MODERN AND FOSSIL POLLEN DATA SYNTHESIS FROM THE INDIAN SUBCONTINENT: AN OVERVIEW OF THE CURRENT STATUS AND FUTURE PERSPECTIVES

The INDSUBIO (INDian SUbcontinent BIOmes) initiative aims to understand the natural dynamics of vegetation on the Indian subcontinent in response to long-term climate changes, using a combination of modeling tools, modern environmental data and pollen-based vegetation reconstructions. INDSUBIO has compiled modern and fossil pollen samples for the mid-Holocene and last glacial maximum (6000 and 18,000 ¹⁴C yr B.P. respectively). The data set comprising 693 modern, 43 mid-Holocene and 18 Last Glacial Maximum records has been used to map palaeovegetation patterns based on an objective classification scheme in which the pollen taxa are first classified into plant functional types (PFTs), and then associations of PFTs are used to determine major vegetation types (biomes). The objective classification scheme was tested using the modern surface sample data, and successfully reproduced observed biomes at nearly 90% of the modern pollen sites. Such an agreement despite a large range of climates (from tropical to alpine and desert) and the complex topography of the region indicates the robustness of the approach and provides a measure of confidence that can be placed in palaeo-reconstructions. The reconstructions show that the vegetation response to mid-Holocene changes in orbital forcing was comparatively muted on the Indian subcontinent, but that large changes in vegetation patterns occurred in response to colder and drier conditions at LGM.

Pollen data is sparse in many areas of the subcontinent. New modern and fossil pollen records are required to improve the current reconstructions. A research project focussing on the forests of the Eastern Ghats in southern India has been initiated involving, a) modern pollen sampling, b) quantitative vegetation studies at pollen-sampling sites, including systematic field studies of plant traits relevant to improving PFT definitions, and c) remote-sensing and GIS for, identifying undisturbed sites using landscape based disturbance analysis and change detection using multi temporal data with a grid hierarchical scheme and mapping the modern vegetation.

To homogenize primary data generation and synthesis and to facilitate better dialogue between data producers and modellers, a new regional online user-interactive database, using MySQL and PHP is currently being set up.