Abstract:
We present a quantitative model for in situ crystallization within a solidification zone (or boundary layer) based on the trace and major element compositions of plutonic igneous rocks and related geochemical data. We developed the model to account for the characteristics of a suite of granitic rocks: nearly uniform mineral compositions in rocks that range widely in bulk composition (e.g., 58-76 wt.% SiO2); linear variation and correlation of all major and trace elements analyzed except Ba; large and apparently random variations in Ba concentration. These characteristics cannot be explained by any standard petrogenetic model involving fractionation, mixing, or restite unmixing, but are successfully reproduced by our model of residual melt migration. We do not attempt to model the entire crystallization history of a pluton but, rather, only that interval during which melt migration processes have recognizable geochemical effects. For the granitic suite, the chemical signature of residual melt migration resulted from the change in Ba compatibility with the onset of orthoclase crystallization at a granite minimum. Our results demonstrate that plutonic rocks can develop large compositional variations, comparable to those expected to result from extreme differentiation within a large magma body, over short distances by melt migration under conditions of high permeability or slow crystallization rate within the solidification zone. Melt migration is probably a common process that could be overlooked in large plutons if sampling is sparse and if variations in some components that appear random are not considered.Our model equations yield estimates for parameters that describe the proportions of residual melt that crystallize within and migrate out of the solidification zone. Values of these parameters can be used to infer information about permeability and melt mobility within the solidification zone. The model parameters derived for the granitic suite have a roughly concentric spatial pattern in the pluton, suggesting that residual melt was trapped near the margins, accumulated in the interior, with a zone of enhanced melt mobility and possibly compositional convection in between. Our model may be of general usefulness because it requires no assumptions about magma chamber geometry or magma dynamics, it is applicable to magmas of any composition, and the equations could be formulated to include those variables best constrained by a particular suite of plutonic igneous rocks.