5. Discussion

5.1 East Side of the Totschunda Fault Thermal Histories

The two samples of Cretaceous gabbro (19SOLO01 and 02) from the east side of the Totschunda fault document mid-Cretaceous (ca. ~95 Ma) cooling of ~5°C/Ma, then ca. 70 million years of relatively slow cooling (<1°C/Ma) followed by an increase in cooling rate to ~2.5°C/Ma in the Late Miocene. Given that the U-Pb zircon emplacement age is ca. 100 Ma (19SOLO01), and AFT ages are 84-87 Ma, we infer that Cretaceous cooling is due to regional exhumation, consistent with similar Cretaceous cooling ages observed west of the central Totschunda fault by Milde (2014). We acknowledge that track shortening from long residence times in the PAZ would have reduced the fission track age, but given the coarse-grained nature of the intrusive rock implying emplacement below the PAZ (~ 5 km) we think Cretaceous exhumation is more apt. These gabbroic rock samples then slowly cooled or were near isothermal from the mid-Late Cretaceous until the Late Miocene where a small pulse of more rapid cooling is evident (~35°C since 6 Ma). This Late Miocene cooling is significantly less than the observed Late Miocene-Present cooling on the west side of the Totschunda fault (~100°C since 6 Ma).

5.2 West side of the Totschunda fault Thermal Histories

U-Pb zircon ages from hypabyssal rocks on the west side of Totschunda fault (samples 19SOLO06; 07; 10 and 11) yield Oligocene ages. Thus, these rocks only constrain post-Oligocene tectonic activity along the Totschunda fault, but we infer their emplacement is related to concurrent transtensional magmatism along the fault (e.g., Blanquat et al., 1998). Given the partial overlap of the AFT ages with U-Pb zircon ages for the hypabyssal samples west of the Totschunda fault (25-28 Ma) and the complex track length distributions, we interpret these rocks to have been emplaced at relatively shallow depths in concurrence with their lithology, followed by residence in the partial annealing zone before later cooling and inferred exhumation. The overlap in fission track and U-Pb zircon ages therefore reflects the timing of crystallization. However, the paucity of longer tracks relative to shorter more-annealed tracks in the Oligocene hypabyssal rocks suggests these rocks were exhumed in the recent geologic past (Figure 5). The HeFTy thermal models further constrain this history with residence in a partial annealing zone from the time of emplacement until ~6 Ma when they were rapidly cooled. The AHe ages (3.23 - 1.07 Ma) also indicate continued rapid cooling through the Pliocene and into the Pleistocene. We interpret this Late Miocene cooling as related to rejuvenated horizontal slip and associated exhumation along the Totschunda Fault.

5.3 The case against thermal perturbation for our samples

The case for thermal perturbation resulting from nearby magmatism (or resultant hydrothermal effects) in our samples is generally poor except for the ZHe results from sample 19SOLO02 as discussed above. While young Wrangell arc lavas nearby have similar Late Miocene - recent ages of eruption compared to AHe ages from the hypabyssal samples (Denton and Armstrong, 1969; Richter et al., 1990; Trop et al., 2022), the fact that the gabbro sample from east of the Totschunda Fault preserves an AHe age that is much older than the nearby lavas (<1 km distance, 27 Ma AHe age vs 2 Ma age of lava) suggests that these lavas did not reset the AHe system in our hypabyssal intrusive rocks. The young AHe ages in our hypabyssal rocks therefore reflect exhumation related cooling. The hypabyssal rocks themselves hypothetically could have reset the AHe ages for the Cretaceous aged gabbro, as the AHe gabbro age (ca. 27 Ma) is the similar as the intrusive ages of the hypabyssal rocks that are only 1.5 km away (25-28 Ma). However, the overall large spread in single grain AHe ages (53.4-9.0 Ma) from the gabbro imply the sample did not experience a significant heating-cooling event at ca. 27 Ma.

5.4 Climatic Forcing of Rock Cooling (?)

The Pleistocene AHe ages from the southern Totschunda fault overlap with the timing of Northern Hemisphere glaciations (e.g., Raymo, 1994). Some researchers have suggested that glaciation can lead to periods of rapid erosion, and thus, rapid exhumation (e.g., Spotila et al., 2004; Huntington et al., 2006; Blythe et al., 2017). If some of the sediments in the White River ‘Tillites’ have a glacial origin, increased climate forcing (i.e., glaciation) could, in part, be responsible for the recorded exhumation. If true, one should expect Pleistocene AHe ages everywhere there was glaciation in the Lime Creek drainage with the youngest cooling ages at the lowest elevations regardless of structural position (i.e., proximity to the fault). That is not the case as shown by the Oligocene AHe ages from Cretaceous gabbro samples collected ~35 m above the valley floor that have ages of ca. 27 Ma as compared to two samples collected ~400 m above the valley floor, that have younger Plio-Pleistocene AHe cooling ages (Figure 6). Additionally, the Totschunda fault is in the rain shadow of the Saint Elias Range and considered an area where surface processes’ control on cooling age patterns are limited during the late Cenozoic (Enkelmann et al., 2017). Overall, the pattern of thermochronologic ages show a structural control-contrasting ages across the Totschunda fault (ages are younger to the west, older to the east) indicative of west-side up displacement. This pattern of younger thermochronologic ages to the west of the fault was also recognized by Milde (2014) on the central Totschunda fault.

5.5 Detrital Cobbles and the White River ‘Tillites’

The different zircon U-Pb and AFT age populations between the two groups of porphyritic andesite and dacite cobbles collected from different stratigraphic positions may reflect a change in provenance during deposition of basin sediments; however the stratigraphic ages of the cobbles are poorly constrained given the deformational history of the basin (Denton and Armstrong, 1969; Eyles and Eyles, 1986). The K-Ar ages on lavas within the stratigraphic section reported by Denton and Armstrong (1969) constrain the timing of sedimentation but given post-depositional deformation are insufficient to resolve the stratigraphic position of the cobbles. Regardless, the cobbles were deposited at the surface sometime between ca. 10 and 2 Ma.
At one outcrop near the base of the section, sample 19SOLO18B has a mean track length of 10.8 mm (n=12) and sample 19SOLO18C from the same location has a mean track length of 14.6 mm (n=15). Overlapping U-Pb and AFT ages for this stratigraphically lowermost outcrop indicate that both clasts originate from coeval subvolcanic or volcanic rocks and both clasts shared the same depositional history. Cobble 19SOLO18C, with little track shortening, had to come from a rock emplaced above the PAZ or erupted on the surface (cooler than 60°C) and never buried to higher temperatures. Significant track shortening in Cobble 19SOLO18B suggests residence in the PAZ, which could potentially have occurred either before source rock exhumation (as in the bedrock samples analyzed) or in the basin during burial. This latter option is unlikely as both cobbles 19SOLO18B and 19SOLO18C must share the same depositional history. Therefore, our cobble AFT data do not record the burial nor exhumation history of this “thin” sedimentary basin. Instead, the cobbles document that there has been little to no thermal resetting in the tillite strata. We therefore cannot constrain the timing of basin deformation and unroofing using the AFT data. Unlike thicker foreland basins where the AFT cobble thermochronology method works well (Beamud et al., 2011; Fitzgerald et al., 2019), we suggest that application of this method to constrain the timing of basin inversion in smaller relatively short-lived basins formed along strike-slip faults is unlikely to work in many cases due to insufficient burial (~3-5 kms) to fully or partially reset the AFT system.

5.6 Tectonic Significance

5.6.1 Mid-Cretaceous to Late Miocene

The two mid-Cretaceous gabbroic samples from east of the Totschunda fault preserve Late Cretaceous cooling which may be related to Mesozoic slip along the Totschunda fault. The thermal history after ca. 85 Ma for our gabbroic rocks on the east side of the Totschunda fault shows prolonged slow cooling until the Late Miocene.
While we observe no pulses of Oligocene rapid cooling in any of our samples, available geologic evidence suggests the Totschunda fault has been continuously active since the Oligocene. The available geologic evidence is as follows: 1) Oligocene-Miocene thrust-top basins along the Totschunda fault after palinspastic reconstruction (Allen et al., 2018); 2) Rapid Oligocene cooling in Cretaceous aged intrusive rocks along the central segments of the Totschunda fault (Milde, 2014); 3) Diking into the central Totschunda fault zone from ca. 29 Ma to 23 Ma linked with fault zone fluid flow and transtension (Brueseke et al., 2019); 4) Magmatism in the Sonya Creek Volcanic field ca. 23 to 19 Ma associated with transtension along the Totschunda fault (Berkelhammer et al., 2019); and 5) Oligocene-early Miocene Wrangell Arc magmatism along the Totschunda fault (Trop et al., 2022, this study).
Therefore, we conclude that the Totschunda fault experienced strike-slip motion ca. 25 to ca. 6 Ma, though at much reduced rates (~2 mm/yr) compared to ca. 6 Ma to modern rates (~14 mm/yr; Marechal et al., 2018; Allen et al., 2022). This slip-rate calculation of ~2 mm/yr is based on 1) ~125 km allowable Totschunda fault slip since 52 Ma (Clearwater-Talkeetna fault to Nutzotin-Totschunda fault offset), and 2) the estimated ~85 km of Totschunda fault slip since 6 Ma (Berkelhammer et al., 2019; Waldien et al., 2021; Allen et al., 2022) (Figures 1, 7). This leaves 40 km of required slip over ~20 million years equating to 2 mm/yr.

5.6.2 Late Miocene - Present: Eastern Denali-Totschunda Fault Reorganization

In the past 20 million years, the most significant plate boundary events along the Alaska southern margin are the following:
The ca. 6 Ma Pacific-Yakutat plate vector change and associated 37% net increase in convergence rate between the Pacific and North American plates (e.g., Engebretsen, 1985; Doubrovine and Tarduno, 2008; Austerman et al., 2011) and
Following 29 million years of the Yakutat flat slab subduction, final collision of the eastern Yakutat ~30 km thick oceanic plateau segment at ca. 1 Ma (e.g., Richter et al., 1990; Gulick et al., 2013; Brueseke et al., 2023).
These tectonic events have reshaped the Alaska plate boundary and the faults which accommodate stresses transferred inboard from this convergent margin.
Numerous authors have referred to the ‘collision’ of the Yakutat oceanic plateau starting as early as the Oligocene and Mio-Pliocene (Chapman et al., 2008; Enkelmann et al., 2010; Lease et al., 2016) and as late as 1 Ma (e.g., Reece et al., 2013; Brueseke et al., 2023). Here, we define collision as the cessation of Wrangell arc magmatism and the abandonment of the subduction zone trench (e.g., McGeary et al., 1985). Evidence against a 6 Ma Yakutat ‘collision’ event are the following: 1) Based on migrating patterns of arc magmatism, the subducting Yakutat slab apparently accelerated in subduction rate (or flattened, but continued subducting) between 6 Ma and 1 Ma (Richter et al., 1990; Trop et al., 2022); 2) The Wrangell arc was still robustly active between 6 Ma and 1 Ma (Richter et al., 1990; Trop et al., 2022); and 3) The now colliding 30 km thick segment of the Yakutat terrane (Worthington et al., 2012) would have been roughly 300 km outboard of the North American margin ~ 6 million years ago based on modern Pacific-Yakutat plate/North America ~50 km/Ma convergence rates (Elliot et al., 2010).