Extended Data Fig. 6: The scaled per-capita population growth rate \({\tilde{{\boldsymbol{\lambda }}}}_{{\boldsymbol{f}}}({{\boldsymbol{N}}}_{{\boldsymbol{s}}})/{{\boldsymbol{r}}}_{{\boldsymbol{f}}}\), as influenced by spatial structure, niche differences and immigration for 21 ForestGEO plots.
From: Latitudinal scaling of aggregation with abundance and coexistence in forests
a, the scaled per capita population growth rate \({\widetilde{\lambda }}_{f}({N}_{s})/{r}_{f}\) for a small abundance of Ns = 10 individuals, per forest plot for scenario 1 (no immigration: vf = 0, no niche differences: βfi = βff) that represents the “pure” effect of spatial structure. b, same as a), but for scenario 2 that adds niche differences (vf = 0, βfi < βff). c, same as a), but for scenario 3 adds immigration to scenario 1 (vf = 0.1, βfi = βff). d, scenario 4 adds niche differences and immigration (vf = 0.1, βfi < βff). e: \({\widetilde{\lambda }}_{f}({N}_{s})/{r}_{f}\) for scenario 1 in dependence on abundance Nf, averaged separately over all species of tropical, subtropical and temperate forests. The dashed vertical line indicates the small abundance of Ns = 10. f, same as e), but for scenario 2. g, same as e), but for scenario 3, and h, same as e), but for scenario 4. In scenarios 2 and 4 we assumed that more closely related species compete more strongly. The models were parameterized for 720 species of the 21 ForestGEO plots. We excluded the temperate forest at CBS with exponent bf < −1. To outline the overall tendency in the data we fitted in panels a) to d) a linear regression to the data.
