Key message Environmental filters select species with different characteristics. However, these do not exhibit the functional integration of the whole plant level. Abstract Functional integration at the whole plant level, between the strategies of the various plant organs, needs to be better elucidated to facilitate the understanding of species ecology over an environmental gradient. Thus, hypotheses were proposed that (1) if functional integration exists between organs, the leaf economic spectrum will be a good indicator of the plant economic spectrum, and that (2) processes of community assembly will change over a successional gradient, generating different economic spectra for the plant. To test these predictions, data on eight functional characteristics (leaf, reproductive, stem and whole plant) was collected in 65 species, distributed in six secondary growth forests (three with approximately 17 years and three with approximately 25 years of abandonment) and three mature forests, located within the dense ombrophilous forest domain in the northern coastal region of Pernambuco, Brazil. The results showed no influence of geographic distance (autocorrelation) or phylogeny. Positive correlations were observed between leaf nutrient concentration (N and P), and negative correlations were observed between leaf nutrient concentration and dry matter content. Functional integration at the whole plant level was not observed. On the primary axis of the principal components analysis, only leaf characteristics and wood density were coordinates, following the same strategy, whilst on the secondary axis, only leaf characteristics and maximum height were coordinates. The observed change in species composition and abundance over the gradient was sufficient for changes in the distribution of functional characteristics and, therefore in strategies, to be observed.

Traits and functional strategies when related to demographic rates provide important information about the distribution of species and their performance in different environments. Therefore, predictions were made that: (1) plant communities in tropical forests, distributed over a chronosequence, have different trade-offs between demographic rates and functional traits due to variations in light availability and (2) acquisitive, intermediate and conservative species have different demographic rates throughout the chronosequence. To test these predictions, data from eight functional traits (maximum plant height; specific leaf area; leaf dry matter content; leaf nitrogen, phosphorus and chlorophyll concentrations; wood density; and seed shape) were collected from 64 tree species, distributed in six secondary (three 17 and three 25 years old) and three mature ([40 years old) Ombrophilous Dense Forest fragments. Demographic rates (mortality, recruitment and growth) of the species were determined for periods of three and five years, in each forest. Mortality was higher and growth rate was lower in the 17-year-old than in the mature forest, and the acquisitive group had the highest recruiting rate, but interactions were not significant. Thus, the majority of relationships between demographic rates and functional traits did not correspond to the predictions, as some traits were not better predictors of the demography of species in a determined forest than others and that these relationships did not vary across the chronosequence. Therefore, the trade-offs between functional and demographic characteristics are related to the species and not to environmental differences across the age gradient, and when the species are separated into groups, the predicted changes in demography are valid across the gradient.

No abstract

No abstract

No abstract