Geologic and Tectonic Evolution of Annette, Gravina, Duke, and Southern Prince of Wales Islands, Southeastern Alaska

Citation

Gehrels, George Ellery (1986) Geologic and Tectonic Evolution of Annette, Gravina, Duke, and Southern Prince of Wales Islands, Southeastern Alaska. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/pavf-pm10. https://resolver.caltech.edu/CaltechTHESIS:03142019-142709345

Abstract

Annette, Gravina, Duke, and southern Prince of Wales Islands are underlain primarily by Cambrian (and perhaps Proterozoic) through Triassic volcanic, sedimentary, plutonic, and metamorphic rocks. These rocks belong to the Alexander terrane, which is a coherent tectonic fragment that underlies much of southeastern (SE) Alaska, the Saint Elias Mountains of British Columbia, Yukon, and eastern Alaska, and coastal regions of west-central British Columbia. Geologic mapping combined with U-Pb (zircon) geochronologic studies have delineated the major geologic units and features of these islands, and contribute to our understanding of the geologic and tectonic evolution of the Alexander terrane.

The oldest rocks recognized on Annette, Gravina, Duke, and southern Prince of Wales Islands consist of greenschist- and amphibolite-facies metavolcanic and metasedimentary rocks of the Wales metamorphic suite. These rocks are locally intruded by dioritic and granodioritic metaplutonic rocks which yield U-Pb apparent ages of approximately 540-520 Ma (Middle and Late Cambrian). Rocks in the Wales suite were therefore deposited, at least in part, prior to Late Cambrian time, but their maximum depositional age is not known. The Wales suite and associated metaplutonic rocks are intruded by large dioritic to granitic plutons which yield U-Pb apparent ages in the 475-425 Ma (Middle Ordovician-Early Silurian) range and are probably overlain by Lower Ordovician-Lower Silurian volcanic and sedimentary rocks of the Descon Formation. However, depositional contacts between the Descon Formation and the older metamorphic rocks have not been demonstrated. Deformation, metamorphism, and uplift of rocks in the Wales metamorphic suite occurred during a Middle Cambrian-Early Ordovician tectonic event which I have referred to as the Wales "orogeny." The term orogeny is used informally in this instance, as little is known about the regional and tectonic significance of this event.

Ordovician-Early Silurian rocks on these islands are interpreted to have formed in an oceanic volcanic arc environment based on similarities with young or presently active volcanic arcs in the Circum-Pacific region. Characteristics of the volcanic-plutonic complex in the Alexander terrane which are similar to those in other magmatic belts include: 1) predominance of basaltic to andesitic volcanic rocks and dioritic to granodioritic and subordinate granitic plutonic rocks, 2) calc-alkaline affinity of the plutonic and volcanic rocks, as defined on AFM, FeO*/MgO versus SiO₂, and La versus Nb diagrams and by an alkali-lime index of 56-62, 3) patterns of strong (50-100 times chondrites) light REE enrichments, moderate (5 to 20 times chondrites) heavy REE enrichments, and strong negative europium anomalies, 4) evolution of the magmatic system over a period of approximately 50 m.y., and 5) increasing potassium content with time in the plutonic rocks.

Facies relations in Ordovician-Silurian strata in the southern part of the terrane generally record northwesterly paleogeographic trends, indicating that the interpreted arc trended oblique to the northnorth-westerly elongation of the terrane. Continuation of Ordovician-Silurian shallow-marine strata for over 600 km to the north-northwest indicates that the interpreted arc probably faced to the southwest and that the strata to the north accumulated in a back-arc environment. Protoliths of the Wales metamorphic suite may also have formed in a volcanic arc environment, but the penetrative deformation and regional metamorphism of these rocks precludes detailed analyses of protolith relations and composition.

During middle Silurian-earliest Devonian time the Early Silurian and older rocks in the area were involved in a major tectonic event which I refer to as the Klakas orogeny. Manifestations of this orogeny on Annette, Gravina, Duke, and Prince of Wales Islands include: 1) cessation of the Ordovician-Early Silurian volcanism and plutonism, 2) deposition of middle and Upper Silurian polymictic conglomerate on northern Prince of Wales Island and regions to the north, 3) erosion or non-deposition of Silurian strata on Annette, Gravina, and Duke Islands and on central and southern Prince of Wales Island, 4) southwest-directed movement on thrust faults on southern Prince of Wales Island and perhaps on Annette Island, 5) deposition and deformation of a Lower Devonian talus breccia and penetrative brecciation of Ordovician rocks along thrust faults on southern Prince of Wales Island, 6) greenschist-and local amphibolite-facies regional metamorphism and penetrative deformation of Ordovician-Early Silurian rocks on Annette, Gravina, and Duke Islands, 7) emplacement, and perhaps generation by anatexis, of Late Silurian trondhjemite, sodic leucodiorite, and subordinate granite plutons, 8) several kilometers (perhaps as much as 10 km) of uplift of Late Silurian and older rocks prior to middle Early Devonian time, and 9) deposition of Lower Devonian conglomeratic red beds of the Karheen Formation in topographically rugged subaerial environments in some regions to the south, and in a northward tapering clastic wedge to the north. Previous workers recognized the stratigraphic manifestations of this orogenic event on central and northern Prince of Wales Island, but most relations to the south were recognized initially during this study.

On southern Prince of Wales Island, Lower Devonian conglomeratic strata are overlain by shallow-marine limestone, mudstone, and siltstone of middle Early Devonian age, which are in turn overlain by deeper-water mudstone and graptolitic shale. Subsidence of the region below sea level following the Klakas orogeny therefore occurred during middle Early Devonian time and produced a marine transgression on a north-facing paleoslope. Lower Devonian strata on Annette, Gravina, and Duke Islands were deposited in shallow-marine environments, and only locally include polymictic conglomerate and coarser clastic strata. Andesitic volcanic rocks of probable Early Devonian age locally overlie the marine clastic strata and are the youngest Paleozoic rocks in the study area.

Triassic strata herein referred to as the Hyd Group unconformably overlie the Devonian and older rocks on Annette and Gravina Islands. At the base of the section in most areas is a thick conglomerate or sedimentary breccia with meter-size clasts of Devonian and older rock in a poorly sorted matrix. These strata are overlain by a sequence, from bottom to top, of rhyolite and rhyolitic tuff, shallow-marine limestone, calcareous siltstone and limestone, and basalt flows and breccia. A UPb apparent age of 225 ± 3 on the rhyolite combined with megafossil and conodont ages demonstrate that these strata were deposited during Late Carnian to Late Norian time, and place a minimum age constraint of 225 ± 3 Ma on the Carnian-Norian boundary. A large body of pyroxene gabbro on Duke Island yields a U-Pb apparent age of 226 ± 3 Ma, which demonstrates that this gabbro is not genetically related to the zoned ("Alaskan-type") ultramafic bodies on Duke Island (assuming that the ultramafic bodies are indeed Cretaceous in age!). Rather, the pyroxene gabbro is interpreted to be genetically related to the Triassic basaltic rocks. Intrusive relations indicate that hornblende gabbro on northeastern Duke Island is pre-Late Silurian in age, and is therefore not genetically related to the pyroxene gabbro or to the Cretaceous(?) ultramafic rocks.

The unconformity at the base of the Triassic section records a major latest Paleozoic(?)-Triassic uplift and erosional event in the Alexander terrane, but this event was not associated with regional deformation or metamorphism. This lack of deformation combined with the occurrence of Triassic strata along the eastern margin of the terrane in SE Alaska and the bimodal (basalt-rhyolite) composition of the volcanic rocks suggest that the Triassic strata and their subjacent unconformity formed in an extensional environment. A major low-angle normal fault on southern Prince of Wales Island (the Keete Inlet fault) may also have moved during this interpreted extensional event.

Jurassic and younger rocks intrude and overlie rocks in various terranes in western British Columbia and southern Alaska and demonstrate that the Alexander terrane has been adjacent to Wrangellia since Middle(?) Jurassic time, and to terranes to the east since Late Cretaceous-early Tertiary time. Regional sub-greenschist- to greenschist-facies metamorphism and moderate deformation of Cretaceous and older rocks along the eastern margin of the terrane are interpreted to have occurred during mid-Cretaceous-early Tertiary juxtaposition of the Alexander terrane against terranes to the east.

North of Annette, Gravina, Duke, and southern Prince of Wales Islands the Alexander terrane is underlain primarily by Paleozoic marine clastic strata and limestone. Lower Paleozoic strata in some regions of the Saint Elias Mountains include Cambrian volcanic rocks which may be correlative with rocks in the Wales metamorphic suite, and a thick section of Ordovician-Devonian clastic strata and limestone. Upper Paleozoic clastic strata are widespread in the Saint Elias Mountains region but occur in only a few areas of southeastern Alaska, where they were generally deposited in tectonically stable, shallow-marine environments. Triassic strata to the north are generally similar to rocks on Annette and Gravina Islands, and are interpreted by other workers to have been deposited in a rift environment.

A variety of geologic, paleomagnetic, and paleobiogeographic evidence suggests that the Alexander terrane occupied low paleolatitudes during much of Paleozoic and Mesozoic time, and did not reach its present latitude in the Cordillera until after Early Cretaceous time. Previous hypotheses were that the Alexander terrane was originally adjacent to rocks in the Sierra-Klamath region of California, and that both assemblages formed and evolved adjacent to the California continental margin. Comparison of the geologic and tectonic evolution of the Alexander terrane with that in the Sierra-Klamath region indicates, however, that the two assemblages have little in common and probably were not closely associated during Paleozoic time.

Alternatively, I suggest that the early Paleozoic geologic and tectonic evolution of the Alexander terrane is remarkably similar to that in a dismembered orogenic belt which occurs in southeastern Australia (Lachlan Fold Belt), New Zealand, the Transantarctic Mountains and Byrd Land of Antarctica, and perhaps in tectonic fragments in Asia. Similarities between the Alexander terrane and the Lachlan Fold Belt include: 1) arc-type(?) volcanism and sedimentation during Cambrian time (and perhaps Proterozoic time in the Alexander terrane), 2) regional deformation and metamorphism of the Cambrian and older(?) rocks during Middle Cambrian-Early Ordovician time, 3) evolution of some regions in a volcanic arc environment during Ordovician time (into Early Silurian time in the Alexander terrane), 4) cessation of this volcanic arc activity during the onset of a Silurian-earliest Devonian orogenic event, which is manifest by regional uplift and erosion, deformation and regional metamorphism, anatectic(?) plutonism (and volcanism in the Lachlan Belt), 5) deposition of Lower Devonian and locally Silurian conglomeratic red beds, and 6) evolution in relatively stable marine environments from middle Early through Middle Devonian time.

Comparison of paleolatitudes of the Alexander terrane (determined from paleomagnetic data) with paleolatitudes of eastern Australia (interpreted from continental reconstructions) indicates that the two regions occupied similar paleolatitudes from Ordovician to Late Devonian time. A similar comparison of declination data from the Alexander terrane indicates that both regions also rotated in a clockwise sense during this period. There are also similarities in lower Paleozoic fossils of the two regions, but some faunas from the Alexander terrane apparently bear stronger affinities with North American or Asian fossils.

Based on the geologic, paleomagnetic, and, to some degree the paleobiogeographic similarities, I raise the possibility that the Alexander terrane formed and evolved along the paleo-Pacific margin of Gondwana, perhaps adjacent to rocks in eastern Australia, during early Paleozoic time. The data are not sufficient to draw correlations between the Alexander terrane and specific regions in this complex orogen, although I note that similarities are strongest with the Molong volcanic province in the Lachlan Belt of eastern Australia. The paleomagnetic data indicate that the terrane could have been associated with these rocks or with potential northern correlatives in tectonicfragments that now reside in Asia.

The geologic, paleomagnetic, and paleobiogeographic(?) similarities between the Alexander terrane and the Lachlan Belt end in Middle Devonian-Early Carboniferous time. During this time the Lachlan Belt apparently underwent a major rifting episode, and the Alexander terrane began to evolve in tectonically stable marine environments. The paleolatitudes of the two regions also diverge at this time, with the Alexander terrane migrating northward toward the paleo-equator and eastern Australia continuing its southward movement. Carboniferous fauna from the Alexander terrane are reported by some workers to have "Tethyan" affinities, a fact that is consistent with the low paleolatitudes determined from the paleomagnetic data. Triassic faunas from the terrane are endemic to equatorial or perhaps more southerly regions in the eastern part of the paleo-Pacific basin, and paleomagnetic data from the terrane are most consistent with a paleolatitude of approximately 43° South. In concert with the hypothesis that the terrane was adjacent to the paleo-Pacific margin of Gondwana during early Paleozoic time, I raise the possibility that the terrane was tectonically removed from the Gondwana margin, perhaps by rifting, during Middle Devonian-Early Carboniferous time, and migrated eastward across the paleo-Pacific basin during late Paleozoic time. Northward displacement apparently began after Late Triassic time, and ended during the mid-Cretaceous to early Tertiary juxtaposition of the terrane against fragments previously accreted to western North America.

Thesis Files