A remote sensing-GIS based approach for assessment of chinook salmon rearing habitat in the Unuk river floodplain

Abstract

Remote sensing offers an alternative method to managers in mapping and monitoring the habitat within large rivers. Large rivers are not accommodating for traditional (foot) fish habitat surveys due to their size and typically complex habitat. This study investigates the use of digital aerial photos and thermal infrared images acquired in spring 2003-2005 to map and quantify juvenile chinook salmon (Oncorhynchus tshawytscha) habitat in a 12-river km section of the Unuk River floodplain in Southeast Alaska. Images were processed and analyzed to produce a fluvial landscape classification (7 landcover classes with an overall classification accuracy of ~84%) using a combination of aerial and thermal images. Change detection of large woody debris (LWD) was also examined and revealed both quantitative and distributional changes during the 3 years. A GIS-based habitat suitability analysis was used to identify potential chinook salmon rearing habitats including: river channel edges, sloughs, braids, pools associated with LWD and primary river channels. Overall 77.82 hectares of potentially medium/high chinook rearing habitats were identified. Results from this study provide a promising foundation towards mapping and monitoring salmon habitat in large river systems for purposes of protection, conservation and monitoring to ensure sustainable stocks of salmon.