Trees and light; Tree development and morphology in relation to light availability in a tropical rain forest in French Guiana.

Tropical rain forest trees spend their life in a heterogeneous light environment. During their life history, they may change their growth in relation to different levels of light availability. Some of their physiological processes (e.g. photosynthesis, carbon allocation, and meristern activity) change with light availability, and tune their development and morphology to the ambient light levels. The underlying physiological processes are not investigated in the present study. The focus is on the development and morphology of trees of canopy species in relation to the light availability in tropical rain forest. The possible consequences for survival, growth, and reproduction of trees are not assessed directly, but are discussed on a speculative basis.The relationships between the light environment, tree development, and morphology are investigated for trees of different size, ranging from small saplings to trees of adult stage. Trees of increasing size are compared in order to explore the changes in tree development and morphology, and their relation to the light environment, with ontogeny. Ontogeny is referred to as the overall growth and development pattern during tree life, both for individual trees and (in more general) for a given tree species.The field work for this thesis was carried out in French Guiana. This country in the north-east of South America has an area of 83.000 km <sup>2</SUP>and is covered by evergreen tropical rain forest. The field work was conducted at two biological stations. 'The Piste St. Elie' station is located 30 km from the coast, south of the town of Sinnamary, and the biological station 'Les Nouragues' is located 100 km from the coast, south of Cayenne. Two canopy tree species were selected for this study: Dicorynia guianensis Amshoff. (Caesalpiniaceae) and Vouacapoua americana Aubl. (Caesalpiniaceae). Both are common species in the forests of French Guiana, and are considered late successionals or shadetolerant species (Schulz 1960). In some chapters, these species are compared with an early successional (light demanding) species, Goupia glabra Aubl. (Celastraceae). Trees of these three species are harvested for their timber in French Guiana and its surrounding countries.The trees that were shorter than 20 rn had not yet reached the open upper canopy. These trees usually occur at relatively low light levels. Although these trees may differ in height (from 0 to 20 m), they usually show the same type of growth response to ambient light levels. They produced more growth units and more leaves at higher light availability. They thus increased their total leaf area and leaf area index (LAI, a measure for the number of leaf layers in the crown) as a response to higher light levels. Under persisting high light levels, the increase in total leaf area may enable these trees to fix more carbohydrates (i.e. carbon) for successful growth and survival in the future. Trees with a high LAI at higher light availability, in combination with more columnar shaped crowns, achieve net photosynthesis (more carbon intake than consumption by leaves) at the least possible cost for leaf area support. In contrast, trees with more planar crowns and lower LAI at lower light availability may avoid self-shading of leaves, but risk higher costs for leaf area support.Trees also produced shorter growth units at lower light availability, and thus spaced their leaves at shorter distances than trees at higher light availability. In more closely spaced leaves, the investments for the support of one leaf are lower. As leaf size did not change in relation to light availability, trees displayed their leaf area more economically (at lower carbon costs) at lower light availability. In this way, they increased light interception per unit of fixed carbon, and they may thus be better able to survive the shade.Dicorynia and Vouacapoua trees also grew faster in height with increasing light availability. In general, trees may reduce their height growth because low light levels simply limit growth. At low light levels, trees are shaded by taller neighbouring trees which intercept the majority of light above them, but they may survive for some time by producing their leaf area slowly and efficiently. When light levels increase because one (or more) of the taller neighbours falls down, trees start to increase their height growth, and may compete with their neighbours for newly available light and space. For both species studied, it was shown that height growth further increased at very high light levels in large gaps through preferential growth of the leader (axis which supports the uppermost apical meristern of the crown) over the other axes in the crown. At lower light levels, individuals did not show preferential growth of the leader. Thus, height growth increased not only because the higher light levels are less growth limiting, but also because of preferential growth of the leader.These growth responses to light refer to trees (up to 20 m tall) that were still heading for the canopy. The taller trees (heights of sampled trees range between 26 and 37 m) at higher light availability in the upper canopy had a larger total leaf area and total branch length than the trees shorter than 20 m. These taller trees also produced larger and more planar shaped crowns, did not further increase their LAI, and decreased their leader growth and the space between leaves, as compared with the smaller individuals. The shift to a wider crown is probably caused by increasing light (and space) availability, and may constrain a further increase in the LAI (the leaves occurred over a much larger horizontal area). The lower leader growth and the production of leaves at shorter distances indicate that these taller trees changed from investments in vertical expansion to investments in the replacement of leaves (and flowers).The increasing stature with ontogeny has to be balanced by mechanical strength (thickness). This strength is needed to carry the increasing tree weight and to resist wind stress. The mechanical design expresses the balance between overall tree stature (in terms of weight or wind force experienced by the tree) and tree (mechanical) strength. The changes in mechanical design with ontogeny were investigated for Goupia, Dicorynia, and Vouacapoua using two models. (1) The elastic-stability model emphasises the mechanical strength against its own weight. Using this model, it was shown that trees of the study species decreased their 'safety margins' (strength) early in ontogeny, but increased their margins of safety later. Trees had their lowest margins at a stem diameter of 15 to 25 em. These margins were close to the theoretical minimum, i.e. trees would buckle under their own weight if they were slightly more slenderly built (taller at a given diameter). In comparison with some temperate tree species, the trees of the present study appeared to have lower safety margins because they were more slender. Slenderness (height/diameter ratio), however, is only one of the factors determining the strength of a tree. The denser and stiffer wood of tropical trees may increase the mechanical strength of tropical trees in comparison with temperate trees. Another explanation for the lower safety margins of tropical trees is that they are exposed to lower external stress forms than temperate trees. Temperate trees experience heavy storms and snow loads during their life, whereas the trees of the present study do not experience such forms of stress. (2) The constant-stress model emphasises the mechanical strength over wind stress on the tree. For the species of study, it was shown that the safety margins against wind stress increased with ontogeny. This was in line with the expectations, because wind stress is likely to increase with increasing tree stature. Finally, the influence of light availability on mechanical tree design could not be investigated. The trees studied had long life-histories under unknown light conditions, and therefore did not show a significant response within the 2-3 years of investigation.The ecological knowledge on commercial tree species presented in this work is thought to be useful for the fine tuning and improvement of forest management systems. In these systems, canopy gaps of different size are created, and they may affect the growth of trees. The results of this thesis indicate that manipulations of light availability (either by killing dominant trees thus inducing light level increase, or by shading) in forests may increase the timber production in trees. Besides this, the follow-up to this study may provi de morphological traits that can be used to indicate the growth potential of trees in relation to the light environment. It is suggested that there is a need for knowledge on the growth response of trees (both in terms of timber production, morphology and development) to the whole range of light availability. Manipulations of the light environment may then be tuned to individual trees of commercial species in order to approach the light conditions that provoke the desired growth response. In general, fundamental research on tree growth in a natural habitat, extended by research on tree growth at higher light levels outside the natural habitat, may provide valuable insights for the improvement of forest management systems