Sie sind hier: Startseite Bonner Bodenk. Abh. Band 61 (2014) - Band 70 (2017) Band 67 (2017) Abstract Band 67

Abstract Band 67

Prost, Katharina (2017): Steroid biomarkers – new insights for detecting and identifying faecal inputs in archaeological soil material.  


Ancient land-use systems and settlement structures may be reconstructed by the investigation of preserved, humus-rich soil, which is presumed to be former topsoil material influenced by animal and plant remains. I hypothesized that former topsoil material, altered by the input of faeces, contains chemical signatures of steroid biomarkers, which might not only be used to detect faecal remains but which might even give information on its source. This implies, however, that faecal steroid patterns are source-specific and that they are neither altered in the course of a treatment of faeces (e.g. by composting) nor after burying the faecal material in soil. My objectives were, therefore, i) to assess, if steroids are appropriate markers to identify an input of different livestock faeces in archaeological soil material, ii) to evaluate their stability in course of composting, and iii) to use the information gained to trace residues of manuring with faeces in Neolithic arable topsoil. 

Previous studies on steroid analysis often failed to indicate the presence of faeces in the environment when using established threshold values. Moreover, steroid analysis could only identify faecal remains of humans, pigs and the group of herbivores. To overcome these shortcomings, I included for the first time all common biomarker steroids (Δ5-sterols, 5α-stanols, 5β-stanols, epi-5β-stanols, stanones, and bile acids) in one analysis. To achieve my objectives, I i) analysed the steroid composition of faeces from old livestock breed (cattle, horse, donkey, sheep, goat, goose, and pig) as well as humans and complemented the results with a literature review on faecal steroids. Furthermore, I ii) investigated the change of steroid patterns of horse and cattle manure in two 168-day lasting composting trials, and iii) traced identified steroid signatures of livestock and humans in preserved archaeological soil material of different age (Early Neolithic, Middle and Younger Neolithic, Bronze Age, Iron Age, and Roman Age). In addition to steroid biomarkers, a multitude of different analyses - i.e. phosphorus (P), black carbon (BC), amino acid, 15N isotopy, micronutrients, and soil colour measurements - complemented the analyses to identify a possible manuring of ancient Neolithic fields. 

By combining the analysis of a multitude of steroids and by using both already established and newly developed steroid ratios, I was able to detect and differentiate each kind of human and livestock faeces considered (with the exception of sheep and cattle). Most remarkable was the identification of horse faeces (via enhanced values for the sum of epi-5β-stigmastanol : 5β-stigmastanol + epicopro-stanol : coprostanol, together with the presence of chenodeoxycholic acid, CDCA) and the successful discrimination between goat (with CDCA) and sheep faeces (without CDCA). As the identified steroid signatures were preserved in soil, they confirmed archaeological find categories of a Roman horse stable, a latrine, and a sewer ditch (dating to 0-450 AD). Furthermore, even much older archaeological finds, like fillings of a water hole and two wells (dating from 5300-389 BC) showed steroid contents large enough to allow the identification of faecal inputs. All inputs identified led to conclusions that were consistent with the archaeological context and the livestock composition for each time period.

However, during composting of cattle manure, steroids were less persistent than in soil. All steroids dissipated nearly completely after 56 days. After 168 days, losses were particularly pronounced for bile acids and 5β-stanols (98.6 - 99.9 % and 99.0 -99.6 % relative to initial contents, respectively), which challenges the assumption that bile acids are per se more resistant towards degradation than other steroids. In contrast, 5α-stanols and epi-5β-stanol contents declined less rapidly and to a lesser extent than those of 5β-stanols and bile acids. As a result, composted and fresh farmyard manure could be distinguished by the sum of epi-5β-stanol to 5β-stanol ratios. Furthermore, by means of this sum of ratios an input of composted faeces into soil could still be identified as faecal material. 

In contrast to the rapid dissipation of bile acids in composting material, for Neolithic topsoil material only bile acids but no faecal stanols and stanones could be detected, confirming the assumed larger persistence of bile acids in the soil mineral matrix. In comparison to reference soils, these enhanced contents of bile acids together with larger contents of BC (38% of soil organic carbon) and plant nu-trients (15N and P) indicated a fertilization of the arable topsoil with manure containing faeces and ash.

In conclusion, by combining the analysis of 5β-stanols, 5α-stanols, epi-5β-stanols, and bile acids, most livestock faeces as well as composted farmyard manure could be identified by their steroid patterns. Moreover, the identified steroid signatures of livestock and human faeces could be well detected in archaeological soil material, proving supposed faecal inputs. Particularly bile acid analysis showed promising results, as bile acids indicated even in up to 7000 year old soil samples a manuring with faeces and thus amelioration measures of ancient topsoil. However, whether a manuring in Neolithic times was done intentionally remains unknown. Furthermore, absolute amounts of applied manure cannot be reconstructed, due to an unknown but - under composting conditions - potentially rapid dissipation of steroids in manure prior to its addition into soil.