You are on page 1of 7

Bull Volcanol (2004) 66:485491 DOI 10.

1007/s00445-003-0332-8

RESEARCH ARTICLE

M. J. Branney T. L. Barry M. Godchaux

Sheathfolds in rheomorphic ignimbrites

Received: 3 July 2003 / Accepted: 3 November 2003 / Published online: 18 February 2004  Springer-Verlag 2004

Abstract Structural reappraisal of several classic rheomorphic ignimbrites in Colorado, Idaho, the Canary Islands and Italy has, for the first time, revealed abundant oblique folds, curvilinear folds and sheathfolds which formed during emplacement. Like their equivalents in tectonic shear-zones, the sheathfold axes lie sub-parallel to a pervasive elongation lineation, and appear as eye structures on rock surfaces normal to the transport direction. With the recognition of sheathfolds, ignimbrites previously inferred to have undergone complex rheomorphic deformation histories are re-interpreted as recording a single, progressive deformation event. In some examples, the trends of sheathfolds and related lineations change with height through a single ignimbrite suggesting that rheomorphism did not affect the entire thickness of ignimbrite synchronously. Instead, we infer that in these ignimbrites a thin ductile shear-zone rose gradually through the aggrading agglutinating mass whilst the flow direction varied with time. This suggests that, in some cases, both welding and rheomorphism can be extremely rapid, with ductile strain rates significantly exceeding rates of ignimbrite aggradation. Keywords Ignimbrite Pyroclastic flows Rheomorphism Shear zone Sheathfold Welded tuff
Editorial responsibility: T. Druitt M. J. Branney ()) Department of Geology, University of Leicester, University Road, Leicester, LEI 7RH, UK e-mail: mjb26@leicester.ac.uk Tel.: +44-0116-2523647 Fax: +44-0116-2523918 T. L. Barry Cardiff University, c/o NIGL/BAS, BGS, Keyworth, NG12 566, UK M. Godchaux Mount Holyoke College & Idaho Geological Survey, Moscow, ID, 83843, USA

Introduction
Eruptions that produce rheomorphic ignimbrites represent a peculiarly awesome and destructive volcanic hazard; they can inundate entire landscapes with searing-hot glass in minutes. Rheomorphism is the ductile deformation of hot, welded pyroclastic material during and just after deposition. Since first descriptions in the 1960s, rheomorphic ignimbrites have increasingly been recognized as a common and widespread volcanic phenomenon, with volumes of some exceeding 100s of km3 (e.g. Bonnichsen et al. 1989). They form a significant component of volcanism in diverse volcanic settings, including intracontinental volcanic provinces (e.g. Leat 1985; Henry et al. 1989; Bonnichsen et al. 1989; Milner et al. 1992); ocean island volcanoes (Wolff and Wright 1981; Schmincke et al. 1993); and in continental arcs (Branney et al. 1992; Kokelaar and Kniger 2000). Ignimbrites undergoing rheomorphism develop a variety of ductile deformation structures, including folded and attenuated pumices, vesicles and welding fabrics (Schmincke and Swanson 1967; Wolff and Wright 1981; Sumner and Branney 2002). These features develop in the viscous welded mass whilst it is still hot and degassing. Some rheomorphic ignimbrites develop flow banding and/or upper and marginal autobreccias similar to those of viscous lavas (Branney et al. 1992). However, most rheomophic ignimbrites can be readily distinguished from lavas on the basis of preserved pyroclastic features (e.g. Henry and Wolff 1992). For rheomorphism to occur, pyroclasts must be sufficiently fluidal at the time of welding to readily deform. This condition is favoured, variously, by high emplacement-temperatures, eruptions of high mass-flux and low pyroclastic fountaining, rapid deposition, strongly peralkaline chemistries, and high contents of dissolved volatiles such as water and/or halogens within the pyroclasts at the time of deposition (e.g. Mahood 1984).

486

Previous work
Few structural studies have been undertaken to determine the nature, timing, and causes of rheomorphic deformation. An early summary reported pumices (fiamme) elongated parallel to the inferred flow direction of a pyroclastic density current, with rheomorphic fold axes commonly oriented perpendicular to this direction (Schmincke and Swanson 1967). An influential structural study was of the Wall Mountain Tuff, an ignimbrite that occupies a paleovalley at Gribbles Run, Colorado, USA (Chapin and Lowell 1979). Complex fold orientations were interpreted as recording two fold generations. The first generation was inferred to have formed with foldaxes perpendicular to the valley axis (i.e. perpendicular to the flow direction of the pyroclastic current), and was taken to indicate that welding (sometimes known as primary welding) and deformation occurred extremely rapidly, during deposition. A second generation of folds was inferred to have developed from hot slumping of already deposited ignimbrite down the valley sides towards the valley floor (secondary viscous mass flowage). This process was invoked to account for folds

whose axes lie sub-parallel to the paleovalley axis. In rheomorphic ignimbrites elsewhere, a similar two-fold deformation history was invoked to account for disparate fold orientations (e.g. Schmincke et al. 1993), and debate has focused upon the extent to which rheomorphism in ignimbrites is syn-depositional or due to post-depositional remobilisation (e.g. Wolff and Wright 1981; Branney and Kokelaar 1992). In the following, we report the results of new structural investigations on 10 rheomorphic ignimbrites, and present a simplified kinematic interpretation of ignimbrite rheomorphism.

Re-appraisal of the Wall Mountain Tuff, Colorado


We revisited the Wall Mountain Tuff at Gribbles Run paleovalley (Fig. 1) to re-assess its rheomorphic deformation. Structural analysis confirms the presence of abundant small-scale rheomorphic folds and a single, pervasive elongation lineation oriented subparallel to the axis of the paleovalley (Fig. 1). By far the majority of fold axes are oriented sub-parallel to the paleovalley axis (Fig. 1); these were the folds that Chapin and Lowell

Fig. 1 Map of rheomorphic elongation lineations and fold axes in the Wall Mountain Tuff (grey) in Gribbles Run paleovalley, northeast of Salida, Colorado. The pyroclastic density current travelled in a southeasterly direction along the Oligocene valley. The lower hemisphere stereogram shows poles to welding folia-

tions (open circles), elongation lineations (crosses) and fold axes (dots) in the eastern part of the area (for western part see Fig. 3). Asterisked fold axis and lineation trends are from Chapin and Lowell (1979)

487

Fig. 2 Sheathfolds in rheomorphic ignimbrites; all viewed along the elongation lineation. Inferred pyroclastic density current flow directions toward the viewer (AD, FH) and away from viewer (E). A Eye structure, Wall Mountain Tuff, Colorado (Fig. 1). Pole is 1 m long. B Eye structure near the base of ignimbrite TL, Gran Canaria. End of ruler shows centimetres. C Curvilinear rheomorphic fold in ignimbrite Q at Bue Marino, Pantelleria, Italy. The fold axis curves through 130 and trends into rock face on both left and right hand side of recumbent culmination (centre). Metre rule.

D Eye structure in ignimbrite D, Gran Canaria. Note pronounced elongation lineation (bottom). Coin for scale. E Eye structure in ignimbrite E, Gran Canaria. F Eye structures in the House Creek ignimbrite, 33 km WSW of Rogerson, Idaho (T15S R13E; Cedar Creek quadrangle). Notebook shows 10 cm. G Eye structures in ignimbrite VI, Gran Canaria. End of metre rule shows centimetres. H Eye structure in the Greys Landing ignimbrite, Idaho (see text). Notebook shows 10 cm. See Fig. 3 for locations of A, B, D, E, G

(1979) had interpreted as secondary, i.e. hot slumps down valley-sides. In contrast, fold-axes perpendicular to the axis of the paleovalley are rare: these folds were inferred by Chapin and Lowell (1979) as primary and formed by shear exerted by the initial pyroclastic current as it flowed along the valley. Our examination revealed that these folds are markedly curvilinear. For example, one photographed by Chapin and Lowell (1979, their Fig. 13) is a recumbent fold with a subhorizontal axial surface and, in plan view, a curvilinear fold-axis that is convex towards the SE. Elongation lineations at this outcrop also trend SE (150). The fold axis curves through 129 before it disappears into the outcrop, and we interpret it as the tight culmination of a sheath fold formed by ductile shear in a down-valley (SE) transport direction. The fold is part of a fold pair that shows vergence to the southeast when viewed perpendicular to

the lineation. Another example of a primary fold photographed by Chapin and Lowell (1979, their Fig. 17) is a small, upright open flexure caused by ductile shear around a rigid agglutinate clast 40 cm in diameter, forming a sigma-object (Passchier and Trouw 1996), again consistent with transport to the SE. Neither fold indicates secondary flow down valley sides (i.e. to the SW or NE). All folds seen with axes at high angles to the valley axis were curvilinear, and most showed vergence to the SE. The presence of a single valley-parallel pervasive elongation lineation (prolate fabric) oriented sub-parallel to the majority of the fold axes (Fig. 1), suggests that rather than registering two fold phases of different origin, all the folds and the lineations record a single, progressive deformation event, caused by ductile shear towards the SE along the Gribbles Run valley. We infer that folds

488

Fig. 3 Lower hemisphere equal-area projections, showing elongation lineations (crosses) fold axes (dots) and poles to welding foliations (circles). Large arrowheads indicate approximate paleoflow directions for each pyroclastic density current in cases where these could be independently inferred from the location of the eruptive source and the topography. In each ignimbrite, the fold axes cluster and lie sub-parallel to the similarly clustered elongation lineations. They define the local rheomorphic transport direction, which is sub-parallel to the inferred flow direction of

each pyroclastic current. Locations: Gran Canarian ignimbrites VI, U, TL, D and E are, respectively, from west Montaa Carboneras, Anden Verde, eastern Montaa Carboneras, Barranco de Taurito, and Barranco de los Medios Almudes; the Green Tuff, Kartibucale, Pantelleria, Italy; the Wall Mountain Tuff, Colorado (west side of Fig. 1); the Greys Landing ignimbrite, the Three Creek Rhyolite and the House Creek ignimbrite were from the Rogerson Graben, south and southwest of Twin Falls in southern Idaho, USA (Fig. 4)

initiated with axial trends at high angles to the valley-axis were transposed into the direction of shear during progressive rheomorphic deformation; that is, the fold axes progressively rotated to become sub-parallel to the trend of the valley along which the pyroclastic density current flowed (Fig. 1). With increasing shear strain, early-formed upright or inclined fold axial surfaces were progressively transposed to become sub-horizontal (recumbent) or with upcurrent dips of less than 10 (see top right of Figs. 1 and 3). Evidence for this style of deformation can be seen on rock surfaces oriented perpendicular to the paleovalley axis and the elongation lineation, exhibiting eye structures (Fig. 2A). Eye structures are sections through sheathfold culminations and saddles, which form by the progressive transposition and attenuation of curvilinear fold axes by ductile shear. Most of the folds at Gribbles Run are highly attenuated isoclinal oblique folds (term from Passchier and Trouw 1996) and sheathfolds, with axes virtually parallel to the

elongation lineation (Figs. 1, 2A and 3). The rare examples of fold axes at higher angles to the valleyparallel elongation lineation are either late gentle flexures, which have been subjected to less shear and so have not become so transposed, or they are highly curvilinear and represent the crest regions of recumbent sheathfolds that verge to the SE. We infer all the folds formed in the same shearing regime with transport to the SE. Overall, the deformation structures in the Wall Mountain Tuff are closely similar to those in tectonic ductile shear-zones (e.g. Alsop and Holdsworth 2002); the hot ignimbrite behaved as a ductile shear zone. The Wall Mountain Tuff at Gribbles Run records no evidence of a second ductile deformation event with downslope transport direction towards the valley axis. This is consistent with the absence of a lineation perpendicular to the paleovalley axis; Chapin and Lowell (1979) astutely noted this absence, despite not being able to account for it. Additional support for our interpretation

489

lies in the relationship between the lineation trends and the folds. Chapin and Lowell (1979) considered that the valley axis-perpendicular folds were the first folds to develop. However, the folds deform the valley-parallel elongation lineation, and so must have formed slightly later. We conclude that all the folds and elongation lineations formed by rheomorphic flow along the valley and may have, in part, initiated as a result of heterogeneous rheomorphic flow (flow perturbation folds of Alsop and Holdsworth 2002).

ited abundant small-scale folds with fold axes that trended predominantly sub-parallel to the elongation lineation (Fig. 3). We conclude that the sheathfolds and related features are common, if not characteristic of, intensely rheomorphic ignimbrites, irrespective of their chemical composition or volcano-tectonic setting. We also found similar relations between fold axes and lineations in the Three Creek Rhyolite of southern Idaho (Fig. 3). This is a lava-like rhyolite sheet with characteristics, such as impersistent basal breccias, intermediate between those characteristic of true lavas and rheomorphic ignimbrites, and whose mode of emplacement is not yet determined.

Sheathfolds in other ignimbrites


A deformation history involving both valley axis-parallel and perpendicular deformation phases had also been inferred for Miocene peralkaline ignimbrite VI on Gran Canaria, Canary Islands (Schmincke et al. 1993). Schmincke had reported orientations of rheomorphic fold axes and lineations indicative of both valley axis-parallel deformation and of local secondary rheomorphic deformation down the valley sides near Montaa Carboneras. We revisited this site (the west side of Montaa Carboneras) where a small SE-trending paleovalley is partly filled by ignimbrite VI. Rheomorphic folds and elongation lineations are abundant, and structural analysis reveals just one progressive deformation event, just as in the Wall Mountain Tuff. Both the lineations and the fold-axes in ignimbrite VI are oriented parallel to the paleovalley axis, i.e. subparallel to the inferred transport direction of the pyroclastic density current (Fig. 3). Virtually all the folds are oblique folds or sheath folds, and eye structures are abundant on rock surfaces oriented at high angles to the elongation lineation (Fig. 2G). To determine whether sheathfolds are widespread in rheomorphic ignimbrites, we examined eight other rheomorphic ignimbrites in different settings in the US, Italy and Canary Islands. These were the ignimbrites TL, U, D and E of Gran Canaria (Schmincke et al. 1993); the Green Tuff and ignimbrite Q at Pantelleria (Mahood and Hildreth 1986); the Greys Landing ignimbrite (see below) and the 3 m-thick House Creek ignimbrite of southwest Idaho. Although the ignimbrites had different thicknesses, compositions, welding intensities, and topographic settings, they had several structural features in common. All were found to exhibit sheathfolds and oblique folds (e.g. Figs. 2 and 3). Each also exhibited a single, pervasive elongation lineation lying along the welding foliation and parallel to the likely flow-direction of the pyroclastic density current where this could be inferred independently from the location of the eruptive source and the regional substrate topography (Fig. 3). In all cases the plunges of the elongation lineations were either sub-parallel to the former land surface or oblique such that they plunged predominantly sourceward. Such obliquity (sometimes loosely termed fabric imbrication) is a well-known kinematic indicator in rheomorphic ignimbrites and indicates the sense of ductile shear (Branney and Kokelaar 1992). All the ignimbrites exhib-

Variations in sheathfold trend with height


To determine how consistent the orientation of sheathfolds are with height through a single ignimbrite, we visited the Greys Landing ignimbrite, which is superbly exposed around Salmon Dam, Twin Falls County, southern Idaho (Fig. 4, inset). It reaches 50 m thick and has a simple welding profile with thin upper and lower vitrophyres separated by a thick, massive devitrified centre. It

Fig. 4 Vertical section through the Greys Landing ignimbrite at Greys Landing, Twin Falls County, Idaho, USA (see inset). Trends of fold axes closely correspond to the trends of elongation lineations and both change gradually with height. The top few metres of the ignimbrite have been removed by erosion at this location

490

is inferred to represent a single cooling unit, and there are no obvious internal breaks or flow-unit boundaries. It has a lava-like appearance and pumiceous and vitroclastic textures are rare. Its ignimbrite origin is inferred on the basis of the combination of criteria used by Bonnichsen and Kauffman (1989) in the Snake River Plain volcanic province, notably the absence of: (1) a basal autobreccia, (2) lobate morphology; (3) an irregular upper surface; and (4) blunt flow margins; rhyolite lavas in the Snake River province of Idaho exhibit all these features. The ignimbrite exhibits a ubiquitous foliation that is folded into abundant, mostly dm-scale to m-scale oblique folds and sheath folds (e.g. Fig. 2H). The folds, and a prominent elongation lineation along the foliation, occur at all levels above 5 m. The trends of the sheathfold axes vary systematically with height through the ignimbrite (Fig. 4). The lowest fold axes trend SSE (168) and the trend shifts gradually to ENE (070) at a height of 31 m. Above this, the trend gradually shifts back to SE (130). The trends of the elongation lineations lie subparallel to the trends of the sheathfold axes, and they closely track the vertical changes in fold trend with height (Fig. 4). These folds and fabrics indicate that the shear-sense azimuth of rheomorphism gradually shifted with time, first anticlockwise through 100 and then clockwise through 60. It has been proposed that, although rheomorphism likely continues after ignimbrite deposition has ceased, rapid welding (agglutination) and rheomorphism may in some cases begin during deposition, while the ignimbrite is gradually aggrading (Branney and Kokelaar 1992). The varying sheathfold orientation data in the Greys Landing ignimbrite (Fig. 4) is probably the best evidence thus far that welding and rheomorphic folding can begin while the ignimbrite is still being deposited. The changes in sheathfold orientations with height may even record temporal changes in the flow direction of the lowermost parts of the sustained, depositing pyroclastic density current. Temporal changes in flow-direction of sustained pyroclastic density currents have been recorded by vertical changes in the trend of sedimentary fabrics in non-welded ignimbrites (e.g. Capaccioni et al. 2001) and may result from unsteady flow and changing topography, for example due to the uneven accumulation of aggrading deposit (Branney and Kokelaar 2002). In the Greys Landing ignimbrite, the changes in sheathfold axes and lineations with height suggest that the shear sense within a relatively thin (ca. 2 m) and rising shear zone changed direction with time. This may be because the pyroclastic current exerted sufficient shear stress upon the successive uppermost parts of the hot agglutinate to cause rheomorphic deformation. Alternatively, it may be because the uppermost parts of the agglutinate were sufficiently fluidal to deform gravitationally in different directions in response to changing slopes produced on an evolving irregular aggradation surface. Either way, it indicates that rheomorphism can occur at surprisingly rapid strain rates; that is, rates that significantly exceed the rate of ignimbrite aggradation.

The Greys Landing ignimbrite is an extensive, sheetlike ignimbrite with a sub-horizontal basal contact. We infer that the pyroclastic density current that deposited it was not confined by topography, and was free to shift and meander so that it fluctuated through a total of 160. In marked contrast, the trends of sheathfold axes and attendant prolate lineations within the valley-infilling Wall Mountain Tuff and ignimbrite VI (described above) remained constant with height at a single geographic location. We infer that this difference is because the latter ignimbrites were deposited from pyroclastic currents that were channelled by the valley topography so that a relatively constant flow direction was maintained, at least in lowermost parts of the currents. A possible alternative explanation for changes in azimuth orientation of folds and lineations with height, e.g. in the Greys Landing Ignimbrite, might be that all the folds and lineations formed with a common orientation (NNWSSE trending), but that the central zone underwent late-stage rheomorphism with different flow direction after the top and base of the ignimbrite sheet had cooled and ceased movement. However, we have not found evidence for any ENEWSW refolding of a former NNWSSE-trending structure within the central zone. In summary, we have reported for the first time the occurrence of rheomorphic sheathfolds in 10 ignimbrites, and have shown how folds with different orientations may form during a single, progressive ductile deformation event (emplacement). Orientations of rheomorphic sheathfolds other curvilinear folds and oblige folds and attendant elongation lineations provide useful information about pyroclastic flow directions, and suggest that in some ignimbrites welding and rheomorphism may occur extremely rapidly, during deposition. We anticipate that sheathfolds in intrusions and in lavas similarly have potential to provide information about magma flow and emplacement.
Acknowledgements Many thanks to Bill Bonnichsen, Mike McCurry, Chuck Chapin, Janet Summer, and Steve Temperley for useful discussion. Laurence Coogan, Graham Andrews, and reviewers John Wolff and Anita Grunder provided helpful comments.

References
Alsop GI, Holdsworth RE (2002) The geometry and kinematics of flow perturbation folds. Tectonophysics 350:99125 Bonnichsen B, Christiansen RL, Morgan LA, Moye FJ, Hackett WR, Leeman WP, Honjo N, Jenks MD, Godchaux MM (1989) Excursion 4A: silicic volcanic rocks in the Snake River Plain Yellowstone Plateau province. New Mexico Bureau of Mines and Mineral Resour Mem 47:135182 Bonnichsen B, Kaufmann DF (1989) Physical features of rhyolite lava flows in the Snake River Plain volcanic province, southwestern Idaho. Geol Soc Am Spec Pap 212:119145 Branney MJ, Kokelaar BP (1992) A reappraisal of ignimbrite emplacement: changes from particulate to non-particulate flow during progressive aggradation of high-grade ignimbrite. Bull Volcanol 54:50420

491 Branney MJ, Kokelaar BP (2002) Pyroclastic density currents and the sedimentation of ignimbrites. Geol Soc Lond Mem 27:1 143 Branney MJ, Kokelaar BP, McConnell BJ (1992) The Bad Step Tuff: a lava-like rheomorphic ignimbrite in a calc-alkaline piecemeal caldera, English Lake District. Bull Volcanol 54:187199 Capaccioni B, Nappi G, Valentini L (2001) Directional fabric measurements: an investigative approach to transport and depositional mechanisms in pyroclastic flows. J Volcanol Geotherm Res 107:275292 Chapin CE, Lowell GR (1979) Primary and secondary flow structures in ash-flow tuffs of the Gribbles Run Paleovalley, central Colorado. In: Chapin CE, Elston WE (eds) Ash flow tuffs. Geol Soc Am Spec Pap 180:137154 Henry CD, Price JG, Parker DF, Wolff JA (1989) Excursion 9A: Mid-Tertiary silicic alkalic volcanism of Trans-Pecos Texas: rheomorphic tuffs and extensive silicic lavas. New Mexico Bureau of Mines and Mineral Resour Mem 46:231274 Henry CD, Wolff JA (1992) Distinguishing strongly rheomorphic tuffs from extensive silicic lavas. Bull Volcanol 54:171186 Kokelaar P, Kniger S (2000) Marine emplacement of welded ignimbrite: the Ordovician Pitts Head Tuff, North Wales. J Geol Soc Lond 157:517536 Leat PT (1985) Facies variations in peralkaline ash-flow tuffs from the Kenya Rift Valley. Geol Mag 122:139150 Mahood GA (1984) Pyroclastic rocks and calderas associated with strongly peralkaline volcanic rocks. J Geophys Res 89:8540 8552 Mahood GA, Hildreth W (1986) Geology of the peralkaline volcano at Pantelleria, Straits of Sicily. Bull Volcanol 48:143 172 Milner SC, Duncan AR, Ewart A (1992) Quartz latite rheoignimbrite flows of the Etendeka Formation, north-western Namibia. Bull Volcanol 54:200219 Passchier CW, Trouw RAJ (1996) Microtectonics. Springer, Berlin Heidelberg New York, 289 pp Schmincke H-U, Swanson DA (1967) Laminar viscous flowage structures in ash-flow tuffs from Gran Canaria, Canary Islands. J Geol 75:641664 Schmincke H-U, with contributions by Freundt A, Ferriz H, Kobberger G, Leat P (1993) Geological field guide to Gran Canaria, 6th edn. Pluto Press, Kiel, FRG, pp 1227 Sumner JM, Branney MJ (2002) Emplacement and deformation of a heterogeneous, chemically zoned, rheomorphic and locally lava-like peralkaline ignimbrite sheet: TL on Gran Canaria. J Volcanol Geotherm Res 115:109138 Wolff JA, Wright JV (1981) Rheomorphism of welded tuffs. J Volcanol Geotherm Res 10:1334

You might also like