WHY DOES THE LOESS HAVE STONES IN IT? AN INVESTIGATION OF LOESS DEPOSITS IN JIUZHAIGOU NATIONAL PARK, SICHUAN, CHINA

Located between the China Loess Plateau and the Chengdu Basin, Jiuzhaigou National Park contains isolated fine sand and silt deposits of unknown origin. The deposits, identified on geologic maps as Pleistocene loess, average 2-3 m but are up to 7 m thick locally. A paleosol in the midst of the loess contains archaeological artifacts. Usually sediment above the paleosol contains matrix supported clasts; whether or not the paleosol is present, stone free sediment is always stratigraphically below stony sediment. The mechanism which incorporated stones into the loess is unknown. Two hypotheses have been proposed: 1) the deposits with stones are slump-earthflow deposits that post-date deposition of all loess; 2) loess was deposited continuously and colluviation during deposition mixed stones in.

Loess below the paleosol is late Pleistocene (15-16 ka; 2 OSL dates at 1 site; Henck et al. 2010). Radiocarbon dates at 5 sites and artifact typology date the paleosol to between 1500 and 4000 ¹⁴C years (Henck et al. 2010; Huang pers. comm.; Lu et al. 2010). 8 radiocarbon ages at 1 site increase linearly with depth from 900 to 4000 ¹⁴C years (Huang pers. comm.) and support the hypothesis that material was continuously deposited and has not significantly remobilized. Grain size distributions of sediment <2 mm from 3 sites, both stony and stone free, further support this hypothesis: grain size gradually transitions up section from coarse silt to a bimodal distribution of coarse and very fine silt. Stony samples are 1.2 to 26.0% stones >2 mm.

Thin section analyses show that samples below and above the paleosol (stony and stone free) are primarily composed of silt-sized quartz and contain shell fragments; stony samples also contain local low-grade metamorphic bedrock clasts and have a blocky structure; stone free samples have no structure. In all samples the average size of the largest 10 quartz and feldspar grains in the matrix is very fine sand. The contrast in structure and composition of the stony and stone-free material supports the hypothesis of an abrupt transition between these layers. More detailed field and analytical work is needed to constrain the provenance of the sediments, mechanism by which the rocks are mixed into it, and how the interplay of pre-historic anthropogenic activity and hillslope processes created the current landscape.