Physical effects of sound exposure from underwater explosions on Pacific mackerel (Scomber japonicus): Effects on the inner ear^a)

Exposure to intense sounds is known to damage sensory hair cells (HCs) of the auditory regions of the inner ear of diverse vertebrate species from fishes to humans, resulting in hearing loss (e.g., chapters in Le Prell et al., 2012; Slabbekoorn et al., 2018a). Such hearing loss has the potential to significantly affect fitness and survival if the animal is impaired in its ability to perceive the sounds around it, such as those of predators and prey or the acoustic scene (e.g., Slabbekoorn, 2018; Slabbekoorn et al., 2018a). While most studies to date on potential effects of human-made (anthropogenic) sound have focused on mammals and birds (e.g., Halfwerk et al., 2018; Slabbekoorn et al., 2018b), there is a growing concern about potential effects of such sounds on aquatic vertebrates, including fishes (Hawkins and Popper, 2018; Popper and Hawkins, 2019; Hawkins et al., 2020).

A number of experiments have shown that long-term exposure to moderate levels of continuous sounds may cause recoverable damage to the sensory HCs of the otolith organs (saccule, lagena, and utricle) of the inner ears of some but not all fish species (reviewed in Smith and Monroe, 2016). Inner ear damage may also occur after exposure to very intense impulsive sounds, such as those from seismic air guns and pile driving (McCauley et al., 2003; Casper et al., 2013), although similar sound sources may not damage the ear in other species, even if exposed under identical conditions (e.g., Popper et al., 2005b; Casper et al., 2013).

Temporary hearing loss [referred to as temporary threshold shift (TTS)] has been demonstrated in response to exposure to continuous sounds in some fish species but not others (e.g., Scholik and Yan, 2002a). Moreover, several studies have combined hearing tests with morphological studies to show that there is a relationship between sensory HC loss and TTS (reviewed in Smith, 2016; Smith and Monroe, 2016). However, only one study has looked at hearing before and after exposure to sounds from an impulsive sound source, a seismic air gun, and found TTS in some but not all of the species tested (Popper et al., 2005b).

All of the studies to date on potential effects of impulsive sounds on inner ear tissues have focused on seismic air guns (McCauley et al., 2003; Popper et al., 2005b) and pile driving (Casper et al., 2013) since both sources are increasingly being used around the world for geologic exploration and construction. At the same time, despite explosions being used in a wide range of underwater activities, including construction, demolition of structures (e.g., oil platforms), etc., nothing is known about the potential effects of explosions on inner ear tissues of fishes. Of course, this is not particularly surprising considering that there have been relatively few studies of the effects of explosive sounds on fishes (e.g., Wiley et al., 1981; Yelverton et al., 1991; Keevin and Hempen, 1997; Govoni et al., 2008; Dahl et al., 2020; Jenkins et al., 2022). Moreover, these earlier studies have primarily focused on damage to internal organs, and none examined inner ear tissues.

One reason for the few studies of effects of explosives on fishes (or, for that matter, any aquatic species) is that logistics of working with explosives in the field are complex and require specialized assistance (e.g., in our case, from the U.S. Navy). Accordingly, the work reported here leveraged working on a broader project on the effect of sounds from explosives on fishes (e.g., Dahl et al., 2020; Jenkins et al., 2022) and added a component that examined potential effects on inner ear tissues. Thus, this report represents the first examination of the potential effects of exposure to sounds from explosives on the inner ears of a fish species.

Prior to the specifics of the paper, we provide a brief introduction to the saccular epithelium for readers who are not familiar with the inner ears of fishes. For a detailed discussion of fish ears and the sensory epithelia, see Schulz-Mirbach and Ladich (2016), Popper and Hawkins (2018), Popper and Hawkins (2019), and Schulz-Mirbach et al. (2019a). The inner ears of fishes are very similar to those of other vertebrates. They have three semicircular canals involved with detection of angular acceleration and three otolithic end organs, the saccule, lagena, and utricle, which are thought to be involved in detection of position relative to gravity, as well as hearing (e.g., Platt, 1983; Popper and Hawkins, 2018). Each otolithic end organ has a sensory epithelium (often called the macula) on its wall and a dense calcareous otolith lying in the end organ lumen and connected to the epithelium by a thin otolithic membrane.

Each epithelium contains sensory HCs (the actual number depends on fish size and can be from thousands to hundreds of thousands, e.g., Lombarte and Popper, 1994) surrounded by supporting cells. The HCs have an apical ciliary bundle that projects into the lumen of the otolith organ. The cilia, which are made of β- and γ-isoforms of the protein F-actin, project below the cell surface into a dense cuticular plate made of the same material (Furness and Hackney, 2006). Relative motion between the sensory epithelium and the far denser overlying otolith during sound exposure results in bending of the ciliary bundle (e.g., Popper and Hawkins, 2018). This produces chemical and electrical events that send neural signals about the sound to the brain (Popper and Fay, 1999).

This study used the general methods and approach developed by Dahl et al. (2020) for studies on Pacific sardines (Sardinops sagax). The approach was slightly modified by the same group of investigators for studies with Pacific mackerel (Scomber japonicus), the species used in this report (Jenkins et al., 2022). For details of fish handling, sound exposure, and other experimental details, please see Jenkins et al. (2022).

In brief, studies were performed at sea using test cages placed at seven distances between 35 and 800 m (a total of seven distances were used over the course of the study) from a single explosion. Five trials (each with a single explosion) were completed over five days. Immediately following each explosion, fish were sacrificed and necropsied to determine effects on external and internal tissues (reported by Jenkins et al., 2022). For the work reported here, inner ears of Pacific mackerel were prepared for microscopic examination from either one or two specimens from a subset of these cages from different distances from the explosions.

Fish were exposed to explosive detonations between September 10 and 18, 2019, within the U.S. Navy's Silver Strand Training Complex, located 5 km offshore from San Diego, CA. The immediate area has a seabed largely consisting of unconsolidated sandy sediments.

The explosions were conducted by Navy explosive ordinance disposal personnel in accordance with the U.S. Navy's Hawaii-SOCAL Environmental Impact Statement (www.hstteis.com) and associated permits. A single explosive detonation consisting of 4.5 kg of C-4 [equivalent to 6.2 kg of trinitrotoluene (TNT)] placed mid-column (at approximately 10 m depth) was conducted each test day.

Pacific mackerel (N = 363 for the overall study, mean standard length ± standard deviation = 211.6 ± 16.8 mm, and mean mass ± standard deviation = 106.8 ± 27.9 g) were wild-caught and held in covered, netted, open ocean enclosures in San Diego Bay. The fish were fed and cared for daily for at least 3 weeks to acclimate them to captivity before the experiment.

All aspects of fish collection, transport, and holding/husbandry were conducted under Scientific Collection Permit No. SC-13924, issued by the California Department of Fish and Wildlife. All procedures were approved by the Naval Information Warfare Center Pacific Institutional Animal Care and Use Committee (IACUC Protocol No. 131–2018), and the Navy Bureau of Medicine and Surgery. The study followed all applicable U.S. Department of Defense guidelines for the care and use of laboratory animals.

Details of fish exposure to explosives are described in Dahl et al. (2020) and Jenkins et al. (2022). Briefly, fish were taken by the primary experimental vessel to the experimental site in barrels of aerated sea water. Randomly selected animals were then placed in cages made of polyester mesh netting suspended within a frame of polyvinyl chloride (PVC) tubing and lowered to a depth of 10.5 m (±0.5 m; see Dahl et al., 2020 for details and pictures). Cages were placed at various distances from the explosive source from 35 to 800 m, and a subset of those ranges was used for this study (Table I). The minimum (from 800 m cages) and maximum (from 35 m cages closest to the detonation) peak pressures, sound exposure levels (SELs), and pressure impulses were 217 and 252 dB re 1 μPa, 193 and 215 dB re 1 μPa² s, and 90 and 1305 Pa s, respectively (Jenkins et al., 2022). Additional control fish were treated identically to the exposure animals, but the cages were removed from the water just prior to the explosion. Shortly after the explosions, the cages were retrieved, the animals removed, and two from each cage were randomly selected and prepared for the ear study.

The fish for the ear study were euthanized using buffered MS-222 (tricaine methanesulfonate, 250–500 mg/L) immediately after being brought to the surface. Ten minutes after cessation of all movement, the animals were decapitated. Heads were separated from the rest of the body far enough behind the opercula to ensure that the ears were not damaged. Excess tissue, including the lower mandible and snout, was trimmed away. The remaining tissues were placed in jars containing 4% paraformaldehyde (Electron Microscopy Sciences, Catalog No. 15710-S, CAS No. 30525–89-4) in filtered San Diego Bay seawater (0.45-μm filtration pore size; pH = 7.2–7.4) at an approximately 20:1 v/v ratio. A pilot study found that this fixative recipe produced the best results for subsequent microscopic examination.

Further work could not be done on the boat for safety purposes. However, within 2 h of the animals getting initial fixation, they were in the land-based laboratory. There, the ears were additionally preserved by using a syringe to inject 10–20 ml (depending on head size) of fixative into the cranial cavities. Injection stopped when the eyes bulged, indicating that the cranial cavity was filled with fixative. The heads were then placed into fresh fixative and left overnight. Fixative was again refreshed, jars carefully labeled, packed, and shipped overnight to the laboratory at Western Kentucky University (Bowling Green, KY) for microdissection, preparation, and examination of whole saccules by M.E.S. M.E.S. did not participate in collection of tissue other than at the beginning to train others, and he was not given the test conditions for each sample until the tissue analyses were completed.

Thirty-three fish (none of which were used for general necropsy) were used for inner ear microscopy from the total set of specimens that were examined. Of these, there were four fish from each of the sampled distances from the explosion (see Table I) except for the exposures at about 100 and 400 m, which had three and two fish, respectively. There were eight control fish.

In final preparation for ear tissue examination, the heads were washed in tap water for at least 30 min and the inner ears removed under a dissecting microscope. The right and left saccules were then trimmed and incubated for 30 min in Alexa Fluor 488 phalloidin (Molecular Probes/Invitrogen, Carlsbad, CA), which stains the F-actin that is found in HC ciliary bundles and subsurface cuticular plates (see results for a description of the sensory cells being considered). Saccules were mounted whole under a coverslip with Prolong Gold antifade reagent (Thermo Fisher Scientific, Waltham, MA) with 4′,6-diamidino-2-phenylindole (DAPI) to stain nuclei.

Images of the saccule were taken with 10×, 20×, and 100× objectives using a Zeiss Axioplan 2 (Carl Zeiss GA, Oberkochen, Germany) epifluorescent microscope and a Zeiss MRm digital CCD camera. HC bundle counts were obtained in ten predetermined locations (2500 μm² boxes) across the length of the saccule using ImageJ software (Fig. 1). This methodology to examine rostral-caudal shifts in saccular HC bundle density has been used previously for fishes (Popper and Hoxter, 1984; Lombarte and Popper, 1994; Mann et al., 2001; Smith et al., 2006) and included some counting locations that were on the edges of the saccule, as well as the midline to obtain greater resolution of HC loss across the epithelia. The precise orientation of the counting boxes varied slightly between saccules because all of the saccules could not be mounted on the microscope slides at exactly the same angle. Nevertheless, the counting locations remained in the same area along the rostral-caudal axis and HC bundle counts remained consistent across saccules. Images of each saccule were merged in Photoshop CS2 (version 9.0.2; Adobe Systems, Santa Clara, CA) or PowerPoint (Microsoft, Redland, WA) to see larger potential patterns of damage.

To confirm the nature of the issue appearing in relatively low magnification studies of the phalloidin-labeled tissue, the saccules of two fish (one control and one exposure at 35 m from the source) were examined using scanning electron microscopy (SEM; JEOL 6510LV, JEOL USA Inc., Peabody, MA). Tissue was fixed as described for the phalloidin staining and then put through a dehydration series in ethanol. The saccules were mounted on SEM stubs and gold-palladium sputter-coated using a vacuum evaporator (Emscope SC500, EMZER Technological Solutions, S.L., Barcelona, Spain) and then examined with the SEM.

Preliminary analysis showed no significant differences between right and left saccules, therefore, data for both were averaged for each individual fish to avoid pseudoreplication in subsequent analysis. Analysis of variance (ANOVA) was used to test for differences in HC density between saccular locations in control fishes. We used a general linear model to determine whether HC bundle loss varied significantly by saccular region and distance from the experimental explosion. Since there was significantly more HC damage with treatments 400 m and closer to the detonation compared to treatments at greater distances, we separated treatments into these two groups and performed additional specific contrasts with treatment distance (near versus far) and saccule location as factors. In addition, 70 specified, Bonferroni-corrected contrasts were made comparing controls to experimental fish for each treatment at each saccular location (7 treatments × 10 locations). An initial statistical analysis was performed with SYSTAT 13.1 (SYSTAT Software, Inc., Chicago, IL). Specific post hoc contrasts were performed with R Studio (version 4.2.0; RStudio, Boston, MA) using the emmeans package following the lmerTest package.

HC bundle densities from control fish varied at the different locations quantified on the saccule, ranging from a mean (± standard error; S.E.) minimum HC bundles per 2500 μm² of epithelia of 37.1 ± 1.2 at location 5 and a maximum of 111.3 ± 4.9 at location 2 (Fig. 2). Two general patterns were evident in saccular HC bundle densities in all of the fish. First, the number of HC bundles was greater toward the rostral and caudal ends compared to the central saccule (locations 1–4, 8–10 versus 5–7; P < 0.001). Second, HC bundle densities were greater on the edges of epithelia (i.e., in locations 2, 4, 8, and 9) compared to locations along the midline of the epithelia (i.e., in locations 1, 3, 5, 6, 7, 10; P < 0.001).

A similar pattern of HC density was found in all of the fish that had been exposed to the explosion, except that HC bundles were often damaged and less dense, suggesting HC loss (Fig. 2). This is based on previous studies that showed a correlation between HC bundle damage and whole HC loss (e.g., Smith et al., 2006).

On higher magnification examination of the phalloidin-labeled saccular epithelia of fish exposed to the explosion, the damaged areas were often obvious (Fig. 3). In these areas, HC bundles were sparse, and holes were evident in the epithelial tissue, which appeared as darkened areas [Fig. 3(B)]. Some HC bundles were completely missing, exposing the underlying cuticular plates, while some of the other remaining HCs exhibited ragged, splayed, and fractured stereocilia (Fig. 4). In addition, small circular actin scars were found, presumably where HCs used to be [Figs. 4(C) and 4(D)].

To confirm the nature of the damage reported in the phalloidin-labeled tissue, we used higher power SEM examination of saccules from one control fish (913B1) and one fish exposed at 35 m from the source (918R2). Indeed, although this tissue was not originally prepared using fixation protocols for SEM examination but for phalloidin labeling, the quality of the tissue was sufficient to see qualitative differences between the control and treated fish, and fixation artifacts did not interfere with interpretation of the tissue.

The SEM confirmed the observations on phalloidin-labeled tissue that the control saccule had HCs with bundles that were generally intact while the 35 m treated saccule had large swaths of damaged or missing HC bundles (Fig. 5). This sensory HC damage included shortened and fused stereocilia and HC bundles missing some or all of their stereocilia (Fig. 6). At higher magnification, divots were evident on the surface of the macula where HCs presumably were located [Figs. 5(F) and 6(D)]. Thus, SEM ultrastructure confirms the validity of using fluorescent phalloidin labeling for these kinds of studies as loss of sensory HCs is detectable with both techniques.

Prior to testing the effect of explosion exposure on sensory HC density, a quantile-quantile plot and Kolmogorov-Smirnov test (P = 0.076) was used to confirm the assumption of normally distributed data required for parametric statistics. A general linear model using Type III Sum-of-Squares was then used to test the effect of treatment (distance from the detonation) and location on the saccule on sensory HC density. HC bundle densities varied significantly by treatment and saccular location with a significant interaction between these two factors (P < 0.0001; Table II).

As our original general linear model showed that location on the saccule played a large explanatory role in density of HCs (F-ratio was 152.5 for location compared to 19.9 for treatment; Table II), and since we were more interested in the effects of the explosion on HC numbers than the natural variation in HC density normally found in different locations on a fish saccule, we converted our raw HC counts to percent HC loss by subtracting the mean HC density for a given saccular location of control fishes from each measurement made of explosion-treated fish. These percentage ranges are visualized in Fig. 7 and were used in post hoc contrasts. These contrasts showed that HC loss was significant primarily in the rostral saccule and only at distances of 400 m or closer to the detonation (Fig. 7). It must be noted that the p-values of these contrasts were Bonferroni-corrected, therefore, with 70 separate contrasts, our p-values are highly conservative estimates.

Although HC bundle loss was often evident in the rostral and caudal ends of the saccule, it was rare in the central saccule (Fig. 7). However, two treatments (35 and 325 m) exhibited HC bundle loss over a larger range of saccular locations with HC loss often exceeding 30% (Figs. 7 and 8). The greatest mean HC loss found was a 71% loss at location 3 in the 325 m treatment saccules [Fig. 8(B)].

As HC bundle damage was generally limited to a range of 400 m and closer to the explosion (see Fig. 7), we separated the percent HC loss data into two groups (treatments between ranges of 35 and 400 m and 600 and 800 m), and examined patterns of HC damage by saccular location. There was a significant overall effect of range group (i.e., distance from the explosion; P < 0.001) and saccule location (P < 0.001). HC loss was greater for the 35–400 m ranges than the two further ranges, but this difference was only significant at saccular locations 1 and 3 (P < 0.05), which are centrally located in the rostral end of the saccule (Fig. 9). Using 400 m as the range limit for significant HC damage, it appears that acoustic measures greater than approximately 220 dB re 1 μPa peak sound pressure level, 225 dB re 1 μPa² s SEL [Fig. 10(A)], and 100 Pa-s pressure impulse [Fig. 10(B)] were necessary to produce explosion-induced HC damage.

This study is the first to examine the potential for fish inner ear damage resulting from exposure to underwater explosions. The most critical finding is that caged Pacific mackerel exposed as far away as 400 m to 4.5 kg of C4 explosives in a shallow-water environment show statistically significant explosive-induced HC damage. The finding of HC damage is not particularly surprising as sensory HC loss has been found in some but not all of the species exposed to a number of different acoustic stimuli, including tones (Enger, 1981; Hastings et al., 1996; Smith et al., 2011), broadband white noise (Smith et al., 2006), and impulsive anthropogenic sources such as seismic air guns (McCauley et al., 2003) and pile driving (Casper et al., 2013).

Besides being the first investigation of the effects of explosions on fish ears, this is also the first investigation of auditory HCs in the inner ear of the Pacific mackerel. Thus, we briefly discuss our observations as part of the interest in comparative inner ear issues in fishes. Very little research has been conducted on the ears of fishes from the family Scombridae. The inner ear sensory epithelia of two other related scombrids, skipjack tuna (Katsuwonus pelamis) and bluefin tuna (Thunnus thynnus), have also been described (Popper, 1981; Song et al., 2006). In addition, ionocytes, specialized cells that maintain optimal ionic conditions, have recently been characterized in the inner ear epithelium of Pacific mackerel (Kwan et al., 2020).

HC densities in Pacific mackerel generally ranged between 50 and 100 HCs/2500 μm² (Fig. 2). These results are similar to those reported for bluefin tuna (29 915 HCs/mm² = 75 HCs/2500 μm²; Song et al., 2006). In addition, patterns of HC density varied by saccular location in bluefin tuna, as we found in Pacific mackerel, with centrally located epithelia exhibiting the lowest density of HCs with shorter stereocilia and more peripheral areas exhibiting higher densities with HCs with longer kinocilia and stereociliary bundles.

A similar pattern is also found in non-scombrid fishes. For example, HC densities of the saccules of goldfish (Carassius auratus) and zebrafish (Danio rerio) were greatest in the rostral and caudal ends and the least centrally (Platt, 1993; Schuck and Smith, 2009; Smith et al., 2011), and HC density was higher closer to all edges of the saccular epithelium in a cichlid, the oscar (Astronotus ocellatus; Popper and Hoxter, 1984). However, with HC counts being recorded in so few teleost species, any generalizations about HC density variance must be made with some caution.

Explosion-induced damage to the ear, as discussed in this paper, is likely from overstimulation of the ciliary bundles because of the intense impulsive sounds. Damage may be caused by a shearing of the cilia, leaving the apical cell membrane and cuticular plate intact (e.g., Schuck and Smith, 2009). It is also possible that the signal results in sufficient mechanical stimulation to remove the entire HC, resulting in holes in the epithelium. A key point to be reiterated, however, is that these injuries documented at the cellular level may accompany an array of physical injuries to abdominal organs observed in the companion necropsy study (Jenkins et al., 2022).

Explosion-induced damage to the sensory HCs of the saccules of Pacific mackerel were not evenly distributed across the sensory epithelia but were localized to general regions. Specifically, damaged areas were greatest in the rostral and caudal ends of the saccule and minimal in the central region (Figs. 7 and 8). Moreover, the rostral end appears to be more prone to damage than the caudal (Fig. 9). A similar trend was found in the saccules of Mozambique tilapia (Oreochromis mossambicus), which had been exposed to pile driving sounds (Casper et al., 2013). In that study, mean HC densities were slightly reduced in exposed fishes relative to controls in the rostral and caudal regions but not in the most central region (although low levels of HCs with sparse or missing stereocilia were found centrally), and the reduction was greatest on the rostral end.

The explanation for the intra-epithelial differences in explosion-induced HC damage, supported by their consistency between fishes at different distances from the explosion, may lie with the mechanics of the otolith organs of the inner ear, which function like accelerometers for detecting particle motion (de Vries, 1956; Popper and Hawkins, 2018; Schulz-Mirbach et al., 2019a). In each end organ, the relative motion between the dense overlying otolith and sensory epithelium during sound stimulation move the ciliary bundles on the HCs. Although the mechanics of the inner ear are not well understood in fishes, there is some evidence that the pattern of movement of the otolith in the sound field is, at least in part, affected by the otolith shape, the fluids surrounding the otolith, and constraints imposed by the connection between the otolith and epithelium (de Vries, 1950; Sand and Michelsen, 1978; Schulz-Mirbach et al., 2018).

Assuming this general mechanism, it is likely important that in most fishes, the saccular otolith (the sagitta) has a complex shape (e.g., Carlstrom, 1963; Popper et al., 2005a; Schulz-Mirbach and Ladich, 2016; Schulz-Mirbach et al., 2019b), and this could lead to complex otolith motions. This holds true for the sagitta of Pacific mackerel (Fig. 11) and other scombrids, such as Pacific sierra (Scomberomorus sierra), with a club-like shape in which the caudal end is much thicker and more robust than the rostral end (Gallardo-Cabello et al., 2011).

Moreover, the caudal end of the sagitta has a deep sulcus in which the caudal saccule sits, while the rostral sagitta is more open with a shallower sulcus. Thus, while there is a lack of experimental evidence for Pacific mackerel (or any other fish species), we hypothesize that an explanation of why the rostral saccule may be more prone to damage than the caudal is that the deeper sulcus of the caudal saccule may provide a stronger otolithic membrane to otolith connection, minimizing the amount of shearing forces on the HC bundles in this area compared to the rostral area, owing to the relative motion between the otolith and epithelium. However, estimating this relative motion from the peak acoustic horizontal displacement on the order of 100 μm requires detailed knowledge of tissue impedances and is beyond the scope of this study.

An important issue to consider is whether the damage to the sensory epithelia for Pacific mackerel will affect hearing ability. Indeed, any hearing loss could decrease the animals' ability to detect sounds of predators and prey and the overall acoustic scene (Hawkins and Popper, 2018; Popper and Hawkins, 2019), and this could decrease fitness. Alternatively, there may be no long-term effects of hearing loss since fishes are able to regenerate the sensory cells of the ear over time following damage (e.g., Lombarte et al., 1993).

Still, even temporary hearing loss could impact fitness and survival. Indeed, there is substantial evidence that fishes exposed to sufficiently high intensity sounds will exhibit TTSs (e.g., Scholik and Yan, 2002b,c; Smith et al., 2004a,b; Popper et al., 2005b), and this loss tends to be a linear function of the sound pressure level of the stimulus (Smith et al., 2004a). Popper et al. (2005b) found an increase in TTS with increased sound pressure level in lake chub (Carcharhinus plumbeus), northern pike (Esox lucius), and broad whitefish (Coregonus nasus) exposed to seismic airgun shots. Subsequently, it was found that there was a significant relationship between HC loss and hearing loss in cyprinid fishes (goldfish and relatives; Smith, 2016).

Based on these observations, we expect that Pacific mackerel with damage to the inner ears would exhibit TTS. Moreover, the loss is likely to increase with decreasing range of the fish to the detonation given pressure goes as R^−1.13 (e.g., Wilson et al., 2020). However, without specific data on hearing sensitivity of Pacific mackerel, it is impossible to try and predict the relationship between sound exposure and TTS or the duration of TTS from exposure to explosive sounds.

While there is clearly significant sensory HC damage occurring at ranges up to 400 m, it is important to evaluate this in context of the overall fitness of the Pacific mackerel. In other words, HC damage could affect the ability of the fish to detect important acoustic signals in its environment. However, this potential decrement in hearing would not be as important to the Pacific mackerel as other physical injuries observed in response to the explosive exposure. These injuries included swim bladder bruising and rupture and kidney rupture, with the latter two being designated as severe injuries (Jenkins et al., 2022). While there is evidence that fishes exposed to pile driving could recover from some of these injuries in the laboratory (Casper et al., 2013), it is likely that suffering these injuries in the wild would make the fishes susceptible to predation or possibly infection regardless of whether or not the fish also suffered TTS.

The severe injuries were observed in the fish at levels significantly above control at ranges up to 325 m, with percent of injuries increasing as fish were closer to the explosive source (Jenkins et al., 2022). Significant HC bundle loss was seen at the 35 m range, where percent physical injury was also high. Approximately one-third of the fish at this range were also dead on retrieval, meaning that when the cages were brought to the surface following the exposure, the fish had no opercular movement and did not respond to physical stimulation.

While there is an overlap of significant HC loss and severe physical injuries in Pacific mackerel at ranges up to 325 m, there was still evidence of HC loss at 400 m. This suggests that there is a range away from the explosive in which the fish may not sustain potentially life-threatening injuries but still show damage to the inner ear HCs. If there is a correlation between HC loss and hearing abilities (see Smith, 2016), then fishes at this distance could still have reduced fitness. This is an important finding as it extends the range of influence that explosive exposure has on injuries in the Pacific mackerel.

A critical related point is that our results reflect HC damage and loss within an hour after exposure to the explosive sound. It is very possible that had we been able to keep the fish alive for extended periods of time post-exposure, as was done by McCauley et al. (2003) after exposure to seismic air guns or longer exposure to lower level sounds, we might have seen greater damage. For example, cell death increased in the lagena of the inner ear of goldfish 24 h after 48 h of noise exposure, showing that acoustically induced damage in the fish ear is not always instantaneous (Smith et al., 2006). At the same time, since fishes replace damaged sensory cells, there might have been repair that would replace damaged cells (e.g., Lombarte et al., 1993) had we sampled the Pacific mackerel days after exposure to the explosion.

The results for Pacific mackerel demonstrate that in addition to internal organ damage to the fish, the HCs of at least the saccule can also be damaged by exposure to explosive energy. It has been suggested that internal damage results from motions of the walls of the air-filled swim bladder, resulting from intense impulsive stimulation, and that these motions impinge on and damage near-by tissues (Halvorsen et al., 2012a; Halvorsen et al., 2012b; Dahl et al., 2020). The mechanism for inner ear damage, in contrast, is likely due to the relative motion between the overlying dense otolith and the sensory epithelium.

A critical question then is whether the ear damage results can be extrapolated to other species. While it is likely that the extent of damage will be species specific, the ears of all teleost species have the same basic configuration as seen in the Pacific mackerel. Thus, it is reasonable to suggest that the same sound source would have resulted in some inner ear damage had we used most any other species. Moreover, the extent of inner ear damage may not be correlated with other tissue damage that is related to proximity to the swim bladder (Dahl et al., 2020). Indeed, we can speculate that even species without a swim bladder, which have been shown not to show any internal damage even at very high levels of impulsive signals that affected species with a swim bladder (Govoni et al., 2008; Halvorsen et al., 2012a), would still show inner ear damage.

At the same time, it is possible that inner ear damage would be even greater in species that have air bubbles within the inner ear, as in the Clupeiformes (herrings, shads, sardines; Denton and Blaxter, 1976). Moreover, there are species where the swim bladder has an anterior extension that comes into close contact with the ear region of the head, and it is possible that re-radiated energy from the swim bladder would indirectly cause excessive motion of the otolith relative to the rest of the inner ear, thereby resulting in inner ear damage.

While our observations suggest that impulsive signals from explosions can damage the inner ears of fishes, it is also important to keep in mind that it is very difficult to extrapolate from our results to other species from the perspective of acoustics. Although our study spanned several range-to-depth ratios (∼1–40) only one bottom type could be considered, and fish were at a single nominal target depth representing mid-water in relatively shallow water. What the results would be like in other acoustic environments and other test depths is not possible to predict without considerably more data.

In summary, underwater explosions can damage the auditory HCs of Pacific mackerel and the relationship between the extent of damage and distance from the detonation (and, thus, strength of the stimulus) was not a very linear and predictable relationship in the test environment. Previous studies on other fishes suggest that sensory HC damage leads to hearing loss, which could have negative effects on the fitness of the fish. Although significant saccular damage was found at 400 m and closer to the detonation, it should be noted that there are many other potential impacts of this explosive sound on these fishes (e.g., damage to internal organs and swim bladder; Jenkins et al., 2022, and behavioral responses). For example, bodily injuries pose a greater threat than HC damage for fish located close to the detonation (35 m); but for fish located far enough away from the explosion to avoid more serious life-threatening injuries, the potential effects of HC damage should be considered. Thus, the impact of these explosive sounds should be considered when environmental impacts are estimated for marine construction projects, such as the demolition or building of structures such as oil platforms or wind farms.

We thank Christiana Boerger for her very substantial help in developing this overall project. We thank Dr. John Andersland and Stephanie Lawrence for their assistance with SEM and Dr. Jarrett Johnson for statistical expertise with the specific linear contrasts. This project could not have been completed without aid from Jennie Shield, Michelle Tishler, Kevin Carlin, Dana Schrimpf, Maria Zapetis, and Cameron Martin, and the overwhelming support of several other Naval Information Warfare Center (NIWC) Pacific teams: facilities (Depot), veterinary technicians, and the ROV team. The authors thank Explosive Ordinance Disposal Mobile Unit Three Detachment Southwest, the U.S. Third Fleet, and the U.S. Pacific Fleet for enabling a safe and successful field effort. The views expressed in this publication reflect the results of research conducted by the author(s) and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, or the U.S. Government. Some of the authors are employees of the U.S. Government. This work was prepared as part of their official duties. Title 17 U.S.C. Sec. 105 provides that “Copyright protection under this title is not available for any work of the United States Government.” Title 17 U.S.C. Sec. 101 defines a U.S. Government work as a work prepared by an employee of the U.S. Government as part of that person's official duties.