Mechanisms of community assembly and ecosystem functioning are often analyzed using community functional composition, quantified as community-weighted mean trait values (CWMs). We have little information, however, about how constituent demographic rates (e.g., recruitment, growth, mortality) ultimately give rise to shifts in community functional composition. A quantitative framework for identifying the role of particular demographic processes in driving community-level trait shifts could facilitate more direct tests of theory and provide novel insight to community assembly. We present a novel conceptual framework to quantify the contribution of key demographic processes to temporal changes in CWM values for four key functional traits (seed size, SLA, leaf P content, and wood density). We used this framework to analyze the mechanisms of secondary succession in wet tropical forests in Mexico. We addressed the following two primary questions: (1) How do the community-weighted mean trait values of recruiting, growing, and dying trees change during succession? (2) What is the relative contribution of recruitment, growth, and mortality to community-weighted mean trait dynamics during succession?
Results/Conclusions
Community-weighted mean seed size increased over time, reflecting a trade-off between colonization by small seeds early in succession, to establishment by large seeds later in succession. SLA and leaf P decreased over time, reflecting a trade-off between fast growth early in succession versus high survival late in succession. On average, CWM trait shifts were driven mainly (66%) by growth of surviving trees that comprise the bulk of standing basal area, then mortality (29%), and weakly by recruitment (5%). Trait shifts of growing and recruiting trees mirrored the CWM trait shifts, and traits of dying trees did not change during succession, indicating that these traits drive recruitment and growth, but not mortality, during the first 30 years of succession. The ultimate goals of trait-based ecology include predicting ecosystem function and community response to environmental change. Advancing predictive ability requires stronger links to underlying demographic processes, and a better understanding of how trait-mediated performance differences scale up to community functional patterns. Our approach provides insight to the demographic processes that lead species with certain traits to become dominant, and subsequently influence ecosystem function. Ultimately, we hope our approach will help in using demography to bridge theory and trait-based ecology.