This is the author’s version of a work that was submitted/accepted for pub-lication in the following source:

Bryan, Scott Edward, Orozco-Esquivel, Teresa, Ferrari, Luca, & Lopez-Martinez, Margarita(2013)Pulling apart the mid to late Cenozoic magmatic record of the Gulf of Cali-fornia : is there a Comondú arc?Orogenic Andesites and Crustal Growth, 385.

This file was downloaded from: https://eprints.qut.edu.au/58549/

c© Copyright 2013 Geological Society of London

Notice: Changes introduced as a result of publishing processes such ascopy-editing and formatting may not be reflected in this document. For adefinitive version of this work, please refer to the published source:

https://doi.org/10.1144/SP385.8

1

Pulling Apart the Mid to Late Cenozoic Magmatic Record of the Gulf of California: Is 1 there a Comondú Arc? 2

3 4 Bryan S.E.1, Orozco-Esquivel T.2, Ferrari L.2,3, López-Martínez, M.4 5 6 1. School of Earth, Environmental and Biological Sciences, Queensland University of 7 Technology, GPO Box 2434, Brisbane, 4001, Australia. Email: scott.bryan@qut.edu.au 8 9 2. Centro de Geociencias, Universidad Nacional Autonoma de Mexico, Blvd Juriquilla 3001, 10 Querétaro, 76230, Mexico. 11 12 3. Instituto de Geología, Universidad Nacional Autonoma de Mexico, Circuito Investigacion 13 Cientifica, Ciudad Universitaria, Mexico City, 04510, Mexico. 14 15 4. CICESE, Carretera Ensenada-Tijuana No 3918, Zona Playitas, Ensenada, Baja 16 California, 22860, México. 17 18 19 Abstract: 20

21

The composition of the lithosphere can be fundamentally altered by long-lived subduction 22

processes such that subduction-modified lithosphere can survive for 100's Myrs. Incorrect 23

petrotectonic interpretations result when spatial-temporal-compositional trends of, and source 24

contributions to, magmatism are not properly considered. Western Mexico has had protracted 25

Cenozoic magmatism developed mostly in-board of active oceanic plate subduction beneath 26

western North America. A broad range of igneous compositions from basalt to high-silica 27

rhyolite were erupted with intermediate to silicic compositions in particular, showing calc-28

alkaline and other typical subduction-related geochemical signatures. A major Oligocene 29

rhyolitic ignimbrite “flare-up” (>300,000 km3) switched to a bimodal volcanic phase in the 30

Early Miocene (~100,000 km3), associated with distributed extension and opening of 31

numerous grabens. Extension became more focussed ~18 Ma resulting in localised volcanic 32

activity along the future site of the Gulf of California. This localised volcanism (known as the 33

Comondú “arc”) was dominantly effusive and andesite-dacite in composition. Past tectonic 34

interpretations of Comondú-age volcanism may have been incorrect as these regional 35

temporal-compositional changes are alternatively interpreted as a result of increased mixing 36

of mantle-derived basaltic and crust-derived rhyolitic magmas in an active rift environment 37

rather than fluid flux melting of the mantle wedge above the subducting Guadalupe Plate. 38

39

2

Supplementary material: References from which whole-rock geochemical and radiometric 40

age data have been compiled in this paper are available at www.geolsoc.org.uk/SUP00000. 41

42

Introduction 43

44

Mid- to Late Tertiary magmatism along the continental margin of western Mexico (Fig. 1) is 45

an example where interpretations on the tectonic setting of magmatism have been strongly 46

influenced by the continent-margin position, calc-alkaline affinity, relatively primitive 47

isotopic characteristics, the presence of andesitic or intermediate composition volcanic rocks, 48

and general association with continued subduction beneath western North America (e.g., 49

Cameron et al., 1980; Lanphere et al., 1980; Hausback, 1984; Wark et al., 1990; Umhoefer et 50

al., 2001). A feature of the magmatic record in western Mexico is the general continuity of 51

erupted/igneous compositions from basalt through to high-silica rhyolite (Fig. 2). This raises 52

the question of whether the presence of calc-alkaline “andesitic” igneous compositions offers 53

any constraint on the tectonic setting of magmatism. Past studies have widely interpreted the 54

mid- to Late Tertiary (~38-12 Ma) volcanism as recording a supra-subduction zone volcanic 55

arc (e.g., Cameron et al., 1980; Hausback, 1984; Sawlan & Smith, 1984; Wark et al., 1990; 56

Ferrari et al., 1999; Martín-Barajas et al., 1995, Martín et al., 2000; Umhoefer et al., 2001). 57

However, for Oligocene to Early Miocene volcanic activity (~38-18 Ma), the overwhelming 58

silicic composition, the eruptive scale, volume and output rate, and the rhyolite ignimbrite-59

dominated character of the erupted products sourced from multiple calderas and fissures (e.g., 60

Swanson & McDowell, 1984; Aguirre-Diaz & Labarthe-Hernandez, 2003; Swanson et al., 61

2006) are inconsistent with modern expressions of supra-subduction zone volcanism. 62

Equally, using narrow, silicic-magma dominated continental arc- to back-arc rifts like the 63

Taupo Volcanic Zone (Cole, 1990; Parson & Wright, 1996) as analogues are also not 64

appropriate (Bryan et al., 2008). 65

66

In contrast, mid-Miocene volcanism (~18-12 Ma) has been reported to be predominantly 67

intermediate in composition, with a restricted occurrence along the margins of the recently 68

opened Gulf of California (Fig. 1). When combined with a facies comparison to 69

stratovolcanic assemblages (Hausback, 1984), this has been the basis for arguing a westward 70

migration and reestablishment of supra-subduction zone arc volcanism along eastern Baja 71

3

California before the final termination of subduction along western Mexico at ~12.3-12.5 Ma 72

(Gastil et al., 1979; Stock & Hodges, 1989; Lonsdale, 1991). This interpretation is significant 73

because: 1) prevailing models for the opening and development of the Gulf of California, 74

despite differing in rifting kinematics, all imply Gulf extension began after the termination of 75

andesitic volcanism and subduction at 12.3 Ma (Stock & Hodges, 1989; Atwater and Stock, 76

1998; Fletcher et al., 2007); and 2) the Gulf of California has been considered an 77

anomalously rapid zone of continental rupture based on the onset of seafloor spreading 78

interpreted to be only ~6-10 Myrs after the cessation of subduction and arc volcanism 79

(Umhoefer, 2011). Such interpretations are predicated on extension beginning after 12 Ma 80

because Middle Miocene andesitic volcanism has been universally considered a nonextending 81

supra-subduction zone volcanic arc. This, however, is at odds with seismic data, well 82

information and paleontological data indicating rift basin development in the Gulf of 83

California began much earlier (see Karig & Jensky, 1972), and now supported by new studies 84

(Ferrari et al., 2012; in preparation) from the Gulf region that demonstrate active extension 85

beginning at least 18 Ma and spatially coincident with the andesitic volcanism. This is 86

consistent with structural studies further east (e.g., Gans, 1997) that indicate extension 87

occurred mainly in the Oligo-Miocene, well before the cessation of subduction (see also 88

Henry, 1989). Several studies have demonstrated how extension can strongly influence 89

magmatism in terms of magma generation (location, processes, rates, storage), and eruptive 90

composition and style (e.g., Gans et al., 1989; Axen et al., 1993), such that intermediate 91

magma compositions can be promoted by active extensional faulting (e.g., Johnson & 92

Grunder, 2000). 93

94

To understand better the origin and tectonic setting of the mid-Miocene andesitic volcanism 95

in western Mexico, it is critical to constrain the spatial-temporal-compositional record and 96

trends of magmatism and how this relates to the timing and location of extension across 97

western Mexico. An important limitation on our understanding of the tectonomagmatic 98

setting of mid- to late Tertiary volcanism in western Mexico has been that all current tectonic 99

and petrogenetic interpretations have relied on “discovery-phase” mapping begun in the 100

1970s, and while at least 90% of the Sierra Madre Occidental (SMO) remains unmapped 101

(Swanson et al., 2006); this has particularly been the case for the western margins of the 102

province through the states of Nayarit and Sinaloa. A recent NSF Margins focus on the 103

4

opening of the Gulf of California has resulted in a closer examination by us of how 104

magmatism transitioned from the huge SMO silicic Large Igneous Province through 105

interpreted arc volcanism to crustal rupturing to open the Gulf of California in the Late 106

Miocene. The examination here will begin by a review of the volcanic record of the SMO 107

silicic Large Igneous Province that immediately preceded the mid-Miocene andesitic 108

volcanism, referred to as the “Comondú “arc” or the middle and upper Comondú Group of 109

Umhoefer et al. (2001). Constraints on the development of the Gulf of California will then be 110

reviewed followed by an examination of the spatial-temporal compositional trends of 111

magmatism across this region from ~40 Ma to better understand the tectonic and structural 112

context of the mid-Miocene andesitic volcanism. 113

114

Geologic Background 115

116

Sierra Madre Occidental 117

118

The Sierra Madre Occidental (SMO, Fig. 1) is the largest silicic igneous province in North 119

America (McDowell & Keizer, 1977; McDowell & Clabaugh, 1979; Ward, 1995; McDowell 120

& McIntosh, 2012), and although principally contained within western Mexico, is contiguous 121

with silicic volcanism through the Basin and Range Province of western USA to the north 122

(Lipman et al., 1972; Gans et al., 1989; Best & Christiansen, 1991) and also with the 123

ignimbrite province of the Sierra Madre Sur, south of the Trans Mexican Volcanic Belt 124

(Martiny et al., 2000; Moran-Zenteno et al., 2007). At least 400,000 km3 of dominantly 125

rhyolitic ignimbrite was erupted, mostly between ~38 and 18 Ma, but age dating over the last 126

40 years has identified two main pulses or ‘flare-ups’ of ignimbrite activity (Fig. 3): at ~34-127

28 Ma and ~24-18 Ma (Ferrari et al., 2002, 2007; Bryan et al., 2008; McDowell & McIntosh, 128

2012). The Oligocene pulse is thought to be responsible for at least three-quarters of this 129

erupted volume, whereas at least 100,000 km3 was erupted in the Early Miocene. 130

131

Rhyolitic to dacitic ignimbrite represents at least 85-90% of the erupted volume with the 132

remaining volume being rhyolitic lavas/domes, basaltic lavas and lesser andesitic lavas. Age 133

dating has revealed brief but intense episodes of volcanism (~1 Myr duration) emplacing 134

kilometre thick sections of ignimbrite across the province (e.g., McDowell & Keizer, 1977; 135

5

Ferrari et al., 2002; Swanson et al., 2006; McDowell & McIntosh, 2012), attesting to rapid 136

rates of silicic magma generation and eruption (Bryan et al., 2008). Ignimbrite sections within 137

the central part of the province are flat-lying and define an unextended core to the SMO (Fig. 138

1), and this region also corresponds to the highest crustal thicknesses, up to 55 km (Bonner 139

and Herrin, 1999). In contrast, ignimbrite sections are faulted and tilted along both the 140

western and eastern flanks of this unextended core (Henry & Aranda-Gomez, 2000; Ferrari et 141

al., 2002). Basaltic andesite lavas, referred to as the Southern Cordilleran Basaltic Andesite, 142

or SCORBA (Cameron et al., 1989), appear to be widespread throughout the SMO, and in the 143

northern SMO, are commonly found intercalated with, or overlying the youngest ignimbrites 144

in each section (Swanson et al., 2006). 145

146

Early Miocene volcanic activity was largely superimposed on the Oligocene pulse of 147

volcanism, except in the northern SMO, but also extended further west (Fig. 1) to be present 148

on Baja California (e.g., Umhoefer et al., 2001). No apparent shift in the eastern limit of 149

volcanism occurred for the central and southern SMO, but a westward shift is more 150

pronounced for the northern SMO (Fig. 1). Recent dredge surveys and age dating of 151

recovered rocks through the southern Gulf of California have confirmed the presence of Early 152

Miocene bimodal volcanic and exhumed intrusive rocks offshore (Fig. 1) improving the pre-153

rift connection between Baja California and mainland Mexico (Orozco-Esquivel et al., 2010; 154

Ferrari et al., 2012; in preparation). The Early Miocene phase shows significant differences 155

from north to south despite showing a similar north-south extent to the previous Oligocene 156

pulse. While silicic volcanism appears to have been more volumetrically dominant in the SW 157

part of the SMO (Ferrari et al., 2002; Bryan et al., 2008), Early Miocene volcanism was less 158

abundant and dominantly mafic in composition across the northern SMO (McDowell et al., 159

1997; Murray et al., 2010, submitted). In the central and southern SMO, volcanism was more 160

bimodal with thick rhyolitic ignimbrite packages, similar to the Oligocene sections, 161

characterising some areas (eg., Espinazo del Diablo and El Salto successions, McDowell & 162

Keizer, 1977), whereas elsewhere, graben-focussed bimodal volcanism is characteristic 163

(Ferrari et al., 2002; Ramos Rosique, 2012). Graben margins in the southern SMO are 164

commonly defined by rhyolite domes, whereas basaltic lava packages up to 200 m thick and 165

rhyolitic ignimbrites partly infill the Early Miocene grabens (Ramos Rosique et al., 2010; 166

Ramos Rosique, 2012). Recent zircon chronochemical studies of the Early Miocene rhyolites 167

6

in the SW SMO have shown a very distinct zircon inheritance signature where the inherited 168

zircon ages indicate remelting of silicic igneous rocks formed during the Oligocene and Early 169

Miocene ignimbrite pulses (Bryan et al., 2008; Ramos Rosique, 2012; Ferrari et al., 2012; in 170

preparation). Mid to upper crustal remelting is interpreted to have been driven by basaltic 171

magmatism invading high structural levels in the crust, which was aided by active crustal 172

extension. 173

174

The Comondú “Arc” 175

176

The eruption of andesitic rocks in the Early to Middle Miocene principally along eastern Baja 177

California (the Comondú arc; Hausback, 1984; Sawlan & Smith, 1984; Sawlan, 1991; 178

Umhoefer et al., 2001) has widely been interpreted to mark the termination of the SMO and 179

its broad zone of silicic dominant magmatism and extension beginning ~40 Ma, with the re-180

establishment of typical supra-subduction zone arc magmatism (e.g., Ferrari et al., 2007). 181

This was a re-establishment because as observed along the length of the western North 182

American margin (McQuarrie & Oskin, 2011), no well-defined frontal ‘andesitic’ arc had 183

existed during the Oligocene, when instead a broad zone (up to 1000 km) of volumetrically 184

dominant silicic volcanism occurred (e.g., Lipman et al., 1972; Gans et al., 1989; McDowell 185

& Mauger, 1994; McDowell, 2007). This was the case for Baja California, which had 186

remained attached to mainland Mexico until the Late Miocene. Along Baja California, 187

Cretaceous rocks are generally overlain by lower Tertiary marine and local nonmarine 188

sedimentary rocks (recording this absence of frontal ‘andesitic’ arc volcanism), or late 189

Oligocene to Miocene volcanic rocks above a regional unconformity (e.g., Beal, 1948; 190

Hausback, 1984; Umhoefer et al., 2001). The Upper Oligocene to Middle Miocene 191

volcanosedimentary units in Baja California are known as the Comondú Group (Umhoefer et 192

al., 2001), and have been widely interpreted to be forearc basin and volcanic arc deposits 193

formed immediately prior to plate boundary reorganization and rifting that opened the Gulf of 194

California (Hausback, 1984; Umhoefer et al., 2001; Conly et al., 2005; Godinez et al., 2010; 195

Umhoefer, 2011). The Comondú Group is currently divided into three informal stratigraphic 196

units (Umhoefer et al., 2001). A lower unit (<500 m thick, ~30 - 19.5 Ma) is dominated by 197

quartz sandstones and conglomerate including aeolian sandstone, and interbedded 198

resedimented pyroclastic units (tuffaceous sandstone), whereas several rhyolitic ignimbrites 199

7

and localised basaltic lavas are prominent in the upper parts. Detrital zircon U-Pb ages on 200

aeolian sandstones suggest a maximum depositional age of ~25 Ma (Godinez et al., 2010), 201

and new U/Pb and Ar/Ar ages bracket ignimbrite emplacement between ~24 and 19.50 ± 0.05 202

Ma (Drake, 2005; Godinez et al., 2010). This lower unit is thus temporally and 203

compositionally correlated with Early Miocene bimodal volcanism offshore in the Gulf 204

(Orozco-Esquivel et al., 2010) and on mainland Mexico. These Comondú Group ignimbrites 205

are outflow facies and generally deposited at medial to distal distances from their sources 206

interpreted to occur to the east (Umhoefer et al., 2001; Drake, 2005). The middle and upper 207

‘Comondú’ units (~19.5 - 12 Ma; ~1 km thick) are more intermediate in composition and 208

dominated by sedimentary breccia with interspersed andesite-dacite lavas/domes, particularly 209

in southern Baja California, and cross-cut by dykes and minor porphyry intrusions (Sawlan & 210

Smith, 1984; Hausback, 1984;Umhoefer et al., 2001; Godinez et al., 2010). A few 211

interbedded rhyolitic ignimbrites have also been reported (Hausback, 1984; ash flow tuff D of 212

Drake, 2005). Some zircon U-Pb ages on biotite granodiorite porphyry intrusions indicate 213

emplacement ages of 19.9 ± 0.73 and 16.3 ± 0.49 Ma (Godinez et al., 2010), whereas 214

andesite-dacite lavas range in age from ~19.5 to ~11 Ma (Sawlan & Smith, 1984; Hausback, 215

1984; Martín-Barajas et al., 1995; Martín et al., 2000; Umhoefer et al., 2001; Drake, 2005). 216

The lava dome and flow units are vent to proximal facies, but critically, several stratigraphic 217

studies have shown the andesite-dacite lavas to be interbedded with sedimentary rocks rather 218

than forming thick stacked piles of lava as expected in the construction of stratocones. 219

220

Gulf of California 221

222

The Gulf of California (GoC) has been one of the focus sites for the Rupturing Continental 223

Lithosphere initiative of the NSF-funded MARGINS program from 2004-2010. It is a ~1400 224

km long, highly sedimented, oblique rift characterized by long transform faults and short 225

spreading centres (Lonsdale, 1989; Lizarralde et al., 2007). Despite different models (see 226

review in Fletcher et al., 2007), rifting has been considered to have developed rapidly 227

following cessation of subduction of the Guadalupe Plate (and arc volcanism) at about ~12.3-228

12.5 Ma (Stock & Hodges, 1989; Ferrari et al., 2007; Fletcher et al., 2007; Lizarralde et al., 229

2007; Umhoefer, 2011). Since rift inception, the Gulf has opened between ~300 and 500 km 230

(Oskin et al., 2001; Fletcher et al., 2007), with the Gulf progressively unzippering to the north 231

8

such that active extension of continental crust is currently occurring at the northern end of the 232

Gulf. The timing of Gulf inception and rifting at ~12.3 Ma is generally consistent with the 233

switch to more alkaline and tholeiitic basaltic volcanism, and the reappearance of more 234

rhyolitic volcanism at ~12-13 Ma (Fig. 3; e.g.., Martín-Barajas et al., 1995; Stock et al., 1999; 235

Martín et al., 2000; Vidal-Solano et al., 2007). Most dates for initiation of marine 236

sedimentation in the Gulf region are at 8 Ma and younger (e.g., Oskin & Stock, 2003), but an 237

earlier marine incursion and the development of a seaway in the middle Miocene has been the 238

basis for the concept of a “Proto-gulf” (Karig & Jensky, 1972). While the timing of marine 239

incursion remains contentious and based principally on onshore exposures, comprehensive 240

stratigraphic revision (Carreño & Smith, 2007) and recent micropaleontological studies of 241

several deep wells drilled in the Wagner, Consag, and Tiburón basins of the GoC document 242

at least 1870 m of marine sediments deposited before 11.2 Ma (Helenes et al., 2009; Helenes, 243

2012). 244

245

Seafloor-spreading centres formed at variable times from ~6 to 2 Ma (Umhoefer, 2011), with 246

a spreading centre in the Guaymas sub-basin (central Gulf) interpreted to have began ~ 6 Ma 247

(Fig. 1) based on the width of the new igneous crust observed in seismic refraction profiles 248

(Lizarralde et al., 2007). The Alarcón spreading center began forming proto-oceanic crust ca. 249

3–3.5 Ma (DeMets, 1995), and true sea-floor spreading at present rates began at 2.4 Ma 250

(Sutherland, 2006; Umhoefer et al., 2008). 251

252

Summary 253

254

Two important outcomes on the Miocene history of the GoC region are apparent from the 255

diversity of recent studies in the region. The first is that recent dating studies from the SMO, 256

the GoC and Baja California confirm eruptive ages on rhyolites and basalts related to the 257

Early Miocene pulse of the SMO extend to as young as ~17 Ma (Hausback, 1984; Martín et 258

al., 2000; Umhoefer et al., 2001; Drake, 2005; Bryan et al., 2008; Ramos Rosique, 2012; 259

Ferrari et al., 2012; in preparation). This continued bimodal volcanism thus overlaps the onset 260

of interpreted arc volcanism along Baja California at ~19.5 Ma (Umhoefer et al., 2001), 261

whereas others have considered ‘arc’ volcanism began earlier in northern Baja California at 262

~21 Ma (e.g., Martín-Barajas et al., 1995). This age overlap thus suggests no abrupt 263

9

termination to SMO bimodal volcanism when rejuvenation of supra-subduction zone arc 264

volcanism was apparently initiated, despite the bimodal and andesitic volcanism spatially 265

overlapping (Fig. 1). However, regional temporal-compositional patterns in erupted magma 266

compositions do indicate a strong compositional shift from dominant bimodal volcanism to 267

more intermediate composition volcanism beginning ~19 Ma (Figs 3, 4). The second 268

outcome is that biostratigraphic studies, particularly from the offshore basins are recording a 269

more protracted history of rift basin subsidence and sediment accumulation in the GoC and 270

that marine sedimentation began at least in the middle Miocene (~12 Ma; Helenes, 2012; cf. 271

Sutherland et al., 2012). This would require unrealistically rapid crustal thinning and 272

subsidence if all extension began ~12 Ma. Collectively, these observations are at odds with 273

the notion of rapid switchings from bimodal volcanism across the western SMO and southern 274

Baja California, to establishing a subduction-related andesitic magmatic arc along the eastern 275

side of Baja California and westernmost mainland Mexico from ~22-19 Ma, and which then 276

abruptly terminated at 13-12 Ma to switch back again to predominantly bimodal alkaline 277

volcanism (Fig. 3), coincident with the onset of rifting of the GoC at this time (e.g., Stock & 278

Hodges, 1989; Umhoefer et al., 2001; Sutherland et al., 2012). 279

280

Temporal-compositional trends of Oligocene-Miocene magmatism across western 281

Mexico 282

283

Published and unpublished whole-rock compositional and age data for 40-0 Ma igneous rocks 284

from Western Mexico, occurring in a northwest-trending belt from ~22°N 103°W, to 32°N 285

116°W have been compiled in this study (see Supplementary Material) and serves as an 286

update to that of Ferrari et al. (1999). The dataset comprises three types of data: 1) whole-287

rock compositional data for which radiometric ages have been obtained from the same 288

samples (n = 461); 2) whole-rock compositional data that have stratigraphic control and a 289

stratigraphic age can be assigned (e.g., Oligocene, Late Miocene; n = 1567); and 3) samples 290

that have radiometric age data and lithological information (e.g., basalt lava, rhyolite 291

ignimbrite), but may lack corresponding whole-rock compositions (n = 1520). For samples 292

with radiometric ages from 40-0 Ma, this includes K/Ar (whole rock, groundmass, feldspar, 293

biotite, hornblende), 40Ar/39Ar (whole rock, groundmass, feldspar, hornblende, biotite), and 294

zircon fission track and U/Pb ages. Of the samples with whole-rock compositional data, 1486 295

10

samples have major element analyses with the remainder having whole-rock isotopic (Sr, Pb, 296

Nd) compositions ± some trace element abundances. Overall, trace element data are more 297

sparse with only 288 samples having full rare earth element analyses. Most of the whole-rock 298

geochemical data have been measured by XRF, but ICP-MS trace element data have 299

increased over the last 10 years (e.g., Martín et al., 2000; Till et al., 2009). In the following 300

figures, igneous compositions are grouped into the following: 1) basalts, with 45-52 wt% 301

SiO2; 2) basaltic andesite, 52-56 wt% SiO2; 3) andesite, 56-63 wt% SiO2; 4) dacite, 63-68 302

wt% SiO2; and 5) rhyolite, 68-79 wt% SiO2. 303

304

Compiled whole-rock compositional data from the SMO for which age data are available 305

(Figs 3, 4) illustrate the dominantly silicic compositional character of the Oligocene 306

ignimbrite pulse, but with an increasing appearance and abundance of basaltic lavas from ~30 307

Ma. In contrast, the Early Miocene pulse was distinctly bimodal with a paucity of volcanic 308

compositions between 64 and 68 wt% SiO2 (Fig. 4). Of note is the broad silicic peak for both 309

the Oligocene and Early Miocene pulses or two compositional groupings of dacitic to low-310

silica rhyolite and high-silica rhyolite suites with the latter being dominant and occurring as 311

lavas/domes and ignimbrites. 312

313

The disappearance of rhyolite and high-silica rhyolite compositions across the region is 314

pronounced at ~19-18 Ma (Fig. 3). From 18-13 Ma, erupted magmas were more intermediate 315

in composition as recognised in many previous studies, with a paucity of both basalt (<52 316

wt% SiO2) and rhyolitic (>68 wt% SiO2) magmas. Basaltic andesite and low-silica andesite 317

compositions dominate with andesite-dacite compositions also important (Fig. 4), and almost 318

all have been emplaced as lavas. Consequently, there is not only a strong compositional 319

change at 19-18 Ma, but also a change from mixed explosive-effusive to dominantly effusive 320

eruptive styles. However, the period of dominant andesitic volcanism (~18-13 Ma) that 321

defines the basis for a subduction-related andesitic magmatic arc is actually much shorter 322

than is widely argued (Fig. 3); andesitic arc volcanism has previously been interpreted from 323

~22-16 Ma and 22-12 Ma across northern and southern Baja California, respectively (Sawlan 324

& Smith, 1984; Hausback, 1984; Sawlan, 1991). The reappearance of rhyolite and high-silica 325

rhyolite at 13-12 Ma (Fig. 3) is distinctive because these magma compositions are also 326

peralkaline and mark the recurrence of explosive eruptions emplacing ignimbrites (Vidal-327

11

Solano et al., 2008) such as the 12.6 Ma alkali rhyolite Tuff of San Felipe (Stock et al., 1999). 328

The alkali rhyolite eruptions coincide with the eruption of true basaltic compositions 329

regionally that are also more alkaline in composition. From 13-5 Ma (Fig. 4) magma 330

compositions became fundamentally more basaltic and alkalic with alkalic and tholeiitic 331

basaltic eruptions around the margins of the GoC eventually coinciding with the onset of 332

seafloor-spreading from 6-3.5 Ma. As with the preceding Miocene phase, dacitic 333

compositions (64-68 wt % SiO2) are significant between 12 and 8 Ma (Figs 3, 4). 334

335

Spatial-Temporal-Compositional Trends 336

337

Important insights can also be gained if the long-term temporal compositional trends 338

discussed above can be placed in a spatial context. The age dataset for which lithological ± 339

whole-rock compositional data are available is used here. Age data are plotted for E-W 340

transects across the GoC for the region between 20° and 26°N and 26° and 32°N as shown in 341

Figure 5. Dated igneous rocks have been grouped into four compositional groupings: basalt 342

(<56 wt% SiO2), andesite (56-63 wt% SiO2), dacite (63-68 wt% SiO2) and rhyolite (>68 wt% 343

SiO2). Several key features are apparent from Figure 5: 344

345

1. As observed in Figures 3 and 4, Oligocene to Early Miocene (~40-20 Ma) volcanism is 346

silicic-dominant, and widely distributed (up to 800 km width). Across this broad zone of 347

silicic volcanism between 40-30 Ma, low volumes of basalts and andesites were also erupted. 348

349

2. By the Early Miocene, volcanism became strongly bimodal, but this switch occurred 350

earlier in the northern region (~30 Ma), and basaltic magmas dominated from 30-20 Ma. In 351

the north (Sonora-Chihuahua), basalts (SCORBA of Cameron et al., 1989) commonly marked 352

the termination of silicic volcanism in an area and since ~28 Ma were emplaced on top of 353

rhyolitic ignimbrites and into developing extensional basins (until ~15 Ma; McDowell et al., 354

1997). In the south (Jalisco, Nayarit, Sinaloa), the basalts appear to represent the triggering of 355

the Early Miocene silicic ignimbrite pulse (Bryan et al., 2008). 356

357

3. Rather than a westward migration of volcanic activity as is widely reported from the 358

Oligocene to Miocene (e.g., Damon et al., 1981; Ferrari et al., 1999), volcanism did extend 359

12

westwards at ~30 Ma, but importantly, there was no corresponding shift of the eastern limit 360

of volcanism at this time, which would be expected if volcanic activity was intimately linked 361

to back-arc extensional processes and driven by slab roll back (e.g., Clift et al., 1995). A 362

westward shift in the location of the eastern limit of volcanism at ~25 Ma is more pronounced 363

in the northern region, and did not occur until ~20 Ma in the southern SMO. 364

365

4. The extensive Early Miocene bimodal volcanism was associated with distributed extension 366

and the formation of numerous grabens along the western and southern SMO (Fig. 1). 367

Inception of most grabens had begun by 24 Ma in the southern SMO (Ferrari et al., 2002; 368

Ramos Rosique, 2012) and by 27 Ma across the northern and western SMO (McDowell et al., 369

1997). Graben development occurred over a horizontal distance of up to 450 km (Fig. 1) and 370

at least 600 km in board of the paleocontinental margin at that time. Despite the size and 371

extent of the grabens, the magnitude of Latest Oligocene-Early Miocene extension associated 372

with them is inferred to be modest (<10% stretching, Ferrari et al., 2002). Soon after the 373

onset of this wide zone of rifting, a relatively narrow belt of high magnitude extension 374

producing metamorphic core complexes from ~25-15 Ma (Nourse et al., 1994; Wong et al., 375

2010) also developed across the NW SMO (Fig. 1). The prominent westward shift in volcanic 376

activity at ~25 Ma and subdued silicic volcanism from ~27 Ma for the northern region may in 377

part reflect suppression of volcanism in areas of high magnitude extension (Gans & Bohrson, 378

1998) that was occurring along the eastern limits of this volcanism. 379

380

5. In both regions, andesite-dacite volcanism becomes more abundant beginning ~19 Ma (see 381

also Fig. 3), and in contrast to earlier episodes of andesite eruption, are much more spatially 382

restricted to the location of the nascent GoC. 383

384

6. Although andesite-dacite volcanism was significant from ~20 to 10 Ma, as observed along 385

Baja California, it was coeval with (basalt-dominant) bimodal volcanism occurring 386

regionally. This is particularly the case during the period 20-16 Ma for the southern region. 387

Rhyolitic volcanism producing ignimbrites and lavas persisted until at least 16 Ma in the 388

southern region with rhyolites dated from Baja California (Hausback, 1984; Hosack, 2006), 389

Nayarit (Ferrari et al.2012; in preparation) and Jalisco (Nieto-Obregón et al., 1981; Bryan et 390

al., 2008). Consequently, spatial-temporal distinction between the termination of syn-391

13

extensional bimodal volcanism across the SMO and the beginnings of more intermediate 392

composition magmatism around the GoC is blurred. 393

394

7. The style, composition and extent of volcanism changes significantly at ~13 Ma. Bimodal 395

volcanism characterises the northern region (coastal Sonora and Baja California) and 396

rhyolitic compositions became more peralkaline (e.g., Tuff of San Felipe, Stock et al., 1999; 397

Vidal-Solano et al., 2007). In contrast, volcanism is more basaltic and widespread in the 398

southern region with few rhyolites dated younger than 13 Ma. Despite the widespread 399

eruption of basalts across the southern region, andesite-dacite volcanism remains principally 400

focussed around the margins of the nascent GoC. Andesitic compositions are particularly 401

spatially restricted in the northern region from 10-2 Ma despite other magma compositions 402

being erupted over a broader region (Figs 4,5). 403

404

Timing Relationships to Extension 405

406

An important constraint in understanding the significance of the spatial-temporal 407

compositional trends described above is understanding how these magmatic trends 408

correspond spatially and temporally to the onset, or changes to the rate and breadth, of 409

extension. Most of the SMO has been affected by different episodes of dominantly 410

extensional deformation that began in the Oligocene and potentially at the end of Eocene 411

(Henry, 1989; Ferrari et al., 2007). However, the central zone to the SMO remains unaffected 412

by extension (Fig. 1) where ignimbrite sections rise to high altitudes (>3000 m ASL), remain 413

flat-lying (Fig. 6) and crustal thicknesses are high (Bonner and Herrin, 1999). This 414

unextended core of the SMO represents a physiographic boundary between what has been 415

defined as the “Mexican Basin and Range,” to the east, and the “Gulf Extensional Province,” 416

to the west (Henry & Aranda-Gómez, 2000). At the northern and southern ends of the SMO 417

(i.e., northern Sonora and Chihuahua and Nayarit-Jalisco, respectively), these two provinces 418

merge where extension has affected the entire width of the SMO (Ferrari et al., 2007). 419

420

The age of extension in the Gulf Extensional Province is thought to have been largely Late 421

Miocene to Recent and associated with the termination of subduction and onset of rifting to 422

open the GoC at ~12.3 Ma (e.g., Stock & Hodges, 1989; Henry & Aranda-Gomez, 2000). 423

14

However, an intense phase of extension resulting in up to 100% crustal thinning and 424

producing metamorphic core complexes occurred from ~25-15 Ma in Sonora (e.g., Nourse et 425

al., 1994; Gans, 1997; Gans et al., 2003; Wong et al., 2010). These core complexes slightly 426

post-date the onset of a milder phase of extension producing sediment and basalt-filled 427

graben depressions (Baucarit Formation of King, 1939) through Sonora. Basaltic to andesitic 428

lavas ranging in age between 27 and 20 Ma are common toward the base of these rift basin 429

successions (McDowell et al., 1997; Paz-Moreno et al., 2003). 430

431

The age of extension along the southeastern margin of the GoC through Sinaloa and Nayarit 432

has been less clear. Recent studies by us (Ferrari et al., 2012; Ferrari et al., 2012; in 433

preparation) have revealed that kilometre-thick ignimbrite sections of the SMO, dated as 434

young as 20 Ma, have been tilted by up to 35° (Fig. 6) with some blocks remaining at high 435

elevations (~2000 m). However, along the low-relief coastal plain are extensive, flat-lying 436

and undeformed basaltic lava fields dated between 12-9 Ma extending for at least 700 km 437

along the eastern margin of the GoC (Fig. 1; Ferrari et al., 2012; in preparation; see also 438

Gastil et al., 1979). Similar aged basalts have also been dredged from the submerged 439

continental margins to the southern GoC (Ferrari et al., 2012; in preparation). 440

441

Important insights also come from the Baja California sections of the Comondú Group where 442

interpreted vent and proximal facies lava flows and domes are buried by thick successions of 443

medial facies bedded sedimentary conglomerate/breccia and sandstone, and distal ignimbrites 444

(e.g., Hausback, 1984; Umhoefer et al., 2001; Drake, 2005). In particular, positive relief lava 445

domes with thicknesses of up to 300 m have been shown to be buried by sedimentary 446

deposits (Drake, 2005). These depositional and facies relationships demonstrate that andesitic 447

to dacitic lavas and domes of the Comondú Group were being emplaced into actively 448

subsiding sedimentary basins. 449

450

In summary, field, stratigraphic and now new age data from the eastern margin of the GoC 451

indicate that large-magnitude extension must have occurred between ~25 and 12 Ma to the 452

north (e.g., Gans, 1997) and between ~18 and 12 Ma in the central and southern Gulf region. 453

This extension must have post-dated the final phases of bimodal and ignimbrite-dominant 454

activity of the Early Miocene pulse of the SMO, and preceded the widespread eruption of 455

15

flat-lying, (undeformed) basaltic lavas along the eastern margin of the GoC beginning ~12 456

Ma (Fig. 1). Importantly, as recognised in previous studies, the present-day crustal thickness 457

beneath the unextended core of the SMO is between 40 and 55 km, whereas along the 458

margins of the GoC it is between 18 and 26 km (Gomberg et al., 1988; Couch et al. 1991; 459

Bonner & Herrin, 1999; Persaud et al., 2006; 2007). Consequently, most of this crustal 460

thinning occurred prior to the termination of subduction at ~12.3-12.5 Ma, and was 461

coincident with the major period of andesitic volcanism (middle and upper Comondú Group 462

of Umhoefer et al., 2001). 463

464

Discussion 465

466

Origin of the Middle Miocene Comondú Andesites 467

468

In determining the origin of the Comondú Group andesites and in particular, why there was a 469

discrete period of dominantly intermediate volcanism between ~18-13 Ma, several key points 470

must be considered from the observations and discussion given above. First, there has been 471

no preceding history of andesitic volcanism or existence of a well-defined frontal arc in the 472

region since at least the Eocene, despite ongoing subduction. Andesitic volcanism occurred, 473

but 1) was diffuse and low-volume occurring over a very broad width in-board of the margin 474

(>600 km) where it was closely associated with silicic and basaltic volcanism (Fig. 5); 2) 475

when eruptions of andesitic-dacitic magma compositions became dominant in the mid-476

Miocene, they were localised in space and time around the nascent GoC (Figs 1, 5); 3) this 477

andesitic phase was coincident with the major period of crustal extension through the Gulf 478

Extensional Province with crustal thinning up to 100%; and 4) both immediately before and 479

after, and regionally, volcanism was clearly bimodal in character (Fig. 3). 480

481

Limited petrologic and geochemical data have been published on the Comondú Group 482

andesites and much interpretation has been based on major element chemistry (Gastil et al., 483

1979; Sawlan & Smith, 1984; Hausback, 1984). These previous studies have established the 484

general ‘calc-alkaline’ chemistry, intermediate composition and similarity to modern 485

subduction-related arc successions for the Comondú Group andesites. However, alkalic 486

basalts have also been reported from these suites (Martín-Barajas et al., 1995), which are 487

16

atypical of modern, unextended arc settings. The Comondú Group andesites are commonly, 488

moderately to very crystal-rich and vary from hornblende to pyroxene andesites (e.g., Sawlan 489

& Smith, 1984; Martín et al., 2000). However, magmatic heterogeneity of these andesites is 490

apparent from the brief petrographic descriptions with reports of resorbed olivine (Sawlan & 491

Smith, 1984), quartz and alkali feldspar (Martín et al., 2000); complexly zoned plagioclase 492

(Sawlan & Smith, 1984); megacrystic hornblende (Drake, 2005); enclaves (Drake, 2005); and 493

granitic and upper crustal xenoliths (Martín et al., 2000). Recent U-Pb zircon dating of a 494

Miocene porphyry intrusion that intrudes the Lower Comondú Group (Godinez et al., 2010) 495

has revealed the presence of abundant Cretaceous zircons indicating remelting of the 496

underlying calc-alkaline Cretaceous batholith. Collectively, these characteristics indicate the 497

importance of open system processes, magma mixing (involving basaltic magmas) and 498

crustal assimilation/partial melting in Comondú Group andesite genesis. Importantly, the 499

crustal materials involved have a strong subduction heritage given the long-lived history of 500

subduction along the western margin of North America since the Triassic. Therefore no 501

reliance should be placed on the whole-rock geochemical signatures for interpreting the 502

tectonic environment in which the magmas were emplaced where crustal assimilation has 503

demonstrably occurred (cf. Conly et al., 2005). Furthermore, recent studies of arc andesites 504

are emphasising that most if not all crystals are xenocrysts or antecrysts, being picked up by 505

initially aphyric mafic melts on their way to the surface (Zellmer et al., submitted) such that 506

the bulk rock compositions will incorrectly image the mantle source, and prove unreliable for 507

tectonic discrimination. Consequently, porphyritic intermediate compositions likely represent 508

remobilized igneous protoliths (Zellmer, 2009), and much of their bulk chemical 509

characteristics can be inherited from those source materials. As has been discussed elsewhere, 510

there is no a priori requirement for I-type calc-alkaline magmatism to be directly related to 511

active subduction processes (e.g., Roberts & Clemens, 1993; Morris & Hooper, 1997). 512

513

Mineralogical evidence suggests the importance of magma mixing being an important 514

process. Previous studies have illustrated linear variation trends for major elements (e.g., 515

Sawlan & Smith, 1984; Hausback, 1984). Plotting of several trace element ratio combinations 516

also reveals linear patterns and apparent mixing trends (Fig. 7). The role of mixing is a 517

dominant process at the regional scale and the existence of a subduction-related signature is 518

examined in Figure 7. The use of Ba/Nb as a subduction-related signature is similar to the 519

17

approach of Till et al. (2009) who used Ba/Ta, but because there are few samples with Ta 520

abundances, we use the Ba/Nb ratio here. However, elevated Ba/Nb ratios are also a crustal 521

signature. The Ba/Nb-La/Nb data show a spectrum from “intraplate” basaltic compositions, 522

as represented by the Early and Late Miocene alkali basalts in the region, to more arc-like 523

and crustal compositions with high Ba/Nb ratios. The trends for the different compositional 524

groupings are not readily explainable by fractional crystallisation processes since the mafic, 525

intermediate and silicic igneous compositions overlap. Of note is that from the Early Miocene 526

onwards, a distinctly high Ba/Nb suite of rocks were emplaced (Ba/Nb>150), which are 527

absent from the Eocene and Oligocene suites. These high Ba/Nb and La/Nb rocks are almost 528

entirely small-volume andesite-dacite lavas geographically restricted to the northern GoC 529

(western Sonora, and Baja California); some published whole-rock isotopic data (e.g., Mora-530

Klepeis & McDowell, 2004) indicate relatively radiogenic Sr (87Sr/86Sr = 0.70564) and 531

unradiogenic Nd (143Nd/144Nd = 0.51258) compositions compared to other volcanics in the 532

region. The high Ba/Nb ratios and Sr and Nd isotopic compositions are most similar to 533

Paleozoic to Proterozoic silicic orthogneisses from north central Chihuahua (Cameron et al., 534

1992). We raise the possibility that these high Ba/Nb intermediate composition rocks 535

emplaced since the Early Miocene are small-volume crustal partial melts of potentially 536

Paleozoic-Proterozoic crustal materials along the southwestern edge of the North American 537

craton. 538

539

We therefore conclude that the dominant period of andesitic volcanism focussed around the 540

margins of the GoC, is at the regional scale, the product of magma mixing. Our model 541

relieves issues around the apparent lack of geochemical discrimination of interpreted syn- and 542

post-subduction (<14 Ma) magmatism in the region (Till et al., 2009) as the entire Oligocene-543

Miocene igneous record was unrelated to subduction. Evidence for this comes from 544

uncontaminated Early Miocene basalts that show within-plate trace element compositions 545

(Fig. 7). A similar asthenospheric fingerprint has also been recognised by Cameron et al. 546

(1989) for Oligocene basaltic andesites (SCORBA) erupted across the northern SMO. Based 547

on the chemical variation within the mid-Miocene (~18-13 Ma) suite, magma end-members 548

are interpreted to be: 1) an asthenosphere-derived, within-plate alkali basalt similar to those 549

erupted in the region during the Late Miocene to Quaternary (e.g., Luhr et al., 1995; Pallares 550

et al., 2007), 2) crustal partial melts derived from Mesozoic-Cenozoic igneous crust, 551

18

including the Jurassic-Cretaceous Peninsular Ranges Batholith (e.g., Silver & Chappell, 552

1998) and Eocene-Oligocene batholithic rocks, as informed by inherited zircon ages (Bryan 553

et al., 2008; Godinez et al., 2010; Ramos Rosique, 2012); and 3) small-volume crustal partial 554

melts of dacitic composition derived from Proterozoic-Paleozoic crust (highlighted by the 555

extreme Ba/Nb ratios, Fig. 7). Importantly, Figure 7 illustrates no apparent temporal change 556

in basaltic magma composition from the Early to Late Miocene. Basaltic rocks with more 557

“arc-like” signatures are interpreted to be either crust/lithospheric mantle contaminated, or 558

possibly derived from a previously subduction-modified lithospheric mantle, but where 559

lithospheric mantle melting was in response to basaltic magma input from the asthenosphere 560

and decompression due to lithospheric thinning. 561

562

How was mixing promoted? We believe that extensional faulting causing the large-563

magnitude crustal thinning during the Middle Miocene began actively disrupting bimodal 564

magma systems in the GoC region (Fig. 8). Magma inputs (i.e. basaltic and rhyolitic magmas 565

erupted during the Early Miocene) had not fundamentally changed, and regionally, away 566

from the Gulf, basaltic and rhyolitic magmas generally continued to erupt (Fig. 5). However, 567

locally around the GoC where extension was now being focussed and was greatest, this 568

tectonic influence had a major effect on the composition of magmas being erupted. 569

Consequently, there was a spatial-temporal focussing of andesitic volcanism along the 570

nascent GoC rift. Importantly, this switch from widespread Early Miocene bimodal 571

volcanism to more localised andesitic volcanism during the Middle Miocene (Fig. 8) appears 572

to correspond to a switch from wide to narrow rifting beginning ~18 Ma. This switch had an 573

important effect on silicic magma generation rates, which appears to have significantly 574

decreased at ~19 Ma (Fig. 3) as mafic magma inputs to the crust became more focussed in the 575

Gulf region. The Early Miocene grabens that had developed across a ~500 km width of 576

western Mexico and which had significant volcanic fills, became magmatically abandoned by 577

~18 Ma (e.g., Bolaños Graben, Ramos Rosique, 2012), and continued magmatism was now 578

largely restricted to the GoC and its immediate margins (Fig. 8). The spatial-temporal-579

compositional patterns of magmatism through the Miocene of western Mexico are thus 580

providing us with an important record of changes to the extensional style, intensity and 581

location. 582

583

19

Active extensional faulting, particularly during the period 18-13 Ma, modified the erupted 584

magma compositions that became more intermediate in composition. Concomitant with this 585

compositional change was a change in eruption styles that were dominantly effusive, 586

producing scattered lavas and domes as well as dyking. Eruption location shifted to be 587

concentrated around the nascent GoC rift. We therefore conclude the Comondú Group 588

andesites are a larger-scale regional example of syn-volcanic extensional fault-driven magma 589

mixing processes such as described by Johnson & Grunder (2000). 590

591

Concluding Remarks 592

593

Magmatism is used widely as an instant tracer of the geodynamic setting, and magma 594

composition as a proxy for a specific tectonic setting (e.g., arc/supra-subduction zone, slab 595

melting, slab tearing, crustal rifting). In particular, calc-alkaline chemistries ± relative Nb 596

depletions and blind use of tectonic discrimination diagrams often form the basis for 597

interpreting an active subduction-related setting (e.g., Zhu et al., 2012). However, subduction 598

leaves a strong chemical imprint on the lithosphere such that subduction-related geochemical 599

signatures can persist for 100's of millions of years (e.g., Morris & Hooper, 1997) and stamp 600

the chemistry of new magmas produced at much younger times and unrelated or far removed 601

from any active subduction zone. The low-Ti flood basalts of the Karoo and Paraná-Etendeka 602

LIPs are classic examples of this (Duncan, 1987; Ewart et al., 1998). 603

604

Our study demonstrates the importance of understanding the regional tectonic setting both at 605

the time of volcanism and for the preceding and following magmatic history in interpreting 606

the origin of the middle Miocene Comondú Group andesites. As shown in many previous 607

studies, when considered in isolation and focussing only on the general chemical 608

characteristics, these andesitic rocks do resemble the products of a supra-subduction zone arc 609

(e.g., Hausback, 1984; Till et al., 2009). However, the andesites are spatially restricted to the 610

nascent GoC rift, there was no preceding history of andesitic arc volcanism in this region, the 611

region was clearly undergoing wide rifting and bimodal volcanism immediately beforehand, 612

and their emplacement coincides in space and time with focussed rifting and large magnitude 613

(~100%) crustal thinning. The stratigraphic relationships and architecture of Middle Miocene 614

andesite-dacite lavas and domes (proximal and vent volcanic facies) being buried by thick 615

20

successions of syn-volcanic sedimentary deposits including sandstones demonstrates lava 616

emplacement into actively subsiding rift basins. This facies architecture is more consistent 617

with biostratigraphic and paleontological evidence for earlier basin rifting and subsidence and 618

a marine incursion into the GoC by 12 Ma (Helenes, 2012). Andesite-dacite lavas and domes 619

are subordinate to the volcanic-derived sedimentary rocks and isolated to scattered 620

lavas/domes thus should not be used to define an arc. We extend the recent arguments by 621

Castillo (2008) that the compositionally diverse suite of mafic rocks including adakitic and 622

high-Nb basalts erupted in Baja California are unrelated to subduction or slab melting in the 623

Late Miocene, but that asthenospheric mantle involvement in regional magmatism occurred 624

much earlier, being prevalent in the Oligocene (Cameron et al., 1989) and Early Miocene 625

(Fig. 7). 626

627

Our fundamental conclusion is that the Comondú Group does not represent a supra-628

subduction zone arc and this challenges all previous studies on the Comondú Group. We 629

caution that andesites in ancient extended continental margins may bear no relationship to 630

active subduction and magma chemistry cannot be solely relied on to interpret tectonic 631

setting. Alternatively, as shown here, the andesitic volcanism and the switch from widespread 632

bimodal to more focussed andesitic volcanism can potentially provide much more important 633

insights into the extensional history of a continental margin. 634

635

Finally, our study raises questions about genetic links between the Sierra Madre Occidental 636

silicic volcanism and the Gulf of California, as these have previously been considered two 637

separate phenomena. We suggest the Sierra Madre Occidental Silicic LIP is the pre- to syn-638

rift volcanic event to the GoC rift. This is because from ~30 Ma and certainly by 24 Ma, 639

SMO volcanism and extension were overlapping spatially and temporally. It therefore raises 640

the question of when did rupturing of the Gulf of California really begin? Was it at the onset 641

of wide rifting and bimodal volcanism at ~30-24 Ma, or at the onset of narrow rifting at ~18 642

Ma? Regardless, the new regional timing and stratigraphic-structural relationships evident 643

along the eastern margin of the GoC show that the onset of rifting to open the GoC was 644

clearly older than the ~12.3 Ma end of subduction along this segment of the western North 645

American margin. Consequently, the GoC has had a much longer history of rifting and 646

21

associated magmatism before the onset of seafloor-spreading, and may not be such an 647

anomalously rapid zone of crustal rupture (cf. Umhoefer, 2011). 648

649

Acknowledgements 650

651

S.B. has been supported by a Vice Chancellor’s Research Fellowship from QUT. We 652

acknowledge support of grant CONACyT 82378 to LF. The submarine samples shown on 653

Figure 1 were collected on cruises supported by the US NSF (grants 0203348 and 0646563 to 654

co-PIs Peter Lonsdale and Paterno Castillo), as well as grants to Peter Lonsdale and Jared 655

Kluesner for the BEKL, ROCA and DANA cruises in the Gulf of California. Discussions 656

with John Clemens, Georg Zellmer and Christoph Schrank on aspects of this manuscript are 657

appreciated. We thank Jim Cole and Paterno Castillo and Susan Straub for constructive 658

reviews of this manuscript. 659

660

22

661

662

Figure Captions 663

664

Figure 1. Tectonic map of northwestern Mexico showing the lithospheric variation across the 665

region including unextended and extended continental regions, and transitional to new 666

oceanic crust formed by the propagating spreading centre in the Gulf of California. 667

Superimposed on this tectonic map are the preserved extents of the Oligocene-Early Miocene 668

silicic-dominant volcanic activity of the SMO (Ferrari et al., 2002; Bryan et al., 2008), and 669

the dominantly bimodal phase during the Early Miocene that coincided with the wide 670

development of grabens and rift basins (McDowell et al., 1997; Ferrari et al., 2002) and a 671

restricted belt of metamorphic core complexes in the state of Sonora (Nourse et al., 1994; 672

Wong et al., 2010). Distribution of Comondú Group andesites from Umhoefer et al. (2001). 673

Offshore Miocene igneous rocks from Ferrari et al. (2012; in preparation). Rift basin 674

segments to the Gulf of California are labelled; Car., Carmen, and Pesc., Pescadero basins. 675

NAY., Nayarit; EPR, East Pacific Rise; H., Hermosillo. Red boxed areas near Mazatlán and 676

Tepic refer to locations of photographs in Figure 6. 677

678

Figure 2. Total alkali-silica diagram (TAS; Le Maitre et al., 1989) for Oligocene-Miocene 679

(~34-13 Ma) igneous rocks from NW Mexico, plotted on an anhydrous basis, showing the 680

spectrum of compositions generated during this broad interval. The field of the Southern 681

cordillera basaltic andesites (SCORBA; Cameron et al., 1989) is from McDowell et al. 682

(1997). Diagram contains 606 analyses. 683

684

Figure 3. Probability density plot of igneous ages from western Mexico for the period 40-0 685

Ma. Dated rocks have been grouped into four main compositional groupings - basalt 686

(includes basaltic andesites and tholeiitic, calc-alkaline and alkaline varieties), andesite 687

(includes calc-alkaline and adakitic-type compositions), dacite and rhyolite (includes high-688

silica rhyolites and peralkaline compositions). Important features of the diagram are: 1) the 689

silicic dominant character of the Oligocene ignimbrite pulse; 2) the appearance of basalts 690

during the Oligocene silicic ignimbrite pulse and increase in the frequency of basaltic 691

eruptions up to the start of the Early Miocene pulse; these basaltic eruptions correspond to 692

23

SCORBA of Cameron et al. (1989); 3) the bimodal character of the Early Miocene pulse; 4) 693

the increase in andesitic compositions beginning ~20 Ma until ~13 Ma; 5) the abrupt decline 694

in rhyolite magma generation and eruption beginning ~19 Ma when andesite-dacite eruptions 695

were more predominant; 6) an abrupt return to bimodal volcanism at ~13 Ma with a 696

concomitant decline in andesitic eruptions; and 7) after 10 Ma, volcanism becomes 697

fundamentally basaltic to coincide with the onset of seafloor spreading in the GoC beginning 698

as early as ~6 Ma (Lizarralde et al., 2007). Diagram based on 1496 radiometric ages, and age 699

data plotted using Isoplot (Ludwig, 2003). 700

701

Figure 4. SiO2 histograms of Oligocene-Miocene igneous rocks of NW Mexico. A) The 702

Oligocene pulse of the Sierra Madre Occidental (n = 346 analyses)is silicic dominant and in 703

particular, high-silica rhyolite-dominant. Many of the ignimbrites and lavas are high-silica 704

rhyolites (>75 wt% SiO2), and it is seldom appreciated that high-silica rhyolites, although 705

common in intracontinental settings, are uncommon in continental margin arcs (Hildreth & 706

Fierstein 2000). B) The Early Miocene bimodal pulse of the Sierra Madre Occidental (n = 707

106 analyses). C) Middle Miocene volcanism that is principally focussed around the margins 708

of the Gulf of California (n = 159 analyses), where in contrast to previous episodes, silicic 709

volcanism was subdued, and basaltic andesite to dacite compositions dominant. D) Late 710

Miocene volcanism (n = 321 analyses) is marked by a return to significant silicic volcanism 711

and true basaltic compositions. Note change in vertical scale for each time interval. 712

713

Figure 5. Time-space plots of igneous activity across western Mexico and around the Gulf of 714

California region. In A) age data are for the region between 20° and 26° N; and in B) between 715

26° and 30°N (see boxed areas in inset). Dated samples from Baja California have been 716

displaced to their “pre-rift” positions whereby for each sample the latitude has been adjusted 717

by 2.3°S and the longitude by 3.9°E. The approximate position of the Gulf of California in 718

both sections is marked by the grey arrow. Igneous rocks have been grouped into four main 719

compositional groupings - basalt (includes basaltic andesites and tholeiitic, calc-alkaline and 720

alkaline varieties), andesite (includes calc-alkaline and adakitic-type compositions), dacite 721

and rhyolite (includes high-silica rhyolites and peralkaline compositions). See text for 722

discussion. Inset map of western Mexico from Google Earth. 723

724

24

Figure 6. Field relationships between the SMO and GoC showing the progressive tilting and 725

deformation of the Early Miocene silicic ignimbrite successions in the central and western 726

SMO. A) view west from the Mazatlán-Durango old highway (23° 39.927'N 105° 43.340'W) 727

to the flat-lying ~23 Ma Espinazo-El Salto sequence of Waitt (1970) and McDowell & Keizer 728

(1977), and part of the undeformed core to the SMO. B) Tilted Early Miocene ignimbrite 729

successions approximately 30 km to the west of A viewed to the north of the Mazatlán-730

Durango old highway (23° 24.329'N 105° 54.634'W). C) Tilt blocks of Early Miocene 731

ignimbrites 27 km east of Tepic with ignimbrites dipping up to 35°E (21° 34.030'N 104° 732

37.880'W). The ignimbrites continue to the SSE below younger lavas and where they crop 733

out again have been dated at 20.46 Ma (Ferrari et al., in preparation). 25 km to the NW are 734

flat-lying (post-extensional) basaltic lavas dated between 11 and 10 Ma and form part of an 735

extensive series of flat-lying 9-12 Ma basaltic lavas along the coastal plain of Nayarit and 736

Sinaloa (see Fig. 1). These field relationships bracket the timing of major extension in the 737

GoC to between ~18 and ~12 Ma. 738

739

Figure 7. Ba/Nb vs La/Nb plots for Oligocene-Miocene (34-13 Ma) igneous rocks from NW 740

Mexico, with data organised into compositional groupings: basalts (<52 wt% SiO2); basaltic 741

andesite (52-56 wt% SiO2); andesite (56-63 wt% SiO2); and dacite-rhyolite (>63 wt% SiO2). 742

Stars labelled 1, 2 and 3 are E-MORB, N-MORB (Sun & McDonough, 1989) and averaged 743

calc-alkali andesite compositions (based on 2865 analyses from the GeoRoc database), 744

respectively. PRB, is the average composition of the Peninsular Ranges Batholith from Silver 745

& Chappell (1988). Grey shaded region labelled LMB represents compositions of Late 746

Miocene alkali basalts from the region (e.g., Henry & Aranda Gomez, 2000; Pallares et al., 747

2007). The anomalously high Ba/Nb ratios (>200) approach values for Paleozoic-Proterozoic 748

silicic orthogneiss and Tertiary plutonic-derived, glass-rich silicic xenoliths analysed from 749

the La Olivina basaltic centre in northern Chihuahua by Cameron et al. (1992), which have 750

Ba/Nb ratios between ~90 and 371 and 100-726, respectively. Compare with figure 12 of 751

Ewart et al. (1992). 752

753

Figure 8. A schematic model for the Early to Middle Miocene tectonomagmatic evolution of 754

western Mexico and the initial stages of rifting and opening of the Gulf of California. The 755

cross-sections illustrate: 1) the change from wide to narrow rifting and the focussing of 756

25

magmatism into the gulf region, which occurred ~18 Ma; 2) the general crustal structure such 757

as the development of thickened crust beneath the Sierra Madre Occidental core, and 758

requisite thinning of the lower crust and mantle lithosphere in the gulf region beginning at 759

least by the Early Miocene; 3) the main mafic magma source regions - the involvement of 760

asthenosphere-derived basaltic magmas is evident from at least the Early Miocene (e.g., see 761

Fig. 7) whereas more "calc-alkali” or subduction signature-bearing basalts are either 762

crust/mantle contaminated, or have been derived from subduction-modified lithospheric 763

mantle; 4) Early Miocene silicic magmas are predominantly mid-upper crustal melts with 764

significant material contributions from earlier formed batholithic type granitic rocks (as 765

informed by the zircon inheritance, Bryan et al., 2008; Ramos Rosique, 2012; Ferrari et al., 766

2012; in preparation); 5) for the Middle Miocene, additional silicic magma source regions 767

include the lower crust to account for the anomalously high Ba/Nb dacites/rhyolites (Fig. 7) 768

where lower crustal melting may have been promoted by thinning, decompression and 769

basaltic intrusion; and 6) active faulting interrupts bimodal magma systems (see inset) 770

existing regionally (red= rhyolite, black = basalt) to produce the andesitic magmas (green) in 771

the Middle Miocene around the margins of the Gulf of California. 772

773

774

775

26

References 776

777 AGUIRRE-DÍAZ, G. J. & MCDOWELL, F. W. 1993. Nature and timing of faulting and 778 synextensional magmatism in the southern Basin and Range, central-eastern Durango, 779 Mexico: Geological Society of America Bulletin, 105, 1435-1444. 780 781 AGUIRRE-DÍAZ, G. & LABARTHE-HERNÁNDEZ, G. 2003. Fissure ignimbrites: Fissure-782 source origin for voluminous ignimbrites of the Sierra Madre Occidental and its relationship 783 with Basin and Range faulting. Geology, 31, 773-776, doi: 10.1130/G19665.1. 784 785 ATWATER, T. & STOCK, J. 1998. Pacific-North America plate tectonics of the Neogene 786 southwestern United States. International Geology Review, 10, 375-402. 787 788 AXEN, G.J., TAYLOR W.J. & BARTLEY, J.M. 1993. Space-time patterns and tectonic 789 controls of Tertiary extension and magmatism in the Great Basin of the Western United 790 States. Geological Society of America Bulletin, 105, 56-76. 791 792 BEAL C.H. 1948. Reconnaissance geology and oil possibilities of Baja California, Mexico. 793 Geological Society of America Memior, 31, 138 pp. 794 795 BEST, M. G. & CHRISTIANSEN, E. H., 1991. Limited extension during peakTertiary 796 volcanism, Great Basin of Nevada and Utah. Journal of Geophysical Research, 96, 13509-797 13528. 798 799 BONNER, J. L. & HERRIN, E. T. 1999. Surface wave study of the Sierra Madre Occidental 800 of northern Mexico. Bulletin of the Seismological Society of America, 89, 1323-1337. 801 802 BRYAN S.E., FERRARI L., REINERS P.W., ALLEN C.M., PETRONE C.M., RAMOS 803 ROSIQUE A., & CAMPBELL I.H., 2008. New insights into crustal contributions to large 804 volume rhyolite generation at the mid-Tertiary Sierra Madre Occidental Province, Mexico, 805 revealed by U-Pb geochronology. Journal of Petrology, 49, 47-77. 806 807 CAMERON, M., BAGBY, W. & CAMERON, K.L. 1980. Petrogenesis of voluminous mid-808 Tertiary ignimbrites of the Sierra Madre Occidental, Chihuahua, Mexico. Contributions to 809 Mineralogy and Petrology, 74, 271-284. 810 811 CAMERON, K. L., NIMZ, G. J., KUENTZ, D., NIEMEYER, S. & GUNN, S. 1989. 812 Southern Cordilleran basaltic andesite suite, southern Chihuahua, Mexico: A link between 813 Tertiary continental arc and flood basalt magmatism in North America. Journal of 814 Geophysical Research, 94, 7817-7840. 815 816 CAMERON, K.L., ROBINSON, J.V., NIEMEYER, S., NIMZ, G.J., KUENTZ, D.C., 817 HARMON, R.S., BOHLEN, S.R. & COLLERSON, K.D. 1992. Contrasting styles of Pre-818 Cenozoic and mid-Tertiary crustal evolution in northern Mexico: Evidence from deep crustal 819 xenoliths from La Olivina. Journal of Geophysical Research, 97, 17353–17376. 820 821

27

CARREÑO A.L. & SMITH J.T. 2007. Stratigraphy and Correlation for the ancient Gulf of 822 California and Baja California Peninsula, Mexico. Bulletin of American Paleontology, 371, 823 158 p. 824 825 CASTILLO, P.R. 2008. Origin of the adakite-high Nb basalt associationand its implications 826 for post-subduction magmatism in Baja California, Mexico. Geological Society of America 827 Bulletin, 120, 451–462. 828 829 CLIFT P.D. & LEG 135 SCIENTIFIC PARTY, 1995. Volcanism and sedimentation in a 830 rifting island arc terrain: an example from Tonga, SW Pacific. Geological Society of London 831 Special Publication 81, 29-51. 832 833 COLE, J. W. 1990. Structural control and origin of volcanism in the Taupo Volcanic Zone, 834 New Zealand. Bulletin of Volcanology ,52, 445-459. 835 836 CONLY, A.G., BRENAN, J.M., BELLON, H. & SCOTT, S.D. 2005. Arc to rift transitional 837 volcanism in the Santa Rosalý´a Region, Baja California Sur, Mexico. Journal of 838 Volcanology and Geothermal Research, 142, 303–341. 839 840 COUCH, R.W, NESS, G.E., SANCHEZ-ZAMORA, O., CALDERON-RIVEROLL, G., 841 DOGUIN, P., PLAWMAN, T., COPERUDE, S., HUEHN, B., & GUMMA, W., 1991. 842 Gravity anomalies and crustal structure of the Gulf and Peninsular Province of the 843 Californias. In DAUPHIN, J.P., & SIMONEIT, B.R.T. (eds), The Gulf and the Peninsular 844 Province of the Californias. American Association of Petroleum Geologists Memoir, 47, 47–845 70. 846 847 DAMON, P.E., SHAFI QULLAH, M. & CLARK, K.F. 1981. Age trends of igneous activity 848 in relation to metallogenesis in the southern Cordillera. Arizona Geological Society Digest, 849 14, 137–154. 850 851 DeMETS, C. 1995. A reappraisal of seafloor spreading lineations in the Gulf of California: 852 Implications for the transfer of Baja California to the Pacific plate and estimates of Pacific-853 North America motion. Geophysical Research Letters, 22, 3545-3548. 854 855 DRAKE, W. 2005. Structural analysis, stratigraphy, and geochronology of the San José 856 Island accommodation zone, Baja California Sur, Mexico. M.Sc. Thesis, Northern Arizona 857 University, 271 p. 858 859 DUNCAN A.R. 1987. The Karoo igneous province - aproblem area for inferring tectonic 860 setting from basalt geochemistry. Journal of Volcanlogy and Geothermal Research, 32, 13-861 34. 862 863 EWART, A., SCHON, R.W. & CHAPPELL, B.W. 1992. The Cretaceous volcanic–plutonic 864 province of the central Queensland (Australia) coast—a rift related “calc-alkaline” province. 865 Transactions of the Royal Society of Edinburgh: Earth Sciences, 83, 327–345. 866 867 EWART, A., MILNER, S.C., ARMSTRONG, R.A., & DUNCAN, A.R., 1998. Etendeka 868 volcanism of the Goboboseb Mountains and Messum Igneous complex Namibia. Part I: 869

28

Geochemical evidence of Early Cretaceous Tristan plume melts and the role of crustal 870 contamination in the Paraná–Etendeka CFB. Journal of Petrology, 39, 191–225. 871 872 FERRARI L., LOPEZ MARTÍNEZ, M., OROZCO, M.T., BRYAN, S.E., DUQUE 873 TRUJILLO, J. & LONSDALE, P. (In preparation) Miocene rifting and syn-extensional 874 magmatism in the southwestern Sierra Madre Occidental (Sinaloa and Nayarit states): the 875 early history of the Gulf of California. Geosphere. 876 877 FERRARI, L., LÓPEZ-MARTINEZ, M., AGUIRRE-DÍAZ, G. & CARRASCO-NÚÑEZ, G. 878 1999. Space time patterns of Cenozoic arc volcanism in central Mexico: From the Sierra 879 Madre Occidental to the Mexican Volcanic Belt. Geology, 27, 303-306. 880 881 FERRARI, L., LÓPEZ-MARTÍNEZ, M. & ROSAS-ELGUERA, J. 2002. Ignimbrite flare-up 882 and deformation in the southern Sierra Madre Occidental, western Mexico-implications for 883 the late subduction history of the Farallon Plate. Tectonics, 21, doi: 10.1029/2001TC001302. 884 885 FERRARI, L., VALENCIA-MORENO, M. & BRYAN, S. 2007. Magmatism and tectonics 886 of the Sierra Madre Occidental and its relation with the evolution of the western margin of 887 North America. In ALANIZ-ÁLVAREZ, S.A., & NIETO-SAMANIEGO, A.F. (eds), 888 Geology of México: Celebrating the Centenary of the Geological Society of México. 889 Geological Society of America Special Paper, 422, 1-39, doi: 10.1130/2007.2422(01). 890 891 FERRARI L.,OROZCO-ESQUIVEL, T., LOPEZ MARTÍNEZ, M., DUQUE TRUJILLO J., 892 BRYAN, S.E. & CERCA, M. 2012. 25 million years to break a continent: early to middle 893 Miocene rifting and syn-extensional magmatism in the southern Gulf of California. 894 Geological Society of America Abstracts with Programs, 44, No. 3, p. 6. 895 896 FLETCHER, J.M., GROVE, M., KIMBROUGH, D., LOVERA, O. & GEHRELS, G.E. 897 2007. Ridge-trench interactions and the Neogene tectonic evolution of the Magdalena Shelf 898 and southern Gulf of California; insights from detrital zircon U-Pb ages from the Magdalena 899 Fan and adjacent areas. Geological Society of America Bulletin, 119, 1313-1336, doi: 900 10.1130/ B26067.1. 901 902 GANS, P. 1997. Large-magnitude Oligo-Miocene extension in southern Sonora: Implications 903 for the tectonic evolution of northwest Mexico. Tectonics, 16, 388-408. 904 905 GANS, P.B. & BOHRSON, W.A. 1998. Suppression of volcanism during rapid extension in 906 the Basin and Range Province, United States. Science, 279, 66–68, doi: 907 10.1126/science.279.5347.66. 908 909 GANS, P. B., BLAIR, K. D., MACMILLAN, I., WONG, M. S. & ROLDAN-QUINTANA, 910 J. 2003. Structural and magmatic evolution of the Sonoran rifted margin: A preliminary 911 report. Geological Society of America Abstracts with Programs, 35 (4), 21. 912 913 GANS, P., MAHOOD, G. & BORSHON, W. 1989. Synextensional magmatism in the Basin 914 and Range Province; a case study from the eastern Great Basin. Geological Society of 915 America Special Paper, 233, 1-53. 916 917

29

GASTIL, R G., KRUMMENACHER, D. & MINCH, J. 1979. The record of Cenozoic 918 volcanism around the Gulf of California. Geological Society of America Bulletin, 90, 839-919 857. 920 921 GODINEZ. N.S., KIMBROUGH, D.L. & KOHEL, C. 2010. Stratigraphy and petrologic 922 evolution of the Oligocene-Miocene Comondu Group near Bahia Concepcion and Loreto, 923 Baja California Sur, Mexico. Geological Society of America Abstracts with Programs, 42, 924 No. 4, 66. 925 926 GOMBERG, J.S., PRIESTLEY, K.F., MASTERS, G. & BRUNE, J. 1988. The structure of 927 the crust and upper mantle in northern Mexico. Geophysical Journal, 94, 1–20. 928 929 HAUSBACK, B.P. 1984. Cenozoic volcanic and tectonic evolution of Baja California Sur, 930 Mexico. In FRIZZEL, V.A., Jr., ed., Geology of the Baja California Peninsula. Pacific 931 Section, Society of Economic Paleontologists and Mineralogists Special Publication 39, 219-932 236. 933 934 HELENES, J. 2012. Neogene micropaleontology and evolution of the Northern Gulf of 935 California. Geological Society of America Abstracts with Programs, 44, No. 3, p. 18. 936 937 HELENES, J., CARREÑO, A.L. & CARRILLO, R.M. 2009. Middle to late Miocene 938 chronostratigraphy and development of the northern Gulf of California. Marine 939 Micropaleontology, 72, 10-25, doi: 10.1016/j.marmicro.2009.02.003. 940 941 HENRY, C.D. & ARANDA-GOMEZ, J.J. 2000. Plate interactions control middle-late 942 Miocene, proto-Gulf and Basin and Range extension in the southern Basin and Range. 943 Tectonophysics, 318, 1-26. 944 945 HILDRETH W. & FIERSTEIN J. 2000. Katmai volcanic cluster and the great eruption of 946 1912. Geological Society of America Bulletin, 112, 1594–1620. 947 948 HOSACK, A. 2006. Structure and stratigraphy of the Agua Verde accommodation zone, Baja 949 California Sur, Mexico. M.Sc. thesis, Northern Arizona University, Flagstaff AZ. 950 951 JOHNSON, J.A. & GRUNDER, A.L., 2000. The making of intermediate composition 952 magma in a bimodal suite: Duck Butte eruptive center, Oregon, USA. Journal of Volcanology 953 and Geothermal Research, 95, 175-195. 954 955 KARIG, D.E. & JENSKY, W. 1972. The Protogulf of California. Earth and Planetary 956 Science Letters, 17, 169–174. 957 958 KING, R. E., 1939. Geological reconnaissance in northern Sierra Madre Occidental of 959 Mexico. 960 Geological Society of America Bulletin, 50, 1625-1722. 961 962 LANPHERE, M. A., CAMERON, K. L. & CAMERON, M. 1980. Sr isotopic geochemistry 963 of voluminous rhyolitic ignimbrites and related rocks, Batopilas area, western Mexico. 964 Nature, 286, 594-596. 965

30

966 Le MAITRE, R.W. 1989. A Classification of Igneous Rocks and Glossary of Terms: 967 Recommendations of the IUGS Subcommission on the Systematics of Igneous Rocks. London, 968 Blackwell Scientific, 193 p. 969 970 LIPMAN, P.W., PROSTKA, H. J. & CHRISTIANSEN, R. L. 1972. Cenozoic volcanism and 971 plate-tectonic evolution of the western United States, Part 1: Early and Middle Cenozoic. 972 Philosophical Transactions of the Royal Society of London Series A, 271, 217-248. 973 974 LIZARRALDE, D., AXEN, G.J., BROWN, H.E., FLETCHER, J.M., GONZÁLEZ-975 FERNÁN- DEZ, A., HARDING, A.J., HOLBROOK, W.S., KENT, G.M., PARAMO, P., 976 SUTHERLAND, F. & UMHOEFER, P.J. 2007. Variable styles of rifting in the Gulf of 977 California. Nature, 448, 466-469. 978 979 LONSDALE, P. 1989. Geology and tectonic history of the Gulf of California. In 980 WINTERER, E. L., HUSSONG, D. M., & DECKER, R. W. (eds.), The eastern Pacific 981 Ocean and Hawaii. Boulder, Colorado, Geological Society of America, Geology of North 982 America, v. N, p. 499-521. 983 984 LONSDALE, P. 1991. Structural patterns of the Pacific floor offshore of peninsula 985 California. In the Gulf and Peninsula Province of the Californias. AAPG Memior, 47, 87-125. 986 987 LUDWIG, K. R. 2003. Isoplot Version 3.00: a Geochronological Tool-kit for Microsoft 988 Excel. Berkeley Geochronology Center, Special Publications. 989 990 LUHR, J.F.,ARANDA-GOMEZ, J.J. & HOUSH, T.B. 1995. San Quintin Volcanic field, 991 Baja California Norte,Me´xico. Geology, petrology, and geochemistry. Journal of 992 Geophysical Research, 100, 10,353–10,380. 993 994 MARTÍN-BARAJAS, A., STOCK, J.M., LAYER, P., HAUSBACK, B., RENNE, P. & 995 LOPEZ-MARTINEZ, M. 1995. Arc-rift transition volcanism in the Puertecitos Volcanic 996 Province, northeastern Baja California, Mexico. Geological Society of America Bulletin, 107, 997 407–424. 998 999 MARTÍN, A., FLETCHER, J.M., LOPEZ-MARTINEZ, M. & MENDOZA-BORUNDA, R. 1000 2000. Waning Miocene subduction and arc volcanism in Baja California: the San Luis 1001 Gonzaga volcanic field. Tectonophysics, 318, 27–51. 1002 1003 MARTINY, B., MARTÍNEZ–SERRANO, R.G., MORÁN–ZENTENO, D.J., MACIAS–1004 ROMO, C. & AYUSO, R.A. 2000. Stratigraphy, geochemistry and tectonic significance of 1005 the Oligocene magmatic rocks of Western Oaxaca, Southern Mexico. Tectonophysics, 318, 1006 71–98. 1007 1008 McDOWELL, F.W. 2007. Geologic transect across the northern Sierra Madre Occidental 1009 volcanic fi eld, Chihuahua and Sonora, Mexico. The Geological Society of America Digital 1010 Map and Chart Series, 6, 70 p., doi: 10.1130/2007.DMCH006.TXT. 1011 1012

31

McDOWELL, F.W. & CLABAUGH, S.E. 1979. Ignimbrites of the Sierra Madre Occidental 1013 and their relation to the tectonic history of western Mexico. In CHAPIN, C.E. & ELSTON, 1014 W.E. (eds.), Ash-Flow Tuffs. Geological Society of America Special Paper, 180, p. 113-124. 1015 1016 McDOWELL, F. & KEIZER, R.P. 1977. Timing of mid-Tertiary volcanism in the Sierra 1017 Madre Occidental between Durango city and Mazatlán, Mexico. Geological Society of 1018 America Bulletin, 88, 1479-1487. 1019 1020 McDOWELL, F.W. & MAUGER, R.L. 1994. K-Ar and U-Pb zircon chronology of Late 1021 Cretaceous and Tertiary magmatism in central Chihuahua State, Mexico. Geological Society 1022 of America Bulletin, 106, 118–132, doi:10.1130/0016-1023 7606(1994)106<0118:KAAUPZ>2.3.CO;2. 1024 1025 McDOWELL, F.W. & McINTOSH, W.C. 2012. Timing of intense magmatic episodes in the 1026 northern and central Sierra Madre Occidental, western Mexico. Geosphere, 8,1505-1526. 1027 1028 McDOWELL, F., ROLDÁN-QUINTANA, J. & AMAYA-MARTÍNEZ, R. 1997. 1029 Interrelationship of sedimentary and volcanic deposits associated with Tertiary extension in 1030 Sonora, Mexico. Geological Society of America Bulletin, 109, 1349-1360. 1031 1032 McQUARRIE, N. & OSKIN M., 2011. Palinspastic restoration of NAVDat and implications 1033 for the origin of magmatism in southwestern North America. Journal of Geophysical 1034 Research, 115, B10401, 16 PP., 2010 doi:10.1029/2009JB006435. 1035 1036 MORA-KLEPEIS, G. & McDOWELL, F.W. 2004. Late Miocene calc-alkalic volcanism in 1037 northwestern Mexico: An expression of rift or subduction related magmatism? Journal of 1038 South American Earth Sciences, 17, 297–310, doi: 10.1016/j.jsames.2004.08.001. 1039 1040 MORÁN–ZENTENO, D. J., ALBA–ALDAVE, L., GONZALEZ–TORRES, E. A., 1041 MARTINY, B., BRAVO–DIAZ, B.A. & SOHN, E. 2007. Ignimbrite flare–up in the north–1042 central Sierra Madre del Sur, Southern Mexico: a continuation of the Sierra Madre Occidental 1043 ignimbrite Province?, American Geophysical Union Joint Assembly Abstracts, Acapulco, 1044 Mexico, Abstract U53A–03. 1045 1046 MORRIS G.A. & HOOPER, P.R. 1997. Petrogenesis of the Colville Igneous Complex, 1047 northeast Washington: Implications for Eocene tectonics in the northern U.S. Cordillera. 1048 Geology, 25, 831-834. 1049 1050 MURRAY B., BUSBY, C.J.,.FERRARI L. & SOLARI, L. 2010. Tectonic controls on Late 1051 Oligocene volcanism in the Guazapares mining district, northwestern Sierra Madre 1052 Occidental, Mexico. Geological Society of America Abstracts with Programs, 42, No. 5, p. 1053 344. 1054 1055 MURRAY B.P., BUSBY, C.J., FERRARI F. & SOLARI, L.A., Submitted. Synvolcanic 1056 crustal extension during the mid-Cenozoic ignimbrite flare-up in the northern Sierra Madre 1057 Occidental, Mexico: Evidence from the Guazapares Mining District region, western 1058 Chihuahua. Geosphere. 1059 1060

32

NIETO-OBREGÓN, J., DELGADO-ARGOTE, L. & DAMON, P.E. 1981. Relaciones 1061 petrológicas y geocronológicas del magmatismo de la Sierra Madre Occidental el Eje 1062 Neovolcánico en Nayarit, Jalisco y Zacatecas. Asociación de Ingenieros de Minas, 1063 Metalurgistas y Geólogos de México, Memoria Técnica, XIV, 327–361. 1064 1065 NOURSE, J.A., ANDERSON, T.H. & SILVER, L.T. 1994. Tertiary metamorphic core 1066 complexes in Sonora, northwestern Mexico. Tectonics, 13, 1161–1182, 1067 doi:10.1029/93TC03324. 1068 1069 OROZCO-ESQUIVEL, T., LÓPEZ-MARTÍNEZ, M., LONSDALE, P., FERRARI, L., 1070 CORNEJO-JIMENEZ, C. PINERO-LAJAS, D., & DUQUE-TRUJILLO, J.F., 2010. Tearing 1071 apart of an Early Miocene silicic igneous province in the southern Gulf of California. 1072 Geological Society of America Abstracts with Programs, 42, No. 5, p. 296. 1073 1074 OSKIN, M.E., & STOCK, J.M., 2003. Marine incursion synchronous with plate-boundary 1075 localization in the Gulf of California. Geology, 31, 23-26. 1076 1077 OSKIN, M.E., STOCK, J.M. & MARTÍN-BARAJAS, A. 2001. Rapid localization of Pacific-1078 North America plate motion in the Gulf of California. Geology, 29, 459-462. 1079 1080 PALLARES, C., MAURY, R.C., BELLON, H., ROYER, J.Y., CALMUS, T., AGUILLON-1081 ROBLES, A., COTTEN, J., BENOIT, M., MICHAUD, F., & BOURGOIS, J., 2007. Slab-1082 tearing following ridge-trench collision: evidence from Miocene volcanism in Baja 1083 California, Mexico. Journal of Volcanology and Geothermal Research, 161, 95–117. 1084 doi:10.1016/j.jvolgeores.2006.11.002. 1085 1086 PARSON, L.M. & WRIGHT, I.C. 1996. The Lau-Havre-Taupo back-arc basin: a southward- 1087 propagating, multi-stage evolution from rifting to spreading. Tectonophysics, 263, 1-22. 1088 1089 PAZ-MORENO, F. A., DEMANT, A., COCHEME J.-J., DOSTAL, J., & MONTIGNY, R., 1090 2003. The Quaternary Moctezuma volcanic field: a tholeiitic to alkali basaltic episode in the 1091 central Sonora Basin and Range Province, Mexico. In: JOHNSON, S.E., PATERSON S.R., 1092 FLETCHER, J.M., GIRTY, G.H., KIMBROUGH, D.L., & MARTÍN-BARAJAS, A. (eds). 1093 Tectonic evolution of northwestern Me´xico and southwestern USA. Geological Society of 1094 America Special Paper, 374, 439–455. 1095 1096 PERSAUD, P., PÉREZ-CAMPOS, X., & CLAYTON, R.W., 2007. Crustal thickness 1097 variations in the margins of the Gulf of California from receiver functions. Geophysical 1098 Journal International, 170, no. 2, 687-699, doi: 10.1111/j.1365-246X.2007.03412.x. 1099 1100 RAMOS ROSIQUE A. 2012. Timing and evolution of Late Oligocene to Early Miocene 1101 magmatism and epithermal mineralization in the central Bolaños graben, southern Sierra 1102 Madre Occidental, México. Unpublished PhD Thesis, Kingston University, London, United 1103 Kingdom. 1104 1105 RAMOS ROSIQUE, A., BRYAN, S., FERRARI, L., ALLEN, C., LOPEZ-MARTINEZ, M., 1106 & RANKIN, A., 2010. Timing and evolution of Late Oligocene to Miocene magmatism in 1107 the southern Sierra Madre Occidental silicic large igneous province: insights from zircon 1108

33

chronochemistry and Ar/Ar geochronology. Geophysical Research Abstracts, 12, EGU2010-1109 9788, 2010. 1110 1111 ROBERTS, M. P. & CLEMENS, J. D. 1993. Origin of high-potassium, calc-alkaline, I-type 1112 granitoids. Geology, 21, 825-828. 1113 1114 SAWLAN, M. G. 1991. Magmatic evolution of the Gulf of California rift. In: DAUPHIN, J. 1115 P. & SIMONEIT, B. R., eds., The Gulf and Peninsular Province of the Californias. American 1116 Association of Petroleum Geologists Memoir 47, p. 301-369. 1117 1118 SAWLAN, M.G. & SMITH, J.G. 1984. Petrologic characteristics, age and tectonic setting of 1119 Neogene volcanic rocks in northern Baja California Sur, Mexico. In FRIZZELL, V.A., Jr. 1120 (ed.), Geology of the Baja California Peninsula. Society of Economic Paleontologists and 1121 Mineralogists, Pacific Section, 39, 237-251. 1122 1123 SILVER, L.T. & CHAPPELL, B.W. 1988. The Peninsular Ranges Batholith: An insight into 1124 the evolution of the Cordilleran batholiths of southwestern North America. Philosophical 1125 Transactions of the Royal Society of London, A79, 105-121. 1126 1127 STOCK, J.M. & HODGES, K.V. 1989. Pre-Pliocene extension around the Gulf of California 1128 and the transfer of Baja California to the Pacific plate. Tectonics, 8, 99-115. 1129 1130 STOCK, J.M. LEWIS C.J. & NAGY E.A. 1999. The Tuff of San Felipe: an extensive middle 1131 Miocene pyroclastic flow deposit in Baja California, Mexico. Journal of Volcanology and 1132 Geothermal Research, 93, 53–74. 1133 SUTHERLAND, F.H. 2006. Continental rifting across the southern Gulf of California. Ph.D. 1134 thesis, University of California at San Diego, La Jolla, California, 189 p. 1135 1136 SUTHERLAND, F.H., KENT, G.M., HARDING, A.J., UMHOEFER, P.J., DRISCOLL, 1137 N.W., LIZARRALDE, D., FLETCHER, J.M., AXEN, G.J., HOLBROOK, W.S., 1138 GONZÁLEZ-FERNÁNDEZ, A. & LONSDALE P. 2012. Middle Miocene to early Pliocene 1139 oblique extension in the southern Gulf of California. Geosphere, 10.1130/GES00770.1 1140 1141 SWANSON, E.R. & McDOWELL, F.W. 1984. Calderas of the Sierra Madre Occidental 1142 volcanic field western Mexico. Journal of Geophysical Research, 89, 8787-8799. 1143 1144 SWANSON, E.R., KEMPTER, K.A., McDOWELL, F.W., & MCINTOSH, W.C., 2006. 1145 Major ignimbrites and volcanic centers of the Copper Canyon area: A view into the core of 1146 Mexico's Sierra Madre Occidental. Geosphere, 2, no. 3, 125, doi: 10.1130/GES00042.1. 1147 1148 TILL, C. B., GANS, P. B., SPERA, F. J., MACMILLAN, I. & BLAIR K.D. 2009. Perils of 1149 petrotectonic modeling: A view from southern Sonora, Mexico. Journal of Volcanology & 1150 Geothermal Research, 186, 160–168, doi:10.1016/j.jvolgeores. 2009.06.014. 1151 1152 UMHOEFER, P.J., DORSEY, R.J., WILLSEY, S., MAYER, L. & RENNE, P. 2001. 1153 Stratigraphy and Geochronology of the Comondú Group near Loreto, Baja California Sur, 1154 Mexico. Sedimentary Geology, 44, 125-147. 1155 1156

34

UMHOEFER, P.J. 2011. Why did the Southern Gulf of California rupture so rapidly?-1157 Oblique divergence across hot, weak lithosphere along a tectonically active margin. GSA 1158 Today, 21, no. 11, 4-10. 1159 1160 UMHOEFER, P.J., SUTHERLAND, F., KENT, G., HARDING, A., LIZARRALDE, D., 1161 SCHWENNICKE, T., FLETCHER, J., HOLBROOK, W.S. & AXEN, G. 2008. Synchronous 1162 changes in rift-margin basins and initiation of the Alarcón spreading ridge and related 1163 transform fault, Southwestern Gulf of California. Geological Society of America Abstracts 1164 with Programs, 40, no. 6, p. 151. 1165 1166 VIDAL-SOLANO J.R., PAZ-MORENO, F.A.,DEMANT, A. & LOPEZ-MARTINEZ, 1167 M.2007. Ignimbritas hiperalcalinas del Mioceno medio en Sonora Central: revaluación de la 1168 estratigrafía y significado del volcanismo terciario. Revista Mexicana de Ciencias 1169 Geológicas, 24, 47-67. 1170 1171 VIDAL-SOLANO, J.R., DEMANT, A., PAZ-MORENO, F.A., LAPIERRE, H., ORTEGA-1172 RIVERA, M.A. & LEE, J.K.W. 2008. Insights into the tectonomagmatic evolution of NW 1173 Mexico: geochronology and geochemistry of the Miocene volcanic rocks from the Pinacate 1174 area, Sonora. Geological Society of America Bulletin, 120, 691–708. 1175 1176 WAITT, R.B. 1970. Ignimbrites of the Sierra Madre Occidental between Durango and 1177 Mazatlán, Mexico. M.A. thesis, University of Texis at Austin, USA, 133 p. 1178 1179 WARD, P. L. 1995. Subduction cycles under western North America during the Mesozoic 1180 and Cenozoic eras. In MILLER, D. M., AND BUSBY, C. (eds.), Jurassic Magmatism and 1181 Tectonics of the North American Cordillera. Geological Society of America Special Paper, 1182 299, 1-45. 1183 1184 WARK, D.A., KEMPTER, K.A. & McDOWELL, F.W. 1990. Evolution of waning, 1185 subduction-related magmatism, northern Sierra Madre Occidental, Mexico. Geological 1186 Society of America Bulletin, 102, 1555–1564. 1187 1188 WONG, M.S., GANS, P.B., & SCHEIER, J., 2010. The 40Ar/39Ar thermochronology of core 1189 complexes and other basement rocks in Sonora, Mexico: Implications for Cenozoic tectonic 1190 evolution of northwestern Mexico. Journal of Geophysical Research, 115, no. B7, doi: 1191 10.1029/2009JB007032. 1192 1193 ZELLMER, G.F., 2009. Petrogenesis of Sr-rich adakitic rocks at volcanic arcs: insights from 1194 global variations of eruptive style with plate convergence rates and surface heat flux. Journal 1195 of the Geological Society, London, 166, 725–734. doi: 10.1144/0016-76492008-0721. 1196 1197 ZELLMER, G.F., SAKAMOTO, N., IIZUKA, Y., MIYOSHI, M., TAMURA, Y., HSIEH 1198 H-H., & YURIMOTO, H., SUBMITTED. Crystal uptake into superheated melts: insights 1199 from two-pyroxene pseudo-decompression paths, plagioclase hygrometry, and measurement 1200 of hydrogen in olivines from mafic volcanic rocks of the southwest Japan arc. Geological 1201 Society of London Special Volume on “Orogenic Andesites and Crustal Growth”. 1202 1203

35

ZHU, D-C., ZHAO Z-D.,, NIU Y., DILEK Y., WANG Q., JI W-H., DONG G-C., SUI Q.L., 1204 LIU Y-S., YUAN H-L., & MO X.-X., 2012. Cambrian bimodal volcanism in the Lhasa 1205 Terrane, southern Tibet: Record of an early Paleozoic Andean-type magmatic arc in the 1206 Australian proto-Tethyan margin. Chemical Geology, doi:10.1016/j.chemgeo.2011.12.024 1207 1208