No abstract

Palatability of outer tissues of a suite (12 species) of Antarctic ascidians was evaluated using omnivorous fish and sea star predators. Tissues of 100% of those tested were unpalatable to fish, while 58% were unpalatable to sea stars. Lipophilic and hydrophilic extracts of 11 species were incorporated into pellets and tested in fish and sea star bioassays. Only the lipophilic extract from Distaplia colligans caused fish feeding deterrence. Organic extracts from 10 ascidian species were also examined in food pellet assays using an omnivorous amphipod. Only the lipophilic extract of D. cylindrica was a deterrent. Five of the ascidians possessed acidified outer tunics (pH < 3). We tested the ability of acidified krill pellets (pH 2 to 7) to deter fish and sea star predators and found that, while fish readily ingested acidified food pellets (pH 2), sea stars were deterred at pH 5 or less. Thus either organic or inorganic chemical defenses explain defense in 5 of the 7 ascidian species found unpalatable to sea stars. In contrast, chemical defenses only explain 1 of 12 species found unpalatable to fish, and only 1 of 10 ascidians tested against an amphipod predator. This predator-specific pattern of chemical defense may reflect greater predation pressure on ascidians from Antarctic sea stars. Alternatively, Antarctic ascidians may rely on other factors such as the toughness of their tunic or sequestration of heavy metals such as vanadium to inhibit feeding by Antarctic fish, a taxonomic group known to lack strong jaws.

Coccolithophores have global ecological and biogeochemical significance as the most important calcifying marine phytoplankton group. The structure and selection of prokaryotic communities associated with the most abundant coccolithophore and bloom-forming species, Emiliania huxleyi, are still poorly known. In this study, we assessed the diversity of bacterial communities associated with an E. huxleyi bloom in the Celtic Sea, exposed axenic E. huxleyi cultures to prokaryotic communities derived from bloom and non-bloom conditions and followed the dynamics of their microbiome composition over one year. Bloom-associated prokaryotic communities were dominated by SAR11, Marine group II Euryarchaeota, Rhodobacterales and contained substantial proportions of known indicators of phytoplankton bloom demises such as Flavobacteriaceae and Pseudoalteromonadaceae. Taxonomic richness of replicated co-cultures resulting from natural communities with axenic E. huxleyi rapidly shifted and then stabilized over time, presumably by ecological selection favoring more beneficial populations. Recruited microbiomes from the environment were consistently dependent on the composition of the initial bacterioplankton community. Phycosphereassociated communities derived from the E. huxleyi bloom depth were highly similar to one another, suggesting deterministic processes, whereas cultures from non-bloom conditions show an effect of both deterministic processes and stochasticity. Overall, this work sheds new light on the importance of the initial inoculum composition in microbiome recruitment and elucidates the temporal dynamics of its composition and long-term stability.