Fissure ignimbrites: Fissure-source origin for voluminous ignimbrites of the Sierra Madre Occidental and its relationship with

Basin and Range faulting

Gerardo J. Aguirre-D az* Centro de Geociencias, Campus-Juriquilla, Universidad Nacional Auto noma de Me xico, Quere taro, Guillermo Labarthe-Herna ndez*
Qro. 76230, Mexico Instituto de Geolog a, Universidad Auto noma de San Luis Potos , San Luis Potos , Mexico ever, along three transects that have been geologically mapped in the Sierra Madre Occidentalin the north, Chihuahua to Hermosillo; in the center, Durango to Mazatla n; and in the south, Zacatecas to Tepic (Fig. 1)and elsewhere throughout the Sierra Madre Occidental (see references in Table DR11), calderas are not evident for most ignimbrites. We present evidence of ignimbrites related to ssures, most of them with the regional northwest to northeast trend of the Basin and Range faults, which served as conduits for the eruption of large-volume ignimbrites. We conclude that ssure-type eruptions served as an alternative, if not the most important, source for the ignimbrites of the Sierra Madre Occidental, which we term ssure ignimbrites. LARGE VOLUME OF THE SIERRA MADRE OCCIDENTAL IGNIMBRITES The area covered by the Sierra Madre Occidental proper, i.e., the largest continuous ignimbrite outcrop shown in Figure 1, is 393,000 km2 (Sierra Madre Occidental main in Table 1). On the basis of several studies in the sierra (Table DR1; see footnote 1), the thickness of the ignimbrite package averages 1000 m. Therefore, a conservative estimate of the volume of the Sierra Madre Occidental ignimbrites is 393,000 km3. However, if we include other areas in Mexico with middle Tertiary ignimbrites, such as the outcrops of eastern Chihuahua or those to the south of the Mexican volcanic belt, the volume estimate increases to 587,000 km3 (Table 1). By comparing this volume with that of the well-known middle Tertiary ignimbrites of the San Juan Mountains of Colorado (20,000 km3; Lipman et al., 1970) and following the proportional approach of Swanson and McDowell (1984), we can deduce that the number of calderas responsible for the Sierra Madre Occidental ignimbrites should be 334 if we only consider the main area (Table 1). Even if we assume that the thickness of the main ignimbrite package is 1 km, e.g., 0.75 km (as occurs in the more eroded margins of the sierra), the number of calderas required is still 250 for the main area. Assuming common San Juan volcanic eld caldera sizes (2030 km) and a 1000 km3 ignimbrite for each caldera, the number of calderas required would be 393. Either the conservative 250 or the larger 393, the number of calderas needed is impressive, especially considering that, at most, 15 calderas have been recognized, some of which are still questionable. Therefore, at least 235 calderas have not yet been found; but 235 large calderas would be difcult to not see, either in satellite images or in the mapped transects across the Sierra Madre Occidental. Do all these calderas really exist? SOUTHERN BASIN AND RANGE AND THE IGNIMBRITE FLARE-UP The southern part of the Basin and Range province is in Mexico (Henry and Aranda-Go mez, 1992), and Basin and Range faulting af1GSA Data Repository item 2003112, Table DR-1, representative geologic studies in the Sierra Madre Occidental, and Figure DR-1, geologic map and photos, is available online at www.geosociety.org/pubs/ft2003.htm, or on request from editing@geosociety.org or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301-9140, USA.

ABSTRACT The Sierra Madre Occidental is mostly composed of middle Tertiary large-volume ignimbrites. From the United StatesMexico border (31N), the Sierra Madre Occidental extends southward to its intersection with the Mexican volcanic belt (21N). Ignimbrites of equivalent age extend into southern Mexico as discontinuous outcrops. Considering the average thickness of 1000 m for these ignimbrites based on representative measured sections, a conservative estimate of their total volume is 393,000 km3. Fewer than 15 calderas have been identied in this province, and the source of most of these ignimbrites has been an unsolved problem. We present geologic evidence indicating that ssures, most of them with the regional trend of Basin and Range faults, served as conduits for the ignimbrites. These ssures can be several kilometers long and are represented by pyroclastic (ignimbrite) dikes, rhyolitic lava dikes, linearly aligned lava domes, and elongated coignimbrite lithic-lag breccias adjacent to Basin and Range faults. Considering that the Basin and Range extension overlapped in time and space with the ignimbrite are-up, we propose a model in which batholith-sized magma chambers reached shallow crustal levels and erupted their contents when they reached Basin and Range normal faults. The faults acted as vents and caused fast decompression when the system was opened, and large volumes of silicic magmas were explosively erupted. Finally, devolatilized rhyolitic magmas were emplaced as domes or dikes. We propose the term ssure ignimbrites for ignimbrites formed in this way. Keywords: ignimbrite, Basin and Range, pyroclastic dikes, Sierra Madre Occidental, Mexico. INTRODUCTION The Sierra Madre Occidental (SMO) is the largest continuous ignimbrite province in the world (Swanson and McDowell, 1984). In order to visualize its size, consider that the Sierra Madre Occidental would cover the equivalent area of peninsular Italy (Fig. 1). The Sierra Madre Occidental is at least 1200 km long and 200500 km wide, extending continuously from the United StatesMexico border (31N) to its intersection with the younger Mexican volcanic belt (21N) (Fig. 1). However, middle Tertiary large-volume ignimbrites continue as discontinuous outcrops into the southwestern United States, as well as into southern Mexico and eastern Chihuahua. Swanson and McDowell (1984) estimated that at least 350 calderas the size of those of the San Juan volcanic eld in Colorado would have been needed to produce these ignimbrites. However, fewer than 15 calderas have been identied in the Sierra Madre Occidental by studies beginning in about 1975. The scarcity of recognized calderas in the Sierra Madre Occidental could be due to several causes, e.g., the relatively poor knowledge of this province, implying that there are more calderas to be recognized, or burial of calderas by younger ignimbrite sheets. How*E-mail: Aguirre-D azger@geociencias.unam.mx; Labarthe-Herna ndez labarthe@uaslp.mx.

2003 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editing@geosociety.org. Geology; September 2003; v. 31; no. 9; p. 773776; 3 gures; 1 table; Data Repository item 2003112.

773

TABLE 1. VOLUME OF MIDDLE TERTIARY IGNIMBRITES IN MEXICO Province Sierra Madre Occidental main Eastern Chihuahua Sonora South of Mexican volcanic belt Baja California Inferred beneath the Mexican volcanic belt Total Total of San Juan Mountains* Area (km2) 392,775 66,583 7982 35,287 36,421 32,136 586,727 Thickness (km) 1 1 1 1 1 1 km km km km km km Volume (km3) 392,775 66,583 7982 35,287 36,421 32,136 586,727 20,000

Note: volume is approximated and related to 17 calderas. *Data from Lipman et al. (1970).

Figure 1. Index map of middle Tertiary ignimbrite outcrops in Mexico and southwestern United States and of Basin and Range structural province. Note that middle Tertiary ignimbrites and Basin and Range province overlap in space (and in time, see text). Also shown for comparison is outline of peninsular Italy at same scale. White circles represent mapped or proposed calderas: 1San Carlos-Palomas, 2 Pastorias, 3Nuevo caldera, 4Calera-Del Nido, 5Corralitos, 6TomochicLas Varas, 7Ocampo, 8Chuchupate (Ye cora), 9Chupaderos, 10El Salto, 11Temoaya, 12 Mesa del Nayar, 13Juchipila, 14Jalpan, 15El Cubo. For references for these calderas see Table DR1 (see footnote 1). Black circles represent towns: HHermosillo, CChihuahua, NNazas, DDurango, MazMazatla n, ZZacatecas, SLP San Luis Potos , TTepic. White bars represent tectonic depressions mentioned in text: AAtengo graben, BBolan os graben, JJuchipila graben, CCalvillo graben.

which can be interpreted as the continuation to the southwest of the ignimbrite are-up. Thus, the ignimbrite are-up can be dened as a period of intense explosive volcanic activity that produced enormous volumes of silicic ignimbrite sheets, which took place mainly between 38 and 23 Ma in Mexico. The ignimbrite are-up coincided in time with peaks in Basin and Range faulting (Aguirre-D az and McDowell, 1991, 1993; Aranda-Go mez et al., 2000), and the ignimbrite activity apparently migrated from the northeast to the southwest, i.e., from central Chihuahua (3827 Ma) to Durango-Tayoltita-Nazas (3229 Ma) to Zacatecas-Tepic (2423 Ma), nishing by 16 Ma at Jalisco-Nayarit, as deduced from the compilation of geologic works listed in Table DR1 (see footnote 1). It is unknown yet whether there was a southward migration of Basin and Range faulting and if the ignimbrite are-up occurred episodically as peaks (3827 Ma, 3229 Ma, and 2423 Ma) or was continuous. Nevertheless, by the time that the ignimbrite areup started, the Basin and Range extension was already active in Mexico. Therefore, it can be concluded that the emplacement of the Sierra Madre Occidental ignimbrites and the Basin and Range extensional regime coincided in time and space, and this coincidence occurred in different times and different places throughout the Sierra Madre Occidental volcanic province. LARGE-VOLUME FISSURE-FED IGNIMBRITES OF THE SIERRA MADRE OCCIDENTAL Field evidence in the Sierra Madre Occidental indicates that many large-volume ignimbrites were derived from linear ssures rather than from typical circular or oval calderas. These include (1) large pyroclastic (ignimbrite or tuff) dikes, (2) coignimbrite lithic-lag breccias (CILBs) next to large faults or grabens, (3) aligned postignimbrite rhyolitic domes and lava dikes following a fault trend, and (4) absence of calderas. We provide a few examples of these types of evidence in the following sections. Pyroclastic Dike Pyroclastic (ignimbrite) dikes have been observed in several places in the Sierra Madre Occidental, particularly in the southern part, which is the best studied by us. Two cases are presented as examples, the 50-m-wide dike that crops out along the western shoulder of the Juchipila graben (Fig. 1), and the dikes found along the Juachin fault next to San Luis Potos (Fig. 1). 1. The dike at Juchipila is composite and includes four ignimbrites (Fig. 2), each with distinct welding characteristics and crystal-lithicpumice (amme) contents. The dike was emplaced along a major normal fault that formed the Juchipila graben, a north-northeasttrending, 110-km-long, 1520-km-wide structure. Fiamme and cooling joints are aligned subvertically with about the same dip as that of the dike. The ignimbrites of the dike show shearing and cooking along their margins (Fig. 2). The ignimbrites cooked both the wall rock and the earlieremplaced ignimbrite dikes. Shearing was formed mainly during ignimbrites intrusion, but postemplacement fault-related shearing also took
GEOLOGY, September 2003

fected most of northwestern Mexico. This extensional province overlaps in space and time with the Sierra Madre Occidental (Fig. 1), which therefore is highly dissected by northwest- to northeast-trending normal faults that bound many large grabens and half-grabens. They form particularly long and deep depressions in the southern Sierra Madre Occidental, such as the north-trending Bolan os graben (1800 m deep, 20 25 km wide, and 150 km long) (Fig. 1). Basin and Range faulting occurred between at least 32 Ma and 12 Ma (Aguirre-D az and McDowell, 1993; Aranda-Go mez et al., 2000), with both limits probably extending until the Eocene and the Quaternary (Aranda-Go mez et al., 2000). Ignimbrite activity can be as old as 51 Ma (Aguirre-D az and McDowell, 1991) and as young as 1716 Ma (Nieto-Obrego n et al., 1981), but most of the ignimbrite volume was erupted in the 3828 Ma period (McDowell et al., 1990; Aguirre-D az and McDowell, 1991; Table DR-1 [see footnote 1]), which has been referred to as the ignimbrite are-up (McDowell et al., 1990; Aguirre-D az and McDowell, 1991). However, intense ignimbrite activity is evident until 23 Ma in the south-central and southwestern portions of the Sierra Madre Occidental (e.g., Scheubell et al., 1988; Table DR-1 [see footnote 1]),
774

Coignimbrite Lithic-Lag Breccia CILBs are the second-best indicators (after the pyroclastic dikes) of the location of the ssure-type vents of ignimbrites. In addition to the CILB examples next to the pyroclastic dikes at Juchipila and at the Juachin fault just described, at the Bolan os graben there is an excellent example for a vent location based on CILB. The Alacra n ignimbrite (Scheubel et al., 1988), dated as 23 Ma, is a voluminous unit to 400 m thick in the proximal facies. It is well exposed at both sides of the north-trending graben. The CILB of the Alacra n ignimbrite is found along the grabens shoulders; i.e., both the Alacra n ignimbrite and its corresponding CILB crop out to the west and to the east of the 25km-wide graben. The CILB gradually changes to an ignimbrite poorer in lithic fragments with distance from the grabens shoulders. Just beneath the Alacra n ignimbrites CILB there is an associated layered pyroclastic sequence. This sequence is mostly composed of surge beds and minor ignimbrites. Within the surge deposits, and 45 km from the actual graben scarps, ballistic blocks, to 40 cm in diameter, produced impact-sag deformations indicating their provenance from the Bolan os graben. These ballistic blocks also occur on both sides of the graben. Aligned Rhyolite Domes and Dike At several places in the Sierra Madre Occidental, rhyolitic lava domes are aligned with the same northwest to northeast trends as the regional Basin and Range normal faults. In some areas, such as at the Bolan os and Atengo grabens and the Juachin fault, it is clear that the lava domes were emplaced just after the large-volume ssure-fed ignimbrites and that they occur close or next to the coarser parts of the CILBs that crop out next to normal faults. At the Atengo graben (Fig. 1), a 12-m-wide rhyolitic dike connects with high-standing rhyolitic lava domes. Both dike and domes are aligned with the north trend of the tectonic depression. In some other places, such as in San Luis Potos , it is possible to see rhyolite dikes interconnecting aligned lava domes or lava domes with shapes of boudins aligned with regional normal faults of the Basin and Range, indicating strong structural control for the emplacement of the rhyolitic magmas. The lava domes and dikes are here interpreted as the devolatilized magma that erupted after the climactic explosive events that produced the major ignimbrites, as occurs in classic caldera cycles with the rim lava domes (Smith and Bailey, 1968), but in this case the domes were emplaced along Basin and Range faults. MODEL Figure 3 schematically depicts a model that summarizes our eld observations for large-volume ssure-fed ignimbrites of the Sierra Madre Occidental and their relationship with the regional extensional regime of the Basin and Range province in Mexico. In order to account for the large volume of ignimbrites in the Sierra Madre Occidental and considering the long grabens or fault zones to which they are related, a large magma chamber complex must be inferred. Thus, in our model we propose an elongated magma chamber or series of magma chambers with batholith-sized dimensions. Late Cretaceousearly Cenozoic representatives of these batholiths related to the lower volcano-plutonic complex are now exposed at the western margin of the Sierra Madre Occidental, e.g., the plutonic rocks in Sinaloa (Henry and Fredrikson, 1987) or at the more eroded northwest part of the Sierra Madre Occidental in Sonora (McDowell et al., 2001). These large magma chambers reached shallow crustal levels mainly because of the extensional regime of the Basin and Range in which they were emplaced; reaching the high crustal level caused magmas to become volatile rich by exsolution of gas bubbles due to decompression. Either because the roof of the chamber reached Basin and Range normal faults, or because the faults formed during a peak in intensity of extension and affected the magma chambers roof, the system was
775

Figure 2. Pyroclastic (ignimbrite) dike at Juchipila graben. Dike is 50 m wide and is composite. It includes four ignimbrites, each one with sheared and/or cooked margins. Ignimbrites are densely welded or unwelded. Welded ones contain amme, which are vertically aligned parallel to dike walls. Ignimbrites that form composite dike were emplaced sequentially through same ssure. Thus, different pyroclasticow pulses were erupted to produce equivalent at-layered units outside graben. Dike follows same trend of graben, N1020E.

place. Each ignimbrite of the dike corresponds to a at-lying ignimbrite outside the graben. The dike can be followed as discontinuous outcrops for at least 25 km. At places along the dike trend, a lithic-rich ignimbrite constitutes the dike, which changes laterally to a CILB. Webber et al. (1994) reported two calderas within the Juchipila graben, but although we looked for them, we did not nd any evidence for caldera structures. 2. The pyroclastic dikes at the Juachin fault (Fig. DR-1; [see footnote 1]) occur as small boudins along the northwest-trending normal fault zone of Juachin. The width of the boudins varies from 1 to 8 m; they are composed of a highly sheared pumiceous ignimbrite. Adjacent to the dikes consisting of the boudins and following the northwest trend of the Juachin fault zone is a major CILB that gradually changes laterally to a large-volume ignimbrite sheet, the 27 Ma Panalillo Tuff, which covers a widespread area of the San Luis Potos volcanic eld (Labarthe-Herna ndez et al., 1982). We have found pyroclastic dikes, some as wide as 100 m, at several other places in the Sierra Madre Occidental, generally next to large Basin and Range faults and grabens; e.g., at the Calvillo, Colotla n, and Atengo grabens (Fig. 1).
GEOLOGY, September 2003

A similar mechanism of ssure-vent ignimbrite emplacement was mentioned by Lipman (1997), who referred to the result as volcanotectonic depressions, and by Scheubel et al. (1988), who interpreted the ignimbrites of the Bolan os graben to have erupted through the grabens faults. Our model for ssure ignimbrites is similar to those proposed for calderas; the main differences are the batholithic size of the magma chamber(s) to account for the impressively large volume of the ignimbrite package of the Sierra Madre Occidental and the ssure-type vent (represented by the pyroclastic or rhyolite lava dikes) and/or vents (represented by the domes) aligned along a normal fault instead of forming an arcuate structure. We do not contend that all ignimbrites in the Sierra Madre Occidental were formed in this way. It is well established that calderas were the sources of several ignimbrites, but we think that the dominant volume of the Sierra Madre Occidental ignimbrites was erupted from ssure-type vents related to Basin and Range faulting episodes.
ACKNOWLEDGMENTS We are grateful to Grupo Me xico and the personnel in the Bolan os mine project and to Minera El Pilo n of San Mart n de Bolan os for logistical help. The manuscript was improved by reviews of Jim Luhr and Joann Stock. We thank Jorge Nieto-Obrego n, Ramo n Torres, and Margarito Trista n for their assistance during eld seasons. Financial support was obtained from Universidad Nacional Auto noma de Me xico grants IN-120999 and IN-115302 and Consejo Nacional de Ciencia y Tecnolog a grant 33084-T to AguirreD az.

REFERENCES CITED
Aguirre-D az, G.J., and McDowell, F.W., 1991, The volcanic section at Nazas, Durango, Mexico, and the possibility of widespread Eocene volcanism within the Sierra Madre Occidental: Journal of Geophysical Research, v. 96, p. 13,37313,388. Aguirre-D az, G.J., and McDowell, F.W., 1993, Nature and timing of faulting and synextensional magmatism in the southern Basin and Range, central-eastern Durango, Mexico: Geological Society of America Bulletin, v. 105, p. 14351444. Aranda-Go mez, J.J., Henry, C.D., and Luhr, J.F., 2000, Evolucio n tectonomagma tica postpaleoce nica de la Sierra Madre Occidental y de la porcio n meridional de la provincia tecto nica de Cuencas y Sierras, Mexico: Bolet n de la Sociedad Geolo gica Mexicana, v. 53, p. 5971. Henry, C.D., and Aranda-Go mez, J.J., 1992, The real southern Basin and Range: Midto late Cenozoic extension in Me xico: Geology, v. 20, p. 701704. Henry, C.D., and Fredrikson, G., 1987, Geology of part of southern Sinaloa, Mexico, adjacent to the Gulf of California: Geological Society of America Map and Chart Series no. MCH063, 14 p. Labarthe-Herna ndez, G., Trista n-Gonza lez, M., and Aranda-Go mez, J.J., 1982, Revisio n estratigra ca del Cenozoico de la parte central del estado de San Luis Potos : Instituto de Geolog a y Metalurg a de la Universidad de San Luis Potos , Me xico, Folleto Te cnico no. 85, 208 p. Lipman, P.W., 1997, Subsidence of ash-ow calderas: Relation to caldera size and magma-chamber geometry: Bulletin Volcanologique, v. 59, p. 198218. Lipman, P.W., Steven, T.A., and Mehnert, H.H., 1970, Volcanic history of the San Juan Mountains, Colorado, as indicated by potassium-argon dating: Geological Society of America Bulletin, v. 81, p. 23292352. McDowell, F.W., Wark, D.A., and Aguirre-D az, G.J., 1990, The Tertiary ignimbrite areup in western Mexico: Geological Society of America Abstracts with Programs, v. 22, no. 3, p. 366. McDowell, F.W., Rolda n-Quintana, J., and Connelly, J.N., 2001, Duration of Late Cretaceousearly Tertiary magmatism in east-central Sonora, Mexico: Geological Society of America Bulletin, v. 113, p. 521531. Nieto-Obrego n, J., Delgado-Argote, L.A., and Damon, P.E., 1981, Relaciones petrolo gicas y geocronolo gicas del magmatismo de la Sierra Madre Occidental y el Eje Neovolca nico en Nayarit, Jalisco y Zacatecas: Asociacio n de Ingenieros de Minas Metalurgistas y Geo logos de Me xico, Memoria, XIV Convencio n Nacional, p. 327361. Scheubel, F.R., Clark, K.F., and Porter, E.W., 1988, Geology, tectonic environment, and structural controls in the San Mart n de Bolan os district, Jalisco, Mexico: Economic Geology, v. 83, p. 17031720. Smith, R.L., and Bailey, R.A., 1968, Resurgent cauldrons, in Coats, R.R., et al., eds., Studies in volcanology: Geological Society of America Memoir 116, p. 613622. Swanson, E.R., and McDowell, F.W., 1984, Calderas of the Sierra Madre Occidental volcanic eld western Mexico: Journal of Geophysical Research, v. 89, p. 87878799. Webber, K.L., Ferna ndez, L.A., and Simmons, W.B., 1994, Geochemistry and mineralogy of the EoceneOligocene volcanic sequence, southern SMO, Juchipila, Zacatecas, Mexico: Geof sica Internacional, v. 33, p. 7789. Manuscript received 19 March 2003 Revised manuscript received 29 May 2003 Manuscript accepted 30 May 2003 Printed in USA

Figure 3. Schematic model explaining mechanism for eruption of ignimbrite-forming pyroclastic ows through major faults of Basin and Range province. A: Batholith-sized magma chamber reached shallow crustal levels, and magmas became volatile rich by exsolution due to decompression. B: Roof of magma chamber reached Basin and Range normal fault and opened magmatic system, resulting in fast decompression and explosive eruption of voluminous pyroclastic ows. C: Lithic load of pyroclastic ows was deposited near fault-related vent to form coignimbrite lithiclag breccia. D: After climactic eruption that formed ignimbrite and coignimbrite lithic-lag breccia, magmas remaining in chamber were volatile poor and erupted as rhyolitic lava domes and/or dikes that closed vent. These lava domes are aligned with trend of fault-related vent.

opened and caused fast decompression and the explosive eruptions of voluminous pyroclastic ows and ssure ignimbrites. The lithic load of the pyroclastic ows was deposited near the fault-related vents, forming the observed CILBs adjacent to a graben or to major normal faults. After the climactic eruption that formed the ignimbrites, the magmas remaining in the chamber were volatile depleted and erupted as rhyolitic lava domes and/or dikes that closed the vents. These lava domes and dikes are aligned with the same trend as the fault-related vents.
776

GEOLOGY, September 2003