TECTONICS, VOL. 28, TC6011, doi: /2008tc002352, 2009

Transcription

1 TECTONICS, VOL. 28,, doi: /2008tc002352, 2009 U-Pb ages of detrital zircons in Pacheco Pass metagraywackes: Sierran-Klamath source of mid-cretaceous and Late Cretaceous Franciscan deposition and underplating W. G. Ernst, 1 Uwe Martens, 1 and Victor Valencia 2 Received 26 June 2008; revised 17 June 2009; accepted 24 August 2009; published 24 December [1] Using laser ablation ICP-MS techniques, U-Pb ages are reported for 338 detrital zircons separated from four Franciscan metagraywacke samples from imbricate thrust sheets cropping out in the Pacheco Pass area, east central California Coast Ranges. Structurally higher slabs are clastic quartz albitebearing rocks typified by neoblastic lawsonite + jadeitic pyroxene; the lowest exposed slab lacks jadeitic pyroxene but contains traces of newly grown pumpellyite and lawsonite in addition to abundant quartz + albite. Studied specimens contain moderate amounts of phengite + titanite chlorite, stilpnomelane, carbonate, iron oxides, rock fragments, and carbonaceous matter. These Diablo Range clastic sediments were deposited in mid-cretaceous and Late Cretaceous time. The highest allochthon, unit I, was deposited after 102 Ma, unit IV was deposited after Ma, and unit V was deposited after 86 Ma. Metagraywacke depositional-accretionary ages within slab IV young upward, but ages of the slabs decrease progressively downward. In the foursample aggregate, most zircons have igneous ages falling in the Ma range, with a smaller population at Ma. The rocks contain rare zircons of Middle Proterozoic Late Archean ages. A minor source in northern California and NW Nevada seems likely for the Pacheco Pass metagraywackes, but the Sierran-Klamath calcalkaline arc provided most of the clastic debris. Judging by the volcanic nature of lithic clasts in these Eastern Belt metagraywackes, and because massive Sierran plutonism occurred at Ma, the abundant Ma zircons probably were derived chiefly from the eroding comagmatic volcanic arc rather than from the less voluminous Jurassic arc plutons. 1 Department of Geological and Environmental Sciences, Stanford University, Stanford, California, USA. 2 Department of Geosciences, University of Arizona, Tucson, Arizona, USA. Copyright 2009 by the American Geophysical Union /09/2008TC Citation: Ernst, W. G., U. Martens, and V. Valencia (2009), U-Pb ages of detrital zircons in Pacheco Pass metagraywackes: Sierran-Klamath source of mid-cretaceous and Late Cretaceous Franciscan deposition and underplating, Tectonics, 28,, doi: /2008tc Introduction to the Problem [2] Except for an important Ma subduction interval, early Mesozoic differential slip along the Californian margin apparently involved mainly transcurrent and transpression plate motions, whereas the latest Mesozoic plate tectonic history was dominated by transpression, then subduction, prior to the return to transform plate motions in late Oligocene time [Atwater, 1970; Dickinson, 1970; Ernst, 1984, Wakabayashi, 1999; Miller, 2004; Ernst et al., 2008]. The Cretaceous rock record of lithospheric underflow consists, from west to east, of (1) clastic trench deposition and accretionary offloadedunderplated sediments from the downgoing oceanic plate composed of the metamorphosed Franciscan Complex, (2) inboard clastic sediments of the Great Valley Group (GVG) fore arc, and (3) massive Sierran-Klamath arc magmatism landward of the fore-arc basin [Hamilton, 1969; Ernst, 1970; Dickinson, 1970, 2008; Blake et al., 1988]. Terrigenous units of the Franciscan and GVG consist largely of first-cycle quartzofeldspathic materials. Older ophiolite and chert-argillite terrane collages that crop out beneath and east of the GVG in the Sierra Nevada Foothills and in the Klamath Mountains record pre-cretaceous episodes of chiefly transpression-transtension [Ernst et al., 2008]. These terranes include late Paleozoic early Mesozoic plutonic, volcanic, and metamorphic rocks dominantly of oceanic affinities [Wright, 1982; Wright and Fahan, 1988; Irwin, 2003; Dickinson, 2008]. Simplified geologic relationships are presented in Figure 1. [3] The Franciscan Complex provides a well-documented record of continental margin accretion and underplating [Hamilton, 1969; Maxwell, 1974]. In northern California, it consists of three subparallel terrane amalgams, Eastern, Central, and Coastal belts, deposited and assembled chiefly during the Early, Late, and latest Cretaceous-Miocene, respectively [Irwin, 1960; Bailey et al., 1964, 1970; Blake et al., 1982; McLaughlin et al., 1982]. In general, tectonic packages decrease in age structurally downward in both times of sedimentation and accretion, consistent with models of successive off- 1of20

2 Figure 1. Geologic map of northern California emphasizing areal distributions of the Eastern, Central, and Coastal belts of the Franciscan Complex, the Great Valley Group, and the Sierran-Klamath arc terrane, much simplified after Jennings [1977] and Silberling et al. [1987]. The Pacheco Pass quadrangle and other areas mentioned in the text are indicated. loadings of units in a subduction zone metamorphic environment [Ernst, 1970, 1975; Blake et al., 1988; Wakabayashi, 1992]. [4] Unresolved questions regarding the Franciscan Complex concern its times of deposition and nearly immediate sequestering beneath the Californian margin, and the nature of the source rocks that supplied detritus to the trench. Here we report U-Pb age data for dominantly igneous zircons extracted from Franciscan lithotectonic units that occupy three different structural horizons in the Pacheco Pass quadrangle of the east central Diablo Range. These new geochronologic data, compatible with those obtained by other research groups investigating the Franciscan Eastern and Central belts allow us to determine the time of trench filling and accretion underplating and to better constrain the nature of the source terranes during Cretaceous time. 2. General Geology of the Franciscan and Related Rocks [5] Although the Franciscan Complex and Great Valley Group have similar ages of clastic sedimentation, deeply buried GVG strata record relatively indistinct, incipient metamorphism [Dickinson et al., 1969; Dickinson and Rich, 1972] and display well-developed bedding, whereas Franciscan rocks exhibit feeble to high-pressure/low-temperature (HP/LT) blueschist-facies recrystallization [Ernst, 1975, 1984]. Much of the Franciscan has been disrupted as block-in-matrix mélanges [Brown, 1964; Bailey et al., 2of20

3 1964; Hsü, 1968]. In the western Sierra Nevada Foothills, Upper Cretaceous continental GVG clastic units apparently rest with angular unconformity on superjacent rocks of an Upper Jurassic and older bedrock sequence [Hackel, 1966; Williams, 1997]. To the north, Lower Cretaceous Great Valley strata overlie ophiolitic rocks of the Eastern Klamath terrane and associated Early Cretaceous granitic stocks [Constenius et al., 2000]. The oldest, locally disrupted sedimentary units in the western GVG were deposited on the Middle Jurassic Coast Range Ophiolite [Bailey et al., 1970; McLaughlin and Ohlin, 1984; Shervais et al., 2005; Hopson and Pessagno, 2005; Hopson et al., 2008]. In contrast, the Franciscan Complex accumulated in the vicinity of an offshore trench as a series of imbricate slices that episodically decoupled from the Farallon oceanic plate as it descended beneath the subducting North American lithosphere. Allochthons include both intact, fault-bounded coherent terranes, and tectonically disrupted mélange belts [Maxwell, 1974; Blake et al., 1984; Wakabayashi, 1992]. The Coast Range Ophiolite and overlying GVG comprise the leading, western edge of the nonsubducted North American crust during this outboard accretion and underplating of the Franciscan Complex. [6] The GVG consists mainly of continent- and arcderived clastic fore-arc basin sediments marking the leading edge of the North American plate. On the other hand, the Franciscan includes far-traveled basalt and deep-water chert + limestone of the Farallon oceanic plate in addition to younger terrigeneous sediments; clastic rocks make up 90 95% of the complex and were laid down on the subjacent oceanic lithosphere as it approached the trench [Ernst, 1965; Blake et al., 1988]. This Franciscan section was incipiently to deeply subducted before being decoupled and transferred into the accretionary margin shortly after deposition. Thus, the age of accumulation of Franciscan clastic strata and the youngest oceanic chert limestone layers probably more closely approximates the time of underflow, recrystallization, and offloading, than the formation of the oceanic crust and immediately overlying chemically precipitated strata, the ages of which significantly predate their arrival at the trench and fragmentary incorporation in the subduction complex [Bailey et al., 1970; Wakabayashi, 1999]. [7] A vexing impediment to determining the accretionary architecture and manner of construction of the Franciscan is posed by the scarcity of both datable macrofossils (some of which we suspect may be reworked) and reliable radiometric ages. The clastic metasedimentary rocks are weakly recrystallized and in general have not reequilibrated chemically or isotopically, so fail to yield datable neoblastic minerals [Blake et al., 1988; Wakabayashi and Dumitru, 2007]. However, new detrital zircon radiometric studies in the South Fork Mountain Yolla Bolly area, the San Francisco Bay area, and the Diablo Range [Joesten et al., 2004; Tripathy et al., 2005; Dumitru et al., 2007; Snow et al., 2009] support the concept of a downward younging stack of successively accreted slices. Although controversial, based on the youngest U-Pb ages of these clastic grains, deposition of the oldest Franciscan metaclastic rocks in these areas appear to be Early Cretaceous rather than Late Jurassic as previously held for these predominantly Eastern Belt units [Bailey et al., 1964]. 3. Times of Deposition and Underplating of Eastern Belt Franciscan Metaclastic Rocks [8] The stratigraphically coherent, east dipping thrust sheets and chaotic mélanges that make up the Franciscan Complex comprise a giant stack of nappes [Maxwell, 1974; Blake et al., 1984; Wakabayashi, 1992, 1999]. An unambiguous indication of the times of nappe accretion would be recorded by the age of penecontemporaneous HP/LT metamorphism. The famous coarse-grained blueschist, amphibolite and eclogite lenses indeed are well dated, but they represent exotic tectonic blocks of uncertain but oceanic origin, and most have Middle Jurassic recrystallization ages, considerably older than those of the enclosing metasedimentary sections [Anczkiewicz et al., 2004; Saha et al., 2005]. Metamorphic ages for the far more voluminous, weakly recrystallized sedimentary units are poorly constrained, reflecting a lack of readily separatable neoblastic minerals [Wakabayashi, 1999; Wakabayashi and Dumitru, 2007]. Judging from the overall map disposition of the complex, metamorphic grade monotonically decreases in structurally lower, depositionally younger units, from blueschist on the east through transitional prehnite-pumpellyite to laumontite-bearing zeolite facies on the west [Ernst, 1975; Blake et al., 1982, 1988]. At Ortigalita Peak in the SE Diablo Range, Mattinson and Echeverria [1980] obtained a 95 Ma U-Pb igneous zircon TIMS age on a gabbroic body intrusive into structurally high metagraywackes, and a 92 Ma U-Pb isochron age on the HP/LT recrystallization of this hypabyssal pluton. However, few other intrusive rocks have been dated in the Franciscan, and overlap strata produced during the postsubduction transform regime are somewhat to considerably younger in age. Thus, the least ambiguous constraints on times of accretionary underplating of most Franciscan clastic rocks are their slightly older depositional ages. [9] In the northern Coast Ranges at South Fork Mountain, Eastern Belt nappes in descending tectonic order are the South Fork Mountain Schist, the Valentine Springs Formation these two entities comprise the Pickett Peak terrane and the Yolla Bolly terrane [Worrall, 1981]. The most intensely recrystallized, structurally highest South Fork Mountain Schist gave a 40 Ar/ 39 Ar step heating metamorphic or cooling age of Ma on neoblastic phengite [Wakabayashi and Dumitru, 2007]. For a structurally high sheet within the Yolla Bolly imbricate terrane, Mertz et al. [2001] obtained a 40 Ar/ 39 Ar step heating age of Ma on hornblende from a premetamorphic intrusion into blueschist-facies metagraywackes. Dumitru et al. [2007] used SIMS methods on detrital zircons separated from seven rocks from the Pickett Peak and Yolla Bolly terranes; the youngest U-Pb depositional ages obtained by these authors decrease downward from a maximum of 137 Ma in the South Fork Mountain Schist, to 120 Ma in the Valentine Springs Formation, and 111 Ma in the Yolla Bolly terrane, in agreement with macrofossil 3of20

4 assemblages noted by Blake et al. [1988] but slightly at odds with the premetamorphic intrusive dated by Mertz et al. [2001]. Dumitru et al. [2007] showed that in this area, a large portion of the detrital zircons had ages in the Ma interval. In summary, detrital U-Pb zircon ages for blueschist-facies metagraywackes define a continuum from 110 to 180 Ma, but with a small number of ages populating the Ma interval, possibly reflecting reduced magmatic activity in the sediment source regions. [10] The Diablo Range in the east central California Coast Ranges is underlain by a lithologic section correlated with the Eastern and Central belts of Franciscan rocks to the north. Recently obtained U-Pb detrital zircon data require a middle to Late Cretaceous depositional age for Diablo Range metagraywackes. Joesten et al. [2004] studied clastic zircons from a blueschist-facies broken formation and a mélange in the NE part of the range; for these jadeitic pyroxene-bearing units, cathodoluminescence imaging of euhedral detrital zircon grains revealed strong oscillatory igneous zoning. Most zircons yielded concordant U-Pb age spans of , and Ma by SIMS methods, implying a Sierran-Klamath source [e.g., Irwin, 2003]. Rare Late Archean and Proterozoic zircon ages suggest a minor North American cratonal provenance. Youngest values constrain the maximum age of broken formation and mélange deposition at 85 and 99 2 Ma, respectively. Unruh et al. [2007] obtained similar SIMS U-Pb data for detrital zircon grains from a coherent Mount Diablo metagraywacke, indicating that Franciscan sedimentation took place later than 108 Ma. Using laser ablation ICPMS methods on zircons from 14 well-layered jadeitic metagraywacke samples from Pacheco Pass, Tripathy et al. [2005] documented a U-Pb age of 95 Ma for the earliest time of deposition. Snow et al. [2009] used SIMS techniques to determine U-Pb ages for detrital zircons from five Franciscan metagraywackes from terranes in the vicinity of San Francisco Bay, demonstrating that depositional ages of individual slices young downward in the sequence, Angel Island nappe, Alcatraz nappe, Hunters Point shear zone, Novato Quarry terrane, and San Bruno Mountain terrane; the apparent maximum depositional ages of these lithotectonic units are 102, 100, 97, 83, and 52 Ma, respectively. If the Eocene formation age of the San Bruno Mountain terrane is corroborated by further work, this tectonic unit may actually belong to the Coastal Belt of the Franciscan Complex. Snow et al. [2009] also showed that the degree of HP/LT metamorphism decreases in the structurally lower slabs, similar to depositional ages and metamorphic relationships of Franciscan clastic units farther north at South Fork Mountain. Most Diablo Range and some of the Bay area Franciscan rocks for which detrital zircon studies have been conducted thus far, including the section at Ortigalita Peak, apparently belong to the Eastern Belt [Blake et al., 1984; Wakabayashi, 1992]. 4. Great Valley Group Deposition [11] At about the end of Jurassic or beginning of Cretaceous time, the fore-arc basin sited between the Franciscan trench and the Klamath-Sierran arc began to receive firstcycle felsic debris in a continental slope, rise, and bathyal setting. The section is paraconformable, and deposition was nearly continuous. Except for thick, deeply buried, albitized-zeolitized strata exposed along the western edge of the basin, the GVG is only incipiently recrystallized [Dickinson et al., 1969]. Current features, detrital minerals, rock clasts, and modal proportions in the turbiditic sandstones indicate a chiefly Sierran-Klamath provenance [Dickinson and Rich, 1972; Dickinson et al., 1982; Ingersoll, 1978, 1983; Suchecki, 1984]. Lower stratigraphic units contain abundant volcanic clasts, whereas overlying strata grade upward to quartz + K-feldspar-rich rocks, reflecting progressive unroofing erosion of the inboard volcanic-plutonic arc [Dickinson and Rich, 1972; Ingersoll, 1978, 1983]. [12] Wright and Wyld [2007] studied pre-mesozoic detrital zircon populations of basal Great Valley Group sandstones and called attention to the fact that they do not correlate well with potential source areas in the Sierran Foothills, the eastern Klamaths, and NW Nevada, but instead are comparable to the Jurassic Erg deposits of the Colorado Plateau. Based on >260 Ma zircon ages and both structural evidence and analogy with modern fore-arc systems, they postulated that the older part of the Great Valley group was translated as a tectonic sliver 500 km in a dextral sense prior to deposition of the Upper Cretaceous section of the GVG. [13] Franciscan petrofacies closely match those of the landward Great Valley Group, supporting a common calcalkaline arc provenance for the predominantly first-cycle clastic material [Dickinson, 1985]. In the Franciscan metagraywackes, more intense recrystallization has obliterated some of the rock fragments, accounting for their smaller proportions relative to clasts in the GVG strata. Current features indicate short transport distances for detritus filling the fore-arc basin [Ingersoll, 1979]. By analogy with modern deposits, axial filling of the oceanic trench is envisioned but not proven for the Franciscan Complex. [14] Using SIMS methods, DeGraaff-Surpless et al. [2002] provided U-Pb age data for detrital zircons separated from 17 GVG sandstones from the Sacramento and San Joaquin valleys. Surpless et al. [2006] later analyzed zircons from seven additional specimens of putative uppermost Jurassic turbidites cropping out along the western edge of the Sacramento Valley. Of the eight samples from the basal GVG Stony Creek petrofacies, youngest ages consistently range from 133 to Ma. We caution that some workers [e.g., Wright and Wyld, 2007] have considered lead loss, analytical errors, and time scale uncertainties to be possible reasons for apparent Early Cretaceous ages of the youngest zircons reported by Surpless et al. [2006], whereas others (e.g., R. J. McLaughlin, personal communication, 2008) have argued that structural complexities may have skewed the sampling to horizons up section from the Stony Creek Formation, at least in the Wilbur Springs area [McLaughlin et al., 1990]. Overlying the Stony Creek unit, youngest zircons investigated by DeGraaff-Surpless et al. [2002] range from 72 to 97 2 Ma. Thus, post-stony Creek Great Valley strata would appear to be mainly Late 4of20

5 Table 1. Youngest Radiometric Ages of Rocks of the Late Mesozoic Fore-Arc Basin and Trench, Northern and Central California Rock Type U-Pb Ages of Deposition Number of Samples Source Great Valley Group detrital zircon 24 Main section DeGraaff-Surpless et al. [2002] Basal Stony Creek Surpless et al. [2006] Franciscan Eastern Belt detrital zircon 28 NE Diablo Range Joesten et al. [2004] SE Diablo Range 95 Tripathy et al. [2005] SE Diablo Range this study Mt. Diablo 108 Unruh et al. [2007] San Francisco Bay area 85 Snow et al. [2009] South Fork Mountain 137 Dumitru et al. [2007] Schist Yolla Bolly terrane 111 Dumitru et al. [2007] Franciscan Eastern Belt igneous/metamorphic 4 SE Diablo Range zircon U-Pb Mattinson and Echeverria [1980] South Fork Mountain phengite Ar/Ar 121 Wakabayashi and Dumitru [2007] Schist Yolla Bolly terrane hornblende Ar/Ar 119 Mertz et al. [2001] Cretaceous in depositional age. The GVG-equivalent Hornbrook Formation of NW California and southern Oregon is of a similar Late Cretaceous age range [Beverly et al., 2008]. Table 1 lists modern radiometric constraints on the beginning of deposition of Eastern Belt Franciscan and Great Valley Group sediments. 5. Pacheco Pass Franciscan Geology and Sample Descriptions [15] Detrital zircon-bearing samples analyzed in the present study were selected from exposures in the Pacheco Pass quadrangle of the east central Diablo Range [Ernst, 1993]. Figure 1 places Pacheco Pass in the regional geologic framework of northern California. The Diablo antiform probably represents the southern continuation of the Franciscan Eastern Belt but also includes weakly recrystallized Central Belt mélanges [Blake et al., 1984] in the core of the antiform. Similar to the isolated Mount Diablo faulted anticline along strike to the north [Schemmann et al., 2008], the Diablo Range Franciscan Complex occurs as a window structurally beneath the overlying Great Valley Group. The two terranes are juxtaposed along the Coast Range fault [Bailey et al., 1970; Ernst, 1970; Blake et al., 1984]. In the Pacheco Pass area, this structural break is termed the Ortigalita fault. [16] Figure 2 presents a simplified geologic map and cross section and studied metagraywacke localities from the Pacheco Pass quadrangle [Ernst, 1993]. The Franciscan section here consists of a well-defined pile of east rooting nappes. Each of five mapped sheets is 1 2 km thick; these units I V are juxtaposed along subparallel bedding plane thrust faults. At its faulted base, the discontinuous presence of pods and lenses of massive-to-pillowed metabasalt overlying metacherty flow tops mark each décollement. The latter are presumably decapitated layers 2 and 3 of the oceanic crust. The blueschists and siliceous metasedimentary lenses of each basal unit are succeeded upward by turbiditic metagraywackes and intercalated dark metashales. The entire section was metamorphosed under HP/LT conditions [Ernst, 1984, 1993]. Because bedding plane faults are difficult to recognize in such monotonous graywacke sections, it is possible that imbrications lacking distinctive basal glaucophane schists metacherts have gone unrecognized; accordingly, more numerous, thinner allochthons may comprise the Pacheco Pass area than have been mapped. [17] The imbricate stack of lithostratigraphic packages is gently folded about E W trending axes. This folding postdated HP/LT recrystallization, but could well have taken place during early stages of exhumation from the subduction zone inasmuch as ptygmatic quartz stringers and smallscale rootless folds suggest ductile rather than brittle deformation. Sheets I, II, and III appear to be intact stratigraphic packages. Sparse graded bedding suggests that these units are right-side-up. Unit IV also seems to be coherent, but the occurrence of abundant blueschist and metachert blocks in it suggest the possibility that it is a broken formation. Unit V is more chaotic, and could represent either a broken formation or a mélange. The tectonically highest slab, unit I (sample Q-58), is confined to the SE quadrant of the area. Similar to it, successively lower allochthons, units II, III, and IV (samples X-206 and A-1), consist mainly of quartzbearing metagraywackes containing relict clastic plagioclase transformed to albite, partially or nearly completely replaced by neoblastic jadeite-rich clinopyroxene, minor quartz, and lawsonite. Virtually all the studied metagraywackes of units I IV contain at least modest amounts of jadeitic pyroxene + quartz [Ernst, 1993], so regardless of the degree of imbrication, all of these sheets must have been carried down to depths of 25 km prior to exhumation. In contrast, the structurally lowest slab, unit V (sample Q-24), lacks jadeitic clinopyroxene but contains neoblastic lawsonite pumpellyite in addition to abundant detrital quartz, albite, and a few flakes of biotite; unit V apparently was exhumed from shallower depths that were units I IV as indicated by the persistence of clastic biotite and lack of sodic clinopyroxene. All studied specimens contain moderate amounts of phengite and titanite chlorite, stilpnome- 5of20

6 Figure 2. Geologic map of the Pacheco Pass quadrangle, SE Diablo Range, California Coast Ranges, and cross section A-B, simplified after Ernst [1993]. Analyzed zircon samples and maximum depositional ages in Ma of the allochthons are projected into the line of cross section in their proper stratigraphic positions. 6of20

7 Figure 3. Photomicrographs (plane light, field of view mm) of the four analyzed metagraywacke samples, illustrating their contrasting textures: (a, b, and c) more intensely recrystallized jadeitic clinopyroxene-bearing, albite-poor metagraywackes of the upper units I and IV; (d) a feebly reconstituted albite-rich metagraywacke of the lowest exposed unit V. See Table 2 for estimated modes and the text for brief descriptions of these four investigated rocks. lane, carbonate, iron oxides, rock fragments, and carbonaceous matter. [18] Textural relationships and estimated modes of the studied metagraywackes are presented in Figure 3 and in Table 2, respectively. Quartz, plagioclase, and most rock fragments are angular, reflecting the immaturity and short transport distances of the detritus. The metagraywackes are characterized by neoblastic jadeitic pyroxene needles and sprays (Q-58, X-206, and A-1) and show relatively indistinct grain boundaries due to relatively thorough recrystallization. However, carbonaceous streaks, shale chips, and a few volcanic rock fragments are preserved. In contrast, the slightly coarser grained, jadeite-absent sample (Q-24) displays clear detrital grain boundaries and a more distinct, fine-grained layer-silicate rich matrix. 6. Analytical Techniques [19] Detrital zircons from the four Franciscan sandstones were analyzed employing a laser ablation multicollector ICP-MS at the University of Arizona LaserChron Center to constrain depositional U-Pb ages and sources of the sediments. Zircons were separated from 2 to 3 kg rock samples using standard techniques. Splits of zircon grains were mounted along with standards on centers of 2.54 cm phenolic O rings and polished to expose grain interiors. Detrital zircons were selected at random for analysis, irrespective of their extent of rounding or fracturing; however, grains smaller than 35 mm were avoided. Each measurement cycle consisted of one 12 s integration on every centered peak with the laser off for background counts, twenty 1 s integrations during mineral blasting, and 20 s of purging. Zircons were ablated with a New Wave Research DUV193 ArF Excimer laser (wavelength of 193 nm). The laser beam was operated at 32 mj energy (at 23 kv), a pulse rate of 9 Hz, and a spot size of 35 or 50 mm, depending on size and complexity of the target zircons. Generated pits were 10 mm deep. Ablated material was carried in helium gas into an Isoprobe ICP-MS manufactured by GV Instruments. U, Th and Pb isotopes were analyzed simultaneously in static mode using Faraday collectors for 238 U, 232 U, 208 Pb, 207 Pb and, and an ion counter for 204 Pb. Contributions to the 204 mass by Hg were removed by subtracting background counts. Common Pb corrections were calculated from 204 Pb measurements assuming an initial isotopic composition based on the Pb evolution curve of Stacey and Kramers [1975]. [20] Five Sri Lanka zircon standards (564 4 Ma, 2s error [Gehrels et al., 2006]) were analyzed before each sample and after every five unknowns. Corrections for 7of20

8 Table 2. Visually Estimated Modes of Pacheco Pass Metagraywackes a Th/U, Pb/U and (minor) Pb/Pb fractionation were carried out by comparison to the moving average of six standards. Due to increment in interelement fractionation with pit depth by as much as 5%, a least squares fit through the measured values was performed to calculate the initial value. Grains that showed >10% change in isotopic ratio during analysis, possibly due to age zoning, were discarded from the study. [21] The measured isotopic ratios and ages are reported in Table 3. Systematic errors were propagated separately, and include the aggregate uncertainties of standard age, calibration correction of the standard, composition of common Pb, and U decay constant. Values selected for concordia and probability density plots were / 207 Pb ages for samples older than 1000 Ma, and / 238 U ages for younger zircons. Analyses with >5% reverse discordance, more than >20% discordance, or high / 238 U and / 207 Pb errors (>10%) are considered unreliable and were ignored [Gehrels et al., 2006]. 7. Analytical Results Q-58 X-206 A-1 Q-24 Tectonic sheet I lower IV upper IV V Depositional age Grains Quartz Albite White mica Stilpnomelane 1 Biotite 1 Chlorite Titanite tr Opaques + limonite tr 1 tr Lawsonite Pumpellyite tr Glaucophane 4 tr Jadeitic pyroxene Rock fragments Carbonate 5 1 Carbonaceous matter tr tr Total a In units of vol %; tr, trace. [22] Figure 4 presents detrital zircon U-Pb age histograms and relative probability curves for the four analyzed Franciscan metagraywackes, plotted from 70 to 230 Ma in order to emphasize the Mesozoic zircon populations. Any or all of these units could have more recent ages of sedimentation than indicated by the youngest analyzed zircons; the latter are a function of the nature of the then-eroding source terrane and the submarine distributary channels supplying the detritus. However, in a study of 5386 zircon U-Pb ages from 61 geologically geochronologically well-dated Colorado Plateau strata, Dickinson and Gehrels [2008] employed a variety of statistical techniques on low-discordance grains and concluded that the youngest analyzed grain provides a good measure of the time of deposition in 95% of the units studied. [23] Nevertheless, to err on the conservative side, we averaged between 2 and 6 of the youngest grain ages in order to arrive at the maximum age of sedimentation. Earliest times of deposition and underplating at Pacheco Pass are as follows (the number of youngest zircon grains averaged is indicated in parentheses following the sample age). The oldest rock, Q-58 (Figure 4a; 21 grains) from the structurally highest, unit I nappe formed at or later than 102 Ma (2). An additional 30 zircon analyses were obtained for this rock, but associated errors are too large for the data to be considered reliable; nevertheless, probability peaks for the ignored U-Pb age data would have simply increased the observed populations of Figure 4a. Thus, conclusions based on the youngest analyzed zircon ages remain valid, because none of the ignored data had ages younger than those reported. A metagraywacke from a lower stratum X-206 (Figure 4b; 84 grains) of the structurally lower allochthon, unit IV, is 102 Ma (3), whereas a higher horizon, A-1 (Figure 4c; 94 grains) of the same unit is 90 Ma (2). It is somewhat reassuring that the zircon age data suggest that basal portions of the sheet were laid down before the upper horizons; nevertheless, we suspect that these first-cycle graywackes were deposited far more rapidly, with the 12 Myr age spread in apparent length of time attending deposition simply reflecting statistical uncertainty in the maximum U-Pb ages of sedimentation of the detrital zircons. The youngest analyzed rock, Q-24 (Figure 4d; 87 grains) at 86 Ma (6) is from the structurally lowest sheet, unit V. Allochthons I IV evidently are members of the Franciscan Eastern Belt, whereas unit V probably correlates with the younger, structurally lower Central Belt [Blake et al., 1984]. [24] These zircon ages indicate that Franciscan metaclastic rocks exposed at Pacheco Pass were laid down on the approaching Farallon oceanic lithosphere, and were sequentially offloaded (unit I first, unit V last) into the growing underplated trench stack during middle to Late Cretaceous time. Although the U-Pb ages do not provide strong constraints, the data are compatible with an accretionary prism model in which individually offloaded units young upward, whereas depositional ages and the intensities of recrystallization of the different imbricate sheets decrease structurally downward. As noted earlier, the mapped 1 2 km thick allochthons actually may consist of more numerous, thinner imbricated slices, but the new zircon U-Pb age data are too sparse to allow the definition of such intraunit structures. Shibata et al. [2008] recently reported comparable results from the Upper Cretaceous Paleogene Shimanto Belt of SW Japan, and concluded that underplating occurred episodically over a period of several million years, with zircon age gaps reflecting parcels of sediment that evidently continued on to greater depth than represented by the present exposures. 8. Geologic Implications of the Zircon Age Data [25] Depositional and immediately following accretionary offloading ages of metaclastic units in the Pacheco Pass 8of20

9 Table 3. U-Pb ICP-MS Analyses of Detrital Zircons, Pacheco Pass Franciscan Complex Isotope Ratios Apparent Ages Sample Analysis U (ppm) U/Th 204 Pb 204 Pb * 235 U * 238 U Error Corrected 238 U * 207Pb* 235 U 207 Pb Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Best Age X X X X X X X X X X X X X X X X X X X X X X of20

10 Table 3. (continued) Isotope Ratios Apparent Ages Sample Analysis U (ppm) U/Th 204 Pb 204 Pb * 235 U * 238 U Error Corrected 238 U * 207Pb* 235 U 207 Pb X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Best Age 10 of 20

11 Table 3. (continued) Isotope Ratios Apparent Ages Sample Analysis U (ppm) U/Th 204 Pb 204 Pb * 235 U * 238 U Error Corrected 238 U * 207Pb* 235 U 207 Pb X X X X X X X X X X X X X X X X X X X Best Age A A A A A A A A A A A A A A A A A A A A A A A A A A A of 20

12 Table 3. (continued) Isotope Ratios Apparent Ages Sample Analysis U (ppm) U/Th 204 Pb 204 Pb * 235 U * 238 U Error Corrected 238 U * 207Pb* 235 U 207 Pb A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A Best Age 12 of 20