Date of Award

1987

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Geology

First Advisor

S.E. Delong

Abstract

Little Sitkin island is an Aleutian calc-alkalic volcanic center that has erupted a suite of lavas ranging from andesite through rhyodacite. Whole-rock chemistry of these lavas indicates contrasting evolutionary processes; major-oxide silica variation diagrams exhibit linear trends that are suggestive of magma-mixing while trace-element trends are largely controlled by accessory-phase fractionation.
Plagioclase, the dominant phenocryst phase in all lavas, commonly occurs in two distinct populations with markedly different compositions and textures. Both normal and reverse zonation is noted in the plagioclase and clinopyroxene of several samples. In addition, clinopyroxene is found as rims on orthopyroxene grains and as cores with orthopyroxene rims in one sample. These inhomogeneities and indications of disequilibrium are supportive of mixing.
The phase chemistry of the Little Sitkin samples indicates that several andesites of intermediate composition formed as a result of mixing. There is considerable overlap of phase compositions in the intermediate andesites and other samples, however the most complete overlap occurs with a rhyodacite. This suggests that the intermediate andesites formed as a result of mixing between a silicic rhyodacite and a more primitive low-silica andesite. The andesitic endmember has not been sampled, although its bulk-rock major-element chemistry has been estimated.
Least-squares modelling of four whole-rock mixes shows close agreement between observed and calculated andesite compositions. The sum of the squares of the residuals for these calculations are all less than one. The best match is given by a mixing pair of rhyodacite and the estimated andesite composition, for which the sum of the squares of the residuals is approximately 0.02.
Comparison of the petrography and whole-rock chemistry of a sample previously described as a basalt suggests that addition of approximately 35 to 45 weight percent clinopyroxene to an andesite resulted in the observed composition. Least-squares analysis of the Little Sitkin samples indicates that incorporation of a smaller proportion of clinopyroxene, approximately 34 weight percent, plus lesser olivine, plagioclase and magnetite by an andesitic liquid formed the "basalt". These phases are possibly cumulate and were incorporated into the andesitic liquid when a fresh pulse of parental magma carried them up to a shallow crustal magma chamber from lower crustal depths. This indicates that andesite, not basalt, is the most primitive composition sampled on the island, and therefore the composition of the parental magma must be inferred. Amphibole compositions indicate that the most reasonable composition of the parent is that of high-alumina basalt (HAB).
A model is proposed in which partial melting of mantle peridotite yields an olivine tholeiite liquid. High-pressure fractionation of this liquid at the base of the crust produces the HAB parent magma. The HAB magma then undergoes low-pressure fractionation at upper crustal depths. This process, along with mixing between fresh HAB and more evolved compositions yields the suite of rocks found on the island. Late stage mixing of magmas intermediate between HAB and the silicic component is recorded by intermediate andesite compositions.

Comments

Wolf, D.A., 1987. Identification of endmembers for magma mixing in Little Sitkin Volcano, Alaska.
Unpublished MSc. thesis, State University of New York at Albany. 201 pp., +xiii
University at Albany Science Library call number: SCIENCE Oversize (*) QE 40 Z899 1987 W65

Share

COinS