ImpaginatoDavanzo MMAAMMMMAALL SSKKEELLEETTAALL RREEMMAAIINNSS FFRROOMM TTHHEE FFUUMMAANNEE CCAAVVEE ((VVEERROONNAA,, NNOORRTTHHEERRNN IITTAALLYY)):: AANN OOXXYYGGEENN IISSOOTTOOPPEE SSTTUUDDYY AANNDD IITTSS PPAALLAAEEOOCCLLIIMMAATTOOLLOOGGIICCAALL IIMMPPLLIICCAATTIIOONNSS SS.. DDaavvaannzzoo,, PP.. IIaaccuummiinn && AA.. LLoonnggiinneellllii University of Parma, Department of Earth Sciences, Parco Area delle Scienze 157/A, 43100 Parma, Italy ABSTRACT A stable isotope study of mammal skeletal remains of Palaeolithic age from the Fumane cave (Verona, N. Italy) was carried out to obtain palaeoclimatological information. The studied samples belong to Capra ibex, Cervus elaphus, Capreolus capreolus and Bos/Bison sp. and come from various levels of the cave deposits of Upper and Mid Palaeolithic age. The δ18O of palaeoenvironmental water was calculated from the δ18Op measured on fossil samples, according to the isotope equations previously calibrated on modern specimens. The results obtained suggest that teeth must be considered with great caution for palaeoclimatic reconstruction: in the case of this study they yield quite unreliable results. In a general way, isotopically well-preserved bones are preferrable since the interpretation of their results is more straightforward, more reliable and more meaningful, the bone phosphate representing a mean value of a conside- rable period of the life of each specimen. The δ18Ow calculated from bone samples indicate that the climatic conditions were colder than at present. Further isotope data are of importance as a database for detailed reconstructions of the past climatic conditions along NS and EW geographic sections in Europe. RIASSUNTO Sono state eseguite misure della composizione isotopica del fosfato (δ18Op) su resti di mammiferi fossili provenienti dalla Grotta di Fumane (Verona) allo scopo di ottenere informazioni paleoclimatiche quantitative o almeno semiquantitative per il Paleolitico Medio- Superiore nella zona in esame. I campioni provengono da vari livelli di una successione stratigrafica divisa in quattro unità principali (S, BR, A e D) ed appartengono alle specie Capra ibex, Cervus elaphus, Capreolus capreolus e Bos/Bison sp.. I valori di δ18O delle “paleo” acque meteoriche sono stati calcolati a partire dai valori di δ18Op misurati sui campioni fossili utilizzando le equazioni isotopiche preven- tivamente calibrate su individui recenti. I risultati ottenuti indicano che i denti non sono sempre utilizzabili per ricostruzioni paleoclimati- che e che, in generale, è preferibile utilizzare le ossa, almeno quando si ha a che fare con materiali non interessati da processi diage- netici. I valori di δ18Ow (acqua paleoambientale) calcolati dai campioni di ossa confermano che il clima, nell’intervallo di tempo relativo alla successione esaminata, era sensibilmente più freddo di quello attuale. Questi risultati costituiscono un utile apporto per dettagliate ricostruzioni delle variazioni climatiche nel passato lungo sezioni NS ed EO in Europa. Keywords: Fumane cave, oxygen isotopes, palaeoclimatology, Upper and Mid-Palaeolithic Parole chiave : Grotta di Fumane, composizione isotopica dell’ossigeno, paleoclimatologia, Paleolitico Medio-Superiore Il Quaternario Italian Journal of Quaternary Sciences 1155(1), 2002, 39-44 11.. IINNTTRROODDUUCCTTIIOONN The Fumane Cave (Verona-Northern Italy) has been studied by archaeologists for its deposits from Palaeolithic age. The cave is located at about 350 m.a.s.l. on the Lessini Mountains (Fig.1). It was repea- tedly excavated throughout the last century and an extensive excavation (10 m of sediments) was carried out by the Historical Museum of Verona and the Sovraintendenza alle Antichità from 1982 to 1983 and by the Prehistoric and Ethnographic Museum “Pigorini” of Rome from 1987 to recent time (Bartolomei et al., 1992a). The 10 m of sediments were divided into four units (D, A, BR and S), each one in turn being subdivided into various levels and sublevels. 29 radiocarbon dates are available for 9 of the stratigraphic levels of the D and A units of the Fumane cave. The D unit dates approxima- tely from 27,000 to 32,000 BP and the A unit from 32,000 to 38,000 BP. (for a complete review see Bartolomei et al., 1992a,b; Cassoli and Tagliacozzo, 1994; Cremaschi et al., 1986; Peresani and Sartorelli, 1996). Three thermoluminescence dates were carried out on burnt flint stones from levels BR11, BR12 and S7 yielding ages of about 55, 57 and 80 ka BP respectively (Broglio, personal communication). Deer and ibex are the most abundant faunal remains at Fumane Cave followed by roe deer, chamois and bovines with subordinated marmot and fox. These remains are essentially human food refuse. Micromammals are also well represented. The study of mammal assemblages and sedimentary analyses from the A and D units point to a wooded landscape and a cool climate at lower levels (A13-A3) and to an open landscape with a cold and arid climate at higher levels (A2-D1) (Bartolomei et al., 1992b). 40 Davanzo-Iacumin-Longinellii While the analyses of faunal assemblages from archaeological sites can yield only qualitative palaeocli- matological information, oxygen isotope analyses can provide a quantitative or semi-quantitative information of past climatic conditions in the case of suitable material. Mammal bone and tooth oxygen isotope data enable the calculation of the mean oxygen isotope composition of environmental water throughout the period covered by the samples studied. In fact, the δ18Op values can be converted into δ18O of palaeoenvironmental water by means of isotope equations calibrated on living speci- mens of the same species from different climatic areas (e.g. Longinelli, 1984; Longinelli, 1995; Delgado Huertas et al., 1995). These equations take into consideration the different metabolic processes characteristic of each species and the overall isotope fractionation of the oxy- gen taken in. In turn, the mean δ18O value of environ- mental water is related quantitatively to the yearly mean temperature at the ground. Consequently, the δ18Op values can be considered variables quantitatively rela- ted to palaeoclimatological conditions. 22.. MMAATTEERRIIAALLSS AANNDD MMEETTHHOODDSS The study was carried out on 119 mammal bone and tooth enamel samples belonging to Cervus elaphus, Capra ibex, Capreolus capreolus and Bos/Bison sp. In the case of teeth, the enamel sample portion analysed was collected in such a way as to represent the whole period of tooth accretion. Consequently, the intra-tooth variation related to seasonal climatic changes is consi- derably reduced. The samples were prepared according to well established procedures (Crowson and Showers, 1991 and Lécuyer et al., 1993). The final sample of the che- mical treatment is Ag3PO4.The phosphate oxygen is extracted from the samples by reacting the Ag3PO4 with BrF5 at 600°C for about 15 hours. The oxygen is then purified and converted to CO2 by cycling over hot graphite in the presence of a platinum catalyst; the CO2 is measured for its 18O/16O ratios by means of a mass spectrometer (in this case a Finnigan Delta S). The iso- tope results are reported in the δ terminology versus the international V-SMOW standard. 33.. IISSOOTTOOPPEE RREESSUULLTTSS AANNDD DDIISSCCUUSSSSIIOONN The phosphate oxygen isotope composition (δ18Op) of the measured samples and the oxygen isoto- pe composition of local environmental water (δ18Ow) are reported in Table 1 along with their stratigraphic position in the cave sedimentary sequence. The δ18Ow was cal- culated from the δ18Op values according to the equa- tions suggested by D’Angela and Longinelli (1990) for Cervus and Bos and by Delgado Huertas et al. (1995) for Capra and Capreolus. The same δ18Ow values are also reported graphi- cally in Figure 2. The δ18Ow calculated from bone sam- ples range from -10.2 to -6.2 per mil with a mean value of -7.8±0.8. The mean δ18Ow calculated from tooth sam- ples (-9.6±2.3) is surprisingly lighter than that of bone samples ranging between -13.4 and -3.9 per mil. In general, tooth enamel is considered the most reliable material for δ18Op analyses due to its capability of retai- ning the pristine isotope value throughout long periods (e.g. Ayliffe et al., 1994). However, tooth phosphate is not renewed during the lifetime of an individual and, consequently, it records the isotopic signature of clima- tic conditions referring to a very short period of the indi- vidual’s life (Bryant et al., 1996; Fricke and O’Neil, 1996). This is particularly true for the teeth that minerali- se during the weaning period which are affected by the nursing processes. In the case of our samples and particularly for the C. ibex from which a number of teeth were measured in the lower units (BR and S) the δ18Ow values calculated from teeth δ18Op are systematically 18O-depleted when compared to δ18Ow calculated from bone values, even from the same level (Fig.2). Only very few teeth values behave in the opposite way being 18O-enriched. Since diagenetic processes normally affect preferentially bones leading to an 18O-depletion, this effect can hardly be ascribed to post-depositional processes. Therefore, according to the results obtained the time period repre- sented by the BR and S units could be considered sli- ghtly colder than the period represented by the D and A units. However, this hypothesis is not confirmed by the measurements obtained from other species and, parti- cularly, by the deer samples, several of which were analysed from the BR and S levels. It may be inferred that: 1) the ibex fawns lived at a higher elevation; 2) they lived in an area where streams carried down isoto- pically lighter water from the nearby alpine glaciers; 3) that they were born in a given period so that the perma- nent teeth mineralised preferentially during the cold sea- son. The 18O-enriched teeth values may be ascribed to the weaning period since milk water is considerably 18O- enriched when compared to environmental water (Luz et al., 1984). Because of the great difference between bone and Fig.1 - Location of the Fumane Cave, North-Eastern Italy. Ubicazione della Grotta di Fumane. 41Mammal skeletal remains ... lleevveell ssppeecciimmeenn sskkeelleettaall δδ 1188OO pp δδ 1188OO ww rreemmaaiinnss D1e C. ibex bone 17.2 -7.9 D1e C. ibex bone 17.7 -7.3 D1e+D1d C. ibex bone 16.8 -8.4 D1e+D1d C. ibex tooth 12.3 -13.4 D1d C. ibex bone 17.7 -7.3 D1d C. ibex bone 17.5 -7.5 D1c C. ibex bone 16.9 -8.2 D1c C. ibex bone 16.9 -8.2 D1c C. ibex bone 18.2 -6.7 D1a C. ibex bone 15.1 -10.2 D1a C. ibex tooth 20.7 -3.9 D3b C. ibex bone 17.4 -7.6 D3b C. ibex bone 17.9 -7.0 D3b C. ibex tooth 19.4 -5.4 D5 C. ibex bone 17.2 -7.8 D5 C. ibex bone 18.4 -6.5 D6 C. ibex bone 18.2 -6.7 D6 C. ibex tooth 15.2 -10.1 D6 C. ibex tooth 15.9 -9.3 D6 Bos/Bison sp. bone 16.7 -8.2 D6 Bos/Bison sp. bone 15.7 -9.1 D6 Bos/Bison sp. bone 17.9 -6.9 D6 Bos/Bison sp. bone 15.5 -9.3 A1 C. ibex bone 17.6 -7.4 A1 C. ibex bone 17.6 -7.4 A1 C. ibex bone 17.6 -7.4 A1 Bos/Bison sp. bone 17.3 -7.6 A1 Bos/Bison sp. tooth 12.8 -12.0 A2 C. ibex bone 17.7 -7.3 A2 C. ibex tooth 16.7 -8.4 A2 C. ibex tooth 15.2 -10.2 A3 C. ibex bone 17.0 -8.1 A3 C. ibex tooth 17.8 -7.2 A3 Bos/Bison sp. tooth 15.1 -9.7 A4 C. elaphus bone 16.5 -8.1 A5+6 C. ibex bone 16.5 -8.6 A5+6 C. ibex tooth 17.4 -7.7 A5+6 C. elaphus bone 16.3 -8.2 A5+6 C. elaphus bone 16.9 -7.7 A6 C. ibex bone 16.8 -8.3 A6 C. capreolus bone 17.9 -7.1 A6 C. capreolus bone 17.7 -7.3 A7 Bos/Bison sp. bone 15.0 -9.8 A10 C. ibex bone 18.4 -6.5 A10 Bos/Bison sp. tooth 14.2 -10.6 A11 Bos/Bison sp. tooth 15.1 -9.7 A11 Bos/Bison sp. tooth 16.2 -8.7 BR1 C. ibex bone 17.6 -7.4 BR1 C. elaphus bone 17.7 -7.0 BR3 C. ibex bone 16.8 -8.3 BR4 C. ibex bone 17.4 -7.7 BR5 C. ibex bone 17.0 -8.1 BR5 C. elaphus bone 16.8 -7.8 BR5 C. elaphus bone 17.5 -7.2 BR6 C. elaphus bone 17.1 -7.5 BR6 C. capreolus bone 16.7 -8.4 BR7b C. elaphus bone 17.4 -7.3 BR7b C. elaphus bone 17.2 -7.4 BR7b C. elaphus bone 16.1 -8.4 lleevveell ssppeecciimmeenn sskkeelleettaall δδ 1188OO pp δδ 1188OO ww rreemmaaiinnss BR7c C. elaphus bone 17.4 -7.3 BR7f C. elaphus bone 17.6 -7.0 BR7f C. elaphus bone 17.7 -6.9 BR7f C. capreolus bone 16.7 -9.2 BR7f C. capreolus bone 15.9 -9.3 BR8 C. ibex bone 17.2 -7.8 BR8 C. elaphus bone 17.3 -7.3 BR9 C. ibex bone 16.9 -8.2 BR9 C. elaphus bone 18.5 -6.2 BR9 C. elaphus bone 17.1 -7.5 BR9 C. elaphus tooth 16.5 -8.1 BR9 C. capreolus bone 16.6 -8.5 BR10 C. ibex bone 17.5 -7.5 BR10 C. ibex tooth 12.4 -13.4 BR10 C. ibex tooth 18.0 -7.0 BR11d C. ibex tooth 15.9 -9.3 BR11d C. elaphus bone 17.0 -7.7 BR11d C. elaphus bone 17.6 -7.0 BR11d C. elaphus bone 15.9 -8.6 BR11d C. capreolus bone 17.3 -7.8 BR11d C. capreolus bone 17.7 -7.3 BR12 C. elaphus bone 16.6 -8.0 BR12 C. elaphus bone 17.8 -6.9 BR12 C. elaphus bone 17.4 -7.3 BR12 C. elaphus tooth 16.1 -9.1 BR12 C. capreolus tooth 14.2 -11.3 BR12 C. capreolus bone 15.7 -9.6 S2 C. elaphus bone 16.5 -8.0 S2 C. elaphus bone 17.1 -7.5 S2 C. elaphus bone 17.0 -7.6 S2 C. ibex tooth 14.9 -10.5 S2 C. ibex tooth 14.8 -10.6 S2 C. capreolus bone 17.5 -7.6 S2 C. capreolus bone 17.9 -7.0 S2 C. capreolus bone 17.6 -7.4 S3 C. elaphus bone 16.8 -7.7 S3 C. elaphus bone 16.9 -7.7 S3 C. elaphus bone 16.1 -8.4 S3 C. elaphus bone 17.3 -7.3 S3 C. ibex tooth 14.8 -10.6 S3 C. ibex tooth 14.3 -11.1 S4 C. elaphus bone 17.0 -7.6 S4 C. elaphus bone 16.3 -8.2 S4 C. elaphus bone 16.8 -7.7 S4 C. ibex tooth 12.5 -13.2 S4 C. capreolus bone 18.4 -6.5 S5 C. elaphus bone 15.3 -9.1 S5 C. elaphus bone 16.3 -8.2 S5 C. ibex tooth 14.8 -10.6 S6 C. elaphus bone 16.0 -8.5 S6 C. elaphus bone 17.4 -7.2 S6 C. elaphus bone 17.0 -7.6 S7 C. elaphus bone 17.2 -7.4 S8 C. elaphus bone 16.5 -8.0 S8 C. elaphus bone 16.8 -7.7 S8 C. elaphus bone 17.8 -6.9 S8 C. elaphus bone 14.1 -10.2 S9 C. elaphus bone 16.8 -7.7 S9 C. elaphus bone 14.4 -9.9 Tab.1 - δ18Op of the samples measured, their stratigraphic position and calculated δ18Ow values. Valori di δ18Op dei campioni misurati, la loro posizione stratigrafica e i valori calcolati di δ18Ow. 42 tooth values and the uncertainty on the meaning of the observed difference, teeth were not used for the recon- struction of the palaeoclimatic curve. Apart from a few cases the δ18Ow values calcula- ted from the δ18Op of C. ibex, C. capreolus, C. elaphus and Bos/Bison sp. bone samples are rather similar along the stratigraphic sequence (Fig.3). The observed similarity suggests similar δ18O values for the water taken in by the specimens of the different species despi- te their different behaviour. If the mean δ18Ow values of all the samples measured are taken into consideration the solid curve can be traced through the sequence (Fig.3). This curve is characterised by very small δ18Ow oscillations and does not show a significant trend, most of the samples yielding results that are not very far from modern mean values of atmospheric precipitation and of environmental water in that area (about -7.5 / -8.0 per mil). This consideration is not acceptable since the time interval considered is known as a cold period belonging to the last glacial. To explain this apparent inconsistency in the results it must be assumed that during glacial periods the isotope composition of ocean water was considerably enriched in heavy isotopes (18O and D) when compared to modern ocean water (Shackleton and Kenneth, 1975). This enrichment is related to the storage of huge amounts of isotopically light water in the thick and expanded ice caps covering large areas of the northern hemisphere. Since the Mediterranean basin is known to amplify the climatic effects considerably, it can be assumed that, during the last glacial period, oceanic water was at least 1 to 1.5 per mil heavier than now and the Mediterranean water was, very likely, heavier by 1.5 to 2 per mil than now. Accordingly, the oxygen isotope composition of the atmospheric water vapour and the atmospheric precipitation were 18O enriched by about the same amount. This means that the mean δ18Ow value obtained from most of the samples measured should be shifted by about 2 per mil towards lighter (col- der) values. Such a shift corresponds to a temperature decrease of about 5°C that can be reasonably accepted to represent the mean difference between the average conditions of the last glacial in this area and modern conditions. If the most negative values obtained are taken into account the δ18Ow obtained from levels S8 and D1a are lighter than the modern mean isotope compo- sition of local meteoric water by about 2/2.5 per mil. On the basis of the modern Mediterranean temperatu- re gradient (0.39 per mil) the difference between the modern value and the lowest results obtained would correspond to a negative gradient in the yearly mean air temperature of about 9.0/11.0°C if the variation of 1.5/2 per mil of the oxygen isotope composition of the Mediterranean water is taken into account. These values seem to be too high and in this case it is worth suggesting once more the possibility that these very negative values may be related to specimens that drank stream water from alpine glaciers or, at least, from high mountains. It is also possible that, given the incompleteness of our curve, the coldest periods are not documented because the cave was not occupied. Davanzo-Iacumin-Longinellii Fig.2 - Oxygen isotope composition of palaeoenvironmental waters calculated from the measured δ18Op values reported according to the stratigraphic position of the samples. Solid symbols are bones, open symbols are teeth. Valori di δ18O dell’acqua paleoambientale, calcolati dai valori di δ18Op dei diversi campioni riportati in funzione della loro posizione strati- grafica. I simboli pieni si riferiscono a campioni di ossa, gli altri a campioni di denti. 43 44.. CCOONNCCLLUUSSIIOONNSS The reported stable isotope study suggests that: - when possible, it is better to avoid the use of fos- sil teeth for palaeoenvironmental reconstruction along a stratigraphic sequence because they record the condi- tions existing during a short period of the individual’s life. This short period can be “isotopically” different in comparison to their overall life; - the studied samples recorded the climatic condi- tions during a period considerably colder than the pre- sent one with colder climatic episodes; - the temperature gradient between the studied period and recent time was not lower than about 5°C, with, probably, considerably colder episodes; - these results are of importance as database of the oxygen isotope composition of mammal skeletal remains of Palaeolithic age. By comparing all the sam- ples from Europe, measured up to now, these new data may help the detailed reconstruction of the impact of each climatic event along NS and EW geographic sec- tions. RREEFFEERREENNCCEESS Ayliffe L.K., Chivas A.R. Leakey M.G. – 1994. The retention primary oxygen isotope composition of fossil elephant skeletal phosphate. Geochim., Cosmochim. Acta, 5588, 5291-5298. Bartolomei G., Broglio A., Cassoli P., Castelletti L., Cremaschi M., Giacobini G., Malerba G., Maspero A., Peresani M., Sartorelli A. and Tagliacozzo A. –1992a. La Grotte-Abri de Fumane. Un site Aurignacien au Sud des Alpes. Preistoria Alpina, Museo Tridentino di Scienze Naturali, 2288, 131- 179. Bartolomei G., Broglio A., Cassoli P., Cremaschi M., Giacobini G., Malerba G., Maspero A., Peresani M. and Tagliacozzo A. - 1992b. Risultati prelimina- ri delle nuove ricerche al Riparo di Fumane. Annuario Storico della Valpolicella, 9-64. Bryant J.D., Froelich P.N., Showers W.J. and Genna B.J. – 1996. Biologic and climatic signals in the oxygen isotopic composition of Eocene-Oligocene equid enamel phosphate. Palaeogeogr., Palaeoclimatol., Palaeoecol., 112266, 75-89. Cassoli P. and Tagliacozzo A., - 1994. Considerazioni paleontologiche, paleoecologiche e archeozologi- che sui mammiferi e gli uccelli dei livelli del Pleistocene superiore del Riparo di Fumane (VR) scavi 1988-1991. Bollettino Museo Civico Storia Naturale Verona, 1188, 349-445. Cremaschi M., Ferrarsi M.R., Scola V. and Sartorelli A., - 1986. Note preliminari sul deposito pleistocenico del riparo di Fumane (Verona). Bollettino Museo Civico Storia Naturale Verona, 1133, 535-567. Crowson R.A. & Showers W.J. (1991). Preparation of phosphate samples for oxygen isotope analysis. Analytical Chemistry 6633, 2397-2400. Mammal skeletal remains ... Fig.3 - Oxygen isotope composition of palaeoenvironmental waters calculated from the δ18Op of bone samples reported according to the stratigraphic position of the samples. The curve suggests the possible trend of the “mean” climatic conditions. Valori di δ18O dell’acqua paleoambientale, calcolati dai valori di δ18Op delle ossa riportati in funzione della posizione stratigrafica. La curva è l’unione dei punti che rappresentano il valore medio dei campioni di ossa presenti nel medesimo strato ed indica il possibile andamento climatico. 44 D’Angela D. & Longinelli A. (1990). Oxygen isotopes in living mammals’ bone phosphate: further results. Chemical Geology (Isotope Geoscience Section) 8866, 75-82. Delgado Huertas A., Iacumin P., Stenni B., Sanchez Chillon B. & Longinelli A. (1995). Oxygen isotope variations in mammalian bone and tooth enamel. Geochimica et Cosmochimica Acta 5599, 4299- 4305. Fricke H.C. and O’Neil J.R., - 1996. Inter- and intra- tooth variation in the oxygen isotope composition of mammalian tooth enamel: some implication for paleoclimatological and paleobiological research. Palaeogeogr., Palaeoclimatol., Palaeoecol., 112266, 91-99. Lécuyer C., Grandjean P., O’Neil . J.R., Cappetta H & Martineau F. (1993). Thermal excursions in the ocean at the Cretaceous-Tertiary boundary (northern Marocco): δ18O record of phosphatic fish debris. Palaeogeography, Palaeoclimatology, Palaeoecology 110055, 235-243. Longinelli A. (1984). Oxygen isotopes in mammal bone phosphate: a new tool for paleohydrological and paleoclimatological research? Geochimica et Cosmochimica Acta 4488, 385-390. Longinelli A. (1995). Stable isotope ratios in phosphate from mammal bone and tooth as climatic indica- tors. In: B. Frenzel Ed., Proceedings of the ESF Workshop, Bern 1993, Palaeoklimaforschung Bd 15, Special Issue: 57-70. Luz B., Kolodny Y. and Horowitz M., - 1984. Fractionation of oxygen isotopes between mam- malian bone phosphate and environmental drinking water. Geochim., Cosmochim. Acta, 4488, 1689-1693. Peresani M. and Sartorelli A., 1996. The lithic assembla- ges at the Cave of Fumane. New evidence of technological variability in the Middle Palaeolithic of Northern Italy. XIII U.I.S.P.P. Congress procee- dings Forli 8-14 novembre 1996 ed. A.B.A.C.O., 22, 269-278. Shackleton N.J. and Kenneth J.P., 1975. Palaeotemperature history of the Cenozoic and the initiation of the Antarctic glaciation: oxygen and carbon isotope analyses. In: DSDP sites 277, 279 and 281: Kenneth et al., (Edrs.). Initial reports of the DSDP, 29. US Gov. Print. Office, 743-755. Davanzo-Iacumin-Longinellii Ms. ricevuto il 31 ottobre 2001 Testo definitivo ricevuto il 19 dicembre 2001 Ms. received: October 31, 2001 Final text received: December 31, 2001