Fig 1. Examples of three scenarios for detecting loss of relative high loads of infectious agents using quantile-quantile plots.

Load distributions are contrasted between sequential seasons. For all three scenarios, the slope is significantly less than 1 (i.e. 1:1 line; dashed line). The different scenarios account for any potential infectious agent dynamics in the host and the consequent displacement relative to the 1:1 line (where loads would be similar between seasons): a) no seasonal incremental change in loads overall; b) an equal seasonal incremental increase in all loads; c) an incremental increase in small loads that is higher relative to larger loads. Grey, arcing arrows show the rotational effect of selective high load loss; black arrows and brackets denote the translational effect of increases in load between season 1 and season 2.