Y
Pi. I: ‘

I 1. r5153

S‘whuur...‘

5....
. u... x.
. an?

 

 

o... A
8 ll}:
.1

.\\«
2;...1

.. . 31...?

I433. .u. Tc;

flbugtz i.

.
t 1..., u

r 3.1).;
x .5.
11- ‘

 

 

 

 

 

 

van:

br‘. \(5.

.3 1.3.”.

;

 

“‘filfis

 

 

 

 

 

 

 

LIBRARY lllallllzlllill’iwwfljmwflflul

 

Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

Reiationship of Body Condition Score With
Milk Yield and Conception in Dairy Cattle

presented by

Joseph John Domecq

has been accepted towards fulfillment
of the requirements for

Ph.DL degree in mil—Ssjfince

We KW~

V Major professor

Date '2 “Mow

 

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

PLACE N RETURN BOX to remove this Mont horn your record.
To AVOID FINES mum on or More out. due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU lo An Affirmative Adlai/Equal Opportunity lnotituion
Wanna-o1

RELATIONSHIP OF BODY CONDITION SCORE WITH MILK YIELD
AND CONCEPTION IN DAIRY CATTLE

By

Joseph John Domecq

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Animal Science

1996

ABSTRACT

RELATIONSHIP OF BODY CONDITION SCORE WITH MILK YIELD AND
CONCEPTION IN DAIRY CATTLE

By

Joseph John Domecq

The goal of this project was to investigate the relationships between body condition
scores, milk yield and conception in high producing Holstein dairy cows. Three
specific objectives were investigated: 1) determine if body condition scores are a valid
and consistent measure of subcutaneous fat, 2) determine the relationship between
body condition scores and milk yield, and 3) determine the relationship between body
condition scores and conception to first service. Data was collected from a
commercial dairy farm (1,000 cows) during an 18 month long period. All multiparous
cows entered the study at dry-off, while all primiparous cows began the study at
parturition. Cows were assigned a body condition score weekly by one individual
throughout the study, continuing until 120 DIM. Milk yield, health, and reproductive
data was also collected. Ultrasound measurements of subcutaneous fat were compared
to body condition scores of 50 cows evaluated during three separate sessions.
Ultrasounnd measurements validated that body condition scores were valid and
consistent measures of subcutaneous fat. Lower body condition scores at dry-off, and
a gain of condition during the dry period were associated with increased milk yield in

the first 120 days of lactation. A loss of body condition in the first month of lactation

was associated with a decreased likelihood of conception. A lower body condition
score at dry-off and increasing body condition during the dry period were associated
with an increased likelihood of conception. Results of this research suggest that body
condition scores are more strongly associated with milk production and the probability
of conception than the occurence of health disorders or other management risk factors.
This research confirmed that body condition scores during the dry period and early

lactation are an important dairy herd monitoring tool.

ACKNOWLEDGMENTS

There are many individuals who have contributed to my life and this work in
the last five years, and in many ways an acknowledgment page in a dissertation seems
so inadequate. My major professor, Dr. Andy Skidmore, provided me the opportunity
to chase many ideas and find my own way. He encouraged my involvement in many
other activities such as teaching, judging and advising students which allowed me to
become a well-rounded individual. Drs. Jim Lloyd and John Kaneene provided
guidance and support above and beyond the call of duty. No question was too stupid
and much of their statistical expertise became a cornerstone of this project. Drs. Herb
Bucholtz and Roy Fogwell made sure that the statistics in the project were combined
with biology and were available to answer any question. Probably more important
than the guidance the committee provided were the friendships that were developed in
the process.

There are several Animal Science faculty members and fellow graduate
students who need to be recognized and were an important part of this project. Many
contributed scientifically and assisted in the massive data collection process, but more
important to me is their friendship. Those individuals include: Dr. Ted Ferris, Dr.
John Shelle, Pat Hady, Steve Mooney, Michelle O'Malley, Mary Beth Roe, and Rick

Dado. A special thanks to Frank Amdt, who contributed in more ways than he will

iv

ever know. Frank probably made the most trips to collect data with me, and during
those trips every possible topic was discussed, including science. Many of his ideas
were incorporated into the project. More than anything else, the friendship we
developed will always be cherished.

Dr. Clifford Jump, Director of the Agricultural Technology program, has had a
tremendous impact on my life. I hope to emulate his ideas and enthusiasm for
teaching and working with young people. He has been a true mentor and friend.
Many Ag Tech Dairy Management students who assisted me during the past three
years, also need to be recognized. They provided help in data collection and were
great early morning entertainment. Many of the questions they asked became a part
of the project.

The owners and employees of Halbert Dairy provided valuable assistance and
without their cows and patience this project would never have been completed. They
allowed visits to their farm every Friday for over a year and collected extra data for
the project.

My mother and father created the enthusiasm I have for the dairy industry and
they are the primary reason I am what I am today. They never stopped persuading me
to be all I could be and supported me in all my crazy efforts, including graduate
school. My brother and sister also provided much support.

Finally, my wife Denise, who has the patience of a saint and is the most
important part of my life, deserves much of the credit for this work. She has seen me

at my worst and best, and still loves me.

TABLE OF CONTENTS

LIST OF TABLES .............................................

LIST OF FIGURES .............................................

LIST OF ABBREVIATIONS ......................................

INTRODUCTION ..............................................

CHAPTER 1 ..................................................

Literature Review .........................................

BODY CONDITION SCORING TECHNIQUES ..............
VALIDATION OF BODY CONDITION SCORES ............
ENERGY BALANCE AND MILK PRODUCTION ...........
BODY CONDITION SCORE AND MILK YIELD ............
ENERGY BALANCE AND CONCEPTION ................
BODY CONDITION SCORE AND REPRODUCTIVE
PERFORMANCE ..............................
BODY CONDITION SCORE AND COW HEALTH ..........
AREAS FOR FURTHER RESEARCH ....................

CHAPTER 2 .................................................

Validation of Body Condition Scores with Ultrasound Measurements of

Subcutaneous Fat of Dairy Cows ........................

ABSTRACT .......................................
INTRODUCTION ...................................
MATERIALS AND METHODS .........................

Data Acquisition ...............................

Statistical Analysis .............................
RESULTS ........................................
DISCUSSION .....................................
CONCLUSIONS ....................................

vi

xi

15

l9

l9

19
20
22
22
23
24

CHAPTER 3 .................................................

Relationship of Body Condition Scores and Milk Yield in High

APPENDIX

CHAPTER 4

Producing Holstein Dairy Cows .........................

ABSTRACT .......................................
INTRODUCTION ...................................
MATERIALS AND METHODS .........................
Cow Selection ................................

BCS .......................................

Milk Yield ...................................
Health Disorders ...............................
STATISTICAL ANALYSIS ............................
BCS and Milk Yield at 120 DIM ...................
Principal Component Analysis .....................

BCS and MYA ................................
RESULTS ........................................
Milk Yield ...................................

BCS .......................................
Regression Analysis ............................
Principal Component Analysis .....................
DISCUSSION .....................................
MYA ......................................

BCS .......................................
Regression Analysis ............................

‘ Principal Component Analysis .....................
SUMMARY .......................................

ooooooooooooooooooooooooooooooooooooooooooooooooo

oooooooooooooooooooooooooooooooooooooooooooooooo

Relationship of Body Condition Score with Conception to First Service

in High Yielding Holstein Dairy Cows ....................

ABSTRACT .......................................
INTRODUCTION ...................................

vii

34

34

34
35
37

39
40
4o
42
43
43
44

45
47

72

72
73

MATERIALS AND METHODS ......................... 75

Cow Selection ................................ 75

Body Condition Scores .......................... 76

Milk Production ............................... 76

Conception to First Service ....................... 77

Health Disorders ............................... 78

Statistical Analysis ............................. 78

Principal Component Analysis ..................... 80

RESULTS ........................................ 81
DISCUSSION ..................................... 83
CONCLUSION ..................................... 87

CHAPTER 5 ................................................. 92
Summary .............................................. 92
RESEARCH GOAL AND HYPOTHESES ................. 92

FINDINGS ........................................ 92

GENERAL DISCUSSION ............................. 94

PRACTICAL IMPLICATIONS ......................... 97

LIST OF REFERENCES ........................................ 100

viii

LIST OF TABLES
CHAPTER 2

Table 1. Number of cows, means, standard deviations, minimum, and maximum body
condition scores and ultrasound measurements ................... 31

Table 2. Pearson correlation coefficients for ultrasound measurements of
specific locations ......................................... 32

Table 3. Comparsion of regression models investigating the relationship and

validity of body condition scores with ultrasound measurements ........ 33
CHAPTER 3
Table 1. Definitions of health disorders .............................. 59

Table 2. Number of cows, means, standard deviations, and ranges of total milk
yield at 120 DIM by lactation ................................ 60

Table 3. Mean and SD of regression coefficients (MY A) obtained by
regressing DIM on daily milk yields for the first 15 DIM by lactation
number ............................................... 61

Table 4. Parameter estimates, standard errors, and probabilities of independent

variables in regression analysis for multiparous cows on total milk yield
at 120 DIM ............................................ 62

Table 5. Parameter estimates, standard errors, and probabilities of independent

variables in regression analysis for primiparous cows on total milk yield
at 120 DIM ............................................ 63

Table 6. Parameter estimates, standard errors, and probabilities of independent
variables in regression analysis for multiparous cows on milk-yield-
acceleration in the first 15 d of lactation ........................ 64

Table 7. Parameter estimates, standard errors, and probabilities of independent

variables in regression analysis for primiparous cows on milk-yield-
acceleration in the first 15 d of lactation ........................ 65

ix

Table 8. Parameter estimates obtained from principal component analysis,
standard deviation, and rank of independent variables for milk yield at
120 DIM of multiparous cows ............................... 66

Table 9. Parameter estimates obtained from principal component analysis,
standard deviation, and rank of independent variables for milk yield at
120 DIM of primiparous cows ............................... 67

Table 10. Parameter estimates obtained from principal component analysis,
standard deviation, and rank of independent variables for MYA of
multiparous cows ........................................ 68

Table 11. Parameter estimates obtained from principal component analysis,
standard deviation, and rank of independent variables for MYA of
primiparous cows ........................................ 69

APPENDIX

Table 1. Correlation coefficients between regression model independent variables for
multiparous cows ........................................ 70

Table 2. Correlation coefficients between regression model independent variables for

primiparous cows ........................................ 71
CHAPTER 4
Table 1. Definitions of health disorders .............................. 88
Table 2. Incidence of health disorders of multiparous cows ................ 89

Table 3. Parameter estimates, odds ratios and 95% confidence interval from
logistic regression analysis of multiparous cows ................... 90

Table 4. Parameter estimates obtained after principal component analysis,
standard deviations, odds ratios, and rank of independent variables for
multiparous cows ........................................ 91

LIST OF FIGURES
CHAPTER 2

Figure l. Scatter plot of body condition scores and mean ultrasound measurement of
right and left lumbar area ................................... 30

CHAPTER 3

Figure 1. Mean BCS at specific points of time during the dry period and
lactation for multiparous and primiparous cows ................... 58

xi

LIST OF ABBREVIATIONS

x = Mean.

p = Mean.

ADBCS = Change in body condition score during dry period.
APBCS = Change in body condition score in first month of lactation.
% = Percent.

BCS = Body condition score.

bST = Bovine somatotropin.

cells/ml = Cells per milliliter.

CFS = Conception to first service.

cm = Centimeter

d = Day.
DA = Displaced abomasum.
DD = Days dry.

DD2 = Days dry squared.

DBCS = Body condition score at dry-off.
DIM = Days in milk.

DYST = Dystocia.

FCM = Fat corrected milk.

GNi = Group number.

hr = Hour.
IND = Index.
k = Estimate.
kg = Kilogram.

kg/yr = Kilogram per year.
LAC = Lactation.

LLUM = Left lumbar.

LTHD = Left tail head.

LTRL = Left thurl.

LUM = Mean lumbar measurement.
MET = Metritis.

ME = Milk fever.

MHZ = Megahertz.

mo = Month.

MYA = Milk yield acceleration.
n = Number.

no. = Number.

xii

NDD = New days dry.

p = Probability.

PBCS = Body condition score at parturition.
PCA = Principal component analysis.
R2 = Coefficient of determination.
RLUM = Right lumbar.

RP = Retained placenta.

RTHD = Right tail head.

RTRL = Right thurl.

SD = Standard deviation.

SE = Standard error.

SPR = Spring.

SUM = Summer.

TRL = Mean thurl measurement.
UM,- = Ultrasound measurement.

US = United States.

WIN = Winter.

wk = Week.

Yijk = Body condition score.

xiii

INTRODUCTION

Dairy cows experience a period of negative energy balance in early lactation
when the energy demands for body maintenance and milk production are not being
met by the diet. To meet the energy demands of milk production, cows will utilize
body fat and protein as an energy source. As milk production decreases during
lactation, the demand for energy becomes less, and lost body fat and protein will be
replaced.

Monitoring and understanding the relationship between milk production, feed
intake, and mobilzation of body reserves is needed when formulating management
decisions for early lactation cows. Monitoring milk production of an individual cow is
usually conducted on modern dairies. However, monitoring an individual cow's feed
intake is not typically done on most large dairy farms. Methods to monitor and
evaluate body condition could give insight to the energy status of cows. Body
condition scores (BCS) are a subjective visual evaluation of the amount of body fat of
cows and can be used to monitor negative energy balance as well as the return to
positive energy balance in later lactation.

Results of previous studies have suggested that BCS are associated with the
nutrition, reproduction, milk yield, and health of cows. Much of the research has been

done in the United Kingdom, making inference of results to the United States difficult.

2

The United Kingdom dairy industry utilizes different feedstuffs and management
practices than the United States. Data from dairy cows producing greater than 10,500
kg of milk annually are also limited. The magnitude and duration of negative energy
balance may be greater with higher production levels. If this is true, there is a need
for BCS research of high producing dairy cows in the United States. Current United
States BCS studies utilize small numbers of cows with relatively low production levels
and provide conflicting results.

Changes in BCS during the dry and early lactation periods may be more
important to production and reproduction than absolute BCS. Body condition score
changes during these periods reflect the energy status of cows. Previous research
indicates that the absolute BCS at dry-off or parturition may not be as important to
milk production as changes in BCS. However, extremes in BCS, cows which are too
thin or too fat at parturition, have been associated with health and production
problems.

Body condition loss in early lactation has been associated with reproductive
problems and conception failure in some studies, while other studies are inconclusive.
Energy balance during early lactation may alter the cows' hormonal status, which in
turn may delay the initiation of estrous cycles after parturition. Follicular and luteal
function may also be negatively altered, which could contribute to reproductive and
conception failure. By utilizing and monitoring BCS in early lactation, management

practices could be implemented to improve reproductive performance.

3

Because of the possible relationships between BCS, milk production,
reproduction, and health, previous statistical procedures may not have yielded reliable
results. If correlations with other variables are not reduced, a reliable association of
one variable with another cannot be determined. By utilizing statistical procedures
that reduce collinearity, or the association between independent variables in statistical
models, not only can the contribution of a specific variable to a particular outcome be
determined, but the relative importance of that variable in comparison to all other
variables measured can be determined without the effects and problems of collinearity.

Charts to aid in assigning BCS have been developed and are widely used.
However, because of the subjective nature of BCS, the validity and consistency of
BCS have been questioned and may limit the usefulness of BCS, particularly in
interpretation of research results and comparison of BCS among cows, herds, and
scorers. Techniques for validating BCS are needed.

The goal of this research project was to determine relationships between BCS,
milk production, and conception to first service in Holstein dairy cows producing
greater than 10,500 kg of milk annually. Specific aims were:

1) Determine if BCS are a valid and consistent measure of body fat.

2) Determine if BCS, or changes in BCS, during the dry period and early
lactation are associated with milk yield or the rate of increase in milk yield.

3) Determine if BCS, or changes in BCS, during the dry period and early
lactation are associated with first service conception.

4

A review of literature is presented in Chapter 1. The BCS validation
procedure is reported in Chapter 2. The investigation of the association between BCS
and milk yield is presented in Chapter 3. Chapter 4 includes the investigation of BCS
and conception. A summary of results from the research project is presented in

Chapter 5.

CHAPTERI

literature Review

BODY CONDITION SCORING TECHNIQUES

Body condition scores (BCS) represent a subjective visual or tactile evaluation
of the amount of subcutaneous fat in a cow. Lowman et al. (43) developed a BCS
technique for dairy cattle. It was based upon experience with sheep, which allowed
for assessment of subcutaneous fat in a simple, quick, and repeatable method. A BCS
between 0 (thin) and 5 (obese) was assigned by feeling the amount of subcutaneous fat
over the spinous processes and around the tail head. Gamsworthy (24) reviewed the
methods and techniques for body condition scoring across species of livestock and
regions of the world. Scores typically range from 0 (thin) to 12 (obese), depending
on the system of evaluation.

Wildman et al. (66) developed a system by which cows were assigned a BCS
ranging from 1 (thin) to 5 (obese) based on a subjective tactile and visual evaluation
of the tail head and lumbar regions of the cow. Braun et al. (4) proposed the addition
of tenth-, quarter-, or half-point increments to the l to 5 range to better characterize
the amount of subcutaneous fat carried by cows. More recently, Edmonson et al. (15)
proposed a system by which cows were visually evaluated at the lumbar, thurl and

tailhead regions and an overall score from 1 to 5 was determined. A chart was

6

developed to aid in assigning BCS to specific body regions. The lumbar, rump, and
tail head regions were evaluated to determine an overall BCS utilizing quarter point
increments. A chart was developed that included descriptions for 8 specific body
locations within the lumbar, rump, and tail head regions.

Ferguson et al. (20) clarified and simplified the BCS system of Edmonson et al.
(15) and developed a decision chart for assigning BCS based upon simple descriptions
of several specific body regions, including the thurls, lumbar vertebrae, coccygeal
ligament and sacral ligament. The simple descriptions developed for the seven body
regions were sufficient to categorize cows by quarter point increments between a BCS
of 2.25 and 4.0. Half-point increments were needed for BCS less than 2.00 and
greater than 4.0.

VALIDATION OF BODY CONDITION SCORES

Gamsworthy and Topp (26) reported a range of correlations from .60 to .76
between ultrasound measurements of subcutaneous fat over the 10th and 13th ribs and
BCS. Gamsworthy and Jones (25) reported similar correlations. These correlations
suggest that BCS are positively associated with the actual amount of subcutaneous fat.

According to a review by Gamsworthy (24) the original scale developed by
Lowman et al. (43) in 1973 was from 1 to 4. However, the Lowman scale (43) was
revised in 1976 and ranged from 0 to 5, thus there is some confusion as to what
scoring scale was used in United Kingdom research. Regardless of the scoring system
used Gamsworthy (24) concluded that BCS were positively correlated with

subcutaneous fat.

7

Neilson et al. (49) used a series of ultrasound measurements over the 10th and
13th ribs and 3rd lumbar vertebra to develop a backfat area index for each cow. The
backfat area index was based upon the length and width of the subcutaneous fat layer
over the area. Correlations for BCS and the backfat area index were over .80.

Ultrasound has been used extensively in other species of livestock to measure
subcutaneous fat and carcass characteristics. Houghton and Turlington (35) reviewed
the use and accuracy of ultrasound in swine, sheep, and beef cattle. Correlations
from .45 to .94 for ultrasound measurements of subcutaneous fat at various body
regions to actual measurements obtained at slaughter indicated that ultrasound
measurements were associated with subcutaneous fat. Faulkner et al. (18) concluded
that ultrasound was an accurate method of measuring 12th rib thickness in beef cattle
based upon comparison to ultrasound measurements obtained after slaughter.

Otto et al. (51) investigated the relationship of BCS and composition of the
9th to 11th rib of dairy cattle and found that BCS was an accurate indicator of the fat
composition of a cow. The ether extract content of the 9th to 11th rib section was
highly correlated with BCS. They reported that a one point change in BCS was
associated with a 56 kg change in live weight. Wright and Russel (67, 68) reported
similar results.

The consistency and repeatability of BCS within and across scorers were
investigated by Edmonson et al. (15). Body condition score did not differ between
scorers with various levels of training during a single scoring session, indicating that

BCS could be repeated consistently. All scorers utilized the chart developed by

8

Edmonson et al. (15) and received some training prior to the study. Ferguson et al.
(20) also investigated the repeatability of BCS across scorers and concluded that BCS
were repeatable. Four scorers independently assigned BCS to 225 cows and agreed
with the modal BCS calculated for each cow 58% of the time, or were a quarter point
different from the modal BCS 33% of the time. The BCS obtained in both studies
were compared between scorers but not to any objective measurements of
subcutaneous fat. Because all scores were obtained during one scoring session,
consistency of BCS between scorers and within scorers across time was not
investigated.
ENERGY BALANCE AND MILK PRODUCTION

After parturition, dairy cows increase milk production. Peak milk production
occurs between 6 and 8 weeks after parturition (6, 21, 24, 32). After peak production,
milk production declines at a rate of 2.5% per week for most cows (32). Dry matter
intake, which is an indicator of appetite, also begins to slowly increase after
parturition, but maximum feed intake occurs 4 to 8 weeks after peak milk yield (6, 21,
24, 32). During this time period, cows experience negative energy balance where the
amount of energy needed to support body maintenance and milk production exceeds
the amount of dietary energy consumed (6, 24, 32). Several researchers indicated that
80% of cows experience negative energy balance in early lactation (6, 64).

The nadir of negative energy balance occurs 1 to 2 weeks after parturition (6).
Energy intake becomes greater than energy output and a positive energy balance

occurs 3 to 4 months after parturition (6). Cows meet the energy demands of body

9

maintenance through different combinations of increased energy intake, mobilization of
body fat, and as a last resort, by decreasing milk production (6, 24, 32, 46). As
energy intake exceeds energy output in mid to late lactation, body reserves lost in
early lactation will be replaced.

Gamsworthy (24) offered an explanation for the difference in the rate of
increase in milk production and feed intake in early lactation. Mammals generally
have a regulatory mechanism which reduces feed intake as body reserves (fat) increase
in order to prevent excessive fatness. Cows that are fatter at parturition have lower
feed intakes and take longer to reach maximum feed intake than thinner cows (24, 26).
Cows may have excessive body reserves at parturition, thus decreasing feed intake at a
time in early lactation when feed intake should be increased, leading to the state of
negative energy balance. Therefore, cows must mobilze body fat to maintain
production during this period of reduced feed intake.

Ferguson and Otto (21) concluded that there is a minimum BCS needed at
parturition to support peak milk yield because the increase in feed intake is slower
than the increase in milk production in early lactation. If feed intake is not ad libitum,
the need for body reserves becomes more important.

BODY CONDITION SCORE AND MILK YIELD

Much of the information concerning BCS and milk yield is based upon studies
conducted in the United Kingdom utilizing small numbers of cows. Also the breeds of
cows, management strategies and BCS scoring scales were different from those found

in the US. These factors make inference to conditions in the United States difficult.

Ill

 

10

Frood and Croxton (23) indicated that cows with a BCS of 2 or less at
parturition produced 400 kg more milk during their lactation than cows with a BCS
greater than 2.0. Several studies (3, 37, 41) found no significant relationship between
BCS at parturition and daily milk yields. Results of a study by Gamsworthy and
Topps (26) and research from Treacher et al. (62) suggest a negative relationship
between BCS at parturition and mean daily milk yield in early lactation. Cows with
lower BCS generally produced more milk than cows with higher BCS. A positive
association between BCS and milk production has also been shown to exist; cows with
higher BCS at calving produced more milk (42). In a review of United Kingdom
research, Gamsworthy (24) suggested that the relationship between BCS at parturition
and milk yield was variable and concluded that other factors must be involved. Cows
with a higher BCS at calving generally lost more body condition during lactation,
which negatively influenced milk yield.

Conclusions of more recent studies investigating BCS and milk yield in the
United States are also variable. Holter et al. (34) reported no relationship between
BCS at calving and mean daily milk yields in the first 12 weeks of lactation for 30
cows. Daily milk yields or cumulative milk yields at 80 DIM between cows calving
above or below a BCS of 3.5 (five-point scale) did not differ in a 66 cow study (55).
Mean daily milk production was not influenced by the amount of condition loss. The
amount of BCS loss was related to BCS at calving. Ferguson and Otto (21) reported
no significant influence of BCS at parturition on milk yield of 1300 cows on one dairy

farm. Results from Italy (52) indicated that BCS at parturition is not important to total

ll

milk yield. However, losses in BCS decreased peak production and the persistency
of the lactation curve. Waltner et al. (65) suggested that both BCS at parturition and
the change in BCS during lactation were related to the total milk yield of 3.5% FCM
at 90 DIM. Ruegg (54) reported no differences in 305 d FCM or peak milk yield
among 429 cows in 13 Canadian herds based upon BCS at parturition. Based upon
the conclusions of these studies, the relationship between BCS at parturition and milk
yield is unclear.

The BCS during the dry period prior to parturition may be important.
However, the influence of BCS at dry-off or during the prepartum dry period has not
been extensively researched, and may have a more direct influence on changes in BCS
and milk production. This may be due to the reduced feed intake and greater losses of
BCS in early lactation associated with BCS at parturition. The influence of obesity, or
overconditioning, has been investigated (22, 27, 47, 63) and suggested to be a problem
in production, reproduction and health of dairy cows.

ENERGY BALANCE AND CONCEPTION

Butler and Smith (6) reviewed studies conducted prior to 1987 and concluded
that higher milk yield may have a negative influence on conception. More recently,
Nebel and McGilliard (48) concluded that higher levels of milk production were
associated negatively with reproductive performance, but indicated that data from cows
producing greater than 13,000 kg/yr is limited. Hillers et al. (33) concluded that
conception was not influenced by production and that it is difficult to separate the

confounding effects of management from the possible biological effects of higher

12

levels of milk yield. They further suggested that dry matter intake and mobilization of
body fat may be more important to conception and reproductive performance than milk
yield.

In a review of the literature, Butler and Smith (6) concluded that negative
energy balance was directly correlated to the postpartum interval to first ovulation and
lower conception rates. Their study found that the number of estrous cycles is reduced
prior to insemination because of a delay in the initiation of estrous cycles during a
state of negative energy balance. Fewer estrous cycles prior to insemination may be
related to conception failure (6). The follicle may not be as responsive to luteinizing
hormone or other gonadotropins because of reduced amounts of glucose and insulin
during the state of negative energy balance. Butler and Smith stated that "once estrous
cycles were initiated, they continue regardless of energy balance". They concluded
that negative energy balance may have reduced the pulsatile secretion of luteinizing
hormone which is needed for ovarian activity. Butler et al. (5) reported that energy
balance during the first 20 d of lactation is important in determining the onset of
ovarian activity following parturition. Canfield and Butler (7) concluded that negative
energy balance influences the initiation of estrous cycles by reducing luteinizing
hormone pulsatility and responsiveness of the ovary to luteinizing hormone.

Canfield et al. (8) indicated that the correlation was .75 between days to first
ovulation and days to energy balance nadir. As the energy balance becomes less

negative, the hypophyseal-ovarian axis becomes more active followed by ovulation

and formation of a corpus luteum. The mean nadir of negative energy balance

 

13

occurred 15 d postpartum. First postpartum ovulation occurred at 29 d. The increase
in luteinizing hormone pulsatility was also observed by Canfield and Butler (7) around
the energy balance nadir.

Villa-Godoy et al. (64) reported that mean daily negative energy balance over
the first 100 days of lactation was not associated with a delay in the initiation of
estrous cycles postpartum. Mean days to first postpartum ovulation was 23.9,
regardless of mean daily energy balance over the first 100 d of lactation. Cows with
the lowest mean daily energy balance in the first 100 d reached the nadir of negative
energy balance at day 12 postpartum, which was sooner than other cows in the study.
First postpartum ovulation for all cows occurred during a state of negative energy
balance. Mean daily negative energy balance over the first 100 days of lactation is a
longer time period than used by Canfield et al. (7, 8), which may have contributed to
the lack of relationship between energy balance and initiation of estrous cycles seen by
Villa-Godoy (64). Canfield et al. (8) concluded that first ovulation occurred when
most cows are still in negative energy balance, but after the negative energy balance
nadir and during the return to positive energy balance.

Butler and Smith (6) concluded that negative energy balance could be
monitored by BCS and that days to first postpartum ovulation were associated with
BCS. Cows which lost more BCS in the first 5 weeks of lactation required greater
days to first service than cows losing less BCS. Their data support the conclusion of

Canfield et al. (7, 8) that delayed initiation of estrous cycles postpartum may be due

14

to negative energy balance. Cows may regulate the effects of negative energy balance
through the increased mobilization of body fat and increased feed intake.

Negative energy balance may influence the secretion of progesterone by the
corpus luteum. Villa-Godoy (64) indicated that progesterone secretion from the corpus
luteum of the second and third postpartum estrous cycles was adversely affected by
negative energy balance, but the first corpus luteum postpartum was not affected.
During the first and second postpartum estrous cycles, Spicer et al. (61) observed
higher levels of progesterone in cows with positive energy balance than in cows with
negative energy balance.

Based on the research of Lucy et al. (44) energy balance influences the number
of ovarian follicles. An increase in the number of large follicles was associated with
an increase in energy balance. They concluded that an increased number of large
follicles leads to earlier ovulation for cows in higher states of energy balance.

After reviewing literature pertaining to the role of negative energy balance and
reproduction in dairy cattle, Nebel and McGilliard (48) concluded that the magnitude
and duration of negative energy balance is associated with reduced conception rates.
They concluded that negative energy balance alters the secretion of gonadotropin
releasing hormone, which effects the secretion of gonadotropins, which in turn
influences the secretion of progesterone. They also indicated that BCS may be a way
to monitor negative energy balance and make management decisions to improve

reproductive performance.

15
BODY CONDITION SCORE AND REPRODUCTIVE PERFORMANCE

The relationship between BCS and reproductive performance is somewhat
unclear. The reproductive performance of 202 cows was not associated with BCS at
parturition in a study conducted in Italy (52). Butler and Smith (6) reported that first
service conception rate was not different between a group of control cows and a group
of overconditioned cows at parturition. Waltner et al. (65) found no relationship
between days to first service, or services per conception with BCS at parturition. In a
study conducted by Ruegg et al. (56) BCS at parturition was not related to services per
conception for 66 Holstein cows. Ruegg and Milton (57) conducted a multi-herd trial
and found no relationship between BCS at parturition and reproductive performance.

In a review of research from the United Kingdom, Gamsworthy (24) concluded
that BCS at the time of insemination is not related to conception success or failure, but
changes in BCS between calving and insemination may influence conception. Butler
and Smith (6) grouped cows based on BCS loss between parturition and insemination
and reported that losses of one point or two points of BCS resulted in conception rates
of 53 and 17%, respectively. Ferguson and Otto (21) reported similar results; cows
losing one point of BCS conceived at a rate of 34%, while cows losing two points of
condition had a conception rate of 21%. Ruegg and Milton (57) found no significant
differences in days to first breeding or days to conception for cows grouped by
condition loss between parturition and insemination. However, condition loss in early
lactation and increased number of services per conception was identified in this multi-

herd trial. In contrast, Ruegg et al. (56) indicated that loss of BCS in early lactation

16

was not related to services per conception for 66 cows, but cows with a BCS of 3.5 or
greater at parturition, or cows that lost greater than .75 BCS in early lactation required
more days to conceive.
BODY CONDITION SCORE AND COW HEALTH

Van Saun (63) concluded that investigating the relationship between BCS and
health is difficult because health problems are not totally independent events. Several
researchers (9, 10, 16, 17) used path analysis and odds ratios to show that most
periparturient health problems are a series of complex disorders that are related. In
addition, there is a wide variation in how health problems are defined and diagnosed,
thus making comparison of results difficult. Gearhart et al. (27) also concluded that
small sample size and the lack of power to detect significant associations contribute to
the difficulty in determining relationships between BCS and health. Because of these
factors, the relationship between BCS and specific health disorders is not clear.

Gearhart et al. (27) investigated BCS and changes in BCS on health and
concluded that BCS was related to health. Overconditioned cows (BCS > 3.75) at dry-
off were at increased risk for reproductive and lameness problems during lactation.
Cows overconditioned at 30 d postpartum were 290 times more likely to have metritis
compared to cows with average BCS. Retained placenta, metabolic disorders, and
mastitis were not associated with BCS. The authors concluded that the small
occurrence of health problems in the 9 herds included in the study may make it

difficult to identify relationships.

17

In another multi-herd trial, Ruegg and Milton (57) found no relationship
between BCS and health. No significant association between BCS and health were
found to exist in several single herd studies (34, 52, 65). Although overconditioning
has been identified as a risk factor for health problems, the relatively low number of
overconditioned cows in these studies may have contributed to the lack of an
association.

AREAS FOR FURTHER RESEARCH

Body condition scores within and across scorers have been determined to be
consistent during a single scoring session. The influence of time between scoring
sessions on consistency of BCS has not been investigated. Although BCS have been
found to be similar among scorers, ultrasound measurements have not been used to
validate BCS over time. Although BCS were similar among scorers, BCS may not
have been as valid and consistent as ultrasound measurements of subcutaneous fat over
time. The first objective of this research was to validate and determine the
consistency of BCS over time by comparing BCS to ultrasound measurements of
subcuttmeous fat.

Dairy cows experience a period of negative energy balance in early lactation,
when feed intake cannot meet the demands of production. During early lactation a-
cow mobilizes body fat and protein as an energy source to support lactation. As
lactation progresses, milk production decreases, feed intake increases and body fat lost
in early lactation is replaced. Body condition scoring can be used to monitor the

changes in body fat during lactation.

18

Results of research investigating the influence of BCS on milk production are
mixed. Research from the United Kingdom may not be similar to the US because of
different breeds of cows and management conditions. Extremes in BCS are associated
negatively with milk yield. Based upon the studies conducted in the US, BCS at
parturition appears to have little influence on milk yield. However, changes in BCS
early in lactation have been associated with milk yield. There has been very limited
research investigating BCS and milk yield for cows producing over 10,500 kg/yr.

Because of the biological changes experienced by cows, total milk yield at a
specific point in time may not be an adequate method of evaluating the association of
BCS with milk yield. The rate of increase in milk yield may be influenced by BCS
and changes in BCS and may reflect the biological changes experienced by cows. The
second objective of this study was to investigate the associations of BCS and changes
of BCS during the dry period and early lactation with total milk yield and the rate of
increase in milk yield in early lactation.

Associations between BCS and reproductive performance are unclear.
Increased milk yields have a slight negative association with conception. Association
of changes in BCS between parturition and insemination and conception success is
equivocal. Further investigation of the role of BCS and milk yield on conception in
cows producing greater than 10,500 kg/yr is needed. The third objective of this study
was to investigate the associations of BCS and changes of BCS during the dry period

and early lactation with conception to first service.

CHAPTER 2
Validation of Body Condition Scores with Ultrasound Measurements of Subcutaneous
Fat of Dairy Cows
ABSTRACT
This research validated body condition scores with ultrasound measurements of
subcutaneous fat. Fifty Holstein cows were evaluated during three sessions in 1993.
Cows scored during each session were divided into three groups of fifteen or twenty
cows. Body condition scores were assigned by one trained individual utilizing a five-
point (1 = thin, 5 = fat) visual scoring technique. Cows were scored to the nearest
quarter point. Ultrasound measurements of subcutaneous fat were obtained by another
individual at the lumbar, thurl, and tailhead areas of both sides of the cow. Body
condition scores and ultrasound measurements were collected on the same day, but
obtained independently. Correlations between ultrasound measurements were
determined and ranged from .36 to .86. Regression models were developed to validate
the body condition scoring technique across the three cow groups. Group number and
different combinations of ultrasound measurements were independent variables, and

body condition score was the dependent variable. Ultrasound measurements were

significantly associated with body condition scores. The coefficients of determination

for the models ranged from .36 to .65, depending on which ultrasound measurements

19

 

 

20

were included in the model. These results suggest that the body condition scoring
technique used in this study was as valid as ultrasound techniques in measurement of
subcutaneous fat.

INTRODUCTION

Body condition scores (BCS) represent a subjective visual or tactile (or both)

evaluation of the amount of subcutaneous fat on a cow (15, 20, 21, 24, 51, 66).
Methods and techniques for body condition scoring differ across species of livestock
and region of the world and have been reviewed by Gamsworthy (24). Scores
typically range from 0 to 12, depending on the system of evaluation being used by the
scorer. Low scores usually represent thin cows, and higher scores represent cows with
more subcutaneous fat.

Two BCS systems have been developed in the US for dairy cows (15, 66).
Wildman et al. (66) developed a system by which cows were assigned a BCS ranging
from 1 to 5 based on a subjective tactile and visual evaluation of the tailhead and
lumbar regions of the cow. Edmonson et al. (15) proposed a system by which cows

were visually evaluated at the lumbar, thurl and tailhead regions to determine an

overall score from 1 to 5. Ferguson et al. (20) explored and analyzed the work of

Edmonson et al. (15) and Wildman et al. (66), which clarified and expanded the
usefulness of BC S.

Regardless of the scoring system used, the subjectivity, repeatability, and
validity of BCS have been questioned. Studies in the United Kingdom (24, 25, 26,

49) have related BCS to ultrasound measurements of subcutaneous fat at the 9th, 10th,

'.o _'1 a it‘.

 

Ir- -.:.- r -
.|
,.

 

21

or 1 1th rib of dairy cows. Correlations between BCS and ultrasound measurements
ranged from .59 to .81, suggesting that BCS reflected the actual amount of
subcutaneous fat. However, cows in those studies were scored on a four-point scale
and were from breeds not normally found in the US; thus, those results may not
represent the subcutaneous fat level of cows in the US. Further, the 9th, 10th, and
11th rib are not normally evaluated using the BCS techniques developed in the US.
The consistency and repeatability of BCS within and among scorers were
investigated by Edmonson et al. (15). There was no difference in the BCS assigned to
cows between scorers with various levels of training, indicating that BCS could be
repeated consistently. Since the work of Edmonson (15), the use of BCS as a herd
management tool has increased. However, its validity has been questioned by
academia and industry, even though data collected from slaughtered animals suggests
that BCS and carcass fat are related (18, 51).
Because of the increased role of BCS in dairy herd management and the
questions of the validity of BCS, this research investigated the association between
BCS and ultrasound measurements of subcutaneous fat. To achieve this goal, the
following objectives were addressed: 1) to determine the relationship of ultrasound
measurements across specific body locations, 2) to determine the relationship between

BCS and ultrasound measurements of fat at specific body locations of the cow, and 3)

to validate the relationship between ultrasound measurements and overall BCS over
time.

22

MATERIALS AND METHODS
Data Acquisition

In 1993, 50 Holstein cows located on a commercial dairy farm were assigned a
BCS and evaluated via ultrasound. The herd (1,000 cows) was milked three times a
day, and mean milk production was over 10,500 kg /yr. Three different groups of
cows were evaluated in March (group 1), April (group 2), and December (group 3).
Fifteen cows were randomly selected for each group during the first two sessions and
twenty for the group in the third session.

The BCS and ultrasound measurements were obtained independently for the
same cows on the same day. Throughout the study, cows were assigned a BCS by
one individual, utilizing the visual technique developed by Edmonson et al. (15).
Cows were scored to the nearest quarter point.

An Aloka 500 (Corometrics Medical Systems, Inc., Wallingford, CT)
ultrasound machine with a S-MI-Iz linear array transducer was used to determine the
amount of subcutaneous fat at six different locations in three areas on each cow.

Body locations for ultrasound were selected based upon body areas evaluated during
body condition scoring and ease of obtaining and reading ultrasound measurements.

One individual operated the machine and obtained ultrasound measurements
throughout the study. The hair coats of cows were not clipped because cows were
housed in a cold free-stall barn on a commercial dairy. Measurements were obtained

by freezing the image on the screen of the ultrasound machine and then measuring the

layer of subcutaneous fat in the center of the screen to the nearest .10 cm.

23

The first ultrasound location was the lumbar area. The transducer was oriented
parallel to the midline over the transverse processes of the lumbar vertebrae midway
between the midline and ends of the transverse processes. The thurl area, the second
area, was located midway between the tuber coxae (hooks) and the tuber ischiae
(pins), 2 to 3 cm above the greater trochanter of the femur. The third area was near
the tailhead, 4 cm off midline, parallel to the sacral vertebrae, and midway between

the tuber coxae and the tuber ischiae. Ultrasound measurements were obtained and

recorded for the right and left side in each area.

Statistical Analysis

Ultrasound measurements for right and left sides were used to determine a

mean measurement for each area. The mean lumbar, thurl, and tailhead area

measurements were used to calculate an overall mean for the areas. Pearson
correlation procedures were used to determine associations between ultrasound

measurements of the six different locations, and correlations were considered to be

significantly different from zero at P < .05.

The relationship between BCS and ultrasound measurements were investigated

using regression models. To validate BCS with ultrasound measurements, group

number 1, 2, or 3 was included in the models. The following general model was used.

where:

Yijk = BCS,
u = mean,
GNi = group number (i = l, 2, or 3),

UM]- = ultrasound measurement j, and
Eijk = error i j k.

24

Each location was modeled and analyzed separately. Three models were
selected for presentation. Model 1 includes the mean lumbar measurement, Model 2
includes the mean thurl measurement, and Model 3 includes the mean ultrasound

measurement for the lumbar and thurl regions.
RESULTS
A clear ultrasound image for measurement could not be obtained for l cow

with a BCS of 4.5, and data for that cow was omitted from the analysis. The tailhead
location was discarded after the first group because of low correlation with BCS and
difficulty in obtaining a clear ultrasound image for measurement. In addition, a clear
ultrasound image for measurement could not be obtained for 2 cows on both the right

and left sides of the thurl region, so a mean ultrasound measurement could not be

calculated.

Means, standard deviations, and ranges of BCS and ultrasound measurements
are presented in Table 1. The number of cows evaluated with ultrasound at each
location is also indicated. The BCS ranged from 1.00 to 4.25, mean was 2.94, and
standard deviation was .62.

The mean ultrasound measurements of the lumbar for the right and left sides of

the cow were 2.30 and 2.35 cm., respectively. Right and left thurl means were 1.95
and 1.96 cm, and right and left tailhead measurements were 1.71 and 1.43 cm.
Ultrasound measurements ranged from 0 to 4.20 cm across all three areas and both

sides of the cow. Standard deviations of ultrasound measurements were near .5 cm for

25
all six sites.

A scatter plot of BCS and the mean ultrasound measurement of the right

and left lumbar area is shown in Figure 1.

Pearson correlation coefficients for comparison of the six locations evaluated
with ultrasound are shown in Table 2. Correlation coefficients ranged from .36 to .86
and were all significantly different from zero (P < .05). The lowest correlation, .36,
was between the right or left side thurl with the left lumbar. Correlation was highest
between the right and left side thurl measurement, .86. The correlation for the left and
right lumbar was .53. These Pearson correlation coefficients were numerically but not
statistically different from one another (P > .05).

Results from three different regression models are shown in Table 3. The
number of cows included in models varies because of the difficulty in obtaining
ultrasound measurements at various locations for cows in each of the three groups.

The BCS was the dependent variable for all models. Model 1 includes the mean
lumbar measurements for 47 cows and group number 1, 2, or 3 as independent
variables. Model 2 includes the mean thurl measurement for 29 cows and group
number 1 or 2; the independent variables for model 3 were the mean ultrasound
measurements for the lumbar and thurl regions for 29 cows and group number 1 or 2.

Results shown in Table 3 are similar for all combinations of ultrasound

measurements included in the other models analyzed but not in Table 3. The R2
ranged from .36 to .65, depending on the ultrasound measurement included in the
model. All models were significant (P < .05) in explaining the variation in BCS.

Ultrasound measurements from all six locations evaluated were significantly associated

 

 

26

with BCS (P < .05). Means of lumbar and thurl measurements as well as the overall
mean of both locations, were also significant (P < . 05). Group number did not
influence BCS in any model (P > .3).
DISCUSSION
In this study, the lumbar, thurl, and tailhead locations were selected for
evaluation for several reasons. These locations are the general areas that are evaluated
when a cow is assigned a BCS using the technique developed by Edmonson et al.
(15). United Kingdom studies (24, 25, 26, 49) have used ultrasound measurements
from the 9th, 10th, or 11th rib for correlations to BCS. However, this rib area is not
evaluated when a cow is assigned a BCS and may not be the ideal body location for
comparison and validation of BCS with ultrasound measurements.
The range of standard deviations of .49 to .73 cm for all six locations evaluated

suggests that some variation is associated with ultrasound measurement. Houghton

and Turlington (3 5) and Faulkner et al. (18) reported similar standard deviations of
ultrasound measurements of the rib area of beef cattle and concluded that ultrasound
measurements could be altered by the location of the transducer, the angle the

transducer was held to obtain a measurement, and the amount of hair coat. The hair
coat was not clipped in this study because the cows were housed in a free-stall barn,

and the farmer preferred not to have the hair coat clipped. Lack of clipping may have
contributed to the variation in measurements. The layer of subcutaneous fat is not

uniform in a given location and could also contribute to measurement error and

increased standard deviation.

27

Ultrasound measurements have been shown to represent amounts of
subcutaneous and carcass fat in studies with slaughtered beef and dairy cattle (18, 35,
51); correlations as high as .90 were achieved. If BCS are significantly associated
with ultrasound measurements, the BCS technique can be assumed as valid as
ultrasound measurements to quantify the amount of subcutaneous fat carried by a cow.

Results from all regression models indicated that all six locations were
significantly associated with BCS and much of the variation associated with BCS is
explained by the models, indicating that BCS is a valid measure of subcutaneous fat
Similar associations for BCS and ultrasound measurements, which utilized data from
the 9th, 10th or 11th rib, have been reported for dairy cows in the United Kingdom
(24, 25, 26, 49). The R2 ranged from .36 to .65 and did not change when ultrasound
measurements from only one side of the cow were included in the model, suggesting
that only one side or location of the cow needs to be evaluated with ultrasound.

This study was not designed to identify an ideal area or location or a

combination of areas and locations to validate BCS with ultrasound measurements,
but results suggest that one location might be adequate because combining ultrasound
measurements from the lumbar and thurl areas did not improve the R2 of any model.
Variation in BCS that is unexplained by these models might be due to variation in
technique of the ultrasound technician or to variation in BCS technique.

Group number had no effect on BCS in any model, regardless of the location
used as an independent variable. The association between ultrasound measurements

and BCS did not change during the study and was always statistically significant. The

28

technique for obtaining ultrasound measurements was assumed to be consistent for the
three groups. These results suggest that BCS are valid measures of subcutaneous fat
over time and that BCS technique is as valid and as consistent as ultrasound

measurements over time. If the BCS technique was not consistent, group number

would have had a statistically significant effect on the BCS.

A criticism of BCS has been that an individual scorer or group of scorers may
not be accurate in BCS technique over time, suggesting that the variation in BCS over
time may not be due to actual changes in the cow, but to inaccuracies associated with
scorer technique because BCS have not been validated over time with a quantitative
measurement of subcutaneous fat. Comparisons of BCS and ultrasound measurements
in the United Kingdom did not address this question (24, 25, 26, 49). Edmonson et al.
(15) developed a BCS system and indicated that BCS were consistent among scorers
utilizing the BCS technique, which became widely accepted as an important tool for
dairy herd management. The accuracy of BCS over time was not investigated.

Although Edmonson et al. (15) found that BCS were consistent, or precise, among
scorers, BCS may not be an accurate measure of subcutaneous fat of the cow. In

other words, BCS were consistent but may not reflect the actual levels of fat carried

by the cow. Results of this study indicate that BCS is a valid measure of
subcutaneous fat over time.

This study investigated the relationship between BCS and ultrasound and
determined whether BCS would be valid over time. In commercial dairy herds, BCS

are assigned regardless of stage of lactation or milk production. Therefore, in this

29

study, stage of lactation, production, and other management considerations were not
evaluated. The BCS technique should be valid across all cows and all BCS, regardless
of other management factors. This study was not designed to investigate differences
in BCS or ultrasound related to stage of lactation, milk production, or other
management factors that may be important in the relationship between BCS and
ultrasound. A possible limitation of this study was that only 50 cows in one herd
were used. The techniques and models developed in this study can be used by other
individuals to determine the validity of BCS they obtain.
CONCLUSIONS
The relationships between BCS and ultrasound measurements obtained in this
study validate that BCS adequately reflect the subcutaneous fat of cows. The
ultrasound measurements of thurl, lumbar, and tailhead regions had a statistically
significant relationship with BCS, and BCS were as valid as ultrasound over time.
Future investigation can use the approaches developed in this study to validate BCS

from many herds and cows.

30

4.50

 

 

4.00 +
3.50 +
E
0_ 3.00 it i
+- +
L? + :l: :l: + +
8 + + + +
g 2.00 + + j;
E + +1 ++i
c?) 1.50 +
1.00 +
0.50 +
0.00
0 1 2 3 4 5
Body Condition Score

Figure l. Scatter plot of body condition scores and mean ultrasound measurement of
right and left lumbar area.

31

Table 1. Number of cows, means, standard deviations, minimum, and maximum body
condition scores and ultrasound measurements.

 

 

Variable Cows x SD Minimum Maximum
(no.)
Body condition scorel 49 2.94 .62 1.00 4.25
Right lumbar, cm 47 2.30 .64 .30 4.20
Left lumbar, cm 48 2.35 .69 .60 3.90
Right thurl,cm 31 1.95 .49 0 2.50
Left thurl, cm 31 1.96 .54 O 2.70
Right tailhead, cm 15 1.71 .63 .40 2.70
Left tailhead, cm 15 1.43 .73 0 2.70

 

' Body condition score 1 = thin, 5 = obese.

32

Table 2. Pearson correlation coefficients for ultrasound measurements of specific

 

 

locations.

RLUM' LLUM RTRL LTRL RTHD LTHD
RLUM 1.00 .53 .60 .54 .50 .56
LLUM 1.00 .36 .36 .53 .42
RTRL 1.00 .86 .65 .51
LTRL 1.00 .66 .63
RTHD 1.00 .65
LTHD 1.00

 

' RLUM = Right lumbar.
LLUM = Left lumbar.
RTRL = Right thurl.
LTRL = Left thurl.
RTHD = Right tailhead.
LTHD = Left tailhead.

33

Table 3. Comparsion of regression models investigating the relationship and validity of

body condition scores with ultrasound measurements.

 

 

 

 

 

Model
1l 22 33
R2 .36 65 .57
F 8.15 23.62 16.46
p .0002 .0001 .0001
n 47 29 29
Independent
variables
P

GN4 .75 .325 .81

LUM .0001

TRL .0001

IND .0001

 

' BCS = Lumbar plus group number.

2 BCS = Thurl plus group number.

3 BCS = Index plus group number.

4 Group number (1,2, or 3).

LUM = Mean lumbar measurement.

TRL = Mean thurl measurement.

IND = Mean of lumbar and thurl measurement.

CHAPTER 3
Relationship of Body Condition Scores and Milk Yield in High Producing Holstein
Dairy Cows
ABSTRACT
The relationship between body condition and milk yield of dairy cows was
investigated in this study. Holstein cows (n = 779) on a commercial dairy farm were
scored for body condition weekly beginning at dry-off and continuing until 120 DIM.
Multiple linear regression and principal component analysis were used to investigate
relationships. Mean body condition scores were 2.77 and 2.66 at dry-off and
parturition, respectively. Principal component analysis was used to reduce the
collinearity among independent variables, calculate new parameter estimates, and rank
the relationship of each variable with milk yield. Body condition change in the dry
period was ranked first, followed by lactation number, and then body condition score
at dry—off for multiparous cows. Increased body conditidn score of one point between
dry-off and parturition was associated with 545.5 kg more milk in the first 120 DIM
than for cows which did not gain one point of BCS. However, each additional point
of body condition at dry-off was associated with 300 kg less milk at 120 DIM. Cows
with lower body condition at dry-off, and that gained condition in the dry period,

produced more milk in the first 120 DIM and had increased rates of milk yield

34

 

35

compared to cows with more body condition at dry-off and cows which did not gain
condition in the dry period.
INTRODUCTION

Body condition scores (BCS) represent a subjective, visual or tactile evaluation
of the amount of subcutaneous fat on a cow (1, 15, 66). Two BCS systems for US
dairy cattle have been reported in the literature (15, 66). Scores range from 1 (thin)
to 5 (obese), with scoring increments of a tenth-, quarter-, or half-point utilized (3).
Ferguson et al. (20) clarified and expanded both BCS systems' usefulness.

Much of the information relating BCS and milk yield is based upon studies
conducted in the United Kingdom (24). These studies utilized small numbers of cows
of breeds and with management strategies that were different from those in the US,
and were scored on a 4-point scale, which is not used in the US. These factors may
make extrapolation of results to conditions in the US difficult. In a review of United
Kingdom research, Gamsworthy (24) suggested that the relationship between BCS at
parturition (PBCS) with milk yield was variable, and that cows with higher BCS at
calving generally lost more body condition during lactation, which negatively
influenced milk yield.

Fat or overconditioned cows at calving may be at risk for lower yield and
increased reproductive and health problems (2, 22, 27, 47). However, the fat or
overconditioned cows may represent extremes in BCS that are not typically seen in

high producing dairy herds. Two studies (2, 4) have found no difference in health or

reproductive problems based on BCS.

 

 

36

Conclusions of more recent studies investigating BCS and milk yield in the US
are also variable. Daily milk yields or cumulative milk yields of cows above or below
a BCS of 3.5 on a 5-point scale did not differ in a 66 cow study (55). Mean daily
milk yield in early lactation was not influenced by the amount of condition loss in the
first weeks of lactation. Ferguson (19) reported no signficant influence of BCS at
parturition on milk yield of 1300 cows. Reseachers from Italy (52) indicated that BCS
at calving was not important to total milk yield, but change in BCS influenced peak
yield and the shape of the lactation curve. In contrast, Waltner et al. (65) stated that
BCS at parturition and change in BCS during lactation were related to total milk yield
of 3.5% FCM at 90 DIM. Ruegg (54) reported no differences in 305-d FCM milk
yields or peak milk yield among 429 cows in 13 Canadian herds based on BCS at
parturition.

Body condition score at calving appears to have little influence on milk yield.
However, changes in BCS, when related to BCS at calving, influenced milk yield.
The effect of BCS at dry-off (DBCS) or changes in BCS during the prepartum period
were not included in previous studies and could influence milk yield. The rate of
increase in milk yield in early lactation is important to total yield and may more
accurately record the biological changes experienced by cows. The rate of increase in
milk yield may be associated with body condition or changes in body condition.

The goal of this study was to investigate the relationship between BCS and
milk yield in high producing Holstein dairy cows. The research hypothesis tested was

that BCS or changes in BCS during the dry period and early lactation are associated

37

with milk yield. Specific objectives were 1) to describe BCS' and changes in BCS in
the dry period and early lactation; 2) to quantitate and rank the associations of BCS
and changes of BCS during the dry period and early lactation with total milk yield up
to 120 DIM, and 3) to quantitate and rank the associations of BCS with the rate of

increased milk yield between 1 and 15 DIM.

MATERIALS AND METHODS

Cow Selection

A commercial dairy farm ( 1,000 cows) located in south central Michigan was
utilized. The herd was housed in a curtain-sided, six-row, free-stall barn, fed a total
mixed ration, and milked three times a day. Lactating cows were grouped based upon
production and lactation number into four groups. First lactation cows were kept in
the same group throughout their lactation. Dry cows were housed in two groups: an
early dry group consisting of cows from dry-off to 3 wk prior to expected calving and

a close-up group from 3 wk prior to calving.

Cows entered the study at dry-off. All cows that were dried off during 1993
were included. All heifers that calved for the first time between February 15, 1993,
and February 15, 1994, entered the study prior to first parturition. Data collection on

all cows terminated at 120 DIM or when they were culled from the herd.

BCS

All BCS was done by one individual utilizing the visual technique developed

by Edmonson et al. (15). Cows were scored to the nearest quarter point. The BCS

38

technique of the individual who scored cows in this study was validated with
ultrasound and was consistent throughout the study (13).

Body condition of cows was scored weekly without scorer knowledge of prior
BCS. The DBCS was obtained 1 d after dry-off. Cows were scored weekly
throughout the dry period. The PBCS was determined by utilizing the BCS closest to
parturition. The BCS at 30 and 15 (1 prior to parturition were determined by utilizing
the BCS closest to 30 and 15 d prepartum. All BCS obtained during the study were
not more than 3 d from any specific time or event, including parturition or specific
DIM.

Typically, heifers were available for BCS 2 to 3 wk prior to expected
parturition. A BCS was assigned weekly during this prepartum period, and a PBCS
was determined in the same manner as described for lactating cows. After parturition,
cows were assigned a BCS during weekly herd visits.

Milk Yield

Daily milk weights for all cows were stored electronically on the farm's
computer. The first 120 DIM were divided into seven time periods to characterize
milk yield acceleration (MY A) and total milk yield. The MYA represents the rate of
milk yield for each time period and may more accurately reflect biological changes
experienced by cows than total milk yield to a specific point in time. There were six

15-d time periods beginning at parturition, and the seventh time period represented 91

to 120 DIM.

39

For each time period, DIM was regressed on daily milk yield. The regression
coefficient obtained in the analysis is referred to as the MYA and represents the rate
of change in milk yield for the respective time period. The regression coefficients
(MY A) were then used to estimate milk yield. Milk yield was estimated for each day
in the time period with the MYA obtained for each specific time period. The
estimated daily milk yields are then added together to obtain total yield for each time
period. This procedure compensates for missing daily milk weights in the first 120
DIM due to missing or malfunctioning electronic transponders, or milk weights not
recorded because cows were missed during milking. Daily milk weights were also
missed as cows moved between production groups, particularly after parturition. Each
time period's total yield was added to obtain total milk yield for 120 DIM. Cows must
have had at least five daily milk weights in a time period for MYA to be calculated
and have had MY A in each of the seven time periods or the cow was omitted from the
analysis.

Health Disorders

All health problems were diagnosed and recorded by herd personnel or
veterinarian. The data were collected during weekly visits to the farm. Each health
problem was coded as occurring or not occurring (no=0; yes=l). Repeat occurrences
of the same health problem in the same cow were recorded but not included for
analysis. Health disorder definitions used are shown in Table 1 and are similar to
those used in previous studies (10, 17, 27). The disorders recorded included dystocia,

twins, retained placenta, metritis, milk fever, displaced abomasum, and lameness.

40
STATISTICAL ANALYSIS

BCS and Milk Yield at 120 DIM

A multiple linear regression model was developed to test the hypothesis that
BCS or changes in BCS during the dry period or early lactation are associated with
milk yield:

Total = intercept + DBCS + ADBCS + APBCS + LAC + WIN + SPR + SUM + DD +
(DD)2 + dystocia + twins + RP + MET + DA + MF + lame + error,

where:

Total = Milk yield at 120 DIM.

DBCS = body condition score at dry-off.

ADBCS = change in body condition during the dry period.

APBCS = change in body condition from parturition to wk 4 of lactation.
LAC = lactation number.

WIN = winter parturition (December, January, or February; yes or no).
SPR = spring parturition (March, April, or May; yes or no).

SUM = summer parturition (June, July, or August; yes or no).

DD = number of days dry.

(DD)2 = number of DD squared.

Dystocia = dystocia (yes or no).

Twins = twins (yes or no).

RP = retained placenta (yes or no).

MET = metritis (yes or no).

DA = displaced abomasum (yes or no).

ME = milk fever (yes or no).

LAME = lameness (yes or no).

Lactation number was included as a continuous variable in the model because a
majority of multiparous cows were second or third lactation. The mean and SD of
lactation number for multiparous cows was 2.75 and 1.15, respectively; 60% of
multiparous cows were second lactation. Milk yield increases linearly between

lactations 2 and 3.

41

Because heifers prior to parturition did not have a BCS equivalent to DBCS
(about 60 d prepartum), primiparous cows were analyzed separately with a similar
model which did not include the DBCS, ADBCS, DD, (DD)2, and LAC variables.
PBCS was included in the model for primiparous cows.

Change in BCS during the dry period was determined by subtracting PBCS
from DBCS; a negative number indicates an increase in BCS from dry-off to
parturition. BCS change during the first month of lactation was calculated by
subtracting the BCS at wk 4 of lactation from the PBCS; a positive number indicates a
decrease in BCS from parturition to wk 4 of lactation.

Calving season was considered categorical and based on the month of
parturition. December, January, and February were winter months; March, April, and
May were spring; summer months were June, July, and August; and fall months
included September, October, and November. By its omission from the model, fall
was the basis for comparison of seasonal effects.

Sire and genetic information was unavailable for over 25% of multiparous cows
completing the study. Comparison of frequency distributions of cows with or without
sire information indicated significant differences in several model variables, including
milk production and health disorders. Therefore, no genetic information was included
in the model and the number of cows in the analysis was maximized and results less

biased by cow selection.

42
Principal Component Analysis

When there is correlation between any set of independent variables, the true
contribution of each variable to the outcome of interest may not be demonstrated. To
avoid this situation, Pearson correlation coefficients and variance inflation factors were
calculated between the independent variables of the regression models to determine the
presence of collinearity.

Principal component analysis (PCA) (36, 40, 50) was used to remove
collinearity between independent variables in the regression model, quantitate the
association of each independent variable, and rank the variables in order of importance
to milk yield. Principal component analysis has been utilized by several researchers
in epidemiology and animal science (20, 39, 50).

A decision was made, prior to'analysis, to utilize the principal component
vectors, which accounted for at least 70% of variation of the correlation matrix of the
independent variables from the regression analysis. A vector of new parameter
estimates for each independent variable in the original regression model was developed
utilizing the methods of O'Brien et al. (50).

The new parameter estimates were multiplied by the standard deviation of the
variable to rank the estimates on a standardized scale. A standardized scale was
necessary so comparison and rankings could be made regardless of the unit of measure
of the independent variable. The absolute values of the standardized estimates were
then ranked; the standardized parameter estimate with the largest contribution to the

dependent variable was ranked first. The ranking reflected which of the independent

43

variables had the strongest relationship to the dependent variable after reducing
collinearity problems.
BCS and MYA

To test the hypothesis that BCS or changes in BCS during the dry period or
early lactation were associated with MY A, regression models and principal component
analysis procedures similar to previously described models were utililzed. The MYA
between 1 and 15 days in milk was used to represent the rate of increase in milk yield
in early lactation. The MYA was used as the dependent variable in the regression
models. Principal component analysis was used to reduce collinearity, to quantify the
association, and to rank each variable.

RESULTS

A total of 1,048 cows entered the study; 141 cows, or 13.5%, were culled or
died prior to 120 DIM and were not included in the analysis. The BCS were collected
weekly for the 907 cows that completed the study (120 DIM). An additional 70 cows
were omitted from the analysis because of a separate bST study being conducted.
Fifty-eight cows had missing data for production or other traits and were also omitted,
leaving a total of 779 cows in the analysis.

The rolling herd average for 1993 was over 10,500 kg. Mean days open was
107, and services per conception was 1.8. Bulk tank SCC were consistently < 100,000
cell/ml. The mean age at first calving was 24.7 mo. The mean previous DD for
second lactation was 61.3 (SD = 9.14; range = 16 to 112). The third and greater

lactation mean for previous DD was 55.7 (SD = 15.2; range = 9 to 112).

44
Milk Yield

Milk yield at 120 DIM and groups of cows divided by lactation are shown in
Table 2. Third and greater lactations are combined into one group. The number of
cows in each lactation group was 316, 275, and 188 for first, second, and third and
greater lactations, respectively. Mean yields by lactation were 3803.2, 4918.2, and
5205.7 kg; cows in third or greater lactation produced the highest amount of milk in
the first 120 DIM. Range of milk yields at 120 DIM among all three groups was
2145.3 to 6805.5 kg.

Mean MYA for the first 15 DIM are reported in Table 3. The mean rate of
increase for first lactation cows was .48 kg/dz. Cows in second and third or greater
lactation increased at .90 and .80 kg/dz. The mean coefficients of determination were
.32, .40, and .37 for first, second, and third or greater lactation, indicating that DIM
explained between 32 and 40% of the variation in daily milk yield in the first 15 DIM.
The largest MYA for all cows occurred in the first 15 DIM. Daily milk production of
first lactation cows did not begin to decrease until 90 DIM, but for multiparous cows,
daily milk production decreased after 45 DIM.

BCS

Mean BCS at specific times during the dry period and lactation for multiparous
and primiparous cows in the study are shown in Figure 1. For multiparous cows,
mean DBCS was 2.77, and cows gained condition in the early dry period before losing
condition in the last 2 wk of the dry period. Mean PBCS was 2.66, which was lower

than DBCS. The lowest BCS occurred between wk 4 and 8 of lactation; mean BCS

45

increased after wk 8 for multiparous cows. Mean loss of BCS in the lst mo of
lactation was .62. Primiparous cows showed a similar pattern of BCS, but did not
drop as low as multiparous cows.

Regression Analysis

All multiparous cows were combined into one group for the regression analysis
because of the similarity of milk yield and MYA. A total of 463 cows were included.
The model explained 20% of the variation in milk yield at 120 DIM and was
signficant (P < .0001). The DBCS, change in BCS during the dry period, change in
BCS during the lst mo of lactation, lactation number, the linear and curvilinear term
for DD, dystocia, lameness, twins, and winter parturition all had a significant
association with milk yield at 120 DIM (P < .05).

Increased DBCS parameter estimates were associated with lower milk yield. A
curvilinear relationship between DD and milk yield at 120 DIM was identified. A
PBCS lower than DBCS and loss of condition in the first month of lactation was
associated with increased yield. Increasing lactation number was associated with
higher milk yield. Dystocia, lameness, and twins were all related to decreased milk
yield. Cows calving in the winter had increased milk yield compared to cows calving
in the fall.

The results of the regression analysis of milk yield at 120 DIM for first
lactation are shown in Table 5. The model explained 8% of the variation in total milk
yield and was significant (P < .005). A total of 316 primiparous cows were included

in the analysis. Change in condition in the first month of lactation, winter and spring

46

parturition tended to be associated with milk yield at 120 DIM (P < .1). Loss of
condition in the first month was related to increased milk yield, and primiparous cows
that calved in the spring and winter produced more milk than those that calved in the
fall.

Regression analysis results for the model investigating MY A for multiparous
cows are displayed in Table 6. The model explained 13% of the variation in MY A in
the first 15 DIM (P < .0001). The curvilinear term for DD was not statistically
associated with MYA (P > .05) and was not included in the analysis. Days dry,
metritis, lameness, and parturient paresis and summer parturition approached
significance (P < .07). As DD increased, MYA decreased. Cows diagnosed with
metritis or milk fever had increased MYA, and lameness was associated with
decreased MYA.

The results of the regression analysis of MYA for first lactation cows are
shown in Table 7. The model explained 11% of the variation in MYA in the first 15
DIM and was significant (P < .0001). The PBCS, change in condition in the lst mo
of lactation, retained placenta, displaced abomasum, and winter and summer calvings
all had significant associations with MYA (P < .05). Winter calvings and an
increase in PBCS were related to increased MY A. A loss of condition in the lst mo
of lactation, retained placenta, displaced abomasum, and summer calvings were
associated with decreased MYA.

The Pearson correlation coefficients for the independent variables in the

regression models are shown in Appendix Table 1 for multiparous cows, and Appendix

47

Table 2 for primiparous cows. Correlations as high as .60 were acheived, and many
correlations were highly significant (P < .05), indicating a high degree of collinearity.
Variance inflation factors approached 3.0, which also suggests a high degree of
collinearity. The parameter estimates and standard errors obtained in the regression
analysis may be unreliable because of the high degree of collinearity. The actual, or
true, associations of an independent variable might be masked or biased because of
their association with other variables in the model.

Principal Component Analysis

Because of the significant associations between independent variables based on
Pearson correlation coefficients and variance inflation factors, PCA was used to
calculate new estimates. The linear and curvilinear terms of DD were combined into
one term for PCA. This was done because DD and DD2 were so highly correlated that
a solution could not be obtained with PCA. For these reasons, DD and DD2 were
combined into a single term with Equation [1].

NDD = (DD - k)2 [1]

For the current study, k was estimated by solving the first order conditions of
the regular regression equation with respect to DD. Though the regression coefficients
employed in this manner were admittedly estimated from a data set containing
substantial collinearity, alternative methods were not found to be superior. The
estimate obtained for k (57.78 d) is consistent with both the optimal dry period length

of 60 (1 based on current recommendations (29), and with the mean DD in this herd.

48

The first eight principal components explained 70% of the variation in the
independent variables for multiparous cows, and the first six principal components
were needed for primiparous cows.

Table 8 contains the new parameter estimates, standard deviations of each
independent variable, and rank of the independent variables for milk yield at 120
DIM of multiparous cows. Change in condition in the dry period was ranked first
among the independent variables; a one point increase in PBCS over the DBCS was
associated with 545 kg more milk in the first 120 DIM. Lactation number was ranked
second, and each additional lactation was associated with 135 kg more milk. The
DBCS was ranked third; an additional point of BCS at dry-off was related to 300 kg
less milk in the first 120 DIM. From Equation [1], the highest milk yield at 120 DIM
in this herd was observed for 58 DD. A loss of .339 kg of milk/d2 above or below 58
d was observed. Cows with 48 or 68 DD produced 33.9 kg less milk in 120 DIM.
This DD variable was ranked fourth among the independent variables. A one-point
loss of BCS between parturition and wk 4 of lactation was associated with 242 kg
more milk in the first 120 DIM of lactation and was ranked fifth among the
independent variables. The sixth ranked variable was winter calving, which was
associated with 288 kg more milk, as compared to fall calving. Health disorders
generally ranked lowest among the independent variables.

Results of PCA of milk yield at 120 DIM for primiparous cows are shown in
Table 9. The variable ranked first was change in BCS in the first month of lactation.

Primiparous cows that lost one point of BCS in the lst m0 of lactation produced 162

49

kg more milk than did primiparous cows that did not lose any condition. The PBCS
ranked second; increased PBCS was associated with 92 kg more milk. Primiparous
cows calving in the summer ranked third among independent variables; they produced
191 kg less milk than cows calving in the fall. Health disorders generally ranked
lowest among independent variables.

The results of PCA for multiparous cows and MYA are shown in Table 10.
The DD was ranked first among the variables; each additional DD was associated with
.06 kg/d2 decrease in MY A. The DBCS ranked second and was related to a .33 kg/d2
decrease in MY A for every one-point increase in DBCS. Change in BCS during the
dry period ranked third. A one-point increase in body condition during the dry period
was associated with a .32 kg/d2 increase in MYA.

Measures of body condition for primiparous cows ranked the lowest among the
independent variables used to investigate MYA. The rankings and parameter estimates
are presented in Table 11. Ranking first was winter calving, which was associated
with a .25 kg/d2 increase in MY A. Retained placenta, displaced abomasum, and
summer and spring calving were all related to decreased MYA. 4

DISCUSSION
MYA

The MY A concept was initially developed to investigate the possible role of
the rapid increase in milk yield on conception to first service (14). The hypothesis in
a preliminary study (14) was that MY A was associated with reduced conception. An

association between MY A between 15 and 30 DIM and conception was identified,

50

while total milk yield at 120 DIM had no association with conception (14). Based on
the results of this preliminary study, MYA in early lactation could be an important
indicator of cow performance.

As cited in previous studies, the association between BCS and milk yield (24,
54, 55) examined rates of milk yield by comparing daily milk yields. This approach
might not reflect the rate of increase in yield experienced by a cow. The rate of
increase may have a greater influence on the association between BCS and milk yield
than originally thought. In the current study, the largest MYA occurred between 1
and 15 DIM for cows. The correlation between MYA in the first 15 d of lactation and
milk yield at 120 DIM was .23, which was the highest correlation acheived among the
seven MYA with milk yield at 120 DIM.
BCS

Mean BCS for cows in the study, graphed in Figure 1, was lower than current
recommendations for BCS at various points in lactation (l, 15, 21, 66). The BCS
reported in this study are generally lower than those reported in previous work (21, 54,
55, 65, 66) investigating the influence of BCS on milk yield. However, in two
previous studies (30, 31), in which cows in different herds were scored by the same
individual as in the current study, BCS were similar at various points of lactation.

There are several possible reasons for the lower BCS observed in this herd. In
general, this herd has a higher production level than reported in some of the previous
studies (54, 55, 65, 66), which may suggest an increased demand for energy reserves

by high producing cows. Published information is limited regarding optimal BCS for

51

high producing dairy cows. The BCS of this commercial herd may be more typical of
high producing dairy cows milked three times a day. Possibly a herd could have
lower BCS because of poor management, but the high milk yield, excellent
reproductive status, and milk quality suggest high management was maintained in the
herd. A comparison of BCS across high producing herds would be useful.

Recommendations for BCS are that cows maintain condition between dry-off
and parturition (l, 15, 19, 21, 66). Cows in this study gained condition early in the
dry period, but the PBCS was lower than the DBCS because of loss of condition late
in the dry period. The loss of BCS during the late dry period corresponds to the lower
feed intake 2 wk prior to parturition (45, 63) and may indicate the nutritional status of
the cow at this time. Information is limited regarding BCS or changes in BCS during
the dry period of high producing cows.

The mean loss of BCS in early lactation was similar to cows in other studies
(6, 19, 21, 54, 55, 65). However, during this study the lowest BCS was reached
between the 4th and 8th week of lactation, which was earlier than previous studies (19,
21, 65). The milk yield or MYA of cows in this study may necessitate earlier and
more rapid mobilization of energy reserves after parturition. The lower BCS at nadir
than previously reported may suggest that a cow may not have had condition to lose
once the nadir BCS was reached.
Regression Analysis

The models used in the regression analysis were significant predictors of both

milk yield and MY A. However, only 8 to 20% of the variation was explained. Milk

52

yield is influenced by many factors that were not considered by this study. Dry matter
intake and nutrient densities of rations are examples of variables that affect milk yield,
but were not included in these regression models because of the difficulty in data
collection. The genetic potential of the cow may also influence milk yield, but was
not included in the models utilized in this study because of a lack of genetic data.

A significant curvilinear relationship existed between DD and milk yield. No
other independent variables, including DBCS, had a significant curvilinear association
with milk yield. Since there were no other significant curvilinear relationships
identified, curvilinear terms for independent variables were not included in the models
before the final regression analysis was performed.

Independent variables from the regression model were associated significantly
with milk yield at 120 DIM in multiparous cows, but not primiparous cows. Only
three of the variables tended to be associated with milk yield at 120 DIM in first
lactation cows. First lactation cows did not produce as much milk as multiparous
cows and may not have been as predisposed to changes in BCS and health problems.

The regression analysis of MYA indicated that BCS and changes in BCS were
not significantly associated with MYA in multiparous cows. The MYA of multiparous
cows was related to DD and disease. Summer calving was significantly related to a
decreased MYA for all cows. Because of the collinearity of the independent variables,
it is difficult to obtain the true influence of a single variable or draw any conclusions

from traditional regression analysis.

53
Principal Component Analysis

The results of the PCA clarified the relationship of an independent variable to
the dependent variable in the regression model. Using PCA, each independent variable
can be considered independently without the correlation, or interaction, of another
variable or variables. Previous research has suggested collinearity among BCS,
changes in BCS, and health (24, 27, 54, 55). There was collinearity among the
independent variables in this study, as indicated by correlation coefficients and
variance inflation factors, thus reaffirming the need for PCA. Even though an
independent variable may not have had a significant association with the dependent
variable, it was not dropped from the model before PCA. This is because each
variable may or may not have a contribution to milk yield after reducing collinearity,
which may be important.

For multiparous cows, BCS, changes in BCS during the dry period, and DD
were more important than any other variable measured in this herd. The parameter
estimate for BCS change in the dry period changed from 237 in the original regression
analysis to -545 after PCA, indicating that a one-point gain in BCS between dry-off
and calving was associated with 545 kg more milk in 120 DIM. The collinearity
between DBCS, change in BCS in the lst mo of lactation, and other independent
variables in the model was reduced with PCA and a more reliable estimate was

determined. Change in BCS during the dry period is correlated highly to

54
DBCS, PBCS, and DD. The length of DD was also associated negatively with milk

yield at 120 DIM in this study. Short or long DD was associated with lower milk
yield.

These results suggest that multiparous cows in this herd should possibly be
dried off at lower BCS, gain BCS in the dry period, and have a PBCS higher than
DBCS. Cows in this study generally calved with a lower PBCS than DBCS. After
gaining body condition in the dry period, cows have condition to lose to support milk
production in early lactation. Cows that lost one point of condition in early lactation
produced 242 kg more milk. The negative association of DD with milk yield
suggests that cows should not have long or short dry periods, and milk yield at 120
DIM was maximal at 58 DD.

Overconditioning of cows has been suggested to be a problem in several
studies (2, 22, 27, 47). The length of dry period has also been investigated (58, 60,
63), and extremely long or short dry periods have been related to milk yield and health
problems. The cows in this study may have been dried off at a low enough DBCS that
overconditioning (BCS > 4.0) was not a problem. In two studies (1, 2), extra energy
dense diets fed to cows in the dry period did not influence yield or health. The
results of this study indicate that increasing condition in a short period of time in the
dry period (58 (1) could be beneficial. Relatively short dry periods may not allow
cows enough time to regain adequate condition before parturition.

As expected, displaced abomasum, retained placenta, dystocia, lameness, and

twins were associated with less milk produced in 120 DIM. However, metritis and

55

milk fever were associated with more milk in 120 DIM. This may be associated with a
higher genetic potential for milk yield, thus possibly explaining the positive association
with yield. The association of an individual health problem with milk yield was not as
great as the association of BCS with milk yield for cows in this study.

Total milk yield analysis for primiparous cows was similar to multiparous
cows. Change in BCS in the lst mo of lactation had a stronger association with milk
yield than PBCS. First lactation cows that lost more condition were associated with
increased milk yield at 120 DIM. Cows that did not lose condition in early lactation
may not have the genetic potential for higher milk yield. Increasing PBCS by one
point was associated with an increase in yield. An investigation of BCS in the 60 d
prior to first parturition would be useful to determine the effect of gaining condition
during this period.

The PCA of MYA for multiparous cows suggested that DD, DBCS, and change
in BCS in the dry period had the greatest association with MYA, which is different
from the results of the regression analysis. Increasing DBCS by one point was
associated with decreased MYA of .33 kg/dz, and increased condition in the dry period
of one point was related to a .30 kg/d2 increase in MY A. Cows were dried off at
lower BCS than recommended, gained condition in the dry period without detrimental
effects on MYA, and, in fact, increased the rate of milk yield in the first 15 DIM. We
could not find studies relating BCS in the dry period to MYA, or a rate of increase in

milk yield.

56

The MYA was associated positively with retained placenta, metritis, lameness,
milk fever, and twins and was negatively related to dystocia and displaced abomasum.
Most of the diseases ranked near the bottom of the list of independent variables,
indicating that other variables, including BCS, were more important. The low
incidence of disease in this herd may have contributed to this positive association.

Principal component analysis of MYA for primiparous cows indicated that
PBCS and the change in BCS in the lst m0 of lactation were the least associated of
model variables. The changes in BCS during the lst mo of lactation and PBCS had a
significant association with MYA through regression analysis. All variables except
winter calving, dystocia, and twins had a negative association on the MYA of first
lactation cows. Because the MYA of first lactation cows is not as high as for
multiparous cows, the association may not be as strong for BCS and changes in BCS.

Results of this study should be interpreted with caution. This study was
conducted in one herd. Future work should be done across herds to see whether the
relationships established in this herd exist for other herds under different management
conditions. Body condition scores and milk yield reflect the cows' energy status,
which is also related to many factors that were not measured in this study. Total milk
yield in the first 120 DIM was used, which may not consider any long term
associations that BCS or health problems may or may not have on milk yield for an
entire lactation. Additional work is needed to explain the possible biological
mechanisms relating BCS and milk yield. Controlled experiments could determine and

identify possible reasons for the results obtained in this observational study.

57

Based on the results of this study, the hypothesis that BCS or changes in BCS
during the dry period or early lactation are associated with milk production is
accepted. BCS are an important dairy management tool which can be used to monitor
the productivity of dairy cows.

SUMMARY

The BCS and changes in BCS observed in this high producing commercial
dairy herd were typically lower than BCS recommended and reported in other studies.
Regression and principal component analyses indicated that, in this herd, changes in
BCS in the dry period, condition score at dry-off, the length of the dry period, and
changes in BCS in the first month of lactation were more strongly associated with
milk yield than with health problems. An increase in BCS in the dry period was
related to an increase in milk yield and MYA, while an increase in DBCS was
associated with a decrease in milk yield and MYA. Since these findings are contrary
to current recommendations, further studies are needed where a broader range of BCS
can be tested. A dry period length shorter or longer than 58 d was negatively
associated with milk yield and MYA. The parameter estimates generated with standard
multiple regression were unreliable due to the collinearity among independent
variables. Principal component analysis was used to determine more reliable estimates
by reducing the effects of collinearity. Dairy consultants and advisors can use
analytical techniques, such as principal component analysis, to determine which
management factors have the largest influence on milk yield to help make better

management decisions.

58

3.25

3.00

2.75

2.50

2.25

Body Condition Score

2.00

 

Dry-off

 

 

1 .75 .
450 -30 -15 0 14 28 so 84 112

DIM

Figure 1. Mean BCS at specific points of time during the dry period and lactation for
multiparous (I) and primiparous (El) cows.

59

Table 1. Definitions of health disorders.

 

Disorder

Definition

 

Dystocia

Retained placenta

Metritis

Lameness

Displaced abomasum

Milk fever

Any assisted delivery

Retention of fetal membranes > 12 hr

Enlarged uterus during rectal exam and/or

cloudy discharge.

Any abnormality in locomotion

"Pinging" sound upon abdominal auscultation and

surgical correction.

Weakness, cold skin, and favorable

response to calcium therapy.

 

60

Table 2. Number of cows, means, standard deviations, and ranges of total milk yield
at 120 DIM by lactation.

 

 

Lactation Cows x SD Minimum Maximum
(no.) (kg)

1 316 3803.2 1151.8 2145.3 5295.1

2 275 4918.2 1498.8 2272.9 6289.9

2 3 188 5205.7 1540.4 3095.2 6805.5

 

61

Table 3. Mean and SD of regression coefficients (MY A) obtained by regressing DIM
on daily milk yields for the first 15 DIM by lactation number.l

 

 

Lactation x SD
1 .48 63
2 .90 .65
Z 3 .80 95

 

' Mean coefficients of determination for the regression analysis were .32, .40, and .37
for first, second, and third or greater lactation, respectively.

62

Table 4. Parameter estimates, standard errors, and probabilities of independent
variables in regression analysis for multiparous cows on total milk yield at 120 DIM.l

 

 

Parameter
Variable2 estimate SE P
DBCS -3l8.72 91.14 .0005
ADBCS 237.19 113.01 .0364
APBCS 357.06 113.72 .0018
Lactation number 139.09 27.39 .0001
DD 42.99 11.73 .0003
(DD)2 -.37 .09 .0001
Retained placenta 9.39 109.61 .9317
Metritis 59.38 79.21 .4539
Dystocia -272.87 117.89 .0211
Displaced abomasum -420.22 457.59 .3589
Lameness -63 0.04 186.61 .0008
Milk fever 193.97 269.36 .4718
Twins -371.68 175.32 .0346
Winter parturition 398.45 80.19 .0001
Spring parturition 123.56 97.96 .2079
Summer parturition -102.41 85.92 .2339

 

‘ Regression model (P < .0001; R2 = .2063; n = 463).

2 DBCS = Body condition score (BCS) at dry-off, ADBCS = change in BCS
during the dry period (DBCS - PBCS), APBCS = change in BCS during lst
month of lactation (PBCS - BCS at wk 4), DD = days dry.

63

Table 5. Parameter estimates, standard errors, and probabilities of independent
variables in regression analysis for primiparous cows on total milk yield at 120 DIM.1

 

 

Parameter
Variable estimate SE P
PBCS 65.14 102.86 .5270
APBCS 172.72 104.60 .0997
Retained placenta 4.28 121.04 .9718
Metritis 94.68 75.32 .1295
Dystocia -15.32 97.49 .8752
Displaced abomasum 172.07 180.30 .3407
Lameness -400.01 263.12 . 1295
Twins -96.86 378.97 .7984
Winter parturition 143.04 82.02 .0822
Spring parturition 156.63 79.81 .0506
Summer parturition -121.10 83.87 .1498

 

' Regression model (P < .0053; R2 = .0826; n = 316).

2 PBCS = Body condition score (BCS) at parturition, APBCS = change in BCS

during lst month of lactation (PBCS - BCS at wk 4).

64

Table 6. Parameter estimates, standard errors, and probabilities of independent
variables in regression analysis for multiparous cows on milk-yield-acceleration in the
first 15 d of lactation.1

 

 

Parameter
Variable2 estimate SE P
DBCS -.127 .107 .2349
ADBCS -.079 .132 .5566
APBCS -.133 .133 .3204
Lactation number -.033 .031 .2870
DD -.005 .003 .0717
Retained placenta .083 .128 .5187
Metritis .185 .093 .0457
Dystocia -.049 .138 .7234
Displaced abomasum -.928 .536 .0844
Lameness -. 83 3 .218 .0002
Milk fever .571 .315 .0712
Twins -.198 .204 .3332
Winter parturition .109 .093 .2405
Spring parturition .050 .115 .6649
Summer parturition -.306 .101 .0025

 

‘ Regression model (P < .0001; R2 = .1256; n = 463).

2 DBCS = Body condition score (BCS) at dry-off, ADBCS = change in BCS
during the dry period (DBCS - PBCS), APBCS = change in BCS during lst
month of lactation (PBCS - BCS at wk 4), DD = days dry.

65

Table 7. Parameter estimates, standard errors, and probabilities of independent
variables in regression analysis for primiparous cows on milk-yield-acceleration in the
first 15 d of lactation.‘

 

 

Parameter
Variable2 estimate SE P
PBCS .343 .122 .0053
APBCS -.339 .124 .0069
Retained placenta -.245 .144 .0894
Metritis -.097 .090 .2788
Dystocia .115 .116 .3205
Displaced abomasum -.540 .214 .0123
Lameness .136 .313 .6650
Twins .158 .451 .7268
Winter parturition .189 .099 .0534
Spring parturition -.090 .095 .3459
Summer parturition -.228 .100 .0232

 

' Regression model (P < .0001; R2 = .1139; n = 316).

2 PBCS = Body condition score (BCS) at parturition, APBCS = change in BCS

during lst month of lactation (PBCS - BCS at wk 4).

66

Table 8. Parameter estimates obtained from principal component analysis, standard

deviation‘, and rank of independent variables for milk yield at 120 DIM of multiparous
2

cows.

 

 

Parameter
Variable3 estimate SD Rank
ADBCS -545.57 .47 1
Lactation number 135.07 1.17 2
DBCS -300.04 .51 3
(DD - 57.78)2 -.339 395.73 4
APBCS 242.01 .40 5
Winter parturition 288.20 .02 6
Milk fever 1074.10 .005 7
Lameness -539.25 .008 8
Retained placenta -269.90 .014 9
Metritis 147.17 .018 10
Summer parturition -86.92 .018 11
Dystocia -110.80 .012 12
Displaced abomasum -442.19 .003 13
Twins -103.83 .008 14
Spring parturition 29.40 .018 15

 

' SD for each variable.

2 Rankings based on absolute value of standardized parameter estimate
(Parameter estimate * SD).

3 DBCS = Body condition score (BCS) at dry-off, ADBCS = change in BCS
during the dry period (DBCS - PBCS), APBCS = change in BCS during lst
month of lactation (PBCS - BCS at wk 4), DD = days dry.

67

Table 9. Parameter estimates obtained from principal component analysis, standard
deviation', and rank of independent variables for milk yield at 120 DIM of
primiparous cows.2

 

 

Parameter
Variable3 estimate SD Rank
APBCS 162.77 .36 1
PBCS 92.43 .37 2
Summer parturition -191.59 .023 3
Spring parturition 104.40 .025 4
Winter parturition 60.15 .023 5
Metritis 54.35 .023 6
Retained placenta 68.45 .014 7
Lameness -85.26 .006 8
Twins -122.44 .004 9
Dystocia 17.84 .018 10
Displaced abomasum 35.50 .009 11

 

‘ SD for each variable.

2 Rankings based on absolute value of standardized parameter estimate
(Parameter estimate * SD).

3 PBCS = Body condition score (BCS) at parturition, APBCS = change in
BCS during lst month of lactation (PBCS - BCS at wk 4).

68

Table 10. Parameter estimates obtained from principal component analysis, standard
deviation', and rank of independent variables for MYA of multiparous cows.2

 

 

Parameter
Variable3 estimate SD Rank
DD -.06 12.27 1
DBCS -.33 .51 2
ADBCS -.32 .47 3
Lactation number -.07 1.17 4
APBCS -.06 .40 5
Summer parturition -.36 .018 6
Spring parturition .24 .018 7
Winter parturition .19 .02 8
Metritis .20 .018 9
Dystocia -.09 .012 10
Retained placenta .06 .014 11
Displaced abomasum -.28 .003 12
Lameness .02 .008 13
Milk fever .02 .005 14
Twins .01 .008 15

 

' SD for each variable.

2 Rankings based on absolute value of standardized parameter estimate
(Parameter estimate * SD).

3 DBCS = Body condition score (BCS) at dry-off, ADBCS = change in BCS
during the dry period (DBCS - PBCS), APBCS = change in BCS during lst
month of lactation (PBCS - BCS at wk 4), DD = days dry.

69

Table 11. Parameter estimates obtained from principal component analysis, standard
deviation‘, and rank of independent variables for MYA of primiparous cows2.

 

 

Parameter
Variable3 estimate SD ' Rank
Winter parturition .25 .023 1
Retained placenta -.28 .014 2
Displaced abomasum -.33 .009 3
Dystocia .16 .018 4
Summer parturition -.l l .023 5
Metritis -.09 .023 6
Spring parturition -.08 .025 7
Twins .21 .004 8
APBCS -.002 .36 9
PBCS -.001 .37 10
Lameness -.03 .006 1 1

 

‘ SD for each variable.

2 Rankings based on absolute value of standardized parameter estimate
(Parameter estimate * SD).

3 PBCS = Body condition score (BCS) at parturition, APBCS = change in BCS
during lst month of lactation (PBCS - BCS at wk 4).

 

70

009 Nov 5.- mo. no. mo. voo. 5. me: No. mo. 5 w. N0; 0 r. no. m§<-_

 

ooé vo. .5. car «or nor mor mo. war no. :2 N0. N0. mo. “.2
ooé mor No.- 5.- «or No.- nor nor .5. 5. Nb.- vo. Nov (0
oo._. mw. mo. mo. vor mor mo. mv. #9. mo. .5.- V0. PM:
00... on. 5.. 50.- No.- mr. vor mo. N...- 00. 5.- am
ooé mo. vf- mo. moo. 50.- #0. NF.- or. 5.- mz=<F
00.? moor mo.- «0.. no. mo. .50: 5.- nor Fm>0
009 nor N F.- 5. 506 m P. so. or. DO
00% mm: mm.- 5. #0. ON. VN. 23m
00... mm; mo. 0F: mNr v _..- mam
00.? mo. 50.- No.- nor 2.))
ooé .5. NO. 0... OS
ooé 3.- vw. w0mn_<
00. w vm. w0m0<
CO. v wOmO
m2<-_ “.2 (0 km: am wz_>>.r kw>o 00 Saw mam z_>> 0<-_ wOmm< w0m0< 8mm.—

 

.mBOU 323238 n8 333:3 “5555365 Econ. cemmoewoc 50253 $56508 coca—880 ._ 2an

Oman—59:4

71

 

 

 

ooé m r. 5. nor moor mor mor mo. 5. mar 50. m2<-_
00. w 5. no. 5.- Nooo. vor mw. mor woo. :6 <0
00% w F. m P. m r. nor 0N. nor 5.- 500,- km:
00.? vr. m r. nor no. vor Orr 00.- am
ooé mm. mo. #0. vor 0 r; vor wz_E
00. P mor m w. no; no.- mo.- Pm>0
006 Nm.- mm; nor m P. 2 3w
ooé Nmr N f- n f- mam
00% 5.. No. 2.3
ooé Nw. wOma<
009 $09...
wig (D Pmi am wZ_E Pm>0 23w mam z_>> wOma< wOma
.maoo mDOHMQme—Q how m0—Dwd5> udOfiGQQOV—um EVOE COmmmOHWOH fiooguonfl maflowogoonu flown—0:00 .N Q—fim-fi

CHAPTER4

Relationship of Body Condition Score with Conception to First Service in High
Yielding Holstein Dairy Cows

ABSTRACT

This study investigated the relationship between changes in body condition
during the dry period and early lactation and conception to first postpartum AI.
Holstein cows (n=720) on a commercial dairy farm were scored weekly for body
condition beginning at dry-off and continuing until first service. Occurrence of
postpartum diseases was recorded. A multiple logistic regression model was a
significant predictor of conception success or failure for multiparous cows, but not
primiparous cows. Principal component analysis reduced collinearity among
independent variables and allowed the variables to be ranked based upon their
contribution to conception to first insemination. The top three ranking variables were
lactation number, milk yield at 120 d of lactation, and the change in body condition
score between parturition and wk 4 of lactation. Increased milk yield at 120 d of
lactation was associated with an increased likelihood of conception, while decreasing
body condition in the lst mo of lactation was associated with a decreased likelihood of
conception. Health problems were less associated with conception than body condition

or milk yield in this herd. Body condition scores during the dry period and during

72

73

the first 30 d of lactation are an important tool which can be used to identify cows at
risk for conception failure at first service.
INTRODUCTION

After parturition, dairy cows experience a slow increase in DMI, a rapid
increase in milk yield, and an increased mobilization of body condition (6, 24, 27, 48,
63). An estimated 80% of cows experience a state of negative energy balance in early
lactation, where the energy demands for milk yield are not being met by the diet (6,
24, 27, 48, 63). Individual cows respond to negative energy balance through different
combinations of feed intake, mobilization of body condition, and lower milk yields (6,
24, 48). The reproductive performance of cows, particularly the probability of
conception, may be associated negatively with the magnitude and severity of negative
energy balance in early lactation (6, 48).

Body condition scores (BCS) are a management tool to monitor the amount and
mobilization of subcutaneous fat and thus indicate negative energy balance of cows in
early lactation (6, 15, 24, 27, 48). If conception is related to BCS, magnitude and
changes in BCS in early lactation could be used to determine whether nutritional goals
are achieved and to identify potential problems (6, 24, 48).

The BCS of a cow at parturition appears to have no relationship with
conception or inseminations per conception (6, 48, 52, 56, 57). Gamsworthy (24)
suggested that the BCS at parturition or BCS at insemination was not related to
conception. However, the loss in BCS between parturition and insemination may

influence conception negatively. Butler and Smith (6), Ferguson (11), and Ferguson

74

and Otto (12) reported similar results of conception to first service (CFS) for cows
grouped by losses of BCS. Cows that lost 0.5 to 1.0 points had a 53% rate of CFS
compared with 17% for cows that lost >1.0 point of BCS between parturition and
insemination. Ruegg and Milton (57) identified a marginal association (P=.09)
between BCS loss and number of services in a multiherd trial. In another study,
Ruegg et al. (56) indicated that loss of BCS was not related to services per conception.

Several of the previously mentioned studies utilized services per conception to
investigate the relationship between BCS and fertility of cows. The number of
services per conception may not be the best indicator of the role of BCS in the success
or failure of conception, because the magnitude and duration of negative energy
balance, as monitored by BCS, changes during lactation. A more direct method to
evaluate the relationship between BCS and conception was utilized by Butler and
Smith (6), Ferguson (11), and Ferguson and Otto (12); only CFS was considered in
those studies.

Typically, as cows move through lactation, the magnitude and duration of
negative energy balance decreases. If negative energy balance, as monitored by
changes in BCS, is associated with CFS, the association would probably be in early
lactation when energy balance is most negative. Use of BCS to monitor changes in
negative energy balance in the dry period and during the first 30 d of lactation
(peripartum) may be important to CF S.

The association between BCS, or changes in BCS, with CFS has not been

investigated in high yielding ( > 11,000 kg/yr) cows that may have greater risk of

75

negative energy balance and decreased BCS (48). Some of the previous studies did
not control for biological and management factors, such as season, milk yield, and
health disorders, which may also influence conception and bias results.

The goal of this research was to investigate the relationship between BCS and
CF S in a herd of high yielding Holstein dairy cows. The research hypothesis tested
was that BCS or changes in BCS in the dry period or early lactation were associated
with reduced probability of CFS. The specific objective was to determine whether
BCS or changes in BCS during the dry period or early lactation were associated with
reduced likelihood of CFS, controlling for management risk factors and postpartum
health disorders.

MATERIALS AND METHODS

Cow Selection

A commercial dairy farm (1,000 cows) located in south central Michigan was
utilized. Facilities and herd management have been described previously (13).
Multiparous cows entered the study at dry-off. All cows that were dried off during
1993 were included in the study. All heifers that calved for the first time between
February 15, 1993, and February 15, 1994, entered the study prior to first parturition.
Data collection terminated on all cows at 120 d of lactation or at first insemination if
cows were at >120 d of lactation. Culled cows or cows that died before 120 d of

lactation were not included in the analysis.

76
Body Condition Scores

All BCS were assigned by one individual utilizing the visual technique
developed by Edmonson et al. (15). Cows were scored to the nearest quarter point.
The BCS in this study were validated with ultrasound measurements of subcutaneous
fat (12). Cows were scored weekly without the knowledge of previous BCS. Once a
cow entered the study, a BCS was assigned every week. All cows were dried off on
Thursdays, and the BCS at dry-off (DBCS) was obtained the next day. Body
condition scores were assigned weekly throughout the dry period. The BCS at
parturition was determined by utilizing the BCS obtained closest to parturition. The
BCS at first postpartum Al was determined by utilizing the weekly BCS obtained
closest to the AI date.

Nulliparous cows entered the study when available for prepartum body
condition scoring, generally 2 to 3 wks prior to expected parturition. A BCS was
assigned weekly during this prepartum period, and the BCS at parturition was
determined in the same manner as described for lactating cows. After parturition,
primiparous cows were assigned a BCS during weekly herd visits.

Milk Production

The daily milk yield for each cow was recorded electronically and stored via
computer on the farm. The first 120 d of lactation were divided into seven periods to
characterize rates of milk yield and to calculate milk yield acceleration (MYA) and
total milk yield. The MYA represents the rate of milk yield for each time period and

may more accurately reflect biological changes experienced by the cow in early

77

lactation. Associations between MYA and CFS have been identified (14). There were
six 15-d periods beginning at parturition, and the seventh period represented 91 to 120
d of lactation.

For each period, days of lactation were regressed on daily milk yield. The
regression coefficient obtained in the analysis was referred to as the MY A and
represented the rate of change in milk yield for the respective time period. The
regression coefficients (MY A) were then used to estimate milk yield for each day in
the period. The estimated daily milk yields were then combined to obtain total yield
for each period. This procedure compensated for missing daily milk yields due to
missing or malfunctioning electronic transponders, computer failures, or cows not
milked or missed on a given day. Total yield for each period was added to obtain
total milk yield at 120 d of lactation. A cow must have had at least five daily milk
weights in a period for MYA to be calculated and must have had MYA in each of the
seven periods or she was omitted from the analysis.

Conception to First Service

If a cow was confirmed pregnant 35 to 40 d after first service, CFS was
achieved. If a cow received a second service, CFS did not occur. A successful CFS
was coded l, and an unsuccessful CFS was coded 0. Cows identified by herd
personnel as not being eligible for breeding and, therefore not inseminated, were

omitted from the analysis. The number of days to first service also was recorded.

78

Health Disorders

All health disorders were diagnosed and recorded by herd personnel or the herd
veterinarian. The data were collected during weekly visits to the farm. Each health
disorder was coded as occurring or not occurring (no = 0; yes = 1). Additional
occurrences or recordings of the same health problem were not included. The
disorders of interest were dystocia, twins, retained placenta, metritis, milk fever,
displaced abomasum, and lameness. Definitions of health disorders are shown in
Table 1 and are similar to those used in previous studies (9, 10, 16, 27, 28).
Statistical Analysis

A multiple logistic regression model was developed to test the hypothesis that
BCS or changes in BCS during the dry period or early lactation were associated with
reduced probability of CFS. Logistic regression was used because the dependent
variable, CFS, is a discrete variable with only two outcomes. To demonstrate
accurately the influence of BCS and changes in BCS on CFS, specific management
and health disorders known to influence CF S were controlled in the analysis. The

statistical model that was used is shown in Equation [1]:

_ 1
Y_ l+e-f(BCS) [1]

 

where:

Y = Probability of conception to first service,

e = Base of natural log, and

f(BCS) = Intercept + DBCS + ADBCS + APBCS + LAC + TOTAL + MY A +
WIN+SPR+SUM+DD+DYST+TWINS +RP+MET+DA+MF+
LAME + error.

where:

79

DBCS = Body condition score at dry-off,

ADBCS = Change in condition during the dry period,

APBCS = Change in condition from parturition to wk 4 of lactation,
LAC = Lactation number,

TOTAL = Total milk production at 120 DIM,

MYA = Milk yield acceleration between 1 and 15 DIM,

WIN = First service occurring in December, January, or February (yes/no),
SPR = First service occurring in March, April, or May (yes/no),

SUM = First service occurring in June, July, or August (yes/no),

DD = Number of days dry,

DYST = Dystocia (yes/no),
TWINS = Twins (yes/no),

RP = Retained Placenta (yes/no),
MET = Metritis (yes/no),

DA = Displaced Abomasum (yes/no),
ME = Milk Fever (yes/no),

LAME = Lameness (yes/no).

Because heifers prior to parturition were not scored for DBCS, primiparous

cows were analyzed separately with a similar model that did not include the DBCS,

ADBCS, DD, and LAC variables. The BCS at parturition was included in the model

for first lactation cows.

Change in BCS during the dry period was determined by subtracting the BCS

at parturition from DBCS; a negative number indicated an increase in BCS from dry-

off to parturition. Change in BCS during the lst mo of lactation was calculated by

subtracting the BCS at wk 4 of lactation from the BCS at parturition. The change in

BCS between parturition and first service was determined by subtracting the BCS

obtained nearest to first service from the BCS at parturition.

80

Season of first postpartum service was considered categorical. December,
January, and February were winter months; March, April, and May were spring;
summer months were June, July, and August; and fall months included September,
October, and November. By its omission from the model, fall was the basis for
comparison of seasonal effects.

Principal Component Analysis

When any set of independent variables are correlated, the true contribution of
each variable to the outcome of interest may not be demonstrated. To avoid this
situation, Pearson and Spearman correlation coefficients were calculated between the
independent variables of the regression model to detect whether a problem of
intercorrelation (collinearity) existed. Because of the high collinearity found between
the independent variables in the regression model, principal component analysis (PCA)
(36, 39, 40, 50) was used to reduce collinearity, to quantitate the association of each
independent variable with the probability of CF S, and to rank the independent
variables in the model in order of contribution to the variance of CFS.

Principal component analysis has been utilized by several researchers in
veterinary and animal science (20, 39, 50). The PCA techniques utilized in this study
have been described elsewhere (13, 36, 40). However, in this study, a logistic
multiple regression model was developed because of the discrete dependent variable.

Odds ratios are a measure of the strength of association between the
independent and dependent variables and are used extensively in epidemiology (17, 27,

38, 53). Odds ratios are the natural log of the parameter estimate obtained in the

81

logistic regression analysis. An odds ratio of 1.0 suggests no association between the
independent and dependent variables. Standardized odds ratios were calculated for the
continuous independent variables. Standardization of odds ratios for continuous
variables allows for the expression of odds ratios for one SD of change in the
continuous variable.

RESULTS

A total of 720 cows were included in the analysis. The rate of CF S for 426
multiparous cows was 46%, and the rate of CFS for 294 primiparous cows was 47%.
Milk yield, MYA, and BCS of this herd during the current study have been previously
reported (13). Incidence of health disorders for multiparous cows completing the study
are presented in Table 2. The health disorder with the highest incidence rate in this
herd was metritis; 19.9% of multiparous cows in this study were diagnosed with
metritis.

The results of the logistic regression analysis for multiparous cows are shown
in Table 3. The logistic regression model was a highly significant predictor of CFS
for multiparous cows (Chi-Square = 34.241, P = .012).

Although the model was highly significant, only one independent variable
showed a significant association with CF S. In this herd, a cow that received first
postpartum service during the summer was three times more likely to conceive to first
service than were cows that were inseminated during the fall (P = .003). The 95%
confidence interval for the odds ratios was 1.4 to 6.2. No other variables in the model

had a significant statistical association with CFS (P > .05). The 95% confidence

82

interval for the odds ratios also suggest no significant relationship between the
dependent and independent variables, because 1.0 is included in each 95% confidence
interval. There was no association between BCS or changes in BCS during the dry
period or early lactation, and CFS based on the results from the logistic regression
analysis.

For primiparous cows, a similar logistic regression model was not a significant
predictor of CFS (Chi-Square = 13.81, P = .4340). No further analysis was performed
for primiparous cows. There was no relationship between BCS or any other variable
included in the model and CFS.

Pearson and Spearman correlation coefficients between the independent
variables indicated a high degree of association, or collinearity. Correlations were as
high as 0.54, indicating the need for PCA to reduce collinearity of the independent
variables in the logistic regression model for multiparous cows.

Results of the logistic regression after PCA for multiparous cows are presented
in Table 4. Ranking indicates the strength of the association of independent variables
in the regression model with CF S after reducing collinearity. The association between
lactation number and CFS was ranked first, followed by total milk yield at 120 d of
lactation, and change in BCS during the lst mo of lactation.

Loss of BCS in the lst mo of lactation was ranked third among the
independent variables. Cows that lost one standard deviation, or .40 points of BCS in
the lst mo of lactation were 1.17 (1 / odds ratio) times less likely to conceive than

cows not losing any points of BCS in the lst mo of lactation. Cows that lost two

83

standard deviations, or .80 points of BCS in the lst month of lactation were 1.36 times
less likely to conceive. The DBCS was ranked seventh, and gaining BCS in the dry
period was ranked eighth; lower DBCS and an increase in BCS during the dry period
increased the likelihood of CF S. Health disorders ranked lower than measures of
BCS, suggesting that they have less association with CFS than BCS.

DISCUSSION

Change in BCS in the lst mo of lactation was included in the logistic
regression model, because the correlation between change in BCS in the lst mo of
lactation and change in BCS between parturition and first postpartum service was .72
and was significant (P < .05). Negative energy balance is most severe in the first 2
wk of lactation and may influence physiological functions during this time period that
are important to conception (6, 48). The likelihood of CFS may be determined by the
loss of BCS in the lst mo of lactation.

Hillers et al. (33) reported CFS rates of 54% for primiparous cows and 50% for
multiparous cows. The rates of CF S for primiparous (47%) and multiparous cows
(46%) in the current study were similar. Gearhart et al. (27) reported a 44% CFS rate
for cows in a multiherd study. Based on the recommendations developed by Domecq
et al. (1 l), the conception rate observed in this herd would be considered only a slight
problem. Any problems related to reduced CF S in this herd are probably not due to
poor reproductive management.

Body condition score and CFS in primiparous cows in this herd were not

associated. Primiparous cows in this herd might not have experienced the magnitude

84

or duration of negative energy balance in early lactation, as monitored by BCS, to
reduce the probability of CF S.

The logistic regression model for multiparous cows was a signficant predictor
of the probability of conception. However, only one variable was significantly
associated with CF S. Cows inseminated in the summer were 3.02 times more likely to
conceive to first service than were cows inseminated during the fall. However,
because of the collinearity among the independent variables, the results of standard
logistic regression might not be accurate. Relationships of an independent variable
with CFS may be masked or inflated because of the effects of collinearity. More
reliable associations can be established with PCA.

The PCA for multiparous cows indicated that the three measures of body
condition were ranked higher than health disorders and had a greater association with
CFS in multiparous cows. Cows that lost one point of BCS in the lst mo of lactation
were 1.5 times less likely to conceive than cows that did not lose one point of BCS.
In other words, if the CFS for a group of cows that did not lose one point of BCS
was 60%, cows that did lose a point of BCS would have a CFS of 45%. The loss of
BCS for cows in this herd did not influence CFS as much as previously reported.
Ferguson (19) and Ferguson and Otto (21) reported that a cow that lost one point of
BCS between parturition and first service was 2.0 times less likely to conceive at first
service. No relationship was found between the number of services per conception
and the change in BCS between parturition and first service in several studies (56, 57,

65). Although services per conception are related to CFS, the relationship between

85

BCS and CF S was not addressed directly in previous studies. By controlling for
management risk factors and health disorders and by utilizing PCA, the odds ratios
obtained in the current study might estimate more accurately the importance of BCS to
conception.

Lower BCS at dry-off and increasing BCS during the dry period increased the
likelihood of successful CFS for cows in this study. Overconditioning in the dry
period has been related to health and reproductive problems (22, 47, 63), but the cows
in this herd were not overconditioned. Cows in this herd could gain condition in the
dry period without becoming overconditioned.

Higher yielding cows in this herd were more likely to conceive than lower
yielding cows. There may be an antagonistic relationship between milk yield and
reproductive performance (6, 24, 48). However, Shanks et a1. (59) suggested that high
yielding cows may be more reproductively sound because they are healthier than lower
yielding cows. The measures of BCS in the logistic regression model in this study
might have accounted for the effects of negative energy balance that influence
conception and removed the negative association of milk yield with reproductive
performance.

After reducing collinearity with PCA, the association between summer
inseminations and CFS became weaker, suggesting that other factors, including BCS,
were more important to CPS. Conception rates in the summer were generally lower
than for other seasons of the year (29). The summer season might not have been

warm enough to have an influence on CFS in this herd. Any possible negative

86

association between summer inseminations and conception might be seen through
changes in BCS and other management factors and might not be as important to CFS
as previously thought.

The incidence of most health disorders in this herd were similar to and within
ranges of incidences reported in other studies (16, 27, 28). However, the metritis rate
of 20% might have been inflated because of the aggressive diagnosis and treatment of
uterine infections by herd personnel.

Health disorders were ranked last among the independent variables. The lack
of association of health disorders with CFS might be due to several factors. The
degree, or severity, of specific health disorders was not accounted for in this study and
might have a stronger association with CFS.. Repeat occurrences of health disorders
were not accounted for, which might also negatively influence CFS. Health disorders
might have a greater influence on CFS in herds in which prevention and treatment are
not considered as important as they were in this herd. The association of health with
CF S might be observed through the influence of BCS.

Some cautions have to be considered when interpreting these results.
Additional work across high yielding herds is needed to determine whether
associations exist under different management situations. Other variables, such as
DMI and genetic potential for milk yield might be included in the logistic regression
models, because they might have an association with CFS. In herds in which the CFS
rate is very low, BCS might have a stronger negative association with CF S than

observed. The association between BCS and CFS might be different in herds with

87

greater variation in BCS, or changes in BCS, or with overconditioned cows.

Therefore, results supported the hypothesis that BCS and changes in BCS
during the dry period and early lactation of multiparous cows are associated with
reduced probability of CFS. The hypothesis, however, was rejected for primiparous
cows. Lower DBCS and a gain of BCS during the dry period were associated with a
greater probability of CF S, and a loss of BCS in the lst mo of lactation was associated
with a reduced probability of CF S. The nutritional status of cows, particularly energy
balance, during the dry period and during the lst 30 d of lactation, as monitored by
BCS, was more important to CF S than were health disorders or other risk factors.

Body condition scores at dry-off and parturition, and the loss of BCS in the lst
mo of lactation should be monitored. The BCS and changes in BCS can be used to
identify cows at risk for CFS failure so that possible management strategies can be
implemented to improve conception.

CONCLUSION

In this high yielding herd, a loss of BCS during the first 30 d of lactation
contributed to the failure of CF S in multiparous cows more than any other variable
measured. Higher total milk yield at 120 d of lactation was positively associated with
CFS. Higher producing cows were more likely to conceive. Body condition score and
CFS were not associated in primiparous cows. Energy balance during the dry period
and early lactation, as monitored by BCS, was more important to CFS than health
disorders or other risk factors. The BCS at dry-off, parturition, and during the lst mo

of lactation should be monitored and used to identify cows at risk for CFS failure.

88

Table 1. Definitions of health disorders.

 

Disorder

Definition

 

Dystocia

Retained placenta

Metritis

Lameness

Displaced abomasum

Milk fever

Any assisted delivery

Retention of fetal membranes > 12 hr

Cloudy discharge and enlarged uterus during

rectal exam.

Any abnormality in locomotion

"Pinging" sound upon abdominal auscultation

and surgical correction.

Weakness, cold skin, and favorable

response to calcium therapy.

 

89

Table 2. Incidence of health disorders of multiparous cows.

 

 

Health Disorder Incidence,%
Metritis 19.9
Retained Placenta 8.7
Lameness 2.8
Milk Fever 1.4
Displaced Abomasum .5
Twins 2.8

Dystocia 6.5

 

90

Table 3. Parameter estimates, odds ratios and 95% confidence interval from logistic regression analysis
of multiparous cows.

 

Odds Ratios‘

 

 

 

Parameter
Variable2 estimate Discrete Continuous 95% Cl3
DBCS .03 .93 (.73, 1.4)
ADBCS .06 1.01 (.71, 1.4)
APBCS -.37 .86 (.62, 1.2)
Lac number .14 1.18 (.94, 1.4)
120 d Milk
Production .0002 1.14 (.91, 1.4)
MYA .10 1.08 (.86, 1.3)
Days to
First Service -.006 .87 (.70, 1.1)
DD .008 1.10 (.89, 1.3)
Retained
Placenta .12 1.13 (.50, 2.5)
Milk Fever .21 1.24 (.20, 7.3)
Lameness .70 2.01 (.49, 8.2)
Metritis .38 1.46 (.85, 2.5)
Displaced
Abomasum -.ll .89 (.05, 3.7)
Twins -l.01 .36 (.09, 1.4)
Dystocia .21 1.23 (.53, 2.8)
Winter .18 1.19 (.70, 2.0)
Spring .53 1.70 (.93, 3.1)
Summer 1.1 3.02 (1.4, 6.2)

 

‘ Odds ratios for continuous independent variables calculated for one SD change.
2 DBCS = Body condition score at dry-off, ADBCS = Change in body condition during dry period,
APBCS = Change in body condition in first month of lactation, DD = Days dry, MYA = Milk yield

acceleration in first 15 d of lactation.

3 95% Confidence interval for odds ratios.

91

Table 4. Parameter estimates obtained after principal component analysis, standard deviations, odds
ratios, and rank of independent variables for multiparous cows.

 

Odds Ratios‘

 

 

Parameter
Variable2 Estimate SD Discrete Continuous Rank
Lactation number .18 1.15 1.22 1
120 (1 Milk
Production .0002 654.4 1.13 2
APBCS -.39 .40 .86 3
DD .01 12.37 1.13 4
Days to
First Service -.005 22.4 .89 5
MYA .10 .80 1.08 6
DBCS -.14 .51 .93 7
ADBCS .03 .48 1.01 8
Summer .55 .018 1.73 9
Winter .32 .02 .72 10
Metritis .37 .02 1.45 11
Lameness .34 .008 1.40 12
Retained Placenta .15 .014 1.16 13
Twins -.19 .008 .83 14
Milk Fever -.17 .006 .84 15
Dystocia .08 .01 1.09 16
Spring .03 .02 1.03 17
Displaced
Abomasum -. 14 .003 .87 18

 

‘ Odds ratios for continuous independent variables calculated for one SD change.

2 DBCS = Body condition score at dry-off, ADBCS = Change in body condition during dry period,
APBCS = Change in body condition in first month of lactation, DD = Days dry, MYA = Milk yield
acceleration in first 15 d of lactation.

CHAPTER 5
Summary
RESEARCH GOAL AND HYPOTHESES

The goal of this research was to investigate relationships between body
condition scores (BCS), milk production, and conception to first service in high
producing Holstein dairy cows. Body condition scoring has become an important and
accepted dairy herd monitoring and management tool. However, there has been a
relatively small amount of research conducted in the United States investigating the
validity and consistency of BCS, or the relationship of BCS to milk production and
conception in high producing Holstein cows. Three specific hypotheses were
developed and investigated in this study.

1. Are BCS valid and consistent measures of body fat.

2. Are BCS, or changes in BCS, during the dry period and early lactation

associated with total milk yield at 120 DIM and the rate of increase in milk

yield in early lactation.

3. Are BCS, or changes in BCS, during the dry period and early lactation
associated with the probability of conception to first service.

FINDINGS
Ultrasound measurements of subcutaneous fat at the thurl, lumbar, and tailhead
regions were significantly associated with the overall BCS assigned to a cow. The

association between BCS and ultrasound measurements remained constant over time,

92

93

suggesting that BCS were as consistent over time as ultrasound measurements of
subcutaneous fat. These results suggest that body condition scores are a valid and
consistent measure of subcutaneous fat of Holstein dairy cows.

Low BCS at dry-off, an increase in BCS between dry-off and parturition, and
loss of BCS in the first month of lactation were associated with increased milk
production in the first 120 d of lactation for multiparous cows. A one point increase
in BCS during the dry period was associated with 545 kg more milk in the first 120 d
of lactation. The BCS at parturition and loss of BCS in the first month of lactation
were associated with 120 d milk production of primiparous cows. The rate of increase
in milk production in the first 15 d of lactation was also associated with BCS of
multiparous cows. Body condition scores were more important to milk production
than health disorders or other management factors. There wm an association between
BCS and changes in BCS during the dry period and eariy lactation with milk
production.

A one point loss of BCS in the first month of lactation was associated with a
1.5 decreased likelihood of conception to first service in multiparous cows compared
to cows which do not lose one point of BCS. Lower BCS at dry-off and increases in
BCS during the dry period were associated with an increased likelihood of conception
to first service, but were not as important as the loss of BCS in the first month of
lactation. The occurrence of health problems or other risk factors were less associated
with the probability of conception failure or success than BCS. No association could

be identified between BCS and conception in primiparous cows. Therefore, an

94

association was found between BCS and changes in BCS during the dry period and
early lactation with conception to first service for multiparous cows in this herd.
GENERAL DISCUSSION

The techniques and statistical methods developed in this study can be used to
determine the validity and consistency of other scorers. The BCS assigned by other
scorers may not be valid or consistent because of the subjectivity of the BCS
technique. If a scorer is an invalid or inconsistent evaluator of body condition, based
on ultrasound measurements, retraining in proper BCS technique may be necessary.

There may be differences in subcutaneous fat deposits due to stage of lactation,
parity, or milk production which may influence validity and consistency of BCS.
Additional work could include an investigation of the role of stage of lactation or other
factors in assigning and validating BCS. Body condition scores are normally assigned
regardless of stage of lactation, or any other factor, but if differences in BCS and
ultrasound measurements are identified due to stage of lactation or other factors, BCS
techniques may need to be adjusted.

The BCS and ultrasound measurements were obtained in one herd in this study.
The validity and consistency of BCS across herds may need further investigation. The
environment where cows are scored may influence consistency. Factors such as the
amount and type of light and where the scorer stands in relation to a cow when
assigning a score may be negatively associated with validity and consistency.

Additional research, across herds and scorers, may be needed.

95

The BCS in this study were lower than current recommendations. There are
few BCS studies which include as many cows, scored as frequently during the dry
period and early lactation, with milk production over 10,500 kg/yr, so it is difficult to
compare results obtained in this study. The BCS observed in this herd may more
accurately represent the BCS of high producing herds. The BCS of cows in the
current study could be lower than similar cows in other herds due to specific herd
management practices, particularly related to nutrition, grouping, and feed
management. Further investigation of BCS across high producing herds is needed to
determine proper BCS recommendations for high producing cows.

The individual assigning BCS in this study may assign lower BCS than other
individuals utilizing the same BCS scoring chart. Thus, the cows in this study may
have been scored higher by other individuals and may actually be meeting BCS
recommendations. A comparison of BCS assigned by different individuals would
further validate BCS.

There may be a relationship between BCS, milk production, and the genetic
potential of the cow for milk production. Because of the lack of genetic information
for over 25% of the multiparous cows in the research herd, it was difficult to include
any genetic measure in statistical models without decreasing the number of cows in the
study and biasing results. The BCS of cows with greater genetic potential for milk
production may be different than cows with less potential for milk production. Higher
producing cows may have lower BCS than lower producing cows, particularly at dry-

off and parturition. Body condition is needed as an energy source for milk production

96

in early lactation, so cows with higher genetic potential for milk production may store,
mobilize, and utilize body condition more efficiently than lower producing cows.
Future investigations of BCS should include measures of genetic potential, such as a
sire's predicted transmitting ability for milk production.

Further research involving the relationship of feed intake and BCS, particularly
during the dry period, is needed. Cows with a BCS below 2.5 at dry-off may eat more
throughout the dry period, which may lead to gain of BCS during the dry period. The
drop in feed intake observed 2 wks prior to parturition may not be as severe for cows
with lower BCS at dry off. Controlled experiments, where feed intake and dietary
nutrients can be monitored closely, could address the relationship between low BCS
and feed intake during the dry period and future performance. The role of prepartum
BCS in primiparous cows may also be important and needs further investigation.

A loss of BCS in the first month of lactation was associated with a reduced
probability of conception to first service in this herd. The BCS during the dry period
was also important. The loss of BCS, an indicator of negative energy balance, may
alter important biological functions associated with conception. Delayed days to first
postpartum ovulation has been suggested to be a problem with reduced conception
because cows have had fewer estrous cycles before insemination. Controlled studies
investigating the possible biological role of BCS and negative energy balance on
conception should focus on the dry period and first 30 d of lactation. Studies should
include monitoring hormones such as LH, F SH among others, and the number of days

to first ovulation.

97

After a cow has been identified as having a reduced probability of conception
to first service due to BCS loss in the first month of lactation, management practices
to improve the probability of conception could be implemented. Future research
should concentrate on management practices that could improve conception. For
example, hormonal treatments, such as prostaglandinF20t, in cows which lose excessive
amounts of BCS ( > 1.0) in the first month of lactation, may be an effective method to
improve conception.

Health problems had less association with milk production or conception than
measures of BCS in this herd. It may be that the severity or the treatment of a health
disorder influences milk production or conception. Additional work, across herds,
where the diagnosis and treatment of health disorders are not a priority, as compared
to the current herd, may yield different results.

PRACTICAL IMPLICATIONS

A quick and easy technique for monitoring the relationship between feed
intake, milk production, and the mobilization of body fat is necessary, particularly as
dairy herds become larger and monitoring feed intake of individual cows becomes
difficult. Body condition scores are a valid and consistent measure of body fat which
are relatively easy to assign and monitor.

Cows experience many dynamic biological changes during the dry period and
early lactation. Feed intake decreases 2 wks prior to calving and then begins a slow
increase after parturition. Milk production begins to increase immediately after

parturition. To meet the demands of milk production, cows utilize different

 

98

combinations of feed intake and body condition. The results of this study indicate that
a critical period to assign and monitor BCS is the dry period and early lactation. The
BCS at dry-off, the change in BCS during the dry period, and the change in BCS in
the first month of lactation are associated with milk production and conception to first
service.

Current recommendations are a BCS of between 3.0 and 3.5 at dry-off and at
parturition, meaning cows should maintain condition during the dry period. The
results of this study suggest that lower BCS at dry-off and increasing condition during
the dry period were associated with increased production. Lower BCS than currently
recommended at dry-off may not be critical, particularly if cows are allowed to gain
condition during the dry period. Cows with lower BCS may have higher feed intakes
than cows with higher BCS. Decreases in BCS during the dry period and
overconditioning of cows (BCS > 3.5) should be avoided. Therefore, the feed intake
and quality of the ration fed to dry cows are critical. This study indicates that BCS
should be used to monitor the nutritional status of dry cows.

Cows lose BCS in the first month of lactation. The loss of BCS in early
lactation was associated with increased production because cows utilize body fat to
support lactation. However, this loss of BCS was also associated with a decreased
probability of conception to first service. By monitoring BCS loss during the dry
period and early lactation, cows which lose excessive amounts of body condition can
be identified and management practices to improve the probability of conception can

be implemented.

99

Body condition scores are an important dairy herd monitoring and management
tool which should be used by dairy producers, nutritionists, and consultants to enhance
the decision making process. Decisions important to the nutrition, reproduction,
health, and profitablility of dairy cows can be made more effectively utilizing body

condition scores.

I:

LIST OF REFERENCES

LIST OF REFERENCES

Boisclar, Y., D.G. Grieve, O.B. Allen, and RA. Curtis. 1987. Effect of
prepartum energy, body condition, and sodium bicarbonate on health and blood
metabolites of Holstein cows in early lactation. J. Dairy Sci. 70:2280.

Boisclar, Y., D.G. Grieve, J.B. Stowe, O.B. Allen, and GK. Macleod. 1986.
Effect of prepartum energy, body condition, and sodium bicarbonate on
production of cows in early lactation. J. Dairy Sci. 69:2636.

Bourchier, C.P., P.C. Gamsworthy, J.M. Hutchinson, and TA. Benson. 1987.
The relationship between milk yield, body condition, and reproductive
performance in high-yielding dairy cows. Anim. Prod. 44:460(abst.).

Braun, R.K., G.A. Donovan, T.Q. Tran, J.K. Shearer, E.L. Bliss, D.W. Webb,
D.K. Beede, and B. Harris. 1986. Body condition scoring as a herd
management tool. Comp. Contin. Educ. Pract. Vet. 8:F62.

Butler, W.R., R.W. Everett, and CE. Coppock. 1981. The relationship between
energy balance, milk production, and ovulation in postpartum Holstein cows. J.
Anim. Sci. 53:742.

Butler, W.R. and RD. Smith. 1989. Interrelationships between energy balance
and postpartum reproductive function in dairy cattle. J. Dairy Sci. 72:767.

Canfield, R.W. and W.R. Butler. 1990. Energy balance and pulsatile luteinizing

hormone secretion in early postpartum dairy cows. Domest. Anim. Endocrinol.
7:323.

Canfield, R.W., C.J. Sniffen, and W.R. Butler. 1990. Effects of excess
degradable protein on postpartum reproduction in dairy cattle. J. Dairy Sci.
7322342.

Correa, M.T., H. Erb, and J. Scarlett. 1993. Path analysis for seven postpartum
disorders of Holstein cows. J. Dairy Sci. 76:1305.

100

10.

ll.

12.

13.

14.

15.

l6.

17.

18.

19.

20.

21.

101

Curtis, C.R., H.N. Erb, C.J. Sniffen, R.D. Smith, and DS. Kronfeld. 1985.
Path analysis of dry period nutrition, postpartum metabolic and reproductive
disorders, and mastitis in Holstein cows. J. Dairy Sci. 68:2347.

Domecq, J.J., R.L. Nebel, M.L. McGilliard, and AT. Pasquino. 1991. Expert
system for evaluation of reproductive performance and management. J. Dairy
Sci. 74:3446.

Domecq, J.J., A.L. Skidmore, J.W. Lloyd, and J.B. Kaneene. 1995.
Relationship of body condition scores and milk yield in Holstein dairy cows.
(Submitted, J. Dairy Sci.)

Domecq, J.J., A.L. Skidmore, J.W. Lloyd, and J.B. Kaneene. 1995. Validation
of body condition scoring with ultrasound measurements of dairy cattle. J.
Dairy Sci. 78:2308.

Domecq, J.J., A.L. Skidmore, F. Ngwerume, and J.B. Kaneene. 1992. The
effect of milk-yield-acceleration in early lactation on first service conception. J.
Dairy Sci. 74(Suppl.1):266(abstr).

Edmonson, Al, 1.]. Lean, L.D. Weaver, T. Farver, and G. Webster. 1989. A
body condition scoring chart of Holstein dairy cows. J. Dairy Sci. 72:68.

Erb, H.N. and Y.T. Grohn. 1988. Epidemiology of metabolic disorders in the
periparturient dairy cow. J. Dairy Sci. 71:2557.

Erb, H.N., R.D. Smith, P.A. Oltenacu, C.L. Guard, R.B. Hillman, P.A. Powers,
M.C. Smith, and ME. White. 1985. Path model of reproductive disorders and

performance, milk fever, mastitis, milk yield, and culling in Holstein cows. J.
Dairy Sci. 68:3337.

Faulkner, DE, DE Parett, F.K. McKeith, and LL. Berger. 1990. Prediction of
fat cover and carcass composition from live and carcass measurements. J.
Anim. Sci. 68:604.

Ferguson, JD. 1992. Body condition scoring. Advanced Dairy Cattle Nutrition
Seminar. Am. Assoc. Bovine Pract. Conf. Minneapolis, MN.

Ferguson, J.D., D.T. Galligan, and N. Thomsen. 1994. Principal descriptors of
body condition score in Holstein cows. J. Dairy Sci. 77:2695.

Ferguson ID. and K.A. Otto. 1989. Managing body condition in dairy cows.
Page 75 in Proc. Cornell Nutr. Conf. Feed Manuf. Syracuse, NY. Cornell
Univ., Ithaca, NY.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

102

Fronk, T.J., L.H. Schultz, and AR. Hardie. 1980. Effect of dry period
overconditioning on subsequent metabolic disorders and performance of dairy
cows. J. Dairy Sci. 6321080.

Frood, MJ. and D. Croxton. 1978. The use of condition scoring in dairy cows
and its relationship with milk yield and live weight. Anim. Prod. 27:285.

Gamsworthy, PC. 1988. The effect of energy reserves at calving on
performance of dairy cows. Page 157 in Nutrition and Lactation in the Dairy
Cow. Butterworths, London, UK.

Gamsworthy, PC. and GP. Jones. 1987. The influence of body condition at
calving and dietary protein supply on voluntary food intake and performance in
dairy cows. Anim. Prod. 44:347.

Gamsworthy, PC. and JR Topps. 1982. The effect of body condition of dairy
cows at calving on their food intake and performance when given complete
diets. Anim. Prod. 35:113.

Gearhart, M.A., C.R. Curtis, H.N. Erb, R.D. Smith, C]. Sniffen, L.E. Chase,
and MD. Cooper. 1990. Relationship of changes in condition score to cow
health in Holsteins. J. Dairy Sci. 7323132.

Groehn, J.A., J.B. Kaneene, and D. Foster. 1992. Risk factors associated with
lameness in lactating dairy cattle in Michigan. Prev. Vet. Med. 14:77.

Gwazdauskas, EC, 0]. Wicox, and W.W. Thatcher. 1974. Environmental and
managemental factors affecting conception rate in a subtropical climate. J.
Dairy Sci. 58:88.

Hady, P.J., J.J. Domecq, and J.B. Kaneene. 1994. Frequency and precision of
body condition scoring in dairy cattle. J. Dairy Sci. 77:1543.

Hady, P.J., J.J. Domecq, and J.B. Kaneene. 1994. A computer based body
condition management system. Comp. Contin. Educ. Pract. Vet. 1611383.

Haresign, W. 1980. Body condition, milk yield, and reproduction in cattle. Page
1 in Recent Developments in Animal Nutrition. Butterworths, London, UK.

Hillers, J.K., P.L. Senger, R.L. Darlington, and W.N. Fleming. 1984. Effects
of production, season, age of cow, days dry, and days in milk on conception to
first service in large commercial dairy herds. J. Dairy Sci. 67:861.

1""

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

103

Holter, J.B., M.J. Slotnick, H.H. Hayes, C.K. Bozak, W. E. Urban, Jr., and
ML. McGilliard. 1990. Effect of prepartum dietary energy on body condition
score, postpartum energy, nitrogen partitions, and lactation production
responses. J. Dairy Sci. 7323502.

Houghton, PL. and L.M. Turlington. 1992. Application of ultrasound for
feeding and finishing animals: A review. 1992. J. Anim. Sci. 70:930.

Jackson, J.B. 1991. A User's Guide to Principal Components. John Wiley and
Sons, Inc., New York, NY.

Jones, GP. and PC. Gamsworthy. 1989. The effects of dietary energy content
on the response of dairy cows to body condition at calving. Anim. Prod.
49:183.

Kleinbaum, D.G., L.L. Kupper, and KB. Muller. 1988. Applied Regression and
Other Multivariable Methods. Duxbury Press, Belmont, CA.

Lafi, S. 1989. Epidemiologic and economic study of repeat breeder syndrome
in Michigan dairy cattle. PhD thesis. Michigan State University.

Lafi, S. and J.B. Kaneene. 1992. An explanation of the use of principal

component analysis to detect and correct multicollinearity. Prev. Vet. Med.
13:261.

Land, C. and JD. Leaver. 1981. The effect of body condition at calving on the
production of Friesian cows and heifers. Anim. Prod. 32:362(abst.).

Land, C. and JD. Leaver. 1980. The effect of body condition on the milk
production and feed intake of dairy cows. Anim. Prod. 30:449(abst.).

Lowman, B.G., N.A. Scott, and SH. Somerville. 1976. Condition scoring of
cattle. Bull. No. 6. E. Scotland Coll. Agric.

Lucy, M.C., C.R. Staples, F.M. Michel, and W.W. Thatcher. 1991. Energy
balance and size and number of ovarian follicles detected by ultrasonography in
early postpartum dairy cows. J. Dairy Sci. 74:473.

Marquardt, J.P., R.L. Horst, and NA. Jorgenson. 1977. Effect of parity on dry
matter intake at parturition in dairy cattle. J. Dairy Sci. 60:929.

Moe, P.W., H.F. Tyrell, and WP. Flatt. 1971. Energetics of body tissue
mobilization. J. Dairy Sci. 54:548.

 

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

104

Morrow, DA. 1975. Fat Cow Syndrome. J. Dairy Sci. 59:1625.

Nebel, R.L, and ML. McGilliard. 1993. Interactions of high milk yield and
reproductive performance in dairy cows. J. Dairy Sci. 7623257.

Neilson, D.R., C.T. Whittemore, M. Lewis, J.C. Alliston, D.J. Roberts, L.S.
Hodgson-Jones, J. Mills, H. Parkinson, and J.H.D. Prescott. 1983. Production
characteristics of high yielding dairy cows. Anim. Prod. 36:321.

O'Brien, D.J., J.W. Lloyd, and J.B. Kaneene. 1992. A principal component
analysis of factors critical to participation in veterinary lifelong education
programs. J. Vet. Med. Educ. 20:124.

Otto, K.L., J.D. Ferguson, D.G. Fox, and OJ. Sniffen. 1991. Relationship
between body condition score and composition of the ninth to eleventh rib
tissue in Holstein dairy cows. J. Dairy Sci. 74:852.

Z."J’Lm'-.U£ "

Pedron, O., F. Cheli, E. Senatore, D. Baroli, and R. Rizzi. 1993. Effect of body
condition score at calving on performance, some blood parameters, and milk
fatty acid composition in dairy cows. J. Dairy Sci. 76:2528.

 

 

.Mr.“ L! ‘

Rothman, K.J. 1986. Modern Epidemiology. Little, Brown, and Company,
Boston and Toronto.

Ruegg, PL. 1994. Body condition scoring in Holstein dairy cows:
Relationships with production, reproduction, and disease. Page 176 in Proc. of
Amer. Assoc. of Bovine Practitioners. Albuquerque, NM.

Ruegg, P.L., W.J. Goodger, C.A. Holmberg, L.D. Weaver, and EM. Huffman.
1992. Relation among body condition score, milk production, and serum urea
nitrogen and cholesterol concentrations in high-producing Holstein dairy cows
in early lactation. Am. J. Vet. Res. 53:5.

Ruegg, P.L., W.J. Goodger, C.A. Holmberg, L.D. Weaver, and EM. Huffman.
1992. Relation among body condition score, serum urea nitrogen and
cholesterol concentrations, and reproductive performance in high-producing
Holstein dairy cows in early lactation. Am. J. Vet. Res. 53:10.

Ruegg, PL, and R.L. Milton. 1995. Body condition scores of Holstein cows on
Prince Edward Island, Canada: Relationships with yield, reproductive
performance, and disease. J. Dairy Sci. 78:552.

Schaeffer, LR. and CR. Henderson. 1971. Effects of days dry and days open
on Holstein milk production. J. Dairy Sci. 55:107.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

105

Shanks, R.D., A.E. Freeman, and P]. Berger. 1979. Relationship of
reproductive factors with interval and rate of conception. J. Dairy Sci. 62:74.

Sorenson, IT, and C. Enevoldsen. 1991. Effect of dry period length on milk
production in subsequent lactation. J. Dairy Sci. 7421277.

Spicer, L.J., W.B. Tucker, and GD. Adams. 1990. Insulin-like growth factor-I
in dairy cows: Relationship among energy balance, body condition, ovarian
activity, and estrous behavior. J. Dairy Sci. 73:929.

Treacher, R.J., I.M. Reid, and OJ. Roberts. 1986. Effect of body condition at
calving on the health and performance of dairy cows. Anim. Prod. 43:1.

Van Saun, RI 1991. Dry cow nutrition: The key to improving fresh cow
performance. Page 59 in Veterinary Clinics of North America: Food Animal
Practice. Vol. 7. W. B. Saunders Company, Philadelphia, PA.

Villa-Godoy, A., T.L. Hughes, R.S. Emery, L.T. Chapin, and R.L. Fogwell.
1988. Association between energy balance and luteal function in lactating cows.
J. Dairy Sci. 71:1063.

Waltner, S.S., J.P. McNamara, and J.K. Hillers. 1993. Relationships of body
condition score to production variables in high producing Holstein dairy cattle.
J. Dairy Sci. 7623410.

Wildman, E..,E GM. Jones, RE. Wagner, R.L. Boman, H.F. Troutt, and TN.
Lesch. 1982. A dairy cow body condition scoring system and its relationship to
selected production characteristics. J. Dairy Sci. 65:495.

Wright, IA. and A.J.F. Russel. 1984. Partition of fat, body composition, and
body condition scoring in mature cows. Anim. Prod. 38:23.

Wright, IA. and A.J.F. Russel. 1984. Estimation in vivo of the chemical
composition of the bodies of mature cows. Anim. Prod. 38.33.

 

"11111111«1111111: