Bull Volcanol (2008) 70:583–603 DOI 10.1007/s00445-007-0155-0

RESEARCH ARTICLE

New insights into Late Pleistocene explosive volcanic activity and caldera formation on Ischia (southern Italy) Richard J. Brown & Giovanni Orsi & Sandro de Vita

Received: 19 April 2006 / Accepted: 13 May 2007 / Published online: 17 July 2007 # Springer-Verlag 2007

Abstract A new pyroclastic stratigraphy is presented for the island of Ischia, Italy, for the period ∼75–50 ka BP. The data indicate that this period bore witness to the largest eruptions recorded on the island and that it was considerably more volcanically active than previously thought. Numerous vents were probably active during this period. The deposits of at least 10 explosive phonolite to basaltictrachyandesite eruptions are described and interpreted. They record a diverse range of explosive volcanic activity including voluminous fountain-fed ignimbrite eruptions, fallout from sustained eruption columns, block-and-ash flows, and phreatomagmatic eruptions. Previously unknown eruptions have been recognised for the first time on the island. Several of the eruptions produced pyroclastic density currents that covered the whole island as well as the neighbouring island of Procida and parts of the mainland. The morphology of Ischia was significantly different to that seen today, with edifices to the south and west and a submerged depression in the centre. The largest volcanic event, the Monte Epomeo Green Tuff (MEGT) resulted in caldera collapse across all or part of the island. It is shown to comprise at least two thick intracaldera ignimbrite flowunits, separated by volcaniclastic sediments that were Editorial responsibility: R. Cioni R. J. Brown : G. Orsi : S. de Vita Instituto Nazionale di Geofisica e Vulcanologia, Osservatorio Vesuviano, Via Diocleziano 328, 80124 Napoli, Italy Present address: R. J. Brown (*) Department of Earth Sciences, University of Bristol, Wills Memorial Building, Queen’s Road, Bristol BS8 1RJ, UK e-mail: [email protected]

deposited during a pause in the eruption. Extracaldera deposits of the MEGT include a pumice fall deposit emplaced during the opening phases of the eruption, a widespread lithic lag breccia outcropping across much of Ischia and Procida, and a distal ignimbrite in south-west Campi Flegrei. During this period the style and magnitude of volcanism was dictated by the dynamics of a large differentiated magma chamber, which was partially destroyed during the MEGT eruption. This contrasts with the small-volume Holocene and historical effusive and explosive activity on Ischia, the timing and distribution of which has been controlled by the resurgence of the Monte Epomeo block. The new data contribute to a clearer understanding of the long-term volcanic and magmatic evolution of Ischia. Keywords Pyroclastic stratigraphy . Explosive volcanism . Caldera collapse . Ischia . Late Pleistocene

Introduction The island of Ischia in the Gulf of Naples (Fig. 1) is home to ∼50,000 inhabitants and a thriving tourism business. Its popularity is due in part to the active volcanic state and the presence of numerous hot springs fed by an active hydrothermal system (e.g., Panichi et al. 1992). Volcanic hazard definition on Ischia and the long-range forecasting of future eruptions are prime objectives of current research. Recent (<10 ka) volcanism on Ischia has been characterised by numerous small-volume explosive and effusive eruptions that have distributed their products over relatively restricted areas of the island (see Vezzoli 1988). This low-magnitude activity belies a violent volcanic past involving voluminous explosive volcanic eruptions and caldera formation (see Vezzoli 1988) that remains incompletely understood. How

584

Bull Volcanol (2008) 70:583–603

Fig. 1 Simplified geological map of Ischia; modified after Orsi et al. (2003). The products erupted between 75–50 ka outcrop along the southern coast and around the Monte Epomeo resurgent block. Inset: the regional setting of Ischia

the present volcanic-magmatic system of the island relates to the long-term evolution is crucial because large-volume eruptions at evolved caldera volcanoes typically have long repose periods (i.e., 103–105 years; e.g., Druitt et al. 1989; Brown et al. 2003). Little remains of the earliest volcanic deposits on Ischia (pre-75 ka BP). These are phonolitic in composition and indicate a protracted early history of magma generation, differentiation and eruption (see Vezzoli 1988). In this paper, we present a stratigraphic framework for the period ∼75–50 ka BP that enlarges the state of knowledge on Ischia and indicates that this period was significantly more volcanically active than previously thought. A diverse series of explosive volcanic phenomena are recorded including caldera-forming eruptions, the generation of widespread pyroclastic density currents, Plinian or sub-Plinian eruption columns and phreatomagmatic activity. Furthermore we have identified the deposits of previously unrecognised eruptions. Several eruptions dispersed pyroclastic density currents over the neighbouring island of Procida and the now heavily populated parts of SW Campi Flegrei. This period saw the largest eruption

recorded on the island, the Monte Epomeo Green Tuff (MEGT; Rittmann 1930) which resulted in caldera subsidence (Vezzoli 1988; Tibaldi and Vezzoli 1998). The data build on past local stratigraphic studies (Forcella et al. 1981, 1982; Rosi et al. 1988a, b) and on the comprehensive study of Vezzoli (1988), and provide a more complete stratigraphic framework that complements recent work on neighbouring Procida (De Astis et al. 2004) and on Mediterranean marine and terrestrial ash layers (e.g., Keller et al. 1978; Paterne et al. 1988; Narcisi 1996; Wulf et al. 2004). Volcanism and deformation on Ischia Ischia sits on the western margin of the Phlegraean Volcanic District (Fig. 1), and comprises alkali-trachyte, and subordinate shoshonite, latite and phonolite volcanic rocks (Poli et al. 1987; Crisci et al. 1989; Civetta et al. 1991; Piochi et al. 1999). Volcanism in the Campanian region accompanied Pliocene and Quaternary extension responsible for the opening of the Tyrrhenian basin (Sartori 2003 and refer-

Bull Volcanol (2008) 70:583–603

ences therein). Ischia has experienced discontinuous volcanic activity for over 150 ka (Gillot et al. 1982; Vezzoli 1988; Orsi et al. 1996, 2003; Civetta et al. 1999; Fig. 1), but the stratigraphically oldest rocks on the island are poorly exposed and have not yet been dated. Geophysical data indicate that the island is the remnant of a once larger, older volcanic complex extending to the west (Orsi et al. 1999; Bruno et al. 2002). Between 150 and 75 ka BP trachytic, alkali-trachytic and phonolitic lava, lava domes and pyroclastic rocks were erupted (Gillot et al. 1982; Vezzoli 1988). Following this (75–50 ka BP), explosive eruptions emplaced a succession of pumice fall deposits and pyroclastic breccias (Pignatiello Formation of Forcella et al. 1982; Rosi et al. 1988a; Vezzoli 1988). Numerous ash layers erupted from Ischia during this period occur in marine sediment cores in the eastern Mediterranean (see Paterne et al. 1988; Narcisi and Vezzoli 1999), in lake sediment cores in southern Italy (e.g., Lago Grande di Monticchio, Narcisi 1996; Wulf et al. 2004) and in the Aeolian Islands (Keller 1980). The last 55 ka of activity on the island have been divided into three magmatic cycles based on stratigraphic, geochemical, isotopic and radiometric age data (Poli et al. 1987, 1989; Vezzoli 1988; Civetta et al. 1991). The first cycle (55–33 ka BP) was characterised by closed system behaviour and explosive volcanic eruptions and commenced with the catastrophic caldera-forming MEGT eruption. This was thought to be responsible for the ∼10×7 km caldera on Ischia (Vezzoli 1988; Tibaldi and Vezzoli 1998). This was followed by several phreatomagmatic eruptions at offshore centres (Vezzoli 1988). The second cycle (30–18 ka BP) was marked by the arrival of shoshonite magma into the main reservoir, which triggered the resurgence of the Monte Epomeo block. The third cycle (10 ka–1302 AD) involved open system behaviour and mixing between two distinct magmas (Civetta et al. 1991; Piochi et al. 1999). The last 5 ka have been characterised by discontinuous refilling of the magma chamber, resulting in alternating periods of resurgence, intense volcanic activity and quiescence (Orsi et al. 1991, 1996; de Vita et al. 2006). The last eruption occurred in 1302 AD in the northeast of the island (the Arso lava; Vezzoli 1988). The Monte Epomeo resurgent block records a net uplift of >900 m over ∼30 ka (see Vezzoli 1988; Orsi et al. 1991; Tibaldi and Vezzoli 1998). Repeated flank failures have occurred during uplift, resulting in debris flows (Tibaldi and Vezzoli 2004; de Vita et al. 2006) that have fed a large submarine debris apron. A widespread (250–300 km2) submarine debris avalanche deposit off the southern coast is thought to have resulted from a large-scale sector collapse event (Chiocci and de Alteriis 2006). The continued active volcanic state of the island is indicated by historical ground movements (Buchner et al. 1996), seismicity (e.g., the 1883 Casamicciola earthquake; Cubellis et al. 2004) and by fumaroles and

585

thermal springs (see Inguaggiato et al. 2000; Caliro et al. 1999; Chiodini et al. 2004).

Materials and methods A regional survey was undertaken to define the stratigraphic sequence on Ischia between ∼75 and 50 ka BP (Fig. 2). The succession has been pieced together from widely scattered outcrops around Ischia, Procida and Monte di Procida (SW Campi Flegrei, Fig. 1). The sequence is complex, but is stratigraphically most complete and most easily accessible on the southeast coast of Ischia (Fig. 3). Palaeosols have been used to divide the sequence into stratigraphic units that, for simplicity, we infer are the products of one single eruption, which may or may not have comprised several eruptive phases. These units are summarised in Fig. 2. The sequence includes pumice fall deposits, block-and-ash flow deposits, thick ignimbrites, pyroclastic breccias and minor volcaniclastic sediments, which are described and interpreted below. The oldest deposits overlie palaeosols developed in pyroclastic deposits, lavas and lava domes extruded between >150 and 75 ka BP (e.g., Monte Sant’Angelo and Monte Vico volcanic centres and Parata Tephra; Vezzoli 1988). Major unconformities occur throughout the sequence and the correlation of individual deposits can be difficult, as is determining their vent locations and dispersals. Representative whole rock analyses of major and trace elements for the erupted products are presented in Table 1 and in Fig. 4. The earlier erupted rocks (pre-MEGT) are phonolite and trachyphonolite in composition, while later rocks, including parts of the MEGT, are trachyte to basaltic trachyandesite. The results are in broad agreement with those obtained by Crisci et al. (1989). Terminology generally follows that used by Fisher and Schmincke (1984) and Cas and Wright (1987). We use the term ‘ignimbrite’ for pyroclastic density current deposits, and use prefixes such as ‘pumice-rich’, ‘lithic-rich’ or ‘ashrich’ to distinguish the many different types present. We use the term ‘block-and-ash flow deposit’ to describe pyroclastic density current deposits comprised of dense juvenile clasts in an ash matrix.

Pre-Monte Epomeo Green Tuff volcanism: stratigraphic units Sant’Angelo Tephra The phonolitic Sant’Angelo Tephra outcrops only on the Sant’Angelo peninsula in the south of the island (Figs. 1 and 5). It comprises parts of the Sant’Angelo volcanic

586

Bull Volcanol (2008) 70:583–603

Fig. 2 Generalized vertical succession of the volcanic deposits erupted 75–50 ka with summary description and interpretation. Gaps in the stratigraphy indicate uncertainties in stratigraphic position due

to incomplete exposure. t = trachyte; ta = trachyandesite; bta = basaltic trachyandesite; p = phonolite; alk fsp = alkali feldspar; plag = plagioclase; cpx = clinopyroxene; bio = biotite; neph = nepheline

centre and the basal MEGT “explosion breccia” of Vezzoli (1988; see below for discussion of correlations). It overlies a palaeosol developed in the rubbly top of a lava dome (∼100 ka, Gillot et al. 1982), but it is not seen in contact with other pre-MEGT deposits and its relative age cannot be fully established (see Fig. 3). The Sant’Angelo Tephra can be broadly divided into two units (Fig. 5). The lower unit comprises interbedded decimetre-thick pumice fall deposits and dark-brown ignimbrites. The ignimbrites contain pumice and scoria lapilli and blocks, some of

which are flattened and imbricated, in a brown ash matrix. Diffuse bedding is defined by discontinuous pumice and scoria lapilli lenses and layers (Fig. 5). The upper unit comprises two monomict breccias separated by a 1 m thick pumice fall deposit (Fig. 5). The breccias comprise angular clast-supported lapilli, blocks and boulders of dense porphyritic phonolite in a fines-poor ash matrix. Lower parts of each breccia comprise pumice and lithic lapilli in an abundant ash matrix. The contacts between each breccia and its lower pumice-rich parts exhibit load-and-flame structures

Bull Volcanol (2008) 70:583–603

sheared in a southerly direction. Large boulders protrude from the top of the breccias. The Sant’Angelo Tephra is unconformably overlain by the extracaldera lithic lag breccia of the MEGT eruption (Figs. 3 and 5; see below). The Sant’Angelo eruption The Sant’Angelo Tephra records a range of eruptive mechanisms. There is no evidence for a significant pause in volcanic output and it is interpreted as one single eruptive unit that may have been emplaced over a protracted period (days to months). Intercalated pumice fall deposits and ignimbrites in the lower unit (Fig. 5) record synchronous pumice fall and pyroclastic density current activity from quasi-sustained eruption columns. The monomict breccias (Fig. 5) are interpreted as block-and-ash flow deposits generated during dome collapse events (e.g., Cas and Wright 1987). A sustained eruption column showered pumice over the region in between the block-and-ash flow-generating events (Fig. 5). Palaeocurrent directions from imbricated clasts and sheared load-and-flame structures in the block-and-ash flow deposits indicate eruption to the north of Sant’Angelo (Fig. 5). The presence of block-and-ash flow deposits, together with the absence of the Sant’Angelo Tephra deposits elsewhere on the island, lead us to tentatively conclude that it was probably erupted from a small volcanic centre subject to dome extrusion and small volume explosive activity. Mago, Olummo, Tisichiello and Porticello Tephras The upper part of the pre-MEGT stratigraphy in southern Ischia (Figs. 3 and 6) comprises a succession of bedded phonoliteto-trachyte pumice fall deposits (Fig. 4). These deposits comprise part of the Pignatiello Formation of Vezzoli (1988). Mago Tephra The Mago Tephra is best exposed at Grotta di Terra (Fig. 1). It overlies a coarse scoria breccia of the 74 ka Parata Formation (Vezzoli 1988). It comprises ∼60 cm of diffuse-bedded well-sorted fine to medium phonolite pumice lapilli and grades up into a palaeosol (Fig. 6). Olummo Tephra The Olummo Tephra unconformably overlies the palaeosol developed in the top of the Mago Tephra. It is well-exposed around Grotta di Terra (Fig. 1) where it comprises up to 6 m of diffuse-stratified clast-supported phonolite pumice lapilli and thin intercalated ash layers overlain by a discontinuous 2 m thick breccia (Fig. 6). A distinct internal stratigraphy is defined by vertical grainsize variations and by the presence in some layers of juvenile obsidian, banded pumice and scoria, hydrothermally oxidised lithic clasts and by intercalated thin lithic lapilli layers, orangepink ash layers and diffuse-stratified matrix-supported lapillituff layers (Fig. 6). Carbonised twig and leaf fragments occur

587

in the ash layers. This passes up into a clast-supported lapilli and block breccia comprised of dense porphyritic phonolite and subordinate pumice lapilli and blocks, some with glassy exteriors. A fines-depleted crystal-lithic ash matrix occurs throughout the breccia and its base comprises discontinuous lenses of stratified, inverse-graded fines-depleted lithic-crystal ash. A palaeosol is developed in the top of the breccia. Tisichiello Tephra The ∼7 m thick Tisichiello Tephra overlies an unconformity and palaeosol in the Olummo Tephra (Fig. 3). It is well exposed east of Monte Vezzi (Fig. 1). It has a distinctive internal lithostratigraphy defined by changes in the vesicularity of the pumice (from highly vesicular pumice through to dense clasts; Fig. 6), vertical changes in grainsize and grading and by variable abundances of lithic lapilli. Several clast-supported pumice beds contain rounded pumice lapilli and two have a sparse ash matrix (Fig. 6). Intercalated centimetre to decimetre thick non-indurated pink-brown ash layers and indurated thicker diffuse-stratified lapilli-tuff layers occur within the sequence (Figs. 6 and 7a). Carbonised twig and leaf fragments occur in the lapilli-tuff layers. The Tisichiello Tephra passes up into a brown palaeosol. A major unconformity occurs above the Tisichiello Tephra (Fig. 6), and locally cuts down to the Mago Tephra. At Grotta di Terra (Fig. 1), this horizon is punctuated by box-canyons filled with bedded and stratified volcaniclastic sediments. Porticello Tephra The Porticello Tephra overlies an unconformity and palaeosol in the Tisichiello Tephra (Figs. 3 and 6). It comprises decimetre-thick beds of clast-supported pumice and subordinate lithic lapilli alternating with thin orange-brown ash and lithic lapilli layers containing carbonised vegetation (Figs. 6 and 7b). It is found on Procida and is the same unit as “Pomici pliniane C” of Rosi et al. (1988a). It passes up into a brown palaeosol. Eruption of the Mago, Olummo, Tisichiello and Porticello Tephras The four pumice fall units exhibit similar lithological characteristics. We interpret them as recording the fallout of lapilli from unsteady eruption columns. The indurated tuff and lapilli-tuff layers are interpreted as ignimbrites deposited by short-lived pyroclastic density currents (Fig. 6). Some of these layers have been previously interpreted as palaeosols (see Rosi et al. 1988a). We consider them to be pyroclastic density current deposits on the basis of their sharp, locally erosive basal contacts, unaltered matrix-supported pumice lapilli, diffuse bedding and stratification defined by horizons of pumice lapilli, the presence of rounded pumice lapilli and because they commonly thicken into small depressions. Carbonised

588

Bull Volcanol (2008) 70:583–603

Bull Volcanol (2008) 70:583–603

RFig. 3

Graphic logs of selected measured sections across Ischia, Procida and Campi Flegrei showing the major stratigraphic units and their correlations. The two most widespread deposits (the MEGT and the Schiappone Tephra) are highlighted in grey. Note the major unconformities in the sequence and that large parts of the stratigraphy are absent at some locations (e.g., at Sant’Angelo and Scarrupata di Barano)

twigs and leaves in these thin ignimbrites may indicate hot emplacement (e.g., Scott and Glasspool 2005). The breccia in the Olummo Tephra is interpreted as a block- and-ash flow deposit based on its juvenile components and sedimentological characteristics. The intercalated nonindurated ash layers (Fig. 6) may record either: (1) fine ash settling through the atmosphere due to the shut down of the column(s); (2) brief phreatomagmatic explosions; or (3) distal co-ignimbrite ash. The pumice fall deposits are traceable across much of the SE coast, although deposition on steep palaeoslopes resulted in the remobilisation of some deposits and their character can vary significantly from outcrop to outcrop. Pumice fall deposits occur in similar stratigraphic horizons in the southwest of the Ischia (Punta Imperatore, Figs. 1 and 3) and on Procida (Fig. 3; see Rosi et al. 1988a, b; De Astis et al. 2004) but correlations between some of these deposits remain uncertain. The dispersals of these pumice fall deposits and their vent locations are not presently known. Palaeosols in the sequence are recognised by their gradational bases into underlying deposits and by the presence of altered pumice lapilli (see Fig. 6).

Monte Epomeo Green Tuff: volcanism and volcano-tectonism Intracaldera deposits The MEGT (Rittmann 1930; Vezzoli 1988) outcrops across Ischia, and on Procida and western Campi Flegrei (Figs. 1 and 3). Intracaldera deposits comprise two indurated trachytic ignimbrite flow-units (Lower and Upper MEGT, Fig. 8) exposed in steep fault scarps around the Monte Epomeo resurgent block (Fig. 1). Vezzoli (1988) obtained a K/Ar age of ∼55 ka for the UMEGT intracaldera ignimbrites. The LMEGT is >70 m thick and is exposed only on the western flank of the most uplifted part of the resurgent block (Rione Bocca; Figs. 8 and 9a). Step-faulting associated with resurgence (Orsi et al. 1991) suggests that it lies below the present level of exposure in northwest parts of the block (Fig. 1). The LMEGT overlies marine volcaniclastic sandstones and conglomerates but the base is not exposed. It comprises a basal heterolithic lithic breccia up to 20 m thick that passes up into a ∼50 m thick massive ignimbrite (Fig. 8). The lithic population in the breccia comprises lapilli, blocks and boulders of phonolite and trachyte lava,

589

tuff, lapilli-tuff, coarse-grained holocrystalline feldspar-phyric sub-volcanic rocks, hydrothermally altered clasts and scattered sandstone clasts. Juvenile material in the LMEGT comprises brown-black irregular to fluidal-shaped dense and poorly vesicular spatter, abundant porphyritic tube pumice, and large feldspar (<15 mm) and biotite (<6 mm) free crystals. The altered ash matrix is strongly enriched in crystals relative to the juvenile material. The UMEGT is >200 m thick and is extensively exposed in the fault scarps around Monte Epomeo (Fig. 1). Lower parts of the UMEGT at Rione Bocca comprise, from base to top, clast-supported lithic breccia, boulder-bearing ignimbrite, cross-stratified and dune-bedded ignimbrite (Fig. 9c) and spatter-bearing ignimbrite (Fig. 8). The upper ∼150 m consists of homogenous normally-graded massive ignimbrite with abundant pumice lapilli and blocks and minor lithic lapilli (Fig. 8). Lithic and pumice lapilli become finer grained and less abundant upwards. Juvenile material is similar to that of the LMEGT. The UMEGT is unconformably overlain by laminated, bedded and locally fossiliferous marine volcaniclastic siltstones and sandstones (Colle Ietto Formation, Vezzoli 1988) which record the reworking of the MEGT tephra during post- or syn-eruptive flooding of the caldera (Vezzoli 1988; Barra et al. 1992). The contact between the Lower and Upper MEGT is discontinuously exposed at Rione Bocca (Fig. 1). It is marked by a ∼1 m thick unit of faintly laminated to massive, moderately to poorly-sorted vitric siltstones and fine- to coarse-grained sandstones (Fig. 9b). These sediments locally exhibit convoluted bedding. Elsewhere, the contact is marked by several metres of chaotically bedded whitish vitric siltstone and sandstone with broad local erosion surfaces and in which it is difficult to distinguish either the top of the LMEGT or the base of the UMEGT. To the north, the contact is marked by an erosion surface that cuts down >20 m into the LMEGT. Coarse lithic breccia lenses and scattered boulders occur in the UMEGT along this scour surface, but where they are absent the two ignimbrite flow-units are indistinguishable. The contact is less obvious on the north-facing scarps of Monte Epomeo (Fig. 1), but it may be marked by a discontinuous clastsupported lithic breccia that overlies a massive yellowweathering lapilli-poor ignimbrite that probably belongs to the LMEGT. Extracaldera deposits We correlate a range of deposits exposed along the coastline of Ischia and on Procida and the mainland with the intracaldera MEGT (Fig. 3). The first is a coarsegrained pumice fall deposit up to 7 m thick at Grotta di Terra (Fig. 3). The lower 4 m are lithic-poor and comprise white, clast-supported highly inflated oval and tube-vesicle

590

Bull Volcanol (2008) 70:583–603

Table 1 Major (XRD) and trace element data (ICP) for representative whole rock samples of pyroclastic deposits erupted on Ischia between 75 and 50 ka Sample no.

OIS 0326

OIS 0301

OIS 0302

OIS 0317

OIS 0309

OIS 0320

MEGT 0302

Rock type

Pumice

Pumice

Pumice

Lava blocks

Pumice

Pumice

Glass

Stratigraphic unit

Sant’Angelo

Mago

Olummo

Olummo

Porticello

Tisichiello

MEGT LMEGT

SiO2 TiO2 Al2O3 Fe2O3 MnO MgO CaO Na2O K2O P2O5 L.O.I. Tot. V Cr Co Ni Rb Sr Y Zr Nb Cs Ba La Ce Nd Sm Eu Tb Dy Yb Lu Hf Ta W Pb Th U

61.2 0.57 18.69 3.06 0.24 0.55 1.13 7.23 6.92 0.12 0.00 99.7 29.6 < L.D. 0.98 < L.D. 391 20.8 68.0 809 98 23.5 17.9 161 291 98.8 16.6 1.33 2.02 11.6 7.1 1.13 16.9 5.93 6.29 56.3 59.4 16.7

58.09 0.53 17.68 3.11 0.3 0.99 0.99 7.41 5.94 0.05 0.00 95 22 5.5 0.74 < L.D. 473 11.8 80.3 1037 119 29 11.1 196 346 110 18.1 0.97 2.25 12.9 8.68 1.36 20.7 6.73 6.98 66.6 69.5 18.4

58.21 0.52 17.93 3.1 0.3 0.68 0.97 7.87 6.28 0.11 0.00 95.97 21.1 < L.D. 0.64 < L.D. 494 5.56 81.5 1058 122 30.8 4.5 202 354 111 18.3 1.01 2.26 13 8.98 1.42 21.9 6.97 7.08 68.1 72.5 20.8

61.2 0.54 18.79 3.38 0.16 0.89 1.74 6.01 6.83 0.18 0.00 99.72 43.5 < L.D. 2.37 < L.D. 267 98.1 38.8 365 48.3 11.5 137 83.3 160 62.3 11 1.63 1.27 7.03 3.71 0.57 7.58 3.07 3.65 35 23.9 6.61

59.44 0.47 18.17 2.91 0.16 0.59 1.37 6.06 7.17 0.12 3.35 99.81 32.2 < L.D. 1.43 5.88 288 22.1 43 414 54.7 13.5 16.6 93.6 180 69.3 12.1 1.4 1.41 7.89 4.24 0.66 8.87 3.64 4.89 38.7 27.2 7.88

60.75 0.53 18.64 2.91 0.24 0.47 1.08 6.95 6.51 0.11 0.00 98.19 24.3 < L.D. 0.96 < L.D. 371 17 61.5 623 81.3 20.5 14.4 131 244 91.8 16.2 1.18 1.93 10.9 6.01 0.93 12.9 5.21 6.05 51.4 40.2 11.9

61.98 0.53 19.09 2.22 0.03 0.39 0.85 4.23 9.57 0.11 0.8 99.8 44.8 5.41 1.67 4.02 458 59.6 48.7 365 47.1 12.9 75.9 162 322 126 20.6 2.12 1.92 9.81 4.16 0.59 8.02 4.09 4.7 56.4 47.4 5.67


pumice lapilli and blocks. This passes up into a more poorly-sorted dirty orange-brown pumice fall breccia which contains common accidental lithic lapilli and blocks (lava, coarse-grained holocrystalline subvolcanic rocks and hydrothermally altered clasts) and porphyritic scoriaceous and glassy juvenile clasts. Thermal alteration halos occur around some lithic blocks. A sparse but pervasive fine ash matrix is present throughout the deposit. At Monte Vezzi (Fig. 1) the pumice fall deposit is overlain by 4 m of dark grey weakly welded diffuse-bedded trachytic ignimbrite

with flattened porphyritic scoria clasts (Fig. 9e). The ignimbrite contains <10–15 vol.% lithic lapilli and blocks up to 30 cm in diameter, which are similar to those in the pumice fall breccia. The most widespread extracaldera deposit is a coarsegrained lithic breccia that outcrops intermittently along the coast of Ischia and on Procida (Fig. 3; Rosi et al. 1988a). The breccia reaches 12 m thick (at Punta Imperatore, Figs. 1 and 3) but is typically <2–4 m thick. It comprises clast-supported lithic lapilli, blocks and boulders,

Bull Volcanol (2008) 70:583–603

591

Table 1 (continued) MEGT 0301

OIS 0321

OIS 0325

OIS 0322

OIS 0323

OIS 0324

SC-MEGT 0313

SC-MEGT 0309

Glass

Pumice

Scoria

Pumice

Pumice

Pumice

Pumice

Pumice

MEGT LMEGT

MEGT pumice fall

MEGT welded ig

La Roia

Capo Grosso

Chiummano

Schiappone

Schiappone

62.19 0.46 18.55 2.65 0.1 0.36 1.18 5.44 7.9 0.08 0.87 99.78 27.6
61.34 0.51 18.55 3.07 0.28 0.43 1 7.7 6.47 0.12 0.00 99.47 21.9
60.97 0.53 18.54 3.04 0.26 0.45 1.02 7.53 6.35 0.06 1.1 99.85 22.4
60.85 0.48 18.83 3.44 0.13 0.95 2.18 5.52 7.25 0.18 0.00 99.81 53.4 9.41 3.8
58.44 0.5 18.21 3.66 0.12 1.05 2.58 4.61 7.35 0.22 3.1 99.84 57.5 9.06 4.06 4.22 207 230 25.7 201 28.2 7.72 281 55.7 105 45.2 8.34 1.73 0.93 5.07 2.42 0.37 4.79 1.98 2.89 28.2 13.4 4.09

54.44 0.72 17.14 6.15 0.12 4.76 7.09 3.3 5.81 0.39 0.00 99.92 145 167 18.7 43.1 153 523 22.8 138 18.5 5.07 658 39.6 78.8 37.3 7.48 1.98 0.86 4.57 1.91 0.29 3.33 1.28 2.15 20.1 8.09 2.58

60.4 0.39 18.12 2.65 0.13 0.59 1.4 5.36 7.32 0.08 3.34 99.78 33.6 6.44 1.57
56.46 0.59 17.63 4.65 0.13 3.27 4.52 4.33 6.39 0.29 0.00 98.26 96.6 64.1 12.3 31.4 209 364 28.5 227 30.3 8.5 501 59.2 107 48.2 8.91 1.81 1.00 5.53 2.78 0.42 5.41 2.08 3.39 32.7 14.7 4.72

subordinate rounded pumice lapilli and juvenile porphyritic glassy lapilli and blocks in a strongly fines-depleted crystallithic matrix (Fig. 9d). The lithic population is similar to that of the breccia in the LMEGT and comprises abundant lava, pyroclastic rocks, hydrothermally altered rocks and coarse-grained feldspar-phyric holocrystalline subvolcanic rocks. Ballistically-emplaced boulders occur at Grotta di Terra (Fig. 1). The breccia locally passes upwards into pumiceous ignimbrite (e.g., at Punta Imperatore, Fig. 3). This breccia is the uppermost unit of the “Unitá di Monte Sant’Angelo” (UMSA) deposit of Rosi et al. (1988a). The

lithic breccia is not seen in context with the welded ignimbrite described above and the depositional chronology remains uncertain. At Porto Santa Maria in SW Campi Flegrei, 2.5 m of thick-bedded greenish ignimbrite outcrops at a similar stratigraphic position to the extracaldera MEGT on Ischia and Procida (Fig. 3; Rosi et al. 1988a). It contains variably vesicular porphyritic pumice lapilli, dense green obsidian blocks and sparse lithic lapilli. Large sanidine (<8 mm) and biotite phenocrysts occur in the ash matrix. Thin discontinuous lenses of lithic-crystal ash are present at the base of

592

Fig. 4 TAS plot for the deposits erupted between 75–50 ka on Ischia. Representative analyses are given in Table 1

the ignimbrite, and several <10 cm long, weakly developed elutriation pipes occur in the middle of the ignimbrite. On the basis of a similar stratigraphic position and similar juvenile material, we include this ignimbrite in the MEGT. The MEGT eruption The MEGT eruption was the largest recorded explosive event on Ischia and is considered to have resulted in caldera subsidence (Vezzoli 1988; Tibaldi and Vezzoli 1998). The extracaldera deposits record phases of the MEGT eruption not found in the exposed intracaldera deposits. We interpret the extracaldera pumice fall deposit in the SE of the island (Figs. 3 and 9e) as recording the opening stages of the eruption and the rapid establishment of a sustained eruption column. The appearance of subvolcanic lithic clasts in this fall deposit may record erosion of the vent or the initial foundering of the caldera. The diffuse-bedded ignimbrite overlying the fall deposit (Figs. 3 and 9d) records fountaining of the eruption column. During this phase, we consider that large-scale caldera subsidence was initiated, dispersing lithic-rich pyroclastic density currents widely across Ischia and Procida. These deposited the coarsegrained lithic breccia both within and outside the caldera. Lithologically-similar lithic breccias are commonly emplaced during caldera collapse (e.g., Breccia Museo, Campi Flegrei, Rosi et al. 1996; Santorini, Druitt and Sparks 1982; Crater Lake, Druitt and Bacon 1986). At least two phases of subsidence are recorded by the intracaldera MEGT: a major phase during the emplacement of the LMEGT, which we correlate with the extracaldera lithic breccia, and a possible second collapse event during the emplacement of the UMEGT. In excess of 200 m of

Bull Volcanol (2008) 70:583–603

ignimbrite (LMEGT and UMEGT) was deposited proximally in the caldera. The partial destruction of the magma chamber is indicated by the abundant coarse-grained holocrystalline subvolcanic rocks, which we interpret as crystal mush from the margins of the chamber. The complex stratigraphy in lower parts of the UMEGT records deposition from an unsteady current(s), which experienced fluctuations in the type of clasts supplied to it through time. Cross-stratified and dune-bedded ignimbrite (Fig. 9c) records emplacement by fully-dilute pyroclastic density currents whose lower flow boundaries were traction-dominated. Pyroclastic density currents generated during this climactic phase of the eruption reached the mainland, 16 km to the NE. The true distribution and volume of the MEGT eruption cannot be established due to limited exposure and the dispersal of most tephra into the sea by pyroclastic density currents and eruption plumes. The presence of convoluted volcaniclastic sediments between the two MEGT intracaldera ignimbrite flow-units (Figs. 8 and 9b) indicates a pause in the eruption, although the duration is not known. The local erosion surface between the LMEGT and UMEGT may record erosion of the unconsolidated upper parts of the LMEGT by UMEGT pyroclastic density currents. The presence of marine sediments beneath the LMEGT perhaps suggests that the centre of the island may have been submerged immediately prior to the MEGT eruption, and it is unclear if the ignimbrite flow-units accumulated subaqueously or subaerially. However, the presence of dune-bedded ignimbrite in the UMEGT requires deposition to have occurred, at least briefly, in a subaerial setting during the eruption. The intracaldera MEGT ignimbrites are similar to the Pavey Ark ignimbrite (Ordovician Scafell caldera, England), which is thought to record the discharge of large volume pyroclastic density currents into a flooded caldera (Kokelaar et al. 2007). A similar scenario might be envisaged for MEGT eruption.

Post-Monte Epomeo Green Tuff volcanism La Roia Tephra The trachyandesite La Roia Tephra is exposed at Monte Sant’Angelo and Punta Imperatore (Fig. 3). It comprises ∼50 cm of well-sorted crudely normally-graded pumice lapilli. It overlies a palaeosol developed in the extracaldera MEGT lithic breccia and passes up into a palaeosol overlain by the distal ashfall deposit of the Chiummano Tephra (Fig. 3). Poor exposure precludes a detailed interpretation. A horizon of weathered pumice lapilli within the upper palaeosol at Monte Sant’Angelo probably represents the remnants of a later (unnamed) pumice fall deposit.

Bull Volcanol (2008) 70:583–603

593

Fig. 5 The Sant’Angelo Tephra. a Graphic log through the Sant’Angelo Tephra as seen on the west side of the Sant’Angelo peninsula (Fig. 1). b Photo of the stratigraphy of the Sant’Angelo Tephra showing relationships to the general stratigraphy (see Log 2, Fig. 3)

The La Roia eruption The La Roia Tephra records fallout from a sustained eruption column. A more detailed interpretation of the eruption dynamics is not presently possible.

porphyritic pumice and dense banded porphyritic scoria clasts and contains abundant hydrothermally altered lithic clasts. Ash-coated pumice lapilli are common. Metre-thick beds of angular clast-supported pumice and scoria are present in the ignimbrite. It is separated from the overlying Chiummano Tephra by a thin brown palaeosol.

Capo Grosso Tephra The Capo Grosso eruption The Capo Grosso Tephra is exposed at Monte Cotto (Fig. 1) and unconformably overlies the >150 ka Lower Scarrupata di Barano Formation (Vezzoli 1988). It is a >40 m thick white-weathering diffuse-bedded trachyandesite ignimbrite (Fig. 10a). It is characterised by variably vesicular

The Capo Grosso Tephra was generated by an explosive eruption that generated at least one sustained pyroclastic density current. Its exact position in the stratigraphy is not well constrained, but the thin palaeosol between it and the

594

Bull Volcanol (2008) 70:583–603

Fig. 6 The lower pumice fall deposits. Measured sections through the major pumice fall deposits (Olummo, Tisichiello and Porticello Tephras) on Ischia as seen at Grotta di Terra, SE Ischia (Fig. 1)

overlying Chiummano Tephra perhaps does not suggest a long repose period; the two deposits also exhibit texturally similar juvenile material and crystal and lithic contents. Its vent position and dispersal are not known. Chiummano Tephra The Chiummano Tephra comprises a succession of bedded, cross-stratified and dune-bedded lithic-rich ignimbrites (Figs. 10b and 11). It is best exposed at Scarrupata di Barano, but it also outcrops discontinuously along the

southern coast where it forms an important marker horizon (see Fig. 3). It varies from several centimetres thick to 7 m thick and is locally cut out by an extensive erosion surface. At Scarrupata di Barano it comprises four units detailed in Fig. 11. Juvenile material comprises banded and homogeneous pale-grey and black basaltic-trachyandesite pumice and scoria (Fig. 4). The lithic population comprises lavas, coarse-grained subvolcanic rocks and hydrothermally altered rocks. Fine-grained accretionary lapilli-bearing tuffs occur at the top of the succession (Fig. 11). At Monte Cotto (Fig. 1), the Chiummano Tephra is comprised of coarse

Bull Volcanol (2008) 70:583–603

595

1980), and instead we interpret the Chiummano Tephra as the flank deposits of a maar-type volcano. Schiappone Tephra

Fig. 7 The lower pumice fall deposits. a Base of Tisichiello Tephra with lapilli-tuff layer interpreted as an ignimbrite. b The Porticello Tephra (see Fig. 6)

pumice and scoria fall deposits, while further west it outcrops as a <12 cm thick layer of normally-graded lithic-rich ash (e.g., Sant’Angelo and Punta Imperatore; Fig. 1). The Chiummano eruption We interpret the abundant stratified, cross-stratified and dune-bedded ignimbrites (Fig. 10b), the high accidental lithic content and the presence of accretionary lapilli as resulting from dominantly phreatomagmatic explosivity that generated numerous fully dilute pyroclastic density currents (Taalian activity of Kokelaar 1986). The high lithic content throughout suggests that the phreatomagmatic explosions occurred within an aquifer. The presence of ballistically emplaced boulders at Scarrupata di Barano (Fig. 1) positions the vent within the south-east of the island. During the eruption, a plume deposited lithic ash over the south and south-west of the island (e.g., Sant’Angelo and Punta Imperatore; Fig. 3). The Chiummano Tephra was previously interpreted as alluvial deposits (see Vezzoli 1988), however, the high lithic content, accretionary lapilli and dune-bedded ignimbrites are typical of phreatomagmatic explosions and maar volcanism (e.g., Self et al.

The Schiappone Tephra outcrops between Monte Cotto and Punta della Pisciazza (Fig. 1), and at Monte Vico, Punta Imperatore and Procida (Fig. 3). It comprises a complex sequence of bedded trachyandesitic-trachytic (Fig. 4) pumice fall deposits and pyroclastic density currents deposits (Member A) overlain by a ∼60 m-thick white-weathering trachytic ignimbrite (Member B; Fig. 12). It was previously considered to be an extracaldera ignimbrite of the MEGT (Rosi et al. 1988a; Vezzoli 1988). Member A is well exposed along the southern coast between Monte Cotto and Punta della Pisciazza (Fig. 1). It is rarely fully preserved, and is commonly cut by an erosion surface at the base of the Member B. Member A is complex and it is difficult to trace individual units around the island due to the rapid alternation of fall and flow deposits (Fig. 12). We mainly describe it with reference to the outcrop at Monte Vezzi (Log 3, Fig. 12). The lowermost pumice fall unit contains cm-thick horizons of pink-ash-coated pumice lapilli, which provide useful marker horizons in the tephra across to Procida (Fig. 12). This is overlain by <30 cm of stratified, cross-stratified and dune-bedded lithic-rich ignimbrite containing rare impact sags and thin ash beds with accretionary lapilli. Lithic clasts are commonly intensely hydrothermally altered. Overlying the cross-stratified layers is a sequence of decimetre-thick pumice lapilli beds separated by centimetre-thick poorly sorted ash beds. These pass up into a laterally discontinuous layer of lenticular bedded and cross-bedded pumice lapilli which contains abundant rounded lapilli (e.g., Punta San Pancrazio; Figs. 12 and 13a). The upper >3 m of Member A comprise a thick bedded pumice breccia, in which individual pumices become progressively denser and more flattened upwards (Fig. 12). At Scarrupata di Barano (Figs. 1 and 13b) the fall deposit is intensely welded and overlies several metres of diffuse-bedded ignimbrite which was coevally emplaced with the upper parts of the Member A (see Fig. 12). Member A reaches >25 m thick in steep cliffs at Punta Imperatore (Fig. 3) and is cut out by a major scour surface at the base of Member B. Here, Member A comprises 2 m of bedded lapilli-tuff with rounded pumice lapilli and blocks, scours and low-angle pumice lenses, overlain by ∼20 m of thick-bedded pumice fall breccia (Fig. 12). Two metre-thick brown-orange scoria layers occur in the upper parts. Correlations between beds in Member A at Punta Imperatore and those in the SE are difficult due to an absence of intermediate exposures. The Member B ignimbrite unconformably overlies Member A. It reaches >60 m thick (Scarrupata di Barano,

596

Bull Volcanol (2008) 70:583–603

Bull Volcanol (2008) 70:583–603

ƒFig. 8

The Monte Epomeo Green Tuff. Composite vertical sequence through the Lower and Upper Monte Epomeo Green Tuff as seen at Rione Bocca up to Monte Epomeo (Fig. 1)

Fig. 1) but is typically 5–20 m thick (see Figs. 12 and 13c). Its upper surface is commonly absent due to erosion. The ignimbrite is partly indurated and generally massive to diffuse bedded. It comprises pumice lapilli and blocks in an abundant ash and crystal matrix. Subordinate lithic lapilli and blocks occur throughout the ignimbrite, and are locally concentrated in metres-wide lenses (e.g., Scarrupata di Barano; Fig. 1). Angular pumice blocks in lenses and large, Fig. 9 Deposits of the Monte Epomeo Green Tuff eruption. a The western side of the Monte Epomeo resurgent block showing the contact between the Lower and Upper MEGT. vs = volcaniclastic sediments; ms = marine sediments. b Convoluted, laminated volcaniclastic sediments in between the Lower and Upper MEGT. c Crossstratified ignimbrite at the base of the UMEGT, Rione Bocca (Fig. 1). 10 cm divisions on rule. d Lithic lag breccia at Ciraccio, Procida. Breccia is approximately 4 m thick and is bound by palaeosols (see Log 6, Fig. 3). e Pumice fall deposit and stratified welded ignimbrite at Cavone dei Camaldoli (Fig. 1). Contact between units dips towards the right. Bag for scale. f Non-welded bedded ignimbrite at Porto Santa Maria, Campi Flegrei (Fig. 1). This is interpreted as a distal equivalent of the MEGT intracaldera ignimbrites (see Log 8, Fig. 3). 10 cm divisions on rule

597

metre-scale dune bedforms occur in the ignimbrite along the eastern coast (Grotta di Terra; Fig. 1). On Procida, Member B comprises 3 m of cross-stratified and diffusebedded pumice lapilli-poor ignimbrite (Fig. 12). The Schiappone eruption The Schiappone Tephra records a complex explosive eruption that showered out pumice and dispersed pyroclastic density currents across much of Ischia and Procida. The lowermost layers of Member A record the synchronous

598

Bull Volcanol (2008) 70:583–603

Fig. 10 The Capo Grosso and Chiummano Tephras. a The Capo Grosso at Monte Cotto (see Fig. 1). The ignimbrite of the Schiappone Tephra is seen in the background. b Dune bedform (arrowed) in the Chiummano Tephra, Scarrupata di Barano. Current towards right. 10 cm divisions on rule

fallout of pumice from a sustained eruption column and the generation of short-lived pyroclastic density currents (Fig. 12). We infer that the abundant hydrothermally altered lithic clasts derive from the fragmentation of conduit walls, which may have been initiated by vent-clearing eruptions or by phreatomagmatic explosions. The upper pumice fall breccia records an increase in the eruption intensity, the growth of the eruption column and the fallout of abundant pumice blocks. The welded horizon (Figs. 12 and 13b) may record either increased deposition rates, fallout from a lower column or higher eruptive temperatures of juvenile material. The limited outcrop precludes the construction of isopach maps and the vent position remains unknown: the thick and coarse grained-nature of Member A at Punta Imperatore (Fig. 3) perhaps indicates that the eruption was sourced in the west of the island. Member B records eruption column fountaining and the dispersal of pyroclastic density currents across much of Ischia and Procida. These currents widely eroded the underlying Member A (Fig. 12). Clast-supported lenses of angular pumice lapilli and blocks are interpreted as either contemporaneous fallout from an eruption column into a moving pyroclastic density current, as material eroded from Member A. The absence of rounding on most of these clasts indicates that they deposited soon after entering the flow. Local lithic breccia horizons record the entrainment of lithic material either at source due to vent erosion or during transport. The finer-grained outcrops of the ignimbrite on Procida (Fig. 12) are considered to be more distal. The volume of Schiappone tephra is not known, but it is one of the most widespread and thickest deposits on the island.

Discussion and conclusions Comparison with previous stratigraphic schemes on Ischia In a seminal study, Vezzoli (1988) established the first substantial island-wide stratigraphic framework for Ischia, which built on previous stratigraphic studies of local parts of the island (Forcella et al. 1981, 1982; Rosi et al. 1988a). The stratigraphy presented in this paper represents a revision

Fig. 11 Measured section through the Chiummano Tephra at Scarrupata di Barano (Fig. 1). The tephra comprises a series of stratified, cross-stratified and dune-bedded lithic-rich ignimbrites. Ballistically-emplaced boulders are present in Member B

Bull Volcanol (2008) 70:583–603

599

Fig. 12 Measured sections through the Schiappone Tephra on Ischia and Procida showing the complex nature of basal fall deposit. The presence of several marker horizons, such as the pink ash-coated layers in the basal fall and the lithic-rich cross-stratified ignimbrites

allow the unit to be confidently traced across the region. Of interest is the intimate interbedding of fallout and pyroclastic density current deposits and the presence of flow-modified fall deposits (e.g., crossstratified pumice lapilli in Log 5; see Fig. 13a). For key see Fig. 11

to the published stratigraphy for the period ∼75–50 ka (see Table 2) and below, we briefly discuss the major changes and outline the supporting evidence. The Sant’Angelo Tephra (Fig. 5) as defined in this work correlates to the pumice fall deposits and ignimbrites of the Monte Sant’Angelo volcanic centre (Vezzoli 1988) and the MEGT “explosion breccias” (Vezzoli 1988; UMSA of Rosi et al. 1988a). As defined in this paper, it does not include the uppermost heterolithic breccia of the UMSA. We recognise an unconformity between this heterolithic breccia and the underlying monomict breccias, or “explosion breccias” (which we here re-interpret as block-and-ash flow deposits). We retain the correlation between the heterolithic lag breccia and the MEGT (Rosi et al. 1988a), but note that the lag breccia passes upwards into a palaeosol that is overlain by the La Roia Tephra, which in turn passes up into a palaeosol that contains a second (unnamed) highly weathered pumice layer (Fig. 3). This is overlain by the Schiappone Tephra.

Previously, all these deposits were included in the MEGT (see Rosi et al. 1988a, b; Vezzoli 1988). However, the discovery of these palaeosols indicates that the sequence comprises several different eruptions, separated in time. The pumice fall deposits (Mago, Olummo, Tisichiello and Porticello Tephras) correlate to the Pignatiello Formation of Vezzoli (1988; Table 2). We reinterpret some palaeosols described by previous workers (Rosi et al. 1988a), as pyroclastic density current deposits (Fig. 6) and have consequently reordered the stratigraphy. The pumice fall deposits are complex and rapidly change their character laterally due to intercalated pyroclastic density current deposits and syn-depositional reworking on steep slopes. The Porticello Tephra correlates with “Pomici pliniane C” of Rosi et al. (1988a). The Capo Grosso and La Roia tephras have not previously been recognised as separate eruptive units and were considered to be part of the MEGT (Vezzoli 1988).

600

Bull Volcanol (2008) 70:583–603

RFig. 13

The Schiappone Tephra. a Dune-bedded lapilli-tuff (arrowed) in the basal pumice fall deposit at Punta di San Pancrazio (Fig. 1). 10 cm divisions on rule. b Transition from diffuse-bedded ignimbrite up into welded fall deposit (black) at Scarrupata di Barano (Fig. 1). c Diffuse-bedded ignimbrite at Punta di San Pancrazio. Basal pumice fall deposits sit in an unconformity cut into the underlying deposits

The Chiummano Tephra is a newly recognised pyroclastic unit (Figs. 10b and 11) that was previously interpreted as epiclastic alluvial deposits (Vezzoli 1988). The Schiappone Tephra was previously considered to be the extracaldera equivalent of the MEGT (Rosi et al. 1988a; Vezzoli 1988). However, it is lithologically different to the intracaldera MEGT ignimbrites, is younger and is separated from the extracaldera MEGT deposits by several intercalated palaeosols and pyroclastic deposits (Fig. 3). The radiometric data published by Gillot et al. (1982) and Vezzoli (1988) back up this assertion, and indicate a clustering of ages at ∼56 ka for the intracaldera MEGT and 50 ka for the extracaldera Schiappone Tephra, suggesting a time gap of 5–6 ka between the two eruptions (Table 3). However, overlapping anomalous dates for one sample in each cluster (see Table 3) led previous workers to consider that the two age sets were in analytical error of each other (see Vezzoli 1988). Notwithstanding this, the stratigraphic and lithologic evidence presented here indicates that the Schiappone Tephra is not related to the MEGT and is a younger deposit. The ∼6 ka difference in the two sets of clustered ages for the two units makes sense in the light of this new data. These changes may require a re-evaluation of correlations between terrestrial deposits and trachytic marine and terrestrial distal ash layers across the Central and Eastern Mediterranean region (see Keller et al. 1978; Paterne et al. 1988; Narcisi 1996). The stratigraphic data contrasts with previous ideas of a protracted period of quiescence on Ischia between ∼75 and 50 ka BP (e.g., Vezzoli 1988; Civetta et al. 1991). Alluvial deposits previously invoked as evidence for a protracted pause in volcanic output are here reinterpreted as pyroclastic deposits. We consider that the ten eruptions represent a minimum because the number of recorded eruptions is proportional to the amount of available exposure, which diminishes rapidly back through geologic time on Ischia (see Fig. 1): the deposits (excluding intracaldera deposits) presently outcrop over <2% (∼1 km2) of the present island. The products from small-volume eruptions vented in other parts of the island may not be recorded in the present exposures. Also, most deposits outcrop on steep topography that is not suitable for the preservation of loose pyroclastic material (e.g., on the Monte Vico, Monte Cotto and Monte Vezzi lava domes; Fig. 1). Additionally, major unconformities occur within the sequence and large chunks of the stratigraphy are absent at many localities (Fig. 3). Finally, the stratigraphic position and provenance of several pyro-

Bull Volcanol (2008) 70:583–603

601

Table 2 Comparisons between the new stratigraphy presented in this work with that of previous workers This work

Vezzoli (1988)

Rosi et al. (1988a)

Schiappone Tephra Chiummano Tephra Capo Grosso Tephra La Roia Tephra Intracaldera MEGT Lithic lag breccia

MEGT (66)

MEGT

Epivolcanic deposits (71b)

N/A

MEGT (67)

N/A

Not recognised MEGT (66)

N/A N/A

Epivolcanic deposits (71a)

Pomici pliniane A e breccia associata/ Unitá di Monte Sant’Angelo

Extracaldera ignimbrite/fall deposit Porticello Tephra Tisichiello Tephra Olummo Tephra Mago Tephra Sant’Angelo Tephra

MEGT (67)/Pignatiello Fm. (69) Pomici pliniane C

Pignatiello Fm. (69) Monte Sant’Angelo volcanic centre (77); MEGT (67)

Unitá di Monte Sant’Angelo

Numbers in parentheses refer to unit numbers of Vezzoli (1988).

clastic units outcropping on Ischia, Procida and Monte di Procida (see Fig. 3) remains uncertain due to poor exposure. Volcanic and volcanotectonic evolution of Ischia The new stratigraphic data contribute to a clearer understanding of the volcanic and structural evolution of Ischia during the Late Pleistocene. The exposed volcanic products

comprise trachytic and phonolitic lava domes, lavas and pyroclastic deposits (Fig. 1; Vezzoli 1988), and must have resulted from a protracted early phase(s) of magma generation, differentiation and eruption. The presence of several scattered volcanic centres fed by evolved magmas suggests the existence of a substantial differentiated magma chamber under the region prior to the MEGT eruption. A caldera may have existed prior to the MEGT eruption, although evidence is sparse (Vezzoli 1988). The Sant’Angelo tephra and the Tisichiello tephra preserve block-and-ash flow deposits suggesting the periodic growth and destruction of lava domes at volcanic centres in the south of the island. Episodic sustained explosive activity at these volcanic centres generated eruption columns that showered out pumice across the island. The palaeogeography at the time would have been substantially different to that seen today. Marine volcaniclastic deposits below the LMEGT (Rione Bocca; Figs. 1 and 8) indicate a pre-existing below-wavebase marine depression in central parts of the island. This was probably surrounded by discrete volcanic centres (e.g., lava domes and flows of Carta Romana-Monte Cotto, Sant’Angelo, Monte Vico and Punta Imperatore; Fig. 1; Vezzoli 1988). Geophysical surveys to the west of Ischia have revealed the submarine remnants of a large volcanic complex (Bruno et al. 2002), which may have been the source for some of the eruptions during this period. Mean global sea levels were 50–80 m below present levels (see McGuire et al. 1997 and references therein) and Ischia may have been connected to the mainland by an isthmus that included parts of the present island of Procida (see isobaths on inset, Fig. 3). This phase of eruptive activity is inferred to have continued up until the MEGT eruption at ∼55 ka. The MEGT eruption was a complex, moderate to large-volume explosive volcanic eruption that most probably resulted in caldera subsidence across all or part of the island and partially destroyed the magma chamber. The MEGT

Table 3 Published K/Ar dates for selected pyroclastic units on Ischia Unit (of Vezzoli 1988)

Sample no.

Location

Sampled material

Age

New unit (this study)

MEGT (extracaldera) MEGT (extracaldera) MEGT (extracaldera)

102 108 35Z

Sant’Angelo peninsula Scarrupata di Barano Scarrupata di Barano

Pumice, sanidine Pumice, sanidine Obsidian

Schiappone Tephra Schiappone Tephra Schiappone Tephra

MEGT (intracaldera) MEGT (intracaldera)

38Eaa 38Eba

Monte Epomeo summit Monte Epomeo summit

Ignimbrite, biotite Ignimbrite, sanidine

MEGT (intracaldera)

114Ba

Pietra dell’Acqua

Ignimbrite, sanidine

49,600±1,200; 50,900±1,500 49,000±1,100 50,100±1,300; 51,600±1,300; 55,800±1,800 56,800±2,800 54,200±2,200; 56,700±23,00; 58,000±41,00 51,500±26,00; 56,700±25,00

UMEGT UMEGT UMEGT

Taken from Vezzoli (1988); dates for the Schiappone Tephra, previously correlated with the MEGT, are 5–6 ka younger than those obtained for the intracaldera MEGT. a Taken from Gillot et al. (1982).

602

eruption comprised two phases of pyroclastic fountaining that deposited >200 m of ignimbrite in the deepening caldera. Volcaniclastic sediments between the two ignimbrite flow-units suggest that the subsiding caldera was partially open to the sea during the eruption. The pyroclastic density currents may have been discharged into a growing caldera basin that was inundated with seawater. Following the MEGT eruption seawater flooded the caldera to 70–120 m depth (Barra et al. 1992). Post-MEGT activity comprised a number of phreatomagmatic and magmatic eruptions that generated sustained eruption columns and pyroclastic density currents (Capo Grosso, Chiummano and Schiappone eruptions). The presence of thick ponded ignimbrites along the SE coast indicates topography (now absent) to the south of Scarrupata di Barano (Fig. 1). This topography may have been destroyed during large-scale sector-collapse of the south side of the island (see Chiocci and de Alteriis 2006). Post-MEGT volcanism differed from pre-MEGT volcanism. Differences include an apparent shift towards phreatomagmatic or partly phreatomagmatic eruptions (e.g., Capo Grosso, Chiummano and Schiappone eruptions; see also younger C. S. Costanzo and Citara eruptions of Vezzoli 1988). This probably resulted from subsidence during caldera collapse which could have facilitated the access of seawater to subsequent erupting magmas. Such processes may explain the common presence in the post-MEGT deposits of hydrothermally-altered rocks, which we infer are the exploded remnants of hydrothermal systems, and holocrystalline subvolcanic accidental lithic clasts from the fragmented magma chamber. Additionally, magma compositions changed from phonolite and trachyte to trachyte-trachyandesite and basaltictrachyandesite following the MEGT eruption: this may indicate the replenishment of the partially destroyed magmatic system by more mafic magmas or changes in magma transport to the surface following subsidence.

Comparison with the present volcanic state of the island The studied period saw the largest volume and most devastating eruptions recorded on Ischia. Several eruptions potentially devastated the entire island and at least two produced pyroclastic density currents that laid down thick pyroclastic density current deposits over neighbouring areas. The MEGT resulted in caldera subsidence across part, or all, of the island. The type and magnitude of volcanic activity outlined here contrasts sharply with more recent (10 ka– 1302 AD) activity on Ischia, which comprised numerous (>40) small-volume effusive and explosive eruptions. These Holocene-to-recent eruptions appear to have impacted only Ischia or only parts of Ischia. The differences in the scale and impact of volcanism on Ischia during the last 75 ka resulted

Bull Volcanol (2008) 70:583–603

from major changes in the magmatic system and volcanotectonic structure of the island. Volcanism ∼75–50 ka BP was influenced by the existence and dynamics of a large differentiated magma chamber(s). Following collapse and partial destruction of this chamber, the timing and distribution of volcanism has been instead dictated by caldera resurgence: volcanism has been concentrated along the margins of the resurgent block (Orsi et al. 1991) and the timing of eruptions is closely linked to periods of uplift (de Vita et al. 2006). A substantial magma chamber, similar to that under Ischia prior to ∼55 ka, is not thought to presently exist at depth (e.g., Piochi et al. 1999). Future eruptions on Ischia will probably be of a similar low magnitude and impact to that which has characterised Holocene volcanism. However, an imperfect understanding of the long-term evolution of the magmatic system beneath Ischia means that we lack information on the repose periods of MEGTtype eruptions on Ischia and also on what may trigger such destructive events. It seems expedient for future research to explore these themes. Acknowledgements RJB acknowledges funding from the EU Volcano Dynamics Research Training Network (Fifth Framework Programme). The authors thank A. Carandente and P. Belviso for the support in the laboratory, and I. Arienzo, L. Civetta, E. Marotta and F. Sansivero for assistance and discussion in the field. R. Cioni, J. Keller, R. Sulpizio and an anonymous reviewer are thanked for comments that substantially improved earlier versions of the manuscript. This research was partly carried out under the framework of the Italian INGV-DPC 2004–2006 program, sub-project V3-3 Ischia.

References Barra D, Cinque A, Italiano A, Scorziello R (1992) Il Pleistocene superiore marino di Ischia paleoecologia e rapporti con l’evoluzione tettonica recente. Studi Geologici Camerti, Volume Speciale:231–243 Brown RJ, Barry TL, Branney MJ, Pringle MS, Bryan SE (2003) The quaternary pyroclastic succession of southern Tenerife, Canary Islands: explosive eruptions, related subsidence and sector collapse. Geol Mag 140:265–288 Bruno P, de Alteriis G, Florio G (2002) The western undersea section of the Ischia volcanic complex (Italy, Tyrrhenian Sea) inferred by marine geophysical data. Geophys Res Let 29(9):57; 1–4 (Art. no. 1343) Buchner G, Italiano A, Vita-Finzi C (1996) Recent uplift of Ischia, Southern Italy. In: Jones WJ, Jones AP, Neuberg J (eds) Volcano instability on the earth and other planets. Geol Soc Spec Pub 110:249–252 Caliro S, Panichi C, Stanzione D (1999) Variation in the TDC isotope composition of thermal waters of the island of Ischia (Italy) and its implications for volcanic surveillance. J Volcanol Geotherm Res 90:219–240 Cas RAF, Wright JV (1987) Volcanic successions. Chapman & Hall London, p 528 Chiocci FL, de Alteriis G (2006) The Ischia debris avalanche: first clear submarine evidence in the Mediterranean of a volcanic island prehistoric collapse. Terra Nova 18:202–209

Bull Volcanol (2008) 70:583–603 Chiodini G, Avino R, Brombach T, Caliro S, Cardellini C, de Vita S, Frondini F, Granirei D, Marotta E, Ventura G (2004) Fumarolic and diffuse soil degassing west of Mount Epomeo, Ischia, Italy. J Volcanol Geotherm Res 133:291–309 Civetta L, Gallo G, Orsi G (1991) Sr- and Nd-isotope and trace element constraints on the chemical evolution of Ischia (Italy) in the last 55 ka. J Volcanol Geotherm Res 46:213–230 Civetta L, De Vivo A, Orsi G, Polara G (1999) Il vulcanismo a Ischia in età greco-romana secondo le evidenze geologiche e le testimonianze storico-letterarie. Vichiana, Loffredo Editore. Napoli 15–32 Crisci GM, de Francesco AM, Mazzuoli R, Poli G, Stanzione D (1989) Geochemistry of the recent volcanics of Ischia Island, Italy: evidences of crystallization and magma mixing. Chem Geol 78:15–33 Cubellis E, Carlino S, Iannuzzi R, Luongo G, Obrizzo F (2004) Management of historical seismic data using GIS: the island of Ischia (southern Italy). Nat Hazard 33:379–393 De Astis G, Pappalardo L, Piochi M (2004) Procida volcanic history: new insights into the evolution of the Phlegraean volcanic district (Campania region, Italy). Bull Volc 66:622–641 de Vita S, Sansivero F, Orsi G, Marotta E (2006) Cyclical slope instability and volcanism related to volcano-tectonism in resurgent calderas: the Ischia island (Italy) case study. Engen Geol 86:148–165 Druitt TH, Bacon CR (1986) Lithic breccia and ignimbrite erupted during the collapse of Crater Lake Caldera, Oregon. J Volcanol Geotherm Res 25:1–32 Druitt TH, Sparks RSJ (1982) A proximal ignimbrite breccia facies on Santorini. J Volcanol Geotherm Res 13:147–171 Druitt TH, Mellors RA, Pyle DM, Sparks RSJ (1989) Explosive volcanism on Santorini, Greece. Geol Mag 126:95–126 Fisher RV, Schmincke H-U (1984) Pyroclastic rocks. Springer, Berlin Heidelberg New York, p 472 Forcella F, Gnaccolini M, Vezzoli L (1981) Stratigrafia e sedimentologia dei depositi piroclastici affioranti nel settore sud-occidentale dell’isola d’Ischia. Riv It Paleaont Strat 87:329–366 Forcella F, Gnaccolini M, Vezzoli L (1982) I depositi piroclastici del settore sud-orientale dell’isola d’Ischia (Italia). Riv It Paleaont Strat 89:135–170 Gillot PY, Chiesa S, Pasquare G, Vezzoli L (1982) <33000 yr K/Ar dating of the volcano-tectonic horst of the isle of Ischia, Gulf of Naples. Nature 229:242 Inguaggiato S, Pecoraino G, D’Amore F (2000) Chemical and isotopical characterisation of fluid manifestations of Ischia Island (Italy). J Volcanol Geotherm Res 99:151–178 Keller J (1980) The island of Salina. Rend Soc It Mineral Petrog 36:489–524 Keller J, Ryan WBF, Ninkovich D, Altherr R (1978) Explosive volcanic activity in the Mediterranean over the past 200,000 yr as recorded in deep-sea sediments. Geol Soc Am Bull 89:591–604 Kokelaar BP (1986) Magma-water interactions in subaqueous and emergent basaltic volcanism. Bull Volcanol 48:275–289 Kokelaar BP, Raine P, Branney MJ (2007) Incursion of a largevolume, spatter-bearing pyroclastic density current into a caldera lake: Pavery Ark ignimbrite, Scafell Caldera, England. Bull Volc (in press) DOI 10.1007/s00445-007-0118-5 McGuire WJ, Howarth RJ, Firth CR, Solow AR, Pullen AD, Saunders SJ, Stewart IS, Vita-Finzi C (1997) Correlation between rate of sea-level change and frequency of explosive volcanism in the Mediterranean. Nature 389:473–476 Narcisi B (1996) Tephrochronology of a Late Quaternary lacustrine record from the Monticchio Maar (Vulture volcano, Southern Italy). Quat Science Rev 15:155–165

603 Narcisi B, Vezzoli L (1999) Quaternary stratigraphy of distal tephra layers in the Mediterranean—an overview. Glob Planet Change 21:31–50 Orsi G, Gallo G, Zanchi A (1991) Simple-shearing block resurgence in caldera depressions. A model from Pantelleria and Ischia. J Volcanol Geotherm Res 47:1–11 Orsi G, Piochi M, Campajola L, D’Onofrio A, Gialanella L, Terrasi F (1996) 14C geochronological constraints for the volcanic history of the island of Ischia (Italy) over the last 5000 years. J Volcanol Geotherm Res 71:249–257 Orsi G, Patella D, Piochi M, Tramacere A (1999) Magnetic modelling of the Phlegraean Volcanic District with extension to the Ponza archipelago, Italy. J Volcanol Geotherm Res 91:345–360 Orsi G, de Vita S, Di Vito M, Isaia R, Nave R, Heiken G (2003) Facing volcanic and related hazards in the Neapolitan area. In: Heiken G, Fakundiny R, Sutter J (eds) Earth sciences in cities. American Geophysical Union, Washington, pp 121–170 Panichi C, Bolognesi L, Ghiara M, Noto P, Stanzione D (1992) Geothermal assessment of the island of Ischia (Southern Italy) from isotopic and chemical-composition of the delivered fluids. J Volcanol Geotherm Res 49:329–348 Paterne M, Guichard F, Labeyrie J (1988) Explosive activity of the South Italian volcanoes during the past 80,000 years as determined by marine tephrochronology. J Volcanol Geotherm Res 34:153–172 Piochi M, Civetta L, Orsi G (1999) Mingling in the magmatic system of Ischia (Italy) in the past 5 ka. Miner Petrol 66:227–258 Poli S, Chiesa S, Gillot P-Y, Gregnanin A, Guichard F (1987) Chemistry versus time in the volcanic complex of Ischia (Gulf of Naples, Italy): evidence of successive magmatic cycles. Contrib Min Pet 95:322–335 Poli S, Chiesa S, Gillot P-Y, Guichard F, Vezzoli L (1989) Time dimension in the geochemical approach and hazard estimates of a volcanic area: the isle of Ischia. J Volcanol Geotherm Res 36:327–335 Rittmann A (1930) Geologie der Insel Ischia. Z Vulk Erg Bd 6:265 Rosi M, Sbrana A, Vezzoli L (1988a) Stratigrafia delle isole di Procida e Vivara. Boll GNV 4:500–525 Rosi M, Sbrana A, Vezzoli L (1988b) Correlazioni tefrostratigrafiche di alcuni livelli di Ischia, Procida e Campi Flegrei. Mem Soc Geol It 41:1015–1027 Rosi M, Vezzoli L, Aleotti P, De Censi M (1996) Interaction between caldera collapse and eruptive dynamics during the Campanian Ignimbrite eruption, Phlegraean Fields, Italy. Bull Volcanol 57:541–554 Sartori R (2003) The Tyrrhenian back-arc basin and subduction of the Ionian lithosphere. Episodes 23:217–221 Scott AC, Glasspool IJ (2005) Charcoal reflectance as a proxy for the emplacement temperature of pyroclastic flow deposits. Geology 33:589–592 Self S, Kienle J, Huot J-P (1980) Ukinrek maars, Alaska, II. Deposits and formation of the 1977 craters. J Volcanol Geotherm Res 7:39–65 Tibaldi A, Vezzoli L (1998) The space problem of caldera resurgence: an example from Ischia Island, Italy. Geol Rund 87:53–66 Tibaldi A, Vezzoli L (2004) A new type of volcano flank failure: the resurgent caldera sector collapse, Ischia, Italy. Geophys Res Let 31:L14605 Vezzoli L (1988) Island of Ischia: Quaderni de La Ricerca Scientifica, 114(10). Consiglio Nazionale delle Ricerca, Rome, p 133 Wulf S, Kraml M, Brauer A, Keller J, Negendank JFW (2004) Tephrochronology of the 100 ka lacustrine sediment record of Lago Grande di Monticchio (southern Italy). Quat Int 122: 7–30