Ages of detrital monazite and zircon from alluvium collected from the French Broad River drainage basin, an orogen-crossing main trunk river, and alluvium in first-order tributary streams, provide an unconventional perspective for examining the regional tectonic and metamorphic history of the southern Appalachian orogen (eastern United States). The French Broad River system samples migmatitic Ashe–Tallulah Falls suite paragneisses (with inferred Neoproterozoic clastic protoliths) of the Eastern Blue Ridge and western Inner Piedmont, Mesoproterozoic basement orthogneisses, numerous Paleozoic metaplutonic gneisses, and tectonite equivalents of these lithologies in the Brevard fault zone. Middle Ordovician ages dominate the monazite age spectrum. Monazite from tributaries has a dominant 208Pb-232Th age peak ca. 463 Ma. Monazite from the French Broad River alluvium suite yields a dominant 208Pb-232Th age mode ca. 450 Ma, but differs from the tributaries in having scattered Mesoproterozoic, Silurian–Devonian, and Carboniferous ages. Electron microprobe total Th-U-Pb chemical ages for selected tributary monazite grains also analyzed by ion microprobe reveal additional monazite growth events (i.e., metamorphic reaction) at 480–475 Ma and 445–440 Ma. Tributary and French Broad River zircon age spectra are dominated by Mesoproterozoic and Ordovician grains. Most Ordovician zircon from the French Broad River has Th/U > 0.1 and is most likely derived from the Henderson orthogneiss (447.6 ± 5.4 Ma), the largest pluton in the French Broad River headwaters region. A minor zircon age population at 450 Ma, represented primarily by metamorphic zircon rims with Th/U < 0.05 on magmatic Mesoproterozoic zircon cores, is present in tributaries and samples of migmatitic Ashe–Tallulah Falls suite paragneiss. Rare Neoproterozoic ages of 800–700 Ma and 600–550 Ma are present in all zircon data sets.

Ordovician monazite ages and zircon rim ages correspond to Taconian metamorphism in the Eastern Blue Ridge province. The dominant monazite age mode (463 Ma) from tributaries is slightly older than the Ordovician metamorphic zircon age mode (450 Ma), which is accounted for by monazite growth primarily via prograde metamorphic reactions, and zircon growth by melt-forming reactions in migmatites at the thermal peak. The scattered middle to late Paleozoic zircon and monazite ages attest to the lack of significant thermotectonic and magmatic events of that age in the Southern Blue Ridge providing sediment to the French Broad River drainage system. This dearth of ages is consistent with the pattern of nonpenetrative late Paleozoic deformation, retrograde metamorphism, and scattered plutonism northwest of the Brevard fault zone. A reasonable source of Neoproterozoic zircon in alluvium or in Ashe–Tallulah Falls paragneisses is Neoproterozoic rift-related magmatic rocks. The rarity of Mesoproterozoic monazite compared to zircon contrasts with the marked abundance of monazite in Ashe–Tallulah Falls paragneisses, and emphasizes the responsiveness of monazite compared to zircon in regional metamorphism. The abundant Mesoproterozoic zircon was inherited from Mesoproterozoic basement lithologies by Neoproterozoic sediments, preserved through regional metamorphism and three phases of orogenesis, and persists in modern alluvium being shed by the orogen.

Direct dating of peak regional metamorphism accompanying collisional orogenesis ideally involves isotopic analysis of metamorphic minerals with high closure temperatures for diffusion of daughter isotopes. Use of metamorphic index minerals such as garnet, staurolite, and kyanite would be most definitive for claiming one is dating metamorphism (e.g., Lanzirotti and Hanson, 1995). However, the relatively low concentration of radiogenic parent isotopes (Sm, U) in these minerals and the prevalence of monazite or zircon inclusions (DeWolf et al., 1996) are limiting issues. Many geochronologic studies of regional metamorphic rocks have resorted to the use of monazite as a geochronometer (Parrish, 1990; Smith and Barreiro, 1990), as Th and U are major or minor elements in monazite, monazite contains very low initial Pb concentrations, and monazite growth can often be correlated with metamorphic reactions or fabric-forming deformation events attending porphyroblast growth (Foster et al., 2002; Williams and Jercinovic, 2002; Pyle and Spear, 2003; Gibson et al., 2004). Zircon is generally not amenable to dating low- and medium-grade regional metamorphism as its growth is usually limited to upper amphibolite to granulite facies conditions (Vavra et al., 1999; Rubatto et al., 2001). In addition to suitability for isotopic analysis, the growth history of monazite and zircon as revealed by microimaging methods can aid in interpreting the context of the metamorphic reaction and deformation history. This type of dating requires analytical methods with high spatial resolution, e.g., electron probe and ion probe microanalysis. Higher age precision may be attained using isotope dilution-thermal ionization mass spectrometry (ID-TIMS), but determination of metamorphic ages across a terrane requires a large number of samples, which is inefficient by ID-TIMS.

This study is an unconventional approach to dating regional metamorphism and contemporaneous magmatism that takes advantage of an extensive set of ages of detrital monazite and zircon separated from alluvium of the French Broad River, which drains an ∼12,000 km2 area of the southern Appalachian Blue Ridge. The samples were collected primarily for a comparison of the utility of detrital monazite versus detrital zircon ages as provenance indicators (Hietpas et al., 2010). Here we examine the monazite and zircon age distributions from this data set in more detail. Our study also incorporates new zircon secondary ion mass spectrometry (SIMS) U-Pb ages and monazite electron microprobe Th-U-Pb chemical ages for alluvium and bedrock samples collected in the context of zircon cathodoluminescence (CL) zoning, zircon Th/U values, and monazite compositional zoning, to determine the timing of regional metamorphic and magmatic events in the Blue Ridge and western Inner Piedmont. Although the original petrofabric and microstructural context of detrital monazite and zircon are unknown, we can tie the provenance of each grain to a probable source lithology based on its sample locality and in some cases on chemical characteristics (i.e., Th/U in zircon). The large number of grains analyzed in the Hietpas et al. (2010) study (1100 zircon analyses; 150 monazite analyses) combined with the wide sampling area and new SIMS and electron microprobe ages increases the likelihood of identifying less extensive metamorphic and magmatic events, and provides a terrane-scale context for those events.

The broad outline of Blue Ridge tectonic history and distribution of regional metamorphic isograds have been known for decades (Hadley and Goldsmith, 1963; Carpenter, 1970; Butler, 1973; Rankin, 1975; Hatcher, 1978). However, the timing of metamorphic and magmatic events, based on high precision and high spatial resolution U-Pb geochronometers, has only recently been determined (Miller et al., 2000; Moecher et al., 2004a; Miller et al., 2006, 2010; Corrie and Kohn, 2007), and then only from scattered outcrops. There is also considerable ambiguity in the extent and intensity of each phase of metamorphism that has affected the southern Appalachians due to a series of overprinting Paleozoic collisional events.

The Blue Ridge in the southern Appalachians of western North Carolina is an amalgamation of Mesoproterozoic (Grenville age) basement rocks (mostly granitic orthogneisses), Neoproterozoic rift-related bimodal volcanic and plutonic rocks, late Neoproterozoic to Early Ordovician rift and passive margin, mostly clastic sedimentary cover sequences (Ocoee Supergroup, Ashe Metamorphic Suite, Chilhowee Group), and Ordovician plutonic rocks (Fig. 1A). These Blue Ridge components were first metamorphosed together in the Middle Ordovician. Regional metamorphism reached upper amphibolite facies with extensive development of migmatites in pelitic and mafic bulk compositions. Granulite facies assemblages in pelitic and mafic bulk compositions occur locally (Absher and McSween, 1985; Eckert et al., 1989), as well as the most extensive tract of eclogite facies rocks in the Appalachian orogen (Willard and Adams, 1994; Miller et al., 2010). Middle to late Paleozoic metamorphism is greenschist (in basement, eastern and western Blue Ridge cover sequences) to amphibolite facies (Blue Ridge thrust complex, western Inner Piedmont) and not penetratively developed, compared to early Paleozoic metamorphism and compared to the extent of Acadian metamorphism in the northern Appalachians. Plutons of early, middle, and late Paleozoic age in the Blue Ridge, emplaced into a compressional setting, were variably deformed by each phase of orogenesis.

Uncertainty regarding the setting for early and middle Paleozoic tectonism and plutonism is largely the result of late Paleozoic collision and formation of crystalline thrust sheets that transported the complex of Precambrian and Paleozoic units westward onto the Laurentian margin (Hatcher, 1987). Greenschist facies mylonite zones define thrust sheet boundaries and occur throughout Blue Ridge basement as 1–10-km-wide anastomosing zones of Mississippian age mylonites (Merschat et al., 2006). Late Paleozoic deformation also resulted in folding at all observable scales and retrograde metamorphism of early Paleozoic assemblages (Massey and Moecher, 2005; Clemons and Moecher, 2009).

The French Broad River headwater streams collect alluvium from both sides of the northeastern-striking Brevard fault zone (Fig. 1B). The Eastern Blue Ridge in the French Broad headwaters includes the Mesoproterozoic Toxaway Gneiss (Carrigan et al., 2003), Neoproterozoic metasediments of the Ashe–Tallulah Falls suite (ATFS, Hatcher, 2002; Bream et al., 2004), the Ordovician Whiteside pluton, and Devonian Looking Glass and Pink Beds plutons (Miller et al., 2000). The southeast side of the Brevard zone includes the western Inner Piedmont, dominantly mica schists, migmatitic gneisses, and amphibolites of the ATFS. From its headwaters near Rosman (North Carolina), the French Broad River flows northeast parallel to the Brevard fault zone, turns north through Asheville and across the migmatitic pelitic gneisses and amphibolites of the ATFS (Rankin et al., 1973; McSween et al., 1989; Hatcher, 2002), then turns northwestward across mostly Blue Ridge basement rocks to the Great Smoky fault (base of the crystalline thrust sheets) in western North Carolina and eastern Tennessee (Fig. 1B). The main tributary of the French Broad is the Pigeon River, which cuts across the eastern end of the Great Smoky Mountains and has local headwater tributaries in the Canton quadrangle (Fig. 1B). Metamorphism in the study area ranges from sillimanite grade in the southeast to chlorite grade in the northwest (Carpenter, 1970; Hadley and Nelson, 1971).

Mesoproterozoic (ca. 1350–900 Ma) basement rocks are dominantly upper amphibolite facies, migmatitic layered gneisses enclosing 1–10-km-scale masses of weakly foliated granulite facies orthogneisses. The granulite assemblages are demonstrably Mesoproterozoic (Carrigan et al., 2003; Ownby et al., 2004), whereas the migmatitic layered gneisses are probably the product of upper amphibolite facies early Paleozoic metamorphism of rocks with Mesoproterozoic protoliths (Merschat et al., 2006). The ATFS in the study area is dominantly migmatitic metapelitic gneisses with conspicuous leucosomes and a mineral assemblage of plagioclase + biotite + kyanite/sillimanite + garnet + quartz + rutile + muscovite (Fig. 2). Minor components of the ATFS include amphibolites and calc-silicates. Trondhjemite dikes and pegmatites of Silurian to Devonian age are volumetrically minor but widespread in the area west and north of Asheville, and could potentially contribute detrital zircon to modern alluvium. The kyanite-sillimanite isograd bisects the Canton quadrangle (Merschat and Wiener, 1988).

Known regional tectonometamorphic and magmatic events in the drainage basin of the French Broad and Pigeon Rivers include Mesoproterozoic magmatism and granulite facies metamorphism, Neoproterozoic continental rift magmatism (Goldberg et al., 1986; Su et al., 1994; Aleinikoff et al., 1995; Tollo et al., 2004) (and potential contact or hydrothermal metamorphism), and Paleozoic plutonism and regional metamorphism. The early Paleozoic Henderson orthogneiss is one of the largest Paleozoic plutons in the Southern Blue Ridge, and is bedrock along much of the upper course of the French Broad River (Fig. 1B).

Based on U-Pb zircon and monazite geochronology from localities elsewhere in the Blue Ridge of western North Carolina, peak Paleozoic amphibolite, granulite, and eclogite facies metamorphism is Middle Ordovician in age (Moecher et al., 2004a, 2005; Corrie and Kohn, 2007; Miller et al., 2010). Hornblende 40Ar-39Ar plateau ages in the study area range from 410 to 450 Ma (Dallmeyer, 1975; Kunk et al., 2006), but variably disturbed hornblende isotopic systematics are widespread (Goldberg and Dallmeyer, 1997). The Rb-Sr and Sm-Nd mineral-rock isochrons from samples in the vicinity of the Lick Ridge eclogite support the interpretation of Middle Ordovician regional metamorphism with variable Devonian resetting (Goldberg and Dallmeyer, 1997). Muscovite 40Ar-39Ar plateau and Rb-Sr mineral-rock isochron ages throughout the entire study area are generally Carboniferous (Dallmeyer, 1988; Connelly and Dallmeyer, 1993; Goldberg and Dallmeyer, 1997; Kunk et al., 2006).

For our earlier study (Hietpas et al., 2010) samples of coarse sand were collected from the French Broad River at 14 sites extending from its headwaters near Rosman, North Carolina, to Douglas Lake near Newport, Tennessee (Fig. 1B). An additional sample was collected from the Pigeon River near its confluence with the French Broad River. Heavy mineral concentrates in alluvium from tributary streams in the Canton quadrangle (Fig. 1B; Table 1) were originally obtained from the North Carolina Geological Survey (Merschat and Wiener, 1988) and made available for the Hietpas et al. (2010) study. The tributary stream samples were collected in the context of 1:24,000 scale bedrock mapping and must be derived from either ATFS schist and migmatitic paragneiss (Fig. 2, A-1–A-3) or Mesoproterozoic Sandymush and Earlies Gap gneisses (Merschat and Cattanach, 2008) (Table 2), as these are the only lithologic units observed in the quadrangle (Merschat and Wiener, 1988). The western half of the Canton quadrangle drains into downstream stretches of the French Broad River via the Pigeon River, and the eastern half drains directly into the French Broad River.

Additionally for this study, fresh samples of migmatitic Ashe paragneiss, Ashe pelitic schist, and Henderson orthogneiss were collected from roadcuts near tributary and alluvium sample sites. The rock samples were crushed and processed using heavy liquids and magnetic separation to concentrate zircon for ion microprobe dating.

Monazite and zircon in river and stream alluvium, and crushed rock samples, were handpicked from all recognizable populations based on size, crystal morphology, and color. Other random approaches for mounting grains for analysis (e.g., pouring grains from a vial) result in a bias for larger grains. Zircon from Mesoproterozoic orthogneisses is generally much larger and more abundant than zircon from Paleozoic Appalachian granitic plutons (Moecher and Samson, 2006). Therefore, any approach for mounting grains that might introduce a size bias would also bias the age distributions. Zircon or monazite crystals, along with age standards, were embedded in epoxy. Cured mounts were then ground to expose the interiors of the crystals. Each monazite crystal was imaged before ion probe analysis by backscattered electron imaging using the JEOL 8600 Superprobe at Syracuse University. Representative monazite grains analyzed by ion microprobe were also selected for electron microprobe wavelength dispersive X-ray elemental mapping to determine the nature and extent of growth zoning, which was used to guide monazite electron probe microanalysis. Every zircon analyzed in this study was imaged using cathodoluminescence to determine relative proportion of cores and rims as a guide for microanalysis, to assess magmatic versus metamorphic growth (Hanchar and Miller, 1993; Corfu et al., 2003), and to ensure that analytical spots did not overlap core-rim boundaries. In most cases the cores of monazite and zircon were analyzed, and rims of zircon were analyzed if of sufficient width to accommodate the laser beam.

Monazite 208Pb/232Th was measured using SIMS on the CAMECA 1270 at the University of California Los Angeles as described in Hietpas et al. (2010) (also see Harrison et al., 1995). The 208Pb-232Th system was used because Th is a major element in monazite, leading to higher signal intensities and shorter analysis times, an important factor when high throughputs are necessary for assembling a detrital age data set.

Some detrital zircon U-Pb ages were also determined by SIMS using the CAMECA 1270 instrument at the University of California Los Angeles (Grove et al., 2003). One important advantage of SIMS analysis is that only a few nanograms of zircon are consumed in the analysis and that no mineralogical or chemical change occurs in the crystal being analyzed. Thus the technique can be viewed as nondestructive. The majority of the detrital zircon crystals, however, were dated using laser ablation–multicollector–inductively coupled plasma mass spectrometry (LA-MC-ICP-MS) techniques on the Isoprobe-P at the LaserChron Center of the University of Arizona (Gehrels et al., 2008). As with SIMS analysis, LA-ICP-MS ages are dependent on comparison of ages of natural zircon standards. The main differences between techniques are that the LA-ICP-MS dating technique is much faster than SIMS (∼1 min compared to ∼15 min), but the former uses considerably more material (at least hundreds of nanograms) and physically alters the zircon by preferentially extracting uranium compared to Pb from the analyzed region (Patchett and Samson, 2004). Thus this technique must be viewed as at least partially destructive. The analytical data and calculated ages for accessory minerals from modern alluvium are available in Hietpas et al. (2010). Accessory mineral ages for potential source rocks are presented here.

Selected monazites from the tributary streams draining areas underlain by Ashe paragneisses that were analyzed by SIMS for 208Pb-232Th geochronology were also analyzed by electron probe microanalysis (EPMA) for total Th-U-Pb chemical age determinations (Montel et al., 1996; Pyle et al., 2005; Williams et al., 2007). This method measures the total Pb concentration in monazite and assumes that all Pb is radiogenic, produced by decay of 235U, 238U, and 232Th. Monazite typically contains higher concentrations of Th than U (usually 1–5 wt% ThO2, 0.1–1 wt% UO2; Montel et al., 1996), so that most of the radiogenic Pb results from Th decay. Total Pb concentrations range from 100 to 1000 ppm, which can be measured precisely via EPMA with high beam currents, long count times, and careful background measurements (Williams et al., 2006). The amount of nonradiogenic Pb in monazite is sufficiently low (1–3 ppm; Parrish, 1990; Hawkins and Bowring, 1997) as to not significantly affect the calculated chemical age of monazite. The greatest limitation of the chemical age method is that discordance of the U-Pb system cannot be assessed. The advantage of EPMA chemical age dating is that the electron microprobe has one to two orders of magnitude higher spatial resolution than the ion microprobe, allowing for analysis of multiple and smaller growth generations in a single monazite grain than with the ion microprobe.

Monazite EPMA for this study was carried out at the University of Massachusetts on the CAMECA Ultrachron microprobe following the analytical protocol of Williams et al. (2006) and Williams and Jercinovic (2002). Semiquantitative X-ray maps of monazite grains were used to determine growth domains based on the distribution of Th, U, and Y, followed by complete EPMA quantitative analysis of 6–8 points in each zone. Measured Th, U, and Pb contents (with Y correction) are used to calculate a date for each point (Table 3). Age precision for each point is based on X-ray counting statistics. Individual point analyses (5–10) are pooled to calculate a weighted mean age and error for a generation of monazite growth that is interpreted to be geologically meaningful (time of growth of that zone, representing a phase of mineral reaction; Foster et al., 2002; Spear and Pyle, 2002; Williams and Jercinovic, 2002; Wing et al., 2003; Gibson et al., 2004; Williams et al., 2007; Corrie and Kohn, 2008).


Monazite grains from the Canton quadrangle tributaries exhibit Th, U, and Y compositional zoning patterns consisting of sector zoned cores that compose the bulk of the grain with thin rims overgrowing embayed margins (Fig. 2B). The 208Pb-232Th ages are Ordovician (Fig. 3A; Table 2). All but three ages are within analytical error [generally ±5–10 m.y. (1σ) for monazite 232Th-208Pb ages] in the range 455–475 Ma. The age distribution is slightly skewed to younger ages and the distribution appears to consist of two age populations if plotted with 5 m.y. bins (Fig. 4A).

Monazite grains from French Broad River alluvium are dominated by Middle Ordovician ages, but yield a greater range of Ordovician ages than those of the Canton quadrangle tributary streams, and include scattered Mesoproterozoic ages, three Neoproterozoic ages, and numerous middle to late Paleozoic ages (Fig. 4). The Ordovician age mode consists of a major peak at 455–460 Ma and a spread of ages from 455 to 425 Ma (Fig. 4C). The French Broad samples also have minor populations of Devonian and Carboniferous ages, which are absent in the tributary alluvium. The majority of the younger ages are from samples FB1 to FB4, which were collected southeast of the Brevard fault zone, but could be derived from streams in the French Broad headwaters region that collect alluvium from either side of the fault zone (Fig. 1B). Sample FB15 (Fig. 1B; Table 2), collected in the floodplain of the Pigeon River, yields an age range similar to the French Broad River and tributary streams.

Electron microprobe total Th-U-Pb chemical ages of detrital monazite from Canton quadrangle tributaries are dominantly Middle Ordovician, with rare Cambrian, Silurian, and Devonian ages (Table 3). Three grains for which the same growth zone was analyzed by both EPMA and ion probe yield identical ages (467 ± 10 versus 463 ± 7; 461 ± 9 versus 464 ± 12; 457 ± 8 versus 457 ± 5 Ma; Fig. 2; Table 3). For other monazite grains, distinctive growth zones as defined by differences in Th or U concentration, may yield the same total Th-U-Pb chemical age, or they may yield analytically different ages, and rims may be older than cores, which is not uncommon in monazite due to resorption and embayment followed by overgrowth by new monazite (Spear and Pyle, 2002). The chemical age of 479 ± 5 Ma (sample CT136, Table 3) is similar to the oldest monazite ion probe ages for tributary samples (475–485 Ma). However, the youngest chemical age (395 ± 5 Ma) is for a narrow Th-rich rim zone that is too thin to analyze precisely by ion probe and younger than the youngest ion probe age for the tributaries (Fig. 2B). Although this Devonian age is not detected by ion probe, similar middle Paleozoic ages occur in upper French Broad River alluvium samples (see following).


Zircon from Canton tributary samples (Tables 4 and 6) has oscillatory zoned cores with Th/U values consistent with magmatic crystallization, and unzoned rims and Th/U consistent with metamorphic growth (Fig. 2C). Dominant age peaks occur at 1190, 1050, and 450 Ma, with minor peaks at 1290 and 1800 Ma (Fig. 5A). No ages younger than ca. 420 Ma were found. The Canton tributary samples are dominated by Mesoproterozoic zircon with age distributions that are identical to French Broad alluvium samples collected in areas draining Mesoproterozoic basement northwest of Asheville (e.g., FB6–FB10, Fig. 1B). Neoproterozoic ages are rare in Canton quadrangle stream samples. Values of Th/U for zircon with Ordovician ages are 0.01–0.06, which contrasts with Th/U values in Mesoproterozoic zircon from the tributaries of 0.11–2.25 (Fig. 5D). Neoproterozoic zircon, although rarer, has Th/U of 0.06–0.41. Rock samples of ATFS migmatitic gneiss and schist exhibit an age distribution similar to those of tributary streams (Table 5; Fig. 5E), with fewer Ordovician zircons and rare Neoproterozoic ages. The former have very low Th/U (0.01–0.03) (Fig. 5F; Table 5).

The dominant zircon age population in French Broad alluvium samples is Middle Ordovician (ca. 450 Ma) (Fig. 6A); secondary populations occur at 1150–1190 and 1000–1050 Ma, similar to tributary zircon age distributions. Neoproterozoic ages are much less abundant than Mesoproterozoic ages, but span the entire eon (Fig. 7A). The Ordovician zircons exhibit a range of Th/U (Fig. 8B), indicating that a mix of metamorphic and magmatic zircon defines that population. The dominant age population in French Broad alluvium is the same as the age of zircon from rock samples of the Henderson orthogneiss, which has relatively high Th/U (0.45–1.14; Table 5) and yields a concordia age of 447.6 ± 5.4 Ma (Fig. 9; Table 5). Zircon in Ashe paragneiss and schist is dominantly Mesoproterozoic, with scattered Neoproterozoic ages, and a distinct Ordovician peak (Fig. 5E; Table 5). The Ordovician zircon has distinctly lower Th/U values than the Mesoproterozoic and Neoproterozoic zircon in the Ashe bedrock samples (Fig. 5F), which are similar to values for zircon in tributary samples (Fig. 5D).

Age Distributions: Canton Tributaries versus French Broad River

The restricted sampling area of the Canton tributary streams (7.5 minute quadrangle [Merschat and Wiener, 1988]) compared to the wide region sampled by the French Broad River is reflected in both monazite and zircon age distributions. Based on abundance of monazite in Canton tributary and Ashe bedrock samples, it appears that Neoproterozoic metasedimentary rocks are the predominant source of monazite in this region of the Blue Ridge province, and the majority of the monazite is metamorphic. Either Ordovician monazite did not form in granitic Mesoproterozoic basement orthogneisses, or those rocks did not undergo Taconian metamorphism; the latter is unlikely, because basement lithologies are intimately interfolded with, and similarly migmatitic as, ATFS rocks that are at sillimanite grade. Although igneous monazite is difficult to distinguish from metamorphic monazite based solely on available ages or composition, given its abundance in ATFS rocks we argue that the monazite age mode in Figure 3 corresponds to the time of peak regional metamorphism.

Age Distributions: Monazite versus Zircon

The minor amount of Precambrian monazite is striking compared to the abundance of Paleozoic monazite (Fig. 3B). The Canton tributary samples illustrate the contrasting responsiveness of monazite in pelitic lithologies versus that in granitic systems during high-grade metamorphism. Apparently, there were no significant monazite-fertile pelitic lithologies among those composing the Mesoproterozoic basement complex, which is consistent with bedrock mapping in the region (Merschat and Cattanach, 2008).

In addition to compositional zoning patterns revealed by CL imaging (oscillatory zoning versus sector zoning versus no zoning), Th/U has been purported to distinguish magmatic zircon from metamorphic zircon. The Th/U value is usually higher in igneous zircon than metamorphic zircon. However, there is not a consensus as to the minimum value considered diagnostic of igneous zircon. Hoskin and Schaltegger (2003) stated that Th/U ≥ 0.5 is typical for igneous zircon, Vavra et al. (1999) stated that Th/U = 0.3 is typical of magmatic zircon, and Rubatto (2002) stated that magmatic zircon may have values as low as 0.18. For example, Mesoproterozoic magmatic zircon in igneous rocks from the Adirondack terrane, a classic locality for Grenville magmatism, has Th/U ranging from 0.1 to 1.5 (Fig. 6C). At upper amphibolite and granulite facies conditions under which anatexis is common, zircon in leucosomes is essentially magmatic (e.g., Moecher et al., 2004a), with Th/U values and zoning patterns indistinguishable from magmatic zircon.

The kyanite-sillimanite isograd is mapped in the Canton quadrangle, and much of the ATFS in the French Broad drainage is migmatitic (Fig. 2A), so some zircon growth is possible. Despite the high grade of metamorphism and onset of melt-forming reactions, tributary alluvium derived from weathering of Ashe lithologies, and direct samples of Ashe metasedimentary rocks, yield only a few Ordovician zircon ages, most of which are for zircon rims that have Th/U = 0.01–0.05 (Fig. 5F). Ashe bedrock and alluvium zircon remain dominated by Mesoproterozoic zircon ages in spite of the presence of widespread high-grade Middle Ordovician metamorphism and the abundance of Ordovician monazite ages.

Although the ranges of monazite and metamorphic zircon ages overlap for both tributary and French Broad River data sets, the peak of the dominant age populations for Ordovician metamorphic zircon in both the tributary and river alluvium samples is younger than the peak age of monazite (Figs. 4A, 5B, and 8A). Zircon and monazite need not date the same metamorphic reaction during a prograde event, and both ages inform our interpretation of the metamorphic history. Monazite growth has been shown to typically begin at lower amphibolite facies (staurolite grade) (Smith and Barreiro, 1990; Kingsbury et al., 1993; Foster et al., 2002; Wing et al., 2003; Gibson et al., 2004; Corrie and Kohn, 2008), whereas zircon growth typically coincides with the onset of partial melting (sillimanite grade and higher in metapelites) (Vavra et al., 1999; Rubatto et al., 2001; Moecher and Samson, 2006). For a high-grade terrane such as the southeastern Blue Ridge, monazite growth would begin early on the prograde P-T-t (pressure-temperature-time) path, but zircon growth would not begin until near the thermal peak of metamorphism, that is, later in the P-T-t path. A phase of monazite growth might occur as each successive isograd is reached (i.e., when prograde mineral reactions occur). Because lead diffusion in monazite is sufficiently slow (Cherniak et al., 2004), each phase of monazite growth would be preserved and is dateable if sufficient volume of monazite were generated by each event and resolvable by X-ray mapping techniques. Metamorphic zircon ages should be younger than most of the monazite ages in the same rock, with the youngest generation of monazite overlapping zircon growth. That is the pattern observed in the data set presented here, and emphasized by the range of monazite ages obtained by high spatial resolution electron microprobe chemical age dating. The most common ion probe and monazite chemical ages (460–465 Ma) are both slightly older than the dominant metamorphic zircon age population in our alluvial samples (ca. 455 Ma) and slightly older than the most precise ages for peak regional metamorphism in the Eastern Blue Ridge [concordant U-Pb zircon ID-TIMS age 458 ± 1 Ma (2σ) from granulite facies migmatites; Moecher et al., 2004a]. We propose that the 460–465 Ma monazite ages correspond to the time of the bulk of monazite growth that resulted from release of light rare earth elements from prograde allanite breakdown. The latter begins at lower amphibolite facies conditions, but before the metamorphic thermal peak (Smith and Barreiro, 1990; Wing et al., 2003; Corrie and Kohn, 2008).

Monazite Ion Microprobe versus Electron Microprobe Ages

Williams et al. (1999) demonstrated that monazite electron microprobe Th-U-Pb chemical ages are in many cases identical within uncertainty to conventional Th-U-Pb isotopic ages of monazite for the same rock, indicating that all of the assumptions of the EPMA technique are valid. The same outcome was obtained here when the same spots or growth zones were analyzed on the same crystals by both methods (Fig. 2B; Table 3). Based on ages discernable only by the order of magnitude higher spatial resolution of the electron microprobe, it is apparent that the monazite growth history, and by extension the metamorphic history of the Blue Ridge, may be more complex than the history resolvable by monazite ion probe methods or zircon U-Pb geochronology. In addition to the dominant Middle Ordovician event, the electron microprobe revealed Middle to Late Cambrian, Early and Late Ordovician, Early Silurian, and Early Devonian monazite crystallization events. The ages correspond to known Appalachian tectonic phases and/or have age analogs among the zircon ages for the tributary samples and among the French Broad monazites.

The Th-U-Pb chemical ages for detrital monazite derived from ATFS paragneisses are the same as chemical, ion microprobe, and TIMS ages in Great Smoky Group schist and migmatitic gneisses to the southwest, which also probably formed from Neoproterozoic clastic rocks derived from Grenville sources (Zo map unit, Fig. 1B). Monazite chemical ages from the Great Smoky Group schist and gneiss define Early, Middle, and Late Ordovician ages (Moecher et al., 2004b, 2005), similar to EPMA and ion microprobe ages for monazite from the tributary stream samples. Some monazite grains contain late Neoproterozoic to early Paleozoic cores, and there are scattered younger ages within the distribution, similar to the pattern of French Broad monazite. Corrie and Kohn (2007) interpreted high mean square of weighted deviate values for U-Pb ID-TIMS analysis of monazite extracted from Great Smoky Group schists by microdrilling, spanning ages from 440 to 470 Ma, as not representing a single population of monazite growth. Corrie and Kohn (2007) established 450 ± 5 Ma as most closely representing the time of the majority of monazite growth, and by extension, the thermal peak of regional metamorphism in the Great Smoky Mountains region. These relations constitute preliminary evidence that Taconian regional metamorphism in the Appalachian Blue Ridge consisted of more than one tectonometamorphic pulse.

Henderson Orthogneiss

The age of the Henderson granite (Table 5), now dominantly an orthogneiss that is locally protomylonitic to ultramylonitic in the vicinity of the Brevard fault zone, is a long-standing problem in Blue Ridge tectonics (Horton and Butler, 1986). The Henderson is one of the largest exposed plutons in the southern Appalachians and probably was emplaced during a major period of granitic magmatism. U-Pb zircon and Rb-Sr isochron dates ranging from 600 to 350 Ma have been reported (e.g., Odom and Fullagar, 1973; Sinha and Glover, 1978), some of which are geologically meaningful. Carrigan et al. (2001) reported a SIMS microprobe U-Pb zircon age of ca. 490 Ma, but Vinson et al. (1999) reported ages for some zircon from the Henderson that are as young as 450 Ma. The concordia age determined here for a sample of Henderson gneiss is 447.6 ± 5.4 Ma, making the Henderson essentially a synmetamorphic Taconian pluton. If the Henderson were significantly older, or consisted of multiple plutons intruded over a span of time, older ages or a wider range of ages should be a prominent component of the alluvium zircon age distributions, as the French Broad headwaters flow across a region underlain almost exclusively by Henderson gneiss (Fig. 1B). The majority of Ordovician zircon from the alluvium samples is ca. 450 Ma with Th/U consistent with magmatic values (>>0.1; data in Hietpas et al., 2010) and with values measured for zircon from the rock sample of Henderson gneiss yielding the ca. 450 Ma age (Table 5).

Neoproterozoic Ages

Neoproterozoic zircon ages are rare but present in all zircon data sets (Figs. 5A, 6A, and 7A). Ages cluster in the ranges 600–550 and 800–700 Ma. As little new zircon was generated in the Ashe paragneisses, and most new zircon occurs as Ordovician rims on detrital Mesoproterozoic zircon, these zircons are most likely detrital and derived from Neoproterozoic rocks. The Th/U values for zircon younger than 800 Ma imply that both magmatic and metamorphic zircon is present (Fig. 7B).

The primary late Neoproterozoic tectonic event in the Blue Ridge was continental rifting, which is interpreted in the central and southern Appalachians to have involved two phases of rift-related magmatism (Crossnore plutonic-volcanic complex, Robertson River granites, Bakersville mafic intrusive suites; Catoctin and Mount Rogers volcanics) at 765–680 Ma and 620–550 Ma (Odom and Fullagar, 1984; Goldberg et al., 1986; Su et al., 1994; Aleinikoff et al., 1995; Tollo et al., 2004). These ages are well represented in French Broad River alluvium samples (Fig. 7A), particularly sample FB6 (data in Hietpas et al., 2010) collected northwest of Asheville in an area containing outcrops of Bakersville dikes (Rb-Sr isochron age of 734 ± 26 Ma; Goldberg et al., 1986). Although Crossnore type granites (U-Pb zircon crystallization ages of 745 ± 5 Ma: Su et al., 1994) do not crop out in the French Broad drainage basin, these rocks were apparently exposed and generating detrital zircon during deposition of ATFS sedimentary protoliths, as Neoproterozoic zircon ages also appear in tributary stream samples and ATFS gneisses.

Depositional Age of Ashe Metamorphic Suite Protoliths

The age distributions of Canton quadrangle tributary zircon and zircon from the sample of ATFS gneiss are very similar to detrital zircon age distributions of other extensive metaclastic rocks in the Western, Central, and Eastern Blue Ridge (Tallulah Falls Formation, Ocoee Supergroup, Coleman River Formation, Otto Formation, Chauga Belt rocks; Bream et al., 2004; Chakraborty et al., 2010). Based on regional bedrock mapping, Tallulah Falls Formation metasedimentary rocks are considered to be broadly time equivalent to Ashe Metamorphic Suite metasediments (Hatcher, 2002; Bream et al., 2004). The inferred equivalency is further supported by the new zircon data set presented here. The maximum depositional age of the Ashe suite protoliths is the age of the youngest detrital zircon, which must be as young as Cambrian, as indicated by the youngest zircon ages with magmatic Th/U values in the samples of ATFS gneiss (Figs. 5E, 5F; Table 5) and tributaries draining areas of ATFS bedrock (Fig. 5A). The maximum age of the Neoproterozoic Ocoee Supergroup, another Ashe suite correlative, based on the age of the Bakersville dikes intruding basement, and the absence of crosscutting relations with overlying Ocoee sediments, must be less than the age of the Bakersville dikes (ca. 735 Ma; Goldberg et al., 1986). This inference is supported by detrital zircon U-Pb ages for the lower Ocoee Supergroup (Wading Branch Formation) of 640 ± 10 Ma (Chakraborty et al., 2010) and diagenetic and/or metamorphic monazite and xenotime U-Pb ages of ca. 570–580 Ma for the Thunderhead Formation of the upper Ocoee Supergroup (Aleinikoff et al., 2010).

The monazite and zircon geochronology presented here corroborates the emerging pattern of metamorphic crystallization and cooling ages in the Southern Blue Ridge. Although Paleozoic orogenesis was a temporally protracted process in the Western and Eastern Blue Ridge in North Carolina, peak Barrovian regional metamorphism northwest of the Brevard zone, as bracketed by ages of metamorphic monazite (growth on the prograde path or at the thermal peak) and ages of zircon rims on Mesoproterozoic zircon cores in metasedimentary rocks (growth at the thermal peak), is definitively Middle Ordovician (Taconian). In contrast, regional metamorphism in the western Inner Piedmont is distinctly younger (Late Devonian to Mississippian; Dennis, 2007); evidence for this includes monazite ID-TIMS U-Pb geochronology (Dennis and Wright, 1997), and SIMS U-Pb ages of zircon rims (Carrigan et al., 2001) in orthogneisses and metasediments of the western Inner Piedmont that are correlated with the Tallulah Falls Formation of the Eastern Blue Ridge (Hatcher, 2002). Middle Paleozoic zircon rim ages are also found in easternmost Blue Ridge rocks near the Brevard fault zone (Carrigan et al., 2001). Albeit much less abundant than Ordovician monazite, most of the late Paleozoic monazite ages in our data set are for monazite grains in alluvium samples collected on the southeast side of the Brevard fault zone. Although the age of the metamorphic peak is different in the Eastern Blue Ridge and western Inner Piedmont, there is abundant evidence for middle and late Paleozoic retrograde metamorphism in the Eastern Blue Ridge (Goldberg and Dallmeyer, 1997; Moecher et al., 2007; Miller et al., 2010), but as yet no evidence for Ordovician metamorphism in the western Inner Piedmont (Dennis, 2007). The only non-Ordovician monazite age in the tributary data set (definitive Eastern Blue Ridge affinity) is from a thin monazite rim detected by high spatial resolution EPMA (Fig. 2B). The pattern and distribution of ages suggest variable and nonpenetrative overprinting of Taconian fabrics and mineral assemblages by younger deformation that is progressively more pervasive to the southeast. Resolution of the manner in which middle to late Paleozoic tectonometamorphism characteristic of the western Inner Piedmont overprinted early Paleozoic effects in the Eastern Blue Ridge would be facilitated by an extensive program of ion probe isotopic analysis and microprobe chemical age dating of rocks across the Brevard fault zone, in the context of the deformation and retrograde metamorphic history of those rocks.

Published 40Ar-39Ar hornblende and muscovite age distributions provide further evidence for this contrasting thermotectonic history for the Eastern Blue Ridge and western Inner Piedmont. Many hornblende 40Ar-39Ar age spectra for Eastern Blue Ridge rocks in western North Carolina fail to define plateaus, suggesting isotopic disturbance (Goldberg and Dallmeyer, 1997), but the few samples that define plateaus yield Ordovician to Silurian ages (Dallmeyer, 1975; Kunk et al., 2006). In contrast, hornblende cooling ages in the westernmost Inner Piedmont (including the Brevard zone) are latest Devonian to Carboniferous (Dallmeyer, 1988), decreasing to Permian in the eastern Inner Piedmont (Dallmeyer et al., 1986). The late Paleozoic hornblende ages are derived from Ar release spectra that generally define plateaus, in contrast to the large degree of disturbance in Eastern Blue Ridge rocks along strike to the northeast. Exceptions to the pattern of disturbed Eastern Blue Ridge hornblende age spectra are the plateau ages of 354 and 362 Ma for the Lake Chatuge mafic-ultramafic complex (LCC, Fig. 1A) in the Eastern Blue Ridge of southernmost North Carolina and northernmost Georgia (Fig. 1A) (Dallmeyer, 1989). Muscovite plateau 40Ar-39Ar ages for schists, gneisses, and retrograde mylonites in the Eastern Blue Ridge tend to have an older upper age limit (370–320 Ma; Connelly and Dallmeyer, 1993; Goldberg and Dallmeyer, 1997; Kunk et al., 2006) than western Inner Piedmont muscovites (Dallmeyer, 1988) (Fig. 10).

Although inferred to not be a suture (Hatcher, 2002), based primarily on the similarity of detrital zircon age distributions in metasedimentary rocks from its contiguous belts (Bream et al., 2004), the Brevard fault zone is clearly a significant tectonic boundary separating rocks of different metamorphic age, grade, structural style, and igneous composition (Miller et al., 2000; Mapes et al., 2001; Merschat et al., 2005; Dennis, 2007). The ductile, peak metamorphic (sillimanite grade), and largely strike-slip component of motion on the Brevard fault zone (Reed and Bryant, 1964; Vauchez, 1987) cannot explain the contrast, requiring a component of significant Pennsylvanian vertical displacement, as indicated by the difference in muscovite cooling ages between the western Blue Ridge and western Inner Piedmont (Fig. 10). The Rosman fault, a late southeast-dipping thrust fault within the Brevard fault zone (Horton and Butler, 1986), may accommodate some of the requisite displacement.

Peak Barrovian regional metamorphism in the Western and Eastern Blue Ridge of the southern Appalachian orogen, as determined by high spatial resolution analysis methods on monazite and zircon in alluvium and bedrock samples, is Middle Ordovician. Details of age distributions and the highest spatial resolution age measurements reveal that tectonometamorphism occurred in three phases that span the Ordovician Period. There is thus a richer history of events that requires deconvolution by microgeochronology in order to construct a more complete and accurate tectonic model for the Taconian.

Lithologically, the Southern Blue Ridge is a relatively magma poor region, dominated by extensive latest Neoproterozoic to Cambrian metaclastic sequences. The fingerprint of the orogen, from a detrital mineral provenance perspective, is dominated by Mesoproterozoic zircon derived from basement that exhibits a very weak response to Paleozoic metamorphism. Monazite in selected lithologies is much more responsive to regional tectonometamorphic events. As both minerals have strengths and limitations as a provenance indicator, a more informative approach to provenance analysis would involve routine analysis of both detrital monazite and detrital zircon. The availability of high throughput microanalysis systems now makes such an approach highly practicable.

We thank Axel Schmitt for supervising the secondary ion mass spectrometry analytical sessions at the University of California, Los Angeles, and for guidance on interpretation of analyses, Victor Valencia for supervising analytical sessions at the University of Arizona LaserChron center, and Mike Jercinovic (University of Massachusetts) for supervising monazite electron probe microanalysis. We also thank Doug Rankin, Calvin Miller, and Brent Miller for helpful critical reviews of the manuscript. This research was supported by National Sciences Foundation grants EAR-0635688, EAR-0001322, EAR-0635643, and EAR-0447154.