Mycoplasma bovis: Reviewing the reviews

MPI’s dedicated M bovis website (www.mbovis.govt.nz) is teeming with the partially redacted musings from the advisors, researchers and reviews of the programme. Guess who foolishly volunteered to read them all?

It’s not all bad, there were some juicy bits left amongst the grey boxes they paste over information deemed too sensitive for public release. I have picked through the reports and tried to put together a summary. A review of the reviews if you will.

SEPTEMBER 12, 2004 A 22-month-old Dexter born to an imported Australian dam dies of pneumonia. This may be NZ’s first case of M bovis. We’ll get back to this later.

2007 A national survey of 244 bulk milk samples were randomly tested for M bovis DNA and living mycoplasmas. Nothing is detected.

JULY 17, 2017 A vet treating sick cattle on a South Canterbury dairy farm owned by the Van Leeuwen Dairy Group contacts the exotic pest and disease hotline about suspected cases of M bovis. The cows have developed alarming high rates of mastitis and many have developed arthritis in one of their front legs.

JULY 21, 2017 MPI confirms that this is a case of M bovis. All 16 farms in the Van Leeuwen dairy group are placed under movement control. This is the first notified case of M.bovis in NZ. Later analysis will reveal that the disease has already been in the country for at least 18 months, but for a while experts believe they have found the epicentre of the outbreak.

JULY 31, 2017 M bovis is detected on one of the other Van Leeuwen Dairy Group farms.

AUGUST 11, 2017 NZ’s M bovis testing creates a worldwide shortage of testing kits for the disease.

AUGUST 29, 2017 Three more South Canterbury farms are designated infected.

OCTOBER 12, 2017 The governance group decides to depopulate 5500 cattle from affected farms at an estimated cost of $16.5m. PGG Wrightson agrees to carry out valuations needed for compensation of culled cattle.

OCTOBER 31, 2017 MPI suspends the live animal Import Health Standard for live cattle. This is mostly symbolic since imports are only allowed from Australia and no one has ordered any Aussie cattle since 2013.

NOVEMBER 2017 There is an “Analysis of risk pathways for the introduction of Mycoplasma bovis into New Zealand” report investigating the ways that M bovis may have entered the country. The authors are redacted in the public copy, probably on account of how much heat the report receives in subsequent reports. At this stage seven farms are known to be infected: five of them part of the same business. The report investigated seven potential pathways that M bovis might have reached NZ: imported live cattle from Australia, imported frozen semen, imported embryos, imported feed, imported used farm equipment and the importation of other (not cattle) animals.

Yes, that only adds up to six, MPI has chosen to redact one of the methods which I assume is “animal remedies”. The report does not rule any of these pathways out. From genomic sequencing it looks like this strain has originated in Europe, although there is no definitive match with overseas strains. The authors seem convinced that the first farm with M bovis’ symptoms (“the index farm”) is the epicentre of the outbreak. With the benefit of hindsight we can see that they are barking up the wrong tree. But to be fair to them, the cows on the index farm were going down with weird symptoms at an alarming rate not seen on any other farm.

The authors reveal that M bovis DNA was detected in imported semen, but they could not culture any living M bovis bacteria from the semen straws. So, it seems likely that M bovis is killed when the semen is processed and frozen prior to importation. The authors muse that the high prevalence of M bovis overseas and the amount of artificial insemination taking place with imported semen mean that M bovis should be a much larger problem if this is a plausible route of infection.

The idea of the disease coming in on imported cattle seems a little far fetched. No live cattle have been imported since 2013 and even in the 10 years leading up to that there were only 110 consignments from Australia, so probably not that. Something has also been redacted from the discussion on imported sheep and goats, but from context clues, I think they view that risk as negligible. They recommended more study into the infection risk of semen and embryos, as well as locating records on imported cattle. Tracking down the imported cattle records was eventually done in June 2019, finding two alarming, but ultimately unimportant, potential cases of historical M bovis infections in NZ.

DECEMBER 2017 Infected cattle are traced back to Alfons Zeestraten’s Southland business, Southern Centre Dairies. From here on in, this will be the new assumed epicentre, taking the heat off the Van Leeuwen Dairy Group. The heat comes on everyone else though when the first North Island farm tests positive in Hastings. The Technical Advisory Group (TAG) is having its first meeting when this bombshell hits.

TAG is not a fan of the risk pathways report, saying it “added little useful information germane to this outbreak”, but agrees that imported semen is an unlikely source of infection. TAG notes that overseas donor bulls are required to have never tested positive for M bovis, but there is no requirement to actually test the bulls in the first place. That seems like an oversight, but the symptoms of the infected cows on the Van Leeuwen property (i.e. limping on one front leg in late pregnancy) doesn’t fit well with being inseminated with infected semen nine months prior.

TAG supports eradication, and reports that in other countries eliminating M bovis from a herd has not been successful. So the group recommends whole-herd culling and the 60-day stand down period that would ultimately be adopted. TAG also notes that the evidence points to one or two disease clusters and that the programme should aim to complete national eradication no later than mid May 2018!

TAG also recommends that MPI resolve “major deficiencies in the practical operation” of the NAIT tracing system.

2018 The year of the drawn-out eradication decision

FEBRUARY 2018 It’s time for another TAG report. It is now hypothesised that the M bovis incursion occurred in mid-2016 or earlier. There is a backlog of 200 herds with animal movements to or from infected properties to be tested. Each of these properties is generating another 20–30 new farms to look at. MPI anticipates that it can complete outstanding trace visits before “Gypsy day”. There is no history of the infected farm using imported semen. The majority of TAG still believe that eradication is feasible and desirable, although the scale of the effort has increased substantially.

TAG seemed to have been asked about the prospect of declaring the North Island to be M bovis free. They are cautious about this suggestion (which puts them on the right side of history). They were somewhat keen on establishing disease-free zones, although they are impaired from doing so because they are still not sure they have found the source of the original infection.

At this stage we are still another three months away from the government committing to an eradication programme, so there is some discussion about different schemes. TAG says that a non-government control programme would need a better bulk milk test; there is a level of crystal-ball gazing needed when interpreting M bovis test results. TAG says a voluntary scheme is unlikely to be adopted. Ignoring the problem is also an option the group says. “There are no historical examples where a national government has funded a disease eradication campaign for a disease not listed by the OIE and there are no human health, food safety or trade implications.” Yes it sounds like TAG is taking a dollar each way about eradication, but doesn’t have enough information to do much else. It’s obvious from this report and the previous TAG report that TAG is a little grumpy about not having access to the government’s cost/benefit analysis and better information about how the disease got into the country.

26 MARCH 2018 MPI declares it will cull 22,300 cattle on the 28 farms that have tested positive so far. An extra $85m (of which $50m is compensation) is found in the government coffers. MPI also provides a short list of responses to some of the TAG’s recommendations. The highlights are that MPI says it can’t investigate how the disease got into NZ in greater detail until it knows it has worked out which farm has started it all. MPI also can’t “complete eradication by mid-May 2018” because it says that the scale of the infections has increased considerably and it can’t make a decision on eradication before mid-May.

APRIL 27, 2018 Another TAG report. It says eradication still appears feasible, given the distribution of cases so far. There has been a study testing cattle carcases from 720 farms at the meat works and so far anything suspicious looks like false-positives. The TAG report concludes that the number of infected beef farms is probably low. (Though hindsight being 20:20, we can see that the majority of infected properties are categorised as “beef” in the M bovis stats).

We are still another month away from the Government making a decision on eradication, so there is plenty of discussion on this. TAG says that the programme will need to run for at least five years and the public will need to be warned that cases will be popping up for years to come. If eradication is to be attempted, TAG recommends a go/no decision point in spring 2018, this is when the bulk milk tank results will start coming in from the dairy farms. That should give a lot more information on the size of the problem.

The TAG recommends someone undertake modelling of disease spread. And yes, they would still like to see the economic impacts of the disease under various control strategies. TAG knows some economic modelling must have been done and they would really like to see it. I would love to see it too. In the hundreds of pages of report releases I had to dig through for this, it never once makes an appearance. Also missing in action is the definition of success for an eradication plan which the TAG repeatedly calls for over the next few years. You might think that the definition of success is something like “ding dong the disease is dead”, but it’s not that simple. You can’t prove the absence of something, especially with M bovis which spends most of the time just chilling. M bovis is a part-time disease with an excellent work-life balance. The TAG is happy to give examples of success, “For example, the objective might be 95% confidence that the national herd prevalence is less than 2%”, they say. The report shows that after the TAG meeting, MPI informed them that cleaning up infected properties is not going well. Considerable effort has been spent cleaning equipment that has had limited contact with animals. Cleaning of yards and buildings has been extremely time-consuming, with it taking more than 60 days on some properties to complete two rounds of cleaning including the two-week drying period, after which time the 60-day stand-down begins. In practice therefore, farms are experiencing a stand-down in excess of 120 days. It is costing $100,000 per farm to dispose of the contents of effluent tanks. MPI has therefore proposed changes largely around when the 60-day stand-down period before repopulation begins. Pasture stand down will now begin once all animals leave and they will move to a single round of cleaning for plant and buildings.

MAY 17, 2018 Another month, another TAG report. This time there is some division in the team. Nearly half the TAG members (4/10) now believe that eradication is not rational given the current number of cases. The six others believe that eradication is still technically feasible since the epidemiology remains unchanged; however they have concerns about the rate of ongoing infection compared with rate of identification, concerns about REDACTED and serious concerns about ongoing social effects of eradication. Of the six TAG members who still believe that eradication is achievable, there is a diversity of opinion about what point it would be deemed unfeasible. At this stage there are 38 identified affected farms and a handful of cases in the North Island. There are 252 high-risk farms yet to be tested. Farm depopulations have thus far been slow and TAG questions whether MPI has enough resources available to undertake depopulation. However, no new clusters have been identified, which is encouraging – although at this point testing is limited to farms identified by animal movements and bulk milk testing. There is still no surveillance of the beef population underway.

MAY 28, 2018 – Today is the day that the Government announces a decision to eradicate Mycoplasma bovis. It is expected to take 10 years and cost $886m and the culling of 126,000 cattle. “Senior MPI leaders and industry participants interviewed [for a July 2021 review] stated that the response grew from nothing, to over 400 staff, within a year of initiation.” A condition of the Government agreeing to eradicate M bovis in May 2018 was that industry bodies signed up to a cost-sharing arrangement under the GIA (Government Industry Agreement (for Biosecurity Readiness and Response)). As a result, DairyNZ and Beef + Lamb New Zealand became signatories to the GIA. The GIA provided a mechanism through the Biosecurity Act 1993 to seek approval from farmers for a biosecurity levy to pay for the M bovis response. The Government commits to meeting a very specific 68% of the eradication cost, with DairyNZ and Beef and Lamb sharing the rest of the costs. This is the first, and last time, that anyone sees any sort of cost-benefit analysis. The Government says that opting for long- term management of M bovis would cost $1.2 billion (presumably over 10 years) and not doing anything would cost $1.3b in lost production over 10 years. We can all agree that $886m in eradication costs is smaller than $1.2 or $1.3b. But that’s all the information we have. Spoiler alert: there are budget blowouts ahead for eradication.

JULY 2018 Government announces an investment of up to $30m over two years for science to support the eradication effort. The dairy cows are back in the milking shed, so the national bulk tank milk testing begins for the season. M bovis pops up in Northland. There are discussions with the four largest processing plants in the South Island regarding costing schedules and extra charges for processing M bovis animals. For whatever reason, MPI has spent a bit of time trying to personally convince me that they have not paid meat processors anything. Government departments have strange hobbies. Anyway, July 2018 is also when it becomes compulsory to tell the neighbours that there is M bovis over the boundary fence.

AUGUST 17, 2018 The date at which MPI’s July 2019 review states that the four main processing plants agree to substantially reduce charges for processing M bovis animals. This would be the processor payments that MPI, under pain of the Official Information Act, deny ever happened.

AUGUST 24, 2018 ANZCO’s Five Star Beef feedlot is put under movement control after testing positive for M bovis.

SEPTEMBER 4, 2018 MPI mounts 10 charges against Alfons Zeestraten, the owner of Southern Centre Dairies in Southland relating to the importation of farm equipment in January 2018. They state that these charges are not related to M bovis, although this farm is believed to be the likely epicentre of the M bovis outbreak. Before the trial, four charges of unpacking a container without the permission of an inspector and introducing goods – an effluent pump – into a restricted place (I assume this is Zeestraten’s depopulated property) were withdrawn by the Crown. And halfway through the trial another four charges were dropped when the Crown told the judge it had no further evidence to give in relation to them. The remaining charges related to a container of fertiliser which was picked up from Port Chalmers in Dunedin without first being inspected by Fertiliser NZ in Winton Southland. Zeestraten is eventually discharged without conviction and sells the dairy company in December 2019.

DECEMBER 2018 The governance group becomes concerned that it is being fed erroneous information. Industry leaders start to hear from their own channels that stock movements are not being traced promptly, that numbers of potentially affected properties are increasing, and this is not being formally reported to governance. This backlog of cases will blow up over the coming months until publicly revealed in April.

2019 The year of the backlog

JANUARY 24, 2019 After seven months of silence, the TAG reports are back and they have something to say about the programme. “While we understand that the scope of our work is primarily technical, we have commented on issues such as communication, social licence and staff fatigue. We have done so as we consider these as potential factors that may slow or derail eradication.” In summary, they feel that they have lost some strategic oversight. “For example, there was a greatly reduced involvement of experienced Wallaceville staff during the November 2018 meeting compared to our initial meeting in November 2017 and there were no experienced veterinarians from MPI’s risk assessment or animal import Directorates.” MPI will later commission many reports after the fact that tell them the same thing – that things are getting strained and confusing within the programme.

On the technical side of things, TAG is divided on whether imported semen should be further investigated as a cause of the outbreak. The NZ strain of the disease is associated with a strain that showed susceptibility to routine antimicrobials used in semen preparations and the TAG has “not been provided with any information to indicate that this is no longer the case”.

TAG is asked to comment on whether M bovis has been in NZ prior to the assumed 2015 kickoff date. To do this they review some historical cases of potential M bovis cases in NZ. They say that evidence of historical M bovis cases in NZ was presented to them in November 2018. This is the 2004 Dexter report I mentioned earlier. Results from his samples suggest he had M bovis. But it is unlikely that this animal led to the current outbreak. Another sample from an adult cow with pneumonia that died in 2015 has tested positive for M bovis antibodies. There are some musings about the accuracy of the tests in these historic samples and neither are plausibly linked to the outbreak. But TAG notes that if these cases had been tested earlier, then the alarm could have been raised about M bovis risk before NZ saw an outbreak.

TAG still has some grumblings about the lack of economic data provided. The group’s been asked to comment on the numbers used for economic models, but members aren’t allowed to review the models themselves. TAG also comments on the costs of disinfecting slaughter plants handling infected cattle, noting there is some “resistance from some [meat] works to handle these stock” but that new protocols for meat plants should help resolve this.

MARCH 26, 2019 M bovis programme director Geoff Gwyn says that there are another 41 farms listed as infected (bringing the total up to 143 at the time). He says they can’t link all the infected farms back to the initial farm, likely due to improper use of the national animal tracking system, NAIT. Farmers would argue that NAIT has never been fit for purpose.

APRIL 1, 2019 The eradication programme now employs 423 staff, of which 110 work in the national control centre and 160 for AsureQuality. West Coast has its first case.

APRIL 12, 2019 DairyNZ CEO writes to the Director-General of MPI formally seeking information on the casing backlog and the Government publicly admits to a huge (666 case) backlog in M bovis case reporting. Later, a follow up report will state that “farmers, veterinarians and MPI field staff and contractors claimed that they were alerting the National Control Centre since spring that farmers that had received animals from Confirmed Places had not yet been contacted (i.e. had not been cased). They believed their concerns were not given due consideration and were not surprised by the announcement of a ‘backlog’.”

APRIL 15, 2019 Three days later the backlog is revised to 1100 properties of which 300 are deemed high risk and 800 medium or low risk. MPI’s science adviser is asked to review the reasons and impact of the backlog, and by the time this happens the backlog has been revised again to 1400 cases.

APRIL 25, 2019 Enough’s enough and the Director General of MPI spends his Anzac day weekend drawing up a “Reset” plan. As part of a ten point remedial action plan to get everyone talking to each other, an independent chairperson is appointed.

JULY 2019 There are two reports into the backlog. It transpires that none of the operational staff are surprised, but the backlog is news to upper management. Key reporting that was being undertaken by a Wallaceville epidemiologist has been lacking ever since they stopped inviting said epidemiologist to meetings. The cause of the backlog is a constellation of problems. There is a silo-type structure to the organisation and this discourages collaboration between different arms of the response. Data management is haphazardly taking place in individual spreadsheets across MPI and AsureQuality. There is no decision making at field level; all decisions are funnelled up to a slow and isolated National Control Centre in Wellington. Hastily recruited staff do not have the skills to deal with a disease response and veterinarians have been cut out of the loop. There’s a lack of a well-trained emergency response team to guide said staff – all in addition to generally poor management from an indecent management system that is only appropriate for short responses of a few weeks or months. Add the dog’s breakfast that is NAIT and farmer compliance to it. It was obvious in December 2018 that the caseload was ramping up, but little work was undertaken over the Christmas holidays and staff contracts expired at this point. The science adviser calls for more 49 extra staff, revealing that it takes one person approximately one-to-two weeks to trace the movements of one infected property.

SEPTEMBER 6, 2019 Milksolid levy comes into effect. Dairy farmers are charged 3.9 cents per milk solid for the M bovis programme. Background surveillance of the beef population begins with the testing of some animals coming onto the Five Star Beef feedlot.

OCTOBER 18, 2019 TAG is back and heartened to see that MPI has committed to changing the directorate structure, employing some additional veterinary epidemiologists, implementing a new information system and increasing the resourcing of communication. But it is too early to say if it is going well. TAG remains convinced that this outbreak is either from a single introduction or from a small number (three or four) closely related introductions of M bovis some time around 2015 or 2016. TAG is concerned there is still little surveillance testing for beef farms. They take an audience with Dr Keith Woodford whose blog on his independent investigation into M bovis is blowing up at the time. TAG notes that a number of clinical veterinarians report frustration with their inability to be able to provide technically sound information to herd owners and to understand the context for decision making about individual herds. While it is recognised that, because of confidentiality and resourcing issues, high levels of detail about every decision cannot necessarily be provided to herd owners or their veterinarians, a system enabling herd owners to provide permission for their veterinarians to be privy to, and involved in, discussions with response staff would be highly desirable.

NOVEMBER 2019 There is a $58.2m shortfall in funds to run the programme for the 2019-2020 financial year.

SOMETIME AROUND THE END OF 2019 Research looking into the onfarm symptoms of infected herds is completed. Of the 25 infected dairy farms surveyed, 13 (52%) had no discernable disease from M bovis, a further four farms were battling with mastitis that appeared to be environmental (bad weather, machine faults, etc), two farms had a couple of cows with arthritis, one had 50% of their dry cow mob affected with mastitis and arthritis in their forelimbs (which must be the first Van Leeuwen property), three other farms had higher than normal rates of mastitis. There were 37 calf rearers surveyed and they all seem to have reported symptoms in their calves to some degree, but they note that calf data is hard to interpret, because there are many common calf diseases and because some of these symptoms were exacerbated by calf movement controls. There was little to nothing to report on dry stock farms, lifestyle blocks and feedlots.

2020 The year of lockdowns and droughts

JANUARY 2020 Routine M bovis sampling at slaughter plants begins; it will last until the national lockdown in March coincides with a widespread drought, putting major pressure on cattle processing.

FEBRUARY 2020 In an update to Cabinet, the programme asks for two years’ funding worth $336.2m. Instead they receive one-year funding at $195.50m, including $53.2m in compensation. By March the programme’s contingency fund has been bled dry.

JULY 2020 On-farm testing of M bovis during TB testing is underway.

SEPTEMBER 2020 As part of the $30m research investment, researchers survey 286 commercial cattle farmers and 152 livestyle owners, revealing NAIT compliance is generally pretty good for lifestyle block owners who own (a median of) two dairy cattle or seven beef cattle that barely go anywhere. More than 70% of lifestyle block owners report never taking cattle off the farm prior to taking them for finishing or slaughter. On a less wholesome note, 8% of lifestylers and 14% of commercial farmers with pigs admit to feeding uncooked meat to pigs, which is a pathway to picking up serious livestock diseases. Both sets of farmers listed veterinarians as their most trusted information source. Leagues ahead of research institutes and industry bodies, which were the next runners up. MPI barely scrapes up a pass mark for trust and relevance.

2021 The year of the reviews

MARCH 24, 2021 After an 18-month silence and a week-long meeting in November, the TAG report is back. The programme is down to 10 infected herds, after depopulating 261 farms. Not every infected farm has been linked back to a known source of infection, however they now have nearly 700 genomic samples (representing ~40% of infected herds) and they all appear to be closely related. This reinforces the idea that M bovis is a relatively new arrival to the country. TAG commends the casing team on their improvements in data handling. There is a discussion about a Canterbury cluster of nine infected farms in 2020 and it seems that the Five Star Beef feedlot is a likely source of infection. But, overall things are looking good and it’s time to start thinking about the next stage of the programme: how is NZ going to go about defining itself when it is free from M bovis?

APRIL 23, 2021 Van Leeuwen Dairy Group, the company with the first identified M bovis cases, goes into receivership.

JULY 2021 A 200 metre minimum cattle-free buffer is placed around the Five Star Beef feedlot as a precautionary measure.

An independent review (196 pages!) of the M bovis eradication programme provides a comprehensive autopsy of everything that has gone wrong. The review finds the programme is on track to eradicate M bovis and that $635.9m or (74%) of the budget has been spent. Of this, there was $208.7m of compensation (or just under a third of the total spend). The report notes that there were tensions between MPI and AsureQuality about the costs incurred. At some point there was a review into the value for money spent by AsureQuality. This has not been released, but we can see that the review finds in AsureQuality’s favour. It seems that they were not ripping MPI off. But the contracting, reporting and oversight in the relationship were not fit for purpose.

Data sharing and communication was an issue. “One farm had stock killed under oral instruction from the ICP [Incident Control Point] manager. This ICP manager didn’t document this and then left AQ [AsureQuality] and no records [were] kept. The farmer sought compensation for these animals, but there was no record of the notice to destroy the stock… We eventually found out the wrong stock were destroyed on the wrong farm.”

Overall the independent review finds that many things have improved, but the programme still has poor strategic oversight and acceptable-but-not-great feedback channels.

In the same month, there is also a report on data sources to “identify gaps and highlight areas of missingness” in NAIT records using complicated maths, and an analysis of NZ stock movements to describe NZ’s cattle movement patterns and better ways of getting them into NAIT.

AUGUST 2021 MPI appoints a chief veterinary officer (CVO) for the first time in a couple of decades. Up until this point “MPI had no formally recognised CVO and no single leader responsible for livestock biosecurity readiness and response who would have provided the link between the Programme and wider MPI biosecurity structure, as well as preparing MPI for a future incursion.” (From the independent review above).

OCTOBER 2021 Results from a survey of 71 of the 4681 farmers affected by M bovis. This documents the stressors from on farm restrictions, financial uncertainty, excessive workload, community perceptions and concern for animal welfare. It highlights some holes in the compensation process. “But the amount of extra work it has taken to get through this is horrendous and we can’t claim for it. Like contractors, if you haven’t used them before, you can’t claim for them now.”

DECEMBER 3, 2021 Government announces that M bovis elimination is “on track” and that there are only four farms with current infections (all in Canterbury). At this time there have been 268 farms culled, totaling 173,000 cattle.

2022 The year of the Wakanui cluster

MAY 5, 2022 Government announces a $68m spend over the coming year to “continue momentum on M bovis programme”. At this point M bovis is reported to be isolated to just one farm (ANZCO’s Five Star Beef feedlot in Ashburton).

JUNE 16, 2022 There is another Mid Canterbury case detected via the bulk tank milk surveillance.

JUNE 30, 2022 MPI announce the cost of the programme is now up to $588m, even though it was at $635.9m in a review a year ago.

SEPTEMBER 4, 2022 Two more Mid Canterbury farms test positive for M bovis.

SEPTEMBER 22, 2022 MPI announces for the second or third time they are going to cull the ANZCO’s Five Star Beef feedlot which has been infected for four years at this point. A controlled area notice is declared for the Wakanui area surrounding the feedlot. Eight farms in the high-risk area will be depopulated as well as the 12,000 cattle on the feedlot.

OCTOBER 2022 A new strain of M bovis is identified in this cluster. Investigations are underway to work out where it has come from, including inspecting frozen semen used by the farm.

NOVEMBER 2022 There is a review of the new Canterbury cluster. There are eight infected properties within the new Controlled Area Notice at Wakanui and there is no evidence to suggest these farms brought in infected cattle. In six of these, cattle may have grazed within 1.5km of the feedlot. This situation raises the question that is central to this review: Is M bovis spreading in Wakanui through one or more less common routes of transmission? During the national programme, it was very rare for M bovis to jump boundary fences. In fact, it was only confirmed in three cases. The disease spread predominantly through cattle purchases or feeding infected milk to calves – two routes that have been ruled out in the Wakanui cluster. It seems that these properties have caught the disease from the nearby feedlot, but it is not obvious how. Groundwater leaching, manure spreading, and vehicles are all ruled as unlikely. This leaves airborne transmission, and contamination from seagulls and flies.