Fig. 3

The distribution of individuals during population spread. a–c High Density treatment and d–f High Density + Diffuse hosts treatments (patterns were similar in other treatments; see Additional file 1: Fig. S2). Histograms show the observed distribution of individual detections over all replicates, at three different times (15, 380 and 460 min). A time window of ± 5 min, centred on the focal time, was used for better visualisation. Curves show the best fitting Gaussian (blue) and 2-component Gaussian mixture (G-mixture (2); orange) models. AIC values of Gaussian, G-mixture (2), G-mixture (3) and Student (t) models are also given in each case (AIC was computed with no time-window, to avoid pseudoreplication). Dashed lines represent the two components of the G-mixture (2) model, i.e. the narrow (“resident”) and broad (“explorer”) components