A general analytical approach for assessing the effects of hydroclimatic variability on fish habitat

Abstract

We propose a novel analytical approach that provides a simple, integrated tool for assessing the effects of hydroclimatically-driven flow regime variations on fish habitat. Average habitat quality metrics can be predicted effectively by an analytical equation. This is the result of the integration of two functions describing (a) the flow regime (the frequency distribution of discharge) and (b) the relationship between discharge and habitat quality. We applied this approach as a “proof of concept” to a simple model of velocity thresholds for juvenile salmon fry. The flow regime was described by a gamma distribution with physically meaningful parameters derived using a mechanistic-stochastic approach that explicitly links the flow regime to hydro-meteorological conditions and catchment wetness. Under different flow conditions, habitat quality was estimated using hydraulic habitat models. An empirical analytical function was then used to describe the relationship between specific discharge and habitat quality. Six river channel reaches characterized by different bed geometry and morphology were investigated. Four reaches are located in the Girnock burn, a natural upland river in the North of Scotland, while the remaining two reaches are located in the River Lyon in the central Highlands of Scotland that is heavily regulated for hydropower production. Results show that the stochastic model and the analytical function applied, each containing three parameters, successfully described Girnock (in both dry and wet years) and Lyon (naturalised) flow regimes and captured the relationship between available habitat and specific discharge reasonably well. For each site we obtained heat maps linking climatic and landscape features to bi-monthly average available habitat. These maps can provide clear generalised insights into the potential effects of hydroclimatic variations on metrics of fish habitat and have potential for wider applications using more complex habitat model.