New U-Pb and 40Ar/39Ar ages integrated with geologic mapping and observations across the western Alaska Range constrain the distribution and tectonic setting of Cretaceous to Oligocene magmatism along an evolving accretionary plate margin in south-central Alaska. These rocks were emplaced across basement domains that include Neoproterozoic to Jurassic carbonate and siliciclastic strata of the Farewell terrane, Triassic and Jurassic plutonic and volcanic rocks of the Peninsular terrane, and Jurassic and Cretaceous siliciclastic strata of the Kahiltna assemblage. Plutonic rocks of different ages also host economic mineralization including intrusion-related Au, porphyry Cu-Mo-Au, polymetallic veins and skarns, and peralkaline intrusion-related rare-earth elements. The oldest intrusive suites were emplaced ca. 104–80 Ma into the Peninsular terrane only prior to final accretion. Deformation of the northern Kahiltna succession and underlying Farewell terrane occurred at ca. 97 Ma, and more widespread deformation ca. 80 Ma involved south-vergent folding and thrusting of the Kahiltna assemblage that records collisional accretion of the Peninsular-Wrangellia terrane and juxtaposition of sediment wedges formed on the inboard and outboard terranes. More widespread magmatism ca. 75–55 Ma occurred in two general pulses, each having distinct styles of localized deformation. Circa 75–65 Ma plutons were emplaced in a transpressional setting and stitch the accreted Peninsular and Wrangellia terranes to the Farewell terrane. Circa 65–55 Ma magmatism occurred across the entire range and extends for more than 200 km inboard from the inferred position of the continental margin. The Paleocene plutonic suite generally reflects shallower emplacement depths relative to older suites and is associated with more abundant andesitic to rhyolitic volcanic rocks. Deformation ca. 58–56 Ma was concentrated along two high-strain zones, the most prominent of which is 1 km wide, strikes east-northeast, and accommodated dextral oblique motion. Emplacement of widespread intermediate to mafic dikes ca. 59–51 Ma occurred before a notable magmatic lull from ca. 51–44 Ma reflecting a late Paleocene to early Eocene slab window. Magmatism resumed ca. 44 Ma, recording the transition from slab window to renewed subduction that formed the Aleutian-Meshik arc to the southwest. In the western Alaska Range, Eocene magmatism included emplacement of the elongate north-south Merrill Pass pluton and large volumes of ca. 44–37 Ma andesitic flows, tuffs, and lahar deposits. Finally, a latest Eocene to Oligocene magmatic pulse involved emplacement of a compositionally variable but spatially concentrated suite of magmas ranging from gabbro to peralkaline granite ca. 35–26 Ma, followed by waning magmatism that coincided with initiation of Yakutat shallow-slab subduction. Cretaceous to Oligocene magmatism throughout the western Alaska Range collectively records terrane accretion, translation, and integration together with evolving subduction dynamics that have shaped the southern Alaska margin since the middle Mesozoic.

The southern margin of Alaska is an active convergent plate boundary where the Pacific plate is subducting beneath the North America plate to the northwest at a rate of ∼60 mm/yr (Gripp and Gordon, 2002). The associated magmatic arc has been continuously active since the Eocene (Jicha et al., 2006) and extends from the Aleutian Islands up the Alaska Peninsula and into the western Alaska Range (Fig. 1). Mount Hayes and Mount Spurr mark the northern extent of the volcanoes associated with the active arc in south-central Alaska (Fig. 1). To the east, only localized arc-related lavas such as the Buzzard Creek basalts (Fig. 1) occur across an ∼500 km gap in arc volcanoes between the western Alaska Range and Oligocene to Holocene volcanoes in the Wrangell Mountains (Richter et al., 1990; Andronikov and Mukasa, 2010; Brueseke et al., 2018). The gap in the continuous chain of arc volcanoes is attributed to collision and shallow subduction of the Yakutat plate, an ∼25-km-thick oceanic plateau, since the Oligocene (Fig. 1; Eberhart-Phillips et al., 2006; Christeson et al., 2010; Bauer et al., 2014). Yakutat subduction has also produced regional Neogene deformation and active seismicity hundreds to thousands of kilometers inboard of the margin and has helped to create the highest topography in North America (Brennan et al., 2011; Lease et al., 2016). The subducting Yakutat plate marks the eastern limit of the modern convergent boundary and a transition to a transcurrent boundary between the North America and Pacific plates throughout southeastern Alaska and western British Columbia (e.g., Finzel et al., 2014).

Magmatism is documented along the North American margin extending from southern Alaska through the northwestern Cordillera since the Late Triassic (e.g., Reed and Lanphere, 1973a; Hudson, 1979; Gehrels et al., 2009). However, the spatiotemporal patterns of magmatism and the tectonic evolution and accretionary history of geologic terranes along the western North American margin are complex and not completely understood. Large, remote areas have been mapped only at the reconnaissance level and have typically been dated by K-Ar techniques. Geologic reconstructions are also hampered by considerable along-strike variation in tectonic environment and significant strike-slip faulting during the Mesozoic and Cenozoic (e.g., Reed and Lanphere, 1974; Cowan, 1982; Redfield and Fitzgerald, 1993; Miller et al., 2002; Pavlis and Roeske, 2007). Nevertheless, detailed studies of igneous rocks and their host terranes along the long-lived margin provide key constraints for unraveling the complex geologic evolution of southern Alaska and the northwestern Cordillera and for reconstructing the tectonic evolution of the margin. These studies also provide a critical framework for understanding the formation and distribution of mineral resources throughout the region and, more broadly, in accretionary orogens worldwide.

Here, we present part of a new study of bedrock exposures covering an area of ∼20,000 km2 of the western Alaska Range in south-central Alaska. Our work integrates field observations, geologic mapping, and geochronology to illuminate the age, petrogenesis, and tectonic setting of middle Cretaceous through Oligocene magmatism. In this paper, we use field observations and new U-Pb and 40Ar/39Ar ages to delineate six magmatic suites among overlapping exposures of plutonic and less abundant volcanic rocks inside the study area. Trends in the spatial and temporal distribution of the magmatic suites are interpreted to record a series of overprinting magmatic belts that record ∼80 million years of changing subduction parameters, terrane accretion, and translation. Mineral deposits such as porphyry Cu-Au, intrusion-related Au, and polymetallic veins are associated with some, but not all, magmatic suites. This relationship suggests an important yet incompletely understood link between igneous petrogenesis, tectonic setting, and metallogeny. Some of the plutonic suites also crosscut a profound crustal boundary between geologic terranes having distinct geologic histories and tectonic affinities. The terranes are regionally deformed and locally metamorphosed, and the older plutonic suites are locally to pervasively deformed and metamorphosed. Thus, the igneous history of the study area provides critical constraints on terrane interactions along the southern Alaska margin and, more broadly, lead to a better understanding of the relationship between magmatism and tectonics in accretionary orogens.

The western Alaska Range (WAR) is a broad north-northeast–striking mountain range across the Cook Inlet to the northwest from Anchorage, Alaska (Fig. 1). Prominent faults mark the edges of the WAR to the northwest and southeast (Fig. 2). On its northwest side, the northeast-striking Denali fault is an active, continental-scale strike-slip fault that has tens to hundreds of kilometers of Cenozoic dextral displacement (Reed and Lanphere, 1974; Miller et al., 2002) and modern slip rates of ∼13 mm/yr (Haeussler et al., 2017). On the southeast side, the east-northeast–striking Castle Mountain fault is also an active strike-slip fault having dextral offset (Haeussler et al., 2000) and a modern slip rate of ∼3 mm/yr (Willis et al., 2007). The western segment of the Castle Mountain fault intersects the northeast-striking Bruin Bay fault and is thought to merge with the Lake Clark fault southwest of Mount Susitna (Figs. 1 and 2). The postulated Telaquana fault (Haeussler and Saltus, 2005) is also inferred to project into the Castle Mountain fault from the west on the basis of aeromagnetic anomalies. The eastern edge of the range is defined by the Cenozoic Susitna Basin (Fig. 2), which is bounded at least in part by a southwest-dipping thrust fault placing Cretaceous igneous rocks on top of younger basinal strata (Saltus et al., 2016). The western edge of the western Alaska Range is defined by a relatively sharp north-south topographic break that is not attributed to any known geologic structure.

Igneous rocks represent more than half of the exposed bedrock in the ∼20,000 km2 WAR study area (Fig. 2), the majority of which are Cretaceous to Paleogene plutons that make up the northern part of the Alaska-Aleutian Range batholith (Reed and Lanphere, 1969, 1972, 1973a). Previous studies by the U.S. Geological Survey (USGS; Reed and Lanphere, 1969, 1973b; Lanphere and Reed, 1985) sampled and mapped igneous rocks from the WAR and northern Aleutian Range (Fig. 1) to inform mineral assessments in the region. Reed and Lanphere (1973a) identified the presence of at least three intrusive suites composing the broader Alaska-Aleutian Range batholith on the basis of K-Ar geochronology on hornblende and biotite: Late Jurassic to Early Cretaceous (ca. 180–130 Ma), Late Cretaceous to early Paleogene (ca. 105–50 Ma), and late Paleogene (ca. 40–25 Ma). Jurassic plutonic rocks do not crop out in the study area but are exposed ∼30 km to the south (Fig. 1). The Late Cretaceous to early Paleogene suite was subdivided locally by Reed and Lanphere (1973a) into Summit Lakes, Yentna, Mount Estelle, Hartman, Crystal Creek, Tired Pup, and McKinley sequences. However, much of this suite remained grouped as “Undivided,” primarily in the eastern half of our study area, because of under-sampling in remote, mountainous regions having extensive glacial cover. Reed and Miller (1980) further subdivided the Yentna sequence into Composite plutons and Kichatna plutons on the basis of their compositions and spatial distribution. The late Paleogene pluton suite was locally subdivided into Merrill Pass, Mount Foraker, Windy Fork, and Snowcap assemblages (Reed and Lanphere, 1973a). The most recent refinements to geologic maps and igneous suite nomenclature in the study area were focused in the southwestern part of our study area (Gamble et al., 2013; Sicard et al., 2013; Todd et al., 2017). Our study builds upon this mapping and geochronological framework, and, in some cases, our new geochronology justifies significant revisions to it.

The WAR is underlain by multiple accreted terranes that have contrasting geologic ages and tectonic affinity. Together with distinct sediment wedges that mantle the terranes and overlapped beginning in the Campanian, these tectonostratigraphic components make up the geologic framework into which younger igneous suites were emplaced (Fig. 1 lower inset). These framework components are described in the following sections.

Parautochthonous Yukon-Tanana Assemblage and Farewell Terrane

The north side of the study area is underlain by rocks of the Farewell terrane and the parautochthonous Yukon-Tanana assemblage (YTA, terminology after Dusel-Bacon and Williams, 2009), both of which are continental in nature and are thought to overlie Proterozoic to early Paleozoic basement (Bradley et al., 2003a, 2007; Dusel-Bacon and Williams, 2009). The YTA comprises pre-Late Devonian siliciclastic strata and Late Devonian to Early Mississippian igneous rocks that are ductilely deformed, regionally metamorphosed, and structurally dismembered. These and correlative assemblages are recognized throughout the Yukon-Tanana upland of interior Alaska and in parts of northwestern Canada (e.g., Dusel-Bacon et al., 2006; Mair et al., 2006). The siliciclastic rocks have detrital zircon ages that indicate North American affinity (Bradley et al., 2007; Dusel-Bacon and Williams, 2009; Piercey and Colpron, 2009; Dusel-Bacon et al., 2017), and they are correlated across the Tintina fault with autochthonous units of the western Selwyn Basin (Mair et al., 2006, and references therein). Thus, the YTA is peri-cratonic and is interpreted to represent the tectonically disrupted edge of the ancient North American continental margin. In the study area, the YTA is only exposed north of the Denali fault where it is faulted between the Farewell terrane and the outboard, allochthonous Wrangellia terrane (Fig. 1 inset, see below).

The Farewell terrane (Decker et al., 1994; Bundtzen et al., 1997; Bradley et al., 2003a) is a non-Laurentian micro-continental fragment that consists of (1) a Proterozoic basement complex; (2) late Neoproterozoic to Devonian platform deposits (Nixon Fork subterrane; Bradley et al., 2003a; Dumoulin et al., 2018b); (3) coeval and related slope and basinal strata (Dillinger subterrane; Dumoulin et al., 2018b); and (4) a Devonian to Jurassic mixed succession of carbonate, siliciclastic, and mafic volcanic rocks (Mystic subterrane; Bradley et al., 2014; Dumoulin et al., 2018a). The Proterozoic basement and platform rocks of the Nixon Fork subterrane are exposed to the southwest and north of the study area in other parts of interior Alaska (Bradley et al., 2014). The Dillinger and Mystic subterranes make up all Farewell exposures in the study area south of the Denali fault (Dumoulin et al., 2018a, 2018b). Layered strata of the Dillinger and Mystic subterranes are deformed into a northwest-vergent, fold-and-thrust belt that is truncated on the north by the Denali fault. The age of folding and thrust faulting is not completely understood but is interpreted to predate final collision and translation of the Peninsular and Wrangellia terranes to the south (Box et al., 2019).

Wrangellia and Peninsular Terranes

The Wrangellia terrane consists of Paleozoic oceanic arc and sedimentary successions that are overlain by a distinctive set of Triassic flood basalts—including the Karmutsen Formation on Vancouver Island and Nikolai Greenstone in Alaska—and Triassic to Jurassic carbonate and siliciclastic strata (Nokleberg et al., 1985; Plafker et al., 1989; Plafker and Berg, 1994; Greene et al., 2010). Wrangellia exposures form a discontinuous linear belt over ∼2000 km from the Talkeetna Mountains of south-central Alaska to Vancouver Island, western British Columbia. The terrane is thought to have evolved outboard of the western North American margin, and then it was accreted during progressive, oblique collision that began in the Late Jurassic (e.g., Csejtey et al., 1982; Jones et al., 1982; Pavlis, 1982; McClelland et al., 1992; Cole et al., 1999; Ridgway et al., 2002; Trop et al., 2002, 2005).

The Peninsular terrane chiefly consists of Mesozoic volcanic, plutonic, and sedimentary rocks that are exposed in a discontinuous belt extending from the Tonsina area of east-central Alaska southwest to the Alaska Peninsula and northwestern part of Kodiak Island. The Late Triassic to Early Jurassic (ca. 207–190 Ma) volcanic-dominated Talkeetna Formation and ca. 202–153 Ma plutonic rocks are signatures of the Talkeetna arc, the defining characteristic of the Peninsular terrane (Nokleberg et al., 1994; Amato et al., 2007b; Rioux et al., 2007). The Talkeetna arc is generally thought to have been intra-oceanic (e.g., Debari and Sleep, 1991; Greene et al., 2006; Rioux et al., 2007), but ca. 160 Ma mafic and intermediate plutons also crosscut distinctive Triassic greenstones of the Nikolai Greenstone (Wrangellia) in the Talkeetna Mountains. Isotopic data suggest interaction between Talkeetna arc plutons and Wrangellia crust in the Late Jurassic (Rioux et al., 2007), and Triassic and older successions among Peninsular terrane exposures on the Alaska Peninsula might represent Wrangellia components (Detterman and Reed, 1980; Amato et al., 2007a). Paleoproterozoic and Paleozoic zircon have also been reported in gabbro xenoliths from Redoubt volcano, a modern arc volcano built on the Peninsular terrane (Fig. 1; Bacon et al., 2012). These relationships raise the possibility that parts of the Talkeetna arc are more primitive but other parts of the Peninsular terrane were built on preexisting crust. This possible relationship is particularly relevant to this study because Wrangellia and Peninsular terrane rocks are exposed east and south of the study area but are obscured in the western Alaska Range by younger igneous successions, sedimentary rocks, and surficial deposits. Thus, we infer that rocks of the Peninsular and Wrangellia terranes make up the basement in the southeastern part of the study area on the basis of regional aeromagnetic data (Glen et al., 2007) but did not identify any Jurassic components among the igneous suites described below.

There is evidence for northward displacement of all the terranes discussed herein both before and after they were accreted along the Cordilleran margin. Minimum estimates for post-Late Cretaceous dextral offset along the Tintina and Denali faults are on the order of hundreds of kilometers (e.g., Eisbacher, 1976; Lowey, 1998; Gabrielse et al., 2006). Other estimates suggest that thousands of kilometers of displacement are required (e.g., Marquis and Globerman, 1988; Irving et al., 1995; Wynne et al., 1995; Ward et al., 1997; Day et al., 2016). The magnitude of syn- and post-accretionary, strike-slip displacement is still highly debated, though, and must be carefully considered when interpreting possible terrane interactions and reconstructing tectonic geometries prior to the Cenozoic (e.g., Cowan, 1982; Csejtey et al., 1982; Umhoefer, 1987; Housen and Beck, 1999; Mahoney et al., 1999; Stamatakos et al., 2001; Pavlis and Roeske, 2007; Day et al., 2016).

Sediment Wedges and Complex Terrane Boundaries

The Kahiltna assemblage comprises several sequences of Jurassic to Cretaceous sedimentary rocks that were deposited on the outboard, allochthonous Wrangellia and Peninsular terranes and the inboard Farewell terrane and parautochthonous YTA. Rocks of the Kahiltna assemblage crop out throughout the Alaska Range in a belt that extends for more than 800 km along strike (Fig. 1 inset). Broadly contemporaneous successions such as the Nutzotin Mountains sequence, Dezedeash Formation, and Gravina belt to the east and southeast are in similar positions with respect to the outboard (i.e., Insular belt) and inboard (i.e., Intermontane belt) terranes and define a discontinuous belt of sedimentary strata spanning thousands of kilometers of the Mesozoic western North American margin (e.g., Kapp and Gehrels, 1998; Lowey, 2019; Trop et al., 2020). In the western Alaska Range, the belt of Kahiltna assemblage exposures is ∼100 km wide and hosts many of the plutonic suites in this study. The Kahiltna assemblage is made up of generally fine grained siliciclastic rocks deposited in a marine environment, and Kalbas et al. (2007) estimated a minimum stratigraphic thickness of 5.5 km. Petrographic observations and detrital zircon data indicate that sediment was derived from distinctive source terranes on the north and south (i.e., Farewell and YTA versus Wrangellia and Peninsular) but that the two depocenters remained completely separated until the Late Cretaceous (Kalbas et al., 2007; Hults et al., 2013; Box et al., 2019).

The Kahiltna assemblage was pervasively deformed by folding and faulting and contains kilometer-scale recumbent folds. Slaty cleavage is locally developed in parts of the outcrop belt, and multiple cleavages are observed in some areas. In the northwestern part of the study area, kilometer-scale, northwest-vergent folds and thrust faults are formed in the older parts of the Kahiltna assemblage (Dalzell petrofacies unit; Box et al., 2019) and underlying rocks of the Farewell terrane. Box et al. (2019) reported a metamorphic titanite U-Pb age of 96.8 ± 2.3 Ma from a mafic dike crosscutting deformed rocks of the Farewell terrane, which is younger than the youngest depositional ages of the Dalzell unit as constrained by detrital zircon. Thus, they interpreted the titanite age to record metamorphism associated with northwest-directed thrusting and folding of the Farewell terrane and older components of the Kahiltna assemblage to the north. In contrast, cleavage and bedding are moderately northwest dipping across much of the outcrop belt to the south, suggesting tight, south-vergent folds (Box et al., 2019). South-vergent deformation affected younger parts of the Kahiltna assemblage that have depositional ages younger than ca. 97 Ma (Box et al., 2019) and is interpreted to record oblique collision of the Wrangellia and Peninsular terranes with the YTA and Farewell terrane.

Across the study area, the complexly deformed Kahiltna assemblage obscures the actual accretionary boundary between the inboard and outboard terranes. Regional aeromagnetic data suggest that there is a contrast between the geophysical character of the continental Farewell and oceanic Wrangellia terranes beneath the Kahiltna assemblage (Glen et al., 2007). However, deformation associated with formation and translation of the collisional terrane boundary is distributed across the ∼200-km-wide belt of folded and faulted Kahiltna assemblage and Farewell terrane rocks south of the Denali fault (Box et al., 2019). Similar patterns are observed along strike to the northeast where the Alaska Range suture zone marks an ∼100-km-wide zone of deformation involving Wrangellia, the Kahiltna assemblage, and strongly deformed rocks of the Yukon-Tanana assemblage (Jones et al., 1982; Ridgway et al., 2002). The complex deformation zone is bounded on the south by the Talkeetna thrust and on the north by the Hines Creek fault (Ridgway et al., 2002; Brennan et al., 2011); parts of the Kahiltna assemblage were metamorphosed to upper amphibolite facies during the Late Cretaceous and exhumed along the Valdez Creek shear zone in the Paleocene (Davidson et al., 1992; Ridgway et al., 2002). The suture zone was then crosscut by the Denali fault as the tectonic regime along the margin shifted from oblique collision to strike-slip translation after the Paleocene (e.g., Waldien et al., 2020).

Bedrock samples for geochronology were collected during USGS field campaigns in the western Alaska Range between 2001 and 2013. Where possible, we also identified remainders of previously collected igneous bedrock samples from the USGS archive and processed select samples for mineral separates. These legacy USGS samples allowed us to greatly expand our data set to include igneous bodies that were not sampled during 2001–2013 fieldwork or were outside the area of focused geologic mapping and project work. For U-Pb geochronology, minerals were separated and concentrated from bulk bedrock samples by standard methods including crushing and pulverizing, water shaking table, heavy liquids, and magnetic susceptibility. Zircon and titanite were hand-picked and analyzed by multiple methods: isotope-dilution thermal ionization mass spectrometry techniques (ID-TIMS; University of British Columbia), laser ablation–inductively coupled plasma–mass spectrometry techniques (LA-ICP-MS; University of California Santa Barbara and USGS Denver), and sensitive high-resolution ion microprobe–reverse geometry techniques (SHRIMP-RG; USGS/Stanford University). All of the samples analyzed at the University of California Santa Barbara were done by laser-ablation split-stream (LASS) methods in which U-Th-Pb and Hf isotopes were measured simultaneously for the same spot. Detailed descriptions of analytical methods for each technique are included in the Supplemental Material1 together with all analytical data (Table S1). Bedrock samples including a distinctive suite of fine-grained mafic dikes were submitted to the University of Alaska Fairbanks (UAF) for 40Ar/39Ar geochronology. These samples were prepared using standard crushing and pulverizing techniques at the UAF Geochronology Laboratory, and detailed analytical techniques are described in the Supplemental Material together with all analytical data (Table S2). All U-Pb and 40Ar/39Ar analytical data are also contained in multiple USGS Data Releases (Holm-Denoma et al., 2020; Jones et al., 2020a, 2020b; Todd et al., 2020). Table 1 summarizes the U-Pb age(s) determined for each sample, brief rock descriptions, and the method and lab that were used. It also includes the few U-Pb ages that were previously published in the area. Table 2 summarizes the 40Ar/39Ar ages determined for several samples of plutonic and volcanic rocks in the study area. All quoted uncertainty for U-Pb and 40Ar/39Ar ages in the tables, figures, and text below is at the 2-sigma level.

For this study, we focused on sampling representative parts of all igneous units that were previously mapped in addition to those that were newly identified during our geologic mapping.

Few U-Pb ages for the study area were published prior to this work (Table 1), but previously published K-Ar and 40Ar/39Ar ages (Wilson et al., 2015, and references therein) indicated ∼80 million years of overlapping magmatism and distinct pulses in the Paleocene and Eocene (Fig. 3). Our new results generally agree with published U-Pb and argon data (with some exceptions discussed below) but also provide a more detailed understanding of the age and distribution of middle and Late Cretaceous magmatic events that are regionally significant from a metallogenic perspective (e.g., Graham et al., 2013).

The results described below are grouped into six age suites that were identified through clustering of ages, geochemical similarities, textural similarities, and/or regional crosscutting relationships. Each sample was assigned an abbreviation corresponding to its age suite, and corresponding field sample numbers are given in Table 1. Two samples whose ages are interpreted to represent metamorphism are labeled with the abbreviation “Met” in Table 1.

Early Late Cretaceous (ELC; >90 Ma) Suite

The oldest igneous rock identified in this study is from a localized outcrop band of flow-banded rhyolite that forms a roof pendant within the Paleocene Tordrillo Mountains batholith (see below). The rhyolite yielded a hornblende 40Ar/39Ar age of 122.4 ± 3.0 Ma (ELCA1; Table 2), which we interpret to represent crystallization. The oldest plutons identified in this study are localized in two parts of the study area (Fig. 4). The Shell Hills pluton (99.4 ± 1.3 Ma, sample ELC1, Table 1) underlies the hills east of Shell Lake on the western edge of Susitna Basin (SH in Fig. 4). The Shell Hills pluton is made up of undeformed, medium-grained biotite granite (Fig. 5), and it is associated with coeval volcanic rocks capping the top of the hills that produced a 40Ar/39Ar whole-rock age of 98.2 ± 0.8 Ma (ELCA2; Table 2). Two additional volcanic beds of this approximate age were identified interlayered with fine-grained siliciclastic rocks of the Kahiltna assemblage in exposures ∼30 km west of Shell Hills. Small, euhedral zircon from 2-cm-thick, light-colored layers have U-Pb ages of 93.3 ± 1.2 Ma and 97.1 ± 0.6 (samples ELC7 and ELC5, Table 1). The zircons are interpreted to be primary igneous grains, and the thin, white layers are interpreted to be ash deposited together with Kahiltna assemblage strata on the northern flank of the Peninsular and Wrangellia terranes (Box et al., 2019). The ca. 97 Ma sample also contained three concordant zircon grains that have ages of ca. 233, 951, and 1427 Ma (Table S1) that we attribute to inheritance and also possible lead loss from older, underlying sources. These older ages are consistent with observations by Bacon et al. (2012) for xenoliths from Redoubt volcano to the south (Fig. 1) and with the older components of detrital zircon U-Pb age populations contained in the Kahiltna assemblage country rocks (Box et al., 2019). Otherwise, both ash layers produced unimodal zircon age populations that we interpret to represent primary igneous ages. The age of the older sample overlaps with volcanic rocks exposed at Shell Hills, and both ages overlap with prominent detrital zircon populations in Late Cretaceous parts of the Kahiltna assemblage (Hampton et al., 2010; Hults et al., 2013; Box et al., 2019). These two ash layers represent the first direct depositional ages on the upper parts of the Kahiltna assemblage in the region and suggest that contemporaneous volcanic rocks were an important local source of detrital zircon in the Late Cretaceous. Otherwise, volcanic rocks in the 100–90 Ma age range are rare in the study area, suggesting they were volumetrically minor or were eroded during younger tectonic events.

The other plutonic rocks making up the ELC suite are exposed along the northern shore of Chakachamna Lake in the southern part of the study area (LCh in Fig. 4). These rocks were originally mapped as Tertiary or Cretaceous (unit TKgm in Wilson et al., 2015, and references therein) on the basis of published latest Cretaceous and Paleocene K-Ar ages (summarized in Wilson et al., 2015, recalculated from Reed and Lanphere, 1972). However, two samples of strongly deformed granite to granodiorite—one exposed in shoreline outcrops (Fig. 6B) and a second exposed 1 km north of and 300 m above the lakeshore (Fig. 6F)—yielded identical zircon U-Pb ages of 98.1 ± 1.1 Ma (Figs. 6D and 6H; samples ELC3 and ELC4, Table 1). A compositionally similar but undeformed granite collected ∼3 km north of and 1000 m above the lakeshore (sample ELC2) yielded zircon U-Pb ages of 99.2 ± 1.3 Ma and 99.9 ± 1.5 Ma by SHRIMP and LASS methods, respectively (Table 1). In the deformed lakeshore outcrops, quartz and feldspar are both extensively recrystallized and define a consistent east-northeast–striking, subvertical foliation that is enhanced by aligned biotite (Fig. 6B). The pervasive planar fabric is approximately parallel to the northern shoreline of the elongated lake. The foliation is enhanced locally by meter-thick zones of mylonite containing S-C fabric, centimeter-wide ductile shear bands, asymmetric porphyroclasts, mica fish, and strongly ribboned quartz (Fig. 6C). Biotite and quartz ribbons define a mineral stretching lineation that is moderately to strongly developed, and it plunges moderately east-northeast. Kinematic indicators consistently suggest oblique, northside-down dextral shear together with a component of flattening across the foliation.

In addition to gneissic granitoids, outcrops along the lake shore also contain panels of fine-grained mafic bodies interpreted to be xenoliths or enclaves and multiple generations of crosscutting aplitic veins and dikes that are all similarly deformed. Some of the aplite dikes and veins crosscut the ductile fabrics in the host granodiorite but are also folded in with the pervasive foliation. This relationship suggests that magmatism might have accompanied deformation and metamorphism. The late-kinematic dikes and veins were sampled for geochronology but did not yield zircon or titanite. Titanite was separated from the two samples of ca. 98 Ma deformed granite. The southern lakeshore sample (ELC4) and the sample ∼1 km to the north (ELC3) yielded titanite U-Pb ages of 56.5 ± 0.5 Ma and 59.6 ± 0.8 Ma, respectively (Figs. 6E and 6I; Table 1). In thin sections for both samples, titanite occurs in elongated clusters parallel to the mylonitic foliation. Therefore, we interpret the titanite ages to represent recrystallization during metamorphism that accompanied deformation (Frost et al., 2001; Kohn, 2017). A mylonitic mafic layer in orthogneiss (sample Met2) yielded a preliminary zircon U-Pb age of ca. 57 Ma based on three grains (Table S1), which we also interpret to represent metamorphism. These metamorphic ages are similar to published K-Ar ages ranging from ca. 65–51 Ma for the same Cretaceous granitic rocks along northern Chakachamna Lake (summarized in Wilson et al., 2015), indicating that argon systematics were completely reset in the deformed granitic rocks during the Paleocene.

Middle Late Cretaceous (MLC; Ca. 90–75 Ma) Suite

Late Cretaceous plutonic rocks of what we call the middle Late Cretaceous (MLC) suite are more abundant than the older ELC suite, but they are still relatively minor in exposures across the study area (Fig. 2). The oldest rocks of this suite are exposed on the southern shore of Chackachamna Lake (Fig. 4), where two samples of diorite and granodiorite yielded zircon U-Pb ages of 85.7 ± 1.4 and 84.9 ± 1.1 Ma (MLC1 and MLC2; Fig. 7 and Table 1). These samples were collected 3–4 km south of the older mylonitic rocks described above, and they are undeformed (Fig. 7A). Titanite from one of the samples (MLC2) yielded a U-Pb age of 77.4 ± 0.9 Ma (Fig. 7C), which is slightly younger than the crystallization age. Overall, the metamorphic and thermal history of these two samples is distinct from the sheared rocks to the north, suggesting a structural break along the lake corridor.

In the central part of the study area, three granitoid plutons exposed on the north side of the Tordrillo Mountains south of the Skwentna River produced zircon U-Pb ages of 80.4 ± 0.2 Ma, 79.5 ± 1.1 Ma, and 78.9 ± 1.1 Ma (MLC3, MLC4, and MLC5; Table 1). These samples are all monzonitic to granodioritic and internally undeformed, and they were all emplaced into rocks of the Kahiltna assemblage. Along the Chickak River (labeled “CR” in Fig. 4) ∼2 km northwest of the 80.4 ± 0.2 Ma pluton, we observed a 1–2-m-thick dike of porphyritic dacite that cuts bedding and slaty cleavage in Kahiltna assemblage sandstone and shale (Figs. 7D and 7E). The intrusion is also folded gently in the limb of a larger upright fold and boudinaged, a relationship we interpret to be syn- to late kinematic with respect to deformation of the surrounding Kahiltna assemblage. We collected the dike for U-Pb geochronologic analyses, but it did not yield zircon. We infer that it is associated with the nearby ca. 80 Ma monzonitic pluton on the basis of its proximity and composition, but it might also be associated with the ca. 76 Ma Whistler intrusion (labeled “W” in Fig. 4; Hames, 2014) described below.

On the southeastern side of the study area, diorite, tonalite, and granodiorite are exposed across Mount Susitna (labeled “MS” in Fig. 4), a prominent topographic feature marking the southwestern edge of the Susitna Basin. Plutonic rocks along the ridge contain localized zones of abundant amphibolite that are either xenoliths or enclaves of more mafic intrusive phases. All the intrusive igneous rock exposures contain a moderate to well-developed subvertical foliation that strikes north-northeast (Fig. 7F). Alignment of blocky feldspar phenocrysts suggests that, in some outcrops, the foliation might have initially formed during emplacement and crystallization. However, in most outcrops, the foliation appears to have been modified or intensified by solid-state deformation as evidenced by aligned biotite and hornblende and recrystallized, elongated quartz and feldspar. Solid-state fabrics are most intense in meter-thick zones rich in more mafic material and display local fabric intensification along the contacts. One sample of granodiorite from near the top of Mount Susitna produced a zircon U-Pb age of 78.2 ± 1.5 Ma and a titanite U-Pb age of 71.7 ± 1.0 Ma (MLC6; Figs. 7G and 7H; Table 1). Quartz diorite from ∼3 km to the north produced a 40Ar/39Ar hornblende age of 75.5 ± 2.6 Ma (MLCA1; Table 2). Titanite in thin sections of sample MLC6 occurs in clusters oriented parallel to the foliation, suggesting the younger age represents recrystallization during metamorphism. A separate, poorly exposed plutonic body along Trail Ridge in the Susitna Basin lowlands to the northeast is tentatively correlated with the Mount Susitna pluton on the basis of a 75.9 ± 6.0 Ma 40Ar/39Ar biotite age (Fig. 4; MLCA2; Table 2).

The youngest rocks of the MLC suite host the Island Mountain and Whistler copper-gold deposits in the central part of the range (labeled IM and W in Fig. 4). The Island Mountain deposit is made up of a cluster of mineralized breccia pipes that are associated with ca. 77 Ma monzodiorite and diorite intrusions dated by Gross (2014). The Whistler deposit is a group of porphyry occurrences associated with the ca. 76 Ma Whistler igneous suite (WIS) of Hames (2014); this suite includes dioritic intrusive rocks, hypabyssal dikes and sills, and andesitic volcanic rocks. Intrusive rocks that host the Whistler and Island Mountain deposits were all emplaced into the Kahiltna assemblage. The WIS is also the oldest igneous unit that cuts the Whistler petrofacies unit of the Kahiltna assemblage (Box et al., 2019). The WIS also intrudes deformational features in the Whistler petrofacies of the Kahiltna, indicating that south-vergent deformation had occurred by this time. The Whistler petrofacies is the younger unit of the northern Kahiltna assemblage, and this crosscutting relationship also links the northerly derived petrofacies to the southern petrofacies of approximately the same age (Old Man petrofacies unit; Box et al., 2019).

Latest Cretaceous (LC; Ca. 75–65 Ma) Suite and Adjacent Trends

Latest Cretaceous plutons are widespread across the study area and define a roughly linear, >70-km-long trend that varies in orientation from northeast to north-south. Plutons in this suite extend to the north into the Farewell terrane and include the Composite suite (Reed and Nelson, 1980; Reiners et al., 1996). The Composite suite includes a northeast-trending array of intrusive igneous bodies (Fig. 4) extending from the northern part of Mount Estelle to the Cascade and Lower Yentna plutons (“C” and “LY” in Fig. 4) and beyond the study area. The Composite suite plutons have an average composition of quartz monzonite (Reed and Nelson, 1980), but they were so named because several (but not all) of the plutons are composed of distinctive ultramafic, intermediate, and felsic zones ranging from peridotite to granite. Geochemical and isotopic signatures of some indicate crustal contamination of parental melts, likely from partial melting of fine-grained siliciclastic Kahiltna assemblage country rock (Reiners et al., 1996). Two of the plutons—the Kohlsaat and Cascade (Ko and C in Fig. 4)—yielded K-Ar ages of ca. 66 and 65 Ma (Wilson et al., 2015, recalculated from Reed and Lanphere, 1972). The Cascade and Lower Yentna plutons from the northeastern end of the Composite suite trend (C and LY in Fig. 4) have zircon U-Pb ages of 66.4 ± 0.6 and 69.6 ± 0.5 Ma, respectively (Table 1; Hung, 2008). The Mount Estelle pluton, a prominent north-south–trending body south of the main Composite suite trend (Fig. 4), contains a similar range of rock types that have an average composition of biotite-hornblende granodiorite and, thus, may be related (Reed and Nelson, 1980). Detailed mapping suggests that the pluton represents a series of coalesced intrusions (Graham et al., 2013) that yielded zircon U-Pb ages ranging from ca. 71–67 Ma (Table 1; Taylor et al., 2014).

Several other intrusions associated with this suite are also exposed across the study area on both sides of the Mount Estelle–Composite suite pluton trend. On the western side of the trend, a 20-km-long body of hornblende granodiorite (Sled Pass pluton, or SP, in Fig. 4) yielded zircon U-Pb ages of 72.1 ± 0.5 Ma and 57.2 ± 0.7 Ma (LC2 and P36; Figs. 8A–8D; Table 1). The older age is from a sample of coarse-grained hornblende granodiorite at the southern end of the pluton; this sample has a weak foliation defined by 1–2 cm acicular hornblende and up to 3 cm recrystallized plagioclase feldspar (Fig. 8A). The younger age is from a sample of strongly foliated and locally mylonitic hornblende granodiorite on the northern end of the pluton along the contact on its western edge (Figs. 8D–8F). Rocks of the Kahiltna assemblage adjacent to the contact are subvertical, tightly folded, and sheared together with meter-sized boudins of more coherent rock (Fig. 8G). The shear zone foliation is subvertical, parallel to the northeast-striking pluton margin, and it contains a steeply plunging lineation defined by hornblende. The asymmetry of the boudins and recrystallized feldspar in the mylonitic granodiorite indicate northwest-side-up (i.e., pluton down) kinematics, and folding at this locality appears to be associated with shear zone formation. Finer-grained phyllitic layers of the Kahiltna assemblage adjacent to the shear zone have abundant 0.5 cm andalusite (Figs. 8H and 8I), which is one of the few known localities in the study area where Kahiltna strata contain metamorphic minerals. The phyllitic foliation wraps around the andalusite (Fig. 8I), which suggests that metamorphism occurred before deformation. A previously published K-Ar age of ca. 76 Ma on hornblende from granodiorite ∼800 m east of the deformed pluton margin (Wilson et al., 2015, recalculated from Reed and Lanphere, 1972) suggests that the zircon date from the southern part of the body represents the age of emplacement (Fig. 8B). The zircon grains from the mylonitic granodiorite do not have obvious metamorphic overgrowths, and they have Th/U ratios ranging from 0.44 to 0.16, which are not indicative of metamorphism (Table S1). Thus, we suspect that sample P36 represents a localized Paleocene intrusion that was emplaced along the margin of the Sled Pass granodiorite and then subsequently deformed. The Paleocene intrusion might be associated with the Tired Pup batholith across the Stony River to the west (see below). No available information ties the ca. 76 Ma hornblende K-Ar age to a foliation; so it is unclear when the foliation in the coarse-grained granodiorite at the southern end of the pluton formed. Other plutons of this suite on the western side of the Mount Estelle–Composite suite trend include three more localized plutons of intermediate to felsic composition emplaced into both the Kahiltna assemblage and rocks of the Farewell terrane. These plutons yielded zircon U-Pb ages of 68.3 ± 1.9 Ma (LC6), 67.2 ± 1.2 Ma (LC10), 66.5 ± 1.3 Ma (LC12), and 65.8 ± 0.9 Ma (LC13). Sample LC6, a 68.3 ± 1.9 Ma biotite-hornblende monzonite (Fig. 8J), is the oldest pluton we dated that was emplaced directly into rocks of the Farewell terrane.

On the eastern side of the Mount Estelle–Composite suite trend, medium-grained biotite granite crops out along Kitty Ridge northeast of Beluga Lake (KR and BL in Fig. 4). Our zircon U-Pb age of 73.6 ± 1.0 Ma for this pluton (LC1; Figs. 8J and 8K) contrasts with a biotite 40Ar/39Ar age of 60.5 ± 0.4 Ma from nearby (LCA1; Table 2), suggesting that the argon systematics at this locality were reset during the Paleocene. The granite does not have an obvious fabric in the one accessible outcrop that we examined (Fig. 8K). The Paleocene biotite age is consistent with metamorphic ages in titanite and argon ages from deformed granitoids along Lake Chakachamna to the west (see above); therefore, the biotite in the Kitty Ridge pluton might have been thermally reset during the same event. In the Little Peters Hills of the northern Susitna Basin (LPH in Fig. 4), a ∼1.5-km-wide stock of equigranular coarse-grained biotite granite yielded a zircon U-Pb age of 70.0 ± 1.0 Ma (LC3; Table 1). The intrusion cuts folded siltstone and sandstone of the Kahiltna assemblage but does not appear to be deformed.

Other igneous rocks that are part of this suite are located east of Mount Estelle on the northern side of the Tordrillo Mountains (Fig. 4). On the northern side of the Skwentna River (Fig. 4) are a series of kilometer-scale dioritic to granodioritic plutons that were emplaced into the Kahiltna assemblage. We sampled a granodiorite pluton on the southern end of the group, and it yielded a zircon U-Pb age of 68.5 ± 1.3 Ma (LC4; Table 1). This age is consistent with a previously published biotite K-Ar date of 69.2 ± 2.0 Ma from a diorite pluton 15 km to the northwest (Wilson et al., 2015, recalculated from Reed and Lanphere, 1972). On the southern side of the Skwentna River, rhyolite that overlies sedimentary rocks of the Kahiltna assemblage yielded a zircon U-Pb age of 67.5 ± 0.5 Ma (LC9; Table 1). The rhyolite only crops out on the north side of Paleocene and early Eocene intrusions (ca. 56–54 Ma hornblende and whole-rock 40Ar/39Ar ages; Layer and Solie, 2008). It is not obviously connected to Paleocene volcanic rocks that are more widespread on the eastern flank of the Tordrillo Mountains (see below).

Paleocene (P; Ca. 65–56 Ma) Suite

Paleocene plutonic rocks have the broadest spatial extent of all the igneous suites across the field area and include three major batholiths—the Tordrillo Mountains batholith, the Tired Pup batholith, and the Middle Fork plutonic complex (TM, TP, and MF in Fig. 4). The Tordrillo Mountains batholith is the largest of the three, roughly equant and having a diameter of at least 60 km, exposed in the Tordrillo Mountains and west along the headwaters of the Skwentna River. The batholith was emplaced into rocks of the Kahiltna assemblage and is mostly felsic, made up of granite and quartz monzonite that have zircon U-Pb ages ranging from 60.3 ± 0.1 Ma (P17) to 57.4 ± 0.2 Ma (P35; Table 1). Also, small, localized occurrences of gabbro and other mafic intrusive phases are exposed along the margins of the batholith that have U-Pb ages of 59.8 ± 1.0 Ma (P22) and 57.1 ± 0.5 (P37). The batholith contains a few localized occurrences of syenite and granodiorite that have previously published hornblende K-Ar ages of ca. 109 and 98 Ma (summarized in Wilson et al., 2015, recalculated from Reed and Lanphere, 1972); we interpret these ages to be xenoliths of older intrusive phases, but the extent and abundance of these older bodies has not been mapped in detail. Ca. 61–55 Ma K-Ar and 40Ar/39Ar ages on biotite and hornblende (summarized in Wilson et al., 2015) from throughout the batholith are not much younger than zircon ages and suggest that cooling occurred shortly after crystallization. On the eastern side of the batholith, a large area of coeval volcanic rocks (Figs. 9A and 9B) yielded zircon U-Pb ages of 60.2 ± 0.9 Ma (P18; Fig. 9C) and 60.1 ± 0.9 Ma (P19 and P20). These volcanic rocks include welded andesitic tuffs and phenocrystic rhyolite where sampled; they are mapped all the way to Beluga Mountain and the range front north of Mount Susitna (BM and MS in Fig. 4; Wilson et al., 2015). Bradley et al. (2017) reported a zircon date of 56.1 ± 1.9 Ma (01ADw7a; Table 1) for felsic tuff from the Sheep Creek volcanic field in the northern part of the project area, indicating that volcanic activity associated with this magmatic suite was widespread, though it is not otherwise observed west of the Tordrillo Mountains.

On the western side of the study area, the Paleocene Tired Pup batholith underlies a large part of the Revelation Mountains (RM in Fig. 4) and is up to 65 km long and 25 km across as mapped (Fig. 4). It is elongated north-south and cuts across the mapped boundary between the Kahiltna assemblage and Farewell terrane strata. Compositions are predominantly felsic, and zircon U-Pb ages range from 63.6 ± 1.2 (P1) to 57.6 ± 0.2 Ma (P34) (Table 1). Locally, megacrystic biotite granite and quartz monzonite (58.4 ± 1.1 Ma; P32; Fig. 9D) are interlayered within a shallowly dipping, sheet-like geometry.

Farther to the north near the headwaters of the Middle Fork of the Kuskokwim River (MFK in Fig. 4), the Middle Fork plutonic complex is ∼13 km in diameter and includes compositions ranging from granite to olivine-clinopyroxene syenite to alkaline gabbro (Solie, 1983, 1988; Gilbert et al., 1985; Bundtzen et al., 1997). The Middle Fork complex (MF in Fig. 4) was initially thought to correlate with the Oligocene Windy Fork pluton ∼8 km to the southeast (Solie, 1983). Our sample from the quartz monzonite phase of the Middle Fork complex produced a zircon U-Pb age of 59.3 ± 0.8 Ma (P24; Figs. 9E and 9F; Table 1). The alkali gabbro phase yielded zircon and titanite U-Pb ages of 56.2 ± 0.3 Ma and 57.0 ± 6.4 Ma, respectively (P38; Table 1). These ages are broadly similar with published ca. 57–56 Ma K-Ar ages on hornblende and biotite from across the plutonic complex (Gilbert et al., 1988). The Middle Fork plutonic complex was emplaced into rocks of the Farewell terrane, and a swarm of subvertical, east-west–striking dikes that have variable composition and size extends more than 15 km east of the pluton margin. A hornfels zone in the surrounding carbonate and siliciclastic strata is up to 5 km across (Bundtzen et al., 1997). Although the dikes are not directly dated, they are spatially associated with the eastern margin of the Middle Fork complex and, thus, are inferred to be Paleocene. No genetic link between rocks of Tired Pup and Middle Fork has yet been identified, but this is largely due to the absence of mafic rocks from the latter.

Our zircon U-Pb ages (Table 1) and/or previously published K-Ar ages (Wilson et al., 2015) indicate that numerous other localized plutons throughout the area are part of this magmatic suite. The contacts of the intrusive bodies are all sharp, and the plutonic rocks are not foliated or otherwise deformed anywhere they were observed.

Paleocene to Eocene (Ca. 58–51 Ma) Mafic Dikes

Paleocene and older granitic rocks in the study area were intruded by a large mafic dike swarm, which was first noted by Solie et al. (1991b) and later described in more detail by Haeussler et al. (2008). These distinctively dark-colored dikes intrude the light-colored granites and are most obvious and common along the eastern side of the Tordrillo Mountains (see patterned area in Figs. 2 and 4). They are not found south of Chakachamna Lake nor much farther north than the Skwentna River. The dikes most commonly strike north to north-northwest and dip steeply westward. They constitute up to ∼30% of the volume of some outcrops, and they typically make up 15%–20% of the outcrop area where they are observed.

We obtained seven 40Ar/39Ar whole-rock ages on the mafic dikes, which range from 59.0 ± 2.4 Ma to 52.2 ± 2.4 Ma (MD1–MD7; Table 2). Representative 39Ar release spectra are shown for three samples in Figures 10E–10G. The kernel density estimate (KDE) diagram in Figure 10D shows a bimodal age distribution having a major mode at ca. 58 Ma and a minor mode at ca. 53 Ma. It is notable that where the mafic dike and granitic rocks were both dated, such as the mountainside in Figure 10A, the dike ages are within a few million years of the granitic rock ages. Moreover, the older dikes overlap with the youngest rocks of the Paleocene suite described above, but the dike emplacement continued after plutonism waned in the study area (Fig. 10D).

There is evidence of dip-slip faulting parallel to the dikes after their emplacement. In some outcrops, the faulting has an inconsistent sense of offset— parallel to the strike of the dikes exposed on the cliff shown on Figure 10A. Also, Haeussler et al. (2008) found evidence for 1–2 km of west-vergent reverse faulting after ca. 6 Ma along the margins of the dikes in several areas of the dike swarm. It seems plausible that late-stage reverse faulting may have occurred elsewhere in the study area.

Eocene (E; Ca. 45–35 Ma) Suite

The Merrill Pass pluton is the most prominent Eocene intrusive body in the study area. It extends for at least 50 km north-south between the Tordrillo Mountains and Tired Pup batholiths (Fig. 4). These two Paleocene batholiths effectively define the eastern and western boundaries to most of the exposed Eocene plutonism in our study area. The Merrill Pass pluton is up to 35 km wide west of Lake Chakachamna, and it was largely emplaced into the Kahiltna assemblage but also contains mapped bodies, possibly pendants, of the Late Triassic Chilikadrotna Greenstone (Gamble et al., 2013). The pluton is chiefly made up of biotite granite and has some hornblende locally present (Gamble et al., 2013), and zircon U-Pb ages range from 42.7 ± 0.7 Ma (E4) to 37.0 ± 0.5 Ma (E16) (Table 1). On the western side of the pluton, a sample from the base and another from perhaps a kilometer higher in the succession of andesitic volcanic rocks both yielded ages of 41.4 ± 0.5 Ma (Fig. 11C; E5 and E6; Table 1), indicating that rapid, voluminous volcanism accompanied emplacement of the pluton. Previously published biotite K-Ar ages ranging from 37.7 ± 1.1–35.6 ± 1.1 Ma (summarized in Wilson et al., 2015) indicate relatively rapid cooling.

Notably, the northern lobe of what Gamble et al. (2013) map as the extent of the Merrill Pass pluton, between the South Fork of the Kuskokwim River and Styx River (SFK and Styx in Fig. 4), yielded two Paleocene ages (P13 and P30; 60.8 ± 0.8 Ma and 58.9 ± 0.8 Ma, respectively). The northernmost Eocene crystallization age from the mapped Merrill Pass pluton comes from a sample south of the Chilligan River (ChR in Fig. 4; R. Cole, 2013, written commun.). This general locality, which is in the Lime Hills C-1 1:63,360-scale quadrangle, was the focus of new geologic mapping and systematic sampling by the Alaska Division of Geological and Geophysical Surveys in 2013. On the basis of new geochemistry from that project (Sicard et al., 2013) and correlations with chemistry from samples that have U-Pb ages (Todd and Jones, 2017), we surmise that most, if not all, of the northern lobe (north of 61.5°N) of Merrill Pass pluton is Paleocene and, thus, more likely associated with the Tordrillo Mountains batholith to the east.

Plutons ranging in age from 38.5 ± 0.5 Ma (E11) to 36.6 ± 0.8 Ma (E18; Fig. 11D; Table 1) are exposed to the west of the Merrill Pass pluton along the Stony River (Fig. 4). The sample on the eastern side of the Stony River is from a hornblende-biotite granite pluton that intrudes the base of the ca. 41 Ma volcanic rocks (Figs. 11B and 11D). The granite contains 1–5 cm miarolitic cavities that are abundant within ∼1 m of the contact zone with the overlying andesite (Fig. 11E), and it also has zones up to 20 m wide that contain 60%–70% xenoliths of fine-grained porphyritic andesite. As with the larger Merrill Pass pluton, a 34.9 ± 1.0 Ma biotite K-Ar age from the same approximate locality indicates relatively rapid cooling after emplacement (Wilson et al., 2015).

Other igneous rocks of this suite are exposed north of the Merrill Pass pluton along an approximate north-south trend. West of the Post River (Fig. 4), a localized body of quartz porphyry called the Bowser Creek intrusive (Solie et al., 1991a) was emplaced into Farewell terrane strata. The porphyry has a K-Ar age of ca. 60 Ma (Sample 1 of Solie et al., 1991a), but thin (10–20 cm) biotite-hornblende granite dikes that crosscut the porphyry yielded a zircon U-Pb age of 44.1 ± 0.8 Ma (E2; Table 1). Outcrop observations suggest the granitic dikes coalesce to form a more coherent intrusion away from the sample locality, but this relationship was not mapped in detail. Other igneous rocks of this suite are chiefly volcanic rocks that make up the Post Lake or Terra Cotta, Veleska, and Sheep Creek volcanic fields (Fig. 4; Solie et al., 1991a; Bundtzen et al., 1997). The Post Lake volcanic field of Solie et al. (1991a) was renamed the Terra Cotta volcanic field by Bundtzen et al. (1997). We use the latter term to be consistent with the more recent published geologic map. The Terra Cotta volcanic rocks contain a mixture of volcanic breccia, massive dacite, and tuff. The succession also contains a distinctive lahar deposit near its base; this deposit contains meter-sized blocks of dark, fine-grained Kahiltna assemblage strata together with volcanic fragments (Bundtzen et al., 1997). Two dacite samples from the southern end of the Terra Cotta volcanic field were dated by K-Ar at 41.1 ± 2.4 and 31.4 ± 1.8 Ma (Post Lake volcanic samples of Solie et al., 1991a). We sampled a crystal-rich tuff from the northern end of the field that produced a zircon U-Pb age of 37.6 ± 0.4 Ma (E14; Table 1). The Veleska Lake volcanic field is ∼45 km northwest of the Terra Cotta field and caps a series of peaks west of the Post River, an area covering ∼15 km. The volcanic strata are essentially flat lying, more than 600 m thick, and emplaced on top of folded, steeply dipping Farewell terrane strata. Solie et al. (1991a) reported a K-Ar date of 39.4 ± 2.4 Ma for a quartz porphyry intrusion at the southern end of the volcanic field together with ages for dikes ranging from ca. 39–21 Ma. We sampled dacitic volcanic rocks from the northern end of the field; these rocks yielded a zircon U-Pb age of 40.9 ± 0.7 Ma (E8; Table 1), consistent with the age of the intrusion to the south and indicating that they were co-magmatic. In the Sheep Creek volcanic field to the northwest, Bradley et al. (2017) reported a zircon U-Pb age of 56.1 ± 1.9 Ma (01ADw7a; Table 1) that is consistent with the K-Ar ages of the nearby Sheep Creek intrusion (Solie et al., 1991a). Bradley et al. (2017) also reported a zircon U-Pb age of 42.8 ± 0.5 Ma (01ADw8A) for a felsic tuff from the highest point at the southern end of the same field, indicating that the Sheep Creek volcanic field was emplaced during two magmatic episodes. Solie et al. (1991a) reported K-Ar ages ranging from 49.0 ± 6.8 Ma to 41.7 ± 2.6 Ma for mafic and intermediate volcanic rocks from the field.

In the southeastern part of the study area and in the Susitna Basin (Fig. 4), siliciclastic rocks were deposited in the Susitna and Cook Inlet basins at the same time as igneous rocks of the Eocene suite were being emplaced to the west. The Eocene West Foreland Formation (Detterman et al., 1969; Magoon et al., 1976; LePain et al., 2013) is made up of pebble to boulder conglomerate, cross-stratified sandstone, and siltstone. It is interpreted to represent alluvial fan and fluvial sediment deposited unconformably on top of exhumed Late Cretaceous and Paleocene plutonic rocks (Gillis et al., 2015). Volcanic and granitic clasts are abundant and were derived locally from the magmatic suites described above (Finzel et al., 2009). Airfall and reworked tuffs within the stratigraphy signify that magmatism was ongoing during deposition (Helmold et al., 2019). One sample of andesitic tuff from east of Mount Spurr yielded a zircon U-Pb age of 43.6 ± 0.1 Ma (E3; Table 1), which agrees with depositional age constraints of ca. 48–39 Ma for the West Foreland Formation reported by Gillis et al. (2015, 2020).

Latest Eocene and Oligocene (EO; Ca. 35–27 Ma) Suite

A series of latest Eocene to Oligocene (EO) igneous intrusions are located in the western part of the study area. One notable intrusion of the EO suite is the Windy Fork pluton (WF in Fig. 4; Reed and Miller, 1980), a 15 km by 6 km granitic body exposed at the head of the Windy Fork of the Kuskokwim River in the northwestern part of the project area (WFK in Fig. 4). The Windy Fork pluton is made up of peralkaline arfvedsonite-bearing granite that has enrichments of rare-earth elements and zirconium (Bundtzen et al., 1997; Barker, 2016). Reed and Miller (1980) reported radioactive zones and anomalous U-Th values from the granite, and Gilbert et al. (1988) reported dikes and veins containing eudialyte and aegirine-augite that crosscut the granite and surrounding country rocks. Barker (2016) reported placer rare-earth element and zirconium enrichment in gravels surrounding the pluton exposures, mostly represented by eudialyte and the rare-earth element-bearing mineral chevkinite. Biotite K-Ar ages range from ca. 31 and 30 Ma for granite from the pluton (summarized in Wilson et al., 2015) to ca. 24 Ma on aegirine augite from peralkaline granite (Gilbert et al., 1988). Bradley et al. (2017) dated zircon from granite exposed in the northern end of the pluton at 31.8 ± 0.4 Ma (11ADw117a; Table 1), and we determined zircon U-Pb ages of 31.6 ± 0.4 Ma and 30.8 ± 0.5 Ma for samples from the western and eastern sides of the pluton (EO4 and EO5; Figs. 2 and 4; Table 1). Bradley et al. (2017) also reported a similar zircon U-Pb age of 30.9 ± 0.6 Ma (11ADw115a; Table 1) for a smaller satellite pluton to the southeast called the Windy Fork annex (Figs. 2 and 4), which we interpret to be part of the same overall intrusion.

Other plutonic bodies of this suite are exposed to the south and southeast of the Windy Fork pluton. On the western side of the Tired Pup batholith (Fig. 4), a ∼4.5-km-wide area of granite yielded zircon U-Pb ages of 31.9 ± 0.7 Ma and 30.4 ± 0.5 Ma (EO3 and EO6; Table 1) within the larger Paleocene body. The extent of the Oligocene granite was not mapped in detail, but we interpret it to be a younger phase that cuts the Tired Pup batholith. In the northern part of the Revelation Mountains (RM in Fig. 4) is another small body of granite that yielded a zircon U-Pb age of 34.8 ± 0.5 Ma (EO1; Table 1). Two additional plutons associated with this suite are mapped at the northern end of the eastern prong of the Tired Pup batholith (Fig. 4). One, called the NEP pluton (Northeast Prong pluton; Gamble et al., 2013), is biotite granite that yielded a zircon age of 29.4 ± 0.4 Ma (EO7; Fig. 12D; Table 1). The other is a coarse-grained, biotite-bearing gabbroic body on the western side of Sled Pass called the Sled Pass gabbro. It yielded a zircon U-Pb age of 28.5 ± 0.5 Ma (EO9; Fig. 12B; Table 1). We observed an extensive network of granitic sills intruding the Sled Pass gabbro near the sampled locality; this network contains abundant meter-sized, rounded bodies of the gabbro entrained in the leucocratic granite (Figs. 12A and 12C). The Sled Pass gabbro is mapped in contact with the southern part of the NEP pluton. Although our zircon date for the NEP pluton is slightly older than the Sled Pass gabbro, multiple K-Ar hornblende and biotite ages ranging from ca. 27–26 Ma (summarized in Wilson et al., 2015) in other parts of the NEP pluton suggest that it was emplaced over a period of ∼3 million years.

The only other known igneous rocks of this age in the western Alaska Range are at Dinglishna Hill, a low rise between Mount Susitna and the Susitna River to the east (Fig. 4). The hill is underlain by andesite tuff that yielded a preliminary zircon U-Pb age of ca. 32 Ma (EO2; Table 1).

Magmatic and Tectonic Evolution of the Western Alaska Range

The igneous suites described above represent more than ∼80 million years of overlapping magmatic activity associated with plate interactions—mainly subduction—along the southern Alaska margin. The annotated age-probability diagram shown in Figure 13 illustrates the temporal relationships between magmatism in the western Alaska Range and tectonic events described below, and Figure 14 shows the spatial distribution of magmatism and other key features during emplacement of the different igneous suites in the study area. We also highlight distinct styles of mineralization associated with the different magmatic suites. There was a hiatus in the early Eocene (ca. 51–45 Ma) and another beginning near the end of the Oligocene (ca. 27 Ma), but the magmatic history otherwise represents a continuum along an evolving continental margin from the middle Cretaceous through the Oligocene. Where obvious age breaks are not present, the igneous suites described above are chiefly defined by their spatial distribution, crosscutting relationships, country rock associations, and/or the presence and style of mineralization.

One of the key periods in the tectonic evolution of the southern Alaska margin was the transition from the Early Jurassic Talkeetna arc to middle to Late Cretaceous magmatism that preceded and overlapped with final accretion and translation of the Peninsular and Wrangellia terranes (Fig. 15). Peninsular terrane magmatism (“Talkeetna arc”) extended from ca. 207–162 Ma, and the arc was partially exhumed in the latest Jurassic to earliest Cretaceous (Reed et al., 1983; Trop et al., 2005; Amato et al., 2007b; Rioux et al., 2007). Magmatism then ceased along much of the southern Alaska margin during the Early Cretaceous (Fig. 15B). Emplacement of the ELC suite beginning as early as ca. 122 Ma is interpreted to represent the end of the hiatus and renewed subduction-related magmatism within the Peninsular and Wrangellia terranes. Deposition in the Cook Inlet forearc basin of south-central Alaska was also renewed at about the same time following a ∼30 m.y. hiatus (LePain et al., 2013). In the study area, ELC suite magmatism was relatively localized, although it is possible that other unrecognized parts of the ELC suite could have been obscured by younger rocks and/or eroded away. ELC suite plutons were only emplaced into the southeastern part of the Kahiltna assemblage that was deposited on the northern, inboard side of the Peninsular and Wrangellia terranes (Fig. 15C). Abundant ca. 120–94 Ma detrital zircon in the upper Kahiltna succession (Old Man petrofacies unit; Box et al., 2019) are interpreted to represent local volcanic input in addition to abundant Jurassic age populations derived from the exhumed, eroded Talkeetna arc (Box et al., 2019). However, volcanic rocks associated with the ELC plutonic suite are only preserved as the ca. 122 Ma rhyolite roof pendant to the younger Tordrillo Mountains batholith and as ca. 98–95 Ma volcanic clasts in conglomerate of the Kahiltna assemblage north of the Hayes River (Layer and Solie, 2008). The ca. 122 Ma rhyolite sample from within the Paleocene batholith yielded a biotite 40Ar/39Ar age of 60.4 ± 1.6 Ma (ELCA1b; Table 2), suggesting that argon systematics were partially reset in the sample during emplacement of the surrounding Paleocene batholith. This relationship raises the possibility that other volcanic rocks that have ca. 60 Ma K-Ar or 40Ar/39Ar ages and are spatially associated with the large batholith might be older successions that were thermally reset.

Magmatism and deformation were widespread ca. 120–90 Ma throughout the northwestern part of the Yukon-Tanana terrane and structurally underlying parautochthonous Yukon-Tanana assemblage (YTA) as the Jurassic to middle Cretaceous orogen was exhumed along orogen-parallel extensional faults and shear zones (Fig. 15C; Hansen and Dusel-Bacon, 1998; Staples et al., 2014). In contrast, ELC magmatism is not documented in the Farewell terrane in the study area. Ca. 120–94 Ma detrital zircon are common in the upper succession of the northwestern Kahiltna assemblage (Whistler petrofacies unit; Box et al., 2019), and they are interpreted to have been derived exclusively from northern sources such as the Yukon-Tanana terrane and environs (Fig. 15C). Ca. 97 Ma metamorphic titanite in a deformed gabbroic dike near the northern Kahiltna contact suggests that regional shortening chiefly affected the lower part of the succession of the northern sequence and the underlying Farewell terrane strata (Fig. 13A; Box et al., 2019). Otherwise, we infer that the Peninsular and Wrangellia terranes remained outboard (i.e., south) of the continental margin during this time, separated by a deep-water basin containing sediment wedges extending into it from each side. Given evidence for contemporaneous magmatism in both regions and deformation that propagated across the continental margin and into the Selwyn Basin, we prefer a model where the Peninsular-Wrangellia terranes were tectonically linked to the Yukon-Tanana upland of eastern Alaska in the middle Cretaceous in a cross-sectional geometry shown schematically in Figure 15C (adapted from Mair et al., 2006, and Trop and Ridgway, 2007). When the Farewell terrane was finally accreted to the western continental margin is not clear, but it was not apparently adjacent to the Peninsular and Wrangellia terranes until later.

Relative to the ELC suite, the ca. 90–75 Ma MLC suite is exposed across a broader area extending from Mount Susitna in the southeastern part of the study area to the Whistler deposit in the central part of the area (Fig. 14B). As with the ELC suite, individual intrusions of the MLC suite are relatively small. Whereas earlier deformation was restricted to the Farewell terrane and lower Kahiltna strata to the north, the ca. 80–75 plutons of the MLC suite were emplaced during widespread deformation across the entire range that culminated in formation of the diachronous Kahiltna collisional basin by ca. 81 Ma (Fig. 15D; Box et al., 2019). At this time, the northern components of the Kahiltna assemblage depositionally overlapped the southern components of the assemblage, and the entire basin was deformed by fold-and-thrust-style deformation that formed kilometer-scale recumbent folds and multiple generations of penetrative cleavage (Fig. 14B; Box et al., 2019). The orientation of fabrics indicates south- to southeast-vergent thrusting (present coordinates) across the study area. In general, most plutons of the MLC suite are not internally deformed (Gross, 2014; Hames, 2014). However, sheared and boudinaged ca. 80 Ma dikes and sills along the Chickak River (CR in Fig. 4) are interpreted to be syn- to late kinematic with respect to deformation of the surrounding Kahiltna assemblage. Approximately 10 km to the northwest, the ca. 76 Ma Whistler intrusive suite of Hames (2014) is the oldest igneous unit that crosscuts deformation features in the Whistler petrofacies unit of the Kahiltna assemblage (Box et al., 2019). This crosscutting relationship provides an important constraint on collisional deformation and links the northerly derived petrofacies to the southern petrofacies of approximately the same age (Old Man petrofacies unit; Box et al., 2019). We interpret ca. 80 Ma deformation of the Kahiltna assemblage to record final accretion of the Peninsular and Wrangellia terranes along the southern Alaska margin (Fig. 15D), and the youngest components of the ca. 90–75 Ma MLC suite mark the onset and ending of regional folding and thrusting (Fig. 14).

Plutons of the ca. 75–65 Ma latest Cretaceous (LC) suite are the oldest igneous rocks that were emplaced across both the Peninsular and Farewell terranes, and, thus, they are interpreted to stitch the newly accreted terranes and to record integration of magmatism inboard across the entire evolving margin (Fig. 14C). The semi-continuous belt of Composite suite plutons cuts across the entire outcrop belt of the Kahiltna assemblage from southwest to northeast, indicating strike-slip displacement has not been significant since they were emplaced. The age of the oldest pluton demonstrably emplaced into Farewell terrane strata is ca. 68 Ma, indicating that any large-scale lateral translation of the Peninsular and Wrangellia terranes relative to the Farewell terrane must have been completed by this time. Plutons of the MLC suite were emplaced more than 50 km inboard relative to the older magmatic suites, reflecting more complete integration of magmatism across the southern Alaska margin (shown schematically in Fig. 15D). Plutons that were emplaced through the different basement terranes (i.e., Farewell versus Peninsular-Wrangellia) have distinct hafnium isotopic signatures that confirm the juxtaposition of contrasting crustal components beneath the Kahiltna assemblage (Todd and Jones, 2017). Regional deformation is not widely recognized in the latest Cretaceous except for ductile fabrics observed along Mount Susitna. The north-south orientation of those fabrics is oblique to the regional northeast-striking structural trend established during deformation of the Kahiltna assemblage (Figs. 14A and 14B). The fabrics were possibly rotated in the hanging wall of the younger Beluga Mountain thrust (Saltus et al., 2016), but the ca. 72 Ma fabrics are approximately aligned with the north-south long axis of the coeval Mount Estelle and other plutons to the west. The Mount Estelle pluton is at an apparent oblique angle to the more north-northeast trend of the Composite suite to the north, and it is crosscut by numerous high-angle, northwest-striking, possibly sinistral faults that might have localized mineralizing fluids (Graham et al., 2013, and references therein). All of these Late Cretaceous geometries are compatible with strain patterns predicted for pure-shear dominated dextral transpression (Fig. 14C; Tikoff and Teyssier, 1994) that would have also driven westward (approximate present coordinates) translation of the accreted outboard terranes prior to final stitching ca. 68 Ma.

The Paleocene (P) magmatic suite is the most widespread throughout the study area, and it also contains the oldest coeval volcanic rocks that are well preserved and more exposed than previous suites (Fig. 14D). Plutons of this age extend for 200 km along and across the entire range and account for more than 15,000 km2 of bedrock exposure in the western Alaska Range. The Tordrillo Mountains batholith is in contact with an aerially extensive suite of coeval volcanic rocks on its eastern side, but Paleocene volcanic rocks are not preserved to the west (Fig. 14D). Unlike the older magmatic suites, published K-Ar and 40Ar/39Ar ages for hornblende and biotite from multiple localities overlap with our zircon ages, suggesting rapid cooling and/or lack of subsequent reheating during younger magmatic events (Fig. 13; Reed and Lanphere, 1973a; Wilson et al., 2015). Similar indications of rapid cooling are seen in lower temperature data from K-feldspar thermochronology and apatite fission-track studies (e.g., Benowitz et al., 2012), suggesting relatively shallow emplacement depths for Paleocene granitoids. Rocks of the Paleocene suite are not deformed internally, but regional deformation during this time is recorded in large but discrete high-strain zones in the Lake Chakachamna and Sled Pass areas (Figs. 4 and 14D).

The Lake Chakachamna shear zone is the more prominent of the two Paleocene deformation zones, and it occupies a broad, linear east-west corridor along the lake trace (Fig. 14D). Pervasively sheared ca. 98 Ma granitoids exhibit similar fabrics and kinematics up to 1 km across strike from the northern lake shore. Thin, localized cataclastic zones cut unfoliated granite 3 km north from the lake and have the same orientation and kinematics as the higher temperature fabrics to the south. Previous workers described foliated granite, gneissic granite, and/or orthogneiss along more than 16 km of discontinuous exposure along the north side of Lake Chakachamna (Hults and Wilson, 2009), indicating that the high strain zone extends along strike and is at least 1 km wide. Titanite and metamorphic zircon ages indicate metamorphism and deformation occurred at ca. 58–56 Ma during regional emplacement of the Paleocene magmatic suite, but the deformation is only observed in older middle Cretaceous plutonic rocks. The trend of the Lake Chakachamna shear zone is covered to the east by Quaternary volcanic rocks of Mount Spurr, is intruded to the west by the Eocene Merrill Pass pluton (see below), and projects to the east beneath younger brittle faults such as the Capps Glacier fault and Castle Mountain fault (Bunds, 2001; Willis et al., 2007; Gillis et al., 2009). The shear zone is also parallel to the trace of the postulated Telaquana fault, which was inferred from regional aeromagnetic data by Haeussler and Saltus (2005). The east-northeast–striking Lake Chakachamna shear zone chiefly has dextral strike-slip kinematics and evidence for a component of oblique, north-side-up displacement as well. In contrast, the north-northeast–striking Sled Pass shear zone accommodated mostly vertical displacement between plutonic rocks on the southeast and andalusite-bearing Kahiltna assemblage strata on the northwest. Both structures are compatible with pure-shear–dominated dextral transpression (Tikoff and Teyssier, 1994) with about a 30° clockwise rotation of the strain axes relative to those of the Late Cretaceous (compare strain ellipses in Fig. 14C and 14D), suggesting that oblique convergence continued into the Paleocene. Paleocene magmatism and deformation were followed by north-south mafic dike emplacement ca. 58–51 Ma across the Tordrillo Mountains batholith, suggesting further clockwise rotation of the regional strain axes. Then there was a ∼6 m.y. hiatus in magmatism across the study area interpreted to represent formation of a slab window beneath the southern Alaska margin (Fig. 13; e.g., Bradley et al., 2003b; Terhune et al., 2019).

Magmatism resumed by ca. 45 Ma as indicated by emplacement of the Eocene suite of plutons and associated volcanic deposits to the north (Fig. 13E). The Merrill Pass pluton is the most prominent intrusive body of this age exposed in the study area, and it was at least ∼50 km long as mapped by previous workers (e.g., Gamble et al., 2013). However, our new geochronology shows that the northern portion of the pluton is actually Paleocene, and the revised area of the Merrill Pass pluton is shown schematically in Figure 14E. The contact between the Merrill Pass pluton and coeval volcanic rocks on the western side is marked by abundant miarolitic cavities, indicating relatively shallow emplacement. Coeval Eocene volcanic centers to the north are more than 1 km thick as exposed. Thick lahar deposits at the base of the Terra Cotta volcanic field appear to mantle topography, and the northernmost volcanic fields are topographically higher and are essentially flat lying.

The latest Eocene to Oligocene (EO) suite is the youngest part of the Mesozoic–Cenozoic magmatic belt in the western Alaska Range, and these intrusions are relatively localized in a ∼100 km2 cluster in the western part of the study area (Fig. 14F). Emplacement of this suite was broadly synchronous with faulting on the western edge of Susitna Basin (e.g., Saltus et al., 2016). Deformation along the basin margin is attributed to counterclockwise rotation of the southern Alaska block in response to initial collision of the Yakutat microplate (Haeussler, 2008). The EO suite is slightly inboard of the Eocene suite, consistent with shallowing of the slab as the Yakutat block was subducted (Eberhart-Phillips et al., 2006). Magmatism effectively shut off in the western Alaska range after ca. 26 Ma with the exception of the two Quaternary Aleutian volcanic centers (Fig. 1).

Mesozoic and Cenozoic Tectonic Connections along the Southern Alaska Margin

The gap in magmatism between the Jurassic Talkeetna arc and the middle Cretaceous plutons that we observe in the study area (Fig. 13) is recognized throughout much of eastern Alaska and the northwestern Cordillera (e.g., Hart et al., 2004; Gehrels et al., 2009; Beranek et al., 2017). In eastern Alaska and beyond to the east and southeast, magmatism was renewed by ca. 120–115 Ma in both inboard and outboard terranes along the southern Alaska and northwestern Cordillera margin. Inboard suites are represented by the ca. 115–98 Ma Gardiner Creek batholith in eastern Alaska and western Yukon Whitehorse plutonic suite that were emplaced across parautochthonous Yukon-Tanana assemblage and the allochthonous Yukon-Tanana terrane (Fig. 15C; Richter et al., 1975; Hart et al., 2004; Allan et al., 2013; Dusel-Bacon et al., 2015). Outboard suites are represented by the ca. 123–105 Ma Nutzotin and Kluane Range magmatic belts that were emplaced into the Wrangellia and Alexander terranes (Fig. 15C; Richter et al., 1975; Hudson, 1979; Dodds and Campbell, 1988; Hart et al., 2004; Graham et al., 2016). Similar age patterns are also reported by Gehrels et al. (2009) from the extensive Coast Mountains batholith in British Columbia where there is a distinct reduction in magmatic flux between ca. 142 and 118 Ma. Tectonic models seeking to explain middle Cretaceous magmatism both inboard and outboard of the intervening basins commonly invoke paired subduction (e.g., Pavlis, 1989; Trop and Ridgway, 2007), although alternative models involving a single subduction system together with complex and widespread collisional and postcollisional intrusive complexes in the inboard regions have also been proposed (Mair et al., 2006, 2011). In the paired subduction models, debate is ongoing about subduction polarity beneath the outboard terranes as outlined in Pavlis et al. (2019) and exemplified by the associated Comment of Sigloch and Mihalynuk (2020). The schematic tectonic model in Figure 15 (adapted and revised from Mair et al., 2006) shows our preferred interpretation of the evolving plate margin configuration and eventual accretion of the outboard terranes. In this scenario, which is consistent with the model of Trop and Ridgway (2007), accretion of the Peninsular and Wrangellia terranes was accomplished by a subduction zone that dipped beneath the inboard margin of northwestern Laurentia. Subduction of the leading edge of the Peninsular-Wrangellia terranes in the Jurassic helped to drive orogenesis in the inboard region, but the remainder of the outboard terranes remained separated from the continental margin by a deep-water basin large enough to accommodate the separated Kahiltna sediment wedges (Fig. 15B). The Early Jurassic Talkeetna arc is interpreted to have formed above a second north-dipping subduction zone on the southern side of the Peninsular terrane, and the waning Jurassic magmatism, exhumation of the arc, and the Early Cretaceous magmatic lull might have resulted from shallowing of the southern (i.e., outboard) slab as the Peninsular-Wrangellia terrane began to be subducted into the northern (i.e., inboard) subduction zone and drive deformation to the north (Figs. 15A and 15B). Renewed magmatism in the late Early Cretaceous in the study area was restricted to the Peninsular-Wrangellia terranes and might have occurred during possible rollback steepening of the southern subducting slab. Slab rollback from an earlier shallow dip could have facilitated extensional collapse of the inboard Yukon-Tanana terrane collisional belt and delayed final accretion and collision of the outboard terranes and narrowing of the older Jurassic suture zone (Fig. 15C). Oblique convergence during this time could have also led to translation of the outboard terranes to the west (approximate present coordinates) relative to the inboard continental margin, eventually resulting in its juxtaposition against the Farewell terrane. Continued oblique convergence across the complex margin resulted in final closure of the Kahiltna basin ca. 80 Ma, final collision and accretion of the outboard terranes, and formation of the broader Alaska Range suture zone (Fig. 15D; Ridgway et al., 2002). Any additional translation of the outboard terranes along the suture zone was completed by ca. 68 Ma, when pluton trends of the LC suite stitched the outboard and inboard terranes and integrated across the entire margin.

Evidence for Late Cretaceous strike-slip displacement within the Kahiltna assemblage is not observed or known, and strong geological tie points are lacking. But the distinct, unrelated sedimentary wedges and the pervasive folding and thrusting of the Kahiltna assemblage raise the possibility of significant lateral separation prior to collision ca. 80 Ma. Given widespread evidence for Jurassic accretion of outboard terranes along the Cordilleran margin to the east and southeast (McClelland and Gehrels, 1990; McClelland et al., 1992; van der Heyden, 1992; Gehrels, 2001), the outboard subduction zone beneath the Peninsular-Wrangellia terrane was likely continuous along the rest of the margin. The inboard subduction zone was likely discontinuous (e.g., Pavlis, 1989) and/or was ultimately progressively consumed during Jurassic to Cretaceous collision and dextral transcurrent displacement of the outboard terranes along the Cordilleran margin (e.g., McClelland et al., 1992; Nelson et al., 2013). Box et al. (2019) showed a possible modern analog for the Kahiltna assemblage involving two separate, north-dipping (approximate present coordinates) subduction zones beneath both the outboard and inboard terranes (cf. fig. 16 of Box et al., 2019), and other tectonic models have used this same geometry (Trop and Ridgway, 2007; Staples et al., 2016).

After emplacement of the LC suite and expansion of magmatism across the newly integrated margin, we interpret the widespread distribution of the Paleocene magmatic suite to indicate continued shallowing of the subducting slab. Ca. 65–55 Ma plutonic rocks are widespread to the northeast and southwest of the western Alaska Range and were regionally named the McKinley suite for making up large parts of the Denali massif in the central Alaska Range (Lanphere and Reed, 1985). Plutonic rocks of this suite are mostly felsic, and their broad distribution more than 200 km inboard from the inferred margin is interpreted to reflect a relatively shallow slab and widespread, low-percentage partial melting of lower crust of both the Peninsular and Farewell terranes (Todd and Jones, 2017). Development of ductile deformational fabrics along two conjugate corridors ca. 58–56 Ma together with emplacement of extensive north- and north-northwest–striking dikes in the Tordrillo Mountains suggests continued oblique convergence and dextral transpression and a regional strain field that was rotated clockwise relative to the Late Cretaceous strain field (Fig. 14). Shallowing of the slab beneath the western Alaska Range in the Paleocene is consistent with regional tectonic models that invoke progressive subduction of younger, more buoyant oceanic lithosphere associated with an approaching spreading ridge (e.g., Bradley et al., 2003b; Terhune et al., 2019). We interpret latest Paleocene and early Eocene mafic dikes to record the waning magmatism above the subducting ridge, and the age of the dikes compares well with the ca. 55 Ma position of a postulated triple junction shown in Bradley et al. (2017). Following emplacement of the youngest dikes ca. 51 Ma, magmatism in the western Alaska Range effectively ceased, and a slab window replaced the subducting slab until the Eocene.

Magmatism was renewed in the western Alaska Range in the Eocene, producing a magmatic belt that continues to the southwest down the Alaska Peninsula and to the northeast where it underlies Mount Foraker in the central Alaska Range (Reed and Lanphere, 1973b, 1974; Regan et al., 2020). Eocene magmatic rocks overlap in age with the oldest known volcanic rocks in the Aleutian-Meshik arc to the southwest (Wilson, 1985; Jicha et al., 2006), indicating that Eocene magmatism records renewed subduction along a margin that has persisted for the past ∼46 m.y. Renewed subduction and formation of the Eocene arc resulted from a major change in plate motion ca. 47 Ma when the Pacific plate shifted to its present west-northwest direction relative to the deep mantle (Duncan and Keller, 2004; Sharp and Clague, 2006). Magmatism in the western Alaska Range persisted from the middle Eocene through the Oligocene and emplacement of the EO suite and then shut down again ca. 27 Ma as the Yakutat block began to subduct beneath the southern Alaska margin. Subduction of the thick oceanic plateau influenced patterns of magmatism, sedimentation, and deformation on both sides of the microplate (Eberhart-Phillips et al., 2006; Haeussler et al., 2008; Benowitz et al., 2012; Finzel et al., 2015; Lease et al., 2016; Helmold et al., 2019). The EO suite represents the western side of a gap in arc volcanoes between the western Alaska Range and the Wrangell Mountains in eastern Alaska that formed in the Oligocene and persists to this day (e.g., Wech, 2016; Brueseke et al., 2018).

Regional Metallogenic Implications

Part of the motivation for this study was to investigate the relationship between igneous activity and metallogenesis in the western Alaska Range. Numerous known mineral deposits, prospects, and occurrences in the project area include a variety of commodities such as gold, copper, molybdenum, and rare-earth elements (Alaska Resource Data File; https://ardf.wr.usgs.gov/). Some of the notable examples are mentioned above, and a more comprehensive summary of the regional metallogenic framework is provided by Graham et al. (2013) and Kreiner et al. (2020). Of particular interest was the possibility that a porphyry copper-gold-molybdenum system similar to the Pebble deposit might be present in the western Alaska Range. The Pebble deposit is located ∼200 km southwest of the project area (Fig. 1). It is one of the largest porphyry deposits known and was only discovered in the late 1980s (Lang et al., 2013). Mineralization occurred ca. 90 Ma in association with porphyry plutons of the Kaskanak batholith and other associated intrusive rocks that were all emplaced into the Kahiltna assemblage (Lang et al., 2013). Anderson et al. (2013) noted a northeast-striking array of aeromagnetic anomaly highs that projected from the Pebble deposit into the study area. Ca. 90 Ma plutonic rocks crop out in a discontinuous belt extending from southwest of the Pebble deposit to the Talkeetna Mountains (TM in Fig. 1; Bleick et al., 2012) and also extend across strike more than 300 km from the inferred margin (Bradley et al., 2017). Documented mineral occurrences including Pebble coincide with multiple anomaly highs along the linear trend from the Aleutian Range into the western Alaska Range, and Anderson et al. (2013) broadly interpreted the anomalies to reflect a Late Cretaceous magmatic arc, favorable for additional discoveries of porphyry copper systems.

Our findings and the discussion above show that, despite its regional extent, ca. 90 Ma igneous activity is not widespread in the exposed bedrock record in the study area. The ca. 99 Ma Shell Hills pluton coincides with the magnetic anomaly high in the western Alaska Range (Anderson et al., 2013), and it overlaps in age with some of the pre-mineralization intrusions surrounding the Pebble deposit. However, the Shell Hills pluton is relatively small as exposed and is made up of homogeneous granite that does not have any observed alteration. Coeval, fine-grained volcanic rocks that cap the pluton suggest that is was emplaced at relatively shallow levels, and ca. 98–97 Ma 40Ar/39Ar ages in the volcanic rocks and granite indicate rapid cooling (Table 2). The relatively small size and absence of mineralization in the Shell Hills pluton does not preclude the possibility of a larger mineralized intrusive system beneath the present level of exposure, but any such system would be buried beneath the western Susitna Basin (Stanley et al., 2014) and/or overthrust by the basin-bounding structures (e.g., Saltus et al., 2016). Other plutons in the study area that were broadly contemporaneous with the Pebble system contain pervasive ductile fabrics and have Paleocene K-Ar and 40Ar/39Ar ages, suggesting that they experienced temperatures of ∼400 °C during Paleocene deformation, likely at mid-crustal depths. Thus, we infer that the ca. 98 Ma plutons in the southern part of the study area were emplaced and resided at crustal levels in the Late Cretaceous that are generally too deep for porphyry-style mineralization until the Paleocene (e.g., Seedorff et al., 2008). Also, ELC plutons emplaced along Lake Chakachamna were possibly emplaced at shallower depths but were more deeply buried, metamorphosed, and deformed during Late Cretaceous deformation of the Kahiltna assemblage.

The main episodes of mineralization in the study area were associated with the MLC and LC intrusive suites between ca. 77 and 67 Ma (Figs. 14B and 14D; Graham et al., 2013). The Whistler deposit and Island Mountain prospect are both porphyry gold-copper systems hosted in ca. 77–76 Ma diorite intrusions (Kreiner et al., 2020; Graham et al., 2013; Gross, 2014; Hames, 2014). The Whistler deposit is part of a volcano-plutonic complex that overlies and intrudes the Kahiltna assemblage, and it is thought to represent the high-level portion of a much larger intrusion beneath the surface (Graham et al., 2013). The Island Mountain prospect is characterized by magmatic and hydrothermal breccias that also formed in the Kahiltna assemblage (Gross, 2014). The intrusive rocks associated with both Whistler and Island Mountain are thought to be oxidized arc-related magmas, and the Whistler deposit contains an oxidized alteration assemblage (Hames, 2014). In contrast, the Island Mountain prospect contains more reduced and lower sulfidation-state sulfide assemblages (Gross, 2014), possibly because of increased fluid interaction with carbonaceous Kahiltna assemblage rocks. Nevertheless, both ca. 77–76 Ma porphyry systems contrast with the younger Mount Estelle system only a few kilometers to the west; this system is associated with the Mount Estelle pluton. The ca. 71–67 Ma pluton is more than 50 km long as exposed and is crosscut by at least ten gold-bearing mineralized prospects (Fig. 14D; Graham et al., 2013). The pluton is distinctly elongated in a north-south direction, and mineralization occurs in association with steeply dipping veins and sheeted vein arrays that strike north-northwest.

The transition from more oxidized porphyry systems ca. 77–76 Ma to an intrusion-related gold system ca. 71–67 Ma broadly coincides with the termination of regional deformation associated with collisional accretion of the Peninsular and Wrangellia terranes. The Mount Estelle pluton was emplaced across a relatively broad swath of the deformed Kahiltna outcrop belt, raising the possibility that the structurally interleaved and thickened sedimentary sequences contributed to the geochemistry of the Mount Estelle pluton, including the reduced metallogenic signature of related prospects (Graham et al., 2013). More broadly, faulting and possible imbrication of the continental Farewell terrane and more oceanic Wrangellia and Peninsular terranes along and across the inferred, underlying suture zone would be expected to produce a more heterogeneous crustal column; Late Cretaceous and younger magmatic suites may have interacted with and been emplaced within this column.

Mineralization associated with Paleocene and younger magmatic suites is widespread throughout the area but is more strongly localized in skarns and along polymetallic veins. This relationship is likely a consequence of the more distributed nature and relatively shallow emplacement depths of the Paleocene suite. The presence of rare-earth element mineralization in the Oligocene Windy Fork pluton is a product of its peralkaline chemistry, and the age range of the latest Eocene and Oligocene suite corresponds with shallowing of the subducting slab as collision of the thick Yakutat microplate commenced. However, the relationship between the peralkaline character of the Windy Fork pluton and the broader tectonic environment is not completely understood.

Cretaceous to Oligocene igneous rocks in the western Alaska Range represent ∼80 m.y. of overlapping magmatism and provide key insights into the tectonic evolution and accretionary history of the southern Alaska margin. New field observations and geochronology allow us to define six igneous suites ranging from ca. 104–26 Ma. The oldest suite (ELC; >90 Ma) records renewed subduction-related magmatism following a regional hiatus through most of the Late Jurassic and Early Cretaceous. The ELC suite also coincides with an important episode of porphyry-style mineralization along strike to the southwest, but plutons of this age in the study area were not mineralized and showed no significant signs of alteration. A middle Late Cretaceous suite (MLC; ca. 90–75 Ma) overlaps with a major episode of regional deformation involving south-vergent folding and thrusting of the Kahiltna assemblage country rocks. Ca. 80–76 Ma deformation is interpreted to represent final collision, accretion, and translation of the outboard Peninsular-Wrangellia terrane against the inboard Yukon-Tanana assemblage and Farewell terrane. The youngest plutons of this suite contain porphyry-style mineralization and are interpreted to represent the continuation of subduction-related magmatism through a thickened wedge of deformed Kahiltna assemblage strata following collisional accretion of the outboard terranes. A latest Cretaceous magmatic suite (LC; ca. 75–65 Ma) contains the oldest plutons that extend across the entire Kahiltna assemblage and “stitch” the outboard Peninsular-Wrangellia terranes and inboard Farewell terrane. The orientation of plutons and deformation during this time indicate dextral transpression along the margin. Multiple mineralized prospects are associated with the LC suite including an array of prospects along the north-south Mount Estelle pluton that all have reduced metallogenic signatures. A Paleocene magmatic suite (P; ca. 65–55 Ma) is widespread across the study area and includes major batholiths such as the Tordrillo Mountains and Tired Pup batholiths and thick, aerially extensive volcanic successions. Plutonic rocks of this suite are mostly felsic, and their broad distribution more than 200 km inboard from the inferred margin is interpreted to reflect shallowing of the slab and widespread, low-percentage partial melting of lower crust of both the Peninsular-Wrangellia and Farewell terranes. Development of ductile deformational fabrics along two conjugate corridors ca. 58–56 Ma suggests continued oblique convergence and dextral transpression. A prominent suite of ca. 57–51 Ma mafic dikes were emplaced across much of the Tordrillo Mountains batholith during and following emplacement of the Paleocene suite. The dike suite generally strikes north and north-northwest and marks the last magmatism before a pronounced ∼6 m.y. hiatus through the end of the Paleocene. Shallowing of the slab and complete shutdown of magmatism in the Paleocene is consistent with regional tectonic models that invoke a slab window either by progressive subduction of a spreading ridge sweeping west to east along the margin or by subduction of a spreading ridge that was parallel to the margin.

Magmatism was renewed in the Eocene, producing an Eocene suite (E; ca. 45–37 Ma) that includes multiple volcanic fields and plutons that are observed to intrude their contemporaneous volcanic cover. Eocene magmatic rocks overlap in age with the oldest known volcanic rocks in the Aleutian-Meshik arc to the southwest, indicating that the Eocene suite records renewed subduction along a margin that has persisted for the past ∼46 m.y. The youngest magmatic suite in the western Alaska Range is latest Eocene to Oligocene (EO; ca. 35– 27 Ma), and emplacement of this suite was broadly synchronous with faulting on the western edge of Susitna Basin. Deformation along the basin margin is attributed to counterclockwise rotation of the southern Alaska block in response to initial collision of the Yakutat microplate. The EO suite is slightly inboard of the Eocene suite, consistent with shallowing of the slab as the Yakutat block was subducted. Magmatism was then effectively shut off in the western Alaska Range for ∼26 m.y. until the development of two Quaternary Aleutian volcanic centers. Collectively, these Cretaceous and Paleogene magmatic suites record the transition from a complex, discontinuous, and heterogeneous accretionary margin to an integrated continental arc that persists to this day.

This work was funded by the U.S. Geological Survey (USGS) Mineral Resources Program. The ideas presented herein were informed by discussions with Julie Dumoulin, Richard Lease, Jonathan Caine, Doug Kreiner, Bob Gillis, Jeff Benowitz, and Sean Regan. Assistance at the Stanford-USGS SHRIMP-RG laboratory was provided by Jorge Vazquez, Matt Coble, and Eric Gottlieb. We appreciate reviewers Jeff Amato, Jay Chapman, and Marwan Wartes for constructive suggestions on an earlier version of the manuscript. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

1Supplemental Material. Complete descriptions of all U-Pb and 40Ar/39Ar analytical methods together with detailed analytical data for all samples. Please visit https://doi.org/10.1130/GEOS.S.13133177 to access the supplemental material, and contact editing@geosociety.org with any questions.
Science Editor: Andrea Hampel
Guest Associate Editor: Jason W. Ricketts
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.