Figure 5. Measures of clustering in a Skeeter Buster simulated population using point-pattern analysis L statistic.

We calculate L values for 20 replicate Skeeter Buster simulations. This statistic is calculated (1) for the number of pupae within a house, noted L_w(d), and (2) for the houses themselves, being non-randomly distributed, noted L(d). Here we plot the difference between the increment in L_w(d) and the increment in L(d) – that is, (L_w(d) - L_w(d-1)) – (L(d) - L(d-1)) for all values of d. The distance d between two houses in this model is defined as the number of steps (horizontal or vertical only) separating these two locations in the grid. The existence of a significant cluster of size d is marked by a positive value of this difference, while a value of 0 is expected under random distribution.