ROOF ROCK CONTAMINATION OF MAGMA ALONG THE TOP OF THE RESERVOIR FOR THE BISHOP-TUFF

The Bishop Tuff, a well known Quaternary high-silica rhyolite in east- central California, is widely considered the type example of a vertica lly and monotonically zoned pyroclastic deposit that represents zoning in the source magma reservoir, inverted during the process of pyrocla stic emplacement. However, the deposit of plinian pumice, which forms the base of the Bishop Tuff and represents the initial 10% or so of al l magma erupted during the event that produced the Bishop Tuff, contai ns features at odds with monotonic zoning for the reservoir, Relative to overlying ignimbrite, the plinian deposit contains a reversal in tr ace-element zoning. Moreover, the Sr-87/Sr-86 is significantly higher than that in overlying ignimbrite (about 0.7084 vs 0.7064), and melt i nclusions trapped in quartz phenocrysts exhibit notable variability of trace-element concentrations, even within a single host crystal (e.g. , U:10.77 to 8.91 ppm). These data have been previously interpreted as due to processes of chemical fractionation and evolution operating wi thin a magma system closed to chemical interactions with its roof rock s. For example, the reversal in trace-element zoning has been explaine d by the first-erupted magma being erupted from somewhat below the top of a monotonically zoned reservoir. However, we submit that the rever sed zoning and other above-noted features can be explained equally wel l as consequences of minor assimilation of roof rocks into a magma res ervoir that was erupted from the top down. The basal part of the Bisho p Tuff exhibits extreme concentrations and depletions of trace element s, relative to the average composition of crustal rocks. For example, the upward decrease of Sr in the Bishop magma reservoir (downward decr ease in the ignimbrite) results in concentrations as low as 2-4 ppm, B ecause of the attendant 'chemical leverage', assimilation of <1 wt.% o f Sierra Nevada bathorith racks typical of the area could readily reve rse an 'uncontaminated' Sr (and other trace elements) trend of zoning and could also substantially raise Sr-87/Sr-86. Small-scale trace-elem ent variability in the uppermost part of the Bishop magma reservoir, a s recorded by the above-mentioned melt inclusions, may simply reflect melt heterogeneity produced by the process of assimilation.