A framework for using small Unmanned Aircraft Systems (sUASs) and SfM photogrammetry to detect salmonid redds

Salmonid populations are widely distributed globally and are of economic, cultural and ecological importance. Evidence suggests that they are in decline in many parts of the world and one of a number of hypotheses for their decline is the degradation of spawning habitat. Knowledge of spawning sites and their evolution through time is a means of estimating regional population dynamics and sizes. Traditionally, spawning sites have been identified visually. However, this may not allow a precise quantification of the real extent of salmonid reproduction and of its evolution through time (i.e. within the spawning season). This paper develops a framework for using small Unmanned Aircraft Systems (sUASs) and Structure from Motion (SfM) photogrammetry to detect salmonid redds, the nests that are the distinctive footprint of spawning, through analysis of inter-epoch Digital Elevation Models (i.e. DEMs of Difference). SfM-derived DEMs of Difference are an effective tool to investigate spawning because of the distinctive ellipsoidal erosion-deposition pattern of salmonid redds, which discriminates them from other stream-bed elevation changes. The method detects more redds (e.g. those covered by algae or biofilm) compared with classical visual observation, allowing for a better and more rigorous detection of spawning grounds. SfM photogrammetry also provides additional information relevant to understanding salmonid spawning, including redd-density and probable female lengths, without disturbance of the spawning sites.