Imp.Bondesan PPAALLEEOOEENNVVIIRROONNMMEENNTTAALL RREECCOONNSSTTRRUUCCTTIIOONN FFRROOMM LLGGMM TTOO HHIISSTTOORRIICCAALL TTIIMMEE IINN TTHHEE LLOOWWEERR CCOOAASSTTAALL PPLLAAIINN OOFF TTHHEE PPIIAAVVEE RRIIVVEERR.. PPRREELLIIMMIINNAARRYY PPOOLLLLEENN AANNAALLYYSSIISS OONN AA 2200 MM CCOORREE OOFF LLAAGGOOOONN AANNDD FFLLUUVVIIAALL SSEEDDIIMMEENNTTSS AAllddiinnoo BBoonnddeessaann11,, MMiirrccoo MMeenneegghheell11,, AAnnttoonneellllaa MMiioollaa22 && GGiiaannnnaa VVaalleennttiinnii22 1 Dept. of Geography, University of Padova, Via del Santo 26 - 35123 Padova 2 Dept. of Biology, University of Padova, Via G. Colombo 3 - 35131 Padova ABSTRACT This work presents the first results of a multidisciplinary study in the flood plain of the Piave River located NE of the Venice Lagoon. The aim of the work is to reconstruct the late Quaternary geomorphological evolution of the area. This study is part of a wider research on the Veneto-Friuli plain based on geomorphological surveys, remote-sensing, sediment analyses, archaeological and historical map analyses. In the frame of this study, three cores of 20 m depth were collected along a N-S transect 20 km long between San Donà di Piave and the coast of the Adriatic Sea. The transect crosses the present course of the Piave River and its fluvial ridge, while the corings were performed on the surface of old flat marshes, now reclaimed. Pollen-stratigraphical data and 14C dates are presented for the intermediate core of Palazzetto. The results of radiocarbon dating indicate that the sequence spans the Last Glacial Maximum (LGM) and the Holocene. At the base of the sequence, a peat layer (21,250±150 yr BP) is characterised by dominance of Cyperaceae (pollen, epidermal and roots fragments, fruits) as many other peat layers formerly described by other Authors in the Veneto and Friuli plain. Fine fluvial sediments attributed to the Piave River bury the basal peat. The changes in the local plant communities from fresh water to brackish/salt water communities have been recognised at the depth of ca. 6 m. This level testifies the Flandrian transgression not yet investigated in the San Donà di Piave area. The geochronological dating and the pollen analysis of the three cores will let us know the dynamic of the ingression of the sea at NE of the Venice Lagoon. At the top of the core the pollen assemblage of a peat layer (500±70 yr BP) suggests the presence of a marsh environment. RIASSUNTO Ricostruzione paleoambientale dell’area planiziale costiera del basso Piave dall’ultimo massimo glaciale all’Olocene. Analisi polliniche preliminari di una carota di 20 m di sedimenti fluviali e lagunari. Questo contributo riporta i primi risultati di un’indagine multidisciplinare condotta nella pianura alluvionale del Fiume Piave a NE della Laguna di Venezia. Lo scopo della ricerca è di comprendere l’evoluzione geomorfologica dell’area nel Quaternario recente. Essa è parte di un vasto studio della Pianura veneto-friulana, condotto attraverso indagini geomorfologiche, remote-sensing, analisi dei sedimenti, analisi di mappe storiche e di reperti archeologici. In questo contesto sono stati eseguiti tre carotaggi fino a 20 m di profondità, collocati lungo un transetto N-S della lunghezza di una ventina di km tra San Donà di Piave e la costa settentrionale del Mare Adriatico. Il transetto attraversa l’attuale corso del Fiume Piave e il suo dosso fluviale. I carotaggi sono stati eseguiti in corrispondenza di antiche aree paludose, oggi bonificate. In questo lavoro vengono presentati i risultati dello studio della sequenza di Palazzetto, collocato circa a metà del transetto. Lo scarso contenuto pollinico dei sedimenti analizzati limita l’utilizzo dei dati pollinici alla ricostruzione dei cambiamenti avvenuti nell’ambiente di sedimentazione. I risultati della radiodatazio- ne di quattro livelli di sedimenti organici fanno ritenere che la sequenza si sia deposta tra l’ultimo massimo glaciale e l’Olocene recente. Alla base sono stati individuati strati torbosi (21.250±150 a BP) a prevalenti Cyperaceae (polline, frammenti di radici e di epidermidi, frutti), caratterizzati dalla stessa flora pollinica di altri depositi torbosi, già descritti da vari Autori in sequenze della pianura veneta e friu- lana, di età compresa tra i 18.000 e i 22.000 anni BP. Le torbe basali sono ricoperte da depositi sabbiosi e limosi fluviali attribuiti alle esondazioni del Piave e depostisi in un breve periodo di tempo, circa 2000 anni. Le analisi polliniche indicano un radicale cambiamento nelle comunità vegetali locali nei sedimenti argillo-torbosi collocati a circa 6 m di profondità (6520±150 a BP): comunità vegetali di acqua dolce sono sostituite da comunità tipiche di ambienti di acque salate/salmastre. Questi sedimenti sono la prima testimonianza della trasgressione Flandriana nella zona di San Donà di Piave. Alla sommità della sequenza è stato analizzato uno strato torboso di età radiometrica di 500±70 a BP (Cal AD 1300-1500); l’associazione pollinica esaminata suggerisce la presenza locale di comunità di piante prevalentemente erbacee di palude e nessun indizio di insediamenti umani. Key words: Late Glacial Maximum, Holocene, stratigraphy, geomorphology, pollen and spores analysis, Piave River, North Adriatic. Parole chiave: Ultimo massimo glaciale, Olocene, stratigrafia, geomorfologia, analisi di polline e spore, fiume Piave, Nord Adriatico. Il Quaternario Italian Journal of Quaternary Sciences 1166(1Bis), 2003, 183-192 IINNTTRROODDUUCCTTIIOONN The paleoenvironmental reconstruction of the Venice area has been the aim of several multidiscipli- nary researches (Bortolami et al., 1977; Castiglioni & Favero, 1987; Blake et al., 1988; Correggiari et al., 1996; Mullenders et al., 1996; Serandrei Barbero et al., 2001; Kent et al., 2002). The main task of these studies is the improved knowledge of subsidence phenomenon in the past and the effects of human activities on its evo- lution. The nearby coast located NE is poorly known, even if its evolution in the past influenced or has been influenced by the hydrological conditions of the Venice area. Therefore a wider research on the Veneto-Friuli plain has started some years ago. It is based on geo- morphologic surveys, remote-sensing, sediment analy- ses, archaeological and historical map analyses (Bondesan et al., 1998; Bondesan, 2000; Bondesan & Furlanetto, 2000; Bondesan et al., 2002a; Bondesan et al., 2002b; Ghedini et al., 2002). In the framework of this study, we collected three continuous cores along a 20 km long N-S transect that goes from San Donà di Piave to the coast of the Adriatic Sea (Fig. 1), hopefully contri- buting to the study of evolution in the Venetian coastal plain of the past by means of pollen and spore analysis. In this paper 14C dates and pollen-stratigraphic data are 184 A. Bondesan et al. presented for the intermediate core of Palazzetto. The studied area is located along the modern course of the Piave River, at the north-eastern border of the Lagoon of Venice, into the coastal plain between Sile and Livenza Rivers. The territory is characterised by a dense hydrographical network, where the main rivers (Piave, Sile and Livenza) which are south-eastwards oriented, flow in connection with the artificial canals dug in XIII century by the Republic of Venice. Due to the hydrau- lic reclamation, a closed network of small canals and ditches crosses the whole plain. There are many evident traces due to the ancient shorelines, back from the present coastline; from them we can establish the existence of at least two paleo-Piave deltas, now extinct and buried. The plain surface is mainly constituted by agricultural fields, most of which laying below the sea level: the lower elevations are found northern of Caorle (Livenza mouth), where they locally reach values lower than 4 m below sea-level. Higher hei- ghts mainly stand along fluvial ridges and old dunes clo- sed to present littoral. All this territory was scattered by wide swamps, marshes and lagoons until the end of XIX century when the reclamation started, joined with a strong reduction of old dunes and fluvial ridges. It is well known that during the last strong cold pul- sation of the Pleistocene the sea level dropped more than one hundred metres below the present one. Because of the shallowness of the Adriatic Sea, the coast in the LGM shifted towards south for more than 200 km to the latitude of Ancona. The present North Adriatic Sea was then a fluvial plain where all the rivers of the Po Plain converged and perhaps joined. The Venetian plain was then rich of marshes and swamps that left extended layers of peat dating from ca. 22,000 to 18,000 yr BP (Bondesan et al., 2002b). At the same time, huge glaciers flowed south from the Alps and the majority of them reached the plain. The glacier of the Piave River basin had two snouts at Quero and Vittorio Veneto where it left its moraines. An end moraine of the Vittorio terminus has been dated to 17,670 ± 320 yr BP (Bondesan, 1999); from that date onwards the glaciers started to retreat as a consequence of the climatic impro- vement. Along with the glaciers retreat, the sea level started to lift and the coast shifted again to the north (Flandrian transgression). According to Brambati (1985) the coastline reached a position similar to the present one at about 7-6000 yr BP, while Correggiari et al. (1996) fixed the maximum sea ingression around 5000 calendars years BP, by the dating of lagoonal sediments. The transgression probably joined with changes in the rivers hydrological state, and this led to an instability of the lower reaches of the rivers that could be affected by avulsions as happened for the Rhine and Mose Delta as described by Stouthamer (2001). An avulsion node is that of San Donà di Piave, from which starts the fluvial ridge crossed by the transect described in this paper. GGEEOOMMOORRPPHHOOLLOOGGIICCAALL OOUUTTLLIINNEESS The plain belonging to the Piave River (Fig. 2) is a complex form, generated by deposition and erosion phenomena mainly due to fluvial sediments, with contri- butions of fluvio-glacial and marine deposits. Looking from the point where the mountains merge into the plain, at the height of Nervesa, we found a progressive change from the large fan of the foothills to the low plain and finally to the proper coastal plain, this last one being the result of the interference of marine and fluvial dyna- mics. Looking at the altimetry the surface lies between 80 m above sea level at Nervesa and 4 m below sea level along the coast. The system of the high plain is for- med by a sequence of great, flat, gravelly alluvial fans that lie one overlaid to the other on the Prealpine border (Bondesan, 2000). The Nervesa fan, which constitutes the Nervesa Unit, is furrowed by the present Piave River bed (Bondesan et al., 2002b); it started to form soon after the Upper-Pleistocene fluvial diversion of the Piave River from the Montebelluna to the Nervesa course. The fan is not very steep, showing a dip of 3-4‰ at the apex and 1.5-2‰ uphill of the alignment Treviso-Ponte di Piave; downhill it is less than 1‰. Its gravelly sandy sediments extend eastward up to Oderzo, coming close to River Livenza. There are five main recognizable flu- vial ridges and all of them start from the fan apex at Nervesa; they are related to paleo-Piave beds and to the present fluvial ridge of the Piave River, which is con- tinuously elevated on the plain as far as the mouth. Almost all the traces, visible on aerial photographs, situated on the left wing of the Nervesa fan, can be clas- sified as braided streams; they have been dated from 4170±60 yr BP to present day. Descending towards the sea, the low plain is found where the hydrographical network becomes denser, the gradient decreases and the sediments become finer. Here we find fluvial ridges both along present river courses and also along aban- doned rivers. Sometimes, a small watercourse interpre- Fig.1 - Location map of the studied area. 185Paleoenvironmental reconstruction from LGM ... ted as relict hydrography remains on the top of the ridge. This process leads to the formation of a network of ridges surrounding closed or semi-closed hollows where the drainage is barred. S. Donà di Piave constitutes an avulsion node, from which different watercourses in diffe- rent times depart. There are four main ridges branching out from this point across the coa- stal plain. The western one is the Piave Vecchia ridge (PV in fig. 2), rimming the Venice Lagoon border. Eastwards, the Taglio da Re ridge (TR in fig. 2) gains the Cavetta Channel, back of Jesolo; it is the outrea- ch of a diversion channel excavated by the Venetians in 1534. Nowadays the ridge is overflowed by a scant recla- mation channel. The present main Piave River bed is artifi- cially excavated and it gets Eraclea flowing straight for some kilometres. It follows the left flank of the ridge (PC in fig. 2), 3 km wide and 3 m height over the surrounding plain, probably created by a former ancient Piave. It shows bran- ches and crevasse splays with traces of ancient channels well documented in antique and modern maps. The Piovan (or Piveran) Channel (CP in fig. 2) is very interesting: it flows along a great ridge starting from S. Donà di Piave and it turns southeast after the confluence with the Grassaga Channel toward Stretti. The ridge is 3 km wide and 2 m high at S. Donà di Piave, but it becomes narrow and flat as far as it loo- ses any morphological eviden- ce. The Via Annia, an ancient Roman Way, follows the path of the ridge and changes its direction when it crosses the mound through a Roman brid- ge whose remains were found along the Grassaga Channel; then the ridge follows on to the archaeological site of Cittanova. The fluvial ridges define large low-laying interfluvial areas, where paleo-river beds and paleo-lagoon channels are clearly visible on aerial photo-Fig. 2 - Geomorphological sketch map of the Piave River plain. 186 graphs. The medieval and modern ingression of the lagoon is marked along the stretch north of Taglio del Sile, going along the line linking S.Donà di Piave to Boccafossa (Livenza River) (Castiglioni & Favero, 1987; MURST, 1997). The lagoon deposits spill at the surface and we assert they are very recent. Organic samples of surface lagoon sediments gave a geo-chronological age of 1200±190 yr BP (Utc - 7801), 955±55 yr BP (Rome – 1180) and 658±39 yr BP (Utc - 7802) (personal commu- nication of Henk Berendsen, 1998; Bondesan et al., in press-b). These data confirm that the lagoon sediment sealed the depressions between the older main fluvial ridges in historical times. Along the coast we find the Littoral Unit (Bondesan et al., 2002b) constituted by ancient coastlines and ali- gnment of paleo-dunes, most of them now artificially levelled. The inner traces could be connected with the littoral deposits formed at Lio Maggiore and Lio Piccolo which are dated about 6500 yr BP (Blake et al., 1988). The ancient coastlines are distributed over the whole district between Rivers Piave and Sile. The gene- ral distribution of alignments shows at least four zones, each representing an advancing phase of the littoral, fol- lowed by an erosion event in all likelihood. The coastal stretch defines a cuspate delta, probably asymmetric, well developed in the area nowadays occupied by the Piave mouth, but perhaps much more protruded into the sea, and lately interested by an erosion phase. The development of the older paleo-delta is ascribed to the upper Atlantic, while during the transition Atlantic- Subboreal, the right portion of the Piave paleo-delta might have started. The formation of the more recent cuspate delta, whose traces are scarce, seems to have started in the second part of the Subboreal (Bondesan et al., in press-a). MMAATTEERRIIAALLSS AANNDD MMEETTHHOODDSS The continuous core of Palazzetto was collected using a mechanic sampler, down to a depth of 21.71 m. The lithology of sediments is indicated at the left side of the diagrams, according to the Troels-Smith System. We selected 42 samples (2 cm3). The uppermost 80 cm (agricultural soil and sand) and the sediments from 6.96 m to 10.10 m and from 18.49 m to 19.39 m (sands) were not considered because unsuitable for pol- len conservation. Tablets of a known concentration of Lycopodium spores were added to each samples prior to preparation so that fossil pollen concentration could be estimated. Samples were treated with HCl (37%), cold HF (50%) and boiling NaOH (10%). A sieve of mesh of about 200 µm was employed to isolate macrofossil and epidermal fragments from peat sedi- ments. In the end, samples were acetolysed. Residues were mounted on slides, in glycerine. All samples were analysed to count pollen and spores of ferns, bryophytes, algae and fungi; epidermal fragments were only identified. Identification and coun- ting were made with a magnification of 400x and critical determination at 1000x. Identification was aided by a modern pollen and spore reference collection, and illu- strations and keys including Reille (1992; 1998), Punt (1976), Punt & Clarke (1980; 1981; 1984), Punt et al. (1988) and Moore et al. (1991). “Saccate grains” indica- tes broken pollen grains of the Pinaceae family, but their genus or species are indeterminable. Type 303, is described by Van Geel et al. (1981). They do not iden- tify it, but they report that it mainly occurs in the Late Glacial (Younger Dryas) and Early Holocene sediments. We record high values of Type 303 in peat that was rich of mosses fragments; therefore we put forward the hypothesis that one/more members of Bryophytes could produce it. Studies are in progress. Type 200 is descri- bed by Van Geel et al. (1989) and Kuhry (1997). The zygospores of the Zygnemataceae are described by Van Geel (1978), Van Geel & Van der Hammen (1978) and Van Geel et al. (1981). Identification of some epi- dermal fragments was aided by a modern collection. Pollen counting mostly continued until pollen sum (trees+shrubs+upland herbs) reached 100÷250 units, according to the pollen content of the sample. The per- centage of helophytes and hydrophytes, of pterydophy- tes, of bryophytes, of algae, of fungi and of foraminifera remains, respectively, is calculated over the pollen sum added to each group’s total. Unknown grains did not show the characteristic features of pollen grains; the- refore they were not included in the pollen sum. Indeterminable grains (degraded, corroded, broken, pyritized) have been counted, their percentage is calcu- lated over the sum of all pollen grains (identified, inde- terminable and hidden). The pollen diagrams were cal- culated and drawn using the programs TILIA and TILIA.GRAPH (Grimm, 1990). We present a pollen pre- servation diagram (Fig. 3), a pollen and spore concen- tration diagram (Fig. 4), a pollen and spore percentage synthetic diagram (Fig. 5), a pollen percentage diagram of selected taxa (Fig. 6) and a pollen percentage dia- gram of helophytes and hydrophytes (Fig. 7). When ter- restrial pollen sum didn’t reach 100 units, we have marked the sporomorphs findings with a plus sign in the diagrams. The occurrence of epidermal fragments has been indicated with circle signs. Four 14C dates were obtained from sediment associated with a specific layer identified from the pollen diagrams (Tab. 1). Radiocarbon dating was conducted at the Beta-Analytic Inc. Miami – Florida. Tab. 1 - Radiocarbon dates from Palazzetto core. Lab. reference Sample depth 14C age method (cm) (yr BP) Beta - 173728 76-80 500±70 Standard Beta - 168127 581-609 6520±50 Standard Beta - 168128 1323-1333 19850±120 Standard Beta - 168129 2049-2059 21250±150 Standard RREESSUULLTTSS We have analysed 42 samples from the Palazzetto core (Fig. 3), but 15 samples were virtually free of pollen (pollen sum<10 grains) and 8 samples were rather poor in pollen (pollen sum<100 grains). Pollen concentration is generally low in the A. Bondesan et al. 187 part of the core the pollen preservation is poor, with up to 40% indeterminable pollen, probably resulting from physical transport of the pollen grains in the silt sedi- ments. Grains infilled with crystals of pyrite (pyritized) are continuously present in the upper part of the core documenting chemically-reducing alluvial environment (Berglund & Ralska-Jasieviczowa, 1986). The development of local and upland vegetation has not been continuously documented in the sequence because of the presence of sterile sand layers (see lithology), and because of the scarcity of pollen grains in fluvial sediments. Nevertheless indicative local environments may be reconstructed from the following stratigraphical-pollen data to recognise transgressive trend of the coast line at the north of Venice Lagoon. The chronology of the sequence is based on radiocarbon dates (Tab.1). Hereafter the most important features of the pollen analysis are described. 2211..4444÷÷1188..1199 mm. The basal peat (20.59÷20.34 m) contains a high percentage of uplands herbs (NAP) including Poaceae (53% average), Artemisia, that rea- ches its highest value for the entire record (12%), Chenopo- diaceae, Asteroideae undiff. and Thalictrum. Trees and shrubs are represented by low percen- tage: Pinus (6% average), Be- tula, Ephedra fragilis type, Larix type, Abies, Fagus sylvatica type and Acer campestre type. The local wetland vegetation is mostly represented by Cypera- ceae pollen (93% of helophy- tes+hydrophytes - Fig. 7) and by aerial fragments, roots and fruits of Carex species. We also find pollen of Lemna, Potamogeton subg. P . type, Nuphar, and Sparganium emersum type. High percentage of algal spores (Zygnema type) is recorded in this layer. The presence of mos- ses in the local environment is indicated by the highest value (67%) of spore percentage in the entire sequence and by many fragments of leaves and little stems. The dominant type of spores is Type 303. Upland local vegetation is dominated by Populus (16%) that could occur in the area surrounding the depressions, among grassy vegetation of Poaceae. The pol- len and bryophytes concentra- tions decrease in the peat layer Paleoenvironmental reconstruction from LGM ... Fig. 4 - Pollen and spores concentration diagram of Palazzetto core. Fig. 3 - Pollen preservation in the sediment of Palazzetto core (Total pollen sum: trees, shrubs, upland and wetland herbs, indeterminable, hidden pollen grains). sequence: it reaches the value of about 100 x 103 grains/cm3 only at the depth of 20.34 m (peat) and at the depth of 5.84 m (organic clay) (Fig. 4). We believe that the scarcity of pollen grains was caused by a fast rate of silt and clay sedimentation in the floodplain environ- ment. In fact unsuitable conditions of preservation are unlikely because of the good state of the rare grains we found, almost from the bottom to the depth of 5.84 m and the percentage of indeterminable grains (degraded, corroded, broken and pyritized) surpassing 20% only in the sample at the depth of 13.00 m (Fig. 3). In the upper 188 A. Bondesan et al. Fig. 5 - Pollen and spore percentage synthetic diagram of Palazzetto core. at 18.34÷18.39 m but the pollen assemblage does not change. The fungal spore concentration increases. The dominant type is Type 200. The occurrence of pollen in the next layers up to 13.84 m is very low. 1133..8844÷÷1122..8811 mm. NAP continues to dominate the pollen assemblage. The higher value of total pollen con- centration (60 x 103 grains/cm3) corresponds with a peat layer (13.18÷13.16 m). Poaceae are the dominant taxon (maximum value 85%). Cyperaceae’s percentage shows a declining trend. The sediments at the top con- tain the highest percentage of hydrophytes (61% of helophytes + hydrophytes). Instead Artemisia and Chenopodiaceae percentages decrease. Arboreal taxa are mostly represented by Pinus (P. cembra, P. mugo/P. sylvestris) and “Saccate grains”. Populus and Salix show increasing percentages, while Alnus is rare. Picea and Castanea sativa type are recorded for the first time. 1122..8811÷÷1100..0000 mm. NAP continues to dominate the pollen assemblage and Poaceae are the dominant taxon: the highest concentration value corresponds to a layer of organic clay at the depth of 11.54÷11.52 m. Cyperaceae pollen percentage increases. Hydrophytes show lower concentration and percentage values if com- pared to the previous layers. We do not find any signifi- cant changes in the other herbaceous pollen taxa, as Artemisia, Chenopodiaceae, Asteroideae undiff. and Cichorioideae undiff. Pinus undiff. prevails even if with low percentages; Betula, Picea, Larix type, Abies are present with only few pollen grains. 77..0000÷÷33..7766 mm. A 3 m thick layer of sand separates these layers by the previous ones. The pollen content of the samples from 7.00 m to 5.76 m is very low (<10 grains per slide). We have identified few pollen taxa: Pinus undiff., Picea, Larix type, Salix, Poaceae and Cyperaceae. The occurrence of 3 inner organic linings of Ammonia beccarii at the depth of 6.20 m, is very inte- resting, mostly because of the poorness of samples. We have continuously found it from this layer up to the top of the core. The pollen assemblage of the organic clay sam- ples (5.86÷5.78 m) is characterised by the dominance of NAP as in the previous layers, but it shows some chan- ges in pollen taxa distribution. Some of the previously most important taxa decrease: helophytes and hydrophytes are rare, Poaceae decrease to a percenta- ge values of about 10%, Chenopodiaceae pollen is abundant for the first time: it attains a maximum of 44%. New taxa of Plumbaginaceae are found: Limonium vul- gare type A and Armeria maritima type A and B, that include species living in salt marshes as L. narbonense Miller. Plumbaginaceae undiff. pollen attains its highest frequency in the core (7%). Arboreal pollen does not dominate in the pollen assemblage but some taxa show their maximum concentration value in the core: Quercus robur group, Corylus, Fagus sylvatica type, Pinus undiff., “Saccate grains”, Betula. Carpinus type firstly occurs. Castanea sativa type (<1%) is continuously represented from here up to the top of the core with values ranging around 2-3%. The inner organic lining of Ammonia beccarii (“Other” in fig. 5) attains a maximum concentration value in the core. Silty grey clay sediments constitute the upper layer and they are virtually barren of sporomorphs. 33..7766mm÷÷11..0088 mm. The pollen assemblages are cha- racterised by: i) the dominance of AP; ii) the first notable increase of several broad-leaf trees taxa; iii) the contem- poraneous presence of Quercetum s.l. elements, of Pinus (decreasing) and Betula, and of Picea, Fagus and Abies. AP percentage value attains 90% at the bot- tom of these layers, but it shows a declining trend in the upper layers. The palynological richness attains its maximum value of 19 arboreal pollen taxa and 14 her- baceous pollen taxa. Riparian plants, such as Alnus, Populus and Salix, show increased percentages. Poaceae slightly increase from the bottom to the top. 189Paleoenvironmental reconstruction from LGM ... There is a reduction of helophytes and hydrophytes. Indeterminate grains (mechanically damaged) and “Saccate grains” are continuously found with the highest concentration and percentage values at the top of the core. There is a notable reduction of pollen content at the top: total pollen concentration drops to 8 x 103 grains/cm3. Ammonia beccarii inner organic linings are found in each sample. 11..0088÷÷00..8822 mm. The pollen concentration increases at the top of the core, but deciduous plants disappear except for Populus and Salix. Helophytes and hydrophy- tes pollen prevail on the other taxa. The spores of ptery- dophytes and bryophytes are very abundant. We do not find pollen that indicates the nearby presence of human activities. IINNTTEERRPPRREETTAATTIIOONN OOFF RREESSUULLTTSS The lithology of sediment, along with pollen and macrofossil records point to the fact that the sedimenta- tion environment at the bottom of Palazzetto core, chan- ged from an open water stage (22.00÷21.00 m), proba- bly with a high rate of sedimentation, to a mire stadium (20.49÷20.34 m), with Carex species and brown mos- ses at 21,250±150 yr BP. The presence of a water table close to the surface or other geomorphological condi- tions, i.e. shallow closed depressions or lasting snow cover, are suggested by the ecological optimum of some Carex species (C. ferruginea Scop., C. limosa L., C. rostrata Stokes), which is characterised by a good water availability (Buffa & Sburlino, 2001; Bragazza, 1999). The assemblage of spores of Zygnema type and aquatic taxa (Lemna, Potamogeton, Callitriche, and Nuphar) pollen points to the local occurrence of pools in some shallow depression of the mire. The conditions of sedimentation change significantly at the depth of 19.30 m. Silt and sand layers indicate an active nearby fluvial system. In the upper peat layer (18.34÷18.39 m) the occurrence of fungal Type 200 could be indicative of marked table fluctuations during the growing season and of relatively dry microhabitats in temporary desicca- ting bottoms of pools. The pollen concentration and assemblage are typical of a cold and dry climate, that corresponds to the radiocarbon dating results. The low pollen concentration found in the analy- sed sediments until 13.84 m could be linked to the high rate of sedimentation suggested by the radiocarbon dates. From the depth of 13.84 m the local conditions gradually change: the total pollen concentration increa- ses in concomitance with the development of a new mire stadium (13.16÷13.18 m). Initially high percentage and concentration values of aquatic plants pollen sugge- st that aquatic conditions prevail. Poaceae (perhaps Phragmites sp.) and Cyperaceae stands, grow around the basin but they gradually fill it up, until the formation of a mire (13.16÷13.18 m). The lithology of sediments from 12.81 m to 7.00 m (roughly represented in this paper) suggests that the sedimentation environment changed many times, but we didn’t note significant changes in pollen associa- tions. From the age of 6520±150 yr BP (6.09÷5.81 m), many records suggest local development of a lagoon environment. Plumbaginaceae (Limonium vulgare type A and Armeria maritima type B include L. narbonense Miller), Poaceae, Cheno-podiaceae and Asteraceae Asteroideae undiff. represent the local vegetation. We suppose that modern analogue could be found in pio- neer vegetation of silt and silty sands, living in salt mar- shes of the lagoons in the northern Adriatic coast. Lymonium narbonense Miller, Spartina maritima (Curtis) Fernald and some species of Chenopodiaceae form the association Limonio narbonensis-Spartinetum maritimae that changes in Limonio narbonensis- Puccinellio festu- Fig.6 - Pollen percentage diagram of selected taxa of Palazzetto core. ciformis when the ground level increases (Caniglia et al., 1997). Pollen morphology does not allow any identification of Spartina maritima (Curtis) Fernald or Puccinellia palustris (Seen.) Hayek or different spe- cies of Chenopodiaceae; the- refore the stage of vegetation dynamics is not given to know. Studies are in progress to reco- gnise epidermal fragments of roots. The presence of Ammonia beccarii inner organic linings is another important feature of the zone, because A. beccarii is the dominant taxon among the fora- minifera that today are living in the water of Venice Lagoon (Serandrei Barbero et al., 2001). Its occurrence is indicative of an open salt/brackish water envi- ronment too. Upland vegetation abruptly changes: we note that pollen percentage of Pinus is overtaken by pollen percentage of thermophilous broad-leaf plants. Rizzi Longo (1996) reviewed pollen records from Friuli plain sequences. She underlies that Pinus decrea- ses during the passage from Pre-Boreal to Boreal period and the association of oaks and other broad-leaf trees grows. In the plain belt mixed oak forest domina- tes until historical times. Radiocarbon dating and the results of our preliminary pollen analysis is in full accor- dance with these data. From the depth of 3.76 m an open water environ- ment covered the area and upland forest vegetation was probably near the banks: riparian plants, such as Alnus, Salix and Populus lived near the basin, mixed oak forest covered the surrounding areas. The low percentage of Cyperaceae, aquatic plant pollen, and fresh-water algal spore, and the constant presence of Ammonia beccarii indicate a salt/brackish quality of water. Dominance of Quercetum s.l. decreases towards the top of the core; Castanea and Corylus pollen are constantly present and do not change their trends. These are common features of Friuli plains pollen diagrams and they are dated to historical times (Rizzi Longo, 1996). Drastic changes in pollen association suggest the development of new environmental conditions at the age of 500÷70 yr BP (Cal AD 1400 to 1450). The disap- pearance of tree pollen and the consistent presence of pterydophytes spores could be the signs of the late Middle Age human activity in the area (Cagnin, 2000). The presence of Type 200 fungal spore indicates that the area was subject to marked fluctuations of the water table, during the growing season. The decline in Ammonia beccarii and increased frequencies of fresh- water plants suggest a lower influence of seawater. CCOONNCCLLUUDDIINNGG RREEMMAARRKKSS The results from radiocarbon dating suggest that the Palazzetto sequence spans the LGM and the histori- cal times. The depositions of silt and sand layers inter- rupted the pollen and lithological sequence, and therefo- re the history of the local environment is fragmented. Probably the studies in progress of the other two cores (Ca’ Fornera and Fiorentina) will add further elements to the history of the area. Cyperaceae peat layers were deposited at the bot- tom of the core during the end of the last glacial period. These layers could be correlated with similar peat layers from the Veneto and Friuli Plain, already described by other Authors (Bortolami et al., 1977; Mullenders et al., 1996; Serandrei Barbero et al., 2001; Bondesan et al., 2002b): studies are in progress. Four events of fluvial sedimentation, attributed to the Piave River, follow the deposition of the basal peats at the depth of about 19 m, 17.5 m, 15 m and 11 m. They occurred before the beginning of the Holocene. The last important fluvial sedimentation (10÷7 m) occurs before the development of brackish/salt communities in the area, at the age of ca. 6520±150 yr BP. Pollen and spore association at the depth of ca. 6 m, documents the development of a lagoon environment, connected to the Flandrian tran- sgression not yet described in the San Donà di Piave plain. In the nearby Venice Lagoon, first lagoon sedi- ments after the Flandrian transgression have been dated to 5090±100 yr BP by Serandrei Barbero et al. (2001). The geochronological dating and pollen analysis of the three cores will let us verify the dynamics of the ingression of the lagoon. AACCKKNNOOWWLLEEDDGGEEMMEENNTTSS We would like to thank the anonymous referees for their comments and useful suggestions on a pre- vious draft and dr. Alessandra Asioli for the identification of Ammonia beccari inner organic lining. We thank dr. Paolo Mozzi for the given help in core analysis. 190 A. Bondesan et al. Fig. 7 - Helophytes and hydrophytes percentage diagram of Palazzetto core. RREEFFEERREENNCCEESS Berglund B.E. & Ralska-Jasiewiczowa M. (1986) - Pollen analysis and pollen diagrams. In: Handbook of Holocene Palaeoecology and Palaeohydrology. Berglund B. E. (ed.), John Wiley & Sons, Chicester, p. 455-484. Blake H., Bondesan A., Favero V., Finzi E. & Salvatori S. (1988) - Cittanova - Heraclia 1987: risultati pre- liminari delle indagini geomorfologiche e paleo- geografiche, Quad. Arch. del Veneto, 44, p. 112- 135. Bondesan A. (1999) - Una nuova data 14C nell’anfiteatro morenico di Vittorio Veneto e la costruzione del suo sandur (pianura del F. Meschio), VIII Convegno Glaciologico Italiano - Bormio, 9-12 set- tembre 1999, Risposta dei ghiacciai alpini ai cam- biamenti climatici, SEMINARIO MURST: “Risposte dei processi geomorfologici alle variazioni climati- che in ambiente montano”. Bondesan A. (2000) - I fiumi, le lagune e il mare. La geomorfologia della laguna, in “Il Piave”, Cierre, Verona, p. 76-82. Bondesan A., Calderoni G. & Rizzetto F. (in press-a) - Geomorphologic Evolution of the Lower Piave River, Suppl. Geogr. Fis. Dinam. Quat. In press. Bondesan A., Caniato G., Valerani V. & Zanetti M. (eds) (1998) - Il Sile, Cierre, UNESCO, Verona, pp. 357. Bondesan A., Caniato G., Valerani V. & Zanetti M. (eds) (in press-b) - Il Brenta, Cierre, Verona. Bondesan A. & Furlanetto P. (2000) - L’uomo e le acque: cartografia storica e indagine geomorfolo- gica a confronto, in “Il Piave”, Cierre, Verona, p. 87-91. Bondesan A., Bertani B., Fontana A., Furlanetto P. & Primon S. (2002a) - Carta della fotointerpretazione e dei siti archeologici della Provincia di Venezia tra i fiumi Livenza e Tagliamento, CD-ROM edito dal Magistrato alle Acque - Consorzio Venezia Nuova. Bondesan A., Mozzi P. & Calderoni G. (2002b) - L’assetto geomorfologico della pianura veneta centro-orientale: stato delle conoscenze e nuovi dati, Scritti in ricordo di Giovanna Brunetta, Dip. Geogr. Univ. Padova, p. 19-38. Bortolami G.C., Fontes J., Markgraf V. & Saliege J.F. (1977) - Land, sea and climate in the Northern Adriatic region during Late Pleistocene and Holocene, Palaeogeography, Palaeoclimatology, Palaeoecology, 2211, p. 139-156. Bragazza L. (1999) - Spatial patterns of plant species in a poor mire on the Southern Alps (Italy), Plant Biosystems, 113333,, no. 1, p. 83-92. Brambati A. (1985) - Modificazioni costiere nell’arco lagunare dell’Adriatico settentrionale, Antichità Altoadriatiche, Studi Jesolani, 2277, p. 13-47. Buffa G. & Sburlino G. (2001) - Carex ferruginea gras- slands in the south-eastern Alps, Plant Biosystems, 113355, no. 2, p. 195-206. Cagnin G., (2000) - “Per molti e notabel danni i qual riceve campi, pradi, ville e vigne per lo corso macor de la Piave”. Il difficile rapporto tra un fiume e il suo territorio nel Medioevo”, in “Il Piave”, Cierre, Verona, p. 212-227. Caniglia G., Contin G., Fusco M., Anoè N. & Zanaboni A. (1997) - Confronto su base vegetazionale tra due barene della laguna di Venezia, Fitosociolo- gia, 3344, p. 111-119. Castiglioni G.B. & Favero V. (1987) - Linee di costa anti- che ai margini orientali della Laguna di Venezia e ai lati della foce attuale del Piave, Ist. Ven. SS. LL. AA., Rapporti e Studi, 1100, p. 17-30. Correggiari A., Roveri M. & Trincardi F. (1996) - Late Pleistocene and Holocene evolution of the north Adriatic sea, II Quaternario, 99, no. 2, p. 697-704. Ghedini F., Bondesan A. & Busana S. (eds) (2002) - La tenuta di Ca’ Tron. Ambiente e storia nella terra dei dogi, Cierre, Verona, pp. 240. Grimm E.C. (1990) - TILIA and TILIA.GRAPH, PC spreadsheet and graphics software for pollen data, INQUA Working Group Data Handling Methods Newsletter 44, p. 5-7. Kent D.V., Rio D., Massari F., Kukla G. & Lanci L. (2002) - Emergence of Venice during the Pleistocene, Quaternary Science Review, 2211, p. 1719-1727. Kuhry P. (1997) - The palaeoecology of a treed bog in western boreal Canada: a study based on microfossils, macrofossils and physico-chemical properties, Rev. Palaeobot. Palynol., 9966, p. 183- 224. Moore P.D., Webb J.A. & Collinson M.E. (1991) - Pollen analysis, II ed., Blackwell Science, Oxford, pp. 216. Mullenders W., Favero V., Ciremans M. & Dirickx M. (1996) - Analyses polliniques de sondages a Venise (VE-I, VE-I bis, VE-II), in "Pleistocene Palynostratigraphy". Gullentips F. (ed), Aardk. Mededel., 77, p. 87-117. MURST (1997) - Carta geomorfologica della Pianura Padana, 3 fogli, scala 1:250.000, S.el.ca., Firenze. Punt W. (1976) - The Northwest European Pollen flora, I. Rev. Palaeobot. Palynol., 1177,, 1199 e 2211. Elsevier, Amsterdam, pp. 145. Punt W., Blackmore S. & Clarke G.C.S. (1988) - The Northwest European Pollen flora, V. Rev. Palaeobot. Palynol., 5577, Elsevier, Amsterdam, pp. 154. Punt W. & Clarke G.C.S. (1980) - The Northwest European Pollen flora, II. Rev. Palaeobot. Palynol., 2266, Elsevier, Amsterdam, pp. 265. Punt W. & Clarke G.C.S. (1981) - The Northwest European Pollen flora, III. Rev. Palaeobot. Palynol., 3333, Elsevier, Amsterdam, pp. 138. Punt W. & Clarke G.C.S. (1984) - The Northwest European Pollen flora, IV. Rev. Palaeobot. Palynol., 4422, Elsevier, Amsterdam, pp. 369. Reille M. (1992) - Pollen et spores d’Europe et d’Afrique du Nord, Marseille, pp. 520. Reille M. (1998) - Pollen et spores d’Europe et d’Afrique du Nord, Supplement 2, Marseille, pp. 521. Rizzi Longo L. (1996) - Palynological evidence of the forest vegetation of the Friuli plain, Allionia, 3344, p. 287-295. Serandrei Barbero R., Lezziero A., Albani A. & Zoppi U. (2001) - Depositi tardo-pleistocenici ed olocenici nel sottosuolo veneziano: paleoambienti e crono- logia, Il Quaternario, 1144, no. 1, p. 9-22. Stouthamer E. (2001) - Holocene avulsions in the Rhine-Meuse delta, The Netherlands, E. Stouthamer, Utrecht, pp. 223 191Paleoenvironmental reconstruction from LGM ... Van Geel B. (1978) - A palaeoecological study of Holocene peat bog sections in Germany and the Netherlands, based on the analysis of pollen, spo- res and macro- and microscopic remains of fungi, algae, cormophytes and animals, Rev. Palaeobot. Palynol., 2255, p. 1-120. Van Geel B., Bohncke S.J.P. & Dee H. (1980-1981) - A palaeoecological study of Upper Late Glacial and Holocene sequence from “De Borchert”, The Netherlands, Rev. Palaeobot. Palynol., 3311, p. 367- 448. Van Geel B., Coope G.R. & Van der Hammen T. (1989) - Palaecology and stratigraphy of the Lateglacial type section at Usselo (The Netherlands), Rev. Palaeobot. Palynol., 6600, p. 25-129. Van Geel B. & Van der Hammen T. (1978) - Zygnemataceae in quaternary colombian sedi- ments, Rev. Palaeobot. Palynol., 2255, p. 377-392. 192 A. Bondesan et al.