Workshop Proceedings: Short Papers

Rodney Rountree¹, Francis Juanes², and Joseph E. Blue³

¹School for Marine Science and Technology, UMASS Dartmouth, 706 Rodney French Blvd., New Bedford, MA 02744-1221 rrountree@UMassD.Edu

²Department of Natural Resources, University of Massachusetts, Amherst, MA 01003

³President, Leviathan Legacy, Inc., 3313 Northglen Drive, Orlando, FL 32806 jblue46498@aol.com

Since the seminal work of Fish and Mowbray (1970), little advancement has been made towards the study of soniferous fishes from the marine waters of the Northeastern United States. A review of the literature suggests at least 51 fishes are vocal in New England waters (Table 1), although many of these species are uncommon stragglers to our waters. Spontaneous sound production is known from only about half of these species. However, laboratory studies are often hampered by the difficulty of maintaining healthy specimens, and the difficulty of inducing natural behaviors such as spawning under confinement. This is further complicated by the fact that many fish are primarily vocal during the spawning season, and may not vocalize until maturity, and because vocal behavior is usually limited to males (e.g., haddock and weakfish). The objectives of this study were to conduct a pilot field survey of soniferous fishes in Massachusetts’s waters to determine what species are vocal and examine temporal patterns in vocal behavior. However, because of the unexpected finding of widespread calls of the striped cusk-eel on Cape Cod, this paper will focus on this enigmatic species.

Recordings of fish sounds were made at 12 different sites across Cape Cod at least once between June and October 2001. However, the primary sampling location was the Cotuit Town Landing which was sampled on 18 different dates, including 5 dates on which monitoring was conducted over the diel cycle. Except for the diel studies, most sampling was conducted around sunset, usually beginning 1 to 2 hours before sunset and continuing for 2 to 3 hours after sunset. To obtain information on the daily pattern of fish calls, diel studies were conducted on five different dates at Cotuit Docks. For these studies, sounds were recorded approximately from 1300-1400, 1900-2300, 0100-0200, and 0400-0600, corresponding to afternoon, sunset, night, and sunrise periods, respectively. Low cost hydrophones (Arretec, PB 3098 Bletchley, Milton Keynes MK2 2AD, United Kingdom) were deployed from docks, piers, jetties and small boats and recorded to a hi-fi VCR. Occasionally, recordings were made to a Sony hand-held tape recorder (model TCM-929). In addition, whenever possible, video recordings were made simultaneously to the VCR using a hand-deployed underwater video camera equipped with infrared lights (models made by Vista Cam, 9911 Goodhue St. NE, Blaine MN 55449, and Aqua vu, Nature Vision Inc., 213 NW 4^th St., Brainerd, MN 56401). Sounds were captured to a PC while playing back from a VCR using Cool Edit 2000 (made by Syntrillium Software Corporation). Some spectral analyses were also conducted using Signal for Windows (Engineering Design, 43 Newton St, Belmont, MA 02478). To quantify call frequency, 1-4 hour sound samples were divided into 10-minute segments and a randomly selected 2 minute sound clip was obtained from each. Calls for toadfish, striped cusk-eel and searobins were identified and counted. Reference sound clips of unknown calls were made and used to make counts of unknown sounds by type (e.g., "grunt-A", etc.).

Over 53 VHS and 12 cassette tapes comprising over 160 hours of recordings were collected. Calls of striped cusk-eels, Ophidion marginatum, oyster toadfish, Opsanus tau, and striped searobin, Prionotus evolans, dominated the observations. Several unidentified calls were also common. We are continuing our efforts to identify these calls. In addition, various sources of natural and man-made noise were also recorded including: outboard boats, barges, jet-skies, dock noises, fishing noises, depth-finders, and gas release from sediments. Based on the occurrence of vocal choruses, we found sunset spawning aggregations of the striped cusk-eels at eight of 12 locations sampled across the length of Cape Cod, including two sites (Barnstable Harbor and Provincetown Harbor) on the north shore. Cusk-eels were recorded from the first sampling date (June 11) through the end of August, but abruptly stopped by early September. Oyster toadfish were also already calling at the start of the field season, but sunset choruses had ceased by mid-July. Striped searobin calls were not associated with sunset, but occurred throughout the night. Sea robin calls were most frequent in August and September but were still present in October. The cusk-eel sounds recorded in MA are nearly identical to striped cusk-eel sounds recorded by the first author under laboratory conditions in New Jersey (Mann et al 1997), and more recent sounds recorded in the field and attributed to stripe cusk-eels in Narragansett Bay (Perkins 2001) and North Carolina (Sprague and Luczkovich 2001). Our attribution of these sounds to the striped cusk-eel is further validated by the capture of a 170 mm TL specimen while recording sounds in Cotuit, MA in July 2001, and by subsequent sightings of a larger individual later that same month. Cusk-eels can sometimes be observed in the shallows at night with the aid of a spot light (Rountree, pers. Observ.). In Figure 1, chatters vary in relative amplitude and

Figure 1. Representative sample of cusk-eel calls recorded in Cotuit, MA on June 20th 2001. The lower panel shows a string of six separate cusk-eel calls, likely from six different individuals. The first call overlaps with that of a toadfish. The upper left figures show the waveform and spectrogram for call 3. The power spectrum of call 3 is shown in the upper right panel.

range form 8 to 16 pulses and call times of 275 msec to 730 msec. The dominant frequency was 1098-1866 Hz (compared to the toadfish call at the beginning of the sequence at 171-585 Hz). A sample call recorded from Provincetown, MA on August 23, 2001 is shown in Figure 2.

Figure 2. Single chatter attributed to the striped cusk-eel, Ophidion marginatum, recorded from Provincetown, MA on 23 August 2001. The lower panel shows the energy spectrum for the entire call, while the upper panels show the waveform, energy spectrum and power spectrum of a single pulse.

This call is considerably longer (31 pulses, 1,715 msec) than those in Figure 1, but is still well within the range characteristic of the species (Mann et al 1997, Sprague and Luczkovich 2001). A single representative pulse has most energy between 914 and 1524 Hz (Fig. 2).

Striped cusk-eel calls can be heard sporadically throughout the day, but calls clearly become more frequent at sunset (Fig. 3). Peak number of calls occurred between 20 to 60 minutes after sunset, and declined to near zero within two hours.

Figure 3. Daily pattern of striped cusk-eel, Ophidion marginatum, calls collected on two dates (20-21 June and 2-3 July, 2001). All calls heard within 2 minutes sound clips were counted. Sample clips were taken randomly from within 10-minute sample bins. The vertical arrow marks the time of sunset as obtained from a hand-held GPS.

In contrast, the oyster toadfish calls more frequently during the day, but also exhibits a strong increase in activity associated with sunset (Figure 4). Although data is more limited, peak activity occur 1-2 hours after sunset, with more gradual declines through the night compared to the striped cusk-eel.

Figure 4. Daily pattern of oyster toadfish, Opsanus tau, call on June 20-21, 2001.

Discussion

It is significant that the striped cusk-eel was the most frequently heard and widely distributed species encountered during this study as it has previously been thought to occur from Block Island south to Florida, with only rare stragglers occurring as far north as Cape Cod (Collette and Klein-MacPhee 2002), despite extensive faunal surveys in the region over several decades. This finding nicely demonstrates the usefulness of passive acoustics as a supplement to traditional survey methods. The seasonal and daily pattern of striped cusk-eel vocal activity agrees with published laboratory findings (Mann et al. 1997, Sprague and Luczkovich 2001). Striped cusk-eels were already chorusing by mid-June when sampling began, but had stopped by mid-September in good agreement with previous studies. Call frequency increases rapidly at sunset developing into a loud chorus that lasts from 1 to 2 hours (Fig. 3). Captive cusk-eels have been observed to chorus after sunset as part of courtship and spawning behavior (Mann et al. 1997, Rountree and Bowers-Altman 2002). We believe that our observations suggest widespread spawning of striped cusk-eels within estuaries of both the north and south shores of Cape Cod. The species cryptic nocturnal behavior, and habit of remaining burrowed during the day likely account for the failure of previous researchers using conventional sampling gears (i.e., trawls and seine sampling mostly limited to daylight hours) to recognize its importance to the region. At this time the northern range of the striped cusk-eel must be reconsidered. How much farther up the cost the species extends is unknown. It is notable that Geoghegan et al. (1998) recorded a single adult stripe cusk-eel at Seabrook, New Hampshire and argued that it might represent a small local population. Therefore, we suspect that reproducing populations of this species may occur at least to New Hampshire waters. However, the scarcity of ophidiid eggs in ichthyoplankton surveys of the region is puzzling (e.g., Fahay 1992) and future studies on the distribution and ecology of this cryptic species are needed. Boat sounds were problematic during the day, sometimes occurring during 50-99% of the sound sample clips. During these time, sounds of fishes could not be heard above the boats noise. Boat noise was rare during the evening hours. The impact of boat-associated noise on the behavior of fishes is poorly known, but it had a strong impact on our ability to record day-time fish sounds. It is hoped that the newly available archive of fish sounds originally published by Fish and Mowbray (1970) and recently repackaged by the University of Rhode Island (Rountree et al 2002) will aid in the identification of the unknown calls recorded on Cape Cod. In summary this study has demonstrated the usefulness of even low-cost passive acoustics technology as a tool to survey estuarine and marine fishes. Information on the temporal and spatial patterns of fish vocal behavior can be used to gain insight into temporal and spatial patterns in habitat use patterns by vocal species. In particular, identification of spawning habitats through passive acoustics surveys is promising.

Megan Hendry-Brogan and Katie Anderson are thanked for diligent work in both the field and laboratory to collect and process fish sound data. This project received major funding from the Northeast and Great Lakes National Undersea Research Center, which also provided extensive logistical support. The Woods Hole Sea Grant College Program also provided supporting funds.

The Sounds Conservancy, Quebec-Labrador Foundation/Atlantic Center for the Environment provided a stipend for Megan’s fieldwork.