The effectiveness of a fish-friendly pumping station for critically endangered European eel; An assessment using live eels, fish-mounted sensors and passive sensors

Fish-friendly pumps may provide safe downstream passage for seaward migrating silver European eel, Anguilla anguilla , at pumping stations but further research into their effectiveness is urgently required. This study uniquely combined health assessments of live fish, including behaviour, delayed (48-h) mortality and post-mortem post-passage, with fish-mounted (fish backpack sensors; FBS) and passive (barotrauma detection system; BDS) sensors that passed through a fish-friendly axial flow pump in situ at a real-world pumping station to comprehensively quantify the effectiveness. Nineteen eels (including two control eels released directly into the recapture net) were recaptured and there was no immediate or delayed ( n = 5, including one control eel) mortality or behavioural change, nor were there any major external or internal injuries. Minor external injuries were newly present on most eels (84.6 %) that underwent a post-mortem post-passage ( n = 13) but were equivalent to the types of pathologies recorded pre-passage and were considered unlikely to affect function or survival. A previously defined acceleration threshold for fish vertebrae damage (>50 m/s 2 ) was exceeded by all FBS and BDS (up to 66.48 m/s 2 ), predominantly (89.15 %) after passing through the pump, but no eels showed any evidence of spinal injury or vertebral deformity, indicating that this threshold may be overly conservative for eels. Extremes of pressure (i.e., nadir) and changes in pressure (i.e., pressure rate of change (PRC) and log ratio of pressure change (LRPC)) recorded by FBS and BDS were well below thresholds known to induce injury in American eels ( Anguilla rostrata ) of similar length and no pressure-related injuries were found during post-mortems. Indeed, it illustrated the importance of using sensors alongside live fish to further our understanding of species-specific injury thresholds. Overall, this study demonstrated the effectiveness of a newly installed axial flow fish-friendly pump (under ‘normal’ operating conditions), using live fish and two types of sensors to quantify conditions during passage. • Fish-friendliness assessed for entire real-world pumping station, not just the pump. • Uniquely combined live fish with fish-mounted and passive sensors. • No immediate or delayed (48-h) eel mortality, behavioural change or major injury. • All sensors exceeded fish damage acceleration threshold, mostly after pump passage. • Pressure extremes and changes recorded by sensors below known eel injury thresholds.