Biogeographical studies frequently reveal positive correlations between species richness and estimates of environmental water and/or energy. A popular interpretation of this relationship relates the supply of water and energy to productivity, and then, in turn, to richness. Productivity–diversity theories are now legion, yet none has proved sufficiently intuitive to gain broad acceptance. Like productivity, heterogeneity is known to influence diversity at fine spatial scales, yet the possibility that richness might relate to water–energy dynamics at coarse spatial scales via a heterogeneity-generating mechanism has received little attention. In this paper we outline such a conceptual model for plants that is internally consistent and testable. We believe it may help to explain the capacity of environments receiving different inputs of water and energy to support variable numbers of species at a range of spatial scales, the pervasive correlation between productivity and richness, some exceptions to the productivity–diversity relationship, the form of productivity–diversity curves and the link between richness and environmental ‘harshness’. The model may also provide an answer to one of the most venerable puzzles in the field of diversity studies: why high inputs of water and energy correspond to more species rather than simply more individuals.