Palaeohydrological mapping. The current resolution of Saharan palaeohydrology is too low to accurately determine the spatial relationships between available water resources, animal and plant distributions and archaeological sites. Hence, a critical aspect of this project will be to model with greater resolution the spatial and temporal extent of the AHP palaeohydrology using satellite imagery and digital topographic data (Figure 1). A recently developed method for rapid mapping of palaeolake and wetland features over large areas (Breeze et al., 2015), will be applied across the Sahara. These analyses combine multispectral classifications to identify surface palaeolake and wetland materials with GIS analyses employing geomorphological rules to identify locations where such features could form. These data will then be integrated with palaeodrainage mapping and palaeoecological and palaeoenvironmental data to produce spatially comprehensive high-resolution maps of hydrological features. We will compile a database of all dated lacustrine, fluvial and spring sites which can then be linked to the paleohydrology maps in order to determine spatiotemporal variations in Holocene climate and when the mapped features were active. These palaeohydrological features will be analysed to determine spatial relationships between surface water, biogeographic and archaeological information, the reconstructed demographic patterns, and the stable isotope results.
Figure 1. Preliminary hydrological reconstruction based on the Hydrosheds dataset with palaeoenvironmental proxies as recorded by Lezine et al. 2011 mapped.
Palaeoecological synthesis. Building on our existing 14C database (Manning and Timpson 2014), we will systematically record all published and archived zoogeographic distributions and archaeobotanical records. Mapping this data will allow us to examine spatiotemporal distributions of species, which represent key ecological markers. For example, deep vs. shallow water fish and molluscs or arid-adapted mammals, such as the Dorcas gazelle, can be mapped for different time periods alongside the palaeohydrology to identify spatial changes in hydrological activity and its relationship to ancient biogeography.
Compound-specific carbon (δ13C) and deuterium (δD) isotope analysis of lipid residues from pottery will be used as proxies to determine spatiotemporal variations in vegetation and humidity. Stable carbon isotopes (δ13C) provide a powerful environmental proxy, through vegetation driven signatures in the form of C3 versus C4 plants (Figure 2). We have also shown that compound-specific δD values of lipid biomarkers, such as fatty acids, are sensitive climate recorders (Outram et al. 2009), complementing plant lipids preserved in sediments, which have proved extremely useful in reconstructing past environmental conditions (see Demography and Human Adaptation page). The use of state-of-the-art gas chromatography-thermal conversion-isotope ratio mass spectrometry (GC-TC-IRMS) to obtain δD values of individual lipid compounds has demonstrated strong correlations with precipitation, temperature and humidity. In particular, it has been shown that an elevated δD signal is consistent with increasing aridity, allowing a compound-specific isotopic approach to record high-resolution climate signals. Our report on early dairying in the Sahara (Dunne et al. 2012) also noted the first identification of plant epicuticular waxes in North African ceramics, recognising the possibility for deriving δD values of the individual n-alkanes. These hydrologic signals will be correlated with dietary information obtained from the same vessels, in order to link changes in subsistence strategies and environmental change. Significantly, the δ13C and δD values will provide high-resolution records of local hydrological change in areas inhabited by Holocene groups, which will be compared to existing lower resolution regional signals obtained from 18O isotopes, providing important information on human adaptation to local, short-term climatic and hydrological change.
Figure 2. Plot of δ13C and Δ13C values of n-alkanoic acids in archaeological animal fat residues from mid to late Holocene Saharan vessels, showing the presence of both dairy and adipose fat residues and distinction between C3 and C4 diets reflecting contrasting ecosystem isoscapes between the three regions.