Navigation – Plan du site

A multiscale geoarchaeological approach from the Laurentine shore (Castelporziano, Lazio, Italy)

Géoarchéologie du littoral laurentin (Castelporziano, Lazio, Italie) : approche multiscalaire
Andrew R. Bicket, Helen M. Rendell, Amanda Claridge, Peter Rose, James Andrews et Fiona S.J. Brown
p. 241-256

Résumés

La pertinence des questions géoarchéologiques est directement conditionnée par le choix d’une échelle de travail adaptée dans le but de proposer par la suite des stratégies d’échantillonnage appropriées et efficaces. Ces dernières sont localement influencées par les caractéristiques propres d’un site telles que sa topographie, la nature des sédiments sur lesquels il a été bâti, son environnement (paramètres géochimiques) et son climat. Le littoral laurentin correspond à l’ancienne ligne de rivage datée de l’époque romaine et qui est aujourd’hui localisé à 1 km du trait de côte actuel, dans la partie méridionale du delta du Tibre. La progradation de ce dernier, au cours de l’Holocène moyen et récent, est à mettre en liaison directe avec l’apport sédimentaire de son bassin versant dans un contexte de stabilisation du niveau relatif de la mer, cela permet notamment d’appréhender l’évolution morphologique de la région concernée (la superficie de la zone considérée de l’ordre de 103 km2). Des vestiges archéologiques particulièrement bien conservés dans la région de Castelporziano doivent être justement replacés dans un contexte morphologique régional pour pouvoir être interprétés sur le plan géoarchéologique. En privilégiant une échelle régionale d’étude et en se fondant sur l’établissement de transects, ainsi que sur la datation par Luminescence SAAD-IRSL de cordons dunaires fossiles, on parvient à fournir un modèle évolutif du delta du Tibre au cours de l’Holocène. Les datations obtenues permettent de reconstituer les différentes phases d’édification des cordons dunaires. En raison de la position d’abri de la région de Castelporziano, la topographie des cordons dunaires d’âge holocène y est mieux conservée que dans la partie centrale du delta, il est d’ailleurs possible d’y observer deux cordons supplémentaires. A une échelle régionale, quatre phases majeures de progradation deltaïque ont ainsi été identifiées avec une hausse particulièrement bien marquée du rythme d’avancée du trait de côte durant la période romaine et au cours des cinq cent dernières années. A une échelle moyenne de travail (superficie de la zone considérée de l’ordre de 102 km2), il est possible d’observer que les villas romaines et le Vicus Augustanus sont implantés sur le littoral laurentin. Cette période d’occupation (environ 2050 BP-1er siècle ap. J.-C.) correspond d’ailleurs à une avancée significative du delta du Tibre. La progradation du trait de côte a eu des conséquences directes sur le développement des sites archéologiques. C’est essentiellement pendant la seconde phase de l’expansion du Vicus Augustanus que la progradation est un phénomène observable à l’échelle décennale voire pluri décennale. En effet, l’expansion vers la mer du Vicus Augustanus est intervenue alors qu’aucun territoire n’existait pendant la première phase. Cela est à mettre en relation directe avec la progradation deltaïque et les processus morphologiques affectant le bassin versant du Tibre pendant la période romaine, incluant en particulier les activités d’origine anthropique.

Haut de page

Notes de la rédaction

Article soumis le 8 juin 2009, accepté le 10 octobre 2009

Texte intégral

The project is funded by AHRC Research Grant 18211, and undertaken in collaboration with the Soprintendenza archeologica di Ostia, Segretariato Generale della Presidenza della Repubblica, Direzione della Tenuta Presidenziale di Castelporziano, and the British School at Rome, we are indebted to all for their support and generous assistance in kind. Invaluable fieldwork assistance was provided by Tim Evans, Christopher Hewitson and Elizabeth Dikstra-McCarthy. The efforts of Giuseppe Mastronuzzi, Nick Marriner and an anonymous reviewer for improving an earlier version of this manuscript are also gratefully acknowledged.

Introduction

1The Laurentine Shore, the palaeo-shoreline that existed during the Imperial Roman period (beginning ca. 27 BC) is recorded by the presence of high-status villas, roads and other infrastructure on the southern, distal end of the Tiber Delta, Lazio, Italy. This shoreline is preserved up to 1 km inland of the present-day coast, within what is now the Italian Presidential Estate of Castelporziano (Fig. 1), due to the progradation of the Tiber Delta since the end of the Roman settlement (Bellotti et al., 2007; Rendell et al., 2007). Given the Roman archaeological remains’ position on a dynamic, prograding shoreline, the meso- (temporal and spatial) scale (102) archaeological record needs to be understood within the broader macro-scale (103+) context. This macro-scale context is primarily driven by a combination of changing relative sea level and delta progradation linked to flooding and sediment supply from the Tiber catchment. This has led to the development of a series of shore-parallel dune ridges (Otvos, 2000; Fig. 2 a and b) that act as repositories for sea level changes, coastline aggradation, progradation or transgression, climate, and coastal morphogenesis data (Mastronuzzi and Sanso, 2002) as well as the sedimentary context to date phases of shoreline progradation (Giraudi et al., 2009). The dune ridge record examined in this paper has developed within the context of the prograding highstand systems tract (HST) sediment sequence of the Tiber Delta (Bellotti et al., 1994; Bellotti et al., 1995; Amorosi and Milli, 2001). The HST sediments initially developed during stillstand sea level conditions between 7-4 ka BP (Lambeck et al., 2004a; Bellotti et al., 2007) and episodically throughout the late Holocene. Similar timing of inception and progradation of Tyrrhenian deltas is noted at several locations on the western coast of Italy (Bellotti et al., 2004; Amorosi et al., 2009) linked to macro-scale processes, especially climate. Regional climate, driving extreme storm events, may also be an important factor for both dune ridge formation and the erosion of beaches and duneforms on micro-scale (100-1) time (Hesp and Martínez, 2007). Within the Imperial Roman period focused upon in this paper, major coastal sites such as Portus and Ostia Antica flourish during the settlement of the Laurentine Shore to the south; sites that have a similarly close relationship to the development of the Tiber river and delta (Meiggs, 1973; Heinzelmann, 1998; Keay, 2005; Giraudi et al., 2009). A geoarchaeological approach (e.g., Kraft et al., 2007; Ghilardi and Desruelles, 2009) to investigating the coastal development of the archaeological record was undertaken. A transect-based approach to sampling has been used to develop a geochronological framework to investigate the timings of sand emplacement and dune ridge formation spanning the late Holocene. For clarity selected samples are presented here, derived from across the site. This approach has enabled the macro-scale context to be established for the site. The Roman settlement (meso-scale interactions) can then be examined in detail but within its broader, regional context. The purpose of this paper is to discuss the implications and the benefits derived from using a multi-scale approach by using the Laurentine Shore as a geoarchaeological case study.

Fig. 1 – Castelporziano Estate location map based on Shuttle Radar Topography Mission (SRTM) 90 m resolution DEM.
Fig. 1 – Carte de localisation de la région de Castelporziano réalisée à partir des données Shuttle Radar Topography Mission (SRTM, précision du MNT : 90 m).

Fig. 1 – Castelporziano Estate location map based on Shuttle Radar Topography Mission (SRTM) 90 m resolution DEM.Fig. 1 – Carte de localisation de la région de Castelporziano réalisée à partir des données Shuttle Radar Topography Mission (SRTM, précision du MNT : 90 m).

Deltaic and fluvial sediments characterise low-lying areas adjacent to sea level and are completely constrained by the altitudinal range 0-17 m as presented here. Altitudinal ranges are categorised by quantiles using ArcGIS 9.2. The most low-lying coastal areas recorded as below sea level also preserve the form of the Tiber Delta lagoons, now drained. WGS ‘84 coordinates are presented.
Les dépôts deltaïques et fluviatiles caractérisent les secteurs de faible dénivellation proches du niveau actuel de la mer et sont compris entre 0 et 17 m d’altitude. Les classes d’altitude ont été définies grâce au logiciel ArcGIS 9.2, sous formes de quantiles. Les zones les plus déprimées d’un point de vue topographique sont localisées le long du littoral et sont localisées sous le niveau moyen de la mer, ce sont d’anciennes lagunes désormais drainées. Le système de coordonnées est exprimé en WGS’84.

Fig. 2 – (A) Coastal dune ridge morphology within the Castelporziano Estate. (B) Sampling locations and dune ridge topography transects A-A’ and B-B’.
Fig. 2 – (A) Morphologie des cordons dunaires dans l’Etat de Castelporziano. (B) Localisation des échantillons et des transects topographiques A-A’ et B-B’ des cordons dunaires.

Fig. 2 – (A) Coastal dune ridge morphology within the Castelporziano Estate. (B) Sampling locations and dune ridge topography transects A-A’ and B-B’.Fig. 2 – (A) Morphologie des cordons dunaires dans l’Etat de Castelporziano. (B) Localisation des échantillons et des transects topographiques A-A’ et B-B’ des cordons dunaires.

Key archaeological sites mentioned in the text are highlighted. 1: Castelporziano Estate; 2: primary dune ridges; 3: secondary dunes. Primary dune ridges (2) are those believed to be directly linked to shoreline progradation driven by the development of the Tiber Delta, i.e., they are relict foredunes. Secondary dunes (3) are characterised by in situ remobilisation (such as the large parabolic dune that interrupts the primary dune ridge record seaward of the archaeological remains), or by dune ridges that are stratigraphically inconsistent following formation in inland locations and are therefore not directly linked to beach processes, i.e., cannot be considered as relict foredunes.
Les principaux sites archéologiques cités dans le texte sont indiqués. 1 : région de Castelporziano ; 2 : cordons dunaires primaires (liés à la progradation du delta du Tibre) ; 3 : dunes secondaires (formation non liée à des dynamiques littorales). Les cordons dunaires primaires (2) sont ceux directement liés à la progradation du delta du Tibre et qui se sont édifiés à mesure que le trait de côte avançait, ils sont désormais fossilisés à l’intérieur des terres. Les dunes secondaires (3) sont caractérisées soit par une remobilisation in situ des sédiments soit par une formation en retrait du littoral, qui ne sont donc pas associés à des dynamiques littorales.

Methods

2Three inland transects, normal to the shoreline, were sampled incorporating the current foredune, and preserved relict dune ridges (Fig. 2). Sampling was undertaken by digging out sections to depths of ~1.0-1.5 m in the dune ridges and also by coring to depths of ~2.0 m using a sand auger. Samples for luminescence dating were collected by hammering lengths of opaque plastic pipe into the sections or by transferring augered samples into opaque plastic bags under cover of a tarpaulin. Sampling locations and schematic sampling logs are presented in Fig. 3. Shading refers to the relative organic content of the horizons. Sampling was preferentially undertaken from the clean underlying dune (and occasionally palaeo-beach) sands. Sampling was undertaken between 2003 and 2009. It became apparent during laboratory analysis that aeolian remobilisation phases were fairly common occurrences and that ‘secondary’ dune ridges not associated with delta progradation had limited the effectiveness of the sampling strategy in some places. Emerging technology should improve the effectiveness of sampling strategies in the field (Sanderson and Murphy, in press). Luminescence age control is provided by 14C chronology constraining the Tiber Delta dune ridge sequence, sampled adjacent to Portus and Ostia (Giraudi et al., 2009). Given that much of the area sampled lies within 5 m of current sea level, detailed topographic survey was also undertaken to establish dune ridge morphologies, particularly heights, relative to both current and Roman sea level of -1m relative to modern sea level (RSL ; Lambeck et al., 2004 a and b). Ground-control points (GCP) for photogrammetry and GIS georeferencing were provided by a Leica SR20 D-GPS system. Post-processing against the Italian GPS reference network (INGR-Roma station) allowed cm to sub-cm point accuracy including for altitude providing a WGS84 coordinate system for the site. This permitted accurate assessment of total station measurements relative to current sea level. A number of transects, normal to the coastline were measured by total station to derive dune ridge cross-sections. Transects from the vicinity of the Vicus Augustanus are shown in Fig. 3 and Fig. 4. Dune ridge crests are typically less than 50 m wide. The altitude in Fig. 2 is corrected to an estimated Roman period sea level of -1 m relative sea level (RSL). In order to relate this meso-scale palaeo-shoreline in the wider archaeological and geomorphological setting, the Roman (in the vicinity of the roman archaeological remains) and post-Roman dune ridges were mapped using a Garmin GPSmap 60csx (WAAS/EGNOS-enabled), which provided 3-6 m latitude/longitude position accuracy under dense forest cover. Joint GCP and dune ridge mapping data underpins the position of the Laurentine Shoreline in Fig. 1. Aerial photos of the Tiber Delta area taken in 1954 by the Istituto Geografico Militare, Florence show visible dune ridges and provide additional information on delta structure. In particular, the lack of extensive building on the Tiber delta at this time provides a wealth of geomorphological information. Recent archaeological reconstruction work from the Portus Project (Keay, 2005) and Ostia Antica (Heinzelmann, 1998) incorporate similarly meso-scale geophysical/geomorphological surveys. Integration of these other regions of the Laurentine Shore, geological, geomorphological mapping and aerial photography facilitates the reconstruction of the Laurentine Shore outside the confines of the Castelporziano Estate.

Fig. 3 – Sampling logs from selected dune ridge crests and palaeo-beach locations (CP04/1).
Fig. 3 – Localisation des profils stratigraphiques des crêtes dunaires et des anciennes plages (CP04/1).

Fig. 3 – Sampling logs from selected dune ridge crests and palaeo-beach locations (CP04/1).Fig. 3 – Localisation des profils stratigraphiques des crêtes dunaires et des anciennes plages (CP04/1).

Samples from remobilised or secondary dune ridges (i.e., not linked to foredune formation during delta progradation) are also included (CP03/3, CP08/2, CP08/3). Samples CP07/15 and CP07/17 are taken from ferricretised relict dune ridges; all other samples are derived from calcium-carbonate cemented aeolianites. Munsell colours refer to wet sediments. Shading refers to relative organic content with sampling primarily undertaken from clean sands. 1: low organic content; 2: moderate organic content; 3: high organic content; 4: surface vegetation; 5: extensive root penetration; 6: some root penetration; 7: pseudo-gley features / mottled appearance; 8: indurated sediment; 9: diagenetic glaebules; 10: building rubble; 11: Roman wall.
Les échantillons provenant des dunes secondaires (formations non directement liées à la progradation deltaïque) sont également intégrés (CP03/3, CP08/2, CP08/3). Les échantillons CP07/15 et CP07/17 ont été prélevés dans des cordons dunaires indurés par l’oxydation du fer ; l’ensemble des autres sédiments sont des éolianites possédant un ciment composé de calcium-carbonates. Les couleurs réfèrent à celles du code Munsell et ont été identifiées à partir de sédiments humides. Les ombrages se réfèrent à la teneur en matière organique des échantillons avec un prélèvement réalisé sur des sables dégagés de toute impureté. 1 : faible teneur en matière organique ; 2 : teneur en matière organique moyenne ; 3 : haute teneur en matière organique ; 4 : végétation ; 5 : pénétration du système racinaire ; 6 : quelques ramifications racinaires ; 7 : pseudo gley bariolées ; 8 : sédiments indurés ; 9 : nodules liés à une diagenèse ; 10 : décombres de mur ; 11 : mur romain.

Fig. 4 – Dune ridge topographic transects.
Fig. 4 – Transects topographiques des cordons dunaires.

Fig. 4 – Dune ridge topographic transects. Fig. 4 – Transects topographiques des cordons dunaires.

Location of transects is highlighted in Fig. 2B. Elevations are corrected to a Roman period sea level of -1 m RSL (Lambeck et al., 2004 a and b) to provide some context for assessing archaeological relationships to their contemporary sea level.
La localisation des transects est mise en évidence sur la Fig. 2B. Les altitudes sont corrigées pour être exprimées par rapport au niveau marin de l’époque romaine, situé 1 m en dessous de l’actuel (Lambeck et al., 2004 a and b).

3Grain-size distributions of bulk sediments were measured by a Coulter LS230 laser particle sizer (LPS). Analysis was provided by GRADISTAT v5 (Blott and Pye, 2001). Mode and Folk and Ward method measurements (in μm) are used here. Preliminary optical- and electron- microscopic analysis had shown that due to the aeolian nature of the studied sands, grain shapes were typically rounded to sub-rounded suggesting that LPS (which assumes a spherical grain) derived grain-size distributions are accurate, simplifying the description of the grain-size (Blott and Pye, 2008). Grain-sizes are reported here primarily for contextualising the modal grain-size used for the luminescence dating geochronology (see below), to assess the effectiveness of environmental dosimetry upon coarse grains (i.e., greater than silt-sized particles; Fleming, 1971; McKeever, 1985; Aitken, 1998). Dune ridge and sandsheet sediments were dated by the single-aliquot additive dose infra-red stimulated luminescence (SAAD-IRSL) method (Duller, 1991; Blair et al., 2005); initial research has shown the suitability of this protocol on Holocene sediments from this site (Rendell et al., 2007). Grain-size fractions approximating the mode were used for each dating sample, recovered by dry-sieving for 180-250 μm or 250-300 μm fractions. This was aimed at incorporating the aeolian sorting recorded by the individual geomorphological features. Samples were treated with 10 % HCl to remove carbonates. Loss-on-ignition analysis had previously shown that organic contents across the sampled site were typically very low (<1-2 %; tab. 1), therefore hydrogen peroxide (H2O2) pre-treatment was not undertaken in this case. K-feldspar was isolated using sodium polytungstate (SPT) heavy liquid at densities of 2.62 g/cm2 and 2.58 g/cm2, to remove quartz and heavy minerals, and plagioclase feldspars, respectively. Quartz supplied to the site from the upper Tiber catchment is Alpine in origin (Garzanti et al., 2002), concerns about the low-sensitivity of geologically young Alpine quartz (Preusser, 2003) make it less suitable for dating young sediments of interest in this paper. Small sample aliquots were used to maximise detection of Equivalent Dose (DE) heterogeneity, with approximately 3-5 mg on each disc mounted with silicon oil. An automated Risø TL/OSL DA 12 reader with 90Sr/90Y beta source was used for all luminescence measurements. Optical stimulation was by 31 TEMT484 infrared diodes producing 40 mWcm-2 with peak emission wavelength of 880 ± 80 nm. The detection system was an EMI9635QA photomultiplier tube with Schott BG-39 and Corning 7-59 colour glass filters, creating a detection window of 320-480 nm. This avoids detection of potentially unstable UV emissions from alkali feldspars (Clarke and Rendell, 1997). All DE determinations were made using 12 small-aliquots. The Single-Aliquot Additive Dose (SAAD) protocol was used for each sample. Aliquots were measured using 0.5 s IR stimulation at 50°C, and using preheats of 220°C for 600 s to remove unstable luminescence signal components (Clarke and Rendell, 1997). Six aliquots were used to correct for signal loss due to repeated heating and measurement during the SAAD protocol. Experiments show that K-feldspar from across the site displays no anomalous fading after 14 weeks of storage (Bicket, 2009). Dosimetry was provided by thick source alpha counting (TSAC; Aitken, 1998) and thick source beta counting (TSBC; Sanderson, 1988) methods to identify external U, Th and K dose rates. An internal a-value of 0.2 was used for each sample (Rendell et al., 2007). DE determinations typically exhibited ≤10 % spread. A conservative estimated water content of 10±5 % was used to account for fluctuating water contents throughout the burial history of the sediments. The very low-altitude coastal setting relative to sea level and seasonally fluctuating non-marine water tables suggests a more conservative estimate is appropriate, especially as sampling was undertaken during the summer months when water tables are generally lower than during the winter months.

Tab. 1 – SAAD-IRSL dates. Modal grain-size and sorting (standard deviation) parameters, environmental dosimetry measurements are presented.
Tab. 1 – Datations par luminescence SAAD-IRSL. Sont présentés les indices granulométriques du mode et du tri et les mesures de dosimétrie environnementale.

Sample

Mode (μm)

Sorting (μm)

OD (%)

DE (Gy)

St. Error (Gy)

n (12)

Doserate (Gy/ka)

U (ppm)

Th (ppm)

U/Th

K (%)

Age BP (10% water cont.)

Error (Years)

Transect 1

CP03/6

269.5

1.49

22.261

0.152

0.010

11

2.887

1.61

7.24

0.222

1.27

52

4

CP07/26

295.9

1.759

9.797

2.045

0.063

10

2.832

1.32

5.44

0.243

1.46

722

53

CP07/25

300

1.49

8.034

4.505

0.104

12

2.758

1.935

8.145

0.238

0.96

1633

123

CP04/1*

296.0

1.759

7.837

5.945

0.140

12

2.923

2.065

7.145

0.289

1.24

2034

146

CP03/2

295.9

1.759

4.258

5.845

0.072

12

2.819

1.97

6.08

0.324

1.22

2073

144

CP07/23

269.5

1.894

7.087

8.414

0.172

12

2.609

1.0235

4.22

0.243

1.44

3224

224

Transect 2 (selected samples)

CP07/14

269.5

1.894

9.482

13.373

0.423

10

3.145

3.075

9.82

0.313

1.01

4252

316

CP07/15

269.5

1.894

7.691

18.406

0.448

10

3.662

2.805

8.385

0.335

1.8

5055

331

CP07/17

269.5

1.894

8.470

20.373

0.520

11

3.756

2.285

7.175

0.318

2.16

5424

343

Transect 3 (selected sample)

CP07/18

245.5

2.028

7.930

1.613

0.039

10

2.774

1.455

6.315

0.230

1.37

581

39

Secondary dune samples (post-abandonment deposition)

CP03/3

295.9

1.759

8.366

4.774

0.115

12

3.207

1.46

6.13

0.238

1.8

1488

100

CP08/2

300

1.624

8.992

5.025

0.151

9

3.156

2.62

7.83

0.335

1.26

1592

116

CP08/3

300

1.624

4.587

6.016

0.08

12

3.475

2.015

7.03

0.287

1.89

1731

109

All ages are calculated with a 10 % water content due to uncertainties over water content throughout the burial of these coastal sediments.
Tous les âges sont calculés avec une teneur en eau de 10 % en raison de l’incertitude liée à l’estimation de la teneur en eau des anciens sédiments littoraux, aujourd’hui fossilisés.

Results and analysis

4The three dated inland transects incorporating the current foredune, relict foredunes post-dating, contemporary with, and predating the Roman occupation of the site, and early Holocene/Late Pleistocene sand sheet samples, are presented in tab. 1. As discussed, age control is provided by 14C chronology from marsh sediments constraining dune ridge phases in the vicinity of Portus and Ostia in the central Tiber Delta (Giraudi et al., 2009). A comparison of the 14C chronology from the central Tiber Delta and the nine phases of dune ridge formation determined from luminescence chronology from Castelporziano are shown in tab. 2. Dune phase date ranges are converted to BP based on AD 2000 to enable comparison with the age range of the luminescence chronology. Both dune ridge chronologies are in good agreement. However the age of dune phase CP I may be underestimated due to the post-burial effects of ferricretisation and groundwater mobilisation of radionuclides during diagenesis (Olley et al., 1996; Bicket, 2009) leading to underestimates of dose-rate. Dune phase CP II is also characterised by ferricretisation but appears to agree well with the age control suggesting that post-burial effects are sample specific (Bicket, 2009). The remaining dune phases categorised by luminescence dating at Castelporziano are sampled from the HST carbonate cemented aeolianites of the coastal dune ridges and are not thought to suffer from these dosimetry issues. The Castelporziano dune ridge geochronology records an additional two phases of coastal dune ridge development linked to Tiber Delta progradation to that of the more-centrally sampled framework of C. Giraudi et al. (2009) (tab. 2). This is likely due to the erosional effects of delta retreat discussed by C. Giraudi et al. (2004; 2009) in the central Tiber delta region, and also the protected and relatively unexploited nature of the Castelporziano Estate which is relatively free from extensive coastal construction and modern development in general. A simple age model was constructed from transect 1 luminescence dates (the most comprehensive sampling transect). The inland distance of dune ridges measured from GIS mapping derived from field measurements, topographic survey, GPS data and aerial photography. Progradation rates during the late Holocene are calculated from the age ranges of coastal dune ridge samples only (tab. 3; Fig. 6). Due to the construction of the modern coastal highway, the Via Litoranea bisects the site at a position in dune phase CP IX likely to record the time period of the Little Ice Age (i.e., during the last 500 years). Dating of this phase is then relatively poorly constrained. The shore-parallel highway (not shown in the geomorphological mapping presented in this paper) lies at a position associated with the seaward edge of dune phase CP IX (Fig. 5) and may be responsible for the formation of the substantial parabolic dune orientated to the dominant onshore westerly wind.

Tab. 2 – Coastal dune ridge phases recorded from the Tiber Delta.
Tab. 2 – Phases d’édification des cordons dunaires enregistrées dans le delta du Tibre.

Giraudi et al., 2009

Castelporziano

Dune phase

Age (BP)

Dune sample

Age (BP)

Dune Phase

I

6000 – 5700

CP07/17

5767 - 5081

CP I

II & III

5275 – 4930

CP07/15

5386 - 4724

CP II

IV

4140 – 3920

CP07/14

4568 – 3936

CP III

-----

-----

CP07/23

3448 – 3000

CP IV

V

2910 – 2800

Unsampled

CP V

VI-a

2400 – 1700

CP03/2

2217 – 1929

CP VI

  

  

CP04/7

2032 - 1794

  

The Laurentine Shore ca. 2100 – 1500 BP

VI-b

1700 – 1000

CP07/25

1756 – 1510

CP VII

-----

-----

CP07/26

775 – 669

CP VIII

VII – VIII

600 – 200

CP07/18

620 – 542

CP IX

Comparison of the radiocarbon constrained dune phase chronology of C. Giraudi et al. (2009; using centennial date ranges where appropriate) and luminescence dune ridge chronology from this study show good agreement. Two additional dune ridge phases are recorded by this study.
La comparaison des résultats obtenus grâce aux datations par le radiocarbone (Giraudi et al., 2009) et par la luminescence indique une bonne corrélation. Deux cordons dunaires supplémentaires sont révélés dans le présent article.

Tab. 3 – Tiber Delta minimum progradation rates calculated from age ranges of luminescence dates from transect 1 that incorporates the palaeo-shoreline position of the Vicus Augustanus (Fig. 2).
Tab. 3 – Taux minimum de progradation du delta du Tibre calculés d’après les datations par luminescence réalisées sur le transect 1. Ce dernier intègre l’ancien trait de côte où était installé le Vicus Augustanus (Fig. 2).

Date Range (BP)

Range (a)

Progradation (m)

Rate (m/a)

3448-1929

1519

250

0.165

2217-1510

707

380

0.537

1756-669

1087

170

0.156

775-48

727

250

0.344

Fig. 5 – Dune ridge phase map from Castelporziano based on luminescence geochronology related to complimentary 14C chronology of C. Giraudi et al.(2009).
Fig. 5 – Carte de localisation des cordons dunaires de la région de Castelporziano fondée sur l’utilisation combinée de datations par luminescence et par le radiocarbone 14C (Giraudi et al., 2009).

Fig. 5 – Dune ridge phase map from Castelporziano based on luminescence geochronology related to complimentary 14C chronology of C. Giraudi et al.(2009).Fig. 5 – Carte de localisation des cordons dunaires de la région de Castelporziano fondée sur l’utilisation combinée de datations par luminescence et par le radiocarbone 14C (Giraudi et al., 2009).

1: Castelporziano Estate; 2: dune ridges; 3: secondary dunes. Archaeological sites mentioned in the text are also highlighted.1 : région de Castelporziano ; 2 : cordons dunaires ; 3 : cordons dunaires secondaires. Les sites archéologiques mentionnés dans le présent article sont également représentés.

Fig. 6 – Transect 1 age model (samples and luminescence age determinations are shown in Fig. 2, Fig. 3, and tab. 1, respectively).
Fig. 6 – Le modèle d’évaluation âge/distance pour le transect 1 (les échantillons prélevés et l’obtention des dates par la méthode de luminescence sont respectivement présentées sur la Fig. 2, la Fig. 3 et le tab. 1).

Fig. 6 – Transect 1 age model (samples and luminescence age determinations are shown in Fig. 2, Fig. 3, and tab. 1, respectively).Fig. 6 – Le modèle d’évaluation âge/distance pour le transect 1 (les échantillons prélevés et l’obtention des dates par la méthode de luminescence sont respectivement présentées sur la Fig. 2, la Fig. 3 et le tab. 1).

1: primary dune ridges; 2: secondary dune formations. Central ages are meaningful with respect to archaeological age control. Secondary dunes are sampled from aeolian sand overlying the Vicus Augustanus.1 : cordons dunaires primaires ; 2 : dunes secondaires. Les datations se corrèlent bien avec les données archéologiques. Des sables éoliens ont été échantillonnés dans des dunes de formation secondaire qui recouvrent le Vicus Augustanus.

Discussion

Macro-scale context

5On the macro-scale (103+), sediment supply can be seen to significantly increase during at least the last 2000 years (Fig. 7) which manifests as a threefold increase in progradation rate of the Tiber Delta developing from around 2217 BP (the maximum age range of sample CP03/2). The historical flooding record of the Tiber Delta has been discussed by several authors (Bersani and Bencivenga, 2001; Aldrete, 2007; Stewart and Morhange, 2009) in the context of the effect of flooding upon ancient Rome (Aldrete, 2007) and the impact of the Little Ice Age upon the Tiber Delta and other western Mediterranean deltas (Arnaud-Fassetta and Provansal, 1999; Stewart and Morhange, 2009). This record of flooding frequency based on documentary evidence as well as flood markers and cemetery surveys, must also be interpreted as a gauge for macro-scale sediment supply flux. The flooding frequency record at Rome and the Tiber Delta progradation rates are presented in Fig. 7 and Fig. 8. It is clear that the progradation rate flux is only sensitive on multi-centennial to millennial timescales, whereas the historical flooding record is at least an order of magnitude more sensitive constructed from annual and sub-annual records and presented as a centennial histogram. There is however general agreement on macro-scale (103) periods of increased flooding frequency and progradation rate suggesting the flood record is an effective guide to sediment supply linked to delta progradation rates. This relationship also suggests that dune ridge formation and sediment supply flux operate on different scales. Fig. 9 can be considered as a preliminary macro-scale reconstruction of the Tiber Delta shoreline ca. 1900-1800 BP (ca. early 2nd century AD), following the expansion of Portus under Trajan. Within this broader spatial and temporal late Holocene geomorphological context, the Laurentine Shore is a relatively transient human occupation in both duration and spatial extent (arguably not in alongshore extent !), between ca. 2050-1580 BP based on artefact and architectural evidence (Lauro, 1985; 1988; 1998). The meso-scale archaeological and geomorphological records investigated at the Laurentine Shore are discussed below.

Fig. 7 – Historical Tiber flooding frequency at Rome (adapted from Bersani and Bencivenga, 2001).
Fig. 7 – Fréquence des épisodes d’inondation du Tibre à Rome (adapté de Bersani et Bencivenga, 2001).

Fig. 7 – Historical Tiber flooding frequency at Rome (adapted from Bersani and Bencivenga, 2001).Fig. 7 – Fréquence des épisodes d’inondation du Tibre à Rome (adapté de Bersani et Bencivenga, 2001).

Fig. 8 – Tiber Delta minimum progradation rates calculated from age ranges of luminescence dates from transect 1 that incorporates the palaeo-shoreline position of the Vicus Augustanus (tab. 1).
Fig. 8 – Taux minimum de progradation du delta du Tibre calculés à partir des datations par luminescence effectuées sur des échantillons prélevés sur le transect 1, qui incorpore l’ancienne position littorale du Vicus Augustanus (tab. 1).

Fig. 8 – Tiber Delta minimum progradation rates calculated from age ranges of luminescence dates from transect 1 that incorporates the palaeo-shoreline position of the Vicus Augustanus (tab. 1).Fig. 8 – Taux minimum de progradation du delta du Tibre calculés à partir des datations par luminescence effectuées sur des échantillons prélevés sur le transect 1, qui incorpore l’ancienne position littorale du Vicus Augustanus (tab. 1).

Fig. 9 – Preliminary reconstruction of the Laurentine Shore and central Tiber Delta during the 2nd century AD, following the expansion of Portus under Trajan incorporating data from S.J. Keay et al. (2005), M. Heinzelmann (1998) and P. Bellotti et al., (2007).
Fig. 9 – Proposition de reconstitution du littoral laurentin et du delta du Tibre au IIè siècle ap. J.-C., après l’expansion de Portus pendant le règne de l’Empereur Trajan (d’après les données de S.J. Keay et al. (2005), M. Heinzelmann (1998) et P. Bellotti et al. (2007).

Fig. 9 – Preliminary reconstruction of the Laurentine Shore and central Tiber Delta during the 2nd century AD, following the expansion of Portus under Trajan incorporating data from S.J. Keay et al. (2005), M. Heinzelmann (1998) and P. Bellotti et al., (2007).Fig. 9 – Proposition de reconstitution du littoral laurentin et du delta du Tibre au IIè siècle ap. J.-C., après l’expansion de Portus pendant le règne de l’Empereur Trajan (d’après les données de S.J. Keay et al. (2005), M. Heinzelmann (1998) et P. Bellotti et al. (2007).

1: Portus; 2: Ostia Antica; 3: villa di Plinio; 4: the Laurentine shore; 5: Tiber lagoons.
1 : Portus ; 2 : Ostia Antica ; 3 : villa di Plinio ; 4 : littoral laurentin ; 5 : système lagunaire consécutif de la progradation du delta du Tibre.

Meso-scale archaeological and geomorphological interactions

6Many of the Roman structures lie directly adjacent to the back-beach of that time. The foundations of the roman buildings lie at ~ 2-3 m above relative sea level during the roman period and dune building is the only natural process that will create an elevated topography in this coastal setting. The typical dune crest is around 50 m wide out-with the seasonally-flooded dune slacks. Individual structures such as small villas and baths have floor plan dimensions that could be accommodated on dune ridges of comparable dimensions to those further inland, but the larger complexes of structures such as the Vicus Augustanus (and Imperial Villa near Tor Paterno) occupy areas greater than 100-150 m wide. Luminescence dating of dune ridge crests, palaeo-beach and beneath roman structures (Rendell et al., 2007) compiled into the transect 1 age model describes a coeval coastal strip developing rapidly from around 2100±200 BP. Inception towards the older extreme of this date is more likely as the Roman archaeological remains are constrained to the seaward edge of this coastal strip requiring some time for the ground the foundations were built on to be deposited. The Roman settlement of the Laurentine Shore occurs during this pronounced progradation phase between 2100-1600±200 BP. We assume the subsequent dune ridge position roughly equates to the berm-terrace on the preceding phase of the shoreline within this context of prograding shoreline conditions (Hesp, 2002). The importance of storminess upon the geomorphological and also archaeological record may also be critical for understanding the interaction of human responses and the landscape (Hesp and Martínez, 2007) but is not the primary focus of this paper. For example, dune ridge formation in some cases may be linked to storm erosion of beach profiles, thereby creating berm terraces. The overlapping dune ridge waveform seaward of the Laurentine Shore archaeological remains suggests that wave-scarping may have been an important process (Fig. 4). In this case, if sand-supply is sufficient then foredune formation may occur in the lee of the berm terrace. During periods of negative sediment budget, considerable dune ridge and beach profile erosion and landward overwash may occur (Hesp, 2002; Psuty, 2008). In many cases secondary dune formation has occurred around archaeological buildings. Occasional parabolic dunes are present, in contrast to the predominant transverse coastal dune ridges. These secondary dune formations have been dated in a phase across the site, to around 1500-1600 BP (tab. 1) developing in a period that saw the end of occupation at the Vicus Augustanus, aeolian inundation of the Piscinae and more widely the collapse of the western Roman Empire. It should be noted that no causal link is currently suggested between dune formation and site abandonment, a lack of site maintenance, post-abandonment, could explain these aeolian deposits. The key human-environment interactions associated with the palaeo-shoreline are more visible on the meso-scale, but driven on the macro-scale. Therefore in order to pose and answer geoarchaeological questions the broader context of the site must be investigated. To illustrate this point the inter-relationship of the macro-scale late Holocene Tiber Delta progradation and meso-scale coastal human settlement predominantly at the Vicus Augustanus is specifically discussed below.

Multi-scale geoarcheological interactions at the Vicus Augustamus

7The Vicus Augustanus displays two phases of construction beginning around 2010-1950 BP (Fig. 10). Archaeological excavation suggests that following a major fire at the Vicus Augustanus between 1950-1925 BP (AD 50-75) a rebuilding effort was undertaken at a higher level and closer to the sea (personal communication of Prof. A. Claridge). The second building phase, on the seaward side of the initial complex is recorded between 1750-1900 BP (Rendell et al., 2007; Fig. 10). Underlying sand from beneath the first phase of the Vicus is dated to 2050±240 BP (CP895; Rendell et al., 2007) around 950 m inland of the current shoreline (Fig. 11). This date is in good agreement with a wall-abutting sample (CP04/1 this study, 2034±146 BP) and inland primary dune ridge (CP03/2, 2073±144 BP) that delineates the coeval coastal plain at least 350m wide at the Vicus Augustanus. It also suggests that central age estimates based on luminescence dating are meaningful in conjunction with well-dated archaeological materials such as tiles stamped with calendar dates. Elevated building sites are provided by the dune ridge topography of at least 2-3 m above the contemporary Roman period sea level. A key constraining factor is that Roman archaeological remains are not found further seaward of this recorded series of buildings. This relationship dates the minimum shoreline position at 830 m inland to 1750 BP, which is in good agreement with the transect 1 age model (Fig. 6). Whether or not a shift in shoreline was directly experienced by the Roman population of the Vicus Augustanus and wider Laurentine Shore properties is dependent on the timing and rate of delta progradation; data not sensitively recorded by macro-scale, inter-dune average progradation rates. Relative sea level rise during the settlement of the VicusAugustanus was on the scale of millimetres so an increase in accommodation space is perhaps unlikely to be a mitigating factor for shoreline advance on the decametre-scale. As has been previously discussed, sediment supply to the site is driven by the sediment supply from the Tiber River with flooding frequency an important factor in sediment supply flux. During the 1st centuries BC and AD (2100-1900 BP) high-magnitude floods within the lower Tiber catchment at Rome increased significantly (Fig. 7). An increased sediment supply during flooding of the Tiber River ca. 2000 BP is argued to drive a positive sediment budget on the delta beaches leading to progradational beach topography (Hesp, 2002; Psuty, 2008). Significant progradation is recorded prior to the Roman occupation ca. 2217 BP (discussed above). This maximum age estimate is in agreement with the transect 1 age model for samples from this position inland (Fig. 6). Within this geomorphological context integrating the archaeological data is useful for assessing possible impacts upon the coastal settlements and their occupants. Based on the 70 m distance between age estimate of the first phase of the Vicus (2010 BP), which is in good agreement with luminescence date CP895 (2050±240 BP) from sediment lying stratigraphically below (Rendell et al., 2007), and the minimum age estimate for the dune ridge coeval with abandonment of the VicusAugustanus in the 5th century AD, i.e., 1510 BP (CP07/25), a minimum progradation rate of 0.14 m/a is derived for the period during the occupation of the Vicus Augustanus. Assuming that progradation was a low-level, and relatively constant process, this suggests that shoreline advance was not necessarily of concern on yearly timescales but is observable on decadal to multi-decadal time-scales; at least during the lifetime of an individual witness and certainly within a few generations. During the occupation of the first phase of the Vicus Augustanus, 2010-1950 BP (10 BC-AD 50), an estimated shoreline progradation of 8 m occurred. During the longer second phase, 1925-1580 BP (AD 75-420), an estimated further 48 m of progradation occurred. This progradation records the total long-term sediment budget flux to the shoreline. The expression of dune progradation is conceptually dependent upon positive sediment budgets in both the beach and foredune (Miyanishi and Johnson, 2007). The presence of the relict dune ridge betrays a reduced sediment supply and transition to negative beach and positive dune budgets driving either foredune aggradation or inland overwash; the latter arguably being more problematic for Roman coastal settlements. What is certain is that the dune ridge coeval with the abandonment of the archaeological sites along the Laurentine Shore at Castelporziano (and also the collapse of the Roman Empire; 1633±123 BP, AD 367±123, dune phase CPVII) formed associated with a beach that was around 70 m seaward of the initial Roman coastal settlement (Fig. 11). The discrepancy between building plots from the first phase of the Vicus settlement and the second is also around 70 m (personal communication of A. Claridge) suggesting a human response to shoreline progradation could be inferred. The ground that the seaward elements of the second phase of the Vicus Augustanus were built onto did not likely exist during the occupation of the first phase of settlement. The foundations of the Vicus Augustanus and all of the coastal Roman archaeological remains on the Laurentine Shore upon the Tiber Delta are constructed upon sediments derived from the Tiber catchment supply; this supply being enhanced by the activity of the Roman culture in the wider Tiber valley hinterland.

Fig. 10 – Vicus Augustanus major building phases (adapted from Rendell et al., 2007).
Fig. 10 – Principales phases de construction du Vicus Augustanus (d’après Rendell et al., 2007).

Fig. 10 – Vicus Augustanus major building phases (adapted from Rendell et al., 2007).Fig. 10 – Principales phases de construction du Vicus Augustanus (d’après Rendell et al., 2007).

Fig. 11 – Post-abandonment shoreline reconstruction at the Vicus Augustanus. Age estimates of dune ridge and palaeo-beach samples included in the transect 1 age model and post-abandonment secondary aeolian deposits burying the archaeological remains are presented.
Fig. 11 – Reconstitution du trait de côte après l’abandon du site du Vicus Augustanus. Sont mentionnés la datation approximative des sédiments prélevés dans les cordons dunaires et les anciennes plages (transect 1) ainsi que les dépôts éoliens recouvrant les vestiges archéologiques.

Fig. 11 – Post-abandonment shoreline reconstruction at the Vicus Augustanus. Age estimates of dune ridge and palaeo-beach samples included in the transect 1 age model and post-abandonment secondary aeolian deposits burying the archaeological remains are presented.Fig. 11 – Reconstitution du trait de côte après l’abandon du site du Vicus Augustanus. Sont mentionnés la datation approximative des sédiments prélevés dans les cordons dunaires et les anciennes plages (transect 1) ainsi que les dépôts éoliens recouvrant les vestiges archéologiques.

1: primary dune ridges; 2: secondary dune formations: 3: Laurentine shore sites, site codes refer to M.G. Lauro and A.J. Claridge (1998).
1 : cordons dunaires primaires ; 2 : dunes secondaires ; 3 : sites du littoral laurentin, les références des sites sont extraites de M.G. Lauro et A.J. Claridge (1998).

Conclusions

8Palaeo-shoreline reconstruction underpinned by a luminescence dated geochronological framework has been used to investigate the Geoarchaeological relationship of the Roman occupation of the Laurentine Shore within the meso- and macro-scale context of the Tiber Delta’s progradational development during the late Holocene. The value of the transect approach is that it has permitted a multi-scale context to be examined, placing the meso-scale archaeological remains within the context of macro-scale sea level change, Tiber catchment sediment supply and regional climate. An analysis entirely focused on the Roman archaeological remains would have failed to yield such useful information thus facilitating increasing confidence in the Geoarchaeological interpretation of the site concerned. In addition specific geoarchaeological and geomorphological conclusions can be drawn: i) Increased phases of Tiber delta progradation occurred during the Roman period (2217-1510 BP) and also within the last 500 years BP; ii) The seaward elements of the second phase, 1900-1750 BP (AD 100-250), of settlement at the Vicus Augustanus are constructed onto ground not yet fully deposited during the first phase of settlement, 2010-1950 BP, 10 BC-AD 50) due to episodic Tiber Delta progradation; iii) At least 9 phases of dune ridge development are recorded on the Tiber delta. A combined geochronology constructed from artefacts, radiocarbon dating of interdune marsh sediments (delineating dune ridge phases; Giraudi et al., 2009) and direct dating of dune ridge sands by optically-stimulated luminescence (this study) has provided a coherent record of dune ridge formation driven by the development of the Tiber delta during the mid- to late Holocene; iv) Due to delta retreat phases in the central Tiber delta area and increased preservation within the Castelporziano Estate, an additional two phases of dune ridge formation have been recovered. These rapid phases occur at 3448-3000 BP (CP IV) and 775-669 BP (CP VIII).

Haut de page

Bibliographie

Aitken M.J. (1998)An Introduction to Optical Dating: The dating of Quaternary sediments by the use of photon-stimulated luminescence. Oxford University Press, Oxford, 267 p.

Aldrete G.S. (2007)Floods of the Tiber in ancient Rome. Johns Hopkins University Press, Baltimore, 339 p.

Amorosi A., Milli S. (2001) – Late Quaternary depositional architecture of Po and Tevere river deltas (Italy) and worldwide comparison with coeval deltaic successions. Sedimentary Geology 144, 357-375.

Amorosi A., Ricci Lucchi M., Rossi V., Sarti G. (2009) – Climate change signature of small-scale parasequences from Lateglacial-Holocene transgressive deposits of the Arno valley fill. Palaeogeography, Palaeoclimatology, Palaeoecology 273, 142-152.

Arnaud-Fassetta G., Provansal M. (1999) – High frequency variations of water flux and sediment discharge during the Little Ice Age (1586-1725 AD) in the Rhône Delta (Mediterranean France). Relationship to the catchment basin. In Garnier J., Mouchel J.-M. (Ed.): Man and River Systems. The Functioning of River Systems at the Basin Scale. Developments in Hydrobiology 146, reprinted from Hydrobiologia 410, 241-250.

Bellotti P., Chiocci F.L., Milli S., Tortora P., Valeri P. (1994) – Sequence Stratigraphy and Depositional Setting of the Tiber Delta - Integration of High-Resolution Seismics, Well Logs, and Archaeological Data. Journal of Sedimentary Research Section B-Stratigraphy and Global Studies 64, 416-432.

Bellotti P., Milli S., Tortora P., Valeri P. (1995) – Physical stratigraphy and sedimentology of the Late Pleistocene-Holocene Tiber Delta depositional sequence. Sedimentology 42, 617-634.

Bellotti P., Caputo C., Davoli L., Evangelista S., Garzanti E., Pugliese F., Publio V. (2004) – Morpho-sedimentary characteristics and Holocene evolution of the emergent part of the Ombrone River delta (southern Tuscany). Geomorphology 61, 71-90.

Bellotti P., Calderoni G., Carboni M.G., Di Bella L., Tortora P., Valeri P., Zernitskaya V. (2007) – Late Quaternary landscape evolution of the Tiber River delta plain (Central Italy): new evidence from pollen data, biostratigraphy and 14C dating. Zeitschrift fur Geomorphologie 51, 505-534.

Bersani P., Bencivenga M. (2001) Le Piene del Tevere a Roma: Dal V secolo a.C. all'anno 2000. S.i.e.m.n. Dipartimento per i servizi tecnici nazionali. Presidenza del Consiglio dei Ministri, Roma, 100 p.

Bicket A.R. (2009) Reconstructing the Holocene coastal development of the Laurentine Shore, Castelporziano, Lazio, Italy. Loughborough University (Ph.D Thesis), Loughborough, 263 p.

Blair M.W., Yukihara E.G., McKeever S.W.S. (2005) – Experiences with single-aliquot OSL procedures using coarse-grain feldspars. Radiation Measurements 39, 361-374.

Blott S.J., Pye K. (2008) – Particle shape: a review and new methods of characterization and classification. Sedimentology 55, 31-63.

Duller G.A.T. (1991) – Equivalent dose determination using single aliquots. International Journal of Radiation Applications and Instrumentation. Part D. Nuclear Tracks and Radiation Measurements 18, 371-378.

Fleming S.J. (1971) – Thermoluminescent Dating Principles and Application. Naturwissenschaften 58, 333-338.

Garzanti E., Canclini S., Foggia F.M., Petrella N. (2002) – Unraveling Magmatic and Orogenic Provenance in Modern Sand: The Back-Arc Side of the Apennine Thrust Belt, Italy. Journal of Sedimentary Research 72, 2-17.

Ghilardi M., Desruelles S. (2009) – Geoarchaeology: where human, social and earth sciences meet with technology. Surveys And Perspectives Integrating Environment and Society (S.A.P.I.E.N.S.), 2.2., 1-9.

Giraudi C., Tata C., Paroli L. (2009) – Late Holocene evolution of Tiber river delta and geoarchaeology of Claudius and Trajan Harbor, Rome. Geoarchaeology 24, 371-382.

Heinzelmann M. (1998) – Beobachtungen zur suburbanen Topographie Ostias. ein orthogonales Strassensystem im Bereich der Pianabella. Römische Mitteiungen 105, 175-225.

Hesp P. (2002) – Foredunes and blowouts: initiation, geomorphology and dynamics. Geomorphology 48, 245-268.

Hesp P.A., Martínez M.L. (2007) – Disturbance processes and dynamics in coastal dunes. In Johnson E.A., Miyanishi K. (Ed.): Plant Disturbance Ecology: the Process and Response. Academic Press, Burlington, 215-247.

Keay S.J. (2005) Portus: an archaeological survey of the port of Imperial Roman. Archaeological monographs of the British School at Rome, 15. British School at Rome: Ministero per i beni e le attivitá culturali, Soprintendenza per i beni archeologici di Ostia, London, 360 p.

Kraft J.C., Bückner H., Kayan I., Engelmann H. (2007) – The geographies of ancient Ephesus and the Artemision in Anatolia. Geoarchaeology 22, 121-149.

Lambeck K., Antonioli F., Purcell A., Silenzi S. (2004a) – Sea-level change along the Italian coast for the past 10,000 yr. Quaternary Science Reviews 23, 1567-1598.

Lambeck K., Anzidei M., Antonioli F., Benini A., Esposito A. (2004b) – Sea level in Roman time in the Central Mediterranean and implications for recent change. Earth and Planetary Science Letters 224, 563-575.

Lauro M.G. (Ed.) (1985)Castelporziano I : Campagna di scavo e restauro 1984. Viella, Rome, 78 p.

Lauro M.G. (Ed.) (1988) – Castelporziano II : Campagna di scavo e restauro 1985-1986. Viella, Rome, 98 p.

Lauro M.G. (Ed.) (1998)Castelporziano III : Campagna di scavo e restauro 1987-1991. Viella, Rome, 174 p.

Lauro M.G., Claridge A.J. (1998) – Litus Laurentium : carta archeologica della zona litoranea a Castelporziano. In Lauro M.G. (Ed.) : Castelporziano III : Campagne di scavo e restauro 1987-1991. Soprintendenza Archeologica di Ostia - Viella s.r.l., Roma, 39-62.

Mastronuzzi G., Sanso P. (2002) – Holocene coastal dune development and environmental changes in Apulia (southern Italy). Sedimentary Geology 150, 139-152.

McKeever S.W.S. (1985) Thermoluminescence of solids. Cambridge Solid State Science Series. Cambridge University Press, Cambridge, 376 p.

Meiggs R. (1973) Roman Ostia. Clarendon Press, Oxford, 622 p.

Miyanishi K., Johnson E.A. (2007) – Coastal Dune Succession and the Reality of Dune Processes. In: Johnson E.A., Miyanishi K. (Ed.): Plant Disturbance Ecology: the Process and the Response. Academic Press, Burlington, 249-282.

Olley J.M., Murray A., Roberts R.G. (1996) – The effects of disequilibria in the uranium and thorium decay chains on burial dose rates in fluvial sediments. Quaternary Science Reviews 15, 751-760.

Otvos E.G. (2000) – Beach ridges - definitions and significance. Geomorphology 32, 83-108.

Preusser F. (2003) – IRSL dating of K-rich feldspars using the SAR protocol: Comparison with independent age control. Ancient TL 21, 17.

Psuty N. (2008) – The Coastal Foredune: A Morphological Basis for Regional Coastal Dune Development. In Martínez M.L., Psuty N.P. (Ed.): Coastal Dunes. Ecology and Conservation. Springer, New York, 11-27.

Rendell H.M., Claridge A.J., Clarke M.L. (2007) – Late Holocene Mediterranean coastal change along the Tiber Delta and Roman occupation of the Laurentine shore, central Italy. Quaternary Geochronology 2, 83-88.

Sanderson D.C.W., Murphy S. (in press) – Using simple portable OSL measurements and laboratory characterisation to help understand complex and heterogeneous sediment sequences for luminescence dating. Quaternary Geochronology.

Stewart I.S., Morhange C. (2009) – Coastal Geomorphology and sea-level change. In Woodward J.C. (Ed.): The Physical Geography of the Mediterranean. Oxford University Press, Oxford, 385-413.

Haut de page

Annexe

  

Version française abrégée

Le littoral laurentin est l’ancienne ligne de rivage datée de l’époque romaine (environ 27 av. J.-C.). Il est matérialisé dans le paysage par la présence d’anciennes villas, de routes et de structures archéologiques diverses. Cet ancien trait de côte est aujourd’hui fossilisé, à environ 1 km à l’intérieur des terres (Fig. 1), en raison de la progradation du delta du Tibre depuis la fin de l’époque romaine (Bellotti et al., 2007, Rendell et al., 2007). En replaçant les sites romains dans un contexte morphologique évolutif lié à la mobilité ponctuelle du trait de côte, il a fallu adopter une échelle de travail plus large que celle de l’intra-site pour cerner précisément l’évolution des paysages littoraux. En particulier, il a fallu prendre en compte les variations eustatiques holocènes régionales ainsi que le phénomène de progradation deltaïque favorisant le développement de cordons littoraux parallèles (Otvos, 2000 ; Mastronuzzi and Sanso, 2000 ; Giraudi et al., 2009 ; Fig. 2).

Trois transects ont été réalisés depuis l’intérieur des terres en direction du littoral et des sédiments ont été prélevés sur le cordon dunaire actuel ainsi que sur les cordons fossilisés par la progradation du delta du Tibre (Fig. 2). La Fig. 3 représente les lieux d’échantillonnage ainsi que les profils stratigraphiques établis d’après les analyses sédimentologiques. Des relevés topographiques ont également été effectués pour révéler la morphologie des cordons dunaires et il a ainsi été possible d’exprimer les altitudes obtenues par rapport au niveau actuel de la mer et par rapport à celui estimé pendant la période romaine (Lambeck et al., 2004a ; Lambeck et al., 2004b). Les transects réalisés dans le secteur du Vicus Augustanus sont représentés sur les Fig. 2 à Fig. 4. Les analyses granulométriques ont été effectuées à l’aide d’un granulomètre laser de type Coulter LS230 et permettent de renseigner sur le mode granulométrique utilisé pour les datations par luminescence. Le traitement des données a été opéré à partir du logiciel GRADISTAT v.5 (Blott and Pye, 2001).

Le Vicus Augustanus a été construit en deux phases principales : la première débute vers 2010-1950 BP (Fig. 10), des dépôts sableux prélevés sous le site archéologique principal, et qui font désormais partie d’un cordon dunaire situé à environ 950 m du littoral actuel (Fig. 11), ont été datés 2050±240 (CP895, Rendell et al., 2007). La seconde phase, essentiellement consacrée à l’extension vers la mer des infrastructures se déroule vers 1750-1900 BP (Rendell et al., 2007 ; Fig. 10). D’après les résultats des datations, il apparait clairement que la ligne de rivage était à 830 m à l’intérieur des côtes actuelles et cette paléo ligne de rivage est datée d’environ 1750 BP. Notre travail a également permis d’obtenir un rythme minimum de progradation de 0,14 m/an pendant la période d’occupation du Vicus Augustanus. La progradation du delta du Tibre a donc été très marquée et a favorisé, à l’échelle décennale, un isolement dans les terres de ce site. Si l’on détaille les différentes phases d’occupation la première (2010-1950 BP) enregistre une avancée du trait de côte d’environ 8 m alors que durant la seconde phase (1925-1580 BP), l’estimation est d’environ 48 m.

Les principales conclusions des études géoarchéologique et géomorphologique menée sur le littoral laurentin sont les suivantes : la période d’avancée significative du delta du Tibre est observée pendant la période d’occupation romaine (2217-1510 BP). En raison de l’avancée rapide du trait de côte, les infrastructures du Vicus Augustanus regardant vers la mer et datées de la seconde phase d’aménagement, reposent sur un substrat peu consolidé et pas entièrement déposé pendant la première phase de construction du site (2010-1950 BP). La formation d’au moins neuf cordons dunaires a été interprétée d’après les résultats des datations par le radiocarbone (Giraudi et al., 2009) et le présent article confirme, par la méthode de datation par luminescence, cette géochronologie. En raison de la position d’abri de la région de Castelporziano, située dans la partie méridionale du delta du Tibre, deux cordons dunaires supplémentaires (par rapport à la partie centrale du delta, Giraudi et al., 2009) ont pu être identifiés et datés précisément. Leur période de formation intervient en 3448-3000 BP (CP IV) and 775-669 BP (CP VIII).

Haut de page

Table des illustrations

Titre Fig. 1 – Castelporziano Estate location map based on Shuttle Radar Topography Mission (SRTM) 90 m resolution DEM.Fig. 1 – Carte de localisation de la région de Castelporziano réalisée à partir des données Shuttle Radar Topography Mission (SRTM, précision du MNT : 90 m).
Légende Deltaic and fluvial sediments characterise low-lying areas adjacent to sea level and are completely constrained by the altitudinal range 0-17 m as presented here. Altitudinal ranges are categorised by quantiles using ArcGIS 9.2. The most low-lying coastal areas recorded as below sea level also preserve the form of the Tiber Delta lagoons, now drained. WGS ‘84 coordinates are presented.Les dépôts deltaïques et fluviatiles caractérisent les secteurs de faible dénivellation proches du niveau actuel de la mer et sont compris entre 0 et 17 m d’altitude. Les classes d’altitude ont été définies grâce au logiciel ArcGIS 9.2, sous formes de quantiles. Les zones les plus déprimées d’un point de vue topographique sont localisées le long du littoral et sont localisées sous le niveau moyen de la mer, ce sont d’anciennes lagunes désormais drainées. Le système de coordonnées est exprimé en WGS’84.
URL http://geomorphologie.revues.org/docannexe/image/7720/img-1.jpg
Fichier image/jpeg, 256k
Titre Fig. 2 – (A) Coastal dune ridge morphology within the Castelporziano Estate. (B) Sampling locations and dune ridge topography transects A-A’ and B-B’.Fig. 2 – (A) Morphologie des cordons dunaires dans l’Etat de Castelporziano. (B) Localisation des échantillons et des transects topographiques A-A’ et B-B’ des cordons dunaires.
Légende Key archaeological sites mentioned in the text are highlighted. 1: Castelporziano Estate; 2: primary dune ridges; 3: secondary dunes. Primary dune ridges (2) are those believed to be directly linked to shoreline progradation driven by the development of the Tiber Delta, i.e., they are relict foredunes. Secondary dunes (3) are characterised by in situ remobilisation (such as the large parabolic dune that interrupts the primary dune ridge record seaward of the archaeological remains), or by dune ridges that are stratigraphically inconsistent following formation in inland locations and are therefore not directly linked to beach processes, i.e., cannot be considered as relict foredunes.Les principaux sites archéologiques cités dans le texte sont indiqués. 1 : région de Castelporziano ; 2 : cordons dunaires primaires (liés à la progradation du delta du Tibre) ; 3 : dunes secondaires (formation non liée à des dynamiques littorales). Les cordons dunaires primaires (2) sont ceux directement liés à la progradation du delta du Tibre et qui se sont édifiés à mesure que le trait de côte avançait, ils sont désormais fossilisés à l’intérieur des terres. Les dunes secondaires (3) sont caractérisées soit par une remobilisation in situ des sédiments soit par une formation en retrait du littoral, qui ne sont donc pas associés à des dynamiques littorales.
URL http://geomorphologie.revues.org/docannexe/image/7720/img-2.png
Fichier image/png, 581k
Titre Fig. 3 – Sampling logs from selected dune ridge crests and palaeo-beach locations (CP04/1).Fig. 3 – Localisation des profils stratigraphiques des crêtes dunaires et des anciennes plages (CP04/1).
Légende Samples from remobilised or secondary dune ridges (i.e., not linked to foredune formation during delta progradation) are also included (CP03/3, CP08/2, CP08/3). Samples CP07/15 and CP07/17 are taken from ferricretised relict dune ridges; all other samples are derived from calcium-carbonate cemented aeolianites. Munsell colours refer to wet sediments. Shading refers to relative organic content with sampling primarily undertaken from clean sands. 1: low organic content; 2: moderate organic content; 3: high organic content; 4: surface vegetation; 5: extensive root penetration; 6: some root penetration; 7: pseudo-gley features / mottled appearance; 8: indurated sediment; 9: diagenetic glaebules; 10: building rubble; 11: Roman wall. Les échantillons provenant des dunes secondaires (formations non directement liées à la progradation deltaïque) sont également intégrés (CP03/3, CP08/2, CP08/3). Les échantillons CP07/15 et CP07/17 ont été prélevés dans des cordons dunaires indurés par l’oxydation du fer ; l’ensemble des autres sédiments sont des éolianites possédant un ciment composé de calcium-carbonates. Les couleurs réfèrent à celles du code Munsell et ont été identifiées à partir de sédiments humides. Les ombrages se réfèrent à la teneur en matière organique des échantillons avec un prélèvement réalisé sur des sables dégagés de toute impureté. 1 : faible teneur en matière organique ; 2 : teneur en matière organique moyenne ; 3 : haute teneur en matière organique ; 4 : végétation ; 5 : pénétration du système racinaire ; 6 : quelques ramifications racinaires ; 7 : pseudo gley bariolées ; 8 : sédiments indurés ; 9 : nodules liés à une diagenèse ; 10 : décombres de mur ; 11 : mur romain.
URL http://geomorphologie.revues.org/docannexe/image/7720/img-3.jpg
Fichier image/jpeg, 164k
Titre Fig. 4 – Dune ridge topographic transects. Fig. 4 – Transects topographiques des cordons dunaires.
Légende Location of transects is highlighted in Fig. 2B. Elevations are corrected to a Roman period sea level of -1 m RSL (Lambeck et al., 2004 a and b) to provide some context for assessing archaeological relationships to their contemporary sea level.La localisation des transects est mise en évidence sur la Fig. 2B. Les altitudes sont corrigées pour être exprimées par rapport au niveau marin de l’époque romaine, situé 1 m en dessous de l’actuel (Lambeck et al., 2004 a and b).
URL http://geomorphologie.revues.org/docannexe/image/7720/img-4.jpg
Fichier image/jpeg, 88k
Titre Fig. 5 – Dune ridge phase map from Castelporziano based on luminescence geochronology related to complimentary 14C chronology of C. Giraudi et al.(2009).Fig. 5 – Carte de localisation des cordons dunaires de la région de Castelporziano fondée sur l’utilisation combinée de datations par luminescence et par le radiocarbone 14C (Giraudi et al., 2009).
Légende 1: Castelporziano Estate; 2: dune ridges; 3: secondary dunes. Archaeological sites mentioned in the text are also highlighted.1 : région de Castelporziano ; 2 : cordons dunaires ; 3 : cordons dunaires secondaires. Les sites archéologiques mentionnés dans le présent article sont également représentés.
URL http://geomorphologie.revues.org/docannexe/image/7720/img-5.jpg
Fichier image/jpeg, 340k
Titre Fig. 6 – Transect 1 age model (samples and luminescence age determinations are shown in Fig. 2, Fig. 3, and tab. 1, respectively).Fig. 6 – Le modèle d’évaluation âge/distance pour le transect 1 (les échantillons prélevés et l’obtention des dates par la méthode de luminescence sont respectivement présentées sur la Fig. 2, la Fig. 3 et le tab. 1).
Légende 1: primary dune ridges; 2: secondary dune formations. Central ages are meaningful with respect to archaeological age control. Secondary dunes are sampled from aeolian sand overlying the Vicus Augustanus.1 : cordons dunaires primaires ; 2 : dunes secondaires. Les datations se corrèlent bien avec les données archéologiques. Des sables éoliens ont été échantillonnés dans des dunes de formation secondaire qui recouvrent le Vicus Augustanus.
URL http://geomorphologie.revues.org/docannexe/image/7720/img-6.jpg
Fichier image/jpeg, 76k
Titre Fig. 7 – Historical Tiber flooding frequency at Rome (adapted from Bersani and Bencivenga, 2001).Fig. 7 – Fréquence des épisodes d’inondation du Tibre à Rome (adapté de Bersani et Bencivenga, 2001).
URL http://geomorphologie.revues.org/docannexe/image/7720/img-7.jpg
Fichier image/jpeg, 104k
Titre Fig. 8 – Tiber Delta minimum progradation rates calculated from age ranges of luminescence dates from transect 1 that incorporates the palaeo-shoreline position of the Vicus Augustanus (tab. 1).Fig. 8 – Taux minimum de progradation du delta du Tibre calculés à partir des datations par luminescence effectuées sur des échantillons prélevés sur le transect 1, qui incorpore l’ancienne position littorale du Vicus Augustanus (tab. 1).
URL http://geomorphologie.revues.org/docannexe/image/7720/img-8.jpg
Fichier image/jpeg, 44k
Titre Fig. 9 – Preliminary reconstruction of the Laurentine Shore and central Tiber Delta during the 2nd century AD, following the expansion of Portus under Trajan incorporating data from S.J. Keay et al. (2005), M. Heinzelmann (1998) and P. Bellotti et al., (2007).Fig. 9 – Proposition de reconstitution du littoral laurentin et du delta du Tibre au IIè siècle ap. J.-C., après l’expansion de Portus pendant le règne de l’Empereur Trajan (d’après les données de S.J. Keay et al. (2005), M. Heinzelmann (1998) et P. Bellotti et al. (2007).
Légende 1: Portus; 2: Ostia Antica; 3: villa di Plinio; 4: the Laurentine shore; 5: Tiber lagoons.1 : Portus ; 2 : Ostia Antica ; 3 : villa di Plinio ; 4 : littoral laurentin ; 5 : système lagunaire consécutif de la progradation du delta du Tibre.
URL http://geomorphologie.revues.org/docannexe/image/7720/img-9.jpg
Fichier image/jpeg, 252k
Titre Fig. 10 – Vicus Augustanus major building phases (adapted from Rendell et al., 2007).Fig. 10 – Principales phases de construction du Vicus Augustanus (d’après Rendell et al., 2007).
URL http://geomorphologie.revues.org/docannexe/image/7720/img-10.jpg
Fichier image/jpeg, 304k
Titre Fig. 11 – Post-abandonment shoreline reconstruction at the Vicus Augustanus. Age estimates of dune ridge and palaeo-beach samples included in the transect 1 age model and post-abandonment secondary aeolian deposits burying the archaeological remains are presented.Fig. 11 – Reconstitution du trait de côte après l’abandon du site du Vicus Augustanus. Sont mentionnés la datation approximative des sédiments prélevés dans les cordons dunaires et les anciennes plages (transect 1) ainsi que les dépôts éoliens recouvrant les vestiges archéologiques.
Légende 1: primary dune ridges; 2: secondary dune formations: 3: Laurentine shore sites, site codes refer to M.G. Lauro and A.J. Claridge (1998).1 : cordons dunaires primaires ; 2 : dunes secondaires ; 3 : sites du littoral laurentin, les références des sites sont extraites de M.G. Lauro et A.J. Claridge (1998).
URL http://geomorphologie.revues.org/docannexe/image/7720/img-11.jpg
Fichier image/jpeg, 205k
Haut de page

Pour citer cet article

Référence papier

Andrew R. Bicket, Helen M. Rendell, Amanda Claridge, Peter Rose, James Andrews et Fiona S.J. Brown, « A multiscale geoarchaeological approach from the Laurentine shore (Castelporziano, Lazio, Italy) », Géomorphologie : relief, processus, environnement, vol. 15 - n° 4 | 2009, 241-256.

Référence électronique

Andrew R. Bicket, Helen M. Rendell, Amanda Claridge, Peter Rose, James Andrews et Fiona S.J. Brown, « A multiscale geoarchaeological approach from the Laurentine shore (Castelporziano, Lazio, Italy) », Géomorphologie : relief, processus, environnement [En ligne], vol. 15 - n° 4 | 2009, mis en ligne le 01 janvier 2012, consulté le 28 avril 2017. URL : http://geomorphologie.revues.org/7720 ; DOI : 10.4000/geomorphologie.7720

Haut de page

Auteurs

Andrew R. Bicket

Department of Geography, Loughborough University, Loughborough, LE11 3TU, UK (A.Bicket@lboro.ac.uk)

Helen M. Rendell

Department of Geography, Loughborough University, Loughborough, LE11 3TU, UK (A.Bicket@lboro.ac.uk)

Amanda Claridge

Department of Classics, Royal Holloway, University of London, Egham, Surrey, TW20 0EX, UK

Peter Rose

Centre for Computing in the Humanities, Kings College London, 26-29 Drury Lane, London, WC2B 5RL, UK

James Andrews

Department of Classics, Royal Holloway, University of London, Egham, Surrey, TW20 0EX, UK

Fiona S.J. Brown

Department of Geography, Loughborough University, Loughborough, LE11 3TU, UK (A.Bicket@lboro.ac.uk)

Haut de page

Droits d’auteur

© Groupe français de géomorphologie

Haut de page
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Groupe français de géomorphologie
  • Revues.org