Basin‐scale methods for predicting salmonid spawning habitat via grain size and riffle spacing,tested in a California coastal drainage

Basin‐scale predictive geomorphic models for river characteristics, particularly grain size, can aid in salmonid habitat identification. However, these basin‐scale methods are largely untested with actual habitat usage data. Here, we develop and test an approach for predicting grain size distributions from high resolution LiDAR (Light Detection and Ranging)‐derived topographic data for a 77 km² watershed along the central California Coast. This approach improves on previous efforts in that it predicts the full grain size distribution and incorporates an empirically calibrated shear stress partitioning factor. The predicted grain size distributions are used to calculate the fraction of the bed area movable by spawning fish. We then compare the ‘movable fraction’ with 7 years of observed spawning data. We find that predicted movable fraction explains the paucity of spawning in the upper reaches of the study drainage, but does not explain variation along the mainstem. In search of another morphologic characteristic that may help explain the variation within the mainstem, we measure riffle density, a proxy for physical habitat complexity. We find that field surveys of riffle density explain 64% of the variation in spawning in these mainstem reaches, suggesting that within reaches of appropriate sized gravel, spawning density is related to riffle density. Because riffle density varies systematically with channel width, predicting riffle spacing is straightforward with LiDAR data. Taken together, these findings demonstrate the efficacy of basin‐scale spawning habitat predictions made using high‐resolution digital elevation models. Copyright © 2016 John Wiley & Sons, Ltd.