A distinctive yet enigmatic suite of fault-bounded ultramafic massifs occurs within accretionary complex mélange of the McHugh Complex on the Kenai Peninsula of southern Alaska. The largest and most significant of these include Red Mountain and the Halibut Cove Complex, consisting of dunite and pyroxenite with chromite seams and lesser quantities of garnet pyroxenite and gabbro. Several different hypotheses have been advanced to explain their origin. Burns (1985) correlated these fault-bounded ultramafic massifs with others known as the Border Ranges Ultramafic-Mafic Complex. Other parts of the Border Ranges Ultramafic-Mafic Complex are located several hundred kilometers away along the Border Ranges fault, marking the boundary between the Chugach terrane and the Wrangellian composite terrane in the northern and eastern Chugach Mountains. Burns (1985) suggested that this entire group of ultramafic bodies represents the deep roots of the Talkeetna arc developed on the southern margin of Wrangellia during Early Jurassic–Cretaceous subduction. In this model, bodies such as Red Mountain represent klippen thrust hundreds of kilometers southward over the McHugh Complex and now preserved as erosional remnants. Bradley and Kusky (1992) suggested alternatively that the Kenai ultramafic massifs may represent segments of a thick oceanic plate offscraped during subduction, and therefore might represent ophiolitic, oceanic plateau, or immature island arc crust as opposed to the roots of the mature Talkeetna arc. In this scenario, the Kenai ultramafic massifs would be correlative with the McHugh Complex, not the Talkeetna arc. A third hypothesis is that the Border Ranges Ultramafic-Mafic Complex may represent forearc or suprasubduction zone ophiolites formed seaward of the Talkeetna arc during early stages of its evolution and incorporated into the accretionary wedge during subsequent accretion tectonics. The implications of which of these models is correct are large because the Talkeetna arc section is the world's premiere example of a complete exposed arc sequence, including the volcanic carapace through deep crustal levels. Many models for the composition and evolution of the crust rely on the interpretation that this is a coherent and cogenetic section of arc crust.

We report six new U/Pb zircon ages that show that at least some of the deep ultra-mafic and mafic complexes of the Border Ranges Ultramafic-Mafic Complex are Triassic (227.7 ± 0.6 Ma; Norian) and significantly older than structurally overlying Jurassic rocks of the Talkeetna arc (201–181 Ma, continuing plutonism until 163 Ma) but the same age as the surrounding Triassic-Jurassic-Cretaceous McHugh Complex. New geochemical data that show that rocks of the Border Ranges Ultramafic-Mafic Complex have ophiolitic affinities, with Cr-chemistry further indicating that the complex's rocks formed in a suprasubduction zone ophiolite. Regional and detailed and field observations show that rocks of the complex are similar to and can be structurally restored with other fault-bounded units in the McHugh Complex mélange, and that a crude ophiolitic stratigraphy can be reconstructed through the Border Ranges Ultramafic-Mafic Complex and McHugh Complex. We suggest that the Border Ranges Ultramafic-Mafic Complex represents the forearc oceanic basement upon which the Talkeetna arc was subsequently built. The conclusion that the Border Ranges Ultramafic-Mafic Complex does not represent the base of the Talkeetna arc but instead contains remnants of a dismembered ophiolitic complex raises questions about the validity of mass balance calculations and bulk crustal compositions, as well as models of arc development used to understand the growth of continental crust.