Impact of Landuse Practices on Ecosystem Diversity using Remotely Sensed Data.

Abstract

Owing to the concerns over the increased emissions of GHGs significant efforts are being made in Canada to produce fuels from biomass to reduce this impact. Such production, however, requires land to produce the biomass with the associated risk of impacts on biodiversity. As a step towards examining this impact, an assessment methodology was developed that integrates statistical modeling, remote sensing, and field studies to characterize landuse changes with respect to ecosystem diversity. This does not address other aspects of biodiversity such as genetic or species diversity. This study investigates the applicability of vegetation fraction derived from spectral reflectance as an indicator of biodiversity. Indices based on regression models between each landuse types and reflectance values from Landsat Thematic Mapper. The Normalized Difference Vegetation Index (NDVI) is used for measurement of green biomass and ecosystem diversity. NDVI values for diversity assessment and monitoring are used to map vegetation land cover with a special emphasis on agriculture. The vegetation dynamics in the context of different land cover types lead to variations in the spectral radiance in ecosystem diversity indicator values. Preliminary results applied to an example from Southeastern Ontario indicated a positive correlation between NDVI values and vegetation diversity.