Vegetation clump size and number as indicators for alternative stable states in semi-arid ecosystems

Dryland ecosystems are vulnerable to desertification, a pressing issue in the face of global climate change. In these ecosystems, vegetation often grows in spatially periodic patterns that differ as aridity increases (gaps, labyrinths and clumps), which has been widely studied theoretically, aiming to assess the proximity of the system to desertification. While some theoretical models predict smooth desertification transitions, most typically predict an abrupt transition linked to the possibility of two alternative stable states (desert and vegetated states), predictions that are yet to be confirmed empirically. If this bistability of alternative stable states occurs, however, environmental fluctuations and the history of the ecosystem determine which state ultimately materializes. This uncertainty makes it harder to predict desertification, compounding the challenges posed by it. Here, we combine empirical data and theoretical methods to investigate the links between bistability and vegetation spatial organization, which can help identify the presence of alternative stable states. We found that, although the presence of vegetation clumps is not indicative of bistability, changes in clump morphology can provide reliable indicators of bistability and an impending desertification transition. Thus, our methodology indirectly identifies whether desertification will occur abruptly, and whether restoration efforts should consider a potential ecosystem history-dependence.