Strontium and oxygen isotopic evidence for strike/slip movement of accreted terranes in the Idaho Batholith

Transcription

1 Lithos 96 (2007) Strontium and oxygen isotopic evidence for strike/slip movement of accreted terranes in the Idaho Batholith Elizabeth M. King a,, Brian L. Beard b, John W. Valley b a Illinois State University, Department of Geography Geology, Campus Box 4400, Normal, IL 61790, United States b University of Wisconsin- Madison, Department of Geology and Geophysics, 1215 West Dayton Street, Madison, WI 53706, United States Received 20 July 2006; accepted 7 November 2006 Available online 20 December 2006 Abstract The oxygen and strontium isotope compositions of granitic rocks of the Idaho Batholith provide insight into the magma source, assimilation processes, and nature of the suture zone between the Precambrian craton and accreted arc terranes. Granitic rocks of the Idaho Batholith intrude basement rocks of different age: Triassic/Jurassic accreted terranes to the west of the Salmon River suture zone and the Precambrian craton to the east. The age difference in the host rocks is reflected in the abrupt increase in the initial 87 Sr/ 86 Sr ratios of granitic rocks in the batholith across the previously defined line. Initial 87 Sr/ 86 Sr ratios of granitic rocks along Slate Creek on the western edge of the batholith jump from less than to greater than along an approximately 700 m transect normal to the Salmon River suture. Initial 87 Sr/ 86 Sr ratios along the Slate Creek transect do not identify a transition zone between accreted arcs and the craton and suggest a unique tectonic history during or after suturing that is not documented along other transects on the west side of the Idaho Batholith. The lack of transition zone along Slate Creek may be a primary structure due to transcurrent/transpressional movement rather than by contractional thrust faulting during suturing or be the result of post-imbrication modification Elsevier B.V. All rights reserved. Keywords: Idaho Batholith; Oxygen isotope; Sr isotope; Zircon; line 1. Introduction Corresponding author. Tel.: ; fax: address: emking@ilstu.edu (E.M. King). First-order constraints on the source materials of magmas can be provided with oxygen and Sr isotope compositions, each of which is sensitive to distinct crustal and mantle reservoirs. Oxygen isotope compositions of magmas are sensitive to the incorporation of material formed by low-temperature, surficial processes (i.e., sediments, hydrothermally altered rocks) and are insensitive to the age of source regions. Strontium isotopes, in contrast, are sensitive to both the age of potential source materials as well as lithology, where, for example, Rb-rich shales would have highly radiogenic Sr isotope compositions, and Rb-poor mafic crust of the same age would be relatively non-radiogenic. Previous strontium and oxygen isotope studies have identified the presence of a crustal boundary, the Salmon River suture zone, on the western side of the Idaho Batholith (Fig. 1) (e.g. Fleck and Criss, 1985; e.g. Criss and Fleck, 1987; Fleck, 1990; Leeman et al., 1992; Manduca et al., 1992). Suturing between the craton and accreted terranes occurred at approximately Ma (Lund and Snee, 1988; Getty et al., 1993). The suture is coincident with the later western Idaho shear /$ - see front matter 2006 Elsevier B.V. All rights reserved. doi: /j.lithos

2 388 E.M. King et al. / Lithos 96 (2007) Fig. 1. Simplified geologic map of the Idaho Batholith (after Bennett and Knowles, 1985) indicating sample localities. All sample locality numbers on this map have the prefix 98IB-. Fig. 2 shows localities of samples with the prefix 01IB-. Sample numbers in circles indicate zircon sample locations. Transects made across the Salmon River suture zone are labeled as the Slate Creek transect, South Fork of the Clearwater River (SFCR), and the McCall area. The SFCR is the location of samples from Fleck and Criss (1985). The McCall area is the location of samples from Manduca et al. (1992). Slate Creek is the area of this study and enlarged in Fig. 2. The Salmon River suture zone is coincident with the younger western Idaho shear zone and the line. zone (WISZ) that is associated with tectonic activity as old as 110 Ma continuing through 88 Ma (Lund and Snee, 1988). Shear indicators such as structural fabrics, fold vergence, and displacement on faults suggest rightlateral transpression for accretion along the Salmon River suture in the Slate Creek area (Lund and Snee, 1988). East west compressive deformation and dextral transpression have also been documented along the batholith margin either during suturing or later shearing (Manduca et al., 1993; McClelland et al., 2000). Cretaceous granitic rocks of the batholith intruded both sides of the Salmon River suture as well as the western Idaho shear zone itself. The isotopic compositions of granitic rocks that intruded east of the boundary indicate mixing of Phanerozoic juvenile arc and Precambrian crustal sources such as metasediments or orthogneisses. The isotopic compositions of granitic rocks on the west side of the crustal boundary are interpreted to represent magmas that ascended through a juvenile volcanic arc terrane (Leeman et al., 1992; Manduca et al., 1992). Magmas within the suture zone record the transition from accreted terrane to cratonic basement rocks as represented by increased initial 87 Sr/ 86 Sr ratios and δ 18 O values eastward (Fleck and Criss, 1985; Criss and Fleck, 1987; Fleck, 1990; Leeman et al., 1992; Manduca et al., 1992). Interpretations of isotopic and structural data from the Salmon River suture zone suggest a sub-vertical terrane boundary that includes a transition zone containing imbricated oceanic and continental lithospheric mantle (Leeman et al., 1992; McClelland et al., 2000). This study examines the Sr and O isotope geochemistry of granitic plutons along a transect across the Salmon River suture zone along Slate Creek (latitude approximately ) (Figs. 1 and 2) to investigate the geometry of the suture. The initial 87 Sr/ 86 Sr ratios along the Slate Creek transect do not identify a transition zone between accreted arcs and the craton. The lack of transition zone suggests either no significant imbrication of the continental and oceanic lithosphere at mid-crustal levels and/or a lack of thrusting during suturing and later shearing events along the western side of the Idaho Batholith. Geochemical data from the Slate Creek transect across the Salmon River suture zone document a very different geologic history in this regard than comparable transects to the north and south along strike of the crustal boundary. 2. Regional geology The Idaho Batholith consists of mesozonal granitic rocks of both Cretaceous and Tertiary age (Vallier and Brooks, 1987). The southern, and larger, portion of the batholith is defined as the Atlanta lobe (25,000 km 2 ), whereas the northern portion is defined as the Bitterroot lobe (14,000 km 2 )(Fig. 1). The emplacement of the Idaho Batholith spans a significant range in ages, generally from 100 to 54 Ma, and many of the mesozonal intrusions contain zircons with inherited cores of Proterozoic age (Armstrong et al., 1977; Chase et al., 1978; Bickford et al., 1981; Chase et al., 1983; Shuster and Bickford, 1985; Lund and Snee, 1988; Toth and Stacey,

3 E.M. King et al. / Lithos 96 (2007) Fig. 2. Simplified geologic map of the Slate Creek area from Lund (1995). Granitic samples with the prefix 01IB- are shown in open circles and granitic samples with the prefix 98IB- are shown in filled circles. Triassic/Permian metamorphic rocks are shown in open squares. The location of the line as shown is based on data from this study. The Salmon River suture zone and western Idaho shear zone are roughly coincident with the line. 1992; Manduca et al., 1993; Mueller et al., 1995; Snee et al., 1995; Foster and Fanning, 1997). Tonalitic rocks and orthogneiss samples from the far western side of the batholith are Ma and the ages for granitic rocks of the Idaho Batholith are generally young to the east (Lund and Snee, 1988; Manduca et al., 1993). Magmatic epidote is found in undeformed rocks with U Pb zircon ages of approximately 90 Ma that occur on the western edge of the batholith suggesting emplacement depth greater than 20 km (Zen, 1985; Lund and Snee, 1988; Snee et al., 1995). Assuming the felsic magmas in the Bitterroot lobe were hydrous and melting was initiated by the dehydration of muscovite from metamorphic basement rocks, the melt likely only rose slightly from the depth at which melting occurred, approximately 30 km. (Hyndman, 1981). Tonalitic units typical of the western edge of the batholith would likely have a deeper source, up to 70 km (Hyndman, 1981). The Cretaceous units of the Bitterroot lobe are thought to have intruded at depths of approximately km, based on metamorphic mineral assemblages in the country rocks and the presence of magmatic muscovite in the granites (Hyndman, 1981; Toth, 1987; Foster and Fanning, 1997). Beginning by 88 Ma, pluton emplacement occurred at successively shallower levels during the development of the batholith to depths of approximately 5 10 km for the youngest plutons (Wiswall and Hyndman, 1987; Snee et al., 1995). Detailed studies of the oxygen isotope compositions of zircon and garnet from igneous rocks of the Idaho Batholith provide strong evidence for assimilation of a high-δ 18 O source, most likely sedimentary rocks, during the crystallization of Cretaceous garnet-bearing granitic magmas (King and Valley, 2001). These minerals were chosen for oxygen analysis because of their resistance to post magmatic alteration, and because garnets are seen texturally to have formed later than zircon. The measured fractionation between coexisting garnet and zircon in the batholith ranges from 0.12 to 0.87, significantly greater than the near-zero fractionation expected under equilibrium conditions, suggesting that the assimilation of a high-δ 18 O component occurred after zircon crystallization, but prior to garnet crystallization (King and Valley, 2001). Incorporation of up to approximately 10% high-δ 18 O metasediments into the granitic magmas after the crystallization of zircon raised the δ 18 O(magma) and subsequently δ 18 O(garnet) (King and Valley, 2001).

4 390 E.M. King et al. / Lithos 96 (2007) The isotopic transition that occurs along the western edge of the Atlanta lobe includes an abrupt jump in initial 87 Sr/ 86 Sr ratios of intrusive rocks, from approximately to greater than eastward (Armstrong et al., 1977; Fleck and Criss, 1985; Criss and Fleck, 1987). The boundary between Precambrian basement on the east side of the boundary and Triassic/Jurassic accreted terranes on the west has been termed the Salmon River suture zone (McClelland et al., 2000), and this zone includes the Late Cretaceous western Idaho shear zone, as well as rocks of the Idaho Batholith. The Triassic and Permian Wallowa Terrane is the basement rock located west of the Precambrian continental margin in Idaho (Siberling and Jones, 1984). The plate is subdivided into different plates and blocks based on structure and lithology. The North Fork block, the farthest east portion of the Wallowa Terrane, is located west of the Salmon River suture zone and east of the North Fork reverse fault. Rocks of the North Fork block are typically Triassic or Permian metamorphosed mafic volcanic rocks (Lund et al., 1993). To the west of the North Fork reverse fault is the Rapid River plate. This plate contains Triassic or Permian metamorphic volcanic rocks of the Wallowa Terrane above the Rapid River thrust fault (Lund et al., 1993). Rocks below the Rapid River thrust fault, not given a unique plate or block name, include rocks of the Late Triassic Martin Bridge Formation. 3. Sample description 3.1. Slate Creek transect Samples for this study were collected from granitic rocks of the Atlanta and Bitterroot lobes of the Idaho Batholith (Fig. 1). A west to east transect on the western edge of the Atlanta lobe was collected along Slate Creek (Figs. 1 and 2) to investigate the transition from accreted terranes on the west to the Precambrian craton on the east. Sample 98IB-31 is a late Cretaceous foliated tonalite intruded into the late Triassic Martin Bridge Formation of the Wallowa terrane (Fig. 2). Samples 98IB-32, 98IB-33, and 01IB-3 to 01IB-8, 01IB-11, and 01IB-15 to 01IB-24 are granitic rocks intruded into Mesozoic volcanic and volcaniclastic rocks of the accreted North Fork block. Samples 98IB-34, 98IB-35, 01IB-1, and 01IB-2 are tonalites and granodiorites intruded into Proterozoic to Paleozoic metasedimentary rocks. Samples of older volcanic arc-related country rocks were also collected along the Slate Creek transect to constrain the isotopic compositions of wall rocks that may have been assimilated into granitic magmas. These samples, 01IB-9, -10, -12, -13, and -14 are Triassic, possibly Permian, rocks of the North Fork block (Lund, 1995). Hornblende gneisses and schists are the dominant rock types with a mafic volcanic protolith and may be correlative with the Seven Devils Group (Lund et al., 1993) Geochronology Crystallization ages for samples in this study from the entire batholith likely range from 120 Ma to 54 Ma, where older units tend to occur on the western margin, decreasing in age eastward (Manduca et al., 1992). Ages for samples of this study are estimated from correlative intrusions along the Salmon River suture zone, or ages determined for nearby samples (Bitterroot lobe samples and 98IB-68). The oldest undeformed plutons on the Slate Creek transect have a hornblende 40 Ar/ 39 Ar plateau-age of 84 Ma, a minimum estimate of emplacement, and U Pb zircon ages for these rocks are 89±5 Ma (Lund and Snee, 1988). Since many Slate Creek samples in this study are deformed, the average age of 100 Ma was assigned for all samples from the western side of the batholith in order to calculate initial 87 Sr/ 86 Sr ratios (Manduca et al., 1992). Other samples from the Atlanta lobe that are not closely tied to localities where age determinations have been made were assigned an age of 80 Ma, an average age of the batholith (Fleck and Criss, 1985; Fleck, 1990). The variations observed in initial 87 Sr/ 86 Sr ratios are well outside the variations due to uncertainties in the age. For further discussion of the uncertainty in initial Sr isotope calculations resulting from age uncertainties, see the Analytical methods section. Inherited Proterozoic cores have been documented in geochronology studies (e.g. Shuster and Bickford, 1985; Toth and Stacey, 1992; Snee et al., 1995; Foster and Fanning, 1997). Massbalance calculations and analytical data indicate that the inherited cores do not have significant impact on δ 18 O (zircon) values (King and Valley, 2001). 4. Analytical methods Mineral separates were prepared at the University of Wisconsin and Illinois State University using standard crushing, hydraulic, magnetic, and heavy-liquid techniques. The least magnetic fraction of zircon from each sample was analyzed for oxygen isotope ratio. Geochemical analyses of 11 major elements and 6 trace elements, including Zr, Rb, and Sr, were obtained on fused whole-rock powders using a Siemens SRS 3000 sequential X-ray fluorescence spectrometer at XRAL Laboratories, Canada. Rb and Sr concentrations,

5 E.M. King et al. / Lithos 96 (2007) reported in Table 1, were determined by isotope dilution at the University of Wisconsin Stable isotope analyses Oxygen isotope ratios were determined on 1 2 mgsize mineral separates of zircon and quartz at the University of Wisconsin by laser fluorination in the presence of BrF 5, and using gas-source mass spectrometry (Valley et al., 1995). Oxygen isotope compositions are reported in the standard δ-notation relative to Vienna Standard Mean Ocean Water (V-SMOW). Zircon grains were analyzed after powdering in a boron carbide mortar and pestle to avoid any possible grain-size effects during analysis and to maximize the efficiency of fluorination. Quartz grains were analyzed using rapid heating and a defocused laser beam (Spicuzza et al., 1998a). Wholerock powders were analyzed using an airlock sample chamber in order to eliminate pre-reaction of powders prior to or during analysis of other samples in the chamber (Spicuzza et al., 1998b). The uncertainty in δ 18 O values is estimated to be ± 0.09 (1 SD), based on 70 analyses of the UWG-2 garnet standard analyzed concurrently with the samples. On several days, daily averages of UWG-2 differed ( 0.13 to ) from the accepted value of (relative to for NBS-28). Analyses of samples have been corrected for these small shifts, following the practice of Valley et al. (1995). The average correction was The average reproducibility of duplicated zircon and quartz analyses is ±0.05 (n = 23 samples) and ±0.10 (n=36 samples), respectively. Magmatic δ 18 O values of the whole rock are calculated from the δ 18 O(Zrc) value and SiO 2 content of the whole-rock sample (Valley et al., 2005) using a regression of measured oxygen isotope data over a range of SiO 2 contents in meta-igneous rocks. This procedure avoids the uncertainties that are unavoidable for measured δ 18 O(WR) values, which have frequently been altered by sub-solidus processes. This calculation is relatively insensitive to small changes in SiO 2 content. For example, assuming a measured δ 18 O(Zrc) value of +7.0, a change in whole-rock SiO 2 content by 5% results in a 0.3 change in the calculated δ 18 O(magma) value. Where δ 18 O(Zrc) values were unavailable, δ 18 O (magma) was calculated from δ 18 O(Qtz) (Valley et al., 2003). In samples with co-existing zircon and quartz where closed-system diffusional exchange of oxygen appears to be the only sub-solidus alteration, the average difference between δ 18 O (magma) calculated from quartz and zircon is 0.45, although this ranges from 0.44 to The calculated δ 18 O(magma) values may be compared to the measured δ 18 O values for whole-rock samples, which, in most cases, provide an assessment of the degree of alteration. The measured whole-rock δ 18 O typically is intermediate between the δ 18 O(magma) calculated from zircon and the δ 18 O (magma) calculated from quartz. The calculated δ 18 O (magma) values based on quartz tend to be the highest. The slightly elevated δ 18 O(magma) values calculated from quartz are likely to the result o sub-solidus exchange of δ 18 O in the mineral during cooling due to the low blocking temperature and fast diffusion rates in quartz compared to zircon (Farver and Yund, 1991; Eiler et al., 1992, 1994) Strontium isotope analyses Strontium isotope analyses were performed at the University of Wisconsin Radiogenic Isotope Laboratory. Whole rock samples, approximately 50 mg, were spiked with a mixed 87 Rb 84 Sr spike and dissolved in Parr bombs using a mixture of HF and HNO 3. Rubidium and Sr were separated using 2.5 M HCl and cation exchange resin. The average procedural blanks for Rb and Sr were 2.9 ng and 0.64 ng, respectively, which are negligible. All isotope measurements were made by thermal ionization mass spectrometry using a 7 collector Micromass Sector 54. Strontium isotope ratios were collected using a 3-jump multi-collector dynamic analysis routine with exponential normalization to 86 Sr/ 88 Sr= The reported 87 Sr/ 86 Sr ratios are the average of 120 ratios per analysis, and the 88 Sr ion intensity was maintained at A. During the course of this study the average 87 Sr/ 86 Sr ratio of NIST SRM-987 was ± (2 SD; n=99). Initial 87 Sr/ 86 Sr ratios were calculated based on ages listed in Table 1. The measured 87 Rb/ 86 Sr ratios for all samples are less than unity (excluding 98IB-24 and 01IB-4), and the average 87 Rb/ 86 Sr is 0.3. Because the 87 Rb/ 86 Sr ratio for nearly all of the samples considered in this study is low, initial 87 Sr/ 86 Sr ratios will not be strongly affected by uncertainties in the crystallization age. For example, assuming an age of 80±30 Ma, the uncertainty in the initial 87 Sr/ 86 Sr ratio would be ± using an average 87 Rb/ 86 Sr of 0.3. The error in the initial 87 Sr/ 86 Sr introduced from the uncertainty in the age of the samples is not sufficiently large to affect interpretations of the data. 5. Results 5.1. Radiogenic isotopes The strontium isotope compositions of whole-rock powders are reported in Table 1. Initial 87 Sr/ 86 Sr ratios

6 392 E.M. King et al. / Lithos 96 (2007) Table 1 Strontium isotope composition of rocks from the Idaho Batholith Sample Rb (ppm) Sr (ppm) 87 Rb/ 86 Sr a 87 Sr/ 86 Sr a 2σ error b Age isr c 2σ error Latitude Longitude Rock type Atlanta Lobe granitic rocks 98IB ʺ ʺ garnet-bearing, foliated granodiorite 98IB ʺ ʺ garnet-bearing, foliated granodiorite 98IB ʺ ʺ foliated granodiorite 98IB ʺ ʺ foliated tonalite 98IB ʺ ʺ garnet-bearing, foliated granodiorite 98IB ʺ ʺ 2 mica, foliated tonalite (477.4) c (.5619) ( ) ( ) (.70952) ( ) 98IB ʺ ʺ tonalite 98IB ʺ ʺ 2 mica granodiorite (336.2) (.8862) ( ) ( ) ( ) ( ) 98IB IB ʺ ʺ granodiorite (889.2) (0.2781) ( ) ( ) ( ) ( ) 98IB ʺ ʺ tonalite Slate Creek transect granitic rocks 98IB ʺ ʺ 2 mica, garnet-bearing, foliated tonalite 98IB ʺ ʺ gneissic tonalite 98IB ʺ ʺ 2 mica, garnet-bearing, gneissic tonalite (792.8) (.03797) (0.7) ( ) ( ) ( ) 98IB ʺ ʺ tonalite 98IB ʺ ʺ granodiorite 01IB ʺ granodiorite 01IB ʺ ʺ tonalite 01IB ʺ ʺ granodiorite 01IB ʺ ʺ quartz monzonite 01IB ʺ ʺ foliated tonalite 01IB ʺ ʺ tonalite 01IB-7 NA NA NA NA NA NA NA ʺ ʺ quartz monzonite 01IB ʺ ʺ tonalite 01IB IB ʺ ʺ tonalite 01IB ʺ ʺ granite 01IB IB ʺ ʺ tonalite 01IB ʺ ʺ tonalite 01IB ʺ ʺ foliated tonalite 01IB ʺ ʺ tonalite 01IB ʺ ʺ tonalite 01IB ʺ ʺ granodiorite 01IB ʺ ʺ granodiorite Triassic/Permian metamorphics 01IB ʺ garnet, hornblende gneiss 01IB ʺ hornblende gneiss 01IB ʺ ʺ hornblende gneiss 01IB ʺ ʺ hornblende gneiss 01IB ʺ ʺ biotite, garnet schist

7 E.M. King et al. / Lithos 96 (2007) Table 1 (continued ) Sample Rb (ppm) Sr (ppm) 87 Rb/ 86 Sr a 87 Sr/ 86 Sr a 2σ error b Age isr c 2σ error Latitude Longitude Rock type Bitterroot Lobe granitic rocks 98IB ʺ ʺ granite 98IB ʺ ʺ 2 mica granite 98IB ʺ ʺ garnet-bearing, granite 98IB ʺ ʺ granodiorite (405.9) (0.4085) ( ) ( ) (.71001) ( ) 98IB ʺ ʺ 2 mica granite 98IB ʺ ʺ 2 mica granodiorite 98IB ʺ ʺ granite a: strontium isotopic data are for whole-rock powders. b: errors for measured 87 Sr/ 86 Sr are 2 SE from internal counting statistics. Errors for initial 87 Sr/ 86 Sr ratios are based on the reported 2 SE for the measured 87 Sr/ 86 Sr ratio and an error of 0.63% in the measured 87 Rb/ 86 Sr as determined from replicate analyses of samples from this study. c: values from replicate mass spectrometry on the same solution are in parentheses. Replicate analyses are plotted as average values. for granitic samples in this study range from to Strontium contents for granitic samples range from 22 to 1848 ppm, and average 626 ppm. These compositions are well within the range of previously published data (Fleck and Criss, 1985; Criss and Fleck, 1987; Fleck, 1990). We do not find, however, a correlation between 87 Sr/ 86 Sr and Sr contents as found by Fleck and Criss (1985) (Fig. 3), suggesting that the data for rocks along Slate Creek reflect multiple processes in addition to mixing between juvenile arc magmas and Precambrian crustal sources. The Triassic/Permian metamorphic rocks have a restricted range in initial 87 Sr/ 86 Sr ratios of ± Strontium isotope compositions were not determined for epizonal plutons because their very high-rb/ Sr ratios would produce large errors in initial 87 Sr/ 86 Sr ratios. Eleven granitic samples from the Slate Creek transect were calculated to have initial 87 Sr/ 86 Sr ratios less than (98IB-31, -32, -33, 01IB-15, -16, -18, -19, -21, -22, -23, and -24). These samples are west of, or lie within, the Salmon River suture zone and therefore intrude the accreted terranes. All other samples are east of the suture zone, intrude the Precambrian craton, and have initial 87 Sr/ 86 Sr ratios greater than (98IB-34, -35, 01IB-1, -2, -3, -4, -5, -6, -7, -8, -11). Sample 01IB-15 is the granitic sample farthest east with an initial 87 Sr/ 86 Sr ratio less than Sample 01IB-11 is the granitic sample farthest west with an initial 87 Sr/ 86 Sr ratio greater than The boundary between granitic rocks along Slate Creek that have initial 87 Sr/ 86 Sr ratios less than and those that are greater occurs between samples 01IB-11 and 01IB-15 with less than 1 km between these two samples Oxygen isotopes Oxygen isotope compositions of zircon, quartz, garnet, and whole-rock powder are reported in Table 2. Fig. 3. Plot of initial 87 Sr/ 86 Sr vs. 1/Sr for rocks of the Slate Creek transect. Granitic rocks intruded into the Precambrian craton are plotted as filled diamonds. Granitic rocks intruded into the accreted terrane are plotted filled circles. Triassic metamorphic rocks are plotted as open squares. Possible mixing components are plotted; juvenile arc magma as the star with an initial 87 Sr/ 86 Sr of and 1200 ppm Sr and the Precambrian components, orthogneiss with an initial 87 Sr/ 86 Sr of and 159 ppm and metasediments with an initial 87 Sr/ 86 Sr of and 221 ppm Sr, are located well off the figure as plotted. There is no linear trend between a Sr-rich, low initial 87 Sr/ 86 Sr primitive magmas and Sr-poor, high initial 87 Sr/ 86 Sr Precambrian orthogneiss suggesting a lack of simple mixing between continental and oceanic magma sources. Sr data for juvenile arc magma and Precambrian components are from Fleck (1990) and Manduca et al. (1992).

8 394 E.M. King et al. / Lithos 96 (2007) Table 2 Measured oxygen isotope composition of rocks and minerals from the Idaho Batholith and calculated δ 18 O of magma Sample Zircon δ 18 O Quartz δ 18 O Garnet δ 18 O Δ(Zrc Gnt) Whole-rock δ 18 O Magma δ 18 O a Magma δ 18 O b SiO 2 (wt.%) Atlanta Lobe granitic rocks 98IB ± IB ± ± IB ± IB ± ± IB ± ± ± IB ± IB ± ± IB ± ± IB ± ± IB ± ± Slate Creek transect granitic rocks 98IB ± Altered quartz IB ± IB ± ± ± IB ± ± IB ± IB ± ± IB ± IB ± ± IB ± ± IB ± ± IB ± ± IB ± ± IB ± ± IB IB ± ± ± IB ± IB ± ± ± IB ± IB ± IB ± ± IB ± ± IB Triassic/Permian metamorphics 01IB ± ± IB ± IB IB ± IB ± ± Bitterroot Lobe granitic rocks 98IB ± ± IB ± ± IB ± ± ± IB ± IB ± ± IB ± ± IB ± ± Altered quartz 70.4 a: magma δ 18 O is calculated from zircon. b: magma δ 18 O is calculated from quartz.

9 E.M. King et al. / Lithos 96 (2007) Zircons from granitic samples in this study have an average δ 18 O value of +7.0±0.8 (n=25), which lies within the average for the Idaho Batholith, including epizonal Tertiary plutons of the batholith (King and Valley, 2001). The average calculated δ 18 O(magma) value for granitic rocks of the Idaho Batholith is +8.9± 0.8 based on zircons, and 9.3 ±0.7 based on quartz. The higher calculated δ 18 O(magma) values based on quartz are likely too high reflecting the effects of hydrothermal fluids or diffusional exchange of oxygen during cooling (King and Valley, 2001). In addition, anomalously high-δ 18 O(quartz) values from samples along Slate Creek may reflect recrystallization during shearing events. The maximum calculated δ 18 O(magma) value is throughout the batholith. Whole-rock powders for samples of granitic rocks along Slate Creek range up to +13.3, similar to the range for whole-rock values determined by previous studies (+ 10 to +12, Criss and Fleck, 1987; Manduca et al., 1992). Based on comparison between minerals and whole-rock powders from this study, it is likely that some of the high-δ 18 O values for whole-rock samples from previous studies reflect post-magmatic alteration. Two samples in this study contain quartz with anomalous δ 18 O values, indicating significant hydrothermal alteration. 98IB-24 is proximal to an epizonal Eocene pluton that contains feldspar and biotite with δ 18 Ovalues much lower than those expected for primary igneous values (King and Valley, 2001). The measured quartz zircon fractionations yield an apparent temperature N 1000 C, which likely reflects an anomalously low δ 18 O(Qtz) value as a result of hydrothermal alteration. Sample 98IB-31 has an extremely large quartz-zircon fractionation (7.3 ); this sample is from a 3 m-wide dike of granite that intrudes calcareous arc sediments of Triassic age. Feldspar and biotite have δ 18 O values that are also elevated relative to zircon in sample 98IB-31, and the biotite zircon fractionation is reversed relative to that expected for equilibrium at igneous temperatures (King and Valley, 2001). This sample has experienced significant exchange of oxygen with the surrounding sedimentary rocks that affected quartz, feldspar, and biotite, but not zircon, which presumably crystallized earlier. In contrast to the relatively homogeneous initial 87 Sr/ 86 Sr ratios, the δ 18 O values of the Triassic/Permian metamorphic rocks of the North Fork block are highly variable. Quartz δ 18 O values from the metamorphic samples range from +6.3 to (Table 2). Wholerock δ 18 O values are equally as variable, ranging from +4.9 to Garnets from two samples (01IB-9 and -14) have δ 18 O of +2.6 and +4.4, significantly lower than the garnet-bearing granitic samples of the batholith. The δ 18 O values of the metamorphic rocks are not representative of the original mafic magmatic values. These samples have been variably altered during metamorphism. 6. Discussion The isotopic compositions of granitic rocks from the Idaho Batholith are tracers of differences in the lower crust through which the magmas ascended. Tonalitic rocks on the western edge of the batholith suggest maximum intrusion depths as deep as 20 km based on the presence of magmatic epidote, water content in the magma, metamorphic grade, and the thickness of the stratigraphic cover (Hyndman, 1981; Zen, 1985; Lund and Snee, 1988; Snee et al., 1995). The generation of magma in the more felsic portions in the northeastern Idaho Batholith has been attributed to dehydration melting due to muscovite breakdown occurring in metamorphic rocks approximately 30 km deep (Hyndman, 1981). Tonalitic rocks common on the western side of the batholith would have higher melting temperatures than the two-mica granites exposed in the northeastern portions and any melting initiated by biotite dehydration would entail magma generation perhaps as shallow as 40 km and as deep as 73 km depending on the local geotherm (Hyndman, 1981). Stable and radiogenic isotope data from the batholith indicate a similar magmatic source for all granitic rocks, likely a mixture of juvenile arc magma and Proterozoic basement rocks (Fleck and Criss, 1985; Criss and Fleck, 1987; Fleck, 1990; Manduca et al., 1992; Toth and Stacey, 1992; Mueller et al., 1995; King and Valley, 2001). Magmas on the west side of the Idaho Batholith passed through a minimum of 10 km crust and a maximum of 53 km crust from the source location to emplacement depth. Isotopic trends can then represent the structure of the suture zone at mid-crustal levels at the time of magma generation Oxygen isotopes across the line The δ 18 O values of quartz, zircon, and garnet are higher east of the line, and this increase begins west of the line (Fig. 4). For granitic rocks located west of the line, δ 18 O(Qtz) and δ 18 O(Zrc) increase nearly 1 and remain high east of the line (Fig. 4). Garnet δ 18 O values are more variable from west to east but high-δ 18 O values are seen both east and west of the line. Zircon has been shown to be refractory and preserve its igneous isotope composition during metamorphism (Page et al., 2006). The similar increase of quartz, zircon, and garnet δ 18 O values across

10 396 E.M. King et al. / Lithos 96 (2007) ascent of the granitic magmas through the crust, the assimilant must have had an isotopic composition that was highly variable in δ 18 O values but uniformly low in initial 87 Sr/ 86 Sr ratios. The initial 87 Sr/ 86 Sr of granitic magmas generated from juvenile arc magmas would be minimally altered if the initial magmatic ratio was approximately and the assimilant was no higher than Depending on the δ 18 O of the incorporated metavolcanics, the δ 18 O of the granitic magmas would be variably affected as assimilation of wall rock proceeded. In this situation, an increase of δ 18 O(Zrc) from 5.7 to 6.7, the range of observed values from samples west of the line, suggests 8% assimilation of metavolcanic rocks with δ 18 O(WR) of 8. The isotopic data can be useful for refining models for crustal structure at depth through which the granitic magmas ascended along Slate Creek. The oxygen isotope trends across the South Fork of the Clearwater River (SFCR) and McCall area transects mimic the west east oxygen isotope trend along Slate Creek. The SFCR and McCall transects are consistent with an imbricated or structurally overlapping oceanic and continental lithospheric mantle (Fleck and Criss, 1985; Manduca et al., 1992). The oxygen isotope trend across the Slate Creek transect is consistent with a similar transition zone if considered in the absence of the Sr isotope data Strontium isotopes across the line Fig. 4. Plots of δ 18 O(zircon), δ 18 O(garnet) and δ 18 O(quartz) vs. longitude for samples from the Slate Creek transect. The observed trend of δ 18 O increasing in all minerals from west to east suggests that shearing along the western Idaho Shear Zone has not significantly altered the magmatic δ 18 O values. The west east increase is not due to the incorporation of high-δ 18 O material from the continent since initial 87 Sr/ 86 Sr ratios don't indicate any ancient component in the granitic magmas until east of (Fig. 5). Volcanic arc-derived high-δ 18 O material must influence the δ 18 O of the granitic magmas. the Salmon River suture indicate that shearing along the western Idaho shear zone has not significantly altered magmatic δ 18 O values for these minerals. The Slate Creek transect contains metavolcanic rocks of the North Fork block that are of island-arc origin (Fig. 2) (Lund, 1995). The δ 18 O values of the metavolcanic rocks along Slate Creek are highly variable with δ 18 O(Qtz) ranging from 6.28 to These island-arc rocks are likely correlative to rocks of the Wallowa terrane, which have an average δ 18 O(WR) of +15 (King and Valley, 2001). If rocks similar to the exposed metavolcanic rocks were available for assimilation at the current level of exposure or during the The initial Sr isotope composition of granitic rocks sampled from three transects across the Salmon River suture zone are plotted in Fig. 5. Transects along the SFCR and directly north of McCall, Idaho each define a similar trend from west to east (data of Fleck and Criss, 1985; Manduca et al., 1992)(Figs. 1, 5A, and C). Initial 87 Sr/ 86 Sr ratios in granitic rocks less than are characteristic of magmas intruded through accreted volcanic-arc rocks on the west. Moving eastward, the initial 87 Sr/ 86 Sr ratios abruptly increase up to 0.710, suggesting an increase of Precambrian crustal material assimilated into magmas. Although this is a nearly vertical suture, the transition zone between granitic rocks that have initial 87 Sr/ 86 Sr ratios between and suggests a zone with lithospheric mantle of intermediate Sr isotope composition that has been inferred to be continental lithospheric mantle interlayered with the oceanic lithosphere (Leeman et al., 1992; McClelland et al., 2000) (Fig. 6A). The sampling transect along Slate Creek (Figs. 1 and 2), across the Salmon River suture zone on the western edge of the Idaho Batholith, reveals a

11 E.M. King et al. / Lithos 96 (2007) transects. The abrupt transition from initial 87 Sr/ 86 Sr ratios representative of volcanic arcs to those representative of Precambrian components, as seen in the Slate Creek transect, is difficult to reconcile with thrusting of accreted rocks along the continental margin. A suturing mechanism involving convergent strike slip faulting, not thrusting, is suggested by Lund and Snee (1988) based on stratigraphic, structural, geochemical, and geochronologic data Crustal cross-section along the Salmon River suture zone Granitic rocks obtain their chemical and isotopic compositions at the magma source and during ascent through the crust, permitting inferences to be made regarding the geometry of the suture zone beneath the current level of exposure. With a minimum depth of melting at 30 km, maximum depth of melting at 73 km, and emplacement depth at 20 km, the magma should reflect the 10 to 53 km of crust through which ascent occurred (Hyndman, 1981; Lund and Snee, 1988; Snee et al., 1995). Current models for the suturing mechanism Fig. 5. Plot of initial 87 Sr/ 86 Sr vs. longitude for three transects across the line. The Slate Creek data (B) are from this study. Data for the South Fork of the Clearwater River (SFCR) are from Fleck and Criss (1985) (A). Data from the McCall area are from Manduca et al. (1992) (C). Intermediate values between and in the SFCR and McCall transects suggest a wedge of continental lithosphere overlapping mantle lithosphere resulting in a gradual, albeit laterally compressed, increase of initial 87 Sr/ 86 Sr ratios. The discontinuous trend on the Slate Creek transect suggests no overlap of mantle and continental lithosphere at depth in the crust. remarkable transition from basement rocks of Triassic/ Jurassic accreted terranes to the Precambrian craton. Eleven granitic samples have initial 87 Sr/ 86 Sr ratios less than (ave. = ± ), similar to Jurassic plutons that intruded the accreted terranes (Johnson et al., 1999) and the Triassic/Permian metamorphic rocks of this study. Over a distance of less than 1.0 km, initial 87 Sr/ 86 Sr ratios in granitic rocks along Slate Creek jump discontinuously from to greater than (Fig. 5B). There is no gradational transition zone along Slate Creek as seen in the SFCR and McCall area Fig. 6. Simplified west to east crustal cross-sections across the Salmon River suture zone (SRSZ) at the time of the main intrusive phase of the Idaho Batholith. A, modified from Leeman et al. (1992), shows a shelf of continental lithospheric mantle under a layer of oceanic lithosphere truncated due to contractional suturing. This is the model for the SFCR and McCall area transects. B shows distinct mantle zones that do not overlap and represents a model for the Slate Creek transect. Magmas generated on either side of the suture zone would have an abrupt isotopic transition from volcanic arc to cratonic affinity.

12 398 E.M. King et al. / Lithos 96 (2007) along the Salmon River suture differ between contractional thrust faulting (Wernicke and Klepacki, 1988; Strayer et al., 1989; Selverstone et al., 1992; Manduca et al., 1993) and transcurrent/transpressional motion (Lund and Snee, 1988; Snee et al., 1995). The Sr isotope data from three transects across the suture zone suggest that both models may be applicable to the suturing event on the western edge of the North American craton. Post accretion modification of isotopic trends due to transpression has also been documented in the McCall area (Giorgis et al., 2005). The granitic rocks along the SFCR transect and in the McCall area, which define a continuous change in initial 87 Sr/ 86 Sr with longitude, suggest an abrupt but gradational transition from accreted arcs to the Precambrian continent (Fig. 5A and C). The transition zone of initial 87 Sr/ 86 Sr ratios can occur where the continental margin was thrust over the accreted volcanic arc. Perhaps the continental margin along these transects contains a wedge of cratonal rock units that thins westward and is intermixed with the accreted rocks to allow for the gradual increase in the availability of ancient crustal rocks eastward with elevated 87 Sr/ 86 Sr (Fig. 6A). The mixing of magma sources suggests interaction between different crustal compositions, but the nearly vertical nature of the transition along the McCall and SFCR transects is caused by compression along the WISZ resulting in anywhere from 10 to 75 km shortening (Giorgis et al., 2005). The granitic rocks exposed along Slate Creek suggest a very different structure for the transition between the arc and the craton. One possible interpretation for lack of a gradational isotopic transition zone is that the transitional magmas never formed. In order for magmas with intermediate initial 87 Sr/ 86 Sr values to not form, minimal to no overlap between continental and oceanic crust must occur between the depth of magma generation and the depth of intrusion. Assuming granitic rocks with intermediate initial 87 Sr/ 86 Sr values never formed, the abrupt increase in initial 87 Sr/ 86 Sr over a distance of less than 1 km as presently exposed indicates a nearly vertical suture with little mingling of oceanic and continental lithosphere (Fig. 6B). Strike slip motion during suturing would result in little overlap of continental and oceanic lithosphere, requiring a different geometry at mid-crustal levels (Fig. 6B). The depth to which the lack of interaction between the oceanic and continental lithosphere extends could be as deep as 73 km based on the estimated depth of magma generation. The increase in δ 18 O eastward along the Slate Creek transect, but constant initial 87 Sr/ 86 Sr west of , indicates the increasing availability of high-δ 18 O rocks for assimilation into the magmas upon their ascent. One potential high-δ 18 O endmember available for assimilation into granitic magmas are rocks on the flanks of the Triassic and Permian accreted volcanic arcs if hydrothermally altered at low temperature. Other possible geologic histories that account for the lack of granitic rocks with intermediate initial 87 Sr/ 86 Sr ratios at Slate Creek could involve magmas that formed in a transition zone between the continent and accreted terranes that were subsequently removed or destroyed during tectonic events. High-angle normal faults occur parallel to the Salmon River suture zone along which uplift occurred during the emplacement of undeformed granitic rocks of the Idaho Batholith (Lund and Snee, 1988). These faults could have uplifted granitic rocks of the transition zone that were subsequently removed during compressional or erosional events. Steeply dipping Cenozoic faults are noted in the McCall region with 100 m up to 300 m offset observed (Manduca et al., 1993; Tikoff et al., 2001). Along Slate Creek, however, the major normal fault occurs approximately two miles east of the location of the discontinuity in initial 87 Sr/ 86 Sr values and if throw along the fault is comparable to the McCall region, 300 m is not a sufficient displacement to remove the granitic rocks within the transition zone. Shearing along the WISZ could also have displaced transitional rock units to the north or south from the original location along the Slate Creek transect. Mapping in the area does not clearly indicate the truncation or pinching out of units along the roughly north south running WISZ (Lund et al., 1993; Lund, 1995). Deformation within the rocks of the shear zone makes the differentiation of individual igneous units difficult. The unit mapped as the tonalite to granodiorite tectonite in Fig. 2 is suggested to be either multiple intrusive units that have been strongly deformed or else compositional variation in a single intrusion (Lund et al., 1993). The ductile nature of deformation within the shear zone makes a definitive interpretation of unit continuity difficult to achieve. Extensive sampling just a few kilometers to the north and south of the Slate Creek transect may be able to distinguish the presence of strike slip displaced granitic rocks with intermediate initial 87 Sr/ 86 Sr values, but exposure of the granitic rocks becomes a difficulty as the Columbia River basalts cover significant portions of the western Idaho Batholith (Fig. 2). 7. Conclusions The Salmon River suture zone on the western edge of the Idaho Batholith has previously been recognized as a nearly vertical suture that contains imbricated sections

13 E.M. King et al. / Lithos 96 (2007) of continental and oceanic mantle components (Armstrong et al., 1977; Fleck and Criss, 1985; Criss and Fleck, 1987; Leeman et al., 1992; Manduca et al., 1993; McClelland et al., 2000). The transect across the line along Slate Creek is distinct in the lack of transitional initial 87 Sr/ 86 Sr ratios. The jump from initial 87 Sr/ 86 Sr ratios less than to greater than occurs over a distance less than 1 km. Oxygen isotope ratios in these samples indicate a more gradual increase from samples with an arc signature on the west to samples influenced by Precambrian basement with higher initial 87 Sr/ 86 Sr and δ 18 O to the east. This contrast in isotopic trends indicates the increased assimilation of metavolcanic components with a higher δ 18 O, but low 87 Sr/ 86 Sr, into the granitic magma within the suture zone. Three possible models are proposed to explain the lack of granitic rocks with intermediate initial 87 Sr/ 86 Sr along Slate Creek. Granitic rocks with transitional initial 87 Sr/ 86 Sr could have been removed during uplift, but no major high-angle faults are coincident with the Sr discontinuity along Slate Creek. The nature of the suture could be such that intermediate rocks never formed. The transition from oceanic lithosphere to continental lithosphere could be not only nearly vertical at significant depths in the crust, but also there could be no interaction between continental and oceanic lithosphere. This type of continental margin with no interlayering of different crust would never allow rocks of intermediate initial 87 Sr/ 86 Sr values to form. Finally, the transitional granitic rocks could have been sheared to the north or south during younger tectonic activity. Significant shortening of the magmatic arc has been documented just to the south of the Slate Creek transect, in the McCall region (Giorgis et al., 2005), and this shortening combined with post-accretion strike slip motion could create a steep transition from accreted terranes to the Precambrian continent with no transitional rocks. The suturing of volcanic arcs onto the Precambrian craton and perhaps subsequent tectonic activity appear to have occurred with different mechanisms along the strike of the craton margin. The SFCR and McCall area sutured with thrusting to overlap oceanic and continental crust, while the Slate Creek suture zone likely occurred with more of a strike slip motion to either prevent the interaction of the crustal fragments at depth or displace rocks containing a transitional radiogenic isotopic signature. Contractional stress during suturing without strike slip motion would mean that rocks with intermediate initial 87 Sr/ 86 Sr values still exist along Slate Creek but the transition zone would be compressed as observed in the McCall and SFCR transects. Acknowledgements We thank M.J. Spicuzza for the help with stable isotope analyses, S.L. Moore and J. Murphey for the sample preparation and petrography, and N. Mahlen for the help with radiogenic isotope analyses. The authors thank C. Johnson for the assistance with Sr isotope analyses and manuscript editing. We thank S. Giorgis, W. Leeman, W. McClelland, P. Mueller, and R. Frost for the helpful reviews of this manuscript. This research was supported by NSF grants EAR and EAR (to JWV), DOE grant FG02-93-ER14389 (to JWV), a University Research Grant from Illinois State University (EMK), and a Dean Morgridge Wisconsin Distinguished Graduate Fellowship (to EMK). The Alumni Board of the Department of Geology and Geophysics at the University of Wisconsin also supported the fieldwork. References Armstrong, R.L., Taubeneck, W.H., Hales, P.O., Rb Sr and K Ar geochronometry of Mesozoic granitic rocks and their Sr isotopic composition, Oregon, Washington, and Idaho. Geological Society of America Bulletin 88, Bennett, E.H., Knowles, C.R., Tertiary plutons and related rocks in central Idaho. United States Geological Survey Bulletin 1658 A-S, Bickford, M.E., Chase, R.B., Nelson, B.K., Shuster, R.D., Arruda, E.C., U Pb studies of zircon cores and overgrowths, and monazite: implications for age and petrogenesis of the Northeastern Idaho Batholith. Journal of Geology 89, Chase, R.B., Bickford, M.E., Tripp, S.E., Rb Sr and U Pb isotopic studies of the northeastern Idaho Batholith and border zone. Geological Society of America Bulletin 89, Chase, R.B., Bickford, M.E., Arruda, E.C., Tectonic implications of Tertiary intrusion and shearing within the Bitterroot Dome, Northeastern Idaho Batholith. Journal of Geology 91, Criss, R.E., Fleck, R.J., Petrogenesis, geochronology, and hydrothermal systems of the northern Idaho Batholith and adjacent areas based on 18 O/ 16 O, D/H, 87 Sr/ 86 Sr, K Ar, and 40 Ar/ 39 Ar studies. In: Vallier, T.L., Brooks, H.C. (Eds.), Geology of the Blue Mountains Region of Oregon, Idaho, and Washington; the Idaho Batholith and its Border Zone. United States Geological Survey Professional Paper, vol. 1436, pp Eiler, J.M., Baumgartner, L.P., Valley, J.W., Intercrystalline stable isotope diffusion; a fast grain boundary model. Contributions to Mineralogy and Petrology 112, Eiler, J.M., Baumgartner, L.P., Valley, J.W., Fast grain boundary; a Fortran-77 program for calculating the effects of retrograde interdiffusion of stable isotopes. Computers & Geosciences 20, Farver, J.R., Yund, R.A., Oxygen diffusion in quartz; dependence on temperature and water fugacity. Chemical Geology 90, Fleck, R.J., Neodymium, strontium, and trace-element evidence of crustal anatexis and magma mixing in the Idaho Batholith. In: Anderson, J.L. (Ed.), The Nature and Origin of Cordilleran