Modelling hydraulic and physical features of riverine fish habitats

In the UK, 43% of river habitats are classified as severely modified and there is continued pressure for further development (abstractions, flood protection etc). Typically, the hydrodynamic environment of a river is enriched by longitudinal and lateral variations in bathymetry, water depth and bed armour material, which promotes the generation of secondary currents, turbulence, and variation in local flow velocity and flow depth. These three-dimensional hydraulic features constitute a variety of habitat conditions essential for freshwater invertebrate and fish diversity. However, hydraulic and hydro-ecological models (e.g. PHABSIM, MesoHABSIM) that define changes in physical habitat availability for target fish species given a change in river flow channel geometry typically use a one-dimensional hydraulic approach and generate only crude velocity and water level predictions in rivers of complex bathymetry. In adapting existing riverine environments to meet the Water Framework Directive, it is essential to maintain the natural character, stability and discharge capacity whilst enhancing recolonisation processes, but in practice the optimal conditions for fish are poorly classified.

Existing protocols for evaluating fish swimming performance are based on the area mean water velocity and are conducted in swimming chambers with smooth bed and wall boundaries. No consideration is given to the position of the fish within the water column and magnitude of the turbulence (e.g. the temporal variation of the velocity) so fish swimming performance and stamina could be currently drastically over-estimated for demersal (bottom dwelling) fish and under-estimated for pelagic (water column) fish. Furthermore, the turbulence signature (magnitude of turbulence in relation to the time-averaged velocity) produced in a natural river (with variable bathymetry; rough bed and banks, bedforms, vegetation, islands and man-made structures) will be distinctly different to that produced in an idealised swimming chamber.

At Cardiff, the student will collect detailed empirical data on the hydraulic conditions of a flume under different flow conditions with various habitat modifications to assess how this affects the swimming performance of demersal and benthic fish. The student will develop a habitat modelling approach whereby a generic open source 3-D Computational Fluids Dynamic code (e.g. FLUENT, Pheonics, OpenFOAM, TELEMAC) is linked to a programming and post-processing software environment, such as MATLAB, in which new habitat suitability rules are implemented based on the new data and existing ecological models.