From the product to the process: Fine-scale skeletal structures of scleractinian corals and their relevance to biomineralization models, geochemical sampling strategies, and phylogenetic reconstructions

Interpretation of biomineralization processes based solely on the end product - the skeleton - may seem abreakneck task. This is because the biologically controlled mineralization involves cascades of intermediatephases and the activity of a myriad of organic macromolecules, of which only a fraction is embedded and/orhave been identified in the biomineral skeleton; the final product of these processes. Conversely, however,often some tiny differences in skeletal structures provide a hint of distinct biomineralization process, whichonly in this way can be better characterized. An example is distinct skeletal microstructures of twoscleractinian coral groups - gardineriids and micrabaciids - which prompted their inclusion in molecularphylogeny studies that resulted in discovery of their basal position among all other clades of Scleractinia.This presentation is devoted to the identification of processes involved in the formation of two distinctskeleton regions traditionally referred to as "centers of calcification" (CoC) and "fibers". These structures canbe distinguished in all well-preserved skeletons of modern and fossil corals as mostly nanogranular CoCs (infossils often diagenetically altered) and fibers composed of larger crystals. They occur in essentially all ofskeletal parts, including septa, thecal and axial structures witnessing the universal mechanism ofhexacorallian mineralization. Elemental labelling of growing skeleton (modern Stylophora and Galaxea)shows that, contrary to the traditional interpretation that CoCs are formed first in a two-step growth processand subsequently overgrown by fibers, both structures are in fact formed simultaneously but they exhibitvery different growth dynamics. The formation rate of CoC is distinctly higher than fibers, which justifiesreferring to these two structural elements as rapid accretion deposists (RADs) and thickening deposits (TDs),respectively. The structure of RADs is remarkably similar in all scleractinian clades, whereas TDs very oftenshow patterns reflecting the complex calicoblastic ectoderm topography. For that reason, TDs patterns oftenbear stronger phylogenetic signal than distribution pattern of RADs (but the latter can also be distinct e.g., inflabelliids or dendrophylliids) and may support some well-defined molecular clades, e.g., acroporiids,micrabaciids, flabellids, lobophylliids, pocilloporiids. Furthermore, contrary to some established coralskeleton growth models, the accretion of skeletal layers in corals living in the photic zone takes placesimultaneously in RADs and TDs areas during the day and night; this has a significant implication forstrategies of precise geochemical skeletal sampling. Tips of septal face granulations, septal edge dentitions,or sharp growing edges, are areas formed primarily by RADs, whereas structures directly adjoining to themare composed of TDs. The skeletal organic matrix in the RADs and TDs regions has different proteomiccharacteristics, which points to biochemically distinct calicoblastic cell activity in these two regions. Theadjacent TDs fibers may show clear boundaries (traditional concept of trabeculae) or show smooth transition(even in the ontogeny of a single skeletal element). Because "trabeculae" are only derivative of differences inthe growth dynamics between RADs and TDs, while most relevant information for biomineralization controlis included in RADs and TDs structural patterns and their biogeochemical composition, it is not surprisingthat phylogenetic reconstructions based on the traditional understanding of microstructures have failed inconfrontation with modern molecular phylogenetics.