Leaf allocation improves predictability of interspecific growth rates in a broadleaf deciduous temperate forest

Understanding the relationships between species' demography and functional traits is crucial for gaining a mechanistic understanding of community dynamics. While leaf morphology represents a key functional dimension for plants worldwide (i.e., the leaf economics spectrum), its ability to explain variation in trees' life history strategies remains limited. Plant growth is influenced by both leaf morphology and allocation; hence, incorporating both dimensions is essential but rarely done. Additionally, trait–performance relationships have mainly been studied in tropical communities, leaving gaps in our understanding of temperate forests where different seasonality patterns may alter these relationships. We examined whether species' leaf area index (leaf area per crown size, LAI), a measure of leaf allocation, explains the variation of juvenile tree species' potential growth rates in a winter-deciduous broadleaf forest. LAI has not been characterized as a species-level trait, but its ability to predict plant productivity at the ecosystem scale highlights its potential for explaining plant growth. We evaluated species' maximum LAI both individually and in conjunction with wood density (WD) and leaf mass per area (LMA). We expected that models would improve when both leaf morphology (LMA) and leaf allocation (LAI) were included and that species with denser crowns would have higher potential growth rates. LAI and LMA were significant predictors of growth but only when both were incorporated, and together explained a high proportion of species' growth variations (R²_adj = 0.59). We found evidence of a trade-off between LAI and LMA, with a negative relationship between them and each having a positive influence on species' growth, suggesting that there are multiple allocation strategies to achieve fast growth. A surprisingly positive LMA–growth relationship contrasts with observations from tropical forests. We did not find significant relationships with WD in this forest. Our results highlight that incorporating leaf allocation improves models of trait–performance relationships. They also suggest, in agreement with the limited literature, that temperate forests may exhibit different trait–performance relationships from those of tropical forests, where LMA is negatively related to growth and WD is often important. Clarifying the details and contexts of trait–performance relationships is crucial for applying the functional trait framework to understanding community structure and dynamics of forests globally.