| "Some people talk to animals. Not many listen though. That's the problem."
Acoustic diversity of the fishes of Stellwagen Bank National Marine Sanctuary
Working in partnership with SBNMS, between January 2006 and February 2007, a number of marine autonomous recording units (MARUs; Cornell University Bioacoustics Research Program (Clark et al., 2002)) were deployed within the Sanctuary to calculate the spatial and temporal variability of the soundscapes and to detect vocally active marine species. The units continuously sampled at a rate of 2000 Hz with a flat (±1.0 dB) frequency response sensitivity of ~151.2 dB re 1 ?Pa between 10 and 585 Hz (HTI-94-SSQ, High Tech Inc., Gulfport, MS, USA). For this study, five of these recording locations were selected, representing the greatest possible diversity of bottom types and temporal coverage in SBNMS during 2006.
Using previously-used sound classification parameters (Casaretto and Hawkins, 2002; Rowe and Hutchings, 2006), Raven Pro 1.5 Software (Bioacoustics Research Program, Cornell Laboratory of Ornithology Cornell, NY, USA) and an Atlantic cod detection algorithm (Urazghildiiev and Van Parijs, 2016), we identified call types and spatiotemporal patterns in the vocalizations of fishes within SBNMS during December 2006.
During the three days over each moon phase, the total numbers of vocalizations at each of the five selected recording sites were classified into call type.
A total of nine unique fish vocalization types were identified over the five recording sites over the three moon phases of December 2006. Generally, the majority of the energy contained within these calls was between 50 - 900 Hz. Overall, there were significant differences in the total number of vocalizations among the moon phases (Figure 2), with significantly higher calls occurring during the Full and Quarter moons, compared to the New moon. There were also significant differences in the total number of vocalizations among sites (Figure 2), with Site 27 and 33 having the highest number of calls, followed by Site 32, and then Sites 13 and 18.
Atlantic Cod (Gadus morhua)
Atlantic cod vocalizations were recorded in four of the five recording sites.
Overall, there were significant differences in the total number of cod vocalizations among moon phases, with a significantly higher number of vocalizations occurring during the full moon (Figure 4). There was also a significant difference in the number of total vocalizations among sites; with Site 27 have the highest number of vocalizations followed by Sites 18 and 13, and then Site 33.
Generally, when the total number of cod vocalizations was low (<10 per hour; new and quarter moon phases) there was no diurnal trend in total number of vocalizations. However, during the full moon at Site 27, the vocalization rate rose dramatically and there was significant differences in vocalization rates between dark and light hours.
Haddock (Melanogrammus aeglefinus)
Haddock vocalizations were recorded in only one of the five recording sites.
Overall, there was a significant difference in the number of haddock vocalizations among moon phases (Kruskal-Wallis; P = 0.004), with a significantly higher number of vocalizations occurring during the full moon, compared to both new and quarter moon phases.
In contrast to cod, haddock vocalizations showed significant diurnal trends during all moon phases (Mann-Whitney; P < 0.01).
Overall the results on the extent and patterns of fish vocalizations within SBNMS observed in the current study align with previous research performed in the area (Rountree and Juanes, 2010; Hernandez et al., 2013). Multiple species of soniferous fishes have been documented to inhabit SBNMS, including but not limited to Atlantic cod, haddock, pollock, cusk, and cusk eel (Lindholm et al., 2007; Rountree and Juanes, 2010; Hernandez et al., 2013). Many of these species were also identified in the current study. Of the vocalizations which could be reliably identified to species, Atlantic cod, haddock, and pollock dominated the recordings from December 2006. There were also several additional vocalization types that were heard consistently throughout the recordings, however, definitive identification of the species was not possible at this time.
Marked differences in faunal assemblages due to varying benthic complexity and/or bottom type is well documented in the literature (Scott, 1982; Friedlander and Parrish, 1998; Kaiser, 1999; Rountree and Juanes, 2010), and is also supported in the results from the current study. The presence of certain vocalization types differed with site and in some cases moon phase, with several call types (e.g., haddock and cod) only being present in a sub-set of the recording sites. The five sites were originally selected to represent the diversity of bottom types found within SBNMS, and the differences in acoustic activity observed are likely due to the preferential habitat selection of certain species (Lindholm and Auster, 2003; Bergmann et al., 2004; Lindholm et al., 2007).
Consistent with previous reports on the temporal trends in fish vocalization patterns, the vocalizations of haddock and cod in the current study were most prevalent and increased in total number during dark periods (dusk and night), and during the full moon phase (when vocalization rate exceeded 10/hour) (Locascio and Mann, 2008; Casaretto et al., 2014; Butler et al., 2016; Mooney et al., 2016). These observations are also consistent with the reproductive behavior of haddock. Males produce acoustic and visual displays during patrolling behavior, which leads to courtship interactions and spawning occurring more often and for longer periods during nighttime hours (Casaretto et al., 2015). The same diurnal trends have been recorded during studies of spawning cod in captivity (Brawn, 1961; Rowe and Hutchings, 2006). Rowe and Hutchings (2006) observed that during the spawning season vocalizations were produced at all times of the day, however, was significantly more frequent after sunset and continued at high levels throughout the dark hours.
The increased acoustic activity during the full moon observed in the current study is likely due to a peak in spawning behavior. The spawning behavior of many fishes is known to be mediated by moon phase (Taylor, 1984; Sala et al., 2003; Gordoa and Carreras, 2014), however, biological or ecological explanations for the observed periodicity vary among species, some of which benefit the progeny and some the adults; although several hypotheses apply to multiple species (Rhodes and Sadovy, 2002). Increased light levels during the full moon are hypothesized to enhance migration to a spawning ground and conspecific coordination of activity (Colin et al., 1987; Gladstone, 2007). Additionally, tidal amplitudes associated with full and new moon phases are thought to benefit long distance transport and dispersal of larvae (Barlow, 1981; Johannes, 1988).
To learn more about the issue of noise in Stellwagen Bank National Marine Sanctuary, click here.
Partners: NOAA Stellwagen Bank National Marine Sanctuary, Cornell University
Primary Funders: NOAA Ocean Acoustics Program, NOAA Stellwagen Bank National Marine Sanctuary, National Oceanographic Partnership Program