No abstract

Carbonate deposits in Roman aqueducts of Patara and Aspendos (southern Turkey) were studied to analyse the nature of their regular layering. Optical microscopy and electron-backscattered diffraction results show an alternation of dense, coarsely crystalline, translucent laminae composed of bundles and fans of elongate calcite crystals with their c-axes parallel to the long axis, and porous, fine-grained laminae with crystals at near-random orientation. The d 18 O and δ 13 C data show a strong cyclicity and anti-correlation, whereby high and low δ 18 O values correspond to dense columnar and porous fine-grained laminae, respectively. Geochemical analyses show similar cyclic changes in carbonate composition. Electron microprobe and laser ablation inductively coupled mass spectrometry analyses show that porous finegrained laminae are enriched in elements associated with detrital material (Fe, Mg, K, Al and Si), whereas the dense columnar laminae are nearly pure calcite. The gradient in major and trace-element distribution, regular changes in crystal type and in oxygen and carbon-isotope composition from porous fine-grained to dense columnar laminae reflect changes in water chemistry, discharge, temperature and biological activity. Because of the strong bimodal cyclicity of the Mediterranean climate in southern Turkey, the observed laminae can be attributed to calcite deposition during the dry (porous finegrained) and wet season (dense columnar), respectively. This observation implies that, with proper geochemical and microstructural control, lamination in carbonate deposits in Roman aqueducts can be used for relative dating of aqueduct construction and maintenance and to obtain data on external factors that influenced the aqueducts, such as palaeoclimate and natural hazards. Carbonate deposits in Roman aqueducts show properties of both flowstone speleothems and riverine fresh water tufa. As many aqueducts of nearly identical channel geometry are present in different climate zones and with different source water characteristics, they can be used as natural experiment setups to test and improve existing models of how laminated fresh water carbonates record climate on time scales ranging from seasonal to millennial.

No abstract

The cause of intermediate-depth (50–300 km) seismicity in subduction zones is uncertain. It is typically attributed either to rock embrittlement associated with fluid pressurization, or to thermal runaway instabilities. Here we document glassy pseudotachylyte fault rocks—the products of frictional melting during coseismic faulting—in the Lanzo Massif ophiolite in the Italian Western Alps. These pseudotachylytes formed at subduction-zone depths of 60–70 km in poorly hydrated to dry oceanic gabbro and mantle peridotite. This rock suite is a fossil analogue to an oceanic lithospheric mantle that undergoes present-day subduction. The pseudotachylytes locally preserve high-pressure minerals that indicate an intermediate-depth seismic environment. These pseudotachylytes are important because they are hosted in a near-anhydrous lithosphere free of coeval ductile deformation, which excludes an origin by dehydration embrittlement or thermal runaway processes. Instead, our observations indicate that seismicity in cold subducting slabs can be explained by the release of differential stresses accumulated in strong dry metastable rocks

The study of a heterogeneous ductile shear zone that developed at ~ 500 °C and 0.2 GPa during post-magmatic cooling of a granodiorite has allowed the effect of strain and recrystallization on Ti re-equilibration of quartz to be assessed. Understanding this effect is critical for applying Ti-in-quartz thermobarometry to mylonites. Differently strained quartz across the shear zone shows a heterogeneous distribution of Ti concentrations ([Ti]) (measured by Secondary Ion Mass Spectrometry, SIMS) ranging between 2 and 45 ppm. Quartz cathodoluminescence (CL) is proven by spectral analysis to be correlated with [Ti], allowing CL images to be calibrated as Ti maps using SIMS measurements. Coarse-grained weakly deformed domains consist of magmatic quartz extensively recrystallized by grain boundary migration (GBM) and mostly (65–75% area) contain 20–38 ppm Ti. Resetting to lower [Ti] occurred locally: (i) in haloes surrounding titanite and biotite inclusions ([Ti] as low as 6 ppm); (ii) along grain boundaries; and (iii) towards the interface of quartz domains with other mineral domains. With increasing strain, quartz underwent progressive grain size reduction and developed a bimodal microstructure with elongate grains (> 100's μm long) surrounded by mantles of new grains (10–30 μm in size) recrystallized by subgrain rotation (SGR). Dynamic recrystallization by SGR, associated with prism < a > slip, became increasingly dominant over GBM as strain increased towards the shear zone core. Significant resetting of Ti in quartz only occurred in high strain domains (at shear strain γ probably >> 10) in the shear zone core where fine recrystallization amounts to 50–60% by area and coarser cores are strongly sub-structured. These domains are not compositionally homogeneous and still show a range of [Ti] mostly between 2 and 10 ppm. In all strain facies of the shear zone quartz-filled pressure shadows associated with feldspar show an almost constant [Ti] of ~ 2 ppm. The pristine Ti content of the magmatic quartz mylonitized in the shear zone core is therefore significantly reset and converges “asymptotically” towards the “equilibrium” 2 ppm [Ti] shown by new quartz precipitated in pressure shadows. It is inferred that extensive recrystallization by SGR and repeated cycles of dislocation creep and rearrangement provided fluid access to quartz grain interiors, promoting chemical buffering and leading to partial re-equilibration to low [Ti]. These observations imply limitations on the use of the Ti-in-quartz thermobarometry to constrain ambient conditions of ductile deformation