| Summary: | Molecular tools have long been recognised as having enormous potential to expand horizons in ecology, but the promise remains substantially unfulfilled. In this thesis, I apply genetic approaches to two ecological problems that have proved difficult to solve using traditional techniques. Chapters 1 and 2 apply molecular tools to a community ecology problem to ask what mechanisms govern the persistence of an ant-plant metacommunity. I first use molecular data to clarify the number of coexisting ant species, and then employ population genetic techniques to investigate dispersal scale and other elements of life-history in the three most common species. Where hostplant density is high, a clear dispersal hierarchy is detected, which correlates positively with ant body size and negatively with fecundity, consistent with the hypothesis of a dispersal-fecundity trade-off. The hierarchy is less clear when hostplant density is low because one species shows dispersal plasticity, dispersing longer distances when hostplants are scarce. Results are discussed in the context of mechanisms that allow the coexistence of multiple symbionts with a single plant host. Chapters 3 to 8 address the use of molecular tools for informing decision-making in environmental management and biodiversity conservation. COI metabarcoding data are used to analyse patterns of arthropod diversity in the contexts of sustainable forest management (Chapter 5), agricultural management (Chapter 6), and habitat restoration (Chapter 7). It is shown that this potentially revolutionary technique can detect even fine-scale environmental changes, accurately characterise the biodiversity response to management variables, and be used to test the usefulness of convenient indicator variables. COI data is shown to outperform 18S data in recovering alpha and beta diversity information, and reference-based OTU-picking is demonstrated to be a useful approach where there is interest in the responses of a particular set of species. Potential applications and current limitations are discussed in Chapter 8.
|
|---|
