Search for: All records

Abstract The interaction between subduction zones and oceanic spreading centers is a common tectonic process, and yet our understanding of how it is manifested in the geologic record is limited to a few well-constrained modern and ancient examples. In the Paleogene, at least one oceanic spreading center interacted with the northwestern margin of North America. Several lines of evidence place this triple junction near Washington (USA) and southern British Columbia (Canada) in the early to middle Eocene, and we summarize a variety of new data sets that permit us to track the plate tectonic setting and geologic evolution of this region from 65 to 40 Ma. The North Cascades segment of the voluminous Coast Mountains continental magmatic arc experienced a magmatic lull between ca. 60 and 50 Ma interpreted to reflect low-angle subduction. During this period of time, the Swauk Basin began to subside inboard of the paleo-trench in Washington, and the Siletzia oceanic plateau began to develop along the Farallon plate–Kula plate or Farallon plate–Resurrection plate spreading center. Farther east, peraluminous magmatism occurred in the Omineca belt and Idaho batholith. Accretion of Siletzia and ridge-trench interaction occurred between ca. 53 and 49 Ma, as indicated by: (1) near-trench magmatism from central Vancouver Island to northwestern Washington, (2) disruption and inversion of the Swauk Basin during a short-lived contractional event, (3) voluminous magmatism in the Kamloops-Challis belt accompanied by major E-W extension east of the North Cascades in metamorphic core complexes and supra-detachment basins and grabens, and (4) southwestward migration of magmatism across northeastern Washington. These events suggest that flat-slab subduction from ca. 60 to 52 Ma was followed by slab rollback and breakoff during accretion of Siletzia. A dramatic magmatic flare-up was associated with rollback and breakoff between ca. 49.4 and 45 Ma and included bimodal volcanism near the eastern edge of Siletzia, intrusion of granodioritic to granitic plutons in the crystalline core of the North Cascades, and extensive dike swarms in the North Cascades. Transtension during and shortly before the flare-up led to >300 km of total offset on dextral strike-slip faults, formation of the Chumstick strike-slip basin, and subhorizontal ductile stretching and rapid exhumation of rocks metamorphosed to 8–10 kbar in the North Cascades crystalline core. By ca. 45 Ma, the Farallon–Kula (or Resurrection)–North American triple junction was likely located in Oregon (USA), subduction of the Kula or Resurrection plate was established outboard of Siletzia, and strike-slip faulting was localized on the north-striking Straight Creek–Fraser River fault. Motion of this structure terminated by 35 Ma. These events culminated in the establishment of the modern Cascadia convergent margin. 

The Eocene Golden Horn batholith (North Cascades, northwestern United States) is an unusual example of coeval ferroan (A-type) and I-type granitoids that formed in a subduction zone setting. Associated mafic to intermediate dikes and stocks (adakites and calc-alkaline basalts to andesites) attest to the presence of mafic magmas in the shallow crust during batholith assembly and imply anomalously hot conditions in the mantle that are attributed to rupture or rollback of the subducting slab. Field, geochemical, and Sr-Nd-Pb isotopic data support generation of the ferroan magmas at midcrustal depth by melting of calc-alkaline plutonic rocks or by protracted differentiation of a parental basalt, and production of the I-type magmas by partial melting of gabbroic rocks at greater depth. Metasomatism by fluids bearing Na, F, and high field strength elements enriched some ferroan rocks to produce peralkaline granite. Mixing between the I-type and ferroan magmas also produced a hornblende-biotite granite with rapakivi feldspars, which is the most voluminous phase of the batholith. Results of this study suggest that ferroan granitoids rarely form in continental arcs because that tectonic regime limits voluminous basalt intrusion into the upper crust, and not because of a lack of suitable source rocks. Their presence in an arc setting likely indicates rollback or tearing of the slab or the presence of a slab window.