Disease Dynamics and Control in the presence of a reservoir

Livestock diseases are often accompanied by a wildlife reservoir in which the disease is not managed or even often observed. These reservoirs can complicate the management or containment of disease outbreaks as they can encompass several farms facilitating an unobserved pathway of transmission between farms. Livestock in the UK are routinely tested for several diseases and undergo pre-movement testing at an individual level (cattle) or batch level (sheep). The records of these tests and movements create valuable resource for understanding the transmission of infection during an outbreak. However, the presence of a wildlife reservoir adds a potential for transmission that is difficult to quantify and cannot be seen in these movement records.

Testing and culling of infected livestock represents a considerable expense for the UK (it costs approx. £100M to deal with bovine TB each year in the UK) and any curtailment in the spread of infectious diseases will have a significant [positive] impact on the UK taxpayer.The aim of this project is to develop a mathematical model of an infectious livestock disease, incorporating a network of farms and reservoirs, to determine the optimal vaccination, testing or eradication programmes to combat the spread of the disease. We will test this model against commercially important livestock diseases (bovine TB, avian influenza, foot and mouth disease) where the farm, movement and outbreak data are available.

Dr O’Hare has contacts within DEFRA and Scottish Government that can supply the disease outbreak data that the model can be tested against. There is understandable public interest in how wildlife is treated, especially when they are culled in favour of protecting livestock. This research will have a non-academic impact within the community of users of the countryside in that we aim to demonstrate regimens that provide maximum protection to livestock and minimum disruption to wildlife.

Proposed research programme
This goal of this research is to model the dynamics of how disease spread in the presence of a reservoir and develop optimal control strategies (for both reservoir and livestock). We will test the model against several important livestock diseases. There are many scales of interaction involved in this approach:
•Within-farm transmission - We will use an agent based model to simulate the transmission between animals on the same farm.
•Between-farm transmission –We will model the transmission between farms via animal movements and an exponential transmission kernel to model over-the-fence transmission. Detailed records are kept for all animal movements within the UK which we will utilise in applying this model to real life situations. We will also simulate the pre-movement testing within our model.
•Within-reservoir transmission –we will model the transmission dynamics within the reservoir. The model may be different to the within-farm model (not least because the wild reservoir is not routinely checked).
•Between-reservoir –we will simulate the migration/territorial fighting of animals within a reservoir to simulate over -the-fence transmission between reservoirs.
•Reservoir-Farm/Farm-Reservoir transmission –we will also simulate the transmission of disease from farm to reservoir and vice-versa. Reservoirs will typically span several farms which gives an unobserved transmission pathway between farms.
• Seasonal effects will be considered as [some] farm animals are housed over winter allowing greater within-herd disease transmission, and animals that make up the reservoir may change their behaviour in each season.
Once we have a model that can simulate a realistic epidemic we will use it to investigate various control strategies given realistic inputs such as culling wildlife, increased testing, different testing strategies, improved test sensitivities, biosecurity (removing or decreasing the interaction between farm animals and the reservoir).

The first milestone will be the creation of a simple SIR model within the farm and an SI within the reservoir and will include farm -farm movements. We will use a deterministic model to observe the effect of the contact between the farm and reservoir and determine the most effective culling, biosecurity, and testing strategies. We will back this up by simulations and publish the results. We will then apply this model to a specific livestock disease given available data to demonstrate the impact of different management/containment strategies and identify the most appropriate.