Marine Geology

Marine Geology

Paleoceanographic Changes During the LateHolocene in the Adriatic Sea (Central Mediterranean) A. Piva1 , A. Asioli2 , F. Trincardi3 , R.R. Schneider4 , L. Vigliotti3 1, ENI E&P Division, Sedimentology, Petrography & Stratigraphy Dpt., S. Donato Milanese, Italy 2, Institute of Geosciences and Earth Resources, CNR, Padova, Italy 3, Institute of Marine Sciences, CNR, Bologna, Italy 4, Institute of Geosciences, Christian-Albrechts-Universit`at zu Kiel, Kiel, Germany Andrea.Piva@eni.com Abstract Sub-millennial scale variability during the last 6000 years is recognized in the planktic and benthic foraminifera records from sediment cores retrieved from the Central Adriatic shelf and the Southern Adriatic deep basin. The δ 18 O shift of benthic foraminifer B. marginata suggests an increase of dense water production around 7500 years BP with the onset of the modern routing of the North Adriatic Dense Water occurring before the end of Mediterranean Sapropel S1. After 5500 years BP, the northward intrusion of the salty Levantine Intermediate Water (LIW) in the Mid Adriatic slope basin (MAD) is recorded by the δ 18 O of intermediate-water dweller G. bulloides. During the late Holocene, short-lived episodes of increased runoff are indicated by δ 18 O values of G. bulloides in the MAD with concurrent drops in G. sacculifer concentration, suggesting either increased fresh water input impacting the entire water column (thereby forcing the LIW to a deeper level) or reduced LIW formation failing to intrude the MAD slope basin. Repeated abundance peaks of the planktic foraminifer Globigerinoides sacculifer represent warm-dry intervals, among which the Medieval Warm Period, the Roman Age, the late Bronze Age and the Copper Age. Moreover, the Little Ice Age (LIA) is approximated at the base by the Last Occurrence of G. sacculifer (550 years BP), and its two main (coldest) phases are recorded on all the shelf cores by two peaks of the benthic foraminifer Valvulineria complanata.

1

Introduction

The recent literature on short-term climate change during the Holocene revealed timing, amplitude and possible mechanisms of sub-Milankovitch millennial to centennial fluctuations. Although the origin of this forcing is not fully understood, some authors suggest that, at least in the

North Atlantic region, the weak quasiperiodic forcing (1500±500 years) of climate changes during the Holocene mirrors the Dansgaard-Oeschger oscillations documented during glacial intervals, though with a substantially attenuated expression [1, 2, 3]. Three of the eight warm-cold events recognized during the Holocene oc-

Marine Geology

curred during the last 5500 years [1, 4].

Among these events the Little Ice Age (LIA) is the coldest (ca. 1500-1880 AD), the best chronologically constrained and documented by historical reports, artistic production, economical and societal changes. The goal of this paper is to review Adriatic Sea records studied within the framework of the two EC projects EURODELTA AND EUROSTRATAFORM and showing high resolution short-term (century- to millennial-scale) climatic oscillations during the last 6000 years within the context of the climate variability at hemispheric scale [5]. The Adriatic Sea is one of the key areas of deep water formation in the Mediterranean [6], it is a land-locked basin providing an excellent record of terrestrial proxies (pollens and magnetic properties) related to soil composition and erosion [7], it provides a continuous (slope) and very expanded (mud belt on the inner shelf) sedimentary record of the late Holocene [8, 9], and it is constrained by refined geochronological control, including tephra layers [10]. Eight sediment cores between 55 m and 1126 m water depth and located along the path impinged by the North Adriatic Dense Waters (NAdDW) are here presented at sufficient resolution to recognize regional-scale events and to evaluate the impact of the same environmental signals within shallow and deeper-water settings.

2

Study area

The Adriatic Sea is a narrow epicontinental basin with a low axial topographic gradient in the North, and a narrower and steeper shelf further South (Figure 1). On the shelf, the late Holocene mud wedge is up to 35 m in thickness above the max318

imum flooding surface (mfs), dated ca.

5500 cal. years BP [11, 8, 9]. Highresolution seismic-stratigraphic correlation and integrated stratigraphy indicate a reduced deposition between 5500 and 3700 cal. years BP over much of the shelf and basin [11, 7, 9, 10]. Above this interval, sediment accumulation rates increased to up to 1.5 cm¡yr−1 [12]. These data indicate an overall shore-parallel advective component controlling sediment dispersal and a significant decrease of the sediment accumulation rates during the last century compared to the Little Ice Age interval [8, 9].

The Mid Adriatic Deep (MAD) is a slope basin about 260 m deep providing an excellent paleoenvironmental record through the last deglaciation [13, 14]. Cores from this area reflect atmospheric forcing, river runoff and water mass intrusion from the open Mediterranean basin.

3

Material and methods

The cores selected for this review are described in Piva et al. ([5] and reference therein) and their results are compared with the ones published for core RF93-30, collected on the seaward pinch-out of the 35-m-thick Late Holocene mud wedge revealing century to millennial-scale environmental changes over the last 6000 years [7]. Additional cores from slope settings, where sediment accumulation rates are high, complement the paleoenvironmental reconstructions from shallow-water sites.

Foraminifera. Cores AN97-15, SA03-9, SA03-11 and PRAD1-2 were sampled every 10 cm, and core AMC99-1 every 6 cm.

For the sample processing and counting method the reader is referred to Piva et al.

[5]. All taxa are quantified as percentages of the total number of planktic and benthic

Marine research at CNR

45˚ -50 -5 0

-50 -50

N

-50

dr

0 -5

c

S

-2

80

PRAD1-2 -1

-160

-1 30

ea -100

-200

-300

-1

-500 -400

-100 30

0

0

-16

-14

-100

20

-1

RF93-30 0

RF93-77

20

Late Holocene mud wedge thickness (ms) 42˚

-12 0

-800

-100

00

CM92-43 AMC99-1

-50

00 -1

0 10 20 30 40 50

-200

43˚

ti

-1

-1

-2

00

ia

-600

A

00

AN97-15

-1 00 0

44˚

-80

-20

-10

SA03-11

-10

0

Rome

Italy

SA03-09

41˚

Bari 0 km 50

40˚ 12˚

13˚

14˚

15˚

16˚

17˚

18˚

19˚

Figure 1: Location map of the cores with reported the Adriatic mud wedge thickness on the west [8]. foraminifera, respectively, while the concentration is reported as number of specimens per gram of dry sediment. Regarding planktic foraminifera Globigerinoides ex gr. ruber includes Globigerinoides ruber and Globigerinoides elongatus; while Globigerinoides sacculifer comprehends Globigerinoides trilobus, Globigerinoides sacculifer and Globigerinoides quadrilobatus [15]. Oxygen and carbon stable isotopes.

Oxygen and carbon stable isotope analyses were obtained from selected monospecific foraminifera specimens, planktic species Globigerina bulloides in cores SA03-9, SA03-11, PRAD1-2, G. sacculifer for core AMC99-1 and the benthic species Bulimina marginata for PRAD1-2 from the size

fraction > 0.180 mm [5]. All the isotopic records discussed are expressed as per mil ( ) deviation with respect to the international V-PDB standard with no correction for the ice-volume effect. Radiocarbon dates. The 14 C AMS dates were performed in cores AMC99-1 and SA039 on (mixed) planktic and benthic samples (monospecific Cibicidoides pachyderma, mixed with Uvigerina peregrina in one sample) at the Poznan Radiocarbon Laboratory, Poland, from the size fraction > 0.250 mm. The data were then calibrated using Calib 5.0.2 Radiocarbon Calibration Program [16]. For further details the reader is referred to Piva et al. [5]. Magnetic measurements. Measurements of the Secular

319

Marine Geology

Variation of the magnetic field were carried out for core AMC99-1 on u-channel samples collected from each section of the core and measured on an automated 2G cryogenic magnetometer at the University of California at Davis [17].

4

Chronology

The chronologic framework relies on the integration of several and independent methods, such as: regional bioevents (LO of the planktic species Globorotalia inflata marking the end of the post-glacial sealevel rise), 14 C AMS dates, secular variation and tephra layer correlation. Fine tuning of the core-top ages relies on the definition of the activity depth of 210 Pb shortlived radionuclide [18]. The chronology of the reference core RF93-30 is described in Oldfield et al. [7]. Full details about the age-depth model adopted for the cores are reported in Piva et al. [5].

5

New biostratigraphic results from the Adriatic reg

G. sacculifer presents a consistent distribution in all the cores (Figure 2), although its abundance is inversely related to the water depth, because planktic foraminifers increase in abundance with the increase of the water column [19, 20]. Core AMC991 shows three major long-term increases of G. sacculifer at 5200-4000, 3800-2400 and 2100- 600 years BP, respectively [5] (Figure 2, dashed line for cores SA039 and AMC99-1). This latter peak appears double phased in core SA03-9, because of the higher resolution consequent 320

to the higher sediment accumulation rate.

Moreover, the age of the main frequency minima of G. sacculifer in core AMC991 (2200-2400 and 3800-4100 years BP) are in good agreement with the minima in the other cores from both deeper and shallower water (SA03-9 and RF93-30) (Figure 2 dashed line). The age of the LO of G. sacculifer is about 550 years BP and it is the average among the values obtained from the chronologically best constrained cores RF93-30, AMC99-1 and SA03-9 [5].

Moreover, the LO of G. sacculifer was recognised also in the Eastern Mediterranean slope off Israel [21, 22] and, although occurring slightly earlier (between 800 and 850 years BP), still represents the planktic foraminifers’ bioevent best approximating the base of the LIA in the whole Eastern Mediterranean. The shallow cores AN97-15 and RF93-30 record two peaks of the benthic foraminifer Valvulineria complanata (Figure 2) dated at 1865AD and 1689AD according to Oldfield et al.

[7]. Although the record of V. complanata is peculiar of the inner-shelf mud-belt environment, it is significant that the peaks can be recognized over at least 300 km, showing that this benthic feature can be used for additional stratigraphic correlation at a regional scale [5].

6

Paleoenvironmental inferences

G. sacculifer is an oligotrophic, shallow water, symbiont-bearing dweller, typical of warm, tropical environments [23, 15, 24].

As this species lives only in the Western Mediterranean at the end of summer [24] the main peaks of this species during the last ca. 6000 years are inter-

Marine research at CNR

warm species C. laev. carinata warm species

0

10

20

30

0

10

20

0

C. laev. carinata 0

20 40 60

20

40

Core AN97-15 (w.d. 55.1 m)

Core RF93-30 (w.d. 77 m)

Core SA03-9 (w.d. 712.3 m) Core AMC 99-1 (w.d. 262 m) 20

warm species 40

60

80

100

ka (BP)

1

0

20 40 60

0

2

4

C2 (IrA)

C3(ABA) W4(CA)

5 6

fo cleararest nce?

W3(LBA)

4

80

st e forearanc cle

3

40

f cleaorest ranc e

C1(DACP) W2(RWP)

0

6

W1(MWP)

2

G. sacculifer V. complanata

C. laev. carinata aq. herbs + hygrophytes

LIA

0 0.5 1

0

10

20

G. sacculifer B. spathulata

0

2

4

6

G. sacculifer

8 0 20 40 60 80

B. spathulata

0 15 30 45

G. sacculifer V. complanata

0

20

40

B. spathulata

0

40

80

Arboreal Pollen %

0 5 10 15

B. spathulata

Figure 2: Major climatic oscillations during the last 6 ka: warm planktic foraminifera sum, G. sacculifer abundance, C. laevigata carinata, Brizalina spathulata, trees sum, aquatic+hygrophytes curve (Pinus excluded) (from [7]). In cores SA03-9 and AMC99-1 a 3-point average smoothed curve (dashed bold line) is plotted for G. sacculifer, in order to better highlight the major long-term increases. Grey and dotted stripes indicate warmdry intervals and the Little Ice Age, respectively. W (1-4) are warm-dry events, C (1-3) are cool-wet events. LIA=Little Ice Age; MWP=Medieval Warm Period; DACP=Dark Age Cold Period; RWP=Roman Warm Period; IrA=Iron Age; LBA=Late Bronze Age; ABA=Ancient Bronze Age; CA=Copper Age (from [5] modified). preted as indicative of relative climatic optimum with low turbidity of the water column and reduced river runoff [5]. According to the age-depth model adopted, the main peaks of G. sacculifer correspond to warm intervals recognised by archaeological studies in the regions surrounding the Mediterranean: the Medieval Warm Period, the Roman Age, the late Bronze Age and the Copper Age (Figure 2). Among the benthic foraminifers, the opportunistic species V. complanata presently lives in the mud belt environment, characterized by large availability of organic matter, mainly brought by river runoff and leading to a poorly oxygenated sea floor [25, 26, 27]. The two peaks of this species present in the shallowest cores RF93-30

and AN97-15, occur above the LO of G. sacculifer, therefore they represent the environmental conditions during the Little Ice Age (LIA). Moreover, the ages of these two peaks are consistent with the coldest phases of the LIA: Fernau (15901630 AD) and Napoleon (1810-1820 AD) [28, 29]. Consequently, high frequencies of V. complanata record two cold and humid intervals, characterized by substantially increased river discharge [5]. In the deeper-water cores the abundances of the main benthic taxa show a generally constant trend, except for minor oscillations.

In core AMC99-1 the decrease of Brizalina dilatata (Figure 2) points out an improvement of bottom floor oxygenation, while the Cassidulina laevigata carinata trend

321

Marine Geology

indicates increased bottom water salinity [30]. Moreover, in the two deeper-water cores benthic taxa do not clearly parallel G. sacculifer oscillations. Therefore, the causes responsible of the changes registered in surface waters did not impact the deeper environment. At last, additional high frequency oscillations are visible within the main trend observed for G. sacculifer, as well as for the warm species (Figure 2 solid lines for G. sacculifer in AMC99-1 and SA03-09 cores).

The minima of these high frequency oscillations parallel the pollen record of core RF93-30 [7], which shows a concurrent increase in humidity-related aquatic herbs/hygrophytes pollen (Figure 2), confirming the cool and rainy conditions inferred by the planktonic foraminifera record.

7

The main paleoceanographic steps

In Figure 3 all the δ 18 O records available and indicative of surface, intermediate and bottom water are plotted against calibrated age according to the adopted age depth model [5]. The B. marginata δ 18 O record is available for three cores from the MAD. The three curves are distributed in a quite narrow range of values up to 7500-7000 years BP, after which the isotope composition of core RF93-77 becomes lighter up to the modern time, while the two records in slightly deeper waters (PRAD1-2 and CM92-43) show substantially higher values. This “permanent separation” between the shallower and deeper records occurs close to the end of the deposition of the Sapropel 1 (ca. 7000 years) and it is interpreted to reflect an increased production 322

of the dense bottom waters forming in the Northern Adriatic [5]. According to Artegiani et al. [31, 32] the dense water, formed during winter time, moves from the North following the western coast of the Adriatic, where it is mixed and stored in the MAD becoming the Middle Adriatic Deep Water. This dense water seems to contribute, together with the inflowing LIW, to the formation of dense water in the Southern Adriatic. The modern δ 18 O composition of the dense bottom water indicates a change towards relatively higher values proceeding southward ca. from 1.00 in northernmost part of the Adriatic to 1.36-1.48 in the MAD and to higher values in the Southern Adriatic [33]. Our isotope record suggests that the shallow site (RF93-77) is not affected by the dense water flow responsible of higher oxygen values in the deepest site (CM92-43). PRAD1-2 site, showing a similar record to CM9243, would locate at the upper limit of the area impacted by the dense bottom water path (Figure 3a). The G. bulloides δ 18 O curves from the Central and Southern Adriatic show almost overlapping trends up to 5500 years BP (Figure 3b). Later, the CM92-43 record shifts towards higher values, while all records from the other cores remain constant and overlapping up to the modern time. G. bulloides is an intermediate dwelling species occurring in winter and spring in the Mediterranean [24] down to 200 m and is prolific below the thermocline. Artegiani et al. [31] and Stenni et al. [33] indicate that the Modified Levantine Intermediate Water (MLIW) enters in the Southern Adriatic basin on the eastern side during spring and the shallowest component of this water mass reaches the Middle and Northern Adriatic. Stenni et al. [33] describe this water mass as characterized by an oxygen isotope composition

Marine research at CNR

a) B. marginata (benthos) 0

(kyrs BP)

cal. age 1 2

3.5

3.0

2.5

2.0

b) G. bulloides (plankton) 1.5 ‰ 2.5

2.0

1.5

2.0

1.5

1.0

0.5

0.0 -0.5 ‰

d) 0

G. sacculifer flux core AMC99-1 4 12 8

LIA W1(MWP) C1(DACP) W2(RWP)

3

C2 (IrA)

4

W3(LBA) C3(ABA)

5

c) G. ruber (plankton)

1.0 ‰

W4(CA)

6 7 8 9 10 11

n° spec. cm-2 100 year -1

CM92-43, w.d. 252 m PRAD1-2, w.d. 186 m RF93-77, w.d. 152 m SA03-09, w.d. 712 m SA03-11, w.d. 1126 m

Figure 3: Major paleoceanographic turnovers in the Adriatic basin during the Holocene.

Stripes mark cold-wet intervals. Grey circles mark critical time intervals of the main δ 18 O events: a) NAdDW dense bottom water achieve the modern route some 7000 years BP; b) Modified LIW intrude the Central Adriatic around 5500 years BP and c) Closelyspaced cool/wet events during the last 5500 years (from [5] modified). heavier than the surrounding surface and bottom waters. Therefore, the shift to relatively higher values in CM92-43 records the intrusion of MLIW in the Central Adriatic around 5000-5500 years BP, creating a circulation pattern similar to the modern one. The occurrence of higher values in CM92-43 does not mark a decrease in temperature, because they correspond to warm intervals (G. sacculifer peaks). The surface-water record of G. ruber for the Central Adriatic shows overlapping trends up to 5500 years BP, regardless of major depth or distance from the coast (Figure 3c). After 5500 years the core in deeper water and more distal location shifts towards higher values compared to the proximal core. Moreover, the more distal site becomes less affected by fluvial runoff, reflecting the increased distance of the river mouths at the end of the sea level rise [34].

The oscillations towards lower isotope val-

ues during the last 5500 years in CM9243 mark short intervals of increased rainfall, which match the minima in G. sacculifer (Figure 3d and dotted stripes in Figure 3). In the upper part of the cores, the interval corresponding to the LIA is characterized by a heavier isotope composition (in particular for core RF93-77), that reflects primarily the temperature decrease rather than a salinity change, in contrast with all the preceding wet/cool intervals, as suggested by the pollen record and by the sum of warm planktic species (Figure 2). The Adriatic stratigraphic record can be compared with several other climatic records based on different proxies in several locations worldwide suggesting that an atmospheric connection is the most likely link among these distant areas. A general correspondence between the observed climate oscillations and recognised archaeological intervals confirms the major role

323

Marine Geology

exerted by climate change in determining rises and declines of civilizations. Oldfield et al. [7] recognised the impact of an anthropogenic signal in proximal (inner shelf) cores. The abundance of tree pollen from Oldfield et al. [7]) in core RF93-30 and the derived intervals of forest clearance around 3600, 2400 and 700 years BP are reported in Figure 2, as the best expression of a possible anthropogenic impact on sedimentation. Human impact during these intervals, however, probably affects only relatively proximal environments and does not extend in deeper-water cores AMC99-1, in the MAD slope basin, and SA03-9, in the South Adriatic slope.

In these distal contexts the surface-water planktic foraminifers display subtle but coherent oscillations in relative abundances, which appear to match the climate variability on super-regional extents.

8

Acknowledgements

This study was supported by the ECEURODELTA, EC-EUROSTRATAFORM and EC-PROMESS1 projects. Nils Andersen performed the isotope analysis and to Anna Maria Mercuri helped on the interpretation of pollen records of core RF9330. This is the ISMAR-Bologna (CNR) contribution n. 1697.

References [1] G. Bond, W. Showers, M. Cheseby, R. Lotti, P. Almasi, P. deMenocal, P. Priore, H. Cullen, I. Hajdas, and G. Bonani. A pervasive millennial-scale cycle in North Atlantic Holocene and Glacial Climates. Science, 278:1257–1266, 1997.

[2] G. G. Bianchi and I. N. McCave. Holocene periodicity in North Atlantic climate and deep-ocean flow south of Iceland. Nature, pages 515–517, 1999.

[3] G. Bond, B. Kromer, J. Beer, R. Muscheler, M. N. Evans, W. Showers, S. Hoffmann, R. Lotti-Bond, I. Hajdas, and G. Bonani. Persistent solar influence on North Atlantic climate during the Holocene. Nature, 294:2130–2136, 2001.

[4] P. deMenocal, J.. Ortiz, T.. Guilderson, and M. Sarnthein. Coherent high- and lowlatitude climate variability during The Holocene warm period. Science, 288:2198– 2202, 2000.

[5] A. Piva, A. Asioli, F. Trincardi, R. R. Schneider, and L. Vigliotti. Late-Holocene climate variability in the Adriatic Sea (Central Mediterranean). The Holocene, 18(1):153–167, 2008.

[6] C. M. Lee, F. Askari, J. Book, S. Carniel, B. Cushman-Roisin, C. Dorman, J. Doyle, P. Flament, C. K. Harris, B. H. Jones, M. Kuzmic, P. Martin, A. Ogston, M. Orlic, H. Perkins, P. M. Poulain, J. Pullen, Russo A., C. Sherwood, Signell. R. P., and D. Thaler. Northern Adriatic response to a wintertime bora wind event. American Geophysical Union, 86(16):157–165, 1999.

324

Marine research at CNR

[7] F. Oldfield, A. Asioli, C. A. Accorsi, A. M. Mercuri, S. Juggins, L. Langone, T. Rolph, F. Trincardi, G. Wolff, Z. Gibbs, L. Vigliotti, M. Frignani, K. van der Post, and N. Branch. A high resolution late Holocene palaeo environmental record from the central Adriatic Sea. Quaternary Science Reviews, 22:319–342, 2003.

[8] A. Cattaneo, A. Correggiari, L. Langone, and F. Trincardi. The late-Holocene Gargano subaqueous delta, Adriatic shelf: Sediment pathways and supply fluctuations. Marine Geology, 193:61–91, 2003.

[9] A. Cattaneo and A. Trincardi F.and Asioli, A.and Correggiari. Clinoform formation in the Adriatic Sea: energy-limited bottomset. Continental Shelf Research, special issue on Sediment Dynamics in the Western Adriatic. (Tim G. Milligan and Antonio Cattaneo guest editors), 27(3-4):506–525, 2007.

[10] J. J. Lowe, S. Blockley, F. Trincardi, A. Asioli, A. Cattaneo, I. P. Matthews, A. M. Pollard, and S. Wulf. Age modelling of late Quaternary marine sequences in the Adriatic: towards improved precision and accuracy using volcanic event stratigraphy. Continental Shelf Research, 27(3-4):560–582, 2007.

[11] A. Correggiari, F. Trincardi, L. Langone, and M. Roveri. Styles of failure in heavily sedimented highstand prodelta wedges on the Adriatic shelf. Journal of Sedimentary Research, 71(2):218–236, 2001.

[12] A. Correggiari, A. Cattaneo, and F. Trincardi. Depositional patterns in the LateHolocene Po delta system. In: Bhattacharya, J. P., Giosan, L. (Eds.), River Deltas: Concepts, Models and Examples. SEPM Special Publication, 83:365–392, 2005.

[13] A. Asioli, F. Trincardi, J. J. Lowe, and F. Oldfield. Rapid Communication - Shortterm climate changes during the Last Glacial–Holocene transition: comparison between Mediterranean records and the GRIP event stratigraphy. Journal of Quaternary Science, 14:373–381, 1999.

[14] A. Asioli, F. Trincardi, D. Lowe, J. J.and Ariztegui, L. Langone, and F. Oldfield. Sub-millennial scale climatic oscillations in the central Adriatic during the Lateglacial: palaeoceanographic implications. Quaternary Science Reviews, 20:1201–1221, 2001.

[15] C. Hemleben, M. Spindler, and O. R. Anderson. Modern planktic foraminifera.

Springer-Verlag, New York (Eds.), 1989.

[16] M. Stuiver and P. J. Reimer. Extended 1 4C data base and revised CALIB 3.0 1 4C age calibration program. Radiocarbon, 35:215–230, 1993.

[17] L. Vigliotti, K. L. Verosub, A. Asioli, A. Cattaneo, A. Piva, and F. Trincardi. Late Pliocene to Early Pleistocene Mediterranean mud-dwelling foraminifera; influence of changing environment on community structure and evolution. The Holocene, 18(1):141–152, 2008.

325

Marine Geology

[18] M. Frignani, L. Langone, M. Ravaioli, D. Sorgente, F. Alvisi, and S. Albertazzi.

Fine-sediment mass balance in the western Adriatic continental shelf over a century time scale. Marine Geology, 222-223:113–133, 2005.

[19] G. J. Van der Zwaan and F. J.and de Stigter H. C. Jorissen. The depth dependency of planktonic/benthic foraminiferal ratio: Constraints and applications. Marine Geology, 95:1–16, 1990.

[20] B. K. Sen Gupta. Modern Foraminifera. Kluwer Academic Publishers, 1999.

[21] B. Schilman, M. Bar-Matthews, A. Almogi-Labin, and B. Luz. Global climate instability reflected by eastern Mediterranean marine records during the late Holocene. Palaeogeography, Palaeoclimatology, Palaeoecology, 176:157–176, 2001.

[22] B. Schilman, A. Almogi-Labin, M. Bar-Matthews, and B. Luz. Late Holocene productivity and hydrographic variability in the eastern Mediterranean inferred from benthic foraminiferal stable isotopes. Palaeogeography doi:10.1029/2002PA000813, 18(3):1064, 2003.

[23] A.W.H. B´e and D. S. Tolderlund. Distribution and ecology of living planktonic Foraminifera in surface waters of the Atlantic and Indian Oceans. In: Funnell, B. M., Riedel, W. R. (Eds.), The Micropaleontology of the Oceans, pages 105–149, 1971.

[24] C. Pujol and C. Vergnaud Grazzini. Distribution patterns of live planktic foraminifera as related to regional hydrography and productive systems of the Mediterranean sea. Marine Micropaleontology, 25:187–217, 1995.

[25] F. J. Jorissen. The distribution of benthic foraminifera in the Adriatic Sea. Marine Micropaleontology, 12:21–48, 1987.

[26] Van der Zwaan G. J. and F. J. Jorissen. Biofacial patterns in river-induced shelf anoxia. In: Modern and Ancient Continental Shelf Anoxia, R.V. Tyson and T.H. Pearson, eds., Geological Society Special Publication, 58:65–82, 1991.

[27] J. W Murray. Ecology and Applications of Benthic Foraminifera. Cambridge University Press, 2006.

[28] H. H. Lamb. Weather and climate patterns of the Little Ice Age. In: Oeschger, H., Messerli, B. and Svilar, M.(eds.): Das Klima, Springer-Verlag, Berlin, 12:149–160, 1980.

[29] A. Veggiani. Le fluttuazioni del clima dal XVIII al XX secolo. I cicli di Bruckner.

Bollettino della societ`a Torricelliana di Scienze e Lettere, 37:1–56, 1986.

[30] P. J. J. Verhallen. Late Pliocene to Early Pleistocene Mediterranean mud-dwelling foraminifera; influence of changing environment on community structure and evolutionr. Utrecht Micropal Bull., 40, 1991.

326

Marine research at CNR

[31] A. Artegiani, E. Bregant, D.and Paschini, N. Pinardi, F. Raicich, and A. Russo.

The Adriatic Sea general circulation. Part I: air-sea interactions and water mass structure. Journal of Physical Oceanography, 27:1492–1514, 1997.

[32] A. Lascaratos, W. Roether, K. Nittis, and B. Klein. Recent changes in deep water formation and spreading in the eastern Mediterranean sea: a review. Progress in Oceanography, 44:5–36, 1999.

[33] P. Stenni, B.and Nichetto, P. Bregant, D.and Scarazzato, and A. Longinelli. The 1 8O signal of the northward flow of Mediterranean waters in the Adriatic Sea. Oceanologica Acta, 18(3):319–328, 1995.

[34] F. Trincardi, A. Asioli, A. Cattaneo, A. Correggiari, and L. Langone. Stratigraphy of the late-Quaternary deposits in the Central Adriatic basin and the record of short-term climatic events. Memorie dell’Istituto Italiano di Idrobiologia, 55:39– 70, 1996.

327

Marine Geology

328

Paleoceanographic Evolution of the Central Adriatic During the Last Four Glacial-Interglacial Cycles (Promess1 borehole PRAD1-2) A. Piva1 , A. Asioli2 , N. Andersen3 , J.O. Grimalt4 , R.R. Schneider5 , F. Trincardi6 1, ENI E&P Division, Sedimentology, Petrography & Stratigraphy Dpt., S. Donato Milanese, Italy 2, Institute of Geosciences and Earth Resources, CNR, Padova, Italy 3, Leibniz Laboratory for Radiometric Dating and Stable Isotope Research, CAU, Kiel, Germany 4, Department of Environmental Chemistry, Institute of Chemical and Environmental Research (IIQAB-CSIC) Girona, Spain 5, Institute of Geosciences, Christian-Albrechts-Universit`at zu Kiel, Kiel, Germany 6, Institute of Marine Sciences, CNR, Bologna, Italy andrea.piva@eni.com Abstract The paleoenvironmental history of the central Adriatic basin is here reconstructed for the last 360 ka BP, based on an integrated approach (planktic and benthic foraminifera, alkenone SST and O and C stable isotope records). There seems to be a general inphase trend in the paleoclimatic changes between the central Adriatic and the north Atlantic climate system, except for the intervals related to the deposition of the sapropel layers in the eastern Mediterranean; in particular, the time period between MIS7.5 and MIS5 results to be strongly influenced by the monsoonal regime. Compared to other Mediterranean records, it can be inferred that the Adriatic was affected by very low SST during glacial times (down to 2°C for MIS2), which is uncommon for the Mediterranean basin. The SST record points out that the Adriatic was not capable to maintain interglacial/interstadial conditions for a duration similar to the western Mediterranean. The landlocked position of this shallow basin, in fact, makes it particularly sensitive to factors such as the strong exposure to atmospheric forcing (e.g.

Siberian High), and the strong influence of the nearby land mass, producing a lag in the demise of glacial intervals. Moreover, the progressively higher values of the δ 18 O records of the glacial intervals, consistently with the SST record and the foraminifera assemblage, imply an increasing impact of the formation of cold and dense water in more recent times.

1

Introduction

ing the Quaternary, between intervals more dominated by the high-latitude obliquity (41 ka)-driven climate system or by the The Mediterranean Sea is a mid-latitude, lower-latitude North African climate sysland-locked marginal basin, switched, dur-

Marine Geology

tem, tightly linked to the precession cycle (21 ka). To address the stability of the Mediterranean climatic scenario, high-resolution paleoceanographic studies have been carried out, unravelling records retrieved from broad shelves and upper slopes, where thick sedimentary sequences deposited during the last 500 ka. This interval encompasses several orders of cyclicity (100, 41, and 21 ka) that are characteristic of past Quaternary climate regimes, with the chance of identifying centennial to millennial-scale episodes of abrupt climate change.

Borehole PRAD1-2 is the first continuous and almost undisturbed marine record spanning the last 370 ka retrieved in the Adriatic Basin [1, 2], in order to study the paleoenvironmental changes of the last four glacial-interglacial cycles in a key area for the oceanographic setting of the whole Mediterranean.

Even if the borehole was drilled in a shallow-water and proximal location (186 m water depth), Piva et al. [1] provided a multi-proxy high-resolution integrated stratigraphy for PRAD1-2, documenting a robust chronologic correlation with other oceanic records within and outside the Mediterranean.

This paper aims at reviewing the climatic trends of glacial and interglacial intervals during the last 370 ka in a high resolution Adriatic setting, integrating independent proxies, such as planktic and benthic foraminifera assemblages, foraminifera-derived O and C stable isotope composition, and alkenone-derived SST records.

Comparison with other Mediterranean records provides information for the recognition of the role this small basin played in the past for cold and dense water production.

330

2

Materials and Methods and Chronological Framework

Borehole PRAD1-2 yielded a continuous, 71.2 m long sediment sequence, collected on the western slope of the Mid-Adriatic Deep (Figure 1). The methods adopted for each proxy considered in the paper followed a standard procedure and are fully described by Piva et al. [2]; the reader is referred to Piva et al. [1] for a complete description of all the other proxies taken into account in the integrated stratigraphy.

A comprehensive description of the age model for PRAD1-2 was provided by Piva et al. [1]. The borehole was ascribed to the last 370 ka, ranging from MIS11.1 to MIS1, by means of an integrated approach, based on Oxygen stable isotope stratigraphy (both on planktic foraminifer Globigerina bulloides and on benthic Bulimina marginata), calcareous nannoplankton biostratigraphy, foraminifera bioevents, magnetostratigraphy, radiocarbon dates, sapropel stratigraphy, and the recognition of Dansgaard-Oeschger events.

3

PRAD1-2 General Climate Trends

PRAD1-2 interglacial and interstadial intervals are characterized by peaks of warmwater planktic foraminifera species, (Figure 2) paralleled by significant shifts in the oxygen stable isotope and alkenonederived SST curves. MIS5.5 is the warmest substage of the entire record (about 22째C).

Similarly high SST values are found in interglacials MIS7 and 9. Most glacialinterglacial transitions exhibit rapid and large SST increases, as for the 19.5째C

Marine research at CNR

10o

0o

10o

20o

40o

Pelagosa sill MD01-2443 MD01-2444

35o

Site 977A PRAD1-2 borehole reference cores

KC01-B

Figure 1: Location of borehole PRAD1-2 (star) and of the core records (circles) discussed in the text. warming during Termination II. An overall cooling trend characterizes both MIS9 and MIS5, while MIS7 shows the opposite. Substages 7.3 and 7.1 are warmer than MIS7.5, based on both SST values and the planktic assemblage composition. This succession of climatic changes matches with the integrated sea-land records (marine and pollen data) described by Roucroux et al. [3], and Desprat et al. [4, 5] on the western Iberian margin. These authors suggested an insolation maximum during MIS7.3 similar to the one in MIS7.5 and a mild stadial of MIS7.2, the latter characterized by reduced ice caps compared to the other stadials [6], and by only a slight decrease of Atlantic sea-surface temperatures [7].

The planktic foraminifera assemblage indicates dominant oligotrophic conditions in the surface water during warm intervals. Higher productivity conditions, either related to the development of a Deep Chlorophyll Maximum or concentrated in the uppermost water column, mainly corre-

spond to the deposition of sapropel equivalent layers. Globorotalia inflata, a winter deep dweller species requiring vertical mixing and a cool and homogeneous water column [8], is generally present during warm substages and also at the onset and/or at the end of cold substages (5.4 and 7.4, Figure 2), but not during the deposition of sapropel equivalents, confirming the strong stratification of the water mass during these events. Therefore, when present, G. inflata can be considered an indicator of deep water production, assuming that during past interglacials and interstadials the northern Adriatic deep-water formation occurred through mechanisms similar to the modern interglacial [9].

The difference between planktic and benthic foraminifera oxygen isotope values (∆18 O B. marginata-δ 18 O G. bulloides) is an indicator of the water mass homogeneity. When ∆values approach zero the water mass tends to become more uniform, while increasing δ values reflect a stronger stratification of the water column. Consistently,

331

0 m 4

Foram + nannopl. bioevents

st M a ra g tig ne ra to ph y

Li

th ol o

gy

Marine Geology

LO G. inflata

Control points

cal age (ka BP)

modern time LO G. i.

S1 eq. top Y. D.

0 6 8.5 12

C AMS C AMS

17.3 19.2

14

C AMS

27.7

LCO G. i. in MIS3

43.1

14

16 20

14

H. balthica common LCO G. inflata in MIS3

H. balthica I. islandica E. excavatum f. clavata

40 44

S1 eq.

2

LGM Chronozone

3

D-O events

4

28

36

1

S. sellii

24

32

5.5 4.5 3.5 2.5 1.5 0.5 -0.5 ‰

LO S. sellii

8 12

δ18O vs VPDB G. bulloides δ18O vs VPDB B. marginata

reversal G. caribbeanicasmall Gephyrocapsa

S3 eq.

MIS5.2

81 91

S4 eq.

MIS5.4 S5 eq.

T II MIS6.2 MIS6.4 S6 eq.

IBE MIS7.0 S7 eq.

MIS7.2 S8 eq.

MIS7.4 S9 eq.

T III

101 111 124 130 135 152.5 172 188 189.5 195 200.5 216 225 239 243

FO E. h.

S' eq.

264 288

5.1 5.2

S3 eq.

5.3

S4 eq.

5.4

5.5

6.2

S5 eq.

6.3

6.4

6.5

7.0 7.2

7.1 7.3

7.4

7.5

S6 eq.

S7 eq.

S8 eq.

S9 eq.

48 FO E. huxleyi 52 56 60 64 68 71.2

homogenuos mud silty mud silty laminae sand tephra layer erosional surface diffused bioturbation burrow organic matter biosome bioclast

S10 eq.

331

MIS10.2?

340

MIS11.1?

364

8.5 9.1

9.2

S’ eq.

9.3

S10 eq.

10.2?

11.1?

Figure 2: PRAD1-2 stratigraphic framework. Grey arrows indicate the control points of Dansgaard-Oeschger events (see [1] for details).

the ∆18 O curve for PRAD1-2 shows the highest values (up to 2.5 ) during the deposition of the Adriatic sapropel equivalent layers, while minima in ∆18 O are mainly recorded during cold intervals. G. inflata is absent during the acme of glacial intervals (MIS10.2, 8.4, 6), when ∆18 O minima are extreme and the water depth was too shallow to allow the appropriate depth habitat of this intermediate water dweller (Figure 2). Excluding the sapropel-equivalent in-

332

tervals, the benthic assemblage indicates an upper slope mesotrophic environment during all interglacials and interstadials, characterized by the accumulation of organic matter on the seafloor and relatively low oxygenation. However, in this mesotrophic environment a small amount of epifaunal species is indicative of well oxygenated bottom waters. This occurrence therefore suggests seasonal bottom ventilation, comparable with modern conditions, character-

Marine research at CNR

ized by winter production of oxygenated, dense water in the north Adriatic.

Glacial stages show distinctive features both in terms of paleodepth and climate trends.

MIS10 shows the shallowest glacial sea level of the whole borehole, as testified by the near-absence of planktic foraminifera and by the high percentage of Elphidium + Ammonia, reflecting a very shallow environment. In contrast, MIS4 is characterized by the deepest glacial basin conditions.

4

Comparison with other records: western Mediterranean

We compare PRAD1-2 δ 18 O G. bulloides and alkenone-derived SST records to the time equivalent succession from ODP Site 977A (Figures 1-2), the western Mediterranean Sea record with the highest resolution for the last 250 ka [10]. The 0 Uk37 SST record reported by Martrat et al. [10] for the Alboran Sea indicates SSTs quite similar to those reported during warm substages in PRAD1-2 (Figure 2). Intervals of abrupt warming are detected in PRAD1-2 by concurrent peaks of δ 18 O, SST and warm planktic foraminifera (Figure 2), in particular during MIS9.3 (12°C shift), 9.1 (7°C), 7.5 (13°C), 5.5 (18.5°C) and 1 (8°C), confirming the conclusion made by Martrat et al. [10] that cold stadials had only limited duration, immediately followed by well-defined returns to interstadials with accelerated warming by positive feedback mechanisms once a threshold was passed. However, PRAD12 reveals several exceptions, with regard to the exact timing of the highest SST values: maximum warming is differently

0

recorded by Uk37 SST or warm planktic foraminifera frequency. Frequency peaks of warm planktic species and minima of δ 18 O values are not in phase with coeval 0 peaks of Uk37 SST. These phase lags seem to reflect decrease in productivity before or 0 after the maximum Uk37 SST, producing a poorer total flux of planktic foraminifera assemblage but a relative increase in warm species. All these observed trends lead to the conclusion that the Adriatic basin is not capable to maintain interglacial and interstadial conditions with a duration similar to the western Mediterranean and eastern Atlantic [10, 11], as suggested by three observations: (1) during MIS7.3, 7.1, 5.3 and 5.1, the decreasing SST trend toward cold substages starts earlier in the Adriatic Basin than in the Alboran Sea; (2) similarly, the subsequent SST increase in warm substages is slower and delayed in PRAD12, resulting in prolonged intervals with low SST and (3) in PRAD1-2 the maximum 0 Uk37 SST within MIS7.3 is achieved later (Figure 2). The overall very low SST, uncommon for the Mediterranean Sea, and the shorter duration of warm intervals documented for the Adriatic, may be explained by three interacting causes, which are here listed on the basis of their inferred relative importance: (1) the landlocked nature of this shallow basin, especially during the glacial stages, when sea level was more than 100 m lower than at present, probably amplified the SST excursion and increased the atmospheric forcing, e.g., through outbreaks of northerly polar continental air masses (Siberian High), as already argued by Rohling et al. [12] for the Holocene cold oscillations; (2) the vicinity of the basin to large Alpine and Apennine glaciers, conveying cold air, an uncommon condition for the Mediterranean and (3) the location at a latitude at least 3° higher 333

Marine Geology

than the best documented Mediterranean sites. Glacial intervals like MIS10, 8.4, 6.2 and 2 experienced the lowest temperatures (2-4°C) and MIS4 was just slightly warmer (5°C). These central Adriatic SST are typically 3-4°C lower than at the Alboran site [10] both during glacial and stadial intervals. Greatest SST differences (up to 5-7°C) between the two sites are recorded during MIS6.2–6.3, MIS4 and MIS2. The comparison of the SST record between PRAD1-2 and the Iberian Margin [11], Figures 1, 2) for the last four climate cycles confirms differences as recognized between the Alboran and Adriatic Sites also for MIS8 as well as for the interstadials MIS9.1 and 9.3.

5

Comparison with other records: eastern Mediterranean

MIS5.1, 5.3, 7.1, 7.5 and 9.3, suggesting an enhanced seasonal contrast during these interstadials. Consequently, during MIS5.5 and MIS7.3, a prolonged summer warm season and a weaker winter cooling are inferred. It seems that the climatic evolution of the two areas particularly matched within the time interval from MIS7.3 to MIS5.3. This interpretation is supported by the δ 18 O data (Figure 2), indicating a higher intensity of sapropelequivalent events from S8 to S4, and reinforced by the presence of deep infaunal taxa only during these sapropel equivalent beds (ODS curve), suggesting that in the time interval between S8 and S4 the conditions of the central Adriatic were more similar to those of the eastern Mediterranean, at least during the sapropel deposition. The stronger surface water dilution along with the relatively large thickness of PRAD1-2 sapropel-equivalent intervals S8 to S4 (typically 50 to 100 cm each) suggest enhanced precipitation over the central Adriatic brought about by a stronger influence of the monsoon system over the Mediterranean. The Adriatic and Ionian microfaunistic records differ more significantly during glacial intervals than during interglacials, but this fact is mainly a consequence of the shallow depth of the central Adriatic, where sea level falls hampered the intrusion of deep-dwelling planktic taxa, leaving a planktic association dominated by shallow- and intermediatewater dwellers.

Sanvoisin et al. [13] analyzed the oscillations in the planktic foraminifera assemblage in the Ionian Basin. Core KC01-B was retrieved in 3643 m water depth (Figure 1), and spans the last circa 1.1 Ma BP. The correlation between PRAD1-2 record and the last circa 340 ka BP of the Ionian core allowed the identification of major similarities and differences between the two basins, reflecting local processes in the Adriatic region. Warm planktic species in core KC01-B allow, despite the lower stratigraphic resolution, recognition of all the major warm oscillations documented in the more detailed PRAD1-2 record, in par6 Bottom Water Formaticular for MIS5 and MIS7, with MIS5.3, tion 5.5, 7.1 and 7.3 as the warmest intervals.

Moreover, planktic deep-dweller species, requiring a well developed winter verti- Sanvoisin et al. [13] analyzed the oscilcal mixing, peak in both records during lations in the planktic foraminifera assem334

Marine research at CNR

δ18O G. bulloides B. marginata 0

5.5

200 250

0

20 40 60 80

5.5 3.5 1.5 -0.5 ‰

0

5

10 15 20 25 S1 eq.

4 5.1 5.2 5.3 5.4 5.5 6.2 6.3 6.4 6.5 7.0 7.1 7.2 7.3 7.4 7.5

S3 eq.

S4 eq.

S5 eq.

S6 eq.

S7 eq.

S8 eq.

S9 eq.

0

5

10 15 20 25

Alkenone SST (°C)

8.5

300

Alkenone SST (°C)

δ18O G. bulloides

3

50

150

warm pl. spec. %

1 2

ka

100

1.5 -0.5‰ 0 0.5 1 1.5 2 2.5 ‰

3.5

ADRIATIC vs ADRIATIC vs ALBORAN WEST IBERIAN MARGIN

ADRIATIC

∆(δ18O) (B. marg.- G. bull.)

9.1 9.2 9.3

S10 eq.

10.2?

350

11.1?

S’ eq.

PRAD1-2

more stratification

ODP Site 977A (after Martrat et al., 2004)

400 0

0.02 0.04 0.06

Eccentricity

0 5 10 15 20 25

0

Alkenone SST (°C)

10

20

30

G. inflata %

MD01-2443 + MD01-2444 (after Martrat et al., 2007)

Figure 3: Synthesis of the main proxies of borehole PRAD1-2 against eccentricity (Be91). The alkenone-derived SST parallels significantly the δ 18 O records. On the right, comparison between δ 18 O G. bulloides records of the last 250 ka of the ODP site 977A 0 and PRAD1-2 records; comparison between Uk37 SST records of the ODP site 977A and PRAD1-2. ODP 977A records are plotted according to the age-depth model by Martrat 0 et al. [10]; comparison between Uk37 SST records of the composite Western Iberian site

by Martrat et al. [11] and of PRAD1-2. Grey areas indicate the sapropel equivalent layers detected in PRAD1-2. blage in the Ionian Basin. Core KC01-B was retrieved in 3643 m water depth (Figure 1), and spans the last circa 1.1 Ma BP. The correlation between PRAD1-2 record and the last circa 340 ka BP of the Ionian core allowed the identification of major similarities and differences between the two basins, reflecting local processes in the Adriatic region. Warm planktic species in core KC01-B allow, despite the lower stratigraphic resolution, recognition of all the major warm oscillations documented in the more detailed PRAD1-2 record, in particular for MIS5 and MIS7, with MIS5.3, 5.5, 7.1 and 7.3 as the warmest intervals.

Moreover, planktic deep-dweller species,

requiring a well developed winter vertical mixing, peak in both records during MIS5.1, 5.3, 7.1, 7.5 and 9.3, suggesting an enhanced seasonal contrast during these interstadials. Consequently, during MIS5.5 and MIS7.3, a prolonged summer warm season and a weaker winter cooling are inferred. It seems that the climatic evolution of the two areas particularly matched within the time interval from MIS7.3 to MIS5.3. This interpretation is supported by the δ 18 O data (Figure 2), indicating a higher intensity of sapropelequivalent events from S8 to S4, and reinforced by the presence of deep infaunal taxa only during these sapropel equivalent

335

Marine Geology

δ18O G. bulloides

δ13C B. marginata

5.5 4.5 3.5 2.5 1.5 0.5 -0.5 ‰

-1.5 -1.0

-0.5

0.0

0.5

1.0 ‰

0

enized water

100

150

200

intensified sapropelic conditions in the Adriatic

co more lder, saltier, oxyge nated water

50

more homog

ka

250

0 0.5 1 1.5 2 2.5 ‰

0

∆(δ18O) (B. marginata-G. bulloides)

20

40

60

80 %

C. laevigata-carinata H. balthica

Figure 4: PRAD1-2 integrated proxies showing the overall trend for the last 250 ka. Light grey stripes mark glacial and stadial intervals. The dotted area indicates the interval when the central Adriatic was more influenced by the monsoon regime. beds (ODS curve), suggesting that in the time interval between S8 and S4 the conditions of the central Adriatic were more similar to those of the eastern Mediterranean, at least during the sapropel deposition. The stronger surface water dilution along with the relatively large thickness of PRAD1-2 sapropel-equivalent intervals S8 to S4 (typically 50 to 100 cm each) suggest enhanced precipitation over the central Adriatic brought about by a stronger influence of the monsoon system over the Mediterranean. The Adriatic and Ionian microfaunistic records differ more significantly during glacial intervals than during interglacials, but this fact is mainly a consequence of the shallow depth of the central Adriatic, where sea level falls hampered the intrusion of deep-dwelling

336

planktic taxa, leaving a planktic association dominated by shallow- and intermediatewater dwellers.

7

Conclusions

The analysis of PRAD1-2 multiproxy record provides new evidence for paleoenvironmental trends that appear consistent with those typical for western and eastern Mediterranean basins, apart from some peculiar characteristics like a higheramplitude temperature excursion during major Terminations (up to 19.5°C during T II) and minor climate transitions, and the deepening trend of glacial intervals from MIS10 to MIS4. Three general conclusions can be drawn from the analysis of the

Marine research at CNR

results of PRAD1-2 and their comparison to other paleoceanographic records in the Mediterranean and north Atlantic.

of the surrounding landmass, including the occurrence of glaciers on its western side, when the sea level was more than 100 m lower than at present dur1. The central Adriatic reflects paleocliing glacial intervals, resulting in a lag matic changes during the last 370 ka in the demise of glacial conditions, and that appear in phase with the north At(3) a higher-latitude position compared lantic climate system, except for the to other Mediterranean sites. interval between MIS7.5 and MIS5.3 3. During the modern interglacial the when the Adriatic basin was more influAdriatic basin is a site for dense deeper enced by the monsoon regime. water formation and was so also dur2. The Adriatic Basin does not seem caing past interglacials, contributing to the pable of maintaining interglacial and inventilation of the deep Mediterranean terstadial warm conditions over an interSea, except when sapropelic conditions val comparable to that reconstructed in became established. the western Mediterranean. The reasons are probably the landlocked position of this shallow basin and the response to 8 Acknowledgements other factors such as (1) a greater exposition to atmospheric forcing, partic- This study was supported by ECularly through northerly polar continen- PROMESS 1 project. This is ISMARtal air outbreaks, (2) a higher influence Bologna (CNR) contribution no. 1698

References [1] A. Piva, A. Asioli, R.R. Schneider, F. Trincardi, N. Andersen, E. ColmeneroHidalgo, B. Dennielou, J.A. Flores, and L. Vigliotti. Climatic cycles as expressed in sediments of the PROMESS1 borehole PRAD1-2, central Adriatic, for the last 370 ka: 1. Integrated stratigraphy. Geochemistry, Geophysics, Geosystems, 9, 2008a.

[2] A. Piva, A. Asioli, N. Andersen, J.O. Grimalt, R.R. Schneider, and F. Trincardi.

Climatic cycles as expressed in sediments of the PROMESS1 borehole PRAD1-2, central Adriatic, for the last 370 ka: 2. Paleoenvironmental evolution. Geochemistry, Geophysics, Geosystems, 9, 2008b.

[3] K.H. Roucroux, P.C. Tzedakis, L. de Abreu, and N.J. Shackleton. Fine-tuning the land ocean correlation for the late middle Pleistocene of southern Europe. In. Sirocko, F., M. Clausen, M.F. Sanchez Goni, and T. Litt (Eds.): “The climate of past interglacials. Developments in Quaternary Sciences, 7:359–373, 2007.

[4] S. Desprat, M.F. Sanchez Goni, J.L. Turon, J. Duprat, B. Malaize, and J.P. Peypouquet. Climatic variability of Marine Isotope Stage 7: direct land–sea–ice correlation from a multiproxy analysis of a north-western Iberian margin deep-sea core. Quaternary Science Reviews, 25:1010–1026, 2006.

337

Marine Geology

[5] S. Desprat, M.F. Sanchez Goni, F. Naughton, J.L. Turon, J. Duprat, B. Malaize, E. Cortijo, and J.P. Peypouquet. Climatic variability of the last five isotopic interglacials: direct land-sea-ice correlation from the multiproxy analysis of northwestern Iberian margin deep-sea cores. In Sirocko, F., Clausen, M., M. F. Sanchez Goni, and T. Litt (Eds.): “The climate of past interglacials. Developments in Quaternary Science, 7:375–386, 2007.

[6] N.J. Shackleton. The 100,000-Year Ice-Age Cycle Identified and Found to Lag Temperature, Carbon Dioxide, and Orbital Eccentricity. Science, 289(2):1897– 1902, 2000.

[7] J. McManus, D.W. Oppo, and J.L. Cullen. A 0.5-million-year record of millennialscale climate variability in the North Atlantic. Science, 283:971–975, 1999.

[8] C. Pujol and C. Vergnaud Grazzini. Distribution patterns of live planktic foraminifera as related to regional hydrography and productive systems of the Mediterranean sea. Marine Micropaleontology, 25:187–217, 1995.

[9] A. Lascaratos, W. Roether, K. Nittis, and B. Klein. Recent changes in deep water formation and spreading in the eastern Mediterranean sea: a review. Progress in Oceanography, 44:5–36, 1999.

[10] B. Martrat, J.O. Grimalt, C. Lopez-Martinez, I. Cacho, F.J. Sierro, J.A. Flores, R. Zahn, M. Canals, J.H. Curtis, and D.A. Hodell. Abrupt Temperature changes in the Western Mediterranean over the Past 250,000 years. Science, 306:1762–1765, 2004.

[11] B. Martrat, J.O. Grimalt, N.J. Shackleton, L. de Abreu, M.A. Hutterli, and T.F. Stocker. Four climate cycles of recurring deep and surface water destabilizations on the Iberian Margin. Science, 317(5837):502–507, 2007.

[12] E.J. Rohling, P.A. Mayewski, R.H. Abu-Zied, J.S.L. Casford, and A. Hayes.

Holocene atmosphere–ocean interactions: records from Greenland and the Aegean Sea. Climate Dynamics, 18:587–593, 2002.

[13] R. Sanvoisin, S. D’Onofrio, R. Lucchi, D. Violanti, and D. Castradori. 1 Ma paleoclimatic record from the eastern Mediterranean- Marflux project: First results of a micropaleontological and sedimentological investigation of a long piston core from the Calabrian Ridge. Il Quaternario, 6(2):169–188, 1993.

338

Bottom Water Production Variability in the Ross Sea Slope During the Late Pleistocene-Holocene as Revealed by Benthic Foraminifera and Sediment Geochemistry A. Asioli1 , L. Langone2 , F. Tateo1 , F. Giglio2 , D. Ridente3 , V. Summa3 , A. Carraro5 , M.L. Giannossi4 , A. Piva6 , F. Trincardi2 1, Institute of Geosciences and Earth Resources, CNR, Padova, Italy 2, Institute of Marine Sciences, CNR, Bologna, Italy 3, Institute of Environmental Geology and Geoengineering, CNR, Roma, Italy 4, Institute of Methodologies for Environmental Analysis, CNR, Potenza, Italy 5, Department of Geosciences, University of Padova, Italy 6, ENI E&P Division, Sedimentology, Petrography & Stratigraphy Dpt., S. Donato Milanese, Italy alessandra.asioli@igg.cnr.it Abstract The Antarctic area produces bottom waters that ventilate the vast majority of the deep basins in the rest of the world ocean. The rate of formation in the source area and the strength of these cold bottom waters are key factors affecting the Global Thermohaline Circulation during modern and past climate conditions. We present the results of a multidisciplinary study carried out on a sediment core collected on the slope off the Drygalski Basin (Ross Sea) for the Late Pleistocene-Holocene. The results obtained allow the following main observations: 1) two main intervals (15-10 and 7.5-6 cal kyr BP) mark subsequent enhanced nutrient supply because of a higher efficiency in the Upper CDW upwelling; 2) within this general context, an oscillatory trend is present from 15 kyr BP to present time, indicated by the measured parameters. A possible hypothesis to interpret these oscillations is that foraminifers concentration minima, corresponding to minima in %OC and to reversal of 14 C (relative increase of older carbon) and to colder (atmospheric) condition, reflect dilution in the sediment because of rapid accumulation of fine sediment re-suspended at the shelf edge by the cascading currents. The minima may represent higher rate of bottom water formation; 3) the detected oscillations (minima) seem to correlate to colder conditions in Adelie Land record where increased sea-ice cover and bottom water formation were suggested.

1

Introduction

ters in the rest of the world ocean. The strength of the source of these cold bottom waters and their flow toward the equaThe Antarctic area produces bottom waters tor are key factors affecting Global Therthat ventilate the vast majority of deep wa-

Marine Geology

mohaline Circulation during present and past climate. Most Antarctic Bottom Water (AABW) is thought to be produced in the Weddell Sea [1, 2], while less constrained is the amount from other Antarctic areas, such as the Ross Sea. According to several authors [3], the Ross Sea contribution may be greater than presently estimated. Moreover, whether AABW production was in a steady state during the Holocene is still debated [3, 4]. The western Ross Sea is considered a formation site for a particularly salty variety of AABW [5, 2] as well as an important area of offshelf transfer of water. In detail, the export of Ross Sea shelf water onto the continental slope occurs within plumes (100-250m thick) descending at moderate angle to isobaths (35°), entraining in Lower Circumpolar Deep Water (CDW). The export is punctuated by rapid downhill cascades (60°) to greater depths. The former (moderate angle) is far more persistent and thus may be of greater significance to ocean ventilation, as a precursor of AABW [6, 7]. The results here presented were obtained within the frame of the PNRA project 4.8 “Bottom water production in the Ross Sea during the late Quaternary: a geochemical and micropaleontological study”. Among the goals of the project, the main is to detect a qualitative signal of possible changes in the rate of bottom water production during the Late Pleistocene-Holocene, on the ventilation of bottom water and, indirectly, on sea ice cover variability by integrating data on modern assemblages with sediment geochemistry (bulk mineralogy, Total Organic Carbon, biogenic silica, C and N stable isotopes, Ice Rafted Debris). To satisfy this goal a core was collected at 2377m water depth off Drygalski Basin on the slope adjacent the western continental shelf of the Ross Sea, along the pathway of bottom 340

water spreading [6, 7] (Figure 1).

2

Materials and methods

The gravity core AS05-10 has been retrieved during the XX Antarctic Italian Cruise (2005) on the slope off Cape Adare.

The core was scanned on board by means of a Bartington ring sensor for whole-core magnetic susceptibility. A provisional estimate of the time interval spanned by the entire core is proposed on the basis of the existing literature of the slope area. Ceccaroni et al. [8] studied a core (ANTA918) collected very close (Lat. 70°47’S, Long. 172°50’E, 2383 m water depth) to core AS05-10 (Figure 2). The chronology of this reference core was obtained by a 230 Thex profile, the isotope stage boundaries and ages were set according to the δ 18 O record of Martinson et al. [9] after tuning with biogenic parameter contents [8]. According to Ceccaroni et al.

[8]) and Quaia and Cespuglio [10], core ANTA91-8 spans at least the last two climate cycles. Ceccaroni et al. [8] report also the magnetic susceptibility curve of core ANTA91-8, which allows a correlation with our core AS05-10 (Figure 1, from [11]), resulting, for the MIS1 and 2, in an almost double sedimentation rate in core AS05-10 than in ANTA918. The core was frozen at -20°C to ensure the preservation of sediments for micropaleontological investigation. Indeed, the general scarcity or lack of agglutinated foraminifera in Antarctic sediments should not only be ascribed to their low preservation potential or to ecological factors alone, as also the routine use of often aggressive laboratory techniques (such as sediment desiccation or the use of peroxide) can be, at least partially, responsible. The

Marine research at CNR

Figure 1: Above: location map of the core AS05-10 (open circle) along with the published core ANTA91-8 (star). Below: 3-D map showing the path of the High Salinity Shelf Water forming in Drygalski Basin and flowing in the continental slope. Persistent plumes descend at moderate angle to isobaths (35°, up to 1 m/s speed), while rapid downhill cascades flow at ca. 90° to isobaths at 1.4 m/s speed (from [6]). sections of the core have been subsequently cut and frozen in slices ca. 1cm thick, then each slice has been divided in two halves (one for foraminifera analysis and the other for geochemical and mineralogical analysis). To get a high temporal resolution the study has been performed on all the slices, and the 182 samples of the uppermost two sections have been analysed.

Samples for foraminifers analysis were soaked in ethanol, never dried, washed with a 0.063mm sieve, split into aliquots with a wet-splitter [12], and examined with a stereomicroscope. Whole aliquots were counted until at least 300 benthic foraminifers were reached. Foraminifers were determined at specific levels and here expressed as concentration (number of specimens per 10cc). The IRD content

has been determined counting at light microscope all the grains excluding the ones of biologic origin. The counting has been performed on three fractions: larger than 2mm, between 1 and 2mm, and between 0.5 and 1mm. Samples for TOC, δ 13 C and δ 15 N were dried at 60°C, finely pounded in an agate mortar. TOC and Nitrogen contents were obtained using a FISONS NA2000 Element Analyzer after removal of the carbonate fraction in Ag capsules in 1.5N HCl. Stable isotope analyses were carried out on the same samples using a FINNIGAN Delta Plus mass spectrometer directly coupled to the Elemental Analyzer by means of a CONFLO interface for continuous flow measurements. The IAEA standards NBS-19 (+1.95 ) and N-1(0.4 ) were used as calibration material for

341

Marine Geology

C and N stable isotope analysis, respectively. Biogenic silica was determined following the progressive dissolution method of DeMaster [13, 14] and the colorimetric analysis of Strickland and Parsons[15]; 0.5 M NaOH was used as extracting agent and the uncertainty is about 10%. Total rock analysis: the semi-quantitative estimation of crystalline phases in the bulk rock was performed using an X-ray Philips diffractometer, based on the area measurements [16]; the bulk rock was grounded by hand in agate mortar and pressed with a frozen glass into an alluminium holder.

3

Chronology

Twenty-two 14 C AMS datings were performed on the bulk organic carbon at National Ocean Sciences AMS Facility (Department of Geology and Geophysics, Woods Hole Oceanographic Institution, USA). Two datings were spent to date the core top of the box-core AS05-10bc (4710±30 yr BP) and of the core AS0510 (6160±35 yr BP), to get a reservoir age for this area. To obtain calibrated ages we subtracted 450 years (average ocean reservoir) to the age of the core top AS05-10bc, then we run CALIB 5.0.2 online [19] using a =4260±30 yr. Among the twentyone datings available for the core AS0510 ten levels were discarded, including the core top. The age of the core top is surely younger than 100-150 yr BP since 210 P b analysis carried out in the topmost levels of core AS05-10 revealed the presence of excess 210 P b. Then, an age of 50 years has been ascribed to the core top of AS0510. The other nine discarded levels present age reversal. In summary, the age-depth model is based on the best fitting of twelve control points grouped in three clusters of 342

levels because of the presence of a possible condensed interval around 29 cm core depth. This level corresponds to a sharp colour change (from grey to red-brownish) and it is lithologically composed by weakly cemented silty clay (crust). The age model suggests that this short interval (cm 29.829) spans about 500-600 years.

4

Discussion

The data obtained for the last 15,000 years BP are shown in Figure 3. Although this study is still in progress, the trend of the parameters allows the following preliminary observations: 1. two main intervals around 15-10 kyr BP and 7.5-6 calib kyr BP mark a subsequent enhanced nutrient supply. Indeed, ∆15 N variations depend on the utilization degree of nitrates, which in turn can reflect productivity or nutrient supply changes. The concurrent increase of the paleoproductivity proxies OC and biogenic silica suggests that the responsible of ∆15 N major variations is an increase of the nutrient availability. The Upper CDW is a water mass rich in nutrients; therefore we interpret the above cited two time intervals as characterized by a higher efficiency in the Upper CDW upwelling ; 2. around 7.5-7kyr BP (part of the Middle Holocene Climatic Optimum, Domack et al.[20]) the IRD content drops, and it is interpreted as an evidence of the retreat of sea-ice/icebergs or a change of the iceberg path.

Within this general context, an oscillatory trend testified by all the parameters is present from 15 kyr BP to present time. The benthic foraminifera assemblage is composed only by agglutinated

Marine research at CNR

Figure 2: Whole core magnetic susceptibility correlation between core AS05-10 and reference core ANTA91-8 [8]. The 14 C AMS age in core ANTA91-8 are uncorrected (yr BP) and were performed on bulk organic carbon [10]. The grey areas mark the time interval investigated in core AS05-10 (last 15 cal kyr BP). Note that the sedimentation rate in core AS05-10 is higher (ca. twice) compared to one of the reference core ANTA91-8. foraminifera (inorganic carbon contents indicate negligible calcareous components) and it is strongly dominated by one species (Trochammina multiloculata, Figure 3).

Two hypotheses are proposed to interpret these oscillations: a) minima in foraminifera concentrations reflect relatively stronger dissolution, weaker bottom currents (testified by minima in dry density indicating higher amount of fine fraction) and lower nutrient supply (lighter values of δ 15 N ). Therefore, these intervals may reflect a lower rate of bottom water formation, considering that these latter spill out of the shelf as plumes or cascading currents ; Consequently, the intervals corresponding to maxima in foraminifers concentration should indicate better preservation, higher benthic productivity and/or better oxygenation at bottom, stronger bottom currents (maxima in dry density) and relatively higher nutrient supply reflecting a relatively higher rate of bottom water forma-

tion. b) alternatively, minima in foraminifers, corresponding to minima in %OC and to reversal of 14 C (relative increase of older carbon), reflect dilution in the sediment because of rapid accumulation of fine sediment re-suspended at the shelf edge by the cascading currents. Therefore, the minima represent higher rate of bottom water formation. We compared our records with the climatic trend proposed by Masson et al. [17] for the Ross Sea sector and represented by the D/H ratio of ice-cores. The comparison of this atmospheric signal, the chronology of which is completely independent from the one adopted for our marine core, with the oscillatory trend of core AS05-10, indicates that the foraminifers minima always correspond to colder (atmospheric) condition, at least from 0 to 12kyr. If this interpretation is correct, the higher rate of bottom water formation occurs within cold intervals. This scenario is coherent with the

343

Marine Geology

core AS05-10 (w.d. 2377 m) dry density (g/cm3) 0.6

0.8

cal kyr BP

0

0.4

5

1.0 40

δ15N (‰)

sheet silicates (%) 50

60

70

80

3

6

core MD03-2601 Adelie Land (w.d. 746 m) biogenic silica (%) 910

20

30

F. curta D/H (‰) F. kerguelensis

0

0 1 2 3 4 -0.5 0 0.5 colder

cal. age (kyr BP)

Late E8 E7 Holocene E6 Neoglacial

2

condensed interval

(colour change)

? ?

4

E5

6

Mid Holocene E4 Hypsithermal 2

8

E3

Cool

E2

Early Holocene

10

10

12 ice free 10 30 70 colder

E1 Hypsithermal 1

ψ

warmer

15 control points 14C reversal

0 5 15 10 20 25 >1mm<2mm IRD (n. grains/10cc)

0

300 600 900 0.2 0.4 0.6 0.8 -0.5 0 0.5 D/H (‰) * total n. benthic OC (%) foram. spec./10cc Ross Sea Sector total n. T. multiloculata spec./10cc

higher lateral sedimentary input from Denis et al. (2009)

* from Masson et al (2000)

Figure 3: Some of the main parameters measured for the uppermost 120cm of core AS0510 studied for the Holocene and plotted vs calibrated age. Sheets silicates, δ 15 N and OC are plotted with a 3-point average smoothed curve. The open arrows mark the control points selected for the age-depth model, while the black ones indicate age reversals.

On the right, in sequence, the climatic trend (D/H) for the Ross Sea sector calculated by Masson et al. [17] and three parameters (ratio diatomsFragilariopsis curta/Fragilariopsis kerguelensis, the focusing factor and the climatic trend for the Eastern Antarctic plateau by [17]) are shown from the core MD03-2601 collected off Wilkes-Adelie Land (modified by [18]). The grey stripes in core AS05-10 indicate the intervals corresponding to minima in foraminifers abundance and are correlated with the cold events (E1-E8) detected for the core MD03-2601 by Denis et al. [18]. record reported by Denis et al. [18] for a core (MD03-2106) retrieved on the slope off Wilkes-Adelie Land, where increased sea-ice cover (=colder conditions) corresponds to enhanced bottom water formation, on the basis of different parameters among which the diatom composition (ratio Fragilariopsis curta/Fragilariopsis kerguelensis as proxy of the sea-ice cover) and the focusing factor, as indicator of lateral sedimentary input (Figure 3). On the basis of the above inferences, we prefer the hypothesis b), although both the two hypothesis support the non-steady state rate

344

of bottom water production of AABW during the Holocene proposed by Harris et al.

[4] in a study carried out on cores and seismic stratigraphy off George V Land. At last, the condensed/hiatus interval centered at ca. 3.5-4 kyr BP does not seems to mark a major change in the general pattern of our variables, apart from biogenic silica and sheets silicates which show an increase of the oscillation amplitude. Nevertheless, this feature is coeval to the base of the Neoglacial and it is time-equivalent to the beginning of major changes in the Antarctic environment: for instance, in the circu-

Marine research at CNR

lation pattern in Antarctic Peninsula (oscil- 5 Acknowledgements lations between Upper Circumpolar Deep Water and shelf water-dominated states, This study was supported by the PNRA [20], and in glacier advance and sea-ice ex- project 4.8.

This is ISMAR-Bologna pansion off Wilkes-Adelie Land [18].

(CNR) contribution no. 1696.

References [1] T. Whitworth III, A.H. Orsi, S.J. Kim, W.D. Nowlin, and R.A. Locarnini. Water masses and mixing near the Antarctic slope front. In: Jacobs, S.S. e Wiess R.F. (Eds.): Ocean, Ice and Atmosphere. Interactions at the Antarctic Continental Margin. Antarctic Research Series, 75:1–27, 1998.

[2] A.H. Orsi, G.C. Johnson, and J.L. Bullister. Circulation, mixing and production of Antarctic Bottom Water. Progress in Oceanograph, 43:55–109, 1999.

[3] W.S. Broecker, S.L. Peackock, S. Walker, R. Weiss, E. Fahrbach, M. Schroeder, U. Mikolajewicz, C. Heinze, R. Key, T.H. Peng, and S. Rubin. How much deep water is formed in the Southern Ocean? Journ. Geophys. Res., 103(C8):15833– 15843, 1998.

[4] P.T. Harris, G. Brancolini, L. Armand, M. Busetti, R. Beaman, G. Giorgetti, M. Presti, and F. Trincardi. Continental shelf drift deposit indicates non-steady state Antarctic bottom water production in the Holocene. Marine Geology, 179:1– 8, 2001.

[5] S. Jacobs, R. Fairbanks, and Y. Horibe. Origin and evolution of water masses near the Antarctic continental margin: Evidence from H218O/H216O ratio in seawater, in Oceanology of the Antarctic Continental Shelf. Antarct. Res. Ser., 43:59– 85, 1985.

[6] A.L. Gordon, E. Zambianchi, A. Orsi, M. Visbeck, C.F. Giulivi, T. Whitworth III, and G. Spezie. Energetic plumes over the western Ross Sea continental slope.

Geophysical Research Letters, 31, 2004.

[7] A.L. Gordon, A. Orsi, R. Muench, B.A., Huber, E. Zambianchi, and M. Visbeck. Western Ross Sea continental slope gravity currents. Deep-Sea Research II, 56:796–817, 2009.

[8] L. Ceccaroni, M. Frank, M. Frignani, L. Langone, M. Ravaioli, and A. Mangini.

Late Quaternary fluctuations of biogenic component fluxes on the continental slope of the Ross Sea, Antarctica. Journal of Marine Systems, 17:515–525, 1998.

[9] D.G. Martinson, N.G. Pisias, J.D. Hays, J. Imbrie, T.C. Moore, and N.J. Shackleton.

Age dating and the orbital theory of the ice ages—Development of a high-resolution 0 to 300,000-year chronostratigraphy. Quaternary Research, 27:1–29, 1987.

345

Marine Geology

[10] T. Quaia and G. Cespuglio. Stable isotope records from the Western Ross Sea continental slope (Antartica): considerations on carbonate preservation. Terra Antartica Reports, 4:199–210, 2000.

[11] A. Piva, A. Asioli, L. Langone, D. Ridente, F. Tateo, and F. Trincardi. Cruise results and preliminary study of living benthic foraminifera assemblages in the western Ross Sea (XX Antarctic Expedition, 2004-2005). Terra Antartica Reports, 14:247– 254, 2008.

[12] D. B. Scott and J.O.R. Hermelin. A device for precision splitting of micropaleontological samples in liquid suspension. Journal of Paleontology, 67:151–154, 1993.

[13] D.J. DeMaster. The marine budgets of silica and 32 Si. JPhD Thesis, Yale University, New Haven., 1979.

[14] D.J. DeMaster. The supply and accumulation of silica in the marine environment.

Geochim. Cosmochim. Acta, 45:1715–1732, 1981.

[15] J.D.H. Strickland and T.R. Parsons. A practical handbook of seawater analysis.

Bull. Fish. Res., 167:311, 1972.

[16] E. Barahona. Arcillas de ladrilleria de la provincia de Granada: evaluaci´on de algunos ensayos de materias primas. Ph.D. Thesis, University of Granada, Spain., 1974.

[17] V. Masson, F. Vimeux, J. Jouzel, V. Morgan, M. Delmotte, P. Ciais, C. Hammer, S. Johnsen, Y. Lipenkov, V.E. Mosley-Thompson, J.R. Petit, E.J. Steig, and M.R.V. Stievenard. Holocene climatic variability in Antarctica: what can be inferred from 11 ice core isotopic records? Quaternary Research, 54:348–358, 2000.

[18] D. Denis, X. Crosta, S. Schmidt, D.S. Carson, R.S. Ganeshram, H. Renssen, V. Bout-Roumazeilles, S. Zaragosi, B. Martin, M. Cremer, and J. Giraudeau.

Holocene glacier and deep water dynamics, Adelie Land region, East Antarctic.

Quaternary Science Reviews, 28:1291–1303, 2009.

[19] M. Stuiver and P. Reimer. Extended 14 C data base and revised CALIB 3.0 14 C age calibration program. Radiocarbon, 35:215–230, 1993.

[20] E. W. Domack, A. Leventer, S. Root, J. Ring, E. Williams, D. Carlson, E. Hirshorn, W. Wright, R. Gilbert, and G. Burr. Marine sedimentary record of natural environmental variability and recent warming in the Antarctic Peninsula. In: “Antarctic Peninsula climate variability; historical and paleoenvironmental perspectives.

Antarctic Research Series, 79:205–224, 2003.

346

River Pattern and Shore Bars Migration in Late Quaternary Regression-Trasgression Continental Deposits, Salerno Bay, Southern Italy A. Conforti Institute for Coastal Marine Environment, CNR, Napoli, Italy alessadro.conforti@gmail.com Abstract The evolution of the northern Sele coastal plain during the Late Quaternary is outlined on the base of the stratigraphic reconstruction of the Late-Quaternary sequence of Salerno Bay. To this purpose thickness and stacking patterns of latest PleistoceneHolocene stratigraphic units and associated erosional – depositional features have been obtained, which record the relative sea level variation between ca.100 ky BP and the present.

In the above study area depositional-erosional features in paralic-continental areas, prograding seaward during slow sea level fall (from about 100 ky BP) have been recognised. Sedimentary structures linked to the development of a fluvial system on the presunt inner continental shelf have been also individuated. After the lowstand and during sea level rise occurred between ca. 18 ky BP and 7-5 ky BP, the stacking transgressive units and progradational paralic deposits, forming elongated prisms, locally are preserved at morphological steps below the transgressive ravinement surface. These bars are mostly represented by sandy bodies deposited during a still standing phase of landward coast line shift.

1

Introduction

The marine “forced regression” deposits [1, 2, 3] are widely studied in the recognisable stratigraphic features and, often in the last 4th order eustatic cycle a “sharp based shoreface” system occurs [4]. In the Adriatic the stratigraphic evolution of forced regressive marine deposits has been compared with the Late Holocene HST progradational wedge [5, 6]. However the coastal continental deposits of the late Quaternary regression are relatively poorly studied on Mediterranean shelves [7]. Regressive continental and transitional deposits have been prevented from being eroded in

the inner-middle shelf sector, due to a slow sea level variation during regression and rapid sea level rise. These deposits are not easy to indentify due to the overburden of thick transgressive and high stand units close to the coast that inhibits coring; these strata can only be reachedby geophysics.

Normally the migration of the erosional surface erases these thin deposits, and only some terminations are visible. The continental deposits are too thin to be readable at low resolution in seismic configurations. However, with high resolution seismics, the coarser sediments, typical of this environment, absorb the signal. In the same way the transgressive deposits are

Marine Geology

Figure 1: Salerno Bay, Eastern Thyrrhenian margin. normally thin and not clearly recognisable, due to a rapid landward shift of the coast line, on a low inclination coastal plain.

This study area provides an expanded section of Late Quaternary deposits [8], with well developed regression and transgression units. High resolution seismic profiles allow to read thin strata terminations and to recognize sedimentary bodies in paralic continental facies. The distinction of these bodies allows to map fluvial patterns and continental deposits, down stepping during regression, and the development of coastal ridges in late regression and lowstand phases. Above the transgressive units, erosive surfaces are also recognizable, as well as some shore deposits trapped between the transgressive and the ravinement surfaces. Also these bodies that are normally thin in the Late Quaternary Sequence, are well developed here in the 348

Salerno Bay, preserved in morphological steps below the transgressive ravinement surface.

2

Regional setting

The study area is located in the northern sector of the Salerno Gulf, Eastern Tyrrhenian Sea (Figure 1). The Salerno Gulf and the Sele coastal plain are part of a Pery-Thyrrenian Basin (Bartole et al 1984), whose evolution is linked to the Late Neogene â&#x20AC;&#x201C; Quaternary tectonics of the Appenninian arc [9, 10, 11, 12]. The half graben that forms the Salerno Gulf has a WSWENE trend [13]. The northern block is characterized by a steep homocline slope due to the main listric fault, where the continental platform is almost absent. In the Positano and Amalfi Bays, a portion

Marine research at CNR

Figure 2: Tectonic sketch of study area, in land and offshore. 1) faults, 2) trascurrent faults, 3) thrusts.

Figure 3: Location of studied seismic reflection profiles, unbroken line represents the location of chirp profiles. of the continental shelf is preserved. In the Salerno Bay the shelf break widens to the South close to the Tusciano and Sele mouths. The Pleistocene tectonic evolution of this sector shows a reactivation of some deep lineaments [14] and these recent faults are recognized, testifying of the resurgence of tectonics in recent times. The main among these recent NW-SE trending faults has a transcurrent component; other NE-SW trending faulting adjusts this main rotation, displacing Late Quaternary regression deposits [8] (Figure 2), and providing a large accommodation space for post glacial sediments (Figure 8). A large

scale tilting of the Salerno Gulf margin also occurred in the middle Pleistocene [15].

The continental shelf is mainly formed by a Pleistocenic sedimentary succession [16, 14]. The general progradation of these units is SW ward, near to the Sele River mouth. Finally the middle Pleistocene succession is truncated by a sharp erosional surface formed during the last regression, of the Late Pleistocene. Thick Holocene deposits drape the entire continental shelf.

349

Marine Geology

Figure 4: Chirp profiles showing strata geometry and termination of continentaltransition facies as well as regression and transgression deposits; ch: channel fill; cb: coastal bars; du: dunal cordons; la: lateral accretion; md: marine deposits (TST+HST).

Figure 5: Chirp profiles showing strata geometry and termination of continentaltransition facies as well as regression and transgression deposits; ch: channel fill; cb: coastal bars; du: dunal cordons; la: lateral accretion; md: marine deposits (TST+HST).

3 3.1

Materials and Methods Geophysic Data

More than 200 km of very high-resolution seismic reflection data were collected in Salerno Bay on the continental shelf area.

Data were acquired with the Chirp Cap II Subbottom profiler and the ship positioning was done by a DGPS system; this seismic system has a frequency modulation source (FM) pinging in a frequency band of 3 to 7 kHz; the bathymetry was acquired by an ATLAS single beam echo sounder and data where linearly interpolated by trian350

gulation (Figure 1). Lines have a prevalent NNW-SSE orientation, interspaced in about 300-500 m and some cross lines were acquired with an ENE-WSW orientation (Figure 3). Conversion of two way travel time to real depth in meters in marine sediments, was obtained assuming an average velocity of about 1600 ms-1 within the first 100 ms of the seismic record below the sea floor [17].

Marine research at CNR

Figure 6: Chirp profiles showing strata geometry and termination of continentaltransition facies as well as regression and transgression deposits; ch: channel fill; cb: coastal bars; du: dunal cordons; la: lateral accretion; md: marine deposits (TST+HST).

3.2

Seismic and stratigraphy

sequence

growing laterally to channel facies, recognized in seismics by thin reflectors and a prolongated acoustic facies signal, due The terminations and the stratigraphic arto coarser (probably sandy) sediments chitecture of the Late Quaternary Sequence (Figure 4, 5, 6). deposits have been recognized previously â&#x20AC;˘ coastal dunes (du) [20], a clear reflecin the Salerno Bay [8], and the stacking tor defines these bodies with a total abpattern of Systems Tracts marine units has sorbance of the signal (due to sandy sedibeen clearly defined (Figure 7). However ments). Morphology and lateral continuthe main difficulties of continental-paralic ity are recognizable in seismic data (Figfacies recognition in high resolution seisure 4). mics data are due to coarse sediment sig- â&#x20AC;˘ estuarine microtidal wave dominated nal absorbance and the recognition of thin systems [21, 22], coastal bar corps (cb) strata termination. The Chirp seismic sysare clearly defined by prograding seatem allows for a sufficient penetration in ward clinoforms with great lateral concoarse (sandy) sediment and a good resotinuity, and prolongated acoustic facies lution of terminations. The closely spaced due to coarser sediments (sandy) (Figure dataset allows to interpolate thin and lim4). ited bodies and to map the distribution of acoustic facies corresponding to various continental and transitional deposits.

4 Discussion The main continental-paralic facies recognized are the following: In seismic profiles the geometrically deepâ&#x20AC;˘ fluvial sedimentary elements [18, 19], est unit (Figure 7) called c FSST represents channel cut and channel fills deposits continental deposits of sea level fall; this (ch) that are characterized by concavely unit is characterized by a sharp truncation shaped reflectors, testifying of erosion , surface at the top and organized in contiand filling corps, mainly with parallel re- nental parasequences of coastal plain deflectors. Lateral accretions (la) are char- posits and river facies; these bodies proacterized by progradational clinoforms gressively prograde seaward on the FSST 351

Marine Geology

Figure 7: Schematic section of Late Quaternary Sequence in Salerno Bay. FSST: Falling Stage Systems Tract. LST: Lowstand Systems Tract; TST: Transgressive Systems Tract; HST: Highstand Systems Tract; sb: sequence boundary; mfs: maximum flooding surface; c FSST: continental deposits of Falling Stage; s TST: shore bars of Transgressive Systems Tract. posited. The morphology of the conmarine clinoforms truncated at the top. tinental shelf had a low gradient penA coastal prograding wedge of lowstand dency and the top of the offlap breaks of (LST) stacks on the Late Quaternary Sethese units were eroded by slow sea level quence boundary (sb) and is located at the fall and the relative displacement of the shelf break. The transgressive units are shore and transitional environments; the mainly marine (TST). Two main coastal progradation of the coastal plain on mabars also formed prograding wedges (s rine sediments is recognizable by mapTST) during sea level rise. The High stand units (HST) are characterized by progradaping of alluvial deposits and fluvial facies; the lowering of the base level and tional deposits located close to the coast the seaward migration of fluvial patterns and a thick drape of sediments along the caused a cutting of coastal plain bodies shelf. Post glacial deposits, (transgressive and progressive resedimentation of alluand highstand units) reach a high thickness particularly in the western sector of vial deposits (c FSST, Figure 7).

Salerno Bay, where accommodation space â&#x20AC;˘ On this alluvial plain the probable paleochannel pattern of the Picentino River was larger due to recent tectonics (Figure 8). and the Tusciano River in the South, that probably became one stream (Figure 9), Mapping of marine and continental facies are recognizable. During the slowing linked to sea level variation stages, allowed down of sea level fall some dune cordons to delineate the depositional environment were formed on the coastal plain at about evolution in Late Quaternary as follows: -105 m, maximally 6 m high. The euâ&#x20AC;˘ After 100 ky bp, with a fluctuant seastatic minimum was reached at about ward migration of the coast line, pro120 m, which allowed for the deposition grading marine parasequences were de352

Marine research at CNR

Figure 8: Isopaches of postglacial sediments in Salerno Bay (thickness is expressed in metres). of a prograding wedge (LST) (in Figure 10, D0). Due to sea level standing still, the shore environment formed an erosional flat surface. The coastal and alluvial plain had some wide terraces at about -115 m (E2), -105 m (E3), and -95 m (E4). The rapid sea level rise after the eustatic minimum (about 18ky bp), allowed the coast line to shift rapidly landward, also due to the low inclination of the plain. A few coastal deposits are preserved during sea level rise (D3), trapped in some morphologic steps; close to the terrace at -95 m (E4) a large prograding shore body was deposited (D5) probably during a reduction of sea level rise

velocity (Younger Dryas). Other terraces are recognizable higher up at -60 m (E6), -50 m (E7) and -30 m (E8); the E6 terrace also preserves a well developed shore body (D7). After D5 deposition, mainly aggradation is recognizable in marine deposits up to the present sea level reached at about 5-6 ky bp. Sedimentation rates became exceedingly low in periods around deposition of the maximum flooding surface (mfs, Figure 7).

The first downlaps are recognized in the eastern coastal sectors, and shore prograding bodies developed in these areas with high sedimentation rates.

353

Marine Geology

Figure 9: Reconstruction of river pattern during regression; td thyrrenian dunal cordon on land cutted by actual rivers; ch paleoriver pattern on continental shelf, dashed lines uncertain pattern; ld late regression and lowstand coastal dunes

Figure 10: Morphology of transgressive surface in study area, reconstruction of the landward coastal shift.

354

Marine research at CNR

5

Conclusions

The attempt of correlating acoustic facies in geophysics with sedimentary continental facies, in a defined and constrained sequence stratigraphy, provided the outline of the evolution of a depositional environment, during the Late Quaternary. The consequent river pattern and alluvial plain deposits during regression, are probably more developed close to the present coast.

The presence of recognisable fluvial facies preserved along the inner shelf testify of a probably slow and fluctuating sea level fall during early stages of regression. Also strong river activity and a great sediment supply are also suggested by the thickness of preserved continental deposits. High sediment accommodation space in the western sector was probably induced by recent mainly transcurrent tectonics, that were active until regression

stages. In transgressive stages a rapid sea level rise was recognizable allowing for preservation of coastal deposits , except for coastal bars, that are probably correlated to a slowing down of the sea level rise velocity in the Younger Dryas. In highstand stages the sedimentation rate mainly increased in the eastern sector close to river mouths.

6

Acknowledgements

I would like to mention CA.R.G. P. (experimental marine geologic mapping project) carried out by the IAMC CNR (Institute for the Marine Coastal Environment, National Research Council) of Naples for the Italian Geological Survey (ISPRA); all the data were acquired in the frame of this project between 1998-2000. I would also like to thank Prof. B. D’Argenio, who substained this study with insightful suggestments.

References [1] H.W. Posamentier, G.P. Allen, D.P. James, , and M. Tesson and. Forced regressions in a sequence stratigraphic framework: Concepts, examples, and exploration significance. American Association of Petroleum Geologists Bulletin, (76):1687–1709, 1992.

[2] D. Hunt and M.E. Tucker and. Stranded parasequences and the forced regressive wedge systems tract: deposition during base-level fall. Sedimentary Geology, (81):1–9, 1992.

[3] H.W. Posamentier and W.R. Morris. Aspect of the stratal architecture of forced regressive deposits. Sedimentary responses to forced regressions, (172):19–46, 2000.

[4] A.G. Plint. Sharp-based shoreface sequences and offshore bars in the Cardium Formation of Alberta; their relationship to relative changes in sea level. Sea Level Changes––An Integrated Approach; SEPM Special Publication,, 42(42):357–370., 1988.

[5] F. Trincardi and A. Correggiari. Quaternary forced-regression deposits in the Adri-

355

Marine Geology

atic basin and the record of composite sea level cycles. Geological Society Special Publication, 172:247–271, 2000.

[6] D. Ridente and F. Trincardi. Pleistocene “muddy” forced-regression deposits on the Adriatic shelf: a comparison with prodelta deposits of the late Holocene highstand mud wedge. Marine Geology, (222–223):213–233, 2005.

[7] A. Amorosi, M.L. Colalongo, and F. Fusco. Glacio-eustatic control of continentalshallow marine cyclicity from Late Quaternary deposits of the south-eastern Po Plain (Northern Italy). Quaternary Res., 52:1–13, 1999.

[8] A. Conforti. Stratigrafia integrata della Sequenza Tardo-Quaternaria nel settore settentrionale del Golfo di Salerno ed in quello meridionale del Golfo di Napoli Phd Thesis. Universit`a Federico II di Napoli, 2003.

[9] I.R. Finetti & A. Del Ben. Geophysical study of the thyrrenian opening. Boll. Geof.

Teor. Appl., (28):75–155, 1986.

[10] A. Malinverno & W.B.F. Ryan. Extension in the Tyrrenian Sea and shortening in the Appennines as a result of arc migration driven by sinking of the litoshere. Tectonics, (5), 1986.

[11] E. Patacca and R. Sartori & P. Scandone. Tyrrhenian basin and Appennic arc: cinematic relations since late Tortonians time. Mem. Soc. Geol. It, 45:425–451, 1990.

[12] L. Ferranti, P. R. Gialanella, A. Incoronato, and F. Heller. Localized strain zone during polyphase non coaxial deformation in the Lagonegro rochs, southern apennines: inferences from structural and palaeomagnetic data. Structures and Properties of High Strain Zones in Rocks. Ric. Sc. Educ. Perm, (107), 1996.

[13] M. Marani, F. Gamberi, and E. Bonatti. From seafloor to deep mantle: architecture of the Tyrrhenian backarc basin. Mem.Descr.Carta geol.d’Italia, Vol.LXIV, 2004.

[14] M. Sacchi, S. Infuso, and E. Marsella. Pleistocene compressional tectonics in offshore Campania. Bollettino di geofisica ed Applicata, Vol. XXXVI:141–144, 1994.

[15] F. Budillon, T. Pescatore, and M.R. Senatore. Cicli deposizionali del Pleistocene superiore-Olocene sulla piattaforma continentale del golfo di Salerno (Tirreno Meridionale). Boll. Soc. geol. It., (113):303–316, 1994.

[16] R. Bartole, D. Salvelli, and M. Tramontana & C. Forese. Structural and sedimentary features in the tyrrenian margin off campania , southern Italy,. Marine Geology, (55):163–180, 1984.

[17] R.L. Carlson, A.F. Gangi, and K.L. Snow. Empirical reflection travel time versus depth and velocity versus depth functions for the deep sea sediments column.

Journal of Geophysical Research, (91):8249–8266, 1986.

356

Marine research at CNR

[18] A.D Miall. Architectural-element analysis: a new method of facies analysis applied to fluvial deposits. Earth Science Rewiews, (22):261–308, 1985.

[19] A.D. Miall. Reservoir heterogeneities in fluvial sandstones . ;AAPG , Bulletin, (72):682–297, 1988.

[20] G. Einsele and S.K. Chough & T. Shiki. Depositional events and their records-an introduction. Sedimentary Geology, (104):1–9, 1996.

[21] G.M. Ashley and M.L. Zeff. Tidal channel classification for a low-mesotidal salt marsh. Marine Geology, (82):17–32, 1988.

[22] G.E. Reison. Transgressive barriers Island and Estuarine systems. Facies Models: Response to Sea level Changes., QE 651:179–194, 1992.

357

Marine Geology

358

The Tectono-Stratigraphic Evolution of the Eastern Tyrrhenian Sea (from Calabria to Campania Margins) and its Geodynamic Implications A. Milia, Institute for Coastal Marine Environment, CNR, Napoli, Italy alfonsa.milia@iamc.cnr.it Abstract The interpretation of seismic reflection profiles is a fundamental step in the analysis of sedimentary basins. Over the last two decades we studied the basins off Campania and Calabria located in the Eastern Tyrrhenian Sea. The application of the sequence stratigraphy and structural geology methodologies permitted us to recognize tectonically enhanced unconformities, uplifting and subsiding areas, fault architecture and kinematics to reconstruct a detailed chronostratigraphic framework.

For the first time we identified the main tectonic and sedimentary events that characterize the Eastern Tyrrhenian Sea margin. Because the opening of the Tyrrhenian basin is linked to the formation of the Apennine thrust belt, the results of this work have important implication for the tectonic history and geodynamic evolution of the Tyrrhenian Sea-Apennine thrust belt system. In conclusion the post-700 ka geologic evolution of the Eastern Tyrrhenian Sea features an intense Quaternary volcanism in Campania the coeval activity of NNE and NE-trending normal faults off Campania and NW-trending right lateral faults active off Calabria. Such faults are compatible with the Southeast migration of the Calabria Arc occurred in the last 700 ka.

1

Introduction

Plate reconstructions of the central Mediterranean region have always been controversial due to the lack of welldefined oceanic magnetic anomalies and to a complex tectonic history. The opening of the Tyrrhenian Basin has been the subject of numerous papers. However, these kinematic models have to be confirmed by the structural and stratigraphic evidence derived from marine geology and field data. The Tyrrhenian Basin is a Neogene back-arc basin formed by extensional tectonics within the overriding plate of the eastward migrating subduc-

tion system, giving rise to the arcuate fold and thrust belt running from the centralsouthern Apennines to Sicily (Figure 1).

According to several authors [2, 3, 4], the extensional processes in the Southern Tyrrhenian Sea started during the Tortonian stage and took place in a region previously affected by contractional tectonics and by the stacking of tectonic units now belonging to the Southern Appenine Belt. Nevertheless, a possible Serravallian age for extension onset has been recently suggested by stratigraphic and structural studies carried out by Mattei et al. [5] in the Amantea area along the Tyrrhenian Margin of the Calabrian Arc. The East-

Marine Geology

Figure 1: Seafloor morphology of the Thyrrenian Sea and location of the study areas (from [1]). NB=Naples Bay, PB=Paola Basin, AB=Amantea Basin The upper inset shows a schematic structural map of Italy. TS=Thyrrhenian Sea; NAA=Northern Apennine Arc; SAA=Southern Apennine Arc; AF=Adriatic Foreland. ern Tyrrhenian Sea Margin off Campania and Calabria is characterized by the presence of deep peri-Tyrrhenian basins (Figure 1) filled by an up to 5000 m thick sedimentary succession deposited during the Serravallian-Pleistocene. In particular the Paola Basin (off Calabria) features an older Serravallian-Pleistocene sedimentary infill, whereas the younger Naples Bay Basin (off Campania) was filled by clastic sediments and volcanics Lower Pleistocene to Present in age. The study of the stratigraphic succession of these peri360

Tyrrhenian basins revealed the signature of the tectonic events responsible for the opening of the Tyrrhenian Sea. The study of the Eastern Tyrrhenian Basins is crucial for the kinematic analysis of the opening of the Tyrrhenian Sea as this area lies between the Southern Apennine and Calabrian Arcs and records the sedimentary and tectonic history from the Serravallian to the Present.

The four-dimensional structural analysis of the area reveals a complex framework resulting from the superposition of different tectonic events, the ages of which are

Marine research at CNR

constrained by stratigraphic analysis. Our study was based on a high quality coverage of two dimensional seismic data, cores and boreholes. We integrated high resolution seismic reflection profiles, CROP seismic lines [6] and detailed bathymetric map [1, 7] to provide a three-dimensional image of the geological units. The basin analysis of the deepest pery Tyrrhenian basins revealed a geologic evolution characterized by the opening of the southeast Tyrrhenian margin (off Calabria) starting in the Serravallian stage and successively affected by transcurrent faulting. Instead, the opening of the Campania margin occurred in the Pleistocene and from the Middle Pleistocene both basins migrated toward southeast [8, 9, 10, 11, 12, 13]. These results provide fundamental geological constrains on the geodynamic evolution of the Tyrrhenian â&#x20AC;&#x201C; Apennine system.

2

Methodology

Seismic reflection profiles of different resolution and penetration together with deep stratigraphic logs and piston cores were used to investigate the Eastern Tyrrhenian continental margin (Figure 1). In particular, our work is based on the CROP seismic profiles that were acquired for the Italian deep crust project, characterized by deep penetration and low resolution (for further details see [6]), and Sparker data, acquired with a Multispot Extended Array System.

The output power of the MEAS, transmitted through a 36-tip array, was 16 kJ. Vertical recording scales were 2.0 sec with a maximum vertical resolution of 6 m. Our interpretation of the seismic data set includes the recognition of individual seismic units that were characterized on the basis of the contact relations and internal lay-

ering properties [14]. Structures were correlated between seismic profiles with confidence. A stratigraphic framework was established using unconformity bounded units and key reflectors traced across the deformation zone. In addition, a detailed seismic facies analysis of the undeformed succession was carried out using the sequence stratigraphy approach. The seismic interpretation was calibrated by deep wells, piston cores and the stratigraphic succession outcropping in the coastal area. The mapping of structures and isopachs of the sedimentary units were carried out[8, 9, 10, 11, 12, 13]. The seismic stratigraphy interpretation permitted the individuation of unconformity-bounded units that correspond to tectono-stratigraphic units (Figure 2). During the late Quaternary, sea level fluctuations occur with a cycle of 100 ky as a response to climatic variations. Compared to the tectonic events, the eustatic sea level fluctuations display a higher frequency. This is reflected in the stratigraphic succession where it is possible to distinguish the high frequency depositional sequences associated with the eustatic sea level fluctuations. Several depositional sequences were recognized. Based on the number of recognized depositional sequences and on the age constraints provided by the age of volcanoes, well data and dated geologic events on the coastal area, we tentatively attributed an age to the recognized seismic units (Figure 2). The main results of this work will be schematically illustrated for the Paola Basin and Naples Bay Basin (Figures 1-3)

361

Marine Geology

3

Geologic evolution of the Calabria Margin

The structural and stratigraphic analyses of the basins along the Margin of Calabria (Figure 2 [13]) reveal the presence of two sedimentary depocenters, the Paola Basin and the St. Eufemia Trough, that correspond to the oldest basins of the Eastern Tyrrhenian margin.

Stage 1: The first deposits overlying the metamorphic substrate (Unconformity U1) corresponds to Unit SU1, SerravallianTortonian in age. This sedimentary unit indicates the onset of basin formation whereas the onlap of the strata overlying the metamorphic substrate indicates the progressive filling of the basin. This unit can be interpreted as the stratigraphic signature of the opening of the Tyrrhenian Basin. This Stage is well documented onshore at Amantea [5].

Stage 2: The Messinian stage is represented locally by sediments and by an erosional surface (unconformity U2). The seismic interpretation and well data indicate lateral changes in the thickness of the Messinian succession. Because the thickness increases on hangingwall fault blocks, we suggest that the E-W trending normal faults have been active since the Messinian age. Besides the lateral juxtaposition of two approximately N-S substrate blocks with different stratigraphic successions is in agreement with the presence of a transfer fault during the Tortonian and Messinian stages.

Stage 3: Unit SU3 (Pliocene-Lower Pleistocene) corresponds to deposits overlying the Messinian horizon. This unit witnesses a period of subsidence that is associated with the E-W trending normal fault activity. At the end of this stage the sediments 362

completely filled the basin as suggested by the upper strata characterized by the development of prograding unit in a shallow water environment.

Stage 4: The occurrence of the unconformity U4 is interpreted as an abrupt change in the basin sedimentation. The deposition of SU4, characterized by the presence of deep water turbidites and slumps on the slope, indicates a rapid deepening of the basin. The deposition of a wedge seismic unit thinning toward the West on an uplifting area indicates the formation of synsedimentary growth folds with a source area for the sediments from the West and North.

The map of these fold axes indicates an en-echelon geometry approximately N-S. During this stage the Paola Basin physiography changes, assuming approximately a N-S trending orientation.

Stage 5: Unit SU5 indicates a dramatic change in the basin fill. The occurrence of the unconformity U5 indicates an abrupt change in the stratigraphic architecture of the basin. This event occurs in correspondence to the sequence boundary 700 ka B.P.. The distal stratal termination onlaps the slope of the fold and indicates the infill of the basin corresponding to the previously formed syncline. In the stratigraphic succession we distinguish two substages, represented by SU5 and SU6 units. At the beginning the parallel seismic configuration indicates a horizontal infill of the basin. Successively in the eastern part of the basin, the strata were folded , whereas the deposition of the thicker depositional sequences, made up mainly of turbidites, forms a front fill. These features indicate, for the first time in the evolution of the basin, the presence of a continental slope and an important sedimentary source to the East. At the same time, the post-700 ka uplift of Calabria is documented by the up-

Marine research at CNR

lift of the marine terraces along the coast, Stage 3: During this stage the overall conwhereas offshore the activity of the NW-SE figuration of the shelf and basin determines dextral faults is documented. the architecture of the sedimentary infill.

The latter corresponds to three depositional sequences (B1, B2, B3; Figure 2) that form unit in the shelf slope area and 4 Geologic evolution of the aanprograding aggrading unit in the basin. The transCampania Margin gressive trend of the first unit is related to the creation of accommodation space durThe structural and stratigraphic analysis ing the rapid basin subsidence associated to of the Bay of Naples permitted the re- the tectonic activity [8]. At that time the construction of the Quaternary tectono- continental margin was characterized by stratigraphic evolution of this sector of the high tectonic subsidence related to the fault Eastern peri-Tyrrhenian margin. A five- activity and low sediment supply. This latstage scenario of this evolution will now be ter was enable to fill the accommodation space and as a result the basin was mainly presented (Figure 2).

Stage 1: The oldest sedimentary unit, Unit filled with water.

A, corresponds to the first deposits overly- Stage 4: The end of fault activity is marked ing the Meso-Cenozoic substrate. This unit by a rapid decrease in the accommodation marks the beginning of a period of sub- space creation. This implies that the desidence associated with the activity of the pocentre of sequence set C migrate laterNW-SE trending normal faults. The seis- ally from the shelf toward the basin filling mic configuration and distribution of unit the previously formed basin. This stage A suggest that the rate of sediment supply was characterized by a relatively tectonic was high enough to fill the accommodation stability and high sediment supply that was responsible for the formation of a regresspace creation.

Stage 2: At 700 ka the formation of the sive trend [8]. unconformity U1 and the rapid variation Stage 5: After the deposition of the sein the physiographic configuration of the quence set C the physiography of the bay bay is related to a tectonic event. The NE- was similar to the present day one. ApSW oriented normal faults are character- proximately 100 ky ago the formation of ized by segments linked by relay zones and the unconformity U2 in the shelf repreresponsible for the basin formation. The sent a tectonic event. Indeed the reactifault blocks tilting and the relative posi- vation of the pre-existing faults occur contion of the sea level produced an emerged temporaneously to an intense volcanic acarea and deep hangingwall basins: the Bay tivity [10]. An aggradational depositional of Salerno Basin, the Campi Flegrei Basin, sequence (D) successively formed in the and the Capri Basin. The distribution of shelf area documenting a rapid increase in sediments was affected by the structural the accommodation rate, coeval with the pattern. Indeed sediments produced by the large volume pyroclastic eruptions [11]. shelf erosion are directly deposited in the adjacent basin west of Capri via the relay zone and Campi Flegrei area via slope basin.

363

Marine Geology

5

Conclusion

The structural and stratigraphic analysis of the deeper basins located in the Eastern Tyrrhenian Margin permitted the reconstruction of a detailed basin evolution allowing tectonic and eustatic signatures to be identified. It was seen that, tectonics controlled basin formation, whereas the interaction between tectonic structures and relative sea level position controlled the environment of sedimentation and unconformities. In the Eastern Tyrrhenian Sea off Calabria we document the presence of the oldest stratigraphic unit, SerravallianTortonian in age, that onlaps the substrate.

Successively a N-S extension took place giving rise to the Paola Basin and St. Eufemia Trough (Figure 3) that were com-

pletely filled by sediments. This sedimentary succession witnesses the onset of the Tyrrhenian basin opening. Successively the activity of NW-SE left lateral faults are related to the opening of the Marsili basin (Figure 3). Starting approximately at 700 ka, the formation of pull-apart basins, that overprint en-echelon folds, documents a change in the slip sense of the NW-SE trascurrent faults from a left-lateral to a right-lateral (Figure 3). The opening of the deep basins in the Campania Margin occurs approximately 700 ka. The SE oriented extension, the intense volcanism and the activity of the NW-SE oriented right lateral fault off Calabria suggest a jump toward the north in the extension and the migration of the eastern Tyrrhenian margin sector toward SE (Figure 3).

References [1] M.P. Marani and F. Gamberi. Distribution and nature of submarine volcanic landforms in the Tyrrhenian Sea: the arc vs the backarc. 2004.

[2] A. Argnani and F. Trincardi. Paola slope basin: evidence of regional contraction on the eastern tyrrhenian margin. 1988.

[3] K.A. Kastens, J. Mascle, et al. Leg 107 in the Tyrrhenian Sea: insights into passive margin and back-arc basin evolution. 1988.

[4] R. Sartori. Main results of ODP Leg 107 in the frame of Neogene to Recent geology of perityrrhenian areas. 1990.

[5] M. Mattei, P. Cipollari, D. Cosentino, A. Argentieri, F. Rossetti, F. Speranza, and L. Di Bella. The Miocene tectono-sedimentary evolution of the southern Tyrrhenian Sea: stratigraphy, structural and paleomagnetic data from the on-shore Amantea Basin (Calabrian Arc Italy). 2002.

[6] D. Scrocca. et al. Crop Atlas. Seismic reflection profiles of the Italian crust. 2003.

[7] M. P. Marani et al. Seafloor morphology of the Tyhhernian Sea. Plate 1. 2004.

[8] A. Milia. Aggrading and prograding infill of a pery-tyrrhenian basin (Naples Bay, Italy). 1999.

364

Marine research at CNR

[9] A. Milia and M.M. Torrente. Tectonics and stratigraphic architecture of a peryTyrrhenian half-graben (Bay of Naples, Italy). 1999.

[10] A. Milia and M. Torrente. Late Quaternary volcanism and transtensional tectonics at the Campania continental margin Bay of Naples, Italy. 2003.

[11] A. Milia and M. Torrente. The influence of paleogeographic setting and crustal subsidence on the architecture of ignimbrites in the Bay of Naples (Italy). 2007.

[12] A. Milia et al. Tectonics and crustal structure of the Campania continental margin: relationships with volcanism. 2003.

[13] A. Milia et al. Four-dimensional tectonic-stratigraphic evolution of the Southeastern peri-Tyrrhenian Basins (Margin of Calabria, Italy). 2009.

[14] R.M.Jr. Mitchum, P.R. Vail, and J.B. Sangree. Seismic stratigraphy and global changes of sea level Part 6, Stratigraphic interpretation of seismic reflection patterns in depositional sequences. 1977.

365

Marine Geology

366

Marine research at CNR

Figure 2: Tables showing the results of the seismo-stratigraphic interpretation in the basins offshore Calabria and Campania.

367

Marine Geology

Figure 3: Three evolutionary steps (based on the structural and stratigraphic data, regional geology) linked to the geodynamics of the region (modified from [8]). A: The NS extension is the result of the different orientation and similar magnitude of the velocity vectors of the Southern Appenine and Calabrian Arcs. B: NW left lateral faults is the result of the decrease in the magnitude of the the velocity vectors of the Southern Appenine Arc. C: The NW right lateral faults in the Eastern Tyrrhenian Margin and the stop of the Southern Appenine Arc towards NE indicate a movement of this latter toward SE. In particular the complex tectonic evolution can be interpreted as the response in the velocity variation between the Southern Apenninic Arc and the Calabrian Arc during the thrust belt formation.

368

The Role of Marine Geology in the Reconstruction of Vesuvius Volcano History A. Milia1 , A. Raspini2 , M.M. Torrente3 1, Institute for Coastal Marine Environment, CNR, Napoli, Italy 2, Institute of Geosciences and Earth Resources, CNR, Firenze, Italy 3, Department of Geological and Environmental Studies, University of Sannio, Benevento, Italy alfonsa.milia@iamc.cnr.it Abstract Vesuvius in southern Italy has been one of the most active volcanoes in the world.

Although its eruptions have been recognized through several geologic studies of subaerial exposures, recent marine geological investigations that documented offshore Vesuvius products greatly improved the reconstruction of the volcano history. In detail, three thick stratigraphic units and criptodomes were identified and mapped for the first time on the continental shelf of the Bay of Naples through the interpretation of high-resolution seismic reflection profiles calibrated by offshore gravity cores.

These stratigraphic units are genetically linked to three plinian volcanic eruptions and interlayered between the systems tracts of the last depositional sequence. Two units, coeval with the 18 ka-old Pomici di Base eruption and 3.5 ka-old Avellino eruptions, correspond to debris avalanches resulting by flank collapses whereas the third unit, coeval with the 79 A.D. Pompei eruption, represents submarine pyroclastic flow deposits. The documentation of the rapid entrance of debris avalanches and pyroclastic flows originating from Vesuvius into the sea-water may be relevant for studies of volcanoes in close proximity of coastlines and imply that debris avalanche-generated and pyroclastic flow-generated tsunamis need to be taken into account for hazard evaluation in the management of the densely populated coastal zone of Naples Bay.

1

Introduction

The Somma-Vesuvius volcanic complex is located in a highly populated area on the northeastern coast of the Bay of Naples that is characterized by crustal thinning and severe rifting processes affecting the Neogene southern Apennines thrust belt.

The Bay of Naples is a Middle Pleistocene NE-trending half graben with its depocenter located in the north-west. It is filled by fourth-order depositional sequences that are arranged in sequence

sets displaying long-term aggradationalprogradational stacking patterns [1, 2]. The Mesozoic-Cenozoic carbonate substratum, cropping out on the Sorrento Peninsula, dips towards the NW and is overlain by Quaternary sediments and volcanic products [3]. The Somma-Vesuvius edifice corresponds to the breached crater of the Mt. Somma stratovolcano and the recent cone of the active Vesuvius, which grew within Mt. Somma. It began its eruptive activity at least 35 ka BP [4]. The volcanic succession of the Somma-Vesuvius

Marine Geology

complex mainly consists of lava flows and minor scoria fall deposits [4] overlain by the products of four main plinian eruptions (e.g., [5]): the Pomici di Base (18 ka BP), the Mercato Pumice (8 ka BP), the Avellino Pumice (3.5 ka BP) and the Pompeii Pumice (AD 79). The last eruption of the volcanic complex occurred in 1944 after a period (1631-1944) of semi-persistent activity. At the present, the volcano shows a moderate seismic and fumarolic activity.

The region around Vesuvius has been investigated by means of a lithostratigraphic analysis of boreholes drilled by private companies and local authorities, and/or by the reinterpretation of data from the literature [6, 7, 8, 9]. Locally, seismic reflection data have been calibrated using gravity cores whose sediment description was carried out at centimetre scale using both a 10x hand lens and microscope observations of sieved窶電ried sediment [10]. The interpretation of high-resolution seismic reflection profiles carried out on the continental shelf of the Bay of Naples and calibrated by gravity cores, allowed the identification, for the first time, of cryptodomes and of three thick stratigraphic units (Figure 1) that are interlayered in the systems tracts of the depositional sequence following the Last Glacial Maximum. Two units are coeval with, respectively, the 18 kaold Pomici di Base eruption and 3.5 kaold Avellino eruptions, and correspond to debris avalanches resulting by flank collapses. The third unit, instead, represents the underwater components of the onshore pyroclastic current deposits that buried the Roman town of Herculaneum during the AD 79 Pompeii eruption of Vesuvius. The statigraphy offshore Vesuvius is characterized by the Campania Ignimbrite (CI, 39 ka) covered by the Late Quaternary depositional sequence. Between the systems 370

tracts of the depositional sequence different volcanic features are present [11, 12, 7, 10].

2

Cryptodomes Faults

and

Mounds characterized by reflection free and chaotic reflection configuration were recognized off Vesuvius. They are positioned at water depths ranging between 80 and 110 m and result aligned parallel to the NW-SE coastline (Figure 1). The early Lowstand Systems Tract deposits overlying these mounds appear warped and their upper strata were eroded and re-deposited during the Lowstand of the sea level successively the deposits of the Transgressive and Highstand Systems Tracts (characterized by parallel configuration) draped the previous morphology. These deposits were gently folded, disrupted and pierced by the mounds. This overall stratigraphic architecture suggests a localized uplift in correspondence of the mounds that were consequently interpreted as cryptodomes [11].

These cryptodomes result characterized by at least two events of uplift during the Lowstand and Highstand of the sea level.

Moreover, a tectonic uplift offshore Pompeii predates the deposition of the 79 AD pumice fall [14]. These stratigraphic constraints suggest that the cryptodome intrusions correspond to repeated events possibly coeval with Vesuvius plinian eruptions (Pomici di Base, 18 ka; Avellino, 3.5 ka; A.D. 79). Our cryptodome interpretation was recently confirmed as the four mounds mapped off Vesuvius (Figure 1) exactly correspond to positive anomalies (up to 300 nT) of the magnetic map of Naples Bay [15]. A 34 km-long NE-SW trending re-

Marine research at CNR

Figure 1: Physiography of Naples Bay and map distribution of the DA1 (yellow), and DA2 (light blue) debris avalanches, the submarine volcanoclastic fan associated to the 79 AD eruption (green) and the cryptodomes (red). Dashed lines mark the presumed onshore boundaries of these units. Tectonic framework is from [11] and [13]. gional section from the slope and shelf of the Bay of Naples to the southern Campania Plain was reconstructed [13] by fitting offshore seismic stratigraphy [11, 3] and drill hole stratigraphy [6, 16] in the southern Campanian Plain. The sequence consists of Middle Pleistocene marine sediments overlain by Upper Pleistocene ignimbrites (pre-CI tuffs and CI) and post25 ka Somma-Vesuvius products. Tracing of the Upper Pleistocene ignimbrite marker horizons from the shelf to the southern Campania Plain permitted the imaging of two normal faults (one located on the coast and another below the Vesuvius cone).

Both normal faults displace the pre-CI tuffs and the top of the Campania Ignimbrite.

They downthrow to the southwest. The thicknesses of the pre-CI tuffs and CI unit is greater in the footwall than in the hangingwall block and fault throws are greater for older marker levels. On the basis of these features we maintain that normal fault activity started during the emplacement of the Upper Pleistocene ignimbrites.

An evidence of the recent activity of the NW-SE Vesuvius faults comes from historical data. Indeed an earthquake, which seriously damaged Pompeii, occurred on February 5th 62 AD, with a Magnitude of 371

Marine Geology

5.9. Because of the high value of the magnitude and of the NWâ&#x20AC;&#x201C;SE trend of the isosists it was suggested a tectonic origin for the earthquake and hypothesized that the tectonic event triggered an upward surge of new magma, that 17 years later was responsible of the 79 AD Plinian eruption [17].

3

The DA1 debris avalanche (18ky-old)

The Campana Ignimbrite is capped by an unconformity covered by an approximately 20 m thick younger unit showing parallel reflectors and a tabular external form. This unit, made up of sediment deposited during the fall of the sea level between 39 and 18 ka ago [12], is covered landward

by Unit DA1 that features a chaotic seismic facies. A younger seismic unit, characterized by parallel reflectors with low amplitude and low frequency and a thickness of approximately 30 m, onlaps unit DA1.

At its seaward margin Unit DA1 is covered by a progradational seismic unit corresponding to the Late Lowstand Systems Tract. The seismic unit DA1 shows chaotic facies and hummocky surfaces, and is interlayered within the systems tracts of the Late Quaternary depositional sequence on continental shelf between Portici and Torre Annunziata, until 100 m of water depth (Figure 1). Its top features a staircase morphology with flat surfaces overlain by marine sediments. The seaward termination displays a steep 60 m scarp with an irregular pattern characterized by elongated lobes that lie perpendicular to the coast. Unit DA1 has an areal distribution of 36 km2 , with a volume of approximately 2.9 km3 , and, assuming a mean velocity of 1600 m/s for the depth conversion, its thickness 372

ranges from 100 m close to the volcano to 60 m far from the coast [12]. As Unit DA1 lies above the unconformity overlying the 39 ka Campana Ignimbrite and below the wedge that formed during the sea level lowstand, [12] suggested that it settled about 18 ka ago in a subaerial environment (see also [18]). Unit DA1 cannot be associated with normal sedimentation that occurred on the continental margin or with any Somma-Vesuvius eruption because of its unusually large thickness, lateral extension and internal and external seismic facies. Reflection-free seismic facies (suggesting a massive sedimentary body), hummocky external seismic facies and a 60 mhigh terminal scarp of unit DA1 are typical of a dry debris avalanche with a poorly sorted mixture of brecciated debris (e.g., [19, 20, 21, 12]). Based on 1) the good spatial correlation between the offshore lateral boundary of unit DA1 (depositional area) and the presumed onshore scarp of the Mt. Somma breached crater (source area); 2) the stratigraphic relationships between the debris avalanche products and the systems tracts of the Late Quaternary depositional sequence; 3) the ages of the dated SommaVesuvius plinian eruptions, [12] associated the debris avalanche to the collapse of Mt. Somma at 18 ka shortly before the Pomici di Base eruption.

4

The DA2 debris avalanche (3.5 ka-old)

The deposits related to the Avellino eruption show a chaotic seismic facies, reach a water depth of 150 m, have a total volume exceeding 1 km3 and are sandwiched by marine sediments [7, 8, 9], (Figure 1). They consist of a 15â&#x20AC;&#x201C;20 m-

Marine research at CNR

thick basal loose gravels and sands with subangular pumice, lava and scoria fragments overlain by 5 m-thick deposit consisting of yellow to green lithoid tuff including pumice, scoriae and mainly calcareous lithic fragments. The volcanoclastic deposits are interpreted as the product of a debris avalanche related to the collapse of the western seaward-facing flank of the Somma窶天esuvius, while the tuff deposits are related to pyroclastic gravity currents. Locally, the latter deposit is abruptly covered by a 2 m-thick layer consisting of sand and gravel rich in marine shell fragments and rounded pumice. The succession terminates with shallowing and coarsening upward marine sediments consisting of muds, silts and sands. A comparable stratigraphic succession of the Avellino unit as detected on the eastern coast of Naples is also present near the vent at Novelle [22] and at the Herculaneum excavation site [6]. The correlation between the offshore geological section, reconstructed by means of the interpretation of seismic reflection profiles, and the onshore geological section, reconstructed by the interpretation of borehole data, reveals a physical continuity between the Avellino unit recognized offshore, consisting of a debris avalanche covered by deposits related to pyroclastic gravity currents, and the previously recognized thick tuff deposits, in places overlying breccia deposits in the Volla plain [6]. Offshore, the distribution of the Avellino unit documents a debris avalanche that travelled westward over a distance of approximately 10 km before entering the sea and the subsequent westward flow of pyroclastic currents. The emplacement of this voluminous unit induced an instantaneous inner shelf progradation which resulted in an abrupt shift of the inner shelf edge off San Giovanni a

Teduccio, where the mound-shaped Avellino unit reaches thicknesses ranging from 50 to 70 m [9]. Here the Avellino unit is bounded at the top by a surface that displays an irregular morphology with troughs and highs of up to 20 m relief. This, in turn, is abruptly overlain by an approximately 2 m-thick layer consisting of loose sand and gravel with abundant marine shell fragments and rounded pumice. These features document a high-energy event in coastal areas which produced a deep erosion of the substratum up to a water depth of 30 m and the winnowing of fine sediments across the coastal setting before the settlement of the coarse-grained bioclastic deposit. Such high-energy depositional conditions have been correlated to a tsunami which was induced by the displacement of seawater away from the shore after the sudden entrance into the sea of the voluminous Avellino volcanoclastic unit [7, 9].

Below 40 m water depth, a slump affects the volcaniclastic debris avalanche and the overlying marine sediments. Although parameters that generally favour or trigger slope instabilities include seismic activity, angle of slope of the margin, basement architecture, sea current and high pore-fluid pressure [23], it cannot be excluded that, in such a scenario, the slump may have been triggered by tsunami-induced porepressure changes of the voluminous volcaniclastic deposit that was instantaneously emplaced under the sea, producing a rapid physiographic change and a potential area of disequilibrium [9].

373

Marine Geology

Figure 2: Cartoon depicting the syn-eruptive sedimentary processes associated to the entrance of hot pyroclastic flows into the sea during the A.D. 79 eruption of Vesuvius. For further explanations see text.

5

The AD 79 submarine volcaniclastic fan

During the Pompeii Pumice plinian eruption the deposition of the products of the subaerially-generated pyroclastic currents rapidly entered the sea. They formed a wide volcaniclastic fan that is located off Herculaneum, at 10-140 m water depth [10],( Figure 1), and induced a coastal progradation of approximately 400 m (cf.

[24]). The fan, 0.3 km3 in volume, displays a chaotic seismic facies that gradually changes seaward to parallel reflec374

tors and then to wavy reflectors. Wavy facies form dunes that are aligned roughly parallel to the coast and feature bedforms displaying marked asymmetry and irregular pattern, passing seawards to parallel reflectors. Gravity cores reveal a succession consisting of cm-thick sand/silt-sized ash couplets followed by an up to 180 cmthick graded gravely sand-sized bed with shell fragments and beach-derived pebbles that is overlain by dm-thick graded and laminated bioclastic sandy ash layers [10].

The depositional textures and sedimentary structures of the volcaniclastic fan imply

Marine research at CNR

that high-energy events repeatedly and sequentially affected the shallow continental shelf where they settled. Such an organization in multiple scour-and-graded layers of the topmost part of the AD 79 eruption deposit have been also documented in the southern sector of the Bay of Naples, while it is lacking in the adjacent Bay of Salerno [14]. This confirms that the aforementioned stacking pattern has to be related to an exceptional process that interested the Bay of Naples only, reworking substantially the AD 79 pyroclastic deposits [10]. Also, seismic interpretation showed that most of the pyroclastic products forming the submarine fan are arranged in minor units forming a retrogradational stacking pattern and wavy stratification [10]. In particular, the occurrence of dunes and the observed sedimentary structures reflect events of rapid deposition under the effect of tractive currents and excess of suspended sediment load. The features characterizing the volcaniclastic fan have been interpreted as the product of the interactions between hot pyroclastic density currents entered into the sea and tsunamiinduced water waves [10]. Such event was also observed and described by Pliny the Younger during the 79 AD eruption: “the seashore withdrew and fishes and other marine faunas laid on the emerged seafloor”.

Based on our findings, the geological literature (e.g., [24, 25, 26, 27, 28, 29]), and the description of Pliny the Younger, we put forward a speculative model of syneruptive sedimentary processes that occurred off the city of Herculaneum during the AD 79 plinian eruption (Figure 2). In the morning of 25 August, hot (350/500 °C [30, 31], pyroclastic currents transited over the city of Herculaneum and entered the sea. Near shore steam explosions and, possibly, ash fountains and convectively ris-

ing fine ash plumes were generated, as usually produced experimentally with hot ash flows into water [28]. Due to the density contrast at the air/water interface and high velocity (∼100 km·h−1 ), the currents likely segregated into a basal dense portion, that continue subaqueously, and a more dilute turbulent finer ash upper fraction, that travelled on water for a great distance from the volcano [32, 27, 28]. Once the latter sinking and interacting with seawater, finally deposited, the fractionation of the components, according to their settling velocities, led the formation of coarser ash under traction currents with excess of suspended sediment load, while the finer portion settled from suspension under calmer conditions. It allowed couplets consisting of multiple laminated-and-graded coarser layers overlying by finer material to be deposited. Further pyroclastic flows were delivered towards the sea. The denser basal portion of the flows sank and a degree of interaction between its surface and seawater developed, giving rise to density currents. As a result, chaotic seismic unit on the inner-middle continental shelf passing seaward to parallel reflector facies were laid down. Massive facies also characterize the 79 AD eruption deposits cropping out on roman Herculaneum beach suggesting a rapid fall-out of dense pyroclastic currents when they arrived on the beach and entered the sea (e.g., [33]). Later in the morning of 25 August, tsunami waves induced by the entrance into the sea of a coarse grained pyroclastic flow struck the coast of Naples Bay. The waves partly reworked the previously deposited volcaniclastic material and redistributed it to deeper areas as a marine gravity flow, inducing the settlement of graded gravely sand-sized deposit overlain by dm-thick graded and laminated bioclastic sandy ash layers.

375

Marine Geology

6

Conclusions

Recent marine geology investigations allowed us to document cryptodomes and offshore volcaniclastic deposits related to the Vesuvius volcanic activity.

Three thick stratigraphic units were identified and mapped for the first time on the continental shelf of the Bay of Naples through the interpretation of high-resolution seismic reflection profiles calibrated by offshore gravity cores. Two units, coeval with the 18 ka-old Pomici di Base eruption and 3.5 ka-old Avellino eruptions, correspond to debris avalanches resulting by flank collapses whereas the third unit, coeval with the A.D. 79 Pompeii eruption, represents a submarine volcaniclastic fan related to complex interactions between hot pyroclastic density currents entered into the sea

and the induced tsunami waves. The contemporaneous occurence of normal faults, cryptodomes and debris avalanches in the Vesuvius history suggests a genetic link between these processes. It is possible that normal faulting and cryptodome intrusion were responsible of the volcano instability producing lateral collapse and debris avalanche. The results of our research demonstrate that marine geology may offer a fundamental contribution for greatly improving the reconstruction of coastal volcanoes’ activity. In particular, our findings imply that voluminous mass-flow entering into the sea may generate tsunamis that have to be taken into account for hazard evaluation and disaster management planning of the densely populated region around the coastal active Vesuvius volcano.

References [1] A. Milia. Aggrading and prograding infill of a peri-Tyrrhenian Basin (Naples Bay, Italy). Geo-Marine Letters, 19:237–244, 1999.

[2] A. Milia and M.M. Torrente. Tectonics and stratigraphic architecture of a periTyrrhenian half-graben (Bay of Naples, Italy). Tectonophysics, 315:301–318, 1999.

[3] A. Milia, M.M. Torrente, and A. Zuppetta. Offshore debris avalanches at SommaVesuvius volcano (Italy): implications for hazard evaluation. Journal of the Geological Society, London, 160:309–317, 2003.

[4] R. Santacroce and A. Sbrana. Carta geologica del Vesuvio. 2003.

[5] D. Andronico, G. Calderoni, R. Cioni, A. Sbrana, R. Supplizio, and R. Santacroce.

Geological map of Somma-Vesuvius Volcano. Periodico di Mineralogia, 64:77–78, 1995.

[6] F. Bellucci. Nuove conoscenze stratigrafiche sui depositi effusivi ed esplosivi nel sottosuolo dell’area del Somma-Vesuvio. Bollettino della Societa’ Geologica Italiana, 117:385–405, 1998.

[7] A. Milia, A. Raspini, and M.M. Torrente. The dark nature of Somma-Vesuvius volcano as evidenced from the 3.5 ka, B.P. Avellino eruption. Quaternary International, 173-174:57–66, 2007.

376

Marine research at CNR

[8] A. Milia, A. Raspini, and M.M. Torrente. The dark nature of Somma-Vesuvius volcano: evidence from the 3.5 ka, B.P. Avellino eruption. Reply. Quaternary International, 192:110–115, 2008b.

[9] A. Milia, A. Raspini, and M.M. Torrente. Evidence of slope instabilities and tsunami associated with the 3.5 ka Avellino eruption of Somma-Vesuvius volcano, Italy. The Geological Society, London, Special Publications, 322:105–119, 2009.

[10] A. Milia, F. Molisso, A. Raspini, M. Sacchi, and M.M. Torrente. Syneruptive features and sedimentary processes associated to pyroclastic flows entering the sea: the AD 79 eruption of Vesuvius, Bay of Naples, Italy. Journal of the Geological Society, London, 165:839–848, 2008a.

[11] A. Milia, L. Mirabile, M.M. Torrente, and J.J. Dvorak. Volcanism offshore of Vesuvius Volcano in Naples Bay. Bulletin of Volcanology, 59:404–413, 1998.

[12] A. Milia, M.M. Torrente, M. Russo, and A. Zuppetta. Tectonics and crustal structure of the Campania continental margin: relationships with volcanism. Mineralogy and Petrolology, 79:33–47, 2003.

[13] F. Bellucci, A. Milia, G. Rolandi, and M.M. Torrente. Structural control on the Upper Pleistocene ignimbrite eruptions in the Neapolitan area (Italy): volcano tectonic faults versus caldera faults. Elsevier, B.V., pages 165–182, 2006.

[14] M. Sacchi, D. Insinga, A. Milia, F. Molisso, A. Raspini, M.M. Torrente, and A. Conforti. Stratigraphic signature of the Vesuvius 79 AD event off the Sarno prodelta system, Naples Bay. Marine Geology, 222-223:443–469, 2005.

[15] G. Aiello, A. Angelino, E. Marsella, S. Ruggieri, and A. Siniscalchi. Carta magnetica di alta risoluzione del Golfo di Napoli (Tirreno meridionale). Bollettino della Societa’ Geologica Italiana, 123:333–342, 2004.

[16] F. Brocchini, C. Principe, D. Castratori, M.A. Laurenzi, and L. Gorla. Quaternary evolution of the southern sector of the Campanian Plain and early Somma-Vesuvius activity: insights from the Trecase well. Mineralogy and Petrology, 73:67–91, 2001.

[17] A. Lima, B. De Vivo, L. Fedele, F. Sintoni, and A. Milia. Geochemical variations between the 79 AD and 1944 AD Somma-Vesuvius volcanic products: Constraints on the evolution of the hydrothermal system based on fluid and melt inclusions.

Chemical Geology, 237:401–417, 2006.

[18] A. Milia and M.M. Torrente. The influence of paleogeographic setting and crustal subsidence on the architecture of ignimbrites in the Bay of Naples (Italy). Earth and Planetary Science Letters, 263:192–206, 2007.

[19] H. Glicken. Criteria for identification of large volcanic debris avalanches. EOS Transactions, AGU, 63:1141, 1982.

377

Marine Geology

[20] P.F. Ballance and M.R. Gregory. Parnell Grift - Large subaqueous volcanoclastic gravity with multiple particle-support mechanisms. SEPM Special Publications, 45:189–200, 1991.

[21] T.B. Fortuin, A.R. Roep, P.A. Sumususastro, T.C.E. van Weering, and W. van der Werff. Slumping and sliding in Miocene and Recent developing arc basins, onshore and offshore Sumba (Indonesia). Marine Geology, 108:345–363, 1992.

[22] G. Rolandi, F. Bellucci, and M. Cortini. A new model for the formation of the Somma Caldera. Mineralogy and Petrology, 80:27–44, 2004.

[23] B.C. Vendeville and V. Gaullier. Role of pore-fluid pressures and slope angle in triggering submarine mass movements: natural examples and pilot experimental models. Kluwer, Dordrecht, pages 137–144, 2003.

[24] H. Sigurdsson, S. Carey, W. Cornell, and T. Pescatore. The eruption of Vesuvius in AD 79. National Geographic Research, 1:332–387, 1985.

[25] N.S. Carey, H. Sigurdsson, C.W. Mandeville, and S. Bronto. Pyroclastic flows and surges over water: an example from 1883 Krakatau eruption. Bulletin of Volcanology, 57:493–511, 1996.

[26] C.W. Mandeville, S.N. Carey, and H. Sigurdsson. Sedimentology of the Krakatau 1883 submarine pyroclastic deposits. Bulletin of Volcanology, 57:512–529, 1996.

[27] F. Legros and T.H. Druitt. Shoreline displacement as a mechanism for the emplacement of ignimbrite in shallow marine environments. Journal of Volcanology and Geothermal Research, 95:9–22, 2000.

[28] A. Freundt. Entrance of hot pyroclastic flows into the sea: experimental observations. Bulletin of Volcanology, 65:144–164, 2003.

[29] M. Edmonds and R.A. Herd. Inland-directed base surge generated by the explosive interaction of pyroclastic flows and sea. Geology, 33:245–248, 2005.

[30] D.V. Kent, D. Ninkovitch, T. Pescatore, and R.S.J. Sparks. Paleomagnetic determination of emplacement temperature of Vesuvius, A.D. 79 pyroclastic deposits.

Nature, 290:393–396, 1981.

[31] G. Mastrolorenzo, P.P. Petrone, M. Pagano, A. Incoronato, P.J. Baxter, A. Canzanella, and L. Fattore. Ercolano victims of Vesuvius in AD 79. Nature, 410:769– 770, 2001.

[32] H. Sigurdsson, S. Cashdollar, and R.S.J. Sparks. The eruption of Vesuvius in AD 79: reconstruction from historical and volcanological evidence. American Journal of Archeology, 86:39–51, 1982.

[33] L. Gurioli, R. Cioni, A. Sbrana, and E. Zanella. Transport and deposition of pyroclastic density currents over an inhabited area: the deposits of the AD 79 eruption of Vesuvius at Herculaneum, Italy. Sedimentology, 49:929–953, 2002.

378

Volcanism, Sedimentation and Tectonics in the Campi Flegrei Area (Italy): an Outlook from Marine Geology A. Milia1 , M.M. Torrente2 1, Institute for Coastal Marine Environment, CNR, Napoli, Italy 2, Department of Geological and Environmental Studies, University of Sannio, Benevento, Italy alfonsa.milia@iamc.cnr.it Abstract The interpretation of a strictly spaced seismic grid (made up of several deep-low resolution multichannel seismic lines and shallow-very high resolution monochannel seismic lines) permitted us to investigate the offshore counterpart of the Campi Flegrei. For the first time we outlined: 1) the interplay between clastic and volcanic units in the stratigraphic succession; 2) numerous previously unknown volcanic units; 3) the fault architecture and kinematics. We identified distribution and typologies (monogenetic volcanoes, ignimbrite units and volcanic domes) of these volcanic units and the three-dimensional architecture of inter-eruptive stratigraphic succession. We detected faults and folds and we recognized fault kinematics, slip rates and fold uplift rates. We were able to reconstruct the paleogeographic evolution and calculate the accommodation curves of this region. The structural pattern features NE-trending normal faults, left-lateral E-W faults and right lateral NW-trending faults. Our work documents the enormous capability of marine geology to investigate submarine volcanic fields as well it has important implications for the identification of the active tectonic structures of Campi Flegrei and for the interpretation of the bradyseism episodes. Finally our basin analysis can play a role in the site selection and lithostratigraphic interpretation of the ICDP Campi Flegrei project.

1

Introduction

Volcanism at Campi Flegrei has been active over the past several hundred thousand years and a huge amount of pyroclastics and lavas was erupted (Figure 1).

This volcanism is associated with Quaternary extension along the Eastern Tyrrhenian margin. Although this region has been the target of intense research, the origin of the volcanism and the geodynamic framework remain a matter of spirited de-

bate. Approximately three million people reside in the greater Naples area, representing one of the most densely populated volcanically active regions on Earth.

Of particular interest regarding volcanic, seismic and geodetic hazards at Campi Flegrei is the phenomenon of slow, vertical ground movements and earthquakes referred to as bradyseism that have affected Campi Flegrei since before Roman times. The Tyrrhenian Sea is an extensional back-arc basin formed at the rear of

Marine Geology

the Neogene Apennine thrust belt. During the Quaternary the eastern Tyrrhenian margin was affected by normal faults that controlled basin architecture and volcanism. Milia [1] and Milia and Torrente [2] reconstructed a detailed chronology of faults characterized by Lower Pleistocene NW-SE normal fault followed

by Middle Pleistocene NE-SW normal faults and NW-SE reactived transfer faults.

The NE-trending extensional fault generated the Naples Bay half graben filled by fourth order depositional sequences arranged in sequence sets that display long term aggradational-progradationalaggradational stacking patterns[3]. Among the controversial themes concerning the Campi Flegrei are the ignimbrite source areas, the occurrence of a Caldera, the relationships between regional faults and ignimbrite emission, the meaning of the bradyseismic episodes as precursor of volcanic eruption. The submerged part of Campi Flegrei was investigated by means of more than 4000 km of single-channel and multi-channel seismic reflection profiles (for technical details and interpretation methodologies of the seismic data see [4]). The interpretation of the seismic reflection profiles permitted us to investigate the relation between the volcanism, stratigraphy and tectonics, a fundamental approach that reveals a new complex evolution of the volcanic area. In particular a detailed seismo-stratigraphic interpretation permit: 1. to investigate the stratigraphic succession in terms of syn-eruptive and intereruptive units; 2. to recognize global and local processes; 3. to interpret the origin of the volcanic units; 4. to individuate faults and folds; 5. to reconstruct their kinematics; 380

6. to find the link, in term of space and time, between faulting and volcanism.

2

Volcanic succession

Our marine geological studies of the submarine part of Campi Flegrei [8, 9, 10, 7, 11, 12, 4, 13, 14, 2, 5, 6, 15] documented and mapped for the first time the occurrence of several syn-eruptive volcanoclastic units mainly formed by voluminous ignimbrite deposits and monogenic volcanoes (Figure 1, Figure 2). These deposits were cross-correlated to the stratigraphic succession outcropping at Naples and to the boreholes drilled on the coast. In particular four superposed wedge-shaped units characterised by a chaotic seismic facies (units Pre-CI, CI, F and NYT) occur in the Eastern part of Campi Flegrei and taper seaward. Unit Pre-CI was linked to the PreCampania Ignimbrite tuffs; it has a maximum thickness of approximately 550 m.

Unit CI corresponds to the Campania Ignimbrite (39 ka) that reaches a maximum thickness of approximately 200 m. Unit F has a variable thickness between 60 and 10 m and features diapiric mounds. Unit NYT corresponds to the Neapolitan Yellow Tuff (15 ka) and is approximately 100 m thick.

The overall pyroclastic wedge thicken toward Campi Flegrei reaching a maximum thickness near the coast where a 850 mthick succession of tuffs was drilled by a geothermal deep well. Monogenic volcanoes of the Campi Flegrei offshore are characterized by chaotic or non-coherent reflections seismic facies and by a mound external form; they will be described from older to younger. The oldest volcanic mound (V5) reposes on a subparallel prevolcanic sedimentary unit (C3a). It is approximately 150 m high, has a diameter

Marine research at CNR

Figure 1: Fisiographic map illustrating the distribution of the syn-eruptive units (monogenic volcanoes and ignimbrites) and faults off Campi Flegrei. GB: Gaia Bank, PB: Mariapia Bank, PP: Penta Palummo Bank, NB: Nisida Bank, MB: Miseno Bank, MD: Monte Dolce Dome, MV: Mirabile volcano, WV: Walther volcano (modified from [5, 6, 7]. of ca 1 kilometre and lies in the Penta Palummo area. Volcanic mounds (150 ka) correspond to the Gaia, Walther, Mariapia, Mirabile and V4 volcanoes and are located in the southern part of the continental shelf.

These volcanoes share similar dimensions featuring diameters on the order of 2 km, average heights of approximately 150-200 m and slopes ranging between 18° and 40°. Two volcanic mounds located in the middle part of the continental shelf correspond to the Penta Palummo (V3, 100 kaold) and the Miseno Bank (39 ka) volcanoes. During the same eruptive period of the Neapolitan Yellow Tuff (15 ka) several

small mounds (NYTtcs, YT1, YT2, YT3) formed in the northern part of the continental shelf. The youngest volcanic units (≤ 6 ka V1, PF; 4.1 ka Nisida bank, [16] occur in the northern part of the continental shelf and correspond to pyroclastic flow deposits. The Monte Dolce volcano (V1) reaches the sea floor and locally emerges with up to 60 m-high relief; it was interpreted as shallow magmatic body overlain by a lava-sediment breccia [7]. Pyroclastic flows (Unit PF) younger than 5 ka where present off Capo Miseno and off Monte Spina. The Nisida Bank volcano, the Nisida shoal and the Nisida island form 381

Marine Geology

the Nisida Complex volcano. In conclusion off Campi Flegrei, various types of submarine monogenetic volcanoes were recognized. The specific environmental conditions suggest that they were generated by phreatomagmatic eruptions. Indeed, external water is easily the most important environmental factor influencing the various evolutionary stages of magma/volcanic systems. Based on (1) the internal and external seismic configuration, (2) the submarine environment of volcano formation, (3) the dimensions of the volcanoes and (4) the comparison with younger volcanoes formed in shallow water in Campi Flegrei, we argue that the submarine volcanoes that lie off Campi Flegrei are scoria cones. Gaia Bank, Mariapia Bank, Walther, Mirabile, Miseno Bank and Penta Palummo volcanoes are characterized by a lateral overlap along an E-W direction, similar dimensions (in the order of 2 km in diameter and 150-200 m in height), and the presence of normal faults along the base of the volcanoes. Positive magnetic anomalies corresponding to these volcanoes suggest the presence of cooled lava bodies. These submarine volcanoes are similar to that of Monte Nuovo, a volcano that formed in 1538 during a one week-long hydromagmatic eruption. Nisida Bank is an example of a submarine stratified volcano. It has a diameter of 1.6 km and height of 80 suggesting diffusely bedded, water-settled ash and lapilli. The top of Nisida Bank Volcano is characterized by a flat, wave-cut surface, where rounded gravel were dredged, indicating the position of the sea level at the time of the volcano formation. The Nisida Complex includes the emergent Nisida island, 600 m in diameter and 110 m in height, which is a tuff cone. The evolution of the Nisida Complex tuff cones can be compared with that of the well studied 382

case of Surtsey (Iceland), a subaqueous to emergent tuff cone [17]. After the eruption ceases, the volcanic cone is rapidly destroyed by subaerial and or wave erosion. Removal of the top of these volcanic centers produces a fairly flat, wavecut plateau and widens the volcanic edifice through the deposition of a surrounding apron of volcaniclastic and bioclastic talus and mass-flow deposits. Dikes and domes are present below a water depth of 115 m and form minor volcanic units. The chaotic facies at the lateral margin of the sills can be interpreted as a texturally complex lava-sediment breccia (peperite) produced by the mixing of domes with subjacent wet sediments. A partially extrusive cryptodome may locally break through the cover and emerge at the surface. In particular the Monte Dolce extrusion occurs in association with normal faults downthrown toward the northeast and southeast.

3

Eustatism and Intereruptive sedimentary deposits

Climatic cycles during the Quaternary affected all the continental margins with a stepwise fall in eustatic sea level that culminated in the Last Glacial Maximum of oxygen isotope stage 2; a much faster sea level rise occurred after 18 ka [18]. Sedimentary deposits on continental margins can readily be subdivided on their internal geometry and stacking pattern and referred to distinct phases of sea level fluctuations (e.g. [19]). Our marine geological studies of the submarine part of Campi Flegrei [8, 9, 10, 1, 7, 11, 12, 4, 13, 14, 2, 5, 6, 15] reported for the first time the occurrence of several clastic units (Figure 2). Epiclastic

Marine research at CNR

processes are extremely significant in volcanic terrains, both in terms of their duration and the volume of sediments that may be transported. The interpretation of the seismic reflection profiles off Campi Flegrei reveals that volcanoes are covered by inter-eruptive volcaniclastic units. These units present an extremely lateral variability due to the isolated source areas, their location and their influence on the hydrodynamic currents. Indeed these latter influence the direction of the transported sediment and the deposition of the suspended sediments that occurs in low energy environments. Slope instability processes have sculpted numerous morphological features on the flanks of the submerged volcanoes in the Bay of Naples off Campi Flegrei. The timing and spatial evolution of sediment failures were defined using seismic profiles and multibeam bathymetry [20, 21]. Rotational slump complexes occur immediately after the growth of volcanoes, with some of them remaining active for a long time after their formation. The depth of the main scars occurring on the volcano flanks are deeper than 150 m. The volcanic landslides that developed above a depth of 150 m were characterized by different triggering factors (angle of slope margin, seismic activity, basement architecture, rapid sea level change, sea currents and high pore-fluid pressure). Instead below a water depth of 150 m, the main factor controlling inter-eruptive erosional/depositional processes is probably gravity. Indeed, the flanks of the volcanoes were seen to be instable, and slumps eroded deposits on the slope and deposited material in the adjacent basin at the slope base. Concerning the erosional-depositional processes affecting volcanoes above a water depth of 150 m, the major controls on offshore clastic facies were sediment supply, wa-

ter depth, relative sea level fluctuation, and hydraulic regime. The development of the volcanoes in the northern part of Naples Bay basin created E-W aligned relief and seaways between these volcanoes. The part of each volcano above wave base was affected by erosion. Shoreface erosion is reflected by flat erosional surfaces located on the flanks or at the top of these volcanoes. The occurrence of the prograding wedge adjacent to the erosional surface suggests that the eroded sediments were contemporaneously deposited below the wave base level. Thicker clastic wedges develop radially from Penta Palummo and Mirabile volcanoes along an approximately 15-km-long coastal area and prograde toward the sea. The thickness of these wedges reaches a maximum of 120 m and extends approximately 4 km away from the volcano, suggesting voluminous source areas. The wedges filled the isolated basins between the volcanoes and formed the continental shelf. Weathering and erosion of pre-existing, poorly or non welded, syn-eruptive deposits can simply release the original pyroclasts or autoclasts and rapidly provide large volumes of recycled material. Consequently the great amount of volcaniclastic products suggests that the original dimensions of Penta Palummo and Mirabile volcanoes were much larger than at present. The sequence stratigraphic framework is important for establishing both the relationship of prograding wedges to continuous shoreface areas and the significance of bounding erosional surfaces in relation to relative sea level changes. The paleo-geographic environment, stratal architecture and plausible physical mechanism of sand transport and deposition suggest that the initial deposition of the sedimentary banks was accompanied by a relative sea level fall 383

Marine Geology

(FST/LST), as documented by the downward shift of the toplap surfaces. This caused a sudden and rapid influx of coarse grained sediment. The mechanism of sand emplacement included direct fluvial supply and reworking by coastal currents within a narrow seaway. The growth of the sedimentary banks was marked by a very rapid (750 cm·yr−1 ) basinward shift in the foresets. The rapid rise of relative sea level produced a rapid landward shift of sedimentation toward the volcanic centers and consequently the wide surface at the top of the FST/LST prograding wedge, which, at this stage, corresponds to the maximum flooding surface. The TST is formed by up to 30-m-thick minor wedges that onlap on the volcano craters and overlie the FST/LST. The sedimentary input diminished and formed a condensed section in the distal area. When the sedimentary units covered the volcanoes, all of the continental shelf simultaneously experienced nondeposition due to the absence of source areas and was rapidly drowned. This analysis leads us to conclude that relative sea level fluctuations were the main process controlling the architecture of the intereruptive sedimentary units. This reconstruction, based on the assumption that erosion occurred gradually during sea level fall and rise, suggests that, during the intereruptive periods, sediment delivery diminished and non-volcanic sediment transport and deposition processes dominated. New volcaniclastic particles created solely by surface weathering and erosion (epiclasts), and biogenic particles, may also have been deposited at this time.

384

4

Tectonics

The structure of the Campi Flegrei offshore is characterized by folds and faults (Figure 1) that formed during Late Quaternary [8, 9, 10, 1, 7, 11, 12, 4, 13, 14, 2, 5, 15]. In Pozzuoli Bay three NW-SE folds (two anticlines and an intervening syncline) have been identified: the syncline controls the physiography of the deepest part of the bay, the northern anticline extends along the coast of Pozzuoli (Figure 1). The folding began 8 Ka ago and calculated uplift rates are up to 20 mm·yr−1 [5, 6]. Three main fault systems occur: northeast-southwest, east-west, and northwest-southeast trending faults (Figure 1). The NE-trending fault swarm off Posillipo Hill and southeast of Campi Flegrei features a normal throw and down throw the southeast block.

It displace the 150 ka-old Gaia Bank volcano, the strictly controlled the location of both the Pia Bank volcano and the Monte Dolce dome and offset the Neapolitan Yellow Tuff with a cumulative throw is approximately 50 m. The east-trending faults are left lateral. One of these structures occurs in the northern part of the Penta Palummo continental shelf (Figure 1) and abruptly offsets the 12-9 Ka old prograding wedge of the TST unit. The northwestsoutheast faults display normal and rightlateral slip components. In the Bay of Pozzuoli northwest-southeast-trending en echelon segments of right-lateral normal faults downthrow the central area of the basin by a few meters at the same time as the deposition of the first sedimentary unit above the Neapolitan Yellow Tuff.

One of the northwest-southeast faults displays a dextral offset of the oldest volcanic unit V4. The activity of this fault was contemporaneous with the deposition of the 12-9 Ka old prograding wedge of the

Marine research at CNR

TST unit. In conclusion East-trending leftlateral faults, Northeast-trending normal faults, Northwest-southeast right-lateral faults and Northwest- trending folds are compatible with the activity of an easttrending left-lateral transtensional shear zone [5, 6].

5

Discussion

The architecture of the Campania Ignimbrites in the Bay of Naples is characterised by a thick wedge that overlies Middle Pleistocene marine sediments. In the Eastern part of Campi Flegrei the superimposition of the volcanoclastic wedges occurred in the slope/basin area. The latter resulted from three geologically instantaneous episodes that produced dramatic changes in the physiographic environment.

In particular, the syn-eruptive and intereruptive products of the Pre-CI tuffs and CI tuffs completely filled the accommodation space in the slope/basin area. In succession the products of the NYT were initially deposited in shallow water and in a sub-aerial setting, where the thickness of the volcano was greater than the water depth. Within the Bay of Pozzuoli the pyroclastic wedge deposited in the slope/basin environment thickens toward ENE, and its base dips by approximately 11° landward. We maintain that the volcano-tectonic subsidence is controlled by the activity of a NW-trending fault indicated by the elongation of the PreCI tuffs depocenter at Campi Flegrei. This architecture suggests an increase in the accommodation space balanced by the deposition of the syn-eruptive and inter-eruptive products of the volcano. In this case the paleo-basin (approximately 350 m deep before the deposition of the ignimbrite units) results filled. Consequently, we conclude,

that in area West of Naples the thick sineruptive and inter-eruptive ignimbrite deposits fill both the accommodation space of the basinal environment and that created by a tectonic-controlled subsidence. In order to understand the relationship between volcanic activity and tectonic subsidence during the Late Quaternary, the authors reconstructed the accommodation curve for two specific points of the Naples Bay basin (Penta Palummo and Pozzuoli Bay). The accommodation space curve is the result of the algebraic sum of the eustatic curve (Martinson et al., 1987), the curve of space filled by volcanics and sediments (calculated assuming in the depth conversion a velocity of 1600-1800 m¡sâ&#x2C6;&#x2019;1 compatible with very shallow mainly medium-grained sediments and pyroclastic rocks) and the subsidence curve (associated with faulting and folding). A basin analysis was performed [1, 22, 6, 13, 14] for the last 200 ka and the paleobathymetry was estimated by reconstructing the stratigraphic architecture and depositional environments. Sediment accumulation is plotted through time.

In the Penta Palummo area there was no tectonic subsidence and the rate of the accommodation space changes according to the eustatic curve minus the space filled by volcanics and sediments. Here Middle Pleistocene sedimentary strata extend horizontally from the intraslope basin northward, thus supporting a paleogeographic scenario preceding volcanism. The latter is characterized by a paleo-water depth of approximately 520 m (top of C2 sequence).

The space between the paleo-sea-floor and the sea level was filled by isolated monogenetic volcanoes (V5, V4, V3) and clastic marine sediments until the present water depth of 75 m. Sedimentary units prograde north- and southward from these volcanoes, indicating that the latter became 385

Marine Geology

the source areas for the sediments during the intervals of volcanic standstill. The vertical aggradation of both volcanic and clastic deposits gave rise to the emersion of this area during the last glacial maximum and to subaerial erosion when the curve displaying the vertical aggradation of the volcanic and sedimentary deposits intersected the accommodation curve. The paleogeography of Pozzuoli Bay before the onset of volcanism is that of the adjacent intraslope basin and Penta Palummo area characterized by a paleowater depth of approximately 520 m. The subsidence curve shows a negligible value until 15 ka. Afterward an instantaneous increase of subsidence due to faulting and the Holocene enucleation of the syncline can be seen. The accommodation curve presents a rapid increase over the last 15 ka corresponding to basin subsidence. The space filled by sediments increased linearly producing a gradual decrease in the water depth until this curve intersected the accommodation curve during the last glacial maximum producing an emersion of the area and subaerial erosion (documented by the unconformity at the top of the Forced regression system tract Lowstand system tract, FST-LST, [13]).

The mean rate of sedimentary supply increased over the last 15 ka as indicated by the thick Holocene succession (G3, G2, G1). Repeated volcanic events and a rapid infill of the basin were documented for the Penta Palummo area that experienced a dramatic physiographic change, from a

slope-basin (in the Middle Pleistocene) to a shelf (during Late Quaternary) due to volcanic vertical aggradation. A more gradual physiographic change from a slope-basin to a shelf occurred in Pozzuoli Bay where the basin infill, caused by clastic vertical aggradation, was later followed by a localized tectonic subsidence.

6

Conclusions

These marine geological studies produced key results for the understanding of the geologic evolution of Campi Flegrei. We provided a very detailed picture of the volcanic activity in the marine area, started 200 ka, that records the underwater component of onshore flows entering the sea.

We question the existence of a caldera collapse on the basis of our structural and stratigraphic findings (the presumed ring fault is absent whereas a complex fault pattern does occur and the computed local volcano-tectonic subsidence coeval to the CI and NYT eruptions resulted much smaller than that postulated for the subsiding caldera block).

We fornished geological constrains (stratigraphic sequence characterized by impermeable layers and active compressional structure at Pozzuoli producing fluid overpressures) to a quantitative model for bradyseism at Campi Flegrei with implications for future volcanic eruptions [23, 24, 25].

References [1] A. Milia. Aggrading and prograding infill of a pery-tyrrhenian basin (Naples Bay, Italy). Geo-Marine Letters, 19:237â&#x20AC;&#x201C;244, 1999.

[2] A. Milia and M.M. Torrente. Tectonics and stratigraphic architecture of a pery-

386

Marine research at CNR

Tyrrhenian half-graben (Bay of Naples, Italy). Tectonophysics, 315:297–314, 1999.

[3] A. Milia and F. Giordano. Holocene stratigraphy and depositional architecture of eastern Pozzuoli Bay (Eastern Tyrrhenian Sea margin, Italy): the influence of tectonics and wave-induced currents. Geo-Marine Letters, 22:42–50, 2002.

[4] A. Milia and M.M. Torrente. The influence of paleogeographic setting and crustal subsidence on the architecture of ignimbrites in the Bay of Naples (Italy). Earth and Planetary Science Letters, 263:192–206, 2007.

[5] A. Milia and M.M. Torrente. Fold uplift and syn-kinematic stratal architectures in a region of active transtensional tectonics and volcanism, Eastern Tyrrhenian Sea. Bulletin of the Geological Society of America, 112:1531–1542, 2000.

[6] A. Milia and M.M. Torrente. Late Quaternary volcanism and transtensional tectonics at the Campania continental margin Bay of Naples, Italy. Mineralogy and Petrology, 79:49–65, 2003.

[7] A. Milia. The stratigraphic signature of volcanism off Campi Flegrei (Bay of Naples Italy). Geological Society of America SP Pubbl., 464:155–170, 2010.

[8] A. Milia. Evoluzione tettono-stratigrafica di un bacino peritirrenico: Il Golfo di Napoli. PhD Thesis, page 184 p, 1996.

[9] A. Milia. Le unit`a piroclastiche tardo-quaternarie nel Golfo di Napoli. Geografia Fisica Dinamica Quaternaria, 21:147–153, 1998.

[10] A. Milia. Stratigrafia, strutture deformative e considerazioni sull’origine delle unit`a deposizionali oloceniche del Golfo di Pozzuoli (Napoli). Bolletino Societ`a Geologica Italiana, 117:777–787, 1998.

[11] A. Milia, F. Giordano, and G. Nardi. Stratigraphic and structural evolution of Naples Harbour over the last 12 ka. Giornale di Geologia, 60:41–52, 1998.

[12] A. Milia, M.M Torrente, and G. Nardi. Recent tectonic and magmatic features off the coast of Naples. Giornale di Geologia, 60:27–39, 1998.

[13] A. Milia, M.M. Torrente, and L. Giordano, F. and. Mirabile. Rapid changes of the accommodation space in the Late Quaternary succession of Naples Bay, Italy: the influence of volcanism and tectonics. Development in volcanology, 9:53–68, 2006.

[14] A. Milia, M.M. Torrente, and F. Giordano. Active deformations and volcanism offshore Campi Flegrei, Italy: new data from high resolution seismic reflection profiles. Marine Geology, 171:61–73, 2000.

[15] F. Bellucci, A. Milia, G. Rolandi, and M.M. Torrente. Structural control on the Upper Pleistocene ignimbrite eruptions in the Neapolitan area (Italy): planar volcano tectonic faults versus caldera faults. 9:165–182, 2006.

387

Marine Geology

[16] L. Fedele, D. Insinga, A.T. Calvert, V. Morra, A. Perrotta, C. Scarpati, and S. Lepore. 40Ar/39Ar dating of tuff vents in the Campi Flegrei caldera: towards a new chronostratigraphic reconstruction of the volcanic activity during the Holocene.

Epitome Geoitalia, (3):445, 2009.

[17] B.P. Kokelaar. The mechanism of Surtseyan volcanism. Journal of the Geological Society of London, 140:939–944, 1983.

[18] D.G. Martinson, N.G. Pisias, J.D. Hays, J. Imbrie, T.C. Moore, and N.J. Shackleton.

Age dating and orbital theory of the Ice Ages: Development of a high resolution 0 to 300 000 year chronostratigraphy. Quaternary Research, 27:1–29, 1987.

[19] H.W. Posamentier and P. Vail. Eustatic control on clastic deposition. II. Sequence and system tract models. SEPM, 42:125–154, 1988.

[20] A. Milia, F. Giordano, G. Nardi, and M.M. Torrente. Submarine slides off Posillipo Hill (Naples, Italy). Giornale di Geologia, 60:17–25, 1998.

[21] A. Milia, M.M. Torrente, and F. Giordano. Gravitational instability of submarine volcanoes offshore Campi Flegrei (Naples Bay, Italy). Development in volcanology, 9:69–83, 2006.

[22] A. Milia and M.M. Torrente. Evoluzione tettonica della Penisola Sorrentina (margine peritirrenico campano). Bolletino Societ`a Geologica Italiana, 116:487– 502, 1997.

[23] R.J. Bodnar, B. DeVivo, A. Lima, H.E. Belkin, C. Cannatelli, and A. Milia. Quantitative evaluation of magma degassing and ground deformation (Bradyseism) at Campi, Flegrei, Italy. Geology, 35(9):791–794, 2007.

[24] B. De Vivo, A. Lima, R.I. Bodnar, A. Milia, and F.J. Spera. Il rischio eruzione nei Campi Flegrei. Le Scienze, (settembre):96–103, 2009.

[25] A. Lima, B DeVivo, F.J. Spera, R.J. Bodnar, A. Milia, C. Nunziata, H.E. Belkin, and C. Cannatelli. Thermodynamic model for uplift and deflation episodes (bradyseism) associated with magmatic–hydrothermal activity at the Campi Flegrei (Italy). Earth Science Review, 97:44–58, 2009.

388

Marine research at CNR

Figure 2: Description and interpretation of the eruptive and inter-eruptive units of Campi Flegrei.

389

Research of Marine Sand Resources for Beach Nourishment: an Applied result of Geological Map of the Adriatic Sea (1:250000) A. Correggiari1 , M. Aguzzi1 , F. Foglini1 , A. Gallerani1 , A. Remia1 1, Institute of Marine Sciences, CNR, Bologna, Italy 2, ARPA Environmental Agency of Emilia Romagna, Bologna, Italy anna.correggiari@bo.ismar.cnr.it Abstract Shorelines and coastal development will be even more vulnerable to hazards in the future. Need for offshore sand for nourishment will increase but volumes for sustainable shore protection are uncertain for many regions. Beach nourishment with sand derived from river or coastal borrow sites has been the preferred most common method of shoreline stabilization method in Italy for several decades. This practice has increased rapidly over the last decade to the point that the search of alternative sources of sand became an issue. Better understanding of the shelf geology can aid our ability to plan for sustainable use of offshore sands. Since the â&#x20AC;&#x2122;80 the Adriatic shelf has been studied to identify potential sand deposits available for extraction. Information about the geology of adriatic shelf regions, the character of the seafloor, and samples comprising the seafloor and subbottom have been aquired by ISMARCNR-Bologna as the result of several national and international projects included the Geological Adriatic Map (at the 1:250000 scale). Pleistocene/Holocene relative sea level rise submerged a wide portion of the northern and central Adriatic paleoalluvial plain has been progressively drowned. For each step of the relative sea level rise a barrier lagoon system has been identifed and the amount of sand has been quantified.The potentially sand resources are available to a confined area in the central portion of the basin. The average grain size is fine sand, well to moderately well sorted.

1

Introduction

Several factors may cause beach erosion, most of which are natural.

Beaches are constantly moving, building up here and eroding there, in response to oceanographic factors (waves, winds, storms, relative sea level change and supply fluctuation). Some beaches are also destroyed by human activity when harbors or any other anthropic structure disrupt the fragile balance of erosion and deposition in the

coastal environment. Out of the 5961 km coastline referred to in the Italian Coastal Atlas, 3612 km are sandy beaches, of which more than 960 km are classified as subject to erosion [1]. Many coastal areas are facing chronic long-term shoreline erosion problems and eroding beaches require periodic beach replenishment to maintain stable berms for coastal protection and recreation. This paper reports on sand search investigations offshore as

Marine research at CNR

strategic resources for beach restoration.

Geologic factors control the location of offshore sand resources. For this reason, putting sand resource studies in the context of the geological setting of the area is critical for developing an ability to predict sand potential availability. Demand is growing in Mediterranean and worldwide for information about the geology of offshore continental shelf regions, the character of the seafloor and sediments, comprising the seafloor and subbottom. The Adriatic shelf, particularly offshore Emilia- Romagna and Veneto regions, has been the focus of these studies for the past 25 years with widely varying results. Geophysical studies to locate potential borrow areas, identify sediment quantities, investigate sediment characteristics, and rank candidate sites are usually undertaken in multiple phases. The method draws together local geological information and data to generate the final sand search deliverables.

2

Adriatic coastal setting

On much of its extent the Adriatic coast has been steadily eroding in the recent past.

Sediment transport by the Po and other Apennine rivers between 1960-1980, was reduced to about 30% of its previous values [2], due to the hydrogeological control of the catchment areas and to dam construction (artificial lakes, etc.) accompained by intense quarrying of graveland sand from riverbeds. These factors have led to an enhanced erosive activity in the coastal environment. The decrease of land erodibility resulting from reforestation of the Italian peninsula after the second Word War contributed to a reduced sediment supply to the sea. Growing shores are “nourished”

by material that has been eroded from somewhere else. Any attempt to reduce coastal erosion in one area will result in reduced deposition elsewhere, “starving” another shoreline. Erosion and accretion are therefore two faces of the same process, and their balance may change at extremely slow rates or make dramatic changes in the shoreline within a human lifetime. Sand bedload that used to flow down rivers contributing to beach maintenanceis no longer reaching the coast, resulting in a substantial narrowing of the beach area thereby affecting also the recreational opportunities. Since the 1960s several strategies to prevent beach erosion have been applied on the Italian coasts. Attached and detached breakwaters, groins, jetties built on the coast have changed the natural physiography of the shoreline partially inhibiting the flow of beach sand by longshore drift. These features may actually accelerate erosion or change the ways in which the shoreline can be used. In fact in some areas of the north Adriatic coast, the breakwaters installed to stop beach erosion have been so effective that the beach has become a mud flat with severe environmental pollution problems [3]. Along the Adriatic coast, in addition to the construction of stable beach protections, nourishment operations contributed to preserve beaches from erosion. Protective structures occupy about 35% of the 1630 km of the Italian Adriatic coast and large supply of suitable sand used for beaches accretion was mined from river beds until 1984 when this activity was almost forbidden. In a period of decreasing sediment supply a good strategy is to search an alternative source of sand outside the coastal and river system. In Italy only in the recent years offshore sand become the main source for beach nourishment.

391

Marine Geology

Figure 1: Wide portion of the northern and central Adriatic alluvial plain of the glacial time has been progressively drowned during the late Pleistocene/Holocene relative sea level rise with concurrent widening of the continental shelf area of the Adriatic. The depth/time plot is referred to the most importat drowned barrier lagoon systems in the Adriatic basin.

3

Late Quaternary Geological setting of the Adriatic

The Synthetic geological mapping provides the basis for any applied environmental or natural resources study by making basis information available to wide range of end users. In offshore geological mapping, extensive areas can be covered at much smaller scale than those used for geological mapping on land because of the large costs implied. We must im392

prove knowledge about the marine environment in order to understand the impact of human activities on the sea and to ensure that various human projects are implemented in a sustainable way. Most European countries have extensive geological mapping project for their offhore Exclusive Economic Zone. With the Geological mapping project of the Italian Seas, in collaboration with ISPRA (ex APAT) Servizio Geologico dâ&#x20AC;&#x2122;Italia, ISMAR CNR Bologna extended the geological mapping to the entire italian portion of the Adriatic basin.

The modern Adriatic Sea is a narrow epi-

Marine research at CNR

Figure 2: Schematic evolution of a barrier lagoon system during trangressive erosional processes. continental basin with a low longitudinal topographic gradient (ca 0.02°), whereas the maximum shelf gradient along the central Adriatic is on the order of 0.5°. During the late Pleistocene/Holocene relative sea level rise (between ca 18 and 5.5 kyr BP) [4, 5, 6, 7, 8] wide portion of the northern and central Adriatic alluvial plain of the glacial time has been progressively drowned, with concurrent widening of the continental shelf area of the Adriatic (Figure 1). Across the low-gradient northern shelf, the stepwise, high-amplitude relative sea-level rise favoured the deposition and in-place drowning of different generations of transgressive barrierâ&#x20AC;&#x201C;lagoon systems.

Along the western side of the Adriatic shelf and seaward of the modern shoreline, the late-Holocene mud wedge, a continu-

ous belt of deltaic and shallow-marine deposits, overlies the available transgressive sand deposits. The exploitable trangressive sand bodies comprising old beach deposits, are outcrop only in the axial part of the basin far from the recent muddy sediment.

4

Description of potential beach compatible deposits

The comprehensive review and analysis of geological, geophysical, and geospatial data provides new insight into sand search methodologies for offshore sand resources [9]. Using the Adriatic shelf geological mapping project, benefits of the review and analysis of acquired geological 393

Marine Geology

Figure 3: The ravinement surface, a peculiar sedimentary shell debris in the transgressive deposit caused by waves erosion as relative sea level rises and the shoreline moves landwards. The vibracores were sampled in a drowened barrier lagoon system marine sand offshore Emilia-Romagna region. ([8], Correggiari et al. in press) datasets become manifest. During the last decades the increasing amount of data acquired by ISMAR CNR Bologna provides unique opportunities to summarize knowledge of geology and shelf geomorphology with existing geotechnical and geophysical data that facilitate identification of sand resources. This kind of understanding abbreviates the need to conduct random geophysical and sampling surveys over large expanses of the seabed and is more efficient and economical because only potential deposits are targeted. Some types of offshore sites can be described as linear sand bodies, including remnant shoal features, ebb or flood tidal shoals, drowned barrier islands, oblique sand ridges, longshore bars, trough sand accumulations, and migratory

394

sand spits attached or unattached to tidal inlets. In the northern Adriatic coastal areas several morphological and sedimentological investigations have been carried out along active barrier islands in order to evaluate the sand reservoir potential associated with ebb-tidal deltas [10].

The preservation potential of a trangressive barrier/lagoon system in the low gradient adriatic shelf is limited to its basal portion and is function of its original thickness ad volume. The Figure 2 shows the schematic evolution of a barrier lagoon system and points out the importance of sediment facies characterisation of each sand deposit [11]. The ravinement surface, a peculiar sedimentary signature in the transgressive deposit, forms through waves erosion

Marine research at CNR

Figure 4: Chirp sonar profiles, and vibracore of a sand deposit in the central Adriatic around 80-90 metres water depth. In yellow the sand available. The deposit thickness riches up to 4 m and is considerably homogeneus in terms of composition and grain size as shown in plots and map. as relative sea level rises and the shoreline moves landwards, causing reworking of both older and newly deposited sediment on the shoreface, and separating non marine (below) from marine (above) deposits (Figure 2). The ravinement surface is typically associated to shelly sands and hash deposits largely made up by mollusc remain sourced from paralic to fully marine environments. In the transgressive sand deposit used for nourishment this bioclastic lag may create some problems during dredging operations: the high shells concentration can block the fluidized sediment during the pumping process. [8].

The Adriatic available borrow sand deposits lays in patches inside paleo-barrier

lagoon systems and comprise sediment classified as fine sand (D50=0,160 mm) well to moderately well sorted. The mineralogical and chemical composition of Adriatic sand deposits classifies them as litharenites, with a variable mixture of cabonates, and silicates. Carbonates occurred in sediments as detrital granules of calcite or dolomite frequently associated with shell fragments [8, 12, 13]. Three example of sand research results are shown in Figure 4 and Figure 5. The first one comprise several patches of transgressive deposits with positive bathymetry at 80-90 m water depth in central Adriatic (Figure 4). They represent a reworked complex coastal wedge with barriers lagoon envi-

395

Marine Geology

Figure 5: In the map 3 examples of transgressive deposits where borrow sites for sand extraction have been identified. The number 1 is described in Figure 4. The number 2 is related to offshore of Emilia-Romagna region and identified with A and C, while number 3 describes the starved northern shelf were asymmetric large sand waves contribute to the sand reservoir of the Adriatic basin. ronment dated around 14-16 calibrated kyr BP [14, 15]. The second one is represented by another group of transgressive deposits offshore Emilia-Romagna region studied

by ARPA Ingegneria Ambientale EmiliaRomagna and ISMAR since 1984 already used as borrow sites for beaches nourishment [16, 17, 18, 19]. Several outcrops of trangressive lithosomes, located from 36 m to 42 m water depth, has been dated between 8 to 12 calibrated kyrs BP (Figure 5). The third relict system of starved and reworked sand deposits is located offshore northern coast of the Adriatic basin and has been studied by a collaborative project by Regione del Veneto and ISMAR CNR Bologna with geophysical and geognostic surveys [6, 20, 13]. It is important to note that while the majority of potential 396

Adriatic borrow deposits identified consist of ridges and remnants of barrier lagoonal system, in the future it is likely that the range of deposit types will be expanded to include paleo-channels, paleo-deltas, and other buried sand deposits.

The volume of the entire potental fine sand available reservoir outcropping in the italian portion of the adriatic shelf seafloor has been extimated in ca. 240 x106 cubic metres.

5

Conclusions

Compilation of the surficial marine geological maps, using high-resolution seismic profiles, with the addition of new narrow grid detailed geophysical suveys an sedi-

Marine research at CNR

ment vibracores in targeted locations, is an effective way to evaluate potential borrow sites on any continental shelf.

Offshore sand resource, as any other nonrenewable resource, must be managed on a long-term, large scale, wide basis to ensure that environmental damage will not occur as a result of continual and prolonged use. Sand sources that are to be used re-

peatedly may require additional biological and physical monitoring to avoid unacceptable impacts to the marine and coastal environments. For these reasons it is necessary to address future efforts through improving monitoring protocols to evaluate the long-term impact of offshore dredging operations on the marine environments.

References [1] E. Valpreda and U. Simeoni. Assessment of coastal erosion susceptibility at the national scale: the Italian case. J Coastal Conserv, 9(1):43–48, 2003.

[2] AQUATER. Studio generale per la difesa delle coste prima fase. Rapporti di settore.

Regione Marche. II, 1984.

[3] A. Correggiari, F. Frascari, S. Miserocchi, and D. Fontana. Breakwaters and eutrophication along the Emilia-Romagna coast. pages 277–290, 1992.

[4] F. Trincardi, A. Correggiari, and M. Roveri. Late Quaternary transgressive erosion and deposition in a modern epicontinental shelf: The Adriatic semienclosed basin.

Geo-Marine Letters, 14(1):41–51, 1994.

[5] F. Trincardi, A. Cattaneo, A. Asioli, A. Correggiari, and L. Langone. Stratigraphy of the late-Quaternary deposits in the central Adriatic basin and the record of short-term climatic events. Palaeoenvironmental Analysis of Italian Crater Lake and Adriatic Sediments. Memorie dell’Istituto Italiano di Idrobiologia,, 55:39–70, 1996.

[6] A. Correggiari, M.E. Field, and F. Trincardi. Late-Quaternary Transgressive large dunes on the sediment-starved Adriatic shelf. in M. De Batist and P. Jacobs, (eds.)

Geology of Siliciclastic Shelf Seas Geological Society of London Special Pubbl., 117:155–169, 1996.

[7] A. Correggiari, M. Roveri, and F. Trincardi. Late Pleistocene and Holocene evolution of the north Adriatic sea. II Quaternario - Italian Journal of Quaternary Sciences AIQUA, 9(2):697–704, 1996.

[8] A. Correggiari, D. Carr`a, A. Cattaneo, D. Penitenti, and F. Trincardi. Caratterizzazione delle aree di prelievo dei sedimenti a mare. Programma di messa in sicurezza dei tratti critici del litorale Emiliano-Romagnolo mediante ripascimento con sabbie sottomarine. Arpa Ingegneria Ambientale - Regione Emilia Romagna, Relazione Specialiastica, page 71, 2001.

397

Marine Geology

[9] C.W. Finkl, J.L. Andrews, T.J. Campbell, L. Benedet, and J.P. Waters. Applying geologic concepts coupled with historical data sets in a MIS framework to prospect for beach-compatible sands on the inner continental shelf: the eastern Texas Gulf coast. Journal of Coastal Research, 20(2):533–549, 2004.

[10] G. Fontolan, S. Pillon, F.Delli Quadri, and A. Bezzi. Sediment storage in the northern Adriatic ebb-tidal deltas, Italy: sand use potential and GIS database. Journal of Coastal Research, 50:917–921, 2007.

[11] S. Penland, R. Boyd, and J.R. Suter. Transgressive depositional systems of the Mississipi delta plain: a model for barrier shoreline shelf sand evolution. Journal of Sedimentary Petrology, 58(6):932–949, 1988.

[12] A. Correggiari, A. Remia, and F. Foglini. Progetto di caratterizzazione dei depositi sabbiosi sommersi presenti sulla piattaforma alto adriatica potenzialmente sfruttabili come cave di prestito per il ripascimento costiero nell’ambito del progetto PRASTAVO. Primo SAL stato avanzamento lavori. Relazione tecnica, page 38, 2007.

[13] A. Correggiari, A. Remia, and A. Gallerani. Esecuzione di analisi granulometriche e di suscettivit`a magnetica nell’ambito di Progetti per il ripascimento costiero. Relazione finale, 2008.

[14] A. Correggiari and A. Remia. Ricerca ed individuazione di depositi sabbiosi in Adriatico Centrale. Relazione Tecnica CNR, page 51, 2005.

[15] A. Correggiari and A. Cattaneo. Quality and quantity of sand deposits on the Adriatic continental shelf. EMSAGG-CRIA Conference 7-8 Maggio 2009, Roma, 2009.

[16] P. Colantoni, M. Preti, and B. Villani. Sistema deposizionale e linea di riva olocenica sommersi in Adriatico al largo di Ravenna. Giornale di Geologia, 52(1):1–18, 1990.

[17] M. Preti. Ripascimento di spiagge con sabbie sottomarine in Emilia-Romagna.

Studi Costieri, 5:107–134, 2002.

[18] Various Authors. BEACHMED-e http://www.beachmed.it/Beachmede/Pubblications /tabid/115/Default.aspx. 2008.

[19] A. Correggiari, M. Aguzzi, A. Remia, and M. Preti. Caratteristiche sedimentologiche e stratigrafiche dei giacimenti sabbiosi in Mare Adriatico Settentrionale utilizzabili per il ripascimento costiero. Studi costieri, in press, 2011.

[20] G. Cecconi and G. Ardone. La protezione delle spiagge della laguna di Venezia. in Atti seminario Riqualificazionee salvaguiardia dei litorali: idee proposte confronti tra esperienze mediterranee, pages 58–65, 2003.

398

Influence of Seawater Circulation on the Evolution of Ultramafic Spreading Ridges C. Boschi1 , A. Dini1 , G. Frueh-Green2 1, Institute of Geosciences and Earth Resources, CNR, Pisa, Italy 2, Department of Earth Sciences, ETH-Zurich, Zurich, Switzerland c.boschi@igg.cnr.it Abstract A mid-oceanic ridge is an underwater mountain range formed by the diverging motion of tectonic plates that triggers the production of new lithosphere and the spreading of seafloor. The mid-ocean ridges of the world are connected and form a single global ridge system. It represents the longest mountain chain in the world, totaling 60,000 km. Mid-ocean ridges are geologically active, with new magma constantly intruded, and erupted onto the ocean floor, along the ridge axes. If for some reason the magma supply stops for a period of time, the crust must stretch to match the plate motion producing multiple cracks (i.e. faults) in the lithosphere and large exposure of deeper mantle rocks at the seafloor. These rocks, known as abyssal peridotites, due to the continuous interaction with seawater, are variably hydrated and transformed in serpentinite. Low-angle detachment faults, exposing the serpentinized mantle rocks at the seafloor, act as pathways for seawater allowing complex interactions with ultramafic/mafic rocks at depth with formation of serpentinitic and talc-rich rocks. Here, we illustrate our recent petrographic, geochemical and isotopic studies on a peridotite-dominated underwater massif, the Atlantis Massif, located at 30째N on the Mid-Atlantic Ridge. The consequences of its hydratation and alteration in relation to the geochemical and geodynamical evolution of this area are discussed in details.

1

Introduction

Following the theory of plate tectonics, the lithosphere is broken up into what are called tectonic plates that move in relation to one another at one of three types of plate boundaries: convergent (or collisional), divergent (also called spreading centers) and transform boundaries. Earthquakes, volcanic activity, mountain-building, and oceanic trench formation occur along plate boundaries.

At divergent boundaries, two plates move apart from each other and the space that

this creates is filled with new lithosphere sourced from magma that forms below, forming a massive underwater mountain range, the oceanic ridge system. If for some reason the magma supply stops for a period of time the crust must stretch to match the plate motion. If the crack is not vertical, it almost never is, the lower part of the crust can be pulled sideways out from under the upper layer along a dipping fault (so called detachment fault) and expose deeper mantle rocks at the seafloor, the abyssal peridotite. This type of rocks crop out frequently in slow to ultraslow spread-

Marine Geology

Figure 1: Schematic representation of the history of oceanic lithosphere, from its birth at spreading ridge to the â&#x20AC;?graveâ&#x20AC;? at convergent margin (subduction zone). Major processes discussed in the text are reported. ing ridge environments that are characterized by low magma supply and/or complex tectonic processes related to spreading. In these environments, they may comprise â&#x2C6;ź20% or more of the oceanic crust. Oceanic lithosphere newly formed at spreading ridges is variably geochemically and mineralogically modified by interaction with seawater. Then it moves, as a belt conveyor, for hundreds of kilometers to the convergent boundaries, where during subduction, it releases several elements to the upper plate (Figure 1). For this reason, the understanding of seawateroceanic lithosphere interaction is fundamental not only for processes occurring at spreading ridges (hydration of lithosphere, changes in lithosphere rheology, modification of ocean chemistry, formation of black-smokers and ore deposits, etc.), but also for the spectacular geological processes that take place where the oceanic 400

plates subduct into the mantle (hydration of upper plate by fluids released from the subducted lithosphere, production of arc magmatism, formation of giant metal deposits, etc.).

Taking into account this scenario, the interaction with seawater and the consequent hydration of the abyssal peridotites at slow spreading ridges is a fundamental process (so called serpentinization) that has significant geophysical, geochemical and biological importance for both the global marine system and the subduction zone environments. Hydration is accompanied by a decrease in bulk density and a change in rheology that directly affect the strength and physical properties of the mantle, the magnetic and gravity signatures, and the seismic velocities [3].

The depth to which serpentinization may occur is controlled by the depth to which seawater can penetrate into the oceanic

Marine research at CNR

Figure 2: Plate boundary geometry and seafloor morphology near the intersection of the Mid-Atlantic Ridge (MAR) and the Atlantis Transform Fault. Broad, elevated massifs with spreading-parallel corrugations in this area are interpreted as oceanic core complexes (modified from [1]). crust and upper mantle. Tectonic processes and fracture permeabilities largely control the depth of seawater circulation.

Extreme extension of oceanic lithosphere during seafloor spreading at slow to ultraslow spreading ridges creates oceanic core complexes (OCCs) that are broad, elevated massifs (few tens of kilometers across) where deep crustal and/or upper mantle rocks have been unroofed, uplifted and consequently altered. One of the most studied oceanic core complex is the Atlantis Massif, located at 30째N at the intersection of the Mid-Atlantic Ridge (MAR) and the Atlantis Transform Fault (Figure 2). This dome-like massif has been the subject of different oceanic expeditions:

in particular, the southern and the LCHF were mapped and sampled during three cruises: AT3-60 (MARVEL expedition, 2000, [4]) and AT7-34 [5] using the submersible Alvin with the R/V Atlantis and in 2005 using the remotely operated vehicles (ROV) Hercules and Argus onboard the R/V R.H. Brown.

The multidisciplinary project, led by Debbie Kelley in collaboration with different institutes of research including IMP-ETH (Switzerland) and IGG-CNR (Pisa, Italy) focused on the geological and biological consequences of the interaction between seawater and the Atlantis massif. One of the aspects that we investigated concerns the interaction between seawater and peri401

Marine Geology

Figure 3: Hydrothermal vents in white and black. Photomosaic (courtesy of M. Elend and D. Kelley, U. Washington) of a structure, composed of carbonate and over 10 m in height, typical of those at Lost City identified by [2]. Water depth here is 700–800 m (modified from [2]). dotites along the 100 m thick detachment fault responsible for the uplift of the massif. Here, we summarize our results highlighting the importance of our research.

2

The Atlantis Massif

The dome-like Atlantis Massif is located at 30°N at the intersection of the MidAtlantic Ridge (MAR) and the Atlantis Transform Fault [6]. The central dome 402

of the massif shows a distinct corrugated morphology, which is interpreted as the surface expression of a low-angle detachment fault that led to uplift and exposure of lower crustal and upper mantle sequences over the past 1.5–2 Myr. The domed morphology and exposure of variably altered and deformed gabbros and peridotites during phases of asymmetric extension and detachment faulting are considered key components that define oceanic core complexes (OCCs). To the south,

Marine research at CNR

the Atlantis Massif is truncated by a steep, faulted escarpment with >3800 m of relief adjacent to the Atlantis Transform Fault valley. The southern ridge has experienced significantly greater uplift than the central portion of the massif and shoals to 730 m below sea level. An impressive hydrothermal field, the so called Lost City Hydrothermal Field (LCHF), lies on a terrace at the top of this southern ridge and hosts numerous active and inactive carbonate-brucite chimneys that tower up to 60 m above the seafloor (Figure 3).

The LCHF is the surface expression of warm (40–90°C), high pH (9–11) fluids emanating from fault zones that tap a region of active serpentinization in the underlying peridotites [2, 5]. The southern ridge of the massif consists primarily of variably altered and deformed peridotites (70% of total samples) that have been intruded by smaller bodies of gabbro, pyroxenite, and minor basalt, and are covered with a flat-lying, 1–2 m thick sedimentary sequence. The peridotites are primarily depleted spinel harzburgites, consisting of olivine, orthopyroxene, and chromium spinel with minor clinopyroxene. The rocks are highly serpentinized (from 70% to 100%) primary olivine and orthopyroxene are replaced by serpentine and magnetite, and minor (<5 vol%) chlorite, amphibole and talc. Field studies in 2003 provided detailed observations and samples of a well-defined, continuous, lowangle detachment fault at the top of the south wall of the AM [1]. The fault zone is 100 m thick and is marked by variably deformed, talc- and/or amphibole-rich fault rocks. The alteration to form talcand amphibole-rich rocks occurs locally throughout the southern wall but is a key aspect of the detachment fault (Figure 4).

3

The detachment zone

The detachment fault zone is a fractured zone in which rocks on one side of the fracture move with respect to rocks on the other side. This is also a locus where fluids are focused and rock alteration is pervasive. Deformation along detachment shear zones is intimately connected to the tectonic evolution of the core complex and to the evolving geometry of the shear zone.

The Atlantis detachment fault zone is characterized by the occurrences of a peculiar rock types mostly made by a very soft mineral, the talc (Mg3 Si4 O10 (OH)2 ). This distinctly mechanically weak phase facilitates the dislocation along the detachment and lowers the shear strength. Recently, new experiments to characterize the mechanical behavior of talc show that the presence of even small amounts of talc can substantially weaken the volumes of rock, affecting its behavior and evolution of shear zones [8]. Our petrological and geochemical studies revealed those talcrich rocks are resulting from a dehydration and Si-metasomatism of previous serpentinized peridotite rocks [7], following the reaction: Serpentine + 2SiO2 = Talc + H2 O (1) Such a reaction is generally considered a prograde reaction taking place in the presence of SiO2 -rich fluids at temperatures of about 300-350°C ([9] and references therein). Furthermore, [10] presented experiments of the permeability evolution of an actively dehydrating serpentinite and showed that dehydration of serpentinite leads to a large transitory increase in porosity and permeability, an observation that has important implications for naturally dehydrating systems in subducting slabs 403

Marine Geology

Figure 4: Interpretative cross section of the Atlantis Massif (no vertical exaggeration), with detailed areas showing fluid pathways, metasomatic zones, and extent of serpentinization (light green shaded region) related to detachment faulting and steep normal faults (modified from Boschi et al. [7]). Box A: detail of the narrow shear zone (<100 m) along the detachment surface (in red-yellow) characterized by heterogeneous, variably altered and deformed gabbroic and peridotite lithologies and with extensive synkinematic metasomatism. Fluids were focused along the detachment and through discrete ductile shear zones triggering metasomatism of serpentinites and gabbros. The resulting talc-amphibole schists enclose lenses of relic, locally less deformed serpentinite and gabbroic rocks. The footwall was affected by a diffuse and static metasomatism mainly driven by a cataclastic network of fractures. and in middle to lower-crustal metamorphic regimes. Together with these peculiar rocks, the Atlantis detachment zone is characterized by amphibole-rich rocks (with mafic protolith). These close spatial association of talc- and amphibole-rich rocks suggested a local hydrothermal reaction between mafic rocks (gabbros) and seawater under greenschist-facies conditions along the detachment forming the amphibole schist, releasing silica and causing Si metasomatism and talc formation in the

404

neighboring serpentinites (forming the talc schist Figure 4A and 5).

4

Boron isotope study

During the last 20 years boron isotopes have increasingly been used as geochemical tracers in petrogenetic and ore deposit studies, and for modeling geochemical cycles in the Earthâ&#x20AC;&#x2122;s mantle, crust and oceans.

In spite of the increasing interest in understanding serpentinites geochemistry, the

Marine research at CNR

Figure 5: Three-dimensional representation of major element distribution of talc-rich (blue) and amphibole-rich (red) rocks together with Atlantis Massif serpentinites (green) and gabbros (yellow, see Boschi et al.[7] for details). The talc and amphibole schists show MgO and Al2 O3 between those of serpentinites and gabbros. In the binary plot MgO versus Al2 O3 (projection xz), there is a clear inverse correlation between the two elements. Moreover, concentrations of amphibole-rich samples are closer to the Atlantis Massif gabbros and concentrations of talc-rich samples to the Atlantis Massif serpentinites. By contrast, SiO2 shows no correlation with the other two elements (projections xy and yz), suggesting strong remobilization during the metasomatism. Talc-rich rocks, due to the elevated stoichiometric concentration of Si in the crystal structure, are most enriched in SiO2 .

B-isotope composition of oceanic serpentinites remains poorly documented. Isotopic data on this potentially important boron reservoir is crucial to properly model boron isotope systematics during metasomatic/magmatic processes in subduction zones and to estimate the global flux of boron and its isotopic fractionation during seawater-rock interaction in oceanic environments. Since the 1990’s, researchers of IGG-CNR (Pisa) developed an analythical method for the determination of boron isotopes. Boron isotope compositions are

measured at IGG-CNR (Pisa) using a VG Isomass 54E positive thermal ionization mass spectrometer, following separation of boron by ion-exchange procedures as described by [11]. Boron isotope composition is expressed by the notation: 11 10

( B/ B)sample δ 11 B = 11 10 − 1 × 103 , ( B/ B)stdSRM951

where 11 B/10 B stdSRM951 indicates the isotopic ratio of the international standard NIST SRM 951 (4.04362 ± 0.00137 2s). During serpentinization of peridotites, 405

Marine Geology

part of the boron contained in seawater (ca. 4.5 µg · g−1 most of it as trigonal species with δ 11 B = +39.5 ) is incorporated in serpentine minerals as tetrahedral species, producing variably enriched serpentinites (boron in the range 34-91 µg · g−1 ) with a fractionated isotopic composition (δ 11 B = +11 ÷ +16 ). In this way, the originally boron-poor peridotites (ca. 0.05-1 µg · g−1 ) were transformed in a strongly enriched boron reservoir. Our studies [12] indicate that extensive serpentinization of abyssal peridotites was a seawater-dominated process that occurred predominately at temperatures of 150–250 °C and at high integrated water/rock ratios. Talc and amphibole formation along the detachment fault was controlled by the access of Si-rich fluids derived through seawater–gabbro interactions. Replacement of serpentine by talc and amphibole resulted in boron loss (talc- and amphibole-rich fault rocks contain only 3-30 µg · g−1 ) and significant lowering of δ 11 Bvalues(9to10 ), which we modeled as the product of progressive extraction of boron. Completion of these processes produced the peculiar feature of the Atlantis Massif, where seawater continued to circulate through the serpentinitic basement introducing additional complexities. In fact, present-day seawater circulation is responsible for the formation of the spectacular low-temperature (max. 90°C), high-pH (up to 11) Lost City Hydrothermal Field [13]. These peculiar fluids form carbonate-brucite veins and fissures into the serpentinitic basement and, when discharged at seafloor, produce beautiful carbonate-brucite chimneys up to 60 m in height (Figure 3). Fluids issued at Lost City display the lowest boron concentrations ever measured in marine hydrothermal fluids (0.4 µg · g−1

406

with respect to the average 5 µg · g−1 of ”normal” black smokers) coupled with significantly low boron isotope composition (δ 11 B = +25 ). This behaviour is controlled by the precipitation of brucite, an efficient sink for boron, that is stable only under particular physico-chemical conditions. Our results have strong implications on understanding differences between chemistry of hydrothermal fluids at slow- and fast-spreading ridges that seems to be controlled respectively by the presence/absence of serpentinized peridotites at seafloor. Marine hydrothermal vents represent one of the main inputs to the oceans and the observed chemical differences may significantly influence the secular chemical variations of oceans.

5

Concluding remarks

The integrated field, petrological, geochemical and isotopic study of mafic and ultramafic rocks exposed on the south wall of the Atlantis Massif provides new insights into how major detachment shear zones evolve during the development of oceanic core complexes and demonstrates the interplay of fluids, mass transfer, and metamorphism in strain localization associated with this process. The talc-amphibole-chlorite-mineral assemblages and microstructures in the fault rocks indicate multiple stages of fluid infiltration and high strain deformation in limited domains. These minerals may contribute to softening and lubricate mylonitic fault zones, facilitate dislocation along the detachment and lower its shear strength, concentrate movement along the faults, and allow these faults to remain active as “detachment faults”. Seawater/rock interaction at ultramafic spreading ridges,

Marine research at CNR

like Atlantis Massif, produces a complex geochemical evolution of the original abyssal peridotites characterised by an initial boron content enrichment and isotope composition increase (serpentinization) followed by a later boron content depletion and isotope composition low-

ering (Si-metasomatism along detachment zone). Our research provides new knowledge on the geological evolution of oceanic lithosphere, how it is affected by seawater circulation and how it affects the geochemistry of marine hydrothermal vents.

References [1] J. A. Karson, E. A. Williams, G. L. Frueh-Green, D. S. Kelley, D. R. Yoerger, and M. Jakuba. Detachment Shear Zone on the Atlantis Massif Core Complex, MidAtlantic Ridge 30°N. G3, 7(6), 2006.

[2] Kelley D.S. and et al. An off-axis hydrothermal vent field near the Mid-Atlantic Ridge at 30°N. Nature, 412:145–149, 2001.

[3] G.L. Frueh-Green, J.A.D. Connolly, D.S. Kelley, A. Plas, and B. Grob´ety. Serpentinization of oceanic peridotites: Implications for geochemical cycles and biological activity. The Subseafloor Biosphere at Mid-Ocean Ridges, Geophys. Monogr. Ser., 144:119 – 136, 2004.

[4] D. K. Blackman and et al. Geology of the Atlantis Massif (Mid-Atlantic Ridge, 30°N): Implications for the evolution of an ultramafic oceanic core complex. Mar. Geophys. Res., 23:443–469, 2004.

[5] D.S. Kelley and et al. A serpentinite-hosted ecosystem: The Lost City Hydrothermal Field. Science, 307:1428 –1434, 2005.

[6] J.R. Cann, D.K. Blackman, D.K. Smith, E. McAllister, B. Janssen, S. Mello, E. Avgerinos, A. R. Pascoe, and J. Escartın. Corrugated slip surfaces formed at North Atlantic ridge-transform intersections. Nature, 385:329–332, 1997.

[7] C. Boschi, G.L. Frueh-Green J.A. Karson, D.S. Kelley, and A. Delacour. Mass transfer and fluid flow during detachment faulting and development of an oceanic core complex, Atlantis Massif (MAR 30°N). G3, 7, 2006.

[8] J. Escart´ın, M. Andreani, G. Hirth, and B. Evans. Relationships between the microstructural evolution and the rheology of talc at elevated pressures and temperatures. Earth and Planetary Science Letters, 268:463–475, 2008.

[9] C. Boschi, G.L. Frueh-Green, and J. Escart´ın. Occurrence and significance of serpentinite-hosted, talc-rich fault rocks in modern oceanic settings and ophiolite complexes. Ofioliti, 31(2):123–134, 2006.

[10] E. Tenthorey and S.F. Cox. Reaction-enhanced permeability during serpentinite dehydration. Geology, 31(10):921–924, 2003.

407

Marine Geology

[11] S. Tonarini, M. Pennisi, and W.P. Leeman. Precise boron isotopic analysis of complex silicate (rock) samples using alkali carbonate fusion and ion-exchange separation. Chem. Geol., 142:129–137, 1997.

[12] C. Boschi, A. Dini, G.L. Fruh-Green, , and D.S. Kelley. Isotopic and element exchange during serpentinization and metasomatism at the Atlantis Massif: Insights from B and Sr isotope data. Geochim. Cosmochim. Acta, 72:1801–1823, 2008.

[13] G.L. Fruh-Green, D.S. Kelley, S.M. Bernasconi, J.A. Karson, K.A. Ludwig, D.A. Butterfield, C. Boschi, , and G. Proskuroski. 30,000 years of hydrothermal activity at the Lost City vent field. Science, 301:495–498, 2003.

408

Tephrochronology of Deep Sea Marine Successions: Unlocking the Last 1 Myrs of Explosive Volcanic History in the Central Mediterranean Area D.D. Insinga1 , S. Tamburrino1 , M. Sprovieri2 1, Institute for Coastal Marine Environment, CNR, Napoli, Italy 2, Institute for Coastal Marine Environment, CNR, Capo Granitola (TP), Italy donatella.insinga@iamc.cnr.it Abstract Tephrochronological research at IAMC is currently focusing on the study of deep sea marine successions cored in the southern Tyrrhenian Sea, the Sicily Channel and the Ionian Basin. The composite record results in a 1 Myr long succession with a high-resolution and refined dating framework. The main scientific objective of the work is to extract timing and consequences of volcanic activity in the central Mediterranean area from the marine record, by integrating accurate dating techniques: tephrochronology, astronomical tuning and radiometric methods (40Ar39Ar). Advanced analytical approaches to tephrostratigraphy that include chemical characterisation through WDS-EPMA and LA-ICP-MS on single glass shards have been used and preliminary results were obtained on a number of the most recent tephra layers (<200 kyr) among the 70 recovered along the composite succession.

This allowed us to start off with a new reference database of chemical analyses to characterize the distal products of major volcanic events occurred during the late Pleistocene-Holocene and related to the explosive activity of the Campania Plain, the Aeolian islands, Pantelleria island and Mount Etna.

1

Introduction

Tephra layers originate from explosive volcanic eruptions and are distributed by wind or current action over wide areas. As the deposition of a tephra layer is essentially instantaneous on a geological timescale, these deposits are of major importance in stratigraphy because they provide a sound basis for dating (tephrochronometry) and also for correlation (tephrostratigraphy). Tephrochronology (tephrostratigraphy + tephrochronometry; [1]) can be applied in several geological settings where

fresh volcaniclastic materials are interbedded in continental, lacustrine, marine platform and deep sea sedimentary sequences.

In particular, the study of these horizons is a powerful tool to correlate and provide absolute ages in marine successions and to establish a high-resolution event stratigraphy and chronology in the framework of basin evolution. Tephra (from the Greek word meaning â&#x20AC;&#x153;ashâ&#x20AC;?) is actually used as a collective term for all airborne pyroclasts, including both airfall and pyroclastic flow material. Tephrochronologists who generally work with materials that are de-

Marine Geology

posited at considerable distance away from the volcanic source (distal tephras) are primarily concerned with tephra in ash-size fragments (<2 mm). Tephrochronological research at the IAMC is currently focusing on the study of deep sea marine successions cored in the southern Tyrrhenian Sea (core MD012474G), in the Sicily Channel (ODP Leg 160 Site 963A) and in the Ionian Basin (core KC01B) (Figure 1). The main scientific objective of this work is to â&#x20AC;&#x153;extractâ&#x20AC;? the timing and consequences of volcanic activity in the central Mediterranean area from the marine record since the middle Pleistocene. The great potential of these cored successions in terms of tephrochronology is given by the 1.1 My-long composite record with a high-resolution and refined dating framework, the downwind position of the cores with respect to the major volcanic sources of the area and the recognition of about 70 tephra layers suitable for chemical analyses and radiometric dating. This type of work could lead to several implications that include, for example: 1. the availability of a new analytical reference database; 2. the possibility of unravelling the history of poorly known or unknown explosive activity on land from the marine record; 3. the patterns of ash dispersal in the central Mediterranean; 4. the intercalibration of independent radioisotopic and astronomical dating methods.

Here we present some preliminary results concerning the chemical characterisation of five tephra layers that are found in the studied record and have correlated with major events occurred during the last 200 kyrs in the Campania Plain, at the Aeolian Islands, at Pantelleria island and at Mount Etna.

410

2

Tephra layers in the central Mediterranean area

The central Mediterranean area is an outstanding natural laboratory for tephra studies because of the occurrence of a number of highly explosive volcanoes active during the late Cenozoic. These vents are mostly known from the central-southern Italy, the Hellenic arc (e.g. Santorini) and Central Anatolia in Turkey. The tephra fallout of Hellenic and Anatolian volcanic eruptions is dispersed only in the very eastern part of the Mediterranean [5, 4] while the Italian volcanoes provide a more likely source for tephras in the central Mediterranean area due to their relative proximity and the favourable westerly blowing winds (Figure 1). The Plio-Quaternary magmatism in Italy occurred: a) along a NW-SE trending extensional zone on the Tyrrhenian border of the Italian Peninsula that includes the Roman Province (Vulsini, Vico, Sabatini and Colli Albani volcanoes), the Ernici-Roccamonfina Province, the Campania Province (Somma-Vesuvius, Campi Flegrei, Procida and Ischia volcanoes) and Mount Vulture, located east of the southern Apennines ; b) in the southern Tyrrhenian sea with the Aeolian volcanic arc and seamounts ; c) in the Sicily Province, which includes Mount Etna and Pantelleria island in the Sicily channel ([2] and references therein).

A chronogram of the activity of the volcanic centers is shown in Figure 2. The erupted products range from subalkaline (tholeiitic and calc-alkaline) to alkaline (sodic, potassic and ultrapotassic), from mafic to silicic, and from oversaturated to strongly undersaturated in silica. Calcalkaline and shoshonitic rocks are concentrated in the Aeolian arc, Na-alkaline rocks

Marine research at CNR

Figure 1: Location of the studied cores and volcanic centers of the central Mediterranean area. occur at Etna and Pantelleria while potassic and ultrapotassic rocks represent the most typical composition in the Roman, Roccamonfina and Campania provinces. Silica undersaturated rocks rich in both K2 O and Na2 O are found in the Mount Vulture products [6]. These different rocks of different magmatic series developed in response to the various geodynamic settings and mantle sources in the central Mediterranean region. Since the 1970s-1980s, numerous studies have produced tephrostratigraphic frameworks for the last 200 kyrs in the Ionian and Tyrrhenian sea (e.g.

[5, 7, 8, 9, 10]). They have been progressively refined over the years [11] (Table 1) with the most recent papers dealing with very high-resolution tephrostratigraphic studies of deep basin [12, 13, 14], shallow water [15] and lacustrine [16, 17, 18, 19, 20, 21, 22] archives. Recently, tephrostratigraphic studies in this region have been addressed towards the refinement of palaeoclimatic studies (e.g.

[23, 24, 25, 13, 26]). At the present,

there is no geochronological method better than a tephra marker for synchronising palaeoenvironmental records across the regions and for integrating different â&#x20AC;&#x153;event stratigraphiesâ&#x20AC;?. Eruptive centers from the Italian volcanic regions, particularly active over the last 200 kyrs, have produced very numerous pyroclastic deposits on land and off-shore that are well suited for 40Ar/39Ar dating (sanidine-bearing tephras). Some of them are of particular interest as they are contemporaneous to major climatic changes. This is especially true for the Campanian Ignimbrite (CI), dated at 39 ka on proximal deposits [27, 28] and 40 ka in the marine setting [25].

This major volcanic event has produced a marker layer throughout the central-eastern Mediterranean Sea (tephra layer Y5, Table 1), and possibly as far as the Greenland ice cap, as testified by the marked peak in SO4 found in GISP2 at the end of interstadial 9 (just after the Laschamp event coinciding with interstadial 10) [29]. Furthermore, a large number of tephras are interbedded in 411

Marine Geology

Figure 2: Chronogram of the volcanic activity in the central Mediterranean area during the last 1.1 Myrs. Data from [2] and references therein. CVZ: Campania Volcanic Zone after [3]. the Mediterranean sediments ranging from 0 to 50 Ka. These tephras were not dated before with the highly precise 40Ar/39Ar method, [30, 13]. The direct 40Ar/39Ar dating of these young ash layers (<100 ka) will open the way for the high-resolution tephrochronologic study of proximal and distal volcanic deposits interlayered within continental and marine sequences used in paleoclimatic reconstructions.

Finally, tephras have proved to provide robust absolute tie points for astronomically derived core chronology in several areas, in particular in the Mediterranean region (e.g.

CT 85-5 core, Ton-That et al., 2001), the northeast Indian Ocean (e.g. OPD site 758, [31]) and the New Zealand margin (DSDP 594, [32]).

412

3

Materials and methods

The sample set of this work is represented by three sedimentary successions cored in the southern Tyrrhenian Sea, in the Sicily Channel and in the Ionian basin down to a maximum depth of â&#x2C6;ź3600 meter. The composite record results in a 1.1 Myr long succession with a high-resolution and refined dating framework mostly yielded by the combination of astronomical tuning, environmental magnetism, stable isotope stratigraphy and radiometric methods. About 70 tephras, mostly composed of pumices and glass shards were found along the entire sequence (Figure 3). The tephrochronological results here presented deal with five layers of which three are

Marine research at CNR

Table 1: Tephrostratigraphic framework for the Mediterranean during the last 200 kyrs.

Modified from [4]. from the Ionian basin (tephras I1, I3 and I9), one from the Sicily Channel (tephra SC1) and one from southern Tyrrhenian Sea (tephra T22). They were labelled following an alpha numerical code indicating the location of the drilling site and their stratigraphic position with respect to the other tephras recognised in the core. Piston core KC01B (37.04 m long) is considered to be a reference core for the Pleistocene record in the Mediterranean. It was collected by the French R/V Marion Dufresne in 1991 (MD69 cruise) at a small ridge (the Pisano Plateau-36°15.25’N, 17°34.44’E, 3643 m depth) from the lower slope of the Calabrian Ridge in the Ionian Basin.

Lithology [38], paleoclimatic record (Sanvoisin et al., 1993), high-resolution iso-

tope stratigraphy [39], magnetostratigraphy [40] and astronomical calibration for the last 1.1 Myrs [33] were performed on the cored succession. Coring disturbances did not allow to study the uppermost 231 cm corresponding to the last 14 kyr. An attempt of tephrochronology was proposed by [33]: the author correlated 33 tephra layers found in the core with known events on land or with other distal markers just on the basis of their astronomical age (Table 2). Recently, core KC01B was entirely re-sampled and more than 33 tephras were recognised. Preliminary geochemical data obtained from several layers of the last 200 kyr, allowed to classify them for the first time thus recognising their source event, [41]. Mixed geochemi-

413

Marine Geology

Table 2: Tephrochronology of core KCO1B proposed by [33]. cal populations related to two different volcanic sources, moreover, have been often detected in the same tephras. Gravity core MD01-2474G (13.9 m long) was retrieved by the French R/V Marion Dufresne in 2001 at a structural high from the border of the Marsili Basin (39°10.44’N, 15°2.72’E, 2131 m depth) in southern Tyrrhenian Sea. Ecobiostratigraphy, isotope stratigraphy, 14C AMS dating and paleomagnetism were performed on the top 9 meters of core MD01-2474G and an agedepth model was constructed for the last 70 kyr [34]. ODP site 963 (37°02.148’ N, 13°10.686’ E 470.5 m depth) was drilled in the Sicily Channel on a short ridge between the Adventure Bank to the northwest and 414

the Gela basin to the southeast. The age model was assessed through oxygen isotope data of the 963 Site composite section [42] coupled with high resolution ecobiostratigraphy studies [35, 36] for the last 430 kyr.

3.1

Sampling and chemistry of tephra layers

Tephra samples were disaggregated in water and wet sieved at intervals of 63, 90, 125 and 250 µm. Glass concentrates were obtained through hand-picking with a microscope after a lithological analysis of the materials, avoiding samples with vesicles, crystalline intergrowth and

Marine research at CNR

alteration. Mineral phases were taken mainly from those tephra layers where fresh glass was not found. All these grains were then rinsed in distilled water, cleaned with an ultrasonic probe for two to three minutes and mounted on epoxy resin and then polished. The major element content was obtained through electron probe micro-analysis (EPMA). This enables grain-discrete determinations of the major elements within an individual glass shard. Measurements were made using a CAMECA SX-50 electron microprobe (WDS) at the IGAG-CNR (Istituto di Geologia Ambientale e GeoingegneriaConsiglio Nazionale delle Ricerche) in Rome. The accelerating voltage was 15 kV, the beam current 15 µA and the spot size 10 µm. Peak counting times for major elements were 20 seconds except for Fe which was analysed for 30 seconds. Instrument calibration was based on international glass and mineral standards. Individual analyses of glass shards with total oxide sums lower than 95 wt% were excluded.

The data of accepted analyses of individual tephra layers were then recalculated to 100 wt%. Although WDS requires a higher beam current and a longer counting times than EDS (energy dispersive spectrometry), the former offers the advantage of sequential acquisition of elemental data, so that the degree of sodium loss (frequent during EDS acquisition) can be tracked.

Actually, this methodology is the most recommended tool for determining the geochemical spectra of samples [30]. During EPMA analysis, back-scattered electron images are usually taken to record the position of each shard analysed and to carry out morphological observations which can be used as correlation tools. Trace element analysis on each single shard was obtained through Laser Ablation-Inductively

Coupled Plasma-Mass Spectrometry (LAICP-MS) methodology using a Q-switched Nd:YAG laser, model Quantel (Brilliant) at the IGG (Istituto di Geoscienze e Georisorse)-CNR in Pavia. The spot size was 25µm.

Several criteria were used to identify the primary origin of the studied tephras and they include: (1) well-recognisable peak abundance of the glass fraction above the whole lithic, crystal and bioclastic content coarser than 63 µm and (2) the identification, within each layer, of sizeable glass populations (>10 measurements) from the major element content.

4 4.1

Results and discussion Tephrochronology: tephrostratigraphy and tephrochronometry of the studied layers

The analysed tephras are represented by medium-to-fine dark grey ashes where the glass fraction is dominant. According to their major element content and by using the TAS diagram and the CIPW norms, these layers have compositions ranging from rhyolites to trachytes (qz-norm) and trachyphonolites with tephra T22 showing a wide range of variability (Table 3 and Figure 4). The recognition of the source area for each analysed tephra was mainly based on the comparison with published SEM-EDS and WDS (wavelength dispersive spectrometry) data on single glass shards. In those cases where single glass shard chemical data were lacking, XRF (Xray fluorescence) and ICP-MS whole rock data were used. The combination of EPMA and LA-ICP-MS results finally allowed to better characterise the source event. This 415

Marine Geology

procedure can be considered still pioneeristic when dealing with tephrostratigraphic analyses of marine successions. The “classical” type of investigation takes only the major oxide content into account. This approach has often proved to be inadequate to explain the complexity of some eruptive products and may lead to incorrect conclusions. Otherwise, the availability of a database composed by more than 40 elements for a single shard, may provide a much wider range of information from a single tephra than previous studies, allowing, for example, (1) to detect magmatic evolution of the sources and (2) to recognize chemically different populations of shards within one deposit which sometimes may not be easily distinguishable just from the electron probe data alone. This approach to the study of tephra layers from the cored successions might be necessary, if we consider that in the central Mediterranean area magmatic sources exhibit a very large compositional range of erupted products.

4.1.1

Ionian Basin

Tephra I1 is a 4 cm-thick layer composed of light grey micropumices, black poorly vesiculated obsidian, brown blocky glass shards and loose crystals of clinopyroxene and plagioclase. The glass fraction has a trachytic (qz-normative) composition with a Na-alkaline affinity (Figure 5).

REE (Rare earth Elements) are fractionated while a Ti trough and a Pb spike can be observed in the incompatible element patterns normalised to the primitive mantle (Figure 5). These chemical features strongly suggest an origin for tephra I1 from the Mount Etna volcano. This result, combined with an astronomical age of 16.7 ka B.P. for I1 [33], allowed to 416

correlate this tephra with the Y1 distal marker [5] or Et-1 [9], widespread both in marine [12, 13] and lacustrine [16, 18] archives. The source event is retained to be the Biancavilla Montalto Ignimbrite (or Unit D [43]) dated at ca. 15-16 ka B.P. (14C age [43]). Tephra I3 is about 4 cm-thick layer formed of tubular and Yshaped glass shards and subordinate micropumices. Loose crystals of k-feldspar, biotite and clinopyroxene are also present.

Tephra I3 has a trachytic-trachyphonolitic composition which is typical of the potassic series erupted in the Campania Plain during the Late Pleistocene-Holocene (Figure 6). The overall REE pattern analysed for these deposits is characterised by a high degree of fractionation with a (La/Yb)n ratio (where the subscript ”n” indicates normalization to chondritic abundances) ranging from ∼10 to∼17. Light rare earth elements (LREE) are strongly fractionated whilst heavy rare earth elements (HREE) give almost flat patterns (Figure 6). Eu shows marked variations, with both positive and negative anomalies [Eu/Eu*=0.301.4; Eu/Eu*=Eun /(Smn Gdn )1/2] which is typical in these rocks, where fractional crystallisation of K-feldspar is dominant (e.g.,[28]). The primitive mantlenormalised diagram (Figure 6) shows progressively higher incompatible element abundances in the analysed samples and deeper Ba, Sr, and Ti troughs as the degree of differentiation increases. Tephra I3, dated at 39.1 ka B.P. in core KC01B [33], can be correlated with the Y5 distal marker [5] related to the Campanian Ignimbrite event that occurred at Campi Flegrei (∼39 ka B.P. [27, 28]). This eruption is considered to be the most catastrophic eruption during the Quaternary in the Mediterranean area. Analysis of single shards allowed us to distinguish in tephra

Marine research at CNR

I3 between a “primitive” and “evolved” trachytic composition which indeed occurs in the proximal outcrops [28]. Tephra I9 is a very pronounced layer in core KC01B (8 cm thick) and it is represented by thin elongated, tubular and platy glass shards.

Loose crystals of biotite, clinpyroxene and K-feldspar are present. This layer has a trachytic composition and the major, trace and rare element content indicates a Campania Plain provenance (Figure 7). This result, combined with an age of 110.5 ka B.P. for the studied layer, suggest a correlation with the X6 tephra [5] dated approximately at 110 ka B.P. [52, 24]. This distal marker, widespread in the southern Italy domain [53, 19, 45], has been related to a major event occurred in the Campania Plain which experienced intense explosive volcanism during that period [27].

4.1.2

Southern Tyrrhenian Sea

Tephra T22 (8 cm thick) is formed mainly of dark scoria, light brown curvy, elongate and pumiceous glass shards. The shards representative of this layer fall into a large compositional field ranging from basaltic trachy-andesite to rhyolite with a high-potassium calc-alcaline affinity (Figure 8). REE of mafic and intermediate rocks are fractionated with a flat HREE pattern. Incompatible elements values normalized to primordial mantle give patterns with troughs at Nb and Sr, and peaks at Th, La, Nd and Gd. These chemical features suggest Lipari island (Aeolian Arc) as the source area for tephra T22 (40.6 ka) which might be related to the Brown Tuff units [54, 55] and, in particular, with the “Punta del Perciato” formation (ca. 41 ka B.P. [56]).

4.1.3

Sicily Channel

Tephra SC1 (tephra ODP1 in [57], in press), 4 cm thick, is represented mostly by light grey tubular and Y-shaped glass shards and loose crystals (feldspar). The major element content of the glass fraction shows a bimodal composition: peralkaline rhyolites (Agpaitic Index >1) for the bottom samples and trachytes for the top samples (Figure 9). Rhyolitic glass is pantelleritic in composition. Overall, these felsic volcanic products are characterised by an enrichment in LREE, Rb, Th, Nb and Zr (Figure 9). The main features of the chondrite-normalized REE patterns as the differentiation increases, are represented by the increase of total concentration of REE, a slight increase of LREE/HREE ratios ([La/Yb]n =8.8) and a marked Eu negative anomaly (Eu/Eu∗ = 0.6). These features are typical of Pantelleria products to which tephra SC1 can easily be correlated. In detail, this layer, dated at 42.5 ka B.P. in the ODP record, may represent the signature in the Sicily Channel of the Green Tuff event [58, 59] that occurred at ca. 45-50 ka B.P. [60]. According to this, tephra SC1 corresponds to the distal marine tephra layer Y6 [5], Table 1). Deposits related to this major event have been found extensively in lacustrine archives towards the northeast [50].

5

Conclusion

The tephrochronological analysis of a 1.1 Myr-long deep sea record, characterised by a refined dating framework, allowed us to recognise and correlate a number of tephra layers, out of the 70 found within the composite succession, with volcanic events on land. In this work we presented, in partic417

Marine Geology

ular, results concerning the major and trace element content detected on single glass shards from five tephra layers that span the time period ranging from ∼16 ka B.P. to ∼110 ka B.P.. This allowed us to start off with a new reference database of chemical analyses to characterize the distal products of major volcanic events that occurred during the late Pleistocene-Holocene and that are related to the explosive activity of the Campania Plain, the Aeolian islands, Pantelleria island and Mount Etna.

The main conclusions can be outlined as follows: • the most recent tephra I1, dated at 16.7 ka B.P. in the KC01B astronomicallytuned succession, was related to Mount

418

•

•

•

•

Etna activity and correlated with the Biancavilla Montalto Ignimbrite event; tephra I3, dated at 39.1 ka B.P., has been correlated with the major event of the Campanian Ignimbrite eruption occurred at Campi Flegrei in the Campania Plain; tephra T22, dated at 40.6 ka B.P., has been correlated with Lipari activity in the Aeolian arc; tephra SC1, dated at 42.5 ka B.P., has been correlated with the Green Tuff eruption occurred at Pantelleria island; tephra I9, dated at 110.5 ka B.P., represents the most ancient layer here presented. It has been correlated with a preignimbritic event occurred in the Campania Plain during the Late Pleistocene.

Marine research at CNR

Figure 3: Depth and age of tephra layers recovered in the KC01B, MD 012474G and Leg 160 ODP site 963 successions. Tephra layers presented in this work have been highlighted. (1): after[33]; (2): after [34]; (3) after [35] and [36].

419

Marine Geology

Figure 4: Classification of the studied tephra layers from the composite record of the central Mediterranean according to TAS (total alkali/silica) diagram [37].

Figure 5: (a) Tephra I1 interbedded in marine deposits of core KC01B; (b) classification of tephra I1 and comparison with the average compositional fields of tephra Y1 and the onland deposits of Biancavilla Montalto Ignimbrite eruption; (c) Na2 O/K2 O diagram to show the sodic affinity of tephra I1; (d) primitive mantle normalised and chondrite normalised diagrams for tephra I1. Compositional range of proximal deposits are reported for comparison. Data from: [5, 9, 12, 18, 43, 2] and references therein.

420

Marine research at CNR

Table 3: Major (wt%), trace and rare-element (ppm) composition of glasses from the studied tephras. All analyses recalculated water-free to 100. s: scoria; gs: glass shard; alk: Na2 O+K2 O; AI (Agpaitic Index): molar (Na2 O+ K2 O/Al2 O3 ); n: number of shards analysed for each tephra; Ď&#x192;: standard deviation; T: trachyte; TP: trachyphonolite; BTA: basaltic trachyandesite; TA: trachyandesite; A: andesite; TD: trachydacite; D: dacite; Rhy: rhyolite; P: pantellerite.

421

Marine Geology

Figure 6: (a) Tephra I3 interbedded in marine deposits of core KC01B; (b) classification of tephra I3 and comparison with the average compositional fields of tephra Y5 and the onland deposits of the Campanian Ignimbrite eruption; (c) primitive mantle normalised and chondrite normalised diagrams for tephra I3. Compositional range of proximal deposits are reported for comparison. Data from: [5, 18, 19, 28] and references therein.

422

Marine research at CNR

Figure 7: (a) Tephra I9 interbedded in marine deposits of core KC01B; (b) classification of tephra I9 and comparison with the average compositional fields of tephra X6; (c) primitive mantle normalised and chondrite normalised diagrams for tephra I9. Compositional range of proximal deposits are reported for comparison. Data from: [5, 44, 18, 19, 45].

423

Marine Geology

Figure 8: (a) Tephra T22 interbedded in marine deposits of core MD 012474G; (b) classification of tephra T22 and comparison with the average compositional fields of Lipari proximal deposits; (c) primitive mantle normalised and chondrite normalised diagrams for tephra T22. Compositional range of Lipari proximal deposits are reported for comparison. Data from [2] and references therein.

424

Marine research at CNR

Figure 9: (a) Tephra SC1 interbedded in marine deposits of ODP site 963; (b) classification of tephra SC1 and comparison with the average compositional fields of the Green Tuff proximal deposits and tephra Y6; (c) FeOtot-Al2 O3 diagram (after [46]) where tephra SC1 samples have been plotted for the geochemical discrimination between pantellerites and trachytic comendites; (d) primitive mantle normalised and chondrite normalised diagrams for tephra SC1. Compositional range of Pantelleria proximal and distal deposits (pantellerites) are reported for comparison. Data from [5, 47, 48, 49, 50, 51].

425

Marine Geology

References [1] A. Sarna-Wojcicki. Tephrochronology. Quaternary Geochronology: Methods and Application, pages 357–377, 2000.

[2] A. Peccerillo. Plio-Quaternary volcanism in Italy. Petrology, Geochemistry and Geodynamics. page 365, 2005.

[3] G. Rolandi, F. Bellucci, and M.T. Heitzler. Tectonic controls on the genesis of the Ignimbrites from the Campanian volcanic Zone, southern Italy. Special Issue Mineralogy and Petrology, 79:3–31, 2003.

[4] T.H. Druitt. Late Quaternary rhyolitic eruptions from the Acig¨ol Complex, central Turkey. Journal of the Geological Society, 152:655–667, 1995.

[5] J. Keller, W.B.F. Ryan, D. Ninkovich, and R. Altherr. Explosive volcanic activity in the Mediterranean over the past 200,000 yrs as recorded in deep-sea sediments.

Geological Society of American Bulletin, 89:591–604, 1978.

[6] L. Beccaluva, M. Coltorti, P. Di Girolamo, L. Melluso, L. Milani, V. Morra, and F. Siena. Petrogenesis and evolution of Mt. Vulture alkaline volcanism (Southern Italy). Mineralogy and Petrology, 74:277–297, 2002.

[7] R. Thunell, A. Federman, R.S.J. Sparks, and D. Williams. The age and volcanologic significance of the Y-5 ash layer in the Mediterranean. Quaternary Research, 12:241–253, 1979.

[8] A. Vinci. Distribution and chemical composition of tephra layers from Eastern Mediterranean abyssal sediments. Marine Geology, 64:143–155, 1985.

[9] M. Paterne, F. Guichard, and J. Labeyrie. Explosive activity of the south Italian volcanoes during the past 80.000 years as determined by marine tephrochronology.

Journal Volcanology and Geothermal Ressearch, 34:153–172, 1988.

[10] M. Paterne, F. Guichard, J. Labeyrie, P.Y. Gillot, and J.C. Duplessy. Tyrrhenian Sea tephrochronology of the oxigen isotope record for the past 60.000 years. Marine Geology, pages 259–285, 72.

[11] L. Vezzoli. Tephra layers in Bannock Basin (Eastern Mediterranean). Marine Geology, 100:21–34, 1991.

[12] N. Calanchi, A. Cattaneo, E. Dinelli, G. Gasparotto, and F. Lucchini. Tephra layers in Late Quaternary sediments of the central Adriatic Sea. Marine Geology, 149:191–209, 1998.

[13] G. Siani, R. Sulpizio, M. Paterne, and A. Sbrana. Tephrostratigraphy study for the last 18.000 14C years in a deep-sea sediment sequence of the South Adriatic.

Quaternary Science Reviews, 23:2485–2500, 2004.

426

Marine research at CNR

[14] M. Paterne, F. Guichard, J.C. Duplessy, G. Siani, R. Sulpizio, and J. Labeyrie. A 90,000–200,000 yrs marine tephra record of Italian volcanic activity in the Central Mediterranean Sea. Journal Volcanology and Geothermal Research, 177(1):187– 196, 2008.

[15] D. Insinga, F. Molisso, C. Lubritto, M. Sacchi, I. Passariello, and V. Morra. The proximal marine record of Somma-Vesuvius volcanic activity in the naples and salerno bays, eastern tyrrhenian sea, during the last 3 kyrs. Journal of Volcanology and Geothermal Research, 177:170–186, 2008.

[16] N. Calanchi, E. Dinelli, F. Lucchini, and A. Mordenti. Chemostratigraphy of Late Quaternary sediments from Lake Albano and central Adriatic sea cores. Mem. Ist. It. Idrobiol., (55):247–264, 1996.

[17] B. Narcisi. Tephrochronology of a late Quaternary lacustrine record from the Monticchio maar (Vulture volcano, Southern Italy). Quaternary Science Review, 15:155–165, 1996.

[18] S. Wulf, M. Kraml, A. Brauer, J. Keller, and J.F.W. Negendank. Tephrochronology of the 100 ka lacustrine sediment record of Lago Grande di Monticchio (southern Italy). Quaternary International, 122:7–30, 2004.

[19] R. Munno and P. Petrosino. The late Quaternary tephrostratigraphical record of the San Gregorio Magno basin (southern Italy). Journal of Quaternary Science, 22(3):247–266, 2007.

[20] B. Wagner, R. Sulpizio, G. Zanchetta, S. Wulf, M. Wessels, G. Daut, and N. Nowaczyk. The last 40 ka tephrostratigraphic record of Lake Ohrid, Albania and Macedonia: a very distal archive for ash dispersal from Italian volcanoes. Journal Volcanology and Geothermal Research, 177:71–80, 2008.

[21] B. Giaccio, P. Messina, A. Sposato, M. Voltaggio, G., Zanchetta, F. Galadini, S. Gori, and R. Santacroce. Tephra layers from Holocene lake sediments of the Sulmona Basin, central Italy: implications for volcanic activity in Peninsular Italy and tephrostratigraphy in the central Mediterranean area. Quaternary Science Review, 28:2710–2733, 2009.

[22] R. Sulpizio, A. van Welden, B. Caron, and G. Zanchetta. The Holocene tephrostratigraphic record of Lake Shkodra (Albania and Montenegro). Journal of Quaternary Science, DOI: 10.1002/jqs.1334, 2009.

[23] D. B. Karner, E. Juvigne, L. Brancaccio, A. Cinque, E. Russo Ermolli, N. Santangelo, S. Bernasconi, and L. Lirer. A potential early middle Pleistocene tephrostratotype for the Mediterranean basin: The Vallo Di Diano, Campania, Italy. Global Planetary Change, 21:1–15, 1999.

[24] B. Narcisi and L. Vezzoli. Quaternary stratigraphy of distal tephra layers in the Mediterranean—an overview. Global and Planetary Change, 21:31–50, 1999.

427

Marine Geology

[25] T. Ton That, B. Singer, and M. Paterne. 40Ar/39Ar dating of latest Pleistocene (41 ka) marine tephra in the Mediterranean Sea: implications for global climate records. Earth and Planetary Science Letters, 184:645–658, 2001.

[26] J.J. Lowe, S. Blockleya, F. Trincardi, A. Asioli, A. Cattaneo, I.P. Matthews, M. Pollard, and S. Wulf. Age modelling of late Quaternary marine sequences in the Adriatic: Towards improved precision and accuracy using volcanic event stratigraphy.

Continental Shelf Research, 27:560–582, 2007.

[27] B. De Vivo, G. Rolandi, P.B. Gans, A. Calvert, W.A. Bohrson, F.J. Spera, and H.E. Belkin. New constraints on the pyroclastic eruptive history of the Campanian volcanic Plain (Italy). Mineralogy and Petrology, 73:47–65, 2001.

[28] L. Fedele, C. Scarpati, M. Lanphere, L. Melluso, V. Morra, A. Perrotta, and G. Ricci. The Breccia Museo formation, Campi Flegrei, southern Italy: geochronology, chemostratigraphy and relationship with the Campanian Ignimbrite eruption.

Bulletin of Volcanology, 70(10), 2008.

[29] F.G. Fedele, B. Giaccio, R. Isaia, and G. Orsi. The Campanian Ignimbrite eruption, Heinrich event 4 and the Paleolithic change in Europe: a high-resolution investigation. (139):301–325, 2003.

[30] S. M. Davies, N.P. Branch, J.J. Lowe, and S.M.C. Turney. Towards a European tephrochronological framework for Termination 1 and the Early Holocene. Phil.

Trans. R. Soc. London A, 360:767–802, 2002.

[31] C.M. Hall and J.W. Farrel. Laser 40Ar/39Ar ages of tephra from Indian Ocean deep-sea sediments: Tie points for the astronomical and geomagnetic polarity time scales. Earth and Planetary Science Letters, 133:327–338, 1995.

[32] B. Pillans, B.P. Kohn, G. Berger, P. Froggatt, G. Duller, B. Alloway, and P. Hesse.

Multi-method dating comparison for mid-pleistocene Rangitawa Tephra, New Zealand. Quaternary Science Reviews, 15:641–663, 1996.

[33] L.J. Lourens. Revised tuning of Ocena Drilling Program Site 964 and KC01B (Mediterranean) and imp0licetions for the δ18O, tephra, calcareous nannofossil, and geomagnetic reversal chronologies of the past 1.1 Myr. Paleoceanography, 10.1029/2003PA000997, 19(3):PA3010, 2004.

[34] S. Tamburrino. Mediterranean tephrochronology: new insights from highresolution analyses of a 200 ky long composite sedimentary record. PhD Thesis, 2009.

[35] R. Sprovieri, E. Di Stefano, A. Incarbona, and D.W. Oppo. Suborbital climate variability during Marine Isotopic Stage 5 in the central Mediterranean basin: evidence from calcareous plankton record. Quaternary Science Reviews, 25:2332– 2342, 2006.

428

Marine research at CNR

[36] A. Incarbona, E. Di Stefano, and S. Bonomo. Calcareous nannofossil biostratigraphy of the central Mediterranean Basin during the last 430,000 years. Stratigraphy, 6(1), 2009.

[37] R.W. Le Maitre, P. Bateman, A. Dudek, J. Keller, J. Lameyre, M.J. Le Bas, P.A. Sabine, R. Schmid, H. Sorensen, A. Streckeisen, A.R. Woolley, and B. Zanettin. A Classification of Igneous Rocks and Glossary of Terms: Recommendations of the International Union of Geological Sciences Subcommission on the Systematics of Igneous Rocks. 1989.

[38] D. Castradori. Calcareous nannofossil biostratigraphy and biochronology in eastern Mediterranean deep-sea cores. Rivista Italiana Paleontologia e Stratigrafia, 99(1):107–126, 1993.

[39] M. Rossignol-Strick, M. Paterne, F. Bassinot, K.-C. Emeis, and G. J. De Lange. An unusual mid-Pleistocene monsoon period over Africa and Asia. Nature, 392:269– 272, 1998.

[40] C.G. Langereis, M.J. Dekkers, G.J. De Lange, M. Paterne, and P.J.M. Van Santvoort. Magnetostratigraphy and astronomical calibration of the last 1.1 Myr from an eastern Mediterranean piston core and dating of short events in the Brunhes. Geophysical Journal International, 129:75–94, 1997.

[41] D. Insinga, S. Tamburrino, L. Vezzoli, G.J. De Lange, F. Lirer, M. Sprovieri, and M. Tiepolo. Tephrochronology of the past 200 kyrs in the astronomically tuned KC01B sedimentary core (Ionian Basin, eastern Mediterranean). Geoitalia, 2009.

[42] M.W. Howell, R.C. Thunell, E. Di Stefano, R. Sprovieri, E.J. Tappa, and T. Sakamoto. Stable isotope chronology and paleoceanographic history of Sites 963 and 964, Eastern Mediterranean Sea. Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 160, doi:10.2973/odp.proc.sr.160.014.1998, 1998.

[43] M. Coltelli, P. De Carlo, and L. Vezzoli. Stratigraphic constraints for explosive activity in the past 100 ka at Etna volcano, Italy. International Journal of Earth Science, 89:665–667, 2000.

[44] L.. Pappalardo, L. Civetta, M. D’Antonio, M. Deino, M.A. Di Vito, G. Orsi, A. Carandente, S. de Vita, R. Isaia, and M. Piochi. Chemical and Sr isotopical evolution of the Phlegrean magmatic system before the Campanian Ignimbrite and the Neapolitan Yellow Tuff eruptions. Journal of Volcanology and Geothermal Research, 91(2-4):141–166, 1999.

[45] R. Marciano, R. Munno, P. Petrosino, N. Santangelo, A. Santo, and I. Villa. Late quaternary tephra layers along the Cilento coastline (southern Italy). Journal of Volcanology and Geothermal Research, 177(1):227–243, 2008.

[46] R. MacDonald. Nomenclature and Petrochemistry of the Peralkaline Oversaturated Extrusive Rocks. Bulletin of Volcanology, 38:498–516, 1974.

429

Marine Geology

[47] L. Civetta, Y. Cornette, G. Crisci, P.Y. Gillot, G. Orsi, and C.S. Requejo. Geology, geochronology and chemical evolution of the island of Pantelleria. Geology Magazine, 121:541–562, 1984.

[48] R. Avanzinelli, L. Bindia, S. Menchettia, and S. Ponticelli. Crystallisation and genesis of peralkaline magmas from Pantelleria Volcano, Italy: an integrated petrological and crystal-chemical study. Lithos, 73:41–69, 2004.

[49] V. Margari, D. Pyle, C. Bryant, and P.L. Gibbard. Mediterranean tephra Stratigraphy revisited: results from a long terrestrial sequence on Lesvos Island, Greece.

Journal Volcanology and Geothermal Research, 163:34–54, 2007.

[50] H. Vogel, G. Zanchetta, R. Sulpizio, B. Wagner, and N. Nowaczyk. A tephrostratigraphic record for the last glacial–interglacial cycle from Lake Ohrid, Albania and Macedonia. Journal of Quaternary Science, DOI: 10.1002/jqs.1311, 2009.

[51] J.C. White, D.F. Parker, and M. Ren. The origin of trachyte and pantellerite from Pantelleria, Italy: Insights from major element, trace element, and thermodynamic modelling. Journal of Volcanology and Geothermal Research, 179:33–55, 2009.

[52] M. Kraml. Laser 40Ar/39Ar-datierungen an distalen marinen tephren des jungquart¨aren mediterranean vulkanismus (Ionisches Meer, Meteor-Fahrt 25/4). 1997. ¨ [53] W. Morche. Tephrochronologie der Aolischen Inseln. 1988.

[54] A. Gioncada, R. Mazzuoli, M. Bisson, and M.T. Pareschi. Petrology of volcanic products younger than 42 ka on the Lipari-Vulcano complex (Aeolian Islands, Italy): an example of volcanism controlled by tectonics. Journal Volcanology and Geothermal Research, 122:191–220, 2003.

[55] F. Lucchi, C.A. Tranne, G. De Astis, J. Keller, R. Losito, and W. Morche. Stratigraphy and significance of Brown Tuffs on the Aeolian Islands (southern Italy). Journal of Volcanology and Geothermal Research, 177:49–70, 2008.

[56] G.M. Crisci, R. De Rosa, S. Esperanc¸a, R. Mazzuoli, and M. Sonnino. Temporal evolution of a three component system: the Island of Lipari (Aeolian Arc, southern Italy). Bullettin of Volcanology, 53:207–221, 1991.

[57] S. Tamburrino, D.D. Insinga, M. Sprovieri, P. Petrosino, and M. Tiepolo. Major and trace element characterization of tephra layers offshore Pantelleria Island: insights into the last 200 ka of volcanic activity and contribution for the Mediterranean tephrochronology. Journal of Quaternary Science, In press, 2011.

[58] Y. Cornette, G.M. Crisci, P.Y. Gillot, and G. Orsi. Recent volcanic history of Pantelleria: a new interpretation. Journal of Volcanology and Geothermal Research, 17:361–373, 1983.

430

Marine research at CNR

[59] G. Orsi, , and M. F. Sheridan. Green Tuff of Pantelleria: rheoignimbrite or rheomorphic fall? Bulletin Volcanologique, 47:611â&#x20AC;&#x201C;626, 1984.

[60] L. Civetta, Y. Cornette, P.Y. Gillot, and G. Orsi. The eruptive history of Pantelleria (Sicily Channel) in the last 50 ka. Bulletin of Volcanology, 50:47â&#x20AC;&#x201C;57, 1988.

431

Marine Geology

432

Volcanic Islands: the Tips of Large Submerged Volcanoes that only Marine Geology May Reveal (Examples from W-Pontine Archipelago, Ischia, Stromboli and Pantelleria) A. Bosman1 , M. Calarco2 , D. Casalbore1 , A.M. Conte3 , E. Martorelli1 , A. Sposato1 , F. Falese2 , L. Macelloni4 , C. Romagnoli5 , F.L. Chiocci1,2 1, Institute of Environmental Geology and Geoengineering, CNR, Roma, Italy 2, Department of Earth Sciences, University of Roma â&#x20AC;&#x153;La Sapienzaâ&#x20AC;?, Roma, Italy 3, Institute of Geosciences and Earth Resources, CNR, Roma, Italy 4, Center for Marine Resources and Environmental Technology, University of Mississippi, MS, USA 5, Department of Geo-Environmental and Earth Sciences, University of Bologna, Bologna, Italy eleonora.martorelli@igag.cnr.it Abstract Submarine portions of volcanic islands are several times (up to an order of magnitude) larger than subaerial ones but very poorly known. Their knowledge about them is essential for the reconstruction of eruptive history and volcano-tectonic evolution, which are more difficult to be fully witnessed on-land. The offshore investigation of a number of Italian volcanic islands led to the identification of key outcomes, such as the widespread occurrence of mass-wasting features ranging across different scales and occurrence frequency. Large-scale features (i.e. debris avalanche deposits) occur both at Stromboli and Ischia islands, whereas morphologies and deposits related to minor slides and density flows (e.g. debris flows, grain flows, and turbidity flows) are very common in all the study areas. Moreover, several primary volcanic features have been identified, such as volcanic cones, eruptive fissures and lava fields. The results highlight the complex evolution of volcanic edifices, through the alternation of constructive phases, due to magma emission, and destructive processes, with mass wasting at different scales, including flank collapse. Thus, the achieved scientific results also provide a relevant contribution to the instability hazard assessment.

1

Introduction

Marine research is essential to gain a complete geological reconstruction of insular volcanoes, as it enables the characterization of the submarine sectors of the edifices, usually much wider than the

subaerial ones. Moreover, the underwater environment usually displays a finer record of eruptive events compared to the subaerial counterpart, where erosive and anthropic processes may often obliterate the outcrops. In the last decade, the availability of new high-resolution re-

Marine Geology

mote sensing techniques for the investigation of marine areas, such as multibeam echosounder and side scan sonar, has provided a great improvement to morphological studies. Furthermore, the coupling of such surveys with other geophysical methodologies (e.g. seismic survey) and petrographyc-petrological analysis of seafloor samples greatly contributed to the knowledge of insular and oceanic volcanoes. This contribution aims to show the relevant results of a decade of studies carried out offshore some Italian volcanic islands characterized by different volcanic activity and physiographic setting. Four study-cases are here described following an order that takes into account their different settings with respect to the continental margin, i.e. two isolated island volcanoes (Stromboli and Pantelleria Islands) and two other ones lying on the continental margin (Ischia Island and Western Pontine Archipelago). The data were collected during over ten oceanographic cruises carried out onboard research vessels Urania, Tethis and Universit`atis between 1998 and 2008.

The results are more detailed for Pontine, Ischia and Stromboli that have been studied for a longer time than Pantelleria, where investigations began in 2006. Particular emphasis is given to those results that have improved the knowledge of geological setting, previously based only on subaerial data. Notably, these studies also improve the overall understanding of volcanological evolution of studied apparata and, particularly, provide relevant implications on volcano-tectonic events reconstruction.

wide for its persistent Strombolian activity. It consists of periodic low-energy explosions of incandescent scoriae, magma lumps, ashes and blocks with heights of a few tens to hundreds of meters. The island represents the tip of a steep and mostly submerged (98% of the entire area [1]) stratovolcano that rises for â&#x2C6;ź3000 m from the seafloor. New high-resolution multibeam bathymetry and long-range side scan sonar data allowed the characterization of the morpho-structural setting of the edifice [1]. On the whole, this is defined by a quasi-bilateral symmetry of the submarine flanks with respect to the main SW-NE rift axis, similarly to what observed for the subaerial portion, although a more complex structural pattern can be depicted on the basis of new marine evidence. For instance, a N64°E structural trend controls the alignment of Strombolicchio Canyon (Figure 1), a flat-bottomed erosive feature, 8 km long and 500-700 m wide, bounded by steep (up to 45°) and very rectilinear scarps with a relief of about 200 m.

Marine studies have also shown the occurrence of widespread mass-wasting features that affect about the 90% of its submerged extension [2]. These events range from medium-scale submarine landslides (estimated volume of a few 106 m3 ) up to large-scale sector collapses (estimated volume of each being 1-2 km3 ). The latter catastrophic events drove the development of large and well-marked scars on the E and NW submerged flanks and to the emplacement of related debris avalanche deposits at the base of the edifice. In particular, buried chaotic deposits and a large megablocks field (Figure 1) have been recognized at the foot of the older eastern 2 Stromboli Island flank [3]. These features testify the ocStromboli Island is located in the NE part currence of at least two large-scale sector of the Aeolian Arc and it is known world- collapses in this side of volcano, where 434

Marine research at CNR

Figure 1: 3D perspective view of the eastern flank of Stromboli vocano TOBI data are draped over bathymetry. a) Reference map; b) Bathymetric map (equidistance: 200 m, the grey area corresponds to the tridimensional view). Vertical Exaggeration: 1.2 landslide events were previously considered a minor process. Differently, a large volcaniclastic fan-shaped bulge has been identified on the NW slope, representing the sum of the four nested lateral collapses that affected the NW sector of the Stromboli in the last 13 ka [4, 5]. Despite their low frequency of occurrence (in the order of thousands of years), these large-scale sector collapses pose a major threat to the coastal settlement, as they are able to generate important tsunami waves [6].

However, medium-scale submarine landslides represent a major and possibly more significant (with respect to large-scale sector collapses) hazard as they occur at higher frequencies [7, 8], and may become important at human timescales, as demonstrated by the 30 December 2002 tsunamigenic slide [9]. Thus, the Stromboli 2002 landslide represented a unique opportunity to characterize tsunamigenic

medium-scale submarine mass failures at Stromboli, providing useful insights for hazard assessment [7, 10].

Finally, a large spectrum of erosive/depositional features related to the flowing of density currents were identified on Stromboli submerged flanks. Grain flows seem to dominate the upper slope that is characterized by very high gradients (up to 35°), whereas turbidity currents evolve on the lower slope [2].

3

Pantelleria

Pantelleria Island is located in the Sicily Channel Rift Zone (SCRZ) and represents the emergent tip of an underwater volcano structure with 72% lying below sea level, down to a depth of â&#x2C6;ź1200 m. Pantelleria geology is characterized by a bimodal suite of widespread lavas and tuffs, dominated by silicic peralkaline volcanism and 435

Marine Geology

Figure 2: DEM of Pantelleria volcanic edifice. a) Reference map; b) 3D perspective view of the NW offshore volcanic area. Vertical Exaggeration: 1.5. subordinated mafic lavas. The volcanic history started 300 ka ago and alternated between periods of less energetic activity or quiescence and large explosive eruptions, which caused two caldera collapses the latter being related to the Green Tuff eruption (e.g. [11, 12]). The most recent volcanic event occurred in 1891, four km offshore the NW coast [13, 14] with a submarine basaltic eruption (i.e. the Foerstner volcano eruption).

Until recent marine surveys, little was known about the submarine portions of the volcano. The main results of marine surveys led to: i) the identification of about thirty new submarine eruptive vents, at depth ranging between â&#x2C6;ź250 to â&#x2C6;ź800

436

m; ii) the accurate re-location of the 1891 eruptive vent and iii) the recognition of a wide submerged shelf in the NW sector with recent lava flows.

Most of the newly discovered eruptive vents consist of cones and eruptive fissures that rise from the NW submarine flank of the edifice (Figure 2), with the farthest one located 8.5 km offshore the coastline.

The morphology of the cones is similar to the emerged scoria-cones, with heights up to 350 m and basal widths of 0.52.3 km. Sometimes, the coalescence of a few cones along an eruptive fissure creates elongated or complex structures that are mainly aligned with the SCRZ trend. The eruptive centres exhibit a well-preserved

Marine research at CNR

volcanic morphology; two centers present small-scale sector collapses with smaller cones inside (Figure 2b).

Samples collected from the eruptive vents as well as core samples from the NW sector encompass a variety of products which include highly scoriaceous lava fragments, glass-rich lapilli, and coarse ashes.

This highlights the complexity of eruptive mechanisms that characterized the volcanic history of this area. However, all samples revealed a nearly homogeneous basaltichawaiitic composition, similar to that of recent volcanic products on-land reported by Civetta [15]. This evidence significantly extends the areal distribution of basaltic magmatism known on the island.

Marine data also enabled the accurate relocation of the 1891 eruptive event, which was identified only through floating lava bombs emission [13]. The most likely eruptive vent for the 1891 eruption is a small centre located four km northwest of Pantelleria harbour (Figure 2). Samples related to this eruption were collected as well; they are fragments of highly scoriaceous lava bomb that display a porphyritic and holocrystalline-textured portion and a glassy inner layer. They are hawaiithic in composition, as the other eruptive products from the NW area [16]. Of particular interest is the eruptive style of the 1891 event that represents a peculiar case of shallowintermediate submarine eruptions. Similar volcanic events have been rarely observed (i.e. Socorro Island, Mexico [17] and Terceira Island, Azores [18]), and at present, they are still not fully explained. Moreover, the submarine investigation led to the characterization of the wide shelf that extends offshore about 4 km in the NW direction (Figure 2). This submarine shelf was likely built up by lava flows and other volcaniclastic deposits of different ages, with the

sources localized on-land (Mursia, and P.ta San Leonardo basalts) or submerged at the present-day. Contemporary, over this time, the shelf underwent subaerial and marine erosion due to different stages of eustatic sea level fluctuations.

It deserves to mention that most of the discovered volcanic vents occur in the NW area of Pantelleria and show wellpreserved morphologies, whereas the SE sector of the edifice mainly consists of erosional remnants of old volcanic outcrops. Indeed, this last sector was affected by severe and widespread dismantling processes. These morphological differences, as well as the occurrence of the historical eruption in the NW sector, indicate that, during time, the focus of volcanic activity has been propagating northwestward. The submarine evidence for this migration is in agreement with the hypothesis of [12] based on subaerial volcanological data.

4

Ischia

Ischia Island is an active alkalitrachytic volcanic edifice located at the western end of the Bay of Naples. Excluding the M. Epomeo Green Tuff ignimbritic eruption, that occurred around 55 ka, the large part of the activity consisted of small-scale eruptions, the latest of which occurred in 1302 AD [19]. A notable characteristic of this volcanic edifice is the rapid uplift of Mt. Epomeo, that has been raised â&#x2C6;ź700-800 m in the past 33 kyr [20]. The strong volcano-tectonic uplift caused oversteepening and morphologic disequilibrium that, combined with seismicity and hydrothermal rock weathering, produced diffuse slope instability.

With the exception of the southern flank, the island is surrounded by a continental 437

Marine Geology

Figure 3: 3D perspective view of the southern flank of Ischia Island offshore. TOBI data are draped on the DEM. a) Reference map; b) Bathymetric map (equidistance 200 m, the grey area corresponds to the tridimensional view); c) detailed view of megablocks of the debris avalanche deposit.Vertical Exaggeration: 1.5. shelf that partially merges with the continental margin facing the Volturno Plain and the Gulf of Naples. Our research revealed that in the southern flank a steep slope, locally cut by near vertical scarps and canyon heads, connects the coastline to deep areas. In this area it is also evident the continuation, below the sea level, of the 2.2-km-wide horseshoe scar, carving the Mt. Epomeo. This has been interpreted as a sector collapse scar; related debris avalanche deposit (DA) lies at the toe of the southern slope of Ischia. Between 550 m and 1100 m w.d., collected data show a field of thousands of mega blocks, dispersed over an area of â&#x2C6;ź200 km2 , as far as 50 km from the island (Figure 3). The width of individual blocks in the DA vary from several meters to more than 200 m, while the height of the larger blocks range up to 30-50 m. The deposit has a curved

438

areal distribution, partly following the local pre-collapse topography. However, on its eastern side it extends up over a ridge 80 m high and down into the Magnaghi Canyon (Figure 3). This debris avalanche deposit has been related to the catastrophic failure of the southern flank of the island [21].

Samples of the DA matrix and associated debris flow (DF) deposit were recovered through dredging and, in the interblock regions, by gravity coring. The DA matrix consists of a mixture of clasts ranging in size from centimeters to millimeters and an interclast groundmass with a very low clay fraction; the resulting texture is that of a slightly consolidated sandy deposit. On the contrary, the DF is a mud-supported deposit with a significant clay fraction. Cores also recovered a postavalanche hemipelagic mud drape thinner than 1 m.

Marine research at CNR

A reliable estimate of the DA volume would require knowledge of its overall thickness (i.e. blocks + matrix + DF) while an estimate based only on the volume of “outcropping” blocks that are acoustically detectable (numbering around 5000) results in an extremely conservative value.

Therefore the minimum estimated volume of the deposit is 1.5 km3 , whereas the possible mobilized volume in the scar area ranges between 0.3 km3 and 0.7 km3 .

The age of the collapse is constrained both by the occurrence of Mt. Epomeo Green Tuff (55 kyrs BP) fragments inside the avalanche deposit and by two 14C-AMS datings of Posidonia oceanica fragments (4254-4520 and 5050-5480 calibrated years BP, i.e. Neolithic time), at about 5 kyrs BP. The most recent dating [22] gives an age between ∼2,3 cal. ka B.P. and ∼3 cal. ka B.P. Finally, the resulting morphostructural setting of Ischia highlights a significant geo-hazard for the Ischia-Bay in the Naples area.

5

Western Pontine Archipelago

The Western Pontine Archipelago (Ponza, Zannone and Palmarola Islands) is located along the central sector of the Eastern Tyrrhenian Margin at the tip of a morphostructural high, facing the Vavilov Basin.

Ponza and Zannone are mainly made up of rhyolitic volcanites of calcalkaline affinity emplaced during the Pliocene, whereas Palmarola is made up of alkaline to peralkaline rhyolites emplaced during the Pleistocene. Pleistocene trachytic to peralkaline products also outcrop on the Southern portion of Ponza Island (e.g., [23, 24].

The rhyolites occur either as hyaloclastic

domes or massive lava dykes in response to different degrees of water-magma interaction in a shallow environment (e.g.

[25]). Erosion mainly dismantled the hyaloclastite, while dykes form headland, cliffs and submarine shoals. Over time, such a process has been producing an extremely complex morphology on the narrow (2-8 km) continental shelf (e.g. NW and SW Palmarola offshore). On the outer shelf, the morphology becomes more regular, as the rocky substrate is mainly covered

by Late Quaternary deposits, forming a depositional terrace.

Marine studies include remote sensing investigation as well petrographyc and petrological studies. In particular analysis of submarine rock samples provided new elements, i.e.: the presence offshore Ponza of limestone bedrock (previously described only on-land at Ponza and Zannone); the occurrence, in the SW sector of Palmarola Island, of Pliocene calc-alkaline rhyolites similar to those located at Ponza and Zannone Islands. This latter result is relevant for the general volcanological setting, as Palmarola was previously known to be constituted only of Pleistocene alkaline products. Therefore, this new setting significantly extends the distribution of calcalkaline magmatism in the area. In addition, submarine investigations have shown a compositional linkage between the Pleistocene volcanics of Palmarola and those of South Ponza. This finding reinforces the hypothesis that a progressive change of magmatism from calc-alkaline towards alkaline character occurred in the Western Pontine Archipelago during the Pleistocene, probably in response to a tectonic evolution defined by stronger extensional processes at that time [26, 24].

Beyond the shelf break at 120-150 m w.d., the continental slope is characterized by 439

Marine Geology

Figure 4: 3D perspective view of the Western Pontine Islands offshore. TOBI data are draped on the offshore DEM. a) Reference map; b) Bathymetric map (equidistance 200 m, the grey area corresponds to the tridimensional view). Vertical Exaggeration: 2. a suite of massive instability-erosive phenomena (Figure 4). The cannibalization (slope retreat) of most of the margin (98%) is a consequence of extremely high seafloor gradients (5°-10°, locally up to 30°), as well as the presence of tectonic features [27]. Instability-erosive processes differ in dimension (from less than one km2 to several tens of km2 ) and typology (e.g. slides, debris and grain flows). A well defined correlation between the mass movements distribution and the slope gradient exists, in fact, the steepest areas (∼10° up to 30°), such as the upper continental slope and the flanks of main ridges, are characterized by pervasive slides and erosive channels while gentler sloping areas, such as the lower continental slope, are dominated by grain flow and debris flow deposits (Figure 4).

Grain flow deposits (high backscatter facies in Figure 4) are formed by coarse

440

grained debris (e.g. sands and gravels) delivered by a large amount of linear channels. They develop on a quite steep seafloor (3°-20°) and seem to disappear when the gradient decreases below roughly 3°, reaching however a huge extent (about 150 km2 ). Debris flow deposits (medium backscatter facies in Figure 4) are constituted by relatively finer sediments; they are developed on low gradient seafloor (typically 0.5-2.5°) and cover a smaller area (about 65 km2 ) than that of grain flows.

Aside from the well known differences in sediments and transport mechanisms, density flows of the Western Pontine continental slope show different feeding systems. Grain flow deposits are fed by several punctual sources (i.e. the channels and gullies) so that the deposit is the result of amalgamated debris. In contrast, debris flows originate from a singular and distinct

Marine research at CNR

point source; actually, it is an area of coalescing scars.

The pervasive character of instabilityerosive phenomena affecting the continental slope enables the local crop out of the bedrock. Such outcrops are mainly located in areas affected by severe erosion or along fault escarpments. The petrographicpetrochemical characterization of collected samples evidences that most of them are volcanites comparable to the Pliocene calcalkaline rhyolites cropping out on-land at Ponza and Zannone. Taking into account the aforementioned volcanic outcrops observed on the continental shelf, a further significantly wider extent of calcalkaline rhyolites in this area is highlighted by offshore data.

6

Conclusive Remarks

The reported case-studies prove the importance of marine investigations to constrain the geology and evolution of insular volcanoes since these often represent the tip of very large edifices that are mostly developed underwater. In this regard, long range side scan and swath bathymetry seafloor mapping enable the detection of main volcano-tectonic features and masswasting processes at very high detail, also in deep water areas. Thus, marine geology unveils part of the evolution of volcanoes that is often difficult to recognize on the basis of the sole terrestrial geology. As an example, at Stromboli the recognition of the structurally-controlled Strombolicchio Canyon in the NE side of the edifice depicts a tectonic pattern more complex than the quasi-bilateral symmetry observed on the edifice. Furthermore, multistage lateral collapses have been reconstructed for the eastern Stromboli flank. Their volume

is an order of magnitude higher than previously thought, and comparable to that of the younger collapses affecting the NW flank (Sciara del Fuoco). Similarly, offshore the southern coast of Ischia Island, the recognition of a large debris avalanche deposit testifies the occurrence of largescale instability related to the resurgence of M.te Epomeo. This leads to reconsider the geological interpretation of the Ischia southern coast as due to a lateral collapse.

On the NW part of the Pantelleria edifice, marine studies identified a previously unknown large volcanic field, which includes the vent related to the last eruptive event (occurred in 1891). This evidence seems to indicate a northwestward migration of the activity over time.

Overall, the possibility to enlarge the view of a volcanic edifice allows to correlate the development of instability processes and volcano-tectonic structures to the local physiographic and structural setting. In this regard, the presence of a regional main fault system with NE-SW direction in the Stromboli area results in the NE-SW elongated shape of the edifice due to a strong tectonic control. Such setting determines that the NW and SE flanks are prone to collapse as they are steep and structurally unbuttressed. Similarly, at Ischia the structural control plays a major role, since the coalescence with the Campanian continental margin prevents the collapse of buttressed flanks.

As well as Stromboli, Pantelleria is a main central volcano with a prominent structural control due to the SCZR; however, its elongated shape is also due to the aforementioned northwestward migration of eruptive centers.

In the case of old and inactive volcanic edifices, such as the Western Pontine Archipelago, morphological evolution of 441

Marine Geology

submerged flanks is mainly related to erosion and mass wasting as constructive volcanic processes ended. On the Pontine continental slope dismantling processes are strengthened by tectonics that gives rise to very steep gradients.

Besides a fundamental role in the morphostructural reconstruction of insular volcanoes, marine studies also provide important

insights for the assessment of geo-hazards through the recognition, mapping and characterization of widespread mass-wasting features occurring at different scale and frequencies. As most of the instability phenomena evolve upslope, they may affect coastal infrastructures and may also generate tsunamis, as demonstrated by the Stromboli 2002 event.

References [1] A. Bosman, F.L. Chiocci, and C. Romagnoli. Morpho-structural setting of Stromboli volcano, revealed by high-resolution bathymetry and backscatter data of its submarine portions. Bulletin of Volcanology, 71(9):1007–1019, 2009.

[2] D. Casalbore, C. Romagnoli, F.L. Chiocci, and V. Frezza. Morpho-sedimentary characteristics of the volcaniclastic apron around Stromboli volcano (Italy). Marine Geology, in press, 2010.

[3] C. Romagnoli, D. Casalbore, F.L. Chiocci, and A. Bosman. Offshore evidence of large-scale lateral collapses on the eastern flank of Stromboli, Italy, due to structurally-controlled, bilateral flank instability. Marine Geology, 262:1–13, 2009a.

[4] A. Tibaldi. Multiple sector collapses at Stromboli volcano, Italy: how they work.

Bulletin of Volcanology, 63:112–125, 2001.

[5] C. Romagnoli, P. Kokelaar, D. Casalbore, and F.L. Chiocci. Lateral collapses and active sedimentary processes on the northwestern flank of Stromboli volcano, Italy.

Marine Geology, 265:101–119, 2009b.

[6] S. Tinti, E. Bortolucci, and C. Romagnoli. Computer simulations of tsunamis due to sector collapse at Stromboli, Italy. Journal of Volcanology and Geothermal Research, 96:103–128, 2000.

[7] F.L. Chiocci, C. Romagnoli, P. Tommasi, and A. Bosman. Stromboli 2002 tsunamigenic submarine slide: Characteristics and possible failure mechanisms. Journal of Geophysical Research, 113:1–11, 2008a.

[8] P. Tommasi, P. Baldi, F.L. Chiocci, M. Coltelli, M. Marzella, and C. Romagnoli.

Slope failures induced by the December 2002 eruption at Stromboli volcano. Learning from Stromboli and its 2002–03 eruptive crisis, 182:129–146, 2008.

[9] S. Tinti, A. Manucci, G. Pagnoni, A. Armigliato, and F. Zaniboni. The 30 December 2002 landslide-induced tsunamis in Stromboli: sequence of events reconstructed 442

Marine research at CNR

from the eyewitness accounts. Nat. Hazards and Earth System Sciences, 5:763– 775, 2005.

[10] F.L. Chiocci, C. Romagnoli, and A. Bosman. Morphologic resilience and depositional processes due to the rapid evolution of the submerged Sciara del Fuoco (Stromboli Island) after the December 2002 submarine slide and tsunami. Geomorphology, 100:356–365, 2008b.

[11] L. Civetta, Y. Cornette, G. Crisci, P. Gillot, G. Orsi, and C. Requejo. Geology, geochronology and chemical evolution of the island of Pantelleria. Geological Magazine, 121(6):541–562, 1984.

[12] G.A. Mahood and W. Hildreth. Geology of the peralkaline volcano at Pantelleria, Strait of Sicily. Bulletin of Volcanology, 48(2-3):143–172, 1986.

[13] A. Ricc`o. Terremoti, sollevamento ed eruzione sottomarina a Pantelleria nella seconda meta’ dell’ottobre 1891. Boll. Soc. Geogr. Ital., pages 1–31, 1892.

[14] H.S. Washington. The submarine eruption of 1831 and 1891 near Pantelleria. American Journal of Science, 27:131–150, 1909.

[15] L. Civetta, M. D’Antonio, G. Orsi, and G.R. Tilton. The geochemistry of volcanic rocks from Pantelleria Island, Sicily Channel: petrogenesis and characteristics of the mantle source region. Journal of Petrology, 39(8):1453–1491, 1998.

[16] A. Bosman, M. Calarco, D. Casalbore, F.L. Chiocci, M. Coltelli, A.M. Conte, E. Martorelli, C. Romagnoli, and A. Sposato. New insights into the recent submarine volcanism of Pantelleria Island. Abstract 26 Convegno Nazionale GNGTS 13-15/11/2007, pages 177–178, 2007.

[17] C. Siebe, J.C. Komorowski, C. Navarro, J. McHone, H. Delgado, and A. Cortes.

Submarine eruption near Socorro Island, Mexico: Geochemistry and scanning electron microscopy studies of floating scoria and reticulite. Journal of Volcanology and Geothermal Research, 68(4):239–271, 1995.

[18] J.L. Gaspar, G. Queiroz, J. M. Pacheco, T. Ferreira, N. Wallenstein, M.H. Almeida, and R. Coutinho. Basaltic lava balloons produced during the 1998-2001 Serreta Submarine ridge eruption (Azores). Subaqueous Explosive Volcanism, American Geophysical Union, Geophysical Monograph, 140:205–212, 2003.

[19] L. Vezzoli. Island of Ischia. Quaderni de La Ricerca Scientifica, 114(10):122, 1988.

[20] G. Orsi, G. Gallo, and A. Zanchi. Simple shearing block-resurgence in caldera depressions. A model from Pantelleria and Ischia. Journal Volcanol. Geoth. Res., 47:1–11, 1991.

[21] F.L. Chiocci and G. de Alteriis. The Ischia debris Avalanche: first clear submarine evidence in the Mediterranean of a volcanic island prehistorical collapse. Terra Nova, 18:202–209, 2006.

443

Marine Geology

[22] G. de Alteriis, D. Insinga, S. Morabito, V. Morra, F.L. Chiocci, F. Terrasi, C. Lubritto, C. Di Benedetto, and M. Pazzanese. Age of submarine debris avalanches and tephrostratigraphy offshore ischia Island, Tyrrhenian Sea, Italy. Marine Geology (Submitted), 2010.

[23] A.M. Conte and D. Dolfi. Petrological and geochimical characteristics of PlioPleistocene Volcanics from Ponza Island (Tyrrhenian Sea, Italy). Mineralogy and Petrology, 74:75–94, 2002.

[24] A. Cadeaux, D.L. Pinti, C. Aznar, S. Chiesa, and P.Y. Gillot. New chronological and Geochemcal constraints on the genesis and geological evolution of Ponza and Palmarola Volcanic Islands (Tyrrhenian Sea, Italy). Lithos, 81:121–151, 2005.

[25] D. De Rita, G. Giordano, and A. Cecili. A model for submarine rhyolite dome growth: Ponza Island (central Italy). Journal Volcanol. Geotherm. Res., 107:221– 239, 2001.

[26] A.M. Conte, D. Dolfi, E. Martorelli, and F.L. Chiocci. Aspetti petrologici dei prodotti magmatici delle Isole Pontine occidentali in relazione all’ambiente geodinamico. Atti del 4 Forum Italiano di Scienze della Terra - FIST GEOITALIA 2003, Bellaria, 16-18 settembre, 2003.

[27] F.L. Chiocci, E. Martorelli, and A. Bosman. Cannibalization of a continental margin by regional scale mass wasting: an example from the central Tyrrhenian Sea. Submarine Mass Movements and their consequences, 19:409–416, 2003.

444

The Bathymetry of the Adriatic Sea F. Foglini1 , E. Campiani1 , A. Cattaneo3 , A. Correggiari1 , A. Remia1 , D. Ridente2 , F. Trincardi1 1, Institute of Marine Sciences, CNR, Bologna, Italy 2, Institute of Environmental Geology and Geoengineering, CNR, Roma, Italy 3, French Research Institute for Exploration of the Sea, Brest, France federica.foglini@bo.ismar.cnr.it Abstract The Istituto di Scienze Marine (ISMAR-CNR) conducted several research projects in the Italian side of the Adriatic Sea over more than 15 years collecting bathymetric, geophysical, and sediment core data and performing multidisciplinary studies to reconstruct paleoenvironmental changes and sediment dynamics during the last eustatic cycle. A key issue in any marine dataset is the construction of a detailed bathymetry. The Adriatic Sea bathymetry is unique because, due to the shallow water depth of large areas of the basin, standard hydrographic surveys to obtain a comprehensive Multi Beam bathymetry are not applicable. The Adriatic Sea bathymetric map is necessarily based on heterogeneous data with uneven spatial distribution and includes both Single Beam echo-soundings and Multi Beam surveys in key study areas. The main objectives of this work are to illustrate the methodology applied to compile a new bathymetric map of the Adriatic Sea integrating Single Beam and Multi Beam data, to describe the morphological units reflecting the main geological features, and to discuss the limits of reliability of the data when a bathymetric map is to used by oceanographic modellers

1

Introduction

The bathymetry plays a key role not only in geological, geomorphological and geophysical studies, but also in the fields of physical oceanography and habitat mapping in submarine areas. In particular, the bathymetry represents a crucial constrain for oceanographic models in basin-scale circulation and in bottom-boundary-layer studies, and for the simulation of tsunami propagation across continental margins and in shallow areas. The seafloor morphology has been investigated for more than a century, but only with the technologies developed during the last decades it revealed

world-wide scale physiographic features such as mid-ocean ridges, transform faults and deep-sea trenches [2]. Heezen et al. [2] represented the morphology of the seafloor in a semi-pictorial way based on continuous echosoundings profiles together with an intelligent interpretation of the seafloor features to fill-in areas where no such soundings existed. Their purpose was to illustrate the morphology rather than to offer a precise measure of the water depth at any given point.

In 1922, De Marchi provided the first representation of the Adriatic Sea bathymetry giving, in particular, a conceptual image of the network of fluvial valleys incising

systems

surveyed

kHz

width

of Beams surveyed 126

Reson Seabat 8160

100-1200m

50

1.5°X1.5°

Kongsberg EM300

100-1000m

30

1°x 2°

135

Kongsberg EM710

100-900m

70-100

1°X1°

258

Konsberg EM3000

30-100

300

1.5° x 1.5° 127

Konsberg EM3002D 30-100 m

300

1.5° x 1.5° 508

Reson Seabat 8125

455

0.5°x1°

30-100

7700 km2

1600 km2

240

Tab. 1 - Technical characteristics of the Multi Beam echo sounders used to survey the South West Adriatic Marine Geology Margin and some selected areas of the Adriatic continental shelf.

Fig. 1 - General morphology of the Adriatic Sea from Giorgetti and Masetti, 1969 (Mercator Projection -

Figure 1: General morphology of thefrom Adriatic Sea from andwith Masetti [1]oceanographic (MerScale 1:750.000).

Compiled O.G.S Trieste from Giorgetti 1966-67 cruises the CNR vessel cator Projection - Scale from O.G.S Trieste from 1966-67 cruises “Bannok” and1:750.000). integrated withCompiled I.I.M published data. with the CNR oceanographic vessel ”Bannok” and integrated with I.I.M published data. the alluvial plain during the Last Glacial Maximum and drowned during the following sea level rise. Later, Giorgetti and Mosetti [1], constructed a map of the general morphology of the Adriatic basin based on a great number of echosounder records taken during several geophysical cruises that covered nearly the entire Adriatic Sea. The morphological map of Giorgetti and Mosetti [1] was drawn with the purpose of representing a pictorial view of the seafloor structures (Figure 1). Over the last 15 years, the Istituto di Scienze Marine (ISMAR-CNR) conducted several research projects in the Italian side of the Adriatic Sea collecting bathymetric data in order to obtain a high resolution map of

446

the seafloor features [3]. This map represented a key step for multidisciplinary studies aimed to reconstruct paleoenvironmental changes and sediment dynamics during the last glacial-interglacial cycle. The bathymetric map is necessarily based on heterogeneous data with uneven spatial distribution and includes both Single Beam echo-soundings and variablefrequency Multi Beam surveys in key study areas such as the South West Adriatic Margin and part of the continental shelf on the Italian side. The aim of this paper is to present the new bathymetric map of the Adriatic Sea compiled by CNR-ISMAR for the Italian side of the Adriatic Sea and to illustrate the main geological features of the

Marine research at CNR

Western Adriatic Basin. In this framework, we will examine the methodological approaches applied to process and to integrate single beam and multi beam echosoundings and we will discuss the limits of reliability of the data when the bathymetric map is to used by oceanographic modellers interested either in basin-scale circulation or in bottom-boundary-layer studies.

Kalman filter). Afterwards, the bathymetric data were migrated in a GIS (Geographic Information System) Database and plotted as water depth points in maps at different scale depending on the soundings density, dividing the Adriatic in subset areas from north to south. The bathymetric contours were manually drawn and digitised as vector data in a GIS, with variable space according to water depth range (contour every 1 m from 5 to 150 m and every 2 Methodology 20 m from 150 m to 1200 m). The contours were used to generate a uniform grid 2.1 Single Beam bathymetry - (200 m) applying computing technology Data acquisition and pro- (KRIGING algorithm with variable resolution depending on the soundings density) cessing in order to give a more flexible product for Single Beam sonar data were collected us- visualisation and manipulation of data. ing an hull mounted Echo sounder (Atlas Deso25 operating at frequencies of 12, 100, 33 and 210 kHz) along about 2.2 Multi Beam bathymetry 17.830 km of seismic profiles during 22 Data acquisition and procruises performed by ISMAR from 1991 cessing to 2005 on board R/V Urania in the Italian side of the Adriatic Sea (Figure 2). The The Multi Beam data were collected with echo sounding profiles are unevenly dis- a variety of Multi Beam Echo. with varitributed and the seafloor coverage is within able frequency, beam number and beam anthe range of one sounding every 20-40 gles, according to the scientific objectives m, along track. The Echo sounder At- or to the instruments available. The Multi las Deso25 was merged with the naviga- Beam acquisition strategy comprised a full tion system NAV PRO from Communica- coverage survey of the entire South Adrition Technology and with DGPS position- atic Continental slope on the Italian side ing system with metrical accuracy. The (from 200 to 1200 m) and the investigation sound speed was set at 1500 m¡sâ&#x2C6;&#x2019;1 . In ar- of selected areas of the continental shelf eas where the sound speed profiles were (from 10 to 150 m) characterised by comhighly variable, as for example in front plex seafloor morphology. Table 1 sumof the Po River Delta, local values ob- marizes the Multi Beam Echo. technical tained from CTD (Conductivity, Temper- characteristics and the water depth range ature, and Depth) measurements were ap- surveyed. The Multi Beam data processplied. The navigation and water depth data ing was carried out using both PANGEA were stored every 2 seconds in a Database Multi Beam System and CARIS HIPS and and filtered to correct positioning errors SIPS 7.0. The processing methodology and to delete null values using filtering applied implies the creation of a 2D and procedures implemented by ISMAR (e.g. 3D best-fit interpolated surface using dif447

Marine Geology

Fig. 2 â&#x20AC;&#x201C; Single Beam data (grey lines) collected by ISMAR from 1991 to

Figure 2: Single Beam data (grey lines) collected by ISMAR from 1991 to 2005 on board R/V Urania in the Italian side of the Sea. TheSea. blue boxes the areas surItalian side ofAdriatic the Adriatic The represent blue boxes represent the areas surveye veyed with high frequency Multi Beam systems on the continental shelf, the red boxes shelf, red boxes represent the areas surveye represent the areassystems surveyed on withthe lowcontinental frequency Multi Beamthe systems on the Adriatic continental slope.

systems on the Adriatic continental slope.

ferent algorithms and grid resolutions. The first processing step is the analysis of the data errors and the definition of a strategy to solve them. The latter includes: 1) the correction of the sensor angles (multi beam patch-test); 2) the sound speed correction applying the ray-tracing technique after data acquisition; 3) the manual cleaning of the spikes (beam remove) only in the area where they are visible on the 2D and 3D surface; 4) the automatic filtering for a depth window, by beam number or

448

slope between points. The second processing step is the creation of a new surface after the data correction and cleaning using a different resolution grid for a given water depth range and for geologically relevant seafloor features. The processing methodology applied is based on an interpretative approach instead of a traditional line to line cleaning system. The role of the data processing is to understand the reliability of the seafloor features detected on the grid surface and to identify all kinds of

Marine research at CNR

Fig.3 - Single Beam bathymetry contour map of the Italian side of the Adriatic (contour lines every 1 m from

Figure Single bathymetry map the Italian theAdriatic Adriatic 5 to3:150 m and Beam contour every 20 m fromcontour 150 to 1200 m). of A) Incised valleysside on theofnorth shelf,(contour lines 1m from755and to 100 150m m and contour every from exposure 150 to of 1200 m). A) offshoreevery Ancona, between of water depth formed during20 them subaerial this area Incised valleys theMaximum north Adriatic shelf, Ancona, between 75 and during the Laston Glacial and the early stagesoffshore of the post-glacial sea-level rise; B) Sand dunes 100 on them of waternorth depth formed the of 24 this areadepth. during the Last Glacial Adriatic shelf areduring located 20 kmsubaerial SE of Veniceexposure between 20 and m water Maximum and the early stages of the post-glacial sea-level rise; B) Sand dunes on the . north Adriatic shelf are located 20 km SE of Venice between 20 and 24 m water depth. noise and their origin. The interpretative approach leads to achieve a high resolution bathymetry focusing the processing effort in revealing the geologically most significant seafloor features. This method is less time consuming in terms of manual data cleaning and implies a change in the perspective of the entire processing work flow.

2.3

Multi Beam and Single Beam data integration

The Single Beam and Multi Beam data were integrated at regional scale using the software PANGEA MB Manager and its “Tuning filter” tool. The “Tuning filters” represent special areas, drawn by the operator as polygons with variable shapes, where it is possible to apply a specific resolution. The software assigns an ID number to each Tuning filter with its associated resolution value. The Single Beam 449

Marine Geology

Multi Beam systems

Water depth range Frequency Beam surveyed

N°

kHz

width

of Beams surveyed

Reson Seabat 8160

100-1200m

50

1.5°X1.5°

126

Kongsberg EM300

100-1000m

30

1°x 2°

135

Kongsberg EM710

100-900m

70-100

1°X1°

258

Konsberg EM3000

30-100

300

1.5° x 1.5° 127

Konsberg EM3002D 30-100 m

300

1.5° x 1.5° 508

Reson Seabat 8125

455

0.5°x1°

30-100

Total area

7700 km2

1600 km2

240

Tab. 1Table - Technical characteristics of the Multi Beam echoBeam sounders used to survey the to South Westthe Adriatic 1: Technical characteristics of the Multi echo sounders used survey

Westselected Adriatic Margin somecontinental selected areas MarginSouth and some areas of theand Adriatic shelf.of the Adriatic continental shelf. and the Multi Beam DTM (Digital Terrain Model, respectively 200 m and 20 m resolution) were loaded in the software as soundings. According to the sounding density, at a given water depth range, the operator drawn Tuning Filters including areas with homogenous characteristics. For each areas a different resolution was assigned in order to emphasize the most relevant seafloor features and to maintain the details of the Multi Beam DTM. At the border between the Multi Beam surveys and the Single Beam ones, several Tuning filters were drawn increasing progressively the resolution, going from the Single Beam to the Multi Beam data, in order to minimize the differences between the two areas and to avoid the creation of morphological steps. The Tuning filters allowed the creation of a DTM with variable resolution depending on the operator choice, permitting to merge Single Beam and Multi Beam data maintaining the high resolution of the area surveyed with Multi Beam E.

3 3.1

Results and discussion Single Beam bathymetry of the Western Adriatic Sea

The new contour map of the Western Adriatic Sea resulting from the acquisition, processing and interpretation of data collected

by ISMAR over the last 15 years shows in detail the seafloor morphology of the Adriatic from the northern shelf to southern slope (Figure 3). The bathymetry shows that the Northern Adriatic has a low longitudinal topographic gradient (ca 0.02°), whereas the maximum shelf gradient along the central Adriatic is on the order of 0.5°.

The central Adriatic is characterised by a narrower shelf and localised bathymetric irregularities that are the expression of structural highs offshore Punta Penna, the Tremiti Islands and the Gargano promontory, and reaches a maximum depth of 260 m in two remnant slope basins aligned in a SW-NE direction. The Southern Adriatic, beyond the Pelagosa Sill, reaches the of ca m and flanked Projection by a Fig. 1 - General morphology of the Adriatic Sea fromdepth Giorgetti and1200 Masetti, 1969 is(Mercator steep slope.

In this area, the shelf is genScale 1:750.000). Compiled from O.G.S Trieste from 1966-67 cruises with the CNR oceanographic vessel erally narrow except in the Gulf of Man“Bannok” and integrated with I.I.M published data.

450

Marine research at CNR

Figure 4: Slope map of the Late Holocene clinoform on the Adriatic shelf. The offlap 4 - values) Slope occurs map of the Late Holocene clinoform on the Adriatic shelf. The o break (yellow and Fig. red slope in progressively deeper waters from the Po Delta (few meters water depth) to the area offshore Gargano The geometry of the clinoslope values) occurs intheprogressively deeper waters from the Po Delta (few met form varies between Ancona and Ortona where gradient gradually increases.

offshore Gargano The geometry of the clinoform varies between Ancona and fredonia, south of the Gargano Promongradually increases. tory, where it broadens to about 80 km.

The main morphological features detected on the Single Beam bathymetry reflect the following main geological elements of the area:

ure 4). The Late Holocene clinoform on the Adriatic shelf reaches up to 35 m in thickness with a volume of 180 km3 and rests above the maximum flooding surface (mfs), a regional downlap surface dated ca. 5.5 cal kyr BP [4, 5, 6]; • Several incised valleys on the North • The Late Holocene mud wedge clinoAdriatic shelf, offshore Ancona, between form extending over 600 km along the 100 and 75 m of water depth formed durcoast of Italy from the modern Po Delta ing the subaerial exposure of this area to the area south of the Gargano Promonduring the Last Glacial Maximum and in tory, with a characteristic subaqueous ofthe early stages of the post-glacial sea flap break, marking the transition belevel rise (Figure 3A). The valleys are tween topset and foreset deposits (Fig451

Fig. 5 - Subsurface undulations between 28 m and 93 m water depth the foreset progradational clinoform in the area offshore Ortona (Location in Fig 2 -

(maximum slope of 2°) are associated with mud reliefs that occur further seaward

Fig. 4 - Slope map of the Late Holocene clinoform on the Adria slope values) occurs in progressively deeper waters from the Po

offshore Gargano The geometry of the clinoform varies betwee gradually increases.

Marine Geology

Figure 5: Subsurface undulations between 28 m and 93 m water depth the foreset region of the late Holocene progradational clinoform in the area offshore Ortona (Location in Fig. 5 -undulations Subsurface undulations between 28 m Figure 2 - Area 1). The (maximum slope of 2°) are associated with mudand reliefs that occur further seaward (60 m water depth).

93 m water d

progradational clinoform in the area offshore Ortona (Locati

up to 20 km long, several hundred me(maximum slope of tres to a kilometre wide and between 4 and 15 m deep. The orientation of the valleys is predominantly north-south and their sinuosity is low [7]. The valleys are spatially associated to preserved barrierlagoon deposits, which originated during the Late Pleistocene and Holocene sealevel rise [8, 9]; • Sand dunes on the North Adriatic shelf are located 20 km SE of Venice between 20 and 24 m water depth (Figure 3B) [10]. The sand dunes rest on a broad shore-parallel mound bounded landward by an elongated trough. The dunes are up to 2 km long, characterised by low 452

sinuosity and extend across the entire 2°) are associated with mud reliefs that occur width of the underlying mound to the edge of the shoreline parallel trough [10].

The sand dunes off shore the Venice Lagoon are formed from the reworking of a drowned coastal lithosome accompanied by secondary erosion in the troughs and recycling of low stand fluvial sand [10].

Marine research at CNR

Fig. 6 â&#x20AC;&#x201C; 6: DTM (20 m (20 grid) m of the South showing theMargin extreme geological Figure DTM grid) ofWest the Adriatic South Margin West Adriatic showingand the extreme morphological complexity of the slope.

A) Areas with enhanced bottom-current features in the upperbottomslope. geological and morphological complexity of the slope. A) Areas with enhanced B) The Gondola Slide representing the largest mass failure deposit on the SAM. c) The Bari Canyon System, current in the active uppersince slope.

B) Glacial The Gondola the mainfeatures sediment conduit the Last Maximum Slide representing the largest mass failure deposit on the SAM. C) The Bari Canyon System, the main sediment conduit active since the Last Glacial Maximum.

3.2

Multi Beam bathymetry of the Italian side of the central Adriatic continental shelf offshore Ortona

The Multi Beam map the Adriatic Continental shelf offshore Ortona (Figure 5. location in Figure 2) defines the seafloor expression of subsurface undulations between 30 m and 75 m water depth (typically 300 m wide, 2.5 m high, and several km long, parallel to the bathymetric contour) affecting the foreset region of the late Holocene progradational clinoform above a regional downlap surface (the mfs) in areas where it shows evidence of defor-

mation and fluid escape [5, 11, 12, 13].

The undulations are associated with mud reliefs that occur farther seaward in elongated swarms perpendicular to the regional slope and to the crests of the undulations.

Cattaneo et al. [11], suggested that these seafloor undulations evolved in response to sediment deformation and were successively amplified by differential deposition from bottom currents crossing an irregular seafloor. Recently Sultan et al. [14] demonstrated that the basal unit of the Holocene mud wedge immediately above the mfs has coarser grain size than the underlying and overlying units. The latter represents a weak layer where liquefaction

453

Marine Geology

can occur during earthquake of M1≤4.5, 35 m thick, it has a total runout of about typical of this area.

54 km [15] and a volume of the deposit of about 30 km3 . The evacuation zone includes a crescent-shaped headscarp lo3.3 Multi Beam bathymetry of cated at the shelf edge with a maximum the South West Adriatic height of 250 m with several sub-parallel secondary scarps [15]. The morphological slope pattern reflects the interaction between the The high resolution Multi Beam map of complex relief created by down-slope gravthe South West Adriatic Margin (SAM) ity flows and along-slope bottom currents shows the extreme geological and morpho- [15, 18]. The Bari Canyon System (Figlogical complexity of the slope and allows ure 6C) is a peculiar erosional-depositional detailed description of the seafloor fea- feature characterised by two main, almost tures (Figure 6). The SAM slope is gener- parallel, conduits emanating from a broad ally characterised by: 1) widespread mass- crescent-shaped upper slope region [17]. failure features including slide scars up to This setting is consistent with the flow of 10 km wide and extensive slide deposits bottom currents along the shelf from the with runout distances greater than 50 km North entering the canyon and interacting [15]; 2) a large variety of bottom-current with its complex topography, leading to features [16, 17]; 3) the Bari Canyon Sys- preferential deposition on the up-current tem (BCS), the main sediment conduit ac- side of pre-existing morphological relief tive since the Last Glacial Maximum inter- [17]. Today, density-driven bottom curval [17]; 4) the Dauno Seamount, the main rents cascade off shelf and flow both across structural feature on the slope, with a clear the open slope and through the BCS, reachmorphologic expression. A large variety of ing velocities greater than 60 cm·s−1 [19]. bottom current features (Figure 6A) characterises a confined sub-triangular slope area suggesting the constructive interaction 3.4 Bathymetry and Oceanogbetween two distinct southerly bottom waraphers - Multi Beam and ter masses: the contour-parallel Levantine Single beam combined map Intermediate Water and the North Adriatic Dense Water, cascading seasonally across In the case of basin scale circulation modthe slope. By analyzing the large variety els, the Single Beam bathymetry of the of bottom-current features, it was possible Italian side of the Adriatic compiled by to identify areas in the upper slope strongly ISMAR can be applied in oceanographic swept by bottom currents and characterised numerical model as a uniform resolution by predominant erosion. Seaward and east- grid. The main limitation in using this grid ward of the main current path, the bottom comes from the heterogeneity of the bathycurrent progressively loses energy, through metric data in terms of distribution and a field of progressively more continuous quality, and the possible errors generated and aggradational sediment waves [16, 18]. during the interpolation procedures applied The Gondola slide (Figure 6B) is one of to derive a grid with homogeneous resoluthe largest mass failure deposits on the tion. The Multi Beam bathymetry is more SAM. It is 10 km wide on the slope, up to suitable for bottom boundary layer appli454

Marine research at CNR

cations and for tsunami-propagation simulation models in specific areas. The reliability of Multi Beam data is higher because the Multi Beam Sys. guarantee a full coverage of the seafloor ensuring an homogeneous data quality. The processing of this kind of data leads to the reduction of instrumental noises and to the generation of a high resolution DTM where the uncertainty, given by the interpolation procedures, is extremely reduced. The integration of Single Beam with Multi Beam bathymetry, using a variable resolution grid, allows the generation of a complete bathymetric map functional at different scales. The resulting combined bathymetry is useful for the oceanographers in detecting areas of maximum strength of bottomhugging currents and defining the regional morphological trends; for example in areas of flow restriction caused by the presence of narrow passageways or shallow shoals.

4

Conclusions

Due to its large extent (200 x 800 km) and its physiographic setting with a wide shelf area in the North and a slope basin in the South, the Adriatic bathymetric map is the result of a merger of dataset from numerous oceanographic surveys performed during the last decades with variable tools. In particular, large areas at shallow depth have been mapped with Single Beam tools and interpolated, because an extensive Multi Beam mapping would have been too time consuming in such conditions. The Single Beam contour map of the Italian side of the Adriatic Sea compiled by ISMARCNR shows, at basin scale, the following main geological features: 1) the coastparallel extent of the late Holocene mud

wedge; 2) the occurrence of incised valleys on the north Adriatic shelf; 3) the distribution of sand ridges and sand dunes of variable size on the north Adriatic shelf. The Multi Beam maps of the South Adriatic Continental slope on the Italian side (from 200 to 1200 m water depth) and of some selected areas of the continental shelf (from 10 to 150 m) show: 1) the complexity and variability of the progradational clinoforms of the late Holocene prodelta wedge; 2) widespread mass-failure features on the slope; 3) a large variety of bottom-current features; 4) the Bari Canyon System active during the Last Glacial Maximum, but still impinged by shelf density currents; 5) the Dauno Seamount, the main structural feature on the slope, with a clear morphologic expression. The Multi Beam data processing was based on an interpretative approach instead of a traditional line to line cleaning system. This approach allows to achieve a high resolution bathymetry focusing the processing effort in revealing the geologically most significant seafloor features. The integration of Single Beam with Multi Beam bathymetry using a variable resolution DTM (Tuning filter tool) allows the generation of a complete bathymetric map useful at different scales. The resulting combined bathymetry is useful not only for marine geologists, but also for oceanographers in detecting areas of maximum strength of bottom-hugging currents and defining regional morphological trends. In perspective, the methodology illustrated here could be furthered with the acquisition of the bathymetry on the East side of the Adriatic Sea, through international scientific projects in collaboration with eastern Adriatic countries such as Croatia, Slovenia, Montenegro and Albania.

455

Marine Geology

References [1] G. Giorgetti and F. Masetti. General morfology of the Adriatic Sea. Bollettino di Geofisica teorica ed applicata, 11:44–56, 1969.

[2] B.C. Heezen, M. Tharp, and M. Ewing. The floors of the oceans, Part I, The North Atlantic. Geol. Soc. America Special Paper, 65:122, 1959.

[3] A. Correggiari, D. Penitenti, D. Ridente, M. Roveri, and F. Trincardi. La batimetria ad alta risoluzione del Mare Adriatico: una base di lavoro per studi multidisciplinari. Primo Workshop SINAPSI Roma 6-8 aprile. 1998.

[4] A. Asioli. High resolution foraminifera biostratigraphy in the Central Adriatic basin during the last deglaciation: a contribution to the PALICLAS Project. In: Guilizzoni, P., Oldfield, F. (Eds.), PalaeoenvironmentalAnalysis of Italian Crater Lake and Adriatic Sediments (PALICLAS). Memorie del’Istituto Italiano di Idrobiologia, 55:197 – 218, 1996.

[5] A. Correggiari, F. Trincardi, L. Langone, and M. Roveri. Styles of failure in heavilysedimented highstand prodelta wedges on the Adriatic shelf. Journal of Sedimentary Research, 71(2):218–236, 2001.

[6] A. Cattaneo, A. Correggiari, L. Langone, and F. Trincardi. The Late-Holocene Gargano subaqueous delta, Adriatic shelf: sediment pathways and supply fluctuations. Marine Geology, 193:61–91, 2003.

[7] J.E.A. Storms, G.J Weltije, G.J. Terra, A. Cattaneo, and F. Trincardi. Coastal dynamics under conditions of rapid sea-level rise: Late Pleistocene to Early Holocene evolution of barrier-lagoon systems on the Northern Adriatic shelf (Italy). Quat.

Sci. Rev., 27:1107–1123, 2008.

[8] F. Trincardi, A. Correggiari, and M. Roveri. Late Quaternary transgressive erosion and deposition in a modern epicontinental shelf: The Adriatic semienclosed basin.

Geo-Marine Letters, 14:41–51, 1994.

[9] A. Cattaneo and F.Trincardi. The late-Quaternary transgressive record in the Adriatic epicontinental sea: basin widening and facies partitioning. Isolated Shallow Marine Sand Bodies: Sequence Stratigraphic Analysis and Sedimentologic Interpretation. SEPM Spec. Publ, 64:127–146, 1999.

[10] A. Correggiari, M.E. Field, and F. Trincardi. Late Quaternary transgressive large dunes on the sediment-starved Adriatic shelf. Geology of Siliciclastic Shelf Seas.

Geological Society Spec. Publ, 117:155–169, 1996.

[11] A. Cattaneo, F. Trincardi, L.Langone, A. Asioli, and P. Puig. Clinoformation generation on Mediterranean Margins. Oceanography, 17(4):104–117, 2004.

456

Marine research at CNR

[12] T. Marsset, B. Marsset, Y. Thomas, A. Cattaneo, E. Thereau, F. Trincardi, and P. Cochonat. Analysis of Holocene sedimentary features on the Adriatic shelf from 3D very high resolution seismic data (Triad survey). Marine Geology, 213:73–89, 2004.

[13] F. Trincardi, A.Cattaneo, A. Correggiari, and D. Ridente. Evidence of soft-sediment deformation, fluid escape, sediment failure and regional weak layers within the Late-Quaternary mud deposits of the Adriatic Sea. Marine Geology, 213:91–119, 2004.

[14] N. Sultan, A. Cattaneo, R. Urgeles, H. Lee, J. Locat, F. Trincardi, S. Berne, M. Canals, and S. Lafuerza. A geomechanical approach for the genesis of sediment undulations on the adriatic shelf. Geochemistry Geophysics Geosystems, 9(4):1–25, 2008.

[15] D. Minisini, F. Trincardi, and A. Asioli. Evidence of slope instability in the Southwestern Adriatic Margin. Natural Hazards Earth System Sciences, 6(1):1–20, 2006.

[16] G. Verdicchio and F. Trincardi. Short-distance variability in slope bed-forms along the southwestern Adriatic margin (central Mediterranean). Marine Geology, 234(1/4):271–292, 2006.

[17] F. Trincardi, F. Foglini, G. Verdicchio, A. Asioli, A. Correggiari, D. Minisini, A. Piva, A. Remia, D. Ridente, and M. Taviani. The impact of cascading currents on the Bari Canyon System, SW-Adriatic Margin (Central Mediterranean). Marine Geology, 246(2-4):208–230, 2007.

[18] G. Verdicchio, F. Trincardi, and A. Asioli. Mediterranean bottom-current deposits: an example from the Southwestern Adriatic Margin. Geological Society, London, Special Publications, 276:199–224, 2007.

[19] M. Turchetto, A. Boldrin, L. Langone, S. Miserocchi, T. Tesi, and F. Foglini. Particle transport in the Bari canyon (southern Adriatic Sea). Marine Geology (Van Weering and Heussner, Eds.)., 246(Issues 2-4):231–247, 2007.

457

Marine Geology

458

Early Diagenesis of Carbon and Nutrients in Sediments of the Gulf of Manfredonia (Southern Adriatic Sea) F. Spagnoli1 , G. Bartholini1 , E. Dinelli2 , M. Marini1 , P. Giordano3 1, Institute of Marine Sciences, CNR, Ancona, Italy 2, Interdepartmental Research Centre for Environmental Sciences, University of Bologna, Ravenna, Italy 3, Institute of Marine Sciences, CNR, Bologna, Italy f.spagnoli@ismar.cnr.it Abstract The aims of the research presented here were to investigate the carbon and nutrient cycles in sediments of the Gulf of Manfredonia and to understand their role in the water column chemistry. Four cores were collected in two sites of the gulf in early fall 2002 and late winter 2003. The cores were extruded for pore water (TCO2 , DOC, alkalinity, nutrients, Si(OH)4 , sulphate, Fe, Mn, Ca and Mg) and solid (grain size, organic and total C, total N, 210 Pb excess, 137 Cs, 234 Th) analyses. Furthermore, fluxes at sediment-water interface have been measured by benthic chambers and calculated from pore water concentration profiles for O2 , TCO2 , DOC, alkalinity, nutrients, Si(OH)4 , sulphate, Fe and Mn. Pore water data evidenced high inputs of reactive organic matter in the two stations, diagenesis of organic matter progresses through oxygen consumption, denitrification, Mn-Fe-oxy-hydroxide reduction and weak sulphate reduction. Degradation processes are more intense during the warm season. Bio-irrigation seems to be a consistent transport mechanism in both stations, with more evident effects in early fall in the outer station. Measured benthic fluxes showed no clear difference between sites with slight higher values in the offshore site in the warm season. Benthic flux comparison of the Gulf of Manfredonia with the northern Adriatic allowed evaluating the role of the gulf sediments in the chemistry of the south-western Adriatic Sea waters.

1

Introduction

The sediment-water interface is a site of intense chemical, physical and biological reactions that can lead to the formation of new mineralogical phases, alteration of existing minerals and to changes in the composition of pore water and water column themselves. Reactions involving the oxidation of organic matter are carried out

largely by bacteria that use a sequence of electron acceptors at decreasing redox potentials and increasing depth in the sediment. It is well know that different terminal acceptors are used by the microbial community in the order of decreasing free energy production per mole of organic carbon oxidized [1]. These reactions are controlled by external factors, such as temperature, sedimentation rate, organic mat-

Marine Geology

ter inputs, sea bottom water chemistry and hydrodynamics, bioturbation and irrigation [1]. In coastal marine environments, with shallow waters and fine sediments, early diagenesis processes play an important role in the biogeochemical cycling of nutrients, i.e. they determine the amount of nutrients buried versus the amount of nutrients recycled to the water column as benthic fluxes [2, 3]. Such benthic fluxes can represent an input comparable to, or higher than, inputs from terrestrial sources. Hence, nutrient inputs and consequently the primary productivity in shallow coastal ecosystems can be closely associated with sea-floor biogeochemical processes [4].

In this paper the results of a study regarding the interactions between sediments and water column carried out in the PITAGEM Project [5] are presented. The study focuses on early diagenesis processes and benthic flux measurements in the Gulf of Manfredonia, a shallow basin in southwestern Adriatic Sea (Figure 1), where seasonal variation of nutrient inputs and coexistence of multiple nutrient sources can produce diagenetic process variations and influence quality and quantity of decomposing matter. By studying the early diagenesis processes and the benthic fluxes it was also possible to evaluate organic matter and nutrient regeneration in surface sediments and their role in the chemistry of the coastal waters of a confined area of the south-western Adriatic Sea.

2

Study area

The Gulf of Manfredonia is located in the western side of the southern Adriatic Sea (Figure 1). The gulf is delimited to the North by the Gargano Peninsula, which morphologically marks the northern bor460

der of the southern Adriatic.

The general Adriatic Sea hydrodynamic presents a cyclonic circulation in the southern sub-basin with a strong seasonal variability [6] and a southward coastal current, enriched in nutrients, flowing along the western coast of the Adriatic basin (Western Adriatic Current, WAC, [6]). The WAC connects the northern and southern ecosystems and affects the biogeochemical properties of the whole western Adriatic basin. However, in the southern Adriatic Sea, open waters show clearly oligotrophic characteristics and the nutrient supply to the euphotic zone depends strongly on the vertical stratification and mixing processes [7].

This general hydrodynamic pattern strongly affects the biogeochemical properties of the Gulf of Manfredonia that represents a transition zone between the middle and the southern Adriatic circulation.

The gulf presents a main cyclonic gyre [8] that may reverse depending on wind direction: cyclonic and anticyclonic gyres are respectively generated by N-NW and S-SE winds (Signell, personal communication); this circulation is slower in the inner side of the gulf where a high sedimentation rate is present [9]. Furthermore, the inner area of the gulf is characterised by high water column nutrient contents and high primary production (Spagnoli, unpublished data).

As regards bottom sediments, previous studies based on grain size analyses in surface sediments recognized 3 sediment types: silt-sandy, silt and clay [10, 11].

The distribution of these sediment types is essentially the result of wave influence: grain-size in fact decreases with increasing water depth and distance from the shore.

Silt-sandy sediments are common in zones adjacent to the shoreline between 2 and 4m depth. Silty and clay sediments are found

Marine research at CNR

42.1 42.0 400

41.9

Vieste 0 20

41.8 S2 41.7 41.6

Manfredonia Candelaro

S1 12 0

0 10

41.5 41.4 50

Ofanto

41.3

10

20

Barletta

41.2 41.1

Projection ED50

0 41.0 15.0

Bari Bari

20

40

15.2

15.4

km

15.6

15.8

16.0

16.2

16.4

16.6

16.8

17.0

Figure 1: Study area with early fall 2002 and late winter 2003 stations. S1 is the inner station, S2 is the outer station. in the central part of the gulf (> 8m of depth) and in the deeper zones. Furthermore biogenic clastic fragments are found in some coastal sites (> 6m of depth) due to breaking of biogenic concretions.

3

Methods

During two oceanographic cruises carried out in early fall (October 2002) and late winter (March 2003) gravity cores were collected in two stations (Figure 1) by using a SW-104 corer, a device which assures the collection of undisturbed sedimentwater interfaces [12]. The stations are localised in the inner side of the gulf at 16 m depth (S1), supplied by mixed fine sediments coming from northern Apulian

rivers (mainly the Ofanto River) and from the northern Adriatic Sea, and in the outer side of the gulf at 17 m of depth (S2), fed mainly by northern fine-grained sediments [8]. The waters overlying the cores were generally clear, suggesting minimal disturbance and the sediment surface was generally uneven. On board, cores were sectioned and centrifuged in inert atmosphere for pore water and solid phase separation and analyses. Each core was sectioned in a nitrogen-filled glove-box and punch in pH and Eh measurements were made during the sectioning. In order to extract pore waters, the mud was transferred into plastic tubes and centrifuged for about 15 min at 5500 rpm in a refrigerated centrifuge at the in situ temper-

461

Marine Geology

atures. Pore waters were filtered under a nitrogen atmosphere in plastic syringe holders (hydrophilic PTFE 0.45 µm membrane). Four not acidified splits were saved for alkalinity, TCO2 (total dissolved inorganic carbon), DOC (Dissolved Organic − 3− Carbon), NH+ 4 , NO3 , PO4 , Si(OH)4 and 2− SO4 measurements, another aliquot was acidified (to about pH 1.5) and used for the analysis of dissolved Fe, Mn, Ca and Mg. Pore water extraction and filtration was generally completed within 6 hours from core collection. The centrifuged mud was frozen and subsequently dried at room temperature for analysis of solid phases (organic carbon (C-org), total carbon (Ctot) and total nitrogen (N)). Another mud aliquot was stored at 4°C and used for grain-size analysis. In the first cruise replicate cores have been collected to estimate sedimentation rates and bioturbation coefficients by 210 Pb excess, 137 Cs and 234 Th measurements. The excess activities of 210 Pb were calculated from 226 Ra supported 210 Pb deduced from the activities of 214 Pb and 214 Bi. Radionuclides were counted using an HPGe (30-60% relative efficiency, 2 KeV of resolution). In the second cruise replicate cores were collected to measure dissolved oxygen penetration depth by a microelectrode profiler.

Furthermore,in situ benthic flux chambers were deployed in each site to measure the dissolved fluxes of O2 , TCO2 , DOC, alka− 3− linity, NH+ 4 , NO3 , PO4 , Si(OH)4 , Fe and Mn. Two chambers were deployed in each site to replicate the measurements. Benthic chambers capture approximately 39 l of water in contact with 0.25 m2 of sea bottom. Each chamber was sampled six times during deployments of about 24h.

The incubation time of 24h was sufficient to generate measurable changes in concentration, but not enough to produce signifi462

cant changes in fluxes. A CsCl spike was injected in the incubations after the first sampling and the observed dilution of this spike in subsequent sample drawings was used to calculate chamber volume and also as tracer for chamber-water exchanges with pore waters. Chambers were stirred by a rotating paddle so that the diffusive boundary layer thickness was unaffected by the incubation.

In sediment solid phase, porosity was calculated by wet loss after drying each sample at 60°C, total and organic carbon and total nitrogen (all expressed as weight %) were measured by CHN elemental analyzer (Carlo Erba 1500) after removal of the inorganic carbon with HCl [13]. Organic nitrogen was assumed equal to total nitrogen [14].

Dissolved phosphate, ammonia, nitrate and silicate were measured on pore water and benthic chamber samples by colorimetric autoanalyser technique. Alkalinity was determined by Gran titration [15] and TCO2 was determined from alkalinity, pH, salinity and temperature. Additionally, in the second cruise, an aliquot of 5ml was collected for the analyses of TCO2 by manometric measurements [16]. Fe, Mn and Cs were determined on the acidified aliquots

by Flame-AAS. Ca and Mg were determined by HPLC. For each parameter analysed benthic fluxes were calculated as the product of the slope of concentration vs. incubation time and chamber height, without including bottom water data in the calculation. As the incubation time increases, there is the possibility that chamber chemistry or other artefacts do not produce constant fluxes and this is recognizable by data trends. For this reason benthic flux calculations were determined only from the linear portion of the concentration vs. time plots [17] and no flux was reported if we

Marine research at CNR

had fewer than 4 data points. Using two deployments, fluxes were averaged and the uncertainty in the mean was calculated in two ways: (a) as the standard error of the mean; (b) as the square root of the sum of the variance of each flux value, divided by the number of flux measurements. We report the larger of these two uncertainties, following the procedure described by [18].

4 4.1

Results and discussion General setting and pore water profiles

In early fall 2002 the water column was homogeneous with a mean temperature of about 21째C in both stations.

In the late winter 2003 the water column was characterised by homogeneous temperature at 11째C in the inner station and a weak stratification in the upper 2 m in the outer station (bottom water at 12째C).

Water column productivity in the gulf has been shown to be higher in summer months (Spagnoli unpublished data).

Dissolved oxygen, in continental slope and rise sediments underlying well-oxygenated bottom waters, is the most important electron acceptor for the organic matter decomposition. In sediments of these environments anoxic conditions occur from few millimetres to few centimetres of depth depending on the balance between sedimentation rates and organic matter inputs. For this reason the sedimentary oxygen consumption rate is a good indicator of organic matter oxidation rates in most of these sites. In the sampled stations dissolved oxygen decreases exponentially below the sediment-water interface and at a depth of about 1 cm suboxic conditions [20] occur in both sites (Figure 2). Higher oxygen

concentration at the sediment-water interface and sharper gradient are displayed at S2 station. Nitrate pore water profiles (Figure 3 and 4, plot A) show decreasing values up to about zero in the first centimetre of the sediment after an early increase near the sediment-water interface in both stations and seasons. The peaks are weakly higher in early fall. These nitrate concentration trends indicate that already at 1 cm of depth the suboxic conditions prevails for the degradation of the organic matter. At higher depth, up to about 20 cm, some weak and irregular nitrate increases take place; they could be due to nitrification or irrigation processes related to biological activity. Instead initial nitrate peaks are attributed to nitrification processes that origin for the ammonium migration towards the sediment-water interface where it oxidizes for the oxic conditions present near the surface. Down to the nitrate peak a denitrification process take place by bacteria that use the nitrate as final electron acceptor and that produce a nitrogen (N2 ) loss in the system.

Furthermore, immediately below the nitrate peaks, suboxic diagenesis takes place as indicated by the pore water concentration profiles of dissolved manganese and iron (Figure 3 and 4, plots B and C). This agrees with the well-established diagenetic reaction depth sequence that are governed by the preferential use of Mn and Fe oxide phases as electron acceptors after the nitrates because they provide the highest amount of free energy for the bacterially mediated oxidation of organic matter [13].

Dissolved manganese profiles exhibit, in both stations, near surface peaks (between 0.5 cm and 1 cm of depth), just below the oxygen penetration depth and nitrate peak.

A secondary peak in Mn pore water profile is observed in S2 site at 13 cm in early fall.

463

Marine Geology

Figure 2: Microprofiles of dissolved oxygen measured in the inner (S1) and outer (S2) stations of the Gulf of Manfredonia in late winter 2003.

The dissolved Fe shows, in all profiles, a sharp increase just below the depth range in which Mn concentrations reach maximum values. At higher depth dissolved iron concentrations decrease rapidly down to 10â&#x20AC;&#x201C;20cm and then increase in both fall cores and in the S2 winter core. The increase of dissolved Fe below the first peak suggests a complexation with dissolved refractory organic matter that prevents the precipitation of dissolved Fe with sulphur to form FeS, thermodynamically preferred, or with carbonate, as siderite. The peaks of dissolved Mn and Fe near the interface are attributed to the dissimilated reduction of manganese and iron oxy-hydroxides by bacteria activity in anaerobic condition: these oxide phases are used as terminal electron acceptors by bacteria for the organic carbon oxidation.

About the sulphate pore water profiles (Figure 3 and 4, plot D), the trends are

464

rather constant displaying a weak decrease below 40 cm in S1 station (Figure 3, plot D) and below 20 cm in S2 station (Figure 4, plot D). These trends indicate that a slight organic matter mineralization by sulphate reduction takes place down these depths without however reaching suboxic sulphidic conditions.

The organic matter decomposition using, in order, dissolved oxygen, nitrates, Mn and Fe oxy-hydroxides and sulphates as electron acceptors, results in the release of ammonia and TCO2 into the pore waters, other than the phosphate which concentrations are affected also by other complex processes. Pore water data (Figure 3 and 4) reflect this process, as they show NH+ 4 and TCO2 concentration increases below 20 cm for S1 station and below 10 cm for S2 station. From the sediment-water interface to 10-20 cm of depth, the distribution of both organic matter degrada-

Marine research at CNR

NO3 (uM)

Fe (uM)

Mn (uM)

SO4 (mM)

0

0

0

20

20

20

40

40

40

60

60

60

80

80

80

80

100

100

100

100

0

Depth (cm)

20 40 60

A 0

1

2

0

3

NH4 (uM)

Depth (cm)

C

B

120

120

50

120 50

TCO2 (uM)

100

150

0

PO4 (uM)

0

0

0

20

20

20

20

40

40

40

40

60

60

60

60

80

80

80

80

100

100

100

F

E

120

120 0

300

600

900

G

120

0

5000

10000

15000

0

0,5

1

1,5

50

100

Si(OH)4 (uM)

0

100

D

120

0

100

H

120 2

0

50

100

150

200

2+ Figure 3: Pore water profiles of electron final acceptors (NOâ&#x2C6;&#x2019; , Fe2+ and SO2â&#x2C6;&#x2019; 3 , Mn 4 ) + and main organic matter degradation products (NH4 , TCO2 and phosphate) and silica in cores collected in Gulf of Manfredonia in S1 station in early fall 2002 (yellow) and late winter 2003 (red). Depth intervals were 0.5 cm for the first four layers, 1 cm for the next 4 layers, 3 and 5 cm for the subsequent data; data are plotted at the mid-point of the sampling layer. Data at â&#x2C6;&#x2019;1 cm indicate bottom water values.

tion products are scattered suggesting irrigation and bio-mixing processes. In this case macrofauna activity deepens, channelizes and increases the solute fluxes across the sediment-water interface. At greater depths the constant increase of ammonium and TCO2 indicates that only diffusive processes take place. The NH+ 4 released to the pore waters by suboxic diagenesis diffuses upwards in the sediment column, where part of it is oxidized in the first millimetres before reaching the sediment-water interface giving rise to the nitrate peaks.

On the whole the organic matter degrada-

tion product concentrations (range values between 2.3 and 11 mM for TCO2 and 0.004 and 0.8 mM for ammonium) and their distributions in the pore waters highlight high reactive organic matter inputs into the sediments without a strong difference between the two stations. The higher degradation product concentrations in early fall in both stations (Figure 3 and 4, plots E and F) suggest stronger degradation processes driven by the higher temperature; this is inferred both from the high depths of the higher concentrations and also from satellite data that exclude important fresh

465

Marine Geology

Depth (cm)

NO3 (uM)

Mn (uM)

SO4 (mM)

0

0

0

0

20

20

20

20

40

40

40

40

60

60

60

60

80

80

80

80

100

100

100

100

A

B

120

120 0

1

2

3

0

NH4 (uM)

Depth (cm)

Fe (uM)

50

D

C 120

120

0

100

50

TCO2 (uM)

100

150

0

PO4 (uM)

0

0

0

0

20

20

20

40

40

40

40

60

60

60

60

80

80

80

80

100

100

100

100

E

F

120 0

300

600

900

0

5000

10000

120 15000 0

100

Si(OH)4 (uM)

20

120

50

G

H 120

0,5

1

1,5

2

0

50

100

150

200

2+ Figure 4: Pore water profiles of electron final acceptors (NO− , Fe2+ and SO2− 3 , Mn 4 ), + main organic matter degradation products (NH4 , TCO2 and phosphate) and silica in cores collected in Gulf of Manfredonia in S2 station in early fall 2002 (yellow) and late winter 2003 (red). Depth intervals were 0.5 cm for the first four layers, 1 cm for the next 4 layers, 3 and 5 cm for the subsequent data; data are plotted at the mid-point of the sampling layer. Data at −1 cm indicate bottom water values.

marine organic matter inputs.

The phosphate concentrations in pore waters in all cores have similar trends (Figure 3 and 4, plot G): they show a rapid and irregular increase near the surface (about 3 cm depth) and a second stronger peak at about 17-20 cm, at higher depth the profiles approach the analytical detection limit. The higher values are shown in S1 cores. These variable trends are due to the complexity of the phosphorous cycle: the low concentrations near the sediment-water interface are due to the co-precipitation of the phosphate with the

466

Fe oxy-hydroxides in the oxic environment presents near the sediment-water interface; the phosphate concentration increases from the sub-surface to about 20 cm depth are the result of release by organic matter decomposition and Fe-oxy-hydroxide dissolution in suboxic environment; at higher depths the phosphate decreases result from the authigenic mineral phosphate precipitation (apatite or fluoroapatite).

The silica concentration profiles present distinctive and always similar patterns since they reflect a different process respect to the degradation of organic matter:

Marine research at CNR

2500

a)

2000 1500 1000 500 0 -500

S1 Early fall S1 Late winter

-1000 -1500 -2000

O2 x100 2500 2000

NO3

Mn x10

Fe x10

NH4

TCO2 x10 PO4 x10

Si(OH)4

b)

1500 1000 500 0 -500

S2 Early fall S2 Late winter

-1000 -1500 -2000 O2 x100

NO3

Mn x10

Fe x10

NH4

TCO2 x10

PO4 x10

Si(OH)4

Figure 5: Dissolved benthic fluxes measured by benthic chamber in Gulf of Manfredonia in S1 (a) and S2 (b) stations in early fall 2002 (yellow) and late winter 2003 (red).

3− 2− 2+ Units are µmol·m−2 · d−1 for NO− , Fe2+ , NH+ 3 , Mn 4 , PO4 , Si(OH)4 and SO4 and −2 −1 m·mol·m · d for O2 and TCO2 . mainly they are the consequence of the dissolution of the diatom silicatic exoskeletons. The silica profiles in both stations and seasons show a sharp increase below the sediment-water interface up to values that remain constant with depth (Figure 3 and 4, plot H). This is due to the diatom skeleton dissolution that produces dissolved silica up to the reaching of the equilibrium between the solid and liquid phase at depth.

seasons (Figure 5a and Figure 5b).

In the case of oxygen a general similar uptake (negative fluxes) occurred in both stations with a slight increase in early fall.

These patterns confirm the aerobic organic matter degradation near the sediment-water interface with the higher values in early fall due to the higher temperatures that promote the bacterial activity.

Nitrate fluxes were very low and directed into the sediment in S1 site and higher and out of the sediment in S2, greatest vari4.2 Chamber data ability is evident in S2. This complex behaviour is connected with the nitrification Benthic chamber fluxes were measured for and denitrification processes that take place oxygen, nitrate, Mn, Fe ammonium, TCO2 , near the interface and inside the benthic phosphate and silicate, in both sites and 467

Marine Geology

25,00 20,00 15,00 10,00 5,00 0,00 -5,00 -10,00 Manfredonia Gulf

-15,00

Po River Prodelta Emilia Romagna offshore

-20,00 O2

NO3 x 10

Alk

TCO2

NH4

PO4 x 100

Si(OH)4

Figure 6: Comparison of benthic fluxes measured in the Gulf of Manfredonia and in the Northern Adriatic Sea [19]. Fluxes are in m·mol·m−2 · d−1 . chamber: in case of low fluxes, as our case, nitrification and denitrification mask the fluxes due to diffusion and irrigation. Dissolved manganese and iron show positive fluxes with stronger seasonality in S1 site.

These fluxes are produced by the dissolution of Mn and Fe oxy-hydroxides at the change from oxic to suboxic conditions.

Organic matter degradation products (Ammonium and TCO2 ) fluxes were always directed out of the sediment, with higher fluxes occurring in early fall, this is a further confirmation of the higher organic matter degradation in the warmer season.

Furthermore, in the case of the ammonium, calculated fluxes are always higher than the measured ones. This difference can be explained with a nitrification processes near

468

the oxic sediment-water interface (as supported by the strong nitrate peaks near the interface) that remove part of the ammonia produced by the organic matter degradation.

As regards the phosphate, positive fluxes were always measured, they are mainly due to the anoxic Fe-oxy-hydroxide dissolution near the sediment-water interface that releases the phosphate co-precipitated with Fe-oxy-hydroxides in oxic environment.

Also the silica shows positive fluxes generated by the dissolution of the silicatic exoskeleton of diatoms.

Other major findings can be deduced by the comparison between the measured and the calculated fluxes (Table 1). By comparing the measured and calculated TCO2

Marine research at CNR

Stations Benthic fluxes (mmol/m2 d)

S1 Early fall

TCO2 Measured by chamber Calculated diffusive Constant Rz Exponential Rz Diffusive/Measured %

17.95±2.05 22.62 126

Ammonium Measured by chamber Calculated diffusive Constant Rz Exponential Rz Diffusive/Measured % Phosphate Measured by chamber Calculated diffusive Constant Rz Exponential Rz Diffusive/Measured % Silica Measured by chamber Calculated diffusive Constant Rz Exponential Rz Diffusive/Measured %

S1 Late winter

S2 Early fall

S2 Late winter

4.84±1.49

46.94±8.58

4.65±0.56

7.67 9.76 202

15.04 16.77 36

8.22 8.08 174

0.19

0.15

1.05

0.12

0.71

2.16

2.45

0.56

374

1440

233

467

0.01

0.001

0.02

0.002

0.005

0.003

0.01

-0.003

50

300

50

-150

1.02±0.05

1.01±0.08

2.47±0.37

1.07±0.24

0.66 1.06 104

0.42 0.81 80

0.51 0.99 40

1.24 1.64 153

Table 1: Comparison of measured and calculated benthic fluxes in the S1 and S2 stations in early fall 2002 and late winter 2003. fluxes it raises that diffusive fluxes prevail in both stations and seasons (diffusive fluxes higher or equal to measured fluxes) except for the outer stations (S2) in fall, where the lower calculated flux respect to the measured one suggest the presence of irrigation processes.

Also in the case of the silica and phosphate the irrigation clearly prevails in the early fall outer station (lower calculated flux respect to the measured ones). Finally, measured benthic fluxes in the Gulf of Manfredonia and in the northern Adriatic Sea have been compared (Figure 6) to establish the role of the sediments of the gulf in the chemistry of the western Adriatic Sea waters. From this comparison it results that the benthic fluxes of nutrients and of oxygen in the Gulf of Manfredonia are lower than those of the northern Adriatic Sea and this is attributed to the higher inputs of fresh marine organic matter, fresh and old

continental organic matter and inorganic solid inputs to the sediments in the northern Adriatic Sea connected to the higher dissolved and solid load of the Po River.

In any case the results indicate that also in the Gulf of Manfredonia the fluxes of nutrients to the water column and the oxygen consumption are able to affect the water column chemistry contributing to the eutrophication and dystrophic processes of the southern Adriatic Sea.

4.3

Solid phase

In the two sites surface sediments have a porosity of 0.75-0.8 with decreasing values up to 0.60-0.65 at 25-30 cm depth (Figure 7, plot A and D). The excess activity of 210 Pb, decreasing exponentially below the interface in both stations (Figure 8, plots A and D), may be used to estimate the sediment accumulation rates. Applying 469

Marine Geology

late winter early fall

late winter early fall

Figure 7: Porosity (A and D), C-org (B and E) and C-org vs. N total (C and F) from cores collected in Gulf of Manfredonia in early fall 2002 (red triangles) and late winter 2003 (yellow squares) in S1 and S2 stations.

Constant Flux-Constant Supply (CF-CS) model to the 210 Pb excess profiles [21], sedimentation rates (w) from each stations were estimated to be 0.52 cm·y−1 in station S1 and 0.49 cm cm·y−1 in station S2.

These values are similar to those calculated by [19] and by [22] in the northern Adriatic Sea. In order to check sedimentation rates the vertical profile of 137 Cs (Figure 8, plots B and E) is used. It is well know that a significant input of this radionuclide in the environment occurred in the 1963 related to atmospheric weapon tests. Hence max-

470

ima concentrations in caesium vertical profiles correspond to sediments buried in this time [23], supporting the dating calculated by 210 Pb excess.

Thorium activities were also measured as useful tracer of sediment mixing (Figure 8, plots C and F). The mixing of the upper 10 cm of sediment caused by macrofauna is referred to as bioturbation; it can alter the physical and chemical properties of sediment affecting both particle and solute transport [24, 2, 25]. In order to calculate bioturbation coefficients (Db) the re-

Marine research at CNR

Figure 8: Profiles of solid phase parameters in the inner station (S1: A, B and C) and outer station (S2: D, E and F) of the Gulf of Manfredonia. The solid line shows the exponential function fit of excess 210 Pb vs. mass depth. lation calculated by [26] was applied to 234 Th exponential profile (S2 station) and Db was estimated to be 0.2 cm2 · y−1 . This value falls at the low end of the compilation of [27] relating Db to w for similar environments; this could be explained by the episodic presence of low oxygen bottom water. The profiles of 234 Th in station S1 is constant from sediment-water interface down to 5 cm, suggesting a bioturbated layer too.

Organic carbon concentrations in all cores (Figure 7, plots B and E) rapidly decrease from the sediment-water interface to 5-7 cm, at greater depths C-org reaches low constant values and exhibits a regular trend interrupted only by relatively high values at about 20 cm in both sites. Total N concentrations also show a down-core decreasing trend. These trends indicate the degradation in the first centimetres of the more la-

bile fraction of the total organic matter content of the sediments.

Marine continental margin sediments are a mixture of terrestrial and marine components. As regards the C-org/N, the Redfield ratio of C:N:P in marine sediments, 106:16:1 (C/N=7), is considered representative of settled marine phytoplankton, whereas higher values may reflect inputs of terrestrial derived particulate organic carbon or the breakdown marine organic matter. For these reasons the C-org/N ratio might be used to identify the source of the organic matter undergoing diagenesis [13].

The C-org/N ratios range from 5±2 to 7±2 in late winter and in early fall cruises respectively (Figure 7, plots C and F), with higher values shown by surface sediments collected in early fall that are similar those calculated in North Adriatic Sea by [22].

Therefore, the origin of organic matter in

471

Marine Geology

the gulf is prevalently marine.

5

Conclusions

Pore water concentration profiles of final electron acceptors for organic carbon oxidation indicate that in the Gulf of Manfredonia early diagenesis progresses trough oxygen consumption, denitrification, manganese and iron oxy-hydroxide reduction and slight sulphate reduction while methanogenesis does not takes place.

The high concentrations of the organic matter degradation products in the gulf support high inputs of reactive organic matter without a strong difference among the two areas. In the warmer season the degradation processes are more intense in both stations for the higher temperature that enhances microbial activity.

Pore water concentration profiles also indicate the presence of other processes near the sediment-water interface such as the coupled phosphorous and iron cycling, by repeated iron oxy-hydroxide dissolution and precipitation in anoxic and oxic environment, the nitrification and denitrification processes, the irrigation processes and, at greater depths, the phosphorous, manganese and iron precipitation in authigenic minerals.

Solid phase concentration profiles highlight inputs of prevalently marine organic matter (low C-org/N ratio) in both stations and high sedimentation rate in the gulf.

The measured benthic fluxes disclosed an

important role of the sediments in the Gulf of Manfredonia as supplier of nutrients to the water column, in particular for the TCO2 , and as sinker of dissolved oxygen, contributing in this way to increase the eutrophication and anoxic processes and to affect the carbon dioxide dynamics in the water column.

From the comparison between the measured and calculated benthic fluxes it raised the confirmation of the irrigation processes in the outer station in the warmer season and the nitrification processes near the sediment-water interface.

Finally, by comparing measured benthic fluxes of the Gulf of Manfredonia and of the northern Adriatic Sea it follows that lower organic matter degradation processes in Manfredonia Gulf take place respect to the North Adriatic Sea mainly for the lower organic matter and inorganic solid inputs to the sediments in the southern Adriatic Sea, however these fluxes are able to affect the water column chemistry.

6

Acknowledgements

We wish to acknowledge Giovanni Casolino for the invaluable assistance to carry out the on-board operations and Piero Trentini and R. Belastock for the help respectively in metal and TCO2 analyses. Financial support for this research was provided by the Consiglio Nazionale delle Ricerche trough the PITAGEM project.

References [1] R.A. Berner. Early Diagenesis: A Theoretical Approach. page 241, 1980.

[2] R.C. Aller. Quantifying solute distribution in the bioturbated zone of marine sed-

472

Marine research at CNR

iments by defining an average microenvironment. Geochimica et Cosmochimica Acta, 44:955–1965, 1980.

[3] D.E. Hammond, C. Fuller, D. Harmon, B. Hartman, M. Korosec, L. Miller, R. Rea, W. Berelson, and S. Hager. Benthic fluxes in San Francisco Bay. Hydrobiologia, 129:69–90, 1985.

[4] J.E. Cloern. Phytoplankton bloom dynamics in coastal ecosystems: a review with some general lessons from sustained investigation of San Francisco Bay, California.

Reviews of Geophysics, 34:127–168, 1996.

[5] A. Specchiulli, F. Spagnoli, F. Dicembrini, and A. Conversi. The PITAGEM Project: an example of integrated, continuous and punctual coastal monitoring in the Gulf of Manfredonia, adjacent to the Gargano National Park. Bollettino di geofisica teorica ed applicata, 44:69–77, 2003.

[6] A. Artegiani, D. Bregant, E. Paschini, N. Pinardi, F. Raichic, and A. Russo. The Adriatic Sea general circulation. Part II: Baroclinic Circulation Structure. Journal of Physical Oceanography, 27:1515–1532, 1997.

[7] D. Vilicic, Z. Vucak, A. Skrivanic, and Z. Grzetic. Phytoplankton blooms in oligotrophic open South Adriatic waters. Marine Chemistry, 28:89–107, 1989.

[8] F. Spagnoli, G. Bartholini, E. Dinelli, and P. Giordano. Geochemistry and particles size of surface sediments of Gulf of Manfredonia (Southern Adriatic Sea). Estuarine Coastal and Shelf Science, 80:21–30, 2008.

[9] V. Damiani, C.N. Bianchi, O. Ferretti, D. Bedulli, C. Morri, M. Viel, and G. Zurlini.

Risultati di una ricerca ecologica sul sistema marino pugliese. Thalassia Salentina, 18:153–169, 1988.

[10] W. Sigl. Der Golf von Manfredonia (Sudliche Adria). I. Die fazielle Differenzierung der Sedimente. Senckenbergiana marittima, 5:3–49, 1973.

[11] M. Viel, V. Damiani, and M. Setti. Caratteristiche granulometriche e composizione mineralogica dei sedimenti della piattaforma pugliese. pages 127–147, 1986.

[12] A. Magagnoli and M. Mengoli. Carotiere a gravit´a SW-104. (27):1–45, 1995.

[13] P.N. Froelich, G. Klinkhammer, M.L. Bender, N.A. Luedtki, G.R. Heath, D. Cullen, P. Dauphin, D. Hammond, and B. Hartman. Early oxidation of organic matter in pelagic sediments of the eastern equatorial Atlantic: suboxic diagenesis. Geochimica et Cosmochimica Acta, 43:1075–1095, 1979.

[14] P. Giordani and L. Angiolini. Chemical parameters characterizing a NW Adriatic coastal area. Coastal Shelf Science, 17:159–167, 1983.

[15] J.M. Gieskes and W.C. Rogers. Measurements of total carbon dioxide and alkalinity. Journal of Sedimentary Petrology, 43:272–277, 1973.

473

Marine Geology

[16] D.C. McCorkle and S.R. Emerson. The relationship between pore water carbon isotopic composition and bottom water oxygen concentration. Geochimica et Cosmochimica Acta, 52:1169–1178, 1988.

[17] W.M. Berelson, D. Heggie, A. Longmore, T. Kilgore, G. Nicholson, and G. Skyring.

Benthic Nutrient Recycling in Port Phillip Bay, Australia Estuarine. Coastal Shelf Science, 46:917–934, 1998.

[18] D.E. Hammond, J. McManus, W. Berelson, T. Kilgore, and R. Pope. Early diagenesis of organic carbon in the equatorial Pacific: rates and kinetics. Deep Sea Research, 43:1365–1412, 1996.

[19] P. Giordani, D.E. Hammond, W.M. Berelson, R. Poletti, G. Montanari, A. Milandri, M. Frignani, L. Langone, M. Ravaioli, and E. Rabbi. Benthic fluxes and nutrient budgets for sediments in the Northern Adriatic Sea: burial and recycling efficiencies. Science of the Total Environment, 5:251–269, 1992.

[20] R.A. Berner. A new geochemical classification of sedimentary environments. Journal of Sedimentary Petrology, 51:359–365, 1981.

[21] C. Lalou. Sediments and sedimentation processes. pages 384–406, 1982.

[22] D.E. Hammond, P. Giordani, W. Berelson, and R. Poletti. Diagenesis of carbon and nutrients in sediments of the Northern Adriatic Sea. Marine Chemistry, 66:53–79, 1999.

[23] J.K. Cochran. Particle mixing rates in sediments of the eastern equatorial Pacific: evidence 15 from 210 Pb, 239,240 Pu, and 137 Cs distributions at MANOP sites.

Geochimica et Cosmochimica Acta, 49:1195–1210, 1985.

[24] D.R. Schink and N.L.J. Guinasso. Redistribution of dissolved and adsorbed materials in abyssal marine sediments undergoing biological stirring. American Journal of Science, 278:687–702, 1977.

[25] J.Y. Aller and R.C. Aller. Evidence for localized enhancement of biological activity associated with tube and burrow structures in deep-sea sediments at the HEBBLE site, western North Atlantic. Deep-Sea Research, 33:755–790, 1986.

[26] H.P. Pope, D.J. Demaster, C.R. Smith, and H. Seltman. Rapid bioturbation in equatorial Pacific sediment: evidence from excess 234 Th measurements. Deep Sea Research II, 43:1339–1364, 1996.

[27] T.K. Tromp, P. Van Cappellen, and R.M. Key. A global model for the early diagenesis of organic carbon and organic phosphorus in marine sediments. Geochimica et Cosmochimica Acta, 59:1259–1284, 1995.

474

Deep Sea Depositional Systems: Processes, Architecture and Controls F. Gamberi, G. Dalla Valle, M. Marani, M. Rovere, F. Trincardi, F. Foglini Institute of Marine Sciences, CNR, Bologna, Italy fabiano.gamberi@bo.ismar.cnr.it Abstract Due to the needs for exploration and exploitation of resources in the deep marine environment and thanks to the development of new investigation techniques, the study of deep sea depositional systems has flourished in the last years. ISMAR has contributed to this field of research through the acquisition of a large amount of high resolution data (multibeam, sidescan sonar, high resolution profiles, seafloor samplings) and through their integrated interpretation aimed at defining the characteristics of the deep sea environments. In this paper, we present the results of recent researches carried out at ISMAR dealing with the geomorphological analysis of canyons, channel levee complexes, mass transport complexes, depositional lobes and countourite drifts that are the main architectural elements of deep sea depositional systems. Their large scale geometry and the distribution of their components and facies are reconstructed and defined with a process sedimentology approach to determine the character of their genetic sedimentary processes. Studies are also focussed at the definition of the factors that influence the transfer of sediment from the coastal areas to the deep sea. Recent studies have pointed out the relative importance of the characteristics of the source area, of sea level variations, of the initiating mechanisms of sediment gravity flows, of seafloor topography and gradient on the evolution of deep sea depositional systems on a range of margins with largely different geological setting. The results of the studies have applications in the definition of analogues for hydrocarbon reservoir characterization and for the study of geomarine hazards.

1

Introduction

In the recent years, the comprehension of deep-sea depositional systems has advanced significantly. Much of the new ideas come from studies of outcrops, of 3-D seismic volumes with the development of the concepts of the seismic geomorphology interpretation, of drilling data and of modelling of the processes that control sediment transport and deposition.

Also, the study of modern systems has much contributed to the revitalization of

the researches dealing with deep-sea depositional systems. Although ISMAR is actively working in many of the above fields of research, its main contributions to the understanding of deep-sea depositional systems come from the study of the modern environment, in particular in the marine areas surrounding the Italian peninsula. In this paper, we describe the methodology used to acquire and interpret data in the deep-sea environment. Then we focus on the studies aimed at the characterization of sedimentary processes and at the definition

Marine Geology

Figure 1: Shaded relief map from multibeam bathymetric data showing the extent of acquired data in the Tyrrhenian Sea [1] and in the in the southern Adriatic margin [2]. of the geomorphic elements that represent the basic components of the deep-sea depositional environment. Finally, we describe how the spatial and temporal arrangement of the basic geomorphic elements can be used to infer variations of controlling factors.

2

Imaging and interpretation of the deep-sea deposit

In the recent year, ISMAR has acquired a large amount of multibeam data in the Italian Sea [1, 3] (Figure 1). The multibeam acquisition technology furnishes bathymetry data and gives images of the seafloor topography with a detail comparable to the observations possible on land (Figure 1). The multibeam technology also 476

offers the possibility of acquiring seafloor reflectivity data that are used to qualitatively map sediment grain size and texture at the seafloor (Figure 2). Deep-tow sidescan sonars furnish data that are also used to map sediment distribution at the seafloor but they give a much higher resolution than multibeam reflectivity often allowing the interpreter to map features at a metric scale (Figure 3). Subbottom profiles gives images of the first tenth of ms below the seafloor with a vertical resolution which can be below 1 m and are used to reconstruct the distribution of depositional bodies and erosional surfaces in the near subseafloor section (Figure 3). Finally, seismic data that image subseafloor stratigraphy down to kms below the seafloor, are used to extend the scale of temporal observation and to analyze the evolution of deep sea depositional environments at a basin scale and for temporal interval that can

Marine research at CNR

Figure 2: Mulibeam seafloor reflectivity draped over bathymetric data. In the Gioia basin a meandering channel is flanked by a straight one. The bathymetric data show the finescale detail of meandering channel evolution in a way similar to rivers. Crevasse splays are located on the outer side of bends along the channel courses. span millions of years (Figure 4). Seafloor samples are used to groundtruth the observations made with all the above geophysical data and through bio-stratigraphic analysis to date particular events.

3

Sedimentary processes in the deep-sea depositional

A variety of gravity-driven flows is responsible for the transfer of sediment from the shallow water, coastal and shelf areas though the continental slope to the basin plain. Debris/rock fall consists in the free downslope tumbling of hard sediment or fragmented bedrocks that fail suddenly from steep slopes. They usually re477

Marine Geology

Figure 3: Deep towed sidescan sonar data on the Stromboli slope valley. The resolution of the image allows to define a series of small scale intra-channel erosional and depositional features. The corresponding subbottom profile gives an high resolution image of the near subseafloor stratigraphy. In particular, it enables to define a fine grained area with parallel reflectors and good penetration to west (left) of the coarse grained channel floor characterized by low penetration. sult in local, small areal extent deposits.

Debris/rock avalanches are similar to debris/rock falls but are larger flows in which clasts collide and share their momentum. Submarine debris/rock avalanches are mostly observed in steep volcanic slopes, such as that studied in the Stromboli Island submarine flank that was at the origin of a tsunami in 2002 [4]. Slumps and slides are movement of coherent sediment masses above discrete basal shear surfaces.

Many examples of slumps and slides are

478

currently studied at ISMAR; they are not described here since they are the focus of a dedicated contribution in this volume [5].

Debris flows have plastic rehology and are poorly sorted mixtures in which clasts float in a fine-grained matrix with finite shear strength. Debris flow result in chaotic or poorly organised depositional bodies that are found along many tracts of deep sea depositional systems recently being recognized as a component of frontal splay basin plain deposition [6, 7]. Turbidity currents

Marine research at CNR

Figure 4: Seismic data showing the Milazzo and Niceto channels and their levee wedge with a train of bedforms. Some portion of the levee wedge area affected by incipient instability. are flows of mixed sediment that are maintained in suspension by fluid turbulence and can have strong erosional character.

Deposition from turbidity currents can be associated with traction at the seafloor or to rapid fall out from suspension depending on seafloor gradient and on the degree of flow confinement (Figure 5). Sediment gravity flows are often part of a continuum where one flow can transform into another.

The investigations of the causes that drive the various type of flow transformation in the deep-sea environment is part of the ongoing research at ISMAR [8, 4, 9].

4

Deep-sea depositional systems: basic geomorphic elements

Depending on the character of sediment gravity flows, six main key geomorphic elements are recognized in the deep-sea depositional environment: canyons, leveed channels, channel overbank levees, frontal splays, contourite mounds and drifts, mass transport complexes. The recent researches carried out at ISMAR on mass-transport complexes are not included in the following description since they will be treated elsewhere in this volume [5].

Canyons. Submarine canyons, the proximal trunk of deep sea depositional systems, are amongst the largest features on the earth and are mainly sites of sediment 479

Marine Geology

Figure 5: Reconstruction of longitudinal and transversal variations in channel floor features and grain-size in the Stromboli slope valley. The arrangement of erosional and traction fuatures and featureless seafloor portions is due to different flow behaviour in response to variation in confinement degree and seafloor gradient. transport and erosion. Usually they are located in the continental slope, that connects the shelf with the basin plain, where the steep gradient allow flows to maintain an erosional behaviour. In the continental slope, the locations of canyons is often controlled by river distribution on land [9], simple v-shaped or flat bottomed canyons are typical river-connected features. However, canyons can also develop away from river mouths in areas affected by seafloor instability resulting in highly degraded fea-

480

tures forming complex embayments often indenting the shelf-break [10]. Recent studies carried out at the Institute have shown that canyons can also form away from the continental slope where steep escarpments are present along topographically complex margins [11, 12]. In this cases, canyons can also propagate upward in relatively flat area where headward erosion induced by downslope eroding sediment gravity flows causes knickpoint retrogradation [12] (Figure 6).

Marine research at CNR

Figure 6: Gradient variations (solid line) and depth of incision (crosses) along the Stromboli slope valley. The steeper sectors (III and IV) correspond with tracts with canyon morphology and are controlled by fault (F) crosscutting the slope valley course. They correspond with knickpoihat Leveed channels. At the base of the continental slope or elsewhere where a sudden decrease in seafloor gradient occurs, the sediment gravity flows that are funnelled within canyons lose energy, become depositional and result in the formation of channels. Channel have a large variety of dimension and planform depending on the character, size and grain size of the flows and on seafloor gradient. A recent study of leveed channels in the Calabrian margin has shown that straight and meandering channels run parallel in a slope sector without gradient variation, pointing to a relative major impact of flow character and grain-

size on channel planform [13] (Figure 2).

It also highlights that the gradient threshold for the transition from straight to sinuous channel planform much depends on the grain size of the flows [13]. Erosional scours, inner levees, longitudinal barforms, inner meander bars are all features that develop within the floor of submarine channels (Figure 5). Their distribution is controlled by longitudinal and transversal variations in flow behaviour that are mainly controlled by channel planform, degree of flow confinement, and seafloor gradient (Figure 5). Researches in the Adriatic sea have shown that submarine channels can 481

Marine Geology

also be affected, or in the extreme case fed mainly by geostrophic currents (Figure 7) [10].

Channel overbank levees. Overbank levees are wedges of sediment constructed alongside channels through the deposition of the finer, upper part of sediment-gravity flows that overspill the channel banks (Figure 4). Flows are mostly unconfined and result in the deposition of thin bedded turbidites with thickness wedging away from the levee crest (Figure 4). Sediment waves with trend perpendicular to the channel are frequent on levee flanks (Figure 4). Recent studies carried out at ISMAR have shown that however, also erosional processes are frequently associated with overbanking flows. Particularly in the outside of channel bends, chutes, scours and overbank channels can develop and can give rise downslope to crevasse splay deposits (Figure 2, 8) [13, 3]. The size and internal character of the crevasse splay deposits is mainly dependent on the size of the breach in the levee and thus correlated with the grain size of the flows that exit the cahnnel [13]. Beside constructional processes levee are also affected by instability processes of largely variable dimensions, from small local failure (Figure 4) to margin-wide large collapse often facilitated by lithologic boundaries developed during period of channel inactivity [7].

Frontal splays. At the mouth of channels, sediment-gravity flows lose confinement and become strongly depositional resulting in the formation of frontal splay deposits. Frontal splay deposits stack to

482

originate lobate features that are the main constituent of the basin plain environment.

They can be featureless lobes or be characterized by a distributary channel pattern [9]. They are mainly made up by turbidites but also debris flow lobe are present [9]. Beside basin plain environment frontal splay deposit can be also developed in the slope where low efficiency hyperpycnal derived turbidite currents are unable to cross the whole length of the slope (Figure 9) [13].

Contourite mounds and drifts. Deep contour-parallel currents develop where differences in temperature and salinity characterize different water masses. Where currents intersect the basin margin or interact with bottom-topography they may erode rework transport and deposit sediment. Sediment mounds and drifts and migrating very large bedforms (Figure 10) are the most frequently observed geometry of contourite deposits [14, 15]. A a particular case of the interaction of margin morphology and water masses has been recently evidenced in the southern Adriatic margin where currents developed in the northern Adriatic shelf are forced downslope and originate large bedform fields and contourite mounds and drifts (Figure 7) [10, 16]. In the same area the importance of seafloor topography in determining the location of counturite deposits has also been shown [14]. Data interpretation in the Sicily channel an Adriatic sea have in addition shown that countourite depositional bodies are frequently affected by instability and failure [15].

Marine research at CNR

Figure 7: A submarine channel is fed mainly by cascading current in the southern Adriatic sea. It develops above a canyon that formed through repeted slope and shelf instability and resulted inan embayment along the margin. The temporal evolution of the depositional system are linked with sea level variations and related changes in the character of the sediment feeding mechanism.

483

Marine Geology

Figure 8: An erosional chute feeds a crevasse splay deposit in the Caprera deep sea fan.

5

Deep sea depositional systems in space and in time

The spatial arrangement of geomorphic elements in the deep-sea environment is mainly the result of the geology of the source area, of the initiating mechanisms of sediment gravity flows, of geostrophic currents and of seafloor gradient and topography. The Italian peninsula and its continental margins comprise different geodynamic provinces and thus the depositional systems of the Italian deep water regions are characterized by source areas with highly variable dimensions, type and volume of river sediment transport, variable rate of relative vertical movements and different degree of tectonic activity and seismicity.

The large variability of the geodynamic setting of the Italian region has also a di484

rect impact on the deep sea depositional environment controlling the gradient of the seafloor, the dimensions and shape of the continental slope, of intraslope basins and of basin plain regions and the ultimate base level for sediment-gravity flows. Thus the Italian sea offers an unique setting that has allowed ISMAR to investigate the factors that control the evolution of depositional systems under a large variety of geologic scenarios. A variety of initiating mechanism of sediment gravity flows has been determined to feed sediment to the deep sea areas of the Italian Sea. In the southern Adriatic margin, sediment supply is at present mainly derived from cascading currents (Figure 7) [10, 16]. In the Sardinian margin where a wide shelf trap much of the sediment, along shore currents deflected by across shelf ones feed the Caprera channel [3]. Direct discharge of river hyperpycnal flows has been inter-

Marine research at CNR

Figure 9: Geomorphic element reconstruction in the Gioia basin as a function of source area character. preted to feed the Gioia basin where the lack of the continental shelf allow a direct connection between rivers and submarine canyons. However, a basin-wide channel/canyon system is developed only in connection with large rivers whereas small channels that die out in the slope face small rivers (Figure 9) [13]. Shelf width has also been determined to control the shape and dimensions of intraslope basin fans shown in the Sardinian margin where a large fine-grained fan is developed exclusively in the areas where the shelf is wide and can trap coarse-grained sediment and contrasts with small fans made up of coarser grained material where the shelf is narrower [11]. In the Sicilian margin, differential rates of vertical uplift have been determined to control the location of submarine leveed channels and more importantly to influence their destruction through mass-wasting processes that par-

ticularly affect the portions of channellevee complexes facing areas with high uplift rates [17]. Slope gradient has been recognized as the main factor that control the location of canyon and channel tracts along single basin-wide sedimentary pathways in topographically complex slopes.

Canyon and channel tracts alternate along the length of the Stromboli and Sardinia slope valley in response to gradient variations due to the presence of active faults (Figure 6) [11, 12]. The importance of gradient has also been recognized in relatively flat basin plain areas where slight increase of seafloor dip cause rechannelization of flows downslope from area where they are unconfined (Figure 11) [9]. Within submarine channel, gradient reduction has been recognized as driving deposition with development a multi-thalweg floor and bars and inner levees (Figure 5). The temporal arrangement is a function of all the above 485

Marine Geology

factors that however, to a large extent are controlled by cyclic nature of sea level variation. The Gioia mesisma system show both the importance of sea level in controlling the deactivation of submarine channels where the shelf is wide [13] In the Gioia Basin, only the channels facing the areas lacking the continental shelf are active during the present-day high stand. In the Sardinia Caprera fan lower efficiency flows during the present-high stand result in the retrogradation of the fan that also pass to a depositional phase in regions previously affected by erosion [3]. Sea level variations have also demonstrated to be instrumental in controlling the activation of geostrophic and cascading currents thus impacting on the temporal evolution of continental margin as shown in the southern Adriatic Sea (Figure 7) [10, 16].

486

6

Concluding remarks

ISMAR is actively involved in the studies of deep-sea depositional systems. Large data set have been acquired in many of the submarine areas of the Italian seas. Important results of the recent researches carried out at ISMAR are meaningful steps toward the understanding of the sedimentary processes that occur in the deep sea environment and how they contribute in the shaping of the deep-water landscape. They also show how the deep marine environment represent an archive of a variety of processes that occur in the surrounding areas and of climatic oscillations. Furthermore the results of the researches represent key issue when considering the exploitation of deep sea resources, the action needed for the preservation of the environment and the definition of the geological hazard.

Marine research at CNR

Figure 10: Shaded relief (left) and deep towed sidescan sonar of a sediment wave field formed due to downslope cascading currents in the southern Adriatic Sea. In the same area, chirp profiles (below) show contourite drifts with variable shape often controlled by preexisting seafloor topography are also formed.

487

Marine Geology

Figure 11: Reconstruction of basin plain deposition at about 300m depth in the Tyrrhenian Sea. The influence of the geological character of the different margins of the Tyrrhenian Sea is reflected in the architecture of the two basin plain.

References [1] M. Marani, F. Gamberi, and E. Bonatti. From seafloor to deep mantle: architecture of the Tyrrhenian backarc basin. Memorie descrittive della carta geologica d’Italia, 44, 44:1–194, 2004.

[2] F. Foglini, E. Campiani, A. Cattaneo, Correggiari A., et al. The bathymetry of the Adriatic Sea. this volume, 2011.

[3] G. Dalla Valle and F. Gamberi. Erosional sculpting of the Caprera confined deep-sea fan as a result of distal basin-spilling processes (eastern sardinian margin, Tyrrhenian Sea). Marine Geology, 268:55–66, 2010.

[4] M.P. Marani, F. Gamberi, M. Rosi, A. Bertagnini, and A. Di Roberto. Subaqueous density flow processes and deposits of an island volcano landslide (Stromboli Island, Italy). Sedimentology, 56:1488–1504, 2009.

[5] M. Rovere, G. Dalla Valle, F. Foglini, F. Gamberi, M. Marani, and F. Trincardi.

Submarine landslides: case studies in the Mediterranean Sea. This volume, 2011.

[6] F. Gamberi, M. Marani, V. Landuzzi, A. Magagnoli, et al. Sedimentologic and volcanologic investigation of the deep Tyrrhenian Sea; preliminary results of cruise VST02. Annals of Geophysics, 49:767–781, 2006.

[7] F. Gamberi, V. Marani, M. Landuzzi, A. Magagnoli, et al. Sedimentologic and volcanologic investigation of the dep Tyrrhenian Sea: preliminary results of cruise VST02. Annals of Geophysics, 49(2/3):767–781, 2006.

488

Marine research at CNR

[8] F. Gamberi. Volcanic facies associations in a modern volcaniclastic apron (Lipari and Vulcano offshore, Aeolian island arc). Bulletin of Volcanology, 63:264–273, 2001.

[9] F. Gamberi and M. Marani. Regional geology control on the style of basin plain depositional systems in the Thrrenian Sea. SEPM Special Puplication, External controls on deep water depositional systems, 92:221–232, 2009.

[10] F. Trincardi, G. Verdicchio, and S. Miserocchi. Seafloor evidence for the interaction between cascading and along-slope bottom water masses. Journal of Geophysical Research, (112):1450–1467, 2007.

[11] F. Gamberi and M. Marani. Deep-sea depositional systems of the Tyrrhenian Basin.

Memorie descrittive della carta geologica d’italia, 44:127–146, 2004.

[12] F. Gamberi and M. Marani. Downstream evolution of the Stromboli slope valley (southeastern Tyrrhenian Sea). Marine Geology, 243(1-4):180–199, 2007.

[13] F. Gamberi and M. Marani. Controls on Holocene deep-water sedimentation in the northernGioia Basin, Tyrrhenian Sea. Sedimentology, 55:1889–1903, 2008.

[14] G. Verdicchio and F. Trincardi. Short-distance variability in slope bed-forms along the Southwestern Adriatic Margin (Central Mediterranean). Marine Geology, 234:271–292, 2006.

[15] G. Verdicchio and F. Trincardi. Mediterranean shelf-edge muddy contourites: examples from the Gela and South Adriatic basins. Geo-marine Letters, 28:137–151, 2008.

[16] F. Trincardi, F. Foglini, G. Verdicchio, A. Asioli, et al. The impact of cascading currents on the Bari Canyon System, SW-Adriatic Margin (Central Mediterranean).

Marine Geology, (246):208–230, 2007.

[17] F. Gamberi and M. Marani. Hinterland geology and continental margin growth: the case of the Gioia Basin (southeastern Tyrrhenian Sea). Special publication geological society of London, 262:349–363, 2006.

489

Marine Geology

490

Deep Marine Record from Sentinelle Valley (Sardinia Channel): Integrated Stratigraphy of A Key Site for Western Mediterranean Paleoceanographic and Geological Reconstruction F. Budillon1 , F. Lirer1 , M. Iorio1 , P. Macr`覺2 , L. Sagnotti1 , M. Vallefuoco1 , L. Ferraro1 , S. Innangi1 , G. Di Martino1 , R. Tonielli1 1, Institute for Coastal Marine Environment, CNR, Napoli, Italy 2, National Institute of Geophysics and Volcanology, Roma, Italy francesca.budillon@iamc.cnr.it Abstract Biotic, petrophysical, paleomagnetic proxies combined with 14C AMS data allowed us to produce, for deep marine record of Sentinelle Valley (Sardinia Channel), a detailed integrated stratigraphic time-framework for the last 80 kyr.

Major planktonic foraminiferal changes in quantitative distribution of selected climate sensitive species allowed the identification of 10 eco-biozones and the main climatic global events (Sapropel S1, Younger Dryas, Greenland Isotope Interstadial 1, Greenland Isotope Stadial 2, Heinrich events H1-H6).

An accurate age-depth profile has been proposed for the studied record which spans between 2 and 83 kyr cal. BP. The adopted age model was successively confirmed by comparing the colour reflectance data of the studied record with the astronomically tuned deep marine record ODP-Site 964 from the Ionian Sea. Three turbidite event layers were chronologically constrained within the relative low stand and lowering sea level phases associated to the MIS 4 and MIS 3.

1

Introduction

Since climate excursions recorded in Northern Hemisphere in the Greenland GISP and GRIP ice cores [1] over the last 100 kyr had more or less synchronous effects in the Mediterranean area, many researches have focused on Mediterranean marine cores, with the aim to detect their intensity and the impact on the marine environment. During the last glacial period the Mediterranean region experienced rapid modifications in hydrographic conditions in response to fast climatic excur-

sions, known as Heinrich events (HE) and Dansgaard-Oeschger (D-O) Stadials (cold) and Interstadials (warm) [2, 3]. In particular, [4, 5] and [6] have proved that the millennial scale D-O and HE directly control the winds and precipitation system on the Northern Mediterranean basin. Even during the Holocene the principal climatic events and oscillations of the Northern Hemisphere were clearly traceable in different sectors of the Mediterranean Basin sedimentary records [5, 6, 7, 8, 9, 10, 11, 12, 13].

A detailed outline of the paleoenviron-

Marine Geology

mental changes and their control on marine communities, calibrated by several independent proxies (tephra, sapropel, 14C geochronology), is available for the Mediterranean area (i.e. [14, 15, 16, 17, 18, 19, 9, 10], and reference therein). Several codified eco-bioevents, if clearly detected in marine records, can be used as tie points to chronologically constrain the late Pleistocene-Holocene Mediterranean marine sequences. Nevertheless, even if many reference records are available from deep-sea sites, most of them span a short time interval and lack a high resolution detail of the paleo-environmental and paleo-ecological changes before 40 kyr. Recently, [11, 12] carried out a highresolution study of the tuned ODP-Site 977, located in the Western part of the Alboran Sea, and identified several planktonic foraminiferal eco-bioevents occurred during the marine isotope stages (MIS) 1 to 5.

These eco-bioevents represent the best tool to correlate deep marine records from different Mediterranean sites.

Many recent studies emphasize the challenge when studying deep sea records to establish a reliable chronology even for the deposition of turbidites [20, 21, 15] and underline the utility to support conventional dating methodologies with different constraints. It is widely accepted that one of the main factor controlling and enhancing turbidite deposition along deep sea fan is the fall and lowstand of sealevel, whereas sea-level rise and highstand phases reduce terrigenous supply to deep sea systems [20, 22, 23].

The studied CIESM core C08 is located in the Sardinia Channel in a key position of paleoceanographic and geological significance. In fact, the Sardinia Channel connects the Alboran to the Tyrrhenian Basin and offers a stratigraphic record with 492

the potential to link the eco-stratigraphic and paleoceanographic observations between the Western, the Central and Eastern Mediterranean late Pleistocene Holocene marine records [12, 24, 13, 25, 26]. In fact, a portion of the Modified Atlantic Water (MAW) coming from the Strait of Gibraltar [27], diverges from the part that enters the Eastern Mediterranean and flows through the Sardinia Channel into the Tyrrhenian Sea along the northern Sicilian coast [28], forming a secondary circulation gyre. The circulation system in this sector of the Tyrrhenian Sea is counter-clockwise with the Levantine Intermediate Water (LIW) inflows lapping on the northern Sicilian coast and the outflow occurring along the eastern Sardinia coast ([29] and references therein).

The core site is even in a strategic position to check the efficiency of a submarine canyon in driving density flow to the deep sea environment, even if not directly connected to any emerged sector nor to continental shelf areas [30]. Thus the possibility that such a type of canyon would form a fan can be evaluated, even verifying the significance and the timing of the turbidite deposition.

The aim of this study is to provide a record of integrated stratigraphic data spanning the last 80 kyr, relatively to a deep basin area, based on eco-biozones, 14C-dated ages, Event Stratigraphy, lithostratigraphy and petrophysical properties.

This work represents a short version of the article â&#x20AC;&#x153;Integrated stratigraphic reconstruction for the last 80 kyr in a deep sector of the Sardinia Channel (Western Mediterranean)â&#x20AC;? published on Deep - Sea Research II, v°56, 725-735 doi:10.1016/j.dsr2.2008.07.026, by Budillon et al. (2009, [31]).

Marine research at CNR

Figure 1: A) Location map of CIESM Core C08. Bathymetry from The General Bathymetric Chart of the Oceans (GEBCO, 1997); B) bathymetric detail of core site, close to a channel South-North oriented (green arrow).

2

Geological setting

The core C08 was collected in the Sentinelle Valley of the Sardinia Channel (Figure 1), during the cruise CIESM Sub2 onboard the R/V Urania in December 2005 (38°38.5364’N, 10°21.5576’E - 2370 m below sea level), and it recovered about 5.40 m of hemipelagic mud interlayered with three fine to medium sand turbidite layers of increasing thickness towards the top of the core (Figure 2). In the studied area a 400 km long submerged sector of the Apennine-Maghrebian branch of the Alpine orogen separates the Tyrrhenian (Plio-Pleistocene in age) and the Algero Provenca¸l (Miocene in age) oceanic basins. This sector of the chain was not completely fragmented during the opening of the basins [32]. Due to the relatively minor post-orogenic extension and the good preservation of morpho-structural features, the Sardinia Channel is an important area for the reconstruction of the geodynamic evolution of the Western Mediterranean sector and was recently investigated

through submersible surveys [33, 32]. The triangular shaped valley is bounded by a NE-SW oriented Median Ridge on its north-western side and by the South Cornaglia slope on the south-eastern one. The south- western sector of the Sentinelle Valley receives the sedimentary contribution of the Bizerte Canyon, which engraves the Tunisian Plateau and the south margin of the Sentinelle Bank [34]. The canyon head appears disconnected from the Tunisian shelf margin and extends over an area of about 1000 km2 at an average depth of about 500 m (Figure 1). It represents a particular type of canyons since it is not fed by an emerged areas or by a fluvio-deltaic systems (Reading and Richards, 1994, [35, 36], but it drains a wide submarine plateau.

3

Material and methods

The physical properties of the core were measured at 1 cm step in a fully automated GEOTEK Multi-Sensor Core Logger (MSCL), in the petrophysical laborato493

Marine Geology

Figure 2: The core CIESM 08: photography, lithologic log, petrophysical properties curves (magnetic susceptibility, Grape density, reflectance 550 nm %) plotted against depth (cm below sea floor). ries of IAMC in Naples (Italy). The MSCL system includes a Bartington MS2E Point sensor, to measure the low-field magnetic susceptibility (MS) with a spatial resolution of 0.4 cm and a Minolta Spectrophotometer CM 2002 which records at 0.8 cm step, the percentage of reflected energy (RSC) at 31 wavelengths in 10-nm steps, over the visible spectrum (from 400 to 700 nm).

The analysis of planktonic foraminifera

494

was conducted on 216 samples. Sampling spacing was 2 cm from the top of the core down to the base. Each wet sample of about 20 g was dried at 50° C and washed over sieves with mesh-width size of 63 microns. Quantitative planktonic analyses were carried out on the fraction >125¾m.

The adopted taxonomic units were those reported by [37, 38]. According to [37] and [38], we introduced some supraspecific categories (which remain unchanged even

Marine research at CNR

Figure 3: Relative abundance of selected planktonic foraminifera from core C08 plotted vs depth (m bsf). Eco-biozones 1 to 10 are pointed out, according to the Ecozonal scheme of [9] slightly modified. Grey bands show the selected eco-bioevents proposed by [11, 12]. Event Stratigraphy according to GRIP scheme. MIS and age scale according to [11, 12]. Associated to G. ruber curve, two grey bands interbedded with the black one correspond to the position of sapropel 1 equivalent with the two interval S1a and S1b.

The three banded areas indicate the position of the turbidite layers. under bad preservation conditions), reducing the number of species actually occurring in the planktonic foraminiferal assemblages.

The AMS 14C radiocarbon analysis, was performed at the CIRCE (Centre for Isotopic Research for Cultural and Environmental heritage) laboratory in Caserta (Italy).

In particular, two AMS 14C analysis (at the core top and at 0.44 mbsf) were carried out on mixed planktonic foraminifera (Globigerina bulloides and Globorotalia inflata). All radiocarbon dates were corrected using a reservoir age of 48 Âą 21 yr (a mean DR value calculated among six of the Tyrrhenian Sea) and calibrated using the marine data base and the CALIB 5.0 Program of [35].

4

Results: lithostratigraphy and planktonic foram

The sediment consists of hemipelagic mud, ranging in colour from reddish and ochre to light, olive and dark grey, punctuated by three turbiditic sand layers (Figure 2) from 1.82 to 2.08 m (T1), from 3.10 to 3.17 m (T2), and from 4.33 to 4.34 m (T3). The turbidite layer T1 is marked by a sharp erosive contact and consists of a thin layer of oxidized sand, and then of a massive fine to medium sand with high percentage of shell fragments; the upper boundary is sharp and the grain size populations comprised between fine sand and clay, that usually pertain to â&#x20AC;&#x153;b, c, d, eâ&#x20AC;? divisions of the classical Bouma sequence [39], are missing, as evidenced by the abrupt decrease of MS and Grape density values, which are a function of grain size and lithology. A sharp contact marks the onset of the turbidite layer T2 495

Marine Geology

which contains a thin layer of well sorted dark fine sand passing to a massive, well sorted bioclastic fine sand; also in this case the upper boundary is sharp and the passage from massive and structureless sand to the hemipelagic mud is abrupt. Turbidite layer T3 has well defined sharp boundaries and starts at the base with a thin layer of dark sand passing to light grey bioclastic sand. The thickness of the turbidites increases upward, but no sand layers occur in the uppermost 0.18 m of the core.

The planktonic foraminifera, characterized by modern assemblages, are abundant and well-preserved and the percentages of foraminiferal fragments are very low and do not alter the composition of the planktonic assemblage.

In terms of quantitative distribution of planktonic foraminifera a number of 13 species or groups of species were distinguished: Globigerina bulloides (including extremely rare specimens of G. falconensis), Globigerinoides quadrilobatus (including G. trilobus and very subordinate G. sacculifer), G. ruber white and pink (always extremely rare), G. elongatus (very rare) and G. gomitolus (very rare), Globigerinoides tenellus (rare), Globorotalia truncatulinoides sl. right (very rare) and left coiling, Neogloboquadrina pachyderma right and left coiling (extremely rare), N. dutetrei right and left coiling (extremely rare), Globigerinita glutinata, Orbulina universa, Turborotalita quinqueloba, Globigerinatella siphoniphera (rare) including G. calida (very rare), Globoturborotalita rubescens (rare).

The long-term trend in planktonic foraminifera reveal that the faunal composition of the studied interval does not show drastic changes in the abundance patterns (Figure 3). In particular, among the taxa that have a continuous distribu496

tion patterns, G. bulloides, G. ruber, G. inflata left coiled, G. scitula right coiled, N. pachyderma right coiled and T. quinqueloba show long-term oscillation (trend) superimposed on short-term fluctuations possibly related to high-frequency climatic oscillations (Figure 3). Among the planktonic species having discontinuous distribution, G. quadrilobatus, G. truncatulinoides left coiled and G. tenellus only occasionally reach significant percentages (Figure 3).

5

Planktonic foraminiferal eco-biozonation

Significant changes in quantitative distribution of the planktonic foraminifera species allowed several authors [40, 25, 38, 41, 17, 9, 15] to define eco-biozones useful for fine subdivisions of the stratigraphic record. The eco-biozone boundaries are characterized by temporary appearance or disappearance and/or evident abundance peaks of different taxa. In the present work, we refer the stratigraphic record to eco-biostratigraphic classification of [9], slightly modified. At present, the ecobiostratigraphic classification of [9]) proposes 9 eco-biozones over the last 23 kyr.

In this work, we propose to mark the base of eco-biozone 9 with the strong increase of G. inflata, occurring in the Mediterranean area at about 30 ka [15, 13], and to extend the eco-biozone 10 back to â&#x2C6;ź80 kyr.

Actually, using the quantitative distribution pattern of the most abundant planktonic foraminifera species counted throughout the C08 core, we identified 10 evident ecobiozones from top to bottom (Figure 3).

The uppermost part of the studied record attributed to the eco-biozone 1, which

Marine research at CNR

Table 1: Tie-points used for age-depth profile of core C08. The Globigerina bulloides eco-bioevents B3 to B11 are coded according to [11, 12]. bases at 0.08 m, is characterized by a decrease in abundance of G. quadrilobatus and the end of G. truncatulinoides right coiled. This attribution is also supported by a 14C â&#x20AC;&#x201C; AMS calibration, which dates 2.18 ka cal. BP (Figure 3, Table 1).

The eco-biozone 2 is defined by the concomitant abundance of G. quadrilobatus and G. truncatulinoides left and right coiled and by low abundance values of N. pachyderma right coiled in the lower part.

Besides, the strong increase of G. truncatulinoides left coiled marks the base of the eco-biozone (Figure 3).

The short interval represented by ecobiozone 3, whose base is at 0.31 m, is marked by the end of micropaleontological signature of sapropel S1 and is character-

ized by low abundance values of T. quinqueloba, G. quadrilobatus, and G. truncatulinoides left coiled.

The eco-biozone 4 corresponds to the time interval of sapropel S1 deposition, although no lithological evidence was found, but for colour shades (Figure 2). G. ruber oscillations allowed a reliable identification of the faunal signature of the climatic events associated to the deposition of sapropel S1 (Figure 3). In particular, two distinct peaks in G. ruber mark the two short-term warm oscillations (S1a and S1b, [9]), separated by a cold phase between them (Figure 3). The event is supported by a 14C-AMS datum at 0.44 mbsf (within the cold phase) with an age of 8.79 ka cal. BP (Table 1) as well as an increase in abun-

497

Marine Geology

dance of G. tenellus and G. quadrilobatus [9].

The eco-biozone 5 is defined by the concomitant occurrence of G. ruber and G. inflata, by a distinct peak of G. truncatulinoides left coiled and absence of N. pachyderma right coiled and very low value of T. quinqueloba. The eco-biozone 6 is marked by the absence of G. ruber and G. inflata left coiled, by a distinct peak of T. quinqueloba and of N. pachyderma right coiled. This eco-biozone corresponds to the Younger Dryas event, according to [9].

The eco-biozone 7 is defined by the increase in abundance of G. ruber and G. inflata, by the absence of T. quinqueloba and by a distinct peak of G. quadrilobatus and corresponds to the warmer (interstadial) GI-1. According to [9] in the ecobiozone 8 the persistent high abundance of cold species permit correlation of this interval with the GRIP GS-2 period. In particular this eco-biozone is dominated by T. quinqueloba, N. pachyderma right coiled, G. scitula and by absence of G. inflata left coiled and rare G. ruber. The base of this eco-biozone approximates to the base of MIS 2 [11] (Figure 3).

The eco-biozone 9 is characterized by the concomitant absence of G. inflata left coiled and G. ruber, by low abundance value of N. pachyderma right coiled and high abundance of T. quinqueloba and G. scitula right coiled. The eco-biozone 10 is clearly marked by the progressive downward increase in abundance of G. inflata left coiled and G. ruber and the progressive decrease of G. scitula right coiled, T. quinqueloba and N. pachyderma right coiled (Figure 3). No distinctive or drastic events in the planktonic faunal patterns are visible towards the base of the studied record but only short-term oscillation in G. bulloides (B3-B11 eco-bioevents; the adopted sam498

pling resolution does not allow the recognition of G. bulloides B8 eco-bioevent), G. inflata left coiled (I3-I5 eco-bioevents), T. quinqueloba (Q4-Q9 eco-bioevents) and N. pachyderma right coiled (P5-P7 ecobioevents), clearly associated to the millennial climate oscillations occurring in the last 80 kyr [12]. According to [12] close to the G. bulloides eco-bioevents B7 and B10 are placed the MIS 4/MIS 3 and MIS4/MIS 5a transition, respectively. Finally, according to [12] the lowermost part of the studied record lies within the eco-bioevent B11 and within the uppermost part of MIS 5b (Figure 3).

6

Age model

The identified foraminiferal marker events, regarded as to reflect major changes in oceanographic conditions and already recognised for the Central and Western Mediterranean [37, 38, 25, 24, 9, 12, 13] were used, in combination with two 14C â&#x20AC;&#x201C; AMS data (Table 1), to constrain the age of core CIESM-C08 and strengthen the correlations between the Mediterranean sites. In particular, we used the age model proposed by [11, 12] for the Alboran Sea, to recognize the top and base of the eco-bioevents recorded in the core C08, the age model proposed by [25] for the Adriatic Sea, to identify the Younger Dryas and the base of the Greenland isotope interstadial 1 (GI-1) and the age model of [43] for the NGRIP record to distinguish the Greenland isotope stadial 2 (GS-2) (Table 1). The turbidite layers T1, T2 and T3 have been taken into account to construct the age model curve and to estimate the sedimentation rate. A second-order polynomial is needed to describe the age-depth relationship for the studied record, indicating an average sed-

Marine research at CNR

Figure 4: From left to right: comparison in time domain between the distribution pattern of G. bulloides from ODP-Site 977 [12] and the studied core CIESM C08 (the red curve represents a 3-points average). The grey bands and the labels B1 to B11 are from [12].

Age-Depth profile and Sedimentation Rate of core CIESM C08. The adopted tie-points by eco-bioevents and by 14C – AMS data are shown respectively with black boxes and grey boxes. imentation rate of ∼7 cm·kyr−1 from the base to the top and four main excursions (Figure 4).

In order to confirm the reliability of the proposed age model, a three step validation process have been performed. First, the visual comparison, in time domain, between G. bulloides distribution pattern of [12] for the Alboran Sea and the patterns in core CIESM C08 (Figure 4), which confirmed the tuning accuracy.

The second step consisted of the comparison of the colour reflectance record at 550 nm (%) of the core CIESM C08, with the record of the ODP-Site 964 (Figure 5), drilled in the Ionian Sea at 3650 mbsl and astronomically calibrated [16]. This

comparison shows that the large-scale reflectance fluctuations in core CIESM C08 not only have similar pattern to those reported in the Ionian sea record but also encompass absolute values in the same range (±10 nm). Remarkable is also the strong similarity between the colour reflectance signature of the sapropel S1 equivalent, recorded in the studied core, with the sapropel S1 colour reflectance signature in the Ionian basin. On the whole, the good visual correlation obtained between the two records supports the validity of the age model based on the identified ecobioevents and 14C AMS calibrations.

Finally, the third control step consisted in the comparison of the G. ruber distri499

Marine Geology

Figure 5: Comparison in time domain of colour reflectance data of core CIESM C08 (black curve, with 3 points average, red dotted curve) with reflectance data after[33] for ODP-Site 964 (thin black curve) in the Ionian Sea. The numbers I1-I6 indicate tephra layers in ODP-Site 964 and the code 1 and X indicate the position of the sapropel 1 and X, respectively. bution pattern chronology with the δ18O NGRIP ice core record. The ecological features of G. ruber associated to warm and oligotrophic surface waters has been established in several oceanographic settings [44, 45, 46, 47] and it is considered an useful tool as recorders of climatic variability [48, 49, 9, 10]. The G. ruber and δ18O NGRIP ice core records exhibit a remarkable agreement, with the identification of the Heinrich events (H1 to H6) and of the Younger Dryas (YD) in the studied record which further support the reliability of our tuning (Figure 6).

pod, bivalve and echinoderm debris) and the features of the surfaces bounding T1, T2 and T3 turbidites lead us to infer a distant source of transported material, since it seems to be remobilized from high productivity areas. Both the slope of the median Ridge, and the southern Cornaglia slope can be ruled out as possible source areas for this bioclastic sand rich density currents, since their top is located respectively at about 1300 and 1000 m bsl. Namely, Sartucya 6 diving survey [33] showed that the base of the southern slope of the median Ridge (investigate from 2270 to 1940 mbsl) is draped with mud, shaped by current bedforms, while the upper slope (investigated 7 Ages and provenance of from 1990 to 1640 m bsl) is characterized by conglomerates and sandstone layturbidite events ers outcropping from the mud, then volcanic rocks. The lithology pertaining to The size population of grains, the grain fab- the southern Cornaglia slope has been deric, the high content in bioclasts (gastro500

Marine research at CNR

Figure 6: Distribution pattern of G. ruber (thin black line) with 3 points average (red line) of core CIESM C08 plotted versus δ18O NGRIP [42] record, with 7 points average, in time domain. Labels H1 to H6 indicate the position of Heinrich events and the label YD the position of the Younger Dryas. scribed during the Sarcya 2 submersible diving [33], which evidenced the occurrence of pelagic mud along the plain (surveyed from 2500 to 2250 mbsl) and cemented coarse material in correspondence of the steep slopes (from 2250 to 2060 mbsl). Although the occurrence of several canyons along the Southern Cornaglia slope was evidenced by the bathymetric surveys [51], nevertheless they enter the Sentinelle Valley seaward to the core site and thus they possibly fed a deepest sector of the basin. Thus, we have to infer that the Bizerte Canyon may have acted as the main conduit to transport the bioclastic sand from the high productivity areas of the Tunisian plateau (Figure 1); then the flows had to cover more than 50 km before settling. As highlighted by many authors over the years [52, 53, 30] gentle gradient slopes or pre-existing slope conduits can drive very efficient density currents

[54, 22] able to cover long distances in relatively “instantaneous” time and to “segregate the original grain populations into distinct and relatively well-sorted facies types with distances” [55]. The beds occurring in C08 core and in particular T1 and T2, could correspond to the facies tract F7 in [55], composed predominantly of medium to fine grained well sorted sand. In fact in this model, which associates the horizontal grain size partition of the deposit with the different degrees of flow efficiency, the F7 facies tract consists of medium to fine sand overlying a mm thick traction carpet that accounts for the development of an erosion surface at the base. This model may explain why T1 and T2 turbidites lack parallel and ripple lamination and pelitic divisions.

The turbidite layers in CIESM core C08 are confined in the lower 4 m of the core and their thickness increases upward peaking

501

Marine Geology

Figure 7: The inferred age of turbitite sand beds is plotted on curves of sea level variations over the last 100 kyr (from [50], modified). T2 and T3 turbidites occur during two relative lowstand phases of the MIS 4 and MIS 3, while the T1 turbidite falls during the last dropping phase of sea level of MIS 3. at T1 (Figure 2) and pointing to a general regressive trend. Through the age model scheme of CIESM C08 core here proposed, it is possible to date T layers respectively at 28 ka cal. BP (T1), 48 ka cal. BP (T2) and 63 ka cal. BP (T3) (Figure 5), thus during the MIS 4 and MIS 3. The emplacement of T1 event bed caused the removal of an undefined thickness of hemipelagic mud correspondent to a time span of about 4 kyr (Figure 4).

Plotting their inferred age on the sea level variation curve relative to the last 100 kyr [56, 57], a strong correlation of T2 and T3 turbidite event beds deposited in this sector of the Sentinelle Valley emerges with two relative sea level low stand phases, while

502

the T1 corresponds to the falling stage of sea level that led to the maximum sea level low stand at about 20 ka BP (Figure 7).

This observation seems in agreement with the most accepted stratigraphic models of deep sea deposition [23]. Thus, the part of the basin fan intercepted by the CIESM core C08 was active and fed with bioclastic sand deposition during the relative sea level minimum and increased its transport efficiency following the sea level lowering.

Nevertheless the CIESM core C08 does not record any further event of sand deposition during the maximum sea level low stand, relative to the MIS 2. A rapid starvation in detritus supply occurred in this area starting from 28 kyr and the middle fan fos-

Marine research at CNR

silized below a drape of hemipelagic mud.

It is reasonable to suppose that during this phase, this sector of the fan acted mainly as bypass area. In this hypothesis any possible sand flow would have been deposited basinward moving possibly through the channel showed in Figure1b.

8

Conclusions

The multidisciplinary study of core C08, recovered from the deep sector of the Sardinia Channel, based on planktonic foraminiferal assemblages and petrophysical data, provides an integrated stratigraphic reference record for the Western Mediterranean Sea that spans back for about 83 kyr. The most important ecobioevents widely used for large scale correlation in the Western Mediterranean area were recognized providing a detailed correlation with the eco-stratigraphic reconstruction proposed by [11, 12] for the Alboran Sea. According to [11, 12], the documented short-term oscillations in the planktonic foraminiferal fauna are clearly associated to the stadial/interstadial excursions occurring over the last 80 kyr and allowed the identification in core C08 of the S1, YD, GI-1 and GS-2 climatic events in the last 23 kyr. Furthermore, the comparison between the δ18O NGRIP ice core record with G. ruber oscillation of core C08, suggests the identification in the studied record of the Heinrich events (H1 to H6) and of

the Younger Dryas (YD).

The eco-bioevents chronology combined with 14C – AMS data were used to define a detailed age model which was compared by means of reflectance parameters to the astronomically tuned age model proposed for the Ionian Sea ODP-Site 964 [16].

The similarity between the two reflectance records, validated the age model of the studied record especially in the time intervals between 2-25 kyr and 60-83 kyr. This methodology, if confirmed with further evidences, proved a powerful tool for reliably correlating marine records between comparable deep sea environment settings. The sector of the Sentinelle Valley intercepted by the CIESM core C08 has been sporadically fed by sand turbidite flows, likely driven along the Bizerte Canyon from the northern sector of the Tunisian Plateau, during relative sea level minimum and sea level drop of MIS 4 and 3. This sector of the basin was reached by three sand deposition events of increasing thickness in the time interval from 64 to 28 kyr cal. BP following a regressive trend. Starting from about 28 kyr this part of the fan was deactivated and fossilized beneath a carpet of hemipelagic mud at a sedimentation rate of about 7 cm·kyr−1 .

The combined logging of sedimentological and petrophysical data of core CIESM-C08, integrated with the ecobiozone stratigraphy could provide an important source of information useful to improve the confidence of correlations in the Mediterranean for the last 83 kyr.

References [1] GRIP members. Greenland Ice-Core Project (GRIP) members, Climate instability during the last interglacial period recorded in the GRIP ice core. Nature, 364:203–207, 1993.

503

Marine Geology

[2] H. Heinrich. Origin and consequences of cyclic ice rafting in the Northeast Atlantic Ocean during the past 130,000 years. Quaternary Research, 29:142–152, 1988.

[3] W. Dansgaard, S.J. Johnsen, H.B. Clausen, D. Dahl-Jensen, N.S. Gunderstrup, C.V. Hammer, C.S. Hvidberg, J.P. Steffensen, A.E. Sveinbjornsdottin, and J. Jouzel G. Bard. Evidence for general instability of past climate from a 250 kyr ice-core record. Nature, 364:218–220, 1993.

[4] E.J. Rohling, A. Hayes, D. Kroon, S. De Rijk, W.J. Zachariasse, and D. Eisma.

Abrupt cold spells in the NW Mediterranean. Paleoceanography, 13:316–322, 1998.

[5] I. Cacho, J.O. Grimalt, C. Pelejero, M. Canals, F.J. Sierro, and J.A. Flores N. Shackleton. Dansgaard–Oeschger and Heinrich event imprints in Alboran Sea paleotemperatures. Paleoceanography, 14(6):698–705, 1999.

[6] I. Cacho, J.O. Grimalt, F.J, Sierro, and N.J. Shackleton M. Canals. Evidence for enhanced Mediterranean thermohaline circulation during rapid climate coolings.

Earth and Planetary Science Letters, 183:417–429, 2000.

[7] I. Cacho, J.O. Grimalt, M. Canals, L. Sbaffi, N.J. Shackleton, J. Sch¨onfeld, and R. Zahn. Variability of the western Mediterranean Sea surface temperature during the last 25,000 years and its connection with the Northern Hemisphere climatic changes. Paleoceanography, 16:40–52, 2001.

[8] E.J. Rohling, T.R. Cane, S. Cooke, M. Sprovieri, I. Boulabassi, K.C. Emeis, R. Schiebel, D. Kroon, F.J. Jorissen, A. Lorre, and A.E.S. Kemp. African monsoon variability during the previous interglacial maximum. Earth and Planetary Science Letters, 202:61–75, 2002.

[9] R. Sprovieri, E. Di Stefano, A. Incarbona, and M.E. Gargano. A high-resolution of the last deglaciation in the Sicily Channel based on foraminiferal and calcareous nannofossil quantitative distribution. Palaeogeography, Palaeoclimatology, Palaeoecology, 202:119–142., 2003.

[10] R. Sprovieri, E. Di Stefano, A. Incarbona, and D.W. Oppo. Suborbital climate variability during Marine Isotopic Stage 5 in the central Mediterranean basin: evidence from calcareous plankton record. Quaternary Science Reviews, 25:2332–2342, 2006.

[11] M. P´erez-Folgado, F.J. Sierro, J.A. Flores, I. Cacho, J.O. Grimalt, R. Zahn, and N. Shackleton. Western Mediterranean planktonic foraminifera events and millennial climatic variability during the last 70 kyr. Marine Micropaleontology, 1-2:49– 70, 2003.

[12] M. P´erez-Folgado, F.J. Sierro, J.A. Flores, J.O. Grimalt, and R. Zahn. Paleoclimatic variations in foraminifer assemblages from the Alboran Sea (Western Mediterranean) during the last 150 ka in ODP Site 977. Marine Geology, 212:113–131, 2004.

504

Marine research at CNR

[13] M. Geraga, S. Tsaila-Monopolis, C. Ioakim, G. Papatheodorou, and G. Ferentinos.

Short-term climate changes in the southern Aegean Sea over the last 48,000 years.

Palaeogeography, Palaeoclimatology, Palaeoecology, 220:311–332, 2005.

[14] G. Buccheri, G. Capretto V. Di Donato, P. Esposito G. Ferruzza T.S. Pescatore, E. Russo Ermolli, M.R. Senatore, M. Sprovieri, M. Bertoldo, D. Carella, and G. Madonna. A high resolution record of the last deglaciation in the southern Tyrrhenian Sea: environmental and climatic evolution. Marine Geology, 186:447– 470, 2002.

[15] E. Ducassou, L. Capotondi, A. Murat, S. Bernasconi, T. Mulder, E. Gonthier, S. Migeon, J. Duprat, J. Giraudeau, and J. Mascle. Multiproxy Late Quaternary stratigraphy of the Nile deep-sea turbidite system- Towards a chronology of deep-sea terrigeneous systems. Sedimentary Geology, 200(1-2):1–13, 2007.

[16] L.J. Lourens. Revised tuning of Ocean Drilling Program Site 964 and KC01B (Mediterranean) and implication for the d18O, tephra, calcareous nannofossil, and geomagnetic reversal chronologies of the past 1.1 Myr. Paleocenography, 19:1–20, 2004.

[17] M.S. Principato, S. Giunta, C. Corselli, and A. Negri. Late Pleistocene–Holocene planktonic assemblages in three box-cores from the Mediterranean ridge area (west–southwest of Crete): palaeoecological and palaeoceanographic reconstruction of sapropel S1 interval. Palaeogeography, Palaeoclimatology, Palaeoecology, 190:61–77, 2003.

[18] A. Piva, A. Asioli, R.R. Schneider, F. Trincardi, N. Andersen, E. ColmeneroHidalgo, B. Dennielou, J.A. Flores, and L. Vigliotti. Climatic cycles as expressed in sediments of the PROMESS1 borehole PRAD1-2, central Adriatic, for the last 370 ka: Integrated stratigraphy. Geochemistry, Geophysics, Geosystems, 9(1), 2008.

[19] E.J. Rohling, P.A. Mayewski, and P. Challenor. On the timing and mechanism of millennial-scale climate variability during the last glacial cycle. Climate Dynamics, 20:257–267, 2003.

[20] R.G. Walker. Turbidites and submarine fans. Geological Association of Canada, pages 239–263., 1992.

[21] C. Beaudouin, B. Dennielou, T. Melki, F. Guichard, N. Kallel, S. Bern´e, and A. Huchon. The Late-Quaternary climatic signal recorded in a deep-sea turbiditic levee (Rhˆone Neofan, Gulf of Lions, NW Mediterranean): palynological constraints. Sedimentary Geology, 172(1-2):85–97, 2004.

[22] M. Richards, M. Bowman, and H. Reading. Submarine-fan systems I: characterization and stratigraphic prediction. Marine and Petroleum Geology, 15(7):689–710, 1998.

505

Marine Geology

[23] W.R. Normark, D.J.W. Piper, and R.N. Hiscott. Sea level control on the textural characteristics and depositional architecture of the Hueneme and associated fan systems, Santa Monica Basin, California. Sedimentology, 26:749–774, 1998.

[24] L. Sbaffi, F.C. Wezel, G. Curzi, and U. Zoppi. Millennial- to centennial-scale palaeoclimatic variations during Termination I and the Holocene in the central Mediterranean Sea. Global and Planetary Change, 40:201–217, 2004.

[25] A. Asioli, F. Trincardi, J.J. Lowe, D. Ariztegui, L. Langone, and F. Oldfield. Submillennial scale climatic oscillations in the central Adriatic during the Lateglacial.

Paleoceanographic implications. Quaternary Science Review, 20:1201–1221, 2001.

[26] D. Ariztegui, A. Asioli, J.J. Lowe, F. Trincardi, L. Vigliotti, et al. Palaeoclimate and the formation of sapropel S1: inferences from Late Quaternary lacustrine and marine sequences in the central Mediterranean region. Palaeogeography, Palaeoclimatology, Palaeoecology, 158:215–240, 2000.

[27] H. L. Bryden and T. H. Kinder. Steady two-layer exchange through the Strait of Gibraltar. Deep-Sea Research, 38(1):445–463, 1991.

[28] C. Millot. Circulation in the western Mediterranean Sea. 10:143–149, 1987.

Oceanology Acta,

[29] N. Pinardi and E. Masetti. Variability of the large scale general circulation of the Mediterranean Sea from observations and modelling: a review. Palaeogeography, Palaeoclimatology, Palaeoecology, 158:153–173., 2000.

[30] H.G. Reading and M.T. Richards. The classification of deep-water siliciclastic depositional systems by grain size and feeder system. American Association Petroleum Geologists Bulletin, 78:792–822, 1994.

[31] F. Budillon, F. Lirer, M. Iorio P. Macr`ı, L. Sagnotti, M. Vallefuoco, L. Ferraro, S. Innanzi, M. Sahabi, and R. Tonielli. Integrated stratigraphic reconstruction for the last 80 kyr in a deep sector of the Sardinia Channel (Western Mediterranean).

Deep - Sea Research II, 2009.

[32] G.H. Mascle, P. Tricart, L. Torelli, J.P. Bouillin, R. Compagnoni, S. Depardon, J. Mascle, A. Pecher, D. Peis, F. Rekhiss, F. Rolfo, H. Bellon, G. Brocard, H. Lapierre, P. Moni´e, and G. Poupeau. Structure of the Sardinia Channel: crustal thinning and tardi-orogenic extension in the Apenninic-Maghrebian orogen and results of the Cyana submersible survey (SARCYA and SARTUCYA) in the Western Mediterranean. Bull. Soc. Geol. Fr.,, 175(6):607–627, 2004.

[33] G. Mascle, P. Tricart, J.P. Bouillin, R. Compagnoni, S. Depardon, J. Mascle, A. Pecher, D. Peis, F. Rekhiss, F. Rolfo, and L. Torelli. Donn´ees de campagne des plong´ees Cyana Sarcya-SartucyaEnhanced aridity and atmospheric high-pressure stability over the western Mediterranean during the North Atlantic cold events of the past 50 ky. Geologie Alpine, M´em. H.S.,, 34:7–113, 2001.

506

Marine research at CNR

[34] G.H. Mascle, P. Tricart, L. Torelli, J.P. Bouillin, F. Rolfo, H. Lapierre, P. Moni´e, S. Depardon, J. Mascle, and D. Peis. Evolution of the Sardinia Channel (Western Mediterranean): new constraints from a diving survey on Cornacya seamount off SE Sardinia. Marine Geology, 179(3-4):179–201, 2001.

[35] M.A. Prins, G. Postma, J. Cleveringa, A. Cramp, and N.H. Kenyon. Controls on terrigenous supply to the Arabian Sea during the Late Quaternary: the Indus Fan. Marine Geology, 169:327–349, 2000.

[36] N.H. Kenyon, I. Klaucke, J. Millington, and M.K. Ivanov. Sandy submarine canyon-mouth lobes on the western margin of Corsica and Sardinia, Mediterranean Sea. Marine Geology, 184:69–84, 2002.

[37] F.J. Jorissen, A. Asioli, A.M. Borsetti, L. de Visser, F.J. Hilgen, E.J. Rohiling, K. van der Borg, C. Vergnaud-Grazzini, and W.J. Zachariasse. Late Quaternary central Mediterranean biochronology. Marine Micropaleontology, 21:169–189, 1993.

[38] L. Capotondi, A.M. Borsetti, and C. Morigi. Foraminiferal eco-biozones, a high resolution proxy for the late Quaternary biochronology in the central Mediterranean.

Marine Geology, 153:253–274, 1999.

[39] A.H. Bouma. Sedimentology of some flysch deposits, a graphic approach to facies interpretation. page 168, 1962.

[40] A. Asioli, F. Trincardi, J.J. Lowe, and F. Oldfield. Short-term climate changes during the last Glacial-Holocene transition: comparison between Mediterranean and North Atlantic records. Journal of Quaternary Science, 4:3732–3781, 1999.

[41] J.S.L. Casford, E.J. Rohling, R. Abu-Zied, S. Cooke, C. Fontanier, M. Leng, and V. Lykousis. Circulation changes and nutrient concentrations in the late Quaternary Aegean Sea: A nonsteady state concept for sapropel formation. Paleoceanography, 17:1–11, 2002.

[42] NGRIP members. High-resolution record of Northern Hemisphere climate extending into the last interglacial period. Nature, 431:147–151, 2004.

[43] K.K. Andersen, A. Svensson, S.J.S.O. Johnsen Rasmussen, M. Bigler, R. Rothlisberger, U. Ruth, M.L. Siggaard-Andersen, J.P. Steffensen D. Dahl-Jensen, B.M. Vinther, and H.B.Clausen. The Greenland Ice Core Chronology 2005, 15-42 ka.

Part 1: constructing the time scale. Quaternary Science Reviews, 25:3246–3257, 2006.

[44] C. Hemleben, M. Spindler, and O.R. Anderson. Modern Planktonic Foraminifera. page 1–363, 1989.

[45] C. Pujol and C. Vergnaud-Grazzini. Distribution patterns of live planktonic foraminifers as related to regional hydrography and productive system of the Mediterranean Sea. Marine Micropaleontology, 25:187–217, 1995.

507

Marine Geology

[46] J.M. Watkins, A.C. Mix, and J. Wilson. Living planktic foraminifera: tracers of circulation and productivity regimes in the central equatorial Pacific. Deep-Sea Research II, 43:1257–1282, 1996.

[47] S. Zaric, B. Donner, G. Fischer, S. Mulitza, and G. Wefer. Sensitivity of planktic foraminifera to sea surface temperature and export production as derived from sediment trap data. Marine Micropaleontology, 55:75–105, 2005.

[48] R. Sprovieri. Plio-Pleistocene paleoclimatic evolution at ODP Leg 107 Site 653 (Tyrrhenian sea, Western Mediterranean). Memorie della Societ`a Geologica Italiana, 44:135–144, 1991.

[49] R. Sanvoisin, S. D’Onofrio, R. Lucchi, D. Violanti, and D. Castradori. 1Ma Paleoclimatic record from the Eastern Mediterranean- Marflux Project: the first results of micropaleontological and sedimentological investigation of a long piston core from the Calabrian Ridge. Il Quaternario, 6:169–188, 1993.

[50] F. Antonioli, E. Bard, E.K. Potter, and S. Silenzi S. Improta. 215-ka History of sea-level oscillations from marine and continental layers in Argentarola Cave speleothems (Italy). Global and Planetary Change, 34(1-2):57–78, 2004.

[51] G. Brocard. Le canal de Sardaigne au N´eog`ene: analyse morphologique et structurale. Apports de la bathym´etrie multifaisceaux et des plong´ees Sarcya et Sartucya.

Geologie Alpine, M´em. H.S., 34:115–166, 2001.

[52] R.G. Walker. Turbidite sedimentary structures and their relationship to proximal and distal depositional environments. Journal of Sedimentary Petrology, 37(1):25– 37., 1967.

[53] W. R. Normark, N. E. Barnes, and F. Coumes. Rhone Deep-Sea Fan: A review.

Geo-Marine letters,, 3(2-4):155–160, 1984.

[54] J.E. Damuth and R.D. Flood. Morphology, sedimentation processes, and growth pattern of the Amazon Deep-Sea Fan. Geo-Marine Letters, 3(2-4):109–117, 1984.

[55] E. Mutti, R. Tinterri, E. Remacha, N. Mavilla, Angella, and L. Fava. An introduction to the analysis of ancient turbidite basins from an outcrop perspective. American Association Petroleum Geologists, Continuing Education Course, 39, 1999.

[56] C. Waelbroeck, L. Labeyrie, E. Michel, J.C. Duplessy, K. Lambeck, J.F. McManus, E. Balbon, and M. Labracherie. Sea-level and deep water temperature changes derived from benthic foraminifera isotopic records. Quaternary Science Reviews, 21:295– 305, 2002.

[57] J. Imbrie, J.D. Hays, D.G. Martinson, A. McIntyre, A.C. Mix, J.J. Morley, N.G. Pisias, W.L. Prell, and N.J. Shackleton. The orbital theory of Pleistocene climate: support from a revised chronology of the marine δ18O record. page 269– 305, 1984.

508

Holocene Palaeo-Geographical Evolution of the Sele River Alluvial-Coastal Plain: New MorphoSedimentary Data from Poseidonia-Paestum Area B. D’Argenio1 , V. Amato2 , E. Anzalone1 , P.P.C. Aucell3 , M. Cesarano2 , A. Cinque4 , S. Da Prato5 , G. Di Paola2 , L. Ferraro1 , G. Pappone3 , P. Petrosino4 , C. M. Rosskopf2 , E. Russo Ermolli4 1, Institute for Coastal Marine Environment, CNR, Napoli, Italy 2, Department of Science and Technology for the Environment and Territory, University of Molise, Pesche (IS), Italy 3, Department of Environmental Sciences, University of Napoli “Parthenope”, Italy 4, Department of Earth Sciences, University of Napoli “Federico II”, Napoli, Italy 5, Institute of Geosciences and Earth Resources, CNR, Pisa, Italy b.dargenio@iamc.cnr.it Abstract The Sele river plain is located along the western Tyrrhenian margin of the southern Apennine Chain and is defined seawards by a straight sandy coast formed during the Last Interglacial (Tyrrhenian stage, OIS 5e) and the Holocene. It is characterized by the presence of beach-dune ridges which to the rear interfinger with lagoonal and fluvio-palustrine deposits. This belt was accreted progressively and represents the evolution of a barrier-lagoon system shifted alternatively landward and seaward.

In this work, we summarize the main results relative to the Holocene evolution of the Sele river coastal plain along the coast in front of the archaeological area of Poseidonia-Paestum, where the knowledge has been improved by two new cores and by many collected archaeo-tephro-stratigraphical data. The area was affected by the Holocene marine transgression that formed cliffs that cut the travertine deposits. During the second part of the Holocene the shoreline shifted seaward and a lagoonal-beach bar system (Fossa Lupata) formed. The archaeological remains of Poseidonia (VI cent. B.C.) and the Agnano Monte Spina tephra layer (4.1 ky BP) confirm the presence of this morpho-sedimentary system during this time interval.

After this period, and mostly after the deposition of the A.D. 79 tephra, the shoreline shifted seaward and an additional beach ridge was formed, while the flat area of Fossa Lupata, was rapidly aggraded and dried up.

1

Introduction

The Holocene glacio-eustatic sea level rise after the Last Glacial Maximum (LGM) led to a worldwide flooding of shelf areas and controlled the evolution of marine em-

bayments, fluvial mouths and rocky coasts, while its significant deceleration in midHolocene times resulted in the overcompensation by sediment yields and shoreline progradation in many Mediterranean alluvial-coastal plains. The shoreline shifts

Marine Geology

Figure 1: Schematic geological and geomorphological maps of the Sele river alluvial coastal plain. A: Simplifed geomorphological map; Bâ&#x20AC;&#x201C; Simplifed geological map of southern Apennine with in evidence the Tyrrhenian graben of the Sele-plain. 1) Neogene clastic and volcanic deposits; 2) Eboli Conglomerates (Pleistocene); 3) Miocene deposits; 4) Ligurian Units (Cretaceous-Oligocene); 5) Meso-Cenozoic apennine carbonatic shelfs; 6) Lagonegro Units (Trias-Miocene); 7) Meso-Cenozoic apulian carbonatic shelfs; 8) Thrusts; 9) Main faults; 10) Somma-Vesuvius volcano. forced ancient societies to continuously adapt their lives and infrastructures to ever changing natural factors. These rapid changes of sedimentary environments have been investigated throughout the Mediterranean in detail, combining methods and research designs of a great number of disciplines, including geomorphology, geology, palaeobiology, archaeology (e.g. [1, 2, 3, 4, 5, 6, 7, 8]). All the studies assert that the Holocene sea level rise caused, firstly, a general marine transgression into the alluvial-coastal plains of the Mediterranean Sea, and successively, a strong

510

progradational trend of shorelines. The coastal progradation is relative to the decrease in the rate of sea level rise and to the increase in the sediment load of rivers.

It was more marked during the last 2.5 ky, because favored by the increase in anthropic impacts on vegetation and rivers [9, 10, 11, 12]. Similarly, it is known that the Tyrrhenian alluvial-coastal plain of the Sele river was interested by the same morpho-sedimentary behavior during the Holocene, with a transgressive trend during the early Holocene and with a progradational trend of shorelines starting from

Marine research at CNR

middle Holocene ([13, 14, 15, 16, 17, 18] and references herein). The plain, located along the western Tyrrhenian margin of the southern Apennine Chain, is defined seawards by a straight sandy coast formed during the Last Interglacial (Tyrrhenian stage, OIS 5e) and the Holocene. It is characterized by the presence of beachdune ridges which to the rear interfinger with lagoonal and fluvio-palustrine deposits. This belt was accreted progressively and represents the evolution of a barrier-lagoon system shifted alternatively landward and seaward (Figure 1a). The coastal belt shows the geomorphological evidences (dune ridges and flat depression areas) of the late-Quaternary glacioeustatic sea level changes and in particular of the sea level high stand of the last interglacial periods: Tyrrhenian High Stand Sea Levels - HSSLs of MIS 5 - and Holocene High Stand Sea Level â&#x20AC;&#x201C; HSSL of MIS 1. In fact, the coastal belt was accreted progressively during the late Quaternary and represents the evolution of a barrierlagoon system shifted alternatively landward and seaward as a result of the eustatic sea level changes. In the SE portion of the Sele Plain, near the archeological area of Poseidonia-Paestum, numerous generations of travertine deposits crop out. The depositional system of the â&#x20AC;&#x153;Travertini di Paestumâ&#x20AC;? was active during the late Quaternary, especially during the Last Interglacial period and the Holocene [19], forming a plateaux hanging over the plain. In this sector of the coastal plain the morphologies linked to the eustatic changes are not recognizable, because absent and/or interfingered and/or covered by travertine deposits. Using geomorphological and stratigraphic methods, integrated by geo-archeological and tephro-stratigraphic data, we focused the researches on the

Holocene morpho-stratigraphy changes of this sector of the Sele plain, in order to decipher the local sea level rise history and horizontal shoreline changes. Sedimentary evidence is presented for the Holocene marine transgression due to Postglacial sea level rise and the shoreline progradation, respectively, resulting from reduced eustatic effects and, to a minor extent, increased sediment loads.

2

Material and methods

As regards the Vektor-Vulkost Project (thematic line 2), a multidisciplinary approach, based on a detailed sedimentological, geomorphological and structural characterization of the Sele plain, was carried out in order to reconstruct the paleoenvironmental and landscape changes that occurred during the Holocene, and the related chronological framework. The sedimentary infilling of the SE sector of the Sele plain was studied in detail through two new cores, 15 m long, which were drilled in the coastal sector in front of the archaeological area of Poseidonia-Paestum. Further stratigraphic data were obtained from the interpretation of ca. 200 stratigraphic logs of cores drilled for geotechnical purposes and from some archaeological trenches. On the basis of lithofacies, unconformities, presence of tephra layers and paleosoils, the core and trench successions were subdivided into sedimentary units, by using the Unconformity Boundary Stratigraphic Unit method (UBSU, after Salvador, 1994). The most important layers of the cores were sampled and subjected to laboratory analyses such as pollen and fossil content and tephra-stratigraphy, in order to reconstruct the UBSU chronology. Contemporary, a large scaled geomorphologic analysis of 511

Marine Geology

the study area, based on field surveys and on airphoto and topographic map interpretation (CasMez 1:5.000 and I.G.M.I. 1:25.000), was carried out. This analysis allowed us to identify landscape features of different types and ages and to define the present-day geomorphologic setting.

3

Geological and geomorphological setting

The Sele plain derives from the aggradation of a Pliocene-Quaternary depression located along the western Tyrrhenian margin of the southern Apennine Chain (Figure 1b). It is about 400 km2 wide and presents a triangular plan outline, which is defined seawards by a straight sand coast stretching between the towns of Salerno and Agropoli. The boundaries of the plain are defined by NW-SE and NE-SW trending faults, active during the Early and Middle Pleistocene. The easternmost portion of this structural depression was characterised by continental conditions as testified by the huge phases of clastic sedimentary aggradation of the â&#x20AC;&#x153;Eboli Conglomeratesâ&#x20AC;? auct., which compensated the Quaternary tectonic subsidence [20, 13, 21, 22]. Further seawards, there is a strip of coastal plain formed during the Last Interglacial (MIS 5e) characterized by the presence of three orders of beach-dune ridges which to the rear interfinger with lagoonal and fluvio-palustrine deposits (Figure 1a).

Only the youngest and most external of the Tyrrhenian coastal ridges still has a good morphological evidence (GromolaArenosola palaeo-ridges). The present elevation a. s.l. of these Tyrrhenian deposits proves that the plain has been moderately uplifted since the last interglacial pe512

riod [13, 23, 14, 15, 16, 17, 18]. Between the Tyrrhenian sandy-coastal ridge and the present shoreline, a younger coastal sector occurs, which is elevated up to 5 m a.s.l.

This belt was accreted progressively and represents the evolution of a barrier-lagoon system shifted alternatively landward and seaward during the Holocene. It includes a composite sandy ridge which is partly exposed along the present coast and disappears inland under a muddy, substantially flat depression. After being exposed to subaerial conditions during the last glacial regression, this belt gradually entered brackish water conditions at the beginning of the Holocene, when transgressive trend occurred as the effect of rapid sea level rise.

The inversion of tendency, from retrogradational to progradational, most probably can be ascribed to a decline of the rate of sea level rise under the threshold of balance with the progradation due to fluvial sedimentation. The progradational trend was interrupted by at least three phases of formation of sandy coastal ridges, known as Laura ridge (dated from 5.3 to 3.6 ky BP) and Sterpina ridges (I and II, dated to before 2.6 ky BP and to about 2.0 ky BP, respectively) [23, 15, 16, 17]. During these intervals the flat depression behind the ridges was interested by palustrine conditions which persisted partially until very recent times, when it was interested by anthropic reclamations. In the southern sector of the Sele plain, near the Greek-Roman archaeological area of Poseidonia-Paestum, lobate and self-terraced morphologies of the Paestum Travertine (Figure 2), composed by several polyphasic deposits of travertine, derived by in situ carbonate encrustation of vegetables and/or by fluvial reworking, and then lithified phytoclasts. They were generated by highly charged calcium carbonate waters of the

Marine research at CNR

Figure 2: Litho-facies map (1:5.000) of the coastal sector of the southern Sele river plain springs located at the base of the Soprano Mt. The recognized lithofacies of travertine (stromatolitic travertines, microhermal travertines, phytohermal travertines, phytoclastic travertines and calcareous tufa) allowed us to refer the depositional system of Paestum Travertine to fluvial marshy conditions that favored the emergence of a large sector hanging over the surrounding plain [24, 25]. In this sector of the coastal plain the morphologies linked to the eustatic changes are not recognizable, because absents and/or interfingered and/or covered by travertine deposits. Recently, [19] have provided a

detailed chronological reconstruction of the various stages of the Paestum depositional system, based on radiometric dating, archeo-tephra-stratigraphic data, and geomorphological constraints (Figure 2). The ages of the deposits show that the depositional systems have migrated from NE to S from the Pleistocene to the Holocene, up to deposit thick travertine successions in the valley of Capodifiume river even in modern times. In particular, the deposition was active during the Last Interglacial (Tyrrhenian) and the early Holocene until about 5000 years ago, and in the historical period during the Late-Ancient and Mid-

513

Marine Geology

Figure 3: Schematic geological and geomorphological map of the coastal sector of the southern Sele river plain, in front of the archaeological area of Poseidonia-Paestum dle Ages (V-IX century AD). Further seaward, downstream of the hanging travertine plateaux, a low coastal belt is present.

It is constituted by a continuous sandy dune ridge (of not more than 6 m a.s.l.), which characterize the sector near the actual beach. The dune belt stands behind a depressed area, situated at an altitude not exceeding 3 m, only recently reclaimed by a complex system of anthropic drainage.

In the coastal sector, in front of the archaeological area of Poseidonia-Paestum, Lippmann-Provansal (1987) proposed, on the basis of finds of pottery fragments at Porta Marina (the port of the Greek-Roman town of Paestum looking to the sea), that a coastal lagoon had already established in the Iron Age (3.0 ky ago) while Guy (1990), on the basis of surveys and the interpretation of satellite images and aerial photographs, suggested that, during the Classic period (2.5 ky ago), there was only a small lagoon (pond or artificially preserved and open to the sea). Precisely in this area of the Sele River alluvialcoastal plain we have focused the morphostratigraphic investigations in order to understand how the Holocene sea level rise changed the environments and the coastal

514

landscape.

4

New morpho-stratigrafical data

The coastal strip in front of the archaeological area of Paestum presents a very articulated landscape, consisting in an inner area situated at an altitude between 10 and 20 m a.s.l., some meters higher than the average level of the plain, that does not exceed 5 m a.s.l. (Figure 3).

Such morphological high, slightly sloping to the sea, appears to be composed of travertine deposits, belonging to different depositional bodies (as previously described). These polyphasic depositional bodies, generated during the late Quaternary, now form self-terraced bodies hanging above the average level of the plain.

For this reason the travertine plateaux developed above the tracks and remains of archaeological settlements, particularly those of the Greek-Roman town of PoseidoniaPaestum. Landward, the hanging travertine bodies are connected to the piedmont belt of the Capaccio hills and seaward to the coastal strip by a steep escarpment cut

Marine research at CNR

Figure 4: Stratigraphic logs of the S2 and S3 cores into travertine, whose remains are still visible at Porta Marina of Paestum. This steep escarpment gently downgrades to a depressed area, situated at an altitude of about 4 m a.s.l. (Fossa Lupata), located behind a large sand dune ridge, which reachs an altitude of about 6 m a.s.l., located about 1 km from Porta Marina (Figure 3). In addition to the numerous data from archaeological excavations, collected during multidisciplinary geoarchaeological collaborations, and to new stratigraphic data derived by collected cores, two new cores (Vektor S2 and S3) were drilled that reached the depth of 15 m. S2 and S3 were performed, respectively, in the area immediately located westwards of the escarpment cut into travertine of the Porta Marina (5.5 m a.s.l.), and in the outer dune ridge, immediately behind the coastal road (2.5 m

a.s.l.) (for location see Figures 2 and 3).

The deposits of the S2 core are constituted of, from the top to the bottom (Figure 4): • Unit 1: (5 m thick) Colluvial deposits, buried soils, anthropogenic restores and silt-clay layers of continental environments, subject to marshy episodes. In this unit the tephra of 79 AD Vesuvius eruption is intercalated at a depth of about 2 m.

• Unit 2: (4,5 m thick) Peaty silt and clay layers of marshy and lagoon environments. At the top of the unit, potteries of the VI cent. B.C. are present, while at 7 m of depth (-1,5 a.s.l.) the fall deposits (pumices and ashes) of the Agnano Monte Spina eruption (4.1 ky BP, Di Vito et al, 1999) are present.

• Unit 3: (4 m thick) Coarse sand and gravel layers of high-energy environ515

Marine Geology

Figure 5: Schematic geological section passing through S2, S3 and other cores. ments (cliff toe with proximity of river mouths).

This succession covers the travertine deposits with a clear unconformity marked by an abrasion surface at -7,5 m a.s.l. The deposits of the S3 core are constituted of, from the top to the bottom (Figure 4): â&#x20AC;˘ Unit 1: (6 m thick) Coastal dune sand layers. This unit be divided into two sub-units (1a for the sandy layers of the upper part and 1b for the sandy layers of the lower part) according to the presence of a buried soil, at the depth of 3 m. This palaeosoil, in some archaeological excavations close to the core, holds archaeological materials of the VI-V cent. B.C. and is partly covered by the fall deposits of the 79 AD eruption.

ers of clays and peaty silts, sands and calcareous tufa and gravels of fluvial-marshy environments. In this unit the fall volcanoclastic deposits of the Neapolitan Yellow Tuff (15 ky BP, Deino et al, 2004) and of the Y3 (30 ky BP Munno and Petrosino, 2004) are intercalated at a depth of 8 m (-5,5 m a.s.l.) and 11 m (-8,5 m a.s.l.), respectively.

5

Discussion

The schematic geological section of Figure 5, that passes through the S2, S3 and other cores, was drawn perpendicular to the actual shoreline, from the sea to the escarpment of Porta Marina, where the remains of the Greek-Roman walls of PoseidoniaPaestum are located. Correlating tephra layers, such as the 79 AD tephra, the Agâ&#x20AC;˘ Unit 2: (9 m thick) Alternating lay- nano Monte Spina tephra (4.1 ky BP), 516

Marine research at CNR

Figure 6: Paleogeographical sketchs of holocene morfo-sedimentary trends of coastal area of Paestum. of the Postglacial period led to a rapid submergence of the area of Porta Marina of Paestum, modeling a steep cliff. At the foot of the latter the deposits of cliff toe of Unit 3 of the S2 core were accumulated. The presence of fluvial clusters in the Unit 3 of the S2 core could witness the proximity of one or more river • The sea level low-stand of the Last mouths.

Glacial Maximum (20 ky BP) led to • As soon as the rate of the Holocene a strong progradation of the shoreline. sea level rise decreased, a rapid shoreTherefore, the whole studied area was line progradational trend started, and a interested by continental environments, barrier-lagoon coastal system formed. In represented in the S3 core by deposits fact, the sands of Unit 1 of the S3 (Unit 2) containing the Neapolitan Yelcore represent the barrier beach that isolow Tuff tephra and the Y3 tephra. In the lated the depressed area of Fossa Lupata, S2 core, the deposits of this phase may where the lagoon-marshy clays and silts have been eroded by subsequent transof Unit 2 of the S2 core were deposited. gressive trend and/or have been obliterPollen analysis revealed the presence of ated by the deposition of travertine boda rich and diversified forest in which the ies. dominant trees and shrubs (Alnus, Cory• The rapid sea level rise of the first part the Neapolitan Yellow Tuff tephra (15 ky BP), the Y3 tephra (30 ky BP), the layers with archaeological remains, and referring the data to the known Holocene morpho-sedimentary trends, it was possible to scan the late Quaternary palaeogeographical evolution of this sector:

517

Marine Geology

lus, Quercus, Carpinus, Vitis) are indicative of high soil moisture. The presence of the Agnano Monte Spina Tephra (4.1 ky BP) and the archaeological remains of the VI cent B.C. in Unit 2 of the S2 core, allowed us to hypothesize the presence of the barrier-lagoon coastal system during this period. The collected chrono-data are in agreement with the dating of the Laura paleoridge coastal deposits (5.32.5 ky BP) [15, 16]. â&#x20AC;˘ During the period between the foundation of the Greek-Roman town of Poseidonia-Paestum (540 B.C.) and the 79 AD, the area of S2 core was interested by continental environments subject to marshy episodes (Unit 1) while the area of S3 core was interested by coastal dune environments (Unit 1). During this period the shoreline prograded a few hundred meters because an additional sandy dune ridge was generated seaward. â&#x20AC;˘ After 79 AD and up to now, the progradational trend of shoreline was emphasized through the addition of another sand dune ridge, testified in the S3 core by the deposits of sub-Unit 1a. In the area of the S2 core, a strong aggradation of the ground level took place due to anthropogenic fills, reworked volcanoclastic deposits of 79 AD and historical deposition of travertines.

6

Conclusion

The stratigraphic data, obtained through the study of the cores, and the chronological framework, derived from the archaeology and tephro-stratigraphy and supported by geomorphological indications, allowed us to outline some important stages of the Holocene palaeogeographical evolution of the SE sector of the alluvialcoastal plain of 518

the Sele River. The main results of the research are relative to the Holocene evolution of the plain along the coast in front of the archaeological site of PoseidoniaPaestum, where the knowledge has been improved by two new cores. In particular, by integrating geomorphological and archeo-tephro-biostratigraphical studies, it was possible to characterize the Holocene evolution of this sector with more detail for the last 6.0 ky BP (Figure 6). In particular, the research shows that the sea level changes and shoreline shifts characterized the morpho-sedimentary trend of the coastal areas: during the early Holocene the morpho-sedimentary trend shows a clear transgressive trend, while the late Holocene shows a progradational trend. In the area of Paestum, the transgressive trend has favored the formation of a cliff cut in travertine, now partly buried by travertine deposits of mostly medieval age. The progradational trend started when an extensive sandy dune ridge (Laura Paleoridge) was generated, now located about 0.6 km from the Porta Marina palaeocliff. This ridge isolated a large depression at its back (Fossa Lupata depression). The archaeological remains related to the VI-V cent.

B.C. and the Agnano Monte Spina tephra (4.1 ka BP) confirm the presence, during this period, of a barrier-lagoon morphosedimentary system that shifted alternatively landward and seaward. The Fossa Lupata depression may have been connected with the sea through fluvial mouths, and it was probably used as a natural port and/or sea port of the Greek city. After this period, and mainly after the deposition of the 79 AD volcanoclastic fall deposits, the shoreline shifted seaward a few hundred meters through the formation of another dune ridge (Sterpina ridge). The Fossa Lupata depression, no longer con-

Marine research at CNR

nected with the sea, was filled by fluvial- in particular Dr. Marina Cipriani to have marshy deposits and slowly dried up. kindly granted the permission for the coring within the Porta Marina archaeological area and Dr. Alfonso Santoriello of the 7 Aknowledgments Salerno Universit`a (Beni Culturali Department) for the interpretation of the ceramic We thank the Soprintendenza Archeolog- materials of the cores. ica of Avellino-Salerno (Paestum Office),

References [1] G. Schmiedt. Antichi porti d’Italia. L’Universo, XLVI(2):297–353, 1971.

[2] J.C. Kraft, G. Rapp Jr., and S.E. Aschenbrenner. Late Holocene palaeogeography of the coastal plain of the Gulf of Messenia, Greece, and its relationships to archaeological settings and coastal change. Geological Society of America Bulletin, 86:1191–1208, 1975.

[3] P. A. Pirazzoli. Sea level variations in the northwest Mediterranean during Roman times. Science, 194:519–521, 1976.

[4] P.A. PIRAZZOLI. Sea-Level Changes. The Last 20.000 Years. page pp.211, 1996.

[5] J. La Borel, C. Morhange, R. Lafont, J. Le Campion, and F. Laborel Deguen. Biological evidence of sea level rise during the last 4500 years on the rocky coasts of continental France and Corsica. Marine Geology, 120:203–223, 1994.

[6] G. Leoni and G. Dai Pra. Variazioni di livello del mare nel tardo Olocene lungo la costa del Lazio in base ad indicatori geoarcheologici (pubblicazione ENEA-CNR). page 127, 1997.

[7] K. Lambeck, F. Antonioli, A. Purcell, and S. Silenzi. Sea level change along the Italian coast for the past 10,000 years. Quaternary Science Reviews, 23:1567–1598, 2004.

[8] K. Lambeck, M. Anzidei, F. Antonioli, A. Benini, and A. Esposito. Sea level in Roman time in the Central Mediterranean and implications for recent change. Earth and Planetary Science Letters, 224:563–575, 2005.

[9] G. Vita-Finzi. The mediterranean valleys: geological changes in historical times.

Cambridge University Press. page 186, 1969.

[10] R.S. Bradley. Palaeoclimatology. International Geophysic Series 64, Harcourt Academic Press. page 325, 1999.

[11] B. Messerli, M. Grosjean, T. Hofer, L. Nunez, and C. Pfister. From naturedominated to human-dominated environmental changes. Quaternary Science Reviews, 19:459–479, 2000.

519

Marine Geology

[12] V. Amato. La risposta di alcuni tipici sistemi morfodinamici della Campania (Italia meridionale) alle variazioni climatiche oloceniche. PhD thesis www.fedoa.unina.it, page 405, 2006.

[13] A. Cinque. Guida alle escursioni geomorfologiche(Penisola Sorrentina, Capri, Piana del Sele e Monti Picentini). Gruppo Nazionale Geografia Fisica e Geomorfologia, Amalfi 1986. page 119.

[14] L. Brancaccio, A. Cinque, G. D’Angelo, F. Russo, N. Santangelo, and I. Sgrosso.

Evoluzione tettonica e geomorfologica della Piana del Sele (Campania, Appennino meridionale). Geogr. Fis. e Dinam. Quat., 10:47 – 55, 1987.

[15] L. Brancaccio, A. Cinque, F. Russo, N. Santangelo, M. Alessio, L. Allegri, S. Improta, G. Belluomini, M. Branca, and L. Delitala. Nuovi dati cronologici sui depositi marini e continentali della Piana del F. Sele e della costa settentrionale del Cilento (Campania, Appennino meridionale). Atti del 74.mo Congr. Naz. della Soc. Geol. It., A:55 – 62, 1988.

[16] D. Barra, G. Calderoni, A. Cinque, P. De Vita, C. Rosskopf, and E. Russo Ermolli.

New data on the evolution of the Sele River coastal plain (Southern Italy) during the Holocene. Il Quaternario, 11:287–299, 1998.

[17] D. Barra, G. Calderoni, M . Cipriani, J. De La Genier`e, L. Fiorillo, G. Greco, M. Mariotti Lippi, M. Mori Secci, T. Pescatore, B. Russo, M.R. Senatore, G. Tocco Sciarelli, and J. Thorez. Depositional history and palaeogeographic reconstruction of Sele coastal plain during Magna Grecia settlement of Hera Argiva (Southern Italy). Geologica Romana, 35:151–166., 1999.

[18] A. Cinque and P. Romano. Note illustrative della Carta Geologica d’Italia alla scala 1:50.000 foglio 486 Foce del Sele. ISPRA-Servizio Geologico d’Italia. page 83, 2008.

[19] V. Amato, G. Avagliano, A. Cinque, M. Cipriani, G. Di Paola, A. Pontrandolfo, M. C. Rosskopf, and A. Santoriello. Geomorphology and geoarchaeology of the Paestum area: modification of the physical environment in historical times.

Mediterranee, 112:129–135, 2009.

[20] M. Lippmann-Baggioni and G.Gars. La bordure sud des Monts Picentini: un jalon dans l’evolution neotectonique et paleoclimatique de l’Apennin Meridional. Geogr.

Fis. Dinam. Quat., 7:49–58, 1984.

[21] A. Cinque, F. Guida, F. Russo, and N. Santangelo. Dati cronologici e stratigrafici su alcuni depositi continentali della Piana del Sele (Campania): i ”Conglomerati di Eboli”. Geogr. Fis. e Dinam. Quatern., 11:39 – 44, 1988.

[22] A. Zuppetta and A. Sava. Pleistocene brittle deformation in the Eboli Conglomerates (Sele Plain – Campanian Apennines). Boll. Soc. Geol. It., 111:273 – 281, 1992.

520

Marine research at CNR

[23] L. Brancaccio, A. Cinque, G. Belluomini, M. Branca, and L. Delitala. Isoleucine epimerization dating and tectonic significance of upper Pleistocene sea level features of the Sele Plain (Southern Italy). Zeit. Geomorph. N.F., Suppl. Bd.,, 62:159 – 166, 1986.

[24] B. D’Argenio, V. Ferreri, and C. Violante. Travertine in the rise and decline of the ancient town of Paestum (2500-1000 BP). GEOBEN Torino. page 16, 1999.

[25] B. D’Argenio, V. Ferreri, and E. Anzalone. I travertini di Paestum: breve guida al periplo geoarcheologico della citt`a. Guida all’escursione della IX Borse del Turismo Mediteraneo di Paestum, page 25, 2007.

521

Marine Geology

522

Facies Analysis of Flood-Dominated Fan-Deltas off the Amalfi Coast, Eastern Tyrrhenian Sea F. Molisso1 , E. Esposito1 , D. Insinga1 , C. Lubritto2 , S. Porfido1 , M. Sacchi1 , T. T´oth3 , C. Violante1 1, Institute for Coastal Marine Environment, CNR, Napoli, Italy 2, Department of Environmental Sciences, Second University of Napoli, SUN, Caserta, Italy 3, E¨otv¨os Lor´and University, Budapest, Hungary flavia.molisso@iamc.cnr.it Abstract A stratigraphic study of marine gravity cores, complemented by high-resolution seismic profiles acquired off the Amalfi coast, a rocky coastal area on the southern flank of the Sorrento peninsula (Italy), documents the facies associations and the internal stratigraphic architecture of a series of small fan-deltas that develop at the mouth of major bedrock streams. Integrated stratigraphy and correlation of gravitycores allowed for a bed-to-bed calibration of seismic reflectors. Accurate dating and correlation have been essential for the construction of reliable models of the sediment architecture and influx rates in this area, as well as for establishing the links between changes in sedimentation and palaeoenvironmental events. Our research indicates that the Amalfi fan-delta system largely postdates the Plinian eruption of the Vesuvius of AD 79 and displays various phases of development that were ostensibly associated with periods of high sediment supply from the adjacent river basins, under varying climatic conditions. During these periods landscape-mantling loose pyroclastic deposits were quickly eroded and delivered to the continental shelf by sheet wash and flash floods events. This in turn created favourable conditions for seafloor instability, soft sediment failure, slumping and sliding that characterize the deltaic stratigraphic architecture.

1

Introduction

The aim of the study is the detailed reconstruction of stratal architecture of the fandeltas and the interpretation of seismic facies in terms of depositional processes and environmental setting off the Amalfi coast, eastern Tyrrhenian Sea. In recent years, deltaic depositional settings at the mouth of small rivers of the Mediterranean and other temperate regions have received growing attention, due to the relevance of these fa-

cies associations in the understanding of the late Quaternary evolution of inner shelf depositional systems and their interaction with fluvio deltaic processes, seafloor instability of delta slopes, coastal volcanism, active tectonics, and the climatic regime [1, 2, 3]. Sediment dispersal underwater is directly related to supply by rivers.

In the case of bedrock rivers and streams of temperate regions that form fan deltas along high-relief seacliffed coasts [4, 5, 1], the fluvial regime is basically controlled by

Marine Geology

Figure 1: Tectonic sketch-map of the Campania continental margin (Eastern Tyrrhenian Sea) and location of the study area. structural unit of the western flank of Southern Apennines and forms a narrow and elevated mountain range (up to 1444 m) that separates two major embayments of the eastern Tyrrhenian margin, namely the Naples and Salerno Bays (Figure 1).

It is mostly formed by a pile of Mesozoic carbonate rocks, covered by Tertiary to Quaternary siliciclastic and pyroclastic units and is deeply cut by a complex pattern of bedrock rivers and channels characterized by relatively small catchment areas and pronounced disequilibrium of the stream profiles. These rivers show a distinct seasonality and a torrential regime [10, 11]. Coarse-grained coastal alluvial fans confined by narrow valleys at the mouth of the major streams are relatively common in this setting (Figure 2). They are formed by deposition from flash floods, 2 Geological framework that occur during heavy rain storms. The The study area is located on the south- delivery of sediments towards the coastal ern slope of the Sorrento Peninsula. The fans is favoured by the steep slopes and peninsula is a major Quaternary morpho- the loose material of a wide size range episodic, and sometimes catastrophic discharges which cause flooding of the fans.

Long-term development of fan deltas obviously reflects a wide range of processes but variations in sediment supply and in the morphoclimatic regime appear to be major controls (e.g., [6]). Among the factors that may have significant impact on fan delta construction, are, hence, the frequency of recurrence of exceptional river floods, mudflows and explosive eruptions (pyroclastic falls, surges and flows) from coastal volcanoes. All these processes can induce the supply of large volumes of loose or poorly consolidated sediment into the delta system and over vast areas of the continental shelf [2, 7, 8, 9].

524

Marine research at CNR

Figure 2: Shaded relief map of the Sorrento Peninsula and location of the fan delta systems fed by the main streams of the Amalfi coast. Location of some of Seistec profiles and gravity cores used in this study (see Figure 1 for location) that includes, bedrock river gravel, slopeweathering products, soil, and unconsolidated volcaniclastics deriving from the explosive activity of the Vesuvius and Campi Flegrei Volcanoes [12, 13, 8].

3

cludes multibeam bathymetry, Side Scan Sonar imagery, single-channel Sparker and Chirp-sonar profiles, acquired by the IAMC-CNR between 1997 and 2004.

The sequence stratigraphic nomenclature adopted for seismic interpretation is after [16](1992).

Material and methods

This study is based on integrated stratigraphy and correlation of gravity-cores allowed for a bed-to-bed calibration of seismic reflectors and interpretation of very high-resolution (IKB-Seistec), singlechannel reflection seismic survey carried out on the Amalfi inner shelf, between Salerno and Amalfi, in July 2004 (Figure 2) [14, 15]. The overall control for the stratigraphy and depositional setting of the late Quaternary depositional sequence comes from facies analysis of sediment cores (Figure 3), integrated biostratigraphic (Figure 4) and chronologic data (Figure 5), analysis of an extended dataset that in-

4 4.1

Data and results Gravity-core stratigraphy, 14C chronology and tephra layers

Gravity-cores C90, C106, C106 12 were collected on the outer shelf of Salerno Bay, between Capo dâ&#x20AC;&#x2122;Orso and Amalfi and provide a calibration of the entire last post glacial succession. Stratigraphic correlation of the gravity-cores from the sedimentological analysis is robust, thus allowing the construction of a composite stratigraphic section that represented the 525

Marine Geology

Figure 3: Lithology, textures, sedimentary structures, magnetic susceptibility, calcium carbonate content and facies associations of gravity cores C90 and C106 12. Grain size statistical parameters: Mz: mean; s: sorting; SK: skewness; KG: kurtosis. See text for more information on chronology and nature of event beds. base for the geological calibration adopted in the seismic stratigraphic interpretation.

The cored succession consists of c. 5 m thick transgressive sequence that overlies the major erosional surface (ES) associated with the sea-level drop and lowstand of the last glacial maximum. Below this unconformity, core C106 12 sampled a sandy silt succession older than 50 ka, within Upper Pleistocene deposits (Figures 6 and 7). The cored sequence is punctuated by at least ten tephra layers, we have labelled from top to bottom as tS1, tS1-α, tS1-β, tS1γ, tS2, tS3, tS3-α, and tS4 to tS6 (Figure 3). Tephra nomenclature is after Insinga et al. [17], Sacchi et al.[15] and Molisso et al. [14]. On the basis of sedimentological analysis and quantitative changes in benthic foraminiferal assemblages of the

526

core samples, three main lithofacies associations can be recognized (Figures 3 and 4). A composite stratigraphic section, from bottom to top consists of: a) Shelf mud with volcaniclasts and bioclasts, b) Shoreface sand and pebble, c) Bioturbated prodelta mud.

4.1.1

Facies A) - Shelf mud with volcaniclasts and bioclasts (Upper Pleistocene > c. 50 ka BP)

This facies is represented only in core C106 12 and displays a thickness of c. 110 cm. From base to top, it consists of poorly sorted olive gray (5Y4/2) clayey sandy silt and dark gray (5Y4/1) sandy clayey silt with very thin volcaniclastic lenses interbedded. At 612 cm and 580 cm bsf

Marine research at CNR

two pyroclastic layers occur, namely a 4 cm thick tephra (tS6) and a 9 cm thick tephra (tS5), respectively. The sandy fraction is represented by volcaniclasts and bioclasts and displays a gradual increase, especially in the volcaniclastic component towards the top (Figure 3). The benthic foraminifera assemblage (Figure 4) and the occurrence of epipelagic boreal guest (Limacina retroversa) suggest a cold period of the glacial Pleistocene. The uppermost 20 cm of unit A) is characterized by sparsely-branching burrows of ichnogenus Thalassinoides. Burrows have welldefined circular boundaries, with diameter ranging from 1 to 2 cm, and are passively infilled with the above lithofacies B).

This ichnofabric corresponds to the Glossifungites ichnofacies [18]. The unit is bounded at the top by an erosional surface that correlates with the unconformity (ES) of seismic profiles, and is characterized by the occurrence of shell debris. Glossifungites ichnofabric is commonly taken to indicate a firmground. In this context it likely represents colonisation of the Pleistocene eroded substrate during minor breaks in sedimentation following storm events, below storm wave base in the foreshoreoffshore transition zone, during the early TST. The lithofacies assemblage of this succession, coupled with the seismic evidence of thick Upper Pleistocene forestepping parasequences beneath unconformity (ES), suggests that this unit may be interpreted as a progradational deltaic sequence characterised by shelf mud deposit with volcaniclasts and bioclasts.

4.1.2

Facies B) – Shoreface sand and pebble (Uppermost Pleistocene c.

18.0 – 10.2 ka BP)

These deposits directly overlies facies A) and consist of a 70 cm thick unit in core C106 12. The grain-sizes of this unit range from silty sand to pebble, with very poor sorting. The lowermost 10 cm are represented by medium sandy pebble with inverse gradation that is bounded at the top by an erosive surface that correlates with the ravinement surface recognized on seismic profiles. Towards its top, the deposit is characterized by coarse-grained constituents represented by volcaniclasts, lithoclasts and bioclasts often fragmented.

Among bioclasts are bivalve and gastropod shell fragments, sometimes abraded, bioeroded and encrusted, solitary corals, bryozoans and echinoid fragments. Plant debris (Posidonia oceanica) are also found (Figure 3). The foraminifers assemblages (Figure 4) along with the occurrence of rhodolith-bearing pebbles and shells indicate a infralittoral to circalittoral zone associated with relatively low-salinity conditions. On the basis of lithofacies assemblages and the seismic-stratigraphic architecture this unit can be interpreted as a transgressive lag containing reworked material from the substrate and/or the early shoreface deposits (“healing phase deposits”) (e.g. [19]). A negative peak of the CaCO3 curve in the interval between 495 and 505 cm (C106 12), which suggests a relatively low water temperature, may be tentatively correlated with the Younger Dryas [20]. Analogously, the underlying interval between 505 and 525 cm (C106 12), marked by a positive peak of the CaCO3 , may be taken as corresponding to the B¨olling–Allerød event (Figures 3 and 4 and 7).

527

Marine Geology

528

Figure 4: Relative frequency (%) of selected benthic foraminifera and event beds of core C106 12. See text for more information on chronology and nature of event beds.

Marine research at CNR

Figure 5: Stratigraphic depth versus calibrated radiocarbon ages plots for gravity cores C90 (a) and C106 12 (b).

529

Marine Geology

Figure 6: Seistec profile 3006 showing the general architecture of the continental shelf off the Amalfi coast and the location of C106 12 core site. See the insert for location of profile.

Figure 7: Geological calibration and labelling of seismic reflectors of Seistec profile at C106 12 core site. See text and Figures 3 and 4 for more information on chronology and nature of event beds.

530

Marine research at CNR

4.1.3

Facies C) – Bioturbated prodelta that this lithofacies corresponds to prodelta mud (Uppermost Pleistocene – mud deposits associated with the modern Holocene < c. 10.2 ka BP) fan-delta system of the Amalfi shelf. According to the 14C calibrated age-depth Facies C) is represented in all the study plots we get interpolated ages that procores. It consists of a mud-supported vide chronologic reference for the recoglithofacies characterized by a gray to olive nized event beds and can be used for tengray homogeneous, bioturbated clayey silt tative correlation of the studied tephra laywith at least eight tephra (or cryptothepra) ers with age dated volcanic events onland and a few thin layers of fine-grained tur- (Figures 5 and 7). 14C AMS datings were bidites interbedded. The sandy fraction obtained on carbonate specimens of plankis rare and consists of fine to very fine tic foraminifera, mollusc shells and wood volcaniclasts, sub-rounded to sub-angular remains [15, 14]. Tephra layers tS1 to tS4 grey pumices, minerals, scoriae and glass, are interbedded within the last post-glacial and bioclasts, represented by fragments of succession, while tS5 and tS6 are interbedgastropod, bivalves, bryozoans and echi- ded within the Upper Pleistocene deposits noid. Turbidite layers are a few centimetres underlying unconformity ES ((Figures 6 thick and consist of volcaniclasts (rounded and 7). Most tephras have a sharp base, pumice, locally reddened scoriae, frag- normal or inverse grading, poor sorting and ments of minerals), subrounded lithics and typically a gradual transition to the overreworked bioclasts (fragments of bivalves, lying deposits. All cored tephras layers gastropods, bryozoan, phanerogamous sea- display a thickness in the order of 3-10 grass remains) (Figure 3). Facies C) marks cm, with exception of tephra tS2 that cora significant change in the relative abun- respond to a 80-100 cm thick pyroclastic dance of many benthic taxa, accompanied bed deposited during the AD 79 Plinian by the disappearance and/or abrupt de- eruption of the Vesuvius. This teprha can crease of some species and the parallel in- be subdivided into three major horizons, crease of others (Figure 4). Peaks of max- namely a) white pumice, b) gray pumice imum abundance of some taxa, along with and c) gray lapilli [2, 17] that correspond a relative maximum in the CaCO3 curve, to different phases during the AD 79 erupseem to concentrate around 450 cm, at a tion (Figures 3 and 7). stratigraphic horizon we interpret as the maximum flooding surface (mfs) (Figures 4.2 Seismic interpretation 3 and 4). The AD 79 Pompeii tephra bed (tS2) represents a marked environmental The interpretation of the IKB–Seistec change in the faunistic assemblage (Figure seismic profiles acquired off the Amalfi 4). In general, the benthic assemblage of coast was conducted according to sethis unit indicates a rapid increase in sed- quence stratigraphic principles. Termiimentation rate with respect to the under- nology adopted for sytems tracts is after lying facies B and a predominantly muddy [16], 1992. Core-to-seismics correlation setting, characterized by relatively high or- allowed for calibration of the major seganic matter and low oxygen concentra- quence stratigraphic surfaces on seismic tion, typical of the modern “mud belt” [21]. profiles (e.g. transgressive surface, ravineCorrelation with seismic profiles suggests ment surface, maximum flooding surface).

531

Marine Geology

Figure 8: Detail of Seistec profile 3012 showing the slumping of the Pompeii tephra bed (AD 79). See Figure 7 for key to reflectors labels.

Moreover, it indicated that tephra layers tend to correspond to high amplitude, well developed seismic reflectors. This is clearly the case, for instance, of the Vesuvius early Medieval tephra tS1-Îł, the tS2 (AD 79) pumice fall layer, the tS3 and tS3Îą, AP inter-Plinian deposits), and tephra tS4 (Astroni-Averno). Seistec profiles indicate that unconformity ES separates two main seismic stratigraphic units (Figures 6 and 7). The lower one is represented by a prograding succession truncated at the top by a dramatic erosional surface and mostly consists of Upper Pleistocene Forced Regressive Wedge Systems Tracts (FRWST) deposits. Above unconformity ES, seismic profiles show relatively continuous, parallel and sub-parallel reflectors, gently inclined towards SE, as result of low-angle backstepping and aggrading of layers. This unit is represented from bottom to top by the Transgressive Systems

532

Tract (TST) and Highstand Systems Tract (HST) deposits which formed in response to the time-transgressive landward shift of the coastline during the rapid sea-level rise that accompanied the last deglaciation (c.

18-6 ka). The thickness of the uppermost Pleistocene-Holocene shelf wedge varies in the study area from 35-40 m in the inner-mid shelf, to a minimum of 4-2 m at the shelf edge. The HST deposits of the southern shelf of the Sorrento Peninsula between Amalfi and Capo dâ&#x20AC;&#x2122;Orso, are characterized by the occurrence of a number of remarkably developed reflectors that can be correlated with different pyroclastic layers interbedded mostly within the upper Holocene succession. Seismic profiles show that the late HST (Upper Holocene) succession of the Amalfi shelf is characterized by a number of small prograding deltas that develop at the mouth of the small bedrock rivers with torrential regime.

Marine research at CNR

Figure 9: Detail of Seistec profile 5006 off the Canneto stream, Amalfi, showing approximate correlation of prograding subaqueous delta units with major climatic changes of the last 2000 years. See insert and Figure 2 for location.

The best developed deltaic wedges in the study area occur offshore Maiori at the mouth of the Reginna Major torrent off the village of Minori at the mouth of the Reginna Minor at the mouth of Canneto, Dragone and Cappuccini torrents (Figure 2). These deltaic bodies represent the subaqueous components of the confined alluvial fans that developed in the narrow coastal plain and pocket beaches of the Amalfi coast. Seistec profile 3012 (Figure 8) documents the occurrence of sedimentary structures at the base of the Reginna Major delta front that may be associated with a general gravity-driven instability and soft sediment deformation above distinct stratigraphic surfaces, namely represented by reflector H (base of AD 79 pumice layer) and reflector I. Minor, but still clear evidence of seafloor instability can be recognized above reflectors L and E (Figure 8). Seismic interpretation suggests that soft sediment deformation above the

base of tephra tS2, mostly involves the pyroclastic layer itself and consists of slumpslide folding and slump-fault rupture of the tephra layer (Figure 8).

5

Discussion and Conclusion

The IKB-Seistec seismic reflection profiles and gravity-core data used in this study have revealed unprecedented detailed views of the inner-mid shelf depositional system of the northern Salerno Bay, allowing for the recognition of a number of fan-deltas that developed mostly during the last two thousand years, at the mouth of small rivers of the Amalfi cliffed coast. During this time interval of c. 2000 yrs, both sea-level oscillation and tectonic subsidence/uplift were practically negligible in terms of influence on the overall stratigraphic architecture of the inner shelf 533

Marine Geology

system ([22] and references therein) and the main factor controlling stratal geometries and patterns, were likely the rates and modes of sediment supply. The deltaic bodies imaged by seismic interpretation represent the underwater counterparts of coastal alluvial fans fed by small bedrock rivers with torrential regime that are part of the hydrographic network of the Sorrento Peninsula. In this context subaerial delta-plain components are practically absent and the narrow space at the exit of the valleys is filled up with a coarse-grained alluvial prism up to a few tens of metres thick, whereas at the seashore the alluvial deposits are reworked into pocketbeach settings. Most of the stratigraphic sequence imaged by seismic profiles is represented by HST deposits and the general stratigraphic architecture of the studied fan-deltas is of two main types. The deltaic bodies off Maiori and Minori display relatively steep and long sigmoidal foresets, commonly associated with topset layers.

The fan-deltas off the Amalfi-Atrani coast show a clear variation in the dip angle of foresets and a very reduced or absent topset towards the delta fronts. All the fan-deltas described in this study started to develop above the pyroclastic bed deposited by the Vesuvius during the â&#x20AC;&#x153;Pompeiiâ&#x20AC;? Plinian eruption of AD 79 which corresponds to a major downlap surface. A significant change in the stratal architecture of the fan-deltas occurred after another eruption of the Vesuvius, during the early Medieval period (c. AD 512-685).

This is documented by the development of another remarkable downlap surface that can be correlated with the oldest Medieval products of the Vesuvius preserved in the Salerno Bay (tephra layer tS1-Îł). Seismic reflector (E) that correlates with these products consistently separates the fan-deltas 534

of all the study areas into two sub-units showing distinct stratal patterns (Figure 9).

Seistec profiles reveal evidence of gravitydriven instability at various stratigraphic levels within the fan-deltas. Likely, more than one mechanism of sediment deformation or failure is behind the variety of the features described in the fan-deltas of the Amalfi coast. The available data allow for the recognition of (1) crenulation in muddominated prodelta slopes, possibly associated with shear deformation of sediments by creeping, (2) slide/slump deformation of the AD 79 pyroclastic deposits (3) gravity (inertia/turbidity/debris) flow deposits associated with selected stratigraphic intervals. Seismic stratigraphic interpretation showed that most of the gravity-driven instability processes are not diffused across the fan-delta but indeed concentrated on a few stratigraphic horizons that invariably correspond to major tephra layers or tephra clusters. This observation, coupled with the recognition that the sediment supply to the fan-delta system is largely affected by high-energy river floodings suggests a direct relationship between the rates of erosion of the river basin slopes that follow deposition of landscape-mantling volcaniclastic deposits and the rates of underwater sediment that are delivered to the fan-deltas. The interpretation proposed in this study, implies that the growth rates of the fan-deltas of the Amalfi coast were primarily controlled by the average recurrence and magnitude of river flooding episodes that have provided conspicuous sediment yields to the delta system, concomitant with periods during which abundant, erosion-prone (volcani)clastic material was available on the slopes of the feeding river basins. Accordingly it may be proposed that the amount of sediments delivered to the coastline and hence the rates

Marine research at CNR

of development of the Amalfi fan-deltas in the last 2000 years were possibly dictated by the interplay of the availability of loose pyroclastic covers on the slopes of the alluvial basins, on one hand, and on the other by the varying erosional rates on the slopes due to the climatic oscillations that have occurred during the last millennia. The illustration of the fan delta developing at the mouth of the Canneto stream (Amalfi) (Figure 9), shows an attempt of correlation between the seismic stratigraphic framework described in this study and the major climatic fluctuations of the last millennia (e.g. [23]). This tentative correlation, derived from mere chronologic basis, sug-

gests that the major change detectable in the stratal geometries of the fan-deltas occurring in the early Medieval time (tephra tS1-γ) may be associated with the onset of a period of climatic cooling, known as Early Medieval cool period (c. AD 500AD 800), that developed immediately after the Roman Warm period. Similarly, further minor changes in the stratal patterns of the delta foresets, that are consistently imaged by the seismic record in all the individual fan-deltas of the Amalfi coast may be correlated, in turn, with the Medieval Warm Period (c. AD 900-AD 1100) and the Little Ice Age (c. AD 1400-AD 1850).

References [1] F.J. Lobo, L.M. Fern´andez-Salas, I. Moreno, J.L. Sanz, et al. The sea-floor morphology of a Mediterranean shelf fed by small rivers, northern Alboran Sea margin.

Continental Shelf Research, 26:2607–2628, 2006.

[2] M. Sacchi, D. Insinga, A. Milia, F. Molisso, et al. Stratigraphic signature of the Vesuvius 79 AD event off the Sarno prodelta system, Naples Bay. Marine Geology, 222-223:443–469, 2005.

[3] T.S. McConnico and N. Bassett Kari. Gravelly Gilbert-type fan-delta on the Conway Coast, New Zealand: Foreset depositional processes and clast imbrications.

Sedimentary Geology, 198:147–166, 2007.

[4] L.M. Fern´andez-Salas, F.J. Lobo, F.J. Hern´andez-Molina, et al. High-resolution architecture of late Holocene highstand Prodeltaic deposits from southern Spain: the imprint of high-frequency climatic and relative sea-level changes. Continental Shelf Research, 23:1037–1054, 2003.

[5] T. Hasiotis, M. Charalampakis, A. Stefatos, G. Papatheodorou, et al. Fan-delta development and processes offshore a seasonal river in a seismically active region, NW Gulf of Corinth. Geo-Marine Letters, 26:199–211, 2006.

[6] A. Colella and B.D. Prior, (Eds.). Coarse-grained deltas. International Association of Sedimentologists. Special Publication, 10:357 pp., 1990.

[7] R. Sulpizio, G. Zanchetta, F. Demi, M. Di Vito, et al. The Holocene syneruptive volcaniclastic debris flowin theVesuvian area: geological data as a guide for haz-

535

Marine Geology

ard assessment. In: Siebe, C.and Macias, J.L. and Aguirre-Diaz, G.J. (Eds.) Neogene–Quaternary continental margin volcanism: a perspective from Mexico. Geol.

Soc. Am. Special Paper, 402:203–221, 2006.

[8] M. Bisson, M.T. Pareschi, G. Zanchetta, R. Sulpizio, et al. Volcaniclastic debrisflow occurrences in the Campania region (Southern Italy) and their relation to Holocene–Late Pleistocene pyroclastic fall deposits: implications for large-scale hazard mapping. Bulletin of Volcanology, 70:157–167, 2007.

[9] E. Esposito, G. Foscari, F. Molisso, S. Porfido, M. Sacchi, et al. Flood risk estimation through document sources analysis: the case of the Amalfi Rocky Coast. This volume, 2010.

[10] E. Esposito, S. Porfido, and C., (Eds.) Violante. Il nubifragio dell’ottobre 1954 a Vietri sul Mare-Costa di Amalfi Salerno. CNR GNDCI, 2870:381, 2004.

[11] C. Liquete, P. Arnau, M. Canals, and S. Colas. Mediterranean river systems of Andalusia, southern Spain, and associated deltas: a source to sink approach. Marine Geology, 222–223:471–495, 2005.

[12] H. Sigurdsson, S. Carey, W. Cornell, and T. Pescatore. The eruption of Vesuvius in AD 79. National Geographic Research, 1:332–387, 1985.

[13] R. Sulpizio, G. Zanchetta, F. Demi, M. Di Vito, et al. The Holocene syneruptive volcaniclastic debris flowin theVesuvian area: geological data as a guide for hazard assessment. Geol. Soc. Am. Special Paper, 402:203–221, 2006.

[14] F. Molisso, D. Insinga, F. Marzaioli, M. Sacchi, et al. Radiocarbon dating versus volcanic event stratigraphy: age modelling of Quaternary marine sequences in the coastal region of the Eastern Tyrrhenian Sea. Nucl. Instr. and Meth. in Phys. Res. B., 268(7-8):1236–1240, 2010.

[15] M. Sacchi, F. Molisso, C. Violante, E. Esposito, et al. Insights into flood-dominated fan-deltas: very high-resolution seismic examples off the Amalfi cliffed coasts, eastern Tyrrhenian Sea. The Geological Society London, Spec. Publ., 322:33–71, 2009.

[16] D. Hunt and M. E. Tucker. Stranded parasequences and the forced regressive wedge systems tract: deposition during base-level fall. Sedimentary Geology, 81:1–9, 1992.

[17] D. Insinga, F. Molisso, C. Lubritto, M. Sacchi, I. Passariello, and V. Morra. The proximal marine record of Somma–Vesuvius volcanic activity in the Naples and Salerno bays (eastern Tyrrhenian Sea) during the last 3 kyrs. Journal of Volcanology and Geothermal Research, 177:170–186, 2008.

536

Marine research at CNR

[18] M.K Gingras, S.G Pemberton, and T. Saundners. Bathymetry, sediment texture and substrate cohesiveness their impact on modern Glossifungites trace assemblages at Willapa Bay, Washington. Palaeogeogr. Palaeoclimatol. Palaeoecol., 169:1–21, 2001.

[19] H.W. Posamentier and G.P. Allen. Variability of the sequence stratigraphic model: effects of local basin factors. Sedimentary Geology, 86:91–109, 1993.

[20] N. Kallel, M. Paterne, L. Labeyrie, J.C. Duplessy, and M. Arnold. Temperature and salinity records of the Tyrrhenian Sea during the last 18000 years. Paleogeography, Paleoclimatology, Palaeoecology,, 135:97–108, 1997.

[21] F. J. Jorissen. Benthic foraminifera from the Adriatic Sea principles of phenotypic variation. Utrecth Micropaleontol. Bull., 37:176, 1988.

[22] C. Caiazzo, A. Ascione, and A. Cinque. Late Tertiary–Quaternary tectonics of the Southern Apennines (Italy): New evidences from the Tyrrhenian slope. Tectonophysics, 421:23–51, 2006.

[23] R. Br´azdil, C. Pfister, H. Wanner, H. von Storch, and J. Luterbacher. Historical climatology in Europe–the state of the art. Climatic Change, 70:363–430, 2005.

537

Marine Geology

538

The Coastal Depositional Systems along the Campania Continental Margin (Italy, Southern Tyrrhenian Sea) since the Late Pleistocene: New Information Gathered in the Frame of the CARG Project F. Budillon1 , G. Aiello1 , A. Conforti1 , B. Dâ&#x20AC;&#x2122;Argenio1 , L. Ferraro1 , E. Marsella1 , L. Monti2 , N. Pelosi1 , R. Tonielli1 1, Institute for Coastal Marine Environment, CNR, Napoli, Italy 2, Regione Campania, Settore Geotecnica, Geotermia e Difesa del Suolo, Napoli, Italy francesca.budillon@cnr.it Abstract Extensive high-resolution mapping of the continental margin off the southern Campania region (eastern Tyrrhenian Sea) has revealed significant morphological and geological features which allowed us to outline the gradual modification of the coastal domains since the Late Pleistocene. Swath bathymetry, acoustic images of the seafloor, seismic acquisition, core and bottom samples were used to implement a large database.

Shore bodies ranging in age from pre- to post- last-glacial times have been identified in the uppermost 100 ms of the seismic stratigraphic record off the Sele and Bussento river mouths. The oldest bodies formed during the seaward retreat of the shoreline during the Late Pleistocene sea level drop. The peak of the retreat accounts for the growth of a shelf-margin, in the Salerno and Policastro Bays, and a mid-shelf, off Cilento, littoral body, at least 100 km long, during the last maximum lowstand phase. At that time, the River Sele flowed directly on the upper slope and formed a channel drainage system, still preserved between a depth of 180 and 500 m, due to density flows that have transferred sediment from the coastal area directly into the Salerno Valley, an intraslope basin of the Eastern Tyrrhenian Margin. The postglacial sea level rise caused the fast drowning of the shelf and a partial preservation of the transgressive deposits. However a prograding wedge 1.5 km long and about 10 ms thick, which lies above the transgressive surface (90/60 m below the present day sea level), could represent a trace of the Younger Dryas climatic event. The rapid shore progradation during the â&#x2C6;ź12 ky B.P. cold event testifies the sensitivity of the Sele coastal system even to minor climatic oscillations.

1

Introduction

The geological mapping Project of the Campania offshore (CARG project) allowed us to gather new geological (Chirp sonar Subbottom and Uniboom profiles,

gravity cores and grabs) and morphological (Swath-bathymetric soundings and Sidescan Sonar images of seafloor) information, valuable to create maps of the seafloor at scales of 1:10000 and 1:25000, down to the 200 m isobath. The project was commis-

Marine Geology

sioned to the IAMC- CNR by the Regione Campania (Settore Geotecnica, Geotermia e Difesa Suolo) about 8 years ago, to be realized according to the steering lines of the Italian Geological Survey, now ISPRA, Institute for Environmental Protection and Research.

Mapping criteria focussed mainly on the physiographic features of the shelf/slope sector, the lithology and textures of sediment at the seafloor, the stratigraphic stacking pattern within the Late Quaternary depositional sequence (SDTQ).

This project has been a valuable opportunity to investigate the morphology and stratigraphy and to map the southern Campania offshore thoroughly, implementing the high resolution geological and morphological data set of the CNR-IAMC. This report summarizes the main outcome of the surveys.

2

Geological setting

The surveyed marine area is about 1500 km2 and falls within four sheets of the National Geological Map of Italy (n. 486 Foce del Sele, n. 502 Agropoli, n. 519 Palinuro, n. 520 Sapri, Figure 1). It pertains to the Tyrrhenian side of the Southern Apennine range, which has been forming since the Late Pliocene- Early Pleistocene, along with the opening of the Marsili basin, within the Tyrrhenian Sea back-arc basin [1]. The Tyrrhenian border of the chain achieved most of its actual configuration by the Early Pleistocene, when NE-SW ori-

540

ented alternating structural depressions and highs began to delineate. A set of lowangle, south-east verging faults, and associated minor faults, led to the extensional deformation of the Meso-Cenozoic carbonate basement and the Miocene nappes and the formation of asymmetrical halfgrabens [2, 3]. Throughout the Mid- and Late Pleistocene the extensional regime produced high-angle NW-SE oriented normal faults and trans-tensional faults that enhanced the vertical displacement of the margin of the chain with respect of its axial portion [4], [3]. A thick syn-tectonic Pleistocene sedimentary succession made of continental, volcanoclastic and marine deposits filled the structural depressions [5, 6]. The Sele Plain-Salerno Gulf halfgraben is a ENE-WSW oriented structural depression, which displays up to 2400 ms of post-orogenic sediment infill [5] and about 1000 m of Pleistocene sediment documented by offshore well stratigraphy. Seismic reflectorsâ&#x20AC;&#x2122; geometries show local evidences of transpressive deformation and tectonic inversion due to the effect of strike-slip faulting [7]. The deepest portion of the gulf, the Salerno Valley, sits at the foot of the Sorrento-Amalfi Peninsula and exceeds a depth of 1000 m to the south of Capri Island [8].

The shelf areas of the Salerno Gulf contain the seaward front of the Sele, Tusciano, Picentino and Solofrone alluvial plains which prograded seaward by about 15 km starting from the MIS 5a and following the general retreat of the sea level during the Late Pleistocene [9, 10], (Figure 2).

Marine research at CNR

Figure 1: The continental margin off southern Campania region, between Salerno and Sapri, extensively surveyed by IAMC â&#x20AC;&#x201C;CNR since the 2003, to accomplish the CARG project aims (bold numbers refer to the geological maps of Carta Geologica dâ&#x20AC;&#x2122;Italia, scala 1:50000). Submarine topography is by swath bathymetry. 1) submerged beach; 2) continental shelf; 3) upper continental slope; 4) intraslope ridges; 5) acoustic substratum; 6) Infreschi canyon-fan system; 7) slope failure; 8) dismantling slope areas ; 9) intraslope basin; 10) gas charged sediment; 11) sediment waves field; 12 ) water escape features and plastic deformation of subbottom reflectors; 13) outer shelf marine erosional surface; 14) paleo-drainage system; 15) physiographic shelf margin; 16) buried lowstand deposits; 17) relic features; 18) withdrawing slope; 19) last glacial sea level terrace; 20) furrows.

541

Marine Geology

Figure 2: Schematic line drawing of Uniboom line off the River Sele, showing the regressive trend of shore deposits during Late Pleistocene sea level fall (modified after [9]).

The offshore area included in the Agropoli and Palinuro sheets is the result of the Pleistocene evolution of the Cilento margin. The shallow position of the acoustic basement, and therefore the scarce accommodation space for sediment deposition [11], accounts for the formation of seaward prograding wedges, bounded by marine and subaerial erosive unconformities [12]. The basement consists of siliciclastic sequences pertaining to the internal Ligurides and Cilento Group units and outcrops mainly along the Cilento coast [13, 14] and across the shelf [7].

The acoustic substratum outcrops off Mt. Bulgheria coast (Palinuro and Sapri sheets)

542

consist of carbonate rocks, Upper Triassic to Lower Miocene in age, unconformably covered by Lower Miocene siliciclastic deposits [3]. On land this succession forms a N-verging fold thrust over the internal Ligurides nappe [15], cut by a complex pattern of faults. A major NEâ&#x20AC;&#x201C; SW trending fault borders Mt. Bulgheria towards the Gulf of Policastro graben, where the Meso-Cenozoic basement is downthrown to about 3000 m below the sea level [16].

The shelf area (about 1200 km2 ) between Salerno and Sapri underwent drowning, following the Holocene transgression [9, 17, 10, 18].

Marine research at CNR

543

Figure 3: Subbottom chirp line across the Salerno Bay shelf and slope (from [10], modified).

Marine Geology

3

The shelf

The continental shelf widens roughly following the coastline, while it enlarges by more than 23 km seaward off the Licosa and Palinuro promontories (Cilento coast) (Figure 1). Off the Sele and Bussento alluvial plains loose and smooth seabed prevails, whereas off the Cilento coast, the rocky substratum largely controls the morphological features and the seafloor lithology. Two rocky ridges crop out off Licosa Cape and Acciaroli, respectively E-W and N-S oriented (Figure 1). Several orders of terraces shape the seabed at various depth intervals: 160/140 m, 44/46 m, 18/24 m, 12/14 and 7/8 m [19, 20, 11]. The last three surfaces have been largely reported by scuba dive surveys and have been tentatively related to Tyrrhenian sea-level stages [20].

The shore belt is affected by both erosive and depositional processes, due to wave action and alongshore currents and is generally characterized by the occurrence of well sorted sand deposits. They extend down to a water depth of 10-12 m and include the outermost sand bars. Three different types of submerged beach can be distinguished along the Sele shore, based on the topographic profile: shoal bar, bartrough and mixed types with cells less than 100 m wide [21]. The Alento submerged shore displays a bar-trough system and the Marina di Ascea shore a shoal bar [22]. These features are controlled by the coastal morpho-dynamics and therefore may change rapidly. Sediments consist of coarse to medium, well sorted sand down to 3 m and, medium to fine, wellsorted fine sand down to10-12 m. The fine-grained fraction is less than 20 Finegrained, poorly sorted deposits occur between the outer limit of the submerged 544

shore and water depth of 40-50 m and form the inner shelf depositional system where sandy pelite lithofacies prevails. The fine-grained fraction increases beneath the Cymodocea nodosa meadows which trap the loose sediment at the seabed and with Caulerpa racemosa, recently introduced in the Tyrrhenian Sea [23]. Posidonia oceanica meadows occur down to 25 m in sectors where terrigenous supply is scarce.

The seismic acoustic profiles off the main river mouths show beneath the seabed a shallow unit with fluid escape and plastic deformation features (Figure 3). This unit is bounded at the base by a regular and conformable reflector lying halfway between the 79 A.D. Vesuvius tephra and the present-day seabed, between 40 and 70 m bsl (Figure 3). The unit, which lies seaward of shallow biogenic gas pockets can be associated to the estuarine depositional environment and possibly marks the boundary between the silty and the muddy prodelta system.

The outer shelf environment ranges between 40-50 m bsl and the shelf break.

Fine grained textures prevail, however a variable but valuable fraction of fine sand, mostly pumices, scoria and bioclasts, is common at the seabed where sediment waves and terraced areas occur (Figures 1 and 2). Conversely, in the same depth range off the Cilento Promontory authigenic bioclastic coarse sand and gravel lies eteropically to siliciclastic deposit. Shell fragments, bioclasts and rhodholits form the coarse fraction that largely tapers the rocky seabed and the terraced surfaces.

The bioclastic and organogenic coarse sand and fine gravel (maerl facies) pertain to the â&#x20AC;&#x153;coastal detritus assemblageâ&#x20AC;? Auct., and form decimetre thick patches at a water depth between 25 and 70 m; this lithofacies changes seaward into mud supported

Marine research at CNR

Figure 4: Subbottom chirp line across the Cilento shelf and slope. organogenic gravel and bioclastic sand typical of the “muddy coastal detrital assemblage” and may occur down to a depth of 90 m. Organogenic mounds, made of fouling organisms, encrust the travertine outcrops in the Salerno Bay and build up pinnacles on the rocky seabed off Acciaroli and Palinuro offshore.

The outermost sector of the shelf, south of the River Sele to the Acciaroli Cape, corresponds to a morpho-structural high that lacks a modern terrigenous supply from the mainland. High resolution seismic sections show that the most shallow units pertain to the distal marine segment of the wedge which prograded seaward during the Late Pleistocene sea level fall and lowstand phases. They lie conformably on marine units possibly Middle Pleistocene in age (Figure 3). A set of furrows scratched into the seabed parallel to the isobaths could be due to present day seabed currents which border the shelf margin.

Pleistocene relic morphologies outcrop at the seabed along the Cilento shelf (Figures 1 and 4) and consist of large irregular reliefs in the first 80 m of depth or regu-

lar ridges sub-parallel to the isobaths and overlying the prograding reflectors of the outer shelf (Figure 1). Sediment texture consists of well sorted, coarse to medium sand and contains Arctica islandica shells and corals. These morphologies have been interpreted as relic ridges and bars formed during the maximum glacial lowstand [19, 12, 11].

The most evident morphological element of the Campania margin is the physiographic edge of the shelf, which occurs between 100 m and 230 m bsl Figure 1; beyond it, lies the slope sector with a gradient of more than 1.5°. The varying extension, the different depth and gradient of the shelf could be accounted for by the stacking pattern of the systems tract pertaining to the Middle - Late Pleistocene depositional sequences, which allowed this sector of the Campania margin to expand. Indeed Sparker seismic profiles in the Cilento offshore show a deep stratigraphic unconformity below the Tyrrhenian marine unit, which develops down to a depth of 180 m and is largely affected by sub-vertical faulting [11], therefore possibly pertaining to

545

Marine Geology

Figure 5: Relict channel system at the paleo-mouth of the Rivers Sele and Solofrone/Capo di Fiume, carved during the last lowstand phase of sea level. the Middle Pleistocene lowstand phase.

4

The slope

Slope morphologies, with gradients of between 1.3째 and 6째 are quite uneven and include large erosive sectors shaped by slides or deeply engraved by gullies and canyon heads, alternating to slope ridges bordered by structural lineaments (Figure 1). Dismantling slope sectors are evident off Salerno and Sapri, where erosional processes produced a dense network of downstream gullies with herringbone patterns and caused the shelf retreat (Figure 5).

The slope sector which forms the southern boundary of the Salerno Valley is shaped by several slide headscarps, whose deposits reached great distances due to the high gradient of the slope. Off the Cilento coast, slide scars are confined to the upper slope 546

and relative deposits accumulated into the elongated depocentres which are bounded seaward by intra-slope ridges [24, 25].

Seaward of the 180 m isobath (where the HST wedge thins out) a paleo-channel system, possibly engraved by the outflows of the Sele, Solofrone/Capo di Fiume and Alento rivers, is still preserved. Indeed, during the lowstand stage of sea level these rivers flowed directly at the shelf edge forming small shelf-margin deltas and sediment underflows were transferred along the upper slope down to the Salerno Valley and Cilento offshore intraslope basin (Figures 1 and 5). A wavy unit developed at the bend of the upper slope in the Salerno Gulf (Figures 1 and 3). The geometries of the wavy reflectors and the location of the unit point to sediment drift structures and thus possibly to depositional processes related to bottom current dynamics [26]. It has been observed that the strati-

Marine research at CNR

graphic discontinuity within this unit may cause sediment failures along weak layers.

A slope canyon-fan system with meandering thalwegs, levees, overflowing channels and fan lobes has been developing on the southern slope of Monte Bulgheria (Figure 1), possibly being fed by coastal sediment drift alongshore (Budillon et al., in press).

5

Depositional environments within the SDTQ

Several key elements for a sequence stratigraphic rendering were recognized and mapped within the SDTQ, from the coastline down to the water depth of 250 m, based on submarine morphology interpretation, sediment core stratigraphy, high resolution seismic records and bottom samples analysis. In particular, stillforming and completely-formed units were distinguished, according to analogous inland mapping criteria and steering lines of Servizio Geologico dâ&#x20AC;&#x2122;Italia. The first corresponds to the highstand systems tract (hst), which has been developing since 56 kyr BP, while the second represents the transgressive, lowstand and falling-stage systems tracts (TST, LST and FST) [27], stacked following the Late Pleistocene â&#x20AC;&#x201C; Holocene sea level variations. Specifically, three still-forming depositional environments (beach, shelf and upper slope), two completely-formed depositional environments (relic shelf and relic slope) and one completely-formed unit, continental in origin, (a relic tabular plate of travertine, correlated with the inland outcrop of Travertini di Paestum) have been mapped. The most represented unit within the SDTQ is the highstand systems tract. It includes, landwards, the present-day coastal system

and typically consists of a tapering seaward wedge. The maximum thickness is reached on the inner shelf off the river mouths (Figure 3). In the Salerno Bay the HST depocentre is located off the Sele River 40 m deep, where its thickness exceeds 10 m and rapidly thins out seaward to less than 1 m, about 20 km off the coast (Figure 6). It therefore goes beyond the shelfbreak in the northern sector, while it tapers in the southern part of the bay at about 160 m, onto the relict outer shelf off Licosa Cape, due to the low sediment supply. Pre- and postglacial shore units, featuring prograding geometries with offlap terminations, were identified off the Sele and Bussento river mouths. The oldest ones (Figure 2), lying below the maximum glacial unconformity, formed as a consequence of the seaward retreat of the shoreline during the last stages of the Late Pleistocene sea level drop and therefore show a regressive trend [9]. The regression of the shore system culminated, as largely reported at a global scale, during the 20-18 ky lowstand stage, which accounts for the prograding units commonly stacking at the shelf margin. Nevertheless the depositional terrace linked to the last glacial maximum is not always evident, since shore deposits relative to this phase are preserved in morphological steps (green dashed line, Figure 1) [12]. A midshelf prograding wedge marks the lowstand stage of the last glacial peak off the Agropoli, Licosa and Acciaroli coasts. On the whole, relic deposits and coastal morphologies related to the last glacial maximum form a discontinuous belt, cropping out or draped by Holocene sediments, that can be followed for more than 100 km along the southern Campania margin (Figure 1). The post-glacial sea level rise resulted in a rapid drowning of the shelf, with a limited preservation of the transgressive 547

Marine Geology

Figure 6: Schematic block diagram of the main stacking pattern and morphologies in the Salerno Bay (from [10] modified). units. Transgressive lithosomes are poorly represented and resolvable due to their low thickness. However, a shore system, 1.5 km wide and 5-10 m thick (Figures 3 and 6), which lies above the transgressive surface, 90/60 m below the sea level, could be the remains of the Younger Dryas climatic event [28]. These bodies consist of a continuous set of prograding reflectors, with offlap and downlap lateral terminations, overlain by onlapping sub-horizontal reflectors, topped in turn by the maximum flooding surface. Other lithosomes relative to the transgressive systems tract occur seaward to the shelf break and represent the healing phase of postglacial transgression [27]; besides they occur in morphological steps on the shelf between the transgressive and the ravinement surfaces, showing acoustic facies typical of transitional shore deposits.

548

6

Conclusions

The CARG Project allowed the acquisition of regularly spaced, high resolution data set valuable for geological, morphological and cartographic purposes. The large amount of data led to the redaction of four geological sheets in the Southern Campania marine area down to the upper slope environment, yet published [10] or in press.

In particular, within the Late Pleistocene Depositional Sequence, still-forming units and completely-formed units were distinguished. The first ones correspond to the highstand wedge (HST), which has been developing since 5-6 kyr BP, while the second ones represent the transgressive, lowstand and falling-stage systems tracts, stacked following the Late Pleistocene â&#x20AC;&#x201C; Holocene sea level variations. Pre- and post-glacial shore units, featuring prograding geometries with offlap terminations, were identified off the Sele and Bussento

Marine research at CNR

river mouths. The one lying above the transgressive surface, 90/60 m below the sea level, could be the effect of the Younger Dryas climatic event. A paleo-channel system along the upper slope indicates the position of Sele, Solofrone/Capo di Fiume and Alento river mouths during the glacial

maximum retreat of the sea level and testify the density flows passage down to the intraslope basins. However, if the main target has been achieved by mapping the seabed features and lithologies, a large number of further scientific questions remain, to be addressed at a later date.

References [1] M. Marani and F. Gamberi. Structural framework of the Tyrrhenian Sea unveiled by seafloor morphology. Mem. Descr. Carta Geol. D’It., LXIV:97–108, 2004.

[2] A. Milia and M.M. Torrente. Tectonics and stratigraphic architecture of a periTyrrhenian half-graben (Bay of Naples, Italy). Tectonophysics, 315:301–318, 1999.

[3] C. Caiazzo, A. Ascione, and A. Cinque. Late Tertiary-Quaternary tectonics of the Southern Apennines (Italy): new evidences from the Tyrrhenian slope. Tectonophysics, 421:23–51, 2006.

[4] E. Casciello, M. Cesarano, and G. Pappone. Extensional detachment faulting on the Tyrrhenian margin of the Southern Apennines contractional belt (Italy). Journal of the Geological Society, London, 163:617–629, 2006.

[5] R. Bartole, D. Savelli, M. Tramontana, and F.C. Wezel. Structural and sedimentary features in the Tyrrhenian margin off Campania, southern Italy. Marine Geology, (55):163–180, 1984.

[6] L. Brancaccio, A. Cinque, P. Romano, and C. Rosskopf. Geomorphology and neotectonic evolution of a sector of the Tyrrhenian flank of the Southern Apennines, (Region of Naples, Italy). Zeit. Geomorph., 82:47–58, 1991.

[7] M. Sacchi, S. Infuso, and E. Marsella. Late Pliocene-Early Pleistocene compressional tectonicsm offshore Campania (Eastern Tyrrhenian Sea). Boll. Geof. Teor. ed Appl., 36(141-144), 1994.

[8] G. Aiello, E. Marsella, V. Di Fiore, and C. D’Isanto. Stratigraphic and structural styles of half-graben offshore basins in Southern Italy: multichannel seismic and Multibeam morpho-bathymetric evidences on the Salerno Valley (Southern Campania continental margin, Italy). Quaderni di Geofisica, (77):1–33, 2009.

[9] F. Budillon, T. Pescatore, and M.R. Senatore. Cicli deposizionali del Pleistocene superiore-Olocene sulla piattaforma continentale del Golfo di Salerno (Tirreno meridionale). Boll. Soc. Geol. It., 113:303–316, 1994.

[10] Carta Geologica d’Italia, scala 1:50000, Foglio 486 Foce del Sele. System Cart, Roma, 2009.

549

Marine Geology

[11] L. Ferraro, T. Pescatore, B. Russo, M.R. Senatore, et al. Studio di Geologia Marina del Margine Tirrenico: la piattaforma continentale tra P.ta licosa e Capo Palinuro (Tirreno meridionale). Boll. Soc. Geol. It., 116:473–485, 1997.

[12] F. Trincardi and M.E. Field. Geometry, lateral variation and preservation of downlapping regressive shelf deposits: Eastern Tyrrhenian Sea Margin, Italy. J. of Sedimentary Petrology, 61(5):775–790, 1991.

[13] G. Bonardi, O. Amore, G. Ciampo, P. de Capoa, et al. Il complesso Liguride Auct.: stato delle conoscenze e problemi aperti sulla sua evoluzione pre-appenninica ed I suoi rapporti con l’Arco calabro. Mem. Soc. Geol. Ital., 41:17–35, 1988.

[14] G. Bonardi, S. Ciarcia, S. Di Nocera, F. Matano, et al. Carta delle principali Unit`a cinematiche dell’Appennino meridionale. Boll. Soc. Geol. It., 128:47–60, 2009.

[15] M. Tozzi, A. Menconi, and S. Sciamanna. Studio strutturale del M. Bulgheria (Cilento meridionale) e sue implicazioni per la tettogenesi dell’Appennino campano. Boll. Soc. Geol. Ital., 115:249–278, 1996.

[16] G. Bigi, D. Cosentino, M. Parotto, et al. Structural Model of Italy, scale 1:500,000, sheet 4. CNR Progetto Finalizzato Geodinamica, Selca, Firenze, 1992.

[17] G. Buccheri, G. Capretto, V. Di Donato, P. Esposito, et al. A high resolution record of the last deglaciation in the southern Tyrrhenian Sea: environmental and climatic evolution. Marine Geology, 186:447–470, 2002.

[18] M. Iorio, J. Liddicoat, F. Budillon, P. Tiano, et al. Paleomagnetic secular variation time constraints on late Neogene geological events in slope sediment from the eastern Tyrrhenian Sea. In Kneller, et. al (eds.) External Controls on Deepwater Depositional Systems, SEPM. (92):33–243, 2009.

[19] M.G. Coppa, M. Madonna, M. Putignano, P. Russo, et al. Elementi geomorfologici e faunistici del margine continentale tirrenico tra P.ta Campanella e P.ta degli Infreschi (Golfo di Salerno). Soc. Geol. It., Atti del74° Congresso, pages 203–207, 1988.

[20] F. Antonioli, A. Cinque, L. Ferranti, and P. Romano. Emerged and submerged marine terraces of Palinuro Cape (Southern Italy). Mem. Desc. Carta Geol. D’Italia, LII:237–260, 1994.

[21] E. Cocco and S. Iuliano. L’erosione della fascia costiera tra Foce Sele e Paestum (Salerno): dinamica evolutiva ed ipotesi di intervento a difesa e tutela della spiaggia e della pineta litoranea. Il Quaternario, 12(2):125–140, 1999.

[22] E. Cocco and F. Musella. Variazioni della linea di riva e dinamica dei sedimenti fra Marina di Casalvelino e Marina di Ascea (Cilento, Campania). Atti del XII Congresso AIOL (Isola di Vulcano, 18-21 settembre 1996), II:341–350, 1998.

550

Marine research at CNR

[23] M.C. Gambi and A. Terlizzi. Record of a large population of Caulerpa racemosa (Forsskal) J. Agardh (Clorophyceae) in the Gulf of Salerno. (Southern Tyrrhenian Sea, Italy). Biologia Marina Mediterranea, 5:553–556, 1998.

[24] F. Trincardi, A. Cattaneo, A. Correggiari, S. Mongardi, et al. Submarine slides during sea level rise: Two examples from the eastern Tyrrhenian margin, in Submarine Mass Movements and Their Consequences, edited by J. Locat and J. Mienert.

Kluwer Acad., Dordrecht, Netherlands, page 469– 478, 2003.

[25] A. Bellonia, F. Budillon, F. Trincardi, D. Insinga, et al. Licosa and Acciaroli submarine slides, Eastern Tyrrhenian margin: characterisation of a possible common weak layer. Rendiconti online Soc. Geol. It., 3:83–84, 2008.

[26] G. Verdicchio and F. Trincardi. Mediterranean shelf-edge muddy contourites: examples from the Gela and South Adriatic basins. Geo-Mar Lett., 28:137–151, 2008.

[27] D. Hunt and M.E. Tucker. Stranded parasequences and the forced regressive wedge systems tract: deposition during base-level fall. Sedimentary Geology, 81:1–9, 1992.

[28] R.G. Fairbanks. A 17,000 year glacio-eustatic sea-level record: Influence of glacial melting rates on the younger Dryas event and deep ocean circulation. Nature, 342:637–642., 1989.

551

Marine Geology

552

3D Seismic Geomorphology: the Enigma Project, an Encounter between Academia and Industry G. Dalla Valle1 , F. Gamberi1 , F. Trincardi1 , P. Rocchini2 1, Institute of Marine Sciences, CNR, Bologna, Italy 2, ENI E&P Division, Sedimentology, Petrography & Stratigraphy Dpt., S. Donato Milanese, Italy giacomo.dalla.valle@bo.ismar.cnr.it Abstract The advent of the 3D seismic technology has represented a major revolution for the Earth Sciences, in the development of interactive interpretation systems. 3D seismic technology, through the transition from a two-dimensional, cross-section analysis to a real three-dimensional visualization of the entire sedimentary basin, has led to a unique opportunity for the study of submarine geological processes. 3D seismics provide the opportunity to image the geological elements in plainview, furnishing a detailed morphology of the seafloor at any given time/depth. Studies are aimed at tracing the signature of sedimentary processes from shallow water to the deepwater realm, easing the evaluation of architectural elements distribution, their facies components and the making of lithological predictions. The imaging of a variety of features related to the mass-transport complexes (MTCs) and fluid migration offers also the possibility to improve geohazard studies.

The coupling of modern seafloor observations, a field in which ISMAR has a long lasting tradition of researches, with industry-derived three-dimensional seismic data furnished by eni, creates a positive feedback that leads to exciting results regarding the characterization of depositional systems and the develop of sedimentological models.

1

Introduction

Three dimensional (3D) seismic has represented one of the most important revolution in the Earth Sciences of the last forty years. Originally developed by industry to mitigate risk in hydrocarbon exploration, nowadays the 3D seismic is an indispensable tool earth scientists for understanding complex geological phenomena. Seismic geomorphology, that consists in the extraction of geomorphic insight using 3D seismic data, allows the researchers to study the subsurface using plain view image and

making possible the reconstruction of the evolution of the landscapes through time.

Thanks to the powerful spatial resolution furnished by the 3D seismic, comparable, at least for the shallower portion of the sedimentary basin, to that obtainable through modern swath bathymetry, it has been possible to characterized complex geological structures with unprecedent detail. Among these are the investigation of buried sinuous deep-water channels, with a detail that was virtually impossible with the 2D seismics. The advent of 3D seismics increasingly document the wide variety in mor-

Marine Geology

phology and internal organization of large mass-transport complexes (MTCs) on continental margins. The ISMAR-Bologna has started a collaboration with eni-Milan in order to develop a sedimentological conceptual model of continental margin development coupling the industrial-derived 3D seismic database and the know-how of ISMAR regarding the modern seafloor sedimentary processes.

2

3D seismic: background

historical

It was as early as 1970 when G.G. Walton presented the concept of three dimensional seismic (3D seismic) in the seminal paper â&#x20AC;&#x153;Three Dimensional Seismic Methodâ&#x20AC;?

[2], that was the cornerstone for an extraordinary revolution in the Earth Sciences, both in the industrial and in the academic realm. In this paper, in fact, were first presented the concepts that underlie the 3D seismic technology and its possible applications in term of hydrocarbon exploration. Only three years after, in the 1975, the new technology was applied to the real world with the first 3D commercial survey performed in the North Sea. In the early years, the use of the 3D seismic technology was restricted to the petroleum industry realm due to the very high costs of operation both in the planning and in implementation of the seismic survey phase, and in the processing phase, obtainable only through high-expensive performance computing. Whereby the clas-

554

sical 2D surveys consisted of 100 to 200 lines and cross lines, a 3D survey require an higher number of closely spaced crossing seismic traces with a high-precision navigation system. Due these requirements, an important limitation of the early surveys was the small areal coverage due to prohibitive costs per unit area. With the advance of 3D technology, cost was lowered gradually and nowadays is not uncommon to deal with up to 10.000 km2 single surveys, or mega-surveys, obtained to the merging of multiples surveys as in the North Sea Basin, with a coverage of over 100.000 km2 . The resolving power of 3D seismic technology, both in term of vertical and horizontal resolution, and in term of accuracy and precision has promoted an enormous boost on the ability of interpret and characterize complex geological structures at the scale of the entire basin or wider. Until the advent of the 3D technology in fact, the limitation of the spatial resolving power of the 2D seismic technology was the greatest obstacle for the accurate reconstruction of hydrocarbon traps, reservoir continuity and integrity [3]. Despite with the use of 2D seismic, it was possible to delineate the basic framework of the sedimentary basins, and to recognize the main sedimentary and tectonic structures, the detailed reconstruction of the morphology of the depositional systems and the three-dimensional geometries of the fault systems and other tectonic structures (thrust systems, relay ramp, etc) it is possible only with the 3D seismic technology.

Marine research at CNR

Figure 1: a) Example of a 3D seismic amplitude cube. The green reticulate is achieved by the picking of a key horizon on the seismic cube (Courtesy of ENI). b). In a sesmic cube each amplitude trace consists of an array of amplitude samples that is represented as a series of voxels. The gray level associated with each voxel denotes the amplitude intensity of the sample (Modified from [1]).

555

Marine Geology

556 Figure 2: Example of a deep-water channel imagined on the same time-slice through three different seismic attributes: a). Amplitude; b). Coherence; c). Sweetness (Image courtesy of ENI).

Marine research at CNR

3

3D seismic technology

Whereas initially 3D seismic data was displayed and interpreted in a 2D view, with the advent of graphic interactive workstations the interpretation was made into a fully real 3D environment. Some specialized software allows infact to visualize the acquired seismic data as three-dimensional “cube” consisting of millions of voxels, that represents the smallest unit in a 3D dataset that can display a colour corresponding to the seismic sample value. A 3D cube, or volume (Figure 1) is not a simple closer staking of 2D lines, but differs greatly from this: there is no the spatial aliasing as in the 2D data, and the greater density of spatial sampling guarantee a lateral resolution similar to the vertical one.

The sophisticated 3D migration techniques (3D dip move-out as example), permit to image geological structures characterized by steep dips and strong lateral velocity variations with a detail that would be impossible with the 2D seismic. The seismic cube, or volume, can be interrogated by the interpreter using a wide range of mathematical operations as rotation to any angle, flattening, enlarging and zooming. The interpreter may need infact to “stretch” or “shrink” the view, or to squeeze the 3D data to show sedimentary features or tectonic discontinuities in more detail, and this can be accomplished also by changing the vertical and horizontal exaggeration. One of the most useful application of the 3D seismic technology is the ability to incorporate opacity and transparency values in the display of a three-dimensional volume, and the possibility to render the 3D volume or probe. In this way the interpreter can isolate specific range of data values within the volume in order to see exclusively those data within the probe, rather than just see

the data on the faces of the probe.

The interpreter has the possibility to create a variety of seismic volumes using different seismic attributes (Figure 2). A seismic attributes normally provide informations associate to the amplitude, shape, and position of the seismic waveform and permit to resolve geological features and their relationship that otherwise would be omitted, such as channels pathway, channel infill, and presence of fluids or gas. Attributes as envelope, instantaneous phase, and instantaneous frequency allows to discern the presence of gas (bright spots), and changes in lithology. Coherency attribute is used to evidence the lateral continuity of the reflectors, and the presence of numerically separated surfaces, as faults, and fractures.

Nowadays the production of various reflection attributes such as amplitude, dip magnitude, dip azimuth, time/depth structure, polarity sweetness, etc. is increased exponentially; however there is not an allpurpose seismic attribute, but these must be carefully chosen by the interpreter on the basis of his experience and on the basis of the expected goal.

4

3D seismic geomorphology

The study of geological features using plain view images generated through the manipulation of 3D seismic data is called seismic geomorphology [4]. The opportunity to visualize geological features in map view has represent the most valuable progress that the 3D technology has provide to the geological prediction, establish a marked step beyond the seismic stratigraphy mindset.

557

Marine Geology

Figure 3: Example of the same sinuous deep-water channel and its main architectural elements, investigated trough the traditional 2D seismic (a), and imaged using plan view images with the 3D seismic technology (b) (Image courtesy of ENI).

558

Marine research at CNR

Figure 4: a) An example of the resolution obtainable with the 3D seismic on a series of buried deep-water channels (Image courtesy of ENI); b) Modern deep-water channel imagined with multibeam swath bathymetry by ISMAR Bologna.

Figure 5: Example of a slope channel developed on a topographically complex slope.

The channel pathway is controlled by a series of topographic highs, related to tectonic structures. In this case is the paleo-seafloor is imagined with a series of time-equivalent, coherence probes (Image courtesy of ENI).

559

Marine Geology

4.1

Methodologies

Various planes, surfaces and time slices can be extracted from the different types of 3D volume. The interpreter can slice a plane surface through the depth/time direction of the 3D cube, without reference to any stratigraphic horizon. In this way, a time-slice surface is extracted from a 3D cube at a constant depth/time coordinate, and permits to the interpreter to looking for geological features, above or below the surface event.

However, geological features rarely show up completely at a spatially consistent time/depth within the seismic volume, and they could be also deformed by postdepositional structural events, which make them even more complex. To outflanking these limitations, the interpreter can create a series surfaces, through the picking (manually or automatically) of spatially consistent reflectors such as thick shale horizons, in order to obtain a time-equivalent surface from the 3D cube. The interpreter can analyze the spatial variations of any seismic attribute data along the selected surfaces and through time. Moreover, in the case of tectonic deformation, the picked time-equivalent horizon can be flattened, and the resulting image is an approximation of the morphology as it existed at the time of deposition. Through the seismic geomorphologies methods, the interpreter has a more realistic view of the sedimentary environments, and can directly identify the process-derived facies distribution on map view [5]. When used in conjunction with seismic stratigraphy, and with correlated borehole data, seismic geomorphology, represent the state of the art approach to extracting stratigraphic insights from 3D seismic data of subsurface sedimentary bodies [6].

560

4.2

Submarine channels

One of the most striking success of the seismic geomorphology was the imagining of buried sinuous submarine channels (Figure 3). Before the advent of the 3D seismic in fact, no imaging of the planform of the depositional systems was available, and only qualitative 2D description of seismic reflection geometries and terminations captured planview distributions of depositional elements. A further advantage furnished by the 3D seismics is that the planview image of the submarine channels and others sedimentary elements can be extracted at any given depth. As a consequence the interpreter can observe the morphological evolution of the depositional elements through time, and examine the response of the sedimentary systems to the changes of the depositional environment. Using a geomorphic approach, the sedimentary architectural elements can be mapped accurately both in time and space, observing the variations of the depositional systems at in response of controlling factors changes.

It can be evaluated how submarine channels evolve in topographically complex slopes (Figure 5) resulting from shale/salt tectonics or from tilted basement faultblocks, during the infilling of the receiving basin. The progressive healing of the seafloor topography drives the evolution and the pathway of the channel, by changing in the slope gradient, basin geometry and the degree of confinement, and eventually, in promoting fill-and-spill processes from upper intra-slope basins to lower ones. The availability of large 3D surveys has been also very useful in showing that those who were considered ponded turbidite systems, in some cases, do not terminate in intra-slope basins, but they have continuous but very convoluted

Marine research at CNR

courses which takes them through and beyond complex slope topography. Intrachannel knickpoints, meandering cut-off, meander-loop migration or abrupt channel avulsions due to variations in the surrounding seafloor topography can be not only observed and described, but they can also quantified in space and in time. The evolution of these sedimentary architectural elements that characterize the submarine channels can be afforded also by statistical analysis of different parameters, as variations of the sinuosity index, the numbers of bend and their radius of curvature, the meander length and width, the ratio between channel depth and width, levee heights and width during the whole evolution of the systems.

These parameter variations can be used to predict the nature and the distribution of the deposits, obtaining important information in the reservoir heterogeneity and lithological prediction [7].

4.3

along the basal shear surface, it can be also reconstructed the type and direction of motion and the model of emplacement of the various MTCs [9].

The study of MTCs through 3D seismic geomorphological analysis has been useful for those researcher who seek to understand the mechanisms of submarine slope failures that can generate tsunamis in coastal regions. The 3D seismic geomorphology has bought to light the complex behaviour of large mass transport deposits, often characterized by multiple failure events, showing different rheology and mechanisms of initiation. Submarine failures along the continental margins can also affect and destroy submarine cables, pipeline and other engineered structures for hydrocarbon exploration, and the increasing studies lead with a seismic geomorphological approach is gradually improving our ability to predict and mitigate the impact of these events.

Mass Transport Complexes 4.4 (MTCs)

3D seismic geomorphology has been recently also used on the imaging of masstransport complexes (Figure 6)([8]). The possibility of study and analyze from a geomorphic perspective, large-scale events in the off-shore environment, has lead to a better comprehension of the mechanisms controlling the initiation and the propagation of the MTCs. In particular, through the characterization of the external geometries and internal distribution of deformational structures that forms in headwall, in the sidewalls and in the toe-region of the MTCs, with the imaging of complex imbricate thrusts and fold systems. Through the recognition of specific kinematic indicators

Other research avenues

The 3D seismic have made great contribution to a wide range of additional avenues of research ranging from structural geology, fluid-rocks interactions and the study of igneous systems. Three-dimensional analysis has allowed to study how faults grow and link [10] and to discover new type of fault systems as polygonal faults [11]. Other application of 3D seismic regard the study of subsurface plumbing systems as gas-blow out pipes and mud volcanoes [12], and the imaging of large-scale sand intrusions and collapse craters, that plays a significant role in controlling and affecting reservoir geometry and permeability [13].

561

Marine Geology

Figure 6: Example of a buried slope failure imaged through time equivalent, coherence probes (Image courtesy of ENI).

5

The Enigma Project: an overview

age of eni 3D seismic dataset make possible to obtain morphometric observations and to trace the signature of sedimentary processes, at the scale of the entire basin, In the last year, the ISMAR of Bologna from shallow to deep-water areas, outflankhas started a collaboration with the Ex- ing the spatial limitation linked to tradiploration& Production division of ENIi tional 2D seismic data (Figure 7).

(Ente Nazione Idrocarburi) of San Donato Milanese. The intent of this join venture (Enigma Project) is to couple the 6 The Enigma project: knowledge of the two institutions in order to develop a sedimentological concepmain results tual model on silicoclastic continental margins. ISMAR has provided its data set of The eni 3D seismic volume has been acmodern seafloor of the Adriatic sea that quired along the central Adriatic conticonsist of multibeam swath bathymetry, nental margin, focussing in particular in sidescan sonar and sedimentological and the Plio-Pleistocene sedimentary successtratigraphic coresm whereas eni has fur- sion that consists mainly of prograding nished its large, high-resolution 3D seis- slope clinoforms. The seismic geomormic dataset acquired in recent years along phology techniques has allowed to asses the Adriatic basin. This approach has pro- how sediment is dispersed and partitioned vided the exciting opportunity to investi- across the slope clinoforms during their degate an industrial-oriented data set with an velopment. The changes of the clinoforms academic point of view. The wide cover- characteristics during the infilling of the 562

Marine research at CNR

Figure 7: Example of a series of regional scale surfaces with different ages, obtained through the interpolation of selected horizons in order to approximate the paleo-lanscapes at the time of deposition. sedimentary basin form the basis for the construction of the conceptual model. An examination of the geometries of the slope clinoforms and of their internal stacking patterns has allowed to in order to shed light on the morphological differences and the linkages between the different sectors of the clinoforms and the related sedimentary bodies.

By comparison with similar geomorphic elements present on the modern seafloor of the Adriatic and Tyrrhenian basin, that can

provide indications on the geological processes responsible for their formation, it has been possible to develop a conceptual tool to predict when sands are likely to be connected or disconnected along slope segments of the different types of shelf-margin clinoforms. The model has been successively generalized as much as possible in order to be exported and applied in different but similar contexts, with the aim of reduce uncertainties in exploration and reservoir evaluation.

References [1] D. Gao. 3D seismic volume visualization and interpretation: An integrated workflow with case studies. Geophysics, 74:W1â&#x20AC;&#x201C;W12, 2009.

[2] G.G. Walton. Three dimensional seismic method. Geophysics, 37(3):418â&#x20AC;&#x201C;430, 1972.

563

Marine Geology

[3] H.W. Posamentier. Application of 3D seismic visualization techniques for seismic stratigraphy, seismic geomorphology and depositional systems analysis. 29:11–24, 2004.

[4] H.W. Posamentier. Seismic stratigraphy into the next millennium. A focus on 3D seismic data. Abstract from AAPG Annual conf. New Orleans, April 16-19, 2000, page A118, 2000.

[5] J. Cartwright and M. Huuse. 3D seismic technology: the geological ‘Hubble’.

Basin Research, 17(1):1–20, 2005.

[6] H.W. Posamentier and V. Kolla. Seismic geomorphology and stratigraphy of depositional elements in deep-water settings. Journal of Sedimentary Research, 73(3):367–388, 2003.

[7] L.J. Wood and K.L. Mize-Spansky. Quantitative seismic geomorphology of a quaternary leveed-channel system, offshore eastern Trinidad and Tobago, northeastern South America. AAPG Bullettin, 93(1):101–125, 2009.

[8] J. Frey Martinez, J. Cartwright, and B. Hall. 3D seismic interpretation of slump complexes: examples from the continental margin of Israel. Basin Research, 17(1):83–108, 2005.

[9] S. Bull, J. Cartwright, and M. Huuse. A review of kinematic indicators from mass-transport complexes using 3D seismic data. Marine and Petroleum Geology, 26(7):1132–1151, 2009.

[10] C. Mansfield and J.A. Cartwright. Fault growth by linkage: observation and implications from analogue models. Journal of Structural Geology, 23:745–763, 2001.

[11] J.A. Cartwright. Episodic basin-wide expulsion from geopressured shale sequences in the North Sea Basin. Geology, 22:447–450, 1994.

[12] R.J. Davies and S.A. Stewart. Emplacement of giant mud volcanoes in the South Caspian basin: 3D seismic reflection imaging of their root zone. Journal of the Geological Society, 162:1–4, 2005.

[13] M. Huuse, D. Duranti, N. Steinsland, C.G. Guargena, P. Prat, K. Holm, J.A. Cartwright, and A. Hurst. Seismic characteristics of large-scale sandstone intrusions in the Paleogene of the South Viking Graben, UK, and the Norwegina North Sea. Memoirs of the Geological Society of London, 29:262–277, 2004.

564