Stress fields in central portions of the Pacific Plate: Delineated in time by linear volcanic chains

Abstract

Many linear island chains in the Pacific appear to consist of individual volcanic shields that lie on relatively short, curved loci sometimes of sigmoidal form. These loci, in turn, lie at an angle to the directions of propagation of the chains. Those chains that now trend just south of east (the Pratt-Welker, Hawaiian, Tuamotu, and Austral chains) consist of en echelon sets of loci with right lateral (clockwise) sense of stepwise overlap, whereas the supposed older extensions of these chains that lie nearly north-south (the Emperor and Ellice-Gilbert-Marshall chains) consist of sets of loci with left lateral (counterclockwise) sense of stepwise overlap. Rift zones of isolated volcanoes in the chains, that is, those volcanoes whose local stress fields were not influenced by the buttressing effects of near-neighbor volcanoes at the time of their formation, show the same orientation as the loci. First-motion studies of earthquakes in the Pacific lithosphere to date are few in number and give a variety of solutions with inferred directions of maximum principal compression (P) nearly horizontal and trending in both northeasterly-southwesterly and northwesterly-southeasterly directions. However, the number of such solutions is small, in some cases related to edifice effects, and interpretations relating compression and dilatation to principal stress directions in magmatic source regions are open to question. We consider it likely that the dominant orientation patterns have been due to dynamic effects related to the overall kinematic patterns of Pacific plate motions. Although a large number of different factors can influence the inherent and transmitted stresses in the lithosphere, and thereby influence the locally dominant stress field, we conclude that the effective stress orientations in the very recent past and for about the last 40–50 m.y. can be considered to have been caused by dynamic forces reflected in a right lateral rotational couple acting within the plane of the Pacific plate. These forces have induced maximum (S1) principal compressional stress directions that have had surface traces trending about northwest-southeast, relating to minimum (S3) principal stress directions that have trended northeast-southwest. Prior to 40–50 m.y. ago the dominant stress orientations in the plate were caused by a tendency for left lateral stress rotations, which induced maximum (S1) principal compressional stress directions with surface traces that trended just west of north and minimum (S3) principal stress directions that trended just north of east. These data and interpretations are independent of whether or not a change in direction of motion of the Pacific lithospheric plate took place roughly at the time represented by the bends in the island chains and are also independent of arguments as to whether melting anomalies of the Pacific are rigidly fixed or whether they are better explained in terms of thermal plumes or gravitational anchors. We conclude that the trends and age correlations of volcanic loci in the Pacific accurately track and identify the evolution of states of stress in the Pacific lithosphere with time. As more age data for linear island edifices become available, it should be possible to construct a Pacific-wide chronology of volcanism independent of but similar to that developed from magnetic reversals. Further, it appears that changes in stress directions in the plate have been episodic and that they may correlate with episodic magma generation not only in the central part of the Pacific plate but also around its margins.