To date, climate change has had a relatively modest effect on ecosystems and biodiversity, compared to direct anthropogenic actions such as overharvesting and land use change resulting in habitat loss. This relative importance is already changing, and the negative ecological impacts of climate change are becoming more apparent and very likely to intensify over the coming decades. On land, climate change is increasing precipitation variability and the probability of extreme dry and wet events, and long-term warming and increasing atmospheric water deficits are increasing physiological and hydrological stress and ecosystem flammability. In the ocean, an increased occurrence of heatwaves and long-term trends of acidification increase physiological stress on many organisms and ecosystems. Interaction of other anthropogenic stressors such as defaunation, overfishing, invasive species, fragmentation and direct habitat degradation tend to amplify the sensitivity of ecosystems to climate change. It is extremely challenging to predict the patterns and probabilities of biodiversity loss, both from the subtle effects on individual species within complex multi-trophic ecosystems and the more abrupt effects of ecosystem degradation.
In the context of the complexity of ecosystems and a vast shortfall in the understanding of how specific species, and interspecific interactions, will respond to climate change, there is a need to adopt a strategy of adaptive ecosystem research, in addition to adaptive ecosystem management. There are many aspects of ecosystem science where we will not know enough in sufficient time. Ecosystems are changing so rapidly in response to global change drivers that our research and modelling frameworks are overtaken by empirical, system-altering changes. New frameworks for modelling and monitoring highly dynamic complex systems need to be applied. We need improved ways to implement adaptive ecosystem management under uncertainty.