TESTING COMPOSITE DIKE INTRUSION MODELS WITH CHEMICAL ANALYSIS

Recent work on dike swarms indicates that many single dikes are composed of multiple injections. For example, in the 148 Ma Independence dike swarm (IDS) of eastern California, a typical one meter dike is composite and composed of roughly 10 dikelets. Whether each dikelet represents a separate injection of magma depends on the mode of injection. There are two end-member models for dike intrusion—antitaxial and syntaxial. In antitaxial intrusion, each dikelet injects the dike-wall interface, perhaps because it is mechanically weaker than the dike or wall rock. In syntaxial intrusion, new injections split preexisting dikes, as in the classic sheeted dike model for mid-ocean ridges. In this case all dikelets but the last are mirror images of another. This model requires that subsequent injections fracture the interior of the previous intrusion. By field examination of composite dikes and chemical analyses of dikelets, it should be possible to determine if one model is more widely represented than the other. To test this hypothesis, we collected suites of samples from several composite dikes in the east-central Sierra Nevada.

Sampling locations included dikes along the south fork of Big Pine Creek and along the John Muir Trail below Mt. Cedric Wright. We made strip maps of dikes perpendicular to strike and collected samples for XRF analysis. Strip maps and field observations indicate that strong mirror symmetry across dike centers is common. For antitaxial injection, the number of injections (n) is equal to the number of dikelets (D); for syntaxial injection, D=2n-1. Most of the composite dikes we observed are composed of an odd number of dikelets consistent with syntaxial intrusion. Chemical data from suites of samples collected near Big Pine Creek show chemical variation of up to 10 wt% silica along with strikingly symmetrical compositional profiles of several elements from margin to margin.

Both field relations and major-oxide data demonstrate symmetry consistent with syntaxial intrusion, contrary to expectations that subsequent magmatic intrusions would exploit the theoretically weaker contacts between adjacent intrusive bodies or the surrounding wall rock.