Yet Another Study Shows Dramatic Decrease in Dolphin Forgaing Near Boats
Antonella Arcangeli and Robert Crosti. The short-term impact of dolphin-watchingo n the behaviour of bottlenose dolphins (Tursiops truncatus) in western Australia. Journal of Marine Animals and Their Ecology. Vol 2, No. 1. 2009. [DOWNLOAD PAPER]
Over the past few years, a steady trickle of new studies have indicated that many species may decrease foraging activity in the presence of boats. In many situations, it can be difficult to have enough sightings without boats to create a data set large enough to have statistically signifcant results (e.g, orcas in the Pacific northwest, or dolphins in the Mediterranean). Also, the presence of the observation boat can confound results. This study took place in a bay near Bunbury, 200km south of Perth, western Australia, a location where these limitations could be overcome. The weather conditions in summer are generally calm, allowing excellent observing conditions, with the observation inflatable's motor often shut down; observations were only made in quite calm seas (Beaufort 2 or lower). A large local population of dolphins offered plenty of data, and the population is generally stable in place, engaging in a series of regular behaviors over relatively prolonged timeframes. Behavior "without" boats present was noted when no boats (other than the observation inflatable) were within 350m; behavior "with" boats present was noted when tour boats came within 150m. (Ed. note: it would be interesting to consider the data when tour boats were 150-350m distant, to assess whether the observed effects extend to greater distances). Over the course of 65 hours of observation by a single observer, 1028 3-minute behavioral samples were taken, 781 with no boat (average duration 69 min.), and 247 with a tour boat present (average duration 47 min.). As usual with such studies, several behavioral states were analyzed: Traveling, Socializing, Resting, Milling, Feeding, and Diving. All showed some change in the presence of boats, with Traveling, Milling, and Diving all increasing in the presence of boats, and Socializing, Feeding, and Resting decreasing. Of these, the most relevant to an animal's energy budget all were negatively affected: Traveling increased from 27% of the time to 46%, Resting decreased from 31% to 20%, and by far the largest energetic impact was seen in Feeding, which reduced from 20% of the time to just 7.6% of the time. This is the most dramatic decrease in feeding seen in such studies (others have tended to find a 25-35% decrease in foraging). This may indicate that the relatively localized population had plenty of time to eat when the boats were not present, and chose to either leave or play with the boats when they were present. The study did look at attraction/deterrence, finding that in in 20% of cases, dolphins moved closer to the boats, while in 28% of encounters, they moved away (about half the time, there was no spatial difference, though behaviors may still have changed). The Discussion section of this paper includes an unusually detailed comparison of the findings and the contexts of this and several similar studies, as well as consideration of the ways that in other settings, were human activity is more prevalent, these behavioral changes could have longer-term, population-level impacts.

Note: Following are some highlights of the 2008 Research Summaries; see archive pages linked above for more entries

Pinger Study Illustrates Extreme Variabitlity of Real-world Sound Propagation
Shapiro, Tougaard, Jorgensen, Kyne, Balle, Bernardez, Fjalling, Karlsen, Wahlberg. Transmission loss patterns from acoustic harassment and deterrent devices do not always follow geometrical spreading predictions. Marine Mammal Science, 25(1): 53067 (January 2009)
This study was designed to investigate why acoustic harassment devices, or pingers, on fishing nets do not always succeed in keeping dolphins away. The researchers did a series of recordings around pingers of several different styles, and found that the received sound levels did not simply increase as they came closer, but fluctuated dramatically; thus, animals approaching the nets may not have clear sense of how far away they are. While the results here are somewhat mixed for pingers (overall, the average sound level DOES go up as the animals approach, albeit mixed with large variations), the study offers a fascinating glimpse into the real-world factors that will often confound simple geometric or spherical-spreading models generally used to predict sound exposure levels around any sound source. The researchers suggest that many factors likely contribute to the variation: changes in directitivty of the source, subtle differences in source level for each ping, and, most influentially, as a result of the constructive and destructive interference of direct, surface-reflected, and bottom-reflected sound waves.

Ed. note: This graph is a great picture of the degree of varation seen. Note a routine variation of 10-15dB, with some more dramatic shifts: at 400m, the sound shifted from around 127dB to over 150dB; and again, at just under 200m, from just above 130dB to over 160dB (around a general average at that distance of perhaps 150db). Given the increasing concern among biologists about behavioral disruptions triggered by noise, it is extremely important to acknowledge that real-world exposures can vary to this degree; many more animals respond behaviorally at 150dB and above, than down at 120-140dB.
Related: Beaked Whales Very Senstive to Pingers; By-Catch Reduced to Zero over 11 Years
Carretta, Barlow, Enriquez. Acoustic pingers eliminate beaked whale bycatch in a gill net fishery. Marine Mammal Science, 24(4):956-961 (October 2008).
This study documents a dramatic shift in beaked whale bycatch by gillnets set for sharks and swordfish in the California Current. Over the first 6 years of data, 1990-1996, 33 beaked whales died in the nets; after installation of pingers in 1996, NO beaked whales were caught by the nets over the succeeeding 11 years, though 260 cetaceans of 11 other species did drown in the nets. Ed. note: this result seems to dovetail well with observed reactions of beaked whales to other sounds; it appears they are espeically sensitive to noise, perhaps at a wide range of frequencies.

Passive Acoustic Detection of Beaked Whales: Easy Within 1km, Nearly Impossible Beyond 4km
Zimmer, Harwood, Tyack, Johnson, Madsen. Passive acoustic detection of deep-diving beaked whales. The Journal of the Acoustical Society of America, November 2008, Volume 124, Issue 5, pp. 2823-2832.
Beaked whales spend very little time at the surface, with foraging dives that last an hour or more, including about 30 minutes of active echolocation at the feeding depth. Visual detection is very difficult, so the possibility of using Passive Acoustic Detection (PAD) is an attractive complement to visual spotting. However, these high-frequency clicks attenuate rapidly. According to this paper, acoustic modeling suggests that in "good conditions,"e.g. wind speed of 2 m/s, a hydrophone close to the surface should detect beaked whales with a high probability within .7km. At the other end of the detection range, no whales would be detected at distances greater than 4km, except in very low ambient noise or unusually good propagation conditions. The detection curve generated by the models, drops steeply as distance increases, with a relative leveling off that suggest roughly 50% detection when whales are 1.5-3.5km distant. Additionally, the researchers substituted some actual dive profile data obtained in D-Tag studies, and found that in these instances, detection probability may rise somewhat, with 80% detection being possible at distances of 1.5-2.5 km. Actual detection of real whales is complicated by the fact that the sound of the clicks is highly directional; only clicks directed nearly directly toward the receiver will be heard at the modeled distances, but echolocating whales do scan in many directions, so at least some clicks from any whale should be detected. Finally, the relatively quiet interval between deep foraging dives can be as long as 110 minutes, meaning that listening time should be roughly 140 minutes to have a high probability of detecting if beaked whales are present. This, in turn, suggests that a slow-moving vessel (such as gliders or drifting buoys) will be more successful than a fast-moving active sonar vessel. (That is, if detections are only going to occur within roughly 4km, the listening platform should not move out of a relatively similar area during the 2-hour-plus PAD session.) Ed. note: While it has still not been confirmed to all observers' satisfaction, there is some evidence that beaked whales affected by mid-frequency active sonar may well have been tens of km from sonar vessels. Thus, PAD may be useful for avoiding nearby exposure, but may not fully mitigate for effects that could occur at greater distances, including fleeing in ways that disrupt the whales' post-dive recovery and oxygen/nitrogen rebalancing.

Detailed Review of Sound Exposure Impacts on Puget Sound Orcas Released by NOAA
Marla Holt, Sound Exposure and Southern Resident Killer Whales (Orcinus orca): A Review of Current Knowledge and Data Gaps. NOAA Technical Memorandum NMFS-NWFSC-89, February 2008. 77pp. [DOWNLOAD REPORT(pdf)]
This review includes summaries of most of the studies of ambient and shipping noise in the range of the southern resident orca pods of the Pacific Northwest. Topics covered include ambient sound patterns, shipping noise, whalewatching noise, orca auditory capabilities and auditory effects of sound exposure, including masking and hearing loss, and behavioral changes in the presence of excess sound. An important section looks at the emerging concept of "zones of influence" of sound: the zones of audibility, response, masking, and hearing loss or injury. The report concludes with a section that assesses the likelihood of acoustic impacts and suggestions for future research.

The report is a treasure-trove of clear information on noise/sound impacts. Among the points that stood out in my reading are the following. Ambient noise is generally elevated at least 10-15dB by small boats, and 20-25dB by large ships (another study indicates increases more in the range of 50-70dB in the presence of various boats). The section on hearing senstivity makes a point about captive animal testing that is rarely noted: when animals receive negative reinforcement for reporting a "false positive" (i.e. responding as if hearing a sound, when no sound was played), they may well become conditioned to be conservative in their response to faint sounds, and so the results can easily underestimate the hearing sensitivity of the species (for orcas, only two studies have been done on captive whales to assess hearing senstivity curves). Tentative measurements of the critical ratios for orcas (tentative because, again, there are few test subjects) suggest that they are more affected by masking than many other toothed whales: to hear a target sound in noise, the target sound must be 20dB louder than ambient at 10kHz, and 20dB louder than the noise levels at 80kHz. There are some indications that orcas spend less time feeding and more time traveling or resting when boats are nearby (within 100m or 400m); likewise, they may travel in less direct paths in the presence of boats within 1000m (though interestingly, one study suggests that when there are many boats present, paths are more direct; perhaps there is little use in trying to avoid boats in this instance). An interesting study (based on models, not measured/observed results in the field) suggest that the presence of boats can significantly decrease the detection range for an orca ecolocating on a salmon: boats at 400m decreased the detection range by roungly 35-90%, and boats at 100m decreased the detection range to as little as 0-20m, or 90-100%.

European Cetacean Society Workshop Addresses Offshore Wind Turbine Impacts on Marine Mammals
Evans, P.G., Ed. ASCOBANS/ECS Workshop: Offshore Wind Farms and Marine Mammals: Impacts and Methodologies for Assessing Impacts. ECS Special Publication Series No. 49, February 2008. 70pp. [DOWNLOAD REPORT]
Prospect of Widespread Offshore Development Requires Cumulative Impact Analysis
Simmonds, Dolman. All at sea: renewalbe energy production in the context of marine nature conservation.
This survey of potential conservation impacts of offshore wind development quotes several previous studies that address specific concerns. Madsen et al. (2006) comment that “if the very large offshore wind farms are realised... this could involve construction activities at several locations in the area [of the German Bight] simultaneously every summer for the next decade.”
Madsen, P.T., Wahlberg, M., Tougaard, J., Lucke, K. and Tyack, P. 2006. Wind turbine underwater noise and marine mammals: implications of current knowledge and data needs. Mar Ecol Prog Ser., 309: 279-295. Wahlberg and Westerberg (2005) concluded that fish behaviour could be affected at ranges of several kilometres. Wahlberg, M. and Westerberg, H. 2005. Hearing in fish and their reactions to sounds from offshore wind farms. Mar Ecol Prog Ser., 288: 295-309. Carstensen et al. (2007) commented that ‘given the extensive plans for expanding the offshore wind energy sector, it is important to know the effect of single wind farms and well as the cumulative affects of several wind farms within the range of each marine mammal species”. Indeed, it would make good sense to consider the cumulative impact of all new developments on
populations and yet, in our experience, this rarely seems to happen. Carstensen, J., Henriksen, O.D. and Teilmann, J. 2007. Impacts on harbour porpoises from offshore wind farm construction: acoustic monitoring of echolocation activity using porpoise detectors (T-PODs). Marine Ecology Progress Series 321: 295–308.
UK Offshore Development Scaling Rapidly to Much Larger Facilities
Prior, McMath. Marine mammals and noise form offshore renewable energy projects—UK developments.
This paper notes that offshore wind farm development in the UK has taken place in licensing "rounds," with rapid increases in the size of wind farms being built. Round 1 was an initial demonstration phase, with eleven projects permitted so far (and five in operation), each relatively small in scale (less than 30 turbines) and relatively short construction phases (as short as one month). Fifteen Round 2 projects have been proposed (with four approved so far); these are larger developments often including hundreds of turbines. Round 2 construction will likely take place over a number of seasons, so construction disturbance will be longer and the impacts could be different than those seen in Round 1. In addition, piles of up to 6 feet in diameter have been proposed to support larger turbines; this will lead to substantially increased noise during construction, with disturbance impacts extending over tens of kilometers. Proposals for Round 3 are currently undergoing Strategic Environmental Assessment; however, the proposed capacity of development may be as much as four times the eventual level built out under Rounds 1 and 2 combined. Concern over cumulative impacts may be prominent during the consenting process, particularly if construction operations on adjacent sites takes place concurrently, giving rise to the potential for longer term and geographically widespread increases in underwater noise.
Porpoises Numbers in Wind Farms Can Vary Dramatically
Teilmann, Tougaard, Carstensen. Effects from offshore wind farms on harbour porpoises in Denmark. Included in:
In a paper summarizing the results from acoustic monitoring at two Danish offshore wind farms, results clearly show a dramatic difference in harbour porpoise responses. At one wind farm, Horns Reef, porpoises numbers decreased modestly during construction (both inside the farm in in monitoring sites 10km away), and decreased a bit more during early start-up, when maintenance activity was highest, but rebounded to baseline (pre-construction) levels early in the first year of normal operations. However, at the Nysted wind farm, the decrease in porpoise numbers that likewise began during construction was not reversed during the first two years of operations: porpoises mostly avoid the wind farm (though numbers were slowly increasing by the end of the second year), while they returned to the areas 10km away once construction ended. The authors note two possible explanations for the different long-term impacts of the wind farms. The area around Horns Reef is an important harbour porpoise habitat, with a larger natural population density; it is possible that the "the porpoises around Nysted are not particularly interested in the area and will simply avoid it if disturbed." In addition, the area around Nysted is relatively sheltered, whereas Horns Reef has higher wind and waves, and thus a higher background natural ambient noise level, so that "the relative noise level from the turbines is higher and audible to the porpoises at greater distances at Nysted than at Horns Reef." The authors conclude that "it can be concluded that the same species may react differently to similar types of disturbance (wind farms) in different localities. This is an important conclusion for future monitoring programmes of wind farms and other offshore installations."

Orcas Avoid Mid-frequency Sonar Signals, but not LFAS; Herring Unaffected
Kvadsheim, Benders, Miller, Doksaeter, Knudsen, Tyack, Nordlund, Lam, Samarra, Kleivane, Godo. Herring (slid), killer whales (spekknogger) and sonar - the 3S-2006 cruise report with preliminary results. Norwegian Defence Research Establishment (FFI). 30 April 2007
This paper reports preliminary results from an ambitious multi-national Controlled Exposure Experiment using acoustic D-tags, which allow researchers to record received sound levels while charting dive patterns. Six orcas were tagged, with 17 hours of data collected, with one animal exposed to LFAS signals, two to mid-frequency active sonar signals, and one used as a control, fewer samples than hoped. The whale exposed to LFAS signal did not change its behavior, nor did its companions. However, the group exposed to mid-frequency sonar signals ceased feeding and moved rapidly away; in addition, they exhibited an unusual dive pattern, diving roughly twice as deep (over 60m) as normal (20-45m), including a highly unusual reversal of their ascent (coming up to 15m from the surface, then retreating again to 60m). [Ed. note, not part of the report narrative, based on observation of included diagrams: Four other dives during the 35-minute sonar transmission were within the normal range of depth (the deep dive was the third of five dives during sonar transmission), with three longer than normal rest periods at the surface during the hour after sonar transmission ceased.] Received levels did not exceed 150dB re 1uPa. Looking at longer-term responses, while the experimental use of sonar in the herring trials did not seem to deter orcas from the general area, by contrast, no orcas were seen during three days of a planned Navy exercise in the area that included mid-frequency sonar use. The herring trials used bottom-mounted sensors to chart the mean depth of a school of herring as a sonar vessels passed nearby. Herring at shallow depths (25-50m) tended to show a minor downward reaction as the sonar source ship passed overhead, while herring located deeper, or in less dense schools, showed no detectable reaction; no horizontal avoidance was seen. While herring hearing curves suggest they should be able to clearly hear LFAS signals, but not mid-frequency signals, the responses were the same for all signals (including ship-only, no sonar), suggesting that the herring may have been responding to the sound of the ship, rather than the sonar signal; such reactions to ships have been seen for many fish species. A related trial using orca echolocation signals on the herring showed more of a response, suggesting that the herring could distinguish between the mid-frequency sonar and the orca sounds, but the three trials all produced slightly different response patterns, suggesting the need for further study.

Review of Possible Ways that Increasing Ocean Noise May Affect Marine Mammals
Peter L. Tyack. Implications for marine mammals of large-scale changes in the marine acoustic environment. Journal of Mammalogy, 89(3): 549-558, 2008. [DOWNLOAD PAPER(pdf)]
In this wide-ranging literature review, Peter Tyack of Woods Hole Oceanographic Institute sketches the history of research into the effects of noise on marine life, with some references as well to effects seen in terrestrial creatures. He begins by noting that while acute disturbance of individuals attracts the most attention, the likely more profound effects of chronic disturbance on population vitality (success in foraging and mating) are much harder to discern. Several examples are presented of studies that documented both temporary and long-term abandonment of key habitat when loud noise was present (including grey whales abandoning a birthing lagoon for several years, then returning when the salt production facility was abandoned, and dolphins moving away from foraging habitat when shipping traffic is heavy).

Next, Tyack turns to a detailed examination of the question of whether global shipping may be dramatically decreasing the area within which whales can hear each other's calls, beginning with the thought that the unintentional consequences of increased shipping noise may be creating unexpected problems analogous to those created by the introduction of industrial-waste gasses into the atmosphere, which went unnoticed for decades. Following on models created in the 1970s, updated to take into account the hundred-fold increase in shipping noise since then, he notes that "the increase in ambient noise from shipping seems to have reduced the detectable range of low frequency whale calls from many hundreds of kilometers in the prepropeller ocean down to tens of kilometers in many settings today." (For example, the finback whale range shrank from at least 400km in the pre-engine ocean to 90km in the 1960s, down to 32km today.) He notes that, as populations of great whales fall, the separation between them may increase, with these increases in shipping noise compounding the challenge of finding mates or sharing information about active feeding grounds. However, he then goes on to point out that there is, so far, no clear evidence that the great whales do indeed communicate over long distances; clear responses to the calls of other whales have been seen only at ranges of 10km or less; the fact that a human acoustic sensor can detect a signal at 400km does not necessarily mean that the whales themselves rely on hearing signals at such distances. He suggests that acoustic tags may help to clarify whether distant, faint signals from conspecifics (whales of the same species) do in fact trigger any discernable reaction (calling in response, or moving toward the distant whale).

While noting that it may be impossible to design scientifically valid studies to uncover the possible cost of "lost opportunities" when communication is drowned out by shipping noise, an indirect way to get at this question is within reach of researchers: if animals alter their calls in noisy conditions, we can infer that the noise is disrupting their normal communication channels. And indeed, Tyack notes a long list of studies that show such changes, such as beluga whales and manatees increasing the volume of their calls in noisy conditions, and an apparently dramatic increase in the frequency (pitch) of right whale calls in sections of the ocean where low-frequency shipping noise is more intense. While noting that these and other studies "suggest that vessel noise clearly does interfere with communication in marine mammals," Tyack also notes that we do not know how costly these adaptations are, or what noise level would preclude such compensation. Also, he asks, "When does noise so degrade the usefulness of a habitat that animals leave? Can this level be predicted by the compensation behavior?" As of yet, these are unanswered, and difficult to answer, questions.

Finally, Tyack turns to research that show clear disturbance reactions to ocean noise, including killer whales staying 4km away from acoustic harassment signals near fish farms, dolphin numbers dropping to 8% of normal within 3.5 km of similar noise-makers on other fish farms (with those small numbers implying that the avoidance distance was far greater). He notes that the degree of displacement or behavioral response is not necessarily a direct indicator of the severity of impact, suggesting that "if an animal is in bad enough condition that the risk of altering behavior is high, then it may be less likely to show a disturbance response." For example, hungry animals will linger in a feeding area the longest. He also notes that some responses to noise may be caused by noise sources that resemble a predator's call (as in recent modeling Tyack has done that suggests beaked whale decompression sickness may result from a long series of near-surface dives as the whales flee sonar signals that they mistake for orca calls). He cites some compelling studies on terrestrial animals showing that repeated disturbance exacts high costs in reproductive success and overall health (including a study of geese that showed that when undisturbed, geese increased their body mass and had a 46% breeding success, whereas in nearby areas where farmers scared them off their fields, they did not gain mass and had a breeding success of only 17%).

To conclude, Tyack suggests that there are several lines of research that have so far received little attention, which could help to move key understanding of noise impacts forward, including: focusing on the most vulnerable animals as subject of study into the effects of disturbance, further study of the possibility that predator responses underlay key behavioral impacts (including fleeing, increased vigilance, and avoiding habitats), and following up on the recent theory of allostasis (behavioral changes that allow an animal to maintain equilibrium in the face of external environmental changes or stressors) as a way of understanding the costs and benefits of changing behavior in the face of noise.

Workshop Report: Encouraging Establishment of Noise-Free Zones in and Around Marine Protected Areas
Agardy, Aguilar, Canadas, Engel, Frantzis, Hatch, Hoyt, Kashner, LaBrecque, Martin, Notarbartolo di Sciara, Pavan, Servidio, Smith, Want, Weilgart, Wintle, Wright. 2007. A Global Scientific Workshop on Spatio-Temporal Management of Noise. Report of the Scientific Workshop. 44 pages. [DOWNLOAD(pdf)]
In June 2007, a workshop was held in the Canary Islands to consider the potentials for extending the management principles used in Marine Protected Areas to provide some protection from anthropogenic noise. Fundamental to the purpose and effectiveness of MPAs are "spatio-temporal restrictions" (STRs) of specific human activities: for example, excluding fishing, from a specific area (spatial restriction), or sometimes at times of special biological importance, such as spawning (temporal restriction). Few of today's MPAs are large enough to provide protection from "elevated levels of ensonification:" buffers of tens of kilometers would be necessary for protection from mid-frequency sound, and a hundred or more kilometers from low-frequency sound. Of today's 350 MPAs that include some cetacean habitat, 64 are large enough to provide some mid-frequency protection, 20 are large enough to provide at least some low-frequency protection (e.g. shipping), and only the 6 largest are probably sufficient to protect from shipping noise. (of course, shipping is not generally excluded from MPAs: this is merely a hint at the scale of noise STRs that would be useful.) The Workshop report includes several key sections:

• Descriptions of some existing attempts to provide STRs focused on noise: In 2003, Brazil established a large buffer zone around a small existing MPA, to exclude the sounds of seismic survey airguns from entering the MPA (the buffer was withdrawn by a court due to jurisdictional issues; attempts are underway to re-establish it). In the Canary Islands, a 50 nautical mile buffer zone has been established around the islands, in which active sonar is not allowed (there have been some subsequent strandings that raise questions whether this is large enough a buffer).
• Recommendations for MPA managers, centered on a framework for making management decisions regarding possible noise-related STRs.
• Recommendation that noise-producers provide more information, and longer lead times prior to operations near MPAs, to allow for analysis of effects of proposed noise. A far-reaching element of this is a call for use of "detectability curves" to reflect how easy it is to find various species of whales, and to insure that more effort is made to find hard-to-detect whales, rather than assuming that they are not present if none are found using standard observational techniques.
• Recommended measures that could reduce the noise impacts of the primary noise producing activities.
• Suggested MPAs or proposed MPAs where noise-oriented STRs could be introduced, as case studies for future MPA managment protocols. These include the PELAGOS Sancturay in the northwestern Mediterranean, off France, Italy, and Monaco, the Alborian Sea/Strait of Gibraltar, the Bay of Bengal, and East Asian waters off Japan, China, and the Philippines.

Ocean Fish Vary Widely in Startle Response to Noise
Kastelein, van der Heul, Verboom, Jennings, van der Veen, de Haan. Startle response of captive North Sea fish species to underwater tones between 0.1 and 64kHz. Marine Environmental Research 65 (2008) 369-377.
This study exposed eight marine fish species to pure tones ranging from 100Hz to 64kHz. The tests took place in specially designed quiet tanks; species were chosen in part due to their economic importance to fisheries. Some species did not respond to the sound at all, even at the highest dB levels that could be produced, while others exhibited very clear startle responses to a relatively narrow range of frequencies (generally 100-700Hz), at received levels of about 100dB (re 1uPa, rms) for the lowest frequencies, with the startle threshold increasing to the range of 160dB as frequency increased to 700Hz. Only one species responded to higher frequencies than this: Horse mackeral responses extended up to 2kHz. The species that did not startle at all were Atlantic cod, Pollack, Common eel, and Atlantic herring. Horse mackeral startled to the widest range of frequencies, with Sea bass also quite responsive; Thicklip mullet and Pout both startled to a narrower range of frequencies. Interestingly, for fish that have established audiograms, it appears that the startle response does not begin until the noise is 10-30dB above the hearing threshold; and, again, some fish showed no startle even at levels up to 45dB above their presumed hearing threshold. The researchers note that these results on captive fish to pure tones can not be reliably extraploated to wild fish in varied contexts or to more complicated sounds, but suggest that the extreme variability seen here is an important consideration in more natural situations as well. They suggest futher study using "sounds more similar to anthropogenic noise, to more complicated sounds, such as sweeps, and to the actual broad-band noise of, for instance, wind turbines and shipping..."
Related research: Some Fish Affected by Sound of Pingers
R.A. Kastelein et al. Effects of acoustic alarms, designed to reduce small cetacean bycatch in gillnet fisheries, on the behavior of North Sea fish species. Marine Environmental Research 64 (2007) 160–180
Pingers are used in large fishing nets, to alert dolphins so they avoid entanglement and drowning. This study found that some pinger signals seem to affect he behavior of Sea bass, mullet, and herring (increased swimming speeds and/or movement toward or away from surface). The researchers note that "Pingers developed to reduce odontocete bycatch should not deter the fisheries’ target species from the gillnets, and should not deter target or non-target fish from ecologically important areas such as feeding and breeding grounds, or mask their communication sounds"