The environment they lived in: anthropogenic changes in local and regional vegetation composition in eastern Fennoscandia during the Neolithic

Vegetation History and Archaeobotany https://doi.org/10.1007/s00334-020-00796-w ORIGINAL ARTICLE The environment they lived in: anthropogenic changes in local and regional vegetation composition in eastern Fennoscandia during the Neolithic Teija Alenius1,2 · Laurent Marquer3 · Chiara Molinari4 · Maija Heikkilä5 · Antti Ojala6 Received: 6 October 2019 / Accepted: 12 August 2020 © The Author(s) 2020 Abstract Understanding about regional versus local changes in vegetation is critical in answering archaeological questions, in particular at a time when humans are assumed to have caused higher disturbances at local scales rather than regional scales; this is the case during the Neolithic. The aim of this paper is to assess the impact of Neolithic land use on regional and local vegetation dynamics, plant composition and disturbance processes (e.g. fire) in eastern Fennoscandia. We apply the Landscape Reconstruction Algorithm (LRA) to high-resolution pollen records from three lacustrine sediment cores that cover the Neolithic period. We calculate changes in vegetation composition and the rate of plant compositional change. Fire dynamics are estimated as an indicator of land use, although fire can result from both natural and anthropogenic disturbances. Our results show that during the Early Neolithic, changes were mainly driven by natural and climate-induced factors and vegetation composition and fire activity were similar at both regional and local scales. From ca. 4000 bc onwards, trends in vegetation and fire dynamics start to differ between regional and local scales. This is due to local land uses that are overshadowed at the regional scale by climate-induced factors. The use of the LOVE model in pollen analyses is therefore very useful to highlight local land uses that are not visible by using REVEALS. Keywords Land cover · Plant compositional change · Pollen · Landscape reconstruction algorithm (LRA) · Fire · Environmental history · Human–environment interactions Introduction Communicated by M.-J. Gaillard. Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00334-020-00796-w) contains supplementary material, which is available to authorized users. * Teija Alenius 1 Turku Institute for Advanced Studies, TIAS (Department of Archaeology), University of Turku, 20014 Turku, Finland 2 Department of Cultures, University of Helsinki, Unioninkatu 38F, P.O. Box 59, 00014 Helsinki, Finland 3 Research Group for Terrestrial Palaeoclimates, Max Planck Institute for Chemistry, Hahn‑Meitner‑Weg 1, 55128 Mainz, Germany Holocene sediment archives from lakes, ponds and peat bogs in Europe provide evidence of human activities during the last ca. 12,000 years. Deforestation practices aimed at expanding areas for agriculture are commonly identified in pollen records from these archives via a decrease in tree pollen percentages and an increase in herbs and crop pollen types (e.g. Ojala and Alenius 2005; Gaillard et al. 4 Department of Physical Geography and Ecosystem Science, Lund University, Sölvegatan 12, 223 62 Lund, Sweden 5 Ecosystems and Environment Research Programme, Faculty of Biological and Environmental Sciences, and Helsinki Institute of Sustainability Science, University of Helsinki, Viikinkaari 1, P.O. Box 65, 00014 Helsinki, Finland 6 Geological Survey of Finland, Vuorimiehentie 5, P.O. Box 96, 02151 Espoo, Finland 13 Vol.:(0123456789) Vegetation History and Archaeobotany 2015; Dubois et al. 2017). Additionally, deforestation and land use cause erosion processes that can be observed in sediment archives (e.g. disturbances in accumulation rates). Anthropogenic impact on vegetation is related to the spread of human activities in a catchment area. It is related to local activities that can expand through time to a regional scale. Based on pollen-based modelling and anthropogenic land cover change scenarios, we have increase our understanding about regional and subcontinental-scale vegetation and land use patterning in Europe, although this information mostly refers to deforestation (e.g. Trondman et al. 2015; Kaplan et al. 2017; Marquer et al. 2017; Roberts et al. 2018; Zanon et al. 2018). The development of pollen databases (e.g. Fyfe et al. 2009; Giesecke et al. 2014) allows the assessment of changes in Holocene vegetation composition and dynamics at regional and subcontinental scales (e.g. Marquer et al. 2014, 2017; Giesecke et al. 2017). These variations are basically due to climate, environmental (e.g. soil characteristics, characteristic traits of plant species, extreme disturbance events) and/or land use factors. Recent studies have shown that climate and land use are important drivers of vegetation change during the Holocene in northern Europe and Fennoscandia (Reitalu et al. 2013; Kuosmanen et al. 2016, 2018; Marquer et al. 2017). While climate is the dominant factor during the first half of the Holocene, land use becomes one of the primary drivers during the last 2,500 years. Pollen-based reconstruction of past vegetation at regional and subcontinental scales represents a summary of the outcomes from all the interactions between vegetation dynamics, climate influences and impacts of land use within 104 km2 or larger (Prentice 1985). Land use impacts on vegetation composition and dynamics based on pollen are certainly recorded differently at local scales than at regional scales. The understanding of land use consequences at a local scale is particularly important to explore changes in the landscapescale vegetation cover during the Neolithic; Neolithic populations, with subsistence based on hunting and gathering, are assumed to have had lower impact on vegetation at a regional scale than at a local scale in some areas. Furthermore, recent studies show that forest landscapes may have been shaped by human manipulation over millennia, although to a small extent. These human activities have caused local changes in biodiversity, species distributions and biomass (Dambrine et al. 2007; Innes et al. 2013; Pini et al. 2017). A better knowledge of early anthropogenic impact at a local scale can provide insights into human-nature relationships at a time when populations were highly dependent on climate and environmental variabilities (Boivin et al. 2016). However, this information can only be gained by studying known spatial resolutions and using adequate indicators of vegetation change, such as land cover, plant diversity and composition indices. 13 Traditional pollen data expressed as proportions provide information about a mix of the regional and local pollen assemblages, thus making it difficult to assess the local changes in vegetation. The Landscape Reconstruction Algorithm (LRA, Sugita 2007a, b) was proposed to overcome this issue and provides pollen-based reconstructions of regional and local vegetation in quantitative terms, i.e. plant cover at a defined spatial scale (a few metres to kilometres from the (...truncated)