Journal of Archaeological Science: Reports 37 (2021) 102918 Contents lists available at ScienceDirect Journal of Archaeological Science: Reports journal homepage: www.elsevier.com/locate/jasrep First evidence of vivianite in human bones from a third millennium BC Domus de Janas: Filigosa tomb 1, Macomer (NU), Sardinia Consuelo Rodriguez a, *, Luigi Sanciu b, Alfredo Idini c, Dario Fancello c, Clizia Murgia a, Ilenia Atzori d, Vittorio Mazzarello e, M. Eulalia Subirà a a Unitat d’Antropologia Biològica, Dpt BABVE, Facultat de Biociències, Universitat Autònoma de Barcelona, Spain b Centro Studi di Storia Naturale del Mediterraneo, Masullas, Oristano, Sardinia, Italy c Department of Chemical and Geological Sciences, University of Cagliari, Sardinia, Italy d Independent Researcher e Department of Biomedical Sciences, University of Sassari, Viale San Pietro 43b, Sassari, Sardinia, Italy A R T I C L E I N F O A B S T R A C T Keywords: Sardinia is an island located in the middle of the Mediterranean Sea. Due to its position geologically acquired in Vivianite the Middle Miocene (around 16 Ma ago), this island had a very strategic position in antiquity, as it was involved Sardinia into all the trade routes that crossed the Mediterranean Sea; however, it developed original archaeological Domus de Janas features up to the Iron Age. During the Final Neolithic, the Ozieri’s Culture developed throughout the Sardinian Chalcolithic Geoarchaeology territory, with diffusion of typically hypogea graves named Domus de Janas. The study was conducted inside the Bones Domus de Janas at Filigosa necropolis, located on a tuffaceous hillside near the village of Macomer in the area of Marghine, Central-Western Sardinia. Tomb 1 dating back to the beginning of the third millennium BC had been excavated by Professor E. Contu in 1965 At the time of first excavation, this tomb showed very particular conditions that had enabled an excellent conservation of several osteological and wooden samples. This paper focuses on the presence of vivianite deposits on human bones and its origin. For such a mineral to be formed an interaction between phosphate, iron and water has to occur. These findings can be considered the first evidence of such mineral in a Sardinian archaeological site, and one of the most ancient findings of this mineral in Italy, as well as the first evidence observed in prehistoric sites related to a period before the intro­ duction of iron use. 1. Introduction. with the iceman mummy Oetzi (Pabst and Hofer, 1998), who presented small amounts of this mineral into his lungs and on his skin. Vivianite is a hydrated iron phosphate mineral (chemically, Fe In Italy, other archaeological evidence about this mineral relates to (PO4)2⋅8H2O) which crystallizes in a monoclinic system. Colourless Bronze Age, Iron Age and Roman contexts; in two of these sites, the when unaltered, vivianite ranges from grey to indigo, green or dark blue Terramare Bronze Age site of Fondo Paviani, Verona, (Nicosia et al., when subjected to oxidation whereas, depending on its oxidation state, 2011) and the Roman harbor of Como (Ferrario et al., 2015), vivianite it ranges from a vitreous to a pearly lustre. was found in sedimentary stratigraphic soils, whereas in three Iron Age This mineral was first described in 1817 by Abraham Gottlob sites in Campania, Longola-Poggiomarino, Sarno river valley and Castel Werner, who named it ‘Vivianite’ after John Henry Vivian, the first Vetrano necropolis, (Balassone et al., 2009), it was detected in metal mineralogist to discover such specimen in Cornwall (UK) (Werner, artefacts. 1817). The ideal conditions for this mineral’s formation in archaeological Vivianite findings in archaeological contexts have been reported contexts require aqueous environment and the presence of phosphates since 1800, although only in 2006 McGowan and Prangnell reported on and iron ores. Therefore, archaeological sites usually presenting viv­ its relevance in archaeology (McGowan and Prangnell, 2006). ianite pertain to Ancient or Modern History, where human and/or faunal The oldest known evidence of vivianite in archaeology is associated remains in swampy environments are associated to iron goods: * Corresponding author. E-mail address: consuelorodriguez80@yahoo.it (C. Rodriguez). https://doi.org/10.1016/j.jasrep.2021.102918 Received 1 July 2019; Received in revised form 29 January 2021; Accepted 5 March 2021 Available online 21 March 2021 2352-409X/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). C. Rodriguez et al. Journal of Archaeological Science: Reports 37 (2021) 102918 formation of vivianite could not be formed unless such conditions are part of the island called Sulcis-Iglesiente and can be accurately dated to met. that era. The dating was possible due to the presence of relevant fossils This research represents the first archaeological evidence of vivianite such as Trilobites and Archaeocyatha in those predominantly arena­ in Sardinia found in a prehistoric grave cut (usually called Domus de ceous and carbonate rocks. Janas) dating from the Chalcolithic period, around the early third mil­ Since that period, the marine sedimentation had consistently settled lennium BC, located in the archaeological site of Filigosa. The relevance until the Middle Ordovician, when the sea levels dropped; upon the pre- of this research is the absence of ferrous objects among goods retrieved Variscan tectonic processes, which can be included in the so-called from Tomb 1: thus it might be considered the first evidence in prehis­ ‘Sardinian Phase’, those sediments that had previously settled were toric archaeological context. Therefore, as in the case of Oetzi, the iron deformed and uplifted resulting into the first ancient group of emerged ore should rather be identified with the bedrock composition of the continental layers heavily eroded by exogenous agents during the tomb. Cambrian-Ordovician series. Domus de Janas are prehistoric rock-cut graves found specifically in In the Early Ordovician, the sea spilling again over into the whole Sardinia, dating from the Recent Neolithic (also locally known as the San island resulted in the recovery of the sedimentation activity, predomi­ Ciriaco culture), namely around the fourth millennium BC (Tanda, nantly involving carbonate and clay-rich rocks that can be clearly 2009). The subsequent Ozieri culture, which developed during the Final identified in the Silurian and Devonian rocky outcrops. By the end of the Neolithic, represents the first Sardinian culture with a regional diffusion, Lower Carboniferous, a new phase of tectonic processes occurred in which sees homogeneous rituals and ceramic traditions from the north Sardinia, which grew in power and intensity in the Mid-Carboniferous: to the south of the Island (Melis, 2009). the thousands-metres thick successions from the Palaeozoic that had The most recent surveys reported about 3,500 Domus de Janas previously sedimented, underwent pressure and temperature increase, (Tanda, 2009), including both single tombs and necropolis, carved out bending and overturning movements, taking on a schistose structure and of different bedrocks and provided with either a single or multiple cell, a Barrovian metamorphism. By the last phases of the orogenesis thus presenting different sizes. In addition, these tombs might or may (320–315 Ma), massive magma bodies climbed up through the fractures not present sculptures and mural decorations, which often imitated the and faults of the deformed rocks and crystallized as granite at middle structure of ordinary houses. crustal levels, locally leading to the deformation of Variscan rocks and The Domus de Janas necropolis of Filigosa, located over a hillside triggering thermic metamorphism phenomena. near the village of Macomer, Central-Western Sardinia, in the sub-region At the end of the Mid-Carboniferous Sardinia’s emerged areas were called Marghine (Fig. 1) consists of 4 rock-cut graves. It represents the mountains, part of that Variscan chain extending across Central Europe first necropolis where a specific Eneolithic culture from 3264 to 2915 cal and the Western Mediterranean, to which the island was firmly con­ BC1, later indicated as Filigosa culture, was first acknowledged (See nected. Finally, the severe Permian erosion activity, at the end of the Figs. 2 and 3). Palaeozoic Era, completely levelled the Sardinian area. The site was discovered in 1960 following some illegal excavations that affected a small portion of the dromos in Tomb 1 and it was exca­ 2.1.2. The mesozoic era vated again by Contu in 1965 (Contu, 1965; Foschi, 1986). The end of the Variscan orogeny marked the beginning of a long At the moment of its discovery, the tomb presented with some water volcanic and tectonic stagnation in the Mesozoic Era (251–66 Ma ago). and a thin layer of mud, that had likely seeped through the tuffaceous During this Era, the sea levels raised again, although submerging Sar­ rock of the necropolis; these conditions, noted in the relevant excavation dinia only for short periods during the Jurassic Era. The rocks formed log and confirmed by its associated pictures (Fig. 4), allowed the ideal during this age, predominantly limestones, can be identified in large conditions for the conservation of any organic materials, including the outcrops located in the sub-region called Nurra, in the Supramonte, in renowned miniature wooden vase currently exhibited at the National Monte Albo, in the small island of Tavolara, and in the sub-regions called Archaeological Museum “G. A. Sanna” in Sassari. Ogliastra and Sarcidano. The specificity of the materials found in this tomb lies in the ideal conditions that allowed the conservation of different materials, such as 2.1.3. The cenozoic era human remains as well as biological material from plants and of ento­ In the Cenozoic Era (66 Ma ago to present day), Sardinia started to mological and mineral origin, which represent unique finding from get its own characteristic appearance, differentiating itself from the rest Sardinian prehistory to date. of Europe. Constant dislocations due to fractures and orogenic uplifts The focus of this paper is the analysis of skeletal remains found in related to the third of the major orogeny phases, called the Alpine Filigosa Tomb 1 presenting crystal-shaped encrustations, grey/black/ orogeny, split the Sardinian crust into different blocks. bluish in colour. In the Early Miocene, the block formed by modern Sardinia and Corsica broke away from Continental Europe and moved toward the 2. A geological perspective Western Mediterranean until it reached its current position (Middle Miocene), finally becoming an island. 2.1. Sardinian geology Sardinia’s drifting is marked by various volcanic activities, pre­ dominantly including andesitic lava which became increasingly massive 2.1.1. The Palaeozoic Era while amassing together with other marly and arenaceous sediments The geological history of Sardinia (Carmignani et al., 1992) started into the ‘Sardinian Rift’, namely a large depression which goes through at the beginning of the Palaeozoic Era (541–251 Ma ago), more specif­ the island, from the Gulf of Asinara to the Gulf of Cagliari. ically in the Cambrian series 2: the large majority of the rocks which Following a relative calm during the Mid-Upper Pliocene, in Sardinia sedimented when Sardinia was still submerged crops out in the Southern the Rift Campidano Valley started to lower and to take its current shape, while the volcanic activity was resumed especially on Monte Arci and Monte Ferru, followed by the abundant basaltic emissions from which 1 originated the large plateaus in the Central and Northern part of the Data obtained as a part of the project "Nutrition and society at the origins of island (i.e. Gesturi and Campeda). nuragic civilization" (CRP_661), funded by a Grant to Young Researchers, Autonomous Region of Sardinia, awarded to Dr. Luca Lai, presented during the During the Quaternary (1,8 Ma ago to present day) the erosion following study: Sample UniCa 143, individual 5, Oxford laboratory code accumulation activity operated by the sea, rivers and brooks, the wind Radiocarbon Accelerator Unit, Oxford University OxA-25337, ™13C‰ = and other exogenous agents have shaped the current conformation of the − 19.51; conventional years BP = 4401 ± 32. island, which had already been broadly set up during the Pliocene. 2 C. Rodriguez et al. Journal of Archaeological Science: Reports 37 (2021) 102918 Fig. 1. a) The position of Sardinia in the Mediterranean; b) the position of the Filigosa necropolis on the Island map. form making every trace of it disappear by sublimation. The lithic product that is generated by cooling is ignimbrite characterized by a typical bank geometry and recognizable in the countryside by the characteristic “flames” of light color generated by hot collapse of the pumice. Alternating with these paroxysmal phenomena there were other more “peaceful” ones characterized by the ejection activity of ashes and pumice. The final result is an alternation of reddish hard rocks, the ig­ nimbrites, and softer rocks, the tuffs, gray, gray greenish and pinkish. The tertiary volcanites, inclined towards the NW as already mentioned, occupy the whole top part of the chain and characterize the profile of the mountain with broken slopes due to the heads of the ignimbritic banks, alternating with gentler slopes due to the softer rocks. During the tec­ tonic movement, substantial flows of basaltic chemistry flowed from the primary and secondary faults. Some emission centers are present in the studied area and have been mapped, the outcrops of these old volcanoes are characterized by the presence of considerable quantities of boiling and very light slag. Fig. 2. Tomb 1, Filigosa (ph. Donna Nuragica). As already mentioned, the part lowered by the fault was subse­ quently filled with continental deposits. First of all, sediments of alluvial 2.2. Geology of Filigosa’s territory and fluvio-lacustrine origin were deposited, generated by the disman­ tling of the important chain just raised. Above these sediments, during The geology of the area can be distinguished in two fundamental the Plio - Quaternary, basaltic lavas were placed in expansions that gave elements of the landscape: the Abbasanta plateau and the Marghine rise to the Abbasanta plateau. The basaltic lavas have the characteristic chain. The first is an important basaltic plateau of Quaternary age, of being, at the time of placement, very fluid therefore they often filled constituted through mainly linear emission centers. The second is a paleo depressions leaving a tabular morphology at the top. Erosion then mountain range due to an important OSO - ENE tectonic dislocation that affects the edges of the expansion selectively, especially if the substrate determined the tilting towards the NW of the ignimbritic coulters of the is soft or even loose as in this case, thus implementing the inversion of Oligo-Miocene. From a geological and morphological point of view, the the relief: what was previously at the bottom of the valley becomes a region has a limited variability; the substrate includes exclusively vol­ summit area and vice versa. In the recent quaternary, the modelling canic lithologies, covered by sedimentary deposits due to the modelling phenomena of the slopes were completed, with the accumulation of of the slopes occurred in the Quaternary. In its reciprocal movement, the detrital deposits. The erosion, transport and sedimentation phenomena fault distinguished a lowered part, subsequently covered by continental were triggered by the newly raised volcano-tectonic reliefs. The so and volcanic deposits, and a raised part, which now forms the chain. The formed debris accumulated at the foot of the slopes, filled the concavity raised part was further tilted towards NO. In fact, the main peaks of the of the ground or accumulated in the valley bottoms. In general, these Marghine are facing SE, or in the diametrically opposite direction where sediments are classified in relation to the morphogenetic action that more steep slopes are found, while on the other side the chain slopes determined them, the extent and type of transport. more gently. The oldest part of the chain, not included in the municipal Into the detail of our study area, the lithologies observed are part of area, consists of a crystalline basement of intrusive and metamorphic the Volcanic District of Bonorva - Macomer Unit (OER). In the outcrop, rocks dating back to the Paleozoic era. In particular, granodiorites pyroclastic flow deposits can be distinguished in ignimbritic facies, whose placement is to be traced back to the Variscan orogeny are found, welded, with structures from vitroclastic to eutaxitic; in banks alter­ encased in a metamorphic complex consisting of schists, phyllites and nating with pyroclastic flow, fall and basal wave deposits. (K / Ar 21.6 ± marbles. Tertiary volcanics, due to highly explosive volcanic phenom­ 1.1 Ma (Aquitanian): Lecca et al., 1997). Pyroclastic flow, fall and surge ena are found above the Paleozoic rocks. Fiery clouds are made up of an levels (OERb). emulsion of gas, lava and solid parts that emerge from linear fractures at temperatures around 900 ◦ C with speeds of several tens of kilometers per hour, these paroxysmal activities are capable of destroying any life 3 C. Rodriguez et al. Journal of Archaeological Science: Reports 37 (2021) 102918 Fig. 3. Tomb 1 plan and section (Moravetti, 1998). Fig. 4. Tomb 1 - first excavation in progress: a) flooding of internal burial rooms, view from cell A; b) deposits in the tomb’s dromos, view from cell A; c) cell A Eastern wall’s deposits; d-e) cell A Eastern wall’s excavation in progress: water and mud; bones and vases still in situ. (note: On the concession of the Ministero per i beni e le attività culturali – Soprintendenza archeologia, belle arti e paesaggio per le provincie di Sassari e Nuoro. Any further reproduction or duplication is forbidden). 4 C. Rodriguez et al. Journal of Archaeological Science: Reports 37 (2021) 102918 3. Materials and methods analyses were performed through polarizing microscope. Moreover, some bone fragments have been analysed in low vacuum 3.1. Material with Fei Quanta 200 Scanning electron microscope (SEM) operating at 20 kV. The human remains recovered from the tomb consist on 1700 frag­ Mineralogical characterization was made through XRD powder ments of human bones, 103 of which were covered by a mineral diffraction analyses, after pulverizing small portion of the dark green/ encrustation (Fig. 5); seven of them were subject to further in-depth blue mineralization collected in contact on the bones. Experimental analysis (XRD, petrographic thin sections and SEM analysis). setting of the diffractometer PANalytical X’Pert Pro was 40 kV and 40 Samples were prepared through conventional laboratory procedures, mA, theta-theta geometry, radiation Cu-Kα1 (1.540598 Å), scan range including dry cleaning or with cold water without direct immersion, 3.484–70.998, step-size 0.167113, and 0.19 s per step. The software with soft bristle brushes and wood utensils, drying at air without heat used for mineralogical identification and analysis was X’Pert HighScore source (to avoid deformation of the sample). Reconstitution of the Plus version 2.1b (Degen et al. 2014). fragmented parts when possible using the tape “Scotch 3 M”, reversible with alcohol, or Paraloid B72 in 20% solution for the consolidation of 4. Results fragile pieces. Once clean, bones were catalogued and inventoried by a permanent The anthropological study (Rodriguez doctoral dissertation in pro­ marker on enamel base an acronym identifying each piece with the code cess) showed an NMI (White and Folkens, 1991) of 49 adults in Tomb I, of the reservoir (e.g. FIL by Filigosa), the number of the tomb, the calculated based on their most significant bones (complete preserved number of the box or the number or letter of identification of the cell (if right tibia, or at least complete for 2/3 proximal and distal extensions); and when we have it indicated) and they are entered in a register to and 33 for non-adults, as follows: facilitate subsequent reviews or studies. Since the graves corresponded to multiple burials, first a calculation • foetal: probably 1 individual to whom a scapular fragment may of the minimum number of individuals (NMI) was performed (Duday, belong. 2000, 2005). The first step to calculate the NMI is to group the bones by • Infant I (0–6 years old): 11; typology, distinguish them by laterality, define their sex (Acsádi and • infant II (7–12 years old): 10; Nemeskéri, 1970; Borrini, 2007; Buikstra and Ubelaker, 1994; Fer­ • 13–16 years old: 11 individuals. embach et al., 1977–79; Rodríguez Cuenca, 1994; Walker, 2008) and age range (Black and Scheuer, 1996; Brooks and Suchey, 1990; Broth­ A post-depositional taphonomic study was carried out on all in­ well, 1981; Campillo and Subirà, 2004; Gindhart, 1973; Langley-Shirley dividuals (Rodriguez doctoral dissertation in process). The analysis and Jantz, 2010; Lovejoy et al., 1985; Maresh, 1970; Scheuer and Black, showed that the taphonomic syndromes mainly observed on the exam­ 2000; Stloukal and Hanáková, 1978; Ubelaker, 1989). Then we pro­ ined skeletal material include a dark colour of the samples, both biotic ceeded following a criterion of “maximum parsimony” attempting to (of biological, plant and entomological origin) and abiotic encrustations eliminate any possibility of double counting of the same individual. The (e.g. minerals, carbonated deposits) such as post-mortem fractures, taphonomic characteristics have been highlighted, useful for interpret­ erosion and exfoliation. In addition, it should be noted that the weight of ing the post-depositional events of the bone remains (Borrini et al. each specimen is above average, in accordance with their long under­ 2011). water permanence. As previously remarked, water was constantly pre­ Vivianite encrustations were also observed over a wooden fragment sent in Tomb 1: the examined samples were found covered in mud, and recovered in the same site (Fig. 6). regardless the fact that the tomb is currently opened and empty, seepage still occurs and humidity exudes from the rocks, as evidenced by musk on the walls and on the floors in different burial rooms. 3.2. Mineralogical and petrographic methods Out of the 1,700 bone fragments of various sizes found on-site, about 6% (1 0 3) showed evidence of crystal presence. The excavation log To perform the petrographic observation, some bones fragments hand-written by the archaeologist during his inspection did not mention hosting submillimeter- to millimeter-sized dark green/blue crystals were such crystals probably because of the vivianite’s lack of visibility until its cut in thin slices, incorporated in epoxy resins and bound on a petro­ oxidation, or likely due to the presence of mud over the bone samples; graphic glass slide. Bone slices were then reduced up to 30 µm in unfortunately, it did not describe where the bones’ surface appeared, as thickness and polished with 1 µm alumina powder and the optical Fig. 5. a) Calcaneus vivianite encrustation; b) Vivianite encrustation on humeral shaft. 5 C. Rodriguez et al. Journal of Archaeological Science: Reports 37 (2021) 102918 Fig. 6. Various magnifications of SEM images showing a wooden fragment presenting vivianite encrustations. well as it did not specifically mention the place where the samples were characterized by concentric lamellae around the pores. collected or the different conditions presented by the whole archaeo­ Dark green/blue crystals develop within the pores commonly filling logical context. The bones examined showed some crystals only over a them and assuming their shapes (Fig. 7a, b) or more rarely preserving a portion of their surface, sometimes closely connected to the inner tra­ euhedral shape (Fig. 7c). In the sections perpendicular to the rhombic beculated bone tissue: this might suggest that on such portion of the prism (00 l) a perfect cleavage oriented parallel to the major diagonal is bone surface where the encrustations appeared might have occurred the observed. The same cleavage system can be observed parallel to the ideal conditions for a greater water saturation and the best anaerobic faces of the prism (hk0). A fracture system, perpendicular to the c-axis, is conditions for the formation of vivianite crystals. Furthermore, the rarely found. The most striking feature of these crystals is the strong presence of vivianite on these samples might also help in the taphonomic pleochroism that changes from pale yellow to green/brownish to cobalt interpretation of human remains since it can provide relevant informa­ blue (Fig. 7a, b). Some crystals are patchy-colored from brown to blue. tion about the methods used for deceased individuals deposition. At crossed polar interference colors are masked by the absorption of the mineral and extinction angle ranges between first and third. The patchy- colored zones observed at single polar can be detected also in crossed 4.1. Mineralogical characterization polars, manifested by faint bluish to pinkish third order colors (Fig. 7d). The above described features are compatible with vivianite (General Thin section observation shows the high porosity of the bones due to Formula: Fe2+3 (PO4)2 8⋅H2O) and/or metavivianite (General Formula: the decay of soft tissues. The pores are separated by osseous septa, Fe2+ Fe3+2 (PO4)2 (OH)2 6⋅H2O), two iron phosphates sometimes found yellowish to brownish, with irregular shapes and a structure Fig. 7. Photomicrograph of bones with vivianite/metavivianite crystals: a) panoramic image showing crystals that fill the pores in sharp contact with osseous trabeculae (single polar); b) the same portion of bone shown in a but rotated of about 45◦ , note the strong pleochroism (single polar); c) detail of euhedral crystal with perfect cleavage traces (single polar); d) complex interaction between vivianite and bone with some reaction border highlighted by with arrows (crossed polars). 6 C. Rodriguez et al. Journal of Archaeological Science: Reports 37 (2021) 102918 in human and animal bones. Diffractometric analyses showed a pattern that is compatible to the two above mentioned species, vivianite (ref. code ICSD 00–030-0662) and metavivianite (ref. code ICSD 00–026-1137). Both phases share the higher intensity peaks located at 13.0750◦ 2θ (rel. int. 100%), at 13.2160◦ 2θ (rel. int. 27.68%) and at 18.1546◦ 2θ (rel. int. 5.02%). However, the presence of both phases can be inferred by the position of minor peaks that can be attributed exclusively to one or to the other phase (Fig. 8). 5. Discussion During the anthropological study (Rodriguez doctoral dissertation in progress) evidence emerged that they demonstrate a collective burial, as claimed by the NMI, while there is no trace of scarification rituals, in fact the remains of skin adhering to the bones indicate a primary type burial. So, the vivianite’s concretions might help in the interpretation of this archaeological context. The identified mineral has a colour ranging from Fig. 9. Photomicrograph showing the relation between bones and vivianite/ white to blue, turning into ‘blue-green’ when exposed to air. It can be metavivianite crystals: in the area dotted by red circles the reactions of bones formed in archaeological contexts rich in iron, phosphate and water, but composed by hydroxylapatite and neo-formation of vivianite/metavivianite can poor in sulphides and oxygen, with a stable pH range between 5.5 and be recognized. 8.5 (McGowan and Prangnell, 2006); it can also be “vivianite formed after bone hydroxylapatite in a relatively acidic and reduced environ­ iron phosphate phases like vivianite and metavivianite. The presence of ment, controlled by complex relations among organic matter distribu­ both phases, detected by XRD analyses, is further supported by optical tion and degradation, water supply and circulation rate. Slag and iron observation; indeed, the patchy change of color is likely due to the flakes provided abundant Fe2, which was mobilized in acidic conditions” different oxidation state of iron. (Maritan and Mazzoli, 2004). The finding of vivianite or metavivianite in the archaeological In the case of the samples collected from the Filigosa necropolis, the context of Filigosa Tomb 1 confirms that all the necessary requirements contact between the vivianite/metavivianite and the bones is commonly for formation of this mineral were met: presence of water: namely a sharp and in some cases there is a border of reaction (Fig. 7d, Fig. 9), swamp or submerged ambience, either intermittently or continuously. suggesting that vivianite forms at the expense of the bones (McGowan In fact, at the time of its first excavation, the tomb contained a large and Prangnell, 2006). The red circle in Fig. 9 highlights the mineral­ amount of water and mud (Fig. 4); an environment with reduced oxygen ogical neo-formation of vivianite/metavivianite (general formula Fe2+ concentrations, preserving artefacts and/or human remains that would Fe3+2 (PO4)2 (OH)2 6⋅H2O), the reaction could be triggered due to the otherwise have decomposed in case of higher concentrations. The phosphate (PO4)3- release from the hydroxylapatite (general formula finding of the miniature wooden vase currently displayed at the National Ca5(PO4)3OH), that is the main inorganic constituent of the bones. Archaeological Museum “G. A. Sanna” in Sassari, as well as of other The mineralogical characterization clearly indicates the presence of materials of plant origin that are currently subjected to further studies, Fig. 8. XRD pattern of vivianite/metavivianite crystals. Note that the minor peaks can be assigned to one or to the other phase exclusively. 7 C. Rodriguez et al. Journal of Archaeological Science: Reports 37 (2021) 102918 are a clear evidence of the reduced oxygen concentration in this Vittorio Mazzarello: Writing - review & editing, Resources, Supervi­ archaeological context; sion. M. Eulalia Subirà: Methodology, Writing - review & editing, Su­ close proximity with phosphates: this includes human and/or faunal pervision, Project administration. remains, and leather. In the archaeological context examined herein, human remains were the largest finding (outnumbered only by ceramic Declaration of Competing Interest remains which, however, are not included in this study), whereas the tomb also presented a small number of animal remains (sheeps, goats The authors declare that they have no known competing financial and lagomorphs), as well as biological materials of human origin which interests or personal relationships that could have appeared to influence will be subjected to further studies; close proximity with iron ores: the work reported in this paper. namely, iron artefacts or materials directly originating from the enclosing rocks. In this case, the examined tomb pertains to the Filigosa culture, dating back to Copper Age, when iron was still an unknown References material, thus suggesting the iron enrichment occurred in accordance Acsádi, G., Nemeskéri, J., 1970. History of Human Life Span and Mortality. Akadémiai with the latter hypothesis (originating from the surrounding rocks); Kiadó, Budapest. Vivianite formation pattern is crucial; this is a in situ formation occurred Balassone, G., Boni, M., Di Maio, G., Villa, I.M., 2009. Characterization of metallic in close proximity with phosphates and iron ores or at the edge of artefacts from the Iron Age culture in Campania (Italy): a multi-analytical study. Periodico di Mineralogia 78 (1), 45–63. excavated facilities from which it might have been transferred as a so­ Black, Sue, Scheuer, Louise, 1996. Age changes in the clavicle: from the early neonatal lution containing phosphates and iron. In this case, the conversion of period to skeletal maturity. Int. J. Osteoarchaeol. 6 (5), 425–434. hydroxyapatite contained in the bones -major source of phosphate- Borrini, M., Mariani, P.P., Murgia, C., Rodriguez, C., Tumbarello, M.V. (2011). Contextual Taphonomy: Superficial Bone Alterations as Contextual Indicators. reacting with water inside the tomb and with the iron contained by the Journal of Biological Research, 1, vol. LXXXIV, 217–219. surrounding rocks enabled vivianite crystal formation. This might also Brooks, S., Suchey, J.M., 1990. Skeletal age determination based on the os pubis: A provide further information for the interpretation of depositional rituals, comparison of the Acsádi-Nemeskéri and Suchey-Brooks methods. Hum. Evol. 5 (3), 227–238. hopefully confirming also the anatomical connection of the available Brothwell, D.R. (1981). Digging up bones. Ed. Oxford University press. skeletal remains. Unfortunately, at the time of writing, this corrobora­ Buikstra, J.E., Ubelaker, D.H. (1994). Standards for data collection from human skeletal tion is not possible for Tomb 1 since this tomb was used by two different remains. Fayetteville, Arkansas Archaeological Survey Research Series, 44. Campillo, D., Subirà, M.E., 2004. Antropología Física para arqueólogos. Editorial Ariel, cultures in chronological succession (the Filigosa culture and the Barcelona. following Abealzu culture), condition which may have partially altered Carmignani, L., Carosi, R., Disperati, L., Funedda, A., Musumeci, G., Pasci, S., Pertusati, the previously existing burials either because of the post-depositional P.C. (1992) - Tertiary transpressional tectonics in NE Sardinia, Italy. In Carmignani movements caused by the mud penetrating the grave cuts; another L. & Sassi F. P. (Eds), Contributions to the Geology of Italy with special regard to the Paleozoic Basements. A volume dedicated to Tommaso Cocozza, IGCP No. 276, limitation in confirming such hypothesis is caused by the lack of NEWSLETTER, 5, Siena. pp. 83-96. appropriate documentation pertaining the tomb’s excavation (e.g.: Contu E. (1965). Filigosa, Macomer. Rivista di Scienze Preistoriche, XX, 377–379. plants, georeferenced pictures), which might have allowed a later Degen , T., Sadki, M., Bron, E., König, U., Nénert ,G. (2014). The HighScore suite.Powder Diffr 29(S2), S13–S18. https://doi.org/10.1017/S0885715614000840. interpretation of this archaeological context. However, there is no evi­ Duday, H. (2000). Antropología biológica “de campo”, tafonomía y arqueología de la dence of the presence of iron elements from Abealzu culture. muerte. In Malvido E., Pereira G., Tiesler V. (Eds.). (1997), El cuerpo humano y su tratamiento mortuorio. Mexico City: Colección Científica, Instituto Nacional de Antropología e Historia. 6. Conclusions Duday, H. (2005). Lezioni di archeotanatologia (archeologia funeraria e archeologia da campo), Roma: ed. Erma. For the vivianite formation interaction between phosphate, iron and Ferembach, D., Schwidetzky, I., Stloukal, M. (1977-79). Raccomandazioni per la determinazione dell’età e sesso sullo scheletro. Rivista di Antropologia, 60, 5-51. water is needed. An interaction clearly exists between phosphate con­ Ferrario, M.F., Brunamonte, F., Caccia, A., Livio, F., Martinelli, E., Mazzola, E., Michetti, tained in bones and water, but explaining the interaction with iron it is A.M., Terrana, S. (2015). Buried landscapes: geoarchaeology of the Roman harbor of more difficult at archaeological level. In any cases exist iron elements as Como (Italy). Alpine and Mediterranean Quaternary, 28 (2), 11–120. Foschi, A. (1986). La tomba I di Filigosa (Macomer - Nuoro). Alcune considerazioni sulla a goods. This aspect differentiated vivianite formation in this setting cultura di Abealzu-Filigosa nel contesto eneolitico della Sardegna. Nuoro: Editrice compared to other archaeological settings. Coop. Grafica Nuorese. For vivianite to be stable there should be low sulphide rates; this Gindhart, Patricia S., 1973. Growth standards for the tibia and radius in children aged would allow the conservation of metallic objects, which could have been one month through eighteen years. Am. J. Phys. Anthropol. 39 (1), 41–48. Langley-Shirley, N., Jantz, R.L., 2010. A bayesian approach to age estimation in modern corroded by the sulphuric acid produced by sulphate-reducing bacteria. americans from the clavicle*. J. Foren. Sci. 55, 571–583. https://doi.org/10.1111/ Specifically, this archaeological site did not include any metallic arte­ j.1556-4029.2010.01089.x. facts; however, the presence of specimens of plant origin, as the Lecca, L., Lonis, R., Luxoro, S., Melis, E., Secchi, F., Brotzu, P., 1997. Oligo-miocene volcanic sequences and rifting stages in sardinia: a review. Periodico di Mineral. 66, abovementioned wooden miniature vase and the other specimens 7–61. currently subjected to further studies, in case of higher tannate levels Lovejoy, C.O., Meindl, R.S., Pryzbeck, T.R., Mensforth, R.P., 1985. Chronological might have contributed to the sulphide reduction, thus enabling viv­ metamorphosis of the auricular surface of the ilium: a new method for the determination of age at death. Am. J. Phys. Anthropol. 68, 15–28. ianite formation. Maresh, M.M., 1970. Measurements from Roentgenograms. In: McCammon, R.W. (Ed.), Although the source of the phosphate seems to be clear from the Human Growth and Development. C.C. Thomas, Springfield, IL, pp. 157–200. petrographic evidences, the source of iron remains unclear and more Maritan, L., Mazzoli, C., 2004. Phosphates in archaeological finds: implications for environmental conditions of burial. Archaeometry 46 (4), 673–683. analysis and field observation are warranted to find, for instance, a McGowan, Glenys, Prangnell, Jonathan, 2006. The Significance of Vivianite in lithogenic source. Archaeological Settings. Geoarchaeology: An. Int. J. 21 (1), 93–111. Melis, M.G. (2009). L’Eneolitico antico, medio ed evoluto in Sardegna: dalla fine dell’Ozieri all’Abealzu, in Atti delle XLIV Riunione Scientifica IIPP, La preistoria CRediT authorship contribution statement della Sardegna, Cagliari-Barumini 2009, I, 81-95. Moravetti, A., 1998. Ricerche archeologiche nel Marghine - Planargia, Sardegna Consuelo Rodriguez: Conceptualization, Methodology, Validation, Archeologica, Studi e Monumenti 5. Carlo Delfino editore, Sassari. Nicosia, C., Balista, C., Cupitò, M., Ertani, A., Leonardi, G., Nardi, S., Vidale, M., 2011. Investigation, Resources, Data curation, Writing - original draft. Luigi Anthropogenic deposits from the Bronze Age site of Fondo Paviani (Verona, Italy): Sanciu: Writing - original draft, Validation, Investigation, Visualization. pedochemical and micropedological characteristics. Quat. Int. 243 (2), 280–292. Alfredo Idini: Writing - original draft, Investigation, Formal analysis, Pabst, M., Hofer, F., 1998. Deposits of different origin in the lungs of the 5,300-year-old Writing - review & editing. Dario Fancello: Writing - original draft, Tyrolean ice- man. Am. J. Phys. Anthropol. 107 (1), 1–12. Rodríguez Cuenca, J.V. (1994) Introducción a la antropología forense. Análisis e Investigation, Formal analysis, Writing - review & editing. Clizia Mur­ identificación de restos óseos humanos. gia: Methodology, Validation. Ilenia Atzori: Writing - original draft. Scheuer, L., Black, S., 2000. Developmental Juvenile Osteology. Academic Press, London. 8 C. Rodriguez et al. Journal of Archaeological Science: Reports 37 (2021) 102918 Stloukal, M., Hanáková, H., 1978. Die Länge der Längsknochen altslawischer Werner, A.G. (1817). Letztes Mineral-System, Freiberg/Wien. Belvökerun-gen, unter besonderer Berücksichtigung von wachstumsfragen. Homo White, T.D., Folkens, P.A., 1991. Human Osteology. Academic Press, San Francisco. 29, 53–69. Walker, Phillip L., 2008. Sexing skulls using discriminant function analysis of visually Tanda, G. (2009). Il Neolitico recente. In Atti delle XLIV Riunione Scientifica IIPP, La assessed traits. Am. J. Phys. Anthropol. 136 (1), 39–50. https://doi.org/10.1002/ preistoria della Sardegna, Cagliari-Barumini 2009, I, 59-61. ajpa.20776. Ubelaker, D.H., 1989. Human skeletal remains. Smithsonian Institution, Washington. 9