Citation
Value of the Dairy Herd Improvement (DHI) hot list as a dairy management tool

Material Information

Title:
Value of the Dairy Herd Improvement (DHI) hot list as a dairy management tool
Creator:
Belsito, Jessica Elizabeth ( Dissertant )
Natzke, Roger P. ( Thesis advisor )
Vries, Albert de ( Reviewer )
Place, Nick ( Reviewer )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Copyright Date:
2005
Language:
English
Physical Description:
xi, 61 p.

Subjects

Subjects / Keywords:
Bacteria ( jstor )
Herds ( jstor )
Infections ( jstor )
Lactation ( jstor )
Mastitis ( jstor )
Milk ( jstor )
Milking ( jstor )
Somatic cells ( jstor )
Teats ( jstor )
Udders ( jstor )
Animal Sciences thesis, M.S ( local )
Dissertations, Academic -- UF -- Animal Sciences ( local )
City of Gainesville ( local )

Notes

Abstract:
Mastitis, an inflammation of the udder, is the most costly disease in the dairy industry. Somatic cells (which are mainly white blood cells) increase during an infection. For this reason, and because bacteriological procedures are time consuming and expensive, the somatic cell counts (SCC) are used as a general indicator of udder health. Currently, Dairy Herd Improvement (DHI) offers milk sampling to determine individual cow SCC. Many dairy producers use these SCCs to make management decisions such as culling, treatment or early dry-off. However, it is not clear what the value is of these SCC to base these management decisions on. The objectives of the following studies were to analyze the accuracy of the DHI labs in Florida in determining SCC, observe more closely milking to milking SCC variation and to use DHI records to determine the value of the DHI hot list, which ranks the highest SCC cows in the herd each test day. Chapter 3 was completed to assess the accuracy of our lab. Milk samples were analyzed in duplicates by electronic somatic cell counting. The variance of the mean differences between a sample and its duplicate was not statistically significant between the labs (P=.098). In addition to this, variation from the standard was not significant (P=.387). These two factors indicate that only a small portion of the milking-to-milking variation can be attributed to the sampling and cell counting procedure. In Chapter 4, 390 cows were sampled for 15 consecutive milkings. Hot lists were created for each milking and each day to evaluate the repeatability of cows on the list. The repeatability was also very low. When individual cow SCC was compared to the standard deviation for all 15 milkings, we saw standard deviation increase with average SCC. An additional study was completed by Southeast DHI (Gainesville, FL) over a period of five milkings so there would be a comparison for the first study. Many of the results in the two studies in Chapter 2 were similar. Chapter 5 contains a statistical analysis of lactation records from 1998 until 2003 for two Florida dairy herds. The objective of the study was to determine the value of the use of the hot list to reduce the bulk tank SCC over time. The hot list identifies the 20 cows in the herd that contribute the most cells to the bulk tank. Very few cows repeated on the hot list month after month. In addition to this, statistical analysis of daily bulk tank SCC showed no positive effect from hot list use. Therefore, if the SCC of many cows decreases without intervention (based on the results from the hot list analysis), one must question the economic value of intervening. We concluded that SCCs are highly variable. These data indicate that the hot list may not be as useful as previously thought and dairy producers should evaluate and utilize other methods to manage their bulk tank SCC.
Subject:
cell, count, DHI, hot, list, somatic, variability
General Note:
Title from title page of source document.
General Note:
Document formatted into pages; contains 72 pages.
General Note:
Includes vita.
Thesis:
Thesis (M.S.)--University of Florida, 2005.
Bibliography:
Includes bibliographical references.
General Note:
Text (Electronic thesis) in PDF format.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Belsito, Jessica Elizabeth. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
7/30/2007

Downloads

This item has the following downloads:


Full Text












VALUE OF THE DAIRY HERD IMPROVEMENT (DHI) HOT LIST AS A DAIRY
MANAGEMENT TOOL
















By

JESSICA ELIZABETH BELSITO


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Jessica Elizabeth Belsito
































This document is dedicated to the greatest man I never met, Robert John Belsito (1925-
1973) and to the strongest woman I ever knew, Marcia Walsh (1920-2004).















ACKNOWLEDGMENTS

Gratefulness is extended to my supervisory committee, Dr. Roger Natzke, Dr

Albert deVries and Dr Nick Place, for their time and effort.

Many thanks are given to Dave Bray, for taking me under his wing and giving me a

wealth of knowledge about the more practical side of the dairy industry.

I Thank my fellow students Bruno Amaral and his wife Michelle, Ben Butler, Liz

Johnson, Christy Bratcher, Nathan Krueger and Wimberly Krueger. Lastly, I thank Jamie

Foster for being a great Sigma Alpha sister.

I thank my officemates Kelly Jimenez and Tiffany Herrera, for the company and

camaraderie. I would be remiss if I did not thank Kasey Moyes, for inspiring me to go to

graduate school in the first place and being a sincere friend.

I also wish to thank my parents for their support and love throughout this process.

It was their guidance and the example they set forth which allowed me to become a

successful person. I wish to extend my love to the rest of my family, Kerri, Bobby, Trish,

Cassandra, Becky, Jerimy, Tatum, Colby, and the little one who is on her way.

Lastly, I wish to thank Sergio, who has been understanding, patient, supportive and

has made my life complete.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TA BLE S ....................................................... .. .......... ............ .. vii

L IST O F FIG U R E S .............. ............................ ............. ........... ... ........ viii

A B STR A C T ................................................. ..................................... .. x

CHAPTER

1 IN TRODU CTION ................................................. ...... .................

2 LITERATURE REVIEW ........................................................................4

B ack g rou n d ................................................................................................. . 4
P athogenesis ................................................................. 5
M ode of Infection .................. ..................................... .. ........ .... 5
Im m u nity .................................................................... 6
P a th o g e n s .................. ... .................................................................. ............. 9
M astitis Prevention and Treatm ent.........................................................................9
E co n o m ics ................................................. ................................15
Milk Composition and Quality .............. .............................................. .......17
Milk Culturing and Somatic Cell Counting Techniques .........................................18
Factors Affecting Somatic Cell Counts .......................................................19
V ariability in Som atic C ell C ounts....................................... ................................2 1

3 FLORIDA MILK QUALITY LAB ANALYSIS ................................................. 23

Intro du action ...................................... ................................................ 2 3
M materials and M methods ....................................................................... ..................24
R results and D iscu ssion .............................. ......................... ... ........ .... ............24

4 VARIATIONS IN SOMATIC CELL COUNTS FROM MILKING TO MILKING 29

In tro d u ctio n ........................................................................................................... 2 9
M materials and M methods ....................................................................... ..................30
R results and D discussion ..................................... ......... ...... .. ............... 1




v









5 ANALYSIS OF THE VALUE OF HOT LIST USE ..............................................40

M materials an d M eth od s ...................................................................... ....................42
R results and D iscu ssion ............................. ...... .................. .. .......... ...... ...........43

6 GENERAL DISCUSSION AND CONCLUSIONS..............................55

L IST O F R E F E R E N C E S ...................... .. ............. .. ....................................................58

B IO G R A PH IC A L SK E T C H ...................................................................... ..................61
















LIST OF TABLES


Table page

3-1 Lab Analysis of Official Samples. ........................................ ....................... 28

4-1 Raw and adjusted (for milk production) SCC correlations for consecutive
m ilk in g s .......................................................................... 3 6

4-2 Raw and adjusted (for milk production) SCC correlations for morning, afternoon
and evening m ilkings. ................................... .. .. .. ... ................... 37

5-1 Number of times cows were on hot lists for study one and study two...................50

5-2 Results of the Proc Mixed Procedure on daily bulk tank data. .............................52

5-3 Results of the LS means procedure on daily bulk tank data. ..................................53
















LIST OF FIGURES


Figure page

3-1 SCC for Samples and their duplicates analyzed at lab 1.......................................25

3-2 SCC for Samples and their duplicates analyzed at lab 2 .......................................26

3-3 SCC for Samples and their duplicates analyzed at Lab 1 and Lab 2......................27

4-1 Average SCC for the 15 milkings on an individual cow basis..............................31

4-2 Number of cows over or under certain somatic cell counts at all 15 milkings .......32

4-3 Number of cows over or under certain daily somatic cell counts each day ............32

4-4 Number of cows with a daily SCC over one million ..........................................34

4-5 Number of cows with a daily SCC over 750,000 cells/ml ......................................34

4-6 Number of cows with a daily SCC over 500,000 cells/ml ......................................35

4-7 Number of cows with a daily SCC over 250,000 cells/ml ......................................35

4-8 Individual cow standard deviation versus her average SCC .................................38

4-9 Individual cow coefficient of variation versus her average SCC .............................38

5-1 Exam ple of a m monthly D H I hot list. ............................................... .....................41

5-2 Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year Period
with Hot List Cows Removed for Dairy A. ................................. .................44

5-3 Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year Period
with Hot List Cows Removed for Dairy B.................................... .................45

5-4 Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year Period
with Hot List Cows in the Previous Month Excluded for Dairy A........................46

5-5 Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year Period
with Hot List Cows in the Previous Month Excluded for Dairy B ........................46









5-6 Frequency of number of appearances on the hot list..............................................48

5-7 Frequency of number of appearances on the hot list..............................................49















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

VALUE OF THE DAIRY HERD IMPROVEMENT (DHI) HOT LIST AS A DAIRY
MANAGEMENT TOOL

By

Jessica Elizabeth Belsito

August 2005

Chair: Roger P Natzke
Major Department: Animal Sciences

Mastitis, an inflammation of the udder, is the most costly disease in the dairy

industry. Somatic cells (which are mainly white blood cells) increase during an infection.

For this reason, and because bacteriological procedures are time consuming and

expensive, the somatic cell counts (SCC) are used as a general indicator of udder health.

Currently, Dairy Herd Improvement (DHI) offers milk sampling to determine individual

cow SCC. Many dairy producers use these SCCs to make management decisions such as

culling, treatment or early dry-off. However, it is not clear what the value is of these SCC

to base these management decisions on. The objectives of the following studies were to

analyze the accuracy of the DHI labs in Florida in determining SCC, observe more

closely milking to milking SCC variation and to use DHI records to determine the value

of the DHI hot list, which ranks the highest SCC cows in the herd each test day.

Chapter 3 was completed to assess the accuracy of our lab. Milk samples were

analyzed in duplicates by electronic somatic cell counting. The variance of the mean









differences between a sample and its duplicate was not statistically significant between

the labs (P=.098). In addition to this, variation from the standard was not significant

(P=.387). These two factors indicate that only a small portion of the milking-to-milking

variation can be attributed to the sampling and cell counting procedure.

In Chapter 4, 390 cows were sampled for 15 consecutive milkings. Hot lists were

created for each milking and each day to evaluate the repeatability of cows on the list.

The repeatability was also very low. When individual cow SCC was compared to the

standard deviation for all 15 milkings, we saw standard deviation increase with average

SCC. An additional study was completed by Southeast DHI (Gainesville, FL) over a

period of five milkings so there would be a comparison for the first study. Many of the

results in the two studies in Chapter 2 were similar.

Chapter 5 contains a statistical analysis of lactation records from 1998 until 2003

for two Florida dairy herds. The objective of the study was to determine the value of the

use of the hot list to reduce the bulk tank SCC over time. The hot list identifies the 20

cows in the herd that contribute the most cells to the bulk tank. Very few cows repeated

on the hot list month after month. In addition to this, statistical analysis of daily bulk

tank SCC showed no positive effect from hot list use. Therefore, if the SCC of many

cows decreases without intervention (based on the results from the hot list analysis), one

must question the economic value of intervening.

We concluded that SCCs are highly variable. These data indicate that the hot list

may not be as useful as previously thought and dairy producers should evaluate and

utilize other methods to manage their bulk tank SCC.














CHAPTER 1
INTRODUCTION

Quality food products are always a concern in the United States for producers,

consumers and the government alike. One way milk quality is measured is through the

counting of somatic cells. The word somatic means of the body. Therefore, it follows

that somatic cells in milk are any kind of cells from the body other than milk. Somatic

cells could possibly be white blood cells, epithelial cells or other types of cells. There are

many different reasons for keeping somatic cell counts (SCC) low. Producers, consumers

and other groups each have their own reason for wanting low SCC.

Producers are concerned about quality. A recent study (Ma et al., 2000) concluded

that lower SCC leads to an increase in the shelf life of milk. Producing a product that

lasts longer is economically beneficial to all consumers. Some consumers may also be

concerned about the health of the animal producing the milk and the cleanliness of farms

(both of these factors can affect SCC)

The National Mastitis Council (NMC), a well known organization that focuses on

mastitis and milk quality, has been pushing the US government to lower the legal limit of

somatic cells from 750,000 cells/ml to 400,000 (NMC, 2003). The reason is that many

countries in other areas of the world have lowered their legal limit to 400,000 and some

people believe that for the United States to successfully sell milk products on an

international level we need to lower our legal somatic cell limit. Clearly SCC is currently

a pressing issue for dairy producers.









One reason many dairy producers try to lower somatic cell counts is premiums.

Many milk buying organizations offer monetary premiums for lower SCC. Perhaps even

more important is the issue of mastitis (which is a costly disease). It is generally believed

that if a herd SCC is low less mastitis will be found on the farm. These are just some of

the reasons that the Dairy Herd Improvement Association (DHI), an association which

aids dairy producers in record keeping and data collection, has tried to develop and

implement tools to aid dairy producers in monitoring their SCC.

One of the first things that DHI did was to set up record keeping systems and labs

to evaluate the SCC of milk samples. Many herds across the US participate in DHI

monthly somatic cell count testing, where a DHI technician comes to their farm and takes

a milk sample from every cow. These samples are sent to DHI labs where they are

analyzed and the results are sent to the producer and stored in DHI databases. This

provides the dairy producer with a milk weight and SCC for each individual cow. The

goal of monthly somatic cell counts is to help producers identify individual cows with

high SCC. By identifying these cows, managers can then make a management decision

(culling, antibiotic therapy, etc.) to try to lower the SCC if necessary. To help managers

identify these cows more efficiently, DHI provides a "hot list" which ranks the top

twenty cows in the herd by total cells contributed to the bulk tank (the bulk tank SCC

refers to the average somatic cell count of the entire herd as the bulk tank is where all the

milk is stored, kept cool and agitated). This "hot list," in theory, aids dairy producers in

quickly identifying problem cows based on their SCC.

The literature lacks peer-reviewed articles on milking to milking SCC variability

and hot list analysis. Because of the problems outlined above and because of the lack of









literature on the subjects, the goal of this thesis was to observe and describe patterns

and/or changes in individual cow SCC over time and to analyze the effectiveness of the

hot list in managing bulk tank SCC.

Through a combination of these studies we were able to observe variability in SCC,

the value of the hot list to the dairy producer and the effect of the hot list on the bulk tank

SCC over time. Our hypothesis for the study was that the labs would be accurate and the

milking to milking somatic cell counts would show slight variability. It was also thought

that if the hot list was improved by adding other important information it would then

become a more useful tool for dairy producers.














CHAPTER 2
LITERATURE REVIEW

Mastitis is one of the most costly diseases to milk producers. Losses are

associated with decreased milk production, discarding milk from cows treated with

antibiotics, the cost of the antibiotics and increased labor expenses (Ravinderpal et al.,

1990). In addition to this, milk quality can be negatively affected by high somatic cell

counts, which indicate the presence of infection (Ma et al., 2000). It is therefore the

purpose of this literature review to describe the events leading to mastitis, economic

losses due to infection, methods used to detect mastitis, variability in SCC and milk

quality and compositional changes.

Background

The National Mastitis Council defines mastitis as an "inflammation of the udder,

most commonly caused by infecting microorganisms." The word mastitis itself means

inflammation of the mammary gland. Giesecke (1975) gives a much more specific

definition of mastitis. His definition states that mastitis is indeed an infection of the

mammary gland. First, there is damage to the mammary epithelium. An inflammatory

reaction (either clinical or subclinical) follows. Depending on the severity of the

infection, local or widespread changes can take place in the animal.

Giesecke adds to his definition the distinction between subclinical and clinical

mastitis. Subclinical mastitis involves subtle physiological changes (Giesecke, 1975).

Clinical mastitis, however, is a more serious inflammatory response. Signs of clinical









mastitis include: changes in milk composition (watery milk, flakes or clots), visual signs

of infection (red, swollen quarters) and a more dramatic increase in SCC (Neave, 1975).

Pathogenesis

Mode of Infection

Mastitis begins at the teat end. The bacteria must first come in contact with the

sphincter. Once the bacteria pass through the sphincter it must travel up the teat canal

which is lined with keratin. Keratin is a wax-like substance designed for trapping

bacteria and also contains substances with antimicrobial properties (Sordillo et al., 1997).

If the bacteria pass by the keratin in the teat canal it can gain access to the

mammary gland. There are many ways in which bacteria may be introduced into the

mammary gland. Pressure on the teat caused by cow movement may aid the bacteria in

traveling up the teat canal. In addition to this, bacteria may multiply continually until

they have reached milk producing tissues, allowing them to establish themselves more

permanently (Kehrli and Shuster, 1994).

The milking machine also aids the bacteria in gaining access into the teat. The

exposure of the teat to the vacuum (usually between 12-17 Hg) can cause damage to the

skin on the teat. Even more damage is caused when the teat cup liners do not collapse

completely around the cow's teat. When the liner is open, the pressure inside the teat is

greater than the ambient pressure due to the vacuum, allowing the milk to be expelled.

When the liner collapses and the milk is no longer being expelled, milk from other

quarters may splash back and infect uninfected quarters. Large vacuum fluctuations also

can affect the new infection rate. There are many causes for these fluctuations. All

milking machines have vacuum fluctuations because of the pulsation cycle, or the cycle

in which the vacuum is turned on and off to open and close the teat canal. Pulsation is









used to avoid exposing the teat to a constant vacuum. It allows the teat canal to be open

and closed rhythmically and prevents the congestion of blood in the teat. These

fluctuations are caused on an irregular basis because of air that may get into the vacuum

system via the inflations. Additionally, the movement of milk causes vacuum fluctuation

(Kingwill et al., 1979).

Data on these fluctuations shows that when teats are contaminated with mastitis

causing pathogens the rate of new infection is higher when exaggerated vacuum

fluctuations are applied to the cow's teats. In addition to this, contaminated milk coming

in contact with the cow's teat could be the cause of a new infection. When both causes of

vacuum fluctuation occur at the same time the rates of new infection are at their highest

(Kingwill et al., 1979).

Bacteria that are able to reach milk producing tissues, or alveoli, now must find

ways to stay in the tissue. Some bacteria, such as Streptococcus agalactiae and

Staphylococcus aureus are well adapted to attach themselves to the epithelial lining of the

alveoli. This prevents the bacteria from being flushed out during the milking process.

Other types of bacteria do not adhere as well to tissue. These bacteria have adapted other

ways of staying in the udder, such as extremely fast multiplication (Kehrli and Shuster,

1994).

Immunity

Once in the gland, there are two main ways in which the cow's immune system will

try to clear the bacteria. The two basic groups of immunity in the cow's udder are innate

and specific immunity. Innate immunity is more nonspecific and responds to the

bacterial challenges almost immediately. Innate immunity involves macrophages,

neutrophils, polymorphonucleur neutrophils (PMN), and natural killer cells. Specific









immunity requires time to develop after the initial infection. This response involves

antibody molecules, macrophages and lymphoids. These cells attack only the type of

bacteria they are programmed to eliminate (Sordillo et al., 1997).

If the leukocytes in the udder are not able to eradicate the bacteria, inflammation

and infection will result. The milk in the gland provides an optimal environment for

bacteria multiplication. If the infection becomes severe enough, the cow may show

systemic signs of infection including a red, hard udder and a fever. Inflammation is

caused by the flood of cells involved in the immune response into the udder (Paape et al.,

1979).

Bacteria cells that adhere and multiply in the udder can cause a variety of negative

effects. Perhaps the worst is the formation of scar tissue in the udder. This is a result of

bacteria adhering themselves to the lining of the alveoli (Sordillo et al., 1997). Bacteria

are not the only organisms that can cause scar tissue. The movement and adhesion of

PMNs to milk producing tissues may damage the tissue, form scar tissue, and lead to an

increase in milk SCC (Harmon, 1994). This scar tissue, or fibrosis, inhibits milk

production in the current lactation as well as in subsequent lactations. In addition to this,

bacteria have the ability to produce toxins and other substances that can kill milk

producing cells, again reducing milk production. Membrane permeability is also

affected, allowing more blood components into the milk. If the infection becomes severe

enough, visible changes may occur in the cow's milk (Kehrli and Shuster, 1994).

In addition to the effects already listed, mastitis pathogens can cause healthy tissue

to involute, or go into a state of resting. Usually, it is not until the next lactation, if ever,









that this tissue begins producing milk again. The blood and immune system components

can also cause secretary ducts to become blocked (Sordillo et al., 1997).

PMNs are one of the first responders in the non-specific immune response. PMNs

are phagocytic cells that engulf and digest bacteria. Another non-specific immune

response involves neutrophils. They account for greater than 90% of leukocytes in the

mammary gland. Neutrophils are bactericidal and travel from the blood to the mammary

gland after substances such as cytokines (an inflammatory mediator) are present in the

blood. Neutrophils also have small antibacterial peptides (Sordillo et al., 1997).

Macrophages, cells which are present in healthy mammary glands, are also

phagocytic. These cells also engulf and digest foreign bacteria. Neutrophils and

macrophages both destroy milk solids, making them less desirable than blood leukocytes

(Sordillo et al., 1997).

The specific immune response is more complex. Lymphocytes are the only

immune cells with the ability to identify foreign bacteria with specific receptors on their

membranes. The other cells involved in specific immunity are antigen-presenting cells

(Sordillo et al., 1997).

There are two types of lymphocytes, T and B. T lymphocytes produce and secrete

cytokines. There is much speculation surrounding the function of T lymphocytes. What

is known, however, is they play a role in protecting epithelial surfaces. They may also

mediate the activity of other cells involved in the immune system. B lymphocytes

produce antibodies specific to the bacteria that are invading the udder. B lymphocytes

present cells to T lymphocytes to be eliminated (Sordillo et al., 1997).









Pathogens

There are a variety of pathogens that cause mastitis. The most common pathogens

include Staphylococcus aureus, Streptococcus agalactiae, coliforms, and other

streptococci. These species are considered to be the major mastitis pathogens because

they are the most common. Other noteworthy bacteria include Pseudomonas,

Corynebacterium bovis, Mycoplasma and coagulase-negative staphylococci. In addition

to being major pathogens, Staphylococcus aureus and Streptococcus agalactiae are also

contagious. These bacteria can be spread to other cows through unsanitary milking

procedures (Harmon, 1994).

Environmental pathogens also cause acute clinical mastitis. These infections tend

to affect the cow more quickly and severely than other mastitis pathogens. The duration

of these infections also tends to be shorter. These pathogens are found in various places

in the cow's environment. Most of these infections are caused by cows laying in manure

or other types of unclean bedding, mud, or standing in unclean ponds or cooling ponds.

The coliforms are Gram-negative and make up the largest group of environmental

pathogens they include: Escheria coli, Klebsiella spp., Streptococcus bovis, Enterococcus

faecium, Enterobacter spp., Citrobacter spp., Streptococcus dysgalactiae, Streptococcus

uberis and Enterococcusfaecalis (Harmon, 1994).

Mastitis Prevention and Treatment

To avoid losing money because of mastitis, prevention is the key. The rate of

clinical mastitis varies greatly from farm to farm. In 62 citations from 1982 to 1996 the

incidence of clinical mastitis ranged from less than 2.5% to over 50% (Kelton et al.,

1998). There are many ways in which dairy producers attempt to lower the incidence of

mastitis on their farm. These include pre and post dipping of teats with a disinfectant









using antibiotics at dry-off to prevent future infections, preventing the transmission of

bacteria from one cow to another and keeping freestalls clean (Oliver et al., 1993).

An additional study evaluating the efficacy of pre-dipping was completed in

Tennessee in 1993. This study was a natural exposure study following the procedures set

by the National Mastitis Council. The active ingredients in the teat dip were sodium

chlorite (0.64%) and lactic acid (2.64%). The study was conducted over a period of

fifteen months and the average number of cows enrolled in the study at any given time

was 175. There was a total of 423 cows involved in the study. Cows were milked twice

daily and housed in a freestall with sawdust bedding (Oliver et al., 1993).

The study was a negative control using a split udder design. The left teats were the

controls. They were stripped and dried. Both right teats were forestripped, dipped and

wiped. Duplicate samples of foremilk were collected aseptically from all quarters

monthly, twice within ten days after parturition, and when clinical mastitis was observed

(Oliver et al., 1993).

New intramammary infections caused by Staph aureus (P < 0.05) and Strep species

(P < 0.025) were both lower in teats that were treated with the predip. In addition to this,

the total number of quarters with new infections caused by other major pathogens was

also lower in the treatment group (P < 0.01) (Oliver et al., 1993).

Besides teat dipping, another method that can help control mastitis is antibiotic use.

In 1966, Smith wanted to evaluate the effectiveness of intramammary antibiotic use at the

start of the dry period. They had observed previously that half of the cows in most

dairies (excluding first calf heifers) calved with one quarter infected. Smith believed that









eliminating late lactation infections as well as preventing new infections during the dry

period would help to decrease the number of infections present at calving.

Smith used 900 cows from 36 herds. Cows were chosen by duplicate

bacteriological milk samples taken in the last week of the cow's lactation. Cows were

assigned randomly at dry-off into three groups 1) control (no disinfectant no antibiotics;

2) .2g Cloxicillin intramammary and dipped with 5% hypochlorite; 3) intramammary

infusion of 1 g Cloxicillin and teats dipped with the 5% hypochlorite solution.

When the cows were dried off, half of them had positive bacteriological milk

samples. At calving, bacteriological samples were taken within 7 days. Over 60% of the

control group had positive bacteriological milk samples compared to only 23% of group

2 and 15% of group three. This indicates that the use of a 1 gram dose of Cloxicillin is

advantageous when given at dry off (Smith et al., 1966).

Perhaps the most complete assessment of advantageous mastitis prevention

practices is a study which was completed by England's National Institute for Research in

Dairying in 1975. This study included thirty herds over a three year period. The herds

were divided randomly and assigned to one of two treatment groups. The first group

employed several mastitis prevention strategies. They utilized post-dipping, antibiotic

therapy for infected lactating cows and antibiotic therapy for all cows at dry-off. The

second group used only teat dip hypochloritee containing 4% available chlorine). In

addition to these factors, milking machines were evaluated annually at each herd and

adjusted as necessary. The researchers visited each herd weekly to assess the implication

of the mastitis prevention strategy and to take milk samples for culture (Wilson and

Kingwell, 1975).









At the beginning of the study a "blitz" therapy was utilized. All infected quarters

in half the herds in each group were treated with an antibiotic. This was done to evaluate

the initial reduction in the number of subclinical cows (Wilson and Kingwell, 1975).

Percent of cows infected during lactation and percent of quarters infected during

lactation both dropped steadily each year. The average reduction of udder disease in all

herds was about 70%. The blitz therapy eliminated 68% of all infections in the cows

which received it. However, a larger number of subclinicals were cured by the use of

antibiotics at dry off in the non-blitzed herds. At the end of the first year the percentages

of infection were relatively similar in the blitzed and non-blitzed herds (Wilson, 1975).

The reduced rate of infection was shown in the mean bulk tank SCC, which fell from

680,000 to 310,000 cells/ml over the three year period. Treatment of clinical cases of

mastitis fell by about one half over the three year period (Asby et al., 1975).

The new infection rate of each pathogen declined each year except for Coliforms

and others. The infection rates of these two groups remained relatively the same over the

three years. Streptococcus agalactiae was virtually eliminated from all herds (Wilson

and Kingwell, 1975).

In addition to teat dipping and use of antibiotics, nutrition may also play a small

role in mastitis management. In 1991 Oldham et al. conducted a study looking at the

effects of vitamin A and P-carotene during the dry period and early lactation on udder

health. This study had three treatment groups. The first received 50,000 IU of vitamin

A, the second 170,000 IU of vitamin A and the third 50,000 IU of vitamin A plus 300 mg

of P-carotene (per cow per day). This group of 82 Holsteins was supplemented starting

two weeks before dry off until 6 weeks post calving. Blood samples were taken to









measure the serum concentrations of vitamin A and P-carotene 2 weeks before dry off, at

dry off, the 28th day of the dry period, once between day 10 to 4 pre-calving, within 24

hours of calving, once during week 3 of lactation and once during the sixth week of

lactation. Milk was also sampled for culture. All samples were in duplicate. They were

collected at dry off, the 28th day of the dry period, once between day 10 to 4 before

calving, within 24 hours of calving and once each week during week three and six of

lactation. After culturing, the remainder of milk was used to determine somatic cell

count. Quarters with clinical mastitis were sampled also (Oldham et al., 1991).

Oldham found no difference in serum vitamin A levels between treatments

throughout his study. Treatment effect on serum p-carotene showed a tendency towards

being significant (p<.08). Both serum p-carotene levels and vitamin A levels tended to

drop from 2 weeks prior to dry off until calving (Oldham et al., 1991).

To determine if P-carotene and/or vitamin A were effective in treating mastitis

frequency of new intramammary infections (IMI), SCC, and frequency of clinical

mastitis were used. It was observed by these authors that treatment had no effect on all

three of these parameters. It was also noted that treatment had no effect on which

pathogens were the cause of the mastitis (Oldham et al., 1991).

Vitamin E and selenium have both been shown to have positive effects on udder

health when supplemented alone and together. In Ohio, nine dairy herds were used to

evaluate the effects of supplementing vitamin E and selenium. Bulk tank SCC score was

measured as well as clinical mastitis rate. Serum levels were also observed. The herds

involved with the study were monitored for one year. Strep agalactiae and Staph aureus

mastitis had to be well controlled for a herd to be enrolled in the study. Rations were









sampled for analysis three times during the year. Plasma samples were also drawn from

10 cows in each herd three times during the year. Samples were taken from cows who

were 60 days pre-calving up until 60 days post-calving. Duplicate quarter samples were

taken for bacteriological culture from all quarters of all cows that were diagnosed with

clinical mastitis before treatment was given. Bulk tank samples were also taken weekly

(Weiss et al., 1990).

Increasing the amount of vitamin E and selenium in the diet raised blood serum

levels. Blood serum levels for selenium, however, were independent of supplementation

once intake exceeded 5 mg/day. It was also observed in this study that increased levels

of blood serum selenium had a highly significant positive effect on bulk tank SCC (p <

.005). In addition to this, cows fed a higher diet of vitamin E tended to have less clinical

mastitis (p <0.1) (Weiss et al., 1990). It must be noted, however, that approximately two

thirds of the soil in the United States is selenium deficient. If the animals were selenium

deficient at the start of the study, the results may be biased (Hogan et al., 1993).

Researchers in Ohio found that 24 hours after an intramammary injection with E.

coli plasma vitamin C concentration decreased by 39% (p < 0.01). Vitamin C and

ascorbic acid concentrations in milk from the infected quarters decreased by 52 and 62%

respectively. Unchallenged quarters were unaffected. Twenty-one Holsteins were used

in this study. Blood and milk was sampled before challenge, 24 hours after the challenge

and 7 days after challenge to be analyzed for vitamin C concentration. Because vitamin

C concentrations decreased so dramatically following the E. coli challenge, these results

indicate that vitamin C may play an important role in helping cows recover from mastitis

as well as possibly help to prevent it (Weiss et al., 2004).









Economics

Mastitis is the most economically devastating disease facing dairy farmers. There

have been several publications on the subject. All agree that mastitis is very costly but,

because all the estimations use different methods and different numbers, the end results

can vary greatly. The following estimation is one of the more common found in the

literature. It is estimated that at a milk price of $12.07 per hundred weight, $184 dollars

are lost per cow in the herd per year (Blosser, 1979). This number agrees with

Ravinderpal (1990) who projected that mastitis accounted for 70 to 80% of the $140-

$300 dollars lost per cow per year due to disease. Thus, the annual loss per year due to

mastitis in the United States can easily exceed 2 billion dollars. The majority of this loss

(estimated to be about 2/3rds) is accounted for by reduced milk production. Other factors

include cost of treatment, discarded milk, and increased labor. It is projected that if a

quarter becomes infected the average milk loss for that quarter is anywhere from .34 to

2.66 kg per day (Janzen, 1970).

In 1983, McDermott et al. analyzed the economics of treating every cow with a

SCC above 400,000. This fifteen month study analyzed milk production and possible

profits resulting from treating early to obtain a higher response rate. Lactating dairy

cows were used from five commercial herds that all used teat dip and dry cow therapy.

Composite milk samples were taken monthly for electronic SCCs. The somatic cell

sampling was unannounced. Within seven days of the sampling individual quarter

samples were collected and the bacteriology was completed. Cows were assigned

randomly to the control or experimental group based on their ear tag number. In the

experimental group all cows were infused with a lactating antibiotic (cephapirin) the first

time their SCC rose above 400,000 cells/ml. This was done only once in the lactation.









Clinical cases in both the control and experimental groups were treated by the farm

managers themselves. Treatment was based on SCC. Bacteriology was completed to

assess any possible benefits of the antibiotic in eliminating infection.

Treating cows with SCC of 400,000 and above had no effect on milk production in

that lactation when compared to the control group. In addition to this, the cost of treating

cows in this manner translated into almost $20 per cow. Most of this cost was incurred

because 49 false positive cows were treated. Included in the $20 were SCC survey, labor,

discarded milk and drugs (McDermott et al., 1983).

The study completed in 1975 at the University of Reading also included an

economic analysis. It was calculated that dairy producers who followed a basic routine

of post-dipping with an iodine teat dip, treating infected lactating cows with antibiotics

and treating all cows at dry-off with an antibiotic enjoyed a benefit/cost ratio of 2.55 to 1.

That is for every one dollar spent on mastitis prevention, the dairy producer should see a

$2.55 profit. The results of this study suggest that mastitis management procedures are

economically beneficial but should be used as part of the regular routine on the farm and

should be viewed as a long term application and not a quick fix for a mastitis problem

(Asby et al., 1975).

More recently, other mastits control and treatment techniques have been evaluated.

A study done in 1998 used computer modeling to test the effectiveness of several mastitis

control strategies. This study evaluated control strategies for the following bacteria

types: Streptococcus agalactiae, Streptococcus spp, Staphylococcus aureus, coagulase-

negative staphylococci, and Escheria coli. Prevention of mastitis (forestripping,









predipping and postdipping), vaccination for E. coli, lactation therapy and dry cow

therapy were all evaluated for economic efficiency (Allore and Erb, 1998).

Each strategy evaluated in the computer model was compared to a control herd to

find the annual benefit (dollars per cow per year). The criteria for budgets were

determined by changes in milk composition and production due to mastitis as well as the

number of cows that were culled for mastitis reasons. Using this criteria, it was

determined that prevention and dry cow therapy were beneficial for all pathogen groups.

In addition to this it was found that in herds where environmental mastitis was dominant,

vaccination for the prevention ofE. Coli is also beneficial (Allore and Erb, 1998).

Milk Composition and Quality

Elevated SCCs are not only accompanied by milk loss; compositional changes also

take place. Milk fat tends to show a slight decrease as does lactose (although some

studies show no change in fat content). The protein content of milk fluctuates very little.

The types of protein present, however, tend to change. Casein, the desirable milk protein

for making cheese, decreases. Whey, an undesirable protein for cheese making, tends to

increase. Blood components such as albumins and immunoglobulins find their way into

the milk because of an increase in permeability of the milk blood barrier (Harmon, 1994).

Mineral balance in the milk is also affected. Potassium and calcium are markedly

reduced in milk with a high SCC. Potassium leaks through damaged epithelial cells.

These same cells also allow sodium and chloride from the blood into the milk. This

increase of blood components in the milk also causes the pH of milk to rise to 6.9 or even

higher (Harmon, 1994).

Quality of milk is also affected in the presence of mastitis. In a study comparing

shelf life of high SCC milk (about 750,000 cells/mL) to low SCC milk (about 45,000









cells/ml) it was found that after pasteurization the high SCC milk had a much shorter

shelf life than the low SCC milk by as many as seven days. It is thought that there are

free fatty acids in high SCC milk that increase in number and cause an increase in casein

hydrolysis. After 21 days the organoleptic quality of the low SCC milk remained high.

The high SCC milk was rancid, as determined by a sensory panel (Ma et al., 2000).

Milk Culturing and Somatic Cell Counting Techniques

Milk samples are cultured for a variety of reasons. Culturing milk identifies which

organism or organisms the quarter is infected with. This can help in identifying the cause

of the mastitis and which drugs to use during the treatment process. The most widely

used culturing technique involves the use of blood agar plates. Aseptic milk samples are

plated (.01 ml of milk) on agar plates containing 5% sheep blood and .1% esculin.

Samples are then incubated at 37C for 48 hours. Usually the plates are read at 24 hours

of incubation and after the full 48 hours. The most effective culture results are obtained

from aseptic quarter milk samples (Dinsmore et al., 1992).

Somatic cells are mainly white blood cells. Because white blood cells often appear

in response to infection, SCC are used as a general indicator of udder health. There are

several methods in which SCC can be determined.

The most popular on farm test is the California Mastitis Test (CMT). This test

gives an estimation of the SCC for each quarter. The milk sample is mixed with a

solution that congeals in the presence of somatic cells. The more gel-like the sample

becomes the higher the SCC is thought to be. When this reaction occurs it is assumed

that the cow has mastitis in that quarter. Another test that is very similar to this is the

Wisconsin Mastitis Test (Kitchen, 1981).









In laboratories somatic cells can be determined either by a technician counting cells

in a small milk sample (DMSCC) or by an automated machine (ESCC). The DMSCC

has several recognized drawbacks. These include non-homogenous distribution of cells

in the sample, subjective decisions and human error (Kitchen, 1981).

There are two main ways to count somatic cells electronically. They are the use of

the Coulter Counter and the Fossomatic. The coulter counter uses particle size to identify

somatic cells and count them (Phipps, 1968). The Fossomatic uses fluorescence to count

cells. Once the milk is diluted with a buffer, ethidium bromide is added to make the

somatic cells fluorescent. They are then counted (Heeschen, 1975). The Bentley

Somacount (Bentley Instruments Inc, Chaska, MN), an instrument used presently, also

counts cells using fluorescence.

Factors Affecting Somatic Cell Counts

Somatic cell counts are highly variable and can fluctuate immensely in a short

period of time. There are many factors which affect SCC. The most important factors

are the infection status and the type of organism which causes the infection. SCC tends

to peak shortly after the cow has been challenged by a pathogen. Depending on the

organism present, this peak may occur hours or days after the initial challenge. In

addition to this, the level of the response varies from cow to cow. The time it takes for

SCC to return to normal can vary from days to months or never if the cow becomes

chronically infected (Harmon, 1994).

It is generally thought that older cows and cows in late lactation have higher

somatic cell counts. In many cases this is true because the cow has had more time to be

exposed to pathogens. However, if lactating cows remain uninfected throughout the

course of their lactation, SCC varies very little (Sheldrake et al., 1983).









Another important aspect of SCC and stage of lactation is fresh cows. After

calving, cows tend to have a very high SCC. If the cow does not become infected during

her fresh period, her SCC should decrease rapidly and return to normal within 35 days in

milk (Reneau, 1986).

There is also thought to be a "dilution" factor that affects SCC. This means that if

a cow were to experience a reduction in milk production and the SCC remained the same,

we would observe an increase in SCC. It has been suggested that this phenomenon is

responsible for the increase in SCC towards the end of lactation (Schultz, 1977). A

reduction in water or feed has also been shown to increase SCC, probably because of the

reduction in milk production (Martin, 1973).

Various sources of stress can also cause an increase in SCC (Dohoo and Meek,

1982). Heat stress has been shown to have dramatic effects on SCC (Elvinger et al.,

1991). Heat stress also causes a marked decrease in milk production; therefore, it is

unclear whether we can attribute a rise in SCC to the stress itself or a decrease in

production, which is thought to cause the "dilution factor."

More recently, researchers in Minnesota utilized DHI records to evaluate somatic

cell variation. DHI records with at least 220 days in milk and at least 4 test days were

used. Provisions were added into the data analysis to ensure that records with

questionable data were not utilized. SCC was transformed into a somatic cell score

(SCS) using a base two log scale. Scores range from 0 to 9. Lactation records were

separated by parity and by breed (Schutz et al., 1990).

The mean SCS increased with age, this finding agrees with previous research

(Emanuelsson and Persson, 1984 and Miller et al., 1983). The general trend observed in









this study was a high SCS at calving which steadily decreased until peak milk production.

After peak milk production, SCS tended to increase throughout the remainder of the

lactation. The authors admit, however, that some of this increase in SCS could be due to

the "dilution factor" associated with a decrease in milk production (Schutz et al., 1990).

Variability in Somatic Cell Counts

Individual cow somatic cell counts have been used as a management tool for many

years, mainly due to the relatively cheap availability of this data due to DHI (Reneau,

1986). However, it is suspected that individual cow somatic cell counts can be highly

variably due to sampling errors or more within cow variation than previously thought.

In 2004, a study was conducted in the United Kingdom which attempted to observe

the variation in individual cow somatic cell counts. This study utilized aseptic milk

samples as well as daily SCC from the morning milking to determine SCC variations.

The results implied that monthly individual cow SCC would not reflect all of the

infections that affected the herd in that month. However, the most interesting observation

is between the within cow variance for uninfected and infected cows. The variance for

both groups was about 0.5. This is somewhat of a surprise, as most would expect the

variance of uninfected cows to be lower because it is generally thought that uninfected

cows do not experience spikes in their SCC. This study suggests that to accurately assess

the mastitis status of a cow multiple somatic cell counts from a short period of time are

more accurate (Berry et al., 2004).

In 1972, researchers in Canada observed variations of somatic cells throughout

lactation on a weekly basis. Eleven Holsteins were used and samples were collected on a

weekly basis using DHI approved sample collection jars. The sample time alternated

between morning and evening milkings. Differential somatic cell counts were









performed. Aseptic milk samples were also taken for bacteriological cultures when

visual evidence of inflammation was observed or when the SCC doubled for any one cow

(Duitschaver and Ashton, 1972).

Large variations in SCC were observed throughout the lactations. The amount of

neutrophils present mirrored the rise and fall of total somatic cells present. It can be

interpreted from this data that fluctuations in SCC are due mainly to white blood cells.

The average sample to sample coefficient of variation for AM milkings was 136%, for

PM milkings it was 98%. The sample to sample coefficient of variation was lower for

PM milkings, however the average SCC was higher for evening milkings (509,000 versus

557,000). It is thought that differences in milk yield are not responsible for the

differences in SCC (Duitschaver and Ashton, 1972).

Clearly, the most effective and inexpensive way to prevent mastitis is through using

teat dip, antibiotic therapy, and keeping cows clean. Other factors, such as nutrition, may

play a small role as well. Mastitis is the most costly disease that affects dairy cattle

therefore preventing it is economically advantageous to all dairy farmers.














CHAPTER 3
FLORIDA MILK QUALITY LAB ANALYSIS

Introduction

As stated in Chapter 1, the goal of this thesis is to describe the patterns of

fluctuation in individual cow SCC over relatively short periods of time. Traditionally,

most dairy producers use monthly milk SCC from DHI. In addition to this, these

monthly SCC usually are not daily composites, meaning that many dairy producers only

see one SCC from one milking per cow per month. In a herd that is milked two times

daily, a dairy producer is then using only 1 milking out of 60 to evaluate the udder health

of individual cows. If the herd is milked 3 times daily the producer is then only using

one SCC out of 90 milkings. SCC testing costs money, therefore it is easily

understandable that dairy producers do not want to pay for excessive testing. The

problem, however, is that more money may be lost when cows are treated with antibiotics

or culled unnecessarily. If the samples that were tested for SCC were a composite sample

from 2 or more milkings the variation would be reduced.

These observations were the basis for this thesis. There is very little data, however,

on milking to milking variation in SCC. Therefore, we believed the appropriate first step

would be to evaluate the lab that would be used in the study. With little to no data to

compare ours to, if the lab was found to be accurate then we know that any variation that

we identified is due to factors other than the accuracy of the laboratory. The objective of

this study was to determine the accuracy and the precision of the DHI labs in Florida that









analyze milk samples for SCC. Our hypothesis was that the lab would prove to be

accurate and precise.

Materials and Methods

Four labs in the state of Florida were evaluated for accuracy and repeatability of

SCC results. Thirty milk samples (one gallon each) were obtained from 30 different

dairies. The samples were stored below 40 degrees F. These samples were then split into

8 sub samples. Four laboratories received sixty samples each (two laboratories are used

in this analysis), the duplicates were randomly assigned a number from 31 to 60 so

technicians would not know which sample it was a duplicate of. In addition to the sixty

samples, each laboratory also received four standard samples that were purchased from a

commercial laboratory.

SCC were counted electronically with a Bentley Somacount cell counter (Bentley

Instruments Inc, Chaska, MN). The Somacount utilizes laser based flow cytometry to

count cells one by one. First the DNA of the somatic cells is stained with ethium

bromide, which makes them fluorescent. This enables the laser to then count them. The

machine is calibrated at start up and every hour thereafter using a standard. The study

was conducted as a blind trial so lab personnel were unaware which samples were

duplicates. The sample and its duplicate were not run one after the other. Duplicate

samples of four standards were also run at the same two labs using the same procedures.

Data was analyzed using SAS (version 9.0) and Microsoft Excel to perform f-

distributions and t-tests.

Results and Discussion

Figures 3-1 through 3-3 illustrate the values for the sample and its duplicate sample

at both of the labs.







































Figure 3-1. SCC for Samples and their duplicates analyzed at lab 1.

In Figure 3-1 duplicate samples are being compared in the same lab. Careful

observation of the data will point to the conclusion that most of the duplicate samples

analyzed were extremely close in value, if not exactly the same. Sample nine seems to

be the only outlier. The average difference between the sample and its duplicate was

12,600 cells/ml at lab one. The difference between sample nine and its duplicate was

97,000 at lab one. Most likely, sample nine was not mixed properly before it was split

into sub-samples. Milk tends to settle after short periods of time if the milk is not

homogenizd. Therefore, not mixing the milk thoroughly before the sample was split

could easily have caused this error.

































Figure 3-2. SCC tor Samples and their duplicates analyzed at lab 2.

Many of the trends in Figure 3-2 are the same as those in Figure 3-1. Most of the

samples were very close in value if not exactly the same. In Figure 3-1, however, the

value for the duplicate sample is higher than sample 1 and the opposite is true in Figure

3-2. Again one can see that there seems to be a problem with sample 9. The difference

between sample 9 and its duplicate at this lab is 141,000. The average difference

between a sample and its duplicate at lab two was 16,000 cells/ml. These differences are

so slight, however, that they are not even significant. Small variations like this one may

be due to the way in which the sample was mixed before it was split and the small

amount of variation that is possible with electronic somatic cell counting techniques.

Again, one can visually observe that there was an error with sample nine. Aside from

sample nine, the majority of the data seems very accurate.



































Figure 3-3. SCC for Samples and their duplicates analyzed at Lab 1 and Lab 2.

These graphs illustrate that the variation within lab was very small as well as the

variation between labs. The variance between the means in lab one was not significantly

different than the variance between means in lab two (P = .098) and there was no lab bias

(P = .684). Looking at this data it is easily observed that almost every sample has equal

values for its duplicate at each lab. It can be seen even more clearly now that sample

nine is erroneous. Many of the values, such as those for sample 3, 7, 29, and 30 are

almost exactly the same. There are no significant differences between them.

The four standards were also run at each lab so that there would be samples in the

analysis where the value was known before the trial was conducted. The value the lab

obtained for each standard was not statistically different from the actual value of the

standard (P = .387).









Table 3-1. Lab Analysis of Official Samples.
Lab Replicate Official SCC (in 1000s) Lab SCC (in Difference (in 1000s)
1000s) (Official Lab)
1 1 1182 1327 -145
1 1 670 702 -32
1 1 127 121 6
1 1 370 377 -7
1 2 1182 1305 -123
1 2 670 680 -10
1 2 127 125 2
1 2 370 373 -3
2 1 1182 1344 -162
2 1 670 681 -11
2 1 127 142 -15
2 1 370 375 -5
2 2 1182 1351 -169
2 2 670 682 -12
2 2 127 125 2
2 2 370 374 -4

The results from this trial show that the results from both of the labs are accurate.

Very similar results were obtained from both of the labs and the results were very

repeatable. Not only were the individual labs very precise and able to obtain similar

results from the same sample of milk, but both of the labs were accurate, meaning they

both obtained similar results from the same sample, and also were accurate when they

were checked against standards.














CHAPTER 4
VARIATIONS IN SOMATIC CELL COUNTS FROM MILKING TO MILKING

Introduction

SCC are often used as an indicator of udder health. It is important, therefore, that

we understand more scientifically the variations in SCC. Understanding these variations

will help scientists develop the best way in which to use somatic cell counts as an udder

health indicator.

Very little is known about SCC variation from milking to milking or even day to

day because limited daily or even weekly SCC variation data is available. There are two

studies worth mentioning. The first looked at SCC variation on a weekly basis and the

second on a daily basis.

In 1972, researchers in Canada evaluated SCC fluctuations from week to week

(Duitschaverand Ashton, 1972). Large variations in SCC were observed throughout the

lactations. The average sample to sample coefficient of variation for AM milkings was

136%, for PM milkings it was 98%.

In 2004, a study was conducted in the United Kingdom which attempted to observe

the variation in individual cow somatic cell counts. This study utilized aseptic milk

samples as well as daily SCC from the morning milking to determine SCC variations.

The most interesting observation is that which compares the within cow variance for

uninfected cows and the within cow variance for infected cows. The variance for both

groups was about 0.5%. This is somewhat of a surprise, as most would expect the

variance of uninfected cows to be lower because it is generally thought that uninfected









cows do not experience fluctuations in their SCC. This study suggests that to accurately

assess the mastitis status of a cow multiple somatic cell counts from a short period of

time reduce variation (Berry et al., 2004). The objective of this study was to observe the

variation of somatic cell counts from milking to milking. The expected results were

slight variations in SCC from milking to milking.

Materials and Methods

This study was conducted at the University of Florida Dairy Research Unit. Cows

were housed in free-stall facilities and milked 3 times daily. Milkings were evenly

spaced eight hours apart. 380 lactating cows were sampled for 15 consecutive milkings

over a period of 5 days. At each milking one milker was present. Three different people

milked the herd each day.

Milk samples were collected according to DHI sampling procedures. In-line milk

sampling devices were attached to the milk meters to take a composite sample from the

entire milking. Milk from the sampling device was mixed with a preservative in a milk

vial and analyzed within 2 days. Samples were collected by seven different persons

working in pairs. Two were trained DHI technicians. The others attended a mandatory

orientation before the start of the trial to outline sampling procedures. The quality of the

sample taken could potentially be affected by the amount of milk in the vial, a missing

preservative pill and inadequate mixing of the milk with the preservative (which causes

the fat to separate).

Samples were analyzed using electronic somatic cell counting (Bentley Somacount,

Bentley Instruments Inc, Chaska, MN). Electronic cell counting uses ethium bromide to

stain the DNA of somatic cells. The instrument then uses laser flow cytometry to count









the cells one by one. Anywhere from 100 to over 500 samples per hour can be analyzed

using electronic somatic cell counting.

The data collected was imported to Microsoft Excel. It was evaluated by two

different persons for unusual data or errors in the data (cow numbers that may have been

reversed etc.) Cows without 15 observations were dropped from the study. Data was

then analyzed using Microsoft Excel to calculate correlations and adjusted correlations.

Results and Discussion

Throughout the five day period, the average bulk tank SCC was 450,000 cells/ml.

Although the bulk tank SCC remained fairly constant, large variations in SCC from

milking to milking on an individual cow level were observed.

Figure 4-1 depicts the average SCC on an individual cow basis for the 15 milkings.

The majority of the cows have averages of less than 500,000 cells/ml. Almost half of the

cows, however, have an average above 500,000. The somatic cell counts range from very

low (22,000 cells/ml) to extremely high (9,000,000 cells/ml).


Average SCC on an Individual Cow Basis

100
3 80-
4-60
0


S20-
0
0-250,000 251,000- 500,001- 751,000- 1,000,000+
500,000 750,000 1,000,000
SCC (cells/ml)

Figure 4-1. Average SCC for the 15 milkings on an individual cow basis (380 cows
total).





























Figure 4-2. Number of cows over or under certain somatic cell
(380 cows total).


counts at all


15 milkings


Number of Cows Above or Below a Certain Daily
SCC All 5 Days


a 200

o 150

o 100

E 50
z 0
> < < > > > >
200,000 400,000 750,000 750,000 750,000 500,000 250000
SCC (cells/ml)

Figure 4-3. Number of cows over or under certain daily somatic cell counts each day
(380 cows total).

Figures 4-2 and 4-3 describe how many cows were over or under randomly chosen

SCC thresholds. This analysis was done on an individual milking basis and on a daily

basis. For the duration of this reserach, daily SCC will refer to the average SCC for all


Number of Cows Over or Under Threshold SCC at
all 15 Milkings


120
o 100
80 -
o 60 -
40
E 20
z 0

200000 400000 750000 1000000 750000 500000 250000
SCC









three milkings in a 24 hour period for each cow. The majority of the cows tended to stay

under 750,000 cells/ml throughout the duration of the study. The number of cows over

one million or even 750,000 at all milkings or every day is very low (0 and 2,

respectively). Particularly interesting is that only one cow had a SCC higher than

250,000 at every milking. This tells us that even though some cows had extremely high

SCC (>1,000,000), within the five day period of the study that same cow also

demonstrated very low somatic cell counts (< 250,000). Most cows never even had a

daily SCC above one million (189 cows). One hundred and twelve cows had one daily

SCC above one million, 60 cows had two, 15 cows had three, 4 cows had four and no

cows had all 5 daily SCC above one million.

Figures 4-4 to 4-7 indicate the frequency in which cows have daily SCC over a

certain threshold. For example, Figure 4-4 shows that almost 200 cows had no daily SCC

above one million, over 100 cows had one daily SCC above one million, about 60 cows

had two daily SCC over one million and so on. From these figures we can see that most

cows had very few, if any, daily SCC above 750,000. In Figure 4-6 we can see that more

cows had SCC above 500,000 two and three times. Figure 4-7 shows that the amount of

times a cow had a SCC above 250,000 is more spread out, but four and fives occurrences

still show the least amount of animals.

The conclusion that can be drawn from these figures is that, in general, most cows

did not have daily SCC that were extremely high. Many cows experienced consistently

low SCC (see Figure 4-7, almost 80 cows never had a daily SCC above 250,000). It is

peculiar, however, that no cows had a SCC above one million every day in the 5 day trial,

but we see over 100 cows with at least one daily average above one million. This is most









likely due to the variation in SCC that has been observed consistently throughout this

research.


Number of Cows With A Daily SCC Above One
Million


200

150

100

50
n


0 1 2 3 4 5
Number of Occurences Over One Million

Figure 4-4. Number of cows with a daily SCC over one million (380 cows total).


Number of Cows With A Daily SCC Above 750,000

200

150
0
4-)
o 100

S50 -
z
0 "-
0 1 2 3 4 5
Number of Occurences

Figure 4-5. Number of cows with a daily SCC over 750,000 cells/ml (380 cows total).


- -










Number of Cows With A Daily SCC Above 500,000


140
120
100
80
60
40
20
0


* Series


0 1 2 3 4 5
Number of Occurences


Figure 4-6. Number of cows with a daily SCC over 500,000 cells/ml (380 cows total).


Number of Cows With A Daily SCC Above 250,000


100

80

60

40

20

0


2 3
Number of Occurences


4 5


Figure 4-7. Number of cows with a daily SCC over 250,000 cells/ml (380 cows total).

Several correlations were calculated to better describe the relationship of SCC

fluctuations. Correlations based strictly on the SCC were calculated from milking to









milking as well as from morning to morning, afternoon to afternoon, etc. Correlations

with the SCC multiplied by milk production were also calculated to adjust for different

levels of milk production. The following tables display the values. The values in Table

4-1 and Table 4-2 are not the same; however they remain in a relatively small range of

0.3 to -0.01. These values reflect a very weak positive correlation and in some cases

even a negative correlation. Taking into account milk production did not seem to have

much of an effect on the correlations.

Table 4-1. Raw and adjusted (for milk production) SCC correlations for consecutive
milkings.
Milking Raw Correlation Adjusted
Correlation
1 to 2 0.005 0.009
2 to 3 0.055 0.039
3 to 4 0.143 0.128
4 to 5 0.104 0.085
5 to 6 0.075 0.044
6 to 7 0.024 0.095
7 to 8 -0.011 0.166
8 to 9 0.148 0.133
9 to 10 -0.011 -0.026
10 to 11 0.075 0.058
11 to 12 0.311 0.266
12 to 13 0.215 0.136
13 to 14 0.070 0.063
14 to 15 0.079 0.084

Correlations were also calculated by shift, which were each milked by a different

operator, to observe the relationship between SCC in the morning, afternoon and evening.

Milkings 1,4,7,10 and 13 were morning milkings (approximately 5 AM). Milkings

2,5,8,11 and 14 were afternoon milkings (1 PM) and milkings 3,6,9,12, and 15 were

evening milkings (9 PM). These correlations were calculated to look for differences in

SCC due to time of day.









The results of these correlations do not allow us to draw many conclusions. The

time of day in which the cow was milked does not seem to have an effect on the SCC.

Furthermore, the correlations that are calculated taking the milk weights into

consideration do not seem to give a clearer picture than the others. At first glance, Table

4-2 might seem to indicate that the correlations for morning milkings are more positively

related than the correlations between successive milkings. However, the correlation

between the fourth morning (milking 10) and the fifth morning (milking 13) is actually a

negative value in both tables. The evening milkings seem to be the most highly

correlated as there are no negative values in Table 4-2.

Table 4-2. Raw and adjusted (for milk production) SCC correlations for morning,
afternoon and evening milkings.
Shift Milking Raw Correlation Adjusted
Correlation
Morning 1 to 4 0.227 0.256
4 to 7 0.089 0.104
7 tol0 0.096 0.017
10to13 -0.004 -0.035
Afternoon 2 to 5 -0.006 0.008
5 to 8 0.043 0.057
8 toll 0.179 0.164
11 to 14 0.008 0.032
Evening 3 to 6 0.216 0.282
6 to 9 0.060 0.052
9to2 0.142 0.112
12 to 15 0.101 0.141

In a further attempt to describe the data, standard deviations and coefficients of

variation were calculated for each cow. Scatter plots were then constructed to observe

the coefficient of variation and standard deviation versus the cow's average somatic cell

count. Figures 4-8 and 4-9 display the results.











SCC Standard Deviation


4000
a 3500
0
= 3000
> 2500
o 2000
S1500
S1000
, 500
0


0 500 1000 1500 2000
Average SCC


2500


Figure 4-8. Individual cow standard deviation versus her average SCC (380 cows total).


Coefficient of Variation Versus Average SCC


- 4
0 3.5
*" 3
> 2.5
05 2
4-W
S 1.5
"I 1
w 0.5
o 0


500


1000 1500
Average SCC


2000


2500


Figure 4-9. Individual cow coefficient of variation versus her average SCC(380 cows
total).

Figure 4-8 illustrates that the higher a cow's average SCC is, the greater the

standard deviation. This means that the cows with higher cell counts tend to vary more

dramatically in their cell counts. Almost the same trend is observed in Figure 4-9. When


AL-
S+NY









the SCC is above a million, however, the coefficient of variation tends to level off at

about 2.

With such large fluctuations in SCC on a daily basis and from one milking to the

next, the hot list is probably not an accurate reflection of the udder health of individual

cows. Observing the data in this chapter it is clear that many cows experience temporary

spikes in SCC. From this analysis, we cannot attempt to explain those events. However,

it is safe to draw the conclusion that if any of the cows on the hot list were experiencing a

temporary spike, they would be treated unnecessarily. Admittedly, more studies will be

needed to insure that these results are repeatable; but the data collected so far, however,

suggests that the hot list may prove to be of no use.














CHAPTER 5
ANALYSIS OF THE VALUE OF HOT LIST USE

Monthly DHI somatic cell counts have been used for many years to assess the

mastitis status of individual cows (Reneau, 1986). More recently, the hot list was created

to aid dairy producers in quickly identifying the top 20 cows with highest number of

somatic cells per milking. The list gives a variety of information on the cow such as her

lactation number, her days in milk, and what her SCC was on that test day. In addition to

this, the hot list also calculates what the bulk tank SCC would be if that cow's milk, and

the cows above her, were not added to the tank on that particular day as well as the

percentage of cells in the bulk tank that the individual cow is responsible for.

To calculate which cows are on the hot list the first step is to calculate the total

amount of somatic cells in the bulk tank. After that is done the total amount of somatic

cells from each cow is calculated. The individual cow SCC is then divided by the bulk

tank SCC and multiplied by 100. This number represents the percentage of somatic cells

in the bulk tank that the individual cow is responsible for. The hot list then ranks the top

20 cows in the herd by percent cells contributed to the bulk tank.

Figure 5-1 is an example of a hot list that is sent monthly to dairy producers by

DHI in the state of Florida. The first column is the cow's identification number.

Following that is her lactation number, days in milk, her milk production for that day

expressed in pounds, and her SCC in thousands. The column labeled W/O is what the

bulk tank SCC would be without that cow's milk, and the cows above her, in the bulk

tank. The percent cells column lets the dairy producer know what percentage of the









somatic cells in the bulk tank that particular cow is responsible for. Only five cows are

shown on this example hot list. Actual hot lists show twenty cows.

Lactation Milk
Cow ID Lactaton DIM Milk SCC (1000s) W/O % Cells
Number (lbs)
A 5 101 79 9052 363 6.3
B 4 154 93 7352 341 6.1
C 3 100 72 9052 319 5.8
D 5 205 78 7352 300 5.1
E 1 29 83 5199 286 3.8
Figure 5-1. Example of a monthly DHI hot list (bulk tank SCC 386,000).

There are two ways in which the hot list could possibly benefit dairy producers.

The first way is lowering the bulk tank SCC over time by removing cows with high SCC.

The second way is more short term. It is thought that some producers may use the hot list

to rapidly reduce their bulk tank SCC if they are in danger of shipping illegal milk. This

could be accomplished by either culling cows on the hot list or purposely withholding

their milk from the bulk tank until their SCC returns to normal levels.

The first goal of this study was to describe cow movement on and off the hot list.

The second goal was to observe the effect on bulk tank SCC over a long period of time if

hot list cows were to be removed from the milking herd. The hypothesis was that cows

would not repeat on the hot list as often if the hot list was used to make management

decisions. Little is known about daily or even weekly variations in SCC, therefore it is

important to describe more scientifically the value of the hot list. The final objective was

to describe the effect of the hot list on the bulk tank SCC over short periods of time. It

was expected that the hot list would prove to be of little value because it only described

the SCC of the cow during one milking.









Materials and Methods

Two commercial dairies (Dairy A and Dairy B) in the state of Florida were used for

the first analysis. Dairy Herd Improvement lactation records from 1998 to 2003 were

utilized. Somatic cell counts were taken monthly by DHI technicians. Both dairies used

in the analysis milked 450 to 550 cows three times daily. In addition to this, both dairies

used floor mounted cow washers, pre-stripped, and post dipped with a teat disinfectant.

Dairy A used the hot list for making management decisions. More specifically,

every month each cow on the hot list was either treated with antibiotics, culled, dried off

early, or another management decision was made. Dairy B did not use the hot list to

make management decisions.

Data was obtained from Dairy Records Management Systems in Raleigh, North

Carolina. Data was analyzed using SAS (Version 9.0) using the mean and rank

procedures.

Data from Chapter 4 was also used and analyzed in Microsoft Excel. Another

study, very similar to the study in Chapter 4, is also included in this analysis. All the

data collectors used in this study were trained DHI personnel. The entire herd (the same

herd as was used in the previous study) was sampled for 5 consecutive milkings,

beginning with an evening milking. Cows that did not have a SCC for all 5 milkings

were dropped from the study. The same sampling devices were used in both studies and

samples were analyzed at the same lab using electronic somatic cell counting. This data

was also analyzed using Microsoft Excel.

Finally, the last data set was obtained from Southeast Milk Inc. (Belleview,

Florida) and was analyzed using the mixed procedure in SAS (version 9.0). Three years

of complete data were available (2002-2004) and forty-three herds were used in the









analysis. Somatic cell count data was available from daily or every other day milk pick

ups. Somatic cell counts were analyzed at the SMI lab in Belleview, Florida. This lab is

one of the labs that was analyzed in Chapter 3. Merging this data with the data obtained

from DHI (Raleigh, NC) we were able to determine when each testday was for each herd.

Results and Discussion

To comprehend the results, the possibilities for cow movement on and off the hot

list must be described. On any test day, there are two options for a cow. She could be

tested or not tested. If she is not tested she cannot appear on the hot list. Cows that are

not tested have been dried off, are in the hospital herd or have a missing sample. If a cow

was tested she could be on the list or off the list. In the following month, the same

options are again present for each cow. As an example, if a cow was tested in the first

month and was on the hot list, she could be tested the following month and still be on the

hot list or she could be tested and dropped off the hot list. The last option is that she was

not tested and therefore, not on the hot list.

Statistical analysis of cow movement on and off the hot list was very similar for

both dairies. On dairy A, 26.1% (SD 12.4) of the cows who were on the hot list during

any given month were on the hot list again during the following month. For dairy B,

26.1% (SD 10.8) of the cows were on the hot list for two successive months. On both

dairies, about 60% of the cows that were on the hot list during any given month dropped

off the hot list in the following month (the remainder of the cows were either not tested or

culled since the last test date).

For cows that were not on the hot list in the first month, about 3.5% of them on

dairy A and dairy B were on the list again in the following month. About 80% on both

dairies were not on the list in the current month and the following month.









Because dairy A made management decisions regarding all the cows on the hot list

every month, the results that were observed were different from what was expected. The

hypothesis was that Dairy A would have fewer cows than Dairy B repeating on the hot

list because more of those cows would be treated. The results suggest that dairy B was

effectively finding and treating mastitis cases using other management tools besides the

hot list. An alternate explanation could be that the fluctuation in cows on and off the hot

list is due to the high variation in individual cow SCC. This is in keeping with what was

reported in Chapter 4.

The following analysis observed the effect of the hot list cows on the bulk tank

SCC over a five year period. Figures 1 and 2 depict the bulk tank SCC of that test date as

well as a weighted bulk tank SCC. The weighted bulk tank SCC represents what the bulk

tank SCC would be if all the cows on the hot list in the previous month were culled. To

elaborate, the weighted bulk tank SCC reflects the current month's SCC with the milk

removed from the cows who were on the hot list the month before.












+Actual SCC
-calculated SCC




Figure 5-2. Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year
Period with Hot List Cows Removed for Dairy A (used hot list).




























Figure 5-3. Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year
Period with Hot List Cows Removed for Dairy B (did not use hot list).

The graphs clearly show that if the milk from all the cows on the hot list was

removed from the bulk tank on the test day there would be an obvious decrease in bulk

tank SCC. Removing the hot list cows from the bulk tank SCC reduces the SCC by

roughly one half, on dairies of about 500 cows. Differences in the amount of reduction

will vary depending on herd size. Smaller herds would see a more dramatic change in

SCC reduction because the 20 cows on the hot list represent a greater percentage of their

total herds. The opposite is also true for large herds. Their reduction in SCC would not

be as great because the 20 cows on the hot list represent a smaller percentage of cows in

their herd.

If we look at the effects of removing hot list cows on bulk tank SCC over time we

do not observe the same results. Figure 5-4 and Figure 5-5 depict actual SCC on a test

day and a calculated SCC where all milk from cows that were on the hot list in the

previous month has been removed.































Figure 5-4. Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year
Period with Hot List Cows in the Previous Month Excluded for Dairy A
(used hot list).


Figure 5-5. Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year
Period with Hot List Cows in the Previous Month Excluded for Dairy B (did
not use hot list).











Looking at the long term effects of hot list cows on the bulk tank SCC, it can be

observed that they are not nearly as dramatic as the short term effects. Many of the actual

SCC are very similar to the calculated SCC. Therefore, if all the cows on the hot list

were culled every month, the observed reduction in SCC would be about only 81,800

cells/ml the following month (for Dairy A and Dairy B). This was found by calculating

the differences between the actual SCC and calculated SCC and averaging the

differences.

From the analysis of DHI records, we observed that for both dairies cow movement

on and off the hot list was very similar. It was also calculated that cows on the hot list are

responsible for approximately half of the somatic cells in the bulk tank on the test day.

However, if all the cows on the hot list the previous month were culled, the decrease in

bulk tank SCC would not be nearly as dramatic as expected. We concluded that use, or

non-use, of the hot list did not have a significant effect on bulk tank SCC in the long term

on these two dairies.

This may be because of the high variability in somatic cell counts that was

previously discussed in Chapter 4. If the correlation of SCC from milking to milking is

very weakly correlated and very variable, it is unreasonable to expect that one SCC a

month can give us an accurate picture of what the mastitis status of a cow is. Basing

management decisions on the hot list could prove, in fact, to be detrimental. If a cow

appeared on the hot list one milking, based on the results in Chapter 4, it is possible that

her somatic cell count could return to a normal level within 8 hours. Therefore, a

producer may be treating or even culling a cow that is in a perfectly normal state of










health. Making assumptions about the health of a cow based on a single SCC could

easily be a very expensive mistake.

In an attempt to solidify these statements, hot lists were created from the data set

analyzed in Chapter 4. A hot list was made for each milking and each day. Figures 5-6

and 5-7 depict the possible number of times a cow could be on the hot list and how many

cows were on the hot list for each of those numbers.

The majority of cows never appear on the hot lists. There was one hot list created

for each milking, for a total of 15 hot lists. Each hot list has 20 cows on it. Fifty percent

of the cows on the 15 hot lists only appeared once. Thirty percent of the cows on the hot

list were on the lists twice. This tells us that if a producer is treating all the cows on the

hot list with antibiotics that possibly over 80% of the cows would probably be treated

unnecessarily. Only 21 of the cows were on the list more than twice and a mere eight

cows were on the list 4 or more times. There were no cows on the hot list more than 6

times, meaning there were no cows on the hot list for more than half of the milkings.

These data show great variation in which cows appear on the hot list from milking to

milking.


Frequency of Cows Appearing on the Hot List
(Hot Lists were calculated for each of the 15 milkings)

250
200
150
0
0 100
o M

0 1 2 3 4 5 6
Number of Times on the Hot List


Figure 5-6. Frequency of number of appearances on the hot list (lists calculated at each
milking; 380 cows total).









Daily hot lists were also created and analyzed. Figure 5-7 depicts how many cows

were on the daily hot lists zero times, one time, two times and three times. These hot lists

were calculated by averaging the SCC for all three milkings for each cow and then

finding the top twenty cows who contributed the most to the bulk tank each day. There

were five hot lists created. There potentially could have been 100 cows appearing on the

lists (5 days times 20 cows per day). Most of the cows never appeared on the hot list.

Seventy cows appeared once and only 4 of those 70 cows appeared on the hot list a

second time. That is only 6%. Again it can clearly be seen that if all the cows on the hot

list were treated 94% of them would have been treated needlessly.


Frequency of Cows Appearing on the Hot List
(Hot Lists calculated for each day)

350
300
O 250
0 200
150
E 100
50 -
0
0 1 2 3
Number of times on the Hot list


Figure 5-7. Frequency of number of appearances on the hot list (lists calculated for each
day; 380 cows total).

Because very little literature exists on milking to milking SCC variation, a second

study was conducted by the Southeast DHI (Gainesville, FL). This study will be called

study 2. It was conducted at the same dairy as used in Chapter 4 (Webb, 2005).

To conduct a direct comparison between the first and second study, hot lists were

created for each milking in the second study, for a total of 5 hot lists. These hot lists









were then compared with two sets of 5 hot lists from study one. Each set of hot lists from

the first study began with an evening milking so the analysis would not be confounded by

time of day.

Table 5-1 again depicts the variability in the cows on and off the hot list. It seems

as though the majority of the cows are only on the list once or twice although there are a

few cows that appear 3, 4 and 5 times. The results of this small analysis seem somewhat

inconclusive. If anything, they support the idea that management decisions based solely

on the hot list are not economically sound decisions.

Table 5-1. Number of times cows were on hot lists for study one and study two.
Number of Study 1, First set of Study 1, Second set Study 2 hot lists
Appearances on the hot lists of hot lists
hot list
1 49 61 27
2 17 18 10
3 5 0 6
4 1 0 5
5 0 0 3

The low repeatability of cows on the hot list observed in the original hot list

analysis and from creating hot lists from the two sets of SCC variability data further

indicates that the hot list is not a beneficial tool for dairy producers to utilize. This data

coupled with the data from the previous study (SCC Variability Chapter 4) provides solid

evidence that the hot list is not economically advantageous for the dairies analyzed in this

study.

It could be argued that there are ways in which to improve the hot list. For

example, if the hot list were to tell the dairy producer how many times each cow had

appeared on the hot list and when, producers may be able to identify which cows are truly

the persistent high SCC cows and not the cows that tend to have temporary spikes in their









SCC. This addition to the list might be a good one, in theory. However, referring back to

Figure 5-6, it can be easily observed that most cows appear on the hot list only once in a

period of 15 milkings. Fifty cows do show up on the hot lists twice, but we must

remember that is only twice out of fifteen milkings. On a percentage basis, these cows

only appear on the hot list 13% of the time. An occurrence of high SCC only 13% of the

time is probably not a high enough number to convince dairy producers that a certain cow

should be treated or culled. The cows that repeat on the hot list more than twice are an

insignificant amount.

Table 5-1 takes this analysis one step further and compares three sets of hot lists

which span 5 milkings each. It looks as though there are far more cows repeating on

these lists than in the set of hot lists used in Figure 5-6. However, if things are broken

down to a percentage basis again, averaging the data for these three sets of lists, 33% of

the cows repeat on the hot list twice in five days. A strong argument can be made that

dairy producers again would not find this amount substantial enough to want to utilize the

hot list to make management decisions.

A final analysis was completed to assess the effect of the hot list on the bulk tank

SCC immediately after the dairy producer receives the hot list. To analyze if the hot list

had an immediate effect on the bulk tank SCC, the bulk tank SCC of 43 dairies for thirty

days after the test day was analyzed. The thirty days was divided into six groups of five

days each. Group 1 consists of days 1-5 after the testday. Group two consists of days 6-

10 and so on until day 30. To see if there was an effect of the hot list in herds that had a

higher or lower average bulk tank SCC, the months were also divided into "high months"

or "low months" for each herd. A month is considered to be high if the 10 day average









SCC before the test day was above 500,000. A month was considered a low month if the

10 day average SCC before the test day was less than 500,000. The model inputted into

SAS was difference = herdcode highmonth daysaftergroup highmonth*daysaftergroup

where difference = average SCC in for a group the 10 day average SCC before the test

day, herdcode is each individual herd's identification and daysaftergroup is group 1-6,

depending on how many days after the testday the current SCC is. The variable name

"highmonth" accounts for high months and low months. The results for the proc mixed

procedure are displayed in Table 5-2. The fixed effects of the model that were

significant were herdcode, high month/low month, and group number. The interaction

between month and group was not significant but approached significance therefore it

was included in the model. The most interesting observation here is the value for high

month/low month. This tells us that the herd's SCC before the test day is probably the

most statistically important factor in determining if the herd's SCC will decrease after the

testday.

Table 5-2. Results of the Proc Mixed Procedure on daily bulk tank data.
Effect Pr > F
Herdcode 0.0002
Month <0.0001
Group 0.0184
Month*Group 0.0868

Least squares means were also calculated for high month and low month, groups 1-

6 and the interaction between high month and group number as well as low month and

group number. The results of the LS means procedure are listed in Table 5-3. The

change in SCC was significant statistically in many cases. The concern is, however, that

although these numbers are significant according to statistics, variations in SCC tend to









fluctuate often. Therefore, by just observing the change in SCC visually, the numbers are

not that dramatic.

Table 5-3. Results of the LS means procedure on daily bulk tank data.
Effect Change in SCC (in 1000s) Pr > t
High Month -16.33 <0.0001
Low Month -0.87 0.6618
Group 1 -4.31 0.1965
Group 2 -4.07 0.2250
Group 3 -13.04 0.0002
Group 4 -8.65 0.01711
Group 5 -16.65 <0.0001
Group 6 -5.07 0.2853
High Month Group 1 -11.14 0.0379
High Month Group 2 -4.97 0.3569
High Month Group 3 -20.43 0.0003
High Month Group 4 -19.62 0.0011
High Month Group 5 -29.9 <0.0001
High Month Group 6 -12.34 0.1262

High months have very significant (P < 0.0001) effects on the bulk tank SCC after

the testday. Herds that were having a low month had no significant groups and no

significance in the interactions between low months and group number. Groups 3-5

(days 11-25) also had a significant effect on bulk tank SCC after the test day. The

interaction between high months and groups 3-5 were also significant. None of the

interactions between low months and groups were significant.

This data is perplexing. The hypothesis was that if the hot list had an immediate

effect on SCC that the effect would be seen sometime between days 5 and 10. The

rationale for this is that the hot list does not reach the dairy producer until at least 3 days

after the testday because the lab must have time to analyze the samples. Therefore, it was

expected that groups one and two would show significance. Instead, groups 3-5 show

significance with group five being the most significant. The most feasible reason for this









is that the bulk tank SCC was high before the testday and it was in the process of

decreasing. If the reason for the decrease in somatic cell count was due to the use of the

hot list it would most likely be before day 10.

The hot list is a good idea in theory and with little research on daily or milking to

milking variation in SCC the usefulness of it could not be accurately assessed. However,

with the data presented in this thesis we can now question the role the hot lists should

play in making management decisions on the farm. Many more parameters other than

SCC must be analyzed before making a management decision.














CHAPTER 6
GENERAL DISCUSSION AND CONCLUSIONS

In this study, two labs in Florida were evaluated for accuracy and precision. By

testing each lab with thirty duplicate samples and four standards, we found that these labs

were accurately and precisely evaluating these milk samples. The variation in the

duplicates within lab and between labs was extremely low. The labs also accurately

determined the number of somatic cells in the standards.

A common mechanism DHI uses to help dairy producers monitor the somatic cell

count of their cows is the hot list. To evaluate the hot list we sampled 380 cows for 15

consecutive milkings. The results were surprising, as they showed great variability in

SCC over a short period of time. This analysis, combined with the analysis of the hot list,

indicated that SCC are much more variable than previously thought and because of this

maybe the hot list is not the best tool for dairy producers to use. Some argue that the hot

list should be used only in emergency situations when the dairy producer may be in

danger of shipping illegal milk. The theory is to use the hot list to find the highest cows

and hold their milk from the bulk tank (or take another management action) until a legal

limit of somatic cells is reached. However, this may not be the case. On the day the

technician sampled the herd, the cows on the hot list were the highest cows in the herd.

However, by the time the dairy producer receives the hot list it could be days or even a

week or longer, after the herd was sampled. After the technician samples the herd he

needs to send the sample to the lab, the lab needs to analyze it, print the report and send it

back to the producer. This process takes days. By the time the producer receives the hot









list the cows that were the highest may not be the highest anymore. For example, the

data from the study where the cows were sample at 15 consecutive milking shows cows

that drop from one million to under two hundred thousand in a period of eight hours.

Since large variations in somatic cell counts seem to be possible, the hot list may not be

the most beneficial tool for dairy producers to be using in any situation. In addition to all

of these factors, many herds enrolled in the DHI program are not sampled at every

milking during a 24 hour period, they are sampled at only one milking per month or

every other month. The number that the dairymen receive from this sampling is therefore

highly unreliable.

In addition to this, there are many other factors which should be taken into

consideration other than SCC before a cow is culled. Some of these other parameters

include reproductive status, more specifically, is the cow pregnant and how long did it

take to get her pregnant. Cows that take long periods of time to become pregnant again

are not as profitable as cows who are bred quickly. Cows with low milk production (in

what should be their peak phase) also should be considered for culling. Older cows and

cows with other problems such as lameness and frequent metabolic disorders also should

be taken into consideration. Dairy production is certainly not a black and white

operation. There is hardly ever one definite answer to a problem and many things must

be considered when a problem arises. Basing any decision off of one factor, such as

SCC, is almost always imprudent on a dairy. Factors such as SCC should be part of

making a decision but never the sole motivating force.

Extreme variability in SCC can have many implications for dairy producers. Many

producers rely on the once monthly SCC from DHI to make management decisions









regarding the cows on their dairy. Other producers use methods such as the California

Mastitis Test to evaluate the mastitis. This test depends on somatic cell counts for its

results as well. The variability in SCC is a good explanation of why $20 was lost per

cow treated in the McDermott study when every cow with a SCC above 400,000 was

treated with antibiotics (McDermott et al., 1983).

The problem facing dairy producers is that the cheapest and quickest ways to

evaluate mastitis problems on an individual cow level have always been based on somatic

cell counts. Milk cultures to determine if there is bacteria present in the udder take at

least 48 hours and are more costly. What dairy producers need is a test to quickly and

cheaply identify cows with mastitis that is not based on SCC. Currently, the only other

way to detect mastitis besides culturing and somatic cell counts is visual observation.

Perhaps dairy producers would be better suited if they trained their milkers extensively.

Milkers that are able to detect inflammation of the udder and abnormal milk could be a

huge asset on a dairy farm. One other possibility in identifying cows that may be ill is

looking at daily milk weights. Many dairy producers have a herd manager which

examines every cow which drops in production during a 24 hour period. This identifies

cows which may be affected with a number of different ailments, including mastitis.

Future research is necessary to solve this problem. The research must begin with

evaluating the variation in SCC more thoroughly to be certain that the results observed in

this study are repeatable. More research is also needed to discover what the true reason is

for the dramatic swings in somatic cell counts. Ultimately, dairy producers will need a

new way to analyze the mastitis status of individual cows in their herd.
















LIST OF REFERENCES


Allore, H.G. and H.N. Erb. 1998. Partial Budget of the Discounted Annual Benefit of
Mastitis Control Strategies. J. Dairy Sci. 81:2280-2292.

Asby, C.B., P.R. Ellis, T.K. Griffin, R.G. Kingwill. The Benefits and Costs of a System
of Mastitis Control in Individual Herds. University of Reading, Department of
Agriculture and Horticulture. Study No 17, 1975.

Berry, E.A., J.E. Hillerton and M. Gravenor. 2004. Variation of Individual Cow Cell
Count. Proc. 43rd National Mastitis Council Meeting. Verona, WI. P 284.

Blosser, T.H. 1979. Economic Losses From and the National Research Program on
Mastitis in the United States. J. Dairy Sci. 62:119-127.

Dinsmore, R.P., P.B. English, R.N. Gonzalez and P.M. Sears. 1992. Use of Augmented
Cultural Techniques in the Diagnosis of the Bacterial Cause of Clinical Bovine
Mastitis. J. Dairy Sci. 75:2706-2712.

Dohoo, I.R., and A.H. Meek. 1982. Somatic Cell Counts in Bovine Milk. Can. Vet. J.
23:119.

Duitschaever, C.L., and G.C. Ashton. 1972. Variations of Somatic Cells and
Neutrophils in Milk Throughout Lactation. J. Milk Food Technology. 35:197-202.

Elvinger, F., P.J. Hansen, and R.P. Natzke. 1991. Modulation of Function of Bovine
Polymorphonucleur Leukocytes and Lymphocytes by High Temperatures in Vitro
and in Vivo. Am. J. Vet. Res. 52:1962.

Emanuelsson, U. and E. Persson. 1984. Studies on Somatic Cell Counts in Milk from
Swedish Dairy Cows. Nongenetic Causes of Variation in Monthly Test-Day
Results. Acta Agric. Scand. 34:33.

Giesecke, W.H. 1975. The Definition on Bovine Mastitis and the Diagnosis of its
Subclinical Types During Normal Lactation. Proc. Seminar on Mastitis Control.
Int. Dairy Fed. Reading, England. Bull. Doc. 85:62.

Harmon, R.J. 1994. Symposium: Mastitis and Genetic Evaluation for Somatic Cell
Count. J. Dairy Sci. 77:2103-2112.

Heeschen, W. 1975. Determination of Somatic Cells in Milk. Proc. Seminar on Mastitis
Control. Int. Dairy. Fed. Bull. Reading, England. Doc. 85:79.









Hogan, J.S. W.P Weiss, and K.L. Smith. 1993. Role of Vitamin E and Selenium in Host
Defense Against Mastitis. J. Dairy Sci. 76:2795:2803.

Janzen, J.J. 1970. Economic Losses Resulting From Mastitis. A Review. J. Dairy Sci.
53:1151.

Kehrli, M.E. and D.E. Shuster. 1994. Factors Affecting Milk Somatic Cells and Their
Role in Health of the Bovine Mammary Gland. J. Dairy Sci. 77:619-627.

Kelton, D.F., K.D. Lissemore and R.E. Martin. 1998. Recommendations for Recording
and Calculating the Incidence of Selected Clinical Diseases. J. Dairy Sci. 81:2502-
2509.

Kingwill, R.G., F.H. Dodd, and F.K. Neave. 1979. Machine Milking and Mastitis. In
Machine Milking. Nat. Inst. For Res. In Dairying, Reading, England. P. 231.

Kitchen, B.J. 1981. Review of Progress of Dairy Science: Bovine Mastitis: Milk
Compositional Changes and Related Diagnostic Tests. J. Dairy Research. 48:167.

Ma, Y., C. Ryan, D.M. Barbano, D.M. Galton, M.A. Rudan, and K.J. Boor. 2000.
Effetct of Somatic Cell Count on Quality and Shelf-Life of Pasteurized Fluid Milk.
J. Dairy Sci. 83:264-274.

Martin, J.M. 1973. Milk Yield Interrelationships With Somatic Cells and Chemical
Constituents Over a Lactation and During Restricted Water Consumption. M.S.
Thesis, North Carolina State Univ., Raleigh.

McDermott, M.P., H.N. Erb, R.P. Natzke, F.D. Barnes and D.R. Bray. 1983. Cost
Benefit Analysis of Lactation Therapy with Somatic Cell Counts as Indications for
Treatment. J. Dairy Sci. 66:1198-1203.

Miller, R.H., U. Emanuelsson, E. Persson, J. Brolund, Philipsson and E. Funke. 1983.
Relationships of Milk Somatic Cell Counts to Daily Milk Yield and Composition.
Acta Agric. Scand. 33:209.

National Mastitis Council. 2003. SCC Regulatory Limit in US to Remain at 750,000.
Newsletter. Verona, WI. Volume 26, No 2, p 1.

Neave, F.K. 1975. Diagnosis of Mastitis by Bacteriological Methods Alone. Proc.
Seminar on Mastitis Control. Int. Dairy. Fed. Bull. Reading, England. Doc. 85:79.

Oldham, E.R., R.J. Eberhart and L.D. Muller. 1991. Effects of Supplemental Vitamin A
or p-Carotene During the Dry Period and Early Lactation on Udder Health. J.
Dairy Sci. 74:3775-3781.

Oliver, S.P., M.J. Lewis, T.L. Ingle, B.E. Gillespie and K.R. Matthews. 1993.
Prevention of Bovine Mastitis by a Premilking Teat Disinfectant Containing
Chlorous Acid and Chlorine Dioxide. J. Dairy Sci. 76:287:292.










Ott, S.L. 1999. Costs of Herd-Level Production Losses Associated With Subclinical
Mastitis in U.S. Dairy Cows. Proc. Natl. Mastitis Council. Natl. Mastitis Council,
Verona, WI. P 152.

Paape, M.J., A.J. Schultze, A.J. Guidry and R.E. Pearson. 1979. Leukocytes; Second
Line of Defense Against Invading Mastitis Pathogens. J. Dairy Sci. 62:135.

Phipps, L.W. 1968. Electronic Counting of Cells in Milk: Examination of a Chemical
Treatment for Dispersal of Milk Fat. J. Dairy Res. 35:295.

Ravinderpal, G., W.H. Howard, K.E. Leslie, and K. Lissemore. 1990. Economics of
Mastitis Control. J Dairy Sci. 73:3340-3348.

Reneau, J.K. 1986. Effective Use of Dairy Herd Improvement Somatic Cell Counts in
Mastitis Control. J. Dairy Sci. 69:1708.

Schultz, L.H. 1977. Somatic Cells in Milk- Physiological Aspects and Relationship to
Amount and Composition of Milk. J. Food Prot. 40:125.

Schutz, M.M., L.B. Hansen, G.R. Steuemagel and A.L. Kuck. 1990. Variation of Milk,
Fat, Protein, and Somatic Cells for Dairy Cattle. J. Dairy Sci. 73:484-493.

Sheldrake, R.F., R.J.T. Hoare and G.D. McGregor. 1983. Lactation Stage, Parity and
Infection Affecting Somatic Cells, Electrical Conductivity and Serum Albumin in
Milk. J. Dairy Sci. 66:542.

Smith, A., F.K. Neave, F.H. Dodd and G.C. Brander. 1966. Methods of Reducing the
Incidence of Udder Infection in Dry Cows. Vet. Rec. 79:233.

Sordillo, L.M., K. Shafer-Weaver and D. DeRosa. 1997. Immunobiology of the
Mammary Gland. J. Dairy Sci. 80:1851-1865.

Webb, D. Unpublished data. University of Florida. Gainesville, FL. 2004.

Weiss, W.P., J.S. Hogan and K.L. Smith. 2004. Changes in Vitamin C Concentration in
Plasma and Milk from Dairy Cows After an Intramammary Infusion of Escheria
Coli. J. Dairy Sci. 87:32-37

Weiss, W.P., J.S. Hogan, K.L. Smith and K.H. Hoblet. 1990. Relationships Among
Selenium, Vitamin E, and Mammary Gland Health in Commercial Dairy Herds. J.
Dairy Sci. 73:381-390.

Wilson, C.D., R.G. Kingwell. 1975. A Practical Mastitis Control Routine. Proc.
Seminar on Mastitis Control. Int. Dairy Fed. Bull. Reading, England. Doc. 85:62.















BIOGRAPHICAL SKETCH

Jessica Elizabeth Belsito was born in Millbury, Massachusetts, on January 31st,

1981. She grew up surrounded by the dairy industry. While in Connecticut, the author

spent many evenings and weekends milking at the University's dairy, which further

solidified her love of the dairy industry. She was also heavily involved with the

University's dairy club, Block & Bridle and Sigma Alpha. These activities all helped to

further her knowledge of the dairy industry.

The author decided to attend graduate school during her senior year of college. She

graduated in May, 2003, with a Bachelor of Science in animal sciences and a minor in

dairy management. Immediately following her graduation, she moved to Gainesville

where she began work on her Master of Science.




Full Text

PAGE 1

VALUE OF THE DAIRY HERD IMPROVEMENT (DHI) HOT LIST AS A DAIRY MANAGEMENT TOOL By JESSICA ELIZABETH BELSITO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Jessica Elizabeth Belsito

PAGE 3

This document is dedicated to the greatest man I never met, Robert John Belsito (19251973) and to the strongest woman I ever knew, Marcia Walsh (1920-2004).

PAGE 4

iv ACKNOWLEDGMENTS Gratefulness is extended to my superv isory committee, Dr. Roger Natzke, Dr Albert deVries and Dr Nick Place, for their time and effort. Many thanks are given to Dave Bray, fo r taking me under his wing and giving me a wealth of knowledge about the more pract ical side of the dairy industry. I Thank my fellow students Bruno Amaral a nd his wife Michelle, Ben Butler, Liz Johnson, Christy Bratcher, Nathan Krueger and Wimberly Krueger. Lastly, I thank Jamie Foster for being a great Sigma Alpha sister. I thank my officemates Kelly Jimenez and Tiffany Herrera, for the company and camaraderie. I would be remiss if I did not th ank Kasey Moyes, for inspiring me to go to graduate school in the first place and being a sincere friend. I also wish to thank my parents for their support and love thr oughout this process. It was their guidance and the example they set forth which allowed me to become a successful person. I wish to extend my love to the rest of my family, Kerri, Bobby, Trish, Cassandra, Becky, Jerimy, Tatum, Colby, and the little one who is on her way. Lastly, I wish to thank Sergio, who has been understanding, patient, supportive and has made my life complete.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ..x CHAPTER 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................4 Background...................................................................................................................4 Pathogenesis.................................................................................................................5 Mode of Infection..................................................................................................5 Immunity...............................................................................................................6 Pathogens...............................................................................................................9 Mastitis Prevention and Treatment...............................................................................9 Economics...................................................................................................................15 Milk Composition and Quality...................................................................................17 Milk Culturing and Somatic Cell Counting Techniques............................................18 Factors Affecting Somatic Cell Counts......................................................................19 Variability in Somatic Cell Counts.............................................................................21 3 FLORIDA MILK QUALITY LAB ANALYSIS.......................................................23 Introduction.................................................................................................................23 Materials and Methods...............................................................................................24 Results and Discussion...............................................................................................24 4 VARIATIONS IN SOMATIC CELL COUN TS FROM MILKING TO MILKING29 Introduction.................................................................................................................29 Materials and Methods...............................................................................................30 Results and Discussion...............................................................................................31

PAGE 6

vi 5 ANALYSIS OF THE VALU E OF HOT LIST USE..................................................40 Materials and Methods...............................................................................................42 Results and Discussion...............................................................................................43 6 GENERAL DISCUSSION AND CONCLUSIONS...................................................55 LIST OF REFERENCES...................................................................................................58 BIOGRAPHICAL SKETCH.............................................................................................61

PAGE 7

vii LIST OF TABLES Table page 3-1 Lab Analysis of Official Samples............................................................................28 4-1 Raw and adjusted (for milk product ion) SCC correlations for consecutive milkings....................................................................................................................36 4-2 Raw and adjusted (for milk producti on) SCC correlations for morning, afternoon and evening milkings...............................................................................................37 5-1 Number of times cows were on hot lists for study one and study two.....................50 5-2 Results of the Proc Mixed Pr ocedure on daily bulk tank data.................................52 5-3 Results of the LS means pro cedure on daily bulk tank data....................................53

PAGE 8

viii LIST OF FIGURES Figure page 3-1 SCC for Samples and their dup licates analyzed at lab 1..........................................25 3-2 SCC for Samples and their dup licates analyzed at lab 2..........................................26 3-3 SCC for Samples and their duplicat es analyzed at Lab 1 and Lab 2........................27 4-1 Average SCC for the 15 milkings on an individual cow basis.................................31 4-2 Number of cows over or under certain so matic cell counts at all 15 milkings .......32 4-3 Number of cows over or under certain daily somatic cell counts each day ............32 4-4 Number of cows with a daily SCC over one million ..............................................34 4-5 Number of cows with a daily SCC over 750,000 cells/ml ......................................34 4-6 Number of cows with a daily SCC over 500,000 cells/ml ......................................35 4-7 Number of cows with a daily SCC over 250,000 cells/ml ......................................35 4-8 Individual cow standard devi ation versus her average SCC ...................................38 4-9 Individual cow coefficient of variation versus her average SCC.............................38 5-1 Example of a m onthly DHI hot list..........................................................................41 5-2 Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year Period with Hot List Cows Removed for Dairy A.............................................................44 5-3 Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year Period with Hot List Cows Removed for Dairy B...............................................................45 5-4 Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year Period with Hot List Cows in the Prev ious Month Excluded for Dairy A..........................46 5-5 Actual Bulk Tank SCC and Calculated Bulk Tank SCC Over a Five Year Period with Hot List Cows in the Prev ious Month Excluded for Dairy B..........................46

PAGE 9

ix 5-6 Frequency of number of appearances on the hot list................................................48 5-7 Frequency of number of appearances on the hot list................................................49

PAGE 10

x Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science VALUE OF THE DAIRY HERD IMPROVEMENT (DHI) HOT LIST AS A DAIRY MANAGEMENT TOOL By Jessica Elizabeth Belsito August 2005 Chair: Roger P Natzke Major Department: Animal Sciences Mastitis, an inflammation of the udder, is the most costly disease in the dairy industry. Somatic cells (which are mainly wh ite blood cells) increase during an infection. For this reason, and becaus e bacteriological procedures are time consuming and expensive, the somatic cell counts (SCC) are us ed as a general indica tor of udder health. Currently, Dairy Herd Improvement (DHI) offe rs milk sampling to determine individual cow SCC. Many dairy producers use these SCCs to make management decisions such as culling, treatment or early dry-off. However, it is not clear what the value is of these SCC to base these management decisions on. The objectives of the following studies were to analyze the accuracy of the DHI labs in Fl orida in determining SCC, observe more closely milking to milking SCC variation and to use DHI records to determine the value of the DHI hot list, which ranks the highe st SCC cows in the herd each test day. Chapter 3 was completed to assess the accuracy of our lab. Milk samples were analyzed in duplicates by electronic somatic cell counting. The variance of the mean

PAGE 11

xi differences between a sample and its duplicat e was not statistically significant between the labs (P=.098). In addition to this, vari ation from the standard was not significant (P=.387). These two factors indicate that on ly a small portion of the milking-to-milking variation can be attributed to the sampling and cell c ounting procedure. In Chapter 4, 390 cows were sampled for 15 consecutive milkings. Hot lists were created for each milking and each day to evalua te the repeatability of cows on the list. The repeatability was also very low. Wh en individual cow SCC was compared to the standard deviation for all 15 milkings, we sa w standard deviation increase with average SCC. An additional study was completed by Southeast DHI (Gainesville, FL) over a period of five milkings so there would be a comparison for the first study. Many of the results in the two studies in Chapter 2 were similar. Chapter 5 contains a statistical analysis of lactation records from 1998 until 2003 for two Florida dairy herds. The objective of the study was to determine the value of the use of the hot list to reduce the bulk tank S CC over time. The hot list identifies the 20 cows in the herd that contribu te the most cells to the bulk ta nk. Very few cows repeated on the hot list month after month. In addition to this, statistical an alysis of daily bulk tank SCC showed no positive effect from hot list use. Therefore, if the SCC of many cows decreases without interven tion (based on the results from the hot list analysis), one must question the economic value of intervening. We concluded that SCCs are highly variab le. These data indicate that the hot list may not be as useful as previously thought and dairy producers should evaluate and utilize other methods to manage their bulk tank SCC.

PAGE 12

1 CHAPTER 1 INTRODUCTION Quality food products are always a con cern in the United States for producers, consumers and the government alike. One wa y milk quality is meas ured is through the counting of somatic cells. The word somatic means of the body. Therefore, it follows that somatic cells in milk are any kind of cells from the body other than milk. Somatic cells could possibly be white bl ood cells, epithelial cells or othe r types of cells. There are many different reasons for keeping somatic ce ll counts (SCC) low. Producers, consumers and other groups each have their own reason for wanting low SCC. Producers are concerned about quality. A recent study (Ma et al., 2000) concluded that lower SCC leads to an increase in the shelf life of milk. Producing a product that lasts longer is economically beneficial to a ll consumers. Some consumers may also be concerned about the health of the animal producing the milk and the cleanliness of farms (both of these factors can affect SCC) The National Mastitis Counc il (NMC), a well known orga nization that focuses on mastitis and milk quality, has been pushing th e US government to lower the legal limit of somatic cells from 750,000 cells/ml to 400,000 (NMC, 2003). The reason is that many countries in other areas of the world have lowered their legal limit to 400,000 and some people believe that for the United States to successfully sell milk products on an international level we need to lower our legal somatic cell limit. Clearly SCC is currently a pressing issue for dairy producers.

PAGE 13

2 One reason many dairy producers try to lo wer somatic cell counts is premiums. Many milk buying organizations offer monetary premiums for lower SCC. Perhaps even more important is the issue of mastitis (which is a costly disease). It is generally believed that if a herd SCC is low less mastitis will be found on the farm. These are just some of the reasons that the Dairy Herd Improvement Association (DHI), an association which aids dairy producers in record keeping and data collection, has tried to develop and implement tools to aid dairy producers in monitoring their SCC. One of the first things that DHI did was to set up record keeping systems and labs to evaluate the SCC of milk samples. Many herds across the US participate in DHI monthly somatic cell count tes ting, where a DHI technician co mes to their farm and takes a milk sample from every cow. These samp les are sent to DHI labs where they are analyzed and the result s are sent to the producer and st ored in DHI databases. This provides the dairy producer with a milk weig ht and SCC for each individual cow. The goal of monthly somatic cell counts is to help producers identify i ndividual cows with high SCC. By identifying these cows, manage rs can then make a management decision (culling, antibiotic therapy, etc.) to try to lo wer the SCC if necessary. To help managers identify these cows more efficiently, DHI provides a hot list which ranks the top twenty cows in the herd by total cells cont ributed to the bulk tank (the bulk tank SCC refers to the average somatic cell count of the entire herd as the bulk tank is where all the milk is stored, kept cool and agitated). This hot list, in theory, aids dairy producers in quickly identifying problem cows based on their SCC. The literature lacks peer-reviewed articles on milking to milking SCC variability and hot list analysis. Because of the problem s outlined above and because of the lack of

PAGE 14

3 literature on the subjects, the goal of this thesis was to observe and describe patterns and/or changes in individual cow SCC over tim e and to analyze the effectiveness of the hot list in managing bulk tank SCC. Through a combination of these studies we we re able to observe variability in SCC, the value of the hot list to the dairy producer and the effect of the hot list on the bulk tank SCC over time. Our hypothesis for the study was th at the labs would be accurate and the milking to milking somatic cell counts would sh ow slight variability. It was also thought that if the hot list was im proved by adding other important information it would then become a more useful tool for dairy producers.

PAGE 15

4 CHAPTER 2 LITERATURE REVIEW Mastitis is one of the most costly di seases to milk producers. Losses are associated with decreased milk production, discarding milk from cows treated with antibiotics, the cost of the antibiotics and in creased labor expenses (Ravinderpal et al., 1990). In addition to this, milk quality can be negatively affected by high somatic cell counts, which indicate the presence of infection (Ma et al., 2000). It is therefore the purpose of this literature review to descri be the events leading to mastitis, economic losses due to infection, methods used to dete ct mastitis, variability in SCC and milk quality and compositional changes. Background The National Mastitis Council defines mas titis as an inflammation of the udder, most commonly caused by infecting microorgani sms. The word mastitis itself means inflammation of the mammary gland. Gi esecke (1975) gives a much more specific definition of mastitis. His definition states th at mastitis is indeed an infection of the mammary gland. First, there is damage to the mammary epithelium. An inflammatory reaction (either clinical or subclinical) follows. Depending on the severity of the infection, local or widespread chan ges can take place in the animal. Giesecke adds to his definition the distin ction between subclinical and clinical mastitis. Subclinical mastitis involves subt le physiological changes (Giesecke, 1975). Clinical mastitis, however, is a more serious inflammatory response. Signs of clinical

PAGE 16

5 mastitis include: changes in milk composition (watery milk, flakes or clots), visual signs of infection (red, swollen quarter s) and a more dramatic increase in SCC (Neave, 1975). Pathogenesis Mode of Infection Mastitis begins at the teat end. The bact eria must first come in contact with the sphincter. Once the bacteria pass through the sphincter it must trav el up the teat canal which is lined with keratin. Keratin is a wax-like substance designed for trapping bacteria and also contains substances with an timicrobial properties (Sordillo et al., 1997). If the bacteria pass by the keratin in th e teat canal it can gain access to the mammary gland. There are many ways in wh ich bacteria may be introduced into the mammary gland. Pressure on the teat caused by cow movement may aid the bacteria in traveling up the teat canal. In addition to this, bacteria may multiply continually until they have reached milk producing tissues, al lowing them to establish themselves more permanently (Kehrli and Shuster, 1994). The milking machine also aids the bacter ia in gaining access into the teat. The exposure of the teat to the vacuum (usually between 12-17 Hg) can cause damage to the skin on the teat. Even more damage is cau sed when the teat cup liners do not collapse completely around the cows teat. When the line r is open, the pressure inside the teat is greater than the ambient pressure due to the vacuum, allowing the milk to be expelled. When the liner collapses and the milk is no longer being expelled, milk from other quarters may splash back and infect uninfected quarters. Large vacuum fluctuations also can affect the new infection rate. There ar e many causes for these fluctuations. All milking machines have vacuum fluctuations because of the pulsation cycle, or the cycle in which the vacuum is turned on and off to open and close the teat canal. Pulsation is

PAGE 17

6 used to avoid exposing the teat to a constant v acuum. It allows the teat canal to be open and closed rhythmically and prevents th e congestion of blood in the teat. These fluctuations are caused on an irregular basis because of air that may get into the vacuum system via the inflations. Additionally, the movement of milk causes vacuum fluctuation (Kingwill et al., 1979). Data on these fluctuations shows that wh en teats are contaminated with mastitis causing pathogens the rate of new infecti on is higher when exaggerated vacuum fluctuations are applied to the cows teats. In addition to this, contaminated milk coming in contact with the cows teat could be the cause of a new infection. When both causes of vacuum fluctuation occur at the same time the rates of new infection are at their highest (Kingwill et al., 1979). Bacteria that are able to reach milk pr oducing tissues, or alveoli, now must find ways to stay in the tissue. Some bacteria, such as Streptococcus agalactiae and Staphylococcus aureus are well adapted to attach themselves to the epithelial lining of the alveoli. This prevents the bacteria from be ing flushed out during the milking process. Other types of bacteria do not adhere as well to tissue. These bacteria have adapted other ways of staying in the udder, such as extremely fast multiplication (Kehrli and Shuster, 1994). Immunity Once in the gland, there are two main ways in which the cows immune system will try to clear the bacteria. The two basic groups of immunity in the cows udder are innate and specific immunity. Innate immunity is more nonspecific and responds to the bacterial challenges almost immediately. Innate immunity involves macrophages, neutrophils, polymorphonucleur ne utrophils (PMN), and natural killer cells. Specific

PAGE 18

7 immunity requires time to develop after the initial infection. Th is response involves antibody molecules, macrophages and lymphoids. These cells attack only the type of bacteria they are programmed to eliminate (Sordillo et al., 1997). If the leukocytes in the udde r are not able to eradicat e the bacteria, inflammation and infection will result. The milk in th e gland provides an optimal environment for bacteria multiplication. If the infection becomes severe enough, the cow may show systemic signs of infection including a re d, hard udder and a fever. Inflammation is caused by the flood of cells invo lved in the immune response into the udder (Paape et al., 1979). Bacteria cells that adhere and multiply in the udder can cause a variety of negative effects. Perhaps the worst is the formation of scar tissue in the udder. This is a result of bacteria adhering themselves to the lining of the alveoli (Sordillo et al., 1997). Bacteria are not the only organisms that can cause sc ar tissue. The movement and adhesion of PMNs to milk producing tissues may damage the tissue, form scar tissue, and lead to an increase in milk SCC (Harmon, 1994). This scar tissue, or fibrosis, inhibits milk production in the current lactation as well as in subsequent lact ations. In addition to this, bacteria have the ability to produce toxins and other substances that can kill milk producing cells, again reducing milk produc tion. Membrane permeability is also affected, allowing more blood components into th e milk. If the infection becomes severe enough, visible changes may occur in the cows milk (Kehrli and Shuster, 1994). In addition to the effects already listed, mastitis pathogens can cause healthy tissue to involute, or go into a state of resting. Usually, it is not un til the next lactation, if ever,

PAGE 19

8 that this tissue begins producing milk agai n. The blood and immune system components can also cause secretory ducts to be come blocked (Sordillo et al., 1997). PMNs are one of the first responders in the non-specific immune response. PMNs are phagocytic cells that engulf and digest bacteria. Anothe r non-specific immune response involves neutrophils. They account for greater than 90% of leukocytes in the mammary gland. Neutrophils are bactericidal and travel from the blood to the mammary gland after substances such as cytokines (an inflammatory mediator) are present in the blood. Neutrophils also have small antibac terial peptides (S ordillo et al., 1997). Macrophages, cells which are present in healthy mammary glands, are also phagocytic. These cells also engulf and di gest foreign bacteria. Neutrophils and macrophages both destroy milk solids, making them less desirable than blood leukocytes (Sordillo et al., 1997). The specific immune response is more complex. Lymphocytes are the only immune cells with the ability to identify fo reign bacteria with sp ecific receptors on their membranes. The other cells involved in spec ific immunity are antigen-presenting cells (Sordillo et al., 1997). There are two types of lymphocytes, T a nd B. T lymphocytes produce and secrete cytokines. There is much speculation surr ounding the function of T lymphocytes. What is known, however, is they play a role in pr otecting epithelial surface s. They may also mediate the activity of other cells involved in the immune system. B lymphocytes produce antibodies specific to the bacteria that are invadi ng the udder. B lymphocytes present cells to T lymphocytes to be eliminated (Sordillo et al., 1997).

PAGE 20

9 Pathogens There are a variety of pathogens that cau se mastitis. The most common pathogens include Staphylococcus aureus, St reptococcus agalactiae, coliforms, and other streptococci These species are considered to be the major mastitis pathogens because they are the most common. Othe r noteworthy bacteria include Pseudomonas, Corynebacterium bovis, Mycoplasma and coagulase-negative sta phylococci. In addition to being major pathogens, Staphylococcus aureus and Streptococcus agalactiae are also contagious. These bacteria can be spread to other cows through unsanitary milking procedures (Harmon, 1994). Environmental pathogens also cause acute clinical mastitis. These infections tend to affect the cow more quickly and severely than other mastitis pathogens. The duration of these infections also tends to be shorter. These pathogens are found in various places in the cows environment. Most of these in fections are caused by cows laying in manure or other types of unclean be dding, mud, or standing in unclean ponds or cooling ponds. The coliforms are Gram-negative and make up the largest group of environmental pathogens they include: Escheria coli, Klebsiella spp. Streptococcus bovis, Enterococcus faecium, Enterobacter spp., Citrobacter spp., Streptococcus dysgalactiae, Streptococcus uberis and Enterococcus faecalis (Harmon, 1994). Mastitis Prevention and Treatment To avoid losing money because of mastitis prevention is the key. The rate of clinical mastitis varies greatly from farm to farm. In 62 citations from 1982 to 1996 the incidence of clinical mastitis ranged from less than 2.5% to over 50% (Kelton et al., 1998). There are many ways in which dairy pr oducers attempt to lower the incidence of mastitis on their farm. These include pre a nd post dipping of teats with a disinfectant

PAGE 21

10 using antibiotics at dry-off to prevent futu re infections, preventing the transmission of bacteria from one cow to another and k eeping freestalls clean (Oliver et al., 1993). An additional study evaluating the effi cacy of pre-dipping was completed in Tennessee in 1993. This study was a natural ex posure study following the procedures set by the National Mastitis Council. The active ingredients in the teat dip were sodium chlorite (0.64%) and lactic acid (2.64%). The study was conducted over a period of fifteen months and the average number of cows enrolled in the study at any given time was 175. There was a total of 423 cows involved in the study. Cows were milked twice daily and housed in a freestall with sawdust bedding (Oliver et al., 1993). The study was a negative control using a spli t udder design. The left teats were the controls. They were stripped and dried. Bo th right teats were forestripped, dipped and wiped. Duplicate samples of foremilk were collected aseptically from all quarters monthly, twice within ten days after parturition, and when cl inical mastitis was observed (Oliver et al., 1993). New intramammary infections caused by Staph aureus (P < 0.05) and Strep species (P < 0.025) were both lower in teats that were tr eated with the predip. In addition to this, the total number of quarters with new infections caused by other major pathogens was also lower in the treatment group (P < 0.01) (Oliver et al., 1993). Besides teat dipping, another met hod that can help control mas titis is antibiotic use. In 1966, Smith wanted to evaluate the effectiv eness of intramammary antibiotic use at the start of the dry period. They had observed pr eviously that half of the cows in most dairies (excluding first calf heif ers) calved with one quarter infected. Smith believed that

PAGE 22

11 eliminating late lactation infections as we ll as preventing new inf ections during the dry period would help to decrease the number of infections present at calving. Smith used 900 cows from 36 herds. Cows were chosen by duplicate bacteriological milk samples taken in the last week of the cows lactation. Cows were assigned randomly at dry-off into three groups 1) control (no disi nfectant no antibiotics; 2) .2g Cloxicillin intramammary and dipped with 5% hypochlorite; 3) intramammary infusion of 1 g Cloxicillin and teats di pped with the 5% hypo chlorite solution. When the cows were dried off, half of them had positive bacteriological milk samples. At calving, bacteriol ogical samples were taken within 7 days. Over 60% of the control group had positive bacteriological m ilk samples compared to only 23% of group 2 and 15% of group three. This indicates that the use of a 1 gram dose of Cloxicillin is advantageous when given at dry off (Smith et al., 1966). Perhaps the most complete assessment of advantageous mastitis prevention practices is a study which was completed by E nglands National Institute for Research in Dairying in 1975. This study included thirty herds over a three year period. The herds were divided randomly and assigned to one of two treatment groups. The first group employed several mastitis prevention strategi es. They utilized post-dipping, antibiotic therapy for infected lactating cows and antibio tic therapy for all cows at dry-off. The second group used only teat dip (hypochlorite containing 4% availa ble chlorine). In addition to these factors, milking machines were evaluated annually at each herd and adjusted as necessary. The researchers visite d each herd weekly to assess the implication of the mastitis prevention strategy and to take milk samples for culture (Wilson and Kingwell, 1975).

PAGE 23

12 At the beginning of the study a blitz ther apy was utilized. All infected quarters in half the herds in each group were treated w ith an antibiotic. This was done to evaluate the initial reduction in the number of subclinical cows (Wilson and Kingwell, 1975). Percent of cows infected during lactati on and percent of quarters infected during lactation both dropped steadily each year. The average reduction of udder disease in all herds was about 70%. The blitz therapy elimin ated 68% of all infections in the cows which received it. However, a larger numbe r of subclinicals were cured by the use of antibiotics at dry off in the non-blitzed herds. At the end of the firs t year the percentages of infection were relatively similar in th e blitzed and non-blitzed herds (Wilson, 1975). The reduced rate of infection was shown in the mean bulk tank SCC, which fell from 680,000 to 310,000 cells/ml over the three year period. Treatment of clinical cases of mastitis fell by about one half over the three year period (Asby et al., 1975). The new infection rate of each pathogen declined each year except for Coliforms and others. The infection rate s of these two groups remained relatively the same over the three years. Streptococcus agalactiae was virtually eliminated from all herds (Wilson and Kingwell, 1975). In addition to teat dipping and use of an tibiotics, nutrition may also play a small role in mastitis management. In 1991 Ol dham et al. conducted a study looking at the effects of vitamin A and -carotene during the dry peri od and early lactation on udder health. This study had three treatment groups The first received 50,000 IU of vitamin A, the second 170,000 IU of vitamin A and the third 50,000 IU of vitamin A plus 300 mg of -carotene (per cow per day). This group of 82 Holsteins was supplemented starting two weeks before dry off until 6 weeks post calving. Blood samples were taken to

PAGE 24

13 measure the serum concentrations of vitamin A and -carotene 2 weeks before dry off, at dry off, the 28th day of the dry period, once between day 10 to 4 pre-calving, within 24 hours of calving, once during week 3 of lact ation and once during the sixth week of lactation. Milk was also samp led for culture. All samples we re in duplicate. They were collected at dry off, the 28th day of the dry period, once between day 10 to 4 before calving, within 24 hours of calving and once eac h week during week three and six of lactation. After culturing, the remainder of milk was used to determine somatic cell count. Quarters with clinical mastitis were sampled also (Oldham et al., 1991). Oldham found no difference in serum v itamin A levels between treatments throughout his study. Treatment effect on serum -carotene showed a tendency towards being significant (p <.08). Both serum -carotene levels and vitamin A levels tended to drop from 2 weeks prior to dry o ff until calving (Oldham et al., 1991). To determine if -carotene and/or vita min A were effective in treating mastitis frequency of new intramammary infections (IMI), SCC, and frequency of clinical mastitis were used. It was observed by these authors that treatment had no effect on all three of these parameters. It was also noted that treatment had no effect on which pathogens were the cause of the mastitis (Oldham et al., 1991). Vitamin E and selenium have both been shown to have positive effects on udder health when supplemented alone and together. In Ohio, nine dairy herds were used to evaluate the effects of supplementing vitami n E and selenium. Bulk tank SCC score was measured as well as clinical mastitis rate. Serum levels were also observed. The herds involved with the study were monitored for one year. Strep agalactiae and Staph aureus mastitis had to be well controlled for a herd to be enrolled in the study. Rations were

PAGE 25

14 sampled for analysis three times during the year. Plasma sa mples were also drawn from 10 cows in each herd three times during the year. Samples were taken from cows who were 60 days pre-calving up until 60 days post-calving. Duplicate quarter samples were taken for bacteriological culture from all quart ers of all cows that were diagnosed with clinical mastitis before treatment was given. Bulk tank samples were also taken weekly (Weiss et al., 1990). Increasing the amount of vitamin E and sele nium in the diet raised blood serum levels. Blood serum levels for selenium, however, were independent of supplementation once intake exceeded 5 mg/day. It was also observed in this study that increased levels of blood serum selenium had a highly significant positive effect on bulk tank SCC (p < .005). In addition to this, cows fed a higher diet of vitamin E tended to have less clinical mastitis (p <0.1) (Weiss et al., 1990). It must be noted, however, that approximately two thirds of the soil in the United States is seleni um deficient. If the animals were selenium deficient at the start of the study, the results may be biased (Hogan et al., 1993). Researchers in Ohio found that 24 hours after an intramammary injection with E. coli plasma vitamin C concentration decreased by 39% (p < 0.01). Vitamin C and ascorbic acid concentrations in milk from the infected quarters decreased by 52 and 62% respectively. Unchallenged quarters were unaffected. Twenty-one Holsteins were used in this study. Blood and milk was sampled before challenge, 24 hour s after the challenge and 7 days after challenge to be analyzed for vitamin C concentration. Because vitamin C concentrations decreased so dramatically following the E. coli challenge, these results indicate that vitamin C may play an important role in helping cows recover from mastitis as well as possibly help to pr event it (Weiss et al., 2004).

PAGE 26

15 Economics Mastitis is the most economically devasta ting disease facing dairy farmers. There have been several publications on the subject. All agree that mastitis is very costly but, because all the estimations use different me thods and different numbers, the end results can vary greatly. The following estimation is one of the more common found in the literature. It is estimated that at a milk price of $12.07 per hundred weight, $184 dollars are lost per cow in the herd per year (B losser, 1979). This number agrees with Ravinderpal (1990) who projected that ma stitis accounted for 70 to 80% of the $140$300 dollars lost per cow per year due to dis ease. Thus, the annual loss per year due to mastitis in the United States can easily exceed 2 billion dollars. The majority of this loss (estimated to be about 2/3rds) is accounted for by reduced milk production. Other factors include cost of treatment, dis carded milk, and increased labor. It is projected that if a quarter becomes infected the average milk lo ss for that quarter is anywhere from .34 to 2.66 kg per day (Janzen, 1970). In 1983, McDermott et al. analyzed the economics of treating every cow with a SCC above 400,000. This fifteen month study analyzed milk production and possible profits resulting from treating early to obtai n a higher response rate. Lactating dairy cows were used from five commercial herds th at all used teat dip and dry cow therapy. Composite milk samples were taken monthl y for electronic SCCs. The somatic cell sampling was unannounced. Within seven days of the sampling individual quarter samples were collected and the bacterio logy was completed. Cows were assigned randomly to the control or experimental group based on their ear tag number. In the experimental group all cows were infused with a lactating antibiotic (c ephapirin) the first time their SCC rose above 400,000 cells/ml. Th is was done only once in the lactation.

PAGE 27

16 Clinical cases in both the c ontrol and experimental groups were treated by the farm managers themselves. Treatment was base d on SCC. Bacteriology was completed to assess any possible benefits of the antibiotic in eliminating infection. Treating cows with SCC of 400,000 and a bove had no effect on milk production in that lactation when compared to the control group. In addition to this, the cost of treating cows in this manner translated into almost $20 per cow. Most of this cost was incurred because 49 false positive cows were treated. Included in the $20 were SCC survey, labor, discarded milk and drugs (McDermott et al., 1983). The study completed in 1975 at the Univer sity of Reading also included an economic analysis. It was calculated that dairy producers who followed a basic routine of post-dipping with an iodine teat dip, treating infected lactating cows with antibiotics and treating all cows at dry-off with an antib iotic enjoyed a benefit/c ost ratio of 2.55 to 1. That is for every one dollar spent on mastitis prevention, the dairy producer should see a $2.55 profit. The results of this study suggest that mastitis management procedures are economically beneficial but shoul d be used as part of the re gular routine on the farm and should be viewed as a long term applicati on and not a quick fix for a mastitis problem (Asby et al., 1975). More recently, other mastits control and treatment techni ques have been evaluated. A study done in 1998 used computer modeling to test the effectiveness of several mastitis control strategies. This study evaluated cont rol strategies for the following bacteria types: Streptococcus agalactiae, Streptoc occus spp, Staphylococcu s aureus, coagulasenegative staphylococci, and Escheria coli. Prevention of mastitis (forestripping,

PAGE 28

17 predipping and postdipping), vaccination for E. coli, lactation therapy and dry cow therapy were all evaluated for econo mic efficiency (Allore and Erb, 1998). Each strategy evaluated in the computer model was compared to a control herd to find the annual benefit (dollars per cow pe r year). The criteria for budgets were determined by changes in milk composition and production due to mastitis as well as the number of cows that were culled for mastitis reasons. Using this criteria, it was determined that prevention and dry cow thera py were beneficial for all pathogen groups. In addition to this it was found that in herd s where environmental mastitis was dominant, vaccination for the prevention of E. Coli is also beneficial (Allore and Erb, 1998). Milk Composition and Quality Elevated SCCs are not only accompanied by milk loss; compositional changes also take place. Milk fat tends to show a sli ght decrease as does lactose (although some studies show no change in fat c ontent). The protein content of milk fluctuates very little. The types of protein present, however, tend to change. Casein, the desirable milk protein for making cheese, decreases. Whey, an unde sirable protein for cheese making, tends to increase. Blood components such as album ins and immunoglobulins find their way into the milk because of an increase in permeab ility of the milk blood barrier (Harmon, 1994). Mineral balance in the milk is also aff ected. Potassium and calcium are markedly reduced in milk with a high SCC. Potassi um leaks through damaged epithelial cells. These same cells also allow sodium and chlo ride from the blood into the milk. This increase of blood components in the milk also ca uses the pH of milk to rise to 6.9 or even higher (Harmon, 1994). Quality of milk is also affected in the presence of mastitis. In a study comparing shelf life of high SCC milk (about 75 0,000 cells/mL) to low SCC milk (about 45,000

PAGE 29

18 cells/ml) it was found that after pasteurizat ion the high SCC milk had a much shorter shelf life than the low SCC milk by as many as seven days. It is thought that there are free fatty acids in high SCC milk that increas e in number and cause an increase in casein hydrolysis. After 21 days the organoleptic qua lity of the low SCC milk remained high. The high SCC milk was rancid, as determ ined by a sensory panel (Ma et al., 2000). Milk Culturing and Somati c Cell Counting Techniques Milk samples are cultured for a variety of reasons. Culturing milk identifies which organism or organisms the quarter is infected with. This can help in identifying the cause of the mastitis and which drugs to use during the treatment process. The most widely used culturing technique involve s the use of blood agar plates Aseptic milk samples are plated (.01 ml of milk) on ag ar plates containing 5% sheep blood and .1% esculin. Samples are then incubated at 37 C for 48 hours. Usually the plates are read at 24 hours of incubation and after the full 48 hours. Th e most effective culture results are obtained from aseptic quarter milk samples (Dinsmore et al., 1992). Somatic cells are mainly white blood cells Because white blood cells often appear in response to infection, SCC are used as a general indicator of udder health. There are several methods in which SCC can be determined. The most popular on farm test is the California Mastitis Test (CMT). This test gives an estimation of the SCC for each qua rter. The milk sample is mixed with a solution that congeals in the presence of somatic cells. The more gel-like the sample becomes the higher the SCC is thought to be. When this reaction occurs it is assumed that the cow has mastitis in that quarter. Anothe r test that is very similar to this is the Wisconsin Mastitis Test (Kitchen, 1981).

PAGE 30

19 In laboratories somatic cells can be determ ined either by a tec hnician counting cells in a small milk sample (DMSCC) or by an automated machine (ESCC). The DMSCC has several recognized drawbacks. These include non-homogenous distribution of cells in the sample, subjective decisions and human error (Kitchen, 1981). There are two main ways to count somatic cells electronically. They are the use of the Coulter Counter and the Fosso matic. The coulter counter uses particle size to identify somatic cells and count them (Phipps, 1968). The Fossomatic uses fluorescence to count cells. Once the milk is diluted with a buffe r, ethidium bromide is added to make the somatic cells fluorescent. They are th en counted (Heeschen, 1975). The Bentley Somacount (Bentley Instruments Inc, Chaska, MN), an instrument used presently, also counts cells using fluorescence. Factors Affecting Somatic Cell Counts Somatic cell counts are highly variable a nd can fluctuate immensely in a short period of time. There are many factors which affect SCC. The most important factors are the infection status and th e type of organism which causes the infection. SCC tends to peak shortly after the cow has been challenged by a pathogen. Depending on the organism present, this peak may occur hours or days after the in itial challenge. In addition to this, the level of the response vari es from cow to cow. The time it takes for SCC to return to normal can vary from days to months or never if the cow becomes chronically infected (Harmon, 1994). It is generally thought that older cows a nd cows in late lactation have higher somatic cell counts. In many cases this is tr ue because the cow has had more time to be exposed to pathogens. However, if lactat ing cows remain uninfected throughout the course of their lactation, SCC varies very little (Sheldrake et al., 1983).

PAGE 31

20 Another important aspect of SCC and stag e of lactation is fresh cows. After calving, cows tend to have a very high SCC. If the cow does not become infected during her fresh period, her SCC should decrease rapidl y and return to normal within 35 days in milk (Reneau, 1986). There is also thought to be a dilution fact or that affects SCC. This means that if a cow were to experience a reduction in m ilk production and the SCC remained the same, we would observe an increase in SCC. It has been suggested that this phenomenon is responsible for the increase in SCC toward s the end of lactation (Schultz, 1977). A reduction in water or feed has also been show n to increase SCC, probably because of the reduction in milk production (Martin, 1973). Various sources of stress can also ca use an increase in SCC (Dohoo and Meek, 1982). Heat stress has been shown to have dramatic effects on SCC (Elvinger et al., 1991). Heat stress also causes a marked decr ease in milk production; therefore, it is unclear whether we can attribute a rise in SCC to the stress itself or a decrease in production, which is thought to cause the dilution factor. More recently, researchers in Minnesota ut ilized DHI records to evaluate somatic cell variation. DHI records with at least 220 da ys in milk and at least 4 test days were used. Provisions were added into the data analysis to ensure that records with questionable data were not utilized. SCC was transformed into a somatic cell score (SCS) using a base two log scale. Scores range from 0 to 9. Lactation records were separated by parity and by breed (Schutz et al., 1990). The mean SCS increased with age, this finding agrees with previous research (Emanuelsson and Persson, 1984 and Miller et al., 1983). The general trend observed in

PAGE 32

21 this study was a high SCS at calving which st eadily decreased until peak milk production. After peak milk production, SCS tended to increase throughout th e remainder of the lactation. The authors admit, however, that so me of this increase in SCS could be due to the dilution factor associated with a decr ease in milk production (Schutz et al., 1990). Variability in Somatic Cell Counts Individual cow somatic cell counts have been used as a management tool for many years, mainly due to the relatively cheap avai lability of this data due to DHI (Reneau, 1986). However, it is suspect ed that individual cow somatic cell c ounts can be highly variably due to sampling errors or more w ithin cow variation than previously thought. In 2004, a study was conducted in the United Kingdom which attempted to observe the variation in individual cow somatic cell c ounts. This study utilized aseptic milk samples as well as daily SCC from the morning milking to determine SCC variations. The results implied that monthly individua l cow SCC would not reflect all of the infections that affected the herd in that m onth. However, the most interesting observation is between the within cow variance for uninf ected and infected cows. The variance for both groups was about 0.5. This is somewhat of a surprise, as most would expect the variance of uninfected cows to be lower b ecause it is generally thought that uninfected cows do not experience spikes in their SCC. This study suggests that to accurately assess the mastitis status of a cow multiple somatic cell counts from a short period of time are more accurate (Berry et al., 2004). In 1972, researchers in Canada observed va riations of somatic cells throughout lactation on a weekly basis. Eleven Holsteins were used an d samples were collected on a weekly basis using DHI approved sample colle ction jars. The sample time alternated between morning and evening milkings. Differential somatic cell counts were

PAGE 33

22 performed. Aseptic milk samples were also taken for bacteriological cultures when visual evidence of inflammation was observe d or when the SCC doubled for any one cow (Duitschaver and Ashton, 1972). Large variations in SCC were observed th roughout the lactati ons. The amount of neutrophils present mirrored th e rise and fall of total somatic cells present. It can be interpreted from this data that fluctuations in SCC are due mainly to white blood cells. The average sample to sample coefficient of variation for AM milkings was 136%, for PM milkings it was 98%. The sample to sample coefficient of variation was lower for PM milkings, however the average SCC was higher for evening milkings (509,000 versus 557,000). It is thought that differences in milk yield are not responsible for the differences in SCC (Duitschaver and Ashton, 1972). Clearly, the most effective and inexpensive way to prevent mastitis is through using teat dip, antibiotic therapy, and keeping cows clean. Other factors, such as nutrition, may play a small role as well. Mastitis is the most costly disease that affects dairy cattle therefore preventing it is economically advantageous to all dairy farmers.

PAGE 34

23 CHAPTER 3 FLORIDA MILK QUALITY LAB ANALYSIS Introduction As stated in Chapter 1, the goal of this thesis is to describe the patterns of fluctuation in individual cow SCC over relatively short peri ods of time. Traditionally, most dairy producers use monthly milk SCC from DHI. In addition to this, these monthly SCC usually are not daily composite s, meaning that many dairy producers only see one SCC from one milking per cow per mont h. In a herd that is milked two times daily, a dairy producer is then using only 1 m ilking out of 60 to evaluate the udder health of individual cows. If the herd is milked 3 times daily the producer is then only using one SCC out of 90 milkings. SCC testi ng costs money, theref ore it is easily understandable that dairy producers do not wa nt to pay for excessive testing. The problem, however, is that more money may be lo st when cows are treated with antibiotics or culled unnecessarily. If th e samples that were tested for SCC were a composite sample from 2 or more milkings th e variation would be reduced. These observations were the basis for this th esis. There is very little data, however, on milking to milking variation in SCC. Ther efore, we believed the appropriate first step would be to evaluate the lab that would be used in the study. With little to no data to compare ours to, if the lab was found to be accu rate then we know that any variation that we identified is due to factors other than th e accuracy of the laborat ory. The objective of this study was to determine the accuracy and th e precision of the DHI labs in Florida that

PAGE 35

24 analyze milk samples for SCC. Our hypothe sis was that the lab would prove to be accurate and precise. Materials and Methods Four labs in the state of Florida were evaluated for accu racy and repeatability of SCC results. Thirty milk samples (one ga llon each) were obtained from 30 different dairies. The samples were stored below 40 de grees F. These samples were then split into 8 sub samples. Four laboratories received sixty samples each (two laboratories are used in this analysis), the duplicates were ra ndomly assigned a number from 31 to 60 so technicians would not know which sample it wa s a duplicate of. In addition to the sixty samples, each laboratory also received four st andard samples that were purchased from a commercial laboratory. SCC were counted electronically with a Bentley Somacount cell counter (Bentley Instruments Inc, Chaska, MN). The Somacount utilizes laser based flow cytometry to count cells one by one. First the DNA of th e somatic cells is stained with ethium bromide, which makes them fluorescent. This enables the laser to then count them. The machine is calibrated at start up and every hour thereafter using a standard. The study was conducted as a blind trial so lab pe rsonnel were unaware which samples were duplicates. The sample and its duplicate were not run one after the other. Duplicate samples of four standards were also run at the same two labs using the same procedures. Data was analyzed using SAS (version 9.0) and Microsoft Excel to perform fdistributions and t-tests. Results and Discussion Figures 3-1 through 3-3 illustrate the values for the sample and its duplicate sample at both of the labs.

PAGE 36

25 Data Lab 1 0 200 400 600 800 1000 1200 1471013161922252831 Sample NumberSCC (cells/mL) Sample 1 Duplicate Figure 3-1. SCC for Samples and th eir duplicates analyzed at lab 1. In Figure 3-1 duplicate samples are being compared in the same lab. Careful observation of the data will point to the c onclusion that most of the duplicate samples analyzed were extremely close in value, if not exactly the same. Sample nine seems to be the only outlier. The average difference between the sample and its duplicate was 12,600 cells/ml at lab one. The difference between sample nine and its duplicate was 97,000 at lab one. Most likely, sample nine wa s not mixed properly before it was split into sub-samples. Milk tends to settle afte r short periods of time if the milk is not homogenizd. Therefore, not mixing the m ilk thoroughly before the sample was split could easily have caused this error.

PAGE 37

26 0 200 400 600 800 1000 1200 1357911131517192123252729 Sample NumberSCC (cells/mL) Sample 1 Duplicate Figure 3-2. SCC for Samples and th eir duplicates analyzed at lab 2. Many of the trends in Figure 3-2 are the same as those in Figure 3-1. Most of the samples were very close in value if not ex actly the same. In Figure 3-1, however, the value for the duplicate sample is higher than sample 1 and the opposite is true in Figure 3-2. Again one can see that there seems to be a problem with sample 9. The difference between sample 9 and its duplicate at th is lab is 141,000. The average difference between a sample and its duplicate at lab tw o was 16,000 cells/ml. These differences are so slight, however, that they are not even sign ificant. Small variations like this one may be due to the way in which the sample wa s mixed before it was split and the small amount of variation that is possible with electronic soma tic cell counting techniques. Again, one can visually observe that there wa s an error with sample nine. Aside from sample nine, the majority of the data seems very accurate.

PAGE 38

27 Data Both Labs 0 100 200 300 400 500 600 700 800 900 1000 1357911131517192123252729 Sample NumberSCC (cells/mL) Lab 1 Sample 1 Lab 1 Duplicate Lab 2 Sample 1 Lab 2 Duplicate Figure 3-3. SCC for Samples and their dup licates analyzed at Lab 1 and Lab 2. These graphs illustrate that the variation within lab was very small as well as the variation between labs. The va riance between the means in lab one was not significantly different than the variance between means in lab two (P = .098) and there was no lab bias (P = .684). Looking at this data it is easily observed that al most every sample has equal values for its duplicate at each lab. It can be seen even more clearly now that sample nine is erroneous. Many of the values, such as those for sample 3, 7, 29, and 30 are almost exactly the same. There are no significant differenc es between them. The four standards were also run at each lab so that there would be samples in the analysis where the value was known before th e trial was conducted. The value the lab obtained for each standard was not statistically different from the actual value of the standard (P = .387).

PAGE 39

28 Table 3-1. Lab Analysis of Official Samples. Lab Replicate Official SCC (in 1000s) Lab SCC (in 1000s) Difference (in 1000s) (Official Lab) 1 1 1182 1327 -145 1 1 670 702 -32 1 1 127 121 6 1 1 370 377 -7 1 2 1182 1305 -123 1 2 670 680 -10 1 2 127 125 2 1 2 370 373 -3 2 1 1182 1344 -162 2 1 670 681 -11 2 1 127 142 -15 2 1 370 375 -5 2 2 1182 1351 -169 2 2 670 682 -12 2 2 127 125 2 2 2 370 374 -4 The results from this trial show that the results from both of the labs are accurate. Very similar results were obtained from bot h of the labs and the results were very repeatable. Not only were the individual labs very precise and able to obtain similar results from the same sample of milk, but both of the labs were accurate, meaning they both obtained similar results from the same sample, and also were accurate when they were checked against standards.

PAGE 40

29 CHAPTER 4 VARIATIONS IN SOMATIC CELL COUN TS FROM MILKING TO MILKING Introduction SCC are often used as an indicator of udder health. It is important, therefore, that we understand more scientifically the variations in SCC. Un derstanding these variations will help scientists develop the best way in which to use somatic cell counts as an udder health indicator. Very little is known about SCC variation fr om milking to milking or even day to day because limited daily or even weekly SCC va riation data is available. There are two studies worth mentioning. The first looked at SCC variation on a weekly basis and the second on a daily basis. In 1972, researchers in Canada evaluated SCC fluctuations from week to week (Duitschaverand Ashton, 1972). Large varia tions in SCC were observed throughout the lactations. The average sample to sample coefficient of variation for AM milkings was 136%, for PM milkings it was 98%. In 2004, a study was conducted in the United Kingdom which attempted to observe the variation in individual cow somatic cell c ounts. This study utilized aseptic milk samples as well as daily SCC from the morning milking to determine SCC variations. The most interesting observation is that wh ich compares the within cow variance for uninfected cows and the within cow variance for infected cows. The variance for both groups was about 0.5%. This is somewhat of a surprise, as most would expect the variance of uninfected cows to be lower b ecause it is generally thought that uninfected

PAGE 41

30 cows do not experience fluctuations in their S CC. This study suggests that to accurately assess the mastitis status of a cow multiple somatic cell counts from a short period of time reduce variation (Berry et al., 2004). The objective of this study was to observe the variation of somatic cell counts from milki ng to milking. The expected results were slight variations in SCC from milking to milking. Materials and Methods This study was conducted at the University of Florida Dairy Research Unit. Cows were housed in free-stall fac ilities and milked 3 times dail y. Milkings were evenly spaced eight hours apart. 380 lactating cows were sampled for 15 consecutive milkings over a period of 5 days. At each milking one milker was present. Three different people milked the herd each day. Milk samples were collected according to DHI sampling procedures. In-line milk sampling devices were attached to the milk meters to take a composite sample from the entire milking. Milk from the sampling device was mixed with a preservative in a milk vial and analyzed within 2 days. Sample s were collected by seven different persons working in pairs. Two were trained DHI t echnicians. The others attended a mandatory orientation before the start of the trial to out line sampling procedures. The quality of the sample taken could potentially be affected by the amount of milk in the vial, a missing preservative pill and inadequate mixing of the milk with the preservative (which causes the fat to separate). Samples were analyzed using electronic somatic cell counting (Bentley Somacount, Bentley Instruments Inc, Chaska, MN). Elect ronic cell counting uses ethium bromide to stain the DNA of somatic cells. The instrument then uses laser flow cytometry to count

PAGE 42

31 the cells one by one. Anywhere from 100 to over 500 samples per hour can be analyzed using electronic somatic cell counting. The data collected was imported to Micr osoft Excel. It was evaluated by two different persons for unusual data or errors in the data (cow numbers that may have been reversed etc.) Cows without 15 observati ons were dropped from the study. Data was then analyzed using Microsoft Excel to calcu late correlations and adjusted correlations. Results and Discussion Throughout the five day period, the aver age bulk tank SCC was 450,000 cells/ml. Although the bulk tank SCC remained fairly co nstant, large variations in SCC from milking to milking on an individual cow level were observed. Figure 4-1 depicts the average SCC on an i ndividual cow basis for the 15 milkings. The majority of the cows have averages of less than 500,000 cells/ml. Almost half of the cows, however, have an average above 500,000. The somatic cell counts range from very low (22,000 cells/ml) to extremely high (9,000,000 cells/ml). Average SCC on an Individual Cow Basis0 20 40 60 80 100 0-250,000251,000500,000 500,001750,000 751,0001,000,000 1,000,000+ SCC (cells/ml)Number of Cows Figure 4-1. Average SCC for the 15 milki ngs on an individual cow basis (380 cows total).

PAGE 43

32 Number of Cows Over or Under Threshold SCC at all 15 Milkings0 20 40 60 80 100 120 < 200000 < 400000 < 750000 > 1000000 > 750000 > 500000 > 250000 SCCNumber of Cows Figure 4-2. Number of cows over or under certain somatic ce ll counts at all 15 milkings (380 cows total). Number of Cows Above or Below a Certain Daily SCC All 5 Days0 50 100 150 200 > 200,000 < 400,000 < 750,000 > 750,000 > 750,000 > 500,000 > 250000 SCC (cells/ml)Number of Cows Figure 4-3. Number of cows over or under certain daily so matic cell counts each day (380 cows total). Figures 4-2 and 4-3 describe how many co ws were over or under randomly chosen SCC thresholds. This analysis was done on an individual milki ng basis and on a daily basis. For the duration of this reserach, da ily SCC will refer to the average SCC for all

PAGE 44

33 three milkings in a 24 hour period for each cow. The majority of the cows tended to stay under 750,000 cells/ml throughout the duration of the study. The number of cows over one million or even 750,000 at all milkings or every day is very low (0 and 2, respectively). Particularly interesting is that only one cow had a SCC higher than 250,000 at every milking. This tells us that even though some cows had extremely high SCC (>1,000,000), within the five day peri od of the study that same cow also demonstrated very low somatic cell counts ( < 250,000). Most cows never even had a daily SCC above one million (189 cows). On e hundred and twelve cows had one daily SCC above one million, 60 cows had two, 15 cows had three, 4 cows had four and no cows had all 5 daily SCC above one million. Figures 4-4 to 4-7 indicate the frequenc y in which cows have daily SCC over a certain threshold. For example, Figure 4-4 shows that almost 200 cows had no daily SCC above one million, over 100 cows had one da ily SCC above one million, about 60 cows had two daily SCC over one million and so on. From these figures we can see that most cows had very few, if any, daily SCC above 750,000. In Figure 4-6 we can see that more cows had SCC above 500,000 two and three times. Figure 4-7 shows that the amount of times a cow had a SCC above 250,000 is more spread out, but four and fives occurrences still show the least amount of animals. The conclusion that can be drawn from these figures is that, in general, most cows did not have daily SCC that were extremel y high. Many cows experienced consistently low SCC (see Figure 4-7, almost 80 cows ne ver had a daily SCC above 250,000). It is peculiar, however, that no cows had a SCC a bove one million every day in the 5 day trial, but we see over 100 cows with at least one da ily average above one million. This is most

PAGE 45

34 likely due to the variation in SCC that ha s been observed consis tently throughout this research. Number of Cows With A Daily SCC Above One Million0 50 100 150 200 012345 Number of Occurences Over One MillionNumber of Cows Figure 4-4. Number of cows with a da ily SCC over one million (380 cows total). Number of Cows With A Daily SCC Above 750,0000 50 100 150 200 012345 Number of OccurencesNumber of Cows Figure 4-5. Number of cows with a da ily SCC over 750,000 cells/ml (380 cows total).

PAGE 46

35 Number of Cows With A Daily SCC Above 500,0000 20 40 60 80 100 120 140 012345 Number of OccurencesNumber of Cows Series1 Figure 4-6. Number of cows with a da ily SCC over 500,000 cells/ml (380 cows total). Number of Cows With A Daily SCC Above 250,0000 20 40 60 80 100 012345 Number of OccurencesNumber of Cows Figure 4-7. Number of cows with a da ily SCC over 250,000 cells/ml (380 cows total). Several correlations were calculated to better describe the relationship of SCC fluctuations. Correlations based strictly on the SCC were calculated from milking to

PAGE 47

36 milking as well as from morning to morni ng, afternoon to afternoon, etc. Correlations with the SCC multiplied by milk production were also calculated to adjust for different levels of milk production. The following tabl es display the values. The values in Table 4-1 and Table 4-2 are not the same; however th ey remain in a relatively small range of 0.3 to -0.01. These values reflect a very w eak positive correlation and in some cases even a negative correlation. Taking into account milk production did not seem to have much of an effect on the correlations. Table 4-1. Raw and adjusted (for milk pr oduction) SCC correlations for consecutive milkings. Milking Raw Correlation Adjusted Correlation 1 to 2 0.005 0.009 2 to 3 0.055 0.039 3 to 4 0.143 0.128 4 to 5 0.104 0.085 5 to 6 0.075 0.044 6 to 7 0.024 0.095 7 to 8 -0.011 0.166 8 to 9 0.148 0.133 9 to 10 -0.011 -0.026 10 to 11 0.075 0.058 11 to 12 0.311 0.266 12 to 13 0.215 0.136 13 to 14 0.070 0.063 14 to 15 0.079 0.084 Correlations were also calculated by shif t, which were each milked by a different operator, to observe the rela tionship between SCC in the mo rning, afternoon and evening. Milkings 1,4,7,10 and 13 were morning milki ngs (approximately 5 AM). Milkings 2,5,8,11 and 14 were afternoon milkings (1 PM) and milkings 3,6,9,12, and 15 were evening milkings (9 PM). These correlations were calculated to look for differences in SCC due to time of day.

PAGE 48

37 The results of these correlations do not allow us to draw many conclusions. The time of day in which the cow was milked does not seem to have an effect on the SCC. Furthermore, the correlations that are calculated taking the milk weights into consideration do not seem to give a clearer picture than the others. At first glance, Table 4-2 might seem to indicate that the correlati ons for morning milkings are more positively related than the correlations between successive milkings. However, the correlation between the fourth morning (milking 10) and th e fifth morning (milking 13) is actually a negative value in both tables The evening milkings seem to be the most highly correlated as there are no ne gative values in Table 4-2. Table 4-2. Raw and adjusted (for milk production) SCC correlations for morning, afternoon and evening milkings. Shift Milking Raw Correlation Adjusted Correlation 1 to 4 0.227 0.256 4 to 7 0.089 0.104 7 to10 0.096 0.017 Morning 10 to13 -0.004 -0.035 2 to 5 -0.006 0.008 5 to 8 0.043 0.057 8 to11 0.179 0.164 Afternoon 11 to 14 0.008 0.032 3 to 6 0.216 0.282 6 to 9 0.060 0.052 9 to 2 0.142 0.112 Evening 12 to 15 0.101 0.141 In a further attempt to describe the data, standard deviations and coefficients of variation were calculated for each cow. Sca tter plots were then constructed to observe the coefficient of variation and standard de viation versus the cows average somatic cell count. Figures 4-8 and 4-9 display the results.

PAGE 49

38 SCC Standard Deviation0 500 1000 1500 2000 2500 3000 3500 4000 05001000150020002500 Average SCCStandard Deviation Figure 4-8. Individual cow st andard deviation versus her average SCC (380 cows total). Coefficient of Variation Versus Average SCC0 0.5 1 1.5 2 2.5 3 3.5 4 05001000150020002500 Average SCCCoefficient of Variation Figure 4-9. Individual cow co efficient of variation versus her average SCC(380 cows total). Figure 4-8 illustrates that the higher a cows average SCC is, the greater the standard deviation. This means that the cows with higher cell counts tend to vary more dramatically in their cell count s. Almost the same trend is observed in Figure 4-9. When

PAGE 50

39 the SCC is above a million, however, the coeffi cient of variation tends to level off at about 2. With such large fluctuations in SCC on a daily basis and from one milking to the next, the hot list is probably not an accurate reflection of the udder health of individual cows. Observing the data in this chapter it is clear that many cows experience temporary spikes in SCC. From this analysis, we cannot attempt to explain those events. However, it is safe to draw the conclusion that if any of the cows on the hot list were experiencing a temporary spike, they would be treated unnecessarily. Admittedly, more studies will be needed to insure that these results are repeat able; but the data collected so far, however, suggests that the hot list ma y prove to be of no use.

PAGE 51

40 CHAPTER 5 ANALYSIS OF THE VALUE OF HOT LIST USE Monthly DHI somatic cell counts have b een used for many years to assess the mastitis status of individual cows (Reneau, 1986 ). More recently, the hot list was created to aid dairy producers in quick ly identifying the top 20 cows with highest number of somatic cells per milking. The list gives a va riety of information on the cow such as her lactation number, her days in milk, and what he r SCC was on that test day. In addition to this, the hot list also calculates what the bulk tank SCC would be if that cows milk, and the cows above her, were not added to the tank on that particular day as well as the percentage of cells in th e bulk tank that the individua l cow is responsible for. To calculate which cows are on the hot list the first step is to calculate the total amount of somatic cells in the bulk tank. Afte r that is done the total amount of somatic cells from each cow is calculated. The i ndividual cow SCC is then divided by the bulk tank SCC and multiplied by 100. This number repr esents the percentage of somatic cells in the bulk tank that the indivi dual cow is responsible for. The hot list then ranks the top 20 cows in the herd by percent cel ls contributed to the bulk tank. Figure 5-1 is an example of a hot list th at is sent monthly to dairy producers by DHI in the state of Florida. The first column is the cows identification number. Following that is her lactation number, days in milk, her milk production for that day expressed in pounds, and her SCC in thousands The column labeled W/O is what the bulk tank SCC would be without that cows m ilk, and the cows above her, in the bulk tank. The percent cells column lets the da iry producer know what percentage of the

PAGE 52

41 somatic cells in the bulk tank that particular cow is responsible for. Only five cows are shown on this example hot list. Ac tual hot lists s how twenty cows. Cow ID Lactation Number DIM Milk (lbs) SCC (1000s) W/O % Cells A 5 101 79 9052 363 6.3 B 4 154 93 7352 341 6.1 C 3 100 72 9052 319 5.8 D 5 205 78 7352 300 5.1 E 1 29 83 5199 286 3.8 Figure 5-1. Example of a monthly DHI hot list (bulk tank SCC 386,000). There are two ways in which the hot list could possibly benefit dairy producers. The first way is lowering the bulk tank SCC over time by removing cows with high SCC. The second way is more short term. It is t hought that some producers may use the hot list to rapidly reduce their bulk tank SCC if they are in danger of shipping illegal milk. This could be accomplished by either culling co ws on the hot list or purposely withholding their milk from the bulk tank until their SCC returns to normal levels. The first goal of this study was to descri be cow movement on and off the hot list. The second goal was to observe the effect on bulk tank SCC over a long period of time if hot list cows were to be removed from the milking herd. The hypothesis was that cows would not repeat on the hot lis t as often if the hot list was used to make management decisions. Little is known about daily or even weekly variations in SCC, therefore it is important to describe more scientifically the value of the hot list. The final objective was to describe the effect of the hot list on th e bulk tank SCC over short periods of time. It was expected that the hot list would prove to be of little value because it only described the SCC of the cow during one milking.

PAGE 53

42 Materials and Methods Two commercial dairies (Dairy A and Dairy B) in the state of Florida were used for the first analysis. Dairy Herd Improvem ent lactation records from 1998 to 2003 were utilized. Somatic cell counts were taken mont hly by DHI technicians. Both dairies used in the analysis milked 450 to 550 cows three ti mes daily. In addition to this, both dairies used floor mounted cow washers, pre-stripped, and post dipped with a te at disinfectant. Dairy A used the hot list for making mana gement decisions. More specifically, every month each cow on the hot list was either treated with antibiotics, culled, dried off early, or another management decision was made. Dairy B did not use the hot list to make management decisions. Data was obtained from Dairy Records Management Systems in Raleigh, North Carolina. Data was analyzed using SAS (Version 9.0) using the mean and rank procedures. Data from Chapter 4 was also used and analyzed in Microsoft Excel. Another study, very similar to the study in Chapter 4, is also included in this analysis. All the data collectors used in this study were trained DHI personnel. The entire herd (the same herd as was used in the previous study) was sampled for 5 consecutive milkings, beginning with an evening milking. Cows th at did not have a SCC for all 5 milkings were dropped from the study. The same samplin g devices were used in both studies and samples were analyzed at the same lab usi ng electronic somatic cell counting. This data was also analyzed using Microsoft Excel. Finally, the last data se t was obtained from Southeas t Milk Inc. (Belleview, Florida) and was analyzed using the mixed pr ocedure in SAS (version 9.0). Three years of complete data were ava ilable (2002-2004) and forty-three herds were used in the

PAGE 54

43 analysis. Somatic cell count data was availabl e from daily or every other day milk pick ups. Somatic cell counts were analyzed at the SM I lab in Belleview, Florida. This lab is one of the labs that was analy zed in Chapter 3. Merging this data with the data obtained from DHI (Raleigh, NC) we were able to determine when each testday was for each herd. Results and Discussion To comprehend the results, the possibilities for cow movement on and off the hot list must be described. On a ny test day, there are two options for a cow. She could be tested or not tested. If she is not tested sh e cannot appear on the hot list. Cows that are not tested have been dried off, are in the hos pital herd or have a missing sample. If a cow was tested she could be on th e list or off the list. In the following month, the same options are again present for each cow. As an example, if a cow was tested in the first month and was on the hot list, she could be te sted the following month and still be on the hot list or she could be tested and dropped off the hot list. The last option is that she was not tested and therefore, not on the hot list. Statistical analysis of cow movement on and off the hot list was very similar for both dairies. On dairy A, 26.1% (SD 12.4) of the cows who were on the hot list during any given month were on the hot list agai n during the following month. For dairy B, 26.1% (SD 10.8) of the cows were on the hot list for two successive months. On both dairies, about 60% of the cows that were on the hot list during any given month dropped off the hot list in the following month (the rema inder of the cows were either not tested or culled since the last test date). For cows that were not on the hot list in the first month, about 3.5% of them on dairy A and dairy B were on the list again in the follo wing month. About 80% on both dairies were not on the list in the current month and the following month.

PAGE 55

44 Because dairy A made management decisions regarding all the cows on the hot list every month, the results that were observed we re different from what was expected. The hypothesis was that Dairy A w ould have fewer cows than Dairy B repeating on the hot list because more of those cows would be trea ted. The results suggest that dairy B was effectively finding and treating mastitis case s using other management tools besides the hot list. An alternate explanat ion could be that the fluctuation in cows on and off the hot list is due to the high variation in individual cow SCC. This is in keeping with what was reported in Chapter 4. The following analysis observed the effect of the hot list cows on the bulk tank SCC over a five year period. Figures 1 and 2 de pict the bulk tank SCC of that test date as well as a weighted bulk tank SCC. The wei ghted bulk tank SCC represents what the bulk tank SCC would be if all the cows on the hot list in the previous month were culled. To elaborate, the weighted bulk tank SCC refl ects the current months SCC with the milk removed from the cows who were on the hot list the month before. Actual Bulk Tank SCC and Calculated Bulk Tank SCC with Hot List Cows RemovedDairy uses hot list 0 100 200 300 400 500 600 700 800 900 19981999200020012002 YearBulk Tank SCC (1000s) Actual SCC Calculated SCC Figure 5-2. Actual Bulk Tank SCC and Calc ulated Bulk Tank SCC Over a Five Year Period with Hot List Cows Rem oved for Dairy A (used hot list).

PAGE 56

45 Actual Bulk Tank SCC and Calculated Bulk Tank SCC with Hot List Cows Removed 0 100 200 300 400 500 600 700 19981999200020012002YearBulk Tank SCC (1000s) Actual SCC Calculated Dairy does not use hot list Figure 5-3. Actual Bulk Tank SCC and Calc ulated Bulk Tank SCC Over a Five Year Period with Hot List Cows Removed fo r Dairy B (did not use hot list). The graphs clearly show that if the milk from all the cows on the hot list was removed from the bulk tank on the test day th ere would be an obvious decrease in bulk tank SCC. Removing the hot list cows from the bulk tank SCC reduces the SCC by roughly one half, on dairies of about 500 cows Differences in the amount of reduction will vary depending on herd size. Smaller herds would see a more dramatic change in SCC reduction because the 20 cows on the hot list represent a greater percentage of their total herds. The opposite is also true for la rge herds. Their reduc tion in SCC would not be as great because the 20 cows on the hot list represent a smaller pe rcentage of cows in their herd. If we look at the effects of removing hot list cows on bulk tank SCC over time we do not observe the same results. Figure 54 and Figure 5-5 depict actual SCC on a test day and a calculated SCC where all milk fr om cows that were on the hot list in the previous month has been removed.

PAGE 57

46 0 100 200 300 400 500 600 700 800 900 19981999200020012002 YearBulk Tank SCC (1000s) Actual SCC Calculated SCCActual Bulk Tanks SCC and Calculated Bulk Tank SCC with Hot List Cows In the Previous Month Excluded Dairy uses hot list Figure 5-4. Actual Bulk Tank SCC and Calc ulated Bulk Tank SCC Over a Five Year Period with Hot List Cows in the Previous Month Excluded for Dairy A (used hot list). 0 100 200 300 400 500 600 700 19981999200020012002 YearBulk Tank SCC (1000s) Actual SCC Calculated SCCActual Bulk Tanks SCC and Calculated Bulk Tank SCC With Hot List Cows From the Previous Month Excluded Dairy does not use hot list Figure 5-5. Actual Bulk Tank SCC and Calcul ated Bulk Tank SCC Over a Five Year Period with Hot List Cows in the Prev ious Month Excluded for Dairy B (did not use hot list).

PAGE 58

47 Looking at the long term effects of hot list cows on the bulk tank SCC, it can be observed that they are not nearly as dramatic as the short term effect s. Many of the actual SCC are very similar to the calculated SCC. Therefore, if all the cows on the hot list were culled every month, the observed re duction in SCC would be about only 81,800 cells/ml the following month (for Dairy A a nd Dairy B). This was found by calculating the differences between the actual SCC and calculated SCC and averaging the differences. From the analysis of DHI records, we obs erved that for both dairies cow movement on and off the hot list was very similar. It wa s also calculated that cows on the hot list are responsible for approximately ha lf of the somatic cells in the bulk tank on the test day. However, if all the cows on the hot list the previous month were culled, the decrease in bulk tank SCC would not be nearly as dramatic as expected. We concluded that use, or non-use, of the hot list did not have a signi ficant effect on bulk tank SCC in the long term on these two dairies. This may be because of the high variab ility in somatic cell counts that was previously discussed in Chapter 4. If the co rrelation of SCC from milking to milking is very weakly correlated and very variable, it is unreasonable to expect that one SCC a month can give us an accurate picture of what the mastitis status of a cow is. Basing management decisions on the hot list could prove in fact, to be detrimental. If a cow appeared on the hot list one milking, based on th e results in Chapter 4, it is possible that her somatic cell count could re turn to a normal level with in 8 hours. Therefore, a producer may be treating or even culling a co w that is in a perfectly normal state of

PAGE 59

48 health. Making assumptions about the heal th of a cow based on a single SCC could easily be a very expensive mistake. In an attempt to solidify these statements hot lists were created from the data set analyzed in Chapter 4. A hot list was made for each milking and each day. Figures 5-6 and 5-7 depict the possible number of times a cow could be on th e hot list and how many cows were on the hot list for each of those numbers. The majority of cows never appear on the hot lists. There was one hot list created for each milking, for a total of 15 hot lists. E ach hot list has 20 cows on it. Fifty percent of the cows on the 15 hot lists only appeared once. Thirty percent of the cows on the hot list were on the lists twice. Th is tells us that if a producer is treating all the cows on the hot list with antibiotics that possibly over 80% of the cows would probably be treated unnecessarily. Only 21 of the cows were on the list more than twice and a mere eight cows were on the list 4 or more times. Th ere were no cows on the hot list more than 6 times, meaning there were no cows on the hot lis t for more than half of the milkings. These data show great variation in which co ws appear on the hot list from milking to milking. Frequency of Cows Appearing on the Hot List (Hot Lists were calculated for each of the 15 milkings)0 50 100 150 200 250 0123456 Number of Times on the Hot ListNumber Of Cows Figure 5-6. Frequency of number of appearances on th e hot list (lists ca lculated at each milking; 380 cows total).

PAGE 60

49 Daily hot lists were also created and an alyzed. Figure 5-7 depicts how many cows were on the daily hot lists zero times, one time, two times and three times. These hot lists were calculated by averaging the SCC for all three milkings for each cow and then finding the top twenty cows who contributed the most to the bulk tank each day. There were five hot lists created. There potentially could have been 100 cows appearing on the lists (5 days times 20 cows per day). Most of the cows never appeared on the hot list. Seventy cows appeared once and only 4 of those 70 cows appeared on the hot list a second time. That is only 6%. Again it can cl early be seen that if all the cows on the hot list were treated 94% of them woul d have been treated needlessly. Frequency of Cows Appearing on the Hot List (Hot Lists calculated for each day)0 50 100 150 200 250 300 350 0123 Number of times on the Hot listNumber of Cows Figure 5-7. Frequency of number of appearances on th e hot list (lists calculated for each day; 380 cows total). Because very little literature exists on milking to milking SCC variation, a second study was conducted by the Southeast DHI (Gaine sville, FL). This study will be called study 2. It was conducted at the same da iry as used in Chapter 4 (Webb, 2005). To conduct a direct comparison between th e first and second study, hot lists were created for each milking in the second study, for a total of 5 hot lists. These hot lists

PAGE 61

50 were then compared with two se ts of 5 hot lists from study one. Each set of hot lists from the first study began with an evening milki ng so the analysis would not be confounded by time of day. Table 5-1 again depicts the va riability in the cows on and off the hot list. It seems as though the majority of the cows are only on the list once or twi ce although there are a few cows that appear 3, 4 and 5 times. The re sults of this small analysis seem somewhat inconclusive. If anything, th ey support the idea that management decisions based solely on the hot list are not ec onomically sound decisions. Table 5-1. Number of times cows were on hot lists for study one and study two. Number of Appearances on the hot list Study 1, First set of hot lists Study 1, Second set of hot lists Study 2 hot lists 1 49 61 27 2 17 18 10 3 5 0 6 4 1 0 5 5 0 0 3 The low repeatability of co ws on the hot list observed in the original hot list analysis and from creating hot lists from th e two sets of SCC vari ability data further indicates that the hot list is not a beneficial t ool for dairy producers to utilize. This data coupled with the data from the previous st udy (SCC Variability Chap ter 4) provides solid evidence that the hot list is not economically a dvantageous for the dair ies analyzed in this study. It could be argued that there are ways in which to improve the hot list. For example, if the hot list were to tell the dairy producer how many times each cow had appeared on the hot list and when, producers may be able to identify which cows are truly the persistent high SCC cows and not the cows that tend to have temporary spikes in their

PAGE 62

51 SCC. This addition to the list might be a good one, in theory. However, referring back to Figure 5-6, it can be easily observed that most cows appear on the hot list only once in a period of 15 milkings. Fifty cows do show up on the hot lists twice, but we must remember that is only twice out of fifteen m ilkings. On a percentage basis, these cows only appear on the hot list 13% of the time. An occurrence of high SCC only 13% of the time is probably not a high enough number to convince dairy producers that a certain cow should be treated or culled. The cows that re peat on the hot list more than twice are an insignificant amount. Table 5-1 takes this analysis one step furt her and compares three sets of hot lists which span 5 milkings each. It looks as though there are far more cows repeating on these lists than in the set of hot lists used in Figure 5-6. However, if things are broken down to a percentage basis agai n, averaging the data for these three sets of lists, 33% of the cows repeat on the hot list twice in five days. A strong argument can be made that dairy producers again would not find this amount substantial enough to want to utilize the hot list to make management decisions. A final analysis was completed to assess the effect of the hot list on the bulk tank SCC immediately after the dairy producer receive s the hot list. To analyze if the hot list had an immediate effect on the bulk tank SCC, the bulk tank SCC of 43 dairies for thirty days after the test day was analyzed. The thir ty days was divided into six groups of five days each. Group 1 consists of days 1-5 afte r the testday. Group two consists of days 610 and so on until day 30. To see if there was an effect of the hot list in herds that had a higher or lower average bulk ta nk SCC, the months were also divided into high months or low months for each herd. A month is considered to be high if the 10 day average

PAGE 63

52 SCC before the test day was above 500,000. A month was considered a low month if the 10 day average SCC before the test day wa s less than 500,000. The model inputted into SAS was difference = herdcode highmont h daysaftergroup hi ghmonth*daysaftergroup where difference = average SCC in for a group the 10 day average SCC before the test day, herdcode is each individual herds iden tification and daysaftergroup is group 1-6, depending on how many days after the testda y the current SCC is. The variable name highmonth accounts for high months and low months. The results for the proc mixed procedure are displayed in Table 5-2. Th e fixed effects of the model that were significant were herdcode, high month/low mo nth, and group number. The interaction between month and group was not significan t but approached significance therefore it was included in the model. The most intere sting observation here is the value for high month/low month. This tells us that the herd s SCC before the test day is probably the most statistically important factor in determin ing if the herds SCC will decrease after the testday. Table 5-2. Results of the Proc Mixe d Procedure on daily bulk tank data. Effect Pr > F Herdcode 0.0002 Month <0.0001 Group 0.0184 Month*Group 0.0868 Least squares means were also calculated for high month and low month, groups 16 and the interaction between high month and group number as well as low month and group number. The results of the LS means procedure are listed in Table 5-3. The change in SCC was significant statistically in many cases. The concern is, however, that although these numbers are significant according to statistics, varia tions in SCC tend to

PAGE 64

53 fluctuate often. Therefore, by just observing the change in SCC visually, the numbers are not that dramatic. Table 5-3. Results of the LS means procedure on daily bulk tank data. Effect Change in SCC (in 1000s) Pr > t High Month -16.33 <0.0001 Low Month -0.87 0.6618 Group 1 -4.31 0.1965 Group 2 -4.07 0.2250 Group 3 -13.04 0.0002 Group 4 -8.65 0.01711 Group 5 -16.65 <0.0001 Group 6 -5.07 0.2853 High Month Group 1 -11.14 0.0379 High Month Group 2 -4.97 0.3569 High Month Group 3 -20.43 0.0003 High Month Group 4 -19.62 0.0011 High Month Group 5 -29.9 <0.0001 High Month Group 6 -12.34 0.1262 High months have very significant (P < 0.0001) effects on the bulk tank SCC after the testday. Herds that were having a low month had no significant groups and no significance in the interacti ons between low months and group number Groups 3-5 (days 11-25) also had a significant effect on bulk tank SCC after the test day. The interaction between high months and groups 35 were also significant. None of the interactions between low months and groups were significant. This data is perplexing. Th e hypothesis was that if the hot list had an immediate effect on SCC that the effect would be seen sometime between days 5 and 10. The rationale for this is that the hot list does not reach the dairy producer until at least 3 days after the testday because the lab must have tim e to analyze the samples. Therefore, it was expected that groups one and two would show significance. Instead, groups 3-5 show significance with group five being the most significant. The mo st feasible reason for this

PAGE 65

54 is that the bulk tank SCC was high before the testday and it was in the process of decreasing. If the reason for the decrease in somatic cell coun t was due to the use of the hot list it would most likely be before day 10. The hot list is a good idea in theory and w ith little research on daily or milking to milking variation in SCC the usefulness of it could not be accurately assessed. However, with the data presented in this thesis we can now question the role the hot lists should play in making management decisions on the farm. Many more parameters other than SCC must be analyzed before making a management decision.

PAGE 66

55 CHAPTER 6 GENERAL DISCUSSION AND CONCLUSIONS In this study, two labs in Florida were evaluated for accuracy and precision. By testing each lab with thirty dupl icate samples and four standard s, we found that these labs were accurately and precisely evaluating thes e milk samples. The variation in the duplicates within lab and between labs was extremely low. The labs also accurately determined the number of soma tic cells in the standards. A common mechanism DHI uses to help dairy producers monitor the somatic cell count of their cows is the hot list. To ev aluate the hot list we sampled 380 cows for 15 consecutive milkings. The results were surpri sing, as they showed great variability in SCC over a short period of time. This analysis, combined with the analysis of the hot list, indicated that SCC are much mo re variable than previously thought and because of this maybe the hot list is not the best tool for dair y producers to use. Some argue that the hot list should be used only in emergency situa tions when the dairy producer may be in danger of shipping illegal milk. The theory is to use the hot list to find the highest cows and hold their milk from the bulk tank (or ta ke another management action) until a legal limit of somatic cells is reached. However, this may not be the case. On the day the technician sampled the herd, the cows on the hot list were the highest cows in the herd. However, by the time the dairy producer recei ves the hot list it could be days or even a week or longer, after the herd was sampled. After the technician samples the herd he needs to send the sample to the lab, the lab needs to analyze it, print the report and send it back to the producer. This process takes days By the time the produc er receives the hot

PAGE 67

56 list the cows that were the highest may not be the highest anymore. For example, the data from the study where the cows were sample at 15 consecutive milking shows cows that drop from one million to under two hundred thousand in a period of eight hours. Since large variations in soma tic cell counts seem to be possi ble, the hot list may not be the most beneficial tool for dairy producers to be using in any situati on. In addition to all of these factors, many herds enrolled in the DHI program are not sampled at every milking during a 24 hour period, they are sa mpled at only one milking per month or every other month. The number that the dairymen receive from this sampling is therefore highly unreliable. In addition to this, there are many othe r factors which should be taken into consideration other than SCC before a cow is culled. Some of these other parameters include reproductive status, more specifically, is the cow pregnant and how long did it take to get her pregnant. Cows that take long periods of time to become pregnant again are not as profitable as cows who are bred quickly. Cows with low milk production (in what should be their peak phase) also should be considered for culling. Older cows and cows with other problems such as lameness a nd frequent metabolic disorders also should be taken into consideration. Dairy produc tion is certainly not a black and white operation. There is hardly ever one definite answer to a problem and many things must be considered when a problem arises. Basi ng any decision off of one factor, such as SCC, is almost always imprudent on a dair y. Factors such as SCC should be part of making a decision but never th e sole motivating force. Extreme variability in SCC can have many implications for dairy producers. Many producers rely on the once monthly SCC from DHI to make management decisions

PAGE 68

57 regarding the cows on their da iry. Other producers use met hods such as the California Mastitis Test to evaluate the mastitis. This test depends on somatic cell counts for its results as well. The variability in SCC is a good explanation of why $20 was lost per cow treated in the McDermott study when every cow with a SCC above 400,000 was treated with antibiotics (M cDermott et al., 1983). The problem facing dairy producers is that the cheapest and quickest ways to evaluate mastitis problems on an individual cow level have always been based on somatic cell counts. Milk culture s to determine if there is bacter ia present in the udder take at least 48 hours and are more costly. What da iry producers need is a test to quickly and cheaply identify cows with mastitis that is not based on SCC. Currently, the only other way to detect mastitis besides culturing and somatic cell counts is visual observation. Perhaps dairy producers would be better suited if they trained their milkers extensively. Milkers that are able to detect inflammati on of the udder and abnormal milk could be a huge asset on a dairy farm. One other possibility in identifying cows that may be ill is looking at daily milk weights. Many da iry producers have a herd manager which examines every cow which drops in producti on during a 24 hour period. This identifies cows which may be affected with a number of different ailments, including mastitis. Future research is necessary to solve this problem. The research must begin with evaluating the variation in SCC more thoroughly to be certain that the results observed in this study are repeatable. More research is also needed to discover what the true reason is for the dramatic swings in somatic cell c ounts. Ultimately, dair y producers will need a new way to analyze the mastitis status of individual cows in their herd.

PAGE 69

58 LIST OF REFERENCES Allore, H.G. and H.N. Erb. 1998. Partial Budget of the Discounted Annual Benefit of Mastitis Control Strategies. J. Dairy Sci. 81:2280-2292. Asby, C.B., P.R. Ellis, T.K. Griffin, R.G. Kingw ill. The Benefits and Costs of a System of Mastitis Control in Indi vidual Herds. University of Reading, Department of Agriculture and Horticulture. Study No 17, 1975. Berry, E.A., J.E. Hillerton and M. Gravenor. 2004. Variation of Individual Cow Cell Count. Proc. 43rd National Mastitis Council M eeting. Verona, WI. P 284. Blosser, T.H. 1979. Economic Losses From and the National Research Program on Mastitis in the United States. J. Dairy Sci. 62:119-127. Dinsmore, R.P., P.B. English, R.N. Gonzalez and P.M. Sears. 1992. Use of Augmented Cultural Techniques in the Diagnosis of th e Bacterial Cause of Clinical Bovine Mastitis. J. Dair y Sci. 75:2706-2712. Dohoo, I.R., and A.H. Meek. 1982. Somatic Cell Counts in Bovine Milk. Can. Vet. J. 23:119. Duitschaever, C.L., and G.C. Ashton. 1972. Variations of Somatic Cells and Neutrophils in Milk Throughout Lactat ion. J. Milk Food Technology. 35:197-202. Elvinger, F., P.J. Hansen, and R.P. Natzke 1991. Modulation of Function of Bovine Polymorphonucleur Leukocytes and Lym phocytes by High Temperatures in Vitro and in Vivo. Am. J. Vet. Res. 52:1962. Emanuelsson, U. and E. Persson. 1984. St udies on Somatic Cell C ounts in Milk from Swedish Dairy Cows. Nongenetic Causes of Variation in Monthly Test-Day Results. Acta Agric. Scand. 34:33. Giesecke, W.H. 1975. The Definition on Bovine Mastitis and the Diagnosis of its Subclinical Types During Normal Lactati on. Proc. Seminar on Mastitis Control. Int. Dairy Fed. Reading, England. Bull. Doc. 85:62. Harmon, R.J. 1994. Symposium: Mastitis and Genetic Evaluati on for Somatic Cell Count. J. Dairy Sci. 77:2103-2112. Heeschen, W. 1975. Determination of Somatic Cells in Milk. Proc. Seminar on Mastitis Control. Int. Dairy. Fed. Bu ll. Reading, England. Doc. 85:79.

PAGE 70

59 Hogan, J.S. W.P Weiss, and K.L. Smith. 1993. Role of Vitamin E and Selenium in Host Defense Against Mastitis. J. Dairy Sci. 76:2795:2803. Janzen, J.J. 1970. Economic Losses Resulting From Mastitis. A Review. J. Dairy Sci. 53:1151. Kehrli, M.E. and D.E. Shuster. 1994. Fact ors Affecting Milk Somatic Cells and Their Role in Health of the Bovine Mamma ry Gland. J. Dairy Sci. 77:619-627. Kelton, D.F., K.D. Lissemore and R.E. Martin. 1998. Recommendations for Recording and Calculating the Incidence of Selected Clinical Diseases. J. Dairy Sci. 81:25022509. Kingwill, R.G., F.H. Dodd, and F.K. Neave. 1979. Machine Milking and Mastitis. In Machine Milking. Nat. Inst. For Res. In Dairying, Reading, England. P. 231. Kitchen, B.J. 1981. Review of Progress of Dairy Science: Bovine Mastitis: Milk Compositional Changes and Related Diagnosti c Tests. J. Dairy Research. 48:167. Ma, Y., C. Ryan, D.M. Barbano, D.M. Galton, M.A. Rudan, and K.J. Boor. 2000. Effetct of Somatic Cell Count on Quality and Shelf-Life of Pasteurized Fluid Milk. J. Dairy Sci. 83:264-274. Martin, J.M. 1973. Milk Yield Interrelati onships With Somatic Cells and Chemical Constituents Over a Lactation and During Restricted Water Consumption. M.S. Thesis, North Carolina State Univ., Raleigh. McDermott, M.P., H.N. Erb, R.P. Natzke, F.D. Barnes and D.R. Bray. 1983. Cost Benefit Analysis of Lactation Therapy with Somatic Cell Counts as Indications for Treatment. J. Dairy Sci. 66:1198-1203. Miller, R.H., U. Emanuelsson, E. Persson, J. Brolund, Philipsson and E. Funke. 1983. Relationships of Milk Somatic Cell Counts to Daily Milk Yield and Composition. Acta Agric. Scand. 33:209. National Mastitis Council. 2003. SCC Regul atory Limit in US to Remain at 750,000. Newsletter. Verona, WI. Volume 26, No 2, p 1. Neave, F.K. 1975. Diagnosis of Mastitis by Bacteriological Me thods Alone. Proc. Seminar on Mastitis Control. Int. Dair y. Fed. Bull. Reading, England. Doc. 85:79. Oldham, E.R., R.J. Eberhart and L.D. Muller. 1991. Effects of Supplemental Vitamin A or -Carotene During the Dry Period and Earl y Lactation on Udder Health. J. Dairy Sci. 74:3775-3781. Oliver, S.P., M.J. Lewis, T.L. Ingle, B.E. Gillespie and K.R. Matthews. 1993. Prevention of Bovine Mastitis by a Prem ilking Teat Disinfectant Containing Chlorous Acid and Chlorine Dioxi de. J. Dairy Sci. 76:287:292.

PAGE 71

60 Ott, S.L. 1999. Costs of Herd-Level Pr oduction Losses Associated With Subclinical Mastitis in U.S. Dairy Cows. Proc. Natl. Mastitis Council. Natl. Mastitis Council, Verona, WI. P 152. Paape, M.J., A.J. Schultze, A.J. Guidry and R.E. Pearson. 1979. Leukocytes; Second Line of Defense Against Invading Mastitis Pathogens. J. Dairy Sci. 62:135. Phipps, L.W. 1968. Electronic Counting of Ce lls in Milk: Examination of a Chemical Treatment for Dispersal of Milk Fat. J. Dairy Res. 35:295. Ravinderpal, G., W.H. Howard, K.E. Leslie and K. Lissemore. 1990. Economics of Mastitis Control. J Dairy Sci. 73:3340-3348. Reneau, J.K. 1986. Effective Use of Dair y Herd Improvement Somatic Cell Counts in Mastitis Control. J. Dairy Sci. 69:1708. Schultz, L.H. 1977. Somatic Cells in Milk Physiological Aspect s and Relationship to Amount and Composition of Milk. J. Food Prot. 40:125. Schutz, M.M., L.B. Hansen, G.R. Steuernage l and A.L. Kuck. 1990. Variation of Milk, Fat, Protein, and Somatic Cells for Da iry Cattle. J. Dairy Sci. 73:484-493. Sheldrake, R.F., R.J.T. Hoare and G.D. Mc Gregor. 1983. Lactation Stage, Parity and Infection Affecting Somatic Cells, Elect rical Conductivity and Serum Albumin in Milk. J. Dairy Sci. 66:542. Smith, A., F.K. Neave, F.H. Dodd and G.C. Brander. 1966. Methods of Reducing the Incidence of Udder Infection in Dry Cows. Vet. Rec. 79:233. Sordillo, L.M., K. Shafer-Weaver and D. DeRosa. 1997. Immunobiology of the Mammary Gland. J. Dairy Sci. 80:1851-1865. Webb, D. Unpublished data. University of Florida. Gainesville, FL. 2004. Weiss, W.P., J.S. Hogan and K.L. Smith. 2004. Changes in Vitamin C Concentration in Plasma and Milk from Dairy Cows After an Intramammary Infusion of Escheria Coli. J. Dairy Sci. 87:32-37 Weiss, W.P., J.S. Hogan, K.L. Smith a nd K.H. Hoblet. 1990. Relationships Among Selenium, Vitamin E, and Mammary Gland H ealth in Commercial Dairy Herds. J. Dairy Sci. 73:381-390. Wilson, C.D., R.G. Kingwell. 1975. A Practical Mastitis Control Routine. Proc. Seminar on Mastitis Control. Int. Dair y Fed. Bull. Reading, England. Doc. 85:62.

PAGE 72

61 BIOGRAPHICAL SKETCH Jessica Elizabeth Belsito was born in Millbury, Massachusetts, on January 31st, 1981. She grew up surrounded by th e dairy industry. While in Connecticut, the author spent many evenings and weekends milking at the Universitys dairy, which further solidified her love of the dairy industry. She was also heavily involved with the Universitys dairy club, Block & Bridle and Si gma Alpha. These activities all helped to further her knowledge of the dairy industry. The author decided to attend graduate school during her senior year of college. She graduated in May, 2003, with a Bachelor of Sc ience in animal scie nces and a minor in dairy management. Immediately following her graduation, she moved to Gainesville where she began work on her Master of Science.