| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, P.O. Box 7023, SE-750 07 Uppsala, Sweden
Department of Clinical Sciences, Swedish University of Agricultural Sciences, P.O. Box 7054, SE-750 07 Uppsala, Sweden
1 Corresponding author: christel.hagnestam{at}hgen.slu.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: dairy cow • clinical mastitis • yield loss
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
There are numerous studies on the effects of CM on milk production. Yield losses have been studied on a short-term basis (in the weeks before and after clinical onset) as well as with respect to lactational level (mostly 305-d lactations). It is important to analyze short-term losses so that the effects of CM can be assessed accurately. A temporary drop in production associated with a higher yield level before or after the disease might mask the true effects of disease on the 305-d yield (Østergaard and Gröhn, 1999). Furthermore, to make correct management decisions it is essential to know the patterns of short-term loss associated with CM. When deciding whether to cull a cow experiencing reduced production due to CM, it is necessary to know whether she is most likely to regain her premastitic production level, whether she will remain at her current yield, or whether her milk production will become even more reduced (Bartlett et al., 1991).
In a review by Hortet and Seegers (1998), the estimated lactational milk loss was found to vary between 0 and 9.5% across parities. Management, breed, and yield level, as well as the analytical method used, differed between studies and can be assumed to have influenced the results. Analyses of production losses due to CM are thus situation-specific and should be conducted on the breed and in the environment of interest to ensure that relevant estimates are obtained.
It is preferable to analyze repeated measures of milk yield within lactation instead of a summary measure of 305-d milk yield (Gröhn et al., 1999). Monthly data are not relevant for the assessment of short-term losses or lactational losses when CM occurs in early lactation (Hortet and Seegers, 1998). Therefore, it is recommended that weekly production records should be used for studies of the effects of CM on milk yield.
The extent of the short-term losses, as well as the magnitude of the 305-d losses, has been shown to vary according to when in lactation the cow is diseased (Lucey and Rowlands, 1984; Lucey et al., 1986; Rajala-Schultz et al., 1999). To achieve accurate estimates the timing of the disease event must therefore be taken into account. Furthermore, daily yields need to be related to the day of testing with respect to calving as well as to diagnosis.
So far, studies have investigated the effects of CM on milk yield when the disease has occurred in rather crude periods of lactation. No quantification of the losses associated with CM in the different weeks of lactation has been conducted. The aim of this study was to estimate the short-term and long-term production losses associated with CM, taking into consideration the interaction between DIM and day relative to diagnosis for test days.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Animals and Management
Production records were collected between September 1987 and April 2004 at the research farm of the Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences. Detailed information on disease occurrences, calvings, reproductive status, and culling was available. Test-day records from 307 Swedish Red (SR) and 199 Swedish Holstein (SH) cows of parity 1 to 10 were used in the study. In 2004, the average production of the SR cows in the herd was 8,921 kg of milk, 387 kg of fat, and 315 kg of protein. Corresponding yields of SH cows were 10,626, 392, and 358 kg, respectively. These production levels were slightly above the national averages for the breeds. The SR cows were in a selection experiment. They were bred for either high (n = 148) or low (n = 159) fat content in their milk, while maintaining the same total energy production.
Before 1992, the experimental herd was kept on a farm with tie-stall housing (Kungsängen). The current research farm (Jälla) has a free-stall barn (n = 50) and a tie-stall barn (n = 50). On both farms cows were kept in maternity pens at the time of calving.
The cows were fed according to Swedish standards (Spörndly, 1999). In the first 16 wk of lactation, cows were fed ad libitum. Thereafter, they were fed a restricted ration based on production level and weight. In the tie-stall barn, cows were fed the total ration individually, whereas in the free-stall barn, only concentrates were fed individually. During the pasture period (May to September), additional concentrates and roughage were supplied.
All cows were milked twice daily and were test milked once a week. The milk samples were analyzed for fat and protein content according to standard procedures.
Disease Data
Cases of CM were detected, by the milkers, by presence of abnormal milk in the first milk streams or by signs of inflammation in one or more quarters. All cases were diagnosed by a veterinarian; however, not all cases were treated, depending on the stage in lactation at which the cow developed the disease as well as possible designation for culling. The number of cases and the lactational incidence risk (lactations with at least one case of CM divided by the total number of lactations at risk) of CM in primiparous and multiparous cows of the different breed classes are shown in Table 1
.
|
Milk samples for bacteriological culture were also taken from cows with SCC >180,000 cells/mL on 2 subsequent test days. Furthermore, milk samples were taken from each quarter by routine in wk 4 of lactation, as well as before the cows were dried. From the fall of 1997, milk samples were also taken in the first week of lactation. From 1993 to 2004, milk samples were taken from 1,560 quarters, and 40% of the samples had positive cultures. Table 2 shows the pathogens present in the culture-positive milk samples. Pathogen-specific analyses could not be performed because of the limited number of clinical cases.
|
). Separate index categories were created for the 4 wk preceding diagnosis and the 10 wk following diagnosis. All test days occurring earlier than 4 wk before diagnosis were summarized into one category and all test days occurring later than 10 wk after diagnosis were grouped into another category. Nonmastitic cows constituted a separate index category and were used as a reference for the yield level of healthy cows.
|
Dystocia, stillbirth, and retained placenta were diagnosed by the farm personnel. Claw disorders were diagnosed by either a veterinarian or a professional hoof trimmer. Other disorders were diagnosed by a veterinarian.
Indices were created for the 4 complexes of disorders. Due to limitations of the data, these indices were not as detailed as the mastitis index. For reproductive disorders, metabolic disorders, and claw disorders the indices had 5 categories: the first 4 were test-day yields measured more than 1 mo before diagnosis, less than 1 mo before diagnosis, less than 1 mo after diagnosis, and in the period thereafter. Cows that had not contracted any of the disorders included in a complex were given a fifth category. Because calving disorders obviously only appeared early in lactation, that index had 3 categories; within 1 mo after diagnosis, more than 1 mo after diagnosis and not affected.
Statistical Methods
We analyzed the first lactational incidence of CM, regardless of etiology (n = 298). Cases that occurred before first calving or in a dry period were thus excluded from the study.
The effects of CM on the daily milk, fat, and protein production were estimated using a mixed linear model in the MIXED procedure of SAS (SAS Institute, 2001). The data set contained repeated measurements of test-day yields within lactations. The correlations between test days were accounted for by specifying a correlation structure among the residuals. The first-order autoregressive covariance structure was used. It has previously been shown to be appropriate for this type of data (Rajala and Gröhn, 1998; Gröhn et al., 1999; Santos et al., 2004; Wilson et al., 2004). The autoregressive covariance structure assumes a constant variance of measurements and declining correlations between measures as the time interval between them increases. Specifying lactation nested within cow and breed as a subject effect in the repeated-statement in the model for multiparous cows indicated that test-days were correlated within the lactations of cows.
The maximum lactation length was set to 305 d. The proportion of lactations with short duration (last test-day occurring earlier than 300 DIM) was 46.6 and 62.6% in primiparous and multiparous cows, respectively. The median DIM when the last test-day occurred was 300 in primiparous cows and 295 in multiparous cows. Due to different shapes of their lactation curves, primiparous and multiparous cows were analyzed separately (494 and 698 lactations, 17,039 and 21,496 test days, respectively). For multiparous cows the following model was used:
 |
where Y was the milk, fat, or protein production at each test-day. The lactations were divided into 19 stages. Thus, milk records collected within the first 8 wk of lactation were grouped into weekly intervals; records from wk 9 to wk 16 were grouped into 2-wk periods, and records taken later than 17 wk after calving were grouped into 4-wk periods. The mastitis index was defined as above. Breed had 3 levels: SR low-fat line, SR high-fat line, and SH. Parity was defined as 2 or 
3. Reproductive status indicated whether the cow was pregnant on the test day. We distinguished between 3 calving seasons: January to April, May to August, and September to December. Year and season of calving were pooled and analyzed as a combined variable comprising 50 classes. The indices for other disorders were as defined above. Interactions of biological interest were tested in the preliminary analyses, but were excluded from the final models as they were found to be nonsignificant.
In the model for primiparous cows, the index for metabolic disorders was excluded due to there being very few cases, and parity was excluded for obvious reasons.
The residuals were tested for normality using a Q-Q plot, and were judged as being close to the normal distribution.
Estimation of 305-d Yields
To compare the long-term yield losses associated with CM occurring in different stages of lactation, estimates of the 305-d yields for cows diagnosed with CM in wk 1 to 44 of lactation were calculated and compared with the yields of nonmastitic cows.
Least squares means (LSM) of the interaction between lactational stage and mastitis index were used as estimates of the daily production of milk, fat, and protein in a certain week of lactation at a certain time with respect to diagnosis. The LSM comprising the complete lactation of each group of cows (healthy and diseased in a specified week) were then multiplied by 7 to give the weekly yields, and thereafter summed, resulting in estimates for the 305-d yields.
Not all combinations of lactational stage and mastitis index were available from the analyses, because in some weeks of lactation no case of CM occurred. When an LSM was missing for a combination in the middle of a lactation, the estimated daily yield during that week was calculated as the mean value of the LSM from the week before and the week after the missing value. If the missing LSM was from the beginning or the end of lactation, and hence no estimate from the week before or after was available, the mean value of the LSM from the corresponding week of lactation in cows that had developed CM 1 wk earlier and 1 wk later was used.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
. Clinical mastitis was more frequent in multiparous than in primiparous cows. First-parity cows developed CM earlier in lactation than older cows: the median time of diagnosis was d 11 (third quartile = 100) and d 55 (third quartile = 127), respectively. The production of primiparous cows peaked at wk 8, whereas multiparous cows reached peak production at wk 7. Thus, primiparous cows usually developed CM before peak yield and multiparous cows more commonly developed CM after peak yield. As this study focused on the effects of CM on production, no further attention will be paid to the impact of the complexes of disorders.
|
). To illustrate the weeks of diagnosis in which most data were available, lactation curves for cows diagnosed in different weeks were plotted for primiparous and multiparous cows diseased in late lactation. The stair-like shape of the lactation curves is due to the grouping of weeks into lactation stages after wk 8.
|
Short-Term Losses
The pattern and extent of the short-term losses associated with CM depended on the lactational stage of the cow when diagnosed with the disease. Figure 2 shows the patterns of short-term losses when CM occurred in selected weeks of lactation. Milk yield tended to decline 2 to 4 wk before diagnosis. The drop in milk yield was, however, either not significant or significant only in the week preceding diagnosis. A drop in production seemed to occur 4 wk (or more) before diagnosis when cows developed CM around peak yield. When cows were diseased in mid lactation the drop seemed to occur 3 wk before diagnosis and when cows developed CM in late lactation it seemed to start 2 wk before diagnosis.
|
Primiparous and multiparous cows diseased in the first week of lactation did not reach the yield level of nonmastitic cows in the 10-wk period following diagnosis. Cows that were diagnosed with CM in wk 8 of lactation produced significantly less than nonmastitic cows for some weeks after clinical onset. When diseased later in lactation, cows reached the production level of nonmastitic cows approximately 5 wk after diagnosis, but still produced less than they did 3 wk before diagnosis. The differences between daily milk yields of mastitic cows diseased in late lactation and daily milk yields of nonmastitic cows were, however, not significant in the 10-wk period following diagnosis, except for multiparous cows showing a significantly increased milk yield 6 wk after clinical onset.
Losses over 305 d
The extrapolated 305-d yields (based on LSM) of non-mastitic primiparous cows (396 lactations) were 7,675 kg of milk, 330 kg of fat, and 259 kg of protein. The corresponding figures in multiparous cows (498 lactations) were 7,862 kg of milk, 337 kg of fat, and 253 kg of protein. The 305-d yields of mastitic cows diagnosed in different weeks of lactation, expressed as a percentage of the yields of nonmastitic cows, are shown in Figure 3.
|
The 305-d milk, fat, and protein yields of mastitic multiparous cows were more uniform than those of primiparous cows. Multiparous cows diagnosed with CM in the first half of lactation tended to perform below nonmastitic multiparous cows, whereas cows diagnosed with CM in the latter part of lactation had higher 305-d yields. The most severe milk loss was found when multiparous cows were diseased in wk 3. The magnitude of the yield loss varied from 0 to 902 kg, or 0 to 11% of the 305-d milk yield. The reduction in 305-d fat yield was 0 to 41 kg (0 to 12% of the total yield). Multiparous cows diseased in wk 5 experienced the most extensive loss of fat production. The protein yield was decreased by 0 to 27 kg (0 to 11% of the total yield) in mastitic multiparous cows, with the largest reduction occurring when diseased in wk 3.
Multiparous cows diseased before wk 22 appeared to have lower 305-d milk yields than nonmastitic multiparous cows. In primiparous cows, mastitic cows diseased in wk 1 to 38 tended to produce less than nonmastitic cows on a 305-d basis. The actual milk loss as well as the proportional milk loss during the weeks when both groups produced less than nonmastitic cows were smaller in primiparous cows.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
In multiparous cows, the size of the fluctuations in daily milk yield associated with clinical diagnosis tended to depend on the stage of lactation in which the cow was diseased, in accordance with the results of Lucey et al. (1986) and Rajala-Schultz et al. (1999). Hortet and Seegers (1998) concluded that short-term losses were more severe in cows diseased in early lactation than in cows diseased in mid or late lactation. We observed the same trend, which was particularly pronounced in multiparous cows. At the time of clinical onset, we found that mastitic multiparous cows produced 1 to 8 kg less milk per day than nonmastitic multiparous cows. Thus, the fluctuations were larger in our study than they were in previous studies, which demonstrated daily losses of less than 1 to 2.5 kg (Lucey et al., 1986; Rajala-Schultz et al., 1999).
Dohoo and Martin (1984) and Lucey et al. (1986) found that treatment caused a boost in production; subsequently, in mastitic cows, milk yield equaled or exceeded that of nonmastitic cows. This phenomenon was generally not observed in our study, with the exception of the finding that multiparous cows diseased in late lactation had an increased milk yield 6 wk following clinical onset. This result was likely due to uncertainty of the estimate. Cows diseased before or at peak yield did not reach the yield level of nonmastitic cows in the 10-wk period after diagnosis, indicating that CM caused production losses for an extensive period following clinical onset. This was in accordance with Bartlett et al. (1991), who showed that cows experienced an immediate drop in production when diagnosed with CM and had a suppressed milk yield throughout the 60 d following clinical onset.
Substantial Reduction in 305-d Milk Yield
The effects on production were most severe in cows developing CM in early lactation, confirming results from previous studies (Bartlett et al., 1991; Hortet and Seegers, 1998; Rajala-Schultz et al., 1999; Santos et al., 2004). Interestingly, Wilson et al. (2004) found that the production of multiparous cows was more affected when they developed CM before peak yield, whereas primiparous cows were most affected when diseased between peak yield and 120 DIM. In our study, primiparous as well as multiparous cows were most affected when CM occurred before peak yield. This was probably due in part to CM causing more severe losses in early lactation. It was also, we believe, the result of CM being experienced over a longer period when developed in early lactation, as Lescourret and Coulon (1994) suggest.
After being diagnosed with CM, cows produced less milk than their nonmastitic herdmates throughout the rest of the lactation; this was in accordance with Wilson et al. (2004). Bareille et al. (2003) found that mastitic cows suffered from yield losses for a period of 140 d following diagnosis. In Gröhn et al. (2004), mastitic cows did not return to their premastitic milk yield.
The detrimental effect of CM on lactational milk yield has been shown by several studies (Hortet and Seegers, 1998; Rajala-Schultz et al., 1999; Wilson et al., 2004). Comparisons of the actual milk loss between studies are difficult to make. This is partly because there is considerable variation in the data collected (e.g., with regard to period of collection, breed, yield level of cows, country, and whether the data are from a field study or relate to a research herd), but it is also the outcome of differences in the analytical methods used. The purpose of comparing our estimates with those attained in previous studies was mainly to validate the magnitude of the losses. The estimated 305-d milk losses in primiparous and multiparous cows of 0 to 9% and 0 to 11%, respectively, correspond relatively well with previous studies (Hortet and Seegers, 1998; Rajala-Schultz et al., 1999; Wilson et al., 2004). The outlying values were slightly more extreme than those found by others, although these losses occurred only for cows diseased in specific weeks of lactation. Previous studies either did not account for when, during lactation, cows were diseased or used crudely defined periods of lactation. By not distinguishing between cases occurring in the different weeks of lactation, these studies probably ensured that the most severe yield losses were diluted by milder losses in other weeks. The size of the extensive yield losses found in our study can therefore be regarded as reasonable.
Milk, Fat, and Protein Production Are Equally Affected
The losses of fat and protein were of the same relative magnitude as the milk loss, which is supported by Hortet and Seegers (1998) and Santos et al. (2004). Thus, the effect of CM on fat and protein yield seemed to be caused by reduced milk production and not by changes in the fat and protein content of the milk.
High Milk Production: A Risk Factor for CM
In our study, as well as in previous work (Lucey and Rowlands, 1984; Østergaard and Gröhn, 1999; Gröhn et al., 2004), the incidence of CM was higher in multiparous cows than in the lower producing primiparous cows. Moreover, milk loss due to CM has been found to be greater in multiparous cows (Bartlett et al., 1991; Hortet and Seegers, 1998; Østergaard and Gröhn, 1999). Within parity groups, mastitic cows have been shown to have a higher production level than nonmastitic cows (Rajala-Schultz et al., 1999; Gröhn et al., 2004; Wilson et al., 2004). We observed the same trend, and, like Wilson et al. (2004), found it especially pronounced in multiparous cows. Furthermore, unfavorable genetic correlations have been found between milk yield and CM (Bunch et al., 1984; Emanuelson et al., 1988). Thus, it can be concluded that high milk yield is a risk factor for CM, supporting the results of Bigras-Poulin et al. (1990), Rajala and Gröhn (1998), Østergaard and Gröhn, 1999, and Wilson et al. (2004).
The seemingly higher 305-d yields in mastitic cows affected in mid or late lactation were most certainly a consequence of mastitic cows producing above the level of nonmastitic cows to such an extent that the cumulative 305-d yields were higher, even though the cows had suffered from CM. These cows would probably have been even more superior to their nonmastitic herd-mates had they not developed CM, and it is most likely that the estimated losses would have been larger still if we had corrected for the production potential of the cows in the models.
Sources of Possible Bias
Studying a research herd inevitably imposes a certain bias on the data. Closer surveillance of the animals leads to more cases than normal receiving treatment. This is also shown by the high lactational incidence of CM in our study: 19.8 and 28.7% in primiparous and multiparous cows, respectively, compared with incidences of 9.1 and 13.2 to 24.3% in cows enrolled in the milk-recording scheme in 2004–2005 (Swedish Dairy Association, 2005). Some of the reported cases were probably milder than those reported by practicing veterinarians, which might have influenced the size of the yield estimates.
There are other possible sources of bias in our data. By only including the first lactational incidence of CM, all losses were attributed to just one case. Thus, in lactations with recurrent cases, the loss assigned to the first case will have been overestimated. In addition, nonmastitic cows are usually poorer producers. Therefore, using the yield of nonmastitic cows as a reference for the production level in healthy cows certainly results in an underestimation of the actual losses associated with CM. Finally, our estimated losses are valid on a 305-d basis, assuming that mastitic cows have complete lactations. A consequence of CM can be forced drying off and therefore shortened lactation. When this is the case, the yield losses will probably be even more severe than those reported here.
Implications
The results of this study are of practical interest to dairy farmers, showing that preventative measures should be focused on reducing the frequency of CM between calving and peak yield, because this is the period when the majority of cases occur and when yield losses are the most extensive.
The possibilities of utilizing the fact that daily milk yield starts to drop roughly 3 wk before diagnosis should be investigated. It is likely that the detrimental effects of CM can be reduced if the disease is diagnosed and treated when the drop in production first occurs instead of when the clinical signs are manifested.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication September 11, 2006. Accepted for publication December 21, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
