- Email: [email protected]
Prevalence and incidence of subclinical mastitis in goats and dairy ewes in Vermont, USA
Small Ruminant Research 46 (2002) 115–121
Prevalence and incidence of subclinical mastitis in goats and dairy ewes in Vermont, USA Scott McDougall a,∗ , Woody Pankey b , Carol Delaney b , John Barlow b , Patricia A. Murdough b , Dan Scruton c b
a Animal Health Centre, P.O. Box 21, Morrinsville, New Zealand Department of Animal Sciences, University of Vermont, Burlington, VT, USA c Vermont Department of Agriculture, Montpelier, VT, USA
Accepted 28 July 2002
Abstract The prevalence of bacterial isolation and the somatic cell count (SCC) of the milk from goats (n = 110 from six herds) and sheep (n = 153 from three herds) was determined at parturition and approximately 40 days later. Incidence of new intramammary infection (IMI) and the spontaneous cure rate were determined over this time period. In the goats, 27.3 and 25.5% were infected at parturition and 40 days later, respectively, while 15.0 and 9.1% of sheep were infected at parturition and 40 days later, respectively. Incidence of new infection was 0.039 and 0.034 cases/half/30 days for goats and sheep, respectively, and did not differ between the species (P > 0.2). Spontaneous cure occurred in 93.8% of sheep halves infected at parturition but only 50.0% of goat halves (P < 0.05). Coagulase negative staphylococci (CNS) were the most common isolates from both sheep and goats. Milk from bacteriologically positive halves had a significantly higher somatic cell count than halves from which no bacteria were isolated in both sheep and goats. It was concluded that bacterial infection of the mammary gland was associated with an elevated somatic cell count. Additionally, differences were detected between the species in incidence of new infection and spontaneous cure rate which resulted in a decline in prevalence in sheep, but not goats, with time postpartum. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Mastitis; Sheep; Goats; Prevalence; Incidence
1. Introduction Control of mastitis and somatic cell counts (SCC) in dairy animals is aided by an understanding of the pathogens involved, the rate of new intramammary infection (IMI), the spontaneous cure rate and the source of infection. Defining periods of high risk of new IMI ∗ Corresponding author. Tel.: +64-7889-5159; fax: +64-7889-3681. E-mail address: [email protected] (S. McDougall).
will aid in development of management practices that may reduce new IMI. Coagulase negative staphylococci (CNS) are the most prevalent pathogens of the mammary gland of sheep and goats (Poutrel and Lerondelle, 1983; Maisi et al., 1987; de la Cruz et al., 1994; Fthenakis, 1994). However, the dynamics of CNS IMI of goats and sheep are not well documented. Many studies have defined prevalence of individual pathogens at one point of lactation. Prevalence of IMI is dependant on the rate of new IMI (i.e. the incidence of infection; Thrusfield,
0921-4488/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 1 - 4 4 8 8 ( 0 2 ) 0 0 1 9 1 - 8
116
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
1986) and the duration of existing infections, which is affected by spontaneous cure rate and therapy (Lam, 1996). However, few studies have quantified the incidence rate of new IMI in small ruminants. The aim of the present study was to determine the prevalence of infection at parturition and 40 days later and to quantify the incidence and spontaneous cure rate over this period.
2. Materials and methods Goats (n = 110) and sheep (n = 155) from six and third herds and flocks, respectively, in Vermont, USA that underwent parturition between February and April 1999 were enrolled in the study. All animals were housed over winter. The sheep flocks lambed seasonally with all lambings occurring between March and May. Lambs were left on the ewes for approximately 4 weeks after lambing until weaning in April or May following which the ewes were moved to pasture. The ewes were then milked twice daily. All goat herds had multiple kidding periods, with some goats in milk at all times. The goat herds remained housed throughout the study. The kids were removed within 24 h of kidding and does were milked twice daily. Housed animals were fed hay and concentrates and water was available at all times, while the ewes were pasture grazed once weaning had occurred. All sheep flocks and five of the six goat herds machine-milked the animals while the goats in one herd were hand milked. Herdowners were given sterile tubes, alcohol and swabs and instructed in collection of milk samples aseptically and asked to collect milk from both halves within 24 h of kidding or lambing. The teat-end was prepared for milking by the herdowner following the normal herd protocol. The teat-end was then scrubbed with a cotton wool pledget wetted with 70% alcohol. After the teat-end had dried, the first strip of milk was discarded and then approximately 10 ml of milk was expressed into a sterile test tube, the tube was capped and labelled with herd, animal number or name and half. The herdowner-collected samples were frozen at −20 ◦ C for up to 1 M before culture. Between 11 and 50 animals from each herd were sampled. Herdowners were also asked to record and sample any animal which was diagnosed with clinical mastitis. On 1 day, approximately 6 weeks (average = 40.5 ± 8.1
days, mean ± S.D.) after parturition, a milk sample was drawn from each half of each enrolled animal for bacterial culture and somatic cell count determination by the investigators. These samples were kept on ice and transported to the laboratory for culture. 2.1. Milk sampling and bacteriological procedures The samples were thawed (if required), mixed and 0.01 ml was streaked onto a half plate of 5% trypticase washed-sheep-blood agar (TBA; Micro-Diagnostics Inc., Lombard, IL). Plates were incubated at 37 ◦ C for 48 h. The colonies were provisionally speciated as Staphylococcus aureus, CNS, Streptococcus agalactiae, other streptococci, Coliforms or Corynebacterium bovis on morphology, haemolysis pattern and Gram stain. The numbers of each distinct colony type were recorded as <6, >5 and <11, >10 and <51, and >50 colony forming units (CFU) per plate. Where >5 CFU of multiple colony types were found, representative colonies were subcultured on TBA. Gram-positive cocci were tested for catalase and coagulase production and Gram-negative isolates were further differentiated using MIO deeps (motility, indole, ornithine decarboxylase; Difco Laboratories, Detroit, MI). An IMI was diagnosed when >500 CFU/ml of each of 1 to 3 colony types were isolated. This figure was the normal definition of IMI within the laboratory. A new IMI was diagnosed when >500 CFU/ml of 1 to 3 colony types were isolated from a previously uninfected gland. Cure was defined as having occurred when no IMI was detected in a previously infected gland. Samples from which >2 colony types were isolated or, where <5 colonies of any pathogen were isolated were regarded as contaminated or uninfected, respectively. Prevalence was defined as the number of halves (or animals) from which a pathogen was isolated divided by the total number of halves (or animals) sampled at each time period. The incidence of new IMI was calculated as the total number of new IMI divided by the total number of days between samplings for halves with no infection at the initial sampling. The numerator was multiplied by 30 to express the incidence rate as the number of glands newly infected over a 30-day period. Following sampling for bacteriology, the remaining milk was preserved with one drop of 2-bromo-2nitropropane-1,3-diol (Preservo Liquid, D&F Control
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
Systems, San Francisco, CA) and incubated for >16 h at 4 ◦ C. Somatic cell count were determined using 500 l of the preserved milk (Fossomatic-90 A/S N. Foss Electric, Denmark). The effect of age, herd, species, day of year at parturition (Julian date) and days postpartum at second sampling on the prevalence at parturition and 40 days later was examined by logistic regression. Forward stepwise analysis was used with log likelihood as the selection methodology. The χ 2 statistics were used to compare proportions between species. The somatic cell counts were not normally distributed and were log10 transformed before analysis. The somatic cell counts were analysed by ANOVA with animal species, bacterial status and the species by status interaction as factors in the model. The change in somatic cell count from parturition to 40 days postpartum was analysed by subtracting the two figures and the difference was analysed by ANOVA.
3. Results Seven animals (six sheep and one goat) were diagnosed with unilateral clinical mastitis by the herdowners. Samples were submitted for four of these halves and S. aureus was cultured from two samples, Staphylococcus pulveri/vitulus from one sample and no isolation occurred from the fourth sample. One sheep died and a further two were culled before the end of the trial.
117
More goats than sheep were infected at parturition and at 40 days postpartum (P < 0.05; Table 1). Prevalence of infected goats and halves did not change between parturition and 40 days postpartum (P = 0.3; Table 1). The number of sheep with infection tended to decline with time postpartum (P = 0.12; Table 1) and the number of halves of sheep infected declined significantly postpartum (P < 0.01; Table 1). The number of infected halves increased with age group in goats (6.8, 19.4 and 31.6% of halves from 1 + 2, 3 + 4 and >4-year-old animals; P < 0.05), but not sheep (4.7, 4.3 and 0% of halves from 1 + 2, 3 + 4 and >4-year-old animals, respectively; P > 0.4). The proportion of infected animals at 40 days postpartum was affected by age and species (P < 0.05), but not by parturition date or days postpartum. A total of 14 new IMI in 14 separate animals were detected, seven each in sheep and goats. The incidence rate did not differ between sheep and goats (0.039 and 0.034 cases/half/30 days for goats and sheep, respectively, P > 0.2) and averaged 0.037 new cases/half/30 days. Animals in which one-half was infected at parturition were more likely to have a new IMI than animals which did not have an IMI in either half at parturition (5/15 versus 7/106; odds ratio = 7.6; P < 0.01). Spontaneous cure occurred in 26 of the 38 (68.4%) halves infected at parturition. Goats had a lower spontaneous cure rate than sheep (11/22 (50.0%) versus 15/16 (93.8%), respectively; odds ratio = 0.06; P < 0.05).
Table 1 The percentage of animals with zero-, one- or two-halves with infections at parturition and 40 days later, the number of new infections and the incidence rate of new infection Goats
Number of animals Neither half infected (%) One-half infected (%) Two-halves infected (%) Animals infected (%) Halves infected (%) Contaminated sample or minor infectiona (%) New infection at day 40 (number of glands) New infection at day 40 (% of glands) Incidence rate of new infection (new infection/gland/30 days) a
Sheep
Day 0
Day 40
Day 0
Day 40
110 50.0 19.1 8.2 27.3 35.5 22.7
110 66.4 15.5 10.0 25.5 35.5 9.1 7 6.4 0.039
153 52.9 11.1 3.9 15.0 19.0 32.0
132 69.7 9.1 0.0 9.1 9.1 21.2 7 5.3 0.034
Contaminated sample (>2 colony types) or minor infection (i.e. <500 CFU/ml).
118
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
Table 2 Bacterial species isolated from halves of goats and sheep at parturition and 40 days later Goats
Sheep
Day 0 n Coliforms Corynebacterium spp. Enterococcus faecalis Lactococcus garvieae Micrococcus luteus Pasteurella spp. Pseudomonas aeruginosa Staphylococcus aureus CNS Stomatococcus mucilaginosus Streptococcus gallolyticus Other streptococci Total
Day 40 Percentage
2
5.0
1
2.5
2 35
5.0 87.5
40
n
Day 0 Percentage
3 1
7.7 2.6
1 1 1 4 26
2.6 2.6 2.6 10.3 66.7
1 1
2.6 2.6
n
4 8
39
Day 40 Percentage
33.3 66.7
12
n
Percentage 1
3.6
1 1
3.6 3.6
2 22 1
7.1 78.6 3.6
28
Table 3 Geometric mean somatic cell count (mean × 103 cells/ml) and 95% confidence intervals for foremilk samples from halves of goats and sheep at parturition and 40 days later categorised as having no IMI or having an IMI (i.e. >500 CFU/ml of milk) Species
Goats Goats Sheep Sheep
Status
No growth Infected No growth Infected
n
146 40 210 28
Day 0
Day 40
Mean
95% CI Low
High
1,490 2,584 382 2,263
1,199 1,590 342 1,199
1,851 4,200 427 4,267
Coagulase negative staphylococci were the most common isolates from both goats and sheep and at parturition and 40 days later (Table 2). Sheep had a lower somatic cell count at parturition than goats (P < 0.005; Table 3) and infected halves had a significantly higher somatic cell count than halves with no IMI (P < 0.005). However, there was a significant status (infected, not infected) by species interaction with a lower somatic cell count in non-infected halves in sheep than goats (Table 3). Sheep had a higher somatic cell count at 40 days postpartum than goats (P < 0.005; Table 3). Infected halves had a higher somatic cell count than halves from which no pathogen was isolated at 40 days (P < 0.005; Table 3). There was no species by status interaction (P = 0.3).
Mean
211 1,292 467 4,290
95% CI Low
High
179 790 402 1,713
248 2,115 542 10,747
4. Discussion The prevalence of halves with IMI was 35.5 and 19.0% in goats and sheep at parturition which is similar to figures reported previously (Fthenakis, 1994; de la Cruz et al., 1994; Contreras et al., 1995). Prevalence declined in sheep, but not goats, by 40 days postpartum. A decline in prevalence with time postpartum was reported in ewes (Fthenakis, 1994). Differences in management between the ewes and does included time of weaning, seasonality of parturition and differences among herdowners in management practices. All goats in the present trial were machine or hand milked from within 24 h after parturition, whereas the lambs were left on the ewes until about 30–35 days postpartum. The sheep herds were
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
all seasonal lambing and all used intensive pasture grazing for more than 50% of the year. The goats were all housed for the majority of the year and kidded year round. The housing of the goats may have been more heavily contaminated than that of the sheep due to the continual use of the facilities. Additionally, as there were does lactating at all times of the year, the potential for ongoing transmission of infection from late to early lactation goats during milking or in the barns was always present. In contrast, all sheep were non-lactating at the same time for some months, which may have reduced the potential for ewe to ewe transmission of pathogens within the milking parlour. Prevalence of IMI increased with age in goats in agreement with other studies (Sanchez et al., 1999). Increasing prevalence with age may be due to increased length of exposure to pathogens in older compared to younger animals. Additionally, where the duration of infection is long and the spontaneous cure rate low, prevalence will increase. Prevalence in ewes did not increase with age as has been previously reported (Fthenakis, 1994). The high self-cure rate found in the ewes in the present trial may be the explanation for this. Incidence of new bacterial IMI was similar for goats and sheep in the present study and averaged 0.037 cases/half/30 days. This is similar to the incidence rate of 1%/ewe/week for ewes reported by Hueston et al. (1986). Incidence of new IMI does not appear to have been previously reported for dairy goats. Incidence of IMI for CNS in housed dairy cattle is reported as 1.58 cases/1000 quarter days or about 0.5 cases/quarter/year (Lam, 1996). The goats and sheep appear to have a higher incidence rate than that reported for cattle. However, incidence in the present study was calculated only over the first 40 days of lactation. Incidence of new clinical infections was shown to be higher in early lactation in cattle (Hogan et al., 1989; Lam, 1996). Hueston et al. (1986) noted an increase in incidence rate of new IMI following weaning from 1%/ewe/week pre-weaning to 9.7%/ewe/week post-weaning. In that study ewes were not machine-milked following weaning, so the increased incidence of IMI was associated with involution of the mammary gland. In the present study, the majority (15 of 16) sheep with IMI at lambing did not have an IMI at 40 days
119
postpartum. However, half the goat halves with an IMI were still infected at 40 days postpartum. Spontaneous cure in sheep was reported as 9.5 and 21% between 2 and 8 weeks post-weaning and lambing to 6 weeks postpartum (Fthenakis, 1994; Kirk et al., 1996). The high frequency of suckling and the large litter size (average of >2 lambs/ewe) may have resulted in the ewes being frequently milked dry which may have resulted in the high self-cure rate observed. As reported previously (Maisi et al., 1987; de la Cruz et al., 1994; Fthenakis, 1994; Contreras et al., 1995) CNS were the most common pathogens isolated from the goat and sheep milk. CNS are common isolates from the respiratory tract, the teat skin, the teat-end as well as from milk (Devriese et al., 1985). The significance of CNS isolates from milk needs to be examined further. However, heavy pure cultures of these CNS isolates did occur in the present study and from other studies of goat and sheep milk (Fthenakis, 1994; Contreras et al., 1995; Kirk et al., 1996; Burriel, 1997). CNS isolations have been associated with elevated somatic cell count and increases in concentrations of NAGase, albumin and salt which indicate that mammary gland tissue damage has occurred (Maisi et al., 1987). Repeated isolations of the same CNS species from the same glands have occurred for up to 9 M indicating chronic IMI with CNS occur (Maisi and Riipinen, 1991; Contreras et al., 1997). CNS have also been recovered following fine needle aspirate of the teat cistern, indicating that the CNS are established beyond the teat canal (Maisi and Riipinen, 1991). Experimental infection with S. simulans and S. warnieri resulted in repeated isolations of these pathogens from milk, changes in milk composition and histological changes within the infected glands of sheep (Fthenakis and Jones, 1990; Burriel, 1997). These data support the contention that at least some strains of CNS are pathogenic to the caprine and ovine udder. The epidemiology of CNS IMI in cattle has been mathematically modelled (Lam, 1996). Control measures such as teat spraying and treatment of existing IMI have been successful in controlling contagious pathogens (Bramley and Dodd, 1984). However, despite dry goat therapy being effective in reducing somatic cell count, teat spraying was not successful in one trial (Poutrel et al., 1997). However, as there are differences among active constituents and formulations among teat sprays in efficacy (Pankey et al., 1994) more work needs to
120
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
be done to test the effect of teat spraying on CNS in goats and sheep. Infection significantly increased somatic cell count both at parturition and 40 days later in both in sheep and goats. The somatic cell count in goats declined with time postpartum. A similar decline in somatic cell count over the first few days of lactation was observed in dairy cattle (Harmon, 1994; Lacy-Hulbert and Woolford, 1997). The somatic cell count of infected glands was 6 to 10 times higher than that of uninfected glands 40 days postpartum. However at kidding, the infected goats only had a somatic cell count about twice that of the uninfected glands. This indicates that differentiation of uninfected from infected glands may be easier to achieve some time postpartum rather than immediately after parturition in goats. The somatic cell count was reported to either decline or to increase over the first few weeks of lactation in goats (Lerondelle et al., 1992; Rota et al., 1993) and IMI are associated with an increased somatic cell count (Lerondelle et al., 1992). Hence differences among studies and herds in somatic cell count following parturition may be due to variation in the initial prevalence and incidence of IMI between herds. Uninfected goat halves are reported as having somatic cell count of 0.05 and 0.5 × 106 somatic cell count/ml (Lerondelle et al., 1992; Paape and Capuco, 1997). In contrast, infected goat glands had somatic cell count of 0.3 to 15.1 × 106 cells/ml with variation among bacterial species in the degree of somatic cell count elevation (Dulin et al., 1982; Ryan et al., 1993). The somatic cell count in the present study agree with the previously published data for goats. In sheep, the somatic cell count from uninfected halves is reported as ranging from 0.26 to 1.58 × 106 somatic cell count/ml (Fthenakis, 1994; Burriel, 1997). The basal somatic cell count is affected by stage of lactation and the volume of milk produced (Mavrogenis et al., 1995). Infection increased somatic cell count in sheep. All infected glands had a somatic cell count of >2.0 × 106 somatic cell count/ml (Mavrogenis et al., 1995) and an average of between 3.3 and 3.8 × 106 somatic cell count/ml (Fthenakis, 1994; Burriel, 1997). It is concluded that the dynamics of IMI and somatic cell count differ between sheep and goats. The underlying physiological and/or managerial reasons for these differences require further investigation.
Additionally, studies addressing changes in IMI prevalence and incidence for an entire lactation and non-lactation period need to be undertaken to define periods of high risk of new IMI in these species.
Acknowledgements The help of the herdowners in sampling animals at parturition and allowing the subsequent sample collection in gratefully acknowledged. Bacteriology support was provided by Dina Bawdry, Erin Mulloy and Jeff Watts. Financial support was obtained from the Royal Society of New Zealand in the form of a NZ/USA Co-Operative Science Program, ISAT linkage fund (contract number 99-CSP-42MCDO) and an Animal Science Award from the New Zealand Society of Animal Production, both granted to Scott McDougall. References Bramley, A.J., Dodd, F.H., 1984. Reviews of the progress of dairy science: mastitis control—progress and prospects. J. Dairy Res. 51, 481–512. Burriel, A.R., 1997. Dynamics of intramammary infection in the sheep caused by coagulase-negative staphylococci and its influence on udder tissue and milk-composition. Vet. Rec. 140, 419–423. Contreras, A., Corrales, J.C., Sierra, D., Marco, J., 1995. Prevalence and etiology of nonclinical intramammary infection in Murciano–Granadina goats. Small Rumin. Res. 17, 71–78. Contreras, A., Corrales, J.C., Sanchez, A., Sierra, D., 1997. Persistence of subclinical intramammary pathogens in goats throughout lactation. J. Dairy Sci. 80, 2815–2819. de la Cruz, M., Serrano, E., Montoro, V., Marco, J., Romeo, M., Baselga, R., Albizu, I., Amorena, B., 1994. Etiology and prevalence of subclinical mastitis in the Manchega sheep at mid-late lactation. Small Rumin. Res. 14, 175–180. Devriese, L.A., Schleifer, K.H., Adegoke, G.O., 1985. Identification of coagulase-negative staphylococci from farm animals. J. Appl. Bact. 58, 45–55. Dulin, A.M., Paape, M.J., Wergin, W.P., 1982. Differentiation and enumeration of somatic cells in goat milk. J. Food Protection 45, 435–439. Fthenakis, G.C., 1994. Prevalence and etiology of subclinical mastitis in ewes of southern Greece. Small Rumin. Res. 13, 293–300. Fthenakis, G.C., Jones, J.E.T., 1990. The effect of experimentally induced subclinical mastitis on milk yield of ewes and on the growth of lambs. Br. Vet. J. 146, 43–49. Harmon, R.J., 1994. Physiology of mastitis and factors affecting somatic cell counts. J. Dairy Sci. 72, 2103–2112.
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121 Hogan, J.S., Smith, K.L., Hoblet, K.H., Schoenberger, P.S., Todhunter, D.A., Hueston, W.D., Pritchard, D.E., Bowman, G.L., Heider, L.E., Brockett, B.L., Conrad, H.R., 1989. Field survey of clinical mastitis in low somatic cell count herds. J. Dairy Sci. 72, 1547–1556. Hueston, W.D., Hartwig, N.R., Judy, J.K., 1986. Patterns of nonclinical intramammary infection in a ewe flock. J. Am. Vet. Med. Assoc. 188, 170–172. Kirk, J.H., Glenn, J.S., Maas, J.P., 1996. Mastitis in a flock of milking sheep. Small Rumin. Res. 22, 187–191. Lacy-Hulbert, J., Woolford, M., 1997. Mastitis diagnosis and resolution. In: Hayes, D.P., McDougall, S. (Eds.), Proceedings of the Second International Conference for the Society of Dairy Cattle Veterinarians of the New Zealand Veterinary Association. Vanuatu, Noumea, pp. 24–36. Lam, T.J.G.M., 1996. Dynamics of Bovine Mastitis. Ph.D. thesis, Utrecht. Lerondelle, C., Richard, Y., Issartial, J., 1992. Factors affecting somatic cell counts in goat milk. Small Rumin. Res. 8, 129–139. Maisi, P., Riipinen, I., 1991. Pathogenicity of different species of staphylococci in caprine udder. Br. Vet. J. 147, 126–132. Maisi, P., Juntilla, J., Seppanen, J., 1987. Detection of subclinical mastitis in ewes. Br. Vet. J. 143, 402–409. Mavrogenis, A.P., Koumas, A., Kakoyiannis, C.K., Taliotis, C.H., 1995. Use of somatic-cell counts for the detection of subclinical mastitis in sheep. Small Rumin. Res. 17, 79–84.
121
Paape, M.J., Capuco, A.V., 1997. Cellular defence-mechanisms in the udder and lactation of goats. J. Anim. Sci. 75, 556–565. Pankey, J.W., Eberhart, R.J., Cumming, R.D., Daggett, R.D., Farnsworth, R.J., McDuff, C.K., 1994. Update on postmilking teat antisepsis. J. Dairy Sci. 67, 1336–1353. Poutrel, B., Lerondelle, C., 1983. Cell content of goat milk: California mastitis test, coulter counter and fossomatic for predicting half infection. J. Dairy Sci. 66, 2575–2579. Poutrel, B., Decremoux, R., Ducelliez, M., Verneau, D., 1997. Control of intramammary infections in goats—impact on somatic-cell counts. J. Anim. Sci. 75, 566–570. Rota, A.M., Gonzalo, C., Rodriguez, P.L., Rojas, A.I., Martin, L., Tovar, J.J., 1993. Somatic cell types in goats milk in relation to total cell count, stage and number of lactation. Small Rumin. Res. 12, 89–98. Ryan, D.P., Greenwood, P.L., Nicholls, P.J., 1993. Effect of caprine arthritis–encephalitis virus-infection on milk cell count and N-acetyl-beta-glucosaminidase activity in dairy goats. J. Dairy Res. 60, 299–306. Sanchez, A., Contreras, A., Corrales, J.C., 1999. Parity as a risk factor for caprine subclinical intramammary infection. Small Rumin. Res. 31, 197–201. Thrusfield, M., 1986. Veterinary Epidemiology. Butterworths, London.
Prevalence and incidence of subclinical mastitis in goats and dairy ewes in Vermont, USA Scott McDougall a,∗ , Woody Pankey b , Carol Delaney b , John Barlow b , Patricia A. Murdough b , Dan Scruton c b
a Animal Health Centre, P.O. Box 21, Morrinsville, New Zealand Department of Animal Sciences, University of Vermont, Burlington, VT, USA c Vermont Department of Agriculture, Montpelier, VT, USA
Accepted 28 July 2002
Abstract The prevalence of bacterial isolation and the somatic cell count (SCC) of the milk from goats (n = 110 from six herds) and sheep (n = 153 from three herds) was determined at parturition and approximately 40 days later. Incidence of new intramammary infection (IMI) and the spontaneous cure rate were determined over this time period. In the goats, 27.3 and 25.5% were infected at parturition and 40 days later, respectively, while 15.0 and 9.1% of sheep were infected at parturition and 40 days later, respectively. Incidence of new infection was 0.039 and 0.034 cases/half/30 days for goats and sheep, respectively, and did not differ between the species (P > 0.2). Spontaneous cure occurred in 93.8% of sheep halves infected at parturition but only 50.0% of goat halves (P < 0.05). Coagulase negative staphylococci (CNS) were the most common isolates from both sheep and goats. Milk from bacteriologically positive halves had a significantly higher somatic cell count than halves from which no bacteria were isolated in both sheep and goats. It was concluded that bacterial infection of the mammary gland was associated with an elevated somatic cell count. Additionally, differences were detected between the species in incidence of new infection and spontaneous cure rate which resulted in a decline in prevalence in sheep, but not goats, with time postpartum. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Mastitis; Sheep; Goats; Prevalence; Incidence
1. Introduction Control of mastitis and somatic cell counts (SCC) in dairy animals is aided by an understanding of the pathogens involved, the rate of new intramammary infection (IMI), the spontaneous cure rate and the source of infection. Defining periods of high risk of new IMI ∗ Corresponding author. Tel.: +64-7889-5159; fax: +64-7889-3681. E-mail address: [email protected] (S. McDougall).
will aid in development of management practices that may reduce new IMI. Coagulase negative staphylococci (CNS) are the most prevalent pathogens of the mammary gland of sheep and goats (Poutrel and Lerondelle, 1983; Maisi et al., 1987; de la Cruz et al., 1994; Fthenakis, 1994). However, the dynamics of CNS IMI of goats and sheep are not well documented. Many studies have defined prevalence of individual pathogens at one point of lactation. Prevalence of IMI is dependant on the rate of new IMI (i.e. the incidence of infection; Thrusfield,
0921-4488/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 1 - 4 4 8 8 ( 0 2 ) 0 0 1 9 1 - 8
116
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
1986) and the duration of existing infections, which is affected by spontaneous cure rate and therapy (Lam, 1996). However, few studies have quantified the incidence rate of new IMI in small ruminants. The aim of the present study was to determine the prevalence of infection at parturition and 40 days later and to quantify the incidence and spontaneous cure rate over this period.
2. Materials and methods Goats (n = 110) and sheep (n = 155) from six and third herds and flocks, respectively, in Vermont, USA that underwent parturition between February and April 1999 were enrolled in the study. All animals were housed over winter. The sheep flocks lambed seasonally with all lambings occurring between March and May. Lambs were left on the ewes for approximately 4 weeks after lambing until weaning in April or May following which the ewes were moved to pasture. The ewes were then milked twice daily. All goat herds had multiple kidding periods, with some goats in milk at all times. The goat herds remained housed throughout the study. The kids were removed within 24 h of kidding and does were milked twice daily. Housed animals were fed hay and concentrates and water was available at all times, while the ewes were pasture grazed once weaning had occurred. All sheep flocks and five of the six goat herds machine-milked the animals while the goats in one herd were hand milked. Herdowners were given sterile tubes, alcohol and swabs and instructed in collection of milk samples aseptically and asked to collect milk from both halves within 24 h of kidding or lambing. The teat-end was prepared for milking by the herdowner following the normal herd protocol. The teat-end was then scrubbed with a cotton wool pledget wetted with 70% alcohol. After the teat-end had dried, the first strip of milk was discarded and then approximately 10 ml of milk was expressed into a sterile test tube, the tube was capped and labelled with herd, animal number or name and half. The herdowner-collected samples were frozen at −20 ◦ C for up to 1 M before culture. Between 11 and 50 animals from each herd were sampled. Herdowners were also asked to record and sample any animal which was diagnosed with clinical mastitis. On 1 day, approximately 6 weeks (average = 40.5 ± 8.1
days, mean ± S.D.) after parturition, a milk sample was drawn from each half of each enrolled animal for bacterial culture and somatic cell count determination by the investigators. These samples were kept on ice and transported to the laboratory for culture. 2.1. Milk sampling and bacteriological procedures The samples were thawed (if required), mixed and 0.01 ml was streaked onto a half plate of 5% trypticase washed-sheep-blood agar (TBA; Micro-Diagnostics Inc., Lombard, IL). Plates were incubated at 37 ◦ C for 48 h. The colonies were provisionally speciated as Staphylococcus aureus, CNS, Streptococcus agalactiae, other streptococci, Coliforms or Corynebacterium bovis on morphology, haemolysis pattern and Gram stain. The numbers of each distinct colony type were recorded as <6, >5 and <11, >10 and <51, and >50 colony forming units (CFU) per plate. Where >5 CFU of multiple colony types were found, representative colonies were subcultured on TBA. Gram-positive cocci were tested for catalase and coagulase production and Gram-negative isolates were further differentiated using MIO deeps (motility, indole, ornithine decarboxylase; Difco Laboratories, Detroit, MI). An IMI was diagnosed when >500 CFU/ml of each of 1 to 3 colony types were isolated. This figure was the normal definition of IMI within the laboratory. A new IMI was diagnosed when >500 CFU/ml of 1 to 3 colony types were isolated from a previously uninfected gland. Cure was defined as having occurred when no IMI was detected in a previously infected gland. Samples from which >2 colony types were isolated or, where <5 colonies of any pathogen were isolated were regarded as contaminated or uninfected, respectively. Prevalence was defined as the number of halves (or animals) from which a pathogen was isolated divided by the total number of halves (or animals) sampled at each time period. The incidence of new IMI was calculated as the total number of new IMI divided by the total number of days between samplings for halves with no infection at the initial sampling. The numerator was multiplied by 30 to express the incidence rate as the number of glands newly infected over a 30-day period. Following sampling for bacteriology, the remaining milk was preserved with one drop of 2-bromo-2nitropropane-1,3-diol (Preservo Liquid, D&F Control
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
Systems, San Francisco, CA) and incubated for >16 h at 4 ◦ C. Somatic cell count were determined using 500 l of the preserved milk (Fossomatic-90 A/S N. Foss Electric, Denmark). The effect of age, herd, species, day of year at parturition (Julian date) and days postpartum at second sampling on the prevalence at parturition and 40 days later was examined by logistic regression. Forward stepwise analysis was used with log likelihood as the selection methodology. The χ 2 statistics were used to compare proportions between species. The somatic cell counts were not normally distributed and were log10 transformed before analysis. The somatic cell counts were analysed by ANOVA with animal species, bacterial status and the species by status interaction as factors in the model. The change in somatic cell count from parturition to 40 days postpartum was analysed by subtracting the two figures and the difference was analysed by ANOVA.
3. Results Seven animals (six sheep and one goat) were diagnosed with unilateral clinical mastitis by the herdowners. Samples were submitted for four of these halves and S. aureus was cultured from two samples, Staphylococcus pulveri/vitulus from one sample and no isolation occurred from the fourth sample. One sheep died and a further two were culled before the end of the trial.
117
More goats than sheep were infected at parturition and at 40 days postpartum (P < 0.05; Table 1). Prevalence of infected goats and halves did not change between parturition and 40 days postpartum (P = 0.3; Table 1). The number of sheep with infection tended to decline with time postpartum (P = 0.12; Table 1) and the number of halves of sheep infected declined significantly postpartum (P < 0.01; Table 1). The number of infected halves increased with age group in goats (6.8, 19.4 and 31.6% of halves from 1 + 2, 3 + 4 and >4-year-old animals; P < 0.05), but not sheep (4.7, 4.3 and 0% of halves from 1 + 2, 3 + 4 and >4-year-old animals, respectively; P > 0.4). The proportion of infected animals at 40 days postpartum was affected by age and species (P < 0.05), but not by parturition date or days postpartum. A total of 14 new IMI in 14 separate animals were detected, seven each in sheep and goats. The incidence rate did not differ between sheep and goats (0.039 and 0.034 cases/half/30 days for goats and sheep, respectively, P > 0.2) and averaged 0.037 new cases/half/30 days. Animals in which one-half was infected at parturition were more likely to have a new IMI than animals which did not have an IMI in either half at parturition (5/15 versus 7/106; odds ratio = 7.6; P < 0.01). Spontaneous cure occurred in 26 of the 38 (68.4%) halves infected at parturition. Goats had a lower spontaneous cure rate than sheep (11/22 (50.0%) versus 15/16 (93.8%), respectively; odds ratio = 0.06; P < 0.05).
Table 1 The percentage of animals with zero-, one- or two-halves with infections at parturition and 40 days later, the number of new infections and the incidence rate of new infection Goats
Number of animals Neither half infected (%) One-half infected (%) Two-halves infected (%) Animals infected (%) Halves infected (%) Contaminated sample or minor infectiona (%) New infection at day 40 (number of glands) New infection at day 40 (% of glands) Incidence rate of new infection (new infection/gland/30 days) a
Sheep
Day 0
Day 40
Day 0
Day 40
110 50.0 19.1 8.2 27.3 35.5 22.7
110 66.4 15.5 10.0 25.5 35.5 9.1 7 6.4 0.039
153 52.9 11.1 3.9 15.0 19.0 32.0
132 69.7 9.1 0.0 9.1 9.1 21.2 7 5.3 0.034
Contaminated sample (>2 colony types) or minor infection (i.e. <500 CFU/ml).
118
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
Table 2 Bacterial species isolated from halves of goats and sheep at parturition and 40 days later Goats
Sheep
Day 0 n Coliforms Corynebacterium spp. Enterococcus faecalis Lactococcus garvieae Micrococcus luteus Pasteurella spp. Pseudomonas aeruginosa Staphylococcus aureus CNS Stomatococcus mucilaginosus Streptococcus gallolyticus Other streptococci Total
Day 40 Percentage
2
5.0
1
2.5
2 35
5.0 87.5
40
n
Day 0 Percentage
3 1
7.7 2.6
1 1 1 4 26
2.6 2.6 2.6 10.3 66.7
1 1
2.6 2.6
n
4 8
39
Day 40 Percentage
33.3 66.7
12
n
Percentage 1
3.6
1 1
3.6 3.6
2 22 1
7.1 78.6 3.6
28
Table 3 Geometric mean somatic cell count (mean × 103 cells/ml) and 95% confidence intervals for foremilk samples from halves of goats and sheep at parturition and 40 days later categorised as having no IMI or having an IMI (i.e. >500 CFU/ml of milk) Species
Goats Goats Sheep Sheep
Status
No growth Infected No growth Infected
n
146 40 210 28
Day 0
Day 40
Mean
95% CI Low
High
1,490 2,584 382 2,263
1,199 1,590 342 1,199
1,851 4,200 427 4,267
Coagulase negative staphylococci were the most common isolates from both goats and sheep and at parturition and 40 days later (Table 2). Sheep had a lower somatic cell count at parturition than goats (P < 0.005; Table 3) and infected halves had a significantly higher somatic cell count than halves with no IMI (P < 0.005). However, there was a significant status (infected, not infected) by species interaction with a lower somatic cell count in non-infected halves in sheep than goats (Table 3). Sheep had a higher somatic cell count at 40 days postpartum than goats (P < 0.005; Table 3). Infected halves had a higher somatic cell count than halves from which no pathogen was isolated at 40 days (P < 0.005; Table 3). There was no species by status interaction (P = 0.3).
Mean
211 1,292 467 4,290
95% CI Low
High
179 790 402 1,713
248 2,115 542 10,747
4. Discussion The prevalence of halves with IMI was 35.5 and 19.0% in goats and sheep at parturition which is similar to figures reported previously (Fthenakis, 1994; de la Cruz et al., 1994; Contreras et al., 1995). Prevalence declined in sheep, but not goats, by 40 days postpartum. A decline in prevalence with time postpartum was reported in ewes (Fthenakis, 1994). Differences in management between the ewes and does included time of weaning, seasonality of parturition and differences among herdowners in management practices. All goats in the present trial were machine or hand milked from within 24 h after parturition, whereas the lambs were left on the ewes until about 30–35 days postpartum. The sheep herds were
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
all seasonal lambing and all used intensive pasture grazing for more than 50% of the year. The goats were all housed for the majority of the year and kidded year round. The housing of the goats may have been more heavily contaminated than that of the sheep due to the continual use of the facilities. Additionally, as there were does lactating at all times of the year, the potential for ongoing transmission of infection from late to early lactation goats during milking or in the barns was always present. In contrast, all sheep were non-lactating at the same time for some months, which may have reduced the potential for ewe to ewe transmission of pathogens within the milking parlour. Prevalence of IMI increased with age in goats in agreement with other studies (Sanchez et al., 1999). Increasing prevalence with age may be due to increased length of exposure to pathogens in older compared to younger animals. Additionally, where the duration of infection is long and the spontaneous cure rate low, prevalence will increase. Prevalence in ewes did not increase with age as has been previously reported (Fthenakis, 1994). The high self-cure rate found in the ewes in the present trial may be the explanation for this. Incidence of new bacterial IMI was similar for goats and sheep in the present study and averaged 0.037 cases/half/30 days. This is similar to the incidence rate of 1%/ewe/week for ewes reported by Hueston et al. (1986). Incidence of new IMI does not appear to have been previously reported for dairy goats. Incidence of IMI for CNS in housed dairy cattle is reported as 1.58 cases/1000 quarter days or about 0.5 cases/quarter/year (Lam, 1996). The goats and sheep appear to have a higher incidence rate than that reported for cattle. However, incidence in the present study was calculated only over the first 40 days of lactation. Incidence of new clinical infections was shown to be higher in early lactation in cattle (Hogan et al., 1989; Lam, 1996). Hueston et al. (1986) noted an increase in incidence rate of new IMI following weaning from 1%/ewe/week pre-weaning to 9.7%/ewe/week post-weaning. In that study ewes were not machine-milked following weaning, so the increased incidence of IMI was associated with involution of the mammary gland. In the present study, the majority (15 of 16) sheep with IMI at lambing did not have an IMI at 40 days
119
postpartum. However, half the goat halves with an IMI were still infected at 40 days postpartum. Spontaneous cure in sheep was reported as 9.5 and 21% between 2 and 8 weeks post-weaning and lambing to 6 weeks postpartum (Fthenakis, 1994; Kirk et al., 1996). The high frequency of suckling and the large litter size (average of >2 lambs/ewe) may have resulted in the ewes being frequently milked dry which may have resulted in the high self-cure rate observed. As reported previously (Maisi et al., 1987; de la Cruz et al., 1994; Fthenakis, 1994; Contreras et al., 1995) CNS were the most common pathogens isolated from the goat and sheep milk. CNS are common isolates from the respiratory tract, the teat skin, the teat-end as well as from milk (Devriese et al., 1985). The significance of CNS isolates from milk needs to be examined further. However, heavy pure cultures of these CNS isolates did occur in the present study and from other studies of goat and sheep milk (Fthenakis, 1994; Contreras et al., 1995; Kirk et al., 1996; Burriel, 1997). CNS isolations have been associated with elevated somatic cell count and increases in concentrations of NAGase, albumin and salt which indicate that mammary gland tissue damage has occurred (Maisi et al., 1987). Repeated isolations of the same CNS species from the same glands have occurred for up to 9 M indicating chronic IMI with CNS occur (Maisi and Riipinen, 1991; Contreras et al., 1997). CNS have also been recovered following fine needle aspirate of the teat cistern, indicating that the CNS are established beyond the teat canal (Maisi and Riipinen, 1991). Experimental infection with S. simulans and S. warnieri resulted in repeated isolations of these pathogens from milk, changes in milk composition and histological changes within the infected glands of sheep (Fthenakis and Jones, 1990; Burriel, 1997). These data support the contention that at least some strains of CNS are pathogenic to the caprine and ovine udder. The epidemiology of CNS IMI in cattle has been mathematically modelled (Lam, 1996). Control measures such as teat spraying and treatment of existing IMI have been successful in controlling contagious pathogens (Bramley and Dodd, 1984). However, despite dry goat therapy being effective in reducing somatic cell count, teat spraying was not successful in one trial (Poutrel et al., 1997). However, as there are differences among active constituents and formulations among teat sprays in efficacy (Pankey et al., 1994) more work needs to
120
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121
be done to test the effect of teat spraying on CNS in goats and sheep. Infection significantly increased somatic cell count both at parturition and 40 days later in both in sheep and goats. The somatic cell count in goats declined with time postpartum. A similar decline in somatic cell count over the first few days of lactation was observed in dairy cattle (Harmon, 1994; Lacy-Hulbert and Woolford, 1997). The somatic cell count of infected glands was 6 to 10 times higher than that of uninfected glands 40 days postpartum. However at kidding, the infected goats only had a somatic cell count about twice that of the uninfected glands. This indicates that differentiation of uninfected from infected glands may be easier to achieve some time postpartum rather than immediately after parturition in goats. The somatic cell count was reported to either decline or to increase over the first few weeks of lactation in goats (Lerondelle et al., 1992; Rota et al., 1993) and IMI are associated with an increased somatic cell count (Lerondelle et al., 1992). Hence differences among studies and herds in somatic cell count following parturition may be due to variation in the initial prevalence and incidence of IMI between herds. Uninfected goat halves are reported as having somatic cell count of 0.05 and 0.5 × 106 somatic cell count/ml (Lerondelle et al., 1992; Paape and Capuco, 1997). In contrast, infected goat glands had somatic cell count of 0.3 to 15.1 × 106 cells/ml with variation among bacterial species in the degree of somatic cell count elevation (Dulin et al., 1982; Ryan et al., 1993). The somatic cell count in the present study agree with the previously published data for goats. In sheep, the somatic cell count from uninfected halves is reported as ranging from 0.26 to 1.58 × 106 somatic cell count/ml (Fthenakis, 1994; Burriel, 1997). The basal somatic cell count is affected by stage of lactation and the volume of milk produced (Mavrogenis et al., 1995). Infection increased somatic cell count in sheep. All infected glands had a somatic cell count of >2.0 × 106 somatic cell count/ml (Mavrogenis et al., 1995) and an average of between 3.3 and 3.8 × 106 somatic cell count/ml (Fthenakis, 1994; Burriel, 1997). It is concluded that the dynamics of IMI and somatic cell count differ between sheep and goats. The underlying physiological and/or managerial reasons for these differences require further investigation.
Additionally, studies addressing changes in IMI prevalence and incidence for an entire lactation and non-lactation period need to be undertaken to define periods of high risk of new IMI in these species.
Acknowledgements The help of the herdowners in sampling animals at parturition and allowing the subsequent sample collection in gratefully acknowledged. Bacteriology support was provided by Dina Bawdry, Erin Mulloy and Jeff Watts. Financial support was obtained from the Royal Society of New Zealand in the form of a NZ/USA Co-Operative Science Program, ISAT linkage fund (contract number 99-CSP-42MCDO) and an Animal Science Award from the New Zealand Society of Animal Production, both granted to Scott McDougall. References Bramley, A.J., Dodd, F.H., 1984. Reviews of the progress of dairy science: mastitis control—progress and prospects. J. Dairy Res. 51, 481–512. Burriel, A.R., 1997. Dynamics of intramammary infection in the sheep caused by coagulase-negative staphylococci and its influence on udder tissue and milk-composition. Vet. Rec. 140, 419–423. Contreras, A., Corrales, J.C., Sierra, D., Marco, J., 1995. Prevalence and etiology of nonclinical intramammary infection in Murciano–Granadina goats. Small Rumin. Res. 17, 71–78. Contreras, A., Corrales, J.C., Sanchez, A., Sierra, D., 1997. Persistence of subclinical intramammary pathogens in goats throughout lactation. J. Dairy Sci. 80, 2815–2819. de la Cruz, M., Serrano, E., Montoro, V., Marco, J., Romeo, M., Baselga, R., Albizu, I., Amorena, B., 1994. Etiology and prevalence of subclinical mastitis in the Manchega sheep at mid-late lactation. Small Rumin. Res. 14, 175–180. Devriese, L.A., Schleifer, K.H., Adegoke, G.O., 1985. Identification of coagulase-negative staphylococci from farm animals. J. Appl. Bact. 58, 45–55. Dulin, A.M., Paape, M.J., Wergin, W.P., 1982. Differentiation and enumeration of somatic cells in goat milk. J. Food Protection 45, 435–439. Fthenakis, G.C., 1994. Prevalence and etiology of subclinical mastitis in ewes of southern Greece. Small Rumin. Res. 13, 293–300. Fthenakis, G.C., Jones, J.E.T., 1990. The effect of experimentally induced subclinical mastitis on milk yield of ewes and on the growth of lambs. Br. Vet. J. 146, 43–49. Harmon, R.J., 1994. Physiology of mastitis and factors affecting somatic cell counts. J. Dairy Sci. 72, 2103–2112.
S. McDougall et al. / Small Ruminant Research 46 (2002) 115–121 Hogan, J.S., Smith, K.L., Hoblet, K.H., Schoenberger, P.S., Todhunter, D.A., Hueston, W.D., Pritchard, D.E., Bowman, G.L., Heider, L.E., Brockett, B.L., Conrad, H.R., 1989. Field survey of clinical mastitis in low somatic cell count herds. J. Dairy Sci. 72, 1547–1556. Hueston, W.D., Hartwig, N.R., Judy, J.K., 1986. Patterns of nonclinical intramammary infection in a ewe flock. J. Am. Vet. Med. Assoc. 188, 170–172. Kirk, J.H., Glenn, J.S., Maas, J.P., 1996. Mastitis in a flock of milking sheep. Small Rumin. Res. 22, 187–191. Lacy-Hulbert, J., Woolford, M., 1997. Mastitis diagnosis and resolution. In: Hayes, D.P., McDougall, S. (Eds.), Proceedings of the Second International Conference for the Society of Dairy Cattle Veterinarians of the New Zealand Veterinary Association. Vanuatu, Noumea, pp. 24–36. Lam, T.J.G.M., 1996. Dynamics of Bovine Mastitis. Ph.D. thesis, Utrecht. Lerondelle, C., Richard, Y., Issartial, J., 1992. Factors affecting somatic cell counts in goat milk. Small Rumin. Res. 8, 129–139. Maisi, P., Riipinen, I., 1991. Pathogenicity of different species of staphylococci in caprine udder. Br. Vet. J. 147, 126–132. Maisi, P., Juntilla, J., Seppanen, J., 1987. Detection of subclinical mastitis in ewes. Br. Vet. J. 143, 402–409. Mavrogenis, A.P., Koumas, A., Kakoyiannis, C.K., Taliotis, C.H., 1995. Use of somatic-cell counts for the detection of subclinical mastitis in sheep. Small Rumin. Res. 17, 79–84.
121
Paape, M.J., Capuco, A.V., 1997. Cellular defence-mechanisms in the udder and lactation of goats. J. Anim. Sci. 75, 556–565. Pankey, J.W., Eberhart, R.J., Cumming, R.D., Daggett, R.D., Farnsworth, R.J., McDuff, C.K., 1994. Update on postmilking teat antisepsis. J. Dairy Sci. 67, 1336–1353. Poutrel, B., Lerondelle, C., 1983. Cell content of goat milk: California mastitis test, coulter counter and fossomatic for predicting half infection. J. Dairy Sci. 66, 2575–2579. Poutrel, B., Decremoux, R., Ducelliez, M., Verneau, D., 1997. Control of intramammary infections in goats—impact on somatic-cell counts. J. Anim. Sci. 75, 566–570. Rota, A.M., Gonzalo, C., Rodriguez, P.L., Rojas, A.I., Martin, L., Tovar, J.J., 1993. Somatic cell types in goats milk in relation to total cell count, stage and number of lactation. Small Rumin. Res. 12, 89–98. Ryan, D.P., Greenwood, P.L., Nicholls, P.J., 1993. Effect of caprine arthritis–encephalitis virus-infection on milk cell count and N-acetyl-beta-glucosaminidase activity in dairy goats. J. Dairy Res. 60, 299–306. Sanchez, A., Contreras, A., Corrales, J.C., 1999. Parity as a risk factor for caprine subclinical intramammary infection. Small Rumin. Res. 31, 197–201. Thrusfield, M., 1986. Veterinary Epidemiology. Butterworths, London.