Mesozoic thermal evolution of the southern African mantle lithosphere

The thermal structure of Archean and Proterozoic lithospheric terranes in southern Africa during the Mesozoic was evaluated by thermobarometry of mantle peridotite xenoliths erupted in alkaline magmas between 180 and 60 Ma. For cratonic xenoliths, the presence of a 150–200 °C isobaric temperature range at 5–6 GPa confirms original interpretations of a conductive geotherm, which is perturbed at depth, and therefore does not record steady state lithospheric mantle structure.

Xenoliths from both Archean and Proterozoic terranes record conductive limb temperatures characteristic of a “cratonic” geotherm (40 mW m⁻²), indicating cooling of Proterozoic mantle following the last major tectonothermal event in the region at 1 Ga and the probability of thick off-craton lithosphere capable of hosting diamond. This inference is supported by U–Pb thermochronology of lower crustal xenoliths [Schmitz and Bowring, 2003. Contrib. Mineral. Petrol. 144, 592–618].

The entire region then suffered a protracted regional heating event in the Mesozoic, affecting both mantle and lower crust. In the mantle, the event is recorded at 150 Ma to the southeast of the craton, propagating to the west by 108–74 Ma, the craton interior by 85–90 Ma and the far southwest and northwest by 65–70 Ma. The heating penetrated to shallower levels in the off-craton areas than on the craton, and is more apparent on the southern margin of the craton than in its western interior. The focus and spatial progression mimic inferred patterns of plume activity and supercontinent breakup 30–100 Ma earlier and are probably connected.

Contrasting thermal profiles from Archean and Proterozoic mantle result from penetration to shallower levels of the Proterozoic lithosphere by heat transporting magmas. Extent of penetration is related not to original lithospheric thickness, but to its more fertile character and the presence of structurally weak zones of old tectonism. The present day distribution of surface heat flow in southern Africa is related to this dynamic event and is not a direct reflection of the pre-existing lithospheric architecture.     

The Siberian lithosphere traverse: mantle terranes and the assembly of the Siberian Craton

The kimberlite fields scattered across the NE part of the Siberian Craton have been used to map the subcontinental lithospheric mantle (SCLM), as it existed during Devonian to Late Jurassic time, along a 1000-km traverse NE–SW across the Archean Magan and Anabar provinces and into the Proterozoic Olenek Province. 4100 garnets and 260 chromites from 65 kimberlites have been analysed by electron probe (major elements) and proton microprobe (trace elements). These data, and radiometric ages on the kimberlites, have been used to estimate the position of the local (paleo)geotherm and the thickness of the lithosphere, and to map the detailed distribution of specific rock types and mantle processes in space and time. A low geotherm, corresponding approximately to the 35 mW/m² conductive model of Pollack and Chapman [Tectonophysics 38, 279–296, 1977], characterised the Devonian lithosphere beneath the Magan and Anabar crustal provinces. The Devonian geotherm beneath the northern part of the area was higher, rising to near a 40 mW/m² conductive model. Areas intruded by Mesozoic kimberlites are generally characterised by this higher, but still ‘cratonic' geotherm. Lithosphere thickness at the time of kimberlite intrusion varied from ca. 190 to ca. 240 km beneath the Archean Magan and Anabar provinces, but was less (150–180 km) beneath the Proterozoic Olenek Province already in Devonian time. Thinner Devonian lithosphere (140 km) in parts of this area may be related to Riphean rifting. Near the northern end of the traverse, differences in geotherm, lithosphere thickness and composition between the Devonian Toluopka area and the nearby Mesozoic kimberlite fields suggest thinning of the lithosphere by ca. 50–60 km, related to Devonian rifting and Triassic magmatism. A major conclusion of this study is that the crustal terrane boundaries defined by geological mapping and geophysical data (extended from outcrops in the Anabar Shield) represent major lithospheric sutures, which continue through the upper mantle and juxtapose lithospheric domains that differ significantly in composition and rock-type distribution between 100 and 250 km depth. The presence of significant proportions of harzburgitic and depleted lherzolitic garnets beneath the Magan and Anabar provinces is concordant with their Archean surface geology. The lack of harzburgitic garnets, and the chemistry of the lherzolitic garnets, beneath most of the other fields are consistent with the Proterozoic surface rocks. Mantle sections for different terranes within the Archean portion of the craton show pronounced differences in bulk composition, rock-type distribution, metasomatic overprint and lithospheric thickness. These observations suggest that individual crustal terranes, of both Archean and Proterozoic age, had developed their own lithospheric roots, and that these differences were preserved during the Proterozoic assembly of the craton. Data from kimberlite fields near the main Archean–Proterozoic suture (the Billyakh Shear Zone) suggest that reworking and mixing of Archean and Proterozoic mantle was limited to a zone less than 100 km wide.     

The evolution of lithospheric mantle beneath the Kalahari Craton and its margins

The compositional structure and thermal state of the subcontinental lithospheric mantle (SCLM) beneath the Kalahari Craton and the surrounding mobile belts have been mapped in space and time using >3400 garnet xenocrysts from >50 kimberlites intruded over the period 520–80 Ma. The trace-element patterns of many garnets reflect the metasomatic refertilisation of originally highly depleted harzburgites and lherzolites, and much of the lateral and vertical heterogeneity observed in the SCLM within the craton is the product of such metasomatism. The most depleted, and possibly least modified, SCLM was sampled beneath the Limpopo Belt by early Paleozoic kimberlites; the SCLM beneath other parts of the craton may represent similar material modified by metasomatism during Phanerozoic time. In the SW part of the craton, the SCLM sampled by “Group 2” kimberlites (>110 Ma) is thicker, cooler and less metasomatised than that sampled by “Group 1” kimberlites (mostly ≤95 Ma) in the same area. Therefore, the extensively studied xenolith suite from the Group 1 kimberlites probably is not representative of primary Archean SCLM compositions. The relatively fertile SCLM beneath the mobile belts surrounding the craton is interpreted as largely Archean SCLM, metasomatised and mixed with younger material during Paleoproterozoic to Mesoproterozoic rifting and compression. This implies that at least some of the observed secular evolution in SCLM composition worldwide may reflect the reworking of Archean SCLM. There are strong correlations between mantle composition and the lateral variations in seismic velocity shown by detailed tomographic studies. Areas of relatively low Vp within the craton largely reflect the progressive refertilisation of the Archean root during episodes of intraplate magmatism, including the Bushveld (2 Ga) and Karroo (ca. 180 Ma) events; areas of high Vp map out the distribution of relatively less metasomatised Archean SCLM. The relatively low Vp of the SCLM beneath the mobile belts around the craton is consistent with its fertile composition. The seismic data may be used to map the lateral extent of different types of SCLM, taking into account the small lateral variations in the geotherm identified using the techniques described here.     

Lithosphere mapping beneath the North American plate

Major- and trace-element analyses of garnets from heavy-mineral concentrates have been used to derive the compositional and thermal structure of the subcontinental lithospheric mantle (SCLM) beneath 16 areas within the core of the ancient Laurentian continent and 11 areas in the craton margin and fringing mobile belts. Results are presented as stratigraphic sections showing variations in the relative proportions of different rock types and metasomatic styles, and the mean Fo content of olivine, with depth. Detailed comparisons with data from mantle xenoliths demonstrate the reliability of the sections.

In the Slave Province, the SCLM in most areas shows a two-layer structure with a boundary at 140–160 km depth. The upper layer shows pronounced lateral variations, whereas the lower layer, after accounting for different degrees of melt-related metasomatism, shows marked uniformity. The lower layer is interpreted as a subcreted plume head, added at ca. 3.2 Ga; this boundary between the layers rises to <100 km depth toward the northern and southern edges of the craton. Strongly layered SCLM suggests that plume subcretion may also have played a role in the construction of the lithosphere beneath Michigan and Saskatchewan.

Outside the Slave Province, most North American Archon SCLM sections are less depleted than similar sections in southern Africa and Siberia; this may reflect extensive metasomatic modification. In E. Canada, the degree of modification increases toward the craton margin, and the SCLM beneath the Kapuskasing Structural Zone is typical of that beneath Proterozoic to Phanerozoic mobile belts.

SCLM sections from several Proterozoic areas around the margin of the Laurentian continental core (W. Greenland, Colorado–Wyoming district, Arkansas) show discontinuities and gaps that are interpreted as the effects of lithosphere stacking during collisional orogeny. Some areas affected by Proterozoic orogenesis (Wyoming Craton, Alberta, W. Greenland) appear to retain buoyant, modified Archean SCLM. Possible juvenile Proterozoic SCLM beneath the Colorado Plateau is significantly less refractory. The SCLM beneath the Kansas kimberlite field is highly melt-metasomatised, reflecting its proximity to the Mid-Continent Rift System.

A traverse across the continent shows that the upper part of the cratonic SCLM is highly magnesian; the decrease in mg# with depth is interpreted as the cumulative effect of metasomatic modification through time. The relatively small variations in seismic velocity within the continental core largely reflect the thickness of this depleted layer. The larger drop in seismic velocity in the surrounding Proton and Tecton belts reflects the closely coupled changes in SCLM composition and geotherm.     

Regional patterns in the paragenesis and age of inclusions in diamond, diamond composition, and the lithospheric seismic structure of Southern Africa

The Archean lithospheric mantle beneath the Kaapvaal–Zimbabwe craton of Southern Africa shows ±1% variations in seismic P-wave velocity at depths within the diamond stability field (150–250 km) that correlate regionally with differences in the composition of diamonds and their syngenetic inclusions. Seismically slower mantle trends from the mantle below Swaziland to that below southeastern Botswana, roughly following the surface outcrop pattern of the Bushveld-Molopo Farms Complex. Seismically slower mantle also is evident under the southwestern side of the Zimbabwe craton below crust metamorphosed around 2 Ga. Individual eclogitic sulfide inclusions in diamonds from the Kimberley area kimberlites, Koffiefontein, Orapa, and Jwaneng have Re–Os isotopic ages that range from circa 2.9 Ga to the Proterozoic and show little correspondence with these lithospheric variations. However, silicate inclusions in diamonds and their host diamond compositions for the above kimberlites, Finsch, Jagersfontein, Roberts Victor, Premier, Venetia, and Letlhakane do show some regional relationship to the seismic velocity of the lithosphere. Mantle lithosphere with slower P-wave velocity correlates with a greater proportion of eclogitic versus peridotitic silicate inclusions in diamond, a greater incidence of younger Sm–Nd ages of silicate inclusions, a greater proportion of diamonds with lighter C isotopic composition, and a lower percentage of low-N diamonds whereas the converse is true for diamonds from higher velocity mantle. The oldest formation ages of diamonds indicate that the mantle keels which became continental nuclei were created by middle Archean (3.2–3.3 Ga) mantle depletion events with high degrees of melting and early harzburgite formation. The predominance of sulfide inclusions that are eclogitic in the 2.9 Ga age population links late Archean (2.9 Ga) subduction-accretion events involving an oceanic lithosphere component to craton stabilization. These events resulted in a widely distributed younger Archean generation of eclogitic diamonds in the lithospheric mantle. Subsequent Proterozoic tectonic and magmatic events altered the composition of the continental lithosphere and added new lherzolitic and eclogitic diamonds to the already extensive Archean diamond suite.     

Archean geotherms and supracrustal assemblages

Metamorphic mineral assemblages suggest the existence of variable geotherms and lithospheric thicknesses beneath late Archean continental crust. Archean granite-greenstone terranes reflect steep geotherms (50–70°C/km) while high-grade terranes reflect moderate geotherms similar to present continental crust with high heat flow (25–40°C/km). Corresponding lithosphere thicknesses for each terrane during the late Archean are 35–50 km and 50–75 km, respectively.Early Archean ( 3.0 b.y.) greenstones differ from late Archean ( 2.7 b.y.) greenstones by the rarity or absence of andesite and graywacke and the relative abundance of pelite, quartzite, and komatiite. Mature clastic sediments in early greenstones reflect shallow-water, stable-basin deposition. Such rocks, together with granite-bearing conglomerate and felsic volcanics imply the existence of still older granitic source terranes. The absence or rarity of andesite in early greenstones reflects the absence of tectonic conditions in which basaltic and tonalitic magmas are modified to produce andesite.A model is presented in which early Archean greenstones form at the interface between tonalite islands and oceanic lithosphere, over convective downcurrents; high-grade supracrustals form on stable continental edges or interiors; and late Archean greenstones form in intracontinental rifts over mantle plumes.     

The chemical-temporal evolution of lithospheric mantle underlying the North China Craton

Previous studies of samples of subcontinental lithospheric mantle (SCLM) that underlay the North China Craton (NCC) during the Paleozoic have documented the presence of thick Archean SCLM at this time. In contrast, samples of SCLM underlying the NCC during the Cenozoic are characterized by evidence for melt depletion during the Proterozoic, and relatively recent juvenile additions to the lithosphere. These observations, coupled with geophysical evidence for relatively thin lithosphere at present, have led to the conclusion that the SCLM underlying the NCC was thinned and modified subsequent to the late Paleozoic. In order to extend the view into both the Paleozoic and modern SCLM underlying the NCC, we examine mantle xenoliths and xenocrystic chromites extracted from three Paleozoic kimberlites (Tieling, Fuxian and Mengyin), and mantle xenoliths extracted from one Cenozoic basaltic center (Kuandian). Geochemical data suggest that most of the Kuandian xenoliths are residues of small degrees of partial melting from chemically primitive mantle. Sr-Nd-Hf isotopic analyses indicate that the samples were removed from long-term depleted SCLM that had later been variably enriched in incompatible elements. Osmium isotopic compositions of the two most refractory xenoliths are depleted relative to the modern convecting upper mantle and have model melt depletion ages that indicate melt depletion during Paleoproterozoic. Other relatively depleted xenoliths have Os isotopic compositions consistent with the modern convecting upper mantle. This observation is generally consistent with earlier data for xenoliths from other Cenozoic volcanic systems in the NCC and surrounding cratons. Thus, the present SCLM underlying the NCC has a complex age structure, but does not appear to retain materials with Archean melt depletion ages. Results for what are presumed to be early Paleozoic xenoliths from Teiling are generally highly depleted in melt components, e.g. have low Al₂O₃, but have also been metasomatically altered. Enrichment in light rare earth elements, low ε_Nd values (∼−10), and relatively high ⁸⁷Sr/⁸⁶Sr (0.707-0.710) are consistent with a past episode of metasomatism. Despite the metasomatic event, ¹⁸⁷Os/¹⁸⁸Os ratios are low and consistent with a late Archean melt depletion event. Thus, like results for xenoliths from other early Paleozoic volcanic centers within the NCC, these rocks sample dominantly Archean SCLM. The mechanism for lithospheric thinning is still uncertain. The complex age structure currently underlying the NCC requires either variable melt depletion over the entire history of this SCLM, or the present lithospheric material was partly or wholly extruded under the NCC from elsewhere by the plate collisions (collision with the Yangtze Craton and/or NNW subduction of the Pacific plate) that may have caused the thinning to take place.     

Global 1° × 1° thermal model TC1 for the continental lithosphere: Implications for lithosphere secular evolution

This paper reports a new 1° × 1° global thermal model for the continental lithosphere (TC1). Geotherms for continental terranes of different ages (> 3.6 Ga to present) constrained by reliable data on borehole heat flow measurements (Artemieva, I.M., Mooney, W.D. 2001. Thermal structure and evolution of Precambrian lithosphere: a global study. J. Geophys. Res 106, 16387–16414.), are statistically analyzed as a function of age and are used to estimate lithospheric temperatures in continental regions with no or low-quality heat flow data (ca. 60% of the continents). These data are supplemented by cratonic geotherms based on electromagnetic and xenolith data; the latter indicate the existence of Archean cratons with two characteristic thicknesses, ca. 200 and > 250 km. A map of tectono-thermal ages of lithospheric terranes complied for the continents on a 1° × 1° grid and combined with the statistical age relationship of continental geotherms (z = 0.04  t + 93.6, where z is lithospheric thermal thickness in km and t is age in Ma) formed the basis for a new global thermal model of the continental lithosphere (TC1). The TC1 model is presented by a set of maps, which show significant thermal heterogeneity within continental upper mantle, with the strongest lateral temperature variations (as large as 800 °C) in the shallow mantle. A map of the depth to a 550 °C isotherm (Curie isotherm for magnetite) in continental upper mantle is presented as a proxy to the thickness of the magnetic crust; the same map provides a rough estimate of elastic thickness of old (> 200 Ma) continental lithosphere, in which flexural rigidity is dominated by olivine rheology of the mantle.Statistical analysis of continental geotherms reveals that thick (> 250 km) lithosphere is restricted solely to young Archean terranes (3.0–2.6 Ga), while in old Archean cratons (3.6–3.0 Ga) lithospheric roots do not extend deeper than 200–220 km. It is proposed that the former were formed by tectonic stacking and underplating during paleocollision of continental nuclei; it is likely that such exceptionally thick lithospheric roots have a limited lateral extent and are restricted to paleoterrane boundaries. This conclusion is supported by an analysis of the growth rate of the lithosphere since the Archean, which does not reveal a peak in lithospheric volume at 2.7–2.6 Ga as expected from growth curves for juvenile crust.A pronounced peak in the rate of lithospheric growth (10–18 km³/year) at 2.1–1.7 Ga (as compared to 5–8 km³/year in the Archean) well correlates with a peak in the growth of juvenile crust and with a consequent global extraction of massif-type anorthosites. It is proposed that large-scale variations in lithospheric thickness at cratonic margins and at paleoterrane boundaries controlled anorogenic magmatism. In particular, mid-Proterozoic anorogenic magmatism at the cratonic margins was caused by edge-driven convection triggered by a fast growth of the lithospheric mantle at 2.1–1.7 Ga. Belts of anorogenic magmatism within cratonic interiors can be caused by a deflection of mantle heat by a locally thickened lithosphere at paleosutures and, thus, can be surface manifestations of exceptionally thick lithospheric roots. The present volume of continental lithosphere as estimated from the new global map of lithospheric thermal thickness is 27.8 (± 7.0) × 10⁹ km³ (excluding submerged terranes with continental crust); preserved continental crust comprises ca. 7.7 × 10⁹ km³. About 50% of the present continental lithosphere existed by 1.8 Ga.     

The thermal structure and thickness of continental roots

We compare heat flow data from the Precambrian shields in North America and in South Africa. We also review data available in other less well-sampled Shield regions. Variations in crustal heat production account for most of the variability of the heat flow. Because of this variability, it is difficult to define a single average crustal model representative of a whole tectonic province. The average heat flow values of different Archean provinces in Canada, South Africa, Australia and India differ by significant amounts. This is also true for Proterozoic provinces. For example, the heat flow is significantly higher in the Proterozoic Namaqua–Natal Belt of South Africa than in the Grenville Province of the Canadian Shield (61 vs. 41 mW m⁻² on average). These observations indicate that it is not possible to define single value of the average heat flow for all provinces of the same crustal age. Large amplitude short wavelength variations of the heat flow suggest that most of the difference between Proterozoic and Archean heat flow is of crustal origin. In eastern Canada, there is no good correlation between the local values of heat flow and heat production. In the Archean, Proterozoic and Paleozoic provinces of eastern Canada, heat flow values through rocks with the same heat production are not significantly different. There is therefore no evidence for variations of the mantle heat flow beneath these different provinces. After removing the local crustal heat production from the surface heat flow, the mantle (Moho) heat flow was estimated to be between 10–15 mW m⁻² in the Archean, Proterozoic and Paleozoic provinces of eastern Canada. Estimates of the mantle heat flow in the Kaapvaal craton of South Africa may be slightly higher (≈17 mW m⁻²). Large-scale variations of bulk crustal heat production are well-documented in Canada and imply significant differences of deep lithospheric thermal structure. In thick lithosphere, surficial heat flow measurements record a time average of heat production in the lithospheric mantle and are not in equilibrium with the instantaneous heat production. The low mantle heat flow and current estimates of heat production in the lithospheric mantle do not support a mechanical (conductive) lithosphere thinner than 200 km and thicker than 330 km. Temperature anomalies with surrounding oceanic mantle extend to the convective boundary layer below the conductive layer, and hence to depths greater than these estimates. Mechanical and thermal stability of the lithosphere require the mantle part of the lithosphere to be chemically buoyant and depleted in radiogenic elements. Both characteristics are achieved simultaneously by partial melting and melt extraction.     

Origin of layered continental mantle (Karelian craton, Finland): Geochemical and Re–Os isotope constraints

Studies of mantle xenolith and xenocryst studies have indicated that the subcontinental lithospheric mantle (SCLM) at the Karelian Craton margin (Fennoscandian Shield) is stratified into at least three distinct layers cited A, B, and C. The origin and age of this layering has, however, remained unconstrained. In order to address this question, we have determined Re–Os isotope composition and a comprehensive set of major and trace elements, from xenoliths representing all these three layers. These are the first Re–Os data from the SCLM of the vast East European Craton.

Xenoliths derived from the middle layer B (at  110–180 km depth), which is the main source of harzburgitic garnets and peridotitic diamonds in these kimberlites, are characterised by unradiogenic Os isotopic composition. ¹⁸⁷Os/¹⁸⁸Os shows a good correlation with indices of partial melting implying an age of  3.3. Ga for melt extraction. This age corresponds with the oldest formation ages of the overlying crust, suggesting that layer B represents the unmodified SCLM stabilised during the Paleoarchean. Underlying layer C (at 180–250 km depths) is the main source of Ti-rich pyropes of megacrystic composition but is lacking harzburgitic pyropes. The osmium isotopic composition of layer C xenoliths is more radiogenic compared to layer B, yielding only Proterozoic T_RD ages. Layer C is interpreted to represent a melt metasomatised equivalent to layer B. This metasomatism most likely occurred at ca. 2.0 Ga when the present craton margin formed following continental break-up. Shallow layer A (at  60–110 km depth) has knife-sharp lower contact against layer B indicative of shear zone and episodic construction of SCLM. Layer A peridotites have “ultradepleted” arc mantle-type compositions, and have been metasomatised by radiogenic ¹⁸⁷Os/¹⁸⁸Os, presumably from slab-derived fluids. Since layer A is absent in the core of the craton, its origin can be related to Proterozoic processes at the craton margin. We interpret it to represent the lithosphere of a Proterozoic arc complex (subduction wedge mantle) that became underthrusted beneath the craton margin crust during continental collision  1.9 Ga ago.     

Variable involvements of mantle plumes in the genesis of mid-Neoproterozoic basaltic rocks in South China: A review

Ca. 825–720 Ma global continental intraplate magmatism is generally linked to mantle plumes or a mantle superplume that caused rifting and fragmentation of the supercontinent Rodinia. Widespread Neoproterozoic igneous rocks in South China are dated at ca. 825–760 Ma. There is a hot debate on their petrogenesis and tectonic affiliations, i.e., mantle plume/rift settings or collision/arc settings. Such competing interpretations have contrasting implications to the position of South China in the supercontinent Rodinia and in Rodinia reconstruction models.Variations in the bulk-rock compositions of primary basaltic melts can provide first order constraints on the mantle thermal–chemical structure, and thus distinguish between the plume/rift and arc/collision models. Whole-rock geochemical data of 14 mid-Neoproterozoic (825–760 Ma) basaltic successions are reviewed here in order to (1) estimate the primary melts compositions; (2) calculate the melting conditions and mantle potential temperature; and (3) identify the contributions of subcontinental lithosphere mantle (SCLM) and asenthospheric mantles to the generation of these basaltic rocks.In order to quantify the mantle potential temperatures and percentages of decompression melting, the primary MgO, FeO, and SiO₂ contents of basalts are calculated through carefully selecting less-evolved samples using a melting model based on the partitioning of FeO and MgO in olivine. The mid-Neoproterozoic (825–760 Ma) potential temperatures predicted from the primary melts range from 1390 °C to 1630 °C (mostly > 1480 °C), suggesting that most 825–760 Ma basaltic rocks in South China were generated by melting of anomalously hot mantle sources with potential temperatures 80–200 °C higher than the ambient Middle Ocean Ridge Basalt (MORB)-source mantle.The mantle source regions of these Neoproterozoic basaltic rocks have complex histories and heterogeneous compositions. Enriched mantle sources (e.g., pyroxenite and eclogite) are recognized as an important source for the Bikou and Suxiong basalts, suggesting that their generations may have involved recycled components. Trace elements variations show that interactions between asthenospheric mantle (OIB-type mantle) and SCLM played a very important role in generation of the 825–760 Ma basalts. Our results indicate that the SCLM metasomatized by subduction-induced melts/fluids during the 1.0–0.9 Ga orogenesis as a distinct geochemical reservoir that contributed significantly to the trace-elements and isotope inventory of these basalts.The continental intraplate geochemical signatures (e.g., OIB-type), high mantle potential temperatures and recycled components suggest the presence of a mantle plume beneath the Neoproterozoic South China block. We use the available data to develop an integrated plume-lithosphere interaction model for the ca. 825–760 Ma basalts. The early phases of basaltic rocks (825–810 Ma) were most likely formed by melting within the metasomatized SCLM heated by the rising mantle plume. The subsequent continental rift allowed adiabatic decompression partial melting of an upwelling mantle plumes at relatively shallow depth to form the widespread syn-rifting basaltic rocks at ca. 810–800 Ma and 790–760 Ma.     

Microcontinents among the accretionary complexes of the Central Asia Orogenic Belt: In situ Re–Os evidence

Major elements and Re–Os isotope ratios were analysed in situ on individual sulfide grains in spinel peridotite xenoliths hosted by Quaternary intraplate basalts from the Tariat volcanic field, Central Mongolia. The sulfides are dominantly high-temperature (>900 °C) Fe-rich monosulfide solid solution (MSS). Some sulfides with low Ni contents may be residual MSS, whereas other sulfides defining a negative Ni–Cu correlation may be crystallization products of fractionated sulfide melts. The subchondritic ¹⁸⁷Re/¹⁸⁸Os and ¹⁸⁷Os/¹⁸⁸Os of some sulfides also indicate they are residual MSS. Os isotope compositions of sulfides reveal the presence of Archean to Proterozoic lithospheric mantle beneath the region. The sulfides have T_RD model ages ranging from 3.0 to 0.2 Ga, with peaks at 1.5–1.3, 1 and 0.7–0.5 Ga. The peak ages are indicative of significant events in the lithospheric mantle at those times. The timing of these events is remarkably consistent with those of the major crust-building events within the Tarvagatay Terrane where the Tariat volcanic field is located. The similarity in the ranges of crustal U–Pb ages and Nd model ages, and our sulfide Os model ages, suggests that the sulfide ages may date metasomatic events in the underlying lithospheric mantle, which were related to tectonothermal events that affected the overlying crust. Radiometric ages from the Tarvagatay Terrane appear to correspond to the Archean model ages from its SCLM counterpart. The last two events (1.1 and 0.7–0.5 Ga) recorded in the Tarvagatay Terrane suggest involvement of the “CAOB mantle” and development of significant juvenile crustal growth in the orogeny.     

Thermal state of the lithosphere beneath Central Mongolia: evidence from deep-seated xenoliths from the Shavaryn-Saram volcanic centre in the Tariat depression, Hangai, Mongolia

A suite of garnet-two pyroxene granulites, garnet pyroxenites and garnet peridotites from the pyroclastic facies of the Shavaryn-Saram volcanic centre in the Tariat depression in the northern part of the Hangai dome, Central Mongolia, yields pressure and temperature information for the lower crust and upper mantle in that region. Although a real geotherm cannot be constructed because of the common zoning of the minerals in some of the xenoliths, it can be inferred that the P-T locus from about 900 °C at 45 km to 1050 °C at 60 km defines a likely approximate geothermal gradient for the region around the time of entrainment of the xenoliths (about 1 Ma ago). This geothermal gradient is high relative to cratonic geotherms but is 50–100 °C lower than that for typical alkali basaltic provinces worldwide. The crust-mantle boundary inferred from the incoming of ultramafic rock types in this region is located at about 45 km and granulite rock types extend well into the mantle. This interpretation is consistent with the most recent seismic sections for the area.

Analytical data for major and trace elements (by electron- and proton-microprobe respectively) in clinopyroxenes indicate that the Cr-diopside series xenoliths are enriched in basaltic components (including Al₂O₃, Na₂O, TiO₂, Sr, Y and Zr).

The combination of elevated temperature and fertile composition of the uppermost mantle as revealed by the xenoliths could explain the observed anomalous seismic signatures seen beneath this region.     

Mantle structure and rifting processes in the Baikal–Mongolia region: geophysical data and evidence from xenoliths in volcanic rocks

The origin of the Baikal rift zone (BRZ) has been debated between the advocates of passive and active rifting since the 1970s. A re-assessment of the relevant geological and geophysical data from Russian and international literature questions the concept of broad asthenospheric upwelling beneath the rift zone that has been the cornerstone of many “active rifting” models. Results of a large number of early and recent studies favour the role of far-field forces in the opening and development of the BRZ. This study emphasises the data obtained through studies of peridotite and pyroxenite xenoliths brought to the surface by alkali basaltic magmas in southern Siberia and central Mongolia. These xenoliths are direct samples of the upper mantle in the vicinity of the BRZ. Of particular importance are suites of garnet-bearing xenoliths that have been used to construct P–T- composition lithospheric cross-sections in the region for the depth range of 35–80 km.Xenolith studies have shown fundamental differences in the composition and thermal regime between the lithospheric mantle beneath the ancient Siberian platform (sampled by kimberlites) and beneath younger mobile belts south of the platform. The uppermost mantle in southern Siberia and central Mongolia is much hotter at similar levels than the mantle in the Siberian craton and also has significantly higher contents of ‘basaltic’ major elements (Ca, Al, Na) and iron, higher Fe/Si and Fe/Mg. The combination of the moderately high geothermal gradient and the fertile compositions in the off-cratonic mantle appears to be a determining factor controlling differences in sub-Moho seismic velocities relative to the Siberian craton. Chemical and isotopic compositions of the off-cratonic xenoliths indicate small-scale and regional mantle heterogeneities attributed to various partial melting and enrichment events, consistent with long-term evolution in the lithospheric mantle. Age estimates of mantle events based on Os–Sr–Nd isotopic data can be correlated with major regional stages of crustal formation and may indicate long-term crust–mantle coupling. The ratios of ^143/144Nd in many LREE-depleted xenoliths are higher than those in MORB or OIB source regions and are not consistent with a recent origin from asthenospheric mantle.Mantle xenoliths nearest to the rift basins (30–50 km south of southern Lake Baikal) show no unequivocal evidence for strong heating, unusual stress and deformation, solid state flow, magmatic activity or partial melting that could be indicative of an asthenospheric intrusion right below the Moho. Comparisons between xenoliths from older and younger volcanic rocks east of Lake Baikal, together with observations on phase transformations and mineral zoning in individual xenoliths, have indicated recent heating in portions of the lithospheric mantle that may be related to localised magmatic activity or small-scale ascent of deep mantle material. Overall, the petrographic, P–T, chemical and isotopic constraints from mantle xenoliths appear to be consistent with recent geophysical studies, which found no evidence for a large-scale asthenospheric upwarp beneath the rift, and lend support to passive rifting mechanism for the BRZ.     

中国东部显生宙地幔演化的主要样式 :“蘑菇云”模型

从地幔类型、地球物理资料、岩石学研究成果 3方面讨论了中国东部显生宙地幔演化的主要样式 ,认为古老岩石圈地幔物质以亏损玄武质组分的橄榄岩为主 ,岩石密度低 ,具上浮性质 ,是克拉通稳定的主要原因 ,不可能发生因重力诱发的拆沉作用。白垩纪晚期—新生代地幔成分为饱满的二辉橄榄岩 ,地温高 ,明显有别于古老地幔。该热地幔物质呈“蘑菇云”状上涌 ,蘑菇云之间仍有古老地幔的残留体 ,二者多数呈陡边界接触 ,这是东部地幔演化的主要样式。与此同时岩石圈伸展导致岩石圈减薄。中生代中国东部绝大部分地区地温很高、地壳厚度大 ,但许多问题尚待研究。古老岩石圈地幔中的低速、低阻物质为流体和低熔程度的熔体 ,它们呈脉状、透镜状 ;新生代地幔中的该物质为含熔体的软流圈物质 ,它们呈多个蘑菇云状 ;中生代地幔中的低速物质可能为大量的玄武质熔体 ,呈蘑菇云平流层状。中生代壳幔相互作用明显 ,克拉通稳定时期及新生代时期层圈之间相互作用的活跃带位于岩石圈与软流圈之间。     

Geochemical characteristics of lava-field basalts from eastern Australia and inferred sources: Connections with the subcontinental lithospheric mantle?

 A large new database of major, trace elements and Sr-Nd isotopic ratios from 11 lava-field provinces in New South Wales and Queensland, eastern Australia allows detailed interpretation of the origin of these basaltic magmas. Isotopic signatures and trace element patterns identify an OIB-type (oceanic island basalt) source as a dominant component for most of these and some provinces appear to have additional significant components derived from the subcontinental lithospheric mantle (SCLM). The SCLM components have geochemical characteristics that overlap those observed in spinel lherzolite xenoliths (samples of shallow lithospheric mantle) from eastern Australia. These SCLM components show geochemical provinciality that indicates the occurrence of distinct geochemical lithospheric domains reflecting the timing and style of tectonic evolution of different regions. One component reflects modification by subduction-related processes during the late Paleozoic and Mesozoic, one records enrichment by fluids during old metasomatic events and another suggests a metasomatic event involving a distinctive amphibole and apatite-style enrichment. The composition and age distribution of volcanic lava-field provinces older than 10 Ma are consistent with a model involving a regional upwelling (elongated N–S along eastern Australia) of deep hot mantle related to marginal rifting and with OIB-type source geochemical characteristics. Thermal inhomogeneities within this plume swath resulted in small diapirs which may have undergone melt segregation at about 100 km and incorporated varying amounts of SCLM components there or from higher levels of the SCLM during ascent. Subsequent hot-spot generated central volcanoes overprinted this lava-field volcanism, tapped a similar OIB-type source component and truncated the thermal events. Accepted: 15 March 1995     

Xenolith evidence for polybaric melting and stratification of the upper mantle beneath South China

Mantle xenoliths from Hainan and Qilin, South China have been studied to constrain the nature of the upper mantle and mantle processes beneath a continental margin. The extremely low Ti (160–245 ppm) contents in clinopyroxenes from some spinel lherzolites, indicative of high degrees of partial melting are inconsistent with the relatively high clinopyroxene modes (7.4–13%) in these samples. This inconsistency could be due to polybaric melting that started in the garnet stability field, then, after the breakdown of garnet to pyroxene and spinel, continued in the spinel stability field. Polybaric melting, due to adiabatic decompression of upwelling mantle, would leave a residual mantle in which the degree of depletion decreases with depth. The predicted stratified lithospheric mantle is evidenced by the negative correlation between the forsterite content in olivine and the equilibration temperature, proportional to the depth in the lithosphere from which the xenolith was derived. The lower part of the lithospheric mantle beneath South China consists predominantly of fertile and moderately depleted peridotites, which are either devoid of LREE enrichment, or show the trace element signature of incipient metasomatism, and plot within the Phanerozoic mantle domain. In contrast, the upper part of the mantle contains harzburgite and cpx-poor lherzolite, which are strongly affected by metasomatism of melt/fluid of highly variable composition. The anomalously high orthopyroxene mode (up to 47%) makes some of these refractory samples compositionally similar to the Proterozoic/Archean mantle. Their low equilibrium temperature (800–900 °C) points to the presence of old lithospheric relicts in the uppermost mantle beneath South China. Such lithosphere architecture may have resulted from partial replacement of Archean–Proterozoic lithosphere by asthenosphere that rose adiabatically subsequent to lithospheric thinning during the Cenozoic.     

Generation of calc-alkaline andesite of the Tatun volcanic group (Taiwan) within an extensional environment by crystal fractionation

The Pliocene–Pleistocene northern Taiwan volcanic zone (NTVZ) is located within a trench-arc–back-arc basin and oblique arc–continent collision zone. Consequently the origin and tectonic setting of the andesitic rocks within the NTVZ and their relation to other circum-Pacific volcanic island-arc systems is uncertain. Rocks collected from the Tatun volcanic group (TTVG) include basaltic to andesitic rocks. The basalt is compositionally similar to within-plate continental tholeiites whereas the basaltic andesite and andesite are calc-alkaline; however, all rocks show a distinct depletion of Nb-Ta in their normalized incompatible element diagrams. The Sr-Nd isotope compositions of the TTVG rocks are very similar and have a relatively restricted range (i.e. I_Sr = 0.70417–0.70488; εNd_(T) = +2.2 to +3.1), suggesting that they are derived directly or indirectly from the same mantle source. The basalts are likely derived by mixing between melts from the asthenosphere and a subduction-modified subcontinental lithospheric mantle (SCLM) source, whereas the basaltic andesites may be derived by partial melting of pyroxenitic lenses within the SCLM and mixing with asthenospheric melts. MELTS modelling using a starting composition equal to the most primitive basaltic andesite, shallow-pressure (i.e. ≤1 kbar), oxidizing conditions (i.e. FMQ +1), and near water saturation will produce compositions similar to the andesites observed in this study. Petrological modelling and the Sr-Nd isotope results indicate that the volcanic rocks from TTVG, including the spatially and temporally associated Kuanyinshan volcanic rocks, are derived from the same mantle source and that the andesites are the product of fractional crystallization of a parental magma similar in composition to the basaltic andesites. Furthermore, our results indicate that, in some cases, calc-alkaline andesites may be generated by crystal fractionation of mafic magmas derived in an extensional back-arc setting rather than a subduction zone setting.     

Constraints on the thermal evolution of continental lithosphere from U-Pb accessory mineral thermochronometry of lower crustal xenoliths,southern Africa

U-Pb isotopic thermochronometry of rutile, apatite and titanite from kimberlite-borne lower crustal granulite xenoliths has been used to constrain the thermal evolution of Archean cratonic and Proterozoic off-craton continental lithosphere beneath southern Africa. The relatively low closure temperature of the U-Pb rutile thermochronometer (~400-450 °C) allows its use as a particularly sensitive recorder of the establishment of "cratonic" lithospheric geotherms, as well as subsequent thermal perturbations to the lithosphere. Contrasting lower crustal thermal histories are revealed between intracratonic and craton margin regions. Discordant Proterozoic (1.8 to 1.0 Ga) rutile ages in Archean (2.9 to 2.7 Ga) granulites from within the craton are indicative of isotopic resetting by marginal orogenic thermal perturbations influencing the deep crust of the cratonic nucleus. In Proterozoic (1.1 to 1.0 Ga) granulite xenoliths from the craton-bounding orogenic belts, rutiles define discordia arrays with Neoproterozoic (0.8 to 0.6 Ga) upper intercepts and lower intercepts equivalent to Mesozoic exhumation upon kimberlite entrainment. In combination with coexisting titanite and apatite dates, these results are interpreted as a record of postorogenic cooling at an integrated rate of approximately 1 °C/Ma, and subsequent variable Pb loss in the apatite and rutile systems during a Mesozoic thermal perturbation to the deep lithosphere. Closure of the rutile thermochronometer signals temperatures of 𙠂 °C in the lower crust during attainment of cratonic lithospheric conductive geotherms, and such closure in the examined portions of the "off-craton" Proterozoic domains of southern Africa indicates that their lithospheric thermal profiles were essentially cratonic from the Neoproterozoic through to the Late Jurassic. These results suggest similar lithospheric thickness and potential for diamond stability beneath both Proterozoic and Archean domains of southern Africa. Subsequent partial resetting of U-Pb rutile and apatite systematics in the cratonic margin lower crust records a transient Mesozoic thermal modification of the lithosphere, and modeling of the diffusive Pb loss from lower crustal rutile constrains the temperature and duration of Mesozoic heating to 𙡦 °C for ₞ ka. This result indicates that the thermal perturbation is not simply a kimberlite-related magmatic phenomenon, but is rather a more protracted manifestation of lithospheric heating, likely related to mantle upwelling and rifting of Gondwana during the Late Jurassic to Cretaceous. The manifestation of this thermal pulse in the lower crust is spatially and temporally correlated with anomalously elevated and/or kinked Cretaceous mantle paleogeotherms, and evidence for metasomatic modification in cratonic mantle peridotite suites. It is argued that most of the geographic differences in lithospheric thermal structure inferred from mantle xenolith thermobarometry are likewise due to the heterogeneous propagation of this broad upper mantle thermal anomaly. The differential manifestation of heating between cratonic margin and cratonic interior indicates the importance of advective heat transport along pre-existing lithosphere-scale discontinuities. Within this model, kimberlite magmatism was a similarly complex, space- and time-dependent response to Late Mesozoic lithospheric thermal perturbation.     

Subduction geodynamics in Archean and formation of diamond-bearing lithospheric keels and early continental crust of cratons

The diamond-bearing mantle keels underlying Archean cratons are a unique phenomenon of Early Precambrian geology. The common stable assemblage of the Archean TTG early continental crust and underlying subcontinental lithospheric mantle clearly shows their coupled tectogenesis, which was not repeated in younger geological epochs. One of the least studied aspects of this phenomenon is concerned with the eclogitic xenoliths carried up by kimberlite pipes together with mantle-derived nodules. The eclogitic xenoliths reveal evidence for their subduction-related origin, but the Archean crustal counterparts of such xenoliths remained unknown for a long time, and the question of their crustal source and relationships to the formation of early continental crust remained open. The Archean crustal eclogites recently found in the Belomorian Belt of the Baltic Shield are compared in this paper with eclogitic xenoliths from kimberlites in the context of the formation of both Archean subcontinental lithospheric mantle (SCLM) and early continental crust. The crustal eclogites from the Belomorian Belt are identical in mineral and chemical compositions to the eclogite nodules (group B), including their diamond-bearing varieties. The eclogite protoliths are comparable in composition with the primary melts of the Meso- and Neoarchean oceanic crust, which was formed at a potential temperature of the upper mantle which exceeded its present-day temperature by 150–250 K. The reconstructed pathways of the Archean oceanic crust plunging in the upper mantle suggest that the Archean mantle was hotter than in the modern convergence settings. The proposed geodynamic model assumes coupled formation of the Archean diamond-bearing SCLM and growth of early continental crust as a phenomenon related to the specific geodynamics of that time controlled by a higher terrestrial heat flow.