Instream Flow Predictions from Frequency Vs. Bioenergetic-Based Habitat Suitability Curves

Predictions of optimal flows using PHABSIM are extremely sensitive to the shape of velocity and depth habitat suitability curves (HSCs).  HSCs therefore need to accurately reflect the fitness (or population level) consequences of habitat use for model predictions to be meaningful.  Frequency-based habitat suitability curves are based on the observed frequency of use of different velocity and depth micro-habitats by the target species; however, HSCs may be misleading for territorial taxa if competition by dominant fish displaces subordinates into poor quality (e.g. low-velocity) habitat at high densities.  HSCs based on estimates of density-independent growth rate potential from bioenergetic models offers an alternative to frequency-based HSCs. We compared instream flow predictions generated using frequency vs. bioenergetic-based HSCs for juvenile coho to known smolt production to evaluate the potential for bias in estimation of optimal juvenile coho rearing flows. Frequency-based HSCs inflated habitat quality at low velocities and greatly underestimated the consequences of summer low-flow reduction for smolt production relative to the bioenergetic-based HSCs.  Results demonstrate the potential for serious underestimation of optimal flows using frequency-based HSCs, and support the additional use of generic bioenergetic HSCs for more robust flow assessments, particularly for juveniles of pool-rearing salmonids that display strong territoriality.