Abstract
Tree functional traits and their link to patterns of growth and demography are central to informing trait-based analyses of forest communities, and mechanistic models of forest dynamics. However, few data are available on how functional traits in trees vary through ontogeny, particularly in tropical species; and less is known about how patterns of size-dependent changes in traits may differ across species of contrasting life-history strategies. Here we describe size-dependent variation in seven leaf functional traits and four wood chemical traits, in two Dominican rainforest tree species (Dacryodes excelsa Vahl. and Miconia mirabilis (Aubl.) L.O. Williams), ranging from small saplings to the largest canopy trees. With one exception, all traits showed pronounced variation with tree size (diameter at breast height, DBH). Leaf mass per area (LMA), thickness and tissue density increased monotonically with DBH in both species. Leaf area, leaf nitrogen (N) and carbon (C) : nitrogen (N) ratios also varied significantly with DBH; however, these patterns were unimodal, with peak trait values preceding the DBH at reproductive onset in both species. Size-dependent changes in leaf structural traits (LMA and leaf thickness) were generally similar in both species, while traits associated with leaf-level investment in C gain (leaf area, leaf C : N ratio) showed contrasting ontogenetic trends between species. Wood starch concentration varied with DBH in both species, also showing unimodal patterns with peaks preceding size at reproductive onset. Wood C concentration increased linearly with DBH in both species, though significantly only in M. mirabilis. Size-dependent patterns in wood chemical traits were similar between both species. Our data demonstrate pronounced variation in functional traits through tree ontogeny, probably due to a combination of environmental factors and shifts in resource allocation. Such ontogenetic variation is comparable in magnitude with interspecific variation, and so should be accounted for in trait-based studies of forest dynamics, structure and function.