Behaviour of high field strength elements in subduction zones: constraints from Kamchatka–Aleutian arc lavas

Abstract

Models explaining the characteristic depletion of High Field Strength Elements (HFSE) relative to elements of similar compatibility in subduction zone magmas invoke either (1) the presence of HFSE-rich minerals in the subduction regime or (2) a selectively lower mobility of HFSE during subduction metasomatism of the mantle. In order to investigate the properties of HFSE in subduction regimes closer, we performed high precision measurements of Nb/Ta, Zr/Hf, and Lu/Hf ratios together with 176Hf/ 177Hf analyses on arc rocks from Kamchatka and the western Aleutians. The volcanic rocks of the Kamchatka region comprise compositional end members for both fluid and slab melt controlled mantle regimes, thus enabling systematic studies on the HFSE mobility at different conditions in the subarc mantle. Hf–Nd isotope and systematic Zr/Hf and Lu/Hf covariations illustrate that Zr–Hf and Lu are immobile in fluid-dominated regimes. Hf–Nd isotope compositions furthermore indicate the presence of “Indian” type depleted mantle beneath Kamchatka, as previously shown for the Mariana and Izu–Bonin arcs. In addition to a depleted mantle component, Hf–Nd isotope compositions enable identification of a more enriched mantle wedge component in the back-arc (Sredinny Ridge) that most likely consists of mantle lithosphere. The ratios of Nb/Ta and Zr/Hf are decoupled in rocks from fluid-dominated sources, indicating that Nb and Ta can be enriched in the mantle by subduction fluids to a small extent. In contrast to the fluid-dominated regime in Central Kamchatka, the budget of HFSE and Lu in rocks from the Northern Kamchatka Depression and in adakitic rocks from the western Aleutians is significantly affected by slab melts that originate from subducted oceanic crust. Compositions of the rocks with the highest slab melt components in their source (Sr/Y>30) provide no evidence that either Nb/Ta or Zr/Hf ratios are fractionated at a globally significant scale during melting of subducted oceanic crust. Subduction processes are therefore an unlikely candidate to explain the terrestrial Nb–Ta paradox (i.e., the subchondritic Nb/Ta ratios in all accessible terrestrial silicate reservoirs).