Assessing exposure and response of three species of odontocetes to mid-frequency active sonar during submarine commanders courses at the Pacific Missile Range Facility: August 2013 through February 2015


Baird, R.W., S.W. Martin, R. Manzano-Roth, D.L. Webster, and B.L. Southall. 2017. Assessing exposure and response of three species of odontocetes to mid-frequency active sonar during submarine commanders courses at the Pacific Missile Range Facility: August 2013 through February 2015. Prepared for U.S. Pacific Fleet, submitted to NAVFAC PAC by HDR Environmental, Operations and Construction, Inc., Honolulu, Hawai‘i.


The United States Navy’s marine species monitoring program addresses four general topics surrounding the impact of mid-frequency active sonar (MFAS) on protected species: occurrence, exposure, response, and consequences. Occurrence of odontocete cetaceans on and around the Pacific Missile Range Facility (PMRF) has been studied for several years using a combination of satellite tags and photo-identification, demonstrating the existence of resident populations of several species of odontocetes and resulting in considerable information on diving behavior and movement patterns of these species. In an effort to measure and evaluate both exposures and responses to MFAS during actual military training exercises, we used data from 20 satellite tags deployed on odontocetes prior to three Submarine Commanders Courses (SCC) held on PMRF between August 2013 and February 2015.

MFAS use during each SCC occurred under normal operating conditions north of the Kaulakahi Channel and spanned a three-day period. Sonar transmissions were not controlled according to the location of tagged individuals, but this orientation provided a spatial-temporal opportunity for individuals to avoid MFAS exposure by moving to the opposite sides of either Kaua‘i or Ni‘ihau. Eleven of the 20 tags had either stopped transmitting prior to the start of the SCC or the tagged individuals were far from MFAS and thus exposure levels could not be estimated. For the other nine individuals, we combined locations obtained from satellite tags with Navy-provided data on MFAS use and ship tracks to assess MFAS exposure and potential responses. Subjects included: false killer whales, Pseudorca crassidens (n=1), short-finned pilot whales, Globicephala macrorhynchus (n=5), and rough-toothed dolphins, Steno bredanensis (n=3). Individuals from all three species were known to be part of island-resident populations, with false killer whales from the Northwestern Hawaiian Islands population, and pilot whales and rough-toothed dolphins from populations that generally range from Ni‘ihau to O‘ahu, but whose core areas encompass the Kaulakahi Channel and southern portions of PMRF.

Methods used were similar to earlier analyses but with several improvements. Received levels (RLs) were estimated using the Peregrine propagation model for each satellite tag location within 1 hour of an MFAS transmission, allowing for calculations of thousands of RL estimates relatively quickly. We explicitly accounted for known uncertainty associated with Argos location classes (LCs) by calculating 1,000 RL estimates along a radial through the tagged animal location, with the radial length reflecting the LC uncertainty. Both median and mean (with standard deviations [SDs]) RL levels of the 1,000 estimates were used to address variability associated with LCs. Consistency between mean and median RLs and small SDs suggest that resulting estimates are relatively robust. Estimated RLs were determined at tagged animal locations for both 10-meter depths (m) (+/- 5 m) and at depths representing typical dive depths for each of the three species (false killer whale – 50 m; rough-toothed dolphin – 50 m; short-finned pilot whale – 500 m). Estimated RLs at depth representing typical dives were generally lower than at 10-m depth for all species.

The false killer whale was estimated to have been intermittently exposed to MFAS at distances ranging from 6.5 to 75.4 kilometers (km) over a 1.6-day span. During the period of MFAS exposure the false killer whale transited away from an area of relatively low exposure (starting at an estimated RL of mean = 90.9 [7.68]; median = 89.4 decibels referenced to 1 microPascal root mean square (dB re: 1 μPa RMS; hereafter dB)) to the area of highest RL (mean (SD) = 160.2 (9.55); median = 156.6 dB). The individual then moved away from the area where MFAS was being used for several hours, then moved back through the area of exposure (to an estimated maximum mean RL = 150.8 [7.05]; median = 157.6 dB), and then to an area of lower RLs. The three rough-toothed dolphins were exposed to MFAS at ranges of 19.5 to 94.4 km, with maximum estimated mean (SD) RLs at 10 meters of 150.6 (0.96), 155.3 (3.5), and 157.1 (1.5) dB. The individual with highest estimated RLs (SbTag014) moved from an area farther from the MFAS source into an area with the maximum estimated RL before moving into an area with lower RLs (<140 dB). The five short-finned pilot whales represented three different groups. One of the three groups, which included three different tagged individuals, was exposed to MFAS at ranges of 3.2 to 48.1 km, while the others were exposed at distances of 14.9 to 39.5 km and 48.0 to 57.3 km. Two individuals (GmTag081 and GmTag083) exposed at relatively short distances had relatively high estimated RLs at 10 m (GmTag081 mean = 169 [1.41], median = 168.9 dB; GmTag083 mean = 168.3 [1.50], median = 167.9 dB). No large-scale movements of the individuals away from areas of relatively high RLs, for example to areas in the lee of Kaua‘i or Ni‘ihau, were observed. An almost complete dive record spanning the period from before to after the SCC (22, days with 1,363 dives) was obtained for only one individual, the short-finned pilot whale (GmTag081) with the highest estimated RLs.

Clear changes in diving behavior were documented during the SCC in comparison to pre- and post-SCC periods for the short-finned pilot whale GmTag081. Dive rates during the SCC were lower both during the day and night in comparison to the pre- and post-SCC periods. Day-time dive depths were significantly deeper during the SCC, while night-time dive depths were similar for all three periods. These exposure case studies indicated no large-scale avoidance of areas with moderately high (>150 dB) MFAS RLs where responses might be expected to be likely by nine individuals of three species spanning two years. However, clear behavioral changes during the SCC were observed for one individual for which detailed dive data were available. All individuals were from populations that are generally resident to the area. Given that MFAS has been used in Hawai‘i for many years, these individuals have likely been exposed to MFAS on multiple previous occasions.

Download PDF

Associated projects