CONGLOMERATE OR DEFORMED AND METASOMATIZED TONALITE? GEOCHEMICAL EVIDENCE ON THE ORIGIN OF THE “BOULDER BED,” ISUA GREENSTONE BELT, GREENLAND

One of the more widely discussed lithologies in the Isua Greenstone Belt is a unit best exposed along the shore of Lake 978. This debated “boulder bed” exposure has been illustrated many times beginning with early investigations of the supracrustal rocks. The unit consists of cm-scale highly flattened ellipsoids (“clasts”), whose long dimension is parallel with the regional down-dip lineation, set in a matrix of more easily weathered, dark, fine grained rock (“matrix”). This textural relationship is suggestive of a sedimentary or volcaniclastic origin for the unit. We present geochemical analyses of three “clast” samples and three “matrix”samples. The clasts and matrix have broadly similar highly fractionated REE patterns typical of the Archean TTG suite (LaN/YbN = 22-29). All samples have highly fractionated LREE (LaN/SmN = 4.5), a small or absent Eu anomaly, and less fractionated HREE (GdN/YbN = 3.5). Absolute abundances and pattern shapes of the REE are similar within each group, but the clasts have a consistently higher sum REE, as well as a small negative, but persistent Eu anomaly (Eu/Eu* = 0.76). These differences are subtle but indicate there is a compositional variation between the two groups. This difference is best recognized when overlaying representative samples from each group that have the same value at LuN. Matrix samples become increasingly depleted in REE towards LaN. Mixing modeling shows that addition of as little as 0.01% monazite to matrix material produces a REE pattern that is nearly identical with that of the clasts; we postulate that some matrix constituents were lost during mylonitization and metasomatism of matrix material. Consequently, the data suggest that the matrix and the clasts are genetically related, which is reinforced by the Pb isotopic compositions of these samples, which are similar to Amîtsoq gneiss values. These geochemical and isotopic findings are consistent with the boulder bed actually being a severely deformed sheet of Amîtsoq gneiss that has an intrusive relationship to the Isua supracrustal succession. In this scenario, the clasts would represent relatively less deformed tonalitic relics floating in a strongly mylonitized and metasomatized matrix. An altered igneous origin for this lithology removes it from the list of potential candidates suitable for examining Earth's earliest surface conditions.