TY - JOUR
T1 - Primitive off-rift basalts from Iceland and Jan Mayen
T2 - Os-isotopic evidence for a mantle source containing enriched subcontinental lithosphere
AU - Debaille, Vinciane
AU - Trønnes, Reidar G.
AU - Brandon, Alan D.
AU - Waight, Tod Earle
AU - Graham, David W.
AU - Lee, Cin-Ty A.
PY - 2009
Y1 - 2009
N2 - New measurements of Os, He, Sr and Nd isotopes, along with major and trace elements, are presented for basalts from the three volcanic flank zones in Iceland and from Jan Mayen Island. The 187Os/188Os ratios in lavas with <30 ppt Os (n = 4) are elevated compared to ratios in coexisting olivine and appear to be contaminated at a shallow level. The 187Os/188Os ratios in the remaining lavas with >30 ppt Os (n = 17) range between 0.12117 and 0.13324. These values are surprisingly low for oceanic island basalts and include some samples that are less than putative present-day primitive upper mantle (PUM with 187Os/188Os of 0.1296). These low 187Os/188Os preclude significant shallow-level contamination from oceanic crust. The 187Os/188Os ratios for Jan Mayen lavas are less than PUM, severely limiting the presence of any continental crust in their mantle source. A positive correlation between 143Nd/144Nd and 187Os/188Os ratios in Iceland and Jan Mayen lavas likely reflects the presence in their source of ancient subcontinental lithosphere that has undergone incompatible trace element enrichment that did not affect the Re-Os system. In addition, the Jan Mayen lava isotopic signature cannot be explained solely by the presence of subcontinental lithospheric mantle, and the influence of another geochemical component, such as a mantle plume appears required. Combined 87Sr/86Sr, 143Nd/144Nd, 3He/4He and 187Os/188Os data indicate a genetic relationship between Jan Mayen Island and the Iceland mantle plume. Material from the Iceland mantle plume likely migrates at depth until it reaches the tensional setting of the Jan Mayen Fracture Zone, where it undergoes low-degree partial melting. At a first-order, isotopic co-variations can be interpreted as broadly binary mixing curves between two primary end-members. One end-member, characterized in particular by its unradiogenic 187Os/188Os and 143Nd/144Nd, low 3He/4He and high 87Sr/86Sr, is represented by subcontinental lithospheric mantle stranded and disseminated in the upper mantle during the opening of the Atlantic Ocean. The second end-member corresponds to a hybrid mixture between the depleted-MORB mantle and the enriched Iceland mantle plume, itself resulting from mixing between recycled oceanic crust and depleted lower mantle. This hybrid accounts for the high 3He/4He (28 Ra), high 143Nd/144Nd (0.5132), high 187Os/188Os (0.14) and low 87Sr/86Sr (0.7026) composition observed in Iceland. Two different models may account for these observed mixing relationships between the end-members. In this first model, the Iceland mantle entrains pristine depleted material when rising in the upper mantle and allows refractory sub-lithospheric fragments to melt because of excess heat derived from the deep plume material. A second model that may better account for the Pb isotopic variations observed, uses the same components but where the depleted-MORB mantle is already polluted by subcontinental lithospheric mantle material before mixing with the Iceland mantle plume. Both cases likely occur. Though only three principal components are required to explain the isotopic variations of the Iceland-Jan Mayen system, the different possible mixing relationships may be accounted for by potentially a greater number of end-members.
AB - New measurements of Os, He, Sr and Nd isotopes, along with major and trace elements, are presented for basalts from the three volcanic flank zones in Iceland and from Jan Mayen Island. The 187Os/188Os ratios in lavas with <30 ppt Os (n = 4) are elevated compared to ratios in coexisting olivine and appear to be contaminated at a shallow level. The 187Os/188Os ratios in the remaining lavas with >30 ppt Os (n = 17) range between 0.12117 and 0.13324. These values are surprisingly low for oceanic island basalts and include some samples that are less than putative present-day primitive upper mantle (PUM with 187Os/188Os of 0.1296). These low 187Os/188Os preclude significant shallow-level contamination from oceanic crust. The 187Os/188Os ratios for Jan Mayen lavas are less than PUM, severely limiting the presence of any continental crust in their mantle source. A positive correlation between 143Nd/144Nd and 187Os/188Os ratios in Iceland and Jan Mayen lavas likely reflects the presence in their source of ancient subcontinental lithosphere that has undergone incompatible trace element enrichment that did not affect the Re-Os system. In addition, the Jan Mayen lava isotopic signature cannot be explained solely by the presence of subcontinental lithospheric mantle, and the influence of another geochemical component, such as a mantle plume appears required. Combined 87Sr/86Sr, 143Nd/144Nd, 3He/4He and 187Os/188Os data indicate a genetic relationship between Jan Mayen Island and the Iceland mantle plume. Material from the Iceland mantle plume likely migrates at depth until it reaches the tensional setting of the Jan Mayen Fracture Zone, where it undergoes low-degree partial melting. At a first-order, isotopic co-variations can be interpreted as broadly binary mixing curves between two primary end-members. One end-member, characterized in particular by its unradiogenic 187Os/188Os and 143Nd/144Nd, low 3He/4He and high 87Sr/86Sr, is represented by subcontinental lithospheric mantle stranded and disseminated in the upper mantle during the opening of the Atlantic Ocean. The second end-member corresponds to a hybrid mixture between the depleted-MORB mantle and the enriched Iceland mantle plume, itself resulting from mixing between recycled oceanic crust and depleted lower mantle. This hybrid accounts for the high 3He/4He (28 Ra), high 143Nd/144Nd (0.5132), high 187Os/188Os (0.14) and low 87Sr/86Sr (0.7026) composition observed in Iceland. Two different models may account for these observed mixing relationships between the end-members. In this first model, the Iceland mantle entrains pristine depleted material when rising in the upper mantle and allows refractory sub-lithospheric fragments to melt because of excess heat derived from the deep plume material. A second model that may better account for the Pb isotopic variations observed, uses the same components but where the depleted-MORB mantle is already polluted by subcontinental lithospheric mantle material before mixing with the Iceland mantle plume. Both cases likely occur. Though only three principal components are required to explain the isotopic variations of the Iceland-Jan Mayen system, the different possible mixing relationships may be accounted for by potentially a greater number of end-members.
M3 - Journal article
SN - 0016-7037
VL - 73
SP - 3423
EP - 3449
JO - Geochimica et Cosmochimica Acta
JF - Geochimica et Cosmochimica Acta
ER -