Minke Whale (Balaenoptera Acutorostrata) Boings Detected at the Station ALOHA Cabled Observatory

Minke whale (Balaenoptera acutorostrata) boings detected at the Station ALOHA Cabled Observatory

Julie N. Oswalda,b) and Whitlow W. L. Au Hawaii Institute of Marine Biology, University of Hawaii, P.O. Box 1106, Kailua 96374, Hawaii

Fred Duennebier Department of Geology and Geophysics, School of Ocean and Earth Science and Technology, University of Hawaii, Room 701, 1680 East-West Road, Honolulu 96822, Hawaii (Received 23 July 2010; revised 14 March 2011; accepted 15 March 2011) Minke whales (Balaenoptera acutorostrata) in the tropical North Pacific are elusive and difficult to detect visually. The recent association of a unique sound called the “boing” to North Pacific minke whales has made it possible to use passive acoustics to investigate the occurrence of this species in Hawaiian waters. One year of recordings (17 February 2007–18 February 2008) made at the Station ALOHA Cabled Observatory were examined to investigate the characteristics of boings and temporal patterns in their occurrence at this site, located 100 km north of Oahu. Characteristics of boings exhibited low variability. Pulse repetition rate and duration measurements matched those for “central” or “Hawaii” boing types. Boings were detected from October until May, with a peak in March. Although no boings were detected from June to September, the absence of boings does not necessarily indicate the absence of minke whales. Significant diel variation in boing rate was not observed. The absence of a diel pattern in boing production suggests that day- or night-time acoustic surveys are equally acceptable methods for studying minke whale occurrence. Future research should include efforts to determine what other sounds are produced by minke whales in this area, and which age/sex classes produce boings. VC 2011 Acoustical Society of America. [DOI: 10.1121/1.3575555] PACS number(s): 43.80.Ka, 43.30.Sf [JFF] Pages: 3353–3360

I. INTRODUCTION revealed much about the biology of baleen whales, including distribution patterns, migration routes, population structure, North Pacific minke whales (Balaenoptera acutoros- and seasonal and diel behavioral patterns (e.g., Clapham and trata) are among the smallest of the baleen whales. They are Matilla, 1990; Stafford et al., 1999; Burtenshaw et al., 2004; typically encountered individually or in small groups of two Stafford et al., 2005; Oleson et al., 2007a). or three (although large aggregations occasionally form in Minke whales have been reported to produce a variety high latitudes). Their blows are inconspicuous and they sur- of sounds throughout their range. Low frequency down- face for only short periods of time (Perrin and Brownell, sweeps, higher frequency clicks, whistles, grunts, and pulse 2002). Because of these characteristics, minke whales are trains have been recorded in the presence of minke whales in notoriously difficult to detect using visual methods and the the St. Lawrence Estuary (Edds-Walton, 2000) and the Ca- species has been considered rare in Hawaiian waters (Hor- ribbean (Winn and Perkins, 1976; Mellinger et al., 2000)as wood, 1990). A cabled, ocean bottom observatory located at well as in the presence of Antarctic minke whales (Balae- Station ALOHA (A Long-term Oligotrophic Habitat Assess- noptera bonaerensis) in the Ross Sea (Schevill and Watkins, ment) provided an opportunity to study the occurrence of 1972; Leatherwood et al., 1981). Dwarf minke whales on the minke whales at a deep ocean research site located 100 km Great Barrier Reef, Australia produce a distinctive sound north of Oahu, Hawaii. that has been dubbed the “star wars” vocalization due to its Because minke whales are so difficult to observe visu- unusual, synthetic-sounding characteristics (Gedamke et al., ally, other methods must be used to study their distribution, 2001). behavior, and ecology. Most species of baleen whales pro- In contrast to Atlantic and Antarctic minke whales, duce relatively low frequency sounds that propagate well almost nothing was known about the sounds produced by underwater. As a result, passive acoustic methods have been North Pacific minke whales until quite recently. Based on si- used extensively to study many baleen whales, such as blue, multaneous visual observations and acoustic localization, humpback, fin, right, and bowhead whales (e.g., Clark, 1982; Rankin and Barlow (2005) were able to attribute the mysteri- McDonald et al., 1995; Norris et al., 1999; Stafford et al., ous “boing” sound to this species. Boings are relatively ster- 2005; Delarue et al., 2009). Passive acoustic studies have eotyped calls that consist of a brief pulse followed by a long frequency and amplitude modulated call with a pulsed structure and a peak energy of approximately 1.4 kHz (Thompson a) Current address: Oceanwide Science Institute, P.O. Box 61692, Honolulu and Friedl, 1982; Rankin and Barlow, 2005; Figs. 1 and 2). 96839, Hawaii. b)Author to whom correspondence should be addressed. Electronic mail: The structure of boings exhibits geographic variation, and [email protected] two different boing types have been reported in the

J. Acoust. Soc. Am. 129 (5), May 2011 0001-4966/2011/129(5)/3353/8/$30.00 VC 2011 Acoustical Society of America 3353 Author's complimentary copy FIG. 1. Spectrogram of a minke whale boing (1024 point FFT, Hann window). “Precursor” component is indicated by an arrow.

literature. The “eastern” (or “San Diego”) boing has a pulse Cabled Observatory (ACO), we examined the characteristics repetition rate of 91–93 pulses/s, a mean duration of 3.6 s, of minke whale boings in the waters north of Oahu, as well and has been detected east of 138W. The “central” (or as seasonal and diel patterns in their occurrence. “Hawaii”) boing has a pulse repetition rate of 114–118 pulses/s, a mean duration of 2.6 s, and has been detected west of 135W(Wentz, 1964; Rankin and Barlow, 2005). II. MATERIALS AND METHODS The boing was first described from recordings made by U.S. Recordings were made using a seafloor-mounted hydro- navy submarines off San Diego, California and Kaneohe, phone located at the ACO, 100 km north of Oahu, Hawaii Hawaii (Wenz, 1964) and is a common sound in the North (22450N, 158000W, Fig. 3). A retired electro-optical Pacific Ocean. They have been reported to occur seasonally telecommunications cable (HAW-4 SL-280 electro-optical in Hawaiian waters (Thompson and Friedl, 1982). cable) provided power and broadband Ethernet communica- Attributing the source of the boing to North Pacific tions capability to the ACO, allowing real-time, continuous minke whales has finally given researchers a way to study acoustic monitoring. The factory calibrated hydrophone the occurrence, behavior, and ecology of this elusive species. [OAS (Optimum Applied System, Poughkeepsie, NY) Using acoustic recordings made at the Station ALOHA Model E-2PD] used at the ACO was originally installed for

FIG. 2. (Color online) Waveform of a boing. The pulsed structure of the amplitude modulated component can be seen in the expanded waveform (inset).

3354 J. Acoust. Soc. Am., Vol. 129, No. 5, May 2011 Oswald et al.: Boing detections at Station ALOHA Author's complimentary copy gram, and ranked each boing as one of the five quality categories. The five quality categories ranged from 1 (audible, but barely recognizable as a boing on the spectrogram) to 5 (very loud and clear boing). Low quality category one and two boings had SNRs of approximately 6–12 dB above ambient noise in the 1–5 kHz frequency range. The results of the manual detections were then compared to results of detections made using XBAT on the same section of data. A total of 783 boings were manually identified in the 8- h recording that was used for ground-truthing the detector. The automated detector identified 100% of category 5 boings (n ¼ 49), 99% of category 4 boings (n ¼ 78), 91% of category 3 boings (n ¼ 150), 59% of category 2 boings (n ¼ 259), and 22% of category 1 boings (n ¼ 247). Only 5% of detections made by the XBAT detector were false detections. Measurements were taken from a subset of boings randomly selected from the year of data. In order to avoid sampling multiple boings from a single individual and therefore maintain independence of data, one boing was randomly selected from each 24 h period during which boings were detected and only boings that were at least 6 h apart were measured. Measurements were made using OSPREY, an automated measurement software package (Mellinger and Brad- bury, 2007). Boings were divided into two components. The “precursor” component is an initial pulse that precedes the FIG. 3. Map showing the island of Oahu and the location of the Station ALOHA Cabled Observatory (represented by the white star). longer, “amplitude modulated” (AM) component (Fig. 1). The AM component has a pulsed structure, as seen in the seismic work and had a flat frequency response from 10 Hz waveform in Fig. 2. Several measurements were taken from to 1.3 kHz. Oscillations in the frequency response above each component, including minimum and maximum fre- 1.5 kHz were caused by the sound wavelength approaching quency, peak frequency (the frequency at which peak inten- the dimensions of the sensor. These oscillations had a maxi- sity occurs), and duration. In addition, several measurements mum amplitude of 65 dB. The hydrophone sampled up to were taken only from the AM component: SNR of the first 96 kHz with 24-bit resolution, resolving signals ranging 0.5 s, SNR of the last 0.5 s, and pulse repetition rate. Pulse from 0.01 Hz to 40 kHz. It was located at a depth of 4.7 km repetition rate was measured using a custom written add-on and floated approximately 10 m off the seafloor (Duennebier to OSPREY (S. Martin, SPAWAR). et al., 2008). Data from the ACO were transmitted to the To examine seasonal patterns in boing detection, the AT&T Makaha Cable Station on Oahu and then over a T-1 mean number of boings detected per hour was plotted for datalink to the University of Hawaii, where they were each month. Mean number of boings per hour (hourly boing archived and analyzed. The original data were downsampled rate) was examined instead of absolute number of boings to 24 kHz for storage as archived data. Archived data were detected in a day because, due to technical issues, some days stored at a lower sampling rate in order to conserve storage contained fewer than 24 h of recordings. disk space at the University. For analyses of diel patterns in boing occurrence, only The ACO hydrophone was operational from 17 February days during which calls were detected and days with less than 2007 until 22 October 2008. Recordings were continuously half an hour missing from the recordings were included. Each made during this 20 month period, with small breaks due to day was divided into three light regimes defined by the alti- equipment problems. We examined 1 yr of these recordings tude of the sun (obtained from the United States Naval Ob- (17 February 2007–18 February 2008) using a data template servatory Astronomical Applications Department web site): detector created with XBAT (Extensible Bioacoustic Tool) (1) Light: The hours when the altitude of the sun was greater software. XBAT’s data template detector is a spectrogram cor- than 0 above the horizon, from approximately 06:55 to relation detector. It looks at the time cross-correlation 18:15. sequence between an example sound [in this case, both a (2) Dusk: The hours when the altitude of the sun was high signal-to-noise ratio (SNR) boing and a medium SNR between 0 and –12 below the horizon, from approxi- boing were used as example sounds] and the sound file being mately 06:00 to 06:55 and 18:15 to 19:09. analyzed. Events are detected when the correlation exceeds a (3) Dark: The hours when the altitude of the sun was less user-defined threshold. This detector was used to automati- than –12 below the horizon, from approximately 19:09 cally detect and log minke whale boings. The detector was to 06:00. ground-truthed using 8 h of data recorded on 5th March 2007. An experienced acoustician manually identified boings Mean adjusted boing rate (MABR) was then calculated based on listening to the recording and examining a spectro- for each light regime in a given day. Mean adjusted boing

J. Acoust. Soc. Am., Vol. 129, No. 5, May 2011 Oswald et al.: Boing detections at Station ALOHA 3355 Author's complimentary copy rate was based on the mean hourly boing rate for each 24 h day (BR24) and the mean hourly boing rate during each light regime (BRlight,BRdusk, and BRdark) for that day. For example:

MABRlight¼ BRlight BR24:

A negative value for MABR means that the mean hourly boing rate during that part of the day was less than the mean hourly boing rate for that 24 h period and a positive value means that the hourly boing rate during that part of the day was greater than the mean hourly boing rate for that 24 h period. This approach was taken to remove bias caused by the large variation in the total number of calls detected each day throughout the year (Stafford et al., 2005; Wiggins et al., 2005; Oleson et al., 2007a). Because MABR data were not normally distributed, Kruskal–Wallis tests were used to test FIG. 4. Mean and standard deviation of hourly boing rate by month from 17 the null hypothesis that the difference in boing rates among February 2007 to 18 February 2008. Number of days during which recordings were made for each month shown as solid line and triangles. Number light regimes was equal to zero. of days during which boings were detected for each month shown as dashed line and squares. III. RESULTS Boings were commonly heard during the winter and Measurements were taken from a total of 111 boings and spring, with a total of 15 552 boings detected between are presented in Table I. Frequency measurements for both the February 2007 and February 2008. Figure 4 shows the mean precursor and the AM components of boings exhibited a very number of boings detected per hour for each month. Boings low variability. The fundamental frequencies of boings ranged were detected from 22 October to 21 May and not at all dur- from 1 to 1.8 kHz, but harmonics extending to approximately ing the months of June to September. The occurrence of 9 kHz were often observed. Duration exhibited the highest var- boings increased quite suddenly in November, peaked in iability, especially for the AM component, ranging from 1.4 to March, and dropped off quickly after that. High variability 4.2 s. Overall duration (precursor duration plus AM duration) in hourly boing rate was observed from day to day, as shown had a mean value of 2.5 s (SD ¼ 0.4 s) and ranged from 1.6 to by the example in Fig. 5. Mean boing rate ranged from 0 to 4.5 s. SNR decreased an average of 5.3 6 2.2 dB from the 62 boings/h during the month of January 2008 and large dif- beginning to the end of the AM component. ferences were seen from 1 day to the next. Figure 6 shows mean hourly boing rate for each hour of TABLE I. Mean (with standard deviation in parentheses), minimum and the day, averaged over the year (n 154 days). Hourly boing maximum values for measurements taken from the precursor and amplitude ¼ modulated (AM) components of boings. rate was highest in the early morning and the middle of the day, with a dip in the mid-morning and late evening. Mean Variable Precursor AM adjusted boing rates reflected this pattern, with higher values component component during light hours (mean ¼ 0.10, SD ¼ 6.9, n ¼ 154) than Peak frequency (kHz) Mean 1.4 (0.09) 1.4 (0.02) during dark (mean ¼ 0.02, SD ¼ 6.4, n ¼ 154) or dusk Minimum 1.2 1.4 (mean ¼ –0.19, SD ¼ 5.2, n ¼ 154) hours. However, these Maximum 1.6 1.5 differences in MABR were not significant (Kruskal–Wallis Minimum frequency (kHz) Mean 1.2 (0.07) 1.3 (0.04) test, n ¼ 154, df ¼ 2, H ¼ 20.8, p < 0.0001). Minimum 1.0 1.2 Maximum 1.3 1.4 Maximum frequency (kHz) Mean 1.6 (0.08) 1.6 (0.07) Minimum 1.4 1.4 IV. DISCUSSION Maximum 1.8 1.9 A. Ground-truthing Duration (s) Mean 0.28 (0.05) 2.2 (0.4) Minimum 0.15 1.4 Boings are relatively stereotyped calls that are well Maximum 1.6 4.2 suited for detection using techniques such as spectrogram Pulse repetition rate (pulses/s) Mean n/a 116.3 (0.8) correlation. XBAT’s spectrogram correlation detector identi- Minimum n/a 111.7 fied a very high percentage of high quality boings in the 8 h Maximum n/a 117.9 of recordings used for ground-truthing. Most of the boings SNR of first 0.5 s (dB) Mean n/a 16.7 (2.1) that were missed by the detector were low quality, category Minimum n/a 11.6 one and two boings. These boings had poor SNR and only a Maximum n/a 22.6 portion of the boing was visible on the spectrogram. SNR of last 0.5 s (dB) Mean n/a 11.4 (2.3) Minimum n/a 7.8 Thompson and Friedl (1982) recorded boings using two Maximum n/a 18.2 bottom-mounted hydrophones located north of Oahu and used time-of-arrival differences to estimate that boings could

3356 J. Acoust. Soc. Am., Vol. 129, No. 5, May 2011 Oswald et al.: Boing detections at Station ALOHA Author's complimentary copy FIG. 5. Mean and standard deviation of hourly boing rate by day for the month of January 2008.

be detected at a distance of at least 10.5 km. Based on this it would be more important to detect every boing. To work and the fact that there was little ambient noise masking accomplish this, a lower threshold could be set for boing boings in the ground-truth data, it is likely that our category detection, but this would lead to a greater number of false one and two boings were produced by animals that were a positive detections. signiﬁcant distance from the hydrophone. As the goal of this The false positive (false alarm) rate for this study was study was to examine the occurrence patterns of boings, we low. Most of the false positives were triggered by “moans” felt that it was more important to consistently detect high-to- produced by humpback whales, which are most likely part of medium quality boings produced relatively close to the their songs. These sounds are low frequency, broadband, and hydrophone with relatively few false detections, rather than have relatively long durations, similar to boings. The to increase the incidence of false detections in order to detect ground-truth data were taken at the height of the humpback every boing within a larger area. This approach provides a whale wintering season in Hawaii when noise from hump- good basis for relative comparisons among months, days, back whales was loud and relatively constant throughout the and time periods within days. For other questions such 8 h that were analyzed. It is expected that the false positive as estimating absolute numbers of animals or calling rates, rate would be even lower than 5% during days when there

FIG. 6. Mean and standard deviation of number of boings per hour of the day, averaged from 17 February 2007 to 18 February 2008 and including only days during which boings were detected and with less than half an hour missing from the recordings (n ¼ 154 days). The x- axis starts at midnight and ends at 23:00. Bar above histogram indi- cates light regime: black ¼ night; gray ¼ dusk; and white ¼ day.

J. Acoust. Soc. Am., Vol. 129, No. 5, May 2011 Oswald et al.: Boing detections at Station ALOHA 3357 Author's complimentary copy was less humpback whale song, especially near the begin- ACO recordings. The fact that boings were heard for an ning and end of the humpback whale season when there are extended period of time with no secondary peak suggests that fewer whales in the area. Hawaiian waters are an end-point destination for minke whales, rather than a transitional location. If minke whales B. Characteristics of boings were simply passing through Hawaiian waters on their way to some other destination, one would expect to see two peaks in Measurements taken from the precursor and AM com- their occurrence—one as minkes passed through the area on ponents of boings illustrate the stereotyped nature of these their way to their destination, and another as they returned. sounds. Variability of the frequency measurements was on It is unlikely that minke whales are traveling to Hawai- the order of tens of hertz. It is not likely that this low vari- ian waters primarily to feed, as the North Pacific subtropical ability was a result of measuring multiple boings from a sin- gyre is generally considered to be an oligotrophic environ- gle individual, as only boings that were at least 6 h apart ment (Karl and Lukas, 1996; Sakamoto et al., 2004). Hump- were measured. While it is possible that one whale stayed in back whales migrate to Hawaiian waters to breed during the the area for 6 or more hours, measured boings were an aver- winter and spring (Baker and Herman, 1981) and it is possi- age of 72 h apart, increasing the likelihood that boings were ble that minke whales are behaving similarly, as the season- measured from different individuals. ality in boing detections closely matches that of humpback Pulse repetition rate (PRR) and duration measurements whale presence in Hawaiian waters. In addition, it is thought of the AM component matched those reported for “Central” that most baleen whales feed in productive, high-latitude or “Hawaii” boings. Boings recorded at the ACO had a PRR waters during the summer months and spend winters at low- ranging from 112 to 118 pulses/s and a mean overall dura- latitude breeding grounds (Lockyer and Brown, 1981; tion of 2.5 s. Rankin and Barlow (2005) reported a PRR of Bowen and Siniff, 1999). Although no boings were heard 114–118 pulses/s and mean duration of 2.6 s for their central from June to September, it is important to note that the ab- boings, and Thompson and Friedl (1982) reported a mean sence of boings does not necessarily indicate the absence of pulse repetition rate of 115 pulses/s for boings recorded at a minke whales. Our recordings were made using a single, bot- bottom-mounted hydrophone located north of Oahu. These tom-mounted hydrophone and it is possible that minke similar measurements also illustrate the stereotyped and sta- whales were out of range of our hydrophone, but still in Ha- ble nature of boings. waiian waters, during the summer months. In addition, Duration of the AM component was the characteristic of minke whales may be present in Hawaiian waters year-round boings that exhibited the highest variability, ranging from a and only produce boings at certain times of the year, similar minimum of 1.4 s to a maximum of 4.2 s (Table I). Some of to blue whales, which exhibit seasonal separation in the pro- this variability may have been due to the fact that the SNR duction of their B calls and D calls in the Southern Califor- of boings generally decreases from the start of the AM com- nia Bight (Oleson et al., 2007a). A variety of sounds have ponent to the end of the AM component (see Table I). been attributed to minke whales in other locations, including Because boings appear to “fade out,” those produced at low frequency downsweeps, higher frequency clicks, whis- greater distances may appear to be shorter than those pro- tles, grunts, and others (Winn and Perkins, 1976; Edds- duced at closer range to the hydrophone. Another source of Walton, 2000; Mellinger et al., 2000). Based on this knowl- variability among boings is the presence of the precursor edge, it is plausible that minke whales in the tropical north section. As the aim of this study was to examine characteris- Pacific may be producing other sounds in addition to boings. tics of both the precursor and the AM components of boings, Currently, nothing is known about other call types that may only boings that contained both components were measured. be produced by north Pacific minke whales. Efforts should However, not every boing contains a precursor component. be made to make recordings in the presence of minke whales The short duration of this component relative to the AM for extended periods and to attach acoustic or other remote- component as well as the fact that is it not always present locating tags to minke whales in order to determine what suggests that it is less important to the function of boings other sounds they may be producing. When the vocal behav- than the AM component. ior of north Pacific minke whales is more completely known, it will be possible to gain a much deeper understanding of C. Seasonal and daily variability their true seasonal distribution patterns in the Pacific Ocean. The seasonal trend in boings detected at the ACO was In addition to seasonal variability, there was also high similar to that found by Thompson and Friedl (1982). daily variability in the occurrence of boings. For example, Thompson and Friedl (1982) conducted a long-term study of Fig. 5 shows that the mean hourly boing rate on 12th Janu- sounds produced by several species of whales recorded using ary, 2008 was 62 boings/h, while the very next day, mean two bottom-mounted hydrophones located 11.6 km apart and hourly boing rate decreased to zero boings per hour. This at a depth of approximately 800 m off Kahuku Point, Oahu high variability is likely related to the number of minke (McDonald and Fox, 1999). They reported that boings were whales in the area each day. Thompson and Friedl (1982) heard from November to April and were most abundant in reported that when one sound source (minke whale) was February (Thompson and Friedl, 1982). Similarly, boings present, it produced a boing approximately once for every 6 were heard at the ACO from October to May, with a peak in min, but if two or more whales were present, the boing pro- March and a steady drop-off after that. There was no second- duction increased dramatically to an average of once for ev- ary peak observed by Thompson and Friedl (1982) or in the ery 30 s. So, boing rate not only increases because of an

3358 J. Acoust. Soc. Am., Vol. 129, No. 5, May 2011 Oswald et al.: Boing detections at Station ALOHA Author's complimentary copy increase in the number of animals producing boings, but also look at the relationship between the number of whales pres- because each animal is producing boings more frequently. ent, their proximity to each other and boing rate. Thompson and Friedl’s (1982) observation that boing In addition to providing information regarding the distri- rates increase when two or more minke whales are present in bution and abundance of minke whales in Hawaiian waters, an area suggests that boings may function to maintain spac- the study of boings can also provide insight into why minke ing among animals. This function has been suggested for whales are present and producing boings in these waters dur- humpback whale song (Frankel et al., 1995) and for the “star ing the winter and spring. There are near-monthly cruises to wars vocalization” produced by dwarf minke whales off the Station ALOHA as part of the Hawaii Ocean time-series coast of eastern Australia (Gedamke et al., 2003). The struc- (HOT) program to measure a variety of physical, chemical, tural similarities between the boing and the star wars sounds, and biological properties of the water column and provide a and the close evolutionary relationship between northern comprehensive description of the ocean at this site (Karl and hemisphere minke whales and dwarf minke whales Lukas, 1996). An examination of the relationships between (Gedamke et al., 2001) suggest that the star wars sound and variables such as primary production, plankton community the boing may be used in similar ways. Future studies should structure, and temperature and the presence of minke whales include recordings made using multiple hydrophones, acous- would help illuminate the ways in which minke whales may tic tags, and/or focal follows using passive acoustics (e.g., be utilizing Hawaiian waters. towed arrays) to further investigate the relationship between the number of, and spacing among, minke whales and boing ACKNOWLEDGMENTS rate. This information is not only important for determining the function of boings, but also when considering methods The authors would like to extend their thanks to Jim for estimating abundance based on calling rates (see Mar- Jolly for all of his help with accessing ACO data files. We ques et al., 2009). are grateful to Sofie Van Parijs, Denise Risch, and Kate Staf- ford for their tireless help with XBAT. Thank you also to D. Diel variation Charlotte Fabjanczyk, Shannon Coates, and Michael Oswald for their assistance with data analysis and to Steve Martin There was no significant diel variation in mean adjusted for providing the code for measuring pulse repetition rate. boing rates at the ACO. This suggests that towed array sur- We thank Tom Norris, Shannon Rankin, and Michael veys that take place during daylight hours are an appropriate Oswald for their insights and helpful suggestions on drafts of method for studying minke whales. Surveys conducted dur- this manuscript. Paul Johnson kindly produced the map in ing daylight hours will reasonably sample vocal behavior Fig. 3. The ACO is funded by the National Science Founda- and provide opportunities for visual observations, biopsy tion. J.N.O. was supported by the Hunt Fellowship from the sample collection, and tag applications. These additional Acoustical Society of America. This is HIMB contribution data will provide information on how frequently individual No. 1346. This is SOEST contribution No. 8115. minke whales produce boings, what behaviors boing production is related to, and which age/sex classes are producing Baker, C. S., and Herman, L. M. (1981). “Migration and local movement of boings. Knowing which age/sex classes produce boings is humpback whales (Megaptera novaeangliae) through Hawaiian waters,” Can. J. Zool. 59, 460–469. particularly important when using acoustics to determine Bowen, W. D., and Siniff, D. B. (1999). “Distribution, population biology distribution and abundance. For example, if boings are pro- and feeding ecology of marine mammals,” in Biology of Marine Mam- duced only by males, as has been shown for certain sounds mals, edited by J. E. Reynolds III and S. A. Rommel (Smithsonian Institu- produced by other species (e.g., humpback whales, Tyack, tion Press, Washington, DC), pp. 423–484. Burtenshaw, J. C., Oleson, E. M., McDonald, M. A., Hildebrand, J. A., 1981; fin whales, Croll et al., 2002; blue whales, Oleson et Andrew, R. K., Howe, B. A., and Mercer, J. A. (2004). “Acoustic and sat- al., 2007b), then using boings as a proxy for animal occur- ellite remote sensing of blue whale seasonality and habitat in the northeast rence would be possible for only a portion of the population Pacific,” Deep Sea Res. II 51, 967–986. (i.e., males), and nothing could be said about the distribution Clapham, P. J., and Matilla, D. K. (1990). “Humpback whale songs as indicators of migration routes,” Marine Mammal Sci. 6, 155–160. and abundance of females or other individuals that do not Clark, C. W. (1982). “The acoustic repertoire of the southern right whale, a produce boings. quantitative analysis,” Animal Behav. 30, 1060–1071. Croll, D. A., Clark, C. W., Acevedo, A., Tershy, B., Flores, S., Gedamke, J., E. Future research and Urban, J. (2002). “Only male fin whales sing loud songs,” Nature 417, 809. The amount of information that can be gleaned from a Delarue, J., Laurinolli, M., and Martin, B. (2009). “Bowhead whale (Balaena mysticetus) songs in the Chukchi Sea between October 2007 and single, stationary hydrophone is limited. It is difficult to May 2008,” J. Acoust. Soc. Am. 126, 3319–3328. localize animals with accuracy using a single hydrophone Duennebier, F., Harris, D., and Jolly, J. (2008). “Aloha cabled observatory and therefore we do not know how many animals were pres- will monitor ocean in real time,” Sea Tech. 51054. ent or the distance at which boings were detected. Plans are Edds-Walton, P. L. (2000). “Vocalisations of minke whales (Balaenoptera acutorostrata) in the St. Lawrence Estuary,” Bioacoustics 11, 31–50. in place to install two hydrophones at the ACO in 2011 and Frankel, A. S., Clark, C. W., Herman, L. M., and Gabriele, C. M. (1995). therefore it will likely be possible to obtain this information “Spatial distribution, habitat utilization, and social interactions of hump- from future recordings. This will not only allow us to deter- back whales, Megaptera novaeangliae, off Hawai’i, determined using mine how many vocalizing minke whales are in the ACO acoustic and visual techniques,” Can. J. Zool. 73, 1134–1146. Gedamke, J., Costa, D. P., and Dunstan, A. (2001). “Localization and visual area at any given time, but it will also allow us to monitor verification of a complex minke whale vocalization,” J. Acoust. Soc. Am. the movements of whales within the area and to take a closer 109, 3038–3047.

J. Acoust. Soc. Am., Vol. 129, No. 5, May 2011 Oswald et al.: Boing detections at Station ALOHA 3359 Author's complimentary copy Gedamke, J., Costa, D. P., Clark, C. W., Mellinger, D. K., and O’Neill, F. L. Oleson, E. M., Calambokidis, J., Burgess, W. C., McDonald, M. A., LeDuc, (2003).“The role of song in the spacing of singing dwarf minke whales C. A., and Hildebrand, J. A. (2007b). “Behavioral context of call produc- (Balaenoptera acutorostrata): Preliminary results from playback tion by eastern north Pacific blue whales,” Mar. Ecol. Prog. Ser. 330, 269– experiments,” in First International Conference on Acoustic Communica- 284. tion by Animals, July 27–30, College Park, Maryland, pp. 95–96. Perrin, W. F., and Brownell, Jr., R. L. (2002). “Minke whales,” in Horwood, J. (1990). Biology and Exploitation of the Minke Whale (CRC Encyclopedia of Marine Mammals, edited by W. F. Perrin, B. Wursig, Press, Boca Raton, FL), pp. 21–59. and J. G. M. Thewissen (Academic Press, San Diego, CA), pp. 750– Karl, D. M., and Lukas, R. (1996). “The Hawaii Ocean time-series (HOT) 754. program: Background, rationale and field implementation,” Deep Sea Res. Rankin, S., and Barlow, J. (2005). “Source of the North Pacific “boing” II 43, 129–156. sound attributed to minke whales,” J. Acoust. Soc. Am. 118,3346– Leatherwood, S., Thomas, J. A., and Awbrey, F. T. (1981). “Minke whales 3351. off northwestern Ross Island,” Antarctic J. 16, 154–156. Sakamoto, C. M., Karl, D. M., Jannasch, H. W., Bidigare, R. R., Letelier, R. Lockyer, C. H., and Brown, S. G. (1981). “The migration of whales,” in Ani- M., Walz, P. M., Ryan, J. P., Polito, P. S., and Johnson, K. S. (2004). mal Migration, edited by D. J. Aidley (Cambridge University Press, New “Influence of Rossby waves on nutrient dynamics and the plankton com- York, NY), pp. 105–137. munity structure in the North Pacific subtropical gyre,” J. Geophys. Res. Marques, T. A., Thomas, L., Ward, J., DiMarzio, N., and Tyack, P. L. 109, C05032, doi: 10.1029/2003JC001976, 12 pp. (2009). “Estimating cetacean population density using fixed passive acous- Schevill, W. E., and Watkins, W. A. (1972). “Intense low-frequency sounds tic sensors: An example with beaked whales,” J. Acoust. Soc. Am. 125, from an Antarctic minke whale, Balaenoptera acutorostrata,” Breviora 1982–1994. 388, 1–8. McDonald, M. A., and Fox, C. G. (1999). “Passive acoustic methods applied Stafford, K. M., Moore, S. E., and Fox, C. G. (2005). “Diel variation in blue to fin whale population density estimation,” J. Acoust. Soc. Am 105, whale calls recorded in the eastern tropical Pacific,” Animal Behav. 69, 2643–2651. 951–958. McDonald, M. A., Hildebrand, J. A., and Webb, S. C. (1995). “Blue and fin Stafford, K. M., Nieukirk, S. L., and Fox, C. G. (1999). “An acoustic link whales observed on a seafloor array in the northeast Pacific,” J. Acoust. between blue whales in the eastern tropical Pacific and the northeast Pacif- Soc. Am. 109, 1728–1735. ic,” Marine Mammal Sci. 15, 1258–1268. Mellinger, D. K., and J. W. Bradbury (2007). “Acoustic measurement of ma- Thompson, P. O., and Friedl, W. A. (1982). “A long term study of low fre- rine mammal sounds in noisy environments,” Proceeding of 2nd Interna- quency sounds from several species of whales off Oahu, Hawaii,” Cetol- tional Conference on Underwater Acoustic Measurements: Technologies ogy 45, 1–19. and Results, 25–29 June 2007, Herakion, Greece, 8 pp. Tyack, P. (1981). “Interactions between singing humpback whales and con- Mellinger, D. K., Carson, C. D., and Clark, C. W. (2000). “Characteristics specifics nearby,” Behav. Ecol. Sociobiol. 8, 105–116. of minke whale (Balaenoptera acutorostrata) pulse trains recorded near Wenz, G. M. (1964). “Curious noises and the sonic environment in the Puerto Rico,” Marine Mammal Sci. 16, 739–756. ocean,” in Marine Bio-acoustics, edited by W. N. Tavolga (Pergamon, Norris, T. F., McDonald, M., and Barlow, J. (1999). “Acoustic detections of New York), pp. 101–119. singing humpback whales (Megaptera novaeangliae) in the eastern North Wiggins, S. M., Oleson, E. M., McDonald, M. A., and Hildebrand, J. A. Pacific during their northbound migration,” J. Acoust. Soc. Am. 106, (2005). “Blue whale (Balaenoptera musculus) diel call patterns offshore 506–514. of southern California,” Aqat. Mamm. 31, 161–168. Oleson, E. M., Wiggins, S. M., and Hildebrand, J. A. (2007a). “Temporal Winn,H.E.,andPerkins,P.J.(1976). “Distribution and sounds of separation of blue whale call types on a southern California feeding the minke whale, with a review of mysticete sounds,” Cetology 19, ground,” Animal Behav. 74, 881–894. 1–12.

3360 J. Acoust. Soc. Am., Vol. 129, No. 5, May 2011 Oswald et al.: Boing detections at Station ALOHA Author's complimentary copy