INNOVATIONS IN THE FACE OF CLIMATE CHANGE With and without warning: managing ecosystems in a changing world

www.frontiersinecology.org © The Ecological Society of America I 2011, the worst algal bloom in the history of North America’s Lake Erie developed in the western basin of the lake (Stumpf et al. 2012), the result of a combination of agricultural fertilizer runoff, heavy spring rains, and stable summer conditions that favored heavy algal growth (Michalak et al. 2013). Analysis of the dynamics and projections of climate change, including a prediction of increased storm intensity, led Michalak et al. (2013) to call the 2011 Lake Erie bloom “a harbinger of future blooms”. They were right. In the summer of 2014, another massive bloom developed in western Lake Erie, and drinking water drawn from the lake was found to contain unsafe levels of a cyanobacterial toxin. Consequently, the water supply for the city of Toledo, Ohio (population 284 000), was shut down and citizens were soon waiting in long lines for bottled water. In this case, ecologists provided advance warning; in the future, it will be possible to provide even more detailed predictions of the timing, intensity, and even toxicity of algal blooms in Lake Erie because the causes and conditions leading to such blooms are better understood (Obenour et al. 2014). Climate warming and other human-driven forces mean that, in contrast to the Lake Erie algal blooms, some abrupt ecosystem changes – as well as losses of ecosystem services – may arise without apparent warning. Even in hindsight, the causes of such rapid changes will be hard to discern because of multiple interacting forces. Thus, in the future, abrupt changes are likely to occur both with and without warning. This raises two questions. First, can research improve forecasts and the detection of warning signs? Second, can research help foster ecosystem resilience to limit the risk of crossing irreversible thresholds? Maintaining ecosystem services in the future will require a substantial amount of research on both these questions. Improved forecasts and warnings can help in the management of ecosystems and help to sustain ecosystem services by avoiding unwanted changes and by warning of undesirable conditions. Promoting resilience, especially in cases where there is no forewarning of change, can help avoid thresholds or mitigate abrupt change when thresholds are crossed. This paper addresses approaches to anticipating and managing adverse ecosystem changes, specifically those resulting from threats such as climate warming, intensification of agriculture, fisheries exploitation, and the introduction of invasive species. Extreme climate events associated with these drivers are of special interest because they may push ecosystems into new states and impede recovery to desirable states. We consider warnings provided by model forecasts and by statistical anomalies indicating loss of resilience as thresholds are approached. We also discuss changes that may occur without warning, INNOVATIONS IN THE FACE OF CLIMATE CHANGE