<P> Species respond in very different ways to climate change . Variation in the distribution, phenology and abundance of species will lead to inevitable changes in the relative abundance of species and their interactions . These changes will flow on to affect the structure and function of ecosystems . </P> <P> Accurate predictions of the future impacts of climate change on plant diversity are critical to the development of conservation strategies . These predictions have come largely from bioinformatic strategies, involving modeling individual species, groups of species such as' functional types', communities, ecosystems or biomes . They can also involve modeling species observed environmental niches, or observed physiological processes . </P> <P> Although useful, modeling has many limitations . Firstly, there is uncertainty about the future levels of greenhouse gas emissions driving climate change and considerable uncertainty in modeling how this will affect other aspects of climate such as local rainfall or temperatures . For most species the importance of specific climatic variables in defining distribution (e.g. minimum rainfall or maximum temperature) is unknown . It is also difficult to know which aspects of a particular climatic variable are most biologically relevant, such as average vs. maximum or minimum temperatures . Ecological processes such as interactions between species and dispersal rates and distances are also inherently complex, further complicating predictions . </P> <P> Improvement of models is an active area of research, with new models attempting to take factors such as life - history traits of species or processes such as migration into account when predicting distribution changes; though possible trade - offs between regional accuracy and generality are recognised . </P>

Where can you observe environmental changes that are forcing adaptation