Kinematic evolution of a large-offset continental normal fault system, South Virgin Mountains, Nevada

Robert Brady* Brian Wernicke Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, California 91125, USA Joan Fryxell Department of Geological Sciences, California State University, San Bernardino, California 92407, USA

ABSTRACT examination of other areas suggests that the John and Foster, 1993); others may have ini- evolutionary sequence seen in the South tiated with steep dips and remained active as The South Virgin Mountains and Grand Virgin Mountains may, in fact, be widely they tilted to shallow dips. Tilting of these Wash trough comprise a mid-Miocene nor- applicable. faults is probably due to combined domino- mal fault system that de®nes the boundary style tilting and isostatically driven footwall between the unextended Colorado Plateau Keywords: Egan Range, kinematics, Lemi- ¯exure (Fig. 1C). Many gently dipping normal to the east and highly extended crust of the tar Mountains, normal faults, South Virgin faults might be best interpreted this way, rath- central Basin and Range province to the Mountains, Yerington. er than invoking rotation by younger struc- west. In the upper 3 km of the crust, the tures. system developed in subhorizontal cratonic INTRODUCTION Detailed geologic studies were conducted in strata in the foreland of the Cordilleran the South Virgin Mountains of southeastern fold and thrust system. The rugged topog- Geologic mapping in the Basin and Range Nevada and northwestern Arizona (Fig. 2), a raphy and lack of vegetation of the area af- province has shown that gently dipping nor- region well suited for testing existing kine- ford exceptional three-dimensional expo- mal faults are common (e.g., Longwell, 1945; matic models of normal faulting. Extensional sures. Compact stratigraphy and well-de- Anderson, 1971; Proffett, 1977; Miller et al., faults within the Grand Wash trough and ad- ®ned prefaulting con®guration of the rocks 1983; Wernicke et al., 1985; John and Foster, jacent South Virgin Mountains de®ne a rela- permitted a detailed reconstruction of the 1993), but theoretical fault mechanics sug- tively sharp transition from the unextended system. Reconstruction of cross sections gests that such faults should not slip (cf. An- Colorado Plateau to the strongly extended Ba- 2 based on more than 300 km of detailed derson, 1942; Sibson, 1985; Nur et al., 1986; sin and Range (stretching factor ␤ ഠ 3.5; Wer- mapping at a scale of 1:12 000 shows that Agnon and Reches, 1995). These faults have nicke et al., 1988). The plateau consists of the fault system accommodated more than been explained in a manner consistent with high-grade Proterozoic basement nonconform- 15 km of roughly east-west±directed Mio- mechanical theory by invoking tilting to lower ably overlain by a thin (ϳ2 km) Paleozoic cene extension. Extension was initially ac- dips after cessation of slip (e.g., Proffett, cover, and forms a headwall from which steep- commodated on moderately to steeply dip- 1977; Miller et al., 1983; Buck, 1988). For ly east tilted normal fault blocks, bounded by ping listric normal faults. As the early example, in the Snake Range, the Lemitar both steep and shallow normal faults, includ- faults and fault blocks tilted, steeply to Mountains, and the Yerington district, gently ing both basement and cover, were derived. moderately dipping faults initiated within dipping faults have been interpreted as the re- The mapped area comprises a series of eight the fault blocks, soling into the early faults. sult of multiple generations of crosscutting, tilted fault blocks deformed by multiple gen- Some of the early faults were active at dips steeply dipping normal faults, with older erations of normal faults in middle Miocene of Ͻ20؇. Isostatically driven tilting is su- faults being tilted to shallow dips above youn- time (Fig. 3). These fault blocks are bounded perimposed on tilting due to active slip and ger faults (Fig. 1A; Proffett, 1977; Chamber- by normal faults with 1 km or more of throw, domino-style rotation of the fault blocks. lin, 1982, 1983; Miller et al., 1983). Gently and are internally disrupted by numerous Collectively these processes rotated origi- dipping normal faults have also been attribut- smaller offset faults. Four of these blocks, in- nally steeply dipping faults to horizontal ed to the rotation of originally steeply dipping cluding (from east to west) the Wheeler orientations. The kinematics are inconsis- faults after abandonment above an up¯exing, Ridge, Iceberg Ridge, Indian Hills, and Con- tent with the widely accepted view that tectonically denuded, footwall (Fig. 1B; Buck, noly Wash blocks, straddle the eastern end of many near-horizontal normal faults were 1988). Lake Mead, Nevada (Fig. 3). The next block rotated to their present orientations by lat- However, many gently dipping normal to the west, interpreted to be structurally con- er, crosscutting normal faults. However, re- faults can not be explained by either of these tinuous with the Gold Butte crystalline block, mechanisms. Some of them probably initiated is the Azure Ridge block. The next three *E-mail: [email protected]. with shallow dips (e.g., Wernicke et al., 1985; blocks to the west, including Tramp Ridge,

GSA Bulletin; September 2000; v. 112; no. 9; p. 1375±1397; 16 ®gures; 1 table.

1375 BRADY et al.

Figure 2. Location map, showing the South Virgin Mountains (stippled) and surrounding area, as well as major structural features. Light gray indicates the location of mountainous terrane. BMÐBlack Mountains, CPÐColorado Plateau, FMÐFrenchman Mountain, GPTÐGass Peak thrust, KTÐKeystone thrust, VDÐVirgin detachment, LMFSÐLake Mead fault system, LVRÐLas Vegas Range, LVVSZÐLas Vegas Valley shear zone, MMÐMuddy Mountains, MMTÐMuddy Mountains thrust, NVMÐNorth Virgin Moun- tains, SRÐSheep Range, SVMÐSouth Virgin Mountains.

Lime Ridge, and the Maynard Spring block, to the thin-skinned Sevier thrust belt (Burch- are north of and structurally above the crys- ®el et al., 1974), which developed within the talline block, which is a 15-km-wide basement sedimentary rocks of the Paleozoic miogeo- dome, unroofed by a west-dipping normal cline. The easternmost thrusts seem to have fault (Fryxell et al., 1992). been localized along the hinge zone of the Most of the faults within the South Virgin miogeocline, which trends roughly northeast- Figure 1. Schematic cross sections of the up- Mountains cut through Paleozoic to Tertiary southwest, and runs just to the west of Lake per crust showing three models of normal strata. The stratigraphic section includes more Mead. Accordingly, the easternmost thrust fault evolution, all of which result in both than 30 mappable units in about 3 km of sec- sheet at the latitude of Lake Mead is exposed shallowly and more steeply dipping normal tion. The South Virgin Mountains were little nearby, to the west, in the Muddy Mountains faults. Incipient faults are shown as dashed deformed by Mesozoic compression, as evi- (Fig. 2; Longwell, 1949; Longwell et al., lines, active faults are shown as heavy black denced by the lack of contractional structures 1965; Bohannon, 1983). South of Lake Mead, lines, and inactive faults are shown as thin- within the region, and the very gentle sub- the thrust belt changes to a northwest-south- ner black lines. (A) Multiple generations of Tertiary unconformity across the South Virgin east trend, and thrusting involves Precambrian domino faults; only moderately to steeply and Beaver Dam Mountains (Bohannon, 1984; crystalline basement rocks (Burch®el and Da- dipping faults are active; shallowly dipping Wernicke and Axen, 1988). The faulted sec- vis, 1975). faults have been rotated above younger tions of the South Virgin Mountains must re- Following Cretaceous thrusting, the Lake crosscutting faults (cf. Proffett, 1977; Miller store so as to be continuous with the ¯at-lying Mead region seems to have been tectonically et al., 1983). (B) Rolling hinge model; ini- strata of the Colorado Plateau. Because the stable until Neogene time. However, early tially steeply dipping faults are ¯exurally ro- faults within the South Virgin Mountains are Miocene to Eocene deposits ®lling northeast- tated above an uplifting footwall, becoming well exposed, have accommodated large-mag- ward-draining paleocanyons that cut into the inactive at low dips. New, steep faults break through the intact hanging wall as old faults nitude extension, and must restore to a well- southwestern part of the Colorado Plateau become inactive and are abandoned on the de®ned structural datum, the area provides an show that a basement high had formed to the uplifted footwall (Buck, 1988). (C) Synchro- exceptional opportunity to study the kinematic south of Lake Mead by Eocene time (Young, nous slip model; after some slip and rotation evolution of a continental normal fault system. 1966, 1979). Paleozoic and Mesozoic strata of initially steep faults, younger steep faults were erosionally removed from this basement break and sole into the rotated older faults. GEOLOGIC SETTING high. The exact age and extent of this high All faults remain active and rotate domino- remain unclear, due to incomplete understand- style, as well as rotating as a result of iso- During Late Jurassic and Cretaceous time, ing of the age and preextensional con®gura- static uplift of the thinning region. the Lake Mead area was part of the foreland tion of the sub-Tertiary unconformity.

1376 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

The present physiographic features of the area result primarily from Miocene extension. During Miocene time, extension initiated at the edge of what is now the Colorado Plateau, accommodated on a set of west-dipping nor- mal faults. These faults strongly control the topography of the South Virgin Mountains, as discussed above and illustrated in Figure 3. The timing of extension in the South Virgin Mountains is constrained by ®ssion-track ages from across the Gold Butte crystalline block. Fitzgerald et al. (1991) documented apatite ®ssion-track ages, interpreted as extensional unroo®ng ages, of ca. 15 Ma across the north- ern end of the block. Extension in the eastern Lake Mead region was accompanied by little or no magmatism. The only Tertiary volcanic rocks in the region, excluding thin tuff beds and rare ma®c vol- canic vents, are the postextensional basalts of Gold Butte (9.15±9.46 Ma; Cole, 1989) and Grand Wash Trough (3.99±6.9 Ma; Feuerbach et al., 1993).

CRYSTALLINE ROCK UNITS

Proterozoic X

The oldest rock units in the crystalline com- plex include the 1.7±1.8 Ga (Wasserburg and Lanphere, 1965; Bennett and DePaolo, 1987; work of L.T. Silver as discussed by Stewart, 1980) orthogneiss and garnet gneiss, which account for most of the metamorphic country rock. Three other rock types of inferred Pro- terozoic X age include hornblende granite, leucogranite, and ultrama®c rocks; the two granites locally exhibit gneissic high-temper- ature foliation. Garnet gneiss, the most abundant of the metamorphic rocks, contains garnet, biotite, cordierite, sillimanite, plagioclase, quartz, her- cynite, and magnetite (Volborth, 1962; Thom- as et al., 1988; Fryxell et al., 1992). This unit is extensively migmatitic, with 1-cm- to Ͼ10- m-thick layers of coarse-grained quartz, feld- spar, and garnet; the remainder of the unit is gneissic and has a well-developed foliation with no lineation (Fryxell et al., 1992).

Proterozoic Y

The Gold Butte Granite (1.45 Ϯ 0.25 Ga; Silver et al., 1977) is a large pluton of rapakivi

Figure 3. Simpli®ed geologic map of the South Virgin Mountains, Nevada and Arizona. The eight major fault blocks, listed in ascending structural order, are the Wheeler Ridge, the Iceberg Ridge, the Indian Hills, the Connoly Wash, the Azure Ridge, the Tramp Ridge, the Lime Ridge, and the Maynard Spring blocks. Each of these is bounded by a normal fault with 1 km or more of throw.

Geological Society of America Bulletin, September 2000 1377 BRADY et al. granite that forms about 30% of the basement STRATIFIED ROCK UNITS throughout the map area due to variability in outcrop (Volborth, 1962; Fryxell et al., 1992). depositional thickness as well as probable tec- It shows a distinctive texture of large (1±4 Paleozoic±Mesozoic tonic thickening and thinning, particularly cm), early crystallized potassium feldspar within the incompetent mudstones and ®ne phenocrysts, surrounded by a matrix of later A 2.5±3.5-km-thick succession of sedimen- sandstones of the Hermit Formation. Above crystallized minerals, including plagioclase, tary rocks (Fig. 4) is separated from the crys- this is the upper succession of limestone, quartz, biotite, hornblende, and sphene (Vol- talline basement rocks by the sub-Cambrian made up of the Permian Toroweap and Kaibab borth, 1962). Rapakivi texture, with plagio- unconformity. These strata compose a thin Formations (middle Permian; McKee, 1938). clase rimming the alkali feldspar phenocrysts, cratonal section just east of the Cordilleran It has a maximum thickness of ϳ270 m and is common, but not ubiquitous. Potassium miogeocline. thins toward the southeast, reaching zero feldspar phenocryst abundance and the intru- The strata comprise four major sequences; thickness within the southeastern part of the sive geometry of the granite change signi®- the lower two being carbonate dominated and South Virgin Mountains. the upper two being clastic dominated. The The boundary between the second and third cantly from west to east across the pluton lowest sequence is bounded at its base by the sequences is de®ned by the middle Permian± (Fryxell et al., 1992). The phenocryst content Proterozoic to Lower Cambrian unconformity, Lower Triassic unconformity. The third se- increases from typically Ͻ20% in the western and is composed of a basal clastic succession quence is composed entirely of clastic rocks outcrops to nearly 70% in the easternmost out- and an overlying carbonate succession. The with a maximum total thickness of about 1100 crops. The western outcrops form a complex basal succession includes the Lower Cambrian m, and a minimum thickness of 0 m, the wormy pattern within the metamorphic coun- Tapeats Sandstone and overlying shales and southeasternmost outcrops occurring on the try rock, whereas the eastern outcrops are part limestone of the Middle Cambrian Bright An- east side of Azure Ridge. It includes the Low- of a large continuous body, with clearly de- gel Formation (Schenck and Wheeler, 1942; er to Middle Triassic Moenkopi and Chinle ®ned contacts. Wheeler, 1943; McKee, 1945; Wheeler and Formations (Gregory, 1915; Gregory and Wil- Diabase dikes, reported from the southern Beasley, 1948). The thickness of this succes- liams, 1947; Poborski, 1954; Stewart, 1957), part of the crystalline block, are probably of sion is ϳ140 m, with signi®cant variability as well as the Moenave and Kayenta Forma- Middle Proterozoic age, on the basis of re- due to tectonic thickening and thinning of the tions undifferentiated (Marzolf, 1990). The gional correlation with other diabase dikes Bright Angel shales. The overlying carbonate- Moenave and Kayenta Formations undiffer- (Howard, K., unpublished data). Also, numer- dominated succession varies in thickness from entiated were combined with the Chinle For- ous mica-bearing pegmatite dikes intrude ϳ650 m in Lime Ridge to ϳ520 m in Wheeler mation for mapping purposes, because they many of the mappable units but, due to their Ridge. It is entirely composed of the Middle are discontinuous and dif®cult to distinguish generally small outcrop area, they were not to Upper Cambrian (McKee and Resser, 1945; in the ®eld. mapped by Fryxell et al. (1992) or earlier Longwell, 1949; Longwell et al., 1965) Bo- The third and fourth sequences are separat- workers. The age of these dikes is uncertain, nanza King Formation, the lowest member of ed by the basal Jurassic unconformity. The because they have not been reliably isotopi- which is limestone (the Papoose Lake Mem- fourth sequence includes only the Jurassic Az- cally dated (for a more detailed discussion, see ber); the remainder of the unit is dominantly tec Formation. The Aztec Formation is nearly Fryxell et al., 1992). dolostone. 200 m thick in the northern part of the South The top of the ®rst sequence, and the base Virgin Mountains, but pinches out to 0 m of the second sequence, occurs at the Upper thickness within the central South Virgin Mesozoic±Cenozoic Cambrian to Devonian unconformity. The sec- Mountains, and is not exposed anywhere ond sequence is composed of a carbonate south of Tramp Ridge (Fig. 3). dominated succession, overlain by a clastic All thickness estimates presented herein are The biotite-muscovite granite that crops out succession, in turn overlain by another car- based on averaging the mapped thickness near the western edge of the crystalline com- bonate succession. The lower carbonate suc- from several locations. Most of these strata, as plex was previously inferred to have a Prote- cession is ϳ800 m thick, and is composed of they occur in the South Virgin Mountains, rozoic X age, although ®eld relations were the Devonian Mountain Springs and Sultan were described in more detail by Morgan Formations (McNair, 1951), the Mississippian (1968) and Matthews (1976), as well as in a ambiguous (Fryxell et al., 1992). A new U/Pb Monte Cristo Formation (Langenheim, 1963), measured section by S.M. Rowland and V.S. monazite date of 65.2 Ϯ 0.6 Ma shows that the Pennsylvanian Callville Formation (Long- Korolev (unpublished data). this is a latest Cretaceous to earliest Paleocene well, 1921; McNair, 1951; Lumsden et al., intrusion (work of M. Martin, discussed in 1973), and the Permian Pakoon Formation Tertiary Brady, 1998). It is a ®ne- to medium-grained (McNair, 1951; McKee, 1975). All of these equigranular to pegmatitic biotite-muscovite units are generally described as limestones, Two major sequences of Tertiary strata are granite with relatively abundant apatite and but the Devonian and Mississippian units tend present; one is tilted and disconformably over- monazite as accessory minerals. to be heavily altered to dolostone in the South lies Paleozoic and Mesozoic strata, and the In addition to the biotite-muscovite granite, Virgin Mountains. The middle clastic succes- other is more or less ¯at-lying, and in angular a number of small postdiabase leucogranites sion has an average thickness of about 530 m unconformity on all older rocks. The tilted se- crop out in the southern part of the Gold Butte and is composed entirely of Permian rocks quence comprises the Rainbow Gardens and crystalline block. These may be of Mesozoic (White, 1929; McKee, 1933, 1969, 1975). It Thumb Members of the Miocene Horse or Cenozoic age, since regionally there are no includes the Queantoweap Sandstone, the Her- Spring Formation (Bohannon, 1984; Beard, known postdiabase Proterozoic intrusions mit Formation, and the Coconino Sandstone. 1996). The thickest exposed Horse Spring sec- (Howard, unpublished data). All three of these units vary in thickness tion, east of Tramp Ridge, is ϳ600 m. The

1378 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

Figure 4. Stratigraphic sec- tion from the South Virgin Mountains; thicknesses were measured from 1:12 000 scale ®eld maps.

Geological Society of America Bulletin, September 2000 1379 BRADY et al. basal Horse Spring beds dip nearly concor- Wash Trough (ca. 4±7 Ma; Cole, 1989), but is apparently truncated by the middle Miocene dantly with underlying Paleozoic strata, and these basalts are not seen above the Hualapai Bitter Ridge fault (Campagna and Aydin, therefore predate large-magnitude extension Limestone. 1994). and tilting. However, the lowest of these beds Conglomerates similar to those included in The extensional structures within the South exhibit growth-fault fanning of dips where the rocks of the Grand Wash Trough are abun- Virgin Mountains are divisible into three they crop out near the north end of Wheeler dant on the west ¯ank of the South Virgin categories; the ®rst includes large-offset, west- Ridge, and are therefore classi®ed as early Mountains, and are generally included in the dipping normal faults that bound large ridge- syntectonic strata. Bedding within the over- Muddy Creek Formation (Longwell, 1936; forming fault blocks, the second includes nu- lying Thumb Member fans slightly upsection, Bohannon, 1984). Near the western end of the merous, dominantly west dipping imbricate and is also considered to be syntectonic Gold Butte fault, these conglomerates are normal faults that internally disrupt the major (Beard, 1996). overlain by the Gold Butte basalt (ca. 9±9.5 fault blocks and usually sole into the ridge- The more or less ¯at-lying rocks of the Ma; Cole, 1989). Except where overlain by bounding faults, and the third includes steeply Grand Wash trough are in sharp angular un- the Gold Butte basalt, the ages of the con- east dipping faults that crosscut all other struc- conformity above the Horse Spring Formation glomerates are not constrained, and may be as tures (Figs. 5 and 6). In addition to the exten- (Lucchitta, 1967; Bohannon, 1984). They in- young as Quaternary (the Muddy Creek For- sional structures, the South Virgin Mountains clude a unit composed of red sandstone and mation can not be older than 10.6 Ma, and is contain three steeply dipping, southwest- siltstone with occasional tuff beds, overlain by generally younger than 8 Ma; Bohannon, northeast striking faults and three steeply inter®ngering gypsum, conglomerates, and ba- 1984). plunging folds. salts. The red sandstone unit (ca. 11.9±10.6 There are younger conglomerates, gravels, Slip lineations from all three sets of exten- Ma; Bohannon, 1984) is approximately cor- and sands stratigraphically above the rocks of sional faults suggest that they are predomi- relative in age with the red sandstone unit of the Grand Wash Trough and the Muddy nantly dip-slip normal faults (Fig. 7). Al- White Basin and Frenchman Mountain in the Creek(?) conglomerates. Some of these de- though individual slip lineations may not western Lake Mead region, but was probably posits, including an outcrop of Chemehuevi reliably indicate fault slip direction, the large not part of a connected depositional system, Formation at Sandy Point and scattered round- sample size (n ϭ 79), measured on more than because all three localities were apparently stone gravels, are probably ancient Colorado 35 different fault surfaces, with relatively tight isolated basins at the time of deposition (Bo- River deposits. There are also numerous levels clustering of lineation trends, makes the inter- hannon, 1984). A succession of generally ¯at- of Quaternary alluvium. Within individual preted slip direction robust (note that the 2␴ lying conglomeratic units overlies the red drainages two to four different levels of Qua- variation about the mean slip lineation is be- sandstone in the Grand Wash Trough. The ternary alluvial terraces are common, but tween trends of 262Њ and 330Њ). When consid- conglomerates were divided into three types these can not easily be correlated from drain- ered in combination with the mean dip direc- by Lucchitta (1967); one includes clasts of age to drainage. tion of normal faults (302Њ,nϭ 147, 2␴ Gold Butte granite as well as other crystalline variation between 261Њ and 311Њ), the mea- clasts, a second includes crystalline clasts but STRUCTURE sured slip lineations strongly suggest dip-slip no Gold Butte granite, and a third includes motion. dominantly sedimentary clasts. Within the The South Virgin Mountains were essen- mapped area, most of the conglomerates con- tially undeformed by Mesozoic compressional Ridge-Bounding Faults and Related Folds tain clasts of Gold Butte granite. The sedi- events, but strongly deformed by Tertiary ex- mentary-clast±dominated conglomerates oc- tension and strike-slip faulting. The only South of the Lime Ridge fault (Fig. 5), the cur closer to the Grand Wash Cliffs, and the structures within the mapped area that were South Virgin Mountains include seven major conglomerate that has crystalline clasts, but no probably formed by Mesozoic compression ridge-forming fault blocks, bounded by nor- Gold Butte granite clasts, occurs farther south, are a series of small-amplitude (ϳ10 m) folds mal faults that have each accommodated 1 km in the Grand Wash Trough (Lucchitta, 1966; in Permian to Triassic strata of northeastern or more of offset (Figs. 5 and 6). These faults, Bohannon, 1984). The Hualapai Limestone Lime Ridge (Fig. 5)1. North of the mapped listed from east to west, include the Grand overlies these conglomerates; the age of the area (but still within the South Virgin Moun- Wash fault zone, the Wheeler Ridge fault, the limestone is not directly known, but has been tains), there is a thrust fault that places Penn- Iceberg Canyon fault, the Indian Hills fault, inferred based on its relationship to datable sylvanian strata over Triassic and Jurassic the Million Hills Wash fault, the Garden Wash rocks elsewhere (ca. 10.8±3.8 Ma; Blair, 1978; strata, adjacent to the Bitter Ridge fault (Fig. fault, the Lime Canyon fault, and the Maynard Blair and Armstrong, 1979; Bohannon, 1984; 3). This may be a Mesozoic thrust, because it Spring fault. Faulds et al., 1997). In places, the red sand- The Grand Wash fault zone, easternmost of stone and fanglomerates of the Grand Wash 1Figure 5 is on a separate sheet accompanying the major faults, separates the Wheeler Ridge Trough are overlain by the basalts of Grand this issue. block from the adjacent Colorado Plateau

Figure 6. Vertical cross sections and corresponding reconstructions through the South Virgin Mountains, Nevada and Arizona. The -cross sections were constructed parallel to the inferred extension direction. (A) Cross-section A±A؅, extending west-northwest±east -southeast, from western Lime Ridge to the east side of northern Azure Ridge. (B) Cross-section B±B؅, extending west-northwest±east southeast, from the west side of the Connoly Wash block to the western edge of the Colorado Plateau. (C) Restoration of A±A؅, excluding blocks west of the Maynard Spring fault, which have undergone large vertical-axis rotations and do not restore properly under the .assumption of two-dimensional plane strain. (D) Restoration of cross-section B±B؅ M

1380 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

Geological Society of America Bulletin, September 2000 1381 BRADY et al.

(Fig. 5). Because the strata of Wheeler Ridge are downthrown by about 3.5 km relative to the plateau and folded into a roll-over anti- cline, at least one, and possibly more, normal faults must separate Wheeler Ridge from the edge of the plateau (Fig. 6, section B±BЈ). However, the fault or faults are nowhere ex- posed due to burial by sedimentary rocks of the Grand Wash trough. The next ridge-bounding fault to the west of the Grand Wash fault zone is the Wheeler Ridge fault (Fig. 5). It was ®rst mapped by Longwell (1936) as a 60Њ west-dipping fault where it crossed the Colorado River, in out- crops now submerged beneath Lake Mead. The normal separation on this fault where it crosses Lake Mead is ϳ2.1 km. The Wheeler Ridge fault seems to continue to the north as a single fault plane for a minimum of ϳ20 km and to the south for ϳ8 km (Fig. 5). Beyond the north end of the clearly exposed bedrock of Wheeler Ridge, the fault remains visible for about 12 km, juxtaposing strata of the Horse Spring Formation and rocks of the Grand Wash Trough in its footwall against younger (late Tertiary?) sedimentary strata in its hang- ing wall. This is probably a depositional, rath- er than tectonic contact, with the younger stra- ta lapping against the fault scarp. Farther to the north, the Wheeler Ridge fault becomes buried under younger Tertiary to Quaternary conglomerates and gravels, so its location is Figure 7. Fault orientation and slip lineation data. (A) Kamb contoured stereoplot of poles unknown. To the south, the fault splits into to normal faults. These poles are from surfaces that were exposed and measured directly, several strands near the south ends of Wheeler (with no three-point solutions. Note that mean fault dip direction is 302؇, or N58W. (B and Iceberg Ridges. Some of the displacement Kamb contoured stereoplot of measured slip lineations; mean slip direction is ϳ301؇,or on the Wheeler Ridge fault is transferred to N59W. (C) Rose diagram of slip lineation trends. Together with A and B, this suggests the Airport fault (Matthews, 1976), which cuts deformation by dominantly dip-slip normal faulting, with extension oriented ϳ300؇ or across to the east side of Wheeler Ridge, and N60W. (D) Stereoplot of slip lineations and fault planes from the Gold Butte fault, sug- becomes buried under Grapevine Mesa (Fig. gesting left-oblique normal faulting. 5). The rest of the displacement is distributed onto two major, and a number of minor, fault strands that lose displacement southward and eventually merge again near South Cove. One southward continuation of the Wheeler Ridge fault, which dips between 0Њ and 12Њ to the of the major faults is the South Cove fault, fault (Faulds et al., 1997). This southern por- east and therefore has a very sinuous fault which was originally mapped by Matthews tion of the fault offsets the Hualapai Lime- trace that runs generally north-northeast (1976) as a southwest-northeast±striking left- stone by more than 300 m, and folds it into a through the Indian Hills. It forms several klip- lateral strike-slip fault that cut entirely across roll-over anticline. pen of Cambrian rocks on Proterozoic base- southern Iceberg Ridge. More detailed map- The next major fault to the west of the ment and of Mississippian strata on Cambrian ping has shown that this fault turns southward Wheeler Ridge fault is the Iceberg Canyon (Fig. 5). Where it crosses the line of cross- and runs along the east side of southern Ice- fault, ®rst mapped by Longwell (1936, 1945), section A±AЈ shown in Figure 6, it accom- berg Ridge, and is continuous with the Sun®sh prior to the ®lling of Lake Mead. This fault modates a normal separation of the Tapeats Cove fault of Matthews (1976). Both parts of has a normal separation of ϳ1.2 km, is gently Sandstone of ϳ1.5 km. Hanging-wall bedding this fault are included in the South Cove fault listric in cross section, and cuts bedding in its cutoff angles range from about 50Њ to 70Њ, ex- in Figure 5. The other major strand is the hanging wall at 60Њ±90Њ. It dips 35Њ to the cept in part of the western Indian Hills, where Sheep Canyon fault, which merges with the west where it cuts through Permian limestones the fault forms a hanging-wall ¯at for ϳ200 South Cove fault east of South Bay. South of in its hanging wall, and its dip decreases to m in the shales of the Bright Angel Formation. South Bay, these faults are buried under youn- 10Њ to the west where it cuts Mississippian West of this hanging-wall ¯at, the fault again ger sedimentary rocks. Farther to the south, strata in its hanging wall (Longwell, 1936, cuts downsection to the west, giving it an the fault that carries the Gregg basin in its 1945). overall ramp-¯at-ramp geometry. Immediately hanging wall has been interpreted to be the The next fault to the west is the Indian Hills above this ¯at is a tight roll-over anticline that

1382 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

Figure 8. Photograph looking north at a sharp roll-over in the western Indian Hills. The fold formed above a ramp-¯at-ramp in the Indian Hills fault. Strata involved are upper Bright Angel Formation to middle Bonanza King Formation. Field of view across the middle of the photograph is ϳ1 km. See the legend of Figure 9 for an explanation of stratigraphic labels.

affects the upper unit of the Bright Angel For- gently to the northwest; however, there are outcrop immediately east of Horse Spring mation and higher Cambrian units (Fig. 8). few exposures of the fault plane, making di- Ridge, mapped by Fryxell et al. (1992) and The ®fth fault in the stack, and the last one rect measurement of its orientation dif®cult. Morgan (1967) as Permian limestone, is in- to crop out east of the Gold Butte crystalline Furthermore, because the Garden Wash fault stead the Mississippian Anchor Member of the block, is the Million Hills Wash fault, which is linked to the east end of the Gold Butte Monte Cristo Formation and may be part of a apparently increases in offset from south to fault, and splays off it to the north, the ori- megabreccia deposit within the Thumb Mem- north. It has a well-exposed surface trace that entation of the fault is presumably changing ber of the Horse Spring Formation. Just east de®nes the south end and southwest margin of rapidly as it bends northward near the east end of this outcrop, a large area of Thumb Mem- the Azure Ridge block. It has an average dip of Azure Ridge. This bending of the fault ber of the Horse Spring Formation is irregu- of ϳ12Њ to the northwest, and is slightly listric plane, combined with a limited number of larly mantled by a thin veneer of gravel (Fig. in cross section. Bedding cutoff angles in the widely scattered outcrops, makes it dif®cult to 5). Farther east, ϳ1.4 km north of Azure hanging wall are typically between 70Њ and precisely determine the fault-plane orientation Ridge, small outcrops of Permian Toroweap 90Њ. Normal offset on the Million Hills Wash from its surface trace. Detailed mapping of Limestone were found directly along strike fault is Ͻ100 m where it occurs as a small Azure, Tramp, and Lime Ridges generally cor- with those to the south in Azure Ridge (Fig. klippe above the Connoly Wash block, in- roborates the earlier mapping of the Garden 5). These outcrops are interpreted to be part creasing to ϳ1.1 km at the south end of Azure Wash fault by Fryxell et al. (1992), with only of a continuous ridge of Permian strata, which Ridge, and to slightly greater than 2.4 km un- a few modi®cations. At the northern end of is visible as a pale stripe on air photos for at der central Azure Ridge (Figs. 5 and 6). Azure Ridge, Fryxell et al. (1992) mapped a least 3.1 km to the north of Azure Ridge. This Northwest of the Million Hills Wash fault segment of the Garden Wash fault as a low- partially buried ridge of Permian limestone is the Garden Wash fault, which is responsible angle structure placing Mississippian Redwall has an east-west separation of ϳ8 km relative for translating Tramp Ridge, Lime Ridge, and Limestone (equivalent to the Monte Cristo to equivalent strata on Tramp Ridge. the northernmost outcrops of Azure Ridge Formation) over Cambrian dolostone. This It is proposed here that the Garden Wash westward relative to the rest of Azure Ridge was remapped as Permian Toroweap and Kai- and Gold Butte faults comprise part of the (Fryxell et al., 1992). The outcrop pattern of bab Limestones over Cambrian Bonanza King South Virgin±White Hills detachment (Due- the Garden Wash fault suggests that it dips dolostones. In addition to this change, a small bendorfer and Sharp, 1998), along with the

Geological Society of America Bulletin, September 2000 1383 BRADY et al.

Lakeside Mine fault zone (Fryxell et al., 1992), the Salt Spring Wash fault, and the Cy- clopic detachment (Cascadden, 1991), because these form an apparently continuous series of top-to-the-west Miocene normal faults (Fig. 3). The Garden Wash fault accommodated ϳ8

km of westward translation of Tramp Ridge lowly dipping faults relative to Azure Ridge and is probably linked southward, via the Gold Butte fault, with the Lakeside Mine fault zone, which unroofed the Gold Butte crystalline block ca. 15 Ma (Fitz-

gerald et al., 1991; Fryxell et al., 1992). from Figure 6A. Across the entire ؅ Although the Lakeside Mine fault zone is par- tially buried under younger sediments, it ap- pears to be continuous southward with the Salt Spring Wash fault, which also exposes Prote- rozoic crystalline rocks in its footwall and car- ries ca. 16 Ma growth-fault strata in its hang- ing wall (Cascadden, 1991). Still farther to the south, the Cyclopic detachment also has Pro- terozoic crystalline rock in its footwall and is probably continuous with the Salt Spring Wash fault (Cascadden, 1991). The Lime Canyon fault is west of the Gar- den Wash fault. It dips to the west at ϳ10Њ, and accommodates ϳ8 km of westward sep- aration of the Paleozoic section in western Lime Ridge relative to the corresponding sec- tion in Tramp Ridge (Figs. 5 and 6). The Lime Canyon fault appears to have evolved from two major faults and a number of minor anastomosing splays, which all merged downward. This interpretation is based on a retrodeformable cross section of the Lime Canyon and Perkin's Spring area (Figs. 6 and 9). The westernmost of the two main splays crops out clearly around the head of Lime Canyon, where it is a ¯at-lying fault, carrying relatively simple imbricate fault blocks of east-dipping Mississippian and Pennsylvanian strata in its hanging wall, with bedding cutoff angles between 35Њ and 45Њ (Figs. 5 and 9). The eastern splay crops out on westernmost Tramp Ridge, where it carries Lower Cambrian units down against Protero- zoic basement (Fig. 6, cross-section A±AЈ). This splay projects just above the basement block that separates Tramp Ridge from Lime Ridge, and crops out again as a detachment that underlies the Perkin's Spring area (Figs. 5, 6, and 9). In order for this projection to work, the offset on the faults that bound the east and west sides of the basement block be- tween Tramp and Lime Ridges must ®rst be restored, suggesting that the basement block is a later uplifted horst. The offset on these horst-bounding faults is relatively small (Fig. 9), although offset on the eastern fault increas- es signi®cantly to the north. Near the north

end of the basement block, the upper part of Figure 9. Detailed crosssection, section and shallowly restoration dipping of normalare the faults Perkin's all separate Spring considered area. a to This thin be cross veneer part section of of coincides the extensionally with part Lime imbricated of Canyon Paleozoic cross-section fault, rocks A±A although from only Precambrian the basement. western These strand, shal which crops out at Lime Canyon, is labeled on the section.

1384 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

bounding faults in so far as the strike of hang- ing-wall strata is markedly different from footwall strata. Hanging-wall strata strike ϳN65E, whereas strata in the adjacent parts of Lime Ridge strike due north, suggesting clockwise vertical-axis rotation of ϳ65Њ of the hanging-wall rocks in addition to extension.

Minor Imbricating Faults and Related Folds

Figure 10. Schematic illustration of the sequential deformation of small fault slivers within Numerous smaller offset faults are found a west-dipping, top-to-the-west normal fault system. Note that bedding within the initial within each of the major fault blocks. These fault sliver is rotated into a westward dip and broken by a small top-to-the-east normal faults typically have offsets in the range of 5± fault. 100 m; a few have offsets of several hundred meters. They accommodate internal deforma- tion and extension of the major fault blocks the Cambrian section in Tramp Ridge is jux- east-dipping panel of the Permian Toroweap and usually merge downward with the major taposed against basement, requiring ϳ600 m Formation, carried on an east-dipping fault. ridge-bounding normal faults (Fig. 11). They of throw on the fault. The arrangement of fault slices near Per- dip more steeply and have higher bedding cut- The western splay of the Lime Canyon fault kin's Spring may have been achieved by col- off angles than the ridge-bounding faults. may have a ramp-¯at-ramp geometry, which lapsing a stack of imbricate fault slices, Most of these smaller offset faults have hang- would explain the relatively ¯at-lying, little bounded by predominantly west dipping, ing-wall bedding cutoff angles of ϳ90Њ; some extended strata of northern Lime Ridge. At west-side-down normal faults, onto a shallow- are Ͼ115Њ (Fig. 12). Lime Canyon, the fault cuts bedding at a mod- ly dipping detachment (Fig. 9). In order to Nearly all of the smaller offset faults merge erate angle (ϳ40Њ), and therefore cut through achieve their present con®guration, the imbri- with, or are truncated by, shallowly dipping, this part of the section with an initially mod- cate slices must have initiated in an eastward- large-offset, ridge-bounding faults. Of 33 cas- erate west dip. Its initial dip is less well con- stepping and downward-stepping sequence. es where a more steeply dipping fault clearly strained where the fault cuts through the rest This would result in slices of stratigraphically intersects a shallowly dipping ridge-bounding of the Paleozoic section; however, creation of higher Pennsylvanian Callville Formation be- fault, the more steeply dipping fault is trun- an area-balanced cross section seems to re- ing carried downward, and to the west of, the cated by, or merges with, the shallowly dip- quire the fault to have ¯attened to dips as low stratigraphically lower slices from the Banded ping fault in 31 cases. The two more steeply as 15Њ±20Њ where it cut through the lowest part Mountain Member of the Bonanza King For- dipping faults that do cut a shallowly dipping, of the section. Therefore, it was apparently a mation, which would in turn be carried down- block-bounding fault are close together, sub- shallowly dipping listric fault that included a ward and to the west of slices of the Papoose parallel, and each accommodates ϳ30mof ¯at near the base of the Cambrian section (Fig. Lake Member of the Bonanza King Forma- west-side-down normal separation. They cut 6). A northward increase in the length of this tion. All of these fault slivers would then be through the southern end of the Mississippian ¯at might result in a wide block of hanging- carried westward and downward against Pro- on Cambrian klippe formed by the Indian wall strata being carried westward over a ¯at terozoic basement rocks. The occurrence of Hills fault just north of Devil's Cove (Fig. 5). surface, undergoing only minor internal dis- west-dipping panels, separated by east-dip- The intersection of these faults with the un- ruption and rotation. ping, top-to-the-east normal faults, can be ex- derlying Iceberg Canyon fault is not exposed, Near Perkin's Spring, between Tramp Ridge plained by top-to-the-west simple shear within so it is unclear whether they also cut the Ice- and Lime Ridge, the Lime Canyon fault is the fault-bounded imbricate slices (Fig. 10). berg Canyon fault. nearly ¯at-lying, and its hanging wall is struc- Within each fault sliver shown in Figure 9, it Nearly all of the minor imbricating normal turally complex (Figs. 5 and 9). To the west is reasonable to expect a top-to-the-west ro- faults are synthetic to the ridge-bounding de- of Perkin's Spring, a set of west-dipping, top- tation of blocks, bounded by top-to-the-east tachments. There are a few antithetic faults, to-the-west normal faults within the hanging faults. all of which have been rotated so that they wall of the Lime Canyon fault carry east-dip- The westernmost ridge-bounding fault is now appear to be reverse faults (the two most ping panels of Cambrian strata down against the Maynard Spring fault. It separates roughly signi®cant of these can be seen in cross-sec- Proterozoic basement. These panels of east- north striking strata of Lime Ridge from tion A±AЈ of Fig. 6; one under eastern Azure dipping Cambrian strata are overlain by a northeast-striking strata that comprise the Ridge, and one under central Tramp Ridge). west-dipping panel of the Pennsylvanian Call- westernmost outcrops of the range (Fig. 5). Two large-amplitude, broadly folded, roll- ville Formation, carried on a west-dipping This fault is well exposed for about 1 km over anticlines in the strata of the western half normal fault. The structures to the east of Per- south of the Lime Ridge fault. Farther south of Azure Ridge and the eastern part of Lime kin's Spring dip in the opposite direction to it becomes buried under alluvium. For an ad- Ridge are associated with sets of small-offset those just described from the west of Perkin's ditional 1.4 km to the southeast, small expo- synthetic normal faults (cross-section A±AЈ, Spring. To the east, a set of east-dipping, top- sures of fault-bounded Permian Toroweap Fig. 6). These folds are cut by fanning sets of to-the-east normal faults carry west-dipping Limestone are found; these are probably part synthetic normal faults, in which the faults to panels of Cambrian strata down against Pro- of the hanging wall of the Maynard Spring the west dip more steeply than those to the terozoic basement. These are overlain by an fault. This fault differs from other ridge- east. Rather than being due to measurable

Geological Society of America Bulletin, September 2000 1385 BRADY et al. s Wash fault. The to the east. Hanging-wall bedding cut-off angles are ؇ and 60 ؇ to the west; bedding mostly dips between 45 ؇ and 40 ؇ . (B) Photograph looking north at the south end of Azure Ridge, showing approximately the eastern half of the area mapped. The ؇ and 100 ؇ Figure 11. (A) Mapmoderately of dipping the faults southern typically tip dip of Azure between Ridge, 20 showing anastomosing, moderately dipping faults soling into the nearly horizontal Million Hill mostly between about 65 Million Hills Wash fault extends along the break in slope at the base of the Paleozoic strata.

1386 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

Figure 11. (Continued).

Figure 12. Photograph looking north at moderately west dipping imbricate normal faults in the hanging wall of the Indian Hills fault. These faults merge downward with the Indian Hills fault (not visible in this photograph). Note that the bedding cutoff angle is about 115؇ where Pennsylvanian strata are cut by the most steeply dipping fault (photograph is looking approximately along strike of this bedding-fault intersection). See the legend in Figure 9 for an explanation of stratigraphic labels.

Geological Society of America Bulletin, September 2000 1387 BRADY et al.

Figure 13. Photographs of late, east-side-down, east-dipping faults. Stratigraphic labels are explained by the legends shown in Figures 9 and 11. (A) West-dipping fault offset by ϳ40 m of slip along a surface parallel to the base of the Mississippian Anchor Member, central Indian Hills. (B) West-dipping fault offset by ϳ3 m of bedding-parallel slip within the Pennsylvanian Callville Formation, northern Azure Ridge. (C) Sliver of Mississippian Anchor Member along the Indian Hills fault, offset by a slip surface that is parallel to bedding in the structurally lower Cambrian Bonanza King Formation and subparallel to bedding in the structurally higher Penn- sylvanian Callville Formation, northeastern Indian Hills.

folding of any individual panel of strata, these talline block between Tramp Ridge and Lime cally highest Cambrian Bonanza King against antiforms result from a gradual westward de- Ridge. Proterozoic crystalline rocks. crease in dip of bedding from one fault block Many of the steep crosscutting normal A number of steeply dipping faults crosscut to the next. faults are localized along the basement-cover the Paleozoic strata east of the Gold Butte contact. An en echelon series of these faults crystalline complex, as bedding-parallel shear Crosscutting Faults is located along the western edge of Azure surfaces. These faults are dif®cult to observe, Ridge (Fig. 5 and 6). A similar fault was except where they offset younger normal The extensional structures recognized in the mapped by Longwell (1936) near the south faults (Fig. 13). They apparently decrease in South Virgin Mountains include a number of end of Iceberg Canyon, but it is now covered abundance and offset with increasing distance subvertical faults that cut all other normal by the waters of Lake Mead. The faults from the crystalline complex; there are no ex- faults and are upthrown on the side closest to bounding Azure Ridge dip vertically to steep- amples observed east of Iceberg Canyon. nearby crystalline blocks. East of the Gold ly eastward and have east-side-down normal Butte crystalline block these faults are consis- separations ranging from a few meters to Steep Northeast-Striking Faults tently east-side-down, i.e., they are upthrown Ͼ150 m. The fault near the south end of Ice- on the side closest to the Gold Butte crystal- berg Canyon, as described by Longwell Three major northeast-striking faults ap- line block. North of the Gold Butte block, (1936), is a steeply west dipping structure that peared on previous maps of the South Virgin within the Tramp Ridge±Lime Ridge area, places Proterozoic granite against mid(?) Bo- Mountains. From north to south, these include steeply east- and west-dipping crosscutting nanza King, making it a steep reverse fault the Bitter Ridge, Lime Ridge, and Gold Butte faults bound a horst of crystalline rocks be- with a minimum throw of 500 m. The faults faults (Anderson, 1973; Bohannon, 1979, tween Tramp Ridge and Lime Ridge (Fig. 5 that bound the horst between Tramp Ridge 1984; Beard, 1996) (Fig. 3). All were consid- and 6). Also, two steeply west-dipping faults and Lime Ridge accommodate between a few ered to be steeply dipping strands of the left- within Lime Ridge cut across moderately and meters and Ͼ500 m of normal separation, jux- lateral Lake Mead fault system. Campagna shallowly dipping faults and are upthrown on taposing Paleozoic strata ranging from the and Aydin (1994) showed that the Bitter the east side, i.e., the side closest to the crys- Cambrian Tapeats Sandstone to stratigraphi- Ridge fault is a subvertical structure with

1388 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

Figure 13. (Continued).

slickenlines plunging gently to the southwest, steeply tilted ridges adjacent to fault zones. constrained reasonably well using a variety of supporting the interpretation that this is a Two of these occur north of and directly ad- geologic arguments, as well as observations of strand of the Lake Mead fault system. How- jacent to the Gold Butte fault, producing right- Longwell (1936), made prior to the formation ever, Fryxell et al. (1992) reinterpreted the lateral de¯ection of Horse Spring strata in of Lake Mead. Gold Butte fault as a right-lateral tear fault in Horse Spring Ridge and of basal Cambrian The Grand Wash fault zone must include the hanging wall of the Garden Wash fault strata in southernmost Lime Ridge (Fig. 5). one or more west-side-down normal faults. zone, suggesting that it is not a branch of the The third fold is apparently left lateral, and is These faults are inferred to be relatively steep Lake Mead fault system. outlined by a gradual change in strike of bed- listric structures, rotated only a few degrees The Gold Butte fault was completely re- ding in the ridges east of the Gold Butte crys- from their initial orientations. The inference of mapped during this study. The mapping talline block. The average strike of bedding only minor rotation is based on the absence of showed it to be a vertical to steeply north dip- changes northward from ϳN30E in Wheeler structurally lower faults with major offset, and ping structure; only three clear slip lineations Ridge, Iceberg Ridge, and the southern Indian on the observation that the next higher struc- are observed, and these plunge steeply to the Hills to ϳN05E in northern Azure Ridge (Fig. ture (the Wheeler Ridge fault) is apparently west (Fig. 7D). 5). only slightly rotated from its original orien- New mapping along part of the Lime Ridge tation. The inference of initially steep dip and fault shows that it is not a simple, steeply Interpretation of Subsurface Geometries listric geometry is based on the observed roll- northwest dipping plane. Where remapped at of Concealed Faults over geometry of Wheeler Ridge, an inter- the north end of Lime Ridge (Fig. 5), the fault preted listric geometry on seismic re¯ection exhibits a sinuous trace, and based on its out- Most of the major normal faults of the data ϳ30 km farther south, beneath the Hu- crop pattern where it cuts up and over eastern South Virgin Mountains are well exposed, but alapai basin (Faulds et al., 1997), and analogy Lime Ridge, its dip is only ϳ50Њ to the north- three of the major ridge-bounding normal with higher and lower normal fault zones, west. faults or normal fault zones are concealed by where listric curvature is directly observed. alluvium along much or all of their traces. These include the Hurricane fault and Iceberg Steeply Plunging Folds These faults are the Grand Wash fault zone Canyon fault, both of which had initially and the Wheeler Ridge and Garden Wash steeply dipping listric geometries. The Hurri- There are three large steeply plunging folds faults. The late, steep, crosscutting fault of cane fault cuts the Colorado Plateau ϳ65 km within the South Virgin Mountains; all are southern Iceberg Canyon is also covered by east of the Grand Wash fault zone, is vertical clearly visible in map view and appear to be the waters of Lake Mead. The location, ge- where it cuts Permian strata, and shallows to either right-lateral or left-lateral de¯ections of ometry, and offset of all of these faults can be an ϳ60Њ westward dip where it cuts through

Geological Society of America Bulletin, September 2000 1389 BRADY et al. the lower part of the Cambrian section (Ham- blin, 1965). The 30Њ of shallowing across ϳ2 km of section is similar in curvature to the Iceberg Canyon fault, which shallows at a rate of ϳ16Њ/km. Our preferred interpretation of the Grand Wash fault zone is that it includes two or more normal faults, which have served to downdrop the east side of the Wheeler Ridge block by ϳ3.5 km relative to the Colorado Plateau (Fig. 6). Projection of beds from Wheeler Ridge to the inferred location of the westernmost strand of the Grand Wash fault zone suggests that almost all of this 3.5 km offset can be accom- modated on the westernmost strand. The amount off offset on the other fault or faults under Grand Wash Trough is unknown, but is probably not large on any one fault, because large offset on a listric fault should have re- sulted in hanging-wall rotation and conse- quent creation of a tilted fault-block ridge where the Grand Wash Trough currently exists. The location of the western strand of the Grand Wash fault zone is constrained by geo- metric arguments, but it is nowhere observed due to burial beneath younger sediments. This fault is shown in Figures 5 and 6 near its east- ernmost allowable position, because strata are shown as having cutoff angles slightly Ͼ90Њ. Hanging-wall cutoff angles of slightly Ͼ90Њ are reasonable for an initially west dipping fault because a small increase in cutoff angle could result from collapse of the hanging wall into the potential void adjacent to the listric fault surface. However, these angles would be- come unreasonably large if the fault was drawn farther to the east, and the bedding within the Wheeler Ridge block continues Figure 13. (Continued). eastward at a constant dip or with an east- ward-steepening dip due to roll-over. The westernmost allowable position of the fault strand is immediately east of the bedrock ex- that projects to surface near the eastern edge above the Grand Wash fault. This interpreta- posures of Wheeler Ridge, because it clearly of the Grand Wash trough can not be entirely tion is not preferred because: (1) it requires doesn't crop out on Wheeler Ridge. ruled out, but is not preferred. To be consistent the Grand Wash fault to have a geometry The location of the easternmost strand of with the interpreted shallow depth of the much different from that of better exposed the Grand Wash fault zone is inferred from a Grand Wash trough (ϳ1±2 km; Saltus and nearby analogs and different from its geome- number of small scarps in alluvium, ϳ1km Jachens, 1995), and to avoid hanging-wall cut- try as imaged by seismic re¯ection data to the from the exposed bedrock of the Grand Wash off angles of much greater than 90Њ with the south (Faulds et al., 1997); and (2) it requires Cliffs. These scarps are visible on air photos roll-over strata of Wheeler Ridge, the fault a diachroneity of slip on the major ridge- of the region, and are consistent with forma- would have to ¯atten at a depth of about 1 bounding normal faults, which can not be tion by a west-side-down normal fault, but do km. It would be required to slip earlier than ruled out with the current data. However, this not form a continuously exposed feature and the faults farther to the west because palin- seems unlikely, because there is no structural are therefore suggestive, rather than compel- spastic restoration of such a shallow ¯attening evidence for diachronous slip within the better ling, evidence. Other small-offset normal fault requires that the structurally higher faults exposed fault blocks farther to the west. The faults may exist under the Grand Wash be back-rotated through vertical to initial east available geochronologic evidence suggests Trough. If so, no surface expression of these dips. Because it is unlikely that the faults ini- that all of the major ridge-bounding faults faults has been observed. tiated as east-dipping reverse faults, this were active during a very short interval of An alternate interpretation of the Grand would suggest that they initiated after the stra- time (Brady, 1998). Wash fault zone as consisting of only one fault ta in Wheeler Ridge had been rolled over As discussed earlier, the surface trace and

1390 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM dip of the Wheeler Ridge fault, where it is ing-wall block changes northward. This may where it cuts through Mississippian±Pennsyl- now covered by Lake Mead, was constrained indicate a northward change in the initial ge- vanian strata, show that it initiated as a steeply by observations of Longwell (1936). Howev- ometry of the Garden Wash fault. Given that west dipping listric fault. The Wheeler Ridge er, its cross-sectional shape can only be in- the east-dipping strata along the east side of fault also clearly initiated with a steep west ferred. As with the fault(s) of the Grand Wash Tramp Ridge have a more or less constant dip, because it still dips 60Њ to the west. The fault zone, the Wheeler Ridge fault is inferred strike of about N20E, and the correlative strata Indian Hills, Million Hills Wash, and Maynard to have a listric geometry similar to the Hur- in the buried northward continuation of Azure Spring faults all have hanging-wall bedding ricane fault farther east and the Iceberg Can- Ridge apparently maintain a strike of about cutoff angles between 50Њ and 90Њ, consistent yon fault farther west. This interpretation is N10E, the westward displacement of strata with initially steep west dips. The hanging supported by the fact that the Hualapai Lime- within Tramp Ridge relative to strata of Azure walls to these faults have been internally de- stone forms a roll-over anticline in the hang- Ridge changes little from south to north. The formed, so it is dif®cult to establish whether ing wall of the Wheeler Ridge fault farther southern end of Tramp Ridge is signi®cantly they were initially listric, but the mapped re- south in Gregg basin (Faulds et al., 1997). internally extended on moderately west dip- lationships permit this possibility. Because of The Garden Wash fault, the northernmost ping normal faults, but farther north, it is very the preponderance of evidence that the large- part of the South Virgin±White Hills detach- little extended on steeply dipping normal offset, block-bounding normal faults of the ment, is mostly buried, yet its location and faults. If the Garden Wash fault initiated as a South Virgin Mountains initiated as steeply offset are constrained by geologic mapping steeply dipping listric fault near its south end, west dipping listric structures, it is inferred north of Azure Ridge. At least one major nor- but developed a ramp-¯at-ramp geometry far- that the buried Grand Wash fault zone and mal fault is to the east of, and roughly parallel ther to the north, and the entire length of the South Virgin±White Hills detachment also ini- to, Horse Spring Ridge. This fault separates fault accommodated a similar amount of east- tiated as steeply west dipping listric structures. the Tertiary strata that crop out on Horse west extension, it would result in a relatively Absolute ages of initiation are not known Spring Ridge and immediately to its east from highly extended southern Tramp Ridge and a for all of the block-bounding faults; it is there- the partially buried ridge of Permian Torow- wide, relatively untilted northern Tramp fore possible that there was some diachroneity eap Formation that continues north from Ridge, with a roll-over on its east side (Fig. in the timing of their initial slip. Nevertheless, Azure Ridge (Fig. 5). The normal offset on 5). A smaller scale analog is seen in the south- because they do not crosscut any other struc- this fault must be equal to or greater than the ern Indian Hills, discussed above, where the tures, they have similar spacings and offsets, 8 km separation of Permian strata in the Azure Indian Hills fault forms a ¯at in the Bright and other faults with lesser displacements and Ridge block from equivalent strata in Tramp Angel Shale, above which a roll-over anticline generally larger bedding cutoff angles sole Ridge. Farther south, across the Gold Butte is developed (Fig. 8). into them, they are interpreted to be the ear- fault, the South Virgin±White Hills detach- The steep reverse fault of southern Iceberg liest extensional structures to affect the region. ment is interpreted to have unroofed the Gold Canyon, described by Longwell (1936), is The smaller offset imbricating faults that in- Butte crystalline block. The crystalline block now covered by Lake Mead. This fault is the ternally disrupt the major ridge forming is a more or less intact upper crustal section, only signi®cant apparent compressional struc- blocks are interpreted to have initiated later. through which the unroo®ng fault cut with an ture reported from the South Virgin Moun- The hanging-wall cutoff angles for these faults initial dip of ϳ60Њ (Fryxell et al., 1992). The tains. Because all other faults in the area are are typically near 90Њ, but reach values Ͼ120Њ fault is not currently exposed on the west side normal faults, rather than reverse faults, it is (Fig. 12 and Table 1). This variation in cutoff of the crystalline block, except as a zone of reasonable to suggest that this structure was angles may result from variations in initial dip chlorite breccia near the margins of the block, also active as a normal fault. This fault was or time of initiation of these faults, higher cut- but presumably dips gently away from the probably active as a steeply east dipping nor- off angles resulting from faults initiating after domiform Gold Butte crystalline block. In or- mal fault, and was later rotated a few degrees some amount of rotation of the major fault der to denude the crystalline block, assuming into a steeply west dipping, apparently reverse blocks. These two possibilities are indistin- a60Њ initial dip of the denuding fault zone and orientation. This rotation could have occurred guishable, because none of the minor imbri- a currently exposed width of ϳ18 km (base- either by late-stage slip on the Wheeler Ridge cating faults cut each other. Rather, they tend ment plus cover), the minimum throw on the fault and/or Grand Wash fault zone, or as a to anastomose in map view and occasionally denuding fault zone is ϳ15 km (Fryxell et al., result of up¯exing of the crust adjacent to the are seen to splay upward. The cutoff angles 1992). This interpretation of very large normal uplifting Gold Butte crystalline complex. greater than 90Њ require that the faults either offset on the South Virgin±White Hills de- initiated as reverse faults, then rotated to be- tachment is supported by the fact that to the KINEMATIC EVOLUTION OF THE come normal faults, or initiated as normal north of the Gold Butte crystalline block, SOUTH VIRGIN MOUNTAINS faults after some amount of top-to-the-east where hanging-wall blocks are still present, NORMAL FAULT SYSTEM tilting of bedding had already occurred due to the cumulative westward translation of strata slip on the ridge-bounding faults. The latter in the Maynard Spring block relative to equiv- The large-offset, block-bounding normal explanation is preferable, since it simply re- alent strata in the Azure Ridge block is ϳ19 faults of the South Virgin Mountains appar- quires one evolving extensional fault system, km (Fig. 5). ently initiated as subparallel west-dipping lis- whereas the former requires extension fol- The total east-west extension accommodat- tric faults. The Iceberg Canyon fault is the lowed by compression, then resumed exten- ed by the Garden Wash fault apparently re- best documented and least disrupted of the sion, all accommodated on the same set of mains fairly constant from south to north, and large-offset, block-bounding faults. Its current faults, with no other preserved evidence of a the width of its hanging wall (Tramp Ridge) listric geometry and hanging-wall bedding compressional event. also remains constant; however, the amount of cutoff angles, which range from ϳ90Њ where In almost all cases where the intersection extension and structural style within the hang- it cuts through Permian limestones to ϳ70Њ can be seen, imbricating faults clearly merge

Geological Society of America Bulletin, September 2000 1391 BRADY et al. with the block-bounding faults, requiring syn- TABLE 1. MAXIMUM VALUE OF LOWEST ACTIVE DIP FOR THE MAJOR BLOCK-BOUNDING FAULTS OF THE GOLD BUTTE BREAKAWAY ZONE, SOUTH VIRGIN MOUNTAINS, NEVADA chronous slip of the two fault sets. Because they had to slip synchronously, the hanging- Bounding fault Maximum hanging Minimum Maximum value Maximum value wall cutoff angle for required rotation of lowest initial of lowest active wall bedding cutoff angles of the imbricating imbricating faults in of bounding fault dip for bounding dip for bounding faults can be used to constrain the dip at superjacent block (␤) (R) fault* (␣) fault (␪) which the block-bounding faults were active. Grand Wash fault zone 82Њ 0Њ *60Њ 60Њ As the ®rst-generation faults and adjacent Wheeler ridge fault 85Њ 0Њ *60Њ 60Њ fault blocks tilted, the hanging-wall cutoff an- Iceberg Canyon fault 114Њ 24Њ 60Њ 36Њ Indian Hills fault 100Њ 20Њ 48Њ 28Њ gles of later faults would increase. The mini- Million Hills Wash fault 124Њ 34Њ 45Њ 11Њ mum amount of tilting of bedding and adja- Garden Wash fault 99Њ 9Њ *60Њ 51Њ cent block-bounding faults, prior to initiation Lime Canyon fault 78Њ 0Њ 44Њ 44Њ of the later faults, can be calculated by assum- Note: Variables correspond to those used in Figure 13. *Assumes that initial dip did not decrease below base of Cambrian strata. ing that the later faults initiated with 90Њ dips *Assumed, base on geometry of Hurricane fault. (Fig. 14). For example, the later imbricating faults near the southern tip of Azure Ridge have bedding cutoff angles that range from ϳ70Њ to 124Њ. For the 124Њ case, if the initial dip of the splay was 90Њ or less, the Azure Ridge block and the faults bounding it must have tilted by at least 34Њ prior to initiation of the later imbricating fault. Thus, the block- bounding fault below Azure Ridge, which ini- tiated with its deeper portions dipping at about 60Њ, was actively slipping with dips at least as Figure 14. Diagram showing the relationship between the dip of active block-bounding low as 26Њ at the time of initiation of the in- normal faults and hanging-wall cutoff angles of bedding against the block-bounding and ternal imbricating fault. Similar arguments can imbricating normal faults. The ®gure shows the fault system at the time of initiation of be made for other block-bounding faults (Ta- an imbricating fault. As the system continues to extend, fault and bedding dips will change, hanging-wall cutoff؍␣ .ble 1). These data suggest that major block- but the bedding cutoff angles will remain more or less the same ,hanging-wall cutoff against the imbricating fault؍␤ ,bounding faults remained active at dips of against the block-bounding fault .dip of the active block-bounding fault؍initial dip of the imbricating fault, and ␪؍␥ .Ͻ20Њ If the smaller offset imbricating faults ini- The amount of rotation (R) of the block-bounding fault is the same as the amount of tiated only after the major fault blocks had rotation of bedding in its hanging wall. If bedding was initially horizontal, then the amount -؊␥. The dip of the block␤ ؍ been tilted to some extent, then these imbri- of rotation of the block-bounding fault is given by: R cating faults should be absent or at least less bounding fault is equal to its original dip minus its rotation; thus, its dip can be calculated ؊(␤؊␥). Because the imbricating normal faults had to initiate with␣؍ ؊R␣؍common in the relatively less tilted eastern using: ␪ 90؇ or less, a maximum value for the dip of the active block-bounding fault؍␥ fault blocks, and they should be more abun- dips of .(؊(␤؊90؇␣؍ dant to the west. In fact the abundance of (␪max) can be calculated using: ␪max small offset imbricating faults does increase from east to west, or upward in the structural stack of major fault blocks (Figs. 5 and 6). Only a few imbricating faults affect most of blocks, requiring a tear fault within the hang- of Tramp Ridge, and the steep north dip of the Wheeler Ridge and Iceberg Ridge. The struc- ing wall of the South Virgin±White Hills de- fault in the subsurface. These observations can turally higher and more tilted hanging wall to tachment. Because the change in offset on the be explained in at least two ways. (1) The tear the Iceberg Canyon fault includes a number of detachment coincides with the Gold Butte fault in the hanging wall of the South Virgin± imbricating faults but is still a nearly intact fault, the hanging-wall tear fault must have White Hills detachment coincided with a homoclinal block. The still higher and more been immediately above the Gold Butte fault. northward-deepening lateral ramp in the fault tilted fault blocks of the northern Indian Hills, This interpretation was suggested by Fryxell zone and/or (2) isostatically driven uplift of Azure Ridge, Tramp Ridge, and Lime Ridge et al. (1992), and implies that the Gold Butte the Gold Butte crystalline block was focused are all highly disrupted by imbricating faults. fault was part of, or at least connected to, a beneath the hanging-wall tear fault due to the As extension continued in the South Virgin right-lateral structure in the hanging wall of sharp gradient in amount of extension and Mountains, the South Virgin±White Hills de- the South Virgin±White Hills detachment. depth of unroo®ng. In either case, the Paleo- tachment eventually became the dominant However, the Gold Butte fault can not be sat- zoic strata of Tramp Ridge and Lime Ridge extensional structure. It accommodated the isfactorily explained as a tear fault that cut are offset in a left-lateral or left-lateral unroo®ng of the 15-km-wide Gold Butte crys- through only the hanging wall of a near-planar oblique-normal sense relative to the Gold talline block, as well as the ϳ8 km westward normal fault. A slightly more complex inter- Butte crystalline block, and in a right-lateral translation of Tramp Ridge relative to Azure pretation is required to account for the appar- or right-lateral oblique-normal sense relative Ridge. ently right-lateral vertical axis folds in the to the Paleozoic cover that has been trans- The hanging wall to the Gold Butte crys- ridges adjacent to the Gold Butte fault, as well ported westward off of the Gold Butte crys- talline block has been translated farther west- as the apparently left-lateral smearing of thin talline block. ward than the Tramp Ridge and Lime Ridge fault slivers of Permian strata at the south end The late, steep, crosscutting faults observed

1392 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM in the South Virgin Mountains are interpreted explains the steep north dip of the Gold Butte east-west, and was accommodated on steeply to be isostatic rebound structures that allowed fault, the steeply plunging lineations observed west-dipping, listric normal faults. After a sig- uplift of extensionally denuded crystalline on it (Fig. 7D), and the right-lateral map-view ni®cant amount of slip and rotation had been blocks in an area affected by regional exten- folds of southernmost Lime Ridge and Horse accommodated by the initial set of faults, a sional stresses. This interpretation is consis- Spring Ridge. second set of steeply west dipping, anasto- tent with the observed features of these faults, Alternatively, the Gold Butte fault may be mosing normal faults developed in the higher including the following. (1) They have accom- interpreted as a left-lateral strike-slip fault, ap- fault blocks. The later faults soled into the ear- modated vertical movements of as much as parently a splay of a left-lateral system of lier faults, rather than crosscutting them. This several hundred meters, but relatively little ex- faults known as the Lake Mead fault system requires that the early fault surfaces remained tension, as evidenced by their little-rotated (Anderson, 1973; Bohannon, 1979, 1984; active to very shallow dips while the later steep dips. (2) They are invariably upthrown Beard, 1996). This interpretation does not ex- faults were slipping. The early faults have ac- on the side nearest to large denuded crystal- plain the observed right-lateral folds, or the cumulated normal offsets that range from ϳ1 line blocks. (3) Their abundance and magni- slip lineations, and is inconsistent with the ap- km to perhaps Ͼ15 km, and have been rotated tude of offset decrease away from the uplifted parent continuity of Permian strata northward to subhorizontal dips. Some of the rotation of Gold Butte block. (4) They are the latest struc- from Azure Ridge. Furthermore, it does not the originally steeply dipping faults can be at- tures to affect the area, consistent with their explain the observed low-angle contact of Ter- tributed to folding accompanying isostatic up- formation being a response to extensional tiary Horse Spring Formation over Proterozoic lift of the Gold Butte crystalline complex. denudation. basement at the south end of Horse Spring The isostatic uplift of the Gold Butte crys- Ridge (Fig. 5), unless the Gold Butte fault has RELEVANCE TO INTERPRETING talline block was probably accommodated in deviated southward at that point, and is hidden OTHER EXTENDED REGIONS part by ¯exural folding of the upper crust. within the crystalline block. However, if the This folding is evident in the fanning of ®rst- trace of the Gold Butte fault does deviate The observations that (1) slip on normal generation fault dips east of the crystalline southward into the crystalline block south of faults continued to relatively low dips, and (2) complex. The near-surface dip of these faults Horse Spring Ridge, and northward around rotation of faults and fault blocks did not oc- decreases drastically from east to west, al- the Permian subcrop at the north end of Azure cur on crosscutting sets run counter to both though they apparently initiated with parallel Ridge, then it could still be interpreted as a the multiple-domino models (e.g., Morton and dips. Near-surface fault dips for the ®rst-gen- left-lateral fault (with a highly irregular sur- Black, 1975; Miller et al., 1983) and ¯exural eration structures vary from a probable 60Њ± face trace). If this interpretation is correct, rotation models (e.g., Buck, 1988), where all 90Њ for the fault(s) of the Grand Wash fault then Tramp Ridge may be the offset equiva- slip is accommodated on steeply dipping zone and Wheeler Ridge to a 12Њ east dip on lent to Azure Ridge, as suggested by Longwell faults. This raises the question of whether ei- the Indian Hills fault, only 9 km to the west et al. (1965), Bohannon (1984), and Beard ther of these concepts is broadly applicable to (Figs. 5 and 6). While some of this fanning (1996). In this case, the Garden Wash fault the formation of normal fault systems with can be explained by active slip of a stack of would be equivalent to the Million Hills Wash strongly rotated faults and fault blocks. Both listric domino blocks, at least some of it must fault, rather than being continuous with the are based on the assumption that slip on low- be explained by another mechanism, because South Virgin±White Hills detachment. This angle normal faults can not occur, rather than top-to-the-west normal faults, such as the In- interpretation places more emphasis on strike- on kinematics of well documented normal dian Hills fault, were not slipping with east- slip faulting, and reduces the total extension fault systems. Here we evaluate published ward dips. Much of the fanning of dips may calculated across the seven major fault blocks mapping from the highly extended southern be due to ¯exure caused by the uplift of the of the South Virgin Mountains. As measured Cherry Creek Range and Egan Range, Nevada unroofed crystalline block. Unfortunately, the from the cross sections of Figure 6, the total (Gans and Miller, 1983), and from two areas amount of rotation due to ¯exure is dif®cult extension across these blocks would change previously regarded as examples of the mul- to constrain, due to the dif®culty of calculat- from ϳ15 km if the Gold Butte fault is inter- tiple-domino model, the Yerington District, ing the amount of rotation of faults that bound preted as a lateral ramp to ϳ7kmifitisin- Nevada (Proffett, 1977; Proffett and Dilles, internally deforming listric domino blocks. If terpreted as a later left-lateral strike-slip fault. 1984), and the Lemitar Mountains, New Mex- these calculations could be made, it would While this alternative signi®cantly affects ico (Chamberlin, 1982, 1983), in order to test then be possible to accurately separate the ef- the total extension estimated along a transect whether the geology of these regions is con- across the Gold Butte fault, it has no effect on fects of ¯exural rotation from rotation due to sistent with the fault-system evolution de- the total extension estimated along a transect slip on a stack of listric domino blocks. duced for the normal fault system of the South across the Gold Butte crystalline complex. Virgin Mountains. Such a transect would require a minimum of Alternative Interpretations of the Gold The southern Cherry Creek Range and ϳ4 km extension across the fault blocks east Butte Fault Egan Range comprise a series of tilted normal of the crystalline complex, in addition to the fault blocks, each of which is internally ex- ϳ17 km of extension required to unroof the As described above, the preferred interpre- tended by fanning-upward splays and bounded exposed width of the crystalline block, giving tation of the Gold Butte fault is that it coin- by an originally steeply east dipping listric a total extension across the South Virgin cided with a right-lateral tear in the hanging normal fault that has accommodated Ͼ1km Mountains of ϳ21 km. wall of the South Virgin±White Hills detach- of offset (Gans and Miller, 1983). A sequential ment. This tear may coincide with a ramp in Kinematic Summary reconstruction of the region shows that the the footwall and may have been reactivated as deeper portions of the major block-bounding an isostatic uplift fault after the Gold Butte Early extension near the western margin of faults remained active to dips as low as 0Њ crystalline block was unroofed. This scenario the Colorado Plateau was directed roughly (Fig. 12 of Gans and Miller, 1983). Further-

Geological Society of America Bulletin, September 2000 1393 BRADY et al. fett and Dilles dipping faults. This Figure 15. An approximately(1984), west with to steeper eastreinterpretation cross faults honors section all generally through of crosscutting the the central shallower observations Singatse recorded faults; by Range, (B) Proffett just and reinterpreted, north Dilles of assuming (1984). the that Yerington steeper Mine: (A) faults as tend interpreted to by Prof merge into shallowly

1394 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

Figure 16. Approximately west to east cross section through the Lemitar Mountains (A) as interpreted by Chamberlin (1982) and (B) with late-stage (Pliocene) faults restored. Note that the steeply dipping faults merge into the shallowly dipping faults, rather than cutting them.

more, the reconstruction suggests that the the Yerington District. It is possible to honor tation are observed to cut the shallowest block-bounding faults were rotated through all of the geologic data and have the later, faults. horizontal, to gentle westward dips, due to up- steeper normal faults merge into the earlier, Overall, analysis of data sets from the lift of the adjacent highly attenuated Schell more gently dipping structures. Only the Mon- southern Cherry Creek and Egan Ranges, as Creek Range and Snake Range. This structural tana-Yerington fault demonstrably cuts the well as the Yerington District and the Lemitar evolution is virtually identical to that deduced older, gently dipping structures, and this cross- Mountains, suggests that the normal fault sys- for the normal fault system of the South Vir- cutting fault can not explain much rotation of tems responsible for Oligocene extension in gin Mountains. the older structures, as it appears to be nearly these three widely separated regions evolved The primary source of data for the Yering- planar and is shown as accommodating Ͻ400 in the same manner as the Miocene normal ton District was the map and cross sections of m of normal separation. fault system of the South Virgin Mountains. Proffett and Dilles (1984). A roughly east- In examining the Lemitar Mountains, the This suggests that the model of normal fault west±trending cross section was constructed map and cross sections of Chamberlin (1982) evolution (shown schematically in Fig. 1C) through the Singatse Range, just north of the were used. Cross-section D±DЈЈ from Cham- discussed throughout this paper may be appli- Yerington Mine. This cross section coincided berlin (1982) was partially restored by taking cable to many extended continental regions. with the eastern part of cross-section D±DЈ out the displacement on the latest stage of from Proffett and Dilles (1984). This cross steeply dipping normal faults (Fig. 16). These ACKNOWLEDGMENTS section, as interpreted by Proffett and Dilles, faults are probably related to the formation of and as reinterpreted by us, is shown in Figure the modern basins and ranges of the Socorro This research was supported by National 15. The reinterpretation honors the geologic area, and formed ca. 7±4 Ma, whereas most Science Foundation grant EAR-96-28262. We mapping of Proffett and Dilles, as well as the extension in the region probably occurred be- thank G. Spinelli and M. Herrin for ®eld as- contacts intersected by boreholes and other tween 31 and 27 Ma (Chamberlin, 1983). sistance. Discussions and ®eld excursions with contacts drawn as solid lines on their cross With the effects of late-stage faulting re- L.S. Beard, K. Howard, J. Faulds, and S. section. moved, the cross section shows a pattern of Rowland have contributed to the ideas pre- The reinterpretation of the Yerington Dis- normal faulting that is entirely consistent with sented here, although they are in no way liable trict cross section shows that the model of nor- that seen in the South Virgin Mountains and for any errors in fact or interpretation. Ster- mal fault evolution deduced for the South Vir- Yerington, in that no large-displacement faults eonet by R.W. Allmendinger was used for the gin Mountains is consistent with the faults of that could have accommodated signi®cant ro- analysis and plotting of data. Constructive and

Geological Society of America Bulletin, September 2000 1395 BRADY et al.

thoughtful reviews by James Faulds and L. summary, in Chapin, C.E., and Callender, J.F., eds., Lower Permian), southern Clark County, Nevada: New Mexico Geological Society Guidebook, 34th Journal of Sedimentary Petrology, v. 43, p. 655±671. Sue Beard improved the manuscript. Field Conference, Socorro Region II: Socorro, New Marzolf, J.E., 1990, Reconstruction of extensionally dis- Mexico Geological Society, p. 111±118. membered early Mesozoic sedimentary basins; south- Cole, E.D., 1989, Petrogenesis of late Cenozoic alkalic ba- western Colorado Plateau to the eastern Mojave De- REFERENCES CITED salt near the eastern boundary of the Basin-and- sert, in Wernicke, B.P., ed., Basin and Range exten- Range: Upper Grand Wash trough, Arizona and Gold sional tectonics near the latitude of Las Vegas, Ne- Butte, Nevada [Master's thesis]: Las Vegas, University Agnon, A., and Reches, Z., 1995, Frictional rheology: vada: Geological Society of America Memoir 176, p. of Nevada, 68 p. Hardening by rotation of active normal faults: Tecton- 477±500. Duebendorfer, E.M., and Sharp, W.D., 1998, Variation in ophysics, v. 247, p. 239±254. Matthews, J., 1976, Paleozoic stratigraphy and structural displacement along strike of the South Virgin±White Anderson, E.M., 1942, The dynamics of faulting and dyke geology of the Wheeler Ridge area, northwestern Mo- Hills detachment fault: Perspective from the northern formation with application to Britain: Edinburgh, Ol- have County, Arizona [Master's thesis]: Flagstaff, White Hills, northwestern Arizona: Geological Soci- iver and Boyd, 191 p. Northern Arizona University, 145 p. ety of America Bulletin, v. 110, p. 1574±1589. Anderson, R.E., 1971, Thin skin distension in Tertiary McKee, E.D., 1933, The Coconino sandstoneÐIts history Faulds, J.E., Schreiber, B.C., Reynolds, S.J., Gonzalez, rocks of southeastern Nevada: Geological Society of and origin, in Contributions to paleontology: Carnegie L.A., and Okaya, D., 1997, Origin and paleogeogra- America Bulletin, v. 82, p. 42±58. Institution of Washington Publication 440, p. 78±115. phy of an immense, nonmarine Miocene salt deposit Anderson, R.E., 1973, Large-magnitude late Tertiary strike- McKee, E.D., 1938, The environment and history of the in the Basin and Range (western USA): Journal of slip faulting north of Lake Mead, Nevada: U.S. Geo- Toroweap and Kaibab formations of northern Arizona Geology, v. 105, p. 19±36. logical Survey Professional Paper 794, 18 p. and southern Utah: Carnegie Institution of Washington Feuerbach, D.L., Smith, E.I., Walker, J.D., and Tangeman, Beard, L.S., 1996, Paleogeography of the Horse Spring For- Publication 492, 268 p. J.A., 1993, The role of the mantle during crustal ex- mation in relation to the Lake Mead fault system, Vir- McKee, E.D., 1945, Stratigraphy and ecology of the Grand tension: Constraints from geochemistry of volcanic gin Mountains, in Beratan, K.K., ed., Nevada and Ar- Canyon Cambrian, Part I of Cambrian history of the rocks in the Lake Mead area, Nevada and Arizona: izona: Reconstructing the history of Basin and Range Grand Canyon region: Carnegie Institution of Wash- Geological Society of America Bulletin, v. 105, p. extension using sedimentology and stratigraphy, Geo- ington Publication 563, p. 5±168. 1561±1575. logical Society of America Special Paper 303, p. 27± McKee, E.D., 1969, Strati®ed rocks of the Grand Canyon, Fitzgerald, P.G., Fryxell, J.E., and Wernicke, B.P., 1991, 60. in U.S. Geological Survey, The Colorado River region Miocene crustal extension and uplift in southeastern Bennett, V.C., and De Paolo, D.J., 1987, Proterozoic crustal and John Wesley Powell: U.S. Geological Survey Pro- Nevada: Constraints from ®ssion track analysis: Ge- history of the western United States as determined by fessional Paper 669-B, p. 222±249. ology, v. 19, p. 1013±1016. neodymium isotopic mapping: Geological Society of McKee, E.D., 1975, The Supai GroupÐSubdivision and Fryxell, J.E., Salton, G.S., Selverstone, J.E., and Wernicke, America Bulletin, v. 99, p. 674±685. nomenclature: U.S. Geological Survey Bulletin 1395- B., 1992, Gold Butte crustal section, South Virgin Blair, W.N., 1978, Gulf of California in Lake Mead area of J, 11 p. Mountains, Nevada: Tectonics, v. 11, p. 1099±1120. Arizona and Nevada during late Miocene time: Amer- McKee, E.D., and Resser, C.E., 1945, Cambrian history of Gans, P.B., and Miller, E.L., 1983, Style of mid-Tertiary ican Association of Petroleum Geologists Bulletin, v. the Grand Canyon region: Carnegie Institution of extension in east-central Nevada, in Gurgel, K.D., ed., 62, p. 1159±1170. Washington Publication 563, 232 p. Geologic excursions in the overthrust belt and meta- Blair, W.N., and Armstrong, A.K., 1979, Hualapai Lime- McNair, A.H., 1951, Paleozoic stratigraphy of part of north- morphic core complexes of the intermountain region, stone Member of the Muddy Creek FormationÐThe western Arizona: American Association of Petroleum GuidebookÐPart I: Utah Geological and Mineral Sur- youngest deposit predating the Grand Canyon, south- Geologists Bulletin, v. 35, p. 503±541. vey Special Studies 59, p. 107±139. eastern Nevada and northwestern Arizona: U.S. Geo- Miller, E.L., Gans, P.B., and Garing, J., 1983, The Snake Gregory, H.E., 1915, The igneous origin of the ``glacial logical Survey Professional Paper 1111, 14 p. Range decollement; an exhumed mid-Tertiary ductile- deposits'' on the Navajo Reservation, Arizona and Bohannon, R.G., 1979, Strike-slip faults of the Lake Mead brittle transition: Tectonics, v. 2, p. 239±263. Utah: American Journal of Science, ser. 4, v. 40, p. region of southern Nevada, in Armentrout, J.M., et al., Morgan, J.R., 1968, Structure and stratigraphy of the north- 97±115. eds., Cenozoic paleogeography of the Western United ern part of the South Virgin Mountains, Clark County, Gregory, H.E., and Williams, N.E., 1947, Zion National States: Paci®c Coast Paleogeography Symposium 3: Nevada [M.S. thesis]: Albuquerque, New Mexico, Monument, Utah: Geological Society of America Bul- Los Angeles, Paci®c Section, Society of Economic Pa- University of New Mexico, 103 p. letin, v. 58, p. 211±244. leontologists and Mineralogists, p. 129±139. Morton, W.H., and Black, R., 1975, Crustal attenuation in Hamblin, W.K., 1965, Origin of ``reverse drag'' on the Bohannon, R.G., 1983, Mesozoic and Cenozoic tectonic de- Afar, in Pilger, A., and Roesler, A., eds.: Stuttgart, velopment of the Muddy, North Muddy, and northern downthrown side of normal faults: Geological Society of America Bulletin, v. 76, p. 1145±1164. FRG, Verlagsbuchhandl, Afar depression of Ethiopia, Black Mountains, Clark County, Nevada, in Miller, v. 1, p. 55±65. D.M., et al., eds., Tectonic and stratigraphic studies in John, B.E., and Foster, D.A., 1993, Structural and thermal constraints on the initiation of angle of detachment Nur, A., Ron, H., and Scotti, O., 1986, fault mechanics and the eastern Great Basin: Geological Society of Amer- the kinematics of block rotations: Geology, v. 14, p. ica Memoir 157, p. 125±148. faulting in the southern Basin and Range: The Che- mehuevi Mountains case study: Geological Society of 746±749. Bohannon, R.G., 1984, Nonmarine sedimentary rocks of Poborski, S.J., 1954, Virgin Formation (Triassic) of the St. Tertiary age in the Lake Mead region: U.S. Geological America Bulletin, v. 105, p. 1091±1108. Langenheim, R.L., 1963, Mississippian stratigraphy in George, Utah, area: Geological Society of America Survey Professional Paper 1259, 72 p. Bulletin, v. 65, p. 971±1006. Brady, R.J., 1998, The geology of the Gold Butte break- southwestern Utah and adjacent parts of Nevada and Arizona, in Guidebook to the geology of southwestern Proffett, J.M., Jr., 1977, Cenozoic geology of the Yerington away zone and the mechanical evolution of normal district, Nevada, and implications for the nature and fault systems [Ph.D. thesis]: Pasadena, California In- Utah, in Heylmun, E.B.: Salt Lake City, Utah, Inter- mountain Association of Petroleum Geologists, 12th origin of Basin and Range faulting: Geological Soci- stitute of Technology, 200 p. ety of America Bulletin, v. 88, p. 247±266. Buck, W.R., 1988, Flexural rotation of normal faults: Tec- Annual Field Conference, p. 30±41. Proffett, J.M., Jr., and Dilles, J.H., 1984, Geologic map of tonics, v. 7, p. 959±973. Longwell, C.R., 1921, Geology of the Muddy Mountains, the Yerington District, Nevada: Nevada Bureau of Burch®el, B.C., and Davis, G.A., 1975, Nature and controls Nevada: American Journal of Science, ser. 5, v. 1, p. Mines and Geology map 77, scale 1:24 000. of Cordilleran orogenesis, western United States; ex- 39±62. Saltus, R.W., and Jachens, R.C., 1995, Gravity and basin- tensions of an earlier synthesis: American Journal of Longwell, C.R., 1936, Geology of the Boulder Reservoir depth maps of the Basin and Range province, western Science, v. 275-A, p. 363±396. ¯oor: Geological Society of America Bulletin, v. 47, Burch®el, B.C., Fleck, R.J., Secor, D.T., Vincelette, R.R., p. 1393±1476. United States: U.S. Geological Survey Geophysical and Davis, G.A., 1974, Geology of the Spring Moun- Longwell, C.R., 1945, Low angle normal faults in the Basin Investigations Map GP-1012, scale 1:2 500 000. tains, Nevada: Geological Society of America Bulle- and Range province: Eos (Transactions, American Schenck, E.T., and Wheeler, H.E., 1942, Cambrian se- tin, v. 85, p. 1013±1022. Geophysical Union), v. 26, p. 107±118. quence in western Grand Canyon, Arizona: Journal of Campagna, D.J., and Aydin, A., 1994, Basin genesis as- Longwell, C.R., 1949, Structure of the northern Muddy Geology, v. 50, p. 882±899. sociated with strike-slip faulting in the Basin and Mountain area, Nevada: Geological Society of Amer- Sibson, R.H., 1985, A note on fault reactivation: Journal of Range, southeastern Nevada: Tectonics, v. 13, p. 327± ica Bulletin, v. 60, p. 923±966. Structural Geology, v. 7, p. 751±754. 341. Longwell, C.R., Pampeyan, E.H., Bowyer, B., and Roberts, Silver, L.T., Bickford, M.E., Van Schmus, W.R., Anderson, Cascadden, T.E., 1991, Style of volcanism and extensional R.J., 1965, Geology and mineral deposits of Clark J.L., Anderson, T.H., and Medaris, L.G., Jr., 1977, The tectonics in the eastern Basin and Range province; County, Nevada: Nevada Bureau of Mines Bulletin 1.4±1.5 b.y. transcontinental anorogenic plutonic per- northern Mohave County, Arizona [Master's thesis]: 62, 218 p. foration of North America: Geological Society of Las Vegas, University of Nevada, 209 p. Lucchitta, I., 1967, Cenozoic geology of the upper Lake America Abstracts with Programs, v. 9, p. 1176±1177. Chamberlin, R.M., 1982, Geologic map, cross sections, and Mead area adjacent to the Grand Wash Cliffs, Arizona Stewart, J.H., 1957, Proposed nomenclature of part of Up- map units of the Lemitar Mountains, Socorro County, [Ph.D. thesis]: University Park, Pennsylvania State per Triassic strata in southeastern Utah: American As- New Mexico: New Mexico Bureau of Mines and Min- University, 274 p. sociation of Petroleum Geologists Bulletin, v. 41, p. eral Resources Open-File Report 169, scale 1:12 000. Lumsden, D.N., Ledbetter, M.T., and Smith, G.T., 1973, 441±465. Chamberlin, R.M., 1983, Cenozoic domino-style crustal ex- Lithostratigraphic analysis of the Bird Spring±Call- Stewart, J.H., 1980, Geology of Nevada, a discussion to tension in the Lemitar Mountains, New Mexico; a ville Group and Pakoon Formation (Pennsylvanian± accompany the geologic map of Nevada: Nevada Bu-

1396 Geological Society of America Bulletin, September 2000 KINEMATIC EVOLUTION OF A LARGE-OFFSET CONTINENTAL NORMAL FAULT SYSTEM

reau of Mines and Geology Special Publication 4, in the evolution of normal fault systems: Geology, v. Arizona: Carnegie Institution of Washington Publica- 136 p. 16, p. 848±851. tion 405, 219 p. Thomas, W.M., Clark, S.H., Young, E.D., Orrell, S.E., and Wernicke, B., Walker, J.D., and Beaufait, M.S., 1985, Struc- Wolf, R.A., Farley, K.A., and Silver, L.T., 1996, Helium Anderson, J.L., 1988, Proterozoic high-grade meta- tural discordance between Neogene detachments and diffusion and low temperature thermochronometry of morphism in the Colorado River region, Nevada, Ar- frontal Sevier thrusts, central Mormon Mountains, apatite: Geochimica et Cosmochimica Acta, v. 60, p. izona and California, in Ernst, W.G., ed., Metamor- southern Nevada: Tectonics, v. 4, p. 213±246. 4231±4240. Young, R.A., 1966, Cenozoic geology along the edge of phism and crustal evolution of the western United Wernicke, B., Axen, G.J., and Snow, J.K., 1988, Basin and States, Rubey Volume VII: Englewood Cliffs, New the Colorado Plateau in northwestern Arizona [Ph.D. Range extensional tectonics at the latitude of Las Ve- Jersey, Prentice Hall, p. 527±537. dissert.]: St. Louis, Missouri, Washington University, gas, Nevada: Geological Society of America Bulletin, Volborth, A., 1962, Rapakivi-type granites in the Precam- 155 p. brian complex of Gold Butte, Clark County, Nevada: v. 100, p. 1738±1757. Young, R.A., 1979, Laramide deformation, erosion and plu- Geological Society of America Bulletin, v. 73, p. 813± Wheeler, H.E., 1943, Lower and Middle Cambrian stratig- tonism along the southwestern margin of the Colorado 832. raphy in the Great Basin area: Geological Society of Plateau: Tectonophysics, v. 61, p. 25±47. Wasserburg, G.J., and Lanphere, M.A., 1965, Age deter- American Bulletin, v. 54, p. 1781±1822. MANUSCRIPT RECEIVED BY THE SOCIETY JUNE 15, 1998 Wheeler, H.E., and Beasley, E.M., 1948, Critique of the minations in the Precambrian of Arizona and Nevada: REVISED MANUSCRIPT RECEIVED AUGUST 20, 1999 time-stratigraphic concept: Geological Society of Geological Society of America Bulletin, v. 76, p. 735± MANUSCRIPT ACCCEPTED SEPTEMBER 17, 1999 758. America Bulletin, v. 59, p. 75±86. Wernicke, B., and Axen, G.J., 1988, On the role of isostasy White, D., 1929, Flora of the Hermit Shale, Grand Canyon, Printed in the USA

Geological Society of America Bulletin, September 2000 1397