Leptospirosis in dairy herds : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy at Massey University, School of Veterinary Science, Massey University, Palmerston North, New Zealand

The introduction and adoption of Leptospira vaccination in most New Zealand dairy herds in the 1980s was associated with a substantial reduction of the incidence of notified human leptospirosis cases in the population and notably among dairy farm workers. However, 80% of cases notified from 1999 to 2016 with a “farmer-type” occupation were dairy farmers, hence this occupational group continues to be at risk for leptospirosis. Failure to vaccinate dairy herds can have serious public health consequences. An example is described in an opportunistic case study, chapter 3 of this thesis. Within the space of three months in early 2015, three workers from a dairy farm with an unvaccinated dairy herd were hospitalised with leptospirosis caused by serovars Hardjo and Pomona. In young and adult dairy cattle from this farm, Hardjo, Pomona, Copenhageni, Ballum and Tarassovi serovars were all detected serologically. While two of the diseased workers recovered, one remains affected four years after the event being unable to return to work. These circumstances – the continuation of dairy workers among notified human cases and the potentially serious consequences from failure to vaccinate or to achieve effective immunity – have raised concerns about the effectiveness of the long–term vaccination programme in dairy herds. The concerns were further substantiated by an opportunistic pilot study (2011) that found evidence of Leptospira shedding in vaccinated dairy cattle. Therefore, a large cross-sectional study of New Zealand dairy farms was conducted involving 200 dairy farms and 4,000 cows. Farms were randomly selected from the national database and blood and urine was collected from 20 cows per herd. Non- response (30%) was investigated by personal interview which indicated that selection bias (e.g. by selecting only farmers with vaccinated herds) was minimal, if not absent. Shedding was indicated by a positive qPCR at cow-level and by one or more shedders per herd at herd-level. A serological response was considered positive when titres of the microscopic agglutination test (MAT) were at or above 48. Overall shedding rates were 2.4% at cow- and 26.5% at herd-level. Seropositivity to Hardjo, Pomona and, when trivalent vaccines were used, Copenhageni, was most likely a response to vaccination. None of the vaccinal serovars were associated with urine shedding. However, there was a strong linear association at the cow-level between increasing MAT titres to Tarassovi and the likelihood of shedding. Serological evidence for exposure to Tarassovi was observed in 17% of cows and 74% of the herds. Few cows (1%) and 16% herds were sero-positive to Copenhageni when not vaccinated against this serovar which, however, was not associated with cow-level shedding. Similarly, the rodent-related serovar Ballum was not associated with shedding; with positive titres observed in 3% of cows and 38% of herds. Studies in the 1970s and 1980s found little serological evidence of Tarassovi, so we conclude that this serovar has emerged, became endemic and is now probably causing most of the shedding in the dairy cattle population. Considering published evidence that a large proportion of notified cases in dairy farmers were Tarassovi, there is strong corroborative evidence that this serovar poses a public health risk for workers on dairy farms. Our survey administered a questionnaire about vaccination practices and putative risk factors. All but one of the farmers had regular vaccination programmes for calves, heifers and cows using mostly bivalent (80%, 69%, 68% of farms, respectively) and some trivalent vaccines (20%, 31%, 32% of farms, respectively). Regardless of the almost universal practice of Leptospira vaccination in dairy cattle, fewer than 40% famers conformed with Best Practice Guidelines (2012) developed and propagated by the New Zealand Veterinary Association. A further objective was a risk factor analysis (Chapter 6). One cow-level (age) and three herd-level (presence of sheep or dogs, herd size) factors were significantly associated with the risk of shedding. As 93% of the potential factors evaluated were at herd level, and with only 200 herds included in the study, and the shedding rate being relatively low, the statistical power might have been too low to identify other herd-level determinants related to the management and environment of the farms. Nevertheless, a linear negative effect of age suggested that young cows were more likely to shed Leptospira than adult cows, and therefore increase the risk of infection for dairy workers. Evidence from this thesis suggests that current Leptospira vaccination practices are effective for preventing the exposure of farm workers against the serovars most commonly incorporated in vaccines (Hardjo and Pomona), and the less common serovar Copenhageni. Thus, continuation with vaccination is supported. The public health risk arising from Tarassovi that has emerged, and evidence here that this serovar is widely present in the dairy cattle population, justifies raised awareness, the adoption of protection measures additional to vaccination, further research into the epidemiology of Tarassovi and an evaluation of the justification for its inclusion into vaccines. Dairy workers are advised to take extra care and precautions when milking and handling cows, especially first calving heifers irrespective of their vaccination status