LAUSR

Abstract Microbial bioerosion, in all its manifestations, is one of the major factors determining the long-term survival of archaeologically and environmentally important artefacts and ecofacts made from mineralised collagen – bones, antler, teeth and ivory. The bone diagenesis literature contains extensive descriptions of different morphologies and classifications of microbial bioerosion – microscopical focal destruction, Wedl tunnels, linear longitudinal, budded and lamellate tunnelling, etc. but the causative agents remain to be discovered. Palaeontologists are in a similar situation when describing ichnofossils where bioerosion is classified by its characteristic morphology and only tentatively assigned to specific causes. In archaeological bones, Wedl tunnels have traditionally been ascribed to fungi but a re-examination of Wedl's original paper and subsequent literature has shown that all the early specimens examined came from aquatic environments and that euendolithic microflora (cyanobacteria or chlorophytes) might have been responsible. These microorganisms are known to tunnel into marine shells. Linear longitudinal, budded and lamellate tunnelling have all been ascribed to bacteria. However, analyses using a combination of backscatter SEM (BSEM) and mercury intrusion porosimetry (HgIP) of bones excavated from terrestrial soils suggest that the various tunnelling morphologies described by earlier researchers are actually all manifestations of a single architecture, the differences arising from the inherent variability of bone microstructure and the hydrology of the burial environment. An examination of the BSEM and HgIP data also indicates that the bacteria responsible may spread through dead bone tissues by expanding the canalicular network (or dentinal tubules in teeth) rather than creating new tunnels. Long-term field burial experiments using de-fleshed cow bone specimens have now demonstrated that bacterial tunnelling develops over decades rather than months as was previously thought, even in warm tropical soils. This has obvious implications for the origin of the bacteria responsible (soil bacteria versus endogenous gut bacteria) since the bone specimens were quickly isolated from gut contents. Although no attempt is made here to identify a specific organism or organisms responsible for tunnelling in terrestrial archaeological bones a tentative model is proposed for how cycles of wetting and drying, or changes in local dissolved oxygen levels, could lead a common gram positive soil bacterium to produce the pattern of bacterial bioerosion seen in exhumed bones. This soil organism may act in a similar way to Staphylococcus aureus which proliferates in living bone (causing osteomyelitis) by infiltrating and enlarging the canalicular network.