Fig. 3.

(A) The cumulative PDF of tree diameters in the BCI forest (1995) (24). is the fraction of trees with diameter ≥r. The black dots correspond to diameters in the interval from approximately 2.4–31.8 cm (the range over which scaling is expected to hold from the scaling collapse plot in Fig. 2). The solid line indicates our predicted exponent of -4/3 (derived from a power law probability density with exponent -7/3). The dashed line depicts the exponent of -1 (corresponding to a probability density with exponent -2) and is shown for comparison. (B) Plot of the average distance from the nearest neighbor individual in the same size class, d_s, versus the tree diameter. We divided the BCI tree-diameter dataset (24) in 2 cm bin size and for each bin, we calculated the average distance between nearest neighbor trees belonging to the same bin. The solid line shows the predicted power law behavior with exponent 7/6, whereas the dashed line is a power law with exponent 1. Our prediction follows from the assumption that the trees in a given size class are distributed uniformly across the forest, thus implying that d_s ∼ P_r(r|r_c)^-1/2 ∼ r^7/6.