Does adult stature of a tree species determine how its saplings behave in the understory? A long-held hypothesis states that saplings of canopy or emergent trees should invest disproportionally into growing tall (optimistic strategy), while those of understory species should start expanding their crown early on (pessimistic strategy). Implicit is the assumption that there is a height-growth, crown-expansion trade-off. Although prior evidence suggests this might be the case, we lack a community-wide test of this hypothesis and assumption, especially in hyperdiverse forests. We sampled 915 saplings from 396 species with diameter at breast height (D) between 1 and 5 cm found in mature stands on the plateau of a 25-ha permanent plot near Manaus, Brazil. For each sapling, we measured plant height and the crown's major and minor projected horizontal axes. Because larger saplings are taller and have wider crowns, we used D to control for their size. For each species, we estimated adult size as the 97.5 percentile of the D distribution in the plot. To test for a vertical-horizontal growth trade-off, we ran a hierarchical, multivariate Bayesian analysis which estimated species-specific relationships (power-law models) between D and height (H-D), and between D and mean crown width (W-D).
Results/Conclusions
After taking saplings' size into account, height and width were negatively correlated, indicating a within-species height-growth, crown-expansion trade-off. This trade-off, however, is not evident between species as H-D parameters were not correlated to any W-D parameters. This indicates that while some species' saplings are both taller and wider than average given their size, others do poorly in both dimensions, which possibly reflects fitness differences among species. Additionally, species-specific slopes were strongly negatively correlated to their intercepts for both H-D and W-D, indicating that species that have a head start in H or W will have a slower long-term growth in that dimension compared to species that start smaller. At the community level, H-D and W-D estimated slopes were both 0.75, the scaling factor predicted by metabolic theory. Finally, adult size was a poor predictor of sapling growth strategies, in accordance with our own observation that some local understory species show clear optimistic strategies while some canopy species present the opposite strategy. Our results therefore undermine the view that saplings behave according to their future adult stature and nuance an intricate, yet more intriguing story about how saplings grow and interact with each other in the understory of tropical forests.