April 1999
Mi

Repeatabilities, Heritabilities and

Phenotypic, Genetic, and

Environmental Correlations for

Production, Reproduction, and
Somatic Cells in the

University of Florida Jersey Herd

Rafael M. Roman, Charles J. Wilcox,
H. Herbert Head and H. H. (Jack) Van Horn

UNIVERSITY OF
FLORIDA
Agricultural Experiment Station
Institute of Food and Agricultural Sciences

Bulletin (Tech.) 907
1 1- i1I

SDOC
100
F636b
907

-

Rafael M. Roman is a Professor (Geneticist), Faculty of Agronomia, University of Zulia, Maracaibo, Venezuela; Charles J. Wilcox is
Professor Emeritus (Geneticist), H. Herbert Head is a Professor (Physiologist), and H. H. (Jack) Van Horn is a Professor (Nutritionist),
Department of Dairy and Poultry Sciences, University of Florida, Gainesville, FL 32611-0920, U.S.A. This research was conducted
under project RRF-251, FLA-DAS-03197.

Repeatabilities, Heritabilities and Phenotypic, Genetic, and Environmental
Correlations for Production, Reproduction, and Somatic Cells
in the University of Florida Jersey Herd

Rafael M. Romana
Charles J. Wilcoxb
H. Herbert Headb
H.H. (Jack) Van Hornb

Department of Dairy and Poultry Sciences U NVERSI T
University of Florida OF FLORIDA
Gainesville, FL 32611 JUN 30 1999

professor (Geneticist) MARSTON SCIENCE
Faculty of Agronomia LIBR AR
University of Zulia
Maracaibo, Venezuela

bProfessor Emeritus (Geneticist)
Professor (Physiologist), and
Professor (Nutritionist), Department
of Dairy and Poultry Sciences
University of Florida,
Gainesville, FL 32611-0920. U.S.A.

UNIVERSITY OF FLORIDA LUBRARIE

TABLE OF CONTENTS

Page

Abstract ........................................................ ............ iv

Introduction .............................................. .. ............... 1

Objectives ............................................ ....................... 2

Review of literature .......................................................... 3
Genetic aspects of milk yield and composition ................... .. .... ... 3
Heritabilities of somatic cell counts and relationships with milk yield and composition ..... 5
Heritabilities of reproductive traits and relationships with milk yield and composition ...... 7

Materials and Methods ..................................................... 8
The UF Jersey herd ..................... ... .......................... ... ... 8
Herd management ...................................... ...................8
Description of data ....................................... .................... 9
Statistical analyses ............................................ .............. 11
Estimation of genetic parameters .............................. ........... 12
Computing strategies .................................. ...................... 13

Results and Discussion ........................................................ 13
Repeatabilities and heritabilities ...................................... ......... 13
Phenotypic correlations ........................................................ 18
Genetic correlations ........................................................ 22
Environmental correlations .................................................... 29

Conclusions ...............................................................31

References ................................ ................................32

ABSTRACT

Data collected on the University of Florida Agricultural Experiment Station Dairy

Research Unit Jersey herd were analyzed. Responses were 32 measures of lactation milk

composition and yield, 7 measures of reproductive performance and 6 measures of somatic cell

counts for 935 lactations of 374 cows for 1969 through 1987. Genetic correlations between

lactation yields were positive and high, 0.66 to 0.97, and between percentage composition

positive and high, 0.44 to 0.91. Milk yield was negatively correlated with fat, solids-not-fat, total

solids, protein percentage, 0.21 to 0.56, but positively correlated with solids-not-fat to fat and

protein to fat, 0.38 and 0.56. Heritabilities of yields and ratios ranged from 0.10 to 0.30, and

percentages, 0.19 to 0.45. Heritabilities of somatic cell counts were 0.14 to 0.17. Heritabilities

of reproductive performance were 0 to 0.05. Repeatabilities of milk and constituent yields were

0.30 to 0.33 and percentages, 0.30 to 0.61. Phenotypic, genetic, and environmental correlations

between all measures of performance (>3,300 estimates) indicated that selection for increased

milk yield would result in increased milk yield as expected, with very slight decreases in milk

composition and reproductive performance and very slight increases in somatic cells. Results

should be useful in the design of dairy cattle selection projects which would maximize economic

returns to dairymen.

INTRODUCTION

In dairy cattle populations we often are interested in problems such as estimation of

genetic parameters, breeding values, response to selection and correlated changes when

selection has occurred for many years. Considerable research has been completed to obtain

these estimates by using data from commercial herds or from planned selection experiments.

The latter studies, even when they represent smaller populations, have the advantage of

controlling extraneous sources of variation that are more difficult to handle with field data.

Thus, it would be expected that planned selection experiments would give more reliable

results about response to selection and the correlated changes brought about by selecting for

a particular trait.

In a cooperative effort without precedent in the world, several selection experiments

were planned during the late 1960s and early 1970s in the United States, as part of the

regional project S-49, Genetic Methods of Improving Dairy Cattle for the South. Even

though Holsteins were involved in other USA selection projects, the Agricultural Experiment

Stations of Florida, Georgia, North Carolina and Tennessee utilized the Jersey breed.

Leaders at each of their respective experiment stations were C. J. Wilcox, J. C. Johnson Jr.,

J. E. Legates and D. O. Richardson. It is interesting to note that this project had as its

original objective, as Southern Regional Project S-3, to determine whether special strains or

strain crosses among cattle that were adapted to the hot and humid conditions that

characterized much of the tropics and subtropics were required to increase low levels of

production. However, results indicated that genetic improvement should be devoted to

selection within the existing temperate, European breeds. Moreover, an extensive study of

genetic interrelationships between milk composition and yield involving data from 19 states

suggested that selection emphasis should be placed on milk yield, while maintaining

acceptable and legal standards of milk composition. On the other hand, that research showed

that selection for milk composition could be accomplished, though with some reduction in

gain for increased milk yield (Gaunt et al., 1968; Wilcox et al.,1968). Furthermore, a study

involving straightbred Jerseys, Guernseys, and Holsteins, and their crosses, demonstrated

that Jerseys were superior to Holsteins in total and productive life span as well as number of

parturitions (Wilcox et al., 1966). These results, in addition to smaller size, milk yield

ability and pigmentation, pointed to the potential of Jerseys for subtropical conditions not

only in the southern area of the USA but also for developing dairy areas in other countries

that also had tropical and subtropical climates.

OBJECTIVES

Quantitative traits such as milk yield and those related to reproductive performance

and many other economically important traits are the result of the action and interaction of

genes (an unknown number) on several (also an unknown number) loci. Genes exist in

blocks. In addition, some genes simultaneously affect more than one trait because they

control metabolic pathways that are common for the phenotypic expression of several traits.

Objectives of this research were to

1. estimate heritabilities and repeatabilities of milk and constituent production,

reproductive performance and somatic cell counts under designed experimental conditions.

2. estimate genetic, environmental and phenotypic correlations between these

variables.

REVIEW OF LITERATURE

Genetic Aspects of Milk Yield and Composition. The most extensive work on

investigating heritabilities and repeatabilities in dairy cattle populations was performed with

field records and experimental herds in the past for the principal dairy breeds in USA by

using intraclass correlation techniques and least squares analysis of variance procedures.

Wilcox et al. (1971) analyzed 28,395 records representing 2,948 sires in 325 herds

for Ayrshire, Guernsey, Holstein, Jersey and Brown Swiss breeds, for six yield and four

percentage variables, along with solids-not-fat to fat (SNF/F), and protein to fat (P/F) ratios.

For Jerseys, heritability estimates from intraclass correlation techniques for yields ranged

between .37 to .51, for percentages, estimates were in the range .66 to .91, and for SNF/F

and P/F they were .70 and .52. Repeatability estimates were in the range .47 to .54 for

yields, .61 to .76 for percentages, and, for ratios, .68 for SNF/F and .56 for P/F.

Working with the same breeds and a smaller data set, Gacula et al. (1968), studied

the same variables as Wilcox et al. (1971). They reported, for Jerseys, heritabilities in the

range of .10 to .46 for yields and .35 to .61 for percentages. Benya et al. (1976),

investigating the UF Experiment Station Jersey herd, reported heritabilities for 15 milk yield

and composition traits in first lactation cows in the range .11 to .37 for yields, .43 to .68 for

percentages, and .67 for ratios SNF/F and .84 for P/F. Later, Moya et al. (1985) studied

the same 15 responses, using all records, and found heritability estimates in the range .31 to

.48 for yields, .11 to .47 for percentages, .31 for SNF/F and .21 for P/F.

The genetic relationships between milk yield and composition were estimated for five

dairy breeds in USA by Wilcox et al. (1971). In Table 1 are their estimates for Jerseys from

paternal half sib analysis. These relationships were estimated later for Florida conditions

using monthly measurements ( Sharma et al., 1983), in first parity cows (Benya et al.,

Table 1. Genetic correlations of milk composition and yield for Jerseys.a
Yields Percentages Ratios

Milk Fat SNF TS Protein LM Fat SNF TS Protein SNF/F
Yields
Fat .66
SNF .96 .72
TS .92 .87 .97
Protein .90 .82 .95 .96
LM .95 .68 .97 .93 .90
Percentages
Fat -.56 .23 -.45 -.24 -.33 -.46
SNF -.21 .15 .06 .10 -.01 -.01 .44
TS -.52 .21 -.31 -.15 -.29 -.35 .91 .76
Protein -.55 -.05 -.42 -.31 -.15 -.52 .69 .76 .78
Ratios
SNF/F .56 -.19 .53 .31 .41 .51 -.94 -.14 -.73 -.55
Protein/Fat .38 -.23 .33 .15 .36 .32 -.79 -.27 -.67 -.12 .87
SNF = Solids-not-fat, TS = Total solids, LM = Lactose mineral.
aWilcox et al. (1971).

1976), and using all records available (Moya et al., 1985). From these reports the

following conclusions can be made: (a) milk yield is strongly and positively associated from

a genetic standpoint with the constituent yields; (b) milk and composition yields are

negatively related to composition percentages; and (c) percentage traits among themselves, in

general, have positive genetic associations. These results agreed with others investigating the

relationships between milk yield and fat (Albuquerque et al., 1996; Meyer, 1984; Van Vleck

and Dong, 1988), and those working with milk, fat, and protein yield; and fat and protein

percent using other estimation methods (Ahlborn and Dempfle, 1992; Chauhan and Hayes,

1991; Meyer, 1985).

Heritabilities of Somatic Cells Counts and Relationships with Milk Yield and

Composition. Several response variables have been proposed as indirect indicators of health

status of the mammary gland. The most extensively investigated are somatic cell counts

(SCC) and somatic cell scores (SCS). A selection experiment with selection and control

groups showed that most of the per cow differences in expenses between groups were

because of mastitis, 82% for first lactation and 56% across all lactations (Jones et al., 1994).

Ali and Shook (1980) recommended expressing SCC on a logarithmic scale to obtain more

nearly a normal distribution of errors and homogeneous error variances.

Studies on the heritability of SC indicates that the additive genetic variance is quite

low. Monardes et al. (1983) found heritability estimates of .06 for the average SCC for all

lactations and .12 for the same variable on a logarithmic scale. Coffey et al. (1985),

working with the average of the log of SCC, found heritability estimates of .18 for all

records, but they were .09, .10 and .29 for first, second, and third and later lactations.

Emanuelson et al. (1988) with three data sets found estimates in the range .05 to .11. Schutz

et al. (1990) studied both SCC and SCS for first, second and third lactations; their estimates

for SCC were .05 .07 and .08, whereas those for SCS were .10, .15 and .13. Welper and

Freeman (1992), using restricted maximum likelihood procedures (REML) with an

expectation maximization algorithm, reported a heritability of .16 for SCS. Detilleux et al.

(1995) working with an animal model found .07, .50, 0 and .12 for first, second, third and

fourth lactation. Reents et al. (1995) used REML to estimate heritabilities for SCS of .09,

.09, and .11, for first, second, and third lactations.

Some research has been devoted to estimating genetic and phenotypic

interrelationships between milk yield and composition traits and measures of somatic cells of

milk. In general, there appears that there is a small positive correlation between milk yields

and SC; most reports are in the range .01 to .41 (Banos and Shook, 1990; Boettcher et al.,

1992; Emanuelson et al., 1988; Schutz et al., 1990; Welper and Freeman, 1992).

Nevertheless, the reverse also has been observed because genetic correlations in the range

-.21 to -.06 were reported (Schutz et al., 1990) in second parity cows. This also was

observed in second and later parities (Banos and Shook, 1990). The phenotypic correlations,

on the other hand, usually have been reported as negative within the range -.17 to -.03

(Banos and Shook, 1990; Boettcher et al., 1992; Emanuelson et al., 1988; Schutz et al.,

1990; Welper and Freeman, 1992).

Genetic relationships between fat yield and SC most often have been found as positive

in the range .11 to .20 ( Boettcher et al., 1992; Schutz et al., 1990; Welper and Freeman,

1992). As with milk yield, some negative estimates also have been observed, in the range

-.31 to -.14 (Schutz et al., 1990). Phenotypic correlations were negative but close to those

reported for milk yield. Schutz et al. (1990) found positive genetic correlations between

protein yield and SC in the range .06 to .37. Phenotypic counterparts were negative but

quite low. Welper and Freeman (1992) reported a genetic correlation close to zero for

lactose percentage and SCS. Because of the small negative phenotypic correlation they

suggested that lactose would not be a very reliable indicator of mastitis if used alone.

Despite the combination of the low heritabilities and low genetic correlations, breeders

continue to use measures of SC on milk in selection programs of USA and Canada (Schutz et

al., 1995; Shook and Schutz, 1994; Zhang et al., 1994).

Heritabilities of Reproductive Traits and Relationships with Milk Yield and

Composition. For several years researchers have been concerned with the genetic aspects of

reproductive traits and the genetic and phenotypic relationships between these traits. This, in

turn, translates into two different problems, that of adjusting genetic evaluations for

reproductive effects and most often investigating a possible antagonistic effect of milk yield

and reproductive performance. Most researchers agreed that heritabilities of reproductive

responses were low and generally below .10 (Campos et al., 1994; Silva et al., 1992). This

has been interpreted to mean that natural selection has depleted the additive genetic variance

in these traits, and little further improvement through selection can be expected (Oltenacu et

al., 1979). However, Darwash et al. (1997) studied the interval from calving to

commencement of luteal activity post-partum, finding heritabilities of .28 for untransformed

data, .21 for log transformed data and .13 for reciprocals.

Campos et al. (1994) found that the correlations between breeding values for yields

and reproduction were low and generally antagonistic. Raheja et al. (1989) reported that the

genetic and phenotypic correlations between fertility and production in the subsequent

lactation were very low or close to zero.

Objectives of the present research were to estimate the genetic parameters and

interrelationships between milk yield and composition traits, some measures of reproductive

performance and somatic cells in an experimental Jersey herd.

MATERIALS AND METHODS

The UF Jersey Herd. The Jersey herd of the University of Florida Experiment

Station was founded in 1901. It was maintained on campus in Gainesville until 1949 when

the cows were moved to their final location at Hague, Florida, approximately 11 miles from

the downtown campus. Gainesville is located at 290 39.6' N, 83* 49.6' W and by

atmospheric conditions is classified as a subtropical region, characterized by extreme

variations in weather, especially temperature (Moya et al., 1985; Simerl, 1982). Data for

this study covered the years 1969 through 1987.

Herd Management. Even though many changes occurred at the experiment station

during the long history of the herd, general management practices were similar to those of

most commercial dairies in Florida (Simerl, 1982). Every cow received a reasonable

opportunity to complete at least one lactation. Milking was twice a day, and dry periods

preceding the next lactation were scheduled to range between 50 and 60 d.

Description of Data. For purposes of this research two separate files were created:

(a) monthly summaries of milk yield and composition, and (b) reproductive and milk records

from monthly DHI summaries. The file with the recordings of milk yield and composition

included 10,076 records. Each record had, in addition to monthly summaries of yields,

results from the chemical analyses from sampling of milk composition, and some measure of

somatic cell count (SCC). However, recordings of milk yield were on a daily basis as a

routine practice of the experiment station. Records from 1984 were taken from electronically

kept data files of the herd at the Northeast Regional Data Center (NERDC). From these files

more than 50,000 daily measurements were available. After appropriate editing and

screening, there were 935 records of 374 cows by 67 sires.

Monthly analyses of milk composition were as follows: fat was determined by

Babcock method, protein from formol titration, chloride from Mohr direct titration method,

solids-not-fat by equations based on specific gravity determined by the Watson pattern

lactometer and fat %. Lactose-mineral percentage was estimated as the difference between

solids-not-fat and protein percentages (Sharma et al., 1983). Recordings on Feulgen DNA

reflectance were performed until January, 1978, when Somatic Cell Counts (SCC) were

performed. In October 1984 recordings were based on Somatic Cell Scores (SCS). In order

to have a unique response as indicator of SCC, DNA reflectance records were transformed

into SCC by using the relationship determined between reflectance and absorption,

Absorption =log[l/reflection] (Paape et al., 1965; Madsen, 1969). Previous research on

SCC showed a very high (R=.95) relationship between Feulgen DNA and the cell count

number (Thatcher et al., 1971). The equation published by Madsen (1969) was used to

predict the cell number with a correlation coefficient r =.98. SCS were analyzed in the

original form as recorded, by using a transformation of the counts to natural logarithms and

also after being weighted by the amount of monthly milk yield.

Occurrences of reproductive events were recorded daily as they occurred. For

monitoring reproductive performance, variables such as days from parturition to first

observed estrus and parturition to first breeding, and gestation length were included. Criteria

for concatenation of the files into a single data set were the cow identification number along

with her parity number. From this, a subset including only first lactations was created. In

Table 2 are the definitions of the variables included in this research.

Table 2. Response variables measured at the UF Experiment Station Jersey Herd.
Code Definition
1 MY Milk yield truncated at 305d from DRU records
2 FY Fat yield truncated at 305d from DRU records
3 PY Protein yield truncated at 305d from DRU records
4 LMY Lactose-mineral yield truncated at 305d from DRU records
5 SNFY Solids-not-fat yield truncated at 305d from DRU records
6 TSY Total solids yield truncated at 305d from DRU records
7 F% Fat % for truncated at 305d DRU records
8 P% Protein % for truncated at 305d DRU records
9 LM% Lactose-mineral % for truncated at 305d DRU records
10 SNF% Solids-not-fat % truncated at 305d from DRU records
11 TS% Total solids % truncated at 305d from DRU records
12 P/F Protein to fat ratio for truncated at 305d from DRU records
13 SNF/F Solids-not-fat to fat ratio truncated at 305d from DRU records
14 CL% Chloride % truncated at 305d from DRU records
15 AC% Acidity % truncated at 305d from DRU records
16 MY Milk yield for the complete lactation from DRU records
17 FY Fat yield for the complete lactation from DRU records
18 PY Protein yield for the complete lactation from DRU records
19 LMY Lactose-mineral yield for the complete lactation from DRU records
20 SNFY Solids-not-fat yield for the complete lactation from DRU records
21 TSY Total solids yield for the complete lactation from DRU records
22 F% Fat % for the complete lactation from DRU records
23 P% Protein % for the complete lactation from DRU records

Statistical Analyses. Standard procedures were used for the statistical analysis of

mixed models. Two mathematical models were used. In Model I, effects considered to be

fixed included the third degree polynomial regression for year of calving, and the discrete

effects of month of calving, the second degree polynomial regression of lactation length and

the third degree polynomial regression of age at calving. The random part of the model

included sire effects and cows nested within sire, both effects assumed to be normally

distributed with means 0 and variances Iso2,, and Io2 ,, representing the sire and cow

variances. In matrix form the model can be represented as:

Table 2. (continued)
Code
24 LM%
25 SNF%
26 TS%
27 P/F
28 SNF/F
29 CL%
30 AC%
31 NSERV
32 GEST
33 PAFOE
34 PAFBR
35 DMITO
36 DFATO
37 MSCS3
38 NSCST
39 MSCC3
40 MSCCT
41 MLOG3
42 MLOGT
43 WSCS
44 WSCC

Definition
Lactose-mineral % for the complete lactation from DRU records
Solids-not-fat % for the complete lactation from DRU records
Total solids % for the complete lactation from DRU records
Protein to fat ratio for the complete lactation from DRU records
Solids-not-fat to fat ratio for the complete lactation from DRU records
Chloride % for the complete lactation from DRU records
Acidity % for the complete lactation from DRU records
Number of services
Gestation length
Days from parturition to first observed estrus
Days from parturition to first artificial insemination
Total milk yield from DHI records
Total fat yield from DHI records
Average somatic cell scores (305d)
Average somatic cell scores (total)
Average somatic count (305d)
Average somatic count (total)
Natural logarithm of somatic cell count (305d)
Natural logarithm of somatic cell count (total)
Weighted average of somatic cell score
Weighted average of somatic cell count

y=Xb+ Zu+e

where y is the vector of observations, X and Z are incidence matrices assigning elements of

y to fixed and random effects, b and u are unknown vectors representing the least squares

solutions for the fixed effects and random effects, e is an unobservable random vector with

mean zero and variance Inoe. Model II differed from model I since it did not include the

sire random effects. Cows were nested within genetic group, and year effects were fitted as

a discrete variable along with the interaction of genetic group by year. Also included were

month of parturition, the second degree quadratic polynomial effect of lactation length and

the third order (cubic) regression for age at parturition. These models were developed after

a series of preliminary analyses, along with knowledge of relevant published research.

Estimation of Genetic Parameters. Estimates of heritabilities were based on paternal

half sib analyses, which doubtless is the most widely used method for estimating these

parameters in dairy cattle populations and many other species. The heritability is the ratio of

the additive genetic variance to the phenotypic variance. In half sib analysis an estimate of

the additive genetic variance can be obtained from 4a,2, under the assumption that the

observed sire variance component estimates 1/4 of the additive genetic variance (Falconer,

1990). Genetic correlations were estimated by dividing the appropriate sire covariance

estimate for two traits by the geometric mean of the two between sire variances (Harvey,

1990). Repeatabilities were estimated as the ratio of the between cow variance component to

the total variance. It therefore estimates the resemblance between records on the same

animal due to genetic and permanent differences (Falconer, 1990).

Computing Strategies. For estimation of variance components the general purpose

mixed model maximum likelihood computer program was used (LSMLMW) (Harvey, 1990),

with theoretical background given by Harvey (1960). This program was chosen because of

its capability to estimate (co)variance components as well as direct estimates of genetic

parameters with their corresponding standard errors. All computations were performed at

the computing laboratory of the Dairy and Poultry Sciences Department of the UF, on a

personal computer provided with 32 Mb RAM and 1.2 Gb hard drive.

RESULTS AND DISCUSSION

Repeatabilities and Heritabilities. Heritability estimates, standard errors and sire, cow,

and residual variance components for 305-d and total records for yields and milk composition

traits are in Table 3. Estimates for 305-d records and those from lactation totals in general

agreed. For yield traits, heritabilities were in the range .10 to .23 with standard errors in the

range .09 to .11. Percentage traits had estimates considerably larger compared to those of

yields, which is the usual pattern normally found in the literature (Benya et al., 1976; Moya

et al., 1985; Wilcox et al., 1962).

In Table 4 are the heritabilities, standard errors and variance components for

measures of SC in milk. Variance components for cows were negative for SCC in 305-d and

total records, and for the weighted average of SCC in total records. Negative estimates were

considered to be zero. The heritability estimates for these traits were in the range .14 to .17

with standard errors .10. Estimates for SCS agreed with some research (Banos and Shook,

1990; Poso and Mantysaari, 1996; Schutz et al., 1990); however some authors reported

Variance components, standard errors and heritability estimates for milk yield (kg) and composition traits (305-d

and total records) in the experiment station Jersey herd.
305-d Total records
h2 S.E o2, o2 o2e h2 S.E a2s 020 2e
M .18 .10 13830.59 50659.18 239518.38 .20 .11 16868.32 50904.24 267346.99
F .22 .11 37.71 105.63 556.48 .23 .11 43.17 88.86 622.75
P .10 .09 9.98 45.78 350.25 .10 .09 11.71 48.03 388.76
LM .23 .11 52.58 157.75 701.49 .25 .11 61.56 156.57 779.63
SNF .16 .10 93.82 364.00 1873.02 .18 .10 112.02 361.91 2079.71
TS .14 .10 155.32 662.60 3746.01 .15 .10 186.22 664.42 4158.23
F % .42 .14 .0264 .0987 .1279 .40 .14 .0250 .1000 .1231
P % .41 .14 .0067 .0234 .0362 .39 .14 .0064 .0237 .0354
LM % .19 .11 .0021 .0055 .0363 .19 .10 .0020 .0054 .0356
SNF % .42 .14 .0042 .0193 .0166 .40 .14 .0039 .0186 .0166
TS % .47 .15 .0197 .0800 .0700 .45 .14 .0188 .0794 .0692
P/F ratio .13 .10 .0003 .0020 .0061 .12 .09 .0002 .0021 .0057
SNF/F ratio .33 .13 .0032 .0146 .0215 .30 .12 .0029 .0152 .0199
AC % .25 .11 .000009 .000014 .000117 .21 .11 .000007 .000016 .000112
CL % .38 .13 .000037 .000042 .000314 .34 .13 .000034 .000051 .000317

o2,, 2c and o2e are the sire, cow, and residual variance component estimates.
M=Milk, F=Fat, P=Protein, LM=Lactose-mineral, SNF=Solids-not-fat and TS=Total
CL= chloride.

solid yields in kg; AC=titratable acidity;

Table 3.

Table 4.

Variance components, standard errors and heritability estimates for somatic cells in milk for all records truncated
at 305-d and total records.

305-d Total records

h2 S.E 2, o2o o h2 S.E o2, o o2,

SCS .17 .10 .0311 .0455 .6650 .16 .10 .0289 .0436 .6354

SCC .15 .10 4489.69 -1861.15 113970.32 .16 .10 4596.19 -1725.38 110833.51

Log SCC .14 .10 .0127 .0247 .3172 .14 .10 .0120 .0243 .3024

WSCS .16 .10 .0350 .0602 .7890

o2,, o2c, and a2 are the sire, cow, and residual variance component estimates.
SCS=Somatic cell score, SCC=Somatic cell count, Log SCC=Natural logarithm of SCC, WSCS=Weighted average of SCS.

lower estimates (Detilleux et al., 1995). In reference to SCC, estimates are comparable to

other reports (Schutz et al., 1990 ). However, considerably lower estimates have been

reported (Monardes et al., 1983). It may be noted that the rate of genetic improvement

following selection on any of these traits should be slow.

In Table 5 are the estimates for reproductive traits; the sire variance component was

negative for number of services and gestation length. The latter estimate represents measures

of the effect of sires of cows, not the sires of calves. PAFOE and PAFBR had very low

heritability estimates, which agreed with most research in this field (Moore et al., 1990;

Raheja et al., 1989; Silva et al., 1992; Weller, 1989). Thus, the magnitude of additive

genetic variance for these traits indicates that the possibility of change by selection are nil.

Therefore, improvement should be performed on basis of controlling and improving

management systems.

Table 5. Variance components, heritability and standard error estimates for reproductive
responses.
Parameter estimates
h2 S.E o2s o2, 2e
NSERV 0 -.0095 .3957 2.41
GEST 0 -2.49 33.98 102.62
PAFOE .05 .08 10.4992 116.69 770.88
PAFBR .02 .08 2.9037 55.96 534.69

o2, S o, and o2e are the sire, cow and residual variance component estimates
NSERV=Number of services, GEST=Gestation Length, PAFOE=Parturition to first
observed estrus, and PAFBR=Parturition to first breeding period.

Table 6. Variance components, repeatabilities and standard errors for milk yield (kg) and composition traits (305-d and total
records).

305-d Totals
t SE 0o2C t SE 2, 2e
M .32 .04 81137.32 170695.54 .33 .04 91981.21 184866.67
F .33 .04 185.71 386.24 .32 .04 192.80 415.64
P .30 .04 94.25 225.59 .31 .04 109.36 244.67
LM .32 .04 255.12 539.60 .33 .04 282.26 578.27
SNF .31 .04 601.32 1350.21 .32 .04 678.61 1451.21
TS .30 .04 1140.43 2612.38 .32 .04 1295.49 2813.70
F % .56 .03 .1269 .1000 .57 .03 .1265 .0965
P % .53 .03 .0317 .0280 .54 .03 .0318 .0271
LM % .30 .04 .0093 .0214 .31 .04 .0092 .0211
SNF % .60 .03 .0207 .0140 .59 .03 .0202 .0139
TS % .61 .03 .0961 .0628 .61 .03 .0950 .0614
P/F ratio .41 .04 .0027 .0040 .43 .04 .0028 .0037
SNF/F ratio .52 .03 .0181 .0166 .54 .03 .0183 .0154
AC% .46 .04 .000041 .000049 .45 .04 .000040 .000049
CL% .32 .04 .000080 .000174 .34 .04 .000085 .000170
o2e and 2e are the sire and residual variance component estimates.
M=Milk, F=Fat, P=Protein, LM= Lactose-mineral, SNF= Solids-not-fat and TS=Total solid yields in kg; AC=titratable
acidity, CL=chloride.
SE=Standard error of estimate.

Repeatability estimates, and their standard errors and variance components for 305-d

and total milk yield and composition records are in Table 6. They followed the same

pattern as the heritabilities in the sense that yields had lower estimates than percentages.

Repeatabilities for yields ranged in the interval .30 to .33 for 305-d records and were not

different from those of totals which in turn were in the range .31 to .33. In both cases

standard errors were .04. On the other hand, percentages ranged between .30 to .61 for

305-d records and again were similar to those of totals which were in the range .31 to .61;

the lower estimates for 305-d and total records were for LM%.

The general pattern for percentage traits to have higher estimates than yields was

observed. Heritabilities for 305-d records tended to be higher than the same estimates from

complete records; for percentages they in general agreed. There is evidence in the literature

suggesting that heritabilities for milk and fat yields differ for each lactation (Albuquerque et

al., 1996; Powell et al., 1981; Rothschild and Henderson, 1979). Estimated heritability for

milk yield is identical to a report from an animal model for California data and close to

estimate for New York data (Albuquerque et al., 1996).

Phenotypic Correlations. In Table 7 are the estimates of the phenotypic correlations for 44

responses included in analysis. Phenotypically, yield traits were highly and positively

associated. These estimates were in the range .78 to 1.0; they in general agreed with

previous results in USA (Gaunt, 1973; Sharma et al., 1983; Wilcox et al., 1971).

Milk yield was negatively associated with percentage traits except for LM%, for

which the estimate was very low. Further, degree of association between MY and

percentages was low. Other yield traits were most often negatively associated with

percentages (Benya et al., 1976; Moya et al., 1985; Wilcox et al., 1971). Percentage traits

among themselves were positively associated within the range .13 to .92; exception to this

pattern was found in the pairs LM%-P%, and LM%-TS%. The ratios P/F and SNF/F were

negatively associated to variable degrees with all percentage traits within the range -.97 to

-.01. There also was observed a strong positive correlation between P/F and SNF/F.

CL% had negative phenotypic correlations with all yields with range -.28 to -.19.

Correlations of CL% with percentages were lower and most often negative. AC% had

extremely low positive phenotypic correlations with yields. Correlations of AC% with

percentages were positive; the only negative estimate was with LM% (-.06), and the highest

estimate was with P%.

Table 7. Estimates of phenotypic correlations between milk yields, milk constituents,
reproductive traits, and somatic cells.

Responses 1 to 15 are from 305-d records.
Responses 16 to 30 are from total lactations.

M=Milk yield, F=Fat yield, P=Protein yield, LM=Lactose-mineral yield,
SNF=Solids-not-fat yield and TS =Total solids yield.

F% =Fat percent, P% =Protein percent, LM% =Lactose-mineral percent,
SNF% = Solids-not-fat percent, TS % =Total solids percent, Ac% =Acidity percent and
Cl% =Chloride percent.

NSERV=Number of services, GEST=Gestation length, PAFOE=Parturition to first
observed estrus, PAFBR=Parturition to first breeding.

DMITO=Total milk yield and DFATO=Total fat yield (DHI records).

MSCS3 =Average somatic cell score (305-d), MSCST=Average somatic cell score(total) ,
MSCC3=Average somatic cell count (305-d), MSCCT=Average somatic cell count (total),
MLOG3=Natural logarithm of somatic cell count (305-d), MLOGT=Natural logarithm of
somatic cell count (total), WSCS=Weighted average of somatic cell score and
WSCC=Weighted average of somatic cell count.

Table 7. (continued)

Trait () 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

1 MY

2FY

3PY

4LMY

5 SNFY

6 TSY

7F%

8 P

9LM%

10 SNF%

11 TS%

12P/F

13 SNF/F

14 CL%

15 AC%

16 MY

17FY

18 PY

19 LMY

20 SNFY

21 TSY

22 F

23 P%

24 LM%

2SSNF%

26TS%

27 P/

28 SNF/F

29 CL%

30AC%

31 NSERV

32 GEST

33 PAFOE

34 PAFBR

35 DMITO

36DFATO

37 MSCS3

38 MSCST

39 MSCC3

40 MSCCT

41MLOG3

42 MLOGT

43 WSCS

44 WSCC

.41

.13 -.65

.66 .61 .21

.58 .92 -31 .87

-.77 .23 -.59 -32

-.97 -.30 -.15 -.54

.05 .01 -.14 -.14

.03 .48 -.42 .19

-.23 -.28 .06 -.30

35 -.01 .13 .12

-.07 .10 -.19 -.07

-.19 -.40 .26 -.25

-.15 -.21 .08 -.19

-.13 -.12 0 -.16

1.00 .42 .12 .66

.41 1.00 -.65 .60

.13 -.65 .99 .20

.66 .61 .20 1.00

.57 .92 -.31 .86

-.77 .22 -.58 -32

-.97 -31 -.14 -.54

.04 .02 -.16 -.15

.03 .46 -.41 .18

-.06 -.10 .09 -.04

-.02 -.11 .02 -.12

-.06 -.06 -.03 -.10

-.03 -.04 -.01 -.06

-.24 -.26 .05 -.29

.27 -.01 .11 .10

-.06 ..06 -.01 -.08

-.06 -.06 -.01 -.09

-.05 -.07 .04 -.S5

-.05 -.07 .04 -.05

-.06 -.06 0 -.08

-.06 -.06 0 -.08

-.06 -.07 0 -.08

-.06 -.07 .03 -.05

.80

.91 .82

.97 .80 .83

.99 .84 .94

.98 .85 .97

-.22 .35 -.07

-.28 -.02 .10

.07 .14 -.18

-.29 .13 -.06

-32 .05 .03

.04 .28 .12

.18 -.37 .06

-.25 -.19 -.24

.04 .04 .22

.10 .05 .06

.10 .08 .08

.02 -.01 .01

.03 -.01 .02

.90 .69 .83

.71 .87 .75

-.25 -.28 -.27

-.24 -.27 -.27

-.13 -.15 -.16

-.12 -.15 -.15

-.27 -30 -.29

-.26 -.29 -.28

-.23 -.26 -.25

-.12 -.14 -.14

-.43

-.25 .04

-.20 .05

-.25 .23

-.27 -.05

-.28 .06

-.27 .11

.05 .04

.01 .48

-.13 -.42

-.14 .19

-.07 .40

-.04 .28

-.08 .02

.99 -.43

-.44 .99

-.05 -.02

.01 -.06

.04 -.05

.06 -.05

-.24 .01

-.21 .02

.44 -.28

.44 -.28

.28 -.15

.28 -.15

.45 -.26

.44 -.26

.43 -.28

.27 -.15

.97

.94 .99

-.19 -.15

-.41 -.21

.27 .09

-.24 -.18

-.37 -.22

-.08 0

.15 .12

-.27 -.27

-.0 .06

.12 .10

.10 .09

.01 .01

.02 .02

.87 .89

.71 .76

-.25 -.27

-.25 -.27

-.11 -.14

-.11 -.13

-.27 -.29

-.26 -.28

-.23 -.25

-.10 -.13

Table 7. (continued)

Trait (i) Trait 0)
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
23 .41
24 .13 -.65
25 .66 .61 .20
26 .58 .92 -.31 .87
27 -.77 .23 -.59 -.32 -.01
28 -.97 -.31 -.14 -.54 -.45 .84
29 .04 .01 -.15 -.15 -.07 -.02 -.07
30 .03 .47 -.42 .19 .39 .28 .03 -.44
31 -.06 -.11 .09 -.04 -.09 0 .07 -.05 -.02
32 -.02 -.11 .02 -.12 -.12 -.03 .01 .01 -.06 -.08
33 -.05 -.05 -.03 -.10 -.08 .02 .03 .04 -.05 .02 -.01
34 -.03 -.04 -.01 -.05 -.05 0 .01 .06 -.05 .06 .07 .54
35 -.24 -.25 .05 -.28 -.30 .07 .20 -.24 0 .11 .09 .01 .04
36 .28 -.01 .12 .11 .04 -.30 -.29 -.21 .01 .05 .07 -.02 0 .80
37 -.06 -.06 -.01 -.09 -.08 .02 .05 .45 -.28 -.03 0 .03 .03 -.22 -.27
38 -.06 -.06 -.01 -.09 -.08 .02 .05 .45 -.28 -.02 0 .03 .04 -.22 -.27 .99
39 -.05 -.07 .04 -.05 -.07 .01 .05 .28 -.15 .01 0 .06 .06 -.11 -.15 .71 .71
40 -.05 -.07 .04 -.05 -.07 .01 .05 .28 -.15 .01 0 .07 .06 -.10 -.14 .71 .72 .98
41 -.606 -6 0 -.08 -.08 .02 .05 .45 -.27 -.01 .01 .04 .05 -.24 -.29 .98 .98 .70 .70
42 -.06 -.06 0 -.08 -.08 .02 .05 .45 -.27 -.01 .01 .03 .05 -.23 -.28 .98 .98 .70 .71 .99
43 -.06 -.06 0 -.08 -.08 .02 .05 .44 -.28 -.02 0 .03 .04 -.21 -.25 .98 .98 .72 .72 .96 .97
44 -.06 -.06 .04 -.05 -.06 .01 .05 .27 -.15 0 0 .07 .06 -.10 -.13 .68 .69 .96 .97 .67 .67 .70

Phenotypic correlations for variables for total records are in the area of rows and

columns 16 to 30. On the whole, phenotypic correlation estimates based on total records

agreed with those found with 305-d records. This means that decisions made based upon

either group should lead to the same conclusions.

Below the diagonal, in the area of rows 16 to 30 and columns 1 to 15 are the

phenotypic correlations among milk yield and composition traits for 305-d and total lactations

for all records. Along the diagonal of that submatrix estimates were high and positive (close

to or equal to 1.0 for most traits), suggesting that 305-d records and those from totals were

essentially the same traits.

The lack of phenotypic associations between yield and composition traits with those of

reproductive performance can be noted on the area of rows 31 to 34 and columns 1 to 15 for

305-d records on all records data set. Phenotypic correlations were extremely low without

discernible pattern. This suggests that any possible antagonistic effect of yields on

reproduction explainable through genetic mechanisms would be overcome by using

appropriate management practices.

All measures of somatic cells in milk showed negative phenotypic associations with

yield traits in the range -30 to -.05. This agreed with most previous research concerning the

phenotypic relationship between yields and SC in milk (Banos and Shook, 1990; Boettcher et

al., 1992; Emanuelson et al., 1988). This means that on the phenotypic scale cows with

high SC had low yields. One of the possible factors associated with reduction of milk yield

during mastitis is because leukocytic migration into milk may damage mammary gland

epithelial tissue (Akers and Thompson, 1987). The responses F%, LM%, and SNF% also

had negative estimates. Correlations with P%, and TS% were positive but low.

Nevertheless, in light of the degree of these estimates, conclusions should be made

cautiously. Weighting either SCS or SCC did not change the degree of the association

appreciably. Phenotypic correlations for measures of SC in milk among themselves were

positive, high within the range .67 to .99, which agreed with previous research (Boettcher et

al., 1992). In addition, it may be observed that correlations for 305-d and total records in

general agreed for all these responses.

Genetic Correlations. Estimates of the genetic correlations are in Table 8. They are

shown in the same manner as the phenotypic correlations. Estimates for 305-d records are in

rows and columns 1 to 15. As with their phenotypic counterparts, genetic correlations

between yields were high and positive and usually similar or only slightly lower than the

phenotypic correlations, which is the usual pattern observed for estimates of these parameters

(Searle, 1961). Degree of these correlations suggests that, on average, most genes involved

in milk yield processes are the same. As such, correlated responses should be expected

following selection for any of them. The lower correlation coefficients observed in all data

records were those of M-F and M-P yields; these results have been observed previously

(Moya et al., 1985; Wilcox et al., 1971).

Table 8. Estimates of genetic correlations between milk yields, milk constituents,
reproductive traits, and somatic cells.

Responses 1 to 15 are from 305-d records.
Responses 16 to 30 are from total lactations.

M=Milk yield, F=Fat yield, P=Protein yield, LM=Lactose-mineral yield,
SNF=Solids-not-fat yield and TS=Total solids yield.

F % =Fat percent, P% =Protein percent, LM% =Lactose-mineral percent,
SNF% =Solids-not-fat percent, TS% =Total solids percent, Ac% =Acidity percent and
C1% =Chloride percent.

NSERV=Number of services, GEST=Gestation length, PAFOE=Parturition to first
observed estrus, PAFBR=Parturition to first breeding.

DMITO=Total milk yield and DFATO=Total fat yield (DHI records).

MSCS3=Average somatic cell score (305-d), MSCST=Average somatic cell score(total) ,
MSCC3 =Average somatic cell count (305-d), MSCCT= Average somatic cell count (total),
MLOG3 =Natural logarithm of somatic cell count (305-d), MLOGT=Natural logarithm of
somatic cell count (total), WSCS =Weighted average of somatic cell score and
WSCC=Weighted average of somatic cell count.

*Not estimable (variance estimates were negative).

Table 8. (continued)

Trait (i) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

1 MY

2FY

3 PY

4LMY

5 SNFY

6 TSY

7F%

8 P%

9LM

10 SNF%

11 TS%

12P/F

13 SNF/F

14 CL%

15 AC%

16MY

17 FY

18 PY

19LMY

20SNFY

21 TSY

22 F%

23 P%

24LM%

25 SNF%

26TS%

27 P/F

28 SNF/F

29 CL%

30AC%

31 NSERV*

32 GUEST*

33 PAFOE

34 PAFBR

35 DMITO

37 MSCS3

38 MSCS

39 MSCC3

40 MSCC

41 MLOG#

42 MLOGT

43 WSCS

44WSCC

.99

-.11 0

-.64 -.53 .85

.45 35 -.14 -.62

-.50 -.43 .98 .83 -.07

-.61 -.51 .94 .97 -.39 .95

-.48 -.55 -.86 -.43 -.43 -.85 -.64

.03 -.09 -1.0 -.85 .12 -.98 -.94 .85

-.69 -.72 .25 .07 -.02 .08 .07 -.57 -.38

.09 .22 .24 .69 -.80 31 .55 .46 -.15 -.71

.99 .96 -.36 -.77 .47 -.64 -.75 -.19 .29 -.62 -.02

.76 .82 .56 -.04 .34 .19 .06 -1,1 1,68 -.36 .17

.91 .97 .25 -.20 .11 -.18 -.20 -.66 -.36 -.69 .52

.97 .93 -35 -.82 .62 -.61 -.77 -.27 .28 -.59 -.19

1.02 1.00 -.18 -.68 .49 -.52 -.64 -.41 .09 -.66 .03

1.02 1.02 -.08 -.57 .41 -.45 -.55 -.48 -.02 -.69 .15

-.10 .01 1.0 .85 -.14 .97 .94 -.85 -1.0 .26 .25

-.66 -.55 .85 1.00 -.62 .83 .96 -.43 -.85 .10 .69

.46 .36 -.13 -.62 1.00 -.08 -.40 -.43 .12 -.03 -.81

-.52 -.45 .99 .84 -.08 1.00 .95 -.85 -.99 .10 .31

-.63 -.53 .95 .97 -.40 .95 1.00 -.64 -.95 .09 .55

-.56 -.64 ..87 -.41 -.47 -.84 -.62 1.00 .83 -.60 .49

0 -.13 -1.0 -.85 .11 -.99 -.95 .85 1.00 -.40 -.16

-.73 -.76 .29 .08 -.0 .09 .08 -.64 -.42 1.01 -.70

.11 .24 .20 .67 -.81 .27 .52 .50 -.11 -.72 1.00

.57 .38 -1.1 -1.4 1.22 -.91 -1.2 .49 1.2 -.58 -.26

.77 .61 -1.6 -1.4 1.67 -.71 -1.2 1.6 2.0 .22 -.54

1.08 1.05 -.27 -.70 .45 -.57 -.68 -.29 .18 -.77 .15

.37 .40 .06 -.23 .29 -.08 -.17 -39 -.11 .48 -.32

.44 .47 .06 -.24 .34 -.07 -.17 -.42 -.12 .47 -.32

.37 .41 -.08 -.18 -.11 -32 -.25 -.07 .02 .11 .05

.39 .43 -.06 -.20 -.09 -32 -.26 -.11 -.01 .10 .05

.49 .53 -.01 -.25 .40 -.04 -.16 -.30 -.02 .42 -.39

.54 .58 -. -.28 .45 -.04 -.18 -34 -.03 .42 -.40

38 .42 .14 .18 .28 -.03 -.11 -.50 -.21 .52 -.37

.47 .52 .01 -.17 -.11 -30 -.23 -.23 -.10 .07 .07

.58

.80 .86

.98 .58 .72

.99 .71 .86

.97 .76 .92

-.36 .57 .26

-.79 -.05 -.22

.48 35 .11

-.67 .17 -.22

-.77 .04 -.23

-.25 -1.2 -.73

.27 -.71 -.39

-.65 -37 -.73

.01 .18 .55

.82 -.48 -.10

.85 -.94 .02

1.05 .65 .85

.37 31 .50

.42 35 .56

.46 .19 .60

.48 .22 .62

.48 34 .62

.53 37 .68

36 .38 .53

.55 .34 .73

.98

.94 .99

-.35 -.17

-.84 -.70

.63 .50

-.64 -.54

-.79 -.66

-.34 -.49

.26 .07

-.62 -.69

-.16 .06

.90 .64

.91 .70

1.06 1.06

.30 39

.35 .45

.30 .42

33 .44

.42 .51

.47 .6

.30 .40

39 .52

- -- --

Table 8. (continued)

Trait (i) Trait ()
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 *~fl
23 .85
24 -.14 -.62
25 .99 .84 -.09
26 ,95 ,97 -.40 ,95
27 -.85 -.41 -.47 -.85 -.63
28 -1.0 -.85 .12 -1.0 -.95 .83
29 .29 .10 -.03 .11 .10 -.67 -.44
30 .22 .67 -.82 .27 .52 .52 -.12 -.71
31
32 -
33 -1.2 -1.4 1.2 -.92 -1.2 .58 1.29 -.62 -.15
34 -1.6 -1.4 1.7 -.69 -1.2 1.5 2.0 .24 -.72 1.6
35 -.27 -.73 .45 -.60 -.70 -35 .17 -.79 .20 .56 1.0
36 .36 -.05 .34 .18 .05 -1.2 -.70 -.51 .28 -.81 -.63 .65
37 .07 -.20 .29 -.06 -.15 -.41 -.12 .51 -.35 .81 2.2 .40 .27
38 .08 -.22 .34 -.04 -.15 -.45 -.14 .50 -.36 .78 2.3 .45 32 1.00
39 -.08 -.16 -.14 -.31 -.23 -.06 .01 .17 .06 1.4 1.1 .56 .21 .78 .83
40 -.06 -.18 -.11 -.31 -.25 -.11 -.02 .15 .06 1.4 1.1 .57 .23 .76 .80 1.1
41 0 -.23 .41 -.01 -.14 -32 -.03 .43 -.43 .79 2.4 .52 32 1.0 1.0 .75 .73
42 .01 -.26 .46 -.01 -.16 -.37 -.04 .44 -.45 .76 2.6 .57 .36 1.0 1.0 .78 .76 1.0
43 .15 -.16 .28 -. -. -. -.1 -.53 -.22 .56 -.40 .84 1.9 .39 .34 1.0 1.0 .84 .82 1.0 1.02
44 .02 -.16 -.13 -.29 -.22 -.23 -.11 .13 .08 1.5 .66 .65 34 .76 .80 1.2 1.2 .72 .76 83

Most genetic correlations involving yields and percentages were negative. This was

so for both the all records and first lactation data sets. However, estimates involving LM%

were all positive and within the range .01 to .60. This was observed previously for Holsteins

and Jerseys (Sharma et al., 1983). Nevertheless, several authors have reported consistently

negative genetic associations between LM% and yields (Benya et al., 1976; Moya et al.,

1985). It is most likely for the correlation between lactose concentration and yields to be

positive because lactose is related to milk yield volume by affecting osmotic pressure

regulating water transport to the mammary gland. Pleiotropy and linkage have been cited as

the factors determining genetic correlations (Falconer, 1990). The contradictory results

found in the literature could be explained if linkage rather than pleiotropy were a more

important factor in determining this association, although this explanation does not seem

plausible.

Genetic correlations involving LM% and remaining percentages were negative ranging

from -.62 to -.07. Other percentages had strong positive genetic relationships in the range

.83 to .95. High positive correlations between these traits have been cited previously for

Jerseys (Sharma et al., 1983; Wilcox et al., 1971). Nevertheless, lower estimates have been

published (Chauhan and Hayes, 1991; Meyer, 1985) for fat and protein percentages.

Response variables in the area of rows 16 to 30 and columns 1 to 15 shows the

genetic relationships between 305-d and total records. As with the phenotypic correlations,

inspecting the diagonal of that submatrix reveals that from a genetic point of view both

groups are essentially the same with parameter estimates in the range 1.0 to 1.03. Estimated

genetic correlations based on total records are in the area of rows and columns 16 to 30. On

average they most often followed the same pattern observed previously for 305-d records in

both all data and first parity cows analyses.

All genetic correlations involving CL% and yields were negative and most often

intermediate in degree. This is contrary to findings in Florida for Jerseys (Benya et al.,

1976; Moya et al., 1985). However, evidence of negative association for F and TS was

reported for the breed (Sharma et al., 1983).

Genetic correlations between measures of SC in milk and yields and composition traits

are in the lower triangle in the area of rows 37 to 44 and columns 1 to 15. For all records,

genetic correlations between yields and SC were intermediate and positive. Analyses agreed

with what most often is found in the literature, suggesting an intermediate antagonistic

relationship between these variables (Boettcher et al., 1992; Emanuelson et al., 1988; Welper

and Freeman, 1992). On the other hand, evidence indicating a clear antagonic association

between clinical mastitis and milk yield but indicating genetic independence between milk

yield and SCS was reported (Poso and Mantysaari, 1996), since their estimate was not

different from zero.

Genetic correlations between SC measurements and percentages did not show a clear

definite pattern. In all records most correlation coefficients were negative. Estimates of

Table 9. Estimates of environmental correlations between milk yields, milk constituents,
reproductive traits, and somatic cells.

Responses 1 to 15 are from 305-d records.
Responses 16 to 30 are from total lactations.
M=Milk yield, F=Fat yield, P=Protein yield, LM=Lactose-mineral yield,
SNF=Solids-not-fat yield and TS=Total solids yield.
F% =Fat percent, P% =Protein percent, LM% =Lactose-mineral percent,
SNF% =Solids-not-fat percent, TS% =Total solids percent, Ac% =Acidity percent and
Cl% =Chloride percent.
NSERV=Number of services, GEST=Gestation length, PAFOE=Parturition to first
observed estrus, PAFBR=Parturition to first breeding.
DMITO=Total milk yield and DFATO=Total fat yield (DHI records).
MSCS3=Average somatic cell score (305-d), MSCST=Average somatic cell score(total) ,
MSCC3=Average somatic cell count (305-d), MSCCT=Average somatic cell count (total),
MLOG3 =Natural logarithm of somatic cell count (305-d), MLOGT=Natural logarithm of
somatic cell count (total), WSCS =Weighted average of somatic cell score and
WSCC =Weighted average of somatic cell count.
*Not estimable (variance estimates were negative).

Table 9. (continued)

Trt (1) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

1MY

2FY .86

3PY .95 .83

4LMY .97 .87 .87

5SNFY .99 .88 .96 .98

6 TSY .99 .88 .98 .96 1.0

7F% -.18 .26 -.17 -.12 -.15 -.16

8P% -.09 -.02 .19 -.22 -.05 .02 .11

9 LM% -.03 .10 -.22 .17 .01 -.06 .25 -.69

10 SNF% -.15 .10 -.03 -.07 -.05 -.05 .44 .45 33

11TS% -.13 .04 .11 -.18 -.06 -.01 .29 .89 -.29 .81

12P/F .10 -.23 .23 -.02 .09 .13 -.80 .45 -.62 -.18 .22

13 SNF/F .15 -.25 .17 .10 .13 .14 -.95 .01 -.24 -.28 -.14 .88

14CL% -.15 -.15 -.17 -.17 -.18 -.17 -.05 -.02 -.18 -.25 -1.4 .09 .04

15 AC% .04 -.03 .15 -.02 .06 .08 -.10 35 -.29 .12 .30 .25 .12 -31

16MY .97 .83 .92 .94 .96 .96 -.18 -.08 -.05 -.17 -.14 .11 .15 -.15 .07

17 FY .84 .97 .81 .84 .85 .85 .26 -.01 .08 .09 .04 -.22 -.25 -.14 -.01 .85

18 PY .92 .79 .97 .84 .92 .94 -.18 .19 -.24 -.05 .10 .25 .18 -.17 .17 .94 .3

19LMY .94 .84 .84 .97 .95 .92 -.12 -.21 .16 -.08 -.18 -.02 .10 -.17 0 .97 .87 .87

20SNFY .97 .85 .93 .95 .97 -.96 -.15 -.04 -.01 -.07 -.06 .10 .14 -.18 .08 .99 .88 .95 .98

21TSY .96 .84 .95 .92 .96 .97 -.16 .03 -.08 -.06 -.01 .14 .15 -.18 .11 .99 .87 .98 .95 1.0

22F% -.18 .26 -.16 -.13 -.14 -.15 .99 .12 .23 .44 .30 -.79 -.95 -.05 -.10 -.18 .27 -.16 -.12 -.14 -.15

23 P% -.09 -.03 .18 -.22 -.05 .02 .10 .99 -.68 .45 .89 .45 .01 -.04 .35 -.08 0 .20 -.21 -.04 .03

24LM% -.02 .11 -.21 .18 .02 -.05 .25 -.69 .99 .33 -.29 -.62 ..24 -.16 -.29 -.04 .09 -.23 .17 0 -.07

25 SNF -.13 .10 -.01 -.06 -.04 -.03 .43 .46 .32 .99 .80 -.17 -.28 -.25 .12 -.15 .11 -.02 -.06 -.05 -.04

26TS% -.12 .03 .11 -.18 -.06 -.01 .28 .88 -.29 .80 .99 .22 -.13 -.15 .30 -.13 .05 .12 -.17 -.05 0

27P/F .09 -.24 .22 -.02 .08 .12 -.81 .43 -.61 ..18 .20 .99 .88 .08 .24 .10 -.23 .23 -.02 .09 .13

28SNF/F .15 -.25 .15 .11 .13 .14 -.95 -.02 -.22 -.29 -.16 .86 .99 .04 .11 .15 -.25 .16 .10 .13 .14

29CL% -.14 -.15 -.15 -.17 -.17 -.17 -.07 -.01 -.19 -.26 -.14 .11 .06 .99 -32 -.14 -.14 -.16 -.17 -.17 -.17

30 AC% .03 -.04 .13 -.03 .04 .07 -.08 .35 -.28 .14 31 .23 .10 -33 .98 .05 -.01 .16 -.01 .07 .10

31NSERV- -

32 GEST

33PAFOE -.08 .03 .01 -.12 -.07 -.04 .14 .18 -.17 .03 .13 -.02 -.14 .12 -.02 -.07 .05 .02 -.10 -.05 -.03

34PAFBR -.04 .05 .01 -.07 -.04 -.03 .16 .12 -.13 .01 .08 -.08 -.17 .06 0 -.03 -.07 .02 -.05 -.02 -.01

35DMITO .85 .69 .81 .81 .84 .84 -.24 -.09 -.05 -.18 -.15 .16 .22 -.10 -.05 .87 .70 .83 .82 .85 .85

36DFATO .71 .80 .71 .69 .72 .72 .14 -.01 .05 .05 .02 -.12 -.13 -.12 -.09 .72 .82 .72 .71 .74 .74

37MSCS3 -39 -.41 -.39 -.40 -.41 -.41 -.11 0 -.07 -.09 -.05 .09 .10 .44 -.28 -39 -.42 -39 -.40 -.41 -.41

38MSCST -.39 -.41 -39 -.40 -.41 -.41 -.11 0 -.08 -.10 -.05 .09 .10 .44 -.28 -39 -.43 -.39 -.40 -.41 -.41

39 MSCC3 -.24 -.22 -.25 -.21 -.24 -.24 -.05 -.03 .07 .05 0 .02 .06 .32 -.23 -.25 -23 -.27 -.22 -.25 -.26

40 MSCC -.25 -.22 -.25 -23 -.24 ..24 -.05 -.02 .06 .05 .01 .03 .07 32 -.22 -.26 -.24 -.27 -.22 -.25 -.26

41 MLOG3 -.42 -.43 -.41 -.43 -.44 -.43 -.08 0 -.08 -.10 -.05 .07 .07 .46 -.24 -.42 -.44 -.41 -.43 -.44 -,44

42 MLOGT -.42 -.43 -.41 -.43 -.44 -.43 -.09 .01 -.09 -.10 -.05 .07 .07 .45 -.23 -.42 -.44 -AI -.44 -.44 -.44

43WSCS -.36 -.40 -37 -36 -.38 -38 -.14 -.03 -.05 -.10 -.08 .10 .13 .41 -.26 -.36 -.41 -.36 -.36 -.38 -.38

44WSCC -.23 -23 -.24 -.21 -.23 -.23 -.08 -.03 .06 .03 -.01 .05 .09 .31 -.22 -.24 -.24 -.25 -22 -.24 -.24

Table 9. (continued)

correlations between SCS and F%, and lactose have been small and negative (Welper and

Freeman, 1992) whereas, correlations between SCS and P% are low and positive (Schutz et

al., 1990; Welper and Freeman, 1992). All genetic correlations among SC measurements

(rows and columns 37 to 44) were high and positive following the same pattern as their

phenotypic counterparts.

Environmental Correlations. Environmental correlations for the traits included in this

research are in Table 9. They are defined by Falconer (1990). They are in reality residual

correlations and, as such, include the environmental deviations plus non-additive genetic

differences. For this reason it is difficult to find a real interpretation of these coefficients

Trait i) Trait j)
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
23 .12
24 .23 -.69
25 .44 .46 32
26 .30 .89 -.29 .81
27 -.81 .43 -.62 -.18 .20
28 -.96 -.02 -.22 -.29 -.16 .87
29 -.07 -.03 -.18 -.26 -.15 .10 .06
30 -.08 .35 -.29 .14 31 .23 .10 -.34
31 -
32 -
33 .16 .18 -.16 .04 .14 -.03 -.16 .12 -.04
34 .16 .12 -.13 .01 .09 -.08 -.18 .05 .01 .61
35 -.23 -.08 -.04 -.16 -.14 .15 .21 -.10 -.07 -. -.03
36 .16 .01 .07 .08 .04 -.13 -.14 -.13 -.10 .07 .05 .84
37 -.12 -.01 -.07 -.10 -.06 .09 .10 .43 -.27 -.05 -.10 -.36 -.40
38 -.12 -.01 -.08 -.11 -.06 .09 .10 .44 -.27 -.04 -.11 -.36 -.41 .99
39 -.05 -.04 .08 .04 -.01 .02 .06 .30 -.22 -.07 0 -.24 -.23 .70 .69
40 -.06 -.03 .07 .05 0 .03 .07 31 -.22 -.06 0 -.24 -.23 .70 .70 .96
41 -.09 -.01 -.08 -.11 -.06 .07 .07 45 -.23 -.03 -.09 -39 -.42 .98 .97 .69 .69
42 -.09 0 -.09 -.11 -.06 .08 .08 .46 -.22 -.03 -.09 -39 -.43 .97 .98 .69 .70 .99
43 -.14 -.04 -.05 -.11 -.08 .10 .13 .41 -.25 -.05 -.0 -.33 -.40 .98 .98 .69 .70 .95 .97
44 -.08 -.04 .07 .03 -.02 .05 .09 30 -.21 -.05 .03 -.23 -.23 .67 .67 .93 .94 .66 .66 .69

and they are of questionable practical value. In the lower triangle, rows 1 to 15, are the

environmental correlations for milk yield and composition traits. For yield traits,

correlations among them followed the same pattern observed for the genetic correlations.

However, they usually were slightly higher, which is the usual situation (Falconer, 1990).

Environmental correlations of yields and percentages were mostly negative with cases where

signs were reversed. With respect to estimates for percentages among them, most were

smaller than their genetic counterparts, and most were positive. Environmental correlations

involving reproductive measures and yield and composition traits are in the area of rows 31

to 34 and columns 1 to 15. They were considerably lower than their genetic counterparts.

In the area of rows 37 to 44 and columns 1 to 15 are the environmental correlations

between SC variables and those of yield and milk composition. All correlation coefficients

between SC variable and yields had reverse sign in contrast to the genetic correlations.

Falconer (1990) explained that a large difference, and particularly a difference in sign, shows

that genetic and environmental sources of variation affect the characters through different

physiological mechanisms. For the environmental correlations between SC variables and

percentages, no particular pattern was found, but in general, they were most often positive

but small in degree.

All environmental correlations involving CL% and SC were positive and within the

range .31 to .46 (rows 37 to 44 and column 14). On the other hand, estimates involving

Ac% and measures of SC were negative within the range -.28 to -.22 (rows 37 to 44 and

column 15).

CONCLUSIONS

Heritability estimates for milk yield and composition traits were within the ranges

previously reported, with estimates higher for percentages than for yields. Heritabilities of

measures of somatic cells in milk were low suggesting that only small changes may be

achieved by direct selection for any of these traits. Heritability estimates for the periods

from parturition to first observed estrus and for parturition to first artificial insemination

were essentially zero, indicating that for these traits improvement of population parameters

should be based on refining and improving management practices. Genetic correlations

between yields were high and positive suggesting the possibility of favorable correlated

responses if selection is placed on milk, percentage traits were negatively associated with

yields, and they in turn were highly correlated among themselves. Overall evidence

confirms a moderate antagonistic relationship between yields and somatic cells on milk.

Phenotypic correlations between milk yield and composition and reproductive response

variables were negligible.

REFERENCES

Ahlborn, G., and L. Dempfle. 1992. Genetic parameters for milk production and body size
in New Zealand Holstein-Friesian and Jersey. Livest. Prod. Sci. 31:205-219.

Akers, R. M., and W. Thompson. 1987. Effect of induced leucocyte migration on
mammary cell morphology and milk component biosynthesis. J. Dairy Sci.
70:1685-1695.

Albuquerque, L. G., J. F. Keown, and L. D. Van Vleck. 1996. Genetic parameters of milk,
fat, and protein yields in the first three lactations, using an animal model and
restricted maximum likelihood. Brazil. J. Genetics 19:79-86.

Ali, A. K. A., and G. E. Shook. 1980. An optimum transformation for somatic cell
concentration in milk. J. Dairy Sci. 63:487-490.

Banos, G., and G. E.Shook. 1990. Genotype by environment interaction and genetic
correlations among parities for somatic cell count and milk yield. J. Dairy Sci.
73:2563-2573.

Benya, E. G., C. J. Wilcox, F. G.Martin, R. W. Adkinson, W. A. Krienke, and D. E.
Franke. 1976. Parametros geneticos para peso corporal, composition y production
de leche de un reba0o localizado en Florida. ALPA. MEM. 11:163-169.

Boettcher, P. J., L. B. Hansen, P. M. VanRaden, and C. A. Ernst. 1992. Genetic
evaluation of Holstein bulls for somatic cells in milk of daughters. J. Dairy Sci.
75:1127-1137.

Campos, M. S., C. J. Wilcox, C. M. Becerril, and A. Diz. 1994. Genetic parameters for
yield and reproductive traits of Holstein and Jersey cattle in Florida. J. Dairy Sci.
77:867-873.

Chauhan, V. P. S., and J. F. Hayes. 1991. Genetic parameters for first lactation milk
production and composition traits for Holsteins using multivariate restricted maximum
likelihood. J. Dairy Sci. 74:603-610.

Coffey, E. M., W. E. Vinson, and R. E. Pearson. 1985. Heritabilities for lactation average
of somatic cell counts in first, second, and third or later parities. J. Dairy Sci.
68:3360-3362.

Darwash, A. O., G. E. Lamming, and J. A. Woolliams. 1997. Estimation of genetic
variation in the interval from calving to postpartum ovulation of dairy cows. J. Dairy
Sci. 80:1227-1234.

Detilleux, J. C., M. E. Kehrli, Jr., A. E. Freeman, L. K. Fox, and D. H. Kelley. 1995.
Mastitis of periparturient Holstein cattle: A phenotypic and genetic study. J. Dairy
Sci. 78:2285-2293.

Emanuelson. U., B. Danell, and J. Philipsson. 1988. Genetic parameters for clinical
mastitis, somatic cell counts, and milk production estimated by multiple-trait
Restricted Maximum Likelihood. J. Dairy Sci. 71:467-476.

Falconer, D. S. 1990. Introduction to quantitative genetics. (3rd Ed.). Longman Scientific &
Technical. Hong Kong.

Gacula, M. C. Jr., S. N. Gaunt, and R. A. Damon, Jr. 1968. Genetic and environmental
parameters of milk constituents for five breeds. II. Some genetic parameters. J. Dairy
Sci. 51:438-444.

Gaunt, S. N. 1973. Genetic and environmental changes possible in milk composition. J.
Dairy Sci. 56:270-278.

Gaunt, S. N., C. J. Wilcox, B. R. Farthing, and N. R.Thompson. 1968. Genetic
interrelationships of Holstein milk composition and yield. J. Dairy Sci. 51:1396-1402.

Harvey, W. R. 1960. Least-squares analysis of data with unequal subclass numbers
ARS-20-8 USDA.(Mimeo). Beltsville, Md.

Harvey, W. R. 1990. User's guide for LSMLMW and MIXMDL. PC-2 version. Mixed
model least-squares maximum likelihood computer program 1990. (Mimeo).
Columbus, Oh.

Jones, W. P., L. B. Hansen, and H. Chester-Jones. 1994. Response to health care to
selection for milk yield of dairy cattle. J. Dairy Sci. 77:3137-3152.

Madsen, P. S. 1969. Absorption-spectrophotometric determination of DNA
(Deoxyribonucleic acid) in milk at modum Feulgen. Accuracy of analysis, specificity,
correlation to content of cells. Acta Vet. Scan. 10:319-344.

Meyer, K. 1984. Estimates of genetic parameters for milk and fat yield for the first three
lactations in British Friesian cows. Anim. Prod. 38:313-322.

Meyer, K. 1985. Genetic parameters for dairy production of Australian black and white
cows. Livest. Prod. Sci. 12:205-219.

Monardes. H. G., B. W. Kennedy, and J. E. Moxley. 1983. Heritabilities of measures of
somatic cell count per lactation. J. Dairy Sci. 66:1707-1713.

Moore, R. K., B. W. Kennedy, L. R. Schaeffer, and J. E.Moxley. 1990. Relationships
between reproduction traits, age and body weight at calving, and days dry in first
lactation Ayrshires and Holsteins. J. Dairy Sci. 73:835-842.

Moya, J., C. J. Wilcox, K. C. Bachman, and F. G.Martin. 1985. Genetic trends in milk
yield and composition in a subtropical dairy herd. Brazil. J. Genetics. VII:509-521.

Oltenacu, P. A., C. C. Olson., and C. W. Young. 1979. Repeatability of milk and fat yield
for cows with changed environments. J. Dairy Sci. 62:310-315.

Paape, M. J., H. A.Tucker, and H. D. Hafs. 1965. Comparison of methods for estimating
milk somatic cells. J. Dairy Sci. 48:191-196.

Poso, J., and E. A. Mantysaari. 1996. Relationships between clinical mastitis, somatic cell
score, and production for first three lactations of Finnish Ayrshires. J. Dairy Sci.
79:1284-1291.

Powell, R. L., H. D. Norman, and R. M. Elliot. 1981. Different lactations for estimating
genetic merit of dairy cows. J. Dairy Sci. 64:321-330.

Raheja, K. L., E. B. Burnside, and L. R. Schaeffer. 1989. Relationships between fertility
and production in Holstein dairy cattle in different lactations. J. Dairy Sci.
72:2670-2678.

Reents, R., J. C. M. Dekkers, and L. R. Schaeffer. 1995. Estimation of genetic parameters
for test day records of somatic cell score. J. Dairy Sci. 78:2847-2857.

Rothschild, M. F., and C. R. Henderson. 1979. Maximum likelihood estimates of
parameters of first and second lactation milk records. J. Dairy Sci. 62:990-995.

Schutz, M. M., L. B. Hansen, G. R. Steuernagel, J. K. Reneau, and A. L. Kuck. 1990.
Genetic parameters for somatic cells, protein, and fat in milk of Holsteins. J. Dairy
Sci. 73:494-502.

Schutz, M. M., P. M. VanRaden, G. R. Wiggans, and H. D. Norman. 1995.
Standardization of lactation means of somatic cell scores for calculation of genetic
evaluations. J. Dairy Sci. 78:1843-1854.

Searle, S. R. 1961. Phenotypic, genetic and environmental correlations. Biometrics.
17:474-480.

Sharma, A. K., L. A. Rodriguez, G. Mekonnen, C. J. Wilcox, K. C. Bachman, and R. J.
Collier. 1983. Climatological and genetic effects on milk composition and yield. J.
Dairy Sci. 66:119-126.

Shook, G. E., and M. M. Schutz. 1994. Selection on somatic cell score to improve
resistance to mastitis in the United States. J. Dairy Sci. 77:648-658.

Silva, H. M., C. J. Wilcox, W. W. Thatcher, R. B. Becker, and D. Morse. 1992. Factors
affecting days open, gestation length, and calving interval in Florida dairy cattle. J.
Dairy Sci. 75:288-293.

Simerl, N. A. 1982. Effects of age at first freshening on productive and reproductive
performance of dairy heifers. M.S. Thesis. University of Florida, Gainesville.

Thatcher, W. W., E. C. Harland, W. B. Fredriksson, C. J. Wilcox, and K. L. Smith. 1971.
Routine use at Florida of Feulgen: DNA measurements of total milk somatic cells.In:
Min. Ann. Mtg. Southern Regional Tech. Commi. Genetic methods of improving
dairy cattle for the South. S-49 Project. North Carolina State.

Van Vleck, L. D., and M. C. Dong. 1988. Genetic (co)variances for milk, fat, and protein
yield in Holsteins using an animal model. J. Dairy Sci. 71:3040-3046.

Weller, J. I. 1989. Genetic analysis of fertility traits in Israeli dairy cattle. J. Dairy Sci.
72:2644-2650.

Welper, R. D., and A. E. Freeman. 1992. Genetic parameters for yield traits of Holsteins,
including lactose and somatic cell score. J. Dairy Sci. 75:1342-1348.

Wilcox, C. J., K. O. Pfau, R. E. Mather, R. F. Gabriel, and J. W. Bartlett. 1962.
Phenotypic, genetic, and environmental relationships of milk production and type
ratings of Holstein cows. J. Dairy Sci. 45:223-232.

Wilcox, C. J., J. A. Curl, J. Roman, A. H. Spurlock, and R. B. Becker. 1966. Life span
and livability of crossbred dairy cattle. J. Dairy Sci. 49:991-994.

Wilcox, C. J., W. A. Krienke, J. M. Wing, H. H. Head., and E. L. Fouts. 1968. Genetic
and environmental influences upon composition of milk. Annual report. In:Min. Ann.
Mtg. Southern Regional Tech. Comm. Genetic methods of improving dairy cattle for
the South. S-49 Project. University of Arkansas. p 39. Fayetteville. Arkansas.

Wilcox, C. J., S. N. Gaunt, and B. R. Farthing. 1971. Genetic interrelationships of milk
composition and yield. Agricultural Experiment Stations Bull. 155. Institute of Food
and Agricultural Sciences.University of Florida, Gainesville.

Zhang, W. C., J. C. M. Dekkers, G. Banos, and E. B. Burnside. 1994. Adjustment factors
and genetic evaluation for somatic cell score and relationships with other traits of
Canadian Holsteins. J. Dairy Sci. 77:659-665.

Florida Agricultural Experiment Station, Institute of Food and Agricultural Sciences, University of Florida, Richard L. Jones, Dean for
Research, publishes this information to further programs and related activities, available to all persons regardless of race, color, age,
sex, handicap or national origin. Information about alternate formats is available from Educational Media and Services, University of
Florida, PO Box 110810, Gainesville, FL 32611-0810. This information was published April 1999 as Bulletin (Tech.) 907, Florida
Agricultural Experiment Station.
ISSN 0096-607X

	Front Cover
	Front Matter
	Table of Contents
	Abstract
	Main
	Back Cover