
Earth Sciences
Striking forest revival at the end of the Roman Period in north-western Europe
C. Lambert, A. Penaud, et al.
Discover the intriguing '1.7-1.4 ka AP event' in north-western France, where a notable increase in arboreal activity coincides with the retreat of coastal societies. This research by C. Lambert, A. Penaud, M. Vidal, C. Gandini, L. Labeyrie, L. Chauvaud, and A. Ehrhold suggests that climate degradation may have played a pivotal role in the collapse of agrarian societies at the end of the Roman Period.
~3 min • Beginner • English
Introduction
The study addresses how Late Holocene climate variability and ocean-atmosphere dynamics, particularly solar activity cycles, the North Atlantic Oscillation (NAO), and North Atlantic gyre dynamics, influenced environmental change and human societies in north-western Europe. Over the last 2500 years BP, anthropogenic deforestation and land use complicate paleoenvironmental signals. By leveraging palynological proxies (pollen and marine/freshwater microalgae) from a coastal macro-estuarine archive in NW France, the research aims to reconstruct vegetation and sea-surface changes and assess their implications for regional coastal societies. The work focuses on identifying and interpreting a pronounced arboreal increase during 1.7–1.4 ka BP and its linkage to climatic mechanisms (SPG/NAO) and socio-economic dynamics at the end of the Roman Period.
Literature Review
Holocene climate variability includes a ~2500-year solar-related cyclicity and a ~1500-year ocean-related cyclicity emerging in the Mid- to Late-Holocene. Subpolar gyre (SPG) dynamics modulate northward heat transport via the North Atlantic Current (NAC) and influence deep-water formation and SSTs. NAO governs westerly wind position and vigor, with positive modes associated with increased storminess and moisture over northern Europe. Previous records (e.g., density differences south of Iceland, benthic foraminiferal δ18O in the northern Bay of Biscay, European Atlantic storm event stacks, and speleothem records from northern Spain) document millennial to centennial variability in SPG strength, storminess, and hydroclimate. In southern/western Europe, intervals of aridification and forest retraction have been reported, suggesting contrasting north–south hydroclimatic patterns akin to NAO behavior. Land-use change and deforestation intensified during the Late Holocene and especially post-Medieval Climate Optimum, further complicating interpretations of runoff proxies.
Methodology
Study area and cores: Two high-resolution sediment cores from the Bay of Brest (BB), NW France, were used to reconstruct land-sea changes over the last 2500 years BP. Core G (48°19′14″ N; 4°23′5″ W; 7.4 m depth; collected in 2003 by vibrocorer, Défis Golfe de Gascogne cruise) covers 2530–1660 years BP; core KS-02 (48°18′46″ N; 4°24′27″ W; 8 m depth; collected in 2010 by gravity corer, EssCALICO cruise) covers 1660–500 years BP. The records were combined to form a composite sequence with ~35-year temporal resolution.
Chronology: Fourteen radiocarbon AMS dates were measured on marine gastropods. Dates were calibrated using CALIB 7.1 with the Marine13 curve, applying a 400-year ocean reservoir age correction and a local ΔR of 40 ± 23 years for the Brittany region. An age-depth model for the composite core was constructed using the Bacon Bayesian modeling package, yielding a continuous record from 2528 to 488 cal BP.
Palynology: Fifty-eight samples (27 from G, 31 from KS-02) were processed from the <150 µm fraction using standard HCl and HF chemical treatments and 10 µm sieving to isolate organic microfossils. Residues were mounted in glycerine. Pollen and dinoflagellate cysts (dinocysts) were identified under a Leica DM2500 microscope at 630× magnification. For each sample, ≥300 pollen grains and ≥150 dinocysts were counted. Percentages were calculated on total sums without taxon exclusion. Concentrations (palynomorphs/cm³) were determined via the Lycopodium clavatum marker-grain method. Key indicators included total tree pollen, Corylus, riparian taxa (Alnus, Salix, Betula), dinocyst Lingulodinium machaerophorum, Anthropogenic Pollen Indicators (API) with/without Cerealia-type, and Cerealia-type concentrations. Sedimentation rates and total pollen fluxes were also derived.
Comparative records: BB palynological and sedimentological data were compared to regional to North Atlantic records: European Atlantic storm event stacks, northern Europe Holocene storm periods, SPG strength proxies (density differences south of Iceland; δ18O in CBT-CS11 core from the northern Bay of Biscay), speleothem-derived precipitation for northern Spain (Cueva de Asiul), and Loire estuary Ti-XRF (core KV14bis) as a runoff/terrigenous proxy. Patterns were interpreted in the context of NAO-like atmospheric configurations and SPG/NAC dynamics.
Key Findings
- Identification of a 300-year-long 1.7–1.4 ka Arboreal Pollen (AP) rise event (ca. 1700–1400 years BP; AD ~250–550) in the Bay of Brest watersheds. Tree pollen percentages reached up to ~90% (baseline modern BB ~45%), exceeding even nearby Mesolithic (~9000 BP) values (~80%).
- The AP event coincided with a pronounced increase in Corylus (hazel) and riparian taxa (Alnus, Salix), suggesting moister and milder conditions. Total pollen fluxes and dinocyst Lingulodinium machaerophorum increased, indicating enhanced fluvial discharge and estuarine stratification influenced by freshwater input.
- While NW France showed humid conditions, contemporaneous records to the south (Loire estuary Ti-XRF and northern Spain speleothem precipitation reconstructions) indicated reduced runoff/precipitation (aridity) between ~1700 and 1300 BP, implying a NAO-like dipole with contrasting north–south hydroclimate across Europe.
- The climatic interval overlaps intervals of heightened storminess across European Atlantic coasts (1700–1200 BP and broader 1900–1050 BP), consistent with positive NAO-like configurations and a strengthened SPG leading to intensified NAC and northward westerlies.
- Anthropogenic signal: During the peak arboreal interval, cultivated plants (Cerealia-type) and API declined sharply, synchronous with the tree peak, suggesting agricultural abandonment and reforestation of wetlands and fields, i.e., a local collapse/withdrawal of coastal agrarian activity in NW France around ~1500 BP.
- After ~800 BP, Loire Ti-XRF and northern Spain precipitation proxies diverge: increased Loire detrital signal reflects human-driven soil erosion due to extensive deforestation post-Medieval Climatic Optimum rather than purely climatic runoff, while BB tree percentages decline below long-term average, marking intensified Late Holocene anthropogenic impact.
- Sampling resolution ~35 years; composite age coverage 2528–488 cal BP; multiple independent lines of evidence (pollen taxa, dinocysts, pollen fluxes, sedimentation, external marine and speleothem proxies) support the climatic and societal interpretations.
Discussion
The findings reveal that the exceptional forest resurgence during 1.7–1.4 ka BP resulted from both climatic and anthropogenic processes. A strengthened SPG and intensified NAC likely produced warmer, wetter winters in NW Europe, shifting storm tracks northward and enhancing moisture delivery, which promoted riparian and mixed oak forest growth documented by increased Corylus and riparian pollen and higher pollen fluxes. Concurrent reduction in agricultural indicators (API and Cerealia-type) during the tree peak indicates reduced land use and possible abandonment of agricultural plots, allowing reforestation.
The contrasting hydroclimate between NW France (humid) and SW Europe (arid) mirrors NAO-like atmospheric configurations that generate dipolar precipitation patterns across Europe. Elevated storminess inferred from independent coastal and northern European storm records during this interval supports a positive NAO-like state associated with SPG strengthening. These climatic stressors may have exacerbated socio-economic and political disruptions at the end of the Roman Period, contributing to a regional collapse or retreat of coastal agrarian societies in Brittany.
Post-800 BP trends demonstrate the increasing dominance of anthropogenic forcing (deforestation and soil erosion) on terrigenous signals and vegetation openness, underscoring the evolving balance between natural and human drivers through the Late Holocene. Overall, the study links ocean-atmosphere dynamics to terrestrial ecosystems and human land-use trajectories, offering a nuanced view of climate–society interactions in pre-industrial times.
Conclusion
This study documents an unprecedented 300-year arboreal expansion (1.7–1.4 ka AP event) in NW France, surpassing Mesolithic tree cover, during a period of SPG strengthening and NAO-like atmospheric conditions. Multiproxy evidence indicates wetter, milder conditions in NW Europe contrasted with aridity to the south, consistent with a dipolar hydroclimatic pattern. The synchronous decline of cultivated plants and anthropogenic pollen indicators suggests agricultural retreat and reforestation, pointing to a climate-amplified collapse of coastal agrarian societies at the end of the Roman Period. After ~800 BP, intensified human activities dominate landscape opening and erosion signals.
Future research could integrate higher-resolution, multi-site coastal and inland archives with improved chronological control to disentangle climatic versus anthropogenic signals, couple paleoenvironmental data with regional archaeological datasets for finer-scale land-use reconstructions, and apply climate modeling to test the mechanisms linking SPG/NAO variability, storminess, and socio-environmental outcomes.
Limitations
- Chronological uncertainties due to marine reservoir age variability and the application of regional ΔR corrections may affect precise event timing.
- Palynological data provide regional signals that can be influenced by differential pollen production, transport, and preservation; attribution to specific land-use changes may be indirect.
- Archaeological records for the late Roman Period suffer from preservation and documentation biases (e.g., scarcity of diagnostic ceramics, poor conservation of late levels), complicating robust assessments of settlement dynamics.
- Spatial heterogeneity: regional variability in societal responses and hydroclimate means BB findings may not uniformly represent broader NW Europe.
- Proxy interpretation shifts over time (e.g., Ti-XRF transitioning from runoff to human-driven erosion indicator after ~800 BP) introduce complexities in cross-proxy comparisons.
Related Publications
Explore these studies to deepen your understanding of the subject.