Interspecific facilitation and critical transitions in arid ecosystems

Abstract

The human impact on the climate system and on ecosystems has rapidly increased over the last century. Moreover, forecasts by global climate models predict that increasing temperatures and more frequent and longer drought periods will impose an increasing pressure on ecosystems over the coming decades, with arid regions such as the Sahel, the Mediterranean and other (semi-)arid climate zones, among the most impacted regions. How dryland ecosystems will respond to increased drought stress is therefore a central theme in ecology and environmental sciences. Theoretical ecosystem models showed that increasing drought stress or grazing pressure in dryland ecosystems may lead to a critical transition from a vegetated to a non-vegetated state. Such transitions are difficult to reverse and it is therefore of utmost importance to increase our understanding of under what conditions critical transition may occur. Previous ecosystem model studies showed that critical transitions in drylands can occur if an ecosystem is controlled by positive feedbacks between the vegetation and the physical environment, or if plants provide facilitative effects, for example when mature ‘nurse’ plants protect neighbouring ‘protégé’ seedlings from grazing damage. Although facilitation between plants is now recognised as an important process underlying ecosystem stability, diversity and functioning, it is still uncertain how facilitation changes over environmental gradients. In particular, although the combination of drought stress and grazing stress is common in arid ecosystems, it is unclear how facilitative interactions between plants may change along combined drought and grazing gradients. The two main research questions of this thesis are: how do plant-plant interactions change along combined stress gradients consisting of drought and grazing stress? And how do these changes relate to dryland ecosystem stability? To shed light on these questions, in this thesis I combined insights from a conceptual review study, an observational study, an experimental study and an ecohydrological modelling study. Overall, my findings of this thesis underlined the importance of grazing in explaining observed positive interactions or spatial associations between plants in a semi-arid ecosystem. Also I showed that facilitative interactions may shift back to neutral when stressors are combined at the high end of a stress gradient.This thesis also shows that positive interactions in drylands, and an increase in facilitation with increasing aridity should not be expected a priori. Our model results showed that intermittency in rainfall is highly important in driving plant coexistence ranges and in determining the intensity and direction of plant-plant interactions. A future increase in rainfall intermittency in drylands may result in increased positive interactions between plants, thereby possibly partly counteracting the negative effects of increased prolonged dry spells in arid ecosystems. Future model studies should incorporate the more complex nature of interspecific plant-plant interactions to improve our ability to predict how (semi-) arid ecosystems will respond to projected increases in drought conditions. Combining long-term experimental outcomes, to mechanistic model studies will be essential in order to progress our ability to predict which areas of (semi-) arid ecosystems are most vulnerable to future degradation.