Figure 7C clearly shows that mafic contributions to the finer fraction of the detritus were higher during MIS 2 than in MIS 1. The fact that the Cr concentration in MIS 1 is lower than in MIS 2 (Fig. 4D) implies that a weakened mafic contribution also applies to the coarser fraction.
Potential sources of detrital materials for Site 1017 are (1) the Central Californian Coastal Ranges, which comprise mainly the Franciscan Complex with its associated mafic and ultra-mafic rocks; and (2) the Southern Californian Transverse Range, which mainly consists of Mesozoic to Cenozoic sedimentary rocks and granodiorite (Norris and Webb, 1990; Fig. 1). These basic observations imply that input of both fine and coarse detritus from Central California to Site 1017 was relatively higher during MIS 2 and lower during MIS 1.
There are three possible explanations for these variations. The first is that strengthened flow of the California Current during MIS 2 could have transported more detritus southward from a northern Franciscan source. However, previous paleoceanographic studies along the western margin of North America between Oregon and northwest Mexico suggest that the southward-flowing California Current was weaker during glacial stages (Lyle et al., 1992; Dean et al., 1997; Doose et al., 1997; Ganeshram and Pedersen, 1998) as a result of a diminished atmospheric pressure gradient between the Pacific Ocean and the ice-covered northern North American landmass (Kutzbach et al, 1993). A second potential explanation is that there was increased runoff (higher precipitation) from Central California relative to the southern region of the state during MIS 2. Based on reconstruction of lake levels, Benson et al. (1995) claimed that the southward migration of the average position of jet stream during MIS 2 enhanced precipitation in northern California. This is consistent with the observation of Gardner et al. (1997) that the terrigenous mass accumulation rate off Central California was higher during MIS 2, and their inference that precipitation in Central California was enhanced at that time. Gardner et al. (1997) also pointed out that the main detrital source could have been the Santa Maria-Sisquoc River system, which runs through the Transverse Range. However, this would have enhanced the supply of felsic materials, contrary to the observed increase in mafic input during MIS 2. Therefore, enhanced precipitation is unlikely to have provided the major control on the provenance of sediments at Site 1017. The third explanation, which fits better with our data, is that southerly littoral currents during MIS 2 transported more Franciscan-derived mafic detritus southward. In support of this, Trask (1952) showed that the heavy mineral composition of sand on modern beaches along the coast of central to southern California changes systematically from north to south. He inferred that the heavy minerals largely originated from the north and were carried toward the south in the littoral drift. Increased storminess during MIS 2 associated with the southward migration of the jet stream and associated storm tracks (Benson et al., 1995; Kutzbach et al, 1993) would have strengthened the waves that arrive at the Central Californian coast, resulting in stronger littoral drift. Furthermore, sea level was as much as 120 m lower during MIS 2 and this may have had an effect by shifting the shoreline and the zone of littoral drift to what is today the outermost shelf, that is, closer to Site 1017. We speculate that such a migration could have promoted increased deposition of mafic minerals in the area of Hole 1017E during the last glacial stage.
If our hypotheses are valid, element compositions of Site 1017 sediments have been sensitive to past migrations of the jet stream and to storm and sea-level history on the coast of central and southern California. To confirm such basic controls on the composition of terrestrial detritus in the region, the patterns of fluctuations need to be studied over a broader area and further back in time.