Copyright is owned by the Author of the thesis.  Permission is given for 
a copy to be downloaded by an individual for the purpose of research and 
private study only.  The thesis may not be reproduced elsewhere without 
the permission of the Author. 
 



ANTIBIOTIC THERAPY FOR SUBCLINICAL t�ASTITIS 

IN EARLY LACTATION ; EFFECTS ON INFECTION, 

SOMATIC CE LL CO UNTS AND ��ILK PRODUCTION 

A THESIS PRESENTED IN PARTIAL 

FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE 

OF MASTER OF AGRICULTURAL SCIENCE 

IN ANH·1AL SCIENCE AT 

MASSEY UNIVERSITY 

SAMSON MWAKASAGULE MWAKIPESILE 

' '· 
.. 

1 982 



i i  

ABSTRACT 

This study was conducted with the main purpose of  inve stigat ing 

the ef fects of antibiotic therapy applied during lactation on the 

sub c l inica l ly infected quarters in terms of bacteria l infection , 

somatic cell count ( S . C . C . ) , milk yield and composition ( i . e .  mi lkfat 

and protein percentage ) .  The main ob j ective s involved were:-

(1) To identify cows with high somatic cell counts 

( subc l inica l  mastit i s )  in early lactation . 

( 2 )  T o  identify the individua l quar ters with high 

somatic c e l l  counts and bacterial infection . 

(3 )  To treat some of  these high S . C . C .  quarters with 

an antibiotic , and leave some of them untreated 

( c ontrol quarters )  . 

( 4 )  T o  mea sure the effects o f  treatment on S . C . C . , 

bacterial infection , y ie l d  and compo sition of  

milk . 

During weeks four to six of  lactation 1 2  cows from a herd o f  

1 0 0  cows a t  No . 3  Da iry Re search & Deve lopment Uni t ,  Massey Univer sity , 

were identifie d  to have consistently high value s  for somatic c e l l  

counts . Milk samp l e s  were aseptica lly taken from the individual 

quarters of the se cows and subjected to bacteriological test s . E ight 

sub c l inically infec ted quarters with high somatic cell counts were 

sub sequently treated with sodium c loxac i l l in (Orbenin) formulated in 

a s l ow release base . 

Before antib iotic treatment of any quarters, there was a c lose 

association between infection and somatic cell counts in individual 

quarters .  Thus only 12% of  the quarters which showed no infection 

had somatic cel l  c ounts higher than 3 0 0 , 00 0  c e l l s /ml, whereas the 

corresponding proportion of infected quarters was 74% . 



i i i  

Out of the e ight infected quarters which were treated with 

ant ib iotic , f ive were cured o f  infect ion ( 6 2 . 5 % cure rate ) ; in 

the se quarters the average S . C . C .  was greatly reduced from 

4 , 2 0 7 , 000 c e l l s /ml before treatment to about 160 , 000 c e l l s/ml 

afterwards whereas the three quarters wh ich were treated but not 

cured showed a s l ight decrea se from 3 , 99 1 , 20 0  cell s/ml be fore 

trea tment to 2 , 6 3 8 , 800 c e l l s /ml after treatment . The average 

somatic cell counts o f  the five infected control quarters increa sed 

sl ightly from the ir original va lue o f  about 2 , 061 , 500 cells/ml up 

to 2 , 1 1 1 , 000 c e l l s/ml . 

There wa s no s ignificant or cons istent e ffect of antib iot ic 

therapy , succe s s ful or otherwise , on milk yield from individua l 

quarters . However adj usted means showed a 1 0 %  non-signif icant 

dif ference be tween the succes sfully treated quarters and the 

untreated control quarters in favour o f  the cured quarters . The 

effects of treatment on mi lk compos ition were smal l  and were 

s ignif icant only for protein percentage . 

The practical implications o f  the re sul t s  were discus sed . 

From the re sults obtained firm conc lus ions on the effects of  

antibiotic therapy of  subc linica lly infected quarters on mi lk yield , 

and compo sition were not possible . The ma j or benefit i s  likely to 

be the reduc tion in the number of infec ted quarters , and the 

consequent reduction in the risk of  new infection in other quarters . 

It  i s  however sugge sted that more work i s  required to e s tabl i sh the 

actua l effect of antib iotic therapy by obtaining information on the 

performance of the succes sfully treated quarters in the fol lowing 

lactation . 

* * * * * * * * * * * 



ACKNOWLEDGEMENTS 

It is a pl easure to record my grati tude to many people who 

a s s i sted in this work . My sincere thank s are extended to the 

fol lowing : 

Dr C . W .  Ho lme s  for hi s invaluable continuing guidance and 

encouragement as my supervisor . He helped with part o f  

the data c o l lection and of fered c on structive critic i sm 

and guidance in writing up thi s the si s .  His advice , help 

and con cern with my we l fare are highly apprec iated . 

Profes sor D . S .  Flux o ffered advice on the stati stical part 

o f  the work . His wil l ingne s s  to discuss problems on 

stati stical procedure s even at short notice made an 

inva luable c ontr ibution towards the author ' s  readine s s  

t o  tackle the analyses .  

iv 

Mrs Yvonne Moore a s s i sted in data collection and submi s sion 

of milk samp l e s  to the Livestock Improvement Assoc iation 

Laboratory in Pa lmerston North for somatic cell count 

determinat ion s .  Her assi stance and excellent co-operation 

he lped to make this study more en j oyable than it otherwi se 

might have been . 

Mi s s  Janis Frain and Mr Je ffrey Raven helped in the analy s i s  

o f  milk c ompo s ition (milkfat and protein percentage ) .  

Mi s s  Frain a l so he lped in the Bacter io logical analy sis  

work . Her a s si stance i s  highly apprec iated . 

Me ssrs A .  Lowe , M .  Chester f ield and s .  Harmer - sta f f  at No . 3  

Dairy Re search & Deve lopment Unit , a s s i sted in the routine 

separation of the experimenta l cows into dif ferent gra z ing 

paddock s . 

Dr R . B .  Mar sha l l  and his staff in the Bacteriology Laboratory 

at the Faculty of Veterinary Science o f fered advice and 

help in the Bacteriological ana l y s i s  work and �n v�o 

sen sitivity te s t  of the pathogens isolated . 



Sta f f  o f  the L ive s tock Improvement As soc iation Laboratory , 

Palmers ton North , for carrying out somatic c e l l  count 

determinations .  

V 

Dr A . W . F .  Davey and Dr D . D . S .  MacKenzie f or their encouragement 

and concern with my wel l-being . 

Mr s Carolyn Ca stle , Secretary - Anima l Sc ience Department ,  for 

her a s s i s tanc e in c ler ica l work and excel lent co-operat ion . 

I t  i s  not po s s ible to l i s t  the name s o f  a l l  who contributed 

either direc t ly or indirectly towards thi s  study , but the author 

wi s he s to expre s s  hi s thanks to a l l  those who d i scussed with him 

matters related to this work and are not ment ioned individually . 

Notable among the se are the Sta f f  o f  the An ima l Science Department ,  

Ma s s ey Un iver s ity . The exce l l ent co-operat ion of the Library sta f f  

is a l so acknowl edged with thank s .  

The Helen E .  Akers Scholar ship wa s awarded to the author and he 

expresses his apprec iat ion to the Trustee s of the Scho lar ship for 

thi s kind of a s s i stance . 

The author w i shes to record his gratitude to the New Zealand 

and Tanzania Governments for mak ing it po s s ib l e  for him to undertake 

po s tgraduate studies at Ma s sey Univers ity . 

Sincere thanks are due to Mr s Va lerie Oram for the skil ful and 

e f f i cient manner in which she typed the manuscript . 

Fina l ly , I wish to expre s s  special thanks to my family for the ir 

continued encouragement and understanding throughout the period I 

have been away from them . 

* * * * * * * * * * * 



TABLE OF CONTENTS 

TIT LE PAGE 

ABSTRACT 

ACKNOWLEDGEMENTS 

TABLE OF CONTENTS 

LIST OF TABLES 

LIST OF FIGURES 

CHAPTER ONE 

1 : 0  

CHAPTER TWO 

2 : 0  

2 :  1 

2 : 1 : 0 

2 :  1 :  1 

2 : 1 : 2 

INTRO DUCTION 

LITERATURE REVIEW 

INTRODUCTION 

PART I - MASTITIS IN GENERAL 

DEFINITION 

CAUSE 

PAGE 

i 

ii 

iv 

vi 

X 

xiv 

1 

1 

3 

3 

3 

3 

3 

2 : 1 : 3 CLINICAL FORMS OF MASTITIS 5 

2 :  1 : 4  DIAGNOSIS 6 

2 : 1 : 5 PREDISPOSING CAUSES 8 

2 : 1 : 6 MASTITIS CONTROL 1 3  

2 : 2 : 0  PART II - SUBCLINICAL MASTITIS 1 5  

2 : 2 : 1 INTRODUCTION 1 5  

2 : 2 : 2  ECONOMIC I MPORTANCE OF SUBCLINICAL MASTITIS 1 5  

2 : 2 : 3 : 0  DIAGNOSIS 2 2  

2 : 2 : 3 : 1 FACTORS INFLUENCING SOMATIC CELL COUNTS 2 8  

2 : 2 : 3 : 2  DIAGNOSTIC METHODS 3 2  

2 : 2 : 3 : 2 : 1 DIRECT METHODS FOR SOMATIC CELL DETERMINATION 3 2  

2 : 2 : 3 : 2 : 2  INDIRECT METHODS FOR SOMATIC CELL DETERMINATION 3 4  

2 : 2 : 3 : 3  BACTERIOLOGICAL DIAGNOSIS 3 7  

2 : 2 : 4  CONTROL METHODS 4 1  



CHAPTER THREE 

3 : 0  

3 :  1 

3 : 2  

3 : 3  

3 : 4  

CHAPTER FOUR 

4 : 0  

4 :  1 

4 : 2  

4 : 2 :  1 

4 : 3  

4 : 3 : 1 

4 : 3 : 2  

4 : 3 : 3  

4 : 3 : 4  

4 : 3 : 5  

CHAPTER F IVE 

5 : 0  

5 :  1 

512 

5 : 3  

5 : 4  

5 : 5  

CHAPTER S IX 

6 : 0  

MATERIALS AND METHODS 

RATIONALE 

PROCEDURE AND PLAN 

EXPERIMENTAL PROCEDURE 

METHODS 

RESU LTS 

INTRODUCTION 

SOMATIC CELL COUNTS AND INFECTION 

THE RELATIONSHIP BETWEEN SOMATIC CELL COUNTS 
AND QuARTER MILK YIELDS 

EFFECTS OF ANTIBIOTIC TREATMENT 

INTRODUCTION 

BACTERIAL INFECTION 

SOMATIC CELL COUNTS 

MILK YIELD 

MILK COMPOSITION 

D ISCUSSION 

INTRODUCTION 

THE ASSOCIATION BETWEEN S . C . C .  AND BACTERIAL 
INFECTION 

EFFECT OF ANTIBIOTIC THERAPY ON BACTERIAL 
INFECTION 

THE EFFECTS OF ANTIBIOTIC THERAPY ON S . C . C . 
MILK YIELD AND COMPOSITION 

THE PRACTICAL APPLICATION OF ANTIBIOTIC 
THERAPY FOR SUBCLINICAL MASTITIS 

CONCLUSION 

PAGE 

47 

47 

47 

47 

48 

51 

5 5  

5 5  

5 5  

5 5  

58 

59 

59 

59 

61 

65 

6 5  

71 

71 

71 

71 

74 

77 

82 

85 

85 



PAGE 

APPEND I CES 

1:0 FOR DATA COVER I NG F I RST MONTH OF EXPER I MENTAL 
PER I OD .  SUMMARY OF ANALYS I S  OF VARI ANCE AND 
COVAR I ANCE ANALYS IS  OF DATA OF TREATED AND UN­
TREATED ( CONTROL ) QUARTERS I N  RESPECT OF 
SOMAT I C  CELL COUNT , M I LK Y I ELD , FAT AND PROTE I N  

86 

CONCENTRAT ION I N  THE M I LK .  86 

1 :  1 

1:2 

EFFECT OF TREATMENT ON S . C . C .  OF SUBCL I N I CALLY 
I N FECTED QUARTERS ( LOG CELLS/ML ) .  

EFFECT OF ANT I B I OT I C  TREATMENT ON M I LK Y I ELD OF 
SUBCL I N I CALLY INFECTED QUARTERS ( KG QUARTER-1 

DAY- 1 ) .  

1 : 3 EFFECT OF ANT I B I OT I C  TREATMENT ON MI LKFAT CON-

1 : 4  

CENTRAT I ON ( %  FAT ) OF SUBCL I N I CALLY I NFECTED 
QUARTERS. 

EFFECT OF ANT I B I OT I C  TREATMENT ON M I LK PROTE I N  
CONCENTRAT I ON ( %  PROTE I N ) O F  SUBCL I N I CALLY 
I NFECTED QUARTERS . 

2:0 FOR DATA FROM WHOLE EXPER I MENTAL PER I OD .  SUMMARY 
OF ANOVA AND COVAR I ANCE ANALYS I S  OF DATA OF 
TREATED AND CONTROL QUARTERS I N  RESPECT OF S . C . C . , 
M I LK Y I ELD , FAT AND PROTE I N  CONCENTRAT I ON ( % )  I N  

86 

88 

91 

93 

THE M I LK .  96 

2: 1 

2:2 

EFFECT OF TREATMENT ON S . C . C .  OF SUBCL I N I CALLY 
I NFECTED QUARTERS ( LOG CELLS/ML ) .  

EFFECT OF ANT I B I OT I C  TREATMENT ON M I LK Y I ELD 0�
1 

SUBCL I N I CALLY I NFECTED QUARTERS ( KG QUARTER 
DAY- 1 ) .  

2:3 EFFECT OF ANT I B I OT I C  TREATMENT ON M ILKFAT CON-
CENTRATI ON ( %  FAT ) OF SUBCL I N I CALLY I NFECTED 
QUARTERS . 

96 

98 

1 0 1  



2:4 EFFECT OF ANT I B I OT I C  TREATMENT ON MILK 
PROTE I N  CONCENTRAT I ON ( %  PROTE I N )  OF 
SUBCL I N I CALLY I NFECTED QUARTERS . 

2 : 5  FOR DATA COVERI NG THE FIRST MONTH OF EXPERI­
MENTAL PER I OD .  SUMMARY OF COVAR I ANCE ANALYSES 
OF DATA OF SUCCESSFULLY TREATED AND UNTREATED 
QUARTERS I N  RESPECT OF THE I R  SOMAT IC CELL 

PAGE 

103 

COUNT , r.1ILK Y I ELD , FAT AND PROTE I N  PERCENTAGES . 1 0 5  

2:6 FOR DATA COVER I NG THE WHOLE EXPER I MENTAL PER I OD.  
SUMMARY OF COVAR I ANCE ANALYSES OF DATA OF 
SUCCESSFULLY TREATED AND UNTREATED QUARTERS I N  
RESPECT O F  THE I R  SOMAT IC  CELL COUNT , M I LK Y I ELD , 
FAT AND PROTE I N  PERCENTAGES . 

3 : 0  FOR DATA COVER I NG ONE MONTH OF EXPERI MENTAL 
PER IOD .  TESTS OF S I GN I FI CANCE OF MEAN 
D I FFERENCES OF TREATED AND UNTREATED QUARTERS I N  
RESPECT O F  S . C . C . , M I LK YIELD , FAT AND PROTE I N  

107 

CONCENTRAT I ON I N  THE M I LK .  109 

4:0 FOR DATA COVER I NG THE WHOLE EXPER I MENTAL PER I OD .  
TESTS O F  S I GN I F I CANCE OF MEAN D I FFERENCES OF 
TREATED AND UNTREATED QUARTERS IN RESPECT OF 
S . C . C . , M I LK Y I ELD , FAT AND PROTE I N  CONCENTRAT I ON 
IN  THE M I LK .  117 

5:0 COMPOSI TE*M I LK S . C . C .  OF EACH COW TOGETHER W I TH 
HER s . c . c .  FOR I ND I V I DUAL*

QuARTERS . 1 25 

BIBLIOGRAPHY 126 

* * * * * * * * * * * 



TABLE 

2 .  1 

LIST OF TABLES 

ASSESSMENT OF CYTOLOGI CAL-BACTER I AL F I NDINGS 
IN MAST I T I S  D I AGNOS I S  ( TDLLE , 1 9 75 ) .  

2 . 2  THE RELAT I ON BETWEEN S . C . C . OR I NFECT I ON ,  AND 

2 . 3  

M I LK PRODUCT I ON FROM I ND I V I DUAL QUARTERS OF 
AN UDDER OR FROM I ND I V I DUAL COWS . 

CHANGE I N  % COMPOS I T I ON ASSOC I ATED W I TH AN 
I NCREASE OF 1 00 , 000  CELLS/ML I N  BULK SOMAT I C  
CELL COUNT . 

2 . 4  THE RELAT I ON BETWEEN B . S . C . C .  AND CHEDDAR CHEESE 
Y I ELD ( HOARD'S DA I RYMAN , MARCH 25 , 1 980 ) . 

2 . 5 THE RELAT I ONSH I P  BETWEEN THE I NC I DENCE OF MAST I T I S  
PATHOGENS AND CELL COUNT I N  I ND I V I DUAL COWS 

PAGE 

4 

17 

19 

21 

( MACM I LLAN & DU I RS, 1 9 80 ; N.Z. ) .  23 

2 . 6 MEAN GEOMETR I C  S . C . C .  FOR QUARTERS OF KNOWN 
I NFECT I ON STATUS ( CELLS/ML ) X 1 03 ( SHELDRAKE , 
19 82 ; AUSTRAL I A ) . 

2 . 7 THE RELAT I ONSH I P  BETWEEN SOMAT I C  CELL COUNT AND 

2 . 8  

2 . 9  

I NFECT I ON ( HOARD'S DA I RYMAN , MARCH 25 , 1 982 ; 
VIRG I N I A  POLYTECHN I C  INST I TUTE & STATE UN I VERS I TY ,  
u . s . A .  >. 

INFECT I ON RATE OF COWS W I TH I N  VAR IOUS SOMAT I C  CELL 
COUNT RANGES IN A PENNSYLVAN I A  STUDY ( ST I LES & 
RODENBURG , 1 98 1 ) .  

CELL COUNT LEVELS RELATED TO SUBCLI N I CAL MAST I T I S  
< 0U I RS , 1 980 ; N . Z .  ) .  

2 . 10 CELL COUNTS I N  NORMAL M I LK ( M I LL I ONS/ML ) ( CULLEN , 

2 .  1 1  

1966 ) .  

SOMAT I C  CELL COUNTS I N  GLENFI ELD DA I RY AVERAGE 
< HDARE , 1 977 C I TED BY G I LL ,  1 9 77 ) .  

24 

2 5  

26 

27 

28 

30 



' 

TABLE 

2 . 1 2  

2 . 1 3  

RELATI ONSH I P  BETWEEN % GAS AND SOMAT I C  CELL 
NUMBERS PER ML ( FROM SCHALM � a£, 1 97 1 ) .  

EST I MAT I ON OF THE CELL CONTENT OF M I LK BY CMT 
SCORES ( SCHALM � a£, 1 9 7 1 ) .  

2 . 1 4 THE EFFECT OF ANT I B I OT I C  ( CLOXAC I LL I N )  THERAPY 
AT DRYI NG-OFF ON I NFECT I ONS PRESENT AT DRY I NG­
OFF, NEW I NFECT IONS I N  THE DRY PER I OD AND 

PAGE 

3 5  

36 

LEVELS OF I NFECT I ON AT CALVING ( GR I FFIN , 1 9 7 1 ) .  4 2  

2 . 1 5 SUMMARY OF EFF I CACY DATA ON DRY THERAPY < PANKEY 
� a£, 1 98 2 ; N . Z . ) .  

2 . 1 6 THE RESPONSE TO THERAPY G I VEN DUR I NG LACTAT I ON TO 
SUBCL I N I CAL STAPHYLOCOCCAL INFECT I ON I N  RELAT I ON 
TO THE STAGE OF LACTAT I ON AND AGE OF THE TREATED 
COWS AND THE WHI TES I DE SCORE OF THE INFECTED 
QUARTER . THE RESULTS ARE THE MEAN OF VAR I OUS 

4 3  

SODI UM CLOXAC I LL I N  FORMULATI ONS ( GR I FF I N, 1 9 7 1 ) .  4 5  

3 .  1 

3 . 2  

BREED, AGE AND AVERAGE * S . C . C .  I N  I ND I V I DUAL 
QUARTERS OF THE 1 2  SELECTED COWS . 

PA I R I NG OF QUARTERS I N  EACH TREATMENT GROUP . 

4 . 1 MEAN CELL COUNTS OF QUARTERS I DENT I F I ED AS I �FECTED, 
NON- I NFECTED AND DOUBTFUL PR I OR TO ANT I B I OT I C  
THERAPY ( AVERAGE O F  3 I ND I VI DUAL MEASUREMENTS OF 
SOMAT I C  CELL COUNTS TAKEN DUR I NG THE PREL I M I NARY 

4 9  

50  

PER I OD ) . 56 

4 . 2  THE RELAT I ONSH I P  BETWEEN THE INCI DENCE OF MAST I T I S  
PATHOGENS AND CELL COUNTS . 

4 . 3  COVAR I ANCE ANALYS I S  OF QUARTER M I LK Y I ELD I N  RELAT I ON 
TO SOMAT I C  CELL COUNTS . 

5 7  

58 



TAB LE 

4 . 4  THE RESPONSE TO ANT I B I OT I C  THERAPY ( SODIUM 
CLOXAC I LL I N  200 MG ) G I VEN I N  EARLY LACTAT I ON 
TO SUBCL I N I CALLY I NFECTED QUARTERS . 

4.5 SUBCL I N I CALLY I NFECTED QUARTERS AND THE IR MEAN 
CELL COUNTS BEFORE AND AFTER TREATMENT ( MEAN 
CELL COUNTS ARE AVERAGES OF 3 AND 2 7  I ND I V I DUAL 
MEASUREMENTS OF S . C . C . TAKEN DUR I NG THE 
PREL I M I NARY AND EXPER I MENTAL PER I ODS , RESPECT­
I VELY ) . 

4 . 6  MEAN CELL COUNTS OF TREATED AND UNTREATED QUARTERS 
( AVERAGES OF 3 AND 27 I ND I V I DUAL MEASUREMENTS 
OF S . C . C .  TAKEN DUR I NG THE PREL I M I NARY AND 
EXPER I MENTAL PERI ODS , RESPECT I VELY ) .  

4 . 7  EXAMPLE OF ANALYS I S  OF COVAR I ANCE FOR TEST I NG THE 
D I FFERENCE BETWEEN ADJUSTED MEANS OF TREATED AND 
UNTREATED QUARTERS FOR S . C . C . ( FOR DATA COVER I NG 
F I RST MONTH OF THE EXPER I MENTAL PER I OD ) . 

4 . 8  EXAMPLE OF ANALYSI S  OF VAR I ANCE FOR TEST I NG THE 
D I FFERENCE BETWEEN ADJUSTED MEANS OF TREATED AND 
UNTREATED QUARTERS FOR S . C . C . ( FOR DATA COVER I NG 
F I RST MONTH OF THE EXPER I MENTAL PER I OD ) . 

4 . 9  SECOND EXAMPLE OF ANALYS I S  OF COVAR I ANCE FOR TEST I NG 
THE D I FFERENCE BETWEEN ADJUSTED MEANS OF TREATED 
AND UNTREATED QUARTERS FOR S . C . C . ( FOR DATA 
COVER I NG THE WHOLE EXPER I MENTAL PER I OD ) . 

4 . 1 0 MEAN CELL COUNTS OF SUCCESSFULLY AND UNSUCCESSFULLY 
TREATED QUARTERS ( AVERAGES OF 3 AND 27 I ND I V I DUAL 
MEASUREMENTS OF S . C . C .  TAKEN DURI NG THE PREL I M I NARY 
AND EXPERI MENTAL PER I ODS , RESPECTI VELY ) .  

4 . 1 1  MEAN CELL COUNTS OF UNTREATED ( CONTROL ) AND SUCCESS-
FULLY TREATED QUARTERS ( AVERAGES OF 3 AND 27  

PAGE 

60 

62 

63 

64 

64 

64 

66 

I ND I V I DUAL MEASUREMENTS OF S . C . C .  TAKEN DUR I NG THE 
PREL I M I NARY AND EXPER I MENTAL PER I ODS , RESPECT I VELY ) . 67. 



TABLE 

4.12 AVERAGE M I LK Y I ELD OF TREATED AND CONTROL 
( UNTREATED ) QUARTERS ( KG/QUARTER/DAY ) .  

4.13 AVERAGE M I LK Y I ELD OF SUCCESSFULLY TREATED AND 
CONTROL QUARTERS ( KG/QUARTER/DAY ) .  

4.1 4 AVERAGE M I LKFAT AND PROTE I N  CONCENTRAT I ON ( % )  
OF TREATED AND UNTREATED ( CONTROL ) QUARTERS .  

4.15 AVERAGE M I LKFAT AND PROTE I N  CONCENTRAT I ON ( % )  
OF SUCCESSFULLY TREATED AND UNTREATED 
( CONTROL ) QUARTERS . 

5 .1 ACCURACY W I TH WH I CH SUBCL I N I CALLY I NFECTED 

5.2 

5.3 

5 . 4  

QUARTERS CAN BE I DENT I F IED FROM VALUES FOR 
S . C . C . , US I NG FOUR THRESHOLD VALUES TO I ND I CATE 
BACTER I AL I NFECT I ON ,  BASED ON THE AVERAGE S . C . C . 
FROM THREE SAMPLES . 

THE PROPORTION OF SUBCL I NC I AL I NFECT I ONS EL I M I NATED 
WHEN TREATED*W I TH SODI UM CLOXACILL I N  FORMULATED 
I N  A SLOW RELEASE BASE . 

THE PROPORT I ON OF CL I N I CAL I NFECT I ONS EL I M I NATED 
WHEN TREATED*W I TH SODI UM CLOXAC I LL I N  FORMULATED 
I N  FAST RELEASE BASES ( GR I FF I N  1971). 

RETURNS FROM ANT I B I OT I C  THERAPY OF SUBCL I N I CAL 
MAST I T I S  DUR I NG LACTAT I ON .  

* * * * * * * * * * * 

PAGE 

68 

68  

6 9  

70  

72  

76  

76 

8 3  



FIGURE 

2. 1 

LIST OF FIGURES 

RELAT I ONS BETWEEN SOMAT I C  CELL COUNT , M I LK 
COMPOSI T I ON AND M I LK PRODUCTI ON FROM 
I ND I V I DUAL QUARTERS CHEESCHEN , 1975). 

* * * * * * * * * * 

PAGE 

1 6  



1 

CHAPTER ONE 

1 : 0  INTRODUCTION 

The use of somatic c e l l  counting sy stems for mon itoring ma stitis 

status ( the subc l inical form) of herds and the udder health of 

individual cows is being widely accepted in many countries throughout 

the world . A large amount of survey data and analytical results is  

ava i l able from many countries that show the relation between udder 

infec tion and somatic c e l l  count ( S . C . C . ) . Studies in New Zealand 

( re view by Holmes , 198 1 ) , for example , have shown that there is a 

c lear association between udder infection and somatic c e l l  count in 

the milk from individual c ows , and that on average , c ows with high 

S . C . C .  produce less milk than cows with lower S . C . C .  Also work in 

U . S . A .  ( Schultz , 1977 ) has shown that loss o f  milk produc tion due to 

sub c l inical ma stitis is re lated to somatic ce l l  counts . The report 

indicates that on a quarter basis , loss starts at 500 , 00 0  c e l l s/ml , 

progre sses to 7 . 5% at 1,000,000 c e l l s/ml and 3 0 %  at 5 mi l l ion cell s/ml . 

Barnard ( 1 9 8 1 )  reports that a s  a genera l rule , for each increase o f  

100 , 00 0  cells/ml , there will  be about a 3%  increase i n  infected 

quar ters and a decrease o f  about 0 . 4 5 kg of mi lk per cow daily . 

Recent information pre sented by George Shook o f  University of Wi scon sin , 

U . S . A .  at the annual mee ting o f  the National Ma stiti s Council held in 

mid February , 1982 in Louisvi l le , Kentucky , U . S . A .  ( Hoard ' s  Dairyman , 

March 25 , 1 9 8 2 ) , showed that average dai ly production of Wi sconsin 

cows dropped from 2 1 . 9  kg to 2 1 . 3  kg as  the somatic c e l l  count rose 

from 5 0 , 00 0  to 100 , 000 c e l l s/ml . Ali et al ( 1980 ) have reported that 

rennet c lotting time ( R . C . T . ) , los s of fat in whey , curd mo isture and 

los s e s  in curd yield and r igidity were all greater in the milk of 

higher somatic cell counts . S ince somatic c e l l  s tandards may be 

app l ied to milk used in the manufacture of dairy products , it might in 

the long run become increasingly important to l imit the number o f  the se 

cel l s  in milk . 

I t  is  on the basis  o f  such findings that an inc rease in the 

somatic ce l l  count in the milk of an individua l cow is now used as a 

mea surement o f  subclinical mastit i s , and with the introduc tion of 

individual cow S . C . C .  service in countries l ike New Zealand , the 



2 

farmer can u se the cow's S . C . C .  to detec t cows which are subc l inical ly 

affected and take appropriate control measure s .  The loss  in milk 

production c ited above and the probable reduced yield and qua l ity of 

dairy products ( such as  chee se ) made from mi lk o f  high somatic ce l l  

coun t s  is o f  great concern t o  the dairy farmer and the dairy industry 

at l arge and calls  for measures to avo id such a los s .  

The e f fect oftreating subc linical ma stitis during lac tation has 

been investigated in some overseas c ountrie s .  The re sponse obtained 

in U . K .  with antibiotic therapy given to subc l in ically affected 

quarters ha s shown that subclinical mastitis can be treated with 

ant ibiotic quite succe s s ful ly , with the success varying between 

d i f ferent c ausal organ isms ( Gr iffin , 19 7 1 ) . Such information is 

scanty and seems to be l acking in New Zealand . Additiona l information 

on the contribution of this method of mastitis control is therefore 

de sirable , particularly when cons idering the loss  caused by sub­

c l in ical ma stitis in a lactating dairy cow . Kingwill et a£ ( 19 7 9 )  

sugge sted that another method of reduc ing the duration of infections 

is  to treat a l l  new in fections during lactation . Thi s study was 

there fore intended to provide data which should provide information 

about the re sponse to treatment in terms of bacterial infection , 

somatic ce l l  count , milk yield , mi lk fat and mi lk prote in percentage s .  

Thi s  was accomplished by treating with an antibiotic some of the 

quarters found to be sub c l inically af fected and then measuring the 

sub sequent changes in the somatic c e l l  count , bacterial infection , 

mil k  production and percent o f  both milk fat and milk prote in of the se 

quarters and comparing them with quarters which were subc l inically 

a f f ec ted but not treated . 

* * * * * * * * * * * 



3 

CHAPTER TWO 

2 : 0  LITERATURE REVIEW 

2a 1 I NTRODUCTI ON 

Thi s  chapter i s  divided into two parts , I and I I . Part I is  

devoted to a brief discussion o f  ma stitis in genera l , whereas Part 

II dea l s  with subc linic a l  mastitis in some greater deta i l . The 

ava il ab l e  l i terature on ma stitis is voluminous and therefore on ly 

repre sentative reports will  be c ited in this review . 

2 :  1 : 0 PART I - MASTITIS IN GENERAL 

2 : 1 : 1  DEFI N I T I ON :  Mastiti s ,  which can generally be de fined a s  an 

inf lammation of  the mammary gland , i s  a complex disease having 

dif ferent cause s ,  degrees of intensity and variations in duration 

and r e sidua l e f fects (Carrol , 1 977 ; Schalm and Woods , 1 9 5 3 ) . The 

definition o f  mastitis arrived at by the International Dairy 

Federation ( IDF ) require s the f inding of  pathogenic micro-organ i sms 

in the milk and a l so a con centration of somatic c e l l s  in exce s s  of 

500,000 cell s/ml ( Int . Dairy Fed . , 1 97 1 ) . 

Tol le ( 197 1 )  further e l aborate s on the de f inition that : 

"Ma s t it i s  is  an inflammatory change of the mammary gland which , along 

with phy sic a l , chemical and microbiological change s , is charac terized 

by an increase of somatic c e ll s , e specially leucocytes in milk , and 

by patho logical changes in the mammary tis sue . "  

Ob j ections have been raised to the criteria embodied in thi s  

def i n ition ( see Proc . I DF Seminar·on Ma stitis Control , 1 97 5 :  

G i e se cke , 1974 ) ; nevertheless the criteria draw spec ial attention 

to two a spects of masti t i s  which are of ma jor importance , namely : 

a )  the respon se by o r  damage to , the g land t i s sue 

b )  the presence of  pathogen ic bacteria i n  the gland . 



4 

The International Dairy Federation ( IDF ) de finition can be 

summarized as  shown in Table 2. 1.  

TABLE 2 . 1 :  ASSESSMENT OF CYTOLOG ICAL-BACTER IAL F I ND I NGS I N  MAST I T I S  
D I AGNOS IS  ( TOLLE, 1975). 

Cell Count Pathogen ic Micro-organ i sms 

per ml milk 

<500 , 000 

>500 , 000 

Not I so lated 

norma l secretion 

non - specific mastitis 

I solated 

latent infection 

mastitis 

This de finition applies to the examination o f  quarter milk samples 

taken at the usual mi lking time from the foremi lk after di scarding 

the first few squirts of milk . 

The main drawback o f  this de finition , however , is  that udder 

infec tion is not def ined and this has led to various bacterial 

samp l ing methods and cultural procedure s ( see Part I I )  resulting in 

cons iderable con fus ion between workers in the diagnosis of the disease . 

De spite the disparity in the sampl ing technique s involved it can be 

conc luded from the different de finitions appearing in the literature 

that the diagnosis of mastitis in lactating cows i s  based on : 

( i )  the c l inical examination o f  the udder and its 

secretion ; 

( ii )  the bacteriological examination o f  mi lk ; 

( i i i )  the somatic c e l l  count o f  milk . 

2:1:2 CAUSE : Most common ly the disease begins as  a result o f  pene­

tration of pathogen ic bacteria through the teat duct into the inter ior 

of the gland where , if conditions are favourable , they multiply and 



5 

the i r  by-products of growth and metabol i sm irritate the de licate 

t i s sues caus ing an inf lammatory response ( Scha lm et al, 1 97 1 )  . The 

inflammation grossly charac teri zed by high temperature , redne s s , 

pain and swe l l ing together with alterations in the con si stency and 

appearance of  the milk i s  known as " c l inical ma stitis " ,  whereas the 

exi stence of inflammation in the absence of observable signs or any 

visual abnormal ities in the mi lk i s  re ferred to as " subc l inical 

mas t itis" ( Duirs , 1980) . 

Detect ion of c l inical mastitis i s  much s impler because the vi sual 

symptoms can be easily recognized . The main pathogen ic bacteria of 

economic importance in ma stitis are of the S treptococc i  group ( i . e .  

StA eptoeoee� agaiact{ae ,  StAeptoeoee� dy� gaiact{ae and StAeptoeoee� 

ub�) and Staphyfoeoee� aune� (Neave , 197 5 ) . 

2 : 1 : 3  CL I N ICAL FORMS OF MAST I T I S : 

( i )  Peracute Mastitis - In addition to the gross s igns of 

swe l l ing , heat , redne s s  and pain , this form is accompanied by 

sy stemic sign s of fever , depre s sion , shiver ing , loss of 

appetite and rapid loss o f  we ight . 

( ii )  Acute Ma stitis - This i s  also characterized by the 

gross signs of swe l l in g ,  heat , redness and pain which are 

accompanied by fever and mi ld depre ssion . 

( i i i )  Sub-acute or M i l d  Ma stitis - Thi s  is the form of 

mastitis in which milk shows abnormal con s i s tency or colour 

o f  milk . In this form the gros s  s igns of mastitis are 

subdued and not accompanied by sy stemic e f fe c ts . 

( iv) Chronic Mastitis  - Thi s  i s  the form of  mastitis in 

which the cow ' s  milk per iodically shows abnormal ities and 

then returns to normal i . e .  the condition per iodically f lares­

up and then returns to normal a fter some t ime ( Schmidt , 197 1 ) . 

I t  occurs when a quarter fai l s  to respond to treatment over a 

period o f  time . The quar ter may atrophy or show abnormal 

c l inical changes for the rest of  the animal's l i fe . 

Other forms of mastitis  such a s  latent infection s , in which mi lk 



6 

shows the pre sence of  pathogenic micro-organi sms but has a normal c e l l  

count or non-specific or aseptic mastitis i n  which there is n o  recog­

n isable infection and the symptoms may be sub c l inical or c linical , 

could probably be c lassed under the category of  subclinical mastiti s 

because they often do not show overt signs . Subc l inical form of 

ma stitis is di scus sed in Part I I  but it could s imply be de fined here 

as the form of mastitis that shows no macroscopic evidence of in flam­

mation but examination of the milk reveals udder infection , increa sed 

c e l l  count and also alteration s in the chemic a l  propertie s of  milk . 

2 : 1 : 4 DIAGNOSIS : Since ma stitis is  a disease o f  the udder , the milk 

can be used to diagnose the di sease . The results of cytological , 

enzymat ic , b iochemical and bacteriological analysis of mi lk wi l l  show 

the response s of the udder to mastitis and e f fects of the disease on 

the udder . Some worker s  use bacterial examination almost exclusive ly 

in the diagno sis of mastitis ( e . g .  Neave , 1 97 5 )  whereas other s use a 

comb ination of  somatic c e l l  count and bacteriological information ( e . g .  

Pearson and Greer , 1974 ) . 

Early l iterature sugge sts that c l inical mastitis in cows caused 

problems serious enough to attract the attention of sc ientists .  The 

s c ienti fic l iterature reveals a di stinct tendency to de scribe , c l a s s ify 

and def ine mastitis in terms of  ava ilable diagnostic technique s and 

in formation on the pathogene sis ( see review by Gie secke and Van den 

Heeve r , 1974 ) . Early observations on the disease were therefore 

l imited to the c linical s igns observed by examination of  udder tis sue 

and s ecretion . Schalm et al ( 197 1 )  de scribed a method of physically 

e xamining the lactating udder immediately after milking . They reported 

that thorough evacuation of  the udder is nec e s sary because the pre sence 

of milk reduce s the chances of detecting scar t i s sues and f ibrosis . 

King ( 1974 )  us ing c l inical examination by manual palpation showed 

that quarter s  with distinct or marked fibrosis gave lower yie lds . The 

author state s that the e ffect on yield was due to udder damage . He 

further cite s the work o f  Udal l  & Johnson ( 19 3 1 )  who consider that the 

value of a phy sical examination in the diagnos i s  of mas titis rate s above 

any o ther s ingle method . Switzer and Gates ( 19 3 1 )  cited by King ( 19 7 4 )  



7 

c ompared the results of  post-mortem examination of  udders with the 

physical examination note s on the se cows as  made in routine field 

reports . Re sults obtained proved that a physical examination i s  a 

very e f ficient means of detect ing c linical mastitis under practical 

field conditions . 

Giesecke & Van den Heever ( 197 4 )  have c ited the work of a number 

of research workers who regarded c l inical examination of the udder to 

be very use ful as it detected not only acute conditions but a l so 

permanent udder change s due to chronic mastiti s .  Narayan and Iya 

( 19 5 4 )  showed that tissue abnormalities could be detected by deep 

palpations of the udder . In c ontrast , however , Giesecke & Van den 

Heever ( 1 974 )  have reported work of other s who recommended c l inical 

diagnosis mainly for acute mas titis and laboratory te sts as supple­

mentary methods . 

Some workers have cast doubts about the ef fic iency of phy sical 

examinations .  Gie secke & Van den Heever ( 1974 ) report the work of 

Heidrich and Renk ( 1963 ) which showed that with few except ion s ,  the 

phy s ical examination of the udder is inadequate for diagno sis of 

ma stitis because quarters with no induration may be bacteriological ly 

positive . Gie secke & Van den Heever ( 1974)  noted that discrepancie s 

between chron ic tissue alteration s and the presence or absence of 

inf l ammatory change s in the udder seem to result from the fact that 

in some indurated udder s  epithelial les ions persist whereas in others 

they heal complete ly . I t  is  there fore pos sib le that an incorrect 

diagnosis of  mastitis may be made in udders in which inflammation has 

sub s ided and the quarters have simply become indurated rather than 

mastitic . 

The strip-cup repre sents another form of c l in ical examination in 

which the foremi lk is examined using a black plate placed within the 

cup and the pre sence of sma l l  flake s , milk clots and water-like 

con s i s tency suggest the presence of mastitis . Neave et al ( 19 5 4 )  

reported that 1 2  - 1 3 %  o f  quarters showing c lots i n  milk were not 

infected with the more c ommon pathogenic organisms and hence its 

re l i abil ity as  an indication of  infection is  que stionable . Thi s  



indicates that there may be in stance s  where c lots occur in milk 

without actual infection . 

8 

The mo st recently deve loped technique of c l inical examination 

of mastitis is the individual in- l ine f i lters which show c lots 

suggesting the pre sence o f  mastitis in the herd . The only drawback 

of thi s method is that it involve s  going back to the c ow at milking 

time to trace the quarter ( s )  from which the c lots might have come . 

Neverthel e s s  thi s  would probab ly seem to be the most acceptable 

method of c l inical diagnos i s  o f  mastitis in big herds . 

I t  would appear from this discussion that c l inical examination 

of the udder i s  a very sub j ective technique and depends on the 

experience , skill and j udgement of the examiner . However when 

considering individua l farm conditions it is probably the only 

prac ticable method for diagno sing ma stitis though it might soon 

become a lost art in big herds where the mi lker cannot afford the time 

required to carry out an e f fective c l inical examination of  the udder 

of each cow in the herd . 

The importance of  early diagno sis of mastitis is to alert the 

farmer of impending problems before they become acute c l inical case s , 

and since such c l inical cases deve lop from sub c l inical ones , it might 

be bene fic ial to put more emphasis on the diagnosis of ma stitis at 

subc l inical level . Bacterial assessment and somatic c e l l  counting 

are the two main diagnostic criteria used for the asses sment of sub­

c l inical ma stiti s . Thi s type of diagno sis will be discussed at 

length in Part I I . 

2 : 1 : 5  PRED I SPOSI NG CAUSES : The severity of  mastitis is  determined 

to a considerable extent by the nature of the infecting bacteria , by 

natural mechanisms of  re s i s tance o f  the cow , and to some degree by 

str e s se s  p l aced on the mammary glands by milking practices and 

environmental factors ( Schalm et al, 1 9 7 1 )  . Stre s ses placed on the 

ma mmary glands by the mi lking practices and envi ronmental factors 

are the main factors considered as predisposing cause s in thi s  

context .  



9 

( i )  Mi lking Machine 

There is good prac tical and scientific evidence to show that 

the mi lking machine inf luences ma stitis on farms in other way s than 

the obvious one of transferring bacteri a  to the outside of the teat 

from one c ow to another or from one teat to another teat (Thie l , 

197 5 )  . Machine effects can be loose ly classified as those a s soc iated 

with over mi lking , with internal and external damage to teats and 

with vacuum f luctuations . 

Udder damage c an come from over milking , high vacuum level s  and 

inadequate squee ze phase (e . g .  Petersen , 1964) . Constant expo sure 

of the teats to vacuum leve ls above those normally used in milking 

machines i . e .  above 380 mm Hg ( above 1 5" Hg ) causes damage to the 

teat skin particularly when the teat-cup l iners do not c o l l apse 

complete ly during each pulsation ( Kingwill et al, 1979)  . P lastridge 

( 19 58)  c ites data reported by Burkey & Sanders ( 1 949 ) which sugge sted 

that a vacuum leve l  of over 380 mm Hg tends to increase the inc idence 

of ma stitis . Because higher operating vacuum levels w i l l  result in 

increased teat apex damage and/or teat sinus damage at the end of milk 

f low , it is reasonable to expect that thi s  will lead to a r i se in new 

infections . The po ssible explanation for that i s  bacteria might find 

a ready lodgement through teat eros ions where they could mul tiply and 

thereby increase the risk of infection of the udder . Peter sen ( 1 964 ) 

po stulated that severe teat inj ury c ou ld predispose the udder to 

infection and indicated that the teat canal could remain partly open 

i f  the milking machine was left on the teat after the end o f  milkf low 

thus al lowing easy penetration of any bacteria around . Langlois et al 

( 19 8 2 )  observed a s ignificant increase in teat end score a s  vacuum 

increased from 4 2 . 5  to 51 kPa . Total plate counts were s igni ficantly 

higher ( 16 67 c . f . u/ml milk) with 51 kPa than with 34 or 42 kPa vacuum 

level ( 7 7 1  and 7 2 3  c . f . u/ml milk , respective ly ) . The authors have 

also reported that 48 . 9% of quarter milk samples from cows milked at 

51 kPa were pos itive for StaphyiOQOQQ� au4e� ( the predominant cause 

of mastitis)  vs 1 9 . 9  and 16% for 3 4  and 4 2 . 5  kPa , re spectively . 

Reports on the e ffects of over milking as  a predi spos ing cause o f  

mastitis are c ontr adictory . Mochrie et al ( 1 953a, 1 9 5 3b , 1 9 5 5 ) , for 



1 0  

example , failed to re late the vacuum level and duration o f  milking 

with udder health . Petersen ( 1 964 ) reported that repeated over­

mi lking (on 16 to 26 occa s ion s )  for three to f ive minute s caused 

damage to the l ining of the teat s inus . Milking cows for e ight 

minutes every milk ing for a ful l lactat ion caused an increase in 

mastitis when compared with four minute s (Dodd et at, 1 9 5 0 )  , whereas 

milking for 6 . 5  minute s for a full lactation had no effect when 

c ompared with cup s  be ing removed when milkflow stopped (O ' Shea , 1 974 ) . 

Natzke et at ( 1 978 ) reported more new infections in the group o f  cows 

overmi lked for an average of 6 . 5  minute s compared with the c ontrol 

group . 

Another area o f  r i sk i s  the abi l i ty o f  the mi lking machine to 

transmit bac teria between cows ' quarter s .  Thi s  may occur through 

reverse flow o f  mi lk brought about by cup - s l ip , vacuum in stab i l i ty 

or flooding cup s  w ith milk . 

In Irish work (O ' Callaghan et at, 1 9 7 6 )  teat cup slip has been 

impl icated as a ma j or factor influenc ing the incidence of mastitis in 

a herd of cows . Th is i s  in support o f  a short experiment c onducted 

in U . S . A .  (Thomp son and Schultze, 1 97 5 ) . Austral ian work ( G ibb and 

Me in , 197 6 )  found that the percentage o f  cows on which liner s l ip 

o ccurred var ied between 4%  to 59% for individual l iners . Although no 

individual f actors could apparently be re lated to the inc idence o f  

s l ip ,  i t  appeared that the l iner with t h e  barrel of wide st diame ter 

s l ipped lea s t  frequently and older l iners sl ipped less frequently than 

new l iners . Other work in U . S . A .  and Ireland showed that narrow bore 

l iners fell o f f  during milking more frequently than wider bore l iners 

and l iners with no mouthpiece cavity a l so fell off more frequently 

( Schmidt et al, 1 9 6 3 ; McGrath and O ' Shea , 1 9 7 2 ) . Unstable vacuum 

within the machine might be expected to l ead to an increase in cup 

s l ip or fall and there i s  some evidence which shows that thi s i s  

indeed true (O ' Callaghan et at, 1 9 76 ) . Another possible factor i s  

the carele s s  remova l o f  teat-cups which may cause a drop i n  vacuum or 

may have effects w ithin the milking uni t  equivalent to l iner s l ip .  

Vacuum shoul d  therefore be cut-of f  from the mi lking unit before any 

cups are remove d .  Mos t  o f  the research reports indicate that heavy 

milking units , narrow bore l iners with poorly de signed necks and 



1 1  

unstable vacuum in the machine are the factors l ikely t o  cause cup 

s l ip .  Many spec i fic re ferences and comments on the role o f  machine 

mi lking in re lation to udder health and teat damage are given in the 

book titled " Machine Milking" pub l i shed by National Institute for 

Research in Dairying , Reading (NIRD , 1 9 7 9 ) , edited by C . C .  Thiel and 

F . H .  Dodd . 

It would appear from thi s  brief discus sion that the mi lking 

machine factors predi sposing cows to mastitis are more comp lex than 

i s  at first apparent and thi s  explains why the literature appears to 

be con fusing and contradic tory , particularly the literature which 

deal s  with the importance of most physic a l  settings like type and 

des ign of l iner s . Neverthe le ss , there is enough evidence to show 

that many individua l mi lking machine factor s predispose cows to udder 

in fection . I n  general terms the e f fect o f  the mi lking machine factors 

on incidence s  o f  ma stitis was demon strated by Foot ( 19 3 4 )  and Ward 

( 1 94 3 , 194 4 )  c ited by O l iver ( 19 5 5 ) . Foot ' s  data showed a higher herd 

wastage due to mastitis in machine - than in hand-milked herds and , 

Ward showed a mastitis inc idence o f  1 2 . 5 % and 8 . 9% for cows in machine ­

and hand-milked herds , re spective ly . Re sults reported by P earson et al 

( 19 7 2 )  indicate c learly the importance o f  machine effic iency in 

re lation to udder health as  shown by the fact that 68% of the low 

mastitis inc idence herds had mi lking machines which were total ly 

e f f ic ient , in contrast to only 16%  in the high inc idence herds, although 

the herds with e f f icient machines were probably we l l-managed in a l l  

other way s a s  we l l , which could b e  import�nt . 

( i i )  Milking Routine Hygiene 

Bacteria are ubiquitous i . e .  widelyspread in the cow ' s  environ­

ment . Thus the level of hygiene practiced during every mi lking will 

inf luence the spread o f  bacteria from cow to cow . The pos s ible sources 

of microbial contamination and transmis s ion from cow to cow at mi lking 

t ime are : ( a )  the milking unit ( i . e .  teat-cup l iner s ,  c law and long 

milk tube s ) ,  ( b )  skin on the end of the teats , ( c )  udder washing 

c loths and ( d )  hands o f  the milker . 

Contaminated udder , teat-cup l iner s , c law and mi lk tubes and teat 

washing a l l  cows with a c ommon c loth from a bucket of diluted 



12 

d i s infectant solution , di s seminate most of the micro-organisms 

between mammary quarters o f  each cow and between cows ( Bacic et al, 

1 9 6 8 ; Johns , 1966 ; Newbould , 1 9 6 5 ) . 

S ince intramamrnary in fection results from microbial invas ion 

into the gland via the teat cana l ( Gie secke et al, 196 8 )  , it would 

appear that if no micro-organisms contact the teat end , intramamrnary 

infect ion would not result . McDonald ( 19 6 9 )  reports that when 

bacterial population on the skin on the teat end was reduced to a 

very low degree ,  intramamrnary infection was prevented . Other reports 

( e . g .  McDonald , 1970 ; Neave , 1 97 1 )  indic ate that certain milking­

time hygienic procedure s ( such as washing udders with running water 

and dry ing them with individual towe l s  be fore each milking , and teat 

dipping in disin fectant solution after mi lking c oupled with personal 

c leanl ine s s )  preventtransmi ss ion of micro-organi sms on the teat skin . 

However the use of sing le- service paper towe ls might probably prove 

to be economically not a practical propo sit ion in terms of expen ses 

invo lved . Wa shing udders with runn ing water from a hose would seem 

pre ferable to washing with udder c loths because c loths can be a 

potent source of cross- infection particularly when only one c loth i s  

used for a l l  cows ( Benson et al, 197 4 )  . A study b y  Hoare ( 197 4 )  has 

shown that there is a lower inc idence of mastitis in herds where 

udders are wa shed with running water than where buckets and c loths 

are used . The author reports that there is even less mastitis i f  

udder soap i s  used i n  con j unct ion with the running water . The 

inc idence of subc linical mastitis wa s 25% o f  quarters in herds washed 

with buckets and c loths , 21% with running water , and 12% with running 

water and udder soap . It has been suggested in other reports (Neave , 

1 97 1 )  that s ince the usual methods o f  washing udders are not very 

e f fective in washing the ends o f  teats , washing should be restricted 

to the teats and particular ly the tips o f  teats because washing the 

udder i s  like ly to spread micro-organ i sms resid ing on the sur face o f  

the udder ( th i s  i s  with particular re ference to StneptOQOQQ� ub� 

and c o l i forms ) . 

The mi lker's hands are also a possible source of micro-organi sms 

that can be easily spread to the teats if not frequently disinfected 



1 3  

during the milking proc e s s . Bacteria capabl e  o f  c ausing intramammary 

infection have been i so lated from the milker's hands (Newbould ,  1 96 5 ) . 

An average of  30% of hands yie lded pos itive swabs after being washed 

with d i sinfectant compared to 9 5 %  when no d i s infectant was used 

( Report Natn . Inst . Dairy . U . K .  1 962) . 

S ince mastitis i s  sometime s regarded a s  an environmental disease 

( F i sher , 1 98 1 ) , c leanline s s  o f  the dairy premi ses and sanitation of 

milking equipment are some of the factors that would greatly he lp in 

contro l l ing ma stitis ( Dunsmore , 198 1 ) . 

2 : 1 : 6 MASTI TIS CONTROL : In deal ing with c ontro l measure s ,  farmers 

must realize that mastitis not only causes f inanc ial losses due to 

( a )  reduced mi lk secretion , ( b )  discarded milk , ( c )  increa sed replace ­

ment costs , ( d )  drug costs , ( e )  veterinary fee s , and ( f ) high cull ing 

rate ; but it is a l so an increas ingly important mi lk-qua lity problem 

(Brookbanks et al, 1 9 7 1 ; Ali et al, 1980 ) . 

Most of the studie s reported in the l i terature indic ate that an 

e ffective mastitis contro l program inc lude s the following five maj or 

points , namely : -

1 .  Mi lking machine e f ficiency check . 

2 .  U se o f  a teat sanitizer a fter mi lking . 

3 .  Dry -cow therapy . 

4 .  Prompt treatment of  c l inical cases . 

5 .  Cull ing of c hronically infected cows . 

The importance o f  the mi lking machine in mastiti s control can 

not be over emphas i zed . The da iryman should select a milking machine 

system designed to ensure vacuum stability . F luc tuating vacuum ha s 

been found to be the common fault of  the mi lking system in the causa­

tion of mastitis ( Hopkirk et al, 194 3 ;  Nyham & Cowhig , 1967 ) . Thus 

periodic evaluations of a l l  parts of  the mi lking system are needed to 

maintain it in a proper ly functioning condition . 

D i s infection o f  milking teat-cups between milking of  cows can 

not be accomplished in the time avai l able or with systems currently 



14 

in use, particularly in large herds . However, disinfection of teats 

a fter milking is regarded as  an e f fective means for reduc ing the 

spread of infectious mastitis (Neave et al, 196 9 ;  Neave, 197 1 ; 

Sheldrake and Hoare, 1980) . 

Control o f  ma stitis in problem herd s  by the intramammary infus ion 

of antibiotics into all quarter s  of all cows as they enter the dry 

period has been advocated as the mo st practical and least costly means 

for reducing the inc idence of ma stitis ( Roberts et al, 1969)  . To be 

much more e ffective, however, such a program must be associated with 

hygiene during milking, and use of good milk ing equipment . It i s  

suggested i n  some literature (e . g .  Schalm et al ,  1 9 7 1 )  that it wou ld 

perhaps be better to select cows for dry therapy on the basis o f  

either the ir reaction t o  California Mastitis Test (CMT ) applied 

monthly throughout lactation or an average o f  somatic c e l l  count 

leve l s  regularly monitored during lactat ion . Such an approach would 

avoid infusing antibiotic s into the uninfec ted c ows and thus reduce 

expenses incurred in the control program as we l l  as reduc ing the 

probabil ity that bacteria wi ll become re s i s tant to the antibiotic . 

As suming that c ows are be ing tested routine ly for mastiti s so 

that subc linical and chronic problems are identi fied, rigorous cul ling 

programs can e l iminate chronic mastitis cows from the herd . 

Me l lenberger ( 1982)  suggests that o lder cows that are unre sponsive to 

two or three ser i e s  of antibiotic treatment for mastitis during 

lactation should rece ive serious cons ideration for cul ling . A report 

from Dorset, U . K .  (Farmers Weekly, November 27 , 1981)  shows that high 

cull ing rate s coupled with strict hygiene have brought mastiti s leve l s  

down from 526 , 000 t o  28 3 , 00 0  cell s/ml . 

There is  experimental evidence to show that the net gain in 

productivity in the first year o f  the c ontro l program is  small but it 

increases in subsequent years if the farmer perseveres (Asby et al, 

19 7 5 ) . 



2 : 2 : 0  PART I I  - SUBCL I N I CAL MASTITIS 

2 : 2 : 1 INTRODUCTI ON 

1 5  

Subc l inical ma stitis is a wide spread prob lem in dairy cows and 

the fact that it is virtua l ly without readi ly detectable symptoms 

means that the farmer s are unaware of its pre sence . It has been 

e stimated that subc linical ma stit i s  o ccur s in 2 8 %  of quarters and in 

5 5% of cows ( Dodd and Jackson , 1 97 1 )  and chemica l  analys i s  of milk 

indicate s a s imil ar proportion ( Linze l l  and Peaker , 197 2 ) . New 

Zea land results ( Mol ler , 197 8 )  showed that in an average herd with­

out ma stit i s  control , 1 3 %  o f  quarter s  get treated for c l inical 

mastitis and 2 6 %  of the quarter s have sub c l inical ma stitis . A report 

from Sri Lanka ( Rupas inghe and Rul a segaram , 1982 ) indicate s that the 

inc idence o f  subc l inical mastitis range s from 3 9 . 5  to 92 . 2 % of cows 

in Government farms with large un its o f  European breed cows , and 

from 27 . 8  to 9 0 %  of cows in private farms with smaller units of cross­

bred cows . The average preva lenc e rate o f  subclinical mastitis in 

Norwegian dairy c ows ( Bakken , 1 98 2 )  i s  reported to be 3 1 %  o f  the cows 

and 1 1 . 6% o f  quarters .  Stiles and Rodenburg ( 1 981 ) have reported 

that 97% of mastitis infections in Onta r io , Canada , are subc l inical 

invo lving no visib le change s .  Also a recent report compiled by Ho lme s 

( 19 8 1 )  on behal f  o f  the Livestock Improvement As soc iation Movement in 

New Zealand indicates that about 2 0  to 4 0 %  o f  all  cows in seasonal 

supply herds are affected by sub c l in i cal mastitis ; and a survey of 

mastitis in the Brit i sh herd (Wil son & Richards , 1 980 ) revealed that 

the nationa l prevalence of subc linical ma stitis as de f ined by the 

International Da iry Federation wa s 9 . 6% o f  all  quarters . These few 

c ited reports indicate that the condit ion is numerica l ly important . 

2 : 2 : 2  ECONOMI C  I MPORTANCE OF SUBCL I N I CAL MAST I T IS 

Subc l inical mastitis as  reflected in an e levated somatic c e l l  

count , causes decreases i n  the milk y i e l d  o f  cows . Estimated loss 

reported from Canada ( S tiles & Rodenburg, 1 9 8 1 )  ranged from more than 

1 5 %  in herds w ith c e l l  counts above 800 , 000 c e l l s/ml to l e s s  than 

2 . 5% in herds be low 300 , 000 c e l l s /ml . A recent report from Wiscons in, 

U . S . A .  ( Hoard ' s  Dairyman, March 2 5 , 1 98 2 )  shows that o f  the in fected 



1 6  

quarters , 7 5 %  do not show symptoms other than high cell counts and 

that for every quarter which shows c l inical sign s  such as f lake s in 

mi lk or actua l f lare-ups , there are at least 2 0  or 2 5  other infected 

quarters showing no visible symptoms at all . S ince subc linical 

mastitis has no visible symptoms , cows may be cul led for low produc tion 

without the farmer suspecting the pre sence of infection . The forced 

cul l ing o f  these anima l s  substantial ly reduce s  farmer s '  opportun ities 

to cull for low produc tion ba sed on gene tic potent ial ( Hoare , 198 2 ) . 

The economic importance o f  subc linical mast itis can be asse s sed from 

its e f fect on yield and composition of milk and the qual ity of products 

made from mi lk with high somatic c e l l  counts . Stiles and Rodenburg 

( 1 9 8 1 ) have reported that subc l inical ma stit i s  results in a decrea se 

in production repre senting an average los s  o f  more than $ 10 0  per cow 

per year in Ontario , Canada . 

( i )  Effect on milk y ie ld - Individual quarters which have a high 

somatic cell count ( S . C . C . )  produce less milk than other quarters on 

the same udder but with a low S . C . C .  ( Holme s , 1 98 1 ) ; and simi larly 

individual cows which have a high S . C . C .  produce less milk than other 

cows in the same herd but with a low S . C . C .  ( see Table 2 . 2  and Figure 

2 . 1 ) . Since a high S . C . C .  is  indicative of subc l inical mastitis , it 

can be conc luded that subc l inical mastitis causes a decrease in milk 

yield . 

)f.- J< - ;c  Potass ium - X - ;r 

--"' -
- - - X - � 

· -- " 
. ......... 

-- �  
. ,  -

3 

H (!) +l H m 
(!) (/) 0 

· -:-.... - ::J +l 

.-i 
u 

m z 

100 -

so -

0 

... - -
Chloride 

__ ...... . � - · -
Sodium 

20 s o  100 

- ·  

- .  
"'­

,. , 
· ,(,  �0 ... · '-o 

'-':.$'� • 

. .  ....- · 
-·>< . 

_ ..-t9"· . - .  IQ.J." • • • - - . - -<..r 

- 2 

1 

200 500 1000 2 00 0  5000 
Somatic cell counts 
(thousand c e l l s/ml ) 

a ......... 
� - 0 0 0  M ·.-I ::8 
tJ"I :.:: 

- 500 

4000 

F I GURE 2.1: RELATI ONS BETWEEN SOMAT I C  CELL COUNT , MILK COMPOSI TI ON AND 
M ILK PRODUCT I ON FROM I ND I VI DUAL QUARTERS C HEESCHEN , 1 975 ) , 

u 
Ill ...:l 
cJI' tJ"I = 



17 

As regards the corre lation between milk quantity and cell 

content , it  can be seen from Figure 2 . 1  that the milk yield begins 

to decrease at a relative ly low S . C . C .  in agreement with the bio­

chemical change s .  This inherent relationship between severity of 

disease and milk yield al lows a statistical estimate to be made of 

economic losses due to mastitis per quarter , animal and herd with 

suffic ient accuracy . An increased absolute number of somatic c e l l s  

in milk i s  a sens itive indicator o f  change s in chemical composition 

o f  milk or lower milk yie ld ( Hee schen , 197 5 ) . 

TABLE 2 . 2 :  THE RELAT ION BETWEEN S . C . C .  OR INFECT ION ,  AND M I LK 
PRODUCT ION FROM I ND I V I DUAL QUARTERS OF AN UDDER OR 
FROM I ND I V I DUAL COWS . 

( i )  IND I V I DUAL QuARTERS < FORSTER , 1 964 ; U .  S . A . ) 

California Ma stitis Test Re sult Decrease in yield per quarter 

Trace o f  sl ime ( approx . 300 , 00 0  - 0 . 3  kg per quarter per day 
cells/ml ) 

Moderate s l ime ( approx . 900 , 0 00 - 1 . 0 kg 1 1  1 1  1 1  1 1  

cell s/ml ) 

Thick sl ime ( approx . 2 , 000 , 000 - 1 . 8  kg 1 1  1 1  1 1  1 1  

cells/ml ) 

Very thick sl ime - 2 . 7 kg 1 1  1 1  1 1  " 

( ii)  IND I V I DUAL QUARTERS ( MoRR I S ,  1 97 3 ;  AUSTRAL I A )  

Decrease in yield by an infected quarter 

Cows in 1 s t  Lactation 

Cows in 2nd Lactation 

Cows in 3 rd Lactation 

38% 

57 % 

67% 

Note : The S . C . C .  from the se quarters was increased . 

( iii)  IND I VI DUAL COWS < DAN I EL & F I ELDEN , 1 97 1 ; N . Z . ) 

Somatic C e l l  Count Leve l 

2 1 4 , 000 cells/ml 

647 , 000 cells/ml 

1, 4 80 , 000 c e l l s/ml 

Percentage decrease in y ield 

5% 

- 11% 

- 15% 



1 8  

( iv)  IND I VI DUAL COWS ( G I LL & HDLMES , 1 978 ; MACMILLAN 
& OUI RS , 1 98 0 ; N . Z . ) 

Somatic Ce l l  Count 

0 to 2 4 9 , 000 cel l s /ml 

2 5 0  to 4 4 9 , 000 c e l l s /ml 

500 to 749 , 000 c e l l s /ml 

Greater than 750 , 000 c e l l s /ml 

(v) I ND I V I DUAL 

Le s s  than 300 , 000 c e l l s/ml 

3 0 0 , 000 - 500 , 000 c e l l s /ml 

500 , 000 - 800 , 000 c e l l s /ml 

Greater than 800 , 000 c e l l s /ml 

Cows 

% 

1978 1980 

0 0 

- 2 %  - 3% 

- 5 %  - 5 %  

- 7 %  - 1 0 %  

( ST I LES & RODENBURG , 1 98 1 ; CANADA ) 

Estimated Milk Production Los s  

0 2 . 5  

2 . 5  - 7 . 5  

7 . 5  - 1 5 . 0  

1 5 . 0  - 2 5 . 0  

(vi ) AVERAGE M I LK Y I ELD FROM I NFECTED ( 1 )  AND OPPOS I TE 
NORMAL QUARTERS - ( N )  < K ING, 1 978 ) 

Ce ll  count per ml No of Yie ld ( kg )  Decrease in yield ( kg)  i . e .  

for Infected Quarters cows l N 1 - N 

200 , 00 0  to 500 , 00 0  8 3  2 . 04 2 . 1 1 - 0 . 07 

500 , 000 to 1 , 000 , 000 8 5  1 . 8 3 2 . 09 - 0 . 2 6 

lj)OO , 000 to 2 , 000 , 00 0  8 1  1 .  7 4  2 . 1 3 - 0 . 39 

Note : All the ce l l  counts from the oppos ite healthy ( norma l )  
quarters were below 200 , 000 per ml . 

The results in Table 2 . 2  show in general that the higher the c e l l  

count the greater the reduction in yield . Rako et al ( 1 963 ) in a 

study of  1003 cows found that the yie ld of affec ted quarters was de ­

pre ssed ( in compari son with opposite healthy quarters ) by 1 2 . 2 % for 

quarter s  which were bacteriologically positive but showed no c l inical 

s igns i . e .  subclinical case s . Hoare ( 1 9 8 2 )  cites the work reported 

by Re ichmuth et al ( 19 7 0 )  in which quarters with over 500 , 000 c e l l s /ml 

of  foremi lk y ie lded 17 . 1% less than the oppos ite healthy quarters . 

Also data from 9 3 3  animal s ,  mainly heifers , examined by quarter and 

cow comparison , demonstrated sub stantial loss  due to subc l inical 

mastitis (Me i j ering et al, 1 9 7 8 )  . The same authors c i te the work of 



1 9  

d e  Rooy & Jaartsve ld ( 1969 ) who asse ssed the mean loss over the whole 

lactation for cows with subc l inical mastitis at 7 - 9% ( 1 . 2  to 1 . 5  kg/ 

cow/day ) . Using fixed model s  to examine the partia l influence of c e l l  

count on quarter-mi lk yield , Reichmuth ( 1968)  c ited by Me i j ering et al 
( 1 9 7 8 )  found that the drop in yield became signi f ic ant when the cell 

count exceeded 500 , 0 00 c e l l s /ml . He reported that cows with a cell 

count in exc e s s  of 500 , 000 cells/ml in at least one quarter produced 

9% less milk compared with healthy anima l s . Waite & Blackburn ( 1 957 ) 

and McKenzie et al ( 19 5 8 )  noted a drop in the daily milk yield per 

cow when the total ce l l  count increased to an average of 500 , 000 ce l l s /  

ml . Janzen ( 1 970)  reports that infected quarters whether showing 

ma stitis symptoms or not , usually yield less milk of lower total solids 

than the corre sponding uninfected front or hind quarters of  the same 

udder . 

All the se reports confirm the e ffec t of  subc linical ma stitis on 

milk yield which , as  it c an be deduced from Table 2 . 2 ,  appear s to be 

proportional to the severity of the changes occurring in the udder as  

j udged by cell count . 

( i i )  E f fect on milk composition - The composi tion of mi lk from 

individual quarters is known to change in a s soc iat ion with increase s 

in somatic cell count ( see Figure 2 . 1 ) . The fol lowing change s have 

a l so been reported to occur in herd bulk mi lk as shown in Table 2 . 3 .  

TABLE 2 . 3 :  CHANGE I N  % COMPOSI T I ON ASSOCI ATED W I TH AN I NCREASE OF 
1 0 0 , 000 CELLS/ML IN  BULK SoMATI C  CELL COUNT . 

Milk Constituent 

Fat % 

Prote in % 

Solids-not-fat % 

Percentage loss 

- 0 . 010%  

- 0 . 00 1 %  

- 0 . 019%  

- 0 . 0 1 5 %  

Reference 

( Asby et al ,  19 7 7 )  

( Gi l l , 19 7 7 )  

( Gi l l  & Holme s , 197 8 )  

( Asby et al ,  19 7 7 )  

The se are all re latively sma l l  change s in milk compos i tion , 

neverthel e s s  they repre sent adver se e ffects of  high somatic c e l l  

counts ( i . e .  subclinical ma stiti s ) . 



20  

Udder infections change milk composition by altering membrane 

permeabi lity within the mammary gland and by reduc ing synthes i s  of 

milk components ( Schalm,  1977 ) . There are two general phy sio logical 

explanations for changes in milk compo sition assoc iated with mastitis 

and e levated somatic c e l l s , name ly : -

a )  inj ury t o  the udder ce l l s  which reduce s  the synthe s i s  of 

those milk components synthes i zed in the udder ; 

typical examples are lactose and most of the casein , 

and 

b )  change s in permeabil ity of  membrane which permit increa sed 

" leakage " of materia l s  from the b lood to the milk . 

There appears to be increased permeability of both the 

vascular and secretory epithe lia . Typical examples of 

materials which increase due to this phenomenon are 

sodium , chloride and immunoglobulins ( see Figure 2 . 1 ) . 

When the concentration of lactose is decreased , compen­

sation must be made to en sure that mi lk and b lood 

maintain the same osmotic pres sure . Mo st of this 

compensation i s  accomplished by increases in sodium and 

chlor ide ( Schalm , 1977 ) . 

The c hange s in composition are directly proportional to the 

severity of infection as measured by cel l count or CMT (Waite & 

Blackburn , 196 3 ;  Foster et al, 1966 ; McKenzie and Ander son , 1 9 8 1 ) . 

King ( 1 978 )  observed significant changes in composition when the 

infected quarters showed c e l l  counts of 500 , 000 cells/ml or above . 

On a between herd ba s i s  Asby et al ( 1977 )  demon strated a relationship 

between bulk milk compos ition and c e l l  count ; total solids , solids­

not-fat and fat concentration tending to decrease as  ce l l  count 

increased . 

( ii i )  E f fect on yield and qual ity of  dairy product s  - High 

somatic c e l l  leve l s  are a l so assoc iated with the development of 

quality defects in dairy products , particularly on storage ( Brus and 

Jaartsve ld , 197 1 )  and present problems in the manufacture of cheese 

by depre s s ing acid production and reducing the firmnes s  of the curd 



21 

( Hampton & Randolph , 1 9 6 9 ) . The most recent information come s from 

a study which showed that milk with high S . C . C .  had a slower c lotting 

reaction with rennet and gave a lower curd yield per 100 kg milk (Ali 

et ai, 1 98 0 ) . Work in Wi sconsin , U . S . A .  showed that decreases in 

cheese yield per lOO kg milk occurred in assoc iation with increases 

in bulk somatic c e l l  count ( B . S . C . C . ) ; see Table 2 . 4 .  A high somatic 

c e l l  count may affect the growth rate of bac teria added a s  starter 

cultures to milk . In his review Guthy ( 1 9 7 9 )  cite s the work reported 

by Kastli et ai ( 1 966 ) in which the growth of S�epto �o ��� la� 

wa s retarded when the somatic c e l l  count was between 200 , 00 0  and 

5 00 , 00 0  cel l s /ml and the retardation wa s even more pronounced when 

the c e l l  count was above 500 , 000 ce l l s /ml . He a l so reports that 

butter made from milk of high somatic ce l l  count has a le ss pronounced 

aroma and an impaired taste both of  which become marked on storage . 

TABLE 2. 4 :  THE RELATION BETWEEN B . S . C . C .  AND CHEDDAR CHEESE Y I ELD 
( HOARD ' S DA I RYMAN , MARCH 25 , 1 980 ) . 

B . S . C . C .  
( c e l l s /ml ) 

240 , 000 

496 , 000 

640 , 000 

Cheddar chee se yield 
( kg chee se/ lOO kg milk)  

9 . 7 5 

9 . 69 

9 . 43 

The ce l l  count in milk is  used in some countr ies as  one of  the 

cr iteria of the hygienic qua lity of the mi lk ( Di j kman et ai ,  1 9 6 9 )  . 

Mastitic mi lk has a high c e l l  count and thi s  a l so leads to lowering 

o f  qua lity ( Schalm et ai, 1 9 7 1 )  . The Standard P late Count has now 

become the basic bacterial te st for raw mi lk in the New Zealand Dairy 

Industry and thi s  may be affected by the growth of  mastitis-causing 

organi sms which may lead to down-grading o f  the milk thus lower ing 

income if rate of  payment is based on grade . It may therefore be 

necessary to identify infected cows , both c l inical and subc l inical , 

in order to prevent milk from be ing down-graded . 

Milk quality i s  of  national importance because most o f  New 

Zealand ' s  mi lk , for example , is proc e s sed into products which must 



22 

meet rigid standards before they can be sold and exported . Improved 

udder health should lead to the production o f  good quality milk and 

dairy products . 

2 : 2 : 3 : 0  D I AGNOSI S  

Subc linical mastitis usual ly has no apparent sign s and can be 

detected only by spec ial te sts . Measur ing the number of somatic c e l l s  

i n  milk is  one of  such tests . · Somatic o r  "body " c e l l s  found i n  mi lk 

inc lude sma l l  numbers of c e l l s  released as  part of  the normal proce s s  

o f  replac ing milk produc ing tissue . The epithe lial ce l l s  o f  the 

mammary gland are sub j e cted to continuous wear dur ing the ir spec ific 

activitie s of the synthe s i s  of  milk and its constituents . The worn 

c e l l s  and the c e l l  debris are discharged with the milk ( Hee schen , 

1 9 7 5 )  and thus are a regular and norma l component of  it . The remainder 

of the somatic ce l l s  in the milk are leucocytes or white blood c e l l s  

whose role i n  the body is to : ( a )  de stroy bacteria , (b ) prevent or 

e l iminate infec tion and ( c )  repa ir damaged t i ssue . 

Increased somatic ce l l  counts result when the udder is  inj ured , 

severely str e s sed or when it become s infected with mastiti s ( S chalm & 

Lasmanis , 1 9 6 8 ) . The se authors report that the content of ce l l s  

increases in mastitic mi lk primarily due to an overwhe lming number o f  

leucocyte s infiltrating from the b lood and so , by counting the c e l l s  

a n  assessment o f  the probabil ity o f  bacteria l infection can be made 

( MacMi llan & Duirs , 1980 ; Schalm & La sman i s , 1 9 6 8 ) . 

There are two commonly accepted diagnostic signs of subc l inical 

mastiti s :  infection of  the foremi lk and a raised milk c e l l  count 

( Linze l , 1 9 7 5 ) . The use of a somatic c e l l  c ounting system for monitor­

ing mastitis s tatus of herds and the udder health of  individual cows 

is being wide ly accepted in many countries throughout the wor l d . Cell  

counts provide a measure of  inflammatory reaction in the udder and 

taken together with the isolation of bacteri a , they are a guide to the 

prevalence of subc l inical mastitis where there are no gross c hange s in 

the milk or c l inical change s in the udder . The International Dairy 

Federation ( se e  Section 2 : 1 : 1 ) state s that where the milk and udder 

are macroscopically normal a c e l l  count of more than 500 , 000 c e l ls /ml 



2 3  

together with the presence of  pathogenic bacteria signifies sub­

c l inical mastitis . I solation of pathogenic bacteria in the milk 

sample as we l l  as  high somatic cell leve l s  would therefore seem to 

be an important criterion in arriving at a c orrect diagnos i s  of 

subc linical ma stitis due to the fac t that exudation o f  leucocytes 

into the milk could a lso be caused by any mechanical inj ury of  the 

mammary tis sue as we l l  as stress (Afi f i , 1968 ; Schalm & Lasmanis , 

1968 ; Whittlestone et al ,  1969 ; Pear son et al ,  1971) . It  i s  

reported in l i terature ( S cha1m & Lasmanis , 1 9 6 8 )  that leucocyte 

exudation into the milk may cont inue for long periods after the 

c ause of mastitis has been removed .  The re lationship between the 

pre sence of pathogen s and cell count is shown in Tables 2 . 5 ,  2 . 6 ,  

2 . 7  and 2 . 8 .  

TABLE 2 . 5 :  THE RELATI ONSHI P  BETWEEN THE I NC I DENCE OF MAST I T I S  
PATHOGENS AND CELL COUNT I N  I ND I VIDUAL COWS ( MACM I LLAN 
& DUI RS , 1 98 0 ;  N . Z . ) .  

Infec tion Status 

Ce l l  Count Primary Secondary 
( thousands/ml ) Pathogen ( 1 )  Pathogen ( 2 )  Non-infected 

200 or less 18% 7 4 %  86% 

2 0 1  - 500 30%  2 1 %  12% 

501 - 1 , 000 20% 5% 1% 

More than 1 , 000 3 2 %  0 %  1% 

100% 100% 100% 

No . of cows 79  9 9  1 5 2  

( % )  o f  all  cows ) ( 2 4 )  ( 30 )  ( 4 6 )  

( 1 )  S�aphyloQOQQ� �eu6 , Streptococcus sp . 

( 2 )  Co�yneba��m b o v� , Micrococcus sp . 

All  cows 

6 6 %  

1 9 %  

7%  

8 %  

100% 

300 

I t  can be seen from the above table that primary pathogens known 

to cause mastitis were isolated from only 7 9  ( 2 4 % )  of the 3 3 0  samples ; 

and that 8 2 %  o f  the se 7 9  infected cow samples a l so contained over 

200 , 000 c e l l s/ml . In c ontrast , there were mi lk samples from 1 52 c ows 



2 4  

( 46 % )  from which no bacteria were i solated ; 8 6 %  of the se samples 

c ontained less than 2 0 0 , 000 c e l l s/ml . The secondary pathogen s are 

bacteria pre sent in the milk but do not usua l ly cause mastitis . 

Re sults in Table 2 . 5  show that they do not greatly elevate c e l l  counts . 

Some o f  the samp l e s  from which primary pathogens were isolated a s  

shown in Tab le 2 . 5  a l so had low c e l l  counts ( 1 8 % ) . It is reported 

that the se were probably cows with a less severe infection in only 

one quarter , and that higher cell count from that quarter was di luted 

by mi lk from the other three quarters . The author s postulated that 

s ince cell count leve l s  are constantly changing , these samples may 

have been obtained from cows with an early infec tion , or the non ­

infec ted samples which contained high c e l l  c ounts may have been 

obtained just when the leucocytes had overcome the infection ( Schalm 

& Lasmanis , 1968) . I t  i s  for such reason s that more than one ce l l  

count for each individual cow can prove more meaningful than a result 

from a single sample ( Griffin , 197 1 ;  Natzke e:t al ,  1972 ; Neave , 

1 97 5 ;  MacMil lan & Duirs , 1980) . Cows with persi stently high ce l l  

counts are most l ike ly t o  be tho se with infec ted udders ,  and be the 

cause of herd ' s  high bulk milk cell  count ( Ho lme s , 1 9 8 1 ) . 

TABLE 2 . 6 :  MEAN GEOMETR I C  S . C . C .  FOR QUARTERS OF KNOWN I NFECTI ON 
STATUS ( CELLS/ML ) X 1 03 ( SHELDRAKE , 1 982 ; AUSTRAL I A ) .  

Herd V Herd D Herd W 

S:taph . 
a.wz.et.l.-6 

Minor 
pathogens 

Free S:taph . 
a.uJl.et.l.-6 

Minor 
pathogen s  

Free S:taph . 
a.wz.et.l.-6 

Minor 
pathogens 

Free 

860 

b 

2 50 

c 

1 16 

f 

2 5 2 5  

a 

303 

c 

1 4 0  705 

e , f  b 

1 9 3  

e 

Number 

1 2 2  1 7 19 600 1 2 2  2 2 4  1 3  1 3 8  509 

Note : Different letters indicate s igni ficant difference us ing 
student ' s  t - te st , P <0 . 00 1 . 

The above table shows different level s  of  s . c . c .  in quarters 

affected by S:taphyiococct.l.-6 a.uJl. et.l.-6  or minor pathogens infec tions in 

different herds , indicating the problem encountered in c oming up with 

76 

f 

16 



2 5  

a standard thre shold of  s . c . c .  at which bacterial infection of  economic 

importance can be expected with certainty . She ldrake ( 1 982 ) sugge sts 

that i f  a common thre shold i s  to be selected for herds l ike those 

pre sented in Table 2 . 6 ,  then the lower leve l of  S . C . C .  should be 

chosen . This would result in a high proportion of  false positive 

predic tions in Herd D ,  which neverthe less would be preferable to a 

high proportion of false negative predictions in the remaining two 

herds ( She ldrake , 1982 ) . 

TABLE 2.7: THE RELATI ONSHI P  BETWEEN SOMAT I C  CELL COUNT AND I NFECTION 
C HDARD ' S  DA I RYMAN , MARCH 25 , 1982 ; VIRG I N I A  POLYTECHNI C  
INST I TUTE & STATE UN I VERS I TY ,  U . S . A .  ) .  

Somatic cell  count 
range in foremilk Percentage of  cows Percent Pos itive * 

< 100 , 000 37 3 7 . 8  

100 , 000 - 200 , 000 20 6 1 . 0  

200 , 000 - 300 , 000 1 2  66 . 6  

300 , 00 0  - 400 , 000 8 69 . 0  

400 , 000 - 500 , 000 6 7 3 . 8  

500 , 000 - 700 , 000 7 7 2 . 9  

700 , 000 - 1, 000 , 000 6 7 5 . 8  

[ *Bacteria isolated from foremilk in the data given in Table 2 . 7  
inc luded the streptococci sp . ( i . e .  S�epto co cc� agalactiae ,  
Stnepto co cc� dy�galactiae, S�eptococc� ub � )  , Staphylo cocc� 
a�e� , E� chenichia coli; Staphyio cocc� ep�d�� and 
C o�ynebactenium bov� ] 

The results pre sented in Table 2 . 7  show that a s  the somatic c e l l  

counts ro se , the percentage of  c ows infected rose . Similar results 

( see Table 2 . 8 ) have been reported from a study in Pennsylvania ( Stiles 

& Rodenburg , 198 1 ) . 

The percentage of  in fection shown in Table 2 . 7  seems to be very 

high p articular ly under the < 100 , 000 c e l l s/ml range . Although nothing 

about productivity appears in the table , one would expec t  a reduction 

in milk yield particularly i f  a greater proportion of the pathogens 

involved are the primary one s i . e .  Staph . �� and the Streptococ c i  



2 6  

group . That being the c ase then it could b e  deduced that a mild 

e levation in somatic c e l l  concentration is l ike ly to have an adverse 

effect on mi lk produc tion . 

TABLE 2. 8: INFECT ION RATE OF COWS W I TH I N  VAR IOUS SOMAT I C  CELL 
COUNT RANGES IN A PENNSYLVAN I A  STUDY ( STILES & 
RODENBURG , 1 98 1 ) .  

Somatic c e l l  count 

< 100 , 000 

100 , 000 - 200 , 000 

200 , 000 - 300 , 000 

300 , 000 - 400 , 000 

400 , 000 - 500 , 0 00 

500 , 000 - 600 , 000 

>600 , 000 

% of cows infected with Ma j o r  Pathogen 

6 

1 7  

34  

4 5  

5 1  

6 7  

7 9  

The percentage o f  cows in which maj or pathogen was isolated in 

the data presented in Table 2 . 8  seems to sugge st a subclinical 

mastitis problem at a rather low leve l of  somatic cell count . The 

problem is much more obvious at a leve l  > 50 0 , 00 0  cells/ml where 6 7 %  

of the cows were found t o  b e  infected with a ma jor pathogen . 

Duirs ( 1980 ) gave some recommendations about the leve l of  

somatic c e l l  count to be adopted as a basis for diagno sis o f  

subc l inical mastitis ( see Table 2 . 9 ) . 



2 7  

TABLE 2 . 9 :  CELL COUNT LEVELS RELATED TO SUBCL I N I CAL MAST I T I S  
C DUIRS,  1 9 8 0 ; N . Z . ) .  

Ce l l  c ount 
( c e l l s /ml ) 

Bulk-ce l l  count 
from vat 

Individual cow 
c e l l  c ounts 

< 2 5 0 , 000 Indicate s a low leve l of 
subc l inical ma stitis 
( sati sfactory ) . 

Indicates a healthy 
udder . 

2 50 , 00 0  - 500 , 000 Indicate s that the herd 
has a moderate subc l inical 
mastitis problem . Ac tion 
required to contain 
problem . 

Individual c ows with 
cell counts consist­
ently in thi s  category 
have a high chance of  
being infected . 

> 500 , 000 Indicate s a serious sub­
c linical mastiti s problem . 

Indicates that indivi ­
dual cows are infected 
with mastitis . 

In his  review , Holme s ( 1 9 8 1 )  conc luded that : -

a l l  cows with S . C . C .  which are c onsistently higher 

than about 200 , 000 to 300 , 000/ml , but without 

c l inical signs , have subc linic a l  ma stiti s .  

the maj ority of  cows which have bacterial infec tion 

will have high S . C . C .  

the maj ority of cows which have low S . C . C .  w i l l  be 

uninfected . 

The se conc lusions are in agreement with the data presented in 

Tables 2 . 7  and 2 . 8 .  In view o f  thi s  and the fact that mastitis 

accounts for 80% o f  the variation in somatic cell count ( Duir s , 1 9 8 0 ) , 

one woul d  be tempted to conc lude that somatic c e l l  counts in milk 

appear to be more important in subc l in ical ma stitis detec tion than 

probably bacteriological identif ication . 

However , the interpretation of  somatic c e l l  count results needs 

careful thought and has to take into c onsideration the other factors 

l ikely to cause an increase in c e l l  counts . These factors are as  

indicated in the following discuss ion . 



2 : 2 : 3 : 1 FACTORS INFLUENC I NG SOMAT I C  CELL COUNTS 

( i )  Stage o f  Lactation 

2 8  

It i s  genera l ly agreed that the somatic c e l l  count i s  h igh in 

c o lostrum but soon decreases and remain s low for the first few weeks 

a f ter which it gradual ly increases unt i l  the end of lactation 

(Cul len , 1966 ; Schalm et al ,  197 1 ) . The average cell c ount accord­

ing to Waite & B lackburn ( 19 5 7) was lowe st and least variable from 

day 70 to day 1 3 0  of lactation . The reported increase in c e l l  

counts i n  the beginning of lactation i s  due t o  the presence of  large 

amounts of epithe lial c e l l s  and polymorphonuc lear ( PMN )  leucocytes 

(Cullen , 1966 ) ; see Table 2 . 10 .  

TABLE 2 . 1 0 :  CELL COUNTS I N  NORMAL M I LK ( MI LL I ONS/ML ) C CULLEN , 1 966 ) . 

Epithe l i al No . of 
S tage of  Lac tation cel l s  Polymorphs Lymphocyte s Samples 

Colostrum (0 . 5  day ) 1 . 1 9 2 . 2 8 0 .  7 2  4 4  

Early ( 5  days - 8 weeks)  0 . 14 0 . 14 0 . 09 4 0  

Mid- lactation 0 . 1 2 0 . 10 0 . 0 5  84 

Late - last 4 weeks 

Be fore drying 0 . 8 5  0 . 2 8 0 . 5 9 36 

Blackburn ( 1966 ) reported that the increase in cell count at the 

end of lactation i s  probably due to the shedding of epithelial c e l l s  

from the a lveo l i  into milk during the c ourse of  involution o f  the udder 

which seems to be in agreement with the data pre sented in Tab le 2 . 10 .  

This together wi th the deceasing volume of milk towards the end of 

l actation results in the exaggeration o f  cel l concentration in the 

sma l ler volume o f  milk ( Schalm et al ,  1 9 7 1 )  . 

N.atzke et al ( 19 7 2 ) using the presence of  bacteria in two 

consecutive samp l e s  a s  indic ator of  infection showed that in late 

lactation uninfected quarters did not neces sarily show e levated c e l l  

counts i n  c omposite udder milk . Thi s  confirmed the suggestion of 

S chalm and Lasmanis ( 1968)  that the increase in cell counts a t  drying 



2 9  

o f f  was the result of  new infections . The reported trend in increase 

in cell counts with advanc ing lactation is probably due to an increase 

in new infection rate through l actation . Thus a sudden rise in c e l l  

numbers may sometime s b e  attributed to a new infection . 

( ii )  Age or Lactation Number 

It is indic ated in literature that the average c e l l  count 

increases with the lactation age of a cow ( B lackburn , 1966 , 1968 ; 

Cul len , 1966 ; Weihe , 1969 ) . Thi s could probably be a re flection of  

the past infection hi story of the udder as  Natzke et al ( 1 972 ) found 

that uninfected anima l s  showed no upward trend over five lactations . 

Ear lier , Blackburn ( 1 966)  over a seven year period found that the 

increase in ce l l s  from one lactation to the next was ma inly due to an 

increase in PMN leucocyte s . This increase was associated with bacterial 

infection resul ting in inflammation of the ducts and also an increase 

in the sever ity o f  lobular lesions . 

Tolle ( 19 7 1 )  reports that there is a decrease in c e l l  count 

average s in cows older than six years . She ldrake ( 1982 ) reports that 

if S . C . C .  is to be used to predict the presence of intramammary 

infection the ef fects of stage and number of lactation may be safely 

ignored . In his study the ef fects of stage and number of lactation 

upon S . C . C .  in quarters free from infec tion were found to be sma l l  

when compared to the e levation i n  s . c . c .  caused b y  infect ion with 

Staphytoeoee� au4e� . It would appear from this discus s ion that the 

increase in c e l l  count with age is sma l l  in uninfected cows but past 

infections play an important role . 

( ii i )  Diurnal Variation and Milking Interval 

Considerab l e  variations in c e l l  counts have been observed between 

morning and evening milk samples particular ly when there are irregular 

milking interva l s  ( Cu l len , 1 96 7 ) . These variations were found at both 

high and low somatic c e l l  counts ; although the greate st variation 

occurred when the counts were greater than one mil l ion . 

Schalm et al ( 19 7 1 )  are of the opinion that the diurnal variation 

in c e l l  count is  the result of  pre ssure c hange s in the a lveo l i . With 

. I 



3 0  

the fall in internal g l and pre ssure , a s  mi lk i s  removed dur ing the 

milking proce ss , ce l l s  are released into the milk . Thus , milk taken 

from the gland several hours after a milking contains the greate st 

concentration of c e l l s  o f  any of  the milk fractions . In his review 

G i l l  ( 19 7 7 )  c ite s the work of Hoare ( 19 7 7 )  who reported that there 

may not be a true diurnal variation in c e l l  counts .  Hoare ( 1 9 7 7 )  

showed that the number of  c e l l s  shed per milking tends t o  b e  constant 

and that the higher milk production genera l ly obtained at the morning 

milking results in a lower cell concentration ( see Table 2 . 1 1 ) . 

TABLE 2 . 1 1 :  SoMATI C  CELL COUNTS I N  GLENF I ELD DA I RY AVERAGE 
( HOARE , 1 977 C I TED BY G I LL ,  1 977 ) , 

Production ( litre s )  
3 

Log cell count (per ml x 10 ) 

Total No . of cells/milking 

a . m .  

7 . 8  

1 2 3  
8 

9 . 84 X 10  

p . m .  

4 . 1  

246 

10 . 8  X 10
8 

It would appear from the above table that the somatic c e l l  count 

at evening milkings w i l l  be higher than at morning milkings particu­

larly when there is  a long interva l preceding the morning milking . 

Bulk samples for testing should thus be compo sites of both morning 

and evening milkings . 

( iv )  Mi lk Frac tion 

The somatic c e l l  count of milk varies among different fraction s 

o f  milk obtained at a s ingle mi lking ( S chalm et al ,  197 1 )  . These 

authors reported that c e l l  counts in foremi lk are usual ly lower than 

those in strippings milk and the counts in middle milk are the lowe s t . 

Both PMN leucocyte s and mononuc lear c e l l s  are greater in strippings 

than in foremilk but the differenc e s  with regard to mononuc lear cel l s  

are not a s  great a s  with PMN leucocytes .  

Paape and Tucker ( 1966 ) showed that the c e l l  c ounts in residua l  

milk are higher than in foremilk but lower than in strippings . On 



the contrary Natzke & Schultz ( 1 96 7 )  reported higher counts in 

re s idual milk than in str ippings . 

3 1  

I n  view o f  thi s  i t  would seem important to standardize the time 

of samp l ing in the milking routine when taking samples from individual 

\..- c ows . 

( v )  Stre ss Re l ated Factor s 

The effect of " stre s s "  on the cell counts in cow ' s milk is  quite 

c ontroversial . Macadam ( 1 954 ) reported a f ive - fo ld increase in c e l l  

c ounts when milk was drawn from cows after the ir transportation by 

truck to an abattoir . Whi ttle stone et al ( 1 970 ) showed that stre s s ­

f u l  situations such as  isolation from the herd , chasing by a dog and 

sudden thunderstorms can result in an increa se in cell counts .  The se 

s tudie s however have not indicated that cows were free of infection . 

Stress results in the release of  ACTH from the anterior pituitary 

g land which activate s the adrenal cortex to release corticosteroids 

( Paape et al, 1 9 7 3 )  . Thus the administration of ACTH mimic s  some o f  

the phy siological re spons e s  to stre s s . 

Schalm et al ( 1 9 6 5 )  found that in j ecting c orticoids caused a 

r i s e  in leucocyte in c irculating b lood but not in the milk . Paape 

et al ( 1 9 7 3 )  observed no increase in cell c ounts in non-infected 

quarters when cows were inj ected with ACTH . Earlier Whittle stone 

et al ( 1 970 ) had shown that quarters which had a previous history o f  

mastiti s re sulted i n  increases i n  milk somatic c e l l  counts fol lowing 

inj ection of ACTH . Wegner et ai ( 1 9 7 6 )  recorded positive correlations 

between b lood leucocyte s and somatic cell c ounts o f  milk in ma stiti s ­

free cows in j ected with corticotropin and be tween percent b lood 

neutrophi l s  and somatic c e l l  counts of milk in environmental-heat 

s tre ssed cows with no evidence of  current mastiti s . 

It  seems from some o f  the work c ited above that stre s s  may have 

an e ffect on cell count of the udder with previous history of infect ion 

or existing infection . Thi s  means that under stre s s , infections 

otherwise not detected w i l l  show up as  increa s e s  in somatic c e l l  leve l . 

I t  would thus appear that stre s s - induced ce l l  c ount increases have 



3 2  

diagnostic value bec ause they occur i n  animal s  harbouring infection . 

I t  is  apparent from this review of  factors a ffecting c e l l  

counts that there is  a general l ack of agreement about conc lusions 

to be drawn from somatic c e l l  counts .  Neverthe l e s s  there i s  sub­

stantial experimenta l evidence to show that high cell  counts are 

associated with reduced mi lk yield regardle s s  of whether or not they 

are caused by pathogenic organisms . This together with change s in 

milk composition indicate s that there is some di sturbance in the 

milk secretory tissue s . Ce l l  numbers in milk should therefore be 

c onsidered as a measure o f  the functional status o f  the secretory 

t i ssue . For secretory activity , mastitis is probably the most 

important factor causing disturbances in the secretory t i s sue . I t  

would appear that the source s o f  error in interpreting c e l l  counts 

can be reduced if c e l l  counts are carried out at regular interva ls 

over a period and running means calculated to minimize the anomal ies . 

2:2:3: 2 D I AGNOSTI C  METHODS 

Various methods have been deve loped for estimating the c e l l  

content o f  milk ( Scha lm et al ,  1 97 1 )  . Depending o n  whether they are 

used for precise counting or as screening tests , the methods are 

re ferred to as direct or indirec t methods .  

Direct methods involve either the microscopic or electronic 

counting of  somatic c e l l s . The indirect methods depend on te sts for 

the detection of DNA from cellular nuc lear material or the reaction 

between cell DNA and certain chemical reagents which bring s about a 

marked change in viscos ity . Other indirect methods test for change 

in milk composition , e lectr ical conduc tivity , pH measurements and 

enzyrnatic procedure s .  Some of the se methods are briefly de scribed 

in the following discussion ; for deta i l s  see Schalm et al ( 1 9 7 1 ) . 

2 : 2:3 : 2 : 1  D I RECT METHODS FOR SoMAT I C  CELL DETERM I NATI ON 

( i )  Microscopic Counting 

This is a direct v isual method for determining the number of 

c e l l s  in milk . I t  involve s spreading 0 . 0 1 ml of  mi lk sample over 



3 3  

1 cm2 of  a g l a s s  slide and staining the dried smear with methylene 

b lue . The c e l l s  are then counted in a number of  fields with an oil 

immersion lens and expre s sed as the number o f  c e l l s . per ml . Schalm 

et ai ( 1 9 7 1 )  have reviewed this technique fully . The method was 

first described by Pre scott and Breed in 1910 . They expre s sed each 

cell  counted to repre sent 5000 c e l l s/ml . An improved vers ion of 

this method is now re ferred to as  the D irect Micro scopic Somatic 

Cell Count ( D . M . S . C . C . ) . I t  involve s counting the stained ce l l s  

present i n  strips acro s s  the milk film rather than selected micro­

scopic f ie lds and use s a sy stem of  counting dependent on an estimation 

o f  c e l l s  pre sent in the sample (Braz is et ai ,  196 7 ; Ward & Postle , 

1 9 7 0 )  . 

The direct visual identification and counting of somatic c e l l s  

appears t o  b e  the b e s t  re ference method . There is however ,  an 

inherently large variation within one sample associated with micro­

scopic counting due to the fact that the smears may not be homogeneous 

( Heeschen , 1 9 7 5 ) . Moreover the optical and manual operations involved 

are tedious and time consuming which virtually tend to prec lude the ir 

use in large scale screening programme s .  

( i i )  Electronic Somatic Ce l l  Counting 

Elec tronic cell counting is now often used for estimating somatic 

ce l l s  in cow ' s  milk . The method has been successfully used by several 

workers and is reported to be a very rapid and suff iciently accurate 

procedure . The e lectronic counters include the particle counter which 

uses current interruption ( Coulter Sy stem) or l ight beam interruption 

( Auto analyser)  and Fossomatic for detection of stained c e l l s . 

The s igni ficance o f  the Coulter principle l ie s  in its abi l ity to 

record s ingle c e l l s  one by one without sub j e ctive errors when the se 

c e l l s  are suspended as the only partic l e s  in the ir size range in an 

electrolyte ( Hee schen , 1 9 7 5 ) . S ince the main problem encoun tered in 

determining somatic c e l l s  in milk is the el imination of fat partic l e s  

overlapping the s ize range of  c e l l s , it is  neces sary to ensure that 

fat globules are removed be fore somatic cel l s  are counted . Thi s  is  

accomplished by e ither centrifuging ( Cul len , 196 5 ;  Phipps & Newbould , 



34  

196 5 ;  Read et al, 196 7 )  or di spersing the milkfat using a non-ionic 

wetting agent ( Cul len , 196 7 ) . Chemical isolat ion of intact c e l l s  

f rom the fat i s  cons idered more suitable than the centrifugal method 

for regular examination of large number of milk sample s us ing the 

e lectronic c e l l  count because it is quicker , more economica l  in terms 

o f  cost , and more adaptable to automation ( D i j kman et al ,  1969 ; 

Pear son et al ,  1 9 7 0 )  . 

A more de scriptive account on how to measure somatic c e l l s  in 

c ow ' s  milk using a Coulter Counter has been reported by Tolle ( 1 9 7 1 ) . 

Other workers have reported the corre lation coe f f icient between 

e lectronic c e l l  counts and direct microscopic c e l l  counts .  Mitche l 

et al ( 1 96 7 )  reported a correlation o f  0 . 9 7 8 ; Pearson et al ( 1 9 7 0 )  

0 . 9 78 ; Read et al ( 196 7 )  0 . 997 ; Philpot & Pankey ( 19 7 3 )  0 . 8 5 4  to 

0 . 9 7 7  and Newbould ( 1 974 ) 0 . 8 5 to 0 . 92 4 . 

Heeschen ( 1 9 7 5 )  has reported a method whereby procedure s  us ing 

e lectronic c ount ing can be standardized . He has a l so de scribed the 

c ontinuous f low analysi s  (Auto analyse System) which is e s sentia l ly 

an Automated Coul ter Counter and the Optic a l  Fossomatic Instrument . 

The latter i s  an automated microscopic c e l l  count using a DNA-specific 

f luorescent dye . A l so Schmidt Madsen ( 1 9 7 5 )  compared the Fos somatic 

with the Coulter Counter and direc t microscopic count . Re sults 

obtained indicate that the Fossomatic appears s l ightly superior to 

Coulter in pre c i s ion and in ease of operation . 

Although the e lectronic counting system o f  somatic ce l l s  by 

d i fferent instruments is  comparative ly more accurate and capable of  

handl ing large numbers of samp les in a day , standardization continue s 

to be a problem in addition to the high initial cost of equipment . 

2 : 2 : 3 : 2 : 2  IND I RECT METHODS FOR SoMATI C  CELL DETERM INATI ON 

( i )  The Catalase Test 

Catalase is an enzyme found in the ce l l s  of  both anima l s  and 

plants . Mea surement o f  catalase activity was the first among the 

methods in use today for the indirect e stimation of somatic c e l l  

/ 



3 5  

content of milk . The c atalase content o f  normal mi lk is  low excep t  

a t  the beginning and e n d  of lactation ; it increases in mastitis 

( Schalm et a£ ,  1 9 7 1 )  . 

This  test i s  based on the principle that catalase l iberate s 

molecular oxygen and water from hydrogen peroxide . This gas i s  

measured and expressed i n  vo lume percentage which i s  correlated with 

somatic cell count (Paape et al, 1965 ; and Schalm et al ,  1 9 7 1 )  as 

shown in Table 2 . 12 .  For de tailed description of  the te st see Schalm 

et al ( 19 7 1 )  . 

TABLE 2 . 1 2 1  RELATIONSH I P  BETWEEN % GAS AND SOMAT I C  CELL NUMBERS PER 
ML ( FROM SCHALM et a£, 1 97 1 ) .  

% Oxygen Corre lated Somatic Cell  Range ( ce l l s /m l )  

< 2 0  < 5 00 , 000 

20 - 30 500 , 000 - 1 , 000 , 0 00 

30 - 40 1 , 000 , 000 - 2 , 000 , 000 

over 40 over 2 , 000 , 000 

Normal milk genera l ly y ields less than 10% . The method is  not 

suited for use at the side of the cow but is a useful laboratory te s t .  

( i i )  White side Test 

The reaction between mi lk obtained from mastitic cows and sodium 

hydroxide (NaOH ) re sulting in the formation of  a sticky mas s  was first 

descr ibed by Whiteside in 1 9 3 9 . At room temperature the reaction 

takes place rapidly but when mastitic milk is  heated or when milk 

contains pre servatives such as formalin , mercuric chloride and 

potassium dichromate , no reaction take s place ( S chalm et al ,  1 9 7 1 ) . 

The degree o f  reaction i s  graded from 0 - 4 and the se grades are 

correlated with somatic c e l l  count range s as de scribed by Schalm 

et al ( 19 7 1 )  . 

Murphy and Han son ( 194 1 )  modified the te st to the form in which . 

i t  is used today which involve s the addition o f  one drop of  4% sodium 

hydroxide to five drops of cold milk on a glass plate followed by 



vigorous stirring for about 20 seconds . The reaction i s  ba sed on 

the fact that nuc l e i c  acid from somatic ce l l s  forms a sodium salt 

36 

in the pre sence of  sodium hydroxide produc ing a gelatinous mas s  to 

which serum sol ids and fat globules become absorbed to produce a 

characteristic prec ipitate (Dun et al ,  194 3 )  . 

( i i i )  California Ma stitis Test 

The California Ma stiti s Te st ( CMT ) is  an indirect method widely 

used for routine purpo ses in laboratories and under field condition s .  

The test is  sub j ective and has strict l imitation s mainly because o f  

the wide overlap of c e l l  range s ( see Table 2 . 1 3 ) . 

The test wa s fir st des cribed by Schalm & Noor lander ( 1 9 5 7 )  who 

showed that when an an ionic surface active agent i s  added to milk 

and stirred it results in a viscous mixture . The degree of  viscos ity 

depends on the number of somatic c e l l s  pre sent in the milk i . e .  as 

the somatic cell content increase s ,  viscos ity increases . The reagent 

propo sed by Schalm & Noor1ander ( 1 9 5 7 )  wa s alky1 -aryl su1phonate , with 

the addition of  bromcresol purple to indicate the pH of the mixture . 

The test wa s des igned for use as a cow ' s  side test where mi lk from 

four quarters is squirted into four dif ferent depre s sions on a plastic 

paddle . Thereafter an equa l amount of  CMT reagent i s  added and the 

paddle is swirled approximate ly 10 time s and the degree of viscosity 

o f  the contents asse s sed a s  shown in Table 2 . 1 3 . 

TABLE 2 . 1 3 :  EST I MAT I ON OF THE CELL CONTENT OF M I LK BY CMT SCORES 
C SCHALM et al, 1 97 1 ) .  

CMT 

-ve 

Trace 

1 

2 

3 

Cell  Range ( c e l l s /m! ) 

0 -

150 , 000 -

200 , 00 0  

500 , 000 

400 , 000 - 1 , 50 0 , 000 

8 0 0 , 000 - 5 , 000 , 000 

5 , 000 , 000 

Description o f  Vi sible Reaction 

No evidenc e of precipitate 

Sl ight precipitate which may di s­
appear . 

Di stinct pre c ipitate , no gel 
formation . 

Mixture thickens immediately with 
some gel formation . 

A thick gel forms . 



3 7  

Pettersen ( 19 8 1 )  found a highly significant correlation ( P < O . OO l )  

be tween the ce l l  c ount results obtained b y  the Direct Microscop ic 

Count (DMC ) , Electronic Ce l l  Count ( ECC ) and California Mastitis Te st 

(CMT ) . 

Other ind irect te sts inc lude Brabant Mastitis Test , The Michigan 

Ma stitis Test , The Wi sconsin Mastiti s Test and Rolling Bal l  Viscometer . 

Al l these te sts work on a similar princ iple as the CMT i . e .  a s se s sing 

the viscosity that deve lops in mastitic mi lk fol lowing the addition of 

specific reagents to a measured quantity o f  the milk sample . The 

grades are then corre lated with somatic c e l l  count as de scribed by 

Schalm et al ( 19 7 1 ) . 

2 : 2 : 3 : 3  BACTER I OLOGI CAL D I AGNOSI S  

Bacteriological methods are advocated for diagnostic purpose s 

bec ause the cause of  about 90% of mastiti s of economic importance 

(Neave , 1 9 7 5 )  is due to the main pathogens , Staphyio QOQQU4 �� , 

StneptOQOQQU4 agaia�ae, S�eptO QOQQU4 dy�gala�ae, and Stnepto QO QQU4 

ub � ;  and the mo st common methods detect the se bacteria only . The se 

bacteria have been called ' primary ma stitis pathogens '  and the intra­

mammary infections for which the se organisms are re spons ible , as 

' pr imary infections ' .  The se organ i sms and the col iforms c ause 

in fections with marked cellular response , c l inical symptoms and ma jor 

change s in milk yield and composition ( Gri ffin et al , 19 7 7 )  . The 

other bac teria such as coagulase-negative staphy lococcus ( e . g .  

StaphyfOQOQQU4 ep�d�� ) ,  micrococci and corynebacterium have been 

re ferred to as ' secondary mastitis pathogens ' and intramammary 

infections for which the se organi sms are respons ible , " secondary 

infections ' .  These normally cause relatively sma l l  cel lular reactions 

in udder quarters and are rare ly a ssociated with c l inical disease 

( Gr iffin et al ,  1 9 7 7 ) . As the terms ' primary ' and ' secondary ' c ould 

imply sequence rather than severity of  infection of economic importance ,  

these are gradual ly be ing replaced by the terms ' major ' and ' minor ' 

re spective ly (Griffin et al , 197 7 )  . 

Pearson & Greer ( 1974 ) observed in the ir study that 8 0 %  o f  the 

pathogens were primary and 1 5 %  were secondary . In New Zeal and , 



3 8  

Staphylococci and Streptococci have been reported t o  be more commonly 

a s sociated with abnormal udders and high somatic cell counts than 

other bacteria (El liot et al ,  1976 ) . 

The aim in bacteriological examination i s  to obtain a mi lk sample 

free of contamination by extraneous organi sms . Precautions must 

therefore be taken when milk samples are drawn to ensure that bacteria 

do not enter the milk from extraneous s