Trait variation observed across the Kingdom Plantae is the outcome of dynamic evolutionary histories and present-day ecological and evolutionary processes that act to generate and maintain trait diversity across biological scales. Traits that consistently relate to performance across taxa and environments – often referred to as ‘functional traits’ – provide a functional link between plants and their environments giving context for how these traits reflect adaptation to aspects of the environment. Within a given community, we often observe substantial variation in functional traits among co-occurring species and this variation is assumed to reflect some aspect of niche partitioning that promotes species co-existence. If these functional traits are related to performance, and there is large among species variation in these traits, we might then also expect large differences in performance among species that could lead to competitive asymmetries and potentially competitive exclusion. Marks and Lechowicz (2006) presented a possible explanation for this paradox. Using a simulation approach, they demonstrated that selection on phenotypes in a single environment can lead to ‘alternative designs,’ such that different combinations of traits (with high variation in traits underlying performance) can achieve relatively similar performance (i.e., little variation in performance itself). Here we expand upon this observation by demonstrating that the presence of ‘alternative designs’ can occur under a unique scenario: specifically, plants that differ in traits affecting performance can have similar performance, if the major axes of trait variation are not aligned with the functional axis representing the link between the functional traits and performance. While our model is perhaps one of several explanations to help explain the presence of ‘alternative designs,’ it has important implications for community ecology and for phenotypic integration. First, our solution is consistent with the expectation that co-existing species differ in traits that relate to differential niche usage, while also consistent with the expectation that in order to co-exist, species differ minimally in performance thus preventing competitive exclusion. This model also suggests that ecological selection operating to maintain ‘alternative designs’ among species within communities might in part contribute to the stability of phenotypic integration of populations (i.e., an emergent property that reflects the presence and strength of multivariate trait correlations). The degree to which phenotypes are integrated can both facilitate and hinder plant ecological and evolutionary responses to environmental changes.