The Upper Cretaceous Valdez Group represents the flysch facies of the Mesozoic Chugach terrane accretionary complex in southern Alaska. The Valdez Group is dominated by litharenite sandstone and argillite deposited as coherent beds, unlike the older McHugh Complex mélange and massive sandstones. Detrital zircons from five sandstones sampled along an ∼55 km transect through the Valdez Group were dated using U-Pb laser ablation–multicollector–inductively coupled plasma–mass spectrometry (LA-MC-ICP-MS). The youngest populations from the two oldest samples, located along strike from each other, were 82–81 Ma. Three samples across strike and outboard of the others are separated by ∼50 km, but each has a youngest population dated at ca. 68 Ma. All of these samples have major grain population ages that suggest erosion from the Coast Mountains Batholith, consistent with petrography and grain modes suggesting an arc source. No apparent age gap exists between the youngest McHugh Complex samples and the oldest Valdez Group samples, suggesting continuous deposition despite the different depositional and tectonic style. We propose a model in which the onset of coherently bedded flysch marks the transition from deposition in the trench or trench slope to deposition on the oceanic plate beyond the trench after it was filled at ca. 84 Ma, i.e., the time of the youngest mélange sedimentation. Preservation of coherent bedding resulted as large coherent blocks of Valdez Group rocks were imbricated into the subduction complex during continued subduction in Paleogene time.


Active continental margins are complicated tectonic settings, but they are recognized as the primary locations at which continental crust is both created and recycled back to the mantle (e.g., Clift et al., 2009). Although several factors control whether a margin is tectonically erosive or accretionary, the flux of sediment to the trench is known to be a particularly important variable (e.g., von Huene and Scholl, 1991; Clift and Vannucchi, 2004). Margins of either type are not steady state over periods of more than a million years, and in accretionary margins, crustal growth may have been intermittent because of seamount or ridge collisions. If we are to quantify mass recycling processes at the margin or global scale, then we need to know the ways in which accretionary wedge construction can vary through time. Moreover, the sedimentary rocks in an accretionary wedge not only tell us how fast crustal growth was occurring, but provenance studies can shed light on the regional tectonics of the margin and show how changes in trench dynamics are linked to processes within the wider forearc, the magmatic arc itself, or elsewhere in the continental interior (Amato and Pavlis, 2010; Dumitru et al., 2010). Certainly, the sedimentary record deposited at active plate margins holds out the best prospect of reconstructing past tectonic or magmatic activity in these settings.

While some sediment accumulates in perched forearc, trench-slope basins, accretionary complexes can also potentially provide a detailed record of mass flux to the convergent plate margin. These bodies are constructed of sedimentary rock, typically mostly eroded from the upper plate (i.e., the arc and forearc), as well as lesser volumes of deep-marine sedimentary rocks that have been offscraped from the subducting plate. However, dating sedimentary rocks in accretionary complexes has historically been difficult because of sparse fossil assemblages and the possibility that the fossils that are present represent the time of sedimentation on the oceanic plate, when it was still far from the margin, and thus are much older than the actual accretionary event. Detrital zircon ages can allow for a more continuous record of uplift and erosion of the arc region. If the time between erosion and ultimate deposition is assumed to be brief, as is likely in a rapidly uplifting margin with an overfilled, bypassed, or absent forearc basin, then the crystallization age of zircons in trench and trench-slope basin deposits may closely approximate the depositional age of strata.

The Upper Cretaceous Valdez Group on the southern coast of Alaska is the youngest part of the Chugach terrane accretionary complex (Plafker et al., 1994). The Valdez Group originated as a system of submarine fans that received sediment derived from active magmatic arcs and an exhumed batholith (Nilsen and Zuffa, 1982; Plafker et al., 1994). This sediment was deposited into and beyond the paleo–Aleutian Trench and was subsequently accreted onto the Alaskan margin before the Eocene (Fig. 1; Nilsen and Zuffa, 1982; Dumoulin, 1987; Plafker et al., 1989). The depositional interval of the Valdez Group is constrained to Campanian–Maastrichtian (83.5–65.5 Ma; time scale of Walker and Geissman, 2009) by a few fossil locations in the Valdez Group (Jones and Clark, 1973). Correlative and coeval formations (e.g., the Kodiak Formation, Shumagin Formation, Sitka Graywacke, and Matanuska Formation) are considered correlative and coeval to the Valdez Group based on similar fossils and detrital zircon ages that also indicate depositional intervals in the Campanian–Maastrichtian (see summary in Plafker et al., 1994; Haeussler et al., 2004).

In this study, we attempt to reconstruct the evolution of the active continental margin of southern Alaska during the Cretaceous–Paleogene. We aim to understand the ways in which the varying tectonic and magmatic processes in the subduction system have controlled the rates and styles of tectonic subduction accretion. We present U-Pb detrital zircon ages and sandstone petrography of the Valdez Group along Turnagain Arm near Anchorage, Alaska (Fig. 2). These data are used to: (1) compare ages to fossil data that indicate a maximum depositional age of the Valdez Group in the Campanian–Maastrichtian (83.5–65.5 Ma; time scale of Walker and Geissman, 2009); (2) test previous models suggesting derivation of material from multiple source regions; (3) determine if a period of tectonic erosion separated deposition of the McHugh Complex and Valdez Group; and (4) develop a new model for the deposition of the McHugh Complex and Valdez Group.


The Southern Margin composite terrane of southern Alaska is an amalgamation of three terranes, the Chugach, Ghost Rocks, and Prince William terranes, each consisting of marine sedimentary rocks with variable amounts of metamorphosed sedimentary rocks and numerous Paleogene igneous intrusions (Plafker et al., 1989). The Wrangellia composite terrane consists of the Wrangellia, Peninsular, Alexander, and Taku terranes, which include magmatic arc, oceanic, and rift-fill assemblages (Plafker et al., 1989). The Southern Margin composite terrane is located south of the Border Ranges fault system, a complex structure with both dip-slip and strike-slip history (MacKevett and Plafker, 1974; Little, 1988; Pavlis and Roeske, 2007) and north of the modern Aleutian Trench in the Gulf of Alaska (Fig. 1). Total dextral offset of the Chugach terrane may be 600 km or more (Pavlis and Roeske, 2007), and its original location may have been farther south proximal to the Coast Mountains of western Canada and southeastern Alaska (van der Heyden, 1992).

The Mesozoic volcanic complexes that were active during the formation of the Chugach terrane include the Talkeetna arc, Chitina arc, Chisana arc, and the Coast Mountains Batholith (Fig. 1; Berg et al., 1972; Hudson, 1983; Plafker et al., 1994; Monger et al., 1994). The Talkeetna volcanic arc of the Peninsular terrane was active from the Late Triassic to Middle Jurassic as a result of left-oblique subduction of the Farallon plate under the Alaskan margin (Plafker et al., 1994). The Chitina arc was active mainly in the Late Jurassic and spans the margin of the northern Wrangellia and Alexander terranes (Plafker et al., 1989; Trop et al., 2002, 2005). The Chisana arc was active in Early Cretaceous time and is also located on the margins of the Alexander and Wrangellia terranes (Berg et al., 1972; Trop et al., 2002, 2005). The Coast Mountains Batholith, located in southeastern Alaska and southwestern British Columbia, is a Middle Jurassic to Eocene body interpreted to have likely contributed sediment to the paleo–Aleutian Trench from the Middle Jurassic to Paleocene (Monger et al., 1994).

These arcs are interpreted to be the source of abundant volcaniclastic detritus deposited as submarine fans with intertonguing oceanic pelagic sediment and minor basalt (Nilsen and Zuffa, 1982). Some of this material is preserved in the Mesozoic forearc in the Matanuska Valley–Talkeetna Mountains Basin (Trop and Ridgway, 2007; Trop, 2008). The Matanuska Valley–Talkeetna Mountains Basin is a belt of Middle Jurassic–Upper Cretaceous marine and Paleocene–Oligocene terrestrial sedimentary strata unconformably overlying igneous rocks of the Talkeetna arc (Trop and Ridgway, 2007). The remainder of this sediment was transported either northward into late Mesozoic syncollisional basins (Gravina and Kahiltna Basins) or offshore to the Pacific Ocean floor, where it was swept into the North Pacific trench system to form the Southern Margin composite terrane.

Chugach Terrane

The Mesozoic accretionary complex of southern Alaska is known as the Chugach terrane (Plafker et al., 1989). The Chugach terrane developed along the southern edge of the Wrangellia composite terrane and contains three fault-bounded units (Fig. 3; Plafker et al., 1989): a Triassic–Jurassic blueschist unit, a Jurassic–Cretaceous mélange unit, and a Cretaceous–Eocene flysch unit (Berg et al., 1972; Manuszak et al., 2007). The boundaries of the Chugach terrane are the Border Ranges fault system to the north and the Contact fault system to the south (Plafker et al., 1989). Within the Chugach terrane, the Eagle River fault system separates the Jurassic–Cretaceous mélange unit (locally the McHugh Complex) from the Cretaceous–Eocene flysch unit (locally the Valdez Group; Winkler, 1992).

The two oldest units of the Chugach terrane are blueschist-facies metamorphic rocks and mélange rocks that are found along and near the Border Ranges fault system. These are interpreted as an oceanic assemblage of basalts and marine sedimentary rocks that were underthrust and metamorphosed during the Jurassic–Cretaceous (Plafker and Berg, 1994; López-Carmona et al., 2011). The mélange unit, which is known as the McHugh Complex in the Chugach and Kenai Mountains, is located outboard of the blueschist unit, but in some localities where blueschists are absent it lies directly against the Border Ranges fault (Fig. 1). The McHugh Complex consists of massive sandstones, conglomerates, and metavolcanic rocks, as well as chert, argillite, and green-tuffaceous rocks that are interleaved in a complex fashion, and that are all stratally disrupted into a mélange assemblage (Clark, 1973; Amato and Pavlis, 2010). The McHugh Complex has been interpreted to record the history of accretion during the Jurassic and Early Cretaceous and (Plafter, 1989; Amato and Pavlis, 2010). Amato and Pavlis (2010) determined the maximum depositional age of the mélange and massive sandstones from U-Pb detrital zircon analyses to be 157–146 Ma and 100–84 Ma, respectively.

The Chugach terrane is interpreted to have a history of both subduction accretion and tectonic erosion (Clift et al., 2005; Pavlis and Roeske, 2007; Amato and Pavlis, 2010). Detrital zircon ages show that the Chugach terrane had at least four distinct periods of accretion represented in the rock record by the blueschists, two distinct units of the McHugh Complex, and the Valdez Group (Amato and Pavlis, 2010). Amato and Pavlis (2010) postulated that these units were separated by two and possibly three periods of tectonic erosion based on gaps in the detrital zircon age progression. The oldest period of erosion corresponds to one proposed by Clift et al. (2005) between the blueschists and the oldest McHugh Complex. The strongest evidence for erosion came from a 40 m.y. age gap between two units within the McHugh Complex (Amato and Pavlis, 2010). An erosive period between accretion of the McHugh Complex and Valdez Group was postulated based on a 6 m.y. age gap between the youngest detrital zircon ages from widely spaced samples in each unit and on their disparate depositional styles (Amato and Pavlis, 2010).

The Valdez Group is the youngest part of the Chugach terrane. It is a lithologically homogeneous belt of Cretaceous sedimentary rocks extending ∼2000 km along strike, and it is as much as 100 km wide (Nilsen and Zuffa, 1982; Dumoulin, 1987; Plafker et al., 1994). Nilsen and Zuffa (1982) and Dumoulin (1987) described the composition of the Valdez Group in the Kenai and Chugach Mountains as largely flysch deposits of sandstone and slate, with minor amounts of conglomerate, mafic intrusive rocks, and volcanic rocks. The Valdez Group is interpreted to be correlative with the Sitka Graywacke, Shumagin Formation, Kodiak Formation, and the flysch facies of the Yakutat Group, based on the similar composition of sandstones observed in thin section and diagnostic age fossils found within each formation (Plafker et al., 1994). These fossils (Jones and Clark, 1973) indicate a depositional age of Campanian to Maastrichtian (83.5–65.5 Ma; time scale of Walker and Geissman, 2009). Sparse U-Pb detrital zircon data exist for the Valdez Group to precisely determine maximum depositional age. Detrital zircons from a sandstone yielded a maximum depositional age of 78 Ma, and a granite clast within the unit has been dated at 221 Ma (Bradley et al., 2009). The potentially correlative Sitka Graywacke, near Sitka, Alaska (Fig. 1), yielded samples with youngest peak ages of 74–72 Ma (Haeussler et al., 2004). The source of the Valdez Group is interpreted to be the Coast Mountains Batholith, exposed in southeastern Alaska and western British Columbia, based on petrographic results (Nilsen and Zuffa, 1982; Dumoulin, 1987), paleocurrent measurements (Nilsen and Zuffa, 1982), and Nd, Sr, and Pb isotopic data (Farmer et al., 1993; Sample and Reid, 2003). However, Dumoulin (1987) and Nilsen and Zuffa (1982) also indicated that the Valdez Group contained variable amounts of sandstone derived from active arc and recycled lithic sources.


Seven sandstone samples were collected from near Turnagain Arm and in the Anchorage region of Alaska. Three of these samples (ANJ35, KED28, and ANJ40) were point counted to determine composition and provenance of the sandstones. Detrital zircons were collected from five of the samples (ANJ06, ANJ13, ANJ34, ANJ40, and KED24) for U-Pb analyses to determine maximum depositional age, maximum depositional interval, and provenance of sandstones in the Valdez Group.

Point counting of sandstone samples to determine composition and provenance of the sandstones was conducted using the frameworks developed by Dickinson (1970) and Folk (1974). In total, 400 individual framework sand grains were identified and counted from each thin section. Data were plotted using ternary diagrams with fields described by Dickinson et al. (1983). Grain parameters (Table 1) and modal data follow standard convention (Dickinson and Suczek, 1979; Dickinson, 1985). In the process of identifying grains, we determined that not all of the constituent grain types within the lithic volcanic category (Lv) (e.g., lathwork grains, microlitic grains, vitric grains, etc.) are readily identifiable within a given sample or in all samples equally because of alteration, devitrification, and recrystallization of the grains. All grain types that comprise the Lv category (as indicated on Table 1) were counted as simply the Lv grain type. Thin sections with 10% matrix, pseudo-matrix, or unidentifiable grains were deemed “uncountable” (samples ANJ06, ANJ13, ANJ34, and KED24) and were disregarded from the interpretation.

U-Pb zircon analyses were conducted on the Nu Plasma HR laser ablation–multicollector–inductively coupled plasma–mass spectrometer (LA-MC-ICP-MS) at the University of Arizona using methods described by Gehrels et al. (2008). Beam diameter was 30 μm, with a pit depth of ∼15 μm. Errors are reported in the text as 2σ. Analyses were conducted without the aid of cathodoluminescence images. For detailed explanation of machine and analytical specifications, as well as an explanation of LA-MC-ICP‑MS methodology, consult Gehrels et al. (2006) and Gehrels (2011). Statistical considerations indicate that at least 50 grains are required to characterize the provenance of a sand sample, and more than 100 are recommended for sands derived from multiple sources (Ruhl and Hodges, 2005).

Analyses with discordance (defined as dissimilarity of the 206Pb/238U, 207Pb/235U age determination schemes for samples younger than 1.0 Ga; defined as dissimilarity of the 206Pb/238U, 206Pb/207Pb age determination schemes within error for samples older than 1.0 Ga; Wetherill, 1956) greater than 30% and 206Pb/238U measurement error greater than 10% were discarded. A 30% discordance cutoff was used because this allows for the determination of relative proportions of grains of a wide range of ages (e.g., Cenozoic–Proterozoic; Gehrels, 2011). Systematic errors must be accounted for in each age calculation to properly incorporate analytical errors attributed to analytical drift within the machine. Systematic errors for each sample were added to the 2σ error determined for each maximum depositional age calculated.

In this study, we defined the maximum depositional age determined from U-Pb detrital zircon ages using the methods described by Dickinson and Gehrels (2009). Youngest U-Pb detrital zircon ages will be reported in several ways, listed from least to most statistically significant: (1) youngest single grain; (2) youngest graphical age peak determined from a relative probability distribution plot (Ludwig, 2003); and (3) mean age of the youngest three or more concordant, coeval grains that overlap at the 2σ level (Dickinson and Gehrels, 2009). The youngest single grain is not used to determine the maximum depositional age because it is possible that a single young grain could have experienced Pb loss. However, if this grain is included in the population of three or more concordant, coeval grains that overlap at the 2σ level, the grain could be a less conservative estimate of the maximum depositional age. The youngest graphical age peak age is not used because it tends to overestimate the maximum depositional age (Dickinson and Gehrels, 2009). Accordingly, in this study, we interpret the mean age of the youngest three or more concordant, coeval grains that overlap at the 2σ level to be the maximum depositional age of the sandstone.


Field Relationships

The majority of outcrops in the Valdez Group examined in this study were located along road cuts and coastlines. Outcrops vary significantly from interbedded argillite and sandstone, to massive sandstone, to exclusively argillite (Fig. 4). Sandstones in the Valdez Group are dark gray in color on the weathered surface and charcoal gray-black on fresh surfaces. The sandstones are well cemented and silicified and are generally cut by a network of calcite and/or quartz veins. Locally, rip-up clasts of argillite are present only at the base of turbidite beds, in contrast to McHugh Complex sandstones, in which pebble-size rip-up clasts are pervasive in sandstone beds. Rip-up clast sizes range from 2 to 20 cm. Ripple cross-laminae are exclusive to the interbedded argillite and sandstone outcrops. Grain size ranges from fine to coarse in the sandstone, with typically subangular grain shapes.

Argillites interbedded with sandstone generally have a slatey to phyllitic cleavage that varies from bed-parallel fabrics associated with tight to subisoclinal folds to weaker slatey cleavages oriented oblique to bedding in association with more open folds. Oriented quartz veins locally cut this cleavage. Argillites are typically devoid of observable grains in hand sample, but in some instances, fine- to very fine-grained sand particles are present as well as mica grains.


The Valdez Group sandstones are classified as volcanic litharenites or lithic arkoses in the description scheme of Folk (1974). All sandstones are matrix rich and some have a trace calcite cement (Fig. 5). Grain sizes range from fine to coarse and are typically angular to subangular. Detrital chert grains are present in all samples. Most thin sections exhibit sedimentological fabrics (e.g., laminations), and only one sample (KED24) exhibits a structural foliation. Feldspars are commonly sericitized, otherwise altered, or silicified. Volcanic lithic fragments are commonly recrystallized and/or diagenetically altered. Matrix and pseudo-matrix are the abundant intergranular material in all samples and are likely derived from the diagenesis of volcanic lithic fragments and K-feldspars. These findings are consistent with both the diagenetic history and low-grade metamorphism known to have affected the Valdez Group (Nilsen and Zuffa, 1982; Dumoulin, 1987). Point-counting results are summarized in Table 2.


We determined U-Pb ages of detrital zircons collected from three samples from Turnagain Arm, one sample from near the town of Whittier, and one sample from Mount Magnificent northeast of Anchorage in the Valdez Group (Table 3). Detrital zircon grains that were analyzed range in size from ∼60 μm to ∼120 μm, with a small percentage (<2%) larger. Zircon grains were largely angular to subangular, again with 10%–20% that were rounded to subrounded. Most grains have retained their prismatic shape, although some of the grains have lost this shape. Early Cretaceous– and Jurassic-age detrital zircons show increased rounding when compared to Late Cretaceous–age detrital zircons (Fig. 6).

Sample MM1 was collected from near Mount Magnificent northeast of Anchorage, Alaska (Fig. 2). Eighty of the 100 grains analyzed from this sample were determined to be concordant. The maximum depositional age of this sample was 84.8 ± 1.1 Ma, as derived from a group of 20 grains (Fig. 7). The youngest grain analyzed was dated at 80.6 ± 2.5 Ma. The majority (87.5%) of the grains analyzed were Mesozoic in age. There were nine Precambrian grains.

Sample TA1 was collected from the north shore of the western (Anchorage) end of Turnagain Arm near the mapped location of the Eagle River fault (Fig. 2). Ninety-eight of the 100 grains analyzed were determined to be concordant. The maximum depositional age was 80.7 ± 1.6 Ma from a group of three grains (Fig. 7). The youngest grain analyzed was 76.9 ± 1.0 Ma. The majority (81.6%) of the grains analyzed were of Mesozoic age, and 13 grains were Precambrian.

Sample TA2 was collected from the north shore at the Anchorage end of Turnagain Arm (Fig. 2). Forty-four of the 50 grains analyzed were determined to be concordant. The maximum depositional age was calculated as 70.7 ± 1.8 Ma from a group of nine grains (Fig. 7). The youngest grain analyzed was 66.4 ± 2.3 Ma. The majority (88.6%) of the grains were of Mesozoic age. Only three grains analyzed were Precambrian in age.

Sample TA3 was collected from the north shore at the eastern end of Turnagain Arm (Fig. 2). Forty of the 50 grains analyzed were determined to be concordant. The maximum depositional age of this sample was 68.6 ± 1.3 Ma, which was derived from a group of nine grains (Fig. 7). The youngest grain analyzed was dated at 65.6 ± 2.1 Ma. Again the majority of the grains analyzed were Mesozoic in age (84%). Only three analyzed grains were found to be Precambrian in age.

Sample W1 was collected from near the town of Whittier (Fig. 2), approximately 28 km northwest of the Contact fault system. Seventy-nine of the 100 grains analyzed were determined to be concordant. The maximum depositional age of this sample was 67.7 ± 1.8 Ma, determined from a group of seven grains (Fig. 7). The youngest grain analyzed was 61.8 ± 5.1 Ma. The majority (74.2%) of the grains were Mesozoic in age, with nine analyzed grains being Precambrian in age.

When all of the data from these samples are combined, we recognize several populations: 90–60 Ma, 105–90 Ma, 145–125 Ma, 170–150 Ma, and 200–170 Ma (Fig. 8). These Mesozoic ages constitute 85% of all of the analyses. Only 16 (5% of total analyses) grains were Paleozoic, and 39 grains (10% of total analyses) were Proterozoic or older.


Constraints on Depositional Ages

We suggest that the youngest age group in each sample closely approximates the age of deposition and subsequent accretion because: (1) the sandstones likely had a geologically rapid and simple depositional path from the source; (2) the majority of the dominant age peaks are composed of the youngest zircon grains, consistent with a source in an active arc and/or rapidly exhuming plutons; and (3) the relatively small number of Precambrian age grains is consistent with a proximal sediment source of sediment within the active margin arc. Bearing this in mind, we consider the maximum depositional age to be a suitable approximation to, or slightly older than, the age of deposition and subsequent accretion.

These detrital zircon data from the five samples in the Valdez Group indicate that deposition of the Valdez Group initiated at ca. 85 Ma (sample MM1) and persisted until ca. 68 Ma (sample W1). The only previously published detrital zircon maximum depositional age from the Valdez Group of 78 Ma from Bradley et al. (2009) falls within this age range. This depositional period is also consistent with fossil data from the Valdez Group and the correlative Kodiak Formation, which are considered to be Campanian–Maastrichtian (83.5–65.5 Ma; time scale of Walker and Geissman, 2009) based on the presence of Inoceramus fossils (Jones and Clark, 1973; Sample and Reid, 2003). Our data suggest that the younger part of the Sitka Graywacke, which has a maximum depositional age of 72 Ma (Haeussler et al., 2004), is likely coeval and correlative to the Valdez Group at Turnagain Arm, but that earlier samples with ages of ca. 105 Ma are likely correlative with the massive sandstones and conglomerates of the Upper Cretaceous McHugh Complex (Amato and Pavlis, 2010).

Provenance of the Valdez Group

Our petrographic data from the Valdez Group sandstones indicate a transitional-undissected arc source (Fig. 8). Previous studies used similar ternary discrimination diagrams and concluded that the Valdez Group was largely derived from an undissected arc locality (Nilsen and Zuffa, 1982; Dumoulin, 1987). Our petrographic data are consistent with these findings. Transitional arc sources are characterized by source terranes where the arc is active but has seen some erosion of the volcanic carapace (Dickinson et al., 1983). Undissected arc sources are characterized by source terranes where the arc is active, but where little to no erosion has occurred to remove the volcanic carapace (Dickinson et al., 1983). We interpret the presence of detrital chert grains to suggest derivation of recycled material from the McHugh Complex mélange, which contains numerous beds of chert (Clark, 1973; Amato and Pavlis, 2010).

The youngest detrital zircon grains (120–65 Ma) are found in all samples from the Valdez Group. Zircons of this age are coeval with plutonism in the Alaska Range, Talkeetna Mountains, and the Coast Mountains Batholith (Nokleberg et al., 1985; Trop et al., 2002, 2005; Ridgway et al., 2002; Gehrels et al., 2009). However, structural reconstructions of strike-slip motion along the Border Ranges fault (e.g., Pavlis and Roeske, 2007) and paleomagnetic data (e.g., Plumley et al., 1983; Cowan, 2003) indicate that the Valdez Group (and the entire Chugach terrane) was originally deposited between ∼200 km and ∼2000 km south of its modern location. Regardless of the magnitude of displacement, the Valdez Group was situated south of its present location. Therefore, it is prudent to compare detrital zircon age populations to the magmatic and tectonic history of the Coast Mountains Batholith.

Detrital zircons with ages 120–65 Ma are consistent with derivation from the Coast Mountains Batholith, based on similar ages of magmatism and uplift within the Coast Mountains Batholith (Fig. 9A). Gehrels et al. (2009) constrained the ages of plutonic activity and uplift within the Coast Mountains Batholith using U-Th-Pb zircon analyses to indicate increased plutonic activity and magmatic flux during the period of 120–65 Ma. Low-temperature thermochronology (Parrish, 1983; Gehrels et al., 2009) indicates increased uplift and rapid exhumation in the Late Cretaceous and early Paleogene. Therefore, it is most likely that the 120–65 Ma detrital zircons and the majority of the sediments in the Valdez Group were derived from the Coast Mountains Batholith. This setting is consistent with the provenance indicated by detrital composition of sandstones of the Valdez Group. This provenance is also consistent with previous conclusions that indicate the Coast Mountains Batholith as the most likely source of the Valdez Group and correlative units (Nilsen and Zuffa, 1982; Dumoulin, 1987; Farmer et al., 1993; Sample and Reid, 2003; Haeussler et al., 2004).

Detrital zircons with ages of 210–170 Ma, 170–150 Ma, and 150–125 Ma may have been derived from several potential source regions, such as the Talkeetna arc (202–153 Ma; Amato et al., 2007; Rioux et al., 2007), Chitina arc (171–140 Ma; Plafker et al., 1989; Roeske et al., 2003), Chisana arc (130–120 Ma; Berg et al., 1972; Manuszak et al., 2007), and the Coast Mountains Batholith (230–45 Ma; Gehrels et al., 2009), as well as in the Upper Jurassic–Lower Cretaceous part of the McHugh Complex (Amato and Pavlis, 2010).

Detrital chert suggests the possibility that Lower Jurassic–Upper Cretaceous McHugh Complex mélange material may have been recycled into the Valdez Group. Amato and Pavlis (2010) analyzed detrital zircons from the Lower Jurassic–Upper Cretaceous McHugh Complex mélange and determined that Jurassic-age zircons comprise virtually the entire age spectrum (ca. 180–140 Ma). The McHugh Complex Jurassic-age spectrum is visually very similar to the range of Jurassic-age detrital zircons from samples in the Valdez Group, further suggesting that the Valdez Group may have received recycled sediment from the McHugh Complex (Fig. 9B). Photomicrographs of detrital zircons indicate higher degrees of rounding in Jurassic–Early Cretaceous age detrital zircons, suggesting that these grains are not first cycle but instead have experienced recycling (Fig. 6). These data, paired with Middle to Late Cretaceous uplift of the forearc and accretionary complex suggested by Trop and Ridgway (2007), support the idea that some Jurassic-age detrital zircons could have been recycled from the Jurassic McHugh Complex. However, it is important to note that Jurassic volcanic and plutonic rocks were also recognized by Gehrels et al. (2009) in the Coast Mountains Batholith, suggesting that this unit could have supplied Jurassic-age detrital zircons to the Valdez Group.

Proterozoic- and Archean-age terranes are not present in south-central Alaska. Proterozoic-age terranes are present in parts of western Canada in British Columbia and Yukon Territories (Yukon-Tanana terrane), as well as interior Alaska (Farewell, Wickersham, Yukon-Tanana, Ruby, and Angayucham-Tozitna terranes; Gehrels and Kapp, 1998; Nelson and Gehrels, 2007; Bradley et al., 2007; Dusel-Bacon and Williams, 2009). Archean-age terranes are limited to the interior of Canada on the craton (Hoffman, 1989). Proterozoic-age detrital zircons cannot be assigned to a single source terrane because of the numerous potential source regions. Archean-age detrital zircons are likely only derived from the Canadian craton unless these grains have experienced multiple generations of recycling. Sources in western Canada are preferred over those from interior Alaska because they are most consistent with derivation of younger detrital zircon populations from the Coast Mountains Batholith region, as well as the presence of Archean-age rocks in western and central Canada. Pre-Mesozoic zircons are almost entirely absent from the older McHugh Complex inboard of the Valdez Group (Amato and Pavlis, 2010). We interpret the relative abundance of pre-Mesozoic zircons in the Valdez Group to suggest increased uplift and deformation in the retroarc region of the Coast Mountains, consistent with the uplift and deformation noted by Gehrels et al. (2009). This uplift would allow for increased erosion of older terranes inboard of the arc. Alternatively, development of fluvial systems draining the Coast Mountains orogen by headward advance into the continent could have resulted in more drainage of the continental interior and would also explain the change in provenance with time.

Depositional-Tectonic Model

We interpret our detrital zircon data, together with data from Amato and Pavlis (2010), to indicate the absence of a tectonically erosive period between deposition of the Upper Cretaceous part of the McHugh Complex and the Valdez Group. The youngest sample from the Upper Cretaceous portion of the McHugh Complex along Turnagain Arm (MS2) is 84 ± 3 Ma (Amato and Pavlis, 2010), and the oldest sample of the Valdez Group on Turnagain Arm from this study (TA1) is dated at 80.7 ± 1.6 Ma (Fig. 7). These samples are within 2σ error of one another, and we interpret them to indicate continuous deposition of the Upper Cretaceous McHugh Complex and the Valdez Group. The age gap postulated to exist (Amato and Pavlis, 2010) was based on a small sample set and low spatial sampling resolution. Detrital zircon grains with ages between ca. 146 to ca. 91 Ma are present in samples from the Valdez Group (Fig. 7), confirming the interpreted period tectonic erosion within the McHugh Complex (e.g., Amato and Pavlis, 2010). If grains of this age range were absent in the Valdez Group, this would suggest that the gap in the McHugh Complex ages was not tectonic erosion and was related to an absence of volcanism or plutonism of this age.

A plot of maximum depositional age of samples collected from along and near Turnagain Arm relative to distance from the Eagle River fault shows an apparent trend of decreasing age with increasing distance east of the Eagle River fault (Fig. 10). These data clearly show that there is no age gap between the Upper Cretaceous McHugh Complex and the Valdez Group. Furthermore, there are no apparent gaps in age within the Valdez Group that would suggest a period of erosion or nondeposition within the Valdez Group. The nearly flat trend of the data farthest east from the Eagle River fault suggests that sedimentation rates were very high, resulting in similar ages across a long distance, or that faulting and folding have repeated sections of the Valdez Group.

An Andean-type model of accretion was previously suggested for the Late Cretaceous subduction and accretionary history of the Coast Mountains (van der Heyden, 1992). We interpret the Upper Cretaceous McHugh Complex to represent sediment deposited directly into the subduction trench, likely by sediment gravity flows. If plate convergence remained constant while sediment delivery increased relative to the Jurassic–Lower Cretaceous McHugh Complex, then the trench would become filled, more closely resembling the classic tapered wedge model (e.g., Clift et al., 2009). Submarine fan systems at this point could have deposited material onto the oceanic plate outboard of the trench. Distal submarine-fan turbidite deposits (e.g., the Valdez Group) have been recognized in modern systems and are incorporated into the frontal portions of the accretionary complex by large listric thrust faults (e.g., Davis and Hyndman, 1989). Significant (>1 m thickness) shale intervals within the Valdez Group support our interpretation that the Valdez Group was deposited in a deep-water setting, where substantial hemipelagic sediment would be capable of accumulating. The interbedding of shales and sandstones represents episodic deposition by sediment gravity or density flows and the switching of depositional lobes on the slope fan surface. Thus, the Upper Cretaceous McHugh Complex and the Valdez Group were deposited without a hiatus and represent sequential parts of the same trench sedimentation system.


Detrital zircon ages collected from the Valdez Group of the Chugach accretionary complex constrain its maximum depositional age to be ca. 85 Ma, with deposition continuing until at most 68 Ma. Based on zircon ages, petrographic characteristics, and structural and paleomagnetic reconstructions, the provenance of the Valdez Group is likely the Coast Mountains Batholith. Detrital zircon ages from the oldest Valdez Group rocks suggest continuous deposition of the Upper Cretaceous McHugh Complex and Valdez Group and do not support a period of tectonic erosion between these two units during the Late Cretaceous. An approximate transect of the McHugh Complex and Valdez Group illustrates this continuity of deposition of the Upper Cretaceous McHugh Complex and Valdez Group and suggests the possibility of high erosion and sedimentation rates and/or significant repetition of sections higher in the Valdez Group. The difference between mélange-style deposition in the McHugh Complex and flysch-style deposition in the Valdez Group is based on a model in which the trench filled up as a result of increasing sediment flux, thus allowing deposition of coherently bedded flysch beyond the trench.

This work was funded by National Science Foundation grants EAR-0809608 to Amato, EAR-0809609 to Pavlis, and a Geological Society of America Graduate Student Research Grant to Kochelek. G. Gehrels and M. Pecha helped acquire detrital zircon data, and National Science Foundation grant EAR-0443387 supported the Arizona LaserChron Center. J. Hecker, C. Worthman, and E. Day helped with sample collection. We acknowledge useful discussions with T. Lawton, G. Mack, D. Bradley, and S. Karl.

1GSA Data Repository Item 2011258, U-Th-Pb isotopic data collected from laser ablation–multicollector–inductively coupled plasma–mass spectrometry for detrital zircons analyzed in this study, is available at www.geosociety.org/pubs/ft2011.htm, or on request from editing@geosociety.org, Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301-9140, USA.