Map showing geographical location of Taiwan. The experimental burial site is at National Yunlin University of Sci- ence and Technology in Douliou (starred). 

Context in source publication

Context 1

... use of electron microscopy, many of these features were found to have a fine structure comprising numerous sub-micron tunnels with diameters between 400 nm and 800 nm. These small tunnels are confined to discrete zones, each 10–40 μ m across, imparting a spongiform appearance to the affected bone and often surrounded by a hypermineralised border ( B ell et al. 1991, 1996; J ackes et al. 2001; T urner -W alker et al. 2002). A more recent review and classification of tun- nelling morphologies is provided by J ans (2008). The most dominant tunnelling morphology, sub-micron spongiform porosity ( T urner -W alker et al. 2002 ) is most commonly identified in archaeological bones excavated from relatively shallow, aerated soils. It is absent from bones excavated from anoxic soils below the water table and in bones from very cold climates. Although several attempts have been made to iden- tify the microorganisms responsible for the above mentioned features, the results have been inconclusive. M archiafava et al. (1974) reported rapid and luxuriant growth of fungal mycelium on specimens of human vertebrae buried in garden soil at room temperature after only a few weeks. Using optical microscopy combined with scanning- and transmission electron microscopy they were able to see that fungal hyphae had penetrated the pore structure of the bone and there was some pitting and tunnelling of the bone matrix. They also reported local demineralization of the bone adjacent to fungal hyphae in an advanced stage of de- generation and speculated that decalcification of the bone may result from the aging of the fungal colony. They identify the organism responsible for tunnelling into soil buried bone as belonging to the genus Mucor but from their methods section it is clear that some of the bones used in their study were autoclaved at 100°C or 200°C for 20 minutes prior to burial. This heat treatment would inevitably compromise the integrity of the protein-mineral bond and make the bone tissue more susceptible to microbial degradation, potentially invalidating their conclusions. Other investigations of early bioerosion in skeletal remains have relied upon field studies in which animal carcasses have been intentionally exposed for several years ( F ernández -J alvo et al. 2010) or where forensic cases have yielded skeletonised remains ( Y oshino et al. 1991; B ell et al. 1996 ). These suggest that micro- bial destruction of histology is limited in sub-aerially exposed bones, more rapid in bones buried in soils and more rapid still in bones exposed in the sea . Y oshino et al. ( 1991, Table 1) report that bacterial alteration of surface exposed bone does not occur until 15 years post-mortem, whereas bone buried in soil exhibits limited bacterial tunnelling around secondary osteons after five years and extensive destruction of the outer surface and inner cortex between six and ten years. For bones exposed to the sea, tunnelling appears only on the outer surfaces after 4-5 years ( Y oshino et al. 1991). F ernández -J alvo et al. ( 2010) report similar findings for animal bones exposed on the surface of Welsh uplands after three decades. The circumscribed timescales and characteristic morphologies of these changes in skeletal remains lend themselves to their use as indicators of the early taphonomic histories of excavated bones and they have been used as clues of early depositional environments and subsequent changes to those environments ( T urner -W alker & J ans 2008 ). Unfortunately, the same timeframes that make microbial bioerosion so useful as indicators of early taphonomy also make them difficult to study un- der controlled conditions. The following study exploits the warm and humid conditions of the tropical climate in Taiwan to accel- erate microbial activity in a closely controlled field setting and thus shed light on the very early coloni- sation of bone by microorganisms. This is important because previous studies of truly ancient bone have demonstrated that once microbial destruction of bone tissues begins, it can radically alter a large percentage of the available cortical bone and destroy any evidence for the initial stages of attack. Thus excavated bones exhibit either well preserved or poorly preserved his- tology and this is reflected in both the measured Ox- ford Histological Index (OHI) and the nitrogen content of archaeological bones ( N ielsen -M arsh & H edges 2000 ) as well as tensile strength ( T urner -W alker & P arry 1995). Similarly, fossil bones rarely exhibit mi- crobial degradation and are characterised by high OHI values ( T rueman & M artill 2002 ). Conversely, this study presents data for bones that do preserve remnant microbial activity in their histology. All the bone specimens were installed within the cam- pus area of National Yunlin University of Science and Technology (NYUST) in Central Taiwan. Taiwan lies off the south east coast of China from which it is sepa- rated by the Taiwan Straits (Fig. 1) and NYUST sits just north of the Tropic of Cancer. It therefore experi- ences a marine tropical climate with average annual temperatures of 22-24 °C and average rainfall of 2100 mm. Humidity varies from 78% to 84%. Around the university campus the rainfall is concentrated into the typhoon season of the summer months during which up to 80% of the annual precipitation occurs (http:// twgeog.geo.ntnu.edu.tw/english/climatology/climatol- ogy.htm). Samples were installed at three locations, all within a radius of approximately 20 metres: dry soil adjacent to a small stream, permanently water- logged soil on the banks of the stream and within the stream water itself. These sites lie within an enclosed area containing the university’s water treatment plant and are set away from human activities. The methods adopted in this study were standardised protocols that have been used successfully in previous studies ( T urn - er -W alker & P eacock 2008 ). Cow legs from below the carpal/tarsal joint were ob- tained frozen from a local slaughterhouse and kept fro- zen at -18 °C until used. The diaphyses of the metacar- pals/metatarsals were sawn from the whole legs and the remains discarded. The diapheses, approximately 120 mm in length were de-fleshed and had the marrow removed mechanically whilst still frozen. They were then washed briefly in warm water with a few drops of non-ionic detergent, rinsed, blotted dry and allowed to dry in air at 20 °C and 50% relative humidity. Once dry they were again frozen until use. In addition to the recent bovine bone samples used in the burial experiments, two additional samples of fossil bone were examined in the SEM. One was a fragment of unidentified fossil mammal rib from China dating to the Miocene (24-5.3 Mya). The sec- ond was a small fragment of parasaurolophus skull from the Late Cretaceous (76.5-73 Mya). Both of these specimens were obtained from a fossil preparator in Banqiao, Taipei City. They were embedded, polished and coated for SEM as described for the bovine bone below. The methods followed those described in T urner - W alker & P eacock (2008 ). Boreholes were hand- drilled through the overburden to the target burial layers using a large Archimedes auger (seen leaning against the wall in Fig. 2A). Sections of metaphyses (approximately 100 mm long) were loaded into sample tubes fabricated from polyethylene pipes, 1300 mm x 160 mm of which the lower 400 mm was divided into sub-compartments using inert nylon mesh partitions. Numerous (c. 200) large holes (25 mm diameter) pro- vided for relatively free circulation of ground waters. The bone sections (samples of natural fibre textile, leather and horn were included in separate compart- ments of the installation module) were packed, togeth- er with sediment recovered during drilling (wetted when necessary to ensure close contact between sedi- ment and sample) into the lower sections of the sample module which was then lowered back into the borehole (Fig. 2B). Approximately 200 mm of the upper part of the tubes remained above ground to facilitate retrieval (Fig. 2C). The interiors of the sample modules were filled with the sediments removed during drilling. A temperature datalogger was installed with one probe measuring air temperature and a second probe insert- ed 2000 mm below ground at the same level as the samples in the dry soil. The datalogger was housed in a shaded and perforated shelter (Fig. 2D). Four pipes were installed, two in dry soil and two in waterlogged soil. One bone sample was also installed in the stream itself (Figs. 3A & 3B). This was secured with a ny- lon lock-tie to a thin polyethylene pipe that was then pushed into the soft stream bed such that the sample was suspended approximately 100 mm above the bot- tom and approximately 200 mm below the surface of the water. One installation module from each site (dry and wa- terlogged soil) was recovered after one year. This was achieved on the dry site by drilling a second borehole immediately adjacent to the installation tube and then digging around the pipe before hauling the whole tube free using large holes drilled at the top of the pipe. The sample module could be retrieved from the stream bank by rocking the pipe from side to side and pulling it free. The modules were transferred to the laboratory and the nylon lock ties cut, thus freeing the soil and samples. The whole block of soil could then be pushed out of the tube using a crude ramrod. Soil blocks from each compartment were carefully excavated and the bone samples cleaned of loosely adhering sediment. The bones were then non-vacuum freeze-dried in a freezer compartment at -18 °C. The bone samples placed in the stream were recovered after 6 months by simply pulling the plastic pipe free of the mud and washing the bone in the stream water. After transferral to the laboratory the bone samples were blotted and solvent-dried through ethanol baths. The principal microscopic examination ...