The marine environment is intrinsically linked to the biotic and abiotic processes that regulate the life support systems of the planet, including nutrient and hydrological cycling, climate patterns, geological processes, oxygen production, and nutritional resourcing. Long-term natural cycles in climate variation have pronounced impacts on weather systems, sea surface temperature and marine food webs. Spatial and temporal heterogeneity in these systems and processes can influence communities directly via modulation of survival, reproductive success, and the distribution of resources, and anthropogenic pressures can contribute to heterogeneity in natural systems, influencing bottom-up and top-down processes. For ecological communities regulated by top-down processes, climate-linked shifts in the distribution, population and community dynamics of predators are likely to have pronounced effects on ecosystem composition and function. The influence of environmental variability upon predator ecology is therefore an area of particular research focus. In marine habitats, spatial and temporal heterogeneity in sea surface temperature has been associated with changes to reproductive phenology in predator and prey species, while spatial and temporal heterogeneity in resource availability may be associated with changes in the survival rates of animals across developmental stages. Likewise, heterogeneity in the approach to data collection, management and analysis may influence the interpretation of results and guide subsequent management decisions. To investigate the role of heterogeneity in marine predator ecology I focus on two apex predators in the Irish Sea: the grey seal (Halichoerus grypus) and the Manx shearwater (Puffinus puffinus). I explore how temporal heterogeneity of environmental conditions may affect reproductive phenology, how spatial and temporal heterogeneity of data collection and analysis methods affects estimates of population dynamics, and how heterogeneity in survival at different developmental stages can be reflected in population-level dynamics. After providing an overview of the focus of my thesis in Chapter one, using multi-decade time series from eight major grey seal and Manx shearwater breeding sites, I use logistic population growth models and generalised additive models in Chapter two to explore how changes in the timing and progression of the grey seal pupping season are dependent on climatic drivers. In Chapters three and four I use matrix population models (MPM) to quantify the effects of data aggregation and substitution of missing model parameters upon estimates of population dynamics over multiple spatial and temporal scales. In Chapter five I then continue the application of MPMs to calculate indices of resilience in scenarios of perturbation, to explore the population-level effects of reduced survival in specific demographic groups, namely fledgling, juvenile and adult Manx shearwater. My results suggest that contemporaneous heterogeneity in broad- and local-scale climate indices is less influential to reproductive phenology than intrinsic drivers, and that variation in survival rates of year-one animals is largely explained by fine-scale spatial heterogeneity. The substitution of demographic information when parameterising population models introduced biases and uncertainty into projections of population dynamics, and the simulated reduction of survival in juvenile animals appeared to have a potential latent effect on population stability – the consequences of reduced juvenile survival being realised as a reduction in recruitment to the breeding adult population. Finally, in Chapter six I summarise the main findings of this mosaic of studies and discuss them in the context of existing research, to identify avenues for future research. These investigations highlight the need for intrinsic and spatial processes to be incorporated into studies of climatic drivers of ecological change, and the importance of ensuring the accuracy and appropriate collection, management and analysis of data sources. They also illustrate the potential population-level effects of perturbations to survival in demographic groups which can be logistically difficult to monitor, and act as a reminder that the challenging option is often the one that is necessary.