Distribution of functional traits in Belize forests of varying successional stages

Background/Question/Methods

Plant functional traits are commonly used as indicators of ecological strategy in studies of community assembly.  Particular traits are associated with pioneer strategies, with regard to the leaf economics spectrum and seed size.  As forests develop throughout succession, the range of traits should expand as more mid and late successional species arrive at the site.  Furthermore, it is possible that as forests become more established, trait spacing may become more regular as niche processes exclude weaker close competitors.  In this work, we ask:

1. Do community aggregate trait values shift with successional age?

2.  How does trait spacing and overlap change with successional age?

We conducted this study in two forests of different successional ages in Cayo District, Belize.  All stems >1.6 cm dbh were censused and leaf specimens were collected from the sunniest available location. We measured specific leaf area, leaf laminar area, and leaf nitrogen content. We calculated community trait means for the entire plot and for quadrats in each forest.  We used mean and standard deviation of next and nearest neighbor distance as measures of differentiation and evenness of spacing, and Mouillot overlap curves for trait overlap.  We ran 9999 randomizations for each to determine statistical significance.

Results/Conclusions

The two communities did differ significantly in community mean trait values, but for specific leaf area, the older site had a higher rather than lower mean community SLA, contrary to our expectations.  Trait ranges were higher in the older forest, but spacing was largely random relative to expectations in both forests.  Overlap was higher in the younger forest, but this may be because of the smaller trait range in that community, or because of less sorting.  Overall, this data suggests that there is some change in trait distributions between communities of differing successional stages, but the change is not always as expected from prior knowledge of traits.  The relative lack of signature of plot age in trait spacing suggests limitations to this approach to understanding community assembly with regards to functional traits.