Terrace Archaeology

and Culture in Europe


Identify past crops, biodiversity and management...

The land use of terraces is often inferred rather than scientifically determined, so e.g. if large quantities of olive oil are being produced by a civilisation and even recorded in writing then it might reasonably be assumed the terraces are under olive groves. Whilst this approach is valid it does not tell us much about the actual use or multiple use of specific terrace systems. Traditional methods such as pollen and plant macrofossils are generally inappropriate due to the oxidising environment. However, phytoliths are often preserved. This can often be related to regional culture such as the expansion in the Sainte Victoire area of oil production in the Roman period (Walsh 2014). This links terrace construction and use with regional cultural developments. Phytoliths are silica microremains formed in living plants and which reproduce their cellular tissue. After the decay of plants, phytoliths become incorporated in soils and sediments as microscopic particles (Piperno 1988). Because phytoliths are resistant to most post-depositional processes they can survive over long periods of time and can be found in contexts over several million years old (Stromberg 2004). Although phytoliths have been used to identify agricultural practices (Rosen 1994; Ball et al. 1999), few studies have been conducted on terrace soils. However, one example was recently carried out by Albert (Boixadera et al. 2016) and the results showed high potential for phytolith analyses since they allowed not only differentiation between cultivars in time, but also the identification of terrace construction processes. Selection and preparation of soil samples for phytolith analysis will be carried out at the Department of History and Archaeology, University of Barcelona. The methods proposed follow Albert et al. (1999) and Katz et al. (2010) and morphological identification will be based on standard literature (Piperno, 1988, 2006; Rosen, 1994; Twiss, 1992; Mulholland & Rapp, 1992), as well as on modern plant reference collections (Albert & Weiner 2001; Tsartsidou et al. 2007; Albert et al. 2016- see www.phytcore.com). Morphometric analysis of phytoliths will be used for differentiating between phytoliths produced by closely related taxa (Ball et al. 1999; Berlin et al. 2003). In addition to the microscopic observation, digital images of phytoliths will be taken to prepare the image catalogue that will be added to the database. Infrared spectroscopy (FT-IR) will be used to identify the mineral components of the sediments. FT-IR analysis provides information on the minerals that constitute the bulk of the sample (Weiner & Goldberg 1990) and helps identify postdepositional processes. This will be assisted by the use of soil micromorphology at each site, which can not only document soil use history but also weathering state so feeding back into WP2.    


Whilst this is a well-developed technique TerrACE will also seek to apply a novel technique that of aDNA preserved in the soil in terrace profiles. Although soil DNA was first used to investigate extreme environments such as Norse farming in Greenland (Hebsgaard et al. 2009) and palaeoecosystems in the tundra (Willerslev et al. 2003; Willerslev et al. 2014) it can also be preserved in warmer soil conditions (Yoccoz et al. 2012; Bremond et al. 2017). This approach has been shown to track the variation of the abundance of plants and domestic animals over millennia, enabling reconstruction of human impacts through time from lakes in cold or temperate environments (Giguet Covex 2014 and recently by the PI in Alsos et al. 2016). The use of this technique will be experimental but aDNA has been successfully extracted from a soil near the French study area (Yoccoz et al. 2012) and lacustrine sediments in Norway (Parducci et al. 2012, Paus et al. 2015, Alsos et al. in prep) and this is one of the reasons for including an Alpine and two northerly sites in TerrACE. Notably, for both the Alps and Norway, a full genome reference library is under production by the Grenoble and Tromsø lab, respectively (www.NORBOL.org). Tromsø will analyse the samples from Norway and England, and Grenoble the Southern sites. To identify vascular plants we will use the g and h universal plant primers amplifying the short and variable P6 loop region of the chloroplast trnL (UAA) intron (Taberlet et al. 2007), whereas mammals will be identified using the MamP007R primer amplifying a fragment of the mitochondrial 16S gene (Giguet-Covex et al. 2014). To maximize the detection of taxa and allowing semi-quantification of taxa, we will run eight PCR analyses on each sample (Ficetola et al. 2015). Negative controls will be analysed along with the samples to detect any contamination (Parducci et al. 2017). As cultivated plants and animals are common contaminants in aDNA analyses (Thomson & Willerslev 2015), authenticity will be investigated by shotgun sequencing and subsequent analyses of damage pattern (Weiss et al. 2015). Analysis of aDNA will be performed in two stages with an initial sample tested for DNA quality and quantity from each site, then where results are good, more samples will be analysed from the buried profile. As DNA in soil may mainly represent the species growing within 1 m from the sampling points (Edwards et al. in prep), 20-30 samples will be analysed from each landscape. Including negative controls, 96 samples will be analysed at both Grenoble and Tromsø