Ecosystems are sensitive to human pressure and climate change. This is a main concern because ecosystems may shift abruptly and irreversibly, with concomitant losses of ecological and economic resources. Such critical transitions and associated thresholds in ecosystem have been theoretically attributed to positive feedback and alternative stable ecosystem states. There is a growing body of literature addressing the question whether there are early warning signals predicting such critical thresholds from time series data. Here we address the question whether early warning signals can be derived from spatial data.

Results/Conclusions Theoretical models predict that so-called regular vegetation patterns observed in large areas in ecosystems world-wide are a result of spatial self-organization, and the shapes of the patterns are associated with approaching critical thresholds. Also, patch-size distribution of the vegetation in various ecosystems follows a power law, and consistent deviations from power laws occur if stress is high. Model analysis suggests that such deviations from power laws may be a warning signal for the onset of critical transition. Common ecological mechanisms that account for these patterns are scale-dependent feedback and local facilitation. Our results are relevant to identify areas that are vulnerable to critical transitions in the face of increased human pressure and ongoing global climate change.