Marine top predators often occupy broad geographical ranges that encompass varied habitats. Therefore, a pre-requisite towards conserving these animals is to determine the components of their realized niche, and investigate whether a species is a specialist or a generalist. For generalist species, it is also necessary to understand if local specialisation occurs. Uncovering these components can allow us to build models of a species realized niche that may then be used to infer habitat use in unsampled locations. However, fully understanding the components of a marine top predators realized niche is challenging owing to the limited opportunity for in situ observations. Overcoming these limitations is a key step in marine top predator research. It will enhance our understanding of trophic coupling in marine systems, and aid in the development of tools to better study these predators in their dynamic environment.
Seabirds, penguins (Spheniscids) in particular, are a group of animals for which investigating their realized niche is of vital importance. This is because numerous species face growing uncertainty in the Anthropocene, and in a time of rapid environmental change there is furthermore a need to better understand the potential use of these birds as indicators of ecosystem health. The aim of this thesis, therefore, is to investigate the foraging ecology of gentoo penguins (Pygoscelis papua) at the Falkland Islands. At the Falkland Islands, limited historical information exists regarding this species foraging ecology, with most information coming from a single location at the Falklands. As the Falkland Islands have the world’s largest population of gentoo penguins, elucidating factors influencing this population will have global relevance. Furthermore, historical information indicated potential competition with fisheries, and with prospecting for hydrocarbons and an inshore fishery, there is a need to understand the distribution of these birds across the islands. Penguins are also well suited to carry biologging devices allowing for in situ observations of inter and intraspecific interactions, as well as habitat specific interactions.
In this study, I sampled birds over three breeding seasons, from four breeding colonies - chosen for their varied surrounding at sea habitat - across the Falkland Islands. I investigated the diet with stomach content and stable isotope analysis, the at-sea distribution with GPS and time depth recorders, and how these birds behaved at sea using custom made animal-borne camera loggers. Furthermore, I developed a method to recognise prey encounter events from back mounted accelerometers, using a supervised machine learning approach.
As part of the first species specific description of diet at this scale for the Falklands, I revealed six key prey items for the birds: rock cod (Patagonotothen spp.), lobster krill (Munida spp.), Falkland herring (Sprattus fuegensis), Patagonian squid (Doryteuthis gahi), juvenile fish (likely all nototheniids), and southern blue whiting (Micromesistius australis). The use of animal-borne camera loggers verified that not only do gentoo penguins consume a diverse array of prey items, but they adopted various methods to capture and pursue prey, with evidence of birds following optimal foraging theory.
Prey composition varied significantly between study sites with the at-sea distribution and habitat use of penguins reflecting that of local prey. Birds from colonies close to gently sloping, shallow waters, foraged primarily in a benthic manner and had larger niche widths. However, those at a colony surrounded by steeply sloping, deeper waters, typically foraged in a pelagic manner. Contrasting diet patterns were also prevalent from stable isotope data, and the niche widths of birds relating to both stomach content and stable isotope data were larger at colonies where benthic foraging was prevalent. Therefore, it was clear that surrounding bathymetry played a key role in shaping this species’ foraging ecology, and that at the population level at the Falkland Islands birds are generalists. However, at individual colonies some specialisation occurs to take advantage of locally available prey.
I developed habitat distribution models - via boosted regression trees – which transferred well in time but poorly across space. Reasons for poor model transfer might relate to the generalist foraging nature of these birds and the reduced availability of environmental predictors owing to the limited range of these birds. I furthermore developed a method to identify prey encounter events that can also, to a degree, distinguish between prey items. This method will be a promising approach to refine habitat distribution models in future. These habitat distribution models could potentially contribute to marine spatial planning at the Falkland Islands.
Footage from animal-borne camera loggers clearly showed that prey behaviour can significantly influence trophic coupling in marine systems and should be accounted for in studies using marine top predators as samplers of mid to lower trophic level species. Ultimately, flexibility in foraging strategies and inter-colony variation will play a critical role when assessing factors such as inter-specific competition or overlap with anthropogenic activities.