Chalmers Conferences, 9th European Conference on Mathematical and Theoretical Biology

The ecology of wildlife disease surveillance
Glenn Marion

Last modified: 2014-03-28


We develop a generic mathematical approach in which surveillance of wildlife disease systems is characterised in terms of key demographic, epidemiological and surveillance parameters. Mathematical analysis and computational tools are deployed to show that demographic fluctuations and stochasticity in transmission dynamics can lead to bias in the estimates of prevalence obtained from disease surveillance in wildlife disease systems. Moreover, such fluctuations also reduce the precision of prevalence estimates compared with what would be expected from standard arguments based on the assumption of constant prevalence.  Bias and precision of surveillance estimates are shown to be co-determined by demographic factors and disease dynamics.  The effect of surveillance design is considered in terms of sample size and the level of effort deployed (capture rate). Bias is unaffected by the number of animals sampled but precision of estimates increases at a faster rate than if fluctuations are ignored. In contrast, increasing the capture rate reduces the bias of prevalence estimates, whilst for very rapid sampling the precision of surveillance estimates corresponds to the variance in prevalence of the underlying wildlife disease system.  Fluctuations in prevalence are shown to reduce the power of surveillance to detect disease in wildlife to a considerable extent. 


wildlife disease surveillance; stochasticity; demographic fluctuations