Climate change is expected to directly alter the composition of communities and the functioning of ecosystems across the globe. Improving our understanding of links between biodiversity and ecosystem functioning across large spatial scales and rapid global change is a major priority to help identify management responses that will retain diverse, functioning systems. Here we address this challenge by linking projected changes in plant community composition and functional attributes (height, leaf area, seed mass) under climate change across Tasmania, Australia. Using correlative community-level modeling, we found that projected changes in plant community composition were not consistently related to projected changes in community mean trait values. In contrast, we identified specific mechanisms through which alternative combinations of projected functional and compositional change across Tasmania could be realized, including loss/replacement of functionally similar species (lowland grasslands/grassy woodlands) and loss of a small number of functionally unique species (lowland forests). Importantly, we demonstrate how these linked projections of change in community composition and functional attributes can be utilized to inform specific management actions that may assist in maintaining diverse, functioning ecosystems under climate change.