We investigate whether predictions of mantle structure from tectonic reconstructions are in agreement with a detailed tomographic image of seismic P wave velocity structure under the Caribbean region. In the upper mantle, positive seismic anomalies are imaged under the Lesser Antilles and Puerto Rico. These anomalies are interpreted as remnants of Atlantic lithosphere subduction and confirm tectonic reconstructions that suggest at least 1100 km of convergence at the Lesser Antilles island arc during the past ~45 Myr. The imaged Lesser Antilles slab consists of a northern and southern anomaly, separated by a low‐velocity anomaly across most of the upper mantle, which we interpret as the subducted North America‐South America plate boundary. The southern edge of the imaged Lesser Antilles slab agrees with vertical tearing of South America lithosphere. The northern Lesser Antilles slab is continuous with the Puerto Rico slab along the northeastern plate boundary. This results in an amphitheater‐shaped slab, and it is interpreted as westward subducting North America lithosphere that remained attached to the surface along the northeastern boundary of the Caribbean plate. At the Muertos Trough, however, material is imaged until a depth of only 100 km, suggesting a small amount of subduction. The location and length of the imaged South Caribbean slab agrees with proposed subduction of Caribbean lithosphere under the northern South America plate. An anomaly related to proposed Oligocene subduction at the Nicaragua rise is absent in the tomographic model. Beneath Panama, a subduction window exists across the upper mantle, which is related to the cessation of subduction of the Nazca plate under Panama since 9.5 Ma and possibly the preceding subduction of the extinct Cocos‐Nazca spreading center. In the lower mantle, two large anomaly patterns are imaged. The westernmost anomaly agrees with the subduction of Farallon lithosphere. The second lower mantle anomaly is found east of the Farallon anomaly and is interpreted as a remnant of the late Mesozoic subduction of North and South America oceanic lithosphere at the Greater Antilles, Aves ridge, and Leeward Antilles. The imaged mantle structure does not allow us to discriminate between an “Intra‐Americas origin” and a “Pacific origin” of the Caribbean plate.

Recent installation of 3‐component broadband digital stations around the Gulf of California allowed us to make shear wave splitting observation from records of 73 SKS and SKKS phases from 30 events recorded at 12 stations since January of 2001. Stations in the southern Basin and Range province yield upper mantle fast shear wave direction from NE‐SW to E‐W, consistent with both local direction of Miocene extension and North American absolute plate motion. The shear at the Pacific‐North American plate limit seems to control tightly the anisotropy pattern of the closest station NE70. The anisotropic pattern along the Peninsular Range, east of the former trench between the North American plate and the now subducted Farallon plate, is almost uniform and is interpreted as asthenospheric flow induced by the sinking fragment of the Farallon plate although frozen anisotropy associated with in situ captured fragment of this plate cannot be ruled out.

[1] Mechanical coupling between the lithosphere and the asthenosphere remains a controversial topic in the geosciences. Beneath the Caribbean plate, shear wave splitting measurements indicate EW strain in the asthenosphere, which can be interpreted as mantle flow driving or resisting motion of the overlying lithosphere. Here, we constrain the average shear traction on the base of the Caribbean plate by balancing all torques. These torques result from body forces that act on the Caribbean (slab pull, ridge push, lateral density variations), from plate boundary friction and from basal shear tractions. We obtain a range of physically realistic torque solutions, which we examine further by computing the corresponding stresses and rotations within the Caribbean plate for comparison with observations. The deformation field for the Caribbean is particularly sensitive to the amount of friction on intraplate faults. Representative models have a good fit with observations and are characterized by (1) a near-zero basal shear traction (≤0.3 MPa), (2) (lithosphereaveraged) plate boundary friction ≤ 10 MPa, (3) local forces due to indenters and trench pull, (4) a net pull by the Caribbean slab and (5) intraplate fault shear stresses on the order of tens of megapascals. We conclude that the mechanical coupling of the Caribbean plate to the underlying asthenosphere is small.

The north Caribbean plate boundary zone is a broad deformation zone with several fault systems and tectonic blocks that move with different velocities. The indentation by the Bahamas Platform (the "Bahamas Collision") is generally invoked as a cause of this fragmentation. We propose that a second driver of deformation is the western edge of the south dipping Puerto Rico slab moving sideways with the North America plate. The westward motion of the slab edge results in a push on the Caribbean plate farther west. We refer to this second mechanism for deformation as "Slab Edge Push." The motion of the North America plate relative to the Caribbean plate causes both drivers to migrate from east to west. The Bahamas Collision and Slab Edge Push have been operating simultaneously since the Miocene. The question is the relative importance of the two mechanisms. We use mechanical finite element models that represent the two mechanisms from the late Oligocene (30 Ma) to the present. For the present, both models successfully reproduce observed deformation, implying that both models are viable. Back in time the Slab Edge Push mechanism better reproduces observations. Neither mechanism successfully reproduces the observed Miocene counterclockwise rotation of Puerto Rico. We use this rotation to tune a final model that includes fractional contributions of both mechanisms. We find that the Slab Edge Push was the dominant driver of deformation in the north Caribbean plate boundary zone since 30 Ma.