Rivers in Prehistory

edited by Andrea Vianello

Archaeopress Archaeology Archaeopress Publishing Gordon House 276 Banbury Road Oxford OX2 7ED www.archaeopress.com

ISBN 978 1 78491 178 2 ISBN 978 1 78491 179 9 (e-Pdf)

Cover photo: Giant king fish (Caranx ignobilis) gather in preparation for a unique migration up the Mtentu river. The Mtentu river is one of few remaining pristine wildernesses in South Africa, so the Africa crew avoided using disruptive motor boats, instead paddling to the location each day in a canoe. Image elaborated by Cornelia Stancu.

All rights reserved. No part of this book may be reproduced, stored in retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the copyright owners.

Printed in England by Oxuniprint, Oxford This book is available direct from Archaeopress or from our website www.archaeopress.com Table of Contents

Preface ��iii Introduction �� 1 Rivers, Where Humankind Meets Nature �� 7 Rivers of civilisation ��23 Case Studies ��35 Dugouts from North Patagonia (center-south of Chile): Sailing on Trees ��37 Nicolás Lira S.

Exploitation of the Aquatic Resources of Lake Lubāns and Its Hydrological Regime during the Stone Age ��51 Ilze B. Loze

A River Runs Through It: the Semiotics of Göbekli Tepe’s Map (an Exercise of Archaeological Imagination) ��65 Dragoş Gheorghiu

Rivers, Human Occupation and Exchanges Around the Late Bronze Age Settlement of Frattesina (ne Italy) ��77 Paolo Bellintani and Massimo Saracino

People of the Waters in Northern Italy ��89 Andrea Vianello

The Perennial Rivers and the Changing Settlement Patterns on the Two Sides of the Tiber in Central Italy – the Case Studies of Nepi and Gabii ��103 Ulla Rajala

Bronze Age Barrow Complexes on the Lincolnshire Fen Margin ��125 Peter Chowne

Roads, Routes and Ceremonies: the Fenland Superhighway ��137 Tim Malim

Continuity of Seasonal Access and Occupation on the Turloughs of Ireland ��151 Amy Bunce

An Approach to the Fluvial Networks of the Papaloapan Basin: the Use of the Lower Papaloapan, México, from the Pre-hispanic Period to Early Xx Century ��159 Edith Ortiz Díaz

List of Authors and Contacts ��166

i Copyright material: no unauthorized reproduction in any medium

The Perennial Rivers and the Changing Settlement Patterns on the Two Sides of the Tiber in Central Italy – the Case Studies of Nepi and Gabii

Ulla Rajala University of Cambridge

Introduction (1980) covered the area of the I.G.M. map sheet of Sutri immediately west from Nepi and several less extensive This paper compares two geographically contrasting areas studies covered selected smaller areas (e.g. Selmi 1978; in central Italy with similarities in fluvial hydrology. di Gennaro and Stoddart 1982; Camilli et al. 1995; Cifani It is based on my PhD thesis (Rajala 2002) that showed and Munzi 1995), there had not been a unified survey of certain correlations between fluvial characteristics and the territory. human settlement in my two case study areas (Fig. 1) on the two sides of the Tiber in central Italy. The original In 1999, the areas in the close proximity to Nepi were material comes from the Nepi Survey Project, whilst the walked, comprising 92 units, mainly fields, parts of fields comparative material originates from a desktop case study or pastures. In September 2000, the main focus was on on the Gabii and Collatia area. selected areas near the boundaries of the study area, and 141 units were sampled. These units were located mainly The town of Nepi (Roman Nepet) is located c. 45 to the west, northwest, northeast and south of the town. The kilometres north-west of Rome (Fig. 1). In Antiquity it was units, 233 in total (Fig. 2), covered 8% of the rectangular on the boundary of the Faliscan area in south-east Etruria territory around Nepi. Available land areas were studied (Francocci 2006). The town overlooks a gorge on the by field walking at intervals of 10–20 metres with all western boundary of a highly dissected landscape that is archaeological material collected and recorded. The main part of the Treia watershed (see Rajala 2012: 112). The site aim was to analyze continuous distributions across the of Gabii (Guaitoli 1981) lies c. 20 kilometres east of Rome field units (for the field methods of the Nepi Survey, see on a volcanic crater rim overlooking the drained lake of Mills and Rajala 2011, 149–151). Castigliano and the drained lowlands of Pantano Borghese in the south (De Santis 1992, 11). It was in the past one of I complemented the data in the Nepi area with the published the cities belonging to the Latin League (see e.g. Smith or edited information (c.f. Potter n.d.; di Gennaro 1995b; 1996, 185–223) but the site is now an archaeological Rajala 2002). I also compiled a comparative data set from heritage area. The Nepi area belongs to the internal hilly the hinterland of Gabii/Collatia from the literature. The zone, whereas Gabii and its surroundings lie on the plains British School at Rome carried out fieldwork in the area (Statistiche ambientali 1998, Fig. 1.1). briefly (Kahane and Ward-Perkins 1972; Kahane 1973), but the main body of rural evidence came from the survey The Nepi Survey Project (di Gennaro et al. 2002; Rajala volume of Collatia (Quilici 1974). Further information was 2006), which provided the original material for my thesis, collected from different sources (e.g. Quilici N.D.; Mari was conducted under the directorship of Simon Stoddart, 1991; Carboni and Ragni 1984; Belardelli et al. 1986). within the umbrella of the Tiber Valley Project (Patterson The town area of Gabii was surveyed by the Department and Millett 1998; Patterson et al. 2000; Patterson 2004). of Ancient Topography of the University of Rome “La The aims of the Nepi Survey Project included collecting Sapienza” (Guaitoli 1981). I also included information a coherent body of data for a study of the long-term from the excavations of the Apennine site at Castiglione development of the Faliscan centre of Nepi and its territory (Bietti Sestieri 1984; De Santis 2001; c.f. Bietti Sestieri (Edwards et al. 1995; Lim et al. 1996; Harrison et al. 2004; and De Santis 2000), the Iron Age cemeteries at Castiglione di Gennaro et al. 2008) with an emphasis on exploring and Osteria dell’Osa (Bietti Sestieri 1992a; Bietti Sestieri pre-Roman distributions (Rajala 2007; 2012) in order to and De Santis 2000; De Santis 2001) and the Archaic and address the bias towards the Roman period during earlier Roman town itself (e.g. Majerini and Musco 2001). blanket surveys (c.f. Rajala et al. 1999). The time period discussed here ranges from the Paleolithic The Nepi Survey recovered samples from all sectors of the to the Roman period. The main discussion concentrates on territory by drawing transects along the cardinal directions the late prehistory and the protohistoric period consisting radiating from the town along the grid of the Istituto of the Orientalising and Archaic periods (see Table 1). The Geografico Militare (I.G.M.) map whereas the earlier absolute chronology of later central Italian prehistory still surveys in the Nepi area had concentrated on certain remains problematic, even if dendrochronology and new parts of the landscape, such as the road network (see dating evidence have brought generally new precision Fig. 2; Fredriksen and Ward-Perkins 1957) or the Treia (c.f. Bietti Sestieri 1996: 185–93; Pacciarelli 1996; 2000). watershed east of Nepi (Potter n.d.). Although Morselli There is still no consensus on the end of the Bronze Age

103 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Fig. 1. The map of central Italy (with the main sites in the text).

(c.f. Pacciarelli 2000, 68, Fig. 38; Nijboer et al. 2002, Table Romans tried to conquer Falerii Veteres (Civita Castellana) 1), although the different arguments have approached each in 394 BC, but this attempt has been interpreted to have other (c.f. Bartoloni and Delpino 2005). As we are mostly been unsuccessful (Scullard 1967, 73; Shotter 1976). The dealing with the results of the analyses of unstratified Romans founded the colonies at Sutrium and Nepet in surface materials, I will only refer to the relative periods order to divide different ethnic areas. Sutrium was founded or phases in the discussion and I leave the discussion on most likely after the end of the battles with the Gauls in the implications of these different chronologies for another the early 4th century BC (D.S. 14.117.4) and Nepet in occasion. The current traditional and calibrated date ranges 383 BC (Livy 6,21,4). The status of the Faliscan area has for the periods are listed in Table 1. been discussed at length (e.g. Harris 1971, 41–43; Shotter 1976, 31–32 vs. Cornell 1995, 313). In any case, the Whilst Gabii apparently surrendered to Rome around the most remaining independent central Etruscan city–states time of the foundation of the Republic (Livy 1,54) and the collapsed quickly after the battle of Sentinum in 295 BC Latins entered into a treaty with Rome in 493 BC (Livy (c.f. Carandini 1985, 49; David 1996, 20). Falerii Veteres 2,33), the beginning of the Roman period in the Faliscan was not ultimately conquered until 241 BC; this defeat led area is somewhat ambivalent. Different city-states fell at to its abolishment and the forced move of any surviving different times and the surrounding main non-Faliscan inhabitants to the new Roman town of Falerii Novi. Thus, cities of Veii and Capena both collapsed in 396 BC. The it can be concluded that Nepi had already surrendered in

104 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

Fig. 2. The survey units in the territory of Nepi (with the main place names). the early 4th century BC and had a privileged position (Fig. 3) and similarly Fosso dell’Osa near Gabii and most in the area (Rizzo 1992, 2). Since the classical material of the main river valleys east and west of the ancient town culture at the settlement sites is very similar to that in are fed by springs as well (Fig. 4). Hydrogeologically, the Rome throughout the region, and thus, the Archaic dates areas near Nepi belong to the groups of the Monti Vulsini, discussed here are more likely to refer to the period pre- Cimini and Sabatini whereas the Gabii area is part of the 530/520 BC, unless exceptionally diagnostic. system of the Colli Albani (Boni et al. 1988). The most important feature in the local hydrology in both areas is the Hydrology in the case study areas perennial nature of the main rivers fed by linear springs.

All main rivers between Civita Castellana and Lake Vico in The canyon-like river valleys in the eastern side of the the Faliscan area, to which Nepi geographically belonged Nepi area (see Fig. 5.A.) show beyond doubt where the (Edwards et al. 1995), discharge themselves first into the river valleys were located for the last tens of thousands of Treia and ultimately to the Tiber. On the other side of the years of the Holocene. They mark the long-term riverine Tiber at Gabii all main rivers discharge to the Aniene and landscape at the large scale in the Faliscan area. However, similarly ultimately to the Tiber. In the Nepi area the main it is almost impossible to know with the current level of rivers, Fosso del Ponte del Castello, Fosso del Cerro, Fosso knowledge where the rivers meandered inside the canyons. della Massa and Fosso del Cerreto, are rich with springs Elsewhere in south Etruria the rivers flow in deep, steep-

105 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Table 1. The periodization of later prehistory and early historical times in central Italy.

Calibrated dates Period Traditional dates Bietti Sestieri 1996/Pacciarelli 1996 [Pacciarelli 2000; 2005]

Chalcolithic (Eneolithic) 2700–1800 BC 3000–2300 bc

Early Bronze Age 1800–1600 BC 2300–1700 bc

Middle Bronze Age 1600–1300 BC 1700–1350 bc

Early Middle BA 1600–1300 BC

Apennine Middle BA 1400–1300 BC

Recent Bronze Age 1300–1200/1150 BC 1350–1200 bc

Final Bronze Age 1200/1150–900 BC 1200–1020/960 [950/925] bc

Early Iron Age 900–730/720 BC 1020/960 [950/925]–780/725 [730/725] bc

Orientalising 730/720 – 580 BC

Early 730/720– BC

Middle –640/630 BC

Recent 640/630–580 BC

Archaic 580– BC

(Classical) (530/520– BC)

Fig. 3. The perennial rivers in the Nepi area (with the main Bronze Age and Iron Age centres).

106 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

Fig. 4. The perennial rivers in the Gabii area (with the reconstructed lakes).

Fig. 5. A. Nepi and the ravine photographed from Castel Sant’Elia. B. The upper town of Gabii across the valley of the former Lago di Castiglione (photographs by U. Rajala).

107 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Table 2. The erosion rates in central Italy (Judson 1968).

Veii 800/600 BC to present 30 cm/1000 years

Treia 1000–450 BC 100 cm/1000 years

Via Prenestina 300/100 BC to present 30 cm/1000 years walled trenches created by recent periods of erosion, but three thousand years earlier. The dating of the beginning Judson (1963, 899) did not hypothesize on the age of the of rapid erosion is problematic, because one particular valleys. The erosion rates suggested by the lake cores from historical event was used to fix it. The start of erosion was the basins of Baccano and Monterosi in southern Etruria linked with the building of the Roman Via Cassia (Cowgill were low, whereas the rates from the archaeological sites and Hutchinson 1970, 43, 42–44, Fig. VII–2; Hutchinson in the area were tenfold (c.f. Judson 1968, 1445). This and Cowgill 1970, 168, Fig. XIII–1; Ward-Perkins 1970). disparity was explained by the assumption that these areas Consecutively, this subjectively chosen dating was later were heavily wooded before the Roman times. mistaken as a proof of the change the Romans brought to the area (Potter 1979). In reality, the whole interpretation The erosion rates in the Treia valley were higher due rested on a circular argument; the erosion could already to the colluvial nature of the area (c.f. Table 2), and the have started during the pre-Roman times. In fact, in Treia system shows a complex history of alluvial episodes the Mediterranean basin the silting of ports had already (Cherkauer 1976). The floodplain at the base of the ravines became a problem in the 8th century BC (Vita-Finzi 1969; is a recent formation and there are indications that the Brückner 1986; Sorriso-Valvo 1998, 43). Roman surface lies as deep as four metres under the modern surface. There are indications of meander migration from On the other side of the Tiber, in the valley of the lake the 7th to the 5th centuries BC (Brown and Ellis 1995, 56). Castiglione (see Fig. 5.B.) the proportion of tree and In the Valchetta river valley at Veii, the late Roman layers shrub pollen kept growing during the period from the later were buried under six metres of silt (Jones 1960) and near Mesolithic until the Middle Bronze Age. The climate was Tomba di Nerone in the Fossa della Crescenza, the alluvium temperate and wet. In the end of the Early Bronze Age, layers were similarly thick (Judson 1963; 1983). The near the beginning of the Middle Bronze Age, the tree and Roman river valleys seem to have been shallow channels, shrub pollen proportions were declining and herbs became actively migrating and depositing bars of gravel (Brown more numerous. Oak (Quercus) was the dominant tree, but and Ellis 1995, 69); this was the situation also during the there were also traces of chestnut (Castanea). The results Archaic period. Only a detailed geoarchaeological coring point to the increase of Mediterranean vegetation and, thus, programme in the river valleys could help to reconstruct probable human impact, although no pollen of Cerealia has the micro fluvial landscape inside the ravines in a more been found. The humidity was decreasing and temperature detailed manner. increasing at the end of the Early Bronze Age (Alessio et al. 1986). During the Final Bronze Age the temperature There were notable global changes in climate during the was constant, but humidity was increasing until the late prehistory. During the second millennium BC, there climax during Latial IIB, the early Villanovian period, were several consecutive trends. Climate was deteriorating that was also the low point of temperatures. Afterwards, during the earlier part of the Bronze Age and was temperatures were increasing and by the Archaic period temporarily worse at the end of the 17th century BC, the the lake was diminishing in size (Segre 1992). worst year having a dendrochronological date of 1628 BC (Baillie 1998). Interestingly, this coincided with the earlier The recent climate data reveal some differences in the part of the Middle Bronze Age, whereas the warmer period case study areas. During the ten–year period 1987– was contemporary with the Apennine Bronze Age. During 1996 the average maximum temperature was 20.85 the Recent Bronze Age climate was deteriorating again degrees and the minimum temperature 9.1 degrees at the and the coldest period seems to have been contemporary Grosseto measuring station in southern Tuscany. During with the Final Bronze Age, the worst year being 1159 BC the summer, temperature can reach the forties, but in (Baillie 1998). However, during the later Final Bronze Age the winter frosty temperatures are recorded every year climate was improving again. The colder and more humid (Statistiche ambientali 1998). In Civita Castellana and subatlantic period ended around 500 BC with the beech Nepi, the July temperatures are higher than at Grosseto, cover reaching its maximum around 600 BC (Lorenzoni whereas the January temperatures are lower (c.f. Lorenzoni et al. 1984, 105). et al. 1984, Tab. 4; Lulli et al. 1988, tabella 1). Grosseto is nearer the coast and thus its location can be taken as The climatic and land use history in the south-eastern comparable to that of Gabii where the temperatures are Etruria is still dependant on one C14 dating from the different from urban centre of modern Rome given in the Lake Monterosi core; crucially this date does not relate statistics. When comparing the average number of sunny to the beginning of rapid erosion but is approximately hours, it is clear that southern Etruria (2443.0 in Vigna di

108 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

Valle) gets slightly less sunlight than Latium (2468.0 in by analysing and comparing the geographic characteristics Roma–Ciampino near Gabii between 1992 and 1996). The of the units from which the finds came. Apart from the figures are equivalent to 54.7 and 55.4% relative sunniness descriptive statistics, a combination of suitable statistical (Statistiche ambientali 1998, tavola 1.5). The Tyrrhenian analyses (Wilcoxon rank-sum test, chi-squared test, coast and the area nearest to the centre of Rome belong Spearman’s Rho and Kendall’s Tau correlation coefficients) to the Mediterranean vegetation area where evergreen oak were used in order to find associations between find dominates and the average temperature is 15 degrees. The distributions and terrain attributes, extracted using Arc/ inner areas in Lazio, such as Nepi, belong to the Samnite Info and ArcView. The analysis of overall distributions system with mixed deciduous oak–dominated forests and complemented the discussions on concentrations and sites the average temperature of 13 degrees (Sorriso-Valvo of different periods. 1998, 46, Fig. 3.3). However, the annual mean temperature measured at Nepi was 15 degrees; this is half a degree less Very few flints collected in the Nepi survey were period- than at Civita Castellana (Lorenzoni et al. 1984, Tab. 4; specific. Most of the flints were from the long period Lulli et al. 1988, tabella 1). Nepi seems to be a winter between the Upper Palaeolithic and the Late Neolithic/ cold spot and summer hot spot in comparison with milder Early Bronze Age. Only a few artefacts were from earlier Gabii. phases of the Palaeolithic. As a consequence one cannot speak of sites in the same way as in relation to the later In recent times both case study areas have mean annual periods. One can, however, discuss concentrations in the precipitation between 751 and 1000 mm. Both areas are areas of Piano di Vallescura, Massa, San Paolo and Casale also surrounded with territories where the precipitations Galeotti (c.f. Fig. 6.A.). These areas represent relatively are higher (Ministero dei Lavori Pubblici 1959). The well the average find locations revealed by the project. 1922–1972 averages at Nepi show the minimum annual Statistical analyses showed that flints were found from precipitation of 670 mm and the maximum of 1329 mm flat areas often facing north and east. They were not found (c.f. Lorenzoni et al. 1984, Tab. 1; Lulli et al. 1988, tabella near the rivers, but from the plains. This explains why 2; Table 2.6). The real annual precipitation in central Italy flints were found nearer later road lines than any other can be considerably less, though; during the ten–year period find group. Clearly, land use was very different in hunter- 1987–1996 the lowest annual precipitation at Grosseto was gatherer communities. only 412.3 mm. During this period there were in average 62 rainy days (Statistiche ambientali 1998, tavola 1.4). Later prehistoric sites were not numerous, but it was The soils around Nepi stay dry from July to August while relatively easy to define a series of sites, to be discussed in in Poggio Mirteto east of the Tiber they are dry only in a separate article (Rajala, in preparation). It is noteworthy August (Lulli et al. 1988, Fig. 2, 89). These figures, even if that the Nepi survey revealed the extensive presence of recent evidence, suggest that the eastern case study area is underlying distribution of low density prehistoric pottery moister even today and may have been in the past as well. finds in the territory. This type of distribution has been The perennial rivers were likely to have more impact in previously observed in Greece (Bintliff et al. 1999) and the the drier Etruria throughout the summer months during the interpretation of the significance of these finds will require earlier periods with lower technical skill. a wider investigation into the character of prehistoric finds in central Italy in general. Settlement, land use and river valleys in the Nepi area All evidence points out that the settlement pattern between the Late Neolithic and early Middle Bronze Age was The results of the Nepi survey were analysed statistically; dispersed and all sectors in the plains were settled at this approach was only possible when all major material least at some point during this time span. However, the groups (flints and the dated pottery from the main periods) concentration of settlement from the Middle Bronze Age were grouped into four chronological groupings. It is very onwards cannot be denied and during the Final Bronze Age simplistic to group all prehistoric, ‘Faliscan’ or Roman the settlement is defined to the naturally defendable sites pottery into wider groupings, but most prehistoric pottery such as Il Pizzo at Nepi. The earlier dense settlement pattern was non-diagnostic belonging generically to the period was obviously connected with local water resources. Most between the Late Neolithic and early Middle Bronze Age, of the previously known sites (c.f. Belardelli et al. 2007; and Orientalizing and Archaic pottery was relatively scarce Rajala, in preparation) were situated in river valleys or by in the research area. Furthermore, Roman pottery was cliffs of the ravines. included in the analysis mainly for comparative purposes and any further future work will include the published Small amounts of prehistoric pottery were found from all Roman pottery assemblage (Mills and Rajala 2011). Thus, sectors of the territory (see Fig. 6.B.). Statistics showed wider groupings made analytical sense and helped to that prehistoric pottery was usually found from flatter characterize the areas of activity – or taskscapes (Ingold areas facing south from the area of higher elevations in 1993) – for different chronological cultural economic the west. Although prehistoric pottery was usually found entities (hunter–gatherers, early farmers, protohistoric at some distance from the water sources, the find locations communities, the Roman Empire; see Rajala 2012). The did not correlate with the proximity to rivers. However, distributions of these four major groups were compared they were significantly connected with the pathways to the

109 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Fig. 6. The find densities in the Nepi Survey Project. A. The weight density distribution of flints. B. The weight density distribution of prehistoric pottery. C. The weight density distribution of Archaic and Orientalizing pottery. D. The density of Roman pottery. river valleys but not to the later road network. The wide- settlement pattern was a consolidation and expansion of the scale distribution and varied find locations separated the Iron Age pattern (see Rajala 2005). There are indications of distribution of prehistoric pottery from the distribution of a two-level settlement pattern both during the 8th century Faliscan material. However, the location of Nepi and Il BC and the Orientalising period, but during the Archaic Pizzo started to be important since less prehistoric pottery period the third level was forming with the number of was found close to Nepi. smaller settlements rising. Although many of the sites were situated along the margins of the plateaux, all sites The density maps created for the Orientalizing and were relatively near the known ways of communications. Archaic material showed that the high concentrations of Furthermore, there seem to have been clusters of rural sites, ‘Faliscan’ pottery were rare (Fig. 6.C.). The comparison for example in the Piano di Vallescura and Grotta Arnaro of the distributions corrected and uncorrected with the areas. The settlement pattern at the end of the Archaic surface visibility showed that high quantities and high period consisted of the centre, boundary villages and rural weight densities were often in different areas. The weight farms at a distance from the centre. The pattern stresses the densities seemed to emphasize significant areas; the maximisation of agricultural production, organised from highest average weights connected with the field units the centre. with low visibility but with larger, diagnostic pieces. The survey evidence showed that the town was surrounded by Statistics showed that the Orientalising and Archaic an area dedicated to cemeteries and farming with little pottery was found from the areas of higher elevations in the domestic pottery in the immediate surroundings of the west where the slopes were gentle and facing south-east. town. The clearest evidence for rural settlements came Pottery finds correlated with small hillocks and ascending from the areas nearer the hypothetical boundaries of Nepi. slopes. Protohistoric pottery was clearly found farther This phenomenon suggests that inhabitants were directly away from the rivers than earlier prehistoric pottery, but involved in agriculture and the social structure was not it correlated both with the road cuttings leading into the fully urbanised. The town preserved its rural character. river valleys and the known road lines. Thus, on the basis The increased need for land was reflected in intensification of the evidence, the importance of the fluvial landscape further afield. in the rural everyday life was diminishing, whereas the importance of overland transport was increasing. The distribution of funerary finds near the naturally defended promontories in the fringes of the territory may The quantity and wide distribution of Roman finds suggest the existence of a series of boundary settlements overshadowed all other periods (Fig. 6.D.). Roman pottery in the territory. This would mean that the Orientalising correlated with lower elevations, gentler slopes and eastern

110 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

Fig. 7. A. Down in a ravine by the perennial river. B. The paved Roman road from Massa into the river valley and towards Castel Sant’Elia. and southern aspects. The amount of pottery collected Munzi 1995), then they may even lie at the foot of the correlated positively with the distances from the rivers cliffs like a cut chamber in the Fossitello, observed during and the road network. The distributions of protohistoric the Nepi Survey Project. Furthermore, at Falerii Veteres pottery and Roman pottery were very similar and showed and Narce, temples were located in the river valleys (c.f. continuity and increasing intensity in the use of the Stefani 1947; 1948, Fig. 1; De Lucia Brolli 1990). same areas. Nevertheless, there were clear differences in comparison with other find groups; flints were found more River valleys were used actively through different often from flatter plains and prehistoric pottery from the archaeological periods, not only in the present as hunting, western parts of the research area. fishing and trekking grounds. One can still see the terraces and stone walls marking a pre-20th -century field system The meaning of fluvial landscapes in the Nepi area in the Fosso di Ponte o del Castello valley below Castel Sant’Elia north/east from Nepi. In one of the river valleys Traditionally, archaeologists (e.g. Frederiksen and north-west from Nepi, in the Fosso Grottabella valley, Ward-Perkins 1957, 184) have underestimated the the Nepi Survey Project found a Medieval rock painting importance of river valleys as routes of communication picturing saints. This kind of shrine suggests that the in the Faliscan area. However, river valleys have been travellers were passing it by in the river valley. used for different purposes over the millennia. Many of the prehistoric settlement sites found by the Gruppo During modern times, the valley bottoms of the deep river Archaeologico Romano (c.f. Pennacchioni 1975; Selmi valleys in the Nepi area are easier to walk than the wooded 1978; Brunetti Nardi 1981) were in the river valleys. The areas in the plains. This is due to the microclimate and famous prehistoric Faliscan caves (Rellini 1920) were vegetation in the river valleys. Treetops create a canopy, located in the river valleys and this connection between which keeps valleys shady and ground vegetation low the prehistoric caves and river valleys also defined (Fig. 7.A.). The air is more humid and vegetation low, archaeological research. The perceived suitable locations but lush, not dry and spiky as on the plains. It is true that for settlement sites were heavily conceptualized; the torrential rains fill the rivers and together with the effects prehistorians were looking for prehistoric caves along the of erosion cause landfalls that can block the way and regional river network, whereas Frederiksen and Ward- change pathways abruptly. The relative difficulty and ease Perkins (1957) looked for Roman and Faliscan sites along of moving in the river valleys depend on the season; the the Roman roads (Rajala et al. 1999). The material from rivers are flooding during the winter rains, but during the the caves ranged from flints and Neolithic pottery to the summer only the rivers fed by perennial springs have some Middle Bronze Age and Recent Bronze Age pottery and water. included some Roman finds (c.f. Rellini 1920, 86–105; di Gennaro 1995b). I created a series of least-cost paths between the main centres in the region around Nepi (Rajala 2002). These During the Iron Age the Etruscans and Faliscans buried paths suggested that the Treia valley could have been an some of their dead in the river valley cemeteries. The important route of communication in the past. The fact cemeteries of Petrina B, Monte Lo Greco D and I Tufi at that no road lines were defined during the South Etruria Narce and Valle La Fata and Piano delle Granate at Veii Survey north of Narce by neither Frederiksen and Ward- were in the river valleys (Baglione and De Lucia Brolli Perkins (1957) or Potter (1992, Fig. 4) further points 1990, 70). Faliscan chamber tombs were also located in to the potential importance of the Treia valley in the the river valleys; if not along the edges of the plateaux communication network. The river terraces in the valleys overlooking the valley like at Tenuta Franca (Cifani and provide a platform for roads; even today the Via Cassia

111 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Fig. 8. Road cuttings in the Nepi area. follows Fosso del Pisciarello at Sutri and both the Roman to the river valleys (Fig. 8); these can be used as a proof of road line and modern main road run parallel on the terrace. the past routes. Many rock cuttings were modified by the Romans, as a section of road paving in a cutting at Castel The analysis of the archaeological finds in the Nepi area Sant’Elia proves (Fig. 7.B.). These access routes into the confirms that the relationship between the settlements river valleys and across the ravines were also important and different land routes was in place at least from the during the Medieval times (c.f. Frederiksen and Ward- Orientalising period onwards. However, these routes may Perkins 1957, 183–186). However, many could be dated have an even older history. The existence of minor Iron to the Faliscan period. At Nepi, a large complex of road Age centres such as La Ferriera (Duncan 1958, 119–120; cuttings existed at I Cavoni south of Pizzo. Another maze di Gennaro 1995a; 1995b, n. 71; Belardelli et al. 2007: of road lines and cuttings was at Pizzo Iella north-west no. 386, 312) near the Lake Monterosi in the south-west, from Castel Sant’Elia. The ancient road cuttings display possibly Casale Filissano (Potter n.d., E36a ) south-east multiple road levels and sometimes other features in the from Nepi and Monte Lombrica in the north-east near rock faces. Some are shallow enough to be supposed to Falerii Veteres (Iaia and Mandolesi 1993, n. 60; Belardelli be Medieval, like the cutting above Il Pizzo northeast et al. 2007, no. 261, 283) suggests that the importance of I Cavoni. However, the cuttings have never been of land transport was growing by the very end of later exhaustively surveyed and studied (although see Cifani prehistory with a need to control entry routes. However, et al. 2007; Tuppi 2010), and their origin may be much even during the Archaic period, people were still travelling earlier. through the river valleys as the existence of a tomb in the Fossitello valley suggests. This cut chamber probably Many road cuttings seem to be connected with tombs relates to a path connecting the river valley to the road in and cemeteries. The cuttings at Tenuta Franca, Poggio Fosso dell’Isola towards Falerii Veteres. Maggiore, Cerro and San Marcello (Frederiksen and Ward- Perkins 1957, Fig. 4) were short with a series of tombs in Road cuttings from different eras cut down into the rock the immediate surroundings or overlooking the cutting. faces in the Nepi area and natural pathways allow access The chamber and trench tombs at Vigna Penteriani (Stefani

112 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

1918) were located in a pre-Roman road cutting. At San Orientalising and Archaic pottery on one hand and Roman Marcello the chamber tombs could not be visited because pottery on another, but the levels of confidence were lower. of the vegetation, but the loculi-like horizontal trenches in the rock face at the beginning of the road cutting were The distribution of prehistoric pottery in relation to the road similar to those at Vigna Penteriani (Stefani 1918, Fig. cuttings and river crossings differed significantly from the 2). At Castello del Ponte Nepesino the road cutting was other distributions and from the distance values of all units. lined on both sides by a series of tombs and cemeteries It was the only find group that had a significantly high chi- (c.f. Frederiksen and Ward-Perkins 1957, Fig. 4). The square value. The distribution of prehistoric pottery was position of the southern tombs overlooking the road shows very similar to the overall background distribution. The that they were part of the communication network between distribution of Orientalising and Archaic pottery was also the Nepi area and the south. Road cuttings, cemeteries and slightly similar to the background distribution. When the river valleys met at the hubs of the local communication distance values were compared, the units with flint had the network. smallest mean and medium, whereas prehistoric pottery had the highest. This means that flints were more likely As suggested above the distances from the least-cost paths to be found near the access to the river valleys, whereas between Archaic towns and river crossings were two of prehistoric pottery was found farther afield on plateaux. the geographic attributes used in the GIS analyses. The However, the emphasis of the distribution of prehistoric descriptive statistics for the distances from the overland pottery was in the west, where flatter terrain and natural least-cost paths revealed that, surprisingly, the flint finds access ensured that only very few road cuttings were were found generally nearer the paths than any other needed. material group. Prehistoric pottery was found the farthest. The flint distribution proves the importance of access to Areas with Roman pottery showed more variability than river valleys. The relationship between the prehistoric those with earlier Orientalising and Archaic pottery. All pottery and least-cost routes and the road cuttings show distributions were skewed towards the lower values, but that the relationship between the Late Neolithic – Early the distribution of prehistoric pottery was the least skewed Middle Bronze Age settlement and land routes was and more even than those of other material groups. complicated. Prehistoric finds were made across wide areas and they did not connect directly to the later routes The chi-squared tests showed that the distribution of of communication. Since the find distributions covered Roman pottery was related to the least-cost paths at the predominantly the less extreme landscapes on the western level of confidence of 97%. There was no significant plains, access and land transport were not such a problem. relationship between the paths and the Orientalising and The relationship between the least cost routes and land Archaic pottery, but the level of confidence was relatively transport on one hand and the Orientalising and Archaic high at 83.1%. The distributions of flint and prehistoric and Roman material on the other is clear. Nevertheless, pottery were not related to the distance from the least-cost the results suggest that the pre-Roman communication paths. This suggests that the prehistoric land use was not network was more complex than the efficient Roman road connected with later land networks. However, all groups network. differed significantly from the background distribution for the area; the Orientalising and Archaic and Roman Not only did the statistical analysis suggest a more distributions differed the most and prehistoric pottery the complex, changing relationship between the fluvial least. This is partly due to the fact that the transect design landscape and land transport network during the late for the choice of research areas partly coincided with the prehistory and protohistory, but also the location of the routes of the Roman road lines. major centres themselves suggests changes in emphasis. When one compares the visible territory around different Road cuttings and river crossings represent long-lived centres, the major differences between the Final Bronze landscape features that are difficult to date, partly Age Pizzo and the Early Iron Age Nepi become apparent. because many of them have been modified during the The visible territory of Il Pizzo was small and narrow with Medieval times and later. The chi-squared test showed the gorge was the central feature (Fig. 9.A.), whereas the that the distribution of the surveyed units differed from visible territory was wider around Nepi (Fig. 9.B.) with the the background area; the reason for these units to locate surrounding plains becoming more dominant. Similarly, near the road cuttings was the choice of the areas farther at the Final Bronze Torre dell’Isola and Vignale, the away from Nepi during the second season of the Nepi river valley stood in the middle of the core visible area. Survey. The comparison between the units with and However, Monte Soratte, a calcareous mountain and a without prehistoric and Orientalising and Archaic pottery regional landmark, was not visible from Il Pizzo. The non- proved that there was a significant difference between the visibility of Monte Soratte is significant since this mountain distributions at the level of confidence of 99.98%. The was visible from almost all other major sites and areas in amount of flints did not increase or decrease in relation the study area. Furthermore, the dislocation of settlement with the cuttings, but the positive correlation was very from the Final Bronze Age Vignale to the site of Faliscan high in the case of prehistoric pottery. The correlation Falerii Veteres at Civita Castellana was analogous to the with the road cutting and river crossing existed with the dislocation to Nepi; Monte Soratte became the central

113 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Fig. 9. A. Visibility from the Final bronze Age Il Pizzo. B. Visibility from the Iron Age Nepi. visible feature in both places during protohistory. The Age cemetery of Osteria dell’Osa (Bietti Sestieri 1992a; changes in the physical locations and visible areas suggest 1992b) was part of the complex at Gabii. Pacciarelli that the moves resulted from the population pressures of (2000, 242) emphasized the obvious family ties between growing communities that needed more control over their the buried individuals at Osteria dell’Osa; this cemetery agricultural land and connectivity, but they also pointed was the burial ground for the inhabitants of the western out to the creation of the symbology of a possible common part of the later lower town. Although the tomb groups in identity. the cemetery represented different families, the deceased seemed to be distantly related inside the community. Settlement, lakes and rivers in the Gabii area Of the major sites in the hinderland of Gabii, the ancient The settlement history of Gabii is not fully known, since town of Collatia has been attributed to the site at Lunghezza the northern side of the Archaic town was destroyed during (Quilici 1974, 199) in the north (Fig. 11). The structure the Roman and later by quarrying activities (Guaitoli of the site is not very well understood since a large part 1981, 23, 46–47). However, the site was settled from the of the settlement site was destroyed when the houses on Neolithic period onwards. The site of Casale del Pescatore the hill east of the castle were built (Quilici n.d., n. 780). (Carboni 1993) was located at the western extreme of the Nevertheless, an antefix found from the existing castello later Archaic lower town (see Fig. 10). During the Middle suggests that there was an Archaic temple (c.f. Quilici Bronze Age, an open settlement was located on a terrace 1974, n. 100). The find has a parallel at Antemnae, a Latin looking over the lake in the crater of Castiglione (Bietti site at the mouth of the Aniene. On the other side of the Sestieri 1984; see also Guaitoli 1981, Fig. 3). The site was Aniene at Lunghezza there was another large ancient site partly destroyed by the Iron Age cemetery of Castiglione. (Quilici n.d., n. 778). This site was in direct contact with Very little remains of the later Bronze Age; some pottery ‘Collatia’, but it is likely that it was a twin site, not an sherds were found from the surface around the site of extension of the main site. Castiglione, but only one diagnostic sherd (Guaitoli 1981, Fig. 8, n. 1) has been found from the southern part of the A further Archaic town seems to have been located at lower town inside the perimeter of the Archaic town. One Corcolle (Fig. 11). By the beginning of the Archaic period could suggest that the Final Bronze Age settlement was it was an important site, as a series of chamber tombs with located at the Medieval Torre Castiglione at the highest bucchero and Etrusco-Corinthian wares from the mid-6th point of the crater on the northern part of the Archaic town century BC show (Reggiani et al. 1998). Two ditches, (c.f. Guaitoli 1981, 28; c.f. Fig. 11), as this geographically dated to the 6th century BC, separated acropolis, town defined defendable hillock is analogous with a typical and cemetery area from each other. Fragments of an altar Final Bronze Age site location (e.g. di Gennaro 1986). The with a rare 5th -century inscription have been found from few sherds by the lake may have been eroded from the the area of the main town. The importance of the site was higher ground. proved by a temple podium and architectural terracottas (Mari 1983, 46–51; c.f. Matronola Giansanti 1980). During the Iron Age, the southern part of the later Archaic town (c.f. Fig. 11) was settled as a series of concentrations The visibility analyses suggest that during most of the showed (Guaitoli 1981, Fig. 10). The excavations at a later later prehistory the inhabitants of the consecutive sites at temple area of Juno Gabina revealed some hut foundations the crater of Castiglione had limited views – literally and cut down to tuff (De Santis 1992, 12). The famous Iron cognitively. The settlements dated between the Neolithic

114 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

Fig. 10. Road cuttings into the river valleys and river crossings in the Nepi area. and the Early Bronze Age seem to have looked first and The early settlement in the area was wide-spread. However, foremost at the crater of Castiglione and to the areas south- the distribution was heavily skewed by the data from the west from the crater (see Rajala 2004a). The Neolithic article published by Carboni and Ragni (1984). Contrary farmers of Casale del Pescatore seem to have wanted to to the situation in the Nepi area, the Middle Palaeolithic control the wetlands in their vicinity (Fig. 12.A.). The finds were common. The distribution of Palaeolithic, visible territory seems to have been particularly restricted Neolithic and Chalcolithic/Early Bronze Age settlement during the Middle Bronze Age when the crater and the lake was concentrated along the main rivers in the west and inside were the centre of the domestic world (Fig. 12.B.). in the more dissected eastern area. The Neolithic and Early Bronze Age settlement was widespread and there The core visible area from the highest point of Torre is a visible correlation between settlement and water (c.f. Castiglione (Gabii), the potential Final Bronze Age site, Fig. 10). However, the core area was very empty with the included the hills around the crater of Castiglione and the exception of the Neolithic site of Casale del Pescatore. It is low lands of Pantano Borghese (Fig. 12.C.). The visible unclear, whether this is due to biases in research strategies, area is only slightly smaller than the hypothetical territory past marshy terrains or the real scarcity of finds in the defined by the Thiessen polygons for the Final Bronze Age areas, where there were fewer rivers fed by the perennial settlement. This suggests a positive correlation between springs. the territory and the visible area. The area controllable by the visual means during the Final Bronze Age was also the The difference in numbers of Chalcolithic/Early Bronze core of the Iron Age and later territory. During the Iron Age and Middle Bronze Age sites is considerable (see Fig. Age and Archaic period the inhabitants at Gabii seem to 13). The importance of water when choosing a settlement have turned their sight more and more towards Pantano site is apparent from the distribution pattern even without Borghese and the northern side of the Alban hills. The statistical analysis. At the site of Castiglione bones of views from the highest point of the lower Archaic town tortoise and water bird were found together with domestic looked predominantly south (Fig. 12.D.). and wild animal species and pottery associated with dairy

115 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Fig. 11. Casale del Pescatore and the Neolithic sites in the Gabii area.

Fig. 12. Archaic centres and rural sites in the Gabii area.

116 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

production (Bietti Sestieri and De Santis 1984; 2000, 74–75). These finds show that both aquatic resources and agriculture were important.

The settlement at Castiglione seems to have remained on the lake terrace during the Recent Bronze Age, although its intensity was reduced (Guaitoli 1981, 28). As during the earlier periods, all sites were located near major water sources. During the Final Bronze Age, when the settlement was more wide-spread than during the previous periods, this pattern changed. Some sites did still lie near the main rivers, but otherwise a series of prominent hilltops were settled (see Fig. 14). Similarly to Torre Castiglione there were some finds from near Corcolle at the foot of a hill that can be related to a typical naturally defended Final Bronze Age location.

The knowledge of Iron Age settlements is dominated by the data from Osteria dell’Osa, Castiglione and Gabii (c.f. Bietti Sestieri 1992b). The minor finds from Osteria Curato and Casale De Luca (Carboni and Ragni 1984), Ponte Nona and Pecoreccia (Belardelli et al. 1986) and Colle Tasso and Colle Bulgarini (Sciarretta 1969; Mari 1983) suggest that there were other settlements in the territory. The cemeteries at Osteria dell’Osa (Bietti Sestieri 1992a) and La Rustica (Carettoni et al. 1976) and the Iron Age and Orientalising finds from Tivoli (Giuliani 1966) imply the existence of wealthy communities, where family hierarchies defined social order. The settlement pattern did not change drastically, as in the Faliscan area during the 9th and 8th century BC. However, in both regions the major centres were located relatively near each other and this Latin and Faliscan pattern differed from the large Etruscan city states. Even the burial customs had similarities between these areas (Baglione 1986, 130, n. 20, 136–137).

The Orientalising period was a time of consolidation. The number of settlements increased slightly. This period saw the beginning of the larger sites at S. Giovanni in Camporazio (c.f. Muzzioli 1970, 129, n. 2) and Ponte northern area. D. From the highest point of the lower town in the south. town of the lower point the highest From D. area. northern Mammolo (Quilici 1974, n. 1) that proves the development of a hierarchical settlement pattern in the region. The area on the hypothetical boundary between Gabii and Tivoli was densely populated and analogously to the developments in the Nepi area, the two-level settlement pattern consisted of a few major centres and minor centres.

The relatively nucleated settlement pattern changed during the Archaic period. A dense rural settlement pattern developed between the urban centres of La Rustica, Lunghezza (Collatia) and Gabii. The settlement pattern was notably different from the one in the Nepi area. Since the rural settlement reached the urban area near Gabii, the settlement pattern looks more urbanised than at Nepi; the town–dwellers at Gabii do not seem to have been involved directly in agriculture but in trade and services. The pattern is similar to one in the surroundings of Crustumerium and

Fig. 13. Visibility from the sites at Gabii (after Rajala 2004, fig. 3). A. From the Neolithic Casale del Pescatore. B. From the Middle Bronze Age Castiglione. C. From the highest point of the point the highest C. From Castiglione. Age the Middle Bronze B. From Gabii (after Rajala del Pescatore. Casale the Neolithic 2004, fig. 3). A. From at the sites 13. Visibility from Fig. Fidenae (c.f. Quilici and Quilici Gigli 1980; 1986) and

117 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Fig. 14. A. The settlement in the Gabii area between the Chalcolithic and Early Bronze Age. B. The Middle Bronze Age settlement. that of the Republican settlement at Nepi (Mills and Rajala slopes on ridges were preferred. The distribution of slope 2011, Fig. 6). values show that 75% of sites were on flat land or on relatively gentle slopes. Steeper slope values were usually The 120 Archaic settlements surveyed by Quilici (1974) from the edges of the ridges. The way the settlement in the area between La Rustica, Lunghezza and Gabii (see followed ridges can be observed from aspect values, Fig. 11) were located predominantly in places defined by which did not differ at all from the background values; the their location on ridges; a quarter of sites was located on overall decrease of altitude towards western and northern ridge (31, 26.05%), while the majority were on hilltops directions is reflected in the values. (46, 38.66%), mainly on ridges. Another quarter was on slopes (30, 25.21%). Thus, the farmers chose places that The dispersed rural settlement avoided the areas near the were dominant and overlooked their surroundings from a Aniene. More sites than expected were located between higher altitude. four and six kilometres away from the river. The difference between the observed and expected distributions was The rural settlement was assessed on the basis 100% significant with a high chi-square value of 51.286. of the basic elevation derivates and a series of The analysis of distances from all tributaries of the Aniene distance attributes. on the other hand and the distances from the rivers fed by perennial springs (Fig. 14) on the other revealed interesting The elevations of the sites differed significantly from the background values with the level of confidence of 100%. phenomena. At the local level, the sites were near rivers, The distribution of elevation values was concentrated but not in the immediate vicinity along them. More often between 50 and 75 metres above sea level. The sites than expected they were 150 – 300 metres from the river; clustered around ridges away from the Aniene. The number the distribution differed significantly from the expected. of sites near lower lying river plains was unexpectedly At the regional level, the site distribution concentrated low. However, the survey was more intense in agricultural, in the vicinity of the rivers fed by the perennial springs; non-built-up areas in the 1970s. In any case, the results significantly less sites than expected located more than reflect the past agricultural land use of ridge areas. 900 metres from such river. These distributions mean that while the availability of water all year round near Slope was not as significant as a criterion, although the the agricultural land was important, the settlement sites chi-square values were almost significant; the levels of were chosen in relation to ridges. This suggests – as the confidence are in the low 90%. It is clear from the values evidence from the Nepi area – that the land transport was that steeper areas were avoided and flatter hilltops and important during the protohistoric period.

118 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

Fig. 15. The distance from the perennial rivers to the Archaic rural settlements in the Gabii area.

The rural sites were located relatively often near Gabii but Conclusions avoided the vicinity of La Rustica and Lunghezza by the Aniene. This observation was proved by the chi-squared I have argued that the fluvial landscape at Nepi played a tests as more sites than expected were located less than role in the cognitive change related with the shift from four kilometres from Gabii and significantly less more the Bronze Age site of Il Pizzo to the Iron Age Nepi. This than eight kilometres from the centre. The situation is change was not only chronological but depended on the almost opposite in the case of La Rustica; less sites than increasing populations and changing social landscape in expected were located in the vicinity of this smaller centre central Italy (Rajala 2004b). The dislocation of settlement and more than expected over six kilometres from the site. was paralleled with the dislocation of the visible territory. In the case of Lunghezza (Collatia), most sites were four The view from the highest point of the promontory of Nepi to six kilometres from the site. This is partly due to the was directed towards the edges of the plateaux whereas the uneven research history in the region; most sites surveyed previous settlement on Il Pizzo controlled the canyon-like by Quilici (1974) fell inside the hypothetical territory of river valley. The river remained under the gaze from the Gabii, whereas the areas outside the modern administrative fringes of Nepi, but seem to have remained marginal to the boundaries of Rome remain less studied. However, the fact core of the site. that there were some larger villages on the other side of the Aniene (c.f. Quilici 1974, n. 1) emphasizes the fact The results from the Nepi Survey Project point to the that the Aniene was a natural boundary, lined by the lesser changing importance of the riverine landscape. The gateway sites. The ‘empty’ areas around smaller centres prehistoric flints were found not near the rivers, but from around Gabii served as the agricultural reserve for the the plains. Although prehistoric pottery was widespread second-rank sites by the main river and formed a buffer and most often found at some distance from a water source, towards the agricultural core of the Gabii territory. The the find locations did not correlate with the proximity to pattern reflects the existence of a three-level settlement rivers. However, they were significantly connected with hierarchy in the area. the pathways to the river valleys.

119 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

The protohistoric finds from the Orientalising and Archaic Acknowledgements periods reflect the increased importance of land routes and farming. The clearest indications of Archaic rural Dr Rajala ran the Nepi Survey Project in the field as part settlements were from near the boundaries of the territory; of her Ph.D. research with financial support from the it seems that the town-dwellers were directly involved in British Academy, the Academy of Finland, New Hall farming the land in the immediate vicinity of Nepi. The (Cambridge), the Cambridge European Trust, the Finnish protohistoric pottery was clearly found farther away from Cultural Foundation, Worts Travelling Scholars’ Fund, the rivers than the prehistoric pottery, but it correlated both the Garrod Fund, the Prehistoric Society, the Society of with the road cuttings leading into the river valleys and Antiquaries of London, George Charles Winter Warr the known road lines. The ever-present Roman pottery Scholarships and Ridgway-Venn Travel Studentships. correlated positively with the shorter distances from both This survey was carried out with a concession from the rivers and the road network reflecting the location of and in close collaboration with the Soprintendenza per the finds near the edges of the plateaux along the efficient l’Etruria Meridionale, under the supervision of Dott.ssa Roman land communication network. Daniela Rizzo. It was supported and assisted by the British School at Rome and its staff, together with the mayor and The results from the Gabii area emphasized the urban personnel of the Comune di Nepi, the Museo Civico di character of the centre and point to the differentiated roles Nepi and its director Stefano Francocci, as well as the between the town-dwellers providing services and rural Museo Archeologico dell’Agro Falisco (Civita Castellana) population providing food. They also showed the visible and its staff. The author wants to thank the survey teams of relationship between the later prehistoric settlement and 1999 and 2000. The author is grateful to Simon Stoddart, the perennial rivers. The visibility analyses from the the director of the original Nepi Survey Project and her different sites at Gabii itself show that during the Neolithic supervisor, for his support and to Dr Andrea Vianello for period the crater of Castiglione and the slopes in the then his invitation to contribute to this volume. marshy lake area in the south-west, i.e. the wetlands, were Bibliography under direct gaze. During the Middle Bronze Age the settlement was placed by the crater lake inside the crater Alessio, M., Allegri, L., Bella, F., Calderoni, G., Cortesi, C., and the lake became the centre of the everyday landscape. Dai Pra, G., De Rita, D., Esu, D., Follieri, M., Improta, The views from the later Iron Age and Archaic settlement S., Magri, D., Narcisi, B., Petrone, V. and Sodori, L. in contrast guarded a large agricultural zone further afield. 1986. 14C dating, geochemical features, faunistic and pollen analyses of the uppermost 10 m core from Valle The distribution of the Archaic settlement emphasized the di Castiglione (Rome, Italy). Geologica romana 25, character of the river Aniene as the natural boundary, lined 287–308. by the second-rank gateway communities. The dense rural Baglione, M. P. 1986. Il Tevere e i falisci. In S. Quilici settlement nearer Gabii followed the ridges emphasizing Gigli (ed), Il Tevere e le altre vie d’acqua del Lazio the importance of agricultural land. At the local level the antic,. Quaderni del centro di studio per l’archeologia sites were not in the immediate vicinity along the rivers etrusco-italica 12. Archeologia laziale 7:2, 124–142. but at a distance, whereas at the regional level, the site Roma: C.N.R. distribution concentrated along the rivers fed by the Baglione, M., and De Lucia Brolli, M. 1990. Nuovi dati perennial springs. sulla necropoli de “I Tufi” di Narce. In G. Maetzke (ed), La civiltà dei falisci: Atti del XV convegno di These results underline the particular importance of these studi etruschi ed italici 28–31 maggio 1987, 61–102. perennial rivers in their respective territories. Naturally, Firenze: Leo S. Olschi. their importance was more marked during the prehistoric Baillie, M. G. L. 1998. Evidence for Climatic Deterioration times, but the analysis of the material distributions in the 12th and 17th Centuries BC. In B. Hänsel (ed). suggests that the significance of the access to the aquatic Mensch und Umwelt in der Bronzezeit Europas, 49–55. resources all year round was needed both during the pre- Kiel: Oetker-Voges Verlag. and protohistoric periods. Bartoloni, G. and Delpino, F. (eds) 2005. Oriente e Occidente: metodi e discipline a confronto. Riflessioni I have also made an attempt to argue the importance of the sulla cronologia dell’età del ferro in Italia, Atti major perennial river valleys in the local communication dell’Incontro di studi, Roma, 30–31 ottobre 2003. networks in the Faliscan area. During the periods, when Mediterranea 1 (2004). Pisa: Istituti Editoriali e the landscape was more wooded with Mediterranean Poligrafici Internazionali. vegetation, the microclimate and the tree canopy in the Belardelli, C., Bietti Sestieri, A. M. and Trufelli, F. (eds) deep ravines could have resulted with low vegetation 1986. Preistoria e protostoria nel territorio di Roma. along the river terraces that allowed an easier route for Modelli di insediamento e vie di comunicazione. In transport than the potentially impenetrable dry woodlands S. Quilici Gigli (ed), Il Tevere e le altre vie d’acqua over land. Understanding the true importance of these del Lazio antico, Quaderni del centro di studio per perennial rivers would require more geoarchaeological l’archeologia etrusco-italica 12, Archeologia laziale work in both areas. VII.2, 30–69.

120 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

Belardelli, C., Angle, M., di Gennaro, F. and Trucco, al. (eds). Civiltà del Lazio primitivo. Catalogo della F. 2007. Repertorio dei siti protostorici del Lazio: mostra. Palazzo delle esposizioni, 153–186. Roma: province di Roma, Viterbo e Frosinone. Borgo San Istituto di Studi Etruschi e Italici e del Comitato per Lorenzo (Firenze): All’Insegna del Giglio. l’Archeologia Laziale and Multigrafica. Bietti Sestieri, A. M. 1984. Castiglione (Roma). Abitato Cherkauer, D. 1976. Site K. In T. W. Potter. A Faliscan dell’età del bronzo e necropoli dell’età del ferro. In Town in South Etruria: Excavations in Narce 1966–71, A. M. Bietti Sestieri (ed), Preistoria e protostoria 106–120. London: The British School at Rome. nel territorio di Roma. Lavori e studi di archeologia Cifani, G. and Munzi, M. 1995. Considerazioni sugli pubblicati dalla Soprintendenza archeologica di Roma insediamenti in area falisca. I periodi arcaico e 3, 160–168. tardoantico. In N. Christie (ed), Settlement and Bietti Sestieri, A. M. (ed) 1992a. La necropoli laziale di Economy in Italy 1500 BC to AD 1500. Papers of Osteria dell’Osa. Roma: Quasar. the Fifth Conference of Italian Archaeology, Oxbow Bietti Sestieri, A. M., 1992b. The Iron Age community Monograph 41, 387–94. Oxford: Oxbow. of Osteria dell’Osa: A study of socio-political Cifani, G., Opitz, R. and Stoddart, S. K. F. 2007. Mapping development in central Tyrrhenian Italy. Cambridge: the Ager Faliscus road–system: The contribution of Cambridge University Press. LiDAR survey. Journal of Roman Archaeology 20, Bietti Sestieri, A. M. 1996. Protostoria: Teoria e pratica. 165–176. Roma: La Nuova Italia Scientifica (NIS). Cornell, T. J. 1995. The Beginnings of Rome. Italy and Bietti Sestieri, A. M. and De Santis, A. 2000. Protostoria Rome from the Bronze Age to the Punic Wars, c. 1000- dei popoli latini. Museo nazionale romane terme di 263 BC. London: Routledge. Diocleziano. Milano: Electa. Cowgill, U. M. and Hutchinson, G. E. 1970. Chemistry Bintliff, J. L., Howard, P. and Snodgrass, A. M. 1999. The and mineralogy of the sediments and their source hidden landscape of prehistoric Greece. Journal of materials. In G. E. Hutchinson (ed), Ianula – an Mediterranean Archaeology 12, 139–168. account of the history and development of the Lago Boni, C., Bono, P. and Capelli, G. 1988. Carta Monterosi, Latium, Italy, Transactions of the American idrogeologica del territorio della regione Lazio. Roma: Philosophical Society 60 (4), 37–101. Regione Lazio and Università degli studi di Roma “La David, J.-M. 1996. The Roman Conquest of Italy. Translated Sapienza”. from the French original La Romanizzazione de l’Italie Brown, A. G. and Ellis, C. 1995. People, climate (1994). Oxford, Cambridge: Blackwell. and alluviation: theory, research design and new De Santis, A. 1992. Storia delle ricerche nel comprensorio sedimentological and stratigraphic data from Etruria. Osteria dell’Osa-Gabii-Castiglione. In A. M. Bietti Papers of the British School at Rome 64, 45–74. Sestieri (ed), La necropoli laziale di Osteria dell’Osa, Brückner, H. 1986. Man’s Impact on the Evolution of the 11–18. Roma: Quasar. Physical Environment in the De Santis, A. 2001. Castiglione. Abitato dell’età del Mediterranean Region in Historical Times. GeoJournal Bronzo e necropoli dell’età del Ferro. In F. Filippi (ed), 13(1), 7–17. Archeologia e Giubileo: Gli interventi a Roma e nel Brunetti Nardi, G. 1981. Repertorio degli scavi e delle Lazio nel piano per il Grande Giubileo del 2000, 485– scoperte archeologiche nell’Etruria meridionale 489. Napoli: Electa. (1971–1975). Roma: Comitato per le attività di Gennaro, F. 1986. Forme di insediamento tra Tevere e archeologiche nella Tuscia and C.N.R. Fiora dal Bronzo Finale al principio dell’Eta del Ferro. Camilli, A., Carta, L., Conti, T. and De Laurenzi, A. 1995. Firenze: Olschki. Ricognizioni nell’Ager Faliscus meridionale. In N. di Gennaro, F. 1995a.La Ferriera di Sutri. In N. Negroni Christie (ed), Settlement and Economy in Italy 1500 Catacchio (ed), Tipologia delle necropoli e rituali di BC to AD 1500. Papers of the Fifth Conference of deposizione: Ricerche e scavi, Preistoria e protostoria Italian Archaeology, Oxbow Monograph 41, 395–402. in Etruria: Secondo incontro di studi, Farnese, 21– Oxford: Oxbow. 23 maggio 1993, 277–279. Milano: Centro studi di Carandini, A. (ed) 1985. La romanizzazione dell’Etruria: preistoria e archeologia. il territorio di Vulci. Milano: Electa and Regione di Gennaro, F. 1995b. Le età del bronzo e del ferro nel Toscana. territorio falisco. An unpublished manuscript for tesi di Carboni, G. 1993. Giacimento neolitico con ceramiche Diploma, Scuola di specializzazzione in Archeologia, della facies del Sasso a Casale del Pescatore Cattedra di Etruscologia e Antichità italiche, Università (Montecompatri – Roma). Documenta Albana 14–15 di Roma ”La Sapienza”. (1992–93), 13–33. di Gennaro, F., Cerasuolo, O., Colonna, C., Rajala, U., Carboni, G., and Ragni, E. 1984. Ricognizioni di Stoddart, S. and Whitehead, N. 2002. The recent superficie F. 25. In A. M. Bietti Sestieri (ed), research on the city and territory of Nepi (VT). Papers Preistoria e protostoria nel territorio di Roma, Lavori of the British School at Rome 70, 29–77. e studi di archeologia pubblicati dalla Soprintendenza di Gennaro, F., Rajala, U., Rizzo, D., Stoddart, S. and archeologica di Roma 3, 47–63. Whitehead, N. 2008. Nepi and territory: 1200 BC Carettoni, G., Zeccagni, P., Bietti Sestieri, A. M. and – 400 AD. In F. Coarelli and H. Patterson (eds). Anzidei, A. P. 1976. La Rustica. In M. O. Acanfora et Mercator Placidissimus – The Tiber Valley in antiquity.

121 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

New research in the upper and middle valley, 879–888. at Fosso della Crescenza 1962. Papers of the British Rome: Quasar. School at Rome 51, 70–72. Duncan, G. C. 1958. Sutri (Sutrium). Papers of the British Kahane, A. 1973. The paved Roman road east from Gabii. School at Rome 26, 63–134. Papers of the British School at Rome 41, 18–44. Edwards, C., Malone, C. and Stoddart, S. 1995. Kahane, A. and Ward-Perkins, J. 1972. The Via Gabina. Reconstructing a gateway city: the place of Nepi in Papers of the British School at Rome 40, 91–126. the study of south–eastern Etruria. In N. Christie (ed), Lim, S.E., Stoddart, S., Harrison, A. and Chalmers, A. Settlement and Economy in Italy 1500 bc to ad 1500, 1996. Recent examples of geographical analysis of Papers of the Fifth Conference of Italian Archaeology, archaeological evidence from central Italy. In H. Oxbow Monograph 41, 431–440. Oxford: Oxbow. Kamermans and K. Fennema (eds), Interfacing the De Lucia Brolli, M. A. 1990. Un nuovo santuario a Narce Past, Computer Applications and Quantitative Methods sulla sponda del Treia. In G. Maetzke (ed), La civiltà in Archaeology CAA95 II, Analecta Praehistorica dei Falisci: Atti del XV convegno di studi Etruschi e Leidensia 28 II, 331–336. Leiden: University of Italici, 173–195. Firenze: Leo S. Olschki. Leiden. di Gennaro, F. and Stoddart, S. 1982. A review of the Lorenzoni, P., Lulli, L. and Raglione, M. 1984. Studio evidence for prehistoric activity in part of South pedologico dell’apparato vulcanico di Vico. I. Indagine Etruria. Papers of the British School at Rome 50, 1–21. preliminare sui principali fattori ambientali. Annali Francocci, S. 2006. L’antica Nepi in età romana. In S. dell’Istituto sperimentale dello studio e della difesa del Francocci (ed), Archeologia e storia a Nepi I, Quaderni suolo 15, 81–110. del Museo Civico di Nepi 1, 45–65. Nepi: Museo Lulli, L., Dowgiallo, G. and Brunelli, L. 1988. I suoli dei Civico di Nepi. rilievi del Monte Soratte e del Monte Piccolo (Lazio) e Frederiksen, M.W. and Ward-Perkins, J.B. 1957. The la loro influenza sulla vegetazione. Annali dell’Istituto ancient road systems of the central and northern Ager sperimentale dello studio e della difesa del suolo 19, Faliscus (Notes on southern Etruria, 2). Papers of the 85–107. British School at Rome 25, 67–203. Majerini, V. and Musco, S. 2001. Gabii. Indagini Giuliani, C. F. 1966. Tibur, Pars altera, Forma Italiae archeologiche: area urbana, area della chiesa di San Primitivo e area del Santuario extraurbano orientale. In I:XVII.3. Firenze: Leo S. Olschki. F. Filippi (ed), Archeologia e Giubileo: Gli interventi a Guaitoli, M. 1981. Gabii: osservazioni sulle fasi di Roma e nel Lazio nel piano per il Grande Giubileo del sviluppo dell’abitato. In Ricognizione archeologica: 2000, 490–499. Napoli: Electa. nuove ricerche nel Lazio, Quaderni dell’istituto di Mari, Z. 1983. Tibur, Pars tertia, Forma Italiae I:XVII.3. topografia antica della università di Roma 9, 23–58. Firenze: Leo S. Olschki. Firenze: L.S. Olschki. Mari, Z. 1991. Tibur, Pars quarta, Forma Italiae I:XVII.4. Harris, W. V. 1971. Rome in Etruria and Umbria. Oxford: Firenze: Leo S. Olschki. Clarendon Press. Matronola Giansanti, M. G. 1980. Iscrizione arcaica da Harrison, A., Rajala, U., Stoddart, S., Witcher, R. and Corcolle. Atti e memorie della Societa Tiburtina di Zubrow, E. 2004. The enhancement of the South storia e d’arte 53, 29–34. Etruria Survey: phase 1. In H. Patterson (ed), Bridging Mills, P. and Rajala, U. 2011. The Roman ceramic material the Tiber. Approaches to Regional Archaeology in the from the fieldwalking in the environs of Nepi. Papers Middle Tiber Valley, Archaeological Monographs of of the British School at Rome 79, 147–240. the British School at Rome 13, 29–36. London: British Ministero dei Lavori Pubblici 1959. Precipitazioni School at Rome. medie mensili ed annue e numero dei giorni piovosi Hutchinson, G. E. and Cowgill, U. M. 1970. The History per il trentenio 1921–1950. Servizio Idrografico. of the lake: a synthesis. In G. E. Hutchinson (ed), Pubblicazione n. 24 del servizio, fasc. XIV, settembre Ianula – an account of the history and development of 1959. Roma: Ministero dei Lavori Pubblici. the Lago Monterosi, Latium, Italy. Transactions of the Morselli, C. 1980. Sutrium, Forma Italiae VII:7. Firenze: American Philosophical Society 60 (4), 163–170. Leo S. Olschki. Iaia, C. and Mandolesi, A. 1993. Topografia Muzzioli, M. P. 1970. Praeneste, Pars altera: Lavoro dell’insediamento dell’ VIII secolo a.C. in Etruria eseguito e pubblicato con il contributo del Consiglio meridionale. Journal of Ancient Topography 3, 17–48. nazionale della ricerche, Istituto de topografia antica Ingold, T. 1993. The Temporality of the Landscape. World dell’Università di Roma, Forma Italiae I:7.2. Roma: Archaeology 25, 152–174. De Luca. Jones, G. D. B. 1960. Veii: the Valchetta baths. Papers of Nijboer, A.J., van der Plicht, J., Bietti Sestieri, A. M. and the British School at Rome 26, 63–134. de Santis, A. 2002. A high chronology for the Early Judson, S. 1963. Erosion and Deposition of Italian Stream Iron Age in central Italy, adapted from Palaeohistoria Valleys During Historic Times. Science 140, 898–899. 41/42, 1999, at http://www.lcm.rug.nl/lcm/teksten/ Judson, S. 1968. Erosion Rates near Rome, Italy. Science teksten_uk/a_high_chronology_uk.htm (accessed in June 160, 1444–1446. 2012). Judson, S. 1983. Alluviation and erosion. In E. Fentress, S. Pacciarelli, M. 1996. Nota sulla cronologia assoluta della Judson, T. Blagg, M. de Vos and P. Arthur, Excavations prima età del ferro in Italia. Ocnus 4, 185–191.

122 Copyright material: no unauthorized reproduction in any medium Rajala: The Perennial Rivers and the Changing Settlement Patterns

Pacciarelli, M. 2000. Dal villaggio alla città: la svolta Rajala, U. 2006. Le ricerche della Scuola Britannica a protourbana del 1000 a.C. nell’Italia tirrenica, Grandi Nepi: indagini e prospettive. In S. Francocci (ed). contesti e problemi della protostoria italiana 4. Archeologia e Storia a Nepi, vol. I. Quaderni del Firenze: All’insegna del Giglio. Museo Civico di Nepi 1, 86–94. Patterson, H. 2004. Introduction. In H. Patterson Rajala, U. 2007. The bronze and iron age finds from Il (ed), Bridging the Tiber, Approaches to Regional Pizzo (Nepi, VT): the results of the intensive survey Archaeology in the Middle Tiber Valley, Archaeological 2000. Papers of the British School at Rome 75, 1–37. Monographs of the British School at Rome 13, 11–28. Rajala, U. 2012. Political landscapes and local identities in London: British School at Rome. Archaic central Italy – Interpreting the material from Patterson, H., di Gennaro, F., Di Giuseppe, H., Fontana, Nepi (VT, Lazio) and Cisterna Grande (Crustumerium, S., Gaffney, V., Harrison, A., Keay, S. J., Millett, M., RM, Lazio). In S. Stoddart and G. Cifani (eds), Rendeli, M., Roberts, P., Stoddart, S. and Witcher, R. Landscape, ethnicity and identity in the archaic 2000. The Tiber Valley Project: the Tiber and Rome Mediterranean area, 120–143. Oxford: Oxbow. through two millennia. Antiquity 74, 395–403. Rajala, U., in preparation. The concentration and Patterson, H. and Millett, M. 1998. The Tiber Valley centralisation of the late prehistoric settlement in Project. Papers of the British School at Rome 66, 1–20. central Italy: the evidence from the Nepi Survey. Pennacchioni, M. 1975. Preliminari di studio Papers of the British School at Rome 81. sull’insediamento preistorico delle ”Solfarate” (Sutri). Rajala, U., Harrison, A. and Stoddart, S. 1999. The Arché 1, 3–15. enhancement of the South Etruria Survey. GIS in the Potter, T.W. n.d. An Archaeological Field Survey of the study of the research history of the Southern Faliscan Central and Southern Ager Faliscus. British School area. In L. Dingwall, S. Exon, V. Gaffney, S. Laflin at Rome Archaeological Archive, unpublished and M. van Leusen (eds), Computer Applications and manuscript. Quantitative Methods in Archaeology, Proceedings Potter, T.W. 1979. The Changing Landscape of South of the 25th Anniversary Conference, University of Etruria. London: Elek. Birmingham, April 1997, BAR International Series 750, Potter, T. W. 1992. Reflections of twenty-five years’ 82 (the full text with the figures on the accompanying fieldwork in the Ager Faliscus. Approaches to CD). Oxford: Archaeopress. Landscape Archaeology. In M. Bernardi (ed), Reggiani, A., Adembri, B., Zevi, F., Benedettini, M. G. Archeologia del paesaggio II, 637–666. Firenze: and Mari, Z. 1998. Corcolle. In L. Drago Troccoli All’Insegna del Giglio. (ed), Scavi e ricerche archeologiche dell’Università di Quilici, L. n.d. Inventario e localizzazione dei beni Roma “La Sapienza”, Studia archeologica 96, 120– culturali archeologici nel territorio del comune di 124. Roma: “L’Erma” di Bretschneider. Roma. Roma: Comune di Roma. Rellini, U.,1920. Cavernette e ripari preistorici nell’Agro Quilici, L. 1974. Collatia. Forma Italiae I:10. Roma: De Falisco. Monumenti antichi XXVI, col. 5–180. Luca. Rizzo, D. 1992. Le necropoli di Nepi: Immagini di 10 anni Quilici, L. and Quilici Gigli, S. 1980. Crustumerium. di ricerche archeologiche. Nepi: Comune di Nepi. Roma: C.N.R. Sciarretta, F. 1969. Contributi alla conoscenza della Quilici, L. and Quilici Gigli, S. 1986. Fidenae. Roma: preistoria e protostoria di Tivoli e del suo territorio. Atti C.N.R. e memorie della Societa Tiburtina di storia e d’arte 42, Rajala, U. 2002. Human Landscapes in Tyrrhenian Italy. 7–113. GIS in the Study of Urbanisation, Settlement Patterns Scullard, H. H. 1967. The Etruscan Cities and Rome. and Land Use in South Etruria and Western Latium London: Thames & Hudson. Vetus. An unpublished PhD thesis, University of Segre, A. G. 1992. Comprensorio di Osteria dell’Osa – Cambridge, Department of Archaeology. Castiglione: Geologia tardo-olocenica. In A. M. Bietti Rajala, U. 2004a. Sense and Sensibility – Reflections Sestieri (ed), La necropoli laziale di Osteria dell’Osa, on the Epistemology and Ontology of GIS. Internet 19–21. Roma: Quasar. Archaeology 16. http://intarch.ac.uk/journal/issue16/ Selmi, R. 1978. Presenze preistoriche nel bacino rajala_index.html (visited in June 2012). idrografico del Treia. In Atti del secondo convegno dei Rajala, U. 2004b. The landscapes of power: visibility, Gruppi Archeologici di Lazio, Tolfa Gennaio 1976, time and (dis)continuity in central Italy. Archeologia e 55–59. Roma: Gruppo Archeologico Romano. calcolatori 15, 393–408. Shotter, D. C. A. 1976. Rome, the Faliscans and the Rajala, U. 2005. From a settlement to an early state? The Roman historians. In T.W. Potter, A Faliscan Town in role of Nepi in the local and regional settlement patterns South Etruria, Excavations in Narce 1966–71, 29–35. of the Faliscan area and inner Etruria during the Iron London: British School at Rome. Age. In P. Attema, A.Nijboer and A. Zifferero (eds), Smith, C., 1996. Early Rome and Latium. Economy and Papers in Italian Archaeology VI: Communities and Society c. 1000–500 BC. Oxford: Clarendon Press. Settlements from the Neolithic to the Early Medieval Sorriso-Valvo, M. 1998. Italy. In A. J. Conarcher and Period, Proceedings of the 6th conference of Italian M. Sala (eds), Land Degredation in Mediterranean Archaeology, Groningen, the Netherlands, April 15 - Environments of the World: Nature and Extent, Causes 17 2003, BAR International Series 1452 (II), 706–712. and Solutions, 41–65. Chicester: Wiley.

123 Copyright material: no unauthorized reproduction in any medium Rivers in Prehistory

Statistiche ambientali 1998, Annuario 5. Roma: ISTAT. Vita-Finzi, C. 1969. The Mediterranean valleys: Stefani, E. 1918. Antico sepolcro della necropoli nepesina. Geological changes in the historical times. Cambridge: Notizie degli Scavi 1918, 16–19. Cambridge University Press. Stefani, E.,1947. Tempio di Giunone Curite. Nuove Ward-Perkins, J. B. 1970. Monterosi in the Etruscan and richerche ed ulteriore osservazioni. Notizie degli Scavi Roman periods. In G. E. Hutchinson (ed), Ianula – an 1947, 69–74. account of the history and development of the Lago Stefani, E. 1948. Avanzi di antiche costruzioni scoperte Monterosi, Latium, Italy. Transactions of the American in vocabolo “Sassi Caduti”. Notizie degli Scavi 1948, Philosophical Society 60 (4), 10–12. 102–109. Tuppi, J. 2010. Traffic bottlenecks in south Etruria? Comparing the Archaic road cutting widths with ancient vehicles. Arctos 44, 265–290.

124