S. J. Morreale Department of Natural Resources, Cornell University, Fernow Hall, Ithaca, New York 14853

Underwater, low-frequency noise in a coastal sea turtle habitat Y. Samuel a) Department of Earth and Atmospheric Sciences, Ocean Resources and Ecosystems Program, Cornell University, Snee Hall, Ithaca, New York 14853 S. J. Morreale Department of Natural Resources, Cornell University, Fernow Hall, Ithaca, New York 14853 C. W. Clark Bioacoustics Research Program, Laboratory of Ornithology, Cornell University, Ithaca, New York 14853 C. H. Greene Department of Earth and Atmospheric Sciences, Ocean Resources and Ecosystems Program, Cornell University, Snee Hall, Ithaca, New York 14853 M. E. Richmond U.S.G.S., New York Cooperative Fish and Wildlife Research Unit, Department of Natural Resources, Cornell University, Fernow Hall, Ithaca, New York 14853 Received 16 June 2004; revised 12 November 2004; accepted 22 November 2004 Underwater sound was recorded in one of the major coastal foraging areas for juvenile sea turtles in the Peconic Bay Estuary system in Long Island, New York. The recording season of the underwater environment coincided with the sea turtle activity season in an inshore area where there is considerable boating and recreational activity, especially during the summer between Independence Day and Labor Day. Within the range of sea turtle hearing, average noise pressure reached 110 db during periods of high human activity and diminished proportionally, down to 80 db, with decreasing human presence. Therefore, during much of the season when sea turtles are actively foraging in New York waters, their coastal habitats are flooded with underwater noise. During the period of highest human activity, average noise pressures within the range of frequencies heard by sea turtles were greater by over two orders of magnitude 26 db than during the lowest period of human activity. Sea turtles undoubtedly are exposed to high levels of noise, most of which is anthropogenic. Results suggest that continued exposure to existing high levels of pervasive anthropogenic noise in vital sea turtle habitats and any increase in noise could affect sea turtle behavior and ecology. 2005 Acoustical Society of America. DOI: 10.1121/1.1847993 PACS numbers: 43.80.Nd, 43.30.Nb, 43.50.Rq WWL Pages: 1465 1472 I. INTRODUCTION Underwater noise levels have increased dramatically in recent decades due to anthropogenic sources, such as commercial, industrial, and recreational maritime activities Richardson et al., 1995; Curtis et al., 1999; Andrew et al., 2002. Notably, a predominant component of sounds from these sources is in the low-frequency range less than 1000 Hz. Currently, low-frequency noise is of great concern for sea turtles because their hearing range of highest sensitivity is confined to low frequencies Ridgway et al., 1969; Bartol et al. 1999, and they are endangered or threatened worldwide Plotkin, 1995. Furthermore, sea turtles often aggregate in coastal areas where human activity, and therefore anthropogenic disturbance and underwater noise, is greatly heightened Morreale and Standora, 1998. Our study focused on quantifying noise in a sea turtle foraging habitat, and identifying the anthropogenic component of underwater noise. Many previous studies have linked anthropogenic noise to effects on the natural ecology of marine organisms. While a Electronic mail: ys88@cornell.edu some of these marine organisms have a different hearing sensitivity than sea turtles, and while some of the noise exposures are to different kinds of anthropogenic noise that are relevant to sea turtles, we present these studies here to demonstrate how different sources of anthropogenic noise have been seen to affect marine animals. Among the higher vertebrates, it has been shown that whales exhibit strong avoidance reactions to the sounds from oil and gas exploration activities Malme et al., 1983 and to seismic exploration noises Richardson et al., 1986; Richardson, 1997. In addition, beluga whales Delphinapterus leucas have been observed to decrease their call rates in response to boats moving closer, presumably responding to the associated noise Lesage et al., 1999. Killer whales Orcinus orca off British Columbia, Canada, when approached by motorboats, adopted a less predictable pathway; female killer whales in particular responded by swimming faster and increasing the angle between successive dives Williams et al., 2002. Further, impact models of boat noises on killer whales in the same area were shown to interfere with killer whale communication, cause behavioral avoidance, and cause a temporary threshold shift TTS in hearing Erbe, 2002. When exposed to low-frequency sonar broadcasts, humpback whales J. Acoust. Soc. Am. 117 (3), Pt. 1, March 2005 0001-4966/2005/117(3)/1465/8/$22.50 2005 Acoustical Society of America 1465

Megaptera novaeangliae often responded by singing longer songs, possibly to compensate for acoustic interference Miller et al., 2000; Fristrup et al., 2003. Among fish, several species, including Pacific herring Clupea harengus, cod Gadus morhua, and capelin Mallosus villosus, have been shown to react to noise stimuli by increasing their swimming speed Olsen et al., 1983, by swimming downward Suzuki et al., 1980, or by moving away from the sound sources Blaxter et al., 1981; Schwarz and Greer, 1984; Engås et al., 1998. Sound can have physical effects, too, causing measurable damage to sensory cells in the ears of fishes Hastings et al., 1996; McCauley et al., 2003. Adverse affects of underwater noise have been demonstrated for invertebrates as well. When exposed to higher pressure levels of noise in an experimental setting, brown shrimp Crangon crangon exhibited increased aggression and higher mortality rates, as well as significant reductions in their food uptake, growth, and reproductive rates Lagardère, 1982. For sea turtles, several studies have identified their ability to perceive low-frequency sounds Ridgway et al., 1969; Lenhardt et al., 1983, 1996; O Hara 1990; Bartol et al., 1999. Hearing range coincides with the predominant frequencies of anthropogenic noise, increasing the likelihood that sea turtles might experience negative effects from noise exposure. At present, sea turtles are known to sense lowfrequency sound, however, little is known about the extent of noise exposure from anthropogenic sources in their natural habitats, or the potential impacts of increased anthropogenic noise exposure on sea turtle biology. Much of the acoustic research on sea turtles has focused on studying sea turtle ear anatomy and auditory sensory capabilities. These studies clearly demonstrated that sea turtles are able to detect and respond to sounds, and that their hearing is limited to low frequencies, with the range of highest sensitivity between 200 and 700 Hz, and with a peak near 400 Hz Ridgway et al., 1969; Bartol et al., 1999. A related study has shown that when presented with acoustic stimuli at 430 Hz and 1.5 db re 1, sea turtles placed in 50-gallon tanks respond with abrupt body movements, such as blinking, head retraction, and flipper movement, all of which were interpreted as startle responses Lenhardt et al., 1996. Similar responses were observed when sinusoidal stimuli at 250 and 500 Hz within the range of 55 59 db acceleration re 10 3 m/s 2 ) were directly transmitted to the head of sea turtles, which were submerged in 0.5 m of sea water in 1.44- m-diam tanks Lenhardt et al., 1983. In a separate study, higher level responses, such as changes in swimming patterns and orientation, were noted when sea turtles in a confined canal 300 m long, 45 m wide, and up to 10 m deep were subjected to high-pressure air gun pulses 120 db re 1 bar at 1 m, suspended at 2-m depth and positioned 33 m inward from one side of the tank, with frequencies ranging from 25 to 750 Hz O Hara, 1990. In the current study, we focused on a near-shore habitat in New York state where there is great overlap between sea turtle activity and human activities that generate high levels of low-frequency noise. Each year, juvenile sea turtles, mostly loggerhead sea turtles Caretta caretta, kemp s ridley sea turtles Lepidochelys kempii, and green sea turtles Chelonia mydas, are found in New York waters and the Peconic Bay Estuary system between July and October Morreale and Standora, 1998. These very productive waters provide important developmental habitat for sea turtles during their early life stages. However, because this region also hosts some of the highest human densities along the coast of the western North Atlantic, as well as shipping lanes and major ports, foraging young sea turtles can be exposed to high levels of human activities that pose a risk of potential disturbance. Furthermore, a major portion of the sea turtle coastal foraging season occurs during the summer months, when recreational boating activities are highest. We report here the results of a seasonal study designed to record underwater sound and to determine the noise levels to which sea turtles are exposed in one of their major coastal foraging areas in the western North Atlantic. The primary objective was to characterize and quantify the noise levels from both ambient and introduced sounds in the underwater environment, especially sounds within the sea turtle hearing range. The secondary objective was to distinguish and quantify the anthropogenic component of the noise to which sea turtles are exposed in their foraging habitat. This study was complementary to a separate behavioral study Samuel, 2004 conducted in experimental tanks, which was designed to detect behavioral responses from sea turtles exposed to sounds similar to the ones recorded in the natural environment. II. METHODS A. Overview Research was conducted in the Peconic Bay Estuary system in Long Island, New York from 20 July to 27 September 2001, a period that coincides with the sea turtle activity season in inshore waters. Over the span of the sea turtle foraging season, sounds were recorded in the underwater environment in Southold Bay 41 03 N, 72 24 W, one of the estuary s major foraging areas for juvenile sea turtles. Simultaneous with sea turtle activity, the waters within the estuary system are the site of much boating and recreational activity during the summer period, starting around Independence Day 4 July and ending on Labor Day first Monday in September. However, by early September, after a final burst of human activity on the Labor Day holiday, recreational boating drops off abruptly. These factors combine for a natural experiment in which the sea turtle season can be separated into five different categories, each of which is characterized by a distinctly different level of human motor-boating activity. The presumption is that the five different levels of human boating activity generate different levels of underwater noise that can be measured and compared. The five levels, ranging from periods of highest to lowest human activity were 1 daytime summer weekends before Labor Day; 2 daytime summer weekdays before Labor Day; 3 daytime weekends after Labor Day; 4 daytime weekdays after Labor Day; and 5 nighttime after Labor Day. 1466 J. Acoust. Soc. Am., Vol. 117, No. 3, Pt. 1, March 2005 Samuel et al.: Underwater noise in sea turtle habitats

TABLE I. Sampling design of the 14 different recording sessions conducted between 20 July 2001 and 27 September 2001. Recording sessions were conducted both before and after Labor Day 3 September 2001. Sessions were further divided into weekday and weekend recordings. In addition, there was one recording conducted during nighttime after Labor Day. All recordings were 1 h long. Time indicates the beginning of each 1-h recording session. Before Labor Day After Labor Day Date Day Time Date Day Time Weekday 20 July Friday 13:05 3 September Monday 15:05 7 August Tuesday 14:40 6 September Thursday 16:15 15 August Wednesday 14:45 18 September Tuesday 14:50 26 September Wednesday 12:45 Weekend 22 July Sunday 16:30 9 September Sunday 15:10 18 August Saturday 14:40 22 September Saturday 15:25 25 August Saturday 15:10 1 September Saturday 14:40 Night 27 September Thursday 19:55 B. Sampling design Monitoring of underwater sound and collection of acoustical data occurred on 14 separate days, spanning all five levels of human activity. Seven recording sessions were conducted during the two summer months of high human activity before Labor Day, and seven sessions were conducted during periods of lower human activity, after the Labor Day holiday. Recording sessions for high and low human activity were further divided into daytime weekdays three recordings before Labor Day and four after Labor Day, daytime weekends four recordings before Labor Day and two after Labor Day, and one nighttime recording during the period of very low human activity after Labor Day Table I. All recordings were conducted from a kayak, anchored in the same location in Southold Bay, 120 m from shore, in water with a mean depth of 5 m. From visual observations, watercrafts in the Bay never exceeded 9 m in length. During data collection, the wind force was mostly zero or one on the Beaufort scale, sometimes between one and three, and once approached five. No precipitation fell during any of the data collection periods. Using this sampling scheme, a total of 14 one-hour samples were collected, for all five different human activity levels, in the same location under similar environmental conditions that differed principally only in the amount of anthropogenic activities. C. Acoustic recording The underwater acoustic environment was monitored and measured using a calibrated Sippican Inc. hydrophone system with a sensitivity of 162 dbv at 1000 Hz re 1 Pa. Sounds were recorded with a Sony TCD-D8 Digital Audio Tape DAT Walkman, which sampled at a rate of 44.1 khz. The lowest significant bit of the DAT recorder s ADC was 70 db re 1 Pa. Acoustic data were recorded on BASF-DAT Master 124-min digital tapes. The recording system s frequency response was flat 2 db in the 100 14 000-Hz frequency range, with a 10 db per octave roll-off in the 15 100-Hz frequency band. During the monitoring sessions, the hydrophone was suspended in mid-water column from a kayak approximately 2.5 m below the surface of the water ; the cable was attached by surgical tubing to reduce flutter. D. Acoustic data processing Digital acoustic data were transferred from digital audio tapes DATs to a computer and low-pass filtered at 1200 Hz to include only sounds within the range of sea turtle hearing, using Cornell Raven software Charif et al., 2003. The filtered data were graphically and statistically analyzed, and spectrograms were created FFT size: 512 samples, overlap: 0%, Window type: Hann using Cornell Canary 1.2.4 software. Spectrograms of noises, typical for each of the five different activity levels, along with their power spectrum statistics of the 5th, 50th, and 95th percentiles are shown in Fig. 1. Graphical representation of spectra was performed using interactive data language IDL software. Band levels of received sound pressure in the band of peak hearing range of sea turtles 200 700 Hz, for each of the five different human activity levels, were calculated in a direct computation from the received sound pressure spectral density levels Table II. Statistical comparisons of noise spectra from the four different daytime levels of human activity were performed using the Mann-Whitney rank sum test in which noise levels in the 50 1000-Hz frequency band were considered to be significantly different at p 0.05) when they were separated by 5 db or more. III. RESULTS There was a distinct increase in noise pressure within the low-frequency range of sea turtle hearing that corresponded to an increase of human boating activity in the bay ecosystem Fig. 2. On weekends during the boating season, noise pressure levels were the highest, and, as boating activity decreased, noise levels systematically diminished. High human activity also was accompanied by an increase in the complexity of noises, as measured by the increase of signals across a broad range of frequencies Figs. 2 b e. Below 50 Hz, noise levels were dominated by cable fluttering for all five levels of human activity. No comparisons were made between the spectral patterns for frequencies be- J. Acoust. Soc. Am., Vol. 117, No. 3, Pt. 1, March 2005 Samuel et al.: Underwater noise in sea turtle habitats 1467

TABLE II. Band levels of received sound pressure within the 200 700-Hz band during each of the five different human activity levels in Southold Bay. The received sound pressure spectral density levels have been integrated over the peak hearing range of sea turtles 200 700 Hz in order to calculate the received sound pressure band levels. Human activity level Band levels of received sound pressure in the 200 700-Hz band db re 1 Pa Before Labor Day Weekend 113 Weekday 102 After Labor Day Weekend 91 Weekday 87 Night 83 FIG. 1. I One-hour noise spectrograms from recordings during each of the five different human activity levels in Southold Bay. These spectrograms are considered to be typical for each activity level. a Lowest human activity nighttime; b Low human activity weekday after Labor Day; c Intermediate human activity weekend after Labor Day; d High human activity weekday before Labor Day; and e Highest human activity weekend before Labor Day. Frequencies are between 0.1 and 1.2 khz. All spectrograms were set to the same brightness and contrast. The darkness in the graphs represents pressure of sound, and shows not only how much noise is in the environment, but also how loud it is. II Power spectrum statistics for each of the one-hour noise spectrograms from recordings during each of the five different human activity levels in Southold Bay. The statistics show the 5th solid light gray line bottom line, 50th thick black line middle line, and 95th solid black line top line percentiles of sound. Below each percentile line is where the corresponding percent of sound is found at those pressures and along the entire frequency range. a Lowest human activity nighttime; b Low human activity weekday after Labor Day; c Intermediate human activity weekend after Labor Day; d High human activity weekday before Labor Day; and e Highest human activity weekend before Labor Day. low 50 Hz. Above 50 Hz, the spectral patterns were dramatically different for each of the different levels of human activity Fig. 2. At the lowest level of human activity, which occurred at night after Labor Day, the majority of the ambient noise occurred in the 50 200-Hz frequency band, with pressure levels below 75 db re 1 Pa. At frequencies above 200 Hz, which include the peak hearing range for sea turtles, there was very little noise, with detectable signals above 70 db in these frequencies comprising less than 5% of the total noise present at night Fig. 2 a. Even the maximum noise pressures in these higher frequencies were moderate, rarely exceeding 90 db Fig. 2 a. At slightly higher levels of human activity, during both weekdays and weekends after Labor Day, there was an increase over nighttime levels in both the total amount of noise and the pressure levels at higher frequencies Figs. 2 b and c. During these daytime recordings, pressure levels for sounds at all frequencies were greater than nighttime levels, with levels rising to as high as 100 db. On weekdays after the Labor Day holiday, the noise was spread widely across the entire hearing range of sea turtles. The pressure levels were roughly equal at middle and higher frequencies, with an approximately uniform amount of noise with levels ranging from 70 to 80 db Fig. 2 b. On weekends after Labor Day, there was a further increase in the amount of noise present at pressures below 80 db across all frequencies, and a much greater amount of noise that was louder than 80 db Fig. 2 c. Overall, in the period after Labor Day, there was nearly a 3-dB increase in the amount of noise levels from the nighttime to the weekday samples, and a further 4-dB increase in noise levels from weekdays to the weekend samples. During the mid-summer levels of higher human activity, the bay ecosystem was flooded with noise across all frequencies. On weekdays, more than 50% of the sound at all frequencies was at pressures between 80 and 100 db, with noise levels peaking up to 110 db Fig. 2 d. On weekends, when boating activity was at its peak, the highest levels of noise also occurred. Along with a shroud of noise across the entire range of sea turtle hearing, 50% of the noise levels were at pressures between 90 and 110 db, with approximately 5% at levels greater than 120 db Fig. 2 e. Using the profile from the weekdays after Labor Day as a baseline reference for daytime noise levels, the mean difference in sound pressure spectral densities was calculated for the three other daytime activity levels Fig. 3. This was used as a general index of anthropogenic noise. When sound pressures varied by more than 5 db from the quiet baseline levels, the difference was considered to be significant. Using this measure, during two of the three daytime human activity levels, there was a significant difference (p 0.05) from the baseline Figs. 3 b e. At frequencies greater than 200 Hz, during levels of higher human activity, the mean sound pressures on weekdays were approximately 15 db greater than the baseline Fig. 3 d, and, on weekends, pressures were 26 db greater than the baseline Fig. 3 e. Even during weekends of the lower human activity levels Fig. 3 c, when the 1468 J. Acoust. Soc. Am., Vol. 117, No. 3, Pt. 1, March 2005 Samuel et al.: Underwater noise in sea turtle habitats

FIG. 3. Sound pressure spectral density level profiles in Southold Bay filtered for frequencies within the sea turtle hearing range below 1200 Hz during five periods of different human activity levels, compared to pressures of known levels of noise and to U.S. Navy LFA signals that elicit a response in sea turtles. a Lowest human activity nighttime number of spectral frames, n 77 504); b Low human activity weekday after Labor Day (n 300 988); c Intermediate human activity weekend after Labor Day (n 159 028); d High human activity weekday before Labor Day (n 232 512); e Highest human activity weekend before Labor Day (n 290 640); and f maximum sound pressure levels recorded within the sea turtle peak hearing range, during the highest human activity level. Vertical dashed lines bracket the peak hearing range of sea turtles 200 700 Hz. The maximum noises recorded in the bay were comparable in pressure to the documented levels, which evoked disturbance responses in sea turtles: g O Hara 1990 and h Samuel et al., in manuscript. Pressure levels measured in the bay ecosystem during high human activity levels were much higher than levels at which sea turtles showed a behavioral response: i Samuel et al. in manuscript. FIG. 2. A graphic depiction of both sound pressure and frequency, within the range of sea turtle hearing, over the entire range of human activity in Southold Bay. Distribution of sound is calculated by the percent of time a sound at a particular frequency throughout the sea turtle hearing range is found within a 2-dB range of received sound pressure at the hydrophone over the total hydrophone sampling period. a Lowest human activity nighttime number of spectral frames, n 77 504); b Low human activity weekday after Labor Day (n 300 988); c Intermediate human activity weekend after Labor Day (n 159 028); d High human activity weekday before Labor Day (n 232 512); and e Highest human activity weekend before Labor Day (n 290 640). difference was not significant, sound pressures were higher in many frequencies than baseline values by as much as 4 db. A direct measure of anthropogenic noise was seen best in a comparison of noise levels between weekdays and weekends within the same human activity season. Any differences observed in mean sound pressures between weekdays and weekends could not readily be attributed to biological or seasonal differences and, therefore, were ascribed to the higher boating activity on weekends. After the main boating season, even though sound levels were not significantly different between weekdays and weekends, there was still a 4-dB increase in noise levels on weekends. During the height of the boating season differences were even more obvious, with weekend mean sound pressure levels significantly higher (p 0.05) than weekdays, often by more than a factor of8db Fig. 3. This was particularly true at frequencies within the peak hearing range of sea turtles during this season. The difference in the band levels of sound pressure within the 200 700-Hz band between weekend and weekday was approximately 11 db Table II. During these extremely noisy conditions, when human activity and mean sound pressures were highest, maximum levels of noise observed Fig. 3 f rose far above the mean noise levels. For a sense of perspective, these observed means and maxima were compared to values from other studies in which sea turtles were subjected to a range of pressure levels to determine their behavioral response. Indeed, the maximum pressures recorded in the bay were comparable to sound pressures produced by high-pressure air gun pulses Fig. 3 g, which evoked changes in swimming patterns and orientation of sea turtles O Hara, 1990. In addition, the maximum levels in the bay also were within the range of pressures Fig. 3 h at which sea turtles in tanks exhibited agitated behavior when subjected to simulated boat signals and recordings of the U.S. Navy s Low Frequency Active sonar signals LFA. Details of these measurements are in a separate manuscript. Similar tank studies demonstrated that noise levels in the bay during the high human activity are at least 10 to 20 db higher than pressures at which sea turtles showed a behavioral response Fig. 3 i. J. Acoust. Soc. Am., Vol. 117, No. 3, Pt. 1, March 2005 Samuel et al.: Underwater noise in sea turtle habitats 1469

IV. DISCUSSION During much of the season when sea turtles are actively foraging in New York waters, their coastal habitats are flooded with underwater noise, a substantial component of which is anthropogenic in origin. Much of this noise occurs at frequencies below 1000 Hz, squarely within the range of highest sensitivity for sea turtle hearing. In mid-summer, during periods of high human activity, levels of noise were correspondingly high, and, as human activity decreased, both the pressure and complexity of underwater noise diminished. In general, daytime noise levels remained high for several weeks, until after the Labor Day holiday, when both recreational activity and noise pressure decreased abruptly. A more detailed analysis of the noise profiles within the seasons highlighted the extent of anthropogenic noise. In addition, the observed patterns indicated that much of the lowfrequency noise stemmed from recreational boating activity. On weekends, there were distinctly higher noise levels than on weekdays, and the same pattern was evident during both summer and fall sampling periods. Furthermore, the measured increases in sound pressures were substantial. At the peak of the recreational boating season, average noise pressures at many frequencies within the range of sea turtle hearing were more than 8 db greater than those recorded during the corresponding weekdays. These levels were more than 26 db greater than noise levels recorded at night after Labor Day Fig. 3. Notwithstanding seasonal and diurnal differences, the noise signature from boats was strikingly evident in the sea turtle foraging habitat. There was a substantial change also in the character of the noise at times of increased boating activity. At the lowest level of boating activity, namely at nighttime after Labor Day, most of the detectable sounds were at frequencies below 200 Hz. At frequencies in the range of peak sea turtle sensitivity 200 700 Hz and higher, there were very few detectable signals throughout the entire monitoring session Fig. 2 a. On weekdays after Labor Day, although there was an increase in signals at higher frequency ranges, most of the underwater sounds similarly were in the lower 50 200 Hz frequencies Fig. 2 b. In contrast, on weekends and during the height of the recreational season, underwater noise was louder and much more complex, extending relatively evenly across the entire measured frequency spectrum ranging up to 1200 Hz Figs. 2 c e. A comparison of the sound pressure levels in the 200 700-Hz band measured under the five different human activity levels Table II, with the corresponding sound pressure band levels of ambient sounds from environmental sources as reported by Wenz 1962, reveals that the noises recorded in Southold Bay, even during the daytime period of lowest human activity, are much higher than the reported levels. In shallow waters, such as the Southold Bay area, the primary source of noise from environmental sources is wind and waves Richardson et al., 1995. According to Wenz 1962, at a wind force of 1, on the Beaufort scale, sound pressure in the 200 700-Hz band is approximately 46 db re 1 Pa. As the wind force and consequently wave height increase, so does the sound pressure. At 2 Beaufort, noise levels rise to 55 db, and at 3 Beaufort reach 59 db. Even during the highest wind force we collected data, at 5 Beaufort, noise levels in the 200 700-Hz band attributed to wind by Wenz 1962 are 66 db. Although wind contributes to underwater noise levels, our high band levels, even during the period with the lowest daytime recreational human activity, indicate that biological noise, distant shipping, or even noise from land are increasing ambient noise levels in Southold Bay throughout our recording period. The contribution of these sources to the noise levels in the bay can be seen in the sound pressure band levels recorded during weekdays after Labor Day, a period which we consider to be the baseline of daytime noise levels Table II. Juvenile sea turtles are found in New York coastal waters and estuarine ecosystems for a period of about 21 weeks each year, starting at the beginning of July Morreale and Standora, 1998. The data from this study indicate that these foraging sea turtles spend nearly half of their activity season surrounded by higher than the baseline levels of underwater noise. Rather than being a unique situation, the Long Island study site is likely to be very representative of numerous other coastal habitats where foraging, mating, and nesting sea turtles congregate during key stages in their lives. Similarly, many of these crucial habitats exist in very populous areas where sea turtle activity coincides with exposure to high human activity and underwater noise. From previous research it is evident that sea turtles can detect sound, and that their hearing is confined to lower frequencies, mainly below 1000 Hz Ridgway et al., 1969; Bartol et al., 1999. Therefore, sea turtles are able to hear the low-frequency sound emitted underwater by anthropogenic sources, and such exposure could be directly influencing their health and ecology. However, little is known about the potential effects of noise exposure on the short-term or longer-term behavior and health of sea turtles. Nevertheless, several studies have shown that sea turtles in experimental tanks respond to signals at pressures well below those recorded in the Southold Bay foraging habitat, and some of the responses indicate at least short-term disturbance. Sea turtles displayed agitated behavior, abrupt body movements, startle responses, and even prolonged inactivity at the bottom of the tank in response to low-frequency signals Lenhardt et al., 1983, 1996. Such responses are similar to the ones we observed during exposure of sea turtles to LFA signals and simulated boat sounds. In a similar study in outdoor enclosures, sea turtles exposed to louder pulses from high-pressure air guns exhibited changes in swimming patterns and orientation O Hara, 1990. Long-term effects on sea turtles after exposure to noise, although more insidious, would be much harder to detect. However, it is possible that prolonged exposure could be highly disruptive to the health and ecology of the animals as it has been recorded to affect other marine animals, by encouraging avoidance behavior Malme et al., 1983; Richardson et al., 1986; Richardson, 1997, increasing stress and aggression levels Lagardère, 1982, causing physiological damage to the ears through either temporary or even permanent threshold shifts Hastings et al., 1996; Scholik and Yan, 2001; Erbe, 2002; McCauley et al., 2003, altering surfacing and diving rates Suzuki et al., 1980; Blaxter et al., 1981; 1470 J. Acoust. Soc. Am., Vol. 117, No. 3, Pt. 1, March 2005 Samuel et al.: Underwater noise in sea turtle habitats

Olsen et al., 1983; Schwarz and Greer, 1984; Engås et al., 1998; Williams et al., 2002, or even confounding orientation cues. Sea turtles have been shown to exhibit strong fidelity to fixed migratory corridors, habitual foraging grounds, and nesting areas Morreale et al., 1996; Morreale and Standora, 1998; Avens et al., 2003, and such apparent inflexibility could prevent sea turtles from selecting alternate, quieter habitats. It is likely that sea turtles and other marine animals are not adapted to contemporary noise levels. Over the past 33 years there has been a documented 9- to 10-dB increase in ambient sound levels in some coastal marine environments Andrew et al., 2002. Furthermore, there has been a 6.3% increase in the usage of recreational boats in the last decade NMMA, 2003. Thus, even with the current noise abatement regulations, which prohibit vessel operation in inland waters that exceed noise levels above 82 db re 1 Pa, measured at a distance of 25 m from the vessel 36CFR3 Title 36, Chapter I, Part 3, Sec. 3.7, cumulative noise levels are likely to keep increasing as recreational motor boat use rises. Additional new sources, such as seismic exploration activities in shallow water, which are increasing worldwide Frisk et al., 1998, also will undoubtedly contribute to the pervasive noise levels. Anthropogenic noise in the underwater environment appears to be more than an abstract issue inevitably related to increasing human populations. Our study in a representative coastal habitat highlights the extent of the problem. Sea turtles tend to cluster, as is the habit of many other marine animals, in continental shelf and pelagic waters of the western North Atlantic where there is extensive human activity. This spatial overlap undoubtedly exposes many of them to high levels of ambient noise, especially during warmer months. Currently, the effects of such exposure are largely unknown. At the least, underwater noise must be accounted for when devising appropriate management strategies for the protection and recovery of rare and endangered organisms such as sea turtles. At best, future marine conservation plans should include careful acoustic monitoring to prevent future unabated noise pollution. ACKNOWLEDGMENTS We would like to thank a number of people for their contribution. From the Cornell Laboratory of Ornithology, we thank Melissa Craven Fowler, Christopher Tessaglia- Hymes, Connie Gordon, Kurt Fristrup, Bob Grotke, and Marguerite McCartney for providing all the equipment and helping with the data retrieval. From the Cornell Ocean Resources and Ecosystems Program, we thank all members of the laboratory group, and especially Bruce Monger, for his invaluable help and contributions to the data analyses. From Suffolk County Cornell Cooperative Extension Marine Program, we thank Chris Smith, Rory MacNish, Stacy Myers, Emerson Hasbrouck, Eileen Brennan, Sonia Tulipano, and Ronnie Matovcik for providing funding, support, and access to the study site. Special thanks to Kim Durham, Robert DiGiovanni, Chris Buckman, Kelly Cantara from New York State Marine Mammal and Sea Turtle Stranding Program for support and access to animals. Bennett S. Orlowski from the Long Island Horticultural Research and Extension Center kindly provided living accommodations. We further thank Lt. Commander USCG, Jeannot Smith, for field and technical support. Our thanks also to Riverhead Foundation for Marine Research and Preservation, and National Marine Fisheries Service for support and funding. 36CFR3 2001. Code of Federal Regulations. Title 36, Volume 1, Chapter I, Part 3, Section 3.7. Revised July 2001. Andrew, R. K., Howe, B. M., Mercer, J. A., and Dzieciuch, M. A. 2002. Ocean ambient sound: Comparing the 1960s with the 1990s for a receiver off the California coast, ARLO 3 2, 65 70 2002. Avens, L., Braun-McNeill, J., Epperly, S., and Lohmann, K. J. 2003. Site fidelity and homing behavior in juvenile sea turtles Caretta caretta, Mar. Biol. Berlin 143 2, 211 220. Bartol, S. M., Musick, J. A., and Lenhardt, M. L. 1999. Auditory evoked potentials of the loggerhead sea turtle, Copeia 3, 836 840. Blaxter, J. H. S., Gray, J. A. B., and Denton, E. J. 1981. Sound and startle responses in herring shoals, J. Mar. Biol. Assoc. U.K. 61, 851 869. Charif, R. A., Clark, C. W., and Fristrup, K. M. 2003. Raven 1.1 User s Manual, Cornell Laboratory of Ornithology, Bioacoustics Research Program, Ithaca, NY. Curtis, K. R., Howe, B. M., and Mercer, J. A. 1999. Low-frequency ambient sound in the North Pacific: Long time series observations, J. Acoust. Soc. Am. 106, 3189 3200. Engås, A., Haugland, E. K., and Øvredal, J. T. 1998. Reactions of cod Gadus morhua L. in the pre-vessel zone to an approaching trawler under different light conditions. Preliminary Results, Hydrobiologia 371Õ372, 199 206. Erbe, C. 2002. Underwater noise of whale-watching boats and potential effects on killer whales Orcinus orca, based on an acoustic impact model, Marine Mammal Sci. 18 2, 394 418. Frisk, G. V., Jackson, D., DeFerrari, H., Simmen, J., Spindel, R. C., Buck, J. R., D Spain, G., Hilderbrand, J. A., and Greene, C. R. 1998. Ocean Acoustics, in Proceedings of the workshop on the effects of anthropogenic noise in the marine environment, edited by R. C. Gisiner, Marine Mammal Science Program, Office of Naval Research, pp. 10 18. Fristrup, K. M., Hatch, L. T., and Clark, C. W. 2003. Variation in humpback whale Megaptera novaeangliae song length in relation to lowfrequency sound broadcasts, J. Acoust. Soc. Am. 113, 3411 3424. Hastings, M. C., Popper, A. N., Finneran, J. J., and Lanford, P. J. 1996. Effects of low-frequency underwater sound on hair cells of the inner ear and lateral line of the teleost fish Astronotus ocellatus, J. Acoust. Soc. Am. 99, 1759 1766. Lagardère, J. P. 1982. Effects of noise on growth and reproduction of Crangon crangon in rearing tanks, Mar. Biol. Berlin 71, 177 185. Lenhardt M., Moein, S., and Musick, J. 1996. A method for determining hearing thresholds in marine turtles, in Proceedings of the fifteenth annual workshop on sea turtle biology and conservation, NOAA technical Memorandum NMFS-SEFSC-387. Lenhardt, M. L., Bellmund, S., Byles, R. A., Harkins, S. W., and Musick, J. A. 1983. Marine turtle reception of bone-conducted sound, J. Aud Res. 23, 119 125. Lesage, V., Barrette, C., Kingsley, M. C. S., and Sjare, B. 1999. The effect of vessel noise on the vocal behavior of belugas in the St. Lawrence river estuary, Canada, Marine Mammal Sci. 15 1, 65 84. Malme, C. I., Miles, P. R., Clark, C. W., Tyack, P., and Bird, J. E. 1983. Investigations of the potentials effects of underwater noise from petroleum industry activities on migrating gray whale behaviour: final report for the period of 7 June 1982 31 July 1983, Report No. 5366, prepared by Bolt, Beranek and Newman Inc., Cambridge, MA, USA, for U.S. Minerals Management Service, Alaska OCS Office, Anchorage, USA. McCauley, R. D., Fewtrell, J., and Popper, A. N. 2003. High pressure anthropogenic sound damages fish ears, J. Acoust. Soc. Am. 113, 638 642. Miller, P. J. O., Biassoni, N., Samuels, A., and Tyack, P. L. 2000. Whale songs lengthen in response to sonar, Nature London 405 6789, 903 903. Morreale, S. J., and Standora, E. A. 1998. Early life stage ecology of sea turtles in Northeastern U.S. waters, NOAA Tech. Mem. NMFS-SEFSC- 413. Morreale, S. J., Standora, E. A., Spotila, J. R., and Paladino, F. V. 1996. J. Acoust. Soc. Am., Vol. 117, No. 3, Pt. 1, March 2005 Samuel et al.: Underwater noise in sea turtle habitats 1471

Migration corridor for sea turtles, Nature London 384, 319 320. National Marine Manufacturers Association 2003. Boating 2002: At a glance Facts & Figures, Marketing Statistics Department, Chicago, IL. O Hara, J. 1990. Avoidance responses of loggerhead turtles, Caretta caretta, to low frequency sounds, Copeia 2, 546 567. Olsen, K., Angell, J., Pettersen, F., and Løvik, A. 1983. Observed fish reactions to a surveying vessel with special reference to herring, cod, capelin and polar cod, FAO Fish. Rep. 300, 131 138. Plotkin, P. T. ed.. 1995. National Marine Fisheries Service and U.S. Fish and Wildlife Service Status Reviews for Sea Turtles Listed under the Endangered Species Act of 1973, National Marine Fisheries Service, Silver Spring, MD. Richardson, W. J. 1997. Marine mammal and acoustical monitoring of BPXA s seismic program in the Alaskan Beaufort Sea, LGL Report TA2150-3, prepared by LGL Ltd. and Greeneridge Sciences, Inc. for BP Explorations Alaska, Inc. and National Marine Fisheries Service. King City, Can., LGL Ltd. Richardson, W. J., Würsig, B., and Greene, Jr., C. R. 1986. Reactions of bowhead whales, Balaena mysticetus, to seismic exploration in the Canadian Beaufort Sea, J. Acoust. Soc. Am. 79, 1117 1128. Richardson, W. J., Greene, Jr., C. R., Malme, C. I., and Thomson, D. H. 1995. Marine Mammals and Noise Academic, San Diego, CA. Ridgway, S. M., Wever, E. G., McCormick, J. G., Palin, J., and Anderson, J. H. 1969. Hearing in the giant sea turtle, Chelonia mydas, Psychology 64, 884 890. Samuel, Y. 2004. Underwater low-frequency noise and anthropogenic disturbance in a critical sea turtle habitat, M.S. thesis, Cornell University. Scholik, A. R., and Yan, H. Y. 2001. Effects of underwater noise on auditory sensitivity of a cyprinid fish, Hear. Res. 152, 17 24. Schwarz, A. L., and Greer, G. L. 1984. Responses of pacific herring, Clupea harengus pallasi, to some underwater sounds, Can. J. Fish. Aquat. Sci. 41, 1183 1192. Suzuki, H., Hamada, E., Saito, K., Maniwa, Y., and Shirai, Y. 1980. The influence of underwater sound on marine organisms, J. Nav. 33, 291 295. Wenz, G. M. 1962. Acoustic ambient noise in the ocean: Spectra and sources, J. Acoust. Soc. Am. 34, 1936 1956. Williams, R., Trites, A. W., and Bain, D. E. 2002. Behavioural responses of killer whales Orcinus orca to whale-watching boats: opportunistic observations and experimental approaches, J. Zool. 256, 255 270. 1472 J. Acoust. Soc. Am., Vol. 117, No. 3, Pt. 1, March 2005 Samuel et al.: Underwater noise in sea turtle habitats