Scotland’s first farmers: new insights into early farming practices in North-west Europe

Thirty years after the discovery of an Early Neolithic timber hall at Balbridie in Scotland was reported in Antiquity, new analysis of the site's archaeobotanical assemblage, featuring 20 000 cereal grains preserved when the building burnt down in the early fourth millennium BC, provide new insights into early farming practices. The results of stable isotope analyses of cereals from Balbridie, alongside archaeobotanical and stable isotope results from three other sites, indicate that while cereals were successfully cultivated in well-established plots without manuring at Balbridie, a variety of manuring strategies was implemented at the other sites. These differences reinforce the picture of variability in cultivation practices across Neolithic North-west Europe.


OSM1: stable isotope analysis
The stable isotope composition of archaeological crop remains provides a direct method of investigating crop growing conditions in the past, potentially allowing variation within and between harvests to be identified (Bogaard et al. 2013). The application of organic matter (manure, household midden, seaweed or fish remains) to crops to increase yields, may significantly increase nitrogen stable isotope values in the soil (δ 15 N: the ratio of the stable isotopes of nitrogen, 15 N to 14 N), increasing crop stable isotope values up to c. +10‰ above wild plant values (Fraser et al. 2011;Bogaard et al. 2013;Blanz et al. 2019;Gröcke et al. 2021). Parameters for identifying the intensity of manuring using crop δ 15 N values have been established from modern field trials and crops grown under traditional cultivation regimes, with higher δ 15 N values expected for fields with high (30+ tonnes per hectare) compared to medium (<20 tonnes per hectare) or low (no manuring in the last three or more years) manuring levels (Bogaard et al. 2013;Styring et al. 2017a). Several other factors may also increase δ 15 N values in plants to some extent, such as waterlogging, burning or naturally high soil nitrogen levels, whereas salinity may decrease δ 15 N values in plants (Szpak 2014;2 Fiorentino et al. 2015: 222). Further research is needed, but a recent study suggests that cereals grown on recently burned land have low δ 15 N values (Styring et al. 2017b), implying that manuring and burning can be distinguished.
Thresholds have also been established for identifying levels of water availability for past crops using carbon isotope discrimination (∆ 13 C) values from modern crops grown under different irrigation regimes (Ferrio et al. 2005;Wallace et al. 2013). ∆ 13 C values are calculated from carbon stable isotope values (δ 13 C: the ratio of the stable isotopes of carbon, 13 C to 12 C), and provide a measure of how much plants avoid using-or discriminate against-13 C, allowing for changes in δ 13 C in the air through time (Fiorentino et al. 2015).
Plants preferentially utilise 12 C for photosynthesis, but when carbon dioxide availability is restricted (e.g. in arid conditions when stomata of plant cell walls are closed to restrict water loss) or the reaction rate is increased, insufficient 12 C is available and there is less discrimination against 13 C (Fiorentino et al. 2015). Thus, higher δ 13 C values (corresponds to lower ∆ 13 C values) reflect drier growing conditions and vice versa (Wallace et al. 2013).
Other factors which increase photosynthesis rates or reduce water availability can also result in higher δ 13 C values (i.e. lower ∆ 13 C values) in crops, such as lower soil water-holding capacities or increased temperatures or light intensities (Heaton 1999). An increase in soil salinity will also cause plants to reduce their stomata apertures, resulting in less discrimination against 13 C and higher δ 13 C values (i.e. lower ∆ 13 C values; Gröcke 1998Gröcke , 2002.

OSM2: stable isotope sample selection and analysis
A total of 196 Neolithic cereal grains and 10 crab apple seeds were selected for stable isotope analysis. We selected key contexts from each site for analysis on the basis that they contained concentrations of well-preserved cereal grains of one or more crop species. The degree of cereal grain preservation was determined using the preservation scale of Boardman & Jones (1990) and only grains in the P1-P3 range were selected for analysis to minimise isotopic offsets caused by charring at high temperatures. An adapted scale was used to record the preservation of the crab apple seeds, based on the degree of seed coat surface covering (see Table S6). Overall, 76.5 per cent of the grains and 100 per cent of the crab apple seeds were very well preserved, falling in the P1 and P2 preservation categories, with a further 23.5 per cent of the grains falling in the P3 preservation category. Where more than one crop species was present in the selected context, we analysed grains from both species to examine any differences in crop husbandry practices between species. Where possible, 10 grains or seeds 3 were analysed from each selected context, though in some cases a smaller number were selected due to the low numbers of grains present and the need to preserve grains for future analyses.
Each grain and seed was individually analysed for stable carbon (δ 13 C) and nitrogen (δ 15 N) to examine the level of variability within contexts, whether there was any relationship between the size of the grain and the stable isotope content, and to reduce the number of archaeological grains destroyed as part of the analyses. Prior to analysis the length, width and depth of each grain and seed was measured to the nearest 0.1mm using the internal graticule of a Leica M80 stereomicroscope under × 7.5-60 magnification and the mass of each grain was measured to the nearest 0.001g using a Mettler PM480 Delta Range balance. Grains with adhering sediment were avoided for analysis, but where necessary grains were carefully scraped with a clean scalpel to remove any adhering sediment to ensure reliable δ 13 C and δ 15 N results were produced (Brinkkemper et al. 2018). Prior to analysis, each grain was crushed using a pestle and mortar. No further pre-treatment of grains was undertaken.
Stable isotope analyses, total organic carbon and total nitrogen content were determined using a Costech Elemental Analyser (ECS 4010) connected to a Thermo Scientific Delta V Advantage isotope ratio mass spectrometer in the Stable Isotope Biochemistry Laboratory N between −7.5‰ and +20.4‰. Analytical variation in carbon and nitrogen isotope analyses was typically ±0.1‰ for replicate analysis of the international standards and typically <0.2‰ on replicate sample analysis. BOH S.41 IS5 and IS6 produced unusually high δ 15 N values (14.19‰ and 9.05‰ respectively before adjusting for charring) and therefore these samples were subsequently reanalysed and a consistent result was produced (14.02‰ and 9.06‰, respectively, before adjusting for charring). Total organic carbon and nitrogen data was obtained as part of the isotopic analysis using an internal standard (Glutamic Acid, 40.82% C, 9.52% N). Stable isotope results were adjusted for potential charring offsets by subtracting 0.11‰ from δ 13 C and 0.31‰ from δ 15 N (Nitsch et al. 2015). δ 13 C values of atmospheric CO2 for each context were established using the radiocarbon dates for each site and the AIRCO2_LOESS system (Ferrio et al. 2005). ∆ 13 C values were calculated following the methodology of Farquhar et al. (1989).
+1‰ was added to the crop δ 15 N values from Bogaard et al. (2013) because the raw values in the publication had been adjusted by subtracting one to allow for charring, and it has subsequently been shown that it is only necessary to subtract 0.31‰ to correct for charring (Nitsch et al. 2015).

OSM3: archaeobotanical methods
The cereal assemblage composition charts in Figure 3 include data for the contexts analysed in the stable isotope study from Dubton Farm (Contexts B215, B037/1, B233/1, B187;  Table S1). The aim was to provide an impression of the overall ecological character of all the wild seed taxa for each site and hence the data were considered on an assemblage rather than sample basis and thresholds for numbers of identifications per sample were not set. The crop compositional data, seed ecological data and context associations for the analysed assemblages from Balbridie and Dubton Farm suggest that the wild seed taxa are 'weed seeds' associated with the arable crops (Fairweather & Ralston 1997;Church 2002). In contrast, the higher proportion of freshwater/heathland taxa and wet ground indicators (e.g. Carex sp.) in the Braes of Ha'Breck assemblage suggests that a mix of 'weed seeds' and seeds deriving from the burning of peaty turf are represented (Bishop 2013; see also Church et al. 2007). The Skara Brae wild seed assemblage post-dates the cereal crop assemblage and may consist mainly of peaty turf burning debris (Rowley-Conwy & Bishop 2021): it is included here as an illustration of the range of wild taxa recovered at the site and to highlight similarities with the Braes of Ha'Breck assemblage. Further experimental investigation is needed to disentangle these data, which is beyond the scope of the present study. Habitat and perennation type were classified for the wild seed taxa using ecological information from Hill et al. (2008), with the general habitat categories following Bogaard & Jones (2007) (see Table S2). No threshold was set for the quantity of wild seed taxa per sample or site phase as this was generally low at Balbridie and the Braes of Ha'Breck, where fully processed crops may be represented. Tree and shrub taxa were excluded from the ecological assessment of wild seed data as they do not grow as weeds in arable fields. Only seeds with species or genus-level identifications with clear habitat associations were included in the ecological analyses (see Table S2). The