The Central Atlantic magmatic province (CAMP) is one of the largest igneous provinces on Earth (> 107 km2), spanning four continents. Recent high-precision 40Ar/39Ar dating of mineral separates has provided important constraints on the age, duration, and geodynamic history of CAMP. Yet the North American CAMP is strikingly under-represented in this dating effort. Here we present 13 new statistically robust plateau, mini-plateau and isochron ages obtained on plagioclase and sericite separates from lava flows from the Fundy (n = 10; Nova Scotia, Canada), Hartford and Deerfield (n = 3; U.S.A.) basins. Ages mostly range from 198.6 ± 1.1 to 201.0 ± 1.4 Ma (2σ), with 1 date substantially younger at 190.6 ± 1.0 Ma. Careful statistical regression shows that ages from the upper (199.7.0 ± 1.5 Ma) and bottom (200.1 ± 0.9 Ma) units of the lava pile in the Fundy basin are statistically indistinguishable, confirming a short duration of emplacement (≪ 1.6 Ma; ≤ 1 Ma). Three ages obtained on the Hartford (198.6 ± 2.0 Ma and 199.8 ± 1.1 Ma) and Deerfield (199.3 ± 1.2 Ma) basins were measured on sericite from the upper lava flow units. We interpret these dates as reflecting syn-emplacement hydrothermal activity within these units. Collectively, CAMP ages gathered so far suggest a short duration of the main magmatic activity (2-3 Ma), but also suggest the possibility of a temporal migration of the active magmatic centers from north to south. Such a migration challenges a plume model that would postulate a radial outward migration of the magmatism and is more compatible with other models, such as the supercontinent global warming hypothesis. When compared to the age of the Triassic-Jurassic boundary, the filtered CAMP age database suggests that the onset of the magmatic activity precedes the limit by at least few hundred thousand years, thereby suggesting a causal relationship between CAMP and the end of Triassic mass extinction. An age at 191 Ma possibly suggests a minor CAMP late tailing activity (190-194 Ma) which has been observed already for dykes and sills in Africa and Brazil. We speculate that, if genuine, this late activity can be due to a major extensional event, possibly heralding the oceanization process at ~ 190 Ma. Comparison between high quality U/Pb and 40Ar/39Ar ages of pegmatite lenses from the North Mountain basalts confirms a ~ 1% bias between the two chronometers. This discrepancy is likely attributed to the miscalibration of the 40K decay constants, in particular the electron capture branch. © 2009 Elsevier B.V. All rights reserved.
