Introduction

Different disciplines within ecology have used a wide range of approaches to predict the impacts of climate change on forest ecosystems. There has been considerable effort by ecosystem scientists to couple vegetation dynamics to atmospheric circulation models, but the spatial scales inherent in those models have required extremely primitive characterisation of population and community dynamics (Sitch et al. 2008; Smith et al. 2001). Biogeographers have used a variety of approaches to project changes in species distribution under climate change, but those models typically represent steady-state ‘potential’ responses based on a mapping of current species distributions onto predicted future climates (Jeschke & Strayer 2008; Loehle & LeBlanc 1996). Population and community ecologists have used both contemporary field data and palaeoecological methods to understand the consequences of variation in climate for the dynamics of tree species, and have attempted to incorporate those responses into formal models of forest community dynamics (e.g. Bugmann 1996). But it would be fair to say that there is little consensus on the predictive power of any of these approaches.

A challenge faced by all of these approaches has been that climate change is taking place in the context of a broad suite of other anthropogenic impacts on forest ecosystems, ranging from invasive species to air pollution. A number of these clearly have had significant impact on the current distribution and abundance of species, and the functioning of forest ecosystems. In the face of the rapid environmental change of the past century, and given the inherently long time scales of tree population dynamics and forest succession, it seems inescapable that forests are increasingly in disequilibrium, lagging in their responses to both perturbations to species abundances and to changes in the physical and biotic environment. As a corollary, it seems reasonable to expect that the responses of forests to climate change will be conditioned by and interact with transient dynamics already triggered by other aspects of global environmental change. There is also ample evidence that physical and biological legacies of human land-use can persist over time scales from decades to centuries (Foster et al. 2003; Katz et al. 2010), and that these impacts will also need to be considered in predictions of forest response to future climates.