Micro-contextual identification of archaeological lipid biomarkers using resin-impregnated sediment slabs

Abstract

Characterizing organic matter preserved in archaeological sediment is crucial to behavioral and paleoenvironmental investigations. This task becomes particularly challenging when considering microstratigraphic complexity. Most of the current analytical methods rely on loose sediment samples lacking spatial and temporal resolution at a microstratigraphic scale, adding uncertainty to the results. Here, we explore the potential of targeted molecular and isotopic biomarker analysis on polyester resin-impregnated sediment slabs from archaeological micromorphology, a technique that provides microstratigraphic control. We performed gas chromatography-mass spectrometry (GC-MS) and gas chromatography-isotope ratio mass spectromety (GC-IRMS) analyses on a set of samples including drill dust from resin-impregnated experimental and archaeological samples, loose samples from the same locations and resin control samples to assess the degree of interference of polyester resin in the GC-MS and Carbon-IRMS signals of different lipid fractions (n-alkanes, aromatics, n-ketones, alcohols, fatty acids and other high polarity lipids). The results show that biomarkers within the n-alkane, aromatic, n-ketone, and alcohol fractions can be identified. Further work is needed to expand the range of identifiable lipid biomarkers. This study represents the first micro-contextual approach to archaeological lipid biomarkers and contributes to the advance of archaeological science by adding a new method to obtain behavioral or paleoenvironmental proxies.

Figure 1

Extracted Ion Chromatograms from samples AX2 BS, AX2 DD and R1. In the n-alkane fraction (F1) of R1 only one phthalate was characterized along with some n-alkanes. The AX2 LS and DD sample chromatograms share similarities among their n-alkanes profiles.

Figure 2

δ¹³C_16:0 and δ¹³C_18:0 values from Axlor, Crvena Stijena LS and DD samples, Cape Espenberg DD and resin control samples. (A) δ¹³C_16:0 before Suess effect correction; (B) δ¹³C_18:0 before Suess effect correction; (C) δ¹³C_16:0 after Suess effect correction and (D) δ¹³C_18:0 after Suess effect correction.

Figure 3

Scatter plots of ∆¹³C vs δ¹³C_16:0 (A) using modern animal samples as reference^, and δ¹³C_16:0/δ¹³C_18:0 ratios (B) and comparison with 95% confidence ellipses plotted using published data on isotopic ratios for C3 leaves and wood, terrestrial herbivores and carnivores and marine carnivores (C)^,,,,–,.

Figure 4

Example of drill dust collected from Salt-10-13 and Axlor slabs: (A) Flatbed scan image of a thin section from Salt-10-13 with area of interest outlined in red. (B) 1 cm-thick resin-impregnated slab of Salt-10-13, which is the mirror image of (A) and (C) Drill dust (0.2 g) obtained from the drilling process. (D) Flatbed scan image of thin section from Axlor-18–1 with areas of interest outlined in red, blue and green. (E) 1 cm-thick resin-impregnated slab of Axlor-18-1 with drill marks and (C) Axlor-18-1 block with the drilling area outlined in pink and resin control areas circled in blue.