Abstract
Understanding the environmental effects of marine renewable energy devices is important for ensuring the responsible development of this new industry and is particularly relevant in ecologically sensitive regions. Minas Passage, Bay of Fundy, Canada, is a highly sought-after location for the development of tidal stream energy and is characterized by the world’s highest tides (15-m range) and tidal flow speeds exceeding 5 m/s [1]. Efficient extraction of energy from these flows could produce over 2.5 GW of electricity each tidal cycle [2]. However, Minas Basin hosts at least 85 fish species, including diadromous and marine fishes for which Minas Passage serves as an important migratory corridor [3]. This includes species protected under Canada’s Species at Risk Act, of cultural relevance to First Nations communities, and comprising important commercial and recreational fisheries.
Collision between marine species and rotating turbine blades is perceived to be the most direct environmental impact of tidal stream devices [4]. Risk of collision is typically framed as encounter risk, and is measured using models derived from predator-prey interactions that assess the probability of an animal entering the turbine’s area of effect [5]. Globally, the majority of research focused on interactions between marine animals and tidal power devices has focused on marine mammals [6]. Direct observation of fish presence is difficult in extreme environments such as Minas Passage, particular given the need for species-specific information on potential encounter risk. Acoustic telemetry, in which transmitters carried by fishes produce ultrasonic signals that can be detected by deployed receivers, can provide presence/absence data for the development of species-specific spatiotemporal distribution models and ultimately encounter rate models.
Acoustic telemetry has become a standard method in fisheries and marine ecology studies, and collaborations with existing projects are possible through acoustic telemetry networks [7]. Acoustic tag detection locations can be matched in space and time with environmental data, which enables identification of associations between environmental conditions and species presence/absence and allows for predictive species distribution modelling [8]. Bangley et al. [9] combined the timing and detection location of acoustically tagged fish with oceanographic variables and used a boosted regression tree (BRT) modelling approach to develop a predictive species distribution model (SDM) for striped bass (Morone saxatilis) in Minas Passage the illustrated species presence probability at each tide stage.
A potential caveat of using acoustic telemetry in Minas Passage is that the extreme current velocities can have a significant effect on the detectability of acoustic signal, particularly from the 69-kHz transmitters used in that study [10]. Due to the longer time intervals typical of 69kHz transmitters (usually 1-2 min), there may only be time for a single ping to reach the receiver while the tag is within detection range, and in some cases a tagged fish may be moved through the receiver’s detection range before the next transmission is made. Here, we review the methods developed by Bangley et al. [9] and expand on this previous effort by incorporating data on tag detection efficiency to account for these environmental effects. This allows us to make conservative and realistic predictions of striped bass presence probability even in conditions where tag detectability is reduced. This is a crucial first step towards the development of statistically robust encounter rate models for quantifying the risk of tidal stream turbines to fish.