See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/305078621

An integrated study of Dark Earth from the alluvial valley of the Senne river (Brussels, Belgium)

Article in Quaternary International · July 2016 DOI: 10.1016/j.quaint.2016.06.025

CITATIONS READS 22 470

12 authors, including:

Yannick Devos Cristiano Nicosia Vrije Universiteit Brussel University of Padova

69 PUBLICATIONS 517 CITATIONS 78 PUBLICATIONS 642 CITATIONS

SEE PROFILE SEE PROFILE

Luc Vrydaghs Lien Speleers Vrije Universiteit Brussel Royal Belgian Institute of Natural Sciences

69 PUBLICATIONS 1,781 CITATIONS 9 PUBLICATIONS 40 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

PROLONG View project

Phytolith online database View project

All content following this page was uploaded by Irene Esteban on 06 November 2017.

The user has requested enhancement of the downloaded file. Quaternary International 460 (2017) 175e197

Contents lists available at ScienceDirect

Quaternary International

journal homepage: www.elsevier.com/locate/quaint

An integrated study of Dark Earth from the alluvial valley of the Senne river (Brussels, Belgium)

Yannick Devos a, *, Cristiano Nicosia a, Luc Vrydaghs a, Lien Speleers b, Jan van der Valk f, Elena Marinova b, Britt Claes c, Rosa Maria Albert d, i, Irene Esteban d, Terry B. Ball g, Mona Court-Picon b, h, Ann Degraeve e a Centre de Recherches en Archeologie et Patrimoine, Universite Libre de Bruxelles, Belgium b Royal Belgian Institute of Natural Sciences, Brussels, Belgium c Royal Museums of Art and History, Brussels, Belgium d ERAAUB, Department of History and Archaeology, University of Barcelona, Spain e Department of Archaeology, Brussels Regional Public Service, Belgium f School of Anthropology and Conservation, University of Kent, UK g Department of Ancient Scripture, 210H JSB, Brigham Young University, Provo, USA h Research Team in Archaeo- and Palaeo-Sciences, Brussels, Belgium i ICREA (Catalan Institution for Research and Advanced Studies), Spain article info abstract

Article history: The present article discusses the integration of urban geoarchaeological and archaeobotanical data of a Available online 7 July 2016 series of Dark Earth deposits situated in the alluvial valley of the Senne River in Brussels, Belgium. Due to their homogeneous character, their interpretation is a huge challenge for archaeologists. Through a case- Keywords: by-case approach, a detailed picture of the sequence of different activities and (semi-) natural events, Urban geoarchaeology leading to the build up of Dark Earth at each individual site has been obtained. Among the activities, Dark Earth agricultural practices and waste management, which are rarely archaeologically recorded in urban Soil micromorphology contexts, have been identified. Despite being situated in a valley context, none of the sites presented an Phytoliths Carpology excellent preservation of the organic plant remains. However, through the integration of geo- Palynology archaeological and botanical data, the taphonomical history of the botanical remains could better be understood, allowing us to document the evolution of the environment surrounding the sites. Addi- tionally, it has been demonstrated that the botanical study of Dark Earth units can also provide valuable information on vegetal consumption, in particular fruits. On a broader scale, this integrated geo- archaeologial and archaeobotanical study sheds some light on the process of urbanisation of the Senne alluvial valley between the 11the12th and the 16th century AD. © 2016 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction deposits in cities (Macphail, 1994; Macphail et al., 2003; Galinie, 2004). Today Dark Earth has become a more general expression Urban sites are characterized by many superimposed phases of used in European urban archaeology to designate thick, dark col- human occupation resulting in the repeated accumulation of cul- oured, humic, homogeneous units that are often rich in anthropo- tural sediments (Butzer, 2008). Such accumulations give rise to genic remains (charcoal, ceramic, brick, bone, mortar, coprolites, complex urban stratigraphic sequences (see Carver, 1987). Dark etc.), regardless of their age or geographical location. Importantly, Earths are a typical component of these sequences encountered in the expression does not correspond to a soil classification name, urban settings. Traditionally, they were defined as the uniform dark nor does it imply a univocal archaeological interpretation (Galinie, coloured units occurring between stratified antique and medieval 2004; Devos et al., 2009, 2011a; Fondrillon, 2009; Nicosia and Devos, 2014). As these units lack internal stratification, the under- standing of their formation and their interpretation is a huge challenge for archaeologists. Geoarchaeological studies involving * Corresponding author. detailed field observations, soil micromorphology, eventually E-mail address: [email protected] (Y. Devos). http://dx.doi.org/10.1016/j.quaint.2016.06.025 1040-6182/© 2016 Elsevier Ltd and INQUA. All rights reserved. 176 Y. Devos et al. / Quaternary International 460 (2017) 175e197 complemented by physico-chemical analyses have been shown to initially situated on several small islands formed by the Senne be a powerful tool for understanding the various processes and River. Archaeological excavations show that the earliest occupation human activities that control their formation (Macphail, 1994; dates back to the 9the10th century and is dispersed over 4 pre- Macphail and Cruise, 2000; Cammas, 2004; Goldberg and urban nuclei: A) around the Saint-Goriks church on one of the Macphail, 2006; Cremaschi and Nicosia, 2010; Borderie, 2011; larger Senne islands, B) on the higher western slopes in the actual Cammas et al., 2011; Nicosia et al., 2012, 2013, in press; Devos Coudenberg, C) Treurenberg and D) the Place de la Vieille Halle- et al., 2013b; Nicosia and Devos, 2014; Borderie et al., 2015). aux-Bles (Fig. 1). At the beginning of the 13th century AD, a first city Taking into account their complex formation, the interpretation wall was built surrounding these nuclei. Intensive urbanisation was of botanical remains observed within Dark Earth can also be quickly followed by a second city wall around the middle of the particularly difficult (Devos et al., 2011a,c; Vrydaghs et al., 2016). 14th century AD (Degraeve et al., 2010). Since 1996, the Heritage Direction of the Brussels Capital Region The oldest site documenting the pre-urban development is the has carried out systematic archaeological assessment on all site of Petite Rue des Bouchers, situated near the Grand-Place, in scheduled construction works. Whenever remains are at risk, the heart of the medieval city within the first city wall at an altitude archaeological research is conducted. Geoarchaeological and of ca. 20 m above sea level (Fig. 1). According to the geotechnical archaeobotanical research has become a systematic part of the map, the archaeological layers rest directly on top of the Ypresian archaeological investigations in the historical centre of Brussels (Eocene) substratum (Dam et al., 1977). On top of the substratum (see Devos et al., 2007a,b, 2009, 2011a, 2013a; Speleers and van der (Unit 178), a truncated (i.e., missing the top part due to erosion as Valk, in press). During the early years, interventions mainly focused indicated by the abrupt planar boundary suggesting an erosional on Dark Earth situated in the higher part of the city, above the pre- discontinuity) organo-mineral layer (Unit 193) is overlain by Holocene terraces onto which the valleys of the Senne River and its truncated sands (Units 177 and 194) (Fig. 2a). These are covered by tributaries are incised. Here the preservation of the botanical re- organic-rich layers (Units 194, 203 and 147), the upper one (Unit mains with the exception of phytoliths and charred remains has 147) presenting traces of puddling (Fig. 2b). 14C dating of carbonised been shown to be very poor, thus limiting any vegetation recon- cereal grains places unit 147 in the 10the11th century AD (Van struction. However, in recent years, Dark Earth has been discovered Strydonck and Boudin, 2011). On top of it, a thick Dark Earth on several archaeological excavations within the Senne alluvial (Units 139, 140, 141, 142, 143 and 144) developed in which 58 valley, where a better preservation of the botanical remains can be potsherds were collected (Fig. 3a). They are all small fragments that expected. In this present article we will present the results of an were found spread throughout the sequence. According to these integrated geoarchaeological and archaeobotanical study of three remains, the Dark Earth can be dated to the 11the12th century AD such Dark Earths situated in the alluvial valley: the Poor Clares, the (Claes, 2011). site of Rue des Pierres nr. 18e38, and the site of Petite Rue des The site of the Poor Clares in the Rue de Laeken documents the Bouchers nr. 29. The three sites have been chosen as they each intensive urbanisation period and is situated in the area between document another part of the urbanisation process. We will the first and the second city wall, on the left bank of the ancient particularly focus on the taphonomical history of the botanical course of the Senne river at an altitude of ca. 18 m above sea level remains observed within the Dark Earths and their potential for (Fig. 1). According to the geotechnical map, the site is situated in a environmental reconstruction. We will also evaluate how these zone characterized by an Ypresian substratum, on top of which a Dark Earths can inform us about different activities that have taken sequence of alluvial gravels, sands, peat and finally clays were place at the individual sites and to what extent they can contribute deposited (Dam et al., 1977). On top of the natural sequence, thick to our understanding of the urbanisation process in the lower part Dark Earth units were observed (Figs. 3b and 4). Within the upper of Brussels and its changing environment. part of the Dark Earth (Unit 412) some small potsherds dating from the 14the15th century AD were recovered (Claes, 2008). The lower 2. The studied sites part (Unit 415) did not contain any datable potsherds, but radio- carbon dating on charcoal fragments from this unit places it in the Hydrographically speaking, Brussels belongs to the basin of the 12the14th century AD. river Scheldt with the river Senne as its most important water- The site of Rue des Pierres, documenting the full urbanisation, is course. The basin was formed at the end of the Tertiary (approx. 2 situated in the centre of the perimeter of the first city wall, but on million years ago). Traditionally, it is thought that the city was the right bank of the ancient course of the Senne river, at an altitude

Fig. 1. 3D image of the alluvial valley of the Senne river according to the LIDAR data. The covered area measures 5.1 by 6.7 km. Height differences have been exaggerated 10 . 1: Â Poor Clares; 2: Rue des Pierres; 3: Petite Rue des Bouchers; 4: Rue de Dinant; 5: Treurenberg; 6: Court of Hoogstraeten. A: nucleus around the Saint-Goriks church; B: nucleus around the Coudenberg; C: nucleus around the Treurenberg; D: nucleus around the Place de la Vieille Halle-aux-Bles. Y. Devos et al. / Quaternary International 460 (2017) 175e197 177

Fig. 2. a: Detail of the natural sequence observed on the site of Petite Rue des Bouchers. Units 193 and 177 are clearly truncated. b: the lower part of the Dark Earth, and the puddled layer (Unit 147) of the site of Petite Rue des Bouchers. The blue arrow points to traces of trampling, whereas the red arrow points rather to puddling/poaching. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. Drawings of the studied profiles. a: the site of Petite Rue des Bouchers; b: the site of the Poor Clares; c: the site of Rue des Pierres (after Pion, 2015). 178 Y. Devos et al. / Quaternary International 460 (2017) 175e197

and with oblique incident light (OIL) at magnifications of 25 ,  100 , 200 , 400 and 500 . Additionally, the thin sections were     equally scanned under UV and blue fluorescence at 400 magni-  fication. The thin sections were described following the interna- tional nomenclature of Stoops (2003).

3.1.3. Physico-chemical analyses On the samples of Poor Clares and Rue des Pierres comple- mentary analyses were performed on the <2 mm fraction obtained after gentle crushing and sieving of air-dried bulk samples at the laboratory of INRA-Arras (Institut National de la Recherche Agronomique). The standard procedures published by the French Association for Normalisation (AFNOR, 1999) were adopted, except where otherwise indicated. The following analyses were carried out:

- Particle size distribution by sieving and sedimentation (see method in Baize, 2000), - Soil pH (H2O) with a soil/water ratio of 1:2, - Organic carbon and total nitrogen after dry combustion, - Heavy metals: Cu, Zn, Ni (total HF ICP-AES) and Pb (total HF ICP-

Fig. 4. The Dark Earth observed on the site of the Poor Clares. Note the presence of MS e see method in Baize, 2000). mottling within Unit 414e416. Organic and inorganic phosphorus were measured at the labo- ratory of Soil Science at the University of Ghent following of ca. 16 m above sea level (Fig. 1). The geotechnical map also places Mikkelsen (1997) (see: Devos et al., 2011b). this site on a Ypresian substratum, on top of which a sequence of Similarity-indices (Comparative Particle Size Distribution index alluvial sediments similar to the site of the Poor Clares was matrices, hereafter: S.I.) were calculated following Langohr et al. deposited (Dam et al., 1977). Dark Earth units rest directly on top of (1976). these natural deposits (Fig. 3c). Two sequences have been studied. The first one (section A: Units 119, 120 and 121) contained a series 3.2. Archaeobotanical studies of potsherds, allowing researchers to date it to the 13the15th century AD. After that, they were covered by a layer of construction 3.2.1. Phytoliths debris (Unit 107). The second sequence (section B: Units 28, 10, 74, Phytolith analysis of thin sections were conducted on petro- 75, 76, 77 and 78) can be dated to the 15the16th century AD (Pion, logical microscopes at magnifications of 25 ,100 and 400 un- Â Â Â 2015). der plain polarized light (PPL) and crossed polarizers (XPL) following Devos et al. (2013a) and Vrydaghs et al. (2016). From two 3. Material and methods up to four squares of 25 mm2 were selected per stratigraphic unit for analysis. Their selection relied on indications deriving from 3.1. Geoarchaeological study micromorphological analysis. Within thin sections, phytoliths were observed either within 3.1.1. Field study organic material or as sedimentary particles. In the latter case, three Field descriptions followed the “Comprehensive Field Data Ba- basic distribution patterns of opal phytoliths have been recorded: ses” (Langohr, 1994) and the “Guidelines for soil profile description” isolated, clustered and articulated. Isolated phytoliths are totally (F.A.O, 2006). Adapted checklists have been applied for the disarticulated and widely separated. Clustered phytoliths consists description of associated soil traces (Fechner et al., 2004). Bulk of a group of disarticulated phytoliths wherein not all phytoliths are samples and undisturbed oriented blocks for soil micromor- necessarily of the same type or share the same orientation. Artic- phology, physico-chemical analyses and archaeobotanical studies ulated phytoliths are those that appear to maintain the relative were taken from the available soil profiles following the guidelines distribution that they had within the plant tissues in which they of Devos (2013) (Fig. 3). Micromorphological samples were chosen were produced (for a discussion of different forms of articulated according to the sampling rationale described in Nicosia et al. (in phytoliths see Albert et al., 2008; Devos et al., 2013a; Shillito, 2011; press). In addition core samples for pollen analysis were also Vrydaghs et al., 2016). taken from the peaty layers of the sites of Poor Clares and Rue des Analysis of the phytoliths observed as sedimentary particles was Pierres. conducted in a three-step approach:

3.1.2. Soil micromorphology 1. a systematic recording of the distribution patterns of the phy- Thin sections (60 80 mm) for the site of the Poor Clares were toliths, aiming at establishing their relative frequency was  prepared from air-dried undisturbed blocks following the standard conducted; laboratory procedures of T. Beckmann (Germany; see Beckmann, 2. a subsequent morphotypological analysis of each distribution 1997). Thin sections (60 90 mm and 90 120 mm) for the sites pattern was conducted;   of Petite Rue des Bouchers and Rue des Pierres, were manufactured 3. a counting of phytoliths within each distribution pattern in or- by the laboratory for Mineralogy and Petrography, Ghent University der to obtain numerical data was conducted. (Belgium) according to the guidelines of Benyarku and Stoops (2005). Observations were made with petrological microscopes The morphotypological analyses were carried out according to a under plain polarised light (PPL), under crossed polarizers (XPL) revised version of Runge (1999) inventory (see Devos et al., 2013a). Y. Devos et al. / Quaternary International 460 (2017) 175e197 179

Naming of the phytoliths followed the nomenclature of ICPN 1.0 des Pierres, although some pollen grains (like Juniperus)couldbe whenever possible (ICPN Working Group et al., 2005). Morpho- lost by this procedure. Instead, the sample of the Poor Clares was metric analyses were performed on a series of articulated systems treated using deflocculation and density separation following following Vrydaghs et al. (2016) and Ball et al. (2016). In addition, Bastin and Coûteaux (1966) and Nakagawa et al. (1998). In order the phytolith content of selected archaeological inclusions (e.g. to evaluate the preservation and content two pollen slides per coprolites, ash remains, ceramics) visible in thin sections was also sample were completely counted in the laboratories of the Royal determined. Belgian Institute of Natural Sciences using corresponding refer- For control purposes relating to the preservation of the phy- ence literature (Moore et al., 1991; Beug, 2004) and reference toliths, some bulk samples were processed following the method collection. The pollen sum we report consists of identifiable developed by Katz et al. (2010) at the Laboratory of Archaeology terrestrial arboreal (AP) and non-arboreal (NAP) pollen and ex- of the University of Barcelona. For the microscopic study, 50 ml of cludes aquatic and marsh plants. In order to avoid the “masking” the supernatant was placed on a slide and covered with a effect of the high abundance of Cerealia-type pollen, they were 24 mm 24 mm coverslip. Phytoliths were counted and excluded from the pollen sum and calculated as percentage from  morphologically identified in 20 random fields at 200 and the total pollen. Excluding Cerealia-type allows to illustrate better  400 magnification respectively. Morphological identifica- the abundance of other relevant types for infering the past  tion followed the www.phytcore.org reference collection and vegetation pollen types. Non-Pollen-Palynomorphs (NPP) were nomenclature followed ICPN Working Group et al., (2005) counted in the samples prepared for routine pollen analysis. The whenever possible. “types” identified follow the type numbers described by van Geel (1986, 2001) and van Geel et al. (1995). 3.2.2. Plant macrofossils Macrofossil samples were wet-sieved using mesh sizes 2 mm, 4. Results and discussion 1 mm, 0.5 mm, and 0.2 mm. The sieving residues were sorted, identified and counted under a binocular microscope. The coarser The field data are summarized in Table 1. Fig. 5 reports the re- fraction (2 mm) was analysed completely, while only representa- sults of the similarity-indices (S.I.) and Table 2 the physico- tive parts of the finer fractions were analysed. The quantities of the chemical data. The main micromorphological characteristics are macrofossils were expressed in numbers of remains counted. The presented in Table 3A, 3B and 3C. The phytolith results are pre- numbers of remains counted in the studied sub-samples of the sented in Tables 4, 5, 6 and 7 and Fig. 6. A comparison of the data finer fractions where extrapolated to the entire sample volumes. from the thin section study of Units 139 and 140 of Petite Rue de Identifications were made using the reference collection of the Bouchers and the control bulk samples indicate that with the Royal Belgian Institute of Natural Sciences, identification keys and exception of types such as the Blocky and the indeterminate seed atlases (Schoch et al., 1988; Cappers et al., 2006; Bojnansky epidermal tissue phytoliths, all phytolith types are recorded by and Fargasova, 2007; Jacomet, 2006; Knorzer,€ 2007; Mauquoy both types of analysis. Clearly, the most significant environmental and Van Geel, 2007). The nomenclature of the plant taxa fol- markers and their prevalent frequencies are recorded by the anal- lowed Lambinon et al. (1998). ysis of the bulk samples and the thin sections. The data of the plant macroremains study are presented in Table 8. A macrobotanical 3.2.3. Pollen evaluation was carried out on a sample of Unit 415 of the Poor On the sites of Petite Rue des Bouchers and Rue des Pierres Clares, but it has not been studied in detail because it contained sub-samples of 3 cm3 were taken for pollen analysis, whereas on only very few plant remains (Houchin, 2009). The same applies for the site of Poor Clares a sample of 31.9 g of sediment was selected. some of the pollen samples: Unit 415 of the Poor Clares, the upper All the samples were processed according to Faegri and Iversen 40 cm of coring 4 ( lower part of Unit 121) of Rue des Pierres and ¼ (1989). In order to calculate the pollen concentration Lycopo- Unit 147 of Petite Rue de Bouchers did not contain a sufficient dium tablets were added to the samples (Stockmarr, 1971). Clay amount of pollen. contamination was removed by ultrasonic sieving (5 mmopening Pollen data are presented in Fig. 7 and in the Electronic of the sieves) for the samples of Petite Rue des Bouchers and Rue Supplementary Data.

Table 1 Basic field descriptions and overview of the analyses (Mi: micromorphology; PC: physico-chemical analyses; Po: pollen; Ph: phytoliths; PM: plant macroremains).

Site Unit Description Mi PC Po Ph PM

Poor Clares 402 Brick wall. 401 Loose bricks and mortar fragments, abrupt lower boundary. 406 Bricks and brick fragments mixed with pale brown sandy loam to sand silt loam, abrupt lower boundary. 412 Pale brown sandy loam to sandy silt loam, humiferous, straight XX XX abrupt lower boundary. Contains charcoal, brick fragments, molluscs, bone and ceramics. 415 Light brownish gray clay loam, humiferous,́ slightly undulating XXXXX abrupt lower boundary. Contains charcoal, rare brick and ceramic fragments. 414e416 Brown clay to clay loam, iron mangese mottling. X X X Rue des Pierres 107 Loose bricks and mortar fragments mixed with dark gray clay loam, straight abrupt lower boundary. 109 Dark gray clay loam, abrupt to clear undulating lower boundary. XX X Contains charcoal. 119 Pale brown to light brownish gray clay loam to sandy silt loam, XX X clear undulating lower boundary. Contains charcoal and burned soil/ceramic fragments. (continued on next page) 180 Y. Devos et al. / Quaternary International 460 (2017) 175e197

Table 1 (continued )

Site Unit Description Mi PC Po Ph PM

120 Pale brown to light brownish gray clay loam to sandy silt loam, XX X clear undulating lower boundary. Contains charcoal and burned soil/ceramic fragments. 121 Light olive gray clay loam to sandy silt loam. X X X 28 Dark gray clay loam, abrupt undulating lower boundary. Contains XXX charcoal, brick and mortar fragments. 10 Whitish mortar/plaster layer. X 74 Dark gray clay loam, olive green mottles, clear undulating lower XX boundary. Contains charcoal. 75 Dark gray clay loam, clear undulating boundary. Contains charcoal. X X 76 Dark gray clay loam, gradual boundary. Contains charcoal. X X 77 Dark gray clay loam, clear undulating boundary. Contains charcoal. X X 78 Dark gray clay loam. Contains charcoal. X X X Petite rue 139 Dark gray loam, humiferous, olive green mottles, clear straight lower XXX des Bouchers boundary. Contains rare charcoal. 140 Gray loam, olive green mottles, clear straight lower boundary. XXX Contains rare charcoal. 141 Gray loam, olive green mottles, clear straight lower boundary. XXXX Contains rare charcoal. 142 Gray loam, olive green mottles, clear straight lower boundary. XX Contains rare charcoal. 143 Gray, loam, olive green mottles, clear straight lower boundary. XX Contains rare charcoal. 144 Gray, sandy loam, olive green mottles, iron nodules, abrupt XXXX slightly undulating lower boundary. Contains rare charcoal. 145 Sand, iron crust at the lower boundary. 146 Sand, iron crust at the lower boundary. X 147 Very dark grayish brown organic silt loam, abrupt lower boundary. X X 148 Involutions of very dark grayish brown and pale yellow sand. Abrupt lower boundary. 203 Gray sand. Clear undulating lower boundary. X X X 194 Light gray sand. Abrupt undulating lower boundary. X X X 177 Pale yellow sand. Abrupt slightly undulating boundary. X 193 Very dark grayish brown sandy clay loam. Abrupt undulating lower boundary. X 178 Gray blue sandy clay loam. X X

Table 2 Physico-chemical data.

Site Unit Clay % Silt % Sand % pH Organic matter g/kg C/N Total P mg/kg Inorganic P mg/kg Organic P mg/kg Cu mg/kg Zn mg/kg Pb mg/kg Ni mg/kg

Poor Clares 412 13 31 56 8.2 15.9 16 10,920 10,560 360 80 100 160 10 415 22 44 34 7.7 17.3 10 5600 5290 310 40 80 40 20 414 34 57 9 7.6 7.5 9 3800 3730 70 20 80 20 30 Rue des 109 19 46 35 7.7 28.8 21 4670 3280 1390 40 130 20 20 Pierres 119 17 46 37 7.6 19.7 18 4430 3960 470 30 100 20 20 120 18 52 30 7.5 22.1 22 4360 4270 90 30 80 20 20 121 17 48 35 7.7 16.7 15 4480 4020 460 20 90 20 10 77 18 45 37 7.9 36.0 17 7060 5860 1200 78 20 46 34 7.8 40.6 17 10,500 8450 2040

Table 3A Petite Rue des Bouchers (Brussels): main micromorphological characteristics.

Unit Microstructure and Groundmass Organic material Anthropogenic material Pedofeatures porosity

178 Weak channel, locally weak Porphyric. Coarse fraction: Frequent roots e Fe hypocoatings platy; silt/silt loam with medium macropores ca. 10% sand bands 193 Weak channel, locally weak Well sorted silt to very fine Root fragments, organic e Fe hypocatings platy; sands fine material macropores ca. 10% 177 Basic microstructure, Well sorted fine and Root fragments ee coarse monic; medium sands construction voids 194 Apedal; construction voids Coarse monic. Organic very Wood fragments ee fine sands and silts; reworked aggregates from underlying units 203 Apedal; construction voids Coarse monic. Organic very Frequent vegetal tissue and ee fine sands and silts organ fragments 147 Weak channel; macropores Coarse monic. Coarse Dominant organic fine ee <5% fraction: sandy silts material; vegetal organ and tissue fragments; rare seed coats and sclerotia Y. Devos et al. / Quaternary International 460 (2017) 175e197 181

Table 3A (continued )

Unit Microstructure and Groundmass Organic material Anthropogenic material Pedofeatures porosity

143 Weak/moderate channel; Porphyric. Coarse fraction: Organic fine material Charcoal Orthic Fe nodules and macropores ca. 10% dominant hypocoatings. Phosphate moderately sorted silt to nodules fine sand quartz, non-calcareous 142 Moderate channel; Coarse monic/porphyric. Organic fine material; Charcoal, one fragment of Orthic Fe nodules. macropores ca. 15e20% Coarse fraction: scarce vegetal tissue ceramic material Phosphate nodules. sandy loam fragments Vivianite crystal intergrowths 141 Weak/moderate channel; Coarse monic/porphyric. Organic fine material; Charcoal Scarce phosphate nodules macropores ca. 10% Coarse fraction: scarce seed coats and hypocoatings sandy loam 140 Weak/moderate channel; Coarse monic/porphyric. Organic fine material; Charcoal, scarce bone, Phosphate dense complete macropores ca. 10% Coarse fraction: scarce seed coats charred grains/seeds, one infillings. Phosphate sandy loam ceramic material fragment, hypocoatings and nodules. one metal slag 139 Moderate to strong Porphyric. Coarse fraction: Organic fine material; Charcoal Fe coatings; hypocoatings channel; sandy loam scarce seed coats and quasi-coatings. macropores ca. 15e25% Phosphate hypocoatings.

Table 3B Poor Clares (Brussels): main micromorphological characteristics.

Unit Microstructure and Groundmass Organic material Anthropogenic material Pedofeatures porosity

412 Granular to subangular Porphyric. Coarse fraction: Organ and tissue fragments, Mortar, eggshell, shell, Rare dusty clay coatings, blocky, locally crumbs, ca. dominant unsorted silt to organic fine material (cells ceramic, glass, bone, fish- rare vivianite crystals, 20% macropores, mainly medium quartz sand, and cell residues, bone, ashes, charcoal, phosphatic pedofeatures, channels calcareous dotted punctuations, fungal charred grains, omnivore excremental organo- micromass sclerotia) coprolites, slag, soil mineral micro-aggregates, fragments calcitic biospheroids, iron nodules 415 Channel to subangular Porphyric. Coarse fraction: Organic fine material (cells Ceramic, glass, bone, fish- Dusty clay coatings, short blocky, ca. 15e20% dominant unsorted silt to and cell residues, bone, charcoal, omnivore wavy lenses of dusty fine macropores, mainly medium quartz sand, non- punctuations, fungal coprolites, slag, soil material, excremental channels calcareous dotted sclerotia) fragments organo-mineral micro- micromass aggregates, phosphatic pedofeatures, passage features, iron nodules, iron pseudomorphs of organic material 416/414 Subangular blocky, ca. 10 Porphyric. Coarse fraction: Rare charcoal fragments Dusty clay silt coatings, iron e15% macropores, mainly dominant moderately nodules, iron depletion/ planes and channels sorted silt to fine quartz accumulation sand, non-calcareous

Table 3C Rue des Pierres (Brussels): main micromorphological characteristics.

Unit Microstructure and Groundmass Organic material Anthropogenic material Pedofeatures porosity

119 Channel; ca. 20% Porphyric. Coarse fraction: Organic fine material, Bone (burned and unburnt), Phosphate nodules, macropores dominant poorly sorted silt frequently Fe-substituted, carnivore-omnivore coatings, dense complete to medium sand quartz peaty fragments, organ and excrements, molten and incomplete infillings. tissue fragments phytoliths, rare coarse Vivianite crystal charcoal; burned soil intergrowths. Very rare thin fragments; reworked soil dusty clay coatings fragments 120 Channel; ca. 20% Porphyric. Coarse fraction: Organic fine material, organ Brick; reworked soil Dense complete and macropores dominant poorly sorted silt and tissue fragments fragments; carnivore- incomplete phosphate to medium sand quartz omnivore excrements, infillings, coatings and weathered mortar, volcanic nodules. Vivianite crystal rock fragments, charcoal intergrowths. Very rare thin dusty clay coatings 121 Channel; ca. 20e25% Porphyric. Coarse fraction: Organic fine material, organ Bone (burned and unburnt), Phosphate nodules, macropores dominant poorly sorted silt and tissue fragments carnivore-omnivore coatings, dense complete to medium sand quartz excrements, molten and incomplete infillings. phytoliths, rare coarse Vivianite crystal charcoal; burned soil intergrowths. Very rare thin fragments; reworked soil dusty clay coatings fragments (continued on next page) 182 Y. Devos et al. / Quaternary International 460 (2017) 175e197

Table 3C (continued )

Unit Microstructure and Groundmass Organic material Anthropogenic material Pedofeatures porosity

78 Weak channel; ca. 5e10% Porphyric. Coarse fraction: Organic fine material, Carnivore-omnivore Phosphate nodules, macropores (channels and dominant poorly sorted silt frequent organ and tissue excrements, reworked soil coatings, dense complete moldic voids) to medium sand quartz fragments fragments, rare charcoal, and incomplete infillings. bone and molten phytoliths Vivianite crystal intergrowths. 77 Weak channel; macropores Porphyric. Coarse fraction: Organic fine material, Dominant carnivore- Phosphate nodules, (channels and moldic dominant poorly sorted silt frequent organ and omnivore excrements, coatings, dense complete voids), ca. 5e10% to medium sand quartz elongated tissue fragments reworked soil fragments, and incomplete infillings. bone fragments, charcoal, Vivianite crystal molten phytoliths, intergrowths. Pyrite metallurgical slags, brick framboids and pottery 76 Channel; macropores,ca. 15 Porphyric. Mixture of Organic fine material, organ Carnivore-omnivore Phosphate nodules, e20% organic-rich aggregates and elongated tissue excrements, charcoal coatings, dense complete similar to US 77 with silt fragments and incomplete infillings. loam (substrate?) Vivianite crystal aggregates intergrowths 75 Channel; macropores Porphyric. Mixture of Organic fine material, organ Carnivore-omnivore Phosphate nodules, (channels and moldic organic-rich aggregates and elongated tissue excrements, reworked soil coatings, dense complete voids), ca. 15e20% similar to US 77 with silt fragments fragments, weathered bone and incomplete infillings. loam (substrate?) fragments, charcoal, Vivianite crystal aggregates weathered mortar intergrowths. Pyrite framboids 74 Channel; macropores Porphyric. Mixture of Organic fine material, organ Carnivore-omnivore Phosphate nodules, (channels and moldic organic-rich aggregates and elongated tissue excrements, reworked soil coatings, dense complete voids), ca. 15e20% similar to US 77 with silt fragments occasionally Fe- fragments, bone fragments, and incomplete infillings. loam (substrate?) substituted charcoal, molten phytoliths, Vivianite crystal aggregates weathered mortar, intergrowths. Pyrite metallurgical slags framboids 10 eeecm-thick mortar layer with localized dissolution traces 28 Apedal; areas with scarce Porphyric. Mixture of Organic fine material, organ Brick, reworked soil Phosphate nodules (mostly moldic voids organic-rich (and charcoal- and elongated tissue fragments, weathered reworked, rarely orthic). rich) with predominantly fragments occasionally Fe- mortar, carnivore- Vivianite crystal mineral domains substituted omnivore excrements, intergrowths metallurgical slags, molten phytoliths

Table 4 Semi-numerical data relative to the phytolith observed as sedimentary particles within the studied soil thin sections.

Site Unit n squares Isolated Clustered Articulated Total

Rue des Pierres 74 2 494 136 77 707 Poor Clares 412 3 202 24 2 228 415 3 678 286 107 1071 414e416 4 140 29 21 190 Petite rue des Bouchers 139 2 121 7 / 128 140 2 85 23 / 108 141 3 101 11 10 122 142 3 201 19 12 232 143 3 96 12 3 109 144 3 93 3 / 96 203 3 / / / / 178 2 6 / / 6

Table 5 Analysis of the phytolith content of the coprolitic material of the sites of Poor Clares and Rue des Pierres.

Site Unit Reference coprolite(s) Present (P)/Not observed (NO) Distribution n Plant type Plant part

Poor Clares 412 1 P Isolated 1 ? Vascular system 2 P Isolated 1 ? ? 415 1, 2, 3, 4 NO 5 P Isolated 3 Pooideae Leaf 6 P Isolated 1 ? ? 7 NO 8 P Isolated 1 ? ? 9 P Isolated 1 ? ? 10 P Articulated ? ? Vascular system Rue des Pierres 28 4 P Isolated 3 Pooideae Leaf Articulated 15 Cerealia Inflorescence 74 3 P Isolated 6 Poaceae Y. Devos et al. / Quaternary International 460 (2017) 175e197 183

Table 5 (continued )

Site Unit Reference coprolite(s) Present (P)/Not observed (NO) Distribution n Plant type Plant part

Clustered 5 Pooideae Leaf Articulated 4 Cerealia Inflorescence ? Poaceae Leaf 5 NO 6 P Isolated 3 Poaceae Leaf 75 12 NO 76e77 1 P Isolated 5 Poaceae and/or Cerealia Leaf and/or Inflorescence Clustered 3 Pooideae Leaf 2 P Isolated 4 Poaceae and Cerealia Leaf and Inflorescence 109 7 P Isolated 3 Pooideae Leaf 8 P Isolated 1 ? ? 119 9 NO 120 10 P Isolated 1 Poaceae Leaf 11 NO

Table 6 4.1. The site of Petite Rue des Bouchers Phytolith results on bulk samples from Petite rue des Bouchers. (phy g/ sed. phytoliths per gram of sediment; n number of morphologically identified ¼ ¼ 4.1.1. The parent material phytoliths; % diatoms percentage of diatoms; % WM the percentage of weath- ¼ ¼ ered morphotypes). The basal part of the studied sequence (Units 178, 193, 177, 194) were devoid of archaeological material. It consisted of a succession Sample Phy/g sed. n % Diatoms % WM of well sorted silts with occasional medium sand bands (Unit 178), SU 139 1,400,000 124 5.47 2.42 the upper part (Unit 193) enriched in fine organic matter, overlain SU 140 1,900,000 207 5.61 2.42 by medium sands containing rounded reworked aggregates of

Table 7 Phytolith morphological results on the studied bulk samples from Petite Rue des Bouchers.

Morphotype Inventory Plant type Plant part Unit

139 140

Trichome G3 Poaceae Leaf 4.96% 4.46% Dentriform G5 Poaceae Inflorescence 19.83% 24.26% Rondel G1.1 Poaceae/Pooideae (C3) Whole plant 15.70% 15.35% Long trapeziform G1.1 Poaceae/Pooideae (C3) Leaf 9.92% 5.94% Long trapeziform sinuous G1.1 Poaceae/Pooideae (C3) Leaf 9.92% 9.90% Bilobate G1.2 Poaceae/Panicoideae (C4) Whole plant 1.65% 0.50% Cross G1.2 Poaceae/Panicoideae (C4) Whole plant 0.83% / Saddle G1.3 Poaceae/Chloritoidea (C4) Whole plant / 0.50% Elongate A Monocotyledons 19.83% 21.29% Elongate facetate A4 Monocotyledons 3.31% 3.47% Blocky Dicotyledons Wood/bark 0.83% 0.50% Epidermal tissues indet F ? ? 0.83% / Irregulars & Indeterminates C (irregular) ? ? / 4.95% Fragments ? ? 12.40% 8.91%

Table 8 Plant macrofossils data (seeds and fruits, unless otherwise specified). w waterlogged, mi mineralized, ch charred, few, tens, hundreds, thousands. ¼ ¼ ¼ þ¼ þþ ¼ þþþ ¼ þþþþ ¼ Site Poor clares Petite Rue des Bouchers Rue des Pierres

Unit 412 194 203 147 144 141e143 140 78 28

Dating (century AD) 14the15th <10the11th <10the11th 10the11th 11the12th 11the12th 11the12th 15the16th 15the16th

Volume (L) 15.1 1 1 1 1 1 1 11 15.8

Economic plants Cereals Avena sp. ch ee e e11137 cf. Avena sp. ch ee e eeeee4 Hordeum vulgare ch ee e ee2 1 3 11 Hordeum sp. ch ee e 1 eeeee Panicum miliaceum (glumes) w ee e 1 eeeee Secale cereale mi 1 eeeeeeee cf. Secale cereale mi 1 eeeeeeee Secale cereale (rachis internodium) w ee e 3 eeeee Secale cereale (rachis internodium) ch 1 eeeee424 Secale cereale ch 1 ee1 e 1247 cf. Secale cereale ch 2 eeeeee47 Triticum aestivum/durum/turgidum ch 5 ee111114 Triticum spelta (glume base) w ee e 1 eeeee (continued on next page) 184 Y. Devos et al. / Quaternary International 460 (2017) 175e197

Table 8 (continued )

Site Poor clares Petite Rue des Bouchers Rue des Pierres

Unit 412 194 203 147 144 141e143 140 78 28

Dating (century AD) 14the15th <10the11th <10the11th 10the11th 11the12th 11the12th 11the12th 15the16th 15the16th

Volume (L) 15.1 1 1 1 1 1 1 11 15.8

Triticum sp. ch ee e 11113e Cerealia indet. ch 10 ee1 1 3 1 18 102

Pulses Lens culinaris ch 1 eeeeeeee cf. Pisum sativum ch ee e eeeee1 Pisum sativum/Vicia sativa ch ee e eeeee1 Vicia faba (hilum) mi 1 eeeeeeee

Collected or cultivated fruits and nuts Corylus avellana w e 231eeeee Corylus avellana ch ee e eeeee3 cf. Corylus avellana ch ee e eeee2 e Ficus carica w3 eeeeeee4 Ficus carica mi 10 eeeeeeee Fragaria vesca w ee e eeee600 e Fragaria vesca mi 58 eeeeee24 e Malus sylvestris (endocarp) w ee e eeeee þ Maloideae mi 2 eeeeeeee cf. Maloideae mi 7 eeeeee6 e Maloideae/Prunoideae mi 4 eeeeeee5 Prunus avium/cerasus mi ee e eeee11 e Prunus sp. mi ee e eeee10 e Prunoideae mi 4 eeeeee2 e Rubus fruticosus w e 13 13 6 7 e 2 241 150 Rubus fruticosus mi 1 eeeeeee2 Rubus idaeus w e 10 20 1 6 ee2 e Rubus sp. w e e 200 182 þþþþþ Rubus sp. mi ee e eeee33 2 Sambucus nigra w/mi e 29 19 32 24 143 77 460 114 Sambucus sp. w/mi 7 e 329 430 þþ þþ þ þþ Sambucus sp. ch ee e eeeee2 Vitis vinifera ch 1 eeeeeeee Vitis vinifera mi 19 eeeeee1 e

Kitchen herbs Foeniculum vulgare mi 2 eeeeeeee Apium graveolens w ee e 2 eeeee

Vegetables Beta vulgaris ch 1 eeeeeeee

Other economic plants Cannabis sativa w ee e 2 eeeee Humulus lupulus w ee e 3 eeeee Linum usitatissimum w ee e 2 eeeee Linum usitatissimum (capsule fragment) w ee e 56 eeeee Papaver somniferum mi ee e eeee16 8 cf. Rosa sp. mi ee e eeeee2 Wild plants Ruderal plants, garden weeds and field weeds Aethusa cynapium w ee e 2 e 11ee Agrostemma githago mi 1 eeeeeee1 Agrostemma githago (fragment) w ee e eeeee þ Anagallis arvensis w ee e 2 eee8 e Anthemis cotula w ee e 12 1 eeee cf. Anthemis cotula mi ee e eeeee4 Bromus secalinus ch ee e ee1 eee Chenopodium album w e 716e 3 3 16 e Chenopodium ficifolium w ee 9 49 e 9 11 ee Chenopodium polyspermum w ee e 18 2 14 23 16 e Chenopodium murale w ee e 2 ee1 ee Cirsium arvense w ee e 1 eeeee Cyperus fuscus w ee e 2 eeeee Echinochloa crus-galli w ee e 2 eeeee Euphorbia helioscopia w e 1 e 2 eeeee Fallopia convolvulus w ee e ee1 eee Fumaria officinalis w ee e 3 eeeee Galium aparine (fragment) w ee e eeeee þ Hyoscyamus niger w ee e eeee1 e Lamium purpureum w ee e 17 1 eeee Malva sylverstris w ee e 1 eeeee Malva sp. vk ee e eeeee2 Y. Devos et al. / Quaternary International 460 (2017) 175e197 185

Table 8 (continued )

Site Poor clares Petite Rue des Bouchers Rue des Pierres

Unit 412 194 203 147 144 141e143 140 78 28

Dating (century AD) 14the15th <10the11th <10the11th 10the11th 11the12th 11the12th 11the12th 15the16th 15the16th

Volume (L) 15.1 1 1 1 1 1 1 11 15.8

Raphanus raphanistrum (fruit fragment) w ee e eeeee þ Reseda luteola w ee e ee1 1 16 e Rumex acetosella w e 314eeeee Rumex acetosella mi ee e eeeee4 Scleranthus annuus w ee e 2 eeeee Setaria pumila w ee e 4 eeeee Solanum nigrum w ee e 14 ee1 ee Sonchus asper w ee 1 19 eeeee Sonchus oleraceus w ee e 37 1 eeee Spergula arvensis w ee e 4 eeeee Stellaria media w ee e 36 2 eeee Thlaspi arvense w ee e 11 eeeee Urtica urens w ee e 32 eeeee Urtica urens mi ee e eeee16 20 Vicia hirsuta ch ee e eeee1 e Vicia tetrasperma ch ee e eeee1 e Vicia hirsuta/tetrasperma w ee e 1 eeeee

Nitrophilous fringes Conium maculatum w ee 11e 1 e 22 Galeopsis bifida/speciosa/tetrahit w ee e 5 eeeee Glechoma hederacea w e 1 e 4 eeeee Lamium album mi ee e eeeee8 Lamium album w e 1 eeeeeee Lamium cf. album w ee e eeee32 e Lapsana communis w ee e 18 eeeee Sambucus ebulus w/mi 4 2 1 ee42 53 104 4 Urtica dioica w e 70 73 329 23 14 15 ee Urtica dioica mi ee e eeee8 e cf. Urtica dioica mi ee e eeeee4

Trodden places Coronopus squamatus w ee e 2 eeeee Capsella bursa-pastoris w e 1 e 5 eeeee Plantago major w ee e 21eeee Poa annua w ee e 2 eeeee Polygonum aviculare w ee 1 10 2 eeee Polygonum aviculare mi ee e eeeee6 Sagina procumbens w ee 1 eeeeee

Grassland Phleum pratense w ee e 2 eeeee Phleum pratense ch ee e ee21ee Prunella vulgaris w ee e 2 eeeee Prunella vulgaris mi ee e eeee1 e Ranunculus acris/bulbosus/repens w ee e eeeeee Stellaria graminea w ee e 5 eeeee Verbena officinalis w ee e 7 eeeee

Wetland Alnus sp. ch 1 eeeeeeee Alnus sp. w e 11eeeeee Apium cf. nodiflorum w e 11e 2 eeee Apium nodiflorum w e 1 e 2 eeeee Batrachium sp. w ee e 17 eeeee Berula erecta w ee e 3 eeeee Bidens cernua w ee e 6 eeeee Bidens tripartita w ee e 3 eeeee Bidens sp. w ee e 19 eeeee Cyperus fuscus w e 1 e 2 ee1 ee Chenopodium rubrum w ee e eee1 ee Eleocharis palustris w e 1 eeeeeee Glyceria fluitans w ee e 9 eeeee Glyceria notata w ee e 4 eeeee Juncus sp. w ee e eee94 þþþþ Lycopus europaeus w e 2 1 59 eeeee Nasturtium officinale w ee e 12 eeeee Mentha aquatica/arvensis w e 11 6 e 1 ee8 e Polygonum hydropiper w ee e 91 eeeee Polygonum lapathifolium w e 1 e 83 eee8 e Polygonum mite w e 1 eeeeeee (continued on next page) 186 Y. Devos et al. / Quaternary International 460 (2017) 175e197

Table 8 (continued )

Site Poor clares Petite Rue des Bouchers Rue des Pierres

Unit 412 194 203 147 144 141e143 140 78 28

Dating (century AD) 14the15th <10the11th <10the11th 10the11th 11the12th 11the12th 11the12th 15the16th 15the16th

Volume (L) 15.1 1 1 1 1 1 1 11 15.8

Ranunculus flammula w ee e 5 eeeee Ranunculus sardous w ee e 3 eeeee Ranunculus sceleratus w ee e 588 1 eeee Rumex obtusifolius w ee e 2 eeeee Rumex crispus/obtusifolius w ee 3 10 eeeee Rumex crispus/obtusifolius mi ee e eeee1 e Salix sp. w ee e 5 eeeee Scirpus sylvaticus w ee 1 eeeeee Scirpus tabernaemontani w ee 1 eeeeee Veronica cf. anagallis-aquatica w e 1 e eeeee þþ Other wild plants (not determined to species level) w e 24 18 301 5 1 4 87 16 ch ee e eee1 20 18 mi 1 eeeeee61 65

Other taxa (not determined to species level) w e 25 15 15 5 eeee ch 4 eeeee1 16 69 mi 86 eeeeee37 208

Fig. 5. Similarity-indices (in %). These SI-indices indicate to which extent the particle size distribution of two samples are identical. Low values indicate significant differences between samples. underlying materials (Unit 177). Above this we observed a series of botanical remains over large distances in alluvial sediments should units of very fine organic sands and silts with rounded reworked be considered (Hall and Huntley, 2007), a large part of the identi- aggregates of underlying materials (Unit 194). The organic fraction fied species represents most likely the extra-local vegetation. Both contained fragments of vegetal organs, tissue fragments, and roots the palynological and plant macrofossil remains fit well to an open (Table 3) ranging in size from 20 mm to a few millimetres, and in wetland landscape, with marshes, wet meadows and patches of quantity from ca. 20% to 40e50% (visual estimate). The plant trees and shrubs characteristic for alder carr (Table 8, Fig. 7). The macrofossil assemblages from Unit 194 consisted exclusively of most abundant macrobotanical remains were seeds of rushes material preserved due to waterlogging. In addition to seeds and (Juncus spp.) observed in the finest fraction. Remains of common fruits, fragments of wood, leaves, mosses, buds, and plant stems nettle (Urtica dioica) suggest zones with tall herb vegetation on were also found. Despite the rather good preservation of the nutrient rich and moist soils. Some fruit fragments of alder (Alnus organic remains, the poor pollen preservation indicates that the sp.) and the predominance of alder in the arboreal pollen, indicate waterlogging was not permanent. The low diversity of the pollen that these trees were growing in the local environment or along the taxa and the high percentage of rather corroded Cerealia- and Tilia- river's borders. The same applies to hazel, of which some fruit scale pollen suggest some selective preservation and physical reworking fragments were found and which is the predominant species in the of the pollen (Fig. 7). Phytoliths are extremely rare in these units recovered wood fragments. Subsequently, the elderberry (Sambu- (Table 4). cus nigra), blackberry (Rubus fruticosus), and raspberry (Rubus The sedimentary characteristics outlined above (texture, sort- idaeus) seeds can also be interpreted as remains of shrubs growing ing, presence of reworked aggregates and of organic detritus) in this alluvial environment. The high percentage of Cerealia in the indicate that the lower part of the sequence is of alluvial origin. pollen diagram indicates cereal cultivation taking place in the These deposits can be interpreted as the outcome of flooding events proximity of the site. Anthropogenic influence is also suggested by in a wetland area with input of detrital sediments eroded along the the presence of seeds of weeds and ruderals (Chenopodium album, valley margins (see Bos et al., 2012; Ismail-Meyer et al., 2013; Euphorbia helioscopia, Rumex acetosella) and trodden plants (Cap- Furlanetto et al., 2014). Although the potential transportation of sella bursa-pastoris). Y. Devos et al. / Quaternary International 460 (2017) 175e197 187

Fig. 6. Morphotypological analysis of the isolated phytoliths according to a revised version of Runge (1999) inventory . A: details the relative frequencies of the three majors classes: Elongate (A), Grass phytoliths (G) and Non-identified (NI); B: details the relative frequencies of the phytoliths composing the Poaceae record (G). G1.1: trapeziform and rondel; G1.2: lobate; G1.3: saddle; G2: bulliforms; G3: Hair cell phytoliths; G 4: Long cell phytolith; G5: Dentriforms.

Fig. 7. Percentage pollen diagram from the studied stratigraphic units from the site of Petite Rue des Bouchers (Brussels). Pollen types with very low abundance are indicated with dots (i.e. expressed as present).

4.1.2. The earliest traces of human activities (<10the11th century becoming more organic silts (Unit 147). The upper part of the AD) organic silts and sands of Unit 147 showed a series of convolutions On top of these alluvial deposits we observed very fine organic (Unit 148) (Figs. 2b and 8a), very evident at the field scale. Their sands with frequent vegetal detritus fragments (Unit 203), light yellowish colour contrasted with the surrounding dark grey 188 Y. Devos et al. / Quaternary International 460 (2017) 175e197

Fig. 8. Petite Rue des Bouchers (Brussels), photomicrographs. a: Unit 144/147: convolution of organic-rich material protruding into a predominantly sandy area, as a result of puddling or poaching (PPL); b: Unit 144: Fe-substituted band of organic material (PPL); c: Unit 140: charred grain/seed (PPL); d: Unit 142: ceramic material fragment. Note darkened/rubefied glauconite grains as part of the mineral temper (PPL). matrix of Unit 147. They were often bordered by iron (hydr-)oxide miliaceum, Triticum spelta and Secale cereale) can be interpreted bands and crusts (Fig. 8b). In thin section, one such feature showed likewise. The remains of flax (Linum usitatissimum) and hemp an abrupt boundary with the surrounding material, and was (Cannabis sativa) suggest that these oil-and fibre plants were pro- composed of well sorted fine and medium sands with bands of silt cessed in the wetlands during the earliest occupation phase. It is also and very fine/fine sand giving rise to convoluted laminations. There from this level upwards that evidence of domestic waste (charred was no fine material in the inter-granular space. These macro- and cereal grains of Hordeum sp., S.ecale cereale, Triticum aestivum/durum/ microscopic features suggest processes of poaching, e.g. trampling turgidum and Triticum sp.) was recorded. on a soaking wet surface in the basal part of the sequence (see Langohr, 1991; Fechner et al., 2004). The preservation of these 4.1.3. The Dark Earth units (11the12th century AD) features suggests a rapid burial. In the absence of clear markers of The poached zone was overlain by a series of dark grayish units human impact like artefacts, it seems plausible to ascribe these (144, 143, 142, 141, 140), with the same sandy loam texture and traces to animals, poaching/puddling in a drinking area. Animal lithological composition. Unit 139 belongs to the same series but it passage is further indicated by the presence of spores of the displayed a slightly finer texture. Despite the limited bioturbation, coprophilous (herbivore dung) fungus Sordaria and parasite eggs in no recognizable sedimentary structures pointing to fluvial sedi- this unit, in combination with Plantago pollen (Unit 203), the mentary processes (grading, laminations) were observed within recording of Rubus, Chenopodium, Rumex taxa as well as seeds from the Dark Earth. The chaotic disposition of anthropogenic inclusions plants resistant to trampling (Units 147/203) (Table 8). (bone, ceramic material fragments, charred seeds, etc. eFig. 8c and The wet character of the environment was further indicated by d), sometimes showing rounded edges, seems not related to bio- the presence of vegetation types typical for marshes and riparian turbation and can thus point to downslope transport. Moreover, the vegetation. The palynological data (Unit 203) pointed to local pres- phytolith signature, based on the analysis of soil thin sections, was ence of riparian woodland (Alnus, Salix, Sambucus) and non-arboreal characterized by a clear dominance of the isolated distribution vegetation consisting of shore plants like those occurring in the pattern and rare clusters (Table 4) suggesting either that the plant transition between wetland and terrestrial environment (e.g. Men- remains decomposed before entering the soil matrix or that post- tha-type, Persicaria amphibia, Lythrum salicaria, Cyperaceae, Equise- depositional disturbance occurred (Vrydaghs et al., 2016). The tum, Polypodium-type), while aquatic plants were nearly absent. In phytolith assemblages, based on the analysis of soil thin sections Unit 147 high densities of well-preserved waterlogged seeds and (Fig. 6), revealed an anomalously high percentage of non-identified fruits (more than 2000 remains/L) were observed. Remains of phytoliths (10e30%) for Brussels. Only a part of the inability to wetland plants represented almost half of the total counted plant identify the phytoliths (15e50%) was due to their orientation in the macrofossils. They consisted largely of pioneer species from wet and thin sections. Material extracted from bulk samples (Units 139 and nitrate rich environments with Ranunculus sceleratus and Persicaria 140) showed high phytolith concentrations with very little etching, hydropiper as dominant species. These species are characteristic of indicating good preservation conditions. Nevertheless, the identi- river valleys and grow often in swampy parts of pastures, especially fication of fragments of phytoliths in relatively high frequencies where the soil is trampled (Weeda et al., 2000e2005). Possibly some supports the idea of disturbance or transport. Taking also into ac- of the wetland and grassland remains entered the contexts with count the footslope position of the site, it can be proposed that herbivore dung. The few cereal chaff remains we observed (Panicum these units are the result of colluvial processes (the term Y. Devos et al. / Quaternary International 460 (2017) 175e197 189

“colluvium” is meant here as “the product of slope wash” e see resistant pollen), although in sample 141 pollen of sedges Starkel, 1987), possibly with occasional input of alluvial sediments. (Cyperaceae), which are more vulnerable to corrosion have been It is not possible to distinguish with certainty between fully found. This combination of rather corroded and well-preserved colluvial or fully alluvial input of sediments, as water transport (e.g. pollen grains suggests a mixed character of the pollen assem- surface runoff, laminar erosion) is part of colluvial processes (see blages from redeposited (anthropogenic) and primary deposi- Bertran and Texier, 1997). The position of the site in the valley tion (wetland) origins. bottom, and the presence of intercalated organic sediments in the (e) The presence of diatoms, although in low numbers, which natural part of the sequence (see above) might also imply that might be indicative of some wet conditions (Table 6). periodic inputs of alluvium took place but no direct traces like flood coatings have been observed (Brammer, 1971; Nicosia et al., 2012). The periodically wet environmental conditions described above The archaeobotanical records (phytoliths, pollen and macro- change towards the top of the sequence (Unit 139), and culminate fossils) indicate open herb vegetation, both grasslands and zones with a phase of topsoil formation. Starting with Unit 140, a weak with nitrophilous tall herb plants as well as wetlands. The phytolith but an increasing signature of C4 grasses of the Panicoidaee and record is dominated by C3 grasses of the Pooideae subfamily Chloridoideae subfamilies was detected (Fig. 6). (Fig. 6). Rare phytoliths markers for cultivated crops (such as den- In Unit 139 we observed a peak in biogenic porosity, in fine dritics (ICPN Working Group et al., 2005)) were also observed. organic matter and micro-charcoal. Bioturbation was also attested Whereas in situ crop cultivation has been identified on the sites of by the presence of bow-like features or infillings with crescentic the Court of Hoogstraeten and Rue de Dinant (Devos et al., 2009, fabric (see Stoops, 2003). Topsoil disturbance, linked more probably 2013a), the almost total absence of articulated dendritics in- to a bare surface (e.g. an urban grassland soil), rather than to a tilled dicates here that their deposition did not result from in situ or ploughed topsoil (see 4.2.2 and references therein), was likely x decomposition of plant material and thus do not constitute a local responsible for the enrichment in dusty clays dispersed in the inter- record (Vrydaghs et al., 2016). The pollen and macrofossils add granular space as opposed to well-delimited textural pedofeatures. taxonomic diversity to the image of the vegetation. The macro- Micro-charcoal could nevertheless also have derived from the botanical and palynological records show a strong anthropogenic erosional or colluvial processes that put in place the parent signal as well, and evidence of domestic waste (charred cereal material. grains) was found. In the palynological record this was indicated by pollen of Cerealia-type and weeds occurring in cultivated fields 4.2. The site of the Poor Clares (like Centaurea cyanus and Scleranthus annuus). Such a flora is consistent with all the ecological phases corresponding to the 4.2.1. The parent material alternation of cultivation and fallow. Domestic waste was indicated The parent material (Units 414e416) has a clay to clay loam by carbonized cereal grains (Avena sp., Hordeum vulgare, Secale texture (Table 1). Traces of mottling point to a fluctuating water cereale, Triticum aestivum/durum/turgidum and Triticum sp.) and table (Fig. 4). Micromorphological observations noted that cereal weeds (Bromus secalinus). anthropogenic materials and phytoliths were rare (Tables 3B and 4) Human activity at the time of Dark Earth formation thus took and probably intrusive. A coring carried out showed the presence of place in a periodically wet environment, with water stagnating a thin layer of peat below this clayey sediment, thus confirming the close to the surface. This is further supported by: sequence proposed by the geotechnical map (Dam et al., 1977). As such, Units 414e416 can be considered as a ‘natural’ alluvial sedi- (a) The morphology and distribution of iron pedofeatures, with ment, almost unaffected by human activities. Indeed, the peat layer coatings, hypo- and quasi-coatings in the top (saturated) part of observed below Units 414e416 provided us the image of a densely the profile (Unit 139), grading to nodules in the interface zone forested area (89,6% arboreal pollen). Alder (Alnus glutinosa 50.9%) (Unit 140) and disappearing in the lower part (Lindbo et al., associated with ash (Fraxinus) and birch (Betula) dominated the 2010; see also; Stoops and Eswaran, 1985). The morphology of tree pollen assemblage (Table e Electronic Supplementary phosphatic neoformations could possibly reflect the redox- Content), characterising a riparian forest environment that line- imorphic processes outlined above, but due to the lack of any arly developed along the Senne River. A large amount of wood re- indicator of faecal matter in primary depositional setting, we mains typical of alder and birch (scalariform plates) indicates that cannot exclude post-depositional phosphate contamination these species were growing in the direct vicinity of the site. The from other portions of the sequence or from leakages (e.g. septic presence of Ferns (including sciaphilous Polypodium vulgare) con- tanks), a common occurrence in urban settings (see Macphail firms this forest ambiance, whereas the quite high frequencies of and Linderholm, 2004). This is especially possible as remains freshwater algae observed (Chrysophyceae, Spirogyra, HdV-128) of a stable were found within 1 m of the studied soil profile highlight the wet conditions and thus the proximity of the river. (Claes, 2011). At the rear of the riparian woodland, a hardwood alluvial forest (b) The plant macrofossils from the Dark Earth having lower composed of lime tree (Tilia), hazel (Corylus) and oak (Quercus) densities and being less diverse compared to the other contexts developed. The non-arboreal pollen assemblage was rather poorly from the same site. In this regard we point also to the domi- diversified with large ecological valence taxa including species nance of resistant seeds such as elderberry (Sambucus nigra, from wetlands and more or less clear undergrowth being observed. Sambucus sp.) and dwarf elder (Sambucus ebulus), probably Although recorded in low amounts, the other pollen taxa observed indicating a poor preservation of organic material (Murphy and suggest the presence of edges or small clearings: heliophilous/ Wiltshire, 1994). xerophilic shrub species, pre-forest species or species belonging to (c) The presence of species typical for wet environments such as oak-forest degradation facies (Rosaceae, Lonicera, Juniperus/ Apium cf. nodiflorum, Ranunculus sceleratus, Cyperus fuscus and Cupressus, Pinus, Ericaceae) were observed. Established on the less Sambucus ebulus, which is often found in alluvial environments fertile sandy and drier soils, these taxa indicate some already with nitrogen-rich, humic, and moist soils (Weeda et al., relatively deforested areas not far from the site. The presence of 2000e2005). some ruderal and nitrophilous plants (Chenopodiaceae, Artemisia, (d) The rather bad preservation of the pollen, which seems to be Rumex, Plantago lanceolata type, Ranunculaceae), as well as few also partly selective (indicated by overrepresentation of more spores of the coprophilous fungus Sordaria and micro-charcoal, also 190 Y. Devos et al. / Quaternary International 460 (2017) 175e197 suggest some pastoral practices in the surroundings. However, The humic character of the unit (see Tables 2 and 3B), in com- these human activities should have been rather limited and/or bination with the presence of microscopic traces of bioturbation distant from the site of Poor Clares given the low percentages these (passage features, organo-mineral excremental micro-aggregates) herbaceous taxa were observed. (see Table 3B) indicate that unit 415 is an ancient topsoil (Gerasimova and Lebedeva-Verba, 2010). Multiple dusty clay coat- ings and wavy lenses of dusty (probably humus rich) clay and silt 4.2.2. The Dark Earth units (12the15th century AD) were observed, pointing to a (temporarily) bare and/or disturbed An abrupt and slightly undulating boundary was observed be- surface, possibly affected by agricultural practices (Courty et al., tween the parent material and the overlying Dark Earth. The lower 1989; Macphail et al., 1990; Carter and Davidson, 1998; Lewis, part of this Dark Earth (Unit 415) was composed of clay loam, while 2012). The identification of agricultural practices in the absence the upper part (Unit 412) showed a lighter sandy loam to sandy silt of clear macroscopic evidence (ploughmarks, cultivation ridges, loam texture. etc.) has sometimes proven to be particularly difficult, as many The Dark Earth (Unit 415) and the underlying substrate (Unit micromorphological features that are often linked to agriculture 414) were composed of different sediments (a similarity index of can also derive from other processes (see Carter and Davidson, 75%; see Fig. 5B). However, by omitting the medium sand fraction, 1998; Goldberg and Macphail, 2006; Lewis, 2012, Table 5; Deak Unit 415 and 414 showed a very high textural similarity (>90%) et al., in press). However, several studies have shown that a set of (Fig. 5C). This indicates that Unit 415 is probably a mixture of Unit clear micromorphological observations, ideally in combination 414 with sandy sediments. One potential source are Brusselian with other lines of evidence, can rule out alternative hypotheses Tertiary sands, cropping out on the steep valley flanks (Dam et al., (see for example Goldberg and Macphail, 2006; Devos et al., 2009; 1977). As such sediments are not present in the immediate vicin- Lewis, 2012). As such, it has been possible to identify medieval ity of the site, they have probably been brought in intentionally. agricultural practices on a series of sites in the higher part of the Taking into account the low overall textural similarity, the input city (Devos et al., 2009, 2011c, 2013a). must have been rather substantial. The macroscopic homogeneity Moreover, the combined presence of excrements, secondary of Unit 415 implies that the two sediments have been thoroughly phosphates, bone fragments, charcoal, and potsherds indicates that mixed. This is also suggested by a series of micromorphological the soil has been enriched with domestic waste and manure observations: an unsorted coarse mineral fraction, randomly (Pieters, 1990; Bakels, 1997; Simpson et al., 2005; Macphail and distributed and unoriented anthropogenic elements (charcoal, Goldberg, 2010). The chemical data (Table 2) also point in this di- potsherds, bone fragments, fishbones, etc.), the presence of soil rection. Enhanced values of phosphorus and organic matter were fragments originating from Unit 414 (Fig. 9a) (see for example observed. As stated earlier, there seems to be a substantial input of Gebhardt, 1995), and the presence of passage features and organo- coarser sediments. This might be related to the improvement of a mineral excremental pedofeatures. Also, of all the phytoliths having “heavy” soil for agricultural processes (see for example: Allen, altered morphologies (n 130), 83% were attributed to physical 1852). ¼ damage resulting in either fragmented phytoliths or being The common iron nodules observed point to a fluctuating water morphologically non-identifiable. Two agents seem to be mainly table and thus alternating reducing and oxidising conditions responsible for the homogenisation: biological and human activity. hampering a good preservation of the organic remains. This is

Fig. 9. Poor Clares (Brussels), photomicrographs. a: Unit 415: dense soil fragment probably originating from Unit 414, witnessing the mixing of the soil (PPL); b: Unit 415: articulated dendritic phytoliths. c: fragmented coprolite. Note the presence of articulated phytoliths (arrow) (diameter 2 mm; PPL); d: A: fragmented eggshell, B: calcitic biospheroid (PPL). ¼ Y. Devos et al. / Quaternary International 460 (2017) 175e197 191 confirmed by the plant macrofossils study of unit 415 that had poor decomposition that was periodically waterlogged (McCobb preservation of uncharred plant remains and pollen (Houchin, et al., 2001; Jacomet, 2007). A large part of these mineralized 2009). This can be expected for an agricultural soil with a rather plant macrofossils (43%) was too badly preserved to allow fast decomposition of the organic remains, as reflected by the low identification. The identified mineralised seeds and fruits con- C/N ratio. However, the phytolith study on thin sections provided sisted mainly of digestible fruit remains, including strawberry additional botanical data. First, a distinction could be made be- (Fragaria vesca), fig(Ficus carica) and grapevine (Vitis vinifera) tween phytoliths within excremental remains, those associated seeds. The few identified mineralized kitchen herbs (Foeniculum with burned material, and those within the soil matrix. Articulated vulgare), cereal bran fragments, fragmented pulses (Vicia faba) phytoliths had their highest percentage within the soil matrix of and a corn cockle (Agrostemma githago) fragment together with Unit 415 (Table 4; Fig. 9b), pointing to in situ decomposition of plant some mineralized textile fragments and fly puparia also can be material (see Vrydaghs et al., 2016). Their morphotypological interpreted as typical elements of cesspits (Smith, 2013). Several analysis revealed a predominance of phytoliths deriving from amorphous mineralized concretions of organic material, some- cultivated crops (dendritics sensu IWG Phytolith Nomenclature, times containing unidentifiable plant remains and remains of fly 2005). Morphometric measurements of the wave lobes of these puparia, were found as well. These types of remains are often articulated dendritics fell within the range of Hordeum vulgare, found in poorly preserved latrine fills (Hall and Huntley, 2007). Triticum aestivum, T. dicoccon and Avena sativa, tending to support As mentioned previously, part of the observed ashes might have the notion this material indeed derived from cultivated crops, been used for hygienic purposes. This has for example been although comprehensive data on the wave lobes from wild grasses demonstrated for ancient turf houses in Iceland, where ash was (such as Hordeum murinum) are so far not available for comparison. put on the floor to repair it and to get rid of bad odours (Milek, However, macrobotanical and palynological evidence for Cerealia 2012). Another example is the so-called nightsoil, the practice of within archaeological units dating back to the 13th century AD putting a layer of ash on top of excremental waste to limit bad supports the presence of cultivated crops. Non-articulated den- odours (Richard Macphail, University College London, personal dritic phytoliths were also observed within the soil matrix. communication). The presence of aggregates composed of Although their origin cannot be established with certainty (see coprolitic material and ash point in this direction. Vrydaghs et al., 2016), their disarticulation is probably due to - Artisanal waste. This is witnessed by enhanced levels of Cu and physical reworking. Pb (Table 2) and the presence of glass like silica pointing to high Slightly elevated levels of Cu and Pb were measured (Table 2), temperature heating (Courty et al., 1989; Cammas, 2010; possibly pointing to some metallurgical activities taking place in Viklund et al., 2013). proximity to the site. This is confirmed by the presence of metal- - Construction and/or destruction debris. Burnt soil fragments, lurgical slag and droplets (Table 3). limestone, mortar, and plaster clearly indicate the input of The very sharp boundary observed between Unit 412 and 415 building material. indicates a new depositional phase. Compositional (Unit 412 pos- sessing a very calcareous matrix, while the matrix of Unit 415 is Although very rich in anthropogenic remains, this upper part of non-calcareous) and textural (e.g. a low S.I.: Fig. 5) differences be- the Dark Earth cannot be identified as one single dump of waste nor tween both units confirm this hypothesis. The anomalously as a simple manure heap. The homogeneity indicates that the rate calcareous matrix seems to be due to the input of a series of of accretion was slower than that of soil homogenisation. One of the calcareous materials. Different types of calcareous building mate- agents responsible for the latter process is biological activity, as rials were observed (limestone, mortar, and plaster), together with witnessed in thin sections by the omnipresence of organo-mineral frequent shell fragments, and ash. Unit 412 was particularly rich in excremental aggregates and by earthworm excreta such as calcite anthropogenic remains (Table 3B). At least four origins can be biospheroids (see Canti, 2003a, Fig. 9d). The higher frequency of identified: isolated phytoliths with respect to Unit 415 (Table 4) could result from a more intense physical reworking deriving from gardening - Domestic (kitchen) waste. The combined presence of (burnt) activities compared to a ploughed field. A similar observation was bone, ceramics, glass, eggshell, charcoal, and ashes points in this done on the 15the16th century garden on the site of the Court of direction. We noted also the presence of fish-bone and shell. In Hoogstraeten (Devos et al., 2013a). Whether this Dark Earth can be this respect we also found charred cereal grains (Secale cereale, interpreted as a garden soil remains hard to prove since many Triticum aestivum/durum/turgidum), together with a grape (Vitis traces of early construction on this particular part of the site have vinifera) seed, a fruit of beet (Beta vulgaris) and a seed fragment been disturbed by the 17th century expansion of the Poor Clares of lentil (Lens culinaris). Also one charred fruit fragment of alder monastery. Moreover, the poor preservation of remains of wild (Alnus sp.) was found. These charred remains may have been plants did not allow us to reconstruct the local vegetation. deposited together with the frequent charcoal fragments observed as waste from hearths, also indicated by the presence 4.3. The site of Rue des Pierres of ashes. Although part of them might have been used for hy- gienic purposes (see below). 4.3.1. The parent material - Excremental waste. The presence of coprolites (Fig. 9c) directly A series of corings carried out showed the presence of grey clay identifies excremental waste. Variations within their phytolith loam on top of a thin layer of peat, sands and gravel, thus con- content (Table 5) suggest various origins. Secondary phosphatic firming the sequence of alluvial deposits proposed by the nodules and enhanced phosphorus levels constitute indirect geotechnical map (Dam et al., 1977). The palynological study evidence. The majority of the reported plant macrofossils are (Table e Electronic Supplementary Content) covers the lower most mineralized (83% of the total remains counted). Such remains part (45e50 cm) from coring nr. 4. The three samples containing are formed when organic matter is partly or completely pollen showed a rather homogenous composition. All of them were replaced by salts deposited around the plant cells or in inner dominated by non-arboreal pollen, mainly grasses and cultivated voids. One of the most common forms of mineralization of cereals as well as plants coming from ruderal and wet habitats. This archaeobotanical material is caused by calcium phosphate de- indicates that the pollen assemblages come from rather open posits from faecal deposits and organic material in disturbed landscape and cultivated fields, with some contribution 192 Y. Devos et al. / Quaternary International 460 (2017) 175e197 of the local wetland vegetation. It is also quite plausible that the within coprolites (Fig. 10b; Table 5). Possible weathered excre- pollen assemblages partly reflect human activities from the sur- ments were observed as rounded and sub-rounded fragments rounding site involving dung or manure being added to fields and with sharp edges, as irregularly-shaped aggregates, and as which would also explain the high content of Cerealia-pollen. deformed and fragmented “sludge.” The presence of aggregates composed of charcoal fragments within a phosphatic matrix 4.3.2. The Dark Earth units (13the16th century AD) might again indicate the use of nightsoil (see Goldberg and All the Dark Earth units (28, 74, 75, 76, 77, 78, 109, 119, 120 and Macphail, 2006). 121) are characterised by a clay loam to sandy silt loam texture. Based on soil micromorphology it was possible to differentiate The high input of faecal material is also indicated by the car- between to types of Dark Earth units. The first group corresponds to pological data: 79% of the counted plant macrofossils were fruit Units 78, 77, 74 in Profile B, and Units 121, 120, 119 in Profile A. The remains, consisting mainly of elderberry (Sambucus nigra), black- presence of dusty clay coatings (Table 3C), in combination with the berry (Rubus fruticosus) and strawberry (Fragaria vesca) seeds. As channel microstructure, suggests that these units are a succession remains of sweet/sour cherry (Prunus avium/cerasus), grape (Vitis of unprotected surfaces (Courty et al., 1989; Stoops et al., 2010). The vinifera), Maloideae and Prunoideae are also found but less limited impact of biological activity, as demonstrated by the pres- frequently, these findings can be interpreted as consumption re- ence of a dominant (sub-) horizontal orientation of inclusions like mains. The same applies to mineralized poppy (Papaver somnife- charcoal, indicates that the accumulation rate must have been fast. rum) seeds, which together with small fruit seeds (strawberry, Although the organic matter is predominantly present in the form raspberry (Rubus idaeus), and blackberry (Rubus fruticosus)) pass of humified material, the high initial input can be inferred by the easily through the digestive system, and are typical cesspit fill high concentration of phytoliths and the abundant phosphatic components (Smith, 2013). plant pseudomorphs. Despite macroscopic homogeneity, the micromorphological - Domestic (kitchen) waste observed included molten phytoliths study showed the presence of heterogeneous aggregates indicating and wood charcoal, commonly found in hearths (see Canti, the fast accretion of different kinds of anthropogenic components: 2003b), bones and pottery fragments. In this respect we also observed charred cereal grains of rye (Secale cereale), naked - Abundant faecal material as reflected by the extremely high wheat (Triticum aestivum/durum/turgidum), oats (Avena sp.) and phosphorus concentrations, comparable to the ones observed in barley (Hordeum vulgare) amidst the macroremains (1.5% of the latrines elsewhere in town. The faecal material is of carnivore- counted plant macrofossils). These are the most frequently omnivore origin, and therefore can be attributed to dogs, pigs, found cereal species in medieval contexts in Brussels (Speleers and/or humans (see Macphail, 2000; Macphail and Goldberg, and van der Valk, in press) and in the wider region (see for 2010). These contain silt-sized mineral grains, Chrysophyceae example De Groote et al., 2009). The very rare igneous volcanic cysts, seed coats, charcoal, dark brown (PPL) organic material rock fragments might be microfragments of a grinding stone. and, in some cases, bone fragments (Fig. 10a), diatoms and - By-products of cereal processing. Whereas the coprolites were pollen. Rare isolated, clustered and articulated phytoliths rather poor in phytoliths, the soil matrix itself was very rich. The deriving from grasses and cultivated crops were observed latter records the presence of bract remains of cereals. The

Fig. 10. Rue des Pierres (Brussels), photomicrographs. a: Unit 74: coprolite of carnivore-omnivore, as witnessed by the presence of bone fragments (PPL); b: Unit 74: articulated phytoliths (arrow) inside a coprolite (PPL); c: Unit 77: reworked soil fragment (brickearth? Result of the digging of underlying horizons?). Note illuvial clay-enriched area (XPL); d: Unit 10: weathered mortar, with release of medium sand-sized grains in the groundmass (XPL). Y. Devos et al. / Quaternary International 460 (2017) 175e197 193

morphometric measurements of the wave lobes of the articu- remains) and higher proportions of carbonized and mineral- lated dendritics fell within the range of several cereals: Triticum replaced seeds and fruits (16% and 23% of the total counted re- dicoccon, T. aestivum, Avena sativa and Secale cereale and can mains respectively) (see Table 8). Due to poor preservation one fifth thus be interpreted as the remains of cereal processing. of the total counted macrofossils could not be identified. Remains of - Waste deriving from artisanal activities was evidenced by the economic plants (88% of the total identified remains) consisted presence of metallurgical slags. Enhanced concentrations of Cu, mainly of charred cereal grains and resistant seeds of Sambucus Ni and Pb were observed (Table 2), but these might also be spp. and Rubus spp. Charred cereal grains were more abundantly related to the excremental waste. found here than in Unit 78, but the majority were too distorted or - Earth-based construction material (reworked aggregates of Bt fragmented to allow identification. Some mineralized seeds of fig horizon with non-deformed clay coatings, possibly “brickearth” (Ficus carica), poppy (Papaver somniferum) and Maloideae/Prunoi- - sensu Goldberg and Macphail, 2006 e Fig. 10c). In Unit 74 we deae probably should be interpreted as the remains from reworked observed the weathering (partial dissolution) of the calcareous excremental waste. Most likely due to poorer preservation condi- binding agent in mortar fragments. This process led to the tions of waterlogged remains, seeds of wild plants were observed release in the groundmass of medium sand-sized quartz grains less than in Unit 78 (12% of the total identified remains). The ma- (Fig. 10d). Therefore, part of the mineral material in this unit was jority (60% of the remains of wild plants) were not preserved well originally contained in mortar fragments that was weathered enough to be identified to species level and provide only limited or and dissolved, according to a mechanism of Dark Earth forma- no ecological information. The most counted remains were tion mentioned by Macphail (1994). mineralized seeds of Chenopodiaceae and Rumex sp. The identified species represent the same ecological groups as found in Unit 78 The phosphatic features observed, especially vivianite, can be (crop-fields/gardens, ruderal places, trodden places, nitrophilous attributed to periodic water-saturation (see Stoops, 1983; Karkanas fringes and wetland). and Goldberg, 2010). This is also reflected by the low concentration This second group of Dark Earth units can be interpreted as of plant macrofossils (for example in US 78: about 200 remains/ backfill made up by material resulting from the digging or cutting litre). Circa half of the preserved remains were waterlogged, 3% was through of underlying deposits. This possibly derives from “nega- carbonized, and 9% mineralized. 37% of the counted plant remains tive” (sensu Harris, 1989; see also Carver, 1987) features, such as consisted of resistant seeds of Sambucus spp. They are very resistant pits, robber trenches, postholes, quarries, etc. This second type of and can be preserved even in dry contexts where other seeds and unit also suggests a fast accretion of the sequence, except in some fruits have undergone chemical changes by charring or minerali- cases where very ephemeral periods of stability e marked by zation (Murphy and Wiltshire, 1994; Jacomet, 2003). The different stronger bioturbation and more pronounced oxygenation e were types of preservation do not necessarily indicate various de- observed (e.g. in Unit 76). positions of material of different origin. In urban contexts with thick stratigraphic sequences, that have not been permanently 5. Archaeological implications water-saturated, mineralized and waterlogged materials are often found together (Hall and Huntley, 2007). 5.1. Petite Rue des Bouchers The remains of wild plants (17% of the total counted remains) represent a mixture of different vegetation types including a wet The beginning of sedimentation is very likely related to alluvial environment as well as a nitrophilous one. Some charred seeds of dynamics, possibly draining through/into a wetland area with a crop-field weeds (Vicia hirsuta and V. tertrasperma) may have marsh, riparian vegetation, and wet meadows. This wet character entered the context with the identified charred cereal remains continues as witnessed by the traces of poaching (puddling) dating (Avena sp., Hordeum vulgare, Secale cereale, Triticum aestivum/ from the 10the11th century. The surroundings of the site were durum/turgidum, Triticum sp.). The uncharred crop-field and garden probably open pastures. On top of the 10the11th century surface a weed seeds identified may have been brought in with cultivated fast accreting sequence of colluvial deposits, possibly mixed with plants as well or they may have grown as pioneer plants on alluvium, were observed. These deposits dating from the reworked soil in the local area of the site. Other plant species found, 11the12th century, might be related to the presence of newly suggest nitrophilous fringes, ruderal, trodden places and wetlands cultivated soils along the steep slopes and on the Brusselian plateau in the vicinity of the site. Dwarf elder (Sambucus ebulus) was the (see Devos et al., 2007a, 2009, 2011c, 2013a), which also explains dominant species. As mentioned above, Sambucus spp. diaspores the presence of small amounts of domestic waste. have a better chance of preservation than most other plant remains The upper part of the Dark Earth (Unit 139) shows the devel- (see Table 8). opment of a topsoil horizon, indicating a diminishing of the ac- Based on the field, geoarchaeological and archaeobotanical cretion rate. The absence of markers of strong disturbance in it characteristics mentioned above, the first group of Dark Earth units suggests the development of a grassland soil. Anthropogenic can be interpreted as formed in open-air context with a fast ac- impact still seems to be low, as we did not observe building debris cretion of waste. The latter reflects the accumulation of different or artisanal waste in the colluvium. The latter rather contained kinds of waste in a periodic waterlogged environment. No traces of cereal phytoliths showing physical damage (due to transport) and post-depositional transformation, such as by agricultural practices small amounts of carbonized grains. This fits well with a part of the or trampling (see higher), were observed in them. city not yet fully urban in character, where colluvium, eroded from The second type of Dark Earth unit (Units 75, 76, 28 in Profile B; the cultivated parcels located upslope, accumulated. The picture Unit 109 in Profile A) was made up of organic-rich aggregates from botanical data is that of a rather open and wet environment. (similar to the units of the first type, described above) mixed with very frequent reworked soil aggregates. Compared to the first 5.2. Poor Clares group of units, we observed here a more pronounced oxygenation of the soil, which was again reflected in the preservation status of The pollen study of the peat layer indicates a rather closed the plant macrofossils: in Unit 28 the concentration of the remains forested environment, represented by a riparian woodland (prob- was only half as large as that in Unit 78 and the sample contained ably alder carr type) close to the site surrounded by hardwood lower proportions of waterlogged seeds (24% of the total counted forests. Limited signs of human activities can however be seen 194 Y. Devos et al. / Quaternary International 460 (2017) 175e197 pointing to pastoral practices in the surroundings. The study of the The second type of Dark Earths is more probably linked to the Dark Earth on the site of the Poor Clares demonstrates that its accumulation of backfill, in relation to the construction/destruction formation is not the result of one single anthropogenic activity, but of insubstantial buildings, to quarrying (e.g. digging away of the rather the outcome of different activity phases, combined with excremental waste), and/or to ground raising in this rather wet substantial bioturbation. environment bordering the Senne river. The rather wet conditions The oldest activities observed on the site are most likely agri- explain the excellent preservation of the excremental waste, evi- cultural, as remains of a cultivated clayey topsoil on which cereals dence that has largely disappeared on the sites in the higher part of were grown have been recognised (Unit 415). Obviously, the the city (see Devos et al., 2009, 2011a, 2013a). These wet conditions, cultivation of such heavy soils implied the use of adapted tools together with the high rate of accretion, probably also hampered (ploughs). People also had to deal with the risk of flooding. The fact bioturbation. However, the incomplete preservation of seeds and that they nevertheless cultivated this zone bordering the river fruits (low densities, overrepresentation of resistant remains, and might be related to the continuous search for new arable land mineralisation), in combination with the presence of vivianite, (Charruadas, 2011). As a matter of fact, so far all agricultural plots points to periodic drying of the deposits. identified in Brussels are situated in higher topographic positions on lighter-textured sandy and sandy loamy soils (Devos et al., 6. Conclusions 2007a, 2009, 2011c, 2013a). They all date to the 10the13th cen- tury AD, and seem to end with the construction of the first city wall. The study of the Dark Earth from these three alluvial sites clearly However, cultivation still went on in the area in between the two illustrates that an approach combining geoarchaeology and city walls, as witnessed by the presence of arable land on the site of archaeobotany is efficient in understanding the formation of de- Poor Clares. The few slag fragments point already to the proximity posits that in the field appear to be very similar. Different sequences of metal working activity. of events, anthropogenic activities, syn- and post-depositional Things change with the formation of the second Dark Earth processes were identified for each of the studied sites. This proves (Unit 412) in the 15th century. An important accumulation of sed- the necessity of studying Dark Earths on a case-by-case base (see iments significantly raised the surface level. The kind of activities Nicosia et al., 2013; Nicosia and Devos, 2014). By following such a taking place and/or the way people dealt with their waste on the case-by-case approach, we gained insight into the spatial organi- site and in its vicinity also clearly changed. The significant accu- zation of the Senne valley through time. Here, we observed the mulation of building debris (brick fragments, mortar, plaster, occurrence of a palimpsest of activities, such as agricultural prac- limestone, etc.) indicates building activities nearby. These obser- tices and waste management, which are rarely archaeologically vations correspond with the historical records suggesting that the recorded with certainty in urban contexts. area, situated northeast of a road leading in direction of the Senne, Through the analyses of samples originating from three was taken in by private housing by the end of the 14th and certainly different archaeological excavations, the urbanisation process of in the 15th century AD (Claes, 2008). The presence of artisanal the city of Brussels can be illustrated: waste and the more elevated levels of Cu and Pb also indicate an intensification of metallurgical activities taking place in the The Dark Earth of Petite Rue de Bouchers draws a picture of the  surroundings. pre-urban phase (11e12th century AD), before the construction On the site itself, we can witness the gradual formation of a very of the first city wall in a time period where human habitation is thick and porous soil horizon, with a high fertility. Such a soil might concentrated in some nuclei and where agricultural activities have been used as a small backyard plot that was occasionally used played an important role. The site and its immediate sur- as a garden. roundings were clearly not cultivated, but rather wet grassland used for pasture. On a broader scale, indirect evidence for 5.3. Rue des Pierres cultivation of the higher grounds is provided. The Dark Earth of Poor Clares (12the15th century AD) confirms  The pollen observed in the alluvial deposits show an environ- the presence of agricultural activities in the Senne valley ment that has already been modified by human activities. However, whereas it was only known previously to be taking place up- a soft- and hardwood alluvial forest is still present close to the site. slope (e.g. on the sites of Treurenberg, Rue de Dinant, impasse The study of the Dark Earth of Rue des Pierres suggests a fast ac- du Papier, Court of Hoogstraeten). Furthermore, it shows the cretion of waste and backfill in this site. All evidence points to an transition of a pre-urban occupation towards an urbanised accumulation of essentially excremental remains. This is especially neighbourhood, with a zone largely dedicated to agriculture true for the first type of Dark Earth, which possibly formed in an shifting towards a densely occupied zone with intense building area where waste from latrines and houses accumulated, and in and artisanal activities. which animals were present. This interpretation is in line with The Dark Earth of Rue des Pierres (13the16th century AD) re-  several other Dark Earth studies, which explain the latter as the flects an urbanised and densely populated environment, prob- outcome of middening in external areas around households and ably a zone where excremental waste accumulated, indicating abandoned buildings (see for example Macphail, 1994; Cammas, the importance of future research on the city's waste 2000; Fondrillon, 2007). management. In this respect we can refer to the presence of ‘rues immondes’, alleys where excremental waste accumulated in the Renaissance By integrating geoarchaeological and botanical data, a better (see Guillerme, 1983). Henne and Wauters (1845) report the pres- insight is provided in the origin and the taphonomical processes of ence of a so-called ‘Pisstraetken’ and a ‘stadsheimelicheit’ going the botanical remains, allowing to better separate the remains back to 15th century. Both toponyms point to loci where excre- related to the natural environment from those related to human mental waste accumulated in the neighbourhood. These data could activities. Although Dark Earth units are often believed to be poor thus fuel future research on the city's medieval waste management. contexts to study vegetal consumption, the site of Rue de Pierres, a The accumulation of construction and/or deconstruction debris place were excremental remains accumulated, provides us valuable (brick fragments, mortar, plaster, limestone, etc.) indicated ongoing information on the consumption of vegetal remains, in particular building activities nearby. on the importance of fruit consumption. Y. Devos et al. / Quaternary International 460 (2017) 175e197 195

Despite being situated in an initially alluvial setting, most of the Beug, H.-J., 2004. Leitfaden der Pollenbestimmung für Mitteleuropa und angren- Dark Earth units did not provide an excellent preservation of the zende Gebiete. Dr. Friedrich Pfeil Verlag, München. Bojnansky , V.T., Fargasova, A., 2007. Atlas of Seeds and Fruits of Central and East- organic botanical remains. Meanwhile, the integration of seeds, european Flora : the Carpathian Mountains Region. Springer, Dordrecht. fruits, pollen and phytolith data from several contexts provides Borderie, Q., 2011. L’espace urbain entre Antiquite et Moyen Age,^ analyse geo- either congruent data for several taxa or allows to expand the flora archeologique des terres noires, etude de cas. Universite de Paris 1 (Unpub- lished PhD dissertation). record. In combination with the geoarchaeological approach, it al- Borderie, Q., Devos, Y., Nicosia, C., Cammas, C., Macphail, R.I., 2015. Les terres noires lows to collect satisfying data to detail an alluvial landscape dans l’approche geoarch eologique des contexts urbains. In: Carcaud, N., affected by human activities and its evolution. Most often fragile Arnaud-Fassetta, G. (Eds.), La geoarch eologie française au XXIe siecle. CNRS editions, pp. 247e258. plant remains (spices and kitchen herbs for example) are decayed Bos, I.J., Busschers, F.S., Hoek, W.Z., 2012. Organic facies determination: a key for and more resistant remains overrepresented. Despite the poor understanding facies distribution in the basal peat layer of the Holocene Rhine- preservation, a lot of economic plants have been recorded, Meuse delta. The Netherlands Sedimentology 59, 679e703. Brammer, H., 1971. Coatings in seasonally flooded soils. Geoderma 6, 5e16. providing a partial image of the cultivated, collected, consumed and Butzer, K.W., 2008. Challenges for a cross-disciplinary geoarchaeology: the inter- processed economic plant species in the alluvial valley trough time. section between environmental history and geomorphology. Geomorphology The palaeobotanical assemblages suggest that the vegetation of 101, 402e411. ’ the alluvial plain during the earliest studied occupation phases was Cammas, C., 2000. Apports et perspectives de l analyse micromorphologique des “terres noires”. In: Terres Noires - 1. Documents Sciences de la Ville, vol. 6. already quite open and exposed to anthropogenic activities. Traces Maison des sciences de la ville, de l’ urbanisme et des paysages, Tours, of domestic waste were observed in the archaeobotanical record pp. 45e60. from the 10the11th century onwards. The natural vegetation con- Cammas, C., 2004. Les « terre noires » urbaines du Nord de la France : premiere typologie pedo-s edimentaire. In: Verslype, L., Brulet, R. (Eds.), Terres Noires e sisted of swamps, wet meadows and alder carr. In the later phases Dark Earth. Actes de la table-ronde internationale tenue a Louvain-la-Neuve, les the human impact is manifested by the development of pasture 09 et 10 novembre 2001. Universite Catholique de Louvain, Louvain-la-Neuve, lands, ruderal places and cultivated land. pp. 43e55. Cammas, C., 2010. Histoire microstratigraphique de structures d’habitat des îlots 1 ́ et 27 de la ville de Lattara (-475/-375). Lattara 21, 289e324. Acknowledgements Cammas, C., Borderie, Q., Desachy, B., Augry, S., 2011. L’approche geoarch eologique de l’urbain. Developpement du concept, methodologie et applications aux operations archeologiques. Archeopages 31, 80e93. The authors wish to thank the Brussels Capital Region for Canti, M.G., 2003a. Earthworm activity and archaeological stratigraphy: a review of financing this research. They also want to thank Dr. Richard I. products and processes. Journal of Archaeological Science 30, 135e148. Macphail (University College London), Hugues Doutrelepont (Roots Canti, M.G., 2003b. Aspects of the chemical and microscopic characteristics of plant ashes found in archaeological soils. Catena 54, 339e361. asbl.) and Kristin Ismael-Meyer€ (IPNA, Basel) for their valuable Cappers, R.T.J., Bekker, R.M., Jans, J.E.A., 2006. Digitale Zadenatlas Van Nederland. advice. They acknowledge Constantin Pion (RPAW-CReA-Patri- Barkhuis Publishing, Groningen. moine) for providing information on the site of Rue des Pierres. The Carter, S.P., Davidson, D.A., 1998. An evaluation of the contribution of soil micro- morphology to the study of ancient arable agriculture. Geoarchaeology 13 (6), authors also acknowledge Welmoed Out (Moesgaard Museum) and 535e547. Dries Tys (Vrije Universiteit Brussel) for their pertinent and helpful Carver, M.O.H., 1987. The nature of urban deposits. In: Schofield, J., Lynch, R. (Eds.), comments on an earlier draft of this document. The anonymous Urban Archaeology in Britain. CBA Research Reports 61. Council for British e reviewers of the manuscript are also thanked for their comments. Archaeology, London, pp. 9 26. Charruadas, P., 2011. Croissance rural et essor urbain a Bruxelles. Les dynamiques d’une societ e entre ville et campagnes (1000e1300). Academie royale de Bel- Appendix A. Supplementary data gique, Bruxelles. Claes, B., 2008. Archeologisch Onderzoek Van Het Brusselse Arme Klarenklooster (BR100) (Unpublished report MBHG-KMKG, Brussels). Supplementary data related to this article can be found at http:// Claes, B., 2011. Archeologische evaluatie ter hoogte van drie huizen in de Korte dx.doi.org/10.1016/j.quaint.2016.06.025. Beenhouwersstraat nr, 29 en de Schuddeveldsteeg nr. 2 (BR229). Unpublished report MBHG-KMKG, Brussels. Courty, M.-A., Goldberg, P., Macphail, R.I., 1989. Soil Micromorphology in Archae- References ology. Cambridge Manuals in Archaeology. Cambridge University Press, Cambridge. AFNOR, 1999. Recueil de normes: qualite des sols. Association Française de Nor- Cremaschi, M., Nicosia, C., 2010. Corso Porta Reno, Ferrara (Northern Italy): a study malisation, Paris. in the formation processes of urban deposits. Il Quaternario e Italian Journal of Albert, R.M., Shahack-Gross, R., Cabanes, D., Gilboa, A., Lev-Yadun, S., Portillo, M., Quaternary Sciences 23 (2Bis), 395e408, 2010 e Volume Speciale. Sharon, I., Boaretto, E., Weiner, S., 2008. Phytolith-rich layers from the late Dam, J.P., Nuyens, J., Parent, J., Thonnard, R., 1977. Carte Geotechnique 31.3.7 bronze and iron ages at Tel dor (Israel): mode of formation and archaeological Bruxelles. Institut geotechnique de l’etat, Bruxelles. significance. Journal of Archaeological Science 35, 57e75. Deak, J., Gebhardt, A., Lewis, H., Usai, M.R., Lee, H., 2016. Soils disturbed by vege- Allen, R.L., 1852. The American Farm Book or Compend of American Agriculture. tation clearance and tillage. In: Nicosia, C., Stoops, G. (Eds.), Archaeological Soil Saxton, New York. and Sediment Micromorphology. Wiley-Blackwell, Chichester (in press). Baize, D., 2000. Guide des analyses en pedologie, choix, expression, presentation, Degraeve, A., Demeter, S., Devos, Y., Modrie, S., Van Bellingen, S., 2010. Brussel voor interpretation. INRA, Paris. 1200. Een archeologische bijdrage. In: Dewilde, M., Ervynck, A., Becuwe, F. Bakels, C.C., 1997. The beginnings of manuring in western Europe. Antiquity 71, (Eds.), Cenulae Recens Factae. Een Huldeboek Voor John Demeulemeester. 442e445. Academia Press, Gent, pp. 141e157. Ball, T., Vrydaghs, L., Mercer, T., Pearce, M., Snyder, S., Lisztes-Szabo, Z., Peto,} A., De Groote, K.D., Moens, J., Caluwe, D., Cooremans, B., Deforce, K., Ervynck, A., 2016. A Morphometric Study of Variance in Articulated Dendritic Phytolith Lentacker, A., Van Neer, W., 2009. Op zoek naar de oudste middeleeuwse Wave Lobes within Selected Species of Triticeae and Aveneae. Vegetation His- bewoning aan de Grote Markt te Aalst (prov. Oost-Vlaanderen). Het onderzoek tory and Archaeobotany. http://dx.doi.org/10.1007/s00334-015-0551-x. van afval- en beerkuilen uit de twaalfde tot de veertiende eeuw. Relicta, 4. VIOE, Bastin, B., Coûteaux, M., 1966. Application de la methode de Frenzel al ’extraction Brussels. des pollens dans les sediments archeologiques pauvres. L'Anthropologie 70 Devos, Y., 2013. Manuel d’echantillonnage pour les sciences naturelles au sein et (1e2), 201e203. aux abords des sites archeologiques en Region de Bruxelles-Capitale (version : Beckmann, T., 1997. Praparation€ bodenkundlicher Dünnschliffe für mikromorpho- 15/07/2013). Unpublished technical report CReA-Patrimoine. Universite Libre logische untersuchungen. In: Stahr, K. (Ed.), Mikromorphologische Methoden in de Bruxelles. der Bodenkunde, Hohenheimer Bodenkundliche Hefte, Heft 40. Universitat€ Devos, Y., Fechner, K., Vrydaghs, L., Degraeve, A., Deligne, F., 2007a. Contribution of Hohenheim, Stuttgart. archaeopedology to the palaeoenvironmental reconstruction of (pre-)urban Benyarku, C.A., Stoops, G., 2005. Guidelines for Preparation of Rock and Soil Thin sites at Brussels (Belgium). The example of the Treurenberg site. In: Boschian, G. Sections and Polished Sections. Quaderns DMACS 33. Universidad de Lleida, (Ed.), Proceedings of the Second International Conference on Soils and Lleida. Archaeology, Pisa, May 12e15, 2003. Atti della Societa Toscana di Scienze Bertran, P., Texier, J.P., 1997. Geoarch eologie des versants: les dep ots^ de pente. In: Naturali e Memorie serie A, 112. Societa Toscana di Scienze Naturali, Pisa, Bravard, J.P., Presteau, M. (Eds.), Dynamique du paysage: entretiens de pp. 145e151. geoarchologie. Documents d'Archeologie en Rhone-Alpes,^ Chatillon-sur-Cha-^ Devos, Y., Vrydaghs, L., Laurent, C., Degraeve, A., Modrie, S., 2007b. L’anthropisation laronne, pp. 59e86. du paysage bruxellois au 10e e 13e siecle. Resultats d’une approche 196 Y. Devos et al. / Quaternary International 460 (2017) 175e197

interdisciplinaire. In: On the road again. L'Europe en mouvement. Medieval Karkanas, P., Goldberg, P., 2010. Phosphatic features. In: Stoops, G., Marcelino, V., Europe Paris 2007. 4e Congres international d’ Archeologie Medi evale et Mees, F. (Eds.), Interpretation of Micromorphological Features of Soils and Moderne, Session 7. Archeologies environnementales. Institut National d’ His- Regoliths. Elsevier, Amsterdam, pp. 521e541. toire de l’Art, Paris, pp. 3e8. http://medieval-europe-paris-2007.univ-paris1.fr/ Katz, O., Cabanes, D., Weiner, S., Maeir, A.M., Boaretto, E., Shahack-Gross, R., 2010. Y.Devos%20et%20al.pdf. Rapid phytolith extraction for analysis of phytolith concentrations and assem- Devos, Y., Vrydaghs, L., Degraeve, A., Fechner, K., 2009. An archaeopedological and blages during an excavation: an application at Tell es-Safi/gath, Israel. Journal of phytolitarian study of the “Dark Earth” on the site of rue de Dinant (Brussels, Archaeological Science 37, 1557e1563. Belgium). Catena 78, 270e284. Knorzer,€ K.-H., 2007. Geschichte der synanthropen Flora im Niederrheingebiet (Vol. Devos, Y., Vrydaghs, L., Degraeve, A., Modrie, S., 2011a. Unravelling urban stratig- 61). Verlag Philipp von Zabern, Mainz am Rhein. raphy; the study of Brussels' (Belgium) dark earth. An archaeopedological Lambinon, J., De Langhe, J.-E., Delvosalle, L., Duvigneaud, J., 1998. Flora van Belgie,€ perspective. Medieval Modern Matters 2, 51e76. het Groothertogdom Luxemburg, Noord-Frankrijk en de aangrenzende gebie- Devos, Y., Fechner, K., Mikkelsen, J.H., 2011b. The application of phosphorus den (Pteridofyten en Spermatofyten), 3 ed. Nationale Plantentuin van Belgi, cartography to archaeological sites and structures: a state of the art and pro- Meise. posal of a protocol applicable for Belgium, Luxemburg and Northern France. In: Langohr, R., 1991. Soil characteristics of the Motte of Werken (west Flanders e Fechner, K., Devos, Y., Leopold, M., Volkel,€ J. (Eds.), Archaeology, Soil- and Life- Belgium). In: Tauber, J. (Ed.), Methoden und Perspektiven der Archaologie€ des Sciences Applied to Enclosures and Fields. BAR International Series, 2222. Mittelalters, Tagungsberichten zum interdisziplinaren€ Kolloquium, Sept. 1989, Archaeopress, Oxford, pp. 9e28. Liestal (Schweiz). Archaologie€ und Museum, 20, Liestal, pp. 209e223. Devos, Y., Vrydaghs, L., Fechner, K., Laurent, C., Degraeve, A., Modrie, S., 2011c. Langohr, R., 1994. Directives and rationale for adequate and comprehensive field Buried anthropic soils in the centre of Brussels (Belgium): looking for fields in a soil data bases. In: New Waves in Soil Science. Refresher Course for Alumni of (proto-) urban context. In: Fechner, K., Devos, Y., Leopold, M., Volkel,€ J. (Eds.), the International Training Centre for Post-Graduate Soil Scientists of the Ghent Archaeology, Soil- and Life- Sciences Applied to Enclosures and Fields. BAR University, Harare 1994. ITC-Gent, Publications Series, n 5. Ghent: Interna- International Series, 2222. Archaeopress, Oxford, pp. 143e161. tional Training Centre University Ghent, Ghent, pp. 176e191. Devos, Y., Nicosia, C., Vrydaghs, L., Modrie, S., 2013a. Studying urban stratigraphy: Langohr, R., Scoppa, C.O., Van Wambeke, A., 1976. The use of a comparative particle dark earth and a microstratified sequence on the site of the Court of Hoog- size distribution index for the numerical classification of soil parent materials: straeten (Brussels, Belgium). Integrating archaeopedology and phytolith anal- application to Mollisols of the Argentinian pampa. Geoderma 15, 305e312. ysis. Quaternary International 315, 147e166. Lewis, H., 2012. Investigating Ancient Tillage an Experimental and Soil Micromor- Devos, Y., Wouters, B., Vrydaghs, L., Tys, D., Bellens, T., Schryvers, A., 2013b. A soil phological Study. British Archaeological Reports, S2388. Archaeopress, Oxford. micromorphological study on the origins of the early medieval trading centre of Lindbo, D.L., Stolt, M.H., Vepraskas, M., 2010. Redoximorphic features. In: Stoops, G., Antwerp (Belgium). Quaternary International 315, 167e183. Marcelino, V., Mees, F. (Eds.), Interpretation of Micromorphological Features of Faegri, K., Iversen, J., 1989. Textbook of Pollen Analysis. Wiley and Sons, Chichester. Soils and Regoliths. Elsevier, Amsterdam, pp. 129e147. F.A.O, 2006. Guidelines for Soil Profile Description, fourth ed. FAO, Rome. Macphail, R.I., 1994. The reworking of urban stratigraphy by human and natural Fechner, K., Langohr, R., Devos, Y., 2004. Archaeopedological checklists. Proposal for processes. In: Hall, A.R., Kenward, H.K. (Eds.), Urban-rural Connexions: Per- a simplified version for the routine archaeological record in Holocene rural and spectives from Environmental Archaeology. Symposia of the Association for urban sites of Nord-Western Europe. In: Carver, G. (Ed.), Digging in the Dirt: Environmental Archaeology No. 12, Oxbow Monograph, Vol. 47. Oxbow, Oxford, Excavations in a New Millennium. British Archaeological Reports, International pp. 13e43. Series, S1256. Oxford. John and Erica Hedges Ltd, Oxford, pp. 240e256. Macphail, R.I., 2000. Soils and microstratigraphy: a soil micromorphological and Fondrillon, M., 2007. La formation du sol urbain: l’etude archeologique des terres microchemical approach. In: Lawson, A.J. (Ed.), Potterne 1982-5: Animal Hus- noires a Tours (4ee12e siecle) (Unpublished PhD thesis). Universite de François bandry in Later Prehistoric Wiltshire, Wessex Archaeology Report Nox 17. Rabelais, Tours. Wessex Archaeology, Salisbury, pp. 47e71. Fondrillon, M., 2009. A propos des recherches sur les terres noiress urbaines: Macphail, R.I., Cruise, G.M., 2000. Rescuing our urban archaeological soil heritage: a depasser le concept d’attente. Archeologie Medievale 39, 1e16. multidisciplinary microstratigraphical approach. In: Burghardt, W., Dornauf, C. Furlanetto, G., Nicosia, C., Ravazzi, C., 2014. Biostratigrafia e sedimentologia dei (Eds.), First International Conference on Soils of Urban, Industrial, Traffic and depositi litorali nell’area della palafitta Bodio centrale (sponda meridionale del Mining Areas. Proceedings, Vol I, The Unknown Urban Soil, Detection, Re- Lago di Varese). In: Grassi, B., Mangani, C. (Eds.), Storie Sommerse. Ricerche alla sources and Faces. Universitat-GH Essen, Essen, pp. 9e14. palafitta di Bodio 150 a anni dalla scoperta. Fantigrafica, Cremona, pp. 171e183. Macphail, R.I., Goldberg, P., 2010. Archaeological materials. In: Stoops, G., Galinie, H., 2004. L’expression “terres noires” un concept d’attente. In: Verslype, L., Marcelino, V., Mees, F. (Eds.), Interpretation of Micromorphological Features of Brulet, R. (Eds.), Terres Noires e Dark Earth. Actes de la table-ronde inter- Soils and Regoliths. Elsevier, Amsterdam, pp. 589e622. nationale tenue a Louvain-la-Neuve, les 09 et 10 novembre 2001. Universite Macphail, R.I., Linderholm, J., 2004. “Dark earth”: recent studies of “Dark earth” and Catholique de Louvain, Louvain-la-Neuve, pp. 1e11. “dark earth-like” microstratigraphy in England, UK. In: Verslype, L., Brulet, R. Gebhardt, A., 1995. Soil micromorphological data from traditional and experimental (Eds.), Terres Noires - Dark Earth. Actes de la table-ronde internationale tenue a agriculture. In: Barham, A.J., Macphail, R.I. (Eds.), Archaeological Sediments and Louvain-la-Neuve, les 09 et 10 novembre 2001. Universite Catholoique de Soils. Analysis, Interpretation and Management. Institute of Archaeology, Uni- Louvain, Louvain-la-Neuve, pp. 35e42. versity College, London, pp. 25e39. Macphail, R.I., Courty, M.-A., Gebhardt, A., 1990. Soil micromorphological evidence Gerasimova, M., Lebedeva-Verba, M., 2010. Topsoils e Mollic, Takyric and Yermic of early agriculture north-west Europe. World Archaeology 22 (1), 52e69. horizons. In: Stoops, G., Marcelino, V., Mees, F. (Eds.), Interpretation of Micro- Macphail, R.I., Galinie, H., Verhaeghe, F., 2003. A future for dark earth? Antiquity 77, morphological Features of Soils and Regoliths. Elsevier, Amsterdam, 349e358. pp. 351e368. Mauquoy, D., Van Geel, B., 2007. Mire and peat macros/Plant macrofossil methods Goldberg, P., Macphail, R.I., 2006. Practical and Theoretical Geoarchaeology. Black- and studies. In: Elais, A.E. (Ed.), Encyclopedia of Quaternary Science. Elsevier, well Publishing, Oxford. London, pp. 2315e2336. Guillerme, A., 1983. Les temps de l’eau. La cite, l’eau et les techniques. Champ McCobb, L.M.E., Briggs, D.E.G., Evershed, R.P., Hall, A.R., Hall, R.A., 2001. Preservation Vallon, Seyssel. of fossil seeds from a 10th century AD Cess pit at Coppergate, York. Journal of Hall, A.R., Huntley, J.P., 2007. A Review of the Evidence for Macrofossil Plant Re- Archaeological Science 28 (9), 929e940. mains from Archaeological Deposits in Northern England. English Heritage. Mikkelsen, J.H., 1997. Laboratory Method for Determining Organic, Inorganic and Harris, E.C., 1989. Principles of Archaeological Stratigraphy, second ed. Academic Total Phosphate. Bulletin 1 of the Archaeological Soil Micromorphology Press, London and San Diego. Working Group. University College of London, London. http://www.gre.ac.uk/ Henne, A., Wauters, A., 1845. Histoire de la ville de Bruxelles. Librairie ency- wat05/micro/soilmain/intro1.htm. clopedique de Perichon, Bruxelles. Milek, K.B., 2012. Floor formation processes and the interpretation of site activity Houchin, R., 2009. Voortgangsrapport (Januari-juni 2009). Archeobotanie. Unedited areas: an ethnoarchaeological study of turf buildings at Thvera, northeast Ice- technical report RBINS, Brussels. land. Journal of Anthropological Archaeology 31, 119e137. ICPN Working Group, Madella, M., Alexandre, A., Ball, T., 2005. International code Moore, P.D., Webb, J.A., Collinson, M.E., 1991. Pollen Analysis. Blackwell Scientific for phytolith nomenclature 1.0. Annals of Botany 96, 253e260. Publications, Oxford. Ismail-Meyer, K., Rentzel, P., Wiemann, P., 2013. Neolithic lakeshore settlements in Murphy, P.L., Wiltshire, P.E.J., 1994. A proposed scheme for evaluating plant Switzerland: new insights on site formation processes from micromorphology. macrofossil preservation in some archaeological deposits. Circaea 11, 1e6. Geoarchaeology: An International Journal 28, 317e339. Nakagawa, T., Brugiapaglia, E., Digerfelt, G., Reille, M., Beaulieu (de), J.-L., Yasuda, Y., Jacomet, S., 2003. Und zum Dessert Granatapfel d Ergebnisse der arch- 1998. Dense media separation as a more efficient pollen extraction method for aobotanischen€ Untersuchungen. In: Hagendorn, A., Doppler, H.W., Huber, A., use with organic sediment/deposit samples: comparison with the conventional Hüster Plagmann, H., Jacomet, S., Meyer-Freuler, C., Pfaf€ fli, B., Schibler, J. (Eds.), method. Boreas 27, 15e24. Zur Frühzeit von Vindonissa. Ausvvertung der Holzbauten der Grabung Nicosia, C., Devos, Y., 2014. Urban dark earth. In: Smith, C. (Ed.), Encyclopedia of Windisch-Breite 1996e1998. Aargauische Kantonsarchaologie,€ Brugg, Global Archaeology, vol. 11. Springer, New York, pp. 7532e7540. pp. 173e229. Nicosia, C., Langohr, R., Mees, F., Arnoldus-Huyzendveld, A., Bruttini, J., Cantini, F., Jacomet, S., 2006. Identification of Cereal Remains from Archaeological Sites, second 2012. Medieval Dark Earth in an active alluvial setting from the Uffizi gallery ed. Archaeobotany Lab IPAS, Basel University, Basel. complex in Florence, Italy. Geoarchaeology 27, 105e122. Jacomet, S., 2007. Plant macrofossil methods and studies/use in environmental Nicosia, C., Devos, Y., Borderie, Q., 2013. The contribution of geosciences to the archaeology. In: Elias, S.A. (Ed.), Encyclopedia of Quaternary Science (Vol. 3). study of European Dark Earths: a review. PCA Post-Classical Archaeologies 3, Elsevier Science, pp. 2384e2412. 145e170. Y. Devos et al. / Quaternary International 460 (2017) 175e197 197

Nicosia, C., Devos, Y., Macphail, R.I., 2016. European dark earth. In: Nicosia, C., Stoops, G., Eswaran, H., 1985. Morphological characteristics of wet soils. In: Wetland Stoops, G. (Eds.), Archaeological Soil and Sediment Micromorphology. Wiley, Soils: Characterization, Classification and Utilization. IRRI, Los Banos,~ Chichester (in press). Philippines, pp. 177e189. Pieters, M., 1990. Bemesting en aanverwante grondverbeteringspraktijken arche- Stoops, G., Marcelino, V., Mees, F., 2010. Micromorphological features and their ologisch bekeken. Archeologie 1990, 5e9. relation to processes and classification: general guidelines and keys. In: Pion, C., 2015. Fouille preventive rue des Pierres nos 18-20, 30, 34e36 et rue Henri Stoops, G., Marcelino, V., Mees, F. (Eds.), Interpretation of Micromorphological Maus nos 33-47 a Bruxelles, Archeologie a Bruxelles/Archeologie in Brussel, 9/1. Features of Soils and Regoliths. Elsevier, Amsterdam, pp. 15e35. Bruxelles Developpement urbain, Bruxelles. van Geel, B., 1986. Application of fungal and algal remains and other microfossils in Runge, F., 1999. The opal phytolith inventory of soils in central Africa - quantities, palynological analyses. In: Berglund, B. (Ed.), Handbook of Holocene Palae- shapes, classification, and spectra. Review of Palaeobotany and Palynology 107, oecology and Palaeohydrology. J. Wiley, Chichester, pp. 497e505. 23e53. van Geel, B., 2001. Non-pollen palynomorphs. In: Smol, J., Birks, J., Last, W. (Eds.), Schoch, W.H., Pawlik, B., Schweingruber, F.H., 1988. Botanical Macro-remains: an Tracking Environmental Changes Using Lake Sediments. vol. 3: Terrestrial, Algal Atlas for the Determination of Frequently Encountered and Ecologically and Siliceous Indicators. Kluwer Academic Press, Dordrecht, pp. 99e119. Important Plant Seeds. Paul Haupt Publishers, Berne/Stuttgart. van Geel, B., Pals, J.P., van Reenen, G.B., van Huissteden, J., 1995. The indicator value Shillito, L.M., 2011. Taphonomic observations of archaeological wheat phytoliths of fossil fungal remains, illustrated by a palaeoecological record of a Late from Neolithic Çatalhoyück,€ Turkey, and the use of conjoined phytolith size as Eemian/Early Weichselian deposit in the Netherlands. In: Herngreen, G., van an indicator of water availability. Archaeometry 53, 631e641. der Valk, L. (Eds.), Neogene and Quaternary Geology of North-West Europe. Simpson, I.A., Barrett, J.H., Milek, K.B., 2005. Interpreting the Viking age to medieval Mededelingen Rijks Geologische Dienst, vol. 52, pp. 297e315. period transition in Norse Orkney through cultural soil and sediment analyses. Van Strydonck, M., Boudin, M., 2011. Radiocarbon dating report. Petite rue des Geoarchaeology 20 (4), 355e377. Bouchers 29 site BR229; Place Royale 11 site BR061. IRPA-KIK. Unpublished Smith, D.N., 2013. Defining an indicator package to allow identification of ‘cesspits’ technical report. in the archaeological record. Journal of Archaeological Science 40, 526e543. Viklund, K., Linderholm, J., Macphail, R.I., 2013. Integrated palaeoenvironmental Speleers, L., van der Valk, J.M.A., 2016. Economic plants from medieval and post- study: micro- and macrofossil analysis and geoarchaeology (soil chemistry, medieval Brussels (Belgium), an overview of the archaeobotanical records. magnetic susceptibility and micromorphology). In: Gjerpe, L.E. (Ed.), E18- Quaternary International. http://dx.doi.org/10.1016/j.quaint.2015.11.025 (in prosjektet Gulli-langåker: Oppsummering Og Arkeometriske Analyser. Fag- press). bokforlaget, Bergen, pp. 25e83. Starkel, L., 1987. Man as cause of sedimentological changes in the Holocene. Striae Vrydaghs, L., Ball, T.B., Devos, Y., 2016. Beyond redundancy and multiplicity. Inte- 26, 5e12. grating phytolith analysis and micromorphology to the study of Brussels Dark Stockmarr, J., 1971. Tablets with spores used in absolute pollen analysis. Pollen et Earth. Journal of Archaeological Science 68, 79e88. Spores 13, 615e621. Weeda, E.J., Schaminee, J.H.J., van Duuren, L., Hennekens, S.M., Hoegen, A.C., Stoops, G., 1983. SEM and light microscopic observations of minerals in bog-ores of Jansen, A.J.M., 2000e2005. Atlas Van Plantengemeenschappen in Nederland. the Belgian Campine. Geoderma 30, 179e186. KNNV Uitgeverij, Utrecht. Stoops, G., 2003. Guidelines for Analysis and Description of Soil and Regolith Thin Sections. Soil Science Society of America, Madison.

View publication stats