A pollen-based land-cover atlas covering the last 200 years of Bassies valley vegetation changes : results from a new spatialization method

2021 
Since the 1950s, the Pyrenean Mountains have undergone abandonment of pasture lands and significant decrease in grazing pressure, leading to a dramatic decline of open lands and their related biodiversity. These findings are mainly based on two types of datasets that remain difficult to combine. On the one hand, they rely on spatially-explicit datasets spanning a few decades, derived from remote sensing (aerial photographs or satellite images). On the other hand, palaeoecological data help to address conservation issues by providing long-term records of land-use and vegetation changes and their possible impacts on biodiversity. However, the spatially explicit, small-scale and long-term reconstruction of land-cover changes, combining both types of data, remains very challenging. The Bassies valley (OHM Pyrenees - haut Vicdessos) is representative of the Pyrenean mountains abandonment scenario. This poster focuses on the land-cover history of this mountainous area, its local variability and spatial patterns over the last 200 years, based on the sedimentary records of pollen from eight lakes and bogs. Changes in percentage cover of plant taxa at 10-20 year intervals within a 1 km radius around each site were quantified by the landscape reconstruction algorithm (LRA). A new probabilistic and statistic method was then applied to the LRA estimates, producing fine-scale maps (20m pixels) for eight land-cover types. These maps will then be used together with a modern floristic dataset to study the legacies of land-cover composition and configuration changes on floristic diversity. Such results will be further compared to pastoral archives to better understand the human impact on land-cover, floristic and landscape diversities in the valley over the last two centuries.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []