

-
t\s4\J\

.
+

/'\

‘k

{—4

!. rv'n‘

7A.J\\-/ .. ‘

 

1 ‘ A ‘\
\s' I‘ssru

p? A "’
Ital [b.LA‘

“
JA“»‘
'l D
u L C
- s 4“

o“ n.
Iii. tiL,

I

r
swag -' L.

4‘

. I“
/f\,,.

 

-.«‘L\‘t\4“

' L2

‘\ h
v I“ q
.s \ u.

I V’
r 7
J- ( b»

I q! s .7
l V
o. 44. -l* (\Au-
(~ ‘ Iru'v 1' “f
A! u.’ L4 ‘lo‘; .

 

Ai.gln } . QLPfiJLS rn Llr til

the relative efffctivrrers cf ccntirurus intrteding
with selecticn ccrnrred to tunw~cycles cf recurrert selrctitn
fcr crrn maturity as reesurcd by silking date was studird in
pcrnleticns derivrd frcm twc dcuble-crcss hybrids end fcur
single-cross nytrids.

Direct comperiscn in the nursery brtween rec rrent

and inbred series wrs difficult due to inbreeding digrersicn

which THSkfd the effectiveness cf the brccding wrttcis. Qrcl'.

in fertility ard vi;cr end dole ed maturity were a cry the
general effects cf intreceing erd rre referred tr :'5 irtrcedi.

{3 t“

’W

1TCFSl n.

'11

Inbreeding deprassicn effected the intred series -rre
tten the rrcurrent selecticn series. In recnrrrrtselectirr

crly one plant in eecn rcw was SpleCth in crder to maintain

inbreeding prrssure at a minimun. With inbreeding énl selecti'

tdznrts were selecdxwi'withi: are ewmnag rcws.
To overcome the effects cf inbreedirg deprecsicr, it

was suggested tn't the recurrent selecticns shculd be intrei

h
>1
r 4v
(—5
W
.e
A
1“
H
('1'

P‘s

selecticn ;ressure fcr "eturit;) erd cc pered
tte inbrrd selecticrs.
nursery rvsnlts indie ted that bctn “ethcis ned beer

very effective in dividing tte criyirel F; xeterizl into two

naturity grcups and that fixation cf genes was rapid. fixnticn

wee ver; rapid in a pcpuleticn frcm an ear] -x-ecrly single
crcss (la.l$3 x W25) and less rapid in a penuleticr frcz an

szrly-x—lete single cross (MSENA x LFlT).

ALAIN F. CLHCCS ass; ACT

In a corn breeding program, the ultirete objective
is to utilize inbred lines in hybrid conbination. Thus,
evaluation of lines for maturity in testcrcsses is inportant.
Cornering inbred and recurrent lines in testcrcsses is possibly
the best method to overcome the inbreeding depression effects
which masked the effectiveness of the breeding cethcds when
ccnpared in the nursery.

Testcross results indicated that tte early inbred
series were earlier than recurrent series in two of the tiree
cases, and the late inbred series were later t an the recurrent
series. However, the differences between the two series were
very small indicating that the two methods were ecually effec-
tive. Very rapid fixation of genes and equal effectiveness of
the two breeding methods indicated that few genes for maturity
were involved.

Lines that were both earlier and later in maturity
than the parental lines of the criginal crosses were obtained
in both recurrent and inbred series.

In all the testcross eXprrinents there was a wide
range for yield among the selections within a maturity group,
indicating that selection for combining ability could be

effective within the maturity groups.

ACE-ILU'NLLLJGMIJD' T3

The writer wishes to express his appreciaticn to
Dr. E. C. hossran for helpful advice during tie course
of this investigation and in preparing the manuscript.
he wants to thank Dr. J. E. Grafius fcr helpful
suggestions during the course of this study.
Acknowledgment is also due to Mrs. Korea hay
If the tabulating department for her assistance with
the IBM punched card system.
The writer appreciies the financial support of the
Michigan Certified Hybrid Seed Corn Producers Association

which made this investiration possible.

EZCUthfiT SELECTICN AND INBHEEDIKG th EATThITY IN RAISE

by

ALAIN FhAECCIS CLECCS

A TELSIS

Submitted to the School of braduate Studies of Michigan
State University cf Agriculture and Applied S ience
in partial fulfillment of the requirements

for the degree of

DLCTCE Lb PHILOSOPHY

Department of Farm Crops

1960

Introduction

heview of Literature . . . . . . . . . . . . . . .

Faterial and Methods . . . . . . . . . . . . . . .

Results

I\:ursery HeSUlts O O O O O O O O I O O O O I O O I I

TBStCI‘CSS 885111123 0 o I o o o o o o o 0

Discussion
Summary .

Literature Cited . . . . . . . . .

Appendix

Days from planting to silking . . . . . . .

General considerations . . . . . . . .

Comparisons between breeding groups in
the early-maturity series . . . . . .

Comparisons between breeding groups in
the late-maturity series . . . . . . .

Moisture content at harvest . . . . . . . .

General considerations . . . . . . . .

Comparisons between breeding groups in
tkle early-Thaturity series 0 o o o o 0

Comparisons between breeding groups in
the late-maturity series . . . . . . .

General considerations . . . . . . . .
1":atllrity O I O O O O O O l O O O O O 0

Differences between breeding groups .

Yield Results . . . . . . . . . . . . . . .

O O O O O O O O O O O O I O O O O O O

18

37
1+5
1+5

“9

A

d

\J
v
1

7‘
Li

INIhLDLCTICh

In northern areas of the United States, pre-mature
killing frosts are a hazard to successful corn production.
Corn is considered mature when the moisture content of the
grain is about 35%. Losses in yield and ouality result when
frost occurs before maturity. The best way to avoid these
losses is to grow early-maturing hybrids that will consistently
nature ahead of frosts. Due to early frost in the North,
these hybrids will usually yield as well as or better than
late-maturing hybrids. Early-maturing hybrids are ready
for harvest earlier in the fall when weather is usually more
favorable for harvest aniharvest losses are lower. Mature
corn makes better ouality feed than immature corr. Lower
moisture content permits safer storage.

Developing inbred lines that transmit early maturity,
high combining ability, and lodging resistance in double-cross
hybrids is a major objective of corn breeding programs in
the northern areas of the United States. harly-maturing
inbred lines as a group generally possess a lower level of
conbining ability and lodging resistance than later-maturing
inbreds. Inbreeding and selection in heterozygous populations
of early X lete parents is an effective rethod to improve
combining ability and lodging resistance of early-maturing
inbreds.

Inbreeding with selection is necessary in corn irprove-

ment to fix desirable characteristics in a line, so thrt it

2
will transmit these characteristics in hybrid combinations each
time seed is produced. hapid fixation of geres occurs with
inbreeding. This seriously restricts the effectiveness of
selection that can be practiced. hecurrent srlection is a
breeding schene that offers some Opportunity to minimize this
disadvantage. Plants from a heterozygous source are evaluated
for the desired attribute. Selfed seed of the small sample
exhibiting superior performance is then planted ezr-to-row.
All possible intercrosses are then made by hanc, or sore type
of bulk pollination may be used. The intercrossod population
serves as the source material for the next cycle of selection.
Opportunity is afforded for cumulative selection.

Four types of recurrent selection are recognired: (l)
recurrent selection for characteristics that can be evalurted
accurately phenotypically as described above without trstcrrsses
reouired for (2), (3), and (h); (2) recurrent selection for
general combining ability; (3) recurrent selection for srecific
coabining ability; and (H) reciprocal recurrent selection using
two heterozygous populations with reciprocal self-pollination and
testcrossing, A x B and B x A.

With corn, recurrent selection has been applied for oil
content, tryptoptan and lysine amino-acid contert in grain,
disease and insect resistance, popping expansion of popcorn,
and Specific combining ability (7, 18, 19, 20, 23). The method
has been applied to sweet clover (12, l3, 1%), to tirsfoot trefoil
(21) and cotton (6).

Silking date is a good criterion of corn maturity, is

easily determined, and can be evaluated phenotypically without

testcrosses in a recurrent selection scheme. Selected plants

3
can be pcllinated the sane generation the characteristic is
measured.

The objective of this study was to determine the relative
effectiveness of continuous inbreeding with selection compared to
three cycles of recurrent selection for corn naturity, as
measured by silking drte. Theoretically, three generations
of selfing in a F2 porulation should reduce heterozygosity
to 6.25% while three cycles of recurrent selection should
leave 35% heterozygosity.

This value 35% is obtained in the following manrer:

The gametes produced by N monoecicus individuals unite
wholly at random. Therefore, the gametes have a chance l/N
of coming from the same individual and of h- l/h of coming
from different individuals. The loss in heterozygosity is
then l/EN or, since ten plants were used in this study, 1/20
per generation. After three cycles of recurrent selection,

it would be possible to continue a selection program.

FEVJLW UF LITLFATUEE

Early maturity in corn is frecuently dominant in crosses
between early and late inbred lines (5). The Fl hybrid was
earlier than either of the two parental inbred lines which
were approximately eoual in maturity (28). Jones (15) in-
vestigated the inheritance of corn maturity in six crosses
of early-k-late inbred lines. Either complete phenotypic
dominance or slight heterosis for earliness was indicated in
all crosses. Complete genie dominance for early silking,
partial-to-complete genie dominance for lower ear moisture
at a uniform time from planting, and variations from none to
complete genie dominance for lower ear moisture fifty days
after silking were indicated for the different crosses.

Jones (15) was not able to conclude whether gene action
was following either arithmetic or geometric scheme-. he
suggested that both types of gene action might be involved.
Using the formula for calculating gene number given to Burton
by Wright (1,2), the calculated number of genes was 5 to
19 for silking date, 2 to 11 for moisture content of ears
harvested at a uniform period from planting, and l to 5% for
moisture content of ears harvested fifty days after silking
depending on the particular cross involved. Yang (2E , in a
study of the nature of genes controlling hybrid vigor as
they affect silking time in corn, concluded that there were
probably 2 or 3 pairs of genes involved. These genes were

independently inherited with effects of comparable magnitude.

heritability values which denote additive genetic
Variance in per cent of the total V<riance are useful to
the plant breeder because they indicate the extent to which
desired characters are transmitted from superior plants to
their progenies. The additive genetic variance reflects the
degree to which the progeny are likely to resemble the parents.
Jones (l5) calculated heritability for silking date and
moisture content according to Wright's formula (27); which
includes both additive and non-additive v riance:

Variance F2 - Variance Fl

 

Variance F2

heritability values ranged from 11 to #1 per cent for silking
date, 36 to 58 per cent for moisture content of cars har-
vested at a uniform period from planting, and 22 to 83 per
cent for moisture content of ears harvested fifty days after
silking for the various crosses studied.

Warner (28) obtained a heritability value of 32 per cent
using the formula:

(i) D
Heritability-

'w'h 9 re
VFa
C
(i) D = the additive genetic component of variance of F2 and
VF2 2 total within variance of F2
and (a) D : 2(VF2) - (val / veg) where

V51 and VB2 are the total within variance of the backcrosses

of the F1 to the respective parents.

6

The general procedure employed in recurrent selection,
which has been effective in improving agronomic characters
in corn, cotton and forage creps, was an outgrowth of the
suggestion made by Jenkins in 1935 (9) to select among
early generation inbreds on the basis of general combining
ability. The essential steps were outlined by him in lQMO
(10):

(l) the isolation of one generation selfed lines; (?)
testing of these lines in top crosses for yield ard other
characters to determine the relative endowments with
respect to genes affecting these characters; and (3)
intercrossing of the better endowed selfed lines to pro-
duce a synthetic variety; (H) repetition of the above
process at intervals after each synthetic varietyhas
had a generation or two of mixing, possibly with inclusion
of lines from unrelated sources.

Sprague and Brimhall (23) showed that one cycle of
recurrent selection for yielding ability resulted in a mean
increase of 7 bushels per acre relative to the tester.
Lonnquist (18) produced a high and low yield synthetic, using
51 lines from Krug yellow dent, an Open pollinated variety.

After two generations of random mating, 5 plants were selfed

o
and crossed to (Wf9 x Nlh) as the tester. The mean of the
testcrosses from the high yielding synthetic was 2.7 bushels
less than the tester and 6.H bushels more than the testcrosses
of the 31 lines used to form the synthetic. Testcrosses

from the low yielding synthetic averaged 1%.5 bushels less
than tester. Thus after one cycle, two distinct groups with

respect to combining ability had been separated from the

original Krug pogulation. McGill and Lonnouist (20) crossed

7
80 plants from Krug and three synthetic varieties developed
from Krug by two cycles of recurrent selection to Ni? x Mlh
as a tester. They concluded that recurrent selection had
been effective in modifying combining ability and that the
high yielding synthetics were better sources of new lines
than Krug. In a more recent paper (19), they showed that
advanced generations of the synthetics maintained their
improved productivity through normal mass selection procedures.
Visual selection of better plarts from synthetic 2 to synthetic
h resulted in a slight improvement in yield; but the plants
were later in mrturity.

A modification of recurrent selection for octbining
ability was suggested by hull (7), who advanced reasons for
believing that genes conditioning corn yield exhibit over-
dominance. Under this assumption the heterozygcte is superior to
the best homozygote at loci at which ovrrdominance is Op‘rative.
Hull's method, called recurrent selection for specific
combining ability, involved the use of a stable inbred line
as the tester. Sprague (2“) reported that "unpublished data
indicate that in each of two open pollinated varieties,
using inbred hy as tester, yield increases of approxinately
5 bushels per acre were obtained vith a single cycle of
selection."

Comstock, Robinson, and Harvey (3) proposed a breeding
procedure to make maximum use of both general and specific

combining ability regardless of overdominance. They designated

F

V

their method "reciprocal recurrent selection". Two unrelated
stocks are used as source material. so plants of a source

A are self—pollinated and crossed to plants of source B as
the tester. Selection is based upon experimental comparisons
of testcross progenies. Selected plants are intercrossed

the third year by use of selfed seed from S0 plants. An(ther
cycle of selfing and crossing is initiated the fourth year.
Plants from source B are tested against plants from source

A as the tester in the same way. From theoretical considera-
tions tiey concluded that ”reciprocal recurrent selection”
would be superior to selection for general combining ability
for loci where overdominance existed and was superior to
selection for specific combining ability for loci where
partial dominance existed.

Sprague and Brimhall (23) found that recurrent selection
for oil percentage in the corn kernel was 2.6 times more
efficient than selection luring inbreeding. Considerable
genetic variability remained in the recurrent series, while
genetic variability was exhausted after five generations of
inbreeding.

Frey et a1. (E) reported results from one cycle of
recurrent selection for low zein and high tryptophan in two
populations of intercrosses among F) progenies of hy x I l@8.
One set of F3 progenies had been selected for low ratio of
zein to total protein, and the second set had been selected

for high tryptophan content of the grain. No inprovement

9‘
was realized in the population selected for low zein—protein
ratio. An increase of 12.75 of tre mean of the F; population
was obtained in the pOpulation selected for high tryptophan
percentage.

Jenkins et al. (11) reported on experiserts designed to
measure the efficiency of recurrent selection in concentrating
genes for resistance to leaf blight of corn caused by
helninthospcriun turcicu". hine single-crosses between
resistant and susceptible lines were used. The materials
used were two-generation-backorosses to the susceptible
parent. Three cycles of recurrent selection were completed
for each of the nine progeny groups, providing a tot l of
27 comparisons. In 2% of these comparisons the differences
between successive cycles were positive indicating an increase
in resistance; in 19 out of EM, these differences were sig-
nificant. In three cases the differences were negative,
indicating an increase in susceptibility. Jerkins concluded
that two cycles of recurrent selection would be warranted in
most of the groups studied. The effectiveness of the third
cycle appeared to be related to progress in the previous
cycles. If the gain has been substantial, little progress
wrs achieved.

henderson (6) found that recurrent selection in cotton
was effective for increasing the freouercy of superior gene
combinations for length and strength of fiber, weight of

seed, and weight of fiber per amount of surface area on the

lO
seed considerably above that found in the FE' After one
cycle the freouency of plants above the average was 23 per
cent. he remarked: “. . . the fifteen intercrosses differed
greatly in relative frequency, the lowest progeny hzving no
superior plants, while the highest intercross progeny had M3
per cent of plants above the parent average in all character-
istics."

Johnson reported results of one cycle (I?) and two
cycles (13) of recurrent selection for general conbining
ability in sweet clover. In both the first and second cycles
of the population means slightly exceeded the neans of the
10 plants whose 81 progenies were used to produce the next
cycle. The open pollinated progeny test mean of the 10
selected Madrid plants was 116% of Madrid and the first—cycle
population was 121% of Madrid. The mean of the 10 plants
chosen from the first cycle as parental lines for the second
cycle was lh6% and the second-cycle population mean was 152%
of Madrid. Four successive cycles of phenotypic recurrent
selection of biennial sweet clover for first year growth were
effective in improving uniformity of growth type and plant
vigor according to Johnson and El Banna (1%) who measured the
change per cycle in terms of variance and pepulation means.
The changes were smaller for a plant characteristic with low
heritability value such as plant weight than for a plant
characteristic with high heritability value such as growth

habit. The mean plant yield in the highest vigor group was

ll
increased 88% over the original population by repeatedly
choosing and recombining ten plants in the upper portion of
the Syn. l pepulation.

Peacock and Wilsie (El) reported that one cycle of
recurrent selection for shattering resistance reduced seed
pod dehiscence of birdsfoot trefoil 17 per cent. The
variability among the synthetics produced appeared to be
great enough to allow further progress in shattering resistance

through additional cycles of recurrent selection.

MAQBRIALS AND MLTHCLS

F2 pOpulations of 500 plants each for two double-cross
hybrids and four single-cross hybrids were planted in 195%.
The two double-crosses were Michigan 518 (HQ x N13) (la 153
x W25) and Michigan 20D.(Oh 51A x Mlh) (Oh MOB x WlO). The
four parental lines for Michigan 518 were classified as
relatively early-maturing lines. The parental inbreds Ch 51A
and Elk for Michigan 20D are ckssified as medium in maturity
while Ch ROB and W10 are classified as relatively late in
maturity. The single-cross hybrids were: one single-cross
of two early inbreds (Ia 153 x W25) and three early X late
single-crosses, (RS3 x 38-11), (MSQHA x 38-11), and (RSQHA
x L317).

The 20 earliest silking plants were selected in each of
the six pOpulations. Ten of these were chosen at random and
self-pollinated for the inbreeding series. Equal amounts
of pollen from the other ten were bulked and used to pcllinate
these plants for the recurrent selection series.

The 20 latest silking plants of each pOpulaticn were
selected and divided into two groups of ten plants each.

One group of 10 was self—pollinated for the late-maturing
inbreeding series. The other group of 10 plants was inter-
crossed by bulking ecual amounts of pollen and constituted

the late-maturing recurrent selection series.

13

For each pOpulation in 1955, 50 seeds of each selected
e'r were planted ear—to-row, making a recurrent selected and
an inbred series of twenty rows each. In the inbreeding
series, the ten earliest silking plants and the ten latest
silking plants were again self-pollinated. The selected
plants in five of the populations were crossed to the single
cross (Ch H3 x A158) as a tester for combining ability. In
the recurrent selected block the earliest plant or two in
each row was selected and the pollen of ten of these plants
was bulked and applied to the silks of these ten plants. The
same procedure was followed for the late silking selections.

Due to heterotic effects, the recurrent selected series
would be expected to be earlier than the inbred series. To
compare inbreeding more effectively with recurrent selection,
10 of the early silking plants and 10 of the late silking
plants were selfed in each recurrent selected block. All
plants of the recurrent selection series were crossed to the
tester (Oh #3 X A158).

The same procedures were followed in 1956 to provide 9

breeding groups as illustrated below for each pepulaticn.

 

PS“; 3 - - - - 1956

 

 

1M

(PS)1 3 Pecurrent selected one cycle
(RS)2 = hecurrent selected two cycles
(HS)3 = hecurrent selected three cycles
(8)1 = Inbred with selection one generation
(8)2 = Inbred with selection two generations
(8)3 2 Inbred with selection three generations
ES-l = hecurrent selected one cycle, then inbred
with selection one generation
HS-l-l = Recurrent selected one cycle, then inbred
with selection two generations
(hS)2-l t Recurrent selected two cycles, then inbred

with selection one generation

These nine breeding groups for each of the six pepu-
lations were evaluated for maturity in 1957 in an experiment
with two replications of 15 plants for each ear-to-rcw. In
each population, the early and late maturity groups were
planted in separate blocks. Within each maturity group, each
of the nine breeding groups consisted of 10 ear-to-rcws of
15 plants in each replication. Some poor stands were ob-
tained in the late maturity groups of the (la 153 x N25)
pOpulation.

Silking date and moisture content of ears at harvest
were determined for each plant in two populatiors, Michigan
518 and (B53 x 38-11). For the other four populations, only

silking date for each plant was recorded. Silking date was

0.)

recorded when the silk was exposed about i inch. Aver ge
days from planting to silking, percentage of moisture content,

variance and covariance were calculated for each row.

15
Analyses of variance for each population were made for

each maturity block. Degrees of freedom for each block were as

 

 

 

follows:
Source of Variation Degrees of Freedom
Total 179
Eeplications 1
Strains 69
Groups 8
Within 9
Interaction 72
Error 89

 

 

Degrees of freedom for each breeding group:

 

 

 

Source of Variation Degrees of Freedom
Total 299
Between 9
Within (estimated from row 290
variance)
Parents n

 

16
The ultimate objective in a corn breeding progrrm is
to put inbred lines into hybrid combinations. Therefore it
is necessary to evaluate the lines for yield and maturity as.
testcrosses. 81 and RSI selections of four poyulations were
testcrossed to (Ohh3 x A158). These testcrosses were evaluated
for yield and maturity in four experiments in 1955 as follows:

Table l. hxperiment number, population, and design of the
1955 testcross experi ents.

 

 

 

preriment Number Population Design
55-902 (HS2HA x 35-11) 6 X 6 simple lattice
k replicates
55-903 (R53 x 38-11) 6 x 7 rectangular
lattice, h replicates
55-90% (MsahA x L317) 5 x 6 rectangular
lattice, H replicates
55-905 (Michigan 20D) 5 x 5 simple lattice

H replicates

 

 

The date when about half the plants had silked in each

plot was recorded.

529 RS; and 88-1 selections of 5 pOpulations were

evaluated in testcrosses to (Ohh3 x A158) in 1956 as follows:

» .

Tatle 2. Experinent number, population, and design of the
1956 testcross experiments.

 

 

 

hxperiment Number FOpulation Design
56-901 Michigan 518 E x 8 simple lattice
h replicates
56-902 Michigan 20D 7 x 7 simple lattice
h replicates
56-903 H53 x 38-11) 8 x 8 simple lattice
h replicates

Table 2. Continued

Experiment Lumber

 

56-90%
56-905

Popplation

 

(hSQHA x L317)
(RSQHA x 38-11)

17

7 x 7 simple lrttice

h rellicates

E x ‘ sirple lattice
h replicates

 

S3, 883, RS-l-l and 582-1 selections from the third

cycle of breeding were evaluated as testcrosses with {the} x

A158) in three experiments in 1957 (Table 3). Silking date

was recorded for each plant.

Table 3. theriment number, population, and design of the

1957 testcross experiments.

 

Experiment Number
57-911
57-912
57-913

Population

(h53 x 38-11)

hichigan 518

(xseuA x L317)

Desi n

10 x 10 triple lattice
3 replicates

10 x 10 triple lattice
3 replicates

9 x 9 triple lattice

3 replicates

 

 

 

18
LSCLTS
NLhSLFY hLSULTS
Days from Planting to Silking

General Considerations

 

The first cycle of breeding by either method was effective
in dividing the original F2 material into two maturity groups
(Table h). Although some strains from the early maturity
groups were as late as some strains from the late maturity
groups, no early-maturity breeding group mean was greater
than any late maturity breeding group mean. The F2 value
for each population was intermediate between early and late
maturity series.

The greatest progress towards early and late maturity
seemed to have been made with the first cycle of breeding as
stown in figures 1, 2, 3, h, 5, 6, Appendix, where values for
parental lines F1, F2, and breeding groups have been graphed.

Differences among breeding groups are highly significant
in all cases (Table 5) indicating a difference in progress
from the different breeding methods. No significant difference
within value means indicated that genetic variance was exhausted
after three cycles of breeding.

Comparison between breeding groups in the early maturity series

The mean values of each breeding group are ranked in
Table 6. Values within a bracket are not significartly dif-
ferent according to Duncan's multiple range test (u).

Since the objective was to compare the effectiveness of

continuous inbreeding (selfing) with selection and the

range and mean days from planting to silking of nine breeding groups from six

pOpulaticns.

Table H.

late Maturity

barly Maturity

 

C
(C
(1)
141

 

I
A

 

CL]
H
Ln.
[3
$1
‘r-i
,C'.
U
.H

V

U\\C‘ O‘r-I\O HOT) 0.1." O

O O O O O O O O 0
\LV ()L'Ll Cu (: Hon“ r4 (II
\0 'L\\O L\L\ L\ L\ (\CLJ C\

I O

WOWWOWOOW

o o o o o o o o a
Ow: {\I (YWH O Q) ‘L’u
i—If—IF'I H

II II II II II II II II II
oommooooo

O O
(VJHL\J\L:L1)C\\L)
[\BBBLJU L\L\L\UCJ

I I
UKOOOOLAOOLA

O O O O O l
(\IcIJ[\U\r—I’_j: O "T m
\O\O\L \C'L\\O\U\OL‘-

JLO\CI.\L\C\IE\\COO
o o o o - 0 O o o O
O ”WONCMOr—Ifw w
\O\C\O\L‘\C\O\o\oxo 3

mmo OLROL "xU\U\

"timfiC‘:CU \L; (‘10 LR

IIIIIIIIIIIIIIIII
moommooom

. .

(\BU‘WL LLILITJOW
m\OVZ(\OVS'(WO()
I I I I I I I I I
O . . . . O C O .

OVr-ICQ (7000000

InclhlnxomxoLC/‘xc
H
I-II
r-I Ir—I
(\II (\Ioflml m
mmmmmmmmm
LE mil: .1.an

CI

0,.

O
--------_----u

f"

--—-—-/’.

v

strains---i,u
2. groups--—----------_1.5

l.

LOSOD.

C
C
K
WOWOO
(L o o o o
CO-J‘\DO
H[\[\(\LIJ O)
r—I .-
C
H" U II II Q)
(U Jo
+-> (VI
1: Ln
0) H
I—I
(U Ln om
5140\NC-Cr‘1
33H}:
C
(U
9
O
o o
m C\
\L \0
II I
FI—I
L14
(I)
H
in
Q
(U
OLD
7-4
.1:
C.)
°r-I
2;

A‘—

x D8-11

 

 

H53.

(\J NJ.) “I. ((WC'J' H (J
o o o o o

.7‘\

ORCU C\-T J Li) nu‘x ,-—4 H

[\UE\LD LOLA) J.) 0003

/

OIDOOLRWLROO O
. . C C O O O

\C OxNQ C) 004nm L\
H

r-Ir—Ir-‘I
I II II II II II II II II

OOmOQOmtn
GOLD low 3 LL'L'L’JO JILL)
I I I I
ONWQOWOOm
O O I I O O O O O
\C FIONA) Q L‘\ L’ -- the”)
L\[\\£>L\LDL\L\COC\

\O (Mr—«us anIno (\uo

\L-LD 3 Ln ; L D'WLn {\U I")

\O\D\C\O\O\L\C\L\L\p

U\OU\OOOU\U\U\

.3‘ (\mfi \L. WL\O\F'W

II II II II II II II II II
Omomomomm

O’\O\L c-Omo-(‘utx
£m\(noo\¢\oouc
II I II I II
U\U\LI’\U\C)U\U\C) O
O O O O O O O O O
,1 Mo on: O N m:
\oLouamnoxcuawxc

H
r-II
r—I Ir-I
(\II moon” on
mmmoimmmovm
.I: 4;: 11:13:13

+.H

strains-—-1
groups

1.

1.8.3.

Ln
I

(TI

”I
C.

76.0
Late Maturity

6F‘ .0

H53 =
36-11 =

72.0
71+.

,.
C

F.
barly Maturity

R53 x 38.11 F1 =

(Continued)

Table h.

 

   

£4 01%“ COD ”)0 O O (I L\ \l" (1: (\I 0‘ mm".
(C o o O o o O o o o o 0 CL. 0 o o o
w oom- (gai)c LAC/x; o'xp rv". 0,, (\Qr-«r'.
:3. Pull LI) LL) CO .100.) O\ C I I: J.) 90 Li; LL)
I
I
I
I
I
I
I
Omommommr—I M 0000 mmoo
ooooooooo o .000 a...
:r-I(\(\J\L.\LJO\CD 1n (\ICT\\L (u mmmox
H H H I E\[\C‘JJO‘\
I 0)
II I II II II II II II II I Q) II II II II
I C
OOmOOmOLnH I *4
Q) 0.0000000 I t—IIIHIIII
' O O»: V.» moo \ou‘) L\ I
i; CDIANIJLVCWD O\O“\O‘x I H
LC I I I I I I I I I CC
(I; QU\U\U\U\U\U\OO +3 (Q
o o o o a o o 0 o C O o o o o
\(;L\L‘\MJL) (\Jr—IO\O\ (1 3 LL) (:2P'I
BBB®B®LXDLD® LJ< 'O WBLLDL'O
(If—IF‘I-CH
LuELnQB
c: (V’I\C\C'(\JJNL\I—I‘Y(fl 0: C (3(me
(n o o o o o o o o o o o (I o o o a
0) O ”‘Imr—I r-I L‘~-OO O O O m IL or Mucu H
5.2 BBQBUQQB L\ : 2.. [\L\ L\
I
I
I
I
I
LnOOLanLnOOLfl U\ O O mmOm
O o o o o o o o O o o o o o I o
mxoqwirmmoom J H (\J in: on:
I [\ [\
IIIIIIIIIIIIIIIII I
I II II II II
I
I
LnOLnOOInInQUN (D II I
o o o o o o o o o C U)
(V‘Iwr—{U\,_3'O\OL{\(U «an. H (\I
BBBBBfiBBB (1323 £14 In
I I I I I I I I I I-IO
OOWLHLNQ {ALDO +9 LI
0 o o o o o o o 0 (am C]
LLOt~-;uO*:U\L\L\ O
\CB\O\O\O\O\L\O\D ’ ° CV
r-ICM
C
CO
H 0 Cu)
H I Q "—4
r—I Ir—I - 1:
(\II NMOJI m U) 0 WI
UlUlUlUlCDE/‘UDKDLO 0 H O’DUlUlt‘
III LEI. .ISIIZSIS I—I Z I.

*‘

 

NSQHA x 18-11

 

C)
C)
(\I
C2
C0
L41
.fi
£1
(J
H
22

(Continued)

Table h.

Late Maturity

iaturity

Early b

Mean

Range

r‘
L:

“ea

I
I

 

18-11

_‘

M ERA x

 

entradra-4

O O O C

\L‘ \; ~L) ‘-LJ A»
U) UDKDQ) v.)

0001“

r-IU\O\oL) (V)
H

II I II II II
OLnlflLflO

O O O C .
,jO‘fY‘ILMJ'
O‘CIJCVO‘ON
I I I I I
OWWOCD
. C O C .
(Y‘IJJJH
CIDLLLLDLUKL)

HO\(\I (fir-1

r-I O~ (\1 D» O Q
(\w (\\0 L‘ [\

O O mlnC)

meO
H
II« III"

MOOLnO

O O O O O
JflIfYHutv
L\L\DC\L\

 

strains-—-?.h

—----—-—5.o

.—

l.

L.S.U.

----2.u

---2.o

grozps

2.

Parental lines

seuA x 3a-11 F1 78.0

I
A

I

7702

 

Mean

C
m
9

‘1
A

 

Vi
w
3
K
(*1
UV
rI

m
H

 

21

.1 Huaxo 0 mefi C?) w

0 O O O O O O O o O
raxfi-;:‘u\_3-u'\o;;;.\g U\
L\[\[\ f\ L\[\LU L\£\ L\

---—----€.o

mmmmmmomo
O

\(JLIJU\\CI\L,\LICVJ (\J t
H

r1
C

Ommoommom

o o o o o o o o o

WNNOCwMCPQ;

L\C\(\(,UC\I.\UOL\L\

I I I I I I I I I
m _

"'“"‘"“"/

6
7
73.

\CDBBB

62.8
6h.5
62.3
6A.S
6%.?
61.2
62.8
63.7
a .1
o3.5
--—-l.3

3

OOOU‘\OU\U\OU\

.3" \O "13' \CA \L 51“

u

,,I "II" III“ N

OUNOUNOLAIAOIA
O O O O O O 0 O O
mwi (\Lu com (”-242
\C\O\L\CI\C\L’\£\O\C
I I I I I I I I I
OLnQOOQLlOO
O O O O O O 0 O O
HOH’Y‘I(\JQ\OQF“)
\O\L\£\Q\LU\\O\C\O

strains---1.
groups

1.

2
l

m
U)

RS

8
R3
R8-

52
R33
RSg-l
RS-l-l
L.S.D.

(Continued)

Table h.

Parental lines

Ia. 153 X W25 Fl

F

iaturity

Late I

O

Maturity

Early

Nean

 

:2
co
(L2
2;

 

N
H
m
,4

K
<1:
J
(\I
to

 

\C (Y‘\O _+ mouoa (\J C

I
O I O O O

L! \L
cowmwuloswc U

OCDVVDCDVunU\Q

O O O O O O I O O

t\ric%DClmMO\CU\
r—I I—Ir-I

II I II II II II II II II
QLn.LnoOU\OOO

0 O O O
mO\Cr—~Ir*"I[\O\r—I:
CDOxwLIJO‘~CU.OOO\
I I I I I I I I I
(.anOIleanUVQ
O O O O O O O O O
\L O L\r—I(Y‘.r-<l(‘u,;t O\
C\BL\(\~CC®CUU$<D

----—-—--e.3

"I
C

LnLr‘mej Inf) 0 O
O O O O O O O O O
U'\_j-,;r nIL\U\mr—IL’D
r-I

OOIDOOOOLAC‘»
- o o o o o o a '
C‘swj‘mu’ FIL1)L1,U\
(\[\BL\L\L\E\L\L\
I I I I I I I I I
mine Lo U\ .7) C: mm
0 I O O O O O O O
c—IrflOr—Itocqlnmln
L\L\£\L\\U\O[\\O\O

strains---5.

br.

1.

3

2

S3
LOSOD.

o
0*.3' URI L\ ( .I T («J

groups

r3
C

Parental lines

RSZHA

.17

an,
(
v

I

hSEHA x L

h)
BI

97.0

L317

81.0

a. «J

 

 

 

 

 

$.wH OHHH
w.wH mmHH
n.3w mHm
«ow.mOH mOmH
..m.am duom
O.n: n:
mma:
m.u HJQ
§§C.ww waH
m.;m Hmw
9*A.QH© wmw:
eew.0m Hmom
o.qw cm
owns
m.” amoa
®.CH mom
w.u: Omj-
em.wHJ mwwm
{*M.um JON:
0.4H :H
mmoo
+.n.x .mwm
monopquufioqug mama

mcfiooopo potpowwfic CH mcfixHHm ou muaucqu song mama no“ ooCmfiac> yo mmmszc<

 

—————-—- —- -————-—v————-——

I:
I‘

*.
O

\Or—I\C

l

(u: [\

Q

0
\L L’KL

A’UW\

.Hw>oH pcmonoa
_ .HoboH pooooou

 

 

 

«.4 m :H:
“ 0mm
0:
_ AVMVM
ee©.OH DUO
0.3a , :H
moja
H.m _ He:
mmm
wma
_ mom -I
§§%.HH ONOH
o.w w
a woma
m.m . mmw
. mmm
m:
mum
§§m.© mmw
0.0 m o
_ mmafi
.n.2 . .c
mwfiopo Wufiooopm wmpmm

 

me

O

i

QDO\(\I

a)0«m

[‘\

 

.1)

H map
m out
Ow

um unmcfiwflu-
am acmtacac

HQ *1
Hm 9
Logan
cofiuompoucH
.wnuHR
. Um “hm
mCHmhum
Hume
HmuOB

\

Lu

masonm

Orr-IO\
f\ LU
H

 

Ha-mm x mmm

popam
Cprompmch
cacaaz
.amtom

mcamppm

a .aaom

fiance

ow

mqsoum

 

‘r‘

gonad
nofluomaoucH
sagas:
.mmoom
. mcaaapm
a .qum
HmpOH

3.
CI)

museum

 

mam ammHaofia

 

mopzom

I14
(3

mommoao Homewamno mam Eopm meson»

.m prma

 

 

H.MH M . HAHH*
m.m . OQH h
u.wo . oaw .
orO.QmHm @JWH
..w.mm mmOm
0.0: o: ,_
53m~ ’
*
m.mfi mjmam
m.m- me M
0.:wa :Hma M
..@.m;: mjmm w
§§J.Hm QKmJH
o.mm om ”
mu:m_
m.w mmm w
§§O.NH Hmw M
m.HH mom m
§§B.®H _ OmmH m
..®.omw wmmww
o.w m w W
m m3mm“
_
.m.y .m.m
mmsopo mcfiummpm mung

m.m
§§m.m
:.m
«1H.mm
§§m.w
o.m
0.®
u.w
m.mm
o¢m.mw
.*m.mfi
O.wm
®.w
.rm.m
H.w
«*H.m®
opm.wd
O.HH

mMSQLU

co:
Hm

Ham

 

aw:
4 0mm
r mom

Him
mm

Hmm

-.—.—_..—.. _____‘».;__ —-
v

.HmpwH
.Ho>ma

m:m

Jaw

h")

CI) J./ \I-
N N (0
(D

(\J H

0mm

mmfia
HH
maid

.mwom

wcflnmwpm magma

pews pun H wzu um ugmuflgficwfim .o

#coo pwu m man an

b“- -_.-_‘-——————-—

 

A

_

CO C)‘\ N
L\

m

(I)C‘\C\J

pcmofiwflawflm .

aw Aopua
coauomgmch
:figufiz
mmzogm .nwwpm
mm mcfimuum
H .Haac
05H HmuOB

 

9%.xmwfimH

Hopgm
cofipompwucH
cfigufi3
mazopo .nmwpm
ow mafimhpm
H Hamm
aha Hmuoe

(3K
K \‘

 

mfimg x <1mm2

0w pogpm
Sufiwompmch
cfigufiZ
masopa .cmmpm
mm mcfimppm
H .Hamm
mha Hmuoa

 

Hanmfljx <1mmz

 

ChSO

rv
» ‘
0
Q
0/

.\.
C

"Um cfipcoov .m mfipme

25
effectiveness of recurrent selection, breeding groups witLin
a cycle can be compared (Table 6a).

In 9 of 18 cases, recurrent selected breeding groups were
earlier than the inbred groups. After three cycles of breeding,
in 2 of 6 cases, the recurrent selected groups were earlier
than the inbred groups. However, the differences between the
recurrent selected and inbred groups were small. These results
indicate that there is very little difference betwren the two
breeding methods.

Table 6A. Comparison for early maturity between intred and
recurrent selected groups within a breeding cycle.

 

 

 

 

 

 

Pcpulaticn First Cycle Second Cycle Third Cycle

yichigan 51B [‘ 8‘) BS 82 > R82 1 83 RSj
E 63.8 6o.u 62.7 60.6 :63.0 60.2

Michigan 200 i s > as ’ s? = hsg s3 =es3
. 73.6 70.3 . 71.u 69.6 70.3 67.2

RS3 x 39-11 ' s = as 82 7 as; 53 =as3
68.2 66.6 67.3 6H.l 65.? 63.5

MS2MA x 38-11 N s = as 82:? hS2 53 =as3
72.2 73.0 I 71.1 66.3 70 2 60.:

12.153 x WQS g s > as s; as2 s3\>as3
' 6h.5 62.8 6n.7 62.3 ( eu.7 61.2

rzs21+A 2: L317 s = as * 52 <RS2 s3 =a53
} 75.0 73.0 71.7 76.2 i 70.5 6“,?

l

> later than
not later than

N) H

In}.
6C

0.00

Gmwerm

0.:0

mefiwm

0.06

cmmsmm

muHLSHmE mfipmm on“ Ca

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.m6 0.06 6.H6 H.H6 0.06 0.0 0.06 m.06 m.u6
mm m H-mx mm mm H-mmx mmm mmm H-H-mm
HH-0m x «1062

«.06 m.\.6 0.66 6.66 0.66 6.66 0.66 Hi6 6.?
0 mm H-H-06 mm a-mm H-060 mm mmm mmm
HH-0M x mwm

0.06 mflm6 6.96 0.06 6.06 6.H6 6.06 +t06 0.06
0 mm mm H-6m H-H-mz H-000 mmx mm mm:

masopm mcfiommnn mafia mgp pow wcfixafim on 6200 came Umxcwm

.mmmmOAo xflm

 

maw.cwwfizofix

Bong mmfipmm

.0 wHDwB

O.Hw

‘J

CmmE(m

0.06

CmmEmm

O.mm

C0msmm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.60 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.66
000 0 0-00 00 0-0-00 00m 00 00 000

00mm x «100;

0.16 0.16 0.16 0.16 0.06 0.0 0.06 0.0. 0.06
00 00 0-06 0 0-0-00 00 0-000 000 000
m0: 0 mm0.60

6.00 1.00 0.00 0.00 0.00 0.00 6.06 0.06 0.06
0 00 0-00 00 00 0-0-06 000 0-000 000

000 0000660;

06626060006 .6 6060.0

28

One of the theoretical advantages cf recurrent selec-
tion over selfing with selection is the slower decrease in
heterozygosity from one cycle to the next. A study of the loss
of variability with each cycle and of expected and actual gains is
important. The form of analysis of variance for each breeding
group is given in Table 7. The within variance was estimated
from the average row variance which was obtained according to
the following formula:

r.
C

2
“151 Xenooooooonlo SlO

 

N-lO
where n and 52 are respectively the number of plants and the

variance of the row, N the number of plants in the 10 rows.

Table 7. Form of analysis of variance for each breeding group
within each maturity group.

 

 

 

Source D.F. prected mean souare
Total 299
Among 9
Within (estimated from L

row variance) 290 (3'2 + Qwithin
Parents n (1’

 

—=

 

The expected gains were calculated after Lerner (16).
Selection differentials were expressed in standard deviations
(i) and in absolute units of measurements (1). Expected gains,
A8, from a generation of selection can be computed. The gains
are equal to the product of heritability (hg) and the selection

differential expressed in actual units (1). The heritability

h)
\D

values were calculated in the following manner:
The coefficient of heritability for the F2 population
was calculated according to the following formula:

Variance F - Variance F
- 2 1
H2- 1

 

Variance F2
which includes the non-additive portion of the genetic
variance.
The coefficients of heritability for the various
breeding groups were calculated from separate analyses of
variance (Table 7).

2
6" within

 

6" + 6'; within

In the three early—x-late crosses (LSQHA x 38-11),
(MSZMA x L317), and (RS3 x 38-11) (Table 7B) the variances
for the late maturing parents were greater than the variances
for the early maturing parents. This indicates that the late
inbred lines were influenced more by the environment than
early inbred lines. Thus, the variance of the early inbred
parents was taken as an estimate of environmental variance in
the early-maturing groups. Likewise, the variance of the late
inbred parent was taken as an estimate of the environmental
variance in the late maturing groups. In the cases of the
other crosses, Michigan SIB, Michigan 20D, and la. 153 x WES,
the average variance for the parental lines was used as an

estimation of the environmental variance.

Within and among variances in the inbred and recurrent series from the early

and late maturity groups.

Table 7A.

 

NS2HA x 38-11

 

Late Naturity

 

Earl' Maturity

 

f1.

 

CI
"—4
,L‘
4.)
-r4
3

C10

 

Within

\OMO
.300“
U\

CD\OU'\

(\mo

26.6
23.0
20.5

(\va

(\J\O(\J

(Dr-+1

Inn-Ln

3.
a.
15.0

\{TMO

(Ur-1H

ROWE
o o 0

HOW

onJ"

Int):

CDJB
O O 0
WOW

13.153 x W25

 

,TWO

N; N

O\\CO
O O O
inc-CL)

r0c>o\
O O O

H\Or—1

BIO
MHM

04mm
.00

Junln

HS
HS
.63

O‘N\O
OHO

O\HU\
[\UNJU

(Doom
0 o o

00oux

 

R53 x 38-11

C‘P'WJ'

(Mr-1N

50101
O O O
H O" H
(~30: rfi

O m['\
(Yunn-

r-iwm
O

LAWNQ

15.%
7
a

LAJ

fYLj' (\J

r—1r-1r'1

\f O '7)

D‘I(\) C\
H (u m

(\1 OX?

[\ C‘\ L;\
N \\J ,j'

Hm\[‘.
o o o

r-‘lr-1Lr\

(.3: \f’

(”"1qu (V ‘.

(Y1\.C.,_j'

C O O
lJ‘\(\i O
Fir“ (\l

Late Maturity

 

turiLy

“
C1

1y h

 

bar

.0

(Continued)

Vichigan 513

 

Table 7A.

CL.

 

f.
H
.C
4..)
T4
3

L14

 

Within

0". (3‘ ("W

H (u --i

comm

:xcw
r-4 (\I

[\ UMJ’

. C .
OVA.) 3‘\
fir-1N

UMOxo
’1' (firm

NW0
0 O O
JL\\O

19.2
26.0
21+.0

(\CC"
0 o o

"Jr-1'0

O
\L)\LUL)
Cu

M\(1:U\

O O O
.J' “N.
(\13 m

W0 01')

HOr—i

MON)
u10£\

LACIJLR

. . C
O (\W‘)
r"! H

 

NS2NA x L31]

0‘3”)

OOC.)

anUQ
01mm

OJCI)
O C O
(MUN—1

\C\L2L\
O O O
\OJB

\CBM

. 0
(mm:
Fir-1H

2h.8

Owfibs

ONO

CO\{_‘1L\
O O O
\LNW
(Vt—NH

25.2
10.7

FORM

(Tar-1 r"!

8.9

5.2
11.6

 

Michigan 20D

0‘ (VB

(5 (\- U\

\OO‘O'\

(V'RU UN

O\C.J‘

U\(AQ

(\J CUM

(\®E\

OJr-i

,J'MJJ'

MMB
N

O(\O
[\mr‘fl
HHH

RS
hS
h

0 O

Q O O

Chin—4

O O O
in (\- ,T
Mai; (\J

(—4 m (T.

O . C
”in C\
MCUH

Ll)\fO
O

U\rfiw

cuqeo
0 c o

Lf\,_j'U\

HJQ

Own-1
(Mr-4H

 

32

Table 7B. Average row variance of the inbred and recurrent
series from the early and late maturing groups.

MSZRA x 36-11

 

 

 

 

16.153pngi

 

 

 

Early late Early Late

3 13.0 36.6 B.1 13.2
as2 1u.u 23.0 6.5 16.0
12.83 lSeO 29.5 5e7 13.2
S 8.8 :le7 (720:1 1.qu
95 9.4 21.3 6.8 13.9
8-1 12.7 26e3 5e; 1505'
RCZHA. ...12.0 Ia.153......l€.3
3&‘llee ee27el+ W2Seeeeeeeeel€e2
Fl ..... . 10.5 r1. ......... 18.3
F2e ee e 29e5 F2. e e ..L+3.3
B53 x 38-11 hS2hA x L312
RS 15.# 31.7 13.6 22.7
H82- 17e7 29’e2 l§e7 £E:e1
S 15.3 27.2 17.2 2h.€
s? 12.6 27.9 15.5 25.3
83 20.4 HS.H 16.3 1C.7
RS3........13.3 usauA.... ..12.0
38'lleeeeee27eL" L3l7eeeeeee elueo
Fleeeeeee 15e2 P‘le ee ee eel+eO
P‘E'e e e e e e e e e2l+05 F12 e e e2C'Je8
hichigan 518 Lichigan 2CD
RS 19.2 19.6 17.0 25.0
RS2 2F‘eo 18e5 1507 2F‘e6
883 24.0 29.H 13.0 30.H
S 10.5 8.3 20.1 33.1
82 7e8 0.0 lE‘eu 27e3
83 13.5 7.3 11.6 19.0

Ia. 153....16.3 Oh.hOB.......22.2
W25........16.2 0h.51A.......6h.u
W9..........E.2 N1H..........36.2
N13.........6.5 W10..........22.9

Fleeeeeeeeeeueo
F2.....o...12.85

Fleeeeeeeeeeee

14.2eeeeeeeeeeel+

(3“
H“)

, expected (in?) and actual gains

A

Selection differential, heritability (h‘)

Table 8.

in the early maturity series of six pepulations.

 

HichigangfilB

actual gain

 

(\Jl

HI

hfl

\c-NJ

¢M?C)

V):

(\ONDBOCIL‘»
H

ERR“?
DJyfiR

10.1
1.
0

\oc: O
F“3C)

5%
R82

3 (\J H (*1:
O O O O O
(fir—1C)

I

(VU\C>C‘

W¥R>Q
OO

75

HCO 0
Q13\n

(\hQD

Fhfiri

 

Michigan 20D

C\D»i
HOW

OWDCDO\

m... 0|...

Cmbtx
riflnfi

882

Who D~
H)

OWDCD

W100

('|\

 

(“W

CDUMO
ee e

HWH

 

MSPHA x L312

i §
001.1 “10“qu
O O O O O O O
(L) ('1‘ r'1 \Q (Hr-1&1
I H H

\C‘

OIDC)

rAPLt

nchv
r1

(Dumb
CVrhfi

(n
<Dr44-HI0\(fVT

2—) \J\ r“\ H (\J
H

‘§V£V§
U‘ cu \f.
BtVri

H0 0)
e 0
(“Hr \0

OUJK'
.
(Ur-1H

(\J (\a
[LUDW

(‘1

\

ZETC

frow

’-
k,

D

artlv differen

Signific

(Ccntinued)

Table 8.

13.153 x WZE

 

actual gain

 

(\J

*4

HI

h”

‘1)
O

(\J

00C)

(VOC)

 

572
0%
0:”

BLDG
e O
th-Ln

\[SCon
e e e

Fir-{P1

fig
H82

VWVOrfl

1:30
I

OCDO€3

(MCDD

L\C\JO\
Inmm

\OU‘M

HCUH

 

e
(U

 

RS3 x 38-11

I
imxf (00H O
L\( V0 UNOCJ
t0 (7‘ OO O CTO
NOr—nn NOON
REV-3‘1 ‘RQYR‘PR
UJWMF~ LUCDO
mam m
.iCDKD JTDCF
[\(\\D L\L\Lr\
U'\O(\ U‘O(\
HNH HNH
(\J

mmm N (\J
Elm-ELI. Cut/1U)

MS2HA X 35-11

!‘

 

,J'Jrfiixjr-iQ
I

d) 3‘01

CDC) (.1

V

75%
7%
20¢

O‘t—‘HD

me
H

O\(\O

HHr-i

e
Ot\mwaqnacro

LA)
4:-

‘U

WOJC%_

2 J
O 0
1.9L) U L» A.)

0%.? u?)
H: J
H

O‘"[\ ’3

rhfirfi

CU (\J
(Amt/i

26:1‘0

iron

ent

£4

Genetical variability was exhausted after one cr
two cycles of breeding. This conclusion was reached from
the following:

(1) The within variance representing the variability
between plants did not decrease (Table 78) from one cycle to
the next as expected, had genetical variability been present.

(2) The F ratios between within and among variances
(Table 7A) were not significant. There was less variability
among strains than within rows.

(3) The variances of the early-maturing groups
(Table 78) were close to the values of the earliest parents.

(h) The F2 heritability values (Table 8A) were high
and expected gains were realized. The heritability values
were much reduced after the first cycle of breeding, and
progress, if any, was much less.

In Table 9 actual gains were considered null if the
differences between cycles were non-significant. to progress
toward early silking date was made after the first breeding
cycle in either the recurrent series or inbred series of the
double cross Michigan 51B, the early single cross (18.153 x
W25), and the double cross Michigan 20D where the parental
lines are similar in maturity. On the contrary, sore progress
toward early silking was made after the first breeding cycle
in the recurrent series of the three early-x-late single crosses
(MS2HA x 38-11), (RS3 x 38-11), and (152411 x L3.7), and in the

inbred series of the single cross (M324A x L317).

36

Table 9. Expected and actual gains in the early-maturing series
of six p0pulations.

 

 

 

 

 

 

 

Groups Expected Actual Expected Actual
bains gains Gains Gains
Kichigan 518 13.153 x W25
1:2 7e5 6e6 200 6‘e2
Eu 0.H 0.1 0.0 0.0
582 2.3 0.0 0.0 0.0
F2 7.5 3.2 2.0 3.5
3 0.0 0.0 0.0 0.0
S) 0.0 V.w 0.0 0.0
hichigan 200 553 x 38-11
F2 3.9 1.7 2.8 7.“
RS 0.0 0.0 0.9 2.4
RS2 0.0 0.0 1.7 0.0
F9 309 “le6 2e8 SeE‘
S' 0.0 0.0 0.0 2.1
82 0.0 0.0 0.0 0.0
Js2uA x L3lZ_ rsaun x 33-11
Fa 9.8 8.0 8.8 M.O
as 0.1 0.0 0.4 u.7
H82 C.# 6.h 0.8 0.0
F? 9.8 6.0 8.8 the
S 1.6 3.3 0.0 0.0
82 1.0 1.2 0.0 J.0

 

 

It is concluded that fixation of genes for early silking
date was very rapid. Very little improvement can be made after
the first cycle of breeding. Progress seemed to be greater
with more diverse crosses of early-x-late lines such as

MS2NA x L317.

37

COMFAEISON BETWEEN BREEDING GEOUPS 1N
THE LATE MATURITY SERIES

The mean values of each breeding group are ranked
in Table 10. Values within a bracket are not significantly
different according to Duncan's multiple range test (4).
Breeding groups within a cycle can be compared (Table 10A).

Table 10A. Comparison between inbred and recurrent selected
groups within a breeding cycle for late maturity.

 

P0pulaticn First cycle Second cycle Third cycle
Michigan 518 s = as 59 as; s ‘> as3

70.6 66.5 76.2 68.9 81 5 71.1
Michigan 200 S RS S RS

86.6> 78.2 85.3 > 8020 93.2 >RS 65. 0
R53 I 38-11 S = RS ESQ S =

78.2 79.2 83. 3>77. 8 81.6 8.036
MS24A x 38-11 s - as s; asg a? = RS3

80.9 82.2 86. 6 81. 5 81.1 06.4
Ia. 153 x w25 s = as $0 = h82 Sq - hS3

72.1 71.u 7fi. 0 7u.8 76.0 75.1
MS24A X L317 S 3 RS 8 3 882 S3 = 83

85.6 79.2 8 .4 82.3 92.2 54. 5

\> later than

- equal in maturity

 

Though the results differed somewhat for each p0pula-
tion, the inbred series tended to be later than the recurrent
series. However, after three cycles of breeding, the differ-
ences between the recurrent and the inbred series were small

and non-significant.

 

 

 

 

 

 

 

 

 

 

 

 

 

6.18 H.8w H.mt c.8m 1.8w m.mm m.mw m.Hw m.ow o.sm
H-H-mm H-wmm mm mm mm; H-mc mm max m we

HH-wm x 42mm;

H.am m.;® m.sw :.mw o.Hm 8.0m m.ou «.ms m.ss o.:n
H-H-ms H-mm mm H-mmm mm mmm mm m mmm mm
Hfi-wm x mwm

m.am m.¢s m.os w.ms H.ms H.Hs o.os o.wc m.oo 0.58
mm mm H-H-mm H-mmm H-mm mm; m mmm mm mm

quLSpms mama

ecu CH mqsoom

 

mfim :mmfinofiz

.momwopo xfim 80am mmaumm
wcfinwwan mad: mnp p09 mcfixafim cu mmmu Cmma vexcmm

.OH manna

 

 

 

 

 

 

 

 

 

 

 

 

mfwo m..\.w :.mw ®.mw mimw m.:w m.m® méK. $.05 015
mm H-H-mm mm H-mmm m mmm mmm mm H-mm mm
mfimg x «:mmz

m.ow «.55 0.3. 9mm. Hfim. @345 0.4m. Tmu {an 0.00
H-mmm H-H-mm mm H-mm mmm mmm mm m mm mm
MN? K mmfi .mH

0.30 m.mm n.0, o.mw w.mw o.mw «.mw o.ow m.ms o.mu
2.~-ui mm mm H-mmm H-mx mms m mmm me we

mom cmmfinoaz

Avmzcfipcoov .OH m3 Em

#0

Observation in the field led to believe that inbreeding de-
pression affected the inbred series more than the recurrent
series, making the comparison very difficult.

Table 78 indicates that variances for the late maturity
groups were close to the late maturity inbred parent variance
and those variances did not decrease from one cycle to the
next except in the inbred series from the cross (MSZHA x L317)
and Michigan 200. This indicates that the genetic variability
uas usually exhausted after the first breeding cycle.

In Table llA actual gains were considered null if the
differences between cycles were non-significant. Gains could
have been expected in the recurrent series in only three crosses,
Hichigan SlB, (MSZHA x L317), and (MSZHA x 38-11). In general,
there was considerable gain when none was expected. These
gains must be interpreted as effects of inbreeding depression.

These results indicate that fixation of late maturity

genes was very rapid by either breeding method.

1+1

 

 

. mflm> mm 0L0 ma chog mocwewwmm
.H0>0H mm exp pm chow 50pm pumpwkafin 5HpC0ofihficmHm

 

 

§
inu.5 mum
0.; 0.0 50 m.0 5.5 mmm
5.0 6.0 «:0 0.ma 0.0 m0
m5 0.5 515 0.5 :4 m
Ha-mm x asmma
m.am 5.0
0.a 0.0 m0 m.q :.H mm
5.0 0.0 mo 5.» m; m
0.0a 5.0 505 m.HH 0.H mm
0.50 Mhm
0.~H 0.0 wo m.HH m.m 005
0.0 0.0 50 5.0 m.H mm
«.0 5.0 505 m.HH 0.H mm
005 cmmacoaz
m._a m._
6.: 0.0 50 5.m :.H mm
0.0 0.0 50 0.5 0.w m
0. 5.; W55 H.0 5.H mm
Hun m.5
m.w mqw awe 5.0 m.H 055
:.m 0.0 55H :.5H 0.m mm
5.H 5.: 555 H.0 5.H 5
2500 Mmm mm M M
H05uo<

 

me1CmMH£on

.msofipmfisaoa xfim go mofiumm muHHSme mama 030
C5 mchw szuom 0cm ANLHV umpomaxw .Amnv mafiafinmufiumn .Hmfiucmueccfiv :ofipomfimm .HH magma

(Continued)

Table 11.

NS QHA x 38-11

N
13

HI

hfl

 

\COO)

WOO

\C(\JCI)

l\O‘-L\

”1'06;
0 I .
HOJH

F2
8

 

Ia.l§3 x WZE

tho orn

(\OOB

NQ 5.9 we
L\ O O
m

NNO

(’KUQ

0mm
NHr—i

F
.8
R82

57%
0'

HO\O
O I O
(Mr-1W

LnOO

BOOB

NO\O‘\

mid)

0mm
WHCU

N N
CuU‘JU)

HS2hA x leZ

 

(DO-'4“
00.0

H (“Nu m
I

wuafia
mu) 0
(\rfim

(\mm
HHH

\O'IDLT) N
O

:HJH
(“I

(Q) «.0 \o \

CO In «7 [1%
C O O C
H N

753
#371,
1+3,0

U\(DU\

O
H
Q:
t»
0)
H
he
Ln
Q)
.13
p
4.)
(U
0
h.
we»
NS
H
Ect.
O>
5,
mm
[1,
4.)
C10.)
0).:
up
(D
(Hm
‘e—a-H
ca
'0...)
C:
>.-.-1
r10
+90.
C.
(ca)
00
HQ
“40)
Ht.
CO)
DDS-4
an)
(DJ;

x)

“3

E: (qubTo

0 O O O I

‘ J()kuf\
I

Quid)
WH(\1
I

 

c503
H9504

[\
0
kV

 

[\ KOO
(MCDQ

0
(V

 

543C?
(MCDU

(\J
.C
H

.mzam> mm mnu mH usfioa wocopwwmm
.H0>ma Rm on» pm oumu Eonm pcopmgmun mapcmofihflcwfim a

50 5.ma 0.0
50 H.0 5.H
50m m.5 5.H
mo 0.0 04
mo 5.0 5.H
mmm m.5 5.H

mm a .5

Aamscapcoov

mm
m
mm

mm;
am
“a

 

Ha-mm x1000

.HH mHan

1+1.

Table 11A. prected and actual gains in the late-maturing
series of six populations.

 

 

 

 

 

 

 

Groups Expected Actual Expected Actual
Gains Gains Gains Gains
Michigan 518 18.153 x W25
F2 1+0; " 105 705 201+
80 2.9 0.0 0.0 0.0
hSQ 1.9 0.0 0.0 0.0
F2 L105 000 705 201
S 0.0 6.0 0.0 0.0
S2 0.0 1+09 000 000
Michigan 20D 853 x 35-11
F 8.5 6.2 2.7 5.2
RS 0.0 0.0 0.0 0.0
RS; 0.0 13.0 0.0 0.0
F2 8.5 10.6 2.7 “.2
S 0.0 6.7 0.0 6.1
82 0.0 3.9 0.0 0.0
as2hA x L312 Ms2uA x 38-11
F2 10.8 - 1.8 506 502
RS 5.1 0.0 2.9 0.0
882 h.0 0.0 0.0 5.9
F2 1008 0.0 506 309
S 8.5 0.0 0.0 5.7
S2 6.6 0.0 0.0 0.0

 

 

HS
MLISZCEE coursnr AT hAEVLST

As another measure of maturity, moisture content of
ears at harvest was determined for each plant in two popula-
tions, Michigan 518 and (853 x 38-11), harvested September 22
and October 5 respectively. Some plants of the Michigan 518
pOpulation silked as late as the first week of September, and
some of the (R53 x 38-11) population silked as late as the
second week of September. Thus, many ears were immature.

Results of moisture content measurements at harvest
were similar to those measurements from planting date to
silking.

Fixation of genes for low and high moisture content
was very rapid. The greatest progress towards early and late
maturity seemed to have been made with the first cycle of
breeding, as shown in figures 7 and 8, where values for
parental lines, F1, F2 and breeding group means have been
graphed. Though a few strains from the early-maturing (Table 12)
were as high in moisture content as some strains from the late-
maturing series, none of the breeding group means from the
early-maturing series was greater than any breeding group mean
from the late-maturing series.

Analyses of variance (Table 13) indicated that there
were differences among early breeding groups of the two crosses

and among the late breeding groups of the Michigan 518 cross.

Range and mean for ear moisture content at harvest of 9 breeding groups

from two populations.

Table 12.

Late Maturity

Ran-e

Early haturity

han e

.ean

I

§\"
1 1

".6811

\
I

R53 1 38-11

 

BNWfiOxOOLnj NH
O O I 0 O C O O O O

\L Q. 011* J (1‘) K10“ Cu r-1

U\LF\U\\O\OU\\L\O\LJ

(”O (\GNWOBJ
. O O O O O O O O
Cir-nucmocxoom
Hr-Ir-INCVr-Ir-‘Ir—Ir-I

IIIIIIIIIIII
HN©NMJOCXDO\
O o

O O O O O 0 o
nuxcuwmiodnjuo
\ou:¢m\o\(m\ouo
1 II I II I II
CD(MOMOriOV r4u\

O O O 0 O O 0 O O
.—4 \L MN r-1 MI): W
injg:u\wunuMoux

[\Q'wV: U\(\J OU\(" 3»

mmmmmmw‘.m M

\00\ 01.1mm
C C O O O O O .
o¢.(")Q\LQHO-fl]
O‘\r—Ir—IHr—I¢—Ir—I r—I

III III! I I" I
«seine-ouwm\OLn

. O . O C . .
~110po WNW/r4
:Jmm "WMJJ
I I I I I I I I I
wfl‘)\CCNO\HNO
. O . O . . O O .
r—I O"~C)LU\-O,J‘ (\OCO
MNNNNNNNfV

RS

8

882
RS-l
RS3
882-1
hS-l-l

Strainsoo..8o6

l.

L.S.D.

............l2.5

groups................H.Q

2.

O-J‘

r40x
in:-

F1:
F2:

39.3
59.7

B53:

Parental lines
38-11

(Continued)

Table 12.

Late Maturity

Ban 9

Early Maturity

Mean

iean

. e

han

 

Michigan 51B

(12) (\JCDLnOxJ H L\

O O O O O O O O
c\ (u o m NJ’ (1) (‘0 Q
j» mm’m\(“.u\m\c

 
  

O\O M(\L\O‘O OJ
0 O O O O O O O O

oxqg 0.6mm (’va
r-‘Ir-‘Ir-Ir—Ir-Ir-I

IIIIIIIIIIIIIIII
U\L\CD\O(\JU\MJO\

O O O O O O 0 O 0
01mm: (MUS) W (\J [\
mmmgmmxofim
I I I I I I I I I
\(1 L\U\ symwc rfltmln
O O O O O O O I O
mmxcmxomooom
1:111“ mixo

o mum-4 r) mrfitn

. . . . . . . C O
B0100 C) (\JJ (Tm-0C1)
(V); (“11.1 mmmn'

    
  

OO NNOmeMO
O O O O O 0 0 I O
O U‘xL\r—+ moo (‘5)ch t\
r-Ir—I Hr-I v—Ir-I

II I II II II I II II II
o o o o o

O O O O

(“NH ('1 U‘\\LOU\LD
'3U\1 r3“?i”i‘§
(\. no H: C \o H in
O O O O O O O O O

(*3ka mm L\\0 O‘r-i
mmmmmm mm m

H

r-II

r-I Ir-I

I MWI
U) (amount/1mm
mmmmmmmxm

30.2

........9.5

strains....6.6

l.

L.S.D.

............7.3

groupSOOOOOOOOOOOOOOI00.0.6.2

2.

Parental lines

83.0
1.8.0

F13
F2:

 

 

1+7

LI 5:.

 

 

 

 

 

iflul

00.00 m0.000.0 :0.00 l 50.m00 00 06800
.. 0.0: 0.0005 .. 0. m “ 04.1510 05 00006800000
5.0.. 10.005 0m.0 M 00.050 0 . 000003
.. 0.580 0.0005 0m.000 _ 00.000 0 800000 .00800

.. 5.050 1 00.055.00 5m.5: 00.0m0: 00 mc0maum

00.2: 1 00.:: - 00.:0m 00.:5m 0 .0000

_ 05.550.00 00.3035 050 08060

000.0000000z

5.50 “ 00.000.0 05.00 30.3000 00 00000
0.00 M00.5500 05.0m 00.5000 05 00006808000
0.80 00.0500 0.c; 50.00: 0 000003
m.0m ”05.5mua . :.mma mm.0mma - m mascam .mmmpm

m.00 om.mnc.0 »* 3:.m: :m.aam: 5w mC0000m

030 00.0 50.000 50.0m0 0 .0080

05.000.00 1 50.5000 050 08000

.0.0 .m.m .m.z _ .m.m .5.0 060200

mmzoam 00002005 0000 mmmoawINM0ASpme mahmm 003mm x mmm

Eogm mgflOhw mcfiummpp ucmpmwv0n c0

 

ucmucoo madpm0oE 000 mocm0pm> Mo mmmAHmc<

 

 

.mom000o qupo0m0U 030

.MH wanna

no

LT LN Midgbllifi GhC/L'ES ll"; Iii}:
V’sLY MATbthY SLEIES

The mean values of each breeding group are ranked
in Table 1%. Values within a bracket are not significantly
different according to Duncan's multiple range test (A).

Breeding groups within a cycle can be Cemparcj, Table 1kg,

Table 1%. Ranked means for ear moisture content at barrest

 

 

of nine brrcdin; groups from two crosses, hicrigan
518 and 853 X 35-11.
853 X 38—11
532-1 R83 882 83 Ps-l 9-1-1 82 53 8 F2
30.8 32.2 32.5 33.5 3u.h 35.3 35.8 37.6 38.7 53.5

 

 

 

Eichigan le

RS3 H3 H3 HS-l-l 53 BSg-l 551 8 S2 F2

2
35.3 37.0 35.5 35.5 36.7 39.3 no.1 us.# 52.9 55.0

 

 

 

 

Table lhA. Comparison between inbred and recurrent selected
series within a breeding cycle for early maturity.

 

 

 

 

Pepulation First c1319 Second cgcle Third ggcle

RS3 x 38-11 S 2 RS 52 : R82 83 = {83
39.7 37.6 35.8 32.4 33.h 32.2

Michigan SIB S 3 RS 52 8 RS2 S3 = R33

u2.5 37.9 52.9 38.5 38.7 35.3

 

 

50

The recurrent series tended to be lower in moisture
content than the inbred series but the differences between
them were non-significant. These results agreed with those
obtained from days to silking where the differences were also
small.

Row variance for the nine early breeding groups
(Table 15) gives an indication of the variability in the
series. Variability in the early recurrent series of the cross
(H53 x 38-11) was similar to the variance of the early inbred
parent, RS3, indicating that genetic variability was about
exhausted. Variability in the early inbred series of the
(RS3 x 38-11) cross decreased when compared with the RS3 value,
indicating that after three cycles of selfing very little
variability existed for moisture content.

The variances of Ia.lS3 and W25 were considerably
greater than the variances of M13 and WQ and should be taken
as an estimate of the environment. The variances of the
early series tended to increase, but not to the extent of the
latest parental line value, WQS. It is safe to assume that the

variability in these series was exhausted.

51

Table 15. Average row variance for the inbred and recurrent
series of the early and late Nichigan 515 and
H53 x 38-11 maturity groups.

 

 

 

 

R53 x 38-11 Nichigan_5lB
Early Late Early Late
RS 30.2 67.# NE.2 36.H
832 3.6 91.5 57.9 50.0
H83 65.6 81.3 86.2 81.h
S 99.1 125.0 38.7 81.6
82 79.2 75.3 90.9 9h.6
33 51.5 86.7 7H.8 72,5
853 50.6 Ia.153 66.0
38-11 79.0 W25 99.0
E13 15.H
Fl “6.8 W9 8.2
F2 115.0 1‘11 50“
F2 32.9

 

 

Table 16. Ranked means for ear moisture content at harvest
for the nine breeding groups from two crosses.

R53 x 38-11
F2 R32 hS H33 8 S3 ESQ-1 82 88-1 hS-l-l
59.5 53.3 56.7 58.0 58.3 62.2 63.5 65.0 65.9 60.h

 

Michigan 518
F2 as as2 8 a33 hS-l hSQ-l hS-l-l 52
HH.O h7.8 50.8 52.2 54.“ 55.5 58.1 62.7 62.9

U)
L»

\J
“v
A:

 

 

 

 

 

\n
to

CUI~2PAhISCI§ Bifl’WLLI‘.‘ BhEnDIIéG URCUPS 11V This

LATE MATURITY SEEIES

The mean values of each breeding group are ranked in
Table 16. Values within a bracket were not significantly dif-
ferent according to Duncan's multiple range test (Q). There
were no significant differences among the breeding groups of
the R53 x 38-11 population. however, the inbred series in the
Michigan 518 population were very high in moisture content
when compared to the recurrent series, Table 16A. These in-
creases in moisture content in the inbred series should be
attributed to inbreeding depression since, referring to Table
15, the variance in the late inbred groups did not decrease
from one cycle to the next, and was inferior to the latest
parental line, W25. This indicates that very little genetical
variability remained after the first cycle of breeding.

Table 16A. Comparison between inbred and recurrent series
in the double cross Michigan 518.

 

 

 

First cycle Second cycle Third cycle
as < 8 R52 < 52 as3 < 33
h7.8 52.2 50.8 62.9 54.5 70.3

 

TEST CROSS LXPLEIMERT kthLTS
General Considerations

The objective of the yield trials was to measure the
differences in maturity between breeding groups when the lines
were in hybrid combinations.

Results show that the most important differences were
between early and late maturity groups. Differences between
recurrent selected and inbred groups were very small, and when
significant did not indicate any trend. Cycle gains were very
small, indicating that genetical variability was exhausted
after three cycles of breeding.

There were significant differences among the selections
for days to silking in all the experiments, Table 17, except
55-90% and 55-905 which tested the pOpulations of the single-
cross (MS2MA x L317) and the double cross Michigan 203. There
were significant differences among the selections for car
moisture content in all the experiments.

The correlation coefficients for silking date with
moisture content were all positively significant (Table 18),
indicating that the early-silking selections were lower in
moisture content than the late silking selections. In general
there were no positive correlations between yield and moisture
content in the 1955 testcross experiments (Table 18). In H of
5 experiments in 1956 the high yielding selections were the late-
maturing selections. In 1957 the high-yielding selections from
two early-x-late single crosses, (MS2HA x L317) and (HS3 x 38-11)

were the early-maturing selections which had time to ripen

ԤԤ
U\(
0

x?“
W \\

T

V9WZ
umurn
O

rV-yU\O U‘\(\~
rhfiCV FHH

c
O .

WUVVQ
FPDL)

tCuU‘
rirhfl

so:
e;.cfi

a\.na
\m .C (

Rm.ma

oaom.:

¢§HL.N

eeom.w
*§®H.m
«omH.w

a. ®.JH

oojm.m
mm.a
towfi.m
m:.o

§§®N.N
*nmm.m

.m

nfiwww

ma.m
wu.:
m:.o
Ro.¢
m:.o
was
No.0
was
am.m
mm.o
aw.o
w:.w
.cmuo

xk)

eta
one

I’ §

i i

.U\\o
U\\Q(\J \[2 O\

«35- UflnJ‘ P\

D
f
U‘

§
§
1- l)
m\(\

C

§

(
(Y. _
U\

monumfiox

WUfl

hm)
OHH

\
0

§

‘fihgl
O
.irfifll

‘Rfifi ‘Rfflwi (“C20 CMC
JWO rHD\O ,1E\U\

I O
(UNC‘U

O

‘&
CUOJ ,jrfi

\ i

‘
\
CD¢J
O

‘

.wwoo

.mtsma pcmo you H on»

\ O \
.50 u

c o
new m

 

 

wwmo mcfixflfim

mam-sm
Hoq-cm

mae-mm
:oo-em
:oo-mm
HHOIB

momuem
moo-mm

moo-em
moo-mm

moe-om
moo-mm

 

.0; .oxm

um pomofimficmfim a.

mam smmfizofiz

ms.mq x «:mmzv

HH-mm x mmm
mom cmmasoaz

Ha-mm x «:mma

 

cofiumazmom

.vflmfih

pom pmo>am£ um pompcoo oezumfioe .mcfixfifim 0p mcfiucmfia some mmmc ow mm mmma com

.emcfl .mmofi :« wouosouoo mpcmEprwa we zpfiafinmfipm> Mo mpCmHofigowoo .mwsfim> :m:

.ma mfinme

55

before killing frost on September 20. The parental lines of

the double-cross Michigan 518 are considered early-maturing

lines, and all the selections from this double-cross popula-

tion had time to ripen before killing frost.

Table 18. Correlation coefficients between days to silking
and ear moisture content at harvest, and between

moisture content and yield in the 1955, 1956 end
1957 testcross experiments.

 

 

 

Experiment Silking date Moisture
Ecpulation Hunber x Moisture: x Yield
(tsehA x 38-11) 55-902 0.757H 0.1uo
(nszuA x 35-11) 56-905 0,5u3.. o.h68**
Michigan 20D 55-905 0.3h5** -0.l50
Michigan 20D 56-902 0.H56" 0.H63**
Michigan 518 56-901 0.828" 0.671**
Michigan 51B 57-912 o.975** 0.79ut:
(853 x 38-11) 55-903 0.680" 0.653**
(853 x 38-11) 56-903 0.738H 0-7O7*‘
(853 x 38-11 57-911 0.892*‘ -0.l96
(usauA x L317) 55-90% 0.777H 0.200
(EsauA x L317) 56-904 O.659** 0.563**
(msauA x L317) 57-913 0.720H -O.268

 

** Significant at the 1 per cent level.

 

 

Differences between breeding groups.

The means of R33, S3, BS? and 82 breeding groups for
days to silking and ear moisture content have been ranked in
Table 19. Values within a bracket were not significant accord-

ing to the following test:

X1 - X2

 

 

sp/
\' l/n (N1 / N2)
I

where x1 and x2 are the means of the breeding groups, Sp the
scuare root of the error mean souare given in the analysis of
variance, N1 and N2 the number of selections in each breeding
group, and n the number of replicates.

In one pepulation (R53 x 38-11) there were no differ-
ences in date of silking or moisture content between the 83
and RS3 breeding groups in either the early or the late maturing
groups. Similar results were obtained by both breeding methods.
In the other two pOpulations the early inbred selections were
earlier in maturity than the recurrent selections, and in the
Michigan 51B pOpulation the late inbred selections were later
in maturity than the recurrent selections. The differences
between those breeding groups were very small.

In Table 20 actual gain cycles were considered null if
the differences between cycles were non-significant. Actual

gains though small seemed to have been made for early silking

\
‘5

date in all the crosses and for lower moisture content in the
hichigan 515 and KSZHA x L317 crosses. Very small gain seemed
to have been made for late maturity in the hichigan 51B and

hSQhA x L317 crosses.

Table 19. Days to silking and ear moisture content of 8;, ES},
82 and RS breeding groups in three l957 testcross’

experimen s.

 

DAYS TO SILKIXG htISTUPE CCTlENT

Early Maturity Late Maturity Early Vaturity Late Maturity

 

Michigan 51B

 

 

 

 

 

 

59.1 60.u ' 6u.7 68.1 ! 27.u 29.8 g 38.5 h2.1
52 as2 , hsg 52 g 52 532 5 use 32
61.0 61.h ? 6h. '. .u 2. i .
\ _wmn__m_A/ , 3 67 3 [‘32 3 9/; 37 1 Mo 0
553 x 38-11
S3 RS3 r as3 33 I 53 as3 as3 33
63.1 63.u f 73.6 73.u § 33.9 3S.h uu.5 hh.6
L_--. _--....._J I L..- 1 I . _j _‘ L---“ . _-..__1 \__..__- . . _/
82 RS2 ; 882 82 g 82 R82 PS2 82
65.1 6n.7 ] 72.5 73.5 g 3h.u 33.5 h4.9 u6.9
useuA x L317
S3 883 ! as3 53 g S3 RS3 1 RS3 s3
66.u 65.5 § 72.9 73.2 E 35.0 37.1 ' u2.3 u3.6
l \«~_._1. _-"im_-_.4/ 4 \__._.- . _J
52 R82 i H82 S2 ' S2 H32 ES2 82
68.1 69.8 [ 71.u 73.1 i 37.0 38.1; 41.9 u2.o

I Lm_.

 

 

\‘n
(' ‘7'

Table 20. Cycle gains in days to silhing and ezr percent
moisture content in three populations tested

 

 

 

 

in 1957.
sany iATthITX 131s hhpthTY
Population Days hoisture Days Roisture
Michigan 518 Inbred 1.9 5.0 0.0 2.1
Hecur. 1.0 2.2 0.0 1.h
E53 x 38-11 Inbred 2.0 0.0 0.0 0.0
Beour. 1.3 0.0 0.0 0.0
NS2MA x L317 Inbred 1.7 2.0 0.0 1.8
Recur. 1.3 1.0 1.5 0.“

 

 

These results indicate that recurrent selection was no
more effective than continuous inbreeding with selection.

This conclusion is important. First, inbred lines are
used to transmit early or lat maturity in hybrid combinations;
second, inbreeding depression in the nursery made very difficult
a direct comparison between inbred and recurrent selected lines.
Therefore, testcrossing lines for maturity may be the only way
to measure the relative effectiveness of the two breeding rethcds.

One important point was demonstrated in experiment 57-91].
It was possible to obtain earlier and later lines than the
parental lines from an early-x-late cross, Table 21. The
parental lines 553 and 38-11 testcrossed to (ohh3 x A150)
were included. One line of the recurrent selected series
(hSL3-RSLlO5-ESL110) and one line of the inbred series (121-2-2)

were later in date of silking and higher moisture content than

the parental line, 38-11. Two lines of the 82 breeding groups
were higher in moisture content than the parental line 39-11.
One of these lines (L2l-2) gave rise to two lines which were
higher in moisture content than the parental line 35-11. One
line of the recurrent selected series (hSh20-hSE1-RSLH) and
one line of the inbred series (E2-2-h) were earlier in date

of silking than the early parental line R53.

.Hm>wa acme poo H who pm pomowhfiowfim a.

 

 

 

 

 

l .H0>0H acme poo m one we ucmofiwfiowflm .
w.w \ mame as Ha-wm
w.m m mama 00 mm; emasaoca macaa Hmacwama
..m.0 m.e0 m.w0-m.s0 0.m 0.H0 0.m0-m.mm mm
.0.m m.:0 0.m0-0.00 0.m :.H0 m.:0-m.0m mmm
..0.0 a.00 m.00-m.00 0.m H.0m 0.00-0.wm mm
..m.: s.:0 0.m0-m.m0 0.m 3.00 0.mc-m.mm mmx mam cawaeoaz
aao.m H.mm 0.05-0.mu o.m H.w0 o.omum.u0 mm
0.H :.Hs o.ms-0.00 0.s 0.00 0.Hs-m.a0 mmm
..:.s m.ms 0.me-0.mu 0.5 1.00 0.0s-m.:0 mm
..:.a 0.mu 0.0m-m.as 0.5 m.w0 0.00-0.00 mmm mama x «ammz
..m.m m.ms 0.ms-0.ms 0.: e.;0 m.m0-m.:0 mm
..:.u m.ws «.ms-m.0s 0.: H.m0 0.50-0.a0 mmx
..m.0a :.ma m.ma-m.wa 0.: H.m0 m.:a-a.ae mm
..H.0a m.ms 0.0s-0.as 0.: :.m0 0.m0-m.a0 mmm Ha-wm x mmm
monopowwflm one; wmcmm .o.m.H one: mwomm cofiumHSMOQ
.ufiozuwz mpmq .pHpSpmz wagon

 

omHVSHm or... mwda

.mpcoefipoaxm mwonouwmu mmma omen“ we mQSoam mcflvwopn wm woo .mmm .mm ammx
mpmfi wow mfipmo pom mofixafim 0p mmmm Cece 0cm .pcmpooo menumfion new omce .mmowm .Hm mfipce

6

.H0>wH acme own H on» um pomonchHm

 

 

 

 

 

l .Hm>cH powo you m who am pomoHMHcmHm a”
55 H :.0 Ha-0m
55 \ H.0m mm; emesfiocfi macaa Hmoumbmm
..e.5 0.0: m.m:-5.5m 0. :.mm m.0m-m.5m mm
..H.m H.5m 5.0m-0.m 0.m 0.mm m.0m-o.5m mm:
..5.Ha H.a: :.m:-a.5m 0.m :.Hm 0.0m-0.5 mm
..5.0 5.0m m.0:-m.em 0.m 0.5m m.mm-a.mm mam mam sewage“;
..6.5 0.a: :.m:-0.5m m.5 0.5m w.om-m.:m mm
.w.m 0.H: m.m:-m.0: 5.5 H.0m m.H:-0.0m mmm
..o.w m.:: 0.0:-H.m: m.5 0.5m 0.H:-m.om mm
.wm.m m.w: :.u:-0.xm m.5 H.5m :.o:-m.wm mmm 5qu x <:mmz
..5.ma 0.0: m.m:-0.m: 0.5 :.:m 0.Lm-:.am mm
..H.aa 5.:: H.05-:.m: 0.5 0.mm 5.0m-m.0m mam
..0.0H m.:: 0.5:-0.m 0.5 0.mm 0.5m-a.mm mm
..5.m 0.:: H.0:-5.0: 0.5 :.5m H.H:-0.mm mmm afi-mm x mmm
ooumpmwofia :00: mwomx .Q.m.H smog mmomm coHumHSQom
synonym; 60mg soapsomz magma
Hzaezoo amoemHsa
fleascfiucoov .Hm magma

’3‘\
h)

YIbLD BhSULTS

Selection was directed toward maturity and not com-
bining ability. hence variability for yield should be ex-
pected within maturity groups. The only trend that could be
expected was that late-maturing selections should yielo more
than early-maturing selections in a year with a long growing
season.

In two 1955 test cross experiments (Table 22) evalua-
ting selected F2's of the MS2HA x 36-11 and MS2HA x L317
populations, the yield ranges within maturity groups were
greater than the L.S.D. indicating high and low yielding
selections within the groups. Though the differences between
early and late selections were significant, the yield ranges
overlapped indicating that some early selections yielded as
much as some late selections. The same conclusions can be
drawn from the 1956 testcross experiments (Table 23). Less
variability for yield seemed to exist within maturity groups
of the 1957 testcross experiments(Table 2%), since the yield
ranges were smaller in many cases than the L.S.D. and when
they were greater the differences were small. Selections
within a maturity group tended to be similar in nature. however,
since there was a very great overlapping for yield among the
maturity groups, it would be possible to select for combining
ability in early and late-maturing lines.

In the R53 x 38-11 p0pu1ation, the parental lines

R53 and 30-11 were included. Four lines from the early-
maturity groups yielder more than the early-parental line

853. No late maturing line yielded more than 38-11.

6M

.mooHpomHmm mop moosm mcoompcono oz 1

.H0>¢H udmo mom m who am pomonHCmHm u

 

 

 

 

- m.H m.5w m.00-0.w0 0.00 H.00-0.5w 00m Casanoaz

- H.0 H.0w m.00-0.m5 m.0m 5.m0-0.w5 Ha-wm x mmm

m.wH .m.: m.50 :.0ma-m.a0 5.:0 0.HHH-:.50 5amq x «:mmz

0.5a .5.0 m.m0a H.5Ha-5.00 0.00 :.:aa-m.m5 Ha-wm x «:mms

.m.m.q ooowomonQ omoz oowm C002 mwomm cofipmquoa
weHmoaex mead wealoeas sheen

.muooEHpodxo mmopOpmop
mde on» CH mm UopomHmm mme can memo mo moHon you :mwe 0cm wmcmm .mm wHomH

6.5

.H0>0H pooo pod m mop pm poaonHcmHm *

 

 

0.ma .5.5 0.m5 5.:0-0.H0 m.:0 m.:5-:.0m mm

0.ma .H.0 m.55 5.50-5.50 a.a5 0.55-0.m0 mmm Ha-um x <:mmx

5.:a .5.ma m.:0 m.05-a.05 5.00 0.55-5.:m mm

5.:H 0.m- 0.55 H.00-5.m5 0.00 0.m0-0.:5 ~00 . 5Hma x «:wms

:.ma .0.HH :.00 H.H5-H.00 0.00 m.55-:.05 mm

:.ma .H.5H 0.m0 5.m0-5.m0 5.00 0.05-:.w0 mm: Ha-0m x mmm

0.0a 5.:- H.50 :.m0-a.:5 0.:5 . :.m0-:.u0 mm

0.0a H.m :.:5 0.05-H.m5 m.m5 H.00-5.50 mmm 000 amaaeuaa

m.:a .:.5 H.m5 H.00-5.:0 5.m0 0.05-0.5m .mm

m.:a .H.5 0.50 5.55-0.05 5.00 m.o0-m.5: mam mam camfieoaz
.Q.m.H mocmnmwwfio Emma mNCmL new; wwccm ooHumHzmom

mac wqh<m

.mpcwEprgxm
mmoeopmop wde o>H5 CH mcHommon mm 0cm mmL mo monHz new cmox 0cm owcmm .

(*1
(\J

0H0 0.5

.Hw>oH pcoo non m 0:» pm

ucaoaoacmam .

 

 

0.5a 0.m- 5.5m m.5a "0.:0-m.5: m.00 0.0a H0.00.005 mm

0.5a :.0 0.H0 m.na “H.0010.05 m.a0 m.0a "0.m5-5.m5 mmm Ha-0m x mmm

0.00 .5.5\ :.m0 0.HH "H.05-m.0m 5.55 0.mm “5.:5-m.am mm

0.0m .5.6- H.0m 0.5m "0.00-0.H: 0.50 a.am 1:.m0-m.00 mum 5Hm0 x «:00.

0.5m .0.ma H.05 0.ma "H.00-m.:0 5.m0 m.5a "0.05-0.55 mm

0.5m .H.5H 0.50 5.:H 1:.05-5.00 0.0: 0.:H um.05.0.H: mmm mam smmanoaz
.Q.m.H wooonoMHHm ommz owcmx smog omowm conmHsmom

weHmseax meaa 55H0paaa agree
.mucoerodxm mmonopmmp

mde mop CH mdzoow mcfiomwnp Hanmm 0:0 mm .mmz 50 meHmHz now some

 

 

one owner

.:m magma

.7“
\ 3
w

DISCUSSIUh

The objective was to study the effectiveness of
continuous inbreeding (selfing) with selection compared
to three cycles of recurrent selection. Direct comparison
in the nursery between recurrent series and inbred series
was difficult due to inbreeding depression which masked the
effectiveness of the breeding methods.

Results with testcrosses indicated that recurrent
selection was no more effective than continuous inbreeding
with selection. These results were not in agreement with
those obtained by others to improve yield, oil content,
and leaf blight resistance in corn. Recurrent selection
was more effective than continuous inbreeding with selection.

Self-fertilization leads to a rapid increase in
homozygosity: F2, 50%; 81, 75%; 82, 67.5%; 53, 93.7%.

Rapid fixation of genes occurs with such a method. Any
breeding scheme which involves a rapid fixation of genes
imposes very definite restrictions Upon the effectiveness

of any selection which may be practiced (22). These limita-
tions are: (1) large number of genes, (2) masking effects
of environment, (3) complicated system of gene interaction,
(h) inadequate methods of isolating and evaluating lines.

Recurrent selection is a breeding scheme that offers
some Opportunity to minimize the rapid increase in homozygosity

which is led by self-fertilization. The basis of recurrent

O\
0".“

selection is to intercross selected plants, or bulked pollen
from the selected plants is used to pollinate them, and the
intercrossed pepulations serve as a source of material for
the next cycle of selection. Theoretically, genes for the
desired characteristic should be more concentrated in the
recurrent selected pOpulation.

The following reasons are proposed to explain why
recurrent selection was not more effective than inbretding
with selection to improve maturity, taking the silking date
as a criterion.

Selection in the recurrent selected series was at
a disadvantage compared to continuous inbreeding because of
the necessity for the ten earliest and ten latest maturing
plants to flower at the same time in order to effect the
intercross. In the recurrent selection program only one
plant in each row was selected to maintain inbreeding pressure
at a minimum. Therefore, a few of the earliest plants in a
particular row shed pollen at a time when there were no plants
in other rows with receptive silks. These early plants were
lost from the population. In the continuous inbreeding pro-
gram, such plants were saved. hecurrent selection for late
maturity presented a similar problem. Bulking pollen of the
ten latest plants required that there be ten plants in flower
at the same time. Therefore, slightly earlier plants were
chosen in the recurrent selection series than in the inhreed-

ing series.

69

Rapid fixation of genes and equal effectiveness of
the two breeding methods tend to indicate that few genes for
maturity were involved. Jones (15) using the formula for
calculating gene number given to Burton by Wright (2) found
that the gene number for silking date ranged from 5 to 19.
hohamed* in a recent article estimated the number of genes
for the same characteristic at three. Such few genes could
be isolated very rapidly with any breeding scheme.

Decline in fertility and vigor and delay in maturity
are among the general effects of inbreeding and are referred
to as inbreeding depression.

In the recurrent selection program only one plant
in each row was selected in order to maintain inbreeding
pressure at a minimum. Therefore, inbreeding depression would
be expected to affect the recurrent series less than the
inbred series and render their comparison very difficult.

This was confirmed by visual observation in the field.

One method to overcome the effects of inbreeding
depression would have been to inbreed the recurrent selections
(without any selection pressure for maturity) and compare them
with the inbred selections (figure A).

This would have reouired self-pollinating the HS
selections once, the ESQ selections twice, the H83 selections
three times, and then comparing the HS-l, ESQ-l-l and the
* Mohamed, Ali. Inheritance of ouantitative characters in

Zea Nays. 1. Estimation of the number of genes controlling
the time of maturity. Genetics. 713-72h, 1959.

hS3-l-l-l groups with the 81, 82 and S3 breeding groups.

This selfing of the recurrent selected groups could have been
done in winter in the greenhouse or in Florida since no selec-
tion was involved.

Figure A. Breeding diagram that should have been used to
compare recurrent selected and inbred series.

 

 

iégi‘ffi; _ , ____f 2
‘4 \\ --------- rirst cxcle of selection
\ h-l::>{gl----Pirst cxcle of selfirg
‘\ ---------- Second cycle of selection
‘ h3;-l-l~j---::>qSQ---—Second cycle of selfing
‘ .x\: --------- Third cycle of selection
hg3-l-l-l ------------ :>3g3----Third cycle of selfihg

In this study, selection pressure and inbreeding
pressure (selfing) were applied at the same time, (figure B)
and their effects could not be separated.

Figure B. Breeding diagram that was used to compare recurrent
selected and inbred series. This was the procedure

that was used.

hS3(—-——- H829“ HS (_..—.132

I>\\e -------- First cycle of selection

\,

\

Ka-l --------- First cycle of selfing

  

\\<---L ----- Second cycle of selection
gs-;-1:\82----Second cycle of selfing
------- Third cycle of selection

\
S3----Third cycle of selfing

Comparing inbred and recurrent lines in testcross is
possibly the best method to overcome the effects of inbreedirg
depression. Furthermore, since the ultimate objective is to
utilize inbred lines in hybrid combinations, their evaluaiion
for maturity is of primary importance.

Study of ear-to-row variance for days to silking
(Table 7B) and for ear moisture content at harvest (table 15)
indicated that gene frequency was at equilibrium after one
or two cycles of breedirf. Such an eouilibrium occurs in
the two following situations;

(I) When the homozygote has an advantage over the
heterozygote as in the case of selection for late-maturity,
the only stable equilibrium is at the point of fixation of
the preferred allele. Thus, if the expression for the rate
of change in allelic freouency per generation

sq (l-q)2

q : is set eoual to zero, it can

 

l - S (1—0)2
readily be seen that for values of 5 other than zero, the
only roots of q are zero or 1.
(2) tn the contrary, when the heterozygote is pre-
ferred over either homozygote, as possibly in the case of
early recurrent series, the eouilibrium value of q is deter-

mined in the following manner (16):

\l
o)

Box 1. Gene frequencies when heterozygcte is favored over
both homozygotes.

Genotype AA Aa aa Total

Initial q

freouency q‘ 20(l-q) 1-q2 l

Selective

value l-SA 1 1-58

Parents

selected q2(l-SA) 2q(1-q) (1-q2)(1_sa) 1-3 q22
-S€1( “(1)

 

The change in q in successive generations is then

q(l-q) Sa(l-o) - SA q

 

l‘SA Q2 - 53 (l-q)2
When this expression is equated to zero, the equili—
brium value of the freouency of A is

S.

-I
LA

 

 

QA -
SA / Se
and that of a is
SA
Ga =
SA / 5a

If there is only slight heterosis, variability
would decrease slowly and some pregress would be expected,

Figure C.

‘Q
Us)

Figure C. Decrease in variability when homozygrtes,
heterozygotes are favored.

  
 
 

>) 59% 4 ------------------- homozygotes favored

:1

r1

B 25% < ------------ partly heterotic effects
CD

2 12.5%

g 6.2 (----homozygotes favored

 

 

F2 01 u2 33

No progress for early silking date (Table 9) was
made after the first breeding cycle in either the recurrent
series or inbred series from the double-cross hichigan 51h.
how variance in the early recurrent series (Table 7b) attained
an eouilibrium at a higher value than that for the inbred
series or than that of the parental lines. It might be
assumed that progress for early maturity in this population
ceased, not because of lack of variability, but because the
heterozygotes were favored. The only way to overcome heterctic
effects would be to resort to testcrossing.

In the other populations, row variance in the early
breeding groups equated the variance of the early parental
lines (Table 7B), indicating that variability was exhausted.

In the late-maturity groups the recessive hcmozygote
was favored regardless of heterosis. After three cycles of
selfing, no variability would remain unless natural selection
played a role in removing the unfit homozygote from the

population (Figure D).

71+

Variability of the late inbred series as affected

 

Figure D. g
by natural selection which removes the unfit
homozygotes.

k 50% ; natural selection
w ,/ ,.

H v

'H A '

p (5% . ,

m , l/-

"—4 n ’f g, /

‘1 l d 0 2/33 // I

in 6 o C J /

 

 

{atural selection played such a role in the la.lf3

“‘7

X W25 population. Progress toward late-maturity ceased.
bany plants died or did not reproduce. These degenerative
aspects of inbreeding also affected late populations of the
emrly-x-late crosses. Lethal factors which may have teen
present in the original population or caused by mutation

were brought to the recessive ccnditicn by selection fcr

late maturity.

\3
\\n

SUHHAhY

The relative effectiveness of continuous inbreeding
with selection compared to three cycles of recurrent selec-
tion for corn maturity as measured by silking date was
studied in pOpulations derived from two double-cross hybrids
and four single-cross hybrids.

Direct comparison in the nursery between recurrent
and inbred series was difficult due to inbreeding depression
which masked the effectiveness of the breeding methods.
Decline in fertility and vigor and iiayed maturity were
among tie general effects of inbreeding and are referred to
as inbreeding depression.

Inbreeding depression affected the inbred series
more than the recurrent selection series. In reou‘rent
selection only one plant in each row was selected in order
to maintain inbreeding pressure at a minimum. With inbreeding
and selection, plants were selected within and among rows.

To overcome the effects of inbreeding depression, it
was suggested that the recurrent selections should be inbred
(without selection pressure for maturity) and compared with
the inbred selections.

Nursery results indicated that both methods had been
very effective in dividing the original F2 material into two
maturity groups and that fixation of genes was rapid. Fixation

of genes was very rapid in a population from an early-x-early

so
JT\

3
‘ Jl
V
"2
”W
C.
1...:
’1‘
m
m
.1
’1)

single cross (la.l§3 x We. .pid in a poyulation
from an early-x-late single cross (hSEHA x L317).

In a corn breeding program, the ultimate objective is
to utilize inbred lines in hybrid combination. Thus, evaluation
of lines for maturity in testcrosses is important. Comparing
inbred and recurrent lines in testcrosses is possibly the
best method to overcome the inbreeding depression effects
which masked the effectiveness of the breeding methods when
compared in the nursery.

Testcross results indicated that the early inbred
series were earlier than recurrent series in two of the
three cases, and the late inbred series were later than the
recurrent series in one of the three cases. however, the
differences between the two series were very small, indicating
that the two methods were equally effective. Very rapid
fixation of genes and ecual effectiveness of the two breeding
methods indicated that few genes for maturity were involved.

Lines that were both earlier and later in maturity
than the parental lines of the original crosses were obtained
in both recurrent and inbred series.

In all the testcross experiments there was a wide
range for yield among the selections within a maturity group,
indicating that selection for combining ability could be

effective within the maturity groups.

l.

0\

10.

ll.

\3
\]

LlTbbATUhL CITLD

Burton, G. W. Quantitative inheritance in Pear millet
(Pennisetum glaucum) Agron. Cour. #3: HOQ-HlT. 1951.

Burton, G. W. Quantitative inheritance in grasses. Sixth
International Grassland Congress. Proceedings:

Comstock, R. E., H. F. Robinson, and P. H. Harvey. A
breeding procedure designed to make maximum use of
both general and specific combining ability. Agrcn.
Jour. #1: 360-367. l9h9.

Duncan, D. B. Biometrics ll: l-h2. 1955.

Eckhart, h.C., and A.A. Bryan. Effect of method of
combining two early and two late inbreds of corn
upon the yield and variability of the resulting
double crosses. Jour. Amer. Scc. Agrcn. 32:
6H5-656. l9ho.

Henderson, M. T. Use of recurrent selection in improve-
ment of a predominantly self fertilized plant, Upland
cotton. Abstracts of the annual meetings of the
American Society of Agronomy, 1953.

Hull, F. H. Recurrent selection for specific combining
ability in corn. Jour. Americ. Soc. Agron. 37:
l3l+-ll+50 191+50

Frey, K. J., B. Brimhall, and G. F. Sprague. The effects
of selection upon protein Quality in the corn
kernel. Agron. Ml: 3Q9-HO3. 1949.

Jenkins, M. T. The effects of inbreeding and of selection
within selfed lines of maize upon the hybrids made
after successive generations of selfing. Iowa State
College Jour. Sci. 9: h29-h50. 1935.

Jenkins, M. T. The segregation of genes affecting yield
of grain in maize. Jour. Amer. Soc. Agron. 32:
55-630 191*00

Jenkins, M. T., Alice Roberts, and William Findley.
Recurrent selection as a method for concentrating
genes for resistance to Helminthosporium turcicum
leaf blight in corn. Agron. qur. #6: 89-9#. 195%.

1#.

15.

16.

17.

18.

19.

20.

21.

23.

21+.

Johnson, I. J. Lifectiveness of recurrent selection
for general combining ability in sweet clover, A
Nelilotus offioinalis. Agron. Jour. ##z #76-#ol.
1952.

 

Johnson, I. J. Further progress in recurrent selection
for general combining ability in sweet clover.
Agron. Jour. #8: 2#O. 1956.

Johnson, I. J. and A. 5. L1 Banna. Effectiveness of
successive cycles of phenotypic recurrent selection
in sweet clover. Agron. Jour. #9: 120-125. 1957.

Jones, Champ. An inheritance study of corn maturity.
Unpublished Ph.D. thesis. Michigan State University.
East Lansing, 1952.

Lerner, E. I. The Genetic Basis of selection. Wiley é
Sons Inc. New York. 1958.

Li, C. C. POpulation genetics. University of Chicago
Press. 1955.

Lonnouist, J. h. Recurrent selections as a means of
modifying combining ability in corn. Agron.
Jour. #3: 311-313. 1951.

Lonnquist, J. H. and D. P. McGill. Performance of
corn synthetics in advanced generations of synthesis
after two cycles of recurrent selection. Agron.
Jour. #8: 2#9-252. 1956.

thill, D. P. and J. B. Lonnquist. Effects of two cycles
of recurrent selection for combining ability in
an open pollinated variety of corn. Agron. Jour.
#7: 319-323. 1055.

Peacock, H. A. and C. P. Wilsie. Selection for resistance
to seed pod shattering in birdsfoot trefoil. Lotus
ccrnicus L. Agron. Jour. #9: #29-#32. 1957.

Sprague, G. F. Corn and corn improvement. Agronomy
monographs 5. Academic Press, New York. 1955.

Sprague, G. F. and B. Brimhall. Relative effectivenes,
of two systems of selection for oil content of the
corn kernel. Agron. Jour. #2: 83-86. 1950.

Sprague, G. F., P. A. Miller and B. Brimhall. Additional
studies of the relative effectiveness of two systems

of selection or oil content of tv .
Agron. JOUP. #: 329-331, lo52.Je corn kernel.

( .

"Student” A calculation of the minimum genes in Winter's
selection experiment. Ann. Bug. 6: '7-o2.

Warner, J. K. A method for estimating heritability.
Agron. qur. ##: #26-#38. 1952.

Wright, Sewall. hvclution in mendelian pepulations,
page 111, .7-159. Genetics. 1930.

Yang, W. K. A study of the nature of genes controlling
hybrid vigor as it affects silking time and plant
height in maize. Agron. Jour. #1: 309-312. 19% .

APPENDIX

to

Figure I.
82
80 ,
78
76
7#
%)
.3 72
,2
H
"—1
U)
0
+9
8. 68
C13
:3
66 -
6h ,
62
60.

7O .

Nean days to silking for the four parental lines,
F1, F2 and the nine inbred and recurrent selection
breeding groups in the early and late maturity

series from kichigan 51B.

 

0 I413
\
\
,\ Ia.153 , HS-l-l
.‘\W25 /
\ // 2 P‘S’2‘l
/
x __/I ’ [183
0 W9 J,"
\ , /f
f f
\ i
H J/
hS-l-l

\ _fl .
\4/2; 223*

P F1 F~ I 11 111

C

generations
recurrent series

,1
/
I:
’I
\h
0 \l .
K
\“\\1,___W.S—
F2 I 11 11:

generations
inbred series

silking

vs to

U

Da

Figure 2. Nean days to silking for the two

F1, F2 and inbred and recurrent selection
groups in the early and the late maturity

from (H53 x 3E-11).

as
. HS-l-l
86 '
at. - .'
, , hSE-l
82 , x’ f
.( {1"
CO 1 ’2' . RS
- _ / , 3
78 \ x"
\.;>‘
76 35-11
7 tf - “\ y, /’X..
72 /
,7

7O .
68 ’

2 , as-I—I

I! ,/
66 , . R53 ‘¥>~\I ’

a “' ESQ-l
6L+ "
HS 3
'62 -

60 ‘ . , .

P F1 F2 I II III

Generations
Recurrent series

Jarental lines,
breeding
series

f\
II \\\
l' \
I x S

\\
kL’

/
,
I)!
1”
/
['1
Q
\
\
'\
\"x
e S ’2
F2 I II III

Generations
Inbred series

Figure 3.

Days to silking

89

87

85

81

79

77

75

73

71

69

67

Kean days to silking for the two parental lines, F1,
F2 and the nine inbred and recurrent selection
breeding groups in the early and late maturity series
from (k82#A x 3b-ll).

P F1 F2

Generations
L AAAAAAA rzf

F2 I II

Generations
Inbred series

111

Figure

to silking

U

Davs

8-3

El

79

75

73

69

67

#. Mean days to silking for the four parental lines, Pl,
F2 and the nine inbred and recurrent selection breed-
ing groups in the early and late maturity series
from hichigan 20D.

. RS-l-l
r /
‘ W10 /'f o S)
.- , ng'l ./.I
1 I , ,
. " Oh ea ,»
., I a £133 //
I! I
1' ,
,i' e
.. " 51A .1
\ .
‘ I
\
- y
- K 7
I E
" ‘ f i////\\\
r;1#._ _ ' \
o I 3‘1“]-
- o S?
‘\
5 4 ‘ a l \: H83 I I
P F1 F2 I II III F2 I II III
Generations Generations

Recurrent series IanEd SCTiQS

Figure 5. Mean days to silking for the two parental lines, F1,
F2 and the nine inbred and recurrent selecticn breed-
ing groups in the early andlate maturity series frcm
(Ia.153 x W25).

80 . , RS-l‘l
.0 1182-1
78 .
76 _ was _. 33
Ia. ‘,‘ i, H82 /’/
71+ - \ / ,x’
1‘ /
72 - = -
no 111‘. / //
L ‘1 ,0 e
«‘1 | 1‘,
Z 70 ' ii /
W4 1:
m a], / e
8 68 . :3 / “x, ~.\
I x l ‘
e v / ‘a
V) 3 ,l g
I ’ , ' 3
‘ A W/
61+ " . ‘3‘ ~77.\ hS-l-l
2‘ /’ |n
62 -c,\"/ 1302-1

60 , , ,
I"""P””"”F1m" Fw ”’I””"II’" Ill" “F;"'I II III
Genera ions Generations
Recurrent series Inbred series

Figure 6. Mean days to silking for the two parental lines, F1,

Days to silking

F2 and the nine inbred and recurrent selections breed-
ing groups in the early and the late maturity series

101 - from (NS2#A x L317).
99 -

97 , \ L317

95 ,

87 a , RS-l-l ,,
. '. ask-1 "
85. g ,’ ‘ -//
\ ,, . as3 f
83 i . /
H » / /
El ‘\ /¥ I \
x ,’ \
79 - x \ \
. \\
77 // ° \,
NS2#A RSg-l ‘\
75 . / \

P r1 r2 I ' II III F5 I II III
Generations Generations
hecurrent series Inbred series

Figure

Moisture content percentage

681

66

62
60
58

56.

an
42

#O

38

u)
0\

r 03
7. bean moisture content for the four parental lines}

F F6 and the nine recurrent selections and
9 C

inbred breeding groups of the early and the
late maturity series from Michigan 515. ,
3 N13 i
i .
E X
i Q 118*]. "’1 ,
S f/ 4
.If i
I I r
. / j
. f 4'
. f , BSg-l ,
C 1. - "Ng / /
e ‘ IN 7 w25 Ii
3 ‘18 . 153 ' c
‘ f
’ ' 'lp RS3
; /
. 7 ’ r
a 5 r
2 f
t ' ‘
i
i x
i. \ -
6 .\“ (I ‘8
e hsfl’l
' d e S 3
. I RS-1-1 J
.. \
\~.
\.
‘5 RS3
”FINN ”F1 F2 I II I l I F2 I I I III
Generations Generations

Recurrent series lnbred series

Loisture content

a.. V

66 lires ' 7: arui the
and inbred bIeeding
and late maturity

L ‘ . + ,2 .hS-l-l 1 _
Pi gure t. 168“ utisture center“t o. / tne two parental

y nine recur rent selectitns
groups of tre earlyi
series from (B53 X

V

\

. x X
6 x ,7 .
in} i ,
CU . t'
+3 \ ' l
g: 5 2 - ‘ ’ A
CD 3 ’/
U \- l
'H .
0) /
E
04 5 O - K 9, .
x/ \
i / ‘,
f '
Lt 8 ' /’ X

at. i l \,

#0

i
\ f fl
38 - \ ‘ D
. H53
36

1‘“ 0 {18-1 1
31+ - .
\
\

(.AJ cw
C) R)
.5- .
/W
./ 1
CI"
3- m
U) W
D
I
H

P F1 F2 I II III F; I II III
Generations AF
necurrent series nbr:

my; USE 0m
APR 4 1962 M

""" "’- 7":a ' yes-:1 v“,
£3" an M5 .3
.n, L”, .. '.,-.'_ .

“,.

 

 

TRTE UNIV. LIBRRRIES

1 lilHdllliiiIId‘H if“ H WNW H 1%
31 29

1,“
IE!
3106921012