Polycrystalline garnets are common in metamorphic rocks and may form as a result of close spacing of nuclei (if clustering is early) or impingement of larger grains (if clustering occurs later in the growth history). The timing of clustering relative to garnet growth is relevant to understanding the formation and evolution of porphyroblasts and evaluating the significance (if any) of clustering. Electron backscattered diffraction (EBSD) analysis of garnet-bearing metamorphic rocks reveals the presence of polycrystalline garnet in nine localities examined in this study: the northern Appalachians (Vermont, Maine, New York, USA); North American Cordillera (North Cascades Range, Washington; Snake Range, Nevada, USA); western Rocky Mountains (British Columbia, Canada); southern Menderes Massif (Turkey); Santander Massif (Colombia); and the Sanandaj-Sirjan zone (Hamadan, Iran). In some samples, polycrystals comprise ~20-30% of garnets analyzed, and chemical and textural evidence suggests that early coalescence of garnet polycrystals is common. Some early-coalescing polycrystals exhibit growth zoning that is concentric about the geometric center of the polycrystal. In thin section, these garnets may be undetectable as polycrystals based on morphology or zoning. In some polycrystals, zoning is unrelated to the location of internal grain boundaries; in others, Fe-Mn-Mg zoning has a different pattern than that of Ca; zoning patterns may vary on the scale of a single thin section. In addition, some polycrystals are characterized by high-angle misorientation boundaries that may be in special (non-random) orientations, an observation that indicates that these polycrystals are not random clusters of grains. The presence of internal grain boundaries may affect diffusion pathways and length scales, and may facilitate communication of porphyroblast interiors with matrix phases, thereby influencing reaction history of the rock and the composition/zoning of garnet.
Bibliographical noteFunding Information:
Acknowledgments This research was supported by NSF grant EAR 0510300. Parts of this work were carried out in the University of Minnesota, IT Characterization Facility, which receives partial support from NSF through the NNIN program. We thank Richard Spiess and an anonymous reviewer for helpful and insightful comments. Studies of the Menderes, Solitude Range, and Dutchess County garnets were carried out by Alice Alpert (Brown University) and Caitlin Anderson (Occidental College) as part of the NSF-REU summer 2008 internship program at the University of Minnesota. We thank John Rosenfeld for kindly providing samples from his classic study of curved inclusion trails in garnet from SE Vermont, and Aude Gébelin (Hannover) for providing the element maps of the Snake Range garnet.