Isolated trees as nuclei of regeneration in tropical pastures: testing the importance of niche-based and landscape factors

Questions

Our study tests the relative importance of different attributes of isolated trees in explaining the properties of the regeneration assemblage beneath their crowns to evaluate their effect on the composition of early successional stages and understand the processes underlying this effect. We asked: (1) does the regeneration assemblage contain a high proportion of individuals conspecific with the isolated tree; (2) how does the landscape surrounding an isolated tree influence the properties of the regeneration assemblage; and (3) what is the respective importance of structural variables and functional traits of isolated trees for their influence on the properties of the regeneration assemblage?
Location

Three pasture sites in Guanacaste Province, a tropical seasonally dry area, Costa Rica.
Methods

Ninety isolated trees were characterized by their position in the landscape relative to patches of trees, individual structural variables related to tree and crown size, and species functional traits. The assemblage of woody plants regenerating under their crowns was characterized by number of individuals and species, and taxonomic and functional composition. To understand the mechanisms underlying the facilitative effect of isolated trees on regeneration, correlations between their attributes and properties of the regeneration assemblage were assessed.
Results

We found that a low proportion of regenerating individuals were conspecific with the isolated trees beneath which they have established. Landscape factors influenced the regeneration but their effects were site-dependent. The attributes of isolated trees correlated well with the functional composition of the regeneration assemblage: their structure and phenological habits with traits associated with drought tolerance, and their dispersal syndrome with the dispersal syndrome of the regeneration assemblage. However, isolated tree attributes correlated poorly with the regeneration assemblage's taxonomic composition.
Conclusions

Our study shows that an approach based on functional composition can provide a better understanding of community assembly mechanisms than does a taxonomic approach. Two main mechanisms are suggested to explain the influence of isolated trees on the regeneration assemblage: modification of micro-scale environmental conditions (shade and moisture) and influence on dispersal agents. The effect of attributes of isolated trees on the properties of the regeneration assemblage predicts a long-term legacy of the early stages of succession.