Evaluating and Assessing the Relative Effectiveness of Acoustic Deterrent Devices and other Non-Lethal Measures on Marine Mammals

1 General Introduction

1.1 Project Overview

There is a long history of using sound to keep animals away from human interests (Bomford and O'Brien, 1990), usually to protect those assets from unwelcome attention but sometimes also to protect the animals themselves from harm. For marine mammals, which have acute hearing, spend most of their time below the surface of the water, and which inhabit a medium (water) through which sound propagation is highly efficient, acoustic methods of deterrence represent an attractive means of behavioural manipulation.

Marine acoustic deterrents were initially developed to keep marine mammals away from fishing gear, but have since found use in other sectors of marine industry. They are now widely used in a variety of applications to keep marine mammals away from specific installations or activities where their presence may result in damage to human interests or to the mammals themselves.

The aquaculture industry in particular makes use of high-powered acoustic deterrents to dissuade seals from damaging fish pens, attacking fish through the meshes or making holes through which fish can escape. In Scotland this principally applies to salmon farming, one of rural Scotland's most important industries.

Much the same principle is also now being applied to wild salmon fisheries, where Acoustic Deterrent Devices ( ADDs) have been used in attempts to reduce seal damage to rod and line fisheries and to salmon bag-net fisheries. A different suite of acoustic deterrent devices is also being used in certain long-line and other fisheries to deter cetacean depredations.

Offshore marine engineering projects have increased greatly in past decades; some associated activities, particularly pile driving at offshore construction sites, have the ability to cause severe hearing damage to marine mammals. Acoustic deterrent devices have the potential to exclude animals from areas where they might be at risk of damage.

Smaller and less powerful devices are used to keep some species of small cetacean away from static net fisheries (gillnets) where there is a danger of them becoming entangled. These devices have been shown to reduce the unintentional entanglement of small cetaceans like harbour porpoises by more than 90% in some situations (Dawson et al., 2013; Kingston and Northridge, 2011; Kraus et al., 1997; Northridge et al., 2011).

There is now a wide range of commercially available devices, each generally marketed toward a specific application of the types described above. These devices are known to differ substantially in their acoustic characteristics (frequency and amplitude) as well as practical aspects such as power supply and cost. In many cases, however, the efficacy of specific devices in their respective applications remains largely untested and poorly understood.

Reducing the risk of entanglement in nets, of hearing damage at construction sites, or reducing the chance of being shot at a fish farm site, are all potential benefits to individual animals. The use of acoustic devices in these cases may therefore have conservation or welfare benefits to animal groups or species. However, excluding animals from parts of their natural habitat, as well as potentially causing reductions in hearing sensitivity due to long term exposure to ADD sounds, are costs that need to be balanced against these potential benefits.

A number of studies have investigated the effects of various noises on marine mammals, both behaviourally and physiologically. Conservation and economic benefits have also been addressed in many of the areas described above. This study attempts to provide an overview of this large body of work, and to identify current uncertainties and directions for future research. It is also intended to provide a comprehensive assessment of the capabilities of current and developing non-lethal measures which are used to alter the behaviour of marine mammals in different scenarios with the purpose of answering several management oriented questions.

The main objective of this research project is to review the literature and data on current and developing acoustic deterrent devices which are used for deterring marine mammals in different scenarios, with the purpose of answering questions regarding design, effectiveness, best practice and impacts of these devices on marine mammals in Scotland. More specifically, through this review we also address the following management related questions:

• What types of ADD are currently employed, or are in development, which are used to deter marine mammals in different scenarios, for example at fish farms, netting stations, rivers, and in/around areas of development ( e.g., oil and gas, renewables)?
• Are these devices fit for purpose and appropriate for deterring marine mammals in a range of scenarios and often at a very local scale? For example, are some commercial devices more applicable for deterring seals in more constrained salmon rivers, while others are more appropriate for deployment in coastal or offshore waters? Will some devices be more appropriate for long-term deployment as opposed to short-term?
• Are certain devices more appropriate to a particular species? Are there different requirements for seals, toothed cetaceans, and baleen whales (dependent on the purpose of deterrence)?
• What is the relative effectiveness of existing ADDs on marine mammals (considering seals and cetaceans separately)? For example, at what range do they exclude mammals? Do certain devices exclude seals and not cetaceans, and vice versa?
• Are there efficiency improvements which could be made by best practice in using existing ADDs? For example, targeted activation of devices when marine mammals are located in the vicinity of the devices (as opposed to continuous use).
• What are the ecological consequences of ADD's in terms of underwater noise? Are some devices more 'noisy' (ecologically disruptive) than others?
• Beyond ADDs, are there any other current or developing technologies for deterring marine mammals? When answering this question, consideration should be given to the reasons for deterrence ( e.g., aquaculture, fisheries, mitigation for renewable development).
• Can baseline information be improved which would benefit developing marine industries?

Many of these remain open questions, and we conclude our review by addressing them in turn and in the light of the information reviewed below.

1.2 Acoustic Deterrence

The mechanisms by which aversive sounds or alerting signals achieve their effect have rarely been elucidated. However, there are a number of models that are likely to apply for different species in different circumstances. An important distinction can be drawn between those signals for which animals have a learned association and those for which they do not. Many responses to sounds can be understood in terms of predator-prey behaviour; most marine mammals are on one hand predators that will use acoustics cues to locate prey and on the other potential prey themselves for which acoustic cues may be important for detecting the presence of predators. Interactions with conspecific competitors and other competitor species may also be harmful, for example harbour porpoises ( Phocoena phocoena) are often harassed and even killed by bottlenose dolphins ( Tursiops truncatus) (Ross and Wilson, 1996) and acoustic cues will also have relevance to these scenarios.

Acoustic signals which have no particular relevance may induce avoidance in neophobic species, or those that tend to show fear of novel stimuli. Dawson et al. (2013) suggested that acoustic devices will be most effective in reducing bycatch of neophobic species, such as porpoises, which are well known for being extremely timid. The degree of neophobia an animal exhibits probably reflects its anti-predatory behaviour, being an important component of strategies that involve simple avoidance of potential predators. The behavioural response may be mediated by other factors such as the likelihood of predators being locally present and the proximity of refuges. By contrast, for opportunistic foragers, novel signals may elicit curiosity and attraction. This might be the case for seals, especially in situations where predation risk is perceived as being low. Marine mammals, with the exception of killer whales ( Orcinus orca), are both predators and prey so we might expect their responses to novel sounds to be complex and sometimes contradictory.

In humans, some signal characteristics seem to be inherently unpleasant due to the way they interact with the auditory system. For example, research summarised by Zwicker and Fastl (2004) showed that acoustical properties such as loudness, fluctuation strength and sharpness correlated well with a sound's perceived 'annoyance'. In addition, consonant sounds, which involve the combination of tones whose frequency ratios are small integers, are perceived as being more pleasant than dissonant sounds. However, experiments indicate these characteristics which make sounds pleasant or unpleasant to humans may not transfer to other primates (McDermott and Hauser, 2004).

Götz (2008) and Götz and Janik (2010) attempted to measure whether signals designed to be aversive based on the characteristics summarised by Zwicker and Fastl (2004), termed Psycho-Physical Model Sounds ( PPM), were in fact more aversive to seals than signals from existing ADDs or control signals (such as white noise). Their results were equivocal. In captive studies, seals were provided with motivation by a feeding station. Initially, animals avoided the feeding station on first exposure to all sound types, but in subsequent trials then largely ignored all sound playbacks equally. However, when playbacks were made at real haulout sites, where motivation for animals to remain in the area was lower, observation of the distances at which seals surfaced from the sound source indicated a stronger avoidance for PPM signals. Thus, this work suggests that a psycho-physical model might be used to generate sounds that are more aversive, but also indicates that this degree of aversion is mediated by motivational state and may be insufficient to overcome a feeding motivation.

A sound received at a very high level becomes unpleasant as it begins to exceed the dynamic range of the auditory system eventually causing physical discomfort and then pain. Some ADD manufacturers claim that their devices operate in this manner; they are simply too powerful for animals to be able to approach. Götz (2008) has stressed that devices that work in this way are inherently flawed because the sound levels that are necessary to cause discomfort are likely to lead to permanent threshold shifts after moderate exposure durations, while those that induce pain are operating very close to levels that would cause immediate damage.

Acoustic signals are likely to induce avoidance if they are similar to signals for which the subject has made a negative association. The most obvious example will be noises associated with predators or predation. Such learned associations will become stronger if they are repeatedly reinforced, if for example predators are often encountered, and weaken if animals are repeatedly exposed to signals without reinforcement. This waning in responsiveness with repeated unreinforced exposure is termed habituation. In a review of the use of aversive sound to exclude terrestrial pests, Bomford and O'Brien (1990) concluded that biologically significant signals showed most promise, especially if measures were taken to minimise habituation. In terrestrial species alarm signals are often used, but such calls are not known to be commonly used by marine mammals. A concern with the routine use of predator sounds is that it could interfere with the target animal's ability to respond appropriately to genuine predators - they may, for example, become habituated to predator calls and thus not show avoidance of actual predators. Effects on the predatory species themselves should also be considered before the widespread use of such signals.

Other than its acoustic characteristics, the behavioural response to an acoustic signal is known to depend upon a variety of psychophysical parameters relating to the way the sound is perceived, such as the animal's motivational state and basic learning processes which may have already occurred ( e.g. habituation or conditioning)(Götz and Janik, 2010). The potential for using stimuli that elicit a startle response as aversive signals was investigated by Götz (2008), as part of a Scottish Government funded project to investigate more effective and discriminatory acoustic deterrent devices for use as salmon farms. The acoustic startle response is one of a suite of autonomous reflexes hypothesised to mediate a rapid flight response. To be capable of doing this a sound must be sufficiently loud (received at 92 dB above sensation level in humans) and also have a fast rise time during onset. This principle was studied in further detail by Götz and Janik (2011) who conducted captive trials with seven grey seals ( Halichoerus grypus). By broadcasting high-intensity bursts of sound with very short rise time of 5 ms (the time between the signal onset and maximum amplitude) they found that five of the test subjects exhibited a startle response, while two did not. Rather than the response diminishing over time, as is often observed as animals habituate, they found increasing responses from all of the startled animals, indicating sensitisation, or increased responsiveness to the signal. After six trials the sensitisation effect was marked, with retrieval of the food source prevented and a flight response reliably induced. They also investigated the potential for behavioural conditioning of seals (as discussed by Pryor in Mate and Harvey, 1986). By pairing the startle stimulus with an originally neutral 'conditioned stimulus', played 2 seconds prior to the startle signal, they were able to show that the pre-sound could be used to induce a conditioned response (avoidance behaviour). Results from farm trials are expected soon and commercial trials are underway.

1.3 History of Development

The idea of using acoustic devices to deter marine mammals has existed for many decades, with one of the earliest reported attempts by Fish and Vania (1971) who used broadcasts of killer whale calls in attempts to keep beluga whales ( Delphinapterus leucas) away from salmon nets. Since this time, several research groups and private companies have sought to utilise aversive underwater sounds for a variety of applications, with mixed results.

Anderson and Hawkins (1978) authored the first publication describing attempts to deter pinnipeds acoustically. During a feasibility trial they broadcast pure tones and killer whale calls at salmon netting stations at the mouth of the river Tweed. However, they were unable to achieve any consistently useful deterrent effects. Elsewhere, a programme to develop a high powered acoustic deterrent to reduce pinniped depredation was undertaken in the 1980s by Bruce Mate and colleagues of Oregon State University. Their early reports (Mate et al., 1986b; Mate and Harvey, 1986) describe the development and testing of a device with peak frequencies of 12 and 17 kHz, around the peak of seal hearing sensitivity. They developed a device called the 'sealchaser', which had a source level of 188 dB (re 1 µPa RMS) and an upsweeping frequency from 11.5-15 kHz (Mate and Harvey, 1986). A separate system, developed by Coastline Environmental Systems in Vancouver, Canada was described by an article in 'Canadian Aquaculture' in 1988 (Smith, 1994). This device used recorded and synthesised tones, including killer whale vocalisations. Shortly after this, two similar acoustic deterrents were first tested in Tasmania (reported by Pemberton and Shaughnessy, 1993), one with peak energy at 10 kHz, the other at 27 kHz. The report of a workshop which focused on the use of acoustics to deter marine mammals (Reeves et al., 1996) stated that recent advances in underwater technology had allowed the development of very high-amplitude devices (such as a device from Airmar), and which appeared to have overcome past problems of declining effectiveness.

Acoustic devices were first introduced to the Scottish salmon aquaculture industry in the mid-1980s, after trials made by the Department of Agriculture and Fisheries for Scotland ( DAFS) and shortly afterwards an Orcadian company began to market a similar device. Since then, several different companies have produced models of acoustic deterrent in attempts to control pinniped depredation.

Some early work attempted to scare seals using the vocalisations of mammal-eating killer whales. Attempts were generally unsuccessful in this context and these devices no longer appear to be in use. The use of killer whale vocalisations is discussed in detail below ( section 4.4.1).

More recently, the work of Thomas Götz and Vincent Janik at St Andrews University has pursued an alternative, more discriminatory, approach to the problem based on the use of the acoustic startle reflex. While earlier devices had aimed to maximise acoustic output within pinnipeds' most sensitive frequency band (around 15 kHz), Götz and Janik aimed to reduce unintended impacts on odontocetes by producing sound at lower frequencies, where odontocete hearing is less sensitive than that of pinnipeds. This device emits high intensity sound (180 dB re 1 µPa peak to peak) with very short rise time and a peak frequency of 950 Hz (much lower than other devices)(Götz, 2008). The device was tested at a fish-farm site in Scotland by Götz (2008) who found that seals were excluded from within 50m of the device and that sighting rate was reduced up to 250m. In contrast to these results, the behaviour of harbour porpoises in the vicinity of the farm appeared to be unaffected. The concept of their device and the investigations they have conducted are discussed throughout this review where appropriate.

The concept of using acoustic deterrence to minimise cetacean entanglement in fishing gear was pioneered by Jon Lien, working in Newfoundland, during the 1980s. Lien and colleagues, after several prototype trials, developed a portable, low power acoustic device (4 kHz fundamental frequency, with a source level of 135 dB re 1 µPa @ 1m) which was successful in reducing whale entanglements (Lien et al., 1992). In the early 1990s, fishermen in the US Gulf of Maine persuaded Jon Lien to deploy some of his 'pingers' in a demersal gillnet fishery, with the aim of reducing porpoise bycatch. Initial trials seemed at least partially successful, but were dogged by poor experimental design. Subsequently a consortium of New England research scientists and fishermen working with the Dukane corporation, used a modified 'pinger' originally designed as an underwater location aid, to demonstrate the effectiveness of such devices in minimising the bycatch of porpoises (Kraus, 1999; Kraus et al., 1997). Porpoise bycatch rates were reduced by more than 90%, and numerous other experiments with these and other low powered devices have demonstrated their effectiveness in minimising the bycatch of porpoises and a few other small cetacean species (Dawson et al., 2013).

Although at one stage there was a marked divergence of marketed devices into high powered seal deterrents, primarily used in the Aquaculture industry, on the one hand and lower powered 'pingers' primarily used to minimise porpoise bycatch in gillnets, several companies have subsequently produced devices that bridge the 'gap' between these two archetypes. Quieter seal scarers and louder cetacean deterrent devices have entered the market. So although it has been suggested that acoustic deterrent devices can be grouped into low amplitude and high amplitude devices (sometimes termed acoustic deterrent devices ( ADDs) and acoustic harassment devices ( AHDs) respectively - see Reeves et al. (2001)), we do not use this distinction as it may imply an unwarranted attribution of intent in the use of higher amplitude devices. In general, all such devices are intended to deter animals and it is the mechanism by which they evoke an aversive response which is important. This will depend on many factors of which source level is just one. The distinction is considered unhelpful given the broad range of source levels employed by the different acoustic devices which has blurred a previously clear cut distinction between two groups of device. Here we refer to all devices as ADDs, while also including the widely used term 'pinger' to describe the subset of small alkaline or lithium-ion cell devices predominantly used to reduce small cetacean bycatch in gillnet fisheries.

1.4 Current Uses

A large number of acoustic devices are now available for commercial use, most of which are marketed toward the mitigation of interactions with a specific species or industrial practice ( e.g. gillnetting, aquaculture, etc.). In the majority of cases the acoustic characteristics of these devices are poorly described, if at all. Table 1 lists all known devices, past and present, and shows the usage and species for which the device was designed, or has subsequently been used. This table is adapted and expanded from Dawson et al. (2013), and more details on some devices are available there, including acoustic characteristics where available.

Table 1 Acoustic devices, their manufacturers and intended use/species. Adapted and expanded from Dawson et al. (2013).

Manufacturer	Device	Use/Species
Mustad	Orcasaver	Longline depredation - Orcinus orca [1]
Ixtrawl	Cetasaver	Pair trawl bycatch - Delphinus delphis (Morizur et al., 2008)
STM	DDD	Set net bycatch - Odontoceti (Northridge et al., 2011) Longline depredation - Odontoceti (Nishida and McPherson, 2011)
	DDD-03H	Pair trawl bycatch - Delphinidae (Northridge et al., 2011)
	DiD	Longline depredation - Odontoceti (Nishida and McPherson, 2011)
Fishtek	Banana Pinger	Set net bycatch - Phocoena phocoena [2]
Dukane	Netmark	Set net bycatch - Phocoena phocoena ( e.g. Kraus et al., 1997) plus various small cetacean species ( e.g. Pro Delphinus, 2010)
Aquatec	Aquamark models: 100,200,210,300, and 848	Set net bycatch - Phocoena phocoena, Stenella ( e.g. Sea Fish Industry Authority, 2005) and gill net depredation - Tursiops truncatus ( e.g. Brotons et al., 2008)
Airmar Technology Corp.	Gillnet Pinger 10 kHz	Set net bycatch - Phocoena phocoena ( e.g. Sea Fish Industry Authority, 2005)
	Gillnet Pinger 70 kHz	Gillnets & Handlines - Pontoporia blainvillei (Bordino et al., 2004)
Fumunda Marine/Future Oceans	3 kHz Whale Pinger	Fishing gear entanglement - Baleen whales [3]
	10 kHz Porpoise Pinger	Set net bycatch - Phocoena phocoena ( e.g. Sea Fish Industry Authority, 2005) Tursiops truncatus
	70 kHz Dolphin Pinger	Set net depredation - Tursiops truncatus (Read and Waples, 2010)
Marexi Marine Technology	Pinger V02	Set net bycatch - Phocoena phocoena ( e.g. Morizur et al., 2009)
	Orca-Stop	Orcinus orca [4]
Seamaster Fishing Supplies	Seamaster Fish Protector	Aquaculture, gill nets, purse seine, squid & fish trawl industries - Tursiops truncatus, Delphinus delphis [5]
Ingenieria y Ciencia Ambiental ( ICA)		Aquaculture interactions - Tursiops truncatus (López and Mariño, 2011)
SaveWave	SaveWave ADD	Gill net depredation - Tursiops truncatus (Waples et al., 2013)
Jon Lien et al.		Static gear entanglement - Megaptera novaeangliae and later set net bycatch - Phocoena phocoena ( e.g. Gearin et al., 1996)
Loughborough University	PICE	Set net bycatch - Phocoena phocoena ( e.g. Culik et al., 2001)
Ferranti-Thomson	Mk2 Seal Scrammer and '4x - 24V'	Aquaculture depredation - Phocidae (Gordon and Northridge, 2002)
Ace Aquatec	Silent Scrammer and Universal Scrammer 3	Aquaculture depredation - Phocidae [6]
	Marine Mammal Deterrent ( MMD)	Mitigation of pile-driving and natural disasters [7]
Terecos Limited	DSMS-4	Aquaculture depredation - Phocidae (Gordon and Northridge, 2002)
Airmar	dB Plus II	Aquaculture depredation - Phocidae & Otariidae ( e.g. Vilata et al., 2010)
Lofitech	Fishguard/Seal Scarer/Universal Scarer	Aquaculture depredation - Phocidae [8] Capture fisheries - Odontoceti [9]
Poseidon	T88	Aquaculture depredation - Otariidae (Stewardson and Cawthorn, 2004)
Kemers Maskin AB	MkII	Capture fisheries depredation - Otariidae (Stewardson and Cawthorn, 2004)
Northern Gulf Natural Resource Management ( NGNRM)	Dugong/Dolphin Acoustic Alarm	Set net bycatch - Dugong dugon (McPherson, 2011)
Edmund Gerstein et al.	Parametric Alarm	Vessel collision - Trichechus manatus (Gerstein et al., 2008)