The thesis combines the field studies on the epidemiology of bovine pestivirus infection and the experimental study on pathogenesis of transplacental bovine pestivirus infection. In chapter four, the horizontal route of pestivirus transmission in groups of young, recently purchased calves was examined. A total of 52 batches of calves from 21 dairy farms was submitted to the Queensland Department of Primary Industries Tick Fever Research Centre (TFRC) between 1990 and 1996. The average age of the calves was 2.8 weeks (0.1-9.0) and the average number per batch was 27 (11-42). The serum virus neutralization test (VNT) and antigen-capture enzyme-link immunosorbent assay (ELISA) were used to determine antibody titres to pestivirus and detect pestivirus antigen, respectively. The VNT and ELISA were performed on blood samples collected at the time of arrival at TFRC. VNT was subsequently performed on samples collected on average 42 days (19-67) after arrival at TFRC. ELISA positive calves were retested on average 25 days (9-70) after initial testing. A total of 14 calves (0.98%) from 13 batches were identified as being persistently infected (PI). In PI calf batches, the seroconversion rate was 58.4%, significantly (p<0.01) different from the seroconversion rate of 1.7% in the non-PI batches (i.e., batches without a PI calf). Nineteen percent of calves with an initial elevated serum VNT titre (≥ 1:16) exhibited prolonged maintenance of serum titre. Data on calf serology was correlated with the pestivirus status of the herd of origin. Eighty-one percent of the calves originating from herds containing a PI animal(s) were seropositive on arrival at TFRC, significantly different (p<0.01) from the 29.4% seropositivity in calves from non-PI herds. It is concluded that: 1. The presence of a PI animal(s) in a group of calves results in significant horizontal transmission. 2. Serologic results in young calves could be used to define herd infection status. Chapter five investigates the epidemiology of pestivirus infection in seven dairy herds. Whole herd (i.e., animals aged ≥ 3 months) blood sampling was conducted. A total of 1,661 cattle were bled between June and July 1997. VNT was performed on all serum samples and antigen-capture ELISA was performed on selected blood clot samples. Average herd seroprevalence was 66% (39-94%) in the study population. The estimated annual incidence of infection was 32%. Herds with PI animals contained a significantly higher proportion of seropositive animals in all age groups (p < 0.01). Four animals (0.24%) from three herds were PI. The pattern of age-specific seroprevalence was similar in both PI and non-PI herds, i.e., antibody prevalence increased with age. Between PI and non-PI herds the percentage of seropositive animals was most significantly different in animals aged < 1 year old. Three major herd categories, i.e., susceptible, partially susceptible and immune herds, could be identified. To investigate the spatial changes in herd infection status, data from the whole herd blood sampling in 1997 were collated with the herd serologic histories from 1990 to 1996. In 1998, a further blood sampling was conducted in selected groups of animals, primarily seronegative hejfers from the 1997 blood sampling and calves aged 5-8 months old. Over the period of investigation from 1990-1996, twenty-two cattle (1.3%) from 6 herds were confirmed PI. When combined with the data obtained from 1997 and 1998, a total of 34 cattle (annual prevalence of 0.24%) were confirmed PI. The herd seroprevalence changed over time in all herds studied. The prevalence of PI animals was the main factor affecting the herd infection status. It is concluded that: 1. Pestivirus infection is likely to be endemic in the dairy cattle population on the eastern Darling Downs, Queensland. 2. Management strategies need to be improved in order to control bovine pestivirus infection in dairy herds. Chapter six determined the economic losses associated with transplacental pestivirus infection. A simple mathematical model was used, and the estimated losses were validated by comparing with actual data from a field outbreak of pestivirus infection. In the model, the losses were assumed to occur from non-cytopathic pestivirus only. The losses during four periods of pregnancy, i.e., 14 days before the insemination, 1 to 24 days, 25 to 100 and 101 to 180 days of gestation, were estimated. In a model herd experiencing a pestivirus epizootic, the model estimated losses of A$ 61 per cow with the birth of 13 PI calves per 100 susceptible cows. The losses occurred over a 21 months period. In the field epizootic, a herd of 65 milking cows and 24 replacement heifers was examined. The losses were estimated to be A$ 94 per cow with 13 PI calves born from 94 calvings between 1995 and 1996. It is concluded that the model used could be useful in predicting the economic losses associated with bovine pestivirus infection. In chapter seven, the mechanism of transplacental infection in early pregnant ewes was investigated. Fourteen seronegative merino ewes were treated to synchronize oestrus, then artificially inseminated and subsequently were confirmed pregnant by demonstration of significantly elevated peripheral progesterone concentration on day 16 post AI. On the 18th day of pregnancy, nine ewes were inoculated with lxl05TCID5o of a noncytopathic strain of bovine pestivirus into each nostril. Five ewes served as non­ infected controls. All ewes were ovariohysterectomized on average I00 hours (75-107 hours) post inoculation. Samples of ovary, caruncle, intercaruncular tissue, foetal membrane, foetal fluid and foetus were collected and tested by reverse transcriptase­ polymerase chain reaction (RT-PCR), nested PCR and immunohistochemistry (IHC) for the presence of pestivirus. Nested PCR revealed pestivirus positivity in the exposed ewes as follows: in the ovary 7/9, caruncle 9/9, intercaruncular tissue 7/9, foetal membrane 6/9, foetal fluid 9/9 and foetus 2/9 ewes. By IHC, pestivirus antigen was demonstrated in the ovary 2/9, caruncle 6/9, intercaruncular tissue 5/9, foetal membrane 4/9, and foetus 3/9 ewes. The positive results from RT-PCR were weak and could only be unambiguously seen using nested PCR. The IHC positive cells were found only in small foci in the tissue sections. All exposed ewes seroconverted (confirmed by VNT) within 2-4 weeks of inoculation, while all control ewes remained seronegative. This study provides the first quantitative data on the dynamics of transplacental pestivirus infection in the ewe. It is concluded that bovine pestivirus rapidly (within 4 days) infects conceptuses following the intranasal exposure of the dams. The route(s) of transplacental pestivirus infection could not be identified since all major tissues of the reproductive tract and conceptuses were infected. However, the finding of pestivirus antigen in binucleate cells (BNCs) of the developing placenta suggests that migration of BNCs to the maternal endometrium may provide a 'bridge' for transmission to the foetuses. Chapter eight investigated the possible role of interferon-tau (IFNi) in prevention of intrauterine pestivirus infection. The RT-PCR was used to determined the presence of IFNt in similar tissue samples tested and described in chapter seven. In pregnant animals, IFNt mRNA was detected in 84.2% of caruncles, 81.2% of intercaruncular tissues, 75.0% of ovaries, 78.6% of allantoic fluid, 100% of foetal membranes and 72.7% of foetuses. The percentages were not significantly different between pestivirus-infected and control ewes and concepti. Samples from non-pregnant animals, both from pestivirus-infected and controls, were negative. The detection of IFNt in ovarian tissue has not been reported elsewhere. It is concluded that although there was evidence of IFNt in the majority of tissues of the reproductive tract, the demonstration of concurrent pestivirus infection suggests that IFNt plays little or no role in preventing pestivirus infection.