Comparison of Conventional, Rake, and Sonar‐Based Biophysical Habitat Measurements in a Shallow Ontario River

ABSTRACT Knowledge of habitat availability is critically important for the management and recovery of freshwater species. Quantifying habitat availability often requires fine‐scale sampling at point‐based locations across a large geographic extent, which can be laboursome. Technological advancements and lower costs make sonar approaches increasingly attractive for quantifying habitat availability, but uncertainty remains about their accuracy and precision in shallow ecosystems. The objective of this study was to determine how well sonar‐derived indices aligned with conventional point‐based measurements for habitat variables that influence the distribution and life‐history functions of freshwater biota (water depth, submerged aquatic macrophyte volume, substrate composition) in a shallow river in Ontario, Canada. Sonar‐derived measurements indicated that 98.5% of the surveyed area was < 2 m depth (mean = 0.74 m ± 0.33 SD) while kriging interpolation of conventional depth measurements indicated the entire survey area was < 2 m depth (mean = 0.97 m ± 0.12 SD). A regression slope of conventional and sonar‐derived depth measurements was not significantly different than one ( = −0.33, = 0.74) but conventional measurements were greater than sonar‐derived estimates at 88% of the sites, reflecting consistent differences in identifying the bed elevation. Weak ( < 0.4) positive correlations were observed between sonar‐derived hardness and biovolume with conventional measurements. Differences between conventional and sonar‐derived metrics related to measurement scale and the error associated with both conventional and sonar results. Overall, decisions about which approach to use for measuring habitat availability will depend on the specificity required for management decisions.