Plants defend themselves against herbivores by adopting a myriad of physical and chemical defenses and/or recruiting mutualistic defenders. Ecologists have proposed multiple different theories to explain the evolution of plant defenses, especially variation in secondary metabolites. However, the same is not true for physical defenses, which may be a more effective defense strategy against generalist herbivores. We propose a new theoretical model that extends the existing plant-defense framework to include all types of defenses. The model is based on consumer-resource interactions between one or more resources, a plant population, and a herbivore population. Any change in resource availability affects herbivores indirectly through changes in plant density and/or plant nutrient content. We assume that plants suffer a reduction in growth by investing in defense but also benefit from a reduction in herbivory. Unlike earlier models, we define fitness over a discrete time interval as the product of plant growth rate (as a function of resource availability) and the probability of survivorship (as influenced by herbivory). We then contrast the predictions of the proposed Resource Defense models with either one or two resources with that of the popular Resource Availability Hypothesis. The hypotheses suggested by these different theoretical models will be tested by sampling plants for defense traits across a resource and grazing gradient in the Serengeti National Park, Tanzania.
Results/Conclusions
The predictions are a result of the theoretical development of the Resource Defense models. Graphical analysis of the model with a single resource predicts that defense traits such as thorn density or phenolic concentration should increase with resource availability. This is in contradiction with the Resource Availability Hypothesis which predicts lower levels of defense in resource-rich environments. Furthermore, the inclusion of two resources (carbon and nitrogen) in the model clarifies the effect of different resources on defense traits. The model predicts that both carbon and nitrogen-based defenses should increase with nitrogen availability but decrease with carbon (water or light) availability. Data from the field shows that defense traits like thorn density are lowest at intermediate levels of rainfall and increase at both low and high rainfall, possibly due to changes in plant quality. This emphasizes the importance of considering relative availabilities of different resources in the environment on defense traits. Our models provide an alternative framework to understand variation in plant defenses in addition to laying a foundation for more complex mechanistic models.