Quantifying human impacts to tease apart cultural and climatic drivers of Holocene vegetation change
Cultural pollen indicators in pollen records are an established tool to reconstruct the history of human impact on vegetation and landscape, particularly useful to disentangle the influence of human activities and climatic variability on ecosystems. To infer human impact from pollen diagrams, Behre (1981) established a list of cultural pollen indicators that represent plant taxa that benefit from human activities directly (crops) or indirectly (weeds). The pollen indicators methodology remains the prevalent approach when assessing land-use and its related anthropogenic impact in pollen diagrams. For this purpose, the list of cultural indicators developed by Behre (1981) and elaborated by following research has been widely used and adapted in very different geographical, cultural and vegetational contexts. More recently, index-like approaches, based on the cultural pollen indicators, aim to adapt the method for numerous geographical locations and land use types. In Manuscript 1, we review the most common European approaches to assess their performance at six sites spanning a continental gradient over the boreal, temperate and Mediterranean biomes. We present new insights on how these methodologies can assist in the interpretation of pollen records and on how a careful selection of pollen types and/or indexes according to the specific geographical scope of each study is key to get meaningful reconstructions of human activity through time. We emphasise how important it is to determine if an indicator is adventive (exotic) or apophyte (native) within a study area. Although we can identify approaches or indexes that perform much better than others, we conclude that alongside the increasing wealth of pollen datasets there is a need for developing novel tools that may assist for numeric human-impact reconstructions.
Another factor that affects the indicative capacity of cultural pollen indicators is the taxonomic resolution in which they are identified. This is because different levels of taxonomic resolution are reached in palynology, depending on several factors such as the analystâ€™s expertise, the palynological school, the aim of the study, the preservation of the pollen grains, the reference collections, the available determination keys and the microscope facilities. In Manuscript 2, we assess the impact of taxonomic resolution on the indicative capacity of cultural pollen types. To achieve this, we attribute the pollen types of sixteen sedimentary records, located along a latitudinal gradient spanning from Switzerland to Italy, to three levels of taxonomic resolution previously proposed at the European scale (Giesecke et al., 2019). Our results show that higher taxonomic resolution improves the identification of human impact by enhancing the indicative power of important pollen indicators widely used in the research field. On this basis, we identify highly resolved sensitive pollen types, especially some most commonly used such as Plantago lanceolata-type, Centaurea cyanus, and Cannabis sativa-type. Our results my contribute to the improvement of palynological reconstructions of land use by identifying key pollen types with important detail of identification. We conclude that it is essential to aim at highest taxonomic resolutions for reconstructing past human impact, even if in specific geographical contexts (e.g., the Mediterranean) this might not be sufficient to fully discriminate human impact from natural processes.
Taken together with these first two studies, we aim to contribute to the improvement of palynological reconstructions of land use and human impact by identifying key pollen types whose microscopic identification and origin and timing of introduction requires particular attention. In order to have a semi quantitative methodology and to further contribute to human pollen reconstructions, we propose a new methodology based on cultural pollen indicators, the agricultural LUP (Land Use Probability Index), in Manuscript 3. We train the novel index by using the most suitable cultural indicators found in twenty Holocene pollen sequences along a latitudinal transect encompassing cool temperate colline/montane, cool to cold subalpine/alpine, warm temperate submediterranean, warm mesomediterranean, and subtropical thermomediterranean vegetation conditions, in Switzerland and Italy. We validate the LUP index by (1) applying the generated LUPs to independent pollen records and (2) with a proxy curve for past human activity generated using radiocarbon data from the archaeological context. We discuss the potential of our method to be applied along Europe, which may significantly contribute to refine pollen-based land-use reconstructions in Europe.
To conclude, the cultural pollen indicators methodology (Behre, 1981) was and still is the most important tool that palynologists have to infer human impact from pollen diagrams. In this thesis, we aim to call attention to how the methodology is used, especially when adapting it to areas outside its original calibration realm. We recommend that when identifying human pollen indicators, the highest possible taxonomic precision should be aimed. Also, it is important to define if a human indicator is an adventive or an apophyte within a study area. This very much depends on the vegetational context in the different biomes of Europe. In future, novel methodologies as aDNA or eDNA approaches may additionally inform palaeoecologists and archaeologists to gain better reconstructions of past human activities.