Cattle Practice 6.2

Cattle Practice APRIL 1998 Volume 6 Part 2

CONTENTS Bovine Tuberculosis: Unresolved Questions and Future Approaches to Control 75 W I Morrison Recent Advances in DNA Fingerprinting using Spoligotyping - Epidemiological 79 Applications in Bovine TB R S Clifton-Hadley Badgers and Bovine Tuberculosis: A Review of Studies in the Ecology of a Wildlife 83 Disease Reservoir. R J Delahay, Animal Welfare: Ethics, Economics and Productivity 89 J McInerney Coliform Mastitis; An Evolving Problem? 91 M Green A Prospective Investigation of Intramammary Infections due to Enterobacteriacae 95 during the Dry Period: A Presentation of Preliminary Findings. A J Bradley A Case Control Study of Toxic Mastitis in Dairy Cows 103 L E Green Robotic Milking at Bridgets 109 R Bull Should we use Antimicrobials for Treatment of Coliform Mastitis in Dairy Cows ? 113 N Y Shpigel Ultrasonographical Examination of the Mammary Gland in Cows with Induced 121 S.aureus Mastitis: A Criteria for Prognosis and Evaluation of Therapy. A Banting Dairy Cattle Welfare – The FAWC Perspective 125 R Macpherson Organic Dairy and Beef Cattle Farming 127 H Browning A Slow-Release Iodine, Selenium And Cobalt Cattle Bolus 129 P A M Rogers Making the Best Use of Bulk Milk Antibody Tests 133 G C Pritchard Testing of Individual Cow’s Milk Samples to Ensure Health and Optimum Productivity 139 M W Allen Use of Trans-Thoracic Fine Needle Aspiration in the Investigation of Chronic 145 Bovine Respiratory Disease B P R Sturgeon Ultrasonographic Examination of the Bovine Thorax 151 P R Scott

CATTLE PRACTICE VOL 6 PART 2 Bovine Tuberculosis: Unresolved Questions and Future Approaches to Control Morrison W.I., Institute for Animal Health, Compton, Berkshire. RG20 7NN ABSTRACT During the last decade there has been a significant increase in the incidence of cattle herds infected with Mycobacterium bovis in England and Wales. A large proportion of herd breakdowns occur in areas that have high badger populations which are infected with M. bovis. While available evidence suggests that the badgers in these areas are a significant source of infection for cattle, conclusive data to support this view are still lacking. The Krebs review has recommended a series of epidemiological studies, together with a large field experiment involving different badger culling strategies, to clarify the role of the badger in bovine tuberculosis and to determine the impact and cost effectiveness of badger control. As a longer term approach to the problem, the review has recommended the establishment of a research programme to develop a vaccine for M. bovis, aimed primarily at vaccination of cattle. KEY WORDS: Cattle, badgers, M. bovis, tuberculosis. INTRODUCTION In the 1930's at least 40% of cows in the UK dairy herd were infected with Mycobacterium bovis. Transmission of infection to humans was common and was estimated to cause about 2000 deaths annually accounting for approximately 6% of total human deaths due to tuberculosis. The subsequent introduction of pasteurisation of milk coupled with tuberculin testing of cattle and compulsory slaughter of infected animals dramatically reduced the prevalence of infection in cattle and the incidence of human disease caused by M. bovis. However, despite a sustained herd testing programme since the 1950's, it has not proved possible to eradicate tuberculosis from the cattle population. In the early 1970's, infection with M. bovis was detected in a number of wildlife species, among which the badger was considered a likely source of infection for cattle. This was based on a high prevalence of infection in badgers in affected areas, the occurrence of diseased animals that excreted large numbers of bacteria and the opportunity for contamination of cattle pasture by badgers. During the last 20 years, a number of different strategies for culling badgers have been implemented with the aim of reducing the incidence of tuberculosis in cattle. These measures have had little impact on the overall incidence of disease and, indeed, there has been an increase in the number of herd breakdowns in England and Wales over the last 10 years. Confronted with this problem, in 1996 the Ministry of Agriculture Fisheries and Food set up a review, under the chairmanship of Professor John Krebs, to consider the role of the badger in bovine tuberculosis. This presentation discusses the main conclusions and recommendations of the review (Krebs et al. 1997). BOVINE TUBERCULOSIS - AN INCREASING THREAT Over the last 10 years there has been a progressive increase in the number of herds affected by tuberculosis in England and Wales. This increase has been particularly apparent in South West England, Gloucestershire, Herefordshire and South Wales, and more recently a focus of breakdowns has appeared in the West Midlands. During the 10 year period 1986- 1996 the proportion of total herds with reactor animals in the South West of England increased from 0.75% to 2.6%. This compares with an increase from 0.10% to 0.51% over the same period in the rest of England and Wales. In addition to the losses incurred by farmers as a result of herd breakdowns, this increase in incidence of tuberculosis represents a serious threat to livestock trade as well as posing a potential human health hazard. TUBERCULOSIS IN BADGERS Badgers appear to be highly susceptible to infection with M. bovis, and their social behaviour coupled with the chronic nature of infection in most animals are conducive to maintenance of infection in badger populations. In many animals, infection is clinically inapparent but gives rise to intermittent excretion of the organism, while a minority of cases develop severe disease characterised by lesions in a wide range of tissues and excretion of large numbers of bacteria in sputum and urine. A survey of badger mortality in Woodchester Park in Gloucestershire has indicated that tuberculosis accounts for only about 7% of deaths. Between 1972 and 1990 badgers killed in road traffic accidents were surveyed for the presence of infection BCVA 1998 75

CATTLE PRACTICE VOL 6 PART 2 with M. bovis. The annual prevalence of infection revealed by this survey in the 1980's ranged from 1.3% to 5%. In the same period, examination of carcasses obtained during the course of MAFF badger removal operations from areas with herd breakdowns revealed much higher levels of infection. While these data suggest a relationship between the prevalence of infection in badgers and herd breakdowns, the collection of the data has been highly biased towards those local areas with a high incidence of herd breakdowns and therefore does not permit firm conclusions to be made on the effect of prevalence of infection in badgers on the risk of herd breakdown. THE IMPACT OF BADGER CULLING STRATEGIES Four different badger culling strategies have been employed:- (i) Gassing of sets (1975-81) involved removal by gassing of all badger social groups impinging on breakdown farms. (ii)The Clean Ring Policy (1982-86) involved progresive trapping and removal of social groups in the affected farms and surrounding lands until a “ring” of uninfected social groups was identified. (iii) In the Interim Strategy (1986-97) badgers were removed by trpping from the breakdown areas of affected farms, if identifiable, or from the whole farm. (iv) The Live Test Trial (1994-96) was run in parallel with the Interim Strategy and involved trapping and removal of all badgers from setts with detectable infection on affected farms and on any neighbouring farms at risk from the same badgers. this trial was suspended in 1996 pending the outcome of the Krebs review. Trapping of badgers in cages, as used in the latter 3 strategies, does not capture all animals. These strategies also involved the release of lactating sows. In the periods following implementation of the Gassing and Clean Ring Strategies, there was a reduced rate of recurrence of herd breakdowns in the cleared areas. However, in none of the culling strategies were there comparable control areas. Therefore, it is difficult to draw conclusions on the effectiveness of the measures since other factors influencing risk of infection could have changed during the periods of monitoring. DIAGNOSIS OF TB IN CATTLE The tuberculin skin test is the primary means of diagnosing TB in cattle. Shortcomings in the test have been suggested as a contributory factor to the recurrence of herd breakdowns. However, it is clear that in many parts of the country the test has been used successfully to eradicate the disease. Moreover, the available data on sensitivity and specificity of the test, coupled with its application as a herd test, indicate that there is unlikely to be a significant residual population of infected cattle that escape detection. The number of false positive reactions is also exceedingly low in relation to the total number of animals that are tested. Therefore, although the development of alternative tests may result in more rapid and cheaper testing regimes, it is unlikely to have a major impact on the current disease problem. MAIN RECOMMENDATIONS OF THE REVIEW The Krebs review recognised the need to respond to the immediate disease problem as well as considering methods that might be developed for control of the disease in the longer term. The review concluded that, while the sum of evidence strongly supports the view that the badger is a significant source of infection for cattle, much of this evidence is correlative and does not demonstrate clearly a causal link between the badger and bovine tuberculosis. Moreover, the absence of controls in previously implemented badger culling operations has prohibited accurate measurement of their impact on the incidence of disease and their cost effectiveness. Three main recommendations have been made to address these problems:- (i) That a detailed epidemiological study be undertaken to evaluate the risk of herd breakdowns in defined areas in relation to the presence or absence of badgers. (ii) That recently developed molecular typing techniques for M. bovis be applied to field isolates to examine the spread of infection between badgers and cattle and within badger and cattle populations. (iii) That a large field experiment to evaluate the impact of badger culling on herd breakdowns be carried out. This is envisaged to involve a series of 10km2 sites, all at high risk of TB, allocated to one of 3 treatments, namely complete removal of badgers at the outset, removal of badger social groups from reactor farms in response to breakdowns and (as a control) no intervention. In the longer term, vaccination was recognised as the most direct means of protecting animals against infection with M. bovis. Although the development of highly effective vaccines against Mycobacteria has proved difficult in the past, recent advances in methods for genetic manipulation of these organisms and in vaccine technology in general, have created new opportunities for vaccine development. Vaccination could be targeted to either cattle or badgers. Because of practical difficulties both in the development and testing of a badger vaccine and in its eventual application, the establishment of a research programme aimed primarily at developing a BCVA 1998 76

CATTLE PRACTICE VOL 6 PART 2 vaccine for cattle was recommended. It was also recognised that the use of such a vaccine would require a diagnostic test capable of distinguishing infected from vaccinated animals. REFERENCES Krebs, J. R., Anderson, R. M., Clutton-Brock, T., Morrison, W. I., Young, D. and Donnely, C. 1997. Bovine Tuberculosis in Cattle and Badgers. MAFF Publications, London, pp 1-191. BCVA 1998 77

CATTLE PRACTICE VOL 6 PART 2 BCVA 1998 78

CATTLE PRACTICE VOL 6 PART 2 Recent Advances in DNA Fingerprinting using Spoligotyping - Epidemiological Applications in Bovine TB Clifton-Hadley R. S., Inwald J., Archer J., Hughes S., Palmer N., Sayers A. R., Sweeney K., Van Embden J. D. A. and Hewinson, R. G. Veterinary Laboratories Agency, Addlestone, Surrey, KT15 3NB. SUMMARY Spoligotyping was applied to 2668 Mycobacterium bovis isolates from cattle and other species in Great Britain. Thirty-four spoligotypes have been defined, 25 occurring in cattle. More than one type may exist in a herd at the same time. Certain spoligotypes predominate. Some types exhibit marked clustering in time and space. The great majority of linked cattle and badger isolates share at least one spoligotype. The method provides useful information for epidemiological investigations of sources of herd infection. KEYWORDS: DNA fingerprinting, spoligotyping, Mycobacterium bovis , bovine tuberculosis, PCR. INTRODUCTION Bovine tuberculosis (TB) in cattle is controlled in Great Britain by identifying infected herds through a combination of regular herd testing, using the comparative tuberculin skin test, and slaughterhouse meat inspection. Herds with suspected infection are then placed under movement restrictions until tested clear. Veterinary investigation of any confirmed case is designed to establish the most likely source of infection, such as purchased animals, contiguous premises, or a potential wildlife reservoir. Despite these investigations, the origin of infection in many cases remains undefined, especially outside areas known from past control to harbour infected badger populations (Report 1997). DNA fingerprinting techniques have been developed to distinguish different strains of Mycobacterium tuberculosis in man and used in epidemiological investigations, for example to distinguish reinfection from reactivation of a previous infection. These techniques are now being applied to M.bovis isolates. Molecular typing is recognised as a powerful means of investigating possible sources of infection and may provide ‘conclusive evidence on whether, and to what extent, badger to cattle transmission takes place’ (Krebs 1997). Molecular typing studies for M.bovis isolates have so far relied on three main methods, Restriction Endonuclease Analysis (REA), Restriction Fragment Length Polymorphism (RFLP) and Spoligotyping (Spacer-Oligonucleotide). REA is a method used exclusively in New Zealand and has resulted in a great degree of differentiation of M.bovis isolates. However, the slight differences in DNA patterns revealed by this method make interpretation more difficult than for the other techniques. Results are now available for isolates from various species, especially cattle, deer and possums (Collins et al 1994). Typing using RFLP initially focused on the insertion element IS6110, since M.tuberculosis was often found to have 25 or more copies of IS6110 in the chromosome, so allowing sufficient differentiation for epidemiological studies. However, M.bovis isolates from cattle often have just one copy. Better discrimination has been achieved using other probes, such as the GC-rich repeat sequence and the direct repeat sequence (Skuce et al 1994). There have been various studies using these techniques, for example the human health risks from cattle infection have been investigated in Argentina and the USA (Cave et al 1992; Romano et al 1996). However, RFLP requires a large amount of M.bovis to be cultured and is a labour-intensive process. Spoligotyping is a polymerase chain reaction-based (PCR) method that detects unique DNA sequences at the direct repeat locus of the M.bovis genome. The technique is relatively quick, simple and cheap to perform and the results are easily interpreted using computer analysis. However, it has been suggested that this method may lack sufficient sensitivity when used on its own (Cousins et al 1998) and a combination of techniques may result in the required degree of differentiation (Krebs 1997). In 1996, the Ministry of Agriculture, Fisheries and Food funded a project to develop and validate spoligotyping for M.bovis isolates and to assess whether it could contribute to bovine TB control. This paper summarises our findings. MATERIALS AND METHODS Isolates of M.bovis were obtained from cattle taken as reactors to the skin test or dangerous contacts, or with possible TB lesions found at slaughter. BCVA 1998 79

CATTLE PRACTICE VOL 6 PART 2 In addition, isolates were derived from badgers, either killed in removal operations or sampled during the course of a prospective study of a naturally infected population (Rogers et al 1997), and from other wildlife and exotic species where TB was suspected. DNA was prepared and spoligotyping performed as previously described (Aranaz et al 1996). Gels were analysed using GelCompar (Applied Maths BVBA, Belgium). BCVA 1998 RESULTS Thirty-four unique spoligotypes were identified from 2668 M.bovis isolates collected mainly during 1996 and 1997 from ten species (Table 1). Sixteen spoligotypes were unique to a particular species, although nine were based on just one isolate. Two spoligotypes (numbers 9 and 17) accounted for 70% of the isolates. Table 1. Spoligotypes in different species Species No of spoligotypes No of unique spoligotypes No of isolates typed Alpaca 1 1 4 Badger 17 2 621 Cat 8 1 17 Cow 25 9 2002 Deer 8 1 16 Fox 1 0 2 Goat 1 0 1 Horse 1 0 1 Human 1 0 1 Tapir 1 0 1 Unknown 2 1 2 Samples were typed from 709 known herds. Figure 1 represents the geographical distribution of the two predominant types, with 20x20 km squares coded according to the more frequently detected type. Figure 2 represents the squares according to the most frequent spoligotype found, excluding spoligotypes 9 and 17 and other types detected less than three times. Certain spoligotypes predominate in particular areas. One spoligotype was detected in 623 herds. Between two and five spoligotypes were detected in the other 86 herds (Table 2). The number of spoligotypes in a herd increased as the number of samples increased. Table 2. Samples typed from cattle Number of spoligotypes Samples per head No of herds 1 2 3 4 5 Samples typed 1 378 378 378 2 119 98 21 238 3 52 39 11 2 156 4 34 22 11 1 136 5 19 16 3 95 6 25 17 5 3 150 7 18 14 3 1 126 8 17 13 4 136 9 14 10 3 1 126 10 9 6 2 1 90 11 6 1 3 2 66 12 3 2 1 36 13 4 3 1 52 14 2 1 1 28 15 1 1 15 16 1 1 16 17 1 1 17 18 3 1 1 1 54 19 1 1 19 20 2 1 1 40 Totals 709 623 69 15 1 1 1974 80

CATTLE PRACTICE VOL 6 PART 2 Forty-three herds had isolates typed from more than one breakdown. Except in five instances at least one spoligotype was the same as in the original breakdown. A considered origin of infection for a breakdown was recorded on 426 occasions; the source of infection remained unassigned at the time of preparing this paper in 308 instances. Badgers were implicated 334 times, purchased cattle 41 times, imported Irish cattle 13 times and spread from contiguous premises five times. A source was considered unknown after investigation on 33 occasions. In seven cases where purchased cattle were incriminated, isolates were typed from both the supplying and receiving herds. The spoligotypes matched in six instances. In 99 cases, isolates from badgers, killed in control operations subsequent to herd breakdowns, were paired with the relevant cattle isolates. At least one spoligotype matched in 90 of the paired samples, while on nine occasions there was no match. Isolates were typed from badgers in 18 of 35 social groups of a known-infected population. Samples spanned a period from 1988 to 1997. Altogether, 65 isolates from 46 animals were spoligotyped. All were type 17 except for one type 11 isolate from a badger in a boundary social group without evidence so far of infection with type 17. DISCUSSION It can be seen from the current study that the number of samples typed from different species varied considerably. Cattle isolates were well represented, at least for the years 1996 and 1997 and the geographical patterns probably reflect the true distribution of spoligotypes in cattle during these years. The results from badger samples will be representative for areas with control operations but cannot be extrapolated to badger populations between the control areas. No conclusions can validly be made about spoligotypes in more exotic species until larger numbers of samples have been typed, although it is interesting, for example, to note that the spoligotype from the alpaca is so far unique to that species. However, the results are starting to provide answers to three essential questions which have to be asked of any typing system if it to be of epidemiological use. i) Is the extent of differentiation sufficient? So far, 34 spoligotypes have been identified, 25 being in cattle. Although this provides a reasonable basis for epidemiological studies, two spoligotypes predominate. Further resolution of these types, in particular, would probably be valuable. This may well be achieved by applying alternative fingerprinting techniques or by enhancing the current range of spacer sequences used in spoligotyping. ii) Are the defined spoligotypes stable in time and space? Data from the prospective study of an infected badger population covers a ten-year period and are derived from 65 isolates from 46 animals. All but one sample yielded the same spoligotype, indicating that spoligotypes may be relatively stable within a location and over a length of time, not only within a defined population but also within individual animals. Nevertheless, it is possible that multiple spoligotypes could exist within an animal if there were no selective advantage between the types. iii) Do the results give any insight into the spread of disease? In the examples of recurrent herd breakdowns, mostly the same spoligotype has been present, with additional spoligotypes found in some cases. This is open to various interpretations, for example that the original source of infection, within the herd or not, was not removed. Any wildlife reservoir could be harbouring more than one spoligotype. Multiple types within a herd would argue for this. However, unexpected results may still occur merely through the effects of sampling. Mismatches may be resolved by further sampling or the findings may suggest other routes of transmission that would otherwise have been ignored (for example, Collins et al 1994). The distribution of typed isolates so far in Great Britain suggests that a marked degree of clustering exists for several of the spoligotypes. For example, one type has principally been found in samples from Staffordshire and Derbyshire but, in addition, there was an isolate with this spoligotype from a Scottish herd, where the origin of infection was considered to have been an Irish import. One explanation of the clustering could be the evolution of strains with different spoligotype patterns in a relatively immobile reservoir species, such as the badger. Sporadic appearances of other types could be explained by longer distance movements, in particular of infected cattle being sold between herds, as in the case of the Scottish herd. This indicates the need to establish in some detail the past and current distribution of M.bovis types within GB so that unusual types within any geographic area and any temporal change in pattern can be identified. Spoligotyping looks a promising fingerprinting technique for M.bovis, and studies are in progress to compare it with other methods and establish its full epidemiological potential. ACKNOWLEDGEMENTS The Society for Veterinary Epidemiology and Preventive Medicine is acknowledged for use of maps and tables in this paper. BCVA 1998 81

CATTLE PRACTICE VOL 6 PART 2 REFERENCES Aranaz, A., Liebana, E., Mateos, A., Dominguez, L., Vidal, D., Domingo, M., Gonzolez, O., Rodriguezferri, E.F., Bunschoten, A.E., van Embden, J.D.A., Cousins, D. (1996). Spacer oligonucleotide typing of Mycobacterium bovis strains from cattle and other animals - A tool for studying epidemiology of tuberculosis. J. Clin. Microbiol. 34 2734-40. Cave, M.D., Eisenach, K.D., Salfinger, M., Bates, J.H. and Crawford, J.T. (1992). Usefulness of IS6110 in fingerprinting DNA of Mycobacterium bovis. Medical Microbiology Letters 1 96-102. Collins, D.M., Radford, A.J., de Lisle, G.W. and Billman-Jacobe, H. (1994). Diagnosis and epidemiology of bovine tuberculosis using molecular biological approaches. Vet. Microbiol. 40 83-94. Cousins, D.V. (1998). Krebs, J.R. (1997). Bovine tuberculosis in cattle and badgers. MAFF Publications, London. 191p. Report (1997). Bovine tuberculosis in badgers. Twentieth report by the Ministry of Agriculture, Fisheries and Food. MAFF Publications, London. 35p. Rogers, L.M., Cheeseman, C.L., Mallinson, P.J. and CliftonHadley, R. (1997). The demography of a high-density badger (Meles meles) population in the west of England. J. Zool. 242 705- 28. Romano, M.I., Alito, A., Fisanotti, J.C., Bigi, F., Kantor, I., Cicuta, M.E., Cataldi, A. (1996). Comparison of different genetic markers for molecular epidemiology of bovine tuberculosis. Vet. Microbiol. 50 59-71. Skuce, R.A., Brittain, D., Hughes, M.S., Beck, L.A. and Neill, S.D. (1994). Genomic fingerprinting of Mycobacterium bovis from cattle by restriction fragment length polymorphism analysis. J. Clin. Microbiol. 32 2387-92. BCVA 1998 82

CATTLE PRACTICE VOL 6 PART 2 Badgers and Bovine Tuberculosis: A Review of Studies in the Ecology of a Wildlife Disease Reservoir. Delahay, R. J., Cheeseman, C. L., Mallinson, P. J., Rogers, L. M. & Smith, G. C. Central Science Laboratory, Sand Hutton, York, YO4 1LZ. ABSTRACT Ecological studies are fundamental to an understanding of the dynamics of disease in wildlife populations and their interactions with domestic stock. A longitudinal study of a naturally infected badger population has been conducted at Woodchester Park, Gloucestershire since 1976, to examine the role of the badger (Meles meles) as a reservoir of TB infection for cattle. The social organisation of the population is monitored, and individuals are repeatedly captured and screened for TB infection. Mathematical models are used to integrate ecological and epidemiological data gathered during fieldwork, and to simulate different disease management strategies. In this review paper we highlight the contribution that ecological research has made to increasing our understanding of the dynamics of TB in badger populations and it’s transmission to cattle. KEYWORDS: Badgers, Meles meles, Cattle, Tuberculosis, Mycobacterium bovis, Wildlife Ecology. INTRODUCTION The term reservoir host refers to an animal population within which an infectious agent is maintained and which may consequently be a source of infection for other animals. There are several well documented instances of wildlife species acting as the source of persistent re-infection of domestic animals. In North America for example, the bison (Bison bison) population of Wood Buffalo National Park, Canada is the largest wildlife reservoir of bovine tuberculosis in North America (Tessaro, 1986). In New Zealand the introduced brushtail possum (Trichosurus vulpecula) is the main source of bovine tuberculosis infection in cattle (see Morris & Pfeiffer, 1995). In contrast, growing interest in the conservation of fragmented and endangered wildlife populations has highlighted the risks of disease transmission from domesticated animals to wildlife. For example, the introduction and movement of cattle in East Africa has been responsible for several outbreaks of rinderpest in wild ungulates (see Dobson & Hudson, 1995), and recent research suggests that domestic dogs may be an important reservoir of rabies infection for wildlife in the Serengeti region of Tanzania (Cleaveland & Dye, 1995). Clearly, there will be ecological constraints on the extent to which transmission of disease may occur between wildlife and domestic animals. Consequently, an understanding of the ecology of the wildlife host is an essential prerequisite to understanding the dynamics of the disease and the risks to domestic stock. In the UK and Ireland the European badger (Meles meles) is implicated in the maintenance and transmission to cattle of Mycobacterium bovis, the causative agent of bovine tuberculosis (TB). The problem is concentrated in the south-west of England where 83% of cattle breakdowns in 1995 were attributed to infection from badgers compared to less than 1% in the rest of England (MAFF, 1996). Since suspicion first fell on the badger as a source of infection in the early 1970s (Muirhead, Gallagher & Burn, 1974), outbreaks of TB in English cattle herds have continued (Wilesmith et al. 1982) and have been associated with pockets of infection in wild badgers (e.g. Little et al., 1982). However, there is as yet no scientific proof of a causative link between badgers and TB in cattle, only a compelling body of circumstantial evidence (see Krebs, 1997). Studies of disease transmission between wildlife and domestic animals require a blend of epidemiological and ecological approaches. Rates of contact between infected and susceptible individuals will be a function of their ecology. In addition, where transmission does not occur through direct contact, environmental influences on the viability of the infective agent will be important. Furthermore, the likelihood of successful disease transmission will be mediated by epidemiological considerations such as host immunity and the infectivity of the causative agent. Our research on badger ecology is broadly aimed at investigating factors that may influence the transmission of M. bovis from badger to badger and from badgers to cattle. The purpose of this paper is to review aspects of badger ecology that may influence these two processes, to describe the techniques employed and present some of the results of research carried out by the Central Science Laboratory and the Veterinary Laboratories Agency. BCVA 1998 83

CATTLE PRACTICE VOL 6 PART 2 TECHNIQUES Study area Since 1976 Woodchester Park, Gloucestershire has been the focus of a longitudinal research project on the epidemiology and ecology of a naturally infected badger population. The study area is situated on the Cotswold limestone escarpment at between 47 m and 210 m above sea level. The total area covers 11 km2 and consists of mainly deciduous woodland and permanent pasture, with smaller areas of arable, rotational grassland, scrub, coniferous and mixed woodland. The study area contains 36 badger social groups, of which 21 contiguous groups form the core study population. BCVA 1998 Study Methods The Woodchester Park badger population is sampled by regular live-trapping, using cage traps baited with peanuts. Each badger social group is trapped 4 times per year, and each individual badger is caught on average 2.5 times per year. Trapped animals are anaesthetised and bodyweight, physical condition, reproductive status and other measurements are recorded. At it’s first capture each badger is given a unique identifying tattoo in the inguinal area (Cheeseman & Harris, 1982). Samples of urine, faeces, sputum and bite wound swabs are taken for bacterial culture (Little et al., 1982). Blood samples are screened for M. bovis antibodies using an indirect ELISA (Goodger et al. 1994). Following a period of recuperation, sampled badgers are released at the point of capture. Radio telemetry and direct observation are used to monitor badger activity in the field. Collars carrying radio transmitters may be fitted to badgers so that their movements can be followed, and direct observation can be aided by the additional attachment of a luminous ‘Betalight’ (Cheeseman & Mallinson, 1979). Excretory behaviour of badgers has been investigated by spool and line tracking and the injection of flourescine dye which can be detected subsequently in urine and faeces under ultra-violet light (Brown, Harris & Cheeseman, 1992). Badger social group organisation is monitored annually by bait marking. Each spring when territorial activity is at a peak (Kruuk, 1978) badgers demarcate their territories by depositing faeces at boundary latrines. A bait of peanuts and syrup laced with indigestible coloured plastic beads is laid at each active main sett. Using beads of different appearance for each sett, it is possible to approximate territorial boundaries from their subsequent distribution in droppings at latrines (see Figure 1). In recent years epidemiological and ecological data have been used to produce computer models of the dynamics of TB in badger populations (Smith et al., 1995). This is sometimes the only way to estimate certain epidemiological parameters. The models have also been used to identify areas of data shortfall and to refocus field work accordingly. These models help to increase understanding of the interaction of epidemiology and ecology, and to predict the likely effects of management strategies which would be expensive and time-consuming to perform as field experiments (Smith, Cheeseman & Clifton-Hadley, 1997). Figure 1. Configuration of badger main setts and social group territories at Woodchester Park in 1997 estimated from a bait marking survey. FACTORS AFFECTING BADGER TO BADGER TRANSMISSION Pathological and bacteriological investigations have shown that the most important route of infection amongst badgers is respiratory, followed by bite wounding (Cheeseman, Wilesmith & Stuart, 1989). Rates of contact between infectious and susceptible badgers will be determined largely by their density, social organisation and movement patterns. In Britain badgers generally live in social groups of 3 to 10 individuals (Neal & Cheeseman, 1996) and occupy a territory which is defended against neighbouring groups (Kruuk, 1978). However, a more flexible social system may operate at lower densities (e.g. Cresswell & Harris, 1988). In the high density badger population at Woodchester Park the number of social groups and their territorial configuration has remained relatively stable throughout the study, whilst the number of individuals in each group has increased from 2.7 adults per group in 1978 to 8.8 in 1993 (Rogers et al., 1997). However, despite a steady increase in badger density, no linear relationship with the prevalence of M. bovis infection has been detected. Instead, prevalence has fluctuated in an apparently cyclical pattern (Cheeseman et al., 1989) but has been maintained in the population at an estimated annual level of 12% to 19% from 1981 to 1995 (R. S. Clifton-Hadley, unpublished data). The transmission of disease within the badger population will be facilitated by movement of individuals between social groups. Studies have shown that movement rates between groups are greater at lower badger densities where the social structure is often less stable (Cheeseman et al., 84

CATTLE PRACTICE VOL 6 PART 2 1988a). However, recent work carried out at Woodchester suggests that in this high density population nearly 50% of badgers trapped repeatedly for three or more years had moved between groups at some time (L. Rogers et al., unpublished data). Nevertheless, infection in the Woodchester population has remained largely confined to a relatively small number of social groups (Cheeseman et al., 1988b). Disruption of badger social organisation could increase rates of movement between groups and exacerbate the risks of disease transfer within the population. Following the complete removal of several social groups from Woodchester Park, the initial phase of re-colonisation involved movements of badgers over greater distances than usual, with a stable social structure only returning to the area 9 to 10 years later (Cheeseman et al., 1993). The extent to which badger culling could exacerbate disease spread through such effects is currently under investigation. Territorial behaviour of badgers reaches a peak during spring when excretory activity at boundary latrines (Kruuk, 1978) and the incidence of bite wounds (Gallagher & Nelson, 1979) are greatest. If bite-wounding is largely the result of territorial disputes with neighbours then it may be an important source of disease transfer between groups. This is consistent with the observation that bite wounds and the prevalence of infection are greatest in male badgers (Gallagher & Nelson, 1979; Cheeseman et al., 1988b) who are also most active in territorial defence (Cheeseman et al., 1988a). However, it is not known to what extent disputes within group hierarchies may also be responsible for such wounds. The intimate association of badgers within a social group will facilitate transmission of M. bovis between group members by the respiratory route. The focus of badger social activity is the main sett, where individuals may often share underground chambers on a regular basis (Roper & Christian, 1992). The constant temperature and humidity in the sett may be conducive to the prolonged survival of M. bovis (Jackson, de Lisle & Morris, 1995) and provide optimal conditions for disease transfer between group members. There is also limited evidence for post natal maternal transmission from sows to their cubs, which if widespread could be important in maintaining infection in the population (Cheeseman et al., 1989) However, within social groups it is not uncommon for only one or two individuals to be infected (Cheeseman et al., 1988b) which suggests that other factors such as immunity or behaviour may regulate exposure to infection within the group. FACTORS AFFECTING BADGER TO CATTLE TRANSMISSION The location of lesions in cattle associated with badger related breakdowns strongly suggests the primary route of infection is respiratory (Gallagher, 1980; Wilesmith et al. 1986). This is likely to take place through inhalation following investigation or incidental ingestion of badger excretory products by grazing cattle. Studies of badger activity and excretory behaviour on cattle pasture therefore provide important information on the potential distribution of M.bovis, and may help to identify methods of reducing infection risks to cattle. Although badgers have a wide omnivorous diet, the single most important food item is earthworms (see Neal & Cheeseman, 1996). Temperatures above 20 C and high humidity provide good ‘worming nights’ for badgers as they bring the earthworm Lumbricus terrestris to the surface to feed (Satchell, 1967). The relatively high densities of earthworms under pasture makes this an attractive habitat for foraging, and brings badgers into the proximity of grazing cattle. Badgers deposit faeces, urine and anal secretions at latrines which are used to demarcate their territories (Kruuk, 1978). Where they are accessible to grazing cattle, badger latrines may constitute a significant risk of infection. In a study at Woodchester Park, Brown, Harris and Cheeseman (1992) determined the distribution of excretory products from several badgers using fluorescine dye. The results confirmed the importance of latrines but also showed that 28% of badger urinations occurred on pasture away from setts or latrines, and were often associated with places where badger paths crossed hedges and fences (Brown, 1993). It has been estimated that up to 300,000 M. bovis bacilli may be excreted in one millilitre of urine from a badger with extensive kidney lesions (MAFF, 1979). Grass contaminated in this way could therefore be an important source of infection to grazing cattle. Research has shown that cattle will avoid feeding near badger latrines but become less discriminating as grazing pressure increases (Benham & Broom, 1991; Hutchins & Harris, 1997). These findings suggest that the manipulation of grazing regimes could be used as a tool to reduce disease risks to cattle. Because badger excretory products are likely to be the source of infection in cattle rather than direct contact with badgers themselves, the rate of infection in cattle will not necessarily be a simple function of the density of infected badgers. Density dependent effects could for example be overwhelmed by the effects of climate on the survival and viability of M. bovis in the environment. The epidemiological importance of climate is greater for infectious agents that may persist outside the host. Research on the survival of M. bovis suggests that under optimal conditions bacilli may persist for up to 11 months (E. King, unpublished data). Another potential source of infection for cattle is the contamination of foodstuffs that may result from badgers entering farm buildings. A Radio-tracking study by Cheeseman and Mallinson (1981) investigated the incidence of badgers entering farm BCVA 1998 85

CATTLE PRACTICE VOL 6 PART 2 buildings and recorded that 64% of those found inextremis were infected. Tuberculous badgers may be attracted to farm buildings to exploit readily available foodstuffs, or may simply seek refuge outside their social group after failing to maintain their former social status. THE BADGER AS A WILDLIFE DISEASE RESERVOIR Badgers are widespread throughout the British Isles, densities being greatest in south-west England (Wilson, Harris & McLaren, 1997). Many areas of high badger density are also associated with cattle farming, indeed the management of grasslands for grazing provides badgers with attractive foraging habitat. Essentially, wherever we farm cattle we may be incidentally farming badgers. TB is endemic in the British badger population. However, badger mortality due to TB is low, so infected individuals may survive for relatively long periods whilst excreting bacilli. Also, infected females can produce cubs and the impact of the disease on population size and structure appears to be minimal (Cheeseman et al., 1989). In addition, the ecological niche of the badger brings it into close and protracted contact with pasture, where infectious excretions may become available to grazing cattle. Together these factors make the badger an ideal maintenance host for M. bovis and an important potential source of persistent re-infection for cattle. However, certain characteristics of badger ecology mitigate against the spread of infection. At Woodchester for example the stable social structure and limited movement of badgers between groups limits contact rates between individuals. Intra-group transmission of M. bovis is facilitated by intimate social contact whereas inter-group transmission is limited by spatial segregation and generally low rates of dispersal. This may be one reason why the distribution of infection is localised within the badger population and is not linearly related to changes in density. Data from Woodchester Park and elsewhere show that despite the presence of M. bovis infection in a badger population, protracted co-existence with cattle is possible without them becoming infected (Muirhead et al., 1974; Wilesmith et al., 1986; Cheeseman et al., 1988b). This suggests that the probability of transmission between badgers and cattle may be influenced by local conditions. CONCLUSIONS BCVA 1998 There is convincing circumstantial evidence that in Britain badgers are an important wildlife disease reservoir for the transmission of bovine tuberculosis to cattle. The badger is an ideal maintenance host because infected individuals can survive for relatively long periods and produce viable young, and infection appears to have no significant effects on population size or structure (Clifton-Hadley, 1996). However, whether badgers are only one component of a multi-species disease reservoir is not clear. There is evidence of infection in badgers from throughout mainland Britain (Cheeseman et al., 1989) including areas of high badger density but low cattle breakdown rates. Local variation in cattle breakdown rates may be related to differences in disease dynamics, social organisation and movement patterns of local badger populations, farm management practices, climate and possibly the role of other wildlife. Ecological studies will play an important role in future research to investigate the influence of these factors. ACKNOWLEDGEMENTS We thank Paul Spyvee, Dave Handoll and John Howell of Woodchester Park, and the staff of the Bacteriology section of CVL for expert technical assistance. We are also indebted to all the farmers and landowners in the Woodchester Park study area for their co-operation during fieldwork. The project was funded by MAFF. REFERENCES Benham, P. F. J. & Broom, D. M. (1991). Responses of dairy cows to badger urine and faeces on pasture with reference to bovine tuberculosis transmission. British Veterinary Journal, 147, 517- 532. Brown, J. A. (1993). Transmission of bovine tuberculosis (Mycobacterium bovis) from badgers (Meles meles) to cattle. PhD. Thesis, University of Bristol. Brown, J. A., Harris, S. & Cheeseman, C. L. (1992). The development of field techniques for studying potential modes of transmission of bovine tuberculosis from badgers to cattle. In Hayden, T. J. (ed.). The Badger. 139-153. Royal Irish Academy, Dublin. Cheeseman, C. L. & Harris, S. (1982). Methods of marking badgers (Meles meles). Journal of Zoology, 197, 289-292. Cheeseman, C. L. & Mallinson, P. J. (1979). Radio tracking in the study of bovine tuberculosis in badgers. In Amlaner Jr, C. J. & Macdonald, D. W. (eds). A Handbook on Biotelemetry and radio tracking. 649-656. Pergamon Press, Oxford and New York. Cheeseman, C. L. & Mallinson, P. J. (1981). Behaviour of badgers (Meles meles) infected with bovine tuberculosis. Journal of Zoology, 194, 284-289. Cheeseman, C. L., Mallinson, P. J., Ryan, J. & Wilesmith, J. W. (1993). Recolonisation by badgers in Gloucestershire. In Hayden, T. J. (ed.). The Badger, 78-93. Royal Irish Academy, Dublin. Cheeseman, C. L., Cresswell, W. J., Harris, S. & Mallinson, P. J. (1988a). Comparison of dispersal and other movements in two badger (Meles meles) populations. Mammal Review, 18, 51-59. Cheeseman, C. L., Wilesmith, J. W., Stuart, F. A. & Mallinson, P. J. (1988b). Dynamics of tuberculosis in a naturally infected badger population. Mammal Review, 18 (1), 61-72. Cheeseman, C. L., Wilesmith, J. W. & Stuart, F. A. (1989). Tuberculosis: the disease and its epidemiology in the badger, a review. Epidemiology and Infection, 103, 113-125. Cleaveland, S. & Dye, C. (1995). Maintenance of a microparasite infecting several host species: rabies in the Serengeti. Parasitology, 111, S33-S47. Clifton-Hadley, R. S. (1996). Badgers, bovine tuberculosis and the age of reason. British Veterinary Journal, 152, 243-246. Cresswell, W. J. & Harris, S. (1988).Foraging behaviour and home range utilization in a suburban badger (Meles meles) population. Mammal Review, 18, 37-49. Dobson, A. P. & Hudson, P. J. (1995). Microparasites: observed patterns in wild animal populations. In Grenfell, B & Dobson, A. 86

CATTLE PRACTICE VOL 6 PART 2 (eds). Ecology of Infectious Diseases in Natural Populations. Cambridge University Press, Cambridge, 52-89. Gallagher, J. (1980). The role of other animals in the epidemiology of TB in the badger. In Badgers, cattle and tuberculosis. Report to Rt. Hon. Peter Walker M. P. Ed. Zuckerman, Lord. pp 86-98. HMSO, London. Gallagher, J. & Nelson, J. (1979). Cause of ill health and natural death in badgers in Gloucestershire. Veterinary Record, 105, 546- 551. Goodger, J., Nolan, A., Russell, W. P., Dalley, D. J., Thorns, C. J., Stuart, F. A., Croston, P. & Newell, D. G. (1994). Serodiagnosis of Mycobacterium bovis infection in badgers: development of an indirect ELISA using a 25 kDa antigen. Veterinary Record, 135, 82-85. Hutchins, M. R. & Harris, S. (1997). Effects of farm management practices on cattle grazing behaviour and the potential for transmission of bovine tuberculosis from badgers to cattle. Veterinary Journal, 153, 149-162. Jackson, R., de Lisle, G. W. & Morris, R. S. (1995). A study of the environmental survival of Mycobacterium bovis on a farm in New Zealand. New Zealand Veterinary Journal, 43 (7), 346-352. Krebs, J. R. (1997). Bovine Tuberculosis in Cattle and Badgers. Report to the Rt Hon. Dr Jack Cunningham MP.The Ministry of Agriculture, Fisheries and Food, London. Kruuk, H. (1978). Spatial organisation and territorial behaviour of the European badger Meles meles. Journal of Zoology, 184, 1-19. Little, T. W. A., Swan, C., Thompson, H. V. & Wilesmith, J. W. (1982). Bovine tuberculosis in domestic and wild mammals in an area of Dorset. II. The badger population, its ecology and tuberculosis status. Journal of Hygiene, Camb. 89, 211-224. MAFF. (1979). Bovine Tuberculosis in Badgers. Third Report. Ministry of Agriculture, Fisheries and Food, London. MAFF. (1996). Bovine Tuberculosis in Badgers. Nineteenth Report. Ministry of Agriculture, Fisheries & Food, London. Morris, R. S. & Pfeiffer, D. U. (1995). Directions and issues in bovine tuberculosis epidemiology and control in New Zealand. New Zealand Veterinary Journal, 43 (7), 256-265. Muirhead, R. H., Gallagher, J. & Burn, K. J. (1974). Tuberculosis in wild badgers in Gloucestershire: Epidemiology. Veterinary Record, 95, 552-555. Neal, E. & Cheeseman, C. L. (1996). Badgers. T & A. D. Poyser Ltd, London. Rogers, L. M., Cheeseman, C. L., Mallinson, P. J. & CliftonHadley, R. (1997). The demography of a high density badger (Meles meles) population in the west of England. Journal of Zoology, 242, 705-728. Roper, T. J. & Christian, S. F. (1992). Sett use in badgers (Meles meles). In Priede, I. G. & Swift, S. H. (eds). Radio telemetry: remote monitoring and tracking of animals. Ellis Harwood, Chichester, 661-669. Satchell, J. E. (1967). Lumbricidae. In Burgess, A. & Raw, F. (eds). Soil Biology. Academic Press, London & New York, 259- 322. Smith, G. C., Richards, M. S., Clifton-Hadley, R. S. & Cheeseman, C. L. (1995). Modelling bovine tuberculosis in badgers in England: preliminary results. Mammalia, 59 (4), 639- 650. Smith, G. C., Cheeseman, C. L. & Clifton-Hadley, R. S. (1997). Modelling the control of bovine tuberculosis in badgers in England: culling and the release of lactating females. Journal of Applied Ecology, 34, 1375-1386. Tessaro, S. V. (1986). The existing and potential importance of Brucellosis and Tuberculosis in Canadian wildlife: A review. Canadian Veterinary Journal, 27, 119-124. Wilesmith, J. W., Little, T. W. A., Thompson, H. V. & Swan, C. (1982). Bovine tuberculosis in domestic and wild mammals in an area of Dorset. I. Tuberculosis in cattle. Journal of Hygiene, 89, 195-210. Wilesmith, J. W., Sayers, P. E., Bode, R., Pritchard, D. G., Stuart, F. A., Brewer, J. I. & Hillman, G. D. B. (1986). Tuberculosis in East Sussex. II. Aspects of badger ecology and surveillance for tuberculosis in badger populations (1976-1984). Journal of Hygiene (Cambridge), 97, 11-26. Wilson, G., Harris, S. & McLaren, G. (1997). Changes in the British badger population, 1988 to 1997.Peoples Trust for Endangered Species, London. BCVA 1998 87

CATTLE PRACTICE VOL 6 PART 2 BCVA 1998 88

CATTLE PRACTICE VOL 6 PART 2 Animal Welfare: Ethics, Economics and Productivity McInerney J., Agricultural Economics Unit, Exeter University, Lafrowda House, St German’s Road, Exeter, Devon. EX4 6TL DISCUSSION Attention to animal welfare has always been an explicit consideration in veterinary practice and an obvious component of good livestock husbandry - i.e. it has been seen as an interest from the production side of livestock in the food chain. However, of recent years it has increasingly become an issue of interest on the demand side, with food consumers and, ostensibly on their behalf, food retailers viewing the conditions under which animals are kept as effectively a quality characteristic of the final livestock product. Independent of these food chain connections, various national and international groups have more actively espoused the cause of farm animal welfare as a political issue about which modern society should be more sensitised and reactive. There are a jumble of forces at work within these varying interests and pressures. At one extreme there are ethical and other value judgements about how human beings should treat other sentient beings, some people talking in terms of inherent ‘animal rights’ and others simply feeling a growing discomfort as they reflect on what they are told about the conditions under which modern farming takes place. At the other extreme are pragmatic issues about the most effective way of exploiting the biological production potential of animals in an environment of market competition for livestock products, without crossing thresholds that might be widely judged to be inhumane. In between these sets of value judgements are impersonal forces of the emergence of new technological possibilities and systems of animal production, evolving national and international trading structures, developments in the demand for food, and changing perceptions and preferences in society. All these facets demonstrate that, while ‘animal welfare’ at first sight appears to be about the wellbeing of the animal, and hence something to be considered within the technical/science context of the animal scientist/vet/farmer, it has increasingly become an issue that requires a social science framework in order to understand its current prominence - and, in particular, to analyse what might/should be done about it. This conclusion is further reinforced by the fact that (a) there is no clear and workable definition in scientific terms of what animal welfare actually is, how to measure it, and how to rank different animal welfare states; and (b) what is actually done about animal welfare is determined by what people decide - so it is their perceptions and preferences (which may, but need not, be informed by science-based information), not the animal’s, which define the parameters in practice, thereby effectively making animal welfare a subset of human welfare (i.e. the guiding influence is that we need to feel comfortable with the way animals are treated). There are several, essentially inconsistent, perceptions of animal welfare that seem to be extant in the current debate - much of which anyway tends to focus just on particular practices (embryo transfer, veal crates, slaughter) rather than on the overall systems of production within which livestock are kept. Between the limits of ‘maximal’ welfare (keeping the animal in a manner that is perceived as the best possible state for it) and ‘cruelty’ (subjecting it to conditions which the society of the day would regard as unacceptable) are a range of possible husbandry conditions. In general, this range reflects a competitive relationship between the animal’s (economic) productivity and its welfare. There is a tendency for agricultural production technique to pursue the animal’s biological capacity for producing economically valuable output - milk, liveweight gain, etc - in a way that creates an increasingly ‘unnatural’ environment for it, and hence is perceived as causing an increasing challenge to its welfare (this is the popular perception of ‘intensity’ in farming). In the past, all the economic incentives in the human food system have fostered this pursuit - indeed it has often been the mark of achievement amongst science and farming peer groups; the (welfare) cost to the animal does not appear in any monetary form, does not get included in the accounting calculations, and hence does not get set against the economic gain from enhanced animal productivity. Consequently, even in an affluent and tolerably civilised society, the impersonal forces of the market can lead to the unacceptable exploitation of farm animals, in the same way as those forces ignore the interests of the poor, the economically vulnerable, the environment, the future - all sorts of groups and elements that are external to the market transaction. These effects are nothing particularly to do with the existence or operation of farming support policies and other distortions of the modern agricultural scene, but are inherent in the workings of a market economy. Economists (as opposed to accountants) have long focussed their analytical attention on these ‘externalities’, those elements of genuine economic cost and benefit to society that do not appear in trading accounts, and their discussion of appropriate BCVA 1998 89

CATTLE PRACTICE VOL 6 PART 2 policies revolves around the means whereby such inefficiencies in resource use can be corrected/ prevented. In the context of social concerns over the welfare of farm animals, it is not evident that government policies are either necessary or the most effective means of bringing about the desired styles and standards of animal husbandry that the wider public is beginning to demand. An affluent, well fed, food secure society that is no longer particularly price sensitive in its food purchase decisions is voicing its perceptions about how farm animals should be kept. (Some will argue whether the ‘consumer’ is the household or the food retailer. It makes little difference, because either way it is a specifically demand-led phenomenon in what has traditionally been a production process driven by supply-side concerns.) If the underlying thesis of a welfareproductivity conflict is valid then higher welfare standards in livestock production inevitably mean foregoing some actual or potential productivity gains, and hence will result in higher supply costs of livestock products. This is not in itself a problem, however. The whole of economic logic - and we each reveal it regularly in our own personal behaviour - says that people seek the products and quality characteristics they prefer, will pay for them if they can afford to, and will feel more satisfied (gain greater economic benefit) as a result. Simply minimising expenditure is not the criterion that guides consumption behaviour. The developing market for livestock products is evidence of this in the higher priced ‘quality assured’ and ‘welfare friendly’ labels that are becoming more common. Nor is the extra cost likely to be all that significant. There is a tendency on the part of those connected with the farming community to sit inside the farmgate, caution against the extra production costs of increased welfare standards, and then to view them as a threat to producers’ incomes. A little thought, and some fairly straightforward calculations, show this to be misguided. From the standpoint of the food system, it is the supply cost of the consumed commodity, not the farmgate cost of the raw material, that is the primary interest. Factored through the value adding economic processes in the food chain from farm to final consumer, any cost increases in livestock production emerge greatly reduced as a proportion of the final food product price. And as a proportion of household food expenditures, such cost increases would appear to be nothing for anyone to get excited about. The above discussion merely sets out the elements of a complex process. Its merit is that it is built upon the logical propositions of economic analysis and the almost tautological proposition that, whatever the price of a (food) commodity, if people are prepared to pay that price and choose to buy the product in preference to another, they must be gaining a net economic value by doing so. This is the case with the growing demand for livestock products from higher cost systems with perceived higher welfare characteristics. Without being too cynical, one can argue that whether such choices are well-informed, and who it is who knows ‘the truth’ about what constitutes good animal welfare - the animal behaviourist, the vet, the animal scientist, the farmer, the animal welfare activist, the supermarket buyer, the consumer - is not necessarily an issue. The implications of this discussion for developments in cattle production are an interesting topic for further exploration. Consumer perceptions that cows should lie on straw, for example, or necessarily require access to green fields, or that any incidence of mastitis or lameness means productivity and profit considerations have over-ridden welfare concerns, may owe as much to Beatrix Potter or simple ignorance of livestock production processes as to ‘real’ information. A popular perception that calves suckled for at least nine months and grass-fed beef is undoubtedly better in welfare terms than early weaning and ‘intensive’ rearing in housed systems may not accord at all with the cattle farmer’s perception of the all-round best method of production. But that is not the point. In the last analysis, livestock production for the food chain does not function for the benefit of (a few) producers; as for every other resource owner in the economy, it is what the (far more numerous) food consumers want that determines what is ‘right’. As the great economist Adam Smith wrote in 1775, the object of economic activity is not production; it is consumption! BCVA 1998 90

CATTLE PRACTICE VOL 6 PART 2 Coliform Mastitis; An Evolving Problem? Green M*., and A J Bradley** * Orchard Veterinary Group, Wirrall Park Road, Glastonbury, Somerset BA69XE. ** University of Bristol, Division of Molecular and Cellular Biology, Department of Clinical Veterinary Science, Langford House, Langford, Bristol. BS18 7DU ABSTRACT E coli and similar organisms are an important cause of bovine mastitis. It is uncertain whether the incidence of E coli mastitis is increasing but it does not seem to be decreasing. This questions the usefulness of current control methods. This paper reviews literature which suggests various possible risk factors for E coli mastitis and introduces some current areas of research. KEYWORDS: E Coli, Mastitis, Incidence, Risk Factors, Somatic Cell Counts CURRENT SITUATION Mastitis caused by ‘coliforms’, sometimes referred to as Enterobacteriacae, continues to be a problem in dairy cows. E coli is the most common mastitis isolate of this group. Other members of the genus include Klebsiella spp., Enterobacter spp., Citrobacter spp and Seratia spp. The current incidence and prevalence of mastitis causing organisms in the UK is not easy to assess because no recent national studies have been carried out. However the following information was presented at the British Mastitis Conference in 1997 (Booth J M. 1997). It documents VIDA figures of mastitis isolates (probably from clinical and subclinical cases) from 1975 and 1995 (Table1) and the results of 2 clinical mastitis surveys, the last of which occurred 15 years ago (Table2). Table 1. Main mastitis isolates (VIDA) (% of diagnoses) Pathogen 1975 1995 Change from 1975 Staph (not specified) 23 2 \ -2 Staph aureus - 19 / Strep dysgalactiae 10 7 -3 Strep uberis 13 17 +4 Strep agalactiae 7 5 -2 E coli 21 28 +7 A pyogenes 6 3 -3 BCVA 1998 Table 2. Incidence of clinical mastitis (cases per 100 cows pa) (Wilson and Kingwill 1975 and Wilesmith and others 1986) Pathogen 1967 1982 Change from 1967 Staph aureus 51 6 -45 Strep dysgalactiae 27 4 -23 Strep uberis 24 7 -17 Strep agalactiae 4 1 -3 E coli 7 7 0 Others 8 10 +2 No isolate 21 6 -15 The organisms associated with mastitis have been identified in a current ongoing study involving 750 cows (Table3). The data are from 6 Somerset dairy herds with bulk cell counts generally rolling between 70,000 and 200,000/ml, (see following paper for details) and are of interest to compare with the figures above. Table 3. Mastitis isolates from six dairy herds in Somerset 1997-8 Cases/100 cows/year * % of total isolates Coag +ve Staphs 2.9 7.3 S. dysgalactia 2.6 6.5 S. uberis 5.9 14.9 S. agalactia 0 0 E. coli 14.2 36.0 All Enterobacteriacae 16.2 41.0 A. pyogenes 0.9 2.3 Others 2.9 7.3 No growth 10.3 26.1 TOTAL 39.4 - All these studies suggest that E coli is an important cause of bovine mastitis. A lack of data means it is difficult to know whether there has been an absolute increase in E coli mastitis in recent years or simply a relative increase due to a reduction in other bacteria. At least most commentators agree there does not seem to have been a reduction in mastitis due to E coli. The traditional concept of E coli mastitis is that the organisms live in the environment and contaminate the teats. Invasion of the udder is considered to occur when the teat orifice is open eg. at or soon after milking or after teat damage. Following rapid bacterial multiplication in the milk, an inflammatory response is mounted. The severity of the disease is then thought to be influenced in part by the speed of the immune response and in particular PMN 91

CATTLE PRACTICE VOL 6 PART 2 migration into the udder. Control of the disease has centred around reducing environmental challenge around parturition and during lactation and ensuring strict hygiene at and after milking. The mystery of E coli mastitis is that while herds have been able to make managemental changes to reduce the amount of contagious mastitis, husbandry improvements have apparently not resulted in a reduction of coliform infections. It has been found that even well managed herds which can maintain low somatic cell counts may find it difficult to control E coli mastitis (eg. Miltenburg and others, 1996, Hogan and others, 1989). Why is it that herds which are capable of reducing contagious pathogens seem less able to reliably prevent E coli mastitis?. Are they not able/prepared to ‘clean up’ sufficiently?. Is it impossible to reduce environmental challenge by the necessary amount in day-to-day farming conditions?. Or is the situation more complicated than this?. The lack of success in reducing E coli mastitis has led to questions being asked concerning other risk factors for the disease. Various studies have implicated factors which may increase the likelihood of E coli mastitis, some of them are discussed below. SOMATIC CELL COUNTS Certainly one of the most controversial areas involving E coli mastitis!. As long ago as the 1960s experimental evidence demonstrated that elevated somatic cell counts in a quarter could give subsequent protection against coliform infection (Schalm and others 1964). Cell counts and E coli mastitis were discussed in the Veterinary Record letters section in 1976 (Jones TO 1976) , it is clearly not a new subject!. A study on individual cow SCCs and mastitis in the USA in the 1980s found that mastitis generally increased as SCCs increased (Coffey and others, 1986) although this may not have been surprising because the type of pathogen was not specified and higher cell count cows will often be prone to recurrent bouts of contagious mastitis. Conversely a more recent study in The Netherlands (Miltenburg and others 1996) found a significant increase in mastitis (the commonest pathogen being E coli) associated with a reduction in herd somatic cell counts. The subject of SCCs and mastitis has come to the fore in the UK, probably because of the reduction in SCCs over the last 5-10 years. In a field study in the UK on cows with toxic mastitis (most commonly caused by E coli) it was found that cases were more likely to occur in herds with a lower bulk milk somatic cell count in that month than control herds (Green and others 1996). This raised the question of causality (ie whether the low cell counts caused the mastitis) and also the issue of how bulk tank cell counts relate to individual cow or quarter cell counts. These questions have been further examined and some results are given in a later paper. Experimental studies have also been carried out and there has been one particularly interesting experimental study recently reported (Shuster and others1996). In this investigation 12 cows were inoculated with E coli to induce mastitis. 6 cows were infected just after calving and 6 in mid lactation with the intention of examining why post-parturient cows are more susceptible to E coli mastitis than those later in lactation. It was found that the postparturient cows did get a more severe mastitis but that it was not due to a slower speed of leukocyte recruitment. In fact post-parturient cows recruited leukocytes faster and to a greater level. The authors did find a lower initial cell count in the cows which had recently calved and suggested that this may have been a reason for the severe mastitis. They attributed the difference in severity of mastitis to an inability of post-parturient cows to control bacterial growth in the first few hours after inoculation, before leukocyte recruitment began. Bacterial numbers escalated greatly in this time and the result was a more severe mastitis. Mid lactation cows which started with more cells in the milk were apparently able to control early bacterial growth to a greater extent. One cow which did have a particularly slow speed of leukocyte recruitment to the udder, however, became the most severely affected of all. This agreed with earlier findings (Hill 1981) that a slow PMN response can be important in dictating the severity of clinical disease. Could it therefore be that both initial numbers and speed of migration of leukocytes are important in a cows defence against E coli mastitis?. Furthermore, does this mean that as somatic cell counts decrease, the number of PMNs in the milk decreases and a cow becomes more likely to succumb to E coli mastitis?. In another experimental study on 12 cows infected with E coli, (Dosogne and others, 1997), a negative correlation was reported between severity of mastitis and the number of blood neutrophils, the number of mature PMNs and the number of phagocytic PMNs in the blood. This again suggests that PMN numbers and function are important in the cows defence against E coli, although the mechanisms of regulation of PMN characteristics (eg abilities to migrate and phagocytose) during mastitis are not completely understood. It also raises the question of whether the PMN number and function in blood or milk is dictated simply by genetics or whether it is also determined by environment i.e. response to previous challenge?. A 3 year PhD based at the Universities of Bristol and Glasgow has been funded by the MDC to examine whether the number of somatic cells in the udder and also the types of cell present are important in protection against E coli mastitis. BCVA 1998 92

CATTLE PRACTICE VOL 6 PART 2 POST MILKING TEAT DIPPING, MINOR PATHOGENS AND ‘BLITZ THERAPY’ Studies in the Netherlands have found that in low somatic cell count herds (<150,000/ml), post milking teat disinfection can increase the risk of E coli mastitis (Schukken and others 1991, Lam 1996). The authors attribute this to a removal of minor pathogens (‘non pathogenic bacteria’) from the mammary gland and therefore the removal of the protective effect these bacteria may provide. Lam calculated a decrease in E coli infections of 71% if post milking teat dipping was discontinued. The role of minor pathogens might be linked to somatic cell count levels as they may stimulate a small cellular response in the udder but it has also been suggested that their protective role could arise from ‘competitive exclusion’. E coli mastitis following blitz therapy (widespread antibacterial treatment for removing Strep agalactiae infection) was the subject of a series of recent letters in the Veterinary Record (Boyer, Biggs, Edmonson, Bradley, July-August 1997). A possible protective effect of minor pathogens has implications for blitz therapy because such bacteria will be removed by it. It is therefore interesting that E coli infections after blitz therapy were reported for weeks after therapy rather than only immediately following tube administration. Perhaps it is possible that, as the bacteriological and immunological environment within the mammary gland changes following blitz therapy, newly acquired or pre-existing infections that would previously have been ‘suppressed’ may become manifest. This also raises the question of dry cow therapy which is also effective at removing minor pathogens. Is it a coincidence that most E coli infections are reported to occur within a week of calving?. DRY PERIOD INFECTIONS A field study on a 190 cow dairy herd in the USA (Todhunter and others 1991) found that many (>60%) new gram negative infections originated in the dry period. The secretion of the non-lactating gland is known to have inhibitory properties to gram negative bacteria and therefore clinical mastitis is not seen until the gland changes to the lactating state even though infection can occur. It is possible that the dry period may be an important time for new E coli infections in UK conditions and this could be a risk factor for E coli mastitis in the subsequent lactation. This area is currently being researched in the UK. OTHER RISK FACTORS Other factors which may increase the likelihood of E coli mastitis include increasing milk flow rates (Grindal and Hillerton 1991), fatty liver disease (Reid 1981) and low vitamin E / selenium in the diet (Smith and Hogan 1993). CONCLUSIONS Prevention of E coli mastitis presents a problem simply because in farm conditions it is impossible to ensure that the organisms do not come into contact with cows teats. Current methods are probably not reducing the incidence of E coli mastitis and this may be partly because cows are becoming more susceptible to the disease even if hygiene is improved. More research is needed to determine reliable methods to control the disease. REFERENCES Booth J M (1997) British Mastitis Conference proceedings 3 Coffey E M, Vinson W E and Pearson R E (1986) Journal Dairy Science 69 552 Dosogne H, Burvenich C, van Werven T, Roets E, NoordhuizenStassen E N and Goddeeris B (1997) Veterinary Immunology and Immunopathology 60 47 Green M J, Green L E and Cripps P J (1996) Veterinary Record 138 305 Hill A W (1981) Research in Veterinary Science 31 107 Grindal R and Hillerton E (1991) Journal Dairy Research 58 263 Hogan J, Smith K L, Hoblet K, Schoenberger P, Todhunter D, Hueston W, Pritchard D, Bowman G, Heider L, Brockett B and Conrad H (1989) Journal Dairy Science 72 1547 Jones T O (1976) Veterinary Record 98 410 Lam T (1996) PhD Thesis, Utrecht University, The Netherlands Miltenburg J, de Lange D, Crauwels A, Bongers J, Tielen M, Schukken Y, and Elbers A (1996) Veterinary Record 139 204 Reid I (1981) BCVA Proceedings 1979-80 65 Schalm O W, Lasmanis J, Carroll E J (1964) American Journal of Veterinary Research 25 83 Schukken Y, Grommers F, Van Der Geer D, Erb H and Brand A, J Dairy Science (1991) 74 3 826 Shuster D E, Lee E K and Kehrli M E (1996) American Journal of Veterinary Research 57 11 1569 Smith K L, and Hogan J (1993) Veterinary Clinics of North America 3 489 Todhunter D, Smith K L, Hogan J, and Schoenberger P (1991) American Journal of Veterinary Research 52 2 184 Wilesmith J W, Francis P G, and Wilson C D (1986) Veterinary Record 106 431 Wilson C D and Kingwill R G (1975) IDF Document 85 422 BCVA 1998 93

CATTLE PRACTICE VOL 6 PART 2 BCVA 1998 94

CATTLE PRACTICE VOL 6 PART 2 A Prospective Investigation of Intramammary Infections due to Enterobacteriacae during the Dry Period: A Presentation of Preliminary Findings. Bradley A.J*., Martin Green ** *University of Bristol, Division of Molecular and Cellular Biology, Department of Clinical Veterinary Science, Langford House, Langford, Bristol. BS48 7DU. ** Orchard Veterinary Group, Wirral Park Road, Glastonbury, Somerset. BA6 9XE. ABSTRACT Enterobacterial mastitis continues to be a major cause of financial loss to the UK dairy industry, despite significant progress in the control of contagious mastitis. This paper presents some preliminary findings from a study designed to investigate the incidence and significance of enterobacterial infections acquired during the non-lactating period. The bacteriological status of three hundred and seventy three cows was assessed through the non-lactating period; mastitis samples were collected from these cows in the subsequent lactation. A significant rise in the incidence of intramammary infection was detected through the dry period. Quarters infected during the dry period, with enterobacteriacae, were significantly more likely to develop mastitis due to the same pathogen, than uninfected quarters. In 52% of enterobacterial mastitis cases the causal organism had been previously isolated, in the same quarter, during the dry period. KEYWORDS: Mastitis, E. coli, Enterobacteriacae, Dry Period. INTRODUCTION In recent years implementation of the five point plan has resulted in a marked decline in the incidence of contagious mastitis1 . This decline has not been accompanied by a comparable fall in the incidence of environmental mastitis; which has therefore become of relatively greater importance. Classically, the non-lactating mammary gland has been considered refractory to ‘coliform’ infection2 . However research in the US, from as early as 1943, has implicated the dry period as being the time of greatest risk for the establishment of new gramnegative intra-mammary infections (IMIs)3,4,5,6,7, with some 61% of new IMIs occurring at this time. Further studies have illustrated the ability of such infections to remain quiescent within the udder until calving, subsequently causing clinical mastitis in early lactation8 . To date, despite some anecdotal evidence, similar studies, to validate these findings, have not been carried out in the UK. As a consequence the importance of the dry period in the control of environmental mastitis remains equivocal. The aim of the research outlined in this paper is to investigate the incidence of intra-mammary infections acquired during the non-lactating period of dairy cattle, under UK field conditions and their subsequent importance in the ensuing lactation. MATERIALS AND METHODS Herd Selection Herds were selected on the basis of location (Somerset), low bulk milk somatic cell count (generally <200,000), calving pattern and likelihood of owner compliance with the study protocol. All herds were milk recorded (NMR/DAISY). All cows calving within the time period selected (02/04/97- 11/11/97) were included in the study. Sampling Strategy Duplicate lacteal secretion samples were collected, by the authors, from all four quarters at ‘drying off’ and during the week following calving. During the dry period, duplicate samples were taken from two quarters (LF and LH (odd numbered cows) or RF and RH (even numbered cows)) once weekly during the two weeks prior to the anticipated calving date. The other two quarters remained as unsampled controls to eliminate the sampling procedure as a cause of new intramammary infections. Any cow not calving by her ‘due date’ was sampled weekly until parturition. During the subsequent lactation milk samples were collected from all mastitic quarters identified by the herdsmen and graded for severity. Sampling Procedure Prior to sampling, teats were cleansed of gross contamination and dipped in a solution containing 2800ppm available chlorine (Agrisept, UpJohn). BCVA 1998 95

CATTLE PRACTICE VOL 6 PART 2 Following a minimum 30 second contact time the teats were wiped dry. Each teat was subsequently scrubbed with surgical spirit and allowed to dry. Immediately before sampling the teat ends were scrubbed for a second time using surgical spirit and foremilk was discarded (Strict foremilk was collected from udders assessed as having little secretion present). Duplicate samples were then collected, following a third scrub of the teat ends. Following sampling, teats were dipped in a solution containing 2800ppm available chlorine (Agrisept, UpJohn) and cows confined to a loafing yard for at least a half hour. Samples were immediately stored in a cool box and maintained at or below 4o C, bacteriology was performed within 24 hours. Mastitic samples were collected by the herdsmen, using the same sampling procedure; these were frozen and batched each week for submission to the laboratory. Aliquots of all samples taken were stored for further use. Dry Cow Therapy Dry cow therapy was administered by the authors following collection of the ‘drying off’ samples. The teat ends were scrubbed with surgical spirit for a fourth time prior to partial insertion of the tube canula. Following treatment teats were dipped a solution containing 2800ppm available chlorine (Agrisept, UpJohn) and the cows confined to a loafing yard for at least a half hour. Dry cow products used contained cloxacillin (Orbenin Extra), cephalonium (Cepravin) or procaine penicillin G (Mylipen). Bacteriology BCVA 1998 Samples were submitted to Langford VIC for bacteriology. 10μl of secretion was innoculated onto sheep blood agar and Edward’s agar; 100μl of secretion was innoculated onto MacConkey agar to enhance the detection of Enterobacteriacae6 . Plates were incubated at 37o C and read at 24 and 48 hours. Organisms were identified and quantified using standard laboratory techniques. E. coli was identified by colony morphology, oxidase and indole tests, other Enterobacteriacae were identified using a microtube identification system (RapiD 20 E, bioMérieux). All pathogens were colony purified and stored using a commercial micro-organism storage system. Definition of an Intramammary Infection Isolation of a recognised pathogen, in pure growth was considered an intramammary infection. If a screening sample was obviously contaminated or contained >1 enterobacterial isolate the duplicate sample was submitted for bacteriological examination and intramammary infections diagnosed on the basis of re-isolation of the organism. Statistical Analysis Results were collated and analysed using Excel (Microsoft Corporation) and Epi-Info (CDC, Alanta, GA). Statistical analysis was performed using the χ2 test, the significance probability was set at < 0.05 for a two tailed test. RESULTS Mastitis Incidence The mastitis isolates from all cows on the farms involved in the current study are shown below (Table 1) and illustrated in Figure 1. Table 1: Mastitis Isolates Cases/100 cows/year* % of total isolates S. agalactiae 0 0 S. dysgalactiae 2.6 6.5 S. uberis 5.9 14.9 Coag +ve Staphs 2.9 7.3 E. coli 14.2 36.0 All Enterobacteriacae 16.2 41.0 A. pyogenes 0.9 2.3 Others 2.9 7.3 No growth 10.3 TOTAL 39.4 *Based on 261 cases of mastitis occurring between 02/04/97 and 17/02/98. Dry Period Infections These preliminary results are based on 373 cows calving between 02/04/97 and 11/11/97. Mastitis results were collated for the first 100 days of the subsequent lactation. Major pathogens The percentage of quarters infected at each sampling point with the major pathogens are shown below (Table 2) and illustrated in Figures 2 and 3. Table 2: Percentage of Quarters Infected with Major Mastitis Pathogens. Drying Off 3 Weeks Precalving 2 Weeks Precalving 1 Week Precalving Calving n 1492 271 605 700 1491 S. agalactiae 0 0 0 0 0 S. dysgalactiae 0.34 0 0.17 0.29 0.40 S. uberis 1.14 1.48 0.50 1.86 0.87 Coag +ve Staphs 1.88 1.85 1.32 1.57 1.54 E. coli 2.75 4.80 3.64 4.86 5.30 All Enterobacteriacae 3.42 6.64 6.45 7.71 6.51 96

CATTLE PRACTICE VOL 6 PART 2 Table 3: Percentage of Quarters Infected with Enterobacteriacae other than E. coli. Drying Off 3 Weeks Precalving 2 Weeks Precalving 1 Week Precalving Calving n 1492 271 605 700 1491 Enterobacter spp. 0.34 0.74 0.99 1.57 0.40 Klebsiella spp. 0 0 0 0.14 0.13 Serratia spp. 0.07 0 0.50 0 0.07 Citerobacter spp. 0 0 0.33 0.14 0.20 Proteus spp. 0.07 0.74 0.99 1.00 0.34 Salmonella spp. 0.07 0 0 0 0 Morganella spp. 0.13 0.37 0 0 0.07 Enterobacteriacae Enterobacteriacae were isolated from 190 out of 1492 quarters at one or more sampling time points (12.7%). Of these quarters, 156 (82%) were infected with E. coli. The percentage of quarters infected with Enterobacteriacae other than E. coli, at each sampling time point, is shown in Table 3 and illustrated in Figure 4. There was a significant increase in the proportion of quarters infected with Enterobacteriacae between drying off (51/1492 quarters) and one week prior to calving (54/700 quarters, p<0.05). There was a similar significant increase in E. coli infections (41/1492 quarters at drying off and 34/700 quarters at one week prior to calving, p<0.05). Of 41 quarters infected with E. coli at drying off only 3 apparently persisted until calving. New intramammary infections with E. coli were detected in 7.9% of quarters (55/700) during the dry period. New quarter infections accounted for 95% of all enterobacterial infections detected during the dry period. Mastitis in Cows on the Study 90 cases of clinical mastitis occurred in cows on the study. In 37 (41%) cases an enterobacteriacae was identified in the milk. E. coli was isolated on 28 occasions, Klebsiella spp. on 4, Serratia spp. on 4 and Enterobacter spp. on 1. Out of 700 quarters sampled during the dry period, 81 (11.6%) quarters were found to be infected with an enterobacteriacae at one or more of the sampling points. 6 (7.4%) of these infected quarters later developed mastitis due to the same species of bacteria; these quarters had 12 cases in total. Conversely 619 quarters were not infected with enterobacteriacae during the dry period and 10 (1.6%) developed enterobacterial mastitis, accounting for 11 cases in total. There was a significantly greater risk of a previously infected quarter later developing mastitis than an uninfected quarter (p<0.05). In quarters which were sampled during the dry period, a total of 23 cases of enterobacterial mastitis occurred in 16 quarters. Of these 23 cases, 12 (52%) occurred in quarters previously infected with the same species of bacteria, during the dry period. Sampled and control quarters There was not a significantly increased incidence of enterobacterial infections in the routine post-calving milk samples of quarters which were sampled during the dry period (47 out of 700 quarters) compared to those which were not (41 out of 791 quarters, p>0.2). There was also no significant increase in incidence of enterobacterial mastitis in quarters sampled during the dry period (16 out of 700 quarters) than those not (10 out of 791, p>0.2). DISCUSSION The preliminary data from this study on a small, though an arguably typical, cohort of dairy herds would tend to support the anecdotal evidence which has suggested an increase in the incidence of mastitis due to enterobacteriacae over the past few decades. The enterobacteriacae have classically been classified as opportunistic environmental pathogens capable of colonising the udder, causing a transient infection and mastitis, not uncommonly accompanied by severe systemic disease. Sub-clinical infections, although recognised9,10 have not been implicated as a significant cause of subsequent disease. The dry period has been implicated as a crucial period for aquisition of new gram -ve intramammary infections (IMIs), with >60% of all new IMIs occurring during this period7 . The initial data from this study would also support this finding as a significant rise in the level of infection was detected during the dry period. Perhaps the most compelling figure generated from this study is that more than 50% of all clinical mastitis due to enterobacteriacae arises from quarters previously infected during the dry period. This phenomenon can be ascribed to two possible scenarios; either the bacteria are residing within the udder awaiting conditions conducive to multiplication and subsequent disease, or certain quarters show an increased susceptibility to reinfection with the same pathogen. The authors feel a combination of these two scenarios is most likely. BCVA 1998 97

CATTLE PRACTICE VOL 6 PART 2 In an attempt to identify whether the same bacteria are residing within the quarter, DNA fingerprinting studies of the enterobacterial isolates are currently being undertaken. Preliminary findings from this study may have far reaching implications for our understanding of ‘environmental’ mastitis and our approach to the control of outbreaks of disease. In the past, control has centred around the management of the lactating and peri-parturient cow; these findings would suggest that significant effort also needs to be directed towards management of the dry cow. Some suggestions for possible control strategies are outlined below. Management of the Environment All too often non-lactating cows are ‘turned away’ to a second rate paddock or accommodation and, relative to the lactating cows, receive little or no attention until immediately pre-partum. There is an apparently high new infection rate from 3 weeks precalving which would imply that the whole of the dry period is critical in preventing new infections and as such needs to be addressed in minimising bacterial challenge. MANAGEMENT OF THE COW Teat dipping An obvious approach to decreasing bacterial challenge to the udder is to teat dip. Some practitioners already advocate this during the periparturient period as an aid to control of ‘down calving’ mastitis. There would obviously be significant client resistance to implementing such a regime throughout the dry period though application of a teat disinfectant with residual activity (eg Agrisept, UpJohn) via a knapsack sprayer may be a viable option. Teat Sealants Opinion is divided on the utility of these products in lactating cows11,12, though they may have a role to play in the dry cow. Dry Cow Therapy Classically dry cow therapy was developed to aid in the control of contagious pathogens and was directed against gram +ve bacteria. More recently, products have been formulated which may protect against the risk of mastitis following accidental introduction of gram -ve pathogens at the time of administration. Few products are available that provide gram -ve ‘cover’ which persists beyond the first few weeks of the dry period, and no published studies have demonstrated the in vivo efficacy of these preparations. Previous studies have been unable to satisfactorily implicate these new infections, acquired during the dry period, in subsequent mastitis7 . Also the absence of unsampled control quarters means that the role of iatrogenic introduction of infection was unknown7 . It is hoped that the design of this investigation will control for the effect of sampling quarters during the dry period and more conclusively implicate dry period infections in subsequent mastitic episodes. As discussed above one shortcoming of the outlined research is the effect of ‘opening’ quarters during the dry period, though we hope to control for this as described. Another problem arises in the definition and detection of IMIs; funding constraints prohibited the culture of duplicate samples, which will inevitably means that some infections will not have been detected, also some isolates may be false positives. In an attempt to overcome this it is hoped to develop molecular techniques for identification of pathogens which can subsequently be applied to samples maintained in storage. OUTSTANDING QUESTIONS Are there strains of E. coli more likely to persist and subsequently cause disease? What is the role played by minor pathogens in enterobacterial mastitis - do they have a protective effect? Why do only some quarters develop IMI’s? Why do only some cows with IMI’s subsequently develop mastitis? Why is there a delay between aquistion of infection and subsequent disease? FUTURE WORK Sub-species typing of enterobacteriacae. Intervention studies to attempt to decrease the incidence of new E. coli IMI’s in the dry period and mastitis incidence in the subsequent lactation. Areas worthy of investigation include dry cow therapy (type), teat sealants and propholactic teat dipping. Another interesting area of study would be the feasibility of seasonal, selective use of dry cow in low somatic cell count cows. ACKNOWLEDGEMENTS The authors gratefully acknowledge the financial support of Leo Laboratories without which this study could not have been undertaken; The Langford Veterinary Investigation Centre for undertaking bacteriology, the patience of farmers for allowing weekly access to their herds and UpJohn for supply of Agrisept Tablets. A. J. Bradley is supported by a stipend paid by the Wellcome Trust. REFERENCES 1. BOOTH J.M. (1997) Progress in Mastitis Control - An Evolving Problem. Proceedings of the British Mastitis Conference. 3-9. 2. JONES T.O. (1990) Escherichia coli Mastitis in Dairy Cattle - A review of the Literature. Veterinary Bulletin 60 205-231. 3. MURPHY J.M. and HANSON J.J. (1943) Infection of the Bovine Udder with Coliform Bacteria. Cornell Vet 33 61-77. 4. EBERHART R.J. and BUCKALEW J.M. (1977) Intramammary Infections in a Dairy Herd with a Low Incidence of Streptococcus BCVA 1998 98

CATTLE PRACTICE VOL 6 PART 2 agalactia and Staphylococcus aureus Infections. JAVMA 171 630- 634. 5. OLIVER S.P. and MITCHELL B.A. (1983) Susceptibility of Bovine Mammary Gland to Infections During the Dry Period. J Dairy Sci 66 1162-1166. 6. SMITH K.L., TODHUNTER D.A. and SCHOENBERGER P.S. (1985) Environmental Pathogens and Intramammary Infection During the Dry Period. J Dairy Sci 68 402-417. 7. TODHUNTER D.A., SMITH K.L., HOGAN J.S. and SCHOENBERGER P.S. (1991) Gram-negative bacterial infections of the mammary gland in cows. Am J Vet Res 52 184-188. 8. McDONALD J.S. and ANDERSON A.J. (1981) Experimental Intramammary Infection of the Dairy Cow with Escherichia coli During the Non-lactating Period. Am J Vet Res 42 229-231. 9. LAM T.J.G.M., LIPMAN L.J.A., SCHUKKEN Y.H., GAASTRA W. and BRAND A. (1996) Epidemiological Characteristics of Bovine Clinical Mastitis caused by Escherichia coli and Staphylococcus aureus studied by DNA fingerprinting. Am J Vet Res 57 39-42. 10. HILL A.W. and SHEARS A.L. (1979) Recurrent Coliform Mastitis in the Dairy Cow. Vet Rec 105 299-301. 11. FARNSWORTH R.J., WYMAN L. AND HAWKINSON R. (1980) Use of a Teat Sealer for Prevention of Intramammary Infections in Lactating Cows. JAVMA 177 441-444. 12. MCAURTHUR B.J., FAIRCHILD T.P. AND MOORE J.J. (1984) Efficacy of a Latex Teat Sealer. J Dairy Sci. 67 1331- 1335. BCVA 1998 99

CATTLE PRACTICE VOL 6 PART 2 Figure 1: Relative Incidence of Mastitis Pathogens S. dysgalactia S. uberis Coag +ve Staphs E. coli Other Enterobacteriacae A. pyogenes Others No growth Figure 2: Percentage of Quarters Infected at Each Sampling Time Point - Non Enterobacteriacae 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Drying Off 3 Weeks Precalving 2 Weeks Precalving 1 Week Precalving Calving % S. agalactia S. dysgalactia S. uberis Coag +ve Staphs BCVA 1998 100

CATTLE PRACTICE VOL 6 PART 2 Figure 3: Percentage of Quarters Infected at Each Sampling Time Point - Enterobacteriacae 0 1 2 3 4 5 6 7 8 Drying Off 3 Weeks Precalving 2 Weeks Precalving 1 Week Precalving Calving % E. coli All Enterobacteriacae Figure 4: Percentage of Quarters Infected at Each Sampling Time Point - Enterobacteriacae 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 Drying Off 3 Weeks Precalving 2 Weeks Precalving 1 Week Precalving Calving % Enterobacter spp. Klebsiella spp. Serratia spp. Citerobacter spp. Proteus spp. Others BCVA 1998 101

CATTLE PRACTICE VOL 6 PART 2 BCVA 1998 102

CATTLE PRACTICE VOL 6 PART 2 A Case Control Study of Toxic Mastitis in Dairy Cows Green L.E.**, Green M.J.*, N.A Tadich, A Carey, R. Porter, J. Ridley., University of Bristol, Dept. of Clinical Veterinary Science, Langford House, Langford, Bristol BS40 5DU. *Orchard Veterinary Group, Wirral Park, Glastonbury, Somerset. **Author for correspondence. ABSTRACT A retrospective case-control study of farm level risk factors for toxic mastitis was carried out in November and December 1996. Twenty six farms from Mid-Somerset were visited: 13 case farms (had had a cow with toxic mastitis in the previous year) and 13 geographically matched controls (no case of toxic mastitis). The farmers were interviewed and the buildings were examined. Information was collected on the type and quality of housing, usual milking routines, milk quality and mastitis prevalence in the previous year. All the data were collected onto pretested recording sheets and loaded into a database. Simple and complex analysis were done. The following variables were significantly (P<0.05) associated with an increased risk of toxic mastitis in the simple analysis: housing cows in October rather than November, a low number of calving boxes per cow, a high proportion of cows with intermediate body condition and low HBMSCC. In the final logistic low HBMSCC and a high proportion of cows with intermediate body condition remained significant. The authors conclude that, despite the small sample size, the results of this study are consistent, plausible and support the information from previous experimental and observational studies about the role of somatic cell counts in toxic mastitis. KEYWORDS: Toxic mastitis, case-control study, HBMSCC. INTRODUCTION Environmental mastitis and contagious mastitis are terms used to classify the primary pathogens causing intramammary infection (Smith and Hogan, 1993). Environmental mastitis affects all dairy herds and its relative importance in the herd increases as contagious mastitis is brought under control (Erskine and others, 1988, Schukken and others, 1991). In contrast to contagious mastitis, environmental mastitis is primarily associated with clinical mastitis; approximately 80% to 90% of gram negative intramammary infections result in clinical mastitis (Smith and Hogan, 1993; Harmon, 1994). The economic impact of environmental mastitis is primarily as a result of the clinical effects of the disease. The majority of the costs are due to loss of milk production, discarded milk, death or premature culling (Dobbins, 1977; Smith and Hogan, 1993). Coliform mastitis is an environmental mastitis and the term includes mastitis caused by Escherichia coli, Klebsiella spp and Enterobacter spp. (Radostits and others, 1994; Smith and Hogan, 1993). Anderson (1989) has classified the manifestation of the disease into three types 1 (local), 2 (systemic) and 3 (toxic). About 10% to 15% of clinical coliform mastitis infections are thought to result in toxic (type 3) mastitis. In recent work, Green and others, (1997) reported that approximately 50% of the cows affected with toxic mastitis survive, independent of the treatment used, indicating that reliable preventive measures are needed. The disease occurs most commonly within a week of calving in well managed herds with seasonal calving patterns (Smith and Hogan, 1993; Radostits and others, 1994). Risk factors for environmental coliform mastitis include infections acquired during the dry period, (Erskine and others, 1988) continuous housing of the cows (Smith and Hogan, 1993), contamination of the udder between milking (Radostits and others, 1994); contaminated bedding, damage of the teats, leaking of milk (Schukken and others, 1991), low vitamin E/Selenium status (Ndiweni and Finch, 1996) and low somatic cell counts. (Booth, 1993; Kehrli and Shuster, 1994; Green and others, 1996). Environmental hygiene, high milk yield and diet have been recognised as risk factors which may precipitate the onset of all forms of clinical mastitis (Barnouin and others, 1986; Shukken and others, 1991). However, there is controversy about the possible role of low somatic cell counts as a risk factor for coliform mastitis (Berry 1994; Logan, 1996). This communication reports the results of a retrospective case-control study carried out to investigate the relationship between farm level risk factors and toxic mastitis in dairy cows. MATERIALS AND METHODS Data were collected in November and December 1996 from 26 dairy herds in Mid-Somerset. There were 13 herds which had had a cow with toxic mastitis in the previous 12 months (case farms) and 13 which had not (control farms). A clinical case of toxic mastitis was defined as a cow with an apparent udder inflammation and abnormal secretion; BCVA 1998 103

CATTLE PRACTICE VOL 6 PART 2 recumbent or severely weak; and displaying clinical signs of endotoxic shock (sunken eye, raised pulse, raised respiratory rate), (Green and others, 1996). Two questionnaires were designed, both to be completed on the farm. One was an interview questionnaire filled in using the responses of the farmer/head cow person and the other was completed by three members of the research team according to observations and measurements taken on the farm. These questionnaires were tested at the university farm before the study started to test their suitability and standardise the criteria and methods used. The first questionnaire was laid out with special introductory remarks to each group of questions being asked. The main subject areas covered by the questions were: number and breed of cows in the herd, milk production and quality, mastitis levels, housing, calving management, dry cow management, the milking parlour and milking routine. The second questionnaire collected data on the general hygiene of the farm (yards, gateways, etc.) scored into one of the three categories: clean, average or dirty. To avoid erroneous observations caused by disturbance of the cows a behavioural assessment was performed immediately the research team entered the buildings where they were housed. For this assessment the number of cows observed in each of the following positions was recorded : lying and standing in the cubicles/kennels, lying and standing in the passageways, lying and standing half in half out of the cubicles/kennels. These figures were used to calculate the proportion of the herd in each category of the total number of cows observed. The body condition score of the milking cows was assessed visually and recorded as percentage of fat, intermediate or thin cows in the herd. The number of buildings and number of cubicles/kennels in each building were recorded and dimensions of the house, cubicles/kennels: width, length, kerb height, height of side and height of drop, were measured. The type of bedding used in the milking herd housing was also noted, together with its depth. The general hygiene of the cubicles/kennels/straw yards was classified as : clean, average or dirty. A "squelch test" was carried out on the bedding in each building used to house the milking herd, the result was considered as positive if a "squelch" sound was heard when the bed was pressed down under foot. The passageways and cubicles/kennels were classified as dry, moist, wet or wet with pooling water. The number of calving boxes on the farm was recorded and whether the dry cows were housed as groups, individuals or both. In both the calving boxes and dry cow housing, the following assessments were made: type and depth of bedding, general hygiene, "squelch test" and wetness of the floor. In the milking parlour the general hygiene and wetness of floor was assessed. The liner usage at each farm was calculated using the formula (Blowey and Edmondson, 1995): Liner usage = Number of cows in herd x 2* x liner life in days number of clusters in parlour * or number of milkings / day Records of the number of intramammary milking cow tubes purchased between November '95 and October '96 were obtained from the Orchard Veterinary Practice. The geometric mean herd bulk milk somatic cell count (HBMSCC) for each of the months (November '95 to October '96) were obtained from the farm's monthly milk statement. Data were entered onto a spread sheet (Microsoft© Excel) and then imported into Epi Info 6 (Dean and others, 1991) for analysis. Differences between proportions were investigated using Yates corrected chi square tests and odds ratios (ORs) with Cornfield 95% confidence intervals (95% CIs). Quantitative variables were categorised and analysed using chi square test (2 x k contingency tables). To adjust for confounding, the relationship between the dependent variable "case" and in-dependent variables with P <0.25 in the bivariate analysis were investigated further using forward stepwise logistic regression analysis (EGRET, 1991). Final models included variables that resulted in a reduction in the likelihood ratio statistic (LRS) with P ≤ 0.05. RESULTS The mean herd size was 140 for the case and 87 for the control farms (Table 1). Twenty three herds had Friesian x Holstein cows, eleven pedigree Friesian cows, nine Holstein and four herds had some cows of another breed (Jersey, Brown Swiss, Ayrshire). The average milk yield ranged from 4000 to 7828 litres per cow/year with a mean value of 6000. The mean milk production/year for case farms was 6168 L and for the control farms 5936 L . Herds were housed for approximately six months of the year (Table 2), there was a significant increase in the risk of mastitis and housing cows in October (OR 6.85; CI 1.02 - 62.6) rather than November. An intermediate BCS was associated with an increase risk of toxic mastitis (OR 1.3; CI 0.949 - 1.83). Three farmers did not have their milking machine serviced. Of the farmers that did, three had it serviced every two years and 20 at least once a year. Only three farmers changed their teat cup liners at the rate currently recommended, every 2500 milkings (Blowey and Edmondson, 1995). BCVA 1998 104

CATTLE PRACTICE VOL 6 PART 2 Table 1. Mean, standard error and significance of continuous variables by case and control farms Variables Case Farms mean (s.e.) Control Farms mean (s.e.) P Herd size 140 (23.1) 87 (8.0) P < 0.07 Milk yield per annum (litres) 6168 (210) 5936 (247) P < 0.49 Mean number of heifers / herd 32.8 (7.6) 14.8 (3.1) P < 0.10 No. cows died with mastitis (1995-1996) 0.6 (0.24) 0.07 (0.07) P ≤ 0.05 No. milkings per cup liner 7129 (848) 5164 (471) P ≤ 0.05 No. mastitis tubes used (1995-1996) 237 (55) 118 (31.2) P ≤ 0.06 Number of cows / calving box 73 (19.7) 38 (9.9) P ≤ 0.02 Intermediate BCS (% of herd) 87.3 (1.43) 75.1 (4.38) P ≤ 0.01 High BCS (% of herd) 6.9± 1.3 14.5 ± 4.8 P ≤ 0.26 Low BCS (% of herd) 5.8 ± 1.12 10.3 ± 3.22 P ≤ 0.23 Height of the cubicle side (m) 0.46 ± 0.056 0.37 ± 0.02 P ≤ 0.02 P = probability value Table 2. OR, 95% confidence limits and significance of discrete variables in case and control farms Exposures OR 95% CI P Cows housed in October = Yes 6.85 1.02 - 62.6 P ≤0.04 Cows housed in November =Yes 0.13 0.01 - 0.94 P ≤ 0.04 Separation of early lactation cows = Yes 9.39 0.87 - 511 P ≤ 0.07 Sprinkle slaked lime onto bedding = Yes 9.39 0.87 - 511 P ≤ 0.07 Adequate number of cubicles = Yes 0.23 0.02 - 1.73 P ≤ 0.19 Milking machine serviced once a year =Yes 0.23 0.02 - 1.83 P ≤ 0.19 Milking machine serviced every two years =Yes 5.24 0.44 - 294 P ≤ 0.17 Use of kennels for dairy herd = Yes 0.00 0.00 - 2.30 P ≤ 0.22 Parlour hygiene clean = Yes 6.95 0.61 - 383 P ≤ 0.16 Parlour hygiene average = Yes 0.17 0.01 - 1.27 P ≤ 0.09 Cubicles / kennels / straw yards dirty = Yes 4.43 0.58 - 57.14 P ≤ 0.20 P = probability value Table 3. Mean, standard error & significance values of the HBMSCCs of case & control farms by month Number of farms Mean (1000s/ml) Standard error P November ’95 CASE CONTROL 11 10 161 243 21.17 18.86 0.08 December CASE CONTROL 11 11 154 205 21.87 20.81 0.10 January ‘96 CASE CONTROL 11 11 138 185 13.65 16.34 0.03 February CASE CONTROL 11 11 126 181 10.75 18.03 0.01 March CASE CONTROL 11 11 135 183 14.19 20.57 0.06 April CASE CONTROL 11 11 135 193 12.65 20.78 0.02 May CASE CONTROL 11 12 144 207 24.25 14.30 0.03 June CASE CONTROL 11 12 156 209 17.55 28.38 0.13 July CASE CONTROL 11 11 157 192 19.57 30.36 0.33 August CASE CONTROL 11 12 139 201 10.42 22.57 0.02 September CASE CONTROL 11 12 134 205 9.57 21.41 0.006 October CASE CONTROL 11 12 130 196 10.45 19.04 0.007 P = probability value BCVA 1998 105

CATTLE PRACTICE VOL 6 PART 2 Table 4. Logistic regression analysis of variables and toxic mastitis Variable Coefficient Odds ratio 95% CI LRS P Constant -14.88 September HBMSCC .0539 0.947 0.893 - 1.00 9.484 0.002 Intermediate BCS .2759 1.318 0.949 - 1.83 7.916 0.005 95% CI = 95percent confidence intervals LRS = Likelihood ratio statistic P = probability value Twenty three farmers did some form of udder preparation, three did a wet preparation, 16 dry wiped only and seven did a wet followed by dry preparation. Eighteen farmers checked cows for mastitis before milking their herd. All the farmers (26) used post milking disinfection of the udder: eleven used a dip, 14 a spray and one used a commercial ointment. Eighteen farmers kept records of mastitis cases. The number of cases per month ranged from eight herds having one case per month to two herds having eight cases per month. All the farmers used dry cow therapy on all their cows. Between three and five farmers failed to provide data of their HBMSCC for one or more months. HBMSCCs were significantly lower in case farms for 8 out of 12 months of the year. September was the most strongly associated with a case of toxic mastitis (P<0.006) in the bivariate analysis. In the remaining four months there was no significant association between the HBMSCC and toxic mastitis. These figures were strongly correlated within farms (Table 3). When logistic regression was performed, September HBMSCC and intermediate BCS stayed in the final model (Table 4). There was a decreased odds of mastitis in herds as HBMSCC increased ( OR=0.9) and an increased odds as the proportion of cows with intermediate BCS increased (OR=1.3). DISCUSSION Though several studies have shown an association between the incidence of environmental clinical mastitis and low somatic cell counts in herds (Kehrli and Shuster, 1994; Miltenburg and others, 1996), as far as the authors are aware, the relationship between low somatic cell count and toxic mastitis has not been reported before, except for a previous work of Green and others, (1996). The possible protective role of increased SCC has been suggested by Schalm and others (1964) and Carroll and others (1973) who reported that quarters with somatic cell counts of 200,000 to 350,000 gave partial protection from coliform infections while counts of 500,000 or more gave complete protection. Schukken and others (1989a) found a large variation in the incidence of clinical mastitis in herds with SCC < 150,000 cells/ml, and suggested that housing and management have a considerable influence on the incidence of the disease. One explanation for the association between low SCC and environmental mastitis has been postulated by (Schukken and others, 1989b; Lam, 1996). These authors suggest that post milking teat disinfection kills the minor pathogens found in the teat canal. These minor pathogens increase SCC slightly, possibly enough to provide protection against environmental organisms. A recent experimental investigation by Shuster and others (1996), in periparturient and mid lactation cows found that periparturient cows with significantly lower SCC were more susceptible to coliform mastitis despite a greater recruitment of leucocytes than mid lactation cows during the first 12 hours after bacterial inoculation. The higher susceptibility was attributed to an inability to reduce bacterial growth during the first few hours after bacterial inoculation because of an absolute low SCC before the challenge and poor up regulation of CD18 expression by stimulated neutrophils. Previously, susceptibility to coliform mastitis had been associated with impaired leukocyte recruitment to the infected gland in cows with low SCC (Hill, 1981). The results of this study agree with previous papers by Erskine and others (1988), Kehrli and Shuster (1994), and Miltenburg and others, (1996) which reported that clinical mastitis caused by coliform pathogens are a major problem in herds with low SCC. They also support the results of Green and others (1996) who found that herds with cases of toxic mastitis were more likely to have low HBMSCCs than control herds, suggesting that low HBMSCCs increased the herd susceptibility to this type of mastitis. Body condition score is used as subjective method to estimate the quantity of fat and muscle in specific regions of the animal. Research and field studies suggest that changes in BCS influence health and productivity in dairy cows (Gearhart and others, 1990). The nutrition of the herd has a role in the occurrence of clinical mastitis. Barnouin and others (1986) found that a positive balance of energy/ protein was risk factor for clinical mastitis, probably due to an increase in the occurrence of mammary gland oedemas and fatty liver disease. It has been suggested that fatty liver and vitamin E and Selenium deficiency have an immunosupressive effect (Reid, 1981; Erskine and others, 1988; Ndiweni and Finch,1996). In the present study the increase in the percentage of cows with condition score intermediate BCVA 1998 106

CATTLE PRACTICE VOL 6 PART 2 (2-3) was associated with an increase risk of toxic mastitis, the explanation for this is unclear, but it may be that keeping cows in adequate body condition is an indication of a well managed herd as is low HBMSCC. In the present study milk yields in case farms were higher than, but not significantly different from, control farms. In most observational studies herd milk production has been positively correlated with the rate of clinical mastitis (Barnouin and others, 1986, Schukken and others, 1990; Grindal and Hillerton, 1991; Shook, 1993). In the bivariate study, housing the cows in October rather than November increased the risk of toxic mastitis. These cows were likely to be housed for a longer period than cows housed in November. The period of housing has been reported as a risk factor for clinical mastitis, cows that are maintained at high stocking density are more susceptible to trauma and expose the mammary gland to a highly contaminated environment (Carrol, 1977; Barnouin and others, 1986). However, this variable did not remain significant in the multivariate analysis. CONCLUSION Epidemiological studies are generally not conclusive in terms of causal relations between observed risk and disease. Instead, studies offer hypotheses that may be elaborated in further clinical or experimental studies (Schukken and others, 1990). The observed statistically significant risk factors associated with toxic mastitis in this study are not necessarily causal. It can also be argued that the small number of farms provided insufficient power to detect associations between some of the other variables measured. However, despite the small sample size, the results of this study that associate low HBMSCCs with toxic mastitis are consistent, plausible and support the information from previous experimental and observational studies about the role of somatic cell counts in toxic mastitis. REFERENCES ANDERSON K.L. (1989) Therapy for acute coliform mastitis. Compendium for Continuing Education: Food Animal, 11 1125 - 1134. BARNOUIN J.C., M. FAYET, M .JAY, M. BROCHART & B FAYE (1986) Enquete eco - pathologique continue: Facteurs de risque des mammites de la vache laitiere II. Analyses complementaires sur donnees individuelles et d'elevage. Canadian Vet.Journal, 27 173-184. BERRY E.A., (1994) Mastitis incidence in low cell count herds. Veterinary Record, 135 479-480. BLOWEY R&P Edmondson. (1995) Mastitis Control in Dairy Herds. Farming Press, UK. 56-57. BOOTH, J.M. (1993) Mastitis in the 1990s. Cattle Practice, 1 125-131. CARROLL E.J., N.C. JAIN., O.W. SCHALM & J. LASMANIS (1973) Experimentally induced coliform mastitis: Inoculation of udders with serum - sensitive and serum resistant organism. American Journal of Veterinary Research, 34 1143-1146. CARROL E.J., (1977) Environmental factors in bovine mastitis. Journal of the American Veterinary Medical Association, 170 1143-1148. DEAN A.D. J.A. DEAN., A.H. BURTON & R.C. DICKERAND (1991) Epi Info 6.03. USD, Incorporated, Stone Mountain, Georgia, USA. DOBBINS N. (1977) Mastitis Losses. Journal of American Vet Medical Assoc., 170 1129-1133. EGRET (1991) Epidemiological Graphics Estimation and Testing Package. Cytel Statistical Software. USA. ERSKINE R.J., R.J. EBERHART L.J. HUTCHINSON S.B. SPENCER & M.A. CAMPBELL (1988) Incidence and types of clinical mastitis in dairy herds with high and low somatic cell counts. Journal of the American Veterinary Medical Association, 192 761-765. GEARHART M.A., C.R. CURTIS, H.N. ERB, R.D. SMITH, C.J. SNIFFEN, L.E. CHASE & M.D. COOPER. (1990) Relationship of changes in condition score to cow health in Holsteins. Journal of Dairy Science, 73 3132-3140. GREEN M.J., L.E. GREEN & P.J. CRIPPS (1996) Low bulk milk somatic cell counts and endotoxic - associated (toxic) mastitis. Veterinary Record, 138 305 - 306. GREEN M.J., L.E. GREEN & P.J.CRIPPS (1997) Comparison of fluid and flunixin meglumine therapy in combination and individually in the treatment of toxic mastitis. Vet. Record, 140 149- 152. GRINDAL R.J. & J.E. HILLERTON (1991) Influence of milk flow rate on new intramammary infection in dairy cows. Journal of Dairy Research, 58 263-268. HARMON R.J. (1994) Symposium: Mastitis and genetic evaluation for somatic cell count : Physiology of mastitis and factors affecting somatic cell counts. Journal of Dairy Science, 77 2103- 2112. HILL A.W. (1981) Factors influencing the outcome of Escherichia coli mastitis in the dairy cow. Research in Veterinary Science, 31 107-112. KEHRLI M.E. & D.E. SHUSTER (1994) Factors affecting milk somatic cells and their role in health of the bovine mammary gland. Journal of Dairy Science, 77 619-627. LAM T., (1996) Dynamics of bovine mastitis: A field study in low somatic cell count herds. Ph.D. Thesis, University of Utrech, Netherlands, pp. 132 - 145. LOGAN E.F. (1996) Low bulk milk SCC and toxic mastitis. Veterinary Record, 138 424. MILTENBURG D.J., D. de LANGE., A.P.P. CRAUWELS., J.H. BONGERS., M.J.M. TIELEN., Y.H. SCHUKKEN & A.R.W. ELBERS (1996) Incidence of clinical mastitis in a random sample of dairy herds in the southern Netherlands. Veterinary Record, 139 204-207. NDIWENI N & J.M. FINCH (1996) Effects of in vitro supplementation with α - tocopherol and selenium on bovine neutrophil functions: implications for resistance to mastitis. Veterinary Immunology and Immunopathology, 51 67-78. RADOSTITS O.M., D.C. BLOOD & C.C. GAY. (1994) Veterinary Medicine, 8th Edition, Bailliere Tindall, London, UK. pp 584-593. REID IM (1981). British Cattle Veterinary Association Proceedings 1979-1980. Pp 65-72 SCHALM O.W., J. LASMANIS & E.J. CARROLL (1964) Effects of pre-existing leukocytosis on experimental coliform (Aerobacter Aerogenes) mastitis in cattle. American Journal Veterinary Research, 25 83 - 89. SCHUKKEN Y.H., F.J. GROMMERS., D. VAN DE GEER & A. BRAND. (1989a) Incidence of clinical mastitis on farms with low somatic cell counts in bulk milk. Veterinary Record, 125 60-63. SCHUKKEN Y.H., D. VAN DE GEER., F.J. GROMMERS., J.A.H. SMITH & A. BRAND (1989b) Intramammary infections and risk factors for clinical mastitis in herds with low somatic cell counts in bulk milk. Veterinary Record, 125, 393- 396. SCHUKKEN Y.H., F.J. GROMMERS., D.VAN DE GEER., H.N.ERB & A. BRAND (1990) Risk factors for clinical mastitis in herds with a low bulk milk somatic cell count. 1. Data and risk factors for all cases. Journal of Dairy Science, 73 3463- 3471. BCVA 1998 107

CATTLE PRACTICE VOL 6 PART 2 SCHUKKEN Y.H., F.J. GROMMERS., D. VAN DE GEER., H.N. ERB & A. BRAND. (1991) Risk factors for clinical mastitis in herds with a low bulk milk somatic cell count. 2. Risk factors for Escherichia coli and Staphylococcus aureus. Journal of Dairy Science, 74 826-832. SHOOK G.E. (1993) Genetic improvement of mastitis through selection on somatic cell count. Veterinary Clinics of North America: Food Animal Practice, 9 563-577. SHUSTER D.E., E.K. LEE & M.E. KEHRLI (1996) Bacterial growth, inflammatory cytokine production and neutrophil recruitment during coliform mastitis in cows within ten days after calving, compared with cows at mid lactation. American Journal of Veterinary Research, 11 1569 -1575. SMITH K.L & J.S. HOGAN (1993) Environmental mastitis. Veterinary Clinics of North America: Food Animal Practice, 9 489-498. BCVA 1998 108

CATTLE PRACTICE VOL 6 PART 2 Robotic Milking at Bridgets Bull R., ADAS Bridgets, Martyr Worthy, Winchester, SO21 1AP INTRODUCTION During 1996 we were approached by Liberty Dairy Systems to install a Robotic Milker at ADAS Bridgets for a two year period to evaluate the machine and use it for any funded research we could obtain for it. The benefits of Robotic Milking over conventionally milking cows are stated as being:- • Higher production per cow • Assisting in meeting the changing demands of Environmental Legislation • Attachment without the farmer • Twenty four hour per day milking • Add lib milking on cow demand • Lower production cost per litre • Improved milk quality • Improved cow welfare and longer productive life • Social improvement for the farmer • Improved labour conditions • Individual management of the cow • A move from muscle to brains • Assist in the move from production minded to market orientated farming PARLOUR OPERATION Cow enters unit and is identified, if she has not been milked within the last five hours (this time can be varied) attachment takes place, if she has been milked within the last five hours she is released unmilked. Pre selection can be fitted so that cows are identified before entry into the unit and those recently milked can be diverted past the box rather than through it. Once the cow is in the box and requires milking, the feed box is adjusted forward or backward for that individual cow to ensure she is standing in the correct position for milking. The robot is activated and collects the milking cluster, using ultra sound it locates the teats and attaches the clusters. Teats are pre washed using water which is dumped along with foremilk. The premilking and milking process is carried out on a quarter basis with each being milked separately. Conductivity of each quarter is recorded during milking, with the computer building up a rolling picture of normal conductivity levels for the cow as well as milk yield. If conductivity deviates by more than 18% it is flagged to the milker. If once the cow has been milked she has not given 90% of her expected yield she is re-attached to ensure she has been milked out correctly, if after this daily rolling yield is down by 20% this again is flagged to the herdman. The robot will attempt to attach the unit three times if unsuccessful the cow is released and diverted into the collection area for another attempt. Once milking is complete the cow is released, no teat disinfection takes place. The placement of the box and cow routing is critical to get the cows to go through the system (see Diagram 1.). Our system is designed so that cows have to go through the system to eat silage or complete diet in front of the feed barrier, most systems work along these lines. Milking started through the unit in early April 1997 with 20 purchased cows which had been managed on a high yielding system and milked 3 x daily. The type of cow which suits the system is very important as udder and teat uniformity are critical. Udders must be well supported and not too low, teats must be positioned well and preferably pointing straight down. Teats which are too close or too wide do not work particularly well, nor do teats which are too long and too thin. One was rejected due to teat conformation. The initial 20 purchased cows adapted amazingly well to the system and after two days of pushing cows through the unit and manually attaching the clawpiece the robot was switched on and the first fully automated milkings started to take place. By the end of the second week all cows except one which proved to be too much of a problem and was lost from the system were being milked by the Robot. Once the system had been operational for a month a further 20 mid lactation cows were introduced to the system in two lots of 10 to give a total of 40 cows on the one box system. The introduction of these additional cows proved to be a major upset, as well as having to train these cows to the system they upset the now established routine of the 20 original cows. As well as having to train the new cows the operators now had to spend time pushing a number of the original cows through the system, it took a lot longer to get the 20 new cows introduced and all cows back into a routine. COW FLOW THROUGH THE SYSTEM Once settled into a routine cows visited the system on average of 5.8 times per day in early lactation and were milked on average 2.9 times per day, with maximum milking intervals set at 14 hours and minimum set at 5 hours. Visit were well spread through the day and night. See Table 1. BCVA 1998 109

CATTLE PRACTICE VOL 6 PART 2 Diagram 1. ADAS Bridgets Robotic Unit 3 2 .1 m FEED PASSAGE 1 2 .2 5 m D A IR Y 4 .0 m R O B O T IC U N IT 1 4 .7 5 m 5 .0 m 9 .0 m Table 1: ADAS Bridgets Robotic Unit Time 0 – 4 4- 8 8- 12 12- 16 16- 20 20- 24 No of visits in time period 18 20 26 20 28 25 Major influences on cow movement through the system were:- Milk yield The higher the milk yield the more often they visited the unit, the lower the milk yield the less often they visited the unit. This had a large effect in late lactation. Temperature At very high temperatures cows were reluctant to move through the milking system. This was somewhat overcome by the use of grazing to encourage the cows through the box at each end of the day. BCVA 1998 Individual cows were probably the largest variable ranging from a reluctant 2 visits to an enthusiastic 15 plus. Using concentrates as bait to encourage cows in when fed a complete diet had little or no effect on cow throughput. Concentrates were finally fed at very low levels to cows in the unit whilst being milked as most was being rejected, reverting to a maintenance plus production complete diet of the feed barrier. Factors effecting success of attachment Udder and teat conformation, which is critical with the machine preferring a well placed upright teat firmly full of milk. Milk yield, the lower the yield the less reliable attachment was due to the fullness of the udder and teats. Milk let down response having an effect in two ways, one with initial attachment and if it was delayed too long reattachment as the ACR would be activated before the cow had started to milk and so reattachment would be required. The effects of late lactation and lower milk yields on throughput and attachment of cows did have a compounding effect, even though milking interval was extended and number of milkings per day was reduced on lower yielding cows. Milk yields were not affected by robotic milking when animals were introduced to it and very high yields of high quality were consistently achieved using robotic milker. See Table 2. Table 2: ADAS Bridgets Robotic Unit Production Stage Of Lactation 0-100 100-200 200-300 Cows Kg 79.0 48.0 36.0 Fat% 3.94 3.69 4.74 Protein % 2.62 3.10 3.26 Heifers Kg 51.0 33.0 Fat% 4.40 3.96 Protein% 3.20 3.26 Milk hygiene did prove to be a problem at the start of the machines operation but was not due solely to the machine. Satisfactory hygiene results were consistently achieved once the critical teething problems were sorted out, but this is an area which can still be improved. Health and Welfare The robot was without doubt the quietest area of the entire cow unit, not unlike walking a cowshed at night. There is no forced cow movement or traffic with everything happening at the cows pace. Clinical lameness was in line with the rest of the herd. Clinical mastitis was the equivalent to 41 cases per 100 cows per year, very similar to the level found in the main herd over the same time period. Once a high conductivity had been flagged to indicate a possibility of mastitis the cow was hand stripped daily to check for clinical signs. No other major health disorders or effects of the Robot were apparent, generally the cows accepted the concept of Robotic milking very well. A greater amount of time did have to be spent on observing stock to ensure they were fit and well on a daily basis. Labour requirement for the successful running of the machine is high and needs to be of a higher than average skill level. One side would argue that the 110

CATTLE PRACTICE VOL 6 PART 2 job never has to be done, the other would argue the job is never done. The lack of an automated sampler did make monthly milk sampling a rather tedious job. CONCLUSIONS In conclusion the robotic milker does work if you have the correct attitude and resource to make it work. It requires a totally different approach to conventionally milking cows as far as the milker is concerned. It appears to be better suited to high yielding cows with very good udder conformation. As ever the cow is flexible enough to adapt to the system quickly. It is very questionable as to whether or not labour would be reduced or whether the social life of the farmer is going to be improved by carrying a pager with him, or having a system which has to run 24 hours per day. There are still a lot of refinements to make but the system does work. The system is not low cost. BCVA 1998 111

CATTLE PRACTICE VOL 6 PART 2 BCVA 1998 112

CATTLE PRACTICE VOL 6 PART 1 Should we use Antimicrobials for Treatment of Coliform Mastitis in Dairy Cows ? Shpigel N.Y., The Koret School of Veterinary Medicine, The Hebrew University of Jerusalem, P.O.B 12, Rehovot, Israel 76100. ABSTRACT Coliform mastitis is the most common severe septic condition in dairy cows. Despite the infectious etiology of the disease treatment with antimicrobials is highly controversial. Three recent studies are presented dealing with the occurrence of bacteremia and the efficacy of antimicrobial therapy in experimentally induced and field cases of coliform mastitis. The paper provides some supportive data in favour of rational use of systemic antimicrobials in the treatment of severe coliform mastitis. Treatment should be based on the following principles:1) Routine milk culture and antibiograms of clinical cases; 2) Early diagnosis and treatment; 3) Pharmacokinetically adequate antimicrobials; 4) Short duration of treatment with an adequate dose. KEYWORDS: Cattle, Mastitis, Treatment. INTRODUCTION Despite worldwide efforts, mastitis has remained the most economically important disease in dairy cattle. The implementation of effective preventive and control measures resulted in marked reduction and even eradication of Staphylococcus aureus and Streptococcus agalactiae subclinical mastitis in some herds. However, in these well managed herds, acute clinical mastitis caused by environmental pathogens is becoming a major problem. Coliforms are probably the major etiologic organisms of severe clinical mastitis in most dairies (Shpigel et al 1998). Although various preventive measures and management practices are effective there is still a great need for effective therapeutic measures for acute and peracute clinical mastitis. It is generally accepted that coliforms infect the bovine udder via the teat canal. The required infective dose is very low, as little as 50 cfus infused into the teat canal consistently elicit severe clinical mastitis (Frost et al 1982). The organisms probably do not adhere to the udder epithelium and undergo rapid proliferation. Rapid bacterial proliferation is prerequisite to disease induction and positively correlated with severity of clinical signs and recovery rate (Lohuis et al 1990). Bacterial counts in the milk of cows experimentally infected with coliform mastitis peaks between 12 and 24 hours after infection (Vandeputte-Van Messom et al 1993, Van Werven et al 1997). In severe cases bacterial counts probably peak later than in mildly affected mastitic cows. Clearly detectable clinical signs appear within 6 to 8 hours after infection, attaining maximum severity 2 to 6 hours thereafter (Vandeputte-Van Messom et al 1993, Shpigel et al 1996, Van Werven et al 1997). Clinical signs probably peak earlier and bacterial counts peak later and sustain longer in severely affected cows. Therefore field cases of coliform mastitis can and should be detected by herdsmen at an early stage of the disease, clearly well before the peak and natural drop of bacterial counts in the milk. Proliferating and phagocytosed organisms are known to release endotoxins and to elicit a local and systemic inflammatory response. The pathogenesis of coliform mastitis cannot be solely attributed to the local or systemic effects of endotoxins released by the organisms in the udder. The disease induced by intramammary infusion of various doses of endotoxins is considerably different from field cases or experimentally induced coliform mastitis. Endotoxin induced mastitis is a milder disease where all infused animals completely recover irrespective of treatment. Furthermore, comparing experimental intramammary infusion of endotoxins to bacterial infection, electron microscopic studies revealed a different pattern of polymorphonuclear leukocytes invasion and udder lesions (Hill 1994). The claim that therapy of coliform mastitis should primarily focus on removing endotoxins from the udder and counteracting their effect, is largely unproven. Frequent milk-out after oxytocin injections and anti-inflammatory therapy are commonly advocated for treatment of coliform mastitis. Oxytocin injection before milking was found as effective as intramammary cephapirin or amoxicillin for treatment of field cases of mild clinical mastitis (Guterbock et al 1993). However, oxytocin injection combined with frequent milk-out for treatment of field cases of mild clinical mastitis was found to be either non-effective or even detrimental when compared to no treatment (Roberson 1997). The efficacy of steroidal antiinflammatory drugs in the treatment of mastitis was never proven in properly designed clinical studies. The efficacy of non-steroidal anti-inflammatory drugs (NSAID) was only proven in combination with antimicrobial therapy (Shpigel et al 1994, 1996). BCVA 1998 113

CATTLE PRACTICE VOL 6 PART 1 The systemic signs in coliform mastitis are generally attributed to the effect of inflammatory mediators and modulators released in the udder. Systemic invasion of coliforms from the udder does not occur and bacteremia has not been demonstrated in field cases of coliform mastitis (Powers et al 1986). However, the occurrence of bacteremia was recently reported in 32% of cows with severe or protracted clinical mastitis (Cebra et al 1996). Consequently, the use of systemic antimicrobial therapy was advocated in the treatment of this disease. The treatment of coliform mastitis is still highly controversial, there is a great need for well designed clinical trials to substantiate appropriate treatment regimens. Clinical trials are clinical research studies designed to assess the efficacy of a treatment protocol by comparing its effects with those of another protocol in a comparable group of client-owned, naturally occurring clinical cases. A well designed clinical trial should be a randomised, blind, controlled clinical trial. Despite ample experimental evidence for the potential efficacy of antimicrobials in the treatment of severe clinical mastitis, this has never been proven in a clinical trial. There is not enough evidence that antimicrobials can improve recovery and survival of clinically mastitic cows and reduce the loss of milk production. The assessment of mastitis therapies should include in vitro studies, experimentally induced mastitis studies and clinical trials. In vitro studies and experimentally induced mastitis studies demonstrated the therapeutic potential of various pharmaceuticals, defined their mode of action and improved our understanding of the pathophysiology of acute mastitis. However, these studies never proved the efficacy of these pharmaceuticals in the treatment of mastitis. This problem was well addressed by Verschueren in 1992 who suggested that "It should be borne in mind that, whereas it is interesting and sometimes helpful to show how a product works, what remains essential is to show that it works". We must remember that medical therapy should be based on the results of clinical field trials rather than on physiological reasoning. Previously published results from field trials and experimentally induced coliform mastitis trials failed to prove the efficacy of antimicrobial therapy (Jones et al 1990, Guterbock et al 1993, Pyörälä et al 1994). Antimicrobial drugs are assumed to exert their beneficial therapeutic effect via bactericidal or bacteriostatic action. In addition to the antimicrobial effect, some drugs have been reported to affect the pathophysiological processes by other modes of actions. One example is the capability of polymyxins to neutralize bacterial endotoxins (Ziv et al 1978). Antimicrobials might also exert detrimental effects by causing massive bacterial endotoxin release (Shenep et al 1984, Shenep et al 1985) and by interfering with phagocytic activity in the udder (Ziv et al 1983, Paape et al 1990). Significant differences were reported in the propensities of antimicrobials to release endotoxins and other bacterial toxins and of some bacterial strains to spontaneously release these toxins (Hurley 1992, Takahashi 1997). The clinical significance of both antimicrobial endotoxin inactivation and drug-induced endotoxin release is yet to be proved for coliform mastitis (Ziv et al 1983, Pyörälä et al 1994). In the three studies presented in this paper the occurrence of bacteremia and the efficacy of antimicrobial therapy were studied in experimentally induced and field cases of coliform mastitis. The clinical and bacteriological efficacy of cefquinome in the treatment of experimentally induced Escherichia coli mastitis was recently reported and will be further discussed in the present paper (Shpigel et al 1996). Cefquinome is an advanced broad spectrum, fourth generation cephalosporin with improved antibacterial activity over the second and third generation cephalosporins (Sader et al 1993). Cefquinome is resistant to β-lactamases produced by the majority of clinically important bacteria. Chemically, cefquinome is a new cephem; its zwitterionic structure can facilitate rapid penetration across biological membranes including the porins of the bacterial cell wall. The in vitro activity of cefquinome against E. coli of bovine mastitis origin is comparable to, or better than, third generation cephalosporins; the MIC90 and MIC50 values are 0.13 μg/ml with resistance rate of 0.4% (Schmid et al 1994). Sulphonamides and trimethoprim combinations have been used for many years for the treatment of infectious conditions caused by sensitive organisms. Although efficacy was never established this combination is frequently used in the treatment of clinical mastitis. The in vitro sensitivity of coliform udder pathogens to sulphadiazine and trimethoprim (STM) combination and its effect on the outcome of STM treated mastitis field cases were recently reported and will also be presented in this paper (Shpigel et al 1998). MATERIAL & METHODS Bacteremia study Bacteriologic blood cultures were performed on 22 cows with experimental Escherichia coli mastitis. Multiparous, Israeli Holstein cows in early lactation producing at least 25 L milk/d were used. Cows were infected with either 750 or 7500 cfu of E. coli (strain P-4, serotype O32:H37) infused into two healthy quarters of each cow. Cows were randomly assigned into one of the following treatment groups: 1) 20 g of sulfadiazine Na and 4 g of trimethoprim intramuscularly and additional intramuscular treatment of 10 g of sulfadiazine Na and 2 g of trimethoprim 24 hours later; 2) non antibiotic treatment. The treatment started 12-h after inoculation when clear clinical signs of acute mastitis were evident. BCVA 1998 114

CATTLE PRACTICE VOL 6 PART 1 For blood culture, the hair was shaved over the jugular vein, the site was disinfected with chlorhexidine and isopropyl alcohol. Five ml of blood were aspirated aseptically, 4, 12, 16 and 24 hours after infection. The blood was injected aseptically through a new needle into a blood culture vial containing 25 ml enrichment broth. The vial was aerated through the needle after blood inoculation, incubated overnight at 370 C in 10% CO2 atmosphere, and streaked onto blood and EMB agar plates. The plates were incubated overnight at 370 C in 10% CO2 atmosphere and colonies identified by standard laboratory methods. Cefquinome study Forty-seven multiparous, Israeli Holstein cows in early lactation that produced at least 25 L/d of milk were used, and 400 to 750 cfu of E. coli (strain P-4, serotype O32:H37) were infused into two healthy quarters of each cow. This strain was sensitive to cefquinome (MIC < 0.1 µg/ml) and ampicillin (MIC = 0.5 µg/ml) in vitro. Cows were randomly assigned to one of the following treatment groups: 1) 75 mg of cefquinome administered intramammarily three times at 12-h intervals, 2) 75 mg of cefquinome administered intramammarily three times at 12-h intervals and 1 mg/kg of cefquinome administered intramuscularly two times at a 24-h interval, 3) 1 mg/kg of cefquinome administered intramuscularly two times at a 24-h interval, and 4) 75 mg of ampicillin and 200 mg of cloxacillin administered intramammarily three times at 12-h intervals. Systemic and local clinical signs were monitored throughout the study period. Rectal temperature, heart rate, respiratory rate and rate of primary rumen contractions were determined once daily for 3 days before infection, immediately before infection, 4, 8, 12, 16 and 24 h after infection and then twice daily (a.m. and p.m.) for 7 d and once daily for another week. Systemic signs and clinical status of inoculated and control quarters were graded clinically. Cows attitude and appetite, milk appearance and quarter size, quarter edema, quarter pain and quarter temperature were graded as previously described [a clinical mastitis score (CMS) with a 7 to 35-point scale was used] (Anderson et al 1986). Duplicate quarter milk samples were aseptically collected from all quarters of each cow for bacteriological culture at 6 hours after infection, immediately pre-treatment time and 7 and 14 days post treatment. Bacteriological examination for udder pathogens was performed as previously described, using standard methods. Concurrent with the clinical examination, the California Mastitis Test (CMT) was performed and recorded using a five-point scale (Schalm et al 1971). Jugular blood samples were collected in plain and EDTA vacuum blood tubes once daily for two consecutive days before infection and 4, 12, 16 and 24 hours after udder inoculation and once daily for the following 7 days. Whole blood was analyzed for complete blood cell counts (CBC), and hematocrit (Ht). Serum samples were analyzed for total calcium, total serum protein (TSP), aspartate serum transaminase (AST), urea, creatinine, inorganic phosphorus and sodium. The biochemical analysis was determined enzymatically by use of an automated analyzer (Kone Autoanalyzer). Quarter bacteriological cure rates among treatment groups were compared by chi-square tests. Treatment differences among all other parameters were tested by least squares ANOVA with data blocked by day of challenge (SAS/STAT, SAS Institute Inc.). Differences among treatment means were determined using Duncan’s Multiple Range Test. For all comparisons, P < 0.05 was considered to be significant. Sulphadiazine trimethoprim study The association between in vitro sensitivity to antimicrobials and the outcome of treatment was studied in 228 cases of cows with coliform mastitis. All cases of mastitis were treated with a sulphonamide trimethoprim preparation, and 197 cases were also treated with a non-steroidal anti-inflammatory drug. Clinical mastitis was defined by the diagnosis of abnormal changes (acute, local, and systemic) in the body, udder, and milk, with concurrent decrease of at least 25% in daily milk production. Changes in the udder included pain, swelling, warmth, and abnormal appearance of milk (i.e., watery, clots, flakes or blood). The outcome of every case was categorized into one of the following groups: 1) recovered, cows that had returned to at least 75% of the premastitis daily milk production; 2) blind quarter, cows for which lactation ceased in the affected quarter(s) for the duration of the present lactation; 3) culled, cows that died, were slaughtered, were culled, or did not return to at least 75% of pre-mastitis daily milk production. The effect of in vitro sensitivity to sulphonamide trimethoprim on recovery was analyzed using multivariate logistic regression. The possible confounding effects of non-steroidal antiinflammatory drug treatment, days in lactation, parity, herd, and type of infecting organism were tested. RESULTS Bacteremia study All cows developed typical signs of acute clinical mastitis by 8 to 12 hours post inoculation. These included typical local udder and milk abnormalities and clinical, hematological and biochemical changes previously described in experimental and field cases of E. coli mastitis. Two Blood samples from 11 cows infected with 7500 cfu were bacteriologically positive; one with coagulase negative staphylococcus and the other BCVA 1998 115

CATTLE PRACTICE VOL 6 PART 1 streptococcus sp. Seven samples from the 11 cows infected with 750 cfu were positive; 2 coagulase negative staphylococcus, 2 bacillus sp. and enterobacter sp., 1 bacillus, 1 enterobacter and 1 coagulase negative staphylococcus and a coliform (non E. coli). All EMB cultures were negative for E. coli. Cefquinome study All cows developed acute clinical mastitis as assessed by the CMS and CMT (Fig. 1). CMS and CMT peaked around 24 hours post inoculation without a significant difference between treatment groups until day 2 post inoculation for CMS and day 7 for CMT. Thereafter, significant differences were observed between groups 1, 4 and groups 2,3. Figure 1: Mean clinical mastitis score (CMS) (top panel) and mean California mastitis test (CMT) score (bottom panel) in experimentally induced Escherichia coli mastitis in 47 cows before and after treatment with cefquinome administered intramammarily (‹), cefquinome administered intramuscularly and intramammarily („), cefquinome administered intramuscularly (σ), and ampicillin and cloxacillin administered intramammarily (z). BCVA 1998 Body temperature, heart rate (Fig. 2) and respiratory rate peaked around 12 to 24 hours after quarter inoculation without a significant difference between treatment groups. Primary rumen contractions (Fig. 3) declined to their lowest around 24 hours post inoculation. This decline was significantly smaller in group 2 cows. These changes were associated with decreased total WBC (Fig.3), TSP, blood calcium, phosphorus (Fig. 4) and sodium levels. The decline in WBC was significantly smaller in group 2 cows while the drop in blood sodium was significantly larger for group 4 cows. Blood urea and creatinin peaked on day 2 post inoculation for urea and around days 1 and 2 for creatinin. These peaks were significantly higher for group 1 and 4 cows. The hematocrit peaked around day 1 post inoculation without a significant difference between the treatment groups. Aspartate serum transaminase (AST) levels showed a small rise to peak around day 1 post inoculation without a significant difference between treatment groups. Figure 2: Mean rectal temperature (top panel) and mean heart rate (bottom panel) in experimentally induced Escherichia coli mastitis in 47 cows before and after treatment with cefquinome administered intramammarily (‹), cefquinome administered intramuscularly and intramammarily („), cefquinome administered intramuscularly (σ), and ampicillin and cloxacillin administered intramammarily (z). 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Days from i f ti CM S 37.5 0 5 0 5 0 5 0 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Days from infection Temperature ( 0 C) 38. 38. 39. 39. 40. 40. 41. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Days from i f ti CMT 60 65 70 75 80 85 90 95 100 105 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Days from infection Heart rate (beats/min.) 116

CATTLE PRACTICE VOL 6 PART 1 Figure 3: Mean primary rumen contractions (top panel) and mean white cell count (bottom panel) in experimentally induced Escherichia coli mastitis in 47 cows before and after treatment with cefquinome administered intramammarily (‹), cefquinome administered intramuscularly and intramammarily („), cefquinome administered intramuscularly (σ), and ampicillin and cloxacillin administered intramammarily (z). BCVA 1998 Figure 4: Mean total serum calcium (top panel) and mean serum inorganic phosphorus (bottom panel) in experimentally induced Escherichia coli mastitis in 47 cows before and after treatment with cefquinome administered intramammarily (‹), cefquinome administered intramuscularly and intramammarily („), cefquinome administered intramuscularly (σ), and ampicillin and cloxacillin administered intramammarily (z). 2 3 4 5 6 7 8 -2 -1 0 1 2 3 4 5 6 7 8 Days from infection Serum phosphorus (mg/dl) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Days from i f ti Contractions / 3 min. Milk production declined to its lowest 24 hours post inoculation for groups 2 and 3 while declining further in groups 1 and 4 to its lowest on day 2. Thereafter, milk production increased rapidly for groups 2 and 3, while milk production was significantly lower for groups 1 and 4 (Fig.5). Figure 5: Mean daily milk production in experimentally induced Escherichia coli mastitis in 47 cows after treatment with cefquinome administered intramammarily (‹), cefquinome administered intramuscularly and intramammarily („), cefquinome administered intramuscularly (Δ), and ampicillin and cloxacillin administered intramammarily (z). a,bMeans on the same day with different letters differ (P < 0.05). 0 5 10 15 20 25 30 35 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Days from i f ti The bacteriological infection success rate was assessed by duplicate culture samples taken 6 hours post inoculation and at pretreatment (Table 1). The bacteriological cure rate was assessed by duplicate culture samples taken at days 7 and 14 post treatment. All samples taken from successfully infected quarters should have been negative for E. coli to achieve bacteriological cure. E.coli were cultured from 95.8% (23/24), 87.5% (21/24), 95.8% (23/24) and 100% (22/22) of inoculated quarters of group 1, 2, 3 and 4 cows respectively (Table 2). The cure rate was 82.6% (19/23), 95.2% (20/21), 82.6% (19/23) and 54.5% Milk production (L/d a b a a b b a a b b a a b b a b b a a b a a a a a a b b b b b b b b ) 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 -2 -1 0 1 2 3 4 5 6 7 8 Days from i f ti Blood Calcium (mg/dl) 117

CATTLE PRACTICE VOL 6 PART 1 (12/22) of infected quarters of group 1, 2, 3 and 4 cows (Table 1). Table 1: Bacteriological infection rates and cure rates in infused quarters among the various treatment groups. Cows were sampled 6 h after infection, immediately before antibiotic treatment, and 7 and 14 d after the end of the antibiotic treatment. Treatment and route Of administration Infected quarters Recovered quarters (%) (no./no.) (%) (no./no.) Cefquinome Intramammary 95.8 23/24 82.6 19/23a Intramammary & injectable 87.5 21/24 95.2 20/21a Injectable 95.8 23/24 82.6 19/23a Ampicillin & cloxacillin Intramammary 100 22/22 54.5 12/22b BCVA 1998 a,b Rates with different letters differ (P < 0.05). The difference in bacteriological cure rates between group 1 and 4 (χ2 =4.132, p=0.042), group 2 and 4 (χ2 =9.345, p=2.24E-03) and group 3 and 4 (χ2 =4.132, p=0.042) are statistically significant. The differences in between the cefquinome treatment groups (groups 1, 2, 3) are not statistically significant. Sulphadiazine trimethoprim study In the STM study recovery rate from clinical mastitis was higher for cows infected by coliforms that were sensitive to STM than for cows infected by coliforms that were resistant to STM, 89.1% (147 of 165) and 74.6% (47 of 63), respectively. The odds ratio of recovery for cases associated with organisms that were sensitive to STM relative to cases associated with organisms that were resistant to STM was 2.75; the 95% confidence interval ranged from 1.25 to 5.85. Only non-steroidal anti-inflammatory drug treatment had a significant confounding effect and was included in the final statistical model. The odds ratio of recovery for cases treated with non-steroidal anti-inflammatory drugs versus cases treated with STM only was 2.76; the 95 % confidence interval ranged from 1.12 to 6.79. DISCUSSION The local and systemic signs observed in the blood culture study and the cefquinome study were similar to those described by others using the same model (Hill et al 1978, Lohuis et al 1990, Pyörälä et al 1994) and could not be differentiated from natural, field cases of acute coliform mastitis. The severity of the systemic involvement and of the sepsis that developed in the infected cows was clearly indicated by the various clinical, biochemical and hematological parameters described previously. The negative blood culture results probably indicate the absence of E. coli bacteremia at the acute stage of severe E. coli mastitis. Bacteremia might still occurs, as previously described, in severely affected cows with protracted disease. However, this phenomenon can probably occur in any severe and protracted septic condition irrespective of cause and affected body system or organ. The isolated organism should be typed and identity with the udder strain should be established. Otherwise, the bacteremic organism might have not originated from the udder and systemic invasion of normal gut organisms is more than likely in severe and protracted septic conditions. The use of antimicrobials in the treatment of severe clinical mastitis cannot be based on systemic bacterial invasion at the acute stage of the disease. Cows affected by any severe and protracted septic condition should probably be treated with systemic antimicrobials. In the cefquinome study no significant differences could be detected among the different groups from the beginning of the trial until the commencement of antimicrobial therapy. Cows in the control group further deteriorated clinically after the commencement of treatment and developed more severe local and systemic signs of clinical mastitis; those differences were present until the end of the study. Although the cows treated with cefquinome, developed a comparable level of disease after the intramammary infection, as indicated by the various clinical, hematological and biochemical parameters, those cows showed a more rapid and complete response to treatment. Cows treated with cefquinome either intramuscularly or intramuscularly and intramammarily returned more rapidly and completely to preinfection milk production with similar disappearance of mastitic changes in the udder and abnormal deviations of the clinical, hematological, and biochemical parameters. The bacteriological results clearly indicated the superiority of the cefquinome treatments over the control treatment. Although infection rates were not significantly different among the treatment groups, important and significant differences existed in bacteriological recovery rates. The combination of intramuscular and intramammary cefquinome treatment resulted in the highest bacteriological cure rate (20 of 21 infected quarters). However, this difference was not significant when compared with the other cefquinome groups and should be reevaluated in larger studies in which more quarters are involved. Treatment for the control group resulted in a bacteriological cure rate of 54.5% (12 of 22 infected quarters); this cure rate was very close to the spontaneous bacteriological cure rate expected in coliform mastitis at 14 d post-infection (Guterbock et al 1993). The significantly higher bacteriological cure rates achieved by the cefquinome treatment not only coincided with the return to production and normalization of the clinical, hematological, and biochemical parameters, but also could have lowered the probability of development of chronic mastitis 118

CATTLE PRACTICE VOL 6 PART 1 conditions with acute relapses in the following period. As previously indicated, results of field trials and trials with experimentally induced coliform mastitis that have been published thus far failed to prove the efficacy of any antimicrobial therapy. Furthermore, none of these trials were able to show the advantage of one antimicrobial treatment over the other. In the present study, negative, untreated controls were not included for ethical reasons, although the use of positive, treated controls was in fact advocated (Dohoo 1987). Nevertheless, the recovery rates and clinical responses achieved in the control group in the present trial were probably similar to what would have been found without treatment. In the present study, the advantage of one antimicrobial therapy over the other was clearly indicated. This difference was indicated in terms of return to milk production, disappearance of clinical signs of acute mastitis, and return to normal of various hematological and biochemical parameters. This study clearly supported the efficacy of cefquinome in the treatment of bovine coliform mastitis and indicated that cefquinome therapy can improve recovery and survival of mastitic cows and reduce the loss of milk production. In the cefquinome study all treatments were administered at the acute stage of the disease, starting 12 hours after infection and for a short duration of time. It should be emphasized that the clinical cases described in this study are of severe coliform mastitis where cows are systemically and locally affected. Comparison with previously published field studies where severe cases were purposely omitted (Guterbock et al 1993) should be cautiously done. Owners are reluctant to avoid treatment of severely affected mastitis cows, therefore, the efficacy of antimicrobial treatment was never evaluated in a negatively controlled, randomized, clinical trial. The STM study provided a unique opportunity to compare the effect of antimicrobial therapy with a virtual negative control group. The STM therapy probably did not have any effect other than the antimicrobial effect. Therefore, cows were randomly allocated into the treatment and control groups by virtue of sensitivity to STM of the infecting organism. Whether infection by strains that are sensitive or resistant to STM is associated with other factors such as age, days in lactation, or herd which may confound recovery, is unknown. Some of these possible confounders were evaluated in the statistical analysis. There is a possibility that organisms that are sensitive to STM induced a milder form of the disease and, hence, a higher recovery rate. In fact, the opposite might be true, the drug-sensitive strains of Gram-negative organisms are known to differ in their capacity to release endotoxins, either spontaneously or upon exposure to bactericidal antimicrobial drugs (Shenep et al 1984, 1985). Therefore, if indeed endotoxin release induced by drugs is of any clinical significance in coliform mastitis, we would have expected a lower rate of therapeutic success for the group of strains that were sensitive to STM. The antimicrobial therapy frequently has been opposed because of the rapid spontaneous drop of bacterial counts 8 to 24 h after infection (Hill et al 1978, Pyörälä et al 1994). Erskine et al (1991) argued that most treatments under field situations are administered too late and may even be detrimental because they further induce endotoxin release. The very high management standards of the participating herds and the use of computerised in and on-line electrical conductivity measurement systems probably ensured early diagnosis of clinical mastitis. The feasibility and advantage of early diagnosis and treatment were recently reported for Staphylococcus aureus and Streptococcus uberis mastitis (Milner et al 1996, 1997). Early diagnosis of Gram-negative clinical mastitis can be further facilitated by the use of cow-side tests such as the Limast® test for detection of endotoxins in the milk (Katholm 1997). The inclusion criteria of the STM study ensured that only severe cases were included and treated for clinical mastitis. Spontaneous cure rates are probably extremely high in mild cases of coliform mastitis. Clinical mastitis field trials are difficult and expensive, and inclusion of mild cases of coliform mastitis in those field trials may dilute the effect of treatment to undetectable levels. CONCLUSION This paper provided some supportive data in favour of rational use of systemic antimicrobials in the treatment of severe coliform mastitis. Treatment should be based on the following principles:- 1. Routine milk cultures and antibiograms of clinical cases 2. Early diagnosis and treatment 3. Pharmacokinetically adequate antimicrobials 4. Short duration of treatment with an adequate dose ACKNOWLEDGEMENT This paper is dedicated to the memory of Gideon Ziv, a teacher and a friend. REFERENCES Anderson K.L., A.R. Smith, R.D. Shanks, L.E. Davis, and B.K. Gustafsson. 1986. Efficacy of flunixin meglumine for the treatment of endotoxininduced bovine mastitis. Am. J. Vet. Res. 47:1366. Cebra C.K., F.B. Garry, and R.P. Dinsmore. 1996. Naturally occurring acute coliform mastitis in Holstein cattle. J. Vet. Internal Med. 10:252. Dohoo I. R. 1987. An assessment on evaluation of mastitis therapy. Page 207 in Proc. Int. Mastitis Symp., Bellevue, Quebec, Canada. Frost A.J., and A.W. Hill, 1982. Pathogenesis of experimental bovine mastitis following a small inoculum of Escherichia coli. Research in Veterinary Medicine 33:105. Guterbock W.M., A.L. van Eenennaam, R.J. Anderson, I.A. Gardner, J.S. Cullor and BCVA 1998 119

CATTLE PRACTICE VOL 6 PART 1 C.A. Holmberg. 1993. Efficacy of intramammary antibiotic therapy for treatment of clinical mastitis caused by environmental pathogens. J.Dairy Sci. 76:3437. Jones G.F., and G.E. Ward. 1990. Evaluation of systemic administration of gentamycin for treatment of coliform mastitis in cows. JAVMA 197:731. Hill A.W. 1994. Escherichia coli mastitis. Page 117 in Gyles, C.L., Escherichia coli in domestic animals and humans, CAB International, Wallingford, UK. Hill A.W., A.L. Shears, and K.G. Hibbitt. 1978. The elimination of serum-resistant Escherichia coli from experimentally infected single mammary glands in healthy cows. Res. Vet. Sci. 25:89. Hurley J.C. 1992. Antibiotic-induced release of endotoxins: a reappraisal. Clin. Infect. Dis. 15:840. Katholm J. 1997 Recommended principles for treatment of mastitis caused by gram negatives (all coliforms). Proceedings of the Nordic Seminar Regarding the Future Use of Antibiotics in Mastitis Therapy, Gothenburg, Sweden. Lohuis J.A.C.M., Y.H. Schukken, J.H.M. Verheijden, A. Brand, and A.J.S.P.A.M. Van Miert. 1990. Effect of severity of systemic signs during the acute phase of experimentally induced Escherichia coli mastitis on milk production losses. J. Dairy Sci. 73:333. Milner P., K.L. Page, A.W. Walton, and J.E. Hillerton. 1996. Detection of clinical mastitis by changes in electrical coductivity of foremilk before visible changes in milk. J. Dairy Sci. 79:83. Milner P., K.L. Page and J.E. Hillerton. 1997. The efects of early antibiotic treatment following diagnosis of mastitis detected by a change in the electrical conductivity of milk. J Dairy Sci. 80:859. Paape M.J., S.C. Nickerson, and G. Ziv. 1990. In vivo effects of chloramphenicol, tetracycline, and gentamicin on bovine neutrophil function and morphologic features. American Journal of Veterinary Research 51:1055. Powers M.S., M.E. White, P. Dinsmore., C. Guard and B., Dimock. 1986. Aerobic blood culturing in cows with coliform mastitis. J. Am. Vet. Med. Assoc. 189:440. Pyörälä S., L. Kaartinen, H. Käck, and V. Rainio. 1994. Efficacy of two therapy regimens for treatment of experimentally induced Escherichia coli mastitis in cows. J. Dairy Sci. 77:453. Roberson J.R., 1977. Frequent milk-out as a treatment for subacute clinical mastitis. Page 152 in Proc. 36th Annual Meeting, National Mastitis Council, Albuquerque, New Mexico. Sader H.S., and R.N. Jones. 1993. The fourth-generation cephalosporins: antimicrobial activity and spectrum definitions using cefpirome as an example. The Antimicrobic Newsletter 9:9. Schalm O.W., E.J. Carroll, and N.C. Jain. 1971. Bovine mastitis. Lea and Febiger, Philadelphia. Schmid P., A. Bottner and R. Humke. 1994. In vitro testing of bacterial field strains from bovine origin for sensitivity to cefquinome. Proc. 18th World Buiatric Congr., Bologna, Italy 1:539. Shenep J.L. and K.A. Mogan. 1984. Kinetics of endotoxin release during antibiotic therapy for experimental Gram-negative bacterial sepsis. J. Infect. Dis. 150:380. Shenep J.L., R.P. Barton and K.A. Mogan., 1985. Role of antibiotic class in the rate of liberation of Endotoxin during therapy for experimental Gram-negative bacterial sepsis. J. Infect. Dis. 151:1012. Shpigel N.Y., R. Chen, M.,Winkler, A. Saran, G. Ziv, and F. Longo., 1994. The anti-inflammatory ketoprofen in the treatment of field cases of bovine mastitis . Research in Veterinary Science 56:62. Shpigel N.Y., M. Winkler, A. Saran and G. Ziv. 1996. The antiinflammatory drugs phenylbutazone and dipyrone in the treatment of field cases of bovine mastitis. Journal of Veterinary Medicine A 43:331. Shpigel N.Y., D. Levin, M. Winkler, A. Saran and G. Ziv. 1996. Efficacy of the fourth generation cephalosporin cefquinome for the treatment of experimentally induced Escherichia coli mastitis in cows. J. Dairy Science 80:318. Shpigel N.Y., M. Winkler, G. Ziv and A. Saran. 1998. The effect of the in vitro sensitivity of coliform udder pathogens on the outcome of clinical mastitis treatment. Veterinary Record 142:135. Shpigel N.Y., M. Winkler, G. Ziv, and A. Saran. 1998. Clinical bacteriological and epidemiological aspects of clinical mastitis in Israeli dairy herds. Preventive Veterinary Medicine (Accepted for publication). Takahashi K. K., Narita Y., Kato T., Sugiyama N., Koide, T. Yoshida and T., Yokochi. 1997. Low-level release of shiga-like toxin (verocytotoxin) and endotoxin from enterohemorrhagic Escherichia coli treated with imipenem. Antimicrobials Agents and Chemotherapy 41:2295. Verschueren, C. 1992. A few thoughts on the demonstration of efficacy and clinical trials. Vet. Drug Reg. Newsletter, 6: 86. Ziv G., I. Hartman, and M. Torten. 1978. In vitro inactivation of endotoxin by polymyxin B and colistin in mastitic milk. Journal of Veterinary Pharmacology and Therapeutics 1:213. Ziv G., M.J. Paape and A.M. Dulin. 1983. Influence of antibiotics and intramammary antibiotic products on phagocytosis of Staphylococcus aureus by milk leukocytes. American Journal of Veterinary Research 44:385. Ziv G., and W.D. Schultze. 1983. Influence of intramammary infusion of polymyxin B on the clinicopathologic course of endotoxin-induced mastitis. American Journal of Veterinary Research 44:1446. BCVA 1998 120

CATTLE PRACTICE VOL 6 PART 2 Ultrasonographical Examination of the Mammary Gland in Cows with Induced S.aureus Mastitis: A Criteria for Prognosis and Evaluation of Therapy. Banting A., Bio Logic, 37340 Savigné sur Lathan, France. ABSTRACT Ultrasonographical examination of the mammary gland is used for the evaluation and follow up of the inflammatory and fibrotic reactions of the udder. 54 lactating cows with induced S.aureus subclinical, subchronical mastitis were used to determine the optimal dose rate of cefquinome administered via the intramammary route based on bacteriological criteria, somatic cell counts and the results of the ultrasonographical assessement. The significance and use of ultrasonography is discussed. KEYWORDS: S. aureus mastitis, Ultrasonography , bacteriology, cefquinome, amoxycillin. INTRODUCTION Mastitis is an inflammatory condition of the udder mainly of microbial origin. Diagnosis and evaluation of therapy are based on bacteriological criteria and somatic cell counts. However, bacteriological examination may be negative due to irregular excretion of certain microorganisms such as S. aureus,(Sears et al - 1990)(5) and somatic cell counts are influenced by many other factors, such as milking machines and procedures. Ultrasonography has been used for many years to examine soft tissues and can in particular reveal the presence of inflammatory reactions and fibrotic tissue in the bovine mammary gland as described by Tranquart et al. (7). Induced S. aureus mastitis models are convenient for dose titration studies of new intramammary products in lactating cows (1) and at drying off . A model in lactating cows has been used to evaluate the optimal dose of cefquinome(INN), a novel broad spectrum 4th generation cephalosporin . The purpose of the present paper is to describe our findings with echographical examination of the mammary gland in cows with induced mastitis and to discuss possible applications in the field. MATERIAL AND METHODS 1.Animals A total of fifty four lactating cows in at least their 2nd month of lactation were selected on a basis of somatic cell counts < 500 000 /ml and the absence of udder pathogens. The cows were in between their second and fifth lactation. The 500 000 cells/ml criteria was the commonly accepted figure when these trials were planned and conducted. The inflammatory and fibrotic status of the mammary parenchyma was also evaluated ultrasonographically on selection and again during the acclimatisation period, preceding the beginning of the challenge procedure described below. 2. Challenge procedure Somatic cell counts and bacteriological examinations were performed on individual quarter samples daily for the five morning milkings prior to challenge. Ultrasonographical examinations were also carried out before inoculation of the test strain. S. aureus 107-59 was provided by Poutrel (3). This coagulase positive strain produces β and δ hemolysins and has a virulence index of 90 according to Postle et al.(2).This virulence index is based on a bacteriogical score that is attributed according to the number of CFU recovered post inoculation for a given strain. The score is multiplied by 100 to determine the susceptibility index of a gland and the virulence index is the mean value of the susceptibility indices. The inoculum was prepared according to the technique described by Poutrel et al. and Postle et al . (3,2). 100 Colony Forming Units are infused into each quarter via the teat canal under a volume of 0.2 ml. 3. Post challenge procedure As from day three post inoculation, daily milk quarter samples are taken for somatic cell counts and bacteriological determinations until at least 80 % of the quarters excrete S.aureus. During this period ultrasonographical examinations are done every two or three days and clinical parameters recorded and scored (appearance of milk, pain, ...) at morning milkings. After the end of treatment the main parameters are determined twice weekly for 4 weeks. Echographical examination was only occasionally performed after treatment in Trial 1 and weekly in Trial 2. BCVA 1998 121

CATTLE PRACTICE VOL 6 PART 2 4. Treatment This study was conducted in two phases. In the first trial five groups of six cows were used and in the second phase one group of six infected non treated controls and two groups of nine treated animals. The cows were allocated to the different treatment groups so that the number of quarters excreting S.aureus was similar in each group. An intramammary formulation of cefquinome was prepared at various potencies (45, 60 and 75 mg/syringe). The reference product chosen was Amoxi Mast® (Pfizer USA) that contains 62.5 mg of amoxicillin per unit. The recommended use is an infusion after three consecutive milkings. In Trial 1 the three different potencies of cefquinome were evaluated, the test product being infused twice at a 24 hour interval . One group was kept as infected controls and the remaining six animals received the reference treatment.This design was chosen so as to define a dose response in comparison with a commercially available product used according to manufacturer’s recommendations. During the second phase, the effects of three infusions of the test product at the highest potency (75 mg/unit) (n = 9) were compared to the reference product (n=9) and the six remaining cows kept as infected controls. 5. Ultrasonographical examination Portable Aloka 210 DX equipment was used with a 5 Mhz linear probe. Examinations were documented with a Sony video printer. The absorption of the ultrasonic waves is modified when the hydric content of tissues is modified (hypoechogenic inflammatory nodules) or in presence of dense stromal tissue (hyperechogenic fibrosis). The hypoechogenic nodules appear as dark circles, distinctly limited by dense white tissue, these are to be seen disseminated in the parenchyma. Dense white tissue in the intermediate or deep part of the parenchyma represent the fibrotic lesions. The presence of dense white tissue in the superficial part of the parenchyma is normal. The following semi quantitative scoring system was used:- * Inflammation index 0: homogeneous aspect of the parenchyma, 1:small hyoechogenic nodules (<1cm) 2: large hypoechogenic nodules (>1cm) BCVA 1998 3: both types of nodules present * Fibrosis index 1: no fibrosis 2: discrete signs of fibrosis 3: extensive fibrosis RESULTS 48 hours post inoculation subacute clinical mastitis may be observed: udders are painful, hot and swollen. Milk clots are observed. These reactions are not seen in all the cows and vary considerably in intensity. At this stage a moderate to severe increase in somatic cell counts is observed (~ 2 million cells per ml) but S.aureus is rarely isolated from milk samples. As from day 3 a subclinical form develops. The aspect of milk returns to normal, the udder is hard but no longer painful or hot. Somatic cell counts stay at the same level or continue to rise. Inflammatory nodules (fig. 1) are regularly observed in practically all the infected quarters and S.aureus starts to be isolated from milk samples. S. aureus is regularly isolated as from day 7 to 10 onwards the inflammatory nodules persist and increase and fibrosis starts to appear. Somatic cell counts are stable ( 2 to 3 million cells per ml). At this stage the percentage of quarters excreting the test organism justifies initiation of treatment (> 80%). In this case bacteriological evidence of success of experimental infection was obtained in 83% of the quarters but modifications of the inflammatory and fibrotic status of the mammary parenchyma were observed in virtually all the animals. Even when S.aureus was never isolated from a given quarter, inflammatory nodules were observed whereas somatic cell counts were variable. Results of the bacteriological determinations are given in Table 1. Bacteriological cure is defined by the total disappearance of S.aureus from the milk samples. Spontaneous bacteriological cure is normally around 15 % in this model but was somewhat higher (21%) in the present study due to normal biological variations. Table 1. Bacteriological results Treatment No. of quarters No. of infected quarters No. of cured quarters % cure rate Control 48 38 8 21 Amoxi Mast x 3 60 52 30 58 Cefquinome 45mg x 2 24 21 7 33 Cefquinome 60mg x 2 24 18 8 44 Cefquinome 75mg x 2 24 21 11 52 Cefquinome 75mg x 3 36 30 21 70 216 180 These results clearly show a significant dose response with the two infusion regimen and the second phase justifies the the three infusions recommended for the commercial product (Cephaguard® LC - Hoecsht Roussel Vet Ltd.) 122