Archaeological and Anthropological Sciences https://doi.org/10.1007/s12520-018-0646-2 ORIGINAL PAPER Heart of darkness: an interdisciplinary investigation of the urban anthropic deposits of the Baptistery of Padua (Italy) Cristiano Nicosia 1 & Andrea Ertani 2 & Alvise Vianello 3,4 & Serenella Nardi 2 & Gian Pietro Brogiolo 1 & Alexandra Chavarria Arnau 1 & Francesca Becherini 5 Received: 3 February 2017 / Accepted: 17 April 2018 # Springer-Verlag GmbH Germany, part of Springer Nature 2018 Abstract Archeological excavations beside the Baptistery of the Dome of Padua (north-eastern Italy) unearthed anthropic deposits formed between the seventh- and tenth-century AD. These were analyzed using soil micromorphology, soil chemical analyses (especially aimed at the definition of organic matter properties and dynamics), and GC/MS analyses of fecal biomarkers, the latter corroborated by principal component analysis. This inter-disciplinary study allowed differentiating between units resulting from in situ accumulation of trampled domestic waste and other, more frequent, units derived from repeated dumping or backfilling episodes. Fast accumulation of organic-rich domestic waste, coupled with an incomplete evolution of organic molecules appears as a fundamental formation process of these anthropic deposits. The overall level of fecal contamination in the Padua Baptistery sediments proved to be very low or absent. Keywords Urban geoarchaeology . Urban deposits . Formation processes . Organic matter Introduction According to geoarchaeological literature, Banthropic deposits,^ Banthropic horizons,^ or Bcultural layers^ form where intense human occupation takes place for prolonged periods of time (see Butzer 1982, 2008, 2011; Moinerau 1970). This is especially the case of cities, towns, or villages, which can be more broadly defined as Burban settings^ (on the definition of Burban,^ see Staski 1982, p. 97; see also Smith 1989, 2008; Heimdahl 2005 and references therein). In many European cities, dark-colored deposits occur directly above the remains of Roman-age structures, often robbed, decayed, or partially destroyed (see Carver 1987; Brogiolo 2011). The term BDark Earth^ was used to address such deposits dating to the Late Roman–Early Medieval period (i.e., fifth- to eleventh-century AD), initially in the UK, and later on in Italy and France (Macphail et al. 2003; see also Nicosia et al. 2013; Nicosia and Devos 2014). An excavation in the historic center of Padua (Veneto region, northeastern Italy, Fig. 1) offered the chance for an inter-disciplinary study of anthropic deposits Electronic supplementary material The online version of this article (https://doi.org/10.1007/s12520-018-0646-2) contains supplementary material, which is available to authorized users. * Cristiano Nicosia cristiano.nicosia@unipd.it 1 Dipartimento dei Beni Culturali, Università di Padova, Padova, Italy 2 Dipartimento di Agronomia, Animali, Alimenti, Risorse Naturali e Ambiente (DAFNAE), Università di Padova, Padova, Italy 3 Section of Water and Environment, Department of Civil Engineering, Aalborg University, Aalborg, Denmark 4 Italian National Research Council-Institute for the Dynamics of Environmental Processes (CNR-IDPA), Padova, Italy 5 Italian National Research Council-Institute of the Atmospheric Sciences and Climate (CNR-ISAC), Padova, Italy * Andrea Ertani andrea.ertani@unipd.it Alvise Vianello av@civil.aau.dk; alvisevianello83@gmail.com Alexandra Chavarria Arnau chavarria@unipd.it Francesca Becherini f.becherini@isac.cnr.it Archaeol Anthropol Sci excavation of crucial importance for the history of Medieval Padua. The analyses described in this article have the main task of reconstructing human activities and the use of space in this part of the city between the seventh- and tenth-century AD. To arrive at reconstructing these aspects, we tried to face the Bheart of darkness,^ that is all traces—in particular those invisible to the naked eye or to optical microscopy—hidden in the dark of Dark Earth. Physical setting and archeological phasing Fig. 1 Top, position of Padua in northeast Italy: 1, Brenta river floodplain; 2, Adige river floodplain; 3, major watercourses; 4, main cities (Image based on Bondesan and Furlanetto 2012); 5, sixteenthcentury city walls; 6, active hydrography; and 7, buried or deviated hydrography that could fit with the characteristics of Dark Earth as reported in literature. The Padua deposits would fit in the Bnarrow use^ of the term, i.e., Dark Earths occurring above Roman remains and belonging to the Late Antique and Early Medieval period (see Macphail et al. 2003, p. 349). The narrow use of the term Dark Earth is opposed to a more generalized one indicating later deposits eventually found in cities with no Roman past (see Nicosia et al. 2013 and references therein) or even outside urban contexts (see Loveluck 2004). In the present study, soil micromorphology, soil chemical analyses, and gas chromatography/mass spectrometry (GC/ MS) analysis of fecal biomarkers were performed in order to corroborate the data gathered from stratigraphic excavations. These were carried out between 2011 and 2013 and allowed to highlight a complex stratigraphic sequence related to the Late Roman and Medieval evolution of a central area of the city of Padua. Their interest lays mainly in the scarce archeological information existing about the Early Medieval evolution of the city as well as about the origins of its main urban church. Moreover, the insistence of some researchers on placing the first cathedral of the city in its suburbs makes the results of this The account of the physical setting is based on Castiglioni et al. (1987), Ferrarese et al. (2006), and Mozzi et al. (2010). The city of Padua is located in the fine-grained distal sector of the Brenta river, a major Alpine watercourse that drains the western part of the Dolomites (Fig. 1). The Brenta floodplain sediments constitute a thick Holocene alluvial sequence, with Last Glacial Maximum sediments outcropping several kilometers north and west of the city. Another river, named Bacchiglione, crosses this portion of the plain, but its geomorphic activity is restricted to its present meander belt. Such river has been flowing through Padua within two large meanders— identified as belonging to a former Brenta course—since the Middle Ages. The Bacchiglione was then diverted artificially outside the city in the twentieth century. Presently, the Brenta follows a NW-SW path ca. 5 km north of Padua. The episcopal complex of Padova comprised in the tenthcentury AD three different sections: the cathedral and its baptistery, the residential palace of the bishop (castrum doioni), and a porticoed structure in the northern area named BChiostro dei Canonici.^ Excavations took place in the area comprised between the Romanic Baptistery and the Chiostro dei Canonici and showed a series of phases that can be summarized as follows (Fig. 2): 1. Construction of a late Roman (probably fourth-century AD) building with at least three different rooms paved with black and white mosaics; 2. Destruction of the building testified by traces of fire on the mosaic floors and large pieces of charred wood; 3. Demolition of the building. A huge quantity of architectonic material, roof elements, and a large number of marble fragments that can be linked to liturgical elements (including a piece of altar table) were recovered. These elements date to the sixth-century AD; 4. Reuse of the area for settlement purposes. This is shown by postholes and traces of buildings made with stones and earth material. Four graves (two infants and two adults dated between the eighth- and the tenth-century AD) can be linked to these houses. The area was intensively occupied, with a quick succession of dwelling surfaces exposed during the open area excavation, and traces of Archaeol Anthropol Sci Fig. 2 The excavated area with respect to the Baptistery, with the position of the three studied sequences frequent reconstruction of the buildings. Dark Earth formation begins in this phase; 5. Construction of a large building, parallel to the northern wall of the Romanic Baptistery, to which a number of infant burials were associated. This marked a new important transformation of the area which assumed again its Christian function, maybe as external area of the Early Medieval (tenth- to eleventh-century AD) Baptistery. In synthesis, the excavations exposed a section of the Late Roman episcopal complex, probably an area close to the main church or its baptistery. This area was affected by destructions at the beginning of the seventh-century AD when it became a marginal area used as a dump of domestic waste and successively as a settlement with insubstantial buildings. The area did not regain importance until the end of the Early Middle Ages (tenth- to eleventh-century AD) when it was occupied by a high-rank cemetery north of the cathedral’s baptistery. Materials and methods Three stratigraphic profiles, named sequences I, II, and III were sampled during the archeological excavations (see Table 1; Figs. 2, 3, 4, and 5) and analyzed for soil micromorphology, soil chemical analyses, and GC/MS analysis of fecal biomarkers, although due to budget constraints, it was not possible to produce the thin sections from sequence II. Samples for soil analyses and GC/MS were labeled with the sequence number first (in Roman number) followed by a progressive number from top to bottom of the sequence (e.g., I/1, I/2, I/3). Thin sections were numbered from 1 to 13. Table 1 Archaeol Anthropol Sci Table 1 List of samples for soil micromorphology, soil chemical, and GC/MS analyses shows the correspondences between samples for soil and GC/ MS analyses, thin sections, and stratigraphic unit numbers. Sequence Unit Thin section Samples for GC/MS and soil analyses I 401a 1 I/1 401b 2 3 4 I/2 I/3 I/4 Profile description was carried out according to the FAO/ WRB (FAO 1990) guidelines. 401c 5 I/5 Soil micromorphology 402 145 6 – I/6 II/1 147A – – II/2 II/3 – II/4 198 147B – II/5 II/6 – – II/7 II/8 Soil analyses 285 – 147C – 323 7 II/9 II/10 III/1 326 330 335 333 294 III/2 III/3 III/4 III/5 III/6 The pH was determined in water with a soil to water ratio of 1:2.5. Total carbonates (TC) content was determined by the calcimeter method and by gravimetric loss of CO2. The organic C and N contents were measured using an element analyzer (vario MACRO CNS, Hanau, Germany). Humic substances (HS) were extracted from air-dried samples using 0.1 M KOH (1:10, w/v) at 50 °C for 16 h in a N2 atmosphere, freed from the suspended material by centrifugation at 7000×g for 20 min and filtered through Whatman No. 42 filter paper. The extract was dialyzed against distilled water with a 14-kDa molecular weight cut-off Visking membrane (Medicell, UK), desalted with ion exchange Amberlite IR-120 (H+ form) (Stevenson 1994). The dialyzed solution was reduced in volume to about 50 mL and freeze dried. The HS molecular weight distribution was determined by gel filtration and chromatography (Nardi et al. 2007). The column calibration was based on previously assessed standard proteins (Kit MS-II, Serva, Heidelberg, Germany; Dell’Agnola and Ferrari 1971). The apparent molecular weight of the fractions was defined as follows: F1, > 100.000 Da; F2, 10.000– 100.000 Da; F3, < 10.000 Da. Field observations and profile description II III 321 346 8 9 10 11 12 13 14 III/7 – Undisturbed oriented blocks were collected for thin section analysis in sequences I and III only. Thin sections were prepared according to the methods of Benyarku and Stoops (2005) and described using the terminology and concepts of Stoops (2003). Gas chromatography/mass spectrometry analyses Fig. 3 Sequence I, with position of thin sections (rectangles) and samples for soil analyses and GC/MS (asterisks) Samples were subsampled from the internal part of the monoliths collected for soil micromorphology using sterile equipment. 5β-Stanols are commonly formed by the reductive action of enteric bacteria on stenol precursors that are either ingested or biosynthesized by mammals, as these compounds pass through the gut and are excreted with feces (Murtaugh and Bunch 1967; Hatcher and McGillivary 1979). Coprostanol has been established as the major 5β-stanol present in human feces, which led to its extensive use in medical (Barker et al. 1993; Midvedt and Midvedt 1993), pollution (Leeming et al. 1997), and archeological studies (Lin et al. Archaeol Anthropol Sci Fig. 4 Sequence II, with position of samples for soil analyses and GC/MS (asterisks) 1978; Knights et al. 1983; Pepe et al. 1989; Pepe and Dizabo 1990; Bethell et al. 1994; Simpson et al. 1998; Bull et al. 1999a, b, 2002). Some specific ratios using the single analyte’s abundance (ratio analyte (A)/internal standard (IS)) were calculated in order to assess potential fecal contamination and try to distinguish fecal inputs. The ratios employed are summarized in Table 2. compounds to trymethyl-silyl (TMSO) derivatives to obtain volatile non-polar compounds (suitable for GC/MS analysis) adding 50 μL of BSTFA/TMCS 99:1 and heating for 1 h at 70 °C (Bull et al. 2001, 2003). The samples were then injected in a GC/MS system for analysis. Sample preparation GC/MS analysis were carried out using an internal standard (IS; 5α-cholestane-6) to identify ten compounds: (1) 3βcholest-5-en-3-ol (cholesterol), (2) 5α-cholestan-3β-ol (5αcholestanol), (3) 24-ethyl-5α-cholestan-3β-ol (5αstigmastanol), (4) 3β-hydroxy-24-ethyl-5,22-cholestadiene (stigmasterol), (5) 5β-cholestan-3α-ol (epicoprostanol), (6) 5β-cholestan-3β-ol (coprostanol), (7) 24-ethylcholest-5-en3β-ol (β-sitosterol), (8) 24α-methyl-5-cholesten-3β-ol (campesterol), (9) 24-ethyl-5β-cholestan-3β-o (5βstigmastanol) l, and (10) 24-ethyl-5β-cholestan-3α-ol (epi5β-stigmastanol). Quantitative determination, performed recalculating concentrations by calibration curves obtained injecting standard solutions of the analytes (range 10– 2000 ng), was not performed on 5β-stigmastanol and epi5β-stigmastanol, as these commercial standards were not available. To include all the investigated compounds in the evaluation of fecal contamination, the ratio analyte/internal standard (A/IS) areas was used to calculate ratios, instead of absolute concentrations. GC/MS analysis was performed using an Agilent Technologies 6890 gas-chromatographer coupled with a Twenty-four samples (one of which a control sample of alluvial sands from the deep substrate for comparison, named C53) were analyzed. Samples consisted of ca. 2 g each (dry weight), previously sieved at 2000 μm. They were first dried in a lab stove, then soxhlet extracted for 20 h using a mixture of acetone/hexane 1:1 to obtain a total lipid extraction (Benfenati et al. 1994) after adding 5α-cholestane as internal standard (Pratt et al. 2008). All the extracts were analyzed to investigate the abundance of 5β-stanols, which are considered to be suitable diagnostic markers for fecal pollution in sediments and even in archeological samples (Bull et al. 2002). Normally, samples would be submitted to some other preparation steps, involving, i.e., sample workup/purification and separation, but due to the small amount available (around 2 g each sample) and to the absence of replicates, it was decided to use directly the TLE extract (Sistiaga et al. 2014). The TLE was reduced in volume by rotating evaporation, then transferred in a 1.5-mL GC vial and dried under a gentle flux of nitrogen. Samples were then derivatized converting target GC/MS analysis Archaeol Anthropol Sci Fig. 5 Sequence III, with position of thin sections (rectangles) and samples for soil analyses and GC/MS (asterisks) quadrupole mass analyzer Agilent 5973N (Agilent Technologies, Santa Clara, CA, USA). The GC/MS was equipped with a HP-5 ms non-polar capillary column (5% phenyl-methyl siloxane, 0.25 mm ID, 0.25 μm film thickness, 30 m length) and a split/splitless injector. The analyses (two duplicate injections per sample) were performed both in full SCAN and in SIM mode, to clearly identify the analytes in combination with retention time (SCAN) and to quantify analytes benefiting from the higher sensitivity of SIM. A list of the ions used to quantify the analytes is reported in Online resource 1. Samples were manually injected (2 μL each one) in split mode (split 1:1, injector temperature of 270 °C) using a 10-μL GC syringe (Agilent Technologies); the oven’s thermal program was set with the following parameters: 150 °C— 2 min isotherm, 15 °C/min to 230 °C, 5 °C/min to 250 °C, 1.5 °C/min to 300 °C, held for 3 min. The total run time was 46 min. Principal component analysis Principal Component Analysis (PCA; Wold et al. 1987) is a statistical exploratory technique useful to capture the relevant information and to visualize major trends and structure of data. PCA was applied to the markers used to assess the potential fecal contamination within the analyzed samples in order to extract information on the samples and identify Authors Ratio Index Ratio No. Threshold Interpretation Seq I Unit 401a 401b II 401c 402 145 147A 198 147B 285 147C III 323 326 330 335 333 294 321 Contr. – Sample I/1 I/2 I/3 I/4 I/5 I/6 II/1 II/2 II/3 II/4 II/5 II/6 II/7 II/8 II/9 II/10 III/1 III/2 III/3 III/4 III/5 III/6 III/7 C53 Grimalt and Albaigés Bull et al. (1999a, b) (1990)/Takada et al. (1994) Coprostanol/Cholesterol Coprostanol + epicoprostanol/coprostanol + epicoprostanol + 5α-cholestanol) 1 2 0.2 0.7 > 0.2 = fecal > 0.7 = fecal contamination contamination (Coprostanol/5β-stigmastanol) Coprostanol + Coprostanol/(coprostanol + epi-coprostanol)/(5β-stigmasta5β-stigmastanol))*100 nol + epi-5β-stigmastanol) Herb% = (73 – Y) × 2,86 3 4 1.5 ~ 0.25 = herbivore, ruminant; > 1.5 = human-derived sewage (human feces = 5.5); 0.25 < x < 1.5 = mixed contribution 5 > 73%; < 38% > 73% = human fecal pollution; < 38% = exclusive herbivore contribution; 38 < x < 73 = mixed contribution 6 – herbivore % contrib. 0.73 0.87 0.25 0.13 0.18 0.04 0.12 0.16 0.18 0.15 0.02 0.20 0.24 0.13 0.06 0.14 0.07 0.14 0.19 0.06 0.17 0.08 0.02 0.02 0.88 0.99 0.97 0.78 0.72 0.69 0.39 0.34 0.41 0.42 0.60 0.61 0.55 0.56 1.08 0.63 0.54 0.83 0.91 1.00 1.55 0.78 0.72 0.19 46.8 49.8 49.2 43.7 41.7 40.9 28.2 25.4 29.0 29.7 37.3 37.8 35.6 35.8 51.9 38.8 35.0 45.4 47.8 50.1 60.8 43.8 42.0 15.7 74,8 66,4 68,0 83,7 89,5 91,8 100 100 100 100 100 100 100 100 60,4 97,8 100,0 78,9 72,1 65,5 34,9 83,6 88,7 100 0.48 0.51 0.45 0.39 0.38 0.45 0.52 0.52 0.51 0.51 0.49 0.57 0.53 0.55 0.52 0.44 0.47 0.50 0.56 0.52 0.63 0.53 0.42 0.31 Evershed and Bethell (1996) Evershed and Bethell (1996) (expansion)/Bull et al. (2002) 1.10 1.19 1.24 1.05 1.14 0.92 0.53 0.48 0.53 0.54 0.76 0.80 0.77 0.74 1.32 0.82 0.76 0.95 1.18 1.31 2.06 1.17 0.93 0.23 Leeming et al. (1997) Archaeol Anthropol Sci Table 2 Markers used to calculate potential fecal contamination in analyzed samples and to distinguish potential sources with indicated fecal contamination ratios obtained for the three studied sequences (see also Online resource 4) Archaeol Anthropol Sci similarities. The calculated ratios were assembled in a single dataset, composed of 24 samples and 12 variables and pretreated with autoscaling (i.e., mean centering and unit variance scaling). PCA was then performed with PLS Toolbox for Matlab R2014b. Results Field observations and profile description All the three studied sequences include stratigraphic units with dark colors in the slightly moist state (very dark gray, dark gray, dark grayish brown). The texture is silt loam or silty clay loam throughout and large quantities of construction rubble (tiles, brick, marble fragments, mortar) are present. The detailed description of the three sequences is available in Online resource 2. Soil micromorphology The detailed description of each thin section can be found in Online resource 3. We begin by discussing the results of profile III as it shows peculiar characteristics worth mentioning first. the low-energy reaches of the Brenta floodplain, as shown also by their calcareous nature (crystallitic b-fabric—see also soil analyses in Table 3). However, these original alluvial sediments have been heavily reworked by human activities. Secondarily, we observe large (i.e., > 0.5 mm to millimeter sized) charcoal fragments, eggshell, fishbone, and bone fragments, occasionally with chromatic variations pointing to burning (see Stiner et al. 1995; Villagran et al. 2017). The important observation is that such components show random orientation and no traces of sorting, and that no fine layering can be discerned. These indicators point to dumping as the main depositional process, as for example deriving from deliberate backfilling, rake-out, or discard (see Matthews et al. 1997, p. 289; Miller et al. 2009, p. 31; Milek 2012, p. 132; Banerjea et al. 2015a, p. 95). The dumped material is of domestic origin. Microstructure, when present, derives from moderate bioturbation with channel and locally coarse monic basic microstructures types (the latter given by mineral excrements such as spheres and ellipsoids). Infillings with crescent fabric (i.e., Bbow-like features^ sensu Bullock et al. 1985) and the presence of earthworm granules (Canti 1998) also indicate bioturbation. Finely comminuted charcoal/charred vegetal matter and fine organic punctuations (not possible to distinguish between the three by optical microscopy alone) vary between 15 and 20% and 25–30% by visual estimate at ×200 magnification. Sequence III Sequence I Under the microscope units 321 and 333 are significantly different from the rest of this same sequence and also from those in sequence I. They are in fact finely layered and have coarser components with their major axis horizontally aligned. Moreover, they contain in situ Bsnapped^ and Bcrushed^ bone and fishbone fragments (Fig. 6). This indicator, together with the fine horizontal layering and the horizontal orientation of coarser components can be regarded as evidence for trampling and compaction of occupation deposits (see Matthews et al. 1997, pp. 289, 300; Miller et al. 2009, pp. 33, 35; Miller and Sievers 2012, pp. 3040, 3049; Banerjea et al. 2015a, pp. 95– 98). In these units, bioturbation is very weak and the microstructure is massive. They contain wood ash aggregates, and phytoliths are rather well attested, whereas they are rare in the rest of this sequence and in sequence I (see below). Above these trampled domestic occupation deposits, we find a series of units (335, 330, 326, 323, 294) which are very homogeneous from the micromorphological standpoint. They are all composed by calcareous silt loam/silty clay loam with large quantities of ceramic material (i.e., brick, tile, pottery fragments), mortar, and rock fragments (oolitic limestone, metamorphic and volcanic rocks). Chrysophycean stomatocysts, the resting stages of chrysophyte algae (Verleyen et al. 2017), are very well attested within all units (Fig. 7a). They suggest a local source for the sediments, i.e., This sequence shows textural and microstructural homogeneity with slight variations in the quantity of some anthropic components. The microstructure is predominantly biogenic (weakly developed channel microstructure) with earthworm granules also indicating bioturbation. Towards the top of the sequence (unit 401a), a subangular blocky microstructure is present. The lithology is the same described in sequence III. The groundmass is calcareous throughout with a crystallitic bfabric. Anthropic components present in all units are mortar fragments, ceramic material, bone (occasionally weathered or burnt—see Stiner et al. 1995; Villagran et al. 2017), charcoal, eggshell, and fishbone fragments. Carnivore-omnivore coprolite fragments (identification based on Goldberg and Macphail 2010) occur in unit 401b and 401c and are scarcely attested or absent in other units (Fig. 7b). Fragments of allochtonous earth material, most likely identifiable as earth-based construction material, are well attested in units 402 and in the top part of unit 401a (thin section 1, Fig. 7c, d). Such fragments are noncalcareous and silty clay textured. They do not show traces of the addition of vegetal temper or of exposure to fire. As described for the upper part of sequence III, no sorting or preferential orientation of coarser component is present, a characteristic pointing to dumping. Dusty/impure clay coatings and infillings (Fig. 7e), with unoriented clays and lack of Archaeol Anthropol Sci laminations or layering, are present in unit 402, at the base of unit 401c (thin section 5), in unit 401b (thin section 4), and in and 401a (thin section 1). These textural pedofeatures might confirm the dumping of loose material or might derive also from post-depositional modification (such as decay of organics or changes in redox conditions—see Banerjea et al. 2015a, 2015b). Vivianite crystal intergrowths are attested throughout the sequence, with the exception of unit 402 (Fig. 7d). This confirms periodic reduction and subsequent oxidation of this part of the sequence. Waterlogging seems to be a mandatory requirement for vivianite formation (see McGowan and Prangnell 2006), but the site is very well drained, and no redoximorphic feature was recorded in any thin section (Lindbo et al. 2010) or in the field. As vivianite was totally absent in the other two sequences, we conclude that its presence in sequence I could derive from very localized water saturation and weathering of organic material. Soil analyses The results are visible in Table 3. In sequences I and II, the soil reaction is always sub-alkaline with the values ranging from 7.6 to 7.8 (sequence I) and between 7.6 and 7.9 (sequence II). In these sequences, the total carbonate value ranges between 160.8 and 231.7 g/kg. In sequence III, the pH is higher, in the alkaline range with values between 8.3 and 8.5. The total carbonate value is between 153.7 and 253.0 g/kg. Sequence I can be split in two parts. The top three samples (I/1, I/2, I/3—unit 401a and top part of 401b) have a higher organic carbon content (above 1.6%) and a lower degree of humification (between ca. 15 and 30% HC/OC ratio) with respect to sample I/4 below. This latter sample marks a change, possibly deriving from a short period of stability, with humification of organic material exposed on the surface (peak value in HC/OC ratio, 82.3%, then decreasing towards the bottom in I/5 and I/6; also highlighted by the lowest C/N ratio in the sequence). This is corroborated by the highest value of Bmature^ F1 molecules (12.52%) in sample I/4 with respect to the rest of the sequence. Also, sequence II can be divided in two parts. The trends of organic C and N are irregular in the top part of the profile (interval II/1-II5), with very low values of humic carbon in the two topmost samples. A difference can be spotted in sample II/6, with a peak in organic C content (1.87%) descending with an almost linear trend downwards. This coincides with the lowest carbonate content (177.33 g/kg in II/6) suggesting slight leaching of carbonates, probably during a brief interval of stability of the surface. In II/6 also the amount of F3 molecules reaches a peak, suggesting weathering of organic matter. In sequence III, the organic carbon and total nitrogen values show an irregular trend with values ranging between 0.72 and 1.87%. The amount of humic carbon also shows an Fig. 6 Examples of in situ snapped or crushed fish bones, indicating trampling. Unit 333 (sequence III). PPL irregular trend with depth and a large variability, ranging between 0.139 and 0.614%. The molecular size distribution shows a predominance of the fraction with intermediate molecular weight (F2, between 100.000 and 10.000 Da) in all sequences. However, some samples deviate markedly from this trend (i.e., sample I/5 in Fig. 7 a Chrysophycean stomatocyst (arrow). Note finely comminuted organic material in the groundmass. Unit 335 (sequence III), PPL. b Fragment of carnivore-omnivore excrement. Unit 401b (sequence III), PPL. c, d Fragments of allochtonous silty clay-textured material, probably earth-based construction material. Unit 401b (sequence III). c PPL, d XPL. e Dusty clay coating. Unit 401b (sequence I), PPL. f Vivianite crystal intergrowth. Unit 401c (sequence I), PPL Archaeol Anthropol Sci Table 3 Results of soil analyses Sequence Unit I II 401a I/1 7.6 177.33 1.71 0.24 7.2 0.251 14.7 3.41 76.70 19.89 I/2 7.5 160.78 1.74 0.21 8.1 0.418 24.0 4.96 75.30 19.74 401b I/3 I/4 7.7 7.7 170.23 165.5 1.66 0.78 0.15 0.15 10.4 4.9 0.515 0.646 31.0 82.8 3.75 81.04 15.21 12.52 23.33 64.15 401c I/5 7.7 177.33 1.15 0.17 6.7 0.711 61.8 3.89 73.37 22.75 402 I/6 145 II/1 147A II/2 7.8 7.8 7.9 189.15 182.05 182.05 1.08 1.39 1.73 0.13 0.159 0.187 7.8 8.7 9.2 0.236 0.084 0.084 21.9 6.0 4.8 3.92 78.29 17.79 3.90 75.76 20.34 3.28 76.29 20.43 II/3 7.8 193.88 1.27 0.164 7.7 0.698 54.9 4.39 76.93 18.67 II/4 II/5 7.7 7.9 179.69 224.61 1.49 1.12 0.176 0.149 8.4 7.5 0.502 0.419 33.8 37.4 3.78 78.74 17.48 5.50 79.45 15.10 147B II/6 II/7 7.6 7.6 177.33 193.88 1.78 1.65 0.201 0.204 8.8 8.1 0.776 0.558 43.7 33.8 5.53 61.77 32.70 16.20 65.62 18.20 II/8 198 III HC/OC ratio Molecular size Sample pH (H2O) Total CaCO3 (g/kg) Organic C (OC) (%) N (%) C/N ratio Humic distribution (%) carbon (HC) (%) F1 F2 F3 7.7 186.78 1.61 0.181 8.9 0.698 43.4 8.80 68.93 22.27 285 II/9 147C II/10 323 III/1 8.0 7.9 7.8 222.25 231.71 191.51 1.21 1.03 1.96 0.101 12.0 0.124 8.3 0.156 12.6 0.447 0.837 0.363 36.8 81.0 18.5 7.65 73.58 18.77 6.79 69.63 23.58 1.96 67.28 30.77 326 330 III/2 III/3 8.5 8.4 153.68 167.87 0.72 1.88 0.13 0.197 5.5 9.5 0.614 0.537 85.4 28.6 3.39 65.98 30.63 4.02 65.93 30.05 335 333 294 III/4 III/5 III/6 321 III/7 8.3 8.5 8.5 8.4 8.4 241.16 241.16 238.8 205.7 252.98 0.82 1.28 1.88 0.99 1.13 0.114 7.3 0.09 14.2 0.147 12.8 0.101 9.8 0.106 10.6 0.558 0.580 0.139 0.169 0.364 67.7 45.2 7.4 17.0 32.2 1.54 3.61 7.72 3.48 2.56 sequence I and II/7 in sequence II), with predominant low molecular weight fraction (F3 < 10.000 Da) and a four- to fivefold increase with respect to the average values in the high molecular weight fraction (F1 > 100.000 Da). 70.27 70.41 76.64 64.90 73.67 28.19 25.98 15.64 31.62 23.77 peaks and those belonging to an actual sample (sample I/2 from sequence I). Sequence I Gas chromatography-mass spectrometry analyses The concentration of target compounds (excluded 24-ethyl5β-cholestan-3β-o (5β-stigmastanol) and 24-ethyl-5βcholestan-3α-ol (epi-5β-stigmastanol)) in analyzed samples can be seen in Online resource 4. The percentage of each compound, calculated using each ratio A/IS versus the total amount of the analyzed compounds, is shown in Online resource 5. The calculated ratios can be found in Table 2. In the studied sample set, the low values of 5β-stanols obtained do not allow to clearly establish a marked fecal contribution. Nevertheless, the calculated ratios proved to be useful to interpret and corroborate data from other analytical techniques. An explanatory chromatogram can be seen in Online resource 6 highlighting the match between the analytical standards Ratio 1 (see Table 2 for numbering of the different ratios) showed a potential fecal contribution for samples from units 401a (I/1and I/2) and 401b (I/3). Considering ratio 2, no sample exceeded the theoretical threshold of 0.7 that indicates fecal pollution. The highest values were measured again for unit 401a (samples I/1, 0.48 and I/2, 0.51). Considering ratio 3 from Evershed and Bethell (1996) and expanding it using the one of Bull et al. (2002) (ratio 4)—i.e., taking into account the possibility of conversion of coprostanol to epicoprostanol under anaerobic conditions—all samples were in the range between 0.25 and 1.5. The highest values were recorded for units 401a (sample I/1—0.88 (ratio 3) and 1.10 (ratio 4); sample I/2, 0.99 (ratio 3) and 1.19 (ratio 4)) and 401b (sample I/3, 0.97 (ratio 3) and 1.24 (ratio 4)). These higher values in ratios 1, 2, 3, and 4 are coherent with the subdivision of this sequence in Archaeol Anthropol Sci two parts highlighted by physico-chemical analyses (see BSoil analyses^ above). Ratio 5 showed that all the samples in sequence I exceeded the threshold of 38%, with values between 40.8% (unit 401c, sample I/5) and 49.8% (unit 401a, sample I/2), a circumstance that can tentatively be interpreted as an indicator of mixed contributions (see Bull et al. 2002). Samples I/1, I/2, and I/3 were characterized by the lowest values of the so-called Bherbivore relative contribution^ according to ratio 6 (74.8, 66.4, and 68%, respectively). For the other samples of the same sequence, the values for this proxy increased with depth, reaching the highest value in sample I/6 (unit 402, 91.9%). Sequence II According to ratio 1, only two samples from unit 147B (II/6, II/7) show a weak potential fecal contribution. No samples exceed the theoretical threshold of 0.7, indicating fecal contribution according to ratio 2. The highest values of this ratio in sequence II were obtained in unit 147B (samples II/6, II/7, and II/8, respectively, 0.57, 0.53, and 0.55). This is in line with the subdivision of the sequence based upon physico-chemical analyses, with two main episodes of sedimentation and with sample II/6 (unit 147B) marking a difference. Proxies 3 and 4 show quite low values for this sequence, in the range between 0.25 and 1.5, with the highest value for unit 285 (sample II/9, 1.08 (ratio 3) and 1.32 (ratio 4)). Considering ratio 5, only samples II/9 and II/10 exceeded the 38% threshold, with values of 51.9 and 38.8%. The fecal contribution appears to be exclusively of herbivore origin according to ratio 6, with the exception of samples II/9 (unit 285) and II/10 (unit 147C). Nevertheless, given the overall low quantities of target compounds, ratios 5 and 6 can only be used tentatively. Sequence III No sample from this sequence showed fecal contribution considering ratio 1 or 3, albeit units 333, 335, and 330 (samples III/5, III/4, and III/3) show the highest values of the latter (0.63, 0.52, and 0.56, respectively—the theoretical threshold being 0.7). These values are the highest measured values among the entire sample batch. Considering ratios 3 and 4, unit 333 (sample III/5) is characterized by values of 1.55 and 2.06, respectively, the highest ones between all the analyzed samples and both above the threshold (1.5), indicating humansourced pollution (Bull et al. 2002). According to ratio 5, all samples exceed the value of 38%, except unit 323 (sample III/ 1, 35%), with all the samples showing mixed contributions. The lowest value in unit 333 (sample III/5, 34.9%), shows the highest value for Bother contributions^ (65%) of all analyzed samples. Overall, in sequence III, ratio 3 indicates a potential fecal contribution in the bottom part of the sequence, Interestingly, values close to fecal contribution according to ratio 2 and above the threshold (1.5) for ratio 4, together with mixed contribution highlighted by the other two ratios (5, 6) occur in unit 333, formed within a household. Carnivoreomnivore excrement fragments have been observed in thin sections from such unit and are also scattered—albeit in low quantities—in units above. Principal component analysis The total variance accounted by the first two principal components (PCs) was around 75%. In the score plot on the left side of Fig. 8, two groups of samples could be identified, related respectively to the sequences II and III, while samples belonging to sequence I were quite dispersed. Exceptions to this were samples C53 (reference sample of natural alluvial sands from the deep substrate), III/5 (unit 333 in sequence III), as well as I/1 and I/2 (unit 401a in sequence I). The loading plot on the right side of Fig. 8 put in evidence the relationship between the variables, which could be grouped as follows: the first group included A–B–G and O–P which had positive loadings for both the first two PCs; the second group was formed by F–I and L–M–N, characterized by positive loadings for PC1 and negative loadings for PC2; the third group was represented by H with negative and positive loadings respectively for PC1 and PC2. Variable D was outside the aforementioned groups and it showed significant (negative) loading only for PC1. Samples from unit 401a in sequence I (I/1 and I/2) resulted isolated from the rest of the samples of the same Sequence, showing the highest positive values of the variables A–B–G and O–P of the whole dataset. This result indicated mixed contributions within these two samples: omnivore feces (relative high concentration of coprostanol/cholesterol and absolute amount of coprostanol), degraded organic material (high values of 5αcholestanol/cholesterol), and potential organic material coming from herbivore fecal inputs (high values of the markers linked to the variables O–P). Most of the samples belonging to sequence II showed negative scores for PC1 and positive scores for PC2, as they were characterized by an exclusive herbivore contribution. Instead, sample II/9 (sequence II, unit 285) was characterized by score values more similar to the ones of samples belonging to sequence III, probably due to its mixed contribution (herbivore, 60.4%; other, 39.6%) and its fecal contamination (I = 0.51). According to their score values, samples related to Sequence III showed mixed contributions, the herbivore one increasing with decreasing score values for PC1. In fact, sample III/1 (unit 323) was 100% herbivore (H), while sample III/ 5 (unit 333) was characterized by the lowest value of H and highest fecal contamination (I = 0.63) within the dataset, albeit possibly of post-depositional origin. Archaeol Anthropol Sci The reference sample C53 showed the highest values of H and D, while the lowest values of all the other variables, indicating no evidence of fecal contamination. Discussion Some characteristics are common for all units. The silty clay loam/silt loam texture, the presence of chrysophycean stomatocysts, and the carbonate content are coherent with sediments of the Brenta/Bacchiglione floodplain. These, however, have been heavily transformed and reworked by humans, and no natural input of alluvial sediments—a common occurrence in many urban sites located in alluvial settings (see Nicosia et al. 2012; Cremaschi and Nicosia 2010)— was observed. A general discourse can also be made for the organic status of the studied units. The quality of HS in the three studied sequences is shown by gel filtration analysis, which reveals their molecular weight distribution. The latter is regulated by the amount of organic inputs and by the balance between mineralization and humification processes. The mineralization process, through microbial activity, decomposes substances and transforms them into simple inorganic substances. On the contrary, the humification process consists in the formation of high molecular weight substances, by means of resynthesis and neogenesis processes, starting from the substances that make up the plant and animal residues. The rate of degradation of the organic starting substances (mineralization process), and its evolution in humic substances (humification) is influenced by the environmental conditions (temperature, presence of water and oxygen) and the recalcitrance of the starting material. The low percentage (around 10%) of HS fraction F1 (> 100 kDa), measured in the majority of samples, suggests a scarce quantity of high-level macromolecules, like alkylic and aromatic compounds, that can be easily accumulated and included in the humic carbon structure (Zech et al. 1997), e.g., through hydrophobic interaction (Piccolo 2009). The intermediate fraction (F2, between 100,000 and 10,000 Da) is the most abundant and represents humic matter that accumulated but has not yet undergone the polycondensation process. The lowest fraction (F3 < 10.000 Da), typical of compounds that have only undergone the hydrolysis process (Dell’Agnola and Nardi 1979), is around 20%, a further demonstration of the difficult transformation of organic substances into HS. In light of these considerations, the pedoclimatic conditions impeded the polymerization and polycondensation of HS and promoted the accumulation of HS of medium molecular weight. This characteristic is due to a scarce oxygenation that prevents the final equilibrium of humic products. Generally, at the end of the first phase, the low molecular weight HS should assemble as associations of self-heterogeneous and relatively small molecules and stabilized predominantly by weak dispersive forces (Nardi et al. 1991). On the contrary, in our samples the self-association of small molecules is low (see values of F1 in Table 3). Field observations and soil micromorphology provided the necessary contextual data (sensu Goldberg and Berna 2010) to interpret the results of soil chemical and GC/MS analyses. It was possible to distinguish two main types of anthropic deposits: 1) In situ, trampled, domestic occupation deposits (base of sequence III, units 333 and 321). These do not show significant differences in soil chemical parameters with respect to units deriving from simple dumping (see below). This is not surprising as most likely occupation deposits like these are the parent material of dumped units. Interestingly, these household deposits are the most polluted by fecal material according to ratios 2, 3, and 4 (Table 2) and exhibit a mixed contribution (even if ratios 5 and 6 in Table 2 must be used with caution). This probably derives from a promiscuous use of the living space, with humans and animals co-existing under the same roof and in extremely poor hygienic conditions. This is in line with the picture of post-roman urban dwelling taking places in insubstantial huts or tuguria, often found in written sources and well documented archeologically (Brogiolo 2011). 2) Units deriving from dumping of domestic waste (all remaining units). The overall irregular variation with depth of soil chemical parameters (see Table 3) confirms several episodes of dumping of domestic waste, with two main phases of accretion observed in both sequences I and II. The irregular distribution with depth concerns also the concentration of fecal biomarkers. In sequence I, unit 401a resulted the only one showing fecal contribution when ratio 1 is used (Table 2). However, as stated by Bull et al. (2002), this proxy should be used with caution because it may be affected by differential losses of cholesterol due to either preferential utilization and/or degradation in aerobic environments or reduction under anaerobic conditions (Wardroper et al. 1978; Quirk et al. 1980; Bull et al. 2000). Unit 401a shows also the highest value according to ratio 2 (Bull et al. 1999a, b—although constantly below the theoretical threshold of 0.7) in sequence I. This could be explained by the presence of several carnivore/omnivore excrement fragments, observed in thin section. In sequence II, based on field observations, on the presence of mixed archeological materials dating from the first- to the tenth-century AD, and on the irregular trend with depth of soil chemical parameters, we can again infer several dumping episodes. As mentioned above, these can be grouped in two Archaeol Anthropol Sci Fig. 8 Score (left) and loading (right) plots of PC2 versus PC1. In the score plot samples from the three sequences are differently colored: I, blue; II, red; III, green. Key (variables from fecal markers): A, coprostanol/total analyzed sterols (%); B, coprostanol/cholesterol; G, 5α-cholestanol/cholesterol; O, (5β-stigmastanol)/(5β-stigmastanol + 5α-stigmastanol); P, (5β-stigmastanol + Epi-5β-stigmastanol)/(5βstigmastanol + epi-5β-stigmastanol + 5α-stigmastanol); F, (coprostanol/ (5α-cholestanol + coprostanol); I, (coprostanol + epicoprostanol)/(5αcholestanol + coprostanol + epicoprostanol); L, coprostanol/5βstigmastanol; M, (coprostanol + epicoprostanol)/(5β-stigmastanol + epi5β-stigmastanol); N, (coprostanol/(coprostanol + 5βstigmastanol))*100); H, herbivore relative contribution (%); D, epicoprostanol/coprostanol main periods of dumping, separated by a brief period of stability occurring on top of unit 147B. Fecal biomarkers appear to confirm this bipartition of the profile, with higher concentrations in unit 147B and in those below when ratios 3 and 4 are employed. PCA applied to fecal markers calculated from GC/MS data was useful as a clustering technique in order to identify homogenous groups of samples within the dataset (Kaufman and Rousseeuw 2005). This clustering reveals that each sequence has Ba story of its own^ (this is particularly true for sequences II and III, while sequence I is less neatly clustered, Fig. 8). From the geoarchaeological standpoint, this means that notwithstanding heavy human reworking (digging, quarrying, dumping, etc.) and bioturbation, homogenization was not complete across the site. throughout this paper to Bdumping^ for lack of a more precisely definable human action. It must be stressed that in Roman cities in north Italy and in the rest of the empire, waste disposal was strictly regulated (see Furlan 2017 for a review of the topic; see also Gelichi 2000). Waste included both Bsolid^ domestic and industrial rubbish as well as liquid and solid fecal residues, the latter disposed of by an efficient sewer system. The overall low levels of fecal contaminations recorded by GC/MS at the Padua Baptistery permitted to exclude a link between Dark Earth formation and the collapse of the roman sewer, a hypothesis postulated for some north Italian cities in the 1980s (see La Rocca 1986; Brogiolo et al. 1988; Brogiolo 2011). It is rather the accumulation of solid rubbish (i.e., artifacts, ecofacts, and their encasing sedimentary matrix) inside the city and not anymore in dumps extra muros that is at the base of the formation of these anthropic deposits and, consequently, of many Dark Earths. In such a framework, these form of anthropic accretion embody the tangible outcome of the change in the life ways of cities after the collapse of the Roman regulatory system, a phenomenon well known by historians (see, among others, Galinié 2004 and Brogiolo 2011). In being basically accumulations of waste, Dark Earth and any other type of anthropic deposit of any age bears no specific differences from (as stated also by Carver 1987). When not marking a change in the style of sedimentation with respect to an underlying Roman Bpast^—that is, outside the narrow use of the term (see Chapter 1)—Dark Earth Conclusion Archeological implications The interdisciplinary analyses at the Baptistery of Padua allowed for the identification of a series of in situ deposits formed in the seventh-century AD as the result of human occupation with markedly Brural^ characteristics. These took place in the frame of the re-use of the area for settlement purposes, after the destruction of the fourth-century AD episcopal building. The remaining units have been interpreted as deriving from the repeated dumping of household waste. We referred Archaeol Anthropol Sci appears to be even less differentiable from any other anthropogenic accretion deposit. The dumping that was observed at the Baptistery of Padua can be the progressive accumulation of discard from living spaces, as for example in sequence III (interval between units 335 and 323). Here, the slow buildup of waste is linked to seventh- to eighth-century AD reuse of the area for settlement purposes, documented during open area archeological excavations. Dumping can also be faster, as for example when backfill is accumulated for ground-raising purposes or to fill a negative structure. This is most likely the case for sequence I (chronology in course of determination) and sequence II (ninth- to tenth-century AD), respectively. In both sequences, two main episodes of dumping have been observed, based on soil and geochemical properties. The presence of materials of different ages mixed together proved to be the most robust indicator to distinguish massive backfill episodes from slower, gradual accretion. Implications for anthropic deposits or Dark Earth formation Thin section analysis proved to be an effective method to differentiate between slowly accreting, trampled in situ deposits and units deriving from dumping. At the Padua baptistery, where Dark Earth formed in response to the fast accumulation of organic-rich domestic waste coupled with a slow rate of evolution of organic to humic matter. The latter phenomenon is demonstrated by the low degree of maturity of humic substances in the vast majority of samples. The impeded humification of organic carbon might be linked to the occurrence of reducing conditions. These are suggested by high values of 5α-cholestan-3β-ol (5αcholestanol) in some samples and by the presence of poorly oriented dusty clay pedofeatures (deriving from oxidation of organic matter and changes in redox conditions according to Banerjea et al. 2015a, b) in others. Reduction seems here primarily related to the decay of organic matter, and not so much to waterlogging. Beside the lack of redoximorphic features, it must be also stressed that the site occurs on a topographic high, occurring 5–6 m higher than the surrounding alluvial plain (formed due to the repeated superimposition of archeological strata—see the high-resolution digital elevation models of Ferrarese et al. 2006; Ninfo et al. 2011). The reducing conditions stressed the microbial community shifting the populations from oxidative to reductive. Anoxic microorganisms possess a great potential to drive the dynamic of humic matter in the soil (Frankerberger Jr and Arshad 1995). The role of the impeded evolution of organic substances in the formation processes of Dark Earth, and possibly in dark-colored anthropic accretion deposits of any age (see Moinerau 1970; Nicosia et al. 2011), deserves further study. Future research could also focus on the identification and quantification through GC/MS of target compounds (such as lignin and waxes) that could confirm the hypothesis, often reported in literature, that Dark Earth derives from the decay of wooden elements in buildings in perishable material (see Macphail 1994; Yule 1990; Courty et al. 1989, p. 268; Macphail et al. 2003; Goldberg and Macphail 2010, pp. 271–273). Acknowledgements As required by the Italian system of research evaluation, we indicate below the relative contributions of each author to the different parts of the article. Author contribution C.N. wrote paragraphs “Introduction,” “Materials and methods,” “Field observations and profile description,” “Soil micromorphology,” “Results,” “Field observations and profile description,” and “Conclusion”. C.N., A.C.A., and G.P.B. co-wrote paragraph “Physical setting and archaeological phasing.” A.E. and S.N. wrote paragraphs “Soil analyses (under “Materials and methods”) and “Soil analyses.” (under “Results”). A.V. wrote “Gas chromatography/mass spectrometry analyses,” “Sample preparation,” (under “Materials and methods”) and “Gas chromatography-mass spectrometry analyses” (under “Results.”) F.B. wrote paragraph “Principal component analysis” (under “Materials and Methods” and sub-paragraph “Principal component analysis” in “Gas chromatography-mass spectrometry analyses.” All authors contributed equally to “Discussion.” References Banerjea RY, Bell M, Matthews W, Brown A (2015a) Applications of micromorphology to understanding activity areas and site formation processes in experimental hut floors. Archaeol Anthropol Sci 7:89–112 Banerjea RY, Fulford M, Bell M, Clarke A, Matthews W (2015b) Using experimental archaeology and micromorphology to reconstruct timber-framed buildings from Roman Silchester: a new approach. Antiquity 89(347):1174–1188 Barker GM, Radley S, Davis A et al (1993) Analysis of faecal neutral sterols in patients with familial adenomatous polyposis by gas chromatographymass spectrometry. Int J Colorectal Dis 8(4):188–192 Benfenati E, Cools E, Fattore E, Fanelli R (1994) A GC-MS method for the analysis of fecal and plant sterols in sediment samples. Chemosphere 29(7):1393–1405 Benyarku CA, Stoops G (2005) Guidelines for preparation of rock and soil thin sections and polished sections. Quaderns DMACS 33. Universidad de Lleida, Lleida Bethell PH, Goad LJ, Evershed RP, Ottaway J (1994) The study of molecular markers of human activity: the use of coprostanol in the soil as an indicator of human faecal material. J Archeol Sci 21:619–632 Bondesan A, Furlanetto P (2012) Artificial fluvial diversions in the mainland of the Lagoon of Venice during the 16th and 17th centuries inferred by historical cartography analysis. Geomorphol Relief Proc Environ 18(2):175–200 Brogiolo GP (2011) Le origini della città medievale. Post-ClassicalArchaeologies studies 1. SAP, Mantova Brogiolo GP, Cremaschi M, Gelichi S (1988) Processi di stratificazione in centri urbani (dalla stratificazione naturale alla stratificazione archeologica). In: di VV AA. (eds.) Archeologia stratigrafica in Italia Settentrionale 1, Atti del Convegno (Brescia, 1 Marzo 1986). New Press, Como, pp. 23–30 Bull ID, Betancourt PP, Evershed RP (1999a) Chemical evidence for a structured agricultural manuring regime on the island of Pseira, Crete during the Minoan period. Aegaeum 20:69–74 Bull ID, Simpson IA, Dockrill SJ, Evershed RP (1999b) Organic geochemical evidence for the origin of ancient anthropogenic soil deposits at Tofts Ness, Sanday, Orkney. Org Geochem 30:535–556. https://doi.org/10.1016/S0146-6380(99)00020-0 Bull ID, Van Bergen PF, Nott CJ et al (2000) Organic geochemical studies of soils from the Rothamsted classical experiments DV. The fate of lipids in different long-term experiments. Org Geochemistry 31:1367–1376 Archaeol Anthropol Sci Bull ID, Betancourt PP, Evershed RP (2001) An organic geochemical investigation of the practice of manuring at a Minoan site on Pseira Island, Crete. Geoarchaeology 16:223–242 Bull ID, Lockheart MJ, Elhmmali MM, Roberts DJ, Evershed RP (2002) The origin of faeces by means of biomarker detection. Environ Int 27:647–654 Bull ID, Elhmmali MM, Roberts DJ, Evershed RP (2003) The application of steroidal biomarkers to track the abandonment of a Roman wastewater course at the Agora (Athens, Greece). Archaeometry 45:149– 161. https://doi.org/10.1111/1475-4754.00101 Bullock P, Fedoroff N, Jongerius A et al (1985) Handbook for soil thin section description. Waine Research Publications, Wolverhampton Butzer KW (1982) Archaeology as human ecology. Cambridge University Press, Cambridge Butzer KW (2008) Challenges for cross-disciplinary archaeology: the intersection between environmental history and geomorphology. Geomorphology 101:402–411 Butzer KW (2011) Geoarchaeology, climate change, and sustainability: a Mediterranean perspective. In: Brown AG, Basell, L., Butzer KW (eds) Geoarchaeology, climate change, and sustainability. GSA Special Papers; Vol. 476. Geological Society of America, pp 1–14 Canti MG (1998) Origin of calcium carbonate granules found in buried soils and quaternary deposits. Boreas 27:275–288 Carver MOH (1987) The nature of urban deposits. In: Schofield J, Lynch R (eds) Urban archaeology in Britain. CBA, London, pp 9–26 Castiglioni GB, Girardi A, Rodolfi G (1987) Le tracce degli antichi percorsi del Brenta per Montà e Arcella nei pressi di Padova: studio geomorfologico. Memor Sci Geol XXXIX:129–149 Cremaschi M, Nicosia C (2010) Corso Porta Reno, Ferrara (Northern Italy): a study in the formation processes of urban deposits. Il Quaternario-Italian. J Quat Sci 23(2Bis):395–408 C o u r t y M A , G o l d b e rg P, M a c p h a i l R I ( 1 9 8 9 ) S o i l s a n d Micromorphology in Archaeology. Cambridge University Press Dell’Agnola G, Ferrari G (1971) Molecular sizes and functional groups of humic substances extracted by 0.1 M pyrophosphate from soil. J Soil Sci 22:342–349 Dell’Agnola G, Nardi S (1979) Umificazione controllata di rifiuti solidi industriali e valutazione dei caratteri chimici ed agronomici del materiale umico prodotto. Agric Ital 108:211–222 Evershed RP, Bethell PH (1996) Application of Multimolecular Biomarker Techniques to the Identification of Fecal Material in Archaeological Soils and Sediments. In Orna MV (ed) Archaeological Chemistry. ACS Symposium Series 625. American Chemical Society, Washington, pp 157–172 FAO (1990) Guidelines for soil description. Food and Agriculture Organisation, Rome Ferrarese F, Mozzi P, Cervo B, Veronese F (2006) High resolution DTM for the geomorphological and geoarchaeological analysis of the city of Padua (Italy). BAR International Series 1568. Oxford, pp 291–296 Frankenberger WT Jr, Arshad M (1995) Phytohormones in soil: microbial production and function. Marcel Dekker, New York Furlan G (2017) When absence means things are going well: waste disposal in roman towns and its impact on the record as observed in Aquileia. Eur J Archaeol 2017:1–29 Galinié H (2004) L’expression Bterres noires^, un concept d’attente. In: Verslype L, Brulet R (eds) Terres noires dark earth, Actes de la tableronde internationale tenue à Louvain-la-Neuve les 09 et 10 Novembre, vol 2001. UCL, Raymond Brulet, Louvain-la-Neuve, pp 1–11 Gelichi S (2000) L’eliminazione dei rifiuti nelle città romane del nord Italia tra antichità ed alto medioevo. In: Raventós XD, Remolà Vallverdú JA (eds) Sordes Urbis: la eliminación de resduos en la ciudad romana. L’Erma di Bretschneider, Rome, pp 13–23 Goldberg P, Berna F (2010) Micromorphology and context. Quat Int 214: 56–62 Goldberg P, Macphail RI (2010) Archaeological materials. In: Stoops G, Marcelino V, Mees F (eds) Interpretation of micromorphological features of soils and regoliths. Elsevier, Amsterdam, pp 589–622 Grimalt JO, Albaigés J (1990) Characterization of the depositional environments of the Ebro Delta (western Mediterranean) by the study of sedimentary lipid markers. Marine Geology 95:207–224 Hatcher PG, McGillivary PA (1979) Sewage contamination in the New York Bight. Coprostanol as an indicator. Environ Sci Technol 13: 1225–1229. https://doi.org/10.1021/es60158a015 Heimdahl J (2005) Urbanised nature in the past. Site formation and environmental development in two Swedish towns, AD1200-1800. Dissertation, University of Stockholm. http://su.diva-portal.org/ smash/get/diva2:197550/FULLTEXT01 Kaufman L, Rousseeuw PJ (2005) Finding groups in data: an introduction to cluster analysis. Wiley-Interscience, Hoboken-NJ Knights BA, Dickson CA, Dickson JH, Breeze DJ (1983) Evidence concerning the Roman military diet at Bearsden, Scotland, in the 2nd century A.D. J Archeol Sci 10:139–152 La Rocca HC (1986) Dark Ages a Verona. Edilizia privata, aree aperte e strutture pubbliche in una città dell’Italia settentrionale. Archeol Medievale 13:31–78 Leeming R, Latham V, Rayner M, Nichols P (1997) Detecting and distinguishing sources of sewage pollution in Australian inland and coastal waters and sediments. ACS Symp Ser 671:306–319 Lin DS, Connor WE, Napton LK, Heizer RF (1978) The steroids of 2000year-old human coprolites. J Lipid Res 19:215–221 Lindbo DL, Stolt MH, Vepraskas M (2010) Redoximorphic features. In: Stoops G, Marcelino V, Mees F (eds) Interpretation of micromorphological features of soils and regoliths. Elsevier, Amsterdam, pp 129–147 Loveluck C (2004) Terres noires and early medieval rural settlement sequences: conceptual problems, descriptive limitations and deposit diversity. In: Verslype L, Brulet R (eds) Terres noires dark earth, Actes de la table-ronde internationale tenue à Louvain-la-Neuve les 09 et 10 Novembre, vol 2001. UCL, Raymond Brulet, Louvain-la-Neuve, pp 86–96 Macphail RI (1994) The reworking of urban stratigraphy by human and natural processes. In: Hall AR, Kenward HK (eds) Urban-rural connexions: perspectives from environmental archaeology. Symposia of the Association for Environmental Archaeology No. 12, Oxbow Monograph 47, Oxford, pp. 13–43 Macphail RI, Galinié H, Verhaeghe F (2003) A future for dark earth? Antiquity 77(296):349–358 Matthews W, French CAI, Lawrence T, Cutler DF, Jones MK (1997) Microstratigraphic traces of site formation processes and human activities. World Archaeol 29:281–308 McGowan G, Prangnell J (2006) The significance of vivianite in archaeological settings. Geoarchaeology 21(1):93–111 Midvedt AC, Midvedt T (1993) Conversion of cholesterol to coprostanol by the intestinal microflora during the first two years of human life. J Pediatr Gastroenterol Nutr 17:161–168 Milek K (2012) Floor formation processes and the interpretation of activity areas: an ethnoarchaeological study of turf buildings at Thverá, Northeast Iceland. J Anthropol Archaeol 31:119–137 Miller CE, Sievers C (2012) An experimental micromorphological investigation of bedding construction in the middle stone age of Sibudu, South Africa. J Archaeol Sci 39:3039–3051 Miller CE, Conard NJ, Goldberg P et al (2009) Dumping, sweeping and trampling: experimental micromorphological analysis of anthropogenically modified combustion features. In: Théry-Parisot I, Chabal L, Costamagno S (eds.) The taphonomy of burned organic residues and combustion features in archaeological contexts. Proceedings of the round table, Valbonne, May 27–29 2008. P@lethnologie 2:25-37 Moinerau J (1970) Aspects pédologiques de la sédentarisation humaine. Apports et limites des études pédologiques dans les sites archéologiques. Quaternaria 13:205–223 Archaeol Anthropol Sci Mozzi P, Piovan S, Rossato S et al (2010) Palaeohydrography and early settlements in Padua (Italy). Il Quaternario 23:387–400 Murtaugh JJ, Bunch RL (1967) Sterols as a measure of fecal pollution. J Water Pollut Control Fed 39:404–409 Nardi S, Concheri G, Dell’Agnola G, Scrimin P (1991) Nitrate uptake and ATPase activity in oat seedlings in the presence of 2 humic fractions. Soil Biol Biochem 23:833–836 Nardi S, Muscolo A, Vaccaro S et al (2007) Relationship between molecular characteristics of soil humic fractions and glycolytic pathway and Krebs cycle in maize seedlings. Soil Biol Biochem 12:3138–3146 Nicosia C, Devos Y (2014) Urban Dark Earth. In: Smith C (ed) Encyclopedia of global archaeology, vol 11. Springer, New York, pp 7532–7540 Nicosia C, Balista C, Cupitò M, Ertani A, Leonardi G, Nardi S, Vidale M (2011) Anthropogenic deposits from the bronze age site of Fondo Paviani (Verona, Italy): Pedochemical and micropedological characteristics. Quat Int 243:280–292 Nicosia C, Langohr R, Mees F, Arnoldus-Huyzendveld A, Bruttini J, Cantini F (2012) Medieval Dark Earth in an active alluvial setting from the Uffizi gallery complex in Florence, Italy. Geoarchaeology 27:105–122 Nicosia C, Devos Y, Borderie Q (2013) The contribution of geosciences to the study of European dark earths: a review. PCA - Post-Classical Archaeologies 3:145–170 Ninfo A, Ferrarese F, Mozzi P, Fontana A (2011) High resolution DEMs for the analysis of fluvial and ancient anthropogenic landforms in the alluvial plain of Padua (Italy). Geogr Fis Din Quat 34:95–104 Pepe C, Dizabo P (1990) Etude d’une fosse du 13ème siècle par les marqueurs biogéochimiques: chantier archéologique du Louvre (Paris). Rev d’Archéometrie 14:23–28 Pepe C, Dizabo P, Scribe P, Dagaut J, Fillaux J, Saliot A (1989) Les marqueurs biogéochimiques: application à l’ archéologie. Rev d’Archéométrie 13:1–12 Piccolo A (2009) The supramolecular structure of humic substances. Soil Science 166(11):810–832 Pratt C, Warnken J, Leeming R, Arthur MJ, Grice DI (2008) Degradation and responses of coprostanol and selected sterol biomarkers in sediments to a simulated major sewage pollution event: a microcosm experiment under sub-tropical estuarine conditions. Org Geochem 39:353–369. https://doi.org/10.1016/j.orggeochem.2007.06.009 Quirk MM, Wardroper AMK, Brooks PW et al (1980) Transformations of acyclic and cyclic isoprenoids in recent and sedimentary environments. In: Daumas R (ed.) Biogeochemie de la Matiere Organique a l’Interface Eau-Sediment Marin. CNRS, pp 225–32 Simpson IA, Bull ID, Dockrill SJ, Evershed RP (1998) Early anthropogenic soil formation at Tofts Ness, Sanday, Orkney. J Archaeol Sci 25:729–746 Sistiaga A, Berna F, Laursen R, Goldberg P (2014) Steroidal biomarker analysis of a 14,000 years old putative human coprolite from Paisley Cave, Oregon. J Archaeol Sci 41:813–817 https://doi.org/10.1016/j. jas.2013.10.016 Smith ME (1989) Cities, towns, and urbanism: comment on Sanders and Webster. Am Anthropol 91:454–461 Smith ME (2008) Aztec city-state capitals. University Press of Florida, Gainesville Staski E (1982) Advances in urban archaeology. In: Schiffer MB (ed) Advances in archaeological method and theory, vol 5. Academic Press, New York, pp 97–149 Stevenson FJ (1994) Humus chemistry: genesis, composition, reactions, second edn. Wiley, New York Stiner MC, Kuhn SL, Weiner S, Bar-Yosef O (1995) Differential burning, recrystallization and fragmentation of archaeological bone. J Archaeol Sci 22:223–237 Stoops G (2003) Guidelines for analysis and description of soil and regolith thin sections. Soil Science Society of America, Madison Takada H, Farrlngton JW, Bothner MH et al (1994) Transport of sludgederived organic pollutants to deep-sea sediments at deep water dump site 106. Environ Sci Technol 28:1062–1072 Verleyen E, Sabbe K, Vyverman W, Nicosia C (2017) Siliceous microfossils from single-celled organisms: diatoms and chrysophycean stomatocysts. In: Nicosia C, Stoops G (eds) Archaeological soil and sediment micromorphology. John Wiley & Sons, Chichester, pp 165–170 Villagran X, Huisman H, Mentzer SM et al (2017) Bone and other skeletal tissues. In: Nicosia C, Stoops G (eds) Archaeological soil and sediment micromorphology. John Wiley & Sons, Ltd, Chichester, pp 11–38 Wardroper AMK, Maxwell JR, Morris RJ (1978) Sterols of a diatomaceous ooze from Walvis Bay. Steroids 32:203–221. https://doi.org/ 10.1016/0039-128X(78)90006-5 Wold S, Esbensen K, Geladi P (1987) Principal component analysis. Chemom Intell Lab Systems 2:37–52 Yule B (1990) The dark earth and lateRoman London. Antiquity 64:620– 628 Zech W, Senesi N, Guggenberger G, Kaiser K, Lehmann J, Miano TM, Miltner A, Schroth G (1997) Factors controlling humification and mineralization of soil organic matter in the tropics. Geoderma 79: 117–161