The Animal Health Board (AHB) - Containing Bovine TB

Date: 25 August 2009 - 2 Comments

Bovine TB is an infectious disease caused by the bacterium Mycobacterium bovis. In New Zealand, contact with wild animals (known as TB vectors) is the major source of TB infection in dairy cattle. The most common TB vector is the Australian brush tail possum. A high prevalence of bovine TB in dairy cattle could result in negative consumer perceptions and market reactions and significant production losses for New Zealand farmers[1]. Potential export trade bans on the dairy, beef and deer industries as a result of bovine TB have been estimated at a cost of $1.3 billion over 5 years[2]. In 1998, the Animal Health Board (AHB) was appointed under the National Pest Management Strategy to protect New Zealand dairy, beef and deer exports and reduce the cost of bovine TB to farmers. Their current objective is to achieve the international standard for TB freedom where 99.8% of domestic cattle and deer herds are free of bovine TB for three years[3].

Disease and vector control of bovine TB has moved from a paper based system in the 1990s to a system that widely adopts modern spatial information and technology. While disease and vector control planning has always been spatially based, systems are now underpinned by digitised geospatial information. This includes the VectorNet and Disease Management Information Systems used by the AHB.

Prior to the formation of AHB, bovine TB control was undertaken by pest destruction boards which were amalgamated into regional councils. Possum control by government is also currently managed on public conservation land by the Department of Conservation[4] (DOC). 

Spending on control of bovine TB in 2007 was $81.92 million; $44.86 million from the private sector and $36.92 million from local and central government. This private sector spending includes at least $16.97 million paid by dairy farmers as levies (Animal Health Board, 2008).

Vector control: spatially based methodology and tools

Vector control involves creating buffer zones of low-density TB vector populations between TB-infected vectors and cattle herds. In order to create buffers, surveying the extent of TB-infected wildlife populations and their habitat close to and near buffers zones has been used to determine the level of pest control needed.

In the 1990s, vector control planning was a paper based exercise involving manually tracing over aerial photographs to highlight possible possum habitats in bush/pasture margins. Sites for control were typically traced on to paper and the areas and perimeters were calculated using a mapping wheel.  The distribution and abundance of possums in these areas would be monitored by observing the rate at which possums were caught in trap lines that were randomly placed[5] in these control areas. 

As well as aerial photos, the addition of satellite images of land cover are now being used as a base for identifying and representing (digitising) possum habitat as a GIS layer. Data from monitoring in the Marlborough high country since 2005 has been used to predict high possum densities, where numbers are likely to host TB. This GIS layer has been overlaid with digital environmental data (including altitude,[6] vegetation class, slope and aspect) to create a model to predict possum densities. This includes predictions in areas that have not been directly surveyed. This type of digital mapping is outlined in Figure 13 below.

This process of mapping possum habitat allows a lower-cost partial control approach where only area of high possum habitat and abundance are selectively targeted. In addition, the ability of the model to predict possum density reduces costs associated with possum monitoring.

Figure 13: Digital mapping for possum control

The impact of the use of GIS based habitat mapping in the Marlborough high country has been to reduce the area requiring possum control by up to 40 per cent, reducing the costs of vector control from $15 per hectare of total area to about $7-8 per hectare.[7] The application of this mapping technology is still in the early adoption stage but is now being widely applied by AHB as discussed later in this report. 

A barrier identified relating to the wider use of habitat mapping involves the inaccuracies of some of the digital environmental data used to create GIS layers. Some of these inaccuracies relate to a lack of real time information. For example, vegetation and forest data uploaded into the Land Cover Database Version 2 (LCDB2) may not show recent changes to vegetation types resulting from land use changes, or may result in the incorrect interpretation of vegetation type when images show vegetation in shadow at certain times of the day.

Methods of Vector Control

Improvements in TB vector control methods have resulted in reduced costs through greater precision in targeting and application. These methods, including the aerial application of bait, and the selection and laying of ground-based trap lines and bait, have increasingly made use of modern spatial information and technology.

In addition to the use of GIS information to determine the most favoured locations for vector control, the use of GPS has allowed greater precision in the site specific application of bait or trap lines.

Hand-held GPS units have been utilised by possum control contractors since the mid 1990s as a means to more quickly and accurately locate, set and check trap lines or bait locations. Prior to this, contractors would be given an initial starting point on the edge of the bush (e.g. track entrance, fence post or some other identifiable position), a direction and a distance to the trap line which had to be located through the use of paper maps and compass bearings.

Hand held GPS is now widely adopted across the industry for possum vector control and monitoring. This has allowed contractors to reduce the time taken to locate trap lines, and reduce the time taken to record location information of traps or baits laid [8] or catch results. This has resulted in an estimated 11 per cent time savings per monitoring contract [9], as well as providing benefits in reducing data entry errors and increasing transparency in the accuracy of reported catch and monitoring results.

GPS guided aerial poison drops by DOC and AHB since the 1990s have also allowed the application of baits at increasingly lower rates. Aerial drops can be sown in narrow strips or clusters of favoured possum habitat rather than being spread more widely at lower densities. Differential GPS systems on aircraft have also increased the precision of application by generating flight paths with calibrated swath widths. Flight data recorded by differential GPS has been used as a check against flight plans to ensure efficient and effective drops. This precision shortens the amount of time an aircraft is in the air, and saves on fuel consumption.

From an average rate of application of 1080 poison of 6.3kg per hectare in 1995[10], application rates had decreased to around 2kg per hectare by 1998 by refining the use of GPS[11]. Application rates have been reduced by a further 60 per cent through the use of the latest digital mapping databases as part of the recent monitoring in the Marlborough high country since 2005. This most recent drop in application rates will equal further cost savings when applied across AHB coordinated operations.

VectorNet

The advances in vector control methods and habitat mapping outlined above are being integrated into the AHB's bovine TB control operations. This is most recently occurring through the use of the VectorNet information system that was completed in early 2008. VectorNet uses a map-based interface to access, query, and report on all aspects of AHB's vector control processes. VectorNet contains a number of geodatabases that staff use for contract management, strategic planning, and reporting purposes.    

With regards to vector control, VectorNet has a central spatial database that is updated with GIS environmental data (including altitude, vegetation class, slope and aspect). Similar to the mapping research outlined earlier in this report, these data are used to calculate areas of possum habitat, to automatically determine the number of traps required in an area to meet the population monitoring requirements, and where they are to be placed in the field. This process is outlined in the figures below.

Figure 14: Generate Trap line - VectorNet

Vector control and monitoring contractors have the capability to use GPS-enabled handheld devices to locate trap lines or bait stations generated by VectorNet, update the database with vector control or monitoring results from the field, then upload information through a Web browser to VectorNet. Data are then validated and added to the reporting geodatabase.

VectorNet has refined the process so that contractor effort is being targeted to habitats and locations where possums are likely to exist. Cost savings are also realised due to ability to predict possum habitat and abundance, which reduces the need to monitor possum abundance and TB incidence in the field to the same extent as before. Preselecting trap lines and numbers will further reduce time taken to locate trap lines and specific trap locations, as these locations are downloaded from VectorNet to handheld devices used in the field. Savings will be achieved in terms of reduced labour time to locate, lay and retrieve traps.

There is an estimated $550,000 annual efficiency gain on the overall vector programme budget from better information provided by VectorNet. AHB has recognised that not all these savings are due to the spatial information or technology per se, but from the coordinative efficiencies provided by VectorNet.

In addition, there is an estimated $1,800,000 annual management cost savings related to the new spatially based VectorNet programme. The number of vector control and monitoring contract arrangements have been reduced from 11 to 6 (reducing the number of contractors from 107 to 45), and staff levels have been increased from 45 to 92 within existing budget levels.

The expected NPV [net present value] for VectorNet is $1.9 million with a payback of 3.3 years. 

These savings are balanced against estimated annual costs relating to spatial information and technology including $70,000 on licences for spatial data, and the necessity to use up to 9 full time equivalent (FTE) employees at a cost of approximately $630,000 per year to maintain data accuracy.     

AHB estimate that additional costs of approximately $5 million would be incurred annually in the absence of the spatially enabled VectorNet system[12]. This cost would include the inability to halt vector control as quickly as currently occurs due to data reporting mechanisms on the proof of control being less effective.

Disease control - the Disease Management Information System (DMIS)

Bovine TB disease control has traditionally been spatially based, involving controls on the movement of cattle from areas containing infected herds, documentation of herd's bovine TB status, and documentation of the history of cattle being moved from any herd or property[13].

A number of bovine TB testing zones have been defined to control the movement of cattle and the spread of disease. These zones are the:

• Movement Control Area (MCA): Areas where more than 1 per cent of herds are infected. Pre movement tests[14] are required plus annual testing of all cattle stock over 3 months. Requires boundaries of the zones to be gazetted and reviewed annually to ensure that at least 1 per cent of herds in each area are infected.
• Special Testing Area (STA): No pre movement tests are required but variable age and frequency testing is dependent on geography and TB risk.
• Surveillance (S): Areas that are TB "free". Triennial testing required over 24 months.

These testing zones historically tended to be hand drawn on physical maps and had text descriptions of physical boundaries. However, with advances in the use of modern spatial information and technology, the process of defining and reviewing boundaries of testing areas and testing the animals within them has been refined.

The Disease Management Information System (DMIS) was created by AHB in 2005, which uses GIS to record the geographic location of herds on each farm as well as each herd's type, test history, TB testing zone, and TB test results by age, sex, and date. DMIS is used to notify farmers when their herd requires testing and records and calculates the number of infected herds in each zone. DMIS has also refined the process for defining and reviewing MCA zone boundaries using an Arc View based GIS package that utilises standard land cover, parcel, and topographical layers. 

Boundary changes are used to make sure that MCA zones are kept as small as possible because of the movement restrictions that apply to them[15].  Arc View is used within DMIS as part of an annual process to review all boundaries by making an assessment based on the herd prevalence and disease incidence in each area, plus the likelihood of any herd becoming infected in the 12 months following a boundary shift. Figures 5 and 6 show how Arc View GIS has been used to present and recommend MCA boundary changes.

VectorNet is also increasingly being used by AHB to coordinate disease control including the administration, movement and control of animals between declared Movement Control Areas.

Cost avoidance

These spatially based GIS tools are being used to assess and define boundary changes to bovine TB testing zones that have varying levels of control within them. This process avoids the cost to the dairy industry of having the restrictions and costs that are applied to farms with infected cattle being applied herd or industry wide.

Costs and restrictions applied to farms with infected cattle include:

• Inability to sell calves off farms
• Inability to milk cows that are reactors
• Inability to graze off herds over winter to other multi-herd grazing units
• Compliance and reporting costs to AHB, including the requirement that two consecutive clear whole herd tests are received, with a minimum of six months between tests and no further evidence of disease. Tests cost $2.30 for a skin test per animal or $35 for a blood test per animal.
• Loss of stock that need to be slaughtered, with 65 per cent of the market value of animals being compensated to farmers. This compensation is paid for by industry levies. There is a population of 5,261,000 dairy cattle in New Zealand in 2007 with a market value ranging from $475 to $1,425 per animal [16].

Rather than an industry-wide shut down, more precise zoning of the disease incidences (e.g. into MCA, STA and surveillance zones) has allowed trade to continue and costs to be avoided compared to what would otherwise have been the case if there was an industry wide shutdown.

Furthermore, farms with non infected herds have avoided the costs and restrictions applied to farms with infected cattle through the definition of the bovine TB testing zones and the management of animals within them.

Figure 15: Movement control areas managed in GIS

This case study has been taken from the Spatial Information in the New Zealand Economy - Realising Productivity Gains report, August 2009.

---

References

[1] http://TBfree.ahb.org.nz/Default.aspx?tabid=118.

[2] http://www.landcareresearch.co.nz/research/programme.asp?Proj_Collab_ID=7.

[3] This figure has been set by the Office Internationale Epizooties (World Organisation for Animal Health).

[4] DOC controls possums on 1 million ha of public conservation lands where its priorities are highest, while the AHB has controlled possums on 4.5 million ha, of which c. 13 per cent has been estimated as being on public conservation lands.

[5] determined by using grid overlays or random number generation.

[6] Possums are likely to live at lower altitudes.

[7] Landcare Research Manaaki Whenua Interview.

[8] Recording trap or bait sites and catch results on GPS units replaces the manual recording of 14 digits on a recording sheet to be transcribed on a summary sheet.

[9] Possum Control and Monitoring Contractor Interview.

[10] Warburton. B, Cullen. R 1995: Cost-effectiveness of Different Possum Control Methods SCIENCE FOR CONSERVATION: 4. Department of Conservation, Wellington.

[11] Gillren, D 1999: GIS and Possum Control on Mount Karioi. Presented at SIRC 99 – The 11th Annual Colloquium of the Spatial Research Centre University of Otago, Dunedin.

[12] Animal Health Board Interview.

[13] National Bovine Tuberculosis Pest Management Strategy National Operational Plan: 1 July 2005 – 30 June 2013.

[14] TB tests required up to 60 days prior to movement of cattle off farms.

[15] For example, because the MCA is defined by having greater than one per cent infected herd prevalence, as the overall number of infected farms in the MCA decreases (and the overall percentage decreases), a non-infected farm near the boundary may cease to be in the MCA where a boundary shift inward will still allow the overall rate of infected herds to remain at over one per cent.

[16] PGG Wrightsons Ltd: Maximum Valuation Limits for Dairy TB Reactors. Dairy valuations from 1st May 2009. Approved PGG Wrightsons Ltd.