Search

Some seismic models derived from tomographic studies indicate elevated shear-wave velocities (≥4.7 km/s) around 120–150 km depth in cratonic lithospheric mantle. These velocities are higher than those of cratonic peridotites, even assuming a cold cratonic geotherm (i.e., 35 mW/m2 surface heat flux) and accounting for compositional heterogeneity in cratonic peridotite xenoliths and the effects of anelasticity. We reviewed various geophysical and petrologic constraints on the nature of cratonic roots (seismic velocities, lithology/mineralogy, electrical conductivity, and gravity) and explored a range of permissible rock and mineral assemblages that can explain the high seismic velocities. These constraints suggest that diamond and eclogite are the most likely high-Vs candidates to explain the observed velocities, but matching the high shear-wave velocities requires either a large proportion of eclogite (>50 vol.%) or the presence of up to 3 vol.% diamond, with the exact values depending on peridotite and eclogite compositions and the geotherm. Both of these estimates are higher than predicted by observations made on natural samples from kimberlites. However, a combination of ≤20 vol.% eclogite and ~2 vol.% diamond may account for high shear-wave velocities, in proportions consistent with multiple geophysical observables, data from natural samples, and within mass balance constraints for global carbon. Our results further show that cratonic thermal structure need not be significantly cooler than determined from xenolith thermobarometry.

We evaluate the factors influencing the abundance, [Li], and isotopic composition of riverine Li delivered to the oceans through analyses and modeling of [Li] and δ7Li in streams and groundwaters draining a single continental lithology, the Columbia River Basalts (CRBs). The streams were sampled in different climate zones that lie east (dry), and west (wet) of the Cascades Mountains, and during two different seasons (summer and late winter) in order to evaluate climatic and seasonal influences on Li isotopes in rivers. Dissolved Li (δLidis7=+9.3 to +30.4) is systematically heavier than that of fresh or weathered CRBs (-4.7 to +6.0, Liu et al., 2013), suspended loads (-5.9 to -0.3), and shallow groundwaters (+6.7 to +9.4), consistent with previous studies showing that Li isotope fractionation is affected by equilibration between stream water and secondary minerals. However, the lack of correlation between δ7Lidis and climate zone, the uniform secondary minerals and bedrock, coupled with the highly variable (>20‰) δLidis7 indicate that other factors exert a strong control on δ7Lidis. In particular, the heavier Li in streams compared to the shallow groundwaters that feed them indicates that continued isotopic fractionation between stream water and suspended and/or bed loads has a major influence on riverine δ7Li. Seasonal δ7Li variation is observed only for streams west of the Cascades, where the difference in precipitation rate between the dry and wet seasons is greatest. Reactive transport model simulations reveal that riverine δ7Li is strongly controlled by subsurface residence times and the Li isotope fractionation occurring within rivers. The latter explains why there is no positive correlation between δ7Li and traditional weathering proxies such as Si or normalized Si in rivers, as riverine Li isotope fractionation drives δ7Li to higher values during transport, whereas the concentrations of major cations and anions are diluted. The varying residence time for groundwaters feeding the western streams in summer (long residence times, higher δ7Li, greater weathering) and winter (short residence times, lower δ7Li, less weathering) explains the observed seasonal variations. A global, negative correlation between δ7Li and Li/Na for streams and rivers draining basaltic catchments reflects the overall transport time, hence the amount of silicate weathering. Based on our results, the increase of δ7Li in seawater during the Cenozoic is unlikely related to changing climate, but may reflect mountain building giving rise to increased silicate weathering.

Mineralogical, chemical, and Li and Nd isotopic compositions of two drill cores (8-9m deep) through bauxites developed on the Miocene Columbia River Basalts document the changes associated with basalt weathering, provide insights into the processes involved, and allow us to examine the overall influence of chemical weathering on juvenile (basaltic) crust. Gibbsite, hematite, ±kaolinite, halloysite, goethite, and maghemite are the weathering products in the bauxites. Quartz is observed near the tops of the cores and its abundance decreases progressively with depth; no quartz is observed below five meters depth in either core. Most major and trace elements, including "mobile" and some "immobile" elements are severely depleted in the bauxites. Niobium is less mobile relative to the rare earth elements, thus chemical weathering attenuates the negative Nb anomaly in the continental crust. Li and Nd are strongly depleted relative to fresh basalt, and both increase systematically towards the surface in the quartz-bearing samples while δ7Li and εNd values decrease systematically towards the surface in these same samples. Both Li and Nd were likely lost from the bauxites through leaching. The systematic enrichment of Li, Nd, and quartz, as well as the less radiogenic Nd isotopic composition at the tops of both profiles reflects 20-60wt.% addition of an eolian component to the soils. The eolian dust is unlikely to have experienced significant post-depositional weathering due to the relatively high Li contents near the tops of the profiles, and, therefore, the low δ7Li and εNd values suggest that the dust came from an old, weathered region of the continent. Our results demonstrate that lithium isotopes are sensitive tracers of chemical weathering, particularly in extreme weathering settings, and support the hypothesis that chemical weathering influences the mass and composition of the continental crust. © 2013 Elsevier Ltd.