Figure 1. Schematic cross section of Chesuncook Dome-Katahdin area showing relationship of mafic and granitic magmas to upper and lower plates of Acadian orogen. Modified from Bradley and Tucker (2002). ABSTRACTDuring collisional convergence, failure in extension of the lithosphere of the lower plate due to slab pull will reduce the thickness or completely remove lower-plate lithosphere and cause decompression melting of the asthenospheric mantle; magmas from this source may subsequently provide enough heat for substantial partial melting of crustal rocks under or beyond the toe of the collisional accretionary system. In central Maine, United States, this type of magmatism is first apparent in the Early Devonian West Branch Volcanics and equivalent mafic volcanics, in the slightly younger voluminous mafic/silicic magmatic event of the Moxie Gabbro-Katahdin batholith and related ignimbrite volcanism, and in other Early Devonian granitic plutons. Similar lower-plate collisional sequences with mafic and related silicic magmatism probably caused by slab breakoff are seen in the Miocene-Holocene Papuan orogen, and the Hercynian-Alleghenian belt. Magmatism of this type is significant because it gives evidence in those examplesof whole-lithosphere extension. We infer that normal fault systems in outer trench slopes of collisional orogens in general, and possibly those of oceanic subduction zones, may not be primarily due to flexural bending, but are also driven by whole-lithosphere extension due to slab pull. The Maine Acadian example suggests that slab failure and this type of magmatism may be promoted by preexisting large margin-parallel faults in the lower plate.

No abstract

Modern and past ridge subduction events are characterized by the intrusion of midocean-ridge basalt (MORB) magmas into an overlying accretionary prism. The fi eld relationships and trace-element geochemistry of Ordovician mafi c igneous rocks of the Weeksboro-Lunksoos and Munsungun anticlinoria of north-central Maine indicate that they resulted from such an event. The Bean Brook gabbro intrudes the Hurricane Mountain mélange and other related sedimentary strata of continental derivation. The gabbro and associated Dry Way volcanics have MORB trace-element chemistries, while the Bluffer Pond and Stacyville volcanics are more enriched (E-MORB), all of which indicate derivation from a mid-ocean ridge. On petrogenetic diagrams, mafi c samples plot in MORB fi elds, or when Th is used as a discriminator, in arc fi elds along trends that originate from MORB fi elds and extend toward the composition of upper continental crust. These trends are consistent with the presence of silicic and metasedimentary xenoliths in the Dry Way volcanics and Bean Brook gabbro and indicate the magmas were not subduction products, but were contaminated by Th-rich upper continental crust. The nearby Chain Lakes Massif likely represents the basement to the "Chain Lakes microcontinent," and the geographic relationship and ages of the massif, Hurricane Mountain mélange, and Dry Way-Bean Brook magmatic rocks indicate northwestdirected subduction (modern coordinates) on the southeastern margin of the Chain Lakes microcontinent, within the Early to Middle Ordovician Taconic ocean. Subduction at this boundary probably terminated because of the ridge subduction, analogous to Neogene California.

No abstract

The Ordovician Stanbridge Group of Quebec has long been considered an allochthonous nappe. It is an internally coherent unit that consists of lower slaty limestone overlain by slate, which is correlated with the Highgate and Morses Line formations, respectively, in Vermont. In Quebec, the basal limestones have been inferred to be thrust over Cambrian dolomites (Gorge Formation in Vermont) of the Rosenberg slice, part of the parauthochthonous shelf, although this contact is not exposed there. In the Missisquoi River gorge of Vermont, a conformable sequence of upper GorgeHighgate-Morses Line formations is exposed. The map distribution of rock units indicates that this conformable relationship probably extends up to at least the middle unit of the Stanbridge Group. Therefore, the relationships in Vermont require that the Stanbridge Group must be part of the parauthochthonous Taconic foreland rather than a far-traveled nappe. The Rosenberg slice in Quebec forms a large anticline (Highgate anticline) whose western limb is truncated by the Rosenberg thrust. In Vermont, the anticline is internally cut by the Highgate Falls Thrust, which is an out-ofsequence thrust that decreases in displacement northwards to the International Border.Résumé : Le Groupe de Stanbridge au Québec (Ordovicien) a longtemps été considéré comme une nappe allochtone. Il s'agit d'une unité interne cohérente qui comprend un calcaire schisteux inférieur recouvert par de l'ardoise. Ces roches sont corrélées respectivement au Vermont avec les formations de Highgate et de Morses Line. Au Québec, les calcaires de base auraient été chevauchés par-dessus les dolomies cambriennes (la Formation de Gorge au Vermont) du copeau tectonique de Rosenberg, lequel fait partie de la plate-forme parautochtone, bien que ce contact n'affleure pas à cet endroit. Une séquence conforme des formations Gorge-Highgate-Morses Line affleure dans la gorge de la rivière Missisquoi au Vermont. Sur les cartes, l'étendue des unités rocheuses indique que cette relation concordante s'étend probablement au moins jusqu'à l'unité centrale du Groupe de Stanbridge. Les relations au Vermont exigent donc que le Groupe de Stanbridge fasse partie de l'avant-pays parautochtone taconique plutôt que d'une nappe provenant de loin. Le copeau tectonique de Rosenberg forme un grand anticlinal au Québec (anticlinal de Highgate) dont le membre ouest est tronqué par le chevauchement de Rosenberg. Au Vermont, le chevauchement de Highgate Falls recoupe l'intérieur de l'anticlinal; ce chevauchement est irrégulier et présente un déplacement qui décroît en direction nord vers la frontière internationale.[Traduit par la Rédaction] Schoonmaker and Kidd 169