Water Management and Land Use in the Terramare

ARTICLE IN PRESS

Quaternary International 151 (2006) 87–98

Water management and land use in the terramare and a possible climatic co-factor in their abandonment: The case study of the terramara of Poviglio Santa Rosa (northern Italy)

Mauro Cremaschia,Ã, Chiara Pizzia,1, Veruska Valsecchib,2

aCNR-IDPA, Dipartimento di Scienze della Terra, v. Mangiagalli 34, 20132 Milano, Italy bInstitute of Plant Sciences, University of Bern, Altenbergrain, 21, 3013 Bern, Switzerland

Abstract

The terramare, archaeological sites generated by a civilisation which settled the centre of the Po plain since the Middle Bronze age, were abruptly abandoned at the end of the Recent Bronze age. The reason for this abandonment is still a matter of debate. In the context of a multi-causal explanation, this paper discusses a possible climatic factor, inferred from recent discoveries at the site of Poviglio Santa Rosa (northern Italy). The culture of the terramare produced a systematic and intensive agricultural and pastoral exploitation of the environment and caused heavy deforestation. Water was a critical resource and was carefully managed. The moat which typically surrounded most of the sites was meant to concentrate and redistribute water to the ﬁelds through a network of irrigation ditches. The study of the water wells discovered at the fringe of the terramara of Santa Rosa reveals a drop in the water table during the late Recent Bronze age, caused by a dry event of regional relevance. Chronological contiguity, supported by the archaeological context between this occurrence and the abandonment of the terramare system, strongly suggests a causative link between the two events. The climatic event, a minor episode in the history of the Holocene climate, nevertheless acted as a catalyst in a stressed environment whose resources were over-exploited by rising demographic pressure. r 2006 Elsevier Ltd and INQUA. All rights reserved.

1. Foreward researches carried out in the 19th century (Fig. 1). The terramare are the archaeological remains of a complex In Italian Prehistory, the term ‘‘terramare’’ is used to society whose subsistence was based on agriculture, indicate the banked and ditched villages, generally quad- pastoralism, handicraft and long-distance trade, connect- rangular in plan, of the Middle and Recent Bronze ages (c. ing the Po plain to northern and eastern Europe and to the 1650–1150 cal BC). Their name is derived from an old Mediterranean. The terramare settled the alluvial plain of Italian term used to indicate the organic rich earth—terra the Po as of the 16th century BC. They underwent marna—of mounds quarried as fertiliser in the 19th century significant demographic growth during the second half of (Bernabo` Brea and Mutti, 1994). The terramare are located the 15th century BC—at their apogee—and were aban- mostly in the modern Emilia sub-region, but also in doned abruptly at about 1150 BC as a consequence of a southern Lombardy and in south-western Veneto. About societal collapse (De Marinis, 1975; Barfield, 1994; Balista 200 sites have been found since the beginning of the and De Guio, 1997; Bernabo` Brea and Cardarelli, 1997; Bernabo` Brea et al., 1997; Cardarelli,1997; Pearce, 1998). It is clear now that the collapse of the terramare system ÃCorresponding author. E-mail address: [email protected] (M. Cremaschi). is not linked to the climatic deterioration at the sub-Boreal/ 1C. Pizzi is responsible of the archaeological data. sub-Atlantic transition (Cremaschi, 1991–1992), as was 2V. Valsecchi is responsible of the palinological data. reputed by Sa¨flund (1939), since this event is at least three

Fig. 1. Distribution of the Middle and Recent Bronze age sites in northern Italy. (1) Archaeological sites, (2) banked and ditched sites, (3) terramare covering pile-dwellings sites, (4) Holocene palaeochannels, (5) Pleistocene moraines, (6) Pre-Holocene terraces, (7) Upper Pleistocene outwash fans and (8) Holocene alluvial plain. Terramare quoted in the text: 1—Motta Balestri; 2—Santa Rosa; 3—Monticelli di Poviglio; 4—Case Cocconi; 5—Case del Lago; 6—Gaggio; 7—Fondo Paviani, 8—Castello del Tartaro; 9—Fabbrica dei Soci.

centuries more recent. The abandonment of the terramare palaeochannel which was active during the life of the site appears to be broadly parallel to the collapse of the late (Ferri, 1996; Cremaschi, 1997; Balista, 2002). Bronze age civilisations in the eastern Mediterranean (De For instance, in the Enza basin, in a central position with Marinis, 1975). More recently, a multi-causal explanation respect to the terramare region, five large sites located at in which environmental stress played a leading role has regular intervals are lined along the palaeochannel of the emerged (Cremaschi, 1991–1992; Barfield, 1994; Bernabo` Enza river, which during the Bronze age connected the Brea et al., 1997; Pearce, 1998). Yet again, a recent Apennine fringe to the Po river (from South to North: Case discovery in the terramara of Poviglio Santa Rosa (Reggio del Lago Case Cocconi, Monticelli di Poviglio, Santa Rosa Emilia—Italy) now suggests that a climatic event causing and Motta Balestri; Cremaschi, 1997, 2004)(Fig. 2). The dry conditions may have contributed to the final abandon- direction of their ramparts, large enough to be perceived on ment of the sites. The environmental context in which this a small-scale map, follows the former hydrographic trend, event occurred, strongly affected by human impact, will be of North-East orientation, as opposed to the present discussed, with emphasis on the water resources, critical to drainage oriented to the North. the subsistence strategy of the terramare. A close relation between individual sites and the hydrography has been stressed since the beginning of the 2. Terramare and hydrography archaeological investigation in the mid-19th century. The traditional ideas are clearly summarised by Chierici (1871). On the regional scale, the connection between the The inner part of the terramare and the outer ditches were terramare and fluvial network is systematic: each site for fed by an artificial channel deriving water from an adjacent which we have geomorphologic evidence is linked to a river. ARTICLE IN PRESS M. Cremaschi et al. / Quaternary International 151 (2006) 87–98 89

Fig. 3. Fabbrica dei Soci site: geomorphologic interpretation; (1) alluvial ridge, (2) Tartaro palaeochannel and artiﬁcial channel, (3) moat and rampart of the terramara, (4) roman centuriation and (5) present irrigation ditches, location in Fig. 2 (after Cremaschi, 1997).

Management of water was not limited to the capture and adaptation of natural streams, but also encompassed artificial ditches and channels for irrigation. In the case of the Castello del Tartaro (Balista, 1997; Cremaschi, Fig. 2. Geomorphology of the area of Enza Basin. (1) S. Ilario alluvial fan, (2) palaeochannel, (3) Praticello palaeochannel, (4) alluvial ridge, (5) 1997)—Grandi Valli Veronesi, North of the Po—the phreatic pool and related drainage, (6) boundary of swamp areas, (7) irrigation channels were conducted outward from the moat Bronze age sites, (8) banked and ditched terramare, (9) roman age of the site to distribute the water to cultivated fields. centuriation and (10) contour lines. Terramare quoted in the text: 1— This case is not isolated, similar evidence having been Motta Balestri; 2—Santa Rosa; 3—Monticelli di Poviglio; 4—Case recently discovered in several sites, as in the terramara di Cocconi; 5—Case del Lago (after Cremaschi, 1997). Gaggio (Modena) (Balista et al., 2003). At Case Cocconi (Bronzoni and Cremaschi, 1989), the terramara, 15 ha wide, is included in a concentric belt about 60 ha wide Recent geomorphologic research based on an examina- (Fig. 4). This area, probably devoted to specialised tion of aerial photographs has confirmed the model which agriculture, was delimited by a deep ditch, excavated can be applied to most of the sites (Mutti, 1993; Ferri, during the Bronze age and still perfectly perceivable in the 1996; Cremaschi, 1997). An impressive example is con- present landscape. A peripheral ditch was recently ob- stituted by the Fabbrica dei Soci site (Balista and De Guio, served (Bernabo` Brea and Mutti, unpublished) also in the 1990–1991, Cremaschi, 1997) which is located North of the Case del Lago terramara. Po in the plain South of Verona. The site (Fig. 3), Water resources appear to have been critical for the rectangular in plan and surrounded by an earth rampart, subsistence of the terramare. In the country outside the was built close to a palaeochannel of the Tartaro river, sites, the existence of a network of irrigation ditches from which an artificial drainage was abducted to feed the originating from the moats demonstrates that these moat surrounding the terramara. structures, beyond defensive purposes, were mostly Where river courses were not easily accessible, minor planned to collect and channel the water, and redistribute drainage and phreatic pools were also used. The terramara it to irrigate the fields. of Case del Lago, the largest (22 ha) of the terramare system, was built in connection with a large phreatic pool, 3. Land use which was not exhausted during the Bronze age. Today it has been harnessed by the waterworks to supply the city of The average density of the terramara settlements is one Reggio Emilia (Cremaschi, 1997). site per 25 km2, and in some areas, it rises to one site per ARTICLE IN PRESS 90 M. Cremaschi et al. / Quaternary International 151 (2006) 87–98

Fig. 4. Case Cocconi terramara. (1) Natural drainage, (2) roman centuriation, (3) contour lines, (4) area of the site delimited by the Bronze age ditch and (5) possible terramara size, location in Fig. 1 (after Cremaschi, 1997).

10 km2. This estimation is to be regarded as minimum as A tendency toward a more open environment appeared, large areas of the Po plain were buried by later flooding during the final Middle Bronze age, in the stratigraphic and they may hide sites at present inaccessible to sequence of the terramara of Montale (Cardarelli, 2004), investigation (Cardarelli, 1989; Cremaschi, 1991–1992). and the pollen record from the infilling of the ditches of the Owing to the high concentration of sites and the fact that terramara Santa Rosa at Poviglio indicates that, during the for each of them the primary economy was intensive Recent Bronze Age, the deforested area spread significantly agriculture (Cremaschi, 1991–1992; Cardarelli, 1997; For- (AP—arboreal pollens—rate decreasing from 27% in ni, 1997) and local pastoralism (De Grossi Mazzorin and Middle Bronze age to 19% in Recent Bronze age) (Ravazzi Riedel, 1997; Riedel, 2004), a strong exploitation of the et al., 2004). land surrounding the sites is to be expected. An increase in the clearance of woods follows a main Pollen analyses indicate that an open landscape sur- change in the settlement pattern. In their early phases, the rounded the terramare from the Middle Bronze age terramare were generally no bigger than 2 ha, while in the onward. This has been proven by the pollen diagram Recent Bronze age, within a general increase in the density recently obtained from the site of Parma (Valsecchi, 2004/ of sites, several reached considerable dimensions: Case del 2005), which dates back to this period both on the base of Lago 22 ha (Bernabo` Brea et al., 1997); Fondo Paviani the archaeological (Catarsi Dall’Aglio, 1989) and of 16 ha (Balista, 1997), Santa Rosa 7 ha (Bernabo` Brea and radiocarbon dates, recently obtained (Fig. 5). The defor- Cremaschi, 1997). estation rate is rather high all along the stratigraphic Recent work in palaeobotany has focused on the concept sequence (NAP—non-arboreal pollen—from 60% to of the pollen source area and its importance when 80%), and anthropogenic indicators, cereals particularly, interpreting spatial representation of human activity (e.g. are well represented everywhere. Sugita, 1994, 1998; Dumayne-Peaty, 2001). The Relevant However, when we compare the deforestation rate in the Source Area (RSA—the area from which the pollen region of the terramare during the Middle and Recent deposited at a site comes from) calculated for the terramara Bronze age (XVI–XII centuries BC) (Ravazzi et al., 1992, of Poviglio indicates that the area deforested and cultivated 2004, Bandini Mazzanti et al., 1996; Cardarelli, 2004)to around the site was 1 km wide during the Middle Bronze that of the surrounding regions—as indicated by the pollen age and expanded to 4 km during the Recent Bronze age records from southern Switzerland, northern Italy and (Ravazzi et al., 1992; Ravazzi et al., 2004). peninsular Italy (Fig. 6)—we perceive that the terramare When we extend the information obtained from Poviglio area appeared as a wide, artificial clearing, surrounded by to the other sites of the Enza basin, we observe that the dense forests. open areas around the individual sites intersect, indicating ARTICLE IN PRESS M. Cremaschi et al. / Quaternary International 151 (2006) 87–98 91

Fig. 5. Pollen diagram of the Parma terramara. The whole sequence dates back to the Middle Bronze age. Radiocarbon dates: cm 28 , Poz-12160 AMS 3205 730 BP; cm 212, Poz-1261 AMS 3190 730 BP. that wood cover became very scarce or completely absent technique occurred in the frame of the same cultural over most of the area (Fig. 7). context and is contemporary to the maximum extent of the Enhanced needs of cleared land for food supply, deforested areas. Therefore, it may be interpreted as compelled by the rising demographic pressure (Bernabo` evidence of strong reduction, or even exhaustion of wood Brea et al.,1997; Cardarelli, 1997; Bernabo` Brea and suitable for building purpose. A similar trend in depletion Cremaschi, 2004a), can be regarded as a main cause of of wood availability, consequent to highly demanding such intense deforestation. construction technique, is also recorded in the early-middle High demand of timber for building purpose was a Bronze age sites of the south-western Germany (Billamboz further determining factor of the deforestation: the et al., 1988). terramare were made mostly of wood. The houses of the All these circumstances persuade to hypothesize an Santa Rosa terramara during the Middle and Recent environmental scenario in which the possibility of ﬁnding Bronze ages were built on posts (Bernabo` Brea and new land to expand agriculture and pastoralism was Cremaschi, 2003), and more than 10 000 trees per hectare strongly reduced and the resources of the already exploited may have been required for their construction. area dramatically subsided. Furthermore, signs point to stressed resources during the late Recent Bronze age. A land evaluation procedure 4. The hydraulic systems in the terramara of Santa Rosa: a (FAO, 1976; Kammersmans et al., 1985) applied to the drop in water availability at the ﬁnal stage of the site soils of the Enza basin for cereals (Cremaschi, 1991–1992) indicates that the productivity was suitable for the Detailed information about the exploitation of water hypothesised annual demand of a village during the Middle resources in the terramare comes from the site of Poviglio Bronze age, but possibly inadequate to answer the Santa Rosa, located 5 km South of the present course of requirement of growing population during the late Recent the Po river, and under excavation since 19843 (Bernabo` Bronze age. Brea and Cremaschi, 2004b). The site, as clearly seen in the During the Recent Bronze age, in the terramara of Santa aerial photograph (Fig. 8), is delimited by two strips of Rosa, the houses were no longer built on piles, but set on the ground according to the block-house technique, which 3The excavation, promoted by the Soprintendenza Archaeologica needs less wood (Bernabo` Brea et al., 2003; Bernabo` Brea dell’Emilia Romagna in cooperation with the University of Milan, is and Cremaschi, 2004a). This change in construction directed by M. Bernabo` Brea and M. Cremaschi. ARTICLE IN PRESS 92 M. Cremaschi et al. / Quaternary International 151 (2006) 87–98

Fig. 6. Wood cover (AP) during the Middle and Recent Bronze ages in Italy. (1) Fiave` (Greig, 1984), (2) Lago del Frassino (Bertoldi, 1968), (3) Biandronno e Ganna (Schneider, 1978), (4) Prato Mollo (Cruise, 1990), (5) Prato Spilla (Lowe, 1992), (6) Poviglio (Ravazzi et al., 2004), (7) Montale Fig. 7. Relevant Source Area (RSA) of the terramare in the lower basin of (Cardarelli, 2004), (8) Monte Castellaccio (Bandini Mazzanti et al., 1996); the Enza river. (1) Large-sized terramare, (2) middle-sized terramare, (3) (9) Lago Padule (Lowe and Watson, 1993), (10) Lagaccione and Lago di small-sized terramare, (4) RSA of the Middle Bronze age, (5) RSA of the Vico (Magri, 1997), (11) Lago Albano (Lowe et al., 1996), (12) Valle di Recent Bronze age and (6) Apennine fringe. Castiglione (Magri, 1997), (13) Canolo Nuovo (Schneider, 1985), (14) Central Adriatic core pal 94-8 (Lowe et al., 1996) and (15) Central Adriatic core rf 93–30 (Lowe et al., 1996). AP—arboreal pollen; NAP— non-arboreal pollen. laminated deposits and material fallen from the walls, which indicates that the wells collapsed and were re- excavated several times during their life. In the last phase, clear-coloured soil consisting of the earth ramparts built they were used as reservoirs and finally sealed by dump. during its last phase. The smaller village (Villaggio Piccolo) The archaeological evidence included in them (Cremaschi to the North is the oldest part, founded during the Middle and Pizzi, in press) indicates that they were used for a long Bronze age. A U-shaped rampart delimits the larger part of period from the late Middle Bronze age (about the site (Villaggio Grande), which mostly dates to the 1400–1330 cal BC), the period of the foundation of the Recent Bronze age (Cremaschi et al., 1994). Villaggio Grande, up to its apogee, in the Recent Bronze The hydraulic system was discovered at the southern age (1330–1170 cal BC). part of the Villaggio Grande in coincidence with the Some of the wells are interconnected by small ditches, palisade and the connected structures located at the very dug out in order to bring the water outside the site, in the fringe of the site and in the peripheral moat, just outside of direction of the peripheral moat, and not inside the village. it (Fig. 9). This is particularly clear for the well-denominated US In correspondence with the palisade, and distributed (Stratigraphic Unit) 2206, which is located close to the along it, 45 large water wells were found, dug inside the south-western gate of the village. A ditch is conducted from loamy clay ground of the terramara. They were excavated the upper limit of the well with the purpose of bringing to reach two sand layers acting as aquifers at the depth of 2 water rising from the water table to fill some cisterns and 3 m, respectively, from the base of archaeological opened inside the moat (Fig. 10). deposit (Pizzi and Cremaschi, 2004). Outside the palisade, to the South, is the moat The fill of these water wells, clayey in texture, displays surrounding the village (Fig. 11). While still apparent in complex stratigraphy consisting of organic material, the present topography, it is almost sealed by sediments ARTICLE IN PRESS M. Cremaschi et al. / Quaternary International 151 (2006) 87–98 93 which were deposited long after the abandonment of the very thin, clayey and deprived of sedimentary structures, terramara, during the Roman and the Medieval periods. indicating both that the ditch was kept clean and that it The Bronze Age stratification included inside the moat is was not occupied by running water. When excavated, the moat appears to be about 30 m wide, but not very deep. While the outer bank is smooth, the inner is steep, in the shape of a slope (glacis) obtained in remodelling the fringe of the fluvial ridge on which the village was built, and is lined by a V-shaped ditch (Fig. 11). Surprisingly, its bottom has been found riddled with pits. On a surface of 1500 m2—which has been unearthed up to now—30 pits have been found. Some are clustered and interconnected and may represent hydraulic systems to distribute the water at the outer part of the terramara. Most of them are water wells: they are about 2.5 m deep, and reach a sandy aquifer, which lies 5 m under the topographic base of the site—this means 2 m deeper than the wells in the enclosure. The fill of these wells is generally rather homogeneous, clayey in texture and poor in organic matter, and it includes very scarce archaeological material. The wells appear to have been excavated in a very short time. In two cases, they are interconnected, and in this case, the deepest well is also the last to have been dug, indicating that they were frequently re-excavated with the purpose of following the descending water table’s fringe (Fig. 12). The archaeological material, while scarce, is rather significant, as it consists exclusively of sherds dating back Fig. 8. Santa Rosa terramara (RE). The aerial photograph. Light- coloured strips indicate the perimeter of the terramara. The small village to the final phase of the Recent Bronze age. It can be (Villaggio Piccolo) to the North and the large village (Villaggio Grande) concluded, therefore, that the wells inside the moat were just below. dug when those of the palisade were already deactivated:

Fig. 9. Santa Rosa terramara (RE). The DTM (Digital Terrain Model) of the southern fringe of the Villaggio Grande after the 2004 field season. The palisade is interrupted by the gate which is 4 m large. Notice the wells in correspondence with the fringe of the village and inside the moat. Darker shadow corresponds to deeper areas (survey C. Putzolu). ARTICLE IN PRESS 94 M. Cremaschi et al. / Quaternary International 151 (2006) 87–98 they might reflect the need to reach a deeper aquifer as the Brea, 2004) and by those of the Villaggio Grande is almost upper ones might have been exhausted. the same. The wells in the moat, on the contrary, reached about 2 m deeper. This means that the level of the water table was almost stable throughout the Middle Bronze age 5. A dry episode at the end of the Santa Rosa terramara and the Recent Bronze age (a time interval about three centuries long), but it dropped abruptly during the final To understand better the environmental meaning of the phase of the village, in the late Recent Bronze age (in a time event which led to the deactivation of the wells on the interval less than 50 years long). As this event cannot be fringe, the whole hydrographic system of the terramara has attributed to a river avulsion, due to the fact that the Po to be taken into consideration (Fig. 13). river during the Bronze age was deeply entrenched inside The Po river, ever neighbouring the site when it was the plain (Cremaschi, 2004), the drop of water tables has to active, was the main source for the aquifers feeding the be attributed to a depletion of the river flow, consequent to terramara. The maximum depth reached by the water wells a dry event which reduced precipitation rate and water found in the Villaggio Piccolo (Cremaschi and Bernabo` availability. As the local level of the water tables is strictly connected with the flow of the Po river, which is the main hydrological collector of a considerable part of northern Italy, the dry event observed in Poviglio is likely to have had a regional relevance. In the whole alpine region, there is evidence of a dry phase during this period. Drop in lake levels is recorded in the Garda region at Lago Lucone (Valsecchi et al., 2006). Lake level falls are also recorded in the Swiss plateau and in the Jura mountains (Magny and Richoz,1998; Magny, 2004; Holzhauser et al., 2005). During this period, the alpine glaciers were in a short withdrawal phase (Rothlis- berger, 1986). A dry event at 3100 cal yr BP is not in contrast with the palaeoclimatic record from ice cores of Greenland. A well-pronounced peak is recorded in the melt water diagram of the Agassiz Ice Cap core (Fisher and Koerner, 2003) and a negative peak of delta O18 can be seen in the GISP and GRIP cores (Dansgaard et al., 1993; Meese, et al., 1994). They both indicate a period with higher temperatures which may have resulted in enhanced aridity at middle latitudes. The end of the occupation of the Santa Rosa site is contemporary with the general abandonment of the terramare. As the drought documented at the site had a regional extension, there is a clear contiguity between these two events, and we may assume a causative link between Fig. 10. Santa Rosa terramara (RE). The well US 2206 carries the rising them (abandonment of the terramare and dry event) and water trough a ditch to some cisterns excavated inside the moat. (1) Limit infer that the climatic event may have acted as a catalyst to of the excavated area (field season 2000), (2) contour of well, (3) scarp and (4) post hole. Numbers and letters at the margin of the drawing indicate promote the abandonment of the terramare. In order to the reference grid of the excavation, the figures inside the drawing explain how a climatic event, quite minor in the frame of denominate the stratigraphic units. the Holocene climatic change, may nonetheless have

Fig. 11. Santa Rosa terramara (RE). The moat surrounding the village during the ﬁeld season 2003. Notice the pits excavated in its bottom. ARTICLE IN PRESS M. Cremaschi et al. / Quaternary International 151 (2006) 87–98 95

Fig. 12. Santa Rosa terramara (RE) Three interconnected wells inside the moat. (a) Plan and (b) DTM. Notice that the later well (US 6133) is also the deepest one (survey C. Putzolu).

Fig. 13. Cross section N-S of the terramara Santa Rosa (RE). Comparison between the water table levels during the Middle Bronze age and the final Recent Bronze age as indicated by the water wells depth. (1) Position of the cross section in the site map, (2) sand layers holding acquifers, (3) fluvial deposits (laminate sand/silt) of the Po palaeochannel, (4) shifting level of the water table: light-grey Middle and Recent Bronze age; dark-grey late Recent Bronze age and (5) the depth of the wells of the terramara. contributed to and probably determined the collapse of a experiment, though unsuccessful, in the context of the well-settled cultural system, we should understand that it European Bronze age. In contrast, with a tendency toward saddled an environment already stressed by human decline in all the regions of Europe (Berger et al., 2000), the exploitation. Several years of drought may have acted culture of the terramare produced a systematic and negatively on the basic sources of food, reducing soil intensive exploitation of the environment, quite early for fertility and rendering the feeding and watering of animals northern Italy. Intensive agriculture, pastoralism and the extremely difficult. A limit may have been reached, beyond architecture of the sites, demanding large amounts of which there was no possible answer to the demand for wood, were made possible by heavy deforestation of the resources by a rising population, and the area became whole Po plain, rendering it an artificial steppe devoted to unsuitable for life and was rapidly abandoned. cereal crops. Water was a critical resource, strictly linked to the 6. Conclusive remarks primary economy. The moats surrounding most of the sites were probably conceived to concentrate water and redis- As far as the management of the environment is tribute it to the surrounding country through a network of concerned, the terramare culture represented a unique irrigation ditches. The hydraulic system discovered at the ARTICLE IN PRESS 96 M. Cremaschi et al. / Quaternary International 151 (2006) 87–98 fringe of the terramara of Santa Rosa, consisting of large Balista, C., Cardarelli A., Steffe´, G., Labate, D., Mazzoni, C., 2003. water wells and interconnecting ditches, suffered a clear Terramara di Gaggio. In Ministero per i Beni e le Attivita` Culturali, drop in the water table level during the late Recent Bronze Soprintendenza Archeologica dell’Emilia Romagna, Museo Arche- age, and in the final phase of the occupation of the site ologico Nazionale di Parma (Eds.). Archeologia ad Alta Velocita` In Emilia, Abstracts of the Meeting: Indagini archeologiche e geologiche coincided with the abandonment of all the sites in the lungo il tracciato ferroviario ad Alta Velocita` , Cepav uno, Parma, pp. terramare region. 25–28. The hydrological crisis recorded at Santa Rosa is the Bandini Mazzanti, M., Mercuri, A.M., Accorsi, A.M., 1996. Primi dati local expression of a regional dry event which is also palinologici sul sito di Monte Castellaccio (76 m. s.l.m., 441210N 0 recorded in the Alpine area and beyond, in the North 11142 E, Imola, Bologna). In: Pacciarelli, M. (Ed.), La collezione Scarabelli.2. La Preistoria, Grafis Edizioni, Imola, pp. 158–175. Atlantic area. Barfield, L., 1994. The Bronze Age of Northern Italy: recent work and The chronological contiguity, supported by the archae- social interpretation. In: Mathers, C., Stoddart, S. (Eds.), The ological context, between the collapse of the terramare Mediterranean Cultures during the Bronze. Development and decline system and the dry episode recorded in Poviglio, suggests a in the Mediterranean Bronze Age. Sheffield Archaeological Mono- causative link between the two events. graphs, vol. 8. Collins J.R. Publications, Sheffield, pp. 129–144. The period of dry climate affected an environment Berger, J.F., Magnin, F., Thiebault, S., Vital, J., 2000. Emprise et de´prise culturelle dans l’Age du Bronze: l’exemple du Bassin Valdainais already stressed by over-exploitation of natural resources (Drome) et de la moyenne valleé du Rhone. Bulletin de la Socie´te´ by a large demographic increase, and its consequences on Pre´historique Franc-aise, tome 97 (1), 95–119. the sources of the primary economy (soils and livestock) Bernabo` Brea, M., Cardarelli, A., 1997. Le terramare nel tempo. In: may have rendered the area unsuitable for living and Bernabo` Brea, M., Cardarelli, A., Cremaschi, M. (Eds.), Le terramare caused its abandonment. La piu` antica civilta` padana. Electa, Milano, pp. 295–301. Bernabo` Brea, M., Cremaschi, M., 1997. Le terramare: ‘palafitte a secco’ o The case of the terramara rise the question of the role of ‘villaggi arginati’. In: Bernabo` Brea, M., Cardarelli, A., Cremaschi, M. short climatic changes in societal collapse which appears to (Eds.), Le terramare La piu` antica civilta` padana. Electa, Milano, pp. have been a primary agent in several cases from the 187–196. Mediterranean area, Middle East and meso-America over Bernabo` Brea, M., Cremaschi, M., 2003. Il Villaggio Piccolo nella the Holocene (Weiss, 2000; De Menocal, 2001; Weiss and terramara S. Rosa di Poviglio (RE): le strutture abitative. In: Atti della Bradley, 2001; Catto and Catto, 2004). XXXV Riunione Scientifica I.I.P.P. Le comunita` della preistoria italiana. Studi e ricerche sul neolitico e le eta` dei metalli (Castello di The present research confirms a multi-causal explanation Lipari, 2–7 giugno 2000), I.I.P.P., Firenze, pp. 145–157. for the terramare disappearance, in which the climatic Bernabo` Brea, M., Cremaschi, M., 2004a. La terramara di Santa Rosa di component, after the recent discovery in Santa Rosa, Poviglio nel corso del Bronzo Recente. In: Atti del Convegno L’eta` del appears likely to have played a role. Nevertheless, the main Bronzo Recente in Italia (Lido di Camaiore 26–29 ottobre 2000), F. factor still remains the environmental stress induced by Baroni, Viareggio, pp. 101–111. anthropogenic over-exploitation of resources and uncon- Bernabo` Brea, M., Cremaschi, M. (Eds.), 2004b. Il Villaggio Piccolo della terramara di S. Rosa di Poviglio (Scavi 1987/1992). I.I.P.P., Firenze, p. trolled demographic pressure. This interpretation is further 890. supported by the fact that in the areas surrounding the Bernabo` Brea, M., Mutti, A., 1994. ‘‘Le terramare si scavano per terramare, which were marginal during the apogee, the concimare I prati’’. La nascita dell’archeologia preistorica in Parma civilisation (specifically the Apennine and the Veneto nella seconda meta` dell’Ottocento. Silva, Parma, p. 482. plain), there was no break in occupation and the local Bernabo` Brea, M., Cardarelli, A., Cremaschi, M., 1997. Il crollo del sistema terramaricolo. In: Bernabo` Brea, M., Cardarelli, A., Cre- cultures of the Final Bronze age overcame the crisis which maschi, M. (Eds.), Le terramare. La piu` antica civilta` padana. Electa, was fatal to the terramare (Bernabo` Brea et al., 1997; Bietti Milano, pp. 745–756. Sestieri, 1997). Bernabo` Brea, M., Cremaschi, M., Pizzi, C., 2003. Le strutture abitative del Villaggio Grande-fase su palafitta- della terramara di S. Rosa di Poviglio (RE), in Atti del Convegno ‘Analisi informatizzata e trattamento dati delle strutture di abitato di eta` Preistorica e References Protostorica in Italia’, I.I.P.P., Firenze, pp. 271–285. Bertoldi, R., 1968. Ricerche pollinologiche sullo sviluppo della vegeta- Balista, C., 1997. Fossati, canali e paleoalvei: connessioni nevralgiche per zione tardiglaciale e postglaciale nella regione del lago di Garda. Studi l’impianto e la sopravvivenza dei grandi siti terramaricoli di bassa Trentini di Scienze Naturali Sez B 45, 87–162. pianura. In: Bernabo` Brea, M., Cardarelli, A., Cremaschi, M. (Eds.), Bietti Sestieri, A.M., 1997. Il territorio padano dopo le terramare. In: Le terramare. La piu` antica civilta` padana. Electa, Milano, pp. Bernabo` Brea, M., Cardarelli, A., Cremaschi, M. (Eds.), Le terramare. 126–136. La piu` antica civilta` padana. Electa, Milano, pp. 757–772. Balista, C., 2002. La paleoidrografia dell’area terramaricola centro- Billamboz, A., Keefer, E., Koninger, J., Torke W., 1988. La Transition padana verso la fine dell’ eta` del Bronzo: inquadramento stratigrafico, Bronze Ancien-Moyen dans le sud-ouest de L’Allegmagne a` l’example cronologico e paleoclimatico. Quaderni della bassa modenese 42,16I,2, de deux stations de l’habitat palustre (Station Forschener, Federsee) et 7–48. littoral ( Bodman-Schachen I, Bodensee), Anonymous, Proceeding of Balista, C., De Guio, A., 1990–1991. Il sito di Fabbrica dei Soci the 1131 Congre` s National des Socie´te´Savantes: Pre´-Protohistoire, (Villabartolomea, VR): oltre la superficie. Padusa 26–27, 111–117. Dynamique du Bronze moyen. Strasbourgh, pp. 51–78. Balista, C., De Guio, A., 1997. Ambienti ed insediamento nell’eta` del Bronzoni, L., Cremaschi, M., 1989. La terramara di Case Cocconi (RE): Bronzo nelle Valli Grandi Veronesi. In: Bernabo` Brea, M., Cardarelli, strutture e materiali. Padusa 25, 173–226. A., Cremaschi, M. (Eds.), Le terramare. La piu` antica civilta` padana. Cardarelli, A., 1989. L’eta` del Bronzo nel Modenese: organizzazione del Electa, Milano, pp. 137–170. territorio, forme economiche, strutture sociali. Modena dalle origini ARTICLE IN PRESS M. Cremaschi et al. / Quaternary International 151 (2006) 87–98 97

all’anno Mille. Studi di archeologia e storia, I. Edizioni Panini, Forni, G., 1997. Le tecniche agricole nelle terramare. In: Bernabo` Brea, Modena, pp. 86–127. M., Cardarelli, A., Cremaschi, M. (Eds.), La Piu` Antica Civilta` Cardarelli, A., 1997. Evolution of settlement and demography in Padana. Electa, Milano, pp. 457–468. terramare culture. In: Rittershofer, K.S., (Ed.), Demographie der Greig, J., 1984. A preliminary report on the pollen diagram and some Bronzen Zeit. Palaeo Demographie Moglichkeiten und Grenzen. results from Palafitta Fiave` . In: Perini, R. (Ed.), Scavi archeologici Internationale Archaeologie, vol. 36. pp. 230–237. nella zona palafitticola di Fiave` Carera, parte I, campagne 1969–1976. Cardarelli, A., 2004. Parco Archeologico e Museo All’aperto Della Patrimonio storico ed artistico del Trentino, vol. 8. Provincia Terramara di Montale. Nuova Grafica, Carpi, p. 42. Autonoma di Trento, Trento, pp. 305–322. Catarsi Dall’Aglio, M., 1989. La terramara di Parma. Padusa 25, 237–337. Holzhauser, H., Magny, M., Zumbuhl, H.J., 2005. Glacier and lake-level Catto, N., Catto, G., 2004. Climate changes, communities and civiliza- variations in western-central Europe over the last 3500 years. Holocene tions: driving force, supporting player, or blackground noise? 15,6, 789–801. Quaternary International 123–125, 7–10. Kammersmans, H., Loving, S., Voorips, A., 1985. Changing patterns of Chierici, G., 1871. Le antichita` preromane della provincia di Reggio prehistoric land use in the Agro Pontino. In: Malone, C., Stoddart, S. nell’Emilia. Tipografia Artigianelli, Reggio Emilia, p. 35. (Eds.), Paper in Italian Archaeology IV, I. The Human Landscape, Cremaschi, M., 1991–1992. Economia ed uso del territorio: possibile crisi B.A.R. International Series 243, 53–68. ambientale durante il Bronzo Recente. In: Atti del Convegno L’eta` del Lowe, J.J., 1992. Late Glacial and early Holocene lake sediments from the Bronzo in Italia nei sec. dal XVI al XIV a.C., (Lido di Camaiore 26–30 Northern Apennine Italy. Pollen stratigraphy and radiocarbon dating. ottobre 1989). Rassegna di Archeologia 10, 180–182. Boreas 21, 193–208. Cremaschi, M., 1997. Terramare e paesaggio padano. In: Bernabo` Brea, Lowe, J.J., Watson, C., 1993. Late Glacial and Early Holocene pollen M., Cardarelli, A., Cremaschi (Eds.), Le Terramare. La piu` antica stratigraphy of the Northern Apennines, Italy. Quaternary Science civilta` padana. Electa, Milano, pp. 107–125. Reviews 12, 727–738. Cremaschi, M., 2004. La terramara di Santa Rosa ed il suo territorio: Lowe, J.J., Accorsi, C.A., Bandini Mazzanti, M., Bishop, A., Van der aspetti geomorfologici. In: Bernabo` Brea, M., Cremaschi, M. (Eds.), Il Kaars, S., Forlani, L., Mercuri, A.M., Rivalenti, C., Torrei, P., Villaggio Piccolo della terramara di S. Rosa di Poviglio (Scavi 1987/ Watson, C., 1996. Pollen stratigraphy of sediment sequences from 1992). I.I.P.P., Firenze, pp. 21–37. lakes Albano and Nemi and from the central Adriatic spanning the Cremaschi, M., Bernabo` Brea, M., 2004. Gli elementi strutturali. In: interval from oxygen isotope stage 2 to present day. Memorie dell’Istituto Italiano di Idrobiologia 55, 71–98. Bernabo` Brea, M., Cremaschi, M. (Eds.), Il Villaggio Piccolo della Magny, M., 2004. Holocene climate variability as reflected by mid- terramara di S. Rosa di Poviglio (Scavi 1987/1992). I.I.P.P., Firenze, European lake-level fluctuations and its probable impact on prehistoric pp. 151–184. human settlements. In: Brauer, A., Guilizzoni, P. (Eds.), The record of Cremaschi, M., Pizzi, C. I pozzi perimetrali del Villaggio Grande di Santa Human/Climate Interaction in Lake Sediments. Quaternary Interna- Rosa di Poviglio (RE). Uso delle risorse idriche tra la fine del Bronzo tional 113, 65–79. Medio e il Bronzo Recente. In: AA. VV., Studi in onore di Renato Magny, M., Richoz, I., 1998. Holocene lake-level fluctuations in Lake Peroni, in press. Seedorf, southern Swiss Plateau. Eclogae Geologicae Helveticae 91, Cremaschi, M., Ferretti, A., Forte, M., 1994. Tecniche digitali e di 345–347. visualizzazione in geoarcheologia: il caso di studio della terramara Magri, D., 1997. Middle and Late Holocene vegetation and climate Santa Rosa di Poviglio. Archeologia e Calcolatori 5, 305–316. changes in peninsular Italy. In: Dalfes, H.N., Weiss H., Kukla, G., Cruise, G.M., 1990. Pollen Stratigraphy of two Holocene peat sites in the (Eds.), Third Millennium BC Climate Change and Old World Collapse Ligurian Apennines, north Italy. Review of Paleobotany and Nato Asi Series 1, vol. 49. pp. 517–530. Palynology 63, 299–313. Meese, D.A., Gow, A.J., Grootes, P., Mayewski, P.A., Ram, M., Stuiver, Dansgaard, W., Johnsen, S.J., Clausen, H.B., Dahl-Jensen, D., Gundestr- M., Taylor, K.C., Waddington, E.D., Zielinski, G.A., 1994. The up, N.S., Hammer, C.U., Hvidberg, C.S., Steffensen, J.P., Sveinb- accumulation record from the GISP2 core as an indicator of climate jornsdottir, A.E., Jouzel, J., Bond, G., 1993. Evidence for general change throughout the Holocene. Science 266, 1680–1682. instability of the past climate from 250-kyr ice core record. Nature 364, Mutti, A., 1993. Caratteristiche e problemi del popolamento terramaricolo 218–220. in Emilia occidentale. University Press, Bologna, p. 400. De Grossi Mazzorin, J., Riedel, A., 1997. La fauna delle terramare. In: Pearce, M., 1998. New research on the terramare of northern Italy. Bernabo` Brea, M., Cardarelli, A., Cremaschi, M. (Eds.), Le Antiquity 72, 743–746. Terramare. La piu` antica civilta` padana. Electa, Milano, pp. 475–481. Pizzi, C., Cremaschi, M., 2004. I pozzi lungo la struttura di recinzione del De Marinis, R., 1975. L’eta` del Bronzo. In: Cremaschi, M. (Ed.), Villaggio Grande di Poviglio: caratteri e processi formativi. In: Cocchi, Preistoria e Protostoria nel Reggiano Ricerche e scavi 1940–1975. L., Genick, D. (Eds.), Atti del Congresso Nazionale ‘‘L’eta` del Bronzo Civici Musei, Reggio Emilia, pp. 31–55. Recente in Italia’’ (Lido di Camaiore 26/29 ottobre 2000). Baroni, De Menocal, P.B., 2001. Cultural responses to climate change during the Viareggio, pp. 512–513. late Holocene. Science 292, 667–673. Ravazzi, C., Cremaschi, M., Forlani, L., 1992. Ricostruzione della storia Dumayne-Peaty, L., 2001. Human impact on vegetation. In: Brothwell, della vegetazione padana tra l’eta` del Bronzo e l’Alto Medioevo in D.R., Pollard, A.M. (Eds.), Handbook of Archaeological Sciences. relazione all’intervento antropico. La successione pollinica del fossato Wiley, Chichester, pp. 379–392. della terramara di Poviglio (RE). Archivio Botanico Italiano 6 (3–4), FAO, 1976. A Framework of Land Evaluation. ILRI, Wageningen, p. 198–220. 198. Ravazzi, C., Cremaschi, M., Forlani, L., 2004. Studio archeopalinologico Ferri, R., 1996. La fotografia aerea in zone di bassa pianura: iconografia della terramara S. Rosa. Nuovi dati, analisi floristica e sintassonomica analitica delle tracce fluviali ed evoluzione idrografica delle Valli della vegetazione nell’eta` del Bronzo. In: Bernabo` Brea, M., Grandi Veronesi. In: AAVV, La ricerca archeologica di superficie in Cremaschi, M. (Eds.), Il Villaggio Piccolo della terramara di S. Rosa area padana. AGS Edizioni, Padova, pp. 25–33. di Poviglio (Scavi 1987/1992). I.I.P.P., Firenze, pp. 703–736. Fisher, D.A., Koerner, R.M., 2003. Holocene ice core climate history: a Riedel, A., 2004. La Fauna. In: Bernabo` Brea, M., Cremaschi, M. (Eds.), multi-variable approach. In: Mackay, A.W., Battabe, R.W., Birks, Il Villaggio Piccolo della terramara di S. Rosa di Poviglio (Scavi 1987/ H.J.B., Oldfield, F. (Eds.), Global Change in the Holocene: 1992). I.I.P.P., Firenze, pp. 703–736. Approaches to Reconstructing Fine—Resolution Climate Change. Rothlisberger, F., 1986. 10,000 Jahre Gletschergeschichte der Erde. Verlag Arnold, London, pp. 228–232. Sauerlander, Arau, p. 52. ARTICLE IN PRESS 98 M. Cremaschi et al. / Quaternary International 151 (2006) 87–98

Sa¨ﬂund, G., 1939. Le Terramare delle province di Modena, Reggio, Parma Valsecchi, V., 2004/2005. Holocene vegetation dynamics in the Northern e Piacenza. Acta Instituti Romani Regni Sueciae 7, 225. Italian peninsula and in Syria using palaeoecological analyses: human Schneider, R., 1978. Pollenanalytische Untersuchungen zur Kenntnis der impact on vegetation versus climate changes. Unpublished Ph.D. pat- und postglazialen Vegetetiongeschichte an Suidrand der Alpen Thesis, University of Milan. zwischen Turin und Varese (Italien). Botanische Jahrbucher 100, Valsecchi, V., Tinner, W., Finsinger, W., Ammann, B., 2006. Human 26–109. impact during the Bronze Age: preliminary results from a key site— Schneider, R., 1985. Analyse palynologique dans l’Aspromonte en Lago Lucone (northern Italy). Vegetation History and Archaeobotany Calabre (Italie Meridionale). Cahiers Ligures de Pre´histoire et de 15 (2), 99–113. Protohistoire, nouvelle se` rie 2, 279–288. Weiss, H., 2000. Beyond the younger Dryas, collapse as adaptation Sugita, S., 1994. Pollen representation of vegetation in Quaternary sediments: theory and method in patchy vegetation. Journal of to abrupt climatic change in ancient west Asia and the Ecology 82, 881–897. eastern Mediterranean. In: Bawden, G., Reycraft, R.M. (Eds.), Sugita, S., 1998. Modelling pollen representation of vegetation. Quatiﬁca- Environment and the Archaeology of Human Response. Anthro- tion of land surfaces cleared of forests during the Holocene. In: pological papers, vol. 7. Maxwell Museum of Anthropology, Gaillard, M.J., Berglund, B.E., Frenzel, B., Huckriede, U. (Eds.), Albuquerques, pp. 75–98. Palaoklimaforschung/Palaeoclimate Research, vol. 27. Gustav Fischer Weiss, H., Bradley, R.S., 2001. What drives societal collapse? Science 291, Verlag, Stuggart, pp. 1–16. 609–610.