.r
f»
a
V7

ii»?
; 2.;

3.-....
K w.»
P.5-
i

 

 

155...... I}...
, ¢

()4

LIBRARY
Michigan State
University

This is to certify that the
dissertation entitled

FACTORS INFLUENCING DAIRY CATTLE CULLING
DECISIONS AND THEIR ECONOMIC IMPLICATIONS

presented by

GREGG L. HADLEY

has been accepted towards fulﬁllment
of the requirements for the

Ph. D. degree in Agricultural Economics

 

 

:I?

   
    
 

I2

   

rofessor’ 5 Signature
i 7/ a l/ 0 3

Date

 

MSU is an Afﬁrmative Action/Equal Opportunity Institution

-.-.-_-.- — - -

PLACE IN REFURN BOX to remove this chedtout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 czlﬁC/Dateouelnddpjs

FACTORS INFLUENCING DAIRY CATTLE CULLING DECISIONS
AND THEIR ECONOMIC IIVIPLICATIONS

By

Gregg L. Hadley

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Agricultural Economics

2003

 

ABSTRACT

FACTORS INFLUENCING DAIRY CATTLE CULLING DECISIONS
AND THEIR ECONONIIC IMPLICATIONS

By

Gregg L. Hadley

Dairy cattle culling is the act of removing dairy cattle from a herd and replacing
them with other cows. Dairy cows can be culled from a herd for production or health
reasons. The average annual culling rate describes the percentage of cattle that are culled
from a herd annually. Determining the optimal culling rate can be difﬁcult for a producer.
If the culling rate is too high, farmers fail to earn an adequate return on their cattle
investment. If too low, the farmer forgoes production and genetic improvement.

This dissertation contains three studies on dairy cattle culling. The ﬁrst study
examines how individual cow and farm characteristics as well as market prices affected
the likelihood of a cow being culled on DHIA participating farms in ﬁve Midwestern and
ﬁve Northeastern states during 1993 through 1999. In general, cow attributes such as age,
calving season, breed and production affected culling likelihood. Farm attributes such as
size, expansion and whether the farm raised registered dairy cattle also affected culling
likelihood. Both the milk to feed price and cull cow to replacement heifer price ratio
affected culling likelihood.

Data from the NAHMS ’96 Dairy Survey was used in the second study to

determine what management factors affected the udder and mastitis, lameness and injury,

disease. a

a slgniﬁc.‘
survey F2
farms whc‘
lameness 2
as did fam

programs

In t

the amount
sailed base.
situations, 1,
and lnlun C
mortalities, l
The (
technologies
Wit) Th.
adOpt 0n the l

hOiVe‘ver‘ aﬁ e

their . ~ .
‘eipemn

Proﬁtable in ti

disease, and reproduction culling rates. Overall, very few management programs showed
a signiﬁcant effect on culling rates. This may be due to the cross sectional design of the
survey. Farms with employee handbooks culled less for udder and mastitis problems. On
farms where cattle had access to soft walking surfaces, fewer cows were culled due to
lameness and injury problems. Farms with herd bulls had lower reproduction culling rates
as did farms that used a combination of an employee handbook and employee incentive
programs.

In the third study, a decision support system (DSS) was developed to determine
the amount producers could afford to pay to reduce health culling rates. This amount
varied based upon the underlying culling probabilities, breed and herd size. For most
situations, the health cull reduction type with the greatest potential returns was lameness
and injury culls. It was more proﬁtable for herds with more than 600 cows to reduce
mortalities, however.

The (DSS) was used to evaluate the adoption of two health cull reduction
technologies, a rubberized cattle alleyway floor and gonadotropin releasing hormone
(GNRH). The DSS determined that a rubberized alleyway ﬂoor was not proﬁtable to
adopt on the basis of reducing lameness and injury culls alone. It was proﬁtable to adopt,
however, after including the savings from an overall reduction in lameness episodes and
their respective treatments. Using GNRH to treat cows with cystic ovaries proved

proﬁtable in the DSS estimation.

 

 

Copyright by
GREGG L. HADLEY
2003

 

1 ti
and Dr. St
above and
me when I
dissertatio
place to st;

1 ti.
Sm'astaya
beyond. I l

onanABL

ACKNOWLEGEMENTS

I thank the members of my advisory committee, especially Dr. Christopher Wolf
and Dr. Stephen Harsh, for their guidance in this dissertation. I also thank them for going
above and beyond the call of duty in taking time out of their busy schedules to meet with
me when I was able to get away from work and come back to campus for intense
dissertation work sessions. I also thank Mrs. Harsh for her hospitality in providing me a
place to stay on many of these visits.

I thank my fellow graduate student colleagues - especially George Young, Lorie
Srivastava, and James McQueen -— for their support throughout graduate school and
beyond. I will cherish their friendship forever. I especially thank George who took pity
on an ABD professor and agreed to submit this dissertation for me.

I also thank my colleagues at UW- River Falls and UW- Extension for their
support during the completion of my degree. Special acknowledgement needs to be given
to Dr. Lewis May for teaching my spring semester course and to the members of the UW-
River Falls Agricultural Resource Center Union: Laura Walsh, Paul Kivlin, Brenda
Boetel, and Alycia Aiken. These friends supplied me with moral support, and, in one
case, even a car while completing this dissertation.

I thank my heroes, Mom and Dad, for always encouraging me to further my
education. I also thank my wife Shelly for her love and support throughout the years. Last
but certainly not least, I also thank Beaufort Franklin Hadley for being my constant
companion in completing my dissertation. We can now get back to ﬁshing, hiking, long

walks, and playing in the snow.

LIST OF I

LIST OF E.
CHAPTER
1. 83¢
II. Cull
CHAPTER
I. lntrt
H Date
HI. The
IV. Ate]

Lev
l" Cull:

and]
\I Lact.
“1 With
lIII. COnc
CHAPTER
I Intrm
H. Merl].
Hi Estim
ll',

SUIT!”

TABLE OF CONTENTS

 

 

LIST OF TABLES .................................................................................... x
LIST OF FIGURES ................................................................................. xv
CHAPTER 1. INTRODUCTION .................................................................. 1
I. Background and Motivation ................................................................. 1
II. Culling Rate Terminology ................................................................... 3
CHAPTER 2. CULLING AND LONGEVITY DESCRIPTIVE STATISTICS ............ S
I. Introduction .................................................................................... 5
II. Data Description ............................................................................. 5
III. The Effect of Cow Longevity and Culling on Production .............................. 6
IV. Average Culling Rates for 1993 — 2000 by State, Breed, Production

Level, and Herd Size ........................................................................ 8
V. Culling Reasons for 1993 - 1999 by State, Lactation, Breed, Herd Size

and Production Level ....................................................................... 12
VI. Lactation Speciﬁc Culling Rates for 1993 — 1999 ...................................... 19
VII. Within Lactation Culling Pattern for Health Culls for 1993 — 1999 ................. 27
VIII. Conclusions ................................................................................. 29

CHAPTER 3. EXPLAINING WHY INDIVIDUAL DAIRY COWS ARE
CULLED IN NORTHEASTERN AND MIDWESTERN

DAIRY HERDS .................................................................. 30
I. Introduction ................................................................................. 30
11. Methods and Model Development ....................................................... 30
III. Estimation Results ........................................................................... 48
IV. Summary and Conclusions ............................................................... 71

vi

 

CHAPI

Int
A (
Infc

Den
Cull

The

CHAPTER 4. MODELS TO DETERMINE THE EFFECTS OF

II.

III.

IV.

V.

SELECT MANAGEMENT FACTORS TO DAIRY

CATTLE HEALTH CULLING RATES .................................... 75
Introduction .................................................................................. 75
Data ........................................................................................... 75
Methods and Model Development ......................................................... 76
Estimation Results ......................................................................... 94
Summary and Discussion .................................................................. 102

CHAPTER 5. A CULLING RATE REDUCTION FINANCIAL FEASIBILITY

II.

III.

IV.

V.

VI.

DECISION SUPPORT SYSTEM ............................................. 106
Introduction ............................................................................... 106
A General Overview of the DSS ........................................................ 107
Information Needed to Run the DSS ................................................... 109
Determining the Financial Returns Associated with Reduced Health
Culling Rates .............................................................................. 11 1
The DSS Input Fields ..................................................................... 114
DSS Calculations and Output Fields .................................................... 129

CHAPTER 6. THE FINANCIAL FEASIBILTIY OF DECREASING CULLING
RATES ON DAIRY FARMS ............................................................. 145
Introduction ................................................................................. 145
General Methods and BEA Estimate Categories ....................................... 146

II.

III.

IV.

The BEA Estimates for Health Culling Rate Reductions Among
Midwestern and Northeastern DHIA Dairy Farms ................................... 147

The Estimated BEA for a Health Culling Rate Reduction for
Holstein, Jersey and Guernsey Farms .................................................. 159

The BEA for a Ten Percent Health Culling Rate Reduction for
Farms that Vary by Size .................................................................. 173

vii

III I

ﬁll S
CHAPTE

BIBLIOC

VI. Using the DSS to Estimate the Proﬁtability of Rubberized Alleyway

Surfaces ..................................................................................... 184
VII. Using the DSS to Estimate the Proﬁtability of Gonadotropin

Releasing Hormones ..................................................................... 186
VIII. Summary and Conclusions .............................................................. 190
CHAPTER 7. SUMMARY ..................................................................... 192
BIBLIOGRAPHY ................................................................................. 196

viii

 

 

 

Table 1

Table 2
Table 3.

Table 4
Table 5,
Table 6
Table 7
Table 8

Table 9

Table 10
Table 11
Table 12
Table 13.
Table 14

Table 15

Table 16

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

Table 8.

Table 9.

Table 10.

Table 1 1.

Table 12.

Table 13.

Table 14.

Table 15.

Table 16.

LIST OF TABLES
Estimated “Optimal” Culling Rates of Select Studies .......................... 2

The Effects of Cow Age on Actual 305 Day Milk Production, Milk
Fat Production, Milk Protein Production, and Somatic Cell Count. . . . . . .7

The SCC of Cattle Culled for Speciﬁc Reasons (% Above or Below

Average SCC Level)> .............................................................. 8
Average Culling Rates for 1993 — 2000 ............................. . .......... 10
Culling Reasons by State ......................................................... 13
Culling Reasons by Breeds ....................................................... 17
Culling Reasons By Herd Size ............................................... 18
Culling Reasons By Rolling Herd Average .................................... 19
Lactation Speciﬁc Total Culling Rate1 by State for 1993 — 1999
........................................................................................ 20
Lactation Speciﬁc Health Culling Ratel by State for 1993 — 1999
........................................................................................ 21
Lactation Speciﬁc Mortality Rate by State for 1993 — 1999
....................................................................................... 22
Lactation Speciﬁc Total Cull (car, Health Cull (HR)2 and

Mortality (MR) Rate by Breed for 1993 — 1999 ............ ' .................. 23
Lactation Speciﬁc Total Cull (CR)1, Health Cull (HR)2 and

Mortality (MR) Rate by Herd Size for 1993 — 1999 .......................... 24
Lactation Speciﬁc Total Cull (CR)‘, Health Cull (HR)2 and

Mortality (MR) Rate by Rolling Herd Average for 1993 — 1999 ........... 25
Within Lactation Cow Removal Pattern for Health Culls (H)

and Mortalities (M) by Lactation for All Ten Midwestern and
Northeastern States 1993 — 1998 ............................................... 28
Independent Variable Descriptions for the Probit Model

Estimation Used to Predict the Probability of a Cow

Becoming A Cull ................................................................ 33, 34

Tablel7

Table 18

Table 19

Table 20

Table 21

Table 22

Table 23

 

Table 1 7.

Table 18.

Table 19.

Table 20.

Table 21.

Table 22.

Table 23.

Table 24.

Table 25.

Table 26.

Table 27.

Table 28.

Table 29.

Table 30.

Table 31.

Table 32.

Table 33.

Overall Results for the Probit Model Estimation of the
Likelihood that a Cow Will Be Culled on Midwestern Dairy
Farms ................................................................................ 48

Coefﬁcient Estimate Results for the Probit Model
Estimation of the Likelihood that a Cow Will Be Culled on
Midwestern Dairy Farms .................................................. 50 — 52

Overall Results for the Probit Model Estimation of the
Likelihood that a Cow Will Be Culled on Northeastern Dairy
Farms ................................................................................ 61

Coefﬁcient Estimate Results for the Probit Model Estimation
of the Likelihood that a Cow Will Be Culled on Northeastern
Dairy Farms .................................................................. 63 - 65

The Udder and Mastitis Culling Rate Model Independent
Variable Descriptions .............................................................. 77

The Lameness and Injury Culling Rate Model Independent
Variable Descriptions .............................................................. 83

The Disease Culling Rate Model Independent Variable

Descriptions .................................................................... 88, 89
The Reproduction Culling Rate Model Independent Variable

Descriptions ........................................................................ 92
Overall Results for The Udder and Mastitis Culling Rate Model .......... 94

Parameter Results for The Udder and Mastitis Culling Rate Model ....... 95
Overall Results for The Lameness and Injury Culling Rate Model ........ 96

Parameter Results for The Lameness and Injury Culling Rate Model. . . ..98

Overall Model Results for The Disease Culling Rate Model ................ 99
Parameter Results for The Disease Culling Rate Model ................... 100
Overall Results for The Reproduction Culling Rate Model ............... 101
Parameter Results for The Reproduction Culling Rate Model ............ 102
Information Needed for the DSS .............................................. 110

xi

Table 34

Table 35

Table 36

Table 37

Table 38
Table 39
Table 40.
Table 41
Table 42

Table 43.

Table 44
Table 45
Table 40.

Table 47

Table 34.

Table 35.

Table 36.

Table 37.

Table 38.

Table 39.

Table 40.

Table 41.

Table 42.

Table 43.

Table 44.

Table 45.

Table 46.

Table 47.

The Expected Rolling Herd Average Per Lactation and

Somatic Cell Count Levels Per Lactation Used to Estimate

the BEA for a Health Cull Reduction for the Typical

Midwestern and Northeastern DHIA Farm Estimation

Category ............................................................................. 148

The Effects of Culling on the RHA and Somatic Cell Count
(SCC) Levels of the Culled Cow Used for the BEA Estimation
of a Health Cull Reduction ...................................................... 149

The Average Lactation Speciﬁc Culling Rate For Midwestern
and Northeastern Dairy Farms ................................................... 150

The Within Lactation Removal Schedule for Cattle Culled due
to Non- Death Health Culls Used for the BEA Estimation of a

Health Cull Reduction ........................................................... 151
The Within Lactation Removal Schedule for Deaths Used for

the BEA Estimation of a Health Cull Reduction ............................ 152
Other Critical Production and Financial Factors for the Typical
Midwestern and Northeastern DHIA Dairy Farm Estimation ............. 154
Lactation Expense Adjustment ................................................ 156

The Estimated BEA for Successive 10 Percent Health Cull

Reductions for a 240 Month Period for a 130 Cow Midwestern

or Northeastern DHIA Dairy Farm... . . . . . .... 157
The BEA of a 10 Percent Reduction in Speciﬁc Health Culls

for a 240 Month Period for a 130 Cow Herd with the Midwestern

and Northeastern DHIA Sample Average Health Culling Rate ........... 158
The Expected Rolling Herd Average (RHA) Per Lactation

and Somatic Cell Count Levels Per Lactation Used for the BEA
Estimation of a Health Cull Reduction for All Scenarios .................. 160
The Average Lactation Speciﬁc Culling Rate For Holstein Cattle ........ 161
The Average Lactation Speciﬁc Culling Rate For Jersey Cattle ............. 162
The Average Lactation Speciﬁc Culling Rate For Guernsey Cattle ......163

Other Critical Production and Production Expense Factors
for the Breed Based BEA Estimations ......................................... 165

xii

Table 48

Table 49

Table 50

Table 51

Table 52

Table 53

Table 54
Table 5 5
Table 56
Table 57_

Table 58

Table 59

Table 60

Table 48.

Table 49.

Table 50.

Table 51.

Table 52.

Table 53.

Table 54.

Table 55.

Table 56.

Table 57.

Table 58.

Table 59.

Table 60.

The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 130 Cow Holstein Herd ....... 167

The BEA of a 10 Percent Reduction in Speciﬁc Health Culls
for a 240 Month Period for a 130 Cow Holstein Herd with the
Breed Average Health Culling Rates .......................................... 168

The BEA Associated with Successive 10 Percent Health
Cull Reductions for a 240 Month Period for a 130 Cow
Jersey Herd1 ........................................................................ 169

The BEA for a 10 Percent Reduction in Speciﬁc Health Culls
for a 240 Month Period for a 130 Cow Jersey Herd with the
Breed Average Health Culling Rates .......................................... 170

The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 130 Cow Guernsey
Herd! ............................................................................... 171

The BEA for a 10 Percent Reduction in Speciﬁc Health Culls
for a 240 Month Period for a 130 Cow Guernsey Herd with the
Breed Average Health Culling Rates .......................................... 172

The Average Lactation Speciﬁc Culling Rate For Herds with
Less Than 150 Cows ............................................................. 175

The Average Lactation Speciﬁc Culling Rate For Herds with
300 to 450 Cows ................................................................. 176

The Average Lactation Speciﬁc Culling Rate For Herds with
More Than 600 Cows ............................................................ 177

The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 130 Cow Herd ................... 179

The BEA for a 10 Percent Reduction in Speciﬁc Health Culls
for a 240 Month Period for a 130 Cow Herd with the 0 to 150
Cow Herd Size Average Health Culling Rate. .............................. 180

The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 390 Cow Herd ................... 181

The BEA for a 10 Percent Reduction in Speciﬁc Health

Culls for a 240 Month Period for a 130 Cow Herd with
the 300 to 450 Cow Herd Size Average Health Culling Rate .............. 182

xiii

Table 61.

Table 62

Table 63

Table 64

 

Table 61.

Table 62.

Table 63.

Table 64.

The BEA Associated with Successive 10 Percent Health
Cull Reductions for a 240 Month Period for a 650 Cow Herd ............. 183

The BEA for a 10 Percent Reduction in Speciﬁc Health Culls
for a 240 Month Period for a 650 Cow Herd with the 600 or
More Cow Herd Size Average Health Culling Rate ........................ 184

The Current Lactation Speciﬁc Culling Rate for The GNRH
Feasibility Estimation ........................................................... 188

The Desired Lactation Speciﬁc Culling for The GNRH
Financial Feasibility Estimation .............................................. 189

xiv

figure 3,
figure 4.
Figure 5
figure 6
figure 7
Figure 8

Figure 9
Figure 10

Fisure ll
Flame 1:

FlStare 13.

Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.

Figure 9.

Figure 10.

Figure 11.
Figure 12.

Figure 13.

Figure 14.
Figure 15.
Figure 16.
Figure 17.
Figure 18.

Figure 19.

LIST OF FIGURES

Overview of the DSS ............................................................ 108
Herd Size and Herd Inventory Input Field ......1 16
Expected Production by Lactation Input Field .............................. 117
Culling Associated Production Effect Input Field ........................... 118
SCC by Lactation Input Field ........................... . ....................... 119
Culling Associated SCC Effect Input Field ................................. 119
Current Lactation Speciﬁc Culling Rates Input Field . 121
Desired Lactation Speciﬁc Culling Rates Input Field ..................... 123
Within Lactation Monthly Removal Schedule for

Non-Mortality Health Culls Input Field ...................................... 124
Within Lactation Monthly Removal Schedule for Mortalities

Input Field ........................................................................ 125
Other Production and Financial Information Input Field .................. 126
Lactation Expense Adjustment Input Field ................................... 128
Health Cull and Mortality Reduction Technology

Investment Input Field ....................... 129
Current Herd Inventory Dynamics Output Field. .. . . . ...... ........... 131
Current Monthly Cash Flows Output Field ................................. 134
The Sold for Dairy Proceeds Value Output Field ........................... 135
Cull Cow Proceeds Value Output Field ....................................... 136
Desired Herd Inventory Dynamics Output Field ........................... 140
Desired Monthly Cash Flows Output Field .................................... 141

XV

 

figure 20

figure 21

figure 22

 

Figure 20.
Figure 21.

Figure 22.

The Cash Flows Associated with the Health Cull and
Mortality Reduction Technology and Programs Output Field ......... 142

The Financial Implications of the Health Cull and Mortality
Reduction Technology and Program Output Field .......................... 143

The Estimated Annual Culling Rates with and without

Adopting the Health Cull and Mortality Reduction
Technology and Program Output Field ....................................... 144

xvi

 

 

I Back:
I

Culllr.
constant or e\

lactation heift
(Dairy Recon:

Many 1
tend to be low

level optimal d

 

culling rates ral
hlgher The 3V6
TepOrled To be _

There a

calling rates. \‘

pTOfllable l0 C0

end ofa llPlCal

tound that it We

oiling the”) OL

CHAPTER 1.
INTRODUCTION
1. Background and Motivation

Culling is the act of identifying and removing a cow from a herd, and, assuming a
constant or expanding herd size, replacing the cow with another cow, usually a ﬁrst
lactation heifer. The culling ratel describes the percentage of cows removed from a herd
(Dairy Records Management Systems, 1999).

Many researchers have shown that estimated optimal average annual culling rates
tend to be lower than those observed. Selected previous work concerning estimated herd
level optimal dairy cattle culling rates can be seen in Table 1. The estimated optimal
culling rates range from 19 to 29 percent. Actual average annual culling rates tend to be
higher. The average annual culling rate for Midwest DHIA herds during 1996 — 2000 was
reported to be 38 percent by Quaiffe (2002).

There are numerous trade articles offering suggestions for reducing culling rates.
There are far fewer articles informing producers how to decide how to proﬁtably reduce
culling rates. Van Arendonk and Dykhuizen (1985) determined that it was more
proﬁtable to continue breeding average producing dairy cattle that do not settle until the
end of a typical lactation before culling the animal. Ngatzke, Harsh, and Kaneene (1990)
found that it was more proﬁtable to treat cattle with cystic ovaries twice rather than
culling them outright. Houben, Huirne, and Dykhuizen (1994) found that it was more

proﬁtable to treat mastitic cattle than to cull them.

 

1 The culling rate for this research was calculated by taking the number of cattle culled in a year divided by
the average number of lactating and dry cows for that year.

Table l.

 

l Study ‘
, Renkema an:
Stelwagen l l A.
\"an Arendor.
Rogers. Van
Arendonk. an;
McDaniel t l
Ngatzke, liars
Kaneene ( 199
Lohr ( l 993 )

\
illouben. Hulrn
Dykhuizen ('19

\K.
.Stott (1994)

 

Jones (2001 )
\

The ultil
l0 enable farme.

Cam bi healtl

understand hou-

Utiiled and h0\\'

for t‘ . .
t llhe \Ildlkes

cullz .

titled. and “he!

7‘"-
UN

. 9" s
‘ M“ mOdel are

likelihood of an

Table 1. Estimated “Optimal” Culling Rates of Select Studies

 

 

 

 

 

 

 

 

 

 

Study Research Question Optimal Culling Rate (%)
Renkema and What is the economic signiﬁcance 19
Stelwagen (19791 of longer herd life?
Van Arendonk and Should an animal be bred, kept 27
Dykhuizen (1985) but leﬁ open, or replaced?
Rogers, Van What is the inﬂuence of 25
Arendonk, and production and price on optimum
McDaniel (1988) culling rates in the United States?
Ngatzke, Harsh, and Should an infertile cow be treated, 22
Kaneene (1990) kept but not bred, or replaced?
Bauer, Mumey, and What is the optimal range of 25
Lohr (1993) culling rates given a planned
lactation removal policy?
Houben, Huirne, and Should a currently mastitic cow 29
Dykhuizen (1994) be bred, kept but not bred, or
replaced?
Stott (1994) When should the typical UK. 21 - 24
cow under typical ﬁnancial
conditions be replaced?
Jones (2001) What is the optimal culling rate 22-25%
given various price and cost
conditions?

 

 

 

 

The ultimate goal of this research is to develop a Decision Support System (DSS)
to enable farmers to determine the ﬁnancial feasibility of reducing the number of culls
caused by health problems. Prior to developing a culling reduction DSS, it is important to
understand how culling affects production, how many cattle are culled, why cattle are
culled and how management programs affect culling rates. In Chapter 2, DHIA records
for ﬁve Midwestern states and ﬁve Northeastern states are examined to determine how
culling affects milk production, the percentage of cattle culled each year, why cattle are
culled, and when cattle are culled within a lactation. In Chapter 3, the DHIA dataset and a
probit model are used to determine how individual cow and herd characteristics affect the

likelihood of an individual cow being culled. Four OLS models are applied to the USDA-

 

APHIS NAF‘
programs aft‘:

using a hype-i .

producer wot.

 

ll. Cullil
Cullir.
Culling due to

dairy purposes

 

culls caused b.‘
actually refers
run. there are g
are n0t

TWO e
disease These
“Omenclature
\actlne but cl
voluntary Al‘
as t-Olumary <
indiyidUal Co
Chose a Sire t
aqua”). beyc

BeCal

Q

“alleges ant

APHIS NAHMS Dairy ’96 Survey dataset in Chapter 4 to determine how management
programs affect culling rates. In Chapter 5, a DSS system is described and explained
using a hypothetical farm situation. In Chapter 6, the DSS is used to determine what a
producer would be willing to pay to reduce his or her culling rates.

II. Culling Rate Terminology

Culling due to health reasons is commonly referred to as involuntary culling.
Culling due to low production, cow aggression, or when a cow is sold to another farm for
dairy purposes is referred to as voluntary culling. Involuntary culling typically refers to
culls caused by health problems, but the name infers something else. “Involuntary”
actually refers to something that cannot be controlled by the principal agent. In the short
run, there are some health problems that are beyond the control of the producer, others
are not.

Two examples illustrate this. Assume that cattle must be culled due to a particular
disease. These culls would be referred to as involuntary culls under the usual
nomenclature. If the producer could have prevented the disease through an available
vaccine but chose not to vaccinate, the underlying culling cause is not involuntary but
voluntary. Alternatively, cattle that are culled for relatively low production are referred to
as voluntary culls, but how much control does an owner have over the ability of an
individual cow to produce relative to the rest of the herd? Assuming that the manager .
chose a sire that offered enhanced genetics, the fact that a cow is a production anomaly is
actually beyond the producer’s control and the cull should be deemed an involuntary cull.

Because this research is concerned with the ﬁnancial feasibility of management

strategies and programs designed to reduce the level of culling due to health problems,

which infers

 

producer. the
Culls due to I
classiﬁed as p

lameness and

 

which infers that at least a portion of the health culls are within the control of the
producer, the usual classiﬁcation of “voluntary” and “involuntary” culls were ignored.
Culls due to low production, aggression, and being sold for dairy purposes were
classiﬁed as production culls. Culls caused by health problems — udder and mastitis,

lameness and injury, disease, and reproduction — were classiﬁed as health culls.

 

I. Intro
I

Tode
needed conce.I

origin. cow ag

 

factors. there i
farm's indix'id
al/‘P‘roach can I

The p;
llh'OUgh 1999
and Northeast
York Perms).
help develop

the CUlllng la:

11. Data
Tnt‘ora
Managemm:

CHAPTER 2.
CULLING AND LONGEVITY DESCRIPTIVE STATISTICS
I. Introduction

To determine the ﬁnancial feasibility of health cull reduction, information is
needed concerning how proﬁtability is affected by cow longevity, health culls, state of
origin, cow age, breed, herd size, and production level. If culling is not affected by such
factors, there is no need to develop a DSS that is ﬂexible enough to accommodate a dairy
farm’s individual culling statistics and culling related information and a “one size ﬁts all”
approach can be used when designing the DSS.

The purpose of this chapter is to examine descriptive culling statistics from 1993
through 1999 for Dairy Herd Improvement participating dairy farms in ten Midwestern
and Northeastern states: Illinois, Indiana, Iowa, Maine, Michigan, New Hampshire, New
York, Pennsylvania, Vermont, and Wisconsin. The information in this chapter is used to
help develop hypothesis tests as well as to provide culling and longevity information for
the culling rate reduction decision model.

11. Data Description

Information for this chapter was taken from data supplied by Dairy Records
Management Systems (DRMS). The data contained production records of Dairy Herd
Improvement (DHI) participating dairy farms in Illinois, Indiana, Iowa, Maine, Michigan,
New Hampshire, New York, Pennsylvania, Vermont, and Wisconsin from 1993 through
1999 for individual cow records and 1995 through 1999 for herd level records. There

were 7,087,699 individual cow lactation observations in the data. Depending on the type

of statistical
used in this :"
III The I

Not c

 

 

 

animals, cull;

 

Midwestern ar
Actual 305 day
elven lactation

fat PTOduction

protein PIOduc
There is a prod;

NeVerll
cell COum l‘alu
m 58995 as a
increased With
price deductior

Tile Oiefajl gem

of statistical analysis, many of the observations were dropped in the various analyses
used in this research due to incomplete information.
111. The Effect of Cow Longevity and Culling on Production

Not only does culling affect farm cash ﬂow through the buying and selling of
animals, culling can also affect farm cash ﬂow through its effect on production. Bauer,
Mumey and Lohr (1993) showed that milk production per lactation for Alberta dairy
cattle increased from the ﬁrst lactation before declining after the sixth lactation. Stott
(1994) showed a similar relationship for United Kingdom cattle. Jones (2001), however,
showed a negative correlation between age and milk production per lactation.

Table 2 shows how cow age affected production on the farms of the ten
Midwestern and Northeastern states. These values were determined by comparing the
Actual 305 day milk, milk fat and milk protein production of cattle that completed a
given lactation with those that completed their ﬁrst lactation. Milk, milk protein, and milk
fat production increased with cow age through the ﬁfth lactation. Milk, milk fat and milk
protein production did not slip below ﬁrst lactation levels until the tenth lactation. Thus,
there is a production-based incentive to keep cows through the ﬁfth lactation.

Nevertheless, there are also disadvantages to increasing cow longevity. Somatic
cell count values, which measure the number of white blood cells per milliliter of milk
and serves as a measure of milk quality (Dairy Records Management Systems, 1999),
increased with each completed lactation. Higher somatic cell count values can lead to
price deductions for dairy farmers. Another problem with increasing cow longevity is that

the overall genetic improvement rate for the herd decreases as less heifers enter the herd.

Table 2.

 

Lactation
3 Number .\

 

 

ii l5 lo;

the herd To d,
culled fut heat.
that “9'6 culle
W‘ci'e COmParel
comIllered the;
from 3 percent
Produced from
lovt procjumo1

milled, Callie

Who
lYer
e nos

 

 

 

 

 

 

 

 

 

 

 

 

Table 2. The Effects of Cow Age on Actual 305 Day Milk Production, Milk
Fat Production, Milk Protein Production, and Somatic Cell Count

Lactation Actual 305 Day Actual 305 Day Actual 305 Day Somatic Cell
Number Milk Production Milk Fat Milk Protein Count

(% of First Production Production (% of First

Lactation (% of First (% of First Lactation

Values) Lactation Values) Lactation Values) Values)
1 100 100 100 100
2 112 112 112 127
3 115 115 115 143
4 116 117 117 154
5 115 116 116 163
6 112 113 113 173
7 109 109 109 170
8 106 106 106 178
9 102 102 102 182
10 99 99 99 181

 

 

 

 

 

 

It is logical to assume that culled cattle produce less milk than cattle retained in

the herd. To determine the extent to which milk production declines when cattle are

culled for health and production problems, the projected 305 day milk for the DHI cattle

that were culled for mortality, health problems, production and sold for dairy purposes

were compared to the actual 305 day milk yield for the DHI cattle that successfully

completed their lactations (Table 3). Cattle that were culled for health reasons produced

from 2 percent more to 6 percent less than their healthy counterparts. Cattle that died

produced from ﬁve to sixteen percent than their non-culled herd mates. Cattle culled for

low production produced from six to twenty nine percent less than cattle that weren’t

culled. Cattle sold for dairy purposes produced between two to 7 percent less than those

who were not culled.

Sam-3

 

for cattle that
3).

Table 3'
Counl

l LactatiOn

 

/

t:/_‘

 

a/o/w/l/l

\

There at"
the reduction of
milk milk fat. a
though the ﬁfth
become a health
intbat somatic c

{h
eherd genetic

ll'.

Somatic cell counts also differed between culled and retained DHI cattle. Except

for cattle that died, culled cattle typically have higher somatic cell counts as well (Table

 

 

 

 

 

 

 

 

 

 

 

 

 

3).
Table 3. The Percentage Difference in Milk Production and Somatic Cell
Counts Between Retained and Culled Cattle
Lactation Health Mortalities Production Sold for Dairy
Purposes
Milk SCC Milk SCC Milk SCC Milk SCC
l 6 19 16 5 26 29 11 4
2 15 7 -1 7 19 4 2
3 0 13 6 -1 11 15 3 3
4 1 l3 6 -4 l l 11 3 6
5 2 12 11 -5 10 13 1 7
6 0 9 7 -10 9 10 1 6
7 -2 9 5 -5 3 6 1 7
8 0 ll 9 -6 8 9 5 7
9 0 11 11 -6 7 11 5 3
10 6 8 14 -8 11 11 6 5

 

 

 

 

 

 

 

 

 

 

There are both incentives and disincentives to increasing cow longevity through
the reduction of health culls. A production incentive for decreasing health culls is that
milk, milk fat, and milk protein production increase with each completed lactation
through the ﬁfth lactation. Another incentive to reduce health culls is that. cattle who
become a health cull are less productive than their healthy herd mates. Disincentives exist
in that somatic cell counts increase with cow age and increased cow longevity means that
the herd genetic improvement rate declines as less heifers enter the herd.

IV. Average Culling Rates for 1993 — 1999 by State, Breed, Production Level,
and Herd Size

Across all ten states over the seven year period, the average culling rate for the
1993 — 1999 period was 35.1 percent (Table 4). These culling rates are higher than the

estimated optimal culling rates determined by Rogers, Van Arendonk, and McDaniel

(1988). 33”“

Optimal herC

 

10 39 percent
purposes in l:
DHIA record
316 percent.

importance or
for the remail'

farms for dairjs

 

Northe

herds had annl.

 

Midwestern (ll
average culling
herds had the It
291 percent .\'
period.
Holsteir
dairy cattle bree
exPerienced a It
“Till Brown SW

Percent: bur this

(1988), Bauer, Mumey, and Lohr (1993) , Stott (1994), and Jones (2001). The estimated
optimal herd-level culling rates determined by these researchers ranged from 19 percent
to 29 percent. These researchers, however, did not include cattle that were sold for dairy
purposes in their estimated optimal culling rates. If such culls are removed from the
DHIA records, the average culling rate for the ten states during the 1993 — 2000 period,
31.6 percent, is much closer to the estimated optimal culling rate values. This shows the
importance of understanding how culling rates are calculated when making comparisons.
For the remainder of this chapter, the culling rate will not include cattle sold to other
farms for dairy purposes unless noted otherwise.

Northeastern (Maine, New Hampshire, New York , Pennsylvania and Vermont)
herds had annual average lower culling rates, ranging from 29.1 to 31.4 percent, than
Midwestern (Illinois, Indiana, Iowa, Michigan, and Wisconsin) herds, whose annual
average culling rates ranged from 33.2 to 35.1 percent. Vermont and New Hampshire
herds had the lowest state level average annual culling rate for the 1995 - 1999 period at
29.1 percent. Michigan had the highest average annual culling rate, 35.1 percent, for the
period.

Holstein herds exhibited the second-highest average annual culling rate among
dairy cattle breeds during the 1993 through 2000 period with 31.9 percent. Jersey herds
experienced a lower average annual culling rate of 27.2 percent for the period. Farms
with Brown Swiss cattle had an average annual culling rate of 30.9

percent, but this was not signiﬁcantly different than the Holstein culling rate (p-value

Table 4.

—_—————

I
All ten states I

T New Hampsl:
Vermont
Maine
New l orlt

 

 

1 Pennsylvania
x

Indiana
Illinois

 

 

 

1 Michigan
Wisconsin

V Iowa
Holstein
lersev

‘Brosm SWiss
\

Guemset
Arreshire
, Milking

ShOrthOm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 4. Average Culling Rates for 1993 - 1999.

Average Culling Rate with Sold Average Culling Rate without

for Dairy Culls Sold for Dairy Culls
Farm Rate Standard Farm Rate Standard

Observations (%) Deviation Observations (%) Deviation
All ten states 58,498 35.1 19.4 58,181 31.6 15.7
New Hampshire 799 33.9 19.5 779 29.1 12.6
Vermont 7,700 32.6 16.6 7,630 29.1 11.4
Maine 994 36.3 22.8 975 31.4 15.8
New York 15,363 33.5 16.2 15,289 30.3 12.0
Pennsylvania 12,068 32.9 13.3 12,034 30.5 11.4
Indiana 7,827 38.9 21.3 7,785 33.2 13.9
Illinois 2,978 33.6 12.7 2,963 33.6 12.7
Michigan 3,302 37.7 15.4 3,287 35.1 12.7
Wisconsin 2,201 38.2 40.4 2,192 34.8 38.6
Iowa 5,266 37.9 26.3 5,247 34.9 24.8
Holstein 50,244 34.8 17.2 50,164 31.9 14.2
Jersey 3,334 33.0 22.1 3,237 27.2 15.3
Brown Swiss 723 38.0 57.0 699 30.9 53.9
Guernsey 621 45.3 30.1 604 42.0 23.2
Ayreshire 816 37.0 24.3 778 29.6 17.9
Milking 104 36.5 26.7 100 26.7 13.1
Shorthorn
RHA < 18,000 22,777 34.6 23.1 22,522 30.8 19.0
lbs.
18,001 - 21,000 19,954 34.8 16.4 19,919 31.9 13.6
lbs RHA
21,001 - 24,000 11,852 35.5 15.9 11,832 32.2 9.8
lbs RHA
24,001 - 27,000 3,292 37.2 19.2 3,287 32.3 11.7
lbs RHA
RHA > 27,001 623 40.9 22.7 621 32.2 13.6
lbs
Cows < 150 53,003 35.0 20.1 52,686 31.4 16.2
150 to 300 3,936 35.1 10.9 3,936 33.4 9.2
Cows
301 to 450 849 35.6 11.3 849 34.0 8.5
Cows
451 to 600 335 35.5 8.6 335 34.2 8.4
Cows
Cows>601 375 38.3 15.8 375 36.8 14.3

 

 

 

 

 

 

 

10

 

Al'TCShlr e ht

 

l’ercem Fart
rate for the p
{ale ofHolSli

HerdS
co“ Per lactai
30.00 period T
Wrcent Her d5

those with a R

 

2100018thds
average annual
pounds of mil is
than 27000 PO

RHA of betwes

culling rate of l

farms \i

rate of 31 4 per

exhibited highe

average annual
ai'erage annual

‘

a 342
percent a

‘5‘]

1h more than .

= 0.6242). Guernsey herds had the highest average annual culling rate of 42 percent.
Ayreshire herds exhibited lower culling rates as compared to Holstein farms with 29.6
percent. Farms with milking shorthom cattle exhibited the lowest average annual culling
rate for the period at 26.7 percent, but this was not signiﬁcantly different than the culling
rate of Holstein cattle.

Herds with a Rolling Herd Average (RHA) of less than 18,000 pounds of milk per
cow per lactation exhibited an average annual culling rate of 30.8 percent for the 1993-
2000 period. This culling rate was lower than the all sample (all ten state) average of 3 1 .6
percent. Herds with higher production levels had higher culling rates as compared to
those with a RHA of less than 18,000 pounds. Herds with a RHA of between 18,000 and
21,000 pounds of milk experienced an average annual culling rate of 3 1 .9 percent. The
average annual culling rate seemed to plateau for herds producing more than 21,000
pounds of milk. Herds producing between 21,000 and 24,000 and herds producing more
than 27,000 pounds had an average annual culling rate of 32.2 percent. Farms with a
RHA of between 24,001 and 27,000 pounds per cow experienced an average annual ’
culling rate of 32.3 percent for the period.

Farms with a herd size of less than 150 cows exhibited an average annual culling
rate of 3 1 .4 percent for the 1993 - 2000 period. Farms with more than 150 cows
exhibited higher culling rates. Farms with 150 to 300 cows exhibited a 33.4 percent
average annual culling rate for the period Farms with 301 to 450 cows exhibited an
average annual culling rate of 34 percent, while herds with 451 to 600 cows experienced
a 34.2 percent average annual culling rate for the period. The largest size group, herds

with more than 600 cows, exhibited the highest average culling rate, 36.8 percent.

11

1’. Cull
and

Linda.
what its culi.
manager or a
phrase “it her
but it is apprt
reduction is :T
having a high
Conversely. a
rate if it has a 1

Table 5
State The mos:

0' Other " With

 

V. Culling Reasons for 1993 - 1999 by State, Lactation, Breed, Herd Size
and Production Level

Understanding why cattle are culled on a farm is just as important as knowing
what its culling rate is (Natzke, 2002). Understanding culling reasons assists the farm
manager or advisor in deciding whether to reduce culling rates and how to do so. The
phrase “whether to reduce culling rates ” is not typical in discussions about culling rates,
but it is appropriate. When an above average culling rate is encountered, a culling rate
reduction is generally prescribed. Nevertheless, a farm that has a high culling rate due to
having a high number of sold for dairy culls may be at its optimal culling rate.
Conversely, a farm with a low average culling rate may need to further reduce its culling
rate if it has a larger-than-normal proportion of health culls and mortalities.

Table 5 displays the percentage of culled cattle removed for particular reasons by
state. The most common reason for culling cattle for all ten states combined was “Injury
or Other” with twenty-seven percent of all culls in the ten states being attributed to this

reason. Reproduction problems was the second-most prevalent

12

     

 

-.....9 EEC 3:5 EEC 3:2,. EEC EEC

    

  

. TL::,~ k: «A:
-.C ox... ..C .....ov ..C ....ch ..3 .3; C: oxov C: ......L ..3 e...» ..C we» .2. we» k3 we» 29:31. 5;.me
--- 1.2.: l- .33 .- i=2 --. .. 1;... . Z... . l <Li >2 .- - ...:2 ..- .-.~..\v l. -. .~ ‘2 l- lo~l:.<. . «wispy

0:27. A: b.3957»: H=m==nv .V. .~=~.~.~.

 

m:20

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ow mm MK 5 Q. E. mm mm K mm on £23: 30.:
w o_ N. m N N N m m N v 333
v n v v N m m m N m m 0835
E m 2 N: 3 3 c: N: E o. N. 22:32
2 o. E 2 o. m o. 2 w o— o. 35
350
ON MN 2 M: m. 0N me :m N N N be E22:
M: 2 m: NN NN NN 2 m; a: _N @— cocogemom
203235
N: a ON 2 C N: e a 3 w 2 Bot—
.025
w m: N. o o. \t o w a m w .5: Eom
mwoq
m m v v m n v v N. m v 3a 232
@300
$300 $300 @300 @300 $300 $300 $260 @300 $300 @300 3:20
3:20 3:20 3:20 3:20 3:20 3:20 3:20 3:20 3:20 3:20 .3 o\o0
co :0 co .5 co .3 co .5 co :0 ao .3 co 3 co .5 Co .5 2o 3 seam 5:32
S :5 :2 .d E <m >2 m: ._.> :2 c: :< w=:_20

 

 

 

 

38m .3 2.83% 95:20

.m as:

13

 

culling reascl
cattle were c
accounting ft
the ten state s
New I
production re
injuries New
Eight
10 mortalities
like ten states
Vermont had 1
self“ Percent
alerage
Eight}.
?ioblemS‘ Mail
left the hem du
cattle in Maine
Next Yc

?

L

r . -
easons at club
injuries at fortV

same removed f

f .
‘ol'da .
if}. Phrpoy

culling reason accounting for nineteen percent of all culls. Thirteen percent of the culled
cattle were culled due to low production. Mastitis was the fourth most common reason
accounting for twelve percent of all culls. Cattle mortalities was the ﬁfth ranked reason in
the ten state sample accounting for ten percent of the culled cattle.

New Hampshire farms had the second lowest proportion of culls due to
production reason. The majority of New Hampshire cull cattle were culled due to
injuries. New Hampshire had the highest proportion of culls caused by mastitis problems.

Eight percent of the cattle culled in Vermont were removed from their herds due
to mortalities. This value tied with Pennsylvania for the lowest mortality removals among
the ten states. Injuries caused the majority (twenty four percent) of culls in Vermont.
Vermont had the highest proportion of culled cattle removed for feet and leg problems at
seven percent. Overall, Vermont culled less cattle for health reasons than the ten state
average.

Eighty three percent of all cattle culled in Maine were culled due to health
problems. Maine exhibited the second and third highest proportion of culled cattle that
left the herd due to injuries and mortalities respectively. Thirty one percent of the culled
cattle in Maine were culled for injuries. Eleven percent of the culled cattle died.

New York exhibited the highest percentage of cull cattle removed for health
reasons at eighty eight percent and the highest percentage of cull cattle being removed for
injuries at forty three percent. New York herds exhibited the lowest percentage of culled
cattle removed for reproduction (sixteen percent), low production (six percent) and sold

for dairy purposes (six percent) reasons.

14

Per

reasons at 5

percentage
the Other ea
Pennsylvanl
with Indiana
for reproduc
highest perce
Indiar
health problel
ofculled cattl
percentage ( se
majority (me

T885005

Eighty
ReprodUCllOn ;
Illinois herds I
Cattle remOVEd
canlecuued do,
lot» PTOdUCtion,
accounted f0r e;

 

Ea“. M.

 

 

 

Pennsylvania had the second lowest percentage of cull cattle removed for health
reasons at seventy four percent. Pennsylvania tied with Vermont for having the lowest
percentage of culled cattle being removed for mortality reasons at eight percent. Unlike
the other eastern states in this study, the primary reason for cattle removal in
Pennsylvania was reproduction problems instead of injuries. In fact, Pennsylvania tied
with Indiana and Illinois for having the highest proportion of cull cattle being removed
for reproduction problems at twenty two percent. Pennsylvania also had the second
highest percentage of cull cattle removed for production reasons at nineteen percent.

Indiana tied with Michigan for the lowest percentage of cull cattle removed for
health problems at seventy three percent. Indiana had the largest percentage (ten percent)
of culled cattle removed for sold for dairy purposes. Indiana also had the third largest
percentage (seventeen percent) of culled cattle removed for production purposes. The
majority (twenty two percent) of Indiana cull cattle were removed for reproduction
reasons.

Eighty one percent of Illinois cull cattle were removed for health reasons.
Reproduction problems were the primary reason why culled cattle were removed from
Illinois herds. Illinois tied with Maine for having the second highest percentage of culled
cattle removed for mortalities at eleven percent.

Michigan tied with Indiana for the lowest percentage, seventy three percent, of
cattle culled due to health problems. The most common culling reason in Michigan was
low production, which accounted for twenty percent of all culls. Reproduction reasons
accounted for eighteen percent of all culls and was the most common health cull reason

in Michigan. Michigan had the highest percentage of cattle removed due to mortalities

15

(fourteen p6
to injury Wit

Ses'e
problems. In,
percent) W1:
(fifteen perce
cattle remove
cattle remove

Eight}
Accounting t}
reason in low;
thirteen percer

HOlSIC:
(Table 6) Set;
canle were col I
overall culling y
TOT (lain pUl’pQ:

 

Percent Oflhe C
due to low prod

J ‘l
‘ " PE.”
he”! MC.

 

(fourteen percent). Michigan also experienced the lowest percentage of cattle culled due
to injury with sixteen percent.

Seventy eight percent of Wisconsin culled cattle were culled due to health
problems. Injuries accounted for most of the Wisconsin cattle culls (twenty eight
percent). Wisconsin had the lowest percentage of cattle culled for reproduction reasons
(ﬁfteen percent), mastitis (nine percent), and tied for the lowest percentage of culled
cattle removed for feet and leg problems. Wisconsin had the highest percentage of culled
cattle removed for udder (ten percent) and sold for dairy purposes (thirteen percent).

Eighty percent of culled cattle were culled due to health problems in Iowa.
Accounting for twenty percent of all cattle culls, injuries was the most common culling
reason in Iowa. Iowa had the second highest mortality rate among the ten states with
thirteen percent.

Holstein cattle were most often culled for injury or other health related problems

(Table 6). Seventy-six percent of Holstein cattle were culled for health reasons. Jersey
cattle were culled less often for health problems. Besides Jersey cattle having the lowest
overall culling rate of the six breeds analyzed, most of the culled Jersey cattle were sold
for dairy purposes. Culled Brown Swiss cattle were primarily plagued by reproduction
problems, but 20 percent were also sold to other farms for dairy purposes. Twenty four
percent of the Guernsey cattle were culled for injury reasons, and 21 percent were sold
due to low production. The majority of Ayreshire cattle were culled for reproduction
problems. Milking Shorthorn cattle had the lowest percentage of cattle culled for health
purposes at 54.9 percent and the highest percentage of cattle sold for dairy purposes at

31.7 percent. Most of the culled Milking Shorthorn cattle were sold to other farms for

16

dairy purpo

injury or on

Table 6.

Culling
‘ Reason

feet and
I Lees
' Sold for
‘ Dairy
‘ Low
1 Production

 

19111?) or
Other
Died
Mastitis
Disease

3 Udder

. Toral Health
I Culls

The p

howefeﬂ exl?

dairy purposes. The most common cause of health culls among Milking Shorthorn was

injury or other health problems.

 

 

 

 

 

 

 

 

 

 

 

 

Table 6. Culling Reasons by Breeds

Culling Holstein Jersey Brown Guernsey Ayreshire Milking

Reason (% of (% of Swiss (% of (% of Shorthorn
Culled Culled (% of Culled Culled (% of
Cows) Cows) Culled Cows) Cows) Culled

Cows) Cows)

Feet and 4 2 4 4 3 2

Legs

Sold for 7 19 17 7 15 28

Dairy

Low l3 I4 13 l8 l3 13

Production

Reproduction 19 16 2O 19 22 15

Injury or 27 20 19 27 25 20

Other

Died 1 1 ll 10 ll 9 8

Mastitis 12 12 9 9 9 7

Disease 3 2 3 2 2 2

Udder 4 4 5 3 2 5

Total Health 80 67 70 77 72 59

Culls

 

 

 

 

 

 

 

The primary reason for cattle being culled on farms with less than 300 cows was

injury (Table 7). Injury problems decreased as herd size increased. Larger herds,

however, experienced proportionately more feet and leg problems as well as mortalities.

The results for the larger herd category are higher than those reported by Quaiffe (2002)

for all DHI herds participating in the DRMS DHI recordkeeping system.

17

 

Table 7-

Iﬁ

 

l

l

1

feet and Le“
1 gold for Dali
1 Low Product
. lnlUTV or 01‘,
Disease
L'dder
\
1 Total Healtn
Culls

 

The in
18.000 pound:
highest propOl
to increase prc
proportionatel
herds also exp

problems pea},-

before declinir

Table 7. Culling Reasons By Herd Size

 

 

 

 

 

 

 

 

 

 

 

 

Culling Reason 0 - 150 151 — 300 301 — 450 451 - 600 More than 601
Cows Cows Cows Cows Cows
(% of (% of (°/o of (°/o of (% of Culled
Culled Culled Culled Culled Cows)
Cows) Cows) Cows) Cows)
Feet and Legs 4 7 7 9 8
Sold for Dairy 8 5 4 3 7
Low Production 13 13 13 13 13
Reproduction 19 1 8 1 7 16 14
Injug or Other 27 26 25 23 23
Died 10 13 14 14 16
Mastitis 12 13 14 15 13
Disease 3 3 4 5 4
Udder 4 2 2 2 2
Total Health 79 82 83 84 8O
Culls

 

 

 

 

 

 

The most common reason an animal was culled on herds producing less than

18,000 pounds of milk per cow per lactation was injury, but these farms also had the

highest proportion of low production culls, possibly indicating that these herds aretrying

to increase production through culling (Table 8). As production levels increased,

proportionately more cattle were culled due to reproduction problems. Higher producing

herds also experienced less injuries per cow. The proportion of cattle culled for mastitis

problems peaked for farms averaging between 18,000 and 21,000 pounds of milk per cow

before declining with higher production levels.

18

 

Table 8.

feet an
M
m

Injury or
N

 

l
\

\M
\D
N

I

“- Lacta

 

 

 

 

 

 

 

 

 

 

 

 

Table 8. Culling Reasons By Rolling Herd Average
0 - 18,000 18,001 — 21,001 - 24,001 — More than

Pounds 21,000 24,000 27,000 27,000

Per Cow Pounds Pounds Per Pounds Per Pounds Per
(% of Per Cow Cow Cow Cow

Culled (% of (% of (% of (% of

Cows) Culled Culled Culled Culled

Cows) Cows) Cows) Cows)
Feet and Legs 2 6 7 7 6
Sold for Dairy 8 7 8 9 10
Low Production 14 12 1 1 11 9
Reproduction 1 9 20 1 8 1 7 1 7
Injury or Other 28 26 24 23 24
Died 10 11 12 12 12
Mastitis 1 1 13 14 15 15
Disease 3 3 4 4 4
Udder 5 2 2 2 3
Total Health Culls 78 81 81 80 81

 

 

 

 

 

 

VI. Lactation Specific Culling Rates for 1993 — 1999

In general, the lactation speciﬁc total culling rate (Table 8), health culling rate

(Table 9), and the mortality rate increased through the tenth lactation (Table 10). This

pattern was similar when the data was sorted by breed (Table 11), herd size (Table 12) ,

and production (Table 13). First lactation culling rates ranged from 18.7 percent in

Vermont to 25.8 percent in Indiana. Vermont also had the lowest second lactation culling

rates. Over forty percent of the second lactation cattle were culled. Although Indiana had

the highest ﬁrst lactation culling rate, it had the second to lowest second lactation culling

rate. Overall, Vermont had the lowest lactation speciﬁc culling rate.

19

 

L. 2: an: .\o\

17:1 .1<,al -.--..vzl :2 l ._.>.l

maﬁa: 1 M06- LG.- Una-Ta

«.1 k: ﬂex k: &O.\ kc @5\ U.:.~::,V
m0-w~rwa
l l l. -. 1&er-.- immohtmwx-

A: .02;- M.~..==..V 220% 9:30.?» =o.:asuat~

23.22320

. .llflllt!l

..m. .zQ 3. F

3882.: >32 SO 23.: 3:8 2 28 wins 2:8 5:? 322683 805 8688 £3 :2 33B:—

 

 

 

 

 

 

 

 

 

 

 

 

89: 5
222302
3 mm 3 on O0 OO 3 we Om 3 NO O:
OO mm 3 NO mm no 3 NO Nm 3 OO O
mm Nm we 3 mm 8 5 mm me 5 mm m
Om Om NO mm Om OO mm Om 9a mm X N.
Vm 3 mm So we Om mm em Ne em 3 O
Om Nv mm 3 N2 E S 3 mm Om 5.. m
we wm Nm EV mm we 3 NV mm 3 3 v
ov X we 2. mm mm ov Om wN Om mm m
em 5N Om Om ON Nm Nm Om MN 3. Z N
O: 2 ON ON _N M: 2.: O. S O: M: _
@500
ca .2
€300 €360 $300 $260 $300 @260 $300 €260 $260 3300 83:
co 3 co :0 co 3 co :0 co 3 ..o .5 co .5 co .5 co 3 co :0 wasao
328m
3 :3 :2 r: E «5 >2 NE ._.> 32 3H :< 202803

 

 

 

 

 

 

 

 

 

 

 

 

 

as. .. 83 to. saw 2. .85. 2.5.5 .35 2.28% 52523

.O 932,—.

20

It ..‘s '\ sti‘ 7‘ \C\ I. \C.‘ . on. '.\

3:25.
...».2 . :2 ._.> :z ......E:w..~.:..2.:o._,...

’15:: -.3 ’ -_2 i .: z. -.<._

see. i no... ...: 857.. A2 .35. 2.5.2.0 523: 3.32.7. =e..::.:§ .2. 2:5.

.eaoo

:0: SO 23an 2032358 .0386 52.3: :23 $6323: £86302: 22:22: :23 332 8 02:6 2:26 3:02:62:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

208 .5
30:33:
3» OO mm mm NV 3 Nm \rm 3 :O Ov O:
me me Nv we NV 3 em mm :v em Ov O
NO Om :v we Om 3 :m we Om Om me w
:v hm Om mv hm mv we we Om to :v n
Ow mm Om :Vm :Vm :2 Ov mv mm we Om O
mm mm mm Om Om hm :v mm ON ov Om m
mm Om mm Om NN Nm ON mm N :m mm v
ON NN ON Nm mN NN Nm N :N :m N m .
N :N ON N O: mN ON ON C mN mN N
v: N: N: m: E N: m: N: N: N: m: :
€260
Oo ea:
@300 $300 @260 €300 $300 $300 $300 $300 $300 @300
3:20 3:20 3:20 3:20 3:20 3:20 3:20 3:20 3:20 3:20 83:
.8 as: o é o 3 Co as o as: o as: to 3 co ,5 Co 3 ..o as: 3:3
83%
<: :3 :2 t: Z: <A: >2 3): .m> I7: 3:. :_< 22.331:

 

 

 

 

 

 

 

 

 

as: 1 n8: 8.. 83m 3 .25: 2.2.5 5.82 2.28% 52.223

.O: 932,—.

21

 

‘ F;
’IV: -1”- .:.>> I _r . :21: . . w: 2: ,1 <2 - -57: - -22 -, ..:_> .222 :< . ‘ , 15552.2:

23.. I n3. ...: 85% b. 3:: b..=:h.=: 5.325 5.25:; .. . 93.;

 

0:08 :0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

: : m N: O O: m O O O O m £8388: O:
O O : : O O O O O O O O O
O O :: O O O O O m O O O
O O :: O O O O O v O O O
O O O: O O O O O v O O O
m O O: O O O O O v O O O
O v O O m O O m m O m v
O v O O v v m m m m m m
w m O v m m v m N v m N
N N m N N N N N : N N :
@300
8:20
mo $0
@300 $300 $300 $300 $300 $250 9260 @300 $300 $300 83:
8:20 8:20 8:20 8:20 8:20 8:20 8:20 8:20 8:20 8:20 3:30
he 8 .8 8 O0 .5 mo .5 be as O0 $0 80 é .o é mo :0 Oo .5 822m
8%
<: :3 :3: A: Z: <n: >2 m2 O> $2 =< 20883

 

 

 

 

 

 

 

 

 

 

32 I 8.: .8 83m 3 as: £2.52 2.5on 52.53

.: 032,—.

22

‘\

 

32:25 \ 5.22:2: \ ::.:...:.:~.:\

: 232 :aéiié 2....,..~.;.<- :2, 202.112--..2:

A: 52525 g: .525 :2 3.3.2:: 23. .. 2.: he
89.: .3 85. ES: @25ch 2.... N3:: :20 5.5.: ...:U. :29 .55. 959:5 29:22.5 .N. 93.2.

228:0
202 :2: 28:00:: 2082:0058 688:0 “52.8: :08 £02222: £22005 2:328 O22 8202 0: 26 2:20 8:02:02: a

.8892: $26 20.: £222.: 850 0: Bow O52: 0:28 :23 822088 80:: 8088 2:20 :w 8:02:02:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

20E .6
20:88:
O O: ON O: mv Om O 3 mm m Om Ov O Nm NO O 3 :O O:
O: mv Om O Nv Om O 3 x O. Om Ow O Om OO O 3 Om O
:: :m 3. O Om :m v ov Ov m mm vv O :N mm O N2 Om O
m Om Om O mm Ov m Om Ov m ON Om O :N cm O Ov mm O
O ON Om O Om OV m Nm :2 m ON Vm O ON Om O Om Ov O
O mN mm m ON Om 2 ON Om m NN : m O ON O2 O Vm Ow m
m :N Om v N mm v :N mm v O: ON O O: mv m :m :2 v
N O: :N v ON ON m : N ON m O: mN m V: O2 m ON Om m
N O: 2N m O: ON m O: ON N m: ON v :: mm 2 NN :m N
: O m: : m: O: : m: O: : O: 2: m O ON N m: O: :
m3: mm :0 #3: m0: m0 5): m2 m0 52 MI :0 52 a: m0 E): m2 m0
@300 O0 $0
502205 @3080 $0 3300 O0 :0 @300 O0 .5 $300 O0 $0 9300 O0 .5
OEVES: mmgm 2305 2:282}: 532. 532020 2:03:03 20:22:03

 

2. 52523 2.. 9.2.5 2. 2.8.9.: 9:: I 3.: .5.
28.2 3 25. 220 0:532 .23 ":25 .26 5.3: ..EB :20 .2; 228% 82.223 .2 as:

23

 

III III I .: n..—...u\r‘;~ UC\ ~>\:..p\.‘a~ ..C\ ~\::-.v\a\a~ ....C\

:i .cm:...:.,_-.l.~:.§.: : @2503 I. 7.? @3000? Em .350 03...... :2 .. §:...,,0Om.: -. O.

1. I|.II l 1W|1yx ‘\|l.'~l I . .I I

=2:...>:...:

QQQ~ anwN:

LO.— esmw the: h: 02;— A120 \nzzuhcz 3:: ~A~=: :39 :35»: ..aZU» 2....» ...-ch. uttoiw COOOQOBRQ £4 32255

.58: 8: :5 £5an 5526058 6386 59:: was $2882 £80an S5338 can .025 8 26 2.5 89:05 N

.8883 3:6 8.: SEE: .550 2 28 mEB 2:8 £5 33688 805 3088 230 ES 89:2;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.5::
SN SS SS 2 SS SN 2 SS SN 2 SS SS S NS S a 25:38. 2
S SS SS 2 SS NS 2 SS SS S SS S S SS S S
S. SN SS S NS NS 2 SS NS S NS S S SS SS S
2 SN SS 2 SS S S SS SS S .S SS S SS .S N
: SN SS S SS S S SS SS S SS NS S SS SS S
2 SN SS S SS SS S SS SS S SS SS S NS SS S
: NN SS S SS SS S _S SS S NS SS S SN SS S
S SN SS S SS .S S SN SS S SN S S SN SS S
N NN SS S SN NS S SN SS S _N SN S 8 SS N
S 2 SN S S_ _N S S_ _N S 3 SN N S_ S: _
S: m: 5 m2 . m: 5 a: m: 5 a: m: 5 SE SE 5
9500
3:20 .«o S\Sv AS260 AS260 AS300 AS300
35 8:5 8 .5 8:5 8 SS: 8:5 .6 SS: 8:5 .6 SSS
SS 55 S82 S35 8S .. .SS 35 SS .. 5S 2:5 8S .. _S_ SS5 :2 u o 8883
SS: u SSS.
.8 85 85 SS 35. 9:5 5:33: ...: N25 :5 5.8: ..88 :5 .53. 228% 5:523 .2 SEE.

 

 

\ : . . i: . . .
r SEEN :52?ng S .33: pm :5 ...m :3: .SN I :5 _m .. 33.5. .m - :5 E.

.1: .f:
\ CE:.%.:.E.~.S.V.D< \ 5...:22:

 

SSS: I SSS: .3: ou5§<
Ste: 9.53: S: 3:: 9.5: 5:532 ...... ~32: ......v ...—.3: 33...! ......v 2.3:. 9592?. .5852... .... 2:5.

.58.. .o: :5 SEoan 5.5.6958 .0386 53...: 95 S5582 .SEoBoa SESSE v.5 .02.: o. 26 3.5 Sous—o... N
S3092: SCH“. 5.. SEES .050 o. 28 mass 2:5 .23 8.58%.. 805 .58.... 2:... =m S2535 .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9.0::
S. SS SS S SS SS S SS SS S SS NS S SS NS 8 22.58. S.
S SS SS S. SS .S S SS SS S SS SS S SS .S S
S SS SS S SS SS S SS SS S SS SS S NS SS S
S SS SS S SS SS S SS SS S .S SS S .S SS S
S NS SS S .S SS S SS NS S SS SS S SS .S S
S SS SS S SS SS S SS SS S SS SS S SS SS S
S SS SS S SS SS S SS SS S SS SS S SS SS S
S SS SS S SN SS S SN SS S SN SS S SN SS S
S SN SS S .N SN S .N SN S .N SN S SN SS N
S S. SN S S. S. N S. SN N S. SN N N. S. .
...). .... 5 as. 5. So as. m: 5 as. S... 5 ms. S... 5
.95 SS $5 :35 S SS. 535 ..S SS. .55 S SS. @358 ..S.
S... 5.. SS. 5.. SS.
.SS.SN 8.: 2S2 SSS.SN .. .SS.SN SSSSN I .SS..N SSS..N . .SS.S. SSS.S.=S_..S SS3 SS:52...
SSS. I SSS. 5. Swen}.
8:. S58... SS 55. 2.2. 5:532 SSS .2... =5 5.8: ...mu. :5 .59.. 5.8% 2:52: .3 S35

GU
analyzed f0
culling rate
first lactaiio
ten were lov
rates than H
Shonhom c1
information
culling rates

Alrhc
herd size (Te
cuiling rates
lowest lactar;

There
Culling rate ('
lactation had
r are of lactati.
prOdUCing mc

mm\‘abiliry <
Winds Of mil

The di

ana ,

Guernsey cattle had the highest lactation speciﬁc culling rates of the breed records
analyzed for each lactation (Table 12). Jersey cattle had the lowest lactation speciﬁc
culling rate for each lactation. Although Brown Swiss and Ayreshire cattle had higher
ﬁrst lactation cattle rates than Holstein cattle, the culling rates for lactations two through
ten were lower than Holsteins. Milking Shorthorns had higher lactation speciﬁc culling
rates than Holsteins in all but the sixth lactation; however, the majority of Milking
Shorthorn culls are due to sales to other farms for dairy purposes. An implication of the
information in Table 12 is that culling reduction programs using the lactation speciﬁc
culling rates of Holstein cattle may not produce accurate results for other breeds.

Although there were exceptions, lactation speciﬁc culling rates increased with
herd size (Table 13). Farms with larger than 600 cows had the highest lactation speciﬁc
culling rates. Farms with less than 150 cows, except for lactations two and three, had the
lowest lactation speciﬁc culling rates.

There was no apparent correlation between milk production and lactation speciﬁc
culling rate (Table 14). Herds producing less than 18,000 pounds of milk per cow per
lactation had the lowest lactation speciﬁc culling rate. These herds had the highest culling
rate of lactation 2 cattle. Third lactation animals were more heavily culled in herds
producing more than 27,000 pounds of milk. Cattle of ten lactations or more had better
survivability on herds producing less than 18,000 pounds of milk and more than 27,000
pounds of milk.

The differences in lactation specific culling rates among states, breeds, herd size

and production level indicates that a DSS needs to accommodate different lactation

26

 

spEClﬁC CU“

regarding I}

l
m MI
The

those associLﬂ

feasibility or
(2002) r6130"
first 21 days
determined ft

non-death he;

 

the cattle cullc
cow disappear
that die during
cattle die with.

that die do so i

speciﬁc culling rates. Without this feature, a DSS will not provide reliable information
regarding the feasibility of reducing health culling rates.
VII. Within Lactation Culling Pattern for Health Culls for 1993 — 1999

The net returns of a cow culled at the end of a lactation are generally higher than
those associated with a cull in early lactation. As such, a DSS designed to estimate the
feasibility of reducing health culls needs to consider when those cull occur. Natzge
(2002) reported that most cattle are culled during the ﬁrst 20 day period following the
ﬁrst 21 days aﬁer calving. In this research, the within lactation culling pattern was
determined for non-death health culls and mortalities. The majority of cattle culled due to
non-death health reasons are culled at the end of a lactation (Table 15). Over a third of
the cattle culled for health reasons are removed in lactation month 11 or beyond. The
cow disappearance pattern for cattle that die appears to be bimodal. The majority of cattle
that die during a lactation die in the early part of the lactation. Forty-two percent of the
cattle die within the ﬁrst sixty days of lactation. Over twenty-three percent of the cattle

that die do so in the eleventh lactation or later.

27

lllh . a h cell‘C1-S-s - sz‘.—.\av~w.~ \ -Av-~w~Sv~w~\

.25— l 6A5— ne.=.% .....Zw....:=..:z 3.... .....S...

.....— ==:=:=S— k: An): S. a.—....U ...—...:— ...¢ ...........~ 23:32.9: 3....» ....

031...: ......P =<
3:73.53. ...... 3: :52... 55.53 .m. ......a...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

S. SN S. SN S. SN S. SN S. SN S. cm .N .S NN SS SN SS SN OS I
S S S S S S S S S S S S S S S S S S S S o.
S S S S S S S S S S S S S S S S S S S S S
S S S S S S N S S S m S S S S S S S S S S .
S S N S m S N S S S S S S S S S S S S S S
N S S S S S S S m S S S S S S S S S S S S
S S S S S S S S S S S S m S S S S S S S S
S S S S S S S S S S S S S S S .S S S S. -S S
S S S S S. S S S S S S. S S S S S. S. S S S S
S S S S S S S S S S S S S S S S S S. S S N
SS I SS N. SS N. SS N. SS S. SS N. .S N. SS 2 SN S SN S .
2 I 2 I S. I 2 I 2 I 2 I 2 I 2 I 2 I .2 I

3.30 8.30 3.30 3.30 3.30 8.30 .230 3.30 .230 3:8

3:..U 3:5 3:5 3:..0 3:20 3:5 3:5 3:5 3:20 3:20

S0 SS: :0 SS: :0 SS: :0 SS: :0 SS: :0 SS: So SS: .8 SS: So SS: S0 SS:

o. S S S S m S S N . 5.52
:05303 €05.30qu :05803 COSSQQJ IBEUNA :03303 :03503 :03303— COSSUS 205303: 205803
SSS: I SSS: 8.3m ...—9.2.2.292 E... ...—38332 :3. =<
.5: 55525 .3 :5 83:38.2 3:. AI: ...—.0 ...—So: ...... .28....— _a>oEo~_ 3.5 5:525 E53.» .m— 03:.

28

WI]. Con

The
factors FirS'
Second, can
milk fat, and
lactation, Th
are reduced 2
genetic impr.
discounts, in;

genetic imprc

 

VIII. Conclusions

The ﬁnancial feasibility of reducing health culls is complicated by two competing
factors. First, reducing culls saves producers in making replacement heifer expenditures.
Second, cattle increase in milk production through the ﬁrst ﬁve lactations. Thus, milk,
milk fat, and milk protein per cow should increase with increased longevity until the ﬁfth
lactation. Third, as health culls are reduced, the milk losses associated with health culls
are reduced as well. Disincentives exist in the form of increasing somatic cell counts and
genetic improvement rates. Somatic cell counts, which can possibly result in milk price
discounts, increase with cow age. As fewer cattle are replaced each year, the overall

genetic improvement rate for the herd decreases.

29

EXP

I. Intri

In th
cow and her.
dairy cow n.
mortality rea
productive cl
salvage Value
I’“Provemen:
ﬁve Midwestc
NOHheastem 5
the 1993 - 19.

11‘ Metho

A varié.
bEing culled l1
cow A mOde] ‘
the explanamn

BOIh [hi

These esrimatecl

 

lot-6r limits Tl.

 

CHAPTER 3.
EXPLAINING WHY INDIVIDUAL DAIRY COWS ARE CULLED IN
NORTHEASTERN AND MIDWESTERN DAIRY HERDS

I. Introduction

In this chapter a model is developed and estimated to determine which individual
cow and herd characteristics signiﬁcantly contributed to the likelihood that an individual
dairy cow was culled due to low production, health (including reproductive health) or
mortality reasons. The independent variables of this model included individual cow
productive characteristics, farm characteristics, output and input prices, as well as cattle
salvage values and acquisition prices. The model was estimated using Dairy Herd
Improvement data from Dairy Management Records Systems for participating herds in
ﬁve Midwestern states (Illinois, Indiana, Iowa, Michigan and Wisconsin) and ﬁve
Northeastern states (Maine, New Hampshire, New Yorlg Pennsylvania, and Vermont) for
the 1993 — 1999 period.
I]. Methods and Model Development

A variety of individual and herd level characteristics can contribute to a cow
being culled. It is the cumulative effect of these reasons that causes a manager to cull a
cow. A model was needed that would predict the probability that the cumulative effect of
the explanatory variables had exceeded the needed threshold value for a cow to be culled.

Both the logit and probit models predict the probability of an event occurring.
These estimated probabilities for the dependent variable remain within the upper and

lower limits. The marginal effects are nonlinear (Gujarati, 1995). Both methods also

30

accomrt

both are

function

dichotorr

the logist

cumulatit
distributic
Emmemz
sample siz
research ;
P( l

“he

Bo

(Bl:

T0 es

the fonOWlnc
Marci

“lime

accommodate non-normally distributed explanatory variables (Maddala, 1992). As such,
both are suitable for this analysis.

The logit and probit model differ in that the normal cumulative distribution
function underlying the probit model approaches the upper and lower limits of the
dichotomous variable more quickly than the logistic cumulative distribution ﬁrnction of
the logistic model (Gujarati, 1995). Thus, the tails of the probit model’s normal
cumulative distribution function are ﬂatter than the logit model’s logistic cumulative
distribution ﬁrnction. This means that the probit model becomes more favorable as
sample size increases and there are more observations in the tail (Maddala, 1992). As the
sample size for this model was large, the probit model method was selected for this
research. A generic probit model can be described as follows:

P(Event=1)=[30 +(B 1X1+ ...+B,,X.,)

where:

P( Event = 1) refers to the predicted probability (range = O to 1) that an event

will occur;
[3 0 refers to the intercept for the probit model;
(B 1 X 1 + ...+ B n X n) refer to a set of independent variables that may

inﬂuence whether a cull will occur.

To calculate the marginal effect of a parameter in a probit model, one must use
the following formula:

Marginal effect of the nth parameter = B“ (D (Yi)

where:

[3,, refers to the coefﬁcient of the nth parameter;

31

(D ('

Yr
SAS statist;
A drawback
baseline situ
Model Desc
The ;
follows:
P( Cu.
Where “P( (‘1
event will on
.. ' l3 "X a) ”
cull will occu
Calving Seasi
The Cll
Fall) affects th

Brian, Beede ‘

 

prOdUCllon‘ CO
decisions fOr F
or Winter ma}.
COW that freShe

hone“ Part oft

<1) (.) is the density function of the standard normal variable; and,

Y, refers to the regression model used in the analysis (Gujarati, 1995).
SAS statistical soﬁware was chosen for this research and calculated the marginal effects.
A drawback of this model is that the marginal effect must be explained in terms of the
baseline situation as opposed to a general situation.
Model Description

The probit model used to estimate the likelihood of a cull is summarized as
follows:

P(Cull=1)=Bo +(BIX1+ ...+[3an)
where “P( Cull = 1) " refers to the predicted probability (range = O to 1) that a culling
event will occur, “,6 0” refers to the intercept for the probit model, and “(ﬂ 1X 1 +
+ ﬂ ”X ,,) ” refer to a set of independent variables that may influence whether a
cull will occur. These variables are described in Table 16.
Calving Season

The climate differences characterizing each season (Spring, Summer, Winter, and
Fall) affects the proﬁtability of cattle producing in those conditions. Delorenzo, Spreen,
Bryan, Beede and Van Arendonk (1992) showed that seasonal variations in milk
production, conception, and milk price affected the proﬁt maximizing replacement
decisions for Florida herds. For example, a cow calving during the extremes of Summer
or Winter may undergo a more stressful calving than in Spring or Fall. Alternatively, a
cow that freshens in late Spring and enters peak milk production and is bred during the

hottest part of the year may not peak as high or

32

Table 16.

W

. Winter Ca
7 Summer C
gFall Calvir
‘ Lactation r
Milk Diffe

Fat Differe

 

W

. Persrstency
SCC Difiere

\
Previous Ser
: Conception

W

 

 

 

 

 

 

 

 

 

 

Table 16. Independent Variable Descriptions for the Probit Model
Estimation Used to Predict the Probability of a Cow Being Culled

Parameter Description

Winter Calving l = yes, 0 = no

Summer Calving 1 = yes, 0 = no

Fall Calving 1 = yes, 0 = no

Lactation n 1 = yes, 0 = no where n = 2 through 10 or more

Milk Difference The 305 Day ME Milk hundredweight difference between

the cow and its herd mates

 

Fat Difference

The 305 Day ME Fat difference in pounds between the
cow and its herd mates

 

Protein Difference

The 305 Day ME F at difference in pounds between the
cow and its herd mates

 

Persistency Percent

A DHI measurement indicating how well the observation
cow maintained its production throughout its lactation as
compared to its herd mates

 

SCC Difference

The difference between the cow’s somatic cell count score
and the average somatic cell count score for its herd

 

Previous Services Per

The difference between the observation cow’s previous

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Conception lactation services per conception and the previous
lactation services per conception of cattle in its herd and
age group (Lactations 1, 2, or 3 or more)

PTA Milk A DHI measure indicating the cow’s ability to transmit
milk production traits to its offspring

PTA Fat A DHI measure indicating the cow’s ability to transmit fat
production traits to its offspring

PTA Protein A DHI measure indicating the cow’s ability to transmit
protein production traits to its offspring

Guernsey 1 = yes, 0 = no

Jersey 1 = yes, 0 = no

Other 1 = yes, 0 = no

Registered Herd 1 = yes, 0 = no

Illinois 1 =Les, 0 = no

Iowa 1 = yes, 0 = no

Michigan 1 = yes, 0 = no

Wisconsin 1 = yes, 0 = no

Maine 1 = yes, 0 = no

New Hampshire 1 = yes, 0 = no

New York 1 = yes, 0 = no

Pennsylvania 1 = yes, 0 = no

Herd Size Indicates the average number of milking and dry cows that

 

were in the cow’s herd

 

33

 

Table 16 (C
W
l

l

 

Heifer Ram
l

l

Replacement
W"
l leaf n

Farm '1

breed back 35
affeCts the P“
Winter, Sumn
the Spring cal .
estimation.
Cow Age Var:

The prc:
produce as mu;
likely to be culI

llllElllllly, and c

 

 

alien the likelii

 

 

Table 16 (cont’d).

 

Small Expansion Year n

Cow observation was from a farm that was in the nth year
of an expansion of more than 20 percent but less than 300
percent; 1 = yes, 0 = no where n = 1 through 5

 

Large Expansion Year n

Cow observation was from a farm that was in the nth year
of an expansion of more than 20 percent but less than 300
percent; 1 = yes, 0 = no where n = 1 through 5

 

Heifer Ratio Year n

One current and four lagged parameters measuring the
proportion of the number of dairy replacement heifers
herd greater than thirteen months of age relative to the
number of the milking and dry cow herd size on the
observation cow’s farm

 

Milk Feed Price Ratio n

The milk feed price ratio during the nth lactation month
where n = 1 through 12 or more

 

Cull Cow to
Replacement Heifer
Price Ratio n

The cull cow to replacement heifer price ratio during the
nth lactation month where n = 1 through 12 or more

 

 

 

 

Year n Cow observation was in year n; 1 = yes, 0 = no where n
equals 1996, 1997, 1998, 1999
Farm n Cow observation was from farm n; l = yes, 0 = no

 

breed back as soon as its herd mates. As such, it is important to address how seasonality

affects the probability of a cow being culled. It was hypothesized that the effects of

Winter, Summer and Fall calving season variables would be signiﬁcantly different from

the Spring calving season effect, which was set as the default calving season in the

estimation.

Cow Age Variables

The proﬁtability of that cow changes as it ages. She may become unable to

produce as much milk as she once did. In Chapter 2, it was shown that cattle are more

likely to be culled as they age. The cow may become more susceptible to injury,

infertility, and disease. All of these factors affect the cow’s proﬁtability, which, in turn,

affect the likelihood of a cull.

34

 

Pa!
reproducll‘
Production
Dykhuizen
reasons (ml
(2001) W
Hume." and
increased “I

One i
cow There a.
Hansen (3002
has not varied
number of law
should be posi
encountered di

The de

 

hypothesized If
the tint lactati.
llilk Productii

Van Arc
Should nor affec
results for L‘nite

~0Und that large

 

 

Past researchers have also shown that the incidence of culling due to health,
reproduction, somatic cell count (a measurement of udder health and milk quality), and
production efficiency increases with age. Dutch data used by Van Arendonk and
Dykhuizen (1985) showed that the probability of an animal being culled due to health
reasons (injury, disease, somatic cell count, and reproduction) increased with age. Jones
(2001) found that the somatic cell count of Wisconsin cattle increased with age. Bauer,
Mumey and Lohr (1993) determined that the costs of producing milk for Alberta cattle
increased with age until the eighth lactation.

One important decision in developing this model was how to value the age of a
cow. There are two alternatives, by normal time (age in months or years) or by lactation.
Hansen (2002) reported that some studies show that the average age of cattle at removal
has not varied signiﬁcantly over time even though culling rates have risen and the
number of lactations completed by removal have decreased. As the number of lactations
should be positively correlated with the number of peak milk production periods
encountered during a cow’s lifetime, lactation number was chosen to represent cow age.

The default cow age variable for the estimation was the ﬁrst lactation. It was
hypothesized that later the effect of later lactations would be signiﬁcantly different than
the ﬁrst lactation.

Milk Production Variables

Van Arendonk (1985) showed that the herd production level of Dutch dairy farms
should not affect the optimal cow longevity and culling rate. Stott (1994) found similar
results for United Kingdom dairy herds. Rogers, Van Arendonk, and McDaniel (1988)

found that larger milk yields supported only slightly higher culling rates for dairy farms

35

in the Um-
supported
Weigel am
culling of}
terms of mi
It is
however-I th
instanCe~ a C
whose Cattle
cow in a he“
cow Becaus|
protein yield
and milk prof
F0, milk prod

its herd mates

 

measurements
production as 1
aday and had
1999). Unless I
protein it is EX.

longer.

 

 

Lactatio

percentage is a ’I

 

in the United States. Work conducted later by Jones (2001) on Wisconsin dairies also
supported the premise that average milk yield had little effect on optimal culling rates.
Weigel and Palmer (2003), however, reported a positive correlation between involuntary
culling of high producing cows, deﬁned as cattle within the top 20 percent of their herd in
terms of milk production and the herd level Rolling Herd Average.

It is the relative productivity of an individual cow as compared to its herd mates,
however, that should affect the probability of whether that cow will be culled. For
instance, a cow that produces 25,000 pounds of milk may be a poor cow to a producer
whose cattle average 27,000 pounds of milk per lactation. Alternatively, a 25,000 pound
cow in a herd averaging 18,000 pounds of milk per lactation would be deemed a good
cow. Because U.S. milk producers are paid on the basis of ﬂuid, butterfat, and milk
protein yield — the effect of the differences in ﬂuid milk production, butterfat production,
and milk protein production on the likelihood of an animal being culled were determined.
For milk production, the difference in the 305 ME Milk between the individual cow and
its herd mates were analyzed. The 305 ME Milk, Fat and Protein are production
measurements that standardize production per cow so that the values represent the
production as if all of the cattle were the same age, from the same location, milked twice
a day and had calved during the same season (Dairy Records Management Systems,
1999). Unless there is a economic disadvantage to produce either milk ﬂuid, fat, or
protein, it is expected that a cow producing more than her herd mates will stay in the herd
longer.

Lactation persistency was also considered in the model. The lactation persistency

percentage is a DHI measure that compares the persistency of a cow’s lactation with

36

thosr
prodl
the ill
howe
short .
milk p
more I
persisr
Renket
long-or
that lacr

Herd H

Count sc
somatic
0f White
05infecr
Udder in
SCOres 1
and me
milk pro
somatic l

I’Ouhds F

those of her herdmates. When trying to judge whether an animal should be culled, a
producer should take into account the expected future proﬁtability of the cow currently in
the herd and its potential replacement. In periods of poor milk price cost margins,
however, producers may be tempted to cull lower producing cattle that are not covering
short run variable costs even if a replacement is unavailable. If two cows have identical
milk production per lactation, but one is more persistent that the other, the one that is
more persistent will have an increased probability of remaining in the herd as the less
persistent cow will be less likely to cover variable costs in late lactation. Dijkhuizen,
Renkema and Stelwagen (1985) determined that it became more advantageous to retain a
long-open cow and continue breeding the animal as persistency increased. It is expected
that lactation persistency is negatively correlated with culling rate.
Herd Health Parameters

Two DHIA herd health measures were used as explanatory variables, somatic cell
count score — which is a measure of milk quality — and services per conception. The
somatic cell count score is based on the somatic cell count, which measures the number
of white blood cells per milliliter of milk. As this number rises, it indicates the presence
of infectious agents that the cow’s immune system is ﬁghting (DRMS, 1999). Mastitis, an
udder infection, is strongly correlated with somatic cell counts and somatic cell count
scores. In the US, milk processors pay producers a premium for low somatic cell counts
and assess a discount for high somatic cell counts. High somatic cell count scores hinder
milk production. For somatic cell scores greater than or equal to three, corresponding to a
somatic cell count of between 72,000 and 141,000 — there is an expected milk loss of 1.5

pounds per cow per day (DRMS, 1999). This decrease in milk production increases

37

through th
4,524,000
(DRMS, 1'
H01
culling Stra
cull them fc
mastitic catt
percent Thi
masritis.
Whil
not effect the
cow ’5 somati
that [he COW
relative to t h t
prOdUCllOn 31".
Parameter Via
the individuaj
diffeierice par

being Culled

CaSsell

 

c. |
138011”), are C

 

through the somatic cell score of 9 - corresponding with a somatic cell count of between
4,524,000 and 9,045,000 and an expected milk loss of 10.5 pounds per cow per day
(DRMS, 1999).

Houben, Huirne and Dijkhuizen (1994) examined the effect of mastitis on optimal
culling strategies for Dutch herds. They found it more proﬁtable to treat animals than to
cull them for mastitis. When the incidence of mastitis was increased by ﬁfty percent and
mastitic cattle were voluntarily removed, the mastitis incidence still increased by ﬁfty
percent. This was due largely to the fact that many of the replacement animals contracted
mastitis.

While the overall incidence of mastitis and high somatic cell count scores should
not effect the average culling rate and optimal cow longevity, the difference between a
cow’s somatic cell count and that of her herd mates should contribute to the likelihood
that the cow will be culled. As an individual cow’s somatic cell count score increases
relative to the rest of the herd, the cow becomes less proﬁtable due to treatment costs, lost
production and somatic cell count discounts. The somatic cell count score difference
parameter was calculated by subtracting the herd average somatic cell count score ﬁ'om
the individual cows somatic cell count score. As such, the somatic cell count score
difference parameter is expected to be positively correlated with the probability of a cow
being culled.

Cassell (2002) noted that cows that become pregnant easier than their herd mates
remain in the herd longer. Thus, a measure was needed to represent fertility. Two
common reproductive capability measures monitored by DHI that proxy reproductive

capability are calving interval and services per conception. Calving interval refers to the

38

length of time I
the decision of
voluntary waiti
time. This mak
across farms,
The set
insemination s
COWS producir
Comparable sk-
Cows should a
conception an
Thedi
CO“ and her I
I«Marion 3 or
The PICVIOug
kEpt f0r “Umi
reasons Other
Neverthelesg
cow that Was
more critical]
As a 1
emClency, ll

length of time between successive calvings. The calving interval is heavily inﬂuenced by
the decision of when to start breeding a cow. A producer may elect to have a longer
voluntary waiting period prior to breeding at one point during the year than at another
time. This makes calving interval a potentially poor parameter of reproductive efficiency
across farms.

The services per conception measure, however, measures how many artiﬁcial
insemination services were required before a cow becomes pregnant. Assuming that two
cows producing the same amount of milk are bred at the same time by technicians with
comparable skill, the difference in the services per conception measurement between the
cows should also proxy the fertility difference between the cows. This makes services per
conception an appropriate measure of reproductive efﬁciency for this study.

The difference in services per conception from the prior lactation between the
cow and her herd mate of the same maturity classiﬁcation (Lactation 1, Lactation 2, or
Lactation 3 or more) was used to represent an individual cow’s reproductive efficiency.
The previous lactation service per conception difference was used because cows that are
kept for numerous services in the current lactation are generally being kept in the herd for
reasons other than reproductive efﬁciency, and this would adversely affect the analysis.
Nevertheless, as eluded to in earlier work by Van Arendonk and Dijkhuizen (1985), a
cow that was a difﬁcult breeder last lactation but ﬁnally became pregnant may be viewed
more critically during the present lactation.

As a larger, more positive number indicates that a cow has less reproductive
efﬁciency, it is expected that the effect of the previous lactation service per conception

parameter is positively correlated with the probability of a cull in the current lactation.

39

Genetic C3931
As dait“,
production and
and cow PTA F
products. PTA
transmit a giver
likelihood of an
milk. fat and pr:
Herd Breed C I
For this
“Other” dairy br
Jersey cattle had
had significantly
have inﬂuenced
Opportunity to CI
cattle from regis

probability is dif
cattle with very c

PYOdUCtion merit

lTSIIrriation. Altho

Genetic Capability

As dairy farmers typically get paid based on ﬂuid milk production, butterfat
production and milk protein production - the measures of cow PTA milk, cow PTA fat
and cow PTA protein were chosen to represent the cow’s genetic ability to produce those
products. PTA stands for the predicted transmitting ability and estimates the ability to
transmit a given trait (DRMS, 1999). As higher genetic capability should reduce the
likelihood of an animal being culled, it was hypothesized that the effects of the cow PTA
milk, fat and protein parameters are negatively correlated with culling probability.
Herd Breed Characteristics

For this model, four breed categories were used: Holstein, Guernsey, Jersey, and
“Other” dairy breeds. Chapter 2 showed that mean culling rates differed across breeds.
Jersey cattle had signiﬁcantly lower culling rates than Holstein cattle. Guernsey cattle
had signiﬁcantly higher culling rates than Holsteins. Nevertheless, farm level effects may
have inﬂuenced those averages. Having long lived breeds gives the manager a greater
opportunity to cull for production or sold for dairy purposes. This is especially true for
cattle from registered herds. However, even with registered herds, predicting culling
probability is difﬁcult. Less desirable cattle may be culled quicker on these herds, and
cattle with very desirable traits may be kept in the herd for longer than their milk
production merits. The Holstein breed was chosen to be the default breed in the
estimation. Although the correlation of the effect of breed on the likelihood that a cow
would be culled could not be hypothesized, it is expected that the effect of the Guernsey,
Jersey, and Other breed variables is signiﬁcantly different from Holstein cattle, which

served as the default breed in the estimation.

4o

Registeer H‘
Renglt

milk and high
culling strategi
registration sta'
hypothesized tl
State of Or igir
The date

Indiana Iowa. l
Hampshire. .\'e\
Northeastern stz
like Indiana. the
Alternatively, th
expensive to rep
afarm resides c2
Two estii
Northeastern star
C0mputational cc
NOhheastern stat
divide the large d
had lower culling
to see how each p.

r . .
egional difference

Registered Herd

Registered herds, as opposed to commercial herds, emphasize producing both
milk and high genetic capability cattle. In the previous section, it was discussed how
culling strategies may differ on registered herds. Although the correlation of the effect of
registration status on the likelihood of a cull could not be hypothesized, it was
hypothesized that registered herds cull differently than commercial herds.

State of Origin

The data contained dairy cattle records from ﬁve Midwestern states (Illinois,
Indiana, Iowa, Michigan and Wisconsin) and ﬁve Northeastern states (Maine, New
Hampshire, New York, Pennsylvania, and Vermont). Chapter 2 indicated that the
Northeastern states had lower culling rates than Midwestern herds. In Corn Belt states
like Indiana, the hot, humid summers may increase the likelihood of a cow being culled.
Alternatively, there is less dairy infrastructure in this state, which may make it more
expensive to replace cattle as opposed to Wisconsin, Michigan or Vermont. Thus, where
a farm resides can affect the likelihood of an animal being culled.

Two estimates were run, one for the Midwestern states and one for the
Northeastern states. This was done for a variety of reasons. First, there were
computational constraints to running the entire dataset at once. As the Midwestern and
Northeastern states were separated geographically from each other, it made sense to
divide the large dataset by geographical region. Second, the Northeastern states generally
had lower culling rates than the Midwestern states (Chapter 2). It was deemed important
to see how each parameter differed between the two regions rather than having the

regional differences absorbed by each state independent variables. Third, the two regions

41

differ in “SOL
decisions-

For thr
and \l’isconSir
as the default 5
Hampshire, Ni
Northeastern d
Herd Size and

There a
First. as herds g
individual care
that larger herd:
large farms to p
DHIA data. nor:
culling rates for
percent Kluth ft

In part to a reluc

cow care (Dairy

PrOducing cattle ‘
th '
an m Smaller he

Exceeded 300 cov

differ in resource endowments and dairy policy, which may also inﬂuence culling
decisions.

For the Midwestern estimation, it was expected that the Illinois, Iowa, Michigan
and Wisconsin culling probability effects differed from the Indiana effect, which served
as the default state for the estimation. It was hypothesized that the Maine, New
Hampshire, New York, and Pennsylvania effects differed from the effect of the
Northeastern default state, Vermont.

Herd Size and Expansion

There are two basic arguments concerning the effects of herd size on culling rates.
First, as herds get larger, the managers cannot devote the time needed to provide the
individual care that each animal needs to achieve a long herd life. The second argument is
that larger herds can afford more specialized labor and technologies that enable these
large farms to provide better overall healthcare for their cattle. Quaiffe (2002), quoting
DHIA data, noted that culling rates for herds with 600 or more cows were larger than the
culling rates for herds with fewer than 100 cows, but that the difference was only 2
percent. Kluth found that herd size was positively correlated with Idaho culling rates due
in part to a reluctance by the dairy owners to hire enough employees to provide adequate
cow care (Dairy Proﬁt Weekly, 2002). Weigel and Palmer (2003) found that high
producing cattle were more likely to be culled in larger herds due to herd health problems
than in smaller herds. Chapter 2 showed that average culling rates increased after herds
exceeded 300 cows. Because of the potential mixed effects of herd size on the probability
of a cull, the correlation of the herd size effect on culling probability could not be

hypothesized.

42

Dairy
rates and high
imponance for
cow being CU“
probabilil.V “in
Wisconsin and
years of an EXP
cull decreases C-
ldaho herds exp
underlying will
further noted the
terminal lactatio
(Dairy Proﬁt We
were more likely
expansion dairies

Two grou

hypothesis that e)

culled. The first e

Increased herd $er

Dairy farm expansion has been cited as a contributing factor to increased culling
rates and high replacement heifer prices (Hoard’s Dairyman, 2003). The issue of
importance for this study is not whether expansion increases the likelihood of an dairy
cow being culled. Instead, the true issue is determining when the increase in culling
probability will occur following an expansion. Hadley (2001) found that dairy farms in
Wisconsin and Michigan actually experienced a decrease in culling rates the ﬁrst two
years of an expansion. However, this does not necessarily mean that the likelihood of a
cull decreases during the ﬁrst two years of an expansion. Kluth noted that expanded
Idaho herds experienced an initial decrease in culling rates but also found that the
underlying culling probability for a given lactation increased following expansion. Kluth
further noted that as the initial replacement heifers that fueled the expansion reached their
terminal lactation, the expansion dairy farm experienced a sharp increase in culling rates
(Dairy Proﬁt Weekly, 2002). Weigel and Palmer (2003) reported that expansion dairies
were more likely to involuntarily cull high producing cattle due to health problems, but
expansion dairies were less likely to cull low producing cattle.

Two groups of ﬁve expansion dummy variables were designed to test the
hypothesis that expansion is positively correlated with the likelihood of a cow being
culled. The ﬁrst expansion group, “Small Expansion n, ” included those herds that
increased herd size by between twenty and three hundred percent. Dummy variables for
this expansion group represented the ﬁrst through ﬁfth years of an expansion. The second
group of expansions, “Large Expansion n, ” consisted of those farms that expanded by

more than three hundred percent. A dummy variable was also assigned to represent the

43

first through ﬁﬁh

cow being cull inc

 

Replacement Hei

As the ratir
number of milking
surplus heifers to
combination of b<
rates to accommc
do cull heavier tc
more proﬁtable t
cattle to other far
producers who c
premiums The a
costly to find an
markets Seconc
than what is typ

1f manag
is less prOﬁIablr
“MW for vari
do so bECause t}

e'.‘ , .
\erl hEIfer IO ,

ﬁrst through ﬁfth years of a large expansion. It was hypothesized that the likelihood of a
cow being cull increases with expansion.
Replacement Heifer Availability

As the ratio of the number of heifers beyond 13 months of age to the average
number of milking and dry cow herd grows, a manager may choose to allow all of the
surplus heifers to enter the herd through increased culling, sell the surplus heifers, or do a
combination of both. Radke and Lloyd (2000) asserted that many producers adjust culling
rates to accommodate the number of springing heifers they have available. If managers
do cull heavier to accommodate surplus heifers, they are indicating that they believe it is
more proﬁtable to have the heifer enter their herd than to sell them as replacement dairy
cattle to other farms. Miranda and Schnitkey (1995) found evidence to suggest that Ohio
producers who cull heavier than predicted do so because of unobserved replacement
premiums. The authors suggested two possible reasons for these premiums. First, it is too
costly to ﬁnd and negotiate with heifer buyers in underdeveloped replacement heifer
markets. Second, the producers may place a larger rate-of-return for genetic improvement
than what is typically used in dairy cattle culling research.

If managers do not allow every heifer to enter the herd, they are indicating that it
is less proﬁtable to do and more proﬁtable to sell them. Managers may engage in this
activity for various reasons. First, managers who engage in this replacement strategy may
do so because they believe that the genetic improvement rate is not high enough to allow
every heifer to enter their herd. Second, these managers may perceive that increased cow
longevity increases their returns on their heifer development investment. A third

justiﬁcation for this behavior may be that the marketing costs involved in selling

44

replacement dairy
that the replacem.
herd, which meat
culling statistics
An expla
heifers over 13 r
months) to the a
to determine the
acow being rep
him the likelihc
make their culli
costs to ﬁnd bu
genetic mm\'
the herd. ”the
small, it either
that the heifers
mature 00w IO
heifer grOuP ir
analysis. It IS 6
The Relative

Renk e:

Optima] r ePlac
\v

replacement dairy cattle are low with regard to the price they receive. A fourth reason is
that the replacements have to be culled due to health reasons prior to entering the milking
herd, which means that these animals never enter the milking herd or the milking herd
culling statistics.

An explanatory variable, “Heifer Ratio n " representing the ratio of the number of
heifers over 13 months of age (corresponding to cattle that will calve within 10 to 13
months) to the average annual milking and dry cow herd size (heifer ratio) was developed
to determine the effect that the number of available replacements has on the likelihood of
a cow being replaced. If the marginal effect of this heifer ratio is positively correlated
with the likelihood of a cull and relatively large, it may indicate that producers tend to
make their culling decisions on the basis of the number of available heifers, that search
costs to ﬁnd buyers for their replacement heifers are too high, and/or that the perceived
genetic improvement rate is high enough to justify letting all replacement heifers to enter
the herd. If the parameter’s effect is insigniﬁcant or signiﬁcantly correlated but relatively
small, it either indicates that producers ﬁnd it more proﬁtable to sell surplus heifers or
that the heifers are culled prior to entering the milking herd. As a culling event for a
mature cow today may be triggered by the fact that the cow was part of a relatively large
heifer group in years past, four lagged heifer proportion variables were also used in the
analysis. It is expected that the culling rate is positively correlated with the heifer ratios.
The Relative Price of Output and Variable Inputs

Renkema and Stelwagen (1979) concluded that the effects of milk price on
optimal replacement decisions were small for Dutch dairy farms. Later work by Rogers,

Van Arendonk and McDaniel (1988) concluded that the same was true for US. dairy

45

farms. Bauer. 5

Alberta herds “

optimal longCVi

conjunction W11

If one vi

sea change in ”

however, a Pmd
return does not i
instance, the mil
will be able to or
lactation month ‘
months of lactati
magnitude of the
feed price ratio r-
each lactation mt
likelihood of a cc

The Relative Pr

farms. Bauer, Mumey and Lohr (1993) found that the optimal terminal lactation for
Alberta herds was not affected by changes in milk price. Jones (2001) did ﬁnd that the
optimal longevity for Wisconsin dairy cows changed with milk price but only when in
conjunction with opposite changes in replacement heifer prices.

Ifone views the culling decision as a cow replacement decision, then the effects
of a change in milk price should be small. During periods of low milk price-cost margins,
however, a producer may cull an animal without a replacement if the cow’s ﬁnancial
return does not exceed or equal the variable costs associated with that production. In this
instance, the milk feed price ratio can serve as an important indicator as to whether a cow
will be able to cover its variable costs. By having milk price ratio variables for each
lactation month from calving through twelve months-and-beyond, one can judge which
months of lactation act as trigger months for a culling decision by looking at the
magnitude of the marginal effect of the signiﬁcant parameters. As an increasing milk to
feed price ratio represents an easier ability to cover variable costs, the marginal effect of
each lactation month’s milk price ratio is expected to be negatively correlated with the
likelihood of a cow being culled.

The Relative Price of Replacement Heifers

While milk and variable input prices have been shown by previous researchers to
have little effect on culling rates, the prices of replacement heifers and cull cattle have
been shown to affect culling rates and optimal cow longevity. Renkema and Stelwagen
found that cull cow prices had a small effect on optimal culling policies for Dutch herds.
(1979). Rogers, Van Arendonk and Stelwagen (1988) showed that replacement heifer

prices should have a large effect on culling decisions for US. dairy farm managers.

46

Bauer, Mqu
and/or a large .
terminal lactat:
UK. herds wa:

As the
replacement he
perspective to r
the cull cow pr
likelihood of a
COW price to re
are important it
etTect and the s
rePlacement he
Trend Variabl
AS the (f
dummy Vari able
Whether there “
associated With

Year for the estir

Bauer, Mumey and Lohr (1993) estimated that changes in replacement heifer prices
and/or a large decrease in cull cow prices should have large effects on the optimal
terminal lactation for Alberta producers. Stott (1994) found that the optimal herd life for
UK. herds was also sensitive to changes in replacement heifer prices.

As the price farmers receive for selling cull cattle increases relative to
replacement heifer price, it becomes less expensive from a capital expenditure
perspective to cull and replace a given animal. It is expected that the marginal effect of
the cull cow price to replacement heifer price ratio to be positively correlated with the
likelihood of a cow being culled. As with the milk feed price ratio, by using monthly cull
cow price to replacement heifer price ratios, one can determine which lactation months
are important in the culling decision process by looking at the magnitude of the marginal
effect and the signiﬁcance of the parameter. It was hypothesized that the cull cow to
replacement heifer price ratio is positively correlated with the likelihood of a cull.
Trend Variable

As the data consists of individual, herd and price information from 1995 — 1999,
dummy variables were established for each year. These annual variables will indicate
whether there was a particular trend in culling rates or whether any particular year was
associated with an increase (decrease) in culling rates. 1995 was chosen to be the default
year for the estimation when the 1996 thorough 1999 dummy variables were set to zero.
It is expected that the effect of 1996, 1997, 1998, and 1999 differs from 1995.

Fixed Effects Variables
The data included multiple observations from individual farms. In order to keep

the previous independent variables from capturing effects caused from farm

47

characteristic:
each farm obs

hypothesized

III. Estim
The Midwest

The pr
cattle from Mi
Wisconsin TI
observations it
estimation had

independent 0:

Table 17.

Obsen'ations
’5 Uare
Likelihood Rat

N

egrees of FIG:

7\\

. umber OfCOr

N

umber Of COI-

N

characteristics not included in the model, a ﬁxed effect dummy variable was included for
each farm observation. The correlation of the ﬁxed effects variable could not be

hypothesized.

HI. Estimation Results
The Midwestern Estimation

The previously described probit dairy cattle culling model was estimated for dairy
cattle from Midwestern DHIA dairy farms in Illinois, Indiana, Iowa Michigan, and
Wisconsin. The overall model results are summarized in Table 2. There were 432,444
observations in which 160,135 culling events and 272,309 nonevents occurred. The
estimation had to drop 2,797,017 observations due to missing information regarding the

independent or dependent variables.

Table 17. Overall Results for the Probit Model Estimation of the Likelihood
that a Cow Will Be Culled on Midwestern Dairy Farms

 

Observations 432,444

 

 

 

 

 

 

 

 

R- Sjuare 0.3196
Likelihood Ratio Test Chi-Square ' 166,508

(p-value <0.0001)
Wald Test Chi-Square 107,853

(p-value <0.000I)
Degrees of Freedom 3,583
Number of Correct Culling Events Predicted 78,430
Number of Correct Culling Non-events Predicted 257,468
Percentage Correct 80.40

 

 

 

48

The e
that all Bu = C
3,583 degree?
distinguish a

304 percent <

Table
model. In ordi
probability of
baseline situat
first lactation l
with the sampl
and genetics. 1
average herd s'
1995 ~ 1999 til
equal to the sar
Chance that a f;
the termination

Calving
likelihood of a c

April. and May)

PIObabIlII}. Of a (

The estimation had a R-square of 0.5462. When testing the global null hypothesis
that all [3,, = 0, the likelihood ratio chi-square score was 166,508 (p-value < 0.0001) with
3,583 degrees of freedom. Using a probability level of 0.60 as a threshold value to
distinguish a culling event from a nonevent, the model was able to successfully classify

80.4 percent of the Midwestern culling events and nonevents.

Table 18 displays the coefﬁcient estimate results for the culling probability
model. In order to calculate the marginal effects of the independent variables on the
probability of a cull, the model was designed to provide a culling probability for a
baseline situation. The baseline situation for the Midwestern estimation consisted of a
ﬁrst lactation Holstein cow calving on a particular farm in Indiana in the Spring of 1995
with the sample average milk production capability, milk quality, reproductive efﬁciency
and genetics. The farm this representative cow came from was equal in size to the sample
average herd size of 193 cows and had not undergone an expansion prior to or during the
1995 — 1999 time period. The representative farm had a current year heifer ratio that was
equal to the sample average of 0.4974. For the baseline situation, there was a 2.58 percent
chance that a ﬁrst lactation, average producing Holstein cow would be culled during or at
the termination of the ﬁrst lactation on a particular farm in Indiana.

Calving season variables did play a signiﬁcant role in contributing to the
likelihood of a cow being culled. As compared to the Spring calving season (March,
April, and May), calving in Winter (December, January, and February) increased the

probability of a cull by 0.23 percent (p-value = 0.0008). Calving in Summer decreased

49

Table 18. C I

{ Parameter =1

l

|
,————¢—
l

lnterce t '
Winter I
lCalving .
j Summer 1
Calving . I
Fall Calving l ..
' Lactation 2
Lactation 3 l
‘ Lactation 4
Lattation 5 l _
W
l Lactation 8 .
Lattation 9
Lactation 10 l

W
‘ Difference l

 

 

.BOSMEpat { (
Difference

. 305 ME

1 -(
Protein 3
‘ Difference :
Percent l
W
ourtt SCOre :
lelitrence l
W

' C

Queeption ;
l h CienCes ‘-
Pldhtirk .-

.. A Pr0tein l
L“ClTlSey l

Chm
6537/4;
”'7'!
94’
OC)

__‘

9'70

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 18. Coefﬁcient Estimate Results for the Probit Model Estimation of
the Likelihood that a Cow Will Be Culled on Midwestern Dairy
Farms ‘

Parameter Estimate Standard Wald Test 1 p-value2 Marginal
Error Chi- Effect
Square (%)

Intercept 7.5011 0.2805 714.9159 2 <0.0001

Winter 0.0374 0.0112 11.1331 2 0.0008 0.2330

Calving

Summer -0.1903 0.0103 338.8005 2 <0.0001 -0.9500

Calving

Fall Calving -O.2666 0.0124 461.6945 2 <0.0001 -l.2360

Lactation 2 0.4513 0.0063 5115.0913 2 <0.0001 4.1660

Lactation 3 0.5983 0.0070 7281.9900 2 <0.0001 6.3040

Lactation 4 0.7222 0.0082 77213005 2 <0.0001 8.4670

Lactation 5 0.8362 0.0102 6658.8324 2 <0.0001 10.7680

Lactation 6 0.9337 0.0137 46698139 2 <0.0001 12.9840

Lactation 7 1.0837 0.0195 3092.2544 2 <0.0001 16.8390

Lactation 8 1.1594 0.0294 1553.7702 2 <0.0001 18.9890

Lactation 9 1.3292 0.0459 838.6036 2 <0.0001 24.2820

Lactation 10 1.5560 0.0632 606.3097 2 <0.0001 32.2400

305 ME Milk -0.0054 0.0002 754.5362 1 <0.0001 -0.0330

Difference

305 ME Fat 0.0003 0.00003 123.7454 1 NS 0.0020

Difference

305 ME -0.0023 0.00007 101.5806 1 <0.0001 -0.0140

Protein

Difference

Persistency -0.0147 0.00007 39346.7812 1 <0.0001 -0.0870

Percent

Somatic Cell 0.0662 0.0015 2036.7475 1 <0.0001 0.4230

Count Score

Difference

Previous 0.0341 0.0016 430.0578 1 <0.0001 0.2110

Services Per

Conception

Differences

PTA Milk 0.0015 0.0001 235.7387 1 NS 0.0090

PTA Fat 0.0281 0.0182 2.3685 1 NS 0.1730

PTA Protein 1.2520 0.0381 10807722 1 NS 21.7960

Guernsey 0.1122 0.0641 3 .0606 2 0.0802 0.7510

Jersey -0. 1743 0.0382 20.8626 2 <0.0001 -0.8840

 

 

 

 

 

 

 

50

 

1 Small
lExpanS
Year 3
______.
Small
'Expansr
tSmall
, Expansic
, Year 4
Small
, I Expansion
Wear 5
Large
. Expansion
Year)
.Large
Erpansron
Year2
\
Large
Emails/on
Year3
Latte ..
Erpansrbn /'
Year4 I
the t o
Ethan/on /
Years l
fie-fir Ratio t .[i
lea/[2
heifer Ratio p
in“ ‘

\

  
 
  
   

Table 18 (cont’d).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Other Breed -0.1601 0.0281 32.4168 2 <0.0001 -0.8230
Registered -0.0406 0.0180 5.0690 2 0.0244 -0.23 50
Herd

Illinois —1.3458 0.2301 34.1965 2 <0.0001 -2.5320
Iowa —0.5514 0.2211 6.2214 2 0.0126 -1.9570
Michigan -0.0945 0.1854 0.2595 2 0.6105 -0.5180
Wisconsin -1.4951 0.4134 13.0811 2 0.0003 -2.5530
Herd Size -0.0002 0.00005 16.3583 2 <0.0001 -0.0010
Small 0.0770 0.1006 0.5862 1 NS 0.4980
Expansion

Year 1

Small 0.1737 0.1020 2.9008 1 0.0443 1.2340
Expansion

Year 2

Small 0.0668 0.1005 0.4423 1 NS 0.4280
Expansion

Year 3

Small 0.1011 0.0968 1.0918 1 NS 0.6695
Expansion

Year 4

Small -0.0353 0.1012 0.1218 1 NS -0.2052
Expansion

Year 5

Large 0.1389 0.1586 0.7671 1 NS 0.9538
Expansion

Year 1

Large -0.4044 0.1587 6.4969 1 NS -1.6448
Expansion

Year 2

Large 0.3276 0.1526 4.6067 1 0.0159 2.6948
Expansion

Year 3

Large -0.0955 0.1667 0.3278 1 NS -0.5226
Expansion

Year 4

Large 0.4467 0.1550 8.3032 1 0.0040 4.1054
Expansion

Year 5

Heifer Ratio —0.0049 0 0202 0.0575 1 NS -0.0040
Year 0

Heifer Ratio 0.0383 0.0138 7.6660 1 0.0028 0.0228
Year -1

 

 

 

 

 

 

 

51

 

Table 18 (cont’d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Heifer Ratio 0.0018 0.0095 0.0365 1 NS 0.0008
Year -2

Table

Heifer Ratio -0.0158 0.0104 2.3196 1 NS -0.0097
Year -3

Heifer Ratio -0.0087 0.0109 0.6400 1 NS -0.0055
Year -4 -

MFR 13 -0.0802 0.0125 41.4293 1 <0.0001 -0.4455
MFR 2 -0. 1003 0.0130 59.1788 1 <0.0001 -0.5465
MFR 3 -0.0967 0.0125 60.1961 1 <0.0001 -0.5288
MFR 4 -0.0671 0.0124 29.0745 1 <0.0001 -0.3774
MFR 5 -0.0550 0.0123 19.9110 1 <0.0001 -0.3134
MFR 6 -0.2015 0.0112 324.3316 1 <0.0001 -0.9950
MFR 7 0.0271 0.0056 23.2047 1 NS 0.1668
IVIFR 8 -0.0144 0.0058 6.1236 1 0.0067 -0.0854
MFR 9 0.0132 0.0056 5.4891 1 NS 0.0801
IVER 10 0.0156 0.0061 6.4884 1 NS 0.0947
MFR 11 -0.0170 0.0061 7.8177 1 0.0026 -0.1006
MFR 12 0.0611 0.0059 107.2325 1 NS 0.3892
CR 1“ -0.4706 0.2100 5.0213 1 NS 02702
CR 2 -1.2174 0.2641 21.2459 1 NS -0.6496
CRR 3 -3.8868 0.2676 210.9043 1 NS -1.6044
CR 4 -2.8063 0.2762 103.2451 1 NS -1.2840
CRR 5 -1.0736 0.2684 16.0069 1 NS -0.5810
CRR 6 -3.8928 0.2657 214.6348 1 NS -l.6060
CR 7 -l .3217 0.2707 23.8453 1 NS -0.6981
CRR 8 -0.0006 0.2862 0.0000 1 NS -0.0007
CRR 9 -l .9446 0.3174 37.5413 1 NS -0.9668
CRR 10 -l.0584 0.3293 10.3291 1 NS -0.5735
CRR 11 -0.0243 0.3376 0.0052 1 NS -0.0148
CR 12 3.3629 0.2567 171.6191 1 <0.0001 2.7892
Year 1996 -0.2019 0.1180 2.9306 2 0.0869 -1.0007
Year 1997 0.2527 0.0133 361.1484 2 <0.0001 1.9359
Year 1998 -0.0551 0.0114 23.2176 2 <0.0001 -0.3136
Year 1999 -0.0926 0.0110 70.9526 2 <0.0001 -0.5086

 

 

T A “1” indicates the test was a one-tailed test. A “2” indicates the test was a two-

tailed test.

2 A “NS” indicates that the parameters effect was not signiﬁcant (p—value S 0.1000).
3 “MFR it” refers to the milk feed price ratio for lactation month 11.

4 “CRR it” refers to the cull cow replacement heifer price ratio for month it.

52

 

and
cor.
cart

hrs

11’ it

35fete

0”? ad:

the likelihood of a cull by 0.95 percent (p-value < 0.0001). Calving in the fall
(September, October, November) decreased the likelihood of a cow being culled by 1.24
percent (p-value = 0.0001). Given these results, if a manager were only

interested in culling rate reduction, he or she could prioritize his or her breeding program
to have proportionally more cattle calve in Fall followed respectively by Summer, Spring
and Winter. Other factors, such as milk and input price seasonal movements should be
considered prior to engaging in such calving scheduling. Because proportionally more
cattle are culled in Winter and Spring, producers may want to consider synchronizing
ﬁrst lactation cow calvings in order to counterbalance the increased culling likelihood for
Winter and Spring calving cattle.

A cow surviving to lactation 2 was 4.17 percent more likely to be culled than in
lactation 1 (p-value < 0.0001). Cattle in lactation 3 were 6.30 percent more likely to be
culled than ﬁrst lactation cattle (p-value < 0.0001). Fourth lactation cattle were 8.47
percent more likely to be culled than ﬁrst lactation cattle (p-value < 0.0001). Cattle in
lactations 5 through 10 also exhibited a signiﬁcant increased likelihood of being culled.
Cattle in lactation 5 exhibited a 10.77 percent higher likelihood of being
culled than their ﬁrst lactation counterparts. Cattle in their tenth lactation experienced a
32.24 percent higher likelihood of being culled that ﬁrst lactation cattle. Thus, as
expected, cattle age signiﬁcantly increased the likelihood of a cow being culled.

The difference in milk production between a cow and its herd mates signiﬁcantly
affected the probability of a cow being culled (p-value < 0.0001). A cow that produces

one additional hundredweight more than the average 305 ME Milk

53

yield was 0.033 percent less likely to be culled than the average producing cow. Higher
fat production actually increased the likelihood of a cow being culled, but the effect was
very small (0.002 percent). Cattle whose average difference in 305 ME

Protein were one pound higher than the average producing cow were 0.014 percent less
likely to be culled (p-value <0.0001). Cattle that were one persistency unit higher than the
average producing cow were 0.087 percent less likely to be culled (p-value < 0.0001).

It is logical that cattle producing more milk and more protein than their herd
mates are more proﬁtable, ceteris paribus, and should remain in the herd longer than the
less proﬁtable cattle. Butterfat production, even though a principal component in milk
price determination, seemed of little importance in culling decisions. In fact, cattle that
produced relatively higher butterfat yields were more likely to be culled. This may result
from managers trying to emphasize milk and protein production when making their
genetic decisions. In the past, dairy farmers primarily got paid on ﬂuid milk, butterfat,
and somatic cell count. Managers currently get paid for ﬂuid, butterfat, milk protein,
solids—non-fat, and somatic cell count. Thus, managers may now place more emphasis on
components other than fat. Nevertheless, butterfat production should be a factor in culling
decisions as it is a principal component in the milk price formula.

Milk production persistency within a lactation seemed to be a much more
important factor affecting Midwestern dairy culling probabilities than actual milk,
butterfat or milk protein production. If managers are keeping more persistent cattle in the
herd longer because they are more proﬁtable in late lactation, this might be an erroneous
strategy. Choosing the more persistent animal negates the proﬁtability of the earlier part

of the lactation curve, which may show that a less persistent cow is more proﬁtable.

54

Dairy farm managers should consider the expected ﬁiture proﬁtability of the potential
cull candidates instead of basing culling decisions on persistency. On the other hand, if
producers are keeping more persistent open cattle around for longer periods of time for
breeding purposes, the strategy of retaining more persistent cattle may be warranted.
Dijkhuizen, Renkema and Stelwagen (1985) found that this was a proﬁtable strategy.

With regard to herd health parameters, cattle exhibiting a unit higher one somatic
cell count score than the average somatic cell count score were 0.423 percent more likely
to be culled than the cow represented by the baseline situation. This is not surprising as
high somatic cell count scores decreases both the milk price per hundredweight as well as
the amount of milk produced per cow. To reduce somatic cell count and mastitis induced
culling, methods to reduce somatic cell count score need to be proﬁtably implemented.
Such methods may include implementing or improving udder preparation and post
milking care, using dry cow treatments at dry off, proper equipment sanitation, and/or
proper milking equipment maintenance.

The other herd health parameter, the number of services per conception in the
previous lactation, also signiﬁcantly increased the likelihood of a cull. Cattle that had to
be serviced one more time than the average of its herd mates in the previous lactation
were 0.211 percent more likely to be culled in the current lactation (p-value < 0.0001). To
reduce culling for reproductive failure, producers should consider proﬁtable strategies to
reduce reproductive failure. Such strategies may include devoting more time to heat
detection activities, training more employees to detect estrus, providing training to

improve artiﬁcial insemination techniques, or using estrus synchronization techniques.

55

It was hypothesized that cattle with a higher genetic capability to produce milk,
butterfat, and milk protein would stay in the herd longer than those with lower genetic
capability. The opposite proved true. Midwestern dairy cattle exhibiting a one unit higher
genetic ability to transmit milk, milk fat, and milk protein production traits were 0.009,
0.1730 and 21.796 percent more likely to be culled respectively.

The Cow PTA Milk and Cow PTA Protein results seem to conﬂict with those of the
actual milk and milk protein production. Nevertheless, Kelm (2003) noted that there was
a strong positive correlation between Type (how closely the cow appears to meet the
breed’s ideal structural standard), Cow PTA Milk, and Cow PTA Protein and a strong
negative correlation between Type and Longevity. Hansen (2002) also noted a strong
negative correlation between type and longevity. These results indicate that placing too
much emphasis on milk and milk protein genetics can decrease cow longevity. The
positive correlation between the ability to transmit milk fat production and culling
probability concurred with the correlation of actual milk fat yield and culling probability.
The fact that cattle with a higher ability to transmit milk fat production is positively
correlated with culling likelihood may indicate that dairy farm managers are de-
emphasizing fat production. Whether this is the correct decision from an economic
perspective should be assessed as fat production is still a major component of milk price
determination.

Cattle breed did seem to inﬂuence the culling probability of Midwestern dairy
cattle. If a cow was a Guernsey, it was 0.751 percent more likely to be culled than a
Holstein. Jersey cattle were 0.884 percent less likely to be culled than a Holstein (p-value

< 0.0001). Dairy cattle breeds other than Guernsey, Jersey and Holstein were 0.823

56

percent less likely to be culled than Midwestern Holstein cattle (p-value < 0.0001). If
researchers and dairy farm managers desire to improve the culling likelihood of Holstein
and Guernsey cattle, there appears to be justiﬁcation to compare and contrast their
longevity characteristics with those of the Jersey and the “other” dairy breeds.

Whether a cow was from a registered herd did affect the probability of a cull.
Cattle from registered herds were 0.235 percent less likely to be culled for reasons other
than being sold for dairy purposes (p-value < 0.0001). As dairies with registered cattle are
typically involved in the sale of cattle for dairy and genetics purposes, maintaining the
health of these cattle is very important. An animal that is a below average producer, lame,
or mastitic is not very desirable to a cattle buyer. As such, Midwestern managers of
registered cattle may give more individualized attention to their registered cattle than
commercial dairy managers.

The average ﬁrst lactation cow in Illinois, Iowa and Wisconsin were less likely to
be culled than similar cattle in Indiana. Illinois ﬁrst lactation Holstein cattle were 2.532
percent less likely to be culled than their Indiana counterparts (p—value < 0.0001). Iowa
ﬁrst lactation cattle were 1.957 percent less likely to be culled (p—value = 0.0126).
Wisconsin cattle were 2.552 percent less likely to be culled than Indiana cattle (p-value =
0.0003).

The effect of herd size on the likelihood of a cow being culled were signiﬁcant
but small. A one cow increase in herd size from the average resulted in a 0.001 percent
decrease in the likelihood of a cull (p-value = 0.0001). This means that for every 100 cow
increase in herd size, the cattle on that farm were 0.10 percent more likely to be culled.

This may be caused by a variety of factors. Managers of smaller farms generally have to

57

be a jack-of-all trades. Although they may give more individualized attention to each cow
in their herd, the manager of a smaller herd may be too constrained by other activities,
such as crop planting and harvesting, to give more individualized attention. The negative
correlation may also indicate that larger farms have an advantage in attracting more
specialized labor. Thus, being of a certain herd size may permit a farm to hire better herd
managers, on-staff veterinarians, and laborers. The negative correlation between herd size
and culling rates may also indicate that managers tend to increase their farm’s herd size
as their management ability improves.

The effect of expanding to a larger herd size on the likelihood of a cull were
mixed, however. For the smaller category of expansions (an increase in herd size of less
than or equal to 300 percent), the effect of the initial expansion year on the probability of
a cull was positive but insigniﬁcant. Cattle in the second expansion year were 1.234
percent more likely to be culled (p-value = 0.0885). The likelihood of a cull increased in
the third and fourth years, but the results were statistically insigniﬁcant. The correlation
between the ﬁfth year of a small expansion and the likelihood of a cull was negative but
also statistically insigniﬁcant.

For herds that had a large increase in herd size (a herd size increase of more than
300 percent), the ﬁrst large expansion year was positively correlated with culling
likelihood but insigniﬁcant. During the second year of a large expansion, the culling rate
actually decreased by 1.645 percent (p-value = 0.0885). The probability of a cow being
culled in the third major expansion year increased by 2.695 percent (p-value = 0.0318).

The effect of the fourth major expansion year was negative but insigniﬁcant. The

58

likelihood of a cull increased by 4.105 percent (p-value = 0.004) during the ﬁfth
expansion year.

There seems to be little statistical support suggesting that expansion of
Midwestern DHIA dairy farms increased their overall culling rates. However, it is likely
that managers, lenders and advisors involved in expansion will see increased culling rates
in year 2 of a small expansion and years 3 and 5 of a large expansion and should budget
their expansions accordingly.

For Midwestern DHIA herds, the effect of the current year heifer ratio on the
likelihood of a cow being culled was negative and insigniﬁcant. However, the effect of
the prior year’s heifer ratio on culling likelihood was positive but small (0.0228 percent).
The effect of the remaining heifer ratios on culling probability were insigniﬁcant. Thus,
there is little evidence suggesting that the number of heifers available inﬂuences culling
rates, and, when there is such evidence, the effect on culling probability is small.
Producers may sell some of the additional heifers and/or some of these heifers are lost
due to health problems prior to entering the milking herd.

The effects of the relative price of milk to feed on the probability that a
Midwestern dairy cow will be culled were mixed but mostly negative as hypothesized.
The marginal effect of a one unit increase in the price of milk relative to feed prices for
months one through six were negatively correlated with culling rates (p-value < 0.0001).
The milk to feed price ratio for lactation months 8 and 11 were also negative and
statistically signiﬁcant. The decreases in the likelihood of a cull for the signiﬁcant
lactation months ranged from 0.085 percent to 0.995 percent. This shows that producers

are milk and variable input price responsive when making their culling decisions for 8 of

59

the 12 lactation months analyzed. However, these results conﬂict with those of earlier
researchers who indicated that the effect of milk and variable input prices on optimal
culling rates should be, at the most, small. While the results of this study may be
inﬂuenced by cattle that are culled without replacement (in the case of a herd size
contraction or dispersal), it could also indicate that producers erroneously adjust culling
rates to the price of milk and variable inputs. Producer education programs may be
warranted to correct these erroneous decisions.

All of the cull cow price to the replacement heifer price ratio marginal effects
were surprisingly negative except for the ratio of lactation month twelve and beyond
(CRR 12). In lactation month twelve, a one tenth increase in the cull cow to replacement
heifer price ratio increased the likelihood that a cow would be culled by 2.789 percent.
The fact that this lactation month ratio is positively correlated with culling probability
makes intuitive sense. Cattle that have been lactating for twelve months or more are
generally cattle who are less fertile than their herd mates. As the price of cull cattle goes
up relative to the price of replacement dairy heifers, it becomes less expensive from a
capital expenditure perspective to replace the animal. It is interesting, however, that the
cull cow price to replacement heifer price ratio for the other lactation months with critical
decision point potential (calving, peak milk production, ﬁrst and second services, and
mid-lactation) were insigniﬁcant.

With the exception of 1997, the likelihood of a Midwestern dairy cow being
culled during the period of 1996 through 1999 was lower than the likelihood of being
culled in 1995. Cattle milking in 1996 were 1.001 percent less likely to be culled than

cattle milking in 1995 (p-value = 0.0869). In 1997 the likelihood of a cow being culled

60

increased by 1.936 percent. In 1998 and 1999 the likelihood of a cow being culled on
Midwestern dairy farms as compared to 1995 decreased by 0.314 and 0.509 percent
respectively. Thus, there is little evidence suggesting that there is a trend of increasing
culling rates for years 1995 through 1999.
The Northeastern Estimation

The probit dairy cattle culling model was estimated for dairy cattle from
Northeastern DHIA dairy farms in Maine, New Hampshire, New York, Pennsylvania,
and Vermont. The overall model results are summarized in Table 19. There were 225,987
culling events and 106,339 culling nonevents that took place in the 332,326 observations.
The estimation had to drop 3,525,912 observations due to missing information regarding
the independent or dependent variables.

The estimation had a R-square of 0.4879. When testing the global null hypothesis
that all [3,, = 0, the likelihood ratio chi-square score was 222,393 (p-value < 0.0001) with
3,933 degrees of freedom. Using a probability level of 0.60 as a threshold value to

Table 19. Overall Results for the Probit Model Estimation of the Likelihood
that a Cow Will Be Culled on Northeastern Dairy Farms

 

 

 

 

 

 

 

 

 

Observations 332,326
R-Square 0.4879
Likelihood Ratio Test Chi-Square 222,393.588
Wald Test Chi-Square 103,962.630
Degrees of Freedom 3,933
Number of Correct Culling Events Predicted 198,982
Number of Correct Culling Non-events Predicted 91,636
Percentage Correct 86.7

 

 

 

61

distinguish a CL"
classify 86.7 P9

Table 2".
using the data fr
estimation consr
Vermont in the 9
quality, reprodUc
from was equal i
undergone an ex;
representative fa:
04830. For the b
average producin
lactation on a part

Calving 5e
likelihood of a cot

April. and May), c

 

probability of 21 cu

the likelihood of a

01369). Calving ir

a . .
cow being culled

d'

tfierences betweer

effects the likelihoc

that calved in the F

 

distinguish a culling event from a culling nonevent, the model was able to successfully
classify 86.7 percent of the culling events and nonevents.

Table 20 displays the coefﬁcient estimate results for the culling probability model
using the data from Northeastern herds. The baseline situation for the Northeastern
estimation consisted of a ﬁrst lactation Holstein cow calving on a particular farm in
Vermont in the Spring of 1995 with the sample average milk production capability, milk
quality, reproductive efﬁciency and genetics. The farm this representative cow came
from was equal in size to the sample average herd size of 200 cows and had not
undergone an expansion prior to or during the 1995 — 1999 time period. The
representative farm had a current year heifer ratio that was equal to the sample average of
0.4830. For the baseline situation, there was an 18.08 percent chance that a ﬁrst lactation,
average producing Holstein cow would be culled during or at the termination of the ﬁrst
lactation on a particular farm in Vermont.

Calving season variables did play a signiﬁcant role in contributing to the
likelihood of a cow being culled. As compared to the Spring calving season (March,
April, and May), calving in Winter (December, January, and February) increased the
probability of a cull by 12.608 percent (p-value < 0.0001). Calving in Summer decreased
the likelihood of a cull by 0.768 percent but was statistically insigniﬁcant (p-value =
0.1369). Calving in the fall (September, October, November) increased the likelihood of
a cow being culled by 1.131 percent (p-value = 0.0805). Thus, there were apparent
differences between the Midwestern and Northeastern regions as to how calving season
effects the likelihood of cow being culled. In the Midwestern region estimation, cattle

that calved in the Fall were less likely to be culled than those that calved in Spring. As

62

Table 20.

r________.
Parameter
l

' Intercept

Winter ~
'Cdnng ;_
Summer .
Calving . .-
Fall C alving : _
Lactation 2
WT
.Lactation 5
, Lactation 6 .
Lactation 7
Lactation 8 ;
Lactation 9 -
Lacration lO - 7
+305 31E Milk I
Difference i
305 ME Fill I
Difference ;
.305 ME 5
hmun I
Difference ;
PersiStech.
Percent ' ;
“matte Cell 1 .
Gum Score ,'
lﬁelence i
Prel’iOus .
Services Per
QﬂcePtion

PT-Miilk . I
PM fat

  

 

 

l
l
l
l
l

-C
PIA Promn l -C
Chemsev I
Jenny O

(
t
O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 20. Coefﬁcient Estimate Results for the Probit Model Estimation of
the Likelihood that a Cow Will Be Culled on Northeastern Dairy
Farms
Parameter Estimate Standard Wald Test I p-value7 Marginal
Error Chi- Effect
Square (%)
Intercept 1.4084 0.4862 8.3910 2 0.0038
Winter 0.4076 0.0207 387.5490 2 <0.0001 12.6080
Calving
Summer -0.0296 0.0199 2.2120 2 0.1369 -0.7680
Calving
Fall Calving 0.0422 0.0241 3.0557 2 0.0805 1.1310
Lactation 2 0.4461 0.0092 2337.3264 2 <0.0001 13.9750
Lactation 3 0.6102 0.0101 3662. 1091 2 <0.0001 20.0500
Lactation 4 0.7028 0.0115 3705.7337 2 <0.0001 23.6240
Lactation 5 0.8139 00139 34437830 2 <0.0001 28.0010
Lactation 6 0.9010 0.0178 2570.1815 2 <0.0001 31.4700
Lactation 7 0.9307 0.0241 1487.7339 2 <0.0001 32.6530
Lactation 8 1.0019 0.0360 773.9011 2 <0.0001 35.4910
Lactation 9 1.1254 0.0535 442.3278 2 <0.0001 40.3580
Lactation 10 1.2021 0.0636 357.8267 2 <0.0001 43.3230
305 ME Milk -0.0057 0.0002 535.9905 1 <0.0001 -0.1480
Difference
305 ME Fat 0 00003 0.00004 0.3617 1 NS 0.0010
Difference
305 ME -0.0008 0.00009 83.0444 1 <0.0001 -0.0210
Protein
Difference
Persistency -0.0092 0.00009 100894595 1 <0.0001 -0.2400
Percent
Somatic Cell 0.0588 0.0020 907.6232 1 <0.0001 1.5900
Count Score
Difference
Previous 0.0222 0.0022 103.8017 1 <0.0001 0.6810
Services Per
Conception
Differences
PTA Milk 0.0008 0.0001 38.7815 1 NS 0.0220
PTA Fat -0.0684 0.0274 6.2550 1 0.0062 -1.7430
PTA Protein -0.0601 0.0531 1.2799 1 NS -1.538O
Guernsey 0.4202 0.1017 17.0586 2 <0.0001 13.0530
Jersey -0.1731 0.0416 17.3391 2 <0.0001 -4.1930

 

 

 

 

 

 

 

63

 

Table 20 (C0111.
Other Breed

I Registered
r Herd
,hhme
I New .
j Ham shire
l New York . .
1 Pennsylvania ' .
Herd Size 1
1 Expansion I
Year I .
Expansion 1'
Year2 l
‘ Expansion ,
:‘Year3 l
Expansion .
Item I
; Expansion .
5 Year 5 l
I Large I
EXpinsion I
' Year 1
Large
EXpansion
Year 2
1 Large
Expansion
: Year 3 i I
Large\‘I\.I
l fpansion .

ear 4 1
Large I .

’XPansion l

Heifer Ratio -
l' 0 l ‘

it” 3
W

van] 1 I

l
l
l
l
l

r" l

7 4 t—f“

 

rife; Ratio

ar.2 C
HEll‘er - Ii
l'tar .3Ratro I ‘0

7

_/li

 

Table 20 (cont’d).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Other Breed -0.2878 0.0458 39.5731 2 <0.0001 -6.5750
Registered 0.1138 0.0421 7.3115 2 0.0069 3.1490
Herd

Maine 0.7255 0.8444 0.7381 2 NS 24.5110
New -0.4707 0.4346 1.1727 2 NS -9.7470
Hampshire

New York -0.6989 0.4968 1.9793 2 NS -127230
Pennsylvania -1.2666 0.4294 8.7016 2 0.0032 -16.6l30
Herd Size 0.0009 0.00004 690.8677 2 <0.0001 0.0250
Expansion 0.4124 0.2028 4.1352 1 0.0210 12.7760
Year 1

Expansion -0.3461 0.1981 3.0526 1 NS -7.6680
Year 2

Expansion 0.2014 0.1959 1.0565 1 NS —0.8611
Year 3

Expansion 0.1438 0.1791 0.6448 1 NS 5.7770
Year 4

Expansion 0.1752 0.1977 0.7851 1 NS 4.0300
Year 5

Large 0.0332 0.2079 0.0256 1 NS 4.9730
Expansion

Year 1

Large 0.0096 0.2076 0.0022 1 NS 0.2550
Expansion

Year 2

Large -0.0332 0.2126 0.0244 1 NS -0.8610
Expansion

Year 3

Large -0.0588 0.2320 0.0642 1 NS -1.5050
Expansion

Year 4

Large 0.3730 0.2141 3.0346 1 0.0408 11.4030
Expansion

Year 5

Heifer Ratio -0.0050 0.0229 0.0470 1 NS -0.0130
Year 0

Heifer Ratio -0.0543 0.0171 10.0431 1 NS -0.1420
Year -1

Heifer Ratio 0.0757 0.0127 35.6629 1 <0.0001 0.2000
Year -2

Heifer Ratio -0.0094 0.0058 2.6397 1 NS -0.0250
Year -3

 

 

 

 

 

 

 

64

 

Table 20 (cont’d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Heifer Ratio 0.00169 0.01 18 0.0205 1 NS 0.0050
Year -4

MFR 13 0.1958 0.0196 99.7975 1 Ns 5.6050
MFR 2 -0.3196 00271 138.7345 1 <0.0001 -7.1810
MFR 3 -00451 0.0271 2.7664 1 0.0482 -1.1610
MFR 4 -0.0146 0.0297 0.2403 1 NS -0.3810
MFR 5 -0.2648 0.0302 77.0879 1 <0.0001 -6.1220
MFR 6 -02407 0.0305 62.4280 1 <0.0001 -5.6340
MFR 7 0.0215 0.0322 0.4446 1 NS 0.5710
MFR 8 -0.0965 00273 12.4946 1 0.0002 .2.4270
MFR 9 -0.0816 0.0286 8.1320 1 0.0022 -2.0660
MFR 10 03152 0.0275 131.5070 1 <0.0001 -7.0980
MFR 11 0.1936 00286 45.9638 1 NS 5.5360
MFR 12 0.0654 0.0245 7.1126 1 NS 1.7730
CR 14 2.8787 0.4749 36.7377 1 <0.0001 8.5370
CRR 2 -51431 0.6240 67.9257 1 NS -10.3960
CR 3 2.9584 0.6123 23.3413 1 <0.0001 8.7990
CRR 4 1.3797 0.5849 5.5635 1 0.0092 3.8570
CR 5 -0.8484 0.5957 2.0282 1 NS -21450
CRR 6 2.3845 0.5598 18.1418 1 <0.0001 6.9400
CR 7 -6.3887 0.5028 161.4687 1 NS 42.0370
CR 8 -1.6454 0.4960 11.0054 1 NS -40030
CRR 9 -20191 0.4699 18.4675 1 NS -4.8210
CRR 10 -1.4617 0.5223 7.8334 1 NS -3.5880
CRR 11 2.8043 05714 24.8044 1 <0.0001 8.2930
CR 12 97353 04382 493.5659 1 <0.0001 34.3610
Year 1996 1.5403 0.0316 23819073 2 <0.0001 55.4210
Year 1997 1.4389 0.0306 2215,2520 2 <0.0001 51.9960
Year 1998 2.0315 0.0234 75190573 2 <0.0001 68.7670
Year 1999 1.0711 00164 4261.7793 2 <0.0001 38.2270

 

I A “1” indicates the test was a one-tailed test. A “2” indicates the test was a two-

tailed test.

2 A “NS” indicates that the parameters effect was not signiﬁcant (p—value S 0.1000).
3 “MFR it” refers to the milk feed price ratio for lactation month n.

4 “CRR n” refers to the cull cow replacement heifer price ratio for month 11.

the marginal effects refer to a speciﬁc situation, the differences in calving season effects

may be caused by the climate differences between Vermont and Indiana or the

differences in facility technology between Vermont and Indiana farms.

65

 

A cow 5
in lactation l (p

be culled than f

 

also exhibited 8
exhibited a 23 f
counterparts. C.
of being culled
of a cow being
were larger tha
by how the bas

The dii
SiSnificantly a
that Produced
was 0.143 per
pmdUCtion ac
Small (0001 ‘.
Pound higher
lIH'alue (QC
Producing Cc
direCtion oil

such, NOIThe

A cow surviving to lactation 2 was 13.975 percent more likely to be culled than
in lactation 1 (p-value < 0.0001). Cattle in lactation 3 were 20.050 percent more likely to
be culled than ﬁrst lactation cattle (p-value < 0.0001). Cattle in lactations 4 through 10
also exhibited a signiﬁcant increased likelihood of being culled. Cattle in lactation 4
exhibited a 23 .624 percent higher likelihood of being culled than their ﬁrst lactation
counterparts. Cattle in their tenth lactation experienced a 43.323 percent higher likelihood
of being culled that ﬁrst lactation cattle. Cattle age signiﬁcantly increased the likelihood
of a cow being culled as expected. The Northeastern cow age parameter marginal effects
were larger than those of the Midwestern estimation. The discrepancy may be explained
by how the baseline situations were described using speciﬁc farms in speciﬁc states.

The difference in milk production between a Northeastern cow and its herd mate
signiﬁcantly affected the probability of a cow being culled (p-value < 0.0001). A cow
that produced one additional hundredweight more than the average 305 ME Milk yield
was 0.148 percent less likely to be culled than the average producing cow. Higher fat
production actually increased the likelihood of a cow being culled but the effect was very
small (0.001 percent). Cattle whose average difference in 305 ME Protein were one
pound higher than the average producing cow were 0.021 percent less likely to be culled
(p-value <0.0001). Cattle that were one persistency unit higher than the average
producing cow were 0.240 percent less likely to be culled (p-value < 0.0001). The
direction of these marginal effects concur with those of the Midwestern estimation. As
such, Northeastern managers may also need educational programs that are designed to

inform them of the economic importance of milk fat production and the importance of

66

looking at the 6
end of the lacta

Like tht’
were more likel
cell count score
to be culled that
parameter, the it
significantly inc
time than the ave
likely to be culle
reproductive fail
count scores and

Northeast
promote milk pro
cow. Cows that h
be culled (p-value
were less likely to
Midwestern estim
more likely to be c
ESSOCiated With hig

Culling

i\Oftheaste:

Holstein cattle (p-\

 

looking at the entire lactation as opposed to how persistent the cow’s production is at the
end of the lactation.

Like the Midwestern estimation, cattle that had higher somatic cell count scores
were more likely to be culled. Northeastern cattle that exhibited a one unit higher somatic
cell count score than the average somatic cell count score were 1.590 percent more likely
to be culled than the cow represented by the baseline situation. The other herd health
parameter, the number of services per conception in the previous lactation, also
signiﬁcantly increased the likelihood of a cull. Cattle that had to be serviced one more
time than the average of its herd mates in the previous lactation were 0.681 percent more
likely to be culled in the current lactation (p-value < 0.0001). To reduce culling for
reproductive failure, producers should consider proﬁtable strategies to lower somatic cell
count scores and reduce the incidence of reproductive failure.

Northeastern cattle that exhibited a one unit higher ability to transmit traits that
promote milk production were 0.022 percent more likely to be culled than the average
cow. Cows that had a PTA Fat that was one unit higher were 1.743 percent less likely to
be culled (p-value = 0.0124). Cattle that had a higher ability to transmit milk protein traits
were less likely to be culled but the results were insigniﬁcant. As was the case with the
Midwestern estimation, cattle that are bred to transmit more milk production traits are
more likely to be culled. Producers may want to determine if the ﬁnancial rewards
associated with high Cow PTA Milk values are worth the ﬁnancial costs associated with
culling.

Northeastern Guernsey cattle were 13.053 percent more likely to be culled than

Holstein cattle (p-value < 0.0001). Jersey and “Other” dairy cattle breeds were 4.193 and

67

6.575143“em ]
Once again: b"
the longevity U
determine lip“
L‘nlike I

be culled. NONI
Northeastern C31
difference betwf
this research. ho
exist in order to
Also unli

likely to be culle
probability of a c
may indicate that
individualized co
Cattle in t
percent but less tr
“it‘ll to be cullec‘

eto ' ‘
,ansion were 7

 

effects of small ex
cipansion manage

lower the likelihoc

 

6.575 percent less likely to be culled than Holsteins respectively (p-value < 0.0001).
Once again, breeders of Guernsey and Holstein cattle may want to compare and contrast
the longevity traits associated with Jersey and the “other” dairy cattle breeds in order to
determine if proﬁtable improvements can be made in the breeds to increase longevity.

Unlike Midwestern cattle, cattle from registered dairy farms were more likely to
be culled. Northeastern dairy cattle were 3.149 percent more likely to be culled than
Northeastern cattle on commercial dairies (p-value = 0.0069). Determining why this
difference between registered dairy farms in the two regions exist is beyond the scope of
this research; however, it would be an interesting to determine why these differences
exist in order to see if improvements are warranted in either region.

Also unlike Midwestern farms, cattle from larger Northeastern farms were more
likely to be culled. For every cow over the sample average herd size of 200 cows, the
probability of a cow being culled increased by 0.025 percent (p-value < 0.0001). This
may indicate that the resource endowments of Northeastern farms favor offering more
individualized cow care than on the more diversiﬁed Midwestern dairy farms.

Cattle in the ﬁrst year of a small expansion (a herd size increase greater than 20
percent but less than 300 percent) in the Northeastern region were 12 .776 percent more
likely to be culled (p-value = 0.0420). Cattle in Northeastern herds one year after a small
expansion were 7.668 percent less likely to be culled (p-value = 0.0806). The marginal
effects of small expansion years 3 — 5 were insigniﬁcant. Extension programs concerning
expansion management may be needed to help Northeastern dairy farm managers to

lower the likelihood of a cull during the ﬁrst expansion year.

68

The It
for rare6 “Pa
likelihood 0‘"
Northeast Wh‘
herd longer th‘

The etl

were mixed. T
0) was insignif
(Heifer Ratio I
Northeastern cc
Ratio Year -2 c
the Heifer Ratic
most recent lagg
Nevertheless, tb
culling rates to 1
There IM
From an overall
estimation as mc

correlated with t'

this indicates a p.
that milk and var;

The ﬁrst It

,._. _
ia’SIEHIDcant A

The marginal effects of the large expansion parameters were insigniﬁcant except
for large expansion year 5. In that year, there was an 11.403 percent increase in the
likelihood of a cow being culled. This may indicate that dairy farm managers in the
Northeast who expand their herds by more than 300 percent try to maintain cattle in the
herd longer than farm managers who did not expand or expanded by a smaller increment.

The effect of the heifer ratio on the likelihood of a Northeastern cow being culled
were mixed. The marginal effect of a one tenth increase in the current (Heifer Ratio Year
0) was insigniﬁcant. Nevertheless, a one tenth increase in the previous year’s heifer ratio
(Heifer Ratio Year -1) caused a 0.142 percent decrease in the likelihood that a
Northeastern cow will be culled in the current year. A one tenth increase in the Heifer
Ratio Year -2 caused a 0.200 percent increase in the culling rate of the current year. Both
the Heifer Ratio Year -3 and -4 parameter estimates were insigniﬁcant. Overall, the two
most recent lagged heifer variables had a signiﬁcant impact on culling likelihood.
Nevertheless, the marginal effects were small. Thus, producers do not totally adjust their
culling rates to totally accommodate higher replacement heifer inventories.

There were mixed effects concerning the milk to feed price ratio parameters.
From an overall perspective, however, the results were similar to those of the Midwestern
estimation as most of the marginal effects of the parameter estimates were negatively
correlated with the likelihood of a cow being culled. Although this was hypothesized,
this indicates a possible need for producer education as past researchers have determined
that milk and variable input prices should not inﬂuence culling decisions.

The ﬁrst lactation month’s cull cow price to replacement dairy heifer price ratio

was signiﬁcant. A one-tenth increase in this ratio caused the likelihood of a cow being

69

 

culled to incre
dairy heifer pr]
with a one-tent
value = 0.0183.
price to replace
one-tenth increa
culled (p-value '
also signiﬁcant 1

cause the proba

  

value < O 0001)
Nonheastem daii
Price responsive

managers. The Si

 

 

 

culled to increase by 8.537 percent (p-value < 0.0001). The cull cow price to replacement
dairy heifer price ratio for lactation months 3 and 4 were also signiﬁcant and positive
with a one-tenth increase causing a respective 8.799 (p-value < 0.0001)and 3.857 (p-
value = 0.0183) percent increase in the probability of a cow being culled. The cull cow
price to replacement heifer price ratio was also signiﬁcant for lactation month 6 with a
one-tenth increase generating 6.94 percent increase in the likelihood that a cow will be
culled (p-value < 0.0001) The cull cow price to replacement dairy heifer price ratio was
also signiﬁcant for lactation months 11 and 12. A one-tenth increase in the ratio would
cause the probability of a cow being culled to increase by 8.293 percent in month 11 (p-
value < 0.0001) and 34.361 percent in lactation month 12 (p-value < 0.0001).
Northeastern dairy farm managers seem to be more cull cow and replacement dairy heifer
price responsive with regard to culling decisions as compared to Midwestern dairy farm
managers. The signiﬁcant cull cow price to replacement heifer price ratios for the
Northeastern dairy cattle coincide with those associated with calving, peak milk
production, initial artiﬁcial insemination service, initial pregnancy examination. the
lactation midpoint, and the lactations end. All of these periods are conducive for a
manager to decide whether to cull a cow, and it makes economic sense that an increase in
the cull cow price to replacement heifer price ratio would increase the likelihood of a cull
during these lactation months. With Midwestern farmers, only the parameter for lactation
month 12 and beyond was signiﬁcant.

Unlike the Midwest, dairy cattle were more likely to be culled from 1996 through
1999. In 1996, a dairy cow was 55.421 percent more likely to be culled than in 1995. In

1997, Northeastern dairy cattle were 51.996 percent more likely to be culled than in

70

1995. The lil
increased CU]
IV. 5”"
A pro
characteristic
This mode1 “
Northeast at 3
Both It
likely to be cu
that calved in
Spring In the '
calving cattle.
heifer calvin gs
calving animal.
schedule their l
associated with
Both Mi
PTObability as [h
lactation differet
margitial effects
Speciﬁc states,

leferenc‘

with culling likeli?

1995. The likelihood increased by 68.767 percent in 1998 prior to decreasing to a 38.227
increased culling likelihood in 1999.
IV. Summary and Conclusions

A probit model was developed to determine how individual cow and herd level
characteristics affect the likelihood of a cull of Midwestern and Northeastern dairy cattle.
This model was ableto predict culling events and nonevents in the Midwest and
Northeast at an 80.4 and 86.7 percent accuracy rate respectively.

Both Midwestern and Northeastern cattle that calved during winter were more
likely to be culled than cattle that calved in Spring. For Midwestern dairy farms, cattle
that calved in Summer and Fall were less likely to be culled than cattle that calved in
Spring. In the Northeast, fall calving cattle were more likely to be culled than spring
calving cattle. Midwestern dairy farm managers may want to consider scheduling their
heifer calvings to accommodate the increased likelihood that more Spring and Winter
calving animals will have to be culled. Northeastern dairy farm managers may want to
schedule their heifer calvings to counterbalance the increased culling probability
associated with calving in Fall and Winter.

Both Midwestern and Northeastern cattle experienced an increased culling
probability as they aged. Nevertheless, the increase in culling probability for each
lactation differed in the two regions. These results may be influenced by how the
marginal effects were calculated using a baseline situation referencing speciﬁc farms in
speciﬁc states.

Differences in ﬂuid milk and milk protein production were negatively correlated

with culling likelihood in both the Midwestern and Northeastern estimations. The

71

difference in r
in both region
cow’s fat yielt
culling probat
indicate that n
term ﬁnancial
more proﬁtab
Differ
conception w
farmers may .
Synchronizati
Variance be“:
In the
with Culling
PIOIein. Wm
Protein prod

between let

difference in milk fat production were insigniﬁcant in each region. Dairy farm managers
in both regions may be erroneously ignoring the economic returns associated with a'
cow’s fat yield when making culling decisions. Cow persistency had a greater effect on
culling probability than actual fluid or milk component yield in both regions. This may
indicate that managers may improperly make culling decisions based on immediate short
term ﬁnancial concerns instead of future expected proﬁts. If this is true, it is possible that
more proﬁtable but less persistent cattle are being culled.

Differences in the somatic cell count score and previous lactation services per
conception were positively correlated with the probability of a cull in both regions. Dairy
farmers may want to incorporate programs, such as better dry cow treatments or estrus
synchronization, to reduce the somatic cell count score and services per conception
variance between cows.

In the Midwest and Northeastern regions, PTA Milk was positively correlated
with culling likelihood. In the Midwest, the likelihood of a cull increased with PTA
Protein. While this seemingly disagrees with the marginal effects of actual milk and milk
protein production, these results are similar to genetic research on the correlations
between type, PTA Milk, PTA Protein, and longevity.

Midwestern and Northeastern Guernsey cattle were more likely to be culled than
Holstein cattle. Midwestern and Northeastern Jersey cattle and the “Other” dairy cattle
breeds were less likely to be culled. If increased longevity is ﬁnancially important for
Holstein and Guernsey breeders, they may want to consider comparing and contrasting
the traits of Holstein and Guernsey cattle with those of the Jersey and “Other” dairy

cattle breeds.

72

There‘
registered he“
herds were les
likely to be all

The ell
correlated with
Northeast. This
specialized labt
the ability to of
expansion 63““
however, that m
longer

The aval
milking herd a }‘
cull This indicat
all of the availah
surplus heifers tc
prior to entering

The milk
Culling likelihooc

r
escarch clearlv ll

 

”195 As the milk

Price ratio should

 

There were mixed results concerning the marginal effect of a cow being ﬁ'om a
registered herd on the likelihood of it being culled. Cattle from Midwestern registered
herds were less likely to be culled. Cattle from Northeastern registered herds were more
likely to be culled.

The effect of herd size on culling probability were mixed. Size was negatively
correlated with the likelihood of a cull in the Midwest but positively correlated in the
Northeast. This may indicate that larger Midwestern farms may be able to procure more
specialized labor, whereas the resource availability in the Northeast allows smaller farms
the ability to offer better individualized care. One could not conclude conclusively that
expansion caused increased culling rates except in select expansion years. It did appear,
however, that managers who engaged in a large expansion in the Northeast retained cattle
longer

The availability of replacement heifers relative to the number of cattle in the
milking herd a year prior to the observation year slightly increased the probability of a
cull. This indicates that producers do not totally adjust their culling rate to accommodate
all of the available replacement heifers. Instead, producers must either sell most of their
surplus heifers to other farmers, or the surplus animals are culled due to health reasons
prior to entering the milking herd.

The milk feed price ratio parameters were mostly negatively correlated with
culling likelihood. Educational programming on culling cattle may be warranted as prior
research clearly indicates that milk and variable input prices should not effect culling
rates. As the milk price and variable production costs are equal for both, the milk feed

price ratio should have little effect on culling rates.

73

Mid
cull cows re
parameter 1
price to repl
peak milk pi
months to cc

In the
Midwestern 1
However, in

Opposed to l‘.

Midwestern dairy producers were not price responsive with regard to the price of
cull cows relative to the price of replacement heifers except in the last lactation month
parameter. Northeastern dairy managers were more price responsive. Signiﬁcant cull cow
price to replacement dairy heifer price ratio parameters coincided the calving month,
peak milk production, critical breeding dates and dry off, all of which are reasonable
months to contemplate a culling decision.

In the Midwest, the effect of observation year and culling likelihood were mixed.
Midwestern cattle in 1996, 1998 and 1999 were less likely to be culled than in 1995.
However, in the Northeast, cattle were more likely to be culled in 1996 through 1999 as

opposed to 1995.

74

 

l.

 

rates is 10 3de
Problems I“ or
to the proporth

developed and ‘

 

esll'US synChTOlll
veterinarians - ll
reproduction hee
IL Data
Data for '
Monitoring Serv
farmers in major
performance, tec
incidence of herd
asrratified randor

representative oft

COWS.

CHAPTER 4.

MODELS TO DETERMINE THE EFFECTS OF SELECT MANAGEMENT
FACTORS ON DAIRY CATTLE HEALTH CULLING RATES

I. Introduction

As most cattle are culled due to health problems, one method to reduce culling
rates is to adopt management strategies to prevent or treat the underlying health
problems. In order to do this, one must understand how management programs contribute
to the proportion of cattle culled for health. Four ordinary least squares models were
developed and estimated to determine the effect that management programs - such as
estrus synchronization, BST, soft textured walking surfaces, and employing on-staff
veterinarians — have on udder and mastitis, lameness and injuries, disease, and
reproduction health culls.
II. Data

Data for this study were collected from the USDA National Animal Health
Monitoring Service’s (N AHMS) Dairy ‘96 Survey. This survey involved interviewing
farmers in major dairy producing states about farm descriptive statistics, farm
performance, technology adoption, management programs implemented, and the
incidence of herd health problems as of January 1, 1996. The survey was conducted using
a stratiﬁed random sample design. The stratiﬁed random sample was designed to be
representative of the entire US. dairy farm industry except those milking less than 30
cows.

One important variable for this research, the number of days a dairy employed an

on-staff veterinarian, had only 57 entries. In order to avoid having a very limited sample

75

size, it WE
veterinari;
III. M
F o:
manageme
mastitis cu
reproductic
The Udder
The
the udder an
and Heillber
was between
With subclini
(1993) fOUnd
PrOgramS and
Mastitis and S
with OIhEr l1 Er
Efficiency PTO'

increased Sew

herd heahh prc

trade and anim,

health issue fac

size, it was assumed that a missing entry meant that the farm did not employ an on-staff
veterinarian and a value of zero days was assigned to those observations.
III. . Methods and Model Development

Four ordinary least squares models were developed to determine the effect that
management factors had on culling rates due to speciﬁc health issues: the udder and
mastitis culling rate, the lameness and injury culling rate, the disease culling rate, and the
reproduction culling rate. These models are describe in the following subsections.
The Udder and Mastitis Culling Rate Model

The ﬁrst model determines the effect of select management controlled factors on
the udder and mastitis culling rate. Udder and mastitis problems are costly. Wells, Ott,
and Heillberg Seitzinger (1998) estimated that the aggregate producer cost of mastitis
was between 1.5 and 2 billion dollars per year. The aggregate producer costs associated
with subclinical mastitis were estimated to be 960 million dollars per year. Miller et a1.
(1993) found that Ohio producers spent $14.50 per cow per year for mastitis prevention
programs and $37.91 per cow per year for costs associated with clinical mastitis cases.
Mastitis and subclinical mastitis may also have other indirect costs through its association
with other herd health problems. Scrick et al. (2001) linked mastitis to reproductive
efﬁciency problems such as increased days to ﬁrst service, increased days open, and
increased services per conception. When ranking
herd health problems on the basis of production loss, zoonotic potential, international
trade and animal welfare issues — mastitis was determined to be the most important herd

health issue facing the US. dairy industry (Wells, Ott, Hillberg Seitzinger, 1998).

76

The 5
the udder an
cow manage
use, and bios
descriptions

Table 21.

Independent
.1 Handbook

~ Incentives
Handbook an
Incentives
; Dry Treatmer

l Pie-Di
. Post Di
Pre- and Post

Free Stall Fac

1 Multiple Ani n
Facility

: Held SiZe
Sand Bedding
. W00d BCddlng

0m 03‘ Bedd
,Rubbﬂ Mat B,

Oiii Bedding
8TB d -
B T e ding

9.2
5553?"?
”75”
4.
5.

D . erlnari;

Milk QUaliry T

/.

e.‘

/

The general management factor categories included as independent variables in
the udder and mastitis culling rate model included the use of an employee handbook, dry
cow management, milking procedures, housing type, herd size, BST use, veterinarian
use, and biosecurity factors. The udder and mastitis culling rate independent variable
descriptions are displayed in Table 21.

Table 21. The Udder and Mastitis Culling Rate Model Independent Variable

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Descriptions

Independent Variable Description

Handbook Indicates if a farm uses an employee handbook

Incentives Indicates if a farm uses employee incentive programs

Handbook and Indicates if a farm uses both an employee handbook and

Incentives incentive programs

Dry Treatment Indicates if a farm dry treats all four udder quarters on
almost all cows at the end of a lactation

Wash Pen Indicates if a farm has apre- milk wash pen

Pre-Dip Indicates if a farm pre-dips all teat ends prior to milking

Post Dip Indicates if a farm post dips all teat ends prior to milking

Pre- and Post Dip Indicates if a farm either has a pre-wash pen or pre—dips all
teat ends prior to milking and post dips all teat ends aﬁer
milking

Free Stall Facility Indicates if a farm uses free stall facilities

Multiple Animal Indicates if a farm uses multiple animal housing

Facility

Herd Size The size of the milking and dry cow herd

Sand Bedding Indicates if a farm uses sand bedding

Wood Bedding Indicates if a farm uses wood-based bedding

Compost Bedding Indicates if a farm uses composted manure bedding

Rubber Mat Bedding Indicates if a farm uses rubber mat bedding

Tire Bedding Indicates if a farm uses tire bedding

Newspaper Bedding Indicates if a farm uses newspaper bedding

Mattress Bedding Indicates if a farm uses mattress bedding

Stalk Bedding Indicates if a farm uses corn stalk bedding

Other Bedding Indicates if a farm uses other bedding types

BST Indicates if a farm uses BST

Veteriniry Visits Refers to the number of veterinamisits needed

Staff Veterinarian Refers to the number of days a veterinarian was on staff

Day

Milk Quality Test Indicates if a farm requires a somatic cell count score or
milk culture test prior to purchasing a cow

 

77

A
two Ways
determine
these tasks
methods C
handb00k f
theifIOb ras

prescribed I
parlor equip
monitor the 1
problems art!
and mastitis r
Perfor
employees to
employees wil
incentive prog
culling rate.

Having

A properly prepared employee handbook enhances farm performance in at least
two ways. First, having to write an employee handbook encourages a manager to
determine the best practices for each task on the farm. This may involve analyzing how
these tasks are being performed currently and determining if there are better alternative
methods. Once the best methods have been chosen for each task and composed into
handbook form, the handbook can then serve as a reference to show employees how to do
their job tasks properly and to inform them why it is important to do these tasks in the
prescribed manner. A handbook informs employees how to use, clean and maintain the
parlor equipment, how to prepare each cow for milking, how to check for mastitis, how to
monitor the milking, how to care for each cow after milking, and what to do when
problems arise. It is expected that farms with employee handbooks will have less udder
and mastitis related problems than those that do not.

Performance based employee incentive programs are used to encourage
employees to improve their performance. Employee incentive programs help insure that
employees will strictly follow milking protocols. The marginal effect of using employee
incentive programs is expected to be negatively correlated with the udder and mastitis
culling rate.

Having both a handbook and an incentive program can further enhance employee
performance beyond what each would alone. A handbook describes to employees how to
conduct activities to enhance udder health, and employee incentive programs help insure
that these protocols are followed. Farms with both an employee handbook and an
employee incentive program are expected to have lower udder and mastitis culling rates

than those without.

78

At a
treat dairy C
than when t
the udder w
cattle is exp
It is 1
reduce masti
pens. water i
ends with a 5
These treatm
(Hoards Dair
pOSIdipping I
either wash p
also expected
Cattle
tmes thousj
(or Stanchion)
for EX€rcise ) .
COWS in a lie 5‘

cattle are free i

 

At a lactation’s completion, a standard protocol on many dairy farms is to dry
treat dairy cattle because the infection rate for mastitis is higher during the dry period
than when the cow is lactating (Hoards Dairyman, 1990). Dry treating involves infusing
the udder with antibiotics within one week of a cow’s last milking. A farm that dry treats
cattle is expected to have a lower mastitis incidence than those that do not.

It is important to clean and sanitize the udder and teat ends of the cattle in order to
reduce mastitis. One method is to have the cattle enter a wash pen prior to milking. In the
pens, water is sprayed upward to clean the udder. Dipping or spraying the cow’s teat
ends with a sanitizing liquid is another method of reducing the incidence of mastitis.
These treatments occur before (predipping) and/or after (postdipping) each milking
(Hoards Dairyman, 1990). It is expected that producers who use wash pens, predipping or
postdipping technologies will have less udder and mastitis culls. Producers who utilize
either wash pens or pre-dipping protocols in conjunction with post dipping protocols are
also expected to experience fewer udder and mastitis culls.

Cattle housing may also affect udder and mastitis culls. There are three primary
types of housing facilities: tie stalls, free stalls and multiple animal facilities. Tie stalls
(or stanchion) barns tend to be older facilities whereby the cows are kept in a stall (except
for exercise ) either by a tether or a head chute. Feed and milking units are brought to the
cows in a tie stall facility. Free stall facilities tend to be more modern facilities where
cattle are free to enter and leave any stall and to go to a feed bunk to eat. The cattle are
moved to a milking parlor to be milked. Multiple animal housing can either be loaﬁng
sheds (essentially free stall facilities without stalls) or westem-style dry lots. It is difﬁcult

to tell which facility type will have an advantage as far as udder and mastitis problems

79

are concerne
they may ha'
size associat
may be more
subclinical a:
culls and fac.
animal facilit
these dummy
equation defa
It is a
culls. Althoug
associated wit
herds may be
size. More Sp e
CUlling rates.
The m:
were assessed
WOOd'baSCd, Cr
and Other bedd.
regression Whei
l:i’hile bedding 1

ConelatIOH is d”

are concerned. Free stall facilities are a newer technology than tie stall facilities. As such,
they may have an advantage concerning udder and mastitis culls. Alternatively, the herd
size associated with tie stall facilities tend to be smaller than free stall facilities, and there
may be more individualized attention given to cattle by the milkers, thereby reducing
subclinical and clinical mastitis episodes. As such, the correlation of udder and mastitis
culls and facility type cannot be hypothesized. In the model, free stall and multiple
animal facility types were differentiated by using two dummy variables. When both of
these dummy variables are set to zero, the udder and mastitis culling rate model’s
equation defaulted to represent a farm with tie stall facilities.

It is also difﬁcult to hypothesize how herd size will affect udder and mastitis
culls. Although it was inferred in the previous paragraph that smaller herd size may be
associated with more individualized care that is conducive to better herd health, larger
herds may be able to hire more specialized labor or adopt technology because of their
size. More specialized labor or better technology may also reduce udder and mastitis
culling rates.

The materials used to bed cattle stalls may also affect mastitis. Ten bedding types
were assessed to determine the effect of bedding types on udder and mastitis culls: sand,
wood-based, composted manure, rubber mats, tire, newspaper, mattress, corn stalk, straw
and other bedding. Straw bedding was represented in the baseline situation of the
regression when the other bedding dummy variables were set to zero.

While bedding type is expected to affect the udder and mastitis culling rate, its

correlation is difﬁcult to predict.

80

BST i

hypothesized

 

 

likelihood of
BST had no e
(1996) found

a cow with su

 

longer periods
economic reas
that the prOpo
BST use.
\"eteri
The number i
culls. Come.
POSitiVely CC
0n the Surve
and diagnOS
the eﬁ‘eCI t}
Never-I h Cleg
California .
produCerg

than dISea,

BST increases milk yield in the latter half of a cow’s lactation. It could be
hypothesized that this additional milk yield places undue stress on the cow, increasing the
likelihood of udder and mastitis problems. Nevertheless, Collier et al. (2001) found that
BST had no effect on mastitis. Baumann et al. (1999) and Ruegg, F abellar, and Hintz
(1996) found that BST did not signiﬁcantly affect overall culling rates. BST may enable
a cow with subclinical mastitis to increase its production enough to remain in the herd for
longer periods of time, giving the manager more time to treat the animal and less
economic reasons for culling the cow for udder and mastitis problems. It is hypothesized
that the proportion of health culls caused by udder and mastitis problems decreases with
BST use.

Veterinary services can be both preventive or diagnostic and treatment in nature.
The number of preventive visits should be negatively correlated with udder and mastitis
culls. Conversely, the number of treatment and diagnostic visits are expected to be
positively correlated with udder and mastitis culls. While the number of veterinary visits
on the survey farms was determined, the number of preventive veterinary service calls
and diagnostic and/or treatment veterinary service calls were not, which makes predicting
the effect that veterinary service calls have on udder and mastitis culls difﬁcult.
Nevertheless, Weiglar et al. (1990) found that disease events accounted for 89 percent of
California dairy veterinary farm expense from 1988 to 1989. This may indicate that
producers primarily use non-staff veterinarians for diagnostic and treatment work rather
than disease prevention programming. Udder and mastitis culls are expected to be

positively correlated with veterinarian service calls.

81

Vt
than a "on
activities 1
number of

Mas
follOw a ma
such a prOgr
Nmsts Dal
somatic Ce” c
cow. P r Oducc

and mastitis C1
The Lamenes.
Cook (:
States. Warnic
cattle was 705 l
Toral lameness
California herds
that lameness cc

Lameness and i.

(2003) reported 1

study of a single

laminitis lesions

 

were culled (Cat‘

 

Veterinarians who are employed by farms should spend more time on that farm
than a non-staff veterinarian. The additional time can be used for disease prevention
activities. It is expected that udder and mastitis culls are negatively correlated with the
number of days a farm employs an on-staff veterinarian.

Mastitis problems can be contagious. Pritchard (2000) suggests that dairy farmers
follow a mastitis biosecurity program when purchasing new cows. One component of
such a program is to receive veriﬁcation of the potential new cow’s mastitic state. The
NAHMS Dairy 1996 survey asked producers if they insisted upon having individual cow
somatic cell count tests and/or individual cow milk culture sample tests prior to buying a
cow. Producers who practice this procurement policy are expected to have lower udder
and mastitis culls.

The Lameness and Injury Culling Rate Model

Cook (2003) estimated that over 23 percent of dairy cows are lame in the United
States. Warnick et al. (1995) found that the 305 day mature equivalent milk yield of lame
cattle was 705 pounds less than non-lame herd mates. Lameness cost estimates vary.
Total lameness treatment and prevention costs averaged $7.83 per cow per year for
California herds in 1988 — 1989 (Weigler et al., 1990). New York researchers determined
that lameness costs New York producers $900 per cow per year (Wallace, 2003).
Lameness and injuries may also have spillover effects concerning reproduction. Sutton
(2003) reported that lame cattle remain unbred for 28 days longer than healthy cattle. In a
study of a single lactation for 5,000 New York cows, forty-four percent of the cattle with
laminitis lesions were culled while only twenty-ﬁve percent of the cattle without laminitis

were culled (Cattell, 2001).

82

and injL
facility ‘
BST use
hairy hee
model ca

Table 22.

l lndepend
' Hairv Hee.
Footbath

l

\
Footbath an
l min
Free Stall
‘ Multiple Am
.1 Facilitv
’ Exercise Lor
. Soft Surface
‘ Textured C0,,
l Surface
Sand Bedding
. Wood Bedding

 

The second model estimates the effect of management factors on the lameness
and injury culling rate. General management controlled factors included surface and
facility technologies, hoof conditioning practices, veterinary care, milk production and
BST use. The infectious agent factor was the proportion of cattle that showed signs of
hairy heel wart. The independent variable descriptions of the injury and culling rate
model can be seen in Table 22.

Table 22. The Lameness and Injury Culling Rate Model Independent
Variable Descriptions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Independent Variable Description
Hairy Heel Wart The proportion of cattle with hairy heal warts
F ootbath Indicates if a farm uses a footbath regularly throughout the
year
Hoof Trim Indicates if a farm trims the hooves of nearly all the cattle
during a year
F ootbath and Hoof Indicates if a farm regularly uses a footbath and trims the
Trim hooves of nearly all the cattle during a year
Free Stall Indicates if a farm houses its cows in a free stall facility
Multiple Animal Indicates if a farm houses its cows in multiple animal
Facility housing
Exercise Lot Indicates if a farm has an exercise lot
Soft Surface Indicates if a farm’s cattle primarily stand on a soft surface
Textured Concrete Indicates if a farm has textured concrete cattle walking
Surface surfaces
Sand Bedding Indicates if a farm uses sand bedding
Wood Bedding Indicates if a farm uses wood-based bedding
Compost Bedding Indicates if a farm uses composted manure bedding
Rubber Mat Bedding Indicates if a farm uses rubber mat bedding
Tire Bedding Indicates if a farm uses tire bedding
Newspaper BeddingL Indicates if a farm uses newspaper bedding
Mattress Bedding Indicates if a farm uses mattress bedding
Stalk Bedding Indicates if a farm uses corn stalk bedding
Other Bedding Indicates if a farm uses other bedding types
Veterinary Visit The number of veterinary visits needed on a farm
Staff Veterinarian Day The number of days a veterinarian was on staff
Rolling Herd Average The rolling herd average of the herd
BST Indicates if a farm uses BST

 

83

 

 

IIai
on the heel
Hairy heel t
wnkcanei
proportion c
lameness an.

lioof
promoting be
recommendet
condition, d€(
treat hairy hef
(Wallace, 2001
injury culling,
COITect hoof srr
mmming can it
oflaminitis, a ,
regular hoof rm
reglllarl). USE f0
lameness and in

implement One (

AS linferr

 

b

 

Hairy heel wart, digital dermatitis, is an infectious condition where a lesion forms
on the heel of cattle. As these lesions grow, they can cause the animal to become lame.
Hairy heel warts induce lameness on 40 percent of the Midwest dairy farms, and infected
cattle can experience a decrease in milk yield of 20 to 50 percent (Wallace, 2003). As the
proportion of infected cows increases in a herd, the proportion of cattle culled due to
lameness and injury is expected to increase.

Hoof care is thought to help reduce the incidence of lameness and injuries by
promoting better hoof conditioning. Medicated footbaths and/or hoof trimming are
recommended protocols on many farms. F ootbaths are thought to improve hoof
condition, decrease the incidence of hoof infections, and are used by many to prevent and
treat hairy heel wart. Footbaths need to be used more than once a month to be effective
(Wallace, 2000). Managers who use footbaths regularly should see less lameness and
injury culling. Farms that trim hooves regularly are expected to have more animals with
correct hoof structure, thereby reducing lameness and injuries. Additionally, hoof
trimming can lead to the early detection of poor hoof health and identify ex-post periods
of laminitis, a nutrition problem which has hoof health repercussions. It is expected that
regular hoof trimming is negatively correlated with lameness and injury culls. Farms that
regularly use footbaths and regularly trim hooves are expected to experience lower
lameness and injury culling rates than ﬁrms who do not or those farms that only
implement one of the preventive measures.

As inferred earlier in this section, facility type may also contribute to lameness.
Cook (2002) found that there was no signiﬁcant difference between tie stall and free stall

barns concerning lameness episodes. As such, it is unlikely that free stall technology will

84

less lamenﬁ‘
Cattle from '

culling rates

 

Cattle

  

stall. especia

able to stand
expected to bt
Cattle
textured surfa
less hoof wea
likely to have
Opposed to Ct
about the use
primary surf:
Pack are sgﬁ
expeCIEG to
Smo
SOHditions i

and lame. F

have a signiﬁcant advantage over tie stall facilities. Multiple animal facilities may have
less lameness and injury problems, however, as the cattle typically stand on a dirt pack.
Cattle from multiple animal facilities are expected to have lower lameness and injury
culling rates than cattle from tie stall facilities.

Cattle housed in tie stall facilities typically spend much of the day tethered to a
stall, especially in winter. Even though a bedding is provided, the surface below the
bedding is hard, usually concrete. Cattle that are housed in free stalls, even though they
can get up and move around their barn, predominantly stand on concrete. Having an
exercise lot (either dry lot and/or pasture) allows a cow to get more exercise and to be
able to stand on a softer surface than concrete. Cattle that have access to exercise lots are
expected to be less prone to lameness and injuries.

Cattle that predominantly stand on softer (as opposed to hard concrete surfaces)
textured surfaces, such as pasture and dirt, should have less pressure on their hooves and
less hoof wear. Bray, Giesey and Bucklin (2002) found that cattle were 43 percent less
likely to have a foot health problem when a rubberized walking surface was used as
opposed to concrete surfaces. While the Nahms Dairy ’96 survey did not ask producers
about the use of rubberized surfaces, the producers were asked to distinguish if their
primary surface for the cattle was concrete or pasture or dirt pack. As pasture and dirt
pack are softer than concrete, cattle that predominantly stand on those surfaces are
expected to be less likely to be culled for lameness and injuries.

Smooth concrete surfaces can be slippery. It is hard to avoid wet and slippery
conditions in dairy housing facilities. Cattle that lose their footing may become injured

and lame. Roenfeldt (2001) reported that Kansas operations that used slip-form concrete

85

surfaces in l
textufed OOH
surfaCes and
The ’

rates. Ten be
and maStitis ‘
mattreSSa CO”
baseline Sirua
zero. AS each
various bedd“
Assun‘:

number 0f "6”
this“? health C
hairy heel wart
can devote m0?
was on staff is
In an ell

are t00 rich in 8
become lame (l1
herd average is l
Collier e?
lameness incider

wa ,
sh} pothesize-'

 

surfaces in their new facilities had lower than expected lameness episodes. Having a
textured concrete surface is expected to reduce the problem of cattle slipping on dairy
surfaces and reduce the number of cattle culled due to lameness and injury.

The materials used to bed cattle stalls may also affect lameness and injury culling
rates. Ten bedding types were assessed to determine the effect of bedding types on udder
and mastitis culls: sand, wood-based, composted manure, rubber mats, tire, newspaper,
mattress, corn stalk, straw and other bedding. Straw bedding was represented in the
baseline situation of the regression when the other bedding dummy variables were set to
zero. As each bedding type has its advantages and disadvantages, the correlation of the
various bedding types could not be hypothesized.

Assuming that veterinarians are called to primarily diagnose and treat cattle, the
number of veterinarian visits is expected to be positively correlated with lameness and
injury health culls. On staff veterinarians, however, may be better able to implement
hairy heel wart, laminitis, and poor hoof condition prevention programs in place as they
can devote more time to prevention protocols. Thus, the number of days a veterinarian
was on staff is expected to be negatively correlated with lameness and injury culls.

In an effort to obtain higher production levels, dairy cattle may be fed diets that
are too rich in grain and protein. In this situation, a cow may develop laminitis and
become lame (Hoard’s Dairyman, 1990). It was hypothesized that the effect of rolling
herd average is positively correlated with lameness and injury culling.

Collier et al. (2001) determined that although BST was correlated with increased
lameness incidents, it did not affect the total number of animals culled for lameness. As

was hypothesized with udder and mastitis culling rates, BST may increase lame cattle

86

milk produc
replace then

injury cullin

The Disease

Wieg

year to preve

Disease, cost

Hillberg Seitz
COW per year 2
to be culled pr
In this

rate. The maria
PTOduction, B l
Disease or para
described in Ta
AS a far

may be f0rced 1
animals may Ca;
managed, Dairy

more diseaSe Cu'

 

 

milk production enough to make it more proﬁtable to treat the animals rather than to
replace them. BST is expected to reduce the culling rates associated with lameness and

injury culling.

The Disease Culling Rate Model

Wiegler et al. (1990) estimated that California producers spent $28.20 per cow per
year to prevent and treat diseases from 1988 and 1989. One disease in particular, Johne’s
Disease, cost US. producers an estimated 222 million dollars in 1995 (Wells, Ott,
Hillberg Seitzinger, 1998). Herds infected with Johne’s Disease can have up to $200 per
cow per year higher veterinary costs (Wells, 2000). If a cow that contracts a disease has
to be culled prematurely, the cost of disease to producers increases.

In this section, we estimate how management factors affect the disease culling
rate. The management factors include biosecurity related factors, facility technologies,
production, BST use, and veterinary use. The infectious agent examined was Johne’s
Disease or paratuberculosis. The disease culling rate model’s independent variables are
described in Table 23.

As a farm’s dependence on purchased dairy replacements increases, dairy farmers
may be forced to procure animals from multiple sources. The commingling of these
animals may cause a major biosecurity breach, a disease outbreak, if not properly
managed. Dairy farms with a high proportion of purchased cattle are expected to exhibit
more disease culls.

There are various methods of reducing the threat of disease. Youngstock can be

kept separate from cattle and other species to limit the spread of disease by nose-to-nose

87

Table 23.

lndepenC
l Var iabl
__________._
. Purchase C
“‘““"""’?
l Nose—to-N
? Cow Conta
Nose-to-Nc
3 Other Cont
F—Té—f
I V'accrnes I;
l

l Vaccines

Required
Tests Requi.

Quarantine
' Vaccines, Tt
and Quarant:

i

l

 

Vaccrnes and
Tests

\accmes and
. Quarantine

; Tests and
Quarantine

thne’ 5
- POSIIIVe
. VEIerinary
- Visits
l Staff

‘ V elerinari an I
aV' |

W

aClllty

 

 

 

Table 23. The Disease Culling Rate Model Independent Variable
Descriptions
Independent Variable Description
Variable

 

Purchase Cows

The proportion of the herd that was purchased

 

Nose-to-Nose
Cow Contact

Indicates if nose- --to -nose contact between heifer calves and other
cattle lS prevented

 

Nose-to-Nose
Other Contact

Indicates if nose-to-nose contact between heifer calves and other
species is prevented

 

Vaccines Used

The proportion of the vaccinations listed in the NAHMS Dairy ’96
Survey used by a farm

 

Vaccines Indicates if a farm requires that purchase cattle have sixty percent
Remired or more of the vaccines listed in the NAHMS Dairy ’96 Survey

 

Tests Required

Indicates if a farm requires that purchase cattle be tested for sixty
percent or more of the diseases listed in the NAHMS Dairy ’96
Survey

 

Quarantine

Indicates if a farm has quarantine protocols

 

Vaccines, Tests
and Quarantine

Indicates if a farm requires purchased animals to be vaccinated by
sixty percent of the vaccines listed in the NAHMS Dairy ’96
Survey and requires that purchase animals be tested for sixty
percent of the diseases listed in the NAHMS Dairy ’96 Survey and
has quarantine protocols

 

Vaccines and
Tests

Indicates if a farm requires purchased animals to be vaccinated by
sixty percent of the vaccines listed in the NAHMS Dairy ’96
Survey and requires that purchase animals be tested for sixty
percent of the diseases listed in the NAHMS Dairy ’96 Survey

 

Vaccines and

Indicates if a farm requires purchased animals to be vaccinated by

 

 

 

 

 

 

 

Quarantine sixty percent of the vaccines listed in the NAHMS Dairy ’96
Survey and has quarantine protocols

Tests and Indicates if a farm requires that purchase animals be tested for sixty

Quarantine percent of the diseases listed in the NAHMS Dairy ’96 Survey and
has quarantine protocols

Johne’s Indicates if a farm has cattle that have tested positive for Johne’s

Positive Disease

Veterinary The number of veterinary visits needed on a farm

Visits

Staff The number of days a veterinarian was on—staff

Veterinarian

Day

Herd Size The size of the milking and dry cow herd

Free Stall Indicates if a farm houses its milking herd in a free stall facility

Facility

 

 

88

 

Table 23 (C
' Animal
l Facilit .
i Rolling Her
l Average
. BST
contact. A cl
from other f2
Farm manag'
diseases prio
are expected
implement th
higher beneﬁ:
The cl
loss, and ever.
Youngstock bi,
PTOblem is COr
disease to othe
Positive for Jot
disease Problet
Like 1h,
veterinary Visit
veterinarian Dre

33 a farm empkl

 

 

Table 23 (cont’d)

 

 

 

 

 

Multiple Indicates if a farm houses its milking herd in multiple animal
Animal housing

Facility

Rolling Herd The rolling herd average of the herd

Average

BST Indicates if a farm uses BST

 

 

contact. A comprehensive vaccination program can be implemented. Animals purchased
from other farms can be quarantined and observed for disease symptoms.

Farm managers can insist that potential replacement cattle be vaccinated and tested for
diseases prior to purchasing them. Farms that engage in these individual practices

are expected to have less disease culls than those that do not. Of course, those who
implement these programs in conjunction with the other programs may see an even
higher beneﬁt.

The clinical signs of Johne’s Disease include watery diarrhea, milk loss, weight
loss, and even death. It is highly contagious. Cattle can contract the disease as
youngstock but not show clinical symptoms until between two and ﬁve years of age. The
problem is compounded by the fact a subclinically Johne’s infected heifer can shed the
disease to other youngstock (Stabel, 1998). As such, the presence of an animal that tests
positive for Johne’s disease increases the likelihood that the farm will have to cull for
disease problems.

Like the previously discussed health cull types, the correlation of the number of
veterinary visits and the disease culling rates should be positive as producers use the
veterinarian predominantly for diagnostic and treatment purposes. Having a veterinarian

as a farm employee, however, increases the likelihood that the farm veterinarian will be

89

able to dev
days workq
disease cull
Catt
outbreaks. .‘
flow is avail
and multiple
occur in freel
The e
expected to p
“pushed" to ;
While it was
PI0p0rti0n ot‘
disease culls g
BST is

BST was nOt a
lives, It “’8 S it
make it ”me e
With disease pl
keep Sick dair}

not help dlSCaSt

 

able to develop and implement disease reduction protocols on the farm. The number of
days worked by an on-staff veterinarian is expected to be negatively correlated with the
disease culling rate.

Cattle grouped together in any facility type offer an opportunity for disease
outbreaks. Nevertheless, some respiratory diseases may be minimized if adequate air
ﬂow is available. In general, tie stalls have poorer ventilation than most modern free stall
and multiple animal facilities. As such, it is expected that less culling due to disease will
occur in free stall and multiple animal facilities.

The effect of rolling herd average is difﬁcult to predict. Healthier herds are
expected to produce more. There is a common belief, however, that cattle that are
“pushed” to produce more will have an increased likelihood of contracting a disease.
While it was hypothesized that milk production levels would signiﬁcantly affect the
proportion of health culls culled for disease, the correlation of rolling herd average and
disease culls could not be hypothesized.

BST is one method to increase milk production. Collier et a1. (2001) showed that
BST was not associated with increase incidences of disease. With the previous health cull
types, it was hypothesized that BST may help lame and mastitic cows produce enough to
make it more economical for the producer to treat the animals instead of culling them.
With disease problems, however, the risk of spreading the disease may be too great to
keep sick dairy cattle in the herd despite their production level. Accordingly, BST may

not help diseased animals remain in the herd.

9O

The Rep“

C0‘

are manage
reproductic
herd averag

bull use.

Alth

with high pr
beyond 390 .
that calving i
cow. Cassell
mates. A m0,
factors that c;
culls Caused t
handbook, the
CSIms Synchrc
variables are C
An em

and outline th

aniliCia] inSem

breeding and tr

 

 

The Reproduction Culling Rate Model

Cows that have problems conceiving will not stay in a herd for very long. There
are management controlled factors that can increase or decrease the likelihood of
reproduction culls. Such factors include having an employee handbook, the herds rolling
herd average, herd size, facility technology, veterinary use, estrus synchronization and
bull use.

Although Galton (1997) concluded that longer calving intervals may be proﬁtable
with high producing herds using BST, Jones (2001) reported that calving intervals
beyond 390 days were less proﬁtable for high producing cows. Keown (1986) reported
that calving intervals beyond 395 days costs $3.00 per day per extended calving interval
cow. Cassell (2002) noted that less fertile cows get culled sooner than more fertile herd
mates. A model was developed to determine the effectiveness of management controlled
factors that can improve reproduction efﬁciency and decrease the proportion of health
culls caused by reproduction problems. Such factors include having an employee
handbook, the herds rolling herd average, herd size, facility technology, veterinary use,
estrus synchronization and bull use. The reproduction culling rate model independent
variables are describe in Table 24.

An employee handbook should show employees how to detect if a cow is in estrus
and outline the procedures for informing the manager about the cow. For employees with
artiﬁcial insemination duties, the handbook should also explain the protocols to use when

breeding and testing for pregnancy. The existence of an employee handbook should

91

Table 24.

 

[ Incentive '
‘j Handbook 2l

1
l

RHA
l Herd Size
l BST
’ Veterinary \
m

l Multiple An

l Indepen

 

 

 

 

l
\
decrease th
emploi’ees
pregnancy
are expeCte
those “he
Art
the “We it
aVail them,
omlatiOn s

[a
\t I" .
p OdUCtlt

Table 24.

The Reproduction Culling Rate Model Independent Variable
Descriptions

 

Independent Variable

Description

 

Handbook

Indicates if a farm uses an employee handbook

 

Incentive

Indicates if a farm uses an employee incentive program

 

Handbook and Incentive

Indicates if a farm uses an employee handbook and an
emplgyee incentive program

 

 

 

RHA The farm’s rolling herd average
Herd Size The farm’s milking and dry cow herd size
BST Indicates if a farm uses BST

 

Veterinary Visits

The number of veterinary visits needed on a farm;

 

Staff Veterinarian Day

The number of days a veterinarian was on-staff

 

Free Stall Facility

Indicates if a farm houses its cattle in a free stall facility

 

Multiple Animal Facility

Indicates if a farm houses its milking herd in multiple
animal housing

 

 

 

 

 

Synchronization Indicates if a farm uses estrus synchronization
techniques

Bull Low Indicates if a farm exposed greater than zero but less
than ten percent of their cattle to a herd bull

Bull Medium Indicates if a farm exposed more than ten percent but
less than or equal to thirty percent of their cattle to a
herd bull

Bull High Indicates if a farm exposed more than thirty percent of

 

their cattle to a herd bull

 

decrease the proportion of health culls due to reproduction problems. Providing

employees with incentives for estrus detection and achieving higher-than-anticipated

pregnancy rates can also decrease reproduction culling rates. Reproduction culling rates

are expected to be lower on farms that use both of these management tools as opposed to

those who do not.

Artiﬁcial insemination requires specialized labor. In general, the larger a farm is

the more it can afford to utilize specialized labor. A larger herd size also allows farms to

avail themselves to other reproductive technologies such as pedometers and estrus and

ovulation synchronization. Herd size is expected to be negatively correlated with the

reproduction culling rate.

92

 

It is more difﬁcult to get dairy cattle to conceive as they produce more. Howard
et al. (1992) reported that a manager can expect a 1.5 days open increase for every
genetically induced 785 pound increase in rolling herd average. Fleiscer et al. (2000)
noted that retained placentas and cystic ovaries, conditions with adverse reproduction
repercussions, are positively correlated with milk yield. Cows that produce more milk are
expected to have a higher reproduction culling rate.

Collier et al. (2001) found that BST had no effect on pregnancy rates, days open,
cystic ovaries or abortions. BST promotes production in the latter half of lactation. Cattle
that have difﬁculty conceiving may be more proﬁtable to keep longer using BST than
without BST. This enables the manager more opportunities to breed the animal or to keep
the cow proﬁtably open for longer periods of time. The reproduction culling rate is
expected to be negatively correlated with BST use. Veterinarians work closely with
managers concerning reproduction issues. As most veterinarians visit farms monthly to
check animals for pregnancy, the veterinarians can quickly respond to reproductive
problems. As such, it is expected that less animals will be culled as the number of
veterinary visits or the number of days a veterinarian is on staff should increase.

While tie stall facilities do allow producers to provide very individualized care to
each animal, they are less conducive to estrus detection. Upper Midwest dairymen who
participated in a study on the before and after effects of expansion commented that heat
detection signiﬁcantly improved on farms when they moved into their expanded dairy
free stall facilities (Hadley, 2001). Free stall and multiple animal facilities are expected to

be negatively correlated with the reproduction culling rate.

93

One of the difﬁculties of breeding cattle is detecting estrus. Estrus
synchronization manipulates the cow’s reproductive cycle so that estrus occurs in a
predictable manner. Cattle can then be scheduled for breeding. Farms that use
synchronization technology are expected to have lower reproduction culling rates.

Some managers ﬁnd it prudent to have bulls in the herd to breed or “clean up”
long open cows. Hadley (2001) noted that some dairy expansion managers started to use
bulls for long open cattle rather than to let their average calving intervals increase. While
the genetic merit of bull bred offspring and the safety of having bulls on a farm may be
debatable, it is commonly thought that bulls are more efficient at breeding dairy cattle.
Farms that rely more heavily on natural service are expected to have a lower reproduction
culling rate.

IV. Estimation Results

The were 1,352 observations for Model 1, the Udder and Mastitis Culling Rate
Model (Table 25). Due to the random stratiﬁed sampling technique employed by the
NAHMS Dairy ’96 Survey, these 1, 352 observations represented 94,150 US. dairy
farms. On average, 6.23 percent of dairy cattle were culled for udder and mastitis reasons.
The R-square for the ddder and mastitis culling rate model estimation was 0.029547. The
model was signiﬁcant (p-value = <0.0001).

Table 25. Overall Results for The Udder and Mastitis Culling Rate Model

 

 

 

 

 

 

 

 

Statistic Value
Observations 1,3 52
Wekghted Observations 94,150
Denominator Degrees of Freedom 1,289
Weighted Mean Response 6.230350

 

94

Farms that used handbooks to inform employees about protocols experienced 3.02

percent lower udder and mastitis culling rates than those who did not (Table 26).

Although employees handbooks proved beneﬁcial in reducing udder and mastitis culling

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 26. Parameter Results for The Udder and Mastitis Culling Rate
Model
Parameter Parameter Standard Test T - test P - value
Estimate Error Type1

Intercept 4.8831 0.5982 2 8.29 <0.0001
Handbook -3 .0220 0.8831 1 -3 .42 0.0003
Incentive 0.0549 0.7207 1 0.08 3
Handbook and -0.4606 1.1419 1 —0.40 0.3434
Incentive
Dry Treatment 0.3790 0.6962 1 0.54 1
Wash Pen 8.5592 3.7803 1 2.26 2
Pre-dip 0.4700 1.7037 1 0.28 2
Post Dip 0.6410 0.8474 1 0.76 2
Pre- and Post Dip 0.0467 0.5425 1 0.09 2
Free Stall FacilitL 0.0634 0.6047 2 0.10 0.9165
Multiple Animal 0.5428 0.9370 2 0.58 0.5625
Facility
Herd Size -0.0006 0.0007 2 -0.84 0.4007
Sand Bedding 0.9632 0.7690 2 1.25 0.2106
Wood Bedding -0.4758 0.4745 2 -1.00 0.3162
Compost Bedding -1.9420 0.8331 2 -2.33 0.0199
Rubber Mat Bedding 0.3519 0.5133 2 0.69 0.4931
Tire Bedding -0.8582 1.5961 2 -0.54 0.5904
News Paper Bedding 1.1350 0.9183 2 1.24 0.2167
Mattress Bedding 0.4136 0.7441 2 0.56 0.5784
Stalk Bedding 0.1625 0.6570 2 0.26 08047
Other Bedding -0.6165 1.3506 2 -0.46 0.6481
BST 1.1892 1.1078 1 1.07
Veterinary Visits 0.0181 0.0143 1 1.26 0.1032
Staff Veterinarian Day -0.0043 0.0088 1 -0.49 0.3122
Milk Quality Test 0.8359 1.6218 1 0.52 2

 

I A “1” signiﬁes a one-tailed test. A “2” indicates a two-tailed test.
2 Insigniﬁcant due to an incorrect sign on a one-tail test.

rates, incentives did not. This may support the premise that incentive programs should

not be used to improve performance in activities such as milking. Milking activities,

95

 

especially in larger herds, are generally conducted by multiple employees, making
individual performance in multiple employee activities is hard to monitor and reward
with incentives.

Surprisingly, none of the udder and teat sanitation parameters (dry treatments,
wash pens, pre-dipping, post dipping, pre- and post dipping) were statistically signiﬁcant.
In fact, these parameters carried the wrong sign. These results may be inﬂuenced,
however, by the cross sectional design of the survey. As such, there was no information
concerning the before and after effects of the implementation of these programs on each
farm. Secondly, the quality of how these activities were conducted on each farm was not
addressed in the survey.

Facility type as well as herd size seemed to have little effect on the udder and
mastitis culling rate. For the most part, bedding type also did not effect udder and mastitis
culling rates. Only composted manure bedding reduced udder and lameness culling rates.
Farms that used composted manure bedding had 1.94 percent lower udder and lameness
culling rates (p-value = 0.0199). The effects associated with BST, veterinary visits, staﬁ‘
veterinarian days, and milk quality test parameters were statistically insigniﬁcant.

Table 27. Overall Results for The Lameness and Injury Culling Rate Model

 

 

 

 

 

 

 

 

Statistic Value
Observations 1,3 52
Weighted Observations 94,150
Denominator Degrees of Freedom 1,289
Weighted Mean Response 3.025042

 

There were 1,3 52 observations for the lameness and injury culling rate model,

representing 94,150 US. dairy farms (Table 27). Three percent of all the cattle were

96

culled due to lameness and injury problems. The model had an R-square of 0. 1 10654.
The model was signiﬁcant (p-value < 0.0001).

As the proportion of cattle with hairy heel warts increased, the lameness and
injured culling rate increased by 4.47 percent (Table 28). The use of footbaths and/or
hoof trimming did not signiﬁcantly affect the lameness and injury culling rate. Once
again, the insigniﬁcant results for these parameters may be due to the cross sectional
nature of the survey. Although the lameness and injury culling rate on farms with free
stall facilities were not signiﬁcantly different than tie stall facilities, multiple animal
facilities had signiﬁcantly lower (-0.79 percent) lameness and injury culling rates. Cattle
that were on farms that provided predominantly soﬁ walking surfaces also experienced
signiﬁcantly lower (-0.53 percent) lameness and injury culling rates. The fact that both
multiple animal facilities, which generally have dirt pack surfaces, and farms with
predominantly soﬁ cattle surfaces have lower culling rates indicates that the hardness of
cattle walking surfaces seems to affect lameness and injury culling rates.

Farms that used composted manure bedding or “other” bedding experienced lower
lameness and injury culling rates. Farms using composted manure bedding experienced
lameness and injury culling rates that were 1.15 percent less than those who bedded with
straw. Farms that used “other” bedding alternatives experienced lameness and injury
culling rates that were 1.49 percent less than those that bedded with straw. There was no

information provided that described what the “other” bedding types were.

97

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 28. Parameter Results for The Lameness and Injury Culling Rate
Model
Parameter Parameter Standard Test T - test P - value
Estimate Error Type1

Intercept 0.4547 0.5741 2 0.79 0.4285
Hairy Heel Wart 4.4680 1.1140 1 4.01 0.0001
Footbath -0. 5346 0.5473 1 -0.98 0.1644
Hoof rtirn 0.1034 0.4074 1 0.26 2
Footbath and Hoof 0.9113 0.8650 1 1.05 I
Trim
Free Stall Facility 0.4434 0.3630 2 1.22 0.2222
Multiple Animal -0.7868 0.3042 1 -2.59 0.0049
Facility
Exercise Lot -0.2222 0.4082 1 -0.54 0.2932
Soﬁ Surface -0.5280 0.4074 1 -1.30 0.0977
Textured Concrete -0.1324 0.3463 1 -0.38 0.3511
Surface
Sand Bedding -0.5422 0.3674 2 -1.48 0.1402
Wood Bedding 0.1892 0.3405 2 0.56 0.5784
Compost Bedding -1.1521 0.4788 2 -2.41 0.0163
Rubber Mat Bedding 0.5773 0.3932 2 1.47 0.1423
Tire Bedding 1.6238 1.3145 2 1.24 0.2169
Newspaper Bedding 0.1925 0.8781 2 0.22 0.8265
Mattress Bedding -0.4434 0.5429 2 -0.82 0.4143
Stalk Bedding 0.7970 0.5382 2 1.48 0.1389
Other Bedding -1.4906 0.4265 2 -3.49 0.0005
Veterinary Visits 0.0141 0.0079 1 1.79 0.0370
Staff Veterinarian Day 0.0602 0.0118 1 5.09 2
RHA 0.0001 <0.0001 1 3.31 0.0005
BST -0.6770 0.6671 1 -1.01 0.1552

 

 

 

 

 

 

I A “1” signiﬁes a one-tailed test. A “2” indicates a two-tailed test.
2 Insigniﬁcant due to an incorrect sign on a one-tail test.

The number of veterinary visits or the number of days a veterinarian was on a

farm’s staff did not seem to affect the lameness and injury culling rates. Lameness and

injury culling rates increased as a farm’s production per cow (RHA) increased. Although

BST was negatively correlated with lameness and injury culling rate, the effect was not

signiﬁcant at a p-value of 0. 10 percent or less.

98

 

 

There were 1,141 observations for the disease culling rate model (Table 25). The
sample represented 93,944 US. dairy farms. A little over one percent of the cattle
inventory were culled due to disease. The model had an R-square of 0.08353. The model

was signiﬁcant (p-value < 0.0001).

Table 29. Overall Model Results for The Disease Culling Rate Model

 

 

 

 

 

 

 

Statistic Value
Observations 1,141
Weighted Count 93,944
Denominator Degrees of Freedom 1,076
Weighted Mean Response 1.061111

 

Although not signiﬁcant, the incidence of disease actually decreased instead of
increased as the proportion of purchase cattle increased on the farms (Table 30). Keeping
youngstock away from nose-to—nose contact with other cattle did not signiﬁcantly reduce
disease culling rates at a p-value of0. 1000 or less, but farms that
were able to keep youngstock away from nose-to-nose contact with other species
experienced 1.59 percent lower disease culling rates. Individually or in combination with
each other, the proportion of vaccines used on a farm, the proportion of vaccines required
prior to purchasing a cow, the proportion of tests required prior to purchasing a cow, and
the presence of a quarantine did not signiﬁcantly reduce disease culling rates at a p-value
of 0. 1000 or less. Farms that had cows that tested positive for Johne’s Disease
experienced 0.96 percent higher disease culling rates. The number of veterinary visits and
the number of days worked by an on staff veterinarian did not signiﬁcantly affect the

disease culling rates. The descriptive farm and production variables also did not

Signiﬁcantly affect disease culling rates.

99

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 30. Parameter Results for The Disease Culling Rate Model

Parameter Parameter Standard Test T - test P - value

Estimate Error Typel

Intercept -0. 5269 0.8319 2 -0.63 0.5266
Purchase -0.1804 0.4158 1 -043 2
Nose to Nose Cow -0.9290 0.9382 1 -0.99 0.1612
Contact
Nose to Nose Other -1.5858 0.8895 1 -1.78 0.0375
Contact
Vaccines Used -1.5966 1.3180 1 -1.21 0.1130
Vaccines Reﬁned -0.1101 0.4971 1 -0.22 0.4124
Test Required -0.5141 1.0260 1 -0. 50 0.3083
Quarantine -0. 6476 0.7218 1 -0.90 0.1849
Vaccines and Tests -0.5284 0.5142 1 -1.03 0.1522
Vaccines and -0.8767 0.8189 1 -l.07 0.1423
Quarantine
Tests and Quarantine -3.5862 2.8452 1 -1.26 0.1039
Vaccines, Tests and 0.3113 0.5511 1 0.56 0.2862
Quarantine
Johne’s Positive 0.9551 0.5925 1 1.61 0.0536
Veterinary Visits -00017 0.0076 1 -022 2
Staff Veterinarian Day 0.0076 0.0063 1 1.20 2
Herd Size 0.0002 0.0006 2 0.26 0.7941
Free Stall -0.0669 0.2916 1 -0.23 0.4093
Multiple Animal 2.8781 3.1988 1 0.90 2
Facility
RHA 0.0001 0.0001 2 1.07 0.2830
BST 0.2219 0.7468 2 0.30 0.7664

 

 

 

 

 

 

I A “1” signiﬁes a one-tailed test. A “2” indicates a two-tailed test.
2 Insigniﬁcant due to an incorrect sign on a one-tail test.

There were 1,337 observations used in the reproduction culling rate model

 

estimation (Table 31). The 1,337 observations represented 92,642 US. dairy farms. The
farms culled 6.7 percent of their cattle due to reproductive problems. The R-square for

the model was 0.057797, and the model estimation was signiﬁcant (p-value < 0.0001).

100

Table 31. Overall Results for The Reproduction Culling Rate Model

 

 

 

 

 

 

 

 

Statistic Value
Observations 1,33 7
Weighted Count 92,642
Denominator Degrees of Freedom 1,289
Weighted Mean Response 6.713675

 

Table 32 shows that individually having a handbook or incentive program did not
signiﬁcantly reduce reproduction culling rates, but, farms that used both experienced 1.23
percent fewer culling rates. As herd size increased, the reproduction culling rate
decreased. This could possibly be due to the ability of larger farms to hire more
specialized labor and adopt more costly reproduction technology. As expected,
reproduction culling rates increased as rolling herd average increased.

Veterinary visits, on staff veterinarian days, facility type, and the use of estrus ‘
synchronization did not signiﬁcantly affect reproduction culling rates. However, bull use
did signiﬁcantly reduce reproduction culling rates. Herds that used bulls but exposed less
that 10 percent of their cattle to the bulls experienced 2.37 percent lower reproduction
culling rates. Farms that exposed between 10 and 30 percent of their cattle to bulls
experienced 2.24 percent lower reproduction culling rates. Herds that used bulls on more

than 30 percent of their cattle experienced 2.80 percent fewer culling rates.

101

Table 32.

Parameter Results for The Reproduction Culling Rate Model

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Parameter Parameter Standard Test T - test P - value
Estimate Error Type1

Intercept 1.6437 2.1955 2 -0.63 0.5266
Handbook -1 .5612 1.9012 1 -0.82 0.2059
Incentive -0.8954 0.7517 1 -1.19 0.1169
Handbook and -1.2254 0.8742 1 -1.40 0.0806
Incentive
Herd Size -0.0036 0.0024 1 -l .46 0.0722
Rolling Herd Average 0.0003 0.0002 1 1.79 0.0370
BST -1.4624 1.7026 1 -0.86 0.1953
Veterinary Visits 0.1202 0.1086 1 1.11 2
Staff Veterinarian Day 0.0036 0.0212 1 0.17 2
Free Stall Facility -0. 1937 1.0913 1 -0.18 0.4296
Multiple Animal -0.5307 0.9629 1 -0.55 0.2908
Facility
Synchronization -l .0605 2.7820 1 -0.38 0.3516
Bull Low -2.3653 1.7585 1 -1.35 0.0894
Bull Medium -2.2354 1.4193 1 -1.57 0.0578
Bull High -2.8015 1.0821 1 -2.59 0.0049

 

I A “l” signiﬁes a one-tailed test. A “2” indicates a two-tailed test.
2 Insigniﬁcant due to an incorrect sign on a one-tail test.

V. Summary and Discussion

Four ordinary least squares were developed to determine the effect that select

management factors had on reducing the udder and mastitis, lameness and injury, disease

and reproduction culling rates. Only a few of the management factors signiﬁcantly

contributed to the udder and mastitis, lameness and injury, disease and reproduction

 

culling rates at a p-value of 0. 1000 or less. It is important to remember, however, that the
survey was cross sectional. As such, the before and after effects of implementing the
management programs and protocols could not be assessed. Additionally, how well the
management programs and policies were implemented was not assessed in the survey. If
information had been available concerning the before and after effects of implementing

management programs and the quality of the management program implementation,

102

some of the management factors, such as the effect of pre-dipping and post dipping on
the udder and mastitis culling rate, may have been signiﬁcant.

Only two management factors signiﬁcantly affected udder and mastitis culling
rates. Although providing employees with an incentive program did not reduce udder and
mastitis culling rates, simply providing employees with a handbook reduced udder and
mastitis culling rates by 3.02 percent. Farms that used composted manure experienced a
1.94 percent lower udder and mastitis culling rates. Although the biosecurity measures of
vaccinations, tests and quarantines did not prove to signiﬁcantly reduce disease culls, the
fact that hairy heel warts and Johne’s Disease signiﬁcantly increased lameness and injury
and disease culling rates indicates that producers should limit the exposure of their cattle
to these contagious diseases.

Three factors were positively correlated with the lameness and injury culling rate.
As the proportion of cattle with hairy heel wart increased, the lameness and injury culling
rate increased by 4.47 percent. Thus, it is important for managers to consider hairy heel
wart prevention programs. The number of veterinary visits was also positively correlated
with lameness and injury culling rate. For every veterinary visit the lameness and injury
culling rate increased by 0.01 percent. This probably indicates that producers are
primarily using veterinary visits for diagnostic and treatment purposes as opposed to
prevention methods. Rolling herd average was also positively correlated with the
lameness and injury culling rate. The lameness and injury culling rate increased with
Rolling Herd Average.

Farms with multiple animal facilities and farms with soﬁ cattle surfaces had

signiﬁcantly lower lameness and injury culling rates. Farms with multiple animal

103

facilities experienced a 0.79 percent lower lameness and injury culling rate than tie stall
farms. Farms with a predominantly soﬁ cattle surface experienced 0.59 percent lower
lameness and injury culling rate. As many multiple animal facilities utilize dirt packs, this
provides evidence that a farmer should consider providing his or her cattle with a soﬁ
walking and standing surface to reduce their lameness and injury culling rate. Composted
manure and “other” bedding was also negatively correlated with the lameness and injury
culling rate. Farms using composted manure bedding experienced a 1.15 percent lower
lameness and injury culling rate. Farms using “other” bedding types experienced a 1.49
percent lower lameness and injury culling rate.

There were only two management programs that signiﬁcantly affected disease
culling rates. Managers who prevented their youngstock from nose-to-nose contact with
other species had 1.59 percent lower disease culling rates than managers who did not.
Farms that had animals that tested positive for Johne’s Disease exhibited 0.96 percent
higher disease culling rates.

Employee handbooks were also effective at reducing the reproduction culling rate
but only when combined with incentives. Farms that used both handbooks and incentives
experienced a 1.23 percent lower reproduction culling rate than farms that did not use
handbooks and incentives. Herd size was also negatively correlated with reproduction
culling rates, possibly indicating that larger farms are more apt to hire specialized labor
and to adopt reproduction technologies. Using herd bulls was an effective method of
reducing reproduction culling rates. Farms that use herd bulls on less than 10 percent of
the herd experienced 2.37 percent lower reproduction culling rates. Farms using bulls on

10 to 30 percent of their cattle experienced 2.24 percent lower culling rates. Farms that

104

used herd bulls on more than 30 percent of their cattle experienced 2.80 percent lower
reproduction culling rates. Rolling herd average was positively correlated with
reproduction culling rates.

Although these models indicated which management factors affected health
culling rates, it did not provide information concerning whether such programs were
ﬁnancially successful. In the next chapter, a decision support system that will enable
advisors and producers to determine the ﬁnancial beneﬁts of reducing health culling rates

is described.

105

CHAPTER 5.
A CULLING RATE REDUCTION FINANCIAL FEASIBILITY DECISION
SUPPORT SYSTEM
1. Introduction

Producers are faced with many decisions. One type of decision is whether to
adopt culling rate reduction technologies (rubber cattle walkways to reduce lameness
culls, pedometers to reduce reproduction culls, etc), or management programs (new
milking protocols to reduce udder and somatic cell count culls, allocating more labor to
repair facilities and stalls to reduce injury culls, etc). The decision to invest in culling
rate reduction technologies or programs can be complex. If the producer over-invests, the
ﬁnancial gains from a culling rate reduction (increased milk revenues, decreased
operating expenses, increased surplus replacement heifer sales) are more than offset by
the investment and required operating costs associated with the new technology or
management program. If the producer under-invests in the technology or program, the
producer may not see a sufficient culling rate reduction to justify the expense.

This chapter proposes a decision aid to assist producers in determining the
ﬁnancial merit of culling rate reduction technologies and management programs. This
decision support system (DSS) is different than previous culling programs. Rather than
determining the economically optimal culling rate, this decision aid determines whether it
is ﬁnancially feasible to reduce a farm’s current culling rate to a targeted level.

The information needed to run the DSS is discussed in Chapter II. A general
overview of the procedure used by the DSS to determine the ﬁnancial feasibility of

reducing culling rates is provided in Section II. Section III discusses the information

106

needed to run the DSS. Section IV details how the ﬁnancial feasibility calculations are
made. Section V describes how to input information into the DSS input ﬁelds .The
calculations used to determine the DSS output and the output ﬁelds themselves are
described in Section VI.
II. A General Overview of the DSS

Figure 1 shows a schematic of how the DSS operates. In Step 1, the proﬁtability
of the current culling rate and pattern of those culls is determined. First, herd information
such as herd size, herd inventory, milk production per lactation, somatic cell counts,
lactation speciﬁc culling rates, the within lactation removal schedule,1 and the effect of
culling on milk production and somatic cell counts are entered into the DSS. Next, price,
cost and other ﬁnancial information are entered into the DSS. Price information needed
includes the expected milk price, somatic cell count premium, heifer and bull calf values,
feed, labor and other direct expenses, cull cow prices, replacement heifer prices, and
treatment expenses for culls due to udder and somatic cell count problems, infertility,
lameness and injury, disease and death. The other ﬁnancial information includes the
farm’s debt-to-asset ratio, an opportunity cost of equity capital, the farm’s interest rate,
marginal tax rate, and capital gains tax rate. Revenues and expenses are assigned to each
cow. These assignments are made based upon their lactation, lactation month, and
whether they will be culled or retained. The DSS then sums all of the revenues and
expenses to determine the proﬁt for that month and are discounted to reﬂect the time
value of money. The monthly values are then summed to determine the present value of

the proﬁts for the current culling rate and pattern.

 

' The distribution of when culls occur during a given lactation.

107

Figure 1. Overview of the DSS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Step 1 Current Price and The
Culling Financial Proﬁtability of
Rate Herd —t Information —+ the Current
Information Input Field Culling Rate
Input Field and Pattern
Step 2 Desired Price and The
Culling Financial Proﬁtability of
Rate Herd —t Information -—-> the Desired
Information Input Field Culling Rate
Input Field and Pattern
Step 3 The The The Cost of the The
Proﬁtability Proﬁtability Health Cull Proﬁtability
of the of the Reduction of
Desired — Current - Technology or = Reducing
Culling Culling Program Health
Rate and Rate and Culls
Pattern Pattern

 

 

 

 

 

 

 

 

 

 

 

 

 

 

In Step 2, the proﬁtability of a desired culling rate and culling pattern is basically
determined in the same manner as the current culling rate and pattern with one exception.
Presumably, the user has information about a technology or program that will reduce the
user’s lactation speciﬁc culling rates by some percentage. The user will input this
information into the desired herd information input ﬁeld. The DSS then calculates the
monthly proﬁtability of the desired culling rate and culling pattern and discounts these
cash ﬂows for the time value of money.

Next, the DSS subtracts both the present value of the proﬁts for the current
culling rate and culling pattern and the present value of the cash ﬂows associated with
adopting the health cull reduction technology and program from the present value of the

proﬁts from the present value of the proﬁts for the desired culling rate and culling

108

pattern. This value represents the net present value associated with reducing the current
culling rate and culling pattern to the desired level with the proposed technology or
program. If this value is greater than or equal to zero, the technology or program should
be adopted. If negative, the technology or program should not be adopted as it is too
costly.
III. Information Needed to Run the DSS

To run the DSS, a user will need to have herd and ﬁnancial records available to
input certain information. The needed information is shown in Table 32. “Critica ” refers
to information in which the DSS provides no sample values. “Helpﬁll " refers to
information whereby the DSS generates sample values, but the DSS estimation results
would be more accurate for the user if he or she would input their own information.

Items 1 through 10 can be found by consulting with DHI records or other
production and herd health records. Items 11 through 15 can be determined through
discussions with milk company representatives, meat industry representatives, or
Extension personnel. The expected heifer and bull calf mortality rate can be determined
by examining production records. The farm’s debt to asset ratio can be found by
analyzing the farm’s current balance sheet. The producer can obtain the cost of debt by
talking with their lender. Determining an expected cost of equity capital can be difﬁcult.
The cost of equity capital should be larger than the interest rate. Extension personnel or
other professionals who specialize in dairy ﬁnance should be able to provide this
information. Treatment costs can be obtained by talking with a veterinarian or by
analyzing a farm’s itemized veterinary expenses. Feed costs, labor costs, and “other”

costs can be determined by analyzing income statements. The genetic improvement rate

109

can be ascertained from dairy geneticists. The inflation rate can serve as a proxy for the

labor and other expense growth rate if a growth rate for each cannot be determined. ‘

Table 33. Information Needed for the DSS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Information Type Need Type
1) Average First Lactation Milk Production Critical
2) Average Production Levels for Other Lactations Helpful
3) Average First Lactation Somatic Cell Count Critical
4) Average Somatic Cell Counts for Other Lactations Helpful
5) Average Annual Culling Rate Critical
6) Current Herd Inventory Critical
7) Lactation Speciﬁc Culling Rates by Cull Type Helpful
8) Within Lactation Removal Schedule Helpful
9y Effect of Culling on Milk Production by Cull Type Helpful
10) Effect of Culling on Somatic Cell Count by Cull Type Helpful
11) Expected Milk Price Critical
12) Expected Somatic Cell Count Premium Critical
13) Expected Cull Cow Price Critical
14) Expected Replacement Heifer Price Critical
15) Expected Heifer and Bull Calf Price Critical
16) Expected Heifer and Bull Calf Mortality Rate Critical
”mebt to Asset Ratio Critical
18) Expected Interest Rate Critical
19) Expected Cost of Equity Capital Critical
20) Expected Treatment Cost Per Health Cull Type Critical
21) Expected Feed Cost Per Cow Critical
22) Expected Labor Cost Per Cow Critical
23) Expected Other Expense Per Cow Critical
24) Expected Genetic Improvement Rate Critical
25) Expected Labor and Other Expense Growth Rate Critical
26) Cull Reduction Technology Purchase and Installation Critical
Expense
27) Tax-based Depreciable Life of the Technology Critical
28) Operating Expenses Associated with the Cull Reduction Critical

Technology or Program

 

 

llO

 

IV. Determining the Financial Returns Associated with Reduced Health
Culling Rates

In Step 1, the DSS estimates how cattle will move in and out of the herd over a
240 month planning horizon using a farm’s current lactation speciﬁc culling rate and
within lactation removal schedule. The number of cattle removed and replaced as well as
the reason for the cull are recorded. All viable cattle at the end of the 240th month are sold
in a herd dispersal sale.

Next, the revenues and expenses for both the retained and culled cattle under the
current culling rate and culling pattern are calculated for each month of the planning
horizon. The revenue and expense items include:

a) the proceeds associated with selling mature cattle for dairy purposes;

b) cull cow receipts (adjusted for age and whether the cull was
lame or injured);

c) calf value credits;

d) milk revenues (adjusted for the genetic milk production
improvement rate and whether a cow is retained or culled);

e) somatic count premiums and discounts (adjusted for whether a
cow is retained or culled); .

f) feed, labor and other variable expenses; and,

g) a charge for the additional veterinary expenses associated with
health cull reasons.

These values are assigned based upon the cow’s lactation number, the lactation
month the cow is in, the typical production for that lactation and lactation month, whether
or not the cow is culled in the lactation month or in a future lactation month in the same

lactation, and based upon culling reason. Cattle that are culled typically produce less than

111

their herdmates. Cattle that produce less also tend to eat less than their herdmates. As
such, both the feed and labor expenses were adjusted per hundredweight produced.
In Step 3, the present value of the cash ﬂows resulting from the current culling

rate and culling pattern, PVCumm, are determined by the following formula:

PV 2 Z 2: { [((Mature Cattle Sold for Dairy Purposes Receiptsn

Current
+ Cull Cow Receiptsn + Calf Value CreditsI1 + Milk
Revenuesn + SCC Premiums“) - (Variable
Expensesn + Additional Veterinary Expensesn))

*(1-t)l/(1- 10"}
where “t " is the farm’s marginal tax rate or capital gains tax rate if a capital asset, “k" is
the farm’s after tax weighted average cost of capital, and “n " is the planning horizon
month.

In Step 4, Step 1 is then repeated using the desired culling rate and culling pattern
offered by the new health cull reduction technology or program. In Step 5, the revenues
and expenses for both the retained and culled cattle under the desired culling rate and
culling pattern are calculated for each month of the planning horizon. The revenue and
expense items of the desired culling rate and culling pattern are identical to those of the
current culling rate and pattern except that the desired calculations include the increase
in surplus replacement heifer sales.2 In the sixth step, the present value of the cash ﬂows
resulting from the desired culling rate and culling pattern, PVDes,,ed, is determined by the

following formula:

PVDmed : Z 2: { [((Mature Cattle Sold for Dairy Purposes Receiptsn

+ Cull Cow Receiptsn + Calf Value Creditsn + Milk
Revenuesn + SCC Premiumsn + A Surplus Replacement

 

2 The increase in surplus heifer sales for any given planning horizon month equals the estimated number of
cows culled under the current culling rate and culling pattern in that month minus the estimated number of
cows culled under the desired culling rate and culling rate pattern in that same month.

112

Heifer Sales") — (Variable Expensesn + Additional Veterinary
Expensesn))*(l-t)] / (l- k)“ }

where “A Surplus Replacement Heifer Sales, " refers to the change in surplus replacement
heifer sales.

The DSS then calculates the present value of the cash ﬂows associated with
investing in and operating the proposed health cull reduction technology or program,

PVMWWW, using the following formula:

PVInvestment = ' 10 + Z 31% [ (' In (if greater than n = 0) —' ((TeChUOIOgy Operating
Expensesn)*(l-t)) + (Technology Depreciation

Expensen“ t + Technology Terminal Value") ]

/ (l- 10" }
where “1,, " refers to an initial investment or reinvestment in the capital assets of the
health reduction technology or program, “Technology Operating Expenses" ” refers to the
operating expenses associated with implementing the new technology or program in
planning horizon month n, “Technology Depreciation Expense n" t" refers to the
depreciation expense tax shield in planning horizon month t, and “Technology Terminal
Value” " refers to the sale of any of the new health reduction technology or programs
capital assets in month 71.3

The DSS then calculates the net present value of the culling rate reduction,

NPVRedmon. This value is calculated by using the following formula:

”I VReduction = PV Desired " PV Current ' PV Investment-

 

3 The DSS automatically assumes that all capital assets of the health reduction technology or program will
be sold at the end of their depreciable asset and sold so that no capital gains or capital losses occur. The
DSS then automatically reinvests in the technology once again at a cost equal to the initial investment
adjusted for inﬂation. At the end of the 240 month planning horizon, the new technology’s capital assets
are sold for a price so no capital gains or losses occur.

113

If the NP VRedmm value is greater than or equal to zero, it indicates that the producer
would be ﬁnancially improved by investing in the new health cull reduction technology
or program.

The NP VRedumn value represents the net present value of the cash ﬂows for the
entire 240 month period. This value may be difficult for a farm manager or advisor to
fully understand. As such, the DSS then makes a subsequent calculation to express the
NPVRedumn value as a monthly equivalent expression. This is done using the following
formula:

Monthly Equivalent Returns = NPVRcdmion/ {[1-(l+k)'"]/k}.

V. The DSS Input Fields

A user starts the DSS by ﬁrst choosing the correct version for his or her average
calving interval.4 The user then enters the herd size and herd inventory information into
the “Herd Size and Herd Inventory Input Fields ” (Figure l). The herd size includes all
milking and dry cows. It does not include replacement heifers that have not calved. The
DSS assumes that the manager has enough replacement cattle to satisfy the replacement
needs of the herd. The herd inventory details the number of cattle within a given lactation
(l — 10) and lactation month (1- 13 including two dry period months). In Figure 2, the
user has 1 10 head of milking ﬁrst lactation cows that are evenly distributed between
lactation months 1 through 1 1. Twenty cows that are waiting to start their second

lactation are evenly distributed between the two dry period months.

 

4 . . . .
Currently, only a tlurteen month verSlon 15 available; however, a twelve, fourteen and ﬁfteen month
version are planned.

114

Next, the user inputs the expected milk production per cow per lactation into the
“Expected Production by Lactation Input Field ” (Figure 3). Upon inputting the ﬁrst
lactation expected milk production, the DSS displays the typical production level for the
second through tenth lactations in the “A verage Production by Lactation Table ” (also
Figure 2.). The milk production estimates in the “A verage Production by Lactation
Table ” are determined by multiplying the inputted ﬁrst lactation production level by the
percentages displayed in Chapter 2 Table l in the 305 Day Actual Milk column. The user
can elect to enter these estimated values for lactations 2 through 10 or enter their own
expected values. Figure 2 shows that the user’s ﬁrst lactation cattle are averaging 20,000
pounds per lactation. The average production for the remaining lactations are set

according the estimates listed in the “A verage Production by Lactation Table. ”

115

 

 

92.3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o o o o o o o o o 2 be 2
$25
o o o o o o o o o 2 e8 2
o o o o o o o o o o_ 2
o o o o o o o o o o_ o_
o o o o o o o o o o_ a
o o o o o o o o o o. w
o o o o o o o o o o_ \i
o o o o o o o o o o. o
o o o o o o o o o o_ m
o o o o o o o o o o_ v
o o o o o o o o o o_ m
o o o o o o o o o o. m
o o o o o o o o o o. _
o_ o w \i o m v m N _ 5:02

€233th CESSNA GOUMHUNJ COCQHUMA =23“an CEHSUGA comwmuowi— €22.30th COS—wwomt— COESUMA :Oﬂwuowta
b953—
Be:

a: 4 Be; 38 New can mac—:5 mo L383: owﬁo>< _

 

 

 

 

 

 

 

22m 2:...— b3=o>5 ...—u: use «am can:

to. 9.sz

116

Figure 3. Expected Production by Lactation Input Field

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Your Expected Pounds
Production by

Lactation

Lactation 1 20000
Lactation 2 22308
Lactation 3 23000
Lactation 4 23136
Lactation 5 22950
Lactation 6 22482
Lactation 7 21790
Lactation 8 21158
Lactation 9 20378
Lactation 10 19736

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Average DHIA Pounds
Production by

Lactation

Lactation 1 20000
Lactation 2 22308
Lactation 3 23000
Lactation 4 23136
Lactation 5 22950
Lactation 6 22482
Lactation 7 21790
Lactation 8 21158
Lactation 9 203 78
Lactation 10 19736

 

 

 

Culled dairy cows — whether sold for dairy purposes, low milk production, or

poor health — typically produce less milk than their herdmates. The next input ﬁeld

concerns the producer’s expectations concerning the anticipated milk production decrease

that a culled animal typically experiences prior to being culled. Figure 4 displays the

“Culling Associated Production Eﬂects Input Field. ” The user enters expected

production decreases associated with health, mortality, production and sold for dairy

culls. A “DHIA Average Culling Associated Production Eﬂects ” is also displayed to

assist the user. These percentages are based upon the comparison of the projected 305

day milk production of healthy cattle versus those of cattle that were culled. This

comparison was reported in Chapter 2 . The user can choose to enter the average DHIA

values or their own expected values. In the example, the user entered the typical values.

117

 

 

Figure 4. Culling Associated Production Effect Input Field

 

 

 

 

 

 

 

 

 

 

 

 

 

Your Culling Associated Production Health Mortality Production Sold for
Effects Culls Culls DairyCulls
Lactation 1 Production "/0 Decrease 6 16 26 11
Lactation 2 Production % Decrease l 7 7 4
Lactation 3 Production % Decrease 0 6 11 3
Lactation 4 Production % Decrease l 6 11 3
Lactation 5 Production % Decrease 2 11 10 1
Lactation 6 Production % Decrease 0 7 9 1
Lactation 7 Production °/o Decrease -2 5 3 1
Lactation 8 Production % Decrease 0 8 5
Lactation 9 Production % Decrease 0 ll 7 5
Lactation 10 Production % Decrease 6 1 ll 6

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Average DHIA Culling Associated Health Mortality Production Sold for
Production Effects Culls Culls Dairy Culls
Lactation 1 Production % Decrease 6 16 26 11
Lactation 2 Production % Decrease l 7 7 4
Lactation 3 Production % Decrease 0 6 ll 3
Lactation 4 Production % Decrease 1 6 11 3
Lactation 5 Production % Decrease 2 ll 10 1
Lactation 6 Production % Decrease 0 7 9 1
Lactation 7 Production % Decrease -2 5 3 1
Lactation 8 Production °/o Decrease 0 9 8 5
Lactation 9 Production % Decrease 0 11 7 5
Lactation 10 Production % Decrease 6 14 11 6

 

 

 

 

 

 

 

 

The ﬁfth and sixth input ﬁelds concern somatic cell counts (SCC). Figure 5 shows
the “SC C by Lactation Input Field. ” In this ﬁeld, the user inputs their expected ﬁrst
lactation SCC. Upon entering their expected ﬁrst lactation SCC, the “Average DHIA SCC
by Lactation” table displays SCC estimates for lactations 2 through 10. These estimates
are made by using the SCC percentage adjustment values shown in Table 1 of Chapter 2.
Once again, the user can elect to input these values or their own expected values. In this

case, the manager decides to enter the average DHIA SCC values.

118

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ﬁgure 5. SCC by Lactation Input Field
Your SCC by lactation SCC Average DHIA SCC by SCC
Lactation
Lactation 1 200000 Lactation 1 200000
Lactation 2 254850 Lactation 2 254850
Lactation 3 285900 Lactation 3 285900
Lactation 4 308700 Lactation 4 308700
Lactation 5 326460 Lactation 5 326460
Lactation 6 346280 Lactation 6 346280
Lactation 7 356460 Lactation 7 356460
Lactation 8 358480 Lactation 8 358480
Lactation 9 363560 Lactation 9 363560
Lactation 10 362060 Lactation 10 362060
Figure 6. Cullirg Associated SCC Effect Input Field
Your Culling Associated SCC Effects Health Mortality Production Sold for
Culls Culls Culls DairyCulls
Lactation 1 SCC % Increase 18.75 5.29 28.56 3.84
Lactation 2 SCC % Increase 14.86 -0.51 19.40 1.90
Lactation 3 SCC % Increase 12.92 -1 .15 14.80 3.06
Lactation 4 SCC % Increase 12.71 —4.64 11.44 5.93
Lactation 5 SCC % Increase 12.41 —5.15 12.52 7.09
Lactation 6 SCC % Increase 9.35 -9.61 10.19 6.25
Lactation 7 SCC % Increase 8.68 -471 5.98 7.37
Lactation 8 SCC % Increase 10.60 -5.71 8.92 7.03
Lactation 9 SCC % Increase 10.96 -6. 13 11.48 3.37
Lactation 10 SCC % Increase 7.83 -7.57 11.11 4.67
Average DHIA Culling Associated Health Mortality Production Sold for
SCC Effects Culls Culls Culls Dairy Culls
Lactation 1 SCC % Increase 18.75 5.29 28.56 3.84
Lactation 2 SCC % Increase 14.86 -0.51 19.40 1.90
Lactation 3 SCC % Increase 12.92 -1.15 14.80 3.06
Lactation 4 SCC % Increase 12.71 -4.64 11.44 5.93
Lactation 5 SCC % Increase 12.41 -5.15 12.52 7.09
Lactation 6 SCC "/0 Increase 9.35 -9.61 10.19 6.25
Lactation 7 SCC % Increase 8.68 -4.71 5.98 7.37
Lactation 8 SCC % Increase 10.60 -5.71 8.92 7.03
Lactation 9 SCC % Increase 10.96 -6.13 11.48 3.37
Lactation 10 SCC °/o Increase J 783 -7.57 11.11 4.67

 

 

 

 

 

119

 

 

 

 

 

Figure 6 displays the “Culling Associated SCC Effects Input Field " and its
accompanying “Average DHIA Culling Associated SCC Eﬂects” table. Cattle that are
culled typically have different SCC than their cohorts (Chapter 2). Thus, the user needs to
input the expected SCC percentage increase or decrease associated with a health,
mortality, production, or sold for dairy cull. The “A verage DHIA Culling Associated SCC
Effects " table lists the typical SCC adjustments discussed in Chapter 2. In this example,
the user chooses to enter the “typical” culling induced SCC increases.

The next two input ﬁelds concern the lactation speciﬁc culling rates. In the
“Current Lactation Speciﬁc Culling Rates Input Field” (Figure 7), the user enters the
current total culling rate and the speciﬁc problem culling rate for each lactation. The
speciﬁc problem culling rates include sold for dairy, low production, udder and SCC,
reproduction, disease, and mortality culling rates. The ”Non-Death Health Cull Rates”
and “Total Culling Rate Check " values are calculated for the user. Culling rates
expressed as a percentage of average annual milking and dry cow herd size as shown in

the following formulas:

Total Culling Rate = (Total number of animals culled per year /
Average annual milking and dry cow herd size)*
100 % ; and,

Speciﬁc Problem

Culling Rate = (Total number of animals culled per year for a given
problem / Average annual milking and dry cow herd
size)* 100%.

In situations where lactation speciﬁc culling rates are unavailable, typical lactation
speciﬁc culling rate patterns for various total culling rates are provided. The user can

choose lactation speciﬁc culling rate patterns based upon breed, herd size, and production

120

 

 

x025 8mm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

moan 3.3 2.3 #52. Sums 31$ 3.3 ooﬁm omcv vodm wE=:U :88.

230 5.8:

no.3 3.00 3.3 mmdm 3.3. 3.9m omwv modm mva 3.2 £30.52

mw.~_ no.2 wmd— 26— mod mmd 3.x No.0 mmh oo.m ﬁnen—

No._ sad 2; am; am; om; 1w.— EL 3.. mod 382D

bid— can

Sam 3.; chum NNDN mm.mm 8.3 m2: 3.: 3.: $8 $25qu

3.: 92: mm: 2&2 3.2 3:: 3.2 29: no.9 owe coco—693%

00m

8.9 8.2 3.2 E .2 3.2 8.3 om.~_ 32: wow M: .m can 535

COGUSUOHQ

Saw wms 3:. Que 390 am Sam xch New NM: .504

889:;

ban

9..— oc; 3.. we; 21,— 3... 9.: mm.~ moN ooh .8 Bow

8mm

33 38 2.; 8,: e52 3% was 8.? 2.9. 3.2 95.5 5:

82mm

mezzo

0:6on

2 a w a a m a m N _ 8:823

5:823 5:823 5:563 5:825 5:325 5:523 8:803 :05823 8:88; 8:823 €2.50
29E «:9: v.8; 95.30 95925 55325 Eat—.0 .N. shaman

121

level. In this instance, the manager and advisor chose the “typical” lactation speciﬁc
culling rate pattern for a Holstein herd with a 40 to 42 percent average culling rate.

In the “Desired Lactation Specific Culling Rates Input Field " (Figure 8), the
manager and advisor input the speciﬁc lactation culling rate pattern they expect to
experience with the new programs the veterinarian has developed. Once again, they can
input their own values, or use one of several “typical” patterns offered by the DSS. They
have chosen a lactation speciﬁc culling rate pattern based upon a culling rate pattern that
approximates a Holstein herd with an average total culling rate of 31 to 33 percent.

In Figures 9 and 10, the “Within Lactation Monthly Removal Schedule for Non-
Mortality Health C ulls Input Field ” and “Within Lactation Monthly Removal Schedule
for Mortalities Input Field " are shown respectively. In these ﬁelds, user inputs the
percentage of cull cattle that leave the herd each lactation month. For instance, Figure 9
shows that 5.35 percent of all cull cattle are removed from the herd in the third lactation
month. Figure 10 shows that that 4.73 percent of the ﬁfth lactation cattle mortalities occur
in the tenth lactation month. The user may use the average DHIA removal schedules that
are displayed next to each of the input ﬁelds but are not shown here. In this example, the

user chose to use the average DHIA values.

122

 

 

x85 2%

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.3 Snow 32% when 8.; mm: 3.2. vowm 8.9 3.2 wa:—=0 35,—.

2:6 5:3:

8.3 8.3.. 2.? an?“ 3.3 21mm 2.3 8.3 3.3 3.2 580.52

mm.w ems Nae 2.0 2.6 mmc 3% Se mfm co.~ 5000

$6 who mud 35 3.0 3.2 no.2 2 .2 22 mod 03005

b??— 05

8.3 8.3 52: w:— mm.2 3.2 -.N_ 3.0 :8 3% 08:05.2

”0: 2:2 an. 2 3.2 Nod— aod 9mm 3.5 mms E .v 55050501

00m

m2: ONE >2: «2: m2: two am.” 36 00m 2; v5 .025

:OZUSUOHQ

mm.” wms 3; Que amo Sh vww mew New Nam 26‘—

0089:;

EEO

m: 00.. :0 0.3 a: a: 2.: m2 SN 3m .8 Bow

and

was 2.8 mm: 3.: ON. _ m on: 35. via 32 32 05:5 .38

. was. $55

0E00am

o. o w a 0 m a. m N _ 8:503

5:803 5:803 5:803 5:803 5:803 5:803 5:503 5:803 5:803 5:803 3:000
Earn :3:— 085— w:_==U 050on 5250.3 .0050: .a 0.5»:—

123

 

 

8E

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 .08 0:
€050

00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 be 8
80.88 00.08 00.8 80.08 0: .8 :08 00.00 :00 88.00 00.8 :
00.0 0:.8 00.8 08.8 00.8 00.8 8:.0 00.0 00.0 00.8 0:
00.0 00.0 0. .8 00.8 8.8 0. .8 00.8 00.8 00.8 00.0 0
00.8 00.0 8.8 00.0 00.0 08.0 00.0 08.0 08.0 00.0 0
00.0 00.8 00.0 88.0 00.0 80.0 00.0 08.0 00.0 88.0 8
00.0 00.0 80.0 8.0 8.0 0: .0 00.0 00.0 00.0 00.0 0
00.8 00.0 00.0 0: .0 0:0 00.0 00.0 80.0 8.0 08.0 0
00.0 00.0 80.0 :0 00.0 00.0 000 8.0.0 00.0 0:0 0
80.0 00.0 00.0 00.0 00.0 00.0 00.0 0_ .0 00.0 0_ .0 0
00.0 8.0 00.0 08.0 00.0 00.0 00.0 88.0 80.0 00.8 8
80.0: 08.: 00.: 3.: 00.: 00.: 00.: 00.0: 00.8 00.0 _
0: 0 0 8 0 0 0 0 .8 _ 5:02
comumaowx— :Oﬂwuowx— €033.03 comumuomm— COSSOMA GOCSUGA comumaomd comumuowx— comumuomu— :Oﬂmuomx— ceagom‘u

 

 

 

 

 

203 2.8.:— 0==U 5.00m 3:382:02 ..8 0_=_008_0m 3653— .2552 553004 55.3

.a 0.5»3

124

 

 

€25

 

 

 

 

 

 

 

 

 

 

 

 

 

 

cod 8d 8.0 cod 8d cod cod 8o cod 8.0 be 2
625

cod 8.0 cod 8c 8d 8d 8d cod cod 8.0 38 2
M5: at: $2 2.2 2.2 2.2 3.8 5.8 52 Sam :
V? a: 3v 3e 03 m: S .m :3 o: «S 2
«3 mam 0:. can 8a 9% .2 £3 in Se a
wow 3m 3; RN 3: SN am an 8e :e m
v2 02 own :A SN 33 :3. can m: w: 5
8a 3m 2 .m 3m 8a SN 8.,“ can m3. 8+ 0
w: 2: v3 3m 2%. N3 Ra mam m3 .3 m
a? N? 9% 3m 8+ 85 2 .v o: 3; Se v
2: gm 2; Sm of. 8m 8w 3 .m 2% :o m
who ts So one «.3 of 3: NEW o2. 8.0 N
on? 8.? $9 8.8 8.? 3w? 8.: $3 3.3 :8 _
2 a w h o m w m N _ £82
Comuﬂuomd COSMHOMJ GOZNHUQA :OCNHUNJ Comumuowxu GOSMHUMJ Ccmumeowx— Comwmuomd COUNHUMJ COmumaomJ comwmuoml—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Eur— ..55 85.332 3.. o_:co:um .535: £5.82 55525 5559

.2 3:5

125

Figure 11. Other Production and Financial Information Input Field

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Other Production and Financial Information Value
Interest Rate (%) 7.5
Debt to Asset Ratio (%) 50
Cost of Equity Capital (%) 10
Equity to Asset Ratio (%) 50
Marginal Tax Rate (%) 0
Capital Gains Tax Rate (%) 0
After Tax Weighted Average Cost of Capital (%) 8.75
Replacement Heifer Price (S/covﬂ 1350
Typical Cull Cow Price j$/cow) 350
Expected Milk Price Received Less SCC Premiums/Discount (S/cwt) 13.5
Expected Heifer Calf Price (S/calf) 200
Heifer Calf Mortality (%) 6
Expected Bull Calf Price QS/calf) 75
Bull Calf Mortality (%) 6
Eigected Feed Cost LS/cow) 982
Expected Labor Cost ($/cow) 646
Expected Other Variable Costs ($/cow) 853
Expected Cost per Dry Cow Month ($/cow) 45
Typical Treatment Cost per Udder and Mastitis Health Problem 133
Episode (S/treated cow culled)

Typical Treatment Cost per Reproductive Health Problem Episode 192
{S/treated cow culledL

Typical Treatment Cost per Lameness Episode (S/treated cow culled) 9
Typical Treatment Cost per Disease Episode LS/treated cow culled) 112
Somatic Cell Premium/Discount (SS/owl) 0.00063
Milk Production Genetic Improvement Rate ("/0 per generation) 1.0
Feed Expense Adjustment Factor Required to Support a 1% Milk 0.9
Revenue Increase Associated with Genetic Improvement

Labor Expense Annual Increase §°/o per year) 0.0
Other Variable Expense Annual Increase (% per year) 0.0

 

 

L

126

 

 

 

Figure 1 1 displays the “Other Production and Financial Information Input
Field. ”This input ﬁeld concerns the other pertinent production and ﬁnancial
information needed to run the DSS. Two of the input ﬁeld items, the Equity to Asset
Ratio and the Aﬁer Tax Weighted Average Cost of Capital are calculated
automatically. The Equity to Asset Ratio is calculated by subtracting the Debt to
Asset Ratio from 1. The After Tax Weighted Average Cost of Capital (WACC) is
calculated according to the following formula:

WACC = { [ (Debt to Asset Ratio)*(Interest Rate/ 100)

+ (Equity to Asset Ratio)*(Cost of Equity

Capital/100) ] * (1- (Marginal Tax Rate / 100) }
In this example, the manager and her advisor have decided to analyze their culling
reduction problem on a before tax basis. As such, they have inputted a “0” into the
Marginal Tax Rate and Capital Gains Tax Rate input rows.

The purpose of the “Lactation Expense Adjustment Input Field " of Figure 12
is to adjust the production costs entered in the “Other Production and Financial
Information Input Field ” for cow age. An “A verage Lactation Expense ” table (also
Figure 12) is displayed by the DSS to provide the user with “typical” lactation
expense adjustments. The values for the average expense adjustments were developed
from work done by Bauer, Mumey, and Lohr (1993). The values listed in the Average
Expense Adjustment Table assume that the costs entered in the Other Production and

Financial Information Input Field are representative of a second lactation cow.

127

Figure 12.

Lactation Expense Adjustment Input Field

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Lactation Expense Adjustment Value
Lactation 1 0.963
Lactation 2 1.000
Lactation 3 1.022
Lactation 4 1.057
Lactation 5 1.075
Lactation 6 1.112
Lactation 7 1.114
Lactation 8 1.116
Lactation 9 1.119
Lactation 10 1.125
Average Lactation Expense Adjustment Value
Lactation 1 0.963
Lactation 2 1.000
Lactation 3 1.022
Lactation 4 1.057
Lactation 5 1.075
Lactation 6 1.112
Lactation 7 1.114
Lactation 8 1.116
Lactation 9 1.119
Lactation 10 1.125

 

 

 

 

 

Figure 13 displays the “Health Cull and Mortality Reduction Technology
Investment Input Field " To indicate that a farm will incur additional annual
operating expenses or whether an investment in a three-, ﬁve-, seven-, or ten year
asset is required, the user would enter a “1” in the appropriate cell(s) of the “New
Technology Investment Type " row. If an increase in annual operating expense or an
investment in a speciﬁc asset is not required, the user enters a “0” in the appropriate
cell(s). In the “C 0st of Purchasing, Installation, and Implementation ” row, the user

enters the increase in annual operating expenses and the investment needed in capital

128

 

assets to achieve the health cull reduction. In this example, the user has entered that
the plan to reduce health culls would cost an addition $10,000 per year and that no
additional investments will be needed.

Figure 13. Health Cull and Mortality Reduction Technology Investment

 

 

 

 

 

 

Input Field
Additional Three Five Seven Ten
Annual Year Year Year Year
Operating Asset Asset Asset Asset
Expense
New Technology Investment 1 0 0 0 0
Type
Yes = 1; No = 0)
Cost of Purchasing, Installation, $10000 0 0 0 0
and Implementation
Operating Expense Growth 0
Rate (%)

 

 

 

 

 

 

 

 

 

VI. DSS Calculations and Output Fields
This section explains the calculations used by the DSS to determine the
ﬁnancial feasibility of a farm reducing culling rates as well as the output ﬁelds the

DSS generates. The section will proceed by describing the calculations and output

ﬁelds for the :
I) herd inventory dynamics for the current culling rate and culling
pattern;
2) ﬁnancial returns of the current culling rate and culling pattern;
3) herd inventory dynamics for the desired culling rate and culling
pattern;
4) ﬁnancial returns of the desired culling rate and culling pattern;

129

 

5) cash ﬂows of the proposed health cull reduction technologies
investments and operating costs; and,

6) net present value of the proposed culling rate reduction.
Calculations and Output Fields for the Current Culling Rate

Figure 14 displays the number of cattle of a particular lactation and lactation
month that are removed from the herd and retained in the herd during each month of a
240 month planning horizon period. One can see that during month 1 of the 240
month planning horizon that the dairy farm manager will have ten ﬁrst lactation cows
in lactation month 1. During this month, a total of 0.429 of the case study farm’s ten

lactation 1 month 1 cows are culled for the following reasons:

Sold for Dairy Purposes: 0.155 cows;

Production Culls: 0.035 cows;
Mortalities: 0.071 cows;
Udder and SCC Culls: 0.043 cows;
Reproduction Culls: 0.053 cows;
Lameness and Injury Culls: 0.064 cows; and,
Disease Culls: 0.008 cows.

These numbers were calculated by applying the following formula:

(Number of Lactation n Month t Cows in Planning Horizon Month n)

(Current Lactation n Culling Rate for a Speciﬁc Culling Reason)

(Within Lactation Monthly Removal Percentage for Lactation n Cattle)
Number of Lactation 1 Month 1 Cows Removed for a Speciﬁc Reason

**

All “Current Lactation n Culling Rate for a Speciﬁc Culling Reason ” values are
found in Figure 7, the "Current Lactation Speciﬁc Culling Rate Input Field The
“Within Lactation Monthly Removal Percentage for Lactation n Cattle ” values are

determined by various methods. It was assumed that cattle sold for dairy purposes are

130

Figure 14. Current Herd Inventory Dynamics Output Field

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Planning Horizon Month: 1 2 -> 240 Dispersal Sale
Lactation 1 Month 1 Cows 10.000 2.994 -v 4.538

Sold for Dairy Purposes 0.155 0.046 —+ 0.070

Production Culls 0.035 0.010 —o 0.016

Mortalities 0.071 0.021 —» 0.013

Udder and SCC Culls 0.043 0.013 —+ 0.020

Reproduction Culls 0.053 0.016 —+ 0.024

Lameness and ijury Culls 0.064 0.019 —§ 0.029

Disease Culls 0.008 0.002 -+ 0.004

Cattle Transferred 9.571 0.050 —v 4.363 4.363
Lactation 1 Month 2 Cows 10.000 9.571 -> 4.344

Sold for Dairy Purposes 0.000 0.000 —» 0.000

Production Culls 0.035 0.033 -+ 0.015

Mortalities 0.028 0.027 —» 0.012

Udder and SCC Culls 0.039 0.038 —» 0.017

Reproduction Culls 0.048 0.045 —> 0.021

Lameness and Injury Culls 0.058 0.055 —-> 0.025

Disease Culls 0.007 0.007 —+ 0.003

Cattle Transferred 9.786 9.367 —. 4.251 4.251

1 l l l 1

Lactation 10 Month 11 Cows 0.000 0.000 —+ 0.005

Sold for Dairy Purposes 0.000 0000 —> 0.000

Production Culls 0.000 0.000 —+ 0.004

Mortalities 0.000 0.000 —> 0.000

Udder and SCC Culls 0.000 0.000 —-> 0.000

Reproduction Culls 0.000 0.000 -+ 0.000

Lameness and Injury Culls 0.000 0.000 —» 0.001

Disease Culls 0.000 0.000 —o 0.000

Cattle Transferred 0.000 0.000 —+ 0.000 0.000
Monthly Totals

Total Beginning Animals 130.00 130.00 —+ 130.00

Total Sold for Dairy 0.31 0.34 —+ 0.29

Total Production Culls 0.38 0.40 -» 0.51

Total Mortalities 0.31 0.40 —-§ 0.56

Total Udder and SCC Culls 0.52 0.54 —-» 0.83

Total Reproduction Culls 0.63 0.66 —+ 0.98

Total Lameness and Injury Culls 0.76 0.79 —+ 1.22

Total Disease Culls 0.09 0.10 —+ 0.14

Total Replacements Required 2.99 3.24 —’ 4.51

Estimate Annual Culling Rate 27.64 28.79 -—i 41.87

 

 

131

 

 

sold either in lactation month 1 or at the end of lactation month 11 in all lactations
except for Lactation 10. As such, the “Within Lactation Monthly Removal
Percentage for Lactation n Cattle " value for lactation months 1 and 11 is 50 percent.
No cattle are sold for dairy purposes during lactation months 2 through 10. In
Lactation 10, all of the sold for dairy purposes culls are removed in the ﬁrst lactation
month. Thus, the “Within Lactation Monthly Removal Percentage for Lactation 1 I
Cattle ” is 100 percent. To calculate the number of lactation 1 cattle that are culled
for dairy purposes in lactation month 1 and planning horizon month 1, the following
calculation is made:

(10 Lactation 1 Month 1 Cows in Planning Horizon Month 1)

* (3.09 Percent Removed for Sold for Dairy Purposes)
"‘ (50 Percent Removed in Lactation Month 1)

 

0.155 Lactation 1 Month 1 Cattle Removed for Sold for Dairy
Purposes in Planning Horizon Month 1.

For production culls, the DSS assumes that there is an equal probability that a
production cull can occur at any time during the eleven lactation months. Thus, the
“ Within Lactation Monthly Removal Percentage for Lactation n Cattle ” for all cattle
is 9.09 percent. For the lactation 1 month 1 cattle in the ﬁrst planning horizon month
of the case study situation there are 0.035 Lactation 1 Monthl cattle culled for
production reasons. This value was calculated in the following manner:

(10 Lactation 1 Month 1 Cows in Planning Horizon Month 1)
* (3.82 Percent Culled for Production Reasons)
* - (9.09 Percent Removed in LaLtation Month 1)

0.035 Lactation 1 Month 1 Cattle Culled for Production Reasons
in Planning Horizon Month 1.

132

For cattle that are culled due to non-mortality health culls, the “Within

3,

Lactation Monthly Removal Percentage for Lactation n Cattle can be found in
Figure 9, the “Within Lactation Monthly Removal Schedule for Non—Mortality Health
Culls Input Field ” Thus, the number of Lactation 1 Month 1 cattle culled for
reproduction is calculated by:
(10 Lactation 1 Month 1 Cows in Planning Horizon Month 1)
(6.26 percent culled for reproduction reasons)
(8.39 percent removed in Lactation Month 1)
0.053 Lactation 1 Month 1 Cattle Culled for Reproduction Reasons
in Planning Horizon Month 1.

To determine the number of Lactation n Month t cattle that die during a
particular planning horizon month, the percentage of lactation n cattle that are
removed due to mortalities can be found in Figure 7, the “Current Lactation Specific
Culling Rate Input Field The appropriate “Within Lactation Monthly Removal
Percentage for Lactation n Cattle ” value can be found in Figure 10, the “Within
Lactation Monthly Removal Schedule for Mortalities Input Field ” For Lactation 1
Month 1 Cows, the number of mortalities in Planning Horizon Month 1 is calculated

as follows: -

(10 Lactation 1 Month 1 Cows in Planning Horizon Month 1)

* (3.09 percent Mortality Rate)

* (23.11 percent removed in Lactation Month 1)

0.071 Lactation 1 Month 1 Cattle Mortalities in Planning Horizon Month 1
This process is repeated for 240 Planning Horizon Months. At the end of the

240un Planning Horizon Month, the surviving cattle are sold for dairy purposes in a

dispersal sale.

133

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 15. Current Monthly Cash Flows Output Field

Month: 1 -> 240
Sold for Dairy Purposes Cull Proceeds $403 -+ $149,845
Production Cull Proceeds $134 -> $177
Non-mortality and Non-lame and Injured Health $433 -+ $666
Cull Proceeds
Lame and Injured and Aged Cow Health Cull $289 -+ $442
Proceeds
Calf Credit Proceeds $1,210 —. $1,485
Base Milk Returns for Lactation 1 Cattle $3,984 —-> $3,224

1 L -r 1
Base Milk Returns for Lactation 10 Cattle $0 —» $1
Lactation 1 Sold for Dairy Cull Return Ad'ﬂstments -$8 —r -$3

1 l -* 1
Lactation 10 Sold for Dairy Cull Return $0 -—» $0
Adjustments
Lactation 1 Production Cull Return Adjustments -$29 -» -$9

1 l -r 1
Lactation 10 Production Cull Return Adjustments $0 -> -$1
Lactation 1 Non-mortality Health Cull Return -$25 —+ -$7
Adjustments

1 l -r 1
Lactation 10 Non-Mortality Health Cull Return $0 —» $0
Adjustments
Lactation 1 Mortaligl Return Adjustments -$14 —» -3

l l -’ 1
Lactation 10 Mortality Return Adjustments $0 —+ $0
Health Cull Treatment Costs -$241 ->- -$421
Net Monthly Returns $6,138 —+ $162,043
Present Value of Monthly Returns $6,093 —+ $28,338
Net Present Value of Total Returns $1,081,513

 

 

 

 

 

The DSS then determines the present values of the revenues and expenses

associated with the current culling rate and culling pattern (Figure 15). The current

culling rate and culling pattern has a net present value of total returns of $1,238,817.

To calculate this value, the DSS ﬁrst determined the proceeds for the mature cattle

134

 

 

that were sold for dairy purposes. The value for a sold for dairy purposes cow was
calculated using the following protocols:

1) If the cow sold for dairy purposes was sold as a lactation 1 month 1
cow, the cow was valued at the replacement heifer price;

2) If the cow was sold for dairy purposes in lactation 10 month 1, the cow
was valued at the cull cow price plus $100;

3) No tenth lactation cattle beyond lactation month 1 can be sold for
dairy purposes by the DSS; and,

4) All other sold for dairy purpose values were determined using the
following formula:

Value 1mm" n = Value mm, a -1 - [(Replacement Heifer
Price — Cull Cow Price + $100)] / 10; and,

5) If the cow was sold in lactation month 11 of any lactation, it was
assigned the sold for dairy proceed value of the subsequent lactation.

In the case study, replacement dairy heifers were valued at $1,350 per cow and the
cull cow value was $350. As such, the Sold for Dairy Purposes Proceeds per cow for
each lactation were calculated as they appear in Figure 16.

Figure 16. The Sold for Dairy Proceeds Value Output Field

 

 

Lactation Number Sold for Dairy Proceeds Value ($/cow)
1 $1,350
$1,260
$1,170
$1,080
$990
$900
$810
$720
$630
$450

 

 

 

 

 

 

 

 

 

\OWNOMAUJN

 

 

 

 

O

 

 

 

 

135

The production and non-mortality health cull proceed values were calculated

by the DSS according to the following protocols:

1) Cattle that had milked less than four lactations and were not culled for
being lame or injured were assigned the ﬁlll “ijical Cull Cow Price ”
value (Figure 11);

2) Cattle that had milked more than three lactations but less than seven
lactations and were not culled due to lameness and injuries were
assigned the typical cull cow price value minus four percent; and,

3) Cattle that had milked more than six lactations or cattle that were
culled due to lameness or injury were assigned the typical cull cow
price value minus seventeen percent.

These protocols were adapted from the work of Mumey, Bauer, and Lohr (1993). For
the case study scenario, the user inputted that the typical cull cow price was $350 per

head. Thus, cull cow values were allocated according to Figure 17.

Figure 17. Cull Cow Proceeds Value Output Field

 

 

 

 

 

 

 

 

 

Cull Cow Type Cull Cow Proceed Value ($/cow)
Less Than Four Lactations Old $3 50
More Than Three Lactations Old But $336
Less Than Seven Lactations Old
More Than Six Lactations Old $291
Lame or Injured Cattle $291

 

 

 

 

To determine the Expected Calf Credit, the following formula was used:

Expected Calf Credit = [( 1- (Bull Calf Mortality/100))*(Expected Bull Calf
Price)] + [(1 — (Heifer Calf Mortality/100))*(Expected
Heifer Calf Price)].

In the case study example, the user speciﬁed an expected bull calf price of $75, a bull

calf mortality rate of 6 percent, a heifer calf price of $200, and an a heifer half

136

mortality of 6 percent. This resulted in an Expected Calf Credit of $121 per cow per
lactation.

“Base Milk Returns" (BMR) were calculated for each cow at the beginning of
each planning horizon month according to the following formula:

BMR = Milk Revenues(Cows) + SCC Premium(Cows) — Feed Cost -
Labor Costs — Other Variable Costs.

The monthly BMR were determined by multiplying the “Expected Production by
Lactation ” (Figure 3) value by the “Expected Milk Price Received Less SC C
Premiums/Discount " (Figure 1 1). In order to do this, the Expected Production by
Lactation Value had to be divided by eleven months and adjusted by the genetic

improvement rate according to the following:

Monthly Production = (Expected Production by Lactation/11)*
(1+MPGIR/100)g,

where “MPGIR " refers to the Milk Production Genetic Improvement Rate (Figure
11) and "g” refers to the cows generation number.5 SCC premiums were calculated
using the following formula:
SCC Premium = ((350-(SCC/ 1000))*(Monthly Production/100)*( Somatic
Cell Premium/Discount).
Feed costs were determined using the feed cost information that the manager
and her advisor inputted into Figure 11, the Other Production and Financial

Information Input Field. As milk production improved through the Milk Production

 

5 For the cattle that are in the herd or enter the herd the ﬁrst 13 planning horizon months, g always
equals 0 throughout their life in the herd. For ﬁrst lactation animals that enter the herd during planning
horizon months 14 — 26, g always equals lthroughout their life in the herd. For ﬁrst lactation animals
that enter the herd during planning horizon months 27 -— 39, g always equals 2 throughout their life in
the herd. This process is repeated throughout the planning horizon months with the fust lactation
animals receiving a one unit g increase every fourteen months.

137

Genetic Improvement Rate, the feed costs were adjusted upward by the “Feed
Expense Adjustment Factor ” (F EAF ; Figure 10) and the “Lactation Expense
Adjustment Factor ” (LEAF ; Figure 11). Thus, feed costs were calculated using the
following formula:

Feed Costs = (Expected Feed Cost)*[(1+[(1\/IPGIR/100)*(FEAF))5] * (LEAF).

Labor and Other Variable Expenses were assigned to each cow based upon the
information supplied by the manager and advisor in Figure 11. These values can be
adjusted annually to reﬂect the manager’s expected “Labor Expense Annual
Increase, " the “Other Variable Expense Annual Increase ” and “LEAF. ”

Once the Base Milk Returns are calculated, the DSS then charges a production
loss charge for each type of cull that occurs in the planning horizon month. These are
called Return Adjustments in Figure 16. The charges are assigned to the cull cows
based upon the information provided by the manager in Figure 4, the “Culling
Associated Production Eﬂects Input Field ’, and Figure 6, the “Culling Associated
SC C Eﬂect Input Fiel . ” A charge is added to the current planning horizon month if a
cow in the current month is removed anytime during its eleven month lactation
period. For example, if a cow began milking as a lactation 1 month 1 cow and is
removed in the 3rd planning horizon month, this charge is assigned to lactation
months 1, 2 and 3.

Another charge was added to reflect the increased veterinary charges that may
occur with cattle prior to them being culled for health purposes, the Health Cull
Treatment Cost. This charge was equal to the number of cattle culled for a particular

reason times the appropriate “Typical Treatment Cost per __ Episode ” that the dairy

138

farm manager and her advisor listed in Figure 11. The DSS assigns the “Typical
Treatment Cost per Disease Episode ” charge for cattle that die during the planning
horizon month. The values for the treatment costs shown in Figure 11 were adapted
from the work of Weigler et al. (1990).

The Net Monthly Returns were calculated by adding up the proceeds, returns,
return adjustments, and health cull treatment costs. The Net Monthly Returns for each
planning horizon month was discounted by the weighted average cost of capital to
determine the Present Value of Monthly returns. The formula for accomplishing this
is as follows:

Present Value of Monthly Returnsn = Net Monthly Returns,/ (l+(k/100))".
Next, all of the Present Value of Monthly Returns are added up to determine the Net

Present Value of Total Returns for the current culling rate and culling pattern.

Calculations and Output Fields for the Desired Culling Rate

Figures 18 and 19 display the output for the Desired Culling Rate and Culling
Pattern. The only difference between these calculations and those for the Current
Culling Rate and Culling pattern are the lower lactation speciﬁc culling rates and the

additional surplus replacement heifer sales.

139

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 18. Desired Herd Inventory Dynamics Output Field
PlanningHorizon Month: 1 2 —+ 240 Dispersal Sale
Lactation 1 Cows Month 1 10.000 2.226 —> 4.538
Sold for Dairy Purposes 0.155 0.0344 -—» 0.070
Production Culls 0.035 0.008 —> 0.016
Mortalities 0.048 0.011 -—> 0.013
Udder and SCC Culls 0.029 0.006 —» 0.020
lgproduction Culls 0.035 0.008 —+ 0.024
Lameness and Injury Culls 0.042 0.009 —> 0.029
Disease Culls 0.005 0.001 —+ 0.004
Cattle Transferred 9.652 2.149 —-> 4.363 4.363
Lactation 1 Cows Month 2 10.000 9.652 -v 4.344
Sold for Dairy Purposes 0.000 0.000 —> 0.000
Production Culls 0.035 0.034 —. 0.015
Mortalities 0.028 0.018 -> 0.012
Udder and SCC Culls 0.039 0.025 —» 0.017
Reproduction Culls 0,048 0.031 —. 0.021
Lameness and Injury Culls 0.058 0.037 -> 0.025
Disease Culls 0.007 0.005 -. 0.003
Cattle Transferred 9.786 9.503 —» 4.251 4.251

.1 l l l 1
Lactation 10 Cows Month 11 0.000 0.000 —. 0.030
Sold for Dairy Purposes 0.000 0.000 —+ 0.000
Production Culls 0.000 0.000 -> 0.026
Mortalities 0.000 0.000 —. 0.001
Udder and SCC Culls 0.000 0.000 —. 0.001
Reproduction Culls 0.000 0.000 -> 0.001
Lameness and Injury Culls 0.000 0.000 -> 0.001
Disease Culls 0.000 0.000 -v 0.000
Cattle Transferred 0.000 0.000 —. 0.000 0.000
Monthly Totals
Total Beginning Animals 130.00 130.00 -+ 130.00
Total Sold for Dairy 0.31 0.33 —~ 0.27
Total Production Culls 0.38 0.40 —o 0.82
Total Mortalities 0.21 0.26 .... 0.41
Total Udder and SCC Culls 0.35 0.36 —o 0.62
Total Reproduction Culls 0.42 0.44 —» 0.71
Total Lameness and Injury Culls 0.51 0.53 —-> 0.92
Total Disease Culls 0.06 0.07 —. 0.10
Total Replacements Required 2.23 2.39 —. 3.84
Estimate Annual Culling Rate 20.55 21.32 —» 33.13

 

140

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 19. Desired Monthly Cash Flows Output Field
Month: 1 -—r 240
Sold for Dairy Pupposes Cull Proceeds $403 —i $145,506
Production Cull Proceeds $134 —» $280
Non-mortality and Non-lame and Injured Health $288 —-> $461
Cull Proceeds
Lame and Injured and Aged Cow Health Cull $193 -—» $359
Proceeds
Calf Credit Proceeds $1,210 -+ $1,426
Milk Returns for Lactation 1 Cattle $3,984 —» $2,632
A J, -r 1

Milk Returns for Lactation 10 Cattle $0 —+ -$8
Lactation 1 Sold for Dairy Cull Return Adjustments -$8 —> -$2

1 l -r 1
Lactation 10 Sold for Dairy Cull Return $0 -> $0
Adjustments
Lactation 1 Production Cull Return Adjustments -$29 —-> -$7

1 l -’ 1
Lactation 10 Production Cull Return Adjustments $0 —» -$6
Lactation 1 Non-mortality Health Cull Return -$17 -+ -$4
Adjustments

l l -* 1
Lactation 10 Non-Mortality Health Cull Return $0 —? $0
Adjustments
Lactation 1 Mortality Return Adjustments -$9 —+ -$2

1 l -r 1
Lactation 10 Mortality Return Adjustments $0 —+ $0
Health Cull Treatment Costs -$161 —. —$3 12
Returns from Additional Surplus Heifer Sales $1,036 —+ $902
Net Monthly Returns $7026 —+ $158,414
Present Value of Monthly Returns $6,975 —. $27,703
Net Present Value of Total Returns $1,204,012

 

 

 

 

 

Calculations and Output Fields for the Investment Cash Flows

Figure 20 displays the output ﬁeld that shows the cash ﬂows associated with

the health reduction technology or program investment. The ﬁgure indicates that the

farm will incur $833 in monthly operational expenses to implement the health

141

 

 

reduction program. The $833 was calculated by dividing the $10,000 “Cost of
Purchasing, Installation, and Implementation ” value listed in Figure 13 by twelve
months. This value was inﬂated each year throughout the 240 planning horizon
months by the "Operating Expense Growth Rate ” also listed in Figure 13.

Figure 20. The Cash Flows Associated with the Health Cull and Mortality
Reduction Technology and Programs Output Field

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Health Cull and Mortality Reduction Initial Month —r Month
Investments and Expenses Investment 1 240
Technology Purchase and Installation 0 0 —> 0
After Tax Operating Expenses 0 $833 -> $833
Depreciation Shield 0 0 —+ 0
Terminal Value 0 0 —r 0
Monthly Health Cull and Mortality 0 -$833 —+ -$833
Reduction Technology and Program

Cash Flows

Present Value of Monthly Health Cull 0 -$827 —r -$146
and Mortality Reduction Technology

and Program Cash Flows

Total Present Value of Health Cull and -$94,299

Mortality Reduction Technology and

Program Cash Flows

 

 

 

 

The technology purchase and installation value, the after tax operating
expenses, the depreciation shield and terminal value were added up in each month to
calculate the “Monthly Health Cull and Mortality Reduction Technology and
Program Cash Flows. ” These values were discounted by the weighted average cost
of capital according to determine the “Present Value of Monthly Health Cull and
Mortality Reduction Technology and Program Cash Flows. " By adding all 240
monthly values up, the “Total Present Value of Health Cull and Mortality Reduction

Technology and Program Cash F lows ” was determined.

142

 

The Calculations and Output Fields For the Net Present Value of the Desired
Culling Rate Pattern

The NPV Reduction for the example is $28,200 (Figure 21). As $28,200 is
greater than zero, the management program should be adopted because, after
discounting all cash ﬂows to today’s values, there are $28,200 after accounting for
the additional revenues and expenses of the program as well as the opportunity cost of
the dairy farm manager’s and lender’s capital. By converting the NPV to a monthly
value, the DSS indicates that the farm manager can expect to return an additional
$249 per month after paying off the variable expenses and her and her lender’s
opportunity cost of capital.

Figure 21. The Financial Implications of the Health Cull and Mortality
Reduction Technology and Program Output Field

 

 

 

 

 

 

 

 

 

Financial Implications of the Health Cull and Mortality

Reduction Technology and Program

Present Value of Returns of the Desired Culling Rate Pattern $1,204,012
Present Value of Returns of the Current Culling Rate Pattern $1,081,513
Present Value of the Health Cull and Mortality Reduction -$94,299
Technology and Program Cash Flows

Net Present Value of Reducing the Desired Culling Rate $28,200
Pattern

Monthly Equivalent Returns $249

 

 

 

 

Figure 22 shows the before and aﬁer effects of the proposed program on total
culling rate. Without the program, the user expected to have an average estimated
culling rate of 41 .49 percent. With the proposed program, the user expects to

experience an average annual culling rate of 32.40 percent.

143

 

 

 

 

 

 

Figure 22. The Estimated Annual Culling Rates with and without Adopting
the Health Cull and Mortality Reduction Technology and
Program Output Field
Culling Rate Type Low Average High
Estimated Estimated Estimated
Annual Annual Annual
Culling Culling Culling Rate
Rate Rate (%) (%)
(%)
Culling Rate with the New Health 24.20 32.40 33.46
Cull and Mortality Reduction
Technology and Program
Culling Rate without the New 32.60 41.49 42.00

 

Health Cull and Mortality Reduction
Technology and Program

 

 

 

 

 

This chapter described how to use a prototype DSS to determine whether it is

ﬁnancially feasible to reduce culling rates through the adoption of technological

investments or management programs. In Chapter VI, this DSS is applied to

determine the potential returns for dairy farms that differ in current culling rate, size,

and breed.

144

 

 

CHAPTER 6.

THE FINANCIAL FEASIBILITY OF DECREASING CULLING RATES ON
DAIRY FARMS

I. Introduction

A DSS was described in the previous chapter. This DSS enables a manager or
advisor to determine whether it is economical to reduce health culling rates. In this
chapter, the DSS is used to estimate the maximum amount that managers of dairy farms
with various characteristics would be willing to pay per month for a ten percent reduction
in their lactation speciﬁc health culling rate and for ten percent reductions in speciﬁc
types of health culls. This monthly amount will be called the breakeven annuity (BEA).
The characteristics include the initial lactation speciﬁc health culling rates, cattle breed,
and herd size of the farms. The DSS will also be used to determine if an investment in
two health cull reduction technologies, rubberized floors for cattle walkways to reduce
lameness and injury culls and using gonadotropin releasing hormone (GNRH) to reduce
reproduction culls, is proﬁtable.

Determining how much a farmer can pay for a health cull reduction is important.
If a producer under invests in health culling rate technologies or programs, the costs
associated with health culls — production losses, treatment expenses, and the cash ﬂows
associated with the replacement of unhealthy cattle — become excessive. On the other
hand, if a producer over invests in health cull reduction technology, the resulting lower
culling rate will be less proﬁtable than the current higher culling rate. Furthermore,
understanding how the BEA varies based upon farm characteristics is important to
understand as well. If the marginal BEA is somewhat constant, producers will have little

need to run the DSS after each successive health cull reduction. If not, the DSS will need

145

to be run after each successive reduction. If cattle breed and or herd size affects the BEA,
advisors will need to make separate BEA estimates for farms that vary by breed and size.
H. General Methods and BEA Estimate Categories

To show how different dairy farm characteristics affect the BEA for a ten percent
health cull reduction, a series of estimates were made using the DSS described in Chapter
5. To estimate the BEA, the costs in the Health Cull and Mortality Reduction
Technology Investment Input Field were set equal to zero. The resulting equivalent value
for the ten percent health culling rate reduction would then represent the maximum
amount a farm manager would be willing to pay monthly for the ten percent reduction in
health culling rates for a 240 month period.

In the ﬁrst series of category of estimates, the BEA is calculated for ten example
farms. One of the eXample farms, “Sample Average, ” exhibited the average lactation
speciﬁc health culling rates of the DHIA farms in the ten Midwestern and Northeastern
states using the DHIA data ﬁrst shown in Chapter 2. Five of the example farms exhibited
from ten to ﬁfty percent higher lactation speciﬁc culling rates than the DHIA sample
average, and four of the example farms exhibited from ten to forty percent lower health
culling rates than the sample average. The methods used to estimate the BEA and the
estimation results for the typical Midwestern and Northeastern DHIA farm category are
discussed further in Section III.

In Section IV, the BEA is estimated for farms that use three different breeds of
cattle: Holstein; Jersey; and Guernsey cattle. Estimates for Holsteins were conducted
because the breed is the predominant breed in the United States. Estimates for the Jersey

breed were chosen because they exhibit a lower likelihood of being culled (Chapter 3),

146

produce less milk, and yield a higher valued milk than Holstein cattle. Estimates for
Guernsey cattle were made because they exhibit a higher likelihood of being culled
(Chapter 3), produce less milk, and yield a higher valued milk than Holstein cattle.
Estimates were made for ten example farms within each the breed category. One farm,
“Breed Average, ” exhibited the breed average lactation speciﬁc health culling rates. Five
of the example farms exhibited from ten to ﬁfty percent higher lactation speciﬁc health
culling rates than the breed average culling rates, and four of the example farms
exhibited lactation speciﬁc culling rates that were from 10 to 40 percent lower than the
breed average. The methods used to estimate the BEA and the estimation results for the
breed categories are further discussed in Section IV.

The BEA was then estimated for a category of dairy farms based on herd size.
The ﬁrst group of farms consisted of ten farms with a herd size of 130 cows. The second
group of farms consisted of ten, 390 cow dairy farms. The third group of farms in the
herd size estimation category consisted of ten, 650 cow dairy farms. The ten farms in
each category had lactation speciﬁc health culling rates that ranged form 40 percent
lower to 50 percent higher than the average lactation speciﬁc culling rates for herds with
less than 150 cows, farms with herds of 300 to 450 cows, and farms with more than 600
cows. Additional methods used to estimate the BEA and the estimation results for the
herd size estimation category are discussed in Section V.
II]. The BEA Estimates for Health Culling Rate Reductions Among

Midwestern and Northeastern DHIA Dairy Farms

In this section, the BEA for a health culling rate reduction are estimated for ten

farms. The farms have initial lactation speciﬁc health culling rates that range from forty

147

percent below to ﬁfty percent above the average Midwestern and Northeastern dairy

farm. In the estimations, the lactation speciﬁc culling rate for sold for dairy purposes and

low production were held constant at the breed average level.

Production and Financial Parameters

It was assumed that the typical Midwestern and Northeastern dairy herds would
consist of 130 milking and dry cows with a thirteen month calving interval. There were
initially 10 ﬁrst lactation cows in each of the eleven lactation months and 10 cows in two

dry period months. The cattle in the two dry period months were waiting for their second

lactation to begin.

Table 34. The Expected Rolling Herd Average Per Lactation and Somatic Cell

Count Levels Per Lactation Used to Estimate the BEA for a Health

Cull Reduction for the Typical Midwestern and Northeastern DHIA

Farm Estimation Category

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Lactation Milk Production per Somatic Cell Count per

Lactation (lbs) Lactation
1 20,000 200,000
2 22,308 254,840
3 23,000 285,900
4 23,136 308,700
5 22,950 326,460
6 22,482 346,280
7 21,790 356,460
8 21,158 358,480
9 20,378 363,560
10 19,736 362,060

 

The ten farms’ cattle would have a 20,000 pound ﬁrst lactation production level
and a 200,000 somatic cell count (Table 29). Culled cattle were subject to the production

and somatic cell count adjustments shown in Table 30. The adjustments were applied to

the entire productive period of the terminal lactation as described in Chapter 5.

148

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 35. The Effects of Culling on the RHA and Somatic Cell Count (SCC)
Levels of the Culled Cow Used for the BEA Estimation of a Health
Cull Reduction
Lactation Sold for Dairy Production Culls Non—Death Deaths
Culled Purposes Health Culls
RHA SCC RHA SCC RHA SCC RHA SCC
(+/- (+/- (+/- %) (+/- %) (+/- %) (+/- (+/- (+/- %)
%) °/o) %) %)
1 -11 +4 -26 +29 -6 +19 -16 +5
2 - 4 + 2 - 7 + 19 - 1 + 15 - 7 - 1
3 - 3 + 3 - 11 + 15 0 + 13 - 6 - l
4 -3 +6 -11 +11 -1 +13 -6 -5
5 -1 +7 -10 +13 -2 +12 -11 -5
6 - 1 + 6 - 9 + 10 0 + 9 - 7 - 10
7 -1 +7 - 3 + 6 2 + 9 - 5 - 5
8 - 5 + 7 - 8 + 9 0 +11 - 9 - 6
9 - 5 + 3 - 7 + 12 0 + 11 - 11 - 6
10 -6 +5 -11 +11 -6 +8 -14 -8

 

 

 

 

 

 

 

 

 

The lactation speciﬁc culling rates for this estimation category are shown in Table

31. The average annual lactation speciﬁc culling rates ranged from 22.31 percent in the

ﬁrst lactation to 58.57 percent in lactation ten. The primary reason for cattle being culled

was lameness or injury. In the ﬁrst and tenth lactations respectively, 4.97 and 21.38

percent of the cattle were culled due to lameness or injury.

Table 32 shows the within lactation removal schedule for non-death health culls

that was used for the typical Midwestern and Northeastern DHIA farm estimations. The

largest percentage of cattle were removed in the eleventh lactation month. The second-

highest percentage of cattle were removed in the ﬁrst month of lactation.

149

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E15 Enumaogtez 9:. 588.522 ..em 35% wE==U 953nm E5523 owaao>< 25.

5: mod wmgm 2.0 mod omw mo; 50$ 2
v0.0 who 3.3 3.: mod 36 v0; 3.3 a
who vmo we: 3.: om? 26 mm: wv.mm m
«me mad No.0. 3.2 mm? Vmo do; 3.? n
3.0 33 owe. wed vwd :0 voN $9» 0
cod 8.. mnem— onw mod ohm moN 8.3 m
7mm 2: on: mmw ﬂw ﬂaw NNN vﬁmv v
2; m: 36 $6 Eve 5mm 3N wusm m
cram w: 9;. Is an ohm EM Sam m
SN cod 3% to own 3:. mm.m _m.mm _
Ase
eé 3.» 2mm
8mm 2mm wcézu
axe 95:6 $55 ex; c5 oceam
gov 8mm DEE AX; 00m 83— 8893 5:803
8mm wa:—:0 new 83. mE==U was @530 580 2352 528:2
£39 88me «.85qu 2.262558; SE5 cozoscoi .8 Bow owEo>< 5:803
25a..—

.en «Bah.

150

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ood ocd 00.0 ocd ood ood cod 00.0 cod cod 2
00¢ cod cod cod ood cod ood cod cod cod 2
hash w0.wm moom >0.wN w_.om .ooN modm :.mm 2.0m ovdv :
mw.w w—S ms. 056 00.5 ows 2.x omw woo 0mg. 2
0v.0 0o.0 2.x. oos :8 0—6 oms Vms mms 000 o
mms om.0 oNs ww0 00.0 oh0 00.0 05.0 $0 vwé w
oww mms 0m.0 No.0 0V0 >20 wm0 om.0 wo.0 Saw N.
mm0 m0.m $0 om.0 3.0 $0 ooh 0w.m wwh V04» 0
mos o0.0 m0.m 0_.0 o_.0 0o.m o0.m Duh omw 05v m
oo.0 00.0 $0 :0 3.0 wwh wvw mmh cod 3.0 v
no.0 N00 vow wow wow wwh mmh o_ .m mmé 2.0 m
own NN.0 no.0 wm0 v0.0 mm.m oo.m NEW how oms N
50.2 wN.N_ omN— EN— 00:9 mm.N_ om.: one mos omw _

O— o m N. 0 n v m N ﬂ 5:02
:Omwwuoni— ccmuduowxu Gomuwuomi— Comuwuowiu comm—mucus.— COUmuomi— :Oﬂwuomi— Comumuowiu Comumuom‘r— :Omumuomr— GOﬂmuodA

 

 

 

 

 

 

552.com ==U 5.3: a he 352—Ema «an
2: ...... can: ...—:6 5.3m 5.39 :32 3 2:. 00:50 23.5 ..8 23.2—um .5653— :32323 5539 2:.

.bn 03:.

151

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8d cod cod 8.0 8.0 8.0 8e cod cod 8.0 2
8.0 cod 85 25 8d cod 8.0 cod 8.0 8.0 N.
2.2 $2 $2 3 .2 NS: 22 EN 82 $.mN NNNN :
2N 23 8.4 3.4 one 2.4 2 .m :3 o: NNN 2
2N 2N SN $3 22 Nam .3 SN 22 $4 a
2N EN 2N NN 2N NwN 8N GNN m3. .3 w
2N 2N QNN @N 8N 3N 5N ONN m: N: N
SN oNN 2 N SN 8N EN EN ONN 3N 8.4 o
2.». EN 3.4 EN EN Nam RN 2: 2.4 :e m
8.4 Now 2% SN No.4 8.... 2.4 one 31. $4 4
cos SN 24 $.m o? 82 Now 42 2% :.$ N
20 :N $0 93 m3 of a: New o: 8N N
9...? $2 8.? $2. 8.3 5.2. N. 3 22 EN :N 2
2 o N N o m 4 m N _ £82
comawaowd Gomuwaomi— :Omumuomi— Gomumwoni— comuwaowi— comuwuowg Gomuwuomd COCQBNJ comwwuowwu :Oﬂauowi— comuwuomi—
Gaza-‘10“
=5 5.8: a 2 52.5.2 Em 2.. .8 85 2:25 ..e 2.62% .5232 5:523 5.23 2F .2 2.3.

152

With the exception of the eleventh month (it was assumed in the estimation that no
health culls would occur during the dry period months, lactation months 12 and 13),
these removals are based on the typical removal schedule for Midwestern and
Northeastern DHIA dairy farms.

There was a different distribution for cattle deaths. With the exception of the
ﬁrst lactation, the majority of cattle that died during a lactation were removed during
the first lactation month (Table 33). The second-highest amount of deaths occurred
during the last lactation month. It was assumed that no deaths would occur during the
dry period months. Otherwise, the within lactation removal schedule for cattle deaths
were based upon the typical removal schedule for Midwestern and Northeastern
DHIA dairy farms.

Table 34 displays the other critical production and ﬁnancial factors used in the
BEA estimation for the typical Midwestern and Northwestern dairy farm estimation.
It was assumed that debt capital could be obtained for 7.50 percent. The opportunity
cost of equity capital was set at 10 percent. The debt to asset ratio was set at 50
percent for all of the example farms in this category. In order to estimate the BEA on
a before tax basis, the marginal tax rate was set at zero percent. Thus, the weighted
average cost of capital was calculated to be 8.7 5 percent for all farms.

A replacement heifer price of $1,350 and a cull cow price of $350 were used
in the estimation. The cull cow price was adjusted for age and whether the animal was
culled due to lameness or injury. These adjustments were explained in Chapter 5. A

milk price of $13.50 per hundredweight was assigned. Heifer calves were valued at

153

$200 per head and bull calves were valued at $75 per head. A death loss of 6 percent

was assumed for each.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 39. Other Critical Production and Financial Factors for the Typical
Midwestern and Northeastern DHIA Dairy Farm Estimation
Item Value
Interest Rate (%) 7.50
Percent of Cattle Funded by Debt (%) 50.00
Cost of Equity Capital (%) 10.00
Percent of Capital Funded by Equity (%) 50.00
Marginal Tax Rate (%) 0.00
Capital Gains Tax Rate (%) 0.00
Weighted Average Cost of Capital (%) 8.75
Replacement Heifer Price (S/cow) 1,350.00
Typical Cull Cow Price (S/cow) 350.00
Average Milk Price (S/cwt) 13.50
Somatic Cell Count Premium (5/ 100 lbs) 0.00063
Average Heifer Calf Price (S/cow) 200.00
Heifer Calf Mortality £96) 6.00
Average Bull Calf Price ($/cow) 75.00
Bull Calf Mortality (%) 6.00
Average Feed Cost LS/cow) 982.00
Average Labor Cost ($/cow) 646.00
Average Other Direct Cost LS/cow) 853.00
Average Cost Per Dry Cow Month (S/cow) 45.00
Typical Treatment Cost per Udder and Mastitis Health 133.00
Problem Episode (S/cow)
Typical Treatment Cost per Reproduction Health Problem 192.00
Episode (S/cow)
Typical Treatment Cost per Lameness and Injury Episode 9.00
(S/cow)
Typical Treatment Cost per Disease Episode (S/cow) 112.00
Milk Production Genetic Improvement Rate (% per 1.00
generation)
Feed Expense Adjustment to Support the Additional 0.90
Revenues Associated with the Milk Production Genetic
Improvement Rate Q6 of Milk Revenue Increase)
Labor Expense Annual Increase (%) 0.00
Other Expense Annual Increase (%) 0.00

 

 

154

 

An average feed cost of $982 per cow was assigned. The labor expense
assigned to each cow was $646. Both feed and labor costs were adjusted to an
individual cow’s production on a per hundredweight basis. Other direct expenses
were valued at $853 per cow. A dry period cost charge of $45 per month was also
assigned.

Treatment costs were also charged to culled cattle. For animals that were
culled due to udder and mastitis problems, a treatment cost of $133 was assigned. For
reproduction culls, a treatment cost of $192 was assigned. Lameness and injury culls
were assigned a charge of $9, and disease culls and deaths were assigned a treatment
expense of $1 12. These values were adapted from the work of Weigler et a1 (1990).

A milk production genetic improvement rate of 1.0 per generation was used in
the estimations. It was assumed that a new generation of heifers would begin milking
every 13 months. A feed expense adjustment of 0.9 was assigned to cover the
increase in feed costs necessary to capture the milk production increase from the
genetic improvement rate. Thus, if milk revenues increased by $1 from genetic
improvement, feed costs increased by $0.90.

Lactation expenses were adjusted by lactation according to the values
expressed in Table 9. The values were adapted from the work of Bauer, Mumey and

Lohr (1993).

155

Table 40. Lactation Expense Adjustment

Lactation Value
0.9627
1.0000
1.0218
1.0572
1.0745
1.1118
1.1136
1.1163
1.1191
1.1245

H

2
3
4
5
6
7
8
9

 

ﬂ
0

BEA Estimate Results for Midwestern and Northeastern DHIA Herds

For the example farm with a lactation speciﬁc health culling rate that was 50
percent higher than the Midwest and Northeast DHIA sample average (Table 36:
“50% Above ” row), a producer would be willing to pay $213 per month for the health
cull reduction. Thus, if the manager was looking at a health cull reduction investment
that would cost $213 or less per month for the 130 cow herd, the manager should
make the investment. This would reduce the projected average annual total culling
rate for this example farm from 41 percent to 39.2 percent. The marginal BEA for the
10 example farms decreased through the remaining farms. For the farm with the
sample average lactation speciﬁc culling rates, the BEA for a 10 percent health cull
reduction was $204 per month. Herds with 40 percent lower lactation speciﬁc culling
rates could afford to pay $172 for a 10 percent health cull reduction based on a 130

cow herd size.

156

 

 

 

 

 

 

 

 

 

 

 

 

Table 41. The Estimated BEA for Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 130 Cow Midwestern or
Northeastern DHIA Dairy Farm1
Example Previous New New NPV of the Estimated
Farm Average Average Annual Health Cull Monthly
(Described Annual Annual Total Reduction BEA for a
by the % Total Total Culling ($) 10% Health
Above or Cullin Culling Rate Range Cull
Below the Rate Rate (%) Reduction3
Sample (%) (%) (S)
Average
Health Cull
Rates)
50% Above 41.0 39.2 30.6 — 39.8 24,156 213
40 % Above 39.2 37.5 29.0 — 38.1 24,099 213
30% Above 37.5 35.7 27.4 — 36.4 23,978 212
20% Above 35.7 33.9 25.7 - 34.8 23,775 210
10% Above 33.9 32.0 24.1 - 33.2 23,471 207
Sample 32.0 30.1 22.4 - 31.7 23,041 204
Average
10% Below 30.1 28.2 20.8 - 30.3 22,459 198
20% Below 28.2 26.3 19.1 — 29.1 21,690 192
30% Below 26.3 24.3 17.5 - 28.2 20,697 183
40% Below 24.3 22.4 15.8 - 28.0 19,436 172

 

 

 

 

 

 

I The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held

constant at the breed average.

3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

Table 37 shows how much a 130 cow DHIA dairy farm with the Midwestern

and Northeastern sample average lactation speciﬁc health culling rate would be

willing to pay for a 10 percent reduction in speciﬁc health culls. The highest

estimated BEA was associated with lameness and injury culls. A farm of this type

would be willing to pay $62 per month to have a ten percent reduction in lameness

and injury culls. The second highest estimated BEA for herds in this category was to

157

 

Table 42.

The BEA of a 10 Percent Reduction in Specific Health Culls for a
240 Month Period for a 130 Cow Herd with the Midwestern and
Northeastern DHIA Sample Average Health Culling Ratel

 

 

 

 

 

 

 

Speciﬁc Previous New New NPV of the Estimated
Health Cull Average Average Annual Health Cull Monthly
Reduced by Annual Annual Total Reduction BEA for a

10 Percent Total Total Culling ($) 10% Health
Culling Culling Rate Range Cull
Rate 2 Rate (%) Reduction3
(%) (%) (S)
Udder and 32.0 31.6 23.7 — 32.8 6,031 53
SCC
Reproduction 32.0 31.5 23.6 - 32.8 7,303 65
Lameness 32.0 31.4 23.6 — 32.7 8,682 77
and Injury
Disease 32.0 31.9 24.0-33.1 1,079 10
Death 32.0 31.7 23.8-32.9 4,812 43

 

 

 

 

 

 

I The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held

constant at the breed average.

3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

spend $53 per month to reduce reproduction culls be ten percent. A producer in this

category would have an estimated BEA of $43 per month to reduce udder and SCC

culls by ten percent. Reducing deaths by ten percent would be worth an additional

$39 per month. A ten percent reduction in disease culls would beneﬁt the manager the

least. The BEA for a ten percent disease cull reduction was $8 per month.

Nevertheless, this estimate should be tempered by the fact that the disease health

culling rate incidence for the Holstein breed average was very low. Herds with

chronic disease problems whereby many animals are culled, such as in the case of

Johne’s Disease, could probably pay more to reduce the disease health culling rate.

158

 

IV. The Estimated BEA for a Health Culling Rate Reduction for Holstein,
Jersey and Guernsey Farms

In Section III, the maximum monthly BEA was calculated for the Midwestern
and Northeastern DHIA farms. However, no distinction was made between farms. Of
course, not all farms are the same. In this section, the monthly BEA for a ten percent
health cull reduction is determined for farms with three different breeds of dairy cattle
— Holstein, Jersey, and Guernsey —- to see how the monthly BEA might vary for farms
with different cattle breeds.

The Holstein breed was chosen as it is the major dairy cattle breed in the
United States. The Jersey breed was chosen because of its reputation for being a long-
lived breed. In Chapter 111, it was shown that Jersey cattle are less likely to be culled
than Holstein cattle. Guernsey cattle have the opposite reputation. In Chapter III,
Guernsey cattle were thirteen percent more likely to be culled in the Eastern region of
this analysis than Holstein cattle. Besides how the breeds vary in culling rates, the
breeds also vary in size, resulting in lower cull cow prices, milk price, and milk
production. As such, the maximum BEA for a ten percent reduction in health culls
should differ between the three breeds.

Breed- based Production and Financial Parameters

For the breed-based analysis, it was once again assumed that the dairy herd
would consist of 130 milking and dry cows with a thirteen month calving interval.
There were initially ten ﬁrst lactation cows in each of the eleven lactation months and
ten cows in each of two dry period months. The cattle in the two dry period months

were waiting for their second lactation to begin.

159

The production parameters used in this analysis can be seen in Table 38.
Holstein cattle tend to produce more pounds of ﬂuid milk whereas Jersey and
Guernsey cattle tend to produce less milk volume with higher milk component yields.
The milk production level of ﬁrst lactation Holstein cattle was set at 20,000 pounds of
milk. The milk production level of ﬁrst lactation Jersey and Guernsey cattle were set
at 14,000 pounds of milk. The lactation-based somatic cell count (SCC) levels were
Table 43. The Expected Rolling Herd Average (RHA) Per Lactation and

Somatic Cell Count Levels Per Lactation Used for the BEA
Estimation of a Health Cull Reduction for All Scenarios

 

 

 

 

 

 

 

 

 

 

 

 

 

Lactation Milk Production per Somatic Cell Count per
Lactation (lbs) Lactation
Holstein Jersey and Holstein, Jersey and

Guernsey Guernsey

1 20,000 14,000 200,000
2 22,308 15,616 254,840
3 23,000 16,100 285,900
4 23,136 16,195 308,700
5 22,950 16,065 326,460
6 22,482 15,737 346,280
7 21,790 15,253 356,460
8 21,158 14,811 358,480
9 20,378 14,265 363,560
10 19,736 13,815 362,060

 

 

 

 

held constant across all three breeds. First lactation cows were assigned a SCC level
of 200,000. The culling related production eﬁ‘ects used in the breed based estimates
were the same as those listed in Table 30 of Section HI.

Tables 39 — 41 show the lactation speciﬁc culling rates for the three dairy
cattle breeds. The lactation speciﬁc culling rates for Holstein cattle ranged from 22.26

percent in lactation 1 to 62.88 percent in lactation 10 (Table 39). The most common

160

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmw mod 3.3.. who 2.3 3.x Q... wwNo o—
Oms who 8.5 sma— omd. m2. 8; 3.8 0
No.0 mud 50w. 3.: 3.2 avg. _m._ 3.? w
one £6 £4: 54.2 2.9 mud $2 3.4% N.
94.0 cwd mmf No.2 3.2 Ono 2... om._m o
mmo 4o; 3.2 mod 9am Eh Q..— amsv m
Sam 3.. Nara.— ovw omw 42m 00.. 44.94 4
3.4 0: god as 00.0 won SN 43:” m
wva mm._ 2.6 Nms cmw New moN no.2” N
ooN No.0 36 S .4 Sam NM; mos omNN _
3°C Ge
32 22 3.2
Ge 255 255 2.5 g 32
@L 22 an? as 60m 22 889.5 2.56

2mm wE=:U 28 85— 22:30 98 mam—30 .030 :35? 23:52
5on 8829 32583 22282553— 325 53.665 8.“ Bow owns}. 5:203
0.3.5 5823— ..om 8am w:_==U 2.29on 55523 «3.83.. 2:. .3. 3:5.

161

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.4 end 4w.om :6 as $6 RN omdm E
2 .m $6 8.2 3.2 cod New SN 3.3 o
3.4 3.0 3.4— bod 3w 3.0 3N. 8.04 w
3.4 44.0 ”0: as oﬁw ooh mm.m 8&4 n
:6 36 05m 93 as how mm.m aﬁwm o
3.4 Gd how men oms N54 Rum 3.2 m
3.4 mud Sac oﬁm 3.0 44.4 3.4 no. _ m 4
mm.m 22o Eh $4 4Nm no.4 $4 3.5 m
9.: and 3.4 3.4 0.6m 3.4 00.0 3.3 m
w: and «W». mw.~ owN mm.m mmo mmdm _
3% 3.4
23. 83— 3.;
3% 3:5 @530 Ge 33 2mm
45 sum as? 33 com 22 8895 3:5

8am wa:—=0 can 84m wa:—:0 was wa:—:0 .680 .4352 89:32
580 8485 $28qu 833881 825 532605 5m 20m oweo>< 8:203
9.3.5 5029.. .8“— 85— wE==U 95925 55325 $5.22 2:. .m4 033—.

162

 

 

 

 

 

 

 

 

 

 

 

 

2.5 cod wohm 552 EL. 2.2 00.0 003 A:
50.0 oo.m m0: ooN— 36 3.: mm.m 3.0m a
one 3.0 3? Rd vow vow wad 3.3 w
36 3.0 3.2 mmN— wmw ace 3N 3.3 N.
who 8.0 002 v2: was Sun m0. v03 0
8.0 mmd no.3 02: cms 3:. 8.». modv m1
ooh ooé 8.2 Now 3.0 $6 gum nmwv v
In and 3 N. vow vw.v mos chm 3.? m
00M 5d 009 Rs mm.m 3w mo.m 8.3 N
va $6 $6 00¢ mow ova 02m :wdm _
3.» A5
22 23. £9
Gav 3:5 3:3 $3 $3 23.
33 23. has 3% 00m 29m 8893 3:5
89: wa:—:0 can 8mm wa:—:0 can wa:—:0 EEO _m::=< 59:52
530 88me $28qu cozozcoaom 525 53.605 Bm Eom omEQ>< c2863

 

 

 

 

 

 

 

 

 

 

23.5 mom=uo=0 .8“— Sam w=_=:U uEuoam 55.325 owaLo>< 2:

.3. 935,—.

163

reason for a Holstein cow to be culled during lactations 1 and 2 was “lameness and
injury" with 5.05 percent of the ﬁrst lactation Holstein cattle and 7.71 percent of the
second lactation Holstein cattle succumbing to lameness and injury problems. As
shown in Table 40, Jersey cattle exhibited the lowest lactation speciﬁc culling rates,
ranging from 20.34 percent in lactation 1 to 50.62 percent in lactation 9. The most
common culling reason for Jersey cattle in lactations 1 and 2 was “sold for dairy
purposes. ” In Table 41, the lactation specific culling rates for Guernsey cattle ranged
from 30.81 percent in lactation 1 to 61.56 percent in lactation 10. The most common
reason for ﬁrst and second lactation Guernsey cattle to be culled was “lameness and
injury” followed closely by “production. ” The within lactation removal schedules
for cattle culled due to non- death health culls and deaths were set equal to the
schedules shown in Tables 32 and 33.

Tables 42 shows the other production and cost parameters used in the breed-
based analysis. All Holstein, Jersey and Guernsey replacement heifers were valued at
$1,3 50 per cow. The typical cull price assigned to Holsteins was $3 50. The typical
value for a Jersey cull cow was set at $265 per cow. Guernsey cattle cull cow values
were set at $310 per cow. All cull cow values were adjusted for age and cull type as
discussed in Chapter 5.

Holstein milk was assigned a value of $13.50 per hundredweight. Jersey and
Guernsey milk was assigned a value of $15.50 per hundredweight. A SCC
premium/discount of $0.00063/cwt was assigned based upon the formula discussed in

Chapter 5. All heifer calves were assigned a value of $200 per calf regardless of

164

Table 47. Other Critical Production and Production Expense Factors for the

Breed Based BEA Estimations

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Item Value

Interest Rate (%) 7.50
Percent of Cattle Funded by Debt (%) 50.00
Cost of Equity Capital (%) 10.00
Percent of Capital Funded byEquity (%) 50.00
Marginal Tax Rate (%) 0.00
Capingains Tax Rate (%) 0.00
Weighted Average Cost of Capital (%) 8.75
Holstein, Jersey and Guernsey Replacement Heifer Price ($/cow) 1,350.00
Typical Holstein Cull Cow Price ($/cow) 350.00
Typical Jersey Cull Cow Price ($/cow) 265.00
Typical Guernsey Cull Cow Price ($/cow) 310.00
Average Holstein Milk Price ($/cwt) 13.50
Average Jersey and Guemsey Milk price ($/cwt) 15.50
Somatic Cell Count Premium ($llOO lbs) 0.00063
Average Heifer Calf Price ($/cow) 200.00
Heifer Calf Mortality (%) 6.00
Average Holstein Bull Calf Price ($/cow) 75.00
Average Jersey Bull Calf Price ($/cow) 55.00
Average Guemsey Bull Calf Price ($/cow) 65.00
Bull Calf Mortality (%) 6.00
Average Holstein Feed Cost ($/cow) 982.00
Average Jersey and Guernsey Feed Cost ($/cow) 683.00
Average Holstein Labor Cost ($/cow) 646.00
Average Jersey and Guemsey Labor Cost ($/cow) 455.00
Average Holstein Other Direct Cost ($/cow) 853.00
Average Jersey and Guernsey Other Direct Cost ($/cow) $95.00
Average Cost Per Dry Cow Month ($/cow) 45.00
Typical Treatment Cost per Udder and Mastitis Health Problem 133.00
Episode ($/cow)

Typical Treatment Cost per Reproduction Health Problem 192.00
Episode ($/cow)

Typical Treatment Cost per Lameness and Injury Episode 9.00
($/cow)

Typical Treatment Cost per Disease Episode ($/cow) 112.00
Milk Production Genetic Improvement Rate (% per generation) 1.00
Feed Expense Adjustment to Support the Additional Revenues 0.90
Associated with the Milk Production Genetic Improvement Rate

(% of Milk Revenue Increase)

Labor Expense Annual Increase (%) 0.00
Other Expense Annual Increase (%) 0.00

 

165

 

 

breed. Holstein, Jersey and Guernsey bull calves were valued at $75, $55 and $65 per
calf respectively.

Feed, labor and other direct costs were adjusted for the diﬂ‘erences in milk
production between the three breeds. Holsteins were assigned a feed cost of $982 per
cow while Jersey and Guernsey cattle were assigned a feed cost of $683 per cow. A
labor charge of $670 was assigned to Holstein cattle, and a charge of $646 per cow
was assigned to Jersey and Guernsey cattle. A charge of $853 and $595 per cow was
assigned to Holstein cattle and Jersey and Guernsey cattle respectively to cover other
direct expenses. Treatment expenses, the milk production genetic growth rate, and
lactation expense adjustments were set equal to those used in the Midwestern and
Northeastern farm estimations.

BEA Estimate Results for Holstein, Jersey and Guernsey Cattle

For the farm with Holstein cattle and a ﬁfty percent higher lactation speciﬁc
health culling rates than the average Holstein herd (production and sold for dairy
purpose culls were held constant at the Holstein breed average), the monthly BEA
associated for reducing the health culling rate by 10 percent was $218 per month
(Table 43). This means that the manager of such a herd could afford to pay up to $218
dollars per month for the ten percent reduction in health culls. Reducing the health
culling rates by ten percent would decrease the average annual total culling rate ﬁ‘om
41.5 percent to 39.7 percent for this farm. The marginal BEA for a ten percent
reduction in health culling rates for the “Breed Average ” example farm was $210 per
month. The marginal BEA continued to diminish through the remaining example

farms. The lowest returns, although still positive, were associated with a ten percent

166

lactation speciﬁc health culling rate reduction for herds with health culling rates that

were forty percent lower than the Holstein breed average. Managers with herds in this

category should be willing to pay up to $178 on a monthly basis in order to achieve

the reduction. A ten percent decrease in health culling rates for herds in this category

would cause the average annual total culling rate to decrease from 24.5 percent to

 

 

 

 

 

 

 

 

 

 

 

 

22.4 percent.
Table 48. The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 130 Cow Holstein Herd1
Example Previous New New NPV of the Estimated
Farm Average Average Annual Health Cull Monthly
(Described Annual Annual Total Reduction BEA for a
by the % Total Total Culling ($) 10% Health
Above or Culling Culling Rate Range Cull
Below the Rate 2 Rate (%) Reduction3
Sample (%) (%) (5)
Average
Health Cull
Rates)
50% Above 41.5 39.7 30.9 — 40.3 24,716 218
40 % Above 39.7 38.0 29.2 - 38.5 24,687 218
30% Above 38.0 36.1 27.5 - 36.8 24,594 217
20% Above 36.1 34.3 25.9 — 35.1 24,420 216
10% Above 34.3 32.4 24.2 - 33.5 24,143 213
Breed 32.4 30.5 22.5 — 31.9 23,738 210
Average '
10% Below 30.5 28.5 20.8 — 30.4 23,174 205
20% Below 28.5 26.5 19.2 — 29.2 22,415 198
30% Below 26.5 24.5 17.5 — 28.2 21,419 189
40% Below 24.5 22.4 15.8 — 27.9 20,137 178

 

 

 

 

 

 

I The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held

constant at the breed average.

3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

167

 

A ten percent reduction in the incidence of lameness and injury culls
generated the highest BEA for Holstein herds with the breed average lactation
speciﬁc health culling rates (Table 44). A farmer in this category would be willing to
pay up to $64 per month to reduce lameness and injury culls by ten percent. He or she
would be willing to pay $54, $44, and $40 per month to reduce reproduction culls
udder and SCC culls, and mortalities respectively. Reducing the incidence of disease
culls generated the lowest NPV of maximum potential returns. A farmer that
experiences the Holstein breed average lactation speciﬁc culling rate would be
willing to pay $8 per month to reduce the disease culling incidence by ten percent.
Table 49. The BEA of a 10 Percent Reduction in Specific Health Calls for a

240 Month Period for a 130 Cow Holstein Herd with the Breed
Average Health Culling Ratesl

 

 

 

 

 

 

 

 

 

 

Speciﬁc Previous New New NPV of the Estimated
Health Cull Average Average Annual Health Cull Monthly
Reduced by Annual Annual Total Reduction BEA for a

10 Percent Total Total Culling ($) 10% Health
Culling Culling Rate Range Cull
Rate 2 Rate (%) Reduction3
(%) (%) (5)
Udder and 32.4 32.0 23.8 — 33.1 5,028 44
SCC '
Reproduction 32.4 31.9 23.7 - 33.0 6,115 54
Lameness 32.4 31.8 23.7 — 32.9 7,277 64
and Injury
Disease 32.4 32.3 24.1 — 33.4 925 8
Death 32.4 32.1 23.9 - 33.2 4,569 40

 

 

 

I The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the breed average.

3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly
value.

168

 

The estimates for the example farms with Holstein cattle did not vary
tremendously from those determined in the estimates for the typical Midwestern and
Northeastern DHIA farms. This is not too surprising as the vast majority of dairy
cattle are Holstein. There was a big difference in the amount a manager with Jersey
cattle would be willing to pay, however (Table 45). Jersey cattle exhibited decreasing
marginal returns throughout the example farms. The example farm represented in the

“5 0% Above ” row of Table 45 exhibited the highest BEA for a ten percent reduction

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 50. The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 130 Cow Jersey Herdl
Example Previous New New NPV of the Estimated
Farm Average Average Annual Health Cull Monthly
(Described Annual Annual Culling Reduction BEA for a
by the % Cullin Culling Rate Range ($) 10% Health
Above or Rate Rate (%) Cull
Below the (%) (%) Reduction3
Sample ($)
Average
Health Cull
Rates)
50% Above 33.4 32.2 25.2 — 33.3 17,264 153
40 % Above 32.2 30.9 24.1 — 32.2 17,170 152
30% Above 30.9 29.6 22.9 -— 31.2 17,042 151
20% Above 29.6 28.3 21.8 - 30.2 16,874 ' 149
10% Above 28.3 27.0 20.7 — 29.4 16,659 147
Breed 27.0 25.6 19.6 — 28.7 16,388 145
Average
10% Below 25.6 24.2 18.5 - 28.2 16,051 142
20% Below 24.2 22.9 17.4 — 28.0 15,638 138
30% Below 22.9 21.5 16.2 — 28.2 15,136 134
40% Below 21.5 20.0 15.1 —28.9 14,530 128

 

 

 

 

 

 

T The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the breed average.

3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly
value.

169

 

in overall health culling rates. This example farm would be willing to pay up to $153

per month to achieve this reduction, which would lower the average annual total

culling rate from 33.4 to 32.2 percent. The manager of the “40% Below ” example

farm would be willing to pay up to $128 per month in order to reduce their health

culling rates by ten percent and their average annual total culling rate from 21.5 to

20.0 percent.

 

 

 

 

 

 

 

Table 51. The BEA for a 10 Percent Reduction in Speciﬁc Health Culls for a
240 Month Period for a 130 Cow Jersey Herd with the Breed
Average Health Culling Rates1
Speciﬁc Previous New New NPV of the Estimated
Health Cull Average Average Annual Health Cull Monthly
Reduced by Annual Annual Total Reduction BEA for a
10 Percent Total Total Culling ($) 10% Health
Cullin Culling Rate Range Cull
Rate Rate (%) Reduction3
(%) (%) (S)
Udder and 27.0 26.6 20.4 — 29.2 4,202 37
SCC
Reproduction 27.0 26.6 20.4 — 29.2 4,101 36
Lameness 27,0 26.6 20.4 — 29.2 4,674 41
and Injury
Disease 27.0 26.9 20.7 - 29.4 591 5
Death 270 26.7 20.6 - 29.3 2,930 26

 

 

 

 

 

 

I The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held

constant at the breed average.

3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

Managers of Jersey herds exhibiting the breed average health culling rate

would be willing to pay the most, $41 per month, for a ten percent reduction in

lameness and injury culls. Unlike Holstein cattle, udder and SCC culls were more

costly than reproduction culls. Jersey dairy farm managers with the breed average

170

 

lactation speciﬁc health culling rates should be willing to pay $37 per month for a ten
percent reduction in udder and SCC culls. They should be willing to pay $36 per
month to reduce reproduction culls by ten percent. Jersey dairy farms would be
willing to pay $26 per month for ten percent reductions death and $5 per month for
disease reductions.

Farm managers with Guernsey cattle would be willing to pay more for a ten
percent reduction in health culls than managers with Jersey and Holstein cattle.
Reductions in the health culling rate showed increasing marginal returns from the

Table 52. The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 130 Cow Guernsey Herdl

 

 

 

 

 

 

 

 

 

 

 

 

Example Farm Previous New New NPV of the Estimated
(Described by Average Average Annual Health Cull Monthly
the % Above Annual Annual Culling Reduction BEA for a
or Below the Culling Culling Rate Range ($) 10% Health
Sample Rate 2 Rate (%) Cull
Average (%) (%) Reduction3
Health Cull ($)
RatesL
50% Above 47.0 45.0 38.4 - 45.4 29,882 264
40 % Above 45.0 43.0 36.5 — 43.4 29,985 265
30% Above 43.0 40.9 34.6 - 41.5 30,049 266
20% Above 40.9 38.8 32.7 — 39.5 30,064 266
10% Above 38.8 36.7 30.7 - 37.6 30,013 265
Breed Average 36.7 34.5 28.8 - 35.7 29,877 264
10% Below 34.5 32.3 26.9 — 33.9 29,634 262
20% Below 32.3 30.1 24.9 — 32.3 29,254 259
30% Below 30.1 30.1 23.0 — 30.9 28,705 254
40% Below 27.8 25.5 21.0 — 30.0 27,947 247

 

 

 

 

 

 

The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the breed average.

3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly
value.

171

 

“50% Above ” through the “20% Above ” example farms (Table 47). The manager

with the example farm represented in the “Breed Average " row would be willing to

pay $264 per month to reduce health culling rates by ten percent, which would cause

the average annual total culling rates to reduce from 36.7 percent to 34.5 percent.

Managers with Guernsey herds in the “40% Below" initial lactation speciﬁc health

culling rate category would be willing to pay up to $247 per month to reduce health

culls be ten percent.

 

 

 

 

 

 

 

Table 53. The BEA for a 10 Percent Reduction in Speciﬁc Health Culls for a
240 Month Period for a 130 Cow Guernsey Herd with the Breed
Average Health Culling Ratesl
Speciﬁc Previous New New NPV of the Estimated
Health Cull Average Average Annual Health Cull Monthly
Reduced by Annual Annual Total Reduction BEA for a
10 Percent Total Total Culling ($) 10% Health
Cullin Culling Rate Range Cull
Rate Rate (%) Reduction’
(%) 1%) ($)
Udder and 31.8 31.4 23.6 — 32.6 4,223 37
SCC
Reproduction 31.8 31.3 23.5 - 32.6 7,766 69
Lameness 31.8 31.2 23.4 — 32.5 11,529 102
and Im'ury
Disease 31.8 31.8 23.9—32.9 756 7
Death 31.8 31.5 23.7 - 32.8 5,663 50

 

 

 

 

 

 

I The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the 0 — 150 cow herd size average.

3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly
value.

For Guernsey herd managers with the “Breed Average " lactation speciﬁc
health culling rates, the managers should be willing to pay up to $102 per month for a

ten percent reduction in lameness and injury culls (Table 48). They would be willing

172

 

to pay $69, $50, and $37 per month for ten percent reductions in reproduction culls,
deaths, and udder and SCC culls respectively. They would be willing to pay $7 per
month to reduce the number of disease culls.

As these results have shown, the estimated BEA for a ten percent reduction in
health culls varies based upon breed type. Farms with Jersey cattle, which generally
have lower culling rates than farms with Holstein cattle, cannot afford to pay as much
as Holstein or Guernsey farms for a ten percent reduction in health culls. Farms with
Guernsey cattle, however, can afford to pay more for a ten percent reduction in health
culling rates than Holstein dairy farms.

V. The BEA for a Ten Percent Health Culling Rate Reduction for Farms
that Vary by Size

In the previous sections, the BEA estimates were made for farms of the same
herd size. In this section, the maximum BEA for a ten percent reduction in health
culls was estimated for herds of three different sizes: 130 cows, 390 cows and 650
cows. A larger herd with equivalent culling rates as a smaller herd should be able to
pay more on an absolute basis for a ten percent reduction in health culls because more
cows are being culled. Besides this herd size difference, it was shown in Chapter 3
that herd size and expansion have some effect on the likelihood of a cow being culled.
As such, the BEA for a ten percent reduction in health culls may also vary due to
different culling characteristics between the three size categories.

Herd Size-based Production and Financial Parameters

With the exception of the herd size, herd inventory and lactation speciﬁc

culling rate parameters, it was assumed that all other parameters would be the same as

those identiﬁed for the Midwestern and Northeastern DHIA Farm estimations in

173

Section III. For the 130 cow herd estimations, it was assumed that the farm would
consist of 110 ﬁrst lactation heifers evenly distributed between eleven lactation
months. The remaining twenty cows would be evenly distributed between two dry
period months while waiting for their second lactation to begin. The 390 cow herd
would consist of 330 ﬁrst lactation cows evenly distributed between the eleven
lactation months. There would also be 60 cows evenly distributed between two dry
period months that are waiting to begin their second lactation. The 650 cow herd
would consist of 550 ﬁrst lactation heifers evenly distributed among the eleven
lactation months and 100 evenly distributed cows awaiting their second lactation in
two dry period months.

For the estimation of a 130 cow herd, the average lactation speciﬁc culling
rate for herds less than 150 cows were used (Table 49). The average culling rates
ranged from 21.91 percent in lactation one up to 59.52 percent in lactation ten. The
most common reason for cattle being culled in this size category was lameness and
injuries.

For the 390 cow herd size estimation, the lactation speciﬁc culling rates for
herds with 300 to 450 cows were used (Table 50 ). The average annual lactation
speciﬁc culling rates were higher than the previous size category. The average annual
lactation culling rate ranged from 22.95 percent in lactation one up to 74.19 percent in
lactation ten. This group culled less than the under 150 cow category for low
production and sold for dairy purposes and more for each health related reason. The

most prevalent culling reason was lameness and injury.

174

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:6 $6 2.5 vwd: cmd Vmw :N dem c:
3.0 wmd 3.8 3.: 35 3.0 v0: mmxm m
3.0 :6 we: 5.: end— 2 .5 mm: 3.3 w
who mwd No.2 2 d— de— Vme om: modm N.
3.0 33 owe: cod vwd :o voN wmdv 0
5w wwd 5.: 00¢ vww 2am SN 3.? m
mod 00.0 mm: va mow mow SN 3.? v
:.v wad 3.0 9:. owe En wnN 38m m
woﬁ 3.: El. was 0mm ooh mm.m 3.3. N
m5: $6 $6 Ed mm.m KM v2“ _o._~ _
Gov @u
and as. 3°C
ex» @530 3:6 3°C as: 23.
Ge 23. has Gav 8m aé 3885 3:5

23— mE==U can 83: wa:—:0 was wa:—=0 EEO .352 5252
Exam 0885 32.253 cocosvoaom .025 55355 5.: Bow owﬁo>< cosmos
950 cm: 52:. amou— 5; 8.5: ..oh 85— w£==U 328.5 55523 owaeo>< 2; .vm 033.

175

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mod 3.: NvNN m: .3 cow 3.: oo.o a: .vn O:
de wed 8.8 5.: 8.2 we? wvd Save a
ﬁnd co: 3 .2 5.x 8.: mNN_ VN: hKN: m
5.x we: 3:— wo.w 3N: 5.0 mod N03 5
EN 3..— m_.o_ 05.x 3.: 8.x 5.: Nm.mm o
it.» 3..— 3: 8.5 00.0: E N 3.: N: .Nm m
owe mm: NVN: mms mm.@ 3.0 a: .N Nosv v
$.m cm: 90.2 oos N: N Na: 2N dev m
EN SN 8.0 000 E .m 3% NmN VNNm N
voa 2.: 3w ome «Wm voN Nw: mmNN _
3.: 3.:
2,2 23: 3.:
3.: 3:5 3:6 ex: 3.: 32
3: a; 3:: 3.: 8m sé 889.5 3:5

23: wa:—=0 98 33— 35:0 28 35:0 EEO 33:5: 83:52
5on Undoﬂa mmocoEmw— :OEQSUOHQQM :OUvD comaosﬁohn— :0.“ 30m oweo>< €053”:wa
P560 cmv Ow can 5:? 29.—om boy.— 36“ w:m=——U uczoum :cmuauuux— owaho>< ugh. .mm o-aaht

176

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

coda cod 8.: 8.0— 8.2 cod ood 00.3 o—
wv.w 00.: no: 3.: 3.2 wvw 02“ EN: a
3.3 EN cod Saw 3.2 max m: .N vmdc w
3.0: ova cow Sum 00: wa 2am vav n
3.0: va on: 0.0.0 N5: and: .o.m wb._m o
3.: SN mwd mm: 00.2 mm: m: .m 3.5. m
8.: EN ems v0.0 wmw mus ovN 8.3 v
8.0 2 .N mm: 8.0 00.: 92. wow 5.; m
cm: 5.: 3:. cm: Sam 3:. RN Suwm N
8% m: ._ :mé mmﬁ Nod 05m ww.m «KEN :
ax: 3:
8mm 83. 3°:
3.: 3:5 3:5 G: :x: 22
3.: 23: has 3.: 8m 29: 88:5: 3:5

23: wa:—:0 ES 83: wa:—:0 was 3:30 EEO .m::c< 598:2
58D 0885 $28qu couosnoaom 825 c2839.: 8.: Bow oweo>< 8:803
350 ace :5; 0.82 :23 2:0! .5:— 85— 3:30 95925 55825 awake}: 2:. .cm 03:.

177

The 650 cow herd size estimation used the lactation speciﬁc culling rates
listed in Table 51 for herds with over 600 cows. Compared to farms with 300 to 450
farms, herds in this category had higher lactation speciﬁc culling rates in lactations
one through three, but lower lactation speciﬁc culling rates in lactations four through
ten. The annual lactation speciﬁc culling rates ranged from 27.79 in lactation one to
62.21 percent in lactation nine.
BEA Estimate Results for Dairy Farms that Vary by Herd Size

The monthly BEA for a ten percent reduction in health culls did vary by herd
size. This occurred on both an absolute basis and on a per cow basis.
For the 130 cow herd size category, the “5 0% Above ” example farm could afford to
pay up to $208 per month for a ten percent reduction in health culls (Table 52). The
average annual total culling rate dropped from 40.6 percent to 38.9 percent. The
marginal ability to pay for an additional 10 percent reduction decreased through the
remainder of the categories. For the “40% Below ” example farm, the manager could
afford to pay $167 per month for a ten percent reduction in health culling rates.
Achieving this reduction would cause their average annual total culling rate to reduce
from 24.2 to 22.2 percent. The “Size Category Average ” example farm could afford

to pay up to $198 ($1.52 per cow) per month for a ten percent health cull reduction.

178

 

 

 

 

 

 

 

 

 

 

 

 

Table 57. . The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 130 Cow Herdl
Example Previous New New NPV of the Estimated
Farm Average Average Annual Health Cull Monthly
(Described Annual Annual Culling Reduction BEA for a
by the % Cullin Culling Rate Range ($) 10% Health
Above or Rate Rate (%) Cull
Below the (%) (%) Reduction3
Sample ($)
Average
Health Cull
Rates)
50% Above 40.6 38.9 30.3 — 39.5 23,528 208
40 % 38.9 37.2 28.7 — 37.8 23,470 207
Above
30% 37.2 35.4 27.1 - 36.2 23,347 206
Above
20% 35.4 33.6 25.5 -34.5 23,144 205
Above
10% 33.6 31.8 23.9-33.0 22,842 202
Above
Size 31.8 30.0 22.3 — 31.5 22,417 198
Category
Average
10% Below 30.0 28.1 20.7 — 30.1 21,842 193
20% Below 28.1 26.1 19.1 — 29.0 21,086 186
30% Below 26.1 24.2 17.5 — 28.2 20,110 178
40% Below 24.2 22.2 15.8 - 28.0 18,874 167

 

 

 

 

 

 

I The initial herd inventory distribution was 1 10 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the O — 150 cow herd size average.
3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

Farmers with a herd size of 130 cows and the average lactation speciﬁc health

culling rates for the 0 to 150 cow herd size could afford to pay the most, $61, for a ten

179

 

 

percent reduction in lameness and injury culls (Table 53). This was followed by

reproduction culls ($53), udder and SCC culls ($43), deaths ($36) and disease ($7).

 

 

 

 

 

 

 

Table 58. The BEA for a 10 Percent Reduction in Speciﬁc Health Culls for a
240 Month Period for a 130 Cow Herd with the 0 to 150 Cow Herd
Size Average Health Culling Rate‘

Speciﬁc Previous New New NPV of the Estimated
Health Cull Average Average Annual Health Cull Monthly
Reduced by Annual Annual Total Reduction BEA for a

10 Percent Total Total Culling ($) 10% Health
Culling Culling Rate Range Cull
Rate 2 Rate (%) Reduction3
(%) (%L ($)
Udder and 31.8 31.4 23.6 — 32.6 4,902 43
SCC
Reproduction 31.8 31.3 23.5 — 32.6 6,032 53
Lameness 31.8 31.2 23.4 - 32.5 6,879 61
and Injury
Disease 31.8 31.8 23.9 — 32.9 756 7
Death 31.8 31.5 23.7—32.8 4,026 36

 

 

 

 

 

 

I The initial herd inventory distribution was 110 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 20 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the 0 - 150 cow herd size average.
3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

For the example farms with a 390 cow herd size, the marginal estimated BEA

for a ten percent decrease in health culling rates decreased through the example farm

categories (Table 54). This size category average farm could pay up to $676 per

month or $1.73 per cow per month for the health cull reduction, which would reduce

its total annual average culling rate from 33.1 percent to 31.2 percent. The “40%

Below ” example farm could afford to pay $585 for the reduction, which would lower

the projected average annual total culling rate from 25.1 percent to 23.0 percent.

180

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 59. The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 390 Cow Herdl
Example Previous New New NPV of the Estimated
Farm Average Average Annual Health Cull Monthly
(Described Annual Annual Culling Reduction BEA for a
by the % Culling Culling Rate Range ($) 10% Health
Above or Rate 2 Rate (%) Cull
Below the (%) (%) Reduction3
Sample ($)
Average
Health Cull
Rates)
50% Above 42.5 40.7 31.4 — 41.2 79,290 701
40 % Above 40.7 38.8 29.6 — 39.4 79,194 700
30% Above 38.8 37.0 27.8 - 37.6 78,932 698
20% Above 37.0 35.1 26.1 — 35.8 78,448 693
10% Above 35.1 33.1 24.3 — 34.1 77,680 686
Size 33.1 31.2 22.5 -— 32.4 76,550 676
Category
Average
10% Below 31.2 29.2 20.7 — 30.7 74,963 662
20% Below 292 27.2 18.9 — 29.3 72,807 643
30% Below 272 25.1 17.1 — 28.1 69,945 618
40% Below 25.1 23.0 15.3 — 27.4 66,218 585

 

 

 

 

 

 

I The initial herd inventory distribution was 330 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 60 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the 300 — 450 cow herd size average.
3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

A 390 cow herd with the 300 — 450 cow herd size average health culling rate

could afford to pay the most, $191 per month, for a ten percent reduction in lameness

and injury culls (Table 55). This was followed by ten percent decreases in

reproduction culls ($154 per month), deaths ($154 per month) udder and SCC culls

($138 per month), and disease culls ($43 per month).

181

 

 

 

 

 

 

 

 

Table 60. The BEA for a 10 Percent Reduction in Specific Health Culls for a
240 Month Period for a 130 Cow Herd with the 300 to 450 Cow
Herd Size Average Health Culling Ratel
Speciﬁc Previous New New NPV of the Estimated
Health Cull Average Average Annual Health Cull Monthly
Reduced by Annual Annual Total Reduction BEA for a
10 Percent Total Total Culling ($) 10% Health
Cullin Culling Rate Range Cull
Rate Rate (%) Reduction3
(%) (%) ($)
Udder and 33.1 32.7 23.9-33.7 15,586 138
SCC
Reproduction 33.1 32.7 23.8 - 33.7 17,477 154
Lameness 33.1 32.6 23.7-33.5 21,613 191
and Injury
Disease 33.1 33.0 24.1 — 34.0 4,907 43
Death 33.] 32.8 23.9—33.7 17,462 154

 

 

 

 

 

 

I The initial herd inventory distribution was 330 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 60 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the 300 - 450 cow herd size average.
3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

For the 650 cow estimation (Table 56) there was a decreasing ability to pay

for a ten percent health culling rate reduction throughout the ten example farms. The

manager of the “50% Above " example farm could afford to pay $1,815 per month for

a ten percent reduction in health culls. This would reduce the average annual total

culling rate from 49.5 percent to 46.7 percent for this example farm. The manager of

the “40% Below ” example farm could afford to pay up to $1,113 per month for a ten

percent health cull reduction. The average annual total culling rate would decrease

from 26.0 to 23 .7 percent for the 240 month period for this farm. The “Size Category

Average” example farm could afford to pay up to $1,499 ($2.72 per cow) per month

to reduce their health culling rate by ten percent.

182

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 61. The BEA Associated with Successive 10 Percent Health Cull
Reductions for a 240 Month Period for a 650 Cow Herdl
Example Previous New New NPV of the Estimated
Farm Average Average Annual Health Cull Monthly
(Described Annual Annual Culling Reduction BEA for a
by the % Cullin Culling Rate Range ($) 10% Health
Above or Rate Rate (%) Cull
Below the (%) (%) Reduction3
Sample ($)
Average
Health Cull
Rates)
50% Above 49.5 46.7 39.5 - 47.0 205,339 1,815
40 % 46.7 43.9 36.8 — 44.3 199,608 1,764
Above
30% 43.9 41.2 34.2 — 41.7 193,250 1,708
Above
20% 41.2 38.5 31.7—39.2 186,189 1,645
Above
10% 38.5 35.9 29.3 - 36.8 178,354 1,576
Above
Size 35.9 33.3 26.9 — 34.6 169,679 1,499
Category
Average
10% Below 33.3 30.8 24.6 — 32.6 160,114 1,415
20% Below 30.8 28.4 22.4 — 31.0 149,629 1,322
30% Below 28.4 26.0 20.2 - 29.8 138,222 1,221
40% Below 26.0 23.7 18.2 - 29.3 125,927 1,113

 

 

 

 

 

 

I The initial herd inventory distribution was 550 first lactation heifers evenly
distributed between 1 1 lactation months and 100 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.

2 Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the 600 plus cow herd size average.
3 The monthly BEA is the equivalent annuity of the NPV expressed as a monthly

value.

Unlike the previous estimations, a 650 cow farm that exhibits the typical

lactation speciﬁc health culling rate for the 600 plus herd size could afford to pay the

most ($435 per month) for a ten percent reduction in deaths (Table 57). Reducing

lameness and injury culls by ten percent would save a manager $274 per month. A

183

 

manager of a 650 cow herd could afford to pay up to $252, $250, and $76 per month
to reduce udder and SCC culls, reproduction culls, and disease culls respectively.
Table 62. The BEA for a 10 Percent Reduction in Specific Health Culls for a

240 Month Period for a 650 Cow Herd with the 600 or More Cow
Herd Size Average Health Culling Rate‘

 

 

 

 

 

 

 

Speciﬁc Previous New New NPV of the Estimated
Health Cull Average Average Annual Health Cull Monthly
Reduced by Annual Annual Total Reduction BEA for a

10 Percent Total Total Culling ($) 10% Health
Culling Culling Rate Range Cull
Rate 2 Rate (%) Reduction3
(%) (%) ($)
Udder and 35.9 35.4 28.9 — 36.4 28,496 252
SCC
Reproduction 35.9 35.5 28.8 — 36.4 28,313 250
Lameness 35.9 35.4 28.8 — 36.4 31,029 274
and Injury
Disease 35.9 35.8 29.1 — 36.7 8,566 76
Death 35.9 35.3 28.7 —36.3 49,253 435

 

 

 

 

 

 

TThe initial herd inventory distribution was 330 ﬁrst lactation heifers evenly
distributed between 11 lactation months and 60 dry cows awaiting their second
lactation evenly distributed between their ﬁrst and second dry cow months.
2Lactation speciﬁc production and sold for dairy purposes culling rates were held
constant at the 300— 450 cow herd size average.
3The monthly BEA IS the equivalent annuity of the NPV expressed as a monthly
value

VI. Using the DSS to Estimate the Profitability of Rubberized Alleyway
Surfaces

In the previous sections, the DSS has been used to determine the BEA for a
general reduction in health culls. Farms participating in the NAHMS Dairy ’96
Survey that had their cows walk primarily on soft surfaces experienced a 17 .5
percent decrease in lameness and injury culls in Chapter 41. Although they didn’t look

at the effect on culling rates per se, Bray, Giesey, and Bucklin examined the effect of

 

‘( Lameness and Culling Rate Reduction / Lameness and Culling Rate Mean Response )* 100 %- -
(0 528/3 025)*100°/o=17 454 percent

184

 

 

 

having rubberized floors on lameness episodes on Florida cattle (2002). The
researchers had two 130 cow groups, a 130 cow control group on concrete and a 130
experimental cow group on rubberized flooring. The experiment lasted one year. The
control group experienced 37 percent fewer lameness episodes. With an average
treatment expense of $300 per episode, the experimental group experienced $6,600
lower treatment costs than the control group in that one year period. In this section,
the DSS will be used to estimate the proﬁtability of installing a rubberized freestall
alleyway ﬂoor.

It was assumed that the hypothetical herd for this estimation would take on the
characteristics used in the estimation for the 130 cow Holstein herd used in the
Holstein breed estimation in Section IV with one exception, the lameness and
treatment episodes were increased to $300 per episode. It was assumed that the barn
would be a three row variety with 4,775 square feet of alleyways to surface. Purchase
and installation costs were set at $4 per square foot (Hadley, 2002). Prior to
installation of the rubberized ﬂoors, the breed average lactation speciﬁc culling rates
were used. Aﬁer the installation, it was assumed that the lameness and injury culling
rate would reduce by 17.5 percent for all lactations. A marginal tax rate of 35 percent
and a capital gains tax rate of 15 percent were used in this analysis. The rubberized
floor was depreciated using the 5 year Half Year Convention MACRS schedule.

The present value of the 240 months of cash ﬂows without the rubberized
floor was $1,450,086. The present value of the 240 months of cash ﬂows with the
rubberized floors was $1,459,554. The present value of the technology cash flows

was -$32,448. Thus, the NPV of the investment in rubberized ﬂoors was:

185

$1,459,554 — $1,450,086 -— $32,448 = -$22,980.

Because the NPV of the rubberized ﬂoor was negative, it indicates that the ﬂoor
should not be adopted on the basis of reduced lameness and injury culls alone. When
converted to a monthly equivalent basis, the rubberized ﬂoor is $161 per month less
proﬁtable than not having a rubberized ﬂoor from a culling perspective.

This value does not, however, include the total reductions in lameness
episodes. According to the research of Bray, Giesey, and Bucklin — the lameness
episodes decreased by 37 percent which saved $6,600 per year in treatment costs for
the 130 cow experimental group (2002). When converting the $6,600 to a monthly
equivalent basis, the savings in treatment expenses was equivalent to $528 per
monthz. If the rubberized ﬂoor proﬁtability analysis is adjusted to not include the
$300 in lameness and injury treatment expenses, the monthly equivalent value is -
$162 per month instead of -$161. Thus, from an overall herd health and culling
expense reduction, adopting the technology will result in $528 - $161 = $367 more
income per month and should be adopted.

VII. Using the DSS to Estimate the Proﬁtability of Gonadotropin Releasing
Hormones

Nagategize examined the proﬁtability of using GNRH and human chronic
gonadotropin (HCG) to treat cystic ovaries in cattle, which generally renders cattle
infertile (1988). Left untreated, 30 percent of the animals with cystic ovaries will
recover. Those that do not recover are generally culled. Cattle with cystic ovaries that
are treated with GNRH are 76 percent likely to recover. Those that fail to recover

after the initial treatment are also 76 percent likely to recover with a follow up

 

2 Annual Amount/ (Future Value of an Annuity...8 m... .12) = $6,600/“(1.00727)” — 1)/.00727] = $528.

186

treatment of GNRH. Cattle treated with HCG are 68 percent likely to recover after the
ﬁrst and second treatments with HCG. Ngategize found that it was more proﬁtable to
treat cattle with GNRH and HCG twice prior to culling than to not treat the animals
and that GNRH is more proﬁtable to use than HCG.

In this section, the ﬁnancial feasibility of using GNRH to reduce the number
of cattle culled due to cystic ovaries is determined using the DSS. Currently, HCG,
which is less effective than GNRH, costs the same as GNRH and is generally not
used for cystic ovary treatments (Bauman, 2003). Thus, only the ﬁnancial feasibility
of GNRH is examined in this section.

Production and Financial Parameters

Based upon a study conducted in Michigan, it was assumed that the incidence
of cystic ovaries was 12.8 percent (N gategize, 1988). For a 130 cow herd, that means
that 16.64 cows would develop cystic ovaries in a given year. Assuming that 30
percent would recover naturally means that 11.65 of these cows would be culled
without treatment. Using GNRH for at the most two treatments would mean that only
0.96 cows would be culled. The cost of these treatments would be $82.52 (Bauman,
2003)

For this analysis, it was assumed that the farm would assume the
characteristics of the farm used in the rubberized alleyway ﬂoor example with two
exceptions. The farm’s current and desired lactation speciﬁc culling rates would be
set according to Tables 58 and 59 respectively. These distribution were chosen to
permit a reproduction culling rate that would permit the cystic ovary incidence rate '

describe by Ngategize (1988). For the current lactation speciﬁc culling rate, 19 cows

187

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Quo— hwd mowm 3.3 8.: Saw 2.: 3.3 o—
hmd Std cosh Quom M5,: wms we: coda o
3.x oo: :08 mmdm cm: 91. .0: 33 m
Kw o: ._ Nvdm 3.8 mm: mm: we: 5:: N.
Kw 0: mod: vmhm 2.2 One on: 3.3. c
was mm: 3.: Noam moa— Sw on: 5.00 m
3.: mm: 3.2 3.8 0:: www om: 8.3 v
New av: mm: 3.2 mow wow MMN :sm m
Sly mm: and 3.»: mm: new mod mode m
VON. end mi: 3.9 m3» mw.m mom 5. _ m _
3.: Q:
so: as. 3.:
3.: 3:5 3:5 3.: ex: ea:
3.: use: has 3.: 8m 23. 38:5: 3:5
83: wa:—:0 use 83: 3:30 can wa:—nu BBQ :35}: 39:32
£309 ammomma mmocoEmA GOSUSUOHQQM uoDﬁD comuozﬁohm ..om 20m owmuv>< comagomd
:cmaaamumm

5:37;"; 32:25: 575 2:. ..8 3.5— wE==U #:9on 53523 25.—.59 2:. .8 «35.—t

188

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.62 5.0 8.3 moan 8.2 2..» mm; 3.3 2
two vbo oosm oosm 8.2 was 8.2 cod» a
5.x 8.2 003 003 092 9; $2 8.3 w
:w o: NVNN NYNN 32 2%. v0.2 :25 n
Ex 2 ._ 32 3.2 2.2 owe 2.2 2.2. c
was mm; 8.2 32 no.2 5w 3.2 made m
3.0 mm; 3.2 3.2 oz: vww 32 82% v
36 22 9.2 3.2 mow wow SN 2.? m
21‘ mm; mwd 2.: 3.0 New m2“ mcdv N
voN owd 220 3.0 2; SM ooh 3.; 2
Nos 25
3% 22 25
2.5 $56 wees $3 2.5 22
Aé 22 as? 2.5 08 2am 889.5 3:5
29m @530 28 232 2.530 new wa:—3U EEO .352 89:52
58a 0385 30:0qu 5538qu 525 55.605 5.2 Eom owﬁo>< c2383
..ccaEcmm

$37.3”— EEEEE EZU 2.. ...:— 83m 9530 #5on 55325 3:89 2; .3 “...—5‘

189

of which 11.65 are culled due to cystic ovaries, are culled for reproduction reasons.
For the desired lactation specific culling rates, 8.43 cattle are culled due to
reproductive failure. The difference between the two distributions are caused by the
successful use of the GNRH treatments.
Financial Feasibility Results for the Use of GNRH

The use of GNRH generated a net present value of $66,233 for the twenty
year period or $463 per month. Thus, a farmer with the characteristics of the one used
in this analysis would be $463 per month more proﬁtable with the GNRH treatments
than without the treatments. The use of GNRH in this estimation resulted in the
average total culling rate decreasing from 43.40 percent to 36.59 percent.
VIII. Summary and Conclusions

The BEA for a health culling rate reduction can vary based on farm
characteristics. In the estimation results described in this chapter, the BEA varied by
health cull incidence. For some very high health cull instances, the BEA actually
increased for each successive ten percent health cull reduction. Over most of the
lactation specific health cull ranges, the marginal BEA diminished with each
successive ten percent health cull reduction. Thus, managers and advisors should not
think it likely that they can afford to pay the same amount for successive reductions
in their health culling rate, and the DSS should be used to estimate the BEA aﬁer
each successful health cull rate reduction.

The BEA differed based upon breed and herd size. In general, a Guernsey
herd could pay the most for a ten percent health culling rate reduction. Although the

BEA was still positive for a Jersey herd, Jersey farm managers could afford to pay

190

the least for a ten percent health cull reduction. As herd size increased, the BEA
increased with herd size on both an absolute and per cow basis. The relative
importance of reducing cattle mortalities increased with herd size. For the 650 cow
herd, a ten percent reduction generated the highest BEA. For the 130 cow herd
estimate, reducing deaths was the fourth highest BEA behind lameness and injury,
reproduction, and udder and SCC culls.

It was determined that it was ﬁnancially infeasible to adopt rubberized cattle
alleyway ﬂoors from a culling reduction perspective. Nevertheless, if the reduction in
non-culling lameness episode treatments are included, the technology is proﬁtable to
adopt.

GNRH proved to be a very proﬁtable technology to use to reduce the
incidence of culling due to cystic ovaries. Treating cattle up to two times generated

positive returns of $487 per month over and above the treatment costs.

191

CHAPTER 7.

SUMNIARY

In Chapter 1, it was stated that the ultimate goal of this research was to develop a
Decision Support System (DSS) to aid producers and advisors in deciding the most
proﬁtable methods to reduce their health culling rate. In order to develop such a DSS, it
was important to understand how culling affects production, how many cattle are culled,
why cattle are culled and how management programs affect culling rates.

In Chapter 2, DHIA records for ten Midwestern and Northeastern states were
examined to determine how culling, on average, affects production, the percentage of
cattle culled each year, why cattle are culled, and when cattle are culled within a
lactation. It appears as though there are both production incentives and disincentives to
reducing culling rates and increasing herd culling rates. Cattle generally increase in milk,
milk fat, and milk protein production throughout the ﬁﬁh lactation. Nevertheless, somatic
cell counts also rise with age and should be accounted for when determining whether to
decrease culling rates. Cattle that were culled for health reasons produced less milk than
their healthy herd mates. The majority of cattle were culled for health reasons.
Midwestern farms tended to have higher culling rates than the Northeastern farms. The
majority of cattle culled for health reasons other than death were culled at the end of a
lactation. Most mortalities, however, occurred at the beginning of a lactation.

In Chapter 3, a probit model was used to determine how individual cow and herd
level characteristics contributed to the likelihood that a cow would be culled. Data for
this model was taken from individual cow and herd level DHIA data for ﬁve Midwestern

I and ﬁve Northeastern states for the period of 1995 — 1999. Because of the difference in

192

culling rates between the two regions, the probit model was applied to each region
separately. The probit model was able to predict culling and non-culling events at an
eighty and eighty six percent accuracy rate for the Midwestern and Northeastern
estimations respectively.

Cattle that calved during Winter were more likely to be culled than cattle that
calved in Spring in both regions. Summer calving Midwestern cattle were less likely to
be culled than their Spring calving counterparts. Cattle calving in Fall were less likely to
be culled than their Spring calving counterparts, but, in the Northeastern region, the
opposite was true. The likelihood of a cull increased with each successive lactation. The
difference between a cow and her herd mate’s milk, milk protein, and her production
persistency was negatively correlated with the likelihood of a cull. The difference
between a cow and her herd mate’s somatic cell count and the previous lactation’s
services per conception was positively correlated with the likelihood of a cull. Cattle with
higher or lower genetic potential to produce milk, milk fat, and milk protein did not have
a signiﬁcantly different likelihood than those with average genetic ability. In the
Northeast, however, cattle that had lower genetic ability to produce fat were less likely to
be culled. Jersey cattle were less likely to be culled and Guernsey cattle more likely to be
culled in both regions. Cattle in Midwestern registered herds were less likely to be culled,
but cattle on registered Northeastern herds were more likely to be culled. State of origin
also had a signiﬁcant effect on culling. Illinois, Iowa, and Wisconsin had signiﬁcantly
lower culling likelihoods than Indiana. In the Northeast, only Pennsylvania had a
signiﬁcantly lower culling likelihood than Vermont. Herd size had a small but signiﬁcant

negative effect on the probability of a cull in the Midwest, but a positive effect in the

193

 

Northeast. There were mixed effects associated with expansion and heifer ratio variables.
The milk feed price and cull cow to replacement heifer price ratio of certain months had a
signiﬁcant effect on the likelihood of a cull.

Four ordinary least squares were developed to determine the effect that select
management factors had on reducing the udder and mastitis, lameness and injury, disease,
and reproduction culling rates. Only a few of the management factors signiﬁcantly
contributed to the udder and mastitis, lameness and injury, disease and reproduction
culling rates at a p-value of O. 1000 or less. Only two management factors signiﬁcantly
affected udder and mastitis culling rates, having an employee handbook and using
composted manure for bedding. Three factors were positively correlated with the
lameness and injury culling rate, the presence of hairy heel warts, the number of
veterinary visits — indicating that managers may primarily use veterinarians for treatment
rather than prevention — and rolling herd average. Farms with multiple animal facilities
and farms with soft cattle surfaces had signiﬁcantly lower lameness and injury culling
rates. There were only two management programs that signiﬁcantly affected disease
culling rates. Managers who prevented their youngstock from nose-to-nOse contact with
other species had lower disease culling rates than managers who did not. Farms that had
animals that tested positive for Johne’s Disease exhibited higher disease culling rates.
Employee handbooks were also effective at reducing the reproduction culling rate but
only when combined with incentives. Herd size was also negatively correlated with
reproduction culling rates, possibly indicating that larger farms are more apt to hire

specialized labor and to adopt reproduction technologies. Using herd bulls was also an

194

effective method of reducing reproduction culling rates. Rolling herd average was
positively correlated with reproduction culling rates.

In Chapter 6, the DSS was used to show that the maximum potential returns for a
reduction in health culls varied by breed and herd size. Farms with Guernsey cattle were
estimated to be able to pay the most for a ten percent health cull reduction followed by
Holstein farms and Jersey farms. In general, producers could pay more for lameness and
injury culls. The potential returns associated with health cull reductions decreased with
each ten percent health cull reduction. Managers with larger herds were able to pay more
per cow per month for health cull reductions than managers of smaller herds. Lameness
and injury cull reductions had the highest potential returns for 100 and 400 cow herds,
but decreasing mortalities created the highest potential returns for herds of 600 or more

COWS.

195

 

Bibliography

196

 

Bauer, L. G. Mumey, and W. Lohr. “Longevity and Genetic Improvement Issues in
Replacing Dairy Cows.” Canadian Journal of Agricultural Economics. Vol. 41. 1993.
pp. 71-80.

Bauman, D. W. Everett, L. Weiland, R. Collier. “Production Responses to Bovine
Somatotropin in Northeast Dairy Herds.” Journal of Dairy Science. Vol. 82. 1999. pp.
2564-2573.

Bauman, L. Personal Correspondence. December 5, 2003.

D. Bray, R. Giesey, R. Bucklin. “Should the Rubber Meet the Road.” Dairy Update.
University of Florida Quarterly Newsletter. Winter 2002.

Cassell, B. “She will live a LOT longer if she is pregnant.” Hoard ’s Dairyman. August
25, 2002.

Cattell, M. “Culling and Laminitis: Real cows, real deaths, real losses.” Dairy Herd
Management. February 2001.

Collier, R. J. Byatt, S. Denham, P. Eppard, A. Fabellar, R. Hintz, M. McGrath, C.
McLaughlin, J. Shearer, J. Veenhuizen, J. Vicini. “Effects of Sustained Release Bovine
Somatotropin (Sometribove) on Animal Health in Commercial Dairy Herds.” Journal of
Dairy Science. Vol. 84. 2001. pp. 1098-1108.

Cook, N. “How Cow Comfort Impacts Hoof Health.” Conference Proceedings: 6th
Annual Hoof Care Seminar. December 11, 2002.

Dairy Proﬁt Weekly. “Idaho Turnover Rate Near 40%.” Dairy Proﬁt Weekly. Vol. 13.
Number 18. May 6, 2002. p. 1.

Dairy Records Management Systems. “DHI Glossary: Fact Sheet: A-4.” 1999. pp. 2-26.

Delorenzo, M. T. Spreen, G. Bryan, and D. Beede. “Optimizing Model: Insemination,
Replacement, Seasonal Production, and Cash Flow.” Journal of Dairy Science. Vol. 75.
1992. pp. 885-895.

Dijkhuizen, A. J. Renkema, and J. Stelwagen. “Economic Aspects of Reproductive
Failure in Dairy Cattle. II. The Decision to Replace Animals.” Preventive Veterinary
Medicine. Vol. 35. 1985. pp. 256-276.

Fleischer, P. M. Metzner, M. Beyerbach, M. Hoedemaker, and W. Klee. “The

Relationship Between Milk Yield and the Incidence of Some Diseases in Dairy Cows.”
Journal of Dairy Science. Vol. 84. 2001. pp. 2025-2035.

197

 

Galton, D. “Extended Calving Intervals, BST May Be Proﬁtable.” Feedstuffs. October
13, 1997. pp. ll-l3.

Gujarati, D. “Basic Econometrics.” Third Edition. McGraw-Hill, Inc. 1995. pp. 540 —
570.

Hadley, G. “The Managerial, Production and Financial Implications of Dairy Farm
Expansion in Michigan and Wisconsin.” Masters Thesis. Michigan State University.
2001.p.62.

Hadley, G. “The Cost of Concrete Alternatives for Cow Traffic.” Conference
Proceedings: 6‘h Annual Hoof Care Seminar. December 11, 2002.

Hansen, L. “The Role of Genetics in Cow Longevity.” Midwest Dairy Herd Health
Conference Proceedings. November 2002. pp. 105-122.

Hoard’s Dairyman. “Why Heifer Prices Have Backed Off.” Hoard 's Dairyman. March
25,2003.p.231.

Houben, E. R. Huirne, and A. J. Dijkhuizen. “Optimal Replacement of Mastitic Cows
Determined by a Markov Process.” Journal of Dairy Science. Vol. 77. 1994. pp. 2975-
2993.

Howard, W. M. Hutjens, J. Reneau, and N. Hartwig. “Dairy Management: The
Reproduction Clinic.” wwwinformumdedu June 1992. pp. 1-15.

Jones, B. “Cow Longevity and Optimal Culling Decisions in Dairy Operations.” 2001
Arlington Dairy Days Proceedings, University of Wisconsin - Madison Department of
Animal Science. www. Wisc.edu/dysci/uwex/brochurcLS/atrlingtonddayOl.pdf December
2001.

Jones, B. “Economic Consequences of Extending the Calving Intervals of Multiparous
Cows” Conference Proceedings. Arlington Dairy Days Proceedings, University of
Wisconsin - Madison Department of Animal Science.

www. Wisc.edu/dysci/uwex/brochures/arlingtonddayO1.pdf December 2001.

Kelm, S. (Steven. Kelm @uwrf. edu). (2003, May 27) RE: Genetics Help Email to
Gregg Hadley (Gregg Hadley@uwrf. edu).

Keown, J. “How to Estimate a Dairy Herd’s Reproductive Losses.”
www.ia_nr.unl.edu/pubsLDairy/g822htm November 1986.

Lehenbauer, T. and J. Oltjen. “Dairy Cattle Culling Strategies: Making Economical
Culling Decisions.” Journal of Dairy Science. Vol. 81. 1998. pp. 264-271.

198

 

Maddala, G. “Limited Dependent and Qualitative Variables in Econometrics.”
Cambridge University Press. 1992. pp. 15-28.

Miller, G. P. Bartlett, S. Lance, J. Anderson, and L. Heider. “Costs of Clinical Mastitis
and Mastitis Prevention in Dairy Herds.” Journal of the Applied Veterinary Medicine
Association. Vol. 202. No. 8. April 15, 1993. pp. 1230 — 1236.

Miranda, M. and G. Schnitkey. “An Empirical Model of Asset Replacement in Dairy
Production.” Journal of Applied Econometrics. Vol. 10. 1995. pp. 541-555.

Natzge, D. “Monitor when and why, then manage.” Midwest Dairy Business. September
2002.p.13.

Ngategize, P. “A Model for Evaluating Animal Health Management Strategies with the
Cow Viewed as a Durable Asset.” Ph.D. Dissertation. Michigan State University. 1998.

Ngategize, P., S. Harsh, and J. Keneene. “Applications of Replacement Theory in Dairy
Cows and Its Use in Disease Treatments.” Agricultural Economics. Elsevier Science
Publishers B.V. 5 (1991) pp.385-399.

Pritchard, D. “Make Sure You Don’t “Buy” Mastitis.” Hoard 's Dairyman. March 10,
2000.p.204.

Quaife, T. “Don’t Blame the Big Guys for Higher Heifer Prices.” Dairy Herd
Management. January 2002. pp. 26-28.

Radke, B. and J. Lloyd. “Sixteen Dairy Culling and Replacement Myths.” Food Animal
Compendium. February 2000.

Renkema, J. and J. Stelwagen. “Economic Evaluation of Replacement Rates in Dairy
Herds 1. Reduction in Replacement Rates Through Improved Health.” Livestock
Production Science. Vol. 6. 1979. pp. 15-27.

Ruegg, P., A. Fabellar, and R. Hintz. “Effect of the Use of Bovine Somatotropin on
Culling Practices in Thirty-two Dairy Herds in Indiana, Michigan and Ohio.” Journal of
Dairy Science. Vol. 81. 1998. pp. 1262-1266.

Roenfeldt, S. “Consider Slip-form Concrete.” Dairy Herd Management. January 2001.
pp. 62-66.

Rogers, G., J. Van Arendonk, and B. McDaniel. “Inﬂuence of Production and Prices on
Optimal Culling Rates and Annualized Net Revenue.” Journal of Dairy Science. Vol. 71.
1988. pp. 3453-3462.

SAS. “SAS Version 8 On Line Doc Version Eight.” V8doc.sas.com/sashtml. 2002.

199

Schrick, F ., M. Hockett, A. Saxton, M. Lewis, H. Dowlen, and S. Oliver. “Inﬂuence of
Subclinical Mastitis During Early Lactation on Reproductive Parameters.” Journal of
Dairy Science. Vol. 84. 2001. pp. 1407-1412.

Stabel, J. “Johne’s Disease: A Hidden Threat.” Journal of Dairy Science. Volume 81.
1998. pp.283-288.

Stott, A. “The Economic Advantage of Longevity in the Dairy Cow.” Journal of
Agricultural Economics. Volume 45. 1994. pp. 113-122.

Sutton, F. “Locomotion Scoring of Dairy Cattle.” Conference Proceedings: 6’h Annual
Hoof Care Seminar. December 11, 2002.

Van Arendonk, J. “Studies on the Replacement Policies in Dairy Cattle: 1. Evaluation of
Techniques to Determine the Optimum Time for Replacement for Replacement and to
Rank Cows on Future Proﬁtability.” Zeitschrift fur T ierzuchtung und Zuchtungsbiologie.
Vol. 101. 1984. pp. 330 — 340.

Van Arendonk, J. “Studies on the Replacement Policies in Dairy Cattle: Summary.”
Thesis of Johan va Arendonk. Department of Animal Breeding. Agricultural University.
Wageningen, Netherlands. 1984. pp. 115 - 118.

Van Arendonk, J. and A. J. Dijkhuizen. “Studies on the Replacement Policies in Dairy
Cattle. III. Inﬂuence of Variation in Reproduction and Production.” Seminar Proceedings.
University of Florida. February 1986.

Wallace, D. “Test Your Footbath — Does It Control Hairy Heel Warts?” Hoard ’s
Dairyman.. May 10, 2000. p. 353.

Wallace, R. “Hairy Heel Warts: Fads and Fashions Conference Proceedings: 6th
Annual Hoof Care Seminar. December 11, 2002.

Warnick, L., K. Pelzer, A. Meadows, K. diLorenzo, and W. Whittier. “The Relationship
of Clinical Lameness with Days in Milk, Lactation Number, and Milk Production in a
Sample of Virginia Dairy Herds. Journal of Dairy Science. Vol. 78. 1995. p. 169.

Weigel, K. and R. Palmer. “Will Your High Producing Cow Be An Early Cull?”
Hoard ’s Dairyman. June 2003. p. 414.

Weigler, B., D. Hird, W. Sischo, J. Holmes, C. Danaye-Elmi, C. Palmer, and W.
Utterback. “Veterinary and Nonveterinary Costs of Disease in 29 California Dairies
Participating in the National Animal Health Monitoring System from 1988 to 1989.”
Journal of the Applied Veterinary Medicine Association. Vol. 196. No. 12. June 15,
1990. pp. 1945 — 1949.

200

Wells, S. “Biosecurity on Dairy Operations: Hazzards and Risks.” Journal of Dairy
Science. Vol. 83. 1998. pp. 2380 - 2386.

Wells, 8., S. Ott, and ‘A. Hillberg Seitzinger. “Key Health Issues — New and Old.”
Journal of Dairy Science. Vol. 81. 1998. pp. 3029-3035.

201

   

111111111111111«11111