A Palaeoenvironmental Investigation of Amazonian Lowland Sensitivity to Climatic Drivers Using Pollen-based Modelling Approaches

This project goes beyond the current state-of-the-art in the field of tropical palaeoecology. So far, sampling of modern pollen in Amazonia has been carried out by several authors, revealing important information about the transport and representation of pollen in tropical environments. Already 2,354 modern samples across Latin America have been gathered (see Figure 1). However, only a fraction (548) of the samples available cover Amazonia. In the European Pollen Database, almost 3,500 samples are included, with about 1,400 for the Mediterranean region. The abundance of modern samples in Europe has allowed both a deep understanding of the relationships between pollen composition and local/regional vegetation, as well as enabling accurate climatic reconstructions. Considering that Europe (10mn km2) covers a little more than the size of the main drainage area of the Amazon River (7mn km2), and that NW Amazon is the same size as the Mediterranean (2.5-2.7mn km2), it appears clear that more samples are necessary to achieve the same level of coverage in Amazonia that we have in Europe (aim of 200-400 new samples including unpublished/non-uploaded datasets from collaborators). Our case study will focus on Ecuador and Peru lowlands because both countries share similar biomes (palm swamps) and samples are already available but their coverage needs enhancing, especially in Peru which is less well-represented compared to Ecuador. This will also be the first attempt at implementing the use of water pollen traps as well as aerial and ground traps, crucial for understanding the potentially key role of water-transported pollen signal, and thus achieve a robust calibration of fossil records.

Furthermore, our aims to implement quantitative pollen-based climate reconstructions in Amazonian lowlands also represent an innovative step further. Quantitative techniques such as transfer functions have been successfully applied in temperate regions in Europe and parts of Asia, allowing palaeoecologists to reconstruct past climates using data from fossil pollen, thereby producing models directly applicable to their study without relying on loose comparisons with regional climatic records. However, these methods have never been attempted for the reconstruction of climate parameters in Amazonia. The main reason for this has been the greater focus of palynology on Europe until recent decades, and the lack of availability of modern pollen datasets for Latin America. Moreover, there are inherent difficulties when working in the tropics, such as the high heterogeneity of the landscape in responses to climate, missing analogues, taxonomic uncertainty and discrepancies in the representation of certain taxa. As the number of palaeoecological records from lakes and peat bogs in this region grows, however, there is need for a clearer way to discern external drivers of vegetation change. Recently, modelling of plant trait composition in Peru using modern pollen has been successfully carried out, showing that challenges faced by palaeoecologists in the tropics can be overcome with enough data and the correct network of experts.