Surviving under the Antarctic sea ice : a study of the feeding ecology of Antarctic krill

Antarctic sea ice covers approximately 4 million km2 during minimum extent in February/ March and grows to 19 million km2 during maximum extent in September/October. This sea ice zone harbours a wide diversity of biota and supports large populations of unique Antarctic organisms. During the satellite era, Antarctic sea ice has shown strong regional changes in its extent, duration, and the timing of the annual advance and retreat. Improving our understanding of the relationships between sea ice and ice associated animals will fill a crucial knowledge gap, which will facilitate the conservation and management of the Southern Ocean ecosystems and resources. This thesis compiles three studies focusing on the feeding ecology of Antarctic ice associated zooplankton during the winter-spring transition. This work concentrates on Antarctic krill (Euphausia superba), a key species with ecological and commercial significance in Southern Ocean ecosystems. Morphology defines an animal's feeding capability. The morphological changes through ontogeny of a species thus are not only important taxonomic information, but also crucial for our understanding regarding its feeding ecology. In Chapter 2, I describe the morphological changes of E. superba through its development from an egg to juvenile. This is the first study since 1936 that fully describes the morphological features of this species, in which I update the existent knowledge with details and high-resolution photographs. Intermediate larval stages are identified and these morphological characteristics are related to the overwintering survival strategies for this species. The food that animals actually obtain from the environment can be investigated with either gut content or biochemical analyses. To determine the dietary preferences and trophic relationships of major zooplankton species (Chapter 3), I use stable isotope analyses (13C/12C and 15N/14N) to compare samples collected from East Antarctic pack ice zone during two winter-spring transitions (2007 and 2012). Interannual dietary differences are determined for larval E. superba, suggesting feeding plasticity, which enables this species to adapt to changing environmental conditions. Larval E. superba are primarily herbivorous while utilising sea-ice biota, and consume a more heterotrophic diet when feeding from the water column. In contrast, post-larval E. superba, and the omnivorous krill Thysanoessa macrura consume a mixed diet from both the water column and the sea ice. The pteropod Limacina helicina, small copepods in the genus Oithona spp., ostracods and amphipods rely heavily on sea-ice biota according to their carbon isotope ratios. Large copepods and chaetognaths consume a water column-based diet during the winter. The comparison of isotopic profiles between years suggests that ice-associated zooplankton gain access to sea-ice biota more easily under warm and permeable ice than under cold ice. Our ability to construct food web models for sea-ice ecosystems are restricted by limited real data and inadequate understanding regarding the system structure. A qualitative modelling approach is applied in Chapter 4 to explore how the increasing model complexity affects model predictions, and to identify key variables and interactions governing krill dynamics in this system. A series of qualitative network models are constructed to represent different theories of krill winter feeding, and winter conditions are simulated by applying a perturbation of an increase in sea-ice algae and a decrease in pelagic phytoplankton concurrently. Results demonstrate the importance of including developmental-stage-specific information during the construction of food web models. In addition, model outcomes suggest that the coupling between primary production (either sea-ice or pelagic) and protozoan production should be investigated in future empirical studies as it is a key process affecting model predicted krill responses. This study also demonstrates the usefulness of qualitative network modelling in hypotheses testing regarding ecological processes. This thesis has contributed to a better understanding of feeding ecology of Antarctic krill and other under-ice zooplankton during the winter-spring transition period. My results suggest that future studies regarding feeding ecology of ice-associated zooplankton should use combined approaches. The use of qualitative modelling could help evaluate assumptions based on limited data, which is advantageous for identifying critical ecological processes and pinpointing the empirical observations that are needed. In addition, advanced technologies, such as high-resolution underwater camera and underwater vehicles, will allow us to obtain habitat-structure information and detailed in situ behavioural observations, which should also be evaluated and considered in future feeding studies.