## Description

Weathering, erosion, and sorting during transportation in the fluvial regime play important integrated roles in the development of quartzofeldspathic sediment. For example, H. W. Nesbitt and colleagues showed that sorting of eroded regolithic material produces a mud-rich component that is more weathered than sand derived from the same source. In order to test this idea, I collected 5 corestone samples and 5 saprock samples from a single site located east of the Clark segment of the San Jacinto fault, SE of Anza, California. Each of the five saprock samples was split into two parts. One part was used as the bulk sample, while the other part was sieved into the following size fractions: >63 microns, 63-45 microns, and <45 microns. Utilizing the XRF facilities at SDSU each sample was analyzed for its major element composition. The corestone is an equigranular hypidiomorphic granodiorite to quartz monzodiorite. In p(A)-p(CN)-p(K) space chemical data derived from corestone and bulk saprock samples spread about a linear trend with saprock samples plotting to the left of corestone samples. Non-central principal component analysis and the linear compositional modeling techniques of H. von Eynatten and colleagues, show that PC1 explains 96.4% of the variability in the calculated linear trend. Using orthogonal projection, weathering intensity factors were calculated for the bulk saprock samples. They ranged from 0.04 to 0.14. Utilizing Al as a reference frame, mass balance calculations indicated that 11% (+6%/-7%) of the mass of K and 15% (± 4%) of the mass of P was lost during the conversion of corestone to bulk saprock. Such losses reflect the alteration of biotite and apatite respectively. In addition, a nominally small increase of Si mass (5% ± 3%) is evident. Thus, both the weathering intensity factors and the mass balance calculations indicate that the bulk saprock samples are poorly weathered, and that the loss of K mass from biotite is a major controlling factor. If erosion removed the regolith at the study site, and fluvial processes size fractionated the resulting sediment, then the >63 micron fraction would be sorted into sand and pebble sized material, while the 63-45 micron fraction would represent coarse silt. The <45 micron fraction would include medium and fine silt along with clay-sized materials. On a p(A)-p(CN)-p(K) ternary diagram, the sieved fractions plot about a linear compositional trend that is similar to that derived solely from the corestone and bulk saprock samples. Moreover, the >63 micron fractions plot at the lower end of this trend while the finer grained fractions plot about the trend but nearer the p(A)-p(CN) join. Weathering intensity factors for the two 45-63 micron fractions are 0.86 and 0.95 while weathering intensity factors for the three <45 micron fractions are 0.85, 0.97, and 0.95. These results contrast markedly with those provided above for the bulk saprock samples, and clearly indicate that the sieved finer grained fractions captured a part of the weathering trend that is masked by the bulk saprock samples. In short, the results of this study document for the first time, a trend derived from the weathering of granodiorite to quartz monzodiorite that is controlled primarily by the loss of K from biotite. This trend has not been previously recognized but given that the composition of the corestone studied during this investigation is relatively common in eroded continental margin magmatic arcs on a global basis it may be more widespread than currently recognized. Clearly more work is warranted.