Primitive He isotopic component shows greater affinity with hotter mantle plume
The geochemical studies of the mid-ocean-ridge basalt (MORB) and that of the volcanic hotspot lava have shown that the Earth's mantle is chemically and isotopically heterogeneous and that each of these types of lava reflects a geochemically distinct mantle source. MORBs come from the passive melting of the depleted upper mantle whereas the hotspots are fed by mantle plumes rooted in the deeper mantle and show more varied geochemical compositions. In addition, hot-spot lavas sometimes show higher 3He/4He isotopic ratios compared to low and uniform 3He/4He ratios found in MORBs. These high 3He/4He ratios reflect a primordial component, a relatively pristine reservoir within the Earth's lower mantle.
A long-standing hypothesis maintains that this primordial domain would be located within the lower mantle and thus the lavas of hotspots with high 3He/4He values would result from sampling plumes originating from the lower mantle. But, not all mantle plumes originating from the lower mantle sample high-3He/4He components. Some hotspots have 3He/4He values ranging from low to high and others show only low 3He/4He values.
Jackson et al. (2017) tried to explain the reason of these differences and bring new information about the depth distribution and thermochemical dynamics governing the composition of the mantle. Starting from the hypothesis that only the plumes sufficiently hot, and therefore sufficiently buoyant can trace the 3He/4He rich primitive material to the surface, they establish a relation between the plumes which feed the hot spots and their high contents in 3He/4He.
In order to study only hotspots connected to plumes that potentially sample the deep mantle, they identified plume-related hotspots from three plume catalogues (The French and Romanowicz model, Boschi-1 and Boschi-2 catalogues) and linked them with the maximum 3He/4He values sampled within the lavas of the corresponding hot spots. Then, to test whether mantle plumes that sample high-3He/4He values are hotter, they use shear wave anomalies (using the updated SMEAN2 model) since mantle regions with slow seismic shear-wave velocity anomalies are attributed to higher mantle temperatures. Because the relationship between maximum 3He/4He and seismic anomalies is clear only in the shallow upper mantle, they limit their shear wave anomalies study to a depth of 200 km[JP1] . They also use the recent hotspot buoyancy flux estimates from the MiFil volume method to compare them to the maximum 3He/4He values of hot spots.
Thus, by linking the database information of three plume catalogues, the associated high 3He / 4He lavas (from an expanded 3He/4He datasets), seismic anomalies to a depth of 200km and the recent hotspot buoyancy flux estimates (figure 1), new observations were made providing additional information on the origin of the high 3He/4He component sampled at hot spots.
Indeed, the shear-wave velocity anomalies at 200km depth and the buoyancy flux show each a relationship with the hotspots previously defined as being connected to a continuous plume (via the three plume catalogues) with high 3He/4He values. The correlation is positive with the hotspot buoyancy flux and the relationship is inverse with seismic anomalies (figure1). The anomalies of the lower shear-waves velocities identified at 200 km below the hotspots are probably related to the temperature increase associated with the plumes as the low shear-wave velocities in the shallow mantle are associated with higher temperatures. In addition, the high 3He/4He domain, the primordial domain is predicted to be Fe-rich, more dense and therefore the cooler plumes are not hot, buoyant enough to raise the denser, rich 3He/4He primordial materials. This study shows that although the hotspots are connected to a plume, only the most buoyant can raise the dense and high 3He/4He materials from the primordial domain into the shallow mantle.
But some hotspots that are associated with well-defined plume have low 3He/4He maximum values, not corresponding to a primitive reservoir. It is therefore possible to say that high 3He / 4He values are not necessarily a geochemical signature for deep-mantle plumes. But, this is true if we consider that the low maximum 3He / 4He values measured at these hotspots are representative. This can be explained if the lower mantle, the source of the plumes, is heterogeneous with domains having high-3He/4He values and others having low-3He/4He, non-primordial. Thus low and high 3He/4He values common to the same hot spot can be observed if the associated plume samples both the primitive domains and the domains associated with the depleted mantle or the recycled lithosphere.
The study of the helium isotope therefore gives us information on the relationships between surface magmatism and the thermo-chemical evolution of the Earth's mantle. Since only plumes sufficiently hot, and therefore sufficiently bouyant, can convey the high 3He/4He and denser primitive materials to the surface. This study brings new information to help us to understand how this denser primitive domain could be preserved over a long period of time.
Figure 1: Maximum 3He/4He values at plume-fed hotspots compared with seismic shear-wave velocity anomalies at 200 km and with hotspot buoyancy flux: The presence or absence of a plume is evaluated using two plume catalogues, Boschi-1 (left panels) and Boschi-2 (right panels)10; the French & Romanowicz plume catalogue is used to separate plume-fed and non-plume-fed hotspots. The Pearson correlation coefficients r are provided in the panels and are calculated from individual observations for plume-fed hotspots (for the red symbols only), non-plume-fed hotspots (blue symbols only), and all hotspots; in order to evaluate the actual statistical relevance of the correlations, the 1σ uncertainty is provided (calculated using bootstrap), as is the significance level of the correlation coefficients (P value in parentheses, calculated with Student’s t-test assuming normally distributed data). The sample sizes for the Boschi-1 and Boschi-2 plume catalogues are as follows: Boschi-1 (n = 23 plumes, n = 9 non-plumes); Boschi-2 (n = 21 plumes, n = 11 non-plumes). The SMEAN2 seismic model is an updated variant of a global composite shear-wave model (SMEAN). It is used because it captures the most common, robust, long-wavelength structure across different model. The hotspot buoyancy flux is from MiFil volume method (bottom panels). The maximum 3He/4He values at hotspots are also compared with shear-wave velocity anomalies (δv) (top panels). H = Hawaii, I = Iceland, S = Samoa, G = Galapagos, E = Easter, Y = Yellowstone, A = Afar, So = Societies, C = Cape Verde, Az = Azores, M = Macdonald, He = Heard, J = Juan Fernandez, Sh = St Helena.
