Intraspecific Variation in Plant-Animal Interactions of the Brassicaceae Family Along a Steep Rainfall Gradient in the Eastern Mediterranean Basin

Abstract

The future climate of dryland ecosystems is predicted to become increasingly arid as rainfall decreases and becomes more unpredictable and extreme. Consequently, dryland plant communities may risk extinction if species are unable to adapt to increasing aridity. A plethora of theoretical and empirical studies has investigated plant adaptations to arid environments, yet, adaptations in plant-animal interactions are often overlooked. Because the abiotic environment can determine the composition of pollinators and herbivores, which, in turn, can drive the coevolution and trade-offs in plant reproductive strategies and plant herbivore defences, it is critical that we investigate the dependencies of plants on the activities of their associated animals under different climatic conditions. Observing intraspecific variation in plant-animal interactions along environmental gradients provides compelling evidence of how the ecosystem may respond to the future climate. Whilst this approach has been exploited for latitudinal and elevation gradients, fewer studies have addressed intraspecific variation in plant-animal interactions along rainfall gradients. In addition, these studies often use a single plant species which reduces our ability to generalise. This dissertation aims to provide much needed evidence for the effects of increasing aridity on plant performance, mediated through their interactions with pollinators and herbivores, as predicted by climate change. A steep rainfall gradient in the Eastern Mediterranean Basin was used to determine the effect of increasing aridity on pollinator services and herbivory pressure. Simultaneously, a pollinator-exclusion experiment in the field determined whether different plant species in the Brassicaceae family could secure their reproduction by self-fertilisation in the absence of pollinators. Furthermore, evidence of increasing self-compatibility as an adaptation to increasing aridity was ascertained by a pollination manipulation experiment. Finally, quantifying the concentration of glucosinolates in plant species highlighted the relationship between herbivory damage and plant chemical defence investment. The abundance and diversity of pollinators decreased with increasing aridity, whilst self-fertilisation was variable across the plant species. This weak relationship suggests pollinator limitation plays a minor role in selecting for self-fertilisation in plants. Genetically, the plant species were more self-compatible in increasingly arid environments, suggesting that the mechanisms of prezygotic self-incompatibility weaken with increasing aridity. In congruence with decreasing pollinator activity, herbivore pressure also decreased with increasing aridity, yet, total leaf glucosinolate concentration in plants did not change. The lack of evidence for a trade-off in leaf damage by herbivory and total leaf glucosinolates concentration suggests herbivore pressure may not be the dominant factor selecting for plant chemical defence.

Taken together, the research of this thesis shows that plant performance is surprisingly robust to variations in the pollinator community and in herbivore pressure. I conclude that, whilst the diversity of the pollinator community and the resilience of herbivores may be under threat in future climates, the reproductive success and resource-allocation into constitutive herbivore defences in plants, at least in the Brassicaceae family, shall remain relatively unaffected by climate change.