PETROLOGY OF THE EARLY PROTEROZOIC BURAKOVSKY LAYERED INTRUSION, SOUTHERN KARELIA, RUSSIA - MINERAL AND WHOLE-ROCK MAJOR-ELEMENT CHEMISTRY

The Early Proterozoic (2449 +/- 1.1 Ma) Burakovsky layered intrusion o f southern Karelia, Russia, is the largest (700 km(2)) mafic pluton in the Fennoscandian Shield. The layered series contains two main pans: (i) an ultramafic series (85% dunite), and (ii) a mafic series (mostly gabbros). Mineral and whole-rock chemical variations in the ultramafi c zone indicate little variation in chemistry over the entire sequence (e.g., in olivine, variations are from Fo(89) to Fo(87) over the uppe r 1600 m). In contrast, the Mg# in pyroxenes from the gabbronorite zon e varies from 85 to 58 over an interval of 500 m. On the basis of the current data, there are no apparent reversals in chemistry in the ultr amafic zone, whereas there is evidence for several in the gabbronorite zone. Furthermore, there is a distinctive reversal to more primitive compositions in mineral and whole-rock chemistry at the top of the ult ramafic zone. These differences-indicate that either (i) the gabbronor ite zone crystallized from the residue left from formation of the ultr amafic zone and the subsequent ''flushing'' of the more evolved portio ns to form the Early Proterozoic Karelian volcanic fields, or (ii) the re were two separate magmas. Plagioclase crystals from the gabbronorit e zone are chemically zoned, providing evidence of the local processes affecting crystallization of this sequence. Plagioclase crystals from near the base commonly display relatively Na-rich cores that are beli eved to result from either supersaturation at the onset of nucleation or the prior crystallization of some other, Ca-rich phase (e.g., clino pyroxene). Many plagioclase grains also contain reversely zoned rims, indicative of either (i) exchange between trapped liquid and the main magma during times of replenishment; or (ii) an increase in the CIPW-n ormative plagioclase-liquid partition coefficient for Ca, due to an in crease in the augite component of the melt. The lack of evidence for a bundant magma influxes in the ultramafic zone contrasts with the chemi cal reversals observed in the gabbronorite zone, suggesting that reple nishment was a prevailing process only in the upper half of the intrus ion.