'Ten Years After'-a long-term settlement and bioerosion experiment in an Arctic rhodolith bed (Mosselbukta, Svalbard).

Rhodolith beds and bioherms formed by ecosystem engineering crustose coralline algae support the northernmost centres of carbonate production, referred to as polar cold-water carbonate factories. Yet, little is known about biodiversity and recruitment of these hard-bottom communities or the bioeroders degrading them, and there is a demand for carbonate budgets to include respective rates of polar carbonate build-up and bioerosion. To address these issues, a 10-year settlement and bioerosion experiment was carried out at the Arctic Svalbard archipelago in and downslope of a rhodolith bed. The calcifiers recorded on experimental settlement tiles (56 taxa) were dominated by bryozoans, serpulids and foraminiferans. The majority of the bioerosion traces (30 ichnotaxa) were microborings, followed by attachment etchings and grazing traces. Biodiversity metrics show that calcifier diversity and bioerosion ichnodiversity are both elevated in the rhodolith bed, if compared to adjacent aphotic waters, but these differences are statistically insignificant. Accordingly, there were only low to moderate dissimilarities in the calcifier community structure and bioerosion trace assemblages between the two depth stations (46 and 127 m), substrate orientations (up- and down-facing) and substrate types (PVC and limestone), in that order of relevance. In contrast, surface coverage as well as the carbonate accretion and bioerosion rates were all significantly elevated in the rhodolith bed, reflecting higher abundance or size of calcifiers and bioerosion traces. All three measures were highest for up-facing substrates at 46 m, with a mean coverage of 78.2% (on PVC substrates), a mean accretion rate of 24.6 g m-2   year-1 (PVC), and a mean bioerosion rate of -35.1 g m-2  year-1 (limestone). Differences in these metrics depend on the same order of factors than the community structure. Considering all limestone substrates of the two platforms, carbonate accretion and bioerosion were nearly in balance at a net rate of -2.5 g m-2  year-1 . A latitudinal comparison with previous settlement studies in the North Atlantic suggests that despite the harsh polar environment there is neither a depletion in the diversity of hard-bottom calcifier communities nor in the ichnodiversity of grazing traces, attachment etchings and microborings formed by organotrophs. In contrast, microborings produced by phototrophs are strongly depleted because of limitations in the availability of light (condensed photic zonation, polar night, shading by sea ice). Also, macroborings were almost absent, surprisingly. With respect to carbonate production, the Svalbard carbonate factory marks the low end of a latitudinal gradient while bioerosion rates are similar or even higher than at comparable depth or photic regime at lower latitudes, although this might not apply to shallow euphotic waters (not covered in our experiment), given the observed depletion in bioeroding microphytes and macroborers. While echinoid grazing is particularly relevant for the bioerosion in the rhodolith bed, respective rates are far lower than those reported from tropical shallow-water coral reefs. The slow pace of carbonate production but relatively high rates of bioerosion (both promoted by low carbonate supersaturation states in Arctic waters), in concert with high retention of skeletal carbonates on the seafloor and no calcite cements forming in open pore space created by microborers, suggest a low fossilisation potential for polar carbonates, such as those formed in the Mosselbukta rhodolith beds.