The influence of the rearing period on intramammary infections in Swiss dairy heifers: A cross-sectional study

M.J. Bludau; A. Maeschli; F. Leiber; P. Klocke; J.A. Berezowski; M. Bodmer; B. Vidondo

Accepted Manuscript

Title: The influence of the rearing period on intramammary infections in Swiss dairy heifers: A cross-sectional study

Author: M.J. Bludau A. Maeschli F. Leiber P. Klocke J.A. Berezowski M. Bodmer B. Vidondo

PII: DOI:

Reference:

S0167-5877(16)30128-3

http://dx.doi.Org/doi:10.1016/j.prevetmed.2016.04.013 PREVET 4023

To appear in:

PREVET

Received date: Revised date: Accepted date:

31-7-2015 19-4-2016 25-4-2016

Please cite this article as: Bludau, M.J., Maeschli, A., Leiber, F., Klocke, P., Berezowski, J.A., Bodmer, M., Vidondo, B., The influence of the rearing period on intramammary infections in Swiss dairy heifers: A cross-sectional study.Preventive Veterinary Medicine http://dx.doi.org/10.10167j.prevetmed.2016.04.013

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The influence of the rearing period on intramammary infections in Swiss

dairy heifers: A cross-sectional study

M. J. Bludau a, A. Maeschli a, F. Leiber a, P. Klocke M, J.A. Berezowski c, M. Bodmer b,*,J B. Vidondo c,J

a Department of Livestock Science, Research Institute of Organic Agriculture (FiBL), Ackerstrasse 113, 5070 Frick, Switzerland

b Clinic for Ruminants, University of Berne, Bremgartenstrasse 109a, 3012 Bern, Switzerland c Veterinary Public Health Institute, Schwarzenburgstrasse 155, 3097 Liebefeld, Switzerland

1, Present address: bovicare GmbH, Hermannswerder, Haus 14, 14473 Potsdam, Germany. J these authors contributed equally

* Corresponding author. Tel.: +41 (0)31 631 23 45 E-mail address: michele.bodmer@vetsuisse.unibe.ch

Abstract

Healthy replacement heifers are one of the foundations of a healthy dairy herd. Farm management and rearing systems in Switzerland provide a wide variety of factors that could potentially be associated with intramammary infections (IMI) in early lactating dairy heifers. In this study, IMI with minor mastitis pathogens such as coagulase-negative staphylococci (CNS), contagious pathogens, and environmental major pathogens were identified. Fifty-four dairy farms were enrolled in the study. A questionnaire was used to collect herd level data on housing, management and welfare of young stock during farm visits and interviews with the farmers. Cow-level data such as breed, age at first calving, udder condition and swelling, and calving ease were also recorded. Data was also collected about young stock that spent a period of at least 3 months on an external rearing farm or on a seasonal alpine farm. At the quarter level, teat conditions such as teat lesions, teat dysfunction, presence of a papilloma and teat length were recorded. Within 24 hours after parturition, samples of colostral milk from 1'564 quarters (391 heifers) were collected aseptically for bacterial culture. Positive bacteriological culture results were found in 49% of quarter samples. Potential risk factors for IMI were identified at the quarter, animal and herd level using multivariable and multilevel logistic regression analysis. At the herd level tie-stalls, and at cow-level the breed category "Brown cattle" were risk factors for IMI caused by contagious major pathogens such as Staphylococcus aureus (S. aureus). At the quarter-level, teat swelling and teat lesions were highly associated with IMI caused by environmental major pathogens. At the herd level heifer rearing at external farms was associated with less IMI caused by major environmental pathogens. Keeping pregnant heifers in a separate group was negatively associated with IMI caused by CNS. The odds of IMI with coagulase-negative staphylococci increased if weaning age was less than 4 months and if concentrates were fed to calves younger than 2 weeks. This study identified herd, cow- and quarter-level risk factors that may be important for IMI prevention in the future.

Keywords: heifer; rearing; mastitis; multilevel logistic regression; risk factor.

1. Introduction

It is well accepted that good udder health is crucial for the economic success of a dairy farm. However, farmers often pay less attention to the rearing of young stock than to the management of adult cows, even though it has been shown that adequate management of young stock can avoid future udder health problems (Le Cozler et al., 2008). Recent studies distinguish between clinical and subclinical heifer mastitis, depending on the presence or absence of inflammatory signs in the mammary gland (Piepers et al., 2010). Heifer mastitis is a disease which may be increasing in importance in different parts of the world. In New Zealand, 21.5 % of quarters of heifers had a positive bacterial culture result (Compton et al., 2007) and in a Belgium study 25% of quarters of early postpartum heifers were culture positive (Piepers et al., 2010). Although CNS is the most frequently isolated pathogen in heifers (Fox, 2009; Piepers et al., 2011) CNS is traditionally categorized as minor pathogen and only in rare cases results in clinical mastitis in heifers (Lam et al., 1997). Piepers et al. (2011) reported that CNS infection in heifers in early lactation was very common (72% of tested quarters infected) and was associated with fewer cases of clinical mastitis (CM) throughout the following lactation compared to non-infected herd mates. In Piepers' study the occurrence of IMI caused by contagious pathogens such as S. aureus and Streptococcus agalactiae (S. agalactiae), and environmental pathogens such as Streptococcus uberis (S. uberis), Streptococcus dysgalactiae (S. dysgalactiae) and Escherichia coli (E. coli) were less prevalent in early lactation heifers.

Several studies have identified potential risk factors for heifer mastitis (De Vliegher et al., 2004; Svensson et al., 2006; Piepers et al., 2011; De Vliegher et al., 2012; Krömker et al., 2012, Archer et al., 2013; Bludau et al., 2014; Abb-Schwedler et al., 2014). It is reported to be a multifactorial disease influenced by climate, season, geographical location and genetic background. In particular management factors such as social stress, type of housing system, lack of

environmental hygiene or inadequate nutrition have been associated with IMI in first lactation heifers (Svensson et al., 2006; Nyman et al., 2009; Santman-Berends et al., 2012). Depending on the housing and management system, heifers may suffer from social stress during multiple group changes around calving; for example being moved from the pregnant heifer group to the dry cow group, then to the transition cow group and finally to a lactating cow group. In addition to these social challenges, heifers must deal with physiological changes related to growing, calving and the first lactation. Collectively these factors can negatively influence their immune system and elevate the risk for all infections, including IMI (Mallard et al., 1998; Hultgren and Svensson, 2009).

In Switzerland, a variety of farm management systems exist, including loose housing systems and traditional tie-stall systems. There is a high degree of animal movement in Switzerland due to young stock rearing on specialized farms in which animals from different farms are comingled (Gloor et al., 2007). Calves are moved to external rearing farms at weaning age and are kept there until shortly before calving. Approximately 25% of young stock (personal communication S. Scharrer, Food Safety and Veterinary Office Switzerland, FSVO) are sent to communal alpine pastures in the summer (June to September) where animals from many different herds of origin are comingled. Mountain pastures offer some challenges to heifers such as the variable nutrient composition of the pasture, steep and heterogeneous topography and a colder, more variable climate than in the lowlands (Ruhland et al., 1999; Leiber et al., 2006). On the other hand, alpine grazing of the young stock has been reported to strengthen the health of heifers. (Kuenzi et al., 1988; Ruhland et al., 1999). Young animals are challenged by transport stress, social stress when introduced to a new group, exposure to a new microbiological flora in the new environment and exposure to the risk of transmission of mastitis pathogens through inter-suckling or flies (De Vliegher et al., 2012).

Most of the heifer mastitis studies reported in the literature focus on the period around calving (De Vliegher et al., 2004; Nyman et al., 2009; Piepers et al., 2011; Kromker et al., 2012). To

our knowledge, only one Swedish study reported different factors associated with the entire rearing period (Hultgren et al., 2009). The purpose of our study was to address this gap by investigating potential associations between herd, animal, and quarter level factors present during the rearing period, and the occurrence of heifer mastitis in Swiss dairy heifers.

2. Materials and methods

The study was conducted in accordance with the animal welfare legislation of Switzerland and all ethical aspects of the study were approved by the Federal Food Safety and Veterinary Office.

2.1 Herd selection and sample size calculation

All herds included in the study were affiliated with one of the main Swiss breeding organizations: Swiss Brown Cattle Breeders' Federation, Zug, Switzerland; Holstein Breeders' Federation, Posieux, Switzerland; and the Swissherdbook, Zollikofen, Switzerland. Herds were included if they had at least 20 lactating cows of one of the breeds: Holstein Friesian, Red Holstein, Brown Swiss, Original Brown, Braunvieh (BS x Original Brown), Swiss Red Pied, Simmental or Montbeliarde, and were located at the north side of the Alps. Participation in the study was voluntary. Owners of 856 farms were invited to participate in the study, of these, 224 farm owners agreed to participate, and 72 of these farms were randomly selected for inclusion in the study, using the RAND function in Excel. Piepers et al. (2011) reported 20% loss of samples due to contamination. In anticipation of this potential loss we adjusted our sample size accordingly. Our farm sample size was estimated to be 60 farms and was adjusted to 72 farms to account for these potential sample losses (60*1.20 = 72). We estimated an average of 6 heifers per farm, providing a minimum sample size of around 1, 440 (60 *6*4 = 1,440) quarters. Logistic regression of a binary response variable (Y) on a binary independent variable (X) with a sample size of 1'564, (final

sample of quarters, of which 50% are in the group X=0 and 50% are in the group X=1) was calculated to have 78% power, at a 0.05 significance level (Hsieh et al., 1998, PASS Software 2014).

2.2 Data collection

The 72 selected dairy farms, together with their cooperating 54 heifer rearing operations, and 33 alpine farms where the cattle spend the summer period, were visited one time between May 2012 and August 2013. Demographic farm data and data about management practices were collected using a questionnaire during a farmer interview conducted by a research team member while visiting the farm. During the farm visit, animals, barns and pastures were inspected and evaluated for hygiene and animal welfare. Variables collected at the herd, heifer- and quarter-level can be found in Tables 1A and 1B.

During farm visits dairy farmers were trained to aseptically collect milk samples according to the guidelines of the National Mastitis Council (National Mastitis Council, 2004). There is no consensus concerning the case definition of IMI. In a New Zealand study, IMI status was assessed with duplicate samples collected at one time point (Compton et al., 2007) whereas in two Belgium studies several consecutive samples were collected to define IMI status with different types of bacteria (Piepers et al., 2010; Piepers et al., 2011). We chose to use duplicate samples from each quarter at the time of calving, and farmers were instructed to aseptically collect duplicate individual quarter milk samples from each heifer immediately after parturition for the duration of the study. All heifers that calved were scored for cow-level risk factors immediately after parturition by farmers All milk samples were frozen immediately after collection. Completed score sheets and frozen milk samples were sent by priority mail to the laboratory the day after collection (ILS, Institute for Food Safety and Hygiene, Zurich, Switzerland). Transport duration was a maximum of 24 hours and

samples were shipped in an insulated styropor box to prevent heat damage. The sampling period was from May 2012 until October 2013.

2.2.1 Herd-level

Herd-level data including demographic information, housing, management and feeding of young stock were collected during an interview with the farmer (Table 1B). Variables originally recorded as continuous data were transformed to categorical variables using their terciles to define three categories. Herd size, average milk production per cow and year and average yield corrected herd somatic cell count (CHSCC) for the 2012 calendar year were calculated from dairy herd improvement (DHI) data obtained from the Swiss breeding organizations, as described by Liveaart et al. (2007) and Ivemeyer et al. (2009).

2.2.1.1 Assessment of the welfare status

Most published protocols for assessing cattle welfare focus on the welfare of producing dairy cattle or fattening cattle. For this reason a specific protocol for evaluation of welfare in rearing dairy cattle was developed for this study (see supplementary online material).

Inspired by the work of Bartussek (1996), Sundrum (2007), Schaeffer and Von Borell (2007), Von Borell et al. (2007), Keyserling et al. (2009) and the Welfare Quality Assessment protocol for cattle (2009), five key areas relating to the natural behaviour of animals were considered: 1) assessment of locomotion area (supplementary material Table S1), 2) assessment of resting area (supplementary online material table S2), 3) assessment of feeding area and feeding management (supplementary online material, table S3), 4) assessment of general climatic aspects (supplementary online material, table S4), 5) assessment of general animal health and hygiene (supplementary

online material, table S5). A total of 20 score points were allocated to each key area providing a maximum of 100 score points.

Welfare was assessed separately in each of the following 5 calf and heifer groups: 1) calves up to 3 weeks of age, 2) calves 4 to 8 weeks of age, 3) calves 9 to 15 weeks of age, 4) young prepuberal heifers and 5) postpuberal and pregnant heifers.

2.2.2 Cow-level

A score sheet for recording information about calving ease, udder health, general health and milking characteristics of the heifer was completed by the farmer at the time that milk samples were collected. Cow-level variables are reported in Table 1A. Variables originally recorded as continuous were transformed into categorical variables with three categories using their terciles.

2.2.3 Quarter-level

Quarter-level data included teat lesions, teat swelling, presence of papilloma, and teat length (categorized as: long = > 7cm, medium = 5—7 cm, short =< 5cm) were recorded immediately after parturition by measuring the length from the teat tip to the base with a measuring tape (Table 1A). Teat dysfunctions, such as teat lacerations or additional teat canals with or without gland tissue were also recorded.

2.2.3.1 Bacteriological analysis and definition of IMI-Status

Bacteriological analysis was performed according to NMC standards (National Mastitis Council, 2004). Coagulase-negative staphylococci were not further specified. A sample with more than 2 different bacterial species was classified as contaminated and excluded from further analysis.

If one of the duplicate samples was contaminated, the findings from the uncontaminated duplicates were used to identify an infection.

Quarters were divided into four groups: non-infected, infected with CNS, infected with contagious major pathogens or infected with environmental major pathogens as previously reported (Piepers et al., 2011). Definition of the infection status of quarters using duplicate samples is explained in Table 2.

2.6 Statistical Analyses

The udder quarter was the unit of this analysis. To take into account clustering of quarters within heifers and heifers within herds, heifer and herd were included as random effects in the multilevel models. Three-level logistic regression mixed models with random intercepts were fit using Stata 13. The log likelihood for this type of model has no closed form, so it was approximated by adaptive Gaussian quadrature (StataCorp, 2013). Prior to multivariable analysis, univariable three-level models were used to test the associations between the binary outcome variables (a) IMI with contagious major pathogens (1= infected; 0= non-infected), (b) IMI with CNS (1= infected; 0= non-infected), (c) IMI with environmental major pathogens (1= infected; 0= non-infected) and potential risk factors. Variables with P <0.20 were kept for further analysis. Collinearity between potential risk factor variables was evaluated using Pearson's and Spearman's rank correlation. If two risk factors had a correlation coefficient of >0.60, the one with the lower P-value in the univariable analysis and considered biologically more plausible was included in the multivariable model. For each outcome (a-c), a separate multivariable model was built, following the recommendations of the online course of the Centre for Multilevel Modelling, Bristol, UK

(http://www.bristol.ac.uk/cmm/learning/online-course/index.html). We first computed the null models, and compared them with models that included every single explanatory variable, one at a

time. Likelihood ratio (LR) tests, which do not rely on the assumption of an asymptotic normal sampling distribution, were used to prove that the additional predictors significantly improved the fit of the models. We then proceeded to build more complex models adding one variable at a time and running successive Likelihood Ratio (LR) tests. We used a significance level of P <0.05 to decide which variables remained in the final models. This model selection procedure accounts, at the same time, as a control for confounding, because any variable that renders a significant LR test, and thus influences the model, is retained. Every three-level model was compared with its single-level model counterpart using LR-tests and found to significantly better fit the data.

The Variance Partition Coefficients (VPC) of IMI with contagious and environmental major pathogens and CNS at the herd, heifer and quarter level for both null and final models were estimated. The variance at the quarter level was fixed (constant) to n /3 = 3.29 (with n=3.1416) for the logistic model (Dohoo et al., 2001; Snijders and Bosker, 2012). Intraclass correlation coefficients (ICCs) that measure the similarity of the observed responses within a given level or cluster were also calculated.

3. Results

3.1 Description of the sample

Colostrum samples were collected from 528 heifers on the original 72 farms. Eighteen farms left the study before it was completed. Twelve of these were excluded because of too few submitted samples, one farm because of contamination of samples, one farm because of contamination of samples and too few submitted samples, and 4 farms had less than 3 heifers that calved during the sampling period. The mean herd size was 31 cows (SD=16). The mean milk yield per herd was 7438 kg/cow/year (SD=940 kg/cow/year). Twenty-five farms (46%) had loose housing systems, 28 farms (52%) had tie-stalls and one had a mixed system. Twenty-six farms (48%) were located in the

lowlands (cadastral zone: lowland), and 28 farms (52%) were located in mountainous areas (cadastral zones: mountain zone1-4). The mean CHSCC in the year 2012 was 134,000 cells/mL (SD=72,000 cells/mL). Young stock from 22 dairy farms (41%) were reared at specialized rearing farms together with animals from other farms. Thirty-two dairy farms (59%) housed their young stock at the farm of origin. The young stock of 38 farms (70%) were sent to alpine pastures during the summer.

3.2 Results of the bacteriological examination

The complete spectrum of detected mastitis pathogens is presented in Tables 3 and 4. A total of 1564 quarters were bacteriologically examined and in 51.0% (n=797) at least one pathogen was detected (Table 3); 42.5% were infected with CNS, 6.1% with one or two environmental major pathogens (3.2% S. uberis, 1.9% coliforms and 1.2% S. dysgalactiae) and 4.8% with a contagious major pathogen. Pathogens categorized as contagious major pathogens (in this study only S. aureus) were found in 10.5% of the heifers and in 15 of 54 (27.8%) dairy farms. Environmental major pathogens were found in 19.2% of the heifers and 34 dairy farms (63%). Coagulase-negative staphylococci were diagnosed in 69.1% of the heifers; there was no farm free of CNS.

3.3 Results of the farm's Welfare status

Farms frequently had deficiencies in the key areas of locomotion area and feeding infrastructure and management (Table 5). There were correlations between the welfare scores of different groups, and for this reason only the total score was tested in the final model.

3.5 Risk factor analysis 3.5.1. Univariable analysis

The results of the univariable analysis are presented in Tables 6A-D, which includes the variables that were considered in the final models. The individual SCC of the heifers at first test day was associated with the prevalence of environmental major pathogens only (P=0.003), and not with the prevalence of contagious major pathogens (P=0.65) or CNS (P=0.29). The association was, however, weak and not significant in the final multivariable model. 3.5.2. Multivariable multilevel logistic regression models and analysis of variance Model Staphylococcus aureus. Heifers of the breed category "Brown cattle" were more likely to have IMIs with S. aureus (odds ratio; OR 11.2) than the other breeds. Heifers housed in tie-stalls (OR 26.9) were more likely to have IMIs with S. aureus than heifers in loose housing systems. The results of the multivariable analysis with the corresponding 95% confidence intervals and P-values are presented in Table 7.

Model environmental major pathogens - Quarters with swollen teats (OR 2.67) or teat lesions (OR 37.7) were more likely to be infected with environmental major pathogens than heifers with normal teats. Teat lesions were recorded in only 7 quarters (0.4%). Heifers raised on specialized rearing farms (OR 0.29) were less likely to be infected with environmental major pathogens than heifers raised at their home farms.

Model CNS. The separation of pregnant heifers from younger replacement animals had the strongest effect on the presence of CNS, showing a protective effect against CNS IMI. Three additional factors were associated with the presence of CNS. During model selection the Log-likelihood ratio tests demonstrated that the following 3 final models with variables were equally valid (Table 7): 1) separation of pregnant heifers and feeding concentrates to calves (P = 0.006), 2) separation of pregnant heifers and the weaning age (P = 0.002) and 3) separation of pregnant heifers

and welfare of calves (P = 0.003), although this last model had a very small effect with an OR very close to one. A special group for heifers decreased the odds of CNS infection in each of the 3 models (OR 0.2-0.32). Feeding concentrates to replacement calves younger than 2 weeks and weaning calves before 4 months of age increased the odds of CNS infection.

Variance components - For contagious pathogens i.e. S. aureus, the random effect variance was more evenly distributed in the three levels than for CNS and environmental pathogens (Table 8), for which most of the unexplained variance remained at the levels of quarter and cow. In the final model, the random variance at herd level still amounts to 24% for contagious pathogens. Two farms stood out with a very high proportion of infected quarters (15 and 19 respectively of the total of 75 infected quarters in all farms).

The ICC at the farm level was higher for contagious (24%) than for CNS (5-6%) or environmental (7%) major pathogens (Table 9). A similar pattern was found at the cow level but with much higher ICC values, 70%, 56% and 51%, for contagious, environmental, and CNS, respectively.

4. Discussion

Prevalence of Staphylococcus aureus and its risk factors

In this study the "housing system" was strongly associated with the prevalence of S. aureus IMI at calving. The prevalence of IMI at calving was higher in heifers housed in tie-stall barns, which, in Switzerland, are often older than loose housing barns. Although S. aureus is classified as an udder-associated pathogen and the primary reservoir is the bovine udder, it has been reported that some mastitis causing strains can be found on body sites close to the udder and in the immediate environment of the cow (Anderson et al., 2012). Recently, strains of S. aureus causing persistent

IMI were reported to have the ability to form biofilms (Veh et al., 2015). Therefore, these pathogens may persist in the environment forming reservoirs more often in old barns, which are usually more difficult to clean. The finding that heifers of the breed category "Brown cattle" were more often infected with contagious major pathogens confirmed a previous Swiss study (Ivemeyer et al., 2009), which found an association between poor udder health and the breed category "Brown cattle." "Brown cattle" farms are traditionally located in Eastern Switzerland, where communal alpine farms tend to be bigger than the ones in central and western Switzerland and are supplied by a high number of different farms of origin. Communal alpine pasturing has previously been reported to be a risk for new infections with S. aureus (Voelk et al., 2014).

The rare occurrence of S. aureus compared to environmental mastitis pathogens in the present study confirms the results of Kretzschmar et al. (2013) who investigated mastitis management in Swiss dairy farms. In our study, only 15 farms with S. aureus infected heifers were identified, and therefore the results of this study relating to S. aureus infection should be interpreted with care. It is possible that some IMI with S. aureus may have been missed in our study. The sensitivity of bacterial culture from a single sample using an inoculum of 0.1ml is reported to be low (74.5%) (Sears et al., 1990). The sensitivity for detection of S. aureus in our study may have been higher than reported by Sears et al. (1990) because duplicate samples were collected. However, the duplicate samples were both collected at one time point and in many cases one sample was contaminated and excluded from our analysis and only a standard inoculum of 0.01ml was used for culture. This has been reported to increase the risk of obtaining false negative results (Sears et al., 1990).

Thirty-five of the 75 quarters infected with contagious mastitis pathogens were found at two farms with a known high rate of infection with S. aureus. Both farms were located in Eastern Switzerland and had "Brown cattle" breed cattle which were housed in tie-stalls. One of these farms

practiced communal alpine farming during the summer with all lactating cows and heifers being moved to a communal pasture. With the results of the present study it is not possible to determine which is the more important risk factor: 1) cattle of the breed "Brown cattle" being more susceptible to contagious mastitis, or 2) the management practices conducted in areas where "Brown cattle" are raised. The risk factors identified for IMI with S. aureus - breed "Brown cattle" and tie-stall housing - are not popular in the EU except for Northern Italy and Austria and they can be considered to be specific for pre-alpine and alpine regions. Today more than 70% of the dairy herds in EU are Holstein-Friesian (EFSA, 2009).

Prevalence of environmental major pathogens and their risk factors

For IMI with environmental pathogens, the condition of the teat (i.e. injured skin of the quarter, teat edema) has been reported to be associated with IMI. De Vliegher et al. (2004) reported that the load of bacteria at the teat end was a crucially important risk factor for IMI with environmental pathogens, and, poor heifer hygiene has been reported to be associated with CNS IMI (Piepers et al., 2011). Waage et al. (2001), reported that teat and udder edema were associated with clinical mastitis. They suggested that this may be due to impairment of blood circulation in the affected area, which in turn impairs the transport of immune cells into the affected area. Mechanical forces during milking may have a much greater effect on edematous teats than on non-edematous ones. The same study reported that udder and teat edema may cause milk leakage which may be associated with clinical mastitis.

In our study heifer rearing on a specialized rearing farm was protective against environmental major pathogens. Reasons for this may be: 1) most heifer rearing farms do not keep lactating cows, so there is no exposure to infected adults, or 2) these specialized farmers pay more

attention to young stock than dairy farmers who rear heifers on their own farm, because rearing for other farmers is their main income.

Prevalence of coagulase-negative staphylococci and their risk factors

The importance of CNS as a cause of bovine mastitis is still uncertain (De Vliegher et al., 2009). More than 45 different species and subspecies of the CNS group exist and 12 of them are regularly found in milk of dairy cows (Piessens et al., 2011). The pathogenicity of this group has yet to be established. It is generally accepted that major pathogens induce clinical heifer mastitis. However, IMI with CNS does not negatively influence subsequent productivity (De Vliegher et al., 2012). Our study is in agreement with other studies that reported CNS as the most prevalent mastitis pathogen in heifers (Fox, 2009; Piepers et al., 2011). In our study the presence of CNS was associated with 4 management factors: no separation of pregnant heifers, early provision of concentrates to calves, low weaning age and better welfare of calves. Separation of pregnant heifers from younger animals and adults may indicate a higher degree of professionalism in the farmer. These farmers may provide better management and feeding of young stock in order to reach set rearing targets. Weaning at an older age and feeding of concentrates later in heifer rearing are management practices that have been reported to be associated with more extensive, pasture-based rearing systems and have been reported to be protective effect against CNS IMI. The very early provision of concentrates to calves is often linked to reduced milk feeding (<10% of body weight) as reviewed by Drackley (2008), and therefore to inadequate nutrition, which might lead to insufficient development of heifers, and to an impaired immune system which may increase udder susceptibility to infection.

In contrast to other studies (Bielfeldt et al., 2006; von Keyserlingk et al., 2009) an unexpected finding in our study was, that higher welfare status of calves was associated with the

presence of CNS IMI. Most of the farms in our study had welfare scores higher than 50% and the presence of CNS was not directly associated with disease.

Partition of variance components

The impact of the herd-level was more important for S. aureus than for the other pathogens, suggesting that there may be other herd-level risk factors not yet explored for this pathogen. The highest variation in IMI with environmental major pathogens remained at the quarter-level, and at the heifer level for CNS IMI.

The herd itself had a higher impact on the risk of IMI with contagious pathogens as previously reported for S. aureus (Voelk et al., 2014) and this may be explained by the probability of transmission of S. aureus being much higher in herds with a greater number of cows shedding the pathogen.

Strengths and Limitation of this study

The case definition for IMI varies across studies and this makes comparison of study results difficult. In order to better compare study results an international consensus for the definition of IMI with different pathogens in cows and heifers is needed, based on the work of Dohoo et al. (2011) and Andersen et al. (2010). A potential sampling bias in our study could have been reduced by limiting the proximity of sampled farms to our institute and collecting data exclusively by trained veterinarians. We opted however for increased representativeness of our sample by random sampling from the volunteer farms regardless of their location. This increased representativeness allowed us to include and detect strong previously unreported risk factors, , such us breed and type of housing.

Since milk sampling by trained veterinarians was not possible, and was done by the farmers, 14 farms had to be excluded because of poor sample quality and too few submitted samples. Up to 20% missing values due to poor sample quality has also been reported by Piepers et al., (2011).

While herd level information was collected by a single trained veterinarian, potential bias might have been introduced during the assessment of cow level variables by the participating farmers. Even though our study sample size was twice as large as the sample reported in previous studies (Piepers et al., 2011), the smallest detected ORs ranged from 2.2 and 2.7 at the farm level. There may still be weaker associations that could be identified with a larger sample and these should be pursued with future research.

Conclusion

In this study breed and type of housing were both associated with increased S. aureus IMI in Swiss heifers. Rearing of heifers on specialized rearing farms, and separating pregnant heifers from younger and adults animals were protective against IMI with environmental and CNS pathogens.

Conflict of interest statement

None of the authors of this paper has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.

Acknowledgements

This research was funded by the Swiss Federal Food Safety and Veterinary Office (FSVO; project 1.11.13). We especially thank all participating farmers for their hospitality and cooperation, Mirjam Holinger and the rest of the Animal Science team of FiBL for their advice and support, and

Gertraud Schupbach and Bart van den Borne from the VPHI Institute in Bern for insightful discussions on multilevel analysis.

References

Abb-Schwedler, K., Maeschli, A., Boss, R., Graber, H.U., Steiner, A., Klocke, P., 2014. Feeding mastitis milk to organic dairy calves: effect on health and performance during suckling and on udder health at first calving. BMC Vet. Res. 10, 267. doi:10.1186/s12917-014-0267-7

Anderson, K. L., Lyman, R., Moury, K., Ray, D. ,Watson, D.W., Correa, M. T., 2012. Molecular epidemiology of Staphylococcus aureus mastitis in dairy heifers. J. Dairy Sci. 95, 4921-4930.

Andersen, S., Dohoo, I.R., Olde Riekerink, R., Stryhn, H., 2010. Diagnosing intramammary infections: Evaluating expert opinions on the definition of intramammary infection using conjoint analysis. J. Dairy Sci. 93, 2966-2975. doi:10.3168/jds.2009-2726

Archer, S.C., Mc Coy, F., Wapenaar, W., Green, M.J., 2013. Association between somatic cell count early in the first lactation and the longevity of Irish dairy cows. J. Dairy Sci. 96, 2939-50. doi:10.3168/jds.2012-6115

Bartussek, H., 1996a. Tiergerechtheitsindex für Kälber, TGI 35/L 1995, Stand Mai 1996,

Veröffentlichungen der Bundesanstalt für alpenländische Landwirtschaft (BAL) Gumpenstein, Irdning, Austria.

Bartussek, H., 1996b. Tiergerechtheitsindex für Rinder, TGI 35/L 1996, Stand Mai 1996,

Veröffentlichungen der Bundesanstalt für alpenländische Landwirtschaft (BAL) Gumpenstein, Irdning, Austria, http://www.bartussek.at/pdf/tgirinder.pdf (accessed November 2015).

Bielfeldt, J.C., Tölle, K.-H., Badertscher, R., Krieter, J., 2006. Longevity of Swiss Brown cattle in different housing systems in Switzerland. Livest. Sci. 101, 134-141. doi: 10.1016/j .livprodsci.2005.10.023

Bludau, M.J., Maeschli, A., Leiber, F., Steiner, A., Klocke, P., 2014. Mastitis in dairy heifers: prevalence and risk factors. Vet. J. 202, 566-72. doi:10.1016/j.tvjl.2014.09.021

Compton, C.W.R., Heuer, C., Parker, K., McDougall, S., 2007. Risk Factors for

Peripartum Mastitis in Pasture-Grazed Dairy Heifers. J. Dairy Sci. 90, 4171-4180.

De Vliegher, S., Laevens, H., Barkema, H.W., Dohoo, I.R., Stryhn, H., Opsomer, G., de Kruif, A., 2004. Management Practices and Heifer Characteristics Associated with Early Lactation Somatic Cell Count of Belgian Dairy Heifers. J. Dairy Sci. 87, 937-947. doi: 10.3168/j ds.S0022-0302(04)73238-5

De Vliegher, S., Zadoks, R., Barkema, H., 2009. Heifer and CNS mastitis. Vet. Microbiol.134, 1-2.

De Vliegher, S., Fox, L.K., Piepers, S., McDougall, S., Barkema, H.W., 2012. Invited review: Mastitis in dairy heifers: Nature of the disease, potential impact, prevention, and control. J. Dairy Sci. 95, 1025-1040. doi:10.3168/jds.2010-4074

Dohoo, I.., Tillard, E., Stryhn, H., Faye, B., 2001. The use of multilevel models to evaluate sources of variation in reproductive performance in dairy cattle in Reunion Island. Prev. Vet. Med. 50, 127-144. doi:10.1016/S0167-5877(01)00191-X

Dohoo, I.R., Smith, J., Andersen, S., Kelton, D.F., Godden, S., 2011. Diagnosing intramammary infections: evaluation of definitions based on a single milk sample. J. Dairy Sci. 94, 250-61. doi:10.3168/jds.2010-3559

Drackley, J. K., 2008. Calf nutrition from birth to breeding. Vet. Clin., North Am. Food Animal

Practice 24, 55-86.

EFSA (European Food Safety Authority), 2009. Scientific report of EFSA prepared by

the Animal Health and Animal Welfare Unit on the effects of farming systems on dairy cow welfare and disease. Annex to the EFSA Journal 1143, 1-7.

Fox, L.K., 2009. Prevalence, incidence and risk factors of heifer mastitis. Vet. Microbiol. 134, 82-8. doi:10.1016/j.vetmic.2008.09.005

Gloor, M., Kaufmann, T., Peterhans, E., Zanoni, R., Steiner, A., Kirchhofer, M., 2007. Heifer

breeding farms as a source for spreading the BVD virus? Schweiz. Arch. Tierheilkd. 149, 34551. doi:10.1024/0036-7281.149.8.345

Hsieh, F.Y., Bloch, D.A., Larsen, M.D., 1998. A simple method of sample size calculation for linear and logistic regression. Stat. Med. 17, 1623-34.

Hultgren, J., Svensson, C., 2009. Heifer rearing conditions affect length of productive life in Swedish dairy cows. Prev. Vet. Med. 89, 255-64. doi:10.1016/j.prevetmed.2009.02.012

Ivemeyer, S., Walkenhorst, M., Heil, F., Notz, C., Maeschli, A., Butler, G., Klocke, P., 2009. Management factors affecting udder health and effects of a one year extension program in organic dairy herds. Animal 3, 1596-604. doi:10.1017/S1751731109990498

Kretzschmar, L., van den Borne, B.H.P., Kaufmann, T., Reist, M., Strabel, D., Harisberger, M., Steiner, A., Bodmer, M., 2013. Mastitis management in Swiss dairy farms with udder health problems. Schweiz. Arch. Tierheilkd. 155, 453-62. doi:10.1024/0036-7281/a000491

Krömker, V., Pfannenschmidt, F., Helmke, K., Andersson, R., Grabowski, N.T., 2012. Risk factors for intramammary infections and subclinical mastitis in post-partum dairy heifers. J. Dairy Res. 79, 304-9. doi:10.1017/S0022029912000222

Kuenzi, N., Leuenberger, H., Michel, A., 1988. Alpine Pastures: an important facet of Swiss cattle production. J. Anim. Breed. Genet. 279-293.

Lam, T.J., Schukken, Y.H., van Vliet, J.H., Grommers, F.J., Tielen, M.J., Brand, A., 1997. 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, 17-22.

Le Cozler, Y., Lollivier, V., Lacasse, P., Disenhaus, C., 2008. Rearing strategy and optimizing first-calving targets in dairy heifers: a review. Animal 2, 1393-404. doi:10.1017/S1751731108002498

Leiber, F., Kreuzer, M., Leuenberger, H., Wettstein, H.R., 2006. Contribution of diet type

and pasture conditions to the influence of high altitude grazing on intake, performance and composition and renneting properties of the milk of cows. Anim. Res. 55, 37-53.

Lievaart, J.J., Kremer, W.D.J., Barkema, H.W., 2007. Short communication: Comparison of bulk milk, yield-corrected, and average somatic cell counts as parameters to summarize the subclinical mastitis situation in a dairy herd. J. Dairy Sci. 90, 4145-4148. doi:10.3168/jds.2006-871

Mallard, B.A., Dekkers, J.C., Ireland, M.J., Leslie, K.E., Sharif, S., Vankampen, C.L., Wagter, L., Wilkie, B.N., 1998. Alteration in immune responsiveness during the peripartum period and its ramification on dairy cow and calf health. J. Dairy Sci. 81, 585-95.

Nyman, A.-K., Emanuelson, U., Gustafsson, A.H., Persson Waller, K., 2009. Management practices associated with udder health of first-parity dairy cows in early lactation. Prev. Vet. Med. 88, 138-49. doi:10.1016/j.prevetmed.2008.08.005

PASS 13 Power Analysis and Sample Size Software (2014). NCSS, LLC. Kaysville, Utah, USA, ncss.com/software/pass.

Piepers, S., Opsomer, G., Barkema, H.W., de Kruif, a, De Vliegher, S., 2010. Heifers infected with coagulase-negative staphylococci in early lactation have fewer cases of clinical mastitis and higher milk production in their first lactation than noninfected heifers. J. Dairy Sci. 93, 20142024. doi:10.3168/jds.2009-2897

Piepers, S., Peeters, K., Opsomer, G., Barkema, H.W., Frankena, K., De Vliegher, S., 2011. Pathogen group specific risk factors at herd, heifer and quarter levels for intramammary infections in early lactating dairy heifers. Prev. Vet. Med. 99, 91 -101. doi:10.1016/j.prevetmed.2011.02.010

Piessens, V., Van Coillie, E., Verbist, B., Supré, K., Braem, G., Van Nuffel, A., De Vuyst, L., Heyndrickx, M., De Vliegher, S., 2011. Distribution of coagulase-negative Staphylococcus species from milk and environment of dairy cows differs between herds. J. Dairy Sci. 94, 2933-2944. doi:10.3168/jds.2010-3956

Ruhland, B.K., Granzer, W., Groth+, W., Pirchner, F., 1999. Blood levels of hormones and

metabolites, erythrocytes and leukocytes and respiration and pulse rate of heifers after alpage. J. Anim. Breed. Genet. 116, 415-423. doi:10.1046/j.1439-0388.1999.00203.x

Santman-Berends, I.M., Olde Riekerink, R.G., Sampimon, O.C., van Schaik, G., Lam, T.J., 2012. Incidence of subclinical mastitis in Dutch dairy heifers in the first 100 days in lactation and associated risk factors. J. Dairy Sci. 95, 2476-2484. doi:10.3168/jds.2011-4766

Schaeffer, D., von Borell, E., Richter, T., 2007. Critical control points (CCP) for the housing and management of calves. Züchtungskunde 363-393.

Sears, P.M., Smith, B.S., English, P.B., Herer, P.S., Gonzalez, R.N., 1990. Shedding Pattern of Staphylococcus aureus from Bovine Intramammary Infections. J. Dairy Sci. 73, 2785-2789. doi:10.3168/jds.S0022-0302(90)78964-3

Snijders T.A.B., Bosker R.J., 2012. Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling. 2nd ed., Sage Pub. Ltd., London.

StataCorp. 2013. Stata 13 Base Reference Manual. College Station, Stata Press.

StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP.

Sundrum, A., 2007. Critical control points (CCP) for housing and management of beef cattle. Züchtungskunde 394-414.

Svensson, C., Nyman, A.-K., Persson Waller, K., Emanuelson, U., 2006. Effects of housing,

management, and health of dairy heifers on first-lactation udder health in southwest Sweden. J. Dairy Sci. 89, 1990-9.

Veh, K. a. a, Klein, R.C.C., Ster, C., Keefe, G., Lacasse, P., Scholl, D., Roy, J.-P.J.-P., Haine, D., Dufour, S., Talbot, B.G.G., Ribon, a. O.B.O.B., Malouin, F., 2015. Genotypic and phenotypic characterization of Staphylococcus aureus causing persistent and nonpersistent subclinical bovine intramammary infections during lactation or the dry period. J. Dairy Sci. 98, 155-168. doi:10.3168/jds.2014-8044

Voelk, V., Graber, H.U., van den Borne, B.H.P., Sartori, C., Steiner, A., Bodmer, M., Haerdi-Landerer, M.C., 2014. A longitudinal study investigating the prevalence of Staphylococcus aureus genotype B in seasonally communal dairy herds. J. Dairy Sci. 97, 4184-92. doi:10.3168/jds.2013-7291

Von Borell, E., Herrmann, H.J., Knierim, U., Muller, C., Richter, T., Sanftleben, P., Schaffer, D., Schulze, V., Sundrum, A., 2007. Critical control points (CCP) for cattle housing and management - a concept for the on-farm assessment of animal welfare, health and management. Züchtungskunde 329-338.

von Keyserlingk, M.A.G., Rushen, J., de Passillé, A.M., Weary, D.M., 2009. Invited review: The welfare of dairy cattle—Key concepts and the role of science. J. Dairy Sci. 92, 4101-4111. doi:10.3168/jds.2009-2326

Waage, S., 0degaard, S.A., Lund, A., Brattgjerd, S., R0the, T., 2001. Case-Control Study of Risk Factors for Clinical Mastitis in Postpartum Dairy Heifers. J. Dairy Sci. 84, 392-399. doi:10.3168/jds.S0022-0302(01)74489-X

Welfare Quality®, 2009. Welfare Quality® Assessment Protocol for Cattle. Welfare Quality®140 Consortium. Lelystad, Netherlands, 180 pp. Publisher: Welfare Quality&REG; Consortium, Lelystad, Pays-Bas, The Netherlands; ISBN number: ISBN/EAN 978-90-78240-04-4.

Web references

International Committee for Animal Recording (ICAR), 2014,

http://www.icar.org/Documents/Rules%20and%20regulations/Guidelines/Guidelines_2014.pdf (accessed November 2015).

National Mastitis Council, 2004, http://www.nmconline.org/sampling.htm (accessed

November 2015). Source: Microbiological Procedures for the Diagnosis of Bovine Udder Infection and Determination of Milk Quality. (NMC publication, 2004)

Swiss Federal Food Safety and Veterinary Office (Switzerland),

http://www.blv.admin.ch/themen/tierschutz/00744/00750/index.html?lang=de (accessed, November 2015, used in supplementary material).

University of Bristol, Center for Multilevel Modelling,

http://www.bristol.ac.uk/cmm/learning/online-course/index.html (accessed November 2015).

Tables

Table 1A. Description of herd level risk factors potentially related to intramammary infections in

Swiss dairy heifers

Independent variable Categories Definition of categories

Variables at herd level

(demographic data)

Herd size 12-24 dairy cows Tercile 1

24-33 dairy cows Tercile 2

34-115 dairy cows Tercile 3

Geographical region of the dairy farm Lowland zone Territorial division of agricultural area with different climate,

(Cadastral zones1) Mountain zone I infrastructure and surface structure

Mountain zone II

Mountain zones III and IV

Average milk production in year 2012 5500-7000 kg Tercile 1

7000-7800 kg Tercile 2

7800-10,000 kg Tercile 3

Yield corrected herd somatic cell count < 100,000 (cells/mL) Average in the year 2012

CHSCC1 > 100,000 (cells/mL)

Yield corrected herd somatic cell count <200,000 (cells/mL) Average in the year 2012

CHSCC2 >200,000 (cells/mL)

Housing system (Dairy cows) Loose housing

Tie-stall barn

Housing young stock

Housing of calves Crate

Group pen

Tie-stall barn

Housing of young cattle Deep straw grouped Deep straw bedded group pens without cubicles

Free-stall with cubicles Free-stall with cubicles

Tie-stall barn Tie-stall barn

Alpine rearing Yes/No Communal alpine pasturing during summer

External rearing Yes/No Raising in specialized farms with animals of other farms

Feeding of rearing cattle

Period of milk feedin <4 months Tercile 1

4 months Tercile 2

>4 months Tercile 3

Amount of whole milk fed L//day Range: 5-8 L/day

Quality of whole milk fed Milk with antibiotic residues

High SCC milk

Bulk tank milk

Feeding of minerals to calves Yes/No

Calf age at the start of additional feeding Directly after birth Tercile 1

After 1 week Tercile 2

After 2 weeks Tercile 3

Feeding concentrates for calves Yes/No

Type of roughage for cattle Only hay

Second cut hay

Corn Silage

Feeding concentrates to heifers Grass silage

Feeding of minerals to heifers Yes/No

Grazing regimen Yes/No

< 6 months Tercile 1

6-7 months Tercile 2

> 7 months Tercile 3

Heifer management

Preconditions for the first insemination Age

Weight

Development

Season

Desired calving age of heifers 24-26 months

27-29 months

>30 months

Adaption time in the productive herd < 2 weeks Tercile 1

2 - 3 weeks Tercile 2

> 3 weeks Tercile 3

Heifers housed with dry cows Yes/No

Table 1B. Description of cow- and quarter-level risk factors potentially related to intramammary infections in Swiss dairy heifers

Independent variable Categories Definition of categories

Variables at quarter level

Teat lesions Yes/ no Presence of lacerations and bruises

Teat swelling Yes/ no Swollen teats

Papilloma Yes/ no Presence of papilloma on the teat skin

Dysfunction or abnormality of the Yes/ no Atrophy of teat canal

tTeeaat t length Short teats < 5 cm

Normal teats 5-7 cm

Long teats > 7cm

Variables at heifer level

Breed Holstein Holstein cows with and without

Brown cattle Swiss brown cattle, Brown Swiss,

Red pied Simmental, Swiss Red Pied,

Age at calving Early calving age < 24 months

Middle calving age 25-30 months

Late calving age >30 months

Season of calving Winter January-March

Spring April-June

Summer July-September

Autumn October-December

Progress of parturition Normal No assistance

Dystocia Assistance of >1 person

Stillbirth If calf was born dead or died within

Assisted calving Yes/No

General condition of the heifer Good animal is attentive, standing, eats

Slightly disturbed annoirmall is attentive, standing, does

Seriously disturbed animal does not get up and is

Udder condition (Scores) Soft

Swollen, firm

Red, swollen, firm, warm

Udder edema Vps/Nn Yes/No Retained fluids in the intracellular

spaces of mammary tissue

Milk flow (Score) assessed by High milk flow Easy milker

hand milking Normal milk flow

Low milk flow Hard milker

Table 2: Definition of infection status of quarters for any of the pathogen groups considered.

Culture result of Culture result of Infection status

Duplicate sample 1 Duplicate sample 2 of quarter

positive positive infected

positive negative infected

negative positive infected

negative negative healthy

positive contaminated1 infected

contaminated positive infected

negative contaminated healthy

contaminated negative healthy

contaminated contaminated exclusion

more than 2 different pathogens present

Table 3. Prevalence (%) of mastitis-relevant pathogens in 391 heifers and their 1564 quarters presented at the cow- and quarter-level.

Non- infected Contagious Environmentala CNSb Others' In total

S. aureus

Number of quarters 767 75 95 664 10 1'564

Prevalence (%) 49.0 4.8 6.1 42.5 0.6 d

Number of cows 96 41 75 270 7 391

Prevalence (%) 24.6 10.5 19.2 69.1 1.8 d

a Environmental: Coliforms, S. dysgalactiae, S. uberis. b CNS: Coagulase- negative staphylococci.

c The category includes Trueperellapyogenens (T. pyogenes), aerobic spore formers, Corynebacterium bovis (C. bovis) and other Gram-positive rods.

d The sum of all percentages is bigger than 100% because cows can have two pathogens at one quarter.

Table 4. Number of farms and prevalence (%) of mastitis-relevant pathogens in heifers, at the herd level in 54 Swiss dairy farms.

The influence of the rearing period on intramammary infections in Swiss dairy heifers: A cross-sectional study Academic research paper on "Animal and dairy science"

Abstract of research paper on Animal and dairy science, author of scientific article — M.J. Bludau, A. Maeschli, F. Leiber, P. Klocke, J.A. Berezowski, et al.

Similar topics of scientific paper in Animal and dairy science , author of scholarly article — M.J. Bludau, A. Maeschli, F. Leiber, P. Klocke, J.A. Berezowski, et al.

Academic research paper on topic "The influence of the rearing period on intramammary infections in Swiss dairy heifers: A cross-sectional study"