llllflllllllfl lllllllflllllll

300667 8852

 

 

 

 

 

- rag-.1, 11"“

#4
3 ms ma?
“2:212? 5.: 5:: :2» 533.9

U4 *1 at”: V? effigy?

THESlS

 

'— p- m “—1" n-v

,-

§
taut. ,I...,pw-‘a v—vv— .. . F.“”

This is to certify that the

dissertation entitled

Genetic Parameter Estimates For Weathering and
Growth in Ring-necked Pheasants (Phasianus
colchicus) Population

presented by

Talal H. Hussein

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Poultry Science

 

Javgiém

Major professor

Date 12/29/82

 

MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

 

MSU

 

 

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

 

 

LIBRARIES .
Algggg.!-;_ your record. FINES w1ll
be charged if book is
returned after the date
stamped below.
“was

 

 

 

GENETIC PARAMETER ESTIMATES FOR FEATHERING AND GROWTH IN RING-
NECKED PHEASANT (PHASIANUS COLCHICUS) POPULATION

 

By

Talal H. Hussein

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Animal Science

1983

/37-2799

ABSTRACT

GENETIC PARAMETER ESTIMATES FOR FEATHERING AND GROWTH IN
RING-NECKED PHEASANT (PHASIANUS COLCHICUS) POPULATION

 

By

Talal H. Hussein

Data of 1,200 pedigreed pheasants produced in three generations
of a bi-directional selection experiment for rapid and slow feathering
rate were analyzed for genetic parameter estimates. Changes in traits.
body weight at various ages and egg production, correlated with
feathering were also assessed.

Realized heritability, analysis of variance and covariance for
hierarchical model with unequal sub-class numbers and intra-sire
regression of offspring on dams were the methods employed to estimate
heritability of feathering and body weights.

0n calculated feathering scores from 80.2 to l72.8, the effect
of upward and downward selection for feathering rate resulted in
average feathering differences of 3.5, 4.9 and 7.4 in male progeny
and 3.1, 5.6 and 9.0 in female progeny between the two divergent
lines, respectively.

Greater progress was attained by selection for rapid feathering
in females than in males, indicating the existence of sexual

dimorphism in feathering in ring-necked pheasants.

The average realized heritability for feathering was .08 for
the slow feathering line and .36 for the rapid feathering line.

Based on intra-sire regression of offspring on dams, the
heritability estimates for feathering were .60 and .ll for the plus
and minus selected lines, respectively. From the averages of full-
sib correlations within each line, the heritability estimates for
feathering in the male progeny were .42 and .33 for the slow and
rapid feathering lines, respectively. In the female progeny, the
heritabilities of the two respective lines were -.08 and .21.

The low to moderate estimates of heritability of feathering
suggest that improvement in feathering in ring-necked pheasants can
be most efficiently achieved by a combination of individual and
family selection at early ages. As for the correlated traits,
heritability estimates were within a range of .22 to .53 for four-
week body eight, .16 to .66 for eight-week body weight and .10 to .72
for twelve-week body weight. These results suggest that mass
mass selection can be efficient in improving body weights of pheasant
birds.

Correlation results between feathering and body weights were
positive, indicating that selection for rapid feathering would improve
growth rate and vice versa.

Egg production significantly (p<.05) declined over generations

due to selection for rapid feathering.

DEDICATED

TO MY ENTIRE BELOVED FAMILY

ACKNOWLEDGEMENTS

The author is forever indebted to his academic advisor, Dr.
Theo H. Coleman, who is one of the very few truly outstanding popele
I have come across during my entire life. His guidance and encourage-
ment coupled with his capacity to reason and understanding will be
an everlasting memory.

Grateful acknowledgements are sincerely expressed to Dr. Cal J.
Flegal for his continuous assitance and concern during my graduate
study and the course of this research.

My deepest gratitude and appreciation are extended to Dr.
William T. Magee for his untiring guidance and inspiration during
my stay at Michigan State University. His critical discussions and
concise appraisal of research projects were most invaluable learning
experiences.

Sincere appreciations are due to Dr. Ronald H. Nelson, chairman
of the Animal Science Department, and the Department of Natural
Resource's Wildlife Division for providing the funds and materials
necessary for conducting this project.

A special acknowledgement of gratitude must go to the faculty
of Poultry Science and the staff of Poultry Science Research and
Teaching Center for making my duration of stay at Michigan State
University and East Lansing so much like home and made it more

enjoyable.

1v
I also wish to thank my fellow graduate students, especially
Glenn Carpenter and M. Villarreal for their assistance and comrade-
ship.
The author is also appreciative to the Iraqi Government for
making available the funds for my graduate study abroad.
And, finally, my love and appreciation go to my entire family

for their support during my endeavor.

TABLE OF CONTENTS

SECTION PAGE
LIST OF TABLES .........................
LIST OF FIGURES ........................

I. INTRODUCTION ....................... 1

II. REVIEW OF LITERATURE ................... 3

Genetics and Mode of Inheritance ............ 3

Feathering and Growth ................. 6

Feathering and Nutrition ................ 8

Feathering and Environment ............... 10

Sexual Dimorphism of Feathering and Growth ....... l4

Heritability ...................... l8

Methods of Estimating Heritability ........... 19

Genetic and Phenotypic Correlations .......... 24

Heritability of Feathering and Growth ......... 26

III. MATERIALS AND METHODS .................. 29

Selection and Subsequent Matings ............ 35

IV. STATISTICAL PROCEDURE .................. 42

Correlations Between Feathering and Correlated Traits . 46

V. RESULTS AND DISCUSSION .................. 48
Effect of Mass Selection ................ 48
Heritability Estimates of Feathering .......... 68
Heritability of Associated Traits ........... 78
Correlations Between Feathering and Other Traits. . . . 80
Egg Production ..................... 84

VI. SUMMARY AND CONCLUSIONS ................. 88
BIBLIOGRAPHY .......................... 93
APPENDIX ............................ l03

TABLE
1.

LIST OF TABLES

Results of Selection for Slow and Rapid Feathering
Within the Control and the Slow and Rapid Feathering
Lines Expressed by Averages of Feathering Score i
Standard Deviation for a Period of Three Generations
by Number, Replicate and Sex ..............

The Averages of Four-week Tail Length in cm i Standard
Deviation for the Control and the Slow and Rapid
Feathering Lines Over Three Generations of Selection
for Slow and Rapid Feathering by Number, Replicate
and Sex .........................

The Average One-day Old Body Weight in gm i Standard
Deviation for the Control and the Slow and Rapid
Feathering Lines Over Three Generations of Selection
for Slow and Rapid Feathering by Number, Replicate
and Sex .........................

The Average Body Weight at Four Weeks of Age in gm i
Standard Deviation for the Control and the Slow and
Rapid Feathering Lines Over Three Generations of
Selection for Slow and Rapid Feathering by Number,
Replicate and Sex ....................

The Average Body Weight at Eight Weeks of Age in gm i
Standard Deviation for the Control and the Slow and
Rapid Feathering Lines Over Three Generations of
Selection for Slow and Rapid Feathering by Number,
Replicate and Sex ....................

The Average Body Weight at 12 Weeks of Age in gm 1
Standard Deviation for the Control and the Slow and
Rapid Feathering Lines Over Three Generations of
Selection for Slow and Rapid Feathering by Number,
Replicate and Sex ....................

The Magnitudes and Signs of the Phenotypic Changes of
Feathering Scores and Associated Traits in Two Sub-
sequent Generations of Selection for Slow and Rapid
Feathering by Lines and Sex ...............

vi

PAGE

49

51

55

56

57

58

65

vii
TABLE PAGE

8. Feathering Score Averages of Three Generations, Selected
Parents and Selection Differentials and the Regression
Coefficients of Feathering Means on Generations ..... 69

9. Regression Estimates of the Realized Heritability on
Generations for Feathering of Selected Lines ...... 70

10. The Heritability Estimates 1 Standard Errors of Feather-
ing and Body Weights by Intra-sire Regression of
Offspring on Dams .................... 71

11. Estimates of Heritability i Standard Errors of Feather-
ing and Body Weights of Male and Female Progeny
from Parent Variance Components ............. 74

12. Estimates of Phenotypic and Genetic Correlations i
Standard Errors Between Feathering and Body Weights
of Male and Female Progeny by Lines and Sex ....... 81

13. The Averages of Egg Production, Fertility and Hatch-
ability of Pheasant Breeders Selected for Slow and
Rapid Feathering Compared to the Non-selected Control
by Lines, Replicates and Generations .......... 85

APPENDIX TABLE

Al. Pheasant Starter Ration Fed to Chicks from One Day to
Six Weeks of Age .................... 103

A2. Pheasant Grower Ration Fed to Chicks from Six Weeks to
13 Weeks of Age ..................... 104

A3. Pheasant Flight Ration Fed to Chicks from 13 Weeks to
24 Weeks of Age ..................... 105

A4. Pheasant Breeder Ration Fed to Birds from Time of
Stimulatory Lighting to End of Egg Production Period . . 106

A5. The Standard Model of the Nested Analysis for the Deter-
mination of Variance Components of Feathering Scores
and Other Traits Among Sires, Dams Within Sires and
,Full-sib Progeny .................... 107

A6. The Variance and Covariance Components of Feathering
Scores and Associated Traits from Selection for Slow
and Rapid Feathering for a Period of Three Generations
by Lines and Sex .................... 108

viii

APPENDIX TABLE PAGE

A7.

A8.

A9.

A10.

A11.

A12.

A13.

The Mean Square Values Between Sires, Dams and Within
Full-sibs for Feathering Scores and Associated Traits
from Selection for Slow and Rapid Feathering for a
Period of Three Generations by Lines and Sex ...... 111

The Calculated Values of the Degrees of Freedom of
Sires, Dams and Progeny and the Coefficient of.
Variance Components by Lines and Sex Over Three

Generations of Selection for Slow and Rapid Feathering . 112

The General Model of Analysis of Variance and Covariance
Components of Feathering and Other Traits ........ 113

The Mean Squares and Mean Cross Products of Dams and
Progeny of Two Generations of Selection for Slow and
Rapid Feathering and Body Weights ............ 114

Analysis of Variance of Egg Production by Lines, Genera-
tions and Interaction .................. 115

Analysis of Variance of Feathering Response to Selection
by Lines, Generations and Sex .............. 116

The Transmitted File for the GENSTAT Analysis of Variance
and.Covariance of Unbalanced Data ............ 117

LIST OF FIGURES

FIGURE

1. The Observed Mean Response to Direct Selection for
Feathering in Males and Females of the Slow and Rapid
Feathering Lines Compared to the Non-selected Control
Line (Random) over Generations ..............

2. The Observed Mean Changes in Tail Length of Males and
Females in the Slow and Rapid Feathering Lines Compared
to the Non-selected Control Line (Random) over
Generations .......................

3. Changes in One-day Body Weight of Males and Females of
the Slow and Rapid Feathering Lines Compared to the
Non-selected Control Line (Random) over Generations . . .

4. The Observed Mean Changes of Four-week Body Weight in
Males and Females of the Slow and Rapid Feathering
Lines Compared to the Non-selected Control Line
(Random) over Generations ................

5. Changes in Eight-week Body Weight of Males and Females
of the Slow and Rapid Feathering Lines Compared to the
Non-selected Control Line (Random) over Generations . . .

6. Changes in 124week Body Weight of Males and Females of
the Slow and Rapid Feathering Lines Compared to the
Non-selected Control Line (Random) over Generations . . .

7. The Observed Mean Changes in Egg Production in the Slow

and Rapid Feathering Lines Compared to the Non-selected
Control Line (Random) over Generations ..........

ix

PAGE

52

53

59

61

62

63

87

I. INTRODUCTION

The ring-necked pheasant, Phasianus colchicus, a native of the
Far East, is becoming one of the most popular game birds in the
world.

Two serious problems confronting pheasant raisers have been poor
feathering and feather picking. In a survey taken of game bird oper-
ations in the United States by Dodsen (1971) and cited by Shellen-
barger (1976), 67% of commercial pheasant raisers encounter the poor
feathering condition and a feather picking problem. These, indeed,
can account for a substantial economic loss to the producers. Poor
feathered birds appear less desirable to the consumer which, in turn,
reduces the interest for these birds. Hence, game farms and shooting
preserve operations must emphasize well-feathered birds to revive
this inudstry. In addition, early feathering is important in that
it enables the birds to be released at younger ages for flighting
purposes. Thus, some economy would be realized by saving the cost
of feed consumed, housing, labor, etc. Growth rate is another char-
acteristic of fundamental importance and of particular interest to
ring-necked pheasant raisers.

Although both feathering and growth rate are different in their
mode of inheritance, several workers have reported that in general
purpose breeds of chickens, rapid bird-feathering is associated with
rapid growth rate (Martin, 1929; Schnetzler, 1936 and Glazener and
dull, 1946).

2

The genetic improvement of any economic trait can be best
achieved by selection. The amount of improvement secured by selection
depends on the effective use of genetic variation in the population.
When two or more traits are involved in the breeding plan, knowledge
of genetic and environmental correlations among the traits is neces-
sary to predict the response of selection of traits not directly
selected. Heritability, which is the fraction of the total variance
attributable to the average effects of additive gene action, is the
key parameter for any given population to improve the desired traits.

The primary objectives of the present study were:

1. to estimate the heritability of feathering rate of ring-
necked pheasants and correlations involving feathering and body weights
at various ages, and

2. to determine the simultaneous responses of unselected traits

as a result of selection for feathering.

II. REVIEW OF LITERATURE

In game birds, the formation of lines from one population with
different feathering and growth response of genetic selection, nutri-
tion and management is a relatively unexplored area of investigation.
However, numerous studies of genetic and environmental influence on
feathering and growth rate have been reported among other species of

poultry.

Genetics and Mode 9: Inheritance

 

From the development of lines of chickens differing in feather-
ing, rapid versus slow, several workers have reported that feathering
is a trait that could be improved through genetical selection plans.

The action of the recessive early feathering sex-linked gene, k,
was first reported by Serebrovsky (1922) and Warren (1925). Plumart
and Mueller (1954) described early feathering broiler males as
superior to late fbathering ones in that they feather more rapidly
and possess few pinfeathers at twelve weeks of age.

Radi and Warren (1938) ascribed the early feathering to a dosage
effect of k gene and to the action of endocrine glands. Jones and
Hutt (1946) demonstrated that in White Leghorns, the tardy-related
series, a multiple allelic series of genes, prvents appearance of
the sex-linked rapid feathering trait, and is responsible for slow
development of the tail feathers, of secondary feathers of the wings

3

4

and of contour feathers over the body up to eight weeks of age. The
same conclusion was reported in Rhode Island Red chicks by McGibbon
and Halpin (1946). Siegel et al. (1957b) studied the phenotypic
expression of the sex-linked late feathering gene in homozygous,
heterozygous and hemizygous (K-) chicks. Differences to five weeks
of age were reported to be significant between homozygous late
featering males compared to the other two groups. It was also de-
termined that the sex-linked gene for late feathering is incompletely
dominant to its allele and shows dosage effects. Significant vari-
ations were observed between KK males and K- females and between KK
males and Kk males as well; the KK males feathered poorer than either
Kk or K- chicks.

Somes (1969) studied the mode of inheritance of a naked adult
female and a no tail and flight feathers male of the same cross.
From the pedigree of those birds, it was known that the feather mutant
character was indeed a dominant sex-linked trait. The mutant allele
at the K locus was responsible for delaying feather growth and reduc-
ing comb size. A series of crosses (mutant male with mutant female;
normal male with mutant female and the reciprocal cross) were made.
From the first cross, an equal proportion of mutant and normal female
progeny were obtained, while all the male progeny were normal. In
the second cross, all the female progeny were normal and all the
male progeny were mutant. However, the reciprocal cross produced 50%
mutant and 50% normal in both male and female offspring. From these

results, Somes concluded that a third allele at the K- locus on the

S
sex-chromosome, symbolized K", is dominant over both K and k alleles
and caused an extremely delayed feathering.

In view of the previous studies, Somes (1970) further investi-
gated the effect of the sex-linked Kn allele on various quantitative
traits from mating mutant and normal White Leghorn chickens. In
addition to the late feathering and smaller comb, birds possessing
the Kn gene showed significantly reduced body weight and less total
egg production than normal birds. Also, a slower metabolic rate,
poor hatchability, poorer livability, smaller eggs, more misshapen
eggs, and later sexual maturity were noticed in the mutant birds,
compared to the normal ones.

In another paper of Somes (1975), the effects of testosterone
propionate, at a rate of 3 mg/lOO gms body weight, on comb size,
uropygial gland and other traits of mutant, Kn chicks were studied.
Birds of both sexes of five genotypes kk, Knk, KnK", K- and K"-
were sacrificed to measure the internal organs as percent of body
weight. Results indicated that K'1 birds had a significant hypertrophy
of the uropygial gland and reduced comb size. The heart and adrenal
gland were significantly larger in birds which carried the kn allele
than in those which lacked the allele. Testosterone-treated birds of
KK" and K"- genetic make up were smaller in comb size than, and
exhibited delayed feathering as compared to, testosterone-treated non-
mutant chicks (KK and K-).

Similar studies by McGibbon (1977) were conducted on slow
feathering birds. He also deduced that the extremely slow feathering

is due to a mutation at the KS locus on the sex-chromosome and that

6
it is dominant to both the late K and to the early k feathering

genes.

Feathering and Growth

 

Although feathering and growth rate are different in their mode
of inheritance in that feathering is affected by few genes whereas
growth rate is a multifactoral or polygenic trait, the relationship
between these two traits has been widely investigated.

Several workers have reported that growth rate and other per-
formances are associated with the degree of feathering. According
to Martin (1939), the degree of back feathering in Barred Plymouth
Rocks were noticed to be closely related with the rate of body growth
and egg production. Heavier chicks and better egg production were-
obtained from early feathering hens. Likewise, Lowe et a1. (1965)
reported that larger eggs, earlier sexual maturity and heavier one-
day old chicks were obtained from rapid feathering hens. Warren and
Payne (1945) using New Hampshire chicks found that those carrying
the sex-linked gene fbr rapid feathering were heavier in weight at
twelve weeks of age than those lacking the gene. Hutt (1949) indi-
cated that birds which feather rapidly, weigh more at broiler age
than those birds which feather slowly. Glazener and dull (1946)
concluded that there is a relationship between the number of second-
ary feathers at hatching and body weight at broiler age. Those in-
dividuals with six or more secondaries were heavier in weight than

those with fewer secondaries.

7

The work of Hurry and Nordskog (1953) with Barred Plymouth Rocks
and New Hampshires showed that feathering and growth rate had a
fairly high phenotypic correlation, .39. According to Goodman and
Muir (1965), sex-linked rapid feathering birds showed significantly
heavier body weight (+.11 pounds) than slow feathering ones. Saeki
and Katsuragi (1961) studied the effect of early and late feathering
genes on weight of newly hatched chicks and post hatching growth
rate of New Hampshires, White Leghorns and their crossbreeds. A
significantly heavier weight at hatching time were found in early
feathering groups of New Hampshires than in late feathering ones.
Leghorn birds did not appear to have a significant difference in the
degree of feathering. However, the crossbred progeny from New
Hampshire males and White Leghorn females and from their reciprocal
cross showed significant differences in one-day body weight.

On the contrary, some reports described the relationship between
feathering and growth rate as not so intimate or correlated. Hays
(1951) reported that in a large population of Rhode Island Reds,
the sex-linked gene for rapid feathering had no effect on male's
body weight. However, female's records showed that as the degree of
back feathering increased, a significant increase in body weights
was observed at eight and twelve weeks of age. Godfrey and Farns-
worth (1952) indicated that the gene k for sex-linked rapid feather-
ing was found to have no relationship to body weight or mortality
rate to ten weeks of age. The lack of an intimate relationship
between feathering and growth rate was also reported by Plumart and

Mueller (1954) and Sheridan and McDonald (1963).

Feathering and Nutrition

It has long been known that the unique covering of feathers is
for the purpose of protection and heat regulation and is necessary
for flight. Feathers make up about 7% of the live weight as reported
by Jull (1938) and account for 4 to 9% of the empty live weight,
depending upon the age and sex of the bird (Card and Nesheim, 1972).

Since feathers contain various amino acids, vitamins and minerals,
an adequate supply of good quality protein and a balanced ration are
essential to secure normal feather growth and coloration.

Donaldson et a1. (1955) reported that as the energy level of

broiler diet increased from 35.7 to 48.6 calories of productive energy
per pound for each percent crude protein, poor feather condition at
market age was observed. Schaible (1970) found that chicks and poults
deflect large portions of their sulphur amino acids intake to the
rapid growth of feathers which contain 10% cystine. He also concluded
that a rapidly feathering Leghorn breed showed a marked feathering
response to the supplementation of arginine and 91YC106 COWPBVEd t0
the slow feathering Plymouth Rock breed which showed less response.
It was suggested that arginine and glycine are factors affecting the
feathering rate of young birds. However, Waterhouse and Scott (1962)
reported that female chicks carrying the sex-linked gene k for rapid
feathering did not require a higher concentration of dietary glycine
than did slow K feathering males.

Schaible et al. (1947) indicated that any kind of malnutrition
may cause feather picking and cannibalism in Leghorn chicks. Scott

and Reynolds (1949) noted that supplying a pheasant diet with 0.5%

9
magnesium results in relative prevention of feather-picking condi-
tion. Day and Hill (1959) reported an improvement in feathering
when turkey poults received a diet of 50 ppm of zinc in their starter
and grower rations. However, Supplee et a1. (1961), feeding zinc in
poults‘ diet at 42 ppm and 50 ppm in a corn-soy ration, showed an
extensive feather abnormality as a result of degenerative changes
following hemorrhage in the pulp of the immature feathers. Daghir
and Balloun (1963) and Gehle and Balloun (1965) reported that chicks
which received a vitamin B-6 deficient diet showed a degenerative
change in the wing feathers which was cured by pyridoxine supple-
mentation. Supplee (1966) also substantiated that an abnormality,
characterized by degenerative changes in the quills following hemor-
rhage in the feather pulp, occurred in the flight feathers of turkey
poults fed a diet deficient in vitamin E and selenium.

Thyroid studies revealed that thyroid secretion might normally
play an important role in the complex of factors responsible for
feather development. Ringer (1965) stated that thyroidectomy results
in feather structure alterations with loss of barbules and color.
Schultze and Turner (1945) reported that, in a fast feathering
Leghorn breed, higher thyroid secretion rate was noticed than in a
slow feathering breed of White Plymouth Rocks. Cole and Reid (1924)
observed, in adult stock, that birds which received desiccated
thyroid in their ration showed more rapid feather replacement than
others. Parker (1943) reported that adding small amounts of thryro-
protein to the chicken ration resulted in an increase in the rate of

feathering. Irwin et al. (1943) noted a slight increase in growth

10
rate and feathering of White Plymouth Rock chicks receiving 36 gm
of thyro-protein per 100 lbs. of mash. In further studies, Turner
et al. (1944) also had some positive response from feeding thyro-
protein at a rate of 45 gm per 100 lbs. of feed to Barred Plymouth
Rock cockerels. Wheeler et a1. (1948) reported that male chicks
fed 0.02% thyro-protein to 12 weeks of age feathered early
compared to the control group. Boone et a1. (1950) indicated that
an average of 23 gm of thyro-protein per 100 lbs. feed significantly
increased the feathering rate of slow feathering Rhode Island Red

chicks.

Feathering and Environment

Of the environmental factors that birds are subjected to, light
is the most variable with a definite influence on feathering and per-
fbrmance. Numerous suggestions have been made by poultry workers for
using different light programs to achieve best feathering and pro-
duction.

Mueller et al. (1951), who presented data based on experiments
with turkeys, indicated that light restriction to 12 hours daily
from 4-16 weeks of age and to 10 hours daily from 17-28 weeks of age
had reduced the molting condition of first juvenile plumage. Less
pinfeathers were shown in these birds than in ones kept under natural
daylight or subjected to 15 hours of light daily. Too, Moultrie et
a1. (1955) found that Beltsville White turkeys, reared under 10 hours
of light daily, molt significantly less than those kept on continuous

light. They also noticed relatively less pinfeathers in the

ll

restricted light group at market age, which was ascribed to the slow
rate of molting condition of the postjuvenile body feathers. Moultrie
et a1. (1954) worked with two groups of New Hampshire-White Plymouth
Rock crossbreeds reared on different light regimes, 10 and 15 hours
a day, during the first six weeks of age. Thereafter, both groups
were divided into four finishing light regimes, 5, 10, 15 and 24
hours of artificial light daily to 12 weeks of age. At the end of
the second rearing period those kept under continuous light were
significantly better feathered than other groups. Schumaier et a1.
(1968), using different kinds of lights, indicated that red light
showed a marked effect in prevention of the feather picking condi-
tion and cannibalism during the rearing period of White Leghorn
chickens. Likewise, McWard et a1. (1972) reported that feather pick-
ing and cannibalism were significantly reduced by the use of red
light in White Leghorn chicken houses. Shellenbarger (1976) con-
ducted a series of experiments to study the effect of lighting
systems and debeaking treatments on feathering and cannibalism of
ring-necked pheasants to 8 weeks of age. Four different light
colors were used, red, blue, subdued white and darkness (0.107 lux).
Results indicated that red, blue and partially dark lights produced
superior plumage with less cannibalism than did the white light
system.

The effects of different lighting systems on egg production and
growth rate of broiler chickens and turkeys have been extensively

studied. Carson et a1. (1956) reported that a supplemental light

12
in poultry houses significantly increased egg production compared
to non-lighted hens during an ll-week period. Dakan (1934) observed
better egg production from pullets reared under red light than from
those reared under blue light. Similarly, Platt (1953) reported a
significant difference in egg production in favor of hens kept with
supplemental dim red light as compared to those exposed to daylight
only. These findings were not in agreement with those reported by
Schumaier et a1. (1968) who reported that white light was superior
to red or green light in increasing egg production of White Leghorn
chickens.

Birds subjected to intermittent light produced more eggs than
their counterparts exposed to the same amount of continuous light
(Wilson and Abplanalp, 1956). However, Bell and Moreng (1973) found
no significant differences in egg production from birds under inter-
mittent and those under continuous light.

Ahemeral studies, light-dark cycle more than 24 hours, revealed
that this sort of lighting system caused a noticeable reduction in
egg production, fertility and hatchability compared to the conven-
tional 14:100 day cycle (Proudfoot, 1980).

In broiler chickens, Kleinpeter and Mixner (1947) and Weaver
and Siegel (1968) reported better growth rate in broiler chickens
raised under continuous lighting regimes than in those subjected to
periods of darkness.

The intensity of light appeared to have marked influence on

growth rate of broiler chickens. Skoglund and Palmer (1961) found

13
better average weights with birds under low intensity, 5.4 lux, than
those under higher intensity, 22, 54, 110 or 1300 lux. Similarly,
Bean et a1. (1965), Deaton et al. (1970) and Bacon and Touchburn
(1976) reported that lower light intensity tends to enhance growth
rate of broiler chickens.

Lighting studies involving intermittent and continuous programs
along with different intensities have produced variable results.
Deaton et al. (1976) found that continuous light of 12.9 lux signif-
icantly improved bird's body weights at five and eight weeks of age
when compared with those of birds subjected to light and dark cycles.
Quarles and Kling (1974) conducted two experiments with three light-
ing regimes on broiler chickens. Continuous illumination, illumina-
tion for 12 hours followed by 15 minutes of illumination each two
hours, or 15 minutes of illumination each two hours were employed.
No significant differences were observed in body weights at four
and seven weeks of age under the three treatments. Likewise, Hulan
et a1. (1980), working with turkeys, reported no significant weight
differences between light treatment of 4L:20, continuous 23L:lD or
total darkness.

Other environmental factors possibly affecting feathering and
growth rate are bird density and litter condition. Wells (1972)
stated that hybrid pullets, light-weight, showed poorer feathering
condition when they were reared in densely populated groups. Harris
et a1. (1980) substantiated that broiler chickens on new litter
showed a significantly better feathering than those reared on built-

up litter. In a small number of publications, the situation with

14
regard to feather condition and cannibalism was found to be attrib-
uted to some genetic variations.

Richter (1954) observed considerable differences in the phenomenon
of feather pecking. From the results of cross breeding experiments,
he concluded that feather eating is a special aptitude which is pri-
marily related to hereditary characteristics. The same conclusion
was reported by Cuthbertson (1980) from an investigation of feather-
pecking behavior. At about four weeks of age, experimental birds
were classified as peckers or pecked. His results revealed that the
incidence of peckers was more genetical than the incidence of pecked
birds among different families. Thus, he suggested that selection
is a feasible means to reduce the occurrence of the feather-pecking
condition. Hughes and Duncan (1972) pointed out that the incidence
of feather pecking occurs in varying degrees between different breeds

of chickens.

Sexual Dimorphism g: Feathering and Growth

The relation of sex to rate of feathering has been noted by
many investigators. As early as 1922, Serebrovsky reported the
action of a recessive sex-linked gene which causes early feathering
in chickens. From a cross of Russian Orloff males with Plymouth
Rock females, all cocks produced showed a slow feather development
whereas hens showed rapid feather development. Later, in 1925,
Warren confirmed this fact by crossing White Leghorn males to Jersey
Black Giant females. He noticed that all males of the first genera-

tion were slow feathering and all females were rapid feathering.

15

However, when the reciprocal cross was made, both male and female
offspring were slow feathering birds. He concluded that rapid
feathering is controlled by a recessive sex-linked gene and its
dominant allelomorph causes slower feather development. Martin
(1929) reported that Barred Plymouth Rock chicks feather dimorph-
ically with respect to rate, the males feathering more slowly.
Jaap and Morris (1937), in a rather extensive study with different
breeds and varieties of chickens, concluded that at eight weeks of
age 20% of feathering variation is due to sex. Radi and Warren
(1938) studied the physiological nature and mode of inheritance of
feathering in a strain of Rhode Island Reds known to be homozygous
fbr late feathering. Through selection, two strains were established
and described as early and late feathering R. I. R. It was observed
that tail feathers developed earlier in females than in males. This
difference was noticeable until broiler age but was entirely elimin-
ated at sexual maturity. Similar observations were reported by Hays
and Sanborn (1942). They noticed that in R. I. R. chicks bred for
rapid feathering, tail feathers of females started to appear three
days earlier than those of males.

The common sex-linked recessive gene sl(k) was the main factor
for early feathering in R. I. R. females according to Hays (1951).
However, in addition to the action of the sex-linked gene 51, he
associated an autosomal gene X as a supplementary gene for better
feathering. Since female birds carry only one dose of a sex-linked

gene, it was presumed that the supplementary gene, X, was responsible

16
for the better feathering seen in females as Opposed to that in
males. To further elucidate the situation, Hays (1952) conducted a
sex-dimorphism study with R. I. R. and Leghorn-Red hybrids with
recessive phenotype for the gene 51. From various crosses, he found
that Leghorn hybrid males had a tail length of 1.87 :_.04 cm and the
females averaged 2.08 i .03 cm at ten days of age. In the Rhode
Island Red flock, the males had a mean tail length of 1.60 t .05 cm
while females averaged 1.92 t 0.21 cm at the same age, ten days.
He concluded that sex-dimorphism in tail length occurs at early ages
in favor of the female birds. However, if the recessive gene 51 has
a definite significant effect on tail length, males will develop
longer tails in later ages, since they carry two doses of the reces-
sive gene 51. The greater tail length in the females was described
as being due to the endocrine system of the females. In contrast,
Hurry and Nordskog (1953) demonstrated conclusively that the con-
dition of longer tails in females at 13 days of age was entirely due
to the action of sex-linked recessive gene, k.

Siegel et al. (1957a) studied the sex differences for six
feathering characteristics in early feathering White Plymouth Rocks.
It was found that the female progeny were significantly superior to
the males on the basis of back scores at ten days of age. Also,
breast feathering conditions were superior in females to those in
the males of three line-generations. With regard to the back area
covered with feathers at five and seven weeks of age, the males

were significantly poorer than the females.

17

Work done by Washburn and Siegel (1963) on White Rock chickens
indicated that sexual dimorphism in the feather development of the
back pterylae was evidenced at four and eight weeks of age, with
females being better.

In general, on basis of feathering rate, it seems quite likely
that females feather more rapidly than males, mainly due to the sex-
linked gene, k, and partially due to some variation in the endocrine
system functioning.

For growth and body weights, sex-dimorphism has long been known
among animals and poultry as well. In an earlier study (Knox and
Marsden, 1944), it was noted that 65% of the variation in body
weights of four varieties of turkeys was due to sex, with males
being larger. Asmundson (1948), likewise,reported a 60% variation
in body weights between sexes in Bronze turkeys, males being heavier
at 24 weeks of age. Shaklee et al. (1952), working with Beltsville
Small White turkeys, found that at 24 weeks of age the males out-
weighed the females by 40%. In chickens, Siegel (1962a), from
selection experiments for body weight at eight weeks of age has
demonstrated that female chickens are 12.5% lighter than the males.

Washburn and Siegel (1963) have shown that, in Plymouth Rocks,
males start to significantly outweigh the females at 20 days of age.
Kinney and Shoffner (1965), who worked with meat type chickens,
found that genetic correlation between male and female weights was
0.66 at eight weeks of age. In a similar study with White Leghorns

by Buvanendran (1969), the genetic correlation between male and

18
female weights was reported to be 0.59. This was considered to be a
genotype-sex interaction phenomenon. The magnitude of such varia-
tions could be ascribed to differences in the number of sex-
chromosomes between males and females. In this regard, Brunson et
a1. (1956) and Thomas et a1. (1958) indicated that sex-linked genes
might be the causitive factors in the sexual dimorphism of the in-
heritance of body weight. Hutt (1949) gave a better elaboration
and stated that, in poultry, females carry one dose of sex-linked
genes while males carry a double dose of the genes that might make

them heavier than females.

Heritability

Heritability is the key parameter for any given population to
improve the desired traits by selection. In poultry as in other
farm animals, some traits, qualitative, such as comb types or
feather pattern are inherited in a simple manner. However, most of
the economic traits, quantitative, such as growth rate and egg pro-
duction are influenced by many genes and inherited in a more complex
manner. The latter traits are markedly influenced by environmental
conditions which contribute variation among individuals fer a trait.
Subsequently, the total variation for any quantitative trait is due
to heredity, 02H, environmental, 02E, and the interaction of both.
ozEH. The hereditary variance is indeed due to additive effect of
the genes, dominance deviation and interaction deviations, epistasis.
Thus, the total variance components which make up the phenotypic

variance of a trait can be formulated as:

l9
02p = 02G + 020 + 021 + 02E + ozEH
In this regard, Abplanalp and Kosin (1952) defined heritability in
a broad sense as "That fraction of observed phenotypic variance
which can be ascribed to known genetic differences between indi-
viduals." In a more precise inheritance manner, the definition of
heritability in a narrow sense is limited to include only the
average gene effects, such as would be expected to appear if the
genes were acting additively. The latter definition expresses the
fraction of the total variance which is attributable to variance
due to the average effects of additive gene actions.

In breeding programs, the heritability values in the "narrow
sense" are the main concern. This is due to the fact that when
selection is practiced, this portion of genetic effect, additive,
is likely to be recovered in successive generations.

The degree of heritability, 0.0-l.O, is the reliable indicator
in determining the breeding procedure in the field. When the
heritability value of a given trait is close to 1.0, mass selection
would be the effective way in improving the trait. 0n the other
hand, when the additive variance makes a relatively low proportion

of the total variance, pedigree, family selection or progeny test

would be considered.

Methods of Estimatinngeritability
Various methods for estimating heritability have been developed

by many investigators, depending on the experimental design and the

\

20
trait of interest. The most common methods in estimating herita-
bility, and that are frequently used in poultry breeding studies,
are presented in considerable detail by Lush (1948), Lerner (1958),
Dickerson (1969) and Falconer (1960a). These methods include the
fbllowing techniques:

1. Parental half-sib correlation: 4325/32p,

2. Maternal half-sib correlation: 432d/32p,
3. Full-sib correlation: 2(32s + 32d)/32p,
4. Parent-offspring regression: 2320P/32p, and
5. Realized heritability: R/S
where

825 and 32d are estimated sire and dam components of variance,
respectively,

32p is the phenotypic variance,

GOP is the parent-offspring covariance,

R is the estimate response to selection, and

S is the selection differential.

The prols and con's of each method of estimation are discussed
by Lerner (1950) and Falconer (1960a).

In general, estimates based on parent-offspring correlation are
considered biased upward if epistatic variance is important. How-
ever, parental half—sib correlation is expected to be less biased
than maternal half-sib because of the confounding of maternal
effects or dominant variance with the dam component of variance in
the hierarchical, nested, analysis procedure. Also, it can be stated

that the method of full-sib correlation for estimating heritability

21
is intermediate in the amount of bias between the parental and
maternal half-sib correlation techniques.

The use of parent-offspring regression technique is the least
biased with a satisfactory estimation of the effective heritability
of traits of poultry. .

A reliable measure of additive gene action of various traits
observed in poultry and livestock animals, as well, is from selec-
tion experiments. As described by Falconer (1960a), the heritability
values obtained from bi-directional selection experiments, called
realized heritability, are more reliable estimates than from nested
variance analysis as they are based on the actual response over
- period of generations. The realized heritability can be obtained
by dividing the difference between the means of two groups selected
in opposite directions by the cumulative selection differentials of
the two lines or by regressing the cumulative response on cumulative
selection differential. The use of a random mating control group
in the two-way selection experiments would enable the refinement of
being able to estimate the heritability in both upward and downward
directions.

The term "selection differential" is defined by Lush (1945) as
"the difference between the average of those selected to be parents
and the average of the population in which they were born." Magee
(1965) pointed out that the term selection differential should be
reserved for the case where there is selection directly for the

trait being considered.

22

The rarely accurate estimate of heritability of any quantitative
trait is due to the relatively small size of the population under
study. Moreover, statistical sampling and differences between popu-
lations or conditions under which the traits are studied contribute
some variation to the estimates under the different conditions.

Lush (1948), Lerner (1958) and Falconer (1960a) discussed
experimental procedures to increase the accuracy of estimating
heritability.

Several investigators evaluated the sampling variance of esti-
mates of heritability. In the contest of the statistical model of
Winsor and Clark (1940) in estimating heritability, in narrow sense,
Osborne and Paterson (1952) derived appropriate formulas to estimate
the standard errors of the heritability coefficients from variance
analysis when subclassnumbers are equal. Graybill et a1. (1956)
have given formulas for estimating standard errors of heritability
and confidence limits.

An excellent example, using poultry breeding data, for esti-
mating confidence limits for heritability estimates from analysis
of variance technique was presented by Graybill and Robertson (1957).

Dickerson (1969) reported standard errors formulas of herita-
bility estimate as those of intra-year variance components (S, D and
F) based on the sampling errors of the components used in the enum-
erator of the estimates.

Nestor et a1. (1967) provided an appropriate fbrmula to calcu-

2
late the standard error (oh ) of heritability based on variance-

23
covariance among full-sib families analysis. Becker (1975) out-
lined the concepts of standard errors of heritability of different
experimental procedures.

Klein et a1. (1973) reported a tabulated quick reference for
standard errors of heritability as estimated by four different
methodologies (regression of offspring on mid-parent, on single
parent, intraclass Correlation of full-sibs and intraclass correla-
tion of half-sibs).

Confidence limits of the estimated heritability of a given
trait are important intervals within which the true values might be,
in effect, considered tenable. A number of suggested formulae for
estimating the confidence intervals for heritability are reported
by Bogyo and Becker (1963), Broemeling (1969) and Becker (1975).

Correlated heritability is a term referring to the heritability
of an unselected trait, which shows an associated change, when
selection is made for a given characteristic. Carte and Seigel
(1968) defined correlated heritability as "A ratio of genetic to
phenotypic covariances of selected and correlated traits."

Despite the fact that the response of an unselected trait can
be measured directly, the deciding factor in finding out its realized
heritability is selection differential. In the report of Magee (1965),
the term "secondary selection differential" for unselected traits is
to be used rather than the usual term selection differential which
should be conserved for directly selected traits. Carte and Siegel

(1968) reported that when correlation between selected and unselected

24
traits exists, the expected secondary selection differential of the
unselected trait is equal to the selection differential of the
selected trait multiplied by the phenotypic regression coefficient

of the correlated trait on the selected one.

Genetic and Phenotypic Correlations

In order to express quantitatively the extent to which two
variables are related, it is necessary to calculate the correlation
coefficient.

Gill (1978) stated that “correlation is a measure of the degree
of association or interdependence of two variables." .

The unit of correlation coefficient is a dimensionless quantity
that may take any value between -1 and +1, inclusive. Both of these
extremes represent perfect relationship between the variables, and
0.00 correlation coefficient means the absence of a relationship.

A positive value indicates that individuals obtaining a high
score on one variable tend to obtain a high score on the second
variable, and those obtaining a low score on the first tend to obtain
a low score on the second. On the other hand, a negative relation-
ship means that as one variable increases, the other decreases.
Genetic correlation refers to the correlation of the breeding values
of two characters.

According to Lasley (1972), pleiotropy (where one gene may affect
one or two traits) is probably the major cause of genetic correlation,
although it is possible for linkage to have a similar transitory

effect. Linkage means two or more non-allelic genes are residents

25
of the same chromosome. Some of the linked genes are located closely
enough not to be separated by crossing over during synapsis in
meiosis, over few or several generations. This linkage would entail
the breeding values of two traits in the same individual to be cor-
related. I

The genetic correlation is used to estimate the response in a
trait caused by selection on another trait. In this field of in-
vestigation and since the pioneer study of Hazel (1943) for estimat-
ing the genetic correlations, numerous studies have been carried out
to elucidate the practical treatment of genetic correlations and
methods of estimates.

Among the reports are those of Hazel et al. (1943) where the
covariance between the additive effects of two traits is an essential
element for estimating the genetic correlations.

Lerner (1950) used variance components and covariances in esti-
mating the genetic correlation. Falconer (1960a) and Becker (1975)
derived the genetic correlation between traits from variance co—
variance components of half-sib families in manners comparable to
the methods of heritability estimates.

The implications of genetic correlation in bidirectional
experiments have also been reported. Siegel (1962b) provided
formula(e) to compute the genetic correlation between selected and
unselected traits. With nested experimental designs with unequal
number of observations, Grossman and Gall (1968) presented the

appropriate formula(e) along with numerical examples for the purpose

26
of illustration. Also, Hammond and Nicholas (1972) showed the
computational formula(e) of genetic, phenotypic and environmental
correlations from nested unbalanced data.

In practice, it has been observed that traits of fitness,
fertility, hatchability, and livability show an adverse correlated
response when selection for other economic traits is exercised in
either direction. Lerner (1954) called this sort of reduction in
fitness characters "genetic homeostasis." Further elaboration on
the consequences of this principle has been discussed by Falconer
(1960a).

Since the genetic correlation is a function of the covariance
and the respective variances of two traits, it is subject to some
bias as in the variance component estimations.

Methods of computing the standard errors of genetic correlation
were reported by Robertson (1959b), Tallis (1959) and Becker (1975).
The phenotypic correlation between two traits is a function of
genetic and environmental correlations. Methods of estimates are,
in fact, based on the variance and covariance components as reported

by Hazel et al. (1943), Hazel (1943) and Lerner (1950).

Heritability gfi_Featheringand Growth
The estimation of heritability of body weight in chickens and
turkeys has been extensively studied by numbers of investigators.
However, few have reported on heritability estimates of feathering.
Jaap and Morris (1937) analyzed data of body weights and feather-

ing scores secured from six standard varieties of chickens. They

27
estimated the total variance for feathering due to differences in
varieties, sires, dams and sex together to be 51%. Their data were
further analyzed by Hurry and Nordskog (1953), employing a full-sib
correlation technique, for heritability estimates within sex and
varieties. Feathering rate at eight weeks of age had a heritability
estimate of 71%. Hurry and Nordskog (1953) reported a genetic
analysis of chick feathering where subjective scores were used to
determine the degree of feathering at 13 days and eight weeks of age,
based on tail feather length and back feathering condition. From
the results of analysis of variance and full-sib correlation tech-
nique, the heritabilities of feathering in eight-week-old New
Hampshires and Barred Plymouth Rocks were estimated to be 0.33 and
0.42, respectively.

In a study to estimate heritabilities for degree of feathering
and phenotypic correlation for six feathering characteristics,
Siegel et al. (1957a) conducted a selection experiment with two
divergent lines, superior and inferior, of feathering in White
Plymouth Rocks. The six feathering characteristics were described
as the degree of juvenile feather development in the breast pterylae
at ten days, the amount of feather development on the back at ten
days, five weeks, and seven weeks, the number of pinfeather in the
sheath stage found in the back area at ten weeks and the amount of
body down present at the same age. Based on intra-sire regression
of offspring on their dams, heritability values were 40, 32, 49, 39.

47 and 32% for ten-day back score, ten-day breast score, five-week

28
percentage of back area covered, seven-week percentage of back area
covered, ten-week pinfeather score and ten-week down score, respec-
tively. Sexual dimorphism in feathering rate was observed in favor
of the females.

Gyles et a1. (1959) used White Rock broiler strain chickens in
a two divergent selected lines experiment to determine the herita-
bility values of breast blister and breast feathering conditions.
The calculated heritabilities from combined sire and dam components
for the breast blister and breast feather condition were 0.44 and
0.01, respectively.

From the literature reviewed, it appears rather cumbersome to
list all reports of heritability estimates of body weight of chickens
and other poultry species.

Siegel (1962a), who reviewed the subject, compiled the results
of 176 published heritability estimates of body weight of 6-12
weeks old chickens. From different methods of estimation, herita-
bility values range from 0.1 to 0.8 with the majority being in the
range of 0.3 to 0.6.

III. MATERIALS AND METHODS

In raising game birds, fast feathering stock would enable birds
to be released at younger ages for flighting purposes. This would
result in savings in feed cost, housing, labor, etc.

Recently, the idea of developing early maturing ring-necked
pheasants has been initiated at Michigan State University (M. S. U.).
In indoor poultry research facilities provided by the Poultry Science
Department at M. S. U., a three-year experiment began September 15,
1978 and concluded August 21, 1981. It was conducted to study the
genetic parameters of feathering and body weights of ring-necked
pheasant.

From a randomly produced population of 6,000 game farm pheasants,
11 weeks old, housed in Mason, Michigan, a total of 60 males and
160 females were selected in a divergent selection program for
feathering rate. Mass selection was utilized in the selection of two
groups of male and female pheasants solely on the criteria of feather
appearance. One group (Line) was designated as "rapid feathering"
and the other as "slow feathering." Birds of the rapid feathering
group were characterized by having no bare back with longer wing
and tail feathers than those of the slow feathering group.

In addition to the bi-directional selected birds, a total of 16

males and 30 females were randomly selected and classified as

29

30
"control" group. Birds of this group, control, were randomly bred
throughout the course of the experiment for the purpose of compari-
son.

Selected birds were all transported to the Poultry Science
Research and Teaching Center, P. S. R. T. C., at Michigan State
University. Birds of each of the three lines were housed in separate
pens of 3.05 x 4.88 m, with a layer of wood shavings on the floor of
each pen. A pheasant flight, maintenance, ration (Appendix Table 3)
was provided ad libitum in two hanging feeders in each pen. Each
pen contained one mechanical cup-like waterer to provide water to
the birds ad libitum. To control cannibalism, birds were kept in a
dim light environment supplied by one 25-watt incandescent red light
bulb in each pen. The proper amount of ventilation was provided in
the house to keep birds comfortable.

The experiment commenced when birds of each group (Line) were
wing banded and individually weighed, to the nearest gram, at 12
weeks of age (September 22, 1978). At 20 weeks of age, all the
pheasants were supposed to be transferred to a cage room where they
were to be housed individually in cages. However, due to the lack
of space in the cage room, all birds were allowed to remain on the
floor with a complete dark environment imposed to delay sexual
maturity. Unfortunately, an outbreak of bacterial arthritis disease
(according to Michigan State University diagnostic report) occurred
with various severity among the three groups of birds. This, indeed,

accounted for a substantial part of the total mortality that occurred

31
in the base flock. All birds were placed on antibiotic treatment
for several days, which alleviated the problem.

On February 10, 1979 the 32-week-old birds were individually
caged in single bird cages. The cages used were 17.8 x 35.6 x 30.5
cm in dimension and arranged in the house in a stair-step style with
1.22 m distance above the ground. Each cage was designed to contain
one "nipple" type waterer and a plastic-type feeder trough was
located in the front side of the cages.

Light was set to supply 14 hours of light per day (14L:100) from
6 to 8 a.m., as provided by six 75-watt incandescent light bulbs, to
stimulate egg production. Manure was removed from the cage room
once every two weeks.

At the beginning of the pre-production period, birds were
switched from a maintenance ration (Appendix Table 3) to a pheasant
breeder ration (Appendix Table 4) in pelleted form.

Four weeks after birds were lit, March 10, 1978, most of the
hens produced eggs and the cocks, by milking, showed a good semen
production. Beginning with the first egg produced, individual egg
production was recorded daily in each treatment. Only hens which
produced eggs at a rate of 50% or more within a six-week period
after lighting, within each line, were saved to be the parents of
the first generation progeny. Likewise, only those males that
showed good semen production were assigned to be the sires. Con-
currently, all poor performance birds were culled. From this prac-
tice, a phenotypic selection for eggs and semen production was

applied on birds which were already selected for feathering.

32

On March 24, 1979 the potential birds in the slow and rapid
feathering groups were each subdivided into two replicates.

In each treatment, a total of four randomly chosen potential
males and 16 potential females from each replicate were used as
parents of the next generation. Consequently, four females were
assigned to mate with one male to constitute a particular family.
The control group was designed to contain one replicate with four
families of the same size as in each of the other replicates of other
treatments.

On March 31, 1979 the first mating was carried out by artifici-
ally inseminating each hen with semen from a specific sire. This was
conducted by following a method similar to that described by Burrows
and Quinn (1935) for chickens. The volume of semen inseminated per
hen, herein, varied but was almost always greater than the minimum
amount generally recommended (.05 cc) for chickens. Three days
after first insemination, a second artificial insemination was
employed, to allow for maximum fertility. All subsequent matings,
however, were conducted once per week and four hatches of pedigreed
chicks were obtained.

The pedigreed eggs from each individual hen were collected
daily and refrigerated at 15.6°C from time of laying to the be-
ginning of incubation. At no time were eggs kept more than seven
days, prior to incubation, so that weekly settings of pedigreed
eggs were made. Approximately 18 hours prior to incubation, eggs

were brought to room temperature and sorted out by dams. A Jamesway

33
252 single stage incubator was used for incubation and hatching as
well. Eggs were incubated for three weeks at 99.5°F (37.30C) and
60% relative humidity. After the 21 days, eggs were placed in pedi-
gree baskets and transferred to a hatching unit operated at 98.5°F
(36.l°C) and 70% relative humidity. On the 25th day Of incubation,
the hatch was taken off and all birds were wing banded, individually
weighed and pedigreed by sires and dams. The body weights were
recorded to within one-tenth of gram.

On the hatching day all unhatched eggs were broken out and the
percents of fertility and hatchability of fertile eggs were deter-
mined on sire-family bases.

All day-old chicks, of each hatch, were transported to the
breeding facilities at P. S. R. T. C. A total of 1,006 day-old
chicks were obtained in four hatches, and called generation one.

The hen breeders were allowed to lay eggs for a total of 60 days to
compare egg production among the three groups (lines). Birds of each
hatch were intermingled for brooding and housed in a 3.05 x 4.88 m
pen with a layer of wood shavings spread evenly over the floor.

A confined area was formed, by corrugated chicks guard, in the
middle of each pen to keep birds close to the heat source for the
first five days of age. Pheasant starter ration (Appendix Table l),
in crumble form in two flat trough feeders, was provided ad libitum.
Water was provided in two 3.75 liter plastic jugs per pen. Also,
each pen contained one cup-like mechanical waterer. When birds were

brought to the brooding pen, the beaks of about five dozen were

34

dipped into the water to aid them in locating the waterers. Also,
two pieces of small shiny tin, wing bands, were placed in each water
jug to attract the birds to the water. Water jugs were cleaned and
filled with fresh water daily. Heat was supplied by three infra-
red heat bulbs hung approximately 70 cm above the floor in the middle
of the confined area. Also, a gas brooder canopy was placed in every
other pen as a supplemental source of heat during the cold days in the
winter of each year. When the chick guard was removed, one of the
two water jugs was moved in the pen to close proximity to the mechan-
ical waterer. When the birds were five weeks of age, they were well
able to use the mechanical waterer and, therefore, the water jugs
were removed from the pen.

At six weeks of age, birds were provided with a pheasant grower
ration (Appendix Table 2).

When pheasants were 12 weeks of age, the population of each
hatch was divided into two groups, which were reared in adjacent
pens to lessen bird density. In the meantime, the ration was switched
from pheasant grower (Appendix Table 2) to pheasant flight, mainten-
ance ration (Appendix Table 3). The brooding mortality was recorded
on sire-family bases, as a percent of total birds died to the total
chicks hatched, over a 12-week period.

Before any feathering scores could be recorded, a method of de-
termining feathering differences had to be devised.

Feathering studies were based upon measuring lengths of the

longest wing primary and secondary feathers (numbering proximally

35
from the axial feather) in centimeters from the base of the feather
to the furtherest point of emergence. The actual length of tail
feathers was also included and measured from the extreme tip of the
pygostyle to the end of the longest feathers. Observations upon
these feathers and body weight were made on birds, individually, at
two, four and eight weeks of age. At 12 weeks of age, only body
weights, to nearest gm, and back feathering condition were studied.
Subjective scores were arbitrarily used to indicate the back feather-
ing criteria in which three scores were assigned with zero for almost
or completely bare backed class and three for full complement feather
classes being the extremes. It should be noted that no measurements

were recorded in the foundation stock.

Selection and Subsequent Matings

In most feathering studies, subjective scores or grades are
used as criteria to evaluate the birds. In this regard, Hurry and
Nordskog (1953) reported that about 20% of the variance in the exper-
imental error was found to be due to errors in grading committed by
the grader. For this reason, it was felt that a tangible criterion
is necessary to deal with the hereditary aspect of feathering rather
than subjective scores as used in the present study.

The length of primary, secondary and tail feathers, at various
ages, and body weights were utilized in establishing a feathering
parameter device, Selection Index. Also it was felt that the
preferences of pheasant hunters should be given due consideration

in the selection index. To accomplish this, a pilot questionnaire

36
survey was conducted. Twenty pheasant hunters were asked to rank
in relative importance feathering factors they most appreciated in
the birds they shoot. Length of wing feathers, tail feathers and
density of feathers on the bird's back were the factors listed on
the questionnaire.

The relative importance, on total of 100% basis, pheasant
hunters assigned to each of these factors was 7.50, 16.25 and 76.25
fbr length of wing feathers, length of tail feathers and back feather
density, respectively. To develop the selection index, the variance
with classification (Snedecor, 1956) of certain traits, longest
primary wing feathers, longest secondary wing feathers and the
longest tail feathers, was computed in birds of each hatch of the
first generation. The following formula was used in computing the

variances of the traits:

5% 2 2
I I Pij - {(PR.)2/nR - [(Pc.) /nC - (PS.) /nS

 

Variance = ‘iFR 1?] N - 3
where
Pij = Phenotypic record of jth individual of the ith group,
R = Rapid feathering birds,
C = Non-selected control birds,
S = Slow feathering birds,
nk = Number of observations in a given line, and
N = Grand total of observations in the three lines.

'37

The standard deviation (square root of the variance) of the rela-
tive importance pheasant hunters assigned to each trait was divided
by the standard deviation of the trait, in order to use a standard
measure. The averages of these figures for the four hatches of
generation one were used in developing the selection index as follows:

Selection Index = 1.21 x length of longest primary feathers

in cm + 1.35 x length of longest secondary feathers in cm +

3.34 x length of longest tail feathers in cm + 6.14 x back

score (1, 2 or 3).

The above selection index was used in this study to evaluate the
feathering rate of each bird with observations until 12 weeks of age,
in each treatment and for all generations.

In selecting the breeders of subsequent generations, some efforts
were made to maintain the egg production at a satisfactory level, by
rejecting any hen which showed very poor performance. In some in-
stances, this was at the expense of the feathering score levels of
selected birds which led to some sacrifices in the level of selection
pressure. Some attention was also given to the use of vigorous
males in the matings. When the selected birds of the first genera-
tion were 24 weeks of age, they were moved to a cage room and housed
individually in the same cages used for the base population.

At caging time, all birds started to receive 14 hours light
(6:00 a.m. to 8:00 p.m.) per day as provided by six 60-watt incandes-
cent light bulbs, for a two weeks period. As in the base flock,
light was then increased to 16 hours per day to better stimulate

semen production.

38

At caging time, birds were switched from pheasant flight
(Appendix Table 3) to pheasant breeder (Appendix Table 4) feed.

Four weeks after lighting, cocks were trained for two successive
weeks to ejaculate the semen prior to being used as sires.

Based on the semen and egg production in male and female birds,
respectively, during the first six weeks post lighting, only those
birds exhibiting a reasonable potential for reproducing were chosen
to be breeders of generation two. Selection was performed within
each replicate of the divergent lines.

Prior to beginning any artificial insemination, matings were
made on paper, based on pedigree information. To prevent inbreed-
ing, matings were performed with only unrelated birds within each
replicate of each line. The standardized selection differentials
in the rapid, slow and control feathering groups were 13.17, -10.72
and 0.62, respectively.

Four sires and twelve dams, constituting four families, were
the parents of generation two in each replicate of the slow feather-
ing group. In the rapid feathering group, three males were assigned
to four females in each replicate. The control group contained four
males to mate with 12 females on a pooled semen basis.

Mating procedure was carried out as for the previous genera-
tion. The assigned hen breeders of the three lines were inseminated
twice in the week where egg production reached approximately 30-40%,
and once a week thereafter. Despite all efforts made to select

potential breeders from egg and semen production standpoint,

39
performances of selected parents were relatively lower than their
parental stock. Therefore, in order to maximize the number of chicks
for the next generation, six weekly hatches were carried out. The
first set of pedigreed eggs went into the incubator on December 9,
1979. A total of 600 day-old chicks were obtained and were desig-
nated generation two.

The hatched chicks were wing banded, weighed to nearest tenth
of a gram and pedigreed on sire-dam family basis. Birds were trans-
ported to the rearing facilities at P. S. R. T. C. where they were
housed in their respective pens. Temperature, humidity and lighting
regime were managed so as to provide environmental conditions that
were as similar as possible among hatches and generations. Feeding
schedule and diet were the same for all hatches as with chicks of
the preceding year.

Hen breeders of generation two were allowed to lay eggs for 60
days so that the response of feathering selection on egg production
could be evaluated. In regard to reproductive traits, all unhatched
eggs were broken out and thoroughly examined for fertility on sire-
dam family basis.

Feathering measurements and body weights were recorded, through-
out the 12 weeks measurement period, using the same methods and
tools as for birds of the first generation.

On July 23, 1980, selected birds were moved to the cage room
where they were individually caged, lit, fed and managed similar to
their parent stock. From only the more productive males and females.

parents of generation three were selected.

40

The third year mating scheme was carried out on a smaller scale
than those of previous years, three females per male instead of the
earlier four females per male in each replicate.

As far as possible, dams were inseminated with semen of unre-
lated sires, ascertained from the short pedigrees. HOwever, due to
the limited number of potential birds in each replicate, mating some
related birds by descent was inevitable in generation three.

Although the amount of inbreeding was believed to be trivial,
it was felt necessary to calculate the coefficient of inbreeding in
the third generation. This was computed from the number of sires

and dams used in the mating according to Wright (1940):

where
"M = Number of males (sires), and
NF = Number of females (dams).

The mating procedure was conducted as in the preceding two
generations. Due to the fairly consistent weekly decrease in egg
production, four lO-day egg settings were employed to enhance number
of progeny. Eggs were pedigreed, collected daily, refrigerated and
incubated as in the preceding generations. The progeny of genera-
tion three were generated from six males and 18 females in the slow
feathering line, four males and 12 females in the control and six
males and 15 females in the rapid feathering line.

Four hatches of pedigreed eggs were set on lO-day interval

basis, with the first incubated on March 30, 1981.

41

A total of 544 day-old chicks were obtained and designed as
generation three. At hatching time, all chicks were pedigreed by
sire and dam and individually weighed to nearest tenth of a gram.
Unhatched eggs were examined for fertility as earlier.

Chicks of the four lO-day hatches were brooded and reared in
the same facilities, with conditions kept as uniform as possible,
as for first and second generation chicks.

Feeding program, management and feathering and body weight
measurements were practiced in the same manner as those employed with
previous generations, as the general plan called for. When the last
feathering and body weights were taken at 12 weeks of age, August 21,

1981, the experiment was terminated.

IV. STATISTICAL PROCEDURE

In dealing with the data collected over the three generation
period, only those birds which survived up to 12 weeks of age and
had records of measurements were subject to the statistical analysis.

To estimate the heritability of the selected trait, feathering,
three methods were utilized.

In the first method of estimate, the realized heritability was
determined within lines. This was obtained as outlined by Falconer
(1960a) by regressing the cumulative differences in response between
the selected lines and the control line on the additive selection
differentials in each of the selected lines. The following formula

was used to compute the realized heritability of feathering:

h2 = R/S
where
h2 = Realized heritability,
R = Response to selection, and
S = Selection differential.

In the second method of analyzing the data, heritability esti-
mates of feathering, tail length at four weeks of age, one-day, four-
week, eight-week, and lZ—week body weights were determined by
generation combinations within a given line based on full-sibs
correlation by means of variance component analysis (Lerner, 1950;
and King and Henderson, 1954). The statistical model for this

42

43
hierarchical nested analysis of variance design was as follows:

Yijkl = M + Gi + $13 + Dijk + eijkl ,
where

Yijkl is the random variables of the 1th progeny of the kth
dam by the jth sire in the 1th generation.

M is the overall true means of the offspring population,

a. is the effect of 1th generation; 2 = l, 2 and 3.

Sij is the effect of jth sire; j = l . . . Sj,

D1jk is the effect of kth dam; k = 1 . . . dij. and

eijkl is-the experimental error; 1 = l . . . "ijk
In this model, it was assumed that all effects have means equal to
zero and are mutually uncorrelated,

E(S§j) = as, E(D§jk)= on .

Eléijki) = 0E
and no interaction existed among generations, sires and dams.

Estimates of sires, S, Dam's, D, and full-sibs progenys, W,
variance components were made using the GENSTAT program (Appendix
Table 13) on the CYBER 750 computer at the M. S. U. computer center.
The standard model of the one-way nested hierarchical analysis for
the determination of variance components is given in Appendix Table
5.

The estimates of variance components in the two replicates of
each of the selected lines were averaged and then combined (Appendix
Table 6).

Heritability estimes, which are considered as a fraction of the

total variance were expressed by the following notations:

44

 

 

 

2
2 48$
1. h3 = A A7 A2
05 + 90 + 0W
2
2 460
2. ho = 2 2 A2
85 + 60 + ON
2 2
3 h2 = 2(65 + 60)
° S+D 2 A2
6% + 60 '1' G”
where
6§ = between sires components of variance,
86 = between dams components of variance,
Bfi = between full-sibs variance.

The standard errors of these estimates from the variance compo-
nent analysis were computed by approximation method presented by
Dickerson (1969) as a square root of variance of estimated herita-
bilities. The general formulae are:

16 Var 052/(052 + 03 + at?) 2

Var (h§)

Var (hfi) 16 Var era/(052 + 06 + 0,3) 2
Var (h§+D) = 4[Var of + Var 06 + 2 Cov (a; 66) l/

(052921-06 +09%) 2
The variances of variance components were estimated by formulae

proposed by Anderson and Bancroft (1952) as follows:

 

 

 

62? =
varls) K1; df(s)+2 d°f(D)+2
2 2 "56 + ”SR

 

 

 

V3? (60) "" K2 d'f"(D) + 2 d'f°(w) '1' 2

Cov (6% 66) = [-2M56 /d-f-(dams) + 2]/K K
1 3

45

The third method of heritability estimates used herein was de-
rived from intra-sire regression of offspring on dams within sires,
with sex effect adjusted among individual means. From this technique,
only the heritabilities of feathering and body weights at 8 and 12
weeks of age were estimated.

The statistical model used was in the manner set by Becker
(1975) as follows:

Yij = M + B (xij -'§'. .) + 9i + eij
where

Yij = phenotypic records of offspring of the ith sire mated to
the jth dam.

M a common mean of offspring population.

6 = regression coefficient of offspring on dams.

xij = observed value of jth dam mated to the ith sire.

i’. . = phenotypic mean of the dam population.

h

oi = the effect of the it sire.

eij = the residual error.

The general model and the computational formulas for variance co-
variance table were adapted from Becker (1975) and are illustrated in
Appendix Table 9.

The standard errors of regression estimates were determined by
means of statistical fbrmulae used in a model similar to the one in

a study by Kinney and Shoffner (1965) and outlined by Becker (1975).

SCP XY
2 22m- .25.:
S. E. (bop) a '

(D-S-l) 550(XX) '

 

 

46
s. E. (h2) = 2 s. E. (Bop)

Correlations Between Feathering and Correlated Traits

The analysis of variance schemes were extended in the computer
program to include covariances between the feathering and the corre-
lated traits: body weights at one day, four weeks, eight weeks and
12 weeks of age and tail length at four weeks of age. The components
of covariances were determined on basis of sires, S; dams, D; and
full-sibs, W. The formulae developed by Hazel et a1. (1943), used
by Peeler et a1. (1955) in a report with a model identical to this
study and outlined by Becker (1975) were used to estimate the pheno-
typic, rPxPy’ and genetic, erGy’ correlations from the variance

covariance analysis. These formulae are:

 

 

 

 

1 r Covs + Covn + GOV”
° PxPy=
\I as. + 66 +62: T62, +61:y +65,
Cov
2. erGyS = Vag &
x
EOVD

 

 

3. r O =
GXGY 'vret;‘ ab,

Covs + Covo

,[az,.sx+afi 76%, +66),

5. r = A C°V'
my (05: + 6’3 )/2

 

4. erGy(S+D)=

 

When one or both of the estimates of the variance components

were negative, the arithmetic mean of the variance components was

47
used in the denominator (5). Standard errors for the genetic corre-
lation were estimated in approximate method outlined by Falconer

(1960a) using the following formulas:

 

" 2
S E (r )9, l - (erGyv s. E. (hx) s. E- (n.2,)
’ ' GXGY v—T— hi hi

As for the reproductive traits, due to the very small number

 

 

of observations in this study, the genetic parameter estimates of
reproductive fitness were discounted. Nevertheless, in order to
shed some light on the response of reproductive fitness to selec-
tion for feathering, hen-housed egg production was statistically
analyzed by two-way analysis of variance as outlined by Gill (1978).

Mortality observations were not dealt with in this study.

V. RESULTS AND DISCUSSION

The data obtained in this study, over three generations, were
such that total variance among individuals was measured in the same
location and environment. This was carefully designed to yield as
unbiased estimates as possible. Nonetheless, considerable fluctua-
tions in both magnitude and sign in the estimates of measured traits
were obtained. The average number of offspring obtained in each
generation was 15.5 per sire and 4.3 per dam. These numbers were
in the range of the optimum number recommended by Robertson (1959a)
to estimated medium size heritability. However, such numbers were

considered inadequate for lower heritability estimates.

Effect of Mass Selection

Means and standard deviations for the selected trait, feather-
ing, are presented by line, sex and generations in Table 1. Dif-
ferences in feathering scores between the two divergent line males
were 3.5, 4.9 and 7.4 for first, second and third generation, re-
spectively, with rapid feathering individuals being better. Female
feathering scores in the rapid line exceeded those in the slow
feathering line by 3.1, 5.0 and 9.0 in the three successive genera-
tions, respectively.

The effectiveness of the two-way mass selection was so pronounced
that differences, in both sexes, between lines for feathering

48

49

 

 

—.~N— a.mm— m.¢mp a.wm— m.e- a.—m— wongm><

v.5 H m.v~p mm ~.~ H ¢.n~— mp m.o— H ~.mmp a— ~.—p H N.m¢— up m.—— H o.¢~p mm m.—— H ¢.¢mp mm w: owns:
n.¢ H ~.ap— mm m.w H m.¢~— co m.—— H m.cmp on o.~p H o.~m— am o.o— H c.¢~— cm ~.m— H ¢.m~— cc pa uwama
m.p— H m.opp mm m.—p H m.-p mm ~.o— H c.pn— cw o.p~ H —.wm— em e.o— H o.-— mu a.m— H ¢.mm_ cm —cgu:ou
—.m—— m.»—— a.mmp o.¢n— _ ~.—~— «.mup maagm><
m.—p H m.m_~ mm a.mp H p.m—p ms o.m H m.m~— no c.m H ~.mm— Ne a.mp H a.c- an m.—— H a.mw— mu Na zo—m
e.op H ¢.~pp mm m.m H m.wp— mv ¢.p— H m.-— m“ m.a H a.¢mp mm a.m— H m.p~— cc c.m H ~.—m— cc pm zo—m
mm—wEmn— : «mm—hi: : mm _,nEw..— : mwpm: : mo poem..— : m0 PM! a 9:...

 

m copuococmu

 

 

~ ro.uogo=mw

— ccfiuacmeoo

 

.xmm new mumowpqmc .cmasac xn meowumcmcmm omega mo uo_cma a com cowpep>mu
ccmccmum H mcoum acacmcpmwu mo mommcw>m An commwxaxm mmcwp mcpcmsummm twang new
30pm on» new pocpcoo on“ :Fsuwz acacmgpmmm cram; use zopm so; cowuumpwm mo mu—amwm

.P mpnmh

50
became progressively larger with each subsequent generation (Table
l).

The patterns of feathering and tail growth are shown in Figures
1 and 2. From the means of tail lengths (Table 2), female tail
lengths in the rapid feathering group out-measured those in the slow
feathering group by .3, .3 and .5 cm for the first, second and third
generations, respectively. These differences in the male progeny,
for the respective generations, were .1, L3 and .2 cm with rapid
feathering individuals barely having longer tail feathers than slow
feathering individuals. These figures of feathering and tail growth
rates in the two selected lines reflect the effectiveness of selec-
tion over generations.

On a within line-generation basis, feathering and tail growth
rates of the rapid feathering and control lines showed a decline in
performance over the third generation period, which might be due to
overall environmental effect. Inbreeding is another possible reason
for the performance decline in the entire stock. The coefficient of
inbreeding in the third generation, where mated parents had common
ancestors, was calculated from the actual number of male and female
parents using the approximation formula by Wright (1940). The in-
breeding coefficients were calculated to be 0.05 and 0.06 for the
slow and rapid feathering lines, respectively. This low level of
inbreeding among individuals is considered to be of little negative
impact on performance rates. It was also noticed that the feather-
ing rate and other performance criteria declined somewhat in the

randomly mated control group among their later generation individuals

51

 

 

 

 

 

s.e ..s ... e.. o.e m.e aaaee><
as. s 5.. NN am. e ... o. 2.. . ... a. 4.. H m.m .. e.~ « o.m mm e.~ 5 a.e mm Na e.ae¢
... . m.o an N.. : e.e o. m.. a e.e on ... . o.e am m.~ a ..e am e.~ e o.e es .a e.ae¢
m.. s ... em ... s ”.9 an ... a N.. as e.. a m.. an e.~ a m.. aw e.~ « ... eN .eeueou
..e m.e o.o n.. ..m ..e emeea><
m.. n 5.. am 4.. a m.e m. ... a o.. as e.. 5 e.. we m.~ H m.m an e.~ H m.e mu w. ze.m
o.. : ~.e mm a. H ..e we m.. e ..o m. ... 4 N.. am m.~ n «.5 ce ~.~ a c.. oe .g ze.m
mmpmsmu .. 32F .. $.95.— .. 3E .. $th .. 35: .. on...
m 53238 N .85328 — 5532.3

.xmm new mumu_.amc ..mnsac an mcwcmguemw owns.

new zo.m cow :o.uum.mm mo mcowpmcmcmm omega .m>o mmcw. mcwcmgummm twang new zapm one
use .ocpcou ecu com =o.uop>mu ugmucwum H Eu c. cameo. .wau xmmzuczom mo mmmacm>m och .N mpaah

52

 

 

 

 

 

 

 

 

I40
I35-
3 I}
a: 13'
c3
c:
a)
(D
Z
.... 12
a:
uJ
I I.
p. t
E t.
u. 120 LEEENP
-e- S. F. 1
J -B- C. F. 2
115 -)(— R. F. 3
-o- S.M.4
-x— R.M.5
-!- C. M. 6
u : : i i i 5 . 000
1.000 2-000 3'

GENERHTIONS

Figure l. The observed mean response to direct selection for
feathering in males and females of the slow and rapid
feathering lines compared to the non-selected control
line (random) over generations.

 

53

 

 

 

 

 

 

 

 

 

 

 

0.00
70500"
2:
c:
ES 7.000-
ch
2:
n:
.J
:3 0.500
a:
.—
s:
1.1.1
04
z 0.00 1.5051113
‘;, ' -e» S.F. I
" 45- C.F. 2
..)(_ R. F. 3
5.50o
+ S. M. 4
+ R. M. 5
-x- C.M. s
5.010.000 2.000 ' 3.000
GENERATIONS
Figure 2. The observed mean changes in tail length of males and

females in the slow and rapid feathering lines compared
to the non-selected control line (random) over genera-
tions.

 

54
as compared to earlier generations. Due to the small size of the
population of the control group, inbreeding could have built up
and cause some reduction in performance. However, the most probable
cause that retarded the improvement of feathering and associated
traits in the third generation would more likely be the effect of
overall environment rather than the inbreeding. In a study by
Glazener et a1. (1951), inbreeding at a rate of 40% had no depress-
ing effect on the improvement of early feathering of broiler
chickens.

As for the other traits, means and standard deviations of one-
day, four-week, eight-week and 12-week body weight are presented in
Tables 3, 4, 5 and 6.

In the slow feathering line, one-day body weight, on the average
of three generations data, was heavier than that of both the rapid
and control groups by .43 and .14 gm, respectively. The two divergent
lines and the control showed a consistent decline for one-day body
weight, in both sexes, over generations (Figure 3).

It is well known that sexual dimorphism in body weight occurs
soon after hatching, and this was quite evident for body weight at
four weeks of age in the three lines (Table 4). Male body weights
at this age, four weeks, in the slow, control and rapid groups were
heavier than those of females by an average of 28.35, 19.66 and
25.52 gm per generation for the three groups, respectively. Con-
sidering both males and females, the rapid feathering individuals

out-weighted the slow feathering ones by an average of 1.3, 6.7 and

5
5

 

 

 

 

 

 

5.5. 5.5. 5.55 5.55 5.55 5.55 555.555
5.. 5 5.5. 55 5.. 5 5.5. 5. 5.. 5 5.55 5. 5.. 5 ..55 5. ..5 5 5.5. 55 5.5 5 5.55 55 55 5.555
5.. 5 5.55 55 5.. 5 5.55 55 5.. 5 5.55 55 5.. 5 5.5. 55 5.. 5 5.55 55 5.. 5 5..5 55 .5 5.555
..5 5 5.5. 55 5.5 5 ..5. 55 5.. 5 5.5. 55 5.5 5 5.55 55 5.5 5 5..5 55 5.5 5 5.55 55 .555555

5.5. 5.5. 5.55 5.55 5..5 5..5 555.555
5.. 5 5.5. 55 5.. 5 5.55 5. 5.. 5 5.5. 55 5.. 5 5.5. 55 5.. 5 5..5 55 5.5 5 5..5 55 55 55.5
..5 5 5.5. 55 ..5 5 5.5. 55 5.. 5 5.55 55 5.5 5 5.55 55 5.. 5 5..5 55 5.5 5 5..5 55 .5 55.5

mopoeou : «ohms. : 5o—5eou : nova: 5 mopoeou : mopozl : o=_4
m :o5uococow N covuococow _ co.uogo:ow
.xom 5:5 opoowpaoc .co5555 >5 mcwcozpooo
55555 555 3555 com :oPHuopom mo mcowuococom oocgu Lo>o 5o:5_ 5:55o555o5 5555. 555 35.5.
oz» 555 .ocucoo o5» coo covuow>oo 55555555 H So :5 ucmwoz 5555 5.5 moouoco omoco>o ogh m oPnoh

6
5

 

 

 

 

 

 

5.55. 5.555 5.55. 5.5.5 5.55. ..555 5555555
5.55 5 5.55. 55 5..5 5 5.55. 5. 5.55 5 5..55 5. ..55 5 5.555 5. 5.55 5 5.55. 55 5.55 5 5.555 55 55 5.555
5.55 5 5.55. 55 5.55 5 5.555 55 5.55 5 ..55. 55 5.55 5 ...55 55 ..55 5 5.55. 55 5.55 5 5.555 55 .5 5.555
5.55 5 5.55. 55 5..5 5 5.55. 55 5.55 5 5.55. 55 5.55 5 5.555 55 5.55 5 5.55. 55 5.55 5 5.555 55 .555555

5.55. 5..55 5.55. 5.5.5 5.55. ..555 5555555
5.55 5 5.55. 55 ..55 5 5.555 55 5.55 5 5.55. 55 5.55 5 5.5.5 55 5.55 5 5..5. 55 5.55 5 5.5.5 55 55 55.5
..55 5 5.55. 55 5.55 5 ..55. 55 5.55 5 5.55. 55 5.55 5 5.5.5 55 5..5 5 5.55. 55 5.55 5 5.555 55 .5 35.5

55 .5555 .. 5o .5: .. 5o .95.. .. mobs; .5 5o .55. 5 5o .5: .. o5 ..
m coincocow ~ .....uococow . .....aocozow
.xom 5:5 ouoo..5oc .5o5555 An 55.5o555o. 5.55. ago 35.5 so.
co.uoo.o5 yo 55o.555o:om ooccu 5o>o 5o:.. 55.5o555o. v.55. 555 35.5 on“ 5:5 .ocucoo
osu co. co.uo.>oo 55555555 H 55 :. omo .o 55oo3 5:55 55 Hgm.o3 moon omoco>5 on. .5 o.55.

57

 

 

 

 

 

 

555 555 5.5 555 .55 555 5555555
55 5 555 55 55 5 555 5. 55 5 555 5. 55 5 555 5. 55 5 555 55 55. 5 555 55 55 5.555
55 5 5.5 55 55 5 555 55 55 5 555 55 55 5 .55 55 55 5 555 55 .5 5 555 55 .5 5.555
55 5 555 55 55 5 555 55 55 5 ..5 55 55 5 555 55 55 5 .55 55 55 5 555 55 .555555

.55 555 555 555 .55 555 5555555
.5 5 555 55 55 5 555 55 55 5 555 55 55 5 555 55 55 5 555 55 55 5 555 55 55 35.5
55 5 555 55 55 5 555 55 55 5 555 55 .5 5 5.5 55 55 5 555 55 55 5 555 55 .5 55.5

55 .555... : 5o .5: : 55 .955. : 55:5: : 55 .95... : 5a .m: : m: ...
5 5.52555. N 535.538 . . 5.33:8
.xo5 5:5 5555..:m: .55555: 5: m:.55:umm. v.55: 5:5 35.5 5:.
:o.5um.m5 we 5:5.»5:m:mm 555:» :m>o 55:.. a:.55:5555 v.55: 5:5 35.5 5:5 5:5 .o:u:ou
5:5 555 :o.55.>mc :55::555 5 Eu :. 505 we 5:553 5:555 55 5:5.53 55o: mmwsm>5 5:. .m 5.55.

523

 

 

mom «ma ”Na ~mc. .mm mac. HHHHH><
mm H mm“ NN cm. H mma a. Na. H an» a. an. H a... h. H_. H ..m mm om H a... m~ Na uHaag
mm H .Pm mm mm. H «no. cc «a H Ham on KN. H «as. aw .m H pm» cm ... H sac. H. .a uHaHm
HNN H How on .m H q~o_ mm NH H mom cc mop H ~mc_ cm «H H Hem mu Na. H c¢__ HN .oHchu

Nos ~_c_ e_m mac. new sme— HmHHH><
mo H mks Hm ma H Hue. ms we. H H_m mc m__ H sec. NH cm H «am an Ne. H mhc_ HN ,Ng zo_m
mm H HHH mm cm H mac. m. ec_ H ”pm ms c__ H mac. mm ow H mmm co c.. H mp__ cc .z :o_m
mmpmsmk c mm Fm: : mm P95... : mm 5F : m0 Pagan : murmur! : 0:...—

 

 

m cowuagmcaw

N :o—uagmcmw

 

— copuagocmw

 

.xmm uco mumowpamg .gmnszc An mcwgmcuomH twang can
zo_m Low copuumpmm Ho mcowpmgmcmm muggy gm>o mmcwp mcvgogpmmw twang ucm sopm use can

Pogucou wsa so; :ovumv>mu ugmvcmum H

Em cw mam Ho mxmwz up an usmwwz zuon mmogm>m ask

.0 m—nmh

59

 

 

 

 

 

 

 

 

 

21.50"
Z
c:
E
a 21.000"
H
LIJ
Z
>-
53 20.50-
(D
>—
G:
D
u'_, 20.000- mo .
gg -er s.r. 1
.5- C. F. 2
...)(_. R. F. 3
19.500-
+ S. M. 4
* R. M. 5
* C. M. 6
.00 i i i i i 1‘ i .
‘9 1.000 . 2.000 3.000
GENERHTIONS

Figure 3. Changes in one-day body weight of maies and femaies of
the siow and rapid feathering lines compared to the
non-selected controi line (random) over generations.

 

60
4.8 gm in the three generations, respectively. The differences of
four-week female body weight between rapid and slow line increased
as selection progressed while these differences in male progeny were
somewhat erratic. Also, the changes in body weight, four-week old,
due to selection was inconsistent among lines and sexes as indicated
in Figure 4.

The associated phenotypic change in body weight at eight weeks
of age is given in Table 5. Males were heavier than females by an
average of l05, 100 and 97 gm per generation in the slow, control
and rapid feathering lines, respectively. With sexes combined,
rapid feathering individuals out-weighed the slow feathering indi-
viduals by an average of 3.0, 22.0 and 24.0 gm for the three suc-
cessive generations.

In each of the selected lines and the non-selected control,
there was a constant decrease in body weights over generatibns
(Figure 5).

As had been true for eight-week body weight, a constant decrease
of the same pattern in l2-week body weight was observed on a line-
sex basis in succeeding generations (Figure 6).

Table 6 contains body weights and standard deviations at l2
weeks of age on sex and line generation basis. The differences in
body weight between sexes were pronounced and systematic by subse-
quent generations. On an average per generation, male individuals
were heavier than females by 250, 247 and 230 gm in the slow, con-

trol and rapid feathering lines, respectively. The differences

61

 

23

 

220’....-\\

 

 

 

 

 

 

LEQEED

Il’HMHP

OG‘UNF‘

 

E
c:
S; 210-
E’.
u;
3 \
ES 200%
D
an
s:
uJ
u;
:z 19'
J,
180-
54000 2,500 3.600
GENERRTIONS
Figure 4. The observed mean changes of four-week body weight in

males and females of the slow and rapid feathering lines

compared to the non-selected control line (random

generations.

over

 

62

 

 

 

 

 

 

 

 

 

 

 

z:
(D
....
a:
0
0—0
04
2:
)_
c:
c:
an
s:
04
04
'7‘ new
a: -er S.F 1
-EP C.F 2
AM- R.F 3
-o- s.M 4,
*KM 5
_ -x— C-" 6
450 : : : : : : H4 .
1.000 2.000 3.000

GENERHTIONS

Figure 5. Changes in eight-week body weight of males and females of
the slow and rapid feathering lines compared to the non-
selected control line (random) over generations.

63

 

 

 

 

 

 

 

 

 

new

HM“:

333-1171;"

QU'IAUNH

 

 

 

11$}
23 1050-
I:
z:
2 u q
UJ
3
E
C: 950-
a:
x:
n:
“J ..
7r
2‘.
as
3E::::II====a====EEEEEEEEE:::::::::““~‘~‘~“::::
fiiooo 2.000 ' ' I 3‘000
GENERHTIONS
Figure 6. Changes in lZ-week body weight of males and females of

the slow and rapid feathering lines compared to the non-

selected control line (random) over generations.

 

64
between the two divergent lines were somewhat erratic over genera-
tions. The data in Table 7 indicate that the observed changes in
body weights at various ages were likely to yield an irregular fit
in both sexes within each line.

Related reports concerning body weight in chickens by Maloney
and Gilbreath (l966), Festing and Nordskog (l967) and Carte and
Siegel (1968) indicated that in divergent selection experiments an
asymmetrical response is likely to be realized fairly frequently.
This is to say, the progeny of plus selection gain more than the
progeny of minus selection lose. Also, they concluded that male
individuals gain or lose more than female individuals do, in divergent
lines. In this study, although the main selected trait was the
feathering rate, not body weight, it can be observed from Table 7.
that when both divergent lines had positive response to selection,
the plus selection line gained more than the minus selection line
did. 0n the other hand, when both selected lines had negative magni-
tude, the rapid feathering line lost less than the slow feathering
line. Indeed, these findings are consistent with the first conclu-
sion of the above investigators.

As for the response of feathering to selection in the two sexes,
when feathering achievement showed a positive magnitude, female
progeny of the two divergent lines had higher feathering scores than
male progeny. However, when the magnitude was negative in both
lines, the loss was almost even in both sexes. These concomitant

differences, in response of feathering to selection between the two

65

Table 7. The magnitudes and signs of the phenotypic changes of
feathering scores and associated traits in two subsequent
generations of selection for slow and rapid feathering by
lines and sex.

 

 

 

 

 

 

 

 

 

From 61 to 62 From 02 to GB
TSlow Feathering
Traits M F’ TCombined7 M 7F7 combined
Feathering + 5.6 + 7.5 + 13.1 ~15.5 -15.6 -3l.1
4-Wk Tail
Length (cm) + .7 + l.l + 1.8 - .9 - .7 - l.6
l-Day 8N (gm) - l.6 - 1.6 - 3.2 - .9 - .8 - 1.7
4-wk 8w (gm) - 21.3 - 2.9 - 24.2 -12.7 -12.6 «-2543
8-Wk 3w (gm) - 70.3 - 76.4 -146.7 -27.4 -l3.6 -4l.0
12-Wk 3w (gm) - 40.3 - 29.5 - 69.8 -38.7 -15.8 -54.5
Unselected Control
Feathering + 4.7 + 8.7 + 13.4 -15.2 -14.8 -30.0
4-wk Tail
Length (cm) + .2 + l 3 + 1.5 - 1.0 - 1.3 - 2.3
l-Day BN (gm) - .2 - 1.6 - l.8 - .9 - .2 - l.l
4-wk 3w (gm) - 15.8 - .5 - 16.3 -l9.8 - 3.5 -23.3
8-Wk 3w (gm) - 84.0 - 40.5 -124.5 -38.9 -ll.6 -50.5
lZ-Wk 3w (gm) -108.0 - 41.0 -149.0 - 1.0 - 8.3 - 0.3
Rapid Feathering
Feathering + 7.0 + 10.0 + 17.0 ~ -l3.0 -12.2 -25.2
4-wk Tail
Length (cm) + .6 + l.l + 1.7 -. .7 - .5 - l.2
l-Day 811 (gm) - .5 - .2 - .7 . - .2 - .7 - .9
4-wk 3w gm) - 4.2 - .3.6 - 7.8 - -19.l - 9.9 -29.0
8-Wk 3w gm) - 51.8 - 48.2 -100.0 -47.8 -11.6 -59.4
12-Nk BW (gm) - 42.1 - 8.9 - 51.0 -62.2 -16.9 -79.1

 

66
sexes of the two selected lines, were found to be significant (p<.05),
Appendix Table 12.

The occurrence of sexual dimorphism in feathering response due
to selection, with females being more responsive than males, in this
study is in agreement with Siegel et al. (1957a) who reported that
female progeny feathered better than males in a two selected lines
(superior and inferior) experiment.

Generally, the variation in feathering ability between males
and fema1es is credited either to a dosage effect of sex-linked
genes or to some physiological differences between the two sexes.

Greneman (1941) reported considerable pituitary activity during
the first few weeks of R. I. R. and White Leghorn chick's life. This
early pituitary activity results in gonadotrophin secretion which
indirectly results in the small amount of production of gonadal
hormones, estrogen and androgen. These hormones are produced due to
stimulation of gonadal growth.

Female chickens have a greater amount of estrogen than males
(Sturkie, 1976). This was reported to be responsible for better
feathering in female than in male chickens by Siegel et al. (1957a).
The unidentical response in divergent lines from bi-directional
experiments has been investigated by Falconer (1953). He ascribed
the asymmetric condition to differences in the gene frequencies for
those genes governing the traits. Falconer (1960a) also suggested
that the asymmetrical response in divergence between upwards and

downwards selected lines are consequences of changes in some

67
parameters such as phenotypic standard deviations and covariances,
due to selection applied. Another possible cause for the asymmetry
reported by Falconer (1960a) is the different selection differentials
in the divergent lines. As for correlated traits in selection
experiments, the condition of asymmetrical response is extremely
frequent according to Bohren et al. (1966).

In the current study, correlated traits, body weights at various
ages and egg production, showed irregular changes in the divergent
lines when compared to the unselected control line (Table 7). The
discordance of the pattern of correlated responses has been studied
by several investigators.

Clayton et a1. (1957) from a bi-directional experiment for high
and low sternital bristle number in Drosophila melanogaster, have
observed an asymmetrical response in sternopleural bristle number.
They attributed the condition to the phenomenon of gene drift when
genetic correlation is low. Siegel (1962a) attributed the cause to
some differences in the genetic variance or heritabilities for the
traits. Bohren et a1. (1966) pointed out that the condition of
asymmetric response for correlated traits is from loci contributing
negatively to the covariance of the traits and having frequencies
other than .5. Accordingly, one or more of the possible reasons for
the asymmetric responses of divergent lines for selected and corre-
lated traits should be the cause(s) of the asymmetrical responses
for feathering, body weights and egg production in the two divergent

lines of this study. However, the most tangible reason for this

68
condition is believed to be the different selection differentials in
the rapid and slow feathering lines (Table 8). Genetic correlations
between feathering and the correlated traits will be discussed

throughout this manuscript.

HeritabilityEstimateslgf.Feathering

Generation trends in heritability estimates were based on the
combined averages, of the sources of estimates, of the replicates
within each line. Realized heritability for the selected trait,
feathering, was estimated for males and females combined of the
rapid and slow selected lines, for two generations of selection.

The structure of response on the cumulative selection differ-
ential and the realized heritabilities of feathering of the selected
lines are presented in Tables 8 and 9.

Falconer (1960a) reports that this method of estimation provides
the most useful description of the effectiveness of selection. Ideta
and Siegel (1966a) in estimating realized heritability of selected
and correlated traits in chickens described it as a very reliable
measure of additive gene action.

Since the heritability values of the slow and rapid feathering
groups found herein fall within the biological range (0.0-1.0),
these estimates are to be considered reliable and realistic values.

The heritability estimates for feathering, eight-week and 12-
week body weight derived from intra-sire regression of offspring on
their dams with sex difference adjusted are presented in Table 10.

In calculating the regression, each dam within a sire was included

69

.mcoHuococoa an acne: =o.»o_=aoa mo mucoHUHHHuou =o_mmocooc as» as» mo=.a>-an

.mmanuH—amc ozu no monoco>u use mm:.— mcvcmsuuom v—noc was :o—m Ho museum mcvcugaemua

 

 

 

 

 

em.o op.—- ~o.e~— pm.“ ~_.¢e. po.um_ ~_.m_ c~.p¢_ no.m~_ apnea
om._ o~.e- sm.m_p mm.~ m~.~m— mm.em. ~o.o m~.m~_ m_.m~_ _ccu:ou
um.w- m¢.m- “a.m., mm.m- pe.m~_ mm._m_ -.o_- oo.e_p m~.e~_ zo_m
.HH.= Haapa>-a cam: no: .HHHa ..Hm HHeacHa com: apa .HHHQ ..am HHemcaa cam: aoa H=HH

:c.uuwpom vouumpmm umuum—mm

mmagm>t mmmcm>< ommcmwn

m copuacoeau N coruucmcow _ :oHHHmemu
.mcowpmcmcmm co mcmwe mcwcmcgmmm mo mangUPHHmou cowmmocmmc on» can mpmvucmcmHHHu

cowuumpmm can mucmgma umpumpmm .m:o_umcm:mm omega Ho mommcm>m acoum m=_cmcuomm .w m—nMH

70

Table 9. Regression estimates of the realized heritabilitya on
generations for feathering of selected lines.

 

 

Actual Ave. Sel..
Line b-Value Response "R"b Diff. "s" Realized h2
Slow -5.45 -.55 -8.323 0.08
Control -4.79
Rapid -1.10 3.69 10.343 0.36

 

aRealized h2 = R/S.

bResponse = Deviation of the regression within a selected line
from the regression within the control line.

71

 

 

 

Table 10. The heritability estimates i standard errors of feathering
and body weights by intra-sire regression of offspring on
dams.a

Traits

Ljne Feathering 8-Wk 8H 124Wk78w

Slow 0.11 i .15 0.66 i .35 0.73 i .37

Control -0.02 i .21 0.18 i .31' - .01 i .32

Rapid 0.60 i 1.06 0.18 i .73 0.67 i 1.29

 

aFor the control group the estimate is for within generations

rather than within sires.

'72
as many times as she had progeny. Whereas, for the control group,
where data on birds were produced from multiple-sire mating, the
regression of offspring on dams within generations were calculated.

Displayed in Appendix Table 10 are the average mean squares
and mean cross products of dams and progeny for feathering and body
weights with the number of sires and dams of replicates pooled
within each line. As shown in Table 10, the heritability estimates
for feathering of the slow, control and rapid feathering groups were
.11 t .15, -.02 i .21 and .60 i 1.06, respectively. The low to high
estimations obtained by intra-sire regression analysis for the slow
and rapid feathering lines are not in extreme disagreement with those
estimates based on the cumulative effects of selection (.08 and .36
for the slow and rapid feathering lines, respectively). The herita-
bility estimate of feathering in the control group was -.02 i .21.
This negative estimation, below the theoretical minimum heritability,
indicates that the heritability of feathering in the unselected line
should be considered a very low estimate.

About the reliability of the estimation from intra-sire regres-
sion of offspring on dams, El-Issawi and Remple (1961) stated that an
advantage of parent-offspring regression as the basis for estimating
heritability is that it is subject to less bias from variance in
genotype-environment interaction than half-sib correlation method.

Additional estimates of heritability for feathering were cal-
culated from sets of full-sibs and half-sibs correlation analysis

technique.

73

The nested model of King and Henderson (1954) was adapted in
this study with male and female data analyzed separately. Appendix
Table 7 contains the variance components of sire (not for control
group), dams and full-sibs for each sex and line, from which esti-
mates of heritability of feathering and body weights were derived.
The parameter estimates of heritability and standard deviations for
feathering obtained from the analysis of variance technique are shown
in Table 11.

The average heritability estimates for feathering based on sire
component of variance, h; , for male progeny were .77 t .30 in the
slow feathering line and .10 i .18 in the rapid feathering line.
From dam variance components, “3’ these estimates were .09 i .14
and .85 i .37 for males of the two respective lines.

Based on female progeny, heritability estimates from (h;)' were
.15 e .17 and -.15 e .24 and from (hfi) were .32 e .20 and .83 e .28
for the slow and rapid feathering lines, respectively. The negative
heritability estimate might be due to some effects of genotype-
environment interaction.

In the randomly bred line, control, heritability was estimated
only from dam variance components, “0' For male progeny it was
1.59 i .64 and was .06 i .4 for female progeny.

Generally, estimates based on male progeny were larger than
those based on female progeny (except ha for slow feathering line).

The standard errors of the heritability estimates from sire

(hg) and from dam (h6)_ variance components are considered extremely

74

 

 

 

 

 

 

 

 

 

HF. H mm. mm. H we. mm. H mm. cm. H m_.- mm. H mo. we. H no._ twang
--- .1. me. H on. mm. H m_.p --- --- Pocacoo
m_. H «H. mo. H em. up. H Pm. mp. H mm. cm. H 50.. mp. H mm. gopm
HemHez Heem Heez-e=oa
mo. H mm.” mp. H so.P mm. H mm.m mm. H mm.m «H. H mm.- cm. H No.1 twang
.1. .1. me. H mm.~ mu. H mm.m .1. 1.. Focgcou
NN. H mp._ up. H ~N.~ mm. H mo._ mp. H cp.m pm. H mN. mm. H mm.- zopm
HemHez Huem Hmo-eeo
mo. H «m. op. H mm. mm. H mm. mm. H mm. em. H m~.- mp. H op. twang
--- --- OH. H no. we. H mm._ 1-- 1.. pocucoo
NF. H mm. mP. H me. ow. H mm. e_. H mo. up. H mp. cm. H mm. zepm
mmpmemu mmpmz mmpmemu mme: mmpmsmul mmbmz map;
Ag + mv e a e
N N
mcwgmgummm
a.mucocanoo mucmwem> pcmcmq Easy Acmmoca mpmsmm
use mpme Ho mugmwmz anon ecu memcwgpmmm mo mcogcm ucmccopm H xuwpwnmuwsw; mo mmueswumu .FP mpam»

75

.mmcwp
umuumfimm on» mo comm Ho mmumum_amc o3» Ho mucwcoasou oucmwcm> wmmcm>m soLH umpzaaou mmuesvpmum

 

 

 

 

 

 

 

OH. H H_. m_. H Pm. .N. H ow. Hm. H Ho.- mm. H mm. mm. H mm. eHaea
--- --- NH. H mm. mm. H om.p --- --- HoeHeeu
m_. H Hm. e_. H PH. em. H mm. mm. H Hm._ mm. H mm. mm. H Ho. zepm
eHmemH .HHH Hmez-eaoa
m_. H om. op. H up. mm. H me. mm. H mu. Hm. H up. ow. H mo. HHQHH
--- --- _m. H mm. _m. H No. --- --- _eHHeeo
N_. H mm. .F. H mm. ON. H me. mm. H Hm. mm. H mm. «N. H HF. zepm
HeoHe: Heem Heez-e>_ezh
o_. H on. em. H HN. mm. H oe._ RN. H NF. _m. H m~.- em. H em. eHaem
--- --- me. H OH. mm. H o..P --- --- HoeHeeu
N_. H m_. Ho. H HF. m_. H .P.- m_. H mm. HN. H me. mp. H mo. zo_m
marmsmu mmpez mmmemu mmpez mmpmsou maps: mew;
3+3 a m
we N; N;

 

HeuHez-Heem Heez-HemHu

 

A.e.HeoeV ._P e_eee

76
large, due to small size of the population from which the estimates
were derived. Therefore, the reliability of these estimates should
be accepted with caution.

From Table 11 it can be seen that the estimates of heritability
from sire plus dam variance components in male progeny were compara-
tively higher than those in female progeny. In any of the methods
used, when the estimate was beyond the biological range of herita-
bility (0.0-l.0), it should be considered too low or too high an
estimate. Heritability from half-sib correlations (h§ ) usually is
considered to give the best estimate because it contains all the
additive genetic variance, under the assumption that there are no
sex-linked genes influencing the trait. 0n the other hand, estimates
from dam components (ha) contain the additive genetic variance and
all of the dominance variance. The full-sib correlation method of
estimate (h§+D) contains one-half of the dominance variance in addi-
tion to the additive genetic variance. Consequently, the latter
methods should be considered biased upward and the half-sib correla-
tions (hg) method is the reliable method of estimation. However,
due to the small number of sires used in this study, the estimates
from full-sib correlation are to be considered better estimates,
since they are based on higher number of degrees of freedom.

From the full-sib correlation analysis, the heritabilities
obtained for tail length at four weeks of age were .71 i .14 and
.34 i .19 for males and females of the slow feathering group while

they were .31 i .19 and .17 i .10 for the respective sexes in the

77
rapid feathering group. The relatively small standard errors along
with the higher number of degrees of freedom in the full-sib correla-
tion method, make those estimates acceptable, despite the upward bias
caused by dominance variance.

Since the population of the current study was a finite size,
there was some mating together of relatives. However, the birds had
a relatively low inbreeding coefficient, thus no adjustments of the
genetic parameter estimates were made.

From the three methods of estimates, one feature seems consis-
tent. When the estimates were combined for both sexes based on (h§+0)’
it would appear, from the average of the estimates by the three
methods applied, that the degree of feathering is heritable within
the range of .33-.36. In the light of these findings, it can be
suggested that individual selection combined with family selection
should be effective in improving the trait, and should be practiced
for best results.

From the literature reviewed, a meager number of reports deal-
ing with heritability estimates of feathering were found.

Heritability of feathering at eight weeks of age in chickens was
.51 according to Jaap and Morris (1937) and .34 according to Siegel
(1963). Hurry and Nordskog (1953) reported a heritability of .42
and .33 for feathering in Barred Rocks and New Hampshires, respec-
tively. An estimate of heritability for early feathering in White

Plymouth Rocks was .40 (Seigel et al., 1957a).

78
The heritability estimates of feathering in rin-necked pheasants
obtained in this study are in fairly good agreement with those re-

ported above.

Heritability 9f Associated Traits

To estimate the heritabilities of other traits, the principal
procedure used was the method of variance component analysis. In
addition, the method of intra-sire regression of offspring on dams
was employed for eight-week and 12-week body weight heritability
estimations. The use of the latter method of analysis upon these
traits was due to the fact that body weights at these ages are con-
sidered crucial traits.

Realized heritabilities of the associated traits were not cal-
culated. This, due to the fact that heritabilities of associated
traits are not equal to the changes in these traits divided by their
secondary selection differentials.

The mean heritability estimates and standard errors of body
weights at one day, four weeks, eight weeks, and 12 weeks of age are
presented in Table 11. By means of dam variance component analysis
and from (h§+D), heritability estimates of one-day body weight were
too high, >l.0 in both sexes. Since heritability cannot exceed
unity as a biological maximum, these estimates should be taken as
indications that one-day body weight is presumably a highly heritable
trait in ring-necked pheasant population. However, the very high
estimations may arise from differences in the variance components
of non-additive effects or may be attributed to the role of sampling

ETT‘OY‘S .

79

Age of birds had an obvious influence on the relative size of
the heritability estimates in which no negative values were obtained
when full-sib correlation method of estimation was used (Table 11).
The heritability estimates for four-week body weight range from
-.18 i .20 to 1.07 i .46 for rapid feathering males; whereas, in the
female progeny of the same group the estimates range from .08 i .33
to .98 i .28. In the slow feathering line, the heritability esti-
mates were from .23 i .18 to .26 i .19 in the male progeny and from
.21 i .16 to .67 i .30 in the female progeny. The control group had
higher heritability estimates (1.15 i .57 and .76 i .49) in both male
and female progeny than the selected lines.

As for eight-week body weight, the pooled heritabilities computed
from sire and dam components of variance were .17 i .07 and .24 i .54
for the males of the slow and rapid feathering lines, respectively;
whereas, for females of the same lines the estimates were .19 i .12
and .66 i .10. On the other hand, based on intra-sire regression
technique, with sex effect adjusted, the estimates were .66 i .35
for the slow feathering line and .18 i .73 for the rapid feathering
line (Table 10). For 12-week body weight, based on full-sib correla-
tion analysis, the heritability estimate of male progeny was .35 i .11
in the slow feathering line and .17 1 .10 in the rapid feathering
line, while females of the two lines had heritabilities of .35 i .12
and .30 i .13, respectively. By means of regresSion analysis, the
estimate for individuals of the slow feathering line was .73 i .37 and

for those of the rapid feathering line it was .67 i 1.29.

80

For these traits that were subjected to the intra-sire regression
analysis, the estimates, by this method, should be considered more
reliable and free of errors. This is due to the fact that no errors
arise from dominance and from the possible existing correlations be-
tween the epistatic deviations of parents and progeny.

The relatively high heritability of ring-necked pheasant body
weights, according to this study, at 12 weeks of age indicates that
mass selection for superior individuals would be effective in further-
ing improvement of this character.

Reports of heritability estimates of body weight in poultry are
too numerous to compare with those found in the present study.
Generally, the estimates found from the current study agree quite
well with those reported in the literature, where the estimates run
as low as .05 f0r early body weights, reported by Goodman and Jaap
(1960), to as high as .88 for broiler chickens by Thomas et a1.
(1958).

Correlations Between Feathering and Other Traits

The components of variances and covariances, averaged over
replications, from which the correlation estimates were computed are
presented in Appendix Table 6. In accordance with those variance
and covariance components, phenotypic and genetic correlations and
standard errors of estimates were calculated and are presented in
Table 12. The calculations were based on the average contribution

of sires, dams and both combined, within each line and sex.

81

 

 

 

 

 

 

 

 

 

N_. H om. ow. H mm. mF. H pm. 1.. H mo.P --- H mo.- mm. H cw. oe. om. apnea

--- .1. mm._ H mm. o_. H mm. .1. 1.. um. mm. Pocucou

ow. H mo. Fm. H we. 1.. H mp.- mm. H mm.. Pm. H mo.- .1. H NP.F mm. om. zo_m
Hemwwz Heom Hmmz-H;uHm

mo. H mm. up. H Po. HF. H mm. -1- H em. --- H mo. pm. H me. we. mm. twang

--- --- ou.~ H mo.- «P. H on. .1. .1. me. mm. Fogucoo

cm. H «o. “F. H mm. m¢.. H mm.. mm. H up..- oH. H mm. mm. H mm. «H. Ne. zepm
HemHez Heem H883-H=oa

50. H Po.- _N. H mo. ob. H No.1 PNH H m_. mm. H o~.- --- H mm.- No.1 No. twang

1.. 1.. --- H Hm._ mm H um. 1.. 1.. m—. o_. —ogH=oo

mm. H PP. «F. H no. «N. H o_. o—. H on. hm. H mp. --- H No. po. PF. zopm

mmrmsmu mmpmzl mmbmemu mmpmz murmsmu Hopes» mmpmsmul Hope: mew;

Ag + mv we may we Amv we
Hgmwmzeavomixmo-mco,.
m.xmm ecu mmcwp an xcmmoca mpmsmw ecu mpms Ho mugmwmz anon use
mcwcwcHemH cmmzpmn meoccm ugmecmum H meowuepmccou ovumcmm ace owazuocmnn Ho mmueawumm .NF mFDHH

82

.m6:.. umuom.mm 6:» Ho mmumu..amc
03H ecu mo mucmcoasou mucewce>ou new moco.cm> Ho mmmmcm>m use seem umuzasou mco.pm.mgcoum

 

 

 

 

 

 

 

Ho. H mm. .N. H OH. --- H mo.. --- H mo. --- H mo. --- H .o.N- OH. NH. e.ae¢
1.. 1-1 mm. H mm. ... H cm. 1-- 1.: me. we. .ochou
cm. H me. 8.. H «H. mm. H 8.. NH. H .m.- 8.. H .N. --- H m..m me. NH. 38....
eHmeeH ..HH xemz-e=oa
NN. H m..- mo. H em. .m. H e..- --- H a... --- H No. mN.. H me.- mN. m.. c.88m
--- --- --- H No.- mo. H mm. --- --- .N. mm. .eeHeoo
mN. H m.. «N. H o..- NN. H .8. mm. H m..- m.. H mm.. OH. H N..- eN. mm. zo.m
mm.msmu mmpmzl mmbmemu mmpmzlli mmpmsmui mmpmz mmpmsmu warm: «cw;
Ac + mv as Amy o; va us a;
Hem.ez Heem Hmmz-e>.ez.
..e.H=oev .N. H.8H.

83
Because of the small number of sire groups, the case of sire compo-
nents plus the dam components analysis are to be considered the most
important estimates.

Phenotypic correlations between feathering and one-day body
weight were so minimal in the two selected lines that they could be
considered trivial values. Genetic correlations between feathering
and one-day body weight based on sire and dam components were
.07 i .14 and .05 i .21 for male progeny in the slow and rapid
feathering lines, respectively. Estimates for female progeny were
.11 t .29 and -.01 i .07 in the two respective lines. Feathering and
body weights at four weeks and eight weeks of age in the rapid feather-
ing group had phenotypically and genetically (r (S + 0)) positive
correlations. Consequently, it can be hypothesized that some of the
genes acting on feathering also affect growth rate at those ages.
Thus, selection for rapid feathering individuals would improve growth
rate at these ages and vice versa. These results and conclusions are
in good agreement with those reported by Hurry and Nordskog (1953)
in chickens. In this respect, the findings of this study were also
consistent with earlier reports by Martin (1929), Schnetzler (1936).
Jaap and Morris (1937) and Glazener and dull (1946) who indicated
that better feathering birds were heavier than poor feathering birds.

In the two selected lines, Table 12 indicates that correlations
between feathering and body weights at 12 weeks of age were somewhat
erratic. The relatively different estimates of genetic correlations
might be ascribed to the small number of observations and thus due

to sampling errors.

84
For the control group, where no selection was practiced, pheno-
typic and genetic correlations between feathering and body weights

were larger for males than for females at older ages.

Egg.Production

 

The two-way analysis of variance was the method used in analyz-
ing the data of egg production by lines and generations. Results of
the analysis, in accordance with this method, are given in Appendix
Table 11.

Although data of fertility and hatchability were not statistically
analyzed, it seems desirable to indicate the response of the repro-
ductive traits to selection for feathering. Table 13 contains data
of egg production, fertility and hatchability by lines and genera-
tions.

In the base population, average hen-housed egg production for
birds selected as rapid feathering excelled that of those selected
as slow feathering birds by 11% and that of unselected control birds
by 13% in a 61—day period.

As a consequence of one generation of selection in opposite
directions for feathering a specific trend for decrease in egg pro-
duction in the two selected lines was observed. .This decrease was
10% and 35% for the slow and rapid feathering lines, respectively.
Also, the slow feathering hens produced 14% more eggs than their
counterpart in the rapid feathering group, in the 61-day period.
When selection for feathering was further practiced, to produce

generation two, concomitant changes were observed with, surprisingly,

85

 

 

 

 

 

 

 

 

O.O. 0.0. 0.0H OH .H. -- OOOHO>H
O..O 0.0. ...H OH OO. O NO O.OOO
O.O. O.N. N.OH OH OO. O .O 0.000
O.ON O.OO 0.00 OH .OO N. .OOHOOO
O.OO N.OO O.H. OH OON O OOOHH>H
O.NO O.N. 0.00 OH .OO O NO 3O.O
..OO H.HO H.OO OH OON O .O 3O.O
N :o.HOcm:mw
O.ON 0.00 ..OH .O ONN O OOOHO><
H.OO 0.00 O..O .O NO. O NO O.OOO
..N. 0.00 O.HO .O OON O .O O.OOO
O.H. ..N. 0.00 .O OOH N. .OOHOOO
N..O ..OO ..OO .O .OH N. OOOHOOH
N.HO O.HO O.HO .O NOH N. NO 3O.O
N... H.OO O.HO .O NNH N. .O 3O.O
. OOHHOLOOOO
O.H. O.ON N..O .O OO. O. HOOHO>H
O.N. ..OO 0.0N .O ... O. NO O.OOO
O.O. ..OO ..HO .O .NO O. .O O.OOO
0.0. ..H. O..O. - .O NOO O. .OHHOOO
..O. H.O. O.OO .O .OO O. OOOHO>H
N.H. 0.0. O.OO .O NNO O. NO zO.O
O... O.ON ..O. .O OOO O. .O 3O.O
OH...OOOOHO: HH...HHOH OO..OOOO.OHO HHOO .Oz HOOO .Ozi.OHOH1‘ Hem: .Oz OO.O
OOHHO.OOOO mem
.OOOHHOLOOOO can Omuoo..Omc
.Om:.. On .ocucou umaum.mmuco= age on umcmasou mcwcmgummw O.OO; new 3o.m so» umuumpmm

mgmcmmcn Hcmmamga mo OHH.HOO;OHO; OOO >u...asmm .OOHHusuoga mom mo Ommncw>O Och .m. m.OOH

86
tremendous increase in egg production for the slow feathering hens
but with a further decrease in egg production for rapid feathering
hens. Accordingly, one might suggest that selection for rapid
feathering depresses egg production from one generation to another.
The possible cause of the consistent decline may have been due to
negative pleiotropic effects between the two traits.

Considering egg production a secondary trait in this two-way
selection study, results herein agree with Falconer (1953) who stated
that "If the secondary character forms an important component of
total fitness it might be expected to decline in response to selec-
tion of the primary character in either direction." One shortcoming
of this study is the lack of phenotypic and genetic correlations be-
tween feathering or body weights and reproductive fitness.

Figure 7 shows changes in egg production for the two selected
lines in comparison to the randomly bred control line over genera-
tions. One essential consideration shown in Figure 7 is that egg
production for hens of the control group was linear over generations.
Statistical analysis indicated that there was a significant differ-
ence (p<.05) in egg production by generations. Too, there was a
significant difference (p<.10) in egg production line-generation
interaction, indicating that differences in performance between

selected lines in a given generation are important variations.

87

 

 

 

 

 

 

 

 

 

 

 

90-
:2
2
1.1.1
0
a:
u:
a.
2:
c3 7
H
.—
c3
2)
c: .
O
a:
0.
:3
“J 50*
Lfififiup
- .9. Slow 1
-&-Control 2
.)(.. Rapid 3
m_ : i i i E .
(8000 2.000 3.000
GENERHTIONS
Figure 7. The observed mean changes in egg production in the slow

and rapid feathering lines compared to the non-selected
control line (random) over generations.

VI. SUMMARY AND CONCLUSIONS

Foundation stock for this study was selected from a large flock
of ll-week-old ring-necked pheasants. Three groups of birds, includ-
ing males and females, were selected for feathering on the basis of
phenotype insofar as feather pattern maturity was concerned as
follows: well developed, poorly developed and a control group which
was random collected.

The objectives of this study were to estimate genetic parameters
of feathering and the response to selection of traits, body weights
at various ages and egg production, correlated with feathering and
to assess the relationship between feathering and body weights at
various ages.

A grand total of 1,220 pedigreed progeny from three generations
were involved in the statistical analysis of this study. Three "-
methods of estimating the heritability for feathering were used.

The first one was based on the ratio of actual responses to selec-
tion over cumulative selection differentials to each of the two
divergent lines adjusted according to the control group (realized
heritability). The second method of estimate was based on offspring-
dam regression and the third on the estimation of variance components
from a hierarchical analysis on a within line-sex basis.

Phenotypic and genetic correlations between feathering and other
traits were calculated on a within line-sex basis from the variance

and covariance components.
88

89

From the data of three generations of selection for feathering

and of associated traits of a small population of ring-necked

pheasants, results clearly demonstrated the following:

1.

The average realized heritability of feathering up to 12
weeks of age was estimated to be .08 for the slow feather-
ing line and .36 for the rapid feathering line. From the
regression of offspring on dams, the heritability of
feathering was found to be .11 i .15 and .60 i 1.06 for the
slow and rapid feathering lines, respectively. Based on
the method of full-sib correlations, heritability of feather-
ing in the male progeny of the slow feathering line was

.43 i .15 and .38 3 .10 for the males of rapid feathering
line. In the female progeny, the heritabilities of the two
respective lines were .23 i .12 and .34 i .08.

Due to the low to moderate estimates of heritability of
feathering, some of the genetic variations are presumably
of the additive type. This would suggest that mass selec-
tion at early ages, four to eight weeks, would be effective
in improving the feathering in ring-necked pheasant popula-
tions.

Based on calculated feathering scores from 80.21 to 172.80,
the mean feathering score achieved, in the rapid feathering
line, from the first generation of selection in the female
individuals was 10.05 compared to 7.04 in the male indi-

viduals. This variation was found to be significant (p<.05),

90
indicating the existence of sexual dimorphism for feathering
response due to selection in pheasants.
The responses of feathering to the plus and minus selection
resulted in mean feathering score differences of female
progeny between the two selected lines of 3.1, 5.6 and 9.1,
with the rapid feathering birds being superior, for the
first, second and third generation, respectively. On the
other hand, these differences in the male progeny were found
to be 3.5, 4.9 and 7.5 for the three respective generations.
Correlated responses of body weights and egg production to
selection for feathering were somewhat asymmetrical. Body
weight of the rapid feathering birds, with both sexes
combined, at four weeks of age had mean differences of 1.3,
6.8 and 4.9 gm compared to their counterparts in the slow
feathering line for the first, second and third generation,
respectively. At eight weeks of age, these differences were
fOund to be 1.7, 21.2 and 12.5 gm for the three respective
generations, with the rapid feathering individuals being
heavier. However, at 12 weeks of age the differences in
body weights between the two selected lines were rather
erratic.
Data of egg production showed a significant (p<0.10) inter-
action between generations and lines with a more rapid de-
cline in egg laying performance in the line selected for

more rapid feathering than the line selected for slow

91
feathering. This would indicate that selection for feather-
ing affects egg production in pheasants. Thus, selection
for rapid feathering should be practiced with birds of high
egg production potentiality.
The heritability of four-week body weight as estimated from
full-sib intra-class correlation was found to be in a range
of .24 i .09 in the male of slow feathering line to .53 i
.14 in the female of rapid feathering line.
As for eight-week body weight, the estimated range was from
.17 i .07 to .66 i .10 in the two respective sexes and lines.
By the same method of analysis, intra-sire correlation, the
heritability of body weight at 12 weeks of age was found to
have a range from .17 i .10, in the males of the rapid
feathering line, to .35 i .12, in the females of the slow
feathering line. From intra-sire regression of offspring
on dams, the heritability of body weight at eight weeks of
age varied from .66 i .35 to .18 i .73 and at 12 weeks of
age from .73 i .37 to .67 i 1.29 in the slow and rapid
feathering lines, respectively. Considering these esti-
mates, phenotypic selection for body weights at four, eight
and 12 weeks of age should be effective in promoting genetic
improvement of these characteristics in pheasant population.
Estimates of the phenotypic correlations between feathering
and four, eight and 12-week body weight obtained from vari-

ance and covariance analysis were positive in the two

92
selected lines. The genetic correlations, as estimated
from sire plus dam variance components, between male's body
weights at these ages and feathering of plus selection line
were .61 i .17, .52 i .40 and .84 i .09, respectively. In
the females of this group, these genetic correlations were
.85 i .05, .30 i .12 and -.13 i .22 at the three respective
ages. Using the same method of calculation, the genetic
correlations between feathering and body weights of the
minus selection line at four, eight and 12 weeks of age
were estimated to be .04 i .30, .08 i .40 and .13 i .29 in
the female progeny and .58 i .07, .48 i .21 and -.10 i .23
in the male progeny.
These figures of the phenotypic and genetic correlations be-
tween feathering and body weights at various ages furnish
good evidence that selection for chicks exhibiting rapid
feathering development would automatically improve body

weight and vice versa.

BIBLIOGRAPHY

BIBLIOGRAPHY

Abplanalp, H., and I. L. Kosin, 1952. Heritability of body measure-
ment in turkeys. Poultry Sci. 31:781-791. '

Anderson, R. L., and T. A. Bancroft, 1952. Statistical Theory in
Research. McGraw-Hill Book Co., New York.

 

Asmundson, V. S., 1948. Inherited differences in weight and conforma-
tion of Bronze turkeys. Poultry Sci. 27:695-708.

Bacon, W. L., and S. P. Touchburn, 1976. Effect of light intensity
on confinement rearing of male turkeys. Poultry Sci. 55:999-
1007.

Bean, W. L., P. B. Siegel and H. S. Siegel, 1965. Light environment
as a factor in growth and feed efficiency of meat-type chickens.
Poultry Sci. 44:1009-1012.

Becker, W. A., 1975. Manual Qf_Ouantitative Genetics. (3rd Ed.).
Washington State University Press, Pullman.

Bell, D. D., and R. E. Moreng, 1973. Intermittent feeding and light-
ing of mature Leghorn hens. Poultry Sci. 52:982-991.

Bogyo, T. P., and W. A. Becker, 1963. Exact confidence intervals
for genetic heritability estimated from paternal half-sib
correlations. Biometrics 19:494-496.

Bohren, B. B., W. G. Hill, and A. Robertson, 1966. Some observations
on asymmetrical correlated responses to selection. Genetic
Research 7:44-57.

Boone, M. A., J. A. Davidson and E. P. Reineke, 1950.. Thyroid
studies in fast and slow feathering Rhode Island Red chicks.
Poultry Sci. 29:195-200.

Breneman, W. R., 1941. Growth of the endocrine glands and viscera
in the chick. Endocrin. 28:946-954.

Broemeling, L. D., 1969. Confidence intervals for measures of
heritability. Biometrics 25:424-427.

Brunson, C. C., G. F. Godfrey and B. L. Goodman, 1956. Types of gene
action in the inheritance of ten-week body weight and breast
angle of broilers. Poultry Sci. 35:524-532.

93

94

Burrows, W. H., and J. P. Quinn, 1935. A method of obtaining
spermatozoa from the domestic fowl. Poultry Sci. 14:251-254.

Buvanendran, V., 1969. The heritability and genetic correlations of
sexual dimorphism for lO-week weight in poultry. Br. Poultry
Sci. 10:321-325.

Card, L. E., and M. C. Nesheim, 1972. Poultry_Production. (11th Ed.).
Lea and Febiger, Philadelphia.

Carson, J. R., B. F. Bacon and W. A. Junnila, 1956. Colored illum-
ination during the rearing period and its effect on sexual
maturity and productivity. Poultry Sci. 35:1136.

Carte, I. F., and P. B. Siegel, 1968. Correlated heritabilities of
unselected traits. A check on procedure. Poultry Sci. 47:1821-
1827.

Clayton, G. A., G. R. Knight, J. A. Morris and A. Robertson, 1957.
An experimental check on quantitative genetic theory. III.
Correlated responses. J. Genetics 55:171-181.

Cole, L. J., and D. H. Reid, 1924. The effect of feeding thyroid
on the plumage of fowl. J. Agr. Res. 29:285-287.

Cuthbertson, G. T., 1980. Genetic variation in feather-pecking
behaviour. Br. Poultry Sci. 21:447-450.

Daghir, N. J., and S. L. Balloun, 1963. Evaluation of the effect of
breed on vitamin B6 requirements of chicks. J. Nutrition 79:
279-288.

Dakan, E. L., 1934. Light absorption, feed consumption, and the urge
to lay. Poultry Sci. 13:317-318. -

Day, E. J., and J. E. Hill, 1959. Zinc, potassium and certain other
minerals as additives to practical turkey starter and grower
rations. Poultry Sci. 38:646-649.

Deaton, J. W., F. N. Reece and J. D. May, 1970. Temperature and
light and broiler growth 2. Poultry Sci. 49:1593-1596.

Deaton, J. W., F. N. Reece, L. F. Kubena and J. D. May, 1976. Effect
of varying light intensity on broiler performance. Poultry Sci.
55:515-519.

Dickerson, G. E., 1969. Techniques and procedures in animal science
research. Techniques for research in quantitative animal
genetics. American Society of Animal Science 36-79.

95

Donaldson, W. E., G. F. Combs, G. L. Romoser and W. C. Supplee, 1955.
Body composition, energy intake, feed efficiency, growth rate
and feather condition of growing chickens as influenced by
calorie-protein ratio of the ration. Poultry Sci. 34:1190.

El-Issawi, H. F., and W. E. Remple, 1961. Heritability of growth
rate in inbred swine based on a crossbred foundation. J.
Animal Sci. 20:593-596.

Falconer, D. S., 1953. Selection for large and small size in mice.
J. Genetics 51:470-501.

Falconer, D. S., 1954. Validity of the theory of genetic correla-
tion. J. Heredity 45:42-45.

Falconer, D. S., 1960a. Introduction tg_Quantitative Genetics.
Ronald Press Co., New York.

Falconer, D. S., 1960b. Selection of mice for growth on high and
low planes of nutrition. Genet. Res. 1:91-113.

Festing, M. F., and A. W. Nordskog, 1967. Response to selection for
body weight and egg weight in chickens. Genetics 55:219-231.

Gehle, M. H., and S. L. Balloun, 1965. Selected hemocytological
effects of vitamin 86 deficiency in chicks. J. Nutrition 87:
197-201.

Gill, J. L., 1978. Design and Analysis gfi Experiments in the Animal
and Medical Sciences. Volume 1. The Iowa State Univ. Press,
Ames, Iowa. U. S.

 

Glazener, E. W., and M. A. Jull, 1946. Rate of feathering and ten-
week body weight observations in strains differing in shank
length. Poultry Sci. 25:433-438.

Glazener, E. W., W. L. Blow, C. H. Bostain and R. S. Dearstyne, 1951.
Effect of inbreeding on broiler weights and feathering in the
fowl. Poultry Sci. 30:108-112.

Godfrey, G. F., and G. M. Farnsworth, 1952. Relation of the sex-
linked rapid feathering gene to chick growth and mortality.
Poultry Sci. 31:65-68.

Goodman, B. L., and R. G. Jaap, 1960. Improving accuracy of herita-
bility estimates from diallel and triallel matings in poultry.
Eight-week body weight in closed-flock strains. Poultry Sci.
39:938-944.

96

Goodman, B. L., and F. V. Muir, 1965. The influence of comb and
feathering phenotypes on body weight and dressing percentage
in broilers. Poultry Sci. 44:644-648.

Graybill, F. A., F. Martin and G. Godfrey, 1956. Confidence inter-
vals for variance ratios specifying genetic heritability.
Biometrics 12:99-109.

Graybill, F. A., and W. H. Robertson, 1957. Calculating confidence
intervals for genetic heritability. Poultry Sci. 36:261-265.

Grossman, M., and G. A. E. Gall, 1968. Covariance analysis with un-
equal subclass numbers: component-estimation in quantitative
genetics. Biometrics 24:49-59.

Gyles, N. R., J. Kand and R. M. Smith, 1959. Heritabilities of
breast blisters and breast feathering in a White Rock broiler
strain. Poultry Sci. 38:1210-1211 (Abstract).

Hammond, K., and F. W. Nicholas, 1972. The sampling variance of the
correction coefficients estimated from two-fold nested and
offspring-parent regression analysis. Theoretical and Applied
Genetics 42:97-100.

Harris, G. C., Jr., M. Musbah, R. Sellers, J. Benson, V. Scogin and
P. W. Waldroup, 1980. Effects on feathering and incidence of
breast blisters in broilers of dietary sorbic acid and litter
treatment with potassium sorbate. Poultry Sci. 59:1564
(Abstract).

Hays, F. A., and R. Sanborn, 1942. Breeding Rhode Island Reds for
rapid feathering. Massachusetts Agri. Expt. Sta., Bull. 396.

Hays, F. A., 1951. Rate of chick feathering and growing chick weight
in Rhode Island Reds. Poultry Sci. 30:866-869.

Hays, F. A., 1952. Sex dimorphism in tail length of chicks at ten
days of age. Poultry Sci. 31:1-93 (Abstract).

Hays, F. A., and R. Sanborn, 1942. Breeding Rhode Island Reds for
rapid feathering. Massachusetts Agri. Exp. Sta., Bull. 396.

Hazel, L. N., 1943. The genetic basis for constructing selection
indexes. Genetics 28:476-490.

Hazel, L. N., M. L. Baker and C. F. Reinmiller, 1943. Genetic and
environmental correlations between the growth rates of pigs at
different ages. J. of Animal Sci. 2:118-128.

97

Hughes, B. 0., and I. J. H. Duncan, 1972. The influence of strain
and environmental factors upon feather pecking and cannibalism
in fowls. Br. Poultry Sci. 13:525-547.

Hulan, H. W., F. G. Proudfoot and K. B. McRae, 1980. The nutritional
value of tower and candle rape seed meals for turkey broilers
housed under different lighting conditions. Poultry Sci. 59:
lOO-lO9. .

Hurry, H. F., and A. W. Nordskog, 1953. A genetic analysis of chick
feathering and its influence on growth rate. Poultry Sci. 2:
18-25.

Hutt, F. B., 1949. Genetics of the Fowl. McGraw-Hill Book Co., Inc.,
New York.

Ideta, G., and P. B. Siegel, 1966.. Selection for body weight at
eight weeks of age. 3. Realized heritabilities of unselected
traits. Poultry Sci. 45:923-933.

Innin, M. R., E. P. Reineke and C. W. Turner, 1943. Effect of feed-
ing thyroactive iodocasin on growth, feathering and weights of
glands of young chicks. Poultry Sci. 22:374-380.

Jaap, R. G., and L. Morris, 1937. Genetical differences in eight
week weight and feathering. Poultry Sci. 16:44-48.

Jones, D. G., and F. B. Hutt, 1946. Multiple alleles affecting
feathering in the fowl. J. Heredity 37:197-205.

Jull, A., 1938. Poultry Husbandry. McGraw-Hill Book Co., Inc.,
New York.

King, S. C., and C. R. Henderson, 1954. Variance component analysis
in heritability studies. Poultry Sci. 33:147-154.

Kinney, T. B., and R. N. Shoffner, 1965. Heritability estimates and
genetic correlations among several traits in a meat-type poultry
population. Poultry Sci. 44:1020-1032.

Klein, T. W., J. C. DeFries and C. T. Finkbeiner, 1973. Heritability
and genetic correlation: Standard errors of estimates and
sample size. Behavior Genetics 3:355-364.

Kleinpeter, M. E., and J. P. Mixner, 1947. The effect of the quantity
and quality of light on thyroid activity of the baby chick.
Poultry Sci. 26:494-498.

98

Knox, C. W., and S. J. Marsden, 1944. The inheritance of some
quantitative characteristics in turkeys. J. Heredity 35:89-96.

Lasley, J. F., 1972. Genetics gf Livestock Improvement. (2nd Ed.).
Prentice-Hall, Inc., Englewood Cliffs, New Jersey.

 

Lerner, I. M., 1950. P0pu1ation Genetics and Animal Improvement.
Cambridge University Press, London.

Lerner, I. M., 1954. Genetic Homeostasis. Oliver and Boyd, Edinburgh.

Lerner, I. M., 1958. The Genetic Basis 9: Selection. John Wiley &
Sons, Inc., New York.

Lowe, P. C., S. P. Wilson and R. B. Harrington, 1965. Association of
some qualitative and quantitative traits in chickens. Poultry
Sci. 44:106-112.

Lush, J. L., 1945. Animal Breedinnglans. (3rd Ed.). Collegiate
Press, Ames, Iowa.

Lush, J. L., 1948. The Genetics gj_Populations. Mimeo book, Iowa
State College, Ames, Iowa.

Magee, W. T., 1965. Estimating response to selection. J. of Animal
Sci. 24:242-247.

Maloney, M. A., Jr., and J. C. Gilbreath, 1966. Selection for dif-
ferential body weight of twelve weeks of age in the domestic
fowl. Poultry Sci. 45:1102 (Abstract).

Martin, J. H., 1929. Rate of feather growth in Barred Plymouth Rock
chicks. Poultry Sci. 8:167-183.

McGibbon, W. H., 1977. A sex-linked mutation affecting rate of
feathering in chickens. Poultry Sci. 56:872-875.

McGibbon, W. H., and J. G. Halpin, 1946. Three alleles affecting
completeness of feathering in the chicken. Poultry Sci. 25:406.

McWard, G. R., K. E. Rinehart and J. M. Vande-populiere, 1972. Fac-
tors affecting broiler feathering. Poultry Digest 31(361):133.

Moultrie, F., C. D. Mueller and L. F. Payne, 1954. The effect of
light on feathering of broiler chicks. Poultry Sci. 33:784-790.

Moultrie, F., C. D. Mueller and L. F. Payne, 1955. Molting and
growth of individual feathers in turkeys exposed to 10 or 24
hours of daily light. Poultry Sci. 34:383-388.

99

Mueller, C. D., F. Moultrie, L. F. Payne, H. D. Smith and R. E. Clegg,
1951. The effect of light and temperature on molting in turkeys.
Poultry Sci. 30:829-838.

Nestor, K. E., M. G. McCartney and W. R. Harvey, 1967. Genetics of
growth and reproduction in the turkey. 1. Genetic and non-
genetic variation in body weight and measurements. Poultry Sci.
46:1374-1384. ,

Osborne, R.,.and W. S. B. Paterson, 1952. On the sampling variance
of heritability estimates derived from variance analysis. Proc.
Roy. Soc. Edin. 864:456-461.

Parker, J. E., 1943. Influence of thyroactive iodocasein on growth
of chicks. Proc. Soc. Exp. Biol. Med. 52:234-236.

Peeler, R. J., E. W. Glazener and W. L. Blow, 1955. The heritability
of broiler weight and age at sexual maturity and the genetic

and environmental correlations between these traits. Poultry
Sci. 34:420-426.

Platt, C. S., 1953. Maintaining winter egg production by the use of
dim red light. Poultry Sci. 32:143-145.

Plumart, P. E., and C. D. Mueller, 1954. Effect of sex-linked early
feathering on the plumage from 6 to 12 weeks of age. Poultry
Sci. 33:715-721.

Proudfoot, F. G., 1980. The effects of dietary protein levels,
ahemeral light and dark cycles, and intermittent photoperiods
on the performance of chicken broiler parent genotypes. Poultry
Sci. 59:1258-1267.

Quarles, C. L., and H. F. Kling, 1974. The effect of three lighting
regimes on broiler performance. Poultry Sci. 53:1435-1438.

Radi, M. H., and D. C. Warren, 1938. Studies on the physiology and
inheritance of feathering in the growing chicks. J. Agr. Res.
56:679-707.

Richter, F., 1954. Experiments to ascertain the causes of feather-
eating in the domestic fowl. 10th World's Poultry Congress,
Edinburgh. 258-262.

Ringer, R. K., 1965. Thyroids jg_"Avian Physiology." (2nd Ed.)
P. D. Sturkie, Ed. Cornell University Press, Ithaca, New York,
Chapter 19.

Robertson, A., 1959a. Experimental design in the evaluation of
genetic parameters. Biometrics 15:219-226.

100

Robertson, A., 1959b. The sampling variance of the genetic correla-
tion coefficient. Biometrics 15:469-485.

Saeki, Y., and T. Katsuragi, 1961. Effect of early and late feather-
ing gene on growth in New Hampshires, Leghorns and their cross-
breeds. Poultry Sci. 40:1612-1616.

Schaible, P. J., 1970. Poultry: Feeds and Nutrition. The AVI
Publishing Co., Inc., Westport, Connecticut.

Schaible, P. J., J. A. Davidson and S. L. Bandemer, 1947. Cannibal-
ism and feather picking in chicks as influenced by certain
changes in a specific ration. Poultry Sci. 26:651-656.

Schnetzler, E. E., 1936. Inheritance of rate of growth in Barred
Plymouth Rocks. Poultry Sci. 15:369-376.

Schultze, A. B., and C. W. Turner, 1945. The determination of the
rate of thyroxine secretion by certain domestic animals. Mo.
Agr. Exp. Sta. Res., Bul. 392.

Schumaier, G., P. C. Harrison and J. McGinnis, 1968. Effect of
colored fluorescent light on growth, cannibalism, and subsequent
egg production of single comb White Leghorn pullets. Poultry
Sci. 47:1599-1604.

Scott, M. L., and R. E. Reynolds, 1949. Studies on the nutrition of
pheasant chicks. Poultry Sci. 28:392-397.

Serebrovsky, H. S., 1922. Crossing over involving three sex-linked
genes in chickens. Amer. Nat. 56:571-572.

Shaklee, W. E., C. W. Knox and S. J. Marsden, 1952. Inheritance of
the sex difference of body weight in turkeys. Poultry Sci. 31:
882-825.

Shellenbarger, R. A., 1976. Cannibalism control of growing ring-
necked pheasants. M. S. thesis, Michigan State University,
East Lansing, Michigan.

Sheridan, A. K., and M. W. McDonald, 1963. The relationship between
feathering and body weight in broiler chickens. Poultry Sci.
42:1468-1470.

Siegel, P. B., 1962a. Selection for weight at eight weeks of age.
1. Short term response and heritabilities. Poultry Sci. 41:
954-962.

Siegel, P. B., 1962b. A double selection experiment for body weight
and breast angle at eight weeks of age in chickens. Genetics
47:1313-1319.

101

Siegel, P. B., 1963. Selection for body weight at eight weeks of
age. 2. Correlated responses of feathering, body weights, and
reproductive characteristics. Poultry Sci. 42:896-905.

Siegel, P. B., J. V. Craig and C. D. Mueller, 1957a. Heritabilities,
sex differences and phenotypic correlations for six feathering
characteristics. Poultry Sci. 36:621-628.

Siegel, P. B., C. D. Mueller and J. V. Craig, 1957b. Some phenotypic
differences among homozygous, heterozygous and hemizygous late
feathering chicks. Poultry Sci. 36:232-239.

Snedecor, G. W., 1956. Statistical Methods. (5th Ed.) The Iowa
State University Press, Ames, Iowa, U. S. A.

Skoglund, W. C., and D. H. Palmer, 1961. Light intensity studies
with broilers. Poultry Sci. 40:1458 (Abstract).

Somes, R. G., Jr., 1969. Delayed feathering, a third allele of the
K locus of the domestic f0wl. J. Hered. 60:281-286.

Somes, R. G., 1970. The influence of the rate of feathering allele
K" on various quantitative traits in chickens. Poultry Sci.
49:1251-1256.

Somes, R. G., 1975. Pleiotrophic effects of the sex-linked delayed
feathering gene, K", in the chicken. Poultry Sci. 54:208-216.

Sturkie, P. D., 1976. Avian Physiology. (3rd Ed.). Springer-Verlag,
New York, Inc., New York, U. S. A.

Supplee, W. C., 1966. Feather abnormality in poults fed a diet
deficient in vitamin E and selenium. Poultry Sci. 54:852-854.

Supplee, W. C., R. D. Creek, G. F. Combs and D. L. Blumberg, 1961.
The zinc requirements of poults receiving practical diets.
Poultry Sci. 40:171-176.

Tallis, G. M. 1959. Sampling errors of genetic correlations coef-
ficients calculated from analysis of variance and covariance.
Australian J. Stat. 1: 34-43.

Thomas, C. H., W. L. Blow, C. C. Cockerham and E. W. Glazener, 1958.
The heritability of body weight gain, feed consumption and
feed conversion in broilers. Poultry Sci. 37:862-869.

Turner, C. W., M. R. Irwin and E. P. Reineke, 1944. Effect of feed-
ing thyro active iodocasein to Barred Rock cockerels. Poultry
Sci. 23:242-246.

102

Warren, 0. C., 1925. Inheritance of rate of feathering in poultry.
J. Heredity 16:13-18.

Warren, D. C., and L. F. Payne, 1945. Influence of the early feather-
ing gene upon a chick's growth rate. Poultry Sci. 24:191-192.

Washburn, K. W., and P. B. Siegel, 1963. Influence of thiouracil on
chickens selected for high and low body weights. Poultry Sci.
42:161-169.

Waterhouse, H. N., and H. M. Scott, 1962. Effect of sex, feathering,
rate of growth and acetates on the chick's need for glycine.
Poultry Sci. 41:1957-1960.

Weaver, W. D., and P. B. Siegel, 1968. Photoperiodism as a factor
in feeding rhythms of broiler chickens. Poultry Sci. 47:1148-
1154.

Wells, R. G., 1972. The effect of varying stocking density on the
development and subsequent laying performance of floor-reared
pullets. Br. Poultry Sci. 13:13-25.

Wheeler, R. S., E. Hoffman and C. L. Graham, 1948. The value of
thyro protein in starting, growing and laying rations. I.
Growth, feathering and feed consumption of Rhode Island Red
broilers. Poultry Sci. 27:103-111.

Wilson, W. 0., and H. Abplanalp, 1956. Intermittent light stimuli
in egg production of chickens. Poultry Sci. 35:532-538.

Winsor, C. P., and G. L. Clark, 1940. A statistical study of vari-
ation in the catch of plankton nets. J. Marine Res. 3:1-34.

Wright, S., 1940. Breeding structure of populations in relation to
Speciation. Am. Naturalist 74:232-248.

Wyatt, A. J., 1954. Genetic variation and covariance in egg produc-
tion and other economic traits in chicken. Poultry Sci. 33:
1266-1274.

Yoshida, S., and W. M. Collins, 1967. Individual selection for 4-
week body weight in Japanese quail in relation to sex dimorphism.
Poultry Sci. 46:134 (Abstract).

APPENDIX

TABLES

106

Table A4. Pheasant breeder ration fed to birds from time of
stimulatory lighting to end of egg production period.

 

 

 

Ingredient Percent
Corn 1 53.25
Soybean meal, 44% 15.00
Oats 7.50
Wheat middling 7.50
Alfalfa, 17% 3.00
Fish meal, 60% 2.50
Meat and bone meal, 50% 3.00
Whey, dried 2.00
Salt 0.25
Dicalcium phosphate 1.25
Limestone 3.75
Premixa 0.75
Calculated Analysis

Crude protein 18.00
Fat 3.44
Fiber 4.65
Calcium 2.40
Phosphorus, available 0.68
M.E. Cal/1b. 1225.00

 

aPremix 5004, available from Dawes.

107

 

 

 

 

 

 

 

Table A5. The standard model of the nested analysis for the deter-
mination of variance components of feathering scores and
other traits among sires, dams within sires and full-sib
progeny.

Source of

Variation d.f. E4(MS)

Generations (G) g-l

Sires (5m; s-g oh + k2 06 + k, a;

Dams (D)/S/G d-s ofi + k1 06

Full-sibs (W) N-d ofi

Total N-l
1 ani'k

k1 = d-s “"1 1(—1‘——.—.J——)l
1 J n13
2 z
1 k n 13k (. k "2.. )
= .. - ' 'I k
k2 5-9 [1: §( n1J.. ) X1 in: J J
in.
l j ij.
k3 = s-g [ N - g ni.. ]

108

 

 

 

 

 

 

 

 

Table A5, The variance and covariance components of feathering

scores and associated traits from selection for slow

and rapid feathering for a period of three generations

by lines and sex.

Feathering,
Source of Slow Control Rapig__
Var. Covar. Male Female Male Female Male Female
8§ 22.5 5.1 --- --- - 3.1 - 4.3
65 2.7 11.0 64.7 1.8 26.4 23.9
Bfi 92.4 122.9 97.8 115.1 100.9 95.5
One-Day Body Weight
2
65 - .6 .3 --- --- - .6 - .8
65 2.9 1.7 4.0 4.6 2.6 2.9
afi 1.4 1.5 2.4 1.7 1.7 1.1
Cov S .2 .2 --- --- 1.5 .S
Cov D 1.9 .4 5.9 3.8 .7 - .6
Cov W - .7 .4 .2 - .2 .2 - .2
Four-Week Body Weight

652 90. 3 165.4 --- --- 424. 5 21.3
65 101.9 52.1 381.9 162.9 - 73.3 270.5
Bfi 1378.6 767.0 949.8 692.4 1231.8 810.4
Cov S 60.0 8.1 --- --- 89.5 ..8
Cov D - 19.4 6.0 119.5 - .8 - 12.6 47.3
Cov W 159.6 154.0 131.4 136.5 91.8 100.0

 

109

Table A6. (cont'd)

 

Eight-Week Body_Weight

 

 

 

 

 

Source of Slow_ ContrBTl Rapid
Var. Covar. Male Female Male Female Male Female
6; 71.5 563.1 --- --- *720.5 -400.9
65 414.9 -126.2 1717.9 530.7 246.2 2334.5
6; 5419.0 4139.1 4559.3 3736.4 7068.3 3886.6
Cov s 64.6 - 2.5 --- --- 86.7 9.9
Cov 0 11.6 9.0 276.8 80.8 96.4 48.5
Cov w 359.8 290.1 314.8 173.8 118.0 266.5
Twelve-Week Body Weight
6; 501.6 656.8 --- --- 168.7 428.2
GB 1601.0 1032.9 1732.2 -4849.9 967.6 1083.5
6; 9786.8 7855.8 8546.5 25972.0 12548.6 8678.6
Cov s - 12.8 - 51.6 --- --- - 53.7 4.8
Cov 0 - 9.8 73.2 294.2 60.4 190.0 - 26.4
1 258.8 383.5 267.5 104.2 270.1

Cov W 471.

 

110

Table A6. (cont'd)

 

Four-Week Tail Length

 

 

 

Source of Slgw Control Rapid

Var. Covar. Male Female Male Female Male Female
6g .02 .30 --- --- , .63 .14
65 .81 .25 .92 .53 .06 .19
6; 1.52 2.55 1.91 3.07 3.04 3.63
Cov S 2.29 1.01 --- --- 2.51 - .10
Cov D - .75 .27 5.39 .89 .68 2.35

Cov W 5.51 7.53 4.91 7.89 5.92 8.44

 

111

.OOO.. nmuum.mm mg» Ho OOHOOH.OON 93H ecu mo OmmONm>O ONO OOO.O> ONOOOO :Oms ..<O

 

 

 

 

 

O.N NNOO. NOOO 0.0 ... 0.00 O OH

O.O OHON. OOON NON. N.. O OO. 2

O.H ONON. 0.0.. NOO. H.O. 0.00. O OOO O.OOO
O.O .OOO. ONON ONO. 0.0 ..ON. 2

N.O OOOO. 0.0.. .HO.N O.N 0.0N. O OH.O

N.N ONOO. .OOO., OO.H .m.m O.OO. 2

..O NNOON OONN NOO N.. ..O.. O NOOOOHO

O.. NHOO OOOH OOO H.N O.NO z .OHHOOO
N.H ONOO. OOOO OO.. 0.0. 0.0N. O OOO

O.H O.OOHT OOHO OHON 0.0., 0.00N z

O.N OOON OO.H NON O.. O.NN. O OH

O.. NONO O.HO ONO. H.. H.NO z

N.O ONON. O.OO NHO 0.0 N .N. O OOO 2O.O
O.H H.OO. OOOO OOO. 0.0 0.00 z

..N OHO.N ON.O. OONN O.H. O.HON N OH.O

N.O HOOON HOON OONN N.O. O.HHO z
OHOOOO 2O H3-N. 3O H3-O 3O ¥:-H 3O NOO-. HOHOOO HOOOOHO HHOOHOO OO.O

..O. H3-H OO.HOOHOO.
HH.OH.

 

O.xmm OOO OOOH. an mco.uecm:mm
mmcsu HO OOHNOO O com OOHNOOHOOH O.OO; OOO 3O.O LO. =o.auo.mm seem OHHOLH OOHOHOOOOO .
OOO OONOOO OOHNOOHOOH com OO.O-..OH =.:u.z OOO OEOO .OOL.O :mmzpma mm=.O> «.OOOO sews OOH .N< O.OO»

112

Table A8. The calculated values of the degrees of freedom of sires,
dams and progeny and the coefficient of variance components
by lines and sex over three generations of selection for
slow and rapid feathering.a

 

 

Line Sex d.f.s d.f.D d.f.w K1 K2 K3
Slow M 17 47 246 4.08 5.15 12.99
Slow F 18 44 243 4.06 5.38 12.35
Control M -- 29 64 2.87 -- --
Control F -- 29 73 3.10 -- --
Rapid M 15 30 123 2.75 4.00 7.84
Rapid F 13 35 164 3.29 5.17 10.97

 

aK values are averages of the two replicates and the degrees of
freedom are the summation of the two replicates of the selected lines.

113

Table A9. The general model of analysis of variance and covariance
components of feathering and other traits.

 

 

 

 

Source of Sum of Products
Variation d.f. Dams XX Dam-Progeny XY Progeny YY
Between Sires S-l
DamS/Sires D-S SSD(XX) SCPD<XY) SSD‘YY)
Total D-l
where 2
2 xi.

 

$5 = x.. -

n1 = number of dams mated to the ith sire.

SSrsyy) are obtained in the same manner using record of each
dam' s p geny,Y

 

 

X1. Yi.
SCPD(XY) = Z X Xij Yij ' g “i
SCP
Then the CovD(xy) = DD(X;) MCP
MCP

Regression = b ="§§5z;;;'

hzop = 2b.

114

 

 

 

 

 

 

 

Table A10. The mean squares and mean cross products of dams and
progeny of two generations of selection for slow and
rapid feathering and body weights.a

Tetal number of:

Line MCPD(Xy) M5PD(XX) MSPD(YY)_ Dams Sires

. 'Feathering Scores

Slow 66.20 1236.25 181.29 39 13

Control -11.27 1439.30 284.61 18 --

Rapid 15.17 50.35 148.21 22 ll

8-Week Body Weight

Slow 10807 32697 26453 39 13

Control 3399 37648 15603 18 --

Rapid 1262 13909 18536 22 11

12-Week Body Weight

Slow 37.754 103465 103555 39 13

Control - 711 139410 62303 18 --

Rapid 5,527 16496 .70991 22 ll

 

aMS and MCP are averages of two replicates in the two selected

lines.

115

Table All. Analysis of variance of egg production by lines,
generations and interaction. .

 

 

Source of Degrees of Sum of Mean
Variation Freedom Squares Squares F-statistic
Lines, L 2 464.9233 232.4616 3.00
Generations, G 2 964.3893 482.1947 6.22**
L x G 4 1252.6031 313.1500 4.04*
Experimental

Error 6 463.5143 77.5230

 

**Significant at F<.05.
*Significant at F<.10.

116

Table A12. Analysis of variance of feathering response to selection
by lines, generations and sex.

 

 

Source of Degrees of Sum of Mean
Variation Freedom Squares Squares F-statistic
Lines 2 14.05 7.02 7.15*
Generations 1 1404.01 1404.01 l428.28**
Sex 1 8.33 ' 8.33 8.48*
Experimental

Error 7 6.88 .78

 

**Significant at F<.01.
*Significant at F<.05.

117

Table A13. The transmitted file for the GENSTAT analysis of variance
and covariance of unbalanced data.

 

1. PW, PN AND USER ID. Pass word, problem number and the
user identification.

2. *JOBCARD*, CM 70000, JC 499, The control card specifying a
R02. central memory of 70000 octal
words with a job cost of'$4.99

running in rate group 2.

3. ATTACH, LFN, DATA-FILE- This is to attach the data file
CONTAINING-DATA. to the program as a local file.

4. HAL, L*UNSUP, GENSTAT, USER 1= HAL is a system library. The pro-
TAPE 11, 12=LFN gram with that system, which
attaches GENSTAT, is designed as
L‘UIVSUP. USER 1 is the command
to attach the MACRO Library as
Tape 11 and Ié=data file attachedL
in line 3 above.

5. *E05 This statement ends the control
cards section.

6. "REFERENCE/NUNN=100, NID= Start of the GENSTAT job, and

1000" ‘ setting of’storage parameters.

7. "FILE" LIB=11 This is to define the Library in
user-file LIB on unit number
Tape 11.

8. "FETCH" LIB $ HIERANOVA This is to retrieve the macros

from the library of'macros as a
subfile name HIERANOVA. $ to in-
dicate the end of’the directive.

9. "GET" HIERANOVA'S Macro subfile name and the pointer
PHIERANO file name to inform the program
about all macros.

10.
11.

ll RUN"
"UNIT"

$96

To execute the proceeding block.

Number of’observations in one par-
ticular set of’data.

 

s‘t“).‘l\l l.l. o

118

 

Table A13. (cont'd.)

12. "INPUT" 2 Tells GENSTAT to read the file
associated'with the I? parameter
of‘the GENSTAT control cards.

13. "SCALAR" NL=4 The scalar statements are to iden-
tify the levels and numbers of

14. NL=2 variables of the sorted data as
matrices and scalars (structures)

15. NL(1)=4 referenced'by their variable
names.

16. NL(2)=5

17. NL(3)=3

18. NAN(1,2,3,4)= The NAN standb for the number of
1 analysis where 1 and 4 to perform

analysis of’variance and 2 and 3
for analysis of’variance. The
option 1 is to perform andflprint
the analysis.

19. "SET/F" FORMAT=F, The format statement tells GENSTAT
2X,(2)3,33X, how to read the data for analysis
3.1,18X,6.3,/ in a particular set of’data.

20. I'USE" HIERANOVA $ This is a macro subfile name.

21. "INPUT" 1 Tells GENSTAT to read the local

file input and to find any more
GENSTAT commands.

22. "RUN" Initiates the execution of’the
previous GENSTAT lines.

23. "CLOSE" Blanks GEWSYMJ’memory locations
used in previous GENSTAT blocks.

24. "STOP" Finishes the whole GENSTAT pro-

gram.