The relationship of strain and preferred orientation of phyllosilicate grains in rocks—a review

Abstract

The calculation of strain from the preferred orientation of platy phyllosilicate grains, taking strain to be homogeneous, is surprisingly often successful from a pragmatic point of view, although numerous theoretical objections to this procedure can be and have been raised. For the calculation to succeed, the original distribution of mica or clay mineral grains must be known or assumed. The usual assumption that the starting distribution was random and therefore nearly uniform implies that the strain history of pelitic sediments is taken to start when the watery sediment was last stirred by waves or bioturbation, or if it was laid down in still, anoxic waters, from the moment of first deposition. It therefore includes compaction as part of the strain history, which is usually neglected by structural geologists. Originally random distribution of clay flake orientations has been observed in recently deposited sediments in many instances and has been theoretically explained in various ways. Exceptions probably exist, but their frequency has not been estimated. Biologically disturbed sediments may be brought nearer to a state of random orientation of clay flakes than they were at deposition. With the exception of whole beds or concretions in which pores are early filled with cement, mudrocks are almost always compacted by the weight of later deposits. The strain so induced is axially symmetric about the vertical and thus about the pole of the bedding planes. Compaction acts on soft, unconsolidated rocks but may continue with increasing overburden on rocks strengthened by cementation and by increasing interlocking of grains. Tectonic deformation may start early, before the rock is fully compacted, or it may begin only much later and the mechanisms of deformation increase in complexity with the increasing temperature and hydrostatic stress of deeper burial. Up to a certain level of mild metamorphism, however, phyllosilicate grains seem to be rather permanent, if not as mineral species then at least by preservation of their constituent sheets of silica tetrahedra. Given this permanence, the mechanisms of rock deformation, whether intergranular slip, cataclasis, or dissolution and reprecipitation and various intragranular processes in the non-phyllosilicate matrix hardly matter as far as the recording of strain is concerned. Even a certain amount of stress solution at especially heavily loaded points of the phyllosilicate grains themselves does not destroy their value as strain markers, as long as the dissolved material is disposed of mainly by crystallographically continuous overgrowth on existing grains. The same holds for local dissolution of phyllosilicates in portions of old grains that have become thermodynamically unstable by bending. Widespread irregular or random nucleation and growth of newly nucleated phyllosilicate grains at intermediate or high grades of metamorphism can reduce the reliability of the strain record, and so can domainal structure of the deformed rock if both the volume fraction of phyllosilicate grains and the intensity of total strain differ sharply from one domain to another, if distinct domains are not equant and have themselves a strong preferred orientation of shape. Nevertheless, rocks containing an anastomosing network of zones rich in phyllosilicate grains, as slates and other deformed mudrocks commonly do, have in many cases been found to have average preferred orientations of mica or chlorite which, if used to estimate the strain, give results similar to those obtained from independent strain markers in the same rocks. Thus the most significant verification of the usefulness of strain estimates by measurement of preferred orientation is the mutual confirmation with independently determined strain. The limit to such verification is the rarity of strain markers, so that the coexistence of two independent ones is exceptionally rare. For this reason, however, it is important to make full use of the orientation of phyllosilicate grains where they are available. The alternative, to have no strain measurements at all, causes structural geologists to miss potential information.