Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
TECTONICS, VOL. 21, NO. 4, 10.1029/2001TC001302, 2002 Ignimbrite flare-up and deformation in the southern Sierra Madre Occidental, western Mexico: Implications for the late subduction history of the Farallon plate Luca Ferrari Centro de Geociencias, Universidad Nacional Autónoma de México, Campus Juriquilla, Queretaro, Mexico Margarita López-Martı́nez Departamento de Geologı́a, Centro de Investigación Cientı́fica y Educación Superior de Ensenada, Ensenada, Mexico José Rosas-Elguera Centro de Ciencias de la Tierra, Universidad de Guadalajara, Guadalajara, México Received 17 May 2001; revised 10 February 2002; accepted 27 February 2002; published 24 August 2002. [1] The Sierra Madre Occidental (SMO) of western Mexico is one of the largest silicic volcanic provinces on Earth, but the mechanism for the generation of such a large volume of ignimbrites has never been clearly defined. We present new 40Ar/39Ar ages, geologic mapping, and structural data for the southern part of the SMO demonstrating that most of this volcanic province was built in two episodes of ignimbrite flare-up in Oligocene (31.5–28 Ma) and early Miocene (23.5 – 20 Ma) time, and that extensional deformation occurred mostly before the transfer of Baja California to the Pacific plate. Extensive ignimbrite successions, with 40Ar/39Ar ages clustering at 23 and 21 Ma, cover most of the southern SMO, thus correlating in age with ignimbrites exposed in southern Baja California and central Mexico. Grabens with a 020 to N-S orientation developed in the east almost concurrently with this volcanic episode. Half grabens and NNW striking listric normal fault systems formed at the end of middle Miocene as far as 150 km from the present coast. A belt of left-lateral transpressional structures formed along the southern boundary of the SMO during the same period. We link these magmatic and tectonic events to the evolution and dynamics of the Farallon and North America plates during the Miocene. Particularly, we propose that a first detachment of the lower part of the Farallon plate in early Miocene time produced a transient thermal event and partial melting of the crust via mafic underplating. Middle Miocene extension would be related to a second detachment event, resulting from the slowing subduction that preceded the final capture of the Magdalena microplate by the Pacific plate at 12.5 Ma. Transpression at the southernmost end of the SMO Copyright 2002 by the American Geophysical Union. 0278-7407/02/2001TC001302$12.00 occurred along the inland projection of the MagdalenaCocos plate boundary and may be explained by a difference in subduction rate and by a temporal convergence between the two plates in the eve of the INDEX end of subduction of the Magdalena plate. TERMS: 5480 Planetology: Solid Surface Planets: Volcanism (8450); 8150 Tectonophysics: Evolution of the Earth: Plate boundary—general (3040); 8109 Tectonophysics: Continental tectonics—extensional (0905); KEYWORDS: ignimbrite flare-up, Sierra Madre Occidental, western Mexico, extensional tectonics, slab detachment 1. Introduction 1.1. Purpose and Objectives [2] The huge silicic volcanic plateau of the Sierra Madre Occidental of western Mexico (SMO) is an enigmatic feature in the geology of the North America plate. The SMO runs for over 2000 km from the U.S.-Mexico border to the Trans-Mexican Volcanic Belt (TMVB) (Figure 1). It mostly consists of silicic ignimbrites and, to a lesser extent, rhyolitic domes that cover about 300,000 km2 with an average thickness of 1 km [McDowell and Keizer, 1977; McDowell and Clabaugh, 1979]. With an estimated volume of 300,000 km3 the SMO ranks among the major silicic large igneous provinces on Earth. These kind of provinces are generally related to continental breakup with variable interaction of mantle plumes [e.g., Bryant et al., 2000; Pankhurst et al., 2000], which provide sufficient thermal energy to thin the continental lithosphere and, eventually, to melt the crust. By contrast, in the SMO the production of massive silicic volcanism occurred during subduction of a young plate with no involvement of mantle plumes and took place 20 to 10 Ma before the breaking of the continental crust which led to the formation of the Gulf of California. The inception of silicic volcanism in the SMO has been generally considered as a marker of the beginning of extension following the Laramide orogeny [e.g., Wark et al., 1990; Luhr et al., 2001]. A plate tectonics mechanism 17 - 1 17 - 2 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Figure 1. Geodynamic map of Mexico showing Tertiary extension and volcanism north of the TransMexican Volcanic Belt (TMVB) and the present configuration of plates. Tertiary extension is from Henry and Aranda-Gomez [2000] and this work. Regional tilt domains are from Stewart et al. [1998]. Abbreviations are as follows: Nay., Nayarit; Jal., Jalisco. for the occurrence of massive ignimbritic volcanism, however, has not been provided. [3] The SMO has been also strongly affected by extensional tectonics during the Tertiary. Extensional structures form two broad, NNW-SSE trending, belts on both sides of the SMO [Stewart et al., 1998], which merge in Sonora and Chihuahua, to the north, and in Nayarit-Jalisco, to the south (Figure 1). The western belt borders the Gulf of California and is also known as the Gulf Extensional Province [Gastil et al., 1975]; the eastern belt extends for several hundreds of kilometers east of the core of the SMO and has been considered the southern continuation of the Basin and Range province of the western United States [Henry and Aranda-Gomez, 1992]. Henry and Aranda-Gomez [2000] further suggested that most of the Mexican Basin and Range was linked to the interaction between the Pacific and the North America plates during late Miocene time, after the end of the subduction of the last remnant of the Farallon plate (12.5 Ma) [Atwater and Stock, 1998]. Several studies, however, indicate that extension initiated as early as late Oligocene time both in the north [Nourse et al., 1994; Gans, 1997; McDowell et al., 1997; Stewart et al., 1998] and in the south [Nieto-Samaniego et al., 1999; Luhr et al., 2001]. Therefore this first part of the extensional history remains to be explained. In summary, despite a number of studies published in the last two decades the causes of the ignimbrite flare-up and extension in the SMO are still not completely understood. [4] This paper presents new geologic, geochronologic, and structural data that define the timing of silicic volcanism and extension as well as the tectonic setting of the SMO south of the Tropic of Cancer (latitude 23300). This vast region, which was previously very poorly known, holds an important piece of information for the geologic reconstruction of the whole province since it faces the last and larger fragment of the Farallon plate (Magdalena microplate) to be subducted before the initial rifting of the Baja California peninsula [Lonsdale, 1991]. Our data fill the last gap in the geologic reconnaissance of the SMO and enable us to analyze at a global scale the temporal pattern of ignimbrite FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO 17 - 3 Figure 2. Regional tectonic map of the southwestern Sierra Madre Occidental with locations of Figures 4 and 5. Shaded area is the late Miocene to Quaternary Trans-Mexican Volcanic Belt. Abbreviations of tectonic structures are as follows: Co, Concordia fault; SP, San Pedro fault system; JM, Jesus Maria fault system; SA, San Agustin graben; Ve, La Ventana graben; At, Atengo half graben; Po, Pochotitán fault system; Hu, Huajimic half graben; Pc, Puente de Camotlán half graben; SMSR, Santa Maria del Oro – Santa Rosa transpressional corridor. flare-up and extension. On the basis of this rationale, we propose a possible mechanism for the occurrence of these phenomena with implications for the dynamics of the Farallon slab during its final stage of subduction. 1.2. Location and Methodology [5] South of the Tropic of Cancer the Sierra Madre Occidental volcanic province can be divided into the physiographic provinces of the Mesa Central high plateau to the east and the Sierra Madre Occidental proper to the west [Nieto-Samaniego et al., 1999]. The Sierra Madre Occidental can be further divided into three domains: (1) an eastern domain affected by several NNE to N-S trending grabens, (2) a western domain where NNW trending half grabens dominate the landscape, and (3) a southern domain characterized by left-lateral transpressional structures (Figure 2). Our fieldwork mostly focused on the latter two domains, bounded by the Guadalajara-Fresnillo highway to the east, the Gulf of California to the west, the Mezquital River to the north, and the Trans-Mexican Volcanic Belt to the south (Figure 2). [6] Only sparse geologic and structural works were available in this region, mainly along its borders [Damon et al., 1979; Clark et al., 1981; Nieto-Obregon et al., 1981; Lyons, 1988; Scheubel et al., 1988; Ferrari, 1995]. The paucity of studies was mostly due to the difficult access to 17 - 4 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Figure 3. Composite stratigraphic columns for the study region indicating typical thickness of units and existing ages. New ages presented in this work are in bold. Full references of other ages are given in the text. this region. The volcanic plateau, with elevations ranging from 2100 to 2900 m, is dissected by valleys with elevations as low as 500 m that cut along the main extensional structures. No paved roads exist, and only two graded roads were recently completed to traverse this part of the SMO in a WSW-ENE direction (Figure 2). To make fieldwork more efficient in this complicated area we have first studied 1:250,000 satellite images, 1:80,000, and 1:50,000 air photos, and digital terrain models at various scales to compile a geologic base map. This was subsequently integrated with fieldwork mainly along two transects (Valparaiso-Estación Ruiz; Bolaños-Tepic) and eventually completed with a few helicopter flights. The resulting volcanic stratigraphy and geologic mapping are illustrated in Figures 3, 4, and 5, and will be described in the next section. Figure 4. Geologic map of the Valparaiso –Estación Ruiz transect. Question marks indicate contacts between units inferred from air photos only or unknown extension of units. Curved dashed lines indicate calderas inferred by aerial photos and topographic features. FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO 17 - 5 Santa Maria Ocotlán 15 20 ? 28 15 San 30 Agustin 2000 10 2300 8 400 10 2200 22° 30' Santa Teresa 1300 2200 2000 Rio San Pedro o dr m Pe te n ys Sa ult s fa 19.9 20.0 1100 30 Santa Fe 500 17.0 El Zopilote Estación Ruiz 105° 00' 11.9 ? ? ? ? ? ? ? 2400 ? ? ? ? 2500 ? Mexquitic ? 1350 1800 ? ? 21.1 ? 2000 2100 ? 1700 ? ? ? 22° 00' ? 0 104° 30' Ferrari et al. - Figure 4 25 km 2500 Monte Escobedo 104° 00’ River 21.1 Ar/Ar date Sierra Los Huicholes 1900 2400 2300 2200 1000 22° 00' 50 ntiago 2100 28 50 ? 1600 ? 2500 700 15 2000 ? Peyotán Jesus Maria Rio Sa 32 San Juan Bautista ? ? 21.3 23.5 Mesa del Nayar 1900 Bolaños graben ? 2600 15 21.2 20.9 2100 2700 Santa Lucia 30 500 100 Huejuquilla 27.9 20 S. Juan 30 500 35 28 ? 12 1900 2000 28.6 31.0 ? 25 25 26 ? 25 18 28 25 21.0 1400 ? ? 1100 ? 13 2000 ? ? 23.5 Jesus Maria half graben 2100 2400 Valparaiso 1850 28 2450 2400 1100 Francisco Madero Huazamota 35 ? 26 20 32 28 3000 ? 18 20 31.5 Las Canoas ? 1700 ? Sierra de Valparaiso ? a 1800 ? ? s su Je o ria Ri Ma ? 20 30 25 20 8 ? 15 103° 45’ ? ? ntan zquital 15 1200 2700 2500 a Ve 10 ? 20 ? 2450 Rio L 500 21.1 ? 22° 45’ ? 1500 Rio Me ? Llano Grande Ameca La Vieja Atengo half graben ng o 800 1900 La Ventana graben Rio Ate San Agustín graben 1700 25 (this work) Paved road 23.5 Published Unpaved road K/Ar date Extensional fault 2300 Elevation (m) Caldera wall Dip of the ignimbrites Horizontal strata Extensional anticline Flexure Miocene subvolcanic bodies Conglomerate and sandstone Miocene basalt and andesite Miocene rhyolitic domes Nayar ignimbrite succession (21 Ma) Oligocene basalt and andesite Pre-Miocene rhyolitic domes Canoas ignimbrite succession (23 Ma) Atengo ignimbrite succession (28 Ma) >28 Ma ignimbrites Figure 5. Geologic map of the Bolaños-Tepic transect. Symbols as in Figure 4. 17 - 6 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO ? 105° 00' 104° 45’ El Zopilote ? ? 104° 30’ 2080 500 ? El Venado C C . La e le t bol Hu Rio 11.9 a Huaynamota c 104° 15’ mi aji ? ? 1840 ? 1920 14.0 2111 1000 10 Huajimic 19.9 2720 21 23.2 amot lán Bolaños 2320 22.2 900 2680 Rio C 2200 Sierra de Pajaritos 18.7 22.4 Sierra Alic a nde d ic Rio Huajim n tá m ti e o st ch sy o t P ul fa s ic an lc a) vo 0 M VB 1 TM (1 Rio Gra 2640 1700 Aguamilpa ago e Santi ? ? Tuxpan de Bolaños Guadalupe Ocotlán 11.5 2400 2040 500 400 ? 104° 00’ 900 1000 1000 2240 Puente de Camotlán 21°45’ San Martin de Bolaños 23.7 2300 19.4 ? 35 River 40 25 El Anticline 21.1 Dip of the ignimbrites 25 20 20 20 26 15 10.9 New age (this work) 19 25 15 30 ? 1100 11.4 19 - 21 Ma ignimbrite succession 23 Ma ignimbrite succession 1640 ? ? 2740 2500 Dated mafic dike Miocene to Quaternary volcano-sedimentary deposits Oligocene andesites and ignimbrites 1500 ? ? ? Scale Santa Fe 25 km 21°15’ 1600 Miocene subvolcanic bodies ? 0 45 ? 50 0 Miocene rhyolitic domes ? 2500 te Santa Maria del Oro Early-middle Miocene basalts 21°30’ ? ? ? be Published K/Ar age be 10 20 20 20 23.5 ? ? 2300 5 40 25 15 25 21.3 gra La Manga 40 15 a Pin El ra Syncline 25 10 r Sie Extensional fault 700 2140 5 45 17.2 15 1960 45 ? Bo 25 35 b Ro s le a he ños TMVB volcanics (11 - 0 Ma) TEPIC r z 10 VOLCAN LAS NAVAJAS e on lañ 500 Bo la Pochotitán Unpaved road 20.7 35 25 Paved road n 21.3 os (m) Rio 2300 Elevation 1700 TMVB volcanics (11 - 0 Ma) 500 45 20 Ferrari et al. - Figure 5 ntiago Rio Sa ? ? TMVB volcanics (11 - 0 Ma) 103°45’ 30.1 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO [7] Our geologic reconnaissance is supported by 17 new Ar/39Ar ages (Table 1), 2 new K-Ar ages (Table 1), and 16 published K-Ar data (details in the following section). The 40 Ar/39Ar ages were obtained at Centro de Investigación Cientı́fica y Educación Superior de Ensenada’s (CICESE) geochronology laboratory using a MS-10 mass spectrometer; details on the methodology are given in Appendix 1, which is available as electronic supporting material.1 The samples were step-heated between 700 and 1500C. A summary of the 40Ar/39Ar is given in Table 1; the plateau and isochron ages calculated are in Appendix A. In general, we obtained good agreement between the plateau and isochron ages. On those samples where duplicate analyses were performed we obtained consistent results; the isochron age calculated for the combined fractions of the duplicate experiments is also reported in Table 1. The majority of the samples yielded well-defined isochron ages which we take as the best estimate of the age for the sample; however, we had cases where the distribution of the data on the correlation diagram did not permit a realistic calculation of the regression line or only the intercept with one of the axes could unequivocally be calculated. This is the case for samples TS-10, TS-21, and TS-46, for which we took the plateau age as our best estimate of the age. For sample TS28 we analyzed concentrates of plagioclase and biotite, and because of the scarcity of these minerals, only a few milligrams could be obtained. Two experiments were conducted on the biotite with one and two fractions collected. The four fractions collected on the plagioclase yielded a saddle-shaped age spectrum. Our best estimate for the age of sample TS-28 is taken from the isochron age calculated with the combined fractions of the biotite and plagioclase. [8] Scarce access and dense vegetation prevented a detailed microtectonic study of the fault systems observed in the region. However, the geometry and sense of slip of over 200 striated faults and 57 dikes could be measured at 18 sites. The local paleostress regime responsible for the observed deformation was computed by fault slip data inversion with the method of Angelier [1990], and the relative results are listed in Table 1. Using different approaches, Pollard et al. [1993], Cashman and Ellis [1994], Nieto-Samaniego and Alaniz-Alvarez [1997], and Twiss and Unruh [1998] demonstrated that the assumptions of the stress inversion methods are not always satisfied and that caution should be used in using them. Particularly simplistic assumptions often not fulfilled in nature are (1) a homogeneous stress field, (2) no dynamic and kinematic interaction between faults, and (3) the parallelism between the maximum resolved shear stress on a fault plane and the slickenline. To minimize this problem, we inverted slip data of faults affecting previously unfaulted rock units and/or the fault planes cutting all the other ones, i.e., those with the higher probability of complying with the assumptions of 40 1 Supporting Appendix A is available via Web browser or via Anonymous FTP from ftp://kosmos.agu.org, directory ‘‘append’’ (Username = "anonymous," Password = "guest"); subdirectories in the ftp site are arranged by paper number. Information on searching and submitting electronic supplements is found at http://www.agu.org/pubs/ esupp_about.html. 17 - 7 the inversion methods. Some faults with incongruous orientations or striations relative to the dominant population appeared in the data set, but they were discarded in the final computation. These results should then be representative of the local paleostress conditions. 2. Geology and Geochronology [9] In this section we present an overview of the geology of the study region on the basis of information gathered mostly along two ENE-WSW transects, 70 to 100 km apart, that cross the southwestern SMO (Figures 2, 4, and 5). Reconnaissance field geologic mapping was often extended for 25 to 50 km to both sides of the transect when dirt roads or tracks were available. Composite stratigraphic sections were elaborated for the four areas with more information and are illustrated in Figure 3. 2.1. Oligocene Volcanism [10] The oldest rocks observed are exposed in the northeastern part of the study region and, possibly, in the west along the coast (Figure 4). In the northeast they are silicic ignimbrites and rhyolitic domes that extend widely to the east outside the study region at elevations ranging between 2100 and 2400 m. In the Sierra de Valparaiso the ignimbrites attain at least 300 m of thickness. We obtained an age of 31.5 ± 0.3 Ma for a sanidine concentrate of a welded ash flow tuff also rich in quartz and plagioclase widely exposed in the central part of Sierra de Valparaiso (sample TS 56, Table 1). Similar ignimbrites are widespread toward the south, in the Huejuquilla area (Figures 3 and 4). We have dated two ignimbrites exposed in the footwall of the Atengo half graben (Figure 4). A feldspar separate yielded an age of 31.0 ± 0.7 Ma (sample TS-5, Table 1), indistinguishable from the Sierra de Valparaiso one. A sanidine concentrate from an ignimbrite at an upper stratigraphic position produced an age of 28.6 ± 0.3 Ma (sample TS-10, Table 1). In the western part of the study region, massive andesitic flows constitute the base of the succession east of Estación Ruiz (Figures 3 and 4). These rocks proved to be too altered to be dated. Farther to the south, in the Santa Maria del Oro area and along the Rio Santiago valley, a succession of ignimbrites and subordinated andesitic lavas underlie the Miocene ignimbrite succession. We have dated a biotite separate from a welded ash flow tuff at the southernmost edge of the SMO, 50 km northwest of Guadalajara, at 30.1 ± 0.8 Ma by K-Ar method (sample ES 1, Table 1; Figure 5). [11] Andesites and Oligocene ignimbrites are locally covered by volcanic conglomerate and/or red sandstone. Outcrops of volcano-sedimentary beds are normally too small to appear on the geologic maps of Figures 3 and 4 but were observed in the Huejuquilla area (Figure 4), 15 km NW of El Zopilote (Figure 4), in the Santa Maria del Oro and Santa Fe areas (Figure 5), and in the Rio Santiago valley southeast of these towns (Figure 5). Small plateaus of basaltic to andesitic flows also cover the late Oligocene ignimbrites in the northwestern part of the study region (Figure 4). These, in turn, are capped by a distinct pyroclastic succession of poorly to mildly welded and crystal Latitude, North Elevation, m Rock Type Material Dated 10425.7000 10450.8920 Mexpan, Nay. Arroyo El Naranjo, Nay. Rio Chico, Jal. Cotija area, Mich. Rock Type ignimbrite rhyolite ignimbrite gms feld bio bio gms feld san pl w.r. pl san feld san san K-Ar Ages Obtained at Geochron Laboratory, Cambridge, Massachussets 2110.4900 ignimbrite biotite 10346.2520 1953.7050 ignimbrite feldspar 10302.4760 Elevation, m 1050 460 1050 ignimbrite granodiorite ignimbrite ignimbrite bas.-and. ignimbrite ignimbrite ignimbrite ignimbrite ignimbrite san ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.5 0.3 0.7 0.3 0.4 0.3 0.2 0.5 0.4 0.3 0.4 0.3 0.3 0.7 0.8 5.967 0.946 Average K, wt % 19.9 ± 0.2 – 19.0 ± 0.6 18.3 ± 0.4 20.7 ± 0.2 19.9 ± 0.3 19.9 ± 0.2 – 21 ± 4 21 ± 1 23.3 21.3 21.1 21.1 21.1 21.1 21.0 20.4 19.7 31.6 31.1 28.6 27.9 27.7 28.1 tp Ma 0.01258 0.001301 Average 40Ar*, ppm 31.5 ± 0.3 31.0 ± 0.7 – 27.9 ± 0.2 27.8 ± 0.8 27.6 ± 0.8 28 ± 2b 23.5 ± 0.4 – – 21.1 ± 0.3 20.9 ± 0.4 21.2 ± 0.3 – 20.0 ± 0.6 19.9 ± 0.4 19.9 ± 0.4b 17 ± 9 22 ± 1 20 ± 2 20 ± 2b 20.6 ± 0.2 20.1 ± 0.6 19.8 ± 0.4 20.0 ± 0.3 20.0 ± 0.3 19 ± 2 – – 17 ± 1b tc Ma 18 44 12 11 15 64 48 41 24 5 8 7 17 7 10 5 5 2 ±2 ± ± ± ± ± ± ± ± ± ± ± ± ± ± 21 ± 14 ±8 ± ± ± ± ± ± 27 ± 21 Ar*, wt % 42.5 19.5 40 306 291 – 324 276 318 319 289 – – 294 302 294 – 318 283 308 304 296 298 298 314 295 307 299 299 313 – – 315 (40Ar/36Ar)i 30.1 ± 0.8 23.5 ± 0.9 Age, Ma 7.6/4 – – 15.5/7 1.2/3 3.4/4 – 0.1/4 0.4/3 2.6/4 4.0/7 1.1/4 – – 2.7/3 2.5/4 8.0/4 – 0.5/3 0.2/4 1.7/7 81/4 1.8/4 6.8/4 109/12 1.2/4 10.6/4 17.1/5 34.3/9 SumS/n All errors are 1s. The abbreviations are exp, experiment(s); tp, plateau age with the error in J included; tc, isochron age calculated from the 36Ar/40Ar versus 39Ar/40Ar correlation diagram; SumS of York [1969] n, number of points fitted; San, sanidine; bio, biotite; feld, feldspar; gms, groundmass; pl, plagioclase; w.r., whole rock. Preferred ages are shown in bold. Detailed results are given in Appendix A, which is available as electronic supporting material. b Isochron age calculated with the combined fractions of more than one experiment. a ES 2 CO 1 2101.8000 2201.9090 10425.7000 Mexpan, Nay. 1710 100 1550 650 550 1780 1120 870 2050 2100 ignimbrite Material Dated 2101.8000 10412.2920 Sierra Los Pajaritos, Nay. Latitude, North 2135.0610 10505.6710 San Juan Bautista, Nay. Longitude, West 2208.6100 10444.6970 SW Mesa del Nayar, Nay. Location 2209.6450 10428.0910 10430.6670 10504.5400 10443.5660 10506.5630 10410.6910 10445.3330 east of Jesus Marı́a, Nay. Jesus Maria, Nay. west of Llano grande, Nay. Arroyo El Frayle, Nay. north San Juan Bautista, Nay. south of Santa Lucia, Nay. west of Mesa del Nayar, Nay. 1330 Sample 2217.4450 2215.1450 2245.6000 2210.3620 2211.0960 2205.6300 2220.6300 10359.195 Rio Atengo, Zac. 2240.089 0 feld+bio 40 Ar/ 39Ar Ages Obtained at CICESE Laboratory, Ensenada, Baja California 2253.2250 2450 ignimbrite san 10343.3050 2239.5540 1480 ignimbrite feld 10359.2510 2240.0890 1330 ignimbrite san 10359.1950 2237.5890 2445 rhyolite san 10413.1030 Longitude, West 0 Sierra Valparaiso, Zac. west of Huejuquilla, Jal. west of Huejuquilla, Jal. NNE of Sta. Lucia, Zac. Location TS 56 TS 5 TS 10 TS 15 TS 11 1st exp TS 11 2nd exp TS 11 all exp TS 22 TS 21 ESC-3 TS 25 ESC-2 ESC-4 TS 46 TS 26 1st exp TS 26 2nd exp TS 26 all exp ESC 1 1st exp ESC 1 2nd exp ESC 1 3rd exp ESC 1 all exp TSS 1 Gdl 228 1st exp Gdl 228 2nd exp Gdl 228 all exp Gdl 228 TS 28 TS 28 1st exp TS 28 2nd exp TS 28 all exp Sample Table 1. Summary of New Isotopic Ages for the Southern Sierra Madre Occidentala 17 - 8 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO poor ash flow tuffs, pumice flows, and ash and pumice falls deposits. The latter was named the Atengo ignimbrite succession and attains a minimum thickness of 200 m. A sanidine concentrate for an ash flow tuff from the middle part of this succession was analyzed twice, yielding an isochron age of 28 ± 2 Ma (sample TS 11, Table 1). Moderately welded conglomerate and sandstone cover the Atengo succession inside the half graben. On the western side of the Atengo half graben a large rhyolitic dome complex was dated at 27.9 ± 0.2 Ma (sample TS-15, Table 1). 2.2. Miocene Volcanism [12] The Oligocene succession of the Huejuquilla-Atengo area is covered by a younger ignimbrite succession 6 km east of the village of Las Canoas (Figure 4). The Las Canoas succession is made of several ash and pumice flows and pumice falls that reach an aggregate thickness of at least 350 m. The most representative unit is a gray to pink, moderately welded ash flow tuff with small phenocrysts of biotite, plagioclase, alkali-feldspar and hornblende. Clark et al. [1981] dated this ignimbrite in the vicinity of Las Canoas by the K-Ar method obtaining an age of 23.5 ± 0.5 Ma for a feldspar separate. We have followed this unit over 30 km to the southwest, where it is down-dropped by the large extensional fault system of the Jesus Maria half graben (Figure 4). Our new 40Ar/39Ar date of 23.5 ± 0.4 Ma (sample TS-22, Table 1) for a plagioclase concentrate from one of the westernmost appearance of the Las Canoas succession is indistinguishable from the previous K-Ar age of Clark et al. [1981]. Toward the south the Las Canoas succession may be correlated with the lower part of the succession exposed in the Bolaños graben (Figure 5). Here Scheubel et al. [1988] report K-Ar ages of 23.7 ± 0.6 Ma for an andesite intercalated in a sequence of ignimbrites and rhyolites forming the lower part of the graben. They also report an age of 23.2 ± 0.5 Ma for an ignimbrite of the same succession (Figure 3). At places, thin mafic flows cap the Las Canoas succession (Figures 3, 4, and 5). We obtained whole rock 40Ar/39Ar ages of 21.3 ± 0.3 Ma for one of these flows exposed close to the town of Jesus Maria (104300W, 22150N; Figure 4; sample TS-21, Table 1). Our 40Ar/39Ar age is in complete agreement with the 40Ar/39Ar age of 21.6 ± 0.3 Ma obtained by Rossotti et al. [2002] for a succession of basaltic flows at the southern end of the Bolaños graben and with the K-Ar dates obtained by Scheubel et al. [1988] and Nieto-Obregon et al. [1981] for basaltic flows and dikes in the Bolaños mining area (Figures 3 and 5). [13] Immediately to the west of Jesus Maria (Figure 4), the Las Canoas succession is covered by another silicic succession that dominates the southwestern half of the study area (Figures 4 and 5). This consists of several whitish to light cream colored welded ash flows and air fall tuffs, and, to a lesser extent, rhyolitic domes. The ignimbrites thicken toward the Mesa del Nayar area (Figure 4). In addition, several subcircular lineaments suggestive of caldera depressions are observed in the satellite images and digital topography model in this area (Figures 2 and 4). We checked in the field the Nayar and Santa Teresa subcircular lineaments and found coignimbritic lag fall and breccia deposits, 17 - 9 lacustrine deposits, and/or rhyolite domes associated with them. These elements allow us to propose the existence of a number of caldera depressions aligned in a NNW direction, (the Nayar caldera field), which are the likely source of the Nayar ignimbrite succession (Figure 4). The latter was likely the product of many explosive eruptions occurred over a very short time span. This is confirmed by the 40 Ar/39Ar geochronology. In the valley of Arroyo el Frayle, which runs along the southern rim of the proposed Nayar caldera, at least 11 ignimbrite sheets with an aggregate thickness of 1000 m are observed. In this area we have dated a number of ignimbrite sheets located at different stratigraphic levels in the succession. The 40Ar/39Ar experiments performed on sanidine separates yielded ages of 21.1 ± 0.3 Ma for the lowermost ignimbrite and 21.0 ± 0.2 Ma for the uppermost sheet, located 1000 m above the previous one (samples TS-25 and TS-46, Table 1, respectively). A third, moderately welded ignimbrite, at 3 km to the west of the Nayar caldera rim, was dated 19.9 ± 0.4 Ma (sample TS 26, Table 1). The Nayar succession can be followed for tens of kilometers from the Mesa del Nayar area along the coastal area at least up to the latitude of Acaponeta (Figure 2). To the north of Mesa del Nayar the succession caps the last 600 to 1000 m of the volcanic plateau. Plagioclase phenocrysts from one of the highest ignimbrites of the succession in the Llano Grande area (105030W, 22450N, Figure 4) produced an age of 21.1 ± 0.7 Ma (sample ESC 3, Table 1). West of Mesa del Nayar the succession is found up to the present coast where we have dated the uppermost ignimbrite of a tilted block north of the village of San Juan Bautista at 20.9 ± 0.4 Ma (Figure 4; sample ESC 2, Table 1). To the east the topmost ignimbrite in the Sierra Los Huicholes (Figure 4) provided an age of 21.2 ± 0.3 Ma (sample ESC 4, Table 1). To the south we have dated again the uppermost ignimbrite of the succession exposed in the Sierra de Pajaritos (Figure 5). Here a biotite separate yielded an age of 20.7 ± 0.2 Ma (sample TSS 1, Table 1). In addition, several K-Ar ages reported in previous works for ignimbrites in the Aguamilpa and Santa Maria del Oro area (Figure 5) fall within the limits of our 40Ar/39Ar ages [Gastil et al., 1979; Damon et al., 1979; Soto and Ortega, 1982]. Similarly, in the Bolaños graben, Scheubel et al. [1988] report K-Ar ages of 21.3 ± 0.5 Ma and 20.1 ± 0.4 Ma for ignimbrites of their Huichol group, made of silicic ash flows and minor basalts (Figures 3 and 5), which dominate the succession exposed in the San Martin de Bolaños area. As a whole, the Nayar ignimbrite succession covers at least 5000 km2. Taking into account erosional removal, we consider an average of 900 m as a reasonable estimate of the original thickness. This would imply an original volume of 4500 km3 emplaced in 1.4 Ma, a value comparable to modern large caldera systems like the Taupo Volcanic Zone and Yellowstone [Christiansen, 1984; Houghton et al., 1995]. [14] Several rhyolitic domes are intercalated with the Nayar ignimbrite succession along the ring fault of the inferred calderas. Toward the coast, rhyolitic domes are more abundant and in some cases postdate the Nayar succession (Figure 4). We obtained an isochron age of 17 17 - 10 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO ± 1 Ma for a biotite and feldspar separate from a rhyolitic dome in the El Zopilote area (Figure 4; sample TS-28, Table 1). [15] The Nayar ignimbrite succession is capped in places by basaltic flows with a thickness of 50 to 100 m. In the Mesa del Nayar area these lavas are undated. However, in the Bolaños areas the uppermost basalts yielded a K-Ar age of 19.9 ± 0.4 Ma [Nieto-Obregon et al., 1981]. Subvolcanic intrusive bodies with granitic to quartz-monzonitic composition crop out on the western and southern margins of the SMO along the San Pedro and Santiago rivers where they flow at 150 to 50 m of elevation (Figures 4 and 5). We have performed three 40Ar/39Ar experiments for a sample of the San Juan Bautista granodiorite, one of the largest bodies in the area (Figure 4). The resulting isochron age of 20 ± 2 Ma (sample ESC 1, Table 1) is comparable to the K-Ar age of 17.2 ± 1.0 Ma reported by Rodriguez-Castañeda and Rodriguez-Torres [1992] for another granitic body 70 km to the south, and indicates that these rocks may be the intrusive equivalent of the Nayar ignimbrite succession. [16] Mafic volcanism represent the last volcanic event in the study region. Many mafic dikes are commonly emplaced parallel to NNW striking extensional faults in the westernmost part of the area (Figures 4 and 5). Previous K-Ar determinations for these dikes have consistently yielded ages between 11.9 Ma and 10.9 Ma [Damon et al., 1979; Clark et al., 1981]. Henry and Aranda-Gómez [2000] obtained nearly identical ages for similar mafic dikes in southern Sinaloa, just to the north of our study area. A thick succession of basaltic flows that marks the inception of the TMVB volcanism is emplaced at the boundary between the SMO and the Jalisco block between Tepic and Ixtlán del Rio and to the north of Guadalajara (Figure 2) [Ferrari et al., 2000b]. The emplacement of these lavas began at 11 Ma [Ferrari et al., 2000b, and references therein]. 2.3. Regional Correlation of the Silicic Volcanism [17] Our geochronology results for the southwestern SMO fit well with previous studies in the surrounding regions. Particularly, part of our stratigraphy and ages resemble closely those reported by McDowell and Keizer [1977] along the Durango-Mazatlán transect, located 100 km to the north. Our ages for the Valparaiso-Huejuquilla area (31.5 –28 Ma) overlap with those obtained by these authors in the Durango area, and our age of 27.9 Ma for the rhyolitic dome on the western side of the Atengo valley is identical to their age for a large rhyolitic dome at Las Adjuntas, 110 km west of Durango city. Similarly, the ages of the Las Canoas succession are identical to the 1000 m thick El Salto-Espinazo del Diablo ignimbrite sequence, for which McDowell and Keizer [1977] obtained eight ages, indistinguishable within error, which cluster at 23.5 Ma. Henry and Fredrikson [1987] also obtained similar ages for rhyolite and dacite lavas in Sinaloa. The El Salto succession can be followed several tens of kilometers to the SSE in the direction of the Las Canoas area. A possible caldera structure lies in between, centered on the village of Temoaya [Swanson and McDowell, 1984] (Figure 2). Inside this area we found a widespread fluvio-lacustrine sedimentary succession, which would confirm the existence of a volcano-tectonic depression. Other early Miocene calderas may be located west of Temoaya and southeast of Santa Lucia (Figure 2), but their existence was only suggested by remotely sensed features and helicopter surveys. Southeast of Bolaños, ignimbrites 23 m.y. old are reported in the Teul area [Moore et al., 1994] (Figure 2) and can be followed to the south up to the Rio Santiago, where Nieto-Obregon et al. [1985] report a K-Ar age of 23.6 ± 0.6 for a plagioclase concentrate from an ignimbrite 15 km to the northwest of the Santa Rosa dam (Figure 2). South of the river the early Miocene ignimbrites are covered by the younger volcanism of the Trans Mexican Volcanic Belt (review by Rosas-Elguera et al. [1997]). However, isolated blocks of ignimbrites tilted to the NNE crop out from the northernmost part of the Pliocene succession of the Trans Mexican Volcanic Belt near the village of Ixtlán del Rio (Figure 2). We have dated a groundmass concentrate from a welded ash flow tuffs in this area at 20.0 ± 0.3 Ma (sample Gdl 228, Table 1). [18] Previous workers reported ignimbrites with early Miocene ages farther east. In the Juchipila area (Figure 2), Webber et al. [1994] describe a succession of 10 ash flow tuffs, four of which were dated by the fission track method and yielded ages between 25.2 ± 2.2 and 24.9 ± 2.7 Ma. However, the tuffs cover mafic lavas with a K-Ar age of 23.7 ± 1.4 Ma, leaving the possibility that the age of the ignimbrite succession could still be 23 Ma. In the Los Altos de Jalisco region, located west of Guadalajara (Figure 2), ignimbrite and rhyolitic topographic highs emerged from a late Miocene basaltic plateau. Castillo-Hernandez and Romero-Rios [1991] obtained K-Ar ages of 24.1 ± 0.8 Ma and 24.7 ± 1.0 Ma, for sanidine separated from an ignimbrite and a rhyolite, respectively. The easternmost occurrence of early Miocene ignimbrites is found in central Guanajuato (Figure 6). Remnants of a single ignimbrite sheet cover Oligocene rhyolitic domes in the La Sauceda area. This ignimbrite, dated by K-Ar at 24.8 ± 0.6 Ma [Nieto-Samaniego et al., 1996], represents a distal facies and probably came from the Los Altos de Jalisco area. [19] Ignimbritic successions are also widespread to the south of the TMVB but are typically Late Cretaceous to early Paleocene in age within the Jalisco block [Ferrari et al., 2000a]. However, Tertiary ignimbrite units are exposed more to the east (Figure 6). Although most of the silicic volcanism in Michoacán and Guerrero is Oligocene [Morán-Zenteno et al., 1999], some younger ignimbrites are reported immediately to the south of the TMVB. About 35 km south of Lake Chapala (Figures 2 and 6), a thick pyroclastic succession overlies Tertiary intrusives and Cretaceous sedimentary rocks from the Michoacán block. The uppermost ignimbrite of this succession is a light brownish moderately welded unit with microphenocrysts of sanidine and white pumice supported by an ash matrix, which we called the Cazos ignimbrite. We have obtained a K-Ar age of 23.5 ± 0.9 Ma for a feldspar separate from the Cazos ignimbrite (sample CO-1, Table 1). In the Morelia area (Figure 6), a 200 – 300 m thick rhyolitic ignimbrite succession forms prominent cliffs 15 km south of the city (Puerto la Sosa area). Pasquarè FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO 17 - 11 Figure 6. Geographic plot of available ages (circles and boxes) of Oligocene to early Miocene silicic rocks in central Mexico (data from Ferrari et al. [1999], and this work). Note the broad NNW trending Oligocene silicic arc and the ESE migration of silicic volcanism between the 23 to the 21 Ma pulse. Cross pattern indicates the extension of the Los Cabos and Jalisco (Puerto Vallarta) late Cretaceous batholiths. Boundaries of Mexican states are also shown. Abbreviations are as follows: BC, Baja California; Nay, Nayarit; Jal, Jalisco; Gto, Guanajuato; Mich, Michoacán; Gro, Guerrero. et al. [1991] dated a ignimbrite from the upper part of the succession by K-Ar at 21 ± 1 Ma. These ages suggest a possible continuation of the 23 and 21 Ma silicic volcanism up to some tens of kilometers south of the TMVB. [20] In southern Baja California, Hausback [1984] and Umhoefer et al. [2001] obtained ages in the range 23– 17 Ma for ignimbrites intercalated in the lower part of the Comondú Formation. Particularly interesting are the ages obtained for the ‘‘La Paz tuff’’, a distinct succession of welded ash flow tuffs exposed in the La Paz-Punta Coyote area. Hausback [1984] reports K-Ar ages ranging between 21.8 ± 0.2 and 20.6 ± 0.2 Ma for this succession, which largely overlap our ages for the Nayar succession. Given the large geographic extent of the Nayar ignimbrite succession, the possibility exists that the La Paz tuff could be part of the latter. This correlation poses an additional constraint on the prerifting position of Baja California and confirms that the Los Cabos block was located immediately to the northwest of the Jalisco block prior to the opening of the Gulf of California [Schaaf et al., 2000] (Figure 6). 3. Tectonics of the Southwestern SMO 3.1. Geometry, Kinematics, and Time of Faulting [21] Several large structures cut the Tertiary succession of the southwestern SMO (Figures 2, 4, and 5). According to the geometry and kinematics they can be grouped in an eastern and a western extensional domains, and in the Santa Maria del Oro-Santa Rosa transpressional corridor (Figure 2). 3.1.1. Eastern Extensional Domain [22] This domain comprises six full grabens 30 to 40 km apart: Aguascalientes, Juchipila, Tlaltenango, Bolaños, La Ventana, and Mezquital (Figure 2). The last two were not studied but are grouped with the others because of their orientation and probable age. The first four grabens trend 010 to N-S and are typically over 100 km long and 20 km wide. The grabens are bounded by high-angle faults with dominant dip-slip displacements. Inside the graben, volcanic or sedimentary beds show tilts of 10 to 20 either to the east and the west. Offset of stratigraphic units can only be observed in the Bolaños graben, where it may reach 1800 m. Nieto-Samaniego et al. [1999] interpreted the four grabens as the results of E-W extension during the 23 to 21 Ma interval. This age for the extensional faulting was mainly inferred from geologic relations in the Tlaltenango graben where the flows of a basaltic shield volcano dated 21.8 ± 1.0 Ma abut against the fault scarps and appear unfaulted [Moore et al., 1994]. During our fieldwork, however, we observed that two normal fault scarps, <50 m high, also affect the shield volcano. Furthermore, we found that the 21 Ma ignimbrite succession is faulted in the Bolaños graben. 17 - 12 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO [23] In the Bolaños mining area (103450W, 21490N; Figure 5) Lyons [1988] estimates 1 to 1.5 km of normal offset of the mineralized body. The mineralization was dated at 20.8 ± 1.0 Ma, but it is unequivocally covered by basaltic flows dated 21.0 ± 0.4 Ma [Nieto-Obregon et al., 1981]. According to our observations in the Bolaños area the offset of the 21 Ma succession does not exceed 400– 500 m. In the San Martı́n de Bolaños area (Figure 5), Scheubel et al. [1988] dated at 21.3 ± 0.5 Ma an ash flow tuff on the floor of the graben, which is also found outside the graben over 1300 m higher. However, the geologic cross section of Scheubel et al. [1988] shows the ash flow tuff emplaced against the upper part of the western graben wall with a subvertical contact, suggesting that it could have flowed into a preexisting depression. Lyons [1988] also observed three generations of faults in the Bolaños area: A 060 normal fault set is cut by a 030 set which is in turn cut by N-S normal faults. Our regional observations, however, seem to indicate that this is not the rule along the Bolaños graben. In fact, the mining area is located within a broad accommodation zone that corresponds to a faulted relay ramp between two right stepping N-S trending normal faults. In this context the progression of faulting observed by Lyons [1988] may be interpreted as the progressive reorientation of secondary extensional structure during the northward propagation of the southernmost fault. [24] We interpret the above observations in the following way. The eastern extensional domain was extended shortly after the emplacement of the 23 Ma ignimbrite succession. The extensional pulse probably start to wane during the emplacement of the shield volcano in the Tlaltenango graben (21.8 ± 1.0 Ma) and after the deposition of the Nayar ignimbrite succession (21 Ma) in the Bolaños graben area. Alternatively, the Bolaños graben could have been involved in the younger episode of extension that affected the western domain. 3.1.2. Western Extensional Domain [25] The region between the Bolaños graben and the coastal plain is affected by N-S to NNW trending major structures that cut the grabens of the eastern domain (Figure 2). They are the Jesus Maria half graben and its northern continuation, the San Agustin graben, the Sierra Alica, Atengo, Sierra de Pajaritos, and Puente de Camotlán half grabens. To the west of these structures lies the relatively undeformed zone of the Nayar caldera field. Farther to the west the Pochotitán and San Pedro normal fault systems bound the Gulf of California (Figures 2, 4, and 5). With the sole exception of the northern part of the Jesus Maria half graben and a small area east of Jesus Maria village (104300W, 22150N; Figure 4), all major faults in this region systematically dip to the west or WSW and tilt the hanging blocks up to 30 to the east or ENE. This broadly extended region is separated from an undeformed zone to the north by the Rio Mezquital accommodation zone (Figure 2). This is also the boundary of two domains with opposite vergence: To the north, in fact, the beds systematically dip to the WSW, and major faults dip to the ENE (Figure 2) as documented by Henry and ArandaGomez [2000]. The Rio Mezquital accommodation zone was not studied in the field, but it could be a fault zone with a right-lateral component of motion. At the latitude of Tepic, the El Roble shear zone is a left-lateral normal accommodation zone that divides the western extensional domain from the Santa Maria del Oro transpressional corridor (Figures 2, 5, and 7). The Jesus Maria and the Atengo half grabens have vertical offset of over 2.2 km and 1 km, respectively. They both end to the west with flexure zones that bound flat lying and undeformed zones (Figure 4). In its northern part the Jesus Maria structure becomes a full graben (San Agustin graben), and the flexure becomes a broad extensional anticline (Figures 2 and 4). The Sierra Alica and Sierra de Pajaritos half grabens have vertical offset of 1.3 km, whereas the Puente de Camotlán half graben has an offset of 700 m. The San Pedro and Pochotitán fault systems form a major breakaway zone bounding the Gulf of California (Figures 2, 4, and 5). They consist of a series of normal faults cutting the flat lying SMO plateau to the east and have a cumulative minimum displacement in the order of 1.7 km. The faults are restricted to a 20 km wide and 160 km long belt, where ignimbrites as young as 19 Ma are found in a step-like structure with tilts up to 35 to the ENE. Outcrop-scale faults and accompanying dikes show a dominant NW to NNW trend (Figure 7). North of the Mezquital accommodation zone these fault systems continue into the Concordia fault [Henry and Aranda-Gomez, 2000], although their dip and the sense of tilting is reversed (Figure 2). [26] On average the direction of extension (sHmin) was 055 ± 17, indicating a consistent SW direction of extension for the whole western domain and part of the Bolaños graben (Table 2 and Figure 7). We interpret the uniform direction of extension as an indication of a common phase of deformation. [27] Available ages of faulted units indicate that the Atengo half graben is younger than 28 ± 2 Ma, the Jesus Maria and Sierra de Pajaritos half graben are younger than 20.6 Ma and the Pochotitán and San Pedro fault systems are younger than 19 Ma (Figures 4 and 5). However, no dated geologic units can constrain the upper age of activity of the former three structures. In the case of the San PedroPochotitán fault system the mafic dikes intruded along the normal faults have been dated at 11.9 ± 0.3 Ma at El Zopilote (Figure 4) [Clark et al., 1981] and 11.5 ± 0.5 Ma at Aguamilpa [Ferrari et al., 2000a] (Figures 2 and 5). At this latter location we measured 39 dikes with an average strike of N238E, some of which are tilted up to 70 (Figure 7 and Table 2, site 2). Between Aguamilpa and Tepic (Figure 4), however, basaltic flows dated 8.93 ± 0.11 Ma [Righter et al., 1995] are horizontal and cover tilted blocks of SMO ignimbrites. In Sinaloa, just to the north of our study region, Henry and Aranda-Gomez [2000] observed a similar situation. They obtained 40Ar/39Ar ages of 10.7 ± 0.2 and 11.0 ± 0.2 Ma for dikes moderately tilted and intruded in the conglomerate filling of the NNW trending half graben bounded by the Concordia fault. Accordingly with the above observations we consider that extension along this part of the eastern margin of the Gulf of California may have initiated a FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO 17 - 13 Figure 7. Tectonic map and microtectonic data collected in the western extensional domain. Bold arrow indicates the inferred direction of motion of a block to the west (Los Cabos block?) needed to produce the observed faulting. few millions of years before the intrusion of the dikes and continued for 1 – 2 m.y. after their emplacement. A middle to late Miocene age for faulting at the southern end of the Gulf of California is also confirmed by recent thermochronologic studies on the exhumation of the Los Cabos block, which would have been located just to the west of our study region at that time [Fletcher et al., 2000]. Particularly, they indicate that fast extensional exhumation of the Los Cabos block commenced by 12 Ma. 3.1.3. Santa Maria del Oro-Santa Rosa Transpressional Corridor [28] The southernmost part of the SMO, located close to the boundary of the Jalisco block, contrasts markedly with areas to the north. Instead of extensional structures contractile deformation dominates. The most striking structures are 10 to 40 km long NNW trending open folds arranged in an en echelon pattern between Santa Maria del Oro and the Santa Rosa dam (103440W, 20550N; Figures 2 and 5) [Ferrari, 1995]. Ignimbrites involved in the folds typically dip 30– 35 with the exception of the western flank of the Sierra El Pinabete asymmetric fold (Figure 5), where strata dip up to 75 – 80 on its western limb. In the field the rocks are affected by many 135 to 150 striking left-lateral oblique thrust and strike-slip faults (Figure 8). All these structures are indicative of a left-lateral transpressional shearing. A second group of folds with N-S trending and parallel axes lies to the WNW of the en echelon folds in the La Manga area (Figures 5 and 8). The strata are only gently tilted with dip of 5 – 10. These folds appear to be formed by low- intensity E-W compression on the rear of the northern part of the transpressional shear zone. [29] Folds and faults cut rocks as young as 19.0 ± 0.4 Ma east of Santa Maria del Oro [Damon et al., 1979] and 16.9 ± 0.5 Ma [Nieto-Obregon et al., 1985] in the Santa Rosa dam area. The folds are cut by vertical mafic dikes dated at 11.4 ± 0.3 Ma and 10.9 ± 0.2 Ma [Damon et al., 1979] (Figure 5). [30] The average direction of horizontal compression (sHmax) was 100 ± 26 (Table 2). However, the style of deformation varies from compression in the northwest (sites 11– 13, Table 2; Figure 8) to left-lateral transpression and oblique thrusting at the center of the area (sites 14– 17), to almost left-lateral transcurrence in the Santa Rosa area (site east of Jesus Maria Aguamilpa Dam Bolaños Graben La Cofradı́a Presa El Bañadero Paso De Lozada 1 Paso De Lozada 3 Sierra Pajaritos east Bolaños Graben Aguamilpa dam El Cajon 1 El Cajon 2 Santa Fe Plan de Barrancas Highway km 115 Plan de Barrancas Highway km 115, 3 Paso de La Yesca Santa Rosa Dam 00 1044500000 1042705500 1042705000 1042600500 1041103000 1041105000 1040403900 10444 30 1043103000 1043101000 1041603000 various sites SW of site 3 1045701000 0 1042605900 1044500000 1034105100 Longitude, West Lithology 101/9 114/2 121/15 121/16 275/15 115/15 278/20 216/60 141/62 early Miocene g 33/81 39/72 323/79 225/69 280/69 84/71 216/60 163/67 s1c 23.2 Ma 11.9 Ma 20.1 Ma early Miocene early Miocene early Miocene 17.2 Ma 20.7 Ma 21.0 Ma Age of Rocksb Southern Domain: Left Lateral Transpression 215100000 ignimbrite 22.4 – 18.7 Ma 212502000 ignimbrite early Miocene 212501500 ignimbrite early Miocene 0 00 2116 05 ignimbrite early Miocene? 0 00 2101 00 ignimbrite early Miocene 210004500 ignimbrite early Miocene 210804800 ignimbrite early Miocene ignimbrite 17 Ma 221802100 215100000 215702800 Western Domain: Extensiónf ignimbrite basaltic dikes ignimbrite ignimbrite 0 00 2136 30 ignimbrite 213003000 andesite 213003000 granite 213803000 ignimbrite various sites basaltic dikes SW of site 3 220002000 ignimbrite Latitude, North 9/10 204/11 214/12 331/72 129/73 2/55 81/69 4/26 315/28 129/1 147/6 140/11 125/4 139/17 326/9 4/26 340/23 s2c 230/76 15/79 342/71 214/8 7/9 214/31 186/6 110/14 47/2 219/9 238/16 230/1 34/21 45/13 234/16 100/14 70/1 235/1 s3c 8 8 8 7 9 12 9 13 9 19 39 20 8 8 9 13 8 15 Nd 0.75 0.87 0.85 0.73 0.39 0.53 0.24 0.52 0.53 0.50 0.27 0.20 0.17 0.52 0.14 0.35 Phie b Number of sites as in Figures 7 and 8. Age of rocks according to dated samples (details in the text) or to stratigraphic relations. c Trend and plunge of stress tensor axes determined by fault slip data inversion according to the method of Angelier [1990] except at sites 2 and 9, where they are eigenvectors of pole to dikes obtained by density analysis with the program Orient [Charlesworth et al., 1988]. d Number of planes used in the computation. e Tensor shape is (s2 – s3)/(s1 – s3). f Average direction sHmin = N55 ± 17. g Average direction of sHmax = N100 ± 26. a 11, 12, 13, 14, 15, 16, 17, 18, 10, east of El Venado 1, 2, 3, 4, 5, 6, 7, 8, 9, Site Number and Locationa Table 2. Kinematics of the Deformation 17 - 14 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO 17 - 15 Figure 8. Tectonic map and microtectonic data collected in the Santa Maria del Oro-Santa Rosa transpressional corridor. The geometry of the northern boundary of the Jalisco block is inferred on the basis of Ferrari et al. [2000a]. Bold arrow indicates the inferred direction of motion of the Jalisco block needed to produce the observed deformation. 18). This variety of structures is likely related to the original geometry of the boundary between the Jalisco block and the SMO, which was reactivated by WNW-ESE compression in middle Miocene time (Figure 8). 3.2. Magnitude of Extension [31] Although the normal faults of the study region are prominent features on satellite images and aerial photos, the deformation they accommodated was modest. Nieto-Samaniego et al. [1999] estimated an 8% of stretching in a nearly E-W direction from a profile crossing the Juchipila, Tlaltenango, and Bolaños graben. In the western extensional domain the relatively steep dipping (70 – 60) of faults and modest (30) tilting of blocks are also indicative of a limited amount of extension. The present detail of map- ping and the lack of a clear stratigraphic marker prevent the construction of a precise retrodeformable cross section in this area. However, we have tried to give a rough estimate of the amount of extension along a profile normal to the major extensional structure, the Jesus Maria half graben, using the area balance method described by Groshong [1994] (Figure 9). In this case we have considered the top of the 23 Ma ignimbrite as the reference level prior to extension and erosion (Figure 9). Since this surface outcrops at 2200 m elevation east of the fault system and at 1400 m west of the deformed zone, an average height of 1800 m was chosen. We took into account different depths of detachment and chose 10 km as a realistic, though conservative, approximation. This value gives a horizontal extension of 9% (Figure 9). In the less likely case of a shallower detachment level the extension would remain 17 - 16 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Figure 9. Interpretative cross section through the Jesus Maria half graben with estimation of minimum amount of extension using the method proposed by Groshong [1994]. The top of the 23 Ma Las Canoas ignimbrite succession was taken as reference level. Vertical exaggeration 1.65X. relatively small: For a detachment at 5 km depth, extension would be 17%. 4. Discussion 4.1. Episodic Nature of Ignimbrite Volcanism in the SMO [32] Previous studies recognized that most of the silicic volcanism in the SMO occurred in early Oligocene time (32 –28 Ma) throughout the province and in early Miocene time (23 Ma) in the southern part of the province [McDowell and Keizer, 1977; McDowell and Clabaugh, 1979; Henry and Fredrikson, 1987; Wark et al., 1990; Aguirre-Dı́az and McDowell, 1993; McDowell et al., 1997; Gans, 1997; Nieto-Samaniego et al., 1999]. Our new geochronologic data confirm these results but also demonstrate the regional significance of the early Miocene episode. Ignimbrites in the southwestern SMO actually cluster in the 32 –28 Ma and 24 –20 Ma time spans, and the latter episode covers about two thirds of the southwestern SMO. Thus our data confirm that ignimbritic volcanism in the SMO was concentrated in two pulses. The episodic nature of this volcanism can be appreciated in the probability density histogram of Figure 10, where all the available ages of silicic rocks of the southern SMO are plotted. Even so, the real size and the short duration of the two pulses of ignimbrite volcanism are not completely evident because available ages between 28 and 24 Ma often show large experimental errors and because the volume of volcanism is only loosely related with the number of available ages. [33] The distribution of this voluminous silicic volcanism in space and time define a clear trend (Figure 6). During the Oligocene ignimbrites were emplaced in a broad NW trending arc with a width of 400 km. The two early Miocene pulses are located in narrower belts within a progressively more southeastern position. It is clear that the locus of ignimbrite flare-up has converged toward the southwest from early Oligocene to early Miocene (24– 22 Ma) and then migrated toward the SSE during the 21 to 17.5 Ma time span. A comparison of the timing and space distribution of the Oligocene and early Miocene ignimbrite flare-up (Figures 6 and 10) indicates that the Oligocene episode was less concentrated in time and space, it occurred over a wider area, and had internal pulses less pronounced than the early Miocene one. 4.2. Generation of Silicic Volcanism in the SMO [34] Voluminous episodes of silicic volcanism have been related to different causes. The most common situation is continental break-up with variable involvement of mantle plumes as in the case of the huge Whitsunday Volcanic Figure 10. Histogram and cumulative probability curve for ages of silicic volcanic rocks in the Sierra Madre Occidental south of the Tropic of Cancer plotted constructed with Isoplot/Ex [Ludwig, 1999]. Note the marked pulse of silicic volcanism at 23 and 21 Ma. FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Province of northeastern Australia [Bryant et al., 2000] or the Chon Aike volcanic province of Patagonia and western Antarctica [Pankhurst et al., 2000]. On the other hand, the late Miocene-Quaternary central Andean silicic province has been explained as a consequence of delamination of the lower lithosphere and, possibly, the lower crust, with subsequent rising of hotter asthenospheric mantle [Kay and Kay, 1993]. In all the above cases a rapid increase in the uppermost mantle temperature is thought to have driven extensive melting of the mantle lithosphere and partial melting of the lower crust via basaltic underplating. [35] Isotope studies indicated contrasting scenarios for the generation of the Oligocene silicic volcanism in the SMO. On one hand, K. L. Cameron and coworkers [e.g., Cameron et al., 1980; Smith et al., 1996, and reference therein] have suggested that fractional crystallization of mafic magmas with relatively small crustal interaction generated the Oligocene ignimbrites in Chihuahua, while Ruiz et al. [1988, 1990] advocate a major involvement of the crust through anatexis. From binary mixing calculations, Verma [1984] also estimated values of involvement as high as 80% for Oligocene ignimbrites in the Zacatecas area. More recently, Albrecht and Goldstein [2000] modeled Sr, Nd, and Pb isotopic data from the central SMO in Chihuahua and obtained values of crustal involvement of up to 70%. Particularly, they propose that the silicic rocks were the result of assimilation/fractional crystallization or melting, assimilation, storage, and homogenization processes affecting first the lowermost crust and later the middle crust, as the thermal anomaly would rise into higher sections of the crust. [36] Although no isotope and geochemical studies are available for the early Miocene ignimbrite flare-up of the southern SMO, several lines of reasoning also suggest that melting of the crust played an important role in generating this massive silicic volcanism. First, the rate of silicic magma production and emplacement in the SMO during the early Miocene episodes is very high. The two silicic successions (23 and 21 Ma) have a thickness of 1000 m close to the source, which gradually decreases to <100 m in distal areas. They form a 700 km long and 120 km wide belt from central Sinaloa to western Guanajuato (Figure 6), covering an area of 84,000 km2. Thus, considering a mean thickness of 500 m, a volume of 42,000 km3 appears to have been emplaced in 2 m.y. These values approximate those of plume-related silicic large igneous provinces and suggest a similar mechanism. Second, partial melting of the crust needs much less amount of primary mafic magmas to produce a rhyolite. Several studies [Huppert and Sparks, 1988; Harry and Leeman, 1995] have shown that one volume unit of basalt may be able to produce from two thirds to one equivalent volume of rhyolite by crustal melting over timescale of 102 to 103 years. By contrast, fractional crystallization needs much more basalt volume, takes longer time periods, and implies the formation of rocks of intermediate composition. Our study of the southern SMO shows that volcanism occurred in fast and short pulses and that rocks of intermediate composition are absent. We thus conclude that the early Miocene ignimbrite flare-up was 17 - 17 primarily produced by a massive and fast generation of subcrustal mafic melts, which provided sufficient thermal energy to melt the crust. 4.3. Plate Tectonic Control of Ignimbrite Flare-Up [37] The rapid generation of mafic melts in the subcrustal lithosphere over several hundreds of kilometers asks for a plate tectonic mechanism. In the western United States, the Tertiary silicic flare-up has been related to the removal of the Farallon slab from beneath the North America plate [e.g., Coney, 1978; Humphreys, 1995] as the slab, which was subhorizontal during the Laramide orogeny, progressively rolled back and foundered in the mantle, exposing the upper plate to hotter asthenospheric mantle. This mechanism may have worked as well in Mexico during the Oligocene episode of ignimbrite flare-up that was coeval over a large area from the U.S.-Mexico border to the south of the Trans-Mexican Volcanic Belt. Indeed, it could have been triggered by a relatively sudden rollback of the subducted plate [Nieto-Samaniego et al., 1999] following the slowing in the Farallon-North America relative convergence that should have occurred some millions of years before the first contact of the East Pacific Rise with the continent [Atwater, 1989]. [38] For the early Miocene episode, however, we propose that the detachment of the Farallon slab (also called ‘‘slab breakoff’’) was the ultimate control over the timing and localization of silicic volcanism and extension in the southern SMO. Seismic tomography studies [Van der Lee and Nolet, 1997] have shown that the upper mantle beneath the SMO and north central Mexico is characterized by two subparallel high velocity anomalies (Figure 11). The most obvious interpretation of this double anomaly is that it represents two fragments of the subducted Farallon slab that detached at different times. Detachment of the subducted slab is well known in the Mediterranean-Carpathian region as a consequence of the arrival of the more buoyant continental crust in the subduction zone (see Wortel and Sparkman [2000] for a review). In a similar way, in western Mexico the initiation of detachment was the natural consequence of the approach of the East Pacific Rise to the paleotrench west of Baja California. The arrival of very young (<10 m.y. old) and buoyant oceanic crust in the subduction zone eventually results in the waning and cessation of subduction of the last remnant of the Farallon plate and their capture by the Pacific plate [Lonsdale, 1991]. The captured portion of the subducted slab began to move in an opposite direction (with the absolute motion of the Pacific plate), whereas the rest of the slab kept on sinking in the mantle. Therefore a tear in the subducted slab should have initiated at the time of the stalling of subduction and should have propagated laterally following the progressive termination of subduction off North America. According to the reconstruction of Atwater and Stock [1998] the first breakoff of the subducted part of the Farallon slab initiated at 28 Ma at the latitude of southern California, after the first interaction between the Pacific and the North America plates. At this time a trench parallel tear started to separate the short slab attached to the young Monterey and Arguello 17 - 18 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Figure 11. Map and cross section showing the main upper mantle seismic anomaly detected by the tomographic model of Van der Lee and Nolet [1997]. A double-high velocity anomaly corresponds broadly to the upper mantle region below the SMO in the early Miocene. Light shaded area in the cross section represents a positive anomaly exceeding 50 m/s, which is interpreted as portions of subducted slabs. Bold dashed line shows a possible geometry of the slab fragments (proposed in this work). Dark shading indicates a low-velocity anomaly below 50 m/s, corresponding to the shallow warm region which underlain the Gulf of California rift. microplates from the deeper and sinking part of the Farallon slab (Figure 12a). We propose that the tear in the subducted slab propagated laterally toward the SSE until reaching the southern SMO at the beginning of early Miocene time (Figures 12a and 13a). The seismic velocity contrast between the slab and the surrounding mantle observed in tomography studies may correspond to a temperature difference of several hundreds of degrees [Van der Lee and Nolet, 1997; Schmid et al., 2001]. Thermomechanical models of slab detachment also confirm that the temperature may increase over 500C above the lithospheric gap filled with upwelling hot asthenosphere [van de Zedde and Wortel, 2001]. Thus the abrupt removal of the slab via its detachment may have increased the temperature at the base of the crust of an amount analogous to that of a mantle plume. Unlike a mantle plume, however, this advective-type source of heat has a transient nature and should produce only a temporal episode of melting [van de Zedde and Wortel, 2001], which is what we observed in the southern SMO. 4.4. What Caused Deformation in the Southern SMO? [39] As outlined in section 3.1, two episodes of extension may be recognized in the southern SMO. The first one produced regularly spaced grabens between 23 and 20 Ma, whereas the second affected the western part of the study region approximately between 15 and 11 Ma. In both cases the magnitude of extension was moderate and below 10– 15%. In addition, left-lateral transpression affected the FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO 17 - 19 Figure 12. Generalized reconstruction of Pacific-North America plate interaction and volcanic and tectonic events in the southern SMO during the Miocene. (a) Reconstruction of plate boundaries and slab windows at 24 Ma (adapted from Nicholson et al. [1994] and Atwater and Stock [1998]) showing the proposed zone of slab breakoff, which is considered to have caused the 23 and 21 Ma episodes of silicic volcanism. Abbreviations are MP, Morro microplate; AP, Arguello microplate. (b) Plate configuration at 16 Ma [from Nicholson et al., 1994]. A second slab detachment event (which possible location is indicated) is postulated to have occurred between 15 and 11 Ma concurrently with the waning of the subduction of the Guadalupe and Magdalena plates. This, in turn, is considered to have produced widespread extension in the southern SMO and at the site of the future Gulf of California. (c) Plate configuration at the end of subduction of the Magdalena plate (12.5 Ma) (partly based on Stock and Lee [1994]). Recognizable magnetic anomaly pattern on the Pacific plate (PAC) are also shown [from Atwater, 1989]. The Jalisco block (JB) is thought to have moved toward the ESE shortly before this time producing a left-lateral tranpressional deformation along the southern margin of the SMO. Note the clockwise reorientation of magnetic anomaly on the Pacific plate north of the 22300N fracture zone during the last period of spreading. southern end of the SMO between 17 and 11 Ma. Although more detailed structural and geochronologic studies are needed to resolve the fine time and space evolution of these episodes, the general geometry, kinematics, and age data summarized in this work may be used to make a preliminary assessment of the causes that generated them. [40] The first episode is bracketed by the two peaks of ignimbrite flare-up at 23 and 21 Ma. The orientation of the grabens produced during this episode defines a sort of fan from the N15E trend of the easternmost graben (Juchipila) to the N-S trend of the westernmost one (La Ventana graben) (Figures 2 and 14). The area affected by 17 - 20 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Figure 13. Schematic cross section of the two slab detachment event postulated in this work that produced the early Miocene episode of ignimbrite flare up (SMO) and the late Miocene mafic volcanism at the eastern side of the Gulf of California (GCMV). Abbreviations are Farallon, Farallon plate; NOAM, North America plate; M, Magdalena microplate; BC, Baja California. this episode has also the highest topography within the southern SMO, with elevation ranging from 2400 to 3000 m (Figures 4 and 14). Moreover, elevation contours are roughly perpendicular to the graben orientation (Figure 14). Considering these facts, we speculate that the first episode was driven by magmatic intrusion into the crust and/or removal of the mantle lithosphere. The volume of granitic magma that remains in the crust is often 4 – 5 times the volume of silicic rocks emplaced at surface [Crisp, 1984]. However, part of this volume was already in place since we propose that a considerable amount of the crust was melted. It is reasonable to think, therefore, that material addition due to intrusion in the crust was about equivalent to the extruded material. Recent studies [Cruden, 1998; Améglio and Vigneresse, 1999; Grocott et al., 1999] showed that granitic batholiths frequently grow by lateral intrusion at the ductile/brittle FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO 17 - 21 Figure 14. Topography and structures of the southern SMO. The NW trending elliptical uplift and the fanning graben broadly coincide with the early Miocene ignimbrite successions and are thought to be due to magmatic intrusion at the ductile/brittle transition in the crust. transition zone by depressing the floor and uplifting the roof. Considering this mechanism one possibility is that the regional topography of part of the southern SMO and the fanning grabens of the eastern extensional domain are related to the intrusion of a tabular granitic batholith that fed the 23 and, particularly, the 21 Ma ignimbrite pulses. In addition, another mechanism capable of producing the observed topography and extension is the thermal removal of the mantle lithosphere beneath the southwestern SMO. In fact the material rising at the base of the plate may have replaced the cold mantle lithosphere with hotter asthenosphere, resulting in a doming and stretching of the crust (Figure 13a). The two mechanisms, however, may have worked concurrently. [41] The second extensional episode overlaps in time with the last subduction of the Guadalupe and Magdalena microplates and the initial transfer of Baja California to the Pacific plate (Figure 12b). Indeed, plate tectonic reconstructions indicate that subduction of the Farallon plate ended along the southern half of Baja California at 12.5 Ma, when the Peninsula and a short slab fragment stalled beneath it began to be captured by the Pacific plate [Atwater, 1989; Lonsdale, 1991; Atwater and Stock, 1998]. We agree with Henry and Aranda-Gomez [2000] in considering this episode of extension as a general manifestation of the ‘‘proto-Gulf extension,’’ driven by the capture of Baja California by the Pacific plate. However, Henry and Fredrikson [1987] and the data presented in this work suggest the possibility that extension may have initiated some time before the end of spreading between the Magdalena and Pacific plates. Indeed the component of active subduction of the Magdalena microplate beneath North America in the few million years before the end of spreading was likely small or close to zero. That is because after 15 Ma the ridge started to rotate clockwise tending to be perpendicular to the absolute motion of the Pacific plate and oblique to the trench. This implies that the spreading was driven by the Pacific motion rather than by slab pull. We propose that trench-normal extension observed in the western extensional domain may have initiated in response to the waning of subduction of the Magdalena microplate in middle Miocene time that, in turn, triggered a second slab detachment event (Figures 12b and 13b). This mechanism would have produced also a partial melting of the mantle below the present Gulf of California. In this case, however, the extensional regime induced by plate boundary forces allowed rapid upraise of mafic melts (12 –10 Ma mafic dikes and flows) rather than the formation of silicic crustal melts (Figure 13b). [42] The transpressional deformation observed at the southern boundary of the SMO has no trivial explanation. The plate configuration west of Baja California for that time is not precisely defined because the amount of right-lateral motion along the Tosco-Abreojos fault zone (Figure 12c) is uncertain [e.g., Fletcher, 2001]. However, it is reasonable to think that the middle Miocene plate boundary between the 17 - 22 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Magdalena and Cocos plates was located in front of the mouth of the Gulf [Lyle and Ness, 1991; Stock and Lee, 1994]. In this case the transpressional deformation zone would be located along the inland projection of the Magdalena-Cocos plate boundary; it also corresponds to the boundary between the SMO and the Jalisco block (Figure 12c). These observations led Ferrari [1995] to relate this event to an east-southeastward motion of the Jalisco block shortly before the end of subduction of the Magdalena microplate. In addition, we note that the magnetic anomaly pattern preserved on the western side of the East Pacific Rise indicates a progressive clockwise reorientation of the spreading ridge between the Magdalena and the Pacific plates between 14.1 and 12.9 Ma (Figure 12c). The contrast between the waning subduction of the Magdalena microplate and the steady subduction of the Cocos plate added to the small convergent direction between them at 14– 12.9 Ma may have produced the left-lateral transpressional deformation observed along the inland continuation of their subducted boundary. 5. Concluding Remarks [43] Massive generation of silicic volcanism occurred in the southern SMO during two episodes at 31.5 –28 Ma and 23.5 – 20 Ma. Widespread but moderate extension affected the upper crust during two periods: 23– 20 Ma in the eastern part and 15– 10 Ma in the western part. The southern end of the SMO underwent left-lateral transpression roughly at the same time. We have linked these magmatic and tectonic events to the evolution and dynamics of the plate motion and the mantle in western North America during the Miocene. Particularly, we propose that two consecutive detachments of the subducted Farallon slab in early and middle Miocene time played an important role in controlling the locus and timing of volcanism and extension. Clearly, the interaction among tectonomagmatic events and plate tectonic history postulated in this paper is somewhat speculative but has a number of implications that could be tested by further studies. The underplating of basaltic magmas must have produced a mafic lower crust that may be detected by seismic and gravimetric studies. The involvement of the asthenosphere after the slab detachment must have left a geochemical and isotope imprint in the early Miocene volcanism that may be quantified by petrologic studies. We hope that this work may stimulate further studies that may unravel the enigma of the Sierra Madre Occidental. [44] Acknowledgments. This research was supported by grant CONACyT P-0152 T and Instituto de Geologı́a, UNAM, to L. Ferrari. We are indebted to Consejo de Recursos Minerales and particularly to J. C. Salinas, Head of Geology and Geochemistry Department, for allowing the use of a helicopter. Suzan van der Lee kindly provided sections of her tomographic model NA97 used to sketch Figure 11. Formal reviews by Joann Stock and Wanda Taylor provided valuable suggestions and improved the clarity of the original manuscript. We also thank A. Nieto-Samaniego, D. MoránZenteno, S. Alaniz-Alvarez, Fred McDowell, Chris Henry, J. ArandaGómez, J. Urrutia Fucugauchi, and G. Aguirre-Diaz for several useful discussions at different stages of the research. G. Aguirre-Diaz also helped in an initial phase of fieldwork. Technical assistance by V. Moreno, A. S. Rosas, and V. Pérez is acknowledged. References Aguirre-Dı́az, G. J., and F. W. McDowell, Nature and timing of faulting and synextensional magmatism in the southern Basin and Range, central-eastern Durango, Mexico, Geol. Soc. Am. Bull., 105, 1435 – 1444, 1993. Albrecht, A., and S. L. Goldstein, Effects of basement composition and age on silicic magmas across an accreted terrane-Precambrian crust boundary, Sierra Madre Occidental, Mexico, J. S. Am. Earth Sci., 13, 255 – 273, 2000. Améglio, L., and J. L. Vigneresse, in Understanding Granites: Integrating New and Classical Techniques, edited by A. Castro, C. Fernandez, and J. L. Vigneresse, Geol. Soc. Spec. Publ., 158, 39 – 54, 1999. Angelier, J., Inversion of field data in fault tectonics to obtain the regional stress, III, A new rapid direction inversion method by analytical means, J. Geophys., 103, 363 – 376, 1990. Atwater, T., Plate tectonic history of northeast Pacific and western North America, in The Geology of North America, vol. N, The Eastern Pacific Ocean and Hawaii, edited by E. L. Winterer, D. M. Hussong, and R. W. Decker, pp. 21 – 71, Geol. Soc. of Am., Boulder, Colo., 1989. Atwater, T., and J. M. Stock, Pacific-North America plate tectonics of the Neogene southwestern United States: An update, Int. Geol. Rev., 40, 375 – 402, 1998. Bryant, S. E., A. Ewart, C. J. Stephens, J. Parianos, and P. J. Downes, The Witsunday volcanic province, central Queensland, Australia: Lithological and stratigraphic investigations of a silicic-dominated large igneous province, J. Volcanol. Geotherm. Res., 99, 55 – 78, 2000. Cameron, M., W. C. Bagby, and K. L. Cameron, Petrogenesis of voluminous mid-Tertiary ignimbrites of the Sierra Madre Occidental, Chihuahua, Mexico, Contrib. Mineral. Petrol., 74, 271 – 284, 1980. Cashman, P. H., and M. A. Ellis, Fault interaction may generate multiple slip vectors on a single fault surface, Geology, 22, 1123 – 1126, 1994. Castillo-Hernandez, D., and F. Romero-Rios, Estudio geológico-regional de Los Altos, Jalisco y El Bajio, Open File Rep., 02-91, 35 pp., Com. Fed. de Electr., Gerencia de Proy. Geotermoelectr., Dep. Explor., Morelia, Mich., Mexico, 1991. Christiansen, R. L., Yellowstone magmatic evolution: Its bearing on understanding large-volume explosive volcanism, in Explosive Volcanism: Inception, Evolution, and Hazards, pp. 84 – 95, Nat. Acad. Press, Washington, D. C., 1984. Clark, K. F., P. E. Damon, M. Shafiquillah, B. F. Ponce, and D. Cardenas, Sección geológica-estructural a través de la parte sur de la Sierra Madre Occidental, entre Fresnillo y la costa de Nayarit, Mem. Téc., XIV, pp. 69 – 99, Asoc. Ing. Mineros, Metal. y Geól. de México, Mexico City, 1981. Coney, P. J., Mesozoic-Cenozoic Cordilleran plate tectonics, in Cenozoic Tectonics and Regional Geophysics of the Western Cordillera, edited by R. B. Smith and G. P. Eaton, Mem. Geol. Soc. Am., 152, 33 – 50, 1978. Crisp, J. A., Rate of magma emplacement and volcanic output, J. Volcanol. Geotherm. Res., 20, 177 – 211, 1984. Cruden, A. R., On the emplacement of tabular granites, J. Geol. Soc. London, 155, 853 – 862, 1998. Damon, P. E., J. Nieto-Obregon, and L. Delgado-Argote, Un plegamiento neogénico en Nayarit y Jalisco y evolución geomórfica del Rio Grande de Santiago, Mem. Téc., XIII, pp. 156 – 191, Asoc. Ing. Mineros, Metal. y Geól. de México, 1979. Ferrari, L., Miocene shearing along the northern boundary of the Jalisco block and the opening of the southern Gulf of California, Geology, 23, 751 – 754, 1995. Ferrari, L., S. A. Nelson, J. Rosas-Elguera, and G. J. Aguirre-Dı́az, Tectonics and volcanism of the western Mexican Volcanic Belt, in Magmatism and Tectonics in Central and Northwestern Mexico.A selection of the 1997 IAVCEI General Assembly excursions, edited by G. J. Aguirre-Diaz, pp. 4 – 25, Inst. de Geol., Univ. Nac. Autón. de Méx., Mexico City, 1997. Ferrari, L., M. Lopez-Martinez, G. Aguirre-Diaz, and G. Carrasco-Nuñez, Space-time patterns of Cenozoic arc volcanism in central Mexico: From the Sierra Madre Occidental to the Mexican Volcanic Belt, Geology, 27, 303 – 307, 1999. Ferrari, L., G. Pasquaré, S. Venegas-Salgado, and F. Romero-Rios, Geology of the western Mexican Volcanic Belt and adjacent Sierra Madre Occidental and Jalisco block, Geol. Soc. Am. Spec. Pap., 334, 65 – 84, 2000a. Ferrari, L., S. Conticelli, G. Vaggelli, C. Petrone, and P. Manetti, Late Miocene mafic volcanism and intraarc tectonics during the early development of the Trans-Mexican Volcanic Belt, Tectonophysics, 318, 161 – 185, 2000b. Fletcher, J. M., The Baja California Borderland and the FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Neogene evolution of the Pacific-North American plate boundary, Eos Trans. American Geophysical Union, 82(47), Fall Meet. Suppl., Abstract T52C0966, 2001. Fletcher, J. M., B. P. Kohn, D. A. Foster, and A. J. W. Gleadow, Heterogeneous Neogene cooling and exhumation of the Los Cabos block, southern Baja California: Evidence from fission-track thermochronology, Geology, 28, 107 – 110, 2000. Francis, P. W., and C. J. Hawkesworth, Late Cenozoic rates of magmatic activity in the Central Andes and their relationships to continental crust formation and thickening, J. Geol. Soc. London, 151, 845 – 854, 1994. Gans, P. B., Large-magnitude Oligo-Miocene extension in southern Sonora: Implications for the tectonic evolution of northwest Mexico, Tectonics, 16, 388 – 408, 1997. Gastil, R. G., R. P. Phillips, and E. C. Allison, Reconnaissance geology of the state of Baja California, Mem. Geol. Soc. Am., 140, 170 pp., 1975. Gastil, R. G., D. Krummenacher, and J. Minch, The record of Cenozoic volcanism around the Gulf of California, Geol. Soc. Am. Bull., 90, 839 – 857, 1979. Grocott, J., A. Garde, B. Chadwick, A. R. Cruden, and C. Swager, Emplacement of Rapakiwi granite and syenite by floor depression and roof uplift in the Paleoproterozoic Ketilidian orogen, South Greenland, J. Geol. Soc. London, 156, 15 – 24, 1999. Groshong, R. H., Area balance, depth of detachment, and strain in extension, Tectonics, 13, 1488 – 1497, 1994. Harry, D. L., and W. P. Leeman, Partial melting of melt metasomatized subcontinental mantle and the magma source potential of the lower lithosphere, J. Geophys. Res., 100, 10,255 – 10,269, 1995. Hausback, B. P., Cenozoic volcanic and tectonic evolution of Baja California Sur, México, in Geology of Baja California Peninsula, vol. 199, edited by V. A. Frizzell, pp. 219 – 236, Soc. for Sediment., Pac. Sec., Geol., Tulsa, Okla., 1984. Henry, C. D., and J. J. Aranda-Gomez, The real southern Basin and Range: Mid- to late-Cenozoic extension in Mexico, Geology, 20, 701 – 704, 1992. Henry, C., and J. J. Aranda-Gomez, Plate interactions control middle-late Miocene, proto-Gulf and Basin and Range extension in the southern Basin and Range, Tectonophysics, 218, 1 – 26, 2000. Henry, C. D., and G. Fredrikson, Geology of part of southern Sinaloa, Mexico, adjacent to the Gulf of California, Geol. Soc. Am. Maps Chart Ser., MCH 063, 1 sheet, 14 pp., Geol. Soc. of Am, Boulder, Colo., 1987. Houghton, B. F., C. J. N. Wilson, M. O. McWilliams, M. A. Lanphere, D. Weaver, R. M. Briggs, and M. S. Pringle, Chronology and dynamics of a large silicic magmatic system: Central Taupo Volcanic Zone, New Zealand, Geology, 23, 13 – 16, 1995. Humphreys, E. D., Post-Laramide removal of the Farallon slab, western United States, Geology, 23, 987 – 990, 1995. Huppert, H. E., and R. S. J. Sparks, The generation of granitic magmas by intrusion of basalt into continental crust, J. Petrol., 29, 599 – 642, 1988. Kay, R. W., and S. Kay, Delamination and delamination magmatism, Tectonophysics, 219, 177 – 189, 1993. Lonsdale, P., Structural pattern of the pacific floor offshore peninsular California, in The Gulf and the Peninsular Province of the Californias, edited by J. P. Dauphin and B. R. T. Simoneit, AAPG Mem., 47, 87 – 125, 1991. Ludwig, K. R., Isoplot/Ex 2.01: A geochronological toolkit for Microsoft Excel, Berkeley Geochronological Cent. Spec. Publ. 1a, 50 pp., 1999. Luhr, J., C. D. Henry, T. B. Housh, J. Aranda-Gómez, and W. C. McIntosh, Early extension and associated mafic alkalic volcanism from the southern Basin and Range province: Geology and petrology of the Rodeo and Nazas volcanic fields, Durango, Mexico, Geol. Soc. Am. Bull., 113, 760 – 773, 2001. Lyle, M., and G. Ness, The opening of the Southern Gulf of California, in The Gulf and Peninsular Provinces of the Californias, edited by J. P. Dauphin and B. R. T. Simoneit, AAPG Mem., 47, 403 – 423, 1991. Lyons, J. I., Geology and ores deposits of the Bolaños silver district, Jalisco, Mexico, Econ. Geol., 83, 1560 – 1582, 1988. McDowell, F. W., and S. E. Clabaugh, Ignimbrites of the Sierra Madre Occidental and their relation to the tectonic history of western Mexico, Geol. Soc. Am. Spec. Pap., 180, 113 – 124, 1979. McDowell, F. W., and R. P. Keizer, Timing of midTertiary volcanism in the Sierra Madre Occidental between Durango city and Mazatlan, Mexico, Geol. Soc. Am. Bull., 88, 1479 – 1487, 1977. McDowell, F. W., J. Roldán-Quintana, and R. AmayaMartı́nez, Interrelationship of sedimentary and volcanic deposits associated with Tertiary extension in Sonora, Mexico, Geol. Soc. Am. Bull., 109, 1349 – 1360, 1997. Moore, G., C. Marone, I. S. E. Carmichael, and P. Renne, Basaltic volcanism and extension near the intersection of the Sierra Madre volcanic province and the Mexican Volcanic Belt, Geol. Soc. Am. Bull., 106, 383 – 394, 1994. Morán-Zenteno, D. J., et al., Tertiary arc-magmatism of the Sierra Madre del Sur, Mexico, and its transition to the volcanic activity of the Trans-Mexican Volcanic Belt, J. S. Am. Earth Sci., 12, 513 – 535, 1999. Nicholson, C., C. C. Sorlien, T. Atwater, J. C. Crowell, and B. P. Luyendyk, Microplate capture, rotation of the western Transverse Ranges, and initiation of the San Andreas transform as a low-angle fault system, Geology, 22, 491 – 495, 1994. Nieto-Obregon, J., L. Delgado-Argote, and P. E. Damon, Relaciones petrológicas y geocronológicas del magmatismo de la Sierra Madre Occidental y el Eje Neovolcánico en Nayarit, Jalisco y Zacatecas, Mem. Téc., XIV, pp. 327 – 361, Asoc. Ing. Mineros, Metal. y Geól. de México, CITY, 1981. Nieto-Obregon, J., L. Delgado-Argote, and P. E. Damon, Geochronologic, petrologic and structural data related to large morphologic features between the Sierra Madre Occidental and the Mexican Volcanic Belt, Geofis. Int., 24, 623 – 663, 1985. Nieto-Samaniego, A. F., and S. A. Alaniz-Alvarez, Origin and tectonic interpretation of multiple fault patterns, Tectonophysics, 270, 197 – 206, 1997. Nieto-Samaniego, A. F., C. Macı́as-Romo, and S. A. Alaniz-Alvarez, Nuevas edades isotópicas de la cubierta volcánica cenozoica de la parte meridional de la Mesa Central, México, Rev. Mex. Cienc. Geol., 13, 117 – 122, 1996. Nieto-Samaniego, A. F., L. Ferrari, S. A. Alaniz-Alvarez, G. Labarthe-Hernandez, and J. Rosas-Elguera, Variation in Cenozoic extension and volcanism across the southern Sierra Madre Occidental volcanic province, western Mexico, Geol. Soc. Am. Bull., 111, 347 – 363, 1999. Nourse, J. A., T. H. Anderson, and L. T. Silver, Tertiary metamorphic core complexes in Sonora, northwestern Mexico, Tectonics, 13, 359 – 366, 1994. Pankhurst, R. J., T. R. Riley, C. M. Fanning, and S. P. Kelley, Episodic silicic volcanism in Patagonia and the Antarctic Peninsula: Chronology of magmatism associated with the break-up of Gondwana, J. Petrol., 41, 605 – 625, 2000. Pasquarè, G., L. Ferrari, V. H. Garduño, A. Tibaldi, and L. Vezzoli, Geological map of the central sector of Mexican Volcanic Belt, States of Guanajuato and Michoacan, Geol. Soc. Am. Map Chart Ser., MCH 072, 1 sheet, 20 pp., Geol. Soc. of Am., Boulder, Colo., 1991. Pollard, D. D., S. D. Saltzer, and A. Rubin, Stress inversion methods: are they based on faulty assumptions?, J. Struct. Geol., 15, 1045 – 1054, 1993. Righter, K., I. S. E. Carmichael, and T. Becker, PlioceneQuaternary faulting and volcanism at the intersection of the Gulf of California and the Mexican Volcanic Belt, Geol. Soc. Am. Bull., 107, 612 – 626, 1995. Rodrı́guez-Castañeda, J. L., and R. Rodrı́guez-Torres, Geologı́a estructural y estratigrafı́a del área entre 17 - 23 Guadalajara y Tepic, Estados de Jalisco y Nayarit, México, Rev. Inst. Geol. UNAM, 10, 99 – 110, 1992. Rosas-Elguera, J., L. Ferrari, M. López-Martinez, and J. Urrutia-Fucugauchi, Stratigraphy and tectonics of the Guadalajara region and triple-junction area, western Mexico, Int. Geol. Rev., 39, 125 – 140, 1997. Rossotti, A., L. Ferrari, M. Martı́nez-López, and J. Rosas-Elguera, Geology of the boundary between the Sierra Madre Occidental and the Trans-Mexican Volcanic Belt in the Guadalajara region, western Mexico, Rev. Mex. Cienc. Geol., 19, 1 – 15, 2002. Ruiz, J., P. J. Patchett, and R. J. Arculus, Nd-Sr isotopic composition of lower crustal xenoliths.Evidence for the origin of mid-Tertiary felsic volcanics in Mexico, Contrib. Mineral. Petrol., 99, 36 – 43, 1988. Ruiz, J., P. J. Patchett, and R. J. Arculus, Reply to "Comments on Nd-Sr isotopic compositions of lower crustal xenoliths: Evidence for the origin of midTertiary felsic volcanics in Mexico," by K. L. Cameron and J. V. Robinson, Contrib. Mineral. Petrol., 104, 615 – 618, 1990. Schaaf, P., H. Bohnel, and J. A. Perez-Venzor, Pre-Miocene palaeogeography of the Los Cabos Block, Baja California Sur: Geochronological and palaeomagnetic constraints, Tectonophysics, 318, 53 – 69, 2000. Scheubel, F. R., K. F. Clark, and E. W. Porter, Geology, tectonic environment and structural controls in the San Martin de Bolaños district, Jalisco, Mexico, Econ. Geol., 83, 1703 – 1720, 1988. Schmid, C., S. Goes, S. van der Lee, and D. Giardini, Thermal and seismic signature of the trailing fragments of the Farallon slab, Eos Trans. American Geophysical Union, 82(47), Fall Meet. Suppl., Abstract F1197-1998, 2001. Smith, R. D., K. L. Cameron, F. W. McDowell, S. Nyemeyer, and D. E. Sampson, Generation of voluminous silicic magmas and formation of mid Cenozoic crust beneath north-central Mexico: Evidence from ignimbrites, associated lavas, deep crustal granulites, and mantle pyroxenites, Contrib. Mineral. Petrol., 123, 375 – 389, 1996. Soto, M. A., and J. G. Ortega, Geologia del Rio Santiago en los estados de Jalisco y Nayarit, México, paper presented at XII Congress of the Sociedad Geológico de México, Mexico City, 1982. Stewart, J. H., et al., Map showing Cenozoic tilt domains and associated structural features, western North America, in Accommodation Zones and Transfer Zones: The Regional Segmentation of the Basin and Range Province, edited by J. E. Faulds and J. H. Stewart, Geol. Soc. Am. Spec. Pap., 323, 1998. Stock, J. M., and J. Lee, Do microplates in subduction zones leave a geologic record?, Tectonics, 13, 1472 – 1487, 1994. Swanson, E. R., and F. W. McDowell, Calderas of the Sierra Madre Occidental volcanic field, western Mexico, J. Geophys. Res., 89, 8787 – 8799, 1984. Twiss, R. J., and J. R. Unruh, Analysis of fault slip inversions: Do they constrain stress or strain rate?, J. Geophys. Res., 103, 12,205 – 12,222, 1998. Umhoefer, P. J., R. Dorsey, S. Willsey, L. Mayer, and P. Renne, Stratigraphy and geochronology of the Comondú Group near Loreto, Baja California sur, Mexico, Sediment. Geol., 144, 125 – 147, 2001. Van der Lee, S., and G. Nolet, Seismic image of the subducted trailing fragments of the Farallon plate, Nature, 386, 266 – 269, 1997. van de Zedde, D. M. A., and M. J. R. Wortel, Shallow slab detachment as a transient source of heat at midlithospheric depth, Tectonics, 20, 868 – 882, 2001. Verma, S. P., Sr and Nd isotopic evidence for petrogenesis of mid-Tertiary felsic volcanism in the mineral district of Zacatecas, Zac. (Sierra Madre Occidental), Mexico, Isot. Geosci., 2, 37 – 53, 1984. Wark, D. A., K. A. Kempter, and F. W. McDowell, Evolution of waning subduction-related magmatism, northern Sierra Madre Occidental, Mexico, Geol. Soc. Am. Bull., 102, 1555 – 1564, 1990. 17 - 24 FERRARI ET AL.: IGNIMBRITE FLARE-UP AND DEFORMATION, WESTERN MEXICO Webber, K., L. A. Fernández, and B. Simmons, Geochemistry and mineralogy of the Eocene-Oligocene volcanic sequence, southern Sierra Madre Occidental, Juchipila, Zacatecas, México, Geofı́s. Int., 33, 77 – 89, 1994. Wortel, M. J. R., and W. Sparkman, Subduction and slab detachment in the Mediterranean-Carpathian region, Science, 290, 1910 – 1917, 2000. York, D., Least squares fitting of a straight line with correlated errors, Earth Planet. Sci. Lett., 5, 320 – 324, 1969. L. Ferrari, Centro de Geociencias, Universidad Nacional Autónoma de México, Campus Juriquilla, Qro. Apdo. Postal 1-742 Centro, Querétaro, 76001 Qro., Mexico. ([email protected]) M. López-Martı́nez, Departamento de Geologı́a, Centro de Investigación Cientı́fica y Educación Superior de Ensenada, Km. 107 Carretera Tijuana-Ensenada, C. P. 22860, Ensenada, Baja California, México. ([email protected]) J. Rosas-Elguera, Centro de Ciencias de la Tierra, Universidad de Guadalajara, Calzada Olı́mpica y Blvd. Marcelino Garcı́a Barragán, 44840 Guadalajara, Jal., México. ( [email protected])