Search

This study aims to clarify the relationship between chemical weathering of rocks and the carbon budget of rivers and better understand the weathering mechanisms of plateau watersheds. We chose to study the Jinsha River, which originates from the Tibetan Plateau and also is in the upper reaches of the Changjiang River. Analysis of hydrochemistry, radiogenic strontium isotope and stable calcium isotopes were conducted of the Jinsha River water samples, which were collected along its mainstream and main tributaries in the summer. The results show that the water chemistry of the mainstream waters is dominated by evaporite weathering, which have low 87Sr/86Sr values (0.7098–0.7108) and wide range of Sr contents (2.70–9.35 μmol/L). In contrast, tributaries of the Jinsha River have higher 87Sr/86Sr (0.7090–0.7157) and lower Sr contents (∼1 μmol/L). Moreover, the Ca isotopic compositions in the mainstream (0.87–1.11‰) are heavier than the tributaries (0.68–0.88‰) and could not be fully explained by the conventional mixing of different sources. We suggest that secondary carbonate precipitation fractionates Ca isotopes in the Jinsha River, and fractionation factors are between 0.99935 and 0.99963. At least 66% of Ca was removed in the mainstream of the Jinsha River through secondary mineral precipitation, and the average value is ∼35% in the tributaries. The results highlight that evaporite weathering results in more carbonate precipitation influencing Ca transportation and cycling in the riverine system constrained by stable Ca isotopic compositions and water chemistry.

Extensional rift systems provide important pathways for the release of large amounts of deeply-sourced CO2 into the atmosphere. Continental rifting zones (e.g., East African rift) are thus invoked to be important for understanding the links between CO2 outgassing and global climate change associated with continental breakup. However, deeply-sourced CO2 emissions from extensional rift systems in continental collision zones remain poorly understood. Here, we focus on hydrothermal CO2 emissions from the southern segment of the Yadong-Gulu rift (YGR), the largest extensional rift in southern Tibetan Plateau, aiming at delineating rift-related CO2 emissions from the India-Asia continental collision zone. In-situ measurements of diffuse soil CO2 emissions indicate that average soil CO2 fluxes from the Kangbu, Mengzha, and Chaduo hydrothermal fields are 40, 700, and 255 g m−2 d−1, respectively. Combined with average soil CO2 fluxes (20–437 g m−2 d−1) from central and northern YGR, we speculate a relatively steady-state CO2 degassing pattern for the entire rift. The magnitude of soil CO2 fluxes of the YGR is higher than that of representative areas of the East African rift system. Evidence from 3He/4He reveals a pure crustal origin for CO2-bearing fluids in southern YGR, while the involvement of mantle CO2 is recognized in the central and northern rift segments. We suggest that metamorphic decarbonation of crustal rocks at variable depths is the primary cause for CO2 origin in southern YGR, which differs from the magma‑carbonate interaction model of central and northern YGR. Ternary mixing calculation based on He-CO2 systematics of hydrothermal gases indicates dominant contributions from carbonate rocks to total carbon inventory, with mantle CO2 contributions absent in southern YGR but discernible in northern YGR. The characteristic high proportions of crustal CO2 (>90%) distinguish the YGR from extensional rifts fed primarily by mantle CO2 in continental rifting zones. Our results reveal a crustal-scale carbon cycle in accretionary wedge of the India-Asia continental collision zone, together with magmatic front setting of the central and northern YGR, outlining a transect of the rift-related CO2 emissions across continental collision zones. This would contribute to better understanding the role of continental assembly in deeply-sourced CO2 emissions and global carbon budget.

Particulate inorganic carbon (PIC) is one of the most important components of carbon cycle in hydropower reservoirs. However, the processes of production, migration, and transformation of PIC in hydropower reservoirs are still unclear. As such, we investigated the concentration and carbon isotope composition of PIC and dissolved inorganic carbon and related environmental parameters of hydropower reservoirs on the Wujiang River, Southwest China, to understand these processes. The observed δ13C datasets demonstrated that the river PIC is derived from exogenous carbonate weathering and endogenous carbonate precipitation, whereas the reservoir PIC is dominated by authigenic PIC driven by phytoplankton. The reservoir PIC concentration and δ13CPIC showed seasonal stratification in the water profile, the extent of which depends on surface photosynthesis and bottom respiration. The influence of phytoplankton on the production and transformation of PIC was different among the reservoirs with different ages and trophic levels, and the δ13C technique can reveal these processes very well. This study improves the comprehensive understanding of carbon cycling in hydropower reservoirs and environmental effects of river damming.

The present study seeks to evaluate the application of the 238U-234U-230Th radioactive disequilibrium methodology for the determination of the regolith production rates in thick weathering profiles marked by long histories, encountered under various climate regimes, but still very little studied by these techniques. For this purpose, 238U-234U-230Th disequilibria have been analyzed in the top 11 m of a lateritic profile developed on a granitic bedrock in south China (Longnan, Jiangxi Province) under a subtropical climate. The results demonstrate that in such a weathering profile the determination of weathering rates from the analysis of U-series nuclides in bulk rock samples cannot be recovered by applying in one step to the entire alteration profile the modeling approach classically used to interpret the U-series nuclides, i.e. the “gain and loss” model. The modeling has to be made on subsections of relatively small size (

Lithium (Li) isotopes have been shown to be fractionated significantly during chemical weathering of silicate rocks. However, previous research has mainly been focused on basalt weathering, whereas the behavior of Li isotopes during the weathering of granite remains unclear. The Li isotopic compositions in the saprolite profile developed on granite was analyzed, as well as several individual primary minerals (K-feldspar, plagioclase, quartz and biotite) in the parent granite. Lithium concentration and the isotopic compositions of the individual primary minerals appear to be heterogeneous. Biotite is rich in Li (369.4 mg/kg), but the Li concentration is very low in other minerals (1.3–11.3 mg/kg). Quartz has the highest δ7Li value (+19.9‰); other minerals contain δ7Li within a fairly narrow range (+5.8‰ to +9.0‰). The saprolite samples were mainly composed of primary minerals and were characterized by low weathering intensity (CIA = 52–60). Lithium was distinctly lost in the saprolites relative to parent granite. Meanwhile, δ7Li values of saprolites were all lower than those of parent granites (−2.0‰ to +3.9‰ vs. +7.1‰, respectively), which systematically decreased with increasing weathering intensity. The mineralogical composition and 87Sr/86Sr ratio in the saprolites suggest that the biotite was preferentially weathered below 120 cm depth in the examined profile, whereas plagioclase weathering mainly occurred above 120 cm depth. The Li released during granite weathering was characteristically different from the isotopes in the primary minerals. Lithium isotopic fractionation during granite weathering is affected by adsorption and/or incorporation of Li by secondary minerals (or clay minerals), and the release of Li from biotite. This study highlights the fact that significant Li isotope fractionation occurs during Li leaching (relatively higher 7Li/6Li ratios) from biotite in the incipient stage of granite weathering, which may help to explain the higher δ7Li values found in the dissolved load from the catchments of low weathering intensity.

Rare earth elements (REEs) and major ions were determined in rainwater samples collected from Guiyang in the acid rain impacted zone of Southwest China. The concentrations of total dissolved and acid-soluble REEs in rainwater are higher than those of most of the world's cities as well as the local surface water. The dissolved REE concentrations are negatively correlated to pH in rainwater. This is consistent with acidification of rainwater resulting in increasing REE concentrations. Speciation calculations using the PHREEQC-Model predict that the free REE metal ion (i.e., Ln3 +), sulfate (LnSO4+), oxalate (LnOx+ and LnOx2−) and fluoride (LnF2 +) complexes were relatively important forms of dissolved REE. Although the contribution to rainwater acidity of oxalic acid is much lower than that of sulfuric acid, the proportion of REE-oxalate complexes is not lower than that of REE-sulfate complexes in rain waters. The rainwater is enriched in the middle REEs (MREE) compared to both the light REEs (LREE) and heavy REEs (HREE). REE-phosphate complexes, phosphatic minerals and a Fe-Mn-oxyhydroxide coating are more likely controls on the development of MREE enrichment in the rainwater. Complexation of sulfate and REE played little role in the development of MREE enrichment, even though sulfate is the most important anion in acidic rainwater. In the shale-normalized REE pattern plot, La showed a clear positive anomaly. La might have two main anthropogenic sources in rainwater. One of the important sources could be automobile emission; another main source of excess La in rainwater could be the REE fertilizer that has been widely used in the agriculture of China for approximately 30 years.

Bulk chemical composition and mineralogy were examined in three soil profiles and a deeper 11-meter profile weathering on a granite ridgeline under subtropical climate conditions in south China. The weathering sequence is delineated by mineralogy and major element variations. Apatite, biotite, hornblende and plagioclase dissolve early during weathering, resulting in nearly 100% Ca and Na loss and significant Mg, Fe and P depletion at depth. The K-feldspar reaction front begins at the depth of depletion of plagioclase, leading to a loss of ~ 80% of the K at the land surface near the bottom of the ridgeline and almost 100% at the top. Dissolution of quartz and other silicates releases about 60% of Si in the profiles. Kaolinite is the dominant clay mineral and it transforms to gibbsite in the uppermost layer. The soil horizon (upper 100 cm) is the zone dominated by pedogenic processes, including active biological activity, physical erosion and influx of high concentrations of atmospheric components (especially CO2 and O2). The pedogenic processes are characterized by low pH (~ 4.54 to 5.85), high clay content (kaolinite: 6–16 wt.%; gibbsite: 2–8 wt.%) and total organic carbon content (0.13–3.93%) and intensive fracturing and dissolution of quartz and K-feldspar compared with the lower horizons. Accumulation of organic material and resistant minerals downslope is attributed to down-ridge movement of water (termed here, interflow) and weathering products in the uppermost 100 cm down the ridgeline. Using a mass balance model calculation for the catena that assumes steady-state soil thickness at all sites, the bottom profile shows the highest apparent total chemical weathering loss rate (~ 14 g m− 2 y− 1) whereas the middle position shows the highest physical erosion loss rate (~ 44 g m− 2 y− 1). SiO2 accounts for about 84% of the chemical weathering outflux from the soil horizons along the ridgeline hillslope. Chemical weathering rates at the bottom of the hillslope may be accelerated by high concentrations of organic material and by dissolution of kaolinite. This study demonstrates that mineral reaction fronts in granite become separated over depth intervals of meters and that elemental fluxes and release mechanisms vary with position along a ridgeline catena.

Iron isotope compositions of suspended particulate matters (SPM) collected from the Aha Lake, an artificial lake in the karst area of Yun-Gui Plateau, and its tributaries in summer and winter were investigated for our understanding of the behavior of Fe isotopes during iron biogeochemical cycling in lake. δ56Fe values of SPM display statistically negative shift relative to IRMM-014. Samples from the lake display a range from −1.36‰ to −0.10‰ in summer and from −0.30‰ to −0.07‰ in winter, while river samples vary from −0.88‰ to 0.07‰ in summer and from −0.35‰ to −0.03‰ in winter. The average iron isotope composition of aerosol samples is + 0.10‰, which is very similar to that of igneous rocks (0.09‰). The SPM in most rivers and water column showed seasonal variation in δ56Fe value: the δ56Fe values of SPM in summer were lower than in winter. The seasonal variation in δ56Fe value of the riverine SPM should be ascribed to the change in source of particulate Fe and geochemical process in the watershed: More particulate Fe was leached from soil and produced by weathering of pyrite widely distributed in coal-containing strata. It is suggested that both allochthonous inputs and the redox iron cycling control the variations of δ56Fe values for SPM in lake. During summer stratification, an Fe cycle named “ferrous wheel” is established near the redox boundary where the upwardly diffusing Fe(II) is oxidized and the reactive Fe oxides formed will continuously sink back into the reduction zone to complete the cycle. The δ56Fe values for SPM reach the minima, −0.88‰ for DB station and −1.36‰ for LJK station, just near the redox boundary as a result of the Fe cycling, where a rough 45% to 76% of Fe in these particles was produced by the repetitive cycle. Due to random transportation and diffusion, δ56Fe values of the particles near the redox zone distributed into approximately a Gaussian shape. The good negative correlation existed between δ56Fe values and Fe/Al ratios for DB station, suggesting that they together can be used as good indicators of the redox-driven Fe transformations.

The sources of dissolved inorganic carbon (DIC) and rock weathering processes were studied in the Houzhai catchment, a typical karstic catchment in the Changjiang River Basin, Southwest China. The carbon isotopic compositions ( δ 13 C DIC ) of DIC in the samples collected from the catchment vary from − 13.5‰ to − 6.9‰, with a mean value of − 9.8‰. The DIC in the catchment is thereby considered to be mainly derived from soil CO 2 and weathering of carbonate rocks. The DIC concentrations and the δ 13 C DIC values show pronounced seasonal variations, with the lowest values being observed during the high flow season (from May to October). The logPCO 2 values in the waters were positively correlated with the DIC contents and negatively correlated to the saturation index of calcite (SIc) and δ 13 C DIC values. These observations indicate that CO 2 derived from organic matter oxidation plays an important role in the dissolution of carbonate for this typical karstic environment. Based on a chemical mass balance, the weathering rate of carbonate rocks in the Houzhai catchment was estimated to be approximately 133 t/km 2 /year or 584 × 10 3 mol/km 2 /year in terms of CO 2 consumption rate, which is higher than the reported values for other rivers in Southwest China.

Chemical and strontium isotopic compositions of Nanpanjiang River and Beipanjiang River in the upper reaches of Xijiang River draining the Yunnan–Guizhou Plateau, Southwest China, were measured with a main purpose to understand the rock weathering and associated CO 2 consumption rates in the karstic landform-dominated area. The chemical ion composition of the river water is characterized by dominance of Ca 2+ , Mg 2+ and HCO 3 − , and SO 4 2− , which account for more than 90% of total ion concentration. 87 Sr/ 86 Sr ratios of Nanpanjiang River water range between 0.7080 and 0.7140, while those of most Beipanjiang River water vary within a limited range from 0.7075 to 0.7085. Weathering of carbonate rocks dominate chemistry of major ions especially in the water of Beipanjiang River, whereas weathering of silicate minerals in the upper reaches of Nanpanjiang River basin is obvious in addition to carbonate weathering, mainly according to the Sr isotopic composition of the river water. Analyses of the stoichiometry of the river water show that the water chemistry is controlled by carbonate dissolution not only by carbonic but also by sulfuric acid, and indicate that sulfuric acid plays an important role in carbonate weathering. The sulfate in river water is mainly from atmospheric input by coal-combustion industries and from oxidation of sulfide minerals during weathering of coal-containing strata and coal mining in the catchment. The chemical weathering rates of silicate and carbonate and associated CO 2 consumption rates by both carbonic and sulfuric acid and by only carbonic acid are respectively estimated. The results show that chemical weathering rates of carbonate and silicate weathering in Beipanjiang River basin are higher than those of Nanpanjiang River basin, showing linkages of geology and geography in the catchment to crustal weathering. The involvement of sulfuric acid in carbonate weathering greatly enhances the carbonate weathering but lowers the CO 2 consumption rates, indicating that sulfide acid is an important agent to rock weathering and the role of sulfide acid should further be clarified in view of the regional and global budget of CO 2 .

Apatite and associated minerals, such as spessartite, tourmaline, beryl and alkali feldspar, from the Altay No. 3 granite pegmatite in Xinjiang, China, have highly variable concentrations of rare earth elements (REEs) and possess chondrite-normalized REE patterns that show a clear convex tetrad effect. The Y/Ho and Sr/Eu ratios of the apatite samples are extremely fractionated relative to the corresponding chondritic values. Based on the La/Lu, Sr/Eu and Y/Ho ratios of these minerals, the crystallization zones of the pegmatite can be divided into two groups that are interpreted to reflect two main stages of magma evolution, namely, a magmatic and a magmatic hydrothermal transition stage. The degree of the lanthanide tetrad effect in REE patterns of the apatite, expressed by the quantification factor TE 1,3 , ranges from 1.3 to 1.9. The size of the tetrads correlates with Y/Ho, Eu/Eu* and Sr/Eu ratios. In addition to the mineral samples from the pegmatite, the REE patterns of the altered wall rocks also show a convex tetrad effect. A mass balance calculation was carried out to constrain the REE partitioning between the aqueous fluid, crystalline phases and the residual melt. It was found that the REE patterns of these three phases are dominated by the concentration of the REEs in the initial pegmatite-forming melt. Accordingly, it is proposed that the REE tetrad effect and the fractionation of the chemically coherent isovalent element pairs in the studied mineral separates represent intrinsic features of the granitic pegmatite-forming melt. It appears unlikely that these geochemical characteristics developed as a result of melt/mineral/fluid interactions taking place during magma crystallization after its intrusion. Interaction of an external fluid with pegmatite magma before emplacement or with parent materials of the highly evolved granitic rocks before melting would be favorable for producing the REE tetrad effect in highly evolved granitic rocks.

Rainwater samples collected from Tokyo and nearby locations in Japan as well as from two locations over the East China Sea have been analyzed for their rare earth elements (REEs) and major ions. The major ion compositions of these samples show the contributions of oceanic, crustal, and anthropogenic sources, and suggest the relative importance of these sources. Rainwater samples collected in Tokyo and Chofu (a town about 8 km from metropolitan Tokyo) have distinctly higher SO42− and NO3− content than in rainwater samples from the coastal area and the East China Sea, whereas the major ion compositions of the latter two areas are dominated by sea salt. The rainwater samples also have variable REE compositions, with various enrichment of light REEs over heavy REEs. Rainwater samples from Chofu are significantly enriched in light REEs, whereas those collected in Tokyo have loess or shale-like REE composition. Rainwater from both places, however, has similar major ion compositions. The enrichment of light REEs in the Chofu rainwater samples corresponds to the typical signature of an anthropogenic source. The Y/Ho ratios in these samples are all higher than average crustal values, and in general, increase with increasing enrichment of light REEs. This indicates that physical mixing of crustal and oceanic sources of the REEs cannot account for the variations of the Y/Ho ratios. Furthermore, although the anthropogenic source of the REEs is generally characterized by a high Y/Ho ratio, factors such as particle/rainwater interactions, pH, and speciation of REEs may also be responsible for the variable abundance and fractionation of REEs and Y/Ho in these rainwaters.

Concentrations of major ions and Sr, and Sr isotopic compositions of two major river systems (Wujiang River and Yuanjiang River) in karst-dominated regions, and during winter were measured. The rivers and their tributaries have high total dissolved solid concentrations, with TZ+ ranging from 2.1 to 6.3 meq/l, which are significantly higher than that of the global river average. Calcium and HCO3−, followed by Mg2+ and SO42−, dominate the chemical composition of major ions in the river waters. Both Na++K+ and Cl−+Si account for 5–10% of the cations and anions, respectively. The waters of the Wujiang River have high Sr2+ concentrations and lower 87Sr/86Sr ratios (0.7077–0.7110), while those of the Qingshuijiang River, one of two major tributaries of the Yuanjiang River system, show higher Sr isotopic ratios (0.7090–0.7145) and low Sr concentrations. The lithology and geography of the river basins largely control the chemical and isotopic compositions of these rivers. Analyses of the stoichiometry of the river waters show that the water chemistry is controlled by carbonate dissolution by both carbonic and sulfuric acid (especially of the Wujiang River), and suggest an origin of sulfate mainly from atmospheric input because of coal combustion and/or oxidation of sulfide minerals in the catchments. We identify three main weathering sources: limestone, dolomite and silicates. Chemical and Sr isotopic composition of Wujiang River waters are characterized mainly by limestone weathering, while those of the Qingshuijiang River waters mainly by dolomite and silicate weathering. Chemical weathering rates of carbonate and silicate are estimated, and the results indicate that the chemical weathering rates of carbonate in the catchments are remarkably high, significantly higher than most large rivers. It is considered that presence of sulfuric acid in the rock weathering, overuse of land and deterioration of vegetation are most likely the key factors responsible for the high weathering rates of carbonate in the catchments.

The concentrations of dissolved ( Acid-soluble heavy metals appear to have similar distributions in the Changjiang Estuary. All of them increase in concentration across almost the full salinity range. The suspended matter-normalized concentrations of acid-soluble heavy metals are negatively correlated with the amount of suspended matter, implying that the distribution of acid-soluble heavy metals in the Changjiang Estuary is mainly controlled by the concentration of suspended matter and salinity. In the Changjiang estuarine waters, the proportion of heavy metals in the acid-soluble phase relative to the total concentration (dissolved plus acid-soluble) decreases in the following sequence: Fe>Co>Zn>Pb>Ni>Cu>Cd. The fraction in the acid-soluble phase ranges from 99% to 4%, indicating considerable diversity of particle reactivity among these heavy metals.

The studied Cenozoic basalts from eastern China have SiO2 contents ranging from 38.63 to 55.24 wt%, and Na2O + K2O from 3.1 to 9.4 wt%. They may be divided into three main types — tholeiitic, alkalic and potassic magmas. Variable trace-element abundances, together with the diverse major-element compositions, suggest that their origins involve both partial melting of mantle rocks at different depths, and subsequent fractional crystallization of clinopyroxene and olivine. Variations in the ratios of highly incompatible elements of the basalts in eastern China demonstrate two mantle components. The enriched mantle component, primarily represented by the basalts in NE China, show high LILE/Nb, Ba/Th, Ba/U, Th/U and low La/Ba ratios, typical of the Walvis Bay Ridge, Gough and Tristan da Cunha oceanic island basalts (OIB). These trace-element features support the identification of EM1 in their mantle source previously made by G.H. Xie and coworkers using Sr, Nd and Pb isotopes. This Dupal mantle source is inferred to be lithospheric and related to metasomatism during subduction, rather than due to the direct incorporation of continental crustal material into the mantle. The second component is represented by basalts from SE China which have the trace-element signatures of normal OIB, i.e. they do not have a Dupal anomaly. It appears most likely that upwelling of the asthenosphere and magma ascent induced the melting of overlying heterogeneous mantle lithosphere to generate the Cenozoic basalts of eastern China.

Chemical compositional variations of desert sands and loess, loess-like deposits, river deposits and saline lacustrine deposits collected mostly from the southern margin of the Taklimakan Desert, China, provide strong constraints on the chemical weathering processes of continental crust, and also on the provenance compositions of these Quaternary deposits. Constant but lower concentrations of Al relative to the average upper continental crust (UCC) in these Quaternary deposits could be attributed to chemical weathering of Al-rich minerals such as feldspar and mica, and removal of clay minerals by wind and water. Large variations of mobile Na, Ca, Mg and Sr compared to relatively immobile Al, K and Ba resulted from the chemical weathering of plagioclase and carbonate minerals, salinization and carbonation of the arid lands. Most trace elements are not significantly fractionated. Their variations in absolute concentration are caused by the dilution of quartz, evaporates and carbonate minerals. Variations in Sn concentrations probably depend on relative contents of some heavy minerals. High Sr/Ba, low Th/U, K/Cs ratios of the saline lake deposits suggest differences in aqueous chemistry of these elements. The occurrence of the M-type tetrad effect (a REE pattern with four convex curves) of REE compositions, found mainly in the river deposits, is considered the result of water-particle interaction during chemical weathering of crustal materials. Y/Ho ratios of these samples are significantly higher than those of chondrite and the UCC. This suggests different chemical behavior of these two elements, and possibly the enrichment of carbonate materials. The low LREE/HREE feature of the saline lake deposits can be interpreted as due to the different aqueous chemical behavior of the REE's. However, it can also be reasonably considered as the result of inheritance from their provenance, since LREE and LREE/HREE variations may be related to the geological setting of the source. According to some elements that are not significantly fractionated, the average provenance composition of these Quaternary deposits are probably characterized by enrichments of Ni, Cr and V over Fe, higher La/Th, lower La/Sc and Th/Sc, and lower Ta/W and Nb/W ratios, compared with the UCC composition.

Trace-element abundances and Sr, Nd isotopic compositions have been measured on three megacrysts of clinopyroxene, three of garnet and one of amphibole, as well as their host alkali basalts from the Huangyishan volcano, Kuandian, NE China. The Ce isotopic compositions of three host basalt samples have been also analyzed. All megacrysts studied here are shown to be cognate with the host basalts mainly based on the isotopic compositions and major-element distributions between silicate crystal and melt. The source mantle of Huangyishan basalts is isotopically and chemically heterogeneous, and must have been enriched in Rb, La and Sm relative to Sr, Ce and Nd, respectively, for a relatively long time. Highly incompatible trace-element ratios are also suggestive of enrichment of large-ion lithophile elements in the mantle compared with the mantle source of MORB and most OIB. The fine structure, i.e. the double-concave shape of REE patterns commonly found in alkali basalts, is considered to be an important clue to the origin of alkali basalts. Based on the REE partition between the megacrysts and host basalts, evaluation of several theoretical models which can account for the double-concave shape of REE patterns shows that alkali basalts with such patterns must have been in equilibrium or quasi-equilibrium with both garnet and clinopyroxene solid phases. The small positive Eu anomaly commonly observed in REE patterns in alkali basalts, in most cases, is actually the turning point of the double-curved REE pattern. Partial melting of a garnet peridotite source is the most probable process which can generate the alkali basalts under consideration.