Magmatism and deformation in a terrane suture zone south of the Denali fault, northern Talkeetna Mountains, Alaska
Published:January 01, 2007
- PDF LinkChapter PDF
Ronald B. Cole, Paul W. Layer, Benjamin Hooks, Andrew Cyr, Julie Turner, 2007. "Magmatism and deformation in a terrane suture zone south of the Denali fault, northern Talkeetna Mountains, Alaska", Tectonic Growth of a Collisional Continental Margin: Crustal Evolution of Southern Alaska, Kenneth D. Ridgway, Jeffrey M. Trop, Jonathan M.G. Glen, J. Michael O'Neill
Download citation file:
Volcanic and granitic rocks of the Jack River igneous field were erupted and emplaced in the suture zone between the accreted Wrangellia composite terrane and the former margin of southern Alaska. The volcanic rocks unconformably overlie Jurassic-Cretaceous shale and sandstone of the Kahiltna assemblage and include 100–300 m of basalt, basaltic andesite, and andesite lava flows overlain by a rhyolite unit that consists of over 900 m of lava flows and pyroclastic deposits. Seven basaltic and rhyolite lava samples yield 40Ar/39Ar ages ranging from 56.0 ± 0.3 to 49.5 ± 0.3 Ma. Two granitic samples yield 40Ar/39Ar ages of 54.6 ± 0.4 and 62.7 ± 0.4 Ma. These age dates indicate that the onset of Jack River magmatism at ca. 62.7 Ma coincided with the terminal phase of terrane accretion and continued after accretion to at least 49.5 Ma.
The volcanic rocks range between tholeiitic and high-K calc-alkaline series and show a bimodal distribution with respect to silica (dacite is absent). The Jack River basalts are tholeiitic, have rare earth element and high field strength element ratios that are in the range between Pacific enriched mid-ocean-ridge basalts and Hawaiian ocean-island basalts (e.g., La/Yb = 5.0–8.4; Nb/Zr = 0.07–0.11), and have a within-plate geochemistry (e.g., Ti/V >50; high Zr/Y). All of the Jack River volcanic rocks exhibit some degree of enrichment in large ion lithophile and/or fluid mobile elements (e.g., Cs, Ba, Th, U, K, and Pb), although the basalts have low ratios between large ion lithophile and high field strength elements (e.g., Ba/Nb as low as 32.7 and Pb/Nb of 0.28–0.35). The granitic rocks (granites to granodiorites) are strongly depleted in the heavy rare earth elements, and most samples exhibit characteristics of adakites (e.g., Al2O3 >15 weight %, Yb = 0.6–1.2 ppm, Y = 5.5–12.5 ppm, and Sr/Y = 20.4–66.2).
The Jack River basalts were derived from partial melts of a mantle source that was more enriched than depleted mid-ocean-ridge basalt mantle and that ranged toward an enriched mantle (EM-I-type) composition.The basalts then evolved by assimilation and fractional crystallization to form intermediate magmas. Rhyolite magmas were formed later as anatectic melts of upper crustal argillaceous rocks (Kahiltna assemblage), resulting in the bimodal volcanism. The granitic adakite magmas may have formed by melting of garnet-bearing metamorphosed sedimentary rocks (meta-Kahiltna assemblage) that formed lower crustal rocks in the suture zone. Although the Jack River igneous rocks do exhibit some arc-like geochemical characteristics (e.g., elevated large ion lithophile elements), they differ from calc-alkaline arc rocks in that (1) they are a bimodal volcanic suite; (2) the rhyolites are not comagmatic with the basaltic and intermediate rocks; (3) the basalts and andesites have higher TiO2 (>1.5 weight %) than is typical for arc basalts and andesites; (4) the basalts do not exhibit depletion of high field strength elements (e.g., Ta and Nb) with respect to large ion lithophile elements; (5) the basalts have an intraplate geochemical affinity; and (6) adakites are present. These characteristics show that the geochemistry of postcollisional suture zone magmatism can be transitional between calc-alkaline arc and intraplate magmatism.
The Jack River volcanic field is deformed into a broad, northeast-trending syncline, which is crosscut by small-scale brittle faults that include northwest- and west-trending normal-slip and oblique-slip faults, and a southeast-dipping reverse fault that places Kahiltna assemblage rocks over the Jack River volcanic rocks. The pattern of Jack River deformation is consistent with right-lateral simple shear along the Denali fault system and indicates an episode of post-49.5 Ma strike-slip along the McKinley strand of the Denali fault. The Jack River rocks, therefore, record the magmatic response to terrane accretion and the kinematics of margin-parallel transport of an accreted terrane assemblage after it was sutured to the continental margin.
Volcanic and granitic rocks in the northern Talkeetna Mountains record the magmatic and deformational responses that occurred during the collisional to postcollisional stages of terrane accretion to southern Alaska. Terrane accretion with concurrent magmatism is a hallmark of the Mesozoic and early Cenozoic tectonic history of southern Alaska. For example, the Wrangellia composite terrane, a late Paleozoic—Mesozoic island-arc assemblage (Jones et al., 1982; Csejtey et al., 1982; Nokleberg et al., 1985) collided with the continental margin of southern Alaska beginning sometime between Late Jurassic and Late Cretaceous time (Jones et al., 1982; McClelland et al., 1992; Trop et al., 2002); deformation due to accretion culminated by early Paleocene time (Cole et al., 1999; Ridgway et al., 2002). This collisional event resulted in a protracted Mesozoic-Cenozoic history of magmatism, deformation, basin development, and metamorphism in southern Alaska that ultimately shaped the modern Alaska Range (Csejtey et al., 1982; Lanphere and Reed, 1985; Plafker and Berg, 1994; Cole et al., 1999; Ridgway et al., 2002). Volcanic and granitic rocks in the northern Talkeetna Mountains (referred to herein as the Jack River igneous field) were erupted and emplaced within the suture zone that formed between the Wrangellia composite terrane and the former continental margin of southern Alaska (Fig. 1). The suture zone (referred to as the Alaska Range suture zone by Ridgway et al., 2002) contains Mesozoic marine sandstone and shale of the Kahiltna assemblage that was deposited along the leading edge of the Wrangellia composite terrane and was deformed during terrane accretion (Csejtey et al., 1982; Wallace et al., 1989; Ridgway et al., 2002). Prior to Wrangellia composite terrane accretion, calc-alkaline arc magmatism occurred across western and south-central Alaska, represented in part by 75–56 Ma plutons of the Alaska Range–Talkeetna Mountains magmatic belt (Fig. 1; Wallace and Engebretson, 1984; Moll-Stalcup, 1994). After Wrangellia composite terrane accretion, there was a hiatus in regional calc-alkaline magmatism that resumed farther south at ca. 45 Ma and continues today with the modern Aleutian arc (Fig. 1). The Jack River igneous field is unique because it represents a combination of mantle-derived magmas and crustal melts (including adakites) that formed within the suture zone closely following terrane accretion and subsequent to arc magmatism. The Jack River volcanic rocks also record an episode of right-lateral strike-slip deformation and so provide an important link between magmatism and the crustal kinematics associated with terrane accretion and the Paleogene tectonic events of southern Alaska.
The Jack River igneous field lies within the Late Cretaceous–Paleocene Alaska Range–Talkeetna Mountains magmatic belt (Fig. 1) that was defined by Wallace and Engebretson (1984) and Moll-Stalcup (1994). As summarized by Moll-Stalcup (1994), the 75–56 Ma rocks of this belt have chemical compositions typical of subduction-related magmatism (e.g., high-K, Ba, Sr, Rb, and Th and low Nb, Ta, and Ti relative to the light rare earth elements). The 56–50 Ma rocks in the western and central parts of this belt “represent a transition from subduction-related magmatism to post-subduction, possibly intraplate, magmatism” (Moll-Stalcup, 1994, p. 603). This transition is marked by a decrease in magma volume, an increase in the proportion of basalt, eruption of alkalic basalt and rhyolite, and emplacement of peraluminous granites (e.g., the McKinley sequence plutons; Moll-Stalcup, 1994). Moll-Stalcup (1994) further noted that insufficient data are available for rocks at the eastern end of the belt, including those in the northern Talkeetna Mountains (i.e., this study), to determine if a similar transition occurred there. Our new data show that Jack River magmatism began at ca. 62.7 Ma with the formation of peraluminous granitic adakites and continued to at least 49.5 Ma with the eruption of basalts, andesites, and rhyolites. The Jack River igneous rocks exhibit geochemical characteristics similar to, and also overlap the period of, postsubduction transitional (intraplate) magmatism defined by Moll-Stalcup (1994), thus confirming that an episode of Paleocene-Eocene transitional magmatism occurred in the northern Talkeetna Mountains as well as across western Alaska. Paleogene volcanic rocks in the central and southern Talkeetna Mountains (Fig. 1) also show age and geochemical characteristics that are anomalous for calc-alkaline arc magmatism and likewise represent episodes of transitional magmatism (Amos and Cole, 2003; Cole, 2003; Cole and Stewart, 2005; Cole et al., 2006).
The Jack River volcanic field is situated along the south side of the Denali fault system, a major right-lateral strike-slip fault that extends along southern Alaska and western Canada (Fig. 1). Hundreds of km of Late Cretaceous–Paleogene dextral offset have been estimated for different segments of the Denali fault system (Eisbacher, 1976; Jones et al., 1982; Nokleberg et al., 1994; Miller et al., 2002). The McKinley fault is the main segment of the Denali fault system in south-central Alaska and is estimated to have experienced less than ∼40 km of dextral displacement since Oligocene time (Reed and Lanphere, 1974; Hickman et al., 1977; Brewer, 1982; Cole, 1999; Trop et al., 2004). Recent large earthquakes confirm that the Denali fault remains an active right-lateral strike-slip system (Eberhart-Phillips et al., 2003).
The Jack River volcanic and granitic rocks overlie and intrude marine sedimentary rocks of the Kahiltna assemblage (Fig. 2), which includes shale, siltstone, graywacke, chert, and conglomerate that were deposited by submarine-fan systems along the leading edge of the Wrangellia composite terrane island-arc assemblage (Eastham and Ridgway, 2002; Ridgway et al., 2002; Hampton et al., this volume). In the northern Talkeetna Mountains the Kahiltna assemblage has a depositional thickness of ∼3–5 km (Ridgway et al., 2002), although elsewhere in the suture zone these rocks may be structurally thickened (Barnes, 1977). The crustal rocks that underlie the Kahiltna assemblage are not well defined, but in places north and northwest of the Jack River volcanic field, the Kahiltna assemblage unconformably overlies or is in fault contact with Late Paleozoic through Triassic rocks of Wrangellia affinity (Clautice et al., 2001; O'Neill et al., 2003; Hampton et al., this volume). These rocks include tholeiitic basalt (geochemically similar to Late Triassic Nikolai basalts of southern Alaska), limestone, chert, tuff, mudstone, sandstone, and conglomerate, all subjected to low-grade regional metamorphism (Clautice et al., 2001).
RADIOMETRIC AGE DATA
We determined new 40Ar/39Ar ages for eight volcanic samples, two granitic pluton samples, and one granitic conglomerate clast from the Jack River volcanic field (Table 1; Figures 3 302 and 4). The 40Ar/39 Ar age determinations were performed in the geochronology laboratory at the Geophysical Institute, University of Alaska, Fairbanks. We report plateau ages where plateaus were identified on the basis of a sample having three or more contiguous fractions constituting greater than 50% of gas release and a mean square weighted deviation of plateau steps less than 2.5. Samples containing fractions that are about the same age but that fail the plateau criteria are termed pseudoplateaus, as noted on Figure 4. Analytical procedures for age determinations are outlined in Layer (2000).
Seven of the Jack River volcanic samples yield plateau ages ranging from 56.0±0.3 to 49.5±0.3. The youngest age (49.5 Ma) is for the stratigraphically highest lava exposed in the central part of the volcanic field (Fig. 3) 302; because there is no upper stratigraphic control (no rocks directly overlie the Jack River volcanic rocks) we consider that the end of Jack River volcanism was ≤49.5 Ma. The stratigraphically lowest lavas across the volcanic field yield an age range of 56–53.5 Ma (Fig. 3) 302, which we consider to be the onset of Jack River volcanism. In addition, a poor-quality plateau age of 76.7±0.6 Ma was obtained on a devitrified tuff that overlies Kahiltna sedimentary rocks and underlies Jack River lavas at the east end of the volcanic field (Figs. 3 302 and 4). This sample yielded an isochron age of 74.9±0.5 and an integrated age of 65.2±0.5 (Table 1). This tuff sample is ambiguous; it may be a Late Cretaceous tuff unrelated to Jack River magmatism, or it may have yielded an erroneously old age date and have been comagmatic with the Jack River rocks. Geochemically, this tuff is similar to the Jack River granitic rocks and so may represent an extrusive equivalent of those plutons.
The granitic samples yield plateau ages of 54.6±0.4 and 62.7± 0.4 Ma for plutons on the north and south sides of the volcanic field, respectively (Figs. 3 302 and 4, Table 1). The granitic conglomerate clast, which was sampled from a conglomerate unit at the base of the volcanic rocks, yields a pseudoplateau age of 64.1± 0.7 Ma.
One previous K-Ar age of 53.9±1.6 Ma on sanidine from rhyolite at the west end of the volcanic field was reported by Csejtey et al. (1992; Fig. 3 302, Table 1). This date is in the range of the ages that we obtained for the Jack River rhyolite unit. Other K-Ar ages for granitic plutons in areas to the north and south of the Jack River igneous field range from ca. 68 to 50 Ma (Csejtey et al., 1992).
ROCK UNITS AND STRATIGRAPHY
There is a mappable stratigraphy among the Jack River volcanic rocks (Figs. 3 302 and 5). In the eastern part of the study area the stratigraphically lowest deposits consist of up to 200 m of volcaniclastic conglomerate and sandstone containing minor interbeds of rhyolite tuff. In the western part of the study area the stratigraphically lowest deposits consist of up to 70 m of rhyolite tuff without conglomerate or sandstone deposits. All of these volcaniclastic deposits unconformably overlie the Kahiltna assemblage. Overlying the volcaniclastic deposits is a mafic-intermediate unit that consists of ∼100–300 m of basalt, basaltic andesite, and andesite lava flows. The mafic lavas are overlain by a rhyolite unit that consists of over 900 m of lava flows and pyroclastic deposits.
Lower Volcaniclastic Unit
Conglomerate and Sandstone
Volcaniclastic conglomerate and sandstone deposits form the base of the Jack River volcanic assemblage in the eastern part of the study area (Figs. 3 302 and 5). Conglomerate deposits include about equal proportions of well-organized pebble-cobble conglomerates and poorly organized cobble-boulder conglomerates. The poorly organized conglomerates are characterized by 1- to 6-m-thick tabular to lenticular beds that are poorly sorted and matrix supported and have subangular to rounded clasts that reach over 1 m in diameter (Fig. 6A). Matrix material is typically medium to coarse sand-sized. The well-organized conglomerates are characterized by 0.5–3-m-thick lenticular beds that are well to moderately sorted and clast supported and have rounded clasts and well-developed clast imbrication. Clast imbrications recorded in these conglomerates show eastward-directed paleoflow (Fig. 3) 302. Clast varieties in both types of conglomerate deposits include mostly mafic and intermediate lava and argillite, but minor amounts of felsic lava, tuff, chert, quartzite, and granite-granodiorite are present (Fig. 5). One granitic clast yields a radiometric age of ca. 64 Ma, but this age is not robust because of large Ar-loss from the sample (Table 1). Medium- to coarse-grained sandstone is found interbedded with both types of conglomerate. Sandstone beds range from 2 to 30 cm thick and are mostly tabular or drape underlying conglomerate beds. These sandstone beds are moderately to well sorted and have subangular to subrounded grains, which include quartz, plagioclase, chert, argillite, mafic volcanic lithics, and rare pumice. Overall the well-organized conglomerates are characteristic of braided fluvial deposits (Miall, 1977), and the poorly organized conglomerates are characteristic of noncohesive debris flow and/or stream flood-flow deposits (Costa, 1988).
The thickest tuff deposits (tens of m) are found in the western part of the study area and thinner tuff deposits (cm's to m's) are interbedded with conglomerate and sandstone deposits in the eastern part of the study area (Figs. 3 302 and 5). Most of the tuff deposits are vitric and crystal rich and form 0.5–5-m-thick beds. Vitric grains include 0.5–2-mm-long cuspate shards and 1-mm to 2-cm-long pumice grains; crystals include euhedral and broken quartz and plagioclase grains ranging from 1 to 2 mm in diameter. The beds are well sorted and vary from being massive (lacking internal stratification) to showing faint horizontal laminations and normal grading. We interpret these as pyroclastic fallout deposits similar to Plinianstyle deposits described by Cas and Wright (1987). A small portion of tuff beds are lithic and crystal rich and exhibit internal stratification (low-angle and horizontal laminations). These are usually found interbedded with sandstone or conglomerate and are not typically well exposed or preserved. The higher proportion of lithic and crystal grains compared with vitric grains, along with the presence of internal laminations, indicates that these are probably fluvially reworked tuff deposits in which the less dense vitric grain population was winnowed out (e.g., Cole and Ridgway, 1993).
Mafic and Intermediate Lava Unit
The basalt lavas are red brown on weathered surfaces and dark gray to black on fresh surfaces. Individual lava flows are aphanitic to porphyritic and range from tens of cm to over 3 m thick. The thicker flows display columnar jointing and tend to be vesicular in their upper parts. Based on visual estimates from thin sections, the basalts have a modal mineralogy of 5–20% olivine (some alteration to iddingsite), 10%–20% clinopyroxene, 40%–60% plagioclase (generally An60 to An70), and 5%–10% iron-titanium oxides (Fig. 6B).
The basaltic andesite and andesite lavas are red brown, gray brown, and green brown on weathered surfaces and dark gray on fresh surfaces. Lava flows are mostly porphyritic and range from ∼0.5 to over 6 m thick and show well-developed columnar jointing in the thicker flows. The upper portions of the thicker lava flows are typically vesicular and amygdaloidal. Based on visual estimates from thin sections these intermediate lavas have a modal mineralogy of 2%–5% olivine, 2%–5% clinopyroxene, 5%–10% orthopyroxene, 70%–90% plagioclase (generally An50 to An70), and 5%–10% iron-titanium oxides (Fig. 6C).
The rhyolite unit typically includes a lower 50–100-m-thick base of tuff and/or lapilli-tuff to tuff-breccia interbedded with thin lava flows overlain by hundreds of meters of massive lava flows. The tuff deposits are light gray, tan, and gray green, vitric- and pumice-rich, typically are thin-bedded, and show low-angle and horizontal laminations. These laminated tuff beds are generally 4–20 cm thick and underlie lapilli-tuff or tuff-breccia deposits. The lapilli-tuff and tuff-breccia deposits are light gray, contain cuspate vitric shards and broken euhedral crystals of quartz and plagioclase, and show lapilli to block sized clasts of pumice and porphyritic rhyolite (Fig. 6D). A small percentage of aphanitic mafic lithic grains are present. Beds are generally 30 cm to 2.5 m thick, massive, poorly sorted, and sometimes have pumice-rich upper portions. Some beds are welded. These deposits are typical of pyroclastic flow units in that the laminated tuff represents low-density basal surge deposits and the lapilli-tuff and tuff-breccia represents high-density upper flow deposits (e.g., Sparks et al., 1973). Also found within this lower interval are beds containing spindle-shaped rhyolite blocks (Fig. 6E) that were most likely formed as fallout bomb deposits.
The rhyolite lava flows are light gray, tan, light green, and red gray on weathered surfaces and are light gray on fresh surfaces. From a distance the rhyolite units can appear black in color due to the abundance of lichens that favor these rocks. Flow contacts among the rhyolite lavas are poorly defined, but individual flows are generally greater than 5 m thick. These lavas commonly exhibit flow banding and columnar jointing (Fig. 6F). The rhyolite lavas are porphyritic and contain ∼40–50% phenocrysts in a glassy groundmass. Based on visual estimates from thin sections, the phenocrysts have a modal mineralogy of 40%–60% quartz, 30%–40% alkali feldspar (orthoclase and sanidine) and 5%–10% biotite (Fig. 6G). The rhyolites also contain a trace percent of mm-to cm-sized mafic xenoliths. Interbedded with the rhyolite lavas are intervals of pyroclastic flow deposits (similar to those described above) that range from a few tens of meters to ∼100 m thick.
Granitic plutons on the scale of tens of km2 in area intrude the Kahiltna assemblage in the study area and lie adjacent to the Jack River volcanic rocks. The timing and contacts between the granitic plutons and volcanic rocks is not everywhere clear. The age dates for the granitic rocks (Table 1) and the presence of granitic clasts in the volcaniclastic conglomerate at the base of the volcanic rocks indicate that some of the plutons were emplaced prior to Jack River volcanism. In some places though, the granitic rocks clearly intrude the volcanic rocks (Fig. 3) 302. These relationships suggest to us that the granitic magmas may have begun to form before volcanism, but their emplacement occurred in phases that spanned the interval of volcanic activity. In this case, these plutons were probably emplaced incrementally over millions of years (e.g., Glazner et al., 2004), but more detailed mapping and additional age dates are necessary to test this hypothesis.
Mineralogically, the Jack River granitic plutons consist of quartz (∼20% modal, estimated visually) + plagioclase (∼40%) + orthoclase (∼35%) + biotite (5%) ± hornblende ± muscovite, and ± accessory phases (Fig. 6H). The texture of the granite is typically a framework of plagioclase crystals, and the other phases occurring in the framework's interstices. Quartz is generally an interstitial phase and sometimes forms larger crystals. Plagioclase is complexly zoned, containing obvious core and rim structures. Internal, resorbed surfaces crosscutting earlier zones are also observed in some of the crystals suggesting a complex growth and dissolution history. Perthitic orthoclase is an interstitial phase. However, some larger poikilitic crystals are also observed. Red brown to green biotite can be found as subhedral laths or as euhedral crystals that contain inclusions of zircon and apatite. The plutons also contain biotite-rich enclaves, although these are rare. Minor, altered amphibole is also present. Muscovite is found as rare anhedral crystals, which may be a reaction product. Accessory mineral phases (much less than 1% modally) include apatite, zircon, titanite, and Fe-Ti oxide. Some samples show extensive alteration of the mafic phases to chlorite and epidote.
The rock and mineral textures of the granitic plutons are spatially variable. Grain size systematically varies from larger cm-scale crystals far from the margins to finer mm-scale crystals near the margins or roof. The complexly zoned plagioclase crystals are common, suggesting a complex cooling history for the granitic magma. Even in a single thin section, plagioclase crystals sometimes show very different histories, suggesting that the crystals may have cycled throughout the magma chamber early in its history. These observations are consistent with the hypothesis that these are amalgamated plutons that had multiple phases of emplacement.
The felsic tuff deposits in the volcaniclastic unit at the base of the Jack River volcanic rocks reveal that Jack River volcanism began with small explosive eruptions that shed pyroclastic fallout deposits across the study area. The tuff deposits are thickest in the western portion of the study area, which indicates that vents were located in this region, but no near-vent pyroclastic deposits were found to confirm this relationship. Mafic and intermediate lava flows followed the small felsic eruptions and inundated the study area with up to 300 m of lava. The absence of interbedded pyroclastic material and the widespread extent of the lavas suggest that this was a period of nonexplosive fissure eruptions.
The volcaniclastic conglomerate and sandstone deposits reveal that portions of the volcanic field were being eroded and contributed detritus to drainages at the onset of volcanism. The conglomerate clast compositions indicate that the source rocks contained mafic and intermediate lavas and argillite (Kahiltna assemblage). Paleoflow data indicate that the source rocks were located in the western part of the study area. This is consistent with our radiometric age dates, which show that the mafic and intermediate lavas are older at the west end of the volcanic field. This is also consistent with the distribution of the fluvial deposits; they are present in the eastern part of the volcanic field and are absent in the western part (Figs. 3 302 and 5). These observations indicate that mafic and intermediate lavas began to erupt in the western portion of the volcanic field. The lavas and underlying Kahiltna assemblage (perhaps uplifted by thermal doming) then served as a sediment source for synvolcanic drainages flowing toward the east. Eventually, the mafic and intermediate lavas became widespread across the volcanic field and covered the volcaniclastic deposits. The presence of granitic clasts (up to boulder-sized) mixed with the basalt clasts indicates that nearby plutons had been uplifted and exhumed by the onset of Jack River volcanism.
The Jack River rhyolite unit represents a major eruptive episode of alternating explosive and effusive phases. The pyroclastic flow and surge deposits at the base of the rhyolite unit indicate that this episode began with explosive eruptions that generated eruption columns. The presence of bomb beds within the lower portion of the rhyolite unit (e.g., Fig. 6E) is consistent with an initial phase of explosive activity. This phase was followed by effusion of hundreds of meters of rhyolite lava. Overall there were multiple episodes of alternating explosive and effusive eruptive phases as recorded by pyroclastic intervals within the lavas. The explosive phases could have occurred in response to repeated cycles of volatile buildup and subsequent degassing of the rhyolite parent magma, each of which was followed by lava flow eruptions. We envision a multivent polygenetic eruptive complex for the rhyolite unit, but the lack of definitive vent facies or dome complexes precludes an interpretation of the precise location of rhyolite vent(s).
GEOCHEMISTRY AND PETROGENESIS
Major and Trace Element Data
Major and trace element data were collected on 49 volcanic and plutonic rock samples, two quartz vein samples, and one Kahiltna assemblage sandstone sample from the Jack River igneous field (Table 2) 202203. Samples were powdered using an alumina ceramic mixer mill at Allegheny College. Major element oxides were determined by X-ray fluorescence, and trace element concentrations were determined by inductively coupled plasma mass spectrometry at ALS-Chemex Labs, Inc., Vancouver. Because of the relatively high loss on ignition (>3%) for some of the samples (Table 2) 202203, the major oxide data presented in this paper are normalized to 100% volatile free (e.g., Johnson et al., 1999).
The volcanic rocks are subalkaline and display a bimodal distribution with respect to silica (Fig. 7). One end-member includes basalt, basaltic andesite, and andesite, and the other end-member includes rhyolite; there is a distinct absence of dacitic volcanic samples. The granitic plutons do, however, fall in the dacitic range of SiO2 (Fig. 7). The basalts are tholeiitic on the basis of FeOt/MgO ratios (Fig. 7), have relatively low to moderate Mg#'s ranging from 0.24 to 0.44 (Table 2) 202203, and range from oceanic within-plate to island arc affinity on various tectonic discrimination diagrams (Fig. 8). The basalts have rare earth element patterns as well as high field strength element (HFSE) ratios, such as Nb/Zr, that are in the range between Pacific enriched mid-ocean-ridge basalt (E-MORB) and Hawaiian ocean island basalt (OIB; Figs. 7, 9 and 10). When normalized to primitive mantle, the basalt samples display trace element patterns with small enrichment peaks in some large-ion-lithophile elements (LILE) including Cs, Ba, K, and Pb (Fig. 11). The basalts do not, however, show a strong decoupling of the LILE from the HFSE (e.g., the basalts have low LILE/HFSE ratios, such as Ba/Nb and Pb/Nb; Figs. 7 and 11). The basalts also have low ratios between the more fluid mobile and the less fluid mobile HFSE (e.g., U/Nb) (Fig. 7).
The Jack River basaltic andesites and andesites range from tholeiitic to calc-alkaline on the basis of FeOt/MgO ratios (Fig. 7) and the basaltic andesites, similarly to the basalts, range from oceanic within-plate to island arc affinity on various tectonic discrimination diagrams (Fig. 8). The basaltic andesites and andesites display normalized trace element patterns that also are similar to the basalts but with greater enrichment of Cs, Ba, K, and Pb and a small but pronounced Ta-Nb depletion pattern (Fig. 11). Geographically, the basalts, basaltic andesites, and andesites are least geochemically evolved in the west and are more geochemically evolved progressively east across the volcanic field. For example, there is a west-to-east increase in La/Yb, K2O, and Rb and a decrease in MgO among the basalt and intermediate samples (Fig. 12).
The Jack River rhyolites have a high percentage of K2O and range between the high-K calc-alkaline and shoshonitic series (Fig. 7). The rhyolites show a relatively wide range of light rare earth element (LREE) to heavy rare earth element (HREE) ratios (e.g., La/Yb) (Figs. 7 and 10) and exhibit negative Euanomalies (Fig. 10). The rhyolites also show a wide range in ratios between HFSE (e.g., Nb/Zr) (Fig. 7). When normalized to primitive mantle, the rhyolites are generally enriched in the LILE (e.g., Cs, Rb, K, and Pb) and in the more incompatible HFSE (e.g., Th, U), are strongly depleted in Ba, Sr, P, and Ti, and display a paired Ta-Nb depletion pattern (Fig. 11).
The Jack River granitic rocks range from granodiorite to granite and are peraluminous (Fig. 13), with up to ∼3% normative corundum. The granitic rocks are transitional between volcanic arc and collisional granites (Fig. 13). Although the granitic samples do fall along the major oxide trends of the volcanic samples (Fig. 7), trace element data reveal that there is not a simple comagmatic relationship between the granitic and volcanic rocks. Most distinctively, the granitic rocks are more depleted in the heavy rare earth elements (HREE), resulting in much higher La/Yb ratios than found in the volcanic rocks (Figs. 7 and 10). As a group, the granitic rocks are distinctive because they exhibit characteristics of adakites. Adakites are characterized as having SiO2 ≥56%, Al2O3 ≥ 15%, MgO <3%, Yb ≤1.9 ppm, Y <18 ppm, Sr > 400 ppm, Sr/Y > 20–40, an absence of a negative Euanomaly, and a high percentage of modal plagioclase (Kay, 1978; Defant and Drummond, 1990). The granitic samples meet most of these criteria (Table 2 202203; Figs. 10 and 14) and are geochemically similar to adakites that formed along the collisional orogen of southern Tibet as described by Chung et al. (2003; Figs. 10 and 11). In addition to the granitic rocks, one sample of rhyolite lava (TSU-00-14) and a rhyolite dike sample (TSU-00-1) are also relatively depleted in the HREE and exhibit REE patterns that are steeper than the remaining rhyolite samples (Fig. 10). Sample TSU-00-14 is from a small exposure of rhyolite lava that overlies the granitic pluton in the southern part of the Jack River igneous field, and TSU-00-1 is representative of a set of northeast-trending rhyolite dikes that crosscut the igneous field (Fig. 3) 302. These two samples do not meet the criteria for adakites but geochemically are transitional between the granitic adakites and the remaining rhyolites.
Mantle Source Characteristics
Trace element characteristics of the basalts (Figs. 7–11) indicate that the composition of their mantle source was in the range between Pacific enriched mid-ocean-ridge basalt (E-MORB) and Hawaiian ocean island basalt (OIB)-type mantle end-members. The basalts also have strontium isotope compositions in the range between Pacific E-MORB and Hawaiian OIB with 87Sr/86Sr(i) ratios between 0.7033 and 0.7049 (Cole and Stewart, 2005). This range of mantle source composition is consistent with results of Reiners et al. (1996), who interpret an E-MORB type mantle source for a series of 64–65 Ma composite plutons in the Alaska Range. These plutons were emplaced through the Kahiltna assemblage within the Alaska Range suture zone ∼100 km west of the Jack River igneous field (Fig. 1). The basis for their interpretation is the presence of hornblende-gabbro xenoliths that have 87Sr/86 Sr(i) values as low as 0.7034 and ϵNd values as high as +6.2. The study of Reiners et al. (1996) is important because it is the first comprehensive analysis of potential mantle reservoirs beneath the Alaska Range suture zone. Our interpretations for mantle source composition(s) of the Jack River volcanic rocks are more limited because none of the Jack River basalt samples can be considered to represent a primary magma (all have low Mg#'s, low Cr, and low Ni compared to primary basalts; e.g., Wilson, 1989, p. 22). Nevertheless, the Nd and Sr isotopic data of the hornblende-gabbro xenoliths reported by Reiners et al. (1996), the Sr isotopic data of the Jack River basalts reported by Cole and Stewart (2005), and the trace element affinity of the Jack River basalts reported in this paper all indicate that the Paleogene mantle beneath the Alaska Range suture zone was more enriched than depleted MORB mantle and ranged between Pacific E-MORB and Hawaiian OIB-type mantle compositions (toward the composition of EMI-type mantle).
The slight enrichments of some large-ion-lithophile elements (LILE; e.g., Cs, Ba, K, and Pb; Fig. 11) among the Jack River basalts indicates metasomatic contamination of the basaltic magmas and/or their mantle source. Although enrichment in LILE can be a typical characteristic of arc magmatism whereby these elements are released during dehydration of the subducted slab (Pearce and Parkinson, 1993; Arculus, 1994; Davidson, 1996; Hochstaedter et al., 2001), the Jack River basalts differ from arc basalts in that they do not exhibit strong decoupling of large ion lithophile elements (LILE) from high field strength elements (HFSE). For example, the Jack River basalts have lower LILE/HFSE ratios (e.g., Ba/Nb, Pb/Nb) than basalts of the modern Aleutian arc (Fig. 7), and they lack the paired Ta-Nb depletion trend that is typical of arc rocks (Fig. 11). The basalts also have lower ratios between the more fluid mobile and the less fluid mobile HFSE (e.g., U/Nb) than Aleutian arc basalts (Fig. 7). Furthermore, the Jack River basalts have TiO2 contents above 1.5%, which is higher than is typical for arc basalts (Fig. 7; Gill, 1981). These observations suggest to us that the Jack River basaltic magmas were not formed above an actively subducting and dehydrating slab. This is consistent with the timing of Jack River volcanism (≤56 Ma), which coincided with the end of the 75–56 Ma episode of subduction-related Alaska Range–Talkeetna Mountains magmatism as defined by Moll-Stalcup (1994). The source for Jack River basaltic magmas may therefore have been an enriched mantle component that was distinctive from what we would predict for a metasomatic mantle reservoir that formed during plate subduction that preceded Jack River volcanism. In this case, the Jack River basaltic magmas, or their source, could have attained their LILE enrichment by contamination from crustal rocks in the Alaska Range suture zone, which include Triassic through Cretaceous sedimentary and igneous rocks with elevated LILE contents (Fig. 11), and/or by mixing with the metasomatized mantle that was formed during preceding plate subduction. This is analogous to other cases of postcollisional volcanism that occurred in settings where there was precollisional subduction (Johnson et al., 1978; Coulon et al., 2002; Ilbeyli et al., 2004; Wang et al., 2004).
Petrogenesis of Intermediate and Felsic Magmas
We applied assimilation-fractional crystallization (AFC) models using the REE data from the Jack River igneous field to test whether or not the basaltic magmas could have been parental to the intermediate and felsic Jack River rocks. Without the inclusion of a wider range of elements and isotopic data, these models are not definitive, but they do illustrate the range and minimum complexity of the data and show a permissible petrogenesis for the Jack River rocks. In the AFC models, we chose basalt sample NTM1-288 as the parent composition because this sample has the highest weight percent MgO and the lowest LREE enrichment among the basalt samples. We chose two possible assimilants in these models representing the two known major crustal components in the Alaska Range suture zone. One assimilant is the Kahiltna assemblage represented by an average geochemical composition compiled from data reported by Lanphere and Reed (1985), Reiners et al. (1996), and this study (Table 2) 202203. The other assimilant is an average composition for Triassic sedimentary rocks and basalts (Table 2) 202203.
Using sample NTM1–288 as the parent composition and the Triassic rocks as the assimilant does not provide an AFC model that is consistent with the Jack River intermediate and felsic samples (Fig. 15). But using the same parent and the Kahiltna assemblage as the assimilant, the AFC models indicate that a wide range and high rates of assimilation to fractional crystallization (r-values of 0.5–0.9) can satisfy the range of the more evolved basalts and some of the intermediate samples (Fig. 15). These models do not account for the evolution of those basaltic andesites and andesites that have relatively high LREE enrichment (i.e., those samples on Fig. 15 with Ce/Yb ratios above 15) or for the rhyolites or granites.
The basaltic andesite and andesite samples that have higher Ce/Yb ratios may represent metasomatic contamination of the basalt parent magma, thereby producing a range of parental magma compositions. For example, mixing between basalt sample NTM1–288 and a metasomatic melt can produce the Ce/Yb ratios found in several basaltic andesite and andesite samples, including NTM3–88, which is one of the least evolved among the high Ce/Yb samples (Fig. 15). The metasomatic melt could have been a product of subduction-related processes that preceded Jack River magmatism. AFC modeling shows that basaltic andesite sample NTM3–88 could have been a parent composition for most of the remaining high Ce/Yb intermediate samples with either the Kahiltna assemblage or Triassic rocks as an assimilant (Fig. 15). With the Kahiltna assemblage as the assimilant, a wide range in the rate of assimilation to fractional crystallization (r-values ranging between 0.1–0.9) is required to satisfy the range of high Ce/Yb intermediate samples. With Triassic rocks as the assimilant, a more narrow range and lower rates of assimilation to fractional crystallization (r-values ranging between 0.1–0.4) can satisfy the range of high Ce/Yb inter-mediate samples. Although these models provide a possible range for evolution of the high Ce/Yb intermediate samples, as stated earlier, the models are not definitive. For example, sample NTM3–4 is shown on liquid lines of descent from parent sample NTM3–88 (Fig. 15), but sample NTM3–4 has a lower content of large ion lithophile elements than NTM3–88 and could not have evolved by simple AFC from this parent. Nevertheless, these models do demonstrate that assimilation of Kahiltna assemblage rocks, and possibly Triassic rocks, in the Alaska Range suture zone was important in the evolution of the Jack River volcanic rocks.
AFC models by Reiners et al. (1996) for the Alaska Range composite plutons also predict assimilation of Kahiltna assemblage rocks and a wide range in r-values, similar to our AFC models described above. Their AFC model involved two stages. During the first stage, mantle-derived mafic magmas evolved with a high ratio of assimilation to fractional crystallization (r ≥ 0.9) to form magmas of intermediate composition. A high degree of assimilation occurred during this early stage when the higher temperature mafic magmas were capable of assimilating larger amounts of crustal rocks. The second stage involved mafic to intermediate magmas evolving with a low ratio of assimilation to fractional crystallization (r = 0.3) to form more felsic magmas. Our AFC models cannot define similar stages of magma evolution, but they do provide a range of possible conditions for Jack River magma evolution. In particular, our AFC models show that a basaltic magma could have been parental to some of the basaltic andesites and andesites but that a more evolved magma, possibly formed by contamination of a basalt parent magma with metasomatic fluids, was parental to those intermediate rocks that have relatively high LREE enrichment. In addition, the range of rhyolite compositions is not adequately modeled from basalt or basaltic andesite parent compositions; for example, the rhyolites that show the lowest and highest Ce/Yb ratios do not fall along any reasonable AFC modeling trends that we attempted using basalt, basaltic andesite, and andesite parent compositions. The AFC models also highlight the difference in evolution of the granitic rocks from the volcanic rocks; the granitic rocks have a much higher range of Ce/Yb than would result by AFC evolution from any of the Jack River basalt or intermediate magmas. In each case, described separately on the following page, the rhyolites and granites could have formed by partial melting of crustal rocks without mafic or intermediate parent magmas.
Equilibrium partial melt models using REE compositions show that most of the Jack River rhyolite samples could have formed by ∼20–60% melting of Kahiltna assemblage rocks (Fig. 16). This hypothesis, as opposed to the rhyolites' having formed by continuous assimilation-fractional crystallization from mafic and intermediate parent magmas, is consistent with the absence of dacites in the Jack River volcanic field (Fig. 7). If continuous AFC had occurred, we would expect a more continuous range in volcanic rock compositions. Partial melting of the Kahiltna assemblage could have occurred during mica-dehydration reactions, which can induce large melt fractions at relatively low temperatures (e.g., 650–750 °C at depths between 15 and 40 km; Thompson, 1982) as demonstrated in the evolution of the Alaska Range composite plutons (Reiners et al., 1996). Mica dehydration reactions should yield large amounts of alkalis, which would be consistent with the high K2O content of the Jack River rhyolites. Knesel and Davidson (2002) also demonstrate that dehydration and fluid-fluxed melting of metapelitic crustal rocks was important in collisional bimodal magmatism of the Himalayan orogen. The negative Euanomalies as well as depletions of Ba, Sr, P, and Ti among the rhyolites (Figs. 10 and 11) suggest that the partial melts of the Kahiltna assemblage underwent fractional crystallization (e.g., hornblende [Ba], plagioclase [Eu, Sr], apatite [P], and Fe-Ti oxides [Ti]; Thompson et al., 1984; Wilson, 1989) before they were erupted to form the rhyolites. Two rhyolite samples (TSU-00-1 and TSU-00-14) have higher degrees of HREE depletion than the remaining rhyolites (Fig. 10). These samples are from a dike (TSU-00-1) and from a small exposure of rhyolite lava that overlies the granitic pluton in the southern part of the study area (TSU-00-14). These field relationships imply that these are relatively young rhyolite samples; they may have formed by mixing between the rhyolite and granitic magmas.
The Jack River granitic samples exhibit greater HREE depletion (higher Ce/Yb ratios) than can be explained by the models that apply to the rhyolites (Fig. 16). Instead, the HREE depletion as well as the other adakite-like characteristics among the granitic samples (Figs. 11 and 14) indicate that the granitic magmas formed by partial melting of garnet-bearing and/or amphibole-bearing rocks (Kay, 1978; Drummond and Defant, 1990). Adakites are commonly associated with the partial melting of subducted oceanic lithosphere (Defant and Drummond, 1990; Yogodzinski et al., 2001) but adakite formation has also been attributed to partial melting of oceanic terranes that may form the upper plate of a subduction zone (Arculus et al., 1999) and to partial melting of eclogite or garnet-bearing crustal rocks along a collisional orogen (Chung et al., 2003). We propose, similarly to Chung et al. (2003), that the Jack River granitic adakites were formed by melting of garnet-bearing crustal rocks in the Alaska Range suture zone. The crustal rocks in the Alaska Range suture zone likely include several km of Kahiltna assemblage rocks that overlie late Paleozoic to Triassic metabasalt and metasedimentary rocks of the Wrangellia composite terrane (Clautice et al., 2001; Ridgway et al., 2002). In addition, Davidson et al. (1992) document kyanite-garnet grade metamorphism of Kahiltna assemblage rocks in the Maclaren Glacier metamorphic belt, located southeast of the Jack River igneous field (Fig. 2). This metamorphic belt dips to the northwest (Nokleberg et al., 1985; Davidson et al., 1992; Ridgway et al., 2002), toward the Jack River igneous field. The presence of the Jack River adakitic granites suggests to us that garnet-bearing meta-Kahiltna assemblage rocks similar to those of the Maclaren Glacier metamorphic belt might extend beneath the Jack River igneous field. Such garnet-bearing argillaceous rocks are a viable crustal component to form the Jack River adakitic granites. Alternatively, if thickened meta-Kahiltna rocks are not present beneath the Jack River igneous field, then an alternative source for the adakitic granites could have been partial melts of pre-Triassic(?) meta-argillaceous rocks; such pre-Triassic rocks are not yet well documented but might exist based upon the presence of argillaceous roof pendants on Late Cretaceous to Paleocene plutons east of the Talkeetna Mountains and outcrops of pre-Triassic(?) aged rocks south of the Talkeetna thrust fault in the central Talkeetna Mountains (K.D. Ridgway, personal commun., 2006). In either case, the Jack River adakitic granites indicate that garnet-bearing crustal rocks likely underlie the northern Talkeetna Mountains and that these rocks underwent partial melting during Paleocene to Eocene time.
The Jack River volcanic field is deformed into a broad, northeast-trending syncline (Fig. 3) 302. Dips of the volcanic strata are relatively shallow on each limb of the syncline, ranging from ∼15 to 30 degrees. Faults that crosscut the volcanic rocks include northwest- and west-trending normal-slip and oblique-slip faults and a southeast-dipping reverse fault, which we refer to as the Meadow Creek fault, located at the east end of the volcanic field (Figs. 3 302 and 17). The volcanic rocks and underlying Kahiltna assemblage are also crosscut locally by west-northwest–trending felsic dikes (Fig. 3) 302. These dikes are up to 10 m wide, are subparallel to the strike of the normal and oblique slip faults, and most likely formed under the same stress field as these faults.
The normal and oblique slip faults have small stratigraphic separations on the order of tens of meters to a couple hundred meters, offsetting mappable volcanic units. Fault surfaces contain mineral fibers and grooves indicating frictional slip at shallow crustal levels. The Meadow Creek fault has a larger stratigraphic separation (at least 100s of m) and places rocks of the Kahiltna assemblage over the volcanic rocks (Fig. 3) 302. The Meadow Creek fault is a 50-cm- to 2-m-thick breccia zone containing pervasive quartz mineralization (Fig. 17A). The fault breccia includes mostly clasts of Kahiltna assemblage rocks and a minor amount of volcanic rocks in a comminuted matrix of the same material. The matrix also contains pervasive black, opaque, glassy material (pseudotachylite). This fault zone was apparently a pathway for hydrothermal fluids that yielded relatively high metal concentrations in the fault zone and up to several meters away in the surrounding rocks. For example, a 10- to 20-cm-wide quartz vein cutting the footwall basalt unit within ∼3 m of the Meadow Creek fault contains relatively high concentrations of Ag, Cu, and Ga (28.9, 440, and 5.35 ppm, respectively; Table 2 202203).
The amount of stratigraphic separation decreases northward along the Meadow Creek fault; at the south end, rocks of the Kahiltna assemblage are emplaced over volcanic conglomerate from the base of the Jack River volcanic rocks, and at the north end, Kahiltna is emplaced over Kahiltna (Fig. 3) 302. This indicates fault motion with greater separation along the southern portion of the fault. The nature of crosscutting relationships of the Meadow Creek fault and the granitic plutons at either end of the mapped fault are uncertain. The fault clearly crosscuts the Jack River volcanic rocks. A date of 54.6 Ma on biotite from the northern granitic pluton (Table 1) shows that the pluton is similar in age to the volcanic rocks, and so we would expect the fault to also crosscut the pluton. We did not, however, locate this fault zone in the pluton, implying that either we didn't find the trace or that the fault didn't propagate through the pluton. The decreased stratigraphic separation along the fault toward the north suggests the latter, whereby the granitic pluton may have been a buttress to fault motion. The exposure of the fault where it approaches the southern granitic pluton was covered by snow fields during three consecutive field seasons, and so this fault-granite crosscutting relationship also remains enigmatic.
Collectively, the trends of the regional fold, faults, and felsic dikes indicate an episode of postvolcanic (<50 Ma) northwest-southeast shortening. A similar direction of shortening recorded by Late Cretaceous and Paleogene structures is reported for the Chulitna district west of the volcanic field (Clautice et al., 2001), in the Cantwell basin northwest of the volcanic field (Cole et al., 1999), in the Reindeer Hills mélange unit north of the volcanic field (Cole et al., 2002; Rothfuss et al., 2006), and along the Talkeetna fault southeast of the volcanic field (O'Neill et al., 2003; Fig. 2).
The Jack River igneous rocks provide a link between terrane accretion, the end of a phase of subduction-related magmatism, and strike-slip faulting in south-central Alaska. Jack River magmatism began at ca. 62.7 Ma with emplacement of granitic plutons. This coincided with the culmination of Wrangellia composite terrane accretion as recorded by uplift of metamorphosed Kahiltna assemblage rocks in the Maclaren glacier area by 62 Ma (Ridgway et al., 2002) and by the end of north-south shortening in the Cantwell basin by ca. 60 Ma (Cole et al., 1999). These geologic constraints for the culmination of Wrangellia composite terrane accretion are consistent with available paleomagnetic data, which show that the Jack River volcanic rocks and volcanic rocks in the central and southern Talkeetna Mountains that crosscut the Wrangellia composite terrane have not experienced any significant latitudinal drift with respect to North America (Hillhouse et al., 1985; Panuska et al., 1990). This timing is important because it establishes that Jack River magmatism began during the terminal stage of the terrane accretion and continued as a post-collisional magmatic event.
Another important aspect about the timing of Jack River magmatism is that it began at about the time when the 75–56 Ma episode of Alaska Range–Talkeetna Mountains arc magmatism was ending (Moll-Stalcup, 1994). The Jack River igneous rocks show geochemical characteristics that are inconsistent with typical arc rocks, confirming that Jack River magmatism was not simply a continuation of preceding arc magmatism. These characteristics include: (1) the Jack River volcanic rocks are a bimodal suite (there is an absence of dacite; Fig. 7); (2) there is not a comagmatic relationship (i.e., liquid line of descent) between the mafic rocks and the rhyolites (Figs. 7 and 15); (3) the Jack River basalts and andesites have higher TiO2 than is typical for arc basalts and andesites (Fig. 7); (4) the basalts do not exhibit a strong depletion of high field strength elements (e.g., Ta and Nb) with respect to large ion lithophile elements (Figs. 7 and 11); (5) the basalts have some trace element characteristics of intra-plate magmas (Fig. 8); and (6) the granitic rocks are adakitic (Figs. 10, 11, and 14). Because Jack River volcanism began at the time when 75–56 Ma arc magmatism in south-central Alaska was ending (Moll-Stalcup, 1994) and because the Jack River basalts lack several key geochemical characteristics of arc basalts, we infer that these rocks did not form by active subduction-related processes. This makes sense given that the Wrangellia composite terrane was accreted and sutured to southern Alaska by ca. 60 Ma; following the terrane accretion, active arc magmatism ended in central Alaska and resumed farther south during Oligocene time as the Alaska-Aleutian arc system (Scholl et al., 1986; Moll-Stalcup, 1994). The Jack River igneous rocks, therefore, signify an episode of increased heat flow and renewed magmatism along the suture zone of the Wrangellia composite terrane even while regional arc magmatism was waning.
In other cases of postcollisional magmatism (e.g., in Tibet, Taiwan, Turkey, and the western Mediterranean region), the heat flow required to generate magmas occurred in response to convective removal of the lower lithospheric mantle and/or by slab break-off (Turner et al., 1996; Miller et al., 1999; Coulon et al., 2002; Chung et al., 2003; Ilbeyli et al., 2004; Wang et al., 2004). Both of these processes would allow upwelling of asthenospheric mantle, raising the geothermal gradient beneath the collisional suture zone. Postcollisional magmas could then form by partial melting of upwelling asthenospheric mantle, remaining lithospheric mantle, and/or crustal rocks (e.g., Turner et al., 1996; Coulon et al., 2002; Chung et al., 2003).
Another possible mechanism for increased heat flow beneath southern Alaska during Paleogene time was spreading ridge subduction (Bradley et al., 2003). A series of near-trench plutons (the Sanak-Baranof magmatic belt; Fig. 1) shows an eastward-younging age progression that, together with other geologic evidence, records the migration of a ridge-trench-trench triple junction between ca. 62 and 50 Ma (Bradley et al., 2003). Formation of a slab window in response to spreading ridge subduction would allow asthenospheric upwelling (Thorkelson, 1996) and subsequent magmatism along the accretionary prism as well as farther inboard as documented for Paleogene volcanism and intrusions in southern Alaska (Lytwyn et al., 1997; Harris et al., 1996; Cole et al., 2006), Cenozoic magmatism in the Pacific Northwest (Bretisprecher et al., 2003), Cenozoic volcanism in western California (Cole and Basu, 1995), Neogene volcanism in southern Patagonia (Gorring and Kay, 2001), and Cretaceous magmatism in Japan (Kinoshita, 2002). Interestingly, a slab window due to spreading ridge subduction could have induced high heat flow and magmatism beneath the Wrangellia composite terrane and its suture zone, independently of collisional tectonic processes.
Although the timing of ridge subduction along southern Alaska is consistent with Jack River magmatism, the geochemical data for the Jack River igneous rocks do not conclusively support or refute a spreading ridge subduction model. In the case of a subducted spreading ridge, we would predict a depleted suboceanic mantle source (depleted MORB mantle) for basaltic magmas, as was the case for the southern and central Talkeetna Mountains volcanic field (Cole et al., 2006). The Jack River basalts are more enriched in incompatible elements than we would predict for basalts that formed from a depleted MORB mantle source. Instead, the Jack River basalts are consistent with a mantle source in the range between enriched mid-ocean-ridge basalt and ocean island basalt mantle sources (Figs. 9 and 10). If the slab window that formed during spreading ridge subduction along southern Alaska did extend northward beneath the Alaska Range suture zone, then there is not a direct geochemical signature of its depleted MORB mantle recorded in the Jack River basalts. Perhaps in this case, far inboard from the trench, a slab window would have served more as a heat source to induce melting of the lithospheric mantle that was present beneath the Wrangellia composite terrane and its suture zone, or maybe there was mixing between that lithospheric mantle and the mantle component that upwelled through the slab window. In addition, the spatial relationship between the Jack River igneous rocks and a potential slab window is complicated by Late Cretaceous to Cenozoic margin-parallel transport of terranes along strike-slip faults and counterclockwise rotation of western Alaska, both of which cause uncertainties in the location of the triple junction with respect to inboard areas (see Bradley et al., 2003, for a summary of these uncertainties). We cannot strongly argue here for or against the influence of spreading ridge subduction on Jack River magmatism, but we can argue for Paleogene high heat flow beneath the Alaska Range suture zone, which is consistent with a slab window model.
Each of the processes described above, whether lithospheric mantle delamination or slab break-off related to Wrangellia composite terrane accretion and/or spreading ridge subduction, could have resulted in increased heat flow beneath the terrane suture zone. The Jack River adakitic plutons reveal that this episode of increased heat flow beneath the suture zone began at ca. 62–63 Ma. Emplacement of the granitic adakites was followed at ca. 56 Ma by basaltic volcanism and minor felsic explosive eruptions. The basaltic magmas then evolved by a combination of assimilation of Kahiltna assemblage rocks and fractional crystallization to form intermediate magmas (basaltic andesites and andesites).
The Jack River rhyolites were formed after the basaltic and intermediate magmas, largely as melts of the Kahiltna assemblage. Interestingly, the Jack River rhyolites are geochemically similar to the 55.9–57.9 Ma McKinley sequence granitic plutons of the Alaska Range (Figs. 7 and 10). Lanphere and Reed (1985) emphasized the importance of partial melts of the Kahiltna assemblage in the origin of the McKinley plutons and interpret them to have formed by mixing of ∼10–30% mantle-derived mafic magma with up to 70–90% Kahiltna assemblage rocks. We concur that the Kahiltna assemblage was important in the petrogenesis of felsic magmas in the Alaska Range suture zone but propose that anatectic melts of the Kahiltna (instead of mixing between mafic parent magmas and partial melts of the Kahiltna) was more important in the formation of the Jack River rhyolites and perhaps also for the McKinley sequence granites. This is similar to petrogenetic models proposed for various examples of rift-related volcanism (e.g., Ankeny et al., 1986; Davies and Macdonald, 1987; Glazner, 1990; Verma, 2001). In these cases, mantle-derived basaltic magmas intrude and become “ponded” in the crust. This raises the local geothermal gradient and produces melts of overlying crustal rocks to form felsic magmas. The felsic magmas have lower density and reside over the basaltic magmas to form a zoned magma system; little mixing occurs between the basaltic and felsic magmas. After the felsic magmas are formed, they can rise rapidly and feed silicic magma to the surface, whereas the more dense basaltic magmas are only occasionally tapped.Thistype of model satisfies the bimodality of the Jack River volcanic rocks and the absence of a clear liquid line of descent between mafic-intermediate and felsic rocks, and it explains why mafic magmatism ended with the start of rhyolite magmatism.
Following Jack River volcanism was an episode of northwest-southeast shortening across the northern Talkeetna Mountains that resulted in regional folding and small-scale faulting of the volcanic rocks (Figs. 3 302 and 17). The northwest-southeast–trending rhyolite dikes that crosscut the volcanic rocks (Fig. 3) 302 are consistent with this episode of deformation and indicate that deformation occurred while there was still a source of felsic magmas. Geochemically, these dikes are transitional between the rhyolite lavas and adakitic granites and may represent mixing between the two magmas. The adakitic granitic pluton on the north side of the volcanic field yields a radiometric age of 54.6 Ma and in places crosscuts the volcanic rocks (Table 1; Fig. 3 302), confirming that adakitic magmas continued to form throughout Jack River volcanism. The pattern of deformation of the Jack River volcanic rocks is similar to the deformation of 60–55 Ma volcanic rocks in the Cantwell basin along the north side of the Denali fault (Cole et al., 1999; Fig. 18). Both volcanic fields record northwest-southeast shortening that was accommodated by regional folding and small-scale brittle faulting. These patterns of deformation are consistent with deformation in a zone of right-lateral simple shear along a fault having the same orientation as the McKinley strand of the Denali fault system (Fig. 18). We propose, therefore, that deformation of the Jack River volcanic field was a result of right-lateral strike slip along the Denali fault. Our youngest age date for the Jack River volcanic rocks (49.5 Ma) reveals that this deformation occurred after middle Eocene time. This is consistent with the offset history of the McKinley strand of the Denali fault, which is interpreted to have had ∼40 km of post-Paleocene right-lateral offset (Reed and Lanphere, 1974; Brewer, 1982; Cole, 1999; Trop et al., 2004).
Also taking place during this time frame was ∼44° ± 11° counterclockwise rotation of western Alaska (Hillhouse and Coe, 1994), which formed the curvature of regional structures in south-central Alaska (Fig. 1), known as the southern Alaska orocline. The Jack River igneous field lies in the hinge area of the orocline (even after restoring 40 km of right-lateral slip along the McKinley strand of the Denali fault) along with volcanic fields in the central and southern Talkeetna Mountains (Fig. 1). The combined timing of strike-slip faulting and oroclinal bending is consistent with the timing of Jack River magmatism, suggesting that these regional tectonic events influenced the timing and locus of magmatism. For example, if crustal extension occurred in the orocline hinge area, then the hinge area may have been a unique pathway for Paleogene magmatism in south central Alaska (Cole et al., 2006). This hypothesis of crustal extension is consistent with our petrogenetic model for the Jack River rhyolites and explains the alignment of Paleogene igneous rocks in southern Alaska along a zone that trends orthogonally to the continental margin (e.g., volcanic rocks of the southern, central, and northern Talkeetna Mountains; Fig. 1).
The Jack River volcanic and granitic rocks were erupted and emplaced in the suture zone between the accreted Wrangellia composite terrane and the former margin of southern Alaska. Jack River magmatism began during the terminal stage of terrane accretion (ca. 62.7 Ma) and continued after accretion to at least 49.5 Ma. Geochemistry of the Jack River basalts indicates that their source was in the range between Pacific enriched mid-ocean-ridge basalt and Hawaiian ocean island basalt type mantle compositions. The basalts evolved to form intermediate magmas by open system processes including fractional crystallization and assimilation of argillaceous sedimentary rocks (Kahiltna assemblage) that comprise a portion of the upper crustal rocks in the suture zone. Rhyolite magmas were formed later as anatectic melts of Kahiltna assemblage rocks.
The Jack River igneous rocks show some enrichment in large ion lithophile elements but have several geochemical characteristics that differ from typical arc rocks: (1) they are a bimodal volcanic suite; (2) the rhyolites are not comagmatic with the basaltic and intermediate rocks; (3) the basalts and andesites have higher TiO2 than is typical for arc basalts and andesites; (4) the basalts do not exhibit a strong depletion of high field strength elements (e.g., Ta and Nb) with respect to large ion lithophile elements; (5) the basalts have an intraplate geochemical affinity; and (6) adakites are present. These characteristics indicate that Jack River magmas were not formed above an actively dehydrating subducted slab and are consistent with the interpretation of Moll-Stalcup (1994) that Late Paleocene to Early Eocene igneous rocks in western and central Alaska represent a transition from subduction-related calc-alkaline magmatism to magmatism that is more typical of intraplate tectonic settings.
A viable source for the Jack River granitic adakites was partial melting of garnet-bearing metamorphosed Kahiltna assemblage rocks (or older meta-argillaceous rocks) that form part of the crustal rocks in the terrane suture zone. The occurrence of adakites, as well as the subsequent onset of Jack River volcanism (even after regional calc-alkaline magmatism had ended), indicates a Paleocene to Eocene episode of high heat flow beneath the suture zone. High heat flow could have occurred by several processes related to terrane accretion and/or regional tectonic events including delamination of lithospheric mantle, slab break-off, and/or spreading ridge subduction. In each of these processes, upwelling of asthenospheric mantle would occur, raising the geothermal gradient beneath the suture zone.
The Jack River volcanic rocks experienced northwest-southeast shortening that was accommodated by a broad, regional fold and small-scale brittle faulting. Hydrothermal mineralization was concentrated along the brittle faults. The pattern of deformation is consistent with right-lateral simple shear along the Denali fault system and indicates an episode of post–49.5 Ma strike-slip along the McKinley strand of the Denali fault.
Accretion of tectonostratigraphic terranes along the North American Cordillera, as well as worldwide (e.g., Debiche et al., 1987), typically may involve the closure of marine basins between the terrane and former continental margin. The sedimentary deposits in these basins are important for magmatic processes because they provide a potential source of argillaceous rocks for upper crustal anatectic melts to form the felsic end-member of a bimodal volcanic suite. The sedimentary rocks could also release fluids and alkalis, which would impart an alkaline affinity and an enrichment in large ion lithophile elements to suture zone igneous rocks.
Collectively, our results indicate that postcollisional igneous rocks that form along terrane suture zones can closely postdate preceding arc magmatism, can have some degree of arc-like affinity (e.g., elevation of large ion lithophile elements) but will likely differ from arc rocks by being bimodal, lacking high LILE/HFSE ratios, exhibiting geochemical characteristics of intraplate magmas, and can include adakites. In addition, we would expect a change in deformation regime along a terrane suture zone if the accreted terrane is transported along margin-parallel fault systems. Deformation recorded in postcollisional suture zone igneous rocks can therefore be useful to constrain the crustal kinematics that are associated with terrane accretion and postaccretion tectonic events.
This research was supported by the National Science Foundation (EAR9814377). We thank Chris Nye, Marti Miller, and Ken Ridgway for detailed and constructive reviews and Dwight Bradley, Mike O'Neill, Ken Ridgway, Jeanine Schmidt, and Alison Till for helpful discussions. We also thank Ray Atkins for his valuable logistical advice and bush-piloting skills and Jim Fitch, Sharon Wesoky, Ted Petrosky, and Dave DuBow for their assistance in the field.
Figures & Tables
Tectonic Growth of a Collisional Continental Margin: Crustal Evolution of Southern Alaska
- continental margin
- Denali Fault
- granitic composition
- igneous rocks
- lava flows
- North America
- partial melting
- Southern Alaska
- suture zones
- Talkeetna Mountains
- United States
- volcanic rocks
- Jack River volcanic field