One size does not fit all: Exploring the economic and non-economic outcomes of on-farm biosecurity for bovine viral diarrhoea virus in Australian beef production.

Bovine viral diarrhoea virus (BVDV) is a disease of global importance, affecting the production and welfare of cattle enterprises through poor reproductive performance and calf mortality. In Australia, the prevention of BVDV introduction and spread is primarily achieved with on-farm biosecurity; however, the use of these practices can vary amongst producers. Economic utility is commonly identified as a contributor to the uptake of on-farm biosecurity, but other factors such as animal welfare, producer priorities and introduction risk also influence farmer behaviour. This study uses an individual-based, stochastic simulation model to examine the economic and non-economic value of 23 on-farm biosecurity combinations for the control of BVDV in Australian beef farms without (N0) and with (N1) a neighbouring population of persistently infected (PI) cattle. Combinations of quarantine of purchased bulls (Q), hygiene during herd health events (H), double-fencing adjacent boundaries with neighbouring farms (F) and vaccination against BVDV (V) were tested. This study is the first to simulate the use of strategic PI exposure (PI) as an alternative to V, a contentious practice performed by some Australian beef farmers. Introduction of BVDV into a naïve 300-breeder self-replacing beef herd was achieved through the purchase of PI bulls (N0 and N1 herds) and over-the-fence contact with neighbouring PI animals (N1 herds only). The predicted median cumulative loss due to BVDV over a 15-year period was AUD$172/breeder and AUD$453/breeder for an N0 and N1 herd, respectively. Early establishment of BVDV in the simulation period was found to be the primary factor contributing to economic loss. Consequently, the Q and QF combinations resulted in the highest predicted average annual cost-benefit for BVDV-free N0 and N1 herds. In the five years following establishment of BVDV, use of QP (N0 herds) and V (N1 herds) combinations were most cost-effective. Combinations that involved V and P (in conjunction with F in N1 herds) also resulted in the lowest number of PI animals sold to other farms or feedlots over the simulation period. However, in both N0 and N1 herds, P resulted in the highest number of infected cattle, which has implications for poor animal welfare and increased antimicrobial use on Australian beef farms. The outcomes reported in this study can guide decisions to prevent BVDV introduction and spread on extensive beef farms using on-farm biosecurity, based on the risk of BVDV exposure and the priorities of the individual farmer.

Stakeholder mapping in animal health surveillance: A comparative assessment of networks in intensive dairy cattle and extensive sheep production in Australia.

The capacity to rapidly identify and respond to suspicion of animal disease is fundamental to protecting the integrity of the Australian livestock industry. An incursion of a nationally significant endemic, emerging or exotic animal disease could be disruptive and economically damaging for the industry, broader community and national economy. To counter this potential threat, a surveillance system that includes general and targeted activities exists at a jurisdictional and national level. Such a system requires a collaborative effort from all involved to work towards a common goal, reflecting the notion of shared responsibility. As in all systems, the animal health surveillance system can be enhanced or constrained by the relationships of the players involved. This study focusses on two livestock industries, dairy cattle and sheep, exploring the interrelationships between all stakeholders, and their role within the Australian animal health surveillance system. A stakeholder mapping exercise was undertaken, including a depiction of the perceived level of stakeholder interest and influence on producers' animal health surveillance practices and/or the surveillance system. Results from these activities were expanded upon through interviews. The findings reveal complex networks and a system that is, at times, constrained by institutional and individual barriers such as communication between and within stakeholders, and uncertainty about the consequences of reporting a suspected emergency disease. Whilst these challenges have the potential to negatively impact the robustness of the animal disease surveillance system, the study also provides clear evidence of strong and effective relationships amongst many of the key individuals and organisations.

Pig traders' networks on the Kenya-Uganda border highlight potential for mitigation of African swine fever virus transmission and improved ASF disease risk management.

We applied social network analysis to pig trader networks on the Kenya-Uganda border. Social network analysis is a recently developed tool, which is useful for understanding value chains and improving disease control policies. We interviewed a sample of 33 traders about their experiences with trade and African swine fever (ASF), analyzed the networks they generated in purchasing pigs and selling pork and their potential contribution to modulating dissemination of the ASF virus (ASFV). The majority of the traders were aware of clinical signs of ASF and the risk of trade transmitting ASFV. Most said they avoided buying pigs from ASF outbreak villages or sick pigs but their experiences also indicated that inadvertent purchase was relatively common. Traders had early knowledge of outbreaks since they were contacted by farmers who had heard rumours and wanted to sell their pigs to avoid the risk of them dying. Individual traders bought pigs in up to nine villages, and up to six traders operated in a village. Although each trade typically spanned less than 5km, networks of the various traders, comprising movements of pigs from source villages to slaughter slabs/sites and retail outlets, and movement of pork to villages where it was consumed, linked up indirectly across the 100km×50km study area and revealed several trade pathways across the Kenya-Uganda border. ASF could potentially spread across this area and beyond through sequential pig and pork transactions. Regulation of the pig and pork trade was minimal in practice. The risk of ASFV being spread by traders was compounded by their use of poorly constructed slaughter slabs/sites with open drainage, ineffective or non-existent meat inspection services, lack of provision for biosecurity in the value chain, and sales of pork to customers who were unaware of the risks to their own pigs from contact with ASF infected pork. More effective regulation is warranted. However, limitations on government capacity, together with the strong self-interest that established traders have in reducing the disruption and financial losses that outbreaks cause, highlight the importance of governments and traders co-developing an approach to ASF control. Formation of trader organizations or common interest groups warrants government support as an important step in engaging traders in developing and implementing effective approaches to reduce the risk of ASF outbreaks.

Social network analysis provides insights into African swine fever epidemiology.

Pig movements play a significant role in the spread of economically important infectious diseases such as the African swine fever. Characterization of movement networks between pig farms and through other types of farm and household enterprises that are involved in pig value chains can provide useful information on the role that different participants in the networks play in pathogen transmission. Analysis of social networks that underpin these pig movements can reveal pathways that are important in the transmission of disease, trade in commodities, the dissemination of information and the influence of behavioural norms. We assessed pig movements among pig keeping households within West Kenya and East Uganda and across the shared Kenya-Uganda border in the study region, to gain insight into within-country and trans-boundary pig movements. Villages were sampled using a randomized cluster design. Data were collected through interviews in 2012 and 2013 from 683 smallholder pig-keeping households in 34 villages. NodeXL software was used to describe pig movement networks at village level. The pig movement and trade networks were localized and based on close social networks involving family ties, friendships and relationships with neighbours. Pig movement network modularity ranged from 0.2 to 0.5 and exhibited good community structure within the network implying an easy flow of knowledge and adoption of new attitudes and beliefs, but also promoting an enhanced rate of disease transmission. The average path length of 5 defined using NodeXL, indicated that disease could easily reach every node in a cluster. Cross-border boar service between Uganda and Kenya was also recorded. Unmonitored trade in both directions was prevalent. While most pig transactions in the absence of disease, were at a small scale (<5km) and characterized by regular agistment, most pig sales during ASF outbreaks were to traders or other farmers from outside the sellers' village at a range of >10km. The close social relationships between actors in pig movement networks indicate the potential for possible interventions to develop shared norms and mutually accepted protocols amongst smallholder pig keepers to better manage the risk of ASF introduction and transmission.