Crown development in tropical rain forest trees in gaps and understorey

Abstract

Crown development was monitored for juvenile trees (4–18 m) of the canopy tree species Dicorynia guianensis and Vouacapoua americana in a tropical rain forest in French Guiana. A comparison was made between crown development in closed understorey forest sites (low light) and in canopy gaps (high light); development was expressed by plant traits at multiple levels of organisation. Dicorynia and Vouacapoua responded to gap creation at all organisational levels, but not for all traits. Both species increased their light interception efficiency in the understorey. Firstly, understorey trees produced relatively wide crowns because they did not favour the growth of shoots in the summit of the crown, as did gap trees. Secondly, they reduced self-shading (expressed by leaf area index) by lower sympodial unit production rates, lower leaf production rates per growth unit (only Vouacapoua), and smaller leaf size (only Dicorynia). The reduction in self-shading was weakened by other traits that had the opposite effect on self shading (e.g. longer leaf life span in understorey), and cannot be considered an adaptive response in itself. (3) Understorey trees of Vouacapoua reduced leaf display costs by producing a higher specific leaf area, a shorter space among leaves, a smaller leaf spacing to leaf size ratio, and a longer leaf life span. Thus, all traits contributed to the more economical use of carbon in the understorey. Dicorynia showed the same trends, but not significantly. These results suggest that light availability plays a major role in the development and morphology of trees through its influences at multiple levels of organisation within the crown hierarchy. The two species studied were rather similar in their response at crown level, while they sometimes differed in their responses at the underlying lower organisational levels. The response directions of some individual plant traits were similar for the tall trees studied here and small saplings studied elsewhere. Some of traits investigated here may not be important in the crown level responses of smaller seedlings and saplings, while they are important mechanisms for crown level responses in taller trees.