Aquaculture - Acoustic Deterrent Device (ADD) use: parliamentary report

Chapter 2. Evidence of impact on marine mammals from acoustic deterrent device use

1. Introduction

There is growing concern that both target species (e.g., seals) and non-target species (e.g., whales, dolphins and porpoises) could be disturbed and/or at risk of auditory injury through exposure to noise from Acoustic Deterrent Devices (ADDs) used at finfish farms. The currently available scientific evidence of impacts to marine mammals in terms of disturbance, temporary hearing change (TTS) and permanent hearing change (PTS) is reviewed and summarised here. This review focuses on ADDs that have been, or are known to be, used in the aquaculture industry. Where an ADD is used in a different industry and there is potential for the device to be used in aquaculture (i.e. if the device could be employed to prevent depredation by seals), this evidence has also been included for completeness. Where a custom built ADD with similar acoustic properties to a known ADD has been reviewed, this has been highlighted.

Disturbance to marine mammals is difficult to define and quantify, but is often regarded as a change in the normal behaviour of an animal. Examples of disturbance include changes in distribution, breathing, breeding, nursing, feeding and/or resting as well as disruption to communication. Many of these responses could cause an increase in energy consumption, which could lead to increased vulnerability of an individual to predators or physical stress. This in turn may reduce their survival or reproductive success, potentially resulting in population-level effects. Currently, there is not an accepted sound threshold level for behavioural disturbance to marine mammals.

Marine mammals exposed to intense sound, either instantaneously or over time, have the potential to exhibit reduced hearing sensitivity (termed "threshold shift"). This may be at a particular frequency or over a range of frequencies. The frequency at which it occurs will determine the extent to which an animal suffers detrimental consequences, depending on how important those frequencies are to the animal for communication or foraging. Hearing changes that recover after exposure are termed a temporary threshold shift (TTS) in hearing. Hearing changes that do not recover are termed a permanent threshold shift (PTS) in hearing, which is considered to be an auditory injury (Southall et al., 2007). Physical harm to a marine mammal, such as damage to the physical structures of the hearing system by extremely loud noises, would also be considered to be an injury.

TTS or PTS can occur by two means; instantaneously, through exposure to sudden onset loud noise (e.g. from pile driving, seismic surveys or explosives), or cumulatively through repeated exposure of sound over time. Cumulative TTS or PTS result through a combination of the level of the noise and the duration of time that the animal is exposed to it. This is assessed with a metric called the sound exposure level (SEL), which is the accumulated sound energy an animal is exposed to over time (usually calculated over 24 hours). The sound energy is weighted for different species groups because they each have different hearing sensitivities over a range of frequencies. The calculated SEL is then compared against internationally recognised threshold levels. The sound levels produced by ADDs would not, at recommended operating source levels, pose a risk of instantaneous PTS. However, due to the ongoing nature of sounds produced by ADDs, there is a risk of PTS through cumulative exposure.

Studies on potential impacts of ADDs on non-target species and their efficacy in relation to reducing seal depredation are few in number and show inconsistent results. It is important to note that a large number of factors are potentially influential in the responses of marine mammals to ADDs. These include factors relating to the ADDs themselves (e.g., source level, duty cycle, number of devices used, frequency content), factors relating to the animals (e.g., context, body condition, age of individual, experience of individual) and factors relating to the manner in which the local environment affects sound propagation through the water (e.g. depth, presence of land, water temperature, sea state). Studies into the effects of ADDs typically vary in location, the manner in which the device is used and whether there is motivation (e.g. the presence or absence of prey) for animals to remain within an area, which makes it challenging to assess and compare the evidence base in some cases.

The signals from most ADDs are not a constant tone, but rather a complex combination of frequency-specific pulses.

The following sections cover disturbance, TTS, and PTS, across the four different marine mammal hearing groups. These are low frequency cetaceans (e.g., minke whale, Balaenoptera acutorostrata) high frequency cetaceans (e.g. bottlenose dolphin, Tursiops truncatus), very high frequency cetaceans (e.g. harbour porpoise, Phocoena phocoena) and seals. For seals, disturbance is incorporated in a section that reviews the evidence for efficacy of ADDs because the purpose of ADDs used around finfish farms is typically to disturb seals to keep them away from the fish. Following this, a section outlining the key knowledge gaps has been provided.

2. Summary of Evidence

This section reviews currently available scientific evidence of impacts to marine mammals in terms of disturbance, TTS and PTS. To note, a number of previous studies have also summarised the existing evidence of ADD effectiveness, including potential impacts on non-target species (e.g., Gordon and Northridge, 2002; Coram et al., 2014; Sparling et al., 2016; McGarry et al., 2020).

More detailed information on the studies reviewed here can be found in the Appendix in Tables A1-1 to A1-11. Only the ADD manufacturer is stated in the following review and not the specific model type because the studies reviewed do not adopt a consistent approach in naming devices. Further information on the device type used in each of the studies reviewed can be found in the relevant tables within the Appendix.

Disturbance to cetaceans

As the most abundant and sensitive non-target species, harbour porpoises have been the focus of a number of studies reporting adverse behavioural responses to sound from ADDs (Table 2-3). Captive experiments with harbour porpoise have demonstrated the types of response that may be seen as a result of exposure to different ADDs (e.g., Kastelein et al., 2010 & 2015a). These behavioural responses included increases in surfacing rate (closely linked to respiration rate), respiration force, moving away from the sound source, swim speed and jumping behaviour.

Within studies on free-ranging harbour porpoise, experimental design, source levels, topography and ambient noise levels all vary, which makes direct comparisons difficult. The majority of field trials with wild harbour porpoises have shown displacement and/or exclusion as a result of exposure to the sound either from, or resembling that of, Airmar ADDs (e.g., Johnston 2002; Kyhn et al., 2015; Northridge et al., 2010; Olesuik et al. 2002) or Lofitech ADDs (e.g. Brandt et al., 2012 & 2013; Dähne et al., 2017; Mikkelsen et al., 2017). Lofitech ADDs have not been reported to be in use at Scottish finfish farms (see Chapter 3 of this report), but are used in other industries to deter animals from activities that could injure them (e.g. blasting and impact pile driving). The one published field study using a Terecos device showed reduced detections along a distance gradient from the device, but overall reported a weak or minimal response from harbour porpoise (Northridge et al., 2013). However, even as regards the devices for which there are multiple studies describing the effects on harbour porpoise, there is considerable variation in the frequency and scale of responses, both within and between studies.

In one field study in the Inner Hebrides, using echolocation click detectors (T-PODs), harbour porpoises were detected feeding within 200 m of ten active Airmar ADDs that had been in use at the site for several months (Northridge et al., 2010), with the authors concluding that these devices did not completely exclude all harbour porpoise from the vicinity. In a neighbouring area, data gathered using towed passive acoustic monitoring (PAM) were used to compare harbour porpoise distribution in 2008, when an Airmar ADD was installed, to the two previous years. The number of acoustic detections in the area was significantly lower in 2008 (following the introduction of the ADD) than 2006-2007, and there were no porpoises detected within 4.3 km of the ADD site in 2008 (Northridge et al., 2010). However, this was not a controlled study and there is the potential that other factors, such as prey availability affected harbour porpoise use of the site.

A field study in Denmark ran two consecutive trials using Airmar ADDs. The first involved periodic activation of ADD sound over 47 days, where the devices were activated and deactivated for periods of between two and nine days (Kyhn et al., 2015). After a recovery period of 40 days, the same area was then exposed to ADD sound continuously for 28 days. For the periodic exposure trial, the study reported a strong aversive response by porpoises (displacement away from the ADDs) to the first and second exposures, but this response was reduced in subsequent periodic exposures. Conversely, when continuously exposed to the ADD sound there was no evidence of a reduced response over the 28 day exposure period (i.e., patterns in displacement were maintained over time). The study reported that both periodic and continuous use of these ADDs led to areas of porpoise displacement no greater than 2.5 km. Interpretation of these findings is challenging; the decreasing response over time may be the result of harbour porpoise no longer considering the ADD to be a threat, or may be because they were compelled to use the area due to a lack of foraging opportunities elsewhere.

Benjamins et al. (2018) developed an example ADD signal, based upon those generated by several ADDs available on the market, including Airmar, Lofitech and Ace Aquatec. They produced both a low and high frequency version of the signal (1-2kHz and 8-18kHz respectively) to represent the current ADD market (high frequency) and the market trend towards lower frequency devices, intended to avoid disturbance to harbour porpoise. They observed responses by harbour porpoise to both signals, using a sound source level that was on average approximately 20-35 dB re 1 μPa lower than the commercially available ADDs. They reported that porpoise detections were substantially reduced within 1 km of the sound source with no significant difference in detections between the low and high frequency playbacks (1-2 kHz and 8-18 kHz, respectively), indicating that lower frequency devices might not reduce disturbance to harbour porpoise.

For the other two cetacean hearing groups, low frequency and high frequency cetaceans, the evidence for disturbance effects is less clear, with very few controlled scientific studies published. For low frequency cetaceans much of the evidence is either anecdotal, collected opportunistically as a by-product of research on other species, or suffers from very small sample sizes (e.g., Fairbairns et al., 1994; Götz & Janik, 2016). The two exceptions to this reported differing responses for two species of baleen whale when exposed to sounds from the Lofitech ADD. This device was reported to have deterred minke whales (McGarry et al., 2017) but had no observable impact on humpback whale (Megaptera novaeangliae) behaviour (Basran et al., 2020). The methodologies used in these two experimental exposure trials, despite having very small sample sizes (7 and 15 animals, respectively), were very similar, however the source levels differed considerably. The measured source level was 10 dB re 1 μPa @ 1 m (RMS) lower in the Basran et al. (2020) study compared with the McGarry et al. (2017) study, which is likely to have affected the responses observed.

For high frequency cetaceans, the evidence base for disturbance from active ADDs is also lacking and the few studies that do exist mainly concern Airmar ADDs. For example, observations from ongoing long-term abundance studies (rather than controlled field trails) suggest that Airmar ADDs have caused habitat exclusion of killer whales (Orcinus orca, Morton and Symonds, 2002) and Pacific white-sided dolphins (Lagenorhynchus obliquidens, Morton, 2000). Desk-based modelling suggests that the sounds from Airmar ADDs, as well as Terecos and Ace Aquatec ADDs, are audible to whales, dolphins, porpoises and seals at ranges up to tens of kilometres, depending on ambient underwater noise levels (Todd et al., 2019). However, these figures should be interpreted with a degree of caution, as the modelling methodology used was relatively simple and did not include environmental factors, which would lead to greater attenuation of the sound at distance. It is therefore likely that the distances presented in Todd et al. (2019) may be overestimates.

The GenusWave device, referred to as a Target Acoustic Startle Technology (TAST) device, which is targeted for seal hearing at frequencies lower than those to which harbour porpoises are sensitive, has been the focus of two published field studies (Götz and Janik, 2015 & 2016). These studies concluded that there was a negligible effect of the device use on porpoise abundance and distribution. Given that these devices operate at lower frequencies, there is a notable knowledge gap regarding the potential for disturbance, temporary hearing changes (TTS) and cumulative permanent hearing changes (PTS) in low frequency cetaceans.

Temporary Threshold Shift (TTS) in marine mammal hearing

Temporary Threshold Shift (TTS) involves a reduction in an animal's hearing sensitivity at particular frequencies, from which the animal subsequently recovers. The likelihood of TTS being induced in a marine mammal as a result of exposure to noise from ADDs is heavily dependent upon the animal's behaviour. If the animal moves away from the noise then TTS is unlikely to occur. However, if the animal remains within the area, then there is the possibility of longer durations of noise exposure which can accumulate to produce a TTS. Animals may choose to stay in areas with higher noise levels depending upon motivation, such as the availability of prey within the area.

There are few published studies detailing instances of a TTS in marine mammal hearing as a result of sound emitted by ADDs, due to the difficulty in measuring this physiological effect in free ranging animals. The evidence compiled for TTS in this report constitutes mainly desk-based underwater noise modelling exercises. The modelled scenarios involve instantaneous exposure or cumulative exposure scenarios, and estimate the time and/or distance at which a marine mammal would be predicted to experience TTS due to ADD noise exposure. As the studies use different modelling techniques, criteria for assuming TTS onset, and weightings to account for animals' hearing ability, it is difficult to compare the results.

Götz and Janik (2013) estimated the potential for TTS in marine mammals through a modelling exercise, assuming that an animal remained stationary for a period of time in the vicinity of three different ADD types; Airmar, Lofitech and Terecos. Distances and durations of exposure at which TTS onset were predicted to occur for a range of marine mammal species were presented. The study incorporated a number of criteria for assessing sound exposure levels that may induce TTS. Some of these were internationally recognised thresholds (e.g., Southall et al., 2007; Lucke et al., 2009), while others were unique to this study. A novel method of auditory weighting was applied to the sound sources based on the most sensitive part of the marine mammal groupings' audiograms, which may be more conservative than weightings such as Southall et al. (2007 & 2019), so caution should be noted in comparing the results with other studies.

Though the effects will vary depending on the source level, number of ADDs and specific duty cycles, Götz and Janik (2013) reported that TTS could be induced after a very short exposure period to all three devices tested. The time taken for onset of TTS ranged from under a minute (Airmar, four devices, duty cycle of 200%), up to 47 minutes (Terecos, one device at a duty cycle of 11%). Distances to TTS onset varied across species and also depending on the TTS threshold criteria used. For example, TTS onset was predicted to occur in harbour seals at 10 m (using Southall et al., 2007 threshold criteria), harbour porpoise from 89 m (using Lucke et al., 2009 threshold criteria) up to 345 m (using SEL_SENS, the threshold criteria developed by the authors of this study), bottlenose dolphin from 2.5 m (using Southall et al., 2007 threshold) up to 175 m (using SEL_SENS threshold criteria) and killer whale at 784 m (using SEL_SENS threshold criteria). The publication did not attribute these distances directly to a particular device, rather a source level of 193 dB re 1 µPa for 10 seconds. As noted, these calculations are also based on a scenario where an animal remained stationary, which is considered to be unrealistic, although it does provide a precautionary worst case scenario.

In a later modelling study of the GenusWave device, which differs from other ADDs in that it produces a short sound intended to startle seals, Götz and Janik (2015) predicted that TTS would only occur in unrealistic exposure scenarios. For example, in order to experience cumulative TTS, harbour porpoise would have to stay within 1 m of the sound source for 21 minutes, or within 20 m of the sound source for five to six days, due to the low duty cycle used.

For low frequency cetaceans, a single modelling-based study (McGarry et al., 2017) estimated no risk of TTS due to either instantaneous or prolonged exposure to a Lofitech ADD (which is not in use at Scottish finfish farm – see Chapter 3 of this report), even at very close distances of 25 m from the device, assuming that the animal moved away from the ADD.

A rare example of TTS directly measured in a marine mammal as a result of exposure to ADD sound was reported by Schaffeld et al. (2019). A captive harbour porpoise exposed to Lofitech signals experienced significant TTS, with subsequent simple noise modelling predicting this effect could occur at distances between 200 m for deep water and 6 km for a shallow water scenario, using different sound propagation spreading values. It should be noted here that the sound propagation used to calculate these distances is very simplified and does not take into account environmental factors. These distances are predicted for instantaneous TTS, the authors did not calculate the potential for cumulative SEL TTS onset.

Permanent Threshold Shift (PTS) in marine mammal hearing

Permanent Threshold Shift (PTS) is similar to TTS in that it involves a reduction in hearing sensitivity at particular frequencies. The difference is that this reduction in sensitivity is not recovered over time and becomes a permanent change to the animal's hearing ability. The frequencies of hearing affected will determine how significant that change is for the animal. As with TTS, the risk of PTS is heavily dependent on the behavioural response of the animal to the ADD, which can greatly reduce or increase exposure time. The risk of instantaneous PTS injury from ADDs is negligible, so long as the devices are operated as recommended. The risk of PTS injury from cumulative exposure over time is the main consideration here.

The evidence base for PTS in marine mammal hearing is, similarly to that for TTS, dominated by modelling studies. This is due to the difficulty in measuring this physiological effect in free ranging animals, but also due to the ethical issues involved in deliberately inducing permanent hearing changes to an animal. Indeed, empirical studies of threshold shifts in captive marine mammals measure TTS, and then apply corrections to estimate the levels required to induce PTS (Southall et al., 2007), rather than directly inducing PTS.

Two studies that have contributed most in this respect are Götz and Janik (2013) and Lepper et al. (2014). The studies are not directly comparable, because many variables, such as the models used, source levels and devices, threshold criteria and auditory weightings differed between them.

Götz and Janik (2013) estimated that a harbour porpoise exposed to sounds from multiple Airmar devices set to the highest duty cycle could experience cumulative PTS at a distance of under 10 m in under one minute. However, they also showed that for species with less sensitive hearing (seals, low frequency and high frequency cetaceans), exposed to lower power devices or lower duty cycles, the exposure scenarios required to induce cumulative PTS were unrealistic, requiring multiple days of exposure at extremely close range (Götz and Janik, 2013). McGarry et al. (2017) modelled the potential for cumulative PTS in minke whales at various distances from a single Lofitech device. The authors concluded that there was no risk of cumulative PTS, even at close proximity (within 25 m) to the ADD.

Lepper et al. (2014) modelled a number of different ADDs using a sophisticated modelling methodology that included environmental parameters which affect propagation of sound. They used Southall et al. (2007) auditory weightings and thresholds which, at the time of publishing, were based on the best available science (these have since been updated in Southall et al., 2019). Lepper et al. (2014) concluded that there was a credible risk of exceeding injury criteria for the cumulative sound exposure level PTS threshold for seals if they remained within 100 m of a device for periods of several hours and, for harbour porpoise, over a similar time period at distances up to 500 m. The modelling simulated a stationary animal which was noted in the report as being unrealistic.

Consideration of multiple finfish farms in an area deploying ADDs is required, as this may lead to larger areas that are emitting sound, resulting in greater potential for extended durations of exposure. From the available PTS evidence base for some devices, if the animal does not move away from the sound source then there is a risk of cumulative PTS onset. This risk will be increased in situations where multiple ADDs are operating.

3. Evidence for efficacy of acoustic deterrent devices in deterring seals

ADDs are used in several marine industry sectors, with purposes ranging from deterrence of seals to reduce predation of fish, to mitigation of auditory injury around marine construction. Their efficacy is therefore defined according their intended use. Few desk-based modelling studies have considered the potential for disturbance, TTS or PTS on seals from ADD sounds (but see Tables A1-7 and A2-1 in the Appendix). Given the context of this report, the studies included in the review below are from those that are relevant to the efficacy of ADDs in reducing seal depredation at aquaculture sites. Information has been included where it may be possible that the ADD could be used for this application. Efficacy metrics presented in the studies include increases in fish yields, reductions in numbers of damaged fish and reductions in the numbers of seals in the vicinity of a farm or fishery. The studies considered to inform this Section are summarised in Tables A3-1, A3-2 and A3-3, in the Appendix.

There is considerable disparity in the reported responses of seals to ADDs. For example, most field trials of the Airmar device (Gordon et al., 2015; Götz & Janik, 2010; Yurk & Trites 2000) and the Lofitech device (Gordon et al., 2015; Götz & Janik, 2010; Graham et al., 2009; Harris et al., 2014) reported a deterrence effect, although the extent of the response and deterrence ranges vary widely between studies. There are also examples from studies that have reported no significant effect when using the Airmar device (Jacobs & Terhune, 2002) and one study using the Lofitech device suggested an attraction effect (Mikkelsen et al., 2017). In field trials with wild seals, Götz & Janik (2010) found no significant deterrence range for the Terecos device, in contrast to other devices tested, where deterrence ranges of 60 m, 60 m and 40 m were reported for the Ace-Aquatec, Lofitech and Airmar devices, respectively. Captive experiments carried out on small numbers of seals, such as those by Kastelein et al. (2010 & 2015b) and Götz and Janik (2010), demonstrated behavioural responses, including displacement, increased swim speed and increased surfacing rate resulting from exposure to signals from Airmar, Terecos, Lofitech and Ace Aquatec ADDs.

Field trials of the GenusWave device reported a startle response and deterrence effect on seals (Götz and Janik, 2015). A later trial of the same device reported that, despite observing a smaller change in the number seals surfacing close to the device, there was still a significant reduction in depredation of over 90% (Götz and Janik, 2016). With respect to peer-reviewed literature investigating the efficacy of ADDs in relation to deterring seals from finfish farms, the majority of publications are focussed on the GenusWave device. Additional peer-reviewed studies would improve our understanding of the efficacy of this, and other devices used in reducing depredation at finfish farms.

Another type of commercially available TAST device is FaunaGuard, which has several models available, depending on the species of interest. For the model used to deter seals, evidence on the efficacy of this device is currently limited to just one published study using two captive harbour seals. Both individuals did demonstrate significant changes in behaviour by keeping their heads out of the water or hauling out when the device was operational at received sound pressure levels (SPL) above 160 dB re 1 μPa, however when 'jumping' behaviour was included, the behavioural response threshold was estimated to be approximately 142 dB re 1 μPa (Kastelein et al., 2017). Using the maximum sound source level of the device, the authors predicted the device had an effective deterrent range of 500 -1000 m. The FaunaGuard device has not been reported to be used at finfish farms in Scotland (see Chapter 3 of this report).

Often adverse sounds, such as those from an ADD, will initially induce a reaction from an animal, but when an individual is repeatedly exposed to the sounds (either continuously or periodically), their response may diminish. This occurs more frequently where there is a motivation for the animal to tolerate the sound, such as the presence of food. Following captive experiments with grey and harbour seals, Götz and Janik (2010) reported that when motivated by the presence of food, seals tolerated the sounds produced by Airmar, Ace-Aquatec and Lofitech ADDs. This reduction in the response of seals to ADDs over time is widely considered one of the main difficulties in using acoustic signals to deter seals from predating finfish farms, with over 80% of farms reporting this as an issue when using ADDs (Northridge et al., 2013); see also Table A2-3, which provides brief summaries of questionnaires on perceived efficacy of ADDs completed by industry.

4. Knowledge Gaps

There remain a number of knowledge gaps which will require both empirical data collection and modelling studies to fill. This section does not aim to detail all potential knowledge gaps, but to give an overview of areas where further work will be required.

There are three key areas of challenge to understanding the impacts and efficacy of ADDs:

1. While several modelling studies suggest the potential for disturbance, TTS, and cumulative PTS to cetaceans from ADDs in situ at finfish farms in Scotland, there is limited (in the case of disturbance) and lacking (in the case of TTS and cumulative PTS) empirical evidence for this. There is no evidence that wide scale exclusion of cetaceans from areas where finfish farms are sited is occurring, but no definitive quantitative mapping and modelling has been carried out to assess this. There is also a knowledge gap relating to the individual consequences of any behavioural response or displacement, or the effect this may ultimately have at a population level.

2. Predicting the likelihood of auditory injury is often constrained through a lack of information on the source level and frequency spectrum of the ADDs being used, as well as the way in which they are used (e.g., duty cycling). With good management practices and well controlled studies, these variables can be recorded and their effect on the likelihood of impacts to marine mammals better understood. However, the behaviour of animals in relation to the ADDs is much less straightforward to control or predict and will have the greatest bearing on whether an animal is exposed to sounds sufficient to cause auditory injury. Measuring environmental variables that could affect animal motivation to be in an area during exposure studies may help to better predict behavioural responses.

3. For a number of reasons, the ability to quantify empirically the occurrence and impact of TTS and PTS remains extremely challenging. The ability to conduct hearing tests on animals is currently very limited, and would likely be restricted to animals in captivity.

More specifically, the review of evidence demonstrates that further work is required to:

• Gather empirical data on sound propagation from ADDs to investigate the impact of different parameters (e.g., water depth, substrate, sea state, number of ADDs) upon sound propagation in realistic usage settings.
• Better understand the noise levels that animals are exposed to both instantaneously and cumulatively. Such information could be gathered through studies that track marine mammals moving around acoustically monitored finfish farms with and without ADDs operating to allow realistic estimates of noise exposure to be calculated.
• Better understand the way in which ADDs potentially impact on a wider range of cetacean species. For the most part, the focus has been on harbour porpoise, and even for this species, significant uncertainties regarding responses and impacts remain. However, there is the potential for impact to other species, and with respect to those devices operating in the lower frequencies, for an impact to low frequency cetaceans, such as minke whale and humpback whale.
• Better understand whether there are population level consequences of displacement of cetaceans around finfish farms. Modelling frameworks for this already exist, but there would be a requirement for empirical data to understand whether there is broad scale displacement of animals and their effects.
• Better understand the reasons why individual seals show different responses to ADDs and the circumstances and deployment methods under which ADDs are likely to be most effective. This will need to focus on variables that affect seal behaviour in the vicinity of finfish farms, not just the type or deployment method of ADD.
• Better define and understand the efficacy of ADDs. This could be considered through a multi-faceted approach assessing size of yields, occurrence of damaged fish, distribution and density of seals in the vicinity of a farm or fishery. The information on fish (yield and damage) is routinely kept by farms. Additional, contemporary records of ADD use (e.g., locations, times the devices were in use, duty cycle, frequency, source level), and seal presence would add a great deal of value to our understanding of how well ADDs work to deter seals and how the effects of deterrence change over time.

5. Conclusion

The review of evidence has concluded that while studies have investigated the potential for impacts from ADD used at finfish farms on marine mammals, a number of knowledge gaps still exist. Current regulatory processes acknowledge these gaps, and make precautionary assumptions based on the best available science to ensure that marine mammals are protected and environmental obligations fulfilled. However, improving the evidence base over time through additional research and/or monitoring will enable a better understanding of ADDs and the manner in which marine mammals interact with them. This will improve the evidence available for any future decision making and policy.

To allow this to happen, Scottish Government will work with industry, Statutory Nature Conservation Bodies and other relevant stakeholders to identify research and/or monitoring programmes required to improve the evidence base, building on the outcomes of the review. By bringing together science, industry and environmental interests, this will allow the scientific evidence base for decision making to improve over time, facilitating transparency and efficiency in the licensing process and support the sustainable development of aquaculture in Scottish waters.