Functional seed traits as predictors of germination and seedling growth for species with potential for restoration in Caquetá, Colombia

Key message Pre-germination requirements are not accurately predicted by seed traits of size and mass. However, small-seeded species show high germination and slow early seedling growth rates. Abstract Deforestation in the tropics is one of the greatest biodiversity crises of the Anthropocene. In response, tropical countries such as Colombia have committed to restoring degraded lands and reconnecting forest fragments. To contribute to this mission, we evaluated how easy-to-measure seed functional traits predicted germination requirements and early seedling growth rates (RGRs) for 11 tree species native to the Andean-Amazon piedmont region. Pre-sowing treatments followed a 3 × 3 factorial design: three light conditions (full, partial, and no light) and three pre-germination treatments (hot water, abrasion, and control). A cluster analysis grouped species into three groups driven chiefly by highly correlated variables of seed size and mass. However, species within and among functional groups differed in their response to pre-germination treatments. One exception includes the small-seeded species group responding with little-to-no germination to the absence of light. Since functional traits were highly correlated, we used Principal Component Regression (PCR) to predict germination metrics (germination percentage, velocity, and vigor) and RGRs. The multivariate trait axis of seed size and seed mass measurements (PC1) correlated negatively with germination metrics and positively with RGRs. Our results highlight the challenges associated with using easy-to-measure functional traits to deduce the most effective pre-sowing treatment. Other more relevant functional traits can be measured in lab, but a shotgun approach to testing pre-germination treatments may be more cost-effective. High germination in small-sized seed species highlights the early-successional niche these species occupy, but reduced RGR may represent a tradeoff associated with having relatively few resources stored in seed.