Measuring stream habitat conditions: Can remote sensing substitute for field data?

Abstract

The ongoing encroachment of agricultural activities into natural areas is a growing problem for the ecological condition of streams. Stream ecological condition is best measured using both biotic and abiotic parameters that reflect different channel, riparian zone and catchment aspects. Multiple physical-chemical measures of water quality have long been widely used to represent the environmental conditions of water bodies. More recently, physical habitat structure, catchment land use and land cover have been employed to better understand water body conditions. Both water quality and physical habitat structure metrics are usually measured in the field and often have strong predictive power to analyze biological assemblage conditions. On the other hand, remote sensing of catchment land use and land cover provide relatively low-cost environmental information at large spatial extents, minimizing the need for fieldwork and reducing analytical time. Given these considerations, our aim in the present study was to evaluate the degree to which stream environmental conditions could be measured reliably via remote sensing. In particular, we assessed whether a remote sensing index (Normalized Difference Vegetation Index) and land use can be used as reliable surrogates for site habitat condition, channel dimensions, and water quality. We found that our remote sensing variables were not sufficient for predicting stream water quality or habitat structure. Therefore, we recommend using remote sensing indicators only when it is impossible to measure water quality and habitat structure in the field directly.