Vegetation Dynamics and Forest Degradation Monitoring: The Case for the World Largest Mangrove
Mahmood, Abu RJ; Donoghue, Daniel NM; Atkins, Peter J
Department of Geography, Durham University, Durham, UNITED KINGDOM

The Sundarbans is a unique landscape and is the largest intact, contiguous mangrove forest on Earth, lying at the Ganges-Brahmaputra delta. It is thought to have the highest mangrove biodiversity in the world. But, noted from various recent studies: sea level rise, increasing intensities of tropical cyclones, changes in degree and level of tidal inundations and rainfall regime, and reduced upstream fresh water flow has resulted to start forest degradation in this important landscape though changes in habitat. Generally, mangroves are sensitive to habitat change as a response of habitat change. Contrary, we are not aware of any ground survey or high spatial resolution based remote sensing based studies or combination of both for the case of the Sundarbans which knowledge is essential for policy and academic ground. This study attempts to fill this knowledge gap where we seeks to assess vegetation dynamics over a period of 87 years (1926-2012) and measure forest deteriorations resulted from increasing intensity and frequency of tropical cyclone impacts on vegetation cover, and status of top-dying due to increasing salinity level in tidal water of the Sundarbans. These objectives were attained through integration of field survey of 113 permanent sample plots, from November 2011 to February 2012, high-quality forest inventory data from 1926 through to 1998, and satellite imagery from 1975 to 2012. Using Landsat MSS, TM and ETM+ and imagery from the RapidEye, GeoEye and IKONOS; these has mapped the Sundarbans by radiometrically normalised, imagery from 1972, 1977, 1989, 2000, 2010 and 2011 with ground verification. The results showed a significant (R² = 9.89) decline of freshwater habitat i.e. Heritiera fomes dominated forests from 1926 through to 2012 while saline habitat i.e. Excoecaria agallocha dominated forests showed significant increase (R²=9.78) despite the inherent dominance of those two forest types has been maintained. Both density and height of vegetation was found related to level of salinity: low in the northern part and high in the southern part. Per hectare basis basal area and stocking has found more consistent at middle part (core areas) of the Sundarbaans which showed a decline towards edges facing sea, inland and highly saline zone areas. Comparative analysis of field survey and secondary data shows significant association with forest types, reduced supply from upstream freshwater, salinity intrusion, salinity induced diseases (top-dying of major species), illegal logging, and eventually cyclones. Testing with different spatial resolution based multi-sensor image analysis integrated with historical and ground survey data explained similar trends with filed plot statistics and spatial trends. Considering limitations of optical sensors with lack of consistent cloud free imagery, few pilot scale study in the Sundarbans has showed potential scope for Radar imagery but wall to wall mapping and forest habitat mapping could provide more insight about on-going processes of this globally significant ecosystem. Again, biomass and carbon assessment based monitoring for this biologically diverse and carbon rich tropical ecosystem needs further studies to measure carbon flux which will certainly add value to the United Nations-led Reduced Emission from Deforestation and Forest Degradation (REDD+) monitoring, reporting and verifications (MRVs) mechanisms and various on-going global and national forest mapping and monitoring programmes. Therefore, we understand the need to investigate the inherent advantages and explore potentiality of SAR imagery from Sentinel sensors by European Space Agency (ESA) for the Sundarbans landscape.