Importance of partial barriers and temporal variation in flow when modelling connectivity in fragmented river systems

(1) The potential for catchment-scale connectivity modelling to help plan the restoration of connectivity in fragmented river systems is not yet well understood. In the present study the importance of two interrelated aspects of such modelling in determining predictions of connectivity are explored: (1) uncertainty in the passability of partial barriers (such as fish passes) and how the passabilities of series of partial barriers combine, and (2) temporal variation in connectivity due to flow.

(2) Connectivity for Atlantic salmon (Salmo salar L.) and European perch (Perca fluviatilis L.) are modelled under alternative restoration strategies in the heavily impounded Don Catchment UK using two different methods for simulating the combined passability of series of partial barriers. Catchment-scale hydraulic and connectivity modelling were integrated using a novel method to account for the effect of flow on connectivity, achieved by consideration of flow-fish pass efficiency relationships and the treatment of gaps between habitat patches as partial barrier.

(3) Modelled connectivity is very sensitive to uncertainty in barrier passability and the method used to simulate the combined passability of series of partial barriers. Flow also has a strong and complex effect on connectivity, with predicted temporal patterns being particularly dependent on how the combined impact of series of barriers is modelled. The sensitivity of the modelling constrains its capacity to predict the outcome of alternative connectivity restoration strategies. Nevertheless it does serve as a tool to think critically about connectivity restoration. If applied thoughtfully in full awareness of its limitations it can still be used assist in the planning and appraisal of alternative restoration options.

(4) The modelling also provides a number of important practical insights. It shows that series of fish passes may be ineffective unless they operate at very high efficiencies. Small changes to flow-fish pass efficiency relationships can have a large effect on temporal patterns in connectivity. Overall fish pass efficiency is comprised of attraction and passage efficiencies which may differ in the extent to which they are determined by random processes. This likely has significant implications for the nature of the combined passability of series of fish passes.