The Picuris-Pecos fault is a major strike-slip fault in northern New Mexico (USA) that exhibits ∼37 km of dextral separation of Proterozoic lithotypes and structures. The timing of dextral slip has been controversial due largely to a lack of definitive piercing points of Phanerozoic age. The Picuris-Pecos fault formed the western boundary of the late Paleozoic Taos trough. A distinctive metasedimentary terrane that shed detritus into the western Taos trough was exposed on the Uncompahgre uplift west of the fault during the early to middle Pennsylvanian. We use the distribution of metasedimentary clasts and the age of monazite grains within clasts from conglomeratic strata of the western Taos trough to determine the paleolocation of the southern boundary of this metasedimentary terrane during the middle Pennsylvanian (Desmoinesian), and thereby quantify the subsequent separation on the fault. The rematching of detrital petrofacies with source terranes in the adjacent uplift requires ∼40–50 km of dextral separation on the Picuris-Pecos fault since the early Desmoinesian. This exceeds the present ∼37 km dextral separation of Proterozoic features by the fault, and thus implies that an ∼3–13 km sinistral separation existed on the fault in the early Desmoinesian. The ∼40–50 km of post–early Desmoinesian dextral separation on the Picuris-Pecos fault is the result of slip that accumulated late in the Ancestral Rocky Mountain deformation and/or during the Laramide orogeny.
The Proterozoic basement of the Southern Rocky Mountains in northern New Mexico, USA, has long been known to be dextrally faulted (Montgomery, 1963). Basement-related aeromagnetic patterns have been interpreted to show net dextral offsets of ∼55–130 km on several north-striking faults (Chapin, 1983; Cordell and Keller, 1984; Karlstrom and Daniel, 1993; Cather et al., 2006). The Picuris-Pecos fault is the largest of these faults, with a dextral separation of ∼37 km (Montgomery, 1963). Although there is agreement that major dextral separations of Proterozoic rocks and structures exist in northern New Mexico, the timing of dextral slip is controversial. This controversy derives in part from the lack of definitive piercing points in Phanerozoic rocks of the region. As a result, dextral slip has variously been inferred to have occurred primarily during the Proterozoic (Montgomery, 1963; Yin and Ingersoll, 1997; Fankhauser and Erslev, 2004; Wawrzyniec et al., 2007), mostly during the late Paleozoic Ancestral Rocky Mountain orogeny (Baars and Stevenson, 1984; Woodward et al., 1999), mostly during the Late Cretaceous–Eocene Laramide orogeny (Chapin and Cather, 1981; Chapin, 1983; Karlstrom and Daniel, 1993; Daniel et al., 1995; Bauer and Ralser, 1995; Cather, 1999), or during both the Ancestral Rocky Mountain and Laramide orogenies (Cather, 2004; Cather et al., 2006). The possible regional kinematic role of the Picuris-Pecos fault during Proterozoic deformations is unclear; regional strain analysis suggests lateral slip on the fault during known Proterozoic deformations was probably sinistral (Cather et al., 2006). During the Laramide and Ancestral Rocky Mountain deformations, dextral slip on the Picuris-Pecos fault potentially contributed to crustal shortening in uplifts and basins northwest of the fault (Fig. 1).
The Picuris-Pecos fault formed the western boundary of the late Paleozoic Taos trough during the early part of the Ancestral Rocky Mountain orogeny. During the Pennsylvanian, terranes of distinctive Proterozoic lithotypes (metasedimentary versus dominantly plutonic and metaplutonic) were exposed in the upthrown western block (the Uncompahgre uplift) of the Picuris-Pecos fault. The paleolocations of these terranes can be ascertained using the provenance characteristics of Pennsylvanian strata in the adjoining Taos trough, an approach that was pioneered by Sutherland (1963). We show that proximal Pennsylvanian (Morrowan? to early Desmoinesian) petrofacies are today mismatched from their distinctive source terranes west of the fault. We utilize clast composition and monazite geochronology as provenance indicators to show that ∼40–50 km of dextral slip has occurred on the Picuris-Pecos fault since the early Pennsylvanian.
The Picuris-Pecos fault is the best exposed and most studied of the dextral faults in northern New Mexico (Fig. 2). The fault strikes north, dips steeply west, and cuts rocks ranging in age from Proterozoic to Paleogene. It is exposed for ∼80 km along strike, and may extend the length of the state (∼600 km) if probable fault linkages to the north and south are considered (Cather and Harrison, 2002; Cather, 2009). With ∼37 km dextral separation of Proterozoic lithotypes and east-west–trending ductile structures, the Picuris-Pecos fault exhibits the largest known separation of any fault in the central and Southern Rocky Mountains. It exceeds that of the next largest fault (the Wind River thrust) by a factor of nearly two.
The Picuris-Pecos fault has been repeatedly reactivated. During the Proterozoic, plutonism, ductile deformation, and peak metamorphism in northern New Mexico occurred ca. 1.4 Ga (Williams et al., 1999a). The lack of mylonites along the Picuris-Pecos fault, however, suggests that no ductile precursor to the fault was active at the time (Cather et al., 2006). Subsequent brittle slip along the Picuris-Pecos fault, however, may have occurred during the Grenville orogeny and during Neoproterozoic deformation related to the breakup of Rodinia (Cather et al., 2006). The earliest undisputable evidence for slip on the fault is early in the Ancestral Rocky Mountain orogeny, as shown by Mississippian marine carbonate mud that fills fissures and the interstices of breccias in the damage zone of the fault (Erslev et al., 2004; Fankhauser, 2005; Cather et al., 2008). Slip on the Picuris-Pecos fault during both the Ancestral Rocky Mountain and Laramide orogenies had a west-up component (Sutherland, 1963), but regional strain balancing suggests that it also may have hosted significant dextral slip during these deformations (Cather et al., 2006).
Several lines of evidence indicate that the Picuris-Pecos fault did not accommodate major dip slip. Metasedimentary rocks exposed in the Picuris and Truchas Mountains are separated by the Picuris-Pecos fault, but show evidence for similar peak metamorphic conditions near the Al-silicate triple point (3.5–4.0 kbar, 500–550 °C; Grambling, 1979, 1981; Daniel et al., 1995). Thermochronologic data from the southern Sangre de Cristo Mountains indicate that rocks at similar modern elevations on both sides of the fault shared similar ca. 1.2–0.4 Ga cooling histories and passed into the purely brittle regime (<300–250 °C) during or soon after the Grenville orogeny (ca. 1.2–0.9 Ga; Sanders et al., 2006). Modest west-up components of slip on the Picuris-Pecos fault defined the western margin of the Ancestral Rocky Mountain Taos trough (Casey, 1980) during the early Pennsylvanian, but topographic relief on this basin margin was low enough that it was lapped over and buried by marine strata during the middle Pennsylvanian (early to middle Desmoinesian; see following). West-up components of slip also occurred on the Picuris-Pecos fault during Laramide deformation, but similar late Laramide apatite fission-track cooling ages occur at similar elevations on both sides of the fault (Kelley and Chapin, 1995, their fig. 8) and suggest that Laramide differential uplift across the fault was not large. The observed ∼37 km of dextral separation on the Picuris-Pecos fault must therefore be largely the result of strike slip.
The study area encompasses the central part of the Picuris-Pecos fault adjacent to the Truchas uplift (Fig. 3). Outcrops north of the Truchas uplift are near roads, but areas to the south are within the Pecos Wilderness and are accessible only by foot or on horseback. Within the study area no major faults intersect the Picuris-Pecos fault, resulting in a relatively simple structural geometry. The Picuris-Pecos fault transects and offsets two distinctive terranes of Proterozoic rocks. The southern terrane consists of ca. 1.72–1.44 Ga metaplutonic, plutonic, and subordinate metavolcanic rocks, granitic gneiss and granite being the volumetrically dominant lithotypes (Karlstrom et al., 2004). The northern terrane is dominated by metasedimentary rocks (Hondo Group, ca. 1.69 Ga) and subordinate metavolcanic rocks (Vadito Group, ca. 1.70 Ga). Metasedimentary rocks of the Hondo Group include the Ortega Quartzite and schist, phyllite, and quartzite of the Rinconada, Pilar, and Piedra Lumbre Formations.
The Proterozoic terranes are separated by a major south-dipping ductile thrust fault. East of the Picuris-Pecos fault this fault is termed the Pecos thrust; to the west it is termed the Plomo fault (Fig. 3). The ∼37 km dextral separation of ductile structures and Proterozoic lithotypes by the Picuris-Pecos fault is the basis for the interpretation by Montgomery (1963) and subsequent workers that the Picuris-Pecos fault is a major strike-slip fault. The originally continuous Pecos and Plomo faults define the steep southern limb of a regional synclinorium that formed ca. 1.65–1.4 Ga (Williams et al., 1999a). This synclinorium, marked by the Hondo syncline on the west side of the Picuris-Pecos fault and the Brazos Cabin syncline on the east, contains mostly metasedimentary rocks. The low aeromagnetic value of the metasedimentary rocks has been used to evaluate strike-slip offsets in the subsurface of northern New Mexico (e.g., Karlstrom and Daniel, 1993; Cather et al., 2006).
The oldest Paleozoic strata in the study area are the Mississippian Arroyo Peñasco Group, which consists of the Espiritu Santo Formation (late Osagean) and the Terrero Formation (Meramecian and early Chesterian) (Armstrong et al., 2004). Mississippian strata unconformably overlie Proterozoic rocks and exhibit a range of thickness (0–45 m) that reflects the underlying, low-relief erosional paleotopography (Sutherland, 1963; Baltz and Myers, 1999). Mississippian rocks in the study area accumulated before and during the earliest stages of the Ancestral Rocky Mountain orogeny, but prior to development of significant orogenic topography (Armstrong, 1967; Baltz and Myers, 1999).
Pennsylvanian strata in the study area were deposited near the paleoequator in the western part of the Taos trough, an Ancestral Rocky Mountain basin that is bounded on the west by the Picuris-Pecos fault (Fig. 4). We divide Pennsylvanian beds using the nomenclature advocated by Sutherland (1963), Sutherland and Harlow (1973), and Kues (2001), who apportioned these strata among four formations (Fig. 5). Early to middle Pennsylvanian beds (Morrowan to middle Desmoinesian) disconformably overlie Mississippian and Proterozoic rocks and consist of quartzose sandstone and conglomerate, mudstone, and limestone. Sutherland (1963) divided these beds into two formations based on their content of sandstone and limestone. The northern sandstone-rich lithofacies is termed the Flechado Formation, and the southern carbonate-dominated beds are the La Pasada Formation. These units are conformably overlain by the middle to late Pennsylvanian (middle Desmoinesian to Virgilian) Alamitos Formation, which consists of arkosic sandstone and conglomerate, marine limestone, and mudstone. The Alamitos Formation, in turn, grades upward into arkosic sandstone, conglomerate, and mudstone of the nonmarine Sangre de Cristo Formation (late Pennsylvanian–Early Permian). The contact between these formations is defined at the top of the stratigraphically highest marine limestone.
The quartzose sandstone that is characteristic of the Flechado and La Pasada Formations was interpreted by Sutherland (1963) and most subsequent workers to have been derived from the quartz-rich metasedimentary rocks west of the Picuris-Pecos fault. This conclusion is supported locally by the composition of associated conglomerate, which is almost entirely composed of clasts of Ortega Quartzite in the western Taos trough. In the central Taos trough, however, conglomerate is rare and quartzose sandstone compositions may have resulted from feldspar destruction in initially arkosic sands by tropical weathering or intrastratal dissolution (e.g., Chandler, 1988; Soegaard, 1990). In the eastern Taos trough early to middle Pennsylvanian sandstones are commonly feldspathic (Baltz and Myers, 1999). In view of the possible modification of initial sandstone compositions, herein we limit our provenance analysis to proximal conglomeratic successions near the Picuris-Pecos fault where pebble- to boulder-size clasts are assumed to reflect the lithotypes of the source region.
Rio Chiquito Area
Near the Rio Chiquito, the Picuris-Pecos fault dips steeply west and consists of several anastomosing fault strands that juxtapose basement rocks of the ca. 1.69 Ga Ortega Quartzite on the east with Proterozoic granite-gneiss on the west (Fig. 6). Basement rocks on both sides of the Picuris-Pecos fault are overlain by Pennsylvanian strata; Mississippian beds are absent in this area. East of the Picuris-Pecos fault, the strike of Pennsylvanian beds is rotated clockwise ∼70°–90° near the fault, consistent with a dextral sense of shear on the fault.
In this section we employ stratigraphic analysis, paleontology, paleocurrent measurement, and petrology to establish the following. (1) The Pennsylvanian succession east of the Picuris-Pecos fault near the Rio Chiquito contains proximal, scarp-related braid-delta deposits that fine to the east and were derived from metasedimentary rocks from nearby to the west of the fault. (2) The closest metasedimentary source rocks for these deposits are in the Picuris Mountains, now dextrally separated from the Rio Chiquito area by at least 20 km. (3) The Pennsylvanian succession west of the Picuris-Pecos fault at the Rio Chiquito, although in part the same age as beds east of the fault, is much finer grained, suggesting post–early Desmoinesian juxtaposition by strike-slip faulting. (4) Dating of monazite grains in quartzite clasts from the Rio Chiquito area suggests that these clasts are dextrally separated from their likely source terranes in the Ortega Quartzite by ∼40–50 km.
Rio Chiquito Area East of the Picuris-Pecos Fault
East of the Picuris-Pecos fault at Rio Chiquito, Pennsylvanian strata are well exposed and unconformably overlie the Proterozoic Ortega Quartzite. We measured and sampled a 479-m-thick section of the uppermost Ortega Quartzite, Flechado Formation, and lower Alamitos Formation (Fig. 7). The Proterozoic Ortega Quartzite is a gray to white, amphibolite facies quartzarenite. It commonly exhibits distinctive bedding and cross-bedding defined by heavy mineral laminae. The Ortega Quartzite is nearly pure quartz (∼98%; Montgomery, 1963), with rare muscovite, schistose rock fragments (quartz, mica, feldspar), hematite, kyanite, sillimanite, andalusite, tourmaline, and zircon. We also identified and dated monazite grains in the Ortega Quartzite and in quartzite clasts in the Pennsylvanian section (see the following).
The lower 230 m of the Phanerozoic section east of the Picuris-Pecos fault at Rio Chiquito were interpreted by Sutherland (1963) to be a conglomeratic facies of the Del Padre Member of the Mississippian Espiritu Santo Formation. Armstrong (1967, p. 9), however, showed that these strata are Pennsylvanian, based in part on fossils from marine shales that interfinger with the conglomeratic section. We also collected Pennsylvanian fossils from these shales.
The Pennsylvanian section east of the Picuris-Pecos fault at Rio Chiquito consists of an upward-fining succession of sandstone, pebbly sandstone, conglomerate, sedimentary breccia, mudstone, and marine limestone. The contact between the Flechado Formation and the underlying Proterozoic Ortega Quartzite is low relief and records only a few meters of local paleotopography. For the purposes of mapping, we placed the contact between the Flechado Formation and overlying Alamitos Formation at the top of the uppermost thick (>5 m) quartzarenitic sandstone (unit 92, Fig. 7), a laterally traceable bed. This differs from the original definition of these formations (Sutherland, 1963) that placed the contact at the base of the first arkose (unit 105, Fig. 7). The lowermost arkosic sandstone, however, is thin and weathers recessively, and thus does not provide a useful datum for mapping. Moreover, several thin quartzarenite beds stratigraphically overlie this arkose, indicating interfingering of compositionally distinct sandstones in the stratigraphic transition between the two formations. Thus, by our mapping criteria the Flechado Formation is 391 m thick, but the thickness of the Pennsylvanian section beneath the first arkose is 477 m.
The lower part of the Flechado Formation east of the Picuris-Pecos fault is dominated by sandstone, pebbly sandstone, conglomerate, and sedimentary breccia (Fig. 8). The measured section fines upward, in both maximum clast size and mudstone content. The Flechado Formation also becomes markedly finer and thinner toward the east, away from the Picuris-Pecos fault. Outcrops are commonly conglomeratic near the fault, and breccia is restricted to an area within ∼250 m of the fault. Sandstone and conglomerate are poorly sorted. Beds are generally tabular, and bed thickness typically ranges between 0.3 and 3 m.
Clasts in the Flechado Formation east of the Picuris-Pecos fault consist almost entirely of metasedimentary rocks; most (∼90%) are Ortega Quartzite. Vein quartz, common in the Ortega Quartzite, composes an estimated 5%–10% of clasts and is more common in the upper part of the Flechado Formation. The remainder of the clasts are schist (some are kyanite bearing) that resemble lithotypes in the upper Ortega Quartzite (P.W. Bauer, 2007, personal commun.). Schist clasts become less abundant to the east, away from the Picuris-Pecos fault. No granite or gneiss clasts are present. Maximum clast size of conglomerates is ∼20 cm. The breccia bed (unit 23, Fig. 7) contains clasts of quartzite and schist as much as 2 m in length. Breccia clasts commonly exhibit a jigsaw fabric where individual large clasts have been broken into smaller, semicoherent fragments between which the orientation of bedding and foliation has not been strongly rotated (Fig. 9).
Sedimentary structures in the Flechado Formation are commonly indistinct, probably due to bioturbation or diagenetic overprinting. Sandstone shows solitary or grouped trough cross-beds (facies St of Miall, 1996), and rare horizontal lamination (facies Sh). Coarse pebbly sandstone and conglomerate also commonly display trough cross-bedding (Gt), and subordinate horizontal bedding (Gh). Some conglomerate beds appear massive and display normal grading (facies Gcm). Rare pebbly sandstones show a matrix-supported fabric with pebbles floating in a coarse, sandy matrix (Gmm). Most conglomerate beds have an erosive lower contact. Clast imbrication in the Flechado Formation and the lower, quartzarenitic part of the Alamitos Formation indicates that paleoflow was toward the southeast (Fig. 10). If the observed clockwise deflection of bedding strike near the Picuris-Pecos fault is the result of vertical-axis rotation near the fault, then the restored paleocurrent direction is approximately eastward.
We examined 14 thin sections from the Flechado Formation and 4 from the Alamitos Formation (Fig. 7). Sandstones throughout the Flechado Formation are quartzarenites (Fig. 11) according to the classification of Folk (1974). Monocrystalline quartz is dominant; a few polycrystalline quartz grains were noted. Many quartz grains exhibit undulose extinction and deformation lamellae. Feldspar grains are rare. Schistose metamorphic rock fragments are rare and composed of quartz, mica, and minor feldspar.
Recessive, regolith-covered slopes in the upper part of the measured section appear to be underlain mostly by drab shale. Where exposed, these shales commonly contain marine fossils. A thin section from one of these marine shales (unit 57, Fig. 7) was examined. It is a gray to dark gray, nodular, bioclastic, bioturbated calcareous siltstone. Unit 57 contains crinoids, brachiopods (Orbiculoidea, Derbyia, Linoproductus, cf. Parajuresania, Desmoinesia cf. missouriensis, Mesolobus striatus, Hustedia, Anthracospirifer, Neospirifer cameratus, Composita, Crurithyris, Phricodothyris), gastropods (Pharkidonotus, Euconospira sp., Glabrocingulum? sp.,), bryozoans (Penniretepora, Fenestella?, Rhombopora?), ostracods, bivalves (Aviculopecten, Acanthopecten, Paleolima?, Streblochondria?), rugose corals, trilobite fragments, agglutinated foraminiferans, a small Eotuberitina foraminiferan, and a nautiloid (Domatoceras?). Based on the biostratigraphy of Sutherland and Harlow (1973), the brachiopod assemblage in unit 57 is suggestive of an age close to the Atokan-Desmoinesian boundary (Table 1, sample A).
The Alamitos Formation east of the Picuris-Pecos fault encompasses the uppermost 84 m of the measured section (Fig. 7) and is poorly exposed with thick covered intervals. These covered intervals are probably mostly underlain by marine shale, as can be demonstrated in some cases by tracing them laterally to areas of partial exposure. Sandstone and conglomerate in the lower part of the unit are quartzose and are similar in detrital composition to those in the underlying Flechado Formation. The top of the measured section contains thick marine limestones, and is capped by a 1-m-thick arkosic sandstone. This arkose contains mostly monocrystalline quartz grains and abundant detrital feldspar, dominantly potassium feldspar (Fig. 12). The feldspars are strongly altered. Rare chert grains and detrital micas are also present.
Because the Alamitos Formation is poorly exposed, we obtained no paleocurrent information from the unit in the study area. In the valley of the Rio Santa Barbara, ∼10 km to the north, however, fluvial deposits in the middle part of the Alamitos Formation are well exposed. Paleocurrent measurements from pebble imbrication and cross-bedding in these deposits show derivation of arkosic sand was from the northeast (Fig. 13), even in outcrops near the Picuris-Pecos fault. These paleocurrent data indicate that arkosic sandstone in the Alamitos Formation was not derived from the Uncompahgre uplift west of the Picuris-Pecos fault. This conclusion is supported by the lack of coarsening of the Alamitos Formation toward the Picuris-Pecos fault in the study area. Unlike the Flechado Formation, the mudstone, limestone, and fine sandstone of the Alamitos Formation are juxtaposed against the Picuris-Pecos fault without intervening fault scarp deposits.
Limestone beds (units 101–104, Fig. 7) beneath the arkose contain diverse marine fossils. These include brachiopod shell and spine fragments, crinoid stems and other echinoderm fragments, smaller foraminiferans (Bradyina magna, Bradyina sp., Calcivertella, Climacammina ex gr. moelleri, Climacammina sp., Endothyra, Eotuberitina, Globivalvulina ex gr. bulloides, Glomospira, Syzrania, Tetrataxis), ostracods, bryozoans, fusulinids [Wedekindellina euthysepta (Fig. 14), Beedeina aff. B. hayensis (Fig. 15)], algae and/or problematic algae (Kamaena, Insolenthica, Anthracopelopsis), small gastropods, and very rare Tubiphytes. The fusulinids from units 101 and 103 are of early to early-middle Desmoinesian age (Table 1, sample B; cf. Wilde, 2006).
Rio Chiquito Area West of the Picuris-Pecos fault
Pennsylvanian strata are poorly exposed west of the Picuris-Pecos fault and consist mostly of mudstone with subordinate sandstone, pebbly sandstone, and minor marine limestone. These strata are faulted down against, and apparently overlie, Proterozoic granitic gneiss, but the basal contact is not exposed. Sandstone in the lower (southern) part of the Pennsylvanian succession is quartzarenitic, and is overlain by a fine-grained succession containing arkosic sandstone. Because of poor exposure, Pennsylvanian strata west of the Picuris-Pecos fault were not differentiated into formal units during mapping. Early to middle Desmoinesian marine fossils occur in the upper arkosic succession (Table 1, samples C and D) and allow correlation to the Alamitos Formation. These fossils include brachiopods (Antiquatonia hermosana, Echinaria knighti, Composita subtilita, Phricodothyris perplexa, Linoproductus sp., Linoproductus sp., Punctospirifer, Neospirifer cameratus, Isogramma sp., Derbyia sp., Anthracospirifer sp.), crinoid stem fragments, encrusting and fenestrate bryozoans, the pectinid bivalve Streb-lochondria, a plant stem (Calamites), a trilobite pygidium, and gastropods. Although no age-diagnostic fossils were found in the lower, quartzarenite-bearing beds, we tentatively correlate these strata to the Flechado Formation based on quartzose sandstone compositions and lack of abundant limestone.
Pennsylvanian quartzarenitic strata west of the Picuris-Pecos fault at the Rio Chiquito are distinctly finer grained than those east of the fault. Maximum clast size in pebbly sandstones is 1–2 cm. Despite being apparently more proximal to the source than deposits east of the Picuris-Pecos fault (as shown by paleocurrent data; Fig. 10), the Flechado Formation west of the fault is much finer grained. This textural mismatch is likely the result of post–early Desmoinesian juxtaposition by strike-slip faulting (discussed in the following).
Depositional Environments and Provenance of Pennsylvanian Strata in the Rio Chiquito Area
Based on textural trends, bedding types, paleontology, and paleocurrent data, we interpret the Pennsylvanian succession east of the Picuris-Pecos fault as proximal braid-delta deposits that grade upsection and eastward into marine mudstone and limestone. Quartzarenitic fluvial deposits of the Flechado Formation and the lower Alamitos Formation were derived from Proterozoic metasedimentary rocks in the adjacent Uncompahgre uplift to the west. Uncompahgre-derived quartzarenitic fluvial deposits were onlapped and buried by marine strata during the early to middle Desmoinesian. Mostly marine deposition again gave way to fluvial and deltaic sedimentation during deposition of the Alamitos Formation. Arkosic fluvial deposits in the Alamitos Formation prograded from the north and east, probably mostly from the Cimarron arch, the Sierra Grande uplift, or from sediments that spilled over the Cimarron arch from the central Colorado basin (Fig. 4). The fluvial and marine deposits of the Alamitos Formation grade upward into the entirely nonmarine Sangre de Cristo Formation, which was also derived largely from the north (Soegaard and Caldwell, 1990). Arkosic fluvial and marine deposits derived from the north and east prograded across and buried part or all of the southeastern Uncompahgre uplift during the middle to late Desmoinesian, although the paleogeographic extent of this burial is uncertain because most Pennsylvanian deposits in the Santa Fe Range west of the Picuris-Pecos fault have been eroded.
The Flechado Formation east of the Picuris-Pecos fault at Rio Chiquito is remarkably coarse grained, more so than any other Pennsylvanian succession we have observed in New Mexico. Pebbly sandstone, conglomerate, and sedimentary breccia constitute ∼20% of the Flechado Formation at Rio Chiquito. Clasts within the breccia bed are generally highly fractured, but the fractures generally do not extend into the matrix between the clasts (Fig. 9). Such jigsaw fracture fabrics are common in rock-avalanche breccia (Yarnold, 1993; Friedmann, 1997). Sedimentary breccia suggests a nearby, steep source terrane.
Textural and facies evidence indicates that fluvial strata of the Flechado Formation east of the Picuris-Pecos fault were deposited in proximity to their source area. Paleocurrent data and the eastward fining and thinning of the Flechado Formation show this source region was to the west, and clast composition indicates the source consisted of metasedimentary rocks. Today, the terrane west of the Picuris-Pecos fault consists of Proterozoic granite and granitic gneiss. This lithologic mismatch provides strong evidence that major strike slip on the Picuris-Pecos fault has occurred since the early Desmoinesian. The closest possible source for the Ortega Quartzite clasts in the Flechado Formation exposed east of the Picuris-Pecos fault is in the Picuris Mountains, now dextrally separated by at least 20 km from the Rio Chiquito exposures.
Monazite Age Constraints on Provenance of Quartzite Clasts at Rio Chiquito
Although it is clear that the source of quartzite clasts in the Pennsylvanian strata at Rio Chiquito is the Ortega Quartzite, a more precise knowledge of which part of the Ortega supplied the detritus is desirable. The Ortega Quartzite is best exposed in the Tusas Mountains ∼70 km northwest of the study area (Fig. 16), where it forms an exposure belt ∼30–40 km wide that is an expression of the structurally complex synclinorium in which the quartzite is preserved. East of the Picuris-Pecos fault, the Ortega Quartzite also forms a structurally complex belt ∼40 km wide that is partly obscured by Paleozoic strata. There, the Ortega is preserved in a complex asymmetrical synclinorium that is bounded by the steep Pecos thrust on the south and a beveled erosional edge beneath Paleozoic strata on the north near the Picuris Mountains (Fig. 3; see fig. 5 of Cather et al., 2006, for cross sections that show the synclinorium geometry). Because the Ortega Quartzite belt east of the Picuris-Pecos fault is still partly mantled by late Paleozoic strata, the areal distribution of the Ortega in this region is an approximation of the geometry of the Ortega paleoexposure belt during the Pennsylvanian. In the Picuris Mountains directly west of the Picuris-Pecos fault, only the southernmost 7 km of the Ortega Quartzite belt is currently exposed; the remainder is presumably faulted beneath the fill of the Neogene San Luis Basin of the Rio Grande Rift, as is suggested by an area of low aeromagnetic value that extends ∼30 km northward from the Plomo fault (Cather et al., 2006).
No systematic lateral variation of lithology is apparent within the Ortega Quartzite (Soegaard and Eriksson, 1985), so conventional petrographic techniques cannot be used to specify a detrital source within the Ortega outcrop belt. The age population of detrital zircons in the Ortega Quartzite is also uniform (1.80–1.70 Ga; Jones et al., 2009). Monazite ages within the unit, however, exhibit a systematic southward younging in the Tusas Mountains (Kopera et al., 2002; Kopera, 2003). In the northern part of the Tusas Mountains, the Ortega Quartzite contains ca. 1.7 Ga (1.85–1.65 Ga) monazite that represents both pre–1.69 Ga detrital grains and monazite that grew during ca. 1.67 Ga metamorphism. In the south, monazite ages are entirely ca. 1.4 Ga and reflect recrystallization and growth of new monazite at deeper structural levels to the south during ca. 1.4 Ga peak metamorphism (Kopera et al., 2002; Kopera, 2003, his fig. 6.19). Mixed ages, commonly with older (ca. 1.7 Ga) cores and younger (ca. 1.4 Ga) rims, exist in the central Tusas Mountains in a zone ∼20–30 km north of the southern structural boundary of the Ortega outcrop belt (Fig. 16). Although our results show that the age zonation of Ortega monazite is less systematic near the Picuris-Pecos fault (see following), we dated monazite grains from clasts sampled from the measured section in the Flechado Formation at Rio Chiquito to better refine the provenance of these clasts.
We analyzed monazite grains from seven quartzite clasts in the Flechado Formation and one quartzite clast from the lower part of the Alamitos Formation (Fig. 7; analytic techniques are presented in the Supplemental File1). These clasts were sampled from the lower 430 m of the Pennsylvanian section from east of the Picuris-Pecos fault at Rio Chiquito. Between 10 and 20 individual age determinations on 5–6 monazite grains were carried out for each sample. Monazite grains yielded ages ranging between ca. 1.2 and 1.8 Ga (Fig. 17; for individual samples, see the Supplemental File [see footnote 1]) and form two distinct modes, one centered ca. 1.4 Ga and the second ca. 1.7 Ga.
Monazite grains in the quartzite clasts are typically fine grained (∼10–20 μm in diameter), and many exhibit multiple compositional domains that in some cases correspond to different age determinations (Fig. 18). Some of the compositional domains are small relative to the 3–5 μm beam size, so some of the determined ages may be a combination of older and younger ages. These may account for some data points that plot between the two age-distribution peaks.
Quartzite clasts from the Pennsylvanian section at Rio Chiquito all contain monazite grains representative of the ca. 1.4 Ga and 1.7 Ga age modes, except 3 clasts sampled from 23 m, 68 m, and 100 m above the base of the Flechado Formation. These yielded only ca. 1.4 Ga ages. In contrast to the dominantly mixed ages in clasts at the Rio Chiquito, the basement rocks in the nearest possible source region, the Picuris Mountains, exhibit nearly all ca. 1.4 Ga monazite ages. Monazite grains from the southern (Wingsted, 1997) and northern (Daniel and Pyle, 2006) parts of the range have yielded only ca. 1.4 Ga ages. We analyzed two Ortega Quartzite samples from the central part of the range near Picuris Peak. One sample yielded only ca. 1.4 Ga ages; the other exhibited a range of ages between ca. 1.1 and 1.7 Ga (samples PPF-35 and PPF-38; the Supplemental File [see footnote 1]). Thus, unlike the southernmost exposures of Ortega Quartzite in the Tusas Mountains that so far have yielded only ca. 1.4 Ga monazite (Kopera, 2003), the Picuris Mountains also contain a small portion of ca. 1.7 Ga monazite. Nonetheless, the dominance of ca. 1.4 Ga monazite ages in the Picuris Mountains suggests that the southward-younging zonation of monazite ages documented in the Tusas Mountains (Kopera, 2003) probably also exists near the Picuris-Pecos fault.
Several conclusions can be drawn from the presence of mixed ca. 1.4 and 1.7 Ga ages of monazites in quartzite clasts from Pennsylvanian strata near the Rio Chiquito. (1) The clasts were probably not derived from the northern part of the Ortega Quartzite outcrop belt in which ca. 1.7 Ga ages predominate. (2) The clasts were probably not derived from the southernmost part of Ortega belt in the Picuris Mountains because most monazite grains there are ca. 1.4 Ga. This conclusion is in need of further testing, however, as one of our samples from the Picuris Mountains did contain ca. 1.7 Ga monazite grains. (3) The clasts most likely were derived from the central, mixed-age part of the Ortega belt. The zone of mixed-age monazite in the Tusas Mountains is ∼20–30 km north of the southern edge of the Ortega belt. If a monazite zonation similar to that documented by Kopera (2003) in the Tusas Mountains existed near the Picuris-Pecos fault, then the most likely source of the quartzite clasts is the central Ortega belt, now downfaulted into the Rio Grande Rift near Taos. If this is correct, then the quartzite clasts in the Rio Chiquito section east of the Picuris-Pecos fault have been separated from their source regions west of the fault by ∼40–50 km of dextral slip since the early Desmoinesian. (4) Clasts containing mixed ca. 1.4 Ga and 1.7 Ga monazite ages are present at both the bottom and top of the 430-m-thick sampled interval and imply that the detritus for this stratigraphic interval was derived primarily from the central mixed-age zone in the Ortega Quartzite belt. The lack of clasts derived from other parts of the Ortega belt implies that major strike slip did not occur during deposition of the Flechado and lower Alamitos Formations (Morrowan(?) to early Desmoinesian; time interval of ∼13 m.y. or less; Davydov et al., 2010).
Cave Creek Area
In this section we document paleocurrent trends and clast lithologies in Pennsylvanian sandstones near Cave Creek, south of the Truchas uplift (Fig. 3). Metasedimentary clasts present in the Flechado and La Pasada Formations were derived from the Uncompahgre uplift west of the Picuris-Pecos fault. The mismatch between these conglomeratic strata and their nearest possible source in the Picuris Mountains requires at least ∼40 km of post-Desmoinesian dextral slip on the fault.
Cave Creek Area Description
South of the Truchas uplift, Pennsylvanian strata overlie Mississippian strata and are juxtaposed against Proterozoic granitic gneiss west of the Picuris-Pecos fault. The fault there dips steeply west (∼80°) and exhibits a linear trace with segments that branch or anastomose (Moench et al., 1988). Exposure quality in this deeply forested area is poor, except locally in river valleys. We examined exposures along Cave Creek and in adjacent drainages to the north (Horsethief Creek) and south (Rito Oscuro and Windsor Creek).
Two distinct clastic petrofacies are present in Pennsylvanian strata near Cave Creek. The quartzarenitic petrofacies is expressed in outcrop and float near the Picuris-Pecos fault along Cave Creek and in areas to the north, and consists of quartzarenitic sandstone, pebbly sandstone, and rare conglomerate. Maximum clast size is ∼8 cm. Clasts are mostly (∼90%) Ortega Quartzite; the remainder are schist and vein quartz (Fig. 19). No granite or gneiss clasts were observed.
The quartzarenitic petrofacies interfingers eastward with mudstone and marine limestone that contain Desmoinesian fossils (Sutherland and Montgomery, 1975, p. 89). Limestone is rare near the Picuris-Pecos fault, but becomes abundant 2–3 km east of the fault. Using the nomenclature of Sutherland (1963), the quartzarenitic Pennsylvanian beds near the fault are the Flechado Formation, and the intercalated limestone and quartzarenitic strata to the east are the La Pasada Formation. Because poor exposure prohibits confident mapping of the lateral gradation between these formations, we refer to their clastic components collectively as the quartzarenitic petrofacies. We collected paleocurrent data from the quartzarenitic petrofacies along Cave Creek ∼2 km east of the fault; these show eastward paleoflow (Fig. 20).
Near the Picuris-Pecos fault, the quartzarenitic petrofacies is overlain by the arkosic petrofacies ∼1 km south of Cave Creek. The arkosic petrofacies consists of feldspathic sandstone and sparse pebbly sandstone that are interbedded with marine limestone and mudstone, a lithology that supports correlation to the Alamitos Formation. A fossiliferous limestone in the northern part of the arkosic petrofacies yielded Desmoinesian brachiopods (Table 1, sample E), and fossils in the arkosic petrofacies ∼5 km to the southeast are middle to late Desmoinesian (Windsor Creek locality of Sutherland and Harlow, 1973, interpreted by them to be lower Alamitos Formation). The maximum clast size is 3 cm; pebbles consist of varying proportions of granite, vein quartz, and feldspar megacrysts.
Cave Creek Area Interpretation
A western source for the quartzarenitic petrofacies is indicated by (1) decreased abundance of limestone and mudstone intercalated within the quartzarenitic petrofacies near the Picuris-Pecos fault, and (2) paleocurrent data from clast imbrication that record paleoflow toward the east. These paleocurrent data indicate that quartzite clasts in the quartzarenitic petrofacies were derived from metasedimentary rocks exposed on the Uncompahgre uplift, west of the Picuris-Pecos fault. Because of poor exposure, we were unable to obtain paleocurrent information from the arkosic petrofacies. The presence of limestone and mudstone and the lack of coarse conglomerate in this facies near the Picuris-Pecos fault, however, indicate it was not derived from the adjacent Uncompahgre uplift. More likely, arkose was distally derived from granitic rocks in the Cimarron arch or Sierra Grande uplift to the north and east, as demonstrated for arkose of the Alamitos Formation at the Rio Santa Barbara (Fig. 13) and arkose of the overlying Sangre de Cristo Formation (Soegaard and Caldwell, 1990).
Regional stratigraphy and field relationships indicate that the arkosic petrofacies (Alamitos Formation) overlies the quartzarenitic petrofacies (Flechado and La Pasada Formations, undivided) south of Cave Creek. The ultimate southward extent of the quartzarenitic petrofacies adjacent to the Picuris-Pecos fault is thus obscured by the overlying arkosic petrofacies (Alamitos Formation) south of Cave Creek. Nonetheless, it is clear that the quartzarenitic petrofacies extends southward along the Picuris-Pecos fault at least to the Cave Creek area. The southern boundary of the metasedimentary terrane west of the fault was therefore also near, or south of, the latitude of Cave Creek area during the Desmoinesian. Today, this boundary (the Plomo fault in the Picuris Mountains) is dextrally separated from the southernmost extent of the quartzarenitic petrofacies by at least 40 km, and perhaps by as much as 50 km if our inferences based on monazite ages are correct. These estimates exceed the ∼37 km dextral separation that exists today (as measured between the originally continuous Hondo and Brazos Cabin synclines; Montgomery, 1963), and imply that a sinistral separation existed on the fault during the middle Pennsylvanian. This possibility is supported by evidence for pre-Permian, ∼20°–30° anticlockwise rotation of gneissic foliation in the damage zone of the Picuris-Pecos fault at Deer Creek ∼35 south of the study area (Cather et al., 2008), and is compatible with the predicted sense of shear on the fault during the Grenville and Neoproterozoic deformations (Cather et al., 2006).
Several workers have advocated a Proterozoic origin for the dextral separation on the Picuris-Pecos fault. The first such interpretation was by Montgomery (1963), who mapped folds near the Picuris-Pecos fault in the Picuris Mountains and interpreted them to have resulted from dextral-sense drag on the fault (the geometry of these folds is also compatible with east-up shear; Luther et al., 2008). These folds refolded metasedimentary rocks in both limbs of the east-west–trending synclinorium exposed in the Picuris and Truchas Mountains. Montgomery (1963, p. 16) reasoned that this refolding required “extreme plasticity” and therefore occurred during Proterozoic regional metamorphism, and that dextral slip on the adjacent Picuris-Pecos fault was therefore probably also Proterozoic. Peak metamorphic conditions were attained ca. 1.4 Ga in northern New Mexico with temperatures of 500–550 °C and pressures of 3.5–4.0 kbar (Williams et al., 1999a; Karlstrom et al., 2004). Fault slip under such conditions should produce mylonitic foliation, but to date no mylonites have been reported anywhere along the Picuris-Pecos fault. Moreover, outcrop-scale fracturing and quartz microstructural fabrics in the refolds developed in brittle and brittle-ductile regimes (McDonald and Nielsen, 2004; Luther et al., 2008), not during peak metamorphism. The significance of these refolds to the slip history of the Picuris-Pecos fault is not yet clear. Although initiation of the fault near peak metamorphic conditions can be ruled out, subsequent Proterozoic (Grenville and younger) initiation remains possible (Cather et al., 2006).
Several workers have interpreted Proterozoic dextral slip on the Picuris-Pecos fault using process-of-elimination arguments against Ancestral Rocky Mountain and Laramide dextral slip. Yin and Ingersoll (1997) interpreted Ancestral Rocky Mountain and Laramide faults in the Southern Rocky Mountains as mostly dip slip, and that the observed dextral basement separations therefore must be Proterozoic. They cited a lack of definitive evidence for strike slip during the Phanerozoic orogenies, but did not produce compelling evidence for their preferred dip-slip model (see Cather, 2004, p. 236, for a critique of their Laramide model).
Erslev (2001) collected numerous kinematic data from minor faults in Permian and younger rocks throughout north-central New Mexico; he calculated an east-northeast Laramide σ1 direction that is nearly perpendicular to the major dextral faults in the region (including the Picuris-Pecos fault), and noted relatively few dextral minor faults parallel to the major faults. From these relationships, Erslev (2001, p. 63) interpreted only “limited” (but unspecified) dextral slip on north-striking faults in the Southern Rocky Mountains during the late Laramide. We note, however, that elsewhere σ1 is nearly perpendicular to known strike-slip faults, such as the San Andreas fault (e.g., Mount and Suppe, 1987; Zoback et al., 1987; Provost and Houston, 2001; Wilson et al., 2003) and the Great Sumatran strike-slip fault (Mount and Suppe, 1992). Moreover, detailed studies have shown that very few dextral minor faults parallel the dextral Punchbowl fault of California, a well-studied, extinct strand of the San Andreas fault system (Chester and Logan, 1986; Chester and Chester, 1998, Wilson et al., 2003). For these reasons, we question the rationale of Erslev's (2001) method for determination of slip of a major fault using adjacent minor faults, particularly if the major fault is weak (e.g., Ghisetti, 2000).
Paleomagnetism has been used to test for vertical-axis rotations associated with dextral faulting along the eastern Colorado Plateau boundary. Although several areas of clockwise vertical-axis rotation in Permian–Triassic rocks have been documented in the Southern Rocky Mountains of New Mexico and Colorado (e.g., Wawrzyniec et al., 2002; Geissman, 2004), a locality on the Picuris-Pecos fault ∼35 km south of the study area shows no evidence of significant vertical-axis rotation since the late Paleozoic and has been interpreted to indicate that no major strike slip has occurred since then (Wawrzyniec et al., 2007). However, several localities that lack paleomagnetically defined, vertical-axis rotations exist along the dextral San Andreas and San Gabriel faults (e.g., Tavarnelli, 1998; Levi et al., 2005), and show that strike-slip faulting is not everywhere associated with large rotations.
The mismatch between the mineralogic composition of Pennsylvanian clastic rocks in the western Taos trough and source terranes west of the Picuris-Pecos fault was first noted by Sutherland (1963). He accepted Montgomery's (1963; see above) hypothesis that dextral separation on the Picuris-Pecos fault probably occurred during the Proterozoic. So, to explain the mismatch, Sutherland (1963, p. 41) argued that west-up Pennsylvanian slip caused the northward erosional retreat of metasedimentary rocks west of the Picuris-Pecos fault. Sutherland (1963), following Montgomery (1963), envisioned the southern boundary of metasedimentary terrane in the Picuris Mountains to be marked by an intrusive contact with younger granitic rocks to the south; if this intrusive contact dipped gently to the north, Sutherland reasoned, then Pennsylvanian uplift and erosion on the west side of the Picuris-Pecos fault would cause the northward retreat of the metasedimentary terrane boundary and could explain the mismatch.
There are several questionable aspects of Sutherland's (1963) erosional retreat model. First, Pennsylvanian uplift and erosion west of the Picuris-Pecos fault would necessarily have been deep and extensive, as no remnants of the metasedimentary rocks are preserved at even the highest elevations of the Santa Fe Range (3847 m above sea level). However, as described previously, a variety of geologic and thermochronologic data argue against major dip slip (Pennsylvanian or otherwise) on the Picuris-Pecos fault. Second, recent mapping and geochronologic studies have revealed errors in the original interpretation by Montgomery (1963) of the Proterozoic geology in the Picuris Mountains. Although local, younger granites (ca. 1.4 and 1.65 Ga) were intruded in the region south of the metasedimentary terrane in the Picuris Mountains, the basement underlying the metasedimentary rocks of northern New Mexico is mostly Yavapai age (ca. 1.76–1.72 Ga; Karlstrom et al., 2004) and thus cannot be in intrusive contact with the overlying, younger (ca. 1.69 Ga) metasedimentary rocks. Moreover, detailed mapping and kinematic analysis in the Picuris Mountains (Bauer, 1993, 2004) have shown that the southern boundary of the metasedimentary terrane is a steep, south-dipping ductile thrust (the Plomo fault), not a gently north-dipping intrusive contact. Thus, uplift and erosion west of the Picuris-Pecos fault would cause a southward migration of the metasedimentary terrane boundary, not northward as envisioned by Sutherland (1963).
Baltz and Myers (1999) speculated that the Pennsylvanian provenance mismatch first noted by Sutherland (1963) could be explained by sediment shed from metasedimentary rocks exposed in a Pennsylvanian precursor to the Laramide Truchas uplift east of the Picuris-Pecos fault, rather than being derived from west of the Picuris-Pecos fault. The Truchas uplift developed in response to west-up slip on the Jicarilla fault, a steep (∼75°) west-dipping reverse fault that juxtaposes metasedimentary rocks (mostly Ortega Quartzite) with folded Pennsylvanian strata on the east (Fig. 3). To test the possibility of Pennsylvanian orogenesis in the Truchas area, we examined well-exposed Pennsylvanian beds cut by the Jicarilla fault north of Pecos Baldy Lake and at Chimayosos Peak. In both areas, these beds contain fossils of early Desmoinesian age (see measured sections 10 and 40 of Sutherland and Harlow, 1973), and thus are the same age as the upper part of the conglomeratic section at Rio Chiquito. Despite their proximity to the Jicarilla fault, the measured sections of Sutherland and Harlow (1973) are dominated by mudstone and marine limestone (Fig. 21). Conglomerate and pebbly sandstone occur only as rare, thin beds in which the maximum size of quartzite clasts is typically 1–3 cm. Such lithologic attributes are atypical for sediments deposited adjacent to an active mountain-front fault, and thus suggest that no early Desmoinesian uplift occurred in the Truchas area. Paleocurrent data provide an additional argument against a Pennsylvanian precursor to the Truchas uplift. Pebble imbrications in metasedimentary-sourced Pennsylvanian deposits show no evidence of paleoflow away from a precursor uplift, but instead show that paleoflow was from the west (Figs. 3, 10, and 20).
Although the slip history of the Picuris-Pecos fault is likely complex (Cather et al., 2006), we now have two snapshots into this history. The first was during regional metamorphism in the Mesoproterozoic (ca. 1.4 Ga). At this time the Picuris-Pecos fault did not exist. The second snapshot is during the middle Pennsylvanian, when there is clear evidence for both the existence of the fault as a basin-bounding structure and for the paleolocation of a distinctive metasedimentary terrane in the uplands west of the fault (Fig. 22). There are three lines of evidence that suggest ∼40–50 km of post–early Desmoinesian dextral slip occurred on the Picuris-Pecos fault: (1) Ortega Quartzite clasts in the lower part of the Rio Chiquito section are separated by at least 20 km from their nearest possible source in the Picuris Mountains; (2) monazite ages within these clasts further suggest they are separated from their probable source in the middle part of the Ortega Quartzite belt by ∼40–50 km; and (3) metasedimentary-derived detritus near Cave Creek is separated from its nearest potential source area by at least ∼40 km. This ∼40–50 km estimate of post–early Desmoinesian dextral slip exceeds the present ∼37 km dextral separation of Proterozoic features by the Picuris-Pecos fault, and suggests that no dextral separation, but rather a small sinistral separation (∼3–13 km), existed on the Picuris-Pecos fault during the early Desmoinesian. Thus, either no major pre-Desmoinesian dextral slip occurred on the Picuris-Pecos fault, or such slip was effectively cancelled by other episodes of sinistral slip prior to the Desmoinesian.
Our results imply that the present ∼37 km dextral separation on the Picuris-Pecos fault accumulated late in the Ancestral Rocky Mountain orogeny and/or during the Laramide orogeny. With the exception of at least a few kilometers of dextral slip that must be Laramide (Cather and Lucas, 2004, based on lithofacies of the Late Cretaceous Dakota Sandstone that are mismatched across the fault), the apportionment of dextral slip between these two orogenies is unresolved.
Research was funded by the New Mexico Bureau of Geology and Mineral Resources (Peter A. Scholle, Director). We thank Bruce Allen and Daniel Vachard for fusulinid identification. We benefitted from discussions with Gary Axen, Elmer Baltz, Paul Bauer, Chris Daniel, Matt Heizler, Mike Jercinovic, Karl Karlstrom, Amy Luther, David McDonald, Mike Timmons, and Mike Williams. We thank Amy Luther for Ortega Quartzite samples from the Picuris Mountains. The Cameca SX-100 electron microprobe at New Mexico Institute of Mining and Technology was partially funded by NSF Grant STI-9413900. The Pecos Wilderness part of this study would not have been possible without the assistance of Curtis Verploegh. The manuscript was improved by informal reviews from Gary Axen, Amy Luther, and Mike Timmons. We appreciate careful editing by Bill Thomas, Associate Editor Mike Williams, and an anonymous reviewer.