<<

Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks

R. V. Fisher

SUMMARY: The type, relative abundance and stratigraphical relationships of volcanic rocks that comprise island volcanoes are a function of (i) depth of extrusion beneath water, (ii) composition, and (iii) -water interactions. The water depth at which explosions can occur is called the pressure compensation level (PCL) and is variable. Explosive eruptions that occur above the PCL and below sealevel can give rise to abundant hydroclastic and pyroclastic debris. Below the PCL, clastic material cannot form explosively; it forms from lava by thermal shock. The volcaniclastic products are widely dispersed in basins adjacent to extrusion sources by three principal kinds of marine transport processes. These are slides, sediment gravity flows and suspension fallout. Volcaniclastic debris can be derived in subaqueous and subaerial-to-subaqueous environments (i) directly from eruptions, (ii) from remobilization of juvenile volcaniclastics, or (iii) from epiclastic material which initially develops above sealevel. Sediment gravity flows (fluids driven by sediment motion) exhibit the phenomenon of flow transformation. This term is used here for the process by which (i) sediment gravity flow behaviour changes from turbulent to laminar, or vice versa, within the body of a flow, (ii) flows separate into laminar and turbulent parts by gravity, and (iii) flows separate by turbulent mixing with ambient fluid into turbulent and laminar parts. Dominant kinds of subaqueous volcaniclastic sediment gravity flows are debris flows, hot or cold pyroclastic flows and turbidites. Fine grained material can be thrown into suspension locally during flow transformations or underwater eruptions, but thin, regionally distributed subaqueous fallout is mostly derived from siliceous Plinian eruptions.

Volcaniclastic rocks in marine sequences occur marine volcaniclastic deposits are most abundant in many kinds of sedimentary environments and near island arcs and ocean islands because large tectonic settings (Table 1). Some are the result volumes of volcaniclastic material are trans- of eruptions on land which deliver fallout ash, ported from land into the sea during eruptions lava and pyroclastic flows into water. Others and during subsequent . Volcanogenic are derived from underwater eruptions which sedimentation and tectonics, an area of consid- extrude lava flows and volcanic fragments of erable research, is reviewed by Mitchell & various kinds, or are pyroclastic, hydroclastic or Reading (1978). epiclastic materials reworked from land or remobilized under water. This review is mostly concerned with subaqueous sediment gravity Underwater eruptions flow deposits, because of their abundance in Whether underwater eruptions are effusive or marine environments adjacent to volcanic explosive is determined by (i) the depth (pres- regions. sure) of the water column (Fig. 1), (ii) the com- The site of most voluminous is at position of the magma, especially amounts of divergent plate boundaries where basaltic sheet , and (iii) the extent of interaction be- flows, pillow and smaller amounts of pil- tween magma and water. Subsequent transport low and hyaloclastites are formed. An depends upon slope, which in turn partly estimated 4-6 km 3 of such material is added depends upon the growth rate of the each year to the Earth's crust at mid-ocean and the initial lava-to-clastic ratio of extruded ridges (Nakamura 1974). The second important products. The manner of transport also depends site of volcanism is at convergent plate bound- upon whether or not eruption columns can aries. Basaltic and andesitic island arcs develop develop. When vapour pressure in the magma at converging oceanic plates; converging exceeds water pressure, vesiculation com- oceanic and continental plates give rise to mences. As depth decreases, vesicles become dominantly andesitic to rhyolitic volcanic chains more abundant (Moore 1965, 1970; Moore & on the edges of continents. Perhaps the most Schilling 1973) and, at a shallow level, explo- varied and complex environments are the dif- sions caused by exsolution of magmatic vol- ferent kinds of back-arc settings. A third import- atiles can occur. This depth, which is variable, ant site of oceanic volcanism is represented by is here called the pressure compensation level intra-plate and ocean islands. Sub- (PCL) and 'volatile fragmentation depth' by Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

TABLE 1. Simplified partitioning of environment, kind of extrusion, transport and emplacement processes and some typical pyroclastic deposits

Directly from eruptions

Environment Eruptions Transport Emplacement Deposits

Subaqueous Effusive Lava flows Congealing flows Massive flows; pillow ; hyaloclastites. (marine, ~ (shallow lacustrine, or deep) sub-ice) Explosive Dilute suspension Suspension fallout Thin, well sorted, normally graded beds. May (shallow) in water be reworked by bottom currents. Local occurrence near source.

Pyroclastic flows Suspension fallout Thin, fine grained well sorted, normally ~ (dilute suspensions graded beds. May be reworked by bottom from tops of flows). currents. Rest on turbidites. Turbulent flows from tops of mass flows. Thin sequence of fairly well sorted beds, may be cross-bedded. Tops may be reworked by Laminar mass flows. bottom currents. Rest on pyroclastic flows.

(Flows also develop from slump and flow of Thick, poorly sorted, poorly bedded, water-logged pyroclastic debris on subaqueous non-welded. Mixing with water can result in slopes of volcanoes to give turbidites and .) lahars. Tops may be reworked by bottom currents. May contain rip-ups. Bases erosive to non-erosive. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Subaerial Effusive Lava flows from Congealing flows, Massive flows. Broken pillow complexes and eruptions with land into water. pillows, broken hyaloclastites. Explosive disruption may subaqueous ~ pillows. Explosive produce littoral cones. deposition disruption.

Explosive Turbulent Fallout on water to Thin, well sorted, normally graded beds. suspensions in air. bottom. Sharp bases, bioturbated tops. May be in \ deep sea 100s of km from source. Pyroclastic flows Dilute suspension Thin, well sorted, normally graded. Partly from land into fallout from tops of derived from air fall and surge from land water (may be flows. which does not enter water. Rest on destroyed by turbidites. May be reworked by bottom explosive Turbulent flows currents. disruption upon from tops of mass entering water). flows. Thin sequence of fairly well sorted beds, may be cross-bedded. Rest on pyroclastic flows. Laminar mass flows. Tops may be reworked by bottom currents.

Thick, poorly sorted, poorly bedded. May be welded to base.

Subaerial Effusive Flows Congealing flows Massive forms, sometimes mostly rubble.

No pillows. r~ Explosive Ballistic; Fallout from air ~ turbulent Thin to thick, well sorted; may show normally suspension in air. graded bedding.

Pyroclastic flows. Fallout. Derived from Thin, well sorted beds. tops of flows.

Turbulent suspensions Thin sequences, fairly well sorted, commonly (.) well bedded, may be cross-bedded.

Laminar mass flows. Thick to thin, poorly sorted, massive to poorly bedded, welded to non-welded.

"--.I Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

8 R.V. Fisher

Volcanic Products

Sea Level EXPLOSIVE DEBRIS CAN BE DOMINANT OVER PILLOWED AND MASSIVE FLOWS. EXPLOSIVE HYDROCLASTIC DEBRIS DOMINANT OVER PYROCLASTIC DEBRIS. POWERFUL ERUPTION COLUMNS CAN BREAK THE WATER SURFACE AND DISTRIBUTE VOLCANICLASTICS SUBAERIALLY. LARGE DRAG FORCES ON SUBAQUEOUS ERUPTION 500 - COLUMNS CAN CAUSE COLLAPSE OF COLUMN TO PRODUCE UNDERWATER SEDIMENT GRAVITY FLOWS. ! Pc _ ! LAVA AND PYROCLASTIC FLOWS FROM LAND INTO WATER ! CAN PRODUCE PILLOW LAVAS, HYDROCLASTICS, HOT - I ? PYROCLASTIC FLOWS AND OTHER SEDIMENT GRAVITY FLOWS. .E I000

NON- EXPLOSIVE HYDROCLASTIC PRODUCTS, * PCL = PRESSURE PILLOWED AND MASSIVE FLOWS. FLOWS INCREASE RELATIVE TO HYDROCLASTICS WITH DEPTH: MAXIMUM 1500 - COM PENSATION RATIO OF FLOWS TO HYDROCLASTICS ~ 5 TDI, FLOW-TO- LEVEL CLASTIC RATIO DECREASES WITH INCREASING SLOPE OWING TO SLUMPS, SLIDES AND OTHER GRAVITY PRO- CESSES, SUCH AS ON GROWING SUBSEA VOLCANOES.

20001Mafic Silicic COMPOSITION FIG. 1. Submarine volcanic products related to water depth and composition. Pressure compensa- tion levels (PCL) are not specifically known and depend upon magmatic gas pressures, volumes and expansion rates relative to pressure exerted by the water column.

(Fisher & Schmincke 1984). It might exceed 1972), kinetic processes of falling, slumping 1000 m for silicic and volatile-rich alkalic and breaking of detached pillows on steep , but apparently is less than 500 m for slopes, rupturing of growing pillows (Rittmann most mafic basaltic magmas (McBirney 1963), 1962) and implosions of evacuated pillows well within the range of marine shelves that where water pressure exceeds internal pillow fringe continents and many islands, and of the pressures (Moore 1975). tops of some submerged islands. Geological Explosive hydroclastic processes occur when evidence suggests that the PCL for explosive (i) pore water in rocks is rapidly vaporized by alkali basaltic volcanism may be 500-1000 m an underlying magma body, or when (ii) water (Staudigel & Schmincke 1981), but it is less becomes mixed with magma beneath the sur- than 200 m below water level for most basaltic face, or under a lava at the surface. Mixing pro- eruptions (Moore & Fiske 1969; Jones 1970; cesses are complex (e.g. see Bennett 1972; Col- Fleet & McKelvey 1978; Allen 1980; Furnes et gate & Sigurgeirsson 1973; Wohletz 1980; al. 1980). Kokelaar 1983). Pillowed lava flows and In addition to pyroclastic fragmentation hyaloclastites, either mafic or silicic in composi- caused by explosive expansion of volatiles from tion, are prime evidence of subaqueous envi- within magma, there are hydroclastic processes, ronments. In areas where fossils or other evi- i.e. fragmentation caused by contact of magma dence (e.g. turbidites, carbonates, and and water (Fisher & Schmincke 1984). Hyd- associations such as shelf, slope or submarine roclastic processes can be non-explosive or fan) are absent, pillow lavas are commonly the explosive. Explosive processes can only take only evidence of a subaqueous environment. place above the PCL, but non-explosive frag- Mafic pillowed and sheet lava are the two mentation can occur at all depths. chief forms of lava flows at mid-ocean ridges, as Non-explosive hydroclastic processes include directly observed by submersibles and inferred granulation and thermal spalling due to stresses from drilling through the ocean crust (Bellaiche set up in magma undergoing rapid quenching et al. 1974; Needham & Francheteau 1974; (Rittmann 1962; Carlisle 1963; Honnorez Ballard & Moore 1977; Lonsdale 1977; Ballard Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 9 et al. 1979). Marie volcaniclastic debris is com- flow, turbidite and pelagic rain (suspension monly overlooked in deep-sea environments fallout) (Fig. 2). In volcanic as well as in non- because recovery from drill cores averages volcanic regions, initial debris can be epiclastic, about 20% and gives a bias in abundance but in volcanic regions there also can be sub- toward massive lava. Robinson et al. (1980), aqueous lava flow, and suspen- however, obtained up to 20% hyaloclastite in sion fallout deposits derived directly from sub- Cretaceous crust on the Bermuda Rise from aqueous or subaerial eruptions (Tables 1 and three drill site cores with an average recovery of 2). Moreover, pyroclastic material deposited on 70%. steep volcano slopes or shelf margins can be Subaqueous mafic pillow flows and hyalo- later remobilized as slides and transformed to clastites are most abundant but silicic pillows and submarine lahars, turbidity currents and sus- hyaloclastites have recently been described pension fallout (Fig. 3). The presence of prim- (Bevins & Roach 1979; Furnes et al. 1980; De ary and reworked pyroclastic and hydroclastic Rosen-Spence et al. 1980). According to Dim- debris indicates penecontemporaneous volcan- roth et al. (1979), in comparison to subaqueous ism (and tectonism) whereas epiclastic volcanic basic lavas, subaqueous silicic flows are thicker debris may not. and less extensive, lava pods and lobes are simi- lar but larger, quickly chilled rinds on pods and lobes are thicker, and vesicles are larger. Flow transformations It is generally thought that massive silica-rich Middleton & Hampton (1973, 1976) distin- lava flow sheets cannot form in submarine envi- guish between (i) fluid gravity flow in which ronments. However, thick, massive and exten- fluid is moved by gravity and entrains or drives sive (up to 2000 km 2) - and the sediment parallel to the bed (e.g. rivers, lava flows (Cas 1978) are interbed- ocean currents), and (ii) sediment gravity flow ded and conformable with marine Palaeozoic in which movement is by gravity and the sedi- flysch deposits in Australia. Cas (1978) sug- ment motion moves the interstitial fluid (e.g. gests that the lava flows were erupted in water grain flow, ) (Fig. 4). Lowe (1979) depths great enough to prevent escape of most and Nardin et al. (1979) have classified sedi- volatiles. The flows remained mobile because ment gravity flows based upon flow rheology volatiles were retained, and the rate of viscosity and particle support mechanism. increase from cooling was considerably less Sediment gravity flows exhibit the phenome- than the rate of emplacement. non of flow transformation (Fisher 1983). This term refers to changes within a flow from lami- nar to turbulent, or vice versa, in turn related Transport processes and deposits chiefly to particle concentration, thickness of the flow and flow velocity (slope). Transforma- There are four principal kinds of elastic (vol- tions from slides or slumps to flows (e.g. Naylor canic or non-volcanic) deposists on the sea 1980) could be considered in another class of floor. These are, from proximal to distal loca- transformations but are not discussed here. tions, slump and slide (olistostromes), mass Stanley et al. (1978) use the word 'transforma-

SOURCE: STEEP SLOPES, ABUNDANT DEBRIS (volcanic slopes, deltas, narrow shelves, active SEA LEVEL faults) P E L A G I C R A I N

SLUMP-SLIDE- "'K \ FLOW ~t I MASS FLOWS ~'~ ames) ( Liqu ified flow, groin flow, debris flow ) TURBIDITY CURRENTS ~,~:-., ( Laminar, liquified, or ~ • r-.. _ fluidized underflows with ~.':~V~...- ~ turbu lent tops }

Scale ranging from lO's of meters to IO0's of kilometers. FIG. 2. Conditions for initiation of flows and spectrum of main subaqueous transport processes in non-volcanic regions, but which also occur on subaqueous slopes of volcanoes (see Fig. 3). Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

10 R. V. Fisher

TABLE 2. Simplified list of particles, processes, environments and rocks in subaqueous settings

Origin of Processes of Sites of particles emplacementb Kinds of deposits Sites of origin b depositionc

Hydroclastic a Slides Olistostromes Volcanic Beach Pyroclastic Laminar flow Debris flows eruption Delta Epiclastic Turbulent flow Turbidites Volcano Shelf Suspension fallout Suspension fallout slopes Slope Pyroclastic flows Sedimentary Fan aprons Plain

aCan be of explosive or non-explosive origin. b Subaerial or subaqueous. cSediments can be supplied directly from an eruption or can be remobilized from steep slopes such as delta fronts, continental slopes, etc. Lava flows can occur in proximal environments.

\ \ o \ \ \ VOLCANO~ ,~ (~~(~@ ~ I"e°/eve/ sLoPE ~~C":>.~.~~ ~ ~-, •":'-":":":'~ ~%~

FIG. 3. Sources for subaqueous volcanic debris, and transport processes. This figure combines conditions given in Fig. 2 with volcanics derived directly from an active volcano. (1) Clastic wedge. Hydroclastic, pyroclastic and epiclastic debris. Dominance of particle type determined by type of volcanic activity and periods of inactivity. (2) Slide-debris flow- transitions. (3) Pyroclastic flow from land into water. Transitions to subaqueous lahars and turbidity currents. (4) Pyroclastic flows and suspension fallout from powerful shallow underwater eruptions. Column may or may not breach water surface. (5) Lava flows from land into water; (5a) massive flows; (5b) pillows and hyaloclastites. (6) Suspension fallout; (6a) from subaerial eruptions into water; (6b) from underwater eruptions or flow transformations. (7) Deep eruptions below PCL; flows and pillow lavas.

GENERAL SEDIMENT GRAVITY FLOWS TERM Low High Concentration Concentration I I I I Type of Turbidity Liquified Groin Debris flow current sediment flow flow flow

Grain Turbulence Upward Grain Matrix support Intergranular Interaction Strength mechanism Flow ~,o~O ~o~ o .c~::- ~.~..

Manner of flow Turbulent \0o , o00,I, ,0 ,n0/

FIG. 4. Classification of sediment gravity flows. Modified from Middleton & Hampton (1973, 1976). Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 11

B. A. pyroclastic flows (Sparks 1976; Fisher 1979; Sheridan 1979; Wilson 1980; Moore & Sisson Gg_)v 1981), and possibly from base surges (Wohletz & Sheridan 1979). Fluidized water-saturated mass flows may occur in subaqueous settings (Mid- C. dleton & Hampton 1973, 1976; Carter 1975), but the products of elutriation have not been described or perhaps not recognized in the sub- aqueous realm. As discussed below, in the sec- FIG. 5. Schematic representation of (a) tion on subaqueous pyroclastic flows, gas-sol- body transformations; (b) gravity trans- ids suspensions might also occur beneath water. formations; and (c) surface transforma- The phenomenon of body transformations is tions. More than one kind of transforma- suggested by the experiments and speculations tion can occur in a single flow and, follow- of Kuenen (1952) and Morgenstern (1967) ing separation of flows, transformations can occur again. who proposed that slumps or debris flows can change to turbidity currents, with no change in water content, when their velocity is great tion' but they describe 'transitions' of deposi- enough to produce internal turbulence. This tional features, from mass flow to turbidity cur- can be related to Reynolds and Bingham num- rent deposits, rather than the transformation bers. Middleton (1970) speculated that processes as is done here. changes, from laminar to turbulent and back to At least three kinds of flow transformations laminar flow can occur in large subaqueous appear to be important in subaqueous envi- grain flows. Fisher's (1971) description of the ronments (Fig. 5): Parnell Grit, , where there are large shale rip-ups in a massive 9 m thick, (i) Body transformations, where flow poorly sorted bed with an inversely graded behaviour changes from turbulent to lami- nar, or vice versa, depending upon the base, suggests that transformation from turbu- Reynolds and Bingham numbers (see lent to laminar flow occurred but does not demonstrate a preceding laminar to turbulent below) within the body of the flow, with- transformation. out significant additions or losses of water. (ii) Gravity transformations, in which flows are Gravity transformations are shown experi- initially turbulent but become gravitation- mentally to occur in high-concentration turbid- ally segregated to form a high- ity currents with the development of a 'quick concentration laminar underflow with an bed' (Middleton 1967). The quick bed is a overriding lower concentration turbulent laminar flowing grain flow and/or fluidized sediment flow (Gonzalez-Bonorino & Middle- flow that can move independently. ton 1976) which forms at the base of an initially (iii) Surface transformations, where flow sep- turbulent high-concentration turbidity current arations arise due to the drag effects of water on the top, or nose, of a flow, or due (solids concentration 1> 30% by volume) after to incorporation of fluid at a hydraulic passage of the head. Lowe (1982) describes gravity transformations and resulting deposits jump or beneath the nose of a flow. in high-concentration turbidity currents. He The concept of flow transformation focuses concludes that individual flows become gravita- upon the mass dynamics of flow which chiefly tionally segregated, with low-density flows determine sedimentary textures, structures and evolving from high-density flows as grain con- sequential relationships of massive and lami- centration increases toward the base of a flow. nated units. Various combinations of these Gravity transformations also occur subaerially three kinds of flow transformations evolving where initially turbulent pyroclastic flows from one another are explained in Fig. 5. Also, develop into low-concentration, turbulent fine grained sediment may be thrown into sus- pyroclastic surges flowing above high- pension during transformation to produce sus- concentration, laminar block-and-ash flows pension fallout. Fluidization transformations (Fisher et al. 1980; Fisher & Heiken 1982). The develop by elutriation of fine particles by phenomenon has also been invoked to explain upward moving fluids from a dense phase bed features of the blast deposits derived from the to produce a turbulent dilute phase above the 18 May 1980 eruption of Mount St Helens bed. This type of transformation is best known (Moore & Sisson 1981). The pyroclastic surge in gas-solids systems and is believed to produce may undergo continued gravity transformation ash cloud surges from the upper parts of hot to develop a secondary block-and-ash flow. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

12 R. V. Fisher

Hampton (1972) reviews various subaqueous portant topic of olistostromes is neglected. They transformations and shows how surface trans- are commonly regarded as bouldery mudstones formations can occur whereby sediment strip- that originate from submarine slides and slumps ped from the nose and surface of a subaqueous (e.g. McBride 1978), although Naylor (1981) debris flow becomes turbulently mixed with cautions that the term should be used descrip- water to form a turbidity current that continues tively for bouldery mudstones without refer- beyond the debris flow. Another kind of surface ence to origin. transformation is described by van Andel & Komar (1969) and Komar (1971). They sug- Debris flows and lahars gest that at the base of slopes, rapid flows slow down and may go from supercritical conditions Lahars are debris flows composed of vol- (Froude No. > 1) to subcritical (Froude No. canic materials. They may form deposits in < 1). Turbulence develops at this transforma- subaerial or subaqueous environments, and can tion which is known as a hydraulic jump, originate in several ways (Table 3). Lahars can thereby incorporating ambient fluid into the develop differently from non-volcanic debris flow and reducing its density. A further type is flows, but flow and depositional processes are described by Allen (1971) who concludes that essentially similar. They are sediment gravity the frontal parts of slumps may be transformed flows in which the large fragments (, into turbidity currents by the inflow and mixing gravel, boulders) are supported mainly by the of water at their bottom through 'tunnels' that yield strength of the matrix (Fig. 4). Yield occur along their fronts. The formation of strength in many debris flows is mainly a lahars from hot pyroclastic flows that flow into function of water content and the type and water (e.g. Macdonald 1972) could be classed amount of which together form the pore as a surface transformation. fluid (Hampton 1975). Debris flows are The manner by which particles are supported modified Bingham plastics (Johnson 1970) in in the final stages of flow determines in large which internal stress (Q must overcome part the textures and structures of a bed. The matrix cohesion, a frictional resistance, and manner of support is a direct function of the internal viscous forces, in order to flow particle concentration and the amount of fine (Middleton & Hampton 1973). Debris flows grained cohesive sediment mixed with pore- have a non-flowing central region (rigid plug) water, which in turn greatly affect whether or not where internal shear stress is less than the yield laminar to turbulent or turbulent to laminar strength, ry. The competence of a debris flow transformations take place. (D) is defined by Hampton (1975) as the The following sections deal mainly with sub- diameter of the largest clast that it can carry aqueous debris flows and pyroclastic flows, and (D = 8.8 ty g-l) (Pclast--Pmatfix)" Much current their flow transformations. The large and im- research on debris flow deals with identifying

TABLE 3. Origin of subaqueous lahars

I. Direct and immediate results of eruption 1. Movement of lahars directly from land into water. Subaerial lahars formed by: (a) Eruption through crater lake, snow or ice. (b) Heavy rain during eruption. (c) Flow of hot pyroclastic material into rivers or onto snow or ice. 2. Hot pyroclastic flow that moves directly from land into water and becomes mixed with the water. II. Indirectly related to eruptions 1. Triggering by or sudden liquefaction of water-soaked, pyrodastic or hydroclastic debris previously deposited in shelf environments, or on steep submerged volcano slopes. 2. Reworking of subaerially deposited pyroclastic debris which enters water from land. Subaerial origin can include: (a) Triggering of water-soaked debris by . (b) Bursting and rapid drainage of crater lakes. (c) Dewatering of large originating from collapse of volcano side. III. Not related to contemporaneous volcano activity 1. Epiclastic or hydrothermally altered material which becomes water-soaked and moves from land into water. 2. Epidastic material on steep coasts and frontal parts of deltas and shells which becomes mobilized during slumps and slides. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 13

I00 environments of deposition (especially marine), I I J as aids in constructing facies models and determining tectonic environments (e.g. Stanley et al. 1978; Leitch & Cawood 1980; -.! ~" s'J/° Nemec et al. 1980; Swarbrick & Naylor 1980; tJ3 Gloppen & Steel 1981). .10-- The plastic behaviour of debris flows is the 03 _c \ result of high concentration of particles relative to trapped water, and the abundance of fine 7" 7 compared with coarse grained particles. A ._o relatively small percentage (< 10%) of clay-size ! g m Data plotted fragments has a surprisingly large effect on ~, by Hampton (1972) inhibiting turbulence (Hampton 1972). High concentration, irrespective of particle sizes, #. inhibits turbulence, results in high bulk density and large apparent viscosity values, and 0.1 contributes to high strength properties. Such I00 I000 I0,000 I00,000 flows can support large particles despite REYNOLDS NUMBER, Re laminar flow, and come to rest ('freeze') with FIG. 6. Relationship of Bingham to steep fronts when internal shear stresses Reynolds number for development of tur- decrease below the threshold strength on low bulence in a Bingham plastic. (After His- slopes. Moreover, they can flow over soft cott & Middleton 1979.) erodible materials without significantly disturbing the underlying bed. Possibly, they can flow beneath water as far as 500 km, over laminar to turbulent flow could be irreversible, slopes as low as 0.1 ° (Embley 1976). although gravity transformations could occur Although debris flows move in laminar later. fashion (Johnson 1970), turbulence might Plugs are easily deformable masses of debris develop within the body of a flow without that form within the interiors of debris flows mixing with ambient water (a body where shear stresses fall below the strength of transformation). The onset of turbulence in the fluid, and within which there is little internal Bingham plastics is a function of the ratio of the shearing motion. The plug rides on an Reynolds number (Re = Udp/l~) to the underflow of mobile debris believed to be Bingham number (B = rfl/llU) where laminar (Johnson 1970, p. 500; Hampton U = average velocity, d = thickness of the 1972). Movement in the undertow occurs flow, p = density, /~ = dynamic viscosity and because stresses are high enough (a function of z c = strength of the plastic. Figure 6 shows velocity and weight). As velocity decreases, that turbulence occurs at Re I> 1000 B. This stresses decrease, and the lower part of the flow is equivalent to pU2/q£c ~ 1000 (i.e. progressively 'freezes' as the yield strength Re/B I> 1000), which Hiscott & Middleton is reached. Thus, the undertow progressively (1979) call the Hampton number. Hiscott & becomes part of the plug. Freezing locks in the Middleton (1979) give an excellent review of textures, structures and fabrics inherited by the the criteria for turbulence, calculations for plug from upstream before the internal yield strength of debris flows, and depositional strength of the flow is reached. processes of mass flows (also see Lowe 1976, Characteristic features of deposits from 1979, 1982). Turbulence, therefore, can subaerial and subaqueous high-concentration develop within the body of a debris flow, flows include poor or no internal bedding, despite high concentration, if it is large and matrix support of large fragments, poor sorting, velocity is high enough (see also Middleton and inverse grading in a narrow zone at their 1970). If turbulence develops, erosion and base. Inverse grading has been recorded and mixing with bottom materials and ambient discussed by many authors. For example in water may occur. This in turn can modify the lahars by Schmincke (1967), in subaqueous behaviour of the fluid. Entrainment of clays debris flows and other kinds of sedimentary would tend to reduce turbulence (Hampton gravity flows by Fisher & Mattinson (1968), 1972), but entrainment of water could lower Hiscott & Middleton (1979) and Naylor the relative solids concentration which would (1980), and in subaerial pyroclastic flows (in favour development of turbulence. With which water or abundant clay is not important enough water, a transformation from in the transport or depositional process) by Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

14 R. V. Fisher

Sparks (1976). Elongate particles in these later reworking. Subaqueous debris flows can deposits are commonly oriented roughly be expected to interfinger with lacustrine or parallel to the depositional surface or are marine sediments. imbricated. Such orientation is most strongly A subaqueous volcanic debris flow in the developed near the base. Basal contacts are Lesser Antilles (Carey & Sigurdsson 1980) is commonly non-erosive, and thick, coarse referred to as a sediment gravity flow or pyro- grained debris flows may overlie fine grained, clastic gravity flow by Sigurdsson et al. (1980), easily erodible material in sharp contact. because the structures and textures are similar to Large particles may be impressed upon fine subaerial debris flows. The deposit originated grained sediments due to loading. These from pyroclastic flows which entered the sea and features occur in subaqueous volcanic debris became mixed with water, as indicated by flows as well as in subaerial counterparts incorporated pelagic clay. The clay content (Fisher 1971), and they are used to recognize increases with distance from the source, thereby debris flows in present-day marine basins (e.g. indicating continued mixing during transport. Carey & Sigurdsson 1980; Sigurdsson et al. Thus it appears that the flow was turbulent 1980) and within the geological record (e.g. during most of its transport but was transformed Cook et al. 1972; Cas 1979; Lewis 1976; (body transformation) t ° laminar flow in the last Hiscott & Middleton 1979; Swarbrick & Naylor stages of movement. There appears to have 1980; Leitch & Cawood 1980)• Subaqueous been little difference between the flow debris flows appear to differ from subaerial processes of the sediment gravity flow debris flows in several respects, including the described by Carey & Sigurdsson (1980) and ratio of maximum particle size (MPS) to bed those of the massive division of subaqueous thickness, textures, degree of grading and pyroclastic flows described by Fiske & Matsuda imbrication, and association with other facies (1964) and Bond (1973) (see below). (Fig. 7). The bed thickness of subaqueous Large amounts of subaqueous volcaniclastic debris flows is commonly 3-10 times greater debris were deposited as debris flows within an than the maximum fragment size, whereas in Archaean volcaniclastic sequence in the Noran- subaerial deposits, bed thickness is only 2-4 da region, Quebec (Tasse et al. 1978). In proxi- times greater (Nemec et al. 1980; Gloppen & mal sections, a significant portion of debris flows Steel 1981). Many subaerial debris flows are (type A beds, Fig. 8) are massive or inversely abruptly overlain by a thinner, finer grained graded, with rare parallel laminations. Distally, and better sorted mantle of massive the sequence has a higher percentage of conglomerate or laminated and cross-bedded normally graded beds and beds with parallel granule sandstone, interpreted to have been laminations. The lateral changes can be winnowed by water flow following deposition of attributed to surface transformations in the debris flow. Commonly the subaqueous developing turbidity currents. Their facies deposits are capped by massive or rippled fine variations in terms of mean bed thicknesses and grained sandstone which is gradational with the primary grading features are diagrammatically underlying conglomerate. The sandstone is illustrated in Fig. 8. apparently emplaced at the same time as the underlying material rather than the result of Subaqueous pyroclastic flows Subaqueous pyroclastic flows can develop

/ .i.. i .i. • •-.-., . :. ,' from pyroclastic flows that move into water .i,,i..l..l, ._,_ - - _ .

• . from land (e.g. 1902 Mt Pel6e: Lacroix 1904). • o OQ-~ • 0.0 ~lJ ,,-~ Also, they are postulated to form from entirely ~ o_..7"3. underwater eruptions (Fiske 1963; Fiske et al. • .i • 1963; Fiske & Matsuda 1964; Kokelaar et al., this volume). Many workers do not believe that hot pyroclastic flows can be deposited under water, although some authors report welded from marine sequences (see Table 4). - I Moreover, Sparks et al. (1980b) present A B theoretical arguments suggesting that welding FIG. 7. Comparison of depositional fea- can occur beneath water. Wright & Mutti tures of (a) subaerial debris flows, and (b) (1981) discuss many of the problems encoun- subaqueous debris flows. (After Nemec et tered in interpreting subaqueous pyroclastic al. 1980.) flows. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 15

/ / DIRECTIONOF TRANSPORT .t/" ,( ___. ----" 2~.~.-.-~::.~/:;:n"~i~k~~~ I / /. \

4- + 4- +

4-

A

/ / 55cm IOOcm ..I'/ ! / ; o I km I

~43 ~ 36 50% + + + ~ .ASSJVF ~43 ~ 62 ~ NOR/~AAL + + REVERSE !~ RIPPLES OBLIQUE + PARALLEL B FIG. 8. Lateral variations of primary structures and mean bed thickness in Archaean subaqueous pyroclastic flow deposits, Canada. (a) Mass flow deposits. (b) Turbidite deposits as interpreted by Tasse et al. (1978). (From Lajoie 1979.)

Non-welded deposits pyroclastic flows (Fiske & Matsuda 1964; Fer- Evidence for subaqueous deposition in nandez 1969; Yamada 1973; Niem 1977; Fig. ancient deposits is determined indirectly on 9), as are glassy fragments produced by sudden stratigraphical grounds, using sedimentary quenching. Fine perlitic cracks are characteris- sequence associations (facies models) or inter- tic of essential, non-vesicular, vitric clasts (Kato bedding with pillow lavas or fossiliferous sedi- et al. 1971). ments (Fiske & Matsuda 1964; Howells et al. Non-welded subaqueous pyroclastic flow 1973; Yamada 1973; Niem 1977). Subaqueous deposits characteristically are massive to poorly pyroclastic flow deposits consist of lithic frag- bedded and poorly sorted in their lower part ments, crystals, glass shards and in vari- (lower division), and are thinly bedded in their able proportions and sizes. Shale rip-ups and upper part (upper division) (Fig. 10). The lower blocks up to several metres in diameter derived division can form 50% or more of the total from underlying beds may be incorporated near sequence (Fiske & Matsuda 1964; Bond 1973; the base of a sequence (Yamada 1973; Niem Niem 1977). The two-division sequence is 1977). interpreted in terms of a waning, initially vol- Shattered crystals are common in subaqueous uminous underwater eruption (Fiske & Mat- Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

TABLE 4. Subaqueous pyroclastic flow occurrences (from Fisher & Schmincke 1984)

Environment Age Location Welding and temperature Remarks Reference

1883 Krakatau, Java Non-welded. Generated destructive . Self & Rampino Ox 1981 28,000 yr BP Grenada Basin, Non-welded. Resemble debris Derived from subaerial hot Carey & Sigurdsson Lesser Antilles flow deposits. pyroclastic flows on Dominica. 1980; Sigurdsson et al. 1980 Late Subaqueous west Deposits not cored. Deposits traceable to subaerial Sparks et al. 1980a flank, Dominica, flows (block and ash; welded Lesser Antilles ). Middle Miocene Santa Cruz Non-welded. Pumice-rich. Within a marine sequence. Fisher & Charleton Island, Calif., Massive beds, some with 1976 U.S.A. pinkish oxidized tops. Miocene Tokiwa Non-welded. Pumice-rich. Interbedded conformably with Fiske & Matsuda Formation, Japan fossil-rich beds. 1964 Oligocene Is. of Rhodes, Non-welded. Deep water, marine. Wright & Mutti Greece 1981 Oligocene-Eocene Mt Rainier Natl Non-welded. Massive lower to Marine shelf deposition inferred. Fiske 1963; Fiske Park, Wash., bedded upper divisions. et al. 1963 U.S.A. Early Oligocene- Philippines Non-welded. Pyroclastic Interbedded with marine limestone. Garrison et al. 1979 Marine Late Eocene turbidites. Paleogene to Philippines Non-welded to welded (?) tuff. Within a marine sequence. Fernandez 1969 Cretaceous Lower Mesozoic Sierra , Metamorphosed (shards not Ponded in shallow marine . Busby-Spera , U.S.A. preserved). 1981a,b Permian Alaska, U.S.A. Non-welded. Massive lower to Structures and sequence similar to Bond 1973 bedded upper divisions. Tokiwa Formation, Japan. Mississippian Oklahoma and Non-welded. Turbidite Interbedded within marine flysch Niem 1977 Arkansas, U.S.A. structures. sequence. Ordovician Ireland Welded. Associated with shallow-water Stanton 1960; deltaic sediments. Dewey 1963 Ordovician , United Welded. In marine sequence. Traceable to Francis & Howells Kingdom subaerial pyroclastic flows. 1973; Howells et al. 1979; Howells & Leveridge 1980 Lower Ordovician S. Wales, U.K. Welded. Within marine sequence. Lowman & Bloxam 1981 Archaean Rouyn-Noranda, Non-welded. Structures and sequence similar to Dimroth & Canada Tokiwa Formation, Japan. Demarcke 1978; Tasse et al. 1978

Plio-Pleistocene Japan Non-welded. Resemble Interbedded with lacustrine rocks. Yamada 1973 turbidites. Flow from land into water. Lacustrine Mio-Pliocene Japan Non-welded. Thermoremanent Massive units resembling Kato et al. 1971; magnetism suggests debris flows. Yamazaki et al. emplacement at 500°C 1973 Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 17

...... J

i:.]/.+:: /,: i??.' . . . ,~ .o.o+ ..o

o.+ o. -° . .o. . - ~- • ° d g "-.','." ; ==

o +i .,p . .... • , . .. " o"O. :o

o o ,, • • • < e'C) o o" ~ o e o . • . . ..

• " .- "- 0 o ".o

0 -" • .°oO o .- . + + , A B FIG. 10. Examples of non-welded suba- queous pyroclastic flow deposits. (a) Example interpreted to have developed from waning eruption by Fiske & Matsuda (1964); massive division (1) overlain by doubly graded sequence (2). (b) Obispo Tuff, interpreted to have developed from passage of one flow or several pulses of flow (Fisher 1977). Massive division (1), overlain by (2) sequence of planar to wavy beds or beds with low-angle cross-beds, fine to medium grained tuff to tuff layers, and (3) capped by a thin, very fine grained bioturbated tuff of probable sus- pension fallout origin. Interval (2) may FIG. 9. Shattered crystals from a reflect several pulses of flow, possibly subaqueous pyroclastic flow deposit, representing retrogressive slumping Eugenia Formation, Baja, California (J. (Morgenstern 1967). Pumice lapilli com- Hickey, pers. comm. 1982). monly decrease in size from bottom to top indicating that the tuff was water logged prior to flow, thereby supporting the idea that flow was remobilized pyroclastic suda 1964). Yamada (1973, this volume), how- debris from an unstable subaqueous slope. ever, recognizes five divisions similar to a Bouma sequence (Bouma 1962). A 2-fold sequence capped by a thin, very fine grained, Dense, lithic fragments may be inversely graded bioturbated pelagic layer occurs in the Obispo near the base, as in debris flows (Yamada Tuff, California (Fig. 10b), where basal con- 1973). tacts are generally sharp on undisturbed strata, The upper division of non-welded subaque- although flute marks, grooves and load casts ous pyroclastic flow deposits is commonly com- occur locally. Where bottom contacts are plane posed of many thin, fine to coarse grained ash and sharp, missing underlying strata may indi- beds. In some sequences, each bed may be cate erosion (Yamada 1973). normally graded, and the entire sequence of The lower division of a subaqueous pyroclas- beds becomes finer grained upward. This is cal- tic flow unit is a single, coarse grained bed led a doubly graded sequence (Fiske & Mat- commonly lacking internal structures or lamina- suda 1964). Not all upper divisions are doubly tions (Fiske & Matsuda 1964; Bond 1973), graded (Niem 1977). although some beds have incipient cross- Double grading is interpreted by Fiske & laminations outlined by a crude orientation of Matsuda (1964) to signify contemporaneous coarse fragments (Kato et al. 1971; Yamazaki et waning subaqueous volcanism with deposition al. 1973). Larger and more dense fragments from thin turbidity flows following deposition generally occur near the base of the massive of the massive bed. Deposits without double divisions, with crystals, glass shards and pumice grading, such as those described by Niem becoming more abundant toward the top. (1977) (see also Fig. 10b), are taken to signify Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

18 R. V. Fisher sloughing from oversteepened pyroclastic enough heat to weld. As shown by Sparks et al. slopes on the edge of a volcano where trans- (1980b), welding is theoretically possible. This formations from slides to sediment gravity flows would be likely to occur where changes in slope could occur. are insufficient to cause laminar to turbulent body transformations. Such transformations are Welded deposits more likely to occur where sudden steepening Most controversial are welded pyroclastic of slope could cause turbulence and mixing, flow deposits interpreted to be subaqueous on thereby producing a cool sediment gravity flow stratigraphic grounds. The best known example (Sigurdsson et al. 1980). Also as shown for case is from the Ordovician of Snowdonia in Wales. A (Fig. 11), turbulent ash cloud surges, which There, a welded tuff within the Capel Curig can develop above dense flows on land, can Formation is interpreted to be the submarine flow over the water (Anderson & Flett 1903). correlative of subaerial welded tuff (Francis & Additionally, surface transformations on the Howells 1973; Howells et aI. 1973; Howells et underwater flow could produce turbidity cur- al. 1979; Kokelaar et al., this volume a,b). rents. In each instance, thin, graded beds with Francis & Howells (1973) suggest that subaeri- small total volume can develop by fallout at the ally produced pyroclastic flows entered the sea top of flow units. and remained hot enough to weld after suba- For case B (Fig. 11), Fiske & Matsuda queous emplacement. The subaerial welded tuff (1964) envisage an emerging consists of an uninterrupted sequence of two or with great force from an underwater vent. more flow units of greater combined thickness Large amounts of material are ejected into (70m) than correlative subaqueous units water, fall back to the sea floor, and form a (< 50 m). The base of the subaerial tuff is pla- nar, contains and overlies a conglomer- ate with a reddened top, and welding is restricted to the middle of the sheet. The sub- aqueous pyroclastic flow deposits are welded to their base, and are interbedded with marine siltstones containing brachiopods. Basal con- ." .i~) A tacts of the subaqueous welded tuffs tend to be irregular with zones resembling large-scale load and flame structures. These flow deposits are massive in the lower and central portions and grade imperceptibly upward through ill- defined, faintly bedded zones to evenly bedded tuff which locally has broad, low-angle cross- bedding suggestive of marine deposition. Their tops are in sharp contact with overlying sand- stone with similar broad cross-bedded struc- tures. Kokelaar (1982) accounts for the large load-structure protrusions at the base of the subaqueous welded tuffs by fluidization of sub- FIG. 11. Means by which subaqueous jacent wet sediments following emplacement of pyroclastic flows can develop. (a) Hot a hot pyroclastic flow. pyroclastic flow from land into water (Francis & Howells 1973; Yamada 1973; Sigurdsson et al. 1980). Density separation Origin of subaqueous pyroclastic flows of flow and ash cloud surge at air-water Three principle ways in which subaqueous interface; surface transformation to form pyroclastic flows can be generated are given in thin turbidite sequence. Flow may weld if Fig. 11. Massive lower divisions with finer not mixed with water, or be transformed to grained bedded tops can develop in each situa- debris flow if mixed. (b) Pyroclastic flow tion. Mixing with water by flow transformation develops from column collapse (Fiske to form sediment gravity flows and turbidity 1963; Fiske & Matsuda 1964); waning currents can also occur in each case. The eruption may produce overlying laminated doubly graded sequence. Suspended mater- occurrence of subaqueous welded tuff in Wales ial can form thin fallout layers and floating correlative with subaerial welded tuff (Francis pumice can form pumice rafts. (c) Sedi- & Howells 1973) suggests that hot pyroclastic ment gravity pyroclastic flows develop flows that enter water from land (case A) may from slumping of unstable slopes com- do so without mixing with water and retain posed of pyroclastic debris. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 19 dense water-rich debris flow deposited as a with lacustrine sediments of Onikobe , thick, massive layer. In the closing stages of Japan (Yamada 1973), closely resemble turbi- eruption, small turbidity currents entrain dite sequences. The flows may have originated pumice lapilli and crystals of plagioclase and from subaerial eruptions at the margin of the quartz which are deposited as thinly graded caldera, but there is little evidence to indicate beds above the massive beds. whether hot pyroclastic flows entered the water Ash from a cloud that from land, or if slumping of unstable slopes on breaks the water surface, or that has spread the margin of the lake initiated the flows. widely away from the vent within the water Sparks et al. (1980b) have proposed on body, continues to settle after the eruption theoretical grounds that if a hot pyroclastic flow ends, but in decreasing amounts and becoming manages to pass through the air-water inter- finer grained. Thus, there is an insufficient sup- face, subaqueous welding is not only possible, ply to maintain the continuous flow that depo- but is enhanced. Hydrostatic pressure increases sited the thick structureless part of the pyroclas- by 1 bar for every 10 m water depth, and if tic flow. Instead, intermittent thin turbidity cur- water is 'drawn' into the hot pyroclastic flow it rents follow one another in close succession. As can be absorbed into glass shards. At increased the particle-rich column continues to settle, pressures, water content of glass can be higher, successively smaller amounts of finer grained resulting in lower viscosities which favour weld- and less-dense ash fall out, and the density cur- ing at lower temperatures (Smith 1960; Riehle rents become less frequent and finally end. 1973). Calculations of the viscosity of glass Thus, the total volume of the upper division is shards as a function of water depth suggest that considerably greater than that developed at the the viscosity of glass decreases substantially top of a single flow entering water from land with depth despite cooling of the flow (Sparks et (case A), although distally Bouma sequences al. 1980b). For subaerial pyroclastic flows at may dominate in case A (Yamada 1973, this 800°C, the viscosity of rhyolitic glass is about volume). 2 x 1011 poise, but, at 755°C under 500 m of A vertical eruption column formed entirely water, this is reduced by dissolution of water to beneath water, such as that proposed by Fiske an estimated 2 x 109 poise. Thus glass under & Matsuda (1964), may not develop into hot the latter conditions would compact much fas- pyroclastic flows upon collapse because turbul- ter than in a subaerial flow under the same ent mixing with water along its margin would be uniaxial stress. A major difficulty with the enhanced by drag effects in water which would theory, however, is the slow diffusion rate of cool the materials. Possibly, however, a vol- water into glass. In addition, there is a question uminous 'boiling-over' type of eruption without of how water could be drawn into a flow and a vertical column (Fisher & Schmincke 1984) into contact with glass participles. Turbulent could be extruded at rates great enough to mixing is the dominant process by which water produce a flow protected from the water by a can gain access to a flow, but this would also carapace of steam and retain enough heat to tend to cool or explosively destroy the flow become welded. As yet, however, no subaque- (Walker 1979). A rapid increase of pressure, ous vents for welded tufts have been positively from loading by the water column, would identfied, although one has been postulated by inhibit volatile escape and is a more likely cause Kokelaar et al. (this volume a, b) to explain a of welding. Water surrounding the hot flows welded tuff sequence on Ramsey Island in SW would create a thin vapour carapace around the Wales. flow during cooling (Francis & Howells 1973; Slumping of unstable material (case C) com- Yamazaki et al. 1973), provided that turbulent posed exclusively of pyroclastic material is mixing does not occur at the surface of the flow. probably common on submerged slopes of vol- canoes, but has rarely been documented. It is to be expected, however, that flow transforma- Submarine ash layers tions from subaqueous lahars to turbidity cur- The initial depositional process of marine rents can develop sequences similar to those fallout ash is by the settling of particles produced directly from volcanic eruptions. A introduced into water from the air or from thick (c. 15 m) coarse grained pyroclastic flow underwater flows or eruptions. The process within the Obispo Tuff, California is inter- therefore depends upon the size, shape and preted to have formed by cold water mobiliza- density of fragments, and the vertical and tion of pyroclastic material off steep slopes lateral distribution is influenced by outside (Fisher 1977). factors such as wind and water current Subaqueous pyroclastic flows interbedded velocities and directions. For more complete Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

20 R. V. Fisher reviews of this subject, the reader is referred to Island volcanoes Kennett (1981) and Fisher & Schmincke (1984). Data from widespread submarine fallout Irrespective of magma composition and margi- layers are useful in many ways. For example, (i) nal or intra-plate setting, volcanoes in the sea in correlating over long distances between can be considered to evolve through two over- sequences deposited in different environments lapping stages (Fig. 12). Stage A develops (Ninkovich & Heezen 1965, 1967; Huang et al. beneath the PCL and stage B above the PCL 1973; Watkins et al. 1978; Keller et al. 1978; (L. Ayres, personal communication). As sugges- Hahn et al. 1979; Thunell et al. 1979; ted by previous discussions, volcanic island Ninkovich 1979), (ii) in solution of problems of foundations (below the PCL) can be composed magma genesis and evolution (Scheidegger et of massive and pillowed lava flows with lesser al. 1978, 1980), (iii) in assessments of volcanic amounts of hyaloclastites (up to 20%). As vol- cycles (Hein et al. 1978), (iv) in derminations of canoes grow and slopes steepen in stage A, production rates of tephra and eruption column talus development and slides cause mechanical heights (Shaw et al. 1974; Huang et al. 1973, fragmentation and movement of debris into 1975, 1979; Watkins & Huang 1977; deeper water, resulting in a decrease in the Ninkovich et al. 1978), (v) in estimations of the lava-to-elastic ratio outward from the source. duration of eruptions (Ledbetter & Sparks This is observed on sea-mounts (Stanley & 1979), (vi) as aids in determining rates of plate Taylor 1977; Lonsdale & Spiess 1979). Above movements and arc polarity (Ninkovich & the PCL, however, there is a wider range of Donn 1976; Sigurdsson et al. 1980), (vii) in transport processes, kinds of fragments and con- determination of prevailing wind and water sequent lateral facies changes in rock types. current directions (Eaton 1964; Nayudu 1964; Ayres (pers. comm.) infers a shield shape for Ninkovich & Shackleton 1975; Hampton et al. the foundations of island volcanoes in stage A, 1979), (viii) in making inferences about with steeper slopes developing in stage B. In Fig. climatic changes (Kennett & Thunell 1975, 12, lava-to-elastic ratio is estimated at up to 1977) and (ix) in solving problems concerning 5:1 in early growth stages and in flatter summit diagenesis (Coombs 1954; Fisher & Schmincke regions of cones, decreasing perhaps to 2:1 or 1984). less away from the source as slopes steepen, to Widespread marine fallout layers can occur entirely elastic on surrounding abyssal plains. In in back-arc and fore-arc regions, depending addition to lava flows, the intrusion of sills may mainly on wind directions, and occur in all be important in the growth of an island volcano. marine sedimentary environments. In the On La Palma, Canary Islands, for example, Lesser Antilles, marine fallout layers are the Staudigal & Schmincke (1981) report 2000 m dominant volcaniclastic type in front of the arc of sills which contributed significantly to the (Sigurdsson et al. 1980), but in the Cordilleran elevation of the volcano. region of western North America they are most Rocks of stage B 1, above the PCL but below abundant behind the Cascades , water level, can be dominantly explosive hyd- presently as well as in ancient deposits (Eaton roclastic debris. At DSDP Site 253 (Ninetyeast 1964; Dickinson 1974). Ridge, Indian Ocean) for example, there is at least 388 m of bedded Middle Eocene basaltic

Terroce~

~~~. ','.~ ! .~, .

~-~_~~ ~Low s .,.....-..

FIG. 12. Generalized model of an island volcano. Stage A, foundation of island volcano beneath pressure compensation level (PCL); lava (L) > hydroclastics (H); may contain a significant volume of sills. Stage B1, beneath sealevel and above PCL, L < H. Stage B 2, subaerial part; pyroclastics (P) > H > L if andesitic, but L > P if basaltic shield. Clastic apron enlarges rapidly in stage B 2 and sediment gravity flows and fallout deposits increase in flanks and surrounding basin. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 21 hydroclastic debris, believed to have been areas of long-continued volcanism and erosion, erupted in about 150 m of water (Fleet & fringing shelves become the sites for develop- McKelvey 1978). ment of deltas and fan-deltas with deep-sea fans In stage B2, lava flows of basaltic islands, and abyssal turbidite plains farther away (Londs- such as , cap the hydroclastic material as dale 1975). the vents become subaerial (Moore & Fiske Facies of volcanic rocks reflect distance to 1969). Along the fringes of such islands, explo- source. Dykes, sills, plutons, lava flows, coarse sion debris, pillows and pillow breccias can be grained pyroclastic rocks and their reworked generated where lavas flow into the sea (Moore equivalents, occur at or near the source. Farther 1975; Peterson 1976; Fisher 1968). Erosion away, lava flows and intrusive rocks disappear commences to produce epiclastic volaniclastic and volcaniclastic sediments become finer debris which increases in volume as the volcano grained (Dickinson 1968; Mitchell 1970). Fall- grows in size. However, at volcanoes where out tufts occur within both of these near- and pyroclastic material is abundant relative to lava intermediate-distance regions, but may be dis- flows, such as andesitic stratocones, emerged persed globally and occur at great distances slopes in stage B 2 are steeper than shield vol- from source volcanoes. These relationships canoes, and pyroclastic debris is delivered to apply to both subaerial and submarine envi- the sea as lahars, pyroclastic flows and fallout ronments. ash directly from eruptions, or by remobiliza- It is also true that (i) volcanic rocks are em- tion of loose pyroclastic deposits. The volumes placed in the order in which they are extruded, of pyroclastic material stored in clastic wedges and (ii) the eroded products of an inactive vol- on submerged slopes of andesite volcanoes is cano tend to be emplaced in reverse order of consequently greater than for shield volcanoes, extrusion. Erosion can extend down to the and so is the volume of debris deliverered to source plutons (Dickinson & Rich 1972; Inger- adjacent basins as sediment gravity flows and soll 1978; Mansfield 1979). This simplified suspension fallout. The rate and volume of scheme (Fig. 13) is only a basic framework, debris supplied during periods of volcanic activ- however, because erosion of a volcanic edifice ity and inactivity is discussed by Kuenzi et al. can alternate with periods of active growth, (1979) and Vessel & Davies (1981). The rate adjacent volcanoes can supply material to and volume of supply from fringing terraces to basins at different times, distant volcanoes can adjacent slopes and basins can also be modified supply ash during any part of the constructive by eustatic sealevel changes. During high stands or destructive cycles, constant settling of pelagic of sealevel, more sediment can be stored in the and hemipelagic material from marine waters clastic wedges and less reaches the basins than can occur, and sediments from land masses and during low stands, when sedimentation rate and islands can be supplied from opposite sides of a volume of sediment increases in the basins basin. Further complications arise, for example, (Klein et al. 1979). Primary and reworked vol- where moving plates carry source areas away canic products become more diverse as an from the depositional site (diverging plates), island emerges, because erosion is contem- toward the depositional site (converging poraneous with eruptions, and sediment aprons plates), or laterally along transform boundaries. with unstable slopes can grow more rapidly Irrespective of lateral facies changes and than before emergence. Reefs can form, and in despite the many complexities, it is possible to

VOLCANO BASIN DEPOSITS Primary end Volconiclestic reworked volcanics products of Sediments during active erosion during from other volcanism inactive volcanism sources

FIG. 13. Simplified stratigraphic scheme for deposition of volcaniclasitc debris from active and inactive volcanoes in adjacent basins. (1) plutonic sources and epiclastic products, (2) and (3) volcanic products in sequence of emplacement from active volcanoes. Erosion of the volcanic pile produces epiclastic debris in inverse order. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

22 R. V. Fisher

Lovo flows

seo level A

Mossive Iovos pillows, ond hyoloclostites

I~------,GNEOUSI+~+'T-HCORE COMPLEX ~ VOLCANIC SUBAERIAL PYRO- [~ SUBAQUEOUSSEDIMENT CLASTIC APRON I"~-----] GRAVITY FLOWS VOLCANIC ~ FALLOUT ASH LAYER CONGLOMERATES I' I PELAGIC AND HEMIPELAGIC SEDIMENT FIG. 14. Facies model of island volcanoes. (a) Newly emergent intra-plate basaltic (from Moore & Fiske 1969). (b) Island-arc andesitic stratocone (from Sigurdsson et al. 1980). gain considerable information, about many of the 98% of the volcaniclastic sediments in the aspects discussed above, from vertical sequ- back-arc basin, and ash-fall layers form 60% of ences (in outcrop or drill-hole samples), if the the volcaniclastic sediments in the fore-arc origin of clasts can be determined (e.g. pyro- region. This asymmetry is attributed to wind elastic, hydroclastic, epiclastic) and depositional directions, arc slopes (steeper toward the processes can be interpreted from rock textures back-arc) and ocean current directions. and structures (see Schmincke & von Rad Research in the Lesser Antilles documents 1979). sedimentary asymmetry associated with arc Work in the Lesser Antilles has led to a facies polarity. Sedimentary asymmetry, however, model of a calc-alkaline island-arc volcano (Fig. depends upon arc geometry (facing direction, 14) (Carey & Sigurdsson 1978; Sigurdsson et al. trend direction of the arc) relative to wind and 1980; Sparks et al. 1980a; see also Carey & ocean current directions. Therefore, volcani- Sigurdsson, this volume). In the Lesser Antilles, elastic facies as related to arc polarity will differ dominant kinds of sediment transport processes in different arcs. and deposit types are highly asymmetric. Vol- caniclastic sediments form 34% of the total ACKNOWLEDGMENTS: I thank James Hickey cored sediment behind the arc and 4% of the (UCSB), Douglas Coombs (Otago), Hans-Ulrich total in front of the arc, a proportion which does Schmincke (Bochum) and Cathy Busby-Spera not significantly vary within 300 km of the arc. (Princeton) for critical, penetrating and extremely Sediment gravity flows of various kinds form helpful reviews of this manuscript.

References

ALLEN, C. C. 1980. Icelandic subglacial volcanism: --, HOLCOMBE, R. T. & VAN ANDEL, T. H. 1979. thermal and physical studies J. Geol. 88, 108-17. The Galapagos Rift at 86°W. III. Sheet flows, ALLEN, J. R. L. 1971. Mixing at turbidity current collapse pits, and lava lakes of the rift valley, jr. heads, and its geological implications. J. sedi- geophys. Res. 84, 5407-22. ment. Petrol. 41, 97-113. BELLAICHE, G., CHEMINEE, J. L., FRANCHETEAU, J., ANDERSON, T. & FLEIT, J. S. 1903. Report on the HEKINIAN, R., LEPICHON, X., NEEDHAM, H. D. eruption of the Soufriere in St Vincent in 1902 BALLARD, R. D. 1974. Rift Valley's inner and on a visit to Montagne Pelee in , floor: First submersible study. Nature, Lond. Part I. Philos. Trans. R. Soc. London. A200, 250, 558-60. 353-553. BENNETr, F. D. 1972. Shallow submarine volcanism. BALLARD, R. D. & MOORE, J. G. 1977. Photographic J. geophys. Res. 77, 5755-60. Atlas of the Mid-Atlantic Ridge. Springer, New BEVINS, R. E. & ROACH, R. A. 1979. and York. isolated pillow of rhyodacitic composi- Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 23

tion from the Fishguard Volcanic Group, Lower in northwest Viti Levu, Fiji. Am. J. Sci. 266, Ordovician, SW Wales, United Kingdom. J. 440-53.

Geol. 87, 193-201. -- 1974. Sedimentation within and beside ancient BOND, G. C. 1973. A late Paleozoic volcanic arc in and modem magmatic arcs. In: DoTr JR, R. H. the eastern Alaska Range, Alaska. J. Geol. 81, t~ SHAVER, R. H. (eds) Modern and Ancient 557-75. Geosynclinal Sedimentation. Spec. Pub. Soc. BOUMA, A. H. 1962. Sedimentology of Some Flysch Econ. Paleont. Miner. 19, 230-9.

Deposits; A Graphic Approach to Facies -- & RICH, E. L. 1972. Petrographic intervals and Interpretation. Elsevier, Amsterdam, Nether- petrofacies in the Great Valley Sequence, Sac- lands. ramento Valley, California. Bull. Geol. Soc. Am. BUSBY-SPERA, C. 1981a. Silicic ash-flow tufts inter- 83, 3007-24. bedded with submarine andesitic and sedimentary DIMROTH, E. • DEMARCKE, J. 1978. Petrography rocks in lower Mesozoic roof pendants, Sierra and mechanism of eruption of the Archean Nevada, California (Abst.). Bull. geol. Soc. Dalembert tuff, Rouyn-Noranda, Quebec, Am. 13, 47. Canada. Can. J. Earth Sci. 15, 1712-23. ,1981b. Early Mesozoic submarine epicontinen- --, COUSINEAU, P., LEDUC, M., SANSCHAGRIN, Y. tal calderas in the southern Sierra Nevada, & PROVOST, G. 1979. Flow mechanisms of California. Los, 62, 1061. Archean subaqueous and flows. CAREY, S. N. t~ SIGURDSSON, H. 1978. Deep-sea evi- In: Current Res. Part A, Geol Surv. Canada, dence for distribution of tephra from the mixed Paper 79-1A, 207-11. magma eruption of the Soufriere on St. Vincent, EATON, G. P. 1964. Windborne : A poss- 1902: ash turbidites and air fall. Geology, 6, ible index to polar wandering, J. Geol. 72, 271-4. 1-35. & SIGURDSSON, H. 1980. The Roseau Ash: EMBLEY, R. W. 1976. New evidence for occurrence deep-sea tephra deposits from a major eruption of debris flow deposits in the deep sea. Geology, on Dominica, Lesser Antilles Arc. J. Volcan. 4, 371-4. geoth. Res. Amsterdam, 7, 67-86. FERNANDEZ, H. E. 1969. Notes on the submarine ash CARLISLE, D. 1983. Pillow breccias and their flow tuff in Siargao Island, Surigao del Norte aquagene tufts, Quadra Island, British Colum- (Philippines). Philipp. Geol. 23, 29-36. bia. J. Geol. 71, 48-71. FISHER, R. V. 1968. Puu Hou littoral cones, Hawaii. CARTER, R. M. 1975. A discussion and classification Geol. Rdsch. 57, 837-64. of subaqueous mass-transport with particular 1971. Features of coarse-grained, high- application to grain-flow, slurry-flow, and fluxo- concentration fluids and their deposits. J. sedi- turbidites. Earth Sci. Rev. ll, 145-77. ment. Petrol. 41, 916-27. CAS, R. 1978. Silicic lavas in Paleozoic flyschlike 1977. Geologic guide to subaqueous volcanic deposits in New South Wales, Australia: rocks in the Nipomo, Pismo Beach and Avila Behaviour of deep subaqueous silicic flows. Bull. Beach areas. Penrose Conf. geol. Soc. Am. The geol. Soc. Am. 89, 1708-14. Geology of Subaqueous Volcanic Rocks. 1-29. 1979. Mass-flow arenites from a Paleozoic -- 1979. Models for pyroclastic surges and pyro- interarc basin, New South Wales, Australia: clastic flows. J. Volcan. geoth. Res. Amsterdam, mode and environment of emplacement. J. sedi- 6, 305-18. ment. Petrol. 49, 29-44. 1983. Flow transformations in sediment gravity COLGATE, S. A. & SIGURGEIRSSON,T. 1973. Dynamic flows. Geology, 11, 273--4. mixing of water and lava. Nature, Lond. 244, & CHARLETON, D. W. 1976. Mid-Miocene 552-5. Blanca Formation, Santa Cruz Island, California. COOK, H. E., MCDANIEL, P. N., MOUNTJOY, E. W. & In: HOWELL, D. G. (ed.)Aspects of the Geologic PRAY, L. C. 1972. Allochthonous carbonate History of the California Continental Borderland. debris flows at bank ('reef') margins, Misc. Pub. Pacific Section Am. Assoc. Petrol. Alberta, Canada. Bull. Can. Pet. GeoL 20, Geol. 24, 228-40.

439-97. -- & HEIKEN, G. H. 1982. Mt Pel6e, Martinique: COOMBS, D. S. 1954. The nature and alteration of May 8 and 20 pyroclastic flows and surges. J. some Triassic sediments from Southland, New Volcan. geoth. Res. Amsterdam 13, 339-71.

Zealand. Trans. R. Soc. N.Z. 82, 65-109. -- & MATI~INSON,J. M. 1968. Wheeler Gorge tur- DE ROSEN-SPENCE, A. F., PROVOST, G., DIMROTH, E., bidite--conglomerate series; inverse grading. J. GOCHNAUER, K. & OWEN, V. 1980. Archean sediment. Petrol. 38, 1013-23. subaqueous flows, Rouyn-Noranda, & SCHMINCKE, H.-U. 1984. Pyroclastic Rocks. Quebec, Canada, and their Quaternary equival- Springer, Heidelberg. ents. Precambrian Res. 12, 43-77. ~, SMITH, A. L. & ROOBOL, M. J. 1980. Destruc- DEWEY, J. F. 1963. The Lower Paleozoic stratigraphy tion of St Pierre, Martinique by ash cloud surges, of central Murrisk County, Mayo, Ireland, and May 8 and 20, 1902. Geology, 8, 472-6. the evolution of the South Mayo Trough. J. geol. FISKE, R. S. 1963. Subaqueous pyrodastic flows in Soc. London, 119, 313-43. the Ohanapecosh Formation, Washington. Bull. DICKINSON, W. R. 1968. Sedimentation of volcani- geol. Soc. Am. 74, 391-406. clastic strata of the Pliocene Koroimavua Group & MATSUDA, T. 1964. Submarine equivalents of Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

24 R. V. Fisher

ash flows in the Tokiwa Formation, Japan. Am. --, ADDISON, R., EVANS, C. D. R. & NUTT, J. Sci. 262, 76-106. ~vl. J. C. 1979. The Capel Curig Volcanic Forma- --, HOPSON, C. A. (~ WATERS, A. C. 1963. Geol- tion, Snowdonia, North Wales; variations in ogy of National Park, Washing- ash-flow tufts related to emplacement environ- ton. Prof. Pap. U.S. geol. Surv. 444, 1-93. ment. In: HARRIS, A. L., HOLLAND, C. H. ,~ FLEET, A. J. & MCKELVEY, B. C. 1978. Eocene LEAKE, B. L (eds) The Caledonides of the British explosive submarine volcanism, Ninetyeast Isles-Reviewed. Spec. Publ. geol. Soc. London, Ridge, Indian Ocean. Mar. Geol. 26, 73-97. 8, 611-8. Blackwell Scientific Publications, FRANCIS, E. H. ,~¢ HOWELLS, M. F. 1973. Transgres- Oxford. sive welded ash-flow tufts among the Ordovician HUANG, T. C., WATKINS, N. D. & SHAW, D. M. 1975. sediments of NE Snowdonia, North Wales. J. Atmospherically transported in geol. Soc. London, 129, 621-41. deep-sea sediments: volcanism in sub-antarctic FURNES, H., FRIDLEIFSSON, I. B. & ATKINS, F. B. latitudes of the south Pacific during late Pleis- 1980. Subglacial volcanics--on the formation of tocene time. Bull. geol. Soc. Am. 86, 1305-15. acid hyaloclastites. J. Volcan. geoth. Res. Ams- ~, CAREY, S., SIGURDSSON, H. & DAVIS, A. 1979. terdam, 8, 95-110. Correlations and contrasts of deep-sea ash GARRISON, R. E., ESPIRITU, E., HORAN, L. J. & deposits from the Lesser Antilles in the Western MACK, L. E. 1979. , sedimentology, Equatorial Atlantic and Eastern at and diagenesis of hemipelagic limestone and tuf- latitude 14°N. Los, 59, 1119. faceous turbidites in the Aksitero Formation, WATKINS, N. D., SHAW, D. M. & KENNETT, J. P. Central Luzon, Philippines. Prof. Pap. U.S. geol. 1973. Atmospherically transported volcanic dust Surv. 1112. in South Pacific deep sea sedimentary cores at GLOPPEN, T. G. & STEEL, R. J. 1981. The deposits, distances over 3000km from the eruptive internal structure and geometry in six alluvial source. Earth planet. Sci. Lett. 20, 119-24. fan-fan delta bodies (Devonian-Norway)--A INGERSOLL, R. V. 1978. Petrofacies and petrologic study in the significance of bedding sequence in evolution of the Late Cretaceous forearc basin, conglomerates. Spec. Publ. Soc. econ. Paleontol. northern and central California. J. Geol. 86, Mineral., Tulsa, 31, 49-69. 335-52. GONZALEZ-BONORINO, G. & MIDDLETON, G. V. JOHNSON, A. M. 1970. Physical Processes in Geology. 1976. A Devonian submarine fan in western Freeman, Cooper & Co., San Francisco. Argentina. J. sediment. Petrol. 46, 56-69. JONES, J. G. 1970 Intraglacial volcanoes of the HAHN, G. A., ROSE JR, W. I. & MEYERS, T. 1979. Laugarvatn region, southwest , II. J. Geochemical correlation of genetically related Geol. 78, 127-40. rhyolitic ash-flow and air-fall ashes, central and KATO, I., MUROI, I., YAMAZAKI,T. & ABE, M. 1971. western Guatemala and the equatorial Pacific. Subaqueous pyroclastic flow deposits in the Spec. Pap. geol. Soc. Am. 180, 101-12. upper Donzurubo Formation, Nijo-san district, HAMPTON, M. A. 1972. The role of subaqueous Osaka, Japan. J. Geol. Soc. Japan, 77, 193-206. debris flow in generating turbidity currents. J. KELLER, J., RYAN, W. B. F., NINKOVICH, D. & sediment. Petrol. 42, 775-93. ALTHERR, R. 1978. Explosive volcanic activity in -- 1975. Competence of fine-grained debris flows. the Mediterranean over the past 2000,000 yr as J. sediment. Petrol. 45, 834-44. recorded in deep-sea sediments. Bull. geol. Soc. --, BOUMA, m. H. & FROST, T. P. 1979. Volcanic Am. 89, 591-604. ash in surficial sediments of the Kodiak KENNETT, J. P. 1981. Marine tephrochronolgy. In: shelf--an indicator of sediment dispersal pat- EMILIANI, C. (ed.) The Oceanic Lithosphere, Vol. terns. Mar. Geol. 29, 347-56. 7, John Wiley & Sons, Inc. Pp. 1373-436. HEIN, J. R., SCHOLL, D. W. & MILLER, J. 1978. (~ THUNELL, R. C. 1975. Global increase in Episodes of Aleutian Ridge explosive volcanism. Quaternary volcanism. Science, 187, 497-503. Science, 199, 137-41. -- & ~ 1977. On explosive Cenozoic volcanism HISCOTT, R. N. & MIDDLETON, G. V. 1979. Deposi- and climatic implications. Science, 196, 1231-4. tional mechanics of thick bedded sandstones at KLEIN, G. DEW., OKADA, H. & MITSUI, K. 1979. the base of a submarine slope, Tourelle Forma- Slope sediments in small basins associated with a tion (Lower Ordovician), Quebec, Canada. Spec. Neogene active margin, western Hokkaido Publ. Soc. econ. Paleontol. Mineral., Tulsa, 27, Island, Japan. Spec. Publ. Soc. econ. Paleontol. 307-26. Mineral., Tulsa, 27, 359-74. HONNOREZ, J. 1972. La Palagonitisation: l'alteration KOKELAAR, B. P. 1982. Fluidization of wet sediments sous-marine du verre volcanique basique de during the emplacement and cooling of various Palagonia (Sidle). Vukan-institut Immanuel igneous bodies. J. geol. Soc. London, 139, 21-33. Friedlaender No. 9, Birkhauser Verlag Basel und -- 1983. The mechanism of Surtseyan Volcanism. Stuttgart. J. geol. Soc. London, 140, 939-44. HOWELLS, M. F. (~ LEVERIDGE, B. E. 1980. The KOMAR, P. D. 1971. Hydraulic jumps in turbidity Capel Curig Volcanic Formation. Rep. Inst. Geol. currents. Bull. geol. Soc. Am. 82, 1477-88. Sci. Lond. 80(6), 1-23. KUENEN, PH. H. 1952. Estimated size of the Grand , -- ,~ EVANS, C. D. R. 1973. Ordovician ash- Banks turbidity current. Am. J. Sci. 250, 874-84. flow tufts in eastern Snowdonia. Rep. Inst. Geol. KUENZI, W. D., HORST, O. H. & MCGEHEE, R. V. Sci. Lond. 73(3), 1-33. 1979. Effect of volcanic activity on fluvial-deltaic Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 25

sedimentation on a modern arc-trench gap, Turbidites and Deep-water Sedimentation. Short southwestern Guatemala. Bull. geol. Soc. Am. Course, Soc. Econ. Paleont. Mineral. Pacific 90, 827-38. Section, May 1973, pp. 1-38. LAEROIX, A. 1904. La Montagne Pelde et Ses & -- 1976. Subaqueous sediment transport Eruptions. Masson, Paris. and deposition by sediment gravity flows. In: LAJOIE, J. 1979. Facies models 15. Volcaniclastic STANLEY, D. J. & SWIFT, D. J. P. (eds) Marine rocks. Geosci. Can. 6, 129-39. Sediment Transport and Environmental LEDBETrER, M. T. & SPARKS, R. S. J. 1979. Duration Management. John Wiley & Sons, pp. 197-218. of large-magnitude explosive eruptions deduced MITCHELL, A. H. G. 1970. Facies of an early Miocene from graded bedding in deep-sea ash layers. volcanic arc, Malekula Island, New Hebrides. Geology, 7, 240-4. Sedimentology, 14, 201-43. LEITCH, E. C. & CAWOOD, P. A. 1980. Olistoliths and & READING, H. G. 1978. Sedimentation and debris flow deposits at ancient consuming plate tectonics. In: READING, H. G. (ed.) Sedimentary margins: an eastern Australian example. J. Environments and Facies, 14, 439-76. Blackwell sediment. Petrol. 25, 5-22. Scientific Publications, Oxford. LEWIS, D. W. 1976. Subaqueous debris flows of early MOORE, J. G. 1965. Petrology of deep-sea basalt near Pleistocene age at Motunau, North Canterbury, Hawaii. Am. J. Sci. 263, 40-52. New Zealand. N.Z. Jl Geol. Geophys. 19, 1970. Water content of basalt erupted on the 535-67. ocean floor. Contrib. Mineral. Petrol. 28, 272-9. LONSDALE, P. F. 1975. Sedimentation and tectonic 1975. Mechanism of formation of pillows. Am. modification of the Samoan archipelagic apron. Sci. 63, 269-77. Bull. Am. Ass. Petrol. Geol. 59, 780-98. & FISKE, R. S. 1969. Volcanic substructure 1977. Abyssal pahoehoe with lava coils at the inferred from dredge samples and ocean-bottom Galapagos Rift. Geology, 5, 147-52. photographs, Hawaii. Bull. geol. Soc. Am. 80, & SPIESS, F. N. 1979. A pair of young cratered 1191-202. volcanoes on the east Pacific rise. J. Geol. 87, -- & SCHILLING, J.-G. 1973. Vesicles, water, and 157-73. sulfur in Reykjanes Ridge basalts. Contrib. LOWE, D. R. 1976. Grain flow and grain flow Mineral. Petrol. 41, 105-18. deposits. J. sediment. Petrol. 46, 188-99. & SlSSON, T. W. 1981. Deposits and effects of 1979. Sediment gravity flows: their classification the May 18 pyroclastic surge. In: LIPMAN, P. W. and some problems of application to natural & MULLINEAUX, D. R. (eds) The 1980 Eruptions flows and deposits. Spec. Publ. Soc. econ. of Mount St Helens, Washington. Prof. Pap. U.S. Paleontol. Mineral., Tulsa, 27, 75-82. geol. Surv. 1250, 421-38. 1982. Sediment gravity flows. II. Depositional MORGENSTERN, N. R. 1967. Submarine slumping and models with special reference to the deposits of the initiation of turbidity currents. In: RICHARDS, high-density turbidity currents. J. sediment. A. F. (ed.) Marine Geotechnique, pp. 189-220. Petrol. 52, 279-97. NAKAMURA, K. 1974. Preliminary estimate of global LOWMAN, R. D. W. & BLOXAM, T. W. 1981. The volcanic production rate. In: FURUMOTO, S., petrology of the lower Paleozoic Fishguard MINIKAMI, T. & YUHARA, K. (eds) The Volcanic Group and associated rocks E of Utilization of Volcano Energy. The Proceedings Fishguard, N Pembrokeshire (Dyfed), South of U.S.-Japan Cooperative Science Seminar, Wales. J. geol. Soc. London, 138, 47-68. Sandia Laboratories, Albuquerque, New MACDONALD, G. A. 1972. Volcanoes. Prentice-Hall, Mexico. Pp. 273-84. New Jersey. NARDIN, T. R., HEIN, F. J., GORSLINE, D. S. & MANSFIELD, C. F. 1979. Upper Mesozoic subsea fan EDWARDS, B. D. 1979. A review of mass deposits in the southern Diablo Range, movement processes, sediment and acoustic California: Record of the Sierra Nevada characteristics, and contrasts in slope and magmatic arc. Bull. geol. Soc. Am. 90, base-of-slope systems versus canyon-fan-basin 1025-46. floor systems. Spec. Publ. Soc. econ. Paleontol. MCBIRNEY, A. R. 1963. Factors governing the nature Mineral., Tulsa, 27, 61-73. of submarine volcanism. Bull. Volcanol. 26, NAVLOR, M. A. 1980. The origin of inverse grading in 455-69. muddy debris flow deposits--a review. J. MCBRIDE, E. F. 1978. Olistostrome in the Tesnus sediment. Petrol. 50, 1111-6. Formation (Mississippian-Pennsylvanian), 1981. Debris flow (olistostromes) and slumping Payne Hills, Marathon region, Texas. Bull. Geol. on a distal passive continental margin: the Soc. Am. 89, 1550-8. Palombini limestone-shale sequence of the MIDDLETON, G. V. 1967. Experiments on density and northern Apennines. Sedimentology, 28, 837-52. turbidity currents. III. Deposition of sediment. NAVUDU, Y. R. 1964. Volcanic ash deposits in the Can. J. Earth Sci. 4, 475-505. Gulf of Alaska and problems of correlation of 1970. Experimental studies related to problems deep-sea ash deposits. Mar. Geol. 1, 194-212. of flysch sedimentation. Spec. Pap. geol. Ass. NEEDHAM, J. D. & FRANCHETEAU, J. 1974. Some Can. 7, 253-72. characteristics of the rift valley in the Atlantic & HAMPTON, M. A. 1973. Sediment gravity Ocean near 36°48'N. Earth planet. Sci. Lett. 22, flows: Mechanics of flow and deposition. In: 29-43. MIDDLETON, G. V. & BOUMA, A. H. (eds) NEMEC, W., POREBSKI, S. J. & STEEL, R. J. 1980. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

26 R. V. Fisher

Texture and structure of resedimented Canary Island volcanism inferred from ash layers conglomerates: examples from Ksiaz Formation and volcaniclastic standstones of DSDP Site 397 (Famenian-Tournasian), southwestern Poland. (Leg 47A). Init. Rep. Deep Sea drill. Proj. 47(1), Sedimentology, 27, 519-38. 703-25. NIEM, m. R. 1977. Mississippian pyroclastic flow and SELF, S. & RAMPINO, M. R. 1981. The 1883 eruption ash-fall deposits in the deep-marine Ouachita of Krakatau. Nature, LoAd. 294, 699-704. flysch basin, Oklahoma and Arkansas. Bull. SHAW, D. M., WATKINS, N. D. & HUANG, T. C. 1974. geol. Soc. Am. 88, 49--61. Atmospherically transported volcanic dust in NINKOVICH, D. 1979. Distribution, age and chemical deep sea sediments: theoretical consideration. J. composition of tephra layers in deep-sea geophys. Res. 79, 3087-97. sediments off western Indonesia. J. Volcan. SHERIDAN, M. F. 1979. Emplacement of pyroclastic geoth. Res. Amsterdam, 5, 67-86. flows: a review. Spec. Pap. geol. Soc. Am. 180, & DONN, W. L. 1976. Explosive Cenozoic 125-36. volcanism and climatic implications. Science, 194, SIGURDSSON, H., SPARKS, R. S. J., CAREY, S. N. t~ 899-906. HUANG, T. C. 1980. Volcanogenic sedimentation & HEEZEN, B. C. 1965. tephra. In: in the Lesser Antilles Arc. J. Geol. 88, 523-40. Submarine Geology and Geophysics (Colston SMITJJ, R. L. 1960. Ash flows. Bull. geol. Soc. Am. Papers No. 17), 27, 413-53. Butterworths 71, 795-842. Scientific Publications, London. SPARKS, R. S. J. 1976. Grain size variations in & ~ 1967. Physical and chemical properties and implications for the transport of of volcanic glass shards from ash, pyroclastic flows. Sedimentology, 23, 147-88. Thera Island, and from upper and lower ash --., SIGURDSSON, H. & EARLY, S. 1980a. The layers in eastern Mediterranean deep-sea cores. entrance of pyroclastic flows into the sea. I. Nature, LoAd. 213, 582-4. Oceanographic and geologic evidence evidence & SHACKLETON, N. J. 1975. Distribution, from Dominica. Lesser Antilles. J. Volcan. geoth. stratigraphic position and age of ash layer 'L', in Res. Amsterdam, 7, 87-96. the Panama Basin region. Earth planet. Sci. Lett. -- &- 1980b. The entrance of pyroclastic 27, 20-34. flows into the sea. II. Theoretical considerations .... , SPARKS, R. S. J. & LEDBETTER, M. T. 1978. The on subaqueous emplacement and welding. J. exceptional magnitude and intensity of the Toba Volcan. geoth. Res. Amsterdam, 7, 97-105. eruption, Sumatra: An example of the use of STANLEY, D. J. & TAYLOR, P. T. 1977. Sediment deep-sea tephra layers as a geological tool. Bull. transport down a flank by a combined Volcanol. 41-3, 286-98. current and gravity process. Mar. Geol. 23, PETERSON, D. W. 1976. Processes of volcanic island 77-88. growth, Kilauea Volcano, Hawaii, 1969-1973. , PALMER, H. D. & DILL, R. F. 1978. Coarse In: FERRAN, O. G. (ed.) Proceedings of the sediment transport by mass flow and turbidity Symposium 'Andean and Antarctic current processes and downslope transforma- Problems'. Spec. Series Internat. Assoc. of tions in Annot Sandstone Canyon-Fan Volcanology and Chemistry of the Earth's Valley systems. In: STANLEY, D. J. & KELLING? Interior, 172-89. G. (eds) Sedimentation in Submarine Canyons, RIEHLE, J. R. 1973. Calculated compaction profiles Fans and Trenches. Dowden, Hutchinson & Ross, of rhyolitic ash-flow tufts. Bull. geol. Soc. Am. Stroudsburg, Penn. 85-115. 84, 2193-216. STANTON, W. I. 1960. The lower Paleozoic rocks of RITI~MANN, A. 1962. Volcanoes and Their Activity. south-west Murrisk, Ireland. J. geol. Soc. John Wiley and Sons, New York. London, 116, 269-96. ROBINSON, P. T., FLOWER, M. F. J., SWANSON, D. A. STAUDIGEL, H. • SCHMINCKE, H.-U. 1981. Structural & STAUDIGEL, H. 1980. Lithology and eruptive evolution of a seamount: evidence from the stratigraphy of Cretaceous oceanic crust, western uplifted intraplate seamount on the island of La Atlantic ocean. Init. Rep. Deep Sea drill. Proj. Palma, Canary Islands. Los, 62, 1075. 51-3, i535-55. SWARBRICK, R. E. & NAYLOR, M. A. 1980. The SCHEIDEGGER, K. F., JEZEK, P. m. & NINKOVICH, D. Kathikas melange, SW Cyprus, Cretaceous 1978. Chemical and optical studies of glass submarine debris flows. Sedimentology, 26, shards in Pleistocene and Pliocene ash layers 63-78. from DSDP Site 192, northwest Pacific Ocean. J. TASSE, N., LAJOIE, J. & DIMROTH, E. 1978. The Volcan. geoth. Res. Amsterdam, 4, 99-116. anatomy and interpretation of an Archean , CORLISS, J. B., JEZEK, P. A. & NINKOVICH, D. volcaniclastic sequence, Noranda region, 1980. Composition of deep-sea ash layers Quebec. Can. J. Earth Sci. 15, 874-88. derived from north Pacific volcanic arcs: THUNELL, R., FEDERMAN, A., SPARKS, S. & WILLAMS, variations in time and space. J. Volcan. geoth. D. 1979. The age, origin and volcanological Res. Amsterdam, 7, 107-37. significance of the Y-5 ash layer in the SCHMINCKE, H.-U. 1967. Graded lahars in the type Mediterranean. Quat. Res. 12, 241-53. section of the Ellensburg Formation, VAN ANDEL, TJ. H. & KOMAR, P. D. 1969. Ponded south-central Washington. J. sediment. Petrol. sediments of the Mid-Atlantic Ridge between 37, 438-48. 22 ° and 23 ° North latitude. Bull. geol. Soc. Am. & YON RAD, U. 1979. Neogene evolution of 80, 1163-90. Downloaded from http://sp.lyellcollection.org/ by guest on September 25, 2021

Submarine volcaniclastic rocks 27

VESSEL, R. K. & DAVIES, D. K. 1981. Nonmarine experimental approach. J. Volcan. geoth. Res. sedimentation in an active forre arc basin. Spec. Amsterdam, 8, 231-49. Publ. Soc. econ. Paleontol. Mineral., Tulsa, 31, WOHLETZ, K. H. 1980. Explosive Hydromagmatic 31-45. Volcanism. Thesis, Ph.D. Arizona State WALKER, G. P. L. 1979. A volcanic ash generated by University (unpub.). explosions where ignimbrite entered the sea. SHERIDAN, M. F. 1979. A model of Nature, Lond. 281, 642-6. pyroclastic surge. Spec. Pap. geol. Soc. Am. 180, WATKINS, N. D. t~ HUANG, T. C. 1977. in 177-94. abyssal sediments east of the North Island, New WRJGHT, J. V. & MUTrI, E. 1981. The Dali Ash, Zealand: Chronology, paleowind velocity, and Island of Rhodes, Greece: a problem in paleoexplosivity. N.Z. Jl Geol. Geophys. 20, interpreting submarine volcanigenic sediments. 179-98. Bull. Volcanol. 44-2, 153-67. , SPARKS, R. S. J., SIGURDSSON, H., HUANG, YAMADA, E. 1973. Subaqueous pumice flow deposits T. C., FEDERMAN,A., CAREY, S. & NINKOVICH,D. in the Onikobe Caldera, Miyagi Prefecture, 1978. Volume and extent of the Minoan tephra Japan. J. geol. Soc. Japan, 79, 589-97. from Santorini Volcano: new evidence from YAMAZAKI, T., KATO, I., MUROI, I. & ABE, M. 1973. deep-sea sediment cores. Nature, Lond. 271, Textural analysis and flow mechanism of the 122-6. Donzurubo subaqueous pyroclastic flow W~LSON, C. J. N. 1980. The role of fluidization in the deposits. Bull. Volcanol. 37, 231-44. emplacement of pyroclastic flows: An

RICHARD V. FISHER, Department of Geological Sciences, University of California, Santa Barbara, CA 93106, U.S.A.