Relative pollen productivity estimates (RPP) of some key pollen taxa in south-east India

International audienceReconstructing the past vegetation conventionally uses pollen percentages (PPs) or pollen accumulation rates (PARs). Qualitative interpretation of PPs and PARs is based on insights gained from modern analogues of the pollen-vegetation relationship or on the comparative approach that compares past and modern pollen assemblages. Quantifying past vegetation/ land cover using pollen data requires other approaches that can correct the non-linear relationship with vegetation and biases due to inter-taxonomic differences in pollen productivity, dispersion and deposition. Relative Pollen Productivity (RPP) can be estimated using the Extended R-Value (ERV) model (Prentice and Parsons, 1983), provided modern pollen assemblages and distance weighted vegetation abundance (DWPA) around the pollen sampling point are available from actual field measurements (e.g. Li et al., 2017). We present for the first time RPPs of some of the key pollen taxa retrieved from the fossil pollen sequences located within the broader Acacia–Albizia biogeographic zone in the drier parts of south-east India along the Coromandel coast, falling in the rain shadow of the southwest or summer monsoon. To achieve this, surface soil samples for pollen analysis and distance-weighted vegetation data were collected from 14 sites in the current mosaic landscape of this biogeographic zone using the Crackles Vegetation Survey protocol (Bunting et al., 2013). RPPs of 6 key pollen taxa, viz.,. Poaceae, Acacia/Albizia, Dodonaea, Justicia, Melastomataceae/ Combretaceae, and Randia were estimated with the ERV model (sub-model 3) for which the best likelihood function curve was obtained using the Prentice Bog (dispersion) model. Results indicate a Relevant Source Area of Pollen (sensu Sugita, 1994) of 164 m radius for the modern pollen samples. The estimated RPPs relative to Poaceae (RPP set to 1) are highest for Acacia/Albizia, Dodonaea, and Justicia (>10), and are