Marine resource abundance drove pre-agricultural population increase in Stone Age Scandinavia

The study investigates how mid-Holocene climate and environmental changes influenced cultural and demographic trajectories in Southern Scandinavia, focusing on the centuries preceding the introduction of agriculture. While agriculture commonly triggers population increases, in Southern Scandinavia a sustained pre-agricultural population growth (~500 years) preceded Neolithisation. The central research question is whether Mesolithic cultures selected marine resources due to their increased abundance (i.e., higher marine production) during the Holocene Thermal Maximum (HTM, ca. 8000–4000 BP), thereby enabling population growth, or whether population packing driven by increased terrestrial production forced greater reliance on marine resources. The paper aims to resolve this by integrating high-resolution, chronologically constrained climate, sea-level, environmental and archaeological proxies from multiple coastal sites to test the linkage between marine productivity, resource use, and human population trends.

Previous research using summed probability distributions (SPDs) of radiocarbon dates has revealed long-term global population trends characterized by low growth rates punctuated by boom-bust cycles, with proposed drivers including climate variability, resource availability, disease, and social factors. In northern Europe, boom-bust patterns are well-documented for agrarian societies, but less so for pre-agricultural hunter-gatherers. Studies in Fennoscandia and the Baltic have linked hunter-gatherer demographic shifts to environmental changes and resource abundance, including intensified use of aquatic resources. Classic models (e.g., Binford) suggest aquatic resources are fallback options when terrestrial resources are limited, particularly at high latitudes. However, in Southern Scandinavia, despite temperate conditions, the mid-Holocene Kongemose (ca. 8400–7400 BP) and Ertebølle (ca. 7400–5900 BP) cultures show predominant marine resource use, contrasting with the earlier Maglemose culture’s mixed terrestrial/freshwater diet. The literature documents rising sea levels (Littorina transgression), increased coastal salinity, and warming during the HTM, alongside archaeological evidence of extensive shell middens and rich marine faunas, suggesting potential marine productivity increases as a driver of subsistence choices and population change.

The authors assembled a multiproxy dataset from six Danish coastal sites to reconstruct mid-Holocene climate, sea level, coastal salinity and productivity, and to compare these with archaeological indicators of marine resource exploitation, technology, diet, and population. Key components included: (1) Climate and sea level: pollen-inferred mean annual temperature from Lake Trehörningen (SW Sweden) and sea-level reconstructions from Blekinge (SE Sweden). (2) Coastal salinity: diatom-inferred salinity (WA-PLS component 2; r² = 0.87, RMSEP = 0.44 sqrt units; 1000× bootstrapping) from five Danish fjord/estuarine sites. (3) Marine productivity: sedimentary pigment biomarkers (algal pigments) and accumulation rates at Kilen, Horsens Fjord, and Tempelkrog; additional biological proxies including foraminiferal and mollusc fluxes; and regional organic carbon proxies (TOC percentages/fluxes) in Baltic sediments. (4) Archaeological resource use: a novel shell midden accumulation indicator using the SPD of 231 calibrated 14C dates on European flat oyster (Ostrea edulis) shells from Danish middens; composition and abundance of faunal remains (fish, birds, mammals). (5) Population trends: regional population density proxies for southern Sweden, Jutland, and the Danish Islands derived from archaeological radiocarbon datasets. (6) Technology: cumulative count and timing of marine exploitation technologies (e.g., fish traps, hooks, nets, dugout canoes) and early agricultural tools (polished flint axe, ard). (7) Diet: stable isotope analyses (δ13C, δ15N, δ2H) of human and dog bone collagen to quantify marine versus terrestrial dietary contributions, with δ2H as an additional trophic indicator. (8) Land-use and erosion: lake sediment accumulation rate (SAR), non-arboreal pollen percentages, and indicators of farming (Plantago lanceolata, cereal pollen) from Lake Gudme Sø to track Neolithic land clearance and nutrient export. Coring, subsampling, and proxy preparation followed established protocols; age-depth models were based on 14C-dated terrestrial plant macrofossils (or molluscs at Korup Sø). For the midden SPD, significance was evaluated against 1000 simulated SPDs generated from 231 randomized samples (uniformly distributed 8100–3500 BP) with similar error structure.

• Two distinct phases of elevated marine productivity were identified across Southern Scandinavia: P1 (ca. 7600–7100 BP) and P2 (ca. 6400–5900 BP), superimposed on generally high productivity from ca. 7600–5000 BP during the HTM and high sea levels.
• These productivity pulses coincide with increased salinity and higher sea levels; coastal diatom-inferred salinities were higher than modern, and warmth-demanding/high-salinity taxa appeared.
• Shell midden abundance (oyster SPD) rose from ca. 7600 BP, peaked ca. 6400–5700 BP, and declined after ca. 4200 BP, indicating intense exploitation of marine resources during the Mesolithic and early Neolithic.
• Population proxies show significant pre-agricultural growth, especially a clear increase during P2 (ca. 6400–5900 BP), culminating in an approximately four-fold Mesolithic-to-Neolithic contrast (P2 baseline ×1 vs Neolithic ×4), with the Mesolithic increase preceding the adoption of agriculture (~5900 BP).
• Stable isotopes indicate a major dietary shift: Kongemose/Ertebølle humans exhibit strong marine signatures (δ13C ≈ −13‰; δ15N ≈ 13‰), shifting to predominantly terrestrial values in the Neolithic (δ13C ≈ −20‰; δ15N ≈ 9.5‰). δ2H shifted from +66‰ (Mesolithic) to −5‰ (Neolithic), corroborating the trophic shift.
• Mean annual temperatures during the HTM peak (ca. 7500–6000 BP) were ~2.5°C higher than recent pre-industrial; highest sea levels occurred ca. 7600–5000 BP.
• Elevated marine production persisted into the early Neolithic until ~5000 BP, likely sustained by increased terrestrial nutrient export following agricultural land clearance.
• The timing of peak shell midden accumulation appears slightly earlier than peak primary production in some records; this offset may reflect dating uncertainties (~200 years) or ecological/human feedbacks (e.g., reaching carrying capacity at higher trophic levels or reduced grazing pressure).

The results directly address the central question by demonstrating that climatically and oceanographically driven increases in marine productivity during the HTM (warmer temperatures, higher sea levels, increased salinity and exchange with the North Sea) created abundant, easily accessible coastal resources. Mesolithic hunter-gatherers in Southern Scandinavia expanded marine exploitation, as evidenced by large shell middens, rich aquatic faunas, maritime technological innovations, and strong marine dietary isotopic signatures. This resource base underpinned a substantial pre-agricultural population increase, particularly during P2 (6400–5900 BP), prior to the arrival of farming. Mechanistically, sea-level transgression likely mobilized nutrients from inundated coastal margins and increased the connection to nutrient-rich, oxygenated North Sea waters; a predominantly positive NAO-like state during the HTM may have further enhanced inflow and productivity. The study contrasts the earlier Maglemose reliance on terrestrial/freshwater resources with Kongemose/Ertebølle maritime adaptations, potentially reinforced by declining terrestrial hunting quality under dense primeval forests and local extinctions of key prey (elk, aurochs) on the Danish islands. After ~5900 BP, agriculture broadened the subsistence base, drove land clearance, increased catchment erosion and nutrient export, and sustained coastal productivity briefly while dietary reliance shifted decisively to terrestrial foods. The observed slight lead of human exploitation over peak primary production may reflect either chronological uncertainties or ecological dynamics/feedbacks within the food web and human harvesting pressure.

This study demonstrates that, under favorable climatic and oceanographic conditions (HTM), enhanced marine productivity can support substantial pre-agricultural population growth and cultural development, including maritime technological innovation, in temperate regions. Two coherent productivity pulses (P1 and P2) align with intensified coastal resource exploitation and a marked population rise that preceded Neolithisation in Southern Scandinavia. The findings challenge assumptions that agriculture is a prerequisite for significant population growth and complexity, highlighting the potential for marine-based subsistence to sustain expanding societies. The HTM provides a useful analogue for understanding future coastal productivity changes under warming and sea-level rise. Future research should expand coastal archaeological excavations, develop larger meta-datasets (e.g., SPDs integrating diverse materials), and apply coordinated multiproxy environmental reconstructions to refine the timing, mechanisms, and socio-ecological feedbacks linking marine resource availability, technology, mobility/sedentism, and demographic change.

• Chronological uncertainties: Differences in 14C calibration and age-depth models across proxies may introduce ~200-year offsets, complicating alignment of ecological and archaeological events.
• Proxy coverage and spatial variability: Not all proxies are available at all sites; site-specific conditions introduce heterogeneity that may limit regional generalizations.
• Limited direct productivity reconstructions: Few independent marine palaeoproductivity records exist for the Kattegat/Baltic, constraining cross-validation.
• Mechanistic ambiguity: While multiple mechanisms (sea-level rise, nutrient inputs, North Sea inflow, NAO state) are plausible, their relative contributions cannot be unambiguously resolved.
• Causality in exploitation–productivity timing: The observed lead of shell midden accumulation over primary productivity peaks cannot be definitively attributed to dating error versus ecological/human feedbacks.
• Cultural inferences: Assessments of societal complexity and mobility/sedentism extend beyond the direct scope of the environmental dataset and remain partly inferential.