Vertebrate DNA is littered with the signatures of past retroviral infections.  Surprisingly, there is little evidence of active endogenous retroviruses.  Using modelling tools from ecology we ask the question why does so much of the vertebrate genome consist of signatures of past epidemics?  Our work reveals how ecological methods can be used to address questions that ecologists rarely think about.

We explore the effects of adding a disease to a stage structured population model with a bottleneck in maturation. The disease can specifically target one ontogenetic stage, or it can affect all stages. We vary disease parameters such as infectiousness and virulence, and we consider cases where the disease decreases fecundity or slows down maturation.

Parasites are a key driver of evolutionary processes in wild animals. However, quantifying endoparasite burdens non-destructively can be problematic. We successfully utilised endoscopy in a wild seabird host in the field as a method for measuring natural nematode burdens and verifying that drug treatment removed parasites. Endoscopy was found to be a rapid, reliable and repeatable method for assessing parasite burdens in this system.

Scientific knowledge of the tree of life is expanding rapidly, but the methods used for visualising it are restrictive. Here we present a new and visually appealing way to explore large phylogenetic trees with metadata, using fractals and a zooming interface.  The software should help scientists explore large datasets and also help communicate ecological and evolutionary concepts to the public.

Here we introduce a novel analytical method in which individual trees (including their arthropod communities) are selected from a source dataset to match communities from a target plot in terms of tree abundance, size distribution and community phylogeny. This allows prediction of the arthropod communities in the target area, with deviations revealing changes in the underlying rules dictating community structure.

Predicting the invasion speed of polymorphic populations has revealed that anomalous speeds may be important.  These occur when a polymorphic population invades faster than a single morph.   Using a discrete model we show that these speeds persist when there is demographic stochasticity if population densities or mutation rates are high.

A typical species distribution model is static. We describe species distribution models for categorical abundance data that have an explicit temporal component. We use our approach to study the distribution and dynamics of two intertidal invertebrates, the top shells Osilinus lineatus and Gibbula umbilicalis, based on data from approximately 100 sites surveyed over 8 years

A model that takes a phenotypic-gambit approach and focuses on changes in the frequency of phenotypes within a population, using as an example seasonal breeding was developed. Results showed that the phenotypic approach has utility when attempting to accommodate evolution within an ecological model. Simplified model variants were explored in order to examine how complex should an eco-evolutionary model be.

In ecology, systems composed of many individuals that combine to exhibit a collective behaviour are prevalent – e.g., insect swarms. Whilst models rooted at the individual level can account for empirical observations, they are often impractical and incapable of predicting macro-scale evolution. We present a mathematical theory that can explain how interactions and changes at the micro-scale affect the global dynamics.

Population cycles in voles and lemmings are often thought to require one-year delayed density-dependence (DD), with specialist predators causing crashes. Re-analysing data, and contrasting popular log-linear models with non-linear ones, we show that direct DD is less stabilising than previously thought, and delayed DD can have a different role than that assumed by current theory.

The ability to infer temperature dependence of biological rates is crucial for predicting the response of ecological systems to environmental change. We simulate temperature dependent population time series and use them to test parameter estimation methods. Comparing different methods we obtain accurate estimates of activation energy for simulated data. We use time series from microbial aquatic communities as case studies. 

Plant community dynamics can be interpreted as the eventual replacement of individual plants by other individual/s of the same or different species. This replacement is the result of interspecific juvenile-adult interactions which configure a “replacement network”. We describe a model that uses this framework to study the network structure and dynamics of woody plant communities.