Despite the inclusion of more and more different ecosystem processes in models, large uncertainties persist among land ecosystem models when predicting the responses of terrestrial ecosystems to environmental change. There is both a crisis, and an opportunity: to develop ‘next-generation’ land ecosystem models upon a firm theoretical and empirical basis that is currently missing.

A highly promising new approach to develop next-generation land ecosystem models draws on the concept of evolutionary optimization, which can generate remarkably simple process formulations that can be observationally tested. The general principle is long-standing: it simply proposes that organisms adjust to environmental variations so as to maximize fitness, and are thereby subject to Darwinian selection. However, this leaves open many practical questions about what can be optimized, and in particular the relationship between vegetative processes and fitness. Thus, specific optimality hypotheses have to be tested empirically. A generalization is that ecosystem function is maintained by acclimation within individuals, adaptation within species, changing abundances within ecosystems, and migration within regions; these ideas potentially form the basis for a comprehensive model in which all of the components are individually testable against data.

Optimality concepts have been applied to aspects of plant and ecosystem function including stomatal behaviour, plant water use and photosynthetic capacity, nitrogen uptake and phenology. They hold great promise for incorporating the optimal allocation strategy of plants into a mechanistic modelling framework for net primary productivity and biomass growth. To seed communication and collaboration among pioneers of this field, including observational scientists and experimental ecologists, we invite contributions on all topics related to optimality principles in plant and ecosystem function including theoretical analyses, experimental tests, and models.