The functional trait values that constitute a whole-plant phenotype interact with the environment to determine demographic rates. Current approaches often fail to explicitly consider trait x trait and trait x environment interactions, which may lead to missed information that is valuable for understanding and predicting the drivers of demographic rates and functional diversity. Here, we consider these interactions by modeling growth performance landscapes that span multidimensional trait spaces along environmental gradients. We utilize individual-level leaf, stem, and root trait data combined with growth data from tree seedlings along soil nutrient and light gradients in a hyper-diverse tropical rainforest. We find that multiple trait combinations in phenotypic space (i.e., alternative designs) lead to multiple growth performance peaks that shift along light and soil axes such that no single or set of interacting traits consistently results in peak growth performance. Evidence from these growth performance peaks also generally indicates frequent independence of above- and belowground resource acquisition strategies. These results help explain how functional diversity is maintained in ecological communities and question the practice of utilizing a single trait or environmental variable, in isolation, to predict the growth performance of individual trees.