- Email: [email protected]
Milk production and reproduction during a subclinical bovine herpesvirus 1 infection on a dairy farm
Preventive Veterinary Medicine 34 Ž1998. 97–106
Milk production and reproduction during a subclinical bovine herpesvirus 1 infection on a dairy farm J.J. Hage
a,)
, Y.H. Schukken b, Th. Dijkstra a , H.W. Barkema a , P.H.R. van Valkengoed b, G.H. Wentink c
a
b
Animal Health SerÕices, P.O. Box 361, 9200 AJ Drachten, Netherlands Department of Herd Health and Reproduction, UniÕersity of Utrecht, P.O. Box 80.151, 3508 TD Utrecht, Netherlands c Holland Genetics, P.O. Box 5073, 6802 EB Arnhem, Netherlands Accepted 29 October 1997
Abstract This study describes an outbreak of bovine herpesvirus 1 ŽBHV1. infections in a dairy herd with special reference to disease symptoms, reproductive performance and milk production losses. The study was carried out with a dairy herd consisting of 98 lactating animals. All animals were housed in the same freestall barn with intensive contact between all animals. An outbreak of BHV1 was induced by injecting three seropositive cows with dexamethasone. During the outbreak, no clinical signs were observed in any of the newly infected animals. At the time of infection, a significant drop in milk production was noted in animals that were initially-seronegative. The production loss was estimated at approximately 9.5 l per infected animal during the infectious period of 14 days. None of the pregnant cows aborted because of BHV1 infection. During 50 days before BHV1 circulation, there was a significant decrease in the number of successful inseminations in both seronegative and seropositive animals. Therefore, it is doubtful that early pregnancies were terminated by BHV1 infection. The proportion of successful inseminations during the BHV1 circulation in this herd, and in the period thereafter, did not significantly differ from the baseline period. q 1998 Elsevier Science B.V. Keywords: Bovine herpesvirus; Cattle, microbiological diseases; Milk production; Reproductive performance
)
Corresponding author. Tel.: q31 512 570700; fax: q31 512 520013.
0167-5877r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 7 - 5 8 7 7 Ž 9 7 . 0 0 0 8 8 - 3
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
98
1. Introduction Bovine herpes virus 1 ŽBHV1. is the etiological agent of infectious bovine rhinotracheitis ŽIBR., infectious pustular vulvovaginitis ŽIPV., and infectious pustular balanoposthitis ŽIPB., abortion, and fatal multi-systemic infection in newborns ŽGibbs and Rweyemamu, 1977; Higgings and Edwards, 1986.. BHV1 is an important pathogen of cattle and causes substantial economic losses. Because of its impact on animal health and farm economics, European countries are making a concerted effort to control or eradicate BHV1 infection. 1 BHV1 infection can also occur as a subclinical infection ŽVan Oirschot et al., 1993; Hage et al., 1997.. However, little is known about the economic impact of a subclinical BHV1 infection on milk production and reproductive performance in a dairy herd. Cortese Ž1991. did not observe a decrease in milk production after vaccinating cows with a genetically altered live IBR vaccine. BHV1 is reported to give rise to high milk cell counts in the absence of mastitis pathogens ŽSiegler et al., 1984.. BHV1 can cause mastitis if inoculated into the udder ŽGreig and Bannister, 1965; Corner et al., 1967., and has been isolated from cattle with mastitis ŽGourlay et al., 1974; Roberts et al., 1974.. Miller and van der Maaten Ž1987. isolated BHV1 from reproductive tissues Žovary, infundibulum, and uterine tube. from intravenously infected cattle, and BHV1 has been detected in follicular fluid and oocytes ŽGibbs and Rweyemamu, 1977; Bielanski et al., 1993.. A decrease in the conception rate of initially-seronegative animals after experimental BHV1 infection has been described ŽMiller et al., 1988; Chiang et al., 1990.. Yet in the presence of BHV1, oocytes can be fertilized in vitro and co-cultured to the blastocyst stage; there was almost no difference with the control group ŽBielanski and Dubuc, 1993.. Experimental BHV1 infection has been described to cause embryonic death in vitro ŽBowen et al., 1985. and in early pregnancy in seronegative cattle ŽMiller and van der Maaten, 1987.. However, Miller et al. Ž1988. showed that initially, seronegative heifers carried their fetuses to term when inoculated with BHV1 14 days after mating. If cattle abort as a result of BHV1 infection, abortion usually occurs between the 4th and 7th month of gestation at a variable period after infection ŽGibbs and Rweyemamu, 1977.. Miller et al. Ž1991. reported that strains of BHV1 had different abortifacient properties. Thus, heifers inoculated with BHV1 subtypes 1 and 2.a aborted between 17 and 85 days after inoculation, whereas heifers inoculated with BHV1 subtype 2.b delivered full-term calves. There is also an inter-experiment difference in the abortifacient property of one particular strain of BHV1 ŽMiller et al., 1988, 1991.. Most of the above-mentioned data were obtained from experimental infections. However, it is not clear whether natural infections will also lead to similar production or fertility losses. In this study, we aim to describe milk production and fertility losses associated with an Žinduced. outbreak of BHV1 infections on a commercial dairy farm.
1
EC directives 88r407 and 93r60.
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
99
2. Materials and methods 2.1. Farm The study was carried out with a Holstein–Friesian dairy herd. The study period started at November 9, 1993 and ended April 29, 1994. The fat Ž4.0%. and protein Ž3.3%. corrected milk production was 8651 kg for 305 days. This study concerns a total of 110 Žincomplete. lactations of 98 animals. Fifty-five animals were seropositive and 43 animals were seronegative for BHV1; 40 of the latter cows were first-lactation animals, whereas seropositive cows were all second lactation or higher. During the study, animals were housed in the same freestall barn with intensive contact between all animals. All animals were identified by freeze-branding. None of the animals had ever been vaccinated against BHV1. No animals were introduced into the herd during the study. The population dynamics of this BHV1 circulation has been described before ŽHage et al., 1997.. Briefly, three seropositive cows were given dexamethasone injections Ž0.1 mgrkg. and subsequently all seronegative animals seroconverted within 5 weeks. All animals were observed daily for occurrence of clinical disease and reproductive problems, such as vaginal discharge or abortions. 2.2. Serology During the experiment, blood samples were collected from all animals at weekly intervals and were screened for antibodies against BHV1, using an undiluted gB-blocking ELISA Žspecificity and sensitivity 0.96 and 0.99, respectively. ŽKramps et al., 1994.. The positive serum samples were titrated by serial dilution. Seroconversion was considered to have occurred when a blood sample was negative for BHV1 in one sample but positive in the next sample. Animals were assumed to seroconvert 8 days after infection ŽKramps et al., 1994.. Re-infection was defined as a significant rise in antibody titre Ža 4-fold or higher rise in antibody titre ŽG 2 dilution steps.. in a 14-day period. Because of the 1-week intervals of bloodsampling, seronegative animals were considered to have been infected 10 days Ž1.5 week. before they seroconverted or showed a significant increase in antibody titre. In initially-seropositive animals, the moment of re-infection was considered to be 10 days before there was a significant rise in antibody titre ŽHage et al., 1997.. To estimate the effect of a BHV1 infection on milk production, the day of BHV1 infection of each initially-seronegative lactating animal was considered to be Day 0 ŽD0.. 2.3. Milk production and fertility The lactating animals were milked twice a day. The milk yield was recorded automatically, and morning and evening yields were summed to obtain daily milk production. Data were checked for missing values. When single measurements were missing, these were replaced by the mean of the previous and next production for the same part of the day Žmorning or evening. ŽDeluyker et al., 1990.. At weekly intervals, the percentages fat and protein and somatic cell counts ŽSCC. were measured via the
100
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
Fig. 1. The mean daily milk production of all 98 lactating Holstein–Friesian cattle Ž1993–1994, Netherlands..
Fig. 2. Daily milk production relative to estimated day of infection of 43 initially seronegative Holstein–Friesian cattle Ž1993–1994, Netherlands..
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
101
Dutch milk-recording system. Records were kept of all insemination and calving data over the year before and the year following the start of this experiment. When an insemination led to calving 9 months later or an observed abortion, it was defined as successful. An insemination that was followed by another insemination, or by a negative pregnancy check, was defined as unsuccessful. 2.4. Statistical methods Milk-production data were analysed by using a regression approach. A linearized model to describe the lactation curve was estimated according to the method described by Ali and Schaeffer Ž1987.. This basic daily milk production model can be expressed by: 2
2
y s b 1 DIMr305q b 2 Ž DIMr305. q b 3 log Ž 305rDIM. q b4 Ž log Ž 305rDIM. . where: y s observed daily milk production, DIM s days in milk, and log is natural logarithm. This model was extended by an extra term describing the length of the infectious period. This variable was coded as 1 during the first days after infection ŽD0–D14. and 0 for the remaining days. The regression coefficient of this parameter gives an estimate of the mean production loss per animal per day during the infectious period. Parity 2 animals were not included since only three animals with parity ) 1 were available for analysis. The error component of the model was divided into a random cow component and a component due to random error. All models were fitted
Fig. 3. The weekly percentages fat and protein, and log-transformed somatic cell counts ŽlnSCC. in milk of 43 initially-seronegative Holstein–Friesian cows relative to estimated day of infection of these animals Ž1993– 1994, Netherlands..
102
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
by the SAS system, using PROC MIXED ŽSAS, 1988.. Data on fat percentage, protein percentage, and somatic cell count ŽSCC. were described using graphical techniques. No statistical tests were applied since on average, there was only one observation during the infectious period. Data on SCC were log-transformed ŽLn SCC. before evaluation. Reproductive performance in time was graphically evaluated using a Q-sum graph. With this graphical system, all inseminations were sorted with regard to the day of insemination. The line makes an upward step when an insemination is successful, and makes a downward step when the insemination is unsuccessful. The slope of the Q-sum graph is indicative of the proportion of successful inseminations. A horizontal slope indicates 50% pregnancy, an upward slope is associated with more than 50% success, a downward slope with less than 50% success. The absolute value of the Q-sum graph has no specific meaning. The difference between the proportion of successful inseminations during or immediately before the infectious period and the proportion of successful inseminations in a baseline period approximately 4 months before infection was analyzed statistically by calculating exact binomial probabilities ŽSAS, 1988.. In all situations, a P-value of less than 0.05 was considered to be significant. 3. Results The outbreak of BHV1 in this herd has been described before ŽHage et al., 1997.. Within a period of 5 weeks, all seronegative animals seroconverted. Six of the 55
Fig. 4. The mean daily milk production of 43 initially-seronegative and 55 initially-seropositive Holstein–Friesian cows Ž1993–1994, Netherlands..
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
103
Table 1 The results of modelling milk production in 43 initially-seronegative animals during the infectious period of BHV1 infected Holstein–Friesian cows Parameter
Estimate
Standard error
P-value
Intercept DIMr305 ŽDIMr305. 2 logŽ305rDIM. ŽlogŽ305rDIM.. 2 Infection
47.45 y45.34 19.59 y7.13 0.26 y0.68
1.21 2.17 1.04 0.62 0.083 0.069
0.0001 0.0001 0.0001 0.0001 0.0015 0.0001
DIM: Days in milk.
initially-seropositive animals showed a significant rise in antibody titre during the outbreak. During the period of virus circulation, no clinical signs were observed in any of the cows. We clearly monitored a subclinical infection of BHV1 caused by subtype 1 ŽHage et al., 1997.. There were no observed abortions, and milk production Žexpressed as calendar-day means. did not decrease ŽFig. 1.. The mean milk production during the outbreak was 22.0 kg ŽSD 4.0 kg. per cow per day. There was no rise in the number of cows with clinical mastitis cases or clinical lameness, and nor did the mean SCC increase when measured per calendar-day.
Fig. 5. The fertility Q-sum prior to, during and after the infectious period for BHV1 in a Holstein–Friesian herd Ž1993–1994, Netherlands..
104
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
However, when data were sorted according to day relative to infection, a clear decrease in daily milk production became evident in initially-seronegative animals ŽFig. 2.. No decrease in mean fat percentage, mean protein percentage, or Ln SCC was observed ŽFig. 3.. Only a few seropositive animals showed a significant rise in antibody titre. Because of small numbers, these animals were left out of statistical analysis. Therefore, seropositive animals were not sorted according to day relative to re-infection. The daily mean milk production of initially-seropositive and initially-seronegative animals is shown in Fig. 4. The results of statistical modelling of daily milk production in initially-seronegative cows are presented in Table 1. The final model explained 87% of the variation in the data. A significant decrease in milk production of 0.68 kg per day was estimated by using this model. Thus, infected animals had a milk production loss of 9.52 kg during the infectious period of 14 days. The slope of the Q-sum graph showed that there was a decrease in fertility before the infectious period ŽFig. 5.. Infection did not seem to affect pregnancies older than 50 days or younger than 10 days. This decrease in pregnancy rate occurred in both initially-seronegative and initially-seropositive animals. Statistical evaluation of the pregnancy data revealed that the proportion of successful inseminations decreased significantly in the 50 days before the infectious period Žbinomial probability, P - 0.01.. The proportion of successful inseminations in the infectious period, and in the period thereafter, did not significantly differ from the baseline period Žbinomial probability, P ) 0.20.. In both these periods, the slope of the Q-sum graph was more-or-less horizontal, indicating a 50% pregnancy success.
4. Discussion In this study, we describe production and fertility losses associated with an induced natural subclinical infection of BHV1 subtype 1 ŽHage et al., 1997. in a commercial dairy herd. Milk-production loss was estimated using a linear model based on data from the same animals that were also used to estimate the lactation curve during a non-infected period. There may be some bias in this method of estimation. The assumed non-infectious period may be affected by the aftermath of infection, and therefore, milk production may be lower than in truly not-infected cows. The resulting milk-production losses would then be an underestimate of the true milk-production loss. As BHV1 spread through the herd in 7 weeks, new seronegative animals became infected with BHV1 ŽHage et al., 1997. every week. Therefore, the decrease in milk production was also spread over this time period. As a result, a decrease in milk production of these animals could not be observed, because the daily milk production of the entire herd fluctuated within a normal range ŽFig. 1.. Obviously, the day of infection per animal has to be known and daily milk yield should be recorded individually for a precise measurement of the decrease in milk production in a herd during an outbreak of BHV1.
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
105
As a group, the initially-seropositive cows were likely well protected against this BHV1 strain, as the mean daily milk production of these animals showed an apparent increase during the period BHV1 circulated. Therefore, we conclude that there was an immediate decrease in the milk production of the initially-seronegative animals due to a subclinical infection with BHV1. We did not find an association between BHV1 infection and mastitis, since SCC was unaltered and clinical mastitis was not observed. Such a relationship between BHV1 infection and clinical mastitis has been reported previously ŽGreig and Bannister, 1965; Corner et al., 1967; Gourlay et al., 1974; Roberts et al., 1974; Siegler et al., 1984., but different BHV1 strains may have different effects on the development of mastitis. Approximately 50% of all inseminations were successful before and after the infectious period, and there was no decrease in the conception rate during the same period. Abortions were also not seen. However, in the 50 days before the infectious period, almost 100% of inseminations were unsuccessful. The fall in the pregnancy rate prior to the infectious period in initially-seronegative animals could mean that early pregnancies were terminated by a BHV1 infection. But the results were virtually the same for initially-seronegative and initially-seropositive animals. This makes a causative relationship Žif any. between BHV1 circulation and pregnancy losses doubtful. Since these data were based on observations on a single farm, we cannot argue with certainty that there is a causal association between BHV1 infection and a decrease in production andror reproduction. It cannot completely be excluded that other farm factors also changed during the observed BHV1 outbreak. Essentially, these data describe a case, and are not necessarily representative of an ‘average’ subclinical BHV1 outbreak. Further observations based on several herds with a similar outbreaks would be required to distinguish between Žrandom. confounding factors and true production and fertility losses due to BHV1 infection.
5. Conclusions Our data show that there was a significant decrease in milk production in initiallyseronegative cows that became infected but not in seropositive cows. A significant decrease in the pregnancy rate was observed in all cows. The effect on milk production is most likely due to the infection with BHV1. Fertility losses are not explained by BHV1 infection, since initially, seropositive animals also showed losses similar to those of cows that later seroconverted. BHV1 infection did not cause clinical disease or abortion, and did not affect the SCC and milk components.
Acknowledgements The authors are very grateful to H. and P. van de Bij, farmers on Taco’s Each, Aldeboarn, Netherlands, G.A. Hooijer, veterinary practitioner, Aldeboarn, The Netherlands, and D. Flapper for their help in sorting data of milk production and fertility.
106
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
References Ali, T.E., Schaeffer, L.R., 1987. Accounting for covariances among test day milk yields in dairy cows. Can. J. Anim. Sci. 67, 637–642. Bielanski, A., Dubuc, C., 1993. In vitro fertilization of bovine oocytes exposed to bovine herpesvirus 1 ŽBHV1.. Reprod. Domestic Anim. 28, 285–288. Bielanski, A., Loewen, K.S., Delcampo, M.R., Sirard, M.A., Willadsen, S., 1993. Isolation of bovine herpesvirus 1 ŽBHV1. and bovine virus diarrhea virus ŽBVDV. in association with the in vitro production of bovine embryos. Theriogenology 40, 531–538. Bowen, R.A., Elsden, R.P., Seidel, G.E. Jr., 1985. Infection of early bovine embryos with bovine herpesvirus 1. Am. J. Vet. Res. 46, 1095–1097. Chiang, B.C., Smith, P.C., Nusbaum, K.E., Stringfellow, D.A., 1990. The effect of infectious bovine rhinotracheitis vaccine on reproductive efficiency in cattle vaccinated during oestrus. Theriogenology 33, 1113–1120. Corner, A.H., Greig, A.S., Hill, D.P., 1967. A histological study of the effects of the herpesvirus of infectious bovine rhinotracheitis in the lactating bovine mammary gland. Can. J. Comp. Med. 21, 320–330. Cortese, V.S., 1991. Effects on milk production by a nonabortigenic combination vaccine for prevention of respiratory and reproductive system diseases in cattle. Agri-Practice 12, 21–26. Deluyker, H.A., Shumway, R.H., Wecker, W.E., Azari, A.S., Weaver, L.D., 1990. Modelling daily milk yield in Holstein cows using time series analysis. J. Dairy Sci. 73, 539–548. Gibbs, E.P.J., Rweyemamu, M.M., 1977. Bovine herpesviruses. Part 1. Vet. Bull. 47, 317–343. Gourlay, R.N., Stott, E.J., Espinasse, J., Barle, C., 1974. Isolation of mycoplasma agalactia var. bovis and infectious bovine rhinotracheitis virus from an outbreak of mastitis in France. Vet. Rec. 95, 534–535. Greig, A.S., Bannister, G.L., 1965. Infection of the bovine udder with bovine herpesvirus. Can. J. Comp. Med. 29, 57–62. Hage, J.J., Schukken, Y.H., Barkema, H.W., Benedictus, G., Rijsewijk, F.A.M., Wentink, G.H., 1997. Population dynamics of BHV1 infection in a dairy herd. Vet. Microbiol. 53, 169–180. Higgings, R.J., Edwards, S., 1986. Systemic neonatal infectious bovine rhinotracheitis virus infection in suckler calves. Vet. Rec. 119, 177–178. Kramps, J.A., Magdalena, J., Quak, J., Weerdmeester, K., Kaashoek, M.J., Maris-Veldhuis, M.A., Rijsewijk, F.A.M., Keil, G., Van Oirschot, J.T., 1994. A simple, specific, and highly sensitive blocking enzyme-linked immunosorbent assay for detection of antibodies to bovine herpesvirus 1. J. Clin. Microbiol. 32, 2175–2181. Miller, J.M., van der Maaten, M.J., 1987. Early embryonic death in heifers after inoculation with bovine herpesvirus 1 and reactivation of latent virus in reproductive tissues. Am. J. Vet. Res. 48, 1555–1558. Miller, J.M., Van Der Maaten, M.J., Whetstone, C.A., 1988. Effects of a bovine herpesvirus 1 isolate on reproductive function in heifers: classification as a type-2 Žinfectious pustular vulvovaginitis. virus by restriction endonuclease analysis of viral DNA. Am. J. Vet. Res. 49, 1653–1656. Miller, J.M., Whetstone, C.A., Van Der Maaten, M.J., 1991. Abortifacient property of bovine herpesvirus type 1 isolates that represent three subtypes determined by restriction endonuclease analysis of viral DNA. Am. J. Vet. Res. 52, 458–461. Roberts, A.W., Carter, G.R., Carter, F.A., 1974. Infectious bovine rhinotracheitis virus recovered from the milk of a cow with mastitis. J. Am. Vet. Med. Assoc. 164, 413. SASrSTAT User’s Guide: release 6.03 edn. 1988. SAS, Cary, NC. Siegler, H.H., Marschang, F., Morscher, H., 1984. Beobachtungen uber zwischen virusinfek¨ zusammenhange ¨ tionen und boviner mastitis. Tierartzl. Umschau 39, 602–604. Van Oirschot, J.T., Straver, P.J., Van Lieshout, J.A.H., Quak, J., Westenbrink, F., Van Exsel, A.C.A., 1993. A subclinical infection of bulls with bovine herpesvirus type 1 at an artificial insemination centre. Vet. Rec. 132, 32–35.
Milk production and reproduction during a subclinical bovine herpesvirus 1 infection on a dairy farm J.J. Hage
a,)
, Y.H. Schukken b, Th. Dijkstra a , H.W. Barkema a , P.H.R. van Valkengoed b, G.H. Wentink c
a
b
Animal Health SerÕices, P.O. Box 361, 9200 AJ Drachten, Netherlands Department of Herd Health and Reproduction, UniÕersity of Utrecht, P.O. Box 80.151, 3508 TD Utrecht, Netherlands c Holland Genetics, P.O. Box 5073, 6802 EB Arnhem, Netherlands Accepted 29 October 1997
Abstract This study describes an outbreak of bovine herpesvirus 1 ŽBHV1. infections in a dairy herd with special reference to disease symptoms, reproductive performance and milk production losses. The study was carried out with a dairy herd consisting of 98 lactating animals. All animals were housed in the same freestall barn with intensive contact between all animals. An outbreak of BHV1 was induced by injecting three seropositive cows with dexamethasone. During the outbreak, no clinical signs were observed in any of the newly infected animals. At the time of infection, a significant drop in milk production was noted in animals that were initially-seronegative. The production loss was estimated at approximately 9.5 l per infected animal during the infectious period of 14 days. None of the pregnant cows aborted because of BHV1 infection. During 50 days before BHV1 circulation, there was a significant decrease in the number of successful inseminations in both seronegative and seropositive animals. Therefore, it is doubtful that early pregnancies were terminated by BHV1 infection. The proportion of successful inseminations during the BHV1 circulation in this herd, and in the period thereafter, did not significantly differ from the baseline period. q 1998 Elsevier Science B.V. Keywords: Bovine herpesvirus; Cattle, microbiological diseases; Milk production; Reproductive performance
)
Corresponding author. Tel.: q31 512 570700; fax: q31 512 520013.
0167-5877r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 7 - 5 8 7 7 Ž 9 7 . 0 0 0 8 8 - 3
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
98
1. Introduction Bovine herpes virus 1 ŽBHV1. is the etiological agent of infectious bovine rhinotracheitis ŽIBR., infectious pustular vulvovaginitis ŽIPV., and infectious pustular balanoposthitis ŽIPB., abortion, and fatal multi-systemic infection in newborns ŽGibbs and Rweyemamu, 1977; Higgings and Edwards, 1986.. BHV1 is an important pathogen of cattle and causes substantial economic losses. Because of its impact on animal health and farm economics, European countries are making a concerted effort to control or eradicate BHV1 infection. 1 BHV1 infection can also occur as a subclinical infection ŽVan Oirschot et al., 1993; Hage et al., 1997.. However, little is known about the economic impact of a subclinical BHV1 infection on milk production and reproductive performance in a dairy herd. Cortese Ž1991. did not observe a decrease in milk production after vaccinating cows with a genetically altered live IBR vaccine. BHV1 is reported to give rise to high milk cell counts in the absence of mastitis pathogens ŽSiegler et al., 1984.. BHV1 can cause mastitis if inoculated into the udder ŽGreig and Bannister, 1965; Corner et al., 1967., and has been isolated from cattle with mastitis ŽGourlay et al., 1974; Roberts et al., 1974.. Miller and van der Maaten Ž1987. isolated BHV1 from reproductive tissues Žovary, infundibulum, and uterine tube. from intravenously infected cattle, and BHV1 has been detected in follicular fluid and oocytes ŽGibbs and Rweyemamu, 1977; Bielanski et al., 1993.. A decrease in the conception rate of initially-seronegative animals after experimental BHV1 infection has been described ŽMiller et al., 1988; Chiang et al., 1990.. Yet in the presence of BHV1, oocytes can be fertilized in vitro and co-cultured to the blastocyst stage; there was almost no difference with the control group ŽBielanski and Dubuc, 1993.. Experimental BHV1 infection has been described to cause embryonic death in vitro ŽBowen et al., 1985. and in early pregnancy in seronegative cattle ŽMiller and van der Maaten, 1987.. However, Miller et al. Ž1988. showed that initially, seronegative heifers carried their fetuses to term when inoculated with BHV1 14 days after mating. If cattle abort as a result of BHV1 infection, abortion usually occurs between the 4th and 7th month of gestation at a variable period after infection ŽGibbs and Rweyemamu, 1977.. Miller et al. Ž1991. reported that strains of BHV1 had different abortifacient properties. Thus, heifers inoculated with BHV1 subtypes 1 and 2.a aborted between 17 and 85 days after inoculation, whereas heifers inoculated with BHV1 subtype 2.b delivered full-term calves. There is also an inter-experiment difference in the abortifacient property of one particular strain of BHV1 ŽMiller et al., 1988, 1991.. Most of the above-mentioned data were obtained from experimental infections. However, it is not clear whether natural infections will also lead to similar production or fertility losses. In this study, we aim to describe milk production and fertility losses associated with an Žinduced. outbreak of BHV1 infections on a commercial dairy farm.
1
EC directives 88r407 and 93r60.
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
99
2. Materials and methods 2.1. Farm The study was carried out with a Holstein–Friesian dairy herd. The study period started at November 9, 1993 and ended April 29, 1994. The fat Ž4.0%. and protein Ž3.3%. corrected milk production was 8651 kg for 305 days. This study concerns a total of 110 Žincomplete. lactations of 98 animals. Fifty-five animals were seropositive and 43 animals were seronegative for BHV1; 40 of the latter cows were first-lactation animals, whereas seropositive cows were all second lactation or higher. During the study, animals were housed in the same freestall barn with intensive contact between all animals. All animals were identified by freeze-branding. None of the animals had ever been vaccinated against BHV1. No animals were introduced into the herd during the study. The population dynamics of this BHV1 circulation has been described before ŽHage et al., 1997.. Briefly, three seropositive cows were given dexamethasone injections Ž0.1 mgrkg. and subsequently all seronegative animals seroconverted within 5 weeks. All animals were observed daily for occurrence of clinical disease and reproductive problems, such as vaginal discharge or abortions. 2.2. Serology During the experiment, blood samples were collected from all animals at weekly intervals and were screened for antibodies against BHV1, using an undiluted gB-blocking ELISA Žspecificity and sensitivity 0.96 and 0.99, respectively. ŽKramps et al., 1994.. The positive serum samples were titrated by serial dilution. Seroconversion was considered to have occurred when a blood sample was negative for BHV1 in one sample but positive in the next sample. Animals were assumed to seroconvert 8 days after infection ŽKramps et al., 1994.. Re-infection was defined as a significant rise in antibody titre Ža 4-fold or higher rise in antibody titre ŽG 2 dilution steps.. in a 14-day period. Because of the 1-week intervals of bloodsampling, seronegative animals were considered to have been infected 10 days Ž1.5 week. before they seroconverted or showed a significant increase in antibody titre. In initially-seropositive animals, the moment of re-infection was considered to be 10 days before there was a significant rise in antibody titre ŽHage et al., 1997.. To estimate the effect of a BHV1 infection on milk production, the day of BHV1 infection of each initially-seronegative lactating animal was considered to be Day 0 ŽD0.. 2.3. Milk production and fertility The lactating animals were milked twice a day. The milk yield was recorded automatically, and morning and evening yields were summed to obtain daily milk production. Data were checked for missing values. When single measurements were missing, these were replaced by the mean of the previous and next production for the same part of the day Žmorning or evening. ŽDeluyker et al., 1990.. At weekly intervals, the percentages fat and protein and somatic cell counts ŽSCC. were measured via the
100
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
Fig. 1. The mean daily milk production of all 98 lactating Holstein–Friesian cattle Ž1993–1994, Netherlands..
Fig. 2. Daily milk production relative to estimated day of infection of 43 initially seronegative Holstein–Friesian cattle Ž1993–1994, Netherlands..
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
101
Dutch milk-recording system. Records were kept of all insemination and calving data over the year before and the year following the start of this experiment. When an insemination led to calving 9 months later or an observed abortion, it was defined as successful. An insemination that was followed by another insemination, or by a negative pregnancy check, was defined as unsuccessful. 2.4. Statistical methods Milk-production data were analysed by using a regression approach. A linearized model to describe the lactation curve was estimated according to the method described by Ali and Schaeffer Ž1987.. This basic daily milk production model can be expressed by: 2
2
y s b 1 DIMr305q b 2 Ž DIMr305. q b 3 log Ž 305rDIM. q b4 Ž log Ž 305rDIM. . where: y s observed daily milk production, DIM s days in milk, and log is natural logarithm. This model was extended by an extra term describing the length of the infectious period. This variable was coded as 1 during the first days after infection ŽD0–D14. and 0 for the remaining days. The regression coefficient of this parameter gives an estimate of the mean production loss per animal per day during the infectious period. Parity 2 animals were not included since only three animals with parity ) 1 were available for analysis. The error component of the model was divided into a random cow component and a component due to random error. All models were fitted
Fig. 3. The weekly percentages fat and protein, and log-transformed somatic cell counts ŽlnSCC. in milk of 43 initially-seronegative Holstein–Friesian cows relative to estimated day of infection of these animals Ž1993– 1994, Netherlands..
102
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
by the SAS system, using PROC MIXED ŽSAS, 1988.. Data on fat percentage, protein percentage, and somatic cell count ŽSCC. were described using graphical techniques. No statistical tests were applied since on average, there was only one observation during the infectious period. Data on SCC were log-transformed ŽLn SCC. before evaluation. Reproductive performance in time was graphically evaluated using a Q-sum graph. With this graphical system, all inseminations were sorted with regard to the day of insemination. The line makes an upward step when an insemination is successful, and makes a downward step when the insemination is unsuccessful. The slope of the Q-sum graph is indicative of the proportion of successful inseminations. A horizontal slope indicates 50% pregnancy, an upward slope is associated with more than 50% success, a downward slope with less than 50% success. The absolute value of the Q-sum graph has no specific meaning. The difference between the proportion of successful inseminations during or immediately before the infectious period and the proportion of successful inseminations in a baseline period approximately 4 months before infection was analyzed statistically by calculating exact binomial probabilities ŽSAS, 1988.. In all situations, a P-value of less than 0.05 was considered to be significant. 3. Results The outbreak of BHV1 in this herd has been described before ŽHage et al., 1997.. Within a period of 5 weeks, all seronegative animals seroconverted. Six of the 55
Fig. 4. The mean daily milk production of 43 initially-seronegative and 55 initially-seropositive Holstein–Friesian cows Ž1993–1994, Netherlands..
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
103
Table 1 The results of modelling milk production in 43 initially-seronegative animals during the infectious period of BHV1 infected Holstein–Friesian cows Parameter
Estimate
Standard error
P-value
Intercept DIMr305 ŽDIMr305. 2 logŽ305rDIM. ŽlogŽ305rDIM.. 2 Infection
47.45 y45.34 19.59 y7.13 0.26 y0.68
1.21 2.17 1.04 0.62 0.083 0.069
0.0001 0.0001 0.0001 0.0001 0.0015 0.0001
DIM: Days in milk.
initially-seropositive animals showed a significant rise in antibody titre during the outbreak. During the period of virus circulation, no clinical signs were observed in any of the cows. We clearly monitored a subclinical infection of BHV1 caused by subtype 1 ŽHage et al., 1997.. There were no observed abortions, and milk production Žexpressed as calendar-day means. did not decrease ŽFig. 1.. The mean milk production during the outbreak was 22.0 kg ŽSD 4.0 kg. per cow per day. There was no rise in the number of cows with clinical mastitis cases or clinical lameness, and nor did the mean SCC increase when measured per calendar-day.
Fig. 5. The fertility Q-sum prior to, during and after the infectious period for BHV1 in a Holstein–Friesian herd Ž1993–1994, Netherlands..
104
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
However, when data were sorted according to day relative to infection, a clear decrease in daily milk production became evident in initially-seronegative animals ŽFig. 2.. No decrease in mean fat percentage, mean protein percentage, or Ln SCC was observed ŽFig. 3.. Only a few seropositive animals showed a significant rise in antibody titre. Because of small numbers, these animals were left out of statistical analysis. Therefore, seropositive animals were not sorted according to day relative to re-infection. The daily mean milk production of initially-seropositive and initially-seronegative animals is shown in Fig. 4. The results of statistical modelling of daily milk production in initially-seronegative cows are presented in Table 1. The final model explained 87% of the variation in the data. A significant decrease in milk production of 0.68 kg per day was estimated by using this model. Thus, infected animals had a milk production loss of 9.52 kg during the infectious period of 14 days. The slope of the Q-sum graph showed that there was a decrease in fertility before the infectious period ŽFig. 5.. Infection did not seem to affect pregnancies older than 50 days or younger than 10 days. This decrease in pregnancy rate occurred in both initially-seronegative and initially-seropositive animals. Statistical evaluation of the pregnancy data revealed that the proportion of successful inseminations decreased significantly in the 50 days before the infectious period Žbinomial probability, P - 0.01.. The proportion of successful inseminations in the infectious period, and in the period thereafter, did not significantly differ from the baseline period Žbinomial probability, P ) 0.20.. In both these periods, the slope of the Q-sum graph was more-or-less horizontal, indicating a 50% pregnancy success.
4. Discussion In this study, we describe production and fertility losses associated with an induced natural subclinical infection of BHV1 subtype 1 ŽHage et al., 1997. in a commercial dairy herd. Milk-production loss was estimated using a linear model based on data from the same animals that were also used to estimate the lactation curve during a non-infected period. There may be some bias in this method of estimation. The assumed non-infectious period may be affected by the aftermath of infection, and therefore, milk production may be lower than in truly not-infected cows. The resulting milk-production losses would then be an underestimate of the true milk-production loss. As BHV1 spread through the herd in 7 weeks, new seronegative animals became infected with BHV1 ŽHage et al., 1997. every week. Therefore, the decrease in milk production was also spread over this time period. As a result, a decrease in milk production of these animals could not be observed, because the daily milk production of the entire herd fluctuated within a normal range ŽFig. 1.. Obviously, the day of infection per animal has to be known and daily milk yield should be recorded individually for a precise measurement of the decrease in milk production in a herd during an outbreak of BHV1.
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
105
As a group, the initially-seropositive cows were likely well protected against this BHV1 strain, as the mean daily milk production of these animals showed an apparent increase during the period BHV1 circulated. Therefore, we conclude that there was an immediate decrease in the milk production of the initially-seronegative animals due to a subclinical infection with BHV1. We did not find an association between BHV1 infection and mastitis, since SCC was unaltered and clinical mastitis was not observed. Such a relationship between BHV1 infection and clinical mastitis has been reported previously ŽGreig and Bannister, 1965; Corner et al., 1967; Gourlay et al., 1974; Roberts et al., 1974; Siegler et al., 1984., but different BHV1 strains may have different effects on the development of mastitis. Approximately 50% of all inseminations were successful before and after the infectious period, and there was no decrease in the conception rate during the same period. Abortions were also not seen. However, in the 50 days before the infectious period, almost 100% of inseminations were unsuccessful. The fall in the pregnancy rate prior to the infectious period in initially-seronegative animals could mean that early pregnancies were terminated by a BHV1 infection. But the results were virtually the same for initially-seronegative and initially-seropositive animals. This makes a causative relationship Žif any. between BHV1 circulation and pregnancy losses doubtful. Since these data were based on observations on a single farm, we cannot argue with certainty that there is a causal association between BHV1 infection and a decrease in production andror reproduction. It cannot completely be excluded that other farm factors also changed during the observed BHV1 outbreak. Essentially, these data describe a case, and are not necessarily representative of an ‘average’ subclinical BHV1 outbreak. Further observations based on several herds with a similar outbreaks would be required to distinguish between Žrandom. confounding factors and true production and fertility losses due to BHV1 infection.
5. Conclusions Our data show that there was a significant decrease in milk production in initiallyseronegative cows that became infected but not in seropositive cows. A significant decrease in the pregnancy rate was observed in all cows. The effect on milk production is most likely due to the infection with BHV1. Fertility losses are not explained by BHV1 infection, since initially, seropositive animals also showed losses similar to those of cows that later seroconverted. BHV1 infection did not cause clinical disease or abortion, and did not affect the SCC and milk components.
Acknowledgements The authors are very grateful to H. and P. van de Bij, farmers on Taco’s Each, Aldeboarn, Netherlands, G.A. Hooijer, veterinary practitioner, Aldeboarn, The Netherlands, and D. Flapper for their help in sorting data of milk production and fertility.
106
J.J. Hage et al.r PreÕentiÕe Veterinary Medicine 34 (1998) 97–106
References Ali, T.E., Schaeffer, L.R., 1987. Accounting for covariances among test day milk yields in dairy cows. Can. J. Anim. Sci. 67, 637–642. Bielanski, A., Dubuc, C., 1993. In vitro fertilization of bovine oocytes exposed to bovine herpesvirus 1 ŽBHV1.. Reprod. Domestic Anim. 28, 285–288. Bielanski, A., Loewen, K.S., Delcampo, M.R., Sirard, M.A., Willadsen, S., 1993. Isolation of bovine herpesvirus 1 ŽBHV1. and bovine virus diarrhea virus ŽBVDV. in association with the in vitro production of bovine embryos. Theriogenology 40, 531–538. Bowen, R.A., Elsden, R.P., Seidel, G.E. Jr., 1985. Infection of early bovine embryos with bovine herpesvirus 1. Am. J. Vet. Res. 46, 1095–1097. Chiang, B.C., Smith, P.C., Nusbaum, K.E., Stringfellow, D.A., 1990. The effect of infectious bovine rhinotracheitis vaccine on reproductive efficiency in cattle vaccinated during oestrus. Theriogenology 33, 1113–1120. Corner, A.H., Greig, A.S., Hill, D.P., 1967. A histological study of the effects of the herpesvirus of infectious bovine rhinotracheitis in the lactating bovine mammary gland. Can. J. Comp. Med. 21, 320–330. Cortese, V.S., 1991. Effects on milk production by a nonabortigenic combination vaccine for prevention of respiratory and reproductive system diseases in cattle. Agri-Practice 12, 21–26. Deluyker, H.A., Shumway, R.H., Wecker, W.E., Azari, A.S., Weaver, L.D., 1990. Modelling daily milk yield in Holstein cows using time series analysis. J. Dairy Sci. 73, 539–548. Gibbs, E.P.J., Rweyemamu, M.M., 1977. Bovine herpesviruses. Part 1. Vet. Bull. 47, 317–343. Gourlay, R.N., Stott, E.J., Espinasse, J., Barle, C., 1974. Isolation of mycoplasma agalactia var. bovis and infectious bovine rhinotracheitis virus from an outbreak of mastitis in France. Vet. Rec. 95, 534–535. Greig, A.S., Bannister, G.L., 1965. Infection of the bovine udder with bovine herpesvirus. Can. J. Comp. Med. 29, 57–62. Hage, J.J., Schukken, Y.H., Barkema, H.W., Benedictus, G., Rijsewijk, F.A.M., Wentink, G.H., 1997. Population dynamics of BHV1 infection in a dairy herd. Vet. Microbiol. 53, 169–180. Higgings, R.J., Edwards, S., 1986. Systemic neonatal infectious bovine rhinotracheitis virus infection in suckler calves. Vet. Rec. 119, 177–178. Kramps, J.A., Magdalena, J., Quak, J., Weerdmeester, K., Kaashoek, M.J., Maris-Veldhuis, M.A., Rijsewijk, F.A.M., Keil, G., Van Oirschot, J.T., 1994. A simple, specific, and highly sensitive blocking enzyme-linked immunosorbent assay for detection of antibodies to bovine herpesvirus 1. J. Clin. Microbiol. 32, 2175–2181. Miller, J.M., van der Maaten, M.J., 1987. Early embryonic death in heifers after inoculation with bovine herpesvirus 1 and reactivation of latent virus in reproductive tissues. Am. J. Vet. Res. 48, 1555–1558. Miller, J.M., Van Der Maaten, M.J., Whetstone, C.A., 1988. Effects of a bovine herpesvirus 1 isolate on reproductive function in heifers: classification as a type-2 Žinfectious pustular vulvovaginitis. virus by restriction endonuclease analysis of viral DNA. Am. J. Vet. Res. 49, 1653–1656. Miller, J.M., Whetstone, C.A., Van Der Maaten, M.J., 1991. Abortifacient property of bovine herpesvirus type 1 isolates that represent three subtypes determined by restriction endonuclease analysis of viral DNA. Am. J. Vet. Res. 52, 458–461. Roberts, A.W., Carter, G.R., Carter, F.A., 1974. Infectious bovine rhinotracheitis virus recovered from the milk of a cow with mastitis. J. Am. Vet. Med. Assoc. 164, 413. SASrSTAT User’s Guide: release 6.03 edn. 1988. SAS, Cary, NC. Siegler, H.H., Marschang, F., Morscher, H., 1984. Beobachtungen uber zwischen virusinfek¨ zusammenhange ¨ tionen und boviner mastitis. Tierartzl. Umschau 39, 602–604. Van Oirschot, J.T., Straver, P.J., Van Lieshout, J.A.H., Quak, J., Westenbrink, F., Van Exsel, A.C.A., 1993. A subclinical infection of bulls with bovine herpesvirus type 1 at an artificial insemination centre. Vet. Rec. 132, 32–35.