Abstract Low-temperature thermochronological data for the Eurasian foreland north of the Bitlis–Zagros suture zone suggest that the tectonic stresses related to the Arabian collision during mid-Miocene time were transmitted efficiently over large distances, focusing preferentially at rheological discontinuities. Since the late Middle Miocene a new tectonic regime has been active as the westwards translation of Anatolia is accommodating most of the Arabia–Eurasia convergence, thus precluding the efficient transfer of stress northwards. Apatite fission-track data from the central Lesser Caucasus show that a portion of this orogen underwent a discrete phase of cooling/exhumation at 18–12 Ma (late Early–early Middle Miocene) as a result of the structural reactivation of a segment of the Late Cretaceous–Palaeogene Sevan–Akera suture zone. This inference contradicts the notion that the post-collisional history of the study area was dominated by strike-slip tectonics with relatively minor dip-slip components. Reactivation and exhumation was focused along those segments of the suture zone at high angles to the inferred collisional stress field; the remaining areas were not exhumed enough to expose a new apatite partial annealing zone and thus retained the thermochronological record of a phase of Late Cretaceous cooling/exhumation associated with ophiolite obduction and the following continental collision along the suture zone.

In this paper, we merge more than 200 new apatite and zircon (U-Th)/He analyses and 21 apatite fission-track analyses from 71 new samples with previous published thermochronologic data using the same systems to understand the growth and large-scale kinematics of the central Andes between 21°S and 28°S. In general, minimum dates decrease and the total range of dates increases from west to east across the range. Large variations in thermochronometer dates on the east side reflect high spatial gradients in depth of recent erosional exhumation. Almost nowhere in this part of the Andes has Cenozoic erosion exceeded ~6–8 km, and in many places in the eastern half of the range, erosion has not exceeded 2–3 km, despite these regions now being 5–6 km above sea level. This means that west of the rapidly deforming and eroding eastern range front, uplift and erosion are largely decoupled as a result of meager precipitation, relatively low relief, internal drainage, and volcanic burial. We interpret the west-to-east pattern of decreasing minimum dates across the range as recording the time-transgressive eastward migration of a focused zone of deformation, erosion, and convergence between the South American plate and the eastern edge of the Andean orogenic plateau. At this scale, the thermochronologic data do not suggest major changes in rates of plateau propagation or shortening/convergence with time. We use the thermochronometer date-distance trend and a simple kinematic model to infer a rate of eastward propagation of deformation and plateau growth of 6–10 km/m.y. This plateau propagation model balances horizontal convergence, erosion, and crustal thickening and predicts rates of shortening and convergence between the Andes block and South American plate that are consistent with geologic and geodetic observations.