The relationship between plant functional traits and demographic performance forms the foundation of traitbased ecology. It also serves as the natural linkage between trait-based ecology and much of evolutionary biology. Despite these important aspects, plant trait-demographic performance relationships reported in the literature are typically weak or nonexistent, and a synthetic picture of how traits are related to ecological and evolutionary patterns remains underdeveloped. Here, we begin by presenting an overview of the shortcomings in functional trait-demographic performance research and why weak results are more common than trait-based ecologists like to admit. We then discuss why there should be a natural synthesis between trait-based ecology and evolutionary ecology and potential reasons for why this synthesis has yet to emerge. Finally, we present a series of conceptual and empirical foci that should be incorporated into future trait-demographic performance research that will hopefully solidify the foundation of trait-based ecology and catalyze a synthesis with evolutionary ecology. These include (1) focusing on individuals as the fundamental unit of study instead of relying on population or species mean values for traits and demographic rates; (2) placing more emphasis on phenotypic integration, alternative designs, and performance landscapes; (3) coming to terms with the importance of regional-and local-scale context on plant performance; (4) an appreciation of the varied drivers of life-stage transitions and what aspects of function should be linked to those transitions; and (5) determining how the drivers of plant mortality act independently and in concert and what aspects of plant function best predict these outcomes. Our goal is to help highlight the shortcomings of trait-demographic performance research as it stands and areas where this research could course correct, ultimately, with the hope of promoting a trait-based research program that speaks to both ecologists and evolutionary biologists.

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 9 trait and trait 9 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.

Investigations of forest community structure and dynamics have been facilitated by the use of neighbourhood models that examine the interactions between a focal tree and its neighbours using a fixed radius. However, different studies have chosen different radii without clear reasons, hampering the understanding of mechanisms structuring tree communities. Using functional trait and tree demography data from the Luquillo subtropical forest in Puerto Rico, we compared fixed‐neighbourhood models with a canopy overlap model, in which tree crown overlap is used as an indicator of neighbourhood crowding. Analyses that combine functional trait and demographic data provide a mechanistic understanding of observed patterns of community structure and dynamics as they provide insights into the linkages between phenotypes and the environment. Overall, canopy overlap models had better support when compared to neighbourhood models using a fixed radius, suggesting that the fixed radius approach does not capture the full extent of competitive interactions among trees. Moreover, the effects of functional neighbourhood on tree survival and growth differ depending on the type of approach used and lead to different conclusions with respect to the drivers of tree community dynamics. Synthesis. In summary, our findings highlight the utility of neighbourhood models based on tree crown overlap, and suggest that applying this same approach to different plots and forests will facilitate comparisons across systems and improve our understanding of the mechanisms that drive the structure and dynamics of tree communities.

1. The structure and dynamics of ecological communities are ultimately the outcome of the differential demographic rates of individuals. Individual growth and mortality rates largely result from the interaction between an organism's phenotype and the abiotic and biotic environment. Functional traits have been used extensively over the past decade to elucidate links among phenotypes, demography and community dynamics. 2. A fundamental weakness of most functional trait approaches is the use of 'soft' traits associated with resource acquisition to examine how neighbourhood similarity affects tree survival and growth. However, these 'soft' traits are unlikely to be good predictors of similarities among co-occurring species. Less easily measured aspects of organismal functionsuch as those related to defencehave frequently gone unmeasured. This is particularly problematic for testing important hypotheses in forest ecology, such as the Janzen-Connell hypothesis where focal trees are expected to be at a disadvantage if their neighbours share the same natural enemies. 3. A potential alternative to functional trait approaches is to quantify the transcriptomic or functional genomic similarity of species. Such analyses are now possible in natural systems where de novo transciptome assemblies can be used to conduct functional phylogenomic analyses where homologous gene trees are produced. Using demographic plot data for 21 species from a North American forest dynamic plot, we conduct a community functional phylogenomic analysis of a plant community to elucidate the similarity in defence response genes across species. This similarity was then used to ask whether the similarity in defence genes of heterospecific species in the local neighbourhood of a focal individual tree influences its growth and mortality rates. 4. The results show that individual growth rates are higher when surrounded by dissimilar heterospecific species for 16 of 27 defence genes analysed. Additionally, survival rates are increased when an individual is in a neighbourhood with dissimilar species for 4 of the 27 defence genes studied. Lastly, strong conspecific effects were found in all analyses, underscoring that future analyses investigating the genetic variation and differential expression of defence-related genes in neighbourhoods may prove important. 5. Synthesis. In summary, this research leverages recent advances in RNA sequencing and bioinformatics to conduct community-wide transcriptomic analyses and analyses of defence-related gene similarity across a tree community. The results demonstrate that defence gene similarity in neighbourhoods often does have negative effects on individual demographic performance as predicted by the Janzen-Connell hypothesis.

Disentangling the processes that drive population, community and whole forest structure and dynamics is a challenge. It becomes a grand challenge in the tropics where there are a large number of species, small population sizes, less research infrastructure, and a relatively smaller number of researchers compared to the temperate zone. Tackling this grand challenge, we argue, requires detailed knowledge of the functioning of individuals and species. To this end, researchers frequently employ plant functional traits to study tree populations and communities. Here, we review this approach by first providing a basic background regarding the major trait axes generally of interest. We then discuss how these axes may be or have been applied from ecosystem to community and population studies. In doing so, we highlight where the functional trait research program has failed in tropical tree ecology and where it can be improved or strengthened. Finally, we provide a perspective regarding how functional trait and emerging 'omics approaches can be integrated to address large questions facing the field. Our intention throughout is to provide an entryway into this literature for an early career researcher rather than a comprehensive review of all possible studies that have taken place in tropical forests.