Skip to main content
. 2016 Feb 1;25(5):575–585. doi: 10.1111/geb.12438

Figure 2.

figure

The ability of the two models to quantitatively predict the observed patterns of the Barro Colorado Island (BCI) forest that correspond to the three rules of McGill (2010). The species aggregation Mi(r) measured at 10 and 50 m represents rule 1 (species are aggregated in space), the species abundance distribution (SAD) represents rule 2, and rule 3 is fulfilled by default since the model does not incorporate species interactions (see text). The panels on the top row show the predictions of the neutral model, while the panels on the bottom show the predictions of the extended model that includes species‐specific dispersal. (a), (d) The SAD: The dashed line and the grey area represent mean and 95% simulation envelopes for the model predictions using the best 1000 parameter sets, while the black solid lines show the six BCI censuses (1985–2010). (b), (e) Distribution of species‐specific aggregation indices M i(r) at the scale of 10 m. The grey lines show the 1000 model predictions and the black solid lines the six BCI censuses. The vertical dashed line indicates complete spatial randomness. See the main text for the interpretation of the M i(r) index. (c), (f) The same as panel (b) but for the neighbourhood scale of 50 m.