Taking and Importing Marine Mammals; Taking Marine Mammals Incidental to the U.S. Navy Operations of Surveillance Towed Array Sensor System Low Frequency Active Sonar

DEPARTMENT OF COMMERCE

National Oceanic and Atmospheric Administration

50 CFR Part 216

[Docket No. 070703226-7461-02; I.D. 062206A]

RIN 0648-AT80

Taking and Importing Marine Mammals; Taking Marine Mammals Incidental to the U.S. Navy Operations of Surveillance Towed Array Sensor System Low Frequency Active Sonar

AGENCY:

National Marine Fisheries Service (NMFS), National Oceanic and Atmospheric Administration (NOAA), Commerce.

ACTION:

Final rule.

SUMMARY:

NMFS, upon application from the U.S. Navy, is issuing regulations to govern the unintentional taking of marine mammals incidental to Navy operation of the Surveillance Towed Array Sensor System Low Frequency Active (SURTASS LFA) Sonar. Issuance of regulations, and Letters of Authorization issued under these regulations, is required by the Marine Mammal Protection Act (MMPA) when the Secretary of Commerce (Secretary), after notice and opportunity for comment, finds, as here, that such takes will have a negligible impact on the affected species or stocks of marine mammals and will not have an unmitigable adverse impact on their availability for taking for subsistence uses. These regulations set forth the permissible methods of take and other means of effecting the least practicable adverse impact on the affected species or stocks of marine mammals and their habitat.

DATES:

Effective from August 16, 2007, through August 15, 2012.

ADDRESSES:

A copy of the application, containing a list of references used in this document, and other documents cited herein, may be obtained by writing to P. Michael Payne, Chief, Permits, Conservation and Education Division, Office of Protected Resources, National Marine Fisheries Service, 1315 East-West Highway, Silver Spring, MD 20910-3225, by telephoning one of the contacts listed under

FOR FURTHER INFORMATION CONTACT

, or at:

http://www.nmfs.noaa.gov/pr/permits/incidental.htm.

A copy of the Navy's Final Supplemental Environmental Impact Statement (Final SEIS) and the Final Environmental Impact Statement (Final EIS) can be downloaded at:

http://www.surtass-lfa-eis.com.

Documents cited in this rule may also be viewed, by appointment, during regular business hours at this address.

FOR FURTHER INFORMATION CONTACT:

Kenneth Hollingshead, NMFS, at 301-713-2289, ext 128.

SUPPLEMENTARY INFORMATION:

Background

Section 101(a)(5)(A) of the Marine Mammal Protection Act (16 U.S.C. 1361

et seq.

) (MMPA) directs the Secretary of Commerce (Secretary) to allow, upon request, the incidental, but not intentional taking of marine mammals by U.S. citizens who engage in a military readiness activity if certain findings are made and regulations are issued.

The MMPA directs the Secretary to allow the requested incidental taking during periods of not more than 5 consecutive years each if the Secretary finds that the total taking will have a negligible impact on the affected species or stock(s), will not have an unmitigable adverse impact on the availability of the species or stock(s) for certain subsistence uses. The Secretary must also issue regulations setting forth the permissible methods of taking and other means of effecting the least practicable adverse impact, including a consideration of personnel safety, the practicality of implementation of any mitigation, and the impact on the effectiveness of the subject military readiness activity, and the requirements pertaining to the monitoring and reporting of such taking. These regulations do not themselves authorize the taking of marine mammals. NMFS authorizes the incidental take through “letters of authorization” (LOAs) (50 CFR 216.106). Prior to issuance of an LOA, NMFS conducts a review of the activity and its impact on marine mammals (via the required monitoring, reporting and research) to ensure that the MMPA findings continue to be valid.

NMFS has defined “negligible impact” in 50 CFR 216.103 as “an impact resulting from the specified activity that cannot be reasonably expected to, and is not reasonably likely to, adversely affect the species or stock through effects on annual rates of recruitment or survival.” For the purposes of “military readiness activities” harassment is defined as:

(i) Any act that injures or has the significant potential to injure a marine mammal or marine mammal stock in the wild [Level A harassment]; or (ii) any act that disturbs or is likely to disturb a marine mammal or marine mammal stock in the wild by causing disruption of natural behavioral patterns, including, but not limited to, migration, surfacing, nursing, breeding, feeding, or sheltering, to a point where such behavioral patterns are abandoned or significantly altered [Level B harassment].

The term “military readiness activity” is defined in Public Law 107-314 (16 U.S.C. 703 note) to include all training and operations of the Armed Forces that relate to combat; and the adequate and realistic testing of military equipment, vehicles, weapons and sensors for proper operation and suitability for combat use. The term expressly does not include the routine operation of installation operating support functions, such as military offices, military exchanges, commissaries, water treatment facilities, storage facilities, schools, housing, motor pools, laundries, morale, welfare and recreation activities, shops, and mess halls; the operation of industrial activities; or the construction or demolition of facilities used for a military readiness activity.

Summary of Request

On May 12, 2006, NMFS received an application from the U.S. Navy requesting an authorization under section 101(a)(5)(A) of the MMPA for the taking of marine mammals by Level A and Level B harassment, incidental to deploying the SURTASS LFA sonar system for military readiness activities to include training, testing and routine military operations within the world's oceans (except for Arctic and Antarctic waters, coastal regions as specified in this rule, and offshore biologically important areas (OBIAs)) for a period of time not to exceed 5 years. According to the Navy's application, the Navy planned to operate the SURTASS LFA sonar system on a maximum of 4 ships in areas potentially including the Pacific, Atlantic, and Indian oceans and the Mediterranean Sea.

SURTASS LFA sonar provides the Navy with a reliable and dependable system for long-range detection of quieter, harder-to-find submarines. Low-frequency (LF) sound travels in seawater for greater distances than higher frequency sound used by most other active sonars. According to the Navy, the SURTASS LFA sonar system would meet the Navy's need for improved detection and tracking of new-generation submarines at a longer range. This would maximize the opportunity for U.S. armed forces to safely react to, and defend against, potential submarine threats while remaining a safe distance beyond a submarine's effective weapons range.

NMFS and the Navy have determined that the Navy's use of SURTASS LFA

sonar testing, training, and routine military operations constitute a military readiness activity because those activities constitute “training and operations of the Armed Forces that relate to combat” and constitute “adequate and realistic testing of military equipment, vehicles, weapons and sensors for proper operation and suitability for combat use.”

NMFS' current regulations governing takings incidental to SURTASS LFA sonar activities and the current LOA extends through August 15, 2007.

On September 28, 2006 (71 FR 56965), NMFS published a Notice of Receipt of Application on the U.S. Navy application and invited interested persons to submit comments, information, and suggestions concerning the application and the structure and contents of regulations. These comments were considered in the development of the proposed and final rules.

Prior Litigation, Involving LFA Sonar

On August 7, 2002, the Natural Resources Defense Council, the U.S. Humane Society and four other plaintiffs filed suit against the Navy and NMFS over SURTASS LFA sonar use and permitting. The U.S. District Court for the Northern District of California (Court) issued its Opinion and Order on the parties' motions for summary judgment in the SURTASS LFA sonar litigation on August 26, 2003. The Court found deficiencies in Navy and NMFS compliance with the MMPA, Endangered Species Act (ESA), and National Environmental Policy Act (NEPA). The Court determined that an injunction was warranted but did not order a complete ban on the use of SURTASS LFA sonar. Specifically, the Court found that a total ban on the employment of SURTASS LFA sonar would interfere with the Navy's ability to ensure military readiness and to protect those serving in the military against the threat posed by hostile submarines. The Court directed the parties to meet and confer on the scope of a tailored permanent injunction, which would allow for continued operation of the system with additional mitigation measures. The parties entered into a Stipulation Regarding Permanent Injunction that allowed the Navy to operate SURTASS LFA sonar from both

R/V Cory Chouest

and USNS IMPECCABLE (T-AGOS 23) in stipulated portions of the Northwest Pacific/Philippine Sea, Sea of Japan, East China Sea, and South China Sea with certain year-round and seasonal restrictions. The Court entered the Stipulation as an Order on October 14, 2003. On July 7, 2005, following mediation by the parties, the Court amended the injunction at Navy's request to expand the potential areas of operation based on real-world contingencies. The Navy began work on an SEIS, in response to the Court's ruling on the motion for preliminary injunction. The Navy's Final SEIS, which was completed in April 2007, not only addresses, the concerns identified by the Court in its ruling on the merits of the parties' summary judgment motions, but it also provides additional information regarding the environment that could potentially be affected by the SURTASS LFA sonar systems, and additional information related to mitigation.

A detailed description of the operations is contained in the Navy's application (DON, 2006) and the Final SEIS (DON, 2007) which are available upon request (see

ADDRESSES

).

Description of the Activity

The SURTASS LFA sonar system is a long-range, LF sonar (between 100 and 500 Hertz (Hz)) that has both active and passive components. It does not have to rely on detection of noise generated by the target. The active component of the system is a set of up to 18 LF acoustic transmitting source elements (called projectors) suspended from a cable underneath a ship. The projectors are devices that transform electrical energy to mechanical energy by setting up vibrations, or pressure disturbances, with the water to produce the pulse or ping. The SURTASS LFA sonar acoustic transmission is an omnidirectional (full 360 degrees) beam in the horizontal. A narrow vertical beamwidth can be steered above or below the horizontal. The source level (SL) of an individual projector in the SURTASS LFA sonar array is approximately 215 decibels (dB), and because of the physics involved in beam forming and transmission loss processes, the array can never have a sound pressure level (SPL) higher than the SPL of an individual projector. The expected water depth at the center of the array is 400 ft (122 m) and the expected minimum water depth at which the SURTASS LFA sonar vessel will operate is 200 m (656.2 ft).

The typical SURTASS LFA sonar signal is not a constant tone, but rather a transmission of various signal types that vary in frequency and duration (including continuous wave (CW) and frequency-modulated (FM) signals). A complete sequence of sound transmissions is referred to by the Navy as a “ping” and can last as short as 6 seconds (sec) to as long as 100 sec, normally with no more than 10 sec at any single frequency. The time between pings is typically from 6 to 15 minutes. Average duty cycle (ratio of sound “on” time to total time) is less than 20 percent; however, the duty cycle, based on historical operating parameters, is normally 7.5 percent.

The passive, or listening, component of the system is SURTASS, which detects returning echoes from submerged objects, such as submarines, through the use of hydrophones. The hydrophones are mounted on a horizontal array that is towed behind the ship. The SURTASS LFA sonar ship maintains a minimum speed of 3.0 knots (5.6 km/hr; 3.4 mi/hr) in order to keep the array deployed.

Because of uncertainties in the world's political climate, a detailed account of future operating locations and conditions cannot be predicted. However, for analytical purposes, a nominal annual deployment schedule and operational concept have been developed, based on current LFA sonar operations since January 2003 and projected Fleet requirements. The Navy anticipates that a normal SURTASS LFA sonar deployment schedule for a single vessel would involve about 294 days/year at sea. A normal at-sea mission would occur over a 49-day period, with 40 days of operations and 9 days transit. Based on a 7.5-percent duty cycle, the system would actually be transmitting for a maximum of 72 hours per 49-day mission and 432 hours per year for each SURTASS LFA sonar system in operation. (In actuality however, the combined number of transmission hours for LFA sonar employed on both the

R/V Cory Chouest

and the USNS IMPECCABLE (TAGOS 23) did not exceed 174 hours annually between August 16, 2002, and August 15, 2006 (Table 4 in the Navy's Final Comprehensive Report (Navy, 2007)).

Annually, each vessel will be expected to spend approximately 54 days in transit and 240 days performing active operations. Between missions, an estimated 71 days will be spent in port for upkeep and repair. The nominal SURTASS LFA Sonar annual and 49-day deployment schedule for a single ship can be seen in Table 2-1 of the Final SEIS.

The two existing operational LFA sonar systems are installed on the SURTASS vessels:

R/V Cory Chouest

and USNS IMPECCABLE (T-AGOS 23). To meet future undersea warfare requirements, the Navy is working to develop and introduce a compact active system deployable from existing, smaller SURTASS Swath-P ships. This smaller system is known as Compact

LFA, or CLFA sonar. CLFA sonar consists of smaller, lighter-weight source elements than the current LFA sonar system, and will be compact enough to be installed on the existing SURTASS platforms, VICTORIOUS Class (T-AGOS 19) vessels. The Navy indicates that the operational characteristics of the compact system are comparable to the existing LFA sonar systems as presented in Subchapter 2.1 of the Final EIS and Final SEIS. Consequently, the potential impacts from CLFA sonar will be similar to the effects from the existing SURTASS LFA sonar systems. Three CLFA sonar systems are planned for installation on T-AGOS 20, 21, and 22. With the

R/V Cory Chouest

retiring in FY 2008, the Navy estimates that there will be two systems in operation in FY 2008 and FY 2009, 3 in FY 2010 and 4 systems in FY 2011 and FY 2012. At no point are there expected to be more than four systems in use, and thus this rule analyzes the impacts on marine mammals due to the deployment of up to three LFA sonar systems through FY 2010 and four systems in FY 2011 and FY 2012.

The SURTASS LFA sonar vessel will operate independently of, or in conjunction with, other naval air, surface or submarine assets. The vessel will generally travel in straight lines or racetrack patterns depending on the operational scenario.

Description of Acoustic Propagation

The following is a very basic and generic description of the propagation of LFA sonar signals in the ocean and is provided to facilitate understanding of this action. However, because the actual physics governing the propagation of SURTASS LFA sound signals is extremely complex and dependent on numerous in-situ environmental factors, the following is for illustrative purposes only.

In actual SURTASS LFA sonar operations, the crew of the SURTASS LFA sonar platform will measure oceanic conditions (such as sea water temperature and salinity versus depth) prior to and during transmissions and at least every 12 hours, but more frequently when meteorological or oceanographic conditions change. These technicians will then use U.S. Navy sonar propagation models to predict and/or update sound propagation characteristics. The short time periods between actual environmental observations and the subsequent model runs further enhance the accuracy of these predictions. Fundamentally, these models are used to determine what path the LF signal will take as it travels through the ocean and how strong the sound signal will be at given ranges along a particular transmission path.

Accurately determining the speed at which sound travels through the water is critical to predicting the path that sound will take. The speed of sound in seawater varies directly with depth, temperature, and salinity. Thus, an increase in depth or temperature or, to a lesser degree, salinity, will increase the speed of sound in seawater. However, the oceans are not homogeneous, and the contribution of each of these individual factors is extremely complex and interrelated. The physical characteristics that determine sound speed change with depth, and in the case of temperature and salinity, season, geographic location, and locally, with time of day. After accurately measuring these factors, mathematical formulas or models can be used to generate a plot of sound speed versus water depth. This type of plot is generally referred to as a sound speed profile (SSP).

Near the surface (variable within the top 1000 ft (305 m)), ocean near-surface water mixing results in a fairly constant temperature and salinity. Below the mixed layer, sea temperature drops rapidly in an area referred to as the thermocline. In this region, temperature influences the SSP, and speed decreases with depth because of the large decrease in temperature (sound speed decreases with decreasing temperature). Finally, beneath the thermocline, the temperature becomes fairly uniform and increasing pressure causes the SSP to increase with depth.

One way to envision sound traveling through the sea is to think of the sound as “rays.” As these rays travel through the sea, their direction of travel changes as a result of speed changes, bending, or refracting, toward areas of lower speed and away from areas of higher speed. Depending on environmental conditions, refraction can either be toward or away from the surface. Additionally, the rays can be reflected or absorbed when they encounter the surface or the bottom. For example, under certain environmental conditions, near-surface sound rays can repeatedly be refracted upward and reflected off the surface and thus become trapped in a duct.

Some of the more prevalent acoustic propagation paths in the ocean include: acoustic ducting; convergence zone (CZ); bottom interaction; and shallow-water propagation.

Acoustic Ducting

There are two types of acoustic ducting: surface ducts and sound channels.

Surface Ducts

As previously discussed, the top layer of the ocean is normally well mixed and has relatively constant temperature and salinity. Because of the effect of depth (pressure), surface layers exhibit a slightly positive sound speed gradient (that is, sound speed increases with depth). Thus, sound transmitted within this layer is refracted upward toward the surface. If sufficient energy is subsequently reflected downward from the surface, the sound can become “trapped” by a series of repeated upward refractions and downward reflections. Under these conditions, a surface duct, or surface channel, is said to exist. Sound trapped in a surface duct can travel for relatively long distances with its maximum range of propagation dependent on the specifics of the SSP, the frequency of the sound (e.g., there is a low-frequency cutoff dependent on the thickness of the duct), and the reflective characteristics of the surface. As a general rule, surface duct propagation will improve as the temperature uniformity and depth of the layer increase. For example, transmission is improved when cloudy, windy conditions create a well-mixed surface layer or in high-latitude midwinter conditions where the mixed layer extends to several hundred feet deep.

Sound Channels

Variation of sound speed, or velocity, with depth causes sound to travel in curved paths. A sound channel is a region in the water column where sound speed first decreases with depth to a minimum value, and then increases. Above the depth of minimum value, sound is refracted downward; below the depth of minimum value, sound is refracted upward. Thus, much of the sound starting in the channel is trapped, and any sound entering the channel from outside its boundaries is also trapped. This mode of propagation is called sound channel propagation. This propagation mode experiences the least transmission loss along the path, thus resulting in long-range transmission.

At low and middle latitudes, the deep sound channel axis varies from 1,970 to 3,940 ft (600 to 1,200 m) below the surface. It is deepest in the subtropics and comes to the surface in the high latitudes, where sound propagates in the surface layer. Because propagating sound waves do not interact with either the sea surface or seafloor, sound

propagation in sound channels does not attenuate as rapidly as bottom- or surface-interacting paths. The most common sound channels used by SURTASS LFA sonar are convergence zones (CZs).

Convergence Zones

CZs are special cases of the sound-channel effect. When the surface layer is narrow or when sound rays are refracted downward, regions are created at or near the ocean surface where sound rays are focused, resulting in elevated sound levels. The existence of CZs depends on the SSP and the depth of the water. Due to downward refraction at shorter ranges, sound rays leaving the near-surface region are refracted back to the surface because of the positive sound speed gradient produced by the greater pressure at deep ocean depths. These deep-refracted rays often become concentrated at or near the surface at some distance from the sound source through the combined effects of downward and upward refraction, thus causing a CZ. CZs may exist whenever the sound speed at the ocean bottom, or at a specific depth, exceeds the sound speed at the source depth. Depth excess, also called sound speed excess, is the difference between the bottom depth and the limiting, or critical depth.

CZs vary in range from approximately 18 to 36 nautical miles (nm) (33 to 67 km), depending upon the SSP. The width of the CZ is a result of complex interrelationships and cannot be correlated with any specific factor. In practice, however, the width of the CZ is usually on the order of 5 to 10 percent of the range. For optimum tactical performance, CZ propagation of SURTASS LFA sonar signals is desired and expected in deep open ocean conditions.

Bottom Interaction

Reflections from the ocean bottom and refraction within the bottom can extend propagation ranges. For mid- to high-level frequency sonars (greater than 1,000 Hz), only minimal energy enters into the bottom; thus reflection is the predominant mechanism for energy return. However, at low frequencies, such as those used by the SURTASS LFA sonar source, significant sound energy can penetrate the ocean floor, and refraction within the seafloor, not reflection, dominates the energy return. Regardless of the actual transmission mode (reflection from the bottom or refraction within the bottom), this interaction is generally referred to as “bottom-bounce” transmission.

Major factors affecting bottom-bounce transmission include the sound frequency, water depth, angle of incidence, bottom composition (e.g., sediments), and bottom roughness. A flat ocean bottom produces the greatest accuracy in estimating range and bearing in the bottom-bounce mode.

For SURTASS LFA sonar transmissions between 100 and 500 Hz, bottom interaction would generally occur in areas of the ocean where depths are between approximately 200 m (660 ft) (average minimum water depth for SURTASS LFA sonar deployment) and 2,000 m (6,600 ft).

Shallow Water Propagation

In shallow water, propagation is usually characterized by multiple reflection paths off the sea floor and sea surface. Thus, most of the water column tends to become ensonified by these overlapping reflection paths. As LFA sonar signals approach the shoreline, they will be affected by shoaling, experiencing high transmission losses through bottom and surface interactions. Therefore, LFA sonar would be less effective in shallow, coastal waters.

In summary, for the SURTASS LFA sonar signal in low- and mid-latitudes, the dominant propagation paths for LFA sonar signals are CZ and bottom interaction (at depths less than 2000 m (6,600 ft)). In high-latitudes, surface ducting provides the best propagation. In most open ocean water, CZ propagation will be most prominent. The SURTASS LFA sonar signals will interact with the bottom, but due to high bottom and surface losses, SURTASS LFA sonar signals will not penetrate coastal waters with appreciable signal strengths.

Comments and Responses

On September 28, 2006 (71 FR 56965), NMFS published a Notice of Receipt of Application on the U.S. Navy SURTASS LFA sonar MMPA application and invited interested persons to submit comments, information, and suggestions concerning the application and the structure and contents of regulations. Those comments were considered in the development of the proposed rule. A proposed rule for renewal of the regulations governing SURTASS LFA sonar MMPA authorization was published on July 9, 2007 (72 FR 37404) with a 15-day public comment period. During the two comment periods, comments were received from a large number of organizations and individuals. Those organizations include the Marine Mammal Commission (Commission), the Natural Resources Defense Council (NRDC), Earth Island Institute (EII), Acoustic Ecology Institute (AEI), Animal Welfare Society (AWI), Cetacean Society Institute (CSI), Seaflow, International Ocean Noise Coalition, Olympic Coast Alliance, Citizens Opposing Active Sonar Threats, Ocean Care, Gesselschaft zur Rettung der Delphine, SBOOHER, Ocean Conservation Research, Friends of the San Juans, World Society for the Protection of Animals. We have addressed all comments on the proposed rule. We also responded to comments that appear to be directed solely at the draft SEIS, although we did not address comments strictly related to non-marine mammal issues. See the Navy's Final SEIS, which NMFS has adopted under NEPA.

Activity Concerns

Comment 1:

The U.S. Navy seeks a blanket exemption to do harm to all marine animals in 80 percent of the world oceans with only minor mitigation measures taken. Expanding the SURTASS program into 80 percent of the world's oceans would make the task of monitoring the impacts impossible. An LOA granted would not meet the “negligible impact” condition and would violate the “unmitigable adverse impact” constraints indicated in the MMPA LOA process.

Response:

The Navy is not seeking a “blanket exemption” from the MMPA, but rather is requesting that NMFS issue regulations to govern the incidental take of marine mammals under Section 101(a)(5)(A) of the MMPA. Under these regulations the Navy must apply annually for a letter of authorization (LOA) that would exempt the taking of marine mammals incidental to the Navy's use of SURTASS LFA sonar from the MMPA's general moratorium on the taking of marine mammals for that year, as long as the sonar use was consistent with these regulations and the terms of the LOA. In its LOA application, the Navy must specify where it will operate SURTASS LFA sonar for that year and take authorization would be limited to that area. Under the regulations, the total area that would be available for SURTASS LFA sonar operations over the five-year period is about 70-75 percent of the world's oceans. This in no way equates to LFA sonar operations affecting even close to 70-75 percent of the world's ocean area at any given time. Each year, based on its projected operational needs, the Navy will identify for which particular geographic areas, out of the total available area, it is requesting take authorization through an LOA. The first authorization is for only two SURTASS LFA sonar vessels both operating in the Western Pacific Ocean. Eventually, the Navy plans to have 4 vessels in operation, but even if

all 4 vessels operated in 4 different oceans, the area ensonified would come nowhere close to 70-75 percent of the world's ocean area. Therefore, SURTASS LFA sonar sound will not simultaneously affect 70-75 percent of the world's oceans. In addition, NMFS has determined that incidental harassment takings by SURTASS LFA sonar operations during the effective time period (1 year) of any LOA issued to the Navy pursuant to these regulations must not exceed 12 percent of any marine mammal stock.

The sound pressure level (SPL) that is capable of potentially causing injury to an animal is within approximately 1 km (0.54 nm) of the ship. For the purposes of analyses using the Acoustic Integration Model (AIM) and the risk continuum, there is a 50 percent risk of significant change in a biologically important behavior for a marine mammal exposed to a received level (RL) of 165 dB RMS. The range from the SURTASS LFA sonar vessel for this received level, which could cause behavioral disruption but not injury, could extend to 25 to 65 km (13.5 to 35.1 nm). The received level at the surface along any straight path away from the ship would not decline logarithmically over distance, as would be expected if the sound spread by spherical spreading alone. The reason is that, for CZ propagation, the sound moves in an undulating path with turning points near the surface and near the bottom, where sound is refracted either downward (near surface) or upward (near bottom). Turning points near the surface, termed caustics, occur approximately every 30 nm (56 km). The received level at the surface would be high at the caustics but low in between them because most of the sound energy there would be found at great depth. While the regulations permit the Navy to seek authorization through an LOA to take marine mammals while operating SURTASS LFA sonar in many of the world's oceans and SURTASS LFA sonar signal can be detected at several hundred miles using sophisticated listening gear, SURTASS LFA sonar's potential to cause injury or affect behavior is limited to relatively close to the ship. Thus, the impact of SURTASS LFA sonar is not global in scope. Moreover, monitoring to ensure that marine mammals are not injured is not impossible, as the commenter suggests, given the limited area around the vessel that is ensonified at decibel levels up to 180 dB, and the demonstrated effectiveness of the Navy's tripartite (visual, acoustic, and HF/M3) monitoring scheme.

Since the SURTASS LFA sonar will not operate in Arctic waters, there will not be an unmitigable adverse impact on relevant subsistence uses of marine mammals. That determination is provided later in this document. NMFS also believes the negligible impact standard has been met, as described in this final rule.

Comment 2:

The Navy is proposing to expand the use of LFA sonar, both through expansion of use areas geographically throughout the world's oceans and through doubling the number of LFA sonar array ships. The Navy is also admitting to the use of CLFA sonar in “shallow littoral ocean regions” and do not discuss the characteristics of CLFA sonar in the Final SEIS.

Response:

While the number of SURTASS LFA sonar vessels will increase from 2 to 4 vessels over the course of the five-year rule, the Navy is not increasing the number of SURTASS LFA sonar systems beyond what was analyzed in the January 2001 Final EIS. That document analyzed the potential impacts of up to four SURTASS LFA sonar systems. As stated in the Navy's Record of Decision (ROD) (67 FR 48145, July 23, 2002), the Navy determined that only two of the four systems would be operational during the timeframe of the 2002-2007 regulations governing the taking of marine mammals incidental to LFA sonar testing and training. For that reason, NMFS addressed taking marine mammals incidental to operation of only two systems under the initial five year Final Rule in 2002. Installation and deployment of the third and fourth LFA sonar systems were postponed until after FY 2007. Because of this delay, the decision in the Navy Record of Decision (ROD) and NMFS' MMPA determinations covered the employment of only two SURTASS LFA sonar systems. Therefore, the use of SURTASS LFA sonar, analyzed here, does not exceed the originally analyzed four systems during the timeframe of the requested second five year set of MMPA regulations.

In addition, the Navy's proposal to deploy SURTASS LFA sonar in a number of oceans is not new. The Navy's Final EIS proposed, and NMFS original Final Rule and regulations addressed, deployment of SURTASS LFA sonar throughout most of the world's oceans. As stated in the Final SEIS, these systems will be employed as required for security operations in the oceanic areas as presented in Figure 1-1 of the Final EIS. Potential operations could occur in the Pacific, Atlantic, and Indian Oceans, and the Mediterranean Sea. Large oceanic areas are restricted from operations, including the Arctic and Antarctic Ocean areas, as are all offshore areas within 12 nm (22 km) of land, and OBIAs (Table 2-4 of the SEIS). The limitation of SURTASS LFA operation to the Western Pacific Ocean was a product of the parties' negotiations over the Stipulated Permanent Injunction.

Nevertheless, while the number of systems may increase under this Final Rule and the Navy may seek authorization to use SURTASS LFA sonar in more places than it could under the terms of the permanent injunction, the maximum permissible impact to any particular species or stock remains the same, since the Navy's overall use of SURTASS LFA sonar can have no more than a negligible impact on marine mammal species and stocks. Consistent with its findings in the original rule, NMFS has determined that takings by SURTASS LFA sonar operations during the effective time period (1 year) of any LOA issued to the Navy pursuant to these regulations must not exceed 12 percent of any marine mammal stock.

As stated in the Final SEIS Subchapter 1.2.3 and 2.1, compact LFA sonar (CLFA sonar) sonar is an upgrade and modification to the SURTASS LFA sonar system necessary to install and operate on the smaller VICTORIOUS Class T-AGOS 19 Class ocean surveillance ships. The operational characteristics of the active system components installed, or to be installed, on the

R/V Cory Chouest,

USNS IMPECCABLE, and VICTORIOUS Class vessels are provided in Final SEIS Subchapter 2.1.1. The characteristics of LFA sonar and the upgrade and modifications for the T-AGOS 19 installations are essentially the same. The frequency requirements for the CLFA to be installed onboard the VICTORIOUS Class (T-AGOS 19 Class) vessels are within the 100 to 500 Hz range for LFA sonar and the transmit array also consists of 18 transducers with a similar source level.

Subchapter 1.1.3 of the Final SEIS provides a definition of the term “littoral” as used by the U.S. Navy and explains the ways in which the use of the term as a tactical designation differs from its use as a geographic term. The littoral operating environment does not necessarily include or exclude any waters because of depth; it can include both deep and shallow water. However, under any of the alternatives analyzed in the Final SEIS, LFA sonar would not operate inside of 12 nm (22 km) from any coastline. The use of SURTASS LFA sonar in coastal environments was discussed in Response to Comments (RTCs) 1-1.4 and 3-2.8 in the Final EIS.

Comment 3:

With regard to noise-producing activities, NMFS must describe source levels, frequency ranges, duty cycles, and other technical parameters relevant to determining the potential impacts of an MMPA authorization.

Response:

The NMFS action is the issuance of regulations and LOAs to the Navy for taking marine mammals incidental to SURTASS LFA sonar operations and determining whether SURTASS LFA sonar is having a negligible impact on affected marine mammal species and stocks, not whether LFA sonar operations and other noise producing activities are having a negligible impact on affected species and stocks of marine mammals (and species/stocks not affected by LFA sonar, but potentially by other noise-producing activities). In that regard, all technical parameters relevant to the impact analysis, including those listed by the commenter, were provided in the project descriptions for SURTASS LFA sonar in both the Final EIS (DON, 2001) Subchapters 2.1.1 and 2.3.2.2 and in RTCs 2-1.1 and 2-1.2a; and in the Final SEIS Subchapter 2.1.1.

Comment 4:

There are at least five Navy SWATH vessels already built and outfitted with operational LFA sonars.

Response:

Four VICTORIOUS class Ocean Surveillance ships were built between 1991 and 1993. As stated in the SEIS Subchapter 2.1, there are no LFA sonar systems deployed on these vessels at this time. The projected LFA sonar/CLFA sonar system availabilities are shown in the Final SEIS Figure 2-2, which includes future installations onboard the VICTORIOUS Class vessels.

Comment 5:

It is only a matter of time before many other industrialized nations follow suit and the oceans become a cacophony of LFA sonar systems using loud noise to try and find each other in an increasingly loud environment. The U.S. should re-examine this “need” and come up with a better way to find these quiet submarines.

Response:

This comment is beyond the scope of this rulemaking. As explained in the Final EIS, subchapter 1.2.1, the Navy has considered other alternatives and determined that SURTASS LFA sonar best addresses its need for reliable long-range detection of potentially hostile quiet submarines.

Comment 6:

At peak power, the Navy's LFA sonar system sends out pulses of sound underwater at least the equivalent of standing five feet away from the Saturn rocket on liftoff.

Response:

While an accurate source level of the Saturn V is not known, the comparison of this, or any other rocket, to LFA sonar is inappropriate. The sound generated by a Saturn V rocket, or any rocket in general, is broadband and generates a different frequency spectrum than that of LFA sonar, and travels in a significantly different transmission pattern. The Saturn C 1 rocket (a predecessor to the Saturn I rocket, which had about 1,600,000 lbs of thrust) was projected to have produced acoustic levels as high as 205 dB (in air) from a distance of 305 meters. Some sources suggest that the sound levels produced by the Saturn V (during the launch of Apollo 15, the first stage of the Saturn V generated 7,823,000 lbs of liftoff thrust) may have been as high as 220 dB (in air) (Benson and Faherty, 1978). As sound is perceived differently underwater than it is in air, sound propagation and transmission losses in each case are subject to differing factors, including terrain, wind, and air temperature, and in the case of LFA, water salinity, temperature and depth. Furthermore, sound levels are typically provided with a reference level, which depends on whether the sound is in air (reference of 20 microPascals) or water (reference of 1 microPascal). Despite it being inappropriate to compare a sound level in air with that in water (or vice versa), some simplified conversion or correction factors are available to provide a very generic comparison. Therefore, when corrected to the equivalent sound levels in water (based on pressure and impedance differences of the two media), the above acoustic levels of 205 dB in air and 220 dB in air would be approximately 266.5 and 281.5 dB in water, respectively (Please see Final EIS Appendix B, Subchapter B.3.2). These sound levels are 100 to 10,000 times louder than the LFA sonar source.

Comment 7:

NMFS should require that the U.S. Navy avoid or eliminate triangulation of sonar whether they are doing exercises with other U.S. Navy ships or with those from other nations.

Response:

Triangulation is only necessary for passive acoustics. Triangulation is not necessary for active acoustics because it gives the operator range and bearing. However, the focus of the comment seems to be on the use of multiple LFA sonar ships, which is discussed in the Final SEIS, (Subchapters 4.4.4 and 4.6.1.2) and in the Final EIS (Subchapter 4.2.7.4). The Final EIS states that the vast majority of operations will involve only one ship. This is due to the limited number of ships of SURTASS LFA sonar systems planned to be built and the limited operational conditions that could warrant the use of two sources in proximity to each other. The remote possibility exists that operational requirements or training exercises could require two sources simultaneously in one geographic region, for example the Northwest Pacific Ocean where LFA sonar vessels have been operating. The effect of the presence of two sources transmitting in one area can be conservatively approximated by doubling the single source potential effects provided for that site. An example of these effects can be seen in Table 4-2.13 of the Final EIS. However, even if more than one source operates in a single geographic area, impacts to marine mammals remain capped by the negligible impact requirement. To ensure that SURTASS LFA sonar operations have no more than a negligible impact over five years, not more than 12 percent of any marine mammal stock may be taken, by harassment, in a single year, regardless of how many SURTASS LFA sonar sources are operating in the area.

Comment 8:

There are plenty of safe alternatives to active sonar that the Navy could pursue, such as passive sonar, non-acoustic sensors, and Integrated Sensory Networks.

Response:

The comment is beyond the scope of NMFS' rulemaking for this action. Non-acoustic alternative underwater detection technologies are discussed in the Final EIS, Subchapter 1.2.1.

MMPA Concerns

Comment 9:

NMFS should consolidate all necessary and relevant information from the multiple existing sources of information describing the proposed actions in the proposed rule.

Response:

NMFS does not consider it necessary to consolidate all necessary and relevant information on LFA sonar and its impacts on marine mammals into the proposed and/or final rules. In the proposed and final rules, NMFS has continued and updated the information contained in the preamble to the 2002 final rule. NMFS believes that this information provides the necessary level of detail needed for it to make the determinations required under the MMPA and for the public to review this information. This document also reflects the findings of the Final EIS, with the data and findings of the Final SEIS. These documents and others, which are available on the Navy SURTASS LFA sonar homepage (see ADDRESSES) provide the “consolidated information” that the commenter requested.

Comment 10:

The Commission states that any regulations proposing to issue an incidental taking authorization should include information on specified geographic locations where sonar is

expected to be deployed and the species and number of marine mammals that may be taken in each of those locations.

Response:

While the NDAA removed references to the specified geographical region and small numbers requirements for military readiness activities, NMFS still needs to know where activities would take place and the estimated level of take to inform its negligible impact determination. In order to do so, NMFS considered “worst-case” estimates for purposes of the negligible impact determination as well as an annual 12 percent per-stock “cap” for marine mammals regardless of where and when LFA sonar will be operating (or even how many LFA sonar systems are in operation annually). This rulemaking also considered the oceans and areas where LFA sonar may and may not operate. The rule does not specify the specific location where LFA sonar will be deployed and the number of marine mammals that may be taken in those locations because these are determined annually through various inputs such as mission duration and season of operation [which are calculated in the annual applications for LOAs].

Comment 11:

The Commission recommends the existing annual review process for LOAs should be expanded to include public review and comment. The NRDC believes issuance of LOAs without notice and comment violates MMPA section 101(a)(5)(A) because, it says, each year's authorization will involve new take and negligible impact analyses and potentially new exercise areas that are not modeled in the Navy's SEIS.

Response:

NMFS does not agree. Under section 101(a)(5)(A), notice and opportunity for public comment must be afforded before the Secretary authorizes the incidental take of marine mammals, makes a negligible impact determination, and issues the required regulations. NMFS published the proposed regulations on July 9, 2007 (72 FR 37404), providing the required notice and opportunity for public comment. That proposed rule contained NMFS' negligible impact determination for the five-year period and proposed mitigation, monitoring, and reporting requirements. It also considered the Navy's estimates of take for the five-year rule period. Section 101(a)(5)(A) of the MMPA does not require the regulations to specify the number of marine mammals that may be taken, only the permissible methods of taking and means of effecting the least practicable adverse impact.

As stated in the proposed rule and the Navy's Final EIS, estimates were derived based on modeling sites, since it was not practical to model all areas where the system might be operated. Final EIS p. 4.2-1. These sites represented the upper bound of impacts expected from operation of SURTASS LFA sonar. Final EIS p. 4.2-3; see Final EIS tables 4.2-1, 4.2-4, 4.2-10, 4.2-11, and 4.2-12. If LFA sonar operations occur in a non-modeled area, the take estimates would most likely be less than those obtained from the most similar site that was modeled. Final EIS p. 4.2-3. As stated in the SEIS, the assumptions of the Final EIS are still valid and have been incorporated by reference into the SEIS p. 4-39, 40. Moreover there are no new data that contradict the assumptions or conclusions made in subchapter 4.2 of the FEIS. Thus, it was not necessary to reanalyze potential acoustic impacts in the SEIS.

The risk assessment for each planned mission site for each vessel is performed annually and is part of the Navy's annual mission intention (LOA application) letter. In its annual LOA applications, the Navy must project where it intends to operate during the period of the annual LOAs and provide NMFS with reasonable and realistic risk estimates of the marine mammal stocks in the proposed areas of operations. This process utilizes the best available data and is detailed in the SEIS including a case study. SEIS pp. 4-37 to 4-51. During the initial steps of the risk analysis process, if the take estimates exceed those required under the regulations (including the annual 12 percent per-stock cap), than the mission areas are changed or refined and the analysis is reinitiated. After receipt of an LOA application, NMFS reviews the activity (and previous annual reports) to ensure it remains within the parameters of the rule and the negligible impact assessment.

NMFS' general implementing regulations for section 101(a)(5)(A) of the MMPA, which have been in effect since 1982 and which governed the last rulemaking for SURTASS LFA sonar incidental take, set up the framework under which NMFS issues LOAs that an applicant must obtain before any incidental take is authorized. 50 CFR 216.106(a). The purpose of the requirement for obtaining LOAs is to ensure the authorized taking will be consistent with the original findings. See 47 FR 21248, 21251 (May 18, 1982). Therefore, issuance of an LOA is based on a determination that the level of taking will be consistent with the findings made for the total taking allowable under the specific regulations for the specified activity. 50 CFR 216.106(b). The reporting requirements under these specific SURTASS LFA sonar regulations and LOAs require the Navy to provide both quarterly and annual reports to NMFS. In these reports, the Navy must provide estimated percentages of marine mammal species/stocks potentially affected for each quarter and annually. NMFS' general implementing regulations do not require the agency to provide notice and comment for LOAs. However, if NMFS were to obtain information that calls into question the validity of its determinations in this rule, the agency could withdraw or suspend authorization to take marine mammals if the Secretary, through the Assistant Administrator for Fisheries, finds, after notice and opportunity for public comment, that the regulations are not being substantially complied with, or the taking allowed pursuant to the regulations is having or may have more than a negligible impact on marine mammal species or stocks. 50 CFR 216.106(e). The requirement for notice and comment does not apply if an emergency exists that poses a significant risk to the wellbeing of the species or stocks of marine mammals concerned. 50 CFR 216.106(f).

Comment 12:

The Commission states that NMFS should address the requirement of the NDAA that personnel safety, practicality of implementation, and impact on the effectiveness of the military readiness activity be considered in making a “least practicable adverse impact” determination in the proposed rulemaking.

Response:

NMFS agrees with the Commission and added a discussion of the NDAA in the proposed and final rules.

Comment 13:

The NRDC states the Navy fails to present evidence of negligible impact. Agencies must make every attempt to obtain and disclose data necessary to their analysis. This is important when the program's impacts depend on newly emerging data. The Navy fails to take account of significant new information that has emerged since January 2001 concerning marine mammal thresholds of injury, hearing loss, and significant behavioral change.

Response:

NMFS believes the MMPA requires a determination of negligible impact to be based on the best available data. NMFS believes the best available data were used in the Final SEIS, NMFS' 2002 final rule, the Navy 2006 MMPA application and this final rule, to estimate the potential impacts on the environment. Information that the commenter (and others) believe

contradict this determination by NMFS is addressed throughout this document.

Comment 14:

A number of commenters were of the opinion that a 15-day comment period for the proposed rule is too short to review the material and not in compliance with the Administrative Procedure Act (APA).

Response:

The 15-day comment period on the proposed rule provided an adquate opportunity for public comment. In addition to the comment period on the proposed rule, members of the public had a 30-day public comment period on the Navy's application for renewal of NMFS' regulations (71 FR 56965, September 28, 2006) and a 92-day public comment period (including three public hearings) for the Navy's Draft SEIS on SURTASS LFA sonar (which contains much of the underlying analysis for this proposed rule, affording significant opportunity for public participation). In addition, the proposed rule is substantially similar to the 2002-2007 rule, which underwent a 75-day public comment period, including public hearings in Los Angeles, CA, Honolulu, HI, and Silver Spring, MD. There have been no significant scientific advancements or other developments since the previous rule that would necessitate a longer period for public comment.

Comment 15:

It is well-established that mid-frequency (MF) sonar negatively impacts marine mammals, even resulting in fatalities, with the U.S. Navy having admitted direct responsibility for past beachings. The effects of LF sonar appear to be less understood at this time, but the enormous range of ocean impacted by sonar makes it incumbent upon us to fully understand its effects before authorizing its widescale use. The Precautionary Principle should be applied before issuing a permit.

Response:

NMFS used conservative assumptions for identifying and analyzing potential impacts to the environment, including marine mammals. SURTASS LFA sonar has been operating under NMFS regulations for the last five years without any reports of Level A harassment. The evidence to date, including recent scientific reports, supports the conclusion that operation of the U.S. Navy's LFA sonar does not result in marine mammal strandings. For further information on strandings and MF sonar, please see comments 8, 32, 33, 47, and 49 for further analyses on strandings.

Comment 16:

I request a moratorium on any use of this technology in the oceans, at the levels currently used, until further tests are conducted on the foundational species in the food chain of the marine environment.

Response:

Research using LFA sonar technology has been conducted on several species in the food chain, including whales (blue, fin, grey, and humpback whales) and on fish (catfish, a hearing specialist, and trout; reference species for salmon and a hearing generalist). This research is discussed later in this document (see Research Concerns). NMFS believes the data are sufficient to go forward, recognizing that more research would be valuable.

Marine Mammal Impact Concerns

Comment 17:

The NRDC states that the Navy sets its threshold for hearing loss or “threshold shift” at 180 dB re: 1 microPa (RMS) for a single 100-second “ping” of exposure. The analysis is based on data from humans and other terrestrial mammals and relies on a limited set of data on marine mammals. The Navy has established a sliding scale for behavioral impacts. The Final SEIS fails to incorporate several recent studies on the effects of low-frequency sound on various marine mammal species. Also, the Navy's standard fails to take proper account of chronic impacts, from behavioral changes as well as from certain non-auditory physiological impacts such as stress. The Final SEIS and MMPA application disregard recent evidence indicated the potential for masking to interfere with long-distance mating behavior in mysticetes. The Navy standard is out of step with how the potential for behavioral impacts has been assessed in other contexts. Last, the Navy does not consider the impact that behavioral changes in species such as fish may have on marine mammals foraging.

Response:

NMFS believes that the latest information on impacts of underwater sounds on marine mammals and fish is contained in the Navy's Draft and Final SEIS, and summarized in the Navy's application. NMFS addresses the masking issue in comment 19 and elsewhere in this document.

As stated in the Final EIS, the 180-dB criterion for the purpose of SURTASS LFA sonar analysis is that all marine animals exposed to received levels (RLs) greater than 180-dB rms are evaluated as if they are injured. In its 2002 Final Rule for SURTASS LFA sonar, NMFS stated that temporary threshold shift (TTS) is not an injury. Since the boundary line between TTS and permanent threshold shift (PTS) is neither clear, definitive, nor predictable for marine mammals, NMFS has adopted (as a conservative estimate) 20 dB of TTS to define the onset of PTS (i.e., a temporary shift of 20 dB in hearing threshold) (67 FR 46711, July 16, 2002). As noted in Schlundt

et al.

(2000), bottlenose dolphins and belugas exposed to 1-sec signals at 400 Hz did not exhibit TTS after exposures to maximum RLs of 193-dB sound exposure level (SEL)) (which would be equivalent to a received level of 193 dB re: 1 microPascal (RMS) since the duration is 1-sec). The point must be made that while dolphins and belugas responses at 400 Hz are valid for those species, these results probably do not generalize to large whales (e.g., baleen whales).

In the Schlundt

et al.

(2000) research, dolphins and belugas did not have TTS in response to 400 Hz at RLs of 193 dB SEL, but they did have TTS in response to higher frequencies (where they are more sensitive) at the same level. It is reasonable to assume that the TTS threshold value from odontocetes at their frequency of highest sensitivity is applicable to larger animals and lower frequencies that are in the range of their best hearing sensitivity. This extrapolation is based on the fundamental similarity of cochlear structure between odontocetes and mysticetes. As a result, if it were assumed that 193 dB SEL was the onset of TTS (a conservative assumption because TTS was not observed at an RL of 193 dB SEL), then onset of PTS would be 20 dB above that, at 213 dB RL (SEL). This number is based on a signal of one second in duration. Using a 10 Log (T/Ti) where Ti is 1 second, then for a maximum 100-sec LFA sonar signal, a 20-dB adjustment must be made, meaning that the onset of PTS would be 193 dB RL (SEL). This value is above the conservative LFA sonar criterion of 180 dB for injury. A more detailed discussion is provided in the Final EIS RTCs 4-6.13 and 4-6.38 and the 2002 Final Rule RTCs MMIC8, MMIC9, SIC40, SIC58, and SIC59.

In addition, recent data on critical ratios (CRs) in pinnipeds is discussed in the Final SEIS Subchapter 4.3.5. A CR is the difference between sound level for a barely audible tone and the spectrum level of background noise at nearby frequencies (Richardson

et al.

, 1995). These data indicate that the CRs for pinnipeds are lower in magnitude than for terrestrial animals (Southall

et al.

2003). Southall

et al.

(2003), in describing their CR results, state that “It is reasonable to speculate that acoustic signal production and reception in typically noisy marine environments have led to selection for enhanced ability to detect signals in noise.” Therefore these new CR data indicate that pinnipeds may be pre-adapted for detecting biologically important signals in high noise environments.

Furthermore, the lower critical bandwidths of the pinniped auditory filters has the effect of decreasing the probability of masking of signals by noise at a different frequency (Southall

et al.

, 2000). Nevertheless, NMFS believes pinnipeds remain as susceptible as any species to masking of signals by noise in the same frequency band.

The Final SEIS also considered recent studies on LF sound and injury. In regard to injury, the issue of resonance is addressed in the Final SEIS (RTC 2.5.2). The analysis by the Navy (Cudahy and Ellison, 2002), reports on two workshops on acoustic impacts (DOC, 2002: Cox,

et al.

2006), and the National Research Council (NRC) Ocean Studies Board (NRC, 2003) support the conclusion that resonance from LFA sonar operations is not a “reasonably foreseeable” impact. Cox

et al.

(2006) stated that gas-bubble disease, induced in supersaturated tissues by a behavioral response to acoustic exposure, is a plausible pathologic mechanism for the morbidity and mortality seen in cetaceans associated with MF sonar exposure. They also stated that it is premature to judge acoustically mediated bubble growth as a potential mechanism and recommended further studies to investigate the possibility.

The NRC Report (2003) discusses acoustically-induced stress in marine mammals. The NRC stated that sounds resulting from one-time exposure are less likely to have population-level effects than sounds that animals are exposed to repeatedly over extended periods of time. The NRC also cited controlled laboratory investigations of the response of cetaceans to noise that have shown cardiac responses (Miksis

et al.

, 2001 IN: NRC, 2003) but have not shown any evidence of physiological effects in the blood chemistry parameters measured. Beluga whales exposed for 30 minutes to 134-153 dB received level (RL) playbacks of noise with a synthesized spectrum matching that of a semisubmersible oil platform (Thomas

et al.

, 1990b IN: NRC, 2003) showed no short-term behavioral responses and no changes in standard blood chemistry parameters or in catecholamines. Preliminary results from exposure of a beluga whale and bottlenose dolphin to a seismic watergun with peak pressure of 226 dB source level (SL) showed no changes in catecholamines, neuroendocrine hormones, serum chemistries, lymphoid cell subsets, or immune function (Romano

et al.

, 2001 IN: NRC, 2003).

The NRC Report (2003) also stated that although techniques are being developed to identify indicators of stress in natural populations, determining the contribution of noise exposure to those stress indicators will be very difficult, but important, to pursue in the future when the techniques are fully refined. There are scientific data gaps regarding the potential for LFA sonar to cause stress in marine animals. Even though an animal's exposure to LFA sonar may be more than one time, the intermittent nature of the LFA sonar signal, its low duty cycle, and the fact that both the vessel and animal are moving, means that there is a very small chance that LFA sonar exposure for individual animals and stocks would be repeated over extended periods of time, such as those caused by shipping noise. There is sufficient information available to permit analysis and decision making. Therefore, impacts from stress are not a reasonably foreseeable significant adverse impact on marine mammals from exposure to LFA sonar.

In studying potential alerting stimuli for North Atlantic right whales, Nowacek

et al.

(2003) found that underwater sounds with an acoustic structure similar to their alert stimulus at RLs of 133-148 dB are likely to disrupt feeding behavior for the duration of the sound exposure, with return to normal behavior within minutes of when the sound was turned off. Their results are consistent with those of the LFS Scientific Research Program (SRP), which exposed baleen whales to RLs ranging from 120 to 155 dB, detecting only minor, short-term behavioral responses (please see Final EIS, Subchapter 4.2.4.3 for more information). The LFA sonar risk function is based on the LFS SRP results.

Concern that the LFA sonar signal may cause right whales to surface and thus be more vulnerable to ship strikes is not well founded because the vessels only move at about 5.6 km/hr (3 knots) (significantly lower than normal ship speeds) and LFA sonar mitigation measures will detect any large whales well before they enter the LFA sonar zone, at which time LFA sonar operations would be suspended.

Comment 18:

A number of incidents of whales becoming stranded and dying have occurred around the world linked with the use of very loud military sonars. To date, none of the many incidents involve LFA sonar, although (1) LFA sonar has not been used in close proximity to whale populations and (2) the Navy continues to deny that any military sonar impacts marine life. EII believes LFA sonar may have more lethal impact over longer distances due to the nature of low frequency sound transmission underwater. The Draft SEIS claims that the association between marine mammal stranding events and military sonar is an issue of “public perception” and specifically that “[a]lthough much of the public have the impression that military sonar usage is a principle cause of marine mammal strandings, the facts that are available indicate otherwise.” While this might be true for mass stranding events of a non-anthropogenic origin, it is a grossly misleading statement. The Navy ignores the scientific record.

Response:

Data indicate that the area in which LFA sonar has been operating (Northwestern Pacific Ocean) has relatively abundant populations of marine mammals, as presented in the SEIS as shown in Tables 4.4-2 to 4.4-10. During the LFS SRP in 1997 and 1998, LFA sonar sources were operated in proximity to marine mammals with only minor behavioral effects. As detailed in SEIS RTC 4.3.1 and later in this document, LFA sonar is not known to have caused any marine mammal strandings or injuries.

The “public perception” referred to in the Draft SEIS (p. 4-55) was one that views LFA sonar the same as any other sonar. The intent of the statement was that there is a public perception that the effects of LFA sonar are the same as any other naval, or loud, sonars. As noted in the discussion in the Final SEIS RTC 4.3.1, the potential for impacts from LFA sonar differs from that of mid-frequency active sonar. The best available scientific evidence to date does not indicate that LFA sonar has the potential to cause strandings based on analyses of existing strandings (ICES, 2005; Cox

et al.

, 2006). This paragraph was rewritten in the Final SEIS based on the latest available scientific data (see SEIS RTC 4.4.13).

Comment 19:

Given the relatively long duration of SURTASS LFA sonar “pings” masking may be more of an issue than it is with impulsive noise sources. While the average signal length is 60 seconds—which is a very long time—for an extremely loud noise each can be up to 90 seconds long and can occur as often as every six minutes. This also does not take into account reverberation which can significantly increase the duty cycles and could result in a near continuous signal. Even temporary masking can be significant as it can compromise an animal's ability to avoid predators, communicate, track and catch food, and avoid dangerous environments such as areas of high intensity noise.

Response:

The masking effects of the SURTASS LFA sonar signal are expected to be limited for a number of

reasons. First, the frequency range (bandwidth) of the system is limited to about 30 Hz, and the instantaneous bandwidth at any given time of the signal is small, on the order of 10 Hz. Second, the average duty cycle is always less than 20 percent and based on past LFA sonar operational parameters (2003 to 2007) is nominally 7.5 to 10 percent, as stated in Chapter 2 of the Final SEIS. Also, given the average maximum pulse length (60 seconds), and the fact that the signals vary and do not remain at a single frequency for more than 10 seconds, SURTASS LFA sonar is not likely to cause significant masking. An analysis of marine mammal hearing and masking are in Subchapter 4.6.1.2 of the Final SEIS. In other words, the LFA sonar transmissions are coherent, narrow bandwidth signals of 6 to 100 seconds in length followed by a quiet period of 6 to 15 minutes. Therefore, the effect of masking will be limited because animals that use this frequency range typically use broader bandwidth signals. As a result, the chances of an LFA sonar sound actually overlapping whale calls at levels that would interfere with their detection and recognition would be extremely low.

It is also unlikely that reverberation will significantly increase the duty cycles and result in a continuous signal. As a general rule, reverberation “dies off” or decreases with distance from the source as an exponent of time after sound transmission. However, this is not instantaneous and, depending on propagation and ocean boundary conditions, reverberation can linger in an area for seconds or minutes after a sound transmission, but at greatly reduced SPLs until it fades into background noise. In special cases (i.e., locations with the correct bathymetry, propagation conditions and signal repetition rates), the reverberation may not completely die off before the next transmission. Generally, however, the reverberation levels several seconds after transmission are so much less than the original signal, (i.e., approaching ambient noise levels) that they do not “add to the duty cycle.” LFA sonar signals have sufficient time to significantly decrease to levels much less than 120 dB in the vicinity of the source, prior to the transmission of the next signal. Additionally, reverberation away from the source's location starts at an even lower level than near the source and generally decreases faster than in proximity of the source, so it is always less than near the source (see Final SEIS comment 4.3.39).

Comment 20:

The Draft SEIS sets a threshold SPL of 145 dB for diving and recreational sites, which is an attempt to be precautionary to humans. This is over 1,000 times less intense than the threshold set for marine mammals. It is irrational to assume that marine mammals are less sensitive to sound in water than humans are. It would make far better sense to adopt a 145 dB as the threshold for all animals, including humans. Human exposure guidelines “were established based on psychological aversion testing,” exposure limits for cetaceans are based on avoiding only physiological injury (TTS) or the most dramatic behavioral responses. What basis justifies providing more protection to humans engaging in recreational diving than to native inhabitants of the sea?

Response:

These values represent different criteria: psychological aversion (a behavioral reaction) from direct measurements using human divers (Technical Report #3 of the Final EIS), and the exposure level at or above an RL of 180 dB, for which all marine mammals are evaluated as if they are injured (Final EIS Subchapter 1.4). However, humans are performing in a foreign medium compared to marine mammals. This suggests that the risk to marine mammals for a psychological response would be less than for humans. Furthermore, data cited in the Final EIS suggest that when operating in the presence of a biological imperative such as feeding, migrating or mating, such sound levels are insufficient to make the marine mammal discontinue their behavior (Technical Report #1 LFS SRP). Behavioral responses for marine mammals utilizing the risk continuum (see Final EIS Subchapter 4.2.3) demonstrate the potential for significant biologically important behavioral reactions from RLs from 120 to 179 dB, but with fewer significant behavioral responses at levels around 145 dB. Therefore, NMFS believes the 145-dB criterion for divers is consistent with the estimates of behavioral reactions to marine mammals, but at this time, it is unnecessary to consider this SPL as being warranted for marine mammals since the LFS SRP indicated that there were no significant behavioral reactions at these low levels and no indication that marine mammals might be seriously injured or killed by LFA sonar.

Comment 21:

The Draft SEIS minimizes impacts by emphasizing the small number of SURTASS LFA sonar systems to be employed and the narrow bandwidth of the active sonar signal. It is the intensity and pervasiveness of the SURTASS LFA sonar systems that is important in the discussion of impacts. The fact that there is more than one system merely compounds the problem. To declare that the low number and narrow bandwidth are mitigation measures is ludicrous.

Response:

Even though the source level of SURTASS LFA sonar is similar in intensity to many anthropogenic underwater sound sources, such as air gun arrays and other military sonars, there are significant differences in their operational characteristics. Table 1 illustrates these differences. Also, please see the Final SEIS RTC 4.3.1 for more information.

In a recent analysis for the Policy on Sound and Marine Mammals: An International Workshop sponsored by the Marine Mammal Commission (U.S.) and the Joint Nature Conservation Committee (UK) in 2004, Dr. John Hildebrand provided a comparison of anthropogenic underwater sound sources by their annual energy output. Dr. Hildebrand reported that the most energetic regularly operated sound sources are seismic air gun arrays from approximately 90 vessels with typically 12 to 48 individual guns per array, firing about every 10 seconds. There are approximately 11,000 super tankers worldwide, each operating 300 days per year, producing constant LF noise at source levels of 198 dB (SEL) (Hildebrand, 2005). Conversely, LFA sonar signals are transmitted for a maximum of 432 hours (18 days) per vessel per year. The signal length is between 6 to 100 seconds with 6 to 15 minutes between transmissions with individual elements source levels of 215 dB. Therefore, LFA sonar contributes less acoustic energy to the oceans than other sources. For more detailed discussions on Hildebrand's (2004) analysis, please see SEIS RTCs 4.6.4 and 4.6.5.

Even though LFA sonar signals are long range, LFA sonar cannot be considered to be pervasive (pervasive means to permeate or be present throughout) because of the nominal 7.5 to 10 percent duty cycle, meaning that during any given mission LFA sonar is not transmitting 90 to 92.5 percent of the time. Moreover, impacts to marine mammals species and stocks must remain negligible and, in that regard, taking by behavioral harassment may not exceed 12 percent of a marine mammal stock in any given year.

Comment 22:

Throughout the document, the Draft SEIS claims that impacts will be negligible because there is no contradictory data. The absence of evidence does not equate to evidence of absence. In the absence of data, precaution should prevail.

Response:

The absence of evidence regarding effects of these actions on marine mammals does not mean we can assume they have not occurred, and will

not occur in the future. However, we are not relying solely on absence of evidence. The agencies used the best information currently available to analyze the impacts to marine mammals as shown in this document and in more detail in Chapter 4.0 of the Final SEIS. Some of the new information used by NMFS to make its determinations under the MMPA are discussed and summarized in this

Federal Register

notice. That evidence includes a 5-year track record of using SURTASS LFA in an area rich in marine life without incident. In addition, NMFS requires the Navy to conduct mitigation and monitoring, including research to further clarify impacts on marine mammals from LFA sonar.

Comment 23:

Throughout the Draft SEIS, the Navy states that the SURTASS LFA sonar ships move in two dimensions, whereas marine animals move in three dimensions. It uses this logic to state that the amount of time that an animal would be in the sonar transit beam is very low. A ship does move in two dimensions, so if ship strikes were the only concern, then this rationale would work. However, sound propagates in three dimensions so the logic is flawed.

Response:

The Navy has clarified the intent of this statement in the Final SEIS. The statement now reads: “[A] Slowly moving ship, coupled with low system duty cycle, would mean that fish and sea turtles would spend less time in the LFA sonar mitigation zone (180 dB sound field); therefore, with a ship speed of less than 5 knots, the potential for animals being in the sonar transmit beam during the estimated 7.5 to 10 percent of the time the sonar is actually transmitting is very low.”

Comment 24:

In its discussion of acoustic impacts, the Draft SEIS is flawed because it centers its entire analysis on a questionable premise, an SPL threshold of 180 dB RL for marine animal impact.

Response:

The SPL threshold of 180 dB RL was only for potential injury impacts and not for other impacts, such as significant behavioral modifications. Please see Final SEIS Comment 4.0.1 for more information.

Comment 25:

In its discussion of acoustic impacts the Draft SEIS is flawed because it chooses to base its entire evaluation of the potential acoustic impacts to marine mammals on selective data, while ignoring more timely, widely accepted and peer reviewed science, including applicability of actual stranding events. In its discussion of acoustic impacts the Draft SEIS is flawed because it chooses to dismiss evidence suggesting behavioral reaction to sound can produce Level “A” harassment.

Response:

The scientific evidence supporting findings that marine mammals will not be injured at received levels less than 180 dB by SURTASS LFA sonar is provided in the Final SEIS (RTCs 4.0.3, 4.3.1, and 4.3.7 through 4.3.15). LFA sonar has not been implicated in any known marine mammal strandings as discussed elsewhere in this

Federal Register

notice and in the Final SEIS RTC 4.4.9 through 4.4.26. NMFS and the Navy have determined that the potential for injury to marine mammals by exposure to LFA sonar signals at received levels below 180 dB is unlikely.

Even though there is the potential for the LFA sonar signal to injure marine mammals at RLs greater than 180 dB, that possibility is highly unlikely given the reliability of the Navy's tripartite monitoring scheme and, in particular, the demonstrated effectiveness of the HF/M3. NMFS does not dismiss the possibility that behavioral reactions to sound can possibly produce Level A harassment; however, the best available scientific evidence strongly suggests that this is a concern primarily for certain species of odontocetes when exposed under particular conditions to mid-frequency sonar. The results of the LFS-SRP strongly indicate that the behavioral reactions of baleen whales, which hear best in the low frequency range, when exposed to SURTASS LFA sonar are minimal. Although there is no evidence that LF sound can cause biologically significant behavioral responses in odontocetes, and several factors including the inability of such species to hear well in the low frequency range contraindicate such responses, NMFS presumes that, while unlikely, it has the potential to occur. As a result, the Navy is presently planning its 2007-2008 field research for deep diving marine mammal behavioral response studies in an attempt to scientifically address this issue for LFA sonar, MFA, and seismic sources. This is discussed later in this document (see Research).

Comment 26:

The “Determination of Risk Function,” suggests that there is a continuum of severity of behavioral responses to SURTASS LFA sonar signals, ranging from 95 percent of those exposed to 180 dB having significant (if temporary) change in biologically important behavior, down to the first evidence of “significant” change occurring at 119 dB. If SURTASS LFA sonar signals are arriving at the 22-km (12-nm) offshore line at a level of just under 180 dB, then it is likely that near shore areas will be experiencing sound levels significantly above 120 dB. It would be helpful in making more biologically sound decisions if NMFS or the Navy clarified the radius within which received levels could be expected to be 120 dB, 145 dB, and/or 160 dB. The AEI suggests these radii not because these numbers have special or well defined significance, but to suggest that such information would give regulators and researchers a better sense of the likely zones of influence within which behavioral responses might be expected to increase or decrease in severity. At the least, AEI would suggest a lower allowable threshold of received levels at 22 km from shore, to protect these biologically important areas from behavioral disruptions in response to moderate noise levels.

Response:

The AEI is correct that the risk continuum provides a method to determine effects from sound exposure based on the fact that various animals will react differently to LFA sonar signals. The data from the LFS SRP support a linear dose response function, also known as the LFA sonar risk continuum, for sound exposure and the potential for significant behavioral effects. This risk continuum was an integral part of the analysis in the Final EIS and 2002 Final Rule of the potential for SURTASS LFA sonar operations to cause significant behavioral effects in marine mammals. The ranges to RL isopleths and the ocean volumes they would encompass vary under different oceanographic conditions and were analyzed in the Final EIS. Detailed results of these analyses are presented in Subchapter 4.2 of the Final EIS and in Technical Report #2 (Acoustic Modeling Results). Figures B-1 through B-31 of TR 2 provide the parabolic equation (PE) transmission loss (TL) plots for each of the 31 sites. These plots provide TL as a function of depth and range from the source. The analysis determined that there is the potential for marine mammals to be affected by SURTASS LFA sonar.

However, an analysis summarized in Final SEIS Subchapter 4.7.6 indicates that, while increasing the coastal standoff range from 12 nm (22 km) to 25 nm (46 km) decreases exposure to higher RLs for marine animals closest to the shore (shelf species), it does so at the expense of increasing exposure levels for shelf break species and pelagic species.

As a result of the Final EIS analysis, mitigation protocols were developed to prevent injury to marine mammals. Mitigation protocols were not deemed necessary or practical for other than Level A harassment (injury) takes. Results from operations under the initial

5-year set of regulations for LFA sonar are presented in the SURTASS LFA sonar Final Comprehensive Report (see

ADDRESSES

for availability) and indicate that the Level B harassment take numbers for individual stocks of marine mammals in the areas of operations are within the values from the Final EIS analyses.

Comment 27:

The association between anthropogenic ocean noise and its impacts on marine mammals is well documented although there is still scientific uncertainty over the actual causal mechanisms of impacts. It is generally accepted that impacts can range from altered behavior through temporary injury to mortality. Altered behavior can include a startle response and can affect an animal's ability to: feed, find mates, stay on a migration path, communicate, stay at or return to a favored feeding area, nurse, care for young, catch prey and escape predators. Mortality can result directly from exposure to sound or indirectly as a consequence of altered behavior or temporary injury.

Response:

While NMFS agrees with the statement, it cautions that it does not necessarily mean that all loud anthropogenic sounds will cause the stated reactions. NMFS details the relationship between events and LFA sonar throughout this document.

Comment 28:

The Draft SEIS states that “the operation of SURTASS LFA sonar with monitoring and mitigation will result in no lethal takes.” The evidence obtained from actual mortality incidents associated with anthropogenic noise suggests that the mechanisms by which animals are impacted by noise are far less straightforward than the Draft SEIS suggests. There is now increasing evidence that non-auditory injury or permanent loss of hearing are not the only mechanisms by which mortality can result from exposure to noise. For example, an alteration of behavior (Level B) such as a startle response leading to breaching can result in death whereas a gash injury (Level A) can heal and have no long term impact. The Draft SEIS should concede that the knowledge base surrounding the causal mechanisms of marine mammal impacts is too scant to be so readily compartmentalized.

Response:

See responses to Comments 24, 25 and 27. As related to LFA sonar, the Navy performed extensive research to determine the potential for LF transmissions to cause significant behavioral effects in whales (the LFS SRP). There is no indication during these tests that whales surfaced rapidly or dove prematurely in response to LFA sonar source transmissions. The mechanisms to cause such events are based on the theory that MF-naval sonar can cause rapid surfacing and diving, thus resulting in acoustically mediated bubble growth. Also please see the discussion in the Final SEIS (RTCs 4.0.3, 4.3.7, and 4.3.12).

Comment 29:

The Draft SEIS uses 180-dB RL as the threshold for impacts to marine animals and persistently reminds the reader that this is a conservative figure. Field data suggest that this figure is much too high. In the Bahamas multi species mass stranding incident of 2000 estimates of the average sound exposure level that caused those animals to strand was around 140 dB re: 1 microPa. The Draft SEIS dismisses the Bahamas stranding event saying that the hemorrhaging in the stranded animals could have been caused by factors other than acoustic trauma. This is not consistent with the actual findings published in the Interim Report on the event which states “all evidence points to acoustic or impulse trauma” and identifies “mid-range tactical Navy sonars operating in the area as the most plausible source of the acoustic or impulse trauma.”

Response:

First, the Bahamas 2000 stranding event did not involve LFA sonar. Based on the best information available at this time, NMFS believes LFA sonar operations will not cause injury to marine mammals at received levels below 180 dB. Second, the commenter's statement regarding the mid-frequency sonar decibel levels to which the stranded animals were exposed is incorrect. No one knows to what maximum decibel level the animals that ultimately stranded were exposed. Estimates were based on prior near-shore sightings of beaked whales at the locations where those whales were sighted, but they do not reflect the actual maximum received decibel levels of the particular animals that stranded. Third, the Bahamas interim report and further subsequent analysis of the event indicate that the strandings were likely caused by mid-frequency sonar in combination with a list of other contributing factors. The list of contributing factors is generally supported by the workshop on understanding the impacts of anthropogenic sound on beaked whales convened by the U.S. Marine Mammal Commission in 2004 (Cox

et al.

, 2006) and the analysis by D'Spain

et al.

(2006). Whether or not surface ducts, one of the listed contributing factors, occurred during other reported strandings is not relevant to LFA sonar operations. The LFA sonar signals are initially transmitted substantially below 10 m (32.8 ft) water depth and are not likely to have signal strength above 180 dB in the surface duct. To ensure a thorough environmental analysis, however, surface ducting conditions were analyzed in the Final EIS at a number of the 31 model sites. Therefore, with LFA sonar mitigation, no marine mammals, in waters either with or without a surface duct, are expected to be exposed to injurious levels by LFA sonar signals.

Comment 30:

Since the FEIS was completed in January 2001, there have been at least five mass stranding incidents associated with ocean noise and several studies and papers related to the range of impacts of noise on marine mammals. To claim that none of this new data contradicts the assumptions or conclusions in the FEIS is questionable. There is more compelling evidence that: (1) The mechanisms by which animals strand as a result of a noise event are very complex; (2) different mechanisms can be involved and different impacts can result depending on the species and the circumstances; (3) the noise intensities at which animals strand are likely lower than those previously assumed; and (4) tissue injury is not necessary to cause animals to strand and die.

Response:

The issue is not whether anthropogenic sound causes marine mammal strandings, but rather does LFA sonar cause marine mammal strandings. The evidence to date, supported by recent scientific reports, supports the conclusion that the U.S. Navy's LFA sonar is not likely to cause marine mammal strandings. However, an ad hoc committee of international experts under the auspices of the ICES has reviewed the impacts of sonar on cetaceans and fish. They concluded, “No stranding, injury, or major behavioral change has yet been associated with the exclusive use of low frequency sonar” (ICES, 2005). This is further supported by 36 scientists in their recently published paper which arose from the Marine Mammal Commission workshop on the impacts of anthropogenic noise on beaked whales (Cox

et al.

, 2006). Therefore, the statement that there are no new data contradicting the assumptions or conclusions in the Final EIS and Final SEIS remain correct. Moreover, five years of SURTASS LFA sonar use without evidence of strandings, injury, or other major behavioral changes support the conclusions of the Final OEIS/EIS and the Final Rule 2002. However, NMFS continues to view this issue seriously and does not dismiss it simply because a stranding has not been observed. For more detailed information, please see the Final SEIS

(RTCs 4.0.3, 4.3.1, 4.3.2, 4.3.7, 4.3.8, 4.3.9, and 4.3.12).

Comment 31:

The Draft SEIS mentions only three noise related marine mammal stranding events under the heading “Strandings potentially related to anthropogenic sound.” There is irrefutable evidence that anthropogenic sound causes marine mammal strandings. What is not known with any scientific certainty is the actual causal mechanisms. In listing only three marine mammal stranding incidents potentially related to anthropogenic sound, the Draft SEIS is being disingenuous. Not only are there many more strandings, but when all atypical mass strandings are tabulated, the overwhelming majority is associated with naval maneuvers, and likely sonar usage. (The commenter also provided the table from the ICES (2005) Report of the Ad hoc Group on the Impact of Sonar on Cetaceans and Fish).

Response:

The Navy's intention was to examine three of the more studied stranding events in which naval sonars were implicated as a potential cause. This subchapter has been expanded in the Final SEIS based on stranding event information cited in more recent reports, such as ICES AGISC Report (ICES, 2005), and reports on the potential causes presented by ICES (2005), Cox

et al.

(2006), and D'Spain

et al.

(2006). NMFS believes that this revision is adequate as related to the potential for SURTASS LFA sonar to cause strandings because LFA sonar was not considered causative in any of these events and, indeed, low frequency sonar has never been implicated in any stranding, with the possible exception of the Greece stranding in 1996, during which mid-frequency sonar was also employed.

Comment 32:

The Navy has not reported any marine mammal stranding incident that has occurred in the vicinity of its activities. The Draft SEIS claims that SURTASS LFA sonar has not been implicated in any stranding event. This is not accurate. An LFA sonar system was implicated in the mass stranding of twelve Cuvier's beaked whales in 1996 in Greece though as the Draft SEIS states, the inner ears were not examined. This does not mean that LFA sonar use did not cause the animals to strand. The usage of LFA sonar has also been far more restricted than mid frequency sonar for which there are more associated mass stranding events.

Response:

The Draft SEIS was correct. SURTASS LFA sonar have never been implicated in a stranding. While there was a LF component of the sonar potentially related to the Greek strandings in 1996, only MF components were implicated in the strandings in the Bahamas in 2000, Madeira 2002, and Canaries in 2002. This suggests that the LF component in the Greek strandings was not causative (Cox

et al.

, 2006; ICES, 2005). In its discussion of the Bahamas stranding, Cox

et al.

(2006) stated, “The event raised the question of whether the mid-frequency component of the sonar in Greece in 1996 was implicated in the stranding, rather than the low frequency component proposed by Frantzis (1998).” The ICES in its “Report of the Ad Hoc Group on the Impacts of Sonar on Cetaceans and Fish” is in agreement with Cox

et al.

(2006) stating that the association of MF sonar in the Bahamas, Madeira, and Canary Island strandings suggest that it was not the LF component in the NATO sonar that triggered the Greece stranding of 1996, but rather the MF component (ICES, 2005). The ICES (2005) report also concluded that no strandings, injury, or major behavioral change have yet to be associated with the exclusive use of LF sonar.

Since October 14, 2003, SURTASS LFA sonar use has been restricted under a permanent injunction to limited areas in the western Pacific Ocean (see Final SEIS, Subchapter 1-2.1, Figures 1-1 and 4-4.2). Since commencing operations in 2003, the

R/V Cory Chouest

and USNS IMPECCABLE have completed 40 missions from January 2003 to August 2006 under the first four LOAs (DON, 2007). The general areas are known to the public because they are based on the Court Order, published in the Draft and Final SEIS, and incorporated into the subsequent NMFS LOAs. The locations and times of LFA sonar active operations are reported to NMFS quarterly (classified report) as required in the Final Rule and annual LOAs. These operations, with mitigation, have produced no known Level A takes on marine mammals as reported in the Annual Reports (DON, 2003a; 2004a; 2005a; 2006a) and the Final Comprehensive Report (DON, 2007). Reviews of stranding reports in the LFA sonar operating area showed that there were a total of 19 strandings reported in Asia (four in Taiwan, nine throughout the Philippines, two in Thailand, two in Indonesia, and two in China) (The Cetacean Stranding Database, accessed: 11/28/2006). None of these strandings were coincident either temporally or spatially with LFA sonar operations.

Moreover, the Northwestern Pacific Ocean areas where SURTASS LFA sonar is presently operating are some of the most heavily populated areas in the world and cannot be considered “remote.”

As to the possibility of unreported strandings, the NMFS and the Navy do not consider that this is a very likely scenario for LFA sonar operations. Even though a visual observer onboard the vessel will be unable to see an animal that strands on the shoreline due to operations being greater than 12 nm (22 km) from land, this is not relevant because LFA sonar is unlikely to cause injury beyond the 180-dB mitigation zone (normally 1 km (0.5 nm) radius). Level A (injury) harassments are determined based on actual observations/detections within the LFA sonar mitigation zone. With passive and active acoustic detection, the probability of detection within this zone is over 95 percent for a single marine mammal (see Final EIS, Subchapters 2.3.2.2 and 4.2.7.1.). For multiple animals, the value is nearly 100 percent. The area of the northwestern Pacific Ocean, where LFA sonar vessels are currently operating, is not a remote area and there are stranding networks in the region. A review of reported strandings in the area does not show any correlations to LFA sonar operations either spatially or temporally (see discussion later in this document on strandings in Taiwan).

Comment 33:

The Draft SEIS states that no Level A harassment incidents have been reported in the area of usage; however, it does not relate the effort undertaken to search for such incidents or mention reports of Level “B” harassment incidents.

Response:

See Comment 32.

Comment 34:

The association between mid frequency sonar usage and strandings was not realized until decades after its introduction.

Response:

NMFS agrees, noting that Balcomb and Claridge (2001) reported that beaked whale strandings have increased since the use of MF sonar in the 1960s. However, the association between MF-sonar and strandings appears limited to a confluence of factors. Stranding networks weren't active until much later than the 1960's, but have been active since SURTASS LFA sonar came into use. Certainly, SURTASS LFA sonar has received great scrutiny with respect to the potential for strandings and none have been observed.

Comment 35:

The Draft SEIS appears to be only concerned about impacts producing Level A harassment which it claims will be negligible. The impacts from behavioral alteration to individual animals are dismissed as inconsequential. Behavioral impacts can not only produce level A harassment, but impacts to individuals are

significant especially for endangered populations, and can have population level consequences no matter what the status of the species.

Response:

There are several types of Level B harassment that can result from anthropogenic sounds. Two types of behavioral effects that have potential for population level effects are masking and stress. These will be addressed here. (also see the Final SEIS RTCs 4.0.3 and 4.3.12 in Comment 1, SEIS RTC 4.3.17 in Comment 5, SEIS RTC 4.3.2 in Comment 6, and SEIS RTC 4.3.33 in Comment 7). Other potential Level B harassment effects are addressed elsewhere in this rulemaking document. Also, please see the Biological Opinion issued under section 7 of the ESA for this action by NMFS (see ESA later in this document).

In regard to masking, the commenter is confusing the avoidance response of migrating gray whales and bowhead whales with masking. There was no evidence of masking in any of the research on these two species. Certainly in the gray whale case, the interpretation by the scientists who conducted the research was that the whales responded but responses were not interpreted as having a significant behavioral impact. Furthermore, a received level of 120 dB for LFA sonar would not mask the species-specific sounds of any low frequency mysticete, although under certain, rare circumstances it might interfere with species recognition. The masking effects of the SURTASS LFA sonar signal are expected to be limited for a number of reasons. First, the frequency range (bandwidth) of the system is limited to about 30 Hz, and the instantaneous bandwidth at any given time of the signal is small, on the order of 10 Hz. Second, the LFA sonar signal is active (or on) only about 7.5 percent of the time (i.e., low duty cycle based on historical LFA sonar operations, but may be on for up to 20 percent of the time) and limited to periods during actual missions. Therefore, the effect of masking will be limited because animals that use this frequency region typically use broader bandwidth signals. As a result, the chances of an LFA sonar sound actually overlapping whale calls at levels that would interfere with their detection and recognition would be extremely low.

Regarding stress, stress can be defined as a threat to homeostasis (Fair and Becker, 2000) and is frequently measured with changes in blood chemistry (Thomas

et al.

, 1990; Romano

et al.

, 2004; Smith

et al.

, 2004a). Thomas

et al.

(1990) exposed captive belugas to recorded industrial noise for 30 minutes at a time, with a total exposure of 4.5 hours over 13 days with a source level of 153 dB. Catecholamine blood levels were checked both before and after noise exposure; however, no significant differences in blood chemistry were observed. Another experiment that measured blood chemistry, but also varied the sound level is described in Romano

et al.

(2004). In this experiment, a beluga whale was exposed to varying levels of an impulsive signal produced by a watergun. The levels of three stress related blood hormones (norepinephrine, epinephrine and dopamine) were measured after control, low level sound (171-181 dB SEL) exposure and high level (184-187 dB SEL) sound exposure. There were no significant differences between low level sound exposure and control, while the high level sound exposure did produce elevated levels for all three hormones. Furthermore, regression analysis demonstrated a linear trend for increased hormone level with sound level.

Less relevant to marine mammals, but still informative, Smith

et al.

(2004a) exposed goldfish (a hearing specialist fish) to continuous background noise of 160-170 dB RL. There was a “transient spike” in blood cortisol levels within 10 minutes of the onset of noise that was loud enough to cause TTS. However, this cortisol spike did not persist and there was no long term physiological stress reaction in the animals.

These data support a linear dose response function (like the LFA sonar risk continuum) for sound exposure and the onset of stress, with only high levels of sound leading to a stress reaction. The extrapolation of the response thresholds from the Romano

et al.

(2004) experiment to the LFA sonar situation is tenuous because of the differences in the signals, but the relationship between sound level and stress is supported by several studies. As mentioned elsewhere, there are some recent data (e.g., Evans, 2003) implicating synergistic effects from multiple stressors, including noise. Although there are no data to support synergistic effects, similar impacts might occur with marine mammals, given the multiple stressors that often occur in their environment. This indicates to NMFS that while stress in marine animals could possibly be caused by operation of the LFA sonar source, it is likely to be constrained to an area much smaller than the zone of audibility, probably closer in size to the mitigation zone around the vessel.

Comment 36:

The LFS SRP Phase II conducted by the Navy to determine LFA sonar impacts on migrating whales found that when the source was located in the whales' migratory path (approximately 1 km (0.54 nm) from shore), gray whales avoided levels below 150 dB. The SRP showed negligible avoidance by the whales when the source was located over 2 km (1.1 nm) from shore. From the results of the LFS SRP Phase II, the Navy concluded no biologically significant response. Perhaps in actuality more sensitive individuals or mother calf pairings tend to hug the coast during migration. For some groups, the most sensitive animals may be crucial to a group's survival as these may be the first individuals to become aware of predators or of dangerous situations. To lose sensitive animals or nursing mothers from a group could have population level consequences.

Response:

NMFS believes the characterization of the Navy's conclusion is out of context. See the Final EIS Subchapter 4.2.4.3. NMFS does not believe that some whales “hugged” the coast of California during the LFS SRP. For this phase of the SRP, the sound source was moored offshore of the central California coast, near Point Buchon. Shore-based observers tracked whales using methods that provided highly sensitive measures for avoidance responses. These observers would have sighted whales along the coast line. Also, observers on the playback vessel also carefully monitored marine mammals in order to stop broadcasting in case of worrisome behavioral reactions or if any marine mammals were sighted at close enough range that the sound level to which they were exposed might exceed the maximum planned exposure level (155 dB).

The issue of potential calf strandings during the LFS SRP in Hawaii was addressed in the Final EIS RTC 4 5.25 where it was concluded that these events were not related to LFA sonar testing. Masking of communications could potentially affect the mother calf bond; however, masking effects from the SURTASS LFA sonar signal are extremely unlikely and are expected to be negligible considering the short duty cycle and other factors discussed in this document. The rationale for this is discussed in Final SEIS RTCs 4.3.23 and 4.3.36. Thus, LFA sonar signals are not expected to disrupt the mother calf bond.

Comment 37:

An aversion response can occur many tens of miles from the source, and father away if it is in the direct path of the beam-formed or ducted signal.

Response:

Given that the LFA sonar sound source can be detected at moderate to low levels over large areas of the ocean, the Navy (and NMFS) had concerns at the initiation of the NEPA process in 1996 that there was the potential for large percentages of species/stocks to be exposed; if animals would be disturbed at these moderate-to-low exposure levels such that they experience a significant change in a biologically important behavior, then such exposures could potentially have an impact on rates of reproduction or survival. Knowing that cetacean responses to LF sound signals needed to be better defined using controlled experiments, the Navy helped develop and supported the independent three-year LFS SRP beginning in 1997. The study analyzed the behavioral responses of whale species that have the greatest sensitivity to low frequency sounds and thus were believed to be the most vulnerable, potentially, to LFA sound. This field research program was designed to address three important behavioral contexts for baleen whales: (1) Blue and fin whales feeding in the southern California Bight, (2) gray whales migrating past the central California coast, and (3) humpback whales breeding off Hawaii. Taken together, the results from the three phases of the LFS SRP do not support the hypothesis that most baleen whales exposed to RLs near 140 dB would exhibit disturbance behavior and avoid the area. These experiments, which exposed baleen whales to RLs ranging from 120 to about 155 dB, detected only minor, short-term behavioral responses. Short-term behavioral responses do not necessarily constitute significant changes in biologically important behaviors.

These results have been supported by recent, peer-reviewed papers. Croll

et al.

(2001a) studied the effects of anthropogenic LF noise (SURTASS LFA sonar) on the foraging ecology of blue and fin whales off San Nicolas Island, California. Overall, the whale encounter rates and diving behavior appeared to be more strongly linked to changes in prey abundance associated with ocean parameters than to LFA sonar transmissions. In some cases, whale vocal behavior was significantly different between experimental and non-experimental periods. However, these differences were not consistent and did not appear to be related to LF sound transmissions. At the spatial and temporal scales examined, Croll

et al.

(2001a) stated that they found no obvious responses of whales to a loud, anthropogenic, LF sound.

Both Miller

et al.

(2000) and Fristrup

et al.

(2003) published on the results of tests conducted with male humpback singers off Hawaii in which they evaluated variation in song length as a function of exposure to LF sounds. In spite of methodological differences, the results of both studies indicated that humpback whales slightly increased their songs in response to LF broadcasts. Fristrup

et al.

(2003) found that the fraction of variation in song length that could be attributed to LF broadcast was low and concluded that the effects of LF broadcast did not impose a risk of dramatic changes in humpback whale singing behavior that would have demographic consequences. For more information please also see SEIS RTC 4.3.30.

Comment 38:

SURTASS LFA sonar impacts the vocalizations and other behavior of humpback whales.

Response:

NMFS does not disagree with the potential impacts of LFA sonar on vocalization and other behavior. The justification for the conclusion that the potential effects on the stocks of marine mammals from behavioral changes would be minimal is discussed in the Final SEIS in RTC 4.3.29. The potential effects of masking are discussed in the Final SEIS RTCs 4.3.1 and 4.3.23. The Miller

et al.

(2000) article “Whale songs lengthen in response to sonar” concerning observations of male humpback whales during Phase III of the LFS SRP was addressed in the Final EIS RTC 4-5.19 and in the NMFS Final Rule RTC SIC16 and SIC17. Fistrup

et al.

(2003) used a larger data set from Phase III to describe song length variability and to explain song length variation in relation to LF broadcasts. In spite of methodological and sample size differences, the results of the two analyses were generally in agreement, and both studies indicated that humpback whales tend to lengthen their songs in response to LF broadcasts.

Fristrup

et al.

(2003) provides a detailed picture of short-term response as compared to behavioral variation observed in the absence of stimuli. These responses were relatively brief in duration, with all observed effects occurring within 2 hours of the last LFA sonar source transmission. It should be noted that these effects were not salient to the acoustic observers on the scene, but were revealed by careful statistical analyses (Fistrup

et al.

, 2003). Aside from the delayed responses, other measures failed to indicate cumulative effects from LF broadcasts, with song-length response being dependent solely on the most recently LF transmission, and not the immediate transmission history. The modeled seasonal factors (changes in surface social activities) did not show trends that could be plausibly explained by cumulative exposure. Increases in song length from early morning to afternoon were the same on days with and without LF transmissions, and the fraction of variation in song length that could be attributed to LF broadcast was low. Fistrup

et al.

(2003) found high levels of natural variability in humpback song length and interpreted the whales' responses to LF broadcasts to indicate that exposure to LFA sonar would not impose a risk of dramatic changes in humpback whale singing behavior that would have demographic consequences.

Comment 39:

It is impossible to comment fully on the Acoustic Integration Model (AIM), the program used by the Navy to calculate the system's impacts, because that model has not been released to the public. Disclosure of the model must occur for public comment to be meaningful under NEPA and the Administrative Procedure Act (APA) to be met.

Response:

The Acoustic Integration Model (AIM) contains proprietary programming that prevents its release to the public. As a result, in response to a different incidental take application (Draft EIS for Gulf of Mexico Seismic Surveys), AIM recently underwent an independent scientific review by the NMFS-sponsored Center for Independent Experts (CIE). The CIE review took place September 25-27, 2006. A report from that review is publicly available on the NMFS Web site (

http://www.nmfs.noaa.gov/pr/permits/incidental.htm

). Additional documentation can be found on the SURTASS LFA sonar Web site (see

ADDRESSES

).

Comment 40:

Models used by the Navy in its applications for LOAs to assess its actual work in the Pacific, and in its Final EIS to estimate impacts in sample coastal areas, in large part assume a fairly even distribution of marine mammals across a wide area of ocean, failing to take the possibility that certain animals, like beaked whales and sperm whales, may be concentrated in particular habitats. Specifically, the Navy has not conducted research on beaked whale habitat preferences. In the limited modeling we have seen, the Navy frequently assumes that populations of marine mammals are relatively unstructured, such that individual animals are improbably considered part of region-wide, basin-wide, or even worldwide stocks. The Navy's stock assessments in its LOA applications are based on incomplete and out-of-date information, leading to

a significant underestimation of species abundance and therefore impacts.

Response:

When there is no specific data on marine mammal distribution, impact prediction modeling uses an even distribution over the ocean area, since offshore concentrations of animals are not fixed in space or time. Nearshore concentrations can be relatively fixed in time or space, due to physical forcing from the steep bathymetry and seasonal variations (

e.g.

, Monterey Canyon or Hudson Canyon). However, LFA sonar operates in deeper, offshore waters where the concentrations are fluid due to changing water mass conditions. Therefore an even distribution of animals is the one with the least assumptions. Basically, the model assumes that individuals of the species can occur anywhere within their ranges with equal probability over a long time. On any given day, the distribution of any given species is likely to be highly non-uniform. Over a long period of time the fluctuations in density are likely to even out. Therefore, assuming an even distribution for the purposes of assessing potential impacts is reasonable and appropriate.

NMFS believes that the latest information available is used by NMFS and the Navy when assessing impacts on marine mammals by LFA sonar. Regarding beaked whale research, NMFS notes that the Office of Naval Research (ONR) and SERDP (Strategic Environmental Research and Development Program) has funded the following beaked whale research:

MacLeod, C. D., and G. Mitchell. 2006. Key areas for beaked whales worldwide. J. Cetacean Res. Manage. 7(3):309-322.

MacLeod, C. D., W. F. Perrin, R. Pitman, J. Barlow, L. Balance, A. D'Amico, T. Gerrodette, G. Joyce, K. D. Mullin, D. L. Palka, and G. T. Waring. 2006. Known and inferred distributions of beaked whale species (Cetacea: Ziphiidae). J. Cetacean Res. Manage. 7(3):271-286.

Redfern, J. V., M. C. Ferguson, E. A. Becker, K. D. Hyrenbach, C. Good, J. Barlow, K. Kaschner, M. F. Baumgartner, K. A. Forney, L. T. Ballance, P. Fauchald, P. Halpin, T. Hamazaki, A. J. Pershing, S. S. Qian, A. Read, S. B. Reilly, L. Torres, and F. Werner. 2006. Techniques for cetacean-habitat modeling. MEPS 310:271-295.

Ferguson, M. C., J. Barlow, B., S. B. Reilly, and T. Gerrodette. 2006. Predicting Cuvier's (Ziphius cavirostris) and Mesoplodon beaked whale population density from habitat characteristics in the Eastern Tropical Pacific Ocean. JCRM 7(3):287-299.

In addition, ONR and SERDP have funded the development and fieldwork for the sound-and-orientation recording tag (DTAG), which has been successfully attached with suction cups to beaked whales (Tyack

et al.

, 2006). These data are providing critically valuable information on the movement and diving behaviors of beaked whales, both of which are important to know in order to understand the acoustic exposure that the animals may receive.

As stated in the Final SEIS Subchapter 2.7, the NMFS initial LOA under Condition 7(d) required the Navy to conduct research in accordance with 50 CFR § 216.185(e). The SURTASS LFA sonar LTM Program has been budgeted by the Navy at a level of approximately $1M per year for five years, starting with the issuance of the first LOA. The status of this research was summarized in Table 2-5 of the Final SEIS. Finally, planning has commenced for a 2007-2008 deep-diving odontocetes behavioral response study (BRS) to determine the potential effects of LFA sonar, MFA, and seismic sources on beaked whales and other deep diving odontocetes at an estimated cost of $3M per year. The BRS study is discussed later in this document.

Regarding stock assessment data, the modeling analysis considers the total amount of risk for each marine mammal species by summing a particular species' risk estimate within that stock, across areas of operation for each mission. This methodology does not assume that populations are unstructured, but includes the best information available on the reproductive behavior of each species at each mission site in order to determine stock affiliation and the total risk to the sustainability of each stock. Stock assessment data within U.S. waters are required to be updated annually under the MMPA, with new stock assessments being published when new data are available. The best available data were used in all instances of the modeling analysis for determining stock abundance and distribution.

The Navy states that it performs regular reviews of the latest research, including updating stock and density data. The Navy's applications for SURTASS LFA sonar LOAs are submitted after conducting a thorough review of the latest data on the marine animals present in the potential operating areas.

The Final EIS states, “The model runs are designed to portray high potential effects for each site. For example, seasons were selected based on the potential for maximum LF-sensitive animal abundance.” (Please see FOEIS/EIS Subchapters 4.2.1 and 4.2.2.2, and RTCs 4-3.8, 4-3.9, and 4-3.11.)

Comment 41:

The Navy incorrectly claims that significant impacts on stocks and populations, as modeled for its LOA applications, would necessarily occur at percentages lower than those assumed in the Navy's modeling of coastal area and NMFS Final Rule, even disregarding the underestimates of take resulting from the other errors described. The Navy's approach to modeling behavioral impacts from multiple exposures is not conservative.

Response:

NMFS disagrees with the commenter's statement regarding the Navy's approach to modeling behavioral impacts from multiple exposures not being conservative. Subchapter 4.2.3.1 of the Final EIS provides details on how the Navy derived the L + 5 log

10

(N) formula for a single ping equivalent (SPE). The SPE concept is related to widely accepted methods for comparing sounds of different durations. It is universally acknowledged that increased exposure duration increases the severity of potential impact. The SPE calculation is conservative in assuming that the increase in potential effects observed by extending the duration of a continuous sound stimulus applies to a sequence of SURTASS LFA sonar pings, even though the transmissions are separated by many minutes when the system is off. This applies to SURTASS LFA sonar-type signals, not continuous sound. In this process, an SPE received level is larger than the maximum RL of any single ping in sequence. Also, the SPE for a sequence consisting of a single loud ping and a long series of much softer pings is almost the same as the level of a single loud ping. A ping duration (length) of 60 seconds was assumed in the modeling and risk assessment calculations using SPE. The adoption of 60 seconds and 20 percent as the standard ping duration and duty cycle, respectively, for calculations in the Final EIS, provides a reasonable estimate of the potential for effects from real-world SURTASS LFA sonar operations without sacrificing the conservative nature of the analysis process.

Comment 42:

There is an unknown history of exposure of animals in an area where active sonar is regularly used.

Response:

The adequacy of scientific information on marine animals is discussed in Subchapter 1.4.2 of the Final EIS. It states that there is an urgent need for better methods for measuring and estimating potential risk. These data gaps have necessitated the use of various models and extrapolations in order to provide a rational basis for the assessment of potential risk from exposure to LF sounds. To address some of these gaps, the Navy performed underwater acoustic modeling and supported the LFS SRP to study the

potential effect of LF sound on free-ranging marine mammals. This research did not specifically address the issue of LF impact on marine mammal hearing; rather, it focused on the behavioral responses of baleen whales to controlled exposure from SURTASS LFA sonar-like signals. In general, understanding the mechanics of hearing and the biological functions of sounds for marine mammals has improved considerably over the past decade. Specific information on the effects of most types of human-made underwater noise on marine animals is incomplete, but has also increased in recent years. However, as the environmental evaluation of the SURTASS LFA sonar system progressed, the Navy recognized that additional research was required in several areas to address some basic gaps in scientific knowledge. This included development of a scientifically reasonable estimate of the underwater sound exposure levels that may cause injury to marine mammals, and research on the potential effects of LF sound on marine mammal behavior. While recognizing that not all of the questions on the potential for LF sound to affect marine life are answered, and may not be answered in the foreseeable future, NMFS believes the Navy has combined scientific methodology with a prudent approach throughout the Final EIS and SEIS to protect the marine environment. Although there are recognized areas of insufficient knowledge that must be accounted for when estimating the potential direct and indirect effects on marine life from SURTASS LFA sonar, the present level of understanding is adequate to place reasonable bounds on potential impacts. Therefore, though data on specific exposure of anthropogenic sounds, particularly sonar, on the marine environment is limited, the Navy and NMFS have taken this into account during their analyses. Moreover, we know much more about the impacts of different types of sonar in the marine environment today than we knew five years ago, when SURTASS LFA went through the environmental compliance process the first time, and the best scientific data that we have indicates that SURTASS LFA can be operated safely with the prescribed mitigation, in a manner that has no more than a negligible impact on marine mammal species and stocks.

Comment 43:

There is a low level of accuracy with which the exposed individuals can be monitored in real time.

Response:

Sound field limits are estimated using near-real-time environmental data and underwater acoustic performance models. These models are an integral part of the SURTASS LFA sonar processing system. The acoustic models help determine the sound field by predicting the SPLs, or RLs, at various distances from the SURTASS LFA sonar source location. Acoustic model updates are nominally made every 12 hours, or more frequently when meteorological or oceanographic conditions change. For further information, please see the Final SEIS, RTC 5.1.1. Though individuals cannot be effectively monitored beyond the reach of the HF/M3, the sound field is monitored in near-real-time.

Comment 44:

The intense sound generated by military active sonar can induce a range of adverse effects in whales and other species, from significant behavioral changes to stranding and death. In a 2004 symposium at the International Whaling Commission (IWC), more than 100 whale biologists concluded that the association between sonar and beaked whale deaths is very convincing and appears overwhelming. Mass mortalities, though an obvious focus of much reporting and concern, are likely only the tip of the iceberg of sonar's harmful effects. Marine mammals are believed to depend on sound to navigate, find food, locate mates, avoid predators, and communicate with each other. Flooding their habitat with man-made, high-intensity noise interferes with these other functions.

In addition to strandings and non-auditory injuries, the harmful effects of high-intensity sonar include (1) temporary or permanent loss of hearing; (2) avoidance behavior; (3) disruption of biologically important behaviors such as mating, feeding, nursing, or migration, or loss of efficiency in conducting those behaviors; (4) aggressive (or agonistic) behavior; (5) masking of biologically meaningful sounds; (6) chronic stress; (7) habituation; and (8) declines in the availability and viability of prey species, such as fish and shrimp.

Response:

The use of the term “sonar” does not reflect what Annex K of the IWC 2004 Scientific Committee Report actually stated. The Report does not implicate LFA sonar in the stranding of beaked whales. The full text of the quoted statement is: “The weight of accumulated evidence now associates mid-frequency, military sonar with atypical beaked whale mass strandings. This evidence is very convincing and appears overwhelming.”

There are different types of anthropogenic sounds associated with possible impacts to and strandings of marine mammals. There are naval sonar and seismic airgun arrays, each with different characteristics and purposes. Many lump these types together. Accordingly, when there is a stranding that may be associated with the use of one type of sonar or sound source, all sources are implicated—a premise that does not stand up to scientific scrutiny in the marine bio-acoustics community. A wide range of naval sonars are used to detect, localize and classify underwater targets. For the purposes of the SURTASS LFA sonar Final SEIS, the MMPA application, and this Final Rule, these systems are categorized as LFA sonar (less than 1000 Hz) and MFA sonar (1 to 10 kHz). Table 1 in this document provides pertinent information on different types of LFA sonar and MFA sonar. General information is also provided on airgun arrays. (We also note that sonar signals are generally coherent while air guns are impulsive.)

Table 1.—Comparison of Underwater Acoustic Source Properties

Source type

SURTASS LFA sonar

AN/SQS 53C (MF)

AN/SQS 56 (MF)

Air gun array (LF)

Source Level

215 dB per element

235 dB

223 dB

260 dB.

Pulse Duration

Variable 6 to 100s. Never longer than 10s at single freq

1-2 s

1-2 s

0.02 s.

Inter-pulse Time

6 to 15 min

24 s

24 s

9-14 s.

Center Frequency

100-500 Hz

2.6 & 3.3 kHz

6.8, 7.5, & 8.2 kHz

Broadband.

Bandwidth

30 Hz

100 Hz

100 Hz

Wideband.

Source Depth

Array 87 to 157 m. Center 122 m

8 m

6 m

6-10 m.

Beamwidth

Omni-directional in horizontal

40 degrees

30 degrees

Function of freq.

Beam Direction

Horizontal

3 degrees down from horizontal

Horizontal

Vertical.

MF = mid frequency; LF = low frequency.

Source: D'Spain

et al.

(2006); DON (2001).

Cox

et al.

(2006) provides a summary of common features shared by the strandings events in Greece (1996), Bahamas (2000), and Canary Islands (2002). In addition to use of MF sonar, these included deep water close to land (such as offshore canyons), presence of an acoustic waveguide (surface duct conditions), and periodic sequences of transient pulses (i.e., rapid onset and decay times) generated at depths less than 10 m (32.8 ft) by sound sources moving at speeds of 2.6 m/s (5.1 knots) or more during sonar operations (D'Spain

et al.

, 2006). A number of these features do not relate to LFA sonar operations. First, the SURTASS LFA sonar vessel operates with a horizontal line array (SURTASS: a passive listening system) of 1,500 m (4,921 ft) length at depths below 150 m (492 ft) and a vertical line array (LFA sonar source) at depths greater than 100 m. Second, operations are limited by mitigation protocols to at least 22 km (12 nm) offshore. Therefore, for these reasons SURTASS LFA sonar cannot be operated in deep water that is close to land. Finally, the LFA sonar signal is transmitted at depths well below 10 m (32.8 ft), and the vessel has a slow speed of advance of 1.5 m/s (3 knots).

While there was a LF component to the sonar potentially related to the Greek stranding in 1996, only mid-frequency components were present in the strandings in the Bahamas in 2000, Madeira in 2002, and Canaries in 2002. This supports the logical conclusion that the LF component in the Greek stranding was not causative (ICES, 2005; Cox

et al.

, 2006). In its discussion of the Bahamas stranding, Cox

et al.

(2006) stated, “The event raised the question of whether the mid-frequency component of the sonar in Greece in 1996 was implicated in the stranding, rather than the low-frequency component proposed by Frantzis (1998).” The ICES in its “Report of the Ad-Hoc Group on the Impacts of Sonar on Cetaceans and Fish” raised the same issue as Cox

et al.

, stating that the consistent association of MF sonar in the Bahamas, Madeira, and Canary Islands strandings suggest that it was the MF component, not the LF component, in the NATO sonar that triggered the Greek stranding of 1996 (ICES, 2005).

Most odontocetes, such as beaked whales, have relatively sharply decreasing hearing sensitivity below 2 kHz. If a cetacean cannot hear a sound of a particular frequency or hears it poorly, then it is unlikely to have a significant behavioral impact (Ketten, 2001). Therefore, it is unlikely that LF transmissions from LFA sonar would induce behavioral reactions from animals that have poor LF hearing, e.g. beaked whales, bottlenose dolphins, striped dolphins, harbor porpoise, belugas, and orcas (summarized in: Nedwell

et al.

, 2004).

New data describing potential mechanisms of harm to marine mammals from sonar are concerned with acoustically mediated bubble growth and resonance. Cox

et al.

(2006) stated that it is premature to judge acoustically mediated bubble growth as a potential mechanism and recommended further studies to investigate the possibility. The analysis by the Navy (Cudahy and Ellison, 2002) and reports from two workshops on acoustic impacts (DOC, 2002; Cox

et al.

, 2006) support the conclusion that resonance from LFA sonar operations is not a “reasonably foreseeable” impact. The ICES (2005) report concluded that no strandings, injury, or major behavioral change has yet to be associated with the exclusive use of LF sonar. Please see Final SEIS RTCs 2.5.2 and 4.0.3 for additional discussions.

Therefore, the numerous scientists, who participated in the 2004 Workshop convened by the U.S. Marine Mammal Commission (Cox

et al.

, 2006), and the ICES AGISC (2005), support the logical conclusion that LFA sonar is not related to marine mammal strandings.

The masking effect of the SURTASS-LFA sonar signal will be limited for a number of reasons. First, the bandwidth of the system is limited (30 Hz), and the instantaneous bandwidth at any given time of the signal is small, on the order of 10 Hz. Therefore, within the frequency range in which masking is possible, the effect will be limited because animals that use this frequency range typically use signals with greater bandwidth. Thus, only a portion of the animal's signal would be masked by LFA sonar. Furthermore, the average duty cycle when LFA sonar is in operation, is always less than 20 percent, and based on past LFA sonar operational parameters (2003 to 2007) is nominally 7.5 to 10 percent (as stated in Chapter 2 of the Final SEIS) which means that for 80-92.5 percent of the time there is no risk of animal signals being masked by the LFA sonar signal when LFA sonar is operating. Therefore, within the area in which masking is possible, the effect will be limited because animals that use this frequency region typically use broader bandwidth signals. As a result, the chances of an LFA sonar sound actually overlapping whale calls at levels that would interfere with their detection and recognition would be extremely low. The potential effects of masking are discussed in the Final SEIS RTCs 4.3.1 and 4.3.23.

In regards to biologically significant behaviors, the risk continuum explicitly represents the potential for significant change in a biologically important behavior within the 119 to 180 dB RL range. For additional information, please see the previous discussion on this issue and also the Final EIS (RTCs 4-5.2, 4-5.6, 4-5.12, 4-5.22, 4-6.2, 4-6.3), and Appendix D. The conclusion that the potential effects on the stocks of marine mammals from behavioral changes would be minimal is discussed in the Final SEIS (RTC 4.3.29). It is reiterated that during Phase I of the LFS SRP research, there were times when the test source level was at the higher, operational level. During such test periods received levels at the subject animals were within the range as specified in the research permit and responses were no different than those observed when using lower source levels.

The Miller

et al.

(2000) article “Whale songs lengthen in response to sonar” concerning observations of male humpback whales during Phase III of the LFS SRP was addressed in the Final OEIS/EIS RTC 4-5.19 and in NMFS Final Rule RTC SIC16 and SIC17. Fristrup

et al.

(2003) used a larger data set from Phase III to describe song length variability and to explain song length variation in relation to LF broadcasts. In spite of methodological and sample size differences, the results of the two analyses were generally in agreement, and both studies indicated

that humpback whales tend to lengthen their songs in response to LF broadcasts.

The Fristrup

et al.

(2003) results provide a detailed picture of short-term response as compared to behavioral variation observed in the absence of the stimuli. These responses were relatively brief in duration, with all observed effects occurring within 2 hours of the last LFA sonar source transmission. It should be noted that these effects were not salient to the acoustic observers on the scene, but were revealed by careful statistical analyses (Fristrup

et al.

, 2003). Aside from the delayed responses, other measures failed to indicate cumulative effects from LF broadcasts, with song-length response being dependent solely on the most recent LF transmission, and not the immediate transmission history. The modeled seasonal factors (changes in density of whales sighted near shore) and diurnal factors (changes in surface social activities) did not show trends that could be plausibly explained by cumulative exposure. Increases in song length from early morning to afternoon were the same on days with and without LF transmissions, and the fraction of variation in song length that could be attributed to LF broadcast was low. Fristrup

et al.

(2003) found high levels of natural variability in humpback song length and interpreted the whales' responses to LF broadcasts to indicate that exposure to LFA sonar would not impose a risk of dramatic changes in humpback whale singing behavior that would have demographic consequences.

The effects of SURTASS LFA sonar on fish are discussed elsewhere in this document. Based on the analysis in the Final SEIS, Chapter 4.1, it is not believed that marine mammal prey species will be affected by SURTASS LFA sonar.

Comment 45:

The proposed rule-making cites the ICES report on sonar (which was written partly by non-independent scientists receiving funding from U.S. or Royal Navy, or working for the U.S. government), but does not cite the conclusions or reports from the IWC Scientific Committee (SC) (which consists of several hundred international, independent scientists), whose concerns include lack of monitoring and inappropriateness of current mitigation measures.

Response:

The SEIS cited the ICES report, which was written by experts in the marine field. The SEIS also cited the Journal of Cetacean Resources Management, which is published by the IWC. Since no citation was provided by the commenter, it is unclear which IWC publication the comment refers to. The SEIS cited Cox

et al.

(2006), which was published in the Journal of Cetacean Resources Management. This article discusses monitoring and mitigation, focusing on beaked whales, but the monitoring and mitigation discussion was not specifically discussed in the Final SEIS. The conclusions on monitoring and mitigation state “Current visual survey efforts to detect beaked whales in areas of acoustic activity are probably ineffective as a mitigation aid. Key limiting factors include sea state, amount of daylight, experience of observers and the diving and surfacing behavior of beaked whales, which makes them either difficult to see or unavailable for visual observation at the surface for long periods of time. For the same reasons, surveys to determine distribution and abundance are also difficult and limited in their reliability. However, additional sensing technologies, such as passive acoustics, active sonar and radar, are currently in development that may increase scientists' abilities to detect beaked whales.” As discussed in the Final SEIS, the Final Comprehensive Report and NMFS’ Proposed Rule, the agencies recognize that visual monitoring is limited, particularly due to the factors such as sea state and daylight, as discussed in Cox

et al.

(2006). The final rule also requires passive acoustics, estimated to be 32 percent effective with visual monitoring and active acoustics, the HF/M3, which has a calculated effectiveness of 95 percent. The use of this tri-partite monitoring raises overall mitigation effectiveness to 98 percent. Therefore, the Navy will conduct the monitoring and mitigation measures recommended in Cox

et al.

(2006).

Comment 46:

The Navy's assessment of the risk of marine mammal injury and mortality from LFA sonar use is deficient. The problems with the Navy's calculation of thresholds for injury and behavioral disturbance, (mentioned previously in their October, 2006 letter) carry through to its analysis of the risk of injury.

Response:

NMFS does not agree. The Navy believes that the unusual or innovative nature of LFA sonar is what sets it apart from other anthropogenic sources, especially tactical, mid-frequency sonar and makes it much less likely to cause strandings of those marine mammals most associated with anthropogenic sound-related strandings (i.e., odontocetes, especially beaked whales). First, odontocetes generally have poor LF hearing. Second, the LFA sonar transmit array depth is well below 10 m (33 ft) and thus not likely to be entrained in a surface duct. Third, the 6 to 15 minute off-time in between 60-second transmissions and narrow bandwidth (30 Hz) generally preclude masking.

SURTASS LFA sonar has been operating since 2003 in a restricted area in the western Pacific Ocean, with approximately 470 hours of transmit time under the first four years of the LOAs. These extensive operations, with mitigation, have produced no known Level A takes on marine mammals. As noted before, LFA sonar is not the same as MFA (please see the Comment 44 in this document and the Final SEIS RTC 4.0.3 and 4.3.7). There is no evidence that SURTASS LFA sonar has caused injuries below or within the 180-dB mitigation zone as verified by mitigation monitoring requirements of the LFA sonar safety zone. Therefore, the 180-dB injury threshold remains valid, as does the effectiveness of the mitigation measures within the 180-dB potential injury zone.

The potential for SURTASS LFA sonar to cause harm to marine mammals and the validity of the 180-dB injury threshold for SURTASS LFA sonar are discussed in the Final SEIS RTCs 4.0.1, 4.0.2, 4.0.3, 4.3.1, 4.3.2, 4.3.7, 4.3.8, 4.3.9, 4.3.10, and 4.3.12. LFA sonar will not cause physical harm to marine mammals below 180 dB RL. Moreover, mitigation within the 180-dB mitigation zone is effective (See the Final EIS Subchapter 2.3.2.2).

Comment 47:

The Navy wrongly dismisses mechanisms of sonar injury to marine mammals that would cause harm independent of stranding events. The Navy portrays a leading theory that whales suffer from bubble growth in organs that is similar to decompression sickness, or “the bends” in human divers as a controversial hypothesis. The Navy and NMFS cannot omit the numerous published, peer-reviewed papers that support this theory, or disregard the recognition bubble growth has received from expert panels, such as the one convened in 2004 by the Marine Mammal Commission to review sonar-related strandings. The Navy's analysis of injuries to whales leaves out a possibility that has been widely noted in literature, that some of the observed injuries are a result of behavioral changes, such as rapid surfacing or premature diving, that sonar could induce. In describing the 2000 Bahamas stranding event, the Navy places undue reliance on a list of “contributory factors” that it feels make a similar event unlikely to reoccur. We do not doubt that certain factors, such as the use of sonar in channels, can increase the risk of harm; but it is abundantly evident from the literature that has emerged since the government's

Bahamas report appeared in 2001 that strandings may well occur in their absence.

Response:

NMFS has not dismissed any of the mechanisms of sonar injury to marine mammals that would cause harm independent of stranding events. One form of injury theorized to be caused by marine mammal reactions to sonar is gas-bubble disease. Cox

et al.

(2006) (which is the only reference cited by the commenter on this issue) stated that gas-bubble disease, induced in supersaturated tissues by a behavioral response to acoustic exposure, is a plausible pathologic mechanism for the morbidity and mortality seen in cetaceans associated with mid-frequency sonar exposure. They also state that it is premature to judge acoustically mediated bubble growth as a potential mechanism and recommended further studies to investigate the possibility. Since the Draft SEIS was published, there has been additional information available on this theory. If acoustically mediated bubble growth does prove to be the mechanism leading to mortality and/or strandings of beaked whales, then the fact that LFA sonar has not been associated with any of these strandings would indicate that it would be less likely to cause this effect.

Comment 48:

In addition, the Navy has failed to consider most of the mass beaked whale strandings that have been identified for their association, or possible association, with sonar and the fact that some marine mammal species are especially vulnerable to acoustical injuries. The Navy overestimates the importance of the fact that the long history of strandings associated with military sonar has usually implicated mid-frequency sonar. Many in the scientific community, including NMFS biologists, have expressed concern, based on the best available evidence, that low frequency sound could potentially induce similar effects. The NRDC believes that the Navy places far too much confidence in its assertion that its use of SURTASS LFA sonar in the last few years has not resulted in marine mammal strandings.

Response:

While NMFS shares this concern, to date, SURTASS LFA sonar has not been linked with any stranding events, other than by name association with MF sonar. This was discussed previously in this document. As related to LFA sonar, the Navy performed extensive research to determine the potential for LF transmissions to cause significant behavioral effects in whales (the LFS SRP).

Given that the LFA sonar sound source can be detected at moderate to low levels over large areas of the ocean, there was concern at the initiation of the Navy's NEPA process in 1996 that there was the potential for large percentages of species stocks to be exposed to moderate-to-low received levels. If animals are disturbed at these moderate-to-low exposure levels such that they experience a significant change in a biologically important behavior, then such exposures could potentially have an impact on rates of reproduction or survival. Knowing that cetacean responses to LF sound signals needed to be better defined using controlled experiments, the Navy helped develop and supported the three-year LFS SRP beginning in 1997. This study focused on baleen whales because, as low frequency hearing specialists they are believed to be the most sensitive to LFA sound and thus most likely to have an adverse behavioral reaction. This field research program was designed to address three important behavioral contexts for baleen whales: (1) Blue and fin whales feeding in the southern California Bight; (2) gray whales migrating past the central California coast; and (3) humpback whales breeding off Hawaii. Taken together, the results from the three phases of the LFS SRP do not support the hypothesis that most baleen whales, who are expected to be most sensitive to LF sounds, exposed to RLs near 140 dB would exhibit disturbance behavior and avoid the area. These experiments, which exposed baleen whales to RLs ranging from 120 to about 155 dB, detected only minor, short-term behavioral responses. Short-term behavioral responses do not necessarily constitute significant changes in biologically important behaviors.

Although the LFS SRP did not involve beaked whales, there was no indication during these tests that whales surfaced rapidly or dove prematurely in response to LFA sonar source transmissions. NMFS believes therefore, it is unlikely that, at least for fin, gray and humpback whales exposed to low levels of LFA sonar sounds will not result in the behavioral reactions theorized for beaked whales exposed to MF sonar signals. However, while this does that mean that LF sonar will not cause similar, but presently unknown, reactions in beaked whales, NMFS believes, that based on the best information available, such information does not currently exist. Therefore, NMFS believes, based on our current state of knowledge, it is unlikely that marine mammals would be severely injured by LFA sonar at great distances from the source.

Comment 49:

The Navy attempts to discount the well-established link between sonar use and marine mammal injuries and mortalities by suggesting (based on data compiled when acoustic impacts were not generally considered as a potential cause of strandings) that a majority of marine mammals strandings are related to natural causes. Finally, the Navy states, incorrectly, that “there are no new data that contradict any of the assumptions or conclusions in the Final EIS.” New data exists linking whale strandings to naval sonar; linking non-stranding injuries in marine mammals to naval sonar; describing mechanisms of harm to marine mammals from sonar; showing unexpectedly high propagation of noise in shallow waters; finding that intense noise sources can mask whale calls over great distances; and revealing the difficulties for noise impacts.

Response:

As indicated elsewhere in this response, most marine mammal strandings are unrelated to the use of sonar. While the recognition that there was a link between tactical sonars and beaked whale strandings was slow to develop, that in no way should be interpreted to mean that strandings involving sonar are either common or long-occurring.

NMFS believes the issue for this rulemaking is not whether sonar causes mass strandings of beaked whales, but whether SURTASS LFA sonar has the potential to cause marine mammal strandings. The evidence to date, supported by scientific reports, such as ICES (2005), Cox

et al.

(2006), and D'Spain

et al.

(2006), is that SURTASS LFA sonar has not caused any strandings. In reference to the contributory factors for strandings, the Bahamas 2000 stranding event did not involve LFA sonar. The list of “contributing factors” is generally supported by the workshop on understanding the impacts of anthropogenic sound on beaked whales convened by the U.S. Marine Mammal Commission in 2004 (Cox

et al.

, 2006) and the analysis by D'Spain

et al.

(2006). Whether or not surface ducts occurred during other reported strandings is not relevant to LFA sonar operations. First, NMFS believes LFA sonar operations will not cause physical injury to marine mammals at received levels below 180 dB. Second, LFA sonar signals are initially transmitted substantially below 10 m (32.8 ft) depth and are not likely to have signal strength above 180 dB in the surface duct. Surface ducting conditions were analyzed in the Final EIS at a number of the 31 model sites. Therefore, with LFA sonar mitigation, no marine

mammals, either with or without a surface duct, are expected to be exposed to injurious levels of LFA sonar signals.

The evidence to date, supported by scientific reports, such as ICES (2005), Cox

et al.

(2006), and D'Spain

et al.

(2006), is that SURTASS LFA sonar has not caused any strandings. Beaked whales, which hear best in the mid-frequency range appear to be most vulnerable to acoustic-induced stranding. These animals hear poorly in the low frequency range. The LFS SRP specifically studied the behavioral reactions of baleen whales, which hear best in the low frequency range, and thus were concluded to be most at risk (potentially) from the operation of LFA sonar. The three-phase LFS SRP involved more than 20 scientists from 6 universities and independent research groups. The results of the LFS SRP demonstrated that behavioral responses predictably occurred at received levels around 140 dB, not at the lower decibel levels that had been previously predicted. Moreover, the results showed that behavioral responses lasted for only a matter of tens of minutes and involved only modest changes in behavior. These results plus a five-year history of safely operating SURTASS LFA sonar without evidence of strandings or injury supports NMFS conclusion that the system can be operated, with appropriate mitigation measures, in manner that has no more than a negligible impact on marine mammal species and stocks.

In the Final SEIS Subchapter 4.4.3, the Navy discusses both anthropogenic and natural causes of marine mammal strandings. In the conclusion in Subchapter 4.4.3.4, it is stated that military sonar is not the principal cause of marine mammal strandings. There was no conclusion that the majority of marine mammal strandings were related to only natural causes. The Navy did not intend to give the impression that it discounts any scientifically-supported links between anthropogenic sources and marine mammal strandings. However, it will point out that there is no known connection between marine mammal strandings and LFA sonar, which is supported by scientific workshops, reports, and published papers (ICES, 2005; Cox

et al.

, 2006; D'Spain

et al.

, 2006).

Finally, to address the comment that there is no new data to contradict any of the assumptions or conclusions in the Final EIS, in order to address the comment, it must be pointed out once again that there are different types of anthropogenic sounds potentially associated with possible impacts to and strandings of marine mammals. These are naval sonar and seismic airgun arrays, each with different characteristics and purposes. Many comments lump these types under one heading, loud naval sonars or military sonars; or loud anthropogenic noise sources including sonars and seismic survey airguns. Thus, when there is a stranding that may be associated with the use of one type of sonar or sound source, it gets blamed on sonar as a whole-a premise that is not true and one that does not stand up to scientific scrutiny from the marine bio-acoustics community. A wide range of naval sonars are used to detect, localize and classify underwater targets. For the purposes of the SURTASS LFA sonar Final SEIS analysis, these systems are categorized as LFA sonar (less than 1000 Hz) and MFA sonar (1 to 10 kHz). Table 1 provides pertinent information on different types of LFA sonar and MFA sonar. General information is also provided on airgun arrays. Sonar signals are generally coherent while air guns are impulsive.

Cox

et al.

(2006) provided a summary of common features shared by the strandings events in Greece (1996), Bahamas (2000), and Canary Islands (2002). These included deep water close to land (such as offshore canyons), presence of an acoustic waveguide (surface duct conditions), and periodic sequences of transient pulses (i.e., rapid onset and decay times) generated at depths less than 10 m (32.8 ft) by sound sources moving at speeds of 2.6 m/s (5.1 knots) or more during sonar operations (D'Spain

et al.

, 2006). A number of these features do not relate to LFA sonar operations. First, the SURTASS LFA sonar vessel operates with a horizontal line array (SURTASS: a passive listening system) of 1,500 m (4,921 ft) length at depths below 150 m (492 ft) and a vertical line array (LFA sonar source) at depths greater than 100 m. Second, operations are limited by mitigation protocols to at least 22 km (12 nm) offshore. Therefore, for these reasons SURTASS LFA sonar cannot be operated in deep water that is close to land. Finally, the LFA sonar signal is transmitted at depths well below 10 m (32.8 ft), and the vessel has a slow speed of advance of 1.5 m/s (3 knots).

While it is true that there was a LF component of the sonar potentially related to the Greek stranding in 1996, only mid-frequency components were present in the strandings in the Bahamas in 2000, Madeira 2002, and Canaries in 2002. This supports the logical conclusion that the LF component in the Greek stranding was not causative (ICES, 2005; Cox

et al.

, 2006). In its discussion of the Bahamas stranding, Cox

et al.

(2006) stated, “The event raised the question of whether the mid-frequency component of the sonar in Greece in 1996 was implicated in the stranding, rather than the low-frequency component proposed by Frantzis (1998).” The ICES in its “Report of the Ad-Hoc Group on the Impacts of Sonar on Cetaceans and Fish” raised the same issue as Cox

et al.

, stating that the consistent association of MF sonar in the Bahamas, Madeira, and Canary Islands strandings suggest that it was the MF component, not the LF component, in the NATO sonar that triggered the Greek stranding of 1996 (ICES, 2005).

Most odontocetes, such as beaked whales, have relatively sharply decreasing hearing sensitivity below 2 kHz. If a cetacean cannot hear a sound of a particular frequency or hears it poorly, then it is unlikely to have a significant behavioral impact (Ketten, 2001). Therefore, it is unlikely that LF transmissions from LFA sonar would induce behavioral reactions from animals that have poor LF hearing, e.g. beaked whales, bottlenose dolphins, striped dolphins, harbor porpoise, belugas, and orcas (summarized in: Nedwell

et al.

, 2004).

New data describing potential mechanisms of harm to marine mammals from sonar are concerned with acoustically mediated bubble growth and resonance. Cox

et al.

(2006) stated that it is premature to judge acoustically mediated bubble growth as a potential mechanism and recommended further studies to investigate the possibility. The analysis by the Navy (Cudahy and Ellison, 2002) and reports from two workshops on acoustic impacts (DOC, 2002; Cox

et al.

, 2006) support the conclusion that resonance from LFA sonar operations is not a “reasonably foreseeable” impact.

The ICES (2005) report concluded that no strandings, injury, or major behavioral change has yet to be associated with the exclusive use of LF sonar.

Based on the above discussions, there are no “new” data: (1) Linking LFA sonar to whale strandings, (2) linking LFA sonar to non-stranding related injuries, or (3) describing mechanisms of harm to marine mammals from LFA sonar.

Regarding unexpectedly high propagation of noise in shallow water, this concerns the measurement of propagation of broadband noise from air gun arrays in both deep and shallow water (Tolstoy

et al.

, 2004). As noted in Table 1, there are substantial differences between the impulsive sounds of air guns and the coherent signals from LFA

sonar, so that one must be careful in how they are compared. First, while Tolstoy

et al.

(2004) found that when their calibrations were conducted in deep water (at 3200 m (10,500 ft)) and slope waters (at 500 m (1641 ft)), the predicted and measured distances to the received level of 160 dB from the air gun arrays indicated that the predicted radii tended to overestimate actual 160 dB RL ranges. (This implied that the 180-dB radii for all arrays should be less than the predicted 1 km (0.54 nm), likely significantly less.) Second, they found that their actual measurements for shallow water (30 m (98 ft)) had been underestimated when compared to the same predicted values used for the deep water comparison. This was due to the model not taking into account interaction with the ocean bottom. In deep, homogenous water, sound initially spreads spherically (spherical spreading) and its intensity decreases in proportion to the square of the range. Once sound has propagated to a distance approximately equal to the water depth, it is physically constrained and propagates cylindrically (cylindrical spreading). When this occurs, its intensity decreases in direct proportion to the range (please see Final EIS, Appendix B). Most importantly, however, SURTASS LFA sonar will not operate in water less than 200 m (656.2 ft), most likely always operating in deep and slope waters. Sound propagation from deep offshore waters onto shallower shelf waters will almost always decrease quickly due to bottom and surface interaction with the sound. This means that LFA sonar sounds will more quickly decrease in intensity in shallow water than in other waters. Lastly, the Tolstoy

et al.

(2004) findings are not applicable to the SURTASS LFA sonar analysis because the propagation models utilized for LFA sonar are empirically validated and correctly account for critical variables, such as water depth (Final EIS Subchapters 4.2, 4.2.1 and 4.2.2; and Technical Report #2).

The masking effect of the SURTASS-LFA sonar signal will be limited for a number of reasons. First, the bandwidth of the system is limited (30 Hz), and the instantaneous bandwidth at any given time of the signal is small, on the order of 10 Hz. Therefore, within the frequency range in which masking is possible, the effect will be limited because animals that use this frequency range typically use signals with greater bandwidth. Thus, only a portion of the animal's signal would be masked by the LFA sonar. Furthermore, when LFA sonar is in operation, the LFA sonar source is active only 7.5 percent of the time (based on historical LFA sonar operational parameters) and no more than 20 percent, which means that for 80-92.5 percent of the time there is no risk of animal signals being masked by the LFA sonar signal when LFA sonar is operating. Therefore, within the area in which masking is possible, the effect will be limited in duration and because animals that use this frequency region typically use broader bandwidth signals that allow them to communicate even when SURTASS LFA sonar is transmitting.

Finally, NMFS does not believe that the Navy has experienced difficulties in executing the mitigation procedures required by NMFS for LFA sonar, which are based on protecting marine animals from injury. Because it is impractical and infeasible for mitigation to cover vast oceanic areas, where the received levels do not cause physical injury to marine mammals or jeopardize threatened or endangered species, the laws provide methods for authorizations for limited non-injurious impacts to marine mammals and listed species. NMFS believes that SURTASS LFA sonar has met all of these requirements and has been operating since 2003-without any known physical injuries to marine animals. Potential non-injurious impacts are estimated based on location and times of operations and best available abundance and density data for the areas and seasons of the operations. These are reported to NMFS both quarterly and annually as required by regulation (50 CFR § 216 Subpart Q).

Comment 50

: We don't know the impact of SURTASS LFA sonar on species, stocks, and ecological processes over time. Therefore, NMFS can't say stock-level effects are “not reasonably likely” to occur.

Response:

When compared to other naturally occurring and anthropogenic sources of noise in the ocean, LFA sonar, barely contributes a measurable portion of acoustic energy in the oceans. Other sources o

This text is long and has been trimmed here. Open the source document for the complete record.