Udder Health in Dairy Heifers

Transcription

1

2

3 Udder Health in Dairy Heifers Some Epidemiological and Microbiological Aspects Sarne De Vliegher Merelbeke, 2004

4

5 "Twijfel er echter niet aan, de weg om te slagen is moed houden en geduld, en stevig doorwerken." - "But do not doubt it, the way to succeed is to keep courage and patience, and to work on energetically." (Vincent Van Gogh aan zijn broer Theo - begin februari 1886, Antwerpen - Vincent Van Gogh to his brother Theo - early February 1886, Antwerp) Voor mijn ouders Voor Sofie

6 Udder Health in Dairy Heifers Some Epidemiological and Microbiological Aspects Sarne De Vliegher Cover: Paly De Vliegher Printing: Plot-it, Merelbeke Printing of this thesis was financially supported by Pfizer, Boehringer-Ingelheim, Intervet, and Elanco.

7 Udder Health in Dairy Heifers Some Epidemiological and Microbiological Aspects Uiergezondheid bij Vaarzen Enkele Epidemiologische en Microbiologische Aspecten (met een samenvatting in het Nederlands) Proefschrift voorgedragen tot het behalen van de graad van Doctor in de Diergeneeskundige Wetenschappen aan de Faculteit Diergeneeskunde, Universiteit Gent, 26 november, 2004 door Sarne De Vliegher Vakgroep Voortplanting, Verloskunde en Bedrijfsdiergeneeskunde, Faculteit Diergeneeskunde, Universiteit Gent Department of Reproduction, Obstetrics, and Herd Health, Faculty of Veterinary Medicine, Ghent University

8 Promotors Prof. Dr. Dr. h. c. A. de Kruif Department of Reproduction, Obstetrics, and Herd Health, Faculty of Veterinary Medicine, Ghent University, Belgium Prof. Dr. H. W. Barkema Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Canada Copromotor Prof. Dr. G. Opsomer Department of Reproduction, Obstetrics, and Herd Health, Faculty of Veterinary Medicine, Ghent University, Belgium Additional Members of the Examination Committee Prof. Dr. F. Haesebrouck, Chairman Faculty of Veterinary Medicine, Ghent University, Belgium Dr. L. A. Devriese Faculty of Veterinary Medicine, Ghent University, Belgium Prof. Dr. C. Burvenich Faculty of Veterinary Medicine, Ghent University, Belgium Prof. Dr. Y. H. Schukken College of Veterinary Medicine, Cornell University, USA Prof. Dr. P. Deprez Faculty of Veterinary Medicine, Ghent University, Belgium Dr. Lic. L. De Meulemeester Animal Health Service Flanders, Belgium

9

10

11 Table of Contents Chapter 1 General Introduction 1 Chapter 2 Aims and Outline of the Thesis 11 Chapter 3 Heifer Mastitis: a Review 15 Chapter 4 Somatic Cell Counts in Dairy Heifers during Early Lactation 41 Chapter 5 Chapter 5.1 Chapter 5.2 Chapter 5.3 Significance of Early Lactation Somatic Count in Dairy Heifers for the Subsequent First Lactation 51 Impact of Early Lactation Somatic Cell Count in Heifers on Somatic Cell Counts over the First Lactation 53 Impact of Early Lactation Somatic Cell Count in Heifers on Milk Yield over the First Lactation 79 Association between Somatic Cell Count in Early Lactation and Culling of Dairy Heifers using Cox Frailty Models 101 Chapter 6 Chapter 7 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count of Belgian Dairy Heifers 123 Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers 149 Chapter 7.1 Prepartum Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers is Associated with Low Somatic Cell Counts in Early Lactation 151 Chapter 7.2 In vitro Growth Inhibition of Major Mastitis Pathogens by Staphylococcus chromogenes originating from Teat Apices of Dairy Heifers 165 Chapter 8 General Discussion 179

12 Summary 203 Samenvatting 211 Acknowledgements - Dankwoord 219 Curriculum Vitae - Publications 227

13

14

15 Chapter 1 General Introduction S. De Vliegher Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium Sarne.Devliegher@UGent.be

16 Chapter 1 2

17 General Introduction Introduction Despite considerable research effort, mastitis remains the most costly disease in dairy cattle (Bradley, 2002). Economical losses are due to - in arbitrary order - suboptimal milk production, extra labour, treatment and drug costs, veterinary fees, early culling, discarded milk, death, production problems encountered by the dairy processors (Smith and Hogan, 2001), decreased fertility (Barker et al., 1998; Schrick et al., 2001; Santos et al., 2004), and reduced shelf life (Barbano, 2004). In addition, the general public has become more and more concerned about animal welfare. In that respect, one should not ignore (per)acute mastitis (Bradley, 2002). Mastitis Control Programs The "standard mastitis control plan", as proposed by Neave et al. (1969), has had considerable success in reducing the prevalence of subclinical and clinical mastitis in dairy herds where it has been applied (Bradley, 2002), and comprises 5 points: appropriate treatment of clinical mastitis, culling of chronically infected cows, post-milking teat disinfection, correct maintenance and use of the milking equipment, and application of blanket dry cow therapy. The "5-point program", as it is usually referred to, has helped to control the contagious mastitis pathogens Streptococcus agalactiae, and to a lesser extent, Streptococcus dysgalactiae and Staphylococcus aureus, resulting in a decrease of the average bulk milk somatic cell count. The program was, however, less effective in the fight against environmental pathogens, and has, for that reason, been extended by the National Mastitis Council (NMC - A global organization for mastitis control and milk quality) towards a 10- point "recommended mastitis control program". This extended program emphasises maintenance of a clean and comfortable environment, besides goal setting for udder health, good record keeping, regular monitoring of the udder health, evaluation of the implemented control programs, and the aforementioned 5 points (NMC, 2001). Heifer Mastitis In the past, heifers udders were generally regarded as uninfected prior to first calving and, in raising replacement heifers, emphasis was placed on genetic improvement, reproduction, nutrition, and immunization against diseases unrelated to the mammary gland 3

18 Chapter 1 (Pankey et al., 1991). Nevertheless, during the last 2 decades, several studies have demonstrated that intramammary infection (IMI) in nulliparous heifers is not uncommon at all, resulting in impairment of mammary development and potential detrimental effects on future milk production (Munch-Petersen, 1970; Meaney, 1981; Oliver and Mitchell, 1983; Daniel et al., 1986; Trinidad et al., 1990a; Pankey et al., 1991; Matthews et al., 1992; Myllys, 1995; Fox et al., 1995; Nickerson et al., 1995; Aarestrup and Jensen, 1997). Coagulasenegative staphylococci (CNS) are the most common cause of IMI in heifers, although S. aureus and environmental pathogens can also play a role. In 1 study, as many as 97% of heifers had IMI at parturition, indicating the problem should not be underestimated (Trinidad et al., 1990a). There exists a large variation between farms in the prevalence of IMI and the distribution of causative bacteria, suggesting management plays an important role (Fox et al., 1995). In addition, the incidence of clinical mastitis in heifers is very high during the first 14 d of lactation (Barkema et al., 1998; Waage et al., 1999). Heifers are the future of the dairy herd; therefore, mastitis control programs should include the prevention of mastitis in heifers. No data are available on udder health in recently calved Flemish (Belgium) dairy heifers, although many farmers and veterinarians have observed heifers calving with nonfunctional quarters, clinical mastitis, or elevated somatic cell counts. Dairy Herd Improvement data are regularly recorded on a large number of herds and are used by both farmers and veterinarians as a practical tool in controlling udder health. Dairy Herd Improvement data can be used to estimate the magnitude and pattern of the heifer mastitis problem. In addition, these records can be used to quantify the effect of udder health disorders at calving on udder health, culling hazard, and production during the first lactation, based on data of a large number of animals. The quantification of these effects is needed to get some insight in the overall negative economic implications of heifer mastitis. Some risk factors for heifer mastitis (clinical and subclinical), both at the herd- and heifer-level, have been identified (Fox et al., 1995; Myllys and Rautala, 1995; Østerås et al., 1997; Owens et al., 1998; Roberson et al., 1998; Waage et al., 1998; Bareille et al., 2000; Waage et al., 2001), and these risk factors need to be addressed in any mastitis prevention program. However, the pathogenesis of heifer mastitis remains largely unknown, making prevention difficult. In the US, prepartum antibiotic treatment of heifers is proposed as a simple and effective method for controlling heifer mastitis (Oliver et al., 2004a). Both lactating (Oliver et al., 1992; Oliver et al., 2004b; Fox et al., 2004) and dry cow products (Trinidad et al., 1990b; Owens et al., 1994; Owens et al. 2001) have been used for this purpose and are apparently successful in reducing the prevalence of infected heifers and 4

19 General Introduction quarters at calving. Only some studies have examined the long-term effect of these treatments on production and udder health, but were either not conclusive (Fox et al., 2004) or were restricted to 2 herds (Oliver et al., 2003). Therefore, extrapolation is difficult at present. Additionally, caution is needed because of the likely risk of antibiotic residues in the milk and emerging antibiotic resistance when this approach is used. Blanket dry cow treatment has been questioned for these reasons and a non-antibiotic approach has been proposed when appropriate (Woolford et al., 1998; Huxley et al., 2002). Whereas the use of antibiotics prepartum can be justified under certain circumstances, management improvements should be implemented first. More studies identifying variables associated with heifers' post-partum udder health are needed to elucidate the management factors important in prevention of heifer mastitis (Barkema, 1998). Factors that affect the incidence and prevalence of IMI in heifers may be herd-level factors (e.g. fly control program in use or not), cow-level variables (e.g. age at first calving) or quarter-level factors (e.g. presence or absence of abrasions of the teat). Knowing where most of the variability in the occurrence of IMI exists (i.e. at which level) can help target research into the appropriate area. For example, if most of the variability exists between quarters within a heifer, then research into quarter-level factors will likely be most productive in terms of innovations in mastitis prevention. Inducing or maintaining a protective flora at the teat end (Woodward et al., 1988) would be such an innovative approach. Although Woodward et al. (1987) demonstrated that 25% of all isolates of the normal teat skin flora of non-lactating heifers were able to inhibit the growth of selected mastitis pathogens, only few studies elaborating on these findings have been conducted so far. In that respect, CNS could play an important role as they are part of the normal teat flora (Devriese and De Keyser, 1980; Boddie et al., 1987; White et al., 1989), are generally considered to be minor pathogens, and possess interesting protective capabilities against the development of IMI with major pathogens (Linde et al., 1980; Brooks and Barnum, 1984; Schukken et al., 1989; Matthews et al., 1990; Matthews et al., 1991; Lam et al., 1997). Identification of in vitro protective CNS species and exploration of possible mechanisms behind the protective characteristics are primary steps in exploring the potential of protective teat flora. Studying the association between prepartum teat apex colonization and postpartum udder health would provide additional information on the potential for this new approach in mastitis prevention. 5

20 Chapter 1 References Aarestrup, F. M., and N. E. Jensen Prevalence and duration of intramammary infection in Danish heifers during the prepartum period. J. Dairy Sci. 80: Barbano, D The role of milk quality in addressing future dairy food marketing opportunities in a global economy. Proc. 43rd Ann. NMC meeting, Charlotte, North Carolina, United States of America: Bareille, N., H. Seegers, M. B. Kiebre-Toe, F. Beaudeau, and C. Fourichon Risk factors for elevated milk somatic cell counts during early lactation in dairy heifers. Proc. 10th Intern. Congr. Anim. Hyg., Maastricht, The Netherlands: Barkema, H. W Udder health on dairy farms. A longitudinal study. PhD thesis, Utrecht, The Netherlands. Barkema, H. W., Y. H. Schukken, T.J.G.M. Lam, M. L. Beiboer, H. Wilmink, G. Benedictus, and A. Brand Incidence of clinical mastitis in dairy herds grouped in 3 categories by bulk milk somatic cell counts. J. Dairy Sci. 81: Barker, A. R., F. N. Schrick, M. J. Lewis, H. H. Dowlen, and S. P. Oliver Influence of clinical mastitis during early lactation on reproductive performance of Jersey cows. J. Dairy Sci. 81: Boddie, R. L., S. C. Nickerson, W. E. Owens, and J. L. Watts Udder microflora in nonlactating heifers. Agri-Practice 8: Bradley, A. J Bovine mastitis: an evolving disease. Vet. J. 164: Brooks, B. W., and D. A. Barnum The susceptibility of bovine udder quarters colonized with Corynebacterium bovis to experimental infection with Staphylococcus aureus or Streptococcus agalactiae. Can. J. Comp. Med. 48: Daniel, R. C., D. A. Barnum, and K. E. Leslie Observations on intramammary infections in first calf heifers in early lactation. Can. Vet. J. 27: Devriese, L. A, and H. De Keyser Prevalence of different species of coagulasenegative staphylococci on teats and in milk samples from dairy cows. J. Dairy Res. 47: Fox, L. K., S. T. Chester, J. W. Hallberg, S. C. Nickerson, J. W. Pankey, and L. D. Weaver Survey of intramammary infections in dairy heifers at breeding age and first parturition. J. Dairy Sci. 78: Fox, L. K., A. A. Borm, K. E. Leslie, J. S. Hogan, S. M. Andrew, S. P. Oliver, Y. H. Schukken, W. E. Owens, and C. Norman Effect of prepartum antibiotic therapy 6

21 General Introduction in heifers on milk production and mastitis postpartum. Proc. 43rd Ann. NMC meeting, Charlotte, North Carolina, United States of America: Huxley J. N., M. J. Green, L. E. Green, and A. J. Bradley Evaluation of the efficacy of an internal teat sealer during the dry period. J. Dairy Sci. 85: Lam, T. J., Y. H. Schukken, J. H. van Vliet, F. J. Grommers, M. J. Tielen., and A. Brand Effect of natural infection with minor pathogens on susceptibility to natural infection with major pathogens in the bovine mammary gland. Am. J. Vet. Res. 58: Linde, C., O. Holmberg, and G. Astrom The interference between coagulase-negative staphylococci and Corynebacterium bovis and the common udder pathogens in the lactating cow. Nord. Vet. Med. 32: Matthews, K. R., R. J. Harmon, and B. A. Smith Protective effect of Staphylococcus chromogenes infection against Staphylococcus aureus infection in the lactating bovine mammary gland. J. Dairy Sci. 73: Matthews, K. R., R. J. Harmon, and B. E. Langlois Effect of naturally occurring coagulase-negative staphylococci infections on new infections by mastitis pathogens in the bovine. J. Dairy Sci. 74: Matthews, K. R., R. J. Harmon, and B. E. Langlois Prevalence of Staphylococcus species during the periparturient period in primiparous and multiparous cows. J. Dairy Sci. 75: Meaney, W. J Mastitis levels in spring-calving dairy heifers. Ir. Vet. J. 35: Munch-Petersen, E Mastitis in bovine primiparae. Vet. Rec. 87: Myllys, V Staphylococci in heifer mastitis before and after parturition. J. Dairy Res. 62: Myllys, V., and H. Rautala Characterization of clinical mastitis in primiparous heifers. J. Dairy Sci. 78: Neave, F. K., F. H. Dodd, R. G. Kingwill, and D. R. Westgarth Control of mastitis in the dairy herd by hygiene and management. J. Dairy Sci. 52: Nickerson, S. C., W. E. Owens, and R. L. Boddie Mastitis in dairy heifers: initial studies on prevalence and control. J. Dairy Sci. 78: NMC - A global organization for mastitis control and milk quality (2001). Recommended mastitis control program. Oliver, S. P., and B. A. Mitchell Intramammary infections in primigravid heifers near parturition. J. Dairy Sci. 66:

22 Chapter 1 Oliver, S. P., M. J. Lewis, B. E. Gillespie, and H. H. Dowlen Influence of prepartum antibiotic therapy on intramammary infections in primigravid heifers during early lactation. J. Dairy Sci. 75: Oliver, S. P., M. J. Lewis, B. E. Gillespie, H. H. Dowlen, E. C. Jaenicke, and R. K. Roberts Prepartum antibiotic treatment of heifers: milk production, milk quality and economic benefit. J. Dairy Sci. 86: Oliver, S. P., B. E. Gillespie, S. J. Headrick, M. J. Lewis, and H. H. Dowlen. 2004a. Heifer mastitis: prevalence, risk factors and control strategies. Proc. 43rd Ann. NMC meeting, Charlotte, North Carolina, United States of America: Oliver, S. P., S. J. Ivey, B. E. Gillespie, M. J. Lewis, D. L. Johnson, K. C. Lamar, H. Moorehead, H. H. Dowlen, S. T. Chester, and J. W. Hallberg. 2004b. Influence of prepartum intramammary infusion of pirlimycin hydrochloride or penicillin-novobiocin on mastitis in heifers during early lactation. J. Dairy Sci. 84: Østerås, O., R. B. Larssen, and E. Simensen Environmental risk factors associated with mastitis in heifers. Proc. 9th Intern. Congr. Anim. Hyg., Helsinki, Finland: Owens, W. E., S. C. Nickerson, P. J. Washburn, and C. H. Ray Prepartum antibiotic therapy with a cephapirin dry-cow product against naturally occurring intramammary infections in heifers. J. Vet. Med. Ser. B. 41: Owens, W. E., S. P. Oliver, B. E. Gillespie, C. H. Ray, and S. C. Nickerson Role of horn flies (Haematobia irritans) in Staphylococcus aureus-induced mastitis in dairy heifers. Am. J. Vet. Res. 59: Owens, W. E., S. C. Nickerson, R. L. Boddie, G. M. Tomita, and C. H. Ray Prevalence of mastitis in dairy heifers and effectiveness of antibiotic therapy. J. Dairy Sci. 84: Pankey, J. W., P. A. Drechsler, and E. E. Wildman Mastitis prevalence in primigravid heifers at parturition. J. Dairy Sci. 74: Roberson, J. R., L. K. Fox, D. D. Hancock, J. M. Gay, and T. E. Besser Sources of intramammary infections from Staphylococcus aureus in dairy heifers at first parturition. J. Dairy Sci. 81: Santos, J.E.P., R.L.A. Cerri, M. A. Ballou, G. E. Higginbotham, and J. H. Kirk Effect of timing of first clinical mastitis occurrence on lactational and reproductive performance of Holstein dairy cows. J. Anim. Repr. Sci. 80:

23 General Introduction Schukken, Y. H., D. Van De Geer, F. J. Grommers, J.A.H. Smit, and A. Brand Intramammary infections and risk factors for clinical mastitis in herds with low somatic cell counts in bulk milk. Vet. Rec. 125: Schrick, F. N., M. E. Hockett, A. M. Saxton, M. J. Lewis, H. H. Dowlen, and S. P. Oliver Influence of subclinical mastitis during early lactation on reproductive parameters. J. Dairy Sci. 84: Smith, K. L., and J. S. Hogan The world of mastitis. Proc. 2nd Intern. Symp. Mastitis and Milk Quality. Vancouver, British Columbia, Canada:1-12. Trinidad, P., S. C. Nickerson, and T. K. Alley. 1990a. Prevalence of intramammary infection and teat canal colonization in unbred and primigravid dairy heifers. J. Dairy Sci. 73: Trinidad, P., S. C. Nickerson, T. K. Alley, and R. W. Adkinson. 1990b. Efficacy of intramammary treatment in unbred and primigravid dairy heifers. J. Am. Vet. Med. Assoc. 197: Waage, S., S. Sviland, and S. A. Ødegaard Identification of risk factors for clinical mastitis in dairy heifers. J. Dairy Sci. 81: Waage, S., T. Mørk, A. Røros, D. Aasland, A. Hunshamar, and S. A. Ødegaard Bacteria associated with clinical mastitis in dairy heifers. J. Dairy Sci. 82: Waage, S., S. A. Ødegaard, A. Lund, S. Brattgjerd, and T. Rothe Case-control study of risk factors for clinical mastitis in postpartum dairy heifers. J. Dairy Sci. 84: White, D. G., R. J. Harmon, J. E. Matos, and B. E. Langlois Isolation and identification of coagulase-negative Staphylococcus species from bovine body sites and streak canals of nulliparous heifers. J. Dairy Sci. 72: Woodward, W. D., T. E. Besser, A. C. S. Ward, and L. B. Corbeil In vitro growth inhibition of mastitis pathogens by bovine teat skin normal flora. Can. J. Vet. Res. 51: Woodward, W. D., A. C. Ward, L. K. Fox, and L. B. Corbeil Teat skin normal flora and colonization with mastitis pathogen inhibitors. Vet. Microbiol. 17: Woolford, M. W., J. H. Williamson, A. M. Dat, and P. J. A. Copeman The prophylactic effect of a teat sealer on bovine mastitis during the dry period and the following lactation. N. Z. Vet. J. 46:

24

25 Chapter 2 Aims and Outline of the Thesis S. De Vliegher Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium Sarne.Devliegher@UGent.be

26 Chapter 2 12

27 Aims and Outline of the Thesis The main aims of this thesis are: To describe somatic cell count in early lactation of dairy heifers as a proxy for peripartum udder health, and to explore variation among farms (Chapter 4). To study the association between somatic cell count in early lactation of heifers and testday somatic cell counts during the first lactation (Chapter 5.1). To study the association between somatic cell count in early lactation of heifers and testday milk yield during the first lactation (Chapter 5.2). To study the association between somatic cell count in early lactation and the culling hazard of heifers (Chapter 5.3). To identify management practices and heifer-level variables associated with somatic cell count in early lactation of heifers and to identify the level at which intervention would be most successful (Chapter 6). To study the association between colonization of teat apices of prepartum heifers by Staphylococcus chromogenes, one of the most prevalent coagulase-negative staphylococci causing intramammary infection in heifers at calving, and udder health in early lactation (Chapter 7.1). To elucidate possible mechanisms behind the potential protective capabilities of S. chromogenes towards major mastitis pathogens in vitro (Chapter 7.2). 13

28

29 Chapter 3 Heifer Mastitis: a Review S. De Vliegher Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium Sarne.Devliegher@UGent.be

30 Chapter 3 16

31 Heifer Mastitis: a Review Abbreviation Key CNS = coagulase-negative staphylococci, CPS = coagulase-positive staphylococci, DIM = days in milk, IMI = intramammary infection, SCC = somatic cell count. 17

32 Chapter 3 18

33 Heifer Mastitis: a Review Introduction Heifers represent the future production of every dairy herd and are expected to produce high quality milk. They should freshen with a healthy and well-developed udder to reach that goal. However, a large proportion of heifers have infected quarters at calving, indicating that the once generally accepted assumption that heifers' udders are sterile, is not true. The term "heifer mastitis" should not be confused with "summer mastitis". The latter is especially seen in primigravid heifers, but also in dry cows, and is caused by Arcanobacterium pyogenes, Peptococcus indolicus and/or Streptococcus dysgalactiae, and spread by the fly Hydrotea irritans (Nickerson et al., 1995). When "heifer mastitis" is discussed, we refer to the large proportion of dairy heifers calving with infected quarters, likely resulting in damaging implications for the future productive life of this important group of animals. This chapter consists of a literature review on heifer mastitis. The different sections focus on: 1) prevalence and incidence of heifer mastitis; 2) persistence of intramammary infection (IMI) into the first lactation; 3) the consequences of heifer mastitis at calving for future udder health, production, longevity, and fertility; 4) risk factors associated with heifer mastitis; and 5) antibiotic treatment with dry cow and lactating cow products as a method for controlling heifer mastitis. Prevalence and Incidence of Heifer Mastitis Although heifer mastitis was not unheard of before (Schalm, 1942; Munch-Petersen, 1970), it was only 20 years ago that the first studies demonstrated that a large proportion of heifers and quarters were infected at calving (Meaney, 1981; Oliver and Mitchell, 1983). These studies were followed by additional studies mapping heifer mastitis in North America (Daniel et al., 1986; Boddie et al., 1987; Trinidad et al., 1990b; Pankey et al., 1991; Miller et al., 1991; Matthews et al., 1992; Oliver et al., 1992; Roberson et al., 1994a; Fox et al., 1995; Roberson et al., 1996; Oliver et al., 1997; Oliver et al., 2004a) and Europe (Sobiraj et al., 1988; van der Meer et al., 1993; Myllys, 1995; Aarestrup and Jensen, 1997). Trinidad et al. (1990b) reported that as many as 94 out of 97 unbred and primigravid heifers and 75% of quarters were infected prepartum with 29% of heifers showing clinical symptoms. In addition, IMI can occur at a very early age (Boddie et al., 1987; Trinidad et al., 19

34 Chapter a). One third of the 2435 quarters sampled from 1583 heifers at breeding age (8 to 19 mo) had IMI (Fox et al., 1995). A wide variation in prevalence of infected heifers and quarters among studies has been reported. Between 71.0 and 13.9% of the sampled quarters were culture negative before calving (Table 1), whereas this was 81.7 to 64.0% at calving (Table 2), and 87.3 to 44.4% in early lactation (Table 3). Table 1. Heifer- and quarter-level prevalence of intramammary infections before calving. Study Sample level 1 n Neg. 2 CNS 3 S. aureus Env. 4 Other 5 Oliver and Mitchell, F H 32 Q Boddie et al., F H 10 Q Trinidad et al., F b H Q Oliver et al., F H Q Fox et al., F H 1583 Q Myllys, F 1995 H Q Aarestrup and Jensen, F H 180 Q Oliver et al., F H 82 Q Oliver et al., F a H Q F= Farm, H = heifer, Q = quarter. 2 Bacteriologically negative. 3 Coagulase-negative staphylococci. 4 Environmental pathogens (streptococci non-agalactiae, coliforms). 5 Other pathogens. 6 Coagulase-positive staphylococci. 20

35 Heifer Mastitis: a Review The distribution of pathogens found in the culture positive samples varies between studies, but a common denominator is the high proportion of IMI caused by coagulasenegative staphylococci (CNS) (Tables 1, 2, and 3). Some studies have found that S. aureus was the most prevalent major pathogen (Boddie et al., 1987; Trinidad et al., 1990b; van der Meer et al., 1993), while other studies report environmental bacteria to be the more prevalent major pathogens (Oliver and Mitchell, 1983; Oliver et al., 1992; Oliver et al., 1997; Aarestrup and Jensen, 1997). In general, environmental pathogens are more important around calving and in early lactation (Tables 1, 2, and 3). Coagulase-negative Staphylococci The group of CNS consists of at least 40 different species and subspecies (Devriese et al., 2002). The species associated with bovine mastitis are generally considered to be minor pathogens. Still, they are capable of increasing quarter milk somatic cell count (SCC) slightly (Timms and Schultz, 1987), and are the leading cause of IMI in modern dairy farms where contagious mastitis has been controlled (Smith and Hogan, 2001; Sears and McCarthy, 2003). Table 2. Heifer- and quarter-level prevalence of intramammary infections in heifers at calving. Study Sample level 1 n Neg. 2 CNS 3 S. aureus Env. 4 Other 5 Oliver and Mitchell, F H Q Pankey et al., F H Q Matthews et al., F H 36 Q Roberson et al., F a H Q Fox et al., F H 1583 Q F= Farm, H = heifer, Q = quarter. 2 Bacteriologically negative. 3 Coagulase-negative staphylococci. 4 Environmental pathogens (streptococci non-agalactiae, coliforms). 5 Other pathogens. 6 Coagulase-positive staphylococci. 21

36 Chapter 3 A wide variation in the distribution of CNS species has been reported. Staphylococcus chromogenes and Staphylococcus hyicus (Boddie et al., 1987; Trinidad et al., 1990b), Staphylococcus xylosus and Staphylococcus simulans (Myllys, 1995), and S. chromogenes, S. simulans, and Staphylococcus epidermidis (Aarestrup and Jensen, 1997), respectively, account for most of the CNS IMI in heifers prepartum. At calving, S. chromogenes and S. simulans are the most prevalent species (Matthews et al., 1992), whereas S. xylosus, S. hyicus, and S. simulans were reported to be the most prevalent CNS in early lactation (van der Meer et al., 1993; Myllys, 1995; Aarestrup and Jensen, 1997). Table 3. Heifer- and quarter-level prevalence of intramammary infections in heifers in early lactation. Study Sample n Neg. 2 CNS 3 S. aureus Env. 4 Other 5 level 1 Oliver and Mitchell, F H 32 Q Timms and Schultz, F H Q Matthews et al., F H 36 Q Oliver et al., F H 41 Q van der Meer et al., F H Q Myllys, F 1995 H Q Aarestrup and Jensen, F H 180 Q Oliver et al., F H 42 Q Oliver et al., F a H 40 Q F= Farm, H = heifer, Q = quarter. 2 Bacteriological negative. 3 Coagulase-negative staphylococci. 4 Environmental pathogens (streptococci non-agalactiae, coliforms). 5 Other pathogens. 6 Coagulase-positive staphylococci. 22

37 Heifer Mastitis: a Review Coagulase-negative staphylococci are teat skin opportunists (Sears and McCarthy, 2003). Staphylococcus xylosus, S. chromogenes, and Staphylococcus warneri were the predominant flora isolated from the teat skin from 10 heifers that were sampled repetitively over time (Boddie et al., 1987). Devriese and De Keyser (1980) sampled teat apices from 30 cows and found that 81, 42, 34, and 21% were colonized with S. xylosus, Staphylococcus sciuri, S. warneri, and Staphylococcus haemolyticus, respectively. In another study, the percentage of teat skin samples colonized by S. xylosus, S. warneri, and S. chromogenes was 23, 23, and 22%, respectively (White et al., 1989). The prevalence of positive teat orifices from heifers sampled approximately 2 wk prior to calving was as follows: 28% S. chromogenes, 3.5% S. simulans, 2.8% Staphylococcus hominis, and 2.1% S. epidermidis (Matthews et al., 1992). Some CNS, e.g. S. xylosus, S. cohnii, S. saprophyticus and S. sciuri, seem to be common in the cows' environment (Matos et al., 1991). Staphylococcus aureus Although prevalence of S. aureus IMI in dairy heifers is generally lower compared with CNS (Tables 1, 2 and 3), its importance should not be denied as this bacterium is the most difficult major contagious mastitis pathogen to control (Sol, 2002). Current mastitis control recommendations may reduce prevalence and incidence of S. aureus IMI, but do not readily lead to eradication (Roberson et al., 1994a). Additionally, recently freshened heifers with S. aureus IMI may represent an important infection source at milking for uninfected lactating herd mates (Roberson et al., 1994a). Roberson et al. (1996) reported that 6.5% of 3095 heifers from 22 herds sampled at calving were infected with coagulase-positive staphylococci (CPS): 136, 62, and 1 heifer(s) were culture-positive for S. aureus, S. hyicus, and Staphylococcus intermedius, respectively. In a more recent North American study, 15.4% of quarters from 233 heifers were infected by S. aureus (Owens et al., 2001). Aarestrup and Jensen (1997), on the other hand, reported a very low S. aureus prevalence in Danish dairy heifers during the peripartum period. Only 2.6% of heifers and 0.7% of quarters were infected by S. aureus (Pankey et al., 1991). Environmental Pathogens Environmental pathogens also cause a significant proportion of IMI in prepartum and recently calved heifers. For instance, Oliver and Mitchell (1983) reported that 14 to 7 d prior to calving, at calving, and 7 to 14 days in milk (DIM), 9.6, 12.5, and 4.8% of quarters, respectively, were infected by coliforms and non-agalactiae streptococci. This was confirmed 23

38 Chapter 3 by another study sampling 382 heifers of 11 dairy herds: 45.6% of heifers and 18.5% of quarters were infected, with CNS causing most of the infections (22.8% of heifers and 11.4% of quarters), followed by coliforms and non-agalactiae streptococci (14.9% of heifers and 4.8% of quarters). Clinical Mastitis Although heifers have a lower incidence rate of clinical mastitis compared with older cows, peak incidence in early lactation is higher in heifers (Barkema et al., 1998a). More than 30% of all clinical cases in first lactation occur during the first 14 DIM. Twenty-two and 42% of the sampled heifers in 2 herds, respectively, had IMI at calving (Meaney, 1981). The majority of the infections were clinical and were associated with environmental bacteria (S. dysgalactiae, Escherichia coli, and aesculine-positive streptococci). Another study reported that pathogens were isolated from 77.4% of 349 quarters with clinical mastitis signs before and after calving (Myllys, 1995). Before calving, 18.9% of the 74 quarter milk samples were culture-negative, and 50, 16.2, and 12.2% were culture-positive for CNS, S. aureus, and environmental pathogens (coliforms, S. dysgalactiae, Streptococcus uberis, and Streptococcus spp.), respectively. After calving, no bacteria were isolated in 23.6% of the 275 clinical samples, whereas 33.8, 20.4, and 15.6% were positive for CNS, S. aureus, and environmental pathogens, respectively (Myllys, 1995). A large study was conducted in Norway, identifying bacteria associated with clinical mastitis in heifers occurring prepartum or within 14 d post partum (Waage et al., 1999). The total number of affected quarters was 1361 from 1040 heifers. Staphylococcus aureus (44.3%), S. dysgalactiae (18.2%), and CNS (12.8%) were the most frequently isolated organisms. Staphylococcus simulans was the predominant CNS, followed by S. hyicus and S. chromogenes. No difference in distribution prepartum and post partum was present except for A. pyogenes. Staphylococcus aureus was more often associated with systemic illness in the affected heifers. Persistence of Intramammary Infection Whether IMI at calving are transient or persistent can play a crucial role for future udder health and production. In that respect, a distinction between major and minor pathogens should be made. 24

39 Heifer Mastitis: a Review Prevalence of CNS infected quarters from 32 heifers on 1 farm decreased from 22.2% before calving (14 to 7 d prior to calving), to 18.8% at calving, and 7.5% in early lactation (7 to 14 DIM) (Oliver and Mitchell, 1983). In another study, quarter prevalence of CNS IMI decreased from 39.0% (data from teat orifice swabs) prepartum, to 27.9% at calving, and towards 15.3, 14.6, and 12.2%, 1, 2, and 3 wk post partum, respectively. Thereafter prevalence stabilized at around 14.5% (Matthews et al., 1992). Intramammary infection caused by S. chromogenes decreased shortly after calving to around 1% of quarters, whereas this was 15% before calving (Aarestrup and Jensen, 1997). A large number of other studies also reported that the majority of CNS IMI in heifers is eliminated in early lactation (Miller et al., 1991; Oliver et al., 1992; Owens et al., 1994; Myllys, 1995; Kirk et al., 1996). In conclusion, most CNS infections detected in the prepartum period can be considered transient. Intramammary infections caused by major pathogens, on the other hand, more often persist into early lactation (Oliver and Mitchell, 1983). Nearly 50% of all CPS IMI in heifers at calving appeared to persist for at least 1 mo, presenting a risk of transmission to uninfected herd mates (Roberson et al., 1994a). Persistence of some S. dysgalactiae IMI was confirmed using DNA-typing and the prevalence of S. aureus IMI greatly increased during the first wk after calving, although it was very low prepartum (Aarestrup and Jensen, 1997). Significance of Heifer Mastitis Histological Changes Intramammary infection with CNS in unbred heifers results in an inflammatory response with a likely detrimental effect on the developing parenchyma (Boddie et al., 1987). Mammary tissue of S. aureus infected quarters is less developed and has more leukocyte infiltration compared with tissue from uninfected quarters. The latter was also present in quarters infected by CNS, suggesting a negative effect on future milk production (Trinidad et al., 1990a). On the other hand, while S. aureus infected mammary tissue of non-pregnant heifers revealed a higher leukocytosis than uninfected quarters, indications of reduced secretion potential or glandular development were not present (Owens et al., 1991). Mammary secretions from infected mammary glands had higher SCC than secretions from uninfected mammary glands (Boddie et al., 1987; Trinidad et al., 1990b; Hallberg et al., 1995). Thus, IMI in heifers may impair mammary growth and development compromising future milk production, although results are not fully consistent. 25

40 Chapter 3 Udder Health and Somatic Cell Count Intramammary infection at calving results in an increased SCC, particularly if it is caused by a major pathogen (Barkema et al., 1999). Based on Dairy Herd Improvement data from 30 herds, Coffey et al. (1986) concluded that for heifers the initial rank of SCC classes in early lactation (<100,000, 100,000 to 400,000, and >400,000 cells/ml) was maintained throughout the remainder of the first and subsequent lactations. A study conducted in a Californian dairy with a low prevalence of major mastitis pathogens found that minor pathogen IMI in early lactation in heifers had no effect on SCC in the first 5 milk recordings (Kirk et al., 1996). The probability of clinical mastitis increased with an increasing SCC in early lactation in French Holstein heifers (Rupp and Boichard, 2000). The same pattern was observed in herds with low or high SCC level. In addition, heifers with a higher mean SCC in the first lactation were at a higher risk of clinical mastitis in the second lactation (Rupp et al., 2000). Production Heifers with a first test-day SCC <100,000, between 100,000 and 400,000, and >400,000 cells/ml produced 6452, 6050, and 5696 kg milk during first lactation, respectively (Coffey et al., 1986), indicating a likely effect of IMI at calving on future production. Surprisingly, the heifers with the lowest first test-day SCC produced less in the subsequent lactation compared to the heifers with a higher first test-day SCC. Kirk et al. (1996) reported that IMI with minor pathogens in heifers in early lactation had no effect on average milk production during early to mid lactation. However, in that study only 1 (large) herd was included. The individual milk yield loss in heifers has been estimated at 1.29 kg/day for each unit increase in the log 10 SCC (Koldeweij et al., 1999). In lactating heifers without clinical mastitis, the reduction in milk production was 0.30, 0.61 and 1.09 kg when SCC was 100,000, 200,000, and 600,000 cells/ml, respectively, compared to a reference of 50,000 cells/ml (Hortet et al., 1999). Heifers in organic dairy herds had a production loss of 0.20 kg/day of energy corrected milk with each twofold increase in SCC between 100,000 and 1,500,000 cells/ml (Bennedsgaard et al., 2003). Culling and Fertility Mastitis is known to be an important culling reason in cows (Beaudeau et al., 1995; Barkema et al., 1998a; Bascom and Young, 1998; Gröhn et al., 1998; Seegers et al., 1998; 26

41 Heifer Mastitis: a Review Rajala-Schultz and Gröhn, 1999a, 1999b; Neerhof et al., 2000; Santos, et al., 2004). Nearly 11% of heifers that were treated for clinical mastitis before calving or within the first 14 DIM were culled within 1 mo after treatment (Waage et al., 2000). The main culling reason of 96% of these heifers was mastitis. Cows with test-day somatic cell scores in the highest classes had almost a 3 times higher rate of culling compared with cows with test-day scores on the average level (Samoré et al., 2003). In cows, clinical and subclinical mastitis early post partum had negative effects on reproductive performances (Barker et al., 1998; Schrick et al., 2001). However, in heifers the association between udder health at calving and fertility has, to our knowledge, never been studied in detail. Factors associated with Heifer Mastitis Probably the first study to report a managerial risk factor concluded that S. agalactiae may be transferred to the udders of calves by sucking pen mates fed infected milk (Schalm, 1942). The resulting IMI could persist until freshening. However, little concern remains about detrimental effects of feeding mastitic milk to calves when they are maintained in individual pens (Barto et al., 1982). Limited data indicate that first-lactation cows fed mastitic milk as calves suffered no more udder problems than did their mates formerly given other liquid feed (Kesler, 1981). Group housing of preweaned calves, mastitic milk feeding and a high prevalence of CPS in the lactating herd, do not necessarily translate into a high prevalence of IMI due to CPS in heifers at calving (Roberson et al., 1994a). According to that study, the infected mammary gland is not the reservoir of S. aureus for primiparous cow IMI at first calving, leaving the exact sources and modes of transmission to be determined. Results from a large survey on 28 dairy farms demonstrated that herd, season, trimester of pregnancy, and location (California, Louisiana, Vermont, and Washington) were associated with prevalence of IMI (Fox et al., 1995). The prevalence of IMI was greatest during the last trimester of pregnancy, suggesting that heifers may be most susceptible during that period of gestation, which is most likely associated with the rapid mammary gland development during that time. The same authors reported the largest proportion of heifers to be free of IMI in fall (Fox et al., 1995). Others also reported seasonal variation in post-partum SCC and an influence of location (Hallberg et al., 1995). Poor hygiene of the calving area was associated with an increased prevalence of elevated SCC in heifers (Bareille et al., 2000). Therefore, pregnant and peripartal heifers 27

42 Chapter 3 should be housed in a clean and dry environment to reduce exposure to environmental pathogens (Shearer and Harmon, 1993). The protective effect of supplementing minerals and vitamins on udder health, particularly of selenium and of vitamin E, is well described (Weiss et al., 1990, 1997; Hogan et al., 1993; Barkema et al. 1998b). A recently conducted large study, however, could not find any beneficial effect of treating cows and heifers parenterally before calving with vitamin E on the incidence of, among other diseases, clinical mastitis (LeBlanc et al., 2002). Flies may be important transmission vehicles (Owens et al., 1998; Roberson et al., 1998). Horn flies (Haematobia irritans) are capable of transmitting S. aureus induced IMI to heifers (Owens et al., 1998). Herds using some form of fly control had a lower percentage of heifers with IMI than those without fly control (Nickerson et al., 1995). A recent study indicated that an increased age at calving is a significant risk factor for S. aureus and environmental pathogen IMI (Bassel et al., 2003). The risk of S. aureus IMI was also affected by the amount of time the heifers were housed with older cows, and the proportion of S. aureus infected cows in the herd (Bassel et al., 2003). This suggests that cows' infected udders are an important infection source for heifers, which contrasts with the conclusion put forward in earlier studies (Roberson et al., 1994a, 1994b). Teat disinfection before calving was associated with a decreased risk of calving with an environmental pathogen IMI (Bassel et al., 2003). A beneficial effect of the use of a barrier teat dip before calving against environmental pathogens was not found in a recently conducted study (Edinger et al., 2000). Vaccination of heifers was successful in 1 study (Giraudo et al., 1997) and not successful in another (Tenhagen et al., 2001) in reducing the incidence of S. aureus IMI. Clinical heifer mastitis occurred more frequently in herds with a high mean production (>7000 kg/yr), a low bulk tank milk SCC, high treatment incidence, calculated optimal feeding, and feeding ample dietary cereals (Myllys and Rautala, 1995). Udder edema, teat edema, blood in the milk and milk leakage at time of calving were significant risk factors for clinical mastitis occurring between 1 and 14 DIM, while the last 3 variables also increased the odds on clinical S. aureus mastitis (Waage et al., 2001). An increased risk of clinical mastitis in heifers was associated with an increase in the incidence of clinical mastitis in the herd, a decrease in the bulk milk SCC, and an increase in the mean herd milk yield. Region and calving season were also significant variables (Waage et al., 1998). Although the major reservoir of S. aureus is the infected udder of lactating heifers and cows, this bacterium also colonizes teat skin, vagina, muzzle, and other body sites, but also bedding, feedstuff, air, equipment (Boddie et al., 1987; Matos et al., 1991; Roberson et al., 28

43 Heifer Mastitis: a Review 1994b, Roberson et al., 1998). Scabs on teats are also potential sources of S. aureus (Owens et al., 1998). Heifers with teat skin colonized by S. aureus were 3.3 times more likely to have IMI with the same bacterium at calving than were non-colonized heifers, but the association was not significant (Roberson et al., 1994b). Treatment as a Method for Controlling Heifer Mastitis Intramammary treatment with antibiotics before calving has frequently been evaluated as a practice to reduce the prevalence of IMI in heifers at calving (Trinidad et al., 1990c; Owens et al., 1991; Oliver et al., 1992; Owens et al., 1994; Oliver et al., 1997; Owens et al., 2001; Oliver et al., 2003; Oliver et al., 2004a; Oliver et al., 2004b). Both dry cow and lactating cow products have been evaluated against naturally occurring and experimentally induced IMI. Dry Cow Products The first study using antibiotic treatment before calving reported a reduction in S. aureus and CNS mastitis at calving with a minimal risk of residues by means of a dry cow product containing penicillin-streptomycin (Trinidad et al., 1990c). The trial was conducted on 1 research and 3 commercial dairy herds. Heifers treated during the second trimester of pregnancy showed the best response to treatment as reflected by a decrease in IMI and SCC. The latter trial was followed by a study evaluating a dry cow product containing cephapirin benzathine for the treatment of experimentally induced S. aureus mastitis in nonpregnant and pregnant heifers approximately 9 wk before calving (Owens et al., 1991). The authors reported an effective elimination of the induced S. aureus IMI at calving, although treated quarters from 1 heifer continued yielding naturally occurring S. aureus from secretion and tissue after treatment. The same product was later evaluated by the same research group against naturally occurring IMI (Owens et al., 1994). Mammary secretions from 75 heifers from a research herd were collected 4 to 5 mo prepartum to determine bacteriological status. Heifers with at least 1 IMI were included in the study and assigned to a treatment or control group. Treatment was very effective as over 90% of IMI were cured, and SCC of cured quarters was comparable with SCC of uninfected control quarters. In addition, all quarters from 40 heifers from a commercial herd with a large proportion of S. aureus infected heifers were treated after determination of the IMI status. Cure rates were 94, 84, 100, and 100% for S. aureus, Streptococcus species, CNS, and coliforms, respectively (Owens et al., 1994). 29

44 Chapter 3 Two-hundred and thirty-three heifers from a research farm were treated in the first, second, or third trimester of gestation with 1 of 5 antibiotics (cephapirin, penicillinnovobiocin, penicillin-streptomycin, experimental tilmicosin, and cephalonium, all dry cow products) (Owens et al., 2001). Cure rates of the 5 products indicated that all were equally effective against S. aureus and all were significantly better than the spontaneous cure rate within the untreated control quarters. Efficacy was not influenced by time of treatment relative to expected calving. Lactating Cow Products Jersey heifers from a research herd were assigned to 3 treatments alternately: 1) negative control, 2) treatment with sodium cloxacillin at 7 d prepartum, and 3) treatment with cephapirin sodium at 7 d prepartum. Treatment resulted in elimination of a high number of IMI, especially those caused by CNS (87.1% of all IMI) (Oliver et al., 1992). The possibility of antibiotic residues in milk was warned for - especially in heifers calving too early - as 17 and 85% of colostrum samples were positive for cloxacillin and cephapirin, respectively. Percentage of samples with mastitis pathogens was higher in untreated controls than in treated quarters throughout first lactation, as reported recently (Oliver et al., 2003). This difference was, however, mainly due to CNS rather than due to Streptococcus species or S. aureus, Because antibiotic residues were detected frequently in milk in early lactation (Oliver et al., 1992), the same research group conducted a subsequent study treating heifers at 14 d instead of 7 d prior to the expected calving date (Oliver et al., 1997). Jersey heifers from a research herd were assigned randomly to 2 groups: 1) untreated control, and 2) intramammary infusion with cephapirin sodium. Treatment 14 d instead of 7 d prepartum markedly reduced the occurrence of residues in early lactation without affecting efficacy. For untreated and treated heifers, mastitis pathogens were isolated from 67.3 and 63.8% of samples obtained 14 d prepartum, 55.6 and 15.1 obtained at 3 DIM, and 36.4 and 7.9% obtained at 30 DIM, respectively. Still, it should be noticed that the difference in prevalence of major pathogens isolations was smaller: at 3 DIM, 11.7 versus 7.9%, and at 30 DIM, 6.8 versus 3.9% in the untreated and treated group, respectively. In addition, prevalence of major mastitis pathogens 14 d prior to calving was higher in the control group (18.5%) compared to the treatment group (15.8%) (Oliver et al., 1997). Results of the microbiological evaluations of milk samples collected from these heifers at intervals throughout first lactation were reported afterwards (Oliver et al., 2003). Percent of samples with mastitis pathogens was higher in untreated controls than in treated quarters at every sampling interval during lactation, although this 30

45 Heifer Mastitis: a Review difference was mainly due to CNS. The difference in percentage of S. aureus positive milk samples between treated and untreated quarters in the ensuing lactation could be due to the large difference already present before treatment. Still, the treated heifers out-produced the control heifers by 531 kg milk (Oliver et al., 2003). A recent study reported treatment of heifers with penicillin-novobiocin or pirlimycin hydrochloride 14 d before expected calving on a research herd with Jersey heifers and a research herd with Holstein-Friesian heifers. Both herds had a variable prevalence of IMI prepartum and a different pathogen distribution (Oliver et al., 2004a). Of the quarters infected at 14 d prepartum, 75, 87, and 56% were uninfected following penicillin-novobiocin, pirlimycin, and no treatment, respectively in the Jersey herd, whereas this was 76, 59, and 26% in the Holstein herd. In both herds, the cure rates in the treated quarters were significantly higher than in the untreated control quarters. However, an equivalent number of new IMI during early lactation was observed in treated and control quarters, suggesting little preventive effect (Oliver et al., 2004a). Evaluation of a more "user friendly" approach was conducted to determine efficacy of antibiotic treatment following the first milking (Oliver et al., 2004b). Jersey and Holstein heifers from 2 research herds were assigned to 1) no treatment after first milking, 2) infusion of all mammary quarters with pirlimycin hydrochloride, or 3) infusion of all mammary quarters with novobiocin sodium plus penicillin G. In the Jersey herd, of the quarters infected at calving, 77 and 62% were cured following treatment with pirlimycin or penicillinnovobiocin, and 40% were uninfected in the control quarters. This was 57, 41, and 23% in the Holstein herd. The SCC score from the Jersey heifers during the first 3 mo of lactation was significantly lower in pirlimycin treated heifers compared with the control heifers but not for the penicillin-novobiocin treated heifers. In addition, there was no production difference due to antibiotic treatment in Jersey heifers during the first 90 DIM. The authors concluded that this user friendly approach, avoiding treating heifers at a time they are generally not worked with, was not as effective as prepartum treatment. This was likely due to timing of treatment, persistence of antibiotics, and the time when new IMI occur during the periparturient period (Oliver et al., 2004b). A recently conducted trial performed on 7 experimental herds indicated that in some herds, non-treated control heifers out-produced heifers being infused with antibiotics 2 wk before calving with a lactating cow product containing cephapirin, during the first 200 DIM (Fox et al., 2004). Overall, SCC and milk production differences between treatment groups were small. 31

46 Chapter 3 Conclusions Heifer mastitis is an important disease incorporating a threat to production and udder health in the first and following lactation(s). Coagulase-negative staphylococci have been identified as the predominant cause of IMI in heifers, but S. aureus and environmental pathogens could play a more significant role as these pathogens are more likely to persist into lactation. Intramammary infection leads to histological changes suggesting a negative influence on future production, although quantification has until now never been done. Minor pathogens do not seem to have a large impact, based on the results of a study conducted on 1 herd. Some risk factors associated with heifer mastitis have been identified. Fly control, maintaining a clean and comfortable environment, and proper hygiene in general are probably the most important factors. Avoiding mineral and vitamin deficiencies, not feeding mastitic milk and avoidance of inter-sucking should be considered important management practices as well. Prepartum treatment with antibiotics has been proposed as a simple and effective way of controlling heifer mastitis resulting in a reduced prevalence of infected heifers and udder quarters at calving. Still, the long-lasting effects have not been extensively studied on a large number of commercial herds. Therefore, general recommendation of this practice remains difficult at present. Besides, use of antibiotics in this matter is extra-label. References Aarestrup, F. M., and N. E. Jensen Prevalence and duration of intramammary infection in Danish heifers during the prepartum period. J. Dairy Sci. 80: Bareille, N., H. Seegers, M. B. Kiebre-Toe, F. Beaudeau, and C. Fourichon Risk factors for elevated milk somatic cell counts during early lactation in dairy heifers. Proc. 10th Intern. Congr. Anim. Hyg., Maastricht, The Netherlands: Barkema, H. W., Y. H. Schukken, T.J.G.M. Lam, M. L. Beiboer, H. Wilmink, G. Benedictus, and A. Brand. 1998a. Incidence of clinical mastitis in dairy herds grouped in 3 categories by bulk milk somatic cell counts. J. Dairy Sci. 81: Barkema, H. W., Y. H. Schukken, T.J.G.M. Lam, M. L. Beiboer, G. Benedictus, and A. Brand. 1998b. Management practices associated with low, medium, and high somatic cell counts in bulk milk. J. Dairy Sci. 81:

47 Heifer Mastitis: a Review Barkema, H. W., H. Deluyker, Y. H. Schukken, and T.J.G.M. Lam Quarter-milk somatic cell count at calving and at the first 6 milkings after calving. Prev. Vet. Med. 38:1-9. Barker, A. R., F. N. Schrick, M. J. Lewis, H. H. Dowlen, and S. P. Oliver Influence of clinical mastitis during early lactation on reproductive performance of Jersey cows. J. Dairy Sci. 81: Barto, P. B., L. J. Bush, and G. D. Adams Feeding milk containing Staphylococcus aureus to calves. J. Dairy Sci. 65: Bascom, S. S., and A. J. Young A summary of the reasons why farmers cull cows. J. Dairy Sci. 81: Bassel, L., D. Kelton, A. Godkin, K. Leslie, and K. Lissemore Risk factors for intramammary infection at first calving in Ontario dairy heifers. Proc. 36th Ann. Conv. AABP, Columbus, Ohio, United States of America: Beaudeau, F., V. Ducrocq, C. Fourichon, and H. Seegers Effect of disease on length of productive life of French Holstein dairy cows assessed by survival analysis. J. Dairy Sci. 78: Bennedsgaard, T. W., C. Enevoldsen, S. M. Thamsborg, and M. Vaarst Effect of mastitis treatment and somatic cell counts on milk yield in Danish organic dairy cows. J. Dairy Sci. 86: Boddie, R. L., S. C. Nickerson, W. E. Owens, and J. L. Watts Udder microflora in nonlactating heifers. Agri-Practice 8: Coffey, E.M., W.E. Vinson, and R.E. Pearson Somatic cell counts and infection rates for cows of varying somatic cell count in initial test of first lactation. J. Dairy Sci. 69: Daniel, R. C., D. A. Barnum, and K. E. Leslie Observations on intramammary infections in first calf heifers in early lactation. Can. Vet. J. 27: Devriese, L. A., and H. De Keyser Prevalence of different species of coagulasenegative staphylococci on teats and in milk samples from dairy cows. J. Dairy Res. 47: Devriese, L. A., M. Baele, M. Vaneechoutte, A. Martel, F. Haesebrouck Identification and antimicrobial susceptibility of Staphylococcus chromogenes isolates from intramammary infections of dairy cows. Vet. Microbiol. 87: Edinger, D., B.-A. Tenhagen, P. Kalbe, G. Klünder, B. Baumgärtner, and W. Heuwieser Effect of teat dipping with a germicide barrier teat dip in late gestation on 33

48 Chapter 3 intramammary infection and clinical mastitis during the first 5 d post-partum in primiparous cows. J. Vet. Med. A. 47: Fox, L. K., S. T. Chester, J. W. Hallberg, S. C. Nickerson, J. W. Pankey, and L. D. Weaver Survey of intramammary infections in dairy heifers at breeding age and first parturition. J. Dairy Sci. 78: Fox, L. K., A. A. Borm, K. E. Leslie, J. S. Hogan, S. M. Andrew, S. P. Oliver, Y. H. Schukken, W. E. Owens, and C. Norman Effect of prepartum antibiotic therapy in heifers on milk production and mastitis post partum. Proc. 43rd Ann. NMC meeting, Charlotte, North Carolina, United States of America: Giraudo, J. A., A. Calzolari, H. Rampone, A. Rampone, A. T. Giraudo, C. Bogni, A. Larriestra, and R. Nagel Field trials of a vaccine against bovine mastitis. 1. Evaluation in heifers. J. Dairy Sci. 80: Gröhn, Y. T., S. W. Eicker, V. Ducrocq, and J. A. Hertl Effect of diseases on the culling of Holstein dairy cows in New York State. J. Dairy Sc. 81: Hallberg, J. W., K. J. Dame, S. T. Chester, and C. C. Miller The visual appearance and somatic cell count of mammary secretions collected from primigravid heifers during gestation and early postpartum. J. Dairy Sci. 78: Hogan, J. S., W. P. Weiss, and K. L. Smith Role of vitamin E and selenium in host defence against mastitis. J. Dairy Sci. 76: Hortet, P., F. Beaudeau, H. Seegers, and C. Fourichon Reduction in milk yield associated with somatic cell counts up to 600,000 cells/ml in French Holstein cows without clinical mastitis. Prev. Vet. Med. 61: Kesler, E. M Feeding mastitic milk to calves: review. J. Dairy Sci. 64: Kirk, J. H., J. C. Wright, S. L. Berry, J. P. Reynolds, J. P. Maas, and A. Ahmadi Relationship of milk culture status at calving with somatic cell counts and milk production of dairy heifers during early lactation on a Californian dairy. Prev. Vet. Med. 28: Koldeweij, E., U. Emanuelson, and L. Janson Relation of milk production loss to milk somatic cell count. Acta Vet. Scand. 40: LeBlanc, S. J., T. F. Duffield, K. E. Leslie, K. G. Bateman, J. TenHag, J. S. Walton, and W. H. Johnson The effect of prepartum injection of vitamin E on health in transition dairy cows. J. Dairy Sci. 85:

49 Heifer Mastitis: a Review Matos, J. S., D. G. White, R. J. Harmon, and B. E. Langlois Isolation of Staphylococcus aureus from sites other than the lactating mammary gland. J. Dairy Sci. 74: Matthews, K. R., R. J. Harmon, and B. E. Langlois Prevalence of Staphylococcus species during the periparturient period in primiparous and multiparous cows. J. Dairy Sci. 75: Meaney, W. J Mastitis levels in spring-calving dairy heifers. Ir. Vet. J. 35: Miller, R. H., M. J. Paape, and L. A. Fulton Variation in milk somatic cells of heifers at first calving. J. Dairy Sci. 74: Munch-Petersen, E Mastitis in bovine primiparae. Vet. Rec. 87: Myllys, V Staphylococci in heifer mastitis before and after parturition. J. Dairy Res. 62: Myllys, V., and H. Rautala Characterization of clinical mastitis in primiparous heifers. J. Dairy Sci. 78: Neerhof, H. J., P. Madsen, V. P. Ducrocq, A. R. Vollema, J. Jensen, and I. R. Korsgaard Relationship between mastitis and functional longevity in Danish Black and White dairy cattle estimated using survival analysis. J. Dairy Sci. 83: Nickerson, S. C., W. E. Owens, and R. L. Boddie Mastitis in dairy heifers: initial studies on prevalence and control. J. Dairy Sci. 78: Oliver, S. P., and B. A. Mitchell Intramammary infections in primigravid heifers near parturition. J. Dairy Sci. 66: Oliver, S. P., M. J. Lewis, B. E. Gillespie, and H. H. Dowlen Influence of prepartum antibiotic therapy on intramammary infections in primigravid heifers during early lactation. J. Dairy Sci. 75: Oliver, S. P., M. J. Lewis, B. E. Gillespie, and H. H. Dowlen Antibiotic residues and prevalence of mastitis pathogen isolation in heifers during early lactation following prepartum antibiotic therapy. J. Vet. Med. B. 44: Oliver, S. P., M. J. Lewis, B. E. Gillespie, H. H. Dowlen, E. C. Jaenicke, and R. K. Roberts Prepartum antibiotic treatment of heifers: milk production, milk quality and economic benefit. J. Dairy Sci. 86: Oliver, S. P., S. J. Ivey, B. E. Gillespie, M. J. Lewis, D. L. Johnson, K. C. Lamar, H. Moorehead, H. H. Dowlen, S. T. Chester, and J. W. Hallberg. 2004a. Influence of prepartum intramammary infusion of pirlimycin hydrochloride or penicillin-novobiocin on mastitis in heifers during early lactation. J. Dairy Sci. 84:

50 Chapter 3 Oliver, S. P., S. J. Ivey, B. E. Gillespie, M. J. Lewis, D. L. Johnson, K. C. Lamar, H. Moorehead, H. H. Dowlen, S. T. Chester, and J. W. Hallberg. 2004b. Heifer mastitis: prevalence, risk factors and control strategies. Proc. 43rd Ann. NMC meeting, Charlotte, North Carolina, United States of America: Owens, W. E., S. C. Nickerson, P. J. Washburn, and C. H. Ray Efficacy of a cephapirin dry cow product for treatment of experimentally induced Staphylococcus aureus mastitis in heifers. J. Dairy Sci. 74: Owens, W. E., S. C. Nickerson, P. J. Washburn, and C. H. Ray Prepartum antibiotic therapy with a cephapirin dry-cow product against naturally occurring intramammary infections in heifers. J. Vet. Med. B. 41: Owens, W. E., S. P. Oliver, B. E. Gillespie, C. H. Ray, and S. C. Nickerson Role of horn flies (Haematobia irritans) in Staphylococcus aureus induced mastitis in dairy heifers. Am. J. Vet. Res. 59: Owens, W. E., S. C. Nickerson, R. L. Boddie, G. M. Tomita, and C. H. Ray Prevalence of mastitis in dairy heifers and effectiveness of antibiotic therapy. J. Dairy Sci. 84: Pankey, J. W., P. A. Drechsler, and E. E. Wildman Mastitis prevalence in primigravid heifers at parturition. J. Dairy Sci. 74: Rajala-Schultz, P. J., and Y. T. Gröhn. 1999a. Culling of dairy cows. Part I. Effects of diseases on culling in Finnish Ayrshire cows. Prev. Vet. Med. 41: Rajala-Schultz, P. J., and Y. T. Gröhn. 1999b. Culling of dairy cows. Part III. Effects of diseases, pregnancy status and milk yield on culling in Finnish Ayrshire cows. Prev. Vet. Med. 41: Roberson, J. R., L. K. Fox, D. D. Hancock, J. M. Gay, and T. E. Besser. 1994a. Coagulasepositive Staphylococcus intramammary infections in primiparous dairy cows. J. Dairy Sci. 77: Roberson, J. R., L. K. Fox, D. D. Hancock, J. M. Gay, and T. E. Besser. 1994b. Ecology of Staphylococcus aureus isolated from various sites on dairy farms. J. Dairy Sci. 77: Roberson, J. R., L. K. Fox, D. D. Hancock, J. M. Gay, and T. E. Besser Prevalence of coagulase-positive staphylococci, other than Staphylococcus aureus, in bovine mastitis. J. Am. Vet. Res. 57:

51 Heifer Mastitis: a Review Roberson, J. R., L. K. Fox, D. D. Hancock, J. M. Gay, and T. E. Besser Sources of intramammary infections from Staphylococcus aureus in dairy heifers at first parturition. J. Dairy Sci. 81: Rupp, R., F. Beaudeau, and D. Boichard Relationship between milk somatic cell counts in the first lactation and clinical mastitis occurrence in the second lactation of French Holstein cows. Prev. Vet. Med. 46: Rupp, R., and D. Boichard Relationship of early first lactation somatic cell count with risk of subsequent first clinical mastitis. Livest. Prod. Sci. 62: Samoré, A. B., M. del P. Schneider, F. Canavesi, A. Bagnato, and A. F. Groen Relationship between somatic cell count and functional longevity assessed using survival analysis in Italian Holstein-Friesian cows. Livest. Prod. Sci. 2003: Santos, J.E.P., R.L.A. Cerri, M. A. Ballou, G. E. Higginbotham, and J. H. Kirk Effect of timing of first clinical mastitis occurrence on lactational and reproductive performance of Holstein dairy cows. Anim. Reprod. Sci. 80: Schalm, O. W Streptococcus agalactiae in the udder of heifers at parturition traced to sucking among calves. Cornell Vet. 32: Schrick, F. N., M. E. Hockett, A. M. Saxton, M. J. Lewis, H. H. Dowlen, and S. P. Oliver Influence of subclinical mastitis during early lactation on reproductive parameters. J. Dairy Sci. 84: Sears, P. M., and K. K. McCarthy Management and treatment of staphylococcal mastitis. Vet. Clin. Food Anim. 19: Seegers, H., F. Beaudeau, C. Fourichon, and N. Bareille Reasons for culling in French Holstein cows. Prev. Vet. Med. 36: Shearer, J. K., and R. J. Harmon Mastitis in heifers. Vet. Clin. Food Anim. 9: Smith, K. L., and J. S. Hogan The world of mastitis. Proc. 2nd Intern. Symp. Mastitis and Milk Quality. Vancouver, British Columbia, Canada:1-12. Sobiraj, A., H.-U Ostertag, D. Peip, H. Bostedt, and G. Kielwein Klinische und bacteriologische Untersuchungsbefunde zur Mastitishäufigkeit erstmalig laktierender Rinder intra und post partum. Tierärztl. Prax. 16: Sol, J Cure of Staphylococcus aureus mastitis in Dutch dairy cows. PhD thesis, Utrecht, The Netherlands. Tenhagen, B. A., D. Edinger, B. Baumgartner, P. Kalbe, G. Klunder, and W. Heuwieser Efficacy of a herd-specific vaccine against Staphylococcus aureus to prevent postpartum mastitis in dairy heifers. J. Vet. Med. A 48:

52 Chapter 3 Timms, L. L., and L. H. Schultz Dynamics and significance of coagulase-negative staphylococcal intramammary infections. J. Dairy Sci. 70: Trinidad, P., S. C. Nickerson, and R. W. Adkinson. 1990a. Histopathology of staphylococcal mastitis in unbred dairy heifers. J. Dairy Sci. 73: Trinidad, P., S. C. Nickerson, and T. K. Alley. 1990b. Prevalence of intramammary infection and teat canal colonization in unbred and primigravid dairy heifers. J. Dairy Sci. 73: Trinidad, P., S. C. Nickerson, T. K. Alley, and R. W. Adkinson. 1990c. Efficacy of intramammary treatment in unbred and primigravid dairy heifers. J. Am. Vet. Med. Assoc. 197: van der Meer, C., J. Tigchelaar, C. Zalsman, P. Meyers, and D. J. Peterse Uierinfecties met coagulase-negatieve stafylokokken (CNS) bij nieuwmelkte vaarzen. Tijdschr. Diergeneeskd. 118: Waage, S., S. Sviland, and S. A. Ødegaard Identification of risk factors for clinical mastitis in dairy heifers. J. Dairy Sci. 81: Waage, S., T. Mørk, A. Røros, D. Aasland, A. Hunshamar, and S. A. Ødegaard Bacteria associated with clinical mastitis in dairy heifers. J. Dairy Sci. 82: Waage, S., H. R. Skei, J. Rise, T. Rogdo, S. Sviland, and S. A. Ødegaard, Outcome of clinical mastitis in dairy heifers assessed by re-examination of cases 1 mo after treatment. J. Dairy Sci. 83: Waage, S., S. A. Ødegaard, A. Lund, S. Brattgjerd, and T. Rothe Case-control study of risk factors for clinical mastitis in postpartum dairy heifers. J. Dairy Sci. 84: Weiss, W. P., J. S. Hogan, K. L. Smith, and K. H. Hoblet Relationships among selenium, vitamin E, and mammary gland health in commercial dairy herds. J. Dairy Sci. 73: Weiss, W. P., J. S. Hogan, D. A. Todhunter, and K. L. Smith Effect of vitamin E supplementation in diets with a low concentration of selenium on mammary gland health of dairy cows. J. Dairy Sci. 80: White, D. G., R. J. Harmon, J. E. Matos, and B. E. Langlois Isolation and identification of coagulase-negative Staphylococcus species from bovine body sites and streak canals of nulliparous heifers. J. Dairy Sci. 72:

53

54

55 Chapter 4 Somatic Cell Counts in Dairy Heifers during Early Lactation S. De Vliegher, 1 H. Laevens, 2 G. Opsomer, 1 E. De Mûelenaere, 3 and A. de Kruif 1 1 Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 2 Coordination Centre for Veterinary Diagnostics, Veterinary and Agrochemical Research Centre, Brussels, Belgium 3 Flemish Cattle Breeding Association, Oosterzele, Belgium Flemish Veterinary Journal, 2001, 70:

56 Chapter 4 42

57 Somatic Cell Counts in Dairy Heifers during Early Lactation Abstract This paper describes the distribution of the first milk somatic cell counts (SCC), measured between 5 and 14 d of lactation (SCCel), during a 1 year period (1999) of 12,994 dairy heifers on 3221 herds in Flanders (Belgium). Somatic cell counts 150,000 cells/ml were allocated to class 1. Somatic cell counts between 151,000 and 300,000 cells/ml, between 301,000 and 1,000,000 cells/ml and >1,000,000 cells/ml were allocated to classes 2, 3, and 4, respectively. The prevalence per class was 65.4, 15.6, 12.9 and 6.1% for the classes 1, 2, 3, and 4, respectively. The within-herd prevalence per SCCel-class was calculated for 137 herds where at least 10 SCCel were available. The within-herd prevalence for class 1 varied from 27.3 to 100% with a median of 66.7%. For the classes 2, 3, and 4, the median within-herd prevalence was 15.4 (range: 0.0 to 50.0%), 10.0 (0.0 to 54.5%), and 0.0% (0.0 to 30.0%), respectively. Key Words Dairy cow, early lactation, heifer, somatic cell count Abbreviation Key DIM = days in milk, IMI = intramammary infection, LnSCCel = natural log-transformed SCCel, SCC = somatic cell count, SCCel = somatic cell count in early lactation (between 5 and 14 DIM). 43

58 Chapter 4 Introduction Mastitis is the most common and expensive disease in dairy cattle (Trinidad et al., 1989). Economic losses are caused by reduced milk production, discarded milk, cost of veterinary services and drugs, and culling of high yielding dairy cows before they have ever reached their top production. Current mastitis control practices are focusing on lactating cows and include prompt treatment of clinical cases, proper milking techniques, use of functionally adequate milking machines, teat dipping after milking, dry cow treatment and culling of chronically infected animals. Little attention has been paid to intramammary infections (IMI) in primigravid dairy heifers and the inspection of their udders is usually restricted to palpation shortly before or even after freshening. As their non-lactating udders have traditionally been regarded as uninfected (Trinidad et al., 1989) it took a long time before it was realised that IMI in dairy heifers were present in far greater numbers than previously accepted (Munch-Petersen, 1970; Oliver and Mitchell, 1983; Daniel et al., 1986; Boddie et al., 1987; Sobiraj et al., 1988; Trinidad et al., 1989; Trinidad et al., 1990; Pankey et al., 1991; Fox et al., 1995; Nickerson et al., 1995). The importance of udder health in heifers upon entering the milking herd can however not be overstated because heifers have an impact on future milk yield and quality in the herd. Milk somatic cell counting is an important instrument for monitoring udder health in lactating cows (Trinidad et al., 1990) and is used worldwide as an indicator of subclinical mastitis (Laevens et al., 1997). However, while SCC is available as a useful tool for the farmer to detect and control udder health on his farm, it is at the same time used for penalization when bulk milk SCC exceeds the threshold level of 400,000 cells/ml (EEC directive). The aim of this study was to describe the distribution of the first lactational SCCel from dairy heifers in early lactation during a 1-year period (1999) by allocating them to 4 classes. As the major factor affecting SCC is IMI (Harmon, 1994; Laevens et al., 1998), this distribution reflects the IMI status of the dairy heifers in Flanders at the start of their first lactation. 44

59 Somatic Cell Counts in Dairy Heifers during Early Lactation Materials and Methods Data Collection The individual SCC of all lactating heifers and multiparous cows from the year 1999 from all herds participating in the Dairy Herd Improvement program in Flanders (Belgium) were made available as 1 dataset by the Flemish Cattle Breeding Association. This dataset contained the following information: herd identification - cow identification - breed code - parity - days in milk (DIM) - cumulative milk production - date of SCC measurement - SCC: measured at monthly intervals of every cow that was at least 5 DIM. Composite milk samples were therefore collected from 2 successive milkings and were analysed with the Fossomatic 5000 (Foss Electric, Hillerød, Denmark). Data Processing Only SCC from heifers measured between day 5 and day 14 post partum were selected. A classification was performed as follows: Class 1: SCCel 150,000 cells/ml Class 2: SCCel between 151,000 and 300,000 cells/ml Class 3: SCCel between 301,000 and 1,000,000 cells/ml Class 4: SCCel >1,000,000 cells/ml. The time-dependent variation of SCCel during the observed time interval was assessed by fitting a regression line through the observed log-transformed SCCel (LnSCCel) using SPSS 9.0 for Windows (SPSS Inc., Chicago, Illinois, USA). The within-herd prevalence per class was calculated when at least 10 SCCel were available per herd. The within-herd prevalence is the percentage of heifers per class in a herd. Results The SCCel of 12,994 heifers from 3321 herds were determined and from 137 herds more than 9 SCCel were available. The overall distribution of SCCel of classes 1, 2, 3, and 4 was 65.4, 15.6, 12.9 and 6.1%, respectively (Table 1). Most of the SCCel were measured during the months July to December (n = 8349, 65%), as the majority of the heifers calved during the second half of the year (Table 1). 45

60 Chapter 4 Table 1. Overall and seasonal prevalence per SCCel-class. Prevalence Season n (%) Class 1 Class 2 Class 3 Class 4 January-February-March 3177 (24) 62.5% 17.2% 14.1% 6.2% April-May-June 1468 (11) 55.9% 19.6% 16.6% 7.8% July-August-September 3935 (30) 64.0% 17.1% 13.3% 5.5% October-November-December 4414 (35) 71.7% 11.8% 10.5% 5.9% Overall 12,994 (100) 65.4% 15.6% 12.9% 6.1% A seasonal SCCel variation was observed as shown in Table 1. The prevalence of class 1 SCCel was lowest in April, May and June (55.9%), whereas it was highest in October, November and December (71.7%). The prevalence of classes 2, 3 and 4 was just the opposite. They were highest in April, May and June (19.6, 16.6, and 7.8%, respectively) and lowest in October, November and December, except for class 4 that had the lowest prevalence in July, August, September (11.8, 10.5 and 5.5%). A time dependent variation was observed. Somatic cell counts in early lactation decreased with increasing DIM (LnSCCel = x DIM). The within-herd prevalence for each class is shown in Figure 1. For the SCCel of class 1 the within-herd prevalence varied from 27.3 to 100% with a median of 66.7%. For the SCCel of classes 2, 3, and 4 the median within-herd prevalence was 15.4 (range: 0.0 to 50.0%), 10.0 (0.0 to 54.5%), and 0.0 (0.0 to 30.0%), respectively. Discussion In this study, nearly 35% of all heifers had a SCCel, measured between day 5 and day 14 post partum, >150,000 cells/ml, and 19% had a first lactational SCCel >300,000 cells/ml. This is high, considering the fact that a heifer is expected to have a SCC of 100,000 cells/ml or lower (O'Rourke and Blowey, 1992) and considering the results from Laevens et al. (1997) who found that 95% of the heifers which were bacteriologically negative during their first lactation, had a SCCel 150,000 cells/ml during the first mo of lactation. The results in this study are to be interpreted with caution as data on bacteriological culture were not available. This means that SCCel were only used as an indicator of udder health problems in heifers. Somatic cell counts are physiologically high during the first wk of lactation according to O'Rourke and Blowey (1992). Dohoo et al. (1993) recommended not to 46

61 Somatic Cell Counts in Dairy Heifers during Early Lactation Herds (n) Class Herds (n) Class Herds (n) Class Herds (n) Class Figure 1. Within-herd prevalence per SCCel-class. 47

62 Chapter 4 consider all individual SCC measured during the first 9 d in order to avoid upwards bias. Barkema et al. (1999), however, stated that quarter-milk SCC was applicable as of day 2 post partum to determine IMI in an udder quarter. They state that high SCC (>250,000 cells/ml) early post partum can hardly ever be considered physiologic. The seasonal variation of SCCel might be explained by the fact that the heifers have been kept indoors during winter, creating a higher risk for IMI. This was also observed by Fox et al. (1995). In contrast, Klaas et al. (1998) have found the lowest prevalence of IMI in heifers between August and September. Besides the seasonal variation, there was also a large between-herd variation. This may be indicative that heifer management plays an important role in the prevention of IMI. Prevention of prepartum IMI is currently based on controlling flies during the summer period, using individual calf hutches to avoid suckling among calves, segregating pregnant heifers from dry cows (Trinidad et al., 1989) and applying prepartum treatment (Shearer and Harmon, 1993). However, further investigations have to be done to determine risk factors associated with increased SCCel and IMI in primigravid dairy heifers. References Barkema, H. W., H. Deluyker, Y. H. Schukken, and T. J. Lam Quarter-milk somatic cell count at calving and at the first 6 milkings after calving. Prev. Vet. Med. 38:1-9. Boddie, R. L., S. C. Nickerson, W. E. Owens, and J. L. Watts Udder microflora in nonlactating heifers. Agri-Pract. 8: Daniel, R. C., D. A. Barnum, and K. E. Leslie Observations on intramammary infections in first calf heifers in early lactation. Can. Vet. J. 27: Dohoo, I. R An evaluation of the validity of individual cow somatic cell counts from cows in early lactation. Prev. Vet. Med. 16: Fox, L. K., S. T. Chester, J. W. Hallberg, S. C. Nickerson, J. W. Pankey, and L. D. Weaver Survey of intramammary infections in dairy heifers at breeding age and first parturition. J. Dairy Sci. 78: Harmon, R. J Physiology of mastitis and factors affecting somatic cell counts. J. Dairy Sci. 77: Klaas, I., U. Wessels, H. Rothfuss, and O. W. Schalm Prevalence of mastitis in dairy heifers. Proceedings of the 10th International Conference on Production Diseases in Farm Animals, Utrecht:

63 Somatic Cell Counts in Dairy Heifers during Early Lactation Laevens, H., H. Deluyker, Y. H. Schukken, L. De Meulemeester, R. Vandermeersch, E. De Mûelenaere, and A. de Kruif Influence of parity and stage of lactation on the somatic cell count in bacteriologically negative cows. J. Dairy Sci. 80: Laevens, H., H. Deluyker, and A. de Kruif Somatic cell count measurements: a diagnostic tool to detect mastitis? Proc. 10th Intern. Conf. Prod. Dis. Farm Anim. Utrecht, The Netherlands: Munch-Petersen, E Mastitis in bovine primiparae. Vet. Rec. 87: Nickerson, S. C., W. E. Owens, and R. L. Boddie Mastitis in dairy heifers: initial studies on prevalence and control. J. Dairy Sci. 78: O'Rourke, D. J., and R. W. Blowey Cell Counts and Mastitis Monitoring. In Bovine Medicine Disease and Husbandry of Cattle. Eds A. H. Andrews, R. W. Blowey, H. Boyd, R. G. Eddy. London, Blackwell Scientific Publications, Oliver, S. P., and B. A. Mitchell Intramammary infections in primigravid heifers near parturition. J. Dairy Sci. 66: Pankey, J. W., P. A. Drechsler, and E. E. Wildman Mastitis prevalence in primigravid heifers at parturition. J. Dairy Sci. 74: Shearer, J. K., and R. J. Harmon Mastitis in heifers. Vet. Clin. Food Anim. 9: Sobiraj, A., H. U. Ostertag, D. Peip, H. Bostedt, and G. Kielwein [Clinical and bacteriologic studies of the frequency of mastitis during and after parturition in heifers lactating for the first time]. Tierärztl. Prax. 16: Trinidad, P., S. C. Nickerson, T. K. Alley, and R. W. Adkinson Mastitis in heifers. Louisiana Agriculture 32:4-5. Trinidad, P., S. C. Nickerson, and T. K. Alley Prevalence of intramammary infection and teat canal colonization in unbred and primigravid dairy heifers. J. Dairy Sci. 73:

64

65 Chapter 5 Significance of Early Lactation Somatic Cell Count in Dairy Heifers for the Subsequent First Lactation

66

67 Chapter 5.1 Impact of Early Lactation Somatic Cell Count in Heifers on Somatic Cell Counts over the First Lactation S. De Vliegher, 1 H. W. Barkema, 2 H. Stryhn, 2 G. Opsomer, 1 and A. de Kruif 1 1 Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 2 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Charlottetown, PEI, Canada Journal of Dairy Science, 2004, 87:

68 Chapter

69 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation Abstract The objective of this study was to estimate the impact of somatic cell count in early lactation (SCCel) from Belgian dairy heifers on test-day somatic cell count (SCC) in the first lactation. Geometric mean SCCel [5 to 14 days in milk (DIM)] of the 14,766 available samples was 104,000 cells/ml, and decreased from 178,000 at 5 DIM to 74,000 cells/ml at 14 DIM. Proportion of SCCel >200,000 cells/ml was Heifers calving in the period April-June had highest SCCel. In total, 117,496 monthly SCCs were measured. A multilevel regression analysis revealed that an increase of the natural log-transformed SCCel (LnSCCel) by 1 unit on average resulted in an increase of test-day natural log-transformed SCC (LnSCC) by 0.22 unit. The impact of LnSCCel on LnSCC depended on when LnSCCel was measured: an elevated LnSCCel at 14 DIM was more consequential than an equally elevated LnSCCel at 5 DIM. The probability of having a test-day SCC >200,000 cells/ml during the first lactation, also increased with an increasing LnSCCel. The negative effect of an elevated LnSCCel was still present, although to a lesser extent, in heifers with a second test-day SCC 50,000 cells/ml. This study indicates that udder health problems in heifers in early lactation have a high prevalence and stresses that heifers should have a low SCCel, as an elevated SCCel will negatively influence test-day SCC during the whole first lactation. Key Words Dairy heifer, early lactation, somatic cell count, udder health Abbreviation Key CNS = coagulase-negative staphylococci, DIM = days in milk, DHI = Dairy Herd Improvement, IMI = intramammary infection, LnSCC = natural log-transformed SCC, LnSCCel = natural log-transformed SCCel, MY = milk yield at test-day (kg), OR = odds ratio, SCC = somatic cell count, SCCel = SCC in early lactation (between 5 and 14 DIM). 55

70 Chapter 5.1 Introduction A mastitis prevention program (Neave et al., 1969) has value in reducing the prevalence and incidence of subclinical mastitis. This type of program has primarily focused on lactating and dry cows and overlooks young, primigravid and recently calved heifers. This should change, as many studies have reported a high prevalence of intramammary infection (IMI) in heifers around calving. Nearly all studies indicate that coagulase-negative staphylococci (CNS) are responsible for the majority of the IMIs in non-lactating and freshly calved heifers. Staphylococcus aureus and environmental pathogens also play an important role (Pankey et al., 1991; Oliver et al., 1992; Roberson et al, 1994; Myllys, 1995; Oliver et al., 2003). Reported prevalence differs between studies, but 1 study reported that as many as 45% of all sampled quarters were infected with CNS (Oliver et al., 2003), whereas prevalence of quarters infected with S. aureus and environmental pathogens ranged between 0.6 and 4.7% (Oliver et al., 1992; Myllys, 1995) and between 4.6 and 8% (Myllys, 1995; Oliver et al., 2003), respectively. Intramammary infection at calving results in an increased SCC, particularly if it is caused by a major pathogen (Barkema et al., 1999). The probability of clinical mastitis increases with an increasing SCCel (Rupp and Boichard, 2000). An elevated SCCel could also result in a permanently elevated SCC and an increased risk of subclinical mastitis during the subsequent first lactation. A higher mean SCC in the first lactation is associated with a higher risk of clinical mastitis in the second lactation (Rupp et al., 2000), stressing the importance of having low SCC from the start until the end of the first lactation. Coffey et al. (1986), using DHI test-day records from heifers from 30 herds, concluded that the initial rank of SCC classes in early lactation (<100,000, 100,000 to 400,000, and >400,000 cells/ml) was maintained throughout the remainder of the first and subsequent lactations. Probably because of a relatively small sample size, initial test-day SCC were categorized in 3 classes, and day of measurement of the first test-day SCC was not part of the evaluation. Because SCC in the first couple of wk after calving rapidly decreases (Dohoo, 1993; Laevens et al., 1997; Barkema et al., 1999), this may have resulted in a biased classification of heifers (Dohoo, 1993; Barkema et al., 1999). Additionally, Coffey et al. (1986) studied the effect of first test-day SCC class on the lactational average SCC, ignoring variation in SCC during lactation. Using the information of all test-day measurements instead of a lactational average makes studying a changing effect of SCCel on SCC throughout lactation possible. 56

71 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation Since the publication of this study (Coffey et al., 1986) statistical software and computer power have progressed considerably, making it possible to account for clustering of heifers within herds (McDermott and Schukken, 1994). Ignoring clustering for continuous outcome variables can result in unbiased estimates of the regression coefficients, but standard errors and associated P values might be strongly affected. Ignoring clustering for discrete data will, in addition, lead to biased estimates, especially with limited samples sizes (Dohoo et al., 2003). Multilevel modelling takes into account clustering of animals within an environment and can be used to identify the level where the greatest variation resides, as interventions at that level would seem to have the greatest chance of success (McDermott and Schukken, 1994; Dohoo et al., 2001). The objective of the present study was to estimate the impact of SCCel on test-day SCC in the ongoing first lactation using multilevel regression analysis. Materials and Methods Initial Dataset and Data Handling Four-weekly milk recordings from 2000 and 2001 of all lactating cows and heifers were used from the herds enrolled in the Belgian DHI program (Flemish Cattle Breeding Association, Oosterzele, Belgium). No data were recorded prior to 5 DIM. The database included: SCC, milk yield on test-day (MY) (kg of milk), breed, DIM, and date of measurement. The latter was categorized into 4 "calving seasons": January March, April June, July September, and October December. Somatic cell count was measured in composite milk samples collected from 2 successive milkings and was analyzed using the Fossomatic 5000 (Foss Electric, Hillerød, Denmark). All dairy heifers of which the first test-day SCC was measured between 5 and 14 DIM (SCCel) in the year 2000, were selected (n = 14,766). Monthly-measured test-day SCCs of the 14,766 heifers during the first lactation (until 365 DIM) were extracted from the database using Microsoft Access (Microsoft Corporation, Mountain View, CA). In total, 117,496 additional test-day SCC (full dataset) from the first lactation (measured after 14 DIM) were available from 14,234 (96.4%) of the 14,766 heifers. These animals belonged to 3264 herds (on average 4.4 heifers per herd) (Table 1). 57

72 Chapter 5.1 A subset of data was created by selecting heifers with a second test-day SCC (measured after 14 and before 75 DIM) 50,000 cells/ml. This resulted in data from 7807 heifers (55% of the initial 14,234 heifers) in 2827 dairy herds, with 65,458 measurements in total (on average 8.4 per heifer, not including SCCel) (Table 1). Table 1. Structure of the data: test-records in early lactation and during the course of the first lactation. Dataset 1 Level n Average number per unit at next-higher level Range Early Herd Heifer 14, to 58 4 Test 14, to 1 5 Full Herd Heifer 14, to 56 4 Test 117, to 2 6 Sub Herd Heifer to 22 4 Test 65, to Early = Test-day records in early lactation (between 5 and 14 DIM) from 14,766 dairy heifers; Full = Test-day records between 15 and 365 DIM from 14,234 heifers (full dataset); Sub = Test-day records between 15 and 365 DIM from 7807 heifers with a second test-day SCC 50,000 cells/ml milk (data subset). 2 Average number of heifers per herd. 3 Average number of tests per heifer. 4 Range of heifers per herd. 5 Range of tests per heifer. 6 Range of tests per heifer within the 30-d DIM intervals. Statistical Analysis To approximate the normal distribution, a natural logarithmic transformation of SCC (LnSCC) and SCCel (LnSCCel) was performed. For presentation purposes, LnSCC and LnSCCel estimates were converted to SCC and SCCel after analyses. First, LnSCCel was analyzed by multilevel linear regression with herd random effects and fixed effects of DIM (10 levels, 5 to 14 DIM), calving season and breed (4 levels both, as listed in Table 2), all entered as categorical variables. Second, the effect of LnSCCel on LnSCC was studied. For analysis of the repeated measures of LnSCC of each heifer, both the full and sub dataset were split into 12 subsets referred to hereafter as "30-d DIM intervals" (15 to 45, 46 to 75, 76 to 105 DIM), reflecting the approximately four-weekly milk recordings. No analysis was performed for the first 2 30-d DIM intervals from the subset because selection of animals was based on the 58

73 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation measurements within those 2 intervals. To account for multiple testing of the same hypotheses in the 30-d DIM intervals, all P values were multiplied by the number of intervals analyzed (Bonferroni correction, Dohoo et al., 2003). This approach to repeated measures data does not lead to bias in the estimates, but implies some loss of power and renounces any information about the correlation structure over time for each individual. The approach was favoured because none of the above issues was considered serious for the study's primary purpose of describing the effect of LnSCCel on LnSCC over time, and because potential misspecification of complex repeated measures models was considered a greater concern for the study's validity. Multilevel linear and logistic regression models, 1 model per 30-d DIM interval, were used to analyze LnSCC and the binary outcome variable SCC200 (0: SCC 200,000 cells/ml milk; 1: >200,000 cells/ml), respectively. These models were analyzed by restricted maximum likelihood and penalized quasi-likelihood estimation algorithms, respectively, as implemented in the MLwiN software (Rasbash et al., 2000). The analyses used a second-order algorithm for the full dataset and a first order algorithm for the subset, and constrained in both analyses the variation at the lowest level to be binomial because no substantial extra-binomial variation was detected in the data. The models included random effects for herds and heifers, the latter to account for a small number of heifers with 2 test recordings within the same 30-d interval [except for the first (15-45 DIM) and twelfth ( DIM) intervals]. The full model included regression terms for LnSCCel (predictor of main interest), DIM (between 5 and 14, the day of assessment of LnSCCel) and MY, and the categorical variables, test-season and breed (4 levels both; Table 2). Moreover, the model included interaction terms between LnSCCel and the predictors DIM and breed, as well as a quadratic term for LnSCCel and its interaction with DIM. The continuous predictors were centred by subtracting their overall mean, providing better interpretations and better numerical stability of estimates (Dohoo et al., 2001). The interactions between DIM and LnSCCel were introduced to allow for variable effects of LnSCCel across the different days of testing. Approximate linearity of the DIM effects was checked by fitting DIM as a categorical variable and examining the estimates. After the assumptions of the full model had been evaluated using the residuals, the interaction terms were tested by likelihood-ratio tests and were removed when non-significant. To facilitate comparison of models for different 30-d DIM intervals, a main effect was kept in all models as soon as it was significant in 1 interval. The significance level for all analyses was set at P The distribution of variance at the hierarchical levels (test, heifer, and herd) was assessed for the null-models (models without fixed effects) with LnSCC as outcome variable. Probabilities of Black Holstein-Friesian heifers of having a SCC >200,000 cells/ml 59

74 Chapter 5.1 per 30- DIM interval were calculated based on the final logistic regression models using the full dataset. For this calculation, the calving date was set at September 15, and seasonal changes in SCC and MY were accounted for. The probability per 30-d DIM interval was calculated as e β /(1+e β ), with β being the predicted value on logistic scale, corresponding to an average heifer and herd (subject-specific interpretation; Dohoo et al., 2003). The sum of all probabilities over the 30-d DIM intervals was interpreted as the expected number of cases of subclinical mastitis between 15 and 365 DIM (more precisely, the number of times the testday SCC would exceed 200,000 cells/ml) per Black Holstein-Friesian heifer with a calving date of September 15. Results Descriptive Analysis of SCCel Geometric mean SCCel of the 14,766 samples was 104,000 cells/ml (Table 2). The interquartile range was 183,000 cells/ml. Approximately 70% of the samples were recorded in July to December, reflecting the Belgian calving pattern (Table 2). A difference in LnSCCel between calving seasons (P < 0.001) was noted, with April-June having higher average LnSCCel compared to July-September and October-December (Table 2). Of the heifers, 55% were Black Holstein-Friesian, and 28% were Red Holstein-Friesian. There was a difference in LnSCCel between breeds (P < 0.001) (Table 2). Somatic cell counts in early lactation progressively decreased from 178,000 cells/ml at 5 DIM to 74,000 cells/ml at 14 DIM (Figure 1) (P < 0.001). The proportion of SCCel exceeding 50,000, 100,000, 150,000, 200,000, 500,000, and 1,000,000 cells/ml was 68.8, 46.4, 34.3, 27.5, 12.3, and 6.5%, respectively. Proportion of SCCel above these thresholds decreased with increasing DIM (Figure 2). Descriptive Analysis of SCC during First Lactation Monthly test-data SCC from the first lactation, measured after 14 DIM, were available from 14,234 (96%) of the aforementioned 14,766 heifers. Geometric mean SCCel from these heifers was 112,000 cells/ml. 60

75 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation 450 SCC in early lactation (x 1000 cells/ml) ccccc Days in milk Figure 1. Geometric mean SCC/DIM in early lactation (x 1000 cells/ml; bars represent the interquartile ranges) from 14,766 heifers in the year Percentage Days in milk SCCel thresholds Figure 2. Proportion of SCC/DIM in early lactation in 2000 from 14,766 heifers exceeding the indicated SCC in early lactation (SCCel) thresholds (x 1000 cells/ml). On average 8.2 test-day measurements per heifer were available between 14 and 365 DIM (Table 1). The average interval between 2 test-day measurements was 34 d with a minimum of 21 d. Heifers maximally had 2 test-day measurements within the same 30-d DIM interval. Geometric mean SCC was 75,000 cells/ml (Table 2). The interquartile range was 109,000 cells/ml. Black Holstein-Friesian heifers had the lowest SCC, and the Belgian White 61

76 Chapter 5.1 Table 2. Descriptive statistics of SCC in early lactation and throughout the course of the first lactation (x 1000 cells/ml). Dataset 1 n GM 2 SD Min Max Early Calving season January - March ,022 April - June ,605 July - September ,088 October - December Breed Belgian White-Blue and unknown breed MRY 3, Red, Red and White, and Jersey ,605 Red Holstein-Friesian ,088 Black Holstein-Friesian ,022 Total 14, ,088 Full Test-season January - March 28, ,398 April - June 37, ,965 July - September 22, ,829 October - December 29, ,677 Breed Belgian White-Blue and unknown breed 11, ,398 MRY, Red, Red and White, and Jersey 11, ,509 Red Holstein-Friesian 32, ,965 Black Holstein-Friesian 61, ,677 Total 117, ,398 Sub Test-season January - March 16, ,199 April - June 20, ,965 July - September 12, October - December 15, ,199 Breed Belgian White-Blue and unknown breed MRY, Red, Red and White, and Jersey ,199 Red Holstein-Friesian 18, ,965 Black Holstein-Friesian 35, ,199 Total 65, ,965 1 Early = SCC in early lactation (between 5 and 14 DIM) from 14,766 dairy heifers; Full = SCC between 15 and 365 DIM from 14,234 heifers (full dataset); Sub = SCC between 15 and 365 DIM from 7807 heifers with a second test-day SCC 50,000 cells/ml milk (subset of data). 2 Geometric mean. 3 Meuse-Rhine-Yssel. 62

77 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation Blue dual-purpose heifers and heifers of unknown breed had the highest SCC (Table 2). Geometric mean SCC was 55,000 in the 15 to 45 and 46 to 75 DIM intervals after which it progressively increased from 60,000 cells/ml in the 76 to 105 DIM interval to 120,000 cells/ml at the end of lactation ( DIM). Of the heifers, 55% had a second test-day SCC 50,000 cells/ml milk. This selection resulted in a subset of heifers with lower geometric mean SCC for the particular heifers compared with all heifers (Table 2). The interquartile range was 58,000 cells/ml. Between 15 and 45 DIM, geometric mean SCC was 25,000 and progressively increased from 31,000 cells/ml at DIM to 91,000 cells/ml at the end of the first lactation. Effect of SCCel on Test-day SCC Somatic cell counts stratified by 5 SCCel-classes are presented in Figure 3. Heifers having a SCCel 50,000 cells/ml, for instance, had on average a SCC 25,000 cells/ml lower during the subsequent lactation compared with heifers starting their lactation with SCCel between 51,000 and 200,000 cells/ml SCC (x 1000 cells/ml)ccc Days in milk Figure 3. Geometric mean SCC (x 1000 cells/ml) during first lactation from 14,234 (full dataset) dairy heifers, stratified by SCC in early lactation (SCCel, x 1000 cells/ml), 0 to 50 ( ), 51 to 200 ( ), 201 to 500 ( ), 501 to 1000 ( ), and >1000 cells/ml ( ). Estimates of LnSCCel corrected for other variables of importance (DIM, MY, breed, and season) are presented in Table 3. On average, if LnSCCel of a heifer was 1 unit higher than for another heifer, the LnSCC was 0.36, 0.31, and 0.26 units higher in the first, second 63

78 Table 3. Final multilevel linear regression models per 30-d DIM interval describing LnSCC (x 1000 cells/ml) during the course of first lactation (between 15 and 365 days in milk) in 2000 and 2001 from 14,234 dairy heifers from 3264 herds (full dataset). Variables 30-d DIM interval β SE P 1 β SE P 1 β SE P 1 β SE P 1 Intercept LnSCCel 2,10, < < < <0.001 DIM 3,10, < < < <0.001 MY 4, < < < <0.001 Test-season <0.001 <0.001 < Apr Jun Ref. Ref. Ref. Ref. Jan Mar Jul Sep Oct Dec Breed NS NS NS NS Breed1 5 Ref. Ref. Ref. Ref. Breed RHF BHF LnSCCel x DIM < < < <0.001

79 Table 3. (continued). Variables 30-d DIM interval β SE P 1 β SE P 1 β SE P 1 β SE P 1 Intercept LnSCCel 2,10, < < < <0.001 DIM 3,10, < < < <0.001 MY 4, < < < <0.001 Test-season <0.001 NS NS Apr Jun Ref. Ref. Ref. Ref. Jan Mar Jul Sep Oct Dec Breed NS NS NS Breed1 5 Ref. Ref. Ref. Ref. Breed RHF BHF LnSCCel x DIM < < < <0.001

80 Table 3. (continued). Variables 30-d DIM interval β SE P 1 β SE P 1 β SE P 1 β SE P 1 Intercept LnSCCel 2,10, < < < <0.001 DIM 3,10, < < MY 4, < < < <0.001 Test-season < NS Apr Jun Ref. Ref. Ref. Ref. Jan Mar Jul Sep Oct Dec Breed NS NS NS Breed1 5 Ref. Ref. Ref. Ref. Breed RHF BHF LnSCCel x DIM < < < Bonferroni corrected P values (original values multiplied by 12). 2 Log-transformed SCC in early lactation (between 5 and 14 DIM) (x 1000 cells/ml milk). 3 Day of assessment of LnSCCel. 4 Daily milk yield (kg milk). 5 Belgian White-Blue double-purpose heifers and heifers of unknown breed. 6 Meuse-Rhine-Yssel, Red, Red and White, Jersey. 7 Red Holstein-Friesian. 8 Black Holstein-Friesian. 9 Interaction term between LnSCCel and DIM. 10 Centred by subtracting the mean value (LnSCCel = 4.7; DIM = 9.5; MY = 22.6). See text for details. 11 P values correspond to test of both main effect and interaction.

81 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation and third 30-d DIM interval, respectively. This interpretation assumes LnSCCel of the 2 heifers to be measured at the same, theoretical DIM of 9.5 (mean value between 5 and 14). The size of the effect decreased (i.e. smaller impact on LnSCC) in the later 30-d DIM intervals but remained significant over the whole lactation. Furthermore, the significant interaction between LnSCCel and DIM showed that the impact of LnSCCel depended on when it was measured in early lactation. For example, for a recording at DIM 12 the effect in the first 30-d DIM interval would be [0.013 x (12-9.5)] = Figure 4 gives model-based estimates of LnSCCel effects (comparing heifers differing by 1 unit) at 7, 9.5, and 12 DIM throughout the lactation. The impact of an elevated SCCel measured at 12 DIM is seen to be higher than of an equally elevated SCCel at an earlier DIM Regression coefficient for LnSCCel cccc Days in milk Figure 4. Regression coefficients for log-transformed SCC in early lactation (LnSCCel, x 1000 cells/ml) per 30-d DIM interval measured at 7 ( ), 9.5 ( ), and 12 ( ) DIM in early lactation. Estimates are based on the final multilevel linear regression models describing SCC (x 1000 cells/ml) during first lactation (between 15 and 365 DIM) in 2000 and 2001 from 14,234 dairy heifers from 3264 herds. The logistic regression models also showed the probability of high test-day SCC values (>200,000 cells/ml, interpretable as a case of subclinical mastitis) to increase with increasing LnSCCel (Table 4). For instance, when comparing heifers with a 1 unit difference in LnSCCel, the heifer with the higher LnSCCel was 1.86, 1.71, and 1.60 times more likely [as measured by the odds-ratio (OR)] to have a SCC >200,000 cells/ml in the first, second, and third 30-d DIM intervals, respectively. In all 30-d DIM intervals, the OR was significantly above 1. In the same manner as above, this interpretation corresponded to both heifers being 67

82 Chapter 5.1 measured at the theoretical mean DIM of 9.5, and the significant interaction between LnSCCel and DIM would alter the actual OR for recordings at other DIM (Table 4). Table 4. Regression coefficients for LnSCCel and the interaction between LnSCCel and DIM per 30-d DIM interval based on the final multilevel logistic regression models describing logit(scc200) during first lactation (between 15 and 365 days in milk) from 14,234 dairy heifers from 3264 herds (full dataset). LnSCCel LnSCCel x DIM 30-d DIM interval β SE OR 1 β SE Odds ratio. A direct estimate of the impact of LnSCCel on the risk of subclinical mastitis (SCC >200,000 cells/ml) throughout the lactation was computed as the expected number of cases of subclinical mastitis, assuming 1 recording per 30-d DIM interval, for an average Black- Holstein-Friesian heifer calving at September 15 and housed in an average herd. The expected number of cases ranged from 0.8 at a low LnSCCel (3.3, corresponding to 27,000 cells/ml, the 10th percentile) to 2.1 at a high LnSCCel (6.5, corresponding to 665,000 cells/ml, 90th percentile) for a theoretical DIM of 9.5. Heifers with a second test-day SCC 50,000 cells/ml (data subset, Figure 5) had lower test-day SCC values over time during first lactation when compared to heifers in the full dataset (Figure 3), and showed smaller differences between heifers stratified based on their SCCel-classes (Figures 3 and 5). This selection procedure resulted also in smaller regression coefficients of LnSCCel on LnSCC (Table 5); the effects of LnSCCel (at DIM 9.5) were reasonable constant at 0.1 over the whole lactation, and significant. Similarly, the estimated ORs were in the range and significant, except for the last 30-d DIM interval (Table 5). 68

83 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation SCC (x 1000 cells/ml)cccc Days in milk Figure 5. Geometric mean SCC (x 1000 cells/ml) during first lactation from 7807 heifers with a second test-day SCC 50,000 cells/ml, stratified by SCC in early lactation (SCCel, x 1000 cells/ml), 0 to 50 ( ), 51 to 200 ( ), 201 to 500 ( ), 501 to 1000 ( ), and >1000 cells/ml ( ). Table 5. Regression coefficients of LnSCCel per 30-d DIM interval based on the final multilevel linear and logistic regression models describing LnSCC (x 1000 cells/ml) and logit(scc200) during first lactation (between 76 and 365 days in milk) from 7807 dairy heifers from 2827 dairy herds (sub dataset). Linear regression models Logistic regression models 30-d DIM interval β SE P 1 β SE OR < < < < < < < < < Bonferroni corrected P values (original values multiplied by 10), corresponding to test of both main effect and interaction. 2 Odds ratio. Season was significantly associated with test-day SCC in all but 3 30-d DIM intervals and a higher MY was always associated with lower SCC (Table 3). Black Holstein-Friesian heifers tended to have lower LnSCC compared with the other breeds, even though the models 69

84 Chapter 5.1 corrected for MY (Table 3). The effect of LnSCCel on LnSCC seemed to be the same within all breeds, because the interaction between LnSCCel and breed was non-significant. The distribution of the variance over the levels of the data hierarchy was assessed in the null models (no fixed effects): between 8.5% (346 to 365 DIM) and 14.2% (256 to 285 DIM) of the variation in LnSCC resided at the herd level, leaving the majority of the variance at the heifer and test level. In all models, the quadratic term for SCCel was significant, indicating that the association between LnSCC and LnSCCel was non-linear. The models without the quadratic term were however, based on inspection of graphs of both the simple and more elaborate models, a very good approximation of the models with the quadratic term. Therefore results are presented only for models without the quadratic term. Discussion In this study, monthly milk-recording data collected through the DHI program were used to study the effect of elevated SCC measured in the first 2 wk after calving on SCC in subsequent months of the first lactation. Elevated SCCel resulted in elevated test-day SCCs, and a higher probability of test-day SCC exceeding 200,000 cells/ml. The impact of SCCel however depended on the day it was measured in the period called "early lactation". Composite milk SCC data are used worldwide as a proxy for udder health at the cow level (Schukken et al., 2003). In the first 2 wk after calving however SCC changes rapidly. Dilution due to an increasing production can contribute to the decrease of SCC (Schepers et al., 1997), but it also suggests spontaneous cure of transient IMI caused by CNS (Oliver and Mitchell, 1983). The SCCel pattern described in our study was similar to the ones seen in other studies also using composite milk samples (Dohoo, 1993; Laevens et al., 1997). Somatic cell counts are elevated shortly after calving (Dohoo and Meek, 1982) and therefore care should be taken in estimating the prevalence of infected heifers in that period of time using a certain SCC threshold. This is also reflected by the rapidly changing proportion of animals exceeding several thresholds per DIM in this study. If 200,000 cells/ml would have been chosen to classify heifers as being infected in early lactation, at 5 DIM 43% of all heifers would be regarded as being subclinically infected, whereas this would have been 24 and 19% at 9 and 14 DIM, respectively. Dohoo (1993) warned about using composite SCC before 9 DIM to estimate the prevalence of infected heifers as this would lead to an overestimation of infected animals. Stating that more than 27% of all heifers in the present study were infected 70

85 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation in early lactation using an infection threshold of 200,000 cells/ml would, consequently, be an overestimation according to the latter study. On the other hand, Barkema et al. (1999) concluded that quarter-milk SCC was applicable as of 2 DIM to determine the IMI status in an udder quarter and that quarter-milk SCC of bacteriological negative quarters was as low as 42,000 cells/ml. As a consequence, a heifer without infected quarters should have an equally low composite milk SCC. A heifer with 1 quarter being infected with a minor pathogen would most probably never have a composite milk SCC >200,000 cells/ml, due to dilution by the non-infected quarters, which would lead to an underestimation of infected heifers in early lactation using the 200,000 cells/ml threshold, contrary to that discussed by Dohoo (1993). No bacteriological cultures were done in this study, making an exact determination of the prevalence of infected heifers in early lactation impossible. More studies on the dynamics of quarter- and composite milk SCC in association with IMI status of quarters and animals in early lactation should be conducted to come to better conclusions. A high proportion of IMI caused by CNS is cleared once the heifer starts lactating (Oliver and Mitchell, 1983), and SCC in recovered quarters will take some time to return to a normal level. This may be the reason why the impact of SCCel on SCC in the rest of the lactation depended on the day it was assessed. A SCCel that is still high 2 wk after calving will have a larger impact on future SCC. It could be hypothesized that a heifer with an elevated SCCel at 5 DIM may have had quarters recovering from CNS infections resulting in a more or less normal SCCel level within the next few days. This is supported by a study done on a Californian dairy with low prevalence of major mastitis pathogens investigating the association between IMI in early lactation and SCC in the first 5 DHI test periods in heifers. Subclinical infections with minor pathogens had no effect on average SCC (Kirk et al., 1996). On the other hand, it could be hypothesized that a heifer with an elevated SCCel at 14 DIM in our study either suffered from a persistent IMI caused by a major pathogen already present before calving or had been infected early post partum. Intramammary infections caused by major mastitis pathogens most likely do not cure as easily as IMI caused by CNS (Oliver and Mitchell, 1983) resulting in elevated SCCel until 14 DIM and thereafter. However, more studies are needed to elucidate these findings as no bacteriological cultures were done in this study. A high SCC shortly after calving resulted in elevated test-day SCC in the subsequent mo, which is in accordance with the findings of Coffey et al. (1986). In that study, SCC was examined during the first and subsequent lactations by classes of SCC in initial days of first lactation to determine if heifers with initially low SCC were more at risk to subsequent 71

86 Chapter 5.1 infections. However, they concluded that the initial rank of SCC classes (<100,000, 100, ,000 and >400,000 cells/ml) was maintained throughout the remainder of first and subsequent lactation. Rupp and Boichard (2000) concluded that heifers with the highest initial SCC had the highest probability of clinical mastitis. A high SCCel shortly after calving also increased the probability of having elevated SCCs (>200,000 cells/ml) during the ongoing first lactation. This threshold was used as a cut-off value for subclinical IMI. Although every threshold has its advantages and disadvantages (Schukken et al., 2003), 200,000 cells/ml is commonly accepted as the threshold for IMI (Hillerton, 1999). Our interpretation was that heifers with an elevated SCCel, on the average, remain more at risk to subsequent subclinical IMI. Coffey et al. (1986) have found that heifers starting their lactation with a higher SCC were more at risk of acquiring new IMI, with a higher prevalence of IMI caused by major pathogens. However, as we had chosen to estimate the impact of SCCel per 30-d DIM interval, rather than using the dataset as a whole and to model the test-day SCC as repeated measures, we were not able, using the 200,000 cells/ml threshold, to distinguish between existing IMI (current and preceding SCC >200,000 cells/ml) or new infections (current SCC >200,000 cells/ml and preceding SCC 200,000 cells/ml). We were, therefore, not able to determine whether heifers with elevated SCCel were more at risk to new IMI or whether they were persistently infected. The results from our study highlight the importance of reducing the prevalence of elevated SCCel in dairy heifers, especially in the second part of the period called "early lactation". Farmers should therefore focus on prevention of IMI before calving rather than on cure of existing IMI after calving because the effect of elevated SCCel was still present and significant, although to a lesser extent, in heifers with a second test-day SCC 50,000 cells/ml. In other words, a heifer with an elevated SCCel (e.g. 750,000 cells/ml) but a healthy udder at the second test-day measurement (e.g. 45,000 cells/ml), maybe because she was successfully treated, still had higher SCC throughout the lactation, and was more at risk for IMIs than a heifer starting with a low SCCel and a similar low second test-data SCC. In a recent study, we investigated the distribution of the variance of SCCel over the herd and heifer level using multilevel regression analysis and concluded that focusing more on differences between heifers than between herds in targeting prevention is probably the way to go, as heifers within the same herd responded different to the same management practices (De Vliegher et al., 2004). This was supported by the findings from the present study, and, although the distribution was not as extreme as in the aforementioned paper, still on average 72

87 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation 10% of the variation resided at the herd level, leaving more room for improvement at the heifer and test level. As test-day records collected through the DHI program were used in our study, some heifers suffering from clinical mastitis in early lactation were missed as probably no samples for SCC measurement were collected at the particular time, at least not from the ill quarter. Therefore, we expect that the negative effect of elevated SCCel on SCC during the rest of the lactation calculated in this study is an underestimation of the true effect. An additional underestimation could be suspected from the fact that from the originally selected 14,766 heifers only 14,234 were available for further analyses. Some of the nearly 500 "missing" heifers could have been culled because of udder health problems in early lactation. However, the geometric mean SCCel of the 14,243 heifers was larger than the geometric mean SCCel from the 14,766 heifers. The heifers that were missing were therefore most probably culled for other reasons or because the farm stopped doing business. In addition, some heifers might have been sold to farmers that were not involved in the DHI program. Current research looks at the effect of elevated SCCel on production and culling hazard. Combining the results from such studies will enable us to estimate the economical losses for herds with a high prevalence of heifers suffering from elevated SCCel. Conclusions A large proportion of Belgian heifers had SCCel that suggested they had an IMI during the peripartum period. An elevated SCCel had a negative effect on udder health throughout the first lactation. The magnitude of the effect was dependent upon the point at which SCCel was measured. Elevated SCCel in the first part of the period called "early lactation" led to a reduced increase in test-day SCC than if the SCCel was elevated later. It could be hypothesized that the prevention of IMI prepartum rather than the cure of existing IMI at the time of calving is needed, as a negative effect of elevated SCCel was still present, in a sub group of heifers having a very low SCC ( 50,000 cells/ml) at the second test-day measurement. 73

88 Chapter 5.1 Acknowledgments The authors would like to thank E. De Mûelenaere and the Flemish Cattle Breeding Association (Oosterzele, Belgium) for providing us with the milk-recording data and to Elanco Belgium for supporting this study financially. References Barkema, H. W., H. Deluyker, Y. H. Schukken, and T.J.G.M. Lam Quarter-milk somatic cell count at calving and at the first 6 milkings after calving. Prev. Vet. Med. 38:1-9. Coffey, E. M., W. E. Vinson, and R. E. Pearson Somatic cell counts and infection rates for cows of varying somatic cell count in initial test of first lactation. J. Dairy Sci. 69: De Vliegher, S., H. Laevens, H. W. Barkema, I. Dohoo, H. Stryhn, G. Opsomer, and A. de Kruif Management practices and heifer characteristics associated with early lactation somatic cell counts of dairy heifers in Belgium. J. Dairy Sci. 87: Dohoo, I. R., and A. H. Meek Somatic cell counts in bovine milk. Can. Vet. J. 23: Dohoo, I. R An evaluation of the validity of individual cow somatic cell counts from cows in early lactation. Prev. Vet. Med. 16: Dohoo, I. R., E. Tillard, H. Stryhn, and B. Faye The use of multilevel models to evaluate sources of variation in reproductive performance in dairy cattle in Reunion Island. Prev. Vet. Med. 50: Dohoo, I. R., S. W. Martin, and H. Stryhn Veterinary Epidemiologic Research. First ed., AVC Inc., Charlottetown, PEI, Canada. Hillerton, J. E Redefining mastitis based on somatic cell count. Bull I. Dairy Fed. 345, 4-6. Kirk, J. H., J. C. Wright, S. L. Berry, J. P. Reynolds, J. P. Maas, and A. Ahmadi Relationship of milk culture status at calving with somatic cell counts and milk production of dairy heifers during early lactation on a Californian dairy. Prev. Vet. Med. 28:

89 Impact of Early Lactation Somatic Cell Count on Somatic Cell Counts over the First Lactation Laevens, H., H. Deluyker, Y. H. Schukken, L. De Meulemeester, R. Vandermeersch, E. De Mûelenaere, and A. de Kruif Influence of parity and stage of lactation on the somatic cell count in bacteriologically negative cows. J. Dairy Sci. 80: McDermott, J. J., and Y. H. Schukken A review of methods used to adjust for cluster effects in explanatory epidemiological studies of animal populations. Prev. Vet. Med. 18: Myllys, V Staphylococci in heifer mastitis before and after parturition. J. Dairy Res. 62: Neave, F. K., F. H. Dodd, R. G. Kingwill, and D. R. Westgarth Control of mastitis in the dairy herd by hygiene and management. J. Dairy Sci. 52: Oliver, S. P., and B. A. Mitchell Intramammary infections in primigravid heifers near parturition. J. Dairy Sci. 66: Oliver, S. P., M. J. Lewis, B. E. Gillespie, and H. H. Dowlen Influence of prepartum antibiotic therapy on intramammary infections in primigravid heifers during early lactation. J. Dairy Sci. 75: Oliver, S. P., M. J. Lewis, B. E. Gillespie, H. H. Dowlen, E. C. Jaenicke, and R. K. Roberts Prepartum antibiotic treatment of heifers: milk production, milk quality and economic benefit. J. Dairy Sci. 86: Pankey, J. W., P. A. Drechsler, and E. E. Wildman Mastitis prevalence in primigravid heifers at parturition. J. Dairy Sci. 74: Rasbash, J., W. Browne, H. Goldstein, M. Yang, I. Plewis, M. Healy, G. Woodhouse, D. Draper, I. Langford, and T. Lewis A user's guide to MlwiN. Second ed., Institute of Education, London, UK. Roberson, J. R., L. K. Fox, D. D. Hancock, and C. C. Gay Coagulase-positive Staphylococcus intramammary infections in dairy heifers. J. Dairy Sci. 77: Rupp, R., F. Beaudeau, and D. Boichard Relationship between milk somatic cell counts in the first lactation and clinical mastitis occurrence in the second lactation of French Holstein cows. Prev. Vet. Med. 46: Rupp, R., and D. Boichard Relationship of early first lactation somatic cell count with risk of subsequent first clinical mastitis. Livest. Prod. Sci. 62: Schepers, A. J., T.J.G.M. Lam, Y. H. Schukken, J.B.M. Wilmink, and W.J.A. Hanekamp Estimation of variance components for somatic cell counts to determine thresholds for uninfected quarters. J. Dairy Sci. 80:

90 Chapter 5.1 Schukken, Y. H., D. J. Wilson, F. Welcome, L. Garrison-Tikofsky, and R. N. Gonzalez Monitoring udder health and milk quality using somatic cell counts. Vet. Res. 34:

91

92

93 Chapter 5.2 Impact of Early Lactation Somatic Cell Count in Heifers on Milk Yield over the First Lactation S. De Vliegher, 1 H. W. Barkema, 2 H. Stryhn, 2 G. Opsomer, 1 and A. de Kruif 1 1 Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 2 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Charlottetown, PEI, Canada Journal of Dairy Science, accepted for publication.

94 Chapter

95 Impact of Early Lactation Somatic Cell Count on Milk Yield over the First Lactation Abstract The objective of this study was to estimate the impact of somatic cell count (SCC) in early lactation (SCCel) [measured between 5-14 days in milk (DIM)] of dairy heifers on test-day milk yield (MY) during the first lactation. In total, 117,496 four-weekly test-day records of 14,243 heifers were used. A multilevel regression analysis, which included test-day SCC among the explanatory variables, revealed that an increase by 1 unit of the natural log-transformed SCCel (LnSCCel) was on average associated with a decrease in MY of 0.13 kg/d later in lactation. As an example, a heifer with a SCCel of 50,000 cells/ml measured at 10 DIM was estimated to produce 119 kg and 155 kg more milk during its first lactation than heifers with a SCCel of 500,000 cells/ml and 1,000,000 cells/ml, respectively. When not accounting for test-day SCC, the effect of LnSCCel on MY was larger, indicating that part of the negative impact of elevated SCCel was associated with elevated test-day SCCs later in lactation. Furthermore, an elevated SCCel at 14 DIM had a larger impact than an equally elevated SCCel measured at an earlier DIM. In addition, the negative effect of an elevated SCCel remained present during almost the entire first lactation in a subgroup of heifers with a second test-day SCC 50,000 cells/ml, suggesting that prevention rather than cure of an elevated SCCel should be preferred. This study stresses the importance of heifers having low SCCel, as an elevated SCCel will negatively affect milk production during the first lactation, probably via impairment of the mammary function and, to a smaller extent, via elevated test-day SCCs later in lactation. Key Words Dairy heifer, early lactation, milk yield, somatic cell count Abbreviation key: CNS = coagulase-negative staphylococci, DIM = days in milk, DHI = Dairy Herd Improvement, IMI = intramammary infection, LnSCC = natural log-transformed SCC, LnSCCel = natural log-transformed SCCel, MY = milk yield at test-day (kg), SCC = somatic cell count, SCCel = SCC in early lactation (between 5 and 14 DIM). 81

96 Chapter 5.2 Introduction Heifer mastitis is a well known problem. Many studies have reported a high prevalence of IMI in heifers around calving (Oliver and Mitchell, 1983; Pankey et al., 1991; Oliver et al., 1992; Roberson et al, 1994; Myllys, 1995; Aarestrup and Jensen 1997; Oliver et al., 2003). More than 27% of 14,766 Belgian dairy heifers had a SCCel >200,000 cells/ml, suggesting they had an IMI during the peripartum period (De Vliegher et al., 2004a). Coagulase-negative staphylococci (CNS) are responsible for the majority of the IMIs in non-lactating and freshly calved heifers, but Staphylococcus aureus and environmental pathogens may also play a role (Oliver and Mitchell, 1983; Pankey et al., 1991; Oliver et al., 1992; Roberson et al, 1994; Myllys, 1995; Aarestrup and Jensen, 1997; Oliver et al., 2003). Intramammary infection at parturition, particularly when caused by a major pathogen, will result in an increased SCC at that moment (Barkema et al., 1999). Heifers with an elevated SCCel on average have elevated test-day SCCs, are more at risk of having test-day SCCs >200,000 cells/ml (De Vliegher et al., 2004a), and have an increased probability of clinical mastitis (Rupp and Boichard, 2000). Future milk production can be compromised as well (Oliver and Jayarao, 1997). The negative association between test-day SCC and MY has been addressed in some recent studies: the individual MY loss in heifers has been estimated at 1.29 kg/d for each unit increase in the log 10 SCC (Koldeweij et al., 1999). In lactating heifers without clinical mastitis, the reduction in MY was 0.30, 0.61 and 1.09 kg when SCC was 100,000, 200,000, and 600,000 cells/ml, respectively (Hortet, 1999). In organic dairy herds, heifers had a production loss of 0.20 kg/d of energy corrected milk with each twofold increase in SCC between 100,000 and 1,500,000 cells/ml (Bennedsgaard et al., 2003). The association between first test-day SCC and SCC later in lactation was investigated by Coffey et al. (1986): heifers with a first test-day SCC <100,000, in the range 100, ,000, and >400,000 cells/ml produced 6452, 6050, and 5696 kg during first lactation, respectively. Kirk et al. (1996), on the other hand, found that IMI with minor pathogens in heifers had no effect on average milk production during early to middle lactation. However, in that study only 1 herd was included. The primary objective of the present study was to estimate the impact of SCCel on MY in the first lactation using multilevel regression analysis, taking into account the day of assessment of SCCel. The study accompanies a previous study (De Vliegher et al., 2004a) on the impact of SCCel on test-day SCCs in the first lactation, based on the same dataset. The 82

97 Impact of Early Lactation Somatic Cell Count on Milk Yield over the First Lactation present analysis extends the previous analysis by focusing on MY, and by investigating both the direct and indirect effects of elevated SCCel, while accounting for the hierarchical structure of the data. Materials and Methods Initial Dataset and Data Handling The data used in this study are described elsewhere (De Vliegher et al., 2004a). In short, approximately monthly milk recordings ( 5 DIM) from 2000 and 2001 of all heifers of all Flemish (Belgium) dairy herds enrolled in the DHI program (Flemish Cattle Breeding Association, Oosterzele, Belgium) were used, including: SCC and MY on test-day (kg milk), breed, DIM, and date of measurement. Heifers were included if their first test-day in 2000 occurred between 5 and 14 DIM. In total, 117,496 test-day records (full dataset) from the first lactation (measured after 14 DIM) were available of 14,234 heifers belonging to 3264 herds. A subset of data was created by selecting heifers with a second test-day SCC (measured after 14 and before 75 DIM) 50,000 cells/ml, resulting in 65,458 test-day records from 7807 heifers on 2827 herds. Statistical Analysis The statistical analysis aimed at quantifying the effect of SCC in early lactation (measured between 5 and 14 DIM and referred to as SCCel throughout the paper) on MY during the first lactation. A natural logarithmic transformation of SCC (LnSCC) and SCCel (LnSCCel) was used (De Vliegher et al., 2004a). For analysis of the repeated measurements of each heifer, both the full data set and subset of data were split into 12 subsets referred to as "30-d DIM intervals" (15-45, 46-75, DIM), reflecting the monthly milk recordings. These intervals were analyzed separately, except that no analysis was performed for the first 2 30-d DIM intervals from the subset of data because selection of animals was based on the measurements within those 2 intervals. To account for multiple testing of the same hypotheses in the 30-d DIM intervals, all P values were multiplied by the number of intervals analyzed (Bonferroni correction, see De Vliegher et al., 2004a, for discussion and justification of the analytical approach). Multilevel linear regression models (1 model per 30-d DIM interval) were used to analyze MY as the outcome variable. The models included random effects for herds and 83

98 Chapter 5.2 heifers, the latter to account for a small number of heifers with 2 test recordings within the same 30-d interval [except for the first (15-45 DIM) and twelfth ( DIM) intervals]. All random effects and error terms were assumed normally distributed. The full model included regression terms for LnSCCel (predictor of main interest), DIM (between 5 and 14, the day of assessment of LnSCCel), and the categorical variables test-season and breed (4 levels both: as listed in Table 1). Moreover, the model included interaction terms between LnSCCel and the predictors DIM and breed, as well as a quadratic term for LnSCCel and its interaction with DIM. All models were fitted with and without LnSCC as predictor, which allowed studying changes in the estimate of LnSCCel when including LnSCC in the model, and helped to estimate the direct and indirect effects of LnSCCel. The continuous predictors were centred by subtracting their overall mean. After the assumptions of the full model had been evaluated using the residuals, the interaction terms were tested and were removed when non-significant (likelihood-ratio test). To facilitate comparison of models for different 30-d DIM intervals, a main effect was kept in all models as soon as it was significant in 1 30-d DIM interval. The distribution of variance of MY at the hierarchical levels (test, heifer, and herd) was assessed for the null-models (without fixed effects). The proportion of variation explained by the fixed effects was calculated as 1 minus the ratio of the total variance of the final model and the total variance of the null model (Snijders and Bosker, 1999). Predicted cumulative production between 15 and 365 DIM was calculated for Black Holstein-Friesian heifers by multiplying the daily average predicted MY per 30-d DIM interval with 30 (20 for the last interval) and summing up these values over all 30-d DIM intervals. An average test-season effect was assumed. All analyses were carried out using the MLwiN software, version 1.2 (Rasbash et al., 2000). The significance level for all analyses was set at P Results Descriptive Analysis of SCCel and MY Geometric mean SCCel of the 14,234 selected heifers was 112,000 cells/ml, ranging from 5000 to 25,000,000 cells/ml. More details are presented elsewhere (De Vliegher et al., 2004a). Average MY between 15 and 365 DIM was 22.7 kg/d, ranging from 0.9 to 47.2 kg/d (Table 1). The inter-quartile range was 6.7 kg. Black Holstein-Friesians had the highest and 84

99 Impact of Early Lactation Somatic Cell Count on Milk Yield over the First Lactation the Belgian White Blue dual-purpose heifers and heifers of unknown breed lowest milk production. Small differences were present between test-seasons (Table 1). Average MY was 23.5 kg/d in the d DIM interval, and increased to 23.8 kg/d in the interval, after which MY gradually decreased to 21 kg/d at the end of the lactation ( DIM). Table 1. Descriptive statistics of milk yield (kg/d) throughout the first lactation. Dataset 1 n Av 2 SD Min Max Full Test-season January - March 28, April - June 37, July - September 22, October - December 29, Breed Belgian White-Blue and unknown breed 11, MRY 3, Red, Red and White, and Jersey 11, Red Holstein-Friesian 32, Black Holstein-Friesian 61, Total 117, Sub Test-season January - March 16, April - June 20, July - September 12, October - December 15, Breed Belgian White-Blue and unknown breed MRY, Red, Red and White, and Jersey Red Holstein-Friesian 18, Black Holstein-Friesian 35, Total 65, Full = Milk yield between 15 and 365 DIM from 14,234 heifers (full dataset). Sub = Milk yield between 15 and 365 DIM from 7807 heifers with a second test-day SCC 50,000 cells/ml (subset of data). 2 Average. 3 Meuse- Rhine-Yssel. Of the 14,234 heifers, 55% had a second test-day SCC 50,000 cells/ml and the 7807 heifers within this subset of data had an average MY 0.4 kg higher than all heifers (Table 1). Table 1 also shows differences between breeds and test-seasons. 85

100 Chapter 5.2 Effect of SCCel on Test-day MY Milk yield stratified by 5 SCCel-classes is presented in Figure 1. Heifers having a SCCel 50,000 cells/ml produced on average 0.26 kg/d more during the subsequent test-days compared with heifers starting their first lactation with a SCCel between 51,000 and 200,000 cells/ml. This difference was on average 1.44 kg/d for heifers in the lowest SCCel-class when compared with heifers with a SCCel >1,000,000 cells/ml. In general, differences between SCCel-classes became smaller towards the end of the lactation Daily milk yield (kg) ccc Days in milk Figure 1. Average milk yield (kg/d) during first lactation from 14,234 (full dataset) dairy heifers, stratified by SCC in early lactation (SCCel, x 1000 cells/ml), 0 to 50 ( ), 51 to 200 ( ), 201 to 500 ( ), 501 to 1000 ( ), and >1000 cells/ml ( ). Regression coefficients of LnSCCel on MY corrected for DIM, breed, season, and LnSCC are presented in Table 2. On average, if LnSCCel of a heifer was 1 unit higher than for another heifer, MY was 0.25, 0.16, and 0.14 kg/d lower in the first, second and third 30-d DIM intervals, respectively. This interpretation assumes LnSCCel of the 2 heifers to be measured at the same, theoretical DIM of 9.5 (mean value between 5 and 14). The size of the effect decreased throughout the lactation (smaller impact in the later 30-d DIM intervals). Furthermore, the interaction between LnSCCel and DIM was significant in some of the 30-d DIM intervals, indicating that the impact of LnSCCel on MY depended on when it was 86

101 Table 2. Final multilevel linear regression models per 30-d DIM intervals describing milk yield (kg/d) during the course of first lactation (between 15 and 365 days in milk) in 2000 and 2001 from 14,234 dairy heifers from 3264 herds, accounting for log-transformed test-day SCC (full dataset). 30-d DIM interval Variables β SE P 1 β SE P 1 β SE P 1 β SE P 1 Intercept LnSCCel 2,10, < < < <0.001 DIM 3,10, < LnSCC 4, < < NS Season <0.001 <0.001 <0.001 <0.001 Apr Jun Ref.... Ref.... Ref.... Ref.... Jan Mar Jul Sep Oct Dec Breed <0.001 <0.001 <0.001 <0.001 Breedc1 5 Ref.... Ref.... Ref.... Ref.... Breed RHF BHF LnSCCel x DIM

102 Table 2. (continued). 30-d DIM interval Variables β SE P 1 β SE P 1 β SE P 1 β SE P 1 Intercept LnSCCel 2,10, < < <0.001 DIM 3,10, NS NS NS NS LnSCC 4, < < < <0.001 Season <0.001 <0.001 <0.001 <0.001 Apr Jun Ref.... Ref.... Ref.... Ref.... Jan Mar Jul Sep Oct Dec Breed <0.001 <0.001 <0.001 <0.001 Breedc1 5 Ref.... Ref.... Ref.... Ref.... Breed RHF BHF LnSCCel x DIM NS NS NS NS

103 Table 2. (continued). 30-d DIM interval Variables β SE P 1 β SE P 1 β SE P 1 β SE P 1 Intercept LnSCCel 2,10, < NS NS DIM 3,10, NS NS NS NS LnSCC 4, < < < <0.001 Season <0.001 <0.001 <0.001 <0.001 Apr Jun Ref.... Ref.... Ref.... Ref.... Jan Mar Jul Sep Oct Dec Breed <0.001 <0.001 <0.001 <0.001 Breedc1 5 Ref.... Ref.... Ref.... Ref.... Breed RHF BHF LnSCCel x DIM NS NS NS NS 1 Bonferroni corrected P values (original values multiplied by 12). 2 Log-transformed SCC in early lactation (between 5 and 14 DIM) (x 1000 cells/ml). 3 Day of assessment of LnSCCel. 4 Log-transformed SCC at test day (x 1000 cells/ml) 5 Belgian White-Blue double-purpose heifers and heifers of unknown breed. 6 Meuse-Rhine-Yssel, Red, Red and White, Jersey. 7 Red Holstein-Friesian. 8 Black Holstein-Friesian. 9 Interaction term between LnSCCel and DIM. 10 Centred by subtracting the mean value (LnSCCel = 4.7; DIM = 9.5, LnSCC = 4.3). See text for details. 11 P values correspond to test of both main effect and interaction.

104 Chapter 5.2 measured in early lactation. As an example, for a recording at DIM 12, the effect in the first 30-d DIM interval would be [-0.039*(12-9.5)] = kg/d, whereas if the recording took place at DIM 7, the effect would be kg/d. This LnSCCel effect throughout the whole lactation comparing heifers differing by 1 unit calculated at 7, 9.5, and 12 DIM is illustrated in Figure 2. The impact of an elevated LnSCCel measured at 12 DIM was higher -0,40-0,35-0,30 Milk loss (kg) -0,25-0,20-0,15-0,10-0,05 0, Days in milk Figure 2. Regression coefficients for log-transformed SCC in early lactation (LnSCCel, x 1000 cells/ml) per 30-d DIM interval measured at 7 ( ), 9.5 ( ), and 12 ( ) DIM. Estimates are based on the final multilevel linear regression models describing milk yield (kg/d) during first lactation (between 15 and 365 days in milk) in 2000 and 2001 from 14,234 dairy heifers from 3264 herds, accounting for log-transformed test-day SCC (LnSCC, x 1000 cells/ml). Regression coefficients for LnSCCel (x 1000 cells/ml) per 30-d DIM interval measured at 9.5, but based on the final multilevel linear regression models not accounting for logtransformed test-day SCC (x 1000 cells/ml) are presented by - -. than of an equally elevated LnSCCel at 7 DIM. Predicting the cumulative milk production of heifers with different SCCel values (50,000, 200,000, 500,000, and 1,000,000 cells/ml, respectively) measured between 5 and 14 DIM also demonstrated this (Figure 3). A higher SCCel was always associated with a lower production, but the predicted differences were smaller for heifers with different SCCels measured at DIM 5 compared with heifers with different SCCels recorded at a later DIM. Test-day MY was significantly associated with season and breed. Black Holstein- Friesian heifers out-produced the other breeds (Table 2). The effect of LnSCCel on MY did not differ between breeds, as the interaction term between LnSCCel and breed was never significant. 90

105 Impact of Early Lactation Somatic Cell Count on Milk Yield over the First Lactation Fitting models that did not include LnSCC as predictor variable resulted in larger estimates of LnSCCel. On average, if LnSCCel of a heifer was 1 unit higher than for another heifer, MY was 0.32, 0.20, and 0.16 kg/d lower in the first, second and third 30-d DIM intervals, respectively. As before, this interpretation assumes LnSCCel of the 2 heifers to be measured at the same, theoretical DIM of 9.5. The model-based estimates for LnSCCel (not correcting for LnSCC) in all 30-d DIM intervals are presented in Figure 2 (- -). The estimates of all other predictor variables are not presented because the changes were minimal. Heifers with a second test-day SCC 50,000 cells/ml (subset of data, Figure 4) had higher test-day MY compared to heifers in the full dataset (Figure 1), and the differences between heifers in different SCCel-classes were smaller (Figures 1 and 3), corresponding to smaller regression coefficients of LnSCCel on MY (Table 3, left column). However, the effect of LnSCCel (at DIM 9.5) was reasonably constant at -0.1 kg/d and significant until 255 DIM. As before, fitting models without LnSCC as a predictor variable resulted in larger LnSCCel estimates (Table 3, right column) Cumulative milk yield (kg) vvvccc Days in milk in early lactation SCCel Figure 3. Estimated cumulative milk production between 15 and 365 DIM from Black Holstein-Friesian heifers with different SCC in early lactation (x 1000 cells/ml) measured at different days in milk, based on the final models accounting for log-transformed SCC. Approximately 55% and 44% of the variation in MY resided at the herd level and heifer level, respectively, with a very small proportion left at the test level. In between 10.3% (46-75 DIM) and 16.2% ( DIM) of the variation was explained by adding the fixed effects. 91

106 Chapter Daily milk yield (kg) ccc Days in milk Figure 4. Average milk yield (kg/kg) during first lactation from 7807 heifers with a second test-day SCC 50,000 cells/ml, stratified by SCC in early lactation (SCCel, x 1000 cells/ml), 0 to 50 ( ), 51 to 200 ( ), 201 to 500 ( ), 501 to 1000 ( ), and >1000 cells/ml ( ). Table 3. Regression coefficients for log-transformed SCC in early lactation (LnSCCel, x 1000 cells/ml) per 30-d DIM interval based on the final multilevel linear regression models [with and without log-transformed test-day SCC (LnSCC, x 1000 cells/ml)] describing milk yield (kg/d) during first lactation (between 76 and 365 days in milk) from 7807 dairy heifers from 2878 herds (subset of data). Models with LnSCC 1 Models without LnSCC 30-d DIM interval β SE P 2 β SE P < < NS NS NS NS NS NS 1 Centred by subtracting the mean value (3.9). See text for details. 2 Bonferonni corrected P values (original values multiplied by 10), corresponding to test of both main effect and interaction. In some 30-d DIM intervals, the quadratic term of LnSCCel was significant, indicating that, at least in those intervals, the association between MY and LnSCCel was non-linear. It was, however, decided to present only models without the quadratic term, making interpretation of the models more straightforward. This approach was also preferred earlier 92

107 Impact of Early Lactation Somatic Cell Count on Milk Yield over the First Lactation (De Vliegher et al., 2004a) and was based on inspection of graphs presenting the linear and non-linear relation, showing a good agreement. Discussion Monthly milk-recording data were used to investigate the impact of SCCel, measured in the first 2 wk after calving, on MY in the subsequent first lactation. The day on which SCCel was assessed was taken into account. Elevated SCCel was associated with a reduced MY and a lower cumulative milk production during the first lactation, resulting in suboptimal revenues for the farmer. Prevalence of IMI is high in peripartum dairy heifers (Oliver and Mitchell, 1983; Pankey et al., 1991; Oliver et al., 1992; Roberson et al, 1994; Myllys, 1995; Aarestrup and Jensen 1997; Oliver et al., 2003). Coagulase-negative staphylococci are the predominant organisms causing these IMI, but the number of CNS-infections decreases markedly in early lactation (Oliver and Mitchell, 1983; Matthews et al., 1992; Myllys, 1995; Kirk et al., 1996; Barkema et al., 1999), suggesting that most teat canal colonisations and subclinical IMI with CNS around calving will not become chronic (Myllys, 1995). However, duration of IMI varies among CNS species. Staphylococcus chromogenes, the most prevalent CNS before parturition, disappears from most quarters shortly after freshening, whereas Staphylococcus simulans persists into lactation (Aarestrup and Jensen, 1997). Intramammary infections caused by major mastitis pathogens will not cure as easily as IMI caused by CNS (Oliver and Mitchell, 1983; Roberson et al., 1994). For instance, 50% of S. aureus IMI in heifers persists for at least 1 mo (Roberson et al., 1994). In addition, the incidence rate of clinical mastitis in heifers is very high during the first 14 d of lactation (Barkema et al., 1998). Bacteriological culture results during early lactation were not available in our study. Rather, SCCel was used to reflect udder health of the heifers shortly after calving and to study the effect on future milk production. This approach allowed for a quantification of the impact based on a large number of animals. An elevated SCCel shortly after calving was associated with lower test-day MY in the subsequent months. This is not surprising as mammary tissue of quarters infected with S. aureus is less developed and has greater leukocyte infiltration compared with uninfected quarters (Trinidad et al., 1990). Quarters infected with CNS also exhibit greater leukocyte infiltration and a higher proportion of interalveolar stroma compared with uninfected controls. Furthermore, presence of IMI in unbred heifers increases leukocytosis into the mammary 93

108 Chapter 5.2 gland and reduces secretory activity in heifers, suggesting an adverse effect on future milk production (Trinidad et al., 1990). Our study shows that an elevated SCCel, reflecting the inflammation process of the mammary gland parenchyma caused by mastitis pathogens present at calving, has indeed a detrimental impact on the lactational milk production. Lower milk production will be the consequence of the pathogenic effect of bacterial virulence factors and damage caused by PMN s and other inflammatory cells on the mammary parenchyma. Part of the negative effect on MY in this study also was associated with elevation of subsequent test-day SCC. This was demonstrated by comparing models with and without LnSCC as a predictor variable: LnSCC, when incorporated into the models, took away part of the effect of LnSCCel. This was not unexpected, as, based on the same dataset, we demonstrated earlier that heifers with elevated LnSCCel will on the average, have elevated SCC later in lactation, and will have more SCC>200,000 cells/ml, suggesting the presence of subclinical mastitis (De Vliegher et al., 2004a). Heifers with a first test-day SCC <100,000, produced approximately 400 and 750 kg more than heifers with a first test-day SCC in the range 100, ,000 and >400,000 cells/ml, respectively (Coffey et al., 1986). These differences were larger than in our study, but the comparison is difficult as, for instance, it was not clear when the first test-day SCCs were on average recorded in that study. Furthermore, culling bias may have influenced our results as heifers with udder disorders at parturition are more likely to be culled during their first lactation (Waage et al., 2000). This has probably resulted in an overall underestimation of the effect of an elevated SCCel on MY throughout the first lactation and also explains the smaller impact of LnSCCel on MY in the later 30-d DIM intervals. Day of measurement of the first test-day SCC was not part of the evaluation by Coffey et al. (1986), although the statistically significant interaction between SCCel and DIM in our study indicates it is important to take this into consideration. The biology behind this finding is associated with the decreasing SCCel between 5 and 14 DIM (Dohoo, 1993; Laevens et al., 1997; Barkema et al., 1999; De Vliegher et al., 2001), reflecting the declining prevalence of CNS-infected quarters in early lactation and an increasing MY (Schepers et al., 1997). In the previous study, the effect of SCCel on test-day SCC also depended on the day of assessment in early lactation (De Vliegher et al., 2004a). As mentioned earlier, a high proportion of IMI caused by CNS is cleared once the heifer starts lactating because of their transient nature, but SCC in recovered quarters will take some time to return to a normal level. A heifer with an elevated SCCel at 5 DIM may have had quarters recovering from CNS-infections resulting in a more or less normal SCCel level within the next few days. Such a heifer will not produce 94

109 Impact of Early Lactation Somatic Cell Count on Milk Yield over the First Lactation much less milk than a heifer with a very low SCCel at 5 DIM. This hypothesis is supported by the findings from a study in which 339 heifers were cultured in early lactation in a large dairy herd with a low prevalence of major mastitis pathogens. Intramammary infection caused by minor pathogens did not have a significant effect on milk production (Kirk et al., 1996). On the other hand, a SCCel that was still elevated 2 w after calving had a larger impact on future milk production. Possibly, heifers with an elevated SCCel later in the early post partum period suffered from a persistent IMI (e.g. caused by S. aureus), already present before calving, or had been infected early post partum, resulting in elevated SCCel until 14 DIM and thereafter. The damaging effect of an S. aureus infection will be more extensive as compared with a CNS IMI simply because S. aureus possesses a wider arsenal of virulence factors that allows this pathogen to cause more damage and, therefore, higher and extended SCC values than CNS. However, more studies are needed to elucidate these hypotheses and bacteriological cultures should be included. The results from this study should motivate dairy farmers to try to reduce the number of heifers with an elevated SCCel. Farmers invest time and money in raising their young stock, but their heifers will not produce according to their expectations if they freshen with an elevated SCCel. Therefore, farmers should focus on prevention of IMI before and at calving rather than on treating IMI after calving as the negative effect of an elevated SCCel on MY was still present, although smaller in size, in heifers with a second test-day SCC 50,000 cells/ml. This means that even if a heifer has a low second test-day SCC it will, on average, be out-produced by a heifer with an equally low SCC at the second test-day but with a lower SCCel. Prevention of IMI prepartum in heifers remains difficult because our knowledge of the pathogenesis is still limited, even though some herd- and heifer-associated factors related to heifer mastitis have been identified (Fox et al., 1995; Myllys and Rautala, 1995; Østerås et al., 1997; Waage et al., 1998; Bareille et al., 2000; Waage et al., 2001; De Vliegher et al., 2004b). Multilevel modelling can be an aid in estimating the relative importance of herd and individual factors with a view to target animal health interventions more effectively (McDermott and Schukken, 1994). Although management is key in mastitis prevention, the findings from another study using this technique indicated that focusing on differences between heifers rather than differences between herds is necessary to reduce the prevalence of heifers with an elevated SCCel as the vast majority of the variation in SCCel was present at the heifer level leaving more room for improvement compared with the herd level (De Vliegher et al., 2004b). The findings for SCCel do not apply for MY as approximately 55% of 95

110 Chapter 5.2 the unexplained variation in MY was present at the herd level, indicating that farmers have more room to improve production in heifers by applying better management practices. Conclusions Elevated SCCel in heifers is associated with a decreased lactational production. The cumulative milk loss over the whole first lactation can be high: a heifer with a SCC of 500,000 cells/ml measured at 10 DIM, for instance, produced 119 kg less during her first lactation than a heifer with a SCC of 50,000 cells/ml also measured at 10 DIM. The effect is, however, modified by the day of assessment of SCCel: an elevated SCCel at 5 DIM is less consequential than an equally high SCCel at 14 DIM. The suboptimal production could partially be explained by detrimental effects of IMI at calving on the secretory tissue but our study also indicated that part of the effect was associated with permanently elevated SCCs throughout lactation. The financial losses for a farm with a high prevalence of heifers with elevated SCCel can be large, resulting in suboptimal revenues. Acknowledgments The authors would like to thank E. De Mûelenaere and the Flemish Cattle Breeding Association (Oosterzele, Belgium) for providing us with the milk-recording data and to Elanco Belgium for supporting this study financially. References Aarestrup, F. M., and N. E. Jensen Prevalence and duration of intramammary infection in Danish dairy heifers during the peripartum period. J. Dairy Sci. 80: Bareille, N., H. Seegers, M. B. Kiebre-Toe, F. Beaudeau, and C. Fourichon Risk factors for elevated milk somatic cell counts during early lactation in dairy heifers. Proc. 10th Intern. Congr. Anim. Hyg., Maastricht, The Netherlands: Barkema, H. W., Y. H. Schukken, T.J.G.M. Lam, M. L. Beiboer, H. Wilmink, G. Benedictus, and A. Brand Incidence of clinical mastitis in dairy herds grouped in 3 categories by bulk milk somatic cell counts. J. Dairy Sci. 81:

111 Impact of Early Lactation Somatic Cell Count on Milk Yield over the First Lactation Barkema, H. W., H. Deluyker, Y. H. Schukken, and T.J.G.M. Lam Quarter-milk somatic cell count at calving and at the first 6 milkings after calving. Prev. Vet. Med. 38:1-9. Bennedsgaard, T. W., C. Enevoldsen, S. M. Thamsborg, and M. Vaarst Effect of mastitis treatment and somatic cell counts on milk yield in Danish organic dairy cows. J. Dairy Sci. 86: Coffey, E. M., W. E. Vinson, and R. E. Pearson Somatic cell counts and infection rates for cows of varying somatic cell count in initial test of first lactation. J. Dairy Sci. 69: De Vliegher, S., H. Laevens, G. Opsomer, E. De Mûelenaere, and A. de Kruif Somatic cell counts in dairy heifers during early lactation. Flem. Vet. J. 70: De Vliegher, S., H. W. Barkema, H. Stryhn, G. Opsomer, and A. de Kruif. 2004a. Impact of early lactation somatic cell count in heifers on somatic cell counts over the first lactation. J. Dairy Sci. 87: De Vliegher, S., H. Laevens, H. W. Barkema, I. Dohoo, H. Stryhn, G. Opsomer, and A. de Kruif. 2004b. Management practices and heifer characteristics associated with early lactation somatic cell count of Belgian dairy heifers. J. Dairy Sci. 87: Dohoo, I. R An evaluation of the validity of individual cow somatic cell counts from cows in early lactation. Prev. Vet. Med. 16: Fox, L. K., S. T. Chester, J. W. Hallberg, S. C. Nickerson, J. W. Pankey, and L. D. Weaver Survey of intramammary infections in dairy heifers at breeding age and first parturition. J. Dairy Sci. 78: Hortet, P., F. Beaudeau, H. Seegers, and C. Fourichon Reduction in milk yield associated with somatic cell counts up to 600,000 cells/ml in French Holstein cows without clinical mastitis. Prev. Vet. Med. 61: Kirk, J. H., J. C. Wright, S. L. Berry, J. P. Reynolds, J. P. Maas, and A. Ahmadi Relationship of milk culture status at calving with somatic cell counts and milk production of dairy heifers during early lactation on a Californian dairy. Prev. Vet. Med. 28: Koldeweij, E., U. Emanuelson, and L. Janson Relation of milk production loss to milk somatic cell count. Acta Vet. Scand. 40: Laevens, H., H. Deluyker, Y. H. Schukken, L. De Meulemeester, R. Vandermeersch, E. De Mûelenaere, and A. de Kruif Influence of parity and stage of lactation on the somatic cell count in bacteriologically negative cows. J. Dairy Sci. 80:

112 Chapter 5.2 Matthews, K. R., R. J. Harmon, and B. E. Langlois Prevalence of Staphylococcus species during the periparturient period in primiparous and multiparous cows. J. Dairy Sci. 75: McDermott, J. J., and Y. H. Schukken A review of methods used to adjust for cluster effects in explanatory epidemiological studies of animal populations. Prev. Vet. Med. 18: Myllys, V Staphylococci in heifer mastitis before and after parturition. J. Dairy Res. 62: Myllys, V., and H. Rautala Characterization of clinical mastitis in primiparous heifers. J. Dairy Sci. 78: Oliver, S. P., and B. A. Mitchell Intramammary infections in primigravid heifers near parturition. J. Dairy Sci. 83: Oliver, S. P., M. J. Lewis, B. E. Gillespie, and H. H. Dowlen Influence of prepartum antibiotic therapy on intramammary infections in primigravid heifers during early lactation. J. Dairy Sci. 75: Oliver, S. P., and B. M. Jayarao Coagulase-negative staphylococcal intramammary infections in cows and heifers during the nonlactating and periparturient periods. J. Vet. Med. B. 44: Oliver, S. P., M. J. Lewis, B. E. Gillespie, H. H. Dowlen, E. C. Jaenicke, and R. K. Roberts Prepartum antibiotic treatment of heifers: milk production, milk quality and economic benefit. J. Dairy Sci. 86: Østerås, O., R. B. Larssen, and E. Simensen Environmental risk factors associated with mastitis in heifers. Proc. 9th Intern. Congr. Anim. Hyg., Helsinki, Finland: Pankey, J. W., P. A. Drechsler, and E. E. Wildman Mastitis prevalence in primigravid heifers at parturition. J. Dairy Sci. 74: Rasbash, J., W. Browne, H. Goldstein, M. Yang, I. Plewis, M. Healy, G. Woodhouse, D. Draper, I. Langford, and T. Lewis A user's guide to MlwiN. Second ed., Institute of Education, London, UK. Roberson, J. R., L. K. Fox, D. D. Hancock, and C. C. Gay Coagulase-positive Staphylococcus intramammary infections in dairy heifers. J. Dairy Sci. 77: Rupp, R., and D. Boichard Relationship of early first lactation somatic cell count with risk of subsequent first clinical mastitis. Livest. Prod. Sci. 62: Schepers, A. J., T.J.G.M. Lam, Y. H. Schukken, J.B.M. Wilmink, and W.J.A. Hanekamp Estimation of variance components for somatic cell counts to determine 98

113 Impact of Early Lactation Somatic Cell Count on Milk Yield over the First Lactation thresholds for uninfected quarters. J. Dairy Sci. 80: Snijders, T.A.B., and R. J. Bosker Multilevel analysis: an introduction to basic and advanced multilevel modelling. Sage, London Trinidad, P., S.C. Nickerson, and R.W. Adkinson Histopathology of staphylococcal mastitis in unbred dairy heifers. J. Dairy Sci. 73: Waage, S., S. Sviland, and S. A. Ødegaard Identification of risk factors for clinical mastitis in dairy heifers. J. Dairy Sci. 81: Waage, S., H. R. Skei, J. Rise, T. Rogdo, S. Sviland, and S. A. Ødegaard Outcome of clinical mastitis in dairy heifers assessed by re-examination of cases 1 mo after treatment. J. Dairy Sci. 83: Waage, S., S. A. Ødegaard, A. Lund, S. Brattgjerd, and T. Rothe Case-control study of risk factors for clinical mastitis in postpartum dairy heifers. J. Dairy Sci. 84:

114

115 Chapter 5.3 Association between Somatic Cell Count in Early Lactation and Culling of Dairy Heifers using Cox Frailty Models S. De Vliegher, 1 H. W. Barkema, 2 G. Opsomer, 1 A. de Kruif, 1 and L. Duchateau 3 1 Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 2 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Charlottetown, PEI, Canada 3 Department of Physiology, Biochemistry and Biometrics, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium Journal of Dairy Science, accepted for publication.

116 Chapter

117 Association between Early Lactation Somatic Cell Count and Culling of Dairy Heifers Abstract The association between somatic cell count (SCC) of dairy heifers in early lactation (SCCel) [measured between 5 and 14 d in milk (DIM)] and the culling hazard during the first lactation was studied using Cox frailty models. Udder health problems were the culling reason for 10% of the culled heifers in this study. For each unit increase in the log-transformed SCCel (LnSCCel), the culling hazard increased by 11% (Hazard ratio (HR) = 1.11). The strength of the association depended on 5 factors. Firstly, the association was stronger when SCCel was recorded after 10 DIM than at an earlier DIM. Secondly, the association was stronger if only culling events for udder disorders were considered (HR = 1.32) instead of all culling events (HR = 1.11). Furthermore, for each unit increase of test-day LnSCC after 14 DIM, modelled as a time-varying covariate, the culling hazard in the first lactation increased by 26% (HR = 1.26). Including LnSCC in the model already containing LnSCCel, reduced the estimate of LnSCCel slightly. Fourth, a higher test-day milk yield, modelled as a timevarying covariate, protected against culling and reduced the magnitude of the effect of LnSCCel as well when taken into account. Finally, the association between LnSCCel and culling was still present, although smaller in size, in the group of heifers with a second testday SCC 50,000 cells/ml. Key Words Cox frailty model, culling, dairy heifer, early lactation somatic cell count Abbreviation Key CNS = coagulase-negative staphylococci, DIM = days in milk, DHI = Dairy Herds Improvement, HR = hazard ratio, IMI = intramammary infection, LnSCC = natural logtransformed SCC, LnSCCel = natural log-transformed SCCel, MY = milk yield at test-day (kg), SCC = somatic cell count, SCCel = SCC in early lactation (measured between 5 and 14 DIM). 103

118 Chapter 5.3 Introduction Mastitis is an important culling reason in cows (Beaudeau et al., 1995; Barkema et al., 1998; Bascom and Young, 1998; Gröhn et al., 1998; Seegers et al., 1998; Rajala-Schultz and Gröhn, 1999a, 1999b; Neerhof et al., 2000). Acute mastitis in the first weeks of lactation has a significant effect on culling (Beaudeau et al., 1995; Rajala-Schultz and Gröhn, 1999a). Nearly 11% of heifers that were treated for clinical mastitis before calving or within the first 14 DIM were culled within 1 mo after treatment (Waage et al., 2000). The main culling reason for 96% of these heifers was mastitis. Cows with test-day somatic cell scores in the highest classes had almost a 3 times higher rate of culling compared with test-day scores on the average level (Samoré et al., 2003). Elevated SCCel in heifers, suggesting presence of intramammary infection (IMI) around calving, is associated with elevated test-day SCC and higher probabilities of test-day SCC >200,000 cells/ml (De Vliegher et al., 2004a), and with an increased probability of clinical mastitis during the first lactation (Rupp and Boichard, 2000). In addition, elevated SCCel is associated with lower milk production (Coffey et al., 1986; De Vliegher et al., accepted). Furthermore, clinical and subclinical mastitis early post partum have negative effects on reproductive performance (Barker et al., 1998; Schrick et al., 2001). As a result, dairy herds with a large number of heifers calving with infected udder quarters will suffer substantial economical losses. Moreover, elevated SCCel in heifers could be associated with an increased culling hazard during the first lactation, possibly to some extent because of the aforementioned effects. Disease (including mastitis) has direct and indirect effects on culling. The indirect effects may be reflected by, for instance, milk yield (MY) (Gröhn et al., 1997), as most diseases cause a decline in MY, either temporarily or longer lasting (Gröhn et al., 1998). Comparing models with and without MY helps to estimate the direct and indirect effects of mastitis in general on culling (Rajala-Schultz and Gröhn, 1999b). Survival analysis is often used to assess the effect of covariates that are measured only once (time-independent covariates, e.g. the effect of SCCel in this study). Some covariates, however, are changing over time (e.g. SCC and MY at the different test-days throughout lactation) and can easily be incorporated into the semi-parametric Cox models as time-depending or time-varying covariates (Gröhn, et al., 1997). The objectives of the study were twofold: 1) to examine the association between SCCel of heifers and culling during the first lactation while accounting for the day of assessment of SCCel and for the variability between herds, and 2) to determine what part of the effect of 104

119 Association between Early Lactation Somatic Cell Count and Culling of Dairy Heifers SCCel on culling acts indirectly through increased test-day SCC and decreased MY, by adjusting the models for test-day SCC and MY, modelled as time-varying covariates. Materials and Methods Data Set and Data Handling The DHI data used in the present study are described in detail elsewhere (De Vliegher et al., 2004a). In short, 14,234 heifers belonging to 3264 herds (enrolled in the Belgian DHI program; Flemish Cattle Breeding Association, Oosterzele, Belgium) that calved between January 1, 2000 and December 31, 2000, and of which the first test-day took place between 5 and 14 DIM, were followed up until 365 DIM, drying off or culling. The primary culling reason was recorded by the farmer (low milk production, reproductive disorders, udder disorders, foot/leg problems, behavioural problems, death, and non-specified reasons). Heifers for which it was unclear whether they were culled or not and heifers belonging to herds that stopped activities during the study period were omitted for further analyses, resulting in data from 13,835 heifers (97.2% of 14,234) belonging to 3192 herds. In total, the dataset contained 114,906 test-day records measured after 14 DIM. A heifer was considered to be culled or censored at its last available test-day. Somatic cell count was measured in composite milk samples collected from two successive milkings and was analyzed using the Fossomatic 5000 (Foss Electric, Hillerød, Denmark). Two datasets were created based on the full dataset containing data from 13,835 heifers. The structure of the 2 datasets is given in Table 1. Table 1. Hierarchy of the data. Dataset 1 Level n Average number per unit at next-higher level Range Full Herd 3192 Heifer 13, Test 114, Sub Herd 2755 Heifer Test 64, Full = test-day records between 15 and 365 DIM from 13,835 heifers; Sub = subset of test-day records from heifers with a second test-day SCC (recorded <75 DIM) of 50,000 cells/ml. 2 Average number of heifers per herd. 3 Average number of records per heifer. 4 Range of heifers per herd. 5 Range of tests per heifer. 105

120 Chapter 5.3 An event was defined as culling for all reasons combined in the first dataset (CULLALL full ; 3204 events), and as culling for udder disorders only in the second dataset (CULLUDD full ; 325 events). A similar approach was followed for a subset of data of 7596 heifers with a second test-day SCC (measured after 14 DIM and before 75 DIM) of 50,000 cells/ml. In the first dataset, an event was defined as culling for all reasons combined (CULLALL sub ; 1587 events), and in the second dataset an event was defined as culling for udder disorders only (CULLUDD sub ; 136 events). Statistical Analysis The association between the natural log-transformed SCCel (LnSCCel) and the culling hazard was studied by a semi-parametric Cox model (Cox, 1972). The Cox model was extended to a frailty model by introducing herd as a random effect to account for the clustering of heifers within herds (Duchateau and Janssens, 2004). The time to culling information for the k th heifer from herd j that was assessed at DIM equal to i was given by (t ijk, δ ijk ), where t ijk stands for the time of culling, and δ ijk was equal to 0 if the heifer was censored and to 1 in the case of culling. Two times 7 different models were fitted using CULLALL full and CULLUDD full, respectively, with breed incorporated as a fixed effect in all models. Both natural logtransformed SCC (LnSCC) and MY at test-day (>14 DIM) were included as time-varying covariates. In order to adjust for the fact that LnSCCel was not assessed at the same day after calving for each heifer, the Cox models were stratified according to DIM (5 to 14) on which LnSCCel was measured in the period called "early lactation". This means that for each DIM value (each stratum) another baseline function was assumed. In the first model, LnSCCel was introduced assuming a constant effect over the different DIM in early lactation (Model 1). In the second model, the same relationship between LnSCCel and the culling hazard was studied, but a different relationship between LnSCCel and the culling hazard was allowed according to DIM on which LnSCCel was recorded (Model 2). These 2 models were compared with each other based on the likelihood-ratio test. Models 3 and 4 contained LnSCC and MY, respectively. Models 5 and 6 contained both LnSCCel (as in Model 2) and LnSCC, and LnSCCel (as in Model 2) and MY, respectively, to evaluate the changes in LnSCCel when the time-varying covariates were included separately. Model 7 was the full model containing LnSCCel (as in Model 2), LnSCC, and MY. Two additional models (similar to Model 2) were fitted using CULLALL sub and CULLUDD sub, respectively, to study the association between the LnSCCel and the culling 106

121 Association between Early Lactation Somatic Cell Count and Culling of Dairy Heifers hazard in the group of heifers with a very low second test-day SCC ( 50,000 cells/ml), suggesting no udder health disorders present at that time. Hazard ratios (HR) with 95% confidence intervals were obtained for all covariates. All models were fitted with S-Plus 6.0 for Windows (Insightful Corp, Seattle, US). Results Descriptive Analysis Geometric mean SCCel of the 13,835 heifers was 111,000 cells/ml, ranging from 5,000 to 25,000,000 cells/ml. In total, 3204 heifers (23.2%) were culled during their first lactation (Table 2). The 2 main specified reasons of culling were reproductive disorders and low MY. In total, 325 heifers (2.3% of all heifers and 10.1% of the cows culled in first lactation) were culled for udder health disorders as the primary reason. No reason was specified in 40.5% of all cullings. Table 2. Overview of the culling reasons of the 3204 culled heifers out of the total study population of 13,835 Flemish (Belgium) heifers. Reason for culling n % of all heifers % of culled heifers Non-specified reason Reproductive disorders Low milk yield Udder disorders Death Lameness and foot/leg defects Behaviour Total Heifers with a higher SCCel were more at risk of being culled during lactation (Figure 1). For instance, at 100 DIM, 3% of the heifers with a SCCel 50,000 cells/ml were culled, whereas 7% of heifers with a SCCel >1,000,000 cells/ml were culled. At 200 DIM, this was 7 and 13%, respectively. The same trends are present in the heifers that were culled for udder disorders (Figure 2), but the differences between the SCCel-levels were smaller. Heifers with a SCCel >500,000 cells/ml were culled earlier in lactation compared with the other heifers (Figure 2). Both Kaplan-Meier graphs present survival until 305 DIM as afterwards too few animals remained at risk. 107

122 Chapter Survival (%) Figure 1. Kaplan-Meier graph of culling of heifers for all reasons (until 305 DIM) with a SCC in early lactation (SCCel, measured between 5 and 14 DIM, x 1000 cells/ml) of 0 to 50 ( ), 51 to 200 ( ), 201 to 500 (+), 501 to 1000 ( ), and >1000 cells/ml ( ). DIM Survival (%) Figure 2. Kaplan-Meier graph of culling of heifers for all reasons udder health disorders (until 305 DIM) with a SCC in early lactation (SCCel, measured between 5 and 14 DIM, x 1000 cells/ml) of 0 to 50 ( ), 51 to 200 ( ), 201 to 500 (+), 501 to 1000 ( ), and >1000 cells/ml ( ). DIM Cox Models For each unit increase in the LnSCCel, the culling hazard of 13,835 dairy heifers in the first lactation increased by 11% (HR = 1.11; 95% CI: ) (Table 3, Model 1). The model allowing for a different association between LnSCCel and the culling hazard per DIM in early lactation (Table 3, Model 2) was significantly better (likelihood ratio test; Chi-square = 41.2, df = 9, P < 0.001). An increase in LnSCCel was associated with an increased culling 108

123 Table 3. Association between log-transformed SCC in early lactation (LnSCCel, measured between 5 and 14 DIM, x 1000 cells/ml), logtransformed test-day SCC (LnSCC, x 1000 cells/ml) and test-day milk yield (MY), and the hazard of being culled for all reasons in 13,835 dairy heifers (CULLALL full ). Model 1 Model 2 Model 3 Model 4 Variable HR 1 95% CI 2 HR 1 95% CI 2 HR 1 95% CI 2 HR 1 95% CI 2 LnSCCel LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM MY LnSCC Breed Breed Breed RHF BHF

124 Table 3. (continued). Model 5 Model 6 Model 7 Variable HR 1 95% CI 2 HR 1 95% CI 2 HR 1 95% CI 2 LnSCCel LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM MY LnSCC Breed Breed Breed RHF BHF HR = Hazard Ratio. 2 95% confidence interval around HR. 3 Assuming a constant effect of LnSCCel over the different DIM in early lactation. 4 Allowing different effects of LnSCCel over the different DIM in early lactation. 5 Time-varying covariate. 6 Belgian White-Blue double-purpose heifers and heifers of unknown breed. 7 Meuse-Rhine-Yssel, Red, Red and White, Jersey. 8 Red Holstein-Friesian. 9 Black Holstein-Friesian.

125 Table 4. Association between log-transformed SCC in early lactation (LnSCCel, measured between 5 and 14 DIM, x 1000 cells/ml), logtransformed test-day SCC (LnSCC, x 1000 cells/ml) and test-day milk yield (MY), and the hazard of being culled for udder health reasons in 13,835 dairy heifers (CULLUDD full ). Model 1 Model 2 Model 3 Model 4 Variable HR 1 95% CI 2 HR 1 95% CI 2 HR 1 95% CI 2 HR 1 95% CI 2 LnSCCel LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM MY LnSCC Breed Breed Breed RHF BHF

126 Table 4. (continued). Model 5 Model 6 Model 7 Variable HR 1 95% CI 2 HR 1 95% CI 2 HR 1 95% CI 2 LnSCCel LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM MY LnSCC Breed Breed Breed RHF BHF HR = Hazard Ratio. 2 95% confidence interval around HR. 3 Assuming a constant effect of LnSCCel over the different DIM in early lactation. 4 Allowing different effects of LnSCCel over the different DIM in early lactation. 5 Time-varying covariate. 6 Belgian White-Blue double-purpose heifers and heifers of unknown breed. 7 Meuse-Rhine-Yssel, Red, Red and White, Jersey. 8 Red Holstein-Friesian. 9 Black Holstein-Friesian.

127 Association between Early Lactation Somatic Cell Count and Culling of Dairy Heifers hazard, except for DIM 5. The association was significant from DIM 10 and onwards (except for DIM 11) (Table 3, Model 2). The association between LnSCCel and culling was stronger (a HR of 1.32 for each unit increase in LnSCCel) if only culling for udder disorders (CULLUDD full ) was considered (Table 4, Model 1). Allowing different effects per DIM in early lactation (Table 4, Model 2), did not significantly improve the model (Table 4, Model 2 versus 1; likelihood ratio test; Chi-square = 12.8, df = 9, P = 0.17), but all HR were >1. Black Holstein-Friesian heifers were culled less frequently compared with Belgian White-Blue double-purpose heifers and heifers of unknown breed (Tables 3 and 4, Models 1 and 2). Log-transformed SCC was significantly related to the culling hazard of dairy heifers with a HR of 1.26 for each unit increase in LnSCC (Table 3, Model 3). The HR increased to 1.80 when only culling for udder disorders specifically (CULLUDD full ) was considered (Table 4, Model 3). Introducing LnSCC into the models that already comprised LnSCCel, reduced the estimate of LnSCCel at every DIM in early lactation (Tables 3 and 4, Model 5 versus Model 2). The reduction was larger when studying the association in heifers that were culled for udder problems. Higher MY protected heifers against culling (Tables 3 and 4, Model 4). The magnitude of LnSCCel at the different DIM in early lactation was slightly reduced when MY was taken into account (Tables 3 and 4, Model 6 versus 2). The changes were smaller compared with the changes due to introducing LnSCC. In addition, incorporating MY into the models took away the breed effect: Black Holstein-Friesian heifers were no longer protected from culling (e.g., Tables 3 and 4, Model 4 versus Model 2). The most elaborate model presents the estimates of LnSCCel adjusted for both LnSCC and MY (Tables 3 and 4, Model 7). Studying the association between LnSCCel and culling in heifers with a second test-day SCC 50,000 cells/ml (CULLALL sub and CULLUDD sub ) revealed that, although the association was smaller, an elevated SCCel was, in general, still associated with a higher culling hazard, especially for SCCel measured in the second part of the period called "early lactation" (Table 5). Discussion In this study, Cox frailty models were used to study the association between SCCel and the culling of heifers belonging to different herds. The influence of test-day SCC and MY on the magnitude of SCCel was assessed. The results of this study add further knowledge on the negative impact of heifer mastitis reflected by elevated SCCel as an indicator of subclinical 113

128 Chapter 5.3 mastitis around calving. Reproduction was the primary culling reason in our study, whereas production was second and mastitis third, which corresponds with the findings from Bascom and Young (1998). Table 5. Association between log-transformed SCC in early lactation (LnSCCel, measured between 5 and 14 DIM, x 1000 cells/ml), and the hazard of being culled for all reasons (CULLALL sub ), and for udder health reasons (CULLUDD sub ) in 7596 dairy heifers. CULLALL sub CULLUDD sub Variable HR 1 95% CI 2 HR 1 95% CI 2 LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM LnSCCel DIM Breed Breed Breed RHF BHF Hazard Ratio. 2 95% confidence interval around HR. 3 Allowing different effects of LnSCCel over the different DIM in early lactation. 4 Belgian White-Blue double-purpose heifers and heifers of unknown breed. 5 Meuse- Rhine-Yssel, Red, Red and White, Jersey. 6 Red Holstein-Friesian. 7 Black Holstein-Friesian. Because SCCel decreases substantially in the first 2 wk after calving (Dohoo, 1993; Laevens et al., 1997; De Vliegher et al., 2001; De Vliegher et al., 2004a; De Vliegher et al., 2004b), models that stratified for day of assessment of SCCel were fitted. This implied that no conclusion on a DIM effect on the culling hazard could be drawn, but as no biological relevant effect could be expected, this was not a problem. Based on our previous findings, an effect was expected related to the infection dynamics during the early lactation period as discussed earlier (De Vliegher et. al., 2004a; De Vliegher et al., accepted). The most prevalent group of mastitis pathogens associated with IMI in heifers at parturition are the coagulasenegative staphylococci that are transient in nature (Oliver and Mitchell, 1983), probably because they are colonizing the teat canal rather than the mammary gland and are washed out during milking. In addition, a high rate of spontaneous IMI elimination occurs (Oliver and 114

129 Association between Early Lactation Somatic Cell Count and Culling of Dairy Heifers Jayarao, 1987). Staphylococcus aureus, on the other hand, can also be associated with IMI in heifers at calving (Fox et al., 1995), but tends to persist in lactation (Roberson et al., 1994). Some of the factors that needed to be evaluated or adjusted for (e.g. test-day SCC and test-day MY) change over the course of lactation. In order to assess their immediate effect, they were introduced as time-varying covariates, because summary measures (e.g. lactation average SCC or MY) are not able to determine the effect of a covariate throughout lactation (Gröhn et al., 1997). Furthermore, the clustering effects due to the fact that heifers belong to different herds had to be taken into account by introducing a random herd effect in the Cox model, rather than including herd as a fixed effect because the individual farm is not of interest by itself (Duchateau and Janssen, 2004). Considering only the heifers that were culled for udder health reasons (CULLUDD) versus all culled heifers (CULLALL) increased the magnitude of association between SCCel and SCC, and the culling hazard. This is comprehensible as the effect is not diluted by the other reasons why heifers are culled. We wanted to present both approaches because the DHI program only allows farmers to identify 1 single culling reason per heifer, whereas farmers usually consider many factors when deciding to cull an animal (Bascom and Young, 1998). In addition, for 40.5% of the culled heifers no specific reason was available, although some of them were probably culled because of udder health problems. Still, even when considering all culling reasons, an elevated SCCel predicted a higher culling hazard, confirming the negative economic consequences of heifer mastitis at freshening, even though the HR was only significantly >1 when recorded after 9 DIM. Confidence intervals around the HR were wider for heifers culled for udder health disorders (CULLUDD) compared with all culled heifers (CULLALL). As a consequence, fewer HRs significantly differed from 1, which also occurred in the analyses in the subsets of data (CULLALL sub and CULLUDD sub ). This is comprehensible as the power in survival analysis is a function of the number of events and not of the number of observations (Freedman, 1982). Beaudeau et al. (1995) included the potential 305-d mature equivalent milk production rather than actual MY in the survival models when studying the effect of disease on culling in French dairy cows. This approach avoided inclusion of part of the impact of disease on culling through their effect on cumulative milk production. However, high yielding cows, even if they are diseased, are more likely to be kept in the herd (Gröhn et al., 1998). Comparing models with and without MY, can therefore help to estimate the direct and indirect effects of mastitis or disease in general on culling (Rajala-Schultz and Gröhn, 1999b). Hence, MY was included in the models to find out whether this changed the magnitude of the 115

130 Chapter 5.3 effect of SCCel on culling. Acute mastitis in the first mo of lactation had a significant effect on culling throughout lactation (Rajala-Schultz and Gröhn, 1999a), but adding MY to the model, in general, reduced the effect (Rajala-Schultz and Gröhn, 1999b). The same was true in our study, indicating that a small part of the effect of an elevated SCCel on test-day SCC was mediated through MY. Moreover, introducing MY in the models influenced the breed effect. In the models without MY, Black Holstein-Friesian were protected from culling, but this effect was due to their higher MY. According to Rajala-Schultz and Gröhn (1999b), the influence of MY on the culling decision also depends on the lactation stage. We, however, did not distinguish whether the effect differed with differing stages of lactation, as this was not the primary aim of the study. Test-day SCC was even more associated with culling than was SCCel. Fitting a model with both LnSCCel and LnSCC showed that part of the LnSCCel effect acted through the associated test-day SCC. This was not unexpected as an elevated SCCel increases the odds on elevated test-day SCC (De Vliegher et al., 2004a). Fitting a model containing LnSCCel, LnSCC, and MY results in estimates for LnSCCel that take into account the SCC effect and adjusts for the protective MY effect. Elevated SCCel is associated with elevated test-day SCC and lower MY, thus part of the association between SCCel and culling was mediated through SCC and MY. Heifers with an excellent udder health at the second test-day but with an elevated SCCel were still more at risk of being culled in their first lactation compared with heifers with an equally low second-test-day SCC but a lower SCCel. Most probably this finding is related to the fact that the latter heifers will have fewer test-day SCC >200,000 cells/ml (De Vliegher et al., 2004a) and will out-produce the heifers with the higher SCCel (De Vliegher et al., accepted). This suggests that prevention against elevated SCCel, especially in the second part of the early lactation period as we defined it, should be preferred over treating an elevated SCCel. Additionally, some of the heifers with (sub)clinical mastitis in early lactation will lose milk production in the affected quarter(s) and will consequently have a low second test-day SCC. In this dataset, no heifers were culled before 32 DIM. This does not reflect the actual situation and is caused by the way the data were collected and handled: only heifers of which the first test-day SCC was measured between 5 and 14 DIM in the year 2000 (n = 14,766) and of which additional test-day SCC were available (n = 14,234; selection procedure outlined in De Vliegher et al., 2004a) were used. Therefore, no heifers that were culled within the first 116

131 Association between Early Lactation Somatic Cell Count and Culling of Dairy Heifers mo of lactation were present in the dataset. This has resulted in an underestimation of the effect of peripartum udder health problems on the hazard of culling. The culling decision process is complex and very farmer-, herd- and time-specific. This is especially true under a quota system (as implemented in Belgium), forcing the farmer to take culling decisions depending on the level of expected fulfillment of the quota. This study ignored some of the points in the complex decision of the farmer as the required herd and time-level information was not available. Economic calculations and implementation of new preventive measures would in addition require cow-specific values. Future studies should collect this information and should combine it with data from other studies looking at the association between udder health in early lactation in heifers and udder health (both clinical and subclinical mastitis), production, and fertility in the first and following lactations. Conclusions Heifers with elevated SCCel were at an increased risk of being culled during first lactation. Part of the effect was associated with the consequential elevation of test-day SCC and suppression of test-day MY. High yielding heifers were, on average, protected against culling, even if their SCCel was elevated. The association between LnSCCel and culling was still present, although smaller in size, in heifers with a second test-day SCC 50,000 cells/ml, suggesting that prevention rather than cure of an elevated SCCel is needed. Acknowledgments The authors would like to thank E. De Mûelenaere and the Flemish Cattle Breeding Association (Oosterzele, Belgium) for providing us with the milk-recording data and to Elanco, Belgium, for supporting this study financially. The valuable help from H. Stryhn (AVC, Charlottetown, Canada) was highly appreciated. References Barkema, H. W., Y. H. Schukken, T.J.G.M. Lam, M. L. Beiboer, H. Wilmink, G. Benedictus, and A. Brand Incidence of clinical mastitis in dairy herds grouped in 3 categories of bulk milk somatic cell counts. J. Dairy Sci. 81:

132 Chapter 5.3 Barker, A. R., F. N. Schrick, M. J. Lewis, H. H. Dowlen, and S. P. Oliver Influence of clinical mastitis during early lactation on reproductive performance of Jersey cows. J. Dairy Sci. 81: Bascom, S. S., and A. J. Young A summary of the reasons why farmers cull cows. J. Dairy Sci. 81: Beaudeau, F., V. Ducrocq, C. Fourichon, and H. Seegers Effect of disease on length of productive life of French Holstein dairy cows assessed by survival analysis. J. Dairy Sci. 78: Coffey, E. M., W. E. Vinson, and R. E. Pearson Somatic cell counts and infection rates for cows of varying somatic cell count in initial test of first lactation. J. Dairy Sci. 69: Cox, D. R Regression models and life-tables (with discussion). J. Royal Stat. Soc. B. 34: De Vliegher, S., H. Laevens, G. Opsomer, E. De Mûelenaere, and A. de Kruif Somatic cell counts in dairy heifers during early lactation. Flem. Vet. J. 70: De Vliegher, S., H. W. Barkema, H. Stryhn, G. Opsomer, and A. de Kruif. 2004a. Impact of early lactation somatic cell count in heifers on somatic cell counts over the first lactation. J. Dairy Sci. 87: De Vliegher, S., H. Laevens, H. W. Barkema, I. Dohoo, H. Stryhn, G. Opsomer, and A. de Kruif. 2004b. Management practices and heifer characteristics associated with early lactation somatic cell count of Belgian dairy heifers. J. Dairy Sci. 87: De Vliegher, S., H. W. Barkema, H. Stryhn, G. Opsomer, and A. de Kruif. Impact of early lactation somatic cell count in heifers on milk yield over the first lactation. J. Dairy Sci., accepted. Dohoo, I. R An evaluation of the validity of individual cow somatic cell counts from cows in early lactation. Prev. Vet. Med. 16: Duchateau, L., and P. Janssens Penalized partial likelihood for frailties and smoothing splines in time to first insemination models for dairy cows. Biometrics. 60: Fox, L. K., S. T. Chester, J. W. Hallberg, S. C. Nickerson, J. W. Pankey, and L. D. Weaver Survey of intramammary infections in dairy heifers at breeding age and first parturition. J. Dairy Sci. 78: Freedman, L. S Tables of the number of patients required in clinical trials using the logrank test. Stat. Med. 1:

133 Association between Early Lactation Somatic Cell Count and Culling of Dairy Heifers Gröhn, Y. T., V. Ducrocq, and J. A. Hertl Modelling the effect of a disease on culling: an illustration of the use of time-dependent covariates in survival analysis. J. Dairy Sci. 80: Gröhn, Y. T., S. W. Eicker, V. Ducrocq, and J. A. Hertl Effect of diseases on the culling of Holstein dairy cows in New York State. J. Dairy Sci. 81: Laevens, H., H. Deluyker, Y. H. Schukken, L. De Meulemeester, R. Vandermeersch, E. De Mûelenaere, and A. de Kruif Influence of parity and stage of lactation on the somatic cell count in bacteriologically negative cows. J. Dairy Sci. 80: Neerhof, H. J., P. Madsen, V. P. Ducrocq, A. R. Vollema, J. Jensen, and I. R. Korsgaard Relationship between mastitis and functional longevity in Danish Black and White dairy cattle estimated using survival analysis. J. Dairy Sci. 83: Oliver, S. P., and B. A. Mitchell Intramammary infections in primigravid heifers near parturition. J. Dairy Sci. 66: Oliver, S. P., and B. M. Jayarao Coagulase-negative staphylococcal intramammary infections in cows and heifers during the nonlactating and periparturient periods. J. Vet. Med. B. 44: Rajala-Schultz, P. J., and Y. T. Gröhn. 1999a. Culling of dairy cows. Part I. Effects of diseases on culling in Finnish Ayrshire cows. Prev. Vet. Med. 41: Rajala-Schultz, P. J., and Y. T. Gröhn. 1999b. Culling of dairy cows. Part III. Effects of diseases, pregnancy status and milk yield on culling in Finnish Ayrshire cows. Prev. Vet. Med. 41: Roberson, J. R., L. K. Fox, D. D. Hancock, and C. C. Gay Coagulase-positive Staphylococcus intramammary infections in dairy heifers. J. Dairy Sci. 77: Rupp, R., and D. Boichard Relationship of early first lactation somatic cell count with risk of subsequent first clinical mastitis. Livest. Prod. Sci. 62: Samoré, A. B., M. del P. Schneider, F. Canavesi, A. Bagnato, and A. F. Groen Relationship between somatic cell count and functional longevity assessed using survival analysis in Italian Holstein-Friesian cows. Livest. Prod. Sci. 2003: Schrick, F. N., M. E. Hockett, A. M. Saxton, M. J. Lewis, H. H. Dowlen, and S. P. Oliver Influence of subclinical mastitis during early lactation on reproductive parameters. J. Dairy Sci. 84: Seegers, H., F. Beaudeau, C. Fourichon, and N. Bareille Reasons for culling in French Holstein cows. Prev. Vet. Med. 36:

134 Chapter 5.3 Waage, S., H. R. Skei, J. Rise, T. Rogdo, S. Sviland, and S. A. Ødegaard Outcome of clinical mastitis in dairy heifers assessed by re-examination of cases 1 mo after treatment. J. Dairy Sci. 83:

135

136

137 Chapter 6 Management Practices and Heifer Characteristics Associated with Early Lactation Somatic Cell Count of Belgian Dairy Heifers S. De Vliegher, 1 H. Laevens, 2 H. W. Barkema, 3 I. R. Dohoo, 3 H. Stryhn, 3 G. Opsomer, 1 and A. de Kruif 1 1 Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 2 Coordination Centre for Veterinary Diagnostics, Veterinary and Agrochemical Research Centre, Brussels, Belgium 3 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Charlottetown, PEI, Canada Journal of Dairy Science, 2004, 87:

138 Chapter 6 124

139 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count Abstract Associations between somatic cell counts (SCC) from heifers between 5 and 14 days in milk (DIM), and both herd management practices and heifer characteristics were studied for 1912 heifers in 159 dairy herds in Flanders (Belgium). In higher producing herds and in herds with an average calving age of heifers >27 mo, SCC in early lactation (SCCel) of heifers was lower than in less producing herds or in herds with an average calving age 27 mo. Heifers raised in herds with a higher bulk milk somatic cell count had higher SCCel. In herds where heifers calved on slatted floors, heifers had lower SCCel compared to herds were heifers calved on not-slatted floors. A significant difference in SCCel was observed between provinces. At the heifer level, SCCel decreased with increasing DIM. Heifers calving in April to June on average had higher SCCel compared to those calving in the other months of the year. In the multilevel regression models, nearly all the variation of SCCel resided at the heifer level, indicating that preventive measures against udder health problems in freshening dairy heifers should, in the short term, focus more on factors that vary between heifers, than on factors that vary between herds. However, on the long term, the need to identify new and to implement known herd level strategies is important. Key Words Dairy heifer, multilevel, risk factor, somatic cell count Abbreviation Key BMSCC = Bulk milk SCC, DIM = days in milk, DHI = Dairy Herd Improvement, IMI = intramammary infection, HSCCel = Herd geometric mean SCCel, LnSCCel = natural logtransformed SCCel, SCC = somatic cell count, SCCel = SCC in early lactation (between 5 and 14 DIM). 125

140 Chapter 6 Introduction Mastitis prevention programs, in general, focus on cows and not on heifers, although the latter are the future of the dairy herd. The goal of having all heifers in the herd freshening with a healthy udder, capable of producing high quality milk, is often not reached. Many studies have reported a high prevalence of intramammary infection (IMI) in non-lactating and freshly calved heifers with mainly coagulase-negative staphylococci, and Staphylococcus aureus (Oliver and Mitchell, 1983; Trinidad et al., 1990; Pankey et al., 1991; Myllys, 1995; Fox et al., 1995; Nickerson et al., 1995). In addition, the incidence rate of clinical mastitis in heifers is very high during the first 14 d of lactation (Barkema et al., 1998a). Intramammary infections around parturition, either clinical or subclinical, will lead to considerable financial losses for the farmer and should therefore be avoided. Only a limited number of studies have looked at this problem leaving a lot of questions on how to prevent IMI in heifers pre-, peri-, and postpartum. To reduce exposure to environmental pathogens, pregnant and peripartal heifers should be housed in a clean and dry environment (Shearer and Harmon, 1993). A poor hygiene of the calving area was associated with an increased prevalence of elevated SCC in heifers (Bareille et al., 2000). Supplementary preventive measures to control heifer mastitis include individual housing of pre-weaned calves to prevent suckling, fly control, supplementation of the diet with minerals, and segregation of pregnant heifers from dry cows (Shearer and Harmon, 1993). Frequent teat disinfection prior to parturition in primigravid dairy heifers did not improve udder health (Edinger et al., 2000; Waage et al., 2001). Vaccination of heifers was successful in 1 study (Giraudo et al., 1997) and not successful in another (Tenhagen et al., 2001) in reducing the incidence of S. aureus IMI. Treating pregnant heifers intramammarily with dry or lactating cow antibiotics reduced the prevalence of IMI at calving and increased the milk production in their first lactation (Trinidad et al., 1990; Oliver et al., 1992; Owens et al., 1994; Owens et al., 2001) and is economically beneficial (Oliver et al., 2003). The use of antibiotics in heifers before calving should, however, be carefully promoted in the light of possible residue and resistance problems. Knowledge of more risk factors associated with high prevalence of IMI around parturition should be used to reach a more acceptable level of udder health in dairy heifers. Moreover, understanding at what level, herd versus individual heifer, preventive measures should be applied, is essential in tackling the problem. The objective of this study was to evaluate differences in herd management practices and heifer characteristics associated with SCCel of heifers. The contribution of herd and 126

141 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count heifer to the total variance of the SCCel was assessed using multilevel, multivariable, linear regression models. Materials and Methods Heifer Characteristics Four-weekly milk recordings for 1999 and 2000 were available for all cows and heifers at least 5 DIM from the herds enrolled in the Dairy Herd Improvement program (Flemish Cattle Breeding Association, Oosterzele, Belgium), and included SCC, milk production at day of SCC measurement (kg), breed, DIM, and date of measurement. The last variable was categorized into 4 calving seasons: January March, April June, July September, and October December. Somatic cell counts were measured based on composite milk samples collected from 2 successive milkings and were analyzed with the Fossomatic 5000 (Foss Electric, Hillerød, Denmark). To approximate the normal distribution, a natural logarithmic transformation of SCCel (LnSCCel) was used. Herd Selection and Management Practices Herds were selected from the 3287 dairy herds in Flanders (Belgium) participating in year 2000 in the DHI program that had at least 1 heifer for which the first SCC was measured between 5 and 14 DIM. From these herds, only those that had at least 10 such SCCel were selected. A total of 254 herds matched this criterion and of these 20 were dropped from the study because inspection of their data for the years 1999 and 2000 indicated that their SCCelstatus had changed. To the 234 dairy herds a questionnaire was sent in the spring of 2001 The questionnaire was tested and fine-tuned on 6 dairy herds not participating in the study. It consisted of 117 main questions concerning general management, milking and dry cow management, and management of calves and heifers, all concerning the year A summary is presented in Table 1. All data were entered in an electronic spreadsheet program (Excel 2000, Microsoft Corporation), and were checked for typing and coding errors. Geometric mean SCCel for all selected heifers in a herd (HSCCel) were calculated for the 159 herds (68% of 234) that returned the completed questionnaire. The overall data structure of the herds answering the questionnaire was as follows: SCCel from 1912 heifers (lowest level of data hierarchy) were available, belonging to 159 herds (average of 12 heifers per herd, range 10 33) in 5 provinces (highest level of data hierarchy). Each province contained on average 32 herds (range 6 60). 127

142 Chapter 6 Table 1. Overview of herd management practices during the year 2000 collected through a mailed questionnaire on 159 dairy herds in Flanders (Belgium). General management Property size, herd size, main breed, other activities on the farm, disinfection procedures, use of clothes for visitors, herd health client, record keeping of disease, AAmilk label 1, TQCM label 2, geometric mean bulk milk SCC, quota, average herd net gain 3, fly control in barn. Management of lactating cows Housing, bedding, cleaning frequency, summer grazing, milking parlour, milking frequency, number of cows milked, treatment of clinical mastitis. Management of dry cows Housing, cleaning frequency, dry cow therapy, mineral supplementation, diet. Management of calves Suckling with mothers, colostrum given in first day, feeding of mastitic milk, feeding high somatic cell count or residue milk, use of buckets for drinking, weaning age, vaccination against respiratory disease, outbreak of respiratory disease, sucking each other, housing, injecting vitamins or antibiotics at birth, buying calves. Management around calving Main calving season for heifers and cows, calving pen, bedding of calving area, cleaning frequency of calving area, time spent by calved animal in calving pen, use of calving pen as sick bay, AI or bull, calving on pasture, percentage of heifers with calving difficulties. Management of heifers Housing, bedding, cleaning frequency, sucking each other, average age at moment of conception, average age at calving, beef bull used, checking for udder problems, milking before parturition, teat disinfection before parturition, fly control on pasture, excessive edema at calving, summer mastitis, deworming program, contact with lactating or dry cows, use of oxytocin for milking, mineral supplementation, diet before, during and after pasture period. 1 Quality label for drinking milk introduced in Belgium in the 1950 s. See text for details. 2 Total Quality Care Milk, recently introduced quality label for Belgian dairy herds with requirements regarding hygiene, animal welfare, animal health, and housing. 3 Average fat and protein corrected production per mo for the whole herd, allowing ranking of herds. Statistical Analysis Prior to statistical analysis, observations were explored and were checked for unlikely values. No data were excluded for this reason. The regression model building process involved several steps, as presented in a flow chart (Figure 1). Initially, unconditional associations were tested between the continuous dependent variable on herd level, HSCCel, and all herd management practices (n = 117, Table 1) as independent variables. An identical procedure was followed for LnSCCel per heifer as dependent variable and the 4 heifer characteristics, DIM, milk production, breed, and calving season, as independent variables. Independent-samples t-tests or one-way ANOVA for categorical independent variables, and 128

143 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count linear regression for continuous independent variables were used. Statistical significance in this step was assessed at P < Reclassification was done if necessary for categorical variables with some low frequency classes, based on biological reasoning. Homogeneity-ofvariances and deviances from normality were constantly checked using Levene statistics, and Kolmogorov-Smirnoff test with Lilliefors correction. No problems were detected. To detect multicollinearity, Pearson correlation coefficients were calculated among the significant (n = 28) management variables in a second step. The same was done among the 3 significant heifer characteristics. If 2 variables had a correlation coefficient 0.6, only 1 was selected for further analysis based on either the strongest association with the dependent variable or for biological reasons. Two herd-level variables and no heifer-level variables were discarded for this reason. The herd-level variable calving season of heifers was, although significant, not withheld for further analysis, as this variable was also available at heifer-level (Tables 2a, b, c). In a third step, a multivariable linear regression model was built with the 25 remaining management practice variables as independent variables, and with HSCCel as the dependent variable. Non-significant variables were removed using backwards elimination at P < Variance inflation factors were computed to detect the presence of multicollinearity. Because all variance inflation factors were close to 1 and as a value >10 is taken as an indication of multicollinearity (Neter et al., 1996), it was concluded that the latter was not present. The third step reduced the number of significant management practices to 7 to be included in the multilevel, multivariable, linear regression model. SPSS 10.0 for Windows (SPSS Inc., Chicago, Illinois, USA) was used for all previous analyses. In the final step, a multilevel model was fit for the continuous dependent variable LnSCCel per heifer using the restricted generalized iterative least-squares (RIGLS) algorithm in MLwiN 1.1 (Goldstein et al., 1998). All previously selected independent variables on the herd level (n = 7) and on the heifer level (n = 3) were added as fixed effects. As the herd-level variable "average herd net gain" and the heifer-level variable "production at the day of SCCel measurement" were highly correlated it was decided to leave the latter out of the analysis as it was also correlated with DIM. To evaluate the proportion of variance occurring at the 3 levels of the data hierarchy, a three-level null model (intercept only) was fit with "province", "herd", and "heifer" as random effects. In addition the variance distribution was calculated for the two-level null model (intercept only) by including "herd" and "heifer" as random effects. For further modelling, "province" was included as a fixed effect. The continuous independent variables were centred by subtracting their overall mean, providing better interpretations and 129

144 Chapter 6 Herd-level variables (HSCCel) n = 117 (Table 1) Heifer-level variables (LnSCCel) n = 4 Unconditional associations (P < 0.15) n = 28 n = 3 Evaluation of collinearity n = 26 (Table 2) 1 Multivariable linear regression (P < 0.1) n = 7 (Table 3) n = 3 2 Multilevel, multivariable linear regression (P < 0.05) n = 7 (4 herd level + 3 heifer level) (Table 4) 1 The herd-level variable "calving season of heifers" was not withheld for further analysis, as this variable was also available at heifer level. 2 The heifer-level variable "production at the day of SCCel measurement" was withdrawn as it was highly correlated with the herd-level variable "average herd net gain" and DIM. Figure 1. Flow chart of variable reduction through different steps in the statistical analysis. better numerical stability of estimates (Dohoo et al., 2001). All first-order interactions were tested and were removed when non-significant (Wald s test, P > 0.05), followed by nonsignificant main effects. The adequacy of the final model was tested by examining normal probability plots of residuals and plots of residuals versus predicted values to check whether 130

145 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count the assumptions of normality and homogeneity-of-variance had been fulfilled. No patterns indicating heteroscedasticity were revealed. Results Description of the Data In 1999 and 2000, SCCel between 5 and 14 DIM were available from 12,994 and 14,766 heifers from 3221 and 3287 herds, respectively. Geometric mean SCCel for 1999 was 110,000 cells/ml (range ,817,000 cells/ml). More details on the SCCel of the year 1999 are presented elsewhere (De Vliegher et al., 2001). Geometric mean SCCel of the 14,766 samples in 2000 was 104,000 cells/ml (range ,088,000 cells/ml). Geometric mean SCCel for the 1912 heifers of the responding 159 herds was 92,000 cells/ml (range ,622,000 cells/ml). The interquartile range was 163,000 cells/ml. At the herd level, HSCCel ranged from 24,000 cells/ml to 256,000 cells/ml. Sixty percent of the 1912 heifers were Black Holstein-Friesian, 27% were Red Holstein Friesian, and 7% were Red and White Holstein crossbreds. Average milk quota on the 159 herds was 537,067 kg milk (± 19,406), and the average milking herd consisted of 75 cows (± 2.5) (range ). Average bulk milk SCC (BMSCC) for 2000 was 175,000 cells/ml (range 73, ,000 cells/ml). Half of the 159 farms only had dairy cows, 38% also had beef cows, and 8% also farmed pigs. Nearly 80% of all farms had an AA-milk label, a label that was introduced in Belgium in the 1950 s to stimulate milk consumption. In addition to the official analyses that are conducted for all producers, specific AA-milk regulations provide for additional counts of coliforms twice a month. A number of other criteria (bacteria count, BMSCC) are also more stringent. By meeting these higher quality requirements farmers are paid a premium price for their milk. Herd-level Analysis In the first step of the analysis, a first reduction of all herd data, based on unconditional associations and highly correlated variables, revealed 26 herd management practices significantly (P < 0.15) associated with HSCCel (Tables 2a, b, c). 131

146 Chapter 6 Table 2a. Significant (P < 0.15) unconditional associations between general herd management practices (categorical) and the geometric herd mean SCCel of all heifers between 5 and 14 DIM per herd (HSCCel, x 1000 cells/ml milk) for 159 high yielding dairy herds in Flanders (Belgium). Categorical variables N HSCCel P Province <0.001 Antwerpen Vlaams-Brabant West-Vlaanderen Oost-Vlaanderen Limburg Milking machine type Low line pipeline system Other types AA milk label Yes No Floor type of calving place or calving pen Not-slatted Slatted Time spent in calving pen after calving by cows and heifers Until just after calving/1 d A few days Calving pen used also as sick bay Yes No/no separate calving pen on farm present Type of barn Free stall with slatted floors Other types Floor bedding stanchion place or cubicle Straw or sawdust Rubber or concrete Mastitis treatment By farmer Veterinarian involved Fly problem during summer in the milking barn Similar to other years More/less than other years Injecting newborn calves With antibiotics only/both with vitamins and antibiotics/ not injected With vitamins only Quality label for drinking milk introduced in Belgium in the 1950 s. See text for details. 132

147 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count Table 2b. Significant (P < 0.15) unconditional associations between heifer management practices (categorical) and the geometric herd mean SCCel of all heifers between 5 and 14 DIM per herd (HSCCel, x 1000 cells/ml milk) for 159 high yielding dairy herds in Flanders (Belgium). Categorical variables N HSCCel P Heifers pregnant from AI Bull, both AI and bull, bull after AI failed Average age of heifers at calving Between 23 and 27 mo Older than 27 mo Clipping of heifers udders Around calving A long period before calving/no clipping at all Fly control in heifers on pasture One time pour-on/more than 1 time pour-on/using two ear tags No fly control/only using 1 ear tag Deworming of heifers All heifers Only some heifers/no heifers End term heifers kept with dry cows when not on pasture No Yes End term heifers supplemented with dry cow minerals No Yes End term heifers during summer supplemented with something else than hay, straw, sugar pulp, silage. No Yes End term heifers supplemented with straw during fall/winter No Yes Calving season of heifers During autumn During spring/during the whole year Withdrawn for further analysis. See text for details. 133

148 Chapter 6 Table 2c. Significant (P < 0.15) unconditional associations between management practices (continuous) and the geometric herd mean SCCel of all heifers between 5 and 14 DIM per herd (HSCCel, x 1000 cells/ml milk) for 159 high yielding dairy herds in Flanders (Belgium). Continuous variables β SE P Average herd net gain 1 (x 10,000) Geometric mean bulk milk SCC (x 1000 cells/ml milk) <0.001 Colostrum fed to newborn calves within first 24h (l) Number of individual pens for calves Percentage of heifers calving with difficulties Average fat and protein corrected production per mo for the whole herd, allowing ranking of herds. Multivariable linear regression analysis resulted in a final model with 7 significant herd practice variables at P < 0.1 significance level (Table 3). Table 3. Final multivariable linear regression model based (P < 0.10) on the previously selected herd level variables for the geometric herd mean SCCel of heifers between 5 and 14 days in milk per herd (HSCCel, x 1000 cells/ml milk) as the dependent variable. Variable β SE P <0.001 Province 1 <0.001 Vlaams-Brabant Ref. 3 Antwerpen West-Vlaanderen Oost-Vlaanderen Limburg Average herd net gain (x 10,000) 2, Geometric mean bulk milk SCC (x 1000 cells/ml milk) <0.001 Floor type of calving place or calving pen Not-slatted Ref. 3 Slatted Average age of heifers at calving Between 23 and 27 mo Ref. 3 Older than 27 mo Floor bedding stanchion place or cubicle Straw or sawdust Ref. 3 Rubber or concrete End term heifers fed straw during fall/winter No Ref. 3 Yes Forced into the model. 2 Average fat and protein corrected production per mo for the whole herd, allowing ranking of herds. 3 Reference category. 4 Withdrawn for further analysis. See text for details. 134

149 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count Heifer-level Analysis Somatic cell counts in early lactation showed a seasonal variation, with the highest values found during May to July (Figure 2) Log-transformed SCC in early lactation Jan Febr Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 2. Log-transformed SCCel (x 1000 cells/ml) (± SEM) between 5 and 14 days in milk in 2000, per mo from 1912 heifers from 159 dairy herds in Belgium Log-transformed SCC in early lactation Days in milk Figure 3. Log-transformed SCCel (x 1000 cells/ml) (± SEM) in 2000 between 5 and 14 days in milk in 2000, per day in milk from 1912 heifers from 159 dairy herds in Belgium. 135

150 Chapter 6 Somatic cell counts in early lactation progressively declined from 133,000 at 5 DIM to 70,000 cells/ml at 14 DIM (Figure 3). Most of the 1912 SCCel were measured in August to December (n = 1431 or 74.8%), with the highest proportion measured in September (n = 407 or 21.3%) and lowest in July (n = 21 or 1.1%). The univariable screening of the 4 heifer factors using LnSCCel for all individual heifers as dependent variable led to 3 significant heifer characteristics: DIM, date at which SCCel was measured (recoded to calving season), and production on day of SCCel measurement (kg milk). All 3 variables were presented to the multilevel analysis. Multilevel Analysis The final multilevel model for the continuously distributed LnSCCel is presented in Table 4. Four management practices and 2 heifer characteristics were significantly associated with LnSCCel of heifers between 5 and 14 DIM. In higher producing herds and in herds with an average calving age of heifers >27 mo, SCCel of heifers was lower than in less producing herds or in herds with an average calving age 27 mo. Dairy heifers raised in herds with a higher BMSCC had higher LnSCCel. In herds were heifers calved on slatted floors, LnSCCel was lower than in herds were heifers calved on not-slatted floors. Province also remained a significant variable in the final model. At the heifer level, LnSCCel decreased with increasing DIM. Heifers calving in April to June ran a higher risk of having elevated LnSCCel opposed to the other 3 defined seasons of the year. The variance components for the different models are presented in Table 5. In the threelevel null model, 1.6, 1.5 and 96.8%, of the variation occurred at the province, herd and heifer level, respectively, whereas in the two-level null-model, 2.6, and 97.3% of the variation occurred at the herd and heifer level, respectively. These proportions changed slightly when adding the significant fixed effects. Nine-and-a-half percent of the total variance was explained by the fixed effects ([ /1.696] * 100). 136

151 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count Table 4. Significant (P < 0.05) herd- and heifer-level variables in the final multilevel linear model describing log-transformed SCCel for 1805 heifers between 5 and 14 days in milk (LnSCCel) on 150 high yielding dairy herds in Flanders (Belgium). Variable β SE P Constant <0.001 Province Vlaams-Brabant Ref. 5 <0.001 Antwerpen West-Vlaanderen Oost-Vlaanderen Limburg Average herd net gain (x 10,000) 1,2, <0.001 Geometric mean bulk milk SCC (x 100,000 cells/ml) 1, <0.001 Floor type of calving place or calving pen Not-slatted Ref. 5 Slatted Average age of heifers at calving Between 23 and 27 mo Ref. 5 Older than 27 mo Days in milk 2, <0.001 Calving date April June Ref. 5 January March July September October December Herd-level data (based on questionnaire data). 2 Centred by subtracting the mean. 3 Average fat and protein corrected production per mo for the whole herd, allowing ranking of herds. 4 Heifer-level data (extracted from milk recording lists). 5 Reference category. Table 5. Variance components at each level of 2 null models and the final model. Data Variance estimate (SE 1 - % 2 ) hierarchy Three-level null model 3 Two-level null model 4 Two-level final model 5 Province ( ) Herd ( ) ( ) ( ) Heifer ( ) ( ) ( ) Total variance Standard error of the variance estimate of the parameter. 2 Variance proportion explained at the level of the data hierarchy. 3 Model containing only the intercept with "province", "herd" and "heifer" as random effects. 4 Model containing only the intercept with "herd" and "heifer" as random effects. 5 Final model containing all fixed effects presented in Table 5, and "herd" and "heifer" as random effects. 137

152 Chapter 6 Discussion Milk somatic cell counting is an important instrument for monitoring udder health in lactating cows and is used worldwide as an indicator of subclinical mastitis (Laevens et al., 1998). Quarter-milk SCC is applicable as of day 2 postpartum to determine IMI (Barkema et al., 1999a). In the present study, SCC measured between 5 and 14 DIM were extracted from four-weekly milk recordings to reflect udder health just after freshening. In culture-negative quarters, geometric mean SCC was as low as 42,000 cells/ml at the sixth milking after calving (Barkema et al., 1999a), whereas 16 heifers that were culture-negative during the whole first lactation had an average composite milk SCC of 49,000 cells/ml during their first mo of lactation (Laevens et al., 1997). To identify management factors associated elevated SCCel, HSCCel was calculated giving an overall idea of udder health in periparturient heifers in a herd. A number of studies were able to find herd management practices and heifer characteristics significantly associated with clinical mastitis (Waage et al., 1998, 2001; Myllys and Rautala, 1995). Not many studies, however, have focused on risk factors for subclinical mastitis in heifers early postpartum. Probably the first study to do so determined how the effect of location, season, stage of pregnancy, and herd influenced IMI in heifers (Fox et al., 1995). Two other studies focused on the issue using SCC of heifers (Østerås et al., 1997; Bareille et al., 2000). The present study is, to our knowledge, the first one on a large number of dairy herds to study both management practices and some heifer characteristics that might influence the prevalence of mainly subclinical mastitis in recently freshened heifers. In addition, it was possible to estimate the distribution of variance on different levels (province, herd, and heifer), using multilevel modelling, a technique that takes into account the hierarchical nature of data. As a result, it was possible to identify the level at which interventions will have the greatest impact on the outcome. Around 97% of the variance of LnSCCel occurred at the heifer level based on the null models with very small variance estimates at the province and/or herd level. All variation at the herd level was explained by including the significant herd-level fixed effects, with all variation left in SCCel at the heifer level. Another study on variance components of SCC also reported that herd only explained a small proportion of the variation in LnSCC compared to cows nested within herds (Schepers et al., 1997). The extreme distribution in the present study is most probably influenced by opting to work with SCC measured between 5 and 14 DIM, a period in which SCC vary more when compared to the rest of the lactation, and by opting for herds that had at least 10 SCCel during The latter 138

153 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count procedure resulted in selecting slightly larger and better producing dairy herds with better udder health: geometric mean SCCel of the 1912 heifers was lower when compared to the value obtained from the 14,766 heifers from 2000, of which they were a subset. This selection bias possibly influenced the variance distribution slightly, whereas if the herds would have been selected randomly, more variation at the herd level could have been expected. The selection procedure used in this present study was, however, necessary to get a reliable impression of udder health in heifers on these herds. Hence, as the vast majority of the variation of SCCel resides at the heifer level, focusing on heifers rather than on the herds seems to be necessary when dealing with udder health in the short term. Lack of substantial effects at the herd level, though, should not be considered as proof that the improvement of udder health of dairy heifers is beyond control of the farmer. This appears also from the fact that the herd-level fixed effects incorporated in the final model completely reduced the herd-level variance. But, although management practices of herds are thought to be of a major importance, a large variation seems to exist among heifers in response to an identical management. A possible explanation for the lack of variation at the herd level may be that strategies with a known positive effect on udder health in freshening dairy heifers are not yet implemented in Belgian dairy herds. Another reason could be that important herd-level factors have not yet been identified and have consequently not been implemented. Therefore, if effective preventive measures leading to substantially different SCCel levels among herds, would be adopted by some herds and not by others, more variation could be expected at the herd level. Although statistically significant associations between SCCel and both management practices and heifer characteristics were found in the present study, this does not proof causality. When studying large datasets with lots of variables as in the present study, the probability of finding associations by chance increase substantially (Dohoo et al, 1997). In addition, the management ability of the farmer, influencing both udder health and other management practices, could be a confounder (Barkema et al., 1998b). Significant differences in SCCel between the 5 provinces in Flanders (Belgium) were seen in the present study. It was, however, not possible to explain this by the differences in management between herds. Surprisingly, the average BMSCC per province were not reflecting the same order in udder health, with the province Vlaams-Brabant having the lowest BMSCC (155,000 cells/ml) and the province West-Vlaanderen having the highest BMSCC (183,000 cells/ml). 139

154 Chapter 6 Better producing dairy herds had an overall better udder health in heifers in early lactation. To reach higher levels of milk production these herds have to be well-managed and, therefore, it is likely that they are more aware of the importance of heifer's udder health. Nevertheless, it seems that the herd management practices associated with this higher production were not captured in the present study. Higher BMSCC, reflecting the prevalence of subclinical mastitis in the lactating cows, was associated with a higher risk for individual heifers in a herd to have an elevated SCCel. This indicates that heifers raised in herds with a higher prevalence of subclinical mastitis run higher risks to calve with IMI or to develop IMI within 1-2 wks postpartum, which agrees with the findings of another study (Bareille et al., 2000). On the other hand, high SCCel of heifers could theoretically have increased the BMSCC of the particular herd, although only 1 SCCel per heifer will not have had a huge impact on the BMSCC as this was calculated as the geometric mean of all months in Calving of heifers on slatted floors as apposed to calving on not-slatted floors, reduced the odds of heifers having a high SCCel. Calving areas that have slatted floors will have drier bedding as opposed to areas with completely closed floors. The same reasoning is behind the finding in another study where not complete slatted floors and alleys increased the incidence rate for Escherichia coli mastitis versus complete slatted floors and alleys (Barkema et al., 1999b). Providing a dry and clean environment to heifers and cows that are calving is important. A poor hygiene of the calving area was associated with an increased prevalence of elevated SCC in heifers (Bareille et al., 2000). Mucking out the calving area less often than once a month was a significant risk for clinical mastitis (Peeler et al., 2000). An increased incidence rate of E. coli mastitis, but not S. aureus mastitis, was found in herds where there was no disinfection of the calving area after parturition (Elbers et al., 1998). In herds where heifers calved at an older age, heifers were at a reduced risk of having a high SCCel. This was a bit surprising as older heifers will have been at risk of contracting IMI during a longer period. The current data were measured at the herd level, which makes conclusions on the individual heifer level invalid. No conclusions can therefore be drawn regarding a possible dilution effect as it is known that that heifers calving at an older age tend to produce a bit more compared to younger heifers. On the heifer level, 2 variables were significantly associated with SCCel. Days in milk, as was expected in early lactation, was negatively correlated with SCCel. A similar pattern was seen in other studies looking at composite milk samples (Dohoo, 1993; Laevens et al., 1997). Dilution due to an increasing production will most probably have contributed to this 140

155 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count effect (Schepers et al., 1997), but this finding also suggests spontaneously cure of IMI caused by coagulase-negative staphylococci, and possibly also treatment of some heifers. Calving season was also associated with SCCel. Heifers calving in April to June had a higher risk of having elevated SCCel opposed to heifers calving in the other 3 defined seasons of the year. The fact hat heifers have been in confined housing for a long period puts them at an increased risk of getting IMI, which is comprehensible. Myllys and Rautala (1995) identified 2 peak periods during which more heifers were treated for clinical mastitis, which is in accordance with the one peak in the SCCel in our data. It should, however, be noticed that the peak during July in our data for the 1912 heifers was only based on 21 values. In other studies, calving in late spring or summer was associated with an increased risk for clinical mastitis (Waage et al., 1998), and there was a significant association between the number of heifers with SCC >200,000 cells/ml and calving month, with May holding the highest and December the lowest risk versus September as reference (Østerås et al., 1997). The significant heifer-level variables in the present study are limited in improving udder health in dairy heifers. Changing the calving season to the most favourable months is possible. As most of the variation was present at the heifer level, it would have been interesting to have more details on that level. Variables like calving problems of individual heifers, prepartum milk leakage, age at calving would have reduced the variance at that level. The present study was, however, designed mainly to focus on management practices, and although only 4 of these practices were significantly associated with SCCel, they totally explained the variation at the herd level. Some variables that were expected to be of importance and were included in the questionnaire were not significant in the present study. From literature, for instance, it is known that fly control in heifers is very important in prevention of summer mastitis. Horn flies, in addition, are capable of spreading S. aureus-induced intramammary infection to heifers (Owens et al., 1998). Another study revealed that heifers from herds using fly control had a lower prevalence of IMI than herds without fly control (Nickerson et al., 1995). In the present study, fly control resulted in lower SCCel in the univariable analysis, but this variable disappeared in further multivariable analyses. Supplementing heifers with minerals was a managerial practice with an expected protective influence (Barkema et al., 1998b) on HSCCel and although a univariable association was present in our study it was not present in the final model. Although analyzing the data on a dichotomous scale based on a threshold reflecting a heifer to either be infected or not infected was an option, it was decided to work on a 141

156 Chapter 6 continuous scale for several reasons. Dichotomizing the data would have reduced the power of the study to find significant variables. In addition, the discussion of which threshold was avoided and the use of a threshold could have resulted in biological fallacy: when using e.g. a cut-off of 200,000 cells/ml classifying heifers with a SCCel of 201,000 and 2,500,000 in the same group is debatable. Additionally, one of the aims of this study was to calculate the variance components, which is less straight-forward in binary models (Dohoo, et al., 2001). Finally, we believe that the actual SCCel reflects better whether a heifer is infected in 0, 1, 2, 3, or 4 quarters, where a binary model would classify a heifer just as infected or not. Based on the data used in the present study it was not possible to decide whether a heifer suffered from an elevated SCCel because of a prepartum infection (most likely to be subclinical) or because of an infection occurring during the first 2 wk after calving (most likely to be clinical). Yet, it was possible to relate the significant management factors to preor postpartum infections. It is known that prevalence of IMI, mainly caused by minor pathogens and S. aureus at calving is very high (Oliver and Mitchell, 1983; Trinidad et al., 1990; Pankey et al., 1991; Myllys, 1995; Fox et al., 1995; Nickerson et al., 1995), indicating that heifers were infected prior to calving. Prepartum management factors will therefore have influenced this phenomenon and we believe that the variables "calving season" or "BMSCC" are most probably related to the aforementioned prepartum IMI. On the other hand, some of the heifers will have been infected in the first couple of days after calving as we know from previous studies that incidence rate of mainly clinical mastitis is high during that period (Barkema, et al., 1998a). We trust, therefore, that the variable "hygiene in the calving pen" is related to new infections in early lactation. Conclusions Multilevel regression modelling revealed some important herd management practices and some heifer characteristics associated with SCCel of dairy heifers in early lactation. Probably the most useful is reducing the BMSCC and thus the prevalence of subclinical mastitis in the herd, in this way reducing the risk of heifers getting infected prior to and at calving. In addition, more attention should go to periparturient hygiene and having as much heifers as possible calving during autumn as this reduces the chance of having elevated SCCel significantly under Belgian conditions. Focusing more on differences between heifers than between herds seems to be important to control the problem of heifers calving with bad udder health in the short term, although the present data suggest both a lack of knowledge on the 142

157 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count pathogenesis of IMI in heifers, and the lack of implementation of strategies with a known beneficial effect. Conducting more studies combining bacteriological culture, SCC recording, registration of clinical cases, recording of detailed heifer information and data on management practices on a large number of herds, are essential in tackling udder health problems in dairy heifers. In addition, the effect of an elevated SCCel on production, udder health, and culling risk throughout the whole first lactation should be studied. Acknowledgments The authors would like to thank E. De Mûelenaere and the Flemish Cattle Breeding Association (Oosterzele, Belgium) for providing us with the milk recording lists and the Committees for Milk Quality for providing us with the BMSCC. All farmers that cooperated by sending back the questionnaire are gratefully acknowledged. Thanks must go to J. Valcour for the excellent assistance in data handling, to S. Willaert for the help in mailing the questionnaires, and to Intervet Belgium for supporting this study. References Bareille, N., H. Seegers, M. B. Kiebre-Toe, F. Beaudeau, and C. Fourichon Risk factors for elevated milk somatic cell counts during early lactation in dairy heifers. Pages in Proc. 10 th Intern. Congr. on Anim. Hyg., Maastricht, The Nederlands. Barkema, H. W., Y. H. Schukken, T.J.G.M. Lam, M. L. Beiboer, H. Wilmink, G. Benedictus, and A. Brand. 1998a. Incidence of clinical mastitis in dairy herds grouped in 3 categories by bulk milk somatic cell counts. J. Dairy Sci. 81: Barkema, H. W., Y. H. Schukken, T.J.G.M. Lam, M. L. Beiboer, G. Benedictus, and A. Brand. 1998b. Management practices associated with low, medium, and high somatic cell counts in bulk milk. J. Dairy Sci. 81: Barkema, H. W., H. Deluyker, Y. H. Schukken, and T.J.G.M. Lam. 1999a. Quarter-milk somatic cell count at calving and at the first 6 milkings after calving. Prev. Vet. Med. 38:

158 Chapter 6 Barkema, H. W., Y. H. Schukken, T.J.G.M. Lam, M. L. Beiboer, G. Benedictus, and A. Brand. 1999b. Management practices associated with the incidence rate of clinical mastitis. J. Dairy Sci. 82: De Vliegher, S., H. Laevens, G. Opsomer, E. De Mûelenaere, and A. de Kruif Somatic cell counts in dairy heifers during early lactation. Vlaams Diergeneeskd. Tijdschr. 70: Dohoo, I. R An evaluation of the validity of individual cow somatic cell counts from cows in early lactation. Prev. Vet. Med. 16: Dohoo, I. R., C. Ducrot, C. Fourichon, A. Donald, D. Hurnik An overview of techniques for dealing with large numbers of independent variables in epidemiologic studies. Prev. Vet. Med. 29: Dohoo, I. R., E. Tillard, H. Stryhn, and B. Faye The use of multilevel models to evaluate sources of variation in reproductive performance in dairy cattle in Reunion Island. Prev. Vet. Med. 50: Edinger, D., B. A. Tenhagen, P. Kalbe, G. Klunder, B. Baumgartner, and W. Heuwieser Effect of teat dipping with a germicide barrier teat dip in late gestation on intramammary infection and clinical mastitis during the first 5 d post-partum in primiparous cows. J. Vet. Med. A 47: Elbers, A.R.W., J. D. Miltenburg, D. De Lange, A.P.P. Crauwels, H. W. Barkema, and Y. H. Schukken Risk factors for clinical mastitis in a random sample of dairy herds from the southern part of The Netherlands. J. Dairy Sci. 81: Fox, L. K., S. T. Chester, J. W. Hallberg, S. C. Nickerson, J. W. Pankey, and L. D. Weaver Survey of intramammary infections in dairy heifers at breeding age and first parturition. J. Dairy Sci. 78: Giraudo, J. A., A. Calzolari, H. Rampone, A. Rampone, A. T. Giraudo, C. Bogni, A. Larriestra, and R. Nagel Field trials of a vaccine against bovine mastitis. 1. Evaluation in heifers. J. Dairy Sci. 80: Goldstein, H., J. Rasbash, I. Plewis, D. Draper, W. Browne, M. Yang, G. Woodhouse, and M. Healy A User s Guide to MlwiN. Institute of Education, London, 178 pp. Laevens, H., H. Deluyker, Y. H. Schukken, L. De Meulemeester, R. Vandermeersch, E. De Mûelenaere, and A. de Kruif Influence of parity and stage of lactation on the somatic cell count in bacteriologically negative cows. J. Dairy Sci. 80:

159 Management Practices and Heifer Characteristics associated with Early Lactation Somatic Cell Count Laevens, H., H. Deluyker, and A. de Kruif Somatic cell count (SCC) measurements: a diagnostic tool to detect mastitis. Pages in Proc. 10 th Int. Conf. on Prod. Dis. in Farm Animals, Utrecht, The Nederlands. Myllys, V Staphylococci in heifer mastitis before and after parturition. J. Dairy Res. 62: Myllys, V., and H. Rautala Characterization of clinical mastitis in primiparous heifers. J. Dairy Sci. 78: Neter, J., M. H. Kutner, C. J. Nachtsheim, and W. Wasserman Applied linear statistical models. 4th ed. Irwin, London, UK. Nickerson, S. C., W. E. Owens, and R. L. Boddie Mastitis in dairy heifers: initial studies on prevalence and control. J. Dairy Sci. 78: Oliver, S. P., and B. A. Mitchell Intramammary infections in primigravid heifers near parturition. J. Dairy Sci. 66: Oliver, S. P., M. J. Lewis, B. E. Gillespie, and H. H. Dowlen Influence of prepartum antibiotic therapy on intramammary infections in primigravid heifers during early lactation. J. Dairy Sci. 75: Oliver, S. P., M. J. Lewis, B. E. Gillespie, H. H. Dowlen, E. C. Jaenicke, and R. K. Roberts Prepartum antibiotic treatment of heifers: milk production, milk quality and economic benefit. J. Dairy Sci. 86: Østerås, O., R. B. Larssen, and E. Simensen Environmental risk factors associated with mastitis in heifers. Pages in Proc. 9 th Intern. Congr. Anim. Hyg., Helsinki, Finland. Owens, W. E., S. C. Nickerson, P. J. Washburn, and C. H. Ray Prepartum antibiotic therapy with a cephapirin dry-cow product against naturally occurring intramammary infections in heifers. J. Vet. Med. B 41: Owens, W. E., S. P. Oliver, B. E. Gillespie, C. H. Ray, and S. C. Nickerson Role of horn flies (Haematobia irritans) in Staphylococcus aureus-induced mastitis in dairy heifers. Am. J. Vet. Res. 59: Owens, W. E., S. C. Nickerson, R. L. Boddie, G. M. Tomita, and C. H. Ray Prevalence of mastitis in dairy heifers and effectiveness of antibiotic therapy. J. Dairy Sci. 84: Pankey, J. W., P. A. Drechsler, and E. E. Wildman Mastitis prevalence in primigravid heifers at parturition. J. Dairy Sci. 74:

160 Chapter 6 Peeler, E. J., M. J. Green, J. L. Fitzpatrick, K. L. Morgan, and L. E. Green Risk factors associated with clinical mastitis in low somatic cell count British dairy herds. J. Dairy Sci. 83: Schepers, A. J., T.J.G.M. Lam, Y. H. Schukken, J.B.M. Wilmink, and W.J.A. Hanekamp Estimation of variance components for somatic cell counts to determine thresholds for uninfected quarters. J. Dairy Sci. 80: Shearer, J. K., and R. J. Harmon Mastitis in heifers. Vet. Clin. North Am., Food Anim. Pract. 9: Tenhagen, B. A., D. Edinger, B. Baumgartner, P. Kalbe, G. Klunder, and W. Heuwieser Efficacy of a herd-specific vaccine against Staphylococcus aureus to prevent post-partum mastitis in dairy heifers. J. Vet. Med. A 48: Trinidad, P., S. C. Nickerson, and T. K. Alley Prevalence of intramammary infection and teat canal colonization in unbred and primigravid dairy heifers. J. Dairy Sci. 73: Waage, S., S. Sviland, and S. A. Ødegaard Identification of risk factors for clinical mastitis in dairy heifers. J. Dairy Sci. 81: Waage, S., S. A. Ødegaard, A. Lund, S. Brattgjerd, and T. Rothe Case-control study of risk factors for clinical mastitis in postpartum dairy heifers. J. Dairy Sci. 84:

161

162

163 Chapter 7 Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers

164

165 Chapter 7.1 Prepartum Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers Is Associated with Low Somatic Cell Count in Early Lactation S. De Vliegher, 1 H. Laevens, 2 L. A. Devriese, 3 G. Opsomer, 1 J.L.M. Leroy, 1 H. W. Barkema, 4 and A. de Kruif 1 1 Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 2 Coordination Centre for Veterinary Diagnostics, Veterinary and Agrochemical Research Centre, Brussels, Belgium 3 Department of Pathology, Bacteriology and Poultry Diseases, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 4 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Charlottetown, PEI, Canada Veterinary Microbiology, 2003, 92:

166 Chapter

167 Prepartum Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers Abstract A high number of dairy heifers freshen with udder health problems. The prevalence of teat apex colonization (TAC) with Staphylococcus chromogenes, one of the most widespread coagulase-negative staphylococci (CNS) in milk samples from freshly calved dairy heifers, was measured cross-sectionally in non-lactating heifers on 8 commercial dairy farms in Belgium. The influence of age on this prevalence, and the association between teat apex colonization with S. chromogenes prepartum and quarter milk somatic cell count (SCC) in early lactation were studied. In total, 492 teat apices were sampled from 123 heifers. The age of the heifers varied from 8 to 34 mo. Overall, 20% of the heifers had at least 1 teat apex colonized with S. chromogenes. Of all teats sampled, 10% were colonized with S. chromogenes. The chance of having at least 1 teat apex colonized with S. chromogenes increased with age of the heifer. The presence of prepartum teat apex colonization with S. chromogenes was not associated with intramammary infection (IMI) early postpartum with the same bacterium. On the contrary, teat apex colonization with S. chromogenes prepartum appeared to protect quarters in the first few days of lactation from having somatic cell count 200,000 cells/ml milk, commonly accepted as the threshold for intramammary infection. Key Words Cattle - bacteria, intramammary infection, mastitis, somatic cell count, Staphylococcus chromogenes, teat apex colonization Abbreviation Key CNS = coagulase-negative staphylococci, IMI = intramammary infection, SCC = somatic cell count, TAC = teat apex colonization. 153

168 Chapter 7.1 Introduction Ideally, dairy heifers should freshen with a healthy udder capable of producing as much high quality milk as possible. This goal is not always reached as many studies have reported a high prevalence of intramammary infections in non-lactating and freshly calved heifers with minor pathogens, mainly coagulase-negative staphylococci, and major pathogens, especially Staphylococcus aureus (Oliver and Mitchell, 1983; Boddie et al., 1987; Trinidad et al., 1990; Pankey et al., 1991; Shearer and Harmon, 1993; Myllys, 1995; Fox et al., 1995; Nickerson et al., 1995). Intramammary infections at parturition, either clinical or subclinical, will lead to considerable financial losses for the farmer and should be avoided by all means. The prevalence of teat canal and teat skin colonization in unbred and primigravid heifers, with either CNS or S. aureus, has been reported to be equally high (Boddie et al., 1987; White et al., 1989; Trinidad et al., 1990; Roberson et al., 1994; Roberson et al., 1998). White et al. (1989) found 62% of the sampled heifers harbouring S. chromogenes on the teat skin, while Trinidad et al. (1990) found 43% of all teat canal samples to be colonized by the same bacterium. Staphylococcus chromogenes is a highly prevalent CNS species in both milk and teat canal samples from unbred and freshly calved heifers with a well-described impact on somatic cell counts (Laevens et al., 1997). Teat apex colonization with S. chromogenes near calving could increase the risk of IMI with the same pathogen in early lactation, compromising current and future milk quality and production. Heifers with teat orifices colonized by S. aureus were 3.3 times more likely to have IMI with S. aureus at parturition than heifers with non-colonized teat orifices (Roberson et al., 1994). On the other hand, teat apex colonization with S. chromogenes could hypothetically protect against IMI with major pathogens. The hypothesis of inhibition of mastitis pathogens by bovine teat skin normal flora was tested in vitro with some success (Woodward et al., 1987). The aims of this study were: 1) to measure the prevalence of TAC by S. chromogenes, in young and primigravid dairy heifers, 2) to determine the effect of age of heifers on this TAC prevalence, and 3) to study the association between TAC prepartum and IMI in early lactation with S. chromogenes, and between TAC and quarter milk SCC in early lactation. 154

169 Prepartum Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers Materials and Methods Herds and Animals Eight Belgian dairy herds, consisting mainly of Black Holstein-Friesian and Holstein crossbreds, were selected according to the willingness of the farmer to cooperate with this study. Average 305-d milk production on the dairy herds for was 8264 kg, ranging from 7000 to 9804 kg. Average mean bulk milk SCC of the herds was 214,000 cells/ml, ranging from 92,000 to 400,000 cells/ml. In total, teat apices (n = 492) of 123 young and primigravid heifers between 8 and 34 mo of age were sampled cross-sectionally. Number of heifers sampled per herd ranged from 3 to 31, with an average of 15. The heifers that were sampled were a convenience sample, stabled on the day the herds were visited. Teat apices from 36 of the 123 heifers were sampled within 7 to 14 d prior to the expected calving date. Quarter milk samples were collected between 3 to 5 d post-partum for bacteriological culture and measurement of SCC from 30 of these 36 heifers. Sampling and Laboratory Procedures The heifers were immobilized and every teat apex was first cleansed with a sterile cloth, in this way removing any visual soil on the teats. Thereafter, they were sampled by rotating gently with a transport swab (Cultiplast Transport Swab, Amies, Milan, Italy) on the teat apex. All swabs were transported under cooled conditions immediately after sampling to the Laboratory of the Herd Health Department (Faculty of Veterinary Medicine, Ghent University, Belgium). The morning after sampling, processing of the swabs was started. Quarter milk samples were collected aseptically in duplicate, for both bacteriological culture and SCC measurement. Swabs were streaked onto a half of a Petri Dish with a blood agar base (Oxoid, Basingstoke, UK), supplemented with 5% bovine blood. The same procedure was followed with an inoculum of 0.01 ml of the milk samples. All agar plates were incubated aerobically at 37 C and bacterial growth was read after 24 and 48 hours of incubation. Only yellow, nonhaemolytic colonies were considered and selected for further identification using DNase tests that were carried out by the standardized procedure described by Devriese and Van de Kerckhove (1979). This involved inoculating up to 8 strains radially as narrow streaks with a straight needle on 90 mm plates containing 20 ml DNase agar (Oxoid) and reading after 24 hours of incubation. The reaction zones of S. chromogenes strains are about twice as broad as 155

170 Chapter 7.1 the growth streaks, while those of S. aureus and Staphylococcus hyicus measure at least fourfold the diameters of the streaks. Other CNS associated with IMI are either negative or show very weak and narrow reaction zones. In a recent study, it was concluded that S. chromogenes can be identified reliably using these methods (Devriese et al., 2002). A teat apex was classified as colonized and a quarter milk sample as infected with S. chromogenes when at least 8 CFU were isolated on the blood agar plate. Somatic cell counts were measured in duplicate by the Flemish Cattle Breeding Association (Oosterzele, Belgium) with the Fossomatic 5000 (Foss Electric, Hillerød, Denmark). Data Processing The heifers were allocated to 4 age classes: class 1: 8-12 mo (young, unbred heifers), class 2: mo (average breeding age), class 3: mo (average calving age), and class 4: >28 mo (oldest), to present the data descriptively and graphically. The following variables were calculated: mean number of colonized teat apices per heifer, total prevalence of TAC, prevalence of heifers with at least 1 S. chromogenescolonized teat apex, and prevalence of colonized heifers sampled per herd. Somatic cell counts were checked for unlikely values and no data were excluded for this reason. Statistical Analysis The association between age of a heifer, and colonization with S. chromogenes was analysed using logistic regression. A model was built with age as continuous independent variables (expressed in mo), and colonized heifer (0: not colonized by S. chromogenes, 1: colonized by S. chromogenes) as binary outcome variable. Herd was included as random effect (MLwiN 1.1; Goldstein et al., 1998). The association between TAC prepartum, and quarter milk SCC in early lactation also was analysed using logistic regression. A model was built with the binary variable TAC prepartum at teat level (0: teat not colonized by S. chromogenes, 1: teat colonized by S. chromogenes) as the independent variable, and SCC (0: quarter milk SCC in early lactation <200,000 cells /ml milk; 1: quarter milk SCC in early lactation 200,000 cells /ml milk), as the binary outcome variable. Herd and heifer were included as random effects, to deal with clustering at these levels (MLwiN 1.1; Goldstein et al., 1998). Odds ratios (OR) with 95% confidence intervals (95% CI) were calculated. Statistical significance was assumed at P

171 Prepartum Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers Results The average age of the heifers at the moment of sampling of teat apices was 18.7 mo. Overall, 20% of all heifers had at least 1 S. chromogenes-colonized teat apex, while 10% of all teat apices were colonized with this bacterium. Mean prevalence of colonized heifers sampled per herd was 33%, ranging from 10% to 83%. The youngest heifer to be classified as colonized with S. chromogenes was 10.9 mo of age. More than one-third of the heifers between 20.1 and 28 mo of age, and half of the heifers older than 28 mo had at least 1 teat apex colonized by S. chromogenes (Figure 1, Table 1). Colonized heifers were on average colonized on 2 teats with S. chromogenes (Table 1). Prevalence (%) Heifers Teat apexes Age at sampling (months) Figure 1. Prevalence of Staphylococcus chromogenes-colonized heifers (with at least 1 colonized teat apex), and teat apices per age class. Table 1. Prevalence of Staphylococcus chromogenes-colonization in 123 dairy heifers. Age class n Prevalence Prevalence Average number of colonized teat apices (mo) (heifers) (% heifers) (% apices) per colonized heifer % 4.5% % 3.4% % 16.2% % 27.1% 2.2 Results of the analysis on the association between age, and colonization with S. chromogenes revealed that with increasing age, the chance of heifers being colonized on at least 1 teat apex by S. chromogenes increased (OR= 1.17, 95% CI: , P 0.001). 157

172 Chapter 7.1 From the 36 heifers that were sampled within 2 wk prior to expected calving date, 16 (44%) had at least 1 teat apex colonized with S. chromogenes. From the 144 teat apices sampled, 32 (22%) were colonized with S. chromogenes. Quarter milk samples from 30 (n = 120) of these 36 animals were available in early lactation. Geometric mean SCC of the 120 quarters was 144,000 cells/ml milk (range: 10,000 23,136,000). Only 1 of 30 heifers had 2 quarter milk samples that were culture-positive for S. chromogenes. The teat apices of these 2 quarters were not colonized prepartum with this bacterium. Seventy-five of the 120 (63%) quarter milk samples were culture-negative for all pathogens, as no growth was detected on the blood agar plates after 48 hours of incubation. Geometric mean SCC for these milk samples was 138,000 cells/ml milk. Twenty-five of the 120 quarters (21%) had teat apices colonized with S. chromogenes before parturition. Teat apex colonization with S. chromogenes prepartum protected quarters against SCC 200,000 cells/ml milk in early lactation (OR = 0.27, 95% CI: , P = 0.048). Discussion Coagulase-negative staphylococci are found in high prevalence in secretion, milk and teat canal samples taken from non-lactating and freshly calved heifers (Oliver and Mitchell, 1983; Boddie et al., 1987, Trinidad et al., 1990; Pankey et al., 1991; Shearer and Harmon, 1993; Myllys, 1995; Nickerson et al., 1995; Barkema et al., 1999). Although they are called minor pathogens, these bacteria may be associated with clinical mastitis (Waage et al., 1999). S. chromogenes is a highly prevalent CNS species, capable of increasing SCC (Laevens et al. 1997), hence causing production losses. An SCC 200,000 cells/ml milk is associated with at least 1 kg decrease in daily milk yield in first lactation animals (Jones et al., 1984). In the present study, 10% of all teat apices sampled were colonized with S. chromogenes. This prevalence is similar to the prevalence in another study in which 10 nonlactating heifers were re-sampled bimonthly from 11 mo of age through freshening: 15% of all teat skin samples were colonized by S. chromogenes (Boddie et al., 1987). In another study, the prevalence was expressed as a percentage of staphylococcal isolates within teat skin isolates: 22% of the teat skin samples were colonized with S. chromogenes (White et al., 1989). Others found S. chromogenes on the teat skin and teat apices from lactating cows, although other CNS were more prevalent (Devriese and De Keyser, 1979). In the present study, 20% of all 123 heifers were colonized on at least 1 teat apex at variable ages, ranging from 10.9 to 34 mo. White et al. (1989) found S. chromogenes isolates on the teat skin of 62% 158

173 Prepartum Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers of the sampled nulliparous heifers ranging from 1 d to 2 yrs of age. In our study, the youngest heifer colonized by S. chromogenes was 10.9 mo old. Some clustering, both at herd and heifer level, was present in the data. On some herds more than 80% of the sampled heifers were colonized, while on other herds this was as low as 10%. However, because the number of herds in the study was small, and because sample sizes were small in some of these herds, these results should be interpreted with caution. A larger study including a high number of herds and heifers is needed for better conclusions. Heifers that are colonized show a tendency of being colonized on more than 1 teat which is in accordance with another study, where the number of teats colonized by staphylococci per heifer was even higher (Woodward et al., 1987). The chance of heifers having at least 1 teat apex colonized with S. chromogenes increased significantly with age. Boddie et al. (1987) re-sampled 10 heifers through freshening during a 1-year period, resulting in 388 teat skin samples. They did not report on the influence of increasing age on the prevalence. They furthermore concluded that infections of teat skin and canals with CNS and S. aureus could persist for 1 year. Distribution of staphylococcal species among age groups was presented by White et al. (1989), showing a relative increase of S. chromogenes populations with increasing age. Regrettably, in their study no absolute figures were presented on percentage of heifers and quarters colonized per pathogen or per age group, making a comparison with our data impossible. A tendency of increasing prevalence of heifers with S. aureus-colonized udders with increasing age was reported (Roberson et al., 1994). In our study, teat apices were sampled, while in other studies either teat skin and/or teat canals were sampled (Boddie et al., 1987; White et al., 1989; Roberson, 1994). The sampling of teat canals is difficult to perform on commercial dairies for obvious reasons, and was therefore avoided. It is, however, thought that sampling of the teat apex, close to the teat sphincter, could reflect a challenge to the teat canal, especially if a pure culture of a mastitisassociated pathogen is obtained. The prevalence of IMI with S. chromogenes in early lactation was low. From 30 heifers sampled, only 1 heifer yielded S. chromogenes in the milk of 2 quarters. These 2 quarters did not have TAC with S. chromogenes prior to calving; therefore no association between TAC and IMI with S. chromogenes could be demonstrated in our data. This might be an indication that TAC with S. chromogenes before calving is not a risk factor for IMI with the same bacterium after calving. As milk samples in our study were taken between 3 and 5 d postpartum, this low prevalence could, however, partially be explained by the fact that part of the 159

174 Chapter 7.1 IMI with S. chromogenes were washed out by milking. One would expect elevated SCC in these recently cleared quarters, however, our data showed the opposite: TAC with S. chromogenes prepartum protected against SCC 200,000 cells/ml. Others have found that heifers with a teat orifice colonized by S. aureus were 3.3 times more likely to have S. aureusinfected quarters at parturition than were non-colonized heifers (Roberson et al., 1994). On the other hand, recent work using pulse-field gel electrophoresis showed that S. aureus strains from bovine teat skin are different from milk strains (Zadoks et al., 2002). More elaborate studies at strain level are therefore needed to study the effect of TAC prepartum with minor or major pathogens on IMI and SCC in early lactation. As quarter milk SCC taken in early lactation reflect the IMI status of the particular quarter (Barkema et al., 1999), and as it is commonly accepted that SCC in quarter milk 200,000 cells/ml are indicative of a quarter either being infected or recently being infected (Hillerton, 1999), we conclude that TAC prepartum with S. chromogenes may be protective against IMI in early lactation. The actual mechanism by which S. chromogenes and other CNS might induce a protective effect remains unclear. It is known that naturally occurring intramammary S. chromogenes infections have a considerable protective effect against intramammary challenge with S. aureus (Matthews et al., 1990). Furthermore, field studies in low bulk milk SCC herds suggest the same effect by IMI with all minor pathogens (Schukken et al., 1989; Lam et al., 1997). We studied colonization of the teat apex rather than IMI, but the results seem to add further evidence of a protective effect of CNS against mastitis. Moreover, our results are in agreement with the findings of an in vitro study that found that 25% of the isolates of normal teat skin flora of non-lactating heifers, of which were some staphylococci genera, inhibited the growth of selected mastitis pathogens (Woodward et al., 1987). Competitive exclusion at the teat apex is the most likely mechanism of protection, but more studies, both experimental and in the field, are needed to clarify this. Acknowledgements The authors would like to thank K. Moerloose and L. Vandevelde for the excellent technical assistance. 160

175 Prepartum Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers References Barkema, H. W., H. Deluyker, Y. H. Schukken, and T. J. Lam Quarter-milk somatic cell count at calving and at the first 6 milkings after calving. Prev. Vet. Med. 38:1-9. Boddie, R. L., S. C. Nickerson, W. E. Owens, and J. L. Watts Udder microflora in nonlactating heifers. Agri-Practice 8: Devriese, L. A, and H. De Keyser Prevalence of different species of coagulasenegative staphylococci on teats and in milk samples from dairy cows. J. Dairy Res. 47: Devriese, L. A., and A. Van de Kerckhove A comparison of methods and the validity of desoxyribonuclease tests for the characterisation of staphylococci isolated from animals. J. Appl. Bacteriol. 46: Devriese, L. A., M. Baele, M. Vaneechoutte, A. Martel, and F. Haesebrouck Identification and antimicrobial susceptibility of Staphylococcus chromogenes isolates from intramammary infections of dairy cows. Vet. Microbiol. 87: Fox, L. K., S. T. Chester, J. W. Hallberg, S. C. Nickerson, J. W. Pankey, and L. D. Weaver Survey of intramammary infections in dairy heifers at breeding age and first parturition. J. Dairy Sci. 78: Goldstein, H., J. Rasbash, I. Plewis, D. Draper, W. Browne, M. Yang, G. Woodhouse, and M. Healy A User s Guide to MlwiN. Institute of Education, London, 178. Hillerton, J. E Redefining mastitis based on somatic cell count. Bulletin of the international Dairy Federation. No 345, 4-6. Jones, G. M., R. E. Pearson, G. A. Clabaugh, and C. W. Heald Relationships between somatic cell counts and milk production. J. Dairy Sci. 67: Laevens, H., H. A. Deluyker, L. A. Devriese, and A. de Kruif The influence of IMI with S. chromogenes or S. warneri/haemolytica on the somatic cell count in dairy cows. Proc. 8th Symp. Intern. Soc. Vet. Epidemiol. Econ. Epidémiol. Santé Anim , Lam, T. J., Y. H. Schukken, J. H. van Vliet, F. J. Grommers, M. J. Tielen, and A. Brand Effect of natural infection with minor pathogens on susceptibility to natural infection with major pathogens in the bovine mammary gland. Am. J. Vet. Res. 58:

176 Chapter 7.1 Matthews, K. R., R. J. Harmon, and B. A. Smith Protective effect of Staphylococcus chromogenes infection against Staphylococcus aureus infection in the lactating bovine mammary gland. J. Dairy Sci. 73: Myllys, V Staphylococci in heifer mastitis before and after parturition. J. Dairy Res. 62: Nickerson, S. C., W. E. Owens, and R. L. Boddie Mastitis in dairy heifers: initial studies on prevalence and control. J. Dairy Sci. 78: Oliver, S. P., and B. A. Mitchell Intramammary infections in primigravid heifers near parturition. J. Dairy Sci. 66: Pankey, J. W., P. A. Drechsler, and E. E. Wildman Mastitis prevalence in primigravid heifers at parturition. J. Dairy Sci. 74: Roberson, J. R., L. K. Fox, D. D. Hancock, J. M. Gay, and T. E. Besser Ecology of Staphylococcus aureus isolated from various sites on dairy farms. J. Dairy Sci. 77: Roberson, J. R., L. K. Fox, D. D. Hancock, J. M. Gay, and T. E. Besser Sources of intramammary infections from Staphylococcus aureus in dairy heifers in early lactation at first parturition. J. Dairy Sci. 81: Schukken, Y. H., D. Van De Geer, F. J. Grommers, J.A.H. Smit, and A. Brand Intramammary infections and risk factors for clinical mastitis in herds with low somatic cell counts in bulk milk. Vet. Rec. 125: Shearer, J. K., and R. J. Harmon Mastitis in heifers. Vet. Clin. North Am., Food Anim. Pract. 9: Trinidad, P., S. C. Nickerson, and T. K. Alley Prevalence of intramammary infection and teat canal colonization in unbred and primigravid dairy heifers. J. Dairy Sci. 73: Waage, S., T. Mork, A. Roros, D. Aasland, A. Hunshamar, and S. A. Odegaard Bacteria associated with clinical mastitis in dairy heifers. J. Dairy Sci. 82: White, D. G., R. J. Harmon, J. E. Matos, and B. E. Langlois Isolation and identification of coagulase-negative Staphylococcus species from bovine body sites and streak canals of nulliparous heifers. J. Dairy Sci. 72: Woodward, W. D., T. E. Besser, A. C. S. Ward, and L. B. Corbeil In vitro growth inhibition of mastitis pathogens by bovine teat skin normal flora. Can. J. Vet. Res. 51:

177 Prepartum Teat Apex Colonization with Staphylococcus chromogenes in Dairy Heifers Zadoks, R. N., W. B. van Leeuwen, D. Kreft, L. K. Fox, H. W. Barkema, Y. H. Schukken, and A. van Belkum Comparison of Staphylococcus aureus from bovine and human skin, milking equipment and bovine milk by phage typing, pulsed-field gel electrophoresis and binary typing. J. Clin. Microbiol. 40:

178

179 Chapter 7.2 In Vitro Growth Inhibition of Major Mastitis Pathogens by Staphylococcus chromogenes originating from Teat Apices of Dairy Heifers S. De Vliegher, 1 G. Opsomer, 1 A. Vanrolleghem, 1 L. A. Devriese, 2 O. C. Sampimon, 3 J. Sol, 3 H. W. Barkema, 4 F. Haesebrouck, 2 and A. de Kruif 1 1 Department of Reproduction, Obstetrics, and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 2 Department of Pathology, Bacteriology, and Poultry Diseases, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium 3 Animal Health Service, Deventer, The Netherlands 4 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Charlottetown, PEI, Canada Veterinary Microbiology, 2004, 101:

180 Chapter

181 In Vitro Growth Inhibition of Major Mastitis Pathogens by Staphylococcus chromogenes Abstract Earlier field observations suggest that teat apex colonization (TAC) by Staphylococcus chromogenes prepartum in dairy heifers protects udder quarters against elevated somatic cell counts (SCC) early post partum. To explain these findings, the in vitro inhibitory capability of S. chromogenes from teat apices of heifers towards some major mastitis pathogens was tested using a modified cross-streaking method. Two out of 10 S. chromogenes isolates, both originating from 2 different teats from the same heifer, consistently inhibited growth of all Staphylococcus aureus, Streptococcus dysgalactiae, and Streptococcus uberis strains, but none of the Escherichia coli strains. The present study, therefore, supports the protective effect of teat apex colonization by S. chromogenes by in vitro production of inhibitory substances. Key Words Cattle - bacteria, growth inhibition, major mastitis pathogen, Staphylococcus chromogenes, teat apex colonization Abbreviation Key CNS = coagulase-negative staphylococci, IMI = intramammary infection, SCC = somatic cell count, TAC = teat apex colonization. 167

182 Chapter 7.2 Introduction Coagulase-negative staphylococci (CNS) are generally considered as minor pathogens in bovine mastitis and are categorized as skin flora opportunists (Devriese and De Keyser, 1980; Boddie et al., 1987). Several studies indicate that CNS are the principal cause of intramammary infection (IMI) on modern dairy farms (Smith and Hogan, 2001). Prevalence of IMI with CNS is especially high in unbred and first lactation heifers (Boddie et al., 1987; Fox et al., 1995). On the other hand, a naturally occurring IMI with a particular CNS, namely Staphylococcus chromogenes, had a protective effect against an intramammary challenge with Staphylococcus aureus (Matthews et al., 1990). Other studies regarding the protective effect of minor pathogens against the development of infections with a major organism were, however, not conclusive, as some showed a protective effect (Linde et al., 1980; Lam et al., 1997), and others did not (Hogan et al., 1988). Quarters of heifers of which the teat apex was colonized with S. chromogenes before calving, had significantly lower odds for an elevated somatic cell count during the first 5 d of lactation compared with quarters of which the teat apex was not colonized with this bacterium (De Vliegher et al., 2003). The aim of the present study, therefore, was to examine the in vitro inhibitory capability of S. chromogenes originating from teat apices of dairy heifers towards S. aureus, Streptococcus dysgalactiae, Streptococcus uberis, and Escherichia coli, as a possible mechanism elucidating the findings in the latter study. Materials and Methods Animals, Swabs, and Milk Samples All S. chromogenes strains (n = 10) were isolated from swabs taken from teat apices of unbred, end-term, and recently freshened Holstein-Friesian and Holstein-Friesian crossbred dairy heifers. Swabbing was done as described by De Vliegher et al. (2003). All swabs were either transported under cooled conditions immediately after sampling to the Laboratory of the Herd Health Department, or frozen at -80 C until transportation on dry ice towards the aforementioned laboratory where they were processed further. In addition, 18 different pathogens associated with clinical or subclinical mastitis (milk samples) were used: 5 S. aureus, 6 S. dysgalactiae, 2 S. uberis, and 5 E. coli isolates. 168

183 In Vitro Growth Inhibition of Major Mastitis Pathogens by Staphylococcus chromogenes Bacteriological Procedures Bacteriological culturing and isolate identification of both milk and swabbing samples was done according to National Mastitis Council procedures (Harmon et al., 1990). A 0.01 ml inoculum of milk was spread on both half of a Colombia sheep blood agar plate and half a plate containing Edwards Medium (Oxoid). Plates were incubated aerobically at 37 C, and examined at 24 and 48h. In addition, DNase tests were carried out for identification of S. chromogenes and S. aureus according to the standardized procedure (Devriese et al., 2002). Identification of E. coli was based on motility indole ornithine decarboxylation and Kliglertests (Oxoid). Identification of S. aureus, S. dysgalactiae, S. chromogenes, and S. uberis was confirmed with tdna intergenic spacer PCR analysis (Baele et al., 2000). In preparation of cross-streaking procedures, all isolates were streaked on Colombia sheep blood agar. After 24hrs at 37 C, several colonies were picked with a sterile inoculation loop and a suspension was prepared by vortexing in 3 ml PBS until a density of McF 0.5 was reached. Earlier described cross-streaking methods (Barefoot and Klaenhammer, 1983; Woodward et al., 1987) were assessed to evaluate the inhibitory interaction between different bacterial isolates in vitro. However, because of disappointing results and technical difficulties [e.g. melting of the polystyrene Petri-dishes when using chloroform to kill residual S. chromogenes as described by Woodward et al. (1987)], both methods were combined into a more practical method. The suspension of S. chromogenes was therefore inoculated as a centre-streak with a width of 5 mm (Figure 1) on to polystyrene Petri-dishes (90 mm in diameter) containing Colombia sheep blood agar, and incubated for 24hrs at 37 C, under aerobic conditions. After inspection of bacterial growth, the agar was circularly loosened from the sides of the dish using a sterile 0.5 mm inoculation needle (Novolab) and was turned upside down into the cover of the dish. Using a sterile cotton swab, a 10-3 dilution of the 0.5 McF suspension of a particular major mastitis pathogen was swabbed on the agar in bottomup position to achieve full coverage. After incubation (24hrs, 37 C, and aerobic conditions) the plates were re-examined for bacterial growth and inhibition of the mastitis pathogen within and adjacent to the centre-streak of the S. chromogenes under study. All mastitis pathogens were tested against the 10 S. chromogenes isolates in twofold. Every experiment included negative and positive control plates. Additionally, all S. chromogenes strains were cross-streaked in twofold against themselves. 169

184 Chapter 7.2 Patterns of Inhibition The possible inhibition patterns encountered in this study are presented in Figure 1. Zones are measured along the axis ( ) in mm: C = Central-streak zone of S. chromogenes at backside of the agar (always 5 mm); T = zone of Total growth inhibition of the major mastitis pathogen; P = zone of Partial growth inhibition of the major mastitis pathogen (smaller colonies and/or less colonies as on positive control plate); N = zone of No growth inhibition of the major mastitis pathogen (same size and numbers of colonies as on positive control plate). The zones T, P, and N are not necessarily present in all samples. Figure 1. Patterns of inhibition of major mastitis pathogen growth by S. chromogenes (versus S. aureus strain as an example). All zones were, if present, measured along an axis across the plate, perpendicularly on the centre-streak (zone C). Measurements were done on the left and right side of the centralstreak zone and averaged, and sum up to 45 mm per default (including 2.5 mm for zone C), half the diameter of the Petri-plates (90 mm in diameter) used. The central-streak zone C (S. chromogenes at the backside of the blood agar) had a width of 5 mm per default, but the mastitis pathogen within this zone could have 2 appearances: p = partial growth inhibition (smaller colonies and/or less colonies as on positive control plate), and t = total growth inhibition. Other possible zones, but not necessarily present in all samples, were: T = zone of Total growth inhibition of the major mastitis pathogen, adjacent to zone C; P = zone of Partial growth inhibition (smaller colonies and/or less colonies as on positive control plate), adjacent 170