Amphibole enrichment weakens the mid-to-lower crust of southern Tibet

A mechanically weak mid-to-lower crust is widely invoked to explain plateau uplift, lateral lower-crustal flow, and lithospheric removal beneath southern Tibet, yet the processes responsible for such long-term weakening remain debated. Classical models relying on thermal softening of plagioclase or widespread partial melting are challenged by the low water solubility of plagioclase and by geotherms that are too cool to sustain pervasive melting. Here, we propose that arc-magmatic amphibole enrichment is primary control on deep-crustal rheology, seismic anisotropy, and tectonic evolution in southern Tibet. Here we show that modest hydration (1.0-1.5 wt.% H2O) stabilizes amphibole as a major phase (50-65 vol.%) at 30-55 km depths, transforming initially strong plagioclase-pyroxene-dominated lithologies into weak, anisotropic amphibolite. Phase‑equilibrium modeling combined with experimentally calibrated mineral rheology demonstrates that such amphibole‑rich assemblages are significantly less dense and up to two orders of magnitude weaker than their anhydrous equivalents. The resulting viscosity structure and seismic anisotropy align with geodetic constraints on ductile crustal flow and with the observed distribution of lower‑crustal earthquakes. At pressures greater than ~1.6–1.8 GPa (~55–60 km depth), amphibole breaks down to garnet + clinopyroxene, increasing density and promoting eclogitization that facilitates lithospheric delamination. Fluids released during amphibole breakdown may transiently weaken the lowermost crust and contribute to lower-crustal seismicity, but do not account for the long-lived regional weakness. Our results provide a unified, process-based frame work linking pre‑collisional arc magmatism to present‑day rheology, seismic structure, and the tectonic evolution of southern Tibet.