Livestock Research for Rural Development 20 (8) 2008 Guide for preparation of papers LRRD News

Citation of this paper

Phenotypic relationships among somatic cell count, milk urea content, test-day milk yield and protein percent in dairy cattle

E Z M Oudah

Department of Animal Production, Faculty of Agriculture, Mansoura University, PC: 35516, Mansoura, Egypt
saidauda@yahoo.com

Abstract

A total of 33881 test-day milk records of somatic cell counts (SCC), milk urea content (MU), test-day milk yield (TDM), and milk protein percentage (PP) obtained monthly from January to December 2001 were analyzed to study the phenotypic relationships between them in dairy cattle reared in Czech Republic using individual test-day records. 

 

The overall mean (±SD) of somatic cell counts was 404 ± 750 thousand cell/ml, that of somatic cell score was 3.76 ± 1.85, that of somatic cell count transformated was 3.74 ± 1.84, that of milk urea content was 31.7 ± 9.99 mg/dl, that of test-day milk yield was 22.6 ± 8.50 kg, that of milk protein percentage was 3.38 ± 0.36% and that of days in milk was 155 ± 90 days.  Significant negative correlation coefficients between test-day milk yield and each of milk protein percentage (-0.42), somatic cell counts (-0.13), somatic cell score (-0.22) and days in milk (-0.40), between protein yield (PY) and each of somatic cell counts (-0.12), somatic cell score (-.018) and days in milk (DIM) (-0.33) and between milk urea content and both somatic cell counts (-0.08) and somatic cell score (-0.09) were found.  Test-day milk yield was decreased gradually with increasing the somatic cell measures.  Prediction regression equations were generated to predict both somatic cell counts and milk urea content using different independent variables.  The significant negative correlations suggest that a lower test-day milk yield is phenotypically associated with higher somatic cell counts. 

 

It is recommended that because of association of somatic cell counts and milk yield, it may be important to consider effects of environment on somatic cell counts.

Key words: correlation, dairy cattle, milk traits, milk urea nitrogen, regression equations, somatic cell count


Introduction

The somatic cell counts can serve as an indication of possible udder problems specially mastitis while the milk urea content can be an indication of the nutritional status of the herd.  Somatic cell counts are widely used in milk recording programs in North America and Europe to indicate milk quality and health status of the cow's udder.  On the other hand, Protein content of milk also has received much attention from producers of milk and processors of dairy products. Increased recognition of the value of protein in milk has led processors to pay a premium for milk exceeding minimum protein percentages. To take advantage of such pricing schemes, producers have considered protein performance for culling decisions and breeding plans (Wiggans 1986). The interpretation of these traits and parameters assists the breeders in effectively selecting cows and also in making important management decisions with regard to the health status and nutrition of his herd.  On the other hand, Jones et al. (1984) detected negative relationships of daily milk yields and somatic cell counts.  Eicher et al. (1999) reported insignificant association of somatic cell counts on milk urea nitrogen.  Godden et al (2001) also reported a slightly negative relationship between milk urea nitrogen and linear score.

 

This study raises important issue about predicting one trait (relatively difficult to measure, somatic cell counts and milk urea nitrogen) from other traits (relatively easy to record, milk yield, protein percentage, … etc).  The objectives of the present study were: 1) to estimate the phenotypic relationships among somatic cell counts, milk urea content, test-day milk yield and milk protein percentage in dairy cattle in Czech Republic using test-day milk records and 2) to generate Prediction regression equations to predict both somatic cell counts and milk urea content using different independent variables. 

 

Materials and methods  

Data

 

A total number of 33881 monthly test-day records including test-day milk yield, somatic cell counts, milk urea content and milk protein percentage from January to December 2001 were analyzed to estimate the correlation and regression coefficients among these traits in the milk of Holstein and Czech Spotted cattle reared in Czech Republic using individual test-day records. The data were obtained from Database Computer Center belonging to Czech Moravian Breeders“ Corporation, Inc., (Czech ICAR member), Hiraditcvhko, Czech Republic. Conversion formula: milk urea content (MU, mg/dl) × 0.467 = MUN (milk urea nitrogen, mg/dl).

 

Statistical analyses  

 

Variables were test-day milk yield, protein percentage, protein yield, milk urea content, somatic cell count and days in milk. The "days in milk" was defined as the interval between date of calving and milk test date). Moreover, to evaluate somatic cell counts, three methods of evaluating were used: the first was the actual somatic cell count (1000 cell/ml milk).  The second method was linear somatic cell count score defined by Shook (1982). A score of 0 to 9 was assigned to each sample day according to somatic cell counts level and each integer increase in linear score is associated with a doubling of the actual somatic cell counts (Shook 1982).  The third method was somatic cell count transformed, where somatic cell counts had been transformed to somatic cell count transformed with the base 2 log scale used by Dabdoub and Shook (1984) using the international formula: [SCT = log2 (SCC / 100,000) + 3].  Data analysis was performed using the Statistical Analysis System (SAS/STAT 9.1 User's Guide, 2004.  The statistical analyses were conducted using PROC FREQ, PROC MEANS, PROC CORR and PROC REG.

 

Results and discussion

Descriptive statistics 

 

Summary of the descriptive statistics of test-day records used in this study is shown in Table 1. 


Table 1.  Unadjusted means, standard deviation (SD), minimum and maximum of studied traits

Trait

No.

Mean

SD

Min.

Max.

test-day milk yield, kg

31848

22.6

8.50

3

59.8

Protein percentage, %

31517

3.38

0.36

2.03

4.98

Protein yield, kg

31502

0.75

0.26

0.07

1.91

Somatic cell count, 1000 cell/ml

22857

404

750

1

6992

Somatic cell score

22857

3.76

1.85

0

9

somatic cell count tranformated

22857

3.74

1.84

-3.64

9.13

Milk urea content, mg/dl

30788

31.7

9.99

10

60

Days in milk, day

31750

155

90.0

5

400


The unadjusted mean of test-day milk yield of 31848 records (±SD) was 22.6 ± 8.50 kg with protein percentage 3.38 ± 0.36%. The high SD of test-day milk yield is expected where the animals used in the study belong to different herds, are in different lactation and from different locations.  According to results of Czech-Moravia Breeders Association (CMBA), Prague,  Hanuš et al (2002) reported that the overall mean of protein percentage of the dairy cows (all breeds and all lactations) of the individual milk samples during standard lactations in the Czech milk recording in 2002 was 3.35 ± 0.30% which is in agreement with the  result of the present study. Johnson and Young (2003) in USA found that milk yield in Holstein and jersey cows were 33.8±10.73 and 22.4 ± 7.2 kg, respectively with PP 3.19 ± 0.37 and 3.70 ± 0.45%, respectively.    

 

The unadjusted mean of somatic cell counts for 22857 records in the present study was 404±750 thousand cell/ml which was also near from the value of Hanuš et al (2002) (400±300 thousand cell/ml).  The overall mean (±SD) of somatic cell counts was 404 ± 750 thousand cell/ml, that of somatic cell score was 3.76 ± 1.85, and that of somatic cell count transformated was 3.74 ± 1.84.  The unadjusted mean of milk urea content was 31.7±9.99 mg/dl.  The results of Czech-Moravia Breeders Association (CMBA), Prague regarding milk urea content for all Czech republic reported by Hanuš et al (2002) was 36 ± 19 mg/dl.  The very high standard deviation of somatic cell counts in this study (750 thousand cell/ml) reflect very high variability between herds. Johnson and Young (2003) found that somatic cell counts in Holstein and jersey cows were 270±741 and 336±909 thousand cell/ml milk, respectively and the corresponding values for linear score were  2.57 ± 2.11 and 2.88 ± 2.11 for the two breeds, respectively.

 

The overall mean of milk urea content in the present study was 31.7 ± 9.99 mg/dl (this amount equal 14.8 mg/dl milk urea nitrogen) which was nearly similar to those reported by Johnson and Young (2003) for Holstein and Jersey cows (15.5 ± 3.73 and 14.1 ± 3.20 mg/dl milk urea nitrogen, respectively).  The overall mean of days in milk was 155 ± 90 days.  This value was lower than that reported by Johnson and Young (2003) in USA who found that Mean DIM was 194 ± 129 days for Holsteins and 180 ± 130 days for Jerseys. 

 

Phenotypic correlation coefficients among traits 

 

Pearson correlation coefficients among test-day milk yield, milk protein percentage, protein yield, somatic cell counts, somatic cell score, milk urea content and days in milk are presented in Table 2. 


Table 2.  Pearson correlation coefficients (below diagonal) and number of observations (above diagonal) among different traits

Trait

1

2

3

4

5

6

7

1 TDM

 

31502

31502

22835

22835

30776

31735

2 PP

- 0.42

 

31502

22816

22816

30741

31404

3 PY

0.96

- 0.17

 

22802

22802

30729

31389

4 SCC

- 0.13

0.11

- 0.12

 

22857

22266

22277

5 SCS

- 0.22

0.19

- 0.18

0.740

 

22266

22777

6 MU

0.24

- 0.004

0.26

- 0.08

-0.093

 

30678

7 DIM

- 0.40

0.44

- 0.33

0.024

0.131

- 0.026

 

All correlation values were significant at P<0.001, except between milk protein percentage and milk urea content was not significant.


There were significant negative correlation coefficients between test-day milk yield and each of milk protein percentage (-0.42), somatic cell counts (-0.13), somatic cell score (-0.22) and days in milk (-0.40), between protein yield and each of somatic cell counts (-0.12), somatic cell score (-.018) and days in milk (-0.33) and between milk urea content and both somatic cell counts (-0.08) and somatic cell score (-0.09).  These significant negative correlations suggest that a lower test-day milk yield is phenotypically associated with higher somatic cell counts.  Moreover, there were significant positive correlation coefficients between test-day milk yield and both protein yield (0.96) and milk urea content (0.24), between milk protein percentage and both somatic cell counts (0.11), somatic cell score (0.19) and days in milk (0.44), and between milk urea content and protein yield (0.26) (Table 2). 

 

Comparing the present results with the other investigators, Kiiman and Kaart (2004) working on Estonian Red cattle fount that phenotypic correlation coefficient between log somatic cell counts and both milk protein percentage and protein yield in the first lactation were 0.195 and -0.334.  Kureoja and Kaart (2004) found that the phenotypic correlation coefficients between milk urea content and milk yield, protein yield and protein percent were 0.20, 0.16, -0.14, respectively using Estonian Red cattle.  They found also that the corresponding values using Estonian Holstein cattle were 0.19, 0.14, -0.06, respectively.  Johnson and Young (2003) concluded that milk urea N concentrations were positively associated with milk yield and negatively associated with milk protein. They added that milk urea N concentration appeared to have an inverse association with somatic cell counts. They suggest that milk urea N concentrations should be evaluated in association with breed, days in milk, milk yield, and protein percentage when determining the efficiency of N utilization. The present results are in close agreement with those of Johnson and Young (2003).

 

Prediction regression equations

 

To predict both somatic cell counts and/or milk urea content in test-day milk recording, three prediction regression equations were generated according to which independent variables are available (Table 3). 


Table 3.  Estimates of intercept and partial regression coefficient (±SE) for somatic cell counts and milk urea content predictions using different independent variables

Dependent variable

Intercept (± SE)

Partial regression coefficient (± SE) of

1st independent variable

2nd independent variable

3rd independent variable

Somatic cell counts (1000/ml)

673±14.0

-12.1±0.59 MY

 

 

771±18.8

-11.0±0.63 MY

-3.80 ± 0.51 MUN

 

146±58.4

-7.65±0.69 MY

-4.42 ± 0.51 MUN

168 ± 15.0 PP

Milk urea content
(mg/dl)

25.4 ± 0.16

0.282±0.007 MY

 

 

24.6 ± 0.19

0.329±0.008 MY

-0.00066±0.00009 SCC

 

9.93 ± 0.77

0.397± 0.009 MY

-0.00077±0.00009 SCC

3.91 ± 0.20 PP

Analyses of variance of all regression models were significant at P<0.001.


For predicting somatic cell counts in test-day milk yield record, using test-day milk yield, milk urea content and/or milk protein percentage and for milk urea content using test-day milk yield, somatic cell counts and/or milk protein percentage.  These prediction regression equations could be applied in the regions where somatic cell counts and/or milk urea content estimation is expensive or need time for analyses. Using this tool also give quick chick about udder health waiting for the results of laboratory analyses.    

 

Conclusions 

 

Acknowledgements 

The author would like to thanks Pavel Bucek, Czech Moravian Breeders“ Corporation, Inc., Hradistko for providing the data. Thanks also are extended to European Commission, (Tempus Project) for financial support through the individual mobility grant (IMG) to Czech Republic (IMG_EG3044-2004).

 

References  

Dabdoub S M and Shook G E 1984 Phenotypic relations among milk yield, somatic cell count and clinical mastitis. Journal of Dairy Science 67(Supplement. 1):163–164

 

Eicher R, Bouchard E and Bigras-Poulin M 1999 Factors affecting milk urea nitrogen and protein concentrations in Quebec dairy cows. Preventive Veterinary Medicine 39:53–63

 

Godden S M, Lissemore K D, Kelton D F, Leslie K E, Walton J S and Lumsden J H 2001. Factors associated with milk urea concentrations in Ontario dairy cows. Journal of Dairy Science 84:107–114 http://jds.fass.org/cgi/reprint/84/1/107.pdf

 

Hanuš O, Pytloun J, RķhaJ, Hering P, Matouš E and Genčurovį V 2002  First meeting of ICAR Reference Laboratory Network (Interlaken, 27 May 2002)–Report http://www.icar.org/Documents/milk_laboratories_leray/NWmeeting-1.pdf

 

Johnson R G and Young A J 2003  The Association Between Milk Urea Nitrogen and DHI Production Variables in Western Commercial Dairy Herds Journal of Dairy Science 86:3008–3015 http://jds.fass.org/cgi/reprint/86/9/3008.pdf        

 

Jones G M, Pearson R E, Clabaugh G A and Heald C W 1984 Relationships between somatic cell counts and milk production. Journal of Dairy Science 67:1823-1831 http://jds.fass.org/cgi/reprint/67/8/1823.pdf

 

Kiiman H and Kaart T 2004 In: Proceedings of the 10th  Animal Breeding in Baltics Conference. 13-14 May, Institute of Animal Science, Estonian Agricultural University, Tartu, Estonia 2004. P: 36-39 http://www.eau.ee/~lki/babc/X/  

 

Kureoja A and Kaart T 2004 Genetic and environmental influences on urea concentration in dairy cows, milk. In: Proceedings of the 10th  Animal Breeding in Baltics Conference. 13-14 May, Institute of Animal Science, Estonian Agricultural University, Tartu, Estonia, 2004. P: 42-47 http://www.eau.ee/~lki/babc/X/      

 

SAS 2004  SAS Institute Inc. 2004. SAS/SAT© 9.1 user's guide Cary. NC: SAS Institute Inc.

 

Shook G E 1982 Relationship between lactation measures of somatic cell concentration and milk yield. Journal of Dairy Science 65: 419-425 http://jds.fass.org/cgi/reprint/65/3/419.pdf

  

Wiggans G R 1986  Projection of Protein Yield to 305 Days Journal of Dairy Science 69: 2908–2914 http://jds.fass.org/cgi/reprint/69/11/2908.pdf



Received 16 May 2008; Accepted 8 June 2008; Published 5 August 2008

Go to top