Extinction and environmental change: testing the predictability of species loss
Current extinction rates are thought to be significantly higher than at any point in the last 65 million years. Such a loss of species could have serious implications for human well being, as humanity depends upon services that are derived from diverse, functioning ecosystems. Given the predicted increase in human alterations to the global biosphere in the next century, minimis- ing the loss of species, both now and in the immediate future, has become a prime concern, and has led to policy driven initiatives designed to halt or slow biodiversity loss. Tackling this loss of diversity requires an understanding of current, and recent, extinction events, as well as how future environmental change may alter the probability of a species persisting.
In this thesis I explore, using small-scale aquatic microcosms, modelling, and data from real-world extirpations, the predictability of extinction events, both historic and future, and how environmental change may alter the persis- tence of populations alone and in a community context.
I present results that suggest our ability to accurately infer the current rate of species loss will depend upon the method used to infer extinction, the amount of a habitat historically searched when looking for a species, and also the underlying population dynamics of that species, which can be altered by environmental change.
I demonstrate that the timing of extinction events, driven by various rates of environmental change, can be predicted using a simple phenomenological model, if a detailed knowledge of how the environment will change over time is known.
Lastly, I show that environmental change can interact with community assembly processes to alter the probability of a species persisting, and thus community composition.
The results of this work contribute to our understanding of current, and future, extinction events, and provide a basis for using quantitative approaches to inform conservation decision-making.
