Hydropower is playing a crucial part in the transition away from fossil fuels by acting as a buffer against the erratic availability of other renewable energy sources like wind or solar. Hydropower turbines may be started and stopped according to fluctuations in electricity demand, therefore stabilizing the power grid. However, this operational flexibility leads to abrupt and frequent variations in water discharge from reservoir hydroelectric power stations. Known as hydro-peaking, this phenomenon creates artificial flow conditions for fish and other organisms caused by the rapid increase or decrease in the release of operating water.
Hydro-peaking is two processes. During "up-ramping," fish are pushed downstream when the discharge first increases quickly. Because they are less equipped to handle a rise in flow velocities, larvae and young fish experience this effect more strongly. Conversely, "down-ramping" is the rapid fall in discharge that follows peak flow. During down-ramping, fish may be injured or die if they become stranded on dewatered shallow areas or when they lose connection to the main channel.
To shed light on how fish respond to hydro-peaking, scientists at ETZ University (Zurich, Switzerland) evaluated the fine movements of trout exposed to up- and down-ramping. In order to replicate river conditions, the scientists created an experimental arena built from clear glass. This arena was laterally inclined, and filled with water, cobble stones, and a gravel layer. Water was continuously pumped into an upstream tank via a recirculation system, where it then flowed into the open-channel flume and finally into a downstream tank. The researchers were able to control the discharge in the flume by adjusting the pump rate, reproducing hydro-peaking.
The arena was illuminated with infrared light from underneath using four LED strips emitting at wavelengths above 750 nm. The IR-light penetrated the arena's glass bottom, a diffusor material, a color foil and the gravel layer. IR-light was used as the light source since it is not visible for many fish species. To imitate daylight, a single white-light LED strip was mounted above the flume.
Once complete, trout were placed inside the arena where they were exposed to the different hydro-peaking treatments. To avoid possible effects of group size on behavior, and because trout are territorial, experiments were conducted with individual fish and each was used only once.
Imaging System
Images of the fish were recorded using two synchronized Mikrotron EoSens CL monochrome cameras equipped with filters to record only infrared light. Both cameras were mounted above the arena at a height of 1.7 meters. Fish movements were acquired at a rate of 15 frames per second and a resolution of 1280 × 1024 pixels. To prepare images for tracking the fish, custom-made Python scripts were developed that merged the two synchronized camera views into one image, which were then rotated and cropped. The goal of this pre-processing was to homogenize image size, orientation, and the origin of the coordinate system across experimental runs. The results were manually reviewed for all videos and tracking errors corrected.
Conclusion
After analyzing the images, ETZ scientists determined that the fish deployed multiple behavioral responses to navigate rapid changes in hydrodynamic conditions. To cope with high flows, fish swiftly relocated laterally and reduced exploratory behavior. Fish also occupied areas downstream of cobble where the lift and wake forces reduce hydrodynamic drag. In contrast, increased down-ramping rates resulted in shorter response times and increased relocation speed for the fish. As shallow areas went dry, fish abandoned these areas to avoid being stranded.
The scientists believe that their results will help to deepen knowledge of the ecological impacts of hydro-peaking. Understanding the specific habitat features that provide effective hydrodynamic shelter can inform river management and conservation strategies to mitigate the impacts of hydro-peaking on aquatic ecosystems.