Plants live in a heterogeneous world, where nutrient and neighbours distributions vary in space and time. Plants can respond to this variation through plastic responses in individual organs, which are assumed to be coordinated among traits to support a coherent, adaptive strategy, maintaining plant growth in varying environments. However, this assumption remains understudied. Here, we test coordination in trait plasticity and its influence on plant performance. We do so among 27 naturally co-occurring species under four different neighbour and soil fertility contexts using data from two mesocosm experiments manipulating nutrients and neighbour presence. We focus on commonly-studied resource-acquisitive traits: specific leaf area (SLA), root tissue density (RTD), and specific root length (SRL). We found species exhibited plasticity in SLA, RTD, and SRL 50-75% of the time and that responses to neighbours were stronger than nutrients for SLA and RTD, but not SRL. However, plastic responses were highly modular, correlated <10% of the time, and had little coordination consistent with the trait economics paradigm. Consequences of plasticity for plant performance depended on the trait, the response direction, and whether plants were responding to nutrients or neighbours. Plasticity in RTD was unrelated to performance, but species that increased their SRL under low nutrient conditions performed worse than those showing no SRL plasticity, while SRL responses to neighbours had no effect. Similarly, species that increased SLA under low nutrients or with neighbours showed reduced performance compared to those that decreased SLA in those same conditions. Overall, plasticity in SLA, RTD, and SRL rarely manifested as a correlated or coordinated response promoting growth maintenance in our study. Instead, plasticity’s consequences hinged on the direction of trait shifts, sometimes incurring costs. Future studies should consider plasticity a directional phenomenon.

A standard belief is that species can be organized in competitive hierarchies, leading to concepts of plant strategies. Based on species' average competition values, this approach may fail to predict specific outcomes and overlook common facilitative interactions, biasing understandings of species interactions. Using a mesocosm experiment, we examined interactions of 13 native grassland species grown alone and in 168 pairwise combinations under two nutrient conditions. We measured the positive and negative effects of plants on neighbours and their responses to neighbours. All species engaged in both facilitative and competitive interactions, and neighbour effects and responses were positively related, suggesting that species specialize at either end of the interaction gradient. Nutrient addition reduced the strength of both competitive and facilitative plant-plant interactions, thereby reducing variation in plant growth. These results suggest the existence of a competition-facilitation continuum, integrating both effects on and from neighbours, along which plant strategies may fall.