Skip to main content
Figure 6 | Veterinary Research

Figure 6

From: Neighbourhood contacts and trade movements drive the regional spread of bovine viral diarrhoea virus (BVDV)

Figure 6

Drivers of the annual probability of acquiring infection per causal transmission route (A and C trade movements, B and D neighbourhood contacts). Newly infected herds had at least one transiently (T) or persistently (P) infected animal over the year while none the previous year (Prob_getting_inf). A, B Simulated vs. predicted probability by the Random Forest regression. Each point represents one herd-year. The green and red solid lines are respectively the regression line and the nonparametric-regression smooth. The red dotted lines are the estimations of the variance function. C, D Explanatory variables ranked by importance as measured by the Random Forest method (using %IncMSE, Additional file 1H). Variables with larger values are more important. Variables are: number of herds in the neighbourhood (nNB), annual number of animals purchased (InStrength), annual number of animals of less than 1 year purchased (InStrength_Less1yr), annual number of source herds (InDegree), annual number of animals purchased by neighbouring herds (InStrengthNB), annual number of animals of less than 1 year purchased by neighbouring herds (InStrengthNB_less1yr), proportion of cows aged 24–36 months (%24_36), herd size (SizeTot), number of females younger than 1 year (Size1yr). Analyses were based on 100 runs.

Back to article page