The Indian subcontinent is expected to witness signifi-cant changes in climatic conditions in the future, but the implications of such changes for future spatial dis-tribution of different biomes in the subcontinent are unclear. We sought to understand the potential shifts in the distribution of biomes in India by 2070 under different emission scenarios, identify biomes and re-gions of the country that are particularly at risk from future changes in climate, and quantify uncertainties associated with the predictions. We used an ensemble classifier (random forest) to model current and poten-tial future distribution of biomes in India for different climate trajectories under the newly developed repre-sentative concentration pathway (RCP) scenarios. Cli-mate projections from 19 and 17 different general circulation models (GCMs) were used to predict future biome distributions in India under the RCP 4.5 and RCP 8.5 scenarios respectively. For each scenario, mod-el outputs from different GCM projections were com-bined using a simple majority voting criterion. Approximately 630,000 sq. km (18%) of the country is predicted to experience biome shifts under the RCP 8.5 scenario and 486,000 sq. km (14%) under the RCP 4.5 scenario by 2070. Drier tropical biomes are likely to be replaced by wetter biomes, while temperate biomes are predicted to be dominated by vegetation characteristic of a warmer climate in the future. There was a high to moderate level of agreement between predictions of different GCMs. Our results suggest that biome shifts will be largely concentrated in the east-central and northern parts of the country, with tropical and sub-tropical dry forests, savannas, grasslands and xeric habitats particularly at risk. Future studies should focus on elucidating the responses of different vegetation sub-formations within individual biomes in order to gain a finer-scale understanding of vegetation responses to future climate change in the Indian subcontinent.
Keywords
Ioclimatic Envelope Models, Biome Shifts, Ensemble Modelling, Random Forests, Representative Concentration Pathways.
User
Font Size
Information