Accuracy of coarse-resolution protocols for assessing fish passability at river infrastructure

Abstract

River infrastructure can present a major barrier to fish movement. While mitigation can be achieved through infrastructure removal or modification, e.g. construction of fish passes, prioritising this work to maximise ecological benefit within budgetary and other constraints remains a substantial management challenge. Several coarse-resolution rapid barrier assessment protocols have been developed to estimate fish passability at river infrastructure to aid decision making in relation to catchment scale barrier management. The outputs of these protocols are rarely validated against empirical data, such as that provided by telemetry or other methods of tracking fish, limiting confidence in whether they provide realistic estimates of fish passability. In this study, the accuracy of two barrier passability assessment protocols yet to be validated against empirical data, SNIFFER and ICE, developed in the UK and France, respectively, was assessed by: 1) collating available empirical multi-species fish passage data at river infrastructure, 2) undertaking field surveys at each structure to quantify SNIFFER and ICE passability scores, and 3) comparing the fish passage data with the estimated passability scores. Fish passage data were obtained for four species (Salmo trutta, Cottus gobio, Lampetra fluviatilis and Thymallus thymallus) and five barrier types (sloped weir, culvert, rock ramp, nature-like bypass, and some classified as ‘complex’) at thirteen sites in England. Both protocols suggest these barriers are a major impediment to upstream moving fish as a classification of ‘impassable’ was the most common. However, agreement between protocols was low for barriers considered ‘high impact’, ‘low impact’ or ‘easily passable’. When compared with empirical fish passage data, there was a positive relationship with the passability scores predicted by SNIFFER, but the protocol tended to be conservative. There was no relationship between the empirical fish passage data and the passability scores predicted by ICE, although field surveys were undertaken outside of the recommended discharge range which may have influenced accuracy. This study provides the first partial validation of two barrier passability assessment protocols and suggests that further detailed validation is needed to calibrate the passability scores and enhance confidence in the planning tools (e.g. optimisation models) that utilise their outputs.