Table S1. Summary of the distributions of the proportion of the population harvested each year and supporting data. Figure S1. Histogram and probability density of the estimated proportion of the population harvested each year, autocorrelation and trace plots. Table S2. Summary of the estimated distributions of abundance and supporting data at each site. The Geweke test was -0.557 (P=0.577) for the wild boar and 1.085 (P=0.278) for the red deer models. Figure S2. Histogram and probability density of the abundance estimated at each site, autocorrelation and trace plots of the Markov Chain Monte Carlo simulations. Table S3. Summary of the estimated distributions and supporting references for the sensitivity and specificity of the diagnostic tests. Figure S3. Histogram and probability density of the estimated sensitivity and specificity of the diagnostic tests employed, autocorrelation and trace plots of the Markov Chain Monte Carlo simulations. Table S4. Summary of the estimated distributions of the true prevalence by host species and supporting data at each study site. Figure S4. Histogram and probability density of the estimated true prevalence at each site, autocorrelation and trace plots of the Markov Chain Monte Carlo simulations. Figure S5. Histogram and probability density of the estimated probability of Mycobacterium tuberculosis complex excretion from infected wild boar and red deer, autocorrelation and trace plots of the Markov Chain Monte Carlo simulations. Table S5. Summary of the estimated distributions of the R0 by host species at each study site. Figure S6. Histogram and probability density of the estimated R0 by host species at each study site, autocorrelation and trace plots of the Markov Chain Monte Carlo simulations. For intelligibility, only R0 estimates 1; Obligatory multi-host: R0 both species1. Highest probability for each site is highlighted in bold. Figure S7. Diagnostic plots of the generalised linear models. Diagnostic plots for the wild boar, red deer, and multi-host models with the whole dataset and after removing one influential outlier (site 15). Table S7. Summary of the generalised linear model selection. Models with ΔAICc<2 from the most supported model. Figure S8. Basic reproduction number estimated for the wild boar and red deer in each of the study sites. Kernel density of 100 estimates of the R0 of Mycobacterium tuberculosis complex in the wild boar and red deer drawn randomly from the 100,000 estimates at each site. Axes in square root scale, R0=1 as blue horizontal and vertical lines.

The datasets are reported in the supplementary material Tables S1-S4.

Peer reviewed