JOURNAL OF QUATERNARY SCIENCE (1997) 12 (6) 507–518  1997 by John Wiley & Sons, Ltd. CCC 0267-8179/97/060507–12$17.50 Cesi, an early M iddle Pleistocene site in the Colfiorito Basin (U mbro-M archean Apennine), central Italy G. FICCARELLI 1,* , L. ABBAZZI 1, A. ALBIAN ELLI 1 , A. BERTIN I 1, M . CO LTO RTI 2 , M . M AGN ATTI 3, F. M ASIN I 4, P. M AZZA5, C. M EZZABO TTA1 , G. N APO LEO N E1 , L. RO O K1 , M . RU STIO N I 1 and D . TO RRE1 1 Department of Earth Sciences, University of Florence, Via La Pira 4, I-50121 Florence, Italy 2 Department of Earth Sciences, University of Siena, Via delle Cerchia 3, I-53100 Siena, Italy 3 Museum of Natural Science, University of Camerino, Via del Bastione 2, I-62032 Camerino, Italy 4 Department of Geology and Geodesy, University of Palermo, C.so Tukory 131, I-90134, Palermo, Italy 5 Museum of Geology and Palaeontology, University of Florence, Via La Pira, I-50121 Florence, Italy Ficcarelli, G., Abbazzi, L., Albianelli, A., Bertini, B., Coltorti, M., Magnatti, M., Masini, F., Mazza, P., Mezzabotta, C., Napoleone, G., Rook, L., Rustioni, M. and Torre, D. 1997. Cesi, an early Middle Pleistocene site in the Colfiorito Basin (Umbro-Marchean Apennine), central Italy. J. Quaternary Sci., Vol. 12 , 000– 000. ISSN 0267-8179 (No. of Figures: 8 No. of Tables: 0 No. of References: 39) Received 3 October 1996 Revised 10 July 1997 Accepted 12 July 1997 ABSTRACT: Near the village of Cesi, at the head of the Chienti River, in the Colfiorito Basin (Umbro-Marchean Apennines, central Italy), fluvio-lacustrine deposits have yielded mammal fossil remains. The results of a multidisciplinary investigation indicate that the vertebrate-bearing sediments date about 700 ka and accordingly provide a fossil assemblage for the Middle–Late Galerian. Palynological investigations carried out from sediments underlying the fossiliferous level suggest predominantly cold and dry conditions, whereas the fauna suggests a slight climatic amelioration towards cool and moist conditions in the uppermost part of the sequence.  1997 by John Wiley & Sons, Ltd. KEYWORDS: geomorphology; Mammalia; palynology; palaeomagnetism; early Middle Pleistocene. Introduction In 1987, researchers from the Department of Earth Sciences and the Museum of Geology and Palaeontology of the University of Florence, the Department of Earth Sciences of the University of Camerino and the Soprintendenza Archeologica of Ancona carried out palaeontological investigations in the Colfiorito Basin in Central Italy. At least five fossiliferous sites were found, the most promising being those at Colle Curti and Cesi. Priority was given to the area of Colle Curti and excavations were carried out for five years at this locality. The results of these investigations, published in a series of papers (Borselli et al ., 1988; Ficcarelli and Mazza, 1990; Ficcarelli et al ., 1990; Ficcarelli and Silvestrini, 1991), were so significant that they led to the establishment of a ‘Colle Curti’ Faunal Unit, which marks the beginning of the Galerian Mammal age in Italy. In their latest paper, Ficcarelli and Silvestrini (1991) tentatively suggested that the Cesi locality, then known from only * Correspondence to: G. Ficcarelli, Universita degli Studi di Firenze, Dipartimento di Scienze della Terra, 50121 Firenze, Via G. la Pira 4, Italy. L. Abbazzi, G. Ficcarelli, M. Magnatti, F. Masini, P. Mazza, C. Mezzabotta, L. Rook, M. Rustioni and D. Torre, dealt with vertebrate palaeontology; Albianelli, A. and Napoleone, G. with magnetostratigraphy; Bertini, A. with palynology; Coltorti, M. with geomorphology and stratigraphy. a few superficial finds, could be approximately coeval with that at Colle Curti and therefore extended the faunal list of the latter site to include also Bison schoetensacki , the occurrence of which was known only from a lower third molar at Cesi (see Fig. 4e). In 1993, the same researchers carried out systematic excavations in the Cesi area. A fossiliferous bed was found, which seems to be as important as that at Colle Curti. The results of the studies refuted Ficcarelli and Silvestrini’s (1991) supposition, that the fossiliferous layer of Cesi could be coeval with that from Colle Curti. Geological and geomorphological background The Cesi site is located at the head of the Chienti River, in the Colfiorito Basin (Umbro-Marchean Apennines), at about 820 m above sea level (Figs 1 and 2). Here the bedrock comprises limestones overlain by pelagic and hemipelagic sediments (Upper Triassic–Lower Miocene), the total thickness of which may exceed 2000 m. Lower and Upper Miocene siliciclastic turbidites, up to hundreds of metres thick, are locally preserved in the chain. The mountain chain 508 JOURNAL OF QUATERNARY SCIENCE ˆ Figure 1 Structural setting of the central northern Apennines with the main relationships between the Tuscany Domain (Unita Falterona– Trasimeno) and the Latium–Abruzzi Platform (PLA) from Calamita, 1990). AM, Amelia; ANT, Antrodoco; BT, Battiferro; CAS, Cascia; CO, Contigliano; CL, Cottanello; CT, Castelluccio; FO, Foligno; GU, Gubbio; MA, Monte Acuto; MAG, Monte Aguzzo; MB, Monte Bove; MC, Monte Cavallo; MCO, Monte Cosce; MCS, Monte Coscerno; MF, Monte Fema; MG, Monte Gorzano; MM, Monte Maggiore; MMA, Monte Malbe; MMR, Monte Martano; MN, Monte Nerone; MP, Monte Patino; MPE, Monte Pellecchia; MPG, Monte Peglia; MPR, Monte Primo; MS, Monte Subasio; MST, Monte della Strega; MSV, Monte San Vicino; MT, Monte Tezio; MTO, Monte Tolentino; MTR, Monte Torricella; MV, Monte Vettore; MZ, Monte Zappi; NA, Narni; NO, Norcia; OR, Orte; PO, Posta; SFE, Sassoferrato; SP, Spoleto; TV, Tivoli. J. Quaternary Sci., Vol. 12(6) 507–518 (1997)  1997 by John Wiley & Sons, Ltd. EARLY MIDDLE PLEISTOCENE SITE IN THE APENNINE is bounded to the east by the Sibillini Mountain thrust, which displaced these sediments over the Lower Miocene to Lower Pliocene turbiditic deposits, exceeding 3000 m in thickness, of the Periadriatic Basin (Fig. 1). Reverse faults and overthrusts affected the sequence in the Early Pliocene (Boccaletti et al ., 1983, 1986; Bally et al ., 1986; Calamita and Deiana, 1988; Lavecchia et al ., 1988; Calamita et al ., 1991; Damiani et al ., 1991). The area was first affected by uplift and finally emerged above sea-level at the end of the Messinian (Ambrosetti et al ., 1978, 1982, 1987; Calamita et al ., 1991, 1995). However, a compressional basin continued to exist in the East Tiber area up to the Late Pliocene and was infilled by fluviatile and lacustrine deposits (Coltorti and Pieruccini, in press). In the Early–Middle Pleistocene, extensional tectonics reached the Umbria-Marchean Apennines, and many basins were formed (Calamita et al ., 1982, 1994, 1995; Raffy, 1981; Cattuto et al ., 1992). Numerous thrusts are present in the study area, to the east of the eastern Tiber Basin (Calamita and Pizzi, 1992; Calamita et al ., 1994), as well as many normal faults with an Apennine direction, which also dissect the thrust planes and create important fault escarpments. A ‘planation surface’ is developed on the units formed during the compressional phase, and normal faults dissect this surface (Fig. 2) (Coltorti and Farabollini, 1995; Coltorti and Pieruccini, in press; Coltorti et al ., in press). In the Spoleto Basin, this surface progressively dips under sediments older than 3.3 Ma and it is therefore dated to the Early Pliocene. At the same time, in the Periadriatic basin, east of the mountain area, thick and coarse fan-delta deposits (Ascensione Mountain, Fig. 1) (Cantalamessa et al ., 1986), which originated from the dissection of the planation surface, were emplaced. Later, wide valleys, separated by smooth relief, were formed across all of the Apennine chain affected by progressive uplifting between the Middle Pliocene and Early Pleistocene. Also, the development of this secondary landscape has been attributed to erosional processes that occurred in proximity to base-level (Desplanques, 1969; Coltorti, 1981; Calamita et al ., 1982; Ciccacci et al ., 1985), with palaeovalleys lacking a definite flow direction (Coltorti and Farabollini, 1995). However, Dramis et al . (1991) and Dramis (1992) suggest the existence of pedimentary processes at the mountain front (Periadriatic Basin, Fig. 1) and therefore the possibility that these valleys also developed at high elevations. Previous studies in the Colfiorito (Coltorti et al ., in press) and recent investigations in Castelluccio di Norcia areas (Fig. 1) (Coltorti and Farabolini, 1995), where there are also wide palaeovalleys lacking a definite flow direction on the watershed, seem to confirm the first hypothesis. Moreover, in the Periadriatic area, littoral sediments coeval with the deposits that contain the mammal fauna, were uplifted to over 400 m a.s.l. Moreover, convergent fluvial terraces reveal that the mountain area was uplifted further than the coastal region (Coltorti et al ., 1991), suggesting that the earlier elevation (today about 800 m) was probably less than 400 m a.s.l. The sequence under investigation here, as well as the nearby Colle Curti deposits, represents the infill of one of these valleys (Coltorti et al ., in press). The transition to the present-day landscape was favoured by the deepening of the valleys that followed the increased uplift after the final part of the Early Pleistocene. The activity of anti-Apennine and extensional Apennine faults on the Tyrrhenian side of the Apennine chain occurred in a subsequent movement, causing a series of captures in the drainage system. The Cesi and Colle Curti basins originally  1997 by John Wiley & Sons, Ltd. 509 belonged to the same hydrographic network, but were later separated by both deformation resulting from the normal faults located east of the upper Chienti valley, and the intensive regressive erosion of the Percanestro stream, a tributary of the Nera–Tiber Basin. The nearby Colfiorito Basin represents a relic of the former landscape that was slightly affected by Middle and Late Pleistocene erosional events. In fact, the whole area is characterised by a very low-energy relief, despite being located high on the Adriatic–Tyrrhenian watershed. Stratigraphy The limited erosion that affected the upper part of the Chienti Basin did not expose any great thickness of Pleistocene sediments. Late Middle and late Upper Pleistocene stratified slope deposits (Coltorti and Dramis, 1988) outcrop along the slopes of the valley and in the nearby Colfiorito Basin. An alluvial fan entered the lacustrine basin that occupied the Colfiorito plain during the late Pleniglacial, and lacustrine sedimentation ended at the beginning of the Holocene (Brugiapaglia and de Beaulieu, 1995). This was associated with the strong solution processes that affected the bedrock, and which is also indicated by fractured and faulted limestones. In fact, karstic wells and dolines are common in the area, but during the cold phases of the Pleistocene were filled with debris. The lacustrine basin was re-established during the third millenium bc (Brugiapaglia and de Beaulieu, 1995), probably due to slope degradation induced by the first clearing of the natural vegetation and then drained again artificially in the Middle Ages. In fact, pollen of cereals are found in the sediments overlying the unconformity/nondepositional surface that separates the lower (Upper Pleistocene) from the upper (Holocene) part of the sequence in the borehole. In the Cesi valley, near Madonna del Piano, the palaeontological excavations exposed a sequence about 12 mm thick, which represents the remnant of a thicker alluvial terrace. The lower part of the outcrop is represented by Lower– Middle Pleistocene sediments, predominantly clay-rich sediments with pebbly lenses. The total thickness of these sediments has never been precisely established. At the palaeontological site (Fig. 3), a trench exposed over 8 m of massive clays (bed US 5); thin stone-lines occur, a few decimetres from one another. Fossil bones are preserved in a fine gravel bed 20–50 cm thick (bed US 4), formed of subangular and angular siliceous gravels, resulting from decalcification of limestone clasts, which were locally affected by load casts. A thin pyroclastic lamina (bed US 4a) discontinuously occurs at the base of the gravels. Oscillation of the water table is responsible for the occurrence of Fe– Mn concretions and nodules in the bed. In the same horizon, centimetre-thick calcareous crusts occur linked with the presence of herbaceous roots. Gravels, weathered in a tropical reddish palaeosol, of which only the flint elements remain, occur at the top of the preserved sequence of the terrace. However, in other parts of the basin, in the same morphoand lithostratigraphic unit, limestone gravels are still preserved locally. The evidence suggests a lacustrine environment succeeded in its upper part by clastic material derived from an alluvial fan prograding into the lake. However, the upper part of this sequence was also affected locally by erosional processes, which removed tens of metres of sediments. J. Quaternary Sci., Vol. 12(6) 507–518 (1997) 510 J. Quaternary Sci., Vol. 12(6) 507–518 (1997) JOURNAL OF QUATERNARY SCIENCE  1997 by John Wiley & Sons, Ltd. EARLY MIDDLE PLEISTOCENE SITE IN THE APENNINE 511 Figure 3 Stratigraphy at the site of Cesi (Madonna del Piano). Legend: 1, clays; 2, pyroclastic layer (US 40); 3, limestone gravels; 4, flint gravels; 5, stone lines; 6, alluvial soils. The stratigraphical position of the samples taken for palynological (*) and palaeomagnetic (o) analyses, is indicated. Analogous deeply weathered gravels also characterize the top of the nearby Colle Curti sequence. Pyroclastic sediments, containing sanidine crystals dated, using the Ar/Ar method, to 424 kyr BP, are reported from a thin palustrine layer (Coltorti et al ., in press). In Colle Curti, these sediments overlie clayey beds over 20 m thick. Beneath the clays, intercalations of gravelly lenses and beds become progressively more abundant, down to the base of the outcropping sequence, which is about 100 m thick. The Colle Curti fossil assemblage was recovered at the base of the clays. The Cesi Figure 2 The Colle Curti and Cesi area in its geomorphological setting: 1, major fluvial escarpment; 2, gullies and streams; 3, alluvian fans; 4, water-fall; 5, major captures; 6, hanging Lower Pleistocene palaeovalleys; 7, trough-floored valleys; 8, major extensional fault at time with associated fault escarpments (tract in the lowered part); 9, major overthrusts; 10, major fault escarpment; 11, ‘flat iron’ forms associated with the main extensional fault of the East Tiber Basin; 12, major structural escarpment; 13, ridges with sharp crest; 14, colluvial sediments; 15, major landslides; 16, final Middle Pleistocene, Upper Pleistocene and Holocene alluvial sediments; 17, final Middle Pleistocene and Holocene travertines; 18, Upper Pleistocene and Holocene lacustrine deposits; 19, final Middle Pleistocene and Upper Pleistocene alluvial and fluvial deposits; 20, Lower Pleistocene alluvial and fluvial deposits; 21, remnants of the planation surface; 22, palaeontological sites of Colle Curti (A) and Cesi (B). Note the very flat nature of the planation surface remnants preserved at the top of the relief. It affects all the formations strongly folded and overthrust during the Messinian. A series of palaeovalleys dissects the planation surface and the Colle Curti deposits bear witness to a period of successive filling. Extensional faults were activated later and created a complex network, which first generated a clear watershed between the Adriatic and the Tyrrhenian side.  1997 by John Wiley & Sons, Ltd. J. Quaternary Sci., Vol. 12(6) 507–518 (1997) 512 JOURNAL OF QUATERNARY SCIENCE sequence shares many sedimentological characters with that at Colle Curti, although most of the upper part is missing, probably as a result of later erosion. M ammal palaeontology The taxa identified from Cesi are an advanced representative of Stephanorhinus hundsheimensis, a caballine equid, Hippopotamus sp., Megaceroides solilhacus, Cervus elaphus, Dama clactoniana, Bison schoetensacki , Homotherium (Fig. 4) and an undetermined elephant. All this material is kept in the Museum of Natural Science, University of Camerino. No micromammals nor any other microvertebrates have been recovered. The rhinoceros material is represented by a few poorly preserved bones of a single individual; a mandible, a right second metacarpal bone, a right femur and a right tibia. The morphological characteristics and proportions of the specimens are suggestive of Stephanorhinus hundsheimensis. The femur is fairly diagnostic for recognising Pleistocene European rhinoceroses. The specimen is massive, with a very broad roundish head, a short sturdy neck, a prominent greater trochanter and a well-developed lesser trochanter. Slight disproportions in the lower cheek teeth distinguish the Cesi rhinoceros from other representatives of S. hundsheimensis. However, as this rhinoceros is one of the largest sized representatives of the species, the differences in the dentition may probably reflect allometry (Fig. 4f). Stephanorhinus hundsheimensis probably inhabited fairly open regions (Mazza, 1993) such as wooded steppes, probably occupying a niche similar to that of the living black rhinoceros, Diceros bicornis. The equid is represented only by a right metatarsal bone, slightly deformed and fractured in its distal epiphysis, which prevents a species determination. The hippopotamus remains consist of very fragmentary material: two lower incisors, two lower canines, two scaplae, a humerus, two fragments of a pelvis, a femur, a patella and two tibiae. Again, the very poor state of preservation prevents a determination to species level. Scanty cervid remains were found. Some large-sized and stout remains (fragments of antlers and of antlered skulls, and a calcaneum) indicate the occurrence of a megalocerine. In particular a distal part of a beam, which expands in a broad palmation characterised by several terminal points at the posterior border, allows reference to Megaceroides solilhacus (Fig. 5). Cervus elaphus is represented only by a fragmentary mandible, still bearing the first and second molar. The occurrence of Dama clactoniana is indicated by a third lower molar, the distal fragments of two humeri, a right one and a left one, and by a third phalanx. Although among these remains diagnostic elements, such as antlers, are absent, their larger size in comparison to those of Dama dama allows them to be referred to Dama clactoniana (Fig. 4g). A large-sized bovid is represented by an incomplete mandible, a fourth lower premolar, a third lower molar and a metacarpal bone, all referable to the genus Bison. Although these specimens are poorly diagnostic for specific determination, their overall morphology and size compare with B. schoetensacki . The metacarpal indicates a particularly slender individual; however, given the high phenotypical plasticity of these animals, the Bison material from Cesi is confidently referred here to B. schoetensacki (Fig. 4a–e). J. Quaternary Sci., Vol. 12(6) 507–518 (1997) The occurrence of Homotherium, the only carnivore hitherto found at Cesi, is documented by a right second upper incisor, but this find is weak evidence for any specific identification. The elephant is represented by a large fragment of ox coxae (ileum and part of the acetabulum), a patella and probably a semilunar. The poor state of preservation of the specimen prevents any determination. The presence of a typical Stephanorhinus hundsheimensis, Bison schoetensacki and Dama clactoniana suggests a more recent age than that of the Colle Curti community. The assemblage is of limited diversity and the specimens are fragmentary and more poorly preserved than the material from Colle Curti. The faunal composition is similar to that from Isernia (Sala, 1983) and here it is referred to an early– middle part of the Middle Pleistocene (according to the location of the Early–Middle Pleistocene boundary, at about Oxygen Isotope Stage 25, cf. Cita and Castradori, 1994). By and large, the assemblage is suggestive of open woodlands characterised by limited areas covered by trees. Moreover, the presence of Hippopotamus and Dama indicates humid and not markedly cold climatic conditions. Palynology The palynological content of the 12 m sequence underlying the Cesi fossiliferous bed taken as reference (0 m; Fig. 3) has been studied (A. Bertini, in progress). These sediments are represented mainly by clays, in which Fe–Mn and CaCO 3 concretions occur frequently, and locally also by thin layers containing calcareous and siliceous gravels. The sedimentological analysis and the predominance of palynologically barren sediments suggest considerable weathering, both in the biostratinomic and in the diagenetic phases. The pollen diagram has been reconstructed on the basis of the data from 12 samples (Fig. 6). In these samples, the pollen concentration is generally high, with values ranging from 6955 to 149 825 grains per gramme of sediment. The palynological data from Cesi have been reported in association with those from Colle Curti (Fig. 6), cited in the literature (Coltorti et al ., in press), due to the geographical and stratigraphical proximity of these two deposits. The Cesi pollen diagram is characterised by the dominance of herbaceous elements, especially represented by Poaceae and Asteraceae. Among Asteraceae, Artemisia is very abundant, sometimes in association with Ephedra, another steppe element. Ranunculaceae, Caryophyllaceae, Plantaginaceae and Polygonaceae pollen is also well represented. Arboreal plants are represented mainly by Pinus pollen, which can reach frequencies of 73.65% TLP. Abies and Picea are less abundant although always present, whereas Tsuga and Cedrus are sporadic. Deciduous broadleaf trees, never more than the 6% TLP, are represented mainly by Quercus, and sometimes also by Acer and Carpinus. Mediterranean xerophytes are rare. Zygnemataceae and Chlorophyceae algae and some local herbaceous plants provide indications of the history of the depositional environment. In particular, the constant presence of Pediastrum, in association with Botryococcus, Spirogyra and Mougeotia, suggests a freshwater environment, sometimes under mesotrophic conditions. The Cesi pollen assemblages suggest the occurrence of a landscape dominated by open vegetation characterised by taxa (Poaceae, Asteraceaeae, including Artemisia and  1997 by John Wiley & Sons, Ltd. EARLY MIDDLE PLEISTOCENE SITE IN THE APENNINE 513 Figure 4 (a) Bison shoetensacki . Left metacarpus, MSNC 255; dorsal view (×1/2). (b) Bison shoetensacki . Right mandibular ramus, MSNC 257, occlusal view (×1/2). (c) Bison shoetensacki . Right mandibular ramus, MSNC 257; lateral view (×1/2). (d) Bison shoetensacki . Right M/3, MSNC 257 (×1). (e) Bison shoetensacki . Right M/3, MSNC 206 (×1). (f) Stephanorhinus hundsheimensis. Left mandibular ramus, MSNC 290a; medial view (×1/4). (g) Dama clactoniana. Distal end of right humerus, MSNC 264; cranial view (×1).  1997 by John Wiley & Sons, Ltd. J. Quaternary Sci., Vol. 12(6) 507–518 (1997) 514 Figure 5 JOURNAL OF QUATERNARY SCIENCE Megaceroides solilhacus. Fragment of antler, MSNC 287, and reconstruction of the same antler. Ephedra, Chenopodiaceae, etc.) that were typical of the glacial phases after 2.6 Ma in the Mediterranean region (Suc et al ., 1995). The markedly arid climate prevented the growth of plants demanding year-long humid conditions. Broad-leaf trees, represented mainly by Quercus pollen grains occurred but always at low frequencies. Woodland was dominated by Pinus. The phases characterised by the expansion of Pinus are probably linked with slight increases of the temperature (wooded steppe). No new interglacial phases have been recorded. The present-day arboreal vegetation in the Colle Curti and Cesi areas is mainly a deciduous thermophilous forest of Quercus pubescens and Ostrya carpinifolia series in the hill belt, and woodland with Fagus sylvatica in the mountain belt (Orsomando, 1993). At Colle Curti, steppe associations (Asteraceae, Cyperaceae, Poaceae) are replaced by a mesophilous forest, characterised principally by Tsuga and Cedrus. These elements, which are more typical of the end of an interglacial, indicate the occurrence of more humid conditions than at Cesi, where the main arboreal taxon is Pinus. In view of the fact that the two sequences jointly span over 0.4 myr, the absence of interglacial phases both at Cesi and Colle Curti suggests the occurrence of several hiatuses. fairly unfavourable sedimentary features, mainly inclusions of siliceous clasts and oxidation of the clayey matrix. The same problem was also encountered in the nearby section at Colle Curti (Coltorti et al ., in press). The polarity pattern is established on the basis of the VGP (virtual geomagnetic pole) latitude values calculated from the characteristic magnetisations, as shown in Fig. 8. From the sample immediately below the fossiliferous bed and down to that at 6.5 m, the magnetisation has a single behavioural component (Fig. 7, sample at 3.10 m). From 7 m downwards, a clearly reversed polarity occurs. The sample at 8.0 m (Fig. 7) shows, under the thermal treatment up to 200°C, the removal of an initially superimposed normal field, which produces an increase of intensity with strongly deviated directions. Thereafter, a regular decrease of the primary magnetisation displays a characteristic vector. In the lowest part of the trench, the signal deteriorates, although the lowermost samples seem still to be in the reversed polarity zone. The distribution of the polarities in this section, together with the stratigraphical and biochronological evidence of the local fauna in the Cesi sequence, suggest that the polarity transition at 7 m corresponds to the Brunhes–Matuyama (B– M) boundary. M agnetostratigraphy D iscussion and conclusions About 11 cm of the sequence was sampled from a trench excavated for palaeomagnetic purposes beneath the fossiliferous bed at Cesi. Thermal demagnetisation was used for assessing the characteristic vector of the primary remanence. Figure 7 shows the demagnetisation pattern for two samples of normal and reversed polarity, respectively: the intensity decreases regularly, but the directions have two distinct orientations. This suggests that in the upper part of the section there is a very well-defined normal polarity, whereas in the lower part the magnetisations seem poorly retained; the material in fact appears to be affected by J. Quaternary Sci., Vol. 12(6) 507–518 (1997) The results of the multidisciplinary investigations on the Cesi sequence suggest that the fossiliferous bed is of early–middle Middle Pleistocene age and that the faunal assemblage can be referred to the Middle–Late Galerian. The Cesi sequence is thought to have accumulated in a lacustrine environment, which, in its upper part, was succeeded by clastic sediments from an alluvial fan prograding into the lake. On the basis of geomorphological evidence, the lacustrine basin was formed by the extensional tectonics that affected the Umbro-Marchean Apennines during the  1997 by John Wiley & Sons, Ltd. EARLY MIDDLE PLEISTOCENE SITE IN THE APENNINE Figure 6 Palynological diagrams from Cesi and Colle Curti. Colle Curti palynological and palaeomagnetic determinations are from Coltorti et al (in press). AP: arboreal plant; in the curves of other AP are included Buxus, Salix, indeterminates and indeterminable pollen grains. NAP: non-arboreal plants. Hydrophytes curve includes: Alismataceae, Potamogeton, Sparganium, Typha latifolia, Myriophyllum and Epilobium. The curve of other NAP includes Apiaceae, Brassicaceae, Convolvulaceae, Dipsacaceae, Euphorbiaceae, Fabaceae, Geraniaceae, Lamiaceae, Liliaceae, Linum, Plumbaginaceae. Bone symbol indicates fossiliferous beds. 515 516 JOURNAL OF QUATERNARY SCIENCE Figure 7 Demagnetisation and Zijderveld diagrams for samples of normal and reversed polarity (samples at 3.1 m and 8.0 m, respectively). On the left, the diagrams relative to sample at 3.10 m show a regular decrease of a single normal component of the characteristic magnetisation. On the right, the diagrams relative to sample at 8.0 m display a more complex behaviour with the removal of a secondary component until 200°C and the consequent enhancement of the characteristic reversed magnetisation. Early–Middle Pleistocene. The fauna from Cesi is more advanced than that found in the adjacent Colle Curti basin, as demonstrated by the stage of evolution of most of its components. Likewise, in the pollen flora, the sporadic occurrence of Tsuga, as well as the absence of other elements such as Carya and Liquidambar, confirm a more recent age for the Cesi sequence. The fossil assemblage was obtained from a positively magnetized section of the local sequence, which is thus attributed to the Brunhes Chron; on the other hand, the Colle Curti fauna was preserved very close to the base of the normal polarity magnetized sediments referred to the Jaramillo Subchrone (C1r1n) (Coltorti et al ., in press). The fact that the Cesi fossiliferous bed lies 7 m above the B–M boundary (780 ka) suggests an age close to the latter. The analyses carried out on the Cesi sequence below the fossil-bearing bed indicate the dominance of elements typical of open vegetation, sometimes interrupted by a temporary spread of pine forest. On the whole, markedly arid conditions and low temperatures typical of a glacial phase seem to have occurred. The fauna from Cesi, obtained from the uppermost part of the sequence, suggest open woodlands, but not under particularly severe climatic conditions, and therefore most probably indicate interstadial conditions. The palynological record from both sites indicates a landscape strongly dominated by an open vegetation represented principally by Asteraceae, Cyperaceae and Poaceae pollen grains. The repeated episodes characterised by such a vegetation typical of markedly arid conditions apparently represent glacial phases. No significant expansions of thermophilous trees, with the exception of temporary spreads of a pine forest at Cesi and of a mesophilous forest with Tsuga and Cedrus at Colle Curti, have been recorded. The absence of new interglacial phases is the striking feature of both the pollen diagrams. J. Quaternary Sci., Vol. 12(6) 507–518 (1997) Previous hypotheses (Coltorti et al ., in press) put forward to explain this peculiarity at Colle Curti, such as the strong influence of geographical factors (e.g. altitude), do not seem satisfactory, particularly in light of the new palynological and palaeomagnetic evidence from Cesi. In fact the pollen record obtained from the nearby extant Colfiorito swamp (Brugiapaglia and de Beaulieu, 1995) reveal a similar picture. It is interesting to note that in this sequence an unconformity/non-depositional surface separates the Upper Pleistocene sediments from those deposited during the Holocene, following the first evidence of human deforestation in the area. In the Early Holocene, as in the previous interglacial, much of the fluvial activity was diverted underground along buried karstic features and fractures. The deforestation activated erosional processes on the slopes, artificially generating much drier conditions, which may have been comparable to those during the glacial phases. During the first phases of slope erosion, the soil profiles were eroded as witnessed by a thin stone-line. Only when the ongoing of slope degradation was able to fill the seepage lines was there again the establishment of swamp conditions in the valley bottoms. In the Colle Curti and Cesi sedimentological records, as well as in the Colfiorito swamp, therefore, the interglacial hiatuses tend to be represented by stone-lines. Acknowledgements The research was financed by Consiglio ` Nazionale delle Ricerche and Ministero dell’Universita e della Ricerca Scientifica e Tecnica 60% grants. Special thanks are due to A. Blasetti, V. Borselli, F. Cozzini, F. Landucci and M. Mazzini for their contribution in the excavation and restoration of the fossil material. We would also thank S. Cucchiari and F. Salvadori for their contribution in the excavation of the site. We are indebted to Dr P. Gibbard for the revision of the English.  1997 by John Wiley & Sons, Ltd. EARLY MIDDLE PLEISTOCENE SITE IN THE APENNINE Figure 8 Latitude of VGP (Virtual Geomagnetic Pole) plotted against the stratigraphical position of the samples collected in the Cesi section, and the relative interpreted polarity. References AMBROSETTI, P., CARBONI, M. G., CONTI, M. A., COSTANTINI, A., ESU, D., GANDIN, A., GIROTTI, O., LAZZAROTTO, A., MAZZANTI, R., NICOSIA, U., PARISI, G. and SANDRELLI, F. 1978. Evoluzione paleogeografica e tettonica nei bacini toscoumbro-laziali nel Pliocene e nel Pleistocene inferiore. Memorie della Societa Geologica Italiana, 19 , 573–580. AMBROSETTI, P., CARRARO, F., DEIANA, G. and DRAMIS, F. 1982. Il sollevamento dell’Italia centrale tra il Pleistocene inferiore e il Pleistocene medio. Contributo conclusivo per la realizzazione della Carta Neotettonica d’Italia (II), Consiglio Nazionale delle Ricerche Progetto Finalizzato ‘Geodinamica’, Sotto-Progetto ‘Neotettonica’ , 356 , 1341–1343. AMBROSETTI, P., CARBONI, M. G., CONTI, M. A., ESU, D., GANDIN, A., GIROTTI, O., LAMONICA, G., LANDINI, B. and PARISI, G. 1987. Il Pliocene ed il Pleistocene inferiore del Bacino del fiume Tevere nell’Umbria meridionale. Geografia Fisica e Dinamica Quaternaria, 10 (1), 10–33. BALLY, A. W., BURBI, L., COOPER, C. and GHELARDONI, P. 1986. Balanced sections and seismic reflection profiles across the Central Apennines. Memoria della Societa Geologica Italiana, 35 , 257–310. BERTINI, A. in progress. Pollen record from Colle Curti and Cesi: two early-middle Pleistocene mammal sites in the Umbro-Marchean Apennine (Central Italy). BOCCALETTI, M., CALAMITA, F., CENTAMORE, E., DEIANA, G. and DRAMIS, F. 1983. The Umbria-Marche Apennine: an example of thrusts and wrenching tectonics in a model of ensialic Neo 1997 by John Wiley & Sons, Ltd. 517 genic–Quaternary deformation. Bolletino della Societa Geologica Italiana, 102 , 582–592. BOCCALETTI, M., CALAMITA, F., CENTAMORE, E., CHIOCCHINI, U., DEIANA, G., MORETTI, G., MICARELLI, A. and POTETTI, M. 1986. Evoluzione dell’Appennino Tosco-Umbro-Marchigiano durante il Neogene. Giornale di Geologia, Bologna, 48 (1–2), 227–233. BORSELLI, V., FICCARELLI, G., LANDUCCI, F., MAGNATTI, M., NAPOLEONE, G. and PAMBIANCHI, G. 1988. Segnalazione di mammiferi pleistocenici nell’area di Colfiorito (Appennino umbro` marchigiano) e valutazione della potenzialita del giacimento con ` metodi geofisica. Bollettino della Societa Paleontologica Italiana, 27 (2), 253–257. ´ BRUGIAPAGLIA, E. and DE BEAULIEU, J. L. 1995. Etude de la ´ ´ ´ dynamique vegetale Tardiglaciaire et Holocene en Italie centrale: le marais de Colfiorito (Ombrie). Comptes Rencches Hebclomad´ aires de Seavices de l’Academie des Sciences, Paris, 321 (serie IIa), 617–622. CALAMITA, F. 1990. Thrust and fold-related structures in the Umbria-marche Apennines (Central Italy). Annales Tectonicae, IV (1), 83–117. CALAMITA, F. and DEIANA, G. 1988. The arcuate shape of the Umbria-Marche-Sabina Apennines (Central Italy). Tectonophysics, 146 , 139–147. CALAMITA, F. and PIZZI, A. 1992. Tettonica quaternaria nella dorsale appenninica umbro-marchigiana e bacini intrappenninici associati. Studi Geologici Camerti, Volume Speciale, 1 , 17–25. CALAMITA, F., COLTORTI, M., DEIANA, G., DRAMIS, F. and PAMBIANCHI, G. 1982. Neotectonic evolution and geomorphology of the Cascia and Norcia depressions (Umbria-Marche Apennine). Geografia Fisica e Dinamica Quaternaria, 5 , 263–276. CALAMITA, F., CELLO, G., CENTAMORE, E., DEIANA, G., MICARELLI, A., PALTRINIERI, W. and RIDOLFI, M. 1991. Stile deformativo e cronologia della deformazione lungo tres ezioni bilanciate dall’Appennino Umbro-Marchigiano alla costa Adriatica. Studi Geologici Camerti, Volume Speciale, 1 , 295–314. CALAMITA, F., COLTORTI, M., FARABOLLINI, P. and PIZZI, A. 1994. Le faglie normali quaternarie nella dorsale appenninica umbro-marchigiana: proposta di un modello di tettonica di inversione. Studi Geologici Camerti, Volume Speciale, CROP, 18 , 211–225. CALAMITA, F., COLTORTI, M., PIERUCCINI, P. and PIZZI, A. 1995. Evoluzione geomorfologica e strutturale plio-quaternaria dell’Appennino Umbro-Marchigiano tra il pre-Appennino umbro e la costa adriatica. Convegno ‘Geodinamica e tettonica attiva del sistema Tirreno-Appennino’, Camerino 9–10 February, Abstracts, 51–54. CANTALAMESSA, G., CENTAMORE, E., CHIOCCHINI, U., COLALONGO, M. L., MICARELLI, A., NANNI, T., PASINI, G., POTETTI, M., RICCI-LUCCHI, F., CRISTALLINI, C. and DI LORITO, L. 1986. Il Plio-Pleistocene delle Marche. IN : Centamore, E. and Deiana, G. (eds), Studi Geologici Camerti, Volume Speciale, 61– 82. La Geologia delle Marche. CATTUTO, C., CENCETTI, C. and GREGORI, L. 1992. Il PlioPleistocene nell’area medio-alta del bacino del fiume Tevere: possibile modello morfotettonico. Studi Geologici Camerti, Volume Speciale, 1 , 103–109. CICCACI, S., D’ALESSANDRO, L., DRAMIS, F., FREDI, P. and PAMBIANCHI, G. 1985. Geomorphological and neotectonic evolution of the Umbria-Marche ridge, northern sector. Studi Geologici Camerti, 10 , 7–15. CITA, M. B. and CASTRADORI, D. 1994. Workshop on marine sections from the Gulf of Taranto (Southern Italy) usable as potential stratotypes for the GSSP of the Lower, Middle and Upper Pleistocene. Il Quaternario, 7 (2), 677–692. COLTORTI, M. 1981. Lo stato attuale delle conoscenze sul Pleistocene ed il Paleolitico inferiore e medio della regione marchigiana. Atti I Convegno Beni Culturali-Ambientali delle Marche, Numana, 8–10 May, 63–122. COLTORTI, M. and DRAMIS, F. 1988. The significance of stratified slope-waste deposit in the Quaternary of Umbria-Marche Apen¨ nines, Central Italy. Zeitschrift fur Geomorphologie, Neue Folge Supplemental Band, 71 , 59–70. J. Quaternary Sci., Vol. 12(6) 507–518 (1997) 518 JOURNAL OF QUATERNARY SCIENCE COLTORTI, M. and FARABOLLINI, P. 1995. Quaternary evolution of the Castelluccio di Norcia basin (Umbria-Marche Apennines, Italy). Il Quaternario, 8 (1), 149–166. COLTORTI, M. and PIERUCCINI, P. in press. The plio-pleistocene sedimentary evolution of the southern east Tiber basin (Spoleto, central Italy). Il Quaternario. COLTORTI, M., CONSOLI, M., DRAMIS, F., GENTILI, B. and PAMBIANCHI, G. 1991. Evoluzione geomorfologica delle piane alluvionali delle Marche centro-meridionali. Geografia Fisica e Dinamica Quaternaria, 14 (1), 73–86. COLTORTI, M., ALBIANELLI, A., BERTINI, A., FICCARELLI, G., LAURENZI, M. A., NAPOLEONE, G. and TORRE, D. in press. The Colle Curti mammal site in the Colfiorito area (UmbriaMarchean Apennine, Italy): geomorphology, stratigraphy, paleomagnetism and palynology. Quaternary International . ` DAMIANI, A. V., MINELLI, G. and PIALLI, G. 1991. L’Unita Falterona–Trasimeno nell’area compresa tra la Val di Chiana e la Val Tiberina: sezione Terontola–Abbazia di Cassiano. Studi Geologici Camerti, Volume Speciale, CROP (Crosta Profonda), 03 , 235–41. DESPLANQUES, H. 1969. Champagne Ombriennes. Centre National de la Recherche Scientifique de Paris, Paris, 544 pp. DRAMIS, F. 1992. Il ruolo dei sollevamenti tettonici a largo raggio nella genesi del rilievo appenninico. Studi Geologici Camerti, Volume Speciale, 1992 (1), 9–16. DRAMIS, F., PAMBIANCHI, G., NESCI, O. and CONSOLI, M. 1991. Il ruolo di elementi strutturali trasversali nell’evoluzione tettonicosedimentria e geomorfologica della regione marchigiana. Studi Geologici Camerti, Volume Speciale, 1991 (2), 287–293. FICCARELLI, G. and MAZZA, P. 1990. New fossil findings from the J. Quaternary Sci., Vol. 12(6) 507–518 (1997) Colfiorito basin (Umbria-Marchean Apennine). Bollettino della ` Societa Paleontologica Italiana, 29 (2), 245–247. FICCARRELLI, G. and SILVESTRINI, M. 1991. Biochronologic remarks on the Local Fauna of Colle Curti (Colfiorito Basin, Umbrian-Marchean Apennine, Central Italy). Bolletino della Soci` eta Paleontologica Italiana, 30 (2), 197–200. FICCARELLI, G., MAGNATTI, M. and MAZZA, P. 1990. Occurrence of Microtus (Allophaiomys) gr. pliocaenicus in the Pleistocene lacustrine basin of Colfiorito (Umbria-Marchean Apennine, Central ` Italy). Bolletino della Societa Paleontologica Italiana, 29 (1), 89–90. LAVECCHIA, G., MINELLI, G. and PIALLI, G. 1988. The UmbriaMarche arcuate fold belt (Italy). Tectonophysics, 146 , 125–137. MAZZA, P. 1993. Ethological inferences on Pleistocene rhinoceroses of Europe. Rendiconti Scienze Fisiche e Naturali dell’Accademia dei Lincei , series 9, 6 (2), 127–137. ORSOMANDO, E. 1993. Carte della vegetazione dei fogli Passignano sul Trasimeno e Foligno. Braun-Blanquetia, 10 , 46 pp. RAFFY, J. 1981. Orogenese et dislocations quaternaires du versant tyrrhenien des Abruzzes (Italie centrale). Revue de Geologie Dynamique et de Geographie Physique, 23 (1), 55–72. SALA, B. 1983. La fauna del giacimento di Isernia La Pineta. Nota preliminare. IN : Calderini, (ed.) Isernia La Pineta, un accampa` mento piu antico di 700.000 anni , Catalogo della mostra, 71– 79, Bologna. SUC, J. P., BERTINI, A., COMBOURIEU-NEBOUT, N., DINIZ, F., LEROY, S., RUSSO-ERMOLLI, E., ZHENG, Z., BESSAIS, E. and FERRIER, J. 1995. Structure of the West Mediterranean vegetation and climate since 5.3 Ma. Acta Zoologica Cracoviensia, 38 (1), 3–16.  1997 by John Wiley & Sons, Ltd.