Introduction
Summary In the Carpathian–Pannonian region in Neogene times, westward-dipping subduction in a land-locked basin caused collision of two lithospheric blocks with the southeastern border of the European plate. Calc-alkaline and alkaline magmatism was closely related to subduction, rollback, collision and extension. Previous paleomagnetic results (e.g. Panaiotu, 1998) have shown that some volcanic area suffered important vertical axes rotations during the rollback. The purpose of our study was to identify if such rotations were present in the final stage of the collision at the south-eastern tip of the volcanic chain.
The Carpathian–Pannonian region of eastern central Europe is a key region for resolving the link between magmatism and tectonics. Miocene-Quaternary magmatic rocks display spatial and temporal geochemical variations, which may result from complex tectonic regimes: subduction, collision, postcollision and extension (e.g. Seghedi et al., 2004). Magmatic activity developed between 20 and<0.1 Ma and consists of (a) large volumes of calc-alkaline magmatic products (basalts, basaltic-andesites, andesites, dacites, rhyolites) and (b) sporadic small volumes of alkaline magma types (alkalic basalts, basanites, shoshonites, lamproites). In the Eastern Carpathians the volcanism was active from 12 to<1 Ma, migrating southeastward and progressively waning (Pecskay et al., 1995).
The last part of the volcanic chain is in the Harghita Mountains were the volcanic activity took place between 5 Ma and<1 Ma. Previous paleomagnetic results in the Eastern Carpathians were obtained from the northern part of the volcanic chain (Patrascu, 1993), the central part (Patrascu, 1976) and the south-western end (Panaiotu et al., 2004). In this paper we present the first paleomagnetic results from the southern part of the volcanic chain (the Harghita Mountains).
Sampling and methods We report results from eight sites sampled in andesites (5), basaltic andesites (2) and dacites (1). All samples were collected as cores drilled using a portable drill and oriented with magnetic and solar compasses. Up to three standard 25 x 22 mm cylinders were prepared from each core. Thin and polished sections were prepared from each site for optical mineralogical analysis. All remanence measurements were performed on a JR5 spiner magnetometer. Progressive demagnetization employed conventional alternating field and thermal methods. Demagnetization data were inspected using orthogonal demagnetization diagrams and individual magnetizations were identified as linear segments in both horizontal and vertical projections defined by three or more demagnetization steps.
Characteristic directions were determined using principal components analysis. All accepted linear segments have maximum angular deviation (MAD) values of less than 10°. The methods of Fisher, assuming circular distribution of individual magnetization directions about a true mean direction, were employed to estimate site-mean directions and associated statistics. Rock magnetic investigations involved acquisition of isothermal remanent magnetization (IRM) and backfield direct field demagnetization of IRM obtained in a peak field of 2.0 T, thermal demagnetization of a composite IRM acquired in two different fashions, and AF demagnetization of ARM, acquired in a DC field of 0.05 mT and a peak alternating field of 100 mT. In addition, we monitored lowfield susceptibility during progressive thermal.