Testing the link between Panthalassa tectonic evolution and subduction-modified mantle heterogeneity

Subduction zones provide a structured pathway for the transfer of Earth-surface materials into the mantle, whereby slab dehydration and melting release water-rich components primarily into the mantle wedge, regulating their initial entry into the convecting upper mantle.

Recent studies (e.g., Yang et al., 2021) suggest that the upper mantle can be broadly divided into subduction-modified and subduction-unmodified domains at a global scale. The former is widely distributed in the Indian Ocean and in parts of the Atlantic, reflecting the asthenospheric metasomatism and recycling associated with subducted materials. In contrast, the subduction-unmodified domain is largely restricted to the Pacific basin and shows little evidence for the involvement of subducted components. This contrast highlights the critical role of circum-Pacific region, which has experienced nearly continuous subduction for at least the past 200 Myr, and may have acted as a long-existing “subduction shield”, limiting the dispersal of slab-derived materials into the Pacific mantle, and providing an ideal setting to examine how long-term subduction processes have contributed to upper mantle heterogeneity.  However, whether such large-scale geometry patterns can be reproduced dynamically, and whether it is geodynamically reasonable to classify the upper mantle into subduction-modified and subduction-unmodified domains mantle, remain open questions.

In this study, we employ CitcomS, a finite-element geodynamic code that solves thermo-chemical convection in a spherical shell, to simulate mantle convection and examine the transport of subduction-modified material through the upper mantle, constrained by GPlates-derived plate velocities based on published plate tectonic reconstruction model. Passive tracers are introduced to track material transport over time. By identifying tracers that pass through the mantle wedge, we determine materials acquire subduction signals and evaluate how they are redistributed within the convective mantle.

This analysis provides a quantitative framework for accessing whether the modeled mantle can be conceptually classified into subduction-modified and subduction-unmodified regions, and for investigating how long-term subduction contributes to global upper-mantle heterogeneity. More broadly, our results offer a new perspective for investigating the long-term dynamic evolution of the circum-Pacific subduction system and its role in shaping mantle structure.