Unique Marine Taphonomy in Human Skeletal Material Recovered from the Medieval Warship Mary Rose L. S. BELL a * AND A. ELKERTON b a School of Criminology, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, V5A 1S6, Canada b The Mary Rose Trust, College Rd, HM Naval Base, Portsmouth, PO1 3LX, UK ABSTRACT The effect of skeletal exposure in a marine environment is an area of taphonomy that has been little investigated at the microscopic level. Understanding the peri-mortem and subsequent post mortem history of deposition and/or redeposition is extremely important for event reconstruction and to identify deliberate or accidental redeposition. The material used for this study comes primarily from the Mary Rose shipwreck (a marine mass fatality dated AD 1545), and forensic material recovered from marine, lacustrine and terrestrial contexts is retrospectively referenced. Work presented here outlines a definitive type of marine exposure seen in temperate shallow off-shore and intertidal marine contexts, and illustrates how it may be differentially identified from terrestrial deposition and exposure. Furthermore, the effects of rapid deposition on skeletal remains have been documented, and results indicate that marine organism fouling activity can be fully inhibited by rapid deposition of sediment. The respon- sible organism itself remains unidentified, but produces tunnels which are peripheral in their distribution and maintain fixed dimensions and morphology and are here associated with marine exposure. This type of microstructural change is unique and is not found in terrestrial or freshwater contexts. The study demonstrates a taphonomic microstructural change to bone and teeth which may be identified microscopically and interpreted as evidence of marine exposure. Secondarily, the history of depositional exposure between the two main Tudor layers has provided a new level of detail concerning exposure and site formation processes. The earliest Tudor layer formed rapidly over a period of months and contained no evidence of microstructural tunnelling, whereas microstructural tunnelling was seen exclusively in the second Tudor layer, formed over a period of decades, a period during which the ship’s hull collapsed and a more open marine environment dominated. Copyright ß 2007 John Wiley & Sons, Ltd. Key words: marine taphonomy; marine decomposition; human remains; mass fatality; exposure; microscopy; skeletal remains; diagenesis Introduction The bulk of taphonomic literature related to marine exposure usually concerns microbial fouling, seen as microscopic tunnelling of either calcareous or wooden substrates. Much of the literature has focused on farming activities of oysters or the protection of coral reefs from the deleterious effects of microbial bioerosion. Whilst the Mary Rose wreck is a historical context, it has provided a unique opportunity to assess fully the outcome of short- and International Journal of Osteoarchaeology Int. J. Osteoarchaeol. (2007) Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/oa.952 * Correspondence to: School of Criminology, Simon Fraser Uni- versity, 8888 University Drive, Burnaby, British Columbia, V5A 1S6, Canada. e-mail: [email protected]Copyright # 2007 John Wiley & Sons, Ltd. Received 22 February 2007 Revised 5 July 2007 Accepted 6 August 2007
Transcript
International Journal of OsteoarchaeologyInt. J. Osteoarchaeol. (2007)Published online in Wiley InterScience
52
(www.interscience.wiley.com) DOI: 10.1002/oa.9
* Correspondence to: Schooversity, 8888 University Dri1S6, Canada.e-mail: [email protected]
Copyright # 2007 Joh
Unique Marine Taphonomy inHuman Skeletal MaterialRecovered from the MedievalWarship Mary Rose
L. S. BELLa* AND A. ELKERTONb
a School of Criminology, Simon Fraser University, 8888 University Drive, Burnaby, British
Columbia, V5A 1S6, Canadab The Mary Rose Trust, College Rd, HM Naval Base, Portsmouth, PO1 3LX, UK
ABSTRACT The effect of skeletal exposure in a marine environment is an area of taphonomy that has beenlittle investigated at the microscopic level. Understanding the peri-mortem and subsequentpost mortem history of deposition and/or redeposition is extremely important for eventreconstruction and to identify deliberate or accidental redeposition. The material used forthis study comes primarily from the Mary Rose shipwreck (a marine mass fatality dated AD1545), and forensic material recovered from marine, lacustrine and terrestrial contexts isretrospectively referenced. Work presented here outlines a definitive type of marine exposureseen in temperate shallow off-shore and intertidal marine contexts, and illustrates how it maybe differentially identified from terrestrial deposition and exposure. Furthermore, the effects ofrapid deposition on skeletal remains have been documented, and results indicate that marineorganism fouling activity can be fully inhibited by rapid deposition of sediment. The respon-sible organism itself remains unidentified, but produces tunnels which are peripheral in theirdistribution and maintain fixed dimensions and morphology and are here associated withmarine exposure. This type of microstructural change is unique and is not found in terrestrial orfreshwater contexts. The study demonstrates a taphonomic microstructural change to boneand teeth which may be identified microscopically and interpreted as evidence of marineexposure. Secondarily, the history of depositional exposure between the two main Tudorlayers has provided a new level of detail concerning exposure and site formation processes.The earliest Tudor layer formed rapidly over a period of months and contained no evidence ofmicrostructural tunnelling, whereas microstructural tunnelling was seen exclusively in thesecond Tudor layer, formed over a period of decades, a period during which the ship’s hullcollapsed and a more open marine environment dominated. Copyright � 2007 John Wiley &Sons, Ltd.
Key words: marine taphonomy; marine decomposition; human remains; mass fatality;
The bulk of taphonomic literature related tomarine exposure usually concerns microbial
l of Criminology, Simon Fraser Uni-ve, Burnaby, British Columbia, V5A
n Wiley & Sons, Ltd.
fouling, seen as microscopic tunnelling of eithercalcareous or wooden substrates. Much of theliterature has focused on farming activities ofoysters or the protection of coral reefs from thedeleterious effects of microbial bioerosion.Whilst the Mary Rose wreck is a historicalcontext, it has provided a unique opportunityto assess fully the outcome of short- and
Received 22 February 2007Revised 5 July 2007
Accepted 6 August 2007
L. S. Bell and A. Elkerton
long-term marine exposure of human skeletalmaterial, and has provided an invaluable tapho-nomic proxy for examining silt progression andsite formative processes.The decompositional stages of a body in
marine and lacustrine environments are generallyunderstood (Simpson & Knight, 1985), andmacroscopic changes to bony surfaces are knownto be mediated by larger marine organisms (Sorget al., 1997; London et al., 1997). The usefulnessand taphonomic significance of marine-relatedmicroscopic changes to bone as a type ofexposure are anecdotally reported. Occasionallyhuman skeletal remains are recovered from exoticcontexts; for example, Iscan & McCabe (1995)reported on the recovery of partially articulatedhuman skeletal remains from the gut of a sharkand described the macroscopic effects on exteriorbone surfaces. Arnaud et al. (1978) reportedmicroscopic tunnelling in human skeletal materialin association with a Mediterranean ship wreck;whilst Ascenzi & Silvestrini (1984) experimentallysubmerged defleshed bovine bone in a similarmarine context for one year and found equivalent
Figure 1. Endolithic micro-boring seen extending into thedistance beneath the enamel. The enamel remains unaffecmicrograph (SEM) using backscattered electron imaging (B
Copyright # 2007 John Wiley & Sons, Ltd.
tunnelling. Bell et al. (1991) made an initialobservation on human dental material from MaryRose, observing a network of tunnels invadingperipherally without any reference to tissueorganisation (Figure 1). Work on forensicmaterial recovered from intertidal Canadianwaters demonstrated peripheral microscopictunnelling (Figure 2; Bell et al., 1996), and inno instance has this type of tunnelling beenassociated with lacustrine or terrestrial depositionand/or exposure (Bell et al., 1991, 1996; Bell &Lee-Thorp, 1998).
Microstructural micro-boring has been docu-mented with greater detail in other marinesubstrates. Living coral reefs suffer cyclical attacksby algal communities between the months of Mayand September (Highsmith, 1981; Perry, 2000).Other micro-organisms such as Polychaetes(Sato-Okoshi et al., 1990) and Thraustochyrids(Porter & Lingle, 1984; Chamberlain & Moss,1988) are silt-sensitive and also colonise a range ofcalcareous substrates cyclically, and produce smallperipheral tunnels. Cyanobacteria have also beenimplicated (Raghukumar et al., 1989), as have some
neck of the tooth peripherally and extending for someted by this tunnelling. Individual MR 4: scanning electronSE). Vertical field width: 430 microns.
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
Figure 2. Endolithic micro-boring seen extending beneath the neck of the tooth, peripherally extending beneath, but notintruding into, the enamel. This material was recovered from a shallow cold-water intertidal Canadian context and hasbeen maximally exposed for seven years since death. Note the exact similarity in size and distribution to Figure 1. SEM/BSE image: see magnification bar.
Mary Rose Taphonomy
marine fungi (Zeff & Perkins, 1979). Golubic et al.(1975) gave a fascinating review of tunnelling inrelation to the fossilisation process and called thistype of invasive change ‘endolithic tunnelling’.They suggested that endoliths will create type-specific tunnelling and that their activity will beaffected by depth, temperature and light penetra-tion. They referred to experimental work whichdemonstrated endolithic micro-boring within12 days to two months of initial colonisation(Golubic et al., 1975). More recently, experimentalwork by Wisshak et al. (2005) extendedGolubic et al.’s earlier observations in a coldtemperature setting, documenting a range ofmicro-boring endoliths penetrating an exper-imental substrate composed of bivalve shells,exposed and observed over a two-year period.The study characterised a number of micro-borers and observed a similar light/depthrelationship observed by Golubic et al. (1975).Other larger marine organisms are known toproduce tunnelling and include octopi (Nixon &Maconnachie, 1988), sea snails (Symth, 1988),
The Mary Rose warship was the Vice Flagship ofHenry VIII and sank with virtually all hands loston 19th July AD 1545. The ship itself was fullyloaded with mariners and soldiers ready to engagethe French fleet in the Solent. She sank suddenlybefore reaching the French fleet, in approxi-mately 12–14m of water, and her demise remainsan unsolved and historically sensitive mystery.Several theories have been advanced, includingsinking from French cannon fire, or an awkwardmanoeuvre which caused seawater to floodinto open gun-ports. She was carrying a crewcomplement of 415 men, all of whom were lostexcept for those few men stationed in the rigging(Stirland, 2000). This marine accident representsa medieval disaster on a parallel with the sinkingof a large aircraft carrier, and in financial terms
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
L. S. Bell and A. Elkerton
was of equal significance. From a taphonomicperspective, Mary Rose represents a uniquehistorical marine mass fatality where most diedof drowning or severe trauma, were immediatelydeposited into a marine environment, decom-posed within the ship itself, and the skeletalremains were subsequently exposed differentiallyto an enclosed and latterly a more open marineenvironment (Marsden, 2003).The depositional history of the wreck was not
revealed until excavation and lifting. It wasdetermined that after sinking, the ship came torest on her starboard side which resulted in ashifting of the internal ballast. The hull rapidlyinfilled with current-borne estuarine grey siltswhich settled as a fine-grained sediment withinthe ship. This distinct layer was determined tohave been deposited rapidly over a period ofmonths and constituted what is referred to as thefirst Tudor layer (Marsden, 2003). During theslower formation of the second Tudor layer,formed over decades, seaweed lenses wereincluded within a light grey, shelly clay mix,indicating a stable and partially exposed sea floor.During the formation of this second layer, theupper aspect of the hull collapsed and the interiorof the ship became exposed to a much more openmarine environment (Marsden, 2003). The shipitself was not considered encased by sedimentuntil the third layer was deposited during thelate 16th and 17th centuries, and consisted of ahard grey clay and broken shelly material(Marsden, 2003). A fourth layer, more mobile,being constantly reworked and redeposited,comprised the modern seafloor. Water tempera-ture during excavation had a winter-summervariation of 12–138C and 18–208C respectively(Rule, 1982).The human remains were found almost entirely
in a disarticulated and commingled state withgood representation of all elements of theskeleton (Stirland, 2000). Of the 415 menrecorded as present, only 179 were accountedfor by minimum number analysis (Stirland, 2000).It is possible that some skeletal material existedoutside of the excavation limit, which extended ametre from the remaining hull. During collapse,skeletal material may have been redepositedoutside of this excavation area and thus neverrecovered. The arrangement of the human
Copyright # 2007 John Wiley & Sons, Ltd.
material was almost entirely within the first twoTudor layers, and the commingling of elements isconsistent with marine decomposition within aconstrained space prior to deposition.
A microscopic study was undertaken todocument and assess the impact of marineexposure, if any, on the microstructural arrange-ment of human skeletal remains, and secondarilyto investigate the impact of silt progression andsite formation.
Materials and methods
A sample of 17mandibles and maxillae were takenfrom all decks and silt phases, excluding themodern seabed. During excavation the ship wasdivided into 3m2 excavation quadrants and thelocation of each specimen which constituted thesample group is represented spatially per exca-vation quadrant in Figure 3.
A single tooth and accompanying socket wasremoved from either the maxilla or mandible bycutting the entire tooth and socket free using anIsomet diamond-edged circular saw. The speci-mens were then embedded in PMMA after Bellet al. (1991), cut longitudinally buccolingually,rotary lap-polished using graded abrasives, andfinished with a 1mm diamond paste.
Uncoated blocks were then dry mounted andindividually examined under a Lasertec ILM11confocal reflection microscope (CRM) using ahelium-neon light source. This microscope hasincreased resolution over standard optical reflec-tion microscopes, and allows for direct obser-vation of tissue morphology and characterisationof any post mortem alteration. Using this system, aslight internal reflection artefact was observedand was measured to a depth of approximately2mm from the block face in depth, and again isa significant improvement over non-confocalmicroscope configurations. Invasive post mortemtunnelling was measured in the x and yorientation using microscope-interfaced Lasertecsoftware. All measurements were collected blindwithout prior knowledge of specimen locationwithin the ship stratigraphy. The distribution ofpost mortem tubule invasion was recorded in termsof total morphology and distribution, maximumingress, and maximum tubule diameter at eight
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
Figure 4. Anatomical site locations used to measureingress of the post mortem tunnelling.
Mary Rose Taphonomy
different predetermined and relative anatomicallocations per tooth (Figure 4). Each measur-ement site was then an anatomical relativeapproximation. Given the disarticulated andnon-anatomical deposition and orientation ofmandibles and maxillae within the ship, thebuccal and medial aspects of specimens wereconsidered to have been rendered anatomicallymeaningless, and so the eight sample sites pertooth were based on block face orientation alone.However, each tooth was viewed anatomicallywithin its own socket and that anatomicalrelationship was maintained. The results fromthe assessment of maximum ingress wereregraded into three arbitrarily predeterminedranges, where grade 1¼ less than 100mm(slight); grade 2¼ greater than or equal to100mm to less than 200mm (moderate); andgrade 3¼ greater than or equal to 200mm to thepulp cavity (deep). Total distribution of invasivetunnelling was recorded as unaffected (0) orbilateral (B). The results from this regrading werethen related to the ship’s stratigraphy and plotted
Figure 3. Cutaway diagram of the excavation quadrants. Shaded areas represent sampling quadrants for this study.The four decks and hold are seen in this view; during the formation of the second Tudor layer these decks becameexposed to an open seabed environment due to the ship partially collapsing in on itself.
Copyright # 2007 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
L. S. Bell and A. Elkerton
into the cumulative silt phase schematic. Theintegrity of the periodontal ligament joint (PDJ)space was also assessed.A subset of specimens affected by post mortem
ingress were examined using SEM/BSE imagingafter a study by Bell et al. (1991) where identicalingress was initially observed in Mary Rose dentalmaterial. Some of this SEM/BSE imagery isincluded for illustrative purposes only, eitherretrospectively or as part of this study.
Results
The Lasertec ILM11 CRM proved to be anexcellent tool for the rapid screening of speci-mens, and enabled the identity, location and
Figure 5. Two SEM/BSE montages of terrestrial diagenetichange driven by bacterial ingress is via the pulp cavity icontrast, marine change is peripheral, affecting the external around the external aspect of the mandible. This is a typicaarchaeological medieval cemetery context. Right:Mary Rosefrom within the ship itself. Post mortem changes are typica
Copyright # 2007 John Wiley & Sons, Ltd.
distribution of post mortem change to be accuratelyassessed and measured. The marine type changedocumented in Bell et al. (1991, 1996) was foundreplicated in all specimens affected by micro-structural post mortem alteration. This was the onlypost mortem microstructural alteration observed inthe sample group.
The distribution of the change varied from onespecimen to another in terms of invasive depthand distribution, but was always peripheral,leaving the PDJ unaffected. This replicated theearlier observation made by Bell et al. (1991, 1996)where distribution was mapped as peripheral(Figure 5). Only in one case (specimen 14) wasthe PDJ invaded. The enamel too (includingcalculus) remained unaffected, although it wascharacteristically undermined by the micro-
c change (left) and marine tunnelling (right). Terrestrialnto the open dentine tubules and branching network. Inspect of the tooth and tracking down the neck and trackingl example of this type of marine change. Left: soil-buriedspecimen without known quadrant context, but recoveredl of all changes observed within the Mary Rose sample.
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
Mary Rose Taphonomy
boring agent (Figures 1 and 2). In intenselytunnelled areas of dentine (tooth neck) noobvious directionality could be observed(Figures 1,2,5 and 6), although at the invasivefront of the tunnels an internal reflection artefactrevealed the direction of single tunnels once theydipped below the block face and bifurcated(Figure 7). This suggested that dentine tubuledirection and branching probably influence andfacilitate the initial direction of tunnelling. It ispostulated that once a post mortem tunnel has beencreated it will itself influence the direction offurther tunnelling. Alveolar bone similarly under-went post mortem peripheral tunnelling whichtypically began at the alveolar crest and trackedaround the external aspect to connect with theopposing alveolar crest (Figure 8). The invasivetunnelling lacked directionality in terms of thebony and vascular microstructure. The totaldepth of invasion varied. What is interesting isthat no remineralisation boundary is observed atthe edges of these tunnels, which is in contrast todiagenetic alteration observed in terrestrialcontexts where bacteria are considered the main
Figure 6. A SEM/BSE image of dentine with invading tunnelsborder of these tunnels and tracks both along and at right angseen in Figure 2: identical to post mortem ingress seen in Mscale bar.
Copyright # 2007 John Wiley & Sons, Ltd.
invading organism (Bell et al., 1991, 1996; Bell &Lee-Thorp, 1998).Maximum tubule diameters of invasive tunnels
were measured and ranged between 5–19mm(nearest 0.5mm). Within this range, two separatesubgroups were distinguishable between 5–8mmand 11–19mm respectively. The commonestmaximum diameter fell within the first subgroup-ing, which represented 84% of total diametersmeasured. Only circular tunnels were used for thismeasurement, although it is acknowledged thatslight sectional obliquity might contribute to aslight increase in diameter. This alone, however,does not account for the larger diametersrecorded for the larger subgrouping.The blind study results of maximum ingress per
specimen were translated into slight (1), moder-ate (2) and deep (3) levels of invasion. Totaldistribution was recorded as either unaffected (0)or bilateral (B) (Table 1). The results were thenplotted onto the silt phase schematic (Figure 9). Itwas found that most of the sample groupdeposited in the first Tudor layer (formed overa period of months) exhibited no post mortem
penetrating. Note that no demineralisation is evident at theles to the dentine tubule direction. From forensic specimenary Rose material, used for illustrative purposes only. See
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
Figure 7. Subsurface endolithic marinemicro-boring seen as an internal reflection artefact. Note that the tunnels appearto bifurcate (see arrow). MR 5: field width is 150 microns.
L. S. Bell and A. Elkerton
alteration. Only two specimens exhibited anytunnelling: specimen 15 (mid silt phase, Holdquadrant 9) exhibited slight bilateral tunnelling,mostly to the mandibular alveolar bone; andspecimen 14 (mid/lower silt phase, Hold quadrant10) which was considered to have poor prove-nance and may have been redeposited due to alocalised scouring affect (Rule, 1982; Marsden,2003). Those specimens deposited in the secondTudor layer were all affected by post mortemtunnelling, which tended to be bilateral indistribution and graded 2 and 3. Two specimensfrom the main and upper decks exhibited slightattack with bilateral distribution, whilst twospecimens from the Hold and Orlop decks wereinvaded fairly equally. One specimen situated inthe third layer was heavily invaded with slightbilateralism. This specimen had been found in anarea of scouring and may have been redeposited(Marsden, 2003). No specimens were recoveredfrom the modern mobile seabed layer.The PDJ was unaffected by post mortem
tunnelling in all specimens except one, and thisconfirms earlier observations made on the Mary
Copyright # 2007 John Wiley & Sons, Ltd.
Rose material. The post mortem tunnelling, whenpresent, consistently crossed from the neck of thetooth and progressed across an invisible line tocontinue peripherally from the alveolar crest(Figures 1 and 5). Occasionally the tunnellingdipped slightly lower than the alveolar crestinto the region of the PDJ, where tunnelling ofthe cementum might be observed. The anom-alous specimen 14, which exhibited pronouncedpost mortem tunnelling throughout the PDJ space,also had peripheral distribution which exhibitedonly slight bilateralism, and the diameters of thisspecimen fell within the 5–8mm range. Giventhat exposure was extended over decades in theformation of this layer, it would appear the PDJwas not attractive to the invading micro-organism.
Discussion
This study has demonstrated for the first time thatpost mortem alteration to skeletal microstructure in amarine context represents important taphonomic
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
Figure 8. Alveolar crest imaged as a SEM/BSE image. The peripheral nature of the post mortem endolithic micro-boringis seen to track along the external aspect. The PDJ is clearly seen as unaffected. This was the common pattern of changeseen in this study. The invasion is seen to penetrate without any reference to bony organisation. Same specimen asFigure 5: field width is 855 microns.
Mary Rose Taphonomy
and environmental indicators of marine exposure.In the context of this medieval shipwreck, sitestratigraphy and post mortem tunnelling are closelyrelated to the speed and formation of the layers,
Copyright # 2007 John Wiley & Sons, Ltd.
and this aspect underscores the need to under-stand the depositional context when human re-mains of archaeological and forensic interest arerecovered from marine, intertidal and shoreline
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
Table 1. Penetration and distribution of post mortem tunnelling per tooth
MR no. Layer Location Anatomical sites Final grade
contexts. The type of post mortem tunnellingdetailed here is unique to marine exposure, andas such, if present in a terrestrial or evenfreshwater context would indicate that the bodyhad spent some time prior to its recovery in amarine or intertidal context (Bell et al., 1996).The formation of the first Tudor layer
represents a period of rapid silting which beganimmediately after the ship sank and is consideredto have lasted only a few months (Marsden,
Figure 9. Silt phase schematic showing the major silt phassample groups both affected and unaffected by this unique
Copyright # 2007 John Wiley & Sons, Ltd.
2003). The specimens examined from this layer(except one specimen) exhibited no microstruc-tural change whatsoever. It is interesting to askwhy this should be so when it is known fromother fouling studies that a range of calcareoussubstrates can be affected. When the ship firstsank, all those individuals who were below topdeck had little chance of escape and eithersuffered death due to drowning or traumaticinjury. The bodies would have cooled rapidly at a
es, representing stratigraphic time. The locations of thetype of marine tunnelling are plotted into the stratigraphy.
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
Mary Rose Taphonomy
depth of 12–14m, and after 2–4 weeks thesodden bodies would have started to disarticulate,where bone and skin separates (Rule, 1982;Simpson & Knight, 1985). During this shortperiod the bodies would have floated freely unlesspinned down by debris, and this is supported byhuman remains being recovered in a commingledand disarticulated state (Stirland, 2000). This alsoindicates that enough time had elapsed during theformation of this first layer for skeletonisation tooccur. Another important factor to understand isthe context of the formation of this first Tudorlayer and the effect of the changed orientation ofthe ship itself. Upon sinking the ship came to reston her starboard side, which meant thatessentially this became the ship’s bottom, andtherefore the coolest and darkest part of the shipinternally. This also meant that the decks were alleffectively orientated verically, with one side ofeach deck touching the ship’s bottom aspect. Theship sank in mid-July, and so the first Tudor layeris considered to have been completed bymid-winter (Marsden, 2003). The fact that nopost mortem tunnelling is seen in this layer is assignificant as its occurrence in the second layer,and suggests that time of year, lack of light andenvironmental conditions were not conducive tothe endolithic micro-organism responsible for themajority of tunnelling observed later.The second Tudor layer formed more slowly
than the first and is considered to have takendecades to form (Marsden, 2003). Specimensfrom this layer were all tunnelled and showedconsiderable bilateralism in the overall distri-bution of peripheral tunnelling. The layerconsisted of a fine grey clay full of seaweedlenses, which suggests that this layer had enoughlight to sustain seaweed growth and that theseaweed would have provided some stability tothe sediment. It was during this period that theupper structure of the ship collapsed, exposing allthe ship’s decks to a more open seabedenvironment. This would have had the net effectof opening up a semi-enclosed system to morelight, heat, increased current-borne fauna andmarine detritus. Hence, human material wouldhave provided an ideal substrate for endolithiccolonisation where light, a slight increase intemperature, a silt-free environment and availabledetritus for feeding were newly available. Longer
Copyright # 2007 John Wiley & Sons, Ltd.
term exposure, possibly annually cyclic, alsoallows for endolithic recolonisation, and thiswould explain the differential tunnelling observedamongst the sample group. Deductively, thisopen environment was what was absent duringthe formation of the first layer, where notunnelling was observed, and is highly suggestiveof a silt-sensitive Polychaete, Thraustochyrid,algae or a cyanobacterium as the responsiblemicro-organism. This type of substrate foulingwould also help to explain the difference ininvasive depth and the bilateralism of invasion,since deeper tunnel ingress could be achieved onthe exposed upside, rather than the downsideaspect of the skull or mandible. It is interesting tonote that the two deeply tunnelled specimensshowed only slight bilateralism, and this couldrepresent movement due to scouring action andredeposition.
Conclusion
The Mary Rose shipwreck is a unique archae-ological and historically documented marinemass fatality. This unique context has allowedan assessment of short- and longer-term marineexposure of human skeletal remains similar to thatof an actualistic longitudinal experiment, wherethe taphonomic affects of marine exposure maybe examined. What is apparent from this study isthat a unique taphonomic change associated witha time period greater than one year (or less) maybe produced by invading endoliths that canproduce characteristic tunnelling, even thoughthe micro-organism responsible remains unknown.It shares identical characteristics to a previousstudy by Bell et al. (1996), where an identicalpost mortem change was documented in forensicmaterial recovered from an intertidal Canadiancontext maximally exposed for a period up toseven years since death. Fouling studies on coralsand shellfish indicate that micro-boring can occurwithin a two-month to one-year period and isseasonally cyclical (Golubic et al., 1975; Zeff &Perkins, 1979; Highsmith, 1981). The detailedexperiment by Wisshak et al. (2005) showedbioerosion within a 1–6month period, implicat-ing cyanobacteria as a strong candidate forbioerosion observed here, particularly in the
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
L. S. Bell and A. Elkerton
orientation of invasion, size and noted bifurcationof invasive tunnels. Short-term rapid siltingappears to inhibit this type of substrate fouling,and so it is important to note the temporalcontext of recovery when human remains areassociated with marine, intertidal or shorelineenvironments. Absence of this taphonomicchange does not mean there was not a periodof exposure, but where this tunnelling is found itis clear biological evidence that it did. Attentionto the peripheral distribution of this change canillustrate the aspect or orientation of depositionand whether this has changed. Invasive depth isprobably associated with cyclical tunnelling andis suggestive of prolonged exposure. The changeitself cannot be seen by eye and must be observedmicroscopically. SEM/BSE imaging for charac-terisation in combination with confocal reflectionmicroscopy allowed for both rapid and accurateassessment and is recommended.
Acknowledgements
We thank the Mary Rose Trust for providing thematerial for this study. We also thank Sheila J.Jones and Alan Boyde for comments and adviceon this study. Thanks too to Eva Snirer for herexpert assistance with editorial and graphics. Thisstudy was funded in part by the Medical ResearchCouncil (LSB).
References
Arnaud G, Arnaud S, Ascenzi A, Bonucci E, GrazianiG. 1978. On the problem of the preservation ofhuman bone in sea-water. Journal of Human Evolution7: 409–420.
Ascenzi A, Silvestrini G. 1984. Bone-boring marinemicro-organisms: an experimental investigation.Journal of Human Evolution 13: 531–536.
Bell LS, Boyde A, Jones SJ. 1991. Diagenetic alterationto teeth in situ illustrated by backscattered electronimaging. Scanning 13: 173–183.
Bell LS, Skinner MF, Jones SJ. 1996. The speed of postmortem change to the human skeleton and its tapho-nomic significance. Forensic Science International 82:129–140.
Bell LS, Lee-Thorp JL. 1998. Advances and constraintsin the study of human skeletal remains: a joint
Copyright # 2007 John Wiley & Sons, Ltd.
perspective. In Grave Concerns, Cox M (ed.). CBA:York; 238–246.
Chamberlain AHL, Moss ST. 1988. The thraustochy-trids: a protist group with mixed affinities. Biosystems21: 341–349.
Golubic S, Perbius RD, Lukas KL. 1975. Boringmicroorganisms and microborings in carbonatesubstrates. In The Study of Trace Fossils, Frey RW(ed.). Springer-Verlag: New York; 229–259.
Highsmith RC. 1981. Lime-boring algae in hermaty-pic coral skeletons. Journal of Experimental Biology andEcology 55: 267–281.
Iscan MY, McCabe BQ. 1995. Analysis of humanremains recovered from a shark. Forensic ScienceInternational 72: 15–23.
LondonMR, Krolikowski FJ, Davis JH. 1997. Burials atsea. In Forensic Taphonomy, Haglund WD, Sorg M(eds). CRC Press: Boca Raton; 615–622.
Marsden P. 2003. Sealed by Time: the Loss and Recovery ofthe Mary Rose. Mary Rose Trust: London.
McClanahan TR, Kurtis JD. 1991. Population regula-tion of the rock-boring sea urchin Echinometramathaei (de Blainville). Journal of Experimental MarineBiology and Ecology 147: 121–146.
Nixon M, Maconnachie E. 1988. Drilling by Octopusvulgaris (Mollusca: Cephalopoda) in the Mediter-ranean. Journal of Zoology, London 216: 687–716.
Perry CT. 2000. Macroboring of Pleistocene commu-nities, Falmouth Formation, Jamaica. Palaios 15:483–491.
Porter D, Lingle WL. 1984. Marine shell boringmicroorganisms include Thraustochytrids and otherheterotrophic Protista. Journal of Protozoology 31(4)20A– 21A.
Raghukumar C, Rao VPC, Iyer SD. 1989. Precipitationof iron in Windowpane oyster shells by marineshell-boring cyanobacteria. Geomicrobiology 7: 235–244.
Rule M. 1982. Mary Rose: The Excavation and Raising ofHenry VIII’s Flagship. Conway Maritime Press:London.
Sato-OkoshiW, Sugawara Y, Nomura T. 1990. Repro-duction of the boring polychaete Polydora varie-gata inhabiting scallops in Abashiri Bay, NorthJapan. Marine Biology 104: 61–66.
Simpson K, Knight B. 1985. Forensic Medicine. EdwardArnold Ltd: London.
Sorg M, Dearborn JH, Monahan EI, Ryan HF, Swee-ney KG, David E. 1997. Forensic taphonomy inmarine contexts. In Forensic Taphonomy, HaglundWD, Sorg M (eds). CRC Press: Boca Raton;567–599.
Stirland AJ. 2000. Raising the Dead. John Wiley andSons: New York.
Int. J. Osteoarchaeol. (2007)DOI: 10.1002/oa
Mary Rose Taphonomy
Symth MJ. 1988. Penetrantia clonoldes, sp. nov.(Bryozoa), a boring bryozoan in gastropod shellsfrom Guam. Biology Bulletin 174: 276–286.
Wisshak M, Gettidis M, Freiwald A, Lundalv T. 2005.Bioerosion along a bathymetric gradient in a cold-temperate setting (Kosterfjord, SW Sweden): anexperimental study. Facies 51: 93–117.
Copyright # 2007 John Wiley & Sons, Ltd.
Young HR, Nelson CS. 1985. Biodegradation of tem-perate-water skeletal carbonates by boring spongeson the Scott Shelf, British Columbia, Canada.Marine Geology 65: 33–45.
