Search

In the context of geoarchaeological investigations on Easter Island several hundred human-made pits filled with reddish silty material were discovered in fluvial terraces of two valleys on the southern slope of Maunga Terevaka, the highest volcano of the island. Micromorphological analysis of one representative pit filling and comparison of its geochemical and physical properties with sediments in the surrounding terrace was performed in order to reconstruct the probable formation and use of the material in the pits. A hypothesis of pigment production by heating of minerogenic iron-rich substrate with grass fuel resulting in formation of hematite is suggested. It is assumed that the pits represented the places for production and storage of the pigments, which were used by Rapa Nui for cultural and ritual purposes. The ongoing interdisciplinary research will enhance the interpretation of the pits and their fillings and contribute to a better understanding of cultural development on Easter Island.

Phases of formation and erosion of Holocene soils were reconstructed based on a sequence of buried soil horizons and colluvial layers at a lower slope section beneath the Neolithic earthwork of Salzmunde, Germany. The chronology is based on eight radiocarbon ages of charcoal, soil organic matter as well as on archaeological age estimations of embedded artefacts. Until ca. 2200 BC, a chernozemic A–C-soil developed in fluvially redeposited loess. The occurrence of charcoal, the darker colour, and a shift of soil matrix radiocarbon age indicate an alteration of the upper part of the humus horizon probably related to anthropogenic activities at ca. 2200 BC. This soil was probably buried rapidly by a colluvial layer (at ca. 2200 BC). Until ca. 900 AD, a phase of less intensive land use with slope stability followed. During this phase a pedogenetic alteration — probably under forest cover in the watershed and the deposition area — of the upper part of the loess, the buried soil, and the prehistoric sediment occurred (formation of chernozem cambisol). The upper part of a dated humic crotowina fill was involved in pedogenesis as well. Soil formation is indicated by changes in soil colour (brown), pedogenic iron, clay content, the mineral assemblage of the clay fraction, and magnetic susceptibility. During medieval times (ca. 900–1500 AD), two colluvial layers with a cumulative thickness of about 1 m buried this soil. The partial destruction of the archaeological remains at the plateau started at this time. During modern times (since 1500 AD), the deposition of two colluvial layers of a cumulative thickness of about 1 m completed the sequence of Holocene slope deposits. The results imply that the chernozem at the site might have been altered by prehistoric land use and altered further into a chernozem cambisol during a phase of slope stability of about 3000 years between the Bronze Age and medieval times. The detected brunification processes took place at unexpected high rates.

Some morphometric variables store information about past land surfaces longer than others. This property of morphometric variables is recognised as land surface memory. Slope deposits, soils, and vegetation also have this memory. In this study, a memory effect was quantitatively detected by Spearman correlations between thickness of colluvium and morphometric variables of the modern land surface. During long-term sedimentation, the sign of horizontal curvature (kh) may be inverted from minus to plus, suggesting that locations with positive kh values are not accumulation zones. However, the thickness of colluvial deposits at such locations in our study area indicates sediment accumulation. The sign of minimal curvature (kmin) tends to be more stable and remains negative. This difference provides the stronger correlation of colluvial layer thickness with kmin than with kh. The strongest correlation was found for total thickness of the colluvial deposits of the Neolithic and Iron Age with kmin (− 0.84); the correlation with kh was weaker (− 0.71).